metabolic characterization of hepatocellular...

Integrated Systems and Technologies

Metabolic Characterization of Hepatocellular CarcinomaUsing Nontargeted Tissue Metabolomics

Qiang Huang1, Yexiong Tan2, Peiyuan Yin1, Guozhu Ye1, Peng Gao1, Xin Lu1,Hongyang Wang2, and Guowang Xu1

AbstractHepatocellular carcinoma has a poor prognosis due to its rapid development and early metastasis. In this

report, we characterized the metabolic features of hepatocellular carcinoma using a nontargeted metabolicprofiling strategy based on liquid chromatography-mass spectrometry. Fifty pairs of liver cancer samples andmatched normal tissues were collected from patients having hepatocellular carcinoma, including tumortissues, adjacent noncancerous tissues, and distal noncancerous tissues, and 105 metabolites were filteredand identified from the tissue metabolome. The principal metabolic alternations in HCC tumors includedelevated glycolysis, gluconeogenesis, and b-oxidation with reduced tricarboxylic acid cycle and D-12desaturase. Furthermore, increased levels of glutathione and other antioxidative molecules, together withdecreased levels of inflammatory-related polyunsaturated fatty acids and phospholipase A2, were observed.Differential metabolite levels in tissues were tested in 298 serum specimens from patients with chronichepatitis, cirrhosis, and hepatocellular carcinoma. Betaine and propionylcarnitine were confirmed to confergood diagnostic potential to distinguish hepatocellular carcinoma from chronic hepatitis and cirrhosis.External validation of cirrhosis and hepatocellular carcinoma serum specimens further showed that thiscombination biomarker is useful for diagnosis of hepatocellular carcinoma with a supplementary role toa-fetoprotein. Cancer Res; 73(16); 4992–5002. �2013 AACR.

IntroductionHepatocellular carcinoma is one of the most common

malignancies worldwide. In China, chronic hepatitis B virusinfection is the primary risk factor for hepatocellular carci-noma (1, 2), and the majority of hepatocellular carcinomacases develop from hepatitis infections and subsequentcirrhosis. Rapid development and early metastasis are thetypical characteristics of hepatocellular carcinoma, whichalways results in a poor prognosis. The large populationinfected with hepatitis B virus (HBV) makes the prevention

of hepatocellular carcinoma a formidable task. Therefore,investigating the hepatocarcinogenesis mechanism is veryimportant for decreasing the incidence and mortality ofhepatocellular carcinoma.

The abnormal metabolism of cancer has been considered animportant characteristic of tumors, which could clarify thepathogenesis and provide potential therapeutic targets forclinical treatments (3). According to the Warburg effect, thederegulated energymetabolism of cancer cellsmay alsomodifymany related metabolic pathways that influence various bio-logical processes, such as cell proliferation and apoptosis. As acommon characteristic of cancer cells (4, 5), modified meta-bolism has been the focus of cancer research.

Metabolomics is a top–down platform in the field of sys-tems biology, which focuses on the dynamic changes of smallmolecules in response to the disturbance of the organism (6).Nontargeted metabolomic approaches are being widely usedfor the discovery of new biomarkers and investigation of thecarcinogenesis mechanism, and these approaches have alsobeen used to investigate chronic liver diseases (CLD) andhepatocellular carcinoma (7, 8). Some serum or urine meta-bolites, such as sphingosines (7), bile acids, glycine, trimethy-lamine-N-oxide, and dipeptides (8, 9), have been reported tobe differential metabolites or candidate biomarkers for CLDor HCC. Serum and urine are commonly considered to be apool of metabolites that reflect systemic metabolic deregula-tion in patients, and themarkers in these biofluids could reflectthe characteristics of the system during the course of diseases.

Authors' Affiliations: 1CAS Key Laboratory of Separation Science forAnalytical Chemistry, Dalian Institute of Chemical Physics, Chinese Acad-emy of Sciences, Dalian and 2International Cooperation Laboratory onSignal Transduction, Eastern Hepatobiliary Surgery Institute, the SecondMilitary Medical University, Shanghai, China

Q. Huang and Y-X. Tan contributed equally to this work.

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

CorrespondingAuthors:GuowangXu,CASKey Laboratory of SeparationScience for Analytical Chemistry, Dalian Institute of Chemical Physics,Chinese Academy of Sciences, 457 Zhongshan Road, Dalian 116023,China. Phone: 86-411-84379530; Fax: 86-411-84379559; E-mail:[email protected]; and Hongyang Wang, International Cooperation Labo-ratory on Signal Transduction, Eastern Hepatobiliary Surgery Institute, theSecond Military Medical University, Shanghai, China. Phone: 0086-21-25070856; Fax: 0086-21-65566851; E-mail: [email protected]

doi: 10.1158/0008-5472.CAN-13-0308

�2013 American Association for Cancer Research.

CancerResearch

Cancer Res; 73(16) August 15, 20134992

on January 22, 2020. © 2013 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst July 1, 2013; DOI: 10.1158/0008-5472.CAN-13-0308

http://cancerres.aacrjournals.org/

Tissue metabolomics is a useful tool for studying theabnormal metabolisms of diseases, and it can provide infor-mation about the metabolic modifications and the upstreamregulative mechanism in diseases (10, 11). More importantly,the systemic metabolic characteristics of tissues could pro-vide opportunities for exploring novel diagnostic markersor therapeutic targets for clinical applications (12). Tissuemetabolomics is conducted using a pairwise comparisonof different parts of tissue from each patient, which canremove individual differences, such as age, sex, region, andso on. The differences between the tumor cells and theirsurrounding host cells may reflect the interactions of thetumor and the host, which are important clues for studyingthe invasion and metastasis of tumors (13).In this study, 50 sets of matched hepatocellular carcinoma

tissues, including hepatocellular carcinoma tissue (HCT), adja-cent noncancerous tissue (ANT), and distal noncanceroustissue (DNT), were collected. The metabolic characteristics ofthe tumor tissues and the impact of the tumors on thesurrounding host cells were explored through metabolomicsstrategy based on liquid chromatography-mass spectrometry(LC/MS). On the basis of the defined differential metabolites,metabolic pathways and correlation networks were investi-gated and their potential for use in clinical diagnostics wasinvestigated.

