CANCER RESEARCH | TUMOR BIOLOGYAND IMMUNOLOGY

Dysregulation of Glutamate Transport Enhances TregFunction That Promotes VEGF Blockade Resistance inGlioblastoma A C

Yu Long1, Haipeng Tao1, Aida Karachi2, Adam J. Grippin2, Linchun Jin2, Yifan (Emily) Chang2,Wang Zhang1, Kyle A. Dyson2, Alicia Y. Hou2, Meng Na1, Loic P. Deleyrolle2,3, Elias J. Sayour2,3,Maryam Rahman2,3, Duane A. Mitchell2,3, Zhiguo Lin1, and Jianping Huang2,3

ABSTRACT◥

Anti-VEGF therapy prolongs recurrence-free survival in patientswith glioblastoma but does not improve overall survival. To addressthis discrepancy, we investigated immunologic resistance mechan-isms to anti-VEGF therapy in glioma models. A screening ofimmune-associated alterations in tumors after anti-VEGF treat-ment revealed a dose-dependent upregulation of regulatory T-cell(Treg) signature genes. Enhanced numbers of Tregs were observedin spleens of tumor-bearing mice and later in tumors after anti-VEGF treatment. Elimination of Tregs with CD25 blockade beforeanti-VEGF treatment restored IFNg production from CD8þ T cellsand improved antitumor response from anti-VEGF therapy. Thetreated tumors overexpressed the glutamate/cystine antiporter

SLC7A11/ xCT that led to elevated extracellular glutamate in thesetumors. Glutamate promoted Treg proliferation, activation, sup-pressive function, and metabotropic glutamate receptor 1(mGlutR1) expression. We propose that VEGF blockade coupledwith glioma-derived glutamate induces systemic and intratumoralimmunosuppression by promoting Treg overrepresentation andfunction, which can be pre-emptively overcome through Tregdepletion for enhanced antitumor effects.

Significance: Resistance toVEGF therapy in glioblastoma is drivenby upregulation of Tregs, combined blockade of VEGF, and Tregsmay provide an additive antitumor effect for treating glioblastoma.

IntroductionVEGF and its receptor (VEGFR) have been identified as critical

mediators of angiogenesis (1). Interaction ofVEGFwith three differenttyrosine kinase (TK) receptors (VEGFR1–3) plays a significant role inphysiologic and pathologic angiogenesis, including tumor angiogen-esis (2). Evidence suggests that the blockade of VEGF (bevacizumab, ahumanized anti–VEGFA mAb) or VEGFR2 (ramucirumab) reducestumor-induced angiogenesis. Therefore, inhibition of this pathwaywas thought to be a promising treatment modality for patients withhighly vascularized cancers, such as glioblastoma (GBM; ref. 3).However, although several clinical reports using bevacizumab, eitheralone or in combination with other therapeutic approaches, haveshown prolonged recurrence-free survival (RFS), it has failed toimprove overall survival outcomes in patients with newly diagnosedGBM (4, 5). This pattern is consistent across tumors. These counter-intuitive results suggest that certain critical suppressive pathwaysmight become activated and thus jeopardize the antitumor response

mediated by VEGF inhibition. Previous reports have demonstratedthatVEGFblockade results in vascular remodeling, reducing perfusionand increasing hypoxia inside treated tumors (6, 7). A dramaticincrease in tumor cell invasion into the normal brain has also beenobserved (8). Despite substantial efforts to understand the underlyingmechanism of VEGF blockade treatment in these vasculature-richtumors, such as GBM, many questions still need to be answered.

The GBM tumor microenvironment is an intricate network (9)composed of dozens of different cellular and soluble components, suchas tumor cells, vascular endothelial cells, neural precursor cells, tumorstem cells, stromal cells, residential microglia, infiltrating immunecells, cytokines, and extracellular matrix proteins. Among thosecomponents, the vascular endothelial cells and infiltrating immunecells play crucial roles in manipulating the tumor landscape andaffecting tumor progression (10–12). Our recent studies on humangliomas suggest that the interconnection between glioma angiogenesisand intratumoral CD4þ and Foxp3þ T cells influences tumor pro-gression in patients with glioma. Activated CD4þ T cells were signif-icantly elevated in bevacizumab-resistant recurrent tumors. Regula-tory T cells (Treg)were determined to be an independent risk factor forprogression in that patient population (12). Thus, we hypothesizedthat inhibition of the VEGF/VEGFR pathway might negatively affectthe antitumor immunity that contributes to treatment failure.

