Gastroentérologie Clinique et Biologique (2010) 34, 267—269

MINI REVIEW

Human cirrhosis: Monoclonal regenerative nodulesderived from hepatic progenitor cells abuttingductular reactionCirrhose chez l’homme : nodules de régénération monoclonaux dérivés decellules souches hépatiques en connexion avec la réaction ductulaire

P. Bioulac-Sagea,b, C. Balabaudb,∗,c

a Service d’anatomie pathologique, hôpital Pellegrin, CHU de Bordeaux, 33076 Bordeaux, Franceb Inserm U889, université Bordeaux 2, 33076 Bordeaux, Francec Service d’hépatologie, hôpital Saint-André, CHU de Bordeaux, 33075 Bordeaux, France

Available online 28 April 2010

Summary Cirrhosis is a premalignant condition leading to hepatocellular carcinoma. Cirrhoticnodules are surrounded by a rim of CK 7/CK19-positive biliary cells termed ductular reaction.Half of all regenerative cirrhotic nodules are thought to be monoclonal by studying the pat-tern of inactivation of the X-linked human androgen receptor gene (HUMARA). Using a newtechnique for lineage tracing in human liver based on the identification in the mitochondrialDNA of mutations in the cytochrome c oxidase (CCO) gene, the authors discovered that 20% of

regenerative nodules were monoclonal; in addition they showed that hepatic progenitor cellswithin abutting CCO-deficient cells of the ductular reaction had the same mutations as theadjacent regenerative nodule, indicating a common cell origin. It is the first direct evidencethat regenerative nodules in cirrhosis can be derived from hepatic progenitor cells.© 2010 Elsevier Masson SAS. All rights reserved.

b

Cirrhosis is defined histologically as a diffuse process in which the normal anatomical lobules are replaced by archi-tecturally abnormal nodules separated by fibrous tissue.Knowing that cirrhosis is a premalignant condition leadingto hepatocellular carcinoma, a number of attempts have

∗ Corresponding author.E-mail address: charles.balabaud@chu-bordeaux.fr

(C. Balabaud).

(eRi(bacc

0399-8320/$ – see front matter © 2010 Elsevier Masson SAS. All rights redoi:10.1016/j.gcb.2010.03.004

een made to establish the clonality of regenerative nodulesRNs), often by examination of X chromosome-linked mark-rs in females. In liver cirrhosis, approximately half of allNs have been found monoclonal by studying the pattern of

nactivation of the X-linked human androgen receptor geneHUMARA) [1]. However, in this type of approach the distri-

ution of X-inactivated cells in the liver is not taken intoccount (the progeny of a single X-inactivated embryonicell may be clustered together giving a false informationoncerning clonality).

served.

2

7Etodcaaphih

ef

••

•

tacdisidceaasfsto

aiHcobobtti

[

N

t

N

eFtdh

hhtigtichoaamosptpigtiud

rCn(gtwatf

aptim

cc[R

h

68

Cirrhotic nodules are surrounded by a rim of CK/CK19-positive biliary cells termed ductular reaction (DR).xamination of cirrhotic explants of diverse diseases revealshat most intraseptal hepatocytes (ISH), whether as nodulesr as loose clusters, are associated with DR [2]. Three-imensional reconstruction of ISH in hepatitis C relatedirrhosis demonstrates that in that setting virtually all ISHre associated with DRs, which form a structural and prob-bly physiological link to the biliary tree. Therefore, ifroliferation of hepatocytes in the early stages of chronicepatitis C accounts for much of hepatocyte regeneration,n cirrhosis, the biliary tree becomes more proliferative asepatocyte replication diminishes.

A reaction of ductular phenotype, possibly but not nec-ssarily of ductular origin, in chronic liver disease may ariserom:

proliferation of preexisting cholangiocytes;progenitor cells (local so-called hepatic progenitor cells[HPCs] and/or circulating cells probably bone marrow-derived);rarely, biliary metaplasia of hepatocytes.

The progenitor functioning of the DR attracts much atten-ion [3]. In particular, cells of intermediate morphologynd intermediate immunophenotyping are of interest. Theseells are referred to as intermediate hepatobiliary cells,efined as larger than 6 microns in diameter (the approx-mate size of the normal canal of Hering cell, i.e., themallest cholangiocytes), but less than 40 microns (the typ-cal size of a hepatocyte), with other features suggestingual characteristics of both hepatocytes and cholangio-ytes. These include, but are not limited to: simultaneousxpression of biliary antigens (e.g. keratins 19, 7, OV-6)nd hepatocyte antigens (e.g., HepPar1, albumin, alpha-1-ntitrypsin, biliary glycoprotein-1 detected by canaliculartaining with polyclonal anti-CEA, and, occasionally, alpha-etoprotein), other markers such as NCAM-1/CD56, andtructural features such as basement membrane formationypical of cholangiocytes and canalicular membranes typicalf hepatocytes.

Staining with NCAM, CK19, and HepPar1 has revealeddistinctly bipolar structure to DRs that are embedded

n cirrhotic tissue. Spatial analysis of cells that are singlyepPar1-positive, or CK19-positive, has revealed hepato-ytic and biliary poles, respectively, in the DRs. The locationf singly NCAM-positive cells in DRs suggests that they maye bipotent liver stem/progenitor cells. The locations ofther intermediate hepatobiliary cells, which have com-inations of markers, suggest that CK19+/NCAM+ cells areransitional cells in the biliary lineage and that rare cellshat are negative for all three markers are transitional cellsn the hepatocytic lineage.

