Par « Rémi NGHIEM-XUAN

280
THESE DE DOCTORAT DE L'UNIVERSITE DE NANTES COMUE UNIVERSITE BRETAGNE LOIRE ECOLE DOCTORALE N° 602 Sciences pour l'Ingénieur Spécialité : « (Génie des Procédés) » « Optimisation de la culture dHaslea ostrearia en photobioréacteur » Thèse présentée et soutenue à « Nantes », le « 01/07/2019 » Unité de recherche : Laboratoire GEPEA, UMR CNRS 6144 Thèse N° : Par « Rémi NGHIEM-XUAN » Rapporteurs avant soutenance : Réjean Tremblay Professeur, Université du Québec à Rimouski UQAR Ingrid Fruitier-Arnaudin Maître de conférences HDR, Université La rochelle Composition du Jury : Président : Philippe Michaud Professeur, Université Clermont-Auvergne Examinateurs : Philippe Michaud Professeur, Université Clermont-Auvergne Ian Probert Docteur, Station Biologique de Roscoff Dir. de thèse : Pascal Jaouen Professeur, Université de Nantes Co-dir. de thèse : Jérémy Pruvost Professeur, Université de Nantes Co-enc. de thèse : Vincent Turpin Maître de conférences, Université de Nantes Invité(s) Isabelle Rouillard Docteur, école Centrale de Nantes

Transcript of Par « Rémi NGHIEM-XUAN

Page 1: Par « Rémi NGHIEM-XUAN

THESE DE DOCTORAT DE

L'UNIVERSITE DE NANTES

COMUE UNIVERSITE BRETAGNE LOIRE

ECOLE DOCTORALE N° 602

Sciences pour l'Ingénieur

Spécialité : « (Génie des Procédés) »

« Optimisation de la culture d’Haslea ostrearia en photobioréacteur » Thèse présentée et soutenue à « Nantes », le « 01/07/2019 » Unité de recherche : Laboratoire GEPEA, UMR CNRS 6144 Thèse N° :

Par

« Rémi NGHIEM-XUAN »

Rapporteurs avant soutenance : Réjean Tremblay Professeur, Université du Québec à Rimouski UQAR Ingrid Fruitier-Arnaudin Maître de conférences HDR, Université La rochelle

Composition du Jury :

Président : Philippe Michaud Professeur, Université Clermont-Auvergne Examinateurs : Philippe Michaud Professeur, Université Clermont-Auvergne

Ian Probert Docteur, Station Biologique de Roscoff Dir. de thèse : Pascal Jaouen Professeur, Université de Nantes Co-dir. de thèse : Jérémy Pruvost Professeur, Université de Nantes Co-enc. de thèse : Vincent Turpin Maître de conférences, Université de Nantes

Invité(s) Isabelle Rouillard Docteur, école Centrale de Nantes

Page 2: Par « Rémi NGHIEM-XUAN
Page 3: Par « Rémi NGHIEM-XUAN

Remerciements

La thèse est une expérience exceptionnelle et bien plus riche en enseignements que

ce à quoi je m’attendais et remercier toutes les personnes qui m’ont accompagné dans ce

projet me semble être une bonne conclusion, que vous lirez certainement en premier.

Tout d’abord, rien n’aurait pu être possible sans le professeur Pascal Jaouen qui

m’a accueilli au GEPEA et m’a permis de réaliser cette thèse passionnante. Un très grand

merci à mon co-directeur Jérémy Pruvost qui m’a beaucoup appris scientifiquement, m’a

apporté ses conseils et a su m’encourager dans cette épreuve. J’exprime également toute

ma gratitude à mon co-encadrant de thèse le docteur Vincent Turpin pour avoir fourni les

souches d’Haslea ostrearia, son encadrement sur sa culture ainsi que sa patience pour mes

erreurs d’orthographe. J’exprime également toute ma gratitude au professeur Jean-Luc

Mouget pour m’avoir prodigué ses conseils et son expertise sur cette souche de

microalgue.

Je remercie sincèrement les membres du jury de m’avoir fait l’honneur d’examiner

et d’évaluer mon travail. Merci au professeur Réjean Tremblay et au docteur Ingrid

Fruitier-Arnaudin d’avoir accepté d’être rapporteurs de ma thèse. Merci au docteur Ian

Probert et au professeur Philippe Michaud ainsi qu’à notre invité, le docteur Isabelle

Rouillard pour leur intérêt et leur disponibilité.

Merci à l’administration du GEPEA et plus particulièrement à Carole, Laurette et

Marie-Pierre pour leur aide, leur bonne humeur ainsi que leur efficacité.

Je remercie chaleureusement tout le service technique sans qui le laboratoire ne

fonctionnerait pas, Hélene, Emmanuel, Benjamin, Raphaëlle, Catherine, Jordan, Guillaume,

Delphine, Laurence, Franck et Sébastien Chollet, Merci pour tout.

Un grand merci à tous les maîtres de conférences qui m’ont aidé dans les

différentes étapes de cette expérience : Olivier Gonçalves, Luc Marchal, Estelle Couailler

ainsi que Vona Meleder.

Page 4: Par « Rémi NGHIEM-XUAN

Je remercie particulièrement Christophe Lombard, merci de m’avoir permis de

faire cette thèse et pour nos longues discussions scientifiques et personnelles toujours

très riches.

Je tiens à exprimer également tous mes remerciements à la formidable équipe de

doctorants du GEPEA qui ont permis de créer une ambiance inoubliable. Merci à Vladimir,

ce poète philosophe qui reste une des plus belles rencontres de cette expérience et bien

sûr Jérémy sans qui les aventures en terre d’hyborée seraient évidemment moins

amusantes (Pfff se marier à une danseuse itinérante… Même pas Kithane). Je remercie

tous les autres avec qui j’ai vécu aussi d’incroyables moments, Alexandra (pour toute ta

joie), Philippe (j’espère que tu prends bien soin de mes plantes ?!), Lisa, Antoinette (c’est

grâce à toi que les corrections se sont terminées), Aumaya, Astrid, Marlène, Antoine

(Houm posso !), Guillaume, Eglantine, Armel, Dounia, Fernando, Brieuc, Rozine, Razmig,

Rémy, Erica, Charlène, mais aussi les nouveaux venus, Julien, Joris, et Sim, je vous souhaite

tout le meilleur pour vos thèses.

Merci à ceux qui m’ont prêté main forte durant mes expériences, Ikha pour son

positivisme et sa joie de vivre ainsi que tous les week-ends qu’elle a donné pour ce projet

et pour moi (ne perds jamais ça, c’est unique !), Rebiha pour sa bonne humeur et Gaetan

pour son investissement dans le projet. Merci aussi à Edouard Duliège de l’école

supérieure de physique et de chimie industrielle pour ses échanges et encouragements.

Je tiens à remercier Algosource avec Olivier Lepine et Christophe Lombard qui

m’ont orienté vers cette thèse et aussi à mon époque Thomas, Stéphanie, Sébastien

Jubeau, Arnaud, François, Fanny et Jérôme, je vous souhaite le meilleur.

Je vais remercier ma famille pour tout ce qu’elle a fait pour moi. Papa, Maman,

merci pour avoir supporté de toutes vos ressources, mon humeur, mes doutes, mes

excentricités, mes études et mes peines. Anicq ma tati préférée, merci à toi d’être présente

dans les moments douloureux et de rendre les choses plus simples de ton sourire. Zazou,

Nico, merci de votre présence et votre bonheur, qui suffit au mien. Merci de m’avoir donné

ma petite licorne Kim qui me donne le plaisir de la gâter. Merci à feu ma grand-mère

Fernande, qui m’a transmis sur mon visage d’abord timide et fermé, la joie et la chaleur

du nord de la France. Pour finir, Adrick mon ami, tu es celui qui m’a soutenu et inspiré

durant ces trois ans, je ne sais comment te remercier mais sans toi, je n’aurais sûrement

Page 5: Par « Rémi NGHIEM-XUAN

pas été au bout de cette expérience. J’ai conscience de ce que je t’ai fait vivre, ce qui rend

ma gratitude d’autant plus grande, merci pour tout.

Je n’oublie pas mes chers amis des univers parallèles, Incérion, Gendhis, Phygorn,

Meaban, ma sista Phaere, Tenkeï (quand est-ce qu’on la voit ?), Samanosuké, Irinie,

Zenyatta, Ashrogdan, BB-H et tellement d’autres.

Merci Val et Seb d’avoir trouvé et cultivé chez moi quelques choses que je ne

pensais pas posséder.

Bref, alors que ces remerciements semblent annoncer la fin d’une histoire, affûtez

vos lames ! Sortez vos flèches ! Car en réalité, pour moi, elle ne fait que commencer...

Page 6: Par « Rémi NGHIEM-XUAN

Listes des communications

Conférences :

- EXOBIO 10 juin 2016, Nantes, France: R. Nghiem Xuan, J. L. Mouget, J. Pruvost, V.

Turpin, P. Jaouen. Développement d'un bio-procédé d'immobilisation cellulaire

passif et actif d’haslea ostrearia pour la production de marennine en continu

(Orateur)

- AG AMI 22 mars 2016, Nantes, France: R. Nghiem Xuan, J. L. Mouget, J. Pruvost, V.

Turpin, P. Jaouen. Développement d’un bio-procédé de production en continu de

marennine issue d’une culture immobilisée d’Haslea ostrearia (Orateur)

- Séminaire GEPEA 13 octobre 2016, Saint-Nazaire, France: R. Nghiem Xuan, J. L.

Mouget, J. Pruvost, V. Turpin, P. Jaouen. Développement d’un bio-procédé de

production en continu de marennine issue d’une culture immobilisée d’Haslea

ostrearia (Orateur)

- EABA December 2017 Berlin, Allemagne: R. Nghiem Xuan, A. Busnel, J. Pruvost, P.

Jaouen. From lab to industry; a universal methodology to overcome the scaling-up

challenge of sensitive diatoms and cyanobacteria strains cultures (Orateur)

Poster

- International Society of Applied Phycology (ISAP) juin 2017 Nantes: R. Nghiem

Xuan, J. L. Mouget, J. Pruvost, V. Turpin, P. Jaouen. Bioprocess development for

marennine production with the diatom Haslea ostrearia

Publications

- R. Nghiem Xuan, I. Safitri, J. L. Mouget, J. Pruvost, V. Turpin, P. Jaouen. Design of an

artificial culture medium to optimize Haslea ostrearia biomass and marennine

production. Algal Research (Accepté sous conditions de corrections mineures

08/05/2019)

- R. Nghiem Xuan, J. L. Mouget, J. Pruvost, V. Turpin, P. Jaouen. Investigation of Haslea

ostrearia cellular fragility and consequences on stirring strategy in conventional

photobioreactor culture. (En cours de rédaction)

Page 7: Par « Rémi NGHIEM-XUAN

- R. Nghiem Xuan, J. L. Mouget, J. Pruvost, V. Turpin, P. Jaouen. Haslea ostrearia

culture in mixed photobioreactor: growth parameters gathering and marennine

production optimization. (En cours de rédaction)

Page 8: Par « Rémi NGHIEM-XUAN

Table des matières

INTRODUCTION GENERALE 1

Chapitre I : 6

Etude bibliographique 6

7 LA DIATOMEE HASLEA OSTREARIA (SIMONSEN, 1974) 7

7.1 TAXONOMIE 7

7.2 CYTOLOGIE 8

7.3 REPRODUCTION 9

7.3.1 REPRODUCTION ASEXUEE 10

7.3.2 REPRODUCTION SEXUÉE 11

7.4 L’ECOSYSTEME EN CLAIRE OSTREICOLE 11

7.5 PHOTOSYNTHESE 14

7.6 LA COMPOSITION PIGMENTAIRE 19

8 LA MARENNINE 21

8.1 LA STRUCTURE CHIMIQUE 21

8.2 LES PROPRIETES PHYSICO-BIOCHIMIQUES DE LA MARENNINE 22

8.2.1 LES PROPRIETES BIOCHIMIQUES 22

8.2.2 LES PROPRIETES PHYSIQUES 26

8.3 LE METABOLISME DE PRODUCTION DE MARENNINE 27

9 VALORISATION DE LA BIOMASSE 30

9.1 LES PIGMENTS 30

9.1.1 LA MARENNINE INTRACELLULAIRE 30

9.1.2 LA FUCOXANTHINE 30

9.2 LES LIPIDES 31

Page 9: Par « Rémi NGHIEM-XUAN

10 CARACTERISATION DE LA FRAGILITE CELLULAIRE 32

10.1 NOTIONS DE CISAILLEMENT 32

10.1.1 LA DEFORMATION D’UN FLUIDE : L’ECOULEMENT LAMINAIRE ET TURBULENT 32

10.1.2 LES FLUIDES NEWTONIENS ET LA VISCOSITE 33

10.2 LES CONTRAINTES DE CISAILLEMENT EN PHOTOBIOREACTEUR 34

10.2.1 TECHNOLOGIE AIRLIFT 35

10.2.2 TECHNOLOGIE TORIQUE 38

10.3 L’IMPACT DES CISAILLEMENTS SUR LES ORGANISMES MICROSCOPIQUES UNICELLULAIRE LIBRE ET EN TISSUS 41

10.3.1 LA SENSIBILITE DES MICROALGUES/CYANOBACTERIES/TISSUS D’ORGANISMES PLURICELLULAIRES 41

10.3.2 LA CARACTERISATION DE LA FRAGILITE D’HASLEA OSTREARIA 50

11 LA CULTURE MICROALGALE EN PHOTOBIOREACTEUR 59

11.1 LES SYSTEMES DE CULTURE OUVERTS 60

11.2 LES SYSTEMES DE CULTURE FERMES 63

11.3 FACTEURS LIMITANTS DE LA CROISSANCE MICRO-ALGALE 65

11.3.1 LA LUMIERE 65

11.3.2 LA TEMPERATURE 69

11.3.3 LE PH 71

11.3.4 LA COMPOSITION DU MILIEU DE CULTURE 73

11.3.5 L’AGITATION 76

11.4 LA CULTURE D’HASLEA OSTREARIA EN PHOTOBIOREACTEUR 77

11.4.1 CULTURE EN CELLULES LIBRES AGITEE 77

11.4.2 CULTURE EN CELLULE IMMOBILISEES 79

CONCLUSION 86

Chapitre II: 90

Matériel et méthodes 90

1 MATERIEL BIOLOGIQUE 91

Page 10: Par « Rémi NGHIEM-XUAN

1.1 HASLEA OSTREARIA 91

1.2 NANNOCHLOROPSIS GALITANA 92

1.3 CHLORELLA VULGARIS 92

2 TECHNIQUES DE MESURE 93

2.1 SUIVI DE LA DENSITE NUMERIQUE EN CELLULES 93

2.1.1 COMPTAGE CELLULAIRE 93

2.2 SUIVI PIGMENTAIRE 95

2.2.1 MESURE DE LA CHLOROPHYLLE A, C ET DES PIGMENTS CAROTENOÏDES 95

2.2.2 MESURE DE LA MARENNINE 96

2.3 SUIVI DE LA COMPOSITION DES MILIEUX DE CULTURE 97

2.3.1 DOSAGE DES IONS NITRATE ET PHOSPHATE DISSOUS 97

2.3.2 DOSAGE DES IONS SILICATES 97

2.3.3 DOSAGE DU CARBONE ORGANIQUE DISSOUS 97

2.4 MESURE DU RENDEMENT PHOTOSYNTHETIQUE FV/FM 98

3 MATERIEL DE CULTURE 99

3.1 LES MILIEUX DE CULTURES 99

3.2 SYSTEMES DE CULTURE 101

3.2.1 CULTURES EN VERRERIE CHIMIQUE 101

3.2.2 CULTURE EN FLASQUES 101

3.2.3 CULTURE EN PHOTOBIOREACTEUR 102

3.3 MODES DE CULTURE 109

3.3.1 CULTURE EN BATCH 109

3.3.2 CULTURE EN MODE SEMI-CONTINU 110

3.3.3 CULTURE EN CONTINU 110

3.4 APPLICATION DU FLUX D’ENERGIE LUMINEUX 111

3.5 METHODE D’IMMOBILISATION CELLULAIRE PAR INCLUSION EN GEL D’AGAR 112

3.6 METHODE D’IMMOBILISATION CELLULAIRE PAR INCLUSION EN BILLES D’ALGINATE 113

3.7 TECHNIQUES DE STERILISATION DES PBRS 113

3.8 APPLICATION DES CONTRAINTES DE CISAILLEMENT EN RHEOMETRE 113

Page 11: Par « Rémi NGHIEM-XUAN

Chapter III: 115

Design of an artificial culture medium to optimize Haslea

ostrearia biomass and marennine production 115

1 ABSTRACT: 116

2 INTRODUCTION 117

3 MATERIAL AND METHODS 118

3.1 ORGANISM AND CULTURE MAINTENANCE 118

3.2 THE CULTURE MEDIA TESTED 119

3.3 MACRO AND MICRO NUTRIENTS CONCENTRATION TESTS 119

3.4 GENERAL APPROACH TO INVESTIGATE MIXOTROPHIC AND AUTOTROPHIC GROWTH 120

3.5 MATERIAL FOR THE NEW MEDIUM ASSESSMENT AND PRODUCTIVITY VALIDATION 122

3.6 STATISTICS 122

4 RESULTS 123

4.1 DETERMINATION OF KEY NUTRIENTS RESPONSIBLE FOR HASLEA OSTREARIA GROWTH 123

4.2 EFFECT OF ORGANIC AND INORGANIC SOURCES OF CARBON AND PHOSPHORUS ON HASLEA GROWTH 124

4.3 DETERMINATION OF INORGANIC C AND P SOURCE CONCENTRATION IN NATURAL SEAWATER CULTURE MEDIUM

127

4.4 COMPONENT INTERACTION AND MEDIUM PRECIPITATION 130

4.5 STUDY OF THE IMPACT OF MICRONUTRIENTS ON BIOMASS AND MARENNINE PRODUCTION IN ARTIFICIAL

SEAWATER MEDIUM 132

4.6 VALIDATION OF THE ARTIFICIAL SEAWATER NX MEDIUM IN A PHOTOBIOREACTOR 134

5 DISCUSSION 136

6 CONCLUSION 139

7 SUPPLEMENTAL INFORMATION 141

Page 12: Par « Rémi NGHIEM-XUAN

ARTIFICIAL SEAWATER CULTURE MEDIUM (NX) FOR PBR CULTURE 141

8 REFERENCES 146

Chapter IV: 151

Investigation of Haslea ostrearia cellular fragility and

consequences of stirring strategy in conventional

photobioreactor culture 151

1 ABSTRACT: 152

2 INTRODUCTION 153

3 MATERIALS AND METHODS 155

3.1 INOCULUM AND CULTURES FOR PRELIMINARY EXPERIMENTS IN RHEOMETER AND FLASKS 155

3.2 MEASUREMENT OF CELL VIABILITY 156

3.3 PHOTOSYNTHETIC EFFICIENCY AND PIGMENTS MEASUREMENT 157

3.4 DETERMINATION OF MEAN RATE OF PHOTON ABSORPTION (MRPA) 158

3.5 CHARACTERIZATION OF SHEAR STRESS INFLUENCE ON H. OSTREARIA 160

3.5.1 INVESTIGATION IN RHEOMETER 160

3.5.2 EFFECT OF SHEAR STRESS IN STIRRED FLASKS 161

3.5.3 INVESTIGATION IN TORUS-SHAPED PBR 162

3.6 STATISTICS 163

4 RESULTS 164

4.1 EFFECT OF SHORT-TERM APPLICATION OF HOMOGENEOUS AND CONSTANT SHEAR STRESS 164

4.1.1 INVESTIGATION OF EXPERIMENT DURATION IN RHEOMETER TREATMENT 164

4.1.2 EFFECT OF VISCOSITY ON CELL VIABILITY BY RHEOMETER TREATMENT 165

4.1.3 INVESTIGATION IN DAY/NIGHT CYCLES 167

4.2 EFFECT OF LONG-TERM APPLICATION OF HYDRODYNAMICS SHEAR STRESS IN TURBULENT FLOW 169

Page 13: Par « Rémi NGHIEM-XUAN

4.2.1 INVESTIGATION OF STIRRING AND DARK/NIGHT ILLUMINATION EFFECTS ON H. OSTREARIA IN FLASK CULTURE

169

4.2.2 INVESTIGATION OF MIXING EFFECT IN TORUS SHAPED PHOTOBIOREACTOR 170

5 DISCUSSION 176

6 CONCLUSIONS 178

7 REFERENCES 180

Chapter V: 185

Haslea ostrearia culture in photobioreactor: growth

parameters gathering and marennine production

optimization 185

1 ABSTRACT: 186

2 INTRODUCTION 188

3 MATERIALS AND METHODS 189

3.1 CULTURE MEDIA FOR PRODUCTION IN PBR 189

3.2 CELLS DENSITY AND BIOMASS CONCENTRATION MEASUREMENT 189

3.3 DETERMINATION OF PIGMENT CONCENTRATION 190

3.4 NUTRIENTS CONCENTRATION MEASUREMENT 190

3.5 DETERMINATION OF THE MEAN RATE OF PHOTONS ABSORPTION (MRPA) 190

3.6 STATISTICS ANALYSIS 191

3.7 HASLEA OSTREARIA CULTURE 191

3.7.1 INOCULUM 191

3.7.2 IMMOBILIZED CELL PBR 191

3.7.3 AIRLIFT PBR 1L AND 6L 192

3.7.4 AIRLIFT CRYSTAL PBR 194

Page 14: Par « Rémi NGHIEM-XUAN

4 RESULTS 195

4.1 IMMOBILIZED-CELL CULTURE 195

4.2 HASLEA OSTREARIA CULTURE IN CRYSTAL PBR 197

4.2.1 INVESTIGATION OF CO2 TOLERANCE OF HASLEA OSTREARIA IN BATCH CULTURE 197

4.2.2 ASSESSMENT OF HALSEA OSTREARIA GROWTH IN CONTINUOUS MIXED CULTURE 199

4.3 INVESTIGATION OF NUTRIENT CONSUMPTION IN PBR CULTURE 200

4.3.1 CARBON CONSUMPTION 200

4.3.2 SILICON, NITROGEN AND PHOSPHORUS CONSUMPTION 202

4.4 THE INGATHERING OF OPTIMAL GROWTH PARAMETERS FOR BIOMASS AND EMN PRODUCTIONS 205

4.4.1 LIGHT TOLERANCE AND CELL RESPONSE 205

4.4.2 PH 207

4.4.3 TEMPERATURE 209

4.4.4 PHYSIOLOGIC STRESS FOR EMN OVERPRODUCTION 211

4.5 TOWARDS AN OPTIMIZED PRODUCTION PROTOCOL 214

4.5.1 EMN CONCENTRATION IMPACT ON PROCESS PRODUCTIVITY 214

4.5.2 IMPLEMENTATION OF EMN CONTINUOUS PRODUCTION PROCESS 217

5 DISCUSSION 219

6 CONCLUSIONS 222

7 REFERENCES 224

CONCLUSIONS ET PERSPECTIVES 229

NOMENCLATURE 237

BIBLIOGRAPHIE 239

Page 15: Par « Rémi NGHIEM-XUAN

Table des figures

Figure 1: La phylogénie d'Haslea ostrearia au sein des Eucaryotes des grandes Classes

d’algues ................................................................................................................................................................ 7

Figure 2: Haslea ostrearia au microscope optique, légendé ........................................................... 8

Figure 3: Représentation connective du frustule des diatomées (frustule) ............................. 8

Figure 4: zone centrale de la valve (3) extérieure (4) intérieure SEM (Smithsonian Marin

Station, 2015) .................................................................................................................................................... 9

Figure 5: zone apicale du frustule SEM (Smithsonian Marin Station, 2015) ........................... 9

Figure 6: Représentation de la re-distribution du frustule au cours de la croissance

d'Haslea ostrearia ......................................................................................................................................... 10

Figure 7: Haslea ostrearia déformé au microscope optique ........................................................ 10

Figure 8: Courbe de température moyenne annuelle en Baie de Bourgneuf (Climate data)

............................................................................................................................................................................. 12

Figure 9: représentation du cycle tychopélagique d'Haslea ostrearia dans une claire à

huîtres (D’après Robert, 1983) ............................................................................................................... 13

Figure 10: Relation écologique entre Haslea ostrearia et Crassostrea gigas en claire à

huître ................................................................................................................................................................. 14

Figure 11: Représentation de mécanisme de phase claire dans le phénomène de

photosynthèse. .............................................................................................................................................. 15

Figure 12: Représentation schématique du cycle de Calvin. ....................................................... 16

Figure 13: Schémas des cycles de l'activité oxygénase et carboxylase de l'enzyme

RuBisCO. ........................................................................................................................................................... 17

Figure 14: Représentation de la compostions pigmentaire des photosystèmes et la

circulation de l'énergie lumineuse chez une chlorophycée ......................................................... 19

Figure 15: Spectre d'absorption des pigments photosynthétiques (Devred et al., 2014) 20

Figure 16: 2D RMN (a) Spectre 1H–13C EMn d’Haslea ostrearia Correspond au spectre

1D 1H (Gastineau et al., 2014) ................................................................................................................. 21

Figure 17: Observation de marennine EMn à pH basique (gauche) et acide (droite)

(Jubeau, 2009) ............................................................................................................................................... 26

Figure 18: Absorbance de l'IMn (a) et EMn (b) en fonction du pH (Pouvreau et al.,2006)

............................................................................................................................................................................. 26

Page 16: Par « Rémi NGHIEM-XUAN

Figure 19: Courbes de croissance d’Haslea ostrearia sous lumière blanche (WL), bleu

(BL), vert (GL), jaune (YL), ou rouge (RL) à faible (20μmol photons m−2 s−1) et fort

intensité lumineuse (100μmol photons m−2 s−1) (Mouget et al., 2005) .................................. 28

Figure 20: Contenu en marennine de cellules d’Haslea ostrearia de différentes tailles

(longueur de cellule) produites sous lumière bleu (⧯) ou blanche (⧮) (Mouget et al.,

2005) ................................................................................................................................................................. 29

Figure 21: Cinétique de production de marennine et absorption de nitrate par des

cellules d'Haslea ostrearia immobilisées dans le gel d’agar (Lebeau et al., 2000) .............. 30

Figure 22: écoulement autour d’un cylindre, (Gauche : laminaire Rex< 5×105), (Droit :

turbulent Rex> 5×105) ................................................................................................................................ 33

Figure 23: Représentation de la circulation de la phase liquide en photobioréacteur

Airlift 1L ........................................................................................................................................................... 35

Figure 24: Simulation MFN du champ de vitesse de la phase liquide, coloré par intensité

de vitesse superficielle (m. s-1). (a) représente une Airlift 1L, (b) représente un Airlift

100L (Ndiaye et al., 2018) ......................................................................................................................... 37

Figure 25: schéma du pilote laboratoire Torique de production contrôlée de microalgues

(Artu, 2016) .................................................................................................................................................... 38

Figure 26: Représentation de la géométrie (Gauche) du PBR torique avec ses

dimensions, (Droite) de 2 types d'hélice utilisé pour l'agitation du PBR torique (Pruvost

et al., 2006) ..................................................................................................................................................... 39

Figure 27: Prédiction de la vitesse moyenne du fluide en PBR torique UO en fonction de

la vitesse de rotation des 2 types d'hélices (Fig. 25) utilisées en Torique (Pruvost et al.,

2006) ................................................................................................................................................................. 40

Figure 28: Prédiction des valeurs globales (VY), ε et εimp en fonction de la vitesse

moyenne du flux en PBR Torique (Pruvost et al., 2006) ............................................................... 40

Figure 29: (Gauche) plage de la baie Laguna Grande de nuit, Porto Rico et observation de

bioluminescence sur les vagues ; (droite) isolement du consortium bioluminescent en

laboratoire ....................................................................................................................................................... 46

Figure 30: Observation en microscopie optique de la dinoflagellé Pyrocystis lunula,

(Gauche) sans agitation sous lumière ; (Droite) après agitation sans lumière .................... 47

Figure 31: Observation en microscopie optique de la lignée cellulaire Sf9 de tissu

ovarien de Spodoptera frugiperda, (Gauche) culture adhérente en tissu ; (Droite) culture

libre agitée ....................................................................................................................................................... 47

Page 17: Par « Rémi NGHIEM-XUAN

Figure 32: Cinétique de croissance de culture après traitement en pompe péristaltique

(Q=25 L. h-1) selon le nombre de passage (étoile) témoin, + (croix) 10 passages, o

(rond) 100 passages, ◻ (carré) 250 passages (Rossignol, 1999).............................................. 51

Figure 33: Influence de la puissance et de la durée de sonication sur le cassage des

cellules d'Haslea ostrearia, V=300ml, □ (carrée) 550W, △ (triangle) 400W, ○ (rond)

300W (Rossignol 1999) ............................................................................................................................. 52

Figure 34: Influence du volume de suspension traité sur le cassage cellulaire d'Haslea

ostrearia, P=550W, ◼ (Carré) 90s, ◻ (blanc) 60s (Rossignol 1999) ....................................... 52

Figure 35: Observation de cellules d'Haslea ostrearia après traitement aux ultrasons par

microscopie électronique à balayage : t=40s, P=400W, V=200mL (document LBM -

ISOMer) ............................................................................................................................................................ 53

Figure 36: Influence de la pression du système de broyage haute pression sur le cassage

cellulaire et la libération de marennine dans le milieu (1 cycle), ♦ (losange) dommages

cellulaires, + (croix) diamètre moyen des particules, ● (rond) marennine libérer

(Rossignol 1999) .......................................................................................................................................... 53

Figure 37: Influence du nombre de cycle du système de broyage haute pression sur la

répartition de la taille des particules par granulométrie laser, (carré plein) culture

initiale ; (carré vide) 30 MPa, 1 cycle ; (étoile) 30 MPa, 5 cycles ; o (rond) 150 MPa, 1

cycle. (Rossignol 1999) .............................................................................................................................. 54

Figure 38: Observation d'Haslea ostrearia par microscopie électronique à balayage (a)

avant traitement, (b) après traitement (1 cycle 30 MPa) (Rossignol 1999) ......................... 54

Figure 39: Observation de cellules d'Haslea ostrearia au microscope électronique à

balayage avant (Gauche) et après passage dans une pompe à engrenage (Droite) (5L,

15min, 350 L. h-1) (document LBM - ISOMer) ................................................................................... 55

Figure 40: Comparaison des répartitions granulométriques d'une culture d'Haslea

ostrearia avant ⦁ (rond) et après 100 passages ◻ (carré) dans la pompe à engrenages

(300 L. h-1) (Rossignol 1999) ................................................................................................................... 55

Figure 41: Dégradation cellulaire en fonction du nombre de passages et du débit de

circulation ◼ (carré) 200 L. h-1, ▲ (triangle) 300 L h-1, ● (rond) 400 L. h-1 (Rossignol

1999) ................................................................................................................................................................. 56

Figure 42: Influence de l'accélération centrifuge sur la dégradation cellulaire et la

libération de marennine (Rossignol 1999) ........................................................................................ 57

Page 18: Par « Rémi NGHIEM-XUAN

Figure 43: Influence de la durée de centrifugation sur la dégradation cellulaire et la

libération de marennine (Rossignol, 1999) ....................................................................................... 57

Figure 44: Comparaison de la granulométrie des suspensions selon la méthode de

cassage cellulaire (Rossignol 1999) ...................................................................................................... 58

Figure 45: Infrastructures de production de spiruline de la compagnie Sosa Texcoco

1960 ................................................................................................................................................................... 60

Figure 46: Vue aérienne d'une culture de Dunaliella salina à Hutt lagoon, Australie

(infrastructure BASF) ................................................................................................................................. 61

Figure 47:Vue aérienne des raceways de Cyanotech Co Hawaii, USA ..................................... 61

Figure 48: PBR cascade à Trebon, république Tchèque (Malapascua et al., 2014) ............ 62

Figure 49: Système fermé de production d'Algatechnologie, Israël

(http://WWW.algatech.com) ................................................................................................................... 63

Figure 50: Culture d'Arthrospira platensis en photobioréacteur Airlift, GEPEA (Gauche).

Culture de Chlorella vulgaris en photobioréacteur Torique, Algosolis (droite) ................... 64

Figure 51: Profil d'atténuation lumineuse pour un PBR plan en régime photo-inhibé

(Prusvot et Cornet, 2012).......................................................................................................................... 66

Figure 52: Profil d'atténuation lumineuse pour un PBR plan en régime luminostat

(Prusvot et Cornet, 2012).......................................................................................................................... 67

Figure 53: Profil d'atténuation lumineuse pour un PBR plan en régime photolimité

(Pruvost et Cornet, 2012).......................................................................................................................... 67

Figure 54: Evolution de la productivité surfacique en biomasse de Chlorella vulgaris en

fonction de la température (PFD =130 μmolhv. m-2. s-1) (Souliès 2014) ................................. 70

Figure 55: effet de la température sur le taux de croissance de Phaeodactylum

tricornutum, Scenedesmus sp., Dinobryon divergens et Porphyridium cruentum (Ras et al.,

2013) ................................................................................................................................................................. 71

Figure 56: Représentation des fractions molaires de carbone dissous en fonction du pH à

25°C (Kaplan et al., 1986) .......................................................................................................................... 73

Figure 57: Photographie de la première culture d'Haslea ostrearia en photobioréacteur

de technologie Airlift au GEPEA (Jubeau, 2009) .............................................................................. 78

Figure 58: Installation expérimentale d'un photobioréacteur à membrane immergée

pour la culture d'Haslea ostrearia (Rossignol et al., 1999) .......................................................... 79

Page 19: Par « Rémi NGHIEM-XUAN

Figure 59: Photographie de cultures d'Haslea ostrearia en immobilisation cellulaire

passive : (Gauche) en erlenmeyer non agité 500 mL, (Droite) en bonbonne de verre

borosilicaté non agité 10 L ........................................................................................................................ 80

Figure 60: Culture d'Haslea ostrearia dans les PBR expérimentaux UQAR : (a)

photographie de deux des PBR, (b) courbe de la EMn mesuré dans le surnageant durant

35 jours de batch (Gastineau et al., 2014) .......................................................................................... 81

Figure 61: Méthodologie de la préparation de la culture d'Haslea ostrearia en tank de

500 L et en open pond expérimental de 10m3 (Turpin et al., 2001) ........................................ 81

Figure 62: Variation de la densité cellulaire et de la concentration en marennine dans le

surnageant au sein des culture en batch d'Haslea ostrearia (en tank 500 L et pond 10 m3)

(Turpin et al., 2001) .................................................................................................................................... 82

Figure 63: représentation du photobioréacteur à cellules immobilisées expérimental

(Lebeau et al., 2000) .................................................................................................................................... 83

Figure 64: représentation d'une inclusion cellulaire dans un polymère d'alginate de

calcium .............................................................................................................................................................. 84

Figure 65: photographie au microscope optique : (A) de billes d'alginate vide, (B) de

microcapsule d’alginate contenant Nitzschia closterium (Lebeau et al., 1998).................... 84

Figure 66: Représentation chronologique des innovations de productivité en marennine

extracellulaire selon les différentes technologies de systèmes de culture mis en place par

les laboratoires MMS, du GEPEA et de l'UQAR .................................................................................. 87

Figure 67: Cellule de Nageotte Figure 68: Coupe latéral de la cellule de Nageotte ....... 93

Figure 69: Gravure de la cellule de Nageotte ..................................................................................... 94

Figure 70: Mesure des différentes formes de carbones par TOC-L CSH/CSN....................... 98

Figure 71: Enceinte de culture Sanyo MLR-350 .............................................................................101

Figure 72 : Flasque de culture de tissus cellulaire stérile non traitée (25 cm² de surface

de culture) .....................................................................................................................................................101

Figure 73: panneau de LEDs à variateur pour irradiance horizontale ..................................103

Figure 74: Schéma du PBR à cellules immobilisées ......................................................................104

Figure 75: Schémas du PBR Cristal (Moutel, 2016) ......................................................................105

Figure 76: Schémas du PBR Airlift (Souliès, 2014) .......................................................................106

Figure 77: Schémas du PBR Torique (Souliès, 2014) ...................................................................107

Figure 78: Schémas du PBR plan 6L (Souliès, 2014) ....................................................................108

Page 20: Par « Rémi NGHIEM-XUAN

Figure 79: Représentation de la technique de mesure expérimentale du MRPA sur le PBR

Plan-6L ............................................................................................................................................................109

Figure 80: Photos du plateau et du mobile du rhéomètre (1) et dimensions du système

plateau/mobile (2) .....................................................................................................................................114

Figure 81: Représentation chronologique actualisée des travaux successifs visant à

améliorer la production en marennine extracellulaire selon les différentes technologies

de systèmes de culture mises en place par les laboratoires MMS, du GEPEA et de l'UQAR

...........................................................................................................................................................................234

Page 21: Par « Rémi NGHIEM-XUAN

Table of figures

Fig. 1: Culture chamber SANYO MLR-350 (left) and culture flask positions inside the

culture chamber (right). ..........................................................................................................................120

Fig. 2: Cell, Extracellular Marennine (EMn) and specific EMn productivity of Haslea

ostrearia in batch culture after 7 days in natural seawater-enriched (ES 1/3) based

medium with different sources of phosphorus and carbon. P EMn specific and Pv are,

respectively, the specific EMn productivity and the volumetric productivity. Data is

mean ±95% CI for n=3. .............................................................................................................................125

Fig. 3: Impact of organic phosphorus and copper on marennine and biomass

productivities according to carbon and phosphorus sources in ES1/3 based medium.

Data is mean ±95% CI for n=3. ..............................................................................................................126

Fig. 4: Cell, EMn and specific EMn volumic productivities of Haslea ostrearia for various

media (ES 1/3, artificial seawater (ASW) and ES 1/3 based medium with completely

inorganic C and P (NP)). P EMn specific and Pv are, respectively, the specific EMn

productivity and the volumetric productivity. Data is mean ±95% CI for n=3. .................129

Fig. 5: Simulation of precipitated molecules (%) in the 25N100P medium for 3 pH values

(Visual MINTEQ software) ......................................................................................................................131

Fig. 6: Biomass [X] (bars) and EMn (circles) volumetric productivities (Pv) according to

different macro and micro nutrient concentrations in artificial seawater. Data is mean

±95% CI for n=3 ..........................................................................................................................................133

Fig. 7: Comparison of Haslea ostrearia cell density culture in new NX-25N100P6Si

medium with two natural seawater-based media. Data is mean ±95% CI for n=3. .........135

Fig. 8: Comparison of EMn concentration from Haslea ostrearia culture in new NX-

25N100P6Si medium with two natural seawater-based media. Data is mean ±95% CI for

n=3 ...................................................................................................................................................................135

Fig. 9: Microscopic view of Haslea ostrearia. (A) is the contrast of two cells, cell (1) being

alive and cell (2) being dead. (B and C) shows living cells in a bloc of cellular residues,

with an assessment of cell viability by chlorophyll fluorescence under UV .......................157

Fig. 10: Representation of the torus shaped PBR geometry with its dimensions (left), and

2 types of impellers used for it stirring (right) (Pruvost et al., 2006) ...................................163

Fig. 11: Dynamic viscosity of the sample and shear stress applied over time at a constant

shear rate of 30,000 s-1 .............................................................................................................................164

Page 22: Par « Rémi NGHIEM-XUAN

Fig. 12: Evolution of cell viability at a constant shear rate (30,000 s-1) over time. Data is

mean ± 95% CI for n = 3 ...........................................................................................................................165

Fig. 13: Haslea ostrearia cell viability at a constant shear rate of a 30,000 s-1 with

different concentrations of Locust Bean Gum (LBG) to modify viscosities and shear

stress accordingly. Data is mean ± 95% CI for n = 3 .....................................................................166

Fig. 14: Haslea ostrearia cell viability cultivated in day/night culture. Cells were

subjected to a constant shear rate (30,000 s-1) for 1 h. Values are reported for every hour

during the day (14 h) and night (10 h) period of culture. Data is mean ± 95% CI for n = 3

...........................................................................................................................................................................168

Fig. 15: Haslea ostrearia culture with and without stirring in light/dark cycle and

constant 24 h illumination for 8-day batch. Data is mean ± 95% CI for n = 3 ....................169

Fig. 16: Time evolutions of cell density and fluorimetry during semi-continuous culture

of Haslea ostrearia in a torus-shaped PBR. Data is mean ± 95% CI for n = 3 ......................172

Fig. 17: Cell and biomass surface productivity (PS) obtained for semi-continuous culture

of Haslea ostrearia in a torus PBR in continuous light with five different stirring duration

conditions (0; 8; 14; 20; 24 h). Data is mean ± 95% CI for n = 6 ..............................................173

Fig. 18: Total photosynthetic pigment production (%) and Extracellular marennine

(EMn) excretion (%) for semi-continuous culture of Haslea ostrearia in a torus PBR in

constant light and different stirring duration conditions. Data is mean ± 95% CI for n = 6

...........................................................................................................................................................................174

Fig. 19: Estimation of Mean Rate of Photon Absorption (A24H, suspended) for 0, 8, 14, 20 and

24 h stirring duration in torus PBR, according to biomass surface productivity (PS)

achieved (L = 4cm) (q0 = 300 μmol m-2 s-1). Data is mean ± 95% CI for n = 6 ...................175

Fig. 20: Representation of the immobilized cell photobioreactor 500mL, aS=33 m-1 ......192

Fig. 21: Representation of 1 L airlift photobioreactor, aS = 33 m-1 (Souliès, 2014) ..........193

Fig. 22: Representation of 6 L airlift photobioreactor, aS=25 m-1 (Souliès, 2014) ............194

Fig. 23: Representation of Crystal photobioreactor 250 mL, aS = 57 m-1 (Moutel, 2016)

...........................................................................................................................................................................195

Fig. 24: Passive and active continuous culture of Haslea ostrearia in immobilized cell

PBR. Data is mean ± 95% CI for n = 3 and n = 2 for mean DIC. .................................................196

Fig. 25: Batch culture of Haslea ostrearia for 2% and 15% air/CO2 proportion for cell

density (A) and extracellular marennine concentration (EMn) (B), in Crystal PBR. Data

are mean ± IC 95% for n=3 .....................................................................................................................198

Page 23: Par « Rémi NGHIEM-XUAN

Fig. 26: Continuous culture of Haslea ostrearia at 15% CO2 in a Crystal PBR. Cell density,

extracellular marennine (EMn) and dissolved inorganic carbon (DIC) were measured.

Data is mean ± 95% CI for n = 3 ............................................................................................................199

Fig. 27: Semi-continuous culture of Haslea ostrearia in 1 L airlift PBR with carbon fed-

batch and 2% CO2. Dry weight (DW), extracellular marennine concentration (EMn) and

dissolved inorganic carbon (DIC) were measured. Data is mean ± 95% CI for n = 3 ......201

Fig. 28: Evolution of NO3-, PO42- and SiO32-concentration for semi-continuous culture of

Haslea ostrearia in a 1 L airlift PBR. Data is mean ± 95% CI for n = 3 ...................................202

Fig. 29: Semi-continuous Haslea ostrearia cell culture in 1 L airlift PBR with C, P, Si fed-

batch conditions. Dry weight (DW) and extracellular marennine concentration (EMn)

were measured. Data is mean ± 95% CI for n = 3 ..........................................................................204

Fig. 30: Light tolerance curve of Haslea ostrearia, describing the evolution of surface

productivity (PS) depending on MRPA values for biomass and extracellular marennine

(EMn) results. Data is mean ± 95% CI for n = 5 ..............................................................................205

Fig. 31: Haslea ostrearia total pigment, extracellular marennine (EMn) and light

conversion efficiency (η) into biomass as a function of MRPA. Data is mean ± 95% CI for

n = 5 .................................................................................................................................................................206

Fig. 32: Haslea ostrearia total pigment, extracellular marennine (EMn) and light

conversion efficiency (η) into EMn as a function of MRPA. Data is mean ± 95% CI for n =

5 .........................................................................................................................................................................207

Fig. 33: pH effect on Haslea ostrearia total photosynthetic pigment concentration,

biomass and extracellular marennine (EMn] productivitys (PS) in semi-continuous

culture in 1 L airlift PBR. Data is mean ± 95% CI for n = 5 .........................................................208

Fig. 34: Temperature effect on Haslea ostrearia ostrearia total photosynthetic pigment

concentration, biomass and extracellular marennine (EMn] productivity (PS) in semi-

continuous culture in 1 L airlift PBR. Data is mean ± 95% CI for n = 5. ................................210

Fig. 35: Total photosynthetic pigment concentration of Haslea ostrearia under various

nutrient limitations and depletions. Data is mean ± 95% CI for n = 5 ...................................211

Fig. 36: Biomass and EMn productivity of Haslea ostrearia culture under various nutrient

limitations and depletions. Data is mean ± IC 95% for n = 5. ....................................................212

Fig. 37: Biomass and cell productivities (PS) of Haslea ostrearia culture under various

nutrient limitations and depletions. Data is mean ± 95% CI for n = 5. ..................................213

Page 24: Par « Rémi NGHIEM-XUAN

Fig. 38: Evolution of cell density and extracellular marennine concentration (EMn) for

semi-continuous culture of Haslea ostrearia in NX medium in 1 L airlift PBR (D = 0.09 d-1,

corresponding to 10 days of residence time). Data is mean ± 95% CI for n = 3. ...............215

Fig. 39: Evolution of cell density and extracellular marennine concentration (EMn) in a

semi-continuous (D=0.18 d-1 corresponding to 5 days of residence time) culture of

Haslea ostrearia in NX medium in a 1 L airlift PBR. Data is mean ± 95% CI for n=3 .......216

Fig. 40: Evolution of dry weight (DW), extracellular marennine concentration (EMn) and

specific EMn concentration of Haslea ostrearia for semi-continuous culture in NX media

with optimal culture conditions in a 1 L airlift PBR. Data is mean ± 95% CI for n = 3 ....218

Fig. 41: Evolution of dry weight (DW) and extracellular marennine concentration (EMn)

of Haslea ostrearia for a continuous culture in NX media with optimal culture conditions

in a 500 mL immobilized cell PBR. Data is mean ± 95% CI for n = 3 ......................................219

Liste des tableaux

Tableau 1: Activité antivirale de l’EMn et l’IMn purifiées (Gastineau et al., 2012) ............. 23

Tableau 2: Activité antiproliférative de la marennine IMn et EMn purifiées sur

différentes lignées de cellules cancéreuses (Gastineau et al., 2012a) ..................................... 24

Tableau 3: Constantes physiques caractéristiques de la marennine par analyse

spectrophotométrique (Pouvreau et al.,2006) ................................................................................. 27

Tableau 4: Comparaison de la biomasse et la concentration lipidique produite par

différentes microalgues ............................................................................................................................. 31

Tableau 5 : Etude bibliographique de la sensibilité d'organismes cellulaires face aux taux

de cisaillement ............................................................................................................................................... 42

Tableau 6 : Gamme de température supporté par 5 groupes taxonomiques de

microorganismes photosynthétique (Ras et al., 2013) .................................................................. 69

Tableau 7: Evolution de la solubilité du carbone en fonction de la température, PCO2=380

ppm (Le Gouic, 2013) ................................................................................................................................. 71

Tableau 8: liste des macroéléments nécessaires à la croissance microalgale (Grobbelaar

2003) ................................................................................................................................................................. 74

Page 25: Par « Rémi NGHIEM-XUAN

Tableau 9: liste des microéléments nécessaires à la croissance microalgale (Grobbelaar

2003) ................................................................................................................................................................. 75

Tableau 10: Biométrie du clone NCC 495 d'Haslea ostrearia et aspect au début et à la fin

de la thèse ........................................................................................................................................................ 91

Tableau 11: Biométrie du clone NCC 501 d'Haslea ostrearia et aspect à réception du

clone................................................................................................................................................................... 92

Tableau 12: Composition ionique en macro et microéléments de différents milieux de

culture .............................................................................................................................................................100

Tableau 13 : Caractéristiques des PBRs utilisés pour la culture d’Haslea ostrearia ........102

Tableau 14: Comparaison de l'intensité et la capacité à générer des contraintes

hydrodynamiques du PBR Colonne à bulle, Hector et Raceway clos. ....................................236

Page 26: Par « Rémi NGHIEM-XUAN

List of tables

Table 1: Elemental content comparison between the culture media tested .......................121

Table 2: Preliminary results of biomass and Extracellular Marennine (EMn)

productivities in different media. Data is mean ±95% CI for n=3. ..........................................123

Table 3: Salt solution composition for NX medium .......................................................................141

Table 4: Isolated solution for salt solutions realization ..............................................................142

Table 5: Enrichment solution composition for NX medium ......................................................142

Table 6: Iron solution composition for NX medium......................................................................143

Table 7: Metal A stock solution composition for NX medium ...................................................143

Table 8: Isolated solution for Metal A solution realization ........................................................144

Table 9: Metal B stock solution composition for NX medium ...................................................144

Table 10: Isolated solution for establishing Metal B solution ...................................................145

Table 11: Vitamins stock solution composition for NX medium ..............................................145

Table 12: Experimental conditions during the experiment with progressive shear stress

increase for a constant shear rate of 30,000 s-1 ..............................................................................166

Table 13: Nutrient consumption of Haslea ostrearia (rS, QS and YX determination for NO3-,

PO42- and SiO32-) in semi-continuous culture in an airlift PBR .................................................203

Table 14: Repartition and observation under light microscope of the Haslea ostrearia cell

population density after 24 h without manual stirring. Data is mean ± 95% CI for n = 3.

...........................................................................................................................................................................217

Page 27: Par « Rémi NGHIEM-XUAN

1

Introduction générale

Les microalgues et cyanobactéries, parfois appelées microphytes, désignent les

algues unicellulaires capables de produire du dioxygène et de la matière organique grâce

à la lumière. Ces organismes sont consommés depuis des milliers d’années par l’homme.

La première consommation fut décrite en Chine avec une cyanobactérie du genre Nostoc,

puis vînt la découverte de la Spiruline en 1325 ou encore Chlorella en 1890 (Spolaore et

al., 2005). Unicellulaires ou pluricellulaires indifférenciées, ces micro-organismes

photosynthétiques Eucaryotes ou Procaryotes peuvent posséder des flagelles ou non.

Très cosmopolites, on les trouve dans tous les milieux aquatiques exposés à la lumière.

Possédant souvent plusieurs types de reproduction, leur culture monoclonale est réalisée

dans des systèmes de culture appelés photobioréacteurs (PBR), à l’échelle du laboratoire

mais aussi à l’échelle industrielle pour l’intérêt qu’elles représentent vis-à-vis de leur

composition et leur capacité à synthétiser des molécules d’intérêt.

Depuis plus d'un siècle, botanistes et microbiologistes se sont intéressés à une

algue microscopique, une diatomée à protoplasme bleu responsable du phénomène de

verdissement des huîtres. C'est en 1669, que l’Anglais Sprat constate et étudie le

phénomène de verdissement des huîtres. Mais ce n’est seulement qu’en 1820, que Gaillon,

examinant au microscope la matière verte tapissant le sol des parcs à huîtres, constate

qu'elle est formée d'organismes allongés, mobiles, terminés en pointe aux zones apicales,

et chargés d’un pigment bleu-vert. Il leur donne le nom de Vibrio ostrearius qui sera classé

3 ans plus tard dans le genre Navicula par le Français Bory de Saint Vincent. Ainsi,

Navicula ostrearia, apparaît comme l’agent responsable de ce verdissement grâce à sa

sécrétion d’un pigment bleu-vert soluble dans l’eau, dont les premières études chimiques

et spectroscopiques sont effectuées en 1886 par Lankester, qu’il nommera « marennine ».

D’autre scientifiques se succèdent tels que Ranson (1927), Bachrach (1935) et Moreau

(1970) mais les fonctions physiologiques de cette molécule pour la cellule ne reposent, à

l’époque, que sur des hypothèses. C’est seulement en 1978, que Neuville et Daste

attribuent un caractère toxique à la marennine envers d’autres microorganismes. De

nombreuses études suivent et confirment alors les propriétés biologiques :

allélopathique, antioxydante, antibactérienne, antivirale, anti-tumorale et atténuateur de

Page 28: Par « Rémi NGHIEM-XUAN

2

lumière du pigment (Pouvreau et al., 2007 ; Gastineau et al., 2012). Navicula ostrearia sera

placée dans le genre Haslea par Simonsen, en 1974. Presque 10 ans après, Robert (1983),

étudie dans les claires à huîtres du bassin Vendéen, l’apparition de « blooms » saisonniers

de microalgues constitués entre autres d’Haslea ostrearia. Ces travaux sont suivis

notamment par ceux de Turpin (1999), qui étudie les paramètres physicochimiques et

biologiques responsables de la prolifération de cette microalgue lors de ces « blooms »

dans le bassin de Marennes-Oléron.

La présence d’Haslea ostrearia dans les claires à huîtres est responsable du

phénomène de verdissement des huîtres, d’une part lors de l’ingestion de la diatomée par

les huîtres, et d’autre part par la fixation de la marennine extracellulaire lors de la

filtration et de la sélection par les branchies de ces dernières (Gastineau et al., 2012). En

effet, la marennine dissoute dans l’eau des bassins, se fixe préférentiellement alors aux

cellules parenchymateuses branchiales de l’huître, entraînant une amélioration des

qualités organoleptiques de cette dernière (couleur, goût et masse). Cette technique

d’affinage a été valorisée à partir de 1989 à Marennes-Oléron par un label rouge « Fines

de claire verte » qui lui permet de gagner en notoriété et augmenter son prix de vente de

25 % (Robert, 1983 ; CNC, 2014). C’est à partir de 1989 (P. Jaouen, JM. Robert et X.

Gouabau), avec un développement particulier au milieu des années 90’ (via le soutien du

SMIDAP, Syndicat Mixte pour le développement de l’Aquaculture), que le GEPEA et MMS

ont amorcé leurs collaborations en « génie des bioprocédés » sur les thématiques culture

/ séparations / extractions. D’autres expériences, d’abord de R&D, puis entrepreneuriale

de verdissement d’huîtres (« Emeraudes de Vendée », par exemple) ont également été

conduites à Bouin à partir de 1992.

Malgré ces travaux, la culture en conditions intensives d’Haslea ostrearia reste un

défi. Les conditions d’agitation des photobioréacteurs conventionnels ne semblent pas

correspondre à la culture de cette cellule au comportement benthique (Jubeau, 2009). De

plus, des études de résistance cellulaire (Rossignol et al., 1999 ; Rossignol, 1999 ;

Vandanjon et al., 1999) définissent la cellule comme sensible face aux contraintes

physiques de cisaillements hydrodynamiques. Plusieurs systèmes de culture ont alors été

développés par les deux laboratoires GEPEA et MMS, impliqués dans ce travail de thèse,

pour optimiser la croissance et les productivités de cette diatomée. Un photobioréacteur

1L à cellules incluses en gel d’agar a été mis en place de manière à pouvoir agiter le milieu

Page 29: Par « Rémi NGHIEM-XUAN

3

de culture sans perturber la croissance de l’algue (Lebeau et al., 1999). Un autre

photobioréacteur 1L à membrane de filtration immergée a permis le prélèvement de la

marennine extracellulaire en continu sans perturber la croissance des cellules en

préservant la quantité de biomasse au cours de la culture (Rossignol, 1999). Une

technique de culture en bassin extérieur de 10m3 a été mise au point, dans l’optique de

reproduire le phénomène de verdissement en enrichissant de façon optimisée de l’eau de

mer naturelle avec de l’azote, du phosphore, et de la silice (Turpin et al., 2001). Un système

non agité de 100L de culture, a été utilisé par l’Université du Québec à Rimouski (UQAR)

pour la récolte et la concentration de marennine extracellulaire par filtration (Gastineau

et al., 2014). Toutes ces technologies ont permis la récolte d’informations pour la mise en

place de cultures sur plus grands volumes. Cependant, de nombreuses problématiques de

culture, telles que la reproductibilité des résultats de production, et la productivité

globale des systèmes, limitent le changement d’échelle de la production nécessaire au

passage à la phase industrielle. Le design d’un photobioréacteur spécifique pour la

croissance d’Haslea ostrearia, et d’une façon générale, l’optimisation des conditions de

culture, apparaissent ainsi comme nécessaires, à la fois pour valoriser cette souche de

façon plus systématique, mais aussi dans l’optique d’augmenter les productivités en

marennine extracellulaire. Cette production optimisée apporterait d’autres avantages, en

permettant par exemple une analyse plus approfondie de la structure chimique de la

molécule, mais aussi l’étude du potentiel prophylactique de la marennine grâce à une plus

grande quantité de matière première. Cela augmenterait le potentiel d’utilisation en

aquaculture, ainsi que de valorisation de ce pigment dans divers domaines industriels.

C’est dans ce contexte que s’inscrit ce travail de thèse, co-financé par la région des

Pays de la Loire dans le cadre de la contribution pour le projet de dynamique scientifique

collective AMI (Atlantic MIcroalgae). Le projet AMI implique principalement les équipes

de recherche du laboratoire de Génie des Procédés – Environnent et Agroalimentaire

GEPEA, des laboratoires Ifremer Physiologie et Biotechnologie des Algues (PBA) et

Phytotoxines, ainsi que le laboratoire Mer Molécules Santé (MMS). Il a pour objectifs

d’amplifier les collaborations et le rayonnement académique des équipes impliquées et

de contribuer au développement d’une nouvelle filière industrielle. Au sein de ce projet,

cette thèse a eu pour but de définir les conditions et systèmes de culture pour la

production optimisée de la microalgue Haslea ostrearia et de son métabolite d’intérêt

principal, la marennine. Cela a impliqué un travail systématique, sur la définition des

Page 30: Par « Rémi NGHIEM-XUAN

4

paramètres physico-chimiques et biologiques de culture, à la recherche de modes de

culture adaptés. Les résultats ont été comparés aux précédents travaux sur cette espèce

de manière à pouvoir comprendre et lever les problématiques de culture identifiées.

Le chapitre 1 présente l’Etat de l’art sur la production en photobioréacteur

conventionnel, avec un accent sur les conditions spécifiques pour Haslea ostrearia ainsi

que la récupération de la marennine extracellulaire. Les concepts de base de la culture

microalgale et la présentation de l’outil biologique utilisé sont développés dans ce

chapitre. Les études sur la sensibilité cellulaire d’Haslea ostrearia face à l’agitation est

aussi présentée. Une comparaison est effectuée avec des études pour d’autres

microorganismes unicellulaires libres et en tissu, pour une meilleure compréhension du

phénomène de mécano-sensitivité.

Le chapitre 2 présente la description des matériels et méthodes utilisés pour

répondre aux besoins expérimentaux de cette thèse. Les systèmes de cultures, les outils

biologiques ainsi que les méthodes d’expérimentation et de suivi de cultures y sont

décrites.

Le chapitre 3 est le premier chapitre de résultats et a fait l’objet d’une publication

internationale. La problématique physico-chimique des milieux de culture est étudiée et

la conception d’un nouveau milieu type « eau de mer artificielle » est présentée. La

validation de ce milieu artificiel ne sera effectuée qu’à la fin de cette thèse (chapitre final).

Les résultats présentés dans les chapitres suivants ont donc été effectués avec un milieu

d’eau de mer naturelle enrichie « revisité » et également décrit dans le chapitre 3.

Le chapitre 4 concerne l’étude de la sensibilité d’Haslea ostrearia à l’agitation et

plus précisément aux contraintes de cisaillements hydrodynamiques. Des contraintes de

cisaillement ont été appliquées à différentes intensités sur une culture de l’espèce étudiée

pour mesurer sa résistance dans le but d’adapter le système de culture aux particularités

morphologiques d’Haslea ostrearia.

Le chapitre 5 est le dernier chapitre de résultats et vise à optimiser la production

en culture d’Haslea ostrearia. Les résultats sont ainsi analysés en termes de productivité

en biomasse, pigments photosynthétiques et en marennine extracellulaire, face aux

variations des différents paramètres de culture étudiés (pH, T°C, nutriments et lumière).

Les performances de culture de cette espèce en culture agitée et en culture immobilisée

Page 31: Par « Rémi NGHIEM-XUAN

5

sont ainsi comparées. Au final, l’ensemble des données et résultats permet de proposer

des conditions et modes de culture à appliquer pour une production optimisée d’Haslea

ostrearia et de marennine extracellulaire.

Page 32: Par « Rémi NGHIEM-XUAN

6

Chapitre I :

Etude bibliographique

Page 33: Par « Rémi NGHIEM-XUAN

7

7 La diatomée Haslea ostrearia (Simonsen, 1974)

7.1 Taxonomie

Halsea ostrearia est une microalgue eucaryote de la Classe des Bacillariophycées,

aussi appelées diatomées. Le genre Haslea possède à l’heure actuelle 36 espèces décrites,

certaines produisant de la marennine (Gastineau et al., 2012). Ces diatomées

photosynthétiques possèdent un exosquelette siliceux nommé frustule. Ils forment un

maillon fondamental des cycles du carbone et de la silice en milieu aquatique (Kemp et al.,

2006) constituant majeur du phytoplancton (50 % de la production primaire océanique

globale) et responsable d’un quart de l’oxygène que nous respirons. Ces organismes

représentent environ 1 000 000 d’espèces dont 200 000 diatomées et 20 000 seulement

seraient décrites (Mann et Droop, 1996 ; Guiry, 2012).

Selon AlgaeBase, en 2018, 14 803 espèces de diatomée sont décrites. Elles sont très

utilisées en mytiliculture et en ostréiculture. L’élevage de bivalves intègre en effet

couramment des systèmes de cultures de microalgues pour l’amélioration des

rendements.

Figure 1: La phylogénie d'Haslea ostrearia au sein des Eucaryotes des grandes Classes d’algues

Les diatomées forment l'une des classes des algues jaunes et brunes (ou

Chromophytes), caractérisées par la présence de pigments associés à la chlorophylle (a et

c) tels que la fucoxanthine, qui donne aux chloroplastes une couleur brune ou jaune,

contrairement aux algues vertes et rouges du règne des archaeplastides (Fig. 1).

Page 34: Par « Rémi NGHIEM-XUAN

8

7.2 Cytologie

Le genre Haslea se définit d’abord par sa forme pennée. Haslea ostrearia quant à elle,

possède deux vacuoles apicales, lieu de stockage de la marennine (dite intracellulaire

IMn) mais on retrouve aussi l’IMn sous forme de globule dans le cytoplasme (Fig. 2)

autour des chloroplastes. Généralement au nombre de deux, ces chloroplastes en forme

de longues bandes sont disposés de manière pariétale (Nassiri et al.,1998).

Figure 2: Haslea ostrearia au microscope optique, légendé

La taille standard de cellules jeunes est estimée entre 60-120µm de long et 6-12µm

de large. Le sillon est droit, présent sur les deux valves et ses terminaisons sont peu

accentuées. Le frustule siliceux est composé d’une partie inférieure, l’hypothèque incluse

dans la partie supérieure, l’épithèque et reliées par la copulae qui se détache lors de la

division cellulaire (Fig. 3).

De nombreuses diatomées, présentent une fente, appelée raphé, interrompue par

un nodule central de silice. Ce raphé est à l’origine de la communication avec le milieu

extérieur et de l’excrétion de mucilage (composés polysaccharidiques). La formation de

gangues polymériques autour des diatomées va leur conférer des propriétés particulières

d’adhésion, en fonction de la composition de ces exopolysaccharides.

Les diatomées sont capables de moduler la composition de ces exopolymères

pendant la formation des biofilms benthiques selon les besoins structuraux ou

environnementaux. En outre, ces micro-algues peuvent également produire des lipides et

Figure 3: Représentation connective du frustule des diatomées (frustule)

Page 35: Par « Rémi NGHIEM-XUAN

9

acides gras, qui influencent l’hydrophobicité de leur surface. Après nettoyage du mucilage,

en microscopie électronique à balayage, le frustule apparaît comme étant constituée de

bandes longitudinales parallèles au sillon. La couche extérieure du frustule est constitué

de fissures longitudinales (Fig. 4.3) tandis que la couche intérieure est constituée d'une

structure en grille disposée en rangées longitudinales et transversales (fig. 4.4)

(Smithsonian Marin Station, 2015).

Figure 4: zone centrale de la valve (3) extérieure (4) intérieure SEM (Smithsonian Marin Station, 2015)

Figure 5: zone apicale du frustule SEM (Smithsonian Marin Station, 2015)

Le sommet de la valve possède un raphé terminal entouré de micro orifices (Fig. 5)

au niveau la zone de sécrétion de la marennine (dite extracellulaire EMn) (Nassiri et

al.,1998).

7.3 Reproduction

H. ostrearia est une microalgue qui possède deux types de reproduction : asexuée

et sexuée.

Page 36: Par « Rémi NGHIEM-XUAN

10

7.3.1 Reproduction asexuée

La reproduction végétative des diatomées s’effectue par fission binaire

(scissiparité), et deux nouveaux individus sont formés à partir d’une cellule mère. Le

frustule de la cellule mère se sépare en deux pour donner à l’une de ses deux cellules filles

son hypothèque et à l’autre son épithèque. Les deux thèques de la cellule mère forment

alors l’épithèque pour les cellules filles qui vont fabriquer leurs propres hypothèques.

