Cetuximab Reverses the Warburg Effect by Inhibiting HIF-1 ... · Cells were cultured in 24-well plates with low-glucose (1 g/L), low-serum (0.5% FBS) medium (0.5 mL/well) at 37 C

Cancer Therapeutics Insights

Cetuximab Reverses the Warburg Effect by InhibitingHIF-1–Regulated LDH-A

Haiquan Lu1,2, Xinqun Li1, Zhongguang Luo1,3, Jie Liu3, and Zhen Fan1,2

AbstractHypoxia-inducible factor-1 (HIF-1) plays a critical role in reprogramming cancermetabolism toward aerobic

glycolysis (i.e., the Warburg effect), which is critical to supplying cancer cells with the biomass needed for

proliferation. Previous studies have shown that cetuximab, an EGF receptor–blocking monoclonal antibody,

downregulates the alpha subunit of HIF-1 (HIF-1a) through the inhibition of EGF receptor downstream cell

signaling and that downregulation of HIF-1a is required for cetuximab-induced antiproliferative effects.

However, themechanism underlying these actions has yet to be identified. In this study, we used the Seahorse

XF96 extracellular flux analyzer to assess the effect of cetuximab treatment on changes in glycolysis and

mitochondrial respiration, the two major energy-producing pathways, in live cells. We found that cetuximab

downregulated lactate dehydrogenase A (LDH-A) and inhibited glycolysis in cetuximab-sensitive head and

neck squamous cell carcinoma (HNSCC) cells in anHIF-1a downregulation–dependent manner. HNSCC cells

with acquired cetuximab resistance expressed a high level of HIF-1a and were highly glycolytic. Over-

expression of aHIF-1amutant (HIF-1a/DODD) conferred resistance to cetuximab-inducedG1 phase cell-cycle

arrest, which could be overcome by knockdown of LDH-A expression. Inhibition of LDH-A activity with

oxamate enhanced the response of cetuximab-resistant cells to cetuximab. Cetuximab had no noticeable

inhibitory effect on glycolysis in nontransformed cells. These findings provide novel mechanistic insights into

cetuximab-induced cell-cycle arrest from the perspective of cancermetabolism and suggest novel strategies for

enhancing cetuximab response. Mol Cancer Ther; 12(10); 2187–99. �2013 AACR.

IntroductionGlucose is an important source of energy and carbon for

both normal and cancer cells. Unlike most normal cells—which metabolize glucose by a low rate of glycolysisfollowed by oxidative phosphorylation in the mitochon-dria through the tricarboxylic acid cycle (also known asthe citric acid cycle or Krebs cycle)—cancer cells metab-olize glucose by a high rate of glycolysis followed bylactate production in the cytosol even in the presence ofabundant oxygen, a phenomenon known as aerobic gly-colysis or the Warburg effect (1, 2). The Warburg effect isimportant for cancer cell proliferation because this pro-cess generates building blocks and reducing power, bothof which are needed for the biosynthesis that fuels cell

growth and proliferation (3, 4). This alteredmetabolism incancer cells is a direct result of the aberrant cell signalingcaused by overexpression of growth factor receptors,activation of oncogenes, and/or inactivation of tumorsuppressor genes that permits unlimited cancer cell pro-liferation (5–7).

The transcription factor hypoxia-inducible factor-1(HIF-1) plays a key role in reprogramming cell metabo-lism from oxidative phosphorylation to aerobic glycoly-sis. HIF-1 regulates the expression of the genes coding forproteins involved in various steps of cancer metabolism,from glucose uptake and subsequent glycolytic reactionsto the generation of lactate and its secretion by lactatetransporters (8). HIF-1 is a heterodimer consisting of ahighly regulatedHIF-1a subunit and a constitutively exp-ressed HIF-1b subunit (9–12). A high level of HIF-1aprotein is common in many types of solid tumors includ-ing tumors of the colon, lung, breast, stomach, ovary,pancreas, prostate, kidney, and head and neck (13–15).The high level of HIF-1a in cancer cells is caused not onlyby the decreased ubiquitination and degradation of HIF-1a protein via a posttranslational mechanism associatedwith tumor hypoxia (16, 17) but also by aberrant cellsignaling, which increases HIF-1a protein expression viaa translational mechanism (18–22).

Cetuximab is an EGF receptor (EGFR)-blocking mono-clonal antibody approved for treating patients with headand neck cancers and colorectal cancers in combination

Authors' Affiliations: 1Department of Experimental Therapeutics, TheUniversity of Texas MD Anderson Cancer Center; 2Graduate School ofBiomedical Sciences, The University of Texas Health Sciences Center atHouston, Houston, Texas; and 3Department of Digestive Diseases, Hua-shan Hospital, Fudan University, Shanghai, China

Note: Supplementary data for this article are available at Molecular CancerTherapeutics Online (http://mct.aacrjournals.org/).

Corresponding Author: Zhen Fan, Department of Experimental Thera-peutics, Unit 1950, The University of Texas MD Anderson Cancer Center,1515 Holcombe Boulevard, Houston, TX 77030. Phone: 713-745-3560;Fax: 713-745-3562; E-mail: [email protected].

doi: 10.1158/1535-7163.MCT-12-1245

�2013 American Association for Cancer Research.

MolecularCancer

Therapeutics

www.aacrjournals.org 2187

on February 10, 2021. © 2013 American Association for Cancer Research. mct.aacrjournals.org Downloaded from

Published OnlineFirst August 6, 2013; DOI: 10.1158/1535-7163.MCT-12-1245

http://mct.aacrjournals.org/

with radiotherapy and/or chemotherapy (23, 24).We andothers have previously shown that cetuximab binds toEGFR and blocks the ligand-induced activation of EGFRdownstream cell signaling, which leads to G1 phase arrestof cell-cycle traversal and even apoptosis in certain cir-cumstances (25–40). Our previous work showed thatcetuximab candownregulateHIF-1aprotein by inhibitingthe PI3K/Akt and MEK/Erk pathways, and this down-regulation of HIF-1a is required, although may not besufficient, for cetuximab to induce antiproliferativeeffects (41–45). Knockdown of HIF-1a by siRNA partiallyovercame the resistance caused by overexpression ofconstitutively active Ras mutant to cetuximab-inducedantiproliferative effects (43–45).

These previous studies established the importance ofHIF-1a downregulation in mediating cetuximab-inducedantitumor effects; however, to our knowledge, no studieshave carefully examined the mechanism that leads togrowth inhibition after downregulation of HIF-1a bycetuximab. We hypothesized that cetuximab inhibits can-cer cell proliferation through inhibition of glycolysis bydownregulating HIF-1a, thereby reversing the Warburgeffect that is critically important for cancer cell survival andproliferation. To test this hypothesis, we generated andcharacterized 2 pairs of genetically matched cetuximab-sensitive and -resistant head and neck cancer cell lines.Weused theSeahorseXF96extracellularfluxanalyzer toassessthe effect of cetuximab treatment on glycolysis and mito-chondrial respiration, the 2 major energy-producing path-ways, in live cells. We further investigated the effect ofcetuximab on the expression and enzymatic activity oflactate dehydrogenase-A (LDH-A), which regulates theconversion of pyruvate to lactate, and on the levels ofglucose consumption, lactate production, and intracellularATP in cetuximab-sensitive and -resistant cells. Our find-ings provide novel insights into the mechanisms underly-ing cetuximab-induced antiproliferative and apoptotic eff-ects in cancer cells and suggest a novel therapeutic strategyfor improving cetuximab response.

Materials and MethodsReagents

Cetuximab was provided by ImClone Systems, an EliLilly company. All other reagents, including oxamate,were purchased from Sigma-Aldrich unless otherwisespecified.