Materials and MethodsChemicalsAcetonitrile and formic acid [high-performance liquid chro-

matography (HPLC) grade]were purchased fromMerck.Meth-anol (HPLC grade) was purchased from Tedia. Distilled waterwas filtered through a Milli-Q system (EMD Millipore Corpo-ration). Glycerophosphocholine, betaine, carnitine, hypoxan-thine, acetylcarnitine, methionine, niacinamide, propionylcar-nitine, 50-methylthioadenosine, glycoursodeoxycholic acid,glycocholic acid, glycodeoxycholic acid, glutathione disulfide,

fructose-1,6-bisphosphate, glucose-6-phosphate, palmitoy-lethanolamide, malic acid, fumaric acid, N-acetyl-L-alanine,g-aminobutyric acid, aspartic acid, taurine, serine, threonine,succinate, uric acid, 3-hydroxybutyric acid, xanthine, xantho-sine, valine, ethylmalonic acid, proline, glutamine, S-adenosyl-homocysteine, N-acetyl-L-methionine, leucine, phenylalanine,isoleucine, tryptophan, glycochenodeoxycholic acid, free fattyacid (FFA) C14:0, FFA C18:3, FFA C18:2, and FFA C18:1 werepurchased from Sigma-Aldrich. Lysophosphorylcholine (LPC)C18:2, LPC C16:0, and LPC C18:0 were purchased from AvantiPolar Lipids. Ammonium bicarbonate (NH4HCO3, HPLCgrade) was purchased from Fluka.

Clinical sample collection and pretreatmentWritten informed consents were obtained from the pati-

ents, and the study was approved by the ethics committeeof the Eastern Hepatobiliary Surgery Institute. Tissues werecollected from 50 patients (male/female, 39/11) havinghepatocellular carcinoma who ranged in age from 30 to70 years. Of these, 38 of the patients were HBsAg-positive,and 1 patient was anti-HCV-positive. In each patient, threetypes of tissues were resected during the operation, includ-ing HCTs, ANTs, and DNTs. The ANTs were collected fromless than 2 cm into the solid tumor border. During theoperation, DNTs were collected from the distal edge of theresected tissues, which were more than 2 cm from the borderof the solid tumor. The lesions were analyzed using histo-pathology studies. Detailed clinical information about thecollected samples is presented in Table 1. The tissue sampleswere directly placed into liquid nitrogen after surgicalresection and stored at �80�C until analysis.

To investigate the diagnostic potentials of the differentialmetabolites discovered from the tissues, a total of 298 fastingserum samples, including 81 chronic hepatitis, 78 cirrhosis and139 hepatocellular carcinoma samples, were collected andthe detailed clinical characteristics are described in Supple-mentary Table S1. Another batch of serum samples from

Table 1. Clinical characteristics of patients who have hepatocellular carcinoma

Characteristics Male, n ¼ 39 Female, n ¼ 11 P value

AFP (mg/L) – <20/>20 17/22 6/5 0.708Age, y 51.9 (30–70) 51.2 (38–65) 0.845ALT (U/L) 116.8 (20.0–417.0) 88.0 (22.0–495.0) 0.437AST (U/L) 47.3 (18.4–322.2) 37.6 (16.2–64.1) 0.527Cirrhosis/no cirrhosis 23/16 0/11HBsAg-positive/negative 30/9 8/3Anti-HCV-positive/negative 1/38 0/11Tumor diameter (cm) 7.1 (2–18) 7.1 (3.6–11) 0.996II 8 0III 21 7

Edmonson stageII þ III 2 1III þ IV 2 0Not determined 6 3

NOTE: Age, ALT, AST, and tumor diameter in both male and female were expressed as median (range).

Metabolic Modifications of HCC Based on Tissue Metabolomics

www.aacrjournals.org Cancer Res; 73(16) August 15, 2013 4993




25 cirrhosis and 22 patients with hepatocellular carcinomawas employed for external validation (Supplementary TableS2). The serum samples were stored at �80�C until analysis.

Tissue pretreatmentA piece of the tissue (100 mg) was mixed with 1 mL of cold

methanol/water (4:1, v:v) and then homogenized using a high-speed blender. After ultrasonication, the sample was placed onice for 20 minutes and then deproteinized by centrifugation at4�C (15,000 rpm for 10 minutes). Finally, 800 mL of the super-natant was freeze-dried at �52�C and dissolved in 100 mL ofmethanol/water (4:1, v:v) before analysis.

Serum pretreatmentFour hundred microliters of acetonitrile was added to

100 mL of serum and the mixture was vortexed for 1 minute.Then, at 4�C, the mixture was centrifuged for 10 minutes ata rotation speed of 12,000 rpm. After that, 400 mL of superna-tant was transferred and lyophilized in a freeze-dryer. Final-ly, the dried supernatant was dissolved with 100 mL water/acetonitrile (4:1, v:v) solution and 10 mL sample injectionfor LC/MS analysis. The serum pretreatment for capillaryelectrophoresis–mass spectrometry (CE–MS) analysis hasbeen described in supplementary materials.

Metabolic profiling analysisThe tissue metabolic profiling analysis was conducted on a

Thermo-Fisher Accela ultra-high-performance liquid chroma-tography (UHPLC) coupled to a linear ion-trap quadrupole(LTQ) Orbitrap hybrid mass spectrometry (MS) system(Thermo Fisher). A 2.1 � 100 mm BEH 1.7-mm C8 column(Waters) was used. The column oven was set at 60�C, and thesample manager temperature was maintained at 12�C. Theeluents A (water containing 0.1% formic acid) and B (aceto-nitrile) were employed in the electrospray ionization-positive(ESIþ) mode, whereas eluents C (water containing 5 mmol/LNH4HCO3) and D (95% CH3OH and 5% water containing5 mmol/L NH4HCO3) were used in the ESI negative (ESI�)mode. The flow rate was 0.35 mL/min with a linear gradientelution over 30 minutes. From the start to 1 minutes, B forthe ESIþmode (or D for ESI�) was held at 5%, linearly increas-ed to 30% during the next 9 minutes, linearly increased to90% during an additional 10 minutes, and then increased to100% in 2 minutes and kept constant for 4 minutes. Subse-quently, B (or D) was returned to 95% in 0.1 minutes and heldfor an additional 3.9 minutes before returning to the initialconditions. The sample sequence was random. The MS sprayvoltages were 4.5 kV in the ESIþ mode and 3.0 kV in the ESI�

mode. The capillary temperature was set at 300�C with thesheath gas at 35 arbitrary units and the aux gas at 5 arbitraryunits. The tube lens was set to 100 V and the mass scan rangewas set from 100 to 1000 m/z. The resolution of the Orbitrapwas set at 60,000. The tandem mass spectrometry (MS–MS)data were collected with the collision energy between 10 and35 eV.

