Acid-Induced Downregulation of ASS1 ... · Introduction The urea cycle (UC) is the main metabolic...

Metabolism and Chemical Biology

Acid-Induced Downregulation of ASS1Contributes to theMaintenance of Intracellular pHin CancerAlon Silberman1, Omer Goldman1, Odeya Boukobza Assayag1, Adi Jacob2,Shiran Rabinovich1, Lital Adler1, Joo Sang Lee3,4, Rom Keshet1, Alona Sarver1, Julia Frug1,Noa Stettner1,5, Sivan Galai1, Erez Persi6, Keren Bahar Halpern2, Yehudit Zaltsman-Amir1,Ben Pode-Shakked7,8, Raya Eilam5,Yair Anikster7,8, Sandesh C.S. Nagamani9, Igor Ulitsky1,Eytan Ruppin3,4, and Ayelet Erez1

Abstract

Downregulation of the urea cycle enzyme arginino-succinate synthase (ASS1) by either promoter methyl-ation or by HIF1a is associated with increased metas-tasis and poor prognosis in multiple cancers. We havepreviously shown that in normoxic conditions, ASS1downregulation facilitates cancer cell proliferation byincreasing aspartate availability for pyrimidine synthe-sis by the enzyme complex CAD.Here we report that inhypoxia, ASS1 expression in cancerous cells is down-regulated further by HIF1a-mediated induction ofmiR-224-5p, making the cells more invasive anddependent onupstream substrates of ASS1 for survival.ASS1 was downregulated under acidic conditions, andASS1-depleted cancer cells maintained a higher intracel-lular pH (pHi), depended less on extracellular gluta-mine, anddisplayedhigher glutathione levels.Depletionof substrates of urea cycle enzymes in ASS1-deficientcancers decreased cancer cell survival. Thus, ASS1 levelsin cancer are differentially regulated in various environ-mental conditions to metabolically benefit cancer pro-gression. Understanding these alterations may helpuncover specific context-dependent cancer vulnerabil-ities that may be targeted for therapeutic purposes.

Significance: Cancer cells in an acidic or hypoxicenvironment downregulate the expression of the ureacycle enzyme ASS1, which provides them with a redoxand pH advantage, resulting in better survival.

Graphical Abstract: http://cancerres.aacrjournals.org/content/canres/79/3/518/F1.large.jpg.

1Department of Biological Regulation, Weizmann Institute of Science,Rehovot, Israel. 2Department of Molecular Cell Biology, WeizmannInstitute of Science, Rehovot, Israel. 3Center for Bioinformatics andComputational Biology and Dept. of Computer Science, University ofMaryland, College Park, Maryland. 4Cancer Data Science Lab, NationalCancer Institute, NIH, Bethesda, Maryland. 5Department of VeterinaryResources, Weizmann Institute of Science, Rehovot, Israel.6Computational Biology and Bioinformatics Branch (CBB), NationalLibrary of Medicine, National Center for Biotechnology Information(NCBI), NIH, Bethesda, Maryland. 7Metabolic Disease Unit, Edmond andLily Safra Children's Hospital, Sheba Medical Center, Tel-Hashomer,Israel. 8The Sackler Faculty of Medicine, Tel-Aviv University, Tel-Aviv,

Israel. 9Department of Molecular and Human Genetics, Baylor College ofMedicine, Houston, Texas.

Note: Supplementary data for this article are available at Cancer ResearchOnline (http://cancerres.aacrjournals.org/).

O. Goldman and O. Boukobza Assayag contributed equally to this article.

Corresponding Author: Ayelet Erez, Weizmann Institute of Science, Rehovot76100, Israel. Phone: 972-8934-3714; Fax: 972-8934-3739; E-mail:[email protected]

doi: 10.1158/0008-5472.CAN-18-1062

�2018 American Association for Cancer Research.

CancerResearch

Cancer Res; 79(3) February 1, 2019518

on April 22, 2020. © 2019 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst December 20, 2018; DOI: 10.1158/0008-5472.CAN-18-1062

http://crossmark.crossref.org/dialog/?doi=10.1158/0008-5472.CAN-18-1062&domain=pdf&date_stamp=2019-1-22

http://cancerres.aacrjournals.org/content/canres/79/3/518/F1.large.jpg

http://cancerres.aacrjournals.org/

IntroductionThe urea cycle (UC) is the main metabolic pathway in mam-

mals, responsible for detoxifying excess nitrogen, carried in theforms of glutamine and ammonia, by converting it to urea. TheUC enzyme argininosuccinate synthase (ASS1) catalyzes animportant step in this cycle, which conjugates nitrogen from twosources—aspartate, derived from glutamine, and citrulline, con-taining nitrogen from ammonia—to form argininosuccinate(Fig. 1A; refs. 1, 2). The importance of ASS1 for normal nitrogenmetabolism is underscored by manifestations of citrullinemiatype I, an inborn error of metabolism (IEM) caused by biallelicgermlinemutations in ASS1. Individuals with citrullinemia type Ipresent biochemically with elevated plasma levels of ammoniaand citrulline, and clinically with neurologic impairment andeven death, when the enzymatic deficiency is severe. However,counterintuitively to its critical role in normal homeostasis, ASS1is downregulated in multiple diverse cancers, making tumorsauxotrophic for arginine (3, 4). This feature has been exploitedfor therapy in the treatment of individuals with ASS1-deficienttumors, by employing arginine-depleting agents (5).

We recently determined that ASS1 deficiency under normoxicconditions promotes proliferation of cancer cells by increasingthe availability of the ASS1 substrate-aspartate, for pyrimidinesynthesis via the carbamoyl-phosphate synthase 2-aspartatetranscarbamylase-dihydroorotase (CAD) complex (6). Interest-ingly, under hypoxic conditions, when decreased proliferation isseen, hypoxia-inducible factor alpha (HIF1a) has been shown tobind and downregulate the expression levels of both ASS1 andCAD (7–9). Thus, we set out to investigate the metabolic advan-tages that ASS1 downregulation confers on cancer cells, underacidic and hypoxic conditions.

Tumor hypoxia has been shown to promote increased aci-dosis by inducing a shift in cellular metabolism that generates ahigh acid load (10). Recent findings proposed that glutamineplays an important role in increasing cellular compatibilitywith an acidic environment, by releasing ammonia to neutral-ize protons derived from acids (11, 12). In addition, therecycling of glutamine-derived ammonia has been shown tosupport cancer biomass through its incorporation into aminoacids, rather than its disposal as a waste product (13). On thebasis of these findings, and the biochemical abnormalitiesfound in ASS1-deficient patients, we hypothesized that underhypoxia-related acidosis, the metabolic benefits of ASS1 down-regulation, involve the build-up of ASS1 upstreammetabolites—glutamine and glutamine-derived ammonia—thus enablingcancer cell survival (Fig. 1A).

Materials and MethodsMetabolic measurements in human subjects and cells

The forearm ischemic test (FIT) was performed as part ofthe clinical evaluation of patients who presented due tounexplained consistent or recurrent elevations of serum CPKlevels, at the Metabolic Disease Unit of the Sheba MedicalCenter, Israel.

Ammonia measurements from fibroblasts were performed onanonymized cells devoid of all identifiers. Cells were cultured in a10-cmplate to a confluence of 80%.At time 0, the cellswere rinsedwith PBS, fed with 5 mL of the test medium (20% FBS-containedDMEM), and cultured for 24 hours. Ammonia was measured in

the test medium with a colorimetric assay (Abcam). Ammoniaconcentration in the cells' free medium was subtracted.

Metabolic modelingpHprofiles ofmetabolic enzymes of all organismswere fetched

from BRENDA database. These experimental in vitro measure-ments included at most 6 discrete, 3 acidic, and 3 basic pH pointscorresponding to 0%, 50%, and 100% of enzymatic activity.When no experimental value for human was available for oneor more of these points, we used the available data of the closesthomolog. pH activity profiles of enzymes at any given pH werethen linearly interpolated (14).

We applied pH-activity profiles into a panel of genome-scalemetabolic models (GSMM), developed recently based on thehuman GSMM, representing the NCI-60 cell line panel and thenormal healthy lymphoblastic cell lines from the HapMapproject. In a given intracellular pH (pHi), the pH activityprofiles were used to adjust the permissible flux boundsof each reaction in the GSMM, allowing us to estimate themetabolic state of the network at each pHi. Reactions' fluxeswere measured using flux-variability analysis (FVA) understandard constraint of cell viability, of cellular proliferation(Biomass yield) > 80% of its maximum (14).

The Cancer Genome Atlas tumor data analysisWe downloaded The Cancer Genome Atlas (TCGA) gene

expression profiles of 8,224 patients encompassing 27 cancertypes from the Broad Firehose resources (analyzed on January28, 2016, doi:10.7908/C11G0KM9). Gene expression valuesare normalized in each cancer type to account for the distinctbasal expression levels in different cancer types. A gene isdetermined to be downregulated/upregulated in a sample ifits gene expression is below bottom tertile/above top tertileacross samples in each cancer type. For metastasis analysis, weused gene expression profiles in a subset of 648 patients across14 cancer types. A sample is termed metastatic if the tumor isdiagnosed stage IV, and we only considered those cancer typesfor which greater than or equal to 10 such samples are avail-able. To overcome the differences in tumor heterogeneity, weused the tumor samples with high cancer cell fraction (15).

Hypoxia scoreWe used gene expression signature of hypoxia to calculate the

hypoxia score (16). We took the mean expression level of thesignature genes and rank-normalized the values in each cancertype to control for cancer-type specific effect. To make sure ourfindings are not limited to HIF1a target genes, we repeated thesame analysis with the signature genes excluding validatedHIF1atarget genes.

Patient survival analysisWe performed Kaplan–Meier and Cox proportional hazard

analyses to study the association between ASS1 expression togeth-erwithCA9ormonocarboxylate transporter 4 (MCT4) expressionstates and patient survival. We compared the survival of patientswith high CA9 (top tertile, n ¼ 932 patients) and low-CA9(bottom tertile, n ¼ 905 patients) among the ASS1-downregu-lated samples (bottom tertile) using log-rank test (17), and theeffect size was quantified by the difference in the area under thecurves (DAUC). To control for potential confounders, we per-formed a Cox regression analysis, while controlling for patients'

ASS1 Is Downregulated in Acidity to Preserve pHi

www.aacrjournals.org Cancer Res; 79(3) February 1, 2019 519




Figure 1.