Cell lines and cell culturesHN5, FaDu, SqCC/Y1, and TU167 head and neck squa-

mous cell carcinoma (HNSCC) cells were maintained inDulbecco’s modified Eagle’s medium (DMEM)/F12medium supplemented with 10% FBS, 2 mmol/L gluta-mine, 100 U/mL penicillin, and 100 mg/mL streptomycinin a 5% CO2 atmosphere at 37�C as described previously(45, 46). Cetuximab-resistant HN5-R and FaDu-R cellswere generated by exposing parental HN5 and FaDu cellsto stepwise dose increases of cetuximab (up to 20 nmol/L)for 6 months or more. HN5-HIF-1a/DODD and FaDu-

HIF-1a/DODD cells were established by transfectingparental HN5 and FaDu cells with pcDNA3.1 constructcontaining an HIF-1a oxygen–dependent degradation(ODD) domain deletion mutant (referred to as HIF-1a/DODD) using Lipofectamine 2000 (Life Technologies).The HN5-HIF-1a/DODD and FaDu-HIF-1a/DODD cellswere maintained in medium containing 500 mg/mL neo-mycin. Immortalized nontransformed NOM9-TK humanhead and neck epithelial cells were maintained in serum-free keratinocyte basal growth medium supplementedwith components of a keratinocyte growth mediumSingleQuots kit including bovine pituitary extract, recom-binant human EGF, insulin, hydrocortisone, and genta-micin sulfate (Lonza, Inc.; refs. 46, 47). The authenticity ofthe cell lines used in the current study has not beenconfirmed during the past 6 months.

Cell proliferation assayCells were cultured in 24-well plates with low-glucose

(1 g/L), low-serum (0.5% FBS) medium (0.5 mL/well) at37�C. Following the indicated treatments, 10 mg/mLMTT was added (50 mL/well), and the cells were incu-bated for an additional 2 hours. The cells were then lysedwith a lysis buffer (500 mL/well) containing 20% SDS indimethyl formamide/H2O (1:1, v/v; pH 4.7) at 37�C forat least 6 hours. The relative number of surviving cellsin each group was determined by measuring the opticaldensity (OD) of the cell lysates at an absorbance wave-length of 570 nm. The OD value of each treatment groupwas expressed as a percentage of the OD value of theuntreated control cells.

Knockdown of gene expression by siRNALDH-A–targeted siRNA (target DNA sequence #1,

GGAGAAAGCCGTCTTAATT; #2, GATTAAGGGTCTT-TACGGA; #3, CAGATTTAGGGACTGATAA), HIF-1a–targeted siRNA (target DNA sequence #1, CAAAGTT-CACCTGAGCCTA; #2, GATTAACTCAGTTTGAACT),and control siRNA were purchased from Sigma-Aldrich.The siRNA (200 pmol) and Lipofectamine 2000 (5 mL)weremixed in 100 mL ofminimal essential medium (Opti-MEM, Life Technologies) for 15 minutes, and the siRNA/Lipofectamine 2000 mixture was added into the culturemedium for 6 hours. After transfection, the medium wasreplaced with regular medium, and the cells were cul-tured for an additional 48 hours before the detection ofknockdown of LDH-A andHIF-1a expression byWesternblotting.

Western blot analysisAfter the indicated treatments, cellswere lysed in a lysis

buffer containing 50 mmol/L Tris-HCl (pH 7.4), 150mmol/L NaCl, 0.5% NP-40, 50 mmol/L NaF, 1 mmol/LNa3VO4, 1 mmol/L phenylmethylsulfonyl fluoride, 25mg/mL aprotinin, and 25 mg/mL leupeptin and kept onice for 15 minutes. The lysates were cleared by centrifu-gation, and the supernatants were collected. Equal amo-unts of protein lysate, as determined using the Pierce

Lu et al.

Mol Cancer Ther; 12(10) October 2013 Molecular Cancer Therapeutics2188




Coomassie Plus colorimetric protein assay (Thermo Fish-er Scientific), were separated by SDS-PAGE, blotted ontonitrocellulose, and probed with primary antibodiesagainst HIF-1a (BD Biosciences); Akt, S473-phosphory-lated Akt, T202/Y204-phosphorylated Erk, Y1068-phos-phorylated EGFR, and LDH-A (Cell Signaling Technolo-gy); MAPK (Santa Cruz Biotechnology); and EGFR andb-actin (Sigma-Aldrich). The signals were visualizedusing the Enhanced Chemiluminescence Detection Kit(GEHealthcare). Western blotting results were quantifiedusing the ImageJ v1.46 image processing program (NIH,Bethesda, MD).

Metabolic flux assayThe bioenergetic flux of cells in response to cetuximab

orHIF-1a silencingwas assessed using the Seahorse XF96extracellular flux analyzer (Seahorse Biosciences). Forcetuximab treatment, the indicated cells were plated at2 � 104 cells per well in XF96 plates and incubated withDMEM/F12 regular cell culture medium in a 5% CO2

atmosphere at 37�C for 24 hours. The medium was thenchanged to low-glucose, low-serummedium, and the cellswere incubated for another 24 hours and then treatedwith20 nmol/L cetuximab or left untreated for 10 hours. Themedium was then replaced with unbuffered DMEM XFassay medium (pH adjusted to 7.4 using 1 N sodiumhydroxide) supplemented with 2 mmol/L glutamine and1 g/L glucose with or without 20 nmol/L cetuximab. ForHIF-1 silencing, the indicated cells were treatedwithHIF-1a siRNA or control siRNA for 48 hours. The cells werethen plated at 3 � 104 cells per well in XF96 plates andincubated with DMEM/F12 regular cell culture mediumin a 5%CO2 atmosphere at 37�C for another 24 hours. Themedium was then replaced with unbuffered DMEM XFassay medium as described above. After media changes,the cells were placed in a 37�C CO2-free incubator for 1hour. The basal oxygen consumption rate (OCR) andextracellular acidification rate (ECAR) were then deter-mined using the XF96 plate reader with the standardprogram as recommended by the manufacturer.

Glucose consumption assayCellswere seeded in6-well plates at 5� 105 cells perwell

in 3 mL of phenol red–free, low-glucose, low-serum cellculturemediumasdescribed above.At the indicated timesafter treatment, an aliquot of 50 mL of the conditionedmediumwas collected fromeachwell anddilutedwith 950mL of distilled water (1:20). The glucose concentration inthedilutedmediumwasmeasuredusing theGlucose (GO)AssayKit (Sigma-Aldrich)according to themanufacturer’sinstructions. Briefly, samples weremixedwith the provid-ed glucose assay reagent and incubated at 37�C for 30minutes before OD was measured at 540 nm. The mea-sured ODs of the samples were compared with that of thestandard glucose control. Glucose consumption was cal-culated by subtracting the concentration of glucoseremaining in the medium at the indicated time from theconcentration of glucose present in fresh cell culturemedi-

um. Glucose consumption was expressed as mg/106 cellsat the indicated times.

Lactate production assayCells were seeded in 6-well plates at 5 � 105 cells per

well in 3mLof serum-free cell culturemedium containing1 g/L glucose. At the indicated times after treatment, analiquot of 50 mL of the conditioned mediumwas collectedfrom each well and diluted with 950 mL of distilled water(1:20). The lactate concentration in the diluted mediumwasmeasuredusing theLactateAssayKit fromBioVision,Inc. Briefly, the sample was mixed with the providedreaction reagent and incubated at 37�C for 30 minutesbefore the OD was measured at 570 nm. The measuredODs of the samples were compared with that of thestandard lactate control. The levels of lactate wereexpressed as nmol/106 cells at the indicated times.