The same LC/MS system and mobile phases were used forserummetabolic profiling with a 2.1� 100 mm HSS 1.8 mm T3column (Waters, Ireland). Instrument parameters of LC/MS

were similar to our previous work (14). The column temper-ature was maintained at 35�C and the sample manager wasset to 4�C. The analysis time for each serum sample was 22minutes. The resolution of the Orbitrap was set at 60,000 toensure the mass errors were less than 2 ppm. Both betaineand propionylcarnitine are polar metabolites, and the CE–MSsystem is appropriate for polar compounds analysis. Thus,the quantitative analysis in the external validation frompatients with hepatocellular carcinoma to cirrhosis was car-ried out by using the CE–MS system, and the detailed methodiss described in supplementary materials.

To ensure data quality formetabolic profiling, pooled qualitycontrol (QC) samples were prepared by mixing all of thesamples. Before analyzing the sample sequence, five QC sam-ples were run. During analysis of the sample sequence, one QCsample was run after every 10 injections, which is similar to ourprevious method (15).

Data processing and analysisFirst, the raw data were acquired and aligned using the

SIEVE software package (V1.2, Thermo Fisher) based on them/z value and the sample retention time. Before chemometricsanalysis, all of the detected ions in each sample from onesample class were normalized to the sum of the peak areadefined as 10,000 (16). After the missing values for each sampleclass were treated using the 80% rule (17), the statisticalsignificance was calculated using the Student t test (P <0.05) as implemented in the SPSS version 16 for Windowssoftware package.

Ions from both ESIþ and ESI� with statistical significanceweremerged and imported into the SIMCA-P program (version11.0, Umetrics) for multivariate analysis. Principal componentanalysis (PCA) and partial least squares discriminate analysis(PLS-DA) were applied with unit variance (UV) scaling (16).The parameters of the models, such as the R2X, R2Y, and Q2Y,and the R2Y-, Q2Y-intercepts, were analyzed to ensure thequality of the multivariate models and to avoid the risk ofover-fitting. The normalized amount of each metabolite wasplotted in a histogram using the Origin software package(version 8.0). Hierarchical cluster analysis (HCA) was con-ducted using the MeV software package (version 4.5.1), andthe correlation network was constructed using the Cytoscapesoftware package. Receiver operating characteristic curve(ROC) and binary logistic regression were applied to the serumdata using SPSS software.

ResultsA total of 50 patients with hepatocellular carcinoma were

enrolled in this study. The diagnosis of hepatocellular carci-noma was confirmed by histopathologic studies after surgery.Twenty-three patients had serum a-fetoprotein (AFP) levelsthat were less than the cut-off value (20 mg/L). The Edmonsonstages of the patients (nine were not determined) rangedfrom stages II to IV. There was no statistical significanceobserved between male and female patients (Student t test,P > 0.05) in the clinical tests employed. Steatotic liver tissuewas not observed in any of the samples.

Huang et al.

Cancer Res; 73(16) August 15, 2013 Cancer Research4994




Metabolic profiling of liver tissuesIn this study, a nontargeted metabolomics strategy was

applied. The details of the metabolic profiling method,including sample preparation, metabolite extraction, andLC/MS analysis, were reported in our previous study (18).Supplementary Fig. S1 in supplementary materials presentsthe typical base peak chromatograms (BPC) of the HCT,ANT, and DNT groups. The metabolic profiles of the ANTand DNT groups were similar, but they were substantiallydifferent from that of the HCT group in both ESIþ and ESI�

modes.Validations of method, including linearity, precision, sta-

bility, and recovery, were also carried out (18), which indi-cates that this tissue metabolomics method is reliable.Furthermore, the PCA score plot (Supplementary Fig. S2A)exhibited a clear cluster of the pooled QC samples (R2X, 0.647;Q2, 0.404), indicating that the sample analysis sequence hadsatisfactory stability and repeatability.After the peak alignment and the removal of the missing

values (17), a total of 880 ions had significantly changed(Student t test, P < 0.05) between the HCT and DNT group,942 ions between the HCT and ANT group, and only 46 ionsbetween the ANT and DNT group were maintained (Supple-mentary Fig. S2A).For further analysis of the metabolic differences between

the DNT and HCT group, all of the significant ions from theESIþ and ESI� modes were merged and imported into theSIMCA-P software package. Subsequently, PLS-DA was appli-ed to the classification of cancer and noncancerous tissues.As illustrated by the PLS-DA score plot (Supplementary Fig.S2B), the noncancerous tissues were clearly separated fromthe hepatocellular carcinoma tissues. The cumulative R2Y andQ2 were 0.804 and 0.657 respectively, when two componentswere calculated. No over-fitting was observed according tothe results of the chance permutation (R2Y-intercept was0.420 and Q2-intercept was �0.153). However, there was noobvious separation trend between the ANT and DNT groups.

Differential metabolites between HCT and DNTPLS-DA S-plot (Supplementary Fig. S2C) can be used to

define the differential metabolites for distinguishing theDNT group from the HCT group, 390 ions with a variableimportance in the project (VIP) more than one were selectedfor subsequent chemical structure identification (19). Basedon our previously published strategy (20), the followingsteps were used for the identification of chemical structuresin this study. First, the quasimolecular ions were confirmed.Secondly, the exact masses of the monoisotopic molecularweights were used to search the online databases, such asthe Human Metabolome Database (http://www.hmdb.ca/),Metlin (http://metlin.scripps.edu/) and the Mass Bank(http://www.massbank.jp/) (21). Then, the MS/MS spectrawere also analyzed to verify the structure of the identifiedmetabolites. Subsequently, 44 metabolites in ESIþ mode and65 in ESI� mode were identified, and some of them werefurther confirmed using authentic standard samples. Thenames of the compounds and the related enzymes arepresented in Supplementary Table S3.

The relative average normalized quantities of the identi-fied differential metabolites in the HCT and ANT groupscompared to those in the DNT group were plotted in a heatmap (Fig. 1) using the MeV software package. The metabo-lites were clustered according to their Pearson correlationcoefficients. Overall, the changes in these important meta-bolites in the HCT samples were quite different from thosein the noncancerous tissues.