L’hypothèque étant plus courte que l’épithèque, ce type de multiplication a pour

conséquence de réduire la taille de la cellule de générations en générations, à une vitesse

moyenne de 3,74± 0,33 µm/mois en conditions données (Davidovich et al., 2009) (Fig. 6).

Figure 6: Représentation de la re-distribution du frustule au cours de la croissance d'Haslea ostrearia

Chaque partie du frustule (épithèque et hypothèque) de la cellule mère se

transmettent aux cellules filles, l’état des valves de silice persistent théoriquement

« éternellement ». Dans le cas d’Haslea ostrearia l’apparition, assez rapide au cours de la

croissance, de frustules déformés dans les cultures au cours du temps est observé (Tomas,

1996) (Fig.7), et persistent au cours de la culture.

La baisse de taille de la cellule represente une problématique essentielle de la

culture d’H. ostrearia. Cette baisse engendrant une décroissance de performance en

productivité de la cellule expliqué Chapitre I 2.3, et ne pouvant cryopréserver la cellule,

plusieurs souche de génétiques variables sont donc utilisé dans cette thèse. Les résultats

Figure 7: Haslea ostrearia déformé au microscope optique

Page 37: Par « Rémi NGHIEM-XUAN

11

de performances obtenu sont donc théoriquement peu comparablent mais sont comparé

malgré tout due aux conditions de préservations actuelles.

7.3.2 Reproduction sexuée

La reproduction sexuée est importante pour la récupération de la taille originelle

des cellules. Elle est dite hétérothallique et met en jeu deux éléments (mâle et femelle)

morphologiquement semblables, mais physiologiquement différents pour former des

auxospores dans lesquelles les cellules se débarrassent de leurs thèques pour donner une

nouvelle cellule de taille maximale (60 – 140µm). L’auxosporulation s’effectue lorsque la

cellule atteint une taille critique aux alentours de 50-60±0,4µm de longeur (Mouget et al.,

2009). La lumière est un facteur clef de l’auxosporulation. Les faibles luminosités (<50

µmolhv. m-2. s-1) et les faibles photopériodes (6–10h) déclenchent la reproduction sexuée

lorsque la souche est aux alentours de 50% de sa longueur d’origine. La reproduction

sexuée, même isogamique fait intervenir deux cellules (gamétange). À ce moment-là,

aucune présence de mucus n’est observée, les plastes migrent vers le centre de la cellule

au-dessous de la valve, très tôt dans la prophase. Chaque gamétange produit deux

gamètes de taille égale qui fusionnent (cellules + et - de 2 cellules différentes) ; les deux

zygotes gravitent alors près de la ceinture entre les deux thèques de la cellule mère puis

l’auxospore s’étend entre les thèques pour redonner une cellule de grande taille.

7.4 L’écosystème en claire ostréicole

Haslea ostrearia est une souche euryhaline retrouvée dans de nombreux

écosystèmes tempérés et tropicaux (Gastineau, 2011). Les cultures monoclonales

d’Haslea ostrearia conservées à la NCC (Nantes Culture Collection), proviennent en

majorité d’échantillons prélevés en Baie de Bourgneuf. Dans cette ancienne région de

saliculture, l’activité ostréicole est implantée depuis 1956. La croissance d’Haslea dans ces

claires à huîtres apparaît par pic et atteint une densité maximale de 2 x 105 cellules. L-1 au

mois d’Août et jusqu’en Octobre/Novembre (Robert, 1983).

Page 38: Par « Rémi NGHIEM-XUAN

12

Figure 8: Courbe de température moyenne annuelle en Baie de Bourgneuf (Climate data)

Le mois le plus chaud de l'année est celui d’Août avec une température moyenne

de 19,6 °C (Fig. 8). Le laboratoire MMS estime la température de croissance optimale dans

une moyenne de 16°C avec la possibilité pour la souche de tolérer des températures de 8

à 26 °C au cours de l’année (Robert, 1983).

Haslea ostrearia est une microalgue considérée comme benthique qui possède un

cycle tychopélagique en milieu naturel. Cette souche tombe au fond de la claire et produit

en quasi-permanence (sauf durant la reproduction sexuée), un biofilm qui la protège des

agressions et forme un « tissu » de cellules (Rincé et al., 1999). De plus, elle produit de la

marennine au cours de sa croissance. Cependant, la sécrétion semble plus importante en

phase benthique (Fig. 9).

Page 39: Par « Rémi NGHIEM-XUAN

13

Figure 9: représentation du cycle tychopélagique d'Haslea ostrearia dans une claire à huîtres (D’après Robert, 1983)

Dans l’écosystème claire ostréicole (Fig. 9), les cellules d’Haslea commencent en

phase planctonique (1 et 2). Pendant environ un mois durant cette première phase la

production de marennine est faible. S’ensuit une phase benthique (3) pendant 2-3 mois

où elles forment en se multipliant activement un biofilm à l’interface eau-sédiment.

Débute ensuite une phase épipélagique en remontant à la surface sous forme de « crème

flottante » due à l’accumulation intracellulaire de globules lipidiques, et s’acheve par la

mort cellulaire (Robert, 1983).

Page 40: Par « Rémi NGHIEM-XUAN

14

Figure 10: Relation écologique entre Haslea ostrearia et Crassostrea gigas en claire à huître

En claire ostréicole Haslea ostrearia va servir de nourriture à l’huître est va

excréter la marennine qui sera responsable du phénomène de verdissement des huîtres

(Fig. 10). Ce pigment est filtré par l’huître et/ou d’autres bivalves et va se fixer à leurs

branchies qui vont prendre une teinte bleue verte au bout de quelques heures (Gastineau

et al., 2014). Ce pigment offre donc une coloration verte mais surtout de meilleures

qualités organoleptiques grâce à un affinage de 28 jours. La marennine filtrée et fixée

correspond aussi bien à la marennine sécrétée par les cellules (EMn) qu’à celle relâchée

lorsque l’huître consomme la biomasse d’Haslea (IMn). La croissance de l’huître entraîne

la production de fèces, pseudo fèces et mucus libérés dans le milieu qui fournissent les

nutriments pour la croissance des microalgues (Barillé et Cognie, 2000 ; 1998). Une fois

la marennine fixée aux branchies, la capacité de filtration du bivalve est diminuée si bien

que la relation écologique Haslea ostrearia/Crassostrea gigas, n’est pas encore totalement

qualifiée.

7.5 Photosynthèse

Commun au métabolisme d’Haslea ostrearia et à tout organisme végétal, La réaction

de photosynthèse selon l’équation 1 globale est la suivante :

Équation 1

Dans ce phénomène, l’énergie photonique (hv) est convertie par les cellules

végétales en énergie chimique (ATP, NADPH) utilisant le carbone inorganique et l’eau

Page 41: Par « Rémi NGHIEM-XUAN

15

pour la production de matière organique et de dioxygène. Cette conversion s’effectue

grâce à deux étapes essentielles de la photosynthèse. La phase lumineuse ou phase claire,

désigne la première phase de la photosynthèse, elle consiste à la conversion de l'énergie

lumineuse en énergie chimique sous forme d'ADP et le phosphate en ATP, (conversion de

l'énergie lumineuse en énergie chimique).

Figure 11: Représentation de mécanisme de phase claire dans le phénomène de photosynthèse.

Situés dans la membrane thylacoïde (Fig. 11), les complexes de protéines

membranaires contenant des pigments (donnant leurs couleurs aux thylakoïdes) sont

appelés photosystèmes I et II et permettent la réalisation de la réaction en phase claire.

La composition pigmentaire des thylakoïdes est alors essentielle pour les réactions

photosynthétiques. Situé sur les photosystèmes, le complexe de pigments accessoires

(Chlorophylle c et caroténoïdes pour les diatomées (De Reviers, 2002)) permet la

captation de différentes longueurs d’ondes de la lumière blanche, toutes apportées au

centre de réaction composé de chlorophylle a. Grâce à cette composition pigmentaire, le

complexe PSII absorbe l’énergie photonique et le transfert vers la plastoquinone sous

forme d’électrons après extraction de deux molécules d’eau (production de dioxygène)

(Fig. 11). S’en suit une réaction en chaîne d’oxydo-réduction au sein de la membrane

permettant au final, le stockage de cette énergie sous forme de composés chimiques utiles

(NADPH, ATP). Une fois la réaction déclenchée par le PSII, le PSI permet la synthèse de

Page 42: Par « Rémi NGHIEM-XUAN

16

NADPH grâce à un complexe d’enzymes ferriques et à la précédente hydrolyse effectuée

par le PSII. L’ATP est ensuite synthétisée grâce à l’ATP synthase qui utilise les protons et

les phosphates pour transformer l’ADP en ATP (Gest, 2002).

Les molécules de NADPH servent alors, de donneurs d'électrons utiles pour les

réactions ultérieures indépendantes de la lumière, effectués en phase sombre. Cette phase

à l’obscurité permet la synthèse de matières organiques avec utilisation de l’énergie

chimique précédemment synthétisée en phase claire. Cette phase sombre met en œuvre

plusieurs réactions chimiques cycliques en trois étapes appelées cycle de Calvin (Fig. 12).

Figure 12: Représentation schématique du cycle de Calvin.

Le Cycle de Calvin débute par la fixation du CO2 par une enzyme clef de la

photosynthèse, la ribulose 1,5 biphosphate carboxylase/oxygénase (RuBisCO). Cette

enzyme permet la production de deux molécules de 3-phosphoglycérate grâce à la

consommation de deux molécules de NADPH; H+ et de trois molécules d'ATP, démarrant

la phase de fixation du carbone inorganique. Une molécule de 3-phosphoglycérate va alors

être réduite en 1,3-bisphosphoglycerate puis en Glycéraldehyde 3-phosphate (G3P) par

l’utilisation respective d’ATP et de NADPH+H+ en phase de réduction. Une phase de

Page 43: Par « Rémi NGHIEM-XUAN

17

régénération permet ensuite la synthèse de Ribulose 5-phosphate pour le renouvellement

de la RuBiscCO et la reprise du cycle (Karp et al., 2010).

Comme son nom l’indique, la RuBisCO possède deux activités catalytiques :

- L’activité carboxylase qui lui permet la formation de deux molécules d’acide

phosphoglycérique à partir du RuBP.

- L’activité oxygénase qui permet, à partir du RuBP, la formation d’une molécule

d’acide phospho-glycolique et d’une molécule d’acide phosphoglycérique (PGA).

Ce qui a pour conséquence de limiter/stopper la photosynthèse, ne permettant pas

la poursuite du cycle de Calvin.

Les conditions dans lesquelles l’enzyme va privilégier une activité plutôt qu’une autre

dépend de l’environnement dans lequel l’organisme photosynthétique se trouve.

L’activité carboxylase de l’enzyme est dominante en quantité équivalente d’oxygène et de

dioxyde de carbone, car l’affinité de son site actif pour le CO2 est plus importante que pour

l’O2. Cependant, une concentration plus élevée de dioxygène déclenche l’activité

oxygénase. La température faisant varier la solubilité du CO2 et de l’O2 dans l’eau est donc

un paramètre essentiel pour la croissance de la microalgue.

Figure 13: Schémas des cycles de l'activité oxygénase et carboxylase de l'enzyme RuBisCO.

L’activité d’oxygénase forcée de la RuBisCO fait intervenir le phénomène de

photorespiration (Fig. 13), une voie métabolique coûteuse pour la cellule, car elle réduit

le taux de production du 3-phosphoglycérate par rapport à l’activité carboxylase. De plus,

on peut aussi constater une perte nette en carbone et la libération de NH3, un composé à

Page 44: Par « Rémi NGHIEM-XUAN

18

détoxiquer par la cellule grâce à un processus, lui aussi, coûteux en énergie (Leegood,

2007).

Selon des chercheurs de Belgique, France, Italie et des Etats-Unis (Bailleul et al.,

2015), chez les diatomées, le mécanisme photosynthétique est le même que chez les

microalgues de manière générale, mais la capacité des diatomées à dominer la

communauté phytoplanctonique, leurs nombreuses caractéristiques physiologiques et

leur évolution pigmentaire, supposent une évolution différente des algues vertes et

rouges. Les diatomées possèdent une interaction/communication soutenue entre

chloroplastes et mitochondries, respectivement responsable de la photosynthèse et de la

respiration. La gestion de l’énergie chimique (ATP et NADPH) est différente pour la

fixation du carbone, car un échange soutenu de ces molécules permet de combler

rapidement le déficit du rapport ATP/NADPH lorsque celui-ci est inférieur au besoin de

la phase sombre de la photosynthèse où le CO2 est converti en sucre. Cette information ne

permet pas d’anticiper les comportements complexes des diatomées et d’Haslea ostrearia

dans notre cas, mais suppose que cette microalgue a un comportement et une relation à

la lumière différente des croissances de chlorophycées en photobioréacteur, considérées

comme modèles car cultivées communément à l’échelle laboratoire et industrielle.

Dans le cas des plantes supérieures, la régulation Redox via la thioredoxine du cycle

de Calvin est essentielle pour la régulation des dérivés réactifs de l'oxygène (DRO, en

anglais reactive oxygen species, ROS), espèces chimiques oxygénées (radicaux libres, ions

oxygénés (anion superoxyde O2−, de l'oxygène singulet O2), peroxydes d’hydrogène H2O2

etc …) capables d'oxyder les protéines, l'ADN et les membranes des cellules (attaque des

lipides constitutifs par peroxydation lipidique) (Holmgren, 1989). Les DRO peuvent être

d'origine exogène (produits par des rayonnements ionisants) ou bien endogène (sous-

produits du métabolisme normal de l'oxygène souvent importants dans la communication

entre les cellules). Les cellules subissant ce processus de dégradation se défendent à l’aide

souvent, de petites molécules antioxydantes telles que l'acide ascorbique (vitamine C), les

tocophérols (vitamines E), ou encore de molécules polyphénoliques plus grosses.

Cependant, chez les diatomées, la production de thioredoxine semble fortement

réduite par rapport à d’autres mécanismes de régulation. D’après les travaux de Sachse et

al. (2013), la transcription des enzymes intervenant dans le cycle de Calvin de

Phaeodactylum tricornutum cultivé en cycle jour/nuit et nuit complète a été quantifiée.

Page 45: Par « Rémi NGHIEM-XUAN

19

Les gènes codant pour la phosporibulokinase (PRK qui catalyse la phosphorisation du

ribulose 5-phosphate (RuP) en ribulose 1,5-bisphosphate (RuBP) pour la régénération de

la RuBISCO) et la glycéraldéhydedéshydrogénase (GAP-C1 qui catalyse la première étape

de la glycolyse, convertissant le D-glyceraldehyde 3-phosphate (G3P) en 3-phospho-D-

glyceroyl phosphate, essentiel à la maintenance du niveau d’ATP et du métabolisme des

carbohydrates) sont fortement sollicités durant la phase lumineuse. Sachse évoque

l’hypothèse que la surproduction de PRK et GAP-C1 soit un des mécanismes évolutifs des

diatomées pour compenser le manque de thioredoxine et donc optimiser la croissance

cellulaire face aux agressions des DRO.

7.6 La composition pigmentaire

La présence de certains pigments accessoires ou surnuméraires confère aux algues

la capacité d’absorber différentes longueurs d’ondes, et ainsi des propriétés spécifiques

de survie dans différents environnements. Ces pigments forment des complexes

protéiques appelés antennes collectrices, capables de se charger en énergie lumineuse par

l’interception de photons de longueurs d’ondes dépendantes de la composition des

pigments accessoires (Fig. 13).

Figure 14: Représentation de la compostions pigmentaire des photosystèmes et la circulation de l'énergie lumineuse chez une chlorophycée

Les photons sont récoltés par les pigments (chlorophylle b, c, et caroténoïdes de

protection) qui permettent leurs transmissions aux centres réactionnels, composés de

chlorophylle a, (aussi appelé P680). L’apport d’énergie va exciter le P680 qui va ainsi

libérer l’électron à la plastoquinone, démarrant la réaction de photosynthèse.

Page 46: Par « Rémi NGHIEM-XUAN

20

Cette composition pigmentaire a pour rôle d’accroître la gamme d’absorption de la

lumière et est dépendante de l’espèce de l’organisme photosynthétique étudié, mais aussi

de l’intensité lumineuse à laquelle la cellule est soumise (Masojidek et al., 2013). Selon De

Reviers (2002) la majorité des diatomées ont une composition pigmentaire particulière

composée de chlorophylle a, c, et de fucoxanthine. Cela est confirmé par Carreto et al.,

(1976) et Kuczynska et al., (2015), qui précisent que les diatomées possèdent un groupe

de pigments caroténoïdes photoprotecteurs incluant le bêta-carotène, les xanthophylles

comme la diadinoxanthine, la diatoxanthine, la violaxanthine, l’anthéraxanthine et la

zéaxanthine.

Figure 15: Spectre d'absorption des pigments photosynthétiques (Devred et al., 2014)

La figure 15 présente le spectre d’absorption des différents pigments

photosynthétiques pouvant être retrouvés chez les organismes photosynthétiques. Ces

pigments principaux et accessoires sont dits photosynthétiques du fait de leur rôle dans

la réaction de photosynthèse. Cependant, dans le cas d’Haslea ostrearia, la sécrétion de la

marennine n’a pas été identifiée comme liée à la captation d’énergie lumineuse, et n’est

donc pas considérée comme un pigment photosynthétique (Schubert et al., 1995). Sa

production semble toutefois avoir une relation étroite avec la lumière et donc un lien avec

Page 47: Par « Rémi NGHIEM-XUAN

21

le métabolisme photosynthétique (Neville & Daste, 1978 ; Tremblin et al., 2000 ; Mouget

et al., 2005).

8 La marennine

8.1 La structure chimique

L’analyse élémentaire a été effectuée au CNRS par le laboratoire ICP-MS de Nantes,

par spectrométrie de masse ICP (Pouvreau et al., 2006). La marennine est donc composée

ainsi :

- IMn O = 48% (±0.50), C = 39.55% (±0.15), N=5.70% (±0.30), H=4.50% (±0.25),

Na=1.25% (±0.05), Ca = 1% (±0.05)

- EMn O = 53.5% (±0.50), C = 36% (±0.60), H = 3.80% (±0.15), N=1.70% (±0.25), Na =

3% (±0.05), Ca = 2% (±0.05)

Figure 16: 2D RMN (a) Spectre 1H–13C EMn d’Haslea ostrearia Correspond au spectre 1D 1H (Gastineau et al., 2014)

Cependant, Les études de spectrométrie de masse, spectroscopie Raman, RMN et

RMN 2D ne permettent pas de déduire la structure exacte. Les lignes noires (Fig. 15 a)

séparent les zones comportant des carbones reliés à un ou plusieurs atomes d’hydrogène

; le pyranose générique (Fig. 15 b); la région cyclique 1H–1H TOCSY de l’EMn avec 90 ms

Page 48: Par « Rémi NGHIEM-XUAN

22

de temps de mélange est représentée figure 15 c. Le signal fort 1H observé à 3,65 ppm

représente une impureté (Gastineau et al., 2014).

La différence de structure d’IMn/EMn est sensible chez Haslea ostrearia. L’EMn

semble contenir plus de protons anomériques en conformation alpha (4,9 – 5,7 ppm).

Cependant, les deux formes contiennent un cycle organique aromatique (3,4 – 5,4 ppm)

ainsi qu’un fort signal à 1,22 ppm révélant une région aliphatique (c.a.d, une longue chaîne

carbonée). Selon ces résultats, la marennine serait composée d’un cycle carboné

(Aromatique/benzénique) relié à une longue chaîne carbonée par un carbone

anomérique.

8.2 Les propriétés physico-biochimiques de la marennine

8.2.1 Les propriétés biochimiques

- Activité antibactérienne:

L’impact des deux types de marennine IMn et EMn a été testé par le laboratoire du

MMS sur trois espèces Polaribacter irgensii, Pseudoalteromonas elyakowii et Vibrio

aesturianus. Sélectionnés pour leur dangerosité, Pseudoalteromonas elyakowii et Vibrio

aesturianus ont été identifiés comme des souches de bactéries dangereuses pour la

croissance des huîtres, Polibacter étant une souche terrestre. Le potentiel antibactérien

de la marennine a été évalué par la méthode de diffusion de disque en gel d’agar. Il a été

constaté l’augmentation du cercle d’inhibition de croissance sur gélose avec

l’augmentation de la concentration de marennine utilisée. L’inhibition de la croissance des

bactéries est effective même avec une faible concentration de marennine de 1µg.ml-1.

Cependant, aucune zone d’inhibition n’a été observée pour la croissance de bactéries

terrestres (Gastineau et al., 2012).

- Activités antivirale et cytotoxique:

Cette étude a été effectuée sur la lignée de cellules Vero (Vero cell African green

monkey kidney cell), très utilisées dans la recherche de toxines bactériennes ainsi que

comme cellule hôte pour la culture de virus. Cette lignée cellulaire est aneuploïde ce qui

permet de nombreuses réplications cellulaires sans l’apparition de caractéristiques

sénescentes. L’application de IMn et EMn sur la lignée Vero montre des changements de

morphologie après 3 jours de traitement. Cependant, aucun effet cytotoxique n’a été

Page 49: Par « Rémi NGHIEM-XUAN

23

constaté avec la forme EMn. Concernant l’activité antivirale de la marennine, elle a été

comparée à celle du Zovirax (molécule référence contre le virus de l’herpès HSV-1 et sa

faible cytotoxicité, aussi commercialisée sous le nom d’Activir) (Table. 1).

Tableau 1: Activité antivirale de l’EMn et l’IMn purifiées (Gastineau et al., 2012)

Code CC501(µg/mL) EC502 (µg/mL)

Zovirax >200.0 0.2

IMn marennine 107.2 24.0

EMn marennine >200.0 27.0

Cependant, les deux formes EMn et IMn semblent posséder une activité anti

herpétique très efficace. Avec un EC50 de 24-27µg. ml-1, 2 fois plus important que celui du

Zovirax.

- Activité antiproliférative:

L’expérience a été effectuée sur plusieurs lignées de cellules tumorales humaines :

M113 (mélanome) ; SKOV3 et SHIN3 (cancer ovarien) ; SW116 (cancer du côlon) ; R3111

(cancer des reins) ; 1355 (cancer des poumons) ; et MCF7 (cancer du sein) provenant du

département INSERM U463 (Université de Nantes). La présence d’IMn et d’EMn permet

d’engendrer une différence de morphologie chez les cellules cancéreuses, les cellules

n’adhèrent plus sur les surfaces et prennent une forme globulaire sauf pour la lignée SHIN

3. Aucune dégénérescence ou toxicité n’a été observée sur les cultures, mais la marennine

semble inhiber la croissance de certaines lignées cellulaires (Table. 2) (Gastineau et al.,

2012).

1CC50 représente 50% de la concentration cytotoxique 50%, la concentration qui réduit l'absorbance des cellules pseudo-infectées à 50% de celle du témoin. 2EC50 représente 50% de la concentration antivirale, la concentration qui protège de 50% la cellule contre

l’infection virale HSV-1.

Page 50: Par « Rémi NGHIEM-XUAN

24

Tableau 2: Activité antiproliférative de la marennine IMn et EMn purifiées sur différentes lignées de cellules cancéreuses (Gastineau et al., 2012a)

Lignée cellulaire IC50

3 (µg. mL-1±SE)

EMn IMn

Cancer des

poumons

1355 1,10 ± 0,56 0,79 ± 0,35

Cancer des

poumons

NSCLC-N6 14,4 ± 9,8 nd4

Cancer du sein MCF-7 6,5 ± 3,5 22,2 ± 8,7

Cancer du rein R3III 25,9 ± 9,8 36,7 ± 1,1

Mélanome M113 >100 82,7 ± 33,0

Cancer des ovaires SKOV-3 >100 no5

Cancer des ovaires SHIN3 no no

Cancer du colon SW116 >100 >100

La lignée SHIN3 ne semble pas sensible aux différentes doses des deux formes de

marennine. SKOV-3, SW116 et M113 montre une inhibition à grande concentration de

marennine (100μg. mL-1, 31 ± 9%, 36 ± 4%, and 49 ± 3%, respectivement) pour EMn.

Seulement SW116 est sensible à l’IMn à haute concentration (100 µg. mL-1, 38 ± 7%). Les

deux formes de marennine montrent d’intéressants résultats sur l’inhibition des cellules

cancéreuses de poumons (1355 et NSCLC-N6). L’activité antiproliférative illustrée par

l’IC50 se situe entre 0.79 - 82.7 μg. mL-1 pour l’IMn et entre 1.10 - 25.9 μg. mL-1 pour l’EMn.

Les propriétés de la marennine sont actuellement étudiées pour une valorisation

potentielle dans différents domaines. La marennine pourrait être valorisée dans

l’industrie agroalimentaire, en tant que pigment bleu naturel comestible, ainsi que dans

le domaine de l’ostréiculture (CNC). La région de Marenne-Oléron est la principale région

de production française avec les huîtres labellisées « Fines de claire vertes », dont la

3Analyses statistiques des IC50 déterminées avec ± SE pour n = 6 (ANOVA, SNK; p < 0.05). IC50 représente la concentration minimum de marennine nécessaire pour provoquer 50% de l’inhibition de croissance cellulaire. 4nd = non déterminé 5no = pas d’inhibition détectée.

Page 51: Par « Rémi NGHIEM-XUAN

25

marennine donne aux huîtres cultivées dans cette région leur couleur bleu-vert unique.

Le phénomène de verdissement des huîtres est encore peu compris (Moreau, 1968 et

Gastineau et al., 2012b). C’est un phénomène affectant plusieurs tissus des mollusques

(manteau, branchies, épithélium intestinal), se produisant à différentes intensités et

fréquences au cours de l’année et dans de nombreuses régions, mais ne présente une

grande importance que dans le bassin de Marenne. La variation de l’efficacité du

phénomène de verdissement des huitres est importante d’une claire à une autre.

L’utilisation de la marennine pourrait être un moyen de limiter la mortalité des larves de

bivalves, due à son activité antibactérienne contre les pathogènes marins (Turcotte et al.,

2016 ; Latour et al., 2018). Depuis 2008, la filière ostréicole subit une surmortalité des

huîtres jeunes causées par le virus OsHV1. Cette crise est internationale et a entraîné une

forte baisse de la production d’huîtres en France, passant d’environ 126 000 tonnes en

2006 à 84 100 tonnes en 2010/2011 et 79 200 en 2014/2015 (CNC).

Les caractéristiques antibactériennes, anti-tumorales et anti-virales pourraient

être valorisées en industrie pharmaceutique, le marché des produits antiviraux devenant

plus en plus important depuis 2015 dû à l’augmentation de la population. En contrepartie,

l’expiration des brevets d’anciens antiviraux tels que Sustiva, Combivir, Tenofovir,

Telbivudine, Tamiflu et Relenza (pour 2017) va relancer la compétition des médicaments

génériques. En 2013, le marché des antiviraux était estimé à 22,1 milliards de dollars et

espère atteindre les 30,1 milliards de dollars pour 2017 (Franco, 2012). Le marché des

traitements anti-herpès est quant à lui, estimé à 4,9 milliards de dollars pour 2017. Une

portion de 0,7% de la population globale adulte, entre 15-49 ans est touchée par le virus

de l’herpès chaque année. En 2008, environ 600 millions d’adultes sont touchés par le

virus de l’herpès (Yasobiotech). Les dépenses mondiales en anticancéreux quant à eux,

atteignent $100 milliards (90 milliards d'euros), en 2014, en hausse de 10,3% par rapport

à l'année précédente, selon un rapport consacré à l'oncologie du cabinet IMS Health. Ce

montant représente 10,8% de l'ensemble des dépenses de médicaments à travers le

monde et inclue les traitements de support, comme les antinauséeux ou les traitements

de l'anémie. Le cabinet spécialisé prévoit qu'elles atteindront 117 à 147 milliards de

dollars en 2018, soit une croissance annuelle cumulée estimée entre 6% et 8%.

Les pigments bleus naturels étant rares, la marennine est un candidat pour la

production de cosmétiques d’origine naturelle. Le bleu indigo peut être produit à partir

Page 52: Par « Rémi NGHIEM-XUAN

26

de fleurs d’Indigofera tinctoria, commun en Asie et en Afrique et surtout connu en Europe

depuis le 15ème siècle pour la teinture des vêtements. Cependant, en termes de colorant

alimentaire d’origine naturelle, seule la phycocyanine, extraite de la spiruline, est

autorisée par les normes européennes (EFSA 2011).

8.2.2 Les propriétés physiques

La marennine est un pigment bleu-vert hydrosoluble possédant un déplacement

bathochrome réversible légèrement différent entre ses deux formes IMn et EMn.

Figure 17: Observation de marennine EMn à pH basique (gauche) et acide (droite) (Jubeau, 2009)

Ces deux formes changent de couleur en fonction du pH et de leur pKa. Les pH

basiques donnent une couleur vert sombre, alors qu’une partie des pH acides donnent une

couleur bleu ciel à la marennine (Fig. 17). La valeur du pKa de L’IMn et EMn détermine

leur limite de changement de couleur, respectivement 4,74 ± 0,05 et 4,02±0,02 (Fig. 18a

et b).

Figure 18: Absorbance de l'IMn (a) et EMn (b) en fonction du pH (Pouvreau et al.,2006)

Page 53: Par « Rémi NGHIEM-XUAN

27

L’analyse par spectrométrie nous renseigne sur le coefficient d’extinction molaire, le

point isobestique et le pKa de la marennine (Table. 3).

Tableau 3: Constantes physiques caractéristiques de la marennine par analyse spectrophotométrique (Pouvreau et al.,2006)

Les deux types de marennine sont complètement solubles dans l’eau pure avec une

concentration de saturation atteinte à 25g. L−1 à 25°C pour les deux formes ; Insolubles en

solvants organiques avec précipitation de 20 à 60°C. Après étude en spectrométrie de

masse le poids moléculaire de la marennine est estimé à 10,751±1 KDa pour IMn et 9,893

± 1 KDa pour EMn (Pouvreau et al., 2006).

8.3 Le métabolisme de production de marennine

La voie métabolique de production de marennine est encore inconnue. La question

même de l’origine et de la nécessité pour la microalgue du métabolite reste hypothétique,

confrontant la théorie du métabolite primaire à celui du métabolite secondaire.

Un métabolite primaire est directement impliqué dans la croissance, le

développement et la reproduction normale d'un organisme. Indispensable, il a en grande

majorité, une fonction physiologique intrinsèque, souvent partagée par d’autres

organismes taxonomiquement éloignés. A l’inverse, un métabolite secondaire n'est pas

directement impliqué dans ces processus physiologiques fondamentaux indispensables,

mais possède une fonction écologique ou relationnelle importante. De manière générale,

un métabolite secondaire se retrouve présent dans un ensemble taxonomiquement

restreint d'organismes (Plantes, Champignons, Bactéries etc...). Chez les plantes,

beaucoup de métabolites secondaires se retrouvent sous la forme de « polyphénols » ou

« composés phénoliques ». Souvent d'un poids moléculaire élevé, ce sont des molécules

aromatiques constituées d’un groupement phényle (C6) et d’un hydroxyle (-OH) dérivant

d’acides aminés aromatiques (e.g. tyrosine, phénylalanine) comme les flavonoïdes, les

tanins ou encore les lignines (Scriban, 2006).

Page 54: Par « Rémi NGHIEM-XUAN

28

Certaines études effectuées par le laboratoire MMS, soutiennent l’hypothèse d’une

production continue de marennine corrélée à celle de sa biomasse. L’intensité lumineuse

semble jouer un rôle important dans la production de biomasse et de la marennine qui

n’est pas seulement libérée lors d’une limitation en nutriment mais aussi au cours de la

phase exponentielle (Mouget et al., 2004). D’autres travaux montrent l’affinité d’Haslea à

la lumière bleue pour la production de biomasse et de marennine (Mouget et al., 2005)

(Fig.20). D’autre part, les cellules perdant de leur taille au cours du temps, produisent en

moyenne des quantités de marennine plus faibles (Mouget et al., 2005) (Fig.19).

Figure 19: Courbes de croissance d’Haslea ostrearia sous lumière blanche (WL), bleu (BL), vert (GL), jaune (YL), ou rouge (RL) à faible (20μmol photons m−2 s−1) et fort intensité lumineuse (100μmol photons m−2 s−1)

(Mouget et al., 2005)

Page 55: Par « Rémi NGHIEM-XUAN

29

Figure 20: Contenu en marennine de cellules d’Haslea ostrearia de différentes tailles (longueur de cellule) produites sous lumière bleu (⧯) ou blanche (⧮) (Mouget et al., 2005)

Concernant les figures 19 et 20, la concentration cellulaire est plus importante à

haute irradiance avec des courbes de croissance similaires entre elles, indépendamment

du spectre de lumière appliqué. Cependant, en phase exponentielle, la concentration

cellulaire la plus importante reste en lumière blanche. Alors qu’à faible irradiance, la

concentration cellulaire la plus importante est atteinte sous lumière bleue.

Indépendamment de la lumière, la figure 19 (différents clones d’Haslea sont utilisé A-G)

nous montre que les clones les plus longs (53 – 80µm) présentent les résultats de

concentration de marennine les plus importants.

Cependant, d’autres études menées à MMS et au GEPEA, soutiennent l’hypothèse

du métabolisme secondaire. Malgré un constat évident de la capacité d’Haslea ostrearia à

synthétiser de la marennine (même en très faible quantité) tout au long de sa croissance,

Neuville et Daste (1972), ont montré que la carence azotée avait un impact sur la

production d’EMn, confirmée par Rossignol, (1999) et Lebeau, (2000) (Fig. 21).

Page 56: Par « Rémi NGHIEM-XUAN

30

Figure 21: Cinétique de production de marennine et absorption de nitrate par des cellules d'Haslea ostrearia immobilisées dans le gel d’agar (Lebeau et al., 2000)

La figure 19, montre la production de marennine (⦁) et l’absorption de nitrate (o)

au court de la culture d’Haslea ostrearia dans le prototype de PBR à cellules immobilisées

de Lebeau (Présenté Chapitre I.5.4.2.2.1). La culture est en continu avec un taux de

renouvellement de milieux de culture de 0,025 J-1. Son travail ne présente pas d’autres

mesures de macroéléments nous permettant de confirmer avec certitude que seuls les

nitrates impactent la production de marennine. Cependant, ces résultats de corrélation

remettent en question l’hypothèse du métabolite primaire.

9 Valorisation de la biomasse

9.1 Les pigments

9.1.1 La marennine intracellulaire

La production de marennine intracellulaire pour valorisation est actuellement peu

envisagée, considérant les productivités plus importantes apportées par les technologies

de milking (Rossignol, 1999). La valorisation de l’IMn reste cependant une possibilité

après cassage cellulaire.

9.1.2 La fucoxanthine

La fucoxanthine est un pigment d’intérêt qui donne aux micro et macroalgues une

couleur brune. La composition en fucoxanthine d’Haslea ostrearia n’est pas connue mais

les diatomées, les dinophycées et les phéophycées possèdent en moyenne 20,4% (w/w)

du pigment. La fucoxanthine est un pigment de la famille des caroténoïdes

majoritairement répandu chez les macroalgues et les diatomées. Ce pigment, lié à la

Page 57: Par « Rémi NGHIEM-XUAN

31

chlorophylle a, c et aux apoprotéines forme le meilleur complexe photo absorbeur qui

transfère l’énergie lumineuse à la chlorophylle a pour la réaction photosynthétique. La

fucoxanthine possède des liaisons alléniques, un groupe carbonyle, un 5,6-monoepoxide

et un groupe acétyle qui définissent ses activités sur la physiologie humaine. Ce

caroténoïde montre une forte activité antioxydante, anti-inflammatoire, anti-obésité

(Abidov et al., 2010), antidiabétique (réduction du taux de glucose dans le sang),

anticancéreux et anti-hypertensive (Woo et al., 2010).

Le marché de la fucoxanthine est considéré comme un marché de niche en

croissance avec une valeur de la molécule estimé à 3000€/kg. Principalement pour la

consommation animale en aquaculture, elle est aujourd’hui en vente en gélule comme

complément alimentaire pour la nutrition humaine. La fucoxanthine peut être extraite à

partir de macroalgue comme Laminaria japonica, Eisenia bicyclis, et Undaria

pinnatifidacar, sur un marché très verrouillé par de nombreux brevets, car ces algues sont

déjà consommées en Asie du sud-est et en Europe. Cependant depuis 2005, certaines

entreprises comme Setalg, Algosud ou encore Nature-Algue commercialisent la

microalgue Odontella aurita en gélule pour la consommation humaine.

9.2 Les lipides

La composition lipidique d’Haslea représente un atout potentiel pour sa valorisation

économique. En phase épipélagique, Haslea semble produire des globules lipidiques

intracellulaires qui font remonter les cellules à la surface par densité. La composition

chimique des lipides produits est peu connue, mais le pourcentage produit en masse sèche

se rapproche des résultats obtenus sur des souches étudiées au GEPEA (Table. 5).

Tableau 4: Comparaison de la biomasse et la concentration lipidique produite par différentes microalgues

Biomasse (g/L) Lipides (% MS)

Chlorella sp.6 1 - 2 28 - 32

Chlamydomonas reinhardii7

1 – 2,5 19 - 21

Haslea sp.8 3,0 x 10-2 – 3,6 x 10-2 30,6 - 35,8

6 Taleb et al., 2015 ; Liang et al., 2009 7 Lemasson C. 2015 8 Griffiths et al., 2009

Page 58: Par « Rémi NGHIEM-XUAN

32

Les rendements lipidiques chez Haslea sp., sont élevés comparés à d’autres cellules

modèles du laboratoire. Cependant, les rendements de biomasses sont encore très peu

élevés et maîtrisés pour cette diatomée. D’autres études démontrent que Haslea ostrearia

contient principalement des mono- et digalactosyldiacylglycérol (MGDG and DGDG,

respectivement) membranaires sous forme C18/C16 and C18/C18 (Dodson et al., 2013).

Ces acides gras sont d'importants lipides membranaires, où ils remplacent les

phospholipides pour la préservation du phosphate, alors disponibles pour d'autres

processus biologiques essentiels (Peter et Benning 2002). Chez les plantes supérieures,

ses galactolipides (MGDG et DGDG) sont responsables de la régulation des oxydes

nitriques (composés considérés comme DRO) et ont donc été liés à des activités anti-

inflammatoire, diététique et antitumoral chez l’être humain (Gao et al., 2014).

10 Caractérisation de la fragilité cellulaire

L’agitation étant un paramètre de culture identifié comme important pour la

culture d’Haslea ostrearia (Robert, 1989 ; Rossignol et al., 1999 ; Vandanjon et al., 1999),

la conception et l’optimisation d’un système de culture nécessite la compréhension des

bases de la dynamique des fluides. Les phénomènes de cisaillement dûs à une déformation

d’un fluide sont donc détaillés dans cette partie.

10.1 Notions de cisaillement

10.1.1 La déformation d’un fluide : l’écoulement laminaire et turbulent

La déformation d’un fluide peut être observée par l’ajout de marqueurs (fumée

pour les gaz, colorant pour les liquides) permettant la caractérisation de l’écoulement

d’un fluide. Il est constaté que lorsque le marqueur diffuse très lentement au sein du

système, différentes couches (appelée lamelles) glissent les unes par rapport aux autres

sans se mélanger, formant un gradient de vitesse de vecteur parallèle à un instant t

l’écoulement est dit laminaire. Si tous les vecteurs vitesses sont à la fois parallèles et

égaux, l'écoulement laminaire est uniforme. A contrario, si le marqueur est « perturbé de

manière aléatoire et irrégulière dans toutes les directions » : l’écoulement est dit

turbulent. Les vecteurs de vitesse sont inégaux (différents en direction, sens, intensité).

Des tourbillons (vortex) se forment (Fig. 22).

Page 59: Par « Rémi NGHIEM-XUAN

33

Figure 22: écoulement autour d’un cylindre, (Gauche : laminaire Rex< 5×105), (Droit : turbulent Rex> 5×105)

Le passage d’un régime laminaire à un régime turbulent est régi par la vitesse de

déformation, des caractéristiques visqueuses du fluide ainsi que de la forme de

l’écoulement (espace fermé, canalisation, espace ouvert, etc.…). Ainsi, pour un écoulement

turbulent, les grandeurs caractérisant cet écoulement varient de manière aléatoire. La

notion même d’écoulement turbulent permanent ne peut être comprise qu’en moyenne.

A l’inverse, les fluctuations des gradeurs caractérisant un écoulement laminaire sont

négligeables, voire nulles pour les régimes laminaires parfaits.

La distinction entre les différents régimes s’effectue grâce au nombre de Reynolds :

Équation 2

𝑅𝑒 =𝑉𝐿

𝑣

V est la vitesse du (m.s-1), L la longueur spécifique du système (m) et 𝑣 la viscosité

cinématique. La transition d’un régime laminaire à un régime turbulent se fait

graduellement sur une plage critique de nombre de Reynolds, dont les valeurs peuvent

dépendre des systèmes étudiés. Les écoulements laminaires ont lieu à faible Reynolds

(Rex<2000 pour les cas les plus fréquents) tandis que les forts nombres de Reynolds

correspondront aux écoulements turbulents (Rex> 2000 pour les cas les plus fréquents)

(Bergman 2011).

10.1.2 Les fluides newtoniens et la viscosité

Lors de l’écoulement d’un fluide des forces tangentielles de frottement à cet

écoulement s’opposent au mouvement du fluide. Ces forces sont présentes sur les

extrémités non mobiles d’un système (les parois d’une canalisation par exemple) mais

sont aussi dues aux interactions entre les molécules du fluide, appelées forces de viscosité

intrinsèques aux fluides.

Page 60: Par « Rémi NGHIEM-XUAN

34

Plusieurs grandeurs physiques caractérisent la viscosité : la viscosité dynamique

(celle utilisée le plus généralement), la viscosité cinématique, la seconde viscosité et la

viscosité de volume. Nous ne nous intéresserons qu’aux deux premières (Nic et al., 2014) :

- La viscosité dynamique µ, est la grandeur de référence quand on parle de viscosité

sans autre précision. Elle se mesure en Pascal-seconde (Pa. s-1) et traduit le lien

entre la contrainte de cisaillement et le gradient transversal de la vitesse

d'écoulement dans le fluide « dynamique » :

Équation 3

µ =𝜏

γ

τ étant la contrainte de cisaillement (Pa), et γ le gradient de vitesse aussi appelé

taux de cisaillement (s-1).

- La viscosité cinématique ν, s’exprime en m2. s-1 et se déduit en divisant la viscosité

dynamique par la masse volumique ρ soit :

Équation 4

𝑣 =µ

ρ

Cependant, ces équations (3 et 4) ne sont valables que pour les fluides dits

newtoniens. Pour ses fluides, µ est une constante intrinsèque dépendant uniquement de

la température et de la pression. Dans le cas des fluides non newtoniens, la viscosité est

un paramètre dépendant de la température, de la pression mais aussi des forces en action

sur lui-même, faisant varier la rhéologie du fluide en fonction des contraintes. En d’autres

termes, le fluide newtonien continu de s’écouler indépendamment des forces extérieures

qui agissent sur lui (l’eau est un fluide newtonien parce qu’il conserve sa viscosité quelle

que soit la vitesse à laquelle il est agité).

10.2 Les contraintes de cisaillement en photobioréacteur

Le photobioréacteur représente le système le plus optimisé pour la culture de

microalgues, l’analyse des réactions biochimiques et la maîtrise des conditions axéniques.

Ce procédé optimisé implique plusieurs éléments tels que l’agitation, la recirculation,

l’aération, le pompage, tous dans le but de faciliter et d’optimiser le transfert de matière,

de chaleur autant que l’accès à la lumière au sein du photobioréacteur. Ces opérations

Page 61: Par « Rémi NGHIEM-XUAN

35

peuvent être appliquées à plusieurs intensités, entraînant pour le mélange des

contraintes de cisaillements pouvant être considérées comme excessives et causer une

mortalité cellulaire ou une baisse du taux de croissance (Wang & Lan, 2018).

10.2.1 Technologie Airlift

L’ajout de CO2 est essentiel à la culture des microalgues, en tant que source

inorganique de carbone pour la croissance. Il est nécessaire également d’aérer la culture

pour retirer l’excès de O2 dissous produit par cette même croissance (Ugwu et al., 2008).

Cette pression d’O2 peut devenir un problème pour la production de biomasse, faisant

rentrer la microalgue en photo-respiration (présenté Chapitre I.1.5) et ainsi diminuant la

productivité (Kazbar 2018 ; Pruvost et al., 2017 sur la technologie Algofilm ©). La

technologie Airlift (aussi appelé réacteur gaz/liquide) est une technologie d’agitation de

bioréacteur permettant le mélange, la régulation pH et l’injection de CO2, par la seule

injection de gaz.

Figure 23: Représentation de la circulation de la phase liquide en photobioréacteur Airlift 1L

Comme représenté sur la figure 23, l’injection de gaz située au bas du PBR, entraîne

une recirculation du milieu de culture autour des chicanes. Selon les travaux de Olmos et

al., (2003) et Ndiaye et al., (2018), effectués sur colonne à bulle et Airlift GEPEA, la

technologie Airlift permet l’application de trois types de régimes d’écoulement gazeux :

- Le régime homogène, apparaît lorsque de fines bulles (à faible distribution de

taille) progressent verticalement avec de faibles oscillations latérales sans

Page 62: Par « Rémi NGHIEM-XUAN

36

interaction entre elles. Cet écoulement intervient à faible vitesse superficielle de

gaz (Deckwer, 1991) et considère la rupture ou coalescence des bulles négligeable.

- Le régime de transition, apparaît lors de l’augmentation de la vitesse superficielle

augmentant les phénomènes de coalescences et rupture de bulles et ainsi la

distribution de taille des bulles.

- Le régime hétérogène quant à lui, intervient à haute vitesse superficielle de gaz,

lorsque l’écoulement devient instable et la distribution de taille de bulles très large.

Une distribution de gaz centré entraîne ce type de régime même à faible débit de

gaz.

Les performances des PBRs Airlift sont liées à ces types de régime et plus

particulièrement à la taille de bulles distribuées au sein du PBR, conditionnant le transfert

de matière par leurs aires interfaciales (Deckwer, 1991 ; Ruzicka et al., 2001). Selon

Deckwer (1991), l’importante agitation et l’augmentation du coefficient de transfert de

matière « kL » apporté par un régime hétérogène ne compense pas la diminution de l’aire

interfaciale. Il est donc conseillé de travailler en régime homogène pour optimiser le

coefficient de transfert de gaz-liquide « kLa ».

Les contraintes de cisaillements apportées par le débit de gaz en régime homogène

se retrouvent plus importantes au sein du trajet de gaz où le régime du fluide est

turbulent. L’étude en colonne à bulles de Contreras et al. (1998) permet d’approximer la

vitesse superficielle de gaz maximum du régime homogène, estimé à UG = 0.03 m. s-1 en

Airlift sur une culture de Phaedactylum tricornutum et obtient ses meilleurs rendements

de croissance cellulaire en régime transitoire à UG = 0.055 m. s-1.

Le GEPEA confirme ces résultats grâce à la mécanique des fluides numériques

(MFN) (Ansys-Fluent©14.5) Selon la Figure 24, la vitesse superficielle dans l’axe

d’injection du CO2 peut monter jusqu’à UGMAX=0,180 m. s-1, mais permet une moyenne de

UGMOY=0,05 m. s-1 au sein du PBR Airlift.

Page 63: Par « Rémi NGHIEM-XUAN

37

Figure 24: Simulation MFN du champ de vitesse de la phase liquide, coloré par intensité de vitesse superficielle (m. s-1). (a) représente une Airlift 1L, (b) représente un Airlift 100L (Ndiaye et al., 2018)

Du fait du passage des bulles, le taux de cisaillement apporté par cette vitesse

superficielle de gaz est relié par l’équation suivante (Chisti et Murray Moo-Young, 1989) :

Équation 5

�́� = 5000𝑈𝐺

Où UG la vitesse superficielle de gaz (m. s-1) et γ́ taux de cisaillement (s-1).

Le taux de cisaillement maximum moyen apporté par la turbulence de l’agitation

gazeuse est estimé entre 150 s-1 et 275 s-1 au sein du trajet gazeux dans une PBR Airlift.

Cependant, l’éclatement et l’apparition des bulles respectivement en haut et en bas du

PBR favorise l’apparition de micro vortex pouvant produire jusqu’à 14 000 s-1 de taux de

cisaillement moyen (Contreras et al., 1998 ; Michiel et al., 2010 ; Barbosa et al., 2003).

Malgré l’intérêt évident d’une utilisation de fines bulles pour le transfert de matière, la

multiplication de points d’entrée de gaz additionnée à de fines bulles peut être négative

pour la culture de cellules sensibles aux contraintes de cisaillement (Barbosa et al.,

2003 ; Sobczuk et al., 2006). La multiplication des points d’entrées augmente les

cisaillements dus à la génération des bulles, la production de micro vortex et la fréquence

d’explosion des fines bulles en haut du PBR.

Page 64: Par « Rémi NGHIEM-XUAN

38

10.2.2 Technologie Torique

La géométrie particulière de ce photobioréacteur de laboratoire a été mise au point

par le GEPEA dans le but d’étudier la réponse biologique des microalgues face aux

conditions environnementales appliquées et contrôlées. La géométrie du PBR est un

élément important car elle conditionne le comportement de l’algue au sein du PBR. Le

PBR torique est basé sur la forme « torique » permettant un mélange parfait (limitant les

zones mortes). Cette agitation permet un accès amélioré du micro-organisme à la lumière

en générant un mouvement dans le sens du gradient de lumière intra-PBR (présenté

Chapitre I.5.1.1). D’autre part, le procédé est entièrement automatisé pour le contrôle des

conditions environnementales appliquées (Fig. 25).

Figure 25: schéma du pilote laboratoire Torique de production contrôlée de microalgues (Artu, 2016)

L’injection de gaz en point bas du PBR permet une régulation du pH par ajout de

CO2 mais aussi l’agitation du milieu de culture dans le sens anti-trigonométrique (sens

horaire). Le mouvement principal est toutefois obtenu par l’agitation produite par la/les

hélice(s) marine(s) (Fig. 26 Droite). La forme torique (Fig. 26 Gauche) permet alors une

agitation efficace sans zones non agitées favorisant l’apparition de cellules stagnantes et

de biofilms (Pruvost et al., 2004).

Page 65: Par « Rémi NGHIEM-XUAN

39

Figure 26: Représentation de la géométrie (Gauche) du PBR torique avec ses dimensions, (Droite) de 2 types d'hélice utilisé pour l'agitation du PBR torique (Pruvost et al., 2006)

Tout comme la technologie Airlift, le champ de contraintes de cisaillement réparti

au sein du PBR est complexe et ne peut être réduit qu’à un chiffre moyen. Malgré une

bonne agitation sans zones mortes, les contraintes de cisaillement sont hétérogènes dans

l’ensemble du volume du PBR torique. L’agitation par hélice marine, engendre des

turbulences ainsi que des contraintes plus importantes que dans le reste du PBR, de même

que le point d’injection de gaz (malgré un faible débit d’injection maîtrisé par débitmètre

massique) génère des cisaillements ciblés plus importants. Cependant, la configuration de

la géométrie en boucle de ce PBR permet deux avantages. Tout d’abord la recirculation en

boucle entraîne toutes les cellules vers l’hélice qui ainsi subissent le même historique de

cisaillement au sein de la culture. D’autres part, une faible rotation de l’hélice (donc de

faibles contraintes de cisaillements) couplée à l’injection de gaz suffit à obtenir une bonne

efficacité de mélange (Pruvost et al., 2006).

Page 66: Par « Rémi NGHIEM-XUAN

40

Figure 27: Prédiction de la vitesse moyenne du fluide en PBR torique UO en fonction de la vitesse de rotation des 2 types d'hélices (Fig. 25) utilisées en Torique (Pruvost et al., 2006)

La corrélation entre la vitesse moyenne du fluide et la vitesse de rotation de l’hélice (Fig.

27) est comparée aux résultats de Khalid et Legrand (2001) effectués en boucle à section

circulaire. La vitesse moyenne obtenue par la rotation de l’hélice est un peu plus faible

que celle calculée par Khalid et Legrand, due à la section carrée de la géométrie du PBR

réduisant le débit induit de 10-20% des résultats. Comme présenté précédemment, de

fortes vitesses de rotations d’hélice ne sont pas nécessaires pour éviter la sédimentation

chez les microalgues pélagiques (par exemple, la culture de Chlamydomonas reinhardtii

en PBR Torique s’effectue à U0=0,05 m. s-1).

Figure 28: Prédiction des valeurs globales (VY), ε et εimp en fonction de la vitesse moyenne du flux en PBR Torique (Pruvost et al., 2006)

Page 67: Par « Rémi NGHIEM-XUAN

41

La Figure 28, montre l’évolution de la vitesse moyenne en direction de

l’atténuation de lumière au sein du PBR (VY), ainsi que le taux de cisaillement (ε imp et ε)

en fonction de la vitesse moyenne induite dans le PBR (UO) par la vitesse de rotation de

l’hélice. Le taux de cisaillement est séparé en deux pour différencier le cisaillement moyen

provoqué par l’hélice (ε imp) du cisaillement moyen en boucle (ε). Une évolution linéaire

des paramètres de cisaillement et de vitesse est observée. Les deux types d’hélice

permettent une redirection des microalgues dans le gradient d’atténuation de lumière,

qui se fait de plus en plus importante avec la vitesse moyenne engendrée par l’hélice. De

même, le taux de cisaillement augmente proportionnellement à la vitesse moyenne

induite par l’hélice, pouvant atteindre 90 s-1 en moyenne, 30 s-1 dans la boucle et 140 s-1

dans la zone de l’hélice (hélice 1) pour UO=0,4 m. s-1. Les résultats sont légèrement plus

faibles pour l’hélice 2 avec presque 70 s-1 moyen pour 30 s-1 en boucle et 110 s-1 dans la

zone de l’hélice.

10.3 L’impact des cisaillements sur les organismes microscopiques

unicellulaire libre et en tissus

10.3.1 La sensibilité des microalgues/cyanobactéries/tissus d’organismes

pluricellulaires

Malgré une génération de cisaillement au sein du PBR, l’agitation est essentielle

pour la culture de microorganisme non limitée et donc pour la production industrielle.

Cependant, dans la littérature, certains micro-organismes supportent moins bien

l’agitation. La tolérance des micro-organismes face à ces contraintes de cisaillements est

un phénomène complexe dépendant des facteurs physiologiques (benthique, pélagique,

croissance en colonie etc…) autant que des paramètres physiques (tailles, formes,

flagelles, filaments etc…) de chaque micro-organisme. De manière générale,

l’agitation/cisaillement affecte très peu les Chlorophycées mais elle affecte certaines

Cyanobactéries, Haptophytes, algues rouges, beaucoup de Diatomées et surtout les

Dinoflagellés qui sont considérés comme fragiles. Les tissus cellulaires (peau, organes

etc…) sont aussi étudiés dans la littérature pour leur comportement physiologique sous

contraintes. Après la récolte et la comparaison de ces informations (Table. 5), aucune

réelle corrélation n’a été observée entre la fragilité cellulaire, la forme, la taille ou même

la Classe des cellules étudiées face aux contraintes/taux de cisaillements.

Page 68: Par « Rémi NGHIEM-XUAN

42

Tableau 5 : Etude bibliographique de la sensibilité d'organismes cellulaires face aux taux de cisaillement

Type

d’organismes Classe Espèce

Aperçu

microscopique

Système de

culture

Cisaillement

limite

supporté

Références

Algue verte

Chlorodendrophyceae Tetraselmis

suecica

Rhéomètre 75 000 (s-1)

Michiels et al., 2015

Chlorophyceae Chlorella vulgaris

Cuve agitée 0,9 (Pa) Leupold et al., 2013

Chlorophyceae Senedesmus

obliquus

Cuve agitée 0,9 (Pa) Leupold et al., 2013

Chlorophyceae Chlamydomonas

reinhardtii

Cuve agitée 0,2 (Pa) Leupold et al., 2013

Les chiffres présentés proviennent de plusieurs études non liées entre elles et restent subjectifs. Ceci est du aux différences de temps et de techniques d’application des taux de cisaillement, régime d’écoulement des différentes études.

Page 69: Par « Rémi NGHIEM-XUAN

43

Algue rouge Porphyridiophyceae Porphyridium

cruentum

Cuve agitée 2 (Pa)

Cole K. M. & Sheath R. G.,

1990

Sobczuk et al., 2006

Cyanobactérie

Cyanophyceae Arthospira

platensis

Cuve agitée 2,2 (s-1) Mitsuhashi et al., 1995

Cyanophyceae Aphanizomenon

flos-aquae

Cuve agitée 0,45 (Pa)

Leupold et al., 2013

Bronnenmeier et Märkl,

1982

Synechocystis sp.

Barreau

aimanté 0,114 (Pa) Fadlallah et al., 1995

Diatomée Bacillariophyceae Phaeodactylum

tricornutum

Cuve agitée 7 000 (s-1) Contreras et al., 1998

Page 70: Par « Rémi NGHIEM-XUAN

44

Coscinodiscophyceae Skeletonema

costatum

Rhéomètre 5 000 (s-1) Michiels et al., 2015

Mediophyceae Chaetoceros

muelleri

Rhéomètre 930 (s-1) Michiels et al., 2010

Haptophytes Prymnesiophyceae Isochrysis

galbana

Rhéomètre 5 000 (s-1) Michiels et al., 2015

Dinoflagellé

Dinophyceae Protoceratium

reticulatum

Verrerie

agitée

0,12 (Pa)

0,16 (mPa)

Garcia et al., 2007

Camacho et al., 2007

Dinophyceae Lingulodinium

polyedrum

Rhéomètre 0,004 (Pa) Juhl et al., 2000

Dassow et Latz 2002

Dinophyceae Pyrocystis lunula

Microscope

à force

atomique

Light production

at

36 (µPa)

Tesson et Latz 2015

Page 71: Par « Rémi NGHIEM-XUAN

45

Système

immunitaire

vertébré

Mammalia Lymphocyte T

humain

10 – 20 (Pa) Chittur et al., 1988

Système

sanguin Mammalia

Erythrocyte

humain

Rhéomètre ≥ 2 000 (s-1) Mauer et al., 2016

Tissu

épithélial

ovarien

Insecta

Cellules Sf9

Spodoptera

frugiperda

Verrerie

agitée 0,084 (Pa) Connor et al., 2002

Page 72: Par « Rémi NGHIEM-XUAN

46

Après étude du tableau 5, il apparait que le cisaillement n’est pas qu’une fonction

de facteurs physiques mais aussi de facteurs biologiques, rendant la comparaison sur

cellules vivantes plus compliquée qu’une étude de résistance d’un solide inerte. D’autre

part, chaque analyse de fragilité a été effectuée dans différents systèmes menant à des

cisaillements d’intensité différentes, en type et en durée. Hormis l’importance physique

de l’agitation dans la production en bioréacteur, limiter la fragilité cellulaire à un seul

chiffre de contrainte/taux de cisaillement n’est pas suffisant pour comprendre la

complexité des phénomènes mis en jeu. Son impact sur la physiologie cellulaire, positif ou

négatif, est important à comprendre. Ce phénomène appelé la mécano-transduction ou

mécano-sensibilité regroupe les mécanismes physiologiques cellulaires engendrés par

des stimuli mécaniques, transformant ces signaux mécaniques en signaux électriques ou

chimiques (Griffin et al., 1992 ; Tavernarakis & Driscoll 1997). L’un de ces phénomènes le

plus impressionnant s’illustre sur la baie Laguna Grande à Porto Rico (Fig. 29) abritant un

consortium planctonique de dinoflagellés (Lingulodinium polyedrum, Pyrocystis lunula)

bioluminescentes sous stimuli mécaniques (Tesson et al., 2015 ; Dassow et Latz 2002).

Figure 29: (Gauche) plage de la baie Laguna Grande de nuit, Porto Rico et observation de bioluminescence sur les vagues ; (droite) isolement du consortium bioluminescent en laboratoire

Parmi ce consortium se démarque en concentration Pyrocystis lunula (Fig.30),

capable de synthétiser luciférine/luciférase, un complexe protéique bioluminescent dans

une stratégie de protection contre les prédateurs (Esaias & Curl 1972 ; Buskey & Swift

1985). Selon Tesson et Latz (2015), il suffit d’une contrainte de 36 µPa pour que cette

cellule irradie, une stimulation inférieure à 1.8 µPa ne déclenchant pas la synthèse

protéique bioluminescente. Ainsi, cela prouve l’importance des cisaillements et de leur

intensité dans la réponse des microorganismes.

Page 73: Par « Rémi NGHIEM-XUAN

47

Figure 30: Observation en microscopie optique de la dinoflagellé Pyrocystis lunula, (Gauche) sans agitation sous lumière ; (Droite) après agitation sans lumière

Un autre exemple de l’importance physiologique de la sensibilité aux contraintes

de cisaillement est la culture de tissus d’organismes pluricellulaires. La lignée Sf9 de

cellules ovariennes de Spodoptera frugiperda est une lignée de cellules se cultivant,

immobilisées en tissu mais aussi, libres en verrerie agitée (Fig. 31).

Figure 31: Observation en microscopie optique de la lignée cellulaire Sf9 de tissu ovarien de Spodoptera frugiperda, (Gauche) culture adhérente en tissu ; (Droite) culture libre agitée

Les cellules supportent de très faibles contraintes de cisaillement (0,084 Pa) mais

la culture en cellule libre reste possible. L’agitation d’une de ces cultures en tissu

provoque immédiatement la mort cellulaire, même à très faible agitation. Ainsi, le

problème de la sensibilité de cette souche aux contraintes n’est pas tant le fait des

cisaillements que celui de sa physiologie (Akhnoukh et al., 1993 ; Connor et al., 2002).

La structure cellulaire joue un rôle dans la fragilité aux contraintes mécaniques.

Selon les travaux de Barbosa et al. (2004), le mutant de Chlamydomonas reinhardtii

CC1883 cell wall mutant 15, ne possède pas de paroi cellulaire rigide contrairement à son

équivalent souche sauvage. Soumis aux mêmes contraintes liées au bullage en colonne, le

mutant CC1883 montre un taux de mortalité cellulaire de 1,01 ± 0,29 h-1 pour la même

Page 74: Par « Rémi NGHIEM-XUAN

48

vitesse superficielle du gaz mesuré à 0,085 m. s-1. Ainsi la constitution même de la

membrane cellulaire et le type de parois peuvent fortement altérer la réponse de la cellule

face aux contraintes de cisaillements. La majorité des algues vertes possèdent des parois

riches en cellulose, les cyanobactéries possèdent du peptidoglycane permettant une

certaine rigidité alors que les diatomées possèdent le frustule de silice amorphe formant

une protection de verre rigide. Cependant, le frustule ne rend pas les diatomées forcément

plus résistantes comme démontre Sobczuk et al., (2006). De plus, contrairement aux

autres formes de parois cellulaires, le frustule nécessite une étape d’ouverture lors de la

division asexuée et sexuée de la cellule, rendant potentiellement cette cellule sensible lors

de la division. Ainsi, une paroi rigide ne peut assurer à elle seule une grande résistance

aux contraintes de cisaillement.

Quelques études (Leupold et al., 2013 ; Fadlallah et al., 2016) montrent que le

nombre de couches dans la paroi cellulaire peut potentiellement permettre à la cellule de

résister aux contraintes, comme Scenedesmus obliquus et Synechocystis sp. qui compte

parmi les microalgues multi lamellaires. Cependant, Arthrospira platensis compte aussi

parmi ces organismes multi lamellaires sans pour autant être capable de résister à plus

de 2,2 s-1 (Mitsuhashi et al., 1995 ; Leupold et al., 2013). L’épaisseur des membranes

cellulaires semble elle aussi, peu responsable de la résistance de la cellule puisque que

Chlamydomonas reinhardtti qui possède une paroi de 200 nm est moins résistante que

Chlorella vulgaris avec une paroi de 17-21 nm (Wang et Lan, 2018).

La taille de la cellule est un élément sur lequel il est presque possible de corréler la

sensibilité à la contrainte. En effet, même dans un régime laminaire non parfait, il peut

survenir des micros vortex liés à l’imperfection des parois du système, la présence de

particules ou encore la viscosité du milieu liée à la croissance du microorganisme. Plus les

cellules sont petites et plus elles ont peu de risque d’être impactées par ses micro

perturbations hydrodynamiques. Cependant, les cellules plus grosses ou plus larges telles

que les Dinoflagellés, Diatomées ou Haptophytes sont plus enclines à souffrir de ce

phénomène. Ces microalgues font en moyenne, partie des plus sensibles à l’agitation.