In this study, our hypothesis was tested in glioma models using twoVEGF blocking agents (anti-VEGFA or anti-VEGFR2).We found thatVEGF inhibition enhances Treg-suppressive functions by dysregulat-ing the glutamate/cystine antiporter, and that combinatorial use ofTreg blockade can overcome this effect.

Materials and MethodsMurine glioma lines

Murine glioma cell lines, GL-261 or KR-158B, were transducedwith firefly luciferase plasmid pLenti CMV Puro LUC, gifted from

1The Fourth Section ofDepartment ofNeurosurgery, The First AffiliatedHospital,Harbin Medical University, Harbin, China. 2Lillian S. Wells Department of Neu-rosurgery, University of Florida, Gainesville, Florida. 3Preston A.Wells, Jr. Centerfor Brain Tumor Therapy, University of Florida, Gainesville, Florida.

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

CorrespondingAuthors:Zhiguo Lin, The First AffiliatedHospital, HarbinMedicalUniversity, Harbin, China. E-mail: linzhiguo@hotmail.com; and Jianping Huang,University of Florida, P.O. Box 100265, 1149 Newell Drive, Gainesville, FL 32610.Phone: 352-273-6835; Fax: 352-392-8413; E-mail:jianping.huang@neurosurgery.ufl.edu

Cancer Res 2020;80:499–509

doi: 10.1158/0008-5472.CAN-19-1577

�2019 American Association for Cancer Research.

AACRJournals.org | 499

on September 7, 2021. © 2020 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst November 13, 2019; DOI: 10.1158/0008-5472.CAN-19-1577

Eric Campeau (Zenith Epigenetics Corp.; Addgene plasmid#17477). After the lentiviral transfection and puromycin(631305, ClonTech) selection, the cells were cultured in a singleclone by limiting dilution. The selected cells were then expanded ina DMEM supplemented with 10% FBS, penicillin (100 U/mL), andstreptomycin (100 mg/mL; 15140163, Thermo Fisher Scientific),and kept in a 37�C incubator with 5% CO2. All the cell lines used inthis study are Mycoplasma negative tested using PlasmoTestMycoplasma Detection Kit (InvivoGen).

GL261-Luc and KR158B-Luc brain tumor mouse modelsFemale C57BL/6 mice (6- to 8-week-old) were anesthetized, and

5 � 104 GL261-Luc or KR158B-Luc tumor cells in a 2.5 mL volumewere intracranially injected at 2.0-mm lateral to the bregma at a depthof 3.0 mm below the dura mater with a sterile Hamilton syringe fittedwith a 25-gauge needle. The tumor-bearing mice were monitored byIVIS imaging on day 10 after implantation, andmice were randomizedinto treatment groups. The average tumor luciferase intensities wereapproximately equal among groups. Humanitarian endpoints (also astreatment endpoints) were reached when animals exhibited the fol-lowing deficits: (i) reluctant to move, (ii) weight loss >20% bodyweight, (iii) hunched posture, and (iv) lethargy.

In vivo neutralizing and depleting antibody treatmentsThe doses of antibodies used in this study were based on previous

reports (13–16). For tumor-bearing mice, the anti-VEGFA antibody(B20-4.1.1, Genentech) was administered by intraperitoneal injectionat 50 mg, 100 mg, or 200 mg/dose. VEGFR2 mAb (clone DC101,BE0060, BioXCell) was administered intraperitoneally at 800 mg/doseand the PD-L1 mAb (clone 10F.9G2, BE0101, BioXCell) was usedintraperitoneally at 200 mg/dose. For nontumor-burdened mice, theVEGFR2 antibody was administered intraperitoneally at 800 mg/doseevery 3 days for a total of 4 doses. Treg depletion was achievedintraperitoneally. Furthermore, 200 mg of anti-CD25 (clone PC-61.5.3, BE0012, BioXCell) was injected on days 10 and 12 after tumorimplantation (17).

Isolation of TILsThe tumor-infiltrating lymphocytes (TIL) isolation method has

previously been described (18). The brains were removed after PBSperfusion and digested with collagenase (10103578001, Sigma-Aldrich)/DNase-I digestive enzymes (LS002007, Worthington Bio-chemical Corp.). TILs were collected from 37%, and 70% of theinterface of Percoll (17-5445-01, GE Healthcare) gradient, centrifugedat 500 � g for 20 minutes and washed twice.