Based on the knowledge of these differential metabolitesand an online database of metabolic pathways (Kyoto Encyclo-pedia of Genes and Genomes, http://www.genome.jp/kegg/),a map of the hepatocellular carcinoma-related metabolic path-ways was constructed (Fig. 2). Several metabolic pathways weremodified by the tumor cell, for example, an increase in glycolysismetabolites (6-phosphogluconic acid) and a decrease in tricar-boxylic acid cycle (TCA)metabolites, including succinate, fuma-rate, malate, and succinic acid semialdehyde, were observed.Most of the amino acids (AA) and their related metabolitessignificantly increased in the HCT group. The levels of twoglucogenic AAs (serine and threonine), which can be trans-formed into glucose for energy, significantly increased in theHCT group. Glutamine, glutathione, branched-chain aminoacids (BCAA), including valine, leucine, and isoleucine, and thearomatic amino acids (ArAA), such as tryptophan, phenylala-nine, and tyrosine, were also increased in the HCT group. Theincreasedcatabolismmaybe responsible for the increase in totalAAs. These modifications to the host metabolism representedthe influence of the cancer cells on energy metabolism, whichhave also been confirmed in colon and stomach tumor tissues(13).

The change in the free fatty acids was not the same as thatof the AAs. The levels of saturated fatty acids (SFA) andmonounsaturated fatty acids (MFA) increased, whereas thelevels of polyunsaturated fatty acids (PUFA) decreased(Fig. 3A). The ratio of FFA C18:2 to C18:1 also decreasedin the HCT group (Fig. 3B), which suggested inhibitionof D-12 desaturase. Interestingly, the ratio of acetylcarnitineto carnitine (C2/C0) was upregulated in the HCT group,whereas the ratio of propionylcarnitine to carnitine (C3/C0)was downregulated (Fig. 3C). The ratio of C2/C0 indicatedthat the b-oxidation of even-numbered fatty acids wasupregulated in the HCT group. The ratio of C3/C0 conflictedwith BCAA (Fig. 3D), which also reflected an increase in lipidoxidation for the production of energy in the tumor tissue(22). A decreased level of LPC was observed in the cancertissues. However, phosphorylcholine (PC) and phosphoetha-nolamine (PE) levels were significantly greater in the cancertissues (Fig. 2).

Correlation network of differential metabolites in tumortissue

To investigate the latent relationships of the differentialmetabolites, a correlation network diagram was constructedusing the Cytoscape software (Fig. 4). All of the differentialmetabolites (880 ions) were included in the diagram to obtaina global view of tumor metabolism. We observed that themetabolites could be divided into two primary groups: upregu-lated and downregulated. These groups were bridged by lipids.






Fatty acids, including SFA, MFA, and PUFA, were located inthe center of the metabolic network. Levels of glycolysis-relatedmetabolites and some AAs were upregulated in the tumor.Lipids, especially phospholipids, were the most correlatedmetabolites in the downregulated group. There was a decreasedamount of bile acids in the tumor tissues. Because the metab-olism of phospholipids and bile acids, the deregulation ofthese metabolites may reflect low differentiation levels of thetumor cells. Unlike long-chain acylcarnitines, the quantity of

short-chain acylcarnitines decreased in the cancer tissues.The correlation of the metabolites in the TCA cycle andthe short-chain acylcarnitines supported the hypothesis thatenergy metabolism shows low efficacy in tumor tissues.

Investigation of the diagnostic potential of differentialmetabolites found in tissues

The tissue metabolome can more directly reflect meta-bolic deregulation than the body fluid metabolome, but

Figure 1. Heat map of the 105significantly changed metabolitesamong the DNT, ANT, and HCTgroups. The colors from green tored indicate the relative contents ofmetabolites in the ANT or HCTcompared with those in the DNTgroup.

Huang et al.





the biomarkers for clinical diagnosis should return from thetissue to the biofluids. Therefore, we also investigatedthe serum diagnostic potentials of these significantly differ-ent metabolites. Totally, 298 serum samples, including 81chronic hepatitis, 78 cirrhosis, and 139 hepatocellular carci-noma, were enrolled.Considering the complementary role of novel biomarkers

to AFP in clinical diagnosis, we focused more on the meta-bolites that were not related with AFP level. The results of

the Pearson correlation analysis (|Cij| < 0.15) indicated thateight important metabolites were irrelevant to serum AFPlevel, five of them including betaine, propionylcarnitine,proline, methionine, and isoleucyl-L-proline were found tobe greatly different (Student t test, P < 0.001) betweennonmalignant liver diseases and patients having hepatocel-lular carcinoma (Supplementary Table S5). After the binarylogistic regression analysis of the data, betaine and propio-nylcarnitine were selected as the optimal combination

Figure 2. Metabolic network of the significantly changed metabolites. The normalized contents are shown under the chemical name. Blue and redbar chart, normalized content in the DNT group and HCT group, respectively. All the P values were calculated using Student t test. �, P < 0.05;��, P < 0.01; ��, P < 0.001. Nonsignificantly changed metabolites are shown only by name, and the names with a dashed line frame representthe undetected metabolites. PC, phosphorylcholine (PC); PE, o-phosphoethanolamine; GPEA, glycerylphosphorylethanolamine; GPCho,glycerophosphocholine; MTA, 50-methylthioadenosine; SAH, s-adenosylhomocysteine; SAdo, succinyladenosine; SSADH, succinic acidsemialdehyde; GUDCA, glycoursodeoxycholic acid; GCA, glycocholic acid; GDCA, glycodeoxycholic acid; GCDCA, glycochenodeoxycholicacid; GCDCS, glycochenodeoxycholate-3-sulfate; F1,6bisP, fructose-1,6-bisphosphate; G6P, glucose-6-phosphate (G6P).






for diagnosis of hepatocellular carcinoma. As shown inFig. 5A, both betaine and propionylcarnitine were markedlydecreased in the HCC group. According to the data in pati-ents with hepatocellular carcinoma versus nonmalignant

liver diseases, the ROC curve of betaine plus propionylcar-nitine yielded an area under the curve (AUC) of 0.982,whereas the AUC of AFP was only 0.697 (Fig. 5B). Two-thirds of the samples (54 hepatitis, 53 cirrhosis, and 93