The working cell lineage model for DRs is presented below4]:

HepatocyteCAM−/CK19−/HepPar+

← Stem cellNCAM+/CK19−/HEpPar−

→ Bile ductNCAM−/CK19+/HepPar−

Transitional cells in the hepatic lineage are negative forhe three markers.

cstefi

P. Bioulac-Sage, C. Balabaud

Transitional cells for the bile duct lineage are positive forCAM and CK19.

The liver has a multi-tiered, flexible system of regen-ration rather than a single stem/progenitor cell location.our possible hepatic stem cell niches have been identified:he canal of Hering (proximal biliary tree), intralobular bileucts, periductal ‘‘null’’ mononuclear cells, and peribiliaryepatocytes [5].

A new technique for lineage tracing in human liveras been recently discovered [6]. It has been shown thatuman gastrointestinal stem cells and their progeny con-ain non-pathogenic mutations in their mitochondrial DNA,ncluding mutations in the cytochrome c oxidase (CCO)ene, a component of complex IV of the respiratory chain,hat are relatively common. The mitochondrial genomes prone to mutation. Mutations can expand stochasti-ally within a cell and over time cells will become eitheromoplasmic—all the mitochondria in the cell are mutated—r heteroplasmic—the cell contains a mixture of mutatednd wild-type mitochondria. This stochastic expansion is

lengthy process, often taking many years, and for autated cellular phenotype to be observed, homoplasmy

r a high degree of heteroplasmy must be present. Thus,tem cells are the only cells that have a sufficient lifes-an to accumulate these mitochondrial mutations to a levelhat results in a detectable biochemical deficiency. Manyatches of CCO-negative hepatocytes have been identifiedn human liver, invariably connected to the portal areas sug-esting an origin from an area close to the limiting plate,he stem cell niche. mtDNA sequencing of laser-capturedndividual hepatocytes from these patches has providednequivocal proof that each patch was monoclonallyerived.

The same authors using the same approach in human cir-hosis [7] have shown that most RNs are either uniformlyCO-positive or CCO-deficient, less than 3% are CCO-mixedodules, suggesting that most RNs could be monoclonalFig. 1). mtDNA sequence analysis revealed that within aiven CCO-deficient RN, all cells harbored the same muta-ion(s), which proves monoclonality. Furthermore, HPCsithin abutting CCO-deficient DRs had the same mutationss the adjacent RN, indicating a common cell origin. It ishe first direct evidence that RNs in cirrhosis can be derivedrom HPCs.

RNs can therefore be formed by monoclonal expansion,nd not simply by fibrotic dissection of preexisting liverarenchyma and that HPCs within the abutting DRs can havehe same cell of origin as the monoclonal RN. RNs exam-ned histologically revealed no dysplastic features but theiralignant potential is unknown.This study showing that RNs in cirrhosis can be mono-

lonal, with the cell of origin likely to be a facultative stemell from the biliary tree elegantly confirms the first study2]. This obviously changes our concept of the formation ofN.

Cirrhosis formation takes years. After exhaustion of theepatocytic proliferation capability, the potential stem cell

ompartment is activated. This process may perhaps startooner than thought. Indeed, extracellular matrix deposi-ion and activation of matrix-producing cells occurs as anarly phase of chronic liver injury, and it may be that thebrotic environment is important for establishing the niche

Human cirrhosis: Monoclonal regenerative nodules 269

an ci

[

[

[

[

[

Figure 1 Hum

for the process of activation and differentiation of HPCs. Dif-ferentiation of HPCs to hepatocytes occurs at this fibroticinterface, leading to buds of intraseptal hepatocytes, andthe RNs we see may represent further evolution of thisprocess.

This study opens new questions especially in thefield of carcinogenesis. Are those monoclonal nodules theone who will transform into high grade dysplastic nod-ule? We should not wait very long before knowing theanswer.

Conflict of interest statement

The authors have not declared any conflict of interest.

References

[1] Paradis V, Laurendeau I, Vidaud M, Bedossa P. Clonalanalysis of macronodules in cirrhosis. Hepatology 1998;28:953—8.

[

rrhotic nodules.

2] Falkowski O, An HJ, Ianus IA, Chiriboga L, Yee H, West AB, et al.Regeneration of hepatocyte ‘buds’ in cirrhosis from intrabiliarystem cells. J Hepatol 2003;39:357—64.

3] Roskams TA, Theise ND, Balabaud C, Bhagat G, Bhathal PS,Bioulac-Sage P, et al. Nomenclature of the finer branches of thebiliary tree: canals, ductules, and ductular reactions in humanlivers. Hepatology 2004;39:1739—45.

4] Zhou H, Rogler LE, Teperman L, Morgan G, Rogler CE. Identifica-tion of hepatocytic and bile ductular cell lineages and candidatestem cells in bipolar ductular reactions in cirrhotic human liver.Hepatology 2007;45:716—24.

5] Kuwahara R, Kofman AV, Landis CS, Swenson ES, BarendswaardE, Theise ND. The hepatic stem cell niche: identifi-cation by label-retaining cell assay. Hepatology 2008;47:1994—2002.

6] Fellous TG, Islam S, Tadrous PJ, Elia G, Kocher HM, Bhat-tacharya S, et al. Locating the stem cell niche and tracing

hepatocyte lineages in human liver. Hepatology 2009;49:1655—63.

7] Lin WR, Lim SN, McDonald SAC, Graham T, Wright V, Peplow CL,et al. The histogenesis of regenerative nodules in human livercirrhosis. Hepatology 2010;51:1017—26.