ASS1 expression is downregulated during hypoxia and acidic states. A, Illustration of the hypothesized metabolic effects of ASS1 downregulation on cancer pHgradient. During aerobic metabolism, hydrated CO2 is a main cellular source of acidity. Inactivation of ASS1 in cancer would be expected to lead to accumulationof upstreammetabolites such as glutamine and ammonia (derived from glutamine). Increase in these metabolites could be vital for the pH gradient in themaintenance of an alkalized pHi that is crucial for cancer cell survival, invasion, and migration. In contrast, depletion of ASS1's upstream substrates in ASS1-deficient cancer cells would be expected to confer vulnerability. TCA, tricarboxylic acid; NHE-1, Na(þ)/H(þ) exchanger 1; CAIX, carbonic anhydrase 9. B, The FITwas performed for individuals who presented at the Metabolic Disease Unit of the Sheba Medical Center for evaluation of elevated serum creatine phosphokinaselevels. The graph demonstrates the results of seven male subjects tested in the last five years in the age range of 8–28 years. Five out of the tested subjects wereconsidered to have normal results since they had a coupled elevation of both lactate and ammonia in their serum after three minutes of exercise (blue circles) ascompared with baseline levels (orange circles), while two subjects depicted by the filled red circles showed elevation of either lactate or ammonia without theexpected accompanying elevations of the other and was thus suspected to have inborn error of metabolism. (Continued on the following page.)

Silberman et al.

Cancer Res; 79(3) February 1, 2019 Cancer Research520




age, sex, race, and tumor's genomic instability and the tissue oforigin.

hs t; patientð Þ � h0s tð Þ exp�bCA9 CA9þ bage ageþbGII GII

�;

where s is an indicator variable over all possible combinationsof patients' stratifications based on race, sex, and cancer type.hs is the hazard function (defined as the risk of death of patientsper time unit), and h0sðtÞ is the baseline-hazard function at timet of the sth stratification. The model contains three covariates:(i) CA9 expression (CA9), (ii) age of the patient (age), (iii)genomic instability index (GII). GII measures the relativeamplification or deletion of genes in a tumor based on theSCNA. Given si as the absolute of log ratio of SCNA of gene i ina sample relative to normal control, GII of the sample is givenas in Bilal and colleagues (18):

GII ¼ 1=NXN

1

Iðsi > 1Þ:

The bs are the regression coefficients of the covariates, whichquantify the effect of covariates onpatient survival, determined bystandard likelihood maximization (19). We performed a similaranalysis to identify the association with survival of MCT4 (lowMCT4: n ¼ 905 patients; high MCT4: n ¼ 931 patients) andCA9þMCT4 combined expression (low CA9þMCT4: n ¼ 905patients; high CA9þMCT4: n ¼ 932 patients) with patient sur-vival. We performed a control analysis by repeating the sameanalysis with the samples that overexpress ASS1 (top tertile).

Metabolomics analysisOsteosarcoma cells were seeded at 8 � 105 cells per 10-cm

plate and incubated with 6.5-mL fresh HCO3� free RPMI

(Biological Industries 01-104-5A) under a CO2-free atmo-sphere for 72 hours. Subsequently, medium aliquots werecollected for glutamine uptake quantification by amino acidanalyzer, (Biochrome 30þ, Biochrome). For intracellular glu-tamine levels, cells were than washed with ice-cold saline, lysedwith a solution of 50% methanol in water with ribitol asinternal standard (2 mg/mL), and quickly scraped, followed bythree freeze–thaw cycles in liquid nitrogen. The insolublematerial was pelleted in a cooled centrifuge (4�C) and thesupernatant was collected for consequent gas chromatogra-phy/mass spectrometry (GC/MS) analysis. Samples were driedunder airflow at 42�C using Techne Dry-Block Heater withsample concentrator (Bibby Scientific Limited).

GC/MS and cellular glutamine measurementsDried samples were treated with 40 mL of a methoxyamine

hydrochloride solution (20 mg/mL in pyridine) at 37�C for90 minutes while shaking followed by incubation with 70 mLN, O-Bis (trimethylsilyl) trifluoroacetamide (Sigma) at 37�Cfor additional 30 minutes. Samples were then centrifuged andallowed to stand at room temperature for 2hours before injection.The results were normalized to ribitol as an internal standardand to the protein levels of each sample. GC/MS analysiswas performed using a gas chromatograph (7820AN, AgilentTechnologies) interfaced with a mass spectrometer (5975 AgilentTechnologies). A HP-5ms capillary column 30 m � 250 mm �0.25 mm (19091S-433, Agilent Technologies) was used. Heliumcarrier gaswasmaintained at a constantflow rate of 1.0mLmin�1.The GC column temperature was programmed from 70 to 150�Cvia a ramp of 4�C min�1, 150–215�C via a ramp of 9�C min�1,215–310�C via a ramp of 25�C min�1 and maintained at 310�Cfor additional 5 minutes. The electron impact ionization MSoperated in full scan mode from m/z 30 to 500. The inlet andMS transfer line temperatures were maintained at 250�C and310�C, respectively. The ion source temperature was 280�C.Sample injection (1–2 mL) was in splitless mode.

Glutamine uptake measurementsMedium (100 mL) from cell culture was mixed 1:1 with 5%

sulfur salicylic acid (Sigma) containing 250 mmol/L of norleu-cine (Sigma) as internal standard. The mixture was allowed tostand at 4�C for 30 minutes. The precipitate was centrifuged for10 minutes at 10,000� g at 4�C. The supernatant was taken andfiltered through 0.2-mm membrane filter and injected intoamino acid analyzer (Biochrom 30þ). Results were normalizedto norleucine and to the protein levels of each sample. Back-ground glutamine measurements from cell-free culture weresubtracted.

Glutamine measurements in tumorsFollowing euthanization, tumors were removed and incu-

bated for 6 hours with freshly prepared RPMI supplementedwith 4 mmol/L L-glutamine, (ALPHA-15N, 98%, CambridgeIsotope Laboratories, Inc) and dialyzed FBS. Next, tumors weregrinded using a stainless steel mesh grid and ice-cold saline toproduce single-cell suspensions. Cells were further washedtwice with saline and pelleted. The pellet was lysed with asolution of 50% methanol in water with ribitol as internalstandard (2 mg/mL), followed by three freeze–thaw cycles inliquid nitrogen. The insoluble material was pelleted in a cooledcentrifuge (4�C) and the supernatant was collected, dried under

(Continued.)The normal ranges for serum lactate and ammonia levels are 6.0–18.0 mg/dL and 31.0–123.0 mcg/dL, respectively. pCORR, 0.893 was calculated forhealthy subjects after three minutes of exercise. pCORR, 0.311 was calculated for all tested subjects including those suspected to have IEM. C, Extracellularammonia measurements in media of fibroblasts generated from a patient with CTLN I as compared with normal human dermal fibroblasts (NHDF). Theexperiment was performed in triplicates and repeated twice. D, A representative IHC staining (DAB) of a wild-type mouse liver for ASS1 and GS. Magnification,�10,�20. PV, portal vein; CV, central vein. Glutamine synthetase was used as a central vein biomarker. E, The predicted metabolic activity of the three proximalurea cycle enzymes CPS1, OTC, and ASS1 following pHi changes. The activity is predicted to decrease in cancer cells with acidic pHi, in comparison with normalhealthy cells (green), while that of glutathione dehydrogenase (dehydroascorbate reductase) is predicted to increase in acidic pH in both cancer and healthycells. The activity of lactate dehydrogenase, which has been reported to decrease in cancer cells with acidity (31), served as an internal control for the otherpredictions. Fluxes are given in mmol/mgDW/h. Mean values and their corresponding SDs are estimated across a dozens of genome-scale metabolic models,representing cell lines from each of the NCI-60 and HapMap panels. F, A representativeWestern blot image showing ASS1 protein levels extracted fromosteosarcoma cells following 72 hours of incubation with or without bicarbonate in the media and CO2 in the incubator (-CO2/-HCO3

� orþCO2/þHCO3�

conditions). G, A representativeWestern blot image showing ASS1 protein levels extracted from osteosarcoma cells following 24 hours of incubation ineither pHe 7.4 or pHe 6.3 media using 25mmol/L PIPES. We used Student t test (two-tailed) for statistical analysis in C. The graph is presented as anaverage� SD (� , P� 0.05).






air flow at 42�C, and subjected to GC/MS analysis. Totalglutamine masses were normalized to ribitol and to the proteinlevels of each tumor.

Cell culturesAll cell lines were authenticated and cultured using standard

procedures: 37�C humidified incubator with 5% CO2, RPMI(Invitrogen) supplemented with 10% heat-inactivated FBS, 1%penicillin–streptomycin, and 2 mmol/L glutamine. B16-F10 cellswere cultured using DMEM (Gibco) supplemented with the sameingredients. MNNG/HOS, LOX-IMVI, Malme-3M, OVCAR-3,MCF-7, U2OS, and B16-F10 cell lines were purchased from theATCC. Murine colonMC-38 and Lewis lung carcinoma (LLC) celllines were a kind gift from the laboratories of Profs. Eran Elinavand Lea Eisenbach (Weizmann Institute, Rehovot, Israel), respec-tively. All cells were routinely tested for Mycoplasma using aMycoplasma EZ-PCR Test Kit (20–700-20, Biological Industries).