Concernant la diversité de physiologie de croissance des microorganismes, elle est très

importante (libre, filaments, colonies etc…) et ce paramètre semble être déterminant dans

la fragilité cellulaire. De hautes contraintes de cisaillement entraînent la séparation des

filaments tels que ceux d’Arthrospira platensis ou encore Aphanizomenon flos-aquae

Page 75: Par « Rémi NGHIEM-XUAN

49

(Bronnenmeier et Märkl, 1982 ; Leupold et al., 2013). Cependant, une fois de plus, certains

microorganismes tels que Skeletonema costatum poussent en chaînettes et se montrent

2000 fois plus résistants qu’Arthrospira platensis, 5 fois plus que Chaetoceros muelleri et

aussi résistants qu’une cellule isolée d’Isochrysis galbana. Ou encore, Tetraselmis suecica

possède la même forme et la même taille que Chaetoceros muelleri, pourtant ce dernier

est 80 fois moins résistant que Tetraselmis suecica (Michiels et al., 2010 ; Michiels et al.,

2015).

Certaines microalgues sont motiles et possèdent des flagelles. Ces algues sont

souvent présentées comme plus sensibles aux contraintes de cisaillement et la

destruction de ce flagelle entraîne, le plus souvent, la mort de la microalgue. La majorité

des algues vertes ne possèdent pas de flagelle, mais Chlamydomonas reinhardtii est une

exception qui se montre plus sensible aux cisaillements (0.2 Pa) que beaucoup d’algues

vertes. Cependant, Tetraselmis suecica possède aussi jusqu’à 4 flagelles et se montre

particulièrement résistant (75 000 s-1 soit presque 75 Pa dans une viscosité proche de

celle de l’eau) malgré une perte de mobilité (Jaouen, 1988). Les dinoflagellés, organismes

majoritairement flagellés sont aussi considérés comme les plus sensibles des microalgues.

Malgré quelques exceptions telles que Tetraselmis suecica, les flagellés sont des

organismes majoritairement plus sensibles à l’agitation que les non flagellés, ainsi la

présence de flagelles semble être un élément déterminant dans la sensibilité à l’agitation.

Le métabolisme et les cycles de croissances sont aussi des facteurs qui peuvent

jouer un rôle dans le phénomène de sensibilité cellulaire. Beaucoup de microalgues sont

plus sensibles aux cisaillements lors de leurs divisions cellulaires, expliqué par le fait que

les cellules possèdent une taille plus importante et ces parois/membranes plus fines

(pendant un court temps) lors de la division cellulaire, comme Tetraselmis suecica,

Chaetoceros muelleri, Isochrysis galbana et Skeletonema costatum présenté par le travail

de Michels et al., (2016). L’exemple du système de division cellulaire des diatomées

présenté plus haut illustre aussi l’impact que pourrait avoir le cycle de division sur la

sensibilité aux contraintes. Cependant, contrairement à la relation du cycle de division

avec les cisaillements, le constat de l’impact du métabolisme sur la sensibilité aux

cisaillements est aussi important que celui de la sensibilité des cisaillements sur le

métabolisme. L’agitation mécanique peut engendrer la détérioration cellulaire ou

l’interruption de croissance (cas de la majorité des cellules présentées tableau 6). De

même que le tissu cellulaire de lignée Sf9 peut altérer /adapter son métabolisme pour

Page 76: Par « Rémi NGHIEM-XUAN

50

supporter une faible agitation et permettre une croissance plus rapide des cellules libres

agitées (Connor et al., 2002). Pour finir, Les lymphocytes T et érythrocyte humains,

montrent une capacité d’adaptation très importante face à l’altération de l’intensité des

contraintes de cisaillements et la viscosité du fluide en canalisation (représentation d’un

conduit sanguin), faisant varier leurs formes pour supporter les variations (Chittur et al.,

1988 ; Mauer et al., 2016).

10.3.2 La caractérisation de la fragilité d’Haslea ostrearia

Comme la majorité des diatomées, Haslea possède une coque de silice amorphe

appelée frustule. Cette coque apporte une protection aux diatomées mais permet aussi le

maintien de la membrane cellulaire (Crawford et al., 2001). Longtemps considérées

comme fragiles, les premières hypothèses de cette fragilité se sont reposées sur sa

structure de verre ainsi que de sa forme pennée. Cependant, la caractérisation de la

fragilité d’une souche ne dépend pas seulement des paramètres physiques de la structure

mais aussi de sa physiologie. De plus, il existe une grande diversité de systèmes de culture

qui possèdent chacun selon leurs formes, une diversité de mélange hydrodynamique et

donc de cisaillement. L’association entre les résultats de fragilité cellulaire et l’agitation

du réacteur est complexe. Les travaux de Rossignol (1999) ont permis de cerner les

limites physiques d’Haslea ostrearia face à différents systèmes de transferts et de

destruction cellulaire.

10.3.2.1 Passage dans une pompe péristaltique

Les premiers tests de Rossignol (1999), ont été faits dans des systèmes de transfert

de type pompes péristaltiques reconnus pour générer le moins de cisaillements possibles

(Le Borgne 2011). La viabilité cellulaire d’Haslea ostrearia a été mesurée après le passage

dans une pompe péristaltique à 3 temps de passages différents (Fig. 32).

Page 77: Par « Rémi NGHIEM-XUAN

51

Figure 32: Cinétique de croissance de culture après traitement en pompe péristaltique (Q=25 L. h-1) selon le

nombre de passage (étoile) témoin, + (croix) 10 passages, o (rond) 100 passages, ◻ (carré) 250 passages (Rossignol, 1999)

Haslea ostrearia montre une faible sensibilité à ce type de pompe après

250 passages au moins. Cependant, la sensibilité est moindre que chez les autres types de

systèmes testés par Rossignol. Aucune différence de taux de croissance après la remise en

culture des échantillons traités par pompes péristaltiques n’a cependant été constatée

mais les productivités baissent légèrement. On constate aussi, que même si la sensibilité

à 250 passages est faible, le nombre de passages est lui, aussi relativement faible, comparé

à l’utilisation d’une pompe de recirculation dans un PBR en continu où ces différents

passages peuvent durer sur plus d’un mois.

10.3.2.2 Impact des ultrasons

Les travaux de Rossignol se sont aussi concentrés sur l’étude de la sensibilité

d’Haslea aux ultrasons de manière à comprendre sa fragilité et envisager différentes

techniques de bioraffinage. Différentes expériences ont été réalisées permettant de

comprendre l’impact de la sonication sur les cellules d’Haslea (Fig. 33).

Page 78: Par « Rémi NGHIEM-XUAN

52

Figure 33: Influence de la puissance et de la durée de sonication sur le cassage des cellules d'Haslea ostrearia, V=300ml, □ (carrée) 550W, △ (triangle) 400W, ○ (rond) 300W (Rossignol 1999)

Selon ces résultats, la sonication à un effet non immédiat mais létal sur Haslea, avec

100% de mortalité apparaissant presque 2 minutes après le début de traitement à plus

haute intensité (550W). Cependant l’impact du volume de l’échantillon et donc de la

propagation des ultrasons dans le milieu de culture n’est pas négligeable (Fig. 34).

Figure 34: Influence du volume de suspension traité sur le cassage cellulaire d'Haslea ostrearia, P=550W, ◼

(Carré) 90s, ◻ (blanc) 60s (Rossignol 1999)

Cette méthode de cassage cellulaire ne semble pas détruire immédiatement les

cellules mais il est observé, par microscopie électronique à balayage (Fig. 35) que les deux

moitiés du frustule sont déboîtées, les parois sont altérées et visiblement percées par

endroit modifiant la membrane et permettant la libération du contenu cellulaire.

Cependant, les apex ne semblent pas altérés.

Page 79: Par « Rémi NGHIEM-XUAN

53

Figure 35: Observation de cellules d'Haslea ostrearia après traitement aux ultrasons par microscopie électronique à balayage : t=40s, P=400W, V=200mL (document LBM - ISOMer)

La puissance de sonication permet une meilleure efficacité de cassage de cellules

alors que le volume de suspension agit de manière inversement proportionnelle sur

l’efficacité de cassage.

10.3.2.3 Désintégration haute pression

Tout comme pour les travaux de sonication, de manière à pouvoir récupérer la

marennine intracellulaire, le système de broyage à haute pression a été utilisé pour

évaluer la résistance d’Haslea ostrearia à différentes pressions de 30 à 270 MPa (1 cycle)

(Fig. 36).

Figure 36: Influence de la pression du système de broyage haute pression sur le cassage cellulaire et la libération de marennine dans le milieu (1 cycle), ♦ (losange) dommages cellulaires, + (croix) diamètre moyen

des particules, ● (rond) marennine libérer (Rossignol 1999)

30 MPa permettent le broyage d’une partie des cellules mais c’est à partir de

100 MPa (1 cycle) que la totalité des cellules est broyée et que l’on obtient une

récupération optimale de marennine intracellulaire. Le rapport direct du cassage

Page 80: Par « Rémi NGHIEM-XUAN

54

cellulaire et de la dimension cellulaire est inversement proportionnel, dépendant de la

pression et du nombre de cycles (Fig. 37).

Figure 37: Influence du nombre de cycle du système de broyage haute pression sur la répartition de la taille des particules par granulométrie laser, (carré plein) culture initiale ; (carré vide) 30 MPa, 1 cycle ; (étoile)

30 MPa, 5 cycles ; o (rond) 150 MPa, 1 cycle. (Rossignol 1999)

L’analyse microscopique, quant à elle, montre la destruction du frustule après

traitement (Fig. 38), à la plus faible pression du système, allant jusqu’à détruire les

différentes couches du frustule.

Figure 38: Observation d'Haslea ostrearia par microscopie électronique à balayage (a) avant traitement, (b) après traitement (1 cycle 30 MPa) (Rossignol 1999)

Les expériences ont montré qu’Haslea ostrearia ne résiste pas à 30 MPa, la plus

faible pression du broyeur haute pression, ce qui est due selon Rossignol (1999) à sa taille,

sa forme pennée ainsi qu’aux caractéristiques physiques de son frustule.

Page 81: Par « Rémi NGHIEM-XUAN

55

10.3.2.4 Pompe à engrenages

La dernière technique utilisée par Rossignol (1999), pour le bio raffinage d’Haslea

ostrearia est le système de pompe à engrenages connu pour son fort potentiel de

cisaillement (Le Borgne 2011). Après observation au microscope électronique à balayage

(Fig.39), l’impact négatif de la pompe à engrenage est constaté sur la viabilité cellulaire

d’Haslea. L’étude de répartition granulométrique (Fig. 40) confirme ces résultats.

Figure 39: Observation de cellules d'Haslea ostrearia au microscope électronique à balayage avant (Gauche) et après passage dans une pompe à engrenage (Droite) (5L, 15min, 350 L. h-1) (document LBM - ISOMer)

En comparant avant et après traitement (par microscopie), les cellules semblent

brisées majoritairement au niveau des pointes ce qui est confirmé par granulométrie avec

un décalage du pic représentant les cellules entières (A1) vers un diamètre plus faible (A2).

Figure 40: Comparaison des répartitions granulométriques d'une culture d'Haslea ostrearia avant ⦁ (rond) et

après 100 passages ◻ (carré) dans la pompe à engrenages (300 L. h-1) (Rossignol 1999)

De même, la surface du pic correspondant à des fragments de cellules augmente

(B1 et B2) La présence de cellules cassées aux extrémités entraîne une répartition

granulométrique étendue des cellules (entre 5 et 15 µm de diamètre moyen).

Page 82: Par « Rémi NGHIEM-XUAN

56

Comme les travaux précédents de Rossignol, l’impact du cisaillement des pompes

dépend du nombre de cycles et de l’intensité du débit (Fig. 41).

Figure 41: Dégradation cellulaire en fonction du nombre de passages et du débit de circulation ◼ (carré) 200 L. h-1, ▲ (triangle) 300 L h-1, ● (rond) 400 L. h-1 (Rossignol 1999)

Selon la figure 41, plus le nombre de passages est important et plus la dégradation

cellulaire augmente, de même à nombre de passages identiques, l’augmentation du débit

de circulation augmente le taux de dégradation cellulaire. Selon Rossignol, cette

augmentation provient de l’augmentation de l’intensité des contraintes de cisaillement dû

à l’augmentation du débit, du changement de régime d’écoulement au sein de la pompe et

de l’augmentation de la fréquence de passages.

10.3.2.5 Centrifugation

Le procédé de centrifugation étant un procédé très utilisé pour la manipulation

laboratoire et préindustrielle d’une souche. Rossignol a cherché l’accélération et la durée

de centrifugation maximale que la souche pouvait supporter de 500 à 50 000 g sur des

durées de 1 à 90 minutes (Fig. 42).

Page 83: Par « Rémi NGHIEM-XUAN

57

Figure 42: Influence de l'accélération centrifuge sur la dégradation cellulaire et la libération de marennine (Rossignol 1999)

Les résultats montrent l’impact non négligeable de l’accélération centrifuge sur

Haslea ostrearia. En effet, au-dessus de 20 000 g presque 50% des cellules sont détruites,

ce qui est confirmé par une augmentation de la marennine relâchée dans le surnageant

par les cellules mortes. L’accélération maximale (50 000 g) engendre plus de 80% de

mortalité cellulaire ce qui valide l’impact négatif de l’accélération sur la souche.

L’impact de la durée de centrifugation sur le cassage cellulaire a été investigué

pour deux accélérations (15 000 g et 30 000 g) (Fig. 43).

Figure 43: Influence de la durée de centrifugation sur la dégradation cellulaire et la libération de marennine (Rossignol, 1999)

À une accélération <20 000 g, aucune dégradation significative est observée,

cependant, à une accélération >20 000 g, l’impact de la durée de centrifugation sur la

Page 84: Par « Rémi NGHIEM-XUAN

58

dégradation cellulaire est avéré et passe de 50% après 1min à plus de 60% après 45 min.

Une centrifugation <20 000 g pendant 15min est donc conseillée pour la

récolte/séparation d’Haslea ostrearia.

10.3.2.6 Conclusion

Toutes les techniques étudiées par Rossignol sont très efficaces pour le cassage

cellulaire d’Haslea ostrearia et démontrent ainsi que cette souche est particulièrement

sensible face aux systèmes de récolte, transfert et séparation. A l’origine mises en œuvre

pour mesurer l’efficacité d’extraction de la marennine intracellulaire, ces études

démontrent une fragilité liée à différentes caractéristiques physiologiques. D’après les

précédentes études par microscopie et granulométrie, Rossignol présente la Figure 44,

qui permet de comparer la répartition des tailles de cellule après traitement ultrasons,

pompe à engrenage et broyeur haute pression.

Figure 44: Comparaison de la granulométrie des suspensions selon la méthode de cassage cellulaire (Rossignol 1999)

La pompe à engrenages et les ultrasons sont des méthodes aussi efficaces pour la

libération de marennine mais les particules broyées restent de dimensions plus élevées

Page 85: Par « Rémi NGHIEM-XUAN

59

(diamètres compris entre 1 et 3 µm après). Le broyage haute pression, quant à lui, réduit

la cellule à des fragments de très faible diamètre (<1 µm). Le système de pompe à

engrenages semble casser très rapidement les apex cellulaires, alors que les ultrasons

« déboîtent » les deux parties du frustule (épithèque et hypothèque) libérant ainsi la

marennine interne. La technique de centrifugation n’est pas présentée sur la figure 44.

Cependant, les résultats montrent qu’il est nécessaire de manipuler cette souche à

15 000g maximum pendant 15 min pour ne pas observer de mortalité cellulaire.

Rossignol conclut ainsi, que parmi les techniques utilisées pour l’extraction de la

marennine, la pompe à engrenage représente le meilleur compromis : quantité extraite et

simplicité de mise en œuvre. Cependant, cela conclut aussi que l’apport ou le transport

d’Haslea ostrearia ne peut se faire par pompe sans altérer les rendements au sein du

photobioréacteur. Seul le système de pompe péristaltique semble avoir un faible impact

sur les cellules jusqu’à environ 200 passages à Q=25 L. h-1, mais les cultures supposeront

des débits parfois plus grands avec des fréquences de passages plus élevées, selon les

systèmes mis en œuvre

11 La culture microalgale en photobioréacteur

La culture de microalgues est ancestrale, elle avait déjà lieu il y a 2000 ans, en Chine,

où Nostoc, une cyanobactérie était récoltée pour l’élaboration de farine (Spolaore et al.,

2006). Au Mexique, en 1325, les Aztèques récoltaient la Spiruline dans le lac Texcoco pour

former des galettes séchées qu’ils consommaient (Belay et al., 1996). La première

production commerciale de spiruline démarre au Mexique par la compagnie Sosa

Texcoco, possédant plusieurs centaines d’hectares d’évaporateurs solaires en spirales

(pour une précédente activité de production de carbonate de sodium et chlorure de

calcium). Après un développement inattendu d’Arthropsira platensis, l’entreprise décida

d’utiliser en 1940, 20 hectares de ses évaporateurs pour la culture de cette cyanobactérie

(Fig. 45).

Page 86: Par « Rémi NGHIEM-XUAN

60

Figure 45: Infrastructures de production de spiruline de la compagnie Sosa Texcoco 1960

Très vite, le constat du besoin en lumière limitant les microalgues, entraîne

l’élaboration de systèmes spécifiquement pensés pour la production de microalgues,

appelés photobioréacteurs (PBRs) pour les plus contrôlés. Ces procédés de cultures sont

catégorisés par taille, niveau de contrôle des paramètres de culture (pH, température,

lumière, etc…), mode d’agitation, source lumineuse (solaire ou artificielle) ou encore

géométrie (plan, tubulaire, etc…). Aucun PBR de culture universelle n’a été créé pour

l’instant. L’intégration de la lumière dans le procédé le rend plus complexe qu’une

technologie classique. Ainsi, les technologies PBR actuelles, industrielles et de laboratoire,

dépendent de la zone géographique, des ressources à la disposition (eau, surface,

électricité, soleil…), des connaissances sur la souche cultivée (fragilité, extrêmophile,

besoin en lumière…), et du marché visé par l’application du procédé (agro-alimentaire,

cosmétique, pharmaceutique, environnement) (Borowitzka 1999 ; Chen et al., 2011 ;

Gouveia 2011). Cela amène au final à différentes technologies, chacune ayant ses

avantages et inconvénients.

11.1 Les systèmes de culture ouverts

Ces types de système de culture, sont les premiers apparus, permettant un contrôle

des paramètres de cultures moins efficaces (lumière souvent solaire, température peu ou

pas contrôlée, contaminations microbiennes importantes) a contrario des systèmes de

culture fermés. Permettant des productivités souvent plus faibles que les systèmes

fermés, ils sont moins coûteux (économique et énergétique). Ils permettent cependant,

des productions importantes sur de grandes surfaces. Les premiers systèmes sont des

bassins artificiels, aussi appelés lagunes telles que Sosa Texcoco ou encore Hutt Lagoon

en Australie (Fig. 46).

Page 87: Par « Rémi NGHIEM-XUAN

61

Figure 46: Vue aérienne d'une culture de Dunaliella salina à Hutt lagoon, Australie (infrastructure BASF)

Cela consiste en une étendue d’eau d’une profondeur généralement comprise entre

20 et 40 centimètres sans agitation. Dans ce cas, les paramètres de cultures ne sont pas

maîtrisés et les productivités dépendantes des conditions environnementales

(Borowitzka 1989). Par la suite, des systèmes ouverts agités ont été développés pour

éviter la sédimentation, améliorer l’accès à la lumière, aux nutriments mais aussi éviter

les gradients de pH et le phénomène de stratification thermique pouvant altérer la

productivité (Pulz et Scheibenbogen 1998). Les plus célèbres sont circulaires agités par

un bras de même rayon que le bassin, mais la surface d’occupation au sol est non optimale

(Chaumont 1993). Ils ont été très vite remplacés par les bassins agités nommés

« raceways », pour leur forme de « champs de courses » (Fig. 47). Cette technologie

contient environ 20 à 40 centimètres de volume de culture (pour les systèmes industriels)

agitée le plus souvent par l’intermédiaire d’une roue à aubes.

Figure 47:Vue aérienne des raceways de Cyanotech Co Hawaii, USA

Page 88: Par « Rémi NGHIEM-XUAN

62

Une variante sur la base du plan incliné (Doucha et Lívanský 1995). L’inclinaison

de 1,7% permet le ruissellement de la culture qui est réinjectée par des pompes (Fig. 48).

Ce principe permet l’augmentation de la surface spécifique éclairée (aS) limitant ainsi le

volume d’eau utilisé pour une même surface, permettant l’utilisation moins intensive

d’eau ainsi que la réduction du coût de séparation de la biomasse.

Figure 48: PBR cascade à Trebon, république Tchèque (Malapascua et al., 2014)

A l’échelle industrielle, les systèmes ouverts sont encore les plus utilisés dus à leurs

faibles coûts de construction et d’entretien. Cependant, le manque de contrôle de certains

paramètres de culture et la pression constante de la compétition microbienne ne

permettent pas d’obtenir les meilleurs résultats de productivités. De plus, toutes les

souches de microalgues ne sont pas capables de supporter ces conditions. Actuellement,

les extrêmophiles Dunaliella salina et Arthrospira platensis sont principlement cultivées

en condition industrielle dans des systèmes ouverts, car supportant des conditions

limitant la pression microbienne (respectivement hypersalinité et milieu fortement

alcalin). Un autre désavantage du système ouvert est le contact avec l’air ambiant. L’air

étant pauvre en CO2 par rapport aux besoins, le carbone inorganique transféré à la culture

sera faible, apportant une limitation pour la culture d’algue.

Selon Tredici (2003) la productivité d’Arthorspira platensis en système ouvert est

estimée à 40 t.ha-1.an-1, confirmé par une étude effectuée en Espagne (Jiménez et al.,

2003). Cependant, le climat en zone tempérée réduirait les rendements de 50%, confirmé

par l’étude de Tredici et Materassi (1992), avec un rendement de 20 t.ha-1.an-1 obtenu sur

plusieurs années.

Page 89: Par « Rémi NGHIEM-XUAN

63

11.2 Les systèmes de culture fermés

Plus évolué, ce type de système de culture généralement appelé photobioréacteur

(PBR), n’est que peu utilisé à l’échelle industrielle. Malgré des productivités plus

importantes et un niveau de contrôle des paramètres de culture optimaux, le prix en

fonctionnement, de construction et de maintenance ne permet de produire que des

métabolites à très hautes valeurs ajoutées, comme pour l’exemple de la société Alga

Technologies, Ltd, basée en Israël, leader mondial de la production d’astaxanthine

d’origine naturelle, cultivant Haematococcus en PBR tubulaires (Fig. 49) fermés (Prokop

et al., 2015). On note toutefois ces dernières années que plusieurs unités de production

ont été réalisé sur cette base technologique. On peut citer ici les 10 unités A4F au Portugal,

Subitec en Allemagne, Microphyte en France.

Figure 49: Système fermé de production d'Algatechnologie, Israël (http://WWW.algatech.com)

Les PBR sont très utilisés à l’échelle laboratoire, les PBR fermés permettent la

culture de microorganismes photosynthétique en conditions axénique complète. De plus,

la faible surface d’échange avec l’air ambiant permet une désorption du carbone limitée

et limite les carences en carbone. Finalement, ces systèmes de culture ont une surface

spécifique éclairée élevés. Toutes ces conditions permettent l’étude précise des réactions

biochimiques s’effectuant au sein du PBR et du comportement de l’outil biologique. Deux

PBR du GEPEA sont présentés ci-dessous, le photobioréacteur Torique à agitation

mécanique (Fig. 50 (Droite)) et le photobioréacteur Airlift agité par flux d’air, (Fig. 50

(Gauche)).

Page 90: Par « Rémi NGHIEM-XUAN

64

Figure 50: Culture d'Arthrospira platensis en photobioréacteur Airlift, GEPEA (Gauche). Culture de Chlorella vulgaris en photobioréacteur Torique, Algosolis (droite)

Contrairement aux systèmes de culture ouverts, valorisant l’éclairage solaire, les

systèmes fermés ont aussi la possibilité d’être lancés sous éclairage artificiel, permettant,

là encore, l’optimisation de la productivité ainsi que la précision des études scientifiques.

Ces systèmes sont alors séparés en trois catégories :

- Les systèmes classiques à éclairage artificiel : Précédemment présentés, souvent

de faible contenance pour des productions intensifiées et l’étude en laboratoire.

Pouvant, cependant être aussi utilisés pour des productions industrielles, c’est le

cas lorsque la nécessité est d’avoir tout au long de l’année une quantité et une

qualité de biomasse constante.

- Les systèmes classiques à éclairage solaire : Présents à l’échelle pré-industielle

pour le passage de l’échelle laboratoire (système éclairé) à l’industrielle (pilote)

mais aussi à l’échelle industrielle comme Alga Technologies, produisant une

molécule à haute valeur ajoutée.

- Les systèmes de rupture : De nouveaux systèmes au stade de développement,

comme la technologie Algofilm du GEPEA (Pruvost et al., 2017).

Les photobioréacteurs peuvent aussi être classés par technique d’agitation,

technologie Airlift, agitation bras mécanique ou hélice, recirculation, etc… faisant varier

la géométrie des systèmes. La technologie photobioréacteur est encore en pleine

évolution. Dans le but de réduire les coûts de consommations (notamment en eau et

d’électricité (éclairage)) et d’augmenter les productivités en biomasse obtenues, de

nouveaux systèmes de culture sont actuellement en développement. Il n’existe pas de

Page 91: Par « Rémi NGHIEM-XUAN

65

photobioréacteur idéal adapté à la culture de toutes les algues et aux situations

particulières et chaque contrainte, y compris biologique, devra être prise en compte pour

le passage à une production industrielle.

11.3 Facteurs limitants de la croissance micro-algale

11.3.1 La lumière

11.3.1.1 Atténuation de la lumière en PBR et les régimes de production en PBR

La lumière représente le paramètre le plus important d’une production de

microalgues en condition autotrophe car elle conditionne la croissance par la

photosynthèse et donc la productivité (biomasse et/ou molécule d’intérêt) de notre

système. Considérant alors la lumière comme un « nutriment » pour la croissance des

algues, l’irradiance se propageant dans le PBR sera donc « consommée » par la population

de microorganismes photosynthétiques du PBR. Cependant, contrairement aux macros et

micronutriements ajoutés directement au sein du PBR via le milieu de culture, l’énergie

lumineuse est hétérogène dans le volume du PBR. Elle varie en fonction de la

concentration de la biomasse produite (les cellules plus nombreuses vont absorber plus

de photon) et éventuellement de la production de certaines molécules capables

d’absorber/réfracter la lumière. Connaître et contrôler la quantité de lumière au sein du

système de production est donc essentiel pour obtenir des productivités maximales de la

microalgue.

Comme souligné précédemment, l’atténuation de la lumière sera dépendante de la

lumière captée en surface du PBR appelée PFD (Photon Flux Density exprimée en µmolhv.

m-2. s-1), de la concentration en biomasse (Cx), souvent exprimée en g. L-1, des propriétés

radiatives du microorganisme photosynthétique dépendantes elles-mêmes, de la teneur

en pigment (exprimé en %) et de la forme de la cellule cultivée, ainsi que des propriétés

physiques de sa structure cellulaire (Souliès 2014). Trois régimes d’atténuation de la

lumière au sein du PBR ont pu être mis en évidence, donnant lieu à différentes

performances de production pour une même culture (Pruvost et Cornet, 2012).

Représentée par γ, la fraction volumique éclairée du PBR est définie par la lumière reçue

par les microalgues (ou irradiance, noté G) dans un volume donné. Ce paramètre γ, est

donc le ratio entre le volume éclairé et le volume total du PBR. L’intensité lumineuse

diminuant au sein du PBR pour les raisons précédemment citées, il existe alors, au sein du

Page 92: Par « Rémi NGHIEM-XUAN

66

PBR une zone éclairée permettant la croissance microalguale et une zone trop sombre

pour assurer une croissance positive. Ces deux zones sont délimitées par l’irradiance de

compensation notée Gc. Les trois régimes décrits sont :

- Le régime de culture photo-inhibée (Fig. 51) :

Figure 51: Profil d'atténuation lumineuse pour un PBR plan en régime photo-inhibé (Prusvot et Cornet, 2012)

Ce régime aussi appelé régime cinétique, signifie que tout le flux de photons n’est

pas absorbé par la culture, ainsi γ>1. Comme présenté sur la figure 51, toute la lumière

n’est pas absorbée par la biomasse, celle-ci était trop faible en concentration. Cela peut

amener également à une lumière reçue par cellule très importante, pouvant diminuer les

performances en croissance, voire photo-inhiber les cellules. Les microalgues vont alors

diminuer la production de chlorophylle et augmenter celle de caroténoïdes protecteurs

(effet de photo-inhibition). Cette instabilité pigmentaire entraîne, sur de très faibles

écarts de temps de résidence, une instabilité de productivité flagrante. Ce type de culture

n’est donc pas stable dans le temps pour des productions industrielles, avec la

conséquence d’une instabilité des conditions photosynthétiques, ce qui ne permet pas

d’obtenir les meilleurs rendements de productivité.

- Le régime de culture luminostat (Fig. 52) :

Page 93: Par « Rémi NGHIEM-XUAN

67

Figure 52: Profil d'atténuation lumineuse pour un PBR plan en régime luminostat (Prusvot et Cornet, 2012)

Ce régime signifie que l’intégralité de la lumière est absorbée par la culture, sans

apparition d’une zone sombre. Ainsi γ=1 comme présenté sur la figure 52. Cette situation

permet d’atteindre la productivité maximale ainsi qu’une stabilité pigmentaire durable en

limitant l’apparition du phénomène de respiration, pour des productions industrielles de

microalgues.

- Le régime de culture photolimitée (Fig. 53) :

Figure 53: Profil d'atténuation lumineuse pour un PBR plan en régime photolimité (Pruvost et Cornet, 2012)

Ce régime de culture signifie que l’intégralité de la lumière est absorbée par la

culture. Cependant dans ce cas, une zone d’ombre apparaît, favorisant le phénomène de

respiration au sein du PBR, amenant donc à une diminution des performances en culture.

Page 94: Par « Rémi NGHIEM-XUAN

68

Un temps de résidence long va augmenter la concentration en biomasse et augmenter le

volume de la zone sombre. Cela diminuera tout autant la productivité.

11.3.1.2 Caractérisation de l’absorption du flux lumineux par les microalgues

Souvent utilisée pour lier la croissance d’une culture de microalgues au paramètre

lumineux, l’irradiance G ne permet pas, seule, de comprendre le flux de lumière absorbé

par unité de biomasse microalgale. Comme présenté précédemment, l’absorption

lumineuse par les microalgues dépendent de leurs formes, tailles, structure et

composition pigmentaire. Ainsi, il est plus pertinent de considérer le flux absorbé plutôt

que le flux disponible pour la caractérisation du comportement d’une culture en PBR. La

vitesse d’absorption de photons ou RPA (Rate of Photons Absorption) exprimée en

µmolh𝑣. Kgx-1. s-1 et notée A est utilisée pour caractériser l’absorption du flux lumineux

(Cornet et al.,1998 ; Pruvost et al.,2008 ; Kandilian et al., 2014). Cette vitesse spécifique

d’absorption de photons par unité de biomasse est liée à l’irradiance G et au coefficient

massique d’absorption de la biomasse Ea (propriétés radiatives de la biomasse), celui-ci

pouvant être déterminé par spectrophotométrie à sphère d’intégration (Pilon et al., 2011)

ou par prédiction théorique (Pottie et al., 2005).

Les travaux du GEPEA (Pruvost et Cornet, 2012) permettent de considérer

l’atténuation de la lumière sur une seule dimension (épaisseur de la culture notée L) grâce

à l’utilisation de PBR plans et la captation de la lumière en surface de manière homogène.

L’atténuation lumineuse au sein d’une culture est alors représentée par le modèle dit à

deux flux (Cornet et al., 1995 ; Cornet et al., 1992) utilisant le modèle de Schuster (1905)

décrivant le comportement de la lumière dans une atmosphère brumeuse mais en prenant

en compte les propriétés radiatives des microalgues (effets d’absorption et de diffusion

de la lumière par les microalgues). Les travaux de Kandilian (2016) ont permis d’obtenir

les propriétés radiatives de Chlorella vulgaris. Une fois ces données connues, il est possible

de déduire soit le champ de vitesse d’absorption de photon, soit la moyenne sur la valeur

de culture (appelé MRA, pour Mean Rate of Photon Absorption). Cependant, dans le cas

où la forme, la composition pigmentaire, la taille et la structure sont différentes de

Chlorella, ou que ces informations sont difficiles d’accès, il est possible de mesurer la

valeur de MRPA par un bilan sur la phase photonique dans le volume de culture (Pruvost

et al., 2008) :

Page 95: Par « Rémi NGHIEM-XUAN

69

Équation 6

𝑨 =𝑺

𝑽 𝑪𝑿

(𝒒𝑶 − 𝒒𝑳) =𝒂𝑺

𝑪𝑿

(𝒒𝑶 − 𝒒𝑳)

Où q0 et qL représentent respectivement le flux de photon à l’entrée et à la sortie

du système en µmol m-2 s-1. CX représente la concentration de biomasse g. L-1. aS

représente la surface spécifique du PBR, soit le rapport entre la surface éclairée et le

volume du PBR (en m-1). Ce paramètre est lié au système de culture.

11.3.2 La température

La température est une condition essentielle à la culture en photobioréacteur car

elle impacte l’efficacité de l’activité enzymatique cellulaire et indirectement les réactions

photochimiques (Raven et Geider, 1988). Une température de culture non optimale

affecte la croissance de la microalgue et entraîne des productivités plus faibles mais

engendre aussi de possibles stress conjugués lors de l’étude des paramètres de culture,

modifiant la composition et/ou la réaction de la microalgue face aux conditions, comme

Raven et Geider le montrent sur les lipides.

La majorité des microalgues sont mésophiles, ainsi, elles possèdent une

température optimale de croissance comprise entre 15 et 40 °C (Richmond et Hu, 2008).

Les organismes psychrophiles (TOPT<15°C) et thermophiles (TOPT>50°C) sont capables de

croître en conditions extrêmes. Cependant, aucun microorganisme photosynthétique

possédant une température optimale supérieure à 75°C n’a été pour l’instant découvert

(Due à l’instabilité de la chlorophylle à cette température) (Varshney et al., 2015). Les

travaux de Ras et al., (2013) apportent des précisions sur les gammes de températures

supportées par différents groupes taxonomiques de microorganismes photosynthétiques

(Table. 6).

Tableau 6 : Gamme de température supporté par 5 groupes taxonomiques de microorganismes photosynthétique (Ras et al., 2013)

Groupes

Taxonomique Echelle de température (°C)

Bacillariophyceae 5-25

Dinophyceae 15-25

Prymnesiophyceae 16-25

Cyanophyceae 15-36

Chlorophyceae 20-36

Page 96: Par « Rémi NGHIEM-XUAN

70

Peu de microorganismes photosynthétiques connus et cultivés à l’échelle

industrielle, se cultivent à des températures supérieures à 30°C, à part Spirulina.

Lemasson (2011) compare plusieurs modèles de croissance de la microalgue Chlorella

vulgaris par rapport au paramètre température et démontre que la productivité d’une

culture continue pourrait être modélisée par des modèles utilisés pour la croissance des

plantes supérieures (modèle d’Alexandrov) ou celle des bactéries (modèles

d’Hinshelwood et de Ratkowsky) (Fig. 54).

Figure 54: Evolution de la productivité surfacique en biomasse de Chlorella vulgaris en fonction de la

température (PFD =130 μmolhv. m-2. s-1) (Souliès 2014)

D’autres travaux (Butterwick et al., 2005 ; Kudo et al., 2000 ; Xin et al., 2011 ;

Dermoun et Chaumont, 1992) mettent en lumière la diversité du comportement

microalgal face à la température (Fig. 55). Cependant, malgré cette diversité, les résultats

de la culture de Chlorella vulgaris correspond aux modèles d’Alexandrov, d’Hinshelwood

et de Ratkowsky qui centre la température optimale de croissance entre 20 – 25 °C.

Page 97: Par « Rémi NGHIEM-XUAN

71

Figure 55: effet de la température sur le taux de croissance de Phaeodactylum tricornutum, Scenedesmus sp., Dinobryon divergens et Porphyridium cruentum (Ras et al., 2013)

La température ne va pas altérer que les caractéristiques physiologiques de la

souche mais va avoir un impact non négligeable sur la chimie du milieu de culture, et ainsi

altérer les rendements de la culture. Les travaux de Le Gouic (2013) mettent en évidence

cet impact sur la solubilité du carbone, même dans les conditions de températures

supportées par les cellules (Table. 7).

Tableau 7: Evolution de la solubilité du carbone en fonction de la température, PCO2=380 ppm (Le Gouic, 2013)

Température Solubilité (Kg. m-3) 10 2,37 15 2,01 20 1,72 25 1,50

Ainsi, la culture en PBR thermostaté, conduit à de meilleures productivités (à

température optimale de croissance) et à des résultats reproductibles.

11.3.3 Le pH

De même que pour la température, une gamme de pH optimale de croissance de

chaque microalgue existe. Chlorella vulgaris est capable de croître dans des cultures à pH

compris entre 5 et 9. Cependant, les croissances optimales sont obtenues dans des

cultures à pH compris entre 7 et 8. Les cyanobactéries sont souvent capables de supporter

des gammes de pH plus élevées, et donc des conditions de cultures souvent moins

axéniques, comme dans le cas des cultures ouvertes à pH 10 de spiruline (Richmond

1986).

Page 98: Par « Rémi NGHIEM-XUAN

72

Tout comme pour le cas de la température, le contrôle du pH au cours d’une culture

en photobioréacteur est essentiel pour la reproductibilité des résultats et l’assurance

d’obtenir des productivités optimales (à pH optimale de croissance).

Il est possible de réguler le pH de différentes manières :

- Par l’ajout de NaOH ou HCl, à l’image des cultures de microorganismes en

bioréacteurs. Le phénomène de respiration engendrant la décroissance de pH au

cours de la culture (Phase sombre), il est nécessaire de rajouter du NaOH pour

contrôler le pH. Dans le cas d’une culture autotrophe, le phénomène de

photosynthèse engendre une croissance du pH au cours de la culture, l’ajout d’HCl

est nécessaire.

- Par l’ajout de CO2. L’augmentation du pH liée au phénomène de photosynthèse,

peut être régulée par l’ajout de CO2, dont l’acidité tamponne le pH de la culture.

Ainsi, cette technique permet la régulation du pH tout en assurant un apport en

carbone inorganique constant.

Cette montée en pH des cultures en photo-auhotrophie est liée à la consommation

du Carbone Inorganique Dissous (CID), absorbé sous forme de HCO3 dans les

chloroplastes et consommé sous forme de CO2 comme le montre l’équilibre chimique

suivant (Eq. 7) :

Équation 7

𝑯+ + 𝑯𝑪𝑶𝟑 → 𝑪𝑶𝟐 + 𝑯𝟐𝑶

La Rubisco fixe le CO2 produit dans le cycle de Calvin et consomme le proton (H+)

libéré lors du processus photosynthétique. La transformation du HCO3– en CO2 libère alors

un ion OH- au sein de la cellule, qui doit le neutraliser par l’absorption de H+

extracellulaire. La baisse de H+ dans le milieu entraîne donc la montée du pH (Chi et al.,

2011).

En système de culture ouvert, la concentration en carbone inorganique au sein de

la culture et en équilibre avec l’atmosphère contenant des concentrations proches de

380ppm en CO2 (Le Gouic, 2013). Ainsi l’ajout de CO2 en continus justifie pour assurer un

apport en carbone inorganique constant. L’injection de CO2 en PBR assure donc une

Page 99: Par « Rémi NGHIEM-XUAN

73

culture non limitée en carbone inorganique. Les études de Le Gouic, basées sur Chlorella

vulgaris, permettent d’affirmer des besoins de 10 mM en CID pour une culture non limitée

de 1g. L-1 (MS). Ce carbone est apporté en début de culture par du bicarbonate de sodium

et/ou alimenté au cours de la culture par injection de CO2.

Le CID est présent sous différentes formes dont l’équilibre et la physico-chimie est

régie par le pH de la culture (Fig. 56).

Figure 56: Représentation des fractions molaires de carbone dissous en fonction du pH à 25°C (Kaplan et al., 1986)

Seul le CID sous forme de CO2 est soumis au phénomène d’équilibre entre les

phases aqueuses et gazeuses, pouvant engendrer une désorption du CID de la culture.

Ainsi, comme représenté sur la figure 56, les pH élevés à partir de 8 favorisent une faible

désorption du CO2 du fait d’une faible concentration en CO2 dissous dans le milieu. Ce

phénomène est souvent mis à profit dans les cultures de Spiruline en système de culture

ouvert pH>9. Cependant, la proportion plus importante des autres formes de carbone à

pH élevé, peut aussi engendrer des réactions inattendues entre espèces chimiques au sein

du milieu de culture, comme la précipitation des sels minéraux (Olsen et al., 2006).

11.3.4 La composition du milieu de culture

La composition du milieu de culture est essentielle pour la culture non limitée des

microalgues. Les microorganismes autotrophes consomment des nutriments sous forme

de minéraux divisés en deux classes, les macronutriments et les micronutriments.

Page 100: Par « Rémi NGHIEM-XUAN

74

Les macronutriments constituent les éléments que la microalgue requiert (mais ne

consomme pas forcément) une forte concentration (supérieure à 10 mg. g-1 de masse

sèche) tels que le carbone, l’azote, l’hydrogène, le phosphate, le calcium, le magnésium, le

souffre, le potassium et la silice pour les diatomées (Kaplan et al., 1986 ; Grobbelaar

2003). Le carbone est l’élément majoritaire qui constitue la biomasse, il représente

environ la moitié de la masse sèche de la microalgue. L’hydrogène et l'oxygène sont

présents en grande quantité aussi, mais obtenus à partir de la photolyse de l’eau ou par

consommation de l’oxygène dissous. Ainsi, le deuxième élément le plus important à

fournir après le carbone est l’azote dont la proportion peut varier de 1 à 10% de la matière

sèche. Ce macronutriment est indispensable à la production de pigments (essentiel à la

photosynthèse) et aux protéines (multiplication cellulaire). Dans le cas des diatomées, la

proportion de l’azote est souvent de l’ordre de 1 %, mais la cellule consomme aussi en

concentration similaire, la silice pour constituer son frustule. L’azote est consommé sous

forme d’ions nitrates pour la majorité des cas, mais peut aussi être consommé sous forme

d’ammonium ou encore de nitrite (dépend des espèces). Certains microorganismes sont

capables de consommer des formes organiques (urée) ou encore sous forme gazeuse (N2)

(Kaplan et al., 1986).

Le troisième élément essentiel à la croissance cellulaire est le phosphore. Il

représente en moyenne moins de 1% de la masse sèche, mais reste majeur pour la

production d’énergie chimique (ATP). Il est assimilé sous forme inorganique (HPO2-,

H2PO4-) mais certaines algues consomment du phosphate sous forme organique (Na2-β-

Glycérophosphate) (Andersen 2005). Le tableau suivant présente ces nutriments et ce

qu’ils représentent en termes de pourcentage dans la matière sèche selon Grobbelaar

(2003) (Table. 8).

Tableau 8: liste des macroéléments nécessaires à la croissance microalgale (Grobbelaar 2003)

Macroéléments Composés % de la matière sèche

C CO2, HCO3-, CO32-, molécules

organiques 17,5-65

O O2, H2O, molécules organiques 20,5-33

H H2O, H2S, molécules organiques 2,9-10

N N2, NH4+, NO3-, NO2-, aminoacides, urée

etc. 1-14

Page 101: Par « Rémi NGHIEM-XUAN

75

Na Multiples sels inorganiques (NaCl,

Na2SO4…) 0,04-4,7

K Multiples sels inorganiques (KCl,

K2SO4…) 0,1-7,5

Ca Multiples sels inorganiques (CaCl2,

CaCO3…) 0-8

P H2PO4-, HPO4

2- 0,05-3,3

S Multiples sels inorganiques (MgSO4…) 0,15-1,6

Mg Multiples sels inorganiques (MgSO4,

MgCl) 0,05-7,5

Cl NaCl, KCl, CaCl2… -

Si Na3SiO3, 9H2O 0-23

Les micronutriments constituent les éléments que la microalgue requiert (mais ne

consomme pas forcément) à faible concentration (Inférieur à 10 mg. gx-1 de la masse

sèche) tel que le fer, le brome, le manganèse, le cuivre, le molybdène, le valium, le cobalt,

le nickel, le sélénium, l’aluminium, le lithium, le rubidium ainsi que les vitamines et autres

facteurs de croissance (Kaplan et al., 1986 ; Grobbelaar 2003). Bien que présents en

faibles quantités, ces éléments sont essentiels pour la croissance des microalgues.

Cependant, peu d’études déterminent des besoins précis en micronutriments et leur

dosage reste encore à l’heure actuelle très complexe en raison des faibles concentrations.

Le tableau suivant présente ces nutriments et ce qu’ils représentent en termes de

pourcentage de matière sèche selon Grobbelaar (2003) (Table. 9).

Tableau 9: liste des microéléments nécessaires à la croissance microalgale (Grobbelaar 2003)

Microélément Composés % de la matière sèche

Fe FeCl3, Fe(NH4)2SO4 0,02-3,4

Zn, Mn, Br, B, Mo, V, Sr,

Al, Rb, Li, Cu, Co, I, Se - <0,1

Ce manque de connaissances sur les besoins en microéléments des microalgues a

poussé les premiers chercheurs dans le domaine, à baser leurs milieux de culture sur la

composition de leurs environnements naturels respectifs. La composition des

Page 102: Par « Rémi NGHIEM-XUAN

76

échantillons de substrats naturels est analysée et certains milieux de culture sont même

élaborés à partir des substrats naturels stérilisés (eau de mer, eau minérale, terre,

sédiments, etc… (Andersen, 2005)). La culture de microalgue intensive en PBR nécessite

un milieu de culture concentré en macro et macronutriment pour s’assurer d’une culture

non limitée et des productivités importantes. De même, un milieu artificiel est nécessaire

pour s’assurer d’une reproductibilité des résultats de productivités. Dans le cas de la

culture d’Haslea ostrearia, le laboratoire MMS a déjà cherché à optimiser la composition

du milieu de culture pour l’élaboration d’un milieu d’eau de mer artificiel, appelé DAM

(Gagneux et al., 2007).

11.3.5 L’agitation

L’agitation n’est pas considérée, dans la majorité des cas, comme un paramètre

directement lié à la croissance de la cellule. Cependant, comme présenté chapitre I.4.3,

certains microorganismes répondent différemment au facteur agitation.

D’un point de vue biologique, certaines microalgues possèdent une physiologie

nécessitant une agitation plus faible voire aucune agitation. Cette agitation induit des

forces (contraintes) pouvant hypothétiquement altérer l’intégrité physique de la cellule

ou inhiber la croissance par un phénomène de mécano-sensitivité. Il est généralement

constaté que certaines microalgues (dont Haslea ostrearia) ne peuvent supporter une

agitation conventionnelle, sans altérer les productivités, faisant du paramètre agitation,

un paramètre limitant au sein du PBR.

D’un point de vue physique, le paramètre agitation est considéré comme le plus

important depuis la mise en place des premiers réacteurs chimiques et bioréacteurs, au

même titre que la température et le pH (Hochfeld, 2005 ; Gabelle et al., 2012). L’agitation

permet l’optimisation de la cinétique de réaction chimique, favorisant l’homogénéité de

plusieurs paramètres tels que la thermique, la salinité, le pH et la concentration en

biomasse (pour les bioréacteurs et PBR). Ainsi, indirectement liés, le contrôle et le

maintien d’une bonne agitation permettent d’optimiser la productivité en biomasse en

bio-procédé.

Pour la culture de microalgues, les essais de Phillips et Myers en 1950, démontrent

qu’une culture agitée produit une biomasse plus dense. Ils expliquent cette différence par

la présence de gradients thermiques et de pH au sein du système de culture non agité

Page 103: Par « Rémi NGHIEM-XUAN

77

(Richmond et Hu, 2008). De plus, les travaux de Pruvost et Cornet (2012) présentent

l’homogénéité d’une culture comme un aspect essentiel pour de bonnes productivités

dues à un accès homogène des cellules à la lumière. Différents systèmes d’agitation sont

utilisés en PBR, des systèmes mécaniques et non-mécaniques, permettant la mise en

mouvement des microalgues. L’agitation non mécanique est l’agitation par bullage. Ce

type d’agitation à deux avantages : les contraintes de cisaillement générées sont

généralement faibles. De plus, il permet de favoriser le transfert de l’oxygène dissous de

la phase liquide vers la phase gazeuse, ce qui réduit le risque d’inhibition de la croissance

lié à une trop forte concentration en oxygène dans le milieu de culture (Kazbar, 2018).

L’agitation mécanique peut s’effectuer via une pompe de recirculation, une hélice

marine, une roue à aube ou encore par agitation manuelle en fonction de l’application et

de la géométrie du système de culture utilisé.

11.4 La culture d’Haslea ostrearia en photobioréacteur

Malgré la difficulté de cultiver Haslea ostrearia, le GEPEA et le MMS ont mis en place

différents types de systèmes de culture en cellules libres et en cellules immobilisées

(Rossignol 1999 ; Lebeau et al., 2000). Les hypothèses de l’incapacité d’Haslea ostrearia à

croître en système de cultures conventionnelles sont nombreuses. Les besoins en

nutriments de cette souche sont, depuis longtemps remis en question, ainsi que la gestion

des paramètres de cultures (Température, pH et irradiance). Cependant, l’agitation et plus

précisément les contraintes de cisaillement sont souvent citées.

11.4.1 Culture en cellules libres agitée

Peu de photobioréacteurs conventionnels en cultures agitées ont été utilisés pour

cultiver Haslea ostrearia, car la culture en erlenmeyer agité et en PBR montre déjà une

mort cellulaire (Jubeau, 2009). Très vite, l’agitation est apparue comme un paramètre

suspecté d’inhiber la croissance et/ou briser les cellules d’Haslea. Le GEPEA s’est penché

sur l’étude de la résistance d’Haslea en soumettant la cellule à des systèmes de

recirculation et de destruction cellulaire (Présenté chapitre I.4.3.2) et constate une

sensibilité réelle plus importante face à ces systèmes chez Haslea que chez d’autres

microalgues modèles cultivées au laboratoire (Chlorella vulgaris, Chlamydomonas

reinhardtii etc…).

Page 104: Par « Rémi NGHIEM-XUAN

78

Jubeau (2009) a essayé de cultiver Haslea ostrearia au GEPEA dans le premier PBR

de technologie Airlift, malheureusement sans succès (Fig. 57).

Figure 57: Photographie de la première culture d'Haslea ostrearia en photobioréacteur de technologie Airlift au GEPEA (Jubeau, 2009)

L’agitation a donc toujours été une problématique essentielle à la culture d’Haslea

ostrearia poussant les scientifiques du GEPEA et de MMS à se tourner vers d’autres

technologies. Les travaux de Rossignol et al., 1999, permettent ainsi la culture d’Haslea en

cellules libres agitées grâce à la mise en place d’un photobioréacteur à membrane

immergée (Fig. 58).

Page 105: Par « Rémi NGHIEM-XUAN

79

Figure 58: Installation expérimentale d'un photobioréacteur à membrane immergée pour la culture d'Haslea ostrearia (Rossignol et al., 1999)

Le PBR à membrane immergée est constitué d’un cylindre de verre de 40 cm de

hauteur, 15 cm de diamètre d’un volume total de 7 L pour un volume réactionnel de 6 L.

Deux modules plans d’ultrafiltration de 2 x 45 cm2, en polyacrylonitrile, à un seuil de

coupure de 40 KDa (IRIS 3038 Orelis, Miribel-France), sont placés au fond du PBR. Le

principe de ce PBR consiste à filtrer le surnageant au court du temps grâce à la force

hydrostatique de l’eau, puis de renvoyer, une fois par jour, du milieu de culture frais en «

retrolavage » par la membrane sous un gradient hydrostatique de 6 kPa. Pour optimiser

le transfert de nutriments, le PBR est agité par barreau aimanté pendant 5 min après

l’addition du milieu frais. Les faibles coûts et les faibles cisaillements générés par une

circulation de substrats sans pompe, représentent tout l’intérêt de ce système de culture.

Les productivités en marennine de ce système atteignent 2,45 ± 0,25 mgEMn. L-1. J-1

(Rossignol et al., 1999) et son considéré comme un record pour l’époque.

11.4.2 Culture en cellule immobilisées

L’immobilisation cellulaire consiste à fixer une cellule à un support. Ce concept est

divisé en deux types : l’immobilisation passive et active.

Page 106: Par « Rémi NGHIEM-XUAN

80

11.4.2.1 L’immobilisation cellulaire passive

L’immobilisation passive consiste à fixer des cellules sur support en utilisant la

capacité naturelle de ces cellules à produire une matrice adhérente au support (Moreno-

Garrido, 2013). Haslea ostrearia semblant en être capable de par sa physiologie (Rincé et

al., 1999), le laboratoire MMS a développé cette technologie. Dans un premier temps, la

culture de cellules immobilisées passivement s’est effectuée en système de culture fermé

à éclairage artificiel tel que les erlenmeyers et les bonbonnes de verre borosilicatées non

agitées (Fig. 59).

Figure 59: Photographie de cultures d'Haslea ostrearia en immobilisation cellulaire passive : (Gauche) en erlenmeyer non agité 500 mL, (Droite) en bonbonne de verre borosilicaté non agité 10 L

Basées sur les travaux du laboratoire MMS, ces techniques de culture permettent

d’atteindre de 2 à 10 mg. L-1 de marennine extracellulaire en condition de batch

(Rossignol 1999 ; Jubeau 2009 ; Prasetiya et al., 2016). Le laboratoire canadien de l’UQAR

a lui aussi développé des systèmes de cultures fermés à éclairage artificiel expérimental.

Les PBR de 100 L (Fig. 60) permettent de monter à des concentrations en marennine

d’environ 6 mg. L-1 et des productivités en batch de 0,2 mgEMn. L-1. J-1 (Gastineau et al.,

2014 ; Turcotte et al., 2016).

Page 107: Par « Rémi NGHIEM-XUAN

81

Figure 60: Culture d'Haslea ostrearia dans les PBR expérimentaux UQAR : (a) photographie de deux des PBR, (b) courbe de la EMn mesuré dans le surnageant durant 35 jours de batch (Gastineau et al., 2014)

Le laboratoire MMS a de son côté mis en place un système de culture ouvert à

éclairage solaire de 10 m3, ouvrant la voie de la culture d’Haslea ostrearia en extérieur

pour des applications ostréicoles (Fig. 61).

Figure 61: Méthodologie de la préparation de la culture d'Haslea ostrearia en tank de 500 L et en open pond expérimental de 10m3 (Turpin et al., 2001)

Les conditions de culture de ce système sont moins contrôlées que pour le cas de

systèmes fermés. Les productivités obtenues sont présentées ci-dessous (Fig. 62).

Page 108: Par « Rémi NGHIEM-XUAN

82

Figure 62: Variation de la densité cellulaire et de la concentration en marennine dans le surnageant au sein des culture en batch d'Haslea ostrearia (en tank 500 L et pond 10 m3) (Turpin et al., 2001)

Ce système permet d’atteindre une concentration de 2 x 107 cellules. L-1 et une

concentration en marennine extracellulaire de 3,4 mg. L-1, soit une productivité

volumique moyenne de 0,34 mg. L-1. J-1.

11.4.2.2 L’immobilisation cellulaire active

L’immobilisation cellulaire active consiste à fixer des cellules sur support ou

bloquer les cellules dans une matrice, en utilisant une technologie d’immobilisation

biocompatible pour adhérer/ bloquer les cellules sur/dans le support (Moreno-Garrido,

2013). Parmi les méthodes d’immobilisation, beaucoup sont utilisées pour les

microorganismes non photosynthétiques tels que l’immobilisation dans/sur la mousse

synthétique, les billes/le gel de silice, des fibres de matériaux rugueux etc… La contrainte

de la propagation de la lumière, indispensable pour les microorganismes

photoautotrophes, limite les méthodes d’immobilisation active aux gels d’agar-agar,

d’alginate, carraghénane et autres gels transparents.

11.4.2.2.1 Inclusion cellulaire en matrice agar-agar

La technique d’inclusion cellulaire en gel d’agar-agar est très utilisée en

microbiologie pour le comptage et même la culture cellulaire (Lebeau et al., 1999 ;

Moreno-Garrido, 2013). Lebeau valorise ici cette technique pour bloquer les cellules au

centre du PBR dans un gel d’agar, au contact d’un tube percé où débouchaient des fibres

optiques permettant la diffusion de la lumière dans la fine couche d’Agar. Une membrane

poreuse constituée d’une grille en Inox maintenait la structure de gel sur le tube Inox (Fig.

63).

Page 109: Par « Rémi NGHIEM-XUAN

83

Figure 63: représentation du photobioréacteur à cellules immobilisées expérimental (Lebeau et al., 2000)

Le milieu de culture frais est injecté en continu (D=0,25 J-1) et circule dans le PBR

et le gel d’agar grâce à une pompe pour la récupération de la marennine extracellulaire.

Malgré l’ingéniosité du procédé, les résultats de productivité volumique restent faibles et

ne dépassent pas 0,5 mgEMn. L-1. J-1 (Rossignol 1999 ; Lebeau et al., 2000).

11.4.2.2.2 Inclusion cellulaire en billes d’alginates

Le principe de cette technique consiste à inclure des cellules à l’intérieur d’un

polymère d’alginate de calcium. Cette technique de culture cellulaire a été étudiée sur

d’autres microalgues avant Haslea, souvent dans le but de protéger les cellules de

l’agitation, d’optimiser la récupération d’exométabolites et/ou de biomasse (Moreira et

al., 2005), faciliter l’ingestion de biomasse en aquaculture (Espinosa et al., 2006), pour la

gestion en souchotèque (Lebeau et al., 1998) ou encore pour le traitement de l’eau (Wilde

et Benemann, 1993 ; Vilchez et al., 1997).

Page 110: Par « Rémi NGHIEM-XUAN

84

Figure 64: représentation d'une inclusion cellulaire dans un polymère d'alginate de calcium

Comme présenté en figure 64, les cellules (microalgues) mélangées dans la

solution d’alginate de sodium, se retrouvent incluses au sein des billes d’alginate lors de

la polymérisation (instantanée) de l’alginate avec des cations bivalents (Mg2+, Sr2+, ou

Ca2+…).

Quelques études ont été réalisées sur la croissance d’Haslea ostrearia et Nitzschia

closterium en billes d’alginate (Lebeau et al., 1998 ; Espinosa et al., 2006 ;) notamment

pour le stockage de ces souche (Fig. 65).

Figure 65: photographie au microscope optique : (A) de billes d'alginate vide, (B) de microcapsule d’alginate contenant Nitzschia closterium (Lebeau et al., 1998)

Les résultats de production de biomasse/marennine dans ses conditions, ne

montrent pas de productivités importantes (expérimentation pour le stockage) mais

surtout, la conservation du polymère d’alginate ne dépasse pas 15 jours avant de montrer

des signes de dégradation dans l’eau de mer. Cependant, la solution de production en

Page 111: Par « Rémi NGHIEM-XUAN

85

billes d’alginate, pour la protection des cellules contre les cisaillements reste

envisageable. Malgré une dissolution du gel en eau de mer due essentiellement à la

présence de cations, certaines techniques de chélation des cations « parasites » et/ou de

renforcement du gel sont possibles (Moreira et al., 2006 ; Castro-Ceseña & Del Pilar

Sánchez-Saavedra, 2014). D’autre part, il a été constaté que la division cellulaire

engendrait aussi une perturbation de la structure des billes (Lebeau et al., 1998).

Page 112: Par « Rémi NGHIEM-XUAN

86

Conclusion

Haslea ostrearia est une microalgue eucaryote de la classe des Bacillariophyceae

(diatomées), du genre Haslea. Cet organisme photosynthétique est capable de produire et

d’excréter de la marennine (Gastineau et al., 1983), un pigment bleu-vert. Découverte

avec le phénomène de verdissement des huîtres, la marennine est sécrétée par Haslea

ostrearia dans les claires à huîtres et est filtrée par les huîtres (et d’autres bivalves) pour

donner à leurs branchies une couleur verdâtre (2014 ; Barillé et Cognie, 2000 ; 1998 ;

Gastineau et al.,). La marennine a donc un intérêt organoleptique pour la production

d’huîtres mais ce pigment possède aussi des propriétés antibactérienne, antivirale,

antitumorale et allélopathique (Gastineau et al., 2012) à fort potentiel applicatif.

Contrairement aux microorganismes photosynthétiques modèles (Chlorella sp,

Spirulina sp, Chlamydomonas sp etc…) la culture d’H. ostrearia en photobioréacteur dans

des conditions conventionnelles est considérée comme très difficile. La réponse de cette

souche tychopélagique face à l’agitation reste incomprise. Différents essais ont été

cependant réalisés par le laboratoire MMS, du GEPEA ainsi que celui de l’UQAR, comme

présentés ci-dessous (Fig. 66).

Basé sur des constats de croissances observées en erlenmeyers à agitation

magnétique ainsi que dans des bioréacteurs en verre (données non publiées), une mort

cellulaire est très vite observée. De plus, Rossignol et al., (1999), démontre qu’Haslea

ostrearia souffre, plus que d’autres microalgues (modèles) face aux contraintes imposées

par la recirculation en pompe. Ainsi, Lebeau a créé un PBR à cellules immobilisées.

D’autres technologies de cultures immobilisées (actives et/ou passives) ont été

développées telles que la culture en bassin pour le verdissement des huîtres (Turpin et

al., 2001), la culture en billes d’alginate (Lebeau et al., 1998) et encore les PBRs

expérimentaux de l’UQAR au Canada (Gastineau et al., 2014). Le GEPEA se tourne vers des

technologies à cellules agitées (pas en continu) telles que la première culture en PBR

airlift (Jubeau, 2009), sans succès, ou encore le PBR à membrane immergée de Rossignol

(1999), permettant, jusqu’ici, les meilleurs rendements en productivité volumique de

EMn. Malgré ces avancées technologiques conséquentes, différentes problématiques de

cultures restent à résoudre, et ont fait l’objet de cette thèse, dont le but est de créer et/ou

optimiser un photobioréacteur spécifique pour la culture continue industrielle d’Haslea

ostrearia.

Page 113: Par « Rémi NGHIEM-XUAN

87

Figure 66: Représentation chronologique des innovations de productivité en marennine extracellulaire selon les différentes technologies de systèmes de culture mis en place par les laboratoires MMS, du GEPEA et de l'UQAR

Page 114: Par « Rémi NGHIEM-XUAN

88

Ainsi, les différents systèmes de culture, présentés figure 66, n’ont pas utilisé les

mêmes milieux de culture ainsi que les paramètres de cultures. Ces différences

proviennent majoritairement des contraintes de chaque procédé, mais aussi d’un manque

de connaissances sur la réponse d’Haslea ostrearia face aux paramètres de culture.

Beaucoup de travaux ont été mis en place pour déterminer le meilleur milieu de culture

pour Haslea ostrearia (Turpin et al., 1999 ; Gagneux-Moreaux et al., 2007 ; Mouget et al.,

2009). Cependant, les seuls milieux retenus pour la production en continue (industrielle

ou non) sont des milieux à base d’eaux de mer naturelle (F/2, ES1/3 présenté chapitre

III), pouvant être limitant pour la production d’EMn et diminuant la reproductibilité des

résultats de productivité en biomasse et en marennine. D’autre part, le pH et la

température de culture utilisés dans tous les systèmes de cultures sont ingénieusement

basés sur les conditions de culture de l’écosystème d’Haslea ostrearia. Cependant, ces

deux paramètres, très importants pour une culture, peuvent être à l’origine d’une

limitation de la productivité et les informations récoltées sur la réponse d’Haslea ostrearia

à ces paramètres dans la littérature, ne permettent pas de nous assurer des conditions

optimales de production.

La sensibilité d’Haslea ostrearia à l’agitation reste aussi à éclaircir. Cependant,

quelques études ont déjà été publiées essentiellement par le laboratoire du GEPEA. Cette

cellule semble plus sensible que des modèles tels que Chlorella face à des traitements de

broyages, recirculation et séparation (Rossignol, 1999). Aucun type de pompe ne permet

de faire recirculer Haslea ostrearia en continu sans altérer l’intégrité cellulaire. Le

traitement de sonication semble provoquer l’altération du frustule mais surtout déboîte

les deux parties de celui-ci et déclenche la mort cellulaire à partir de 2 minutes de

traitement à 550 W. L’impact du traitement de centrifugation est aussi non négligeable

sur Haslea ostrearia avec 50% de mortalité cellulaire >20 000 g.

Cependant, les contraintes appliquées sur la cellule lors de ces traitements

s’éloignent fortement des contraintes potentielles apportées dans un photobioréacteur.