LDH-A activity assayLDH-A activity was measured using a colorimetric

LDHActivity AssayKit (BioVision, Inc.), whose detectionof LDH-A activity is based on a reaction in which LDH-Aconverts NAD to NADH in a specific time. Briefly, aftervarious treatments, cell pellets were homogenized on icein 0.5mLof cold assay buffer. Supernatantswere collectedby centrifugation. Protein concentrations were deter-mined using the Pierce Coomassie Plus colorimetric pro-tein assay. Fifty-microliter aliquots (volume adjustedusing assay buffer) containing equal amounts of proteinand 50 mL of reaction reagent (48 mL of assay buffer and 2mL of LDH-A substrate mix solution) were added to eachwell of a 96-well plate, and the plates were read under amicroreader at 450 nm immediately (T0) and 10 minutesafter the reaction (T10). LDH-A activity was expressed asan increase in OD values (T10 � T0). The relative LDH-Aactivities in treated and untreated cells were expressed aspercentages of the LDH-A activity in untreated cells after4 hours in culture, which was arbitrarily set as 100.

Intracellular ATP assayIntracellular ATP levels were determined using the

ATP Bioluminescent Assay Kit (Sigma-Aldrich). Briefly,cells were seeded in 6-well plates at 5 � 105 cells per welland treatedwith 20 nmol/L cetuximab or left untreated inlow-glucose, low-serum medium for 4 hours. The cellswere then harvested and resuspended in 1mL of PBS. Analiquot of 50 mL of the cell suspensionwasmixedwith 100mLofATP-releasing reagent and 50mLof distilledwater ineach well of a 96-well plate. The samples (100 mL) in eachwell were then transferred to awhite opaque 96-well platewhosewellswere eachpre-filledwith 100mLofATP assaymix. The amount of light emitted in each well was imme-diatelymeasured using a FLUOstar Omega luminometer.

Flow cytometric analysisAfter the indicated treatments, cells were harvested by

trypsin and fixed in 70% ethanol overnight. After centri-fugationandwashingwithPBS, the cellswere stainedwith

Cetuximab Inhibits Aerobic Glycolysis

www.aacrjournals.org Mol Cancer Ther; 12(10) October 2013 2189




propidium iodide (50 mg/mL) andRNaseA (20mg/mL) onice for 30 minutes and then subjected to flow cytometricanalysis with an LSRFortessa cell analyzer (BD Bios-ciences). Cell-cycle distribution data were analyzed usingthe FlowJo v10 data processing program.

Apoptosis assaysApoptosiswasmeasuredusing a colorimetric ELISAkit

(Cell Death Detection ELISA, Roche Diagnostics Corp.)that quantitatively measures cytoplasmic histone-associ-ated DNA fragments (mononucleosomes and oligonu-cleosomes). Theprocedurewas conducted exactly accord-ing to themanufacturer’s instructions.Apoptosiswas alsodetected byWestern blotting with an antibody that recog-nizes both cleaved and uncleaved PARP (Cell SignalingTechnology) as described previously (43, 44).

Statistical analysisThe Student t test was used for all statistical analyses.

The data are presented as means � SD. All experimentswere repeated at least once with reproducible findings.

ResultsCetuximab inhibits aerobic glycolysis, and resistanceto cetuximab is linked to increased glycolytic flux

To assess the effect of cetuximab treatment on cancercell metabolism and determine whether this effect cor-relates with cetuximab-induced antiproliferative effectsin cancer cells, we generated 2 pairs of isogenic cetux-imab-sensitive and cetuximab-resistant HNSCC celllines by subjecting cetuximab-sensitive HN5 and FaDucells to stepwise dose increases of cetuximab (up to 20nmol/L) in continuous culture for 6 months or more.The resultant sublines, HN5-R and FaDu-R, were con-siderably less sensitive to cetuximab-induced growthinhibition than the parental HN5 and FaDu cells treatedunder identical experimental conditions (Fig. 1A). HN5cells had higher expression levels of EGFR, phosphor-ylated EGFR (represented by phosphorylation of EGFRat Y1068), and HIF-1a than FaDu cells in normoxicculture. Both HN5 and FaDu cells had lower levels ofHIF-1a after cetuximab treatment (Fig. 1B). In contrast,HN5-R and FaDu-R cells had higher levels of HIF-1athan their parental cells in normoxic culture, and thelevels of HIF-1a in HN5-R and FaDu-R cells were onlyminimally affected by cetuximab treatment. The levelof LDH-A was decreased by cetuximab treatment inparental HN5 and FaDu cells but not in cetuximab-resistant HN5-R and FaDu-R cells. Interestingly,HN5-R and FaDu-R cells remained sensitive or partiallysensitive to cetuximab-induced inhibition of cell signal-ing, as revealed by the inhibition of EGFR autopho-sphorylation (in FaDu-R cells, but not apparent in HN5-R cells) and reduced phosphorylation of Akt-S473 andErk T202/Y204 (with slight difference in the extentsbetween FaDu-R and HN5-R cells), 2 important signal-ing molecules downstream of EGFR (Fig. 1B). Theseinteresting findings challenge the conventional concept

that cetuximab resistance is linked to the failure forcetuximab to effectively inhibit EGFR downstream cellsignaling and suggest that the resistance could occurthrough changes in the pattern of cell metabolism with-out significant changes in cell signaling pathways.

We further analyzed themetabolic profiles of the paren-tal cell lines and the cetuximab-resistant sublineswith andwithout cetuximab treatment using the Seahorse XF96extracellular flux analyzer, which simultaneously recordsthe metabolic patterns of aerobic respiration and glyco-lysis, the 2 major energy-yielding pathways in cells.Compared with the parental HN5 and FaDu cells, thecetuximab-resistant HN5-R and FaDu-R cells had higherECARs, indicating that these cells had higher glycolysisrates than the parental cells (Fig. 1C). Cetuximab treat-ment lowered the ECAR and increased the OCR (anindicator of mitochondrial respiration rate) in HN5 andFaDu cells but not in HN5-R and FaDu-R cells. Mitochon-drial stress challenge with oligomycin, an ATP synthaseinhibitor, decreased the OCR in parental cells (althoughmodestly) but not in the cetuximab-resistant cells, sug-gesting that the cetuximab-resistant cells were more gly-colytic. Challenge with carbonyl cyanide p-trifluoro-methoxyphenylhydrazone (FCCP), an uncoupler of oxi-dative phosphorylation and ATP synthesis, resulted inthe cetuximab-resistant cell lines having a higher capa-city than the parental cells to increase ECAR, which wasconsistent with the finding that HIF-1a and LDH-Alevels in the cetuximab-resistant sublines were higherthan those in the parental cells.

To determine a direct role of high level of HIF-1a inpromoting aerobic glycolysis via regulating LDH-A inHN5-R and FaDu-R cells, we knocked down HIF-1awith2 different HIF-1a siRNA in parental and cetuximab-resistant cells and confirmed that the level of LDH-A inthe cetuximab-resistant cells was markedly lowered afterHIF-1a knockdown as did the level of LDH-A in parentalcells after HIF-1a knockdown (Fig. 1D). Knockdown ofHIF-1a in HN5-R and FaDu-R cells with the same 2different HIF-1a siRNA significantly decreased ECARand increased OCR in these cetuximab-resistant cells,indicating that knockdown of HIF-1a can reverse theWarburg effect in cetuximab-resistant cells, just as cetux-imab reversed the Warburg effect in cetuximab-sensitivecells (Fig. 1E).

Taken together, these novel findings indicate that LDH-A was downregulated and glycolysis was inhibited viadownregulation of HIF-1a by cetuximab in cetuximab-sensitive HN5 and FaDu cells but not in the cetuximab-resistant sublines. The resistant sublines exhibitedincreased glycolytic potential that could be tempered byknockdown of HIF-1a.