FFA C16:0

60

30

0 0DNT HCT DNT

-9D

-9D

-12D -6D

HCT

DNT HCTDNT HCTDNT HCTDNT HCT

DNT HCT DNT HCTDNT HCTDNT HCTDNT HCT DNT HCT DNT

1.0

0.5

0.0

100

3001220

10

0

6

0

200

100

0

50

0HCT

DNT HCT

LCE

ELO

VL5LC

E

90

180 100

50

70

35

0

400300

150

0

10

5

0

C18:1/C18:2 C2/C0 C3/C0 C2+C3/C0

200

0

0

FFA C16:1A

B C D E F

FFA C20:3

FFA C18:0

Rel

ativ

e am

ou

nt

FFA Carnitine BCAA ArAA PAH

FFA C18:1 FFA C18:2 FFA C18:3

Figure 3. Metabolic changes offree fatty acids, carnitines, andamino acids in the tumor tissues.A, FFA metabolism. B, ratio ofFFA C18:2 to FFA C18:1. C, ratiosof carnitines. D, BCAA (leucine,isoleucine, and valine). E,ArAA (phenylalanine andtyrosine). F, activity ofphenylalaninehydroxylase (PAH)according to the ratio of tyrosineto phenylalanine. Light and darkchart, relative amount in the DNTand HCT groups, respectively.�, P < 0.05; ��, P < 0.01;��, P < 0.001.

SFA and MUFA

Bile acids

PUFA

SC-ACAmino acids

Glycolysis metabolites

TCA Cycle metabolites

Phospholipids

Figure 4. Metabolic correlationnetworks of the differentialmetabolites and related pathways.Highly correlated metabolites(|Cij| > 0.6) are connected with aline. Red, upregulated in tumortissue; green, downregulatedin tumor tissue. PUFA,polyunsaturated fatty acid;MUFA, monounsaturated fattyacid; SFA, saturated fatty acid;SC-AC, short-chain acylcarnitine.

Huang et al.





hepatocellular carcinoma) were randomly selected fromeach group as the training set, and the remaining one-thirdof the samples (27 hepatitis, 25 cirrhosis, and 46 hepatocel-lular carcinoma) constituted the test set for validation. Inthe training set, the diagnostic accuracy of the chronic liverdiseases and hepatocellular carcinoma patients achieved91.6% and 96.8%, respectively. Meanwhile, the accuracyrate of the chronic liver diseases and patients with hepato-cellular carcinoma also reached 94.2% and 93.5% in the testset (Fig. 5C). Noticeably, the combination of betaine andpropionylcarnitine had a diagnostic accuracy of 92% in theAFP false-negative (AFP < 20 mg/L) patients with hepato-cellular carcinoma. More valuably, these two metaboliteswere also effective for excluding AFP false-positive (AFP >

200 mg/L) patients without malignancy, which attained adiagnostic accuracy of 100% both for hepatitis and cirrhosis(Fig. 5D). These results showed that the capacity of betaineand propionylcarnitine combination as a novel hepato-cellular carcinoma serum biomarker, especially its supple-mentary role to AFP.

To further externally validate the earlier experimentalresults, another batch of serum samples was enrolled from25 patients with cirrhosis and 22 patients with hepato-cellular carcinoma. Quantitative analysis of betaine andpropionylcarnitine in the external validation serum sam-pleswas carried out on the CE–MS system. As shown inSupplementary Fig. S3A, the levels of betaine and propio-nylcarnitine in the hepatocellular carcinoma group were

70,000

60,000

50,000

40,000

30,000

20,000

10,000

0

******

******

A B

C

D

100%

50%

0%

Betaine

Training Test

Hepatitis Cirrhosis HCC

Hepatitis Cirrhosis HCC

100%100%100%100%94% 97%

92%91%86%

AFP <20 (µg/L) AFP 20~200 (µg/L) AFP>200 (µg/L)

Training Test Training Test

Propionylcarnitine

Rel

ativ

e am

ount

Pos

sibi

lity

100.00%

50.00%

0.00%

Dia

gnos

tic a

ccur

acy

Sen

sitiv

ity

1 - Specificity

Hepatitis1.0

0.8

0.6

0.4

0.2

0.00.0 0.2 0.4 0.6 0.8 1.0

Two metabolites

AFP

CirrhosisHCC

Figure 5. Diagnostic potential forhepatocellular carcinoma bybetaine plus propionylcarnitinefrom serum. A, relative amount inpatients who have hepatitis,cirrhosis, and hepatocellularcarcinoma. B, ROC curves fordistinguishing hepatocellularcarcinoma from nonmalignant liverdiseases by AFP and thecombination of two metabolites,respectively. C, diagnosticaccuracy for nonmalignant liverdiseases and hepatocellularcarcinoma in the training set andtest set by using betaine pluspropionylcarnitine. D, diagnosticaccuracy in differentconcentrations of AFP serumsamples among patients whohave hepatitis, cirrhosis, andhepatocellular carcinoma.��, P < 0.001.






significantly lower than those in the cirrhosis group, whichwere consistent with the previous 298 serum samplesstudy. The diagnostic accuracies of betaine plus propionyl-carnitine for cirrhosis and hepatocellular carcinoma were80.0% and 86.3%, respectively (Supplementary Fig. S3B).

DiscussionCancer metabolism has recently become a subject of con-

siderable interest for both the pharmaceutical industry andclinical research. However, a systematic understanding ofcancer metabolism remains a challenge. The primary goal inthis study is to understand the metabolic features in hepato-cellular carcinoma-related tissues by investigating the meta-bolic differences among the HCT, ANT, and DNT groups usinga nontargeted tissue metabolomics method.

Metabolic features of tumor tissueBased on the current results, we infer that the metabo-