Selon le mode d’agitation et le type de photobioréacteur, les contraintes au sein du PBR

peuvent être fortement différentes, mais la déformation du fluide liée à l’agitation d’un

système de culture entraîne inévitablement des contraintes de cisaillement pouvant

impacter l’intégrité cellulaire (Olmos et al., 2003 ; Pruvost et al., 2006 ; Ndiaye et al., 2018

; Wang & Lan, 2018).

Page 115: Par « Rémi NGHIEM-XUAN

89

Ce phénomène de sensibilité cellulaire aux contraintes de cisaillement (ou autres

types de contraintes) est appelé mécano-sensitivité cellulaire (Griffin et al., 1992 ;

Tavernarakis & Driscoll 1997) et peut favoriser (ou non) la production d’un métabolite

chez certaines cellules tout comme déclencher la mort cellulaire, souvent à hauts

cisaillements. L’impact physique du cisaillement déclenché par l’agitation n’est donc le

plus souvent, pas directement responsable de la baisse de productivité. Un lien peut

apparaître avec la physiologie de la cellule. La question est donc ici, de comprendre si

Haslea ostrearia est réellement sensible aux contraintes de cisaillement, ou si elle possède

une physiologie différente des microalgues considérées comme modèles en production

lorsque les cisaillements sont appliqués.

Pour terminer, la production de marennine extracellulaire par Haslea ostrearia est

un phénomène beaucoup étudié, qui pourtant reste encore sujet à débat. Beaucoup

d’hypothèses divergentes sont soulevées dans les travaux existant (Moreau, 1970 ;

Neuville et Daste, 1978 ; Rossignol, 1999 ; Lebeau et al., 2000 ; Tremblin et al., 2000 ;

Robert et al., 2002 ; Mouget et al., 2005). Le mécanisme de synthèse et d’accumulation de

marennine reste globalement inconnu. D’autre part, les connaissances accumulées sur la

réponse d’Haslea ostrearia face à la lumière ne permettent pas d’établir une gamme de

tolérance précise. Les limites lumineuses qu’Haslea est capable de supporter sont encore

méconnues à l’échelle du PBR de laboratoire alors que ces informations sont essentielles

pour l’élaboration/optimisation d’un PBR à échelle industrielle.

Page 116: Par « Rémi NGHIEM-XUAN

90

Chapitre II:

Matériel et méthodes

Page 117: Par « Rémi NGHIEM-XUAN

91

1 Matériel biologique

1.1 Haslea ostrearia

La microalgue utilisée lors des études de production durant cette thèse est Haslea

ostrearia, (diatomée) prélevée à Bouin (France, Vendée) et isolée au laboratoire MMS (EA

2160) de l'Université de Nantes. Cette souche provient de la Nantes Culture Collection

(NCC), elle est référencée sous le numéro NCC 495. Les caractéristiques biométriques de

ce clone sont présentées dans le tableau 10. Ce clone a été sélectionné pour sa production

de marennine.

Tableau 10: Biométrie du clone NCC 495 d'Haslea ostrearia et aspect au début et à la fin de la thèse***

Date de

biométrie

02/03/2016 20/03/2018

Longueur (µm) 73,09±0,72 31,40±0,36

Largeur (µm) 8,80±0,37 6,04±0,15

Aspect

La déformation de la cellule observable le 20/03/2018 (Tableau 10) semble

inévitable et survient dès les premières semaines de la mise en culture en verrerie.

D’autre part, la cellule diminue en taille au cours des divisions successives sur plusieurs

générations, ce qui explique son différentiel de taille deux ans après la première culture

d’Haslea. Cette décroissance de taille est liée à la physiologie des diatomées et finit par

réduire les productivités (biomasse/marennine) d’Haslea. Ainsi, le clone NCC 501

(Tableau 11) a été fourni par le MMS pour terminer les résultats de thèse.

*** Les valeurs sont significatives ± IC 95% pour n=50.

Page 118: Par « Rémi NGHIEM-XUAN

92

Tableau 11: Biométrie du clone NCC 501 d'Haslea ostrearia et aspect à réception du clone†††

Date de biométrie 09/04/2018

Longueur (µm) 65,45±0,32

Largeur (µm) 6,89±0,15

Aspect

1.2 Nannochloropsis galitana

La souche Nannochloropsis gaditana est une Chlorophycée marine. Le clone 527,

provenant de la collection NCMA/CCMP est cultivé depuis 2015 (Taleb et al., 2015) au

GEPEA. Cette souche n’a été utilisée que pour la validation du photobioréacteur à cellules

immobilisées.

1.3 Chlorella vulgaris

La souche Chlorella vulgaris est une Chlorophycée d’eau douce. Le clone F&M-M33,

provenant du Fotosintetica & Microbiological Srl Culture Collection, est cultivé depuis

2015 (Taleb et al., 2015) au GEPEA. Cette souche n’a été utilisée que pour le test de

résistance de la technologie d’immobilisation cellulaire par encapsulation en bille

d’alginate face à l’agitation en airlift.

††† Les valeurs sont significatives ± IC 95% pour n=50.

Page 119: Par « Rémi NGHIEM-XUAN

93

2 Techniques de mesure

2.1 Suivi de la Densité numérique en cellules

2.1.1 Comptage cellulaire

La concentration en nombre de cellules est déterminée pour Haslea ostrearia au

microscope optique (objectif x40) par comptage sur hématimètre de type Nageotte pour

les cultures à faible biomasse. Les cultures en photobioréacteurs, nous permettent de

suivre la concentration de biomasse par la matière sèche car les concentrations en

biomasse sont plus importantes.

2.1.1.1 Comptage à la cellule de Nageotte

La cellule de Nageotte est une lame spéciale que l'on place sous un microscope

pour le comptage de diverses cellules animales et végétales (Fig. 67).

Figure 67: Cellule de Nageotte Figure 68: Coupe latéral de la cellule de Nageotte

Un quadrillage, dont on connait les dimensions, est précisément gravé au laser

(Fig. 68), et une lamelle délimite une hauteur elle aussi connue.

La gravure verticale de la cellule de Nageotte est composée de 40 colonnes, plus

une ligne seule au centre. Les lignes sont espacées de 0,25 mm. La gravure horizontale est

constituée de 2 lignes, espacées de 10 mm. Le volume de chaque colonne représente

1,25µl (Fig. 69).

Page 120: Par « Rémi NGHIEM-XUAN

94

Figure 69: Gravure de la cellule de Nageotte

Pour les éléments chevauchant les lignes, les cellules qui chevauchent les lignes à

gauche et en haut sont comptées. Ne sont comptés aucun élément chevauchant les lignes

à droite et en bas. On peut ensuite extrapoler la densité cellulaire de la culture, après

comptage, grâce au calcul suivant :

Équation 8

𝑁 =n

V

Soit n le nombre d’éléments comptés

N le nombre d’éléments par microlitre (µl)

V le volume de comptage.

Le volume V représente 5 µl pour 4 bandes de 1,25 µl comptées pour être

représentatif de la concentration de l’échantillon. Toutes les mesures ont été réalisées en

triplicats pour un minimum de 300 cellules.

2.1.1.2 Mesure de masse sèche

L’estimation de la matière sèche s’effectue sur filtre Whatman GF/F avec un

diamètre de pore de 0,7µm. Les filtres sont placés pendant 24h à l’étuve à 110°C pour

retirer l’humidité. Ils sont refroidis avant mesure dans un dessiccateur pendant 10

Page 121: Par « Rémi NGHIEM-XUAN

95

minutes, puis pesés sur une balance de précision. Ces filtres sont ensuite utilisés pour

filtrer la biomasse contenue dans un volume donné d’échantillon après étude sur la

souche Haslea ostrearia, le volume minimum pour obtenir un résultat significatif est 10

ml. Le même volume de formiate d’ammonium (NH4HCO2 à 28 g. L-1) est filtré dans le but

de retirer les sels du milieu de culture. Le filtre est ensuite séché 24h à l’étuve, refroidi 10

minutes dans le dessiccateur puis pesé. La différence entre la masse du filtre seul et celle

du filtre colmaté par les algues, divisée par le volume filtré permet d’obtenir la matière

sèche en (g. L-1). Toutes les mesures ont été réalisées en triplicats.

2.2 Suivi pigmentaire

2.2.1 Mesure de la chlorophylle a, c et des pigments caroténoïdes

Les diatomées ne possèdent que de la chlorophylle a et c. Concernant les

caroténoïdes, les diatomées possèdent d’autre formes comme la fucoxanthine, la

diadinoxanthine, le β-carotène et d’autres (Carreto et Catoggio, 1976).

Après comparaison des plusieurs techniques utilisées par le MMS et le GEPEA pour

l’extraction de pigment (Lorenzen, 1966 ; Speziale et al., 1984 ; Ritchie, 2006), la seule

technique permettant d’obtenir, pour Haslea ostrearia des résultats d’extraction des

pigments comparables avec d’autres souches, est la technique basée sur l’extraction à

l’acétone 90%, mise en place par Ritchie (2008). Puis, le contenu en pigments est

déterminé par une méthode spectrophotométrique (Ritchie, 2008). Ainsi, un volume noté

V0 de la culture, placé dans un contenant en verre, est centrifugé pendant 15 minutes à

6000 rpm. Le surnageant est ensuite éliminé et le culot est mis en suspension dans un

volume Vac d’acétone (90.8%) pendant une durée minimale de 30 minutes et maximale de

3 heures, à 3°C et à l’obscurité. Par la suite, les débris cellulaires sont précipités et séparés

par centrifugation pendant 10 minutes à 6000 rpm ; la couleur blanche du culot est un

moyen de vérification que l’extraction est complète. L’absorbance du surnageant

contenant les pigments dissous dans l’acétone est mesurée à différentes longueurs d’onde

(480, 630, 647 et 664 nm) au spectrophotomètre (JASCO V-630). Trois réplicats sont

préparés. Les valeurs moyennes de ces absorbances, sont finalement utilisées pour

calculer les concentrations en chlorophylle a, la chlorophylle c et les pigments

caroténoïdes via les équations de Ritchie, (2008) pour les chlorophylles et de Strickland

et Parsons, (1968) pour les caroténoïdes :

Page 122: Par « Rémi NGHIEM-XUAN

96

Équation 9

[Chl-a] (μg.ml-1) = (-0,3002A630 – 1,7538A647 + 11,9092A664)Vac/V0l Équation 10

[Chl-c] (μg.ml-1) = (23,6723A630 – 7,9057A647 – 1,5467A664)Vac/V0l Équation 11

[PPC] (μg.ml-1) = (4A480)Vac/V0l

Avec l la longueur du trajet optique de la cuve exprimée en cm.

Connaissant la densité numérique en cellules (cellules. mL-1 ou g. L-1) de la

culture, la teneur en pigments en gramme de pigment par quantité de cellule ou gramme

de biomasse sèche peut être calculée.

2.2.2 Mesure de la marennine

La marennine extracellulaire (EMn) est mesurée par spectrophotométrie

(Pouvreau et al., 2006). L’échantillon d’algue d'un volume de 1.5 ml est centrifugé par

MiniSpin® pendant 15 min à 13500 rpm, puis le surnageant est mesuré à 669nm contre

du milieu neuf à l’aide d’un spectrophotomètre JASCO V-630. Les mesures sont réalisées

en triplicat. Après mesure des spectres d’absorption de la marennine extracellulaire à

plusieurs pH, la longueur d’onde pour la mesure de EMn en routine a été fixée à 669 nm

pour un pH de 7,8, Les résultats se recoupent avec ceux de Robert et al., (2002).

La concentration en marennine extracellulaire est alors calculée par l’équation

suivante, en utilisant la loi de Beer-Lambert :

Équation 12

[𝐸𝑀𝑛] =𝐴669

εl

Avec A, l’absorbance de l’échantillon à 669nm, ɛ le coefficient d’extinction molaire,

(ɛ=12x104 mol-1. L. cm-1 pour la marennine extracellulaire) et l le trajet optique de la cuve

en cm égal à 1 pour notre cas (Pouvreau et al., 2006). Le résultat de la concentration

molaire est ensuite multiplié par la masse molaire de la EMn, MEMn=9893 g. mol-1, pour

obtenir le résultat en g. L-1.

Page 123: Par « Rémi NGHIEM-XUAN

97

2.3 Suivi de la composition des milieux de culture

2.3.1 Dosage des ions Nitrate et Phosphate dissous

La mesure des deux principaux macronutriments, N et P, se fait majoritairement

par chromatographie ionique – DIONEX (ICS 110 colonne cationique et ICS 900 colonne

anionique). L’échantillon est injecté dans une colonne cationique à résines échangeuses

d'ions, un flux d'éluant de H2SO4 permettant la migration des ions NH4+ présents. Les ions

sortant de la colonne sont détectés par conductimètre, cette mesure étant directement

reliée à leur concentration. L'identification de l'espèce ionique est assurée par la

comparaison des temps de rétention des différents pics observés avec ceux de standards

; la hauteur ou la surface des pics permet une détermination quantitative.

Les colonnes ne supportant pas plus de 550 mg. L-1 de Na+ et 3900 mg. L-1 de Cl-,

une dilution des échantillons était nécessaire et les premiers résultats de [PO42-] étaient

trop faibles pour être représentatifs. Ainsi l’utilisation des Kit Hach lange®, avec le

système photométrique LCK et les kit Phosphate LCK350 ont permis la récolte

approximative des premiers dosages en milieux de culture peu concentrés.

2.3.2 Dosage des ions silicates

Plusieurs méthodes de dosage de la silice dans l’eau existent, à savoir l’absorption

atomique, la méthode colorimétrique et la spectrométrie de masse au plasma. Seule la

technique colorimétrique au silico-molybdate est réalisable au laboratoire. Elle permet de

doser la silice réactive, soit uniquement la silice en mesure de réagir avec le molybdate.

Le complexe silico-molybdate émet une couleur jaune dans des conditions acides et est

instable dans ces conditions. Des interférences importantes peuvent survenir lors de

l’utilisation de contenant en verre, de tannins, d’une quantité trop importante de fer, de

la couleur de l’échantillon (la couleur bleue de la marennine hydrosoluble), de la turbidité,

des sulfites et surtout dans notre cas des phosphates. Le PO43- lui aussi, réagit avec le

molybdate. Ainsi pour éliminer ses interférences, de l’acide oxalique est ajouté pour

inhiber la réaction phosphate et du molybdate (Grasshoff et al., 2007).

2.3.3 Dosage du Carbone Organique Dissous

Cette mesure est réalisée par la méthode d’oxydation catalytique par combustion

à 680°C, effectuée par le Total Organic Carbon analyser (TOC-L CSH/CSN Shimadzu). Cette

méthode souvent utilisée comme indicateur de qualité de l’eau consiste à mesurer dans

Page 124: Par « Rémi NGHIEM-XUAN

98

un échantillon la concentration de dioxyde de carbone et de sels d’acide carbonique

(Carbonate, bicarbonate…etc) et du carbone total. L’extrapolation des résultats permet la

déduction de la concentration de Carbone organique total.

Figure 70: Mesure des différentes formes de carbones par TOC-L CSH/CSN

La combustion à 680°C de l’échantillon permet la libération de CO2 qui permet la

détection du carbone total (TC) par l’analyseur de gaz infrarouge (NDIR) après

refroidissement (Fig. 70.1). Le carbone inorganique (IC) est lui, oxydé par l’ajout d’acide

sulfurique permettant la libération de CO2 dans le processus de barbotage, puis mesurer

par NDIR (Fig. 70.2). La concentration de Carbone organique total (TOC) est calculée par

la soustraction de la concentration de carbone inorganique (IC) à la concentration de

carbone totale (TC) (Fig. 70.3). Ces analyses sont réalisées en triplicat.

2.4 Mesure du rendement photosynthétique FV/Fm

Chez les plantes supérieures, la mesure du rendement photosynthétique est un

moyen de quantification des effets de stress abiotiques (lumière, température, carences

etc…) sur la santé de la plante, et plus particulièrement son activité photosynthétique. Le

principe de cette mesure repose sur le suivi du transfert d’électrons du photosystème II

par la fluorescence de la chlorophylle face à une intensité lumineuse reçue. L’intensité de

cette fluorescence peut varier en fonction du bien-être ou de la détérioration du PSII liée

aux paramètres abiotiques appliqués. Les mesure de Fv/Fm sont réalisées à l’aide d’un

fluorimètre WaterPam (Walz, Allemagne) après ajout d’un échantillon de microalgues

dans la cuve d’analyse. La fluorescence de départ (Ft) est fixée à 250 (unité arbitraire)

pour toutes les mesures par dilution avec de l’eau de mer de salinité 28 ‰. L’échantillon

Page 125: Par « Rémi NGHIEM-XUAN

99

est alors plongé dans le noir pendant 15 minutes pour désaturer les antennes collectrices

de photons du PSII (pour la mesure de F0) avant l’émission d’un pulse de lumière

saturante et la mesure de la fluorescence maximale de la chlorophylle (Fm). Le rendement

photosynthétique FV/Fm est donné par l’équation suivante :

Équation 13

𝐹𝑣/𝐹𝑚 =𝐹𝑚 − 𝐹0

𝐹𝑚

F0 représente la fluorescence minimale de la chlorophylle pour un centre ouvert (PSII). A

l’état relaxé, c’est le rendement de la fluorescence à l’obscurité.

Fm représente la florescence maximale de la chlorophylle pour un centre fermé (PSII). A

l’état relaxé, cette fluorescence est obtenue en utilisant un pulse saturant de lumière.

FV représente la fluorescence variable, la fraction de l’énergie photonique absorbée

convertible en énergie chimique FV=Fm-F0.

La valeur théorique du FV/Fm maximale est de 0,83 chez les plantes supérieurs, 0,6

en moyenne chez les diatomées et Haslea ostrearia (Rech, 2004).

3 Matériel de culture

3.1 Les milieux de cultures

Plusieurs milieux de culture ont été utilisés, dans un premier temps, pour la

production d’Haslea ostrearia en verrerie puis ensuite retravaillées (Voir Chapitre III)

pour la culture en photobioréacteur. Tous les milieux sont à base d’eau de mer naturelle

enrichie. Le milieu F/2 (Guillard 1975) utilisé pour la culture de diatomées et Enriched

Seawater (ES) modifié ES1/3 par Robert, (1983) recommandé par le Laboratoire MMS est

utilisé uniquement pour la production d’inoculum. Le milieu F/2 a été abandonné au

profit du milieu ES1/3 modifié après étude des sources de phosphates (Voir Chapitre III)

pour de meilleurs rendements de production et une meilleure reproductibilité. Le milieu

Conway enrichi lui aussi à base d’eau mer naturelle a été optimisé pour la croissance de

microalgues salines au GEPEA, et il a été testé pour la production d’Haslea ostrearia en

photobioréacteur. Sans succès pour cette espèce, il a été utilisé pour la production de

Page 126: Par « Rémi NGHIEM-XUAN

100

Nannochloropsis gaditana en PBR immobilisé (voir Chapitre II 9.2.3.1). La composition de

ces milieux est présentée ci-dessous (Tableau 12).

Tableau 12: Composition ionique en macro et microéléments de différents milieux de culture

Ions

F/2

mol. L-1

ES 1/3

mol. L-1

Conway

mol. L-1

Eau de mer

naturelle‡‡‡

mol. L-1

B - 3,23 x 10-5 5,42 x 10-1 4,2 x 10-4

Br - - - 8,4 x 10-4

C 1,36 x 10-4 1,36 x 10-4 1,51 x 10-2 2,0 x 10-3

Ca - - - 1,0 x 10-2

Cl 3,69 x 10-5 4,79 x 10-1 1,85 x 10-2 5,5 x 10-1

Co 4,20 x 10-8 5,79 x 10-8 8,39 x 10-5 -

Cu 3,92 x 10-8 - 8,0 x 10-5 -

F - - - 6,8 x 10-5

Fe 1,17 x 10-5 6,34 x 10-6 4,80 x 10-3 -

K - - - 1,0 x 10-2

Mg - - - 5,3 x 10-2

Mn 9,10 x 10-7 2,33 x 10-6 1,82 x 10-3 -

Mo 5,43 x 10-8 - 5,09 x 10-5 -

N 8,82 x 10-4 3,00 x 10-4 3,52 -

Na 4,60 x 10-4 4,81 x 10-1 4,60 x 10-4 4,7 x 10-1

P 3,62 x 10-5 1,62 x 10-5 3,91 x 10-1 -

‡‡‡ Données analyse eau de mer (Razmig GEPEA) et données Oceanarium Le croisic, France

Page 127: Par « Rémi NGHIEM-XUAN

101

Si 2,00 x 10-4 1,00 x 10-4 2,00 x 10-4 -

S 1,72 x 10-7 1,51 x 10-2 7,99 x 10-5 2,8 x 10-2

Sr - - - 9,0 x 10-5

Zn 1,72 x 10-7 3,01 x 10-7 1,54 x 10-4 -

3.2 Systèmes de culture

3.2.1 Cultures en verrerie chimique

Les pré-cultures sont réalisées en erlenmeyer en verre borosilicaté dans une

enceinte à régulation thermique à cycle de circadien (Fig. 71). La culture d’Haslea est

effectuée à 16°C avec 250 µmol. m-2. s-1 de photons en cycle

jour/nuit 14/10. Le repiquage de la souche s’effectue tous

les 4 jours pour garder la souche en phase exponentielle.

Ce taux de renouvellement rapide permet l’obtention de

biomasse pour une relance fréquente des

photobioréacteurs. Tous les inoculums ont été effectués

sur milieu ES1/3 modifié avec une source de phosphate

inorganique (voir chapitre III.1.4)

3.2.2 Culture en flasques

Toutes les analyses préliminaires ont été effectuées, par manque de biomasse, en

flasques de culture de tissus cellulaires VWR (fig. 72) de 18 ml (volume horizontal). Ce

matériel de culture possède une importante surface spécifique éclairée de aS=180 m-2, et

présente une surface hydrophile permettant une meilleure culture cellulaire immobilisée

en tissus.

Figure 72 : Flasque de culture de tissus cellulaire stérile non traitée (25 cm² de surface de culture)

Figure 71: Enceinte de culture Sanyo MLR-350

Page 128: Par « Rémi NGHIEM-XUAN

102

3.2.3 Culture en photobioréacteur

Cinq types de photobioréacteurs ont été utilisés au cours de cette thèse, dont trois

types de technologie Airlift (PBR Cristal, Plan-6L et Airlift) un PBR torique et un PBR pour

la culture en cellules immobilisées. Le PBR Airlift a été utilisé pour la mise en culture de

la souche, la compréhension des besoins en macronutriments et l’application de forçages

physiologiques. Les PBR Cristal et Torique ont été mis en place dans l’optique d’obtenir

les paramètres optimaux de croissances en conditions photolimitées, mais surtout pour

l’étude des contraintes de cisaillement en PBR avec le PBR Torique. Un photobioréacteur

a été mis en place pour la culture des cellules immobilisées. Un photobioréacteur Plan-6L,

a été utilisé pour l’application des modèles de culture en conditions de pré-production et

mesures de MRPA. Les caractéristiques sont données en tableau 13.

Tableau 13 : Caractéristiques des PBRs utilisés pour la culture d’Haslea ostrearia

PBR Type

as

(m-1)

Vr

(L)

Matériaux Mode

d’agitation

Gestion

thermique

Taux de

cisaillements

(s-1)

Cristal Plan 57 0,25 PMMA,

Inox Airlift Non ≥150

Cellules

immobilisées Plan 33 0,4 PMMA Recirculation Non 0

Airlift Plan 33 1 PMMA,

Inox Airlift Oui 150

Torique Plan 25 1,3 PMMA,

Inox

Hélice

marine Oui 20

Plan-6L Plan 25 6 PMMA Airlift Non ≥275

Tous les PBR sont équipés de sonde pH/température METTLER TOLEDO®. Le pH

est contrôlé par l’ajout de CO2 dans un mélange air/CO2 (5%). La régulation pH s’effectue

à une valeur de consigne donnée, par l’ouverture automatisée de l’électrovanne CO2. La

Page 129: Par « Rémi NGHIEM-XUAN

103

majorité des PBRs (Airlift, Torique, Plan-6L) sont directement gérés via une interface

LabVIEW développée par l’équipe technique du GEPEA.

3.2.3.1 Le photobioréacteur à cellules immobilisées

La géométrie de ce photobioréacteur est pensée pour limiter tout impact des

contraintes de cisaillement dû à une agitation du fluide. Spécifiquement conçu pour les

cultures de cellules benthiques, ce PBR permet aux cellules de tomber au fond du réacteur

pour former un biofilm grâce aux mucopolysaccharides qu'elles produisent. La lumière

est dispensée de manière horizontale par le haut ou le bas du PBR (Fig. 74) grâce à un

panneau de LED (Fig. 73) contrôlé par variateur.

Figure 73: panneau de LEDs à variateur pour irradiance horizontale

Aucune autre agitation que la recirculation de milieu n’est appliquée dans ce PBR

permettant ainsi l’adhésion des cellules en conditions d’immobilisation passive.

Page 130: Par « Rémi NGHIEM-XUAN

104

Figure 74: Schéma du PBR à cellules immobilisées

Bien que ce PBR soit conditionné pour éviter les contraintes de cisaillement,

l’agitation reste indispensable pour l’homogénéisation des nutriments, l’apport de CO2 et

pour éviter la formation de gradient de pH en sortie de gaz et pouvant être dangereux

pour les cellules. D’autre part, les différentiels de pH provoquent dans certains cas des

précipités de nutriments importants pour la croissance cellulaire (SiO32-, PO43- et HCO3-

voir chapitre III.2.4).

Pour cela, un système de recirculation par pompe péristaltique a été mis en place

pour appliquer un faible courant horizontal au sein du PBR, sans engendrer la

perturbation du biofilm. Suite aux expérimentations, un débit de 7 mL. min-1 est appliqué

pour limiter son impact sur la cellule tout en validant la capacité de culture du PBR. Les

performances du PBR ont alors été validées sur une souche de Nannochloropsis gaditana

et comparées avec les performances d’un PBR Airlift. Les résultats de production de

Nannochloropsis (Taleb et al., 2015) se montrent équivalents à ceux obtenus, soit une

productivité surfacique de PS=10,9 g. m-2. j-1. Le PBR à cellules immobilisées est donc

considéré comme fiable pour la culture microalgale.

Page 131: Par « Rémi NGHIEM-XUAN

105

Ce photobioréacteur a aussi été utilisé pour la culture d’Haslea ostrearia en

immobilisation cellulaire dans un gel d’agar, de manière à limiter encore plus l’impact de

l’agitation sur les cellules, celle-ci étant isolées de l’écoulement.

3.2.3.2 Le photobioréacteur Cristal

Ce PBR a été créé au sein du laboratoire GEPEA pour la culture de souches de

microalgues à fort taux de sédimentation (Moutel, 2016). Ce PBR est inspiré des

décanteurs, sa forme resserrée vers la base avec une pente supérieure à 30 degrés, permet

de sédimenter les cellules vers son point d’injection de gaz, permettant la remise en

suspension des microalgues pour garder la culture homogène (Fig. 75).

Figure 75: Schémas du PBR Cristal (Moutel, 2016)

Le réacteur Cristal a été utilisé pour les premiers tests de croissance d’Haslea en

condition d’agitation continue. Le volume réactionnel (250 mL) de ce PBR ne permettant

pas un suivi de culture efficace et continu, les études de contraintes de cisaillement ont

continué avec le PBR Torique. Les contraintes de cisaillements au sein du réacteur Cristal

n’ont pas été mesurées mais estimées. Etant de même technologie que l’airlift, le taux

moyen de cisaillement en régime turbulent au sein de ce réacteur est estimé à 150 s-1

moyen (Ndiaye et al., 2018 ; Chisti and Murray Moo-Young, 1989) avec la présence de

coalescence et d’explosion de bulles localisées, pouvant engendrer jusqu’à 14 000 s-1

(Contreras et al., 1998).

Page 132: Par « Rémi NGHIEM-XUAN

106

3.2.3.3 Le photobioréacteur Airlift 1L

Le photobioréacteur Airlift 1L est le plus utilisé au laboratoire GEPEA, permettant

une agitation douce par l’injection de gaz à sa base en trois points possibles. L’injection

par ses points entraine le liquide vers le haut du réacteur qui redescend par gravité

derrière les séparateurs fixes situés à gauche et à droite du PBR. Cette technologie et

surtout cette géométrie de PBR, n’est pas la plus adaptée pour la croissance

majoritairement benthique d’Haslea, car l’agitation est trop douce et trop localisée pour

empêcher la forte sédimentation due à la physiologie de la souche. Cependant, la majorité

des résultats en croissance ont étaient obtenus dans ces PBRs pour s’assurer que la

souche ne subisse pas un impact trop important des contraintes de cisaillement.

Figure 76: Schémas du PBR Airlift (Souliès, 2014)

Le PBR Airlift 1L (Fig. 76) possède une plaque inox à l’arrière, permettant la

régulation (optionnelle) de la température. Le nombre d’entrées et sorties sont plus

importantes que celles du PBR Cristal, permettant notamment l’ajout de PO4 et SiO2 en

fed-batch (Voir Chapitre V) en plus du milieu de culture. Etant de technologie Airlift, le

taux moyen de cisaillement en régime turbulent au sein de ce réacteur est estimé à 150 s-

1 moyen, avec des maximums de taux de cisaillement estimés 14 000 s-1, et localisés au

niveau des entrées de gaz et de l’éclatement des bulles (Contreras et al., 1998).

Page 133: Par « Rémi NGHIEM-XUAN

107

3.2.3.4 Le photobioréacteur Torique

Le PBR Torique, développé lui aussi au GEPEA, fait partie des réacteurs de

laboratoire les plus intéressants du fait d’un excellent niveau de contrôle des paramètres

de culture et son excellente agitation. La présence d’une hélice marine agite le liquide dans

le sens de rotation horaire (anti-trigonométrique). De plus, l’injection d’air est effectuée

par le piquage du côté gauche du réacteur et va, en remontant le long de la paroi du PBR,

entrainer le liquide et optimiser l’agitation. La forme circulaire du réacteur permet alors

une agitation optimale en limitant les zones « mortes ».

Figure 77: Schémas du PBR Torique (Souliès, 2014)

Le PBR Torique (Fig. 77) possède, aussi une plaque en inox à l’arrière, permettant

la thermorégulation (optionnelle) ainsi qu’un orifice en PMMA à l’arrière, permettant la

mesure du flux incident de lumière absorbé par les algues. Le PBR torique possède une

forme et un système d’agitation d’hélice marine permettant une remise en suspension

permanente d’Haslea en agitation continue au sein d’un pavé circulaire. Pour une rotation

moyenne de 200rpm de l’hélice [permettant un mélange parfait du substrat (Pruvost et

al., 2006)], l’agitation du réacteur torique engendre des contraintes de cisaillement

maximum de 30 s-1 au niveau de l’hélice marine et 10 s-1 dans la zone sans hélice (boucle),

soit une moyenne de 20 s-1 maximum en régime turbulent (Pruvost et al., 2006). L’action

de l’hélice marine génère cependant des vortex ciblés près du mobile d’agitation (Fig. 77).

C’est pourquoi ce réacteur est considéré différemment que les technologies Airlift en

termes de cisaillement, car la circulation en boucle impose à toutes les cellules de passer

Page 134: Par « Rémi NGHIEM-XUAN

108

dans la zone de l’hélice. Ce PBR a donc été utilisé pour l’application de contraintes de

cisaillement sur du long-terme et la mesure de la sensibilité d’Haslea face à ces contraintes

(régimes laminaire et turbulent).

3.2.3.5 Le photobioréacteur Plan-6L

Ce photobioréacteur (Fig. 78) est de technologie Airlift et a été utilisé pour

anticiper les problèmes de sédimentation obtenus en PBR Airlift. L’agitation du réacteur

est assurée par un ensemble de piquages au fond du réacteur, qui permet l’injection de

gaz à travers un « tapis » et donc la répartition homogène du gaz en microbulles sur toute

la largeur du PBR. De plus, ce PBR est entièrement transparent et a permis l’étude du

MRPA avec des prises de mesures de lumière à travers le PBR.

Figure 78: Schémas du PBR plan 6L (Souliès, 2014)

La seule différence notable avec le PBR Airlift est le mode d’agitation qui est

dispensé ici, par une injection du gaz homogène dans le réacteur par le « tapis ». Etant de

technologie Airlift, le taux moyen de cisaillement en régime turbulent au sein de ce

réacteur n’est pas estimé à plus de 14 000 s-1 de manière localisée lors de l’éclatement des

bulles et leurs coalescence (Contreras et al., 1998). Le taux de cisaillement moyen devrait

être estimé à 275 s-1, cependant, la fréquence d’éclatement et de coalescence des bulles

est plus importante avec une répartition des axes d’injections le long du PBR. Ce constat

suppose que le taux de cisaillement moyen de ce PBR est supérieur à 275 s-1 (Barbosa et

Page 135: Par « Rémi NGHIEM-XUAN

109

al., 2003 ; Sobczuk et al., 2006). Ainsi, le débit de gaz a été réduit tous les soirs pour laisser

Haslea ostrearia se fixer et se reproduire dans les zones mortes. Cependant,

contrairement au PBR Airlift 1L, le gaz est réparti que toute la largeur du réacteur

permettant une meilleure agitation pour Haslea en période de jour. C’est grâce à ces

bonnes conditions d’homogénéisation que l’impact du MRPA moyen a pu être mesuré.

Figure 79: Représentation de la technique de mesure expérimentale du MRPA sur le PBR Plan-6L

Contrairement aux autres PBRs utilisé pour la culture d’Haslea ostrearia, le Plan-

6L est entièrement en PMMA ce qui permet la récolte des mesures de MRPA plus précises

(équation 6, chapitre I.5.3.1.2) (Fig.79) grâce à l’utilisation du capteur LI-COR plan en

faces avant et arrière du PBR.

3.3 Modes de culture

La production de microalgues peut être mise en place selon trois modes de culture :

la production en Batch (ou discontinue), semi-continue et continue. Selon les contraintes

associées au raffinage et/ou à la physiologie de la souche, certains modes de culture

seront préférés à d’autres. La productivité de biomasse et de Marennine (dans notre cas)

sera calculée différemment selon le mode de culture utilisé.

3.3.1 Culture en batch

Ce mode de culture représente celui utilisé pour la majorité des productions

d’inoculum, mais il est aussi pratiqué en photobioréacteur. Dans ce cas, le PBR est inoculé

Page 136: Par « Rémi NGHIEM-XUAN

110

à faible concentration d’algues et une forte concentration d’éléments nutritifs. Aucun

volume n’est soustrait ou additionné au PBR et la culture est souvent récoltée lorsque la

productivité faiblit. Le pH et la température sont contrôlés, cependant, la concentration

en nutriment varie au cours de la culture, faisant varier les conditions nutritives dans

lesquelles la microalgue évolue. La productivité volumique est alors calculée ainsi :

Équation 14

𝑃𝑣 =C𝑓 − 𝐶𝑖

𝑇𝑓 − 𝑇𝑖

Avec Cf, la concentration finale, Ci la concentration initiale, Tf le temps à la fin du batch et

Ti le temps au début.

3.3.2 Culture en mode semi-continu

La culture en mode semi-continu est souvent utilisée lors d’un besoin de biomasse

immédiat, par exemple pour le traitement d’un minimum de biomasse fraîche. La récolte

est donc effectuée de façon séquentielle d’un volume calculé pour un taux de dilution

donné. Le volume prélevé est alors immédiatement remplacé par du milieu de culture

frais pour maintenir le volume réactionnel et le taux de dilution fixé. Travailler en semi-

continu nous permet donc surtout de contrôler la récolte du PBR. La productivité

volumique est alors calculée en moyennant les différentes séquences de récoltes

assimilées à une culture en batch sur une certaine période de culture (équation 14). Plus

le volume prélevé (taux de dilution) est important et plus le temps de croissance l’est

aussi, ainsi plus la fréquence de prélèvement est faible et le volume prélevé fort, plus la

culture est assimilée à des batch successifs. De la même manière, plus la fréquence de

prélèvement est forte et le volume prélevé faible, plus la culture se rapproche d’une

culture en continu (équation 15).

3.3.3 Culture en continu

La culture en continu est réalisée par l’ajout en continu de milieu frais et une

récolte de biomasse souvent par surverse. Le volume est donc maintenu, cette récolte en

continu nous permet d’avoir les productivités les plus importantes calculées selon

l’équation suivante :

Équation 15

𝑃𝑣 = D. 𝐶𝑥

Page 137: Par « Rémi NGHIEM-XUAN

111

Avec D le taux de dilution (J-1) :

Équation 16

𝐷 =Q

V𝑟

Avec Q le débit (mL. J-1), Vr le volume réactionnel (ml) et Cx la concentration en biomasse

obtenue (g. L-1).

Dans notre cas très particulier, la caractéristique benthique d’Haslea ostrearia ne

nous permet pas de la cultiver en PBR Airlift sans éviter une forte sédimentation. La mise

en continu de culture d’algues benthiques est impossible dans cette géométrie de

réacteur, car l’ajout en continu de milieu concentrerait les microalgues au lieu de les

diluer. Dans le cas du continu et du semi-continu, il est possible d’appliquer des taux de

dilution directement sur le surnageant sans diluer les microalgues par membrane

(Rossignol, 1999) ou simplement par surverse pour garder/concentrer la biomasse

catalytique.

3.4 Application du flux d’énergie lumineux

Pour la culture de microalgues, la grandeur la plus utilisée en laboratoire est la

densité de flux de photons (Photons Flux Density, PFD) avec pour unité µmolhv. m-2. s-1.

La gestion continue ou dynamique du PFD est effectuée par consignes souvent

automatisées (LabVIEW) de voltages différents aux panneaux de LED. Pour l’agitation de

l’hélice marine du PBR Torique, la gestion est également automatisée.

Le flux lumineux incident présente le paramètre clé pour la modélisation des PBRs,

car il conditionne les performances de l’appareil en termes de cinétique et d’énergétique

(Pruvost et al., 2012). Il doit être connu avec précision pour calculer les profils de l’énergie

radiante à l’intérieur du réacteur. La méthode de mesure physique par capteurs

radiométrique est utilisée sur le cas du PBR à cellules immobilisées, avec un capteur

sphérique (US-SQS/LI (Heinz Walz GmbH, Effeltrich, Allemagne) de 3 mm. Immersible, il

permet d’effectuer des mesures au sein même de la culture. Ce capteur mesure un débit

de fluence photonique (μmol. m-2. s-1), c’est-à-dire la lumière provenant de toutes les

directions de l’espace. En combinaison avec un calculateur (DataLogger) de type LI-COR

1400 (Li-COR, Lincoln USA), ce capteur permet d’obtenir la mesure pour des radiations

Page 138: Par « Rémi NGHIEM-XUAN

112

photo-synthétiquement actives (entre 400 et 700 nm) présentent une erreur relative de

±5% (donnée LI-COR).

Pour la détermination du flux lumineux incident par le capteur sphérique, la

connectique inférieure (prélèvement de biomasse) est retirée pour introduction

perpendiculaire du capteur au fond du PBR. Le dessous du PBR est recouvert d’une

surface noire non-réfléchissante afin de mesurer le flux incident seul (pas de réflexion sur

le fond du PBR). Trois points sont mesurés au sein du réacteur, les valeurs obtenues sont

ensuite moyennées.

Pour le reste des photobioréacteurs, le capteur plan LI-COR 190 SA a été utilisé,

mesurant le flux incident sur une surface plane (angle solide 2π). Ce capteur permet

d’obtenir alors une densité de flux hémisphérique photonique (μmol. m-2. s-1). ). En

combinaison avec un calculateur (DataLogger) de type LI-COR 1400 (Li-COR, Lincoln

USA), ce capteur permet d’obtenir la mesure pour des radiations photosynthétiquement

actives (entre 400 et 700 nm) présentent une erreur relative de ±5% (donnée LI-COR).

Pour la détermination du flux lumineux incident par le capteur plan, on place le PBR dans

sa position de fonctionnement à une distance D du panneau de LEDs. Après démontage

du réacteur, on effectue différentes mesures de flux. Cette mesure est réalisée pour les

différentes positions (le long de la surface éclairée du PBR), les mesures obtenues sont

ensuite moyennées pour calculer une valeur moyenne du flux incident en surface optique

du réacteur.

3.5 Méthode d’immobilisation cellulaire par inclusion en gel d’agar

La méthode d’inclusion cellulaire en gel d’agar est une méthode courante en

microbiologie. Elle est utilisée par Lebeau et al., (2000) pour protéger Haslea ostrearia de

l’agitation en PBR. Le gel d’agar est constitué de 100 ml d’eau distillé (de manière à former

un gel de 1cm d’épaisseur maximum) et d’agar agar à 15% autoclavé à 121°C durant

20min. le flacon stérile est placé dans un bain thermostaté à 40°C maximum pendant une

heure. Les cellules d’Haslea ostrearia sont prélevées en phase exponentielle et

centrifugées durant 10 min à 6000rpm, avant d’être suspendues dans 1mL de solution

saline a 28 g. L-1 de NaCl. Cette solution est rapidement mélangée à la solution d’agar-agar

sous hotte à flux laminaire avant d’être immédiatement coulée dans le PBR à cellules

immobilisées au préalable stérilisé à l’acide peracétique. 30 min est nécessaire sous hotte

Page 139: Par « Rémi NGHIEM-XUAN

113

à flux laminaire pour que le gel solidifie puis les PBR est refermé et placé dans son espace

de culture où le milieu de culture est injecté en conditions stériles.

3.6 Méthode d’immobilisation cellulaire par inclusion en billes d’alginate

Pour l’inclusion cellulaire dans des billes d’alginate 5 mL de solution de culture de

Chlorella vulgaris ont été ajoutés en condition aseptique à une solution d’alginate de

sodium 2%. Après homogénéisation, la solution est extrudée goutte à goutte dans un

tuyau à diamètre constant terminé par une aiguille et poussé par une pompe à débit

constant. La solution d’alginate et d’algue est extrudé dans une solution de chlorure de

calcium à 14,7 g. L-1, sous agitation par barreau aimanté. Les billes sont alors rincées dans

du milieu de culture stérile avant d’être inoculées en PBR.

3.7 Techniques de stérilisation des PBRs

Tous les photobioréacteurs sont stérilisés grâce à l’ajout d’acide peracétique (5‰

dans de l’eau distillé) pendant 30 min. Puis le PBR est rincé (avec son volume) à l’eau

distillée stérile. Enfin, le réacteur doit être encore rincé (dû à l’intolérance aiguë d’Haslea

ostrearia à l’acide peracétique) en continu durant une nuit avec de l’eau distillée stérile et

une agitation gaz pour lessiver et stripper l’acide peracétique. Les connectiques sont

retirées et nettoyées à l’hypochlorite de sodium (Javel) 10% ou autoclavées pour celles

qui peuvent l’être.

3.8 Application des contraintes de cisaillement en Rhéomètre

Le rhéomètre (Physica MCR 500) a été utilisé pour l’application de contraintes dans

le but de mesurer l’impact des cisaillements sur Haslea. Normalement utilisé pour

mesurer la viscosité d’un échantillon, le rhéomètre est capable, sur de faibles volumes

d’échantillon, d’appliquer un taux de cisaillement à une culture de microalgue pour en

mesurer l’effet (constant ou variable). Pour les besoins d’une mesure de viscosité, les

cisaillements ne doivent pas produire de perturbations hydrodynamiques et se doivent

d’être homogènes dans tout l’échantillon. Ce sont ses caractéristiques qui nous

permettent d’appliquer, sur Haslea, des contraintes de cisaillement laminaires

homogènes dans l’échantillon. Le système de mesure avec cylindres coaxiaux de type

Couette (Fig. 80) a été utilisé (plateau TEZ150P et un mobile CC28.7) pour permettre

d’appliquer des plus hauts taux de cisaillement de l’appareil, soit 30 000 s-1.

Page 140: Par « Rémi NGHIEM-XUAN

114

Figure 80: Photos du plateau et du mobile du rhéomètre (1) et dimensions du système plateau/mobile (2)

Dans ces conditions d’utilisation du rhéomètre, le nombre de Taylor est calculé

pour s’assurer de la stabilité hydrodynamique au sein de la cellule de Couette. Un nombre

de Taylor supérieur à 41,3 permet de déceler la présence de micro-turbulence (vortex) au

sein du flux (Mezger 2011). Il est défini ainsi :

Équation 17

𝑇𝑎 =4ω2R4

ν2

Avec ω La vitesse angulaire (m. s-1), R la taille de l’entrefer (m) qui correspond à la

différence du rayon intérieur du cylindre R2 avec celui du mobile R1 (Fig. 80) et ν, la

viscosité cinématique (m2. min-1) de l’eau de mer:

Équation 18

𝑣 =μ

ρ

La masse volumétrique de l’eau de mer est de ρ = 1025 Kg. m-3 et sa viscosité dynamique

µ=1,07 x 10-3 Pa. s-1, ce qui nous donne une viscosité cinématique de ν = 1,04 x 10-6 m2. s-

1.

Page 141: Par « Rémi NGHIEM-XUAN

115

Chapter III:

Design of an artificial culture

medium to optimize Haslea

ostrearia biomass and marennine

production

R. Nghiem Xuan1,3, I. Safitri4, J. L. Mouget1, 2, J. Pruvost 3, V. Turpin1, P. Jaouen3

1 Université de Nantes, MMS, EA 2160, Faculté des Sciences et des Techniques, 2 Rue de la

Houssinière, BP, 92208, 44322 Nantes Cedex, France.

2 MMS (Mer Molécules Santé), Le Mans Université, Ave O. Messiaen, 72085 Le Mans, France.

3 GEPEA, Université de Nantes, CNRS, UMR6144, Bd de l'Université, CRTT - BP 406, 44602 Saint-

Nazaire Cedex, France.

4 Marine Science Department, Faculty of Mathematics and Natural Sciences, Universitas

Tanjungpura, 78124 Pontianak, Indonesia.

Ce chapitre est rédigé sous forme de publication pour soumission dans un journal

scientifique.

Page 142: Par « Rémi NGHIEM-XUAN

116

1 Abstract:

The diatom Haslea ostrearia was first studied by Gaillon in the year 1820 because

of the greening phenomenon of oysters in western France. This microalga has the capacity

to produce and excrete a blue pigment, called marennine, that has antioxidant, anti-

bacterial and anti-viral properties, with possible industrial applications related to

aquaculture and cosmetics. However, it is difficult to produce biomass in large

concentrations in photobioreactors (i.e, usually 1kg m-3 dry weight or higher) due to

stirring sensitivity, and also because diatoms have special requirements, such as a supply

of silica. This work presents a design for a new, completely artificial, seawater medium

named NX (i.e, Nghiem Xuan) for optimizing cultivation of the diatom H. ostrearia in

photobioreactors. NX takes account of the carbon and phosphorus sources in either

organic or inorganic form, and the composition of the main elements (C, H, O, N, P, S, Si).

Optimization of the calcium, magnesium and iron concentrations was found to be

essential in order to achieve the highest productivities. The resulting NX medium was

validated in an airlift photobioreactor. This led to productivities of 4.9 x 108 cell L-1 (ca.

500 mg L-1 of biomass) and 15.7 mg L-1 of marennine, higher than has previously been

reported in the literature.

Keywords: Haslea ostrearia, Marennine, Culture medium, microalgae, Photobioreactor.

Page 143: Par « Rémi NGHIEM-XUAN

117

2 Introduction

Microalgae are increasingly becoming the subject of investigation for several

industrial purposes. They have the potential to produce molecules for pharmacology,

cosmetology, food, bio-based chemicals and biofuels (Spolaore et al., 2006). Over 15,000

valuable chemically-determined compounds, such as antioxidants, carotenoids, fatty

acids, polymers, peptides, enzymes, toxins and sterols (Cardozo et al., 2006), are produced

by all the microalgae that have been identified to date, but only a few species are currently

produced on an industrial scale: Haematococcus (Chlorophyceae), for example, which is

produced for astaxanthin (Panis et al., 2016), and Chlorella (Chlorophyceae) and

Arthrospira platensis (Cyanophyceae), which are produced worldwide for food and

nutraceutical purposes (Iwamoto, 2007, Spolaore et al., 2006; Belay et al., 1996).

Regarding diatoms, which represent the largest group of microalgae (Stoermer and Smol,

1999), applications mainly concern aquaculture (Merz and Main, 2014). Except for some

species like Phaeodactylum tricornutum, which are already produced on a large scale

(Veloso et al., 1991), one of the biggest challenges is the difficulty in setting up a controlled

and optimized large-scale production.

Issues linked to scaling up the cultivation of a given strain usually relate to the lack of

information regarding growth and metabolite production conditions, specific nutrient

needs and responses to light, but also the need to establish adequate tools and methods

for controlled and optimized scaling-up of production. This is generally supported by

bioprocess engineering approaches, such as the design of culture systems, i.e,

photobioreactor engineering (Pruvost et al., 2016), but in some cases, a specific growth

medium has to be specially designed to fulfill conditions of intensified production, which

results in larger biomass concentrations (usually 1kg m-3 dry-weight or higher). The

composition of the medium used in flask culture must be adapted and chemical

dissolution stability can be impaired due to the need for larger nutrient concentrations.

H. ostrearia is a marine diatom that has long been known (Gaillon, 1820) and mainly

studied at laboratory scale. This diatom has the capacity to produce and excrete a blue

pigment named marennine that has industrial potential (Gastineau et al., 2012). There are

two forms of marennine, which differ in their spectral characteristics: the intracellular

marennine (IMn) synthesized in the cell, and the extracellular marennine (EMn) excreted

in the culture medium (Pouvreau et al, 2006). In a culture, the relative abundance of the

Page 144: Par « Rémi NGHIEM-XUAN

118

two forms was found to depend on the medium composition variability, and on abiotic

and biotic factors (Moreau, 1968; Robert, 1983; Robert, et al., 2002).

The Mer-Molécules-Santé (MMS) laboratory has a long experience in culturing H.

ostrearia at laboratory scale, and in recent decades hundreds of strains have been

maintained for years in culture at Le Mans University and in the Nantes Culture Collection

(NCC), using enriched natural seawater media, such as Guillard F/2 (Andersen, 2005), or

Provasoli-based medium ES1/3 modified by Robert (1983). Haslea strains have also been

grown in artificial seawater (ASW), e.g, the diatom artificial medium (DAM) specifically

designed for H. ostrearia culture (Gagneux-Moreau et al., 2007), (not used due to its

composition complexity and cost), or the ASW proposed by Berges et al. (2001), modified

according to Mouget et al. (2009), which results in production as good as with natural

seawater.

Culturing H. ostrearia using conventional photobioreactors (PBRs) is a challenge, so

specific PBRs have been implemented, such as immersed membrane PBRs and agar

immobilization PBRs (Rossignol et al., 1999; Lebeau et al., 2000). In both cases, the yield

of excreted marennine (EMn) per cell remains low (around 25 mg 109 cell-1 d-1 for

immersed membrane PBRs and 5 mg 109 cell-1 d-1 for agar immobilization PBRs). Despite

some outdoor semi-pilot attempts (Turpin et al., 1999), the scaling-up of production is

still an objective for industrial applications.

Our work aims to develop an optimized medium for both biomass and EMn

production in PBRs. A biomass concentration of 1g L-1 in dry weight is targeted, based on

the use of ES1/3, which has been identified as the best medium for EMn production

(Robert, 1983). The newly-designed artificial seawater culture medium NX (i.e. Nghiem

Xuan) was finally tested in a continuous airlift PBR, and revealed the highest EMn yield

for a conventional PBR in the published literature.

3 Material and methods

3.1 Organism and culture maintenance

A strain of the diatom H. ostrearia (NCC 495) was obtained from the Nantes culture

collection of the MMS laboratory and cultivated in natural seawater from Le Croisic

(Loire-Atlantique, France), enriched with Guillard F/2 medium. The salinity was fixed at

Page 145: Par « Rémi NGHIEM-XUAN

119

28 PSU. Cultures of H. ostrearia were maintained in 250 mL Erlenmeyer flasks containing

100 mL medium with 10% dilution after 4 - 5 days of culture. Light was supplied by white

fluorescent tubes (F36W/840 T8) across the day/night cycle 14 h/10 h at 300 µmol m-2

s-1 inside the culture chamber (SANYO MLR-350) at 16°C.

3.2 The culture media tested

Table 1 shows the composition of all the culture media that were compared for this

study. All these media were made from natural seawater provided by the Croisic

Ocearium (Loire Atlantique), except the ASW medium (Mouget et al., 2009), which is an

artificial seawater. Seawater was filtered at 0.7 µm then 0.45 µm before being used as a

medium. The medium was then sterilized in an autoclave at 121°C for 20 min. The average

elemental composition of this water was estimated from a previous measurement of

different seawater composition (data not shown).

The modified Provasoli ES1/3 medium is a natural, enriched seawater not supplied

with copper, unlike Guillard F/2 medium, which is specifically for diatom culture

(Andersen, 2005). The Conway natural seawater medium is a medium that is commonly

used for marine microalgae culture (Taleb et al., 2015).

Because of the precipitation of sodium bicarbonate with phosphate and silica

sources during autoclave sterilization, these elements were added after the sterilization

step by filtration at 0.2 µm under a laminar flow hood.

3.3 Macro and micro nutrients concentration tests

The concentration of nutrients were tested in flat culture flasks of 50ml volume

(flasks for cell culture tissue, VWR collection, France). The flat surface allows better

illumination and light propagation conditions than in conventional Erlenmeyer flasks (i.e,

curved surface). They can, therefore, be considered as a useful alternative for rapid but

reliable screening of culture conditions, such as the impact of various nutrient

concentrations on cell productivity.

Nutrients were divided into two categories, namely macronutrients (C, Ca, Cl, K,

Mg, N, Na, P, S, Si), which represent the main elements involved in cells, macromolecules

synthesis, and micronutrients, which are usually added at small concentration in the

medium (B, Br, Co, Cu, F, Fe, Mn, Mo, Ni, Se, Sr, Zn). It is worth noting that the respective

Page 146: Par « Rémi NGHIEM-XUAN

120

impact of each nutrients is difficult to distinguish, as each addition can involve a counter-

ion and may modify the equilibrium of the medium, producing micro-precipitation, which

may prevent the cells from accessing the nutrient (Andersen, 2005; Miazek et al., 2015).

To avoid chemical contamination with endogenous macro and micronutrient

concentrations from the inoculum, the biomass was centrifuged at 10,500 g (Hettich

Mikro 22R C1110, Marshal scientific, United states) for 10 minutes and the supernatant

was removed. The cells were then rinsed with a sterile solution of 28 g L-1 NaCl at pH 7.8,

prepared beforehand, and re-suspended in it to reach a mean cell density of 3.6 x 105 cell

mL-1. This concentrated culture enabled starting tests in flasks at 104 cell mL-1 with 500

µL inoculum, to prevent dilution of the medium composition. Tests were carried out under

the same conditions as the inoculum culture (SANYO MLR-350 culture chamber,

Richmond Scientific, UK). However, due to the large variation of irradiance distribution

inside the culture chamber, flasks were placed horizontally in column stacks as shown in

Fig. 1 to obtain similar light intensity on each flask.

The cell density was measured using a Nageotte counting chamber (Marienfeld,

AQRA, India) and extracellular marennine was measured by spectrophotometry, as

described by Pouvreau et al. (2006).

Fig. 1: Culture chamber SANYO MLR-350 (left) and culture flask positions inside the culture chamber (right).

3.4 General approach to investigate mixotrophic and autotrophic growth

The role of organic and inorganic carbon and phosphorus sources was investigated

in particular. This consisted of cultivating H. ostrearia cells with different combinations of

sodium acetate (C2H3NaO2 for organic carbon form), sodium bicarbonate (NaHCO3 for

Page 147: Par « Rémi NGHIEM-XUAN

121

inorganic carbon form), Na2-β-glycerophosphate (for organic phosphorus form) and

sodium dihydrogen phosphate (NaH2PO4 for inorganic form) in the same molar

concentrations as in the modified ES1/3 culture medium. The aim of this experiment (Fig.

2) was to check the capacity of H. ostrearia for growth under autotrophic or mixotrophic

conditions and to determine which phosphorus and carbon forms are preferred for

producing biomass and/or EMn.

Special attention was then given to investigating the effect of copper (Cu) on growth.

Copper is an important micronutrient for the growth of microorganisms, such as

microalgae, fungi and bacteria (Cid et al., 1995). It can be lethal, however, in high

concentrations. In addition, heterotrophic metabolism is highly sensitive to Cu

concentrations because it can damage the respiratory chain (in Escherichia coli),

according to Domek et al. (1984; 1987). The Cu is used by enzymes to drive diverse

structures and biochemical reactions. The intracellular Cu used in the respiratory chain

and the process of moving from Cu+ to Cu2+ allows the production of free hydroxyl

radicals. Theses reactive species are responsible for the degradation of proteins, nucleic

acids, lipids, etc, and the death of the microorganism at high copper concentrations (Festa

and Thiele, 2011). The same phenomenon is observed for microalgae, with an impact on

respiration and photosynthesis (Levy et al., 2008). However, in the case of microalgae, the

respiration process is known to be more sensitive than photosynthesis to copper (Wells

et al., 1998).

The presence or absence of Cu in the tested medium can, thus, reveal which

metabolism the cell is forced to use, depending on the medium composition

(photosynthesis or respiration process). As shown in Table 1, F/2 and ASW, the media

possess the same amount of added Cu.

Table 1: Elemental content comparison between the culture media tested

Culture media composition (mol L-1)

Ions

F/2

(Natural

seawater)

ES1/3

(Natural

seawater)

ASW

(Artificial

seawater)

Conway

(Natural

seawater)

25N100P

(Natural

seawater)

B 4.2 x 10-4 4.5 x 10-4 3.7 x 10-4 5.4 x10-1 4.5 x 10-4

Br 8.4 x 10-4 8.4 x 10-4 7.3 x 10-4 8.4 x 10-4 8.4 x 10-4

Page 148: Par « Rémi NGHIEM-XUAN

122

C 1.4 x 10-4 1.4 x 10-4 2.1 x 10-3 1.5 x 10-2 5.8 x 10-2

Ca 1.0 x 10-2 1.0 x 10-2 9.4 x 10-3 1.0 x 10-2 1.0 x 10-2

Cl 5.5 x 10-1 1.0 4.7 x 10-1 5.6 x 10-1 1.0

Co 4.2 x 10-8 5.8 x 10-8 5.8 x 10-8 8.4 x 10-5 5.8 x 10-8

Cu 3.9 x 10-8 - 3.9 x 10-8 8.0 x 10-8 -

F 6.8 x 10-5 6.8 x 10-5 6.7 x 10-5 6.8 x 10-5 6.8 x 10-5

Fe 1.2 x 10-5 6.3 x 10-6 6.6 x 10-6 4.8 x 10-3 6.3 x 10-6

K 1.0 x 10-2 1.0 x 10-2 8.8 x 10-3 1.0 x 10-2 1.0 x 10-2

Mg 5.3 x 10-2 5.3 x 10-2 4.1 x 10-2 5.3 x 10-2 5.3 x 10-2

Mn 9.1 x 10-7 2.3 x 10-6 2.4 x 10-6 1.8 x 10-3 2.3 x 10-6

Mo 5.4 x 10-8 - 6.6 x 10-9 5.1 x 10-5 -

N 8.8 x 10-4 3.0 x 10-4 5.5 x 10-4 3.5 1.1 x 10-2

Na 4.7 x 10-1 4.8 x 10-1 4.2 x 10-1 4.7 x 10-1 4.8 x 10-1

Ni - - 6.3 x 10-9 - -

P 3.6 x 10-5 1.6 x 10-5 2.2 x 10-5 3.9 x 10-1 1.2 x 10-3

S 2.8 x 10-2 4.3 x 10-2 2.5 x 10-2 2.8 x 10-2 4.3 x 10-2

Se - - 1.0 x 10-9 - -

Si 2.0 x 10-4 1.0 x 10-4 1.1 x 10-4 2.0 x 10-4 1.2 x 10-3

Sr 9.0 x 10-5 9.0 x 10-5 8.2 x 10-5 9.0 x 10-5 9.0 x 10-5

Zn 1.7 x 10-7 3.0 x 10-7 2.5 x 10-7 1.5 x 10-4 3.0 x 10-7

3.5 Material for the new medium assessment and productivity validation

Validation of the growth medium on H. ostrearia growth and marennine

productivity was carried out in a lab-scale airlift photobioreactor. This flat-panel PBR has

been used for microalgae strain characterization, as described in Pruvost et al, (2009),

Kandilian et al. (2014), Taleb et al. (2015). The volume was 1 L and the depth 0.03 m

(aLight= 33.3 m-1). Mixing was done by gas injection at the bottom of the PBR. Temperature

was regulated by cool air blowing into the stainless-steel back, and the pH was controlled

by CO2 injection, supplied by an automated regulating valve.

3.6 Statistics

The statistics were compiled using XLSTAT 2019 software (Addinsoft). Normality of

the distribution law for the different samples was identified by the Shapiro-Wilk test to

Page 149: Par « Rémi NGHIEM-XUAN

123

select the tests (T-test). Statistics was measured with triplicate data for all the

experiments presented in this paper.

4 Results

4.1 Determination of key nutrients responsible for Haslea ostrearia

growth

A series of preliminary experiments was conducted to compare H. ostrearia growth

and EMn production in the different culture media to estimate which nutrient

concentrations have the greatest impact. The strain was cultivated in batch conditions in

the Sanyo culture chamber, under culture chamber conditions. The results of biomass and

EMn productivity depending on culture medium are presented in Table 2.

Table 2: Preliminary results of biomass and Extracellular Marennine (EMn) productivities in

different media. Data is mean ±95% CI for n=3.

Medium F/2 ES1/3 ASW Conway

PV Biomass

(106cell L-1 d-1) 19.30 ±11.02 4.38 ±0.22 11.00 ±1.74 0

PSpecific EMn

(mg 106cell-1 d-1) 0.26 ±0.12 0.46 ±0.10 0.27 ±0.02 0

The Conway medium was too concentrated and precipitated during the culture

growth. This resulted in cell death. The three other media resulted in biomass and EMn

production. However, the ASW and F/2 media were found to be more suitable for biomass

production than the ES1/3 medium. It should be noted that ES1/3 medium was

previously identified as the best medium for EMn production, mainly due to its

stimulating effect on cells for EMn production rather than growth (Robert, 1983). This

tends to be confirmed by our results, as higher specific EMn productivity was achieved for

this medium.

Macronutrient elementary concentrations (nitrogen, phosphorus, carbon, sulfate

and silicon) are given in Table 1 for all the growth media. No significant differences were

observed except for a lower nitrogen concentration in the ES1/3 medium. In addition, no

Page 150: Par « Rémi NGHIEM-XUAN

124

copper was added to ES1/3, as it was with the F/2, ASW and Conway media, however,

differences due to the presence of copper in different seawater supplies cannot be

disregarded. Nitrogen concentrations remained proportional to phosphorus

concentration in the three media, and a higher concentration of silica was also found in

ES1/3 than in the other media. The main difference between the media was the supply of

phosphorus, which was organic (Na2-β-glycerophosphate) for the ES1/3 medium.

4.2 Effect of organic and inorganic sources of carbon and phosphorus on

Haslea growth

The various carbon and phosphorus sources (organic or inorganic, noted as org and

inorg respectively), as well as the addition of Cu, were tested in flat flasks (Fig. 2). The

composition of the conventional ES1/3 medium (C-inorg P-org) was shown to strongly

limit cell productivity with only 4.6 x 106 cell L-1 d-1. Biomass productivity was also

strongly inhibited when organic P sources were combined with the addition of Cu, with

2.2 x 105 and 3.3 x 105 cell L-1 d-1 for C-inorg P-org Cu and C-org P-org Cu respectively.

However, when both P and C sources were in organic form, cell productivity was

significantly higher with 17 x 106 cell L-1 d-1. Similar high productivity of was also

observed when using only inorganic sources, with 18 x 106 cell L-1 d-1 for both C-inorg P-

inorg Cu and C-inorg P-inorg.

Page 151: Par « Rémi NGHIEM-XUAN

125

Fig. 2: Cell, Extracellular Marennine (EMn) and specific EMn productivity of Haslea ostrearia in batch culture after 7 days in natural seawater-enriched (ES 1/3) based medium with different sources of phosphorus and

carbon. P EMn specific and Pv are, respectively, the specific EMn productivity and the volumetric productivity. Data is mean ±95% CI for n=3.

Concerning the specific EMn productivity results, carbon and phosphorus sources

also seemed to have a consistent impact. The first apparent observation emphasizes that

all the culture media composed of organic phosphorus seemed to stimulate specific EMn

production. In contrast, the culture media made with inorganic sources of phosphorus

definitely resulted in low levels of specific EMn production. For growth media where

photosynthesis was hypothesized to occur (C-inorg P-org Cu and C-org P-org Cu), low

biomass with no significant differences (T.Test p=0.424) and high specific EMn

productivity was obtained, with respectively 1.37 and 5.38 mg 106 cell-1 d-1. However, the

medium (C-inorg P-org Cu) induced lower EMn productivity than (C-org P-org Cu), with

respectively 0.30 and 1.78 mg L1 d-1. Comparing similar medium compositions with and

without copper, the presence of copper was found to stimulate EMn synthesis in the

presence of organic carbon, whereas the opposite was found with inorganic carbon, with

EMn production significantly lower (T.Test p=0.008).