Viability assaysHCO3

�-free solution contained RPMI solution without glu-tamine and bicarbonate (Biological Industries 01-104-5A) sup-plemented with 4 mmol/L glutamine, 1% penicillin–strepto-mycin, 10% heat-inactivated FBS, and titrated to pH 7.4 using 5mol/L NaOH. For control conditions, we used 22 mmol/LHCO3

�-contained RPMI solution supplemented with the sameingredients. CO2/HCO3

� buffers contained RPMI without glu-tamine and bicarbonate (Sigma R7755) supplemented with 4mmol/L glutamine, 1% penicillin–streptomycin, 10% heat-inactivated FBS, and either 22 mmol/L HCO3

� (pH 7.4) or1.75 mmol/L HCO3

� (pH 6.3) in accordance with the Hen-derson–Hasselbalch equation. Buffers were left under 5% CO2/95% atmosphere and the pH was adjusted using 1 mol/L KOHand 6 mol/L HCl. For cell viability assays done under hypoxia,we used RPMI solution without glutamine and glucose (Sig-ma), supplemented with 11 mmol/L glucose, 1% penicillin–streptomycin, 10% FBS, and with or without 4 mmol/Lglutamine.

For experiments done under �CO2/�HCO3� versus þCO2/

þHCO3�, cells were seeded in 6-well plates at 8� 104 cells/well in

triplicates. After 24 hours for adherence of the cells (time 0), cellswere washed and the medium was replaced to 1.5 mL of eitherfresh HCO3

�-free RPMI or HCO3�-contained RPMI (as described

before), respectively, and the cells were incubated under theindicated atmosphere, for 72 hours. For experiments done underpHe 7.4 (22 mmol/L HCO3

�; 5% CO2) versus pHe 6.3(1.75 mmol/L HCO3

�; 5% CO2), 2 � 105 cells per well wereplated in 6-well plates in triplicates. After 24 hours for adherenceof the cells (time 0), cells were washed once with saline solution(0.9%) and incubated under the bicarbonate buffers at thespecified pHe for 24 hours. At the end of the incubation timingsand following several washings, cells were fixated with 4%praformaldehyde and further stained with crystal violet(0.2%). Color destaining was performed using 10% acetic acidin H2O. Absorbance was measured at 590 nm. For experimentsdone under hypoxic conditions with the five different humancell lines, cells were seeded in 6-well plates at 8 � 104 cells perwell in triplicates. After 24 hours (time 0), cells were washedwith PBS and the medium was replaced to 1.5 mL of freshlyprepared RPMI with or without glutamine, for 48 hours underhypoxic conditions (1% O2) followed by crystal violet stainingand destaining. In the case of control versus shASS1 MCF-7 cells

with either glutamine depletion, 6-Diazo-5-oxo-L-norleucine(DON), S0859, or 4,4'-Diisothiocyanatostilbene-2,2'-disulfo-nic acid (DIDS) under hypoxic conditions, 1.8 � 105 cells perwell in triplicates were seeded. Time 0 was considered after 24hours for adherence of the cells and the cells were incubatedwith freshly prepared RPMI as described before, using theindicated different inhibitors.

Invasion assaysU2OS pLKO empty vector (EV) and shASS1 osteosarcoma cells

were seeded at 1� 105 cells/well in 12-well plates. At the next day,cells were incubated for 24 hours under either normoxic orhypoxic conditions (1% O2). Then, cells were trypsinized andseeded in duplicates or triplicates in 24-well transwell (Corning)coated with 50% Matrigel (BD Biosciences) in the absence ofserum in the top chamber and with 20% serum in the bottomchamber as an attractant, followed by another 24 hours incuba-tion under the same conditions. For the miRNA experiments,U2OS osteosarcoma cells were seeded at 5 � 104 cells/well in12-well plates. At the next day, cells were transfected (usingLipofectamine 2000 transfection reagent) with either 300 pmol(diluted in Dharmacon Si buffer X1) nontargeting control (Dhar-macon miRIDIANmiRNA Mimic Negative Control) or miR-224-5p (Dharmacon miRIDIAN miRNA Mimic). Forty-eight hourspost-transfection, cells were trypsinized and seeded in duplicatesor triplicates in 24-well transwells (Corning) as described previ-ously, followed by another 24 hours incubation under normoxicconditions. For all experiments, we additionally seeded cells inregular 24-well plates in parallel to seeding the cells in 24-welltranswells. At the end of incubation, transwells were washed withPBS,fixated in 4%paraformaldehyde, and then stainedwith 0.2%crystal violet. Cells thatwere stainedon thebottomchamber outerpartwere capturedwith an inverse lightmicroscopeNikon cameraand quantified by ImageJ. The percentage of invading cells was

calculated as average area of cells on the transwellaverage area of cells seeded in regular well � 100%.

Migration assaysU2OS osteosarcoma cells were seeded at 1 � 105 cells per well

in 6-well plates. At the next day, cells were transfected with either300 pmol miR-224-5p or nontargeting (NT) control and incu-bated for 24 hours. Then, cells were trypsinized and equal amountof cells was seeded in the top chamber in a technical triplicateusing a 24-transwell plate (Corning) and in the absence of serum.As an attractant, we used 10% serum in the bottom chamberand incubated the cells for another 24 hours under normoxicconditions. At the end of incubation, transwells werewashedwithPBS,fixated in 4%paraformaldehyde, and then stainedwith 0.2%crystal violet. The migrated cells were stained on the bottomchamber outer part and were captured with an inverse lightmicroscope Nikon camera. Quantification was done usingImageJ.

pHi measurementspH 7.4 HEPES buffer contained 25 mmol/L HEPES,

140 mmol/L NaCl, 5 mmol/L KCl, 1 mmol/L MgSO4,10 mmol/L glucose, and last to add 2 mmol/L CaCl2.CO2/HCO3

� buffers (bicarbonate buffers) were preparedby bubbling 5% CO2/95% air and contained the amountof NaHCO3

� that yielded the desired pH according tothe Henderson–Hasselbalch equation: pH ¼ 10�6.15 þ log

Silberman et al.





([HCO3�]/ [CO2]). Specifically, pH 7.4 CO2/HCO3

� buffercontained 22 mmol/L HCO3

�, 115 mmol/L NaCl,4.5 mmol/L KCl, 11 mmol/L glucose, 1 mmol/L MgCl2,1 mmol/L CaCl2; pH 6.6 CO2/HCO3

� buffer contained:3.5 mmol/L HCO3

�, 133.5 mmol/L NaCl, 4.5 mmol/L KCl,11 mmol/L glucose, 1 mmol/L MgCl2, 1 mmol/L CaCl2. NaClfinal concentration was changed to allow final osmolarity of300 mOsm/kg. Nigericin buffer contained 10 mmol/L nigericinin 25 mmol/L HEPES, 105 mmol/L KCl, and 1 mmol/L MgCl2.Nigericin buffers were titrated to the required pH using 1 mol/LKOH and 1 mol/L HCl.

PIPES buffers contained RPMI solution without glutamine andbicarbonate (Biological Industries 01-104-5A) and supplementedwith 25mmol/LPIPES, 22mmol/LHCO3

�, 4mmol/L glutamine,1% penicillin–streptomycin, 10% heat-inactivated FBS. PIPESbuffers were titrated to pH 7.4 or pH 6.3 using 1 mol/L KOH or6mol/LHCl, respectively. In the case ofU2OSpLKOempty vector(EV) and shASS1 osteosarcoma under �CO2/�HCO3

� condi-tions, cells were washed twice with HEPES buffer and thenCarboxy SNARF-1 AM (5 mmol/L in HEPES buffer) was loadedon the cells as the membrane permeate ester for 5–10 minutes in37�C. Next, cells were washed twice with HEPES buffer beforeincubation with a fresh HEPES buffer in 37�C. Cell imaging wasperformed using 24-well glass bottom black plates (Cellvis) in anOperetta automated imaging system (PerkinElmer) at 40� mag-nification equippedwith a temperature control chamber andwiththe appropriate excitation and emission filters (Ex 520–550 nm;Em 585–605 nm and 650–700 nm). Each field was defined as aregion of interest (ROI). At least five different ROIs with 50%confluence were chosen per cell type. Fluorescencemeasurementswere taken immediately. Mean fluorescence intensity at the twoemission wavelengths was measured per ROI and converted topHi values using nigericin (10 mmol/L; ref. 20) following a ratiocalculation (585–605 nm/650–700 nm). At the end of eachmeasurement, control cells were loaded with EIPA (25 mmol/L)supplemented directly into the well, serving as a positive controlto reduce the pHi. All imaging analysis was done by Harmony-image analysis tool of the Operetta system.

In the case of the five different cell lines expressing differentlevels of ASS1 or in the case of U2OS pLKO empty vector (EV) andshASS1 osteosarcoma incubated under alternating pHe solutions,we used confocal spinning disc (Zeiss) equipped with a temper-ature control, a CO2 chamber and the appropriate excitation andemission filters (Ex 512 nm; Em 580 and 640 nm). Imagingacquisitionwas performed using a�40magnification (oil). In thecase of the five different cell lines, following the incubationperiod, cells were washed twice with a HEPES buffer and loadedwith Carboxy SNARF-1 AM (5 mmol/L), as described previously.Imaging control cells was performed while using pHe 7.4 bicar-bonate buffer (22 mmol/L HCO3

�; 5% CO2). Imaging cells thatwere incubated under�CO2/�HCO3

� conditionswas performedwhile using a pHe 7.4 HEPES buffer. For each of the cell lines, theappropriate calibration curve was generated using nigericin(10 mmol/L). For imaging U2OS pLKO empty vector (EV) andshASS1 osteosarcoma incubated under alternating pHe solutions,we used pHe 7.4 bicarbonate buffer (22 mmol/L HCO3

�; 5%CO2) and pHe 6.6 bicarbonate buffer (3.5 mmol/L HCO3

�; 5%CO2). Eachmeasurement consisted of a fewROIswith n12 cellsper ROI. Mean fluorescence intensity at the two emissionwavelengths was measured per ROI and converted to pHi valuesusing nigericin (10 mmol/L) following a ratio calculation

(580 nm/640 nm). Each measurement represents the average ofa few ROIs. All measurements were done in biological duplicatesor triplicates and were repeated at least three times. Imaginganalysis was performed using ImageJ.