lites of glycolysis in tumor tissues, such as 6-phosphoglu-conic acid, increased while the metabolites in the TCAcycle decreased (Fig. 2). These data revealed the rapidexpenditure of glucose through the glycolytic pathway witha low level of aerobic oxidation through the TCA cycle.Another important feature of the tumor cell can also beobserved from the correlation network (Fig. 4), all of theenriched metabolites in tumor tissues are related to energysupply, such as glycolysis-related metabolites and some AA,SFA, and MFA. These results revealed the high energyrequirements of the tumor cells for proliferation. Glucoseis rapidly consumed in tumor cells through glycolysis.Furthermore, gluconeogenesis and b-oxidation in the tu-mor cells are upregulated for energy supply. In contrast, theTCA cycle is downregulated, and the short-chain andmiddle-chain acylcarnitines are also decreased, whichshows decreased consumption of carboxylic acids in themitochondria. This result is consistent with the Warburgeffect (23, 24). Moreover, the increased consumption ofglucose may upregulate gluconeogenesis from lipids andproteins. Because of the increased catabolism, the quan-tity of AAs and FFAs consequently increases. Therefore,modified metabolism of FFAs and AAs was observed inthe tumor tissues. b-Oxidation of FFA is a multi-step pro-cess in which fatty acids are broken down in various tissuesto produce energy. As shown in Fig. 4, lipids, includingphospholipids and FFA, are the most correlated meta-bolites. First, the tumor cells modified their lipid meta-bolism to satisfy the large energy demand during cellularproliferation, which could be observed in the tumor tis-sue via the increased amounts of SFA and MFA. Further-more, oxidative stress was increased by the effect of lipidperoxidation (25). Therefore, the suppressed functionalenzyme D-12 desaturase (Fig. 3A) might be the mechanismthrough which tumor cells maintain high levels of SFAand MFA. The expression of the desaturase enzyme con-firmed that b-oxidation was accelerated in the hepato-cellular carcinoma tumor tissues compared with noncan-cerous tissues.

Because oxidative damage is the primary driver for cellapoptosis, the level of antioxidative metabolites would beaccordingly elevated. We observed increased levels of somemetabolites that could protect the tumor cells from oxidativedamage in the tumor tissue. The striking elevation of 6-phos-phogluconic acid suggested an increased flux of metabolicfuels into the pentose–phosphate pathway (PPP), perhaps tocreate NADPH to manage the perceived increase in oxidativestress in hepatocellular carcinoma (26). Glutamine is a majornitrogen carrier and a carbon substrate for anabolic processesin cancer cells, and it may protect cancer cells from oxidativestress and apoptosis (27, 28). Similarly, glutathione and glu-tathione disulfide, which are the major redox couple in animalcells (29), were also increased. The increase of glutathionereflected the alteration of the redox state, which is one of thekey performance indicators in pathologic conditions, especial-ly in cancer (30).

Chronic inflammation is an important factor of hepato-carcinogenesis (4), and it could influence tumor lipid meta-bolism. The metabolism of cholesterol and phospholipidsis a primary function of normal liver cells, and the down-regulated synthesis of bile acids may lead to the malabsorp-tion of lipids. The significant increases in taurine and4-hydroxythreonine and decreases in urobilinogen and gly-cine-conjugated bile compounds in hepatocellular carcino-ma suggested that the co-metabolism of gut microbes wasaltered in hepatocellular carcinoma, and potential changesin the enterohepatic circulation may also have occurred(31). Moreover, bile acids are closely correlated with lipids,which may cause potential regulatory effects on nuclearreceptors (32). The abnormal metabolism of phospho-lipids may influence many biological processes, for example,inflammation. PUFA, especially arachidonic acid, is theprecursor for diverse inflammatory molecules (33). Phos-pholipids from the plasma membrane are the primarysource of arachidonic acid. A decreased level of LPC wasobserved in the cancer tissues. However, the levels of PCand PE were significantly greater in the cancer tissues(Fig. 2). PC could be converted into LPC and arachidonicacid by the catalysis of phospholipase A2 (PLA2). To escapepossible damage from the immune system, the tumor cellmay reduce the generation of endogenic inflammationmolecules by downregulating the related enzymes, such asPLA2 (34). According to the current results, it can be hypo-thesized that levels of PLA2 or D-12 desaturase are alteredin tumor cells in an attempt to modify the metabolism oflipids, thereby supporting the proliferation of tumor cellsas well as their escape from immune attack and apoptoticcell death.

In addition to energy metabolism, AAs also serve a varietyof biological functions. For example, several N-acetyl AAs(Ala, Met, and Trp) consistent with the observed increaseof acetylcarnitine were elevated in hepatocellular carcinoma(Fig. 3C). These observations may suggest elevated acetylcoenzyme A concentrations in hepatocellular carcinoma(35). Furthermore, increased levels of BCAAs and ArAAswere observed in the tumor tissues (Fig. 3D and E). Theincrease of BCAAs occurred in concert with increases in

Huang et al.





ArAAs (Trp, Tyr, and Phe), which implied that the enhance-ment of large neutral amino acids (LNAA) was facilitatedby the L-type AA transporter-1 (LAT1; ref. 36).The ratio between Phe and Tyr can be used as an indicator

of phenylalanine hydroxylase (PAH; ref. 37). The decreaseof the Tyr/Phe ratio in the HCT group (Fig. 3F) suggests theinhibition of PAH in tumor tissues. There were also someother metabolic features of hepatocellular carcinoma thatreflected the upregulation of cellular proliferation. Forexample, the level of hypoxanthine increased in the HCTgroup, whereas that of xanthine declined sharply, whichmay suggest decreased activity of xanthine dehydrogenase/oxidase (38) and active cellular proliferation.As indicated by the current results, the rapid expenditure

of energy causes the cancer cells to modify the metabolicpathway to provide a sufficient amount of energy. Moreover,the cancer cells also try to avoid oxidative damage andchronic inflammation, which may reduce the occurrence ofcellular apoptosis.

Diagnostic potentials of differential metabolitesThe metabolic changes observed in biofluids might be

caused by differences among individuals, including diet, sex,age, and so on. In this study, the significant changes in meta-bolites were defined by a pairwise comparison of differentparts of the liver tissue from individuals, and the differences ofother non-disease factors were removed. The relationshipof differential metabolites with the clinical information ofsamples, including HBsAg, tumor diameter, as well as AFP,ALT, and AST levels, was investigated carefully (Supple-mentary Table S4). Thirty-eight patients with hepatocellularcarcinoma were HBsAg-positive, and 12 patients were HBsAg-negative (Table 1). Citramalic acid in tumor tissue was signi-ficantly decreased in patients who were HBsAg positive com-pared with those who were HBsAg negative (SupplementaryFig. S4). This finding may explain the suppression of TCAcycle under inflammatory status.The differential metabolites found in tissues can be used

as candidates biomarkers to investigate their diagnostic po-tential using serum as the sample. According to the resultsfrom Pearson correlation analysis, betaine and propionyl-carnitine were found to be irrelevant to serum AFP level;they were defined to investigate the potential for diagnosinghepatocellular carcinoma (Fig. 5). The result indicated thatbetaine plus propionylcarnitine was efficient for distin-guishing hepatocellular carcinoma from nonmalignant liverdiseases with a 0.982 AUC value of the ROC curve, which wasmuch better than that of AFP (0.697; Fig. 5B). The changes ofbetaine and propionylcarnitine in the external validationhad the same change trend as the first batch of serum samples.In addition, good validation results further confirmed thatthe combination of these two metabolites was effective forhepatocellular carcinoma diagnosis.