Cetuximab reduces glucose consumption, lactateproduction, and intracellular ATP level in an HIF-1adownregulation–dependent manner

To confirm the effect of cetuximab on inhibiting aerobicglycolysis, we directly measured glucose consumption

Lu et al.





and lactate production after cetuximab treatment in HN5and FaDu cells that did or did not overexpress an exper-imentally created HIF-1a mutant (HIF-1a/DODD). Thismutant retains the majority of the transcriptional activity

of full-length HIF-1a and can be stably overexpressedin normoxic culture because of experimental deletion ofthe ODD domain in HIF-1a that renders HIF-1a highlyunstable in normoxic culture (16, 17). In HN5 and FaDu

Figure 1. Cetuximab downregulates LDH-A and inhibits glycolytic flux in HNSCC cells. A, HN5 and FaDu cells and their cetuximab-resistant sublines (HN5-Rand FaDu-R) were treated with the indicated doses of cetuximab for 5 days in low-glucose (1 g/L), low-serum (0.5% FBS) medium. Cell growth responses tocetuximab were measured using an MTT assay. B, the indicated cells were treated with 20 nmol/L cetuximab or left untreated for 24 hours. Cell lysateswere prepared, and equal amounts of cell lysates were subjected to Western blot analysis with the indicated antibodies. C, HN5, HN5-R, FaDu, and FaDu-Rcells were treated with 20 nmol/L cetuximab or left untreated for 10 hours. Metabolic flux analysis was conducted to measure changes in ECAR andOCR and was followed by a mitochondrial stress test as indicated. The final concentrations of oligomycin, FCCP, antimycin, and rotenone were all 1 mmol/L.Error bars, SD.D, the indicatedcell lineswere subjected to knockdownofHIF-1awith eachof 2differentHIF-1a siRNAsor control siRNA for 48hours, followedby Western blotting with the indicated antibodies. Ctr, control. E, HN5-R and FaDu-R cells were subjected to knockdown of HIF-1a as described in D.Metabolic flux analysis was then conducted to measure ECAR and OCR.






cells transfected with a control vector (HN5-neo andFaDu-neo cells), cetuximab inhibited glucose consump-tion and lactate production in a time-dependent manner(Fig. 2A and B). Consistent with the knowledge thatHIF-1 upregulates glycolysis, overexpression of HIF-1a/DODD in HN5 and FaDu cells (HN5-HIF-1a/DODDand FaDu-HIF-1a/DODD cells) significantly increasedthe levels of glucose consumption and lactate productionand conferred resistance to cetuximab-induced inhibitionof glucose consumption and lactate production.

We also assessed the differences in the intracellularATP levels of cells untreated and treated with cetux-imab for 4 hours when no changes in cell numbers weredetected between the 2 groups. The ATP levels of theHN5-neo and FaDu-neo cells treated with cetuximabwere substantially lower than that of the untreated cells(down 39% and 31%, respectively). In contrast, the ATPlevels of the HN5-HIF-1a/DODD and FaDu-HIF-1a/DODD cells treated with cetuximab cells were onlymodestly lower than that of untreated cells (down15% and 11%, respectively; Fig. 2C). These findings areconsistent with the knowledge that ATP generatedduring glycolysis is a major source of energy in cancercells.

Cetuximab inhibits LDH-A expression andenzymatic activity by downregulating HIF-1a

To further elucidate the biochemical mechanismsunderlying cetuximab-mediated inhibition of glycolysisvia HIF-1a downregulation, we compared the effects ofcetuximab on downregulation of LDH-A level and inhi-bition of EGFR downstream signaling pathways in HN5-neo cells with those in HN5-HIF-1a/DODD cells and inFaDu-neo cells with those in FaDu-HIF-1a/DODD cells.We found that cetuximab treatment downregulated thelevel of LDH-A in both HN5-neo and FaDu-neo cells butnot in the cell lines overexpressing HIF-1a/DODD (Fig.3A). Interestingly, similar to our finding that the phos-phorylation levels of Akt and Erk in HN5-R and FaDu-Rcells were sensitive or partially sensitive to cetuximab(Fig. 1B), the phosphorylation levels of these moleculesin HN5-HIF-1a/DODD and FaDu-HIF-1a/DODD cellswere also sensitive to cetuximab. The levels of HIF-1aincreased steadily after the regular culture medium waschanged for low-glucose, low-serum medium, whichlikely caused metabolic changes that can upregulate theHIF-1a level, such as mTOR activation. However, bothHN5-neo and FaDu-neo cells treated with cetuximab hadlower levels of HIF-1a than untreated cells at each time of

Figure 2. Cetuximab inhibits glucose consumption, lactate production, and intracellular ATP levels in a HIF-1a inhibition–dependent manner. A, HN5-neo,HN5-HIF-1a/DODD, FaDu-neo, and FaDu-HIF-1a/DODD cells were seeded in 6-well plates with 3mL of low-glucose (1 g/L), low-serum (0.5%FBS) medium.The cells were treated with 20 nmol/L cetuximab or left untreated. At the indicated times after treatment, the level of glucose remaining in theconditionedmediumwas determined as described inMaterials andMethods. B, the indicated cells were seeded and treated similarly as described in A. At theindicated times after treatment, the level of lactate produced in the conditioned medium was determined as described in Materials and Methods. C, theindicated cells were treated with 20 nmol/L cetuximab or left untreated in low-glucose (1 g/L), low-serum (0.5% FBS) medium for 4 hours. Cell pelletswere harvested, and the intracellular level of ATP was measured using a luciferase-based ATP determination assay. The relative values of ATP in the treatedgroups were expressed as percentages of the value of ATP in the corresponding untreated groups. The data are presented as means � SD.

Lu et al.





measurement. The kinetics of cetuximab-mediated inhi-bition of EGFR downstream pathways was different: Erkphosphorylationwas inhibited earlier thanAktphosphor-ylation; however, Erk phosphorylation recovered sub-stantially 24 hours after treatmentwhereasAkt phosphor-ylation did not. In cells overexpressing HIF-1a/DODD,even the endogenous HIF-1a became resistant to cetux-imab treatment, an interesting phenomenon that wereported previously (42). These findings provide furthersupport that HIF-1 has a fundamental role in regulatingcell response to cetuximab-induced growth inhibition.Consistentwith the decrease in LDH-Aprotein levels in

HN5-neo and FaDu-neo cells after cetuximab treatment,the activity of LDH-A in catalyzing the production oflactatewas also decreased in these cells as early as 4 hoursafter cetuximab treatment (Fig. 3B). In contrast, HN5-HIF-1a/DODD, and FaDu-HIF-1a/DODD cells had muchhigher LDH-A enzymatic activity than HN5-neo andFaDu-neo cells, and the elevated LDH-A activity was notinhibited by cetuximab (Fig. 3B). Together, these findingsindicate that cetuximab downregulates HIF-1a, therebydownregulating LDH-A expression and inhibiting LDH-A activity to inhibit glycolysis.

Knockdown of LDH-A overcomes resistance tocetuximab-induced G1 arrestAerobic glycolysis has been suggested to be a means of

fueling cancer cells with the biomass needed for prolif-eration (3, 48). To test our hypothesis that cetuximab

inhibits cancer cell proliferation by downregulatingHIF-1a and subsequently LDH-A to inhibit glycolysis,we conducted cell-cycle analysis after subjecting cells totreatment with cetuximab and/or knockdown of LDH-Aby siRNA. We used 3 different LDH-A–specific siRNAs,all of which knocked down the level of LDH-A in allparental and sublines tested (Fig. 4A, C and E). LDH-AsiRNA #1 and #2 were used for knockdown of LDH-A inthe cells subjected to cell-cycle analysis.

As shown in Fig. 4B, in HN5-neo and FaDu-neo cells,the percentages of cetuximab-treated cells in G1 phase(85.0% and 83.2%, respectively) were higher than thoseof the untreated cells in G1 phase (67.9% and 64.9%,respectively). LDH-A knockdown alone by siRNA #1increased the percentages of G1 phase cells modestly(73.0% and 70.2%, respectively) but the combinationof cetuximab and LDH-A knockdown had an addi-tive effect on G1 arrest (90.3% and 87.1%, respectively).Similar results were obtained using LDH-A siRNA#2 in HN5-neo and FaDu-neo cells (supplementaryFig. S1).