Fluorescence-activated cell sorting and analysisFluorescence-activated cell sorting (FACS) was performed using

LSRII (BDBiosciences). The viability of TILs was determined using theLive/Dead Cell Violet Viability kit (L34964, Thermo Fisher Scientific).Cell surface staining antibodies, anti-CD3 (145-2C11), anti-CD4(RM4-5), anti-CD8 (53-6.7cd), anti-CD25 (7D4), anti-CD45 (30-F11), anti–LAG-3(C9B7W), anti-CD69 (H1.2F3), and anti-CD154(MR1), were purchased from BD Biosciences. Anti-PD-1 (J43) andanti-Tim3 (RMT3-23) were purchased from eBioscience and BioLe-gend, respectively. The levels of Foxp3þ and IFNgþ T cells weredetermined by intracellular staining using a Foxp3/TranscriptionFactor Staining Buffer Set, anti-IFNg (XMG1.2), and anti-Fop3(FJK-16s; all from eBioscience). Appropriate isotype controls wereused.All FACSdatawere analyzedusing FlowJo 10.3 software (FlowJo).

RNA sequencing and data analysisFresh frozen normal brain or tumor tissues from healthy mice and

GL261-Luc–bearing mice treated with different doses of anti-VEGFA(50, 100, and 200 mg/dose) were sent to Novogene Corporation forRNA sequencing (RNA-seq). The raw data have been deposited inNCBI's Gene Expression Omnibus (GEO; ref. 19), accessible throughGEO Series accession number GSE107423 (https://www.ncbi.nlm.nih.gov/geo/). Paired-end reads were mapped to muscles UCSC mm10and assembled using STAR aligner and Cufflinks Pipeline 2 using theIllumina base space sequence hub (https://basespace.illumina.com/) asthe default setup. Each transcript expression was estimated as anFPKM (fragments per kilobase million) value. Subsequent two-dimensional principal component analysis (2D PCA) was performedusing Subio Platform (Subio). The signature scores of Treg andexhausted CD8þ T cells were evaluated using single sample gene setenrichment analysis (ssGSEA; ref. 20)with gene sets (GSE7852Treg vs.Tconv up, and GSE9650 exhausted vs. memory CD8þ T-cell up,respectively).

Real-time quantitative reverse transcription-PCRTotal RNA was extracted from Tregs cultured in vitro and from

tumor tissue of GL261-Luc-bearing mice with or without VEGFRblockade treatment and TaqMan gene expression assays (SLC7A11,SLC3A2, Ki67, and r18s) were purchased from Thermo Fisher Scien-tific. In addition, r18s was used for the normalization. Relativequantification of the gene expression was carried out via the doubledelta Ct analysis.

Glutamate concentration assayGL261-Luc cells (2.0� 105/well) were cultured in a 6-well plate with

or without 100 mmol/L of CoCl2 (C8661, Sigma-Aldrich). After48 hours, glutamate concentrations in the medium were measuredusing a glutamate assay kit (ab83389, Abcam) according to themanufacturer's instructions. The tumor pieces were resuspended in500 mL of PBS and the supernatant was collected after centrifuging.The glutamate concentration of the supernatant was measured usingthe kit described above.

In vitro GL261-Luc cells xCT expressionGL261-Luc cells (2.0 � 105/well) were cultured in a 6-well plate

according to the following conditions: (i) the cells were treated with orwithout 100mmol/L of CoCl2 (C8661, Sigma-Aldrich), and (ii) the cellswere treated with or without 20 mg/mL VEGFR2 blockade (cloneDC101, BE0060, BioXCell). After 48 hours, the xCT expression on thecells was determined using the xCT antibody (NB300-318, NovusBiologicals) by flow cytometry.

IHC of xCTTumor tissues were resected and frozen, and 5-mm sections were

put on slides. The slides were incubated with an anti-xCT antibody(1:100; NB300-317, Novus Biologicals). Goat anti-rabbit IgG (1:100;A-11011, Alexa Fluor 568, Thermo Fisher Scientific) was used as thesecondary antibody. Images were taken using Olympus U-HGLGPSmicroscopes.

Isolation of CD4þCD25þ Tregs and CD4þCD25� T conventionalcells

Both CD4þCD25þ Tregs and CD4þCD25� Tconvs from spleens ofC57Bl/6 mice were isolated using the Treg isolation kit (130-091-041,Miltenyi Biotec), according to the manufacturer's instructions.

Long et al.