These results lead to the assumption that H. ostrearia has difficulty consuming

inorganic carbon (photosynthesis) in the presence of an organic source of phosphorus,

0

1

2

3

4

5

6

0

5

10

15

20

25

PV

EMn

(m

g L-1

d-1

)P

EM

n s

pec

ific

(m

g 1

06ce

ll-1

d-1

)

PV

cell

(10

6ce

ll L-1

d-1

)

PV cell PV EMn P EMn specific

Page 152: Par « Rémi NGHIEM-XUAN

126

and may, therefore, favor the organic source of carbon for growth, but this hypothesis

cannot be demonstrated without further studies.

Fig. 3: Impact of organic phosphorus and copper on marennine and biomass productivities according to carbon and phosphorus sources in ES1/3 based medium. Data is mean ±95% CI for n=3.

Figure 3 presents a detailed analysis of the influence of copper and P organic

sources. For a better clarity, each result was normalized against its control. P-org C-inorg

and P-org C-org experiments results have been normalized against the same treatment

with inorganic phosphorus (P-inorg C-inorg and P-inorg C-org). Concerning the Cu

experimental results, these have been normalized against the same treatment without

Copper addition (C-inorg P-org, C-inorg P-inorg, C-org Porg, C-org P-inorg). This

emphasized the significant impact that organic phosphorus has on biomass productivity.

Cell productivity lost 74.6% vs the control with an inorganic carbon source, compared to

cell productivity with an inorganic phosphorus medium, whereas, in the presence of

organic carbon, the culture had 17.1% higher productivity with organic phosphorus than

with inorganic. As a result, it can be concluded that H. ostrearia presents significantly

better growth in either a completely organic or inorganic medium (in terms of carbon and

phosphorus), and presents significant difficulties with growth on organic phosphorus

with inorganic carbon. Additionally, the organic phosphorus allows H. ostrearia to achieve

-150

-100

-50

0

50

100

150

200

Bio

mas

s

EMn

Bio

mas

s

EMn

Bio

mas

s

EMn

Bio

mas

s

EMn

Bio

mas

s

EMn

Bio

mas

s

EMn

C-inorg C-org C-inorg P-org C-inorg P-inorg

C-org P-org C-org P-inorg

Imp

act

of

org

anic

PO

42

-an

d C

u2

+ o

n b

iom

ass

and

m

are

nn

ine

pro

du

ctiv

itie

s (%

)

Porg Cu

Page 153: Par « Rémi NGHIEM-XUAN

127

a gain of 94.0% and 122.9% in EMn productivity when combined with an inorganic and

an organic source of carbon, respectively.

Regarding EMn production, two typical media can be considered. C-inorg P-org

medium offers a final EMn concentration of 18.0 ± 0.2 mg L-1 compared to 11.1 ± 3.3 mg

L-1 for the C-inorg P-inorg medium. For the former, a small quantity of biomass was

produced whilst producing a large quantity of EMn, whereas for the latter, a large amount

of biomass was produced but with a small EMn specific productivity.

Concerning the impact of Cu on biomass and EMn productivity, Figure 3 confirms

the previous hypothesis on completely inorganic medium. Copper does not significantly

impact biomass productivity when an inorganic P source is used (C-inorg P-inorg and C-

org P-inorg) in contrast with an organic P source (C-inorg P-org and C-org P-org).

Regarding marennine production, the correlation of the presence of copper with

the resulting productivity is unclear. Copper was found to only stimulate EMn

productivity for a completely inorganic medium (carbon and phosphorus) with 20.4%

more EMn. However, the ES1/3 medium induced a loss of 87.2% EMn productivity when

copper was added, whereas (C-org P-org Cu) medium showed a constant specific EMn

production (Fig. 2 and 3).

Because organic sources were not found to be mandatory for H. oslearia growth, a

fully mineral medium was retained as a basis for the remaining part of this study. At large

scale, a medium such as this will indeed limit bacterial contamination.

4.3 Determination of inorganic C and P source concentration in natural

seawater culture medium

A stoichiometry analysis based on biomass elemental composition (C, H O, N, P, S,

Si) has been carried out with the aim of defining a medium that allows a biomass

concentration of 1 g L-1 to be achieved without macronutrient limitation [33,34]. Failing

to obtain significant biomass for analysis, the elemental composition of H. ostrearia has

been estimated based on the elemental composition of Chlorella vulgaris, which was

grown separately for the purpose (data not shown). Contrary to the Chlorophyceae, H.

ostrearia is a member of the Bacillariophyceae and needs silicon for frustule synthesis.

According to Turpin et al. [35], the macronutrient mass ratio C:N:P:Si, for H. ostrearia

Page 154: Par « Rémi NGHIEM-XUAN

128

growth is 106:16:1:16. The cell, therefore, needs the same mass of silica as nitrogen. Based

on this result, and the molecular weight of Si (28 g mol-1) and N (14 g mol-1), the following

molar elemental composition formula of H. ostrearia is proposed:

C3.805H4.110O1.359N0.688P0.081S0.024Si0.344

The biomass molecular weight (Mbiomass=24.72 g mol-1) and this formula allow us

to theoretically determine the C, N, P, S, Si concentration needed to produce 1g L-1 of

biomass (for example, for 1 g of biomass, H. ostrearia needs 7 mmol of nitrogen). Based

on this analysis, phosphorus and carbon concentrations in the ES1/3 medium were found

to be underestimated compared to the nitrogen and silica available in the medium. The

composition of the medium was, therefore, adjusted to propose a composition more

equilibrated in N and P. By applying a 1.5 coefficient to ensure 1 g L-1 biomass was

achieved without limitation, this led to 25-fold and 100-fold increases in N and P,

respectively (noted 25N100P) (Table 1).

In the first instance, micronutrient concentrations were considered to be in excess

in the medium, and their original concentrations as defined in the ES 1/3 medium were,

therefore, kept, but in completely inorganic source. However, the silica concentration

(Table 1) was found to be impossible to achieve without the precipitation phenomenon.

According to Fournier and Rowe (1977), amorphous silica solubility is restricted to 100 –

150 mg L-1 (1.7 – 2.7 M) at 25°C, which could decrease to 1-7 mg L-1 (1.7 x 10-5 to 1.2 x 10-

4 M) with the presence of calcium, aluminium and iron cations. However, it may increase

with dissolved inorganic matter. Because of this very low solubility, it was decided to

supply silica separately to the photobioreactor experiments.

A new set of experiments in flat flasks was designed to validate the new media

proposed (Fig. 4), namely NP medium (which corresponds to a fully mineral medium but

with the same N and P molar concentration as the previous ES1/3 medium shown in Table

1, but obtained from natural seawater with various enrichments in N and P up to

25N100P). Note that the period of growth for this experiment did not exceed 8 days.

Page 155: Par « Rémi NGHIEM-XUAN

129

Fig. 4: Cell, EMn and specific EMn volumic productivities of Haslea ostrearia for various media (ES 1/3, artificial seawater (ASW) and ES 1/3 based medium with completely inorganic C and P (NP)). P EMn specific

and Pv are, respectively, the specific EMn productivity and the volumetric productivity. Data is mean ±95% CI for n=3.

According to Figure 4, the NP medium was found to be much more efficient than

ES1/3 in terms of both biomass and EMn. Although the ES1/3 medium allowed the best

EMn specific productivity (Fig. 2 and 3), 6N24P allowed higher EMn concentration with

1.32 mg L-1 compared to 0.17 mg L-1 for ES1/3. The ASW medium allowed cell growth

higher than the ES1/3 medium but was still inferior to NP, with respectively 0.9 x 107cell

L-1 d-1 and 1.8 x 107cell L-1 d-1.

The progressive increase in concentration of N and P resulted in an increase of

biomass up to the 6N24P enrichment. This medium composition showed the best biomass

and EMn production, of respectively 3.0 x 107 cell L-1 d-1 and 1.3 mg L-1 d-1. With higher

enrichment (9N36P and more), therefore, a significant decrease in EMn and biomass

productivity was observed, and the 18N72P medium ultimately led to cell death. These

results could be explained by the silica and carbon concentration, which was the same for

all media but not increased proportionally. The rationale was that Si concentration limited

the strain growth at 3.0 x 107 cell L-1 d-1 (see below). A significant precipitation was

observed from the 12N48P concentrations. Note that 3.0 x 107 cell L-1 d-1 also represents

the maximum value of biomass production that is currently reported in the literature

0

0,2

0,4

0,6

0,8

1

1,2

1,4

1,6

0

0,5

1

1,5

2

2,5

3

3,5

ES1/3 NP 6N24P 9N36P 12N48P 18N72P 25N100P ASW

PV

(mgE

Mn

L-1

d-1

)P

EM

n s

pec

ific

(m

g 1

06ce

ll-1

d-1

)

PV

(10

7ce

ll L-1

d-1

)

Medium

Pv Biomasse Pv EMn Pspecifique EMn

Page 156: Par « Rémi NGHIEM-XUAN

130

(Rossignol et al., 1999). Medium stability could, thus, explain the actual low H. ostrearia

biomass production, as precipitates are not biologically available for cell growth.

The 6N24P medium, which led to the best productivity, was used as a basis for

further investigation of H. ostrearia macronutrient needs in PBR (see final section).

4.4 Component interaction and medium precipitation

Nguyen et al. (2013, 2014) explained the precipitation/flocculation phenomenon

of a seawater-based medium by the interaction between high concentrations of Ca2+, PO43-

and Mg2+ ions. By using Visual MINTEQ V3 software (KTH, Sweden), which enables

prediction of dissolution equilibrium, the percentage of chemical species of

macronutrients (N, P, Si, C) was investigated for the 25N100P medium at different pH

levels and at a fixed temperature (22°C) (Fig. 5).

Dissolution equilibrium predictions confirmed the interaction of the Ca2+, PO43-

and Mg2+ ions. The seawater endogenous ions, Ca2+ and Mg2+, were found to react with the

macronutrient enrichment ions in the culture medium. In addition, the extent of

precipitation increased with pH value. The magnesium precipitated with around 0.8% of

carbon and 20% of phosphorus, as well as with the calcium with 0.4% of carbon and 3.4%

of phosphorus at pH 7.8. Phosphorus precipitation with calcium and magnesium was

emphasized, with an important loss of 23.35% of the phosphate even at pH 7.8. Note that

no significant precipitation was predicted with the nitrogen source.

Page 157: Par « Rémi NGHIEM-XUAN

131

Fig. 5: Simulation of precipitated molecules (%) in the 25N100P medium for 3 pH values (Visual MINTEQ software)

Unfortunately, silica interaction could not be predicted with Visual MINTEQ, as the

software was set up for freshwater analysis. However, silica interaction in seawater is a

common matter for diatom culture and has been studied by Vollast et al. (1968). Their

results confirm the precipitation of hydrated magnesium silica, called synthetic sepiolite,

by the addition of metasilicate in seawater (eq. 1):

Equation 1

2𝑀𝑔2+ + 3𝑆𝑖𝑂2𝑎𝑞 + (𝑛 + 2)𝐻2 𝑂 ↔ 𝑀𝑔2𝑆𝑖3𝑂8(𝐻2𝑂)𝑛 + 4𝐻+

This equilibrated equation finds the solubility of amorphous silica to be 115 ppm,

as confirmed by Morey et al. (1995) with the possibility of precipitation of other nutrients

dependent on variations in temperature and pH.

The analysis shows that the precipitation phenomenon in natural seawater could

represent a significant limitation for H. ostrearia cultivation due to the high content of

both Ca2+ and Mg2+, as it limits medium enrichment and, therefore, the achievable biomass

Page 158: Par « Rémi NGHIEM-XUAN

132

concentration and productivity. Reducing calcium and magnesium, as a possible solution

to prevent precipitation, was investigated by designing an artificial medium.

4.5 Study of the impact of micronutrients on biomass and marennine

production in artificial seawater medium

Based on the previous study (Fig. 4), it was assumed that a medium composition of

nitrogen and phosphorus higher than 6N24P (inorganic) could be limited by silica

concentration. The 6N24P (inorganic) composition was, therefore, chosen for

micronutrient addition tests to prevent Si limitation during growth in artificial seawater.

The design for the experiment, which consisted of a large set of 180 flat flasks (Fig. 6),

provided useful information on the importance of each micronutrient for the optimization

of both biomass and EMn production in artificial medium conditions.

Page 159: Par « Rémi NGHIEM-XUAN

133

Fig. 6: Biomass [X] (bars) and EMn (circles) volumetric productivities (Pv) according to different macro and micro nutrient concentrations in artificial seawater. Data is mean ±95% CI for n=3

-0,5

0

0,5

1

1,5

2

2,5

0

5

10

15

20

25

30

35

40

45

4.5

x 1

0-5

4.5

x 1

0-4

4.5

x 1

0-3

4.5

x 1

0-2

8.4

x 1

0-5

8.4

x 1

0-4

8.4

x 1

0-3

8.4

x 1

0-2

1.0

x 1

0-5

1.0

x 1

0-4

1.0

x 1

0-3

1.0

x 1

0-2

5.8

x 1

0-9

5.8

x 1

0-8

5.8

x 1

0-7

5.8

x 1

0-6

3.9

x 1

0-9

3.9

x 1

0-8

3.9

x 1

0-7

3.9

x 1

0-6

6.8

x 1

0-6

6.8

x 1

0-5

6.8

x 1

0-4

6.8

x 1

0-3

1.2

x 1

0-6

1.2

x 1

0-5

1.2

x 1

0-4

8.1

x 1

0-3

9.0

x 1

0-3

1.6

x 1

0-2

9.2

x 1

0-2

5.3

x 1

0-3

5.3

x 1

0-2

5.3

x 1

0-1

2.3

x 1

0-7

2.3

x 1

0-6

2.3

x 1

0-5

2.3

x 1

0-4

6.6

x 1

0-9

5.4

x 1

0-8

5.4

x 1

0-7

5.4

x 1

0-6

3.4

x 1

0-1

0

3.4

x 1

0-9

3.4

x 1

0-8

3.4

x 1

0-7

2.5

x 1

0-3

2.5

x 1

0-2

2.5

x 1

0-1

6.5

x 1

0-1

0

6.5

x 1

0-9

6.5

x 1

0-8

6.5

x 1

0-7

9.0

x 1

0-5

9.0

x 1

0-4

9.0

x 1

0-3

3.0

x 1

0-8

3.0

x 1

0-7

3.0

x 1

0-6

3.0

x 1

0-5

B Br Ca Co Cu F Fe K Mg Mn Mo Ni S Se Sr Zn

PV

EMn

(m

g L-1

d-1

)

PV

bio

mas

s [X

] (1

06 c

ell

L-1d

-1)

Nutrients concentration (mol L-1)

B [X] PV Br [X] PV Ca [X] PV Co [X] PV Cu [X] PV F [X] PV Fe [X] PV K [X] PV Mg [X] PV Mn [X] PV Mo [X] PV

Ni [X] PV S [X] PV Se [X] PV Sr [X] PV Zn [X] PV B [EMn] PV Br [EMn] PV Ca [EMn] PV Co [EMn] PV Cu [EMn] PV F [EMn] PV

Fe [EMn] PV K [EMn] PV Mg [EMn] PV Mn [EMn] PV Mo [EMn] PV Ni [EMn] PV S [EMn] PV Se [EMn] PV Sr [EMn] PV Zn [EMn] PV

Page 160: Par « Rémi NGHIEM-XUAN

134

The first objective was to decrease the magnesium and calcium concentration to

reduce the occurrence of precipitation/flocculation, which was found to lead to culture

death. The best productivity was obtained at a cell density of 35.0±2.2 x 106 cell L-1 d-1,

with a calcium concentration of ca. 1.0 x 10-3 M, which is 9.4 times less than in the ASW

medium. The best productivity for magnesium was found at 6.4±0.4 x 106cell L-1 d-1 for a

magnesium concentration of 5.3 x 10-2 M, which is the same concentration as the ASW

medium composition.

The second objective was to increase micronutrient concentration to ensure a high

productivity with increasing concentrations of nitrogen, phosphorus and silica, thus,

preventing growth limitation. The only micronutrient that led to the highest biomass

productivity of 25.8±4.0 x 106cell L-1 d-1 was iron, with a concentration of 1.2 x 10-4 M.

This concentration represents 10 times that in the ASW medium, and biomass

productivity increased 5-fold. Zinc also seemed to increase biomass and EMn production

when increased 100-fold, compared to ASW. In addition, 10 times less boron than in ASW

increased biomass productivity, reaching 6.9±0.9 x 106cell L-1 d-1. Finally, a 10-fold

increase in nickel nutrient (3.4 x 10-8 M) led to higher EMn and biomass productivity, with

0.61±0.04 mg L-1 d-1 and 7.3±0.4 x 106cell L-1 d-1, respectively. H. ostrearia seemed to

tolerate the ASW range of concentrations for the other nutrients: bromine, cobalt, copper,

fluorine, magnesium (as previously presented), manganese, molybdenum, potassium,

sulfur, selenium and strontium (T.Test with p>0.05). We can note that conclusions were

similar for both biomass and EMn, except for the molybdenum test, which seemed to show

an inverse correlation. This was not investigated further in the present work.

These results enabled the design of the new artificial seawater NX medium based

on the highest productivities previously obtained (Fig. 6).

4.6 Validation of the artificial seawater NX medium in a photobioreactor

Following verification that there was no visual precipitation, three recipes for

culture medium were tested in a 1 L Airlift photobioreactor. The same conditions of light

(300 µmol m-2 s-1), temperature (22°C), pH (7.8) and cell inoculation concentration were

applied. However, to definitely prevent silica and carbonate precipitation in the artificial

medium, the concentration of carbon and silicon added was equilibrated with the

nitrogen and phosphorus concentration following the same approach based on

stoichiometric growth, leading to 25N100P6Si media with a 10 mM of dissolved carbon

Page 161: Par « Rémi NGHIEM-XUAN

135

concentration maintained constant for the entire culture duration with CO2 addition via

the gas phase. Figure 7 presents the results in terms of cell density and EMn concentration

for batch cultures of 22 days.

Fig. 7: Comparison of Haslea ostrearia cell density culture in new NX-25N100P6Si medium with two natural seawater-based media. Data is mean ±95% CI for n=3.

Fig. 8: Comparison of EMn concentration from Haslea ostrearia culture in new NX-25N100P6Si medium with two natural seawater-based media. Data is mean ±95% CI for n=3

The three batches represented in Figures 7 and 8 reveal the efficiency of the new

artificial seawater culture medium NX. In terms of biomass, the ES1/3 medium limited

0,0E+00

1,0E+08

2,0E+08

3,0E+08

4,0E+08

5,0E+08

6,0E+08

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

Cel

l den

sity

(ce

ll L-1

)

Time (days)

Cell density

Artificiel NX-25N100P6Si Cell Naturel 25N100P6Si Cell Naturel ES1/3 Cell

0

2

4

6

8

10

12

14

16

18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

EMn

co

nce

ntr

atio

n (

mg

L-1)

Time (days)

EMn

Artificiel NX-25N100P6Si EMn Naturel 25N100P6Si EMn Naturel ES1/3 EMn

Page 162: Par « Rémi NGHIEM-XUAN

136

cell concentrations at 7.9 x 107 cell L-1, whereas in the NX medium cultures reached 4.9 x

108 cell L-1 (a 6.2-fold increase). The culture in natural seawater 25N100P6Si quickly

decreased after 4 days, containing a whitish mineral precipitation. In terms of

extracellular marennine, ES1/3 medium reached 8.6 mg L-1, while NX allowed the culture

to reach 15.7 mg L-1. Both NX-25N100P6SFi and natural seawater 25N100P6Si media

received 1.4 g L-1 of sodium bicarbonate, 1.2 x 10-3 M of silica and 1.2 x 10-3 M of

phosphorus at the beginning of the culture. As a result, better stability was observed for

the artificial seawater NX than for natural seawater, with negligible precipitation

phenomena under the same conditions.

5 Discussion

Our results demonstrate that H. ostrearia needs either a completely organic or

completely inorganic medium to achieve the highest biomass and EMn concentrations.

Using an organic medium could be an opportunity to cultivate H. ostrearia and to decrease

the risk of culture medium precipitation, for example. According to Andersen (2005), Na2-

β-glycerophosphate from Provasoli ES medium (Provasoli, 1968) could be used as a

substrate by bacteria because of their heterotrophic nature. H. ostrearia has formerly

been suspected of being a mixotrophic strain, able to use different (organic and inorganic)

forms of nutrients (Rech, 2004). However, no proof has been brought to affirm the

mixotrophic nor the heterotrophic growth of the strain for EMn production purpose,

although an organic carbon source could also be used by H. ostrearia, being found in its

natural growing environment (Zhan et al., 2017). The results from this study indicate that

the culture of H. ostrearia in a medium made with organic phosphorus and inorganic

carbon may hypothetically influence H. ostrearia growth (Fig. 2) but also EMn production

per cell (Fig. 2). Both media containing organic phosphorus and carbon sources, as well

as inorganic phosphorus and carbon sources, lead to important yield of EMn and biomass.

This could indicate a mixotrophic metabolism, which is known to provide a better yield of

energy production than autotrophic metabolism (Li et al., 2014; Bouarab et al., 2002),

increasing then the biomass yield and the synthesis of complex compounds, such as

marennine. Note that in addition to modifying the overall physiology of H. ostrearia, using

an organic medium could increase the production costs.

Page 163: Par « Rémi NGHIEM-XUAN

137

If it is assumed that Cu could inhibit respiration and favor photosynthesis process,

the hypothesis from the results (Fig. 2 and 3) is that H. ostrearia could display

heterotrophy or mixotrophy in the presence of organic phosphorus. EMn overproduction

could, therefore, be the result of mixotrophy/heterotrophy metabolism or of a stress

resulting from the organic source of phosphorus in the medium. The impact of copper on

marennine production has already been observed by Minier et al. (1998), who showed

that copper toxicity varies according to the physiological stage of the microalgae,

especially in relation to EMn synthesis. Marennine production could, therefore, play an

important role in copper bio-availability in oyster ponds. It remains difficult to argue

whether H. ostrearia is able to consume organic matter or not, although the presence of

organic carbon seems to positively stimulate its growth. However, this study confirms the

relevant impact of the phosphorus source on this species. H. ostrearia is able to produce

significant biomass with completely organic or completely inorganic sources of carbon

and phosphorus. However, under these conditions, H. ostrearia seems to favor the

production of biomass rather than marennine.

With a mineral medium, the stoichiometric approach based on the elementary

composition of biomass (CHONPSSi method) allowed the macronutrient composition to

be equilibrated, which sustained high H. ostrearia growth. Although the ES1/3 medium

showed higher specific EMn productivity (0.51±0.04 mg 107cell-1 d-1) than with the

inorganic medium (0.07±0.01 mg 107cell-1 d-1) that was studied, the equilibrated medium

6N24P was found to produce more biomass and EMn (3.04±0.04 x 107 cell L-1 d-1 and

1.31±0.07 mg L-1 d1, respectively). However, in the case of this inorganic medium, a severe

precipitation phenomenon was observed, which revealed significant limitation for scaling

up H. ostrearia cultivation and confirmed possible interaction of the endogenous Ca2+ and

Mg2+ ions from seawater with added PO42-, HCO3- and SiO32- from the enrichment solution.

In order to increase the concentration of different nutrient sources in the medium

while preventing precipitation, a new artificial medium adapted to H. ostrearia was

designed. Following a detailed study of the nutrient impact on both biomass and EMn

productivity, calcium concentration was found to be decreased 10-fold (1.0 x 10-3 M),

leading to higher productivities (35.0±2.2 x 106 cell L-1 d-1 and 2.2±0.2 mgEMn L-1 d-1).

However, a higher decrease in calcium concentration led to lower productivities

(12.0±1.6 x 106 cell L-1 d-1 and 1.3±0.2 mgEMn L-1 d-1). As a major cation in cell culture with

Page 164: Par « Rémi NGHIEM-XUAN

138

magnesium and sodium, calcium is a required nutrient for eukaryotic cells (Merchant and

Helmann, 2014). Decreasing the calcium concentration does not necessarily mean that

less calcium is needed for growth, but it was found to definitely decrease the precipitation

phenomenon. However, H. ostrearia seems to need magnesium in the same concentration

as in ASW (5.3 x 10-2 M), since decreasing the magnesium concentration led to lower

productivity than for ASW Mg concentration (Fig. 6).

Surprisingly, by increasing the iron concentration 10-fold compared to the ASW

medium (1.2 x 10-4 M), a 5-fold increase in biomass productivity was observed (25.8±4.0

x 106 cell L-1 d-1 and 0.8±0.1 mgEMn L-1 d-1). This could be explained by the importance of

iron in the iron-sulfur cluster and enzymes structure, present in ferredoxin and

cytochromes, which are responsible for photosynthesis (Merchant and Helmann, 2014).

However, it should be noted that this iron concentration was the highest tested in this

study, since precipitation appeared over 1.2 x 10-4 M. Thus, H. ostrearia might tolerate a

higher iron concentration.

Zinc testing also showed better productivities when 100 times more zinc was

added than in the ASW medium. Zinc plays a vital role in maintaining the integrity of

ribosomes, as a cofactor for enzyme and protein folding (Praske and Plocke, 1971;

Merchant and Helmann, 2014), and seems to be a major requirement for H. ostrearia.

Nickel also yields better productivities when increased by a factor of 10. However,

nickel ion is used as a cofactor by different enzymes, mainly as a substitute for other

metals in phytoplankton (Merchant and Helmann, 2014). These results could also be

explained by a lack of other nutrients (Fe, Zn, Mg).

It was also shown that the other micronutrients were present in sufficient

concentrations in the ASW medium, and these concentrations were the same in the NX

medium. However, compared to other nutrients, increasing molybdenum concentrations

showed better EMn production and lower biomass production than in low concentrations.

No hypothesis could be confirmed, but according to Merchant and Helmann (2014), Mo is

responsible for the activity of nitrogenase and some other enzymes as a cofactor, and low

concentrations may be stressful for marennine production.

After the new artificial medium NX recipe was designed, its efficiency was tested

in batch mode using photobioreactors. NX medium achieved concentrations of 4.9 x 108

Page 165: Par « Rémi NGHIEM-XUAN

139

cell L-1 (±1.1 x 104) and 15.7±0.1 mg L-1 EMn using a conventional agitated

photobioreactor, compared to 8.0 x 107 cell L-1 (±8.8 x 103) and 8.62±0.04 mg L-1 EMn for

the ES1/3 medium. Biomass production was, therefore, increased 5.7-fold and marennine

yield was almost twice as high. This emphasizes the interest of the NX medium for

growing H. ostrearia. However, it was observed that H. ostrearia started to overproduce

EMn when the biomass reached a maximum concentration of 4.9 x 108 cell L-1 (around day

9). This phenomenon could be attributed to cell lysis due to nutrient

deficiency/limitation, and merits further investigation. As a benthic diatom, H. ostrearia

has a particular physiology that could be influenced when maintained in suspension in a

PBR, as in this study. Marennine, which presents a significant bioactivity (Gastineau et al.,

2012), could also inhibit the growth of H. ostrearia cells.

It should be noted that the NX medium did not allow 1 g L-1 of biomass (equivalent

to ca. 8.0 x 108 cells L-1) to be achieved. Since this limit was obtained through

stoichiometric analysis by setting macronutrient requirements, it could be explained by a

possible limitation of certain micronutrients. But considering the previously-described

design strategy, this would be surprising as they were provided in excess. A particular

physiological response could be the explanation: overproduction of EMn only seems to

appear from day 9 (Fig. 7), which also corresponds with higher cell concentration. EMn

concentration, therefore, only increases when biomass starts to decrease. EMn

overproduction or accumulation could be due to cell death, or conversely cell death could

be due to EMn overproduction, as EMn is also suspected as presenting an inhibitory effect

for growth (Rossignol, 1999; Pouvreau et al., 2007; Gastineau et al., 2014). This

phenomenon could possibly have occurred here, given the large EMn concentration that

was achieved, but investigation of the possible inhibitory effect of EMn on H. ostrearia

growth was not within the scope of this study and remains to be further investigated.

6 Conclusion

An artificial growth medium was designed for the cultivation of H. ostrearia in

PBRs. It enables the growth of H. ostrearia to be stimulated with equilibrated nutrient

composition, while preventing the formation of precipitate under normal growth

conditions. This allows the largest reported productivity values, to our knowledge, to be

achieved for the production of both biomass and marennine.

Page 166: Par « Rémi NGHIEM-XUAN

140

Additionally, the NX medium combined with the airlift PBR was found to be

effective for growing H. ostrearia cells in suspension. This would be a great advantage in

comparison with immobilized-cell PBRs, which could also be used for growing these

benthic species, but would certainly increase the precipitation phenomenon due to low

mass transfer without a stirring system. This condition leads to a pH and temperature

gradient within the PBR, which is closely linked to the calcium and magnesium

precipitation phenomenon (Pulz and Scheibenbogen 1998; Hochfeld, 2005; Gabelle et al.,

2012).

The stability of the NX culture medium remains to be optimized, as precipitation

occurred at high pH and temperature levels (data not shown). For exploitation under

industrial conditions, pH or temperature regulation should be efficient and the medium

composition could be simplified by reducing or omitting costly elements. The same

procedure (i.e, testing the effect of each micronutrient on growth) as described in this

study could be applied for the purpose.

In this study, H. ostrearia was able to grow on both carbon and phosphorus organic

sources. This could be further investigated, as mixotrophic culture remains a possible

solution for optimizing biomass and EMn production, but may also reduce the

precipitation phenomenon. Finally, working without natural seawater will facilitate

research by preventing precipitates and, therefore, controlling the nutrient intake by cells.

Seawater would be preferable on a very large scale (for economic reasons), but would

limit biomass and marennine production due to the endogenous calcium and magnesium

concentration, unless adequate medium renewal or an enrichment strategy, such as fed-

batch, were applied.

Page 167: Par « Rémi NGHIEM-XUAN

141

7 Supplemental information

Artificial seawater culture medium (NX) for PBR culture

NX culture medium based on artificial seawater is described below (Tables 3 to

11). The first step in its synthesis is production of the salt solution. For all salts given in

Table 3 except NaHCO3, NaH2PO4, Na2SiO3, each salt must be dissolved individually (in any

order) until it is completely dissolved. NaHCO3, NaH2PO4 and Na2SiO3 may precipitate

greatly, mainly at high temperature and pH levels, so these salts must be added after the

sterilization step (shown as * in Table 3). Additionally, these three molecules trigger

variations in pH, which can be lethal for algae, so they are added directly to the PBR one

by one to ensure pH-regulated conditions. The vitamin solution was also added after

sterilization to prevent denaturation by temperature (Gadient et al., 1992; Reddy and

Love, 1999). All the salts were mixed in a 1 L graduated flask with 400 mL of distilled

water completed at 1 L after dissolution/dilution. Note that some elements in Table 4

were dissolved in separated solutions. The Si solution was also isolated from the other

nutrients because of its high alkalinity.

Table 3: Salt solution composition for NX medium

Salt Solution (1 L)

Quantity used Concentration in Final

medium (M)

NaCl 21.21 g 3.63 x 10-1

Na2SO4 3.55 g 2.50 x 10-2

KCl 0.60 g 8.04 x 10-3

NaHCO3* 1.4 g 1.62 x 10-2

Br solution 10 mL 8.40 x 10-4

B solution 1 mL 4.50 x 10-5

F solution 1 mL 6.80 x 10-5

MgCl2,6H2O 10.77 g 5.30 x 10-2

CaCl2, 2H2O 0.15 g 1.00 x 10-3

Sr solution 10 mL 8.18 x 10-5

Page 168: Par « Rémi NGHIEM-XUAN

142

Si solution* 2.94 mL 1.17 x 10-3

NaNO3 0.85 g 9.95 x 10-3

P solution* 2.08 mL 1.27 x 10-3

Vitamin stock solution*

20 µL

Enrichment solution 7 mL

Table 4: Isolated solution for salt solutions realization

B solution (100 mL)

Quantity used

Concentration in Final medium (M)

H3BO3 0.28 g 4.50 x 10-5

Br solution (100 mL)

KBr 1.00 g 8.40 x 10-4

F solution (100 mL)

NaF 0.29 g 6.80 x 10-5

Sr solution (100 mL) SrCl2·6H2O

0.22 g 8.18 x 10-5

P solution (100 mL)

NaH2PO4 8.43 g 1.27 x 10-3

Si solution (100 mL)

Na2SiO3·5H2O 2.12 g 1.17 x 10-3

The enrichment solution (Table 5) is composed of metal nutrients and an isolated

iron solution. Tris base buffer could be added to this solution to facilitate pH regulation

during PBR culture. However, since the pH is regulated by CO2 injection in the PBR, the

buffer was not used during PBR culture.

Table 5: Enrichment solution composition for NX medium

Enrichment solution (1 L)

Quantity used Concentration in Final

medium (M)

Tris base 5 g 2.89 x 10-4

Page 169: Par « Rémi NGHIEM-XUAN

143

Iron solution 379.61 mL

Metal A stock solution 250 mL

Iron ion is highly reactive inside a neutral and alkaline aqueous medium and

precipitates immediately. Ethylenediaminetetraacetic acid (EDTA) is a chelating agent

able to stabilize iron ion in the culture media (Paolieri, 2017). Na2EDTA was, therefore,

dissolved in the solution first, for this purpose. When it was completely dissolved, the rest

of the iron forms were dissolved one by one. Both iron forms are based on F/2 and ASW

iron forms with new concentrations based on the results in Fig. 5 (table 6). Note that a

higher stability of iron and metals than the one in this recipe is possible with anhydrous

EDTA.

Table 6: Iron solution composition for NX medium.

Iron solution (500 mL)

Quantity used Concentration in Final

medium (M)

Na2EDTA·2H2O 6.97 g 1.37 x 10-4

FeCl3·6H2O 5.07 g 9.96 x 10-5

FeNH4-Citrate 1.03 g 2.04 x 10-5

The metal A stock solution (Tables 7 and 8) is composed of all the metal ions

(except iron) necessary for H. ostrearia culture (based on ASW and F/2 medium). This

solution is also stabilized with Na2EDTA chelating agent to avoid possible metal

precipitation between metal A ions and also between metal A ions and iron ion, when iron

solution and metal A solutions are mixed within the enrichment solution (Andersen,

2005).

Table 7: Metal A stock solution composition for NX medium

Metal A stock solution (500 mL)

Quantity used Concentration in Final

medium (M)

Na2EDTA·2H2O 16.08 g 1.51 x 10-4

Page 170: Par « Rémi NGHIEM-XUAN

144

Zn solution 100 mL 3.00 x 10-5

Co solution 10 mL 5.66 x 10-8

Metal C stock solution 100 mL

Se solution 58 µL 1.00 x 10-9

Mn solution 10 mL 2.42 x 10-6

Table 8: Isolated solution for Metal A solution realization

Zn solution (100 mL)

Quantity used Concentration in Final

medium (M)

ZnSO4·7H2O 2.46 g 3.00 x 10-5

Co solution (100m mL)

CoCl2·6H2O 0.04 g 5.66 x 10-8

Mn solution (100 mL)

MnSO4·H2O 1.17 g 2.42 x 10-6

Se solution (50 mL)

Na2SeO3 0.04 g 1.00 x 10-9

The metal B stock solution (Table 9) is composed of the lowest micronutrient

concentration of the NX medium, which implies multiple dilutions. This solution kept at

3°C was stable and started to precipitate only after 3 weeks. However, the metal ions do

not precipitate when dissolved separately (Table 10). Na2EDTA was not added to the

metal B solution in order to limit the addition of sodium to the NX medium, but it is still

possible to stabilize it with anhydrous EDTA (Andersen, 2005).

Table 9: Metal B stock solution composition for NX medium

Metal B stock solution (1 L)

Quantity used Concentration in Final

medium (M)

CuSO4·5H2O 10 mL 3.90 x 10-8

Na2MoO4·2H2O 10 mL 6.57 x 10-9

Page 171: Par « Rémi NGHIEM-XUAN

145

NiSO4·7H2O 10 mL 3.40 x 10-8

Table 10: Isolated solution for establishing Metal B solution

Cu solution (100 mL)

Quantity used Concentration in Final

medium (M)

CuSO4·5H2O 0.28 g 3.90 x 10-8

Mo solution (100m mL)

Na2MoO4·2H2O 0.05 g 6.57 x 10-9

Ni solution (100m mL)

NiSO4·7H2O 0.27 g 3.40 x 10-8

Finally, the composition of the vitamin stock solution is based on the vitamin

composition of F/2 and ASW (Table 11). After dissolution, the vitamin solution is filtered

at 0.2 µm inside a sterile 2 L bottle and kept at 3°C, covered with aluminum foil to avoid

light degradation of the vitamins. This solution is added under sterile conditions in the

sterile bottle of medium, after sterilization.

Table 11: Vitamins stock solution composition for NX medium

Vitamins stock solution (2 L)

Quantity used Concentration in Final

medium (M)

Thiamine HCl (C12H17N4OS+) 1.00 g 2.97 x 10-7

Biotine (C10H16N2O3S) 0.01 g 4.09 x 10-9

B12 Cyanocobalamine (C63H88CoN14O14P)

0.02 g 1.47 x 10-9

Page 172: Par « Rémi NGHIEM-XUAN

146

8 References

Andersen, R. A. (2005). Algal culturing techniques. Edited by Robert A. Andersen; Elsevier

Academic Press. 578 p.

Belay A., Kato T., Ota Y., (1996). Spirulina (Arthrospira): potential application as an animal

feed supplement. Journal of Applied Phycology, (1990), 303-311.

Berges J.A., Franklin D.J., Harrison P.J., (2001) Evolution of an artificial seawater medium:

improvements in enriched seawater, artificial water over the last two decades. Journal

of Phycology 37:1138–1145.

Bouarab L., Loudiki M., Dauta A. (2002). Croissance en autotrophie et en mixotrophie de la

microalgue Micractinium pusillum Fres. Isolée d´un lagunage naturel : influence de la

lumière et de la température. Revue des sciences de l'eau, 15(1), 73–86.

doi:10.7202/705437ar.

Cardozo K. H.-M., Guaratini T., Barros M. P., Falcão V. R., Tonon A. P., Lopes N. P., Campos

S., Torres M. A., Souza A. O., Colepicolo P., Pinto E., (2006). "Metabolites from algae

with economical impact". Comparative Biochemistry and Physiology C. Elsevier Inc. 146

(1–2): 60–78. doi:10.1016/j.cbpc.2006.05.007.

Cid A., Torres E., & Abalde, J. (1995). Copper toxicity on the marine microalga Phaeodactylum

tricornutum: effects on photosynthesis related parameters. Aquatic Toxicology, 1(94).

Domek M. J., Lechevallier M. W., Cameron S. C., McFeters G. A., (1984). "Evidence for the

role of copper in the injury process of coliform bacteria in drinking water" (PDF). Applied

and Environmental Microbiology. 48 (2): 289–93.

Domek M. J., Robbins J. E., Anderson M. E., McFeters G. A., (1987). Metabolism of

Escherichia coli injured by copper. Canadian journal of microbiology. 33 (1): 57–62.

Festa R. A., & Thiele D. J., (2011). Copper: An essential metal in biology. Current Biology

21(21), R877–R883. http://doi.org/10.1016/j.cub.2011.09.040.

Fournier R. O., & Rowe J. J., (1977). The solubility of amorphous silica in water at high

temperatures and high pressure. American Mineralogist, 62, 1052-1056.

Gabelle J.C., Jourdier E., Licht R., Ben Chaabane F., Henaut I., Morchain J., Augier F., (2012).

Impact of rheology on the mass transfer coefficient during the growth phase of

Trichoderma reesei in stirred bioreactors. Chemistry Engineering Science, 75, 408-417.

DOI:1 0.1016/j.ces.2012.03.053.

Gadient M., Fenster R., and Latscha T., 1992. Vitamin stability in aquaculture feeds. Fish

Farmer, January/February: pp. 27-28.

Gagneux-Moreaux S., Moreau C., Gonzalez J.-L., & Cosson R. P., (2007). Diatom artificial

medium (DAM): a new artificial medium for the diatom Haslea ostrearia and other marine

Page 173: Par « Rémi NGHIEM-XUAN

147

microalgae. Journal of Applied Phycology, 19(5), 549–556.

http://doi.org/10.1007/s10811-007-9169-4.

Gaillon B., (1820). Des huîtres vertes, et des causes de cette coloration. Journal de physique,

de chimie, d’histoire Naturelle et des Arts. 91,222-225.

Gastineau R., Pouvreau J.-B., Hellio C., Moranc M., Gaudin P., Bourgougnon N., & Mouget

J.-L., (2012). Biological Activities of Purified Marennine, the Blue Pigment Responsible

for the Greening of Oysters. Journal of Agriculture Food and Chemistry. 60(14):3599-

605. http://doi.org/10.1021/jf205004x.

Grobbelaar J. U., (2003). Algal Nutrition – Mieral Nutrition. In Handbook of Microalgal

culture. Edited by Amos Richmond, 95-115.Blackwell Publishing Ltd.

Hochfeld W. L., (2005). Producing Biomolecular Substances with Fermenters, Bioreactors, and

Biomolecular Synthesizers, Boca Raton, FL, Taylor & Francis, ISBN 9781420021318.

Iwamoto H., (2007). Industrial production of microalgal cell-mass and secondary products —

major industrial species — Chlorella, p. 255–263. In Richmond, A. (ed.), Handbook of

microalgal culture. Blackwell, Oxford.

Kandilian R., Pruvost J., Legrand J., and Pilon l., (2014). “Influence of Light Absorption Rate

by Nannochloropsis Oculata on Triglyceride Production during Nitrogen Starvation.”

Bioresource Technology 163 (July): 308–19.

Lebeau, T., Gaudin P., Junter G.-A., Mignot L., & Robert J.-M., (2000). Continuous marennin

production by agar-entrapped Haslea ostrearia using a tubular photobioreactor with

internal illumination. Applied Microbiology Biotechnology, 54, 634–640.

Levy J., Angel B., Stauber J. L., Poon W. L., Simpson S. L., Cheng S. and Jolley D. F., (2008).

Uptake and internalization of copper by three marine microalgae: comparison of copper-

sensitive and copper-tolerant species. Aquatic Toxicology, 89, 82-93.

Li Y. R., Tsai W. T., Hsu Y. C., Xie M. Z., Chen J. J., (2014). Comparison of Autotrophic and

Mixotrophic Cultivation of Green Microalgal for Biodiesel Production. Energy Procedia.

52, 371-376. https://doi.org/10.1016/j.egypro.2014.07.088.

Merchant S. S., & Helmann J. D., (2014). Elemental Economy: microbial strategies for

optimizing growth in the face of nutrient limitation. Advance in Microbial Physiology. 60,

91-210. http://doi;org/10.1016/B978-0-12-398264-3.00002-41.

Merz C. R.., Main K. L., (2014). Microalgae (diatom) production – The aquaculture and biofuel

nexus. Oceans-St. John’s, 14-19 Sept. 2014. 10.1109/OCEANS.2014.7003242.

Miazek K., Iwanek W., Remacle C., Richel A. and Goffin D. (2015). Effect of Metals,

Metalloids and Metallic Nanoparticles on Microalgae Growth and Industrial Product

Biosynthesis: A Review. International Journal of Molecular Science, 16(10): 23929–

23969.

Page 174: Par « Rémi NGHIEM-XUAN

148

Minier C., Tutundjian R., Galgani F., & Rober J. M., (1998). Copper Tolerance in Haslea

ostrearia Assessed by Measurements of in vivo Esterase Activity. Marine Environmental

Reasearch, 46(l), 579–582.

Moreau J., (1968). Les facteurs de verdissement de l’huitre en claires: Le milieu

hydrobiologique et benthique et ses variations. Revue des travaux de l'Institut des pêches

maritimes. 32 (i), p. 369-386.

Morey G. W., & Fournier R. & Rowe J., (1995). The Solubility of Amorphous Silica at 25°C.

Journal of Geophysical Research, 69(10), 1995–2002.

Mouget J-L., Gastineau R., Davidovich O., Gaudin P., Davidovch NA (2009) Light is a key

factor in triggering sexual reproduction in the pennate diatom Haslea ostrearia. FEMS

Microbiology Ecology. 69:194-201.

Nguyen, T. D. P., (2013). Récolte de biomasse microalgale par floculation naturelle et procédés

membranaires (Thèse de doctorat ; Université de Nantes).

Nguyen, T. D. P., Frappart M., Jaouen P., Pruvost J., Bourseau P., (2014). Harvesting Chlorella

vulgaris by natural increase in pH: effect of medium composition. Environmental

Technology, 35, 11I: 1378-1388. https://doi.org/10.1080/09593330.2013.868531

Panis G., & Carreon J. R., (2016). Commercial astaxanthin production derived by green alga

Haematococcus pluvialis: A microalgae process model and a techno-economic assessment

all through production line. Algal Research, 18, 175–190.

http://doi.org/10.1016/j.algal.2016.06.007. [1] Spolaore P., Joannis-Cassan C., Duran E.,

Isambert A., Génie L. De, & Paris E. C., (2006). Commercial Applications of Microalgae,

Journal of Bioscience and Bioengineering. 101(2), 87–96.

http://doi.org/10.1263/jbb.101.87.

Paolieri M., (2017). "Ferdinand Münz: EDTA and 40 years of inventions". Bulletin for the

History of Chemistry. 42 (2): 133–140.

Pouvreau J.-B., Morançais E., Fleury F., Rosa P., Thion L., Cahingt B., Zal F., Fleury F., (2006).

Preliminary characterisation of the blue-green pigment “marennine” from the marine

tychopelagic diatom Haslea ostrearia (Gaillon/Bory) Simonsen. Journal of Applied

Phycology, 18, 757–767. http://doi.org/10.1007/s10811-006-9087-x.

Pouvreau J.-B., Housson E., Le Tallec L., Morançais M., Rincé Y., Fleurence J., Pondaven P.,

(2007). Growth inhibition of several marine diatom species induces by the shading effect

and allelopathic activity of marennine, a blue-green polyphenolic pigment of the diatom

Haslea ostrearia (Gaillon/Bory) Simonsen. Journal of Experimental Marine Biology and

Ecology, 352, 212–225. https://doi.org/10.1016/j.jembe.2007.07.011.

Praske J.A. and D.J. Plocke., (1971). A role of zinc in the structural integrity of the cytoplasmic

ribosomes of Euglena gracilis. Plant Physiology. 48 : 150-155.

Provasoli L., (1968). Media and prospects for the cultivation of marine algae. In: Watanabe A.,

and Hatorri A., eds. Cultures and Collection of Algae. Proc. U. S.-Japan Conference,

Hakone, Japanese society plant physiology, 63-75.

Page 175: Par « Rémi NGHIEM-XUAN

149

Pruvost J., Van Vooren G., Cogne G., Legrand J., (2009). Investigation of biomass and lipids

production with Neochloris oleoabundans in photobioreactor. Bioresource technology,

100 (23), 5988-5995.

Pruvost, J., Le Borgne, F., Artu, A., Cornet, J.-F., Legrand, J., (2016). Industrial

Photobioreactors and Scale-Up Concepts. in: Advances in Chemical Engineering -

Photobioreaction Engineering, Vol. 48, Elsevier, pp. 257-310.

Pulz O., and Scheibenbogen K., (1998). Photobioreactors: Design and Performance with

Respect to Light Energy Input. Bioprocess and Algae Reactor Technology, Apoptosis,

123–52. Advances in Biochemical Engineering Biotechnology 59. Springer Berlin

Heidelberg.

Rech M., (2004). Effets de l’éclairement visible et ultraviolet sur la croissance et la

photosynthèse de microalgues : incidences sur l’écophysiologie du phytoplancton des

claires ostréicoles (Thèse de doctorat ; Université du Maine).

Reddy M.B., and M. Love., (1999). The impact of food processing on the nutritional quality of

vitamins and minerals. Advances in Experimental Medicine and Biology. 459 :99-106.

Robert, J.-M. (1983). Fertilité des claires ostréicoles et verdissement: utilisation de l’azote par

les diatomées dominantes. (Thèse de doctorat ; Université de Nantes).

Robert J., Morançais M., Pradier E., Mouget J., & Tremblin G., (2002). Extraction and

quantitative analysis of the blue-green pigment “marennine” synthesized by the diatom

Haslea ostrearia. Journal of Applied Phycology, 14, (I)4, 299–305.

Rossignol, N. (1999). Procédé d’extraction et de séparation par membranes appliqués à la

production du pigment endo-et exocellulaire synthétisé par la diatomée Haslea ostrearia

Simonsen. Mise en œuvre d’un photobioréacteur à membranes à fonctionnement continu.

(Thèse de doctorat ; Université de Nantes).

Stoermer E. F., Smol P., (1999). The diatoms: Applications for the Environmental and Earth

Sciences, Journal of Bioscience and Bioengineering. 101(2), 87–96.

http://doi.org/10.1263/jbb.101.87.

Taleb A., Pruvost J., Legrand J., Marec H., Le-Gouic B., Mirabella B., Legeret B., (2015).

“Development and Validation of a Screening Procedure of Microalgae for Biodiesel

Production: Application to the Genus of Marine Microalgae Nannochloropsis.”

Bioresource Technology 177 (February): 224–32.

Turpin V., Robert J. M., Goulletquer P., (1999). Limiting nutrients of oyster pond seawaters in

the Marennes-Oléron region for Haslea ostrearia: applications to the mass production of

the diatom in mesocosm experiments. Aquatic Living Resources. 12 (5), 335-342.

Turpin V., J-M Robert, P Goulletquer, G. M. & P. R. (2001). Oyster greening by outdoor mass

culture of the diatom Haslea ostrearia Simonsen in enriched seawater. Aquaculture

Research, 32(801-809).

Page 176: Par « Rémi NGHIEM-XUAN

150

Veloso P., Reis A., Gouveia L., Fernandes H. L., Empis J. A., Novais J. M., (1991). Lipid

production by Phaeodactylum tricornutum. Bioresource technology, 38, 2-3I, 115-119.

https://doi.org/10.1016/0960-8524(91)90141-6.

Vollast R., Mackenzie F. T., Bricker O. P., (1968). Experimental precipitation and genesis of

sepiolite at earth-surface conditions. The American Mineralogist, 53.

Wells P. G., Lee K., Blaise C., (1998) Microscale testing in aquatic toxicology: Advances,

techniques and practice. Edited by Peter G. Wells, Kenneth Lee, Christian Blaise with the

assistance of Johanne Gauthier; Boca Raton: CRC Press. 679p.

Zhan J., Rong J., Wang Q., (2017). Mixotrophic cultivation, a preferable microalgae cultivation

mode for biomass/bioenergy production, and bioremediation, advances and prospect. International Journal of Hydrogen Energy. 42, 12(i), 8505-8517.

https://doi.org/10.1016/j.ijhydene.2016.12.021.

Page 177: Par « Rémi NGHIEM-XUAN

151

Chapter IV:

Investigation of Haslea ostrearia

cellular fragility and consequences

of stirring strategy in conventional

photobioreactor culture

R. Nghiem Xuan1,3, J. L. Mouget1, 2, J. Pruvost 3, V. Turpin1, P. Jaouen3

1 Université de Nantes, MMS, EA 2160, Faculté des Sciences et des Techniques, 2 Rue de la

Houssinière, BP, 92208, 44322 Nantes Cedex, France.

2 MMS (Mer Molécules Santé), Le Mans Université, Ave O. Messiaen, 72085 Le Mans, France.

3 GEPEA, Université de Nantes, CNRS, UMR6144, Bd de l'Université, CRTT - BP 406, 44602 Saint-

Nazaire Cedex, France.

Ce chapitre est rédigé sous forme de publication pour soumission dans un journal

scientifique.

Page 178: Par « Rémi NGHIEM-XUAN

152

1 Abstract:

Discovered by Gaillon with the greening oyster phenomenon in 1820, Haslea

ostrearia is a benthic diatom that produces and excretes a blue pigment called Marennine.

This molecule was studied for several years and represents great industrial potential, as

described by Gastineau et al. (2014). Nevertheless, culture of the diatom on an industrial

scale is still challenging. Like many others diatoms, one of the main difficulties is cell

sensitivity to hydrodynamic shear stress, for which cell immobilization technology is

usually applied.

The cellular fragility of H. ostrearia to hydrodynamic shear stress is investigated in

this study. It has been demonstrated that strain is not sensitive to shear rate or shear

stress in laminar flow, even at high shear rate (30,000 s-1) and shear stress (3,180 Pa).

However, cultivation in a torus PBR with turbulent flow conditions showed a serious

negative impact on biomass productivity, with 0.3 ± 0.2 g m-2 d-1 compared to 1.0 ± 0.2 g

m-2 d-1 for an unstirred PBR. As an alternative strategy, it was proposed to apply a limited

mixing period per day, and 8 h of stirring was found to give the best biomass productivity

results with 1.8 ± 0.2 gx m-2 d-1. This demonstrates the need to mix the culture for at least

a few hours per day to avoid too much sedimentation of the benthic strain, which causes

a decrease in productivity.

Finally, a detailed analysis of the results has shown that biomass productivity is

related to the resulting rate of photon absorption received by pelagic (i.e. suspended) cells

over a 24 h period, which was influenced by both mixing and the length of lit periods. For

moderate mixing conditions leading to an acceptable shear rate in turbulent flow, this

parameter was found to be useful in optimizing the mixed H. ostrearia culture.

Keywords: Haslea ostrearia, industrial production, hydrodynamics, shear stress, cell

fragility, light, photobioreactor culture.

Page 179: Par « Rémi NGHIEM-XUAN

153

2 Introduction

Mixing is known to be a relevant operating parameter for any microalgae culture

system (Hochfeld, 2005; Richmond and Hu, 2008; Gabelle et al., 2012). It allows

optimization of the culture conditions in terms of homogeneity, growth nutrient

concentration, pH and temperature, while also allowing cell suspension for good light

access conditions in bulk. It remains, however, a parameter difficult to optimize when

cultivating strains that are sensitive to hydrodynamic shear-stress (Contreras et al., 1998;

Garcia Camacho et al., 2007; Michiels et al., 2015; Wang and Lan, 2018).

Several technologies are available for the culture of microalgae (Spolaore et al.,

2006; Pruvost et al., 2006; Pruvost et al., 2011; Pruvost et al., 2012). As closed systems,

photobioreactors are known to offer better control of culture conditions, enabling

systematic optimization to obtain maximum growth performance in a given strain. This

approach has been previously demonstrated for commonly cultivated genders like

Chlorella, Nannochloropsis and Arthrospira. These strains were grown in various

photobioreactor technologies with different types of mixing, i.e. airlift principle or

mechanical mixing (Pruvost et al., 2006; Ndiaye et al., 2018). They can therefore be

considered as resistant to the range of hydrodynamic shear stress usually encountered in

conventional photobioreactors. This resistance greatly simplifies their cultivation, which

also opens up the possibility of production at industrial scale.

However, some microalgae such as dinoflagellates and diatoms are also known to

be more sensitive to shear stress (Sobczuk et al., 2006; Wang and Lan, 2018). This is the

case with the diatom Haslea ostrearia, which is considered to be a fragile microalga

(Robert, 1989; Rouillard, 1996). The culture of H. ostrearia in conventional

photobioreactors has been reported as challenging, no doubt due to its marked sensitivity

to shear stress (Robert, 1989). Rossignol et al. (1999) attest that this strain is sensitive to

recirculation in pumps, which also reduces the range of photobioreactors (PBRs) that can

be used.

As a diatom, H. ostrearia possesses a solid silica protective structure called frustule,

which is supposed to present low resistance to shear stress. This frustule is composed of

two parts, which separate during cell division (Nassiri et al., 1998), leading to additional

fragility in the cell.

Page 180: Par « Rémi NGHIEM-XUAN

154

As an alternative to a conventional PBR based on cell-suspended culture,

technologies based on immobilized cells, which protect cells from hydrodynamic shear

stress, can offer benefits for fragile species. For example, immerged membrane PBRs

(Vandanjon, 1997; Rossignol et al., 1999; Rossignol, 1999), cell immobilization in agar gel

(Lebeau et al., 2000; Rossignol et al., 2000), and unstirred open pond PBRs (Turpin et al.,

2001; Turcotte et al., 2016). For H. ostrearia, the main molecule of interest of which -

marennine - is exudate, the immobilized-cell culture principle can be associated with

separation processes such as a membrane to continuously recover the extracellular

marennine (EMn).

When grown in an immerged membrane PBR technology, a biomass productivity

of 2.45 ± 0.25 mg L-1 d-1 was obtained for H. ostrearia (Rossignol et al., 1999). This is the

best productivity reported in literature. Despite extensive engineering work, a lower

productivity is reported for open pond culture and agar immobilized cell PBRs with 0.34

± 0.04 mg L-1 d-1 and 0.50 ± 0.20 mg L-1 d-1 respectively (Lebeau et al., 2000; Turpin et al.,

2001). For conventional suspended-cell systems, airlift technology has been tested but no

growth observed (Jubeau, 2009).

Mixing impacts several culture parameters, such as mass and thermal transfers,

and also light access to flowing cells. Regarding the effect on fragile cells, this is often

described as the mechanical impact of hydrodynamic shear stress, or shear rate, on the

physical integrity of cells (Fadlallah et al., 1995; Mitsuhashi et al., 1995; Olmos et al., 2003;

Garcia Camacho et al., 2007; Tesson and Latz 2015; Mauer et al., 2016). Using an adequate

quantity to relate hydrodynamic stress to microorganism fragility remains controversial,

shear stress itself being a complex physical phenomenon that varies widely between

culture systems. From a hydrodynamics point of view, the mechanical effect resulting

from hydrodynamics is represented by both shear stress (denoted τ) and shear rate

(denoted γ̇). In fluids, the former represents the stress (Pa) produced by the force applied,

parallel or tangential to the surface studied, while the latter is the velocity gradient (s-1)

produced by the progressive shear deformation of the fluid. To the author’s knowledge,

no proof has been provided to confirm that one quantity is more related to cell fragility.

As a result, both are used in literature (Fadlallah et al., 1995; Mitsuhashi et al., 1995;

Olmos et al., 2003; Garcia Camacho et al., 2007; Tesson and Latz 2015; Mauer et al., 2016).

Page 181: Par « Rémi NGHIEM-XUAN

155

In addition, the mechanical effect on cells also depends on the shape of the cell and

its mechanical structure. A great diversity of microalgae is known to exist, from the

Chlorella gender, which presents spherical resistant cells, to complex diatom shapes with

silica cell walls. It is also reported that the mechanical resistance of a given strain depends

on its physiological state, which is highly variable (Wang and Lan, 2018). Concerning H.

ostrearia, its special fusiform shape, as well as its cell division process with frustule

separation, could explain the cell’s fragility.

This study is aimed at a better understanding of the response of H. ostrearia cells

to mixing, and their sensitivity. This will be conducted as a two-step approach. The effects

of hydrodynamic shear stress will first be investigated in homogeneous laminar flow (i.e.

rheometer experiments) and in turbulent flow (i.e. a mechanically stirred

photobioreactor). The combination of light access and mixing will then be analyzed by

cultivating H. ostrearia under various light and mixing conditions.

3 Materials and Methods

3.1 Inoculum and cultures for preliminary experiments in rheometer and

flasks

The diatom H. ostrearia NCC 495 and NCC 501 was obtained from MMS laboratory

and cultivated in natural seawater from Croisic (Loire-Atlantique, France), enriched with

Guillard F/2 medium.

NCC 495 was used to analyze the cell response in perfect laminar flow (rheometer

experiment). NCC 501 was used 2 years after the NCC 495 experiment, when we

succeeded in growing H. ostrearia in conventional photobioreactor conditions. The NCC

495 strain was lost in the meantime due to senescence.

Cultures of H. ostrearia were maintained in 250 mL Erlenmeyer flasks containing

100 mL medium with 10% dilution, after 4 - 5 days of culture in Provasoli ES1/3 medium

(Robert, 1983) which was remodified for exclusive use of complete inorganic sources

(Nghiem-Xuan, 2019). This medium is a natural seawater-based medium and water is

filtered at 0.7 µm then 0.45 µm before being used as a medium.

Page 182: Par « Rémi NGHIEM-XUAN

156

The light supplied was white fluorescent light (F36W/840 T8, Sylvania, China)

with a 14/10 photoperiod (denoted L/D) at 300 µmol m-2 s-1 in a culture chamber (SANYO

MLR-350 culture chamber, Richmond Scientific, UK) at 16°C. Fresh natural seawater was

provided by the Croisic Ocearium (Loire Atlantique, France).

In the case of experiments in the rheometer, all samples were taken from the same

Erlenmeyer flasks where cells were grown for 5 days until they reached the exponential

phase. The preliminary results of the mixing effect on H. ostrearia were obtained in stirred

Erlenmeyer flasks. Stirring was carried out using a magnetic stirrer in 500 ml Erlenmeyer

flasks in the same medium and same culture conditions as the inoculum culture (SANYO

MLR-350 culture chamber, Richmond Scientific, UK).

3.2 Measurement of cell viability

Cell viability was measured with a Nageotte counting chamber (Marienfeld, AQRA,

India) and a light microscope (AXIO Scope A1, Zeizz, Germany). The living cells were

counted before and after each physical treatment to calculate cell viability (%). Note that

staining to distinguish living and dead cells was found to be unnecessary due to the

specific shape of dying/dead H. ostrearia cells, as shown in Figure 9. All measurements

were made around 1 x 105 cells mL-1 in the early stage of the stationary step (maximal

concentration usually observed in an Erlenmeyer flask).

Page 183: Par « Rémi NGHIEM-XUAN

157

Fig. 9: Microscopic view of Haslea ostrearia. (A) is the contrast of two cells, cell (1) being alive and cell (2) being dead. (B and C) shows living cells in a bloc of cellular residues, with an assessment of cell viability by

chlorophyll fluorescence under UV

3.3 Photosynthetic efficiency and pigments measurement

Photosynthetic efficiency as described by the PSII quantic efficiency (i.e. Fv/Fm

fluorescence ratio) was measured by Chlorophyll fluorescence to estimate the impact of

shear stress/mixing on the photosynthetic capacity of the cell. This was obtained with a

chlorophyll fluorimeter (Water-Pam, Heinz Walz, Germany). The maximal value of Fv/Fm

fluorescence ratio is reported as reaching 0.6 for diatoms (Rech, 2004).

To measure pigments, Chlorophylls a and c and total carotenoids were measured

after extraction in 90% acetone (Strickland and Parsons, 1968; Ritchie, 2008).

Chlorophylls a and c were measured using Ritchie’s methodology (2008), and carotenoids

by the Strickland and Parsons (1968) methodology. Extracted pigment solution was

measured (480, 630, 647 and 664 nm) by spectrophotometer (V-630 UV-VIS, JASCO,

Germany). Note that these measurements were only made for torus-PBR experiments.

The extracellular marennine (EMn) was measured using a spectrophotometer (V-630 UV-

VIS, JASCO, Germany) in the supernatant after centrifugation of the sample. A wavelength

of 669 nm was selected and the EMn mass concentration measured using the Beer-

Page 184: Par « Rémi NGHIEM-XUAN

158

Lambert equation (ɛ = 12 x 104 mol L-1 cm-1 and MEMn = 9893 g mol-1) in accordance with

Pouvreau et al. (2006).

3.4 Determination of Mean Rate of Photon Absorption (MRPA)

The mean rate of photon absorption noted that “MRPA” refers to the photons

absorbed by cells per unit of time and biomass, and is expressed as µmolh𝑣 gX -1 s-1 (Cornet

et al., 1998; Pruvost et al., 2008; Kandilian et al., 2014). This physical quantity is related

to the light condition and the radiative properties of the cells, which were measured for

the experiment using a spectrophotometer with integrative sphere (Pilon et al., 2011;

Kandilian et al., 2016). MRPA can also be deduced from a balance on the photonic phase

on the culture system (Pruvost et al., 2008):

Equation 2

𝐴𝑙𝑖𝑔ℎ𝑡 =𝑆

𝑉 𝐶𝑋

(𝑞𝑂 − 𝑞𝐿) =𝑎𝑆

𝐶𝑋

(𝑞𝑂 − 𝑞𝐿)

where Alight is the MRPA received during a light period (µmolh𝑣 gX -1 s-1), q0 and qL refer

respectively to input and output photon flux density on the culture system (µmolh𝑣 m-2 s-

1), CX is the biomass concentration (g L-1), and aS the specific illuminated surface of the

culture system (m-1), which is the ratio between the illuminated surface S (m2) and the

culture volume V (m3).