Both the cetuximab-resistant sublines (HN5-R andFaDu-R) and the HIF-1a/DODD–overexpressing cells(HN5-HIF-1a/DODD and FaDu-HIF-1a/DODD cells)were resistant to cetuximab-induced G1 arrest. As shownin Fig. 4D, compared with the percentages of untreatedHN5-R cells and FaDu-R cells in G1 phase (53.3% and60.9%, respectively), the percentages of cetuximab-treat-ed HN5-R cells and FaDu-R cells in G1 phase were 54.0%

Figure 3. Cetuximabdownregulates LDH-A expressionand reduces LDH-A activity in aHIF-1a inhibition–dependentmanner. HN5 and FaDu cellsexpressing control vector (HN5-neo and FaDu-neo) or HIF-1a/DODD (HN5-HIF-1a/DODD andFaDu-HIF-1a/DODD) were treatedwith 20 nmol/L cetuximab or leftuntreated in low-glucose (1 g/L),low-serum (0.5% FBS) medium forthe indicated times. A, cell lysateswere prepared, and equal amountsof cell lysates were subjected toWestern blot analysis with theindicated antibodies. Thequantification of the LDH-A bandwas conducted using ImageJsoftware, and the data wereexpressed as fold increases inreference to the leftmost lane of theblots, the value of which wasarbitrarily set as 1. B, the indicatedcells were harvested andhomogenized, and samplescontaining equal amounts ofproteinswere subjected to anLDH-A activity assay as described inMaterials and Methods. The dataare presented as means � SD.






and 60.7%, respectively. As shown in Fig. 4F, comparedwith the percentages of untreated HN5-HIF-1a/DODDcells and FaDu-HIF-1a/DODD cells in G1 phase (51.7%and 64.4%, respectively), the percentages of cetuximab-treated HN5-HIF-1a/DODD cells and FaDu-HIF-1a/DODD cells in G1 phase were 52.5% and 67.9%, respective-ly. Thesefindingsare inagreementwith the roleofHIF-1a/DODD in counteracting cetuximab-mediated inhibition ofglycolysis and downregulation of LDH-A (Figs. 2 and 3).

Knockdown of LDH-A alone had modest effects onincreasing the percentage of G1 phase cells in the cetux-

imab-resistant cells. As shown in Fig. 4D, compared withthe percentages of untreated HN5-R cells and FaDu-R cellsinG1phase (53.3%and60.9%, respectively), thepercentagesof LDH-A knockdown HN5-R cells and FaDu-R cells in G1

phase were modestly higher (59.6% and 71.0%, respective-ly). As shown in Fig. 4F, compared with the percentages ofuntreated HN5-HIF-1a/DODD cells and FaDu-HIF-1a/DODD cells in G1 phase (51.7% and 64.4%, respectively),the percentages of LDH-A knockdown HN5-HIF-1a/DODD cells and FaDu-HIF-1a/DODD cells in G1 phaseweremodestly also higher (64.3% and 70.3%, respectively).

Figure 4. Overexpression of HIF-1a/DODD confers resistance to cetuximab-induced cell-cycle arrest, which can be overcome by knockdown of LDH-A. A, C,and E, the indicated cells were subjected to knockdown of LDH-A with each of 3 different LDH-A–specific siRNA or control siRNA in low-glucose (1 g/L),low-serum (0.5% FBS) medium for 48 hours. Cell lysates were prepared, and equal amounts of cell lysates were subjected to Western blot analysiswith the indicated antibodies. B, D, and F, the indicated cells were subjected to knockdownof LDH-Awith LDH-A siRNA#1 or control siRNA as described in A,C, and E. The cells were then treated with 20 nmol/L cetuximab or left untreated for 24 hours and then fixed and stained with propidium iodide forcell-cycle distribution analysis by flow cytometry. The data are presented as means � SD.

Lu et al.





Importantly, knockdown of LDH-A by siRNA #1 large-ly restored the activity of cetuximab in inducing G1 cell-cycle arrest in cetuximab-resistant cells. Compared withthe percentage of cells in G1 phase after cetuximab orLDH-A knockdown alone (Fig. 4D and F), the percentageof G1 phase cells after the combination of cetuximab andLDH-Aknockdown increased to 72.3%and 76.5%, respec-tively, inHN5-R and FaDu-R cells, and increased to 84.7%and 83.0%, respectively, in HN5-HIF-1a/DODD andFaDu-HIF-1a/DODD cells. Similar results were obtainedusing LDH-A siRNA #2 (Supplementary Fig. S1).Together, these findings confirm the role of HIF-1a

in cetuximab-induced cell-cycle arrest and suggest thattargeting LDH-A is an effective strategy for improvingcetuximab response.

Knockdown of LDH-A or inhibition of LDH-A withoxamate enhances the therapeutic effect ofcetuximab in cancer cellsTo determine whether knockdown of LDH-A can

also potentiate cetuximab-induced apoptosis, we used2 independent apoptosis assays (one detecting PARPcleavage and the other detecting DNA fragmentation)to assess the induction of apoptosis in HN5 cells (paren-tal HN5, HN5-R, and HN5-HIF-1a/DODD cells) andFaDu cells (parental FaDu, FaDu-R, and FaDu-HIF-1a/DODD cells) after subjecting the cells to treatmentwith cetuximab and/or knockdown of LDH-A. Treat-mentwith cetuximab or knockdownof LDH-Aalone hadlittle effect on inducing apoptosis in parental cells orcetuximab-resistant cells. Cetuximab plus LDH-Aknockdown caused an increase in the levels of PARPcleavage andDNA fragmentation inHN5 and FaDu cellsbut had only minimal or no effect in the cetuximab-resistant cells (Fig. 5A and B).Next, we assessed the effects of oxamate, a small-

molecule inhibitor of LDH-A (49), alone and in combi-nation with cetuximab on the growth and survival ofHNSCC cells after 5-day extended culture (Fig. 5C).MTT assays, which measure the cytostatic and cytotoxiceffects of the treatments, showed that oxamate aloneonly modestly inhibited cell growth and survival inHN5 cells but had a stronger effect in HN5-R andHN5-HIF-1a/DODD cells, which are more glycolyticthan their parental cells. Oxamate alone had a similarlymodest effect on inhibiting cell growth and survival inFaDu, FaDu-R, and FaDu-HIF-1a/DODD cells; howev-er, the combination of cetuximab and oxamate signifi-cantly enhanced the inhibition of cell growth and sur-vival compared with either single treatment alone, par-ticularly in the cetuximab-resistant sublines. The com-bination treatment yielded similar results in another 2HNSCC cell lines, SqCC/Y1 and TU167, which aremuch less sensitive to cetuximab-induced growth inhi-bition than HN5 and FaDu cells (Supplementary Fig.S2). Together, these data suggest that co-targeting LDH-A with oxamate and cetuximab improves the therapeu-tic effect of cetuximab in cancer cells.