ConclusionsIn summary, the metabolic profiling analysis of liver

tissues provided a holistic view of the metabolic features

of hepatocellular carcinoma. The differential metabolitesin the tumor tissues were filtered and identified, and theresults showed that the metabolism in the tumor wasmodified to promote cellular proliferation or escape fromapoptosis. The rapid consumption of energy by the tumorcells upregulated glycolysis and downregulated the TCAcycle, both of which are consistent with the Warburgeffect. Gluconeogenesis and b-oxidation are also upregu-lated for energy supply, which could be observed from theenriched quantities of AAs and fatty acids. Based on thecorrelation network, modified metabolism of lipids is themost important feature of the hepatocellular carci-noma tumor, which may be important to protect the tumorcell from oxidative damage. This protective effect includedincreased levels of antioxidative metabolites and decreas-ed levels of inflammation-related metabolites of PUFA,which is supposed to be related to the upstream enzymes,such as PLA2. This mechanism may be important forthe tumor cell to break the balance of proliferation andapoptosis.

The diagnostic potential of the differential metabolitesfound in tissues, further studied in serum samples, andvalidated in another group of serum samples showed thatthe combination of betaine and propionylcarnitine has abetter ROC curve than AFP; it is useful for both AFP false-positive and false-negative patients in distinguishing hepa-tocellular carcinoma from hepatitis and cirrhosis. These twometabolites have potential as hepatocellular carcinomabiomarkers in clinical diagnosis.

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

Authors' ContributionsConception and design: Q. Huang, Y-X. Tan, P. Yin, G. XuDevelopment of methodology: Q. HuangAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): Y-X. TanAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): Q. Huang, P. Yin, P. Gao, X. Lu, G. XuWriting, review, and/or revision of the manuscript: Q. Huang, Y-X. Tan,P. Yin, X. Lu, G. XuAdministrative, technical, or material support (i.e., reporting or orga-nizing data, constructing databases): Y-X. Tan, G. YeStudy supervision: Y-X. Tan, G. XuOther (sample treatment): G. YeOther: H-Y. Wang

AcknowledgmentsThe authors thank Jun Zeng for the measurement of metabolites by using

CE-MS.

Grant SupportThe study has been financially supported by the State Key Science &

Technology Project for Infectious Diseases (grant nos. 2012ZX10002-009 and2008ZX10002-019), the key foundation (grant no. 21175132), and the creativeresearch group project (grant nos. 81221061 and 21021004) from the NationalNatural Science Foundation of China.

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.

Received January 31, 2013; revised May 27, 2013; accepted June 7, 2013;published OnlineFirst July 1, 2013.






References1. Pineau P, Tiollais P. Hepatitis B vaccination: a major player in the

control of primary liver cancer. Pathol Biol (Paris) 2010;58:444–53.2. West J, Wood H, Logan RF, Quinn M, Aithal GP. Trends in the

incidence of primary liver and biliary tract cancers in England andWales 1971–2001. Br J Cancer 2006;94:1751–8.

3. ChristofkHR, VanderHeidenMG,HarrisMH,RamanathanA,GersztenRE, Wei R, et al. The M2 splice isoform of pyruvate kinase is importantfor cancer metabolism and tumour growth. Nature 2008;452:230–3.

4. Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation.Cell 2011;144:646–74.

5. Lazebnik Y. What are the hallmarks of cancer? Nat Rev Cancer2010;10:232–3.

6. Nicholson JK, Lindon JC, Holmes E. 'Metabonomics': understandingthe metabolic responses of living systems to pathophysiologicalstimuli via multivariate statistical analysis of biological NMR spectro-scopic data. Xenobiotica 1999;29:1181–9.

7. Yin PY, Wan DF, Zhao CX, Chen J, Zhao XJ, Wang WZ, et al. Ametabonomic study of hepatitis B-induced liver cirrhosis and hepa-tocellular carcinoma by using RP-LC and HILIC coupled with massspectrometry. Mol Biosyst 2009;5:868–76.

8. Shariff MI, Gomaa AI, Cox IJ, Patel M, Williams HR, CrosseyMM, et al.Urinary metabolic biomarkers of hepatocellular carcinoma in an Egyp-tian population: a validation study. J Proteome Res 2011;10:1828–36.

9. Soga T, Sugimoto M, HonmaM, Mori M, Igarashi K, Kashikura K, et al.Serum metabolomics reveals gamma-glutamyl dipeptides as biomar-kers for discrimination amongdifferent formsof liver disease. JHepatol2011;55:896–905.

10. Griffin JL, Nicholls AW. Metabolomics as a functional genomic tool forunderstanding lipid dysfunction in diabetes, obesity and related dis-orders. Pharmacogenomics 2006;7:1095–107.

11. Yang YX, Li CL, Nie X, Feng XS, Chen WX, Yue Y, et al. Metabonomicstudies of human hepatocellular carcinoma using high-resolutionmagic-angle spinning H-1 NMR spectroscopy in conjunction withmultivariate data analysis. J Proteome Res 2007;6:2605–14.

12. Sreekumar A, Poisson LM, Rajendiran TM, Khan AP, CaoQ, Yu J, et al.Metabolomic profiles delineate potential role for sarcosine in prostatecancer progression. Nature 2009;457:910–4.

13. Hirayama A, Kami K, Sugimoto M, Sugawara M, Toki N, Onozuka H,et al. Quantitative metabolome profiling of colon and stomach cancermicroenvironment by capillary electrophoresis time-of-flight massspectrometry. Cancer Res 2009;69:4918–25.

14. Tan YX, Yin PY, Tang L, Xing WB, Huang Q, Cao D, et al. Meta-bolomics study of stepwise hepatocarcinogenesis from the modelrats to patients: potential biomarkers effective for small hepatocel-lular carcinoma diagnosis. Mol Cell Proteomics 2012;11:M111010694.

15. Chen J, Zhang XY, Cao R, Lu X, Zhao SM, Fekete A, et al. Serum 27-nor-5 beta-cholestane-3,7,12,24,25 pentol glucuronide discovered bymetabolomics as potential diagnostic biomarker for epithelium ovariancancer. J Proteome Res 2011;10:2625–32.