Determination of the MRPA value was used to characterize the light absorbed by

suspended cells over a 24 h cycle. For this purpose, MRPA was first calculated during the

illuminated period (denoted Alight) and then related to the total culture cycle:

Equation 3

𝐴24ℎ, suspended = 𝑅𝑙𝑖𝑔ℎ𝑡𝐴𝑙𝑖𝑔ℎ𝑡

where RLight refers to the ratio of the light cycle (14/24 in our case) to compare results to

the continuous light case (denoted “Alight”).

The MRPA was then referred to the pelagic population (i.e. suspended cells). As a

consequence, the biomass concentration of pelagic cells Cpelagic was used as a reference

concentration in Eq. 2 with:

Page 185: Par « Rémi NGHIEM-XUAN

159

Equation 4

𝐴𝑙𝑖𝑔ℎ𝑡 =𝑎𝑆

𝐶𝑝𝑒𝑙𝑎𝑔𝑖𝑐

(𝑞𝑂 − 𝑞𝐿)

Because the biomass in suspension was a function of the stirring conditions, Cpelagic

was estimated as a function of the concentration of pelagic cells achieved in stirred

(Cstirred) and unstirred (Cunstirred) conditions. Cpelagic was obtained as a weighted function of

both concentrations to take into account the different stirring periods therefore applied:

Equation 5

𝐶𝑝𝑒𝑙𝑎𝑔𝑖𝑐 = 𝑅𝑠𝑡𝑖𝑟𝑟𝑒𝑑𝐶𝑠𝑡𝑖𝑟𝑟𝑒𝑑 + 𝑅𝑢𝑛𝑠𝑡𝑖𝑟𝑟𝑒𝑑𝐶𝑢𝑛𝑠𝑡𝑖𝑟𝑟𝑒𝑑

where Rstirred and Runstirred refer respectively to the ratio of stirred and unstirred time

during the culture.

Finally, this leads to the MRPA for suspended cells over a 24 h cycle:

Equation 6

𝑨𝟐𝟒𝒉, suspended =𝒂𝑺 ((𝒒𝟎 − 𝒒𝑳)𝑹𝑳𝒊𝒈𝒉𝒕)

(𝑪𝑺𝒕𝒊𝒓𝒓𝒆𝒅𝑹𝑺𝒕𝒊𝒓𝒓𝒆𝒅) + (𝑪𝒖𝒏𝒔𝒕𝒊𝒓𝒓𝒆𝒅𝑹𝑼𝒏𝒔𝒕𝒊𝒓𝒓𝒆𝒅)

CStirred and CUnstirred represent the pelagic biomass concentrations in the PBR which

were measured during stirred and unstirred periods respectively. These were estimated

from the measurement of cell density and total biomass concentration CX:

Equation 7

𝐶𝑆𝑡𝑖𝑟𝑟𝑒𝑑 = 𝐶𝑥𝑃𝑆𝑡𝑖𝑟𝑟𝑒𝑑

where PStirred is the pelagic population in stirred conditions, obtained from the ratio of

suspended cell density to total cell density:

Equation 8

𝑃𝑆𝑡𝑖𝑟𝑟𝑒𝑑 =𝐶𝑒𝑙𝑙𝑆𝑢𝑠𝑝𝑒𝑛𝑑𝑒𝑑 𝑖𝑛 𝑠𝑡𝑖𝑟𝑟𝑒𝑑 𝑐𝑜𝑛𝑑𝑖𝑡𝑖𝑜𝑛𝑠

𝐶𝑒𝑙𝑙𝑇𝑜𝑡𝑎𝑙

Similarly, CUnstirred was obtained:

Page 186: Par « Rémi NGHIEM-XUAN

160

Equation 9

𝐶𝑈𝑛𝑠𝑡𝑖𝑟𝑟𝑒𝑑 = 𝐶𝑥𝑃𝑈𝑛𝑠𝑡𝑖𝑟𝑒𝑑

Equation 10

𝑃𝑈𝑛𝑠𝑡𝑖𝑟𝑟𝑒𝑑 =𝐶𝑒𝑙𝑙𝑆𝑢𝑠𝑝𝑒𝑛𝑑𝑒𝑑 𝑖𝑛 𝑢𝑛𝑠𝑡𝑖𝑟𝑟𝑒𝑑 𝑐𝑜𝑛𝑑𝑖𝑡𝑖𝑜𝑛𝑠

𝐶𝑒𝑙𝑙𝑇𝑜𝑡𝑎𝑙

with PUnstirred the pelagic population in unstirred conditions and referring to the ratio of

suspended cell density to total cell density.

The stirred and unstirred time, represented respectively by Rstirred and Runstirred,

refers to the ratio of stirred and unstirred time (h) to the maximum light time represented

by TLight (daytime period of 14 h). Note that for the day/night cycle, 100% of the stirring

time is reached for 14, 20 and 24 h of stirring because the light period is 14 h (i.e. less than

the stirring time).

3.5 Characterization of shear stress influence on H. ostrearia

3.5.1 Investigation in rheometer

The effect of hydrodynamic shear stress on cell viability was first studied using a

Couette rheometer (PHYSICA MCR 500, Anton-Paar, UK), to apply a constant and

homogeneous shear stress field in perfect laminar flow on the diatom cells.

Shear cylinders of cone and plate types were used. The inner and outer cylinder

radii R1 and R2 were 14.36 mm and 14.46 mm respectively, with an inter-cylinder gap of

0.1 mm and a total volume of 25 mL. The inner cylinder rotates while the outer cylinder

remains stationary.

By modifying the viscosity of the medium, the impacts of both shear stress and

shear rate on the cell viability were investigated. For a Newtonian fluid (as is the case

here), both quantities were related to the dynamic viscosity of the medium:

Equation 11

𝜏 = μγ ̇

where τ is the shear stress (Pa), μ the dynamic viscosity of the fluid (Pa s-1) and γ̇ the shear

rate (s-1).

Page 187: Par « Rémi NGHIEM-XUAN

161

The maximum shear rate value achievable with shear cylinders was 35,000 s-1 as

obtained at a rotational speed of 1990 rpm. The exposure time to the different levels of

shear stress was limited to 1 h to avoid possible bias due to evaporation, and the

temperature was regulated at 20°C.

The experiments were conducted in triplicate on samples obtained on day/night

culture, for each hour of the daytime (14 h) and night periods (10 h). Cell counting was

carried out on the sample before treatment (i.e. control). Different values of shear rate

were tested (1,000; 5,000; 10,000; 20,000 and 30,000 s-1) but only results with a high

shear rate (30,000 s-1) were found to be relevant and are therefore presented here.

Flow laminar stability in the Couette flow was checked by calculating the Taylor

number to check that no Taylor vortices occurred. These vortices are found for Taylor

numbers higher than 41.3 (Michiels et al., 2010; Contreras et al.,1998), which is defined

as:

Equation 12

𝑇𝑎 =4ω2R4

ν2

where ω is the angular velocity (m min-1), R is the inter-cylinder gap (m) and ν = µ/ρ is

the kinematic viscosity (m2 min-1).

At 30,000 s-1, the angular velocity ω was 181.2 m min-1 and the cinematic viscosity

ν of our seawater-based substrate was 1.04 x 10-6 m2 min-1. This led to Ta = 12.13, below

the Taylor regime.

To distinguish shear rate and shear stress effects, the viscosity of the medium was

increased by adding Locust Bean Gum (LBG) (from Ceratonia siliqua seeds, Sigma Aldrich,

USA). Three different viscosity values were tested (8.54 x 10-3 Pa s-1, 2.84 x 10-2 Pa s-1 and

1.06 x 10-1 Pa s-1, corresponding respectively to 0.3, 0.5 and 1% of LBG). Note that for higher

values, a non-Newtonian shear thinning behavior tends to appear (Michiel et al., 2010; El

Batal et al., 2012).

3.5.2 Effect of shear stress in stirred flasks

In complementary experiments in constant and homogeneous shear stress fields,

the effect of mixing as obtained in stirred flasks was studied. The cells were cultivated

Page 188: Par « Rémi NGHIEM-XUAN

162

under the same conditions as the inoculum culture, with constant 24 h light and light

cycles of 14 h/10 h. Some Erlenmeyer flasks were stirred by a magnetic stirrer while

others remained unstirred. For the mixed flasks, shear rate intensity was estimated as

constant at 50 s-1 as fixed by the magnetic stirrer rotation speed. This shear rate value was

estimated according to Halasz et al. (2008). This corresponded to a Reynolds number Re

of around 5,000, corresponding to a turbulent regime (Re >4,000) inside the flasks (Rott,

1990).

3.5.3 Investigation in torus-shaped PBR

A torus-shaped PBR stirred by a marine impeller was used to investigate the long-

term effect of mixing and shear stress on H. ostrearia culture (Fig. 10, impeller no.2 was

used here). As with most mixed bioreactors, the effect of shear stress on living cells is

difficult to characterize. The shear stress field is difficult to determine accurately because

of a marked heterogeneity in the culture volume. This heterogeneity makes the shear

stress effect on living cells dependent on the cells’ history (and therefore trajectory) in

the culture vessel. The torus-shaped PBR has some advantages here. It can be divided into

two zones: the impeller zone and the circular loop zone (Fig. 10 left). This configuration

avoids a dead mixed volume (Khalid and Legrand, 2001). Because of its loop

configuration, the device also allows all the flowing cells under consideration to be

subjected to a similar history up to hydrodynamic shear stress. Although the shear stress

field is heterogeneous in the overall culture volume, all cells periodically flow in the region

of the impeller (approx. every 10 s), where the maximum shear rate occurs. Shear rate

can also be varied by simply modifying the rotation speed of the impeller, and even a low

impeller rotation speed (200 rpm) leads to good mixing conditions that avoids

sedimentation of H. ostrearia cells.

The impeller rotation speed was fixed at 200 rpm for all the experiments.

Following Pruvost et al. (2006) who numerically investigated the hydrodynamics of the

torus-shaped PBR, this leads to a mean bulk velocity of Uo = 0.05 m s-1 which was found

sufficient to avoid cell sedimentation, corresponding to an average shear rate value of 30

s-1 in the impeller zone and 10 s-1 in the loop zone, with an average value of 20 s-1 for the

total volume. Note that unlike the rheometer experiment where laminar conditions were

guaranteed, a turbulent regime was obtained in the torus PBR. These average shear rate

Page 189: Par « Rémi NGHIEM-XUAN

163

values can therefore only be considered as a first indication, as turbulent flows will

generate shear stress time and spatial variation.

Regarding culture conditions, the torus-shaped PBR enables full control of pH,

temperature, light and also mixing, in terms of stirring periods and intensity (i.e. impeller

rotation speed). Air is injected into the PBR to avoid the accumulation of dissolved oxygen,

which contributes to culture mixing. The flow rate was fixed at constant throughout the

experiments. CO2 was used for pH regulation, mixed in the air circuit by a control valve to

maintain a low variation in the total gas flow rate (100 mL min-1). Cells were grown in

non-limited culture medium 12N48P natural seawater (ES/3 complete inorganic

modified medium) to guarantee that any modification of the biomass productivity when

measured was only due to the effect of mixing.

Fig. 10: Representation of the torus shaped PBR geometry with its dimensions (left), and 2 types of impellers used for it stirring (right) (Pruvost et al., 2006)

3.6 Statistics

The statistics were compiled using XLSTAT 2019 software (Addinsoft). The normality

of the distribution law for the different samples was identified by the Shapiro-Wilk test to

select the tests (T-test). Statistics were measured with triplicate data for the experiments

in Figures 12 to 16 and with sextuplicate data for the in experiments Figures 17 to 19.

Page 190: Par « Rémi NGHIEM-XUAN

164

4 Results

4.1 Effect of short-term application of homogeneous and constant shear

stress

4.1.1 Investigation of experiment duration in rheometer treatment

As a first assessment of the experimental characterization in the rheometer, the

culture (i.e. seawater medium and microalgae cells) was first exposed at 20°C to a

constant shear rate of 30,000 s-1 (the highest intensity provided by the rheometer) for

300 min. The high viscosity of the medium was found to decrease after 260 min due to the

small volume and noted evaporation, although the temperature was regulated at 20°C. Up

until 260 min the proportionality of shear stress to sample viscosity for Newtonian fluid

was assessed as expected (Eq. 11) but with a variation in shear stress of 22% over the

total period, indicating a progressive decrease in apparent viscosity of the medium. For

the first 60 min the shear stress variation represented only 0.4%, which was considered

acceptable for our experiment (Fig. 11).

Fig. 11: Dynamic viscosity of the sample and shear stress applied over time at a constant shear rate of 30,000 s-1

In terms of the effect that shear stress has on cell viability, no significant

relationship was observed, even after five hours of treatment at the highest shear stress

of the rheometer (Fig. 12). The progressive decrease in viscosity (before 230 min) cannot

therefore be related to a cell disruption process. This was assumed to be attributable to

the thixotropic behavior of the culture (medium, algae and released exometabolites),

which is usually encountered in the liquid phase with solid inclusions (Reiner and Scott

0,0E+00

2,0E-04

4,0E-04

6,0E-04

8,0E-04

1,0E-03

1,2E-03

1,4E-03

1,6E-03

0

5

10

15

20

25

30

35

40

45

10 30 50 70 90 110 130 150 170 190 210 230 250 270 290

Dyn

amic

vis

cosi

ty µ

(P

a s-1

)

She

ar s

tre

ss τ

(Pa)

Treatment time in shear cylinder (min)

Shear Stress [Pa] Viscosity

Page 191: Par « Rémi NGHIEM-XUAN

165

Blair, 1967). Note that, while decreasing the apparent viscosity and therefore shear stress

value for a constant stress rate, this helps to limit the shear stress impact on cells.

Based on these results, a single hour of treatment was selected as the reference for

experiments. It was also considered unnecessary to analyze treatment at a lower shear

stress or shear rate.

Fig. 12: Evolution of cell viability at a constant shear rate (30,000 s-1) over time. Data is mean ± 95% CI for n = 3

Results are presented in Fig. 12. No apparent cells destruction was observed even

at the highest shear rate value reached by the rheometer.

4.1.2 Effect of viscosity on cell viability by rheometer treatment

The literature reports different quantities when investigating the effect of

hydrodynamics on cell viability. Shear stress (τ) is usually used (Connor et al., 2002;

Michiel et al., 2010; Leupold et al., 2013), but shear rate was also proposed (Barbosa et

al., 2004; Contreras et al., 1998), while some studies report the effects of operating

parameters of the processes used (Rossignol et al., 1999; Jaouen et al., 1998; Vandanjon

et al., 1999).

In our study, the shear rate is considered the reference and viscosity was

progressively increased to modify shear stress accordingly, as described in Table 12.

60

70

80

90

100

110

120

0 60 120 180 240 300 360

Ce

ll vi

abili

ty (

%)

Time (min)

Page 192: Par « Rémi NGHIEM-XUAN

166

Fig. 13: Haslea ostrearia cell viability at a constant shear rate of a 30,000 s-1 with different concentrations of Locust Bean Gum (LBG) to modify viscosities and shear stress accordingly. Data is mean ± 95% CI for n = 3

In all cases, no significant impact of viscosity on H. ostrearia cell viability was

observed, as shown in Fig. 13 (cell viability > 95%). At a constant shear rate, the shear

stress applied can increase up to 3180 Pa for 1% LBG mixture, but even at this value no

cell disruption was observed. Note that the Taylor number remained below the critical

value, indicating that no Taylor vortices occurred during the experiments.

Table 12: Experimental conditions during the experiment with progressive shear stress increase for a constant shear rate of 30,000 s-1

Control 0.3% LBG 0.5% LBG 1% LBG

Shear stress τ (Pa)

32.1 256.2 852 3180

Taylor number Ta

12.13 6.41 0.8 0.04

Dynamic viscosity µ (Pa

s-1) 1.07 x 10-3 8.54 x 10-3

2.84 x 10-2

Shear thinning 1.06 x 10-1

Shear thinning

We can note that these results contrast with those of Michiel et al., (2010). These

authors imposed a much lower shear rate (from 0 to 1000 s-1) and observed 66% cell

death in the rheometer experiment for the diatom Chaetoceros muelleri at 1.3 Pa. This may

be attributed to a difference in fragility of the species.

80

85

90

95

100

105

110

Control 0.3% LBG 0.5% LBG 1% LBG

Ce

ll vi

abili

ty (

%)

Sample of microalgae

Page 193: Par « Rémi NGHIEM-XUAN

167

4.1.3 Investigation in day/night cycles

Like many microalgae cultured in solar conditions, Halsea ostrearia cells seem to

synchronize with day/night cycles, triggering cell division during the dark phase (Mitsue

León-Saiki et al., 2018; De Winter et al., 2017). To study whether shear rate could have

more impact during the cell division stage, cell fragility was measured with a rheometer

on a flask culture subjected to light (14 h) and dark (10 h) periods. Because of the culture

growth and therefore cell concentration increase, cell viability was determined using the

cell concentration measured after treatment (2.41 ± 0.46 x 107 cell L-1) (Fig. 14).

The highest shear rate was first applied in the rheometer (30,000 s-1). Results are

given in Figure 14. H. ostrearia cells were found to be unaffected by the shear rate during

the 14 h of the light period (confidence interval and T-test does not allow variations to be

concluded between control and treated samples P > 0.05). We can note that the cell

viability of control samples is usually inferior to that of treated samples. This could be

attributable to cell homogenization by rheometer treatment compared to the control

(even with manual stirring before sample), by re-suspending some of the immobilized

cells in their own biofilm.

Same observation can be done for the night period. No significant shear stress

impact was observed. In addition, an increase in cell density, definitely linked to cell

division, is present throughout the night period. Note that this confirms that H. ostrearia

synchronizes with the light cycle.

Page 194: Par « Rémi NGHIEM-XUAN

168

Fig. 14: Haslea ostrearia cell viability cultivated in day/night culture. Cells were subjected to a constant shear rate (30,000 s-1) for 1 h. Values are reported for every hour during the day (14 h) and night (10 h) period of culture. Data is mean ± 95% CI for n = 3

Page 195: Par « Rémi NGHIEM-XUAN

169

4.2 Effect of long-term application of hydrodynamics shear stress in

turbulent flow

4.2.1 Investigation of stirring and dark/night illumination effects on H. ostrearia

in flask culture

Before analyzing H. ostrearia growth in a mixed PBR, the cell was cultured

beforehand in flasks. Cell growth with and without a magnetic stirrer was measured for

two illumination conditions: a 14 h/10 h photoperiod and constant illumination. Agitation

conditions produced by the magnetic stirrer (mean rotation 100 rpm) led to an estimated

shear rate of 50 s-1 in turbulent flow conditions (Halasz et al., 2008). Results are reported

in Fig. 15 in terms of cell productivity.

Fig. 15: Haslea ostrearia culture with and without stirring in light/dark cycle and constant 24 h illumination for 8-day batch. Data is mean ± 95% CI for n = 3

Surprisingly, the unstirred constant 24 h light and the unstirred 24 h in the light

cycle revealed the lowest cell productivity whatever the stirring conditions (between 1.2

and 2.0 x 106 cells L-1 d-1), the best results being achieved with the unstirred culture with

day/light cycles (7.5 x 106 cells L-1 d-1). Comparing stirred cultures, the culture in the

day/night cycle led to a 5.7-fold increase under the same light conditions (6.8 x 106 cells

L-1 d-1) as the 24 h lit culture (1.2 x 106 cells L-1 d-1).

Note that even though the day/night stirred culture production was slightly lower

than the unstirred one under same illumination conditions, the T-test confirmed no

0

1

2

3

4

5

6

7

8

9

10

unstirred Stirred light Stirred 24h unstirred

Light cycle L/D Light 24h

Ce

ll p

rod

uct

ivit

y (1

06ce

ll L-1

d-1

)

Light and stirring conditions

Page 196: Par « Rémi NGHIEM-XUAN

170

significant difference between the two results (P > 0.05). However, for constant 24 h

illumination, a significant impact on constant stirring can be concluded. It can therefore

be concluded that day/night periods are required for increased Haslea ostrearia growth.

Constant stirring has a lower effect, although it seems to negatively affect cell productivity

under turbulent flow conditions with magnetic stirring at around 50 s-1.

4.2.2 Investigation of mixing effect in torus shaped photobioreactor

A possible effect of stirring and shear rate on H. ostrearia cells can be highlighted

from previous results in flasks, achieved in turbulent flow and mainly when a night period

was added. These conditions are therefore very different from those applied in rheometer

experiments (homogeneous shear rate condition in laminar flow).

A full set of experiments was then conducted in a torus-shaped PBR to investigate

the possible long-term effect. The PBR was operated in semi-continuous conditions, and

cell density and photosynthetic efficiency were monitored every hour for 48 h (Fig. 16).

Only the two last days of the semi-continuous culture are reported here (after 7 days of

biomass productivity stabilization). The PBR was operated at a dilution rate of D = 0.1 d-1

with and without stirring by the marine impeller. Note that during the 48 h period of

measurement, the dilution rate was stopped for easier analysis of the cell response

(possibly hidden by culture harvesting). Yellow and blue areas represent light and dark

periods respectively throughout the culture. In addition, for the first 24 h of the culture

the PBR was unstirred and the impeller activated only after the first 24 h for the

remainder of the 48 h. Results are given in Figure 16.

Changes in culture conditions were found to affect cell productivity. In addition,

the response could not be described as usual, with successive increases and decreases in

cell density. A small increase from 1.7 x 108 cells L-1 to 2.5 x 108 cells L-1 appeared from 6

h to 8 h respectively on average, then cell density remained stable until 24 h when it

increased greatly from 25 h to 28 h, corresponding to the beginning of the stirring,

reaching 4.0 x 108 cells L-1. However, cell density decreased immediately to 1.7 x 108 cells

L-1 from 28 h to 31 h. Note that contrary to the experiments conducted in flasks, the

synchronized division stage was not observed at night.

Regarding PSII efficiency (Fv/Fm), values were found to range between 0.62 and

0.69, which is considered the optimal photosynthetic capacity for diatoms (Rech, 2004).

Page 197: Par « Rémi NGHIEM-XUAN

171

A decrease of 5% was observed at 26 h of the culture, which could be due to the impact of

mixing. However, this wide variation can also be observed throughout the culture, even

in unstirred conditions. In all cases, variations did not exceed 10%. It can therefore be

concluded that mixing itself has no significant impact on the photosynthetic capacity of H.

ostrearia cells. Only cell division seems to be affected by stirring.

Page 198: Par « Rémi NGHIEM-XUAN

172

Fig. 16: Time evolutions of cell density and fluorimetry during semi-continuous culture of Haslea ostrearia in a torus-shaped PBR. Data is mean ± 95% CI for n = 3

Page 199: Par « Rémi NGHIEM-XUAN

173

In a second experiment, H. ostrearia was cultivated in semi-continuous mode at D

= 0.1d-1 in constant light, but with different stirring durations. The goal was to better

identify cell response to stirring to establish a mixing strategy for H. ostrearia culture in

photobioreactors. Results are given in Fig. 17 & 18.

Fig. 17: Cell and biomass surface productivity (PS) obtained for semi-continuous culture of Haslea ostrearia in a torus PBR in continuous light with five different stirring duration conditions (0; 8; 14; 20; 24 h). Data is

mean ± 95% CI for n = 6

Figure 17 shows that 8, 14 and 20 h of stirring led to the largest biomass

productivities; larger than unstirred culture (i.e. 0 h in Fig. 17). 1.8 ± 0.2 g m-2 d-1 can be

achieved with 8 hours stirring, compared to unstirred culture which reaches 1.0 ± 0.2 g

m-2 d-1. However, 24 h stirring was shown to greatly reduce biomass productivity at 0.3 ±

0.2 g m-2 d-1. The same results profile can be observed for cell productivity, achieving up

to 22.9 ± 0.5 x 108cells m-2 d-1. However, constant stirring for 24 h led to only 4.3 ± 0.3 x

105 cells m-2 d-1 (equivalent to a cell density of 1.1 x 105cell L-1 which is lower than the

inoculation concentration of 1.5 x 105 cells L-1). So, even if a low biomass is still measured

for constant 24 h stirring, it tends to indicate that culture loss will result if the same

conditions are applied over time. Indeed, the 0.3 ± 0.2 g m-2 d-1 measured here can be

assumed to be mainly composed of cellular exometabolites and empty silica frustules as

observed by light microscopy. The T-test revealed no significant differences between the

14 h and 20 h experiments on biomass productivity (P > 0.05) or between 8 h and 14 h on

cell productivity (P > 0.05).

0

5

10

15

20

25

30

0,0

0,5

1,0

1,5

2,0

2,5

0 8 14 20 24

Cel

l PS

(10

8ce

ll m

-2d

-1)

Bio

mas

s P

S(g

m-2

d-1

)

Stirring time (Hours)

Biomass Cell

Page 200: Par « Rémi NGHIEM-XUAN

174

Fig. 18: Total photosynthetic pigment production (%) and Extracellular marennine (EMn) excretion (%) for semi-continuous culture of Haslea ostrearia in a torus PBR in constant light and different stirring duration

conditions. Data is mean ± 95% CI for n = 6

In terms of pigment production, the content of extracellular marennine (% W/W)

as shown in Figure 18 remains almost unaffected when 0, 8, 14 and 20 h of stirring is

applied, with a content of 0.61 ± 0.04, 0.57 ± 0.18, 0.24 ± 0.44 and 0.34 ± 0.09% of biomass

produced, respectively. This absence of significant difference between the 0 h (no

stirring), 8 h, 14 h and 20 h experiments on EMn productivity is confirmed by a confidence

interval. Surprisingly, the constant 24 h stirring led to significant results, with 6.81 ±

1.92%. This can be attributed here to the intracellular marennine which may have been

released after cell disruption, possibly due to shear stress.

In terms of total pigment concentration (% w/w), similar results as for biomass

productivity were observed. Stirring time allowed an increase of total pigment

concentration of 2.2 ± 0.3 % of biomass for 8 h of stirring, against 1.3 ± 0.2 % for unstirred

conditions. However, the increase in stirring time quickly becomes unfavorable for

pigment production, decreasing to 0.8 ± 0.1 % of biomass for a 24 h mixing period.

Our results tend to emphasize that Halsea ostrearia growth rate is favored by

applying mixing, but also that cells possibly disrupt over a 24 h period of stirring (at least

under the hydrodynamic conditions applied here). Where this relates to results achieved

in day/night periods, the cell disruption may be explained by the sensitivity of H. ostrearia

cells at the division stage, which could occur in a 24 h period. However, the wide variation

in pigment content also requires explanation. Mixing may indeed have an impact on total

-1012345678910

0

0,5

1

1,5

2

2,5

3

0 8 14 20 24

EMn

(%

w/w

)

Tota

l ph

oto

syn

thet

ic p

igm

en

t (%

w/w

)

Stirring time (Hours)

Total pigment EMn

Page 201: Par « Rémi NGHIEM-XUAN

175

pigment production by modifying the photosynthetic conditions in bulk culture. It was

actually shown that mixing does not affect the photosynthetic capacity of the cell, so

suspended cells are capable of photosynthetic conversion. Photosynthetic pigments are

known to be closely related to received light, with an overall decrease in pigment with

increasing light overexposure to limit the energy overload.

The results presented in Figures 17 and 18 alone, therefore, do not prove that 8 h

of stirring is the optimal value for H. ostrearia culture. Because of the large sedimentation

rate of H. ostrearia cells, stirring could indeed induce a change in access to light, which

could also benefit growth. To better analyze this, the rate of photon absorption, as

represented by the MRPA, was estimated for each stirring time (Equation 3). Results are

given in Figure 19.

Fig. 19: Estimation of Mean Rate of Photon Absorption (A24H, suspended) for 0, 8, 14, 20 and 24 h stirring duration in torus PBR, according to biomass surface productivity (PS) achieved (L = 4cm) (q0 = 300 μmol m-2 s-1). Data

is mean ± 95% CI for n = 6

Although this result must be considered as a first estimation due to the difficulty

in re-suspending H. ostrearia cells correctly in in the culture volume (measurement of cell

concentration is used in Eq. 8), this curve can be compared with the data obtained for

Chlorella vulgaris and Nannochloropsis gaditana, which possess optimal MRPA ranges 11

- 15 μmol g-1 s-1 and 13 – 15 μmol g-1 s-1 respectively (Artu, 2016). Our analysis reveals

that only 8, 14 and 20 h stirring periods achieve the range of optimal MRPA values for the

two previously-mentioned microalgae species, with respectively 10.8, 15.5 and 14.0 μmol

g-1 s-1 (Fig. 19). On the other hand, 0 h (unstirred culture) and 24 h stirring periods,

0

0,5

1

1,5

2

2,5

0 5 10 15 20 25 30 35 40 45 50

PS

Bio

mas

s (g

m-2

d-1

)

A24h,suspended (µmolhv gx-1 s-1)

0h 8h 14h 20h 24h

Page 202: Par « Rémi NGHIEM-XUAN

176

corresponding to the lowest biomass productivity (respectively 1.0 ± 0.1 and 0.4 ± 0.2 g

m-2 d-1), are found to be well above these values, with respectively 44.4 and 42.6 μmol g-1

s-1. Large values of MRPA are also known to result in a decrease in pigment content due

to overexposure to light. This is in agreement with the pigment analysis presented in

Figure 18. In addition to the possible effect of shear-stress on cells and their division, this

analysis therefore tends to suggest a strong relation between mixing period, light

exposure, pigment acclimation, MRPA value and, eventually, the resulting photosynthetic

growth.

5 Discussion

As a diatom strain and because of its shape and structure (frustule) H. ostrearia is

usually considered to be physically sensitive to shear stresses, hypothetically caused by

culture mixing (Nassiri et al., 1998, Wang and Lan, 2018). The frustule is also related to

cell division, which implies the frustule has to be opened, hypothetically making the cell

more sensitive to hydrodynamic stress.

Our results demonstrated, however, that H. ostrearia cells are more resistant to

shear stress than could be expected according to the literature, as shown, for example, by

our experiments in the rheometer, which allows a shear-rate of up to 30,000 s-1 to be

achieved (Fig. 12 & 14). In addition, contrary to what has been demonstrated elsewhere

(Michiel et al., 2010), increasing the culture viscosity and consequently the shear stress

did not result in any significant variation in cell viability (Fig. 13) in our case. The

experiment in the rheometer also failed to highlight any increase in cell fragility during

cell division. Although we were only able to conduct experiments under large shear-stress

values for 1 hour due to the conditions of the experiment, H. ostrearia can be considered

resistant to high shear-rate values in a laminar well-structured flow (i.e. without vortices),

as obtained with our experiments in the rheometer. Note that this could represent an

opportunity to design a specific PBR enabling this type of hydrodynamics.

In most culture systems, mixing generates vortices and in most cases turbulent

flow. The specific penne shape of the cell can induce a specific response here. Mauer et al.

(2016) attest that red blood cells present anisotropic behavior because of their oblong

shape, which allows mechanical forces to be limited due to the shear rate on the cells, by

orientating parallel to the fluid velocity gradient (i.e. shear rate). Such behavior could

Page 203: Par « Rémi NGHIEM-XUAN

177

explain our results with H. ostrearia in the rheometer. Because of their shape, cells can

orientate in the Couette flow which is thereby generated, leading to a limitation of the

impact that shear rate has on the cell.

However, parallel laminar flow is rarely found inside conventional mixed

photobioreactors. This makes the effect of shear stress difficult to understand due to the

possible large variations in time and space of shear stress fields. Shear stress fields are

therefore the result of the sparge valve, bubble disruption at the liquid/gas interface in a

tubular PBR, the recirculation pumps, etc. (Leupold et al., 2013 ; Connor et al., 2002 ;

Contreras et al.,1998 ; Jaouen et al., 1998 ; Wang and Lan, 2018). All these conditions could

lead to very different hydrodynamics. In other words, experiments in rheometer and

Couette flow are certainly far from being able to represent actual hydrodynamics in a PBR,

and cell fragility cannot be related to shear rate value.

The effects of mixing were further investigated by applying more realistic

conditions, namely turbulent mixing and day/night periods, because of the possible

relation between shear stress effect and cell division. Preliminary results in an

Erlenmeyer flask (Fig. 15) showed a significant negative impact when the culture was

continuously illuminated, especially when combined with constant stirring. This suggests

cell death due to cell disruption. However, additional experiments in semi-continuous

culture (i.e. long-term effect) refuted this hypothesis by emphasizing the positive impact

that certain stirring periods (8, 14, 20 h) have on biomass productivity. It also suggests

the hypothesis of cell division sensitivity by showing a possible larger cell disruption

during the cell division stage (fig. 16). Nevertheless, it was also shown that the effect of

stirring duration could not be separated from a modification of the cells’ light access. H.

ostrearia cells exhibit a fast sedimentation rate. By applying mixing, cells were re-

suspended, which led to better light access and therefore a resulting photosynthetic

growth for moderate values of mixing duration. In fact, pigment content was found to be

closely related to light absorption conditions (as represented by the MRPA), indicating a

photo-acclimation process, and lower productivity was observed at constant illumination,

which could be attributed to light overexposure. All of our results therefore tend to

illustrate the possibility that the sensitivity of H. ostrearia to mixing in a PBR culture could

result from the combination of shear rate (or shear stress), especially during cell division,

and also modification of the cells’ exposure to light because of their sedimentation ability.

Page 204: Par « Rémi NGHIEM-XUAN

178

Finally, the sensitivity of H. ostrearia to mixing is closely related to the physiology of the

cell (cell division and benthic behavior, but also light needs through modification to cell

light exposure), and not only to its physical resistance (shape and silicate structure).

6 Conclusions

H. ostrearia is usually described as a species that is highly sensitive to hydrodynamic

shear stress due to its tychopelagic behavior and rigid cell structure, composed mainly of

silica. Our study, however, attested to its ability to resist both high shear rate and shear

stress. Compared to similar studies available in literature on other diatoms, a higher

mechanical resistance than for Phaeodactylum tricornutum or Chaetoceros muelleri

(Sobczuk et al., 2005; Michiel et al., 2010), could be suggested, at least in laminar flow, as

tested here by using a rheometer.

H. ostrearia was found to be sensitive to long-term application of mixing in turbulent flow,

which could be attributed to the anisotropic behavior of such mixing conditions, resulting

in more complex hydrodynamics and preventing cell orientation along the flow. However,

comparing immobilized with mixed culture, stirring was shown to have a positive effect

on biomass productivity. The duration of the mixing was found to be relevant, a duration

of around 20 h and over resulting in a negative effect. This could be attributed to a more

marked fragility at the cell division stage

It was also suggested that the positive effect of a moderate mixing duration could

be attributed to an enhancement of cell access to light, resulting in better photosynthetic

growth conditions. Biomass productivity was found to be directly related to the rate of

photon absorption (i.e. MRPA), which was influenced by mixing re-suspended cells in

bulk.

This observation opens up the possibility of culture optimization, but it requires

further investigation as large values of MRPA produce a negative effect on biomass

productivity. This study should focus in particular on better characterization of the light

response of the pelagic cells of H. ostrearia. This physiological behavior has mainly been

studied in the ecological state and under benthic conditions (Robert, 1983; Rincé et al.,

1999). Further studies would be of interest in the pelagic conditions involved in a mixed

PBR.

Page 205: Par « Rémi NGHIEM-XUAN

179

Our study attests that H. ostrearia can be cultured under controlled stirring

conditions, thus avoiding limitations due to the cell immobilization process.

Page 206: Par « Rémi NGHIEM-XUAN

180

7 References

Artu A. (2016).: Etude et optimisation de la culture de microalgues en photobioréacteurs

solaires. (Thèse de doctorat ; Université de Nantes).

Barbosa M. J., Hadiyanto, Wijffels R. H. (2004). Overcoming Shear Stress of Microalgae

Cultures in Sparged Photobioreactors. Biotechnology and Bioengineering, 85:78-85.

http://doi.org/10.1002/bit.10862.

Camacho F.G., Gallardo Rodríguez J.J., Sánchez Mirón A., Cerón García, M.C., Belarbi E.H.,

Molina Grima E., (2007). Determination of shear stress thresholds in toxic dinoflagellates

cultured in shaken flasks. Process Biochemistry, 42, 1506–1515.

http://dx.doi.org/10.1016/j.procbio.2007.08.001.

Contreras A., Garcia F., Molina E., Merchuk J. C., (1998). Interaction between CO2-mass

transfer, light availability, and hydrodynamic stress in the growth of Phaeodactylum

tricornutum in a concentric tube airlift photobioreactor. Biotechnology and

Bioengineering, 60: 317 – 325.

Cornet J.-F., Dussap C. G., Gros J-B. (1998). Kinetics and Energetics of Photosynthetic Micro-

Organisms in Photobioreactors. In Bioprocess and Algae Reactor Technology, Apoptosis,

153–224. Advances in Biochemical Engineering Biotechnology 59. Springer Berlin

Heidelberg.

De Winter L., Dominguez Cabanelas I. T., Martens D. E., Wijffels R. H., Barbosa M. J. (2017).

The influence of day/night cycles on biomass yield and composition of Neochloris

oleoabundans. Biotechnology Biofuels, 10: 104.

El Batal H., Hasib A., Ouatmane A., Jaouad A., Naïmi M. (2002). Rheology and influence

factor of Locust Bean Gum solution. Revue de génie industriel, 8, 55-62.

Fadlallah H., Jarrahi M., Herbert E., Ferrari R., Mejean A., Peerhossaini H. (2016). Effects of

Shear Stress on the Growth Rate of Micro-organisms in Agitated Reactors. Journal of

Biotechnology, 20,167(4):370-6. http://dx.doi.org/10.1115/FEDSM2016-7590.

Gabelle J.C., Jourdier E., Licht R., Ben Chaabane F., Henaut I., Morchain J., Augier F., (2012).

Impact of rheology on the mass transfer coefficient during the growth phase of

Trichoderma reesei in stirred bioreactors. Chemistry Engineering Science, 75, 408-417.

DOI :1 0.1016/j.ces.2012.03.053.

Halasz G., Gyüre B., Janosi I. M., Szabo K. G., Tel T. (2008). Vortex flow generated by a

magnetic stirrer. American Journal of Physics, 75(12), 1092-1098.

Hochfeld W. L., (2005). Producing Biomolecular Substances with Fermenters, Bioreactors, and

Biomolecular Synthesizers, Boca Raton, FL, Taylor & Francis, ISBN 9781420021318.

Jaouen P., Vandanjon L., Quéméneur F. (1998). The shear stress of microalgal cell suspensions

(Tetraselmis suecica) in tangential flow filtration systems: the role of pumps. Bioresource

Technology, 68, 149–154.

Page 207: Par « Rémi NGHIEM-XUAN

181

Jubeau S. (2009). Optimisation de la synthèse d’un exométabolite par la diatomée marine

Haslea ostrearia (Stage de master 2 ; Université de Nantes).

Khalid A., Legrand J. (2001). Energy dissipation distribution and mixing in a torus reactor.

Chemical Engineering Communications, 185, 141–164.

Kandilian R., Pruvost J., Legrand J., Pilon L. (2014). Influence of Light Absorption Rate by

Nannochloropsis Oculata on Triglyceride Production during Nitrogen Starvation.

Bioresource Technology ,163 (July): 308–19.

Kandilian R, Pruvost J., Artu A., Lemasson C., Legrand J., Pilon L. (2016). Comparison of

Experimentally and Theoretically Determined Radiation Characteristics of Photosynthetic

Microorganisms. Journal of Quantitative Spectroscopy and Radiative Transfer, 175, 30-

45. https://doi.org/10.1016/j.jqsrt.2016.01.031.

Lebeau T., Gaudin P., Junter G-A., Mignot L., Robert J.M. (2000). Continuous marennin

production by agar-entrapped Haslea ostrearia using a tubular photobioreactor with

internal illumination. Applied Microbiology Biotechnology, 54, 634–640.

Leupold M., Hindersin S., Gust G., Kerner M., Hanelt D. (2013). Influence of mixing and shear

stress on Chlorella vulgaris, Scenedesmus obliquus, and Chlamydomonas reinhardtii.

Journal of Applied Phycology. 25, 485-495. http://doi.org/10.1007/s10811-012-9882-5.

Mauer J., Mendez S., Fedosov D. A., Fromental J., Claveria V., Nicoud F., Claveria V. (2016).

Red cells dynamic morphologies govern blood shear thinning under microcirculatory flow

conditions. Proceeding of the National Academy of Science of the Unite States of America,

113(50). http://doi.org/10.1073/pnas.1618852114.

Michiels, M. H. A., Van Der Goot A.J., Norsker, N-H., Wijffels R. H. (2010). Effects of shear

stress on the microalgae Chaetoceros muelleri, Journal of Applied Phycology, 28(1), 53-

62. http://doi.org/10.1007/s00449-010-0415-9.

Michiels M.H.A., Van Der Goot A.J., Vermuë M.H., Wijffels R.H. (2015). Cultivation of shear

stress sensitive and tolerant microalgal species in a tubular photobioreactor equipped with

a centrifugal pump. Journal of Applied Phycology, 28(1), 53-62. doi: 10.1007/s10811-015-

0559-8.

Mitsue León-Saiki G., Cabrero Martí T., Van der Veen D., Wijffels R. H., Martens D. E. (2018).

The impact of day length on cell division and efficiency of light use in a starchless mutant

of Tetradesmus obliquus. Algal Research, 31, 387-394.

Mitsuhashi S., Hosaka K., Tomonaga E., Muramatsu H., Tanishita K. (1995). Effects of shear

flow on photosynthesis in a dilute suspension of microalgae. Applied Microbiology and

Biotechnology, 42:744–749.

Nassiri Y., Robert J-M., Rincé Y., Ginsburger-Vogel T. (1998). The cytoplasmic fine structure

of the diatom Haslea ostrearia (Bacillariophuceae) in relation to marennine production.

Journal of the International Phycological Society, vol.37 (2), 84-91.

Page 208: Par « Rémi NGHIEM-XUAN

182

Ndiaye M., Gadoin E., Gentric C. (2018). CO2 gas-liquid mass transfer and kLa estimation:

Numerical investigation in the context of airlift photobioreactor scale-up. Chemical

Engineering Reasearch and Design, 133, 90-102.

Nghiem-Xuan R., (2019). Optimisation de la culture d’Haslea ostrearia en photobioréacteur

(Thèse de doctorat ; Université du Nantes).

O’Connor K. C., Cowger N. L., De Kee D. C. R., Schwarz R. P. (2002). Prolonged shearing of

insect cells in a Couette bioreactor. Enzyme and Microbial Technology, 31, 600–608.

Olmos E., Gentric C., Poncin S., Midoux N. (2003). Description of flow regime transitions in

bubble columns via laser Doppler anemometry signals processing. Chemical Engineering

Science, 58(9). DOI: 10.1016/S0009-2509(03)00002-2.

Pilon L., Berberoğlu H., Kandilian R. (2011). Radiation Transfer in Photobiological Carbon

Dioxide Fixation and Fuel Production by Microalgae. Journal of Quantitative

Spectroscopy and Radiative Transfer 112 (17): 2639–60.

Pouvreau J.-B., Morançais E., Fleury F., Rosa P., Thion L., Cahingt B., Zal F., Fleurence J.,

Pondaven P. (2006). Preliminary characterisation of the blue-green pigment “marennine”

from the marine tychopelagic diatom Haslea ostrearia (Gaillon/Bory) Simonsen. Journal

of Applied Phycology, 18, 757–767. http://doi.org/10.1007/s10811-006-9087-x.

Pruvost J., Pottier L., Legrand J. (2006). Numerical investigation of hydrodynamic and mixing

conditions in a torus photobioreactor. Chemical Engineering Science, 61, 4476–4489.

http://doi.org/10.1016/j.ces.2006.02.027.

Pruvost J., Cornet J-F., Legrand J. (2008). Hydrodynamics Influence on Light Conversion in

Photobioreactors: An Energetically Consistent Analysis. Chemical Engineering Science

63 (14): 3679–94.

Pruvost J., Cornet J. F., Goetz V., Legrand J. (2011). “Modeling Dynamic Functioning of

Rectangular Photobioreactors in Solar Conditions.” American Institute of Chemical Engineers

Journal 57 (7): 1947–60.

Pruvost J, Cornet J.F. (2012). Knowledge Models for the Engineering and Optimization of

Photobioreactors. Microalgal Biotechnology: Potential and Production. De Gruyter,

Berlin, Germany, 181–224.

Rech M., (2004). Effets de l’éclairement visible et ultraviolet sur la croissance et la

photosynthèse de microalgues: incidences sur l’écophysiologie du phytoplancton des

claires ostréicoles (Thèse de doctorat ; Université du Maine).

Reiner M., & Scott Blair G. W. (1967) in Eich, F. R., (ed) Rheology, Theory and Applications

Vol 4 p 465 (Academic Press, NY).

Richmond A., Qiang H. (2008). Handbook of Microalgal Culture: Applied Phycology and

Biotechnology. 2nd Edition. Chichester: Wiley-Blackwell.

Page 209: Par « Rémi NGHIEM-XUAN

183

Rincé Y., Lebeau T., Robert J. M. (1999). Artificial cell-immobilization: a model simulating

immobilization in natural environments? Journal of Applied Phycology, 11, 263–272.

Ritchie R. J. (2008). Universal chlorophyll equations for estimating chlorophylls a, b, c, and d

and total chlorophylls in natural assemblages of photosynthetic organisms using acetone,

methanol, or ethanol solvents. Photosynthetica, 46(1), 115–126.

Robert J.M. (1983). Fertilité des claires ostréicoles et verdissement: utilisation de l’azote par

les diatomées dominantes. (Thèse de Doctorat: Université de Nantes).

Robert J-M. (1989). Procédé de culture en masse de la diatomée Haslea ostrearia Simonsen.

European and French patent no 8915314. Nantes University.

Rossignol N., Moan, R., Jaouen P., Robert J. (1999). Continuous high-pressure disruption of

marine diatom Haslea ostrearia. Assessment by laser diffraction particle sizer.

Biotechnology Techniques, 13(12), 909–913.

Rossignol N. (1999). Procédé d’éxtraction et de séparation par membranes appliqués à la

production du pigment endo-et exocellulaire synthétisé par la diatomée Haslea ostrearia

Simonsen. Mise en œuvre d’un photobioréacteur à membranes à fonctionnement continu.

(Thèse de doctorat; Université de Nantes).

Rossignol N., Lebeau T., Jaouen P., Robert J.M. (2000). Comparison of two membrane-

photobioreactors, with free or immobilized cells, for the production of pigments by a

marine diatom. Bioprocess Engineering, 23, 495-501.

Rott N. (1990). Note on the history of the Reynolds number. Annual Review of Fluid

Mechanics. 22 (1): 1–11. doi:10.1146/annurev.fl.22.010190.000245.

Rouillard I. (1996). Optimisation de la production en masse d'Haslea ostrearia Simonsen sur

une eau souterraine salée : importance de la souche et des conditions de culture;

comparaison avec Skeletonema costatum (Grev) Cleve. (Thèse de doctorat; Université de

Nantes).

Sobczuk T.M., Camacho F.G., Grima E.M., Chisti Y. (2006). Effects of agitation on the

microalgae Phaeodactylum tricornutum and Porphyridium cruentum. Bioprocess and

Biosystem Engineering. 28, 243–250. http://dx.doi.org/10.1007/s00449-005-0030-3.

Spolaore P., Joannis-Cassan C., Duran E., Isambert A. (2006). Commercial Applications of

Microalgae. Journal of Bioscience and Bioengineering, 101(2), 87–

96.http://doi.org/10.1263/jbb.101.87.

Strickland J. D., Parsons T. (1968). A Practical Handbook of Seawater Analysis. Fisheries

Reasearch Bulletin. Bd. Canada.

Tesson B., Latz M. I. (2015). Article Mechanosensitivity of a Rapid Bioluminescence Reporter

System Assessed by Atomic Force Microscopy. Biophysical Journal, 108(6), 1341–1351.

http://doi.org/10.1016/j.bpj.2015.02.009.

Page 210: Par « Rémi NGHIEM-XUAN

184

Turcotte F., Mouget J. L., Genard B., Lemarchand K., Deschênes J. S., Tremblay R. (2016).

Prophylactic effect of Haslea ostrearia culture supernatant containing the pigment

marennine to stabilize bivalve hatchery production. Aquatic Living Resources,29, 401.

Turpin V., Robert J.M., Goulletquer P., Massé G., Rosa P. (2001). Oyster greening by outdoor

mass culture of the diatom Haslea ostrearia Simonsen in enriched seawater. Aquaculture

Rasearch, 32(801-809).

Vandanjon L. (1997). Valorization of marine microalgae. Concentration and purification by

membrane processes of pigments extracted from Haslea ostrearia. (Thèse de doctorat;

Université de Nantes).

Vandanjon L., Jaouen P., Rossignol N., Que F. (1999). Concentration and desalting by

membrane processes of a natural pigment produced by the marine diatom Haslea ostrearia

Simonsen. Journal of Biotechnology, 70, 393–402.

Wang C., Lan C. Q. (2018). Effects of shear stress on microalgae – A review. Biotechnology

Advances, (January). http://doi.org/10.1016/j.biotechadv.2018.03.001.

Page 211: Par « Rémi NGHIEM-XUAN

185

Chapter V:

Haslea ostrearia culture in

photobioreactor: growth

parameters gathering and

marennine production optimization

R. Nghiem Xuan1,3, J. L. Mouget1, 2, J. Pruvost 3, V. Turpin1, P. Jaouen3

1 Université de Nantes, MMS, EA 2160, Faculté des Sciences et des Techniques, 2 Rue de la

Houssinière, BP, 92208, 44322 Nantes Cedex, France.

2 MMS (Mer Molécules Santé), Le Mans Université, Ave O. Messiaen, 72085 Le Mans, France.

3 GEPEA, Université de Nantes, CNRS, UMR6144, Bd de l'Université, CRTT - BP 406, 44602 Saint-

Nazaire Cedex, France.

Ce chapitre est rédigé sous forme de publication pour soumission dans un journal

scientifique.

Page 212: Par « Rémi NGHIEM-XUAN

186

1 Abstract:

The marine diatom Haslea ostrearia which has the capacity to excrete a blue

pigment called marennine, described as having industrial potential according to

Gastineau et al., (2012), remains a challenging organism to culture in a photobioreactor

(PBR). Robert (1989) and Rouillard (1996) assumed that benthic behavior and apparent

sensitivity to hydrodynamic shear stress could be responsible for this. Immobilized-cell

protocols proved successful but have rather low extracellular marennine (EMn), i.e. 0.63

± 0.05 mg L-1 d-1 and 0.43 ± 0.05 mg L-1 d-1 for passive immobilized cell and agar gel

immobilization respectively. This could be attributed to a limitation in mass or light

transfer in the biofilm and consequent decreasing kinetics of biosynthesis (i.e. growth and

EMn production).

This study investigates culture in a conventional mixed PBR, which has previously

been proven to overcome culture limitations such as mass transfer, and to provide a high

level of control over light access to cells, enabling the application of systematic

optimization of photosynthesis-related kinetics.

Increasing the dissolved inorganic carbon concentration up to values of around 5-

10 mM was found to increase both growth kinetics and EMn production. For example,

EMn production reached 1.0 mg L-1 and 7.3 mg L-1 for DIC concentrations of 3.75 mM and

4.61 mM (corresponding to gas enrichment of 2% CO2 and 15% CO2 respectively). Growth

medium enrichment with silica (Si) to avoid growth limitation by nutrients was also found

to be challenging due to chemical precipitation. A fed-batch strategy on Si-P supply was

implemented, leading to an increase in biomass productivity (51.8 ± 2.3 mgX L-1 d-1).

Although EMn was produced continuously, the nutrient-limited conditions led to higher

productivity with 9.2 ± 1.7 mgEMn m-2 d-1, 37.9 ± 0.7 mgEMn m-2 d-1 and 54.5 ± 1.8 mgEMn m-

2 d-1 for phosphorus, silicon and nitrogen deprivation respectively.

In addition to the usual culture parameters of optimal pH and temperature (pHOPT

= 7.8 and TOPT = 22°C), the effect of light availability was also investigated, as represented

by the Mean Rate of Photon Absorption (MRPA). A direct relation was shown for both

biomass and EMn production kinetics. However, maximal biomass and EMn

productivities were found at different MRPA values, respectively 1.7 ± 0.1 gx m-2 d-1 at 12.3

µmolh gx-1 s-1 and 11.0 ± 0.5 mgEMn m-2 d-1 at 8.4 µmolh gx-1 s-1.

Page 213: Par « Rémi NGHIEM-XUAN

187

Finally, a continuous EMn production was obtained in optimal conditions, leading

to a productivity of 4.5 ± 0.16 mgEMn L-1 d-1, enabling validation of the conventional mixed

PBR for H. ostrearia culture and continuous EMn production.

.

Page 214: Par « Rémi NGHIEM-XUAN

188

2 Introduction

Microalgae are investigated for industrial purposes due to their potential for

producing molecules for pharmacology, cosmetology, food and biofuel, and for their use

in water treatment and CO2 valorization (Spolaore et al., 2006; Cardozo et al., 2007).

Nowadays however there are few industrialized microalgae, representing only a minor

part of algae biodiversity. There are estimated to be about 200,000 – 800,000 species of

which only 50,000 have been described (Starckx et al., 2012). This could be partly due to

difficulties in scaling-up the culture to industrial levels. A significant lack of knowledge is

usually the main reason, (Spolaore et al., 2006) but in the case of diatoms and

dinoflagellates, high sensitivity to hydrodynamic shear stress and stirring conditions is

often referenced (Wang and Lan, 2018). Additionally, specific culture medium and culture

conditions may be required, as has previously been shown for several species cultivated

on a large scale. This is the case with Haematococcus pluvialis for astaxanthin production,

obtained under intense light, high salinity, and low availability of nutrients (Shah et al.,

2016; Panis et al., 2016), and for Dunaliella salina production of β-carotene, obtained

under intense light and high salinity (Ben-Amotz et al., 1983; Oren, 2005).

Haslea ostrearia is a diatom that has been studied since 1820 (Gaillon, 1820) at

laboratory scale. It has the capacity to produce and excrete a blue pigment called

marennine, which has industrial potential (Gastineau et al., 2012). This microalgae strain

is well known to be shear-stress sensitive and its culture in PBR conditions is therefore

challenging. Culture in a specific photobioreactor (PBR), such as agar immobilization PBR

and immersed membrane PBR (Rossignol et al., 1999; Lebeau, 2000).was implemented.

The approach was successful due to the benthic behavior of H. ostrearia. Cell

immobilization also enabled the prevention of hydrodynamic shear stress on cells

(Robert, 1989; Rouillard, 1996). But although immobilized-cell protocols proved

successful, the extracellular marennine production obtained was fairly low, i.e. 0.63 ± 0.05

mgEMn L-1 d-1 and 0.43 ± 0.05 mgEMn L-1 d-1 for passive immobilized cell and agar gel

immobilization respectively. This can be attributed to limitations in mass or light transfer

in the biofilm, thereby decreasing the kinetics of biosynthesis (i.e. growth and EMn

production).

Medium composition and cell fragility have been investigated in a previous work. A

new NX medium was proposed and it was demonstrated that H. ostrearia cells could

Page 215: Par « Rémi NGHIEM-XUAN

189

tolerate mixing conditions when an appropriate stirring strategy was applied. This opens

up the perspective of a systematic optimization of culture conditions. As already

demonstrated in several publications (Pruvost et al., 2009; Takache et al., 2012),

suspended-cell culture enables the prevention of growth limitation by nutrient access

(including dissolved organic carbon) while optimizing photosynthetic growth by

controlling light availability in bulk.

The aim of this work is to investigate and optimize the parameters (pH, temperature

and light) for H. ostrearia culture in a mixed PBR. Comparison will be made with

immobilized cell culture. The influence on Extracellular Marennine (EMn) productivity

will also be considered. Currently, the

EMn production pathway remains unclear. Robert (1983) attested that EMn is

continuously produced as a primary metabolite, whereas Neuville and Daste (1972),

Rossignol (1999) and Lebeau, (1999) indicated an increased production under nitrogen

deprivation. Both strategies will be investigated here.

3 Materials and Methods

3.1 Culture media for production in PBR

The ES 1/3 medium (Robert 1983), modified with inorganic sources of phosphorus

and carbon was used for the following part of the study:

- Immobilized H. ostrearia cell culture

- Nutrient consumption in PBR culture

- Establishing the optimal growth parameters for biomass and EMn production

The final optimization was carried out using NX artificial seawater medium

(Nghiem-Xuan, 2019). Note that this medium was not used for all the experiments

because previous experiments were carried out before it was available.

3.2 Cells density and biomass concentration measurement

Cell density measurement was carried out by counting the cells using a Nageotte

counting chamber (Marienfeld, AQRA, India) and observed by light microscopy (AXIO

Scope A1, Zeizz, Germany). Dry weight measurement was carried out by filtering a 10 mL

sample on 0.7 µm Whatman GF/F. The same volume of ammonium format (NH4HCO2 at

Page 216: Par « Rémi NGHIEM-XUAN

190

28 g L-1) was added in order to remove the salt weight. The filters were then desiccated

over 24 h at 110°C.

3.3 Determination of pigment concentration

The measurement of pigments, chlorophylls a, c and total carotenoids, were

carried out with the acetone 90% extraction protocol. Chlorophylls a, c were measured

by the Ritchie (2008) protocol, and carotenoids by the Strickland and Parsons (1968)

protocol. The extracted pigment solution was measured (480, 630, 647 and 664 nm) by

spectrophotometer (V-630 UV-VIS, JASCO, Germany). The extracellular marennine (EMn)

was measured by a spectrophotometer (V-630 UV-VIS, JASCO, Germany) in the

supernatant after centrifugation of the sample. A wavelength of 669 nm was selected and

the EMn mass concentration measured using the Beer-Lambert equation (ɛ = 12 x 104 mol

L-1 cm-1 and MEMn = 9893 Da) according to Pouvreau et al. (2006).

3.4 Nutrients concentration measurement

Nitrogen and phosphorus were mainly measured by ionic chromatography

(DIONEX ICS 110 cationic column and ICS 900 anionic column, Thermo-scientific, France).

Silicon nutrient was measured by silicon-molybdate colorimetric methods (Grasshoff et

al., 2007 Revised). A high concentration of extracellular marennine interferes with this

colorimetric reaction and produces an overestimation. Dissolved Inorganic Carbon (DIC)

was measured by catalytic oxidation methods by combustion at 680°C, using a Total

Organic Carbon analyzer (TOC-L CSH/CSN, Shimadzu, UK).

3.5 Determination of the Mean Rate of Photons Absorption (MRPA)

The Mean Rate of Photon Absorption, also denoted “A”, refers to the light absorbed

by cells per unit of time and microalgal biomass. It is expressed in µmolh𝑣 kgx-1 s-1 (Cornet

et al., 1998; Pruvost et al., 2008; Kandilian et al., 2014). The MRPA is related to the

radiative properties of the biomass (Pilon et al., 2011; Kandilian et al., 2016). In our case,

MRPA was measured by photonic balance directly in the experiment culture, as

represented by Equation 13 (Pruvost et al., 2008):

Equation 13

𝐴 =𝑆

𝑉 𝐶𝑋

(𝑞𝑂 − 𝑞𝐿) =𝑎𝑆

𝐶𝑋

(𝑞𝑂 − 𝑞𝐿)

Page 217: Par « Rémi NGHIEM-XUAN

191

where q0 and qL refer respectively to the input and output PFD of the culture in µmolh m-

2 s-1. CX refers to dry biomass concentration in g m-3, and aS refers to the specific surface

of the PBR, which is the ratio between the illuminated surface and the volume of the PBR

in m-1.

3.6 Statistics analysis

The statistics were compiled using XLSTAT 2019 software (Addinsoft). Normality of

the distribution law for the different samples was identified by the Shapiro-Wilk test to

select the tests (T-test). Statistics were measured in triplicate for the experiments in

Figures 24 to 29 and 38 to 41. Statistics were measured in quintuplicate for the

experiments in Figures 30 to 37.

3.7 Haslea ostrearia culture

3.7.1 Inoculum

The diatom Haslea ostrearia NCC 501 was obtained from MMS laboratory and

cultivated in natural seawater from the Croisic (Loire-Atlantique, France), enriched with

Guillard F/2 medium. Cultures of H. ostrearia were maintained in 250 mL Erlenmeyer

flasks containing 100 mL medium with 10% dilution, after 4 - 5 days culture in Provasoli

ES1/3 (inorganic) modified medium (Robert, 1983) and remodified in total inorganic

sources. Light was supplied by white fluorescent light for the day/night cycle 14/10 at

300 µmol m-2 s-1 inside the culture chamber (SANYO MLR-350 culture chamber,

Richmond Scientific, UK) at 16°C. Fresh natural seawater was provided by Le Croisic

Ocearium (Loire Atlantique, France) and filtered at 0.7 µm then 0.45 µm before being used

for the medium.

3.7.2 Immobilized cell PBR

A PBR of 500 mL was designed to culture H. ostrearia in immobilized cell

conditions. It was designed to limit the shear stress impact of stirring by applying a low

liquid recirculation of the medium through the culture (Fig. 20). Recirculation of the

medium was carried out using a peristaltic pump and controlled to obtain a low horizontal

flowrate of 7 mL min-1. This value was set semi-empirically to avoid influencing the

formation of a biofilm of H. ostrearia (data not shown).

Page 218: Par « Rémi NGHIEM-XUAN

192

Fig. 20: Representation of the immobilized cell photobioreactor 500mL, aS=33 m-1

Gas was injected through the external wall of the PBR near to the pH probe to

better control pH regulation. pH was regulated by CO2 injection with an automatized

valve. The PFD was provided by setting a horizontal LED panel above the PBR. This

prevents any increase in temperature in the culture. The PBR was of Polymethyl

methacrylate (PMMA) material, with a depth of 0.03m (aS = 33 m-1).

This culture system was found to be well-adapted for investigating benthic cell

culture, which was obtained by allowing the cells to produce a cell “tissue” naturally at the

bottom of the PBR or by creating cell inclusions in agar gel. Agar gel and cell inclusion of

H. ostrearia was carried out using the Lebeau et al. (1999) protocol.

3.7.3 Airlift PBR 1L and 6L

1 L airlift PBR technology was used to cultivate H. ostrearia in conventional mixed

conditions. This type was proved to lead to sufficiently low shear stress for H. ostrearia

culture, as demonstrated in a previous work (Fig. 21) (data not shown). Gas was injected

at the center bottom of the PBR, which proved efficient for both mixing and gas-liquid

mass transfer (Kazbar, 2018; Ndiaye et al. 2018).

Page 219: Par « Rémi NGHIEM-XUAN

193

Fig. 21: Representation of 1 L airlift photobioreactor, aS = 33 m-1 (Souliès, 2014)

This technology is more adapted to planktonic (i.e. suspended) cells. It was found

that a single localized gas injection was insufficient to prevent cell sedimentation

completely, which occurs in the “dead zone” around the gas injection. Regarding shear

stress, high values were mainly located along the gas path. A value of around 150 s-1 in

this PBR was estimated, with a maximum of 275 s-1 through the injection axis (Ndiaye et

al., 2018; Chisti and Moo-Young, 1989). However, the bubble disruption and coalescence

that are present oat the top and bottom of the PBR respectively are also known to promote

the emergence of micro-vortices capable of producing up to 14,000 s-1 of mean shear rate

(Contreras et al., 1998; Barbosa et al., 2004; Michels et al., 2010).

In addition to the 1 L airlift PBR, a 6 L airlift PBR was used to study the effect of

light on H. ostrearia culture. This PBR was made of PMMA, which enables determination

of the MRPA by measuring PFD q0 and qL on both sides of the PBR (Equation 13). Note

that whereas the 1 L airlift PBR had only one gas injection, the 6 L airlift PBR is based on

multiple injections by injecting gas through a rubber mat placed on the bottom of the PBR

(Fig. 22).

Page 220: Par « Rémi NGHIEM-XUAN

194

Fig. 22: Representation of 6 L airlift photobioreactor, aS=25 m-1 (Souliès, 2014)

This perforated rubber mat allowed a continuous injection of gas along the bottom

of the PBR in the form of microbubbles, and consequently homogeneous aeration and

stirring of the PBR. The drawback is that the higher values of shear rate obtained in bubble

wakes are obtained across almost the entire culture volume, with a higher frequency of

bubble disruption/coalescence of microbubbles than in the 1 L PBR (Barbosa et al.,

2004; Sobczuk et al., 2006). Whereas the 1 L airlift PBR does not allow perfect biomass

homogeneity, the 6 L airlift PBR provides better cell suspension, which is a mandatory

condition for accurate measurement of MRPA in pelagic condition. Note that due to the

absence of a dead zone in this PBR, the gas flow rate was reduced every night to avoid the

stirring impacting the cell culture during cell division, as shown in a previous work (data

not shown).

3.7.4 Airlift Crystal PBR

The Crystal PBR is designed for the culture of microalgae species presenting a high

sedimentation rate (Moutel, 2016). Inspired by a settling tank, the shape of this PBR

allows cells to settle down to the gas injection, leading to cell re-suspension (Fig. 23). This

PBR has a culture volume of 250 mL, is 20 mm thick and has an illuminated surface of

14426.99 mm2. It was used for the first culture test of H. ostrearia in mixed conditions.