Cetuximab has no significant effect on inhibitingglycolysis in nontransformed cells in normoxia

To determine whether cetuximab affects glycolysis innontransformed cells, we assessed changes in cell prolif-eration, glucose consumption, lactate production, LDH-Aactivity, andATP level inNOM9-TKcells, an immortalizedhuman head and neck epithelial cell line, treated withcetuximab or left untreated in normoxic culture. In theabsence of EGF, NOM9-TK proliferation was slow andbarely inhibited by cetuximab (Fig. 6A). Compared withcells treated under hypoxic conditions, cells treated undernormoxic conditions expressed a very low level of HIF-1a.Twenty-four–hour cetuximab treatment inhibited EGFRautophosphorylationandErkphosphorylationbutnotAktphosphorylation (Fig. 6B). Although cetuximab inhibitedLDH-A expression and aerobic glycolysis in HN5 andFaDu cells in normoxic culture (Fig. 1), it did not inhibitLDH-A expression inNOM9-TK cells (Fig. 6B) and did notsignificantly inhibit glucose consumption (Fig. 6C), lactateproduction (Fig. 6D), LDH-A activity (Fig. 6E), or ATPproduction (Fig. 6F) in the cells in normoxic culture. Theabsolute amount of lactate production in the NOM9-TKcells (Fig. 6D) wasmarkedly less than that in the HN5 andFaDucells during the sameperiod (Fig. 2B). Thesefindingsindicate that cetuximab inhibits glycolysis selectively incancer cells through the downregulation of LDH-A,whichcauses cancer cells that rely heavily on aerobic glycolysisfor biosynthetic metabolism to arrest at G1 phase.

DiscussionIn the current work, we showed that HNSCC cells with

acquired resistance to cetuximab have higher levelsof HIF-1a and are more glycolytic than their parentalcells. Overexpression of a degradation-resistant HIF-1a(HIF-1a/DODD) elevated the glycolytic potential of thecells and conferred resistance to cetuximab-induced gly-colysis inhibition and cell-cycle arrest. We also showedthat cetuximab-induced HIF-1a downregulation led toLDH-A downregulation and that downregulation ofLDH-A expression by siRNA or inhibition of LDH-Aactivity improved cancer cell response to cetuximab. Ourfindings provide strong evidence suggesting that glycol-ysis inhibition plays a major role in the cetuximab-induced inhibition of cell proliferation.

A fundamental difference in metabolism between can-cer cells and normal cells is that cancer cells rely on anuninterrupted supply of biomass to fuel their unlimitedproliferation. Cancer cells thus have evolved to adapt tothe energy-less-efficient type of metabolism, that is, aer-obic glycolysis or the Warburg effect, by directing thepyruvate toward production of lactate in the cytosolunder the action of LDH-A, through which an adequatesupply of biomass is produced via an elevated glycolyticflux. Cetuximab reverses theWarburg effect by inhibitingHIF-1–regulated LDH-A in cancer cells, thereby decreas-ing the supply of biomass needed for proliferation; thecells are thus arrested at G1 phase. Unlike the metabolism






Figure 5. LDH-A knockdown or inhibition enhances the therapeutic effect of cetuximab. The indicated cells were subjected to knockdown of LDH-A withspecific siRNA or control siRNA for 48 hours and then treated with 20 nmol/L cetuximab or left untreated for 24 hours in low-glucose (1 g/L), low-serum (0.5%FBS) medium. A, after treatment, cell lysates were prepared, and equal amounts of cell lysates were subjected to Western blot analysis with theindicated antibodies. The quantification of the LDH-A and cleaved PARP bands was conducted using ImageJ software, and the data were expressed as foldincreases in reference to the leftmost lane of the blots, the value of which was arbitrarily set as 1. B, the cell lysates in A were analyzed for the levelof DNA fragmentation with an apoptosis ELISA kit as described inMaterials andMethods. The data in the treated groups were expressed as fold increases inthe optical density value compared with the value in the corresponding untreated cells. C, the indicated cells were treated with 20 nmol/L cetuximab and/or 5mmol/L oxamate or left untreated for 5 days. Cell growth responses to cetuximab were measured using an MTT assay. The growth and survival of thetreated groups were expressed as percentages of the OD value in the corresponding untreated groups. The data are presented as means � SD.

Lu et al.





in cancer cells, the pyruvate in normal cells enters themitochondria and is fully oxidized through the tricarbox-ylic acid cycle, thereby generating large amounts of ATP.This difference may explain why cetuximab inhibits gly-colysis mainly in cancer cells, but not normal cells, afterLDH-A downregulation.We showed that knockdown of LDH-A expression or

inhibition of LDH-A activity inhibited cell proliferation inboth cetuximab-sensitive and cetuximab-resistant cells.Depending on the extent to which glucose-derived pyru-vate is redirected to the tricarboxylic acid cycle for oxi-dative phosphorylation in cancer cells, inhibition of LDH-A could induce apoptosis in addition to cell-cycle arrest.This is because excess oxidative phosphorylation of thehigh level of pyruvate in cancer cells can cause overpro-duction of reactive oxygen species that are cytotoxic tocells, leading to apoptosis. However, in our models, theeffects of knockdownofLDH-Aor inhibitionof LDH-Aoninducing apoptosis seem to be limited. It was observedonly when the targeting of LDH-A was combined withcetuximab in cetuximab-sensitive cells.

Another interesting finding of the current work is thatthe cetuximab-resistant cells (HN5-R, FaDu-R, HN5-HIF-1a/DODD, and FaDu-HIF-1a/DODD cells) remained atleast partially sensitive to cetuximab-induced inhibitionof major downstream cell signaling pathways of EGFR—that is, Akt phosphorylation and Erk phosphorylation—despite their resistance to cetuximab-induced growthinhibition. These findings suggest that, compared withthese downstream signalingmolecules, LDH-A is a betterbiomarker for predicting cancer cell responsiveness tocetuximab treatment. It is noteworthy that, despite thehigh level of HIF-1a or HIF-1a/DODD, the basal level ofLDH-A was only modestly elevated in these cetuximab-resistant cells. This ismost likely because the basal level ofLDH-A was sufficient to drive glycolysis toward lactateproduction. It was when the cells were treated withcetuximab, an elevated HIF-1a or HIF-1a/DODD wasneeded to protect against downregulation of LDH-A bycetuximab.

In the current study, we did not explore the mechan-isms underlying the increase in the basal HIF-1a levels

Figure 6. Cetuximab does not affect glycolysis in nontransformed cells. A, NOM9-TK cells were treatedwith 20 nmol/L cetuximab or left untreated for 5 days inchemically defined keratinocyte growth medium supplemented with components from the SingleQuots Kit except EGF. Cell growth response tocetuximab wasmeasured using anMTT assay. B, NOM9-TK cells were treated with 20 nmol/L cetuximab or left untreated for 24 hours under normoxia. Cellstreated the sameway under hypoxia served as controls. Cell lysateswere subjected toWestern blot analysiswith the indicated antibodies. C andD,NOM9-TKcells were seeded in 6-well plates with 3 mL of low-glucose (1 g/L) cell culture medium and treated with 20 nmol/L cetuximab or left untreated for theindicated times. At the indicated times after treatment, the level of glucose remaining in the conditioned medium (C) and the level of lactate produced in theconditioned medium (D) were determined with assays described in Materials and Methods. E, NOM9-TK cells were treated with 20 nmol/L cetuximab or leftuntreated for the indicated times. Cells were harvested and homogenized, and samples containing equal amounts of proteins were subjected to an LDH-Aactivity assay as described in Materials and Methods. F, NOM9-TK cells were treated with 20 nmol/L cetuximab or left untreated in serum-freekeratinocyte growth medium for 4 hours. Cell pellets were harvested, and the intracellular level of ATP was measured using a luciferase-based ATPdetermination assay. The data are presented as means � SD.






that makes HN5-R and FaDu-R cells highly glycolytic.We found inconsistent changes in the levels of EGFRdownstream cell signaling in these cells; for example,the level of activation-specific Akt phosphorylationwas higher in HN5-R cells than in HN5 cells, whereasthe level of activation-specific Erk phosphorylationwas higher in FaDu-R cells than in FaDu cells. Thesefindings suggest that either the underlying mechanismsare cell context–specific or additional mechanisms, suchas Src activation, also play a role in upregulatingHIF-1a incetuximab-resistant cells. We are currently investigatingthese possibilities.