Protein analyses and Western blottingCells were lysed in RIPA buffer (Sigma-Aldrich) with

0.5% protease inhibitor cocktail (Calbiochem). Followingcentrifugation, the supernatant was collected and protein con-tent was evaluated by the Bradford assay. A total of 70–100 mgfrom each sample under reducing conditions were loaded intoeach lane and separated by electrophoresis on a 10% SDSpolyacrylamide gel. Following electrophoresis, proteins weretransferred to Immobilon transfer membranes (Tamar). Non-specific binding was blocked by incubation with TBST[10 mmol/L Tris–HCl (pH 8.0), 150 mmol/L NaCl, 0.1 %Tween 20) containing 3% BSA for 1 hour at room temperature.Membranes were subsequently incubated with antibodiesagainst ASS1 (1:500, 124465, Abcam), GAPDH (1:5,000,EPR-16891, Abcam), a -Tubulin (1:3,000, T5168, Sigma),Hif1a (1:500, ab2185, Abcam), MCT4 H-90 (1:500, sc-50329,Santa Cruz Biotechnology). Antibody was detected usingperoxidase-conjugated AffiniPure goat anti-rabbit IgG or goatanti-mouse IgG (Jackson ImmunoResearch) and enhancedchemiluminescence western blotting detection reagents(EZ-Gel, Biological Industries).

Gels were quantified by Gel Doc XRþ (Bio-Rad) and analyzedby ImageLab4.1 software (Bio-Rad). Theband areawas calculatedby the intensity of the band. The obtained value was then dividedby the value obtained from the loading control.

Immunofluorescence and tumor stainingHypoxia and ASS1 levels. Pimonidazole (hypoxyprobe-1 kit, hpi)was intraperitoneally injected, according to the manufacturer'sinstructions, 2 hours prior euthanization. Upon euthanization,the mice were rapidly perfused with cold 2.5% PFA in PBS. Thetumor was then excised and transferred to 1% PFA in PBS. Next,four-mm paraffin-embedded tissue sections were deparaffinizedand rehydrated. Sections undergoing for ASS1 staining wereincubated in cold acetone at�20�C for 7minutes and then boiledin 10mmol/L citric acid pH 6, for 10 minutes using a low boilingprogram in the microwave to break protein cross-links andunmask antigens. For hypoxic areas staining, adjacent 4-mmsection was taken and underwent antigen retrieval in10 mmol/L citric acid pH 6 for 10 minutes using a low boilingprogram in themicrowave. After preincubation with 20%normalhorse serum and 0.2% Triton X-100 for 1 hour at room temper-ature, sections were incubated with the primary antibodiesas follow; ASS1 (1:100, Abcam, ab124465); hypoxyprobe-1(1:35, hpi). All antibodies were diluted in PBS containing 2%normal horse serum and 0.2% Triton. Sections were incubatedovernight at room temperature followed by 48 hours at 4�C.Sections were washed three times in PBS and incubated withsecondary biotinylated IgG at room temperature for 1.5 hours,washed three times in PBS, and incubated with streptavidin-conjugated fluorophore (Jackson) at room temperature for addi-tional 90 minutes.

ASS1 and MCT4. Staining was performed on 4-mm paraffin-embedded tissue sections as described above or on 7-mmoptimalcutting temperature–embedded LLC sections. Sections were






incubated in cold acetone at�20�c for 7minutes and then boiledin 10mmol/L citric acid pH 6, for 10 minutes using a low boilingprogram in the microwave to break protein cross-links andunmask antigens. After preincubation with 20% normal horseserum and 0.2% Triton X-100 for 1 hour at room temperature,sections were incubated with both primary antibodies as follow;MCT4 (1:100, SantaCruzBiotechnology, sc-50329); ASS1 (1:100,Abcam, ab124465). All antibodies were diluted in PBS containing2% normal horse serum and 0.2% Triton. Sections were furthertreated as described above.

Liver IHC stainingFollowing euthanization, liver from wild-type mice was har-

vested and transferred to 4% PFA in PBS. Next, 2-mm paraffin-embedded tissue sections were deparaffinized and rehydrated.Endogenous peroxidase was blocked with 3%H2O2 inmethanol.Sections undergoing glutamine synthetase (GS) staining wereincubated in cold acetone at�20�C for 7minutes and then boiledin TRIS-EDTA for 10 minutes using a low boiling program in themicrowave to break protein cross-links and unmask antigens.Sections undergoing for ASS1 staining were incubated asdescribed above. After preincubation with 20% normal horseserum and 0.2% Triton X-100 for 1 hour at room temperature,sectionswere incubatedwith the primary antibodies as follow:GS(1:100, ab73593, Abcam); ASS1 (1:100, ab124465, Abcam).Sections were washed three times in PBS and incubated withsecondary biotinylated IgGat room temperature for 1.5 hours andwashed three times in PBS. Following incubation with avidin–biotin complex, a DAB (Sigma) reaction was performed.

RNA extraction and complementary DNA synthesisU2OS osteosarcoma or B16melanoma cells were seeded at 1�

105 cells/well in a 6-well plate and at the next day were eitherincubated for 24–48 hours under hypoxic conditions (1%O2) orunder normoxic conditions. At the end of incubation, RNA wasextracted from the cells by using PerfectPure RNA Cultured CellKit (50-PRIME). cDNA was synthesized from 1 mg RNA by usingqScript cDNA Synthesis Kit (Quanta).

Quantitative PCRDetection of ASS1 and CAD cDNAs was performed using SYBR

Green PCR Master Mix (Tamar) and the required primers. Thedetection of hsa- or mmu-miR-224-5p was performed usingPerfeCTa SYBR Green SuperMix (Quanta) and Primers for hsa-ormmu-miR-224-5p were supplied byQuanta. Primer sequenceswere as follows: human ASS1: forward, 50-TTGAAATTTGCTG-AGCTGGTGTA-30; reverse, -50-AGCCTGAGGGAATTGATGTT-GAT-30; human HPRT: forward, 50-ATTGACACTGGCAAAA-CAATGC-30; reverse: 50-TCCAACACTTCGTGGGGTCC-30; humanGAPDH: forward, 50-AAATTGAGCCCGCAGCCTCCC-30; reverse:50-AGCGATGTGGCTCGGCTGG;Mouse ASS1: forward, 50-TGTA-CCTGAACGAAGTTGCGG-30; reverse, 50-ACCTCGGGACTT-CATTCCAAT-30; mouse HPRT: forward, 50-GCAGTACAGCCC-CAAAATGG-30; reverse, 50-GGTTCCTTTTCACCAGCAAGCT-30;human CAD: forward, 50-TTTGCAGTCCTTCCCGCTTC-30;reverse, 50-GCCGGTTTGAAACACCACTT-30. For analysis, we usedStepOne Real-Time PCR Technology (Applied Biosystems).

Luciferase assayPrimers were designed by Primer BLAST software. The

primers had forward Xhol (50-TAATCTCGAGACCCGTGTA-

CAATGAGGAGCT-30) and reverse Notl (50-TTATGCGGCCGC-TAGAAATAAAAGACTAGTGTC-30) restriction sites. PCR reac-tion was performed in a final reaction of 50 mL in 0.2-mL PCRtubes containing 250 ng of template DNA, 5 mL of each primerat 10 mmol/L, 25 C GoTaq Green Supper Mix x2 (Promega).Sterile ultrapure deionized water was used as a negativecontrol. PCR reaction was carried out with initial denatur-ation at 95�C for 5 minutes followed by 35 thermal cyclesof denaturation at 95�C for 1 minute, annealing at 57�C for30 seconds and extension at 72�C for 30 seconds. Finalextension was accomplished at 72�C for 5 minutes. Theamplified 30 UTR fragment was analyzed by electrophoresisin 1% agarose gels.

ASS1 30 UTR DNA fragment was purified (using Qiagen gelextraction kit) and ligated in a 20 mL volume containing 2 mLligation buffer, 1 mL ligase (New England Biolabs), 8.5 mL frag-mented ASS1 30 UTR and 1.5 mL psiCHECK2 plasmid (Addgene).The ligated plasmid was transformed into dh5a-competent bac-teria by heat shock transformation. The transformed bacteria wereseeded on LB-ampicillin agar plate. The recombinant plasmidwaspurified (using Wizard plus SV Minipreps DNA PurificationSystem) and the correct sequence and orientation were verifiedby sequencing with forward (50-TTCCAACGCTATTGTCGAGG-30) and reversed (50-AAACACTCTCATTCGGAGGC-30) cloningvalidation primers.

U2OS cell line was seeded at 1 � 105 cells per well in a6-well plate. At the following day, cells were transfected with300 pmol miR-224-5p or nontargeting (NT) control (usingLipofectamine 2000 transfection reagent). Following incuba-tion of 24 hours, cells were transfected again with 1 mgpsiCHECK2-ASS1 30 UTR plasmid and incubated for addition-al 24 hours. Cells were collected and measured (Cytation 5Cell Imaging Multi-Mode Reader) according to the dual-luciferase reporter assay system manufacturer's protocol(Promega).

ASS1 SilencingIn the case of U2OS, cells were infected with pLKO-based

lentiviral vector with or without the human ASS1 short hairpinRNA (shRNA) encoding two separate sequences (RHS4533-EG445, GE Healthcare, Dharmacon). In the case of MCF-7 cells,GFP shRNA (shGFP) was used to infect control cells. Transducedcells were selected with 2 mg/ml puromycin.

miRNA transfection, extraction, and complementary DNAsynthesis

U2OS osteosarcoma or B16 melanoma cells were seeded at1 � 105 cells/well in a 6-well plate. At the next day, cells wereeither transfected with 300 pmol nontargeting control orwith miR-224-5p. After 24 hours, the cells were transfectedagain with a complementary 600 pmol specific miR inhibitor(Dharmacon miRIDIAN miRNA Hairpin Inhibitor) plus anegative nonspecific sequence (Dharmacon miRIDIAN miRNAInhibitor Negative Control). Cells were incubated for 48 hoursunder normoxic conditions. In the case of hypoxic conditions(1% O2) versus normoxic conditions, miR-224-5p was ext-racted after 24 hours (B16) or 48 hours (U2OS). In all cases,at the end of incubation, miRNA was extracted from the cells byusing miRNeasy Mini Kit (Qiagen). cDNA for miRNA wassynthesized from 0.7–1 mg RNA by using qScript MicroRNAcDNA Synthesis Kit (Quanta).