Page 221: Par « Rémi NGHIEM-XUAN

195

All the airlift PBRs (1 L, 6 L, Crystal) were controlled at a temperature at 22°C. This

was achieved by regulating the room temperature and with fans at the back of the Crystal

PBR only. Light was supplied by LED panels on the day/night cycle 14 h/10 h.

Fig. 23: Representation of Crystal photobioreactor 250 mL, aS = 57 m-1 (Moutel, 2016)

4 Results

4.1 Immobilized-cell culture

H. ostrearia was cultivated in passive cell immobilization to protect the cell from

the effects of shear stress. The immobilized-cell PBR culture was carried out in continuous

mode at dilution rate D = 0.1 d-1, 22°C throughout the culture, with pH regulation fixed at

7.8, a PFD of 200 µmol m-2 s-1 in the day/night cycle 14/10 and cultivated in ES1/3

(inorganic) modified medium. As a passive cell immobilization culture, the biomass was

not mixed, which did not require daily measuring of biomass concentration, and only two

points of cell density were carried out. EMn and dissolved inorganic concentrations in the

culture medium were measured. The results were compared with the agar cell inclusion

culture, which was conducted in the same PBR under the same conditions (Fig. 24). An H.

ostrearia cell in agar gel was included to achieve similar cell density to the passive-cell

Page 222: Par « Rémi NGHIEM-XUAN

196

immobilized culture inoculation. Note that cell immobilization in alginate beads was also

carried out but no positive results were obtained (data not shown).

Fig. 24: Passive and active continuous culture of Haslea ostrearia in immobilized cell PBR. Data is mean ± 95% CI for n = 3 and n = 2 for mean DIC.

Time evolutions of cell density, EMn and dissolved organic concentrations are

given in Fig.24. For biomass concentration, only values at to (4.13 x 107 cell L-1 for passive;

4.18 x 107 cell L-1 for active culture) and tend (cell death) were carried out.

Regarding immobilization, cell death was observed at day 9 whatever the passive

or active culture. Immobilized cell culture being demonstrated elsewhere, this could be

attributed to our growing conditions (see below). Our experiments, however, were found

to be sufficient for estimating marennine production. Both culture techniques revealed

maximum production at day 4, with 3.7 ± 0.2 mgEMn L-1 and 4.5 ± 0.2 mgEMn L-1 for active

immobilized and passive immobilized cultures respectively. In either case, EMn

production over time was found to be unstable and led to a progressive decrease in EMn

concentration, doubtless due to cell death.

Regarding culture loss, multiple attempts were carried out and similar results

obtained. This could not be attributed to shear stress, and the culture medium was

validated elsewhere (see below). One explanation could be the DIC concentration, as

0

1

2

3

4

5

6

7

8

0,0E+00

5,0E+06

1,0E+07

1,5E+07

2,0E+07

2,5E+07

3,0E+07

3,5E+07

4,0E+07

4,5E+07

5,0E+07

0 1 2 3 4 5 6 7 8 9

EMn

co

nce

ntr

atio

n (

mg

L-1)

Mea

n D

IC (

mM

)

Ce

ll d

ensi

ty (

cell

L-1)

Time (days)

Cell active Cell passive EMn active EMn passive Mean DIC

Page 223: Par « Rémi NGHIEM-XUAN

197

shown in Fig. 24. Values given here are an average of the concentrations obtained in

passive and active cell immobilization tests, as similar values were achieved. A

progressive decrease was obtained, which could lead to the appearance of carbon

limitation. It is known that immobilization reduces mass transfer in the cell’s biofilm.

Since a decrease in DIC contributes to the decrease of this mass transfer, it could lead to

limitation of the cells’ access to the carbon source in the film, leading to cell death. Because

the immobilization approach was only applied to obtain some preliminary values in EMn

production for future comparisons to other culture techniques, this was not investigated

further. The effect of carbon supply was investigated in suspended cell conditions (see

section below).

4.2 Haslea ostrearia culture in Crystal PBR

4.2.1 Investigation of CO2 tolerance of Haslea ostrearia in batch culture

CO2 intolerance in H. ostrearia was hypothetized by Robert (1983). In a natural

environment, H. ostrearia usually grows on the bottom of a biofilm, and consequently has

natural protection against the gas phase. Two air enrichments of CO2 (2 and 15%) were

tested in suspended cell culture. A Crystal PBR was used because of its efficiency with cells

presenting a high sedimentation rate. Batch cultures with the same conditions as for

immobilized cell culture were conducted (22°C, pH regulation fixed at 7.8) with an

incident PFD of 115 µmolh m-2 s-1 in day/night cycle 14/10. The incident PFD was

modified to take into account the larger specific illuminated surface of the Crystal PBR (i.e.

lower depth of culture). Results are given in Figure 25.

Page 224: Par « Rémi NGHIEM-XUAN

198

Fig. 25: Batch culture of Haslea ostrearia for 2% and 15% air/CO2 proportion for cell density (A) and extracellular marennine concentration (EMn) (B), in Crystal PBR. Data are mean ± IC 95% for n=3

A marked effect of the CO2 supply was observed. Cell densities reached 2.5 x 108

cell L-1 at day 8 and 1.4 x 108 cell L-1 at day 6 for 15% CO2 and 2% CO2 respectively (T-test

p < 0.001). The main impact was on EMn production, which reached 7.3 mg L-1 against 1.0

mg L-1 with 15% CO2 and 2% CO2 at day 10 respectively (T.-test p < 0.001). We noted a

decrease in EMn concentration for the 2% CO2. This could be attributable to EMn

degradation or consumption by bacteria. Finally, H. ostrearia was found to tolerate CO2-

enriched gas injection when combined with pH regulation. Increasing the proportion of

CO2 (and consequently the carbon transferred to the gas phase, as shown in next section)

revealed an increase in both cell density and EMn concentration. The relevance of carbon

limitation on H. ostrearia culture was therefore confirmed.

Page 225: Par « Rémi NGHIEM-XUAN

199

4.2.2 Assessment of Halsea ostrearia growth in continuous mixed culture

To assess the possibility of cultivating H. ostrearia in stirred conditions, a

continuous culture was conducted under mixed conditions in a Crystal PBR. The CO2

supply in the gas phase was set at 15% and the same conditions as in previous batch

culture were applied. The dilution rate was set at D = 0.1 d-1. Results are given in Figure

26.

Fig. 26: Continuous culture of Haslea ostrearia at 15% CO2 in a Crystal PBR. Cell density, extracellular marennine (EMn) and dissolved inorganic carbon (DIC) were measured. Data is mean ± 95% CI for n = 3

A maximum cell density of 1.5 ± 0.1 x 108 cell L-1 was reached at day 5 before

decreasing to 1.1 ± 0.1 x 108 cell L-1 at day 9. Regarding EMn production, the stirred

continuous culture led to a productivity of 6.9 ± 0.1 mgEMn L-1 at day 9 and 8.2 ± 0.2 mgEMn

L-1 at day 22. These values can be compared to immobilized-cell continuous culture which

reached a maximum of 4.5 ± 0.2 mgEMn L-1. The interest of cultivating H. ostrearia in mixed

conditions was here confirmed.

We can note also that both cell density and EMn concentration, as well as inorganic

carbon concentration, stabilized at day 8 for 4 days, before a decrease in cell density and

inorganic carbon. EMn concentration then started to increase (day 15), demonstrating

EMn synthesis or intracellular marennine (IMn) release in the medium. Despite the 15%

CO2 injection, dissolved inorganic carbon was also found to decrease continuously, as

0

1

2

3

4

5

6

7

8

9

0,0E+00

2,0E+07

4,0E+07

6,0E+07

8,0E+07

1,0E+08

1,2E+08

1,4E+08

1,6E+08

1,8E+08

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

EMn

co

nce

ntr

atio

n (

mg

L-1)

DIC

(m

M)

Ce

ll d

ensi

ty (

cell

L-1)

Time (Days)

Cell density EMn DIC

Page 226: Par « Rémi NGHIEM-XUAN

200

observed previously in the immobilized cell PBR culture. This indicated the need for a

further increase in carbon supply to avoid possible growth limitation.

4.3 Investigation of nutrient consumption in PBR culture

4.3.1 Carbon consumption

The DIC concentration was raised by adding sodium bicarbonate at the beginning

of the culture. Le Gouic (2013) estimated that Chlorella vulgaris needs up to 10 mM of DIC

to avoid carbon limitation, which could be given either by a phase of gas injection enriched

with CO2, or by dissolution of a chemical carbon source in the medium. 10 mM

corresponds to a sodium bicarbonate concentration of 1.2 g L-1. This concentration was

found to lead to nutrient precipitation in the seawater-based medium due to the

interaction of silicon, phosphate and carbonate with Ca2+ and Mg2+ endogenous entities in

seawater (Morey et al., 1995; Nguyen, 2013).

H. ostrearia was therefore cultivated in semi-continuous mode in the 1 L airlift

PBR, with a fed-batch strategy conducted on a sodium bicarbonate source. For this

purpose, a solution of sodium bicarbonate (5 mM maximum) was injected punctually

every day. Because of the parallel CO2 injection, this enabled us to raise the final DIC to

around 10 mM while avoiding the precipitation phenomenon. This was added to the

ES1/3 medium (6N24P inorganic medium), which was optimized for H. ostrearia growth

in PBR, as presented previously (Nghiem-Xuan, 2019). The semi-continuous mode was

selected to cultivate H. ostrearia because of its benthic characteristic, which leads to an

inhomogeneous culture. Harvesting was then applied for a limited period of time per day.

Other parameters were similar: temperature was fixed at 22°C throughout the culture, pH

regulation was fixed at 7.8 and PFD at 200 µmol m-2 s-1 (measured to obtain the same

MRPA value according to the different aS of the PBRs, see Eq.1) in day/night cycle 14/10

and dilution rate at D = 0.1 d-1 (Fig. 27).

Page 227: Par « Rémi NGHIEM-XUAN

201

Fig. 27: Semi-continuous culture of Haslea ostrearia in 1 L airlift PBR with carbon fed-batch and 2% CO2. Dry weight (DW), extracellular marennine concentration (EMn) and dissolved inorganic carbon (DIC) were

measured. Data is mean ± 95% CI for n = 3

Unlike with previous experiments, a constant production of biomass and EMn was

obtained throughout the semi-continuous culture. Dry biomass concentration reached

320 mgX L-1, corresponding to 2.6 x 108 cell L-1 and an EMn concentration of up to 8.2

mgEMn L-1. DIC concentration was found to stabilize progressively at the targeted 10 mM

concentration. The carbon-fed batch strategy was therefore found to be suitable to obtain

Page 228: Par « Rémi NGHIEM-XUAN

202

a stable and constant productivity of 31.7 ± 0.3 mgX L-1 d-1 of biomass and 0.8 ± 0.1 mgEMn

L-1 d-1

4.3.2 Silicon, nitrogen and phosphorus consumption

Silicon and phosphorus consumption were investigated during the semi-

continuous culture presented in Figure 27. The main macronutrients over time evolutions

are given in Figure 28.

All nutrients were found to decrease, as expected. Nitrate decreased to 57% of its

initial concentration (1.73 mM) but without deprivation, with a final concentration of

around 1 mM. Phosphate and silicon dioxide concentrations showed a marked decrease

and could therefore be suspected of limiting cell growth (around 0.7% of the initial

concentration after 4 days). A slight increase of PO42- and SiO32- concentration by day 7

was also noted. This could be attributable to cell release or possible dissolution of small

quantities of agglomerated compounds. In the case of the SiO2 colorimetric measure

method, the high production of EMn by day 7 (Fig. 27) is suspected of interacting with the

method and possibly overestimation of SiO2 concentration.

Fig. 28: Evolution of NO3-, PO42- and SiO32-concentration for semi-continuous culture of Haslea ostrearia in a 1 L airlift PBR. Data is mean ± 95% CI for n = 3

Time evolutions of main nutrients enable to determine substrate consumption

rates (rS), specific substrate consumption rates (QS) and the yield factor (YX) for NO3-, PO42-

and SiO32- (Table 13).

0

50

100

150

200

250

0

200

400

600

800

1000

1200

1400

1600

1800

2000

0 2 4 6 8 10 12 14 16

PO

42

-an

d S

iO3

2-co

nce

ntr

atio

n (

µM

)

NO

3-C

on

cen

trat

ion

M)

Time (Days)

N P Si

Page 229: Par « Rémi NGHIEM-XUAN

203

Table 13: Nutrient consumption of Haslea ostrearia (rS, QS and YX determination for NO3-, PO42- and SiO32-) in semi-continuous culture in an airlift PBR

NO3 PO4 SiO2

rS (µmol L-1 d-1) 3.0 x 10-1 1.0 x 10-1 9.2 x 10-2

QS (µmol gX-1 d-

1) 1.3 4.3 x 10-1 3.9 x 10-1

YX (µmol gX-1) 9.4 x 10-3 3.2 x 10-3 2.9 x 10-3

PO42- and SiO32- consumption rates were found to be respectively 1.0 x 10-1 and 9.2

x 10-2 µmol L-1 d-1; 3 times less than the NO3- consumption rate (nitrogen is the main

nutrient source of photosynthetic organisms). The specific substrate consumption rate

enables us to determine the need for these nutrients. The values obtained here emphasize

the need to increase SiO32- and PO42- concentration in the ES 1/3 medium to avoid any

limitation. However, SiO32- and PO42- are highly precipitating molecules with Ca2+ and

Mg2+ ions, such as sodium bicarbonate (Morey, Fournier & Rowe, 1995; Nguyen, 2013). It

was therefore decided to extend the fed-batch strategy on HCO3- to PO42-, SiO32-. The

nutrient feeding rate was based on the specific substrate consumption rate (Table 13) to

ensure no limitation of the nutrient and that no precipitation phenomenon would occur.

According to Figure 8, for a productivity of PV = 28.33 mgX L-1 d-1, H. ostrearia would need,

for example, a fed-batch of 1.1 x 10-2 µmol L-1 d-1 of SiO32- nutrient and 1.2 x 10-2 µmol L-1

d-1 for PO42-. This fed-batch strategy was implemented in a semi-continuous culture on ES

1/3 medium (6N24P inorganic). No further increase in either biomass or EMn

productivity was observed (data not shown). Thus, a further increase in N and P

concentrations was applied, and H. ostrearia was cultivated on 12N48P (Fig. 29). The

temperature was fixed at 22°C throughout the culture, pH regulation was fixed at 7.8, the

PFD at 330 µmol m-2 s-1 in day/night cycle 14/10 and the dilution rate at D = 0.1 d-1.

Results are given in Fig. 29.

Page 230: Par « Rémi NGHIEM-XUAN

204

Fig. 29: Semi-continuous Haslea ostrearia cell culture in 1 L airlift PBR with C, P, Si fed-batch conditions. Dry weight (DW) and extracellular marennine concentration (EMn) were measured. Data is mean ± 95% CI for n =

3

The fed-batch strategy applied to P-C-Si elements was found to prevent any

precipitation phenomenon in semi-continuous mode culture. This allowed the provision

of macronutrients such as N, P, Si and C in higher concentrations, leading to a 12N48P

medium.

A constant biomass and EMn production were obtained, stable over a period of

more than 20 days. However, whereas biomass productivity reached 51.8 ± 2.3 mgX L-1 d-

1, EMn productivity only reached 0.2 ± 0.1 mgEMn L-1 d-1. Medium enrichment in Si and P

could therefore have a negative impact on EMn synthesis, reducing EMn productivity from

0.8 ± 0.1 mgEMn L-1 d-1 (Fig. 27 B) to 0.2 ± 0.1 mgEMn L-1 d-1. On the other hand, it significantly

increased biomass concentration, up to 522 mgX L-1 (equal to 4.22 x 108 cell L-1). This value

can be considered high compared to H. ostrearia standards. Additionally, stable

production such as this can be considered an important methodologic achievement. This

opens up the possibility of further optimization of H. ostrearia culture in suspended cells.

The section below presents the investigation of light tolerance in a 6 L airlift PBR.

0

0,5

1

1,5

2

2,5

3

3,5

4

4,5

5

0

0,1

0,2

0,3

0,4

0,5

0,6

0,7

0 5 10 15 20 25 30

EMn

co

nce

ntr

atio

n (

mg

L-1)

DW

(g

L-1)

Time (Days)

DW EMn

Page 231: Par « Rémi NGHIEM-XUAN

205

4.4 The ingathering of optimal growth parameters for biomass and EMn

productions

4.4.1 Light tolerance and cell response

This study was carried out in the 6 L airlift PBR, in order to measure input and

output PFD for MRPA measurement (Eq. 13). The temperature of the culture was fixed at

22°C, pH regulation at 7.8 and dilution rate at D = 0.1 d-1. The test was carried out with 6

different PFDs ranging from 140 to 400 µmol m-2 s-1 (140, 165, 210, 300, 350, 400 µmol

m-2 s-1), with a day/night cycle 14 h/10 h. Each culture was monitored for 3 times the

residence time to obtain a constant productivity (Fig. 30).

Fig. 30: Light tolerance curve of Haslea ostrearia, describing the evolution of surface productivity (PS) depending on MRPA values for biomass and extracellular marennine (EMn) results. Data is mean ± 95% CI for

n = 5

The results given in Figure 30 reveal that both EMn and biomass were found to be

closely related to MRPA values. This indicates a light tolerance of the cell response in both

growth (as it should be) and also EMn production. An optimal MRPA value was achieved

for both, indicating a light tolerance of biological responses. Concerning biomass

productivity, the highest values were achieved in the range 10.7-12.3 µmolhv gx-1 s-1,

leading respectively to 1.6 ± 0.1 and 1.7 ± 0.1 gx m-2 d-1. However, the highest EMn

productivity was 11.0 ± 0.5 mgEMn m-2 d-1 which was obtained at 8.4 µmolhv gx-1 s-1, with a

different optimal value, therefore, than for biomass growth. With both biomass and EMn,

MRPAs higher and lower than 12.3 µmolhv gx-1 s-1 resulted in lower productivity. We can

also note that the range of optimal MRPA values, as well as the profile of the curves

0

2

4

6

8

10

12

14

0

0,2

0,4

0,6

0,8

1

1,2

1,4

1,6

1,8

2

0 2 4 6 8 10 12 14 16

PS

EMn

(m

g m

-2d

-1)

PS

Bio

mas

s (g

m-2

d-1

)

MRPA (µmol g-1 s-1)

Biomass EMn

Page 232: Par « Rémi NGHIEM-XUAN

206

achieved, are close to the photosynthetic growth response of Chlorella vulgaris and

Nannochloropsis gaditana (Artu, 2016).

The results can be combined to determine light conversion efficiency, which

represents the amount of biomass produced per µmolh absorbed (gx molh-1):

Equation 14

ηX/hν

=𝑆𝑋

(𝑞0 − 𝑞𝐿)

where SX represent the surface productivity in g m-2 d-1.

Fig. 31: Haslea ostrearia total pigment, extracellular marennine (EMn) and light conversion efficiency (η) into biomass as a function of MRPA. Data is mean ± 95% CI for n = 5

Figure 31 presents the results of light conversion efficiency and total pigment

content. Both evolutions as a function of the MRPA are found to be similar here. Light

conversion efficiency was found to decrease for values lower than 6 and higher than 14

µmolh gx-1 s-1, with maximal values around 0.10 ± 0.01 gx molh-1 in the range 8 and 12

µmolh gx-1 s-1.

The highest pigmentation result (2.9 ± 0.1% w/w) was found at 12.3 µmolh gx-1 s-

1, a value similar to the optimal MRPA condition for biomass productivity. We can note,

however, that the reverse was observed for Chlorella vulgaris, where a progressive

decrease in pigment content was obtained with increasing MRPA value, as a result of the

well-known pigment acclimation process (i.e. decrease of pigment content with increased

light absorption). This could be attributed to the effect of marennine. The EMn content,

0

0,02

0,04

0,06

0,08

0,1

0,12

0,14

0

0,5

1

1,5

2

2,5

3

3,5

0 2 4 6 8 10 12 14 16

η(g

mo

l-1)

Pig

me

nt

(% w

/w)

MRPA (µmolhv gx-1 s-1)

Total pigment EMn η biomass

Page 233: Par « Rémi NGHIEM-XUAN

207

which is also presented Figure 31, decreases continuously with the rise in MRPA.

However, the EMn content curves appear to cross the total pigment curve at 8 µmolh gx-1

s-1, representing a maximal light conversion efficiency of biomass for H. ostrearia.

To better understand EMn excretion as a function of light, the light conversion

efficiency of EMn was investigated. Results are given in Fig. 32.

Fig. 32: Haslea ostrearia total pigment, extracellular marennine (EMn) and light conversion efficiency (η) into EMn as a function of MRPA. Data is mean ± 95% CI for n = 5

As well as total photosynthetic pigment content correlated to light conversion into

biomass efficiency, Figure 32 also shows a similar evolution for the EMn produced (%

w/w) and light conversion into EMn efficiency. The rise in MRPA leads to a decrease in

EMn (%) from 1.89 ± 0.08 % to 0.17 ± 0.05%. The decrease in light conversion into EMn

efficiency occurs at 8 µmolh gx-1 s-1, which also leads to an increase in light conversion

into biomass efficiency (Fig. 31). Figures 31 and 32 show that 8 µmolh gx-1 s-1 triggers

lower EMn production in favor of biomass.

4.4.2 pH

The goal of this experiment was to observe and measure the impact of pH

conditions on H. ostrearia culture in order to select the optimal pH (pHOPT) conditions for

EMn production. The culture was carried out in a 1 L airlift PBR in semi-continuous mode

using a P, Si, C fed-batch strategy. The temperature of the culture was fixed at 22°C

throughout, pH regulation was fixed at 6; 7.8; 8.3; 9 and the dilution rate at D = 0.1 d-1.

The PFD was fixed at 300 µmol m-2 s-1 (to obtain AOPT around 12 µmolh gx-1 s-1) in the

0,00E+00

2,00E-04

4,00E-04

6,00E-04

8,00E-04

1,00E-03

1,20E-03

1,40E-03

1,60E-03

1,80E-03

0

0,5

1

1,5

2

2,5

3

3,5

0 2 4 6 8 10 12 14 16

η(g

mo

l-1)

Pig

me

nt

(% w

/w)

MRPA (µmolhv gx-1 s-1)

EMn Total pigment

Page 234: Par « Rémi NGHIEM-XUAN

208

day/night cycle 14/10. All cultures were monitored for 3 times the residence time to

achieve constant biomass productivity. Results are given in Fig.33 (the presence of the *

symbol indicates the presence of visual precipitation throughout the culture).

Fig. 33: pH effect on Haslea ostrearia total photosynthetic pigment concentration, biomass and extracellular marennine (EMn] productivitys (PS) in semi-continuous culture in 1 L airlift PBR. Data is mean ± 95% CI for n

= 5

Figure 33 shows optimum surface biomass productivity (1.63 ± 0.06 gx m-2 d-1) and

total pigment concentration (2.42 ± 0.08 %) at pH 8.3. However, for the same pH EMn

productivity was low, reaching only 5.27 ± 0.24 mgEMn m-2 d-1. pH 6 also showed low EMn

productivity, biomass and pigment concentration respectively at 6.71 ± 0.37 mgEMn m-2 d-

1, 0.63 ± 0.06 gx m-2 d-1 and 0.48 ± 0.05 %. However, pH 7.8 led to interesting EMn results

with 14.10 ± 0.65 mgEMn m-2 d-1; the best results for this study.

Pigment content was found to follow the same evolution as for biomass

productivity, which tends to attest to the direct relation of pH to H. ostrearia

photosynthetic growth. However, the results should be considered in hindsight because

of the consequences of pH variation. pH value is known to greatly impact the medium

chemistry of H. ostrearia, which could induce a limiting of nutrient conditions or

chemistry interaction with EMn. This hypothesis is supported by the increasing

precipitation with increasing pH value. pH 9 was found to lead to a high level of

precipitation, which raises the chemistry instability of the medium (Nghiem-Xuan, 2019).

0

2

4

6

8

10

12

14

16

0

0,2

0,4

0,6

0,8

1

1,2

1,4

1,6

1,8

6 7.8 8.3* 9*

PS

EMn

(m

g m

-2d

-1)

Tota

l ph

oto

syn

thet

ic p

igm

en

t (%

w/w

)

PS

bio

mas

s (g

m-2

d-1

)

pH

Total pigment PS Biomass PS EMn

Page 235: Par « Rémi NGHIEM-XUAN

209

Despite this precipitation, pigment concentration was found to be higher at pH 8.3 with

2.42 ± 0.08% than for pH 7.8. Finally, this raises the possibility of a pHOPT close to 8.3 while

avoiding any high level of precipitation.

Although the best biomass and pigment productivity result was achieved at pH 8.3,

the best EMn productivity was achieved at pH 7.8, which also exhibits no visual

precipitation. Because the culture of H. ostrearia was found to be more convenient at pH

7.8 due to the precipitation phenomenon being amplified in higher pH conditions, the pH

value of 7.8 was selected as pHOPT for the rest of the studies.

4.4.3 Temperature

The goal of this experiment was to observe and measure the impact of temperature

conditions on H. ostrearia culture in order to select the optimal temperature (TOPT)

condition for EMn production. The culture was carried out in a 1 L airlift PBR in semi-

continuous mode with P, Si, C fed-batch. During the culture, the temperature was

regulated at 16; 22; 25; 30°C throughout, pH regulation was fixed at 7.8 and the dilution

rate at D = 0.1 d-1. The PFD was fixed at 300 µmol m-2 s-1 (to obtain AOPT=12 µmolhv gx-1 s-

1) in the day/night cycle 14/10. All cultures were monitored for 3 times the residence time

to obtain a constant productivity. The results are given in Figure 34. The presence of the

* symbol indicates the presence of visual precipitation throughout the culture.

Page 236: Par « Rémi NGHIEM-XUAN

210

Fig. 34: Temperature effect on Haslea ostrearia ostrearia total photosynthetic pigment concentration, biomass and extracellular marennine (EMn] productivity (PS) in semi-continuous culture in 1 L airlift PBR.

Data is mean ± 95% CI for n = 5.

Temperature was found to impact the H. ostrearia culture, but mainly by triggering

the precipitation phenomenon which could negatively impact cell culture, as presented

above for the pH study. Precipitation was found to occur at temperatures higher than

22°C. Considering only a non-precipitate culture initially, the highest biomass results

were achieved at 22°C with 1.63 ± 0.06 gx m-2 d-1. There was therefore no significant

impact (T-test p = 0.748 for n = 5) of temperature on EMn productivity between 16 and

22°C. In terms of pigment concentration, there was no significant impact (T-test p = 0.563

for n = 5) at 16°C with 1.80 ± 0.29% compared to 22°C culture with 2.08 ± 0.15%, but a

higher pigment concentration is reached at 25°C with 2.62 ± 0.20%.

As with the pH study, the precipitation phenomenon definitely impacts biomass

and EMn. It can be noted, however, that pigment concentration is higher at 25°C even

though precipitation occurs. These results raise the possibility of a TOPT = 25°C, however,

precipitation also impacts biomass productivity greatly, decreasing it from 1.63 ± 0.06 gx

m-2 d-1 to 0.73 ± 0.19 gx m-2 d-1 in the 25°C and 30°C experiment. EMn productivity also

decreased with the increase in temperature. There could be various explanations for this:

the molecule could be degraded at a higher temperature, less could be produced by the

cell or it could interact with the precipitation phenomenon. Finally, because of the

precipitation phenomenon, 22°C was considered as TOPT for the rest of the study.

0

1

2

3

4

5

6

7

0

0,2

0,4

0,6

0,8

1

1,2

1,4

1,6

1,8

16 22 25* 30*

PS

EMn

(m

g m

-2d

-1)

Tota

l ph

oto

syn

thet

ic p

igm

en

t (%

w/w

)

PS

bio

mas

s (g

m-2

d-1

)

Temperature (°C)

Total pigment PS biomasse PS Emn

Page 237: Par « Rémi NGHIEM-XUAN

211

4.4.4 Physiologic stress for EMn overproduction

Based on previous results, the optimal values for MRPA, temperature and pH were

gathered (AOPT, TOPT and pHOPT) for H. ostrearia culture. Different conditions were then

tested to induce stressing conditions for EMn overproduction. Figure 29 shows

continuous production of EMn in a semi-continuous culture without nutrient limitation.

This was retained as a reference. Even in unrestricted conditions, the productivity of EMn

was found to be low at 0.2 mgEMn L-1 d-1, corresponding to 6.06 mgEMn m-2 d-1.

The following experiment consists of measuring EMn, biomass, cell and pigment

productivity in order to understand if nutrient depletion or limitation leads to

overproduction of EMn. This study was carried out in a 1 L airlift PBR and under the same

culture conditions as previously, but with TOPT = 22°C, pHOPT = 7.8 and AOPT = 12.3 µmolhv

gx-1 s-1. Depletion conditions were carried out to cultivate H. ostrearia without nitrogen,

phosphorus, silica or carbon nutrients. Nutrient concentration for limitation conditions

was fixed with regard to the 20-times diluted culture of Rossignol (1999). These

concentrations represent 2.39 x 10-4 M for nitrogen, 3.05 x 10-5 M for phosphorus and 5.49

x 10-5 M for silica. All cultures were monitored for 3 times the residence time to confirm

constant productivity (Fig. 35; 36; 37).

Fig. 35: Total photosynthetic pigment concentration of Haslea ostrearia under various nutrient limitations and depletions. Data is mean ± 95% CI for n = 5

The impact of nutrient limitation and depletion on pigment concentration can be

observed in Figure 35. None of the depleted or limited culture offers any increase in

pigment concentration, with values below 2.5 ± 0.2%. H. ostrearia pigment synthesis is

highly reduced by nitrogen depletion and limitation, at 0.7 ± 0.3% and 1.2 ± 0.4%

0

0,5

1

1,5

2

2,5

3

P Si N C P Si N

Depletion Nodepletion

LimitationTota

l ph

oto

syn

thet

ic p

igm

en

t (%

w/w

)

Page 238: Par « Rémi NGHIEM-XUAN

212

respectively. Both results can be explained by the high nitrogen requirement for protein

synthesis. The reduction of pigment content shows the cell’s capacity to adapt to balance

its composition in order to survive, and certainly impacts the cell’s photosynthetic

efficiency.

Fig. 36: Biomass and EMn productivity of Haslea ostrearia culture under various nutrient limitations and depletions. Data is mean ± IC 95% for n = 5.

The results are therefore analyzed in terms of EMn and biomass productivity

(Figure 36). Depletion and limitation were found to reduce biomass productivity by

almost 2.7 times compared to the unrestricted culture. In terms of EMn productivity,

depletion conditions only led to lower productivity than the unrestricted culture, at 5.3 ±

0.2 mgEMn m-2 d-1. However, P, Si and N limitation conditions led to higher productivity,

9.2 ± 1.7 mgEMn m-2 d-1, 37.9 ± 0.7 mgEMn m-2 d-1 and 54.5 ± 1.8 mgEMn m-2 d-1 respectively.

Silicon limitation therefore stimulates EMn release, while nitrogen limitation shows the

best results for EMn productivity, with 54.5 ± 1.8 mgEMn m-2 d-1. Note that there is no

significant difference (T-test p = 0.376 for n = 5) between phosphorus and silicon

limitation biomass productivity, which is not observed with 0.57 ± 0.05 and 0.62 ± 0.08

mgEMn m-2 d-1 respectively. EMn productivity is significantly higher (T-test p=0.002 for

n=5) for silicon limitation, with 37.9 ± 0.7 mgEMn m-2 d-1 than for phosphorus limitation,

reaching only 9.2 ± 1.7 mgEMn m-2 d-1. The opposite situation can be observed with EMn

productivity under replete culture, reaching 5.3 ± 0.2 mgEMn m-2 d-1, and the carbon-

depleted culture, reaching 5.8 ± 0.6 mgEMn m-2 d-1, whereas biomass productivity for

replete culture is 3.5 times higher. Finally, all these results show that nutrient limitation

0

10

20

30

40

50

60

0

0,2

0,4

0,6

0,8

1

1,2

1,4

1,6

1,8

P Si N C P Si N

Depletion Nodepletion

Limitation

PS

EMn

(m

g m

-2d

-1)

PS

bio

mas

s (g

m-2

d-1

)

Biomass EMn

Page 239: Par « Rémi NGHIEM-XUAN

213

(especially in Si and N) leads to a more marked release of EMn, resulting in greater EMn

productivity.

Since nutrient depletion is known to greatly affect cell physiology and morphoply,

Figure 18 presents both cell and biomass productivity for various nutrient depletion and

limitation conditions. Similar trends for both quantities were observed under both

nitrogen depletion and limitation, with respectively 4.3 ± 0.2 x 105 and 5.5 ± 0.2 x 105cell

m-2 d-1, which were found to be related to biomass productivity. However, in phosphorus

depletion and limitation conditions, cell productivity reached respectively 18.4 ± 0.5 x 105

and 9.4 ± 0.1 x 105cell m-2 d-1, whereas biomass productivity was found to be constant (T-

test p = 0.274 for n = 5) with 0.6 ± 0.1 gx m-2 d-1 for P depletion and 0.6 ± 0.1 gx m-2 d-1 for

P limitation.

Fig. 37: Biomass and cell productivities (PS) of Haslea ostrearia culture under various nutrient limitations and depletions. Data is mean ± 95% CI for n = 5.

As was observed between EMn and biomass productivity, nutrient depletion and

limitation led to a low correlation (r = 0.4178) between cell and biomass productivity.

This decorrelation is mainly observed in nutrient depletion for phosphorus and silicon.

The observation of different cell productivities with low biomass productivities attests to

a metabolism reorientation of H. ostrearia in depletion culture conditions. High cell

density in phosphorus-depleted conditions but with 3.5 times lower mass than no-

depleted culture shows that cell variation composition may be linked to metabolism

adaptation (Rodolfi et al., 2009; Goiris et al., 2015; Bajwa et al., 2018).

02468101214161820

0

0,2

0,4

0,6

0,8

1

1,2

1,4

1,6

1,8

P Si N C P Si N

Depletion Nodepletion

LimitationP

SC

ell

(10

5ce

ll m

-2d

-1)

PS

bio

mas

s (g

m-2

d-1

)

Biomass Cells

Page 240: Par « Rémi NGHIEM-XUAN

214

As a conclusion to these studies, no clear correlation can be assessed between

biomass and cell EMn productivity. However, it can be concluded that unrestricted culture

favors biomass production, with 1.6 ± 0.1 gx m-2 d-1, and nitrogen limitation favors EMn

release, with 54.5 ± 1.8 mgEMn m-2 d-1.

4.5 Towards an optimized production protocol

4.5.1 EMn concentration impact on process productivity

Previous studies have exhibited differing optimal conditions (pHOPT, TOPT, AOPT)

between growth and EMn production. This raises the hypothesis of EMn having a negative

impact on biomass productivity, and consequently on EMn productivity. Two tests were

therefore carried out to identify the impact of EMn on productivity.

H. ostrearia was cultivated in semi-continuous mode to remove 90% of culture

medium at each culture medium renewal, taking advantage of the benthic cell

characteristics. For this purpose, stirring was stopped a few hours before culture renewal

to allow the cells to sediment at the bottom of the PBR, and culture was renewed by

pumping culture medium from the top of the culture volume before replacing it with fresh

medium.

Two residence times were used: 10 days and 5 days, corresponding to around D =

0.09 d-1 and D = 0.18 d-1 respectively. The experiments were done with an optimized

formulation of NX medium to avoid Si, P and C fed-batch strategy. The cultures were

carried out in a 1 L airlift PBR. The temperature of the culture was fixed at 22°C, pH

regulation at 7.8 and incident PFD at 300 µmol m-2 s-1 (to obtain AOPT = 12 µmolh gx-1 s-1)

in the day/night cycle 14/10. For the first part of the study, H. ostrearia was cultivated

under nitrogen limitation (NX-5N100P6Si) conditions in order to enhance marennine

production, according to results achieved in the previous section. Results are given in Fig.

38 and 39.

Page 241: Par « Rémi NGHIEM-XUAN

215

Fig. 38: Evolution of cell density and extracellular marennine concentration (EMn) for semi-continuous culture of Haslea ostrearia in NX medium in 1 L airlift PBR (D = 0.09 d-1, corresponding to 10 days of residence

time). Data is mean ± 95% CI for n = 3.

After 10 days, cell density reached 4 ± 0.1 x 108cell L-1, which also corresponds to

maximum EMn concentration (11 mgEMn L-1). Culture medium was then renewed at 90%.

As expected, this resulted in a marked decrease in EMn concentration due to the dilution

(around 2 mgEMn L-1). Similar observation can be made on day 20, when the second

medium renewal was applied. After 10 days of EMn accumulation in the culture medium,

the concentration was decreased by culture renewal. However, this also induced a

continuous decrease in cell density over time. Even though the cell should have been

immobilized at the bottom of the PBR during the sample step, some of the cells were

definitely sampled during medium renewal. Cell density decreased to finally reach 0.4 ±

0.1 x 108cell L-1 by day 31. In addition, EMn concentration was found to remain below 11.6

± 0.1 mgEMn L-1 and decrease greatly over time, reaching 5.7 ± 0.3 mgEMn L-1 at day 31.

The same observation was made for the 5 days’ residence time experiment (Fig.

39). EMn concentration increased between each culture renewal, reaching 4.3 ± 0.1 mgEMn

L-1 and 5 ± 0.1 mgEMn L-1 at day 10 and day 15 respectively, but then decreased over the

culture to reach 1.3 ± 0.3 mgEMn L-1 by day 31. In terms of cell density, each medium

renewal reduced it and the cells barely managed any growth. Cell density reached 3.3 ±

0.1 x 108cell L-1 by day 5 and finally decreased to 1.7 ± 0.1 x 107cell L-1.

0

2

4

6

8

10

12

14

0,0E+00

5,0E+07

1,0E+08

1,5E+08

2,0E+08

2,5E+08

3,0E+08

3,5E+08

4,0E+08

4,5E+08

5,0E+08

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

EMn

co

nce

ntr

atio

n (

mg

L-1)

Cel

l den

sity

(ce

ll L-1

)

Times (Days)

Cell density EMn

Page 242: Par « Rémi NGHIEM-XUAN

216

Fig. 39: Evolution of cell density and extracellular marennine concentration (EMn) in a semi-continuous (D=0.18 d-1 corresponding to 5 days of residence time) culture of Haslea ostrearia in NX medium in a 1 L airlift

PBR. Data is mean ± 95% CI for n=3

Comparing the two experiments, the 10 days’ residence time experiment and 5

days’ residence time experiment led to an average EMn productivity of 1.0 mgEMn L-1 d-1

and 0.7 mgEMn L-1 d-1 respectively. Both these productivities are lower than the results

obtained by Rossignol (1999). In addition, no stable production was obtained for any of

these tests. The dilution step led to progressive cell density decrease, although EMn

concentration was regularly decreased. There was no strong evidence of any relationship

between EMn concentration and biomass productivity. However, a progressive evolution

of H. ostrearia cell population was observed, which was found to be distributed in the

benthic and pelagic phases even after 24 hours without manual stirring (Table 14).

0

1

2

3

4

5

6

0,0E+00

5,0E+07

1,0E+08

1,5E+08

2,0E+08

2,5E+08

3,0E+08

3,5E+08

4,0E+08

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

EMn

co

nce

ntr

atio

n (

mg

L-1)

Cel

l den

sity

(ce

ll L-1

)

Time (Days)

Cell density EMn

Page 243: Par « Rémi NGHIEM-XUAN

217

Table 14: Repartition and observation under light microscope of the Haslea ostrearia cell population density after 24 h without manual stirring. Data is mean ± 95% CI for n = 3.

Pelagic phase Benthic phase

Cell population

density (%)

42.7 ± 0.4 57.3 ± 0.4

Visual

appearance

The results of Table 14 could explain the decrease in cell density at each medium

renewal, as mainly pelagic cells were harvested. The visual appearance of the cells also

raises the hypothesis of a higher IMn production by the benthic cell population than the

pelagic one. As described by Robert (1983) and Rossignol (1999), this cell behavior is

observed in the natural H. ostrearia environment, which confirms the possibility that

pelagic cell population could be constituted mainly of young cells, and the benthic cell

population constituted of older ones. This hypothesis could explain why removing pelagic

cells leads to a progressive decrease in cell density, as young cells were mostly harvested

at each medium renewal.

4.5.2 Implementation of EMn continuous production process

In a final assessment of our optimization procedure, the knowledge and data

gathered from our experiments was combined to carry out continuous EMn production in

a conventional process.

H. ostrearia was cultivated in NX-25N100P6Si medium (unrestricted growth

medium) until day 8, to optimize the biomass produced and increase marennine

production or/and release. NX-5N100P6Si medium (N limited growth medium) was then

injected at D = 0.25 d-1 in order to overproduce EMn under nitrogen stress culture

conditions. During the culture, TOPT was fixed at 22°C and pHOPT regulation at 7.8. The PFD

was fixed at 300 µmol m-2 s-1 (biomass AOPT=12 µmolhv gx-1 s-1) in the day/night cycle

Page 244: Par « Rémi NGHIEM-XUAN

218

14/10 throughout the culture. Note that the PFD fixed at 200 µmol m-2 s-1 (EMn AOPT=8

µmolhv gx-1 s-1) was initially thought to be overproducing marennine, but with the

concentration of marennine in the PBR expected to be significant, light attenuation should

also be increased. The PFD was therefore slightly increased. Results are given in Fig. 40.

Fig. 40: Evolution of dry weight (DW), extracellular marennine concentration (EMn) and specific EMn concentration of Haslea ostrearia for semi-continuous culture in NX media with optimal culture conditions in

a 1 L airlift PBR. Data is mean ± 95% CI for n = 3

The 8-day batch culture with NX-25N100P6Si allowed a maximum biomass

concentration of 0.63±0.02 g L-1 to be reached. Biomass then instantly decreased by day

8 with the addition of nitrogen-limited medium (NX-5N100P6Si), leading to an increase

in EMn production with a maximum EMn concentration of 19.64 ± 0.19 mg L-1 on day 15.

The specific EMn concentration also increased by day 8 as expected, from 0.88 ± 0.05 mg

108cell-1 to 7.81 ± 0.08 mg 108cell-1 by day 15, and attested to the capacity of H. ostrearia

to overproduce EMn in nitrogen-limited complete inorganic medium. To confirm the

efficiency of the airlift stirred culture, H. ostrearia was cultivated in the same conditions

but in the immobilized cell PBR. Results are given in Fig. 41.

0

5

10

15

20

25

0

0,1

0,2

0,3

0,4

0,5

0,6

0,7

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

EMn

co

nce

ntr

atio

n (

mg

L-1)

Spec

ific

EM

n (

mg

10

8ce

ll-1

)

DW

(g

L-1)

Times (Days)

DW EMn EMn spe

Page 245: Par « Rémi NGHIEM-XUAN

219

Fig. 41: Evolution of dry weight (DW) and extracellular marennine concentration (EMn) of Haslea ostrearia for a continuous culture in NX media with optimal culture conditions in a 500 mL immobilized cell PBR. Data

is mean ± 95% CI for n = 3

Similar biomass and EMn evolutions as discussed above (Fig. 24) were obtained.

However, the EMn concentration reached 10.6 ± 0.2 mg L-1, whereas previous

immobilized cell culture remained at around 4.5 ± 0.2 mg L-1. This emphasizes the interest

of our systematic optimization of culture conditions. In terms of biomass, time monitoring

was not allowed by cell immobilization. Only to (0.07 ± 0.02 g L-1) and tend (0.15 ± 0.05 g L-

1) were carried out.

Compared to stirred culture, the interest of producing H. ostrearia in suspended

(pelagic) cells is confirmed here. An EMn productivity of 4.5 ± 0.2 mg L-1 d-1 was obtained,

although the influence of EMn concentration on cell growth or physiology was not

elucidated.

5 Discussion

This study was the result of numerous experiments aimed at elucidating the main

parameters governing H. ostrearia culture in PBR. All immobilization studies resulted in

low productivity (0.63 ± 0.05 mgEMn L-1 d-1 for passive immobilized cell and 0.43 ± 0.05

mgEMn L-1 d-1 for agar gel immobilization). Alginate entrapment bead was not found to be

resistant to stirring and medium chemistry. Carbon nutrient limitation was found to be a

highly relevant limiting parameter. Even though pH is regulated by CO2 injection, the

geometry of the PBR led to a low mass transfer of CO2 in the biofilm.

0

2

4

6

8

10

12

0

0,05

0,1

0,15

0,2

0,25

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

EMn

co

nce

ntr

atio

n (

mg

L-1)

DW

(g

L-1)

Time (Days)Cell Emn

Page 246: Par « Rémi NGHIEM-XUAN

220

A systematic optimization was then conducted in cell-suspended mode. Dissolved

inorganic carbon (DIC) was added by CO2 injection (Fig. 25), eventually combined with

the addition of sodium bicarbonate (Fig. 27). H. ostrearia revealed a high DIC requirement

in the PBR culture, mainly for EMn production, reaching 7.3 mg L-1 with 15% CO2 against

1.0 mg L-1 with 2% CO2. The high requirement of H. ostrearia for carbon supply could be

attributed to the high carbon content (36 ± 0.6%) of EMn (Pouvreau et al., 2006). With

the addition of sodium bicarbonate, it was found necessary to keep DIC constant at 10

mM, to reach 31.7 ± 0.3 mg L-1 d-1 of biomass and 0.8 ± 0.1 mg L-1 d-1 of EMn in semi-

continuous culture mode. However, culture medium precipitation appeared at high DIC,

phosphorus and silicon concentration, which impaired biomass and EMn productivity.

The substrate consumption rate H. (rS) and specific substrate consumption rates

(QS) of NO3-, PO43- and SiO32- (Table 1) were determined. This enabled us to set a fed-batch

strategy for the P, Si, C elements, avoiding precipitation and therefore increasing biomass

productivity to 51.8 ± 2.3 mgx L-1 d-1. In terms of EMn productivity, this leads to the lowest

productivity obtained, with 0.2 ± 0.1 mgEMn L-1 d-1 (Figure 29). As described by Robert

(1983), H. ostrearia excretes marennine continuously, but surprisingly in nutrient

unrestricted conditions this leads to metabolism orientation to reduce EMn production.

As described in previous studies (Neuville & Daste, 1972; Rossignol, 1999; Lebeau et al.,

2000), marennine overproduction can be induced by nitrogen limitation. This was

confirmed by our study (Figure 37). Phosphorus, silicon and nitrogen limitation

conditions lead to higher productivities than the control test, respectively 9.2 ± 1.7 mgEMn

m-2 d-1, 37.9 ± 0.7 mgEMn m-2 d-1 and 54.5 ± 1.8 mgEMn m-2 d-1. Nitrogen is definitely the best

nutrient limitation to overproduce EMn.

Optimal conditions of temperature, pH and light were investigated. pHOPT = 7.8 and

TOPT = 22°C were assessed. These values were not different from the ideal settings

previously established (Robert, 1983; Turpin et al., 1999), but it was found quite difficult

to determine the optimal ones due to the medium chemistry leading to precipitation of

the medium, which started at pH 8 and 25°C (Nguyen, 2013).

Concerning light, it was determined here to be a function of the MRPA (denoted A),

which enables PFD, biomass concentration and PBR geometry to be gathered in a single

representative physical quantity. Different optimal values AOPT were identified for both

biomass and EMn production, with respectively 1.7 ± 0.1 gx m-2 d-1 at 12.3 µmolh gx-1 s-1

Page 247: Par « Rémi NGHIEM-XUAN

221

and 11.0 ± 0.5 mgEMn m-2 d-1 at 8.4 µmolh gx-1 s-1. Biomass production was found to be

positively impacted by high PFD until an MRPA of around 12 µmolh gx-1 s-1 was reached.

We also noted that these values were close to the values for other species known for their

high light tolerance (Chlorella vulgaris and Nannochloropsis gaditana, with 15 μmolhv gx-1

s-1 ≥ AOPT ≤ 20 μmolhv gx-1 s-1 and AOPT = 13 μmolhv gx-1 s-1 respectively) (Artu, 2016).

The conversion efficiency of light into biomass (Figure 31) showed an efficiency of

around 0.10 ± 0.01 gx molh-1 for MRPA values in the range 8-12 µmol h gx-1 s-1, but a low

correlation (r = 0.4564) between MRPA and total pigment concentration evolution was

observed. The lower the MRPA value, the higher the pigment concentration should be,

according to the photoadaptation phenomenon (Pruvost & Cornet, 2012; Artu, 2016). The

reverse behavior was observed here for H. ostrearia, with an increase in photosynthetic

pigment content from 8 to 12 µmolh gx-1 s-1, which was also found to correspond to the

maximum conversion efficiency of light into biomass. This also corresponded to a

simultaneous decrease in EMn concentration (%) with a switch from EMn production to

biomass, which could be estimated at around 8 µmolh gx-1 s-1 (Figure 32). All these results

tend to indicate a possible particular evolution of H. ostrearia pigments towards light (i.e.

different from Chlorophyceae), which could be linked to H. ostrearia metabolism.

In terms of EMn production, a productivity of 11.0 ± 0.5 mgEMn m-2 d-1 produced at

8.4 µmolh gx-1 s-1 was obtained. Light, as represented by the MRPA, was revealed to be an

important parameter for marennine production. But contrary to Mouget et al. (2004;

2005), according to whom the higher the PFD, the higher the EMn produced, our study

showed an optimal MRPA for EMn production at around 8 µmolh gx-1 s-1, which was found

to be lower than the optimal MRPA for biomass production. Given that the origin of

marennine synthesis is still unknown, only a partial hypothesis can be elaborated here. In

our experiments, EMn production was found to be inversely related to biomass

productivity and photosynthetic pigment concentration (figures 30 and 33). Even if EMn

could (directly or indirectly) come from photosynthetic pigment synthesis or be related

to photosynthetic growth metabolism, it certainly follows its own pathway and role in

terms of cell physiology. These pathways could be energetically in conflict. Therefore, the

cell should promote either biomass or marennine. This hypothesis tends to be verified by

our experiment with nitrogen limitation conditions, which revealed the enhancement of

EMn synthesis.

Page 248: Par « Rémi NGHIEM-XUAN

222

EMn is known to be a highly bioactive molecule (Gastineau et al., 2012). Therefore,

when concentration is increased in culture it could have a negative impact on H. ostrearia

growth. If marennine has a negative impact on biomass productivity (and thus the EMn

synthesis itself) during the production process, this molecule should be removed from the

medium during culture. Additionally, the shadowing characteristic of EMn also raises the

possibility of limiting light propagation in the PBR, thus also reducing photosynthetic

growth. We therefore tested an adapted protocol based on medium renewal (+EMn) after

a period of unstirred conditions to allow cells to sediment at the bottom of the PBR.

Unfortunately, the tychopelagic characteristic of the cell prevented it from removing 90%

of EMn without decreasing too much cell density. The hypothesis of controlling the

tychopelagic behavior by culture condition is raised to better control the production

process.

In a final experiment, we finally demonstrated that whatever the impact of

marennine, the culture of H. ostrearia in a stirred conventional PBR managed to produce

continuous EMn at 4.5 ± 0.16 mgEMn L-1 d-1, against 0.7 ± 0.1 mgEMn L-1 d-1 for immobilized

cell culture (Figure 41). The conventional stirred PBR is therefore confirmed for H.

ostrearia culture and continuous EMn production.

6 Conclusions

Cell immobilization has been presented as a solution for H. ostrearia culture for a

long time (Robert, 1983, 1989; Rouillard, 1996; Rincé et al., 1999; Rossignol, 1999;

Lebeau et al., 2000; Turpin et al., 2001). However, it is rather difficult to scale up and

induces possible limitations related to mass or even light transfer in biofilms (Hochfeld,

2005; Gabelle et al., 2012; Souliès, 2014; Kandilian et al., 2017; Ndiaye et al., 2018). In the

case of H. ostrearia, the composition of the culture medium ES1/3 (Robert, 1983), which

is used as a reference, was also found to be prone to chemical precipitation. This

phenomenon limits biomass and EMn productivity.

It was found essential to design a new specific growth medium for H. ostrearia

culture in a PBR in order to understand, investigate and optimize H. ostrearia culture. It

was also found difficult to fully optimize pH and temperature because of their relationship

with the precipitation phenomenon. Further improvement of the culture medium and

Page 249: Par « Rémi NGHIEM-XUAN

223

conditions (being related) is still possible. The use of P, Si and C fed-batch was found to be

a useful strategy and could be used for this purpose.

Among the various parameters, light as represented by the MRPA was found to be

highly relevant. It impacts biomass productivity and also EMn synthesis, pigment and

overall conversion efficiency. However, the particular response to low MRPA value

remains to be further investigated, as its behavior towards pigment evolution was found

to be very different here than with other well-known organisms such as Chlorella vulgaris.

Hypothetically, this bias in cell response could come from the EMn synthesis and be

confirmed by the study of IMn production to light conditions.

The marennine synthesis pathway was not elucidated in our study (as this was not

the main aim of this paper), but it was confirmed that continuous EMn production and

excretion occur in a PBR culture, even without nutrient limitation. Overproduction was

obtained by applying silicate or nitrogen limitation. As a consequence, nutrient

concentration variation and its consequences for culture cell physiology remains to be

investigated.

To conclude, once all optimum parameters are collected and applied, the culture of

H. ostrearia cells in a stirred airlift PBR is possible. An EMn productivity of 4.5 ± 0.2 mgEMn

L-1 d-1 was achieved. The new NX medium was found to be stable at pH 7.8 and the P, Si, C

fed-batch could be avoided. We succeeded in scaling up H. ostrearia culture to a 6 L PBR.

A further study will investigate the continuous recovery of EMn by using, for

example, membrane technology. This will enable investigation of the interest in keeping

pelagic cells in the culture volume to further increase EMn productivity in a mixed PBR.

Page 250: Par « Rémi NGHIEM-XUAN

224

7 References

Artu A. (2016).: Etude et optimisation de la culture de microalgues en photobioréacteurs

solaires. (Thèse de doctorat ; Université de Nantes).

Bajwa K., Bishnoi N. R., Kirrolia A., Tamil Selvan S. (2007). Evaluation of nutrient stress

(nitrogen, phosphorus regimes) on physio-biochemical parameters of oleaginous

microalgal strains and SEM study under nutrient stress. International Journal of

Environmental Sciences & Natural Resources, 10 (1), doi:

10.19080/IJESNR.2018.10.555776.

Barbosa M. J., Hadiyanto, Wijffels R. H. (2004). Overcoming Shear Stress of Microalgae

Cultures in Sparged Photobioreactors. Biotechnology and Bioengineering, 85:78-85.

http://doi.org/10.1002/bit.10862.

Ben-Amotz A., Avron M. (1983). Accumulation of metabolites by halotolerant algae and its

industrial potential. Annual Review of Microbiology, 37: 95-119 doi:

10.1146/annurev.mi37.100183.000523.

Cardozo K. H-M., Guaratini T., Barros M. P., Falcão V. R., Tonon A. P., Lopes N. P., Campos

S., Torres M. A., Souza A. O., Colepicolo P., Pinto E. (2007). Metabolites from algae with

economical impact. Comparative Biochemistry and Physiology C. Elsevier Inc. 146 (1–

2): 60–78. doi:10.1016/j.cbpc.2006.05.007.

Chisti Y., Moo-Young M. (1989). Calculation of shear rate and apparent viscosity in airlift and

bubble column bioreactors. Biotechnology and Bioengineering, 34, 0–1.

Contreras A., Garcia F., Molina E., Merchuk J. C. (1998). Interaction between CO2-mass

transfer, light availability, and hydrodynamic stress in the growth of Phaeodactylum

tricornutum in a concentric tube airlift photobioreactor. Biotechnology and

Bioengineering, 60: 317 – 325.

Cornet J-F., Dussap C. G., Gros. J-B. (1998). Kinetics and energetics of photosynthetic micro-

organisms in photobioreactors. In Bioprocess and Algae reactor technology, apoptosis,

153–224. Advances in Biochemical Engineering Biotechnology 59. Springer Berlin

Heidelberg.

Gabelle J.C., Jourdier E., Licht R., Ben Chaabane F., Henaut I., Morchain J., Augier F. (2012).

Impact of rheology on the mass transfer coefficient during the growth phase of

Trichoderma reesei in stirred bioreactors. Chemistry Engineering Science, 75, 408-417.

doi:1 0.1016/j.ces.2012.03.053.

Gaillon B. (1820). Des huîtres vertes, et des causes de cette coloration. Journal de physique, de

chimie, d’histoire Naturelle et des Arts. 91, 222-225.

Gastineau R., Pouvreau J-B., Hellio C., Morançais M., Gaudin P., Bourgougnon N., Mouget J-

L. (2012). Biological Activities of Purified Marennine, the Blue Pigment Responsible for

the Greening of Oysters. Journal of Agriculture Food and Chemistry. 60(14):3599-605.

http://doi.org/10.1021/jf205004x.

Page 251: Par « Rémi NGHIEM-XUAN

225

Goiris K., Van Colen W., Wilches I., León-Tamariz F., De Cooman L., Muylaert K. (2015).

Impact of nutrient stress on antioxidant production in three species of microalgae. Algal

Research, vol.7, 51-57.

Grasshoff K., Kremling K., Anderson L. G., Andreae M. O., Behrends B., Yhlen B. (2007).

Methods of Seawater Analysis (WILEY‐VCH Verlag GmbH). Revised Completely.

Hochfeld W. L. (2005). Producing Biomolecular Substances with Fermenters, Bioreactors, and

Biomolecular Synthesizers, Boca Raton, FL, Taylor & Francis, ISBN 9781420021318.

Kandilian R., Pruvost J., Legrand J., Pilon L. (2014). Influence of light absorption rate by

Nannochloropsis oculata on Triglyceride production during Nitrogen starvation.

Bioresource Technology, 163 (July): 308–19.

Kandilian R., Pruvost J., Artu A., Lemasson C., Legrand J., Pilon L. (2016). Comparison of

experimentally and theoretically determined radiation characteristics of photosynthetic

microorganisms. Journal of Quantitative Spectroscopy and Radiative Transfer, 175, 30-

45. https://doi.org/10.1016/j.jqsrt.2016.01.031.

Kandilian R., Bruno J., Legrand J., Pilon L., Pruvost J. (2017). Light transfer in agar

immobilized microalgae cell cultures. Journal of Quantitative Spectroscopy and Radiative

Transfer, 198, 81-92.

Kazbar A. (2018). Etude de l’impact de la concentration en oxygène dissous et la présence

d’une fraction sombre sur la performance des photobioréacteurs. (Thèse de doctorat;

Université de Nantes).

Lebeau T., Gaudin P., Junter G.-A., Mignot L., Robert J-M. (2000). Continuous marennin

production by agar-entrapped Haslea ostrearia using a tubular photobioreactor with

internal illumination. Applied Microbiology Biotechnology, 54, 634–640.

Le Gouic B. (2013). Analyse et optimisation de l’apport de carbone en photobioréacteur (Thèse

de doctorat; Université de Nantes).

Michels M. H. A., Atze J., Van Der Goot J., Norsker N-H. R., Wijffels, R. H. (2010). Effects

of shear stress on the microalgae Chaetoceros muelleri, Journal of Applied Phycology,

28(1), 53-62. http://doi.org/10.1007/s00449-010-0415-9.

Mouget J-L., Rosa P., Tremblin G. (2004). Acclimation of Haslea ostrearia to light of different

spectral qualities – confirmation of “chromatic adaptation” in diatoms. Journal of

Photochemistry and Photobiology, 19;75(1-2):1-11.

http://doi.org/10.1016/j.jphotobiol.2004.04.002.

Mouget J-L., Rosa P., Vachoux C., Tremblin G. (2005). Enhancement of marennine production

by blue light in the diatom Haslea ostrearia. Journal of Applied Phycology, 17, (5), 437–

44517, 437–445. http://doi.org/10.1007/s10811-005-0561-7.

Morey G. W., Fournier R., Rowe J. (1995). The solubility of amorphous silica at 25°C. Journal

of Geophysical Research, 69(10), 1995–2002.

Page 252: Par « Rémi NGHIEM-XUAN

226

Moutel B. (2016). Étude de l'intérêt pharmaceutique et d'une production industrielle des lipides

issus de la microalgue Botryococcus braunii (Thèse de doctorat; Université de Nantes).

Neuville D., Daste P. (1972). Production du pigment bleu par la diatomée Navicula ostrearia

(Gaillon) Bory, maintenue en culture uni-algue sur milieu synthétique carencé en azote.

Compte Rendu de l’Académie des Sciences (Paris), 374 : 2030-2033.

Ndiaye M., Gadoin E., Gentric C. (2018). CO2 gas-liquid mass transfer and kLa estimation:

Numerical investigation in the context of airlift photobioreactor scale-up. Chemical

Engineering Reasearch and Design, 133, 90-102.

Nghiem-Xuan R., (2019). Optimisation de la culture d’Haslea ostrearia en photobioréacteur

(Thèse de doctorat ; Université du Nantes).

Nguyen T. D. P. (2013). Récolte de biomasse microalgale par floculation naturelle et procédés

membranaires (Thèse de doctorat; Université de Nantes).

Oren A. (2005). A hundred years of Dunaliella research: 1905-2005. Saline systems, 1:2

Doi:10.1186/1746-1448-1-.

Panis G., Carreon J. R. (2016). Commercial astaxanthin production derived by green alga

Haematococcus pluvialis: A microalgae process model and a techno-economic assessment

all through production line. Algal Research, 18, 175–190.

http://doi.org/10.1016/j.algal.2016.06.007.

Pilon L., Berberoğlu H., Kandilian R. (2011). Radiation Transfer in Photobiological Carbon

Dioxide Fixation and Fuel Production by Microalgae. Journal of Quantitative

Spectroscopy and Radiative Transfer 112 (17): 2639–60.

Pouvreau J-B., Morançais E., Fleury F., Rosa P., Thion L., Cahingt B. (2006). Preliminary

characterisation of the blue-green pigment “marennine” from the marine tychopelagic

diatom Haslea ostrearia (Gaillon/Bory) Simonsen. Journal of Applied Phycology, 18,

757–767. http://doi.org/10.1007/s10811-006-9087-x.

Pruvost J., Cornet J-F. Legrand J. (2008). Hydrodynamics influence on light conversion in

photobioreactors: An energetically consistent analysis. Chemical Engineering Science 63

(14): 3679–94.

Pruvost J., Van Vooren G., Cogne G., Legrand J. (2009). Investigation of biomass and lipids

production with Neochloris oleoabundans in photobioreactor. Bioresource Technology

100 (23): 5988-95.

Pruvost J., Cornet J-F. (2012). Knowledge models for the engineering and optimization of

photobioreactors. Microalgal Biotechnology: Potential and Production. De Gruyter,

Berlin, Germany, 181–224.

Rincé Y., Lebeau T., Robert J. M. (1999). Artificial cell-immobilization: a model simulating

immobilization in natural environments? Journal of Applied Phycology, 11, 263–272.

Page 253: Par « Rémi NGHIEM-XUAN

227

Ritchie R. J. (2008). Universal chlorophyll equations for estimating chlorophylls a, b, c, and d

and total chlorophylls in natural assemblages of photosynthetic organisms using acetone,

methanol, or ethanol solvents. Photosynthetica, 46(1), 115–126.

Robert J-M. (1983). Fertilité des claires ostréicoles et verdissement: utilisation de l’azote par

les diatomées dominantes. (Thèse de Doctorat : Université de Nantes).

Robert J-M. (1989). Procédé de culture en masse de la diatomée Haslea ostrearia Simonsen.

European and French patent no 8915314. Nantes University.

Rodolfi L., Chini Zittelli G., Bassi N., Padovani G., Biondi N., Bonini G. (2009). Microalgae

for oil: strain selection, induction of lipid synthesis and outdoor mass cultivation in a low-

cost photobioreactor. Biotechnology and Bioengineering. 102, 100–112,

http://dx.doi.org/10.1002/bit.22033.

Rossignol N. (1999). Procédés d’extraction et de séparation par membranes appliqués à la

production du pigment endo-et exocellulaire synthétisé par la diatomée Haslea ostrearia

Simonsen. Mise en œuvre d’un photobioréacteur à membranes à fonctionnement continu.

(Thèse de doctorat ; Université de Nantes).

Rouillard I. (1996). Optimisation de la production en masse d'Haslea ostrearia Simonsen sur

une eau souterraine salée: importance de la souche et des conditions de culture;

comparaison avec Skeletonema costatum (Grev) Cleve. (Thèse de doctorat ; Université de

Nantes).

Shah M. R., Liang Y., Cheng J. J., Daroch M. (2016). Astaxanthin-Producing Green Microalga

Haematococcus pluvialis: From Single Cell to High Value Commercial Products.

Frontiers in Plant Science, 7: 531. Doi: 10.3389/fpls.2016.00531.