16. Yin PY, Zhao XJ, Li QR, Wang JS, Li JS, Xu GW. Metabonomics studyof intestinal fistulas based on ultraperformance liquid chromatographycoupled with Q-TOF mass spectrometry (UPLC/Q-TOF MS). J Prote-ome Res 2006;5:2135–43.

17. Smilde AK, van derWerfMJ, BijlsmaS, van derWerff-van-der Vat BJC,Jellema RH. Fusion of mass spectrometry-based metabolomics data.Anal Chem 2005;77:6729–36.

18. Huang Q, Yin PY, Wang J, Chen J, Kong HW, Lu X, et al. Method forliver tissue metabolic profiling study and its application in type 2diabetic rats basedon ultra performance liquid chromatography–massspectrometry. J Chromatogr B 2011;879:961–7.

19. Chen J, Wang WZ, Lv S, Yin PY, Zhao XJ, Lu X, et al. Metabonomicsstudy of liver cancer based on ultra performance liquid chromatogra-

phy coupled to mass spectrometry with HILIC and RPLC separations.Anal Chim Acta 2009;650:3–9.

20. Chen J, ZhaoXj, Fritsche J, Yin PY, Schmitt-Kopplin P,WangWZ, et al.Practical approach for the identification and isomer elucidation ofbiomarkers detected in a metabonomic study for the discovery ofindividuals at risk for diabetes by integrating the chromatographic andmass spectrometric information. Anal Chem 2008;80:1280–9.

21. Go EP. Database resources in metabolomics: an overview. J Neu-roimmune Pharm 2010;5:18–30.

22. NewgardCB,An J,Bain JR,MuehlbauerMJ, StevensRD, LienLF, et al.A branched-chain amino acid-related metabolic signature that differ-entiates obese and lean humans and contributes to insulin resistance.Cell Metabolism 2009;9:311–26.

23. Warburg O. On respiratory impairment in cancer cells. Science1956;124:269–70.

24. Bi XZ, Lin QS, Foo TW, Joshi S, You T, Shen HM, et al. Proteomicanalysis of colorectal cancer reveals alterations in metabolic path-ways: mechanism of tumorigenesis. Mol Cell Proteomics 2006;5:1119–30.

25. Puskas LG, Feher LZ, Vizler C, Ayaydin F, Raso E, Molnar E, et al.Polyunsaturated fatty acids synergize with lipid droplet binding tha-lidomide analogs to induce oxidative stress in cancer cells. LipidsHealth Dis 2010;9:56.

26. Jeon SM, Chandel NS, Hay N. AMPK regulates NADPH homeostasisto promote tumour cell survival during energy stress. Nature 2012;485:661–5.

27. DeBerardinis RJ, Cheng T. Q's next: the diverse functions ofglutamine in metabolism, cell biology and cancer. Oncogene 2010;29:313–24.

28. Gao P, Tchernyshyov I, Chang TC, Lee YS, Kita K, Ochi T, et al.c-Myc suppression of miR-23a/b enhances mitochondrial gluta-minase expression and glutamine metabolism. Nature 2009;458:762–5.

29. Wu GY, Fang YZ, Yang S, Lupton JR, Turner ND. Glutathione metab-olism and its implications for health. J Nutr 2004;134:489–92.

30. Chaiswing L, Oberley TD. Extracellular/microenvironmental redoxstate. Antioxid Redox Signal 2010;13:449–65.

31. Csanaky IL, Aleksunes LM, Tanaka Y, Klaassen CD. Role of hepatictransporters in prevention of bile acid toxicity after partial hepatectomyin mice. Am J Physiol-gastr L 2009;297:G419–G33.

32. Li TG, Chiang JYL. Regulation of bile acid and cholesterol metabolismby PPARs. PPAR Res 2009;2009:501739.

33. Han C, Bowen WC, Li GY, Demetris AJ, Michalopoulos GK, Wu T.Cytosolic phospholipase A2alpha and peroxisome prolife-rator-activated receptor gamma signaling pathway counter-acts transforming growth factor beta-mediated inhibition ofprimary and transformed hepatocyte growth. Hepatology 2010;52:644–55.

34. Taketo MM, Sonoshita M. Phospolipase A2 and apoptosis. BiochimBiophys Acta 2002;1585:72–6.

35. Fong MY, McDunn J, Kakar SS. Identification of metabolites in thenormal ovary and their transformation in primary and metastaticovarian cancer. PLoS ONE 2011;6:e19963.

36. Vogel KR, Arning E, Wasek BL, Bottiglieri T, Gibson KM. Non-phys-iological amino acid (NPAA) therapy targeting brain phenylalaninereduction: pilot studies in PAH (ENU2) mice. J Inherit Metab Dis2013;36:513–23.

37. Matthews DE. An overview of phenylalanine and tyrosine kinetics inhumans. J Nutr 2007;137:1549S–55S; discussion 73S–75S.

38. LinderN,Martelin E, LundinM, Louhimo J,NordlingS,HaglundC, et al.Xanthine oxidoreductase – Clinical significance in colorectal cancerand in vitro expression of the protein in human colon cancer cells. Eur JCancer 2009;45:648–55.

Huang et al.





2013;73:4992-5002. Published OnlineFirst July 1, 2013.Cancer Res Qiang Huang, Yexiong Tan, Peiyuan Yin, et al. Nontargeted Tissue MetabolomicsMetabolic Characterization of Hepatocellular Carcinoma Using

Updated version

10.1158/0008-5472.CAN-13-0308doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerres.aacrjournals.org/content/suppl/2013/07/09/0008-5472.CAN-13-0308.DC1

Access the most recent supplemental material at:

Cited articles

http://cancerres.aacrjournals.org/content/73/16/4992.full#ref-list-1

This article cites 37 articles, 4 of which you can access for free at:

Citing articles

http://cancerres.aacrjournals.org/content/73/16/4992.full#related-urls

This article has been cited by 8 HighWire-hosted articles. Access the articles 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/73/16/4992To 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-13-0308

http://cancerres.aacrjournals.org/content/suppl/2013/07/09/0008-5472.CAN-13-0308.DC1

http://cancerres.aacrjournals.org/content/73/16/4992.full#ref-list-1

http://cancerres.aacrjournals.org/content/73/16/4992.full#related-urls

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

mailto:[email protected]

http://cancerres.aacrjournals.org/content/73/16/4992


metabolic characterization of hepatocellular...

Documents