An important caveat to our findings is that downregu-lation of LDH-A is required but may not be sufficient forcetuximab-induced growth inhibition. The inhibition ofother glycolytic pathway enzymes, which are also subjectto regulation by HIF-1, may also be necessary for cetux-imab-mediated inhibition of cancer cell proliferation.

In summary, we showed that cetuximab inhibits cancercell proliferation by inhibiting glycolysis. Our findingssuggest that LDH-A is a novel biomarker for predictingcetuximab response and a candidate target for sensitizingcetuximab-resistant cells to cetuximab treatment. Ourwork fills a major knowledge gap regarding the linkbetween EGFR-targeted therapy and cancer cell metabo-lism and advances our understanding of how cetuximabinhibits the proliferation of cancer cells.

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

Authors' ContributionsConception and design: H. Lu, X. Li, Z. FanDevelopment of methodology: H. Lu, X. Li, Z. LuoAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): H. Lu, X. Li, Z. LuoAnalysis and interpretation of data (e.g., statistical analysis, biostatis-tics, computational analysis): H. Lu, X. Li, Z. Luo, J. Liu, Z. FanWriting, review, and/or revision of the manuscript: H. Lu, Z. FanAdministrative, technical, or material support (i.e., reporting or orga-nizing data, constructing databases): H. LuStudy supervision: Z. Fan

AcknowledgmentsThe authors thank Dr. Jay Dunn of Seahorse Bioscience for technical

assistance with the Seahorse XF96 extracellular flux analyzer.

Grant SupportThis work was supported in part by U.S. NIH R01 award (CA129036)

and R21 award (DE021883) to Z. Fan. Theworkwas also supported in partby the NIH through MD Anderson’s Cancer Center Support Grant,CA016672.

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 2, 2013; revised July 2, 2013; accepted July 24, 2013;published OnlineFirst August 6, 2013.

References1. Warburg O. On the origin of cancer cells. Science 1956;123:309–14.2. Hsu PP, Sabatini DM. Cancer cell metabolism: Warburg and beyond.

Cell 2008;134:703–7.3. Vander Heiden MG, Cantley LC, Thompson CB. Understanding the

Warburg effect: the metabolic requirements of cell proliferation. Sci-ence 2009;324:1029–33.

4. DeBerardinis RJ, LumJJ, HatzivassiliouG, ThompsonCB. The biologyof cancer: metabolic reprogramming fuels cell growth and prolifera-tion. Cell Metab 2008;7:11–20.

5. Grander D. How do mutated oncogenes and tumor suppressor genescause cancer? Med Oncol 1998;15:20–6.

6. Dang CV, Semenza GL. Oncogenic alterations of metabolism. TrendsBiochem Sci 1999;24:68–72.

7. DeBerardinis RJ, Sayed N, Ditsworth D, Thompson CB. Brick by brick:metabolism and tumor cell growth. Curr Opin Genet Dev 2008;18:54–61.

8. SemenzaGL.HIF-1: upstreamanddownstreamof cancermetabolism.Curr Opin Genet Dev 2010;20:51–6.

9. Maxwell PH, Pugh CW, Ratcliffe PJ. Inducible operation of the eryth-ropoietin 30 enhancer in multiple cell lines: evidence for a widespreadoxygen-sensingmechanism.ProcNatl AcadSciUSA1993;90:2423–7.

10. Firth JD, Ebert BL, Pugh CW, Ratcliffe PJ. Oxygen-regulated controlelements in the phosphoglycerate kinase 1and lactate dehydrogenaseA genes: similarities with the erythropoietin 30 enhancer. Proc NatlAcad Sci U S A 1994;91:6496–500.

11. Semenza GL, Roth PH, Fang HM,Wang GL. Transcriptional regulationof genes encoding glycolytic enzymes by hypoxia-inducible factor 1.J Biol Chem 1994;269:23757–63.

12. Wang GL, Jiang BH, Rue EA, Semenza GL. Hypoxia-inducible factor 1is a basic-helix-loop-helix-PAS heterodimer regulated by cellular O2tension. Proc Natl Acad Sci U S A 1995;92:5510–4.

13. Semenza GL. Hypoxia, clonal selection, and the role of HIF-1 in tumorprogression. Crit Rev Biochem Mol Biol 2000;35:71–103.

14. Semenza GL. Involvement of hypoxia-inducible factor 1 in humancancer. Intern Med 2002;41:79–83.

15. Semenza GL. Defining the role of hypoxia-inducible factor 1 in cancerbiology and therapeutics. Oncogene 2010;29:625–34.

16. Ivan M, Kondo K, Yang H, Kim W, Valiando J, Ohh M, et al. HIFalphatargeted for VHL-mediated destruction by proline hydroxylation: impli-cations for O2 sensing. Science 2001;292:464–8.

17. Jaakkola P, Mole DR, Tian YM, Wilson MI, Gielbert J, Gaskell SJ,et al. Targeting of HIF-alpha to the von Hippel-Lindau ubiquitylationcomplex by O2-regulated prolyl hydroxylation. Science 2001;292:468–72.

18. Jiang BH, Agani F, Passaniti A, Semenza GL. V-SRC inducesexpression of hypoxia-inducible factor 1 (HIF-1) and transcriptionof genes encoding vascular endothelial growth factor and enolase1: involvement of HIF-1 in tumor progression. Cancer Res 1997;57:5328–35.

19. Zhong H, Chiles K, Feldser D, Laughner E, Hanrahan C, GeorgescuMM, et al. Modulation of hypoxia-inducible factor 1alpha expressionby the epidermal growth factor/phosphatidylinositol 3-kinase/PTEN/AKT/FRAP pathway in human prostate cancer cells: implications fortumor angiogenesis and therapeutics. Cancer Res 2000;60:1541–5.

20. Treins C, Giorgetti-Peraldi S, Murdaca J, Semenza GL, Van Obber-ghen E. Insulin stimulates hypoxia-inducible factor 1 through aphosphatidylinositol 3-kinase/target of rapamycin-dependent sig-naling pathway. J Biol Chem 2002;277:27975–81.

21. Fukuda R, Hirota K, Fan F, Jung YD, Ellis LM, Semenza GL. Insulin-likegrowth factor 1 induces hypoxia-inducible factor 1-mediated vascularendothelial growth factor expression, which is dependent on MAPkinase and phosphatidylinositol 3-kinase signaling in colon cancercells. J Biol Chem 2002;277:38205–11.

22. Laughner E, Taghavi P, Chiles K, Mahon PC, Semenza GL. HER2 (neu)signaling increases the rate of hypoxia-inducible factor 1alpha (HIF-1alpha) synthesis: novel mechanism for HIF-1-mediated vascular

Lu et al.





endothelial growth factor expression. Mol Cell Biol 2001;21:3995–4004.

23. Cunningham D, Humblet Y, Siena S, Khayat D, Bleiberg H, Santoro A,et al. Cetuximab monotherapy and cetuximab plus irinotecan in irino-tecan-refractory metastatic colorectal cancer. N Engl J Med 2004;351:337–45.

24. Bonner JA, Harari PM, Giralt J, Azarnia N, Shin DM, Cohen RB, et al.Radiotherapy plus cetuximab for squamous-cell carcinoma of thehead and neck. N Engl J Med 2006;354:567–78.

25. Fan Z, Masui H, Altas I, Mendelsohn J. Blockade of epidermal growthfactor receptor function by bivalent and monovalent fragments of 225anti-epidermal growth factor receptor monoclonal antibodies. CancerRes 1993;53:4322–8.

26. Fan Z, Baselga J, Masui H, Mendelsohn J. Antitumor effect of anti-epidermal growth factor receptor monoclonal antibodies plus cis-diamminedichloroplatinum on well established A431 cell xenografts.Cancer Res 1993;53:4637–42.

27. Fan Z, Mendelsohn J, Masui H, Kumar R. Regulation of epidermalgrowth factor receptor in NIH3T3/HER14 cells by antireceptor mono-clonal antibodies. J Biol Chem 1993;268:21073–9.