Silberman et al.





Animal studiesAll experiments with mice were approved by the Institutional

Animal Care andUse Committee (Weizmann Institute of Science,Rehovot, Israel). Tumors did not exceed the limits of more than10% of the animal weight and were no longer than 1.5 cm inlength in any dimension. For induction of tumors, 10 millionMALME-3M melanoma cells suspended in 400 mL PBS with 5%Matrigel (4132053, Corning) were subcutaneously injected to 8-to 12-week-oldmale SCIDmice thatwere purchased fromHarlan.MC38 tumors were induced by an orthotopic implantation of1 � 105 MC38 CRC into 8- to 10-week-old male C57BL/6 mice.Tumor sectionswere isolated from the colonic injection site at day14. LLC tumorswere induced by subcutaneously injecting 2� 106

LLC-D122 cells in PBS into the left hind limb of 8- to 10-week-oldmale C57BL/6 mice. Tumor sections were isolated at day 21.Mice were monitored for survival and tumor burden twice aweek by a veterinarian investigator who was blinded to theexpected outcome.

ResultsBecause cancer cells frequently hijack physiologic processes

that benefit normal cells' survival and proliferation, we beganour investigations into the relationship between lactate andammonia levels by analyzing biochemical data of subjects with-out cancer who underwent a FIT. This test is used to evaluatemuscular energy production following complaints of either exer-cise-induced muscle pain, or unexplained elevations in serumcreatine phosphokinase levels (21, 22). As expected, in compar-ison with lactate and ammonia levels at baseline, we found thatafter 3 minutes of FIT, a significant correlation exists between theelevations of ammonia and lactate in healthy subjects, whilepatients with IEM demonstrate an abnormal response(Fig. 1B). In addition, we noted higher levels of ammonia, withno significant rise in lactate levels, in primary citrullinemia type Ifibroblasts, compared with our findings in the media of fibro-blasts from healthy controls (Fig. 1C; Supplementary Fig. S1A).These results suggest that ammonia and lactate levels function intandem in normal physiologic states, whereas disruption of thisassociation might signify metabolic dysregulation.

Ex vivo, at the tissue level, we studied livers fromwild-typemicein which a gradient from high oxygen in the portal vein, to lowoxygen levels in the central vein, had been previously established(23). Here, we found that ASS1 expression is confined to theportal vein, and is diminished in the hypoxic regions surroundingthe central vein, where glutamine synthase is expressed (Fig. 1D).Because lactate levels near the central vein are relatively high (24),the downregulation of ASS1 expression, together with the bio-chemical changes associated with its depletion in normal states,suggest that in hypoxic regions, low ASS1 levels might provide ametabolic benefit resulting from increased ammonia/glutaminelevels.

To determine the relevance of these findings to cancer, weperformed an unbiased exploration of the predicted activity of theproximal UC enzymes at different pHi values in normal and incancerous cells, usinggenome-scalemetabolicmodels (seeMaterialsand Methods; ref. 14). Compared with healthy cells, we found thatin cancer cells under acidic pHi conditions, the activity, or "flux," ofASS1, and of the more upstream enzymes in the UC, is predictedto decrease, while the activity of glutathione dehydrogenase (dehy-droascorbate reductase) is predicted to increase (Fig. 1E).

To validate the potential metabolic association between thedownregulation of ASS1 and pHi, we induced either gradualor acute alterations in extracellular pH (pHe), which triggerintracellular pH acidic stress. First, we grew osteosarcoma cellswithout bicarbonate in the culture medium, in a CO2-free atmo-sphere. In this buffer-free system, intracellular acidic stress isincreased, due to a gradual decrease of the pHe, and the lack ofextrinsic buffering (Supplementary Fig. S1B). Consistent with ourhypothesis, and in silico predictions, ASS1 expression in osteosar-coma cells decreased concomitantly with prolonged acidity(Fig. 1F; Supplementary Fig. S1C). We further validated ourresults by inducing acute pHe changes, independent of bicarbon-ate limitation (25). In line with our previous findings, ASS1 levelswere found to be lower under acidic conditions (pHe 6.3, ascomparedwithmeasurements taken under baseline conditions ofpHe 7.4; Fig. 1G; Supplementary Fig. S1D). Thus, inducingacidic stress using two different systems, reduces ASS1 levels inosteosarcoma cells, and corroborates our findings in nontrans-formed cells.

Next, we evaluated whether ASS1 levels facilitate cancer cells'survival under acidic conditions in other malignancies. To do so,we grew five different human cancer cell lines, each of whichexpressed different levels of ASS1, in the buffer-free systemdescribed above. Two of these cancer lines, MNNG/HOS (oste-osarcoma) and Lox IMVI (melanoma), display low ASS1 levels,whereas the other three cancer cell lines, U2OS (osteosarcoma),OVCAR-3 (ovarian carcinoma), and MCF-7 (breast adenocarci-noma) express varying levels of ASS1 (Fig. 2A; Supplementary Fig.S2A). As expected, all cancer cells, irrespective of ASS1 levels,were less viable and demonstrated a lower pHi when grownunder acidic conditions induced by the lack of extrinsic buff-ering (Supplementary Figs. S2B and S2C). Interestingly, thestress induced by the buffer-free conditions highlighted thedifferences in cell viability, favoring survival of cancer cells withlow ASS1 and a higher pHi, as compared with cancers expres-sing higher ASS1 levels (Fig. 2B).

To decipher the biochemical and cellular changes caused byspecific alterations in ASS1 expression levels, we next focused onU2OS osteosarcoma cells in which we knocked down ASS1 usingshASS1 (Supplementary Fig. S2D; ref. 6). To dissect the specificeffect of ASS1 downregulation on pHi, and to avoid any contam-ination of our results due to cell death, we first performed short-term live-cell imaging experiments under both baseline (pHe 7.4;22mmol/L HCO3

�) and acidic conditions (pHe 6.6; 3.5 mmol/LHCO3

�), by using the Henderson–Hasselbalch equationand limiting the amount of bicarbonate in the media (26).Encouragingly, we found that shASS1U2OS cells display a higherpHi in both instances (Fig. 2C), coupledwith a significant survivaladvantage following their prolonged incubation at pHe 6.3(Fig. 2D). By completely depleting bicarbonate from the mediaand incubating the cells in the buffer-free system, we furtherconfirmed the survival advantage and associated higher pHiof shASS1 U2OS grown under acidic conditions (Fig. 2E–F;Supplementary Fig. S2E).

Because glutamine has been associated with cellular defenseagainst acidic stress (25), and because glutamine levels areelevated in instances of inherited ASS1 deficiency (27), we furtherassessed whether ASS1 downregulation affects both intra- andextracellular glutamine levels. We first determined that ASS1-depleted U2OS cells accumulate higher levels of glutaminewhen grown under baseline normoxic conditions (Fig. 2G).






Figure 2.

ASS1-deficient cancer cells have survival advantage in acidic environment. A, A representativeWestern blot analysis showing ASS1 expression levels in fivedifferent cancer cell lines. B, Viability assays of five different cancer cells divided into two groups based on ASS1 expression as either "High ASS1" or "Low ASS1,"following 72-hour incubation under�CO2/�HCO3

� conditions (viability is quantified relative to day 0; left). Right, pHi measurements of five different cancer cellsdivided into two groups based on ASS1 expression as either "High ASS1" or "Low ASS1," following 72-hour incubation under�CO2/�HCO3

� conditions. C,Arepresentative pHi curve based on quantifications of live confocal microscope images of U2OS osteosarcoma cells infected with either empty vector (control) orshASS1 and incubated in alternating pH solutions. (Continued on the following page.)

Silberman et al.





Importantly, we found that when grown in acidic conditions,cancer cells characterized by ASS1 downregulation consume lessglutamine from the medium (Fig. 2H, left), yet retain similarintracellular glutamine levels (Fig. 2H, right). Because glutathioneis synthesized from glutamine, and plays an essential role inantioxidant metabolism (25, 28), we measured its levels underthese acidic conditions. As expected from our enzymatic activityprediction (Fig. 1E), we found higher glutathione levels in ASS1-depleted cells, comparedwith control cells (Fig. 2I).Hence, cancercells in which ASS1 is downregulated display higher intracellularglutamine levels, enabling them, in acidic states, to depend less onexternal sources of glutamine, to better preserve alkalinized pHiand to synthesize glutathione.

Mechanisms to preserve pHi in cancerous cells are necessary toenable the cells to survive, and maintain their carcinogenicproperties (29–31). In recent years, it has been observed thatcancer cells generate a pHe that is lower than their pHi. Thisunique, "reversed" pH gradient across cancer cell membranes hasemerged as a new hallmark of cancer, and is thought to empowercancer cells' metastatic potential, while preserving its normalcellular functions (29–31). One of the main transcription factorsthat regulates the pH gradient in cancer is HIF1a (10). Underhypoxic conditions, HIF1a is upregulated, to promote pHi pres-ervation by inducing, for example, the expression of the acidextruder MCT4, as well as the expression of the membrane CO2-hydrating enzymes carbonic anhydrase 9 and 12 (CA9, CA12;refs. 10, 32). Interestingly, HIF1a has also been reported todownregulate ASS1 levels (33).