Sobczuk T.M., Camacho F.G., Grima E.M., Chisti Y. (2006). Effects of agitation on the

microalgae Phaeodactylum tricornutum and Porphyridium cruentum. Bioprocess and

Biosystem Engineering. 28, 243–250.

Souliès A. (2014). Contribution à l’étude hydrodynamique et à la modélisation des

photobioréacteurs à haute productivité volumique (Thèse de doctorat; Université de

Nantes).

Spolaore P., Joannis-Cassan C., Duran E., Isambert A. (2006). Commercial Applications of

Microalgae. Journal of Bioscience and Bioengineering, 101(2), 87–

96.http://doi.org/10.1263/jbb.101.87.

Starckx S. (2012). A place in the sun - Algae is the crop of the future, according to researchers

in Geel, Flanders Today, Retrieved 8 December 2012.

Strickland J. D., Parsons T. (1968). A Pratical Handbook of Seawater Analysis. Fisheries

Reasearch Bulletin. Bd. Canada.

Takache H., Pruvost J., Cornet J-F. (2012). Kinetic modeling of the photosynthetic growth of

Chlamydomonas reinhardtii in a photobioreactor. Biotechnology Progress, 28 (3): 681-

92.

Page 254: Par « Rémi NGHIEM-XUAN

228

Turpin V., Robert J-M.,Goulletquer P. (1999). Limiting nutrients of oyster pond seawaters in

the Marennes-Oléron region for Haslea ostrearia: applications to the mass production of

the diatom in mesocosm experiments. Aquatic Living Resources 12, 335-342.

Turpin V., Robert J-M., Goulletquer P., Massé G., Rosa P. (2001). Oyster greening by outdoor

mass culture of the diatom Haslea ostrearia Simonsen in enriched seawater. Aquaculture

Rasearch, 32(801-809).

Wang C., Lan C. Q. (2018). Effects of shear stress on microalgae – A review. Biotechnology

Advances, 36(4): 986 - 1002. http://doi.org/10.1016/j.biotechadv.2018.03.001.

Page 255: Par « Rémi NGHIEM-XUAN

229

Conclusions et perspectives

Le travail présenté dans ce manuscrit de thèse a permis de découvrir une partie

des problématiques de la culture d’Haslea ostrearia en photobioréacteur, et ainsi

apprendre à maîtriser et orienter la production de cette microalgue à fort potentiel, mais

dont la culture optimisée reste un enjeu en vue de son exploitation contrôlée.

Une première problématique identifiée durant cette thèse a été la variabilité des

résultats en milieu d’eau de mer enrichie (Turpin et al., 1999 ; Gagneux-Moreaux et al.,

2007 ; Mouget et al., 2009). La composition et la concentration des éléments nutritifs de

ces milieux sur base naturelle entraine un phénomène de précipitation limitant la

croissance d’H. ostrearia en PBR. Les milieux minéraux d’eau de mer artificiels (NX-

5N10P6Si et NX-25N100P6Si) élaboré par une approche stœchiométrique de la

composition élémentaire de la biomasse, ont permis la diminution du phénomène de

précipitation, augmentant respectivement, les productivités en EMn et en biomasse. Les

interactions chimiques entre les ions Ca2+, Mg2+, PO43-, HCO3- et SiO32- provenant de la

solution d’enrichissement, ont été confirmées par simulation des équilibres physico-

chimiques. D’autre part, l’étude de l’impact de chaque nutriment sur les productivités en

biomasse et en EMn démontre que 10 fois moins de Calcium dans le milieu entraîne une

augmentation de productivité d’un facteur 5,5 en cellules, et 2 en EMn. La concentration

en Magnésium n’a pu être ni baissée ni augmentée sans provoquer des pertes de

productivité. L’augmentation des concentrations de Fer et de Zinc ainsi que la baisse de

Bore et de Nickel, ont permis d’obtenir de meilleures productivités en biomasse et en

marennine extracellulaire ainsi qu’un contrôle des conditions nutritionnelles en PBR.

Cette nouvelle composition de milieu nous permet d’envisager l’augmentation de

macronutriments (N, P, Si) dans le but d’augmenter le seuil de productivité que ce milieu

permet d’atteindre en PBR. De plus, il est possible d’utiliser cette approche pour

l’élaboration d’autres milieux de culture. Cependant, à pH>8, une précipitation est encore

présente, et s’avère létale pour les microalgues à pH>9. Une étude physico-chimique du

Page 256: Par « Rémi NGHIEM-XUAN

230

milieu de culture est donc encore nécessaire pour éviter le phénomène de précipitation

avant d’envisager l’augmentation des concentrations de macronutriments.

Le milieu NX est un milieu minéral, mais les études sur l’impact des sources de

Carbone et de Phosphate ont montré qu’il est possible d’obtenir de bons résultats de

productivité en EMn et en biomasse avec des sources organiques de nutriments. La

mixotrophie d’H. ostrearia n’a pas été démontrée dans cette thèse, mais les diatomées

possèdent une gestion de l’énergie chimique (ATP et NADPH) différente de celle des

chlorophycées pour la fixation du Carbone. Un échange rapide entres ces molécules existe

entre le chloroplaste et la mitochondrie, permettant de combler rapidement le déficit du

rapport ATP/NADPH de la photosynthèse grâce à l’énergie produite par respiration

(Bailleul et al., 2015). Le fait qu’H. ostrearia possède potentiellement cette capacité,

soulève l’hypothèse d’une alternative mixotrophe au milieu inorganique NX. Quoi qu’il en

soit, bien que l’utilisation de l’eau de mer naturelle puisse être préférable à une

production à grand volume, le milieu artificiel NX de composition connue et contrôlée,

apporte une reproductibilité dans l’étude de la culture de cette algue et de la production

de la marennine à l’échelle du laboratoire. A l’avenir, la méthodologie d’enrichissement

combinée à une prédiction des limites de dissolution et de précipitations pourra être

étendu au cas de milieux à base d’eau de mer naturelle. Ainsi, même si celle-ci contient

des concentrations élevées en Ca2+ par exemple, il sera possible d’appliquer une stratégie

fed-batch d’enrichissement progressif afin d’apporter les nutriments nécessaires à la

croissance, tout en gardant un milieu de croissance chimiquement stable.

Bien que l’impact du milieu de culture ait été montré dans la limitation de la

croissance d’H. ostrearia en PBR, l’impact des conditions d’agitation soulève encore des

problématiques de culture. Malgré les travaux du GEPEA (Rossignol et al., 1999 ;

Vandanjon et al., 1999) et ceux de cette thèse. L’étude du phénomène de mécano-

sensitivité se heurte à la mesure de réponse biologique d’H. ostrearia, la différence des

conditions appliquées entre les expériences à hydrodynamique maîtrisée et les réelles

conditions turbulentes en PBR. La sensibilité d’H. ostrearia n’a donc pu être rapportée à

une grandeur type liée aux contraintes de cisaillement, car l’origine même de sa sensibilité

ne semble pas être entièrement physique mais est également due à sa physiologie.

La sensibilité générale des diatomées face à l’agitation est connue et présentée

dans la littérature (Wang et Lan, 2018). Dans le cas d’H.ostrearia, l’hypothèse de la

Page 257: Par « Rémi NGHIEM-XUAN

231

sensibilité due à son système de division cellulaire est la plus plausible à notre niveau de

compréhension de cette microalgue. La baisse de productivité en biomasse observée

durant la culture en PBR torique, indique une difficulté d’H. ostrearia à supporter le

régime turbulent, même à faible cisaillement (20 s-1) et en agitation continue. Pourtant, la

même intensité de 8h d’agitation par jour provoque une augmentation de productivité de

biomasse d’un facteur 1,8 par rapport à une culture non agitée. Il a pu être montré que

l’agitation permet aussi d’augmenter l’accessibilité des cellules à la lumière, une

corrélation directe ayant en effet à pu être montrée avec la remise en suspension des

cellules par l’agitation. Ainsi, du fait de la sédimentation rapide des cellules menant

naturellement à protéger les cellules de la lumière, une durée trop longue d’agitation peut

amener les cellules en suspension à recevoir un excès lumineux sur une journée. D’autre

part, si on suppose que la synchronisation de la population cellulaire ne se déroule qu’en

période de nuit, une agitation de 8h alternée n’engendre des contraintes de cisaillement

que seulement durant la période de jour. Ainsi, ce gain de productivité pourrait autant

venir de la sensibilité d’Haslea ostrearia en phase de division cellulaire, que d’une

meilleure exposition à la lumière. La baisse de concentration en pigment

photosynthétique tend à corroborer l’hypothèse d’une réponse à la lumière. Cependant,

ce point mériterait d’être approfondi, notamment en étudiant plus en détail la relation

entre croissance et lumière absorbée (MRPA), elle-même modifiée par la stratégie de mise

en suspension des cellules.

Compte-tenu des résultats obtenus, une stratégie de 8h d’agitation par jour en PBR

torique a été mis en place, avec pour but d’améliorer l’accès à la lumière des cellules tout

en évitant les stress hydrodynamiques pendant la nuit et/ou la surexposition lumineuse.

Même si cela s’est avéré concluant, une optimisation de cette stratégie reste possible par

une étude du comportement d’H. ostrearia en mode tychopélagique de croissance

(cellules en suspension), pour contrôler au mieux l’impact de la lumière sur le

comportement cellulaire et d’isoler ce comportement de celui provoqué par l’agitation.

Concernant la stratégie d’agitation en PBR Airlift, le comportement tychopélagique d’H.

ostrearia peut en effet s’avérer clé, car cette stratégie permet à une partie de la population

cellulaire de se fixer au fond du PBR (comportement benthique) et de limiter l’impact de

l’agitation turbulente générée par le gaz, alors qu’une remise en suspension des cellules

favorise la réception de lumière pour les cellules au comportement tychopélagique.

Visuellement, les cellules se sont révélées différentes, laissant présager une physiologie

Page 258: Par « Rémi NGHIEM-XUAN

232

différente également. Une étude plus précise de la physiologie tychopélagique/benthique

d’H. ostrearia permettrait une meilleure compréhension et un meilleur contrôle de la

stratégie d’agitation (vis-à-vis de la croissance et de la marennine).

Considérant les cisaillements provoqués par un régime laminaire, la résistance sur

courte durée d’H. ostrearia a été confirmée et à de très hauts taux de cisaillement non

présents en PBR (jusqu’à 30 000 s-1). La forme pennée de la microalgue est l’hypothèse la

plus plausible pour expliquer nos résultats, lui donnant des propriétés anisotropes lui

permettant de diminuer l’impact des cisaillements sur elle-même selon son orientation

dans le fluide (Mauer et al., 2016). Cependant, comparée à la résistance d’autres

diatomées en régime laminaire, H. ostrearia est capable de supporter une intensité de

cisaillement d’un facteur 4,5 plus important que Phaeodactylum tricornutum et

Chaetoceros muelleri (Sobczuk et al., 2005 ; Michiel et al., 2010). Le régime laminaire est

très rare en PBRs conventionnels (Olmos et al., 2003 ; Pruvost et al., 2006 ; Ndiaye et al.,

2018 ; Wang et Lan, 2018). Envisager la conception d’un système de culture

spécifiquement dédié pour H. ostrearia en conditions de régime laminaire uniforme,

permettrait l’application d’agitation et l’optimisation du transfert de matière (élimination

du gradient T°C et pH) sans altérer l’intégrité cellulaire. A noter que des réacteurs à flux

laminaire (LFR) reposant sur le contrôle d’un régime laminaire au sein du PBR sont

communément utilisé en génie des procédés et pourraient servir d’inspiration.

Concernant la réponse d’H. ostrearia à la lumière, nos études ont permis de

confirmer des hypothèses précédemment mentionnées dans la littérature mais en

soulèvent aussi d’autres. H. ostrearia possède un besoin en énergie lumineuse (AOPT=12

µmolhv. gx-1. s-1) similaire à ceux de Chlorella vulgaris et Nannochloropsis gaditana (Artu,

2016). Cependant, nous avons observé une variation pigmentaire importante dans une

gamme de MRPA de 8,4 à 12,3 µmolhv. gx-1. s-1, correspondant respectivement aux AOPT

EMn et AOPT biomasse. L’hypothèse d’une utilisation de la marennine (hypothétiquement

intracellulaire) dans la photosynthèse chez H. ostrearia est donc plausible. Comprendre

les conditions de production d’un exo-métabolite tel que la marennine est une

problématique importante pour la conception optimisée d’un procédé de culture et

d’extraction de ce composé.

En débat dans la communauté scientifique (Robert, 1983 ; Neuville et Daste, 1972 ;

Rossignol, 1999 ; Lebeau et al., 1999), l’origine de la marennine reste d’une façon générale

Page 259: Par « Rémi NGHIEM-XUAN

233

à élucider. Les résultats de cette thèse ont permis d’identifier quelques stress

physiologiques induisant la surproduction de l’EMn. Les résultats démontrent que les

carences en nutriments provoquent chez H. ostrearia une forte diminution voire un arrêt

de l’excrétion de l’EMn, alors que les limitations en Silice et en Azote permettent la

stimulation de l’excrétion. Les faibles résultats de productivité (en cellule et en EMn)

obtenus en PBR à cellules immobilisées (actives et passives) ont conduit à remettre en

question ce mode de culture. Au final, H. ostrearia a donc été cultivée en PBR agité airlift

dans un milieu NX pour optimiser la production de biomasse, puis cultivée dans un milieu

NX (limité N) en continu pour stimuler l’excrétion d’EMn. Cette technique de culture

permet d’obtenir des productivités de 4.5±0.16 mg L-1 d-1 en EMn ce qui correspond à 1,8

fois les productivités publiées par Rossignol (1999) en PBR à membrane immergée

(Figure 81).

La comparaison aux PBR à membrane immergée permet d’envisager une autre

piste d’optimisation. En effet, dans notre étude et contrairement aux travaux de Rossignol

(1999), la récolte d’EMn a été effectuée par surverse en semi-continu, ce qui entraîne une

partie des cellules (principalement tychopélagiques). Il est intéressant de constater qu’H.

ostrearia excrète de l’EMn en continu même en conditions non limitées par les nutriments.

Cela nous permet d’envisager d’optimiser les productivités en EMn du procédé (en

adaptant par exemple les concentrations du milieu de culture pour limiter la cellule en

Azote) en maintenant les cellules dans le PBR grâce aux membranes immergées. On peut

aussi ajouter qu’une concentration importante d’une molécule active tel que l’EMn dans

le milieu soulève l’hypothèse d’un impact négatif de la molécule sur le comportement d’H.

ostrearia (Robert et Turpin, 1993 ; Rossignol, 1999). Les travaux deRossignol (1999) sur

la filtration en continu d’EMn par membrane immergé permettrait donc de palier à ce

potentiel impact négatif de la marennine sur la productivité. Nous avons ainsi estimé qu’il

était encore possible d’augmenter d’un facteur 3 nos productivités volumiques en EMn

dans les conditions de culture optimales identifiées dans cette thèse (PVestimé = 13,5 mgEMn

L-1 j-1).

Page 260: Par « Rémi NGHIEM-XUAN

234

Figure 81: Représentation chronologique actualisée des travaux successifs visant à améliorer la production en marennine extracellulaire selon les différentes technologies de systèmes de culture mises en place par les laboratoires MMS, du GEPEA et de l'UQAR

Page 261: Par « Rémi NGHIEM-XUAN

235

Pour conclure de façon générale sur la culture d’H. ostrearia, les cultures effectuées

en PBRs à cellules immobilisées ont permis jusqu’ici de cultiver H. ostrearia (Rincé et al.,

1999 ; Rossignol, 1999 ; Lebeau et al., 2000 ; Turpin et al., 2001), mais la montée en échelle

de ce type de culture reste une problématique, un tel système apporte aussi des

problématiques propres de transfert de matière au sein du biofilm, avec possible

apparition de gradients localisés de pH, de température, de nutriments (Hochfeld, 2005 ;

Gabelle et al., 2012 ; Souliès, 2014 ; Ndiaye et al., 2018). L’utilisation de systèmes de

culture agités conventionnels (tel que airlift) pour la culture d’H. ostrearia s’est avérée

une alternative possible. Cependant cette algue est sensible aux pompes ainsi qu’à

l’agitation en régime turbulent, notamment avec un couplage qui apparait entre cycles

d’éclairement et périodes d’agitation. Un système d’agitation alterné ou spécifique

(régime laminaire) doit être appliqué. Les études présentées dans cette thèse nous

permettent de montrer qu’une agitation locale d’intensité conventionnelle (20 – 15 000 s-

1) en période de jour est favorable à la culture de cette microalgue. Une homogénéisation

du PBR engendre un meilleur transfert de matière, favorablement combiné ici avec

l’utilisation d’un milieu NX stable au niveau physico-chimique. La mise en suspension des

cellules permettra d’optimiser les conditions d’accès à la lumière, et donc une

optimisation globale de la croissance photosynthétique. La solution de récolte de EMn par

membrane de filtration est aussi une piste très intéressante, car cela permet de limiter

l’impact potentiel de l’EMn sur la croissance d’H. ostrearia, tout en évitant le lessivage de

la culture provoqué par un comportement tychopélagique de cette microalgue.

D’un point de vue applicatif, aux vues de nos résultats, on peut conclure que la

montée en échelle de la culture d’H. ostrearia est donc possible. Nous avons ainsi identifié

quelques PBR sélectionnés selon leurs types d’agitation permettant d’imposer des

périodes d’agitation, et leur capacité à générer des taux de cisaillement supportable pour

H. ostrearia (Tableau 14). Cela pourrait être dans un PBR « colonnes à bulles », un PBR

airlift de type Hector, ou un Raceway clos (voir le Tableau ci-dessous). Cependant, chacun

demandera une étude dédiée pour notamment adapter les stratégies d’agitation. Les

données collectées dans nos travaux pourront alors être utiles pour définir les durées

d’agitation (notamment obtenir une valeur de MRPA optimale sur une journée, en jouant

sur la capacité naturelle de la souche à sédimenter, et ainsi se protéger des excès de

lumière). Cela permet d’envisager des solutions de contrôle des conditions de culture d’H.

ostrearia (milieu, stratégie d’agitation et de sous-titrage de la marennine extracellulaire),

Page 262: Par « Rémi NGHIEM-XUAN

236

et d’envisager une monté en échelle rationnelle de la culture pour sa valorisation

industrielle.

Tableau 14: Comparaison de l'intensité et la capacité à générer des contraintes hydrodynamiques du PBR Colonne à bulle, Hector et Raceway clos.

PBR Type VR (L)

γestimé

(s-1)

Agitation Régulation

pH

Régulation

T°C

Colonne à

bulles Cylindre 130 150 - 275 Airlift

CO2 oui Hector Plan 150 150 - 275 Airlift

Raceway

clos Plan 700-1500 <275

Roue à

aubes

Page 263: Par « Rémi NGHIEM-XUAN

237

Nomenclature

Hydrodynamique et cisaillement

γ Taux de cisaillement s-1

μ Viscosité dynamique Pa. s-1

ν

Viscosité cinématique m2. s-1

Q Débit L. h-1

Re Nombre de Reynolds -

τ Contrainte de cisaillement Pa

Ta Nombre de Taylor -

Ug Vitesse superficielle du gaz m. s-1

Transfert de lumière au sein d’un photobioréacteur

A ou Alight Vitesse spécifique

moyenne d’absorption de

photons (MRPA)

μmolhv. gx-1. s-1

A24h, suspended Vitesse spécifique

moyenne d’absorption de

photons (MRPA) estimé en

fonction de la condition

d’agitation

μmolhv. gx-1. s-1

as Surface spécifique éclairée m-1

CX Concentration en biomasse

sèche

g. L-1

γ Fraction volumique

éclairée du PBR

-

Ea Coefficient massique

d’absorption

m². kg-1

FV/Fm Fraction de l’énergie

photonique absorbée

convertible en énergie

-

Page 264: Par « Rémi NGHIEM-XUAN

238

chimique (mesure de

stress chez la plante)

G Irradiance locale μmolhv. m-2. s-1

L Profondeur de culture m

η L’efficacité de conversion

de la lumière

gx. molhv-1

q0 Flux incident en surface μmolhv. m-2. s-1

qL Flux lumineux en sorti de

PBR

µmolhv. m-2. s-1

V Volume du réacteur L

Production en photobioréacteur

PV Productivité volumique Cell. L-1. d-1 or g. L-1. d-1

PS Productivité surfacique Cell. m-2. d-1 or g. m-2. d-1

N Concentration cellulaire Cell. L-1

D Taux de dilution en culture

semi-continue ou continue

j-1

rS Vitesse de consommation

de substrat

µmol. L-1. d-1

QS Vitesse spécifique de

consommation de substrat

µmol. gx-1. d-1

Yx Rendement de conversion

du substrat en biomasse

µmol. gx-1

P EMn specific Productivité spécifique de

marennine extracellulaire

mg. cell-1. d-1

DW or MS Matière sèche g. L-1

Page 265: Par « Rémi NGHIEM-XUAN

239

Bibliographie

Abidov M., Ramazanov Z., Seifulla R., Grachev S., (2010). The effects of Xanthigen™ in the

weight management of obese premenopausal women with non-alcoholic fatty liver disease

and normal liver fat. Diabetes Obesity Metabolism, 12, 72–81.

Akhnoukh R., Kretzmer G. and Schiigerl K., (1993). Investigation of the cultivation of

Spodoptera frugiperda in spinner flasks and in a 10 L bioreactor with bubble-free aeration.

Bioprocess Engineering, 8, 229-233.

Andersen R. A., (2005). Algal culturing techniques. Edited by Robert A. Andersen; Elsevier

Academic Press. 578 p.Bailleul B., Berne N., Murik O., Petroutsos D., Prihoda J., Tanaka

A., Blign V. V. R., Flori S., Falconet D., Krieger-Liszkay A., Santabarbara S., Rappaport

F., Tirichine P. J., & L., Falkowski P. G., Cardol P., C. B. & G. F., (2015). Energetic

coupling between plastids and mitochondria drives CO2 assimilation in diatoms. Nature,

524, 366–369. http://doi.org/10.1038/nature14599

Bachrach E (1935) Le bleuissement des diatomées et le verdissement des huîtres. Revue des

Travaux de l’Institut des Pêches Maritimes 8 :112–123

Barbosa M. J., Hadiyanto & Wijffels R. H., (2004). Overcoming Shear Stress of Microalgae

Cultures in Sparged Photobioreactors. Biotechnology and Bioengineering, 85:78-85.

http://doi.org/10.1002/bit.10862

Barillé L., & Cognie B., (2000). Revival capacity of diatoms in bivalve pseudofaeces and

faeces. Diatom Research, 15, 11–17.

Belay A., Kato T., Ota Y., (1996). Spirulina (Arthrospira): potential application as an animal

feed supplement. Journal of Applied Phycology, (1990), 303-311

Bergman L. T., (2011). Fundamentals of heat and mass transfer - 7th edition. (NJ Wiley, Ed.)

(Hoboken, Vol. xxIII).

Borowitzka L. J., and M. A. Borowitzka, (1989). “β-Carotene (Provitamin A) Production with

Algae.” In Biotechnology of Vitamins, Pigments and Growth Factors, edited by Erick J.

Vandamme, 15–26. Elsevier Applied Biotechnology Series, Springer Netherlands.

Borowitzka Michael A., (1999). “Commercial Production of Microalgae: Ponds, Tanks, Tubes

and Fermenters.” Journal of Biotechnology, Biotechnological Aspects of Marine Sponges,

70 (1–3): 313–21.

Bronnenmeier R., & Markl H., (1982). Hydrodynamic Stress Capacity of Microorganisms.

Biotechnology and Bioengineering, XXIV, 553–578.

Page 266: Par « Rémi NGHIEM-XUAN

240

Butterwick C., Heaney S. I., Talling J. F., (2005). Diversity in the influence of temperature on

the growth rates of freshwater algae, and its ecological relevance. Freshwater Biology, 50

(2): 291-300

Camacho F.G., Gallardo Rodríguez J.J., Sánchez Mirón A., Cerón García, M.C., Belarbi E.H.,

Molina Grima E., (2007). Determination of shear stress thresholds in toxic dinoflagellates

cultured in shaken flasks. Process Biochemistry, 42, 1506–1515.

http://dx.doi.org/10.1016/j.procbio.2007.08.001.

Carreto J. I., & Catoggio J. A., (1976). Variations in Pigment Contents of the Diatom

Phaeodactylum tricornutum during Growth. Marine Biology.112(646), 105–112.

Castro-Ceseña A. B., & Del Pilar Sánchez-Saavedra M., (2014). Evaluation of sodium

tripolyphosphate-alginate coating and re-calcifying on the entrapment of microalgae in

alginate beads. Journal of Apllied Phycology, 27. http://doi.org/10.1007/s10811-014-

0411-6

Chaumont Daniel., (1993). “Biotechnology of Algal Biomass Production: A Review of Systems

for Outdoor Mass Culture.” Journal of Applied Phycology, 5 (6): 593–604.

Chen Chun-Yen, Kuei-Ling Yeh, Rifka Aisyah, Duu-Jong Lee, and Jo-Shu Chang., (2011).

Cultivation, Photobioreactor Design and Harvesting of Microalgae for Biodiesel

Production: A Critical Review. Bioresource Technology, Special Issue: Biofuels - II: Algal

Biofuels and Microbial Fuel Cells, 102 (1): 71–81.

Chi Z., VO’Fallon J., Chen S., (2011). Bicarbonate produced from carbon capture for algae culture.

Trends in Biotechnology, 29, I11,537–541.https://doi.org/10.1016/j.tibtech.2011.06.006

Chisti Y. and Moo-Young M., (1989). Calculation of Shear Rate and Apparent Viscosity in

Airlift and Bubble Column Bioreactors. Biotechnology and Bioengineering, 34, 0–1.

Chittur K. K., McIntire L. V., Rich R. R., (1892). Shear Stress Effects on Human T Cell

Function. Biotechnology Progress, 4(2), 89–96.

Climate Data, Courbe de Température Marennes et Tableau Climatique Marennes. Site :

http://fr.climate-data.org/location/66175/

CNC, Comité national de conchyliculture, les statistiques : La production Française, mondiale

et Européenne. Site : http://www.cnc-france.com/Les-statistiques.aspx

Cognie B., & Barikle L., (1998). Does bivalve mucus favour the growth of their main food

source, microalgae? Oceanologica Acta, 22, 4 (I), 441-450.

https://doi.org/10.1016/S0399-1784(00)88727-7

Cole K. M. & Sheath R. G., (1990). Biology of Red Algae. (Cambridge university press 1990,

Ed.).

Connor K. C. O., Cowger N. L., Kee D. C. R. De & Schwarz R. P., (2002). Prolonged shearing

of insect cells in a couette bioreactor. Enzyme and Microbial Technology, 31, 600–608.

Page 267: Par « Rémi NGHIEM-XUAN

241

Contreras A., Garcia F., Molina E., & Merchuk J. C., (1998). Interaction between CO2-mass transfer,

light availability, and hydrodynamic stress in the growth of Phaeodactylum tricornutum in a

concentric tube airlift photobioreactor. Biotechnology and Bioengineering.

Cornet J. F., C. G. Dussap and G. Dubertret., (1992). A Structured Model for Simulation of

Cultures of the Cyanobacterium Spirulina Platensis in Photobioreactors: I. Coupling

between Light Transfer and Growth Kinetics. Biotechnology and Bioengineering 40 (7):

817–25.

Cornet J. -F., C. G. Dussap, J. -B. Gros, C. Binois, and C. Lasseur., (1995). A Simplified

Monodimensional Approach for Modeling Coupling between Radiant Light Transfer and

Growth Kinetics in Photobioreactors. Chemical Engineering Science 50 (9): 1489–1500.

Cornet J.-F., C. G. Dussap, and J.-B. Gros., (1998). Kinetics and Energetics of Photosynthetic

Micro-Organisms in Photobioreactors. In Bioprocess and Algae Reactor Technology,

Apoptosis, 153–224. Advances in Biochemical Engineering Biotechnology 59. Springer

Berlin Heidelberg.

Crawford S. A., Higgins M. J., Mulvaney P., & Wetherbee R., (2001). Nanostructure of the

diatom frustule as revealed by atomic force and scanning electron microscopy. Journal of

Phycology, 37(4), 543-554).

Dassow P. Von, & Latz M. I., (2002). The role of Ca2+ in stimulated bioluminescence of the

dinoflagellate Lingulodinium polyedrum. The Journal of Experimental Biology, 2986,

2971–2986.

Davidovich N. A., Mouget J.-L., & Gaudin P., (2009). Heterothallism in the pennate diatom

Haslea ostrearia (Bacillariophyta) European Journal of Phycology, 44(2), 37–41.

http://doi.org/10.1080/09670260802710301

Deckwer W.-D., 1991. Bubble Column Reactors. Wiley. New York

De Reviers B., (2002). Biologie et phylogénie des algues, volume 1. Belin Sup Sciences:

Biologie, ISSN 1158-3762.

Dermoun D., Chaumont D., (1992). Modelling of growth of Porphyridium cruentum in

connection with two interdependent factors: light and temperature. Bioresource and

Technology 42:113-117.

Devred E., Turpie K. R., Moses W., Klemas V. V, Moisan T., Babin M., … Jo Y., (2014).

Future Retrievals of Water Column Bio-Optical Properties using the Hyperspectral

Infrared Imager (HyspIRI). Remote Sensing, 5(12), 6812-683.

http://doi.org/10.3390/rs5126812

Dodson V. J., Dahmen J. L., Mouget J. L., & Leblond J. D., (2013). Mono- and

digalactosyldiacylglycerol composition of the marennine-producing diatom, Haslea

ostrearia: Comparison to a selection of pennate and centric diatoms. Phycological

Research, 61, 3(I), 199-207.http://doi.org/10.1111/pre.12015

Page 268: Par « Rémi NGHIEM-XUAN

242

Doucha Jiří, and Karel Lívanský, (1995). Novel Outdoor Thin-Layer High Density Microalgal

Culture System: Productivity and Operational Parameters. Algological Studies/Archiv Für

Hydrobiologie, Supplement Volumes, January, 129–47.

EFSA European food safety authority - UE 1129/2011. EU legislation on food additives –

Commission européenne. Site : https://www.efsa.europa.eu/fr/topics/topic/food-colours

Espinosa E. P., Barillé L., & Allam B., (2006). Use of encapsulated live microalgae to

investigate pre-ingestive selection in the oyster Crassostrea gigas. Journal of

Experimental Marine Biology and Ecology, 343, 1(I), 118-126.

http://doi.org/10.1016/j.jembe.2006.12.002

Esais W. E., and H. C. Curl, Jr (1972). Effect of dinoflagellate bioluminescence on copepod

ingestion rates.Limnology and Oceanography. 17:901-906.

Ewoldt R. H., Johnston M. T., & Caretta L. M., (2015). Experimental challenges of shear

rheology: how to avoid bad data. Biological and Medical Physics, Biomedical Engineering

book series (BIOMEDICAL), 1–36.

Fadlallah H., Jarrahi M., Herbert E., Ferrari R., Mejean A., Peerhossaini H., (1995). Effects of

Shear Stress on the Growth Rate of Micro-organisms in Agitated Reactors. Journal of

Biotechnology, 20,167(4):370-6. http://dx.doi.org/10.1115/FEDSM2016-7590.

Franco Veronica, (2012). Communiqué de presse : marketwired. Site :

http://www.marketwired.com/press-release/market-research-forecasts-antiviral-market-

at-30-billion-by-2017-1653268.htm

Gabelle J.C., Jourdier E., Licht R., Ben Chaabane F., Henaut I., Morchain J., Augier F., (2012).

Impact of rheology on the mass transfer coefficient during the growth phase of

Trichoderma reesei in stirred bioreactors.Chemistry Engineering Science, 75, 408-417.

DOI:1 0.1016/j.ces.2012.03.053

Gagneux-Moreaux S., Moreau C., Gonzalez J.-L., & Cosson R. P., (2007). Diatom artificial

medium (DAM): a new artificial medium for the diatom Haslea ostrearia and other marine

microalgae. Journal of Applied Phycology, 19(5), 549–556.

http://doi.org/10.1007/s10811-007-9169-4

Gao Q., Yu K., Xia Y., Shine M. B., Wang C., Navarre D., & Kachroo A., (2014). Mono- and

Digalactosyldiacylglycerol Lipids Function Nonredundantly to Regulate Systemic

Acquired Resistance in Plants Report Mono- and Digalactosyldiacylglycerol Lipids

Function Nonredundantly to Regulate Systemic Acquired Resistance in Plants. Cell

Reports, 9(5), 1681–1691. http://doi.org/10.1016/j.celrep.2014.10.069

García Camacho F, Gallardo Rodríguez JJ, Sánchez Mirón A, Cerón García MC, Belarbi EH,

Molina Grima E., (2007). Determination of shear stress thresholds in toxic dinoflagellates

cultured in shaken flasks. Implications in bioprocess engineering. Process Biochemistry,

42:1506–1515

Page 269: Par « Rémi NGHIEM-XUAN

243

Gastineau R., Turcotte F. O., Pouvreau J.-B., Morançais M., Fleurence J., Windaro E., Prasetiya

F.S., Arsad S., Jaouen P., Badin M., Coiffard L., Couteau C., Bardeau J.F., Jacquette B,

Leignel V., Hardivillier Y., Marcotte I., Bourgounon N., Tremblay R., Deschênes J.S.,

Dadawy H., Pasetto P., Davidovich N., Hansen G., Dittmer J. and Mouget J.-L., (2014).

Marennine, Promising Blue Pigments from a Widespread Haslea Diatom Species

Complex. Marine Drugs, 12, 3161–3189. http://doi.org/10.3390/md12063161

Gastineau R., Pouvreau J.-B., Hellio C., Morançais M., Gaudin P., Bourgougnon N., & Mouget

J.-L., (2012a). Biological Activities of Purified Marennine, the Blue Pigment Responsible

for the Greening of Oysters.Journal of Agriculture and Food Chemistry, 60(14):3599-

605http://doi.org/10.1021/jf205004x

Gastineau R., Hardivillier Y., Leignel V., Tekaya N., Morançais M., Fleurence J., Mouget J.-

L., (2012b). Greening effect on oysters and biological activities of the blue pigments

produced by diatom Haslea Karadagensis (Naviculaseae). Aquaculture, 368–369, 61-67

http://doi.org/10.1016/j.aquaculture.2012.09.016

Gastineau R., (2011). Biodiversité, reproduction et phylogénie des diatomées bleues du genre

Haslea et valorisation de leurs pigments de type marennine. (Thèse de doctorat; Université

de Nantes).

Gest H., (2002). History of the word photosynthesis and evolution of its definition,

Photosynthesis Research, vol. 73,1-3, 7-10.

Gouveia Luisa, (2011). Microalgae as a Feedstock for Biofuels. Springer.

Grasshoff K., Kremling K., Revised C., Anderson L. G., Andreae M. O., Behrends B., Yhlen

B., (2007). Methods of Seawater Analysis (WILEY‐VCH Verlag GmbH).

Griffin D., Hopkins J., Pickles J. O., Corey D. P., & Hughes H. (1992). Mechanoeletrical

transduction by hair cells. Annual Review of Biophysics and Biophysical Chemistry, 17:99-

124.

Griffiths M. J., & Harrison S. T. L., (2009). Lipid productivity as a key characteristic for

choosing algal species for biodiesel production. Journal of Applied Phycology, 493–507.

http://doi.org/10.1007/s10811-008-9392-7

Grobbelaar Johan U., (2003). Algal Nutrition ‒ Mineral Nutrition. In Handbook of Microalgal

Culture, edited by Amos Richmond, 95–115. Blackwell Publishing Ltd.

Guiry M. D., (2012). How many species of algae are there? Journal of Phycology, 48(5):1057-63. doi:

10.1111/j.1529-8817.2012.01222.x

Hochfeld W. L., (2005). Producing Biomolecular Substances with Fermenters, Bioreactors, and

Biomolecular Synthesizers, Boca Raton, FL, Taylor & Francis, ISBN 9781420021318.

Holmgren A., (1989). Electron Transport to Reductive Enzymes. The Journal of Biological

Chemistry, 264(24), 13963–13966.

Page 270: Par « Rémi NGHIEM-XUAN

244

Ivar Andre Ødegård, (2012). Composition and Surface Modifications of Silica Structures in

Diatom Frustules by Incorporation of Functional Oxides and Nitride Formation. (Thèse de

doctorat; Norwegian University of Science and Technology).

Jaouen P., (1988). Etude des techniques de séparation par membrane dans le domaine des

pêches et des cultures marines : récupération de protéines solubles de poisson par

ultrafiltration, concentration de microalgues marines par microfiltration tangentielle.

(Thèse de doctorat ; Université de Nantes).

Jaouen P., Vandanjon L., & Quéméneur F., (1998). The shear stress of microalgal cell

suspensions (Tetraselmis suecica) in tangential flow filtration systems: the role of pumps.

Bioresource Technology, 68, 149–154.

Jiménez Carlos, Belén R. Cossio, and F. Xavier Niell., (2003). Relationship between

Physicochemical Variables and Productivity in Open Ponds for the Production of

Spirulina: A Predictive Model of Algal Yield. Aquaculture 221 (1–4): 331–45.

Jubeau S., (2009). Optimisation de la synthèse d’un exométabolite par la diatomée marine

Haslea ostrearia (Stage de master 2 ; Université de Nantes).

Juhl A.R., Velazquez V., Latz M.I., (2000). Effect of growth conditions on flow-induced

inhibition of population growth of a red-tide dinoflagellate. Limnology and

Oceanography, 45, 905–915. http://dx.doi.org/10.4319/lo.2000.45.4.0905.

Khalid A., Legrand J., (2001). Energy dissipation distribution and mixing in a torus reactor.

Chemical Engineering Communications, 185, 141–164.

Kandilian Razmig, Jérémy Pruvost, Jack Legrand, and Laurent Pilon., (2014). Influence of

Light Absorption Rate by Nannochloropsis Oculata on Triglyceride Production during

Nitrogen Starvation. Bioresource Technology,163 (July): 308–19.

Kaplan Richmond, Dubinsky and Aaronson., (1986). Algal Nutrition. In Handbook of

Microalgal Mass Culture, Boca Raton, Florida, 147–98. CRC Press, Inc.

Kandilian, Pruvost, Artu, Lemasson, Legrand, and Pilon., (2016). Comparison of

Experimentally and Theoretically Determined Radiation Characteristics of Photosynthetic

Microorganisms. Journal of Quantitative Spectroscopy and Radiative Transfer, 175, 30-

45. https://doi.org/10.1016/j.jqsrt.2016.01.031

Karp P., Paley s., Krummenacker M., Latendresse M., Dale J., Lee T., Kaipa P., Gilham F.,

Spauling A., Popescu L., Altman T., Paulsen I., Keseler I., and Caspi R., (2010). Pathway

tools version 13.0: integrated software for pathway/genome informatics and systems

biology. Briefings in Bioinformatics, 11(1) :40-79.

Kazbar A., (2018). Etude de l’impact de la concentration en oxygène dissous et la présence

d’une fraction sombre sur la performance des photobioréacteurs. (Thèse de doctorat;

Université de Nantes).

Kuczynska P., Jemiola-rzeminska M., & Strzalka K., (2015). Photosynthetic Pigments in

Diatoms. Marine Drugs, 13, 5847–5881. http://doi.org/10.3390/md13095847

Page 271: Par « Rémi NGHIEM-XUAN

245

Kudo I., Miyamoto M., Noiri Y., Maita Y., (2000). Combined effects of temperature and iron

on te groth and physiology of the marine diatom Phaeodactlum tricornutum

(Bacillariophyceae), Journal of Phycology, 36: 1096-1102

Latour J. C.- T., Bernier S.- B., Doiron K., Tremblay R., Lemarchand K., (2018). Controlling

mass mortality events with probiotics during the blue mussels (Mytilus edulis) larvae

rearing process: What role is played by the larval microbiota. Aquaculture Canada

conference and tradeshow 2018.

Lebeau T., Moan R., Turpin V., & Robert J.-M., (1998). Alginate-entrapped Haslea ostrearia

as inoculum for the greening of oysters. Biotechnology Techniques, 12(11), 847–850.

Lebeau T., Gaudin P., Junter G.-A., Mignot L., & Robert J.-M. (2000). Continuous marennin

production by agar-entrapped Haslea ostrearia using a tubular photobioreactor with

internal illumination. Applied Microbiology Biotechnology, 54, 634–640.

Le Borgne F., (2011). Développement d’un photobioréacteur solaire intensifié en vue de la

production à grande échelle de biomasse microalgale (Thèse de doctorat ; Université de

Nantes).

Le Gouic B., (2013). Analyse et optimisation de l’apport de carbone en photobioréacteur (Thèse

de doctorat ; Université de Nantes).

Leegood R. C., (2007). A welcome diversion from photorespiration. Nature Biotechnology, vol.

25, p. 539-540

Lemasson C., (2011). Etude, modélisation et optimisation du comportement thermique des

photobioréacteurs solaires. Rapport de stage M2. Université de Nantes.

Lemasson C., (2015). Contribution à l’étude de la production d’hydrogène par Chlamydomonas

reinhardtii : vers la mise ne place d’une production continue en photobioréacteur bi-étagé

(Thèse de doctorat ; Université de Nantes).

Leupold M., Hindersin S., Gust G., Kerner M., Hanelt D., (2013). Influence of mixing and shear

stress on Chlorella vulgaris, Scenedesmus obliquus, and Chlamydomonas reinhardtii.

Journal of Applied Phycology. 25, 485-495. http://doi.org/10.1007/s10811-012-9882-5

Liang Y., Sarkany N., Cui Y., (2009). Biomass and lipid productivities of Chlorella vulgaris

under autotrophic, heterotrophic and mixotrophic growth conditions. Biotechnology

Letters, 31(7):1043-9. doi: 10.1007/s10529-009-9975-7.

Lorenzen C. J., (1967). Determination of chlorophyll and phaeopigments: Spectrophotometric

equations. Limnology and Oceanography, 12, 343.

Malapascua José RF, Celia G Jerez, Magda Sergejevová, Félix L Figueroa, and JiˇríMasojídek,

(2014). Photosynthesis Monitoring to Optimize Growth of MicroalgalMass Cultures:

Application of Chlorophyll Fluorescence Techniques. Aquatic Biolology, 22:123–40.

Mann D. G., Droop S. J. M., (1996). Biodiversity, biogeography and conservation of diatoms.

Hydrobiologia. 336, 1-3I: 19-32.

Page 272: Par « Rémi NGHIEM-XUAN

246

Masojidek J., Torzillo G. and Koblizek M., (2013). Photosynthesis in Microlagae, In:

Handbook of Microalgal Culture. Applied hycology and Biotechnology., Richmond A. Hu

Q., Eds., 2nd ed. John Wiley & Sons: USA, pp. 21-36.

Mauer J., Mendez S., Fedosov D. A., Fromental J., Claveria V., Nicoud F., Claveria V., (2016).

Red cells dynamic morphologies govern blood shear thinning under microcirculatory flow

conditions. Proceeding of the National Academy of Science of the Unite States of

America,113(50). http://doi.org/10.1073/pnas.1618852114

Mezger Thomas G., (2011). The Rheology Handbook 3rd revised edition. (Vincentz Network

GmbH & Co KG, Ed.). European Coatings Tech Files.

Michels M.H.A., van der Goot A.J., Vermuë M.H. & Wijffels R.H., (2015). Cultivation of shear

stress sensitive and tolerant microalgalspecies in a tubular photobioreactor equipped with

a centrifugal pump. Journal of Applied Phycology, 28(1), 53-62. doi: 10.1007/s10811-015-

0559-8

Michels M. H. A., Atze J., Van Der Goot J., Norsker N.-H., René, & Wijffels H., (2010). Effects

of shear stress on the microalgae Chaetoceros muelleri. Bioprocess and Biosystem

Engineering, 33(8): 921–927. http://doi.org/10.1007/s00449-010-0415-9

Mitsuhashi S., Hosaka K., Tomonaga E., Muramatsu H., Tanishita K., (1995). Effects of shear

flow on photosynthesis in a dilute suspension of microalgae. Applied Microbiology and

Biotechnology, 42:744–749

Moreira S. M., Moreira-Santos M., Guilhermino L., & Ribeiro R., (2006). Immobilization of

the marine microalga Phaeodactylum tricornutum in alginate for in situ experiments: Bead

stability and suitability. Enzyme and Microbial Technology, 38, Issues 1–2, 135-141.

http://doi.org/10.1016/j.enzmictec.2005.05.005

Moreno-Garrido, (2013). Microalgal Immobilization Methods. Immobilization of Enzymes and

Cells. 327-34. Part of the Methods in Molecular Biology book series (MIMB, volume

1051).

Mouget J.-L., Rosa P., & Erard Tremblin G., (2004). Acclimation of Haslea ostrearia to light

of different spectral qualities – confirmation of “chromatic adaptation” in diatoms. Journal

of Photochemistry and Photobiology, 19 ;75 (1-2): 1-11.

http://doi.org/10.1016/j.jphotobiol.2004.04.002

Mouget J.-L., Rosa P., Vachoux C., & Tremblin G., (2005). Enhancement of marennine

production by blue light in the diatom Haslea ostrearia. Journal of Applied Phycology,

17, (5), 437–44517, 437–445. http://doi.org/10.1007/s10811-005-0561-7

Mouget J.-L., Gastineau R., Davidovich O., Gaudin P., & Davidovich N. A., (2009). Light is a

key factor in triggering sexual reproduction in the pennate diatom Haslea ostrearia. FEMS

Microbiology Ecology, 69(2):194-201. http://doi.org/10.1111/j.1574-6941.2009.00700.x

Moutel B., (2016). Étude de l'intérêt pharmaceutique et d'une production industrielle des lipides

issus de la microalgue Botryococcus braunii (Thèse de doctorat ; Université de Nantes).

Page 273: Par « Rémi NGHIEM-XUAN

247

Moreau Je., (1968). Les facteurs de verdissement de l’huître en claires : Le milieu

hydrobiologique et benthique et ses variations. Revue des travaux de l'Institut des pêches

maritimes, 32 (i) : 369-386.

Moreau J., (1970). Contribution aux recherches écologiques sur les claires à huîtres du bassin

de Marennes-Oléron. Revue des travaux de l'Institut des pêches maritimes, 34: 38-462.

Nassiri Y., J-M. RobertY., Rincé and T. Ginsburger-Vogel, (1998). The cytoplasmic fine

structure of the diatom Haslea ostrearia (Bacillariophuceae) in relation to marennine

production. Journal of the International Phycological Society, vol.37 (2), 84-91.

Ndiaye M., Gadoin E., Gentric C., (2018). CO2 gas-liquid mass transfer and kLa estimation:

Numerical investigation in the context of airlift photobioreactor scale-up. Chemical

Engineering Reasearch and Design, 133, 90-102.

Neuville D., Daste P., (1972). Production du pigment bleu par la diatomée Navicula ostrearia

(Gaillon) Bory, maintenue en culture uni-algue sur milieu synthétique carencé en azote.

Compte Rendude l’Académie des Sciences (Paris), 374: 2030-2033

Neuville D., Daste P. H., (1978). Recherches sur le déterminisme de la production de marennine

par la diatomée marine Navicula ostrearia (Gaillon) Bory en culture in vitro. Revue

générale de botanique. 85:255–303

Nic M., Jirat J., Kosata B., Jenkins A. D., (2014). International Union of Pure and Applied

Chemistry Compendium of Chemical Terminology. (Gold Book, Ed.2.3.3).

Olmos E., Caroline Gentric C., Poncin S., Midoux N., (2003). Description of flow regime

transitions in bubble columns via laser Doppler anemometry signals processing. Chemical

Engineering Science, 58(9). DOI: 10.1016/S0009-2509(03)00002-2

Olsen L. M., Öztürk M., Sakshaug E., Johnsen G., (2006). Photosynthesis-induced phosphate

precipitation in seawater: ecological implications for phytoplankton. Marine Ecology

Progress Series. 319: 103–110

Peter D., & Benning C., (2002). Galactolipids rule in seed plants. Trends in Plant Science, 7(3),

112–118.

Pilon Laurent, Halil Berberoğlu, and Razmig Kandilian, (2011). Radiation Transfer in

Photobiological Carbon Dioxide Fixation and Fuel Production by Microalgae. Journal of

Quantitative Spectroscopy and Radiative Transfer 112 (17): 2639–60.

Pottie L., Pruvost J., Deremetz J., Cornet J.F., Legrand J., Dussap C.G., (2005). A fully

predictive model for one-dimensional light attenuation by Chlamydomonas reinhardti in a

torus photobioreactor. Biotechnology and Bioengineering. 12 (5):

https://doi.org/10.1002/bit.20475

Pouvreau J.-B., Morançais E., Fleury F., Rosa P., Thion L., Cahingt B., … Fleury F., (2006).

Preliminary characterisation of the blue-green pigment “marennine” from the marine

tychopelagic diatom Haslea ostrearia (Gaillon/Bory) Simonsen. Journal of Applied

Phycology, 18, 757–767. http://doi.org/10.1007/s10811-006-9087-x

Page 274: Par « Rémi NGHIEM-XUAN

248

Pouvreau J.-B., Morançais M., Housson E., Le Tallec L., (2007). Growth inhibition of several

marine diatom species induced by the shading effect and allelopathic activity of

marennine, a blue-green polyphenolic pigment of the diatom Haslea ostrearia

(Gaillon/Bory). Journal of Experimental Marine Biology and Ecology, 352 (1), 212–225.

DOI:

Prasetiya F. S., Safitri I., Cognie B., Gastineau R., & Morançais M., (2016). Does allelopathy

affect co-culturing Haslea ostrearia with other microalgae relevant to aquaculture?

Journal of Applied Phycology, 28(4):2241-2254. http://doi.org/10.1007/s10811-015-

0779-y

Prokop Aleš, Rakesh K. Bajpai, and Mark E. Zappi., (2015). Algal Biorefineries: Volume 2:

Products and Refinery Design. Springer.

Pruvost J., Pottier L., & Legrand J., (2006). Numerical investigation of hydrodynamic and

mixing conditions in a torus photobioreactor. Chemical Engineering Science, 61, 4476–

4489. http://doi.org/10.1016/j.ces.2006.02.027

Pruvost J., J. -F. Cornet and J. Legrand, (2008). Hydrodynamics Influence on Light Conversion

in Photobioreactors: An Energetically Consistent Analysis. Chemical Engineering Science

63 (14): 3679–94.

Pruvost J, and JF Cornet, (2012). Knowledge Models for the Engineering and Optimization of

Photobioreactors. Microalgal Biotechnology: Potential and Production. De Gruyter,

Berlin, Germany, 181–224.

Pruvost J., Le Borgne F., Artu A., Legrand J., (2017). Development of a thinf-film solar

photobioreactor with high biomass volumetric productivity (Algofilm ©) based on process

intensification principles. Algal Research. Vol. 21, p. 120-137.

Pruvost J., Legrand J., Legentilhomme P., & Muller-feuga A., (2004). Effect of Inlet Type on

Shear Stress and Mixing in an Annular Photobioreactor Involving a Swirling Decaying

Flow. The Canadian Journal of Chemical Engineering, 82(June), 495–503.

Pulz Otto, and Karl Scheibenbogen, (1998). Photobioreactors: Design and Performance with

Respect to Light Energy Input. Bioprocess and Algae Reactor Technology, Apoptosis,

123–52. Advances in Biochemical Engineering Biotechnology 59. Springer Berlin

Heidelberg.

Ras M., Jean-Philippe Steyer, Olivier Bernard, (2013). Temperature effect on microalgae: a

crucial factor for outdoor production. Reviews in Environmental Science and

Biology/Technology, Springer, 12 (2), pp.153-164.

Ranson G., (1927) L’absorption de matières organiques dissoutes par la surface expérieure du

corps chez les animaux aquatiques. Annales de l’Institut Océanographique IV:49–174.

Raven John A., and Richard J. Geider., (1988). Temperature and Algal Growth. The New

Phytologist 110 (4): 441–61.

Page 275: Par « Rémi NGHIEM-XUAN

249

Richmond, (1986). Outdoor mass culture of microalgae. In CRC Handbook of microalgal mass

culture, Boca Raton, 285–329. CRC press, Inc. Florida.

Richmond Amos, and Qiang Hu, (2008). Handbook of Microalgal Culture: Applied Phycology

and Biotechnology. 2nd Edition. Chichester: Wiley-Blackwell.

Rincé Y., Lebeau T., & Robert J. M., (1999). Artificial cell-immobilization: a model simulating

immobilization in natural environments? Journal of Applied Phycology, 11, 263–272.

Ritchie R. J., (2006). Consistent sets of spectrometric chlorophyll equations for acetone,

methanol and ethanol solvents. Photosynthesis Research, 89 : 27-41.

Ritchie R. J., (2008). Universal chlorophyll equations for estimating chlorophylls a, b, c, and

dand total chlorophylls in natural assemblages of photosynthetic organisms using acetone,

methanol, or ethanol solvents. Photosynthetica, 46(1), 115–126.

Robert J.-M., (1983). Fertilité des claires ostréicoles et verdissement : utilisation de l’azote par

les diatomées dominantes. (Thèse de Doctorat. Univ. Nantes).

Robert J.-M., (1989). Procédé de culture en masse de la diatomée Haslea ostrearia Simonsen.

European and French patent no 8915314. Nantes University

Robert J., Morançais M., Pradier E., Mouget J., & Tremblin G., (2002). Extraction and

quantitative analysis of the blue-green pigment “marennine” synthesized by the diatom

Haslea ostrearia. Journal of Applied Phycology, 14, (I)4, 299–305.

Rossignol N. (1999). Procédé d’éxtraction et de séparation par membranes appliqués à la

production du pigment endo-et exocellulaire synthétisé par la diatomée Haslea ostrearia

Simonsen. Mise en œuvre d’un photobioréacteur à membranes à fonctionnement continu.

(Thèse de doctorat ; Université de Nantes).

Rossignol N., Moan R., Jaouen P., & Robert J., (1999). Continuous high-pressure disruption of

marine diatom Haslea ostrearia. Assessment by laser diffraction particle sizer,

Biotechnology techniques (Foster 1995), 909–913.

Rouillard I., (1996). Optimisation de la production en masse d'Haslea ostrearia Simonsen sur

une eau souterraine salée : importance de la souche et des conditions de culture ;

comparaison avec Skeletonema costatum (Grev) Cleve. (Thèse de doctorat ; Université de

Nantes).

Ruzicka M. C., Drahoš J., Fialova M., Thomas N. H., (2001). Effect of bubble column

dimensions on flow regime transition. Chemical Engineering Science. Vol. 56 (21), p.

6117–6124.

Sachse M., Gruber A., Sturm S., Schellenberger-Costa B., Rottberger J., and Kroth P. G.,

(2013). Regulation of the Calvin cycle in the diatom Phaeodactylum tricornutum. Plant

Physiology, (Thèse de doctorat ; Universität Konstanz)

Scriban R., (2009). Biotechnologie5ème édition. (LAVOISIER / TEC ET DOC, Ed.).

Page 276: Par « Rémi NGHIEM-XUAN

250

Schubert H., Tremblin G., Robert J-M., Sagert S. et Rincé Y., (1995) In vivo fluorescence

measurement of photosynthesis of Haslea ostrearia Simonsen in relation to marennine

content. Diatom Research. 10:341–349

Schuster Arthur., (1905). Radiation through a Foggy Atmosphere. The Astrophysical Journal

21: 1.

Smithsonian marine station at Fort Pierce, Indian River Lagoon Species Inventory. Site :

http://www.sms.si.edu/irlspec/Haslea_ostrea.htm

Sobczuk T.M., Camacho F.G., Grima E.M., Chisti Y., (2006). Effects of agitation on the

microalgae Phaeodactylum tricornutum and Porphyridium cruentum. Bioprocess and

Biosystem Engineering. 28, 243–250. http://dx.doi.org/10.1007/s00449-005-0030-3.

Souliès A., (2014). Contribution à l’étude hydrodynamique et à la modélisation des

photobioréacteurs à haute productivité volumique (Thèse de doctorat ; Université de

Nantes).

Speziale B. J., Schreiner S. P., Giammatteo P. A., Schindler J. E., (1984). Comparaison of N,N

dimethyformamide, dimethyl sulfoxide and acetone for extraction of phytoplankton

chlorophyll. Canadian Journal of Fisheries and Aquatic Sciences, 41,519-1522

Spolaore P., Joannis-cassan C., Duran E., Isambert A., Génie L. De & Paris E. C., (2006).

Commercial Applications of Microalgae. Journal of Bioscience and Bioengineering,

101(2), 87–96.http://doi.org/10.1263/jbb.101.87

Sterrenburg FAS, Tiffany MA, Herwig WE, Hargraves PE., (2015). Seven new species expand

the morphological spectrum of Haslea. A comparison with Gyrosigma and Pleurosigma

(Bacillariophyta). Phytotaxa 207: 143-162.

Strickland J. D., and T. Parsons, (1968). A Pratical Handbook of Seawater Analysis.Fisheries

Reasearch Bulletin. Bd. Canada.

Taleb A., Pruvost J., Legrand J., Marec H., Le-gouic B., Mirabella B., … Bouvet S., (2015).

Bioresource Technology Development and validation of a screening procedure of

microalgae for biodiesel production: Application to the genus of marine microalgae

Nannochloropsis. Bioresource Technology, 177, 224–232.

http://doi.org/10.1016/j.biortech.2014.11.068

Tavernarakis N., & Driscoll M., (1997). Molecular modeling of mechanotransduction in the

nematode Caenorhabditis elegans. Annual Review of Physiology, 59.

Tesson B., & Latz M. I., (2015). Article Mechanosensitivity of a Rapid Bioluminescence

Reporter System Assessed by Atomic Force Microscopy. Biophysical Journal, 108(6),

1341–1351. http://doi.org/10.1016/j.bpj.2015.02.009

Tomas C. R., (1996). Identifying Marine Diatoms and Dinoflagellates. Academic Press, Inc.

Page 277: Par « Rémi NGHIEM-XUAN

251

Tredici M. R., and R. Materassi, (1992). From Open Ponds to Vertical Alveolar Panels: The

Italian Experience in the Development of Reactors for the Mass Cultivation of

Phototrophic Microorganisms. Journal of Applied Phycology 4 (3): 221–31.

Tredici Mario R., (2003). Mass Production of Microalgae: Photobioreactors. In Handbook of

Microalgal Culture, edited by Amos Richmond, 178–214. Blackwell Publishing Ltd.

Tremblin G., Cannuel R., Mouget J-L., Rech M., Robert J-M., (2000). Change in light quality

due to a blue-green pigment, marennine, released in oyster-ponds: effect on growth and

photosynthesis in two diatoms, Haslea ostrearia and Skeletonema costatum. Journal of

Applied Phycology. 12:557–566.

Turcotte F., Mouget J. L., Genard B., Lemarchand K., Deschênes J. S. and Tremblay R., (2016).

Prophylactic effect of Haslea ostrearia culture supernatant containing the pigment

marennine to stabilize bivalve hatchery production. Aquatic Living Resources,29,

401.https://doi.org/10.1051/alr/2016032

Turpin V., Robert J.-M. & Goulletquer P., (1999) Limiting nutrients of oyster pond seawaters

in the Marennes-Oléron region for Haslea ostrearia: applications to the mass production

of the diatom in mesocosm experiments. Aquatic Living Resources 12, 335-342.

Turpin V., J-M Robert, P Goulletquer, G. M. & P. R. (2001). Oyster greening by outdoor mass

culture of the diatom Haslea ostrearia Simonsen in enriched seawater. Aquaculture

Research, 32(801-809).

Ugwu C., Aoyagi H., Uchiyama H., (2008). Influence of irradiance, dissolved oxygen

concentration, and temperature on the growth of Chlorella sorokiniana. Photosynthetica,

45, (I)2, 309–311. 45:309–311. doi: 10.1007/s11099-007-0052-y.

Vandanjon L., Jaouen P., Rossignol N., & Que F., (1999). Concentration and desalting by

membrane processes of a natural pigment produced by the marine diatom Haslea ostrearia

Simonsen. Journal of Biotechnology, 70, 393–402.

Vandanjon L., Rossignol N., Jaouen P., Robert J.M. et Quemeneur F., (1999) "Effects of shear

on two microalgae species (Skeletonema costatum and Haslea ostrearia). Contribution of

pumps and valves in tangential flow filtration systems", Biotechnology and

Bioengineering. Vol 63, n°1, pp 1-9.

Varshney Prachi, Paulina Mikulic, Avigad Vonshak, John Beardall, and Pramod P. Wangikar,

(2015). Extremophilic Micro-Algae and Their Potential Contribution in Biotechnology.

Bioresource Technology, Advances in biofuels and chemicals from algae, 184 (May): 363–

72.

Vilchez C., I. Garhayo, M. V. Lobato, and J. M. Vega, (1997). Microalgae-mediated chemicals

production and wastes removal. Enzyme and Microbial Technology, 20 : 562-572.

Wang C., & Lan C. Q., (2018). Effects of shear stress on microalgae – A review. Biotechnology

Advances, ;36(4):986-1002. http://doi.org/10.1016/j.biotechadv.2018.03.001

Page 278: Par « Rémi NGHIEM-XUAN

252

Wilde E. W., Beneman J. R., (1993). Bioremoval of heavy metals by use of microalgae.

Advanced Biotechnology, 11: 781-812.

Woo M.N., Jeon S.M., Kim H.J., Lee M.K., Shin S.K., Shin Y.C., Park Y.B., Choi M.S., (2010)

Fucoxanthin supplementation improves plasma and hepatic lipid metabolism and blood

glucose concentration in high-fat fed C57BL/6N mice. Chemistry Biology Interaction,

2010, 186, 316–322.

Xin L., Hong-ying H., Yu-ping Z., (2011). Growth and lipid accumulation properties of a

freshwater microalga Scenedesmus sp. Under different cultivation temperature.

Bioressource Technology, 102:3098-3102

Yasobiotech. Site : http://www.yasobiotech.com/anti-viral-market.html

Page 279: Par « Rémi NGHIEM-XUAN
Page 280: Par « Rémi NGHIEM-XUAN

Titre : Optimisation de la culture d’Haslea ostrearia en photobioréacteur

Mots clés : Milieu de culture, fragilité cellulaire, conditions de culture, photobioréacteur, marennine extracellulaire, Haslea ostrearia, diatomée.

Résumé : Haslea ostrearia est une diatomée capable d’excréter un pigment bleu-vert appelé marennine d’intérêt industriel en aquaculture, cosmétique et pharmaceutique. Sa culture en laboratoire et à grande échelle est actuellement un défi, cette espèce à comportement benthique étant considérée comme sensible à l’agitation des systèmes de cultures conventionnels. Cette thèse décrit l’identification, l’étude et

l’optimisation des paramètres limitants de la culture en photobioréacteur. Un milieu de culture d’eau de mer artificielle a été conçu pour garantir une production stable et non limitée par les nutriments. La fragilité d’Haslea ostrearia au cisaillement hydrodynamique a été caractérisée en appliquant différentes intensités de taux de cisaillement.

Les paramètres optimaux de culture et les conditions d’excrétion de marennine extracellulaire ont été finalement identifiés en photobioréacteur. Au final, les résultats obtenus ont permis de proposer des protocoles et systèmes optimisés de culture, adaptés aux spécificités et besoins d’Haslea ostrearia. Les résultats de cette thèse permettent d’envisager la culture semi-continue de cette microalgue, avec des productivités supérieures à celles rencontrées jusqu’alors dans la littérature. La méthodologie permet également d’apporter les bases pour une optimisation future de la culture d’Haslea ostrearia contrôlée à plus grande échelle, ainsi que sa production de marennine extracellulaire associée.

Title : Optimization of Haslea ostrearia culture in photobioreactor

Keywords : Culture medium, Cell fragility, growth conditions, photobioreactor, extracellular marennine, Haslea ostrearia, diatom.

Abstract : Haslea ostrearia is a marine diatom able to produce and excrete a blue-green pigment called marennine, exhibiting potential industrial valorization in aquaculture, cosmetic and pharmaceutics. Its laboratory and large-scale culture are challenging because this benthic strain is considered as sensitive to mixing conditions in conventional culture system. This thesis describes the identification, the

study and the optimization of limiting parameters of the cell culture in a photobioreactor. A new artificial seawater medium has been designed in order to guarantee a stable and no-nutrients limited production. The Haslea ostrearia fragility to hydrodynamics shear stress has been characterized by applying different intensities of shear rate.

The optimal culture parameters and excretion conditions of extracellular marennine have been finally identified in a photobioreactor. In the end, the results obtained made it possible to propose protocols and optimized culture systems, adapted to Haslea ostrearia specificities and needs. The results of this thesis make it possible to consider the semi-continuous culture of this microalgae, with higher productivities than those described so far in the literature. The methodology also provides the basis for future optimization of the Haslea ostrearia controlled culture on a larger scale, as well as its asociated extracelular marennine production.