28. Fan Z, Lu Y,WuX,Mendelsohn J. Antibody-induced epidermal growthfactor receptor dimerization mediates inhibition of autocrine prolifer-ation of A431 squamous carcinoma cells. J Biol Chem 1994;269:27595–602.

29. Wu X, Fan Z, Masui H, Rosen N, Mendelsohn J. Apoptosis induced byan anti-epidermal growth factor receptor monoclonal antibody in ahuman colorectal carcinoma cell line and its delay by insulin. J ClinInvest 1995;95:1897–905.

30. Peng D, Fan Z, Lu Y, DeBlasio T, Scher H, Mendelsohn J. Anti-epidermal growth factor receptor monoclonal antibody 225 up-reg-ulates p27KIP1 and induces G1 arrest in prostatic cancer cell lineDU145. Cancer Res 1996;56:3666–9.

31. WuX, RubinM, Fan Z, DeBlasio T, Soos T, Koff A, et al. Involvement ofp27KIP1 in G1 arrest mediated by an anti-epidermal growth factorreceptor monoclonal antibody. Oncogene 1996;12:1397–403.

32. Fan Z, ShangBY, LuY, Chou JL,Mendelsohn J. Reciprocal changes inp27(Kip1) and p21(Cip1) in growth inhibition mediated by blockade oroverstimulation of epidermal growth factor receptors. Clin Cancer Res1997;3:1943–8.

33. Ciardiello F,BiancoR,DamianoV,DeLorenzoS, PepeS,DePlacidoS,et al. Antitumor activity of sequential treatment with topotecan andanti-epidermal growth factor receptormonoclonal antibodyC225. ClinCancer Res 1999;5:909–16.

34. Ye D, Mendelsohn J, Fan Z. Androgen and epidermal growth factordown-regulate cyclin-dependent kinase inhibitor p27Kip1 and costi-mulate proliferation of MDA PCa 2a and MDA PCa 2b prostate cancercells. Clin Cancer Res 1999;5:2171–7.

35. Liu B, Fan Z. The monoclonal antibody 225 activates caspase-8 andinduces apoptosis through a tumor necrosis factor receptor family-independent pathway. Oncogene 2001;20:3726–34.

36. Liu B, FangM, LuY,Mendelsohn J, FanZ. Fibroblast growth factor andinsulin-like growth factor differentially modulate the apoptosis and G1

arrest induced by anti-epidermal growth factor receptor monoclonalantibody. Oncogene 2001;20:1913–22.

37. Nasu S, Ang KK, Fan Z, Milas L. C225 antiepidermal growth factorreceptor antibody enhances tumor radiocurability. Int J Radiat OncolBiol Phys 2001;51:474–7.

38. Karashima T, Sweeney P, Slaton JW, Kim SJ, Kedar D, Izawa JI, et al.Inhibition of angiogenesis by the antiepidermal growth factor receptorantibody ImClone C225 in androgen-independent prostate cancergrowing orthotopically in nudemice. Clin Cancer Res 2002;8:1253–64.

39. Milas L, Fan Z, Mason K, Ang KK. Role of epidermal growth factorreceptor and its inhibition in radiotherapy. InNieder C, Milas L, AngKK, (eds.). Modification of radiation response: cytokines, growthfactors, and other biological targets. Berlin, Germany: Springer;2003. p. 189–204.

40. Lu Y, Li X, Liang K, Luwor R, Siddik ZH, Mills GB, et al. Epidermalgrowth factor receptor (EGFR) ubiquitination as a mechanism ofacquired resistance escaping treatment by the anti-EGFRmonoclonalantibody cetuximab. Cancer Res 2007;67:8240–7.

41. Luwor RB, Lu Y, Li X, Mendelsohn J, Fan Z. The antiepidermal growthfactor receptor monoclonal antibody cetuximab/C225 reduces hyp-oxia-inducible factor-1 alpha, leading to transcriptional inhibition ofvascular endothelial growth factor expression. Oncogene 2005;24:4433–41.

42. Lu Y, Liang K, Li X, Fan Z. Responses of cancer cells with wild-type ortyrosine kinase domain-mutated epidermal growth factor receptor(EGFR) to EGFR-targeted therapy are linked to downregulation ofhypoxia-inducible factor-1alpha. Mol Cancer 2007;6:63.

43. Li X, Lu Y, Liang K, Pan T, Mendelsohn J, Fan Z. Requirement ofhypoxia-inducible factor-1alpha down-regulation in mediating theantitumor activity of the anti-epidermal growth factor receptor mono-clonal antibody cetuximab. Mol Cancer Ther 2008;7:1207–17.

44. Lu Y, Li X, Lu H, Fan Z. 1, 9-Pyrazoloanthrones downregulate HIF-1alpha and sensitize cancer cells to cetuximab-mediated anti-EGFRtherapy. PLoS One 2010;5:e15823.

45. Lu H, Liang K, Lu Y, Fan Z. The anti-EGFR antibody cetuximabsensitizes human head and neck squamous cell carcinoma cells toradiation in part through inhibiting radiation-induced upregulation ofHIF-1alpha. Cancer Lett 2012;322:78–85.

46. Zhao M, Sano D, Pickering CR, Jasser SA, Henderson YC, ClaymanGL, et al. Assembly and initial characterization of a panel of 85genomically validated cell lines from diverse head and neck tumorsites. Clin Cancer Res 2011;17:7248–64.

47. Chakravarti N, Kadara H, Yoon DJ, Shay JW, Myers JN, Lotan D, et al.Differential inhibition of protein translation machinery by curcumin innormal, immortalized, and malignant oral epithelial cells. Cancer PrevRes (Phila) 2010;3:331–8.

48. Vander Heiden MG, Locasale JW, Swanson KD, Sharfi H, Heffron GJ,Amador-Noguez D, et al. Evidence for an alternative glycolytic path-way in rapidly proliferating cells. Science 2010;329:1492–9.

49. Zhao Y, Liu H, Liu Z, Ding Y, Ledoux SP, Wilson GL, et al. Overcomingtrastuzumab resistance in breast cancer by targeting dysregulatedglucose metabolism. Cancer Res 2011;71:4585–97.






2013;12:2187-2199. Published OnlineFirst August 6, 2013.Mol Cancer Ther Haiquan Lu, Xinqun Li, Zhongguang Luo, et al. Regulated LDH-A

−Cetuximab Reverses the Warburg Effect by Inhibiting HIF-1

Updated version

10.1158/1535-7163.MCT-12-1245doi:

Access the most recent version of this article at:

Material

Supplementary

http://mct.aacrjournals.org/content/suppl/2013/08/06/1535-7163.MCT-12-1245.DC1

Access the most recent supplemental material at:

Cited articles

http://mct.aacrjournals.org/content/12/10/2187.full#ref-list-1

This article cites 48 articles, 28 of which you can access for free at:

Citing articles

http://mct.aacrjournals.org/content/12/10/2187.full#related-urls

This article has been cited by 2 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://mct.aacrjournals.org/content/12/10/2187To request permission to re-use all or part of this article, use this link



http://mct.aacrjournals.org/lookup/doi/10.1158/1535-7163.MCT-12-1245

http://mct.aacrjournals.org/content/suppl/2013/08/06/1535-7163.MCT-12-1245.DC1

http://mct.aacrjournals.org/content/12/10/2187.full#ref-list-1

http://mct.aacrjournals.org/content/12/10/2187.full#related-urls

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

mailto:[email protected]

http://mct.aacrjournals.org/content/12/10/2187


Documents

Cetuximab Reverses the Warburg Effect by Inhibiting HIF-1 ... · Cells were cultured in 24-well plates with low-glucose (1 g/L), low-serum (0.5% FBS) medium (0.5 mL/well) at 37 C