To study the association between ASS1 and the hypoxic geneexpression in cancer, we first analyzed the TCGA database fortumors expressing hypoxia-related genes (16), inwhich the cancercell purity was estimated to be >75% (15). Using this dataset, wefound that ASS1 expression levels are significantly lower in tumorswith a high hypoxic gene signature (Fig. 3A). Because cancer cellsmight express HIF1a irrespective of hypoxia (34), we repeated theanalysis, while excluding HIF1a target genes. Again, we found acomparable, significant trend associating low ASS1 levels andhypoxia, independent of HIF1a expression (Supplementary Fig.S3A). Our analysis of the TCGA data also yielded a significantinverse correlation between the expression levels of ASS1 andMCT4 in different tumor types (Fig. 3B). The inverse correlationbetween ASS1 andMCT4 was also observed at the protein level inhuman osteosarcoma cells, following hypoxia induction (Fig. 3C;Supplementary Fig. S3B). As previously reported (8), we con-firmed that HIF1a upregulation associates with the downregula-tion of CAD expression levels (Supplementary Fig. S3C). To test

whether the downregulation of ASS1 induced by HIF1a underhypoxia can be further augmented under acidosis, we incubatedosteosarcoma cells under acute acidic conditions in the presenceor absence of hypoxia. Indeed, ASS1 levels were downregulatedmore under hypoxic pHe 6.3 as comparedwith normoxic pHe 6.3(Supplementary Fig. S3D). Taken together, these results suggestthat under hypoxic and acidic microenvironment, ASS1 down-regulation supports survival mechanisms in cancerous cells thatdiffer from those we previously identified for promoting prolif-eration under normoxic conditions (6).

To test these findings in vivo, we studied human melanomatumors expressing ASS1 implanted in mice, following injectionwith a hypoxia-dependent reactive probe (35). In these tumors,we found a clear demarcation between the hypoxic regions, andregions expressing ASS1 (Fig. 3D). Furthermore,MCT4 expressioninmelanoma tumors, inmurine LLC, and in colonMC38 tumors,was limited to regions that did not express ASS1 (Fig. 3E; Sup-plementary Figs. S3E and S3F).

Next, we measured glutamine levels in melanoma tumors, andfound that shASS1 tumors grown in vivo, containedhigher levels ofintracellular glutamine, compared with tumors expressing ASS1(Fig. 3F). To evaluate the clinical relevance of these findings forhumans with cancer, we performed a pan-cancer Kaplan–Meirsurvival analysis using the TCGA database. We found that indi-viduals with cancer whose levels of the HIF1a target genes CA9and MCT4 are elevated, alongside decreased ASS1 expression,have poorer chances of survival, compared with patients whosecancers are characterized by similar elevations in CA9 andMCT4,but have increased ASS1 expression (Fig. 3G; SupplementaryFig. S3G). Collectively, these results suggest that in vivo, ASS1expression is downregulated in hypoxic regions, whereas expres-sion of HIF1a�pHi–related target genes is upregulated. It wouldappear that this coregulation increases glutamine levels, promotestumor fitness, and worsens patients' survival.

In accordance with the cellular context, HIF1a can downregu-late different target genes either directly, or via miRNA (36).Interestingly, we found that miR-224-5p, which has been foundto be regulated by HIF1a in invasive melanoma and gastriccancers (37, 38), displays a conserved binding site in mammalson the 30UTR site of ASS1 (Supplementary Fig. S4A). Our analysisof the TCGA data suggested opposite trends between ASS1 down-regulation and miR-224-5p upregulation in different cancers(Supplementary Fig. S4B). We confirmed the specific associationof miR-224-5p upregulation and ASS1 downregulation in hyp-oxia in both osteosarcoma and in murine B16 cancer cells (Sup-plementary Figs. S3C and S4C and S4D). Furthermore, we

(Continued.) Using the Henderson-Hasselbach equation, solutions were supplemented with bicarbonate to reach the specified pH. For pH 7.4 buffer, we used 22mmol/L HCO3

�; 5% CO2, and for pH 6.6 buffer, we used 3.5 mmol/L HCO3�; 5% CO2. Each point of a pHi value is an average of n 10 cells (ROIs).D, Viability

experiments showing increased survival for shASS1 U2OS cells as compared with control cells following 24-hour incubation under pHe 6.3 (1.75 mmol/L HCO3�;

5% CO2). Viability at pHe 6.3 (1.75 mmol/L HCO3�; 5% CO2) of each clone was normalized to its viability at pHe 7.4 (22 mmol/L HCO3�; 5% CO2). We did notdetect significant differences in pHi between control and shASS1 U2OS cells under pHe 6.3. E, Viability measurements of U2OS osteosarcoma cells infected witheither empty vector or shASS1 and incubated under�CO2/�HCO3

� conditions (viability was quantified relative to day 0). F,Averaged pHi measurements ofcontrol U2OS osteosarcoma cells infected with either empty vector or shASS1 following incubation under�CO2/�HCO3

� conditions for 72 hours. Allmeasurements were done using pH 7.4 HEPES buffer medium in an incubator without CO2. Results are presented as an average of n 5 independentexperiments.G, Cellular glutamine measurements using GC/MS of control and shASS1 osteosarcoma cells grown underþCO2/þHCO3

� conditions at time point24 hours (baseline). H,Glutamine uptake measurements using amino acid analyzer from the media of either control or shASS1 osteosarcoma cells grown under�CO2/�HCO3

� conditions for 72 hours (left). Right, Cellular glutamine measurements of control and of shASS1 osteosarcoma cells grown under�CO2/�HCO3�

conditions at time point 72 hours. I, Total glutathione levels in control and shASS1 osteosarcoma cells grown under�CO2/�HCO3� conditions at time point 72

hours. For the multiple cell line comparisons in B, we used two-way ANNOVA (with a priori contrast: MNNG/HOS, LOX-IMVI vs. U2OS, MCF-7 and OVCAR-3). Forall other comparisons, Student t test was used (two-tailed) for statistical analysis. Unless otherwise specified, each graph represents the average of n 3experiments� SD (� , P� 0.05; �� , P� 0.01).






Figure 3.

ASS1 levels are regulated together with HIF1a� gene network for pH regulation. A, Pan-cancer analysis of the TCGA database demonstrating that tumors withlow ASS1 expression levels (red) have a higher hypoxic score, based onmeasurement that includes HIF1a target genes' expression (effect size: 0.180; Wilcoxonrank sum P < 0.05, see Materials and Methods). B, Pan-cancer analysis of the TCGA database for ASS1 and MCT4, as compared with the expression of these genesin the respective normal tissues. MCT4; SLC16A3. pCORR¼�0.23. C, A representativeWestern blot for the protein expression of HIF1a target genes in humanosteosarcoma (U2OS) cells exposed to either normoxia or hypoxia for 24 hours. D, A representative immunofluorescence image of a melanoma (Malme-3M)paraffin-embedded tumor section following the administration and staining of the hypoxia-sensitive pimonidazole probe (red) and of ASS1 (green).Magnification,�4. E, A representative immunofluorescence image of a Malme-3Mmelanoma paraffin tumor section stained for MCT4 (green) and ASS1 (red).Magnification,�6. F, Ex vivomeasurement of tumor glutamine levels using GC/MS in control Malme-3M versus shASS1Malme-3M tumor cells. G, Kaplan–Meierplots showing decreased survival in patients with tumors having either overexpression of CA9 (left), MCT4 (middle), or of both (right) when ASS1 expression isdownregulated. CA9 (low CA9, n¼ 905; high CA9, n¼ 932 patients); Cox HR¼ 1.14 (P < 0.06), log rank P < 1E�16 (DAUC¼ 0.24); MCT4 (low MCT4, n¼ 905patients; high MCT4, n¼ 931 patients); Cox HR¼ 1.19 (P < 0.01), log rank P < 1E-16, (DAUC¼ 0.22). CA9 and MCT4 (low CA9þMCT4, n¼ 905 patients; high CA9þMCT4, n¼ 932 patients) - Cox HR¼ 1.23 (P < 0.004), log rank P < 1E�16 (DAUC¼ 0.21). We used Student t test (two-tailed) for statistical analysis in F thatrepresents the average of five mice per group� SD (� , P� 0.05).


Silberman et al.




determined that overexpression ofmiR-224-5p can downregulateASS1 levels at the RNA level in both cancers, independent ofhypoxia (Fig. 4A; Supplementary Fig. S4E), and that this down-regulation also manifests at the protein level (Fig. 4B). Using aluciferase assay,we specifically demonstrated the direct binding ofmiR-224-5p to ASS1 30UTR, independent of hypoxic conditions(Fig. 4C). Importantly, the downregulation of ASS1 solely bymiR-224-5p was sufficient to increase the extent of invasionand migration in osteosarcoma cells (Fig. 4D and E). The cor-egulation of ASS1 levels by HIF1a and miR-224-5p stronglysupport the notion that low ASS1 levels in hypoxia providecancer with a metabolic benefit that has identifiable phenotypicconsequences.

To assess whether the decreased survival of patients with ASS1downregulation in cancers expressing hypoxia genes could be

associated with pH gradient–related clinical phenotypes such asmetastasis formation (31), we utilized the TCGA database tocompare gene expression in metastatic tumor samples expressinghypoxia genes, to those that donot. Similar toourfindings in solidtumors (Fig. 3A), we found that ASS1 expression is significantlylower in metastases expressing hypoxia gene–related signatures(Fig. 5A). Even when excludingHIF1a target genes, we observed acomparable, significant trend correlating low ASS1 expressionlevels and hypoxic metastases (Supplementary Fig. S5A).

To explore these results in greater depth, we compared theinvasive ability of osteosarcoma cells under normoxic and hyp-oxic conditions, and found higher rates of invasion in the latter(Fig. 5B, left). To determine the contribution of ASS1 down-regulation to hypoxia-independent invasion, we evaluated theinvasion of osteosarcoma cells following ASS1 downregulation in

Figure 4.

miR-224-5p downregulates ASS1. A,ASS1 mRNA (left) levels are downregulated in osteosarcoma cells following transfection with miR-224-5p (right) undernormoxia. Addition of anti-miR-224-5p rescued ASS1 levels. B, A representativeWestern blot image showing reduction in ASS1 expression levels inosteosarcoma cells following transfection with miR-224-5p (48 hours; left). Right, averaged gel quantification. C, Luciferase assay showing diminished ratioof ASS1 30 UTR Renilla to firefly luciferases after transfection with miR-224-5p. D,Osteosarcoma cells invade significantly more following transfection withmiR-224-5p under normoxia. E,Osteosarcoma cells migrate significantly more following transfection with miR-224-5p under normoxia. We used Student t test(two-tailed) for statistical analysis. Unless otherwise specified, each graph represents the average of n 3 experiments� SD (� , P� 0.05; �� , P� 0.01).






Figure 5.

Cancer cells expressing low levels of ASS1 have increased sensitivity to glutamine and bicarbonate depletion. A, Pan-cancer analysis based on TCGAmetastaticsamples showing that tumors with low ASS1 levels have a higher hypoxia score (effect size: 0.088; Wilcoxon rank sum, P < 3.35E-3; see Materials and Methods).B, U2OS osteosarcoma cancer cells invade significantly more upon incubation of 48 hours under hypoxic as compared with normoxic conditions (left). Right,U2OS osteosarcoma cells with ASS1 depletion invade significantly more than control cells following incubation of 48 hours under normoxia. C,MCF-7 breastcancer cells with ASS1 downregulation are less viable following incubation under hypoxia for 48 hours with either glutamine-free conditions (left) or DON(50 mmol/L; right). D,MCF-7 breast cancer cells with ASS1 downregulation are less viable following incubation under hypoxia with either DIDS (300 mmol/L;24-hour incubation; top) or S0859 (100 mmol/L; 48-hour incubation; bottom). Viability is normalized to nontreated cells (cells grown in the absence ofinhibitors). The experiments described in C and D have been performed under pHe 7.4. E, Illustration of the different metabolic consequences following ASS1downregulation in normoxia versus hypoxia or acidity. Unless otherwise specified, we used Student t test (two-tailed) for statistical analysis. Each graphrepresents the average of n 3 experiments� SD (� , P� 0.05).


Silberman et al.




normoxia. Similar to our findings following miR-224-5p trans-fection, we determined that even under normoxic conditions,shASS1-osteosarcoma cells are more invasive than controls(Fig. 5B, right).

To test whether the survival and invasive advantages wefound in cancers with low ASS1 expression in hypoxia relateto the increase in ASS1 upstream metabolites, we grew the fivedifferent human cancer cell lines expressing different levels ofASS1 under hypoxic conditions (Fig. 2A), with or withoutglutamine in the culture media. Indeed, we found that cancercells with low levels of ASS1 demonstrate increased sensitivityto glutamine depletion, compared to cells with high ASS1 levels(Supplementary Fig. S5B). To confirm the specific susceptibilityof cancers with low ASS1 levels to ASS1 upstream metabolites,we knocked down ASS1 in MCF-7 breast adenocarcinoma cells(Supplementary Figs. S5C and S5D). We next grew these cells inglutamine-free conditions, with pharmacologic inhibitors thatdecrease the levels of glutamine and bicarbonate: DON,a glutamine antagonist (39), and S0859, which inhibits theNBC sodium bicarbonate cotransporter (40). Because S0859has been found to have nonspecific effects, we repeated ourexperiments using the HCO3

� transport inhibitor, DIDS. Usingall agents, we find that MCF-7 breast cancer cells subjected toASS1 knockdown are more sensitive to the depletion of bothglutamine and the alkali bicarbonate (Fig. 5C and D; Supple-mentary Fig. S5E).

DiscussionCancer cells display a distinct "reversed" pH gradient, with a

constitutively higher pHi, compared with their pHe. This gradientenables disease progression by promoting malignant cell prolif-eration, inmost instances enabling these cells to evade apoptosis,metabolically adapt, migrate, and invade neighboring tissues,irrespective of tumor pathology, genetics, and origins. Not sur-prisingly, pHi is emerging as a newhallmark in cancer studies, anda potential target for therapy (31, 41). One of the outcomes ofthis reversed pH gradient is that it can be thermodynamicallycostly for cells, necessitating that they extrude more acids (42).Hence, cancer cellsmust promote alternative pathways for ventingand buffering acids, to cope with the additional burden ofintracellular acidity. Herein, we show that cancer cells with ASS1downregulation display higher pHi levels as compared withcontrol cells, and that this metabolic benefit is associatedwith higher glutamine levels. Under hypoxic conditions, ASS1 isdirectly downregulated in cancer cells by HIF1a andmiR-224-5p,with phenotypic consequences on cells' survival and invasivecapacity. Clinically, we find that the downregulation of ASS1,together with the induction of HIF1a -pHi related target genes-MCT4 and CA9 are significantly associated with decreased patientsurvival. Collectively, our results suggest that ASS1 downregula-tion in cancerous processes may play an important role in theHIF1a�pHi network.

To clarify our findings concerning the impact of acidityon cancer cells, we used two acidification systems that are inde-pendent of hypoxia. In so doing, we determined that ASS1 isdownregulated under acidic conditions; notably, under suchconditions cancer cells with low ASS1 levels have a higher pHi,and demonstrate increased survival. While ASS1 downregulationby itself reduces the formation of argininosuccinic acid (43), weshow here that the survival benefit of ASS1-depleted cancer

cells results from the increase in ASS1's upstream substrateglutamine, which contributes to the maintenance of pHi andglutathione synthesis. Thus, ASS1 downregulation contributesdirectly to cellular defenses against both hypoxia-dependent and-independent acidic stresses.

The metabolic rewiring that occurs during carcinogenesis ingeneral, and the decreased vascularization that takes place as thetumor grows, generate acidic levels that requires intracellular pHalkalization for the maintenance of protein activities (44).Although protons generated during hypoxia are not the onlysource of acidity in tumors, many of the changes related tohypoxia are induced by acidosis. This is best exemplified byHIF1a, which, on one hand increases acid extrusion, while onthe other hand, is stabilized and protected from degradationby acidity (45). Whereas reduced pHe in the tumor microenvi-ronment has been linked with metastases formation, alkalinizedpHi has been shown to support cell adhesion processes necessaryfor invasion and migration. Here we find that ASS1-depletedcancer cells maintain a higher pHi at both baseline andacidic conditions, which may explain, at least in part, ourfindings of augmented invasion for ASS1-depleted cancer cells.We further demonstrate that ASS1 downregulation is regulated byHIF1a and, independently, by acidity, to promote these cancer-ous phenotypes.

Our findings have translational relevance as they imply thatcancers expressing ASS1 may benefit too from treatment witharginine-depleting agents as theymayhavehypoxic/acidic regionsin which ASS1 is downregulated. In addition, we show that ASS1downregulation confers on cancerous cells a sensitivity to drugstargeting ASS1 upstream substrates. As such, our results exposecancer-specific vulnerabilities with a therapeutic potential.

Disclosure of Potential Conflicts of InterestNo potential conflicts of interest were disclosed.

Authors' ContributionsConception and design: A. Erez, A. Silberman, O. Boukobza AssayagDevelopment of methodology: A. Erez, A. Silberman, O. Boukobza AssayagAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): A. Erez, A. Jacob, R. Keshet, A. Sarver, K.B. Halpern,B. Pode-Shakked, Y. AniksterAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): A. Silberman, O. Boukobza Assayag, A. Jacob, S.Limanovich, J.S. Lee, E. Persi, S.C.S. Nagamani, I. UlitskyWriting, review, and/or revision of the manuscript: A. Erez, A. Silberman,S.C.S. Nagamani, E. RuppinAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases):O. Boukobza Assayag, S. Limanovich, A. Sarver,K.B. HalpernStudy supervision: A. Erez, E. RuppinOther (performed in vitro and in vivo experiments): A. SilbermanOther (performed in vitro miR experiments): O. GoldmanOther (performed in vitro experiments): L. Adler, J. Frug, S. GalaiOther (performed in vivo experiments): N. StettnerOther (performed the metabolic pH-dependent computational analysis):E. PersiOther (exchanging ideas): Y. Zaltsman-AmirOther (supervising IHC studies): R. Eilam

AcknowledgmentsWe acknowledge and thank the Weizmann Institute for providing financial

and infrastructural support. We greatly appreciate the technical support pro-vided by Yogev Sela and the statistical contributions of Ron Rotkopf. A. Erez isincumbent of the Leah Omenn Career Development Chair and is supported by






research grants from the European research program (CIG618113,ERC614204), the Israel Science Foundation (ISF- 1343/13; 1952/13), and fromtheMinerva grant award (711730). A. Erez received additional support from theAdelis Foundation, the Henry S. and Anne S. Reich Research Fund, the DuklerFund for Cancer Research, the Paul Sparr Foundation, the Saul and TheresaEsman Foundation, from Joseph Piko Baruch, and from the estate of FannieSherr. E. Ruppin is supported by grants from the ISF (658/13) and theNIH (R33CA225291-01).

The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby markedadvertisement in accordance with 18 U.S.C. Section 1734 solely to indicatethis fact.

ReceivedApril 9, 2018; revisedAugust 30, 2018; acceptedDecember 17, 2018;published first December 20, 2018.

References1. Quinonez SC, Thoene JG. Citrullinemia type I. 2014.2. Thoene JG. Citrullinemia Type I. In: Pagon RA, Adam MP, Bird TD,

Dolan CR, Fong CT, Stephens K, editors. Seattle (WA): GeneReviews;2004.

3. Long Y, TsaiW-B,Wangpaichitr M, Tsukamotot T, Savaraj N, Feun LG, et al.Arginine deiminase resistance in melanoma cells is associated with met-abolic reprogramming, glucose dependence, and glutamine addiction.Mol Cancer Ther 2015;12;2581–90.

4. Kobayashi E, Masuda M, Nakayama R, Ichikawa H, Satow R, Shitashige M,et al. Reduced argininosuccinate synthetase is a predictive biomarker for thedevelopment of pulmonary metastasis in patients with osteosarcoma.Mol Cancer Ther 2010;9:535–44.

5. Delage B, Luong P, Maharaj L, O'Riain C, Syed N, Crook T, et al. Promotermethylation of argininosuccinate synthetase-1 sensitises lymphomas toarginine deiminase treatment, autophagy and caspase-dependent apopto-sis. Cell Death Dis 2012;3:e342.

6. Rabinovich S, Adler L, Yizhak K, Sarver A, Silberman A, Agron S, et al.Diversion of aspartate in ASS1-deficient tumours fosters de novopyrimidinesynthesis. Nature 2015;527:379–83.

7. Tsai WB, Aiba I, Lee SY, Feun L, Savaraj N, Kuo MT. Resistance to argininedeiminase treatment in melanoma cells is associated with induced argi-ninosuccinate synthetase expression involving c-Myc/HIF-1alpha/Sp4.Mol Cancer Ther 2009;8:3223–33.

8. Chen KF, Lai YY, Sun HS, Tsai SJ. Transcriptional repression of humancad gene by hypoxia inducible factor-1alpha. Nucleic Acids Res2005;33:5190–8.

9. Long Y, Tsai WB, Chang JT, Estecio M, Wangpaichitr M, Savaraj N, et al.Cisplatin-induced synthetic lethality to arginine-starvation therapy bytranscriptional suppression of ASS1 is regulated by DEC1, HIF-1alpha,and c-Myc transcription network and is independent of ASS1 promoterDNA methylation. Oncotarget 2016;7:82658–70.

10. Chiche J, Brahimi-Horn MC, Pouyssegur J. Tumour hypoxia induces ametabolic shift causing acidosis: a common feature in cancer. J Cell MolMed 2010;14:771–94.

11. Nissim I. Newer aspects of glutamine/glutamate metabolism: the role ofacute pH changes. Am J Physiol 1999;277:F493–7.

12. HuangW, ChoiW, Chen Y, Zhang Q, DengH, HeW, et al. A proposed rolefor glutamine in cancer cell growth through acid resistance. Cell Res2013;23:724–7.

13. Spinelli JB, YoonH, Ringel AE, Jeanfavre S, Clish CB, HaigisMC.Metabolicrecycling of ammonia via glutamate dehydrogenase supports breast cancerbiomass. Science 2017;358:941–6.

14. Persi E, Duran-Frigola M, Damaghi M, Roush WR, Aloy P, Cleveland JL,et al. Systems analysis of intracellular pH vulnerabilities for cancer therapy.Nat Commun 2018;9:2997.

15. Yoshihara K, Shahmoradgoli M, Martinez E, Vegesna R, Kim H, Torres-Garcia W, et al. Inferring tumour purity and stromal and immune celladmixture from expression data. Nat Commun 2013;4:2612.

16. Benita Y, Kikuchi H, Smith AD, Zhang MQ, Chung DC, Xavier RJ.An integrative genomics approach identifies Hypoxia Inducible Factor-1(HIF-1)-target genes that form the core response to hypoxia. Nucleic AcidsRes 2009;37:4587–602.

17. Mantel N. Evaluation of survival data and two new rank orderstatistics arising in its consideration. Cancer Chemother Rep 1966;50:163–70.

18. Bilal E, Dutkowski J, Guinney J, Jang IS, Logsdon BA, Pandey G, et al.Improving breast cancer survival analysis through competition-basedmultidimensional modeling. PLoS Comput Biol 2013;9:e1003047.

19. Therneau TM, Grambsch PM. Modeling survival data: extending the Coxmodel. New York, NY: Springer; 2000. p.350.

20. Grillo-Hill BK, Webb BA, Barber DL. Ratiometric imaging of pH probes.Method Cell Biol 2014;123:429–48.

21. Hogrel JY, Laforet P, Ben Yaou R, Chevrot M, Eymard B, Lombes A. A non-ischemic forearm exercise test for the screening of patients with exerciseintolerance. Neurology 2001;56:1733–8.

22. Tarnopolsky M, Stevens L, MacDonald JR, Rodriguez C, Mahoney D,Rush J, et al. Diagnostic utility of a modified forearm ischemic exercisetest and technical issues relevant to exercise testing. Muscle Nerve2003;27:359–66.

23. LeCluyse EL,Witek RP, AndersenME, PowersMJ. Organotypic liver culturemodels: meeting current challenges in toxicity testing. Crit Rev Toxicol2012;42:501–48.

24. KietzmannT.Metabolic zonationof the liver: the oxygen gradient revisited.Redox Biol 2017;11:622–30.

25. Huang WJ, Choi W, Chen YL, Zhang Q, Deng HT, He W, et al. A proposedrole for glutamine in cancer cell growth through acid resistance. Cell Res2013;23:724–7.

26. Hulikova A, Vaughan-Jones RD, Swietach P. Dual role of CO2/HCO3(-)buffer in the regulation of intracellular pH of three-dimensional tumorgrowths. J Biol Chem 2011;286:13815–26.

27. Wilson CJ, Lee PJ, Leonard JV. Plasma glutamine and ammonia concen-trations in ornithine carbamoyltransferase deficiency and citrullinaemia.J Inherited Metab Dis 2001;24:691–5.

28. Lamonte G, Tang X, Chen JL, Wu J, Ding CK, Keenan MM, et al. Acidosisinduces reprogramming of cellularmetabolism tomitigate oxidative stress.Cancer Metab 2013;1:23.

29. Reshkin SJ, GrecoMR, Cardone RA. Role of pHi, and proton transporters inoncogene-drivenneoplastic transformation. Philos Trans R Soc LondBBiolSci 2014;369:20130100.

30. Parks SK, Pouyssegur J. Targeting pH regulating proteins for cancer therapy-Progress and limitations. Semin Cancer Biol 2017;43:66–73.

31. Webb BA, Chimenti M, Jacobson MP, Barber DL. Dysregulated pH:a perfect storm for cancer progression. Nat Rev Cancer 2011;11:671–7.

32. Swietach P, Patiar S, Supuran CT, Harris AL, Vaughan-Jones RD. The role ofcarbonic anhydrase 9 in regulating extracellular and intracellular ph inthree-dimensional tumor cell growths. J Biol Chem 2009;284:20299–310.

33. Qing G, Skuli N, Mayes PA, Pawel B, Martinez D, Maris JM, et al. Com-binatorial regulation of neuroblastoma tumor progression by N-Myc andhypoxia inducible factor HIF-1alpha. Cancer Res 2010;70:10351–61.

34. Krieg M, Haas R, Brauch H, Acker T, Flamme I, Plate KH. Up-regulation ofhypoxia-inducible factors HIF-1alpha and HIF-2alpha under normoxicconditions in renal carcinoma cells by von Hippel-Lindau tumor suppres-sor gene loss of function. Oncogene 2000;19:5435–43.

35. Bayin NSPlacantonakis DG. Evaluation of vascularity, blood perfusion,and oxygen tension in tumor xenografts with fluorescent microscopy.Methods Mol Biol 2018;1741:199–207.

36. Liu W, Shen SM, Zhao XY, Chen GQ. Targeted genes and interactingproteins of hypoxia inducible factor-1. Int J Biochem Mol Biol 2012;3:165–78.

37. Hwang HW, Baxter LL, Loftus SK, Cronin JC, Trivedi NS, Borate B, et al.Distinct microRNA expression signatures are associated with melanomasubtypes and are regulated by HIF1A. Pigment Cell Melanoma Res2014;27:777–87.

38. He C,Wang L, Zhang J, XuH.Hypoxia-inducible microRNA-224 promotesthe cell growth, migration and invasion by directly targeting RASSF8 ingastric cancer. Mol Cancer 2017;16:35.


Silberman et al.




39. Hernandez-Davies JE, Tran TQ, ReidMA, Rosales KR, Lowman XH, PanM,et al. Vemurafenib resistance reprograms melanoma cells towards gluta-mine dependence. J Translat Med 015;13:210.

40. Ch'en FF, Villafuerte FC, Swietach P, Cobden PM, Vaughan-JonesRD. S0859, an N-cyanosulphonamide inhibitor of sodium-bicarbonate cotransport in the heart. Br J Pharmacol 2008;153:972–82.

41. Sharma M, Astekar M, Soi S, Manjunatha BS, Shetty DC, Radhakrishnan R.pH Gradient reversal: an emerging hallmark of cancers. Recent PatAnticancer Drug Discov 2015;10:244–58.

42. Hulikova A, Black N, Hsia LT, Wilding J, Bodmer WF, Swietach P. Stromaluptake and transmission of acid is a pathway for venting cancer cell-generated acid. Proc Natl Acad Sci U S A 2016;113:E5344–E53.

43. Chen BC, Ngu LH, Zabedah MY. Argininosuccinic aciduria: clinical andbiochemical phenotype findings inMalaysian children.Malaysian J Pathol2010;32:87–95.

44. White KA, Grillo-Hill BK, Barber DL. Cancer cell behaviors mediated bydysregulated pH dynamics at a glance. J Cell Sci 2017;130:663–9.

45. Mekhail K, Gunaratnam L, Bonicalzi ME, Lee S. HIF activation by pH-dependent nucleolar sequestration of VHL. Nat Cell Biol 2004;6:642–7.






2019;79:518-533. Published OnlineFirst December 20, 2018.Cancer Res Alon Silberman, Omer Goldman, Odeya Boukobza Assayag, et al. Maintenance of Intracellular pH in CancerAcid-Induced Downregulation of ASS1 Contributes to the

Updated version

10.1158/0008-5472.CAN-18-1062doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerres.aacrjournals.org/content/suppl/2018/12/20/0008-5472.CAN-18-1062.DC1

Access the most recent supplemental material at:

Overview

Visual

http://cancerres.aacrjournals.org/content/79/3/518/F1.large.jpgA diagrammatic summary of the major findings and biological implications:

Cited articles

http://cancerres.aacrjournals.org/content/79/3/518.full#ref-list-1

This article cites 41 articles, 8 of which you can access for free at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected]

To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/79/3/518To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/lookup/doi/10.1158/0008-5472.CAN-18-1062

http://cancerres.aacrjournals.org/content/suppl/2018/12/20/0008-5472.CAN-18-1062.DC1

http://cancerres.aacrjournals.org/content/79/3/518/F1.large.jpg

http://cancerres.aacrjournals.org/content/79/3/518.full#ref-list-1

http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]

http://cancerres.aacrjournals.org/content/79/3/518


Acid-Induced Downregulation of ASS1 ... · Introduction The urea cycle (UC) is the main metabolic...

Documents

Transcript of Acid-Induced Downregulation of ASS1 ... · Introduction The urea cycle (UC) is the main metabolic...