Cancer Res; 80(3) February 1, 2020 CANCER RESEARCH500

on September 7, 2021. © 2020 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst November 13, 2019; DOI: 10.1158/0008-5472.CAN-19-1577

Treg suppression assayViolet (C34557, Invitrogen) labeled Tconv cells (3� 104/well) were

cultured with Tregs at different ratios (1: 1, 1: 0.5, 1: 0.25, and 1: 0.125)in 96-well round-bottom plates in the presence of a T-cell mediumcontaining 0, 175, or 350 mmol/L of L-glutamate (G8415, Sigma-Aldrich). Cells were then stimulated with Dynabeads MouseT-Activator CD3/CD28 (3 � 104/well) in the absence of IL2 for3 days, and a FACS analysis was performed to measure theCellTrace violet dilution. The T-cell medium was Minimum Essen-tial Medium without L-glutamate and was supplemented with anadditional 10% FBS. For the Treg and Tconv proliferation exper-iment, 1 � 105 Tregs and Tconv cells per well were stimulated withDynabeads Mouse T-Activator at a 1:1 cell to bead ratio in theabsence of IL2, under different concentrations of L-glutamate (0,175, or 350 mmol/L), in 96-well round-bottom plates. The absolutecell counts were determined by hemocytometer.

Western blotsTregs and Tconv cells were lysed in RIPA buffer (PI89900, Thermo

Fisher Scientific) containing protease/phosphatase inhibitor cocktail(PI78442, Thermo Fisher Scientific). Protein concentration was mea-sured using a BCA assay (23227, Thermo Fisher Scientific). Anti-GluR1 (NBP1-82396, Novus Biologicals) was used as the primaryantibody, and b-actin (4970S, Cell Signaling Technology) was used asthe loading control. Blots were detected automatically using the SimpleWestern System (Wes, Bio-techne).

Statistical analysisData are presented as mean � SEM. Mann–Whitney U tests

were used to determine significant differences between twogroups. A one-way ANOVA was used to compare more thantwo groups. Survival was analyzed using Kaplan–Meier survivalcurves and log-rank Mantel–Cox tests. Statistical analyses wereperformed using GraphPad Prism 7 software. P values of lessthan 0.05 were considered significant (�, P < 0.05; ��, P < 0.01;���, P < 0.001).

Study approvalIn our murine experiments, mice were handled in accordance with

the animal care policy at the University of Florida (UF); all protocolswere approved by UF's Institutional Animal Care and Use Committee(IACUC). IACUC#201609208: The combination therapies for gliomas(principal investigator: J. Huang).

ResultsVEGF blockade alters tumor immune landscape

To gain insight into the impact of VEGF blockade on the immu-nologic landscape of gliomas, we treated C57BL/6 mice bearingmalignant gliomas with various amounts of anti-VEGFA antibodyuntil the treatment endpoint was reached (Fig. 1A). The antitumorresponse was found to be dose dependent; a higher dose resulted inlonger survival benefits as reported previously (21). However, allanimals succumbed to their tumors before 50 days post tumorimplantation. We resected tumor samples from the mice near thedate of median survival from each group (12 total), and tissues from 3normal brains were also included. RNA samples were sent forsequencing analysis. The data have been deposited in NCBI's GeneExpression Omnibus (19), accessible using GEO Series accessionnumber GSE107423 (https://www.ncbi.nlm.nih.gov/geo). Next, weassessed the major pathways enriched by VEGF treatment using the

data obtained from the RNA-seq analysis. Reactome pathway enrich-ment was first employed to evaluate overall genetic landscape changespost anti-VEGF therapy. The results indicate that 5 of the top 20enriched pathways were immune-related (Fig. 1B). Furthermore, weevaluated the previously published microarray analysis of chemo-na€�ve patients with GBM or patients who had failed bevacizumabtreatment (22–24). The results indicate that the bevacizumab-resistanttumors represent nearly identical genetic changes to those we observedin the mouse model. More immune-related pathways were revealedcompared with the chemo-na€�ve tumors (Supplementary Fig. S1A andS1B). These results suggest that VEGF blockade may alter the GBMimmune landscape. To test our hypothesis that anti-VEGF therapyalters immunologic infrastructure, a 750 immune-related geneset, acommercially available program (nCounter PanCancer Immune Pro-filing Panel, NanoString Company) was used. This panel consists of770 genes (750 immune genes and 20 internal reference genes) from 24different immune cell types, checkpoints, CT antigens, and genesspanning the adaptive and innate immune response network. Adose-dependent separation/cluster was observed in the anti-VEGFAtreatment using 2D PCA, whereas the normal brains were distinctlyseparated from samples treated with anti-VEGFA (Fig. 1C). Next,RNA-seq data were enriched using a ssGSEA for immunosuppressivesubpopulations (e.g., M1/M2 macrophages and Tregs) based onprevious reports that VEGFR2 blockade combined with Ang-2 canalter the polarization of macrophages (25). We did not find changesin the M1:M2 ratio after treatment with VEGF blockade alone.However, we did notice a dose-dependent increase in Tregs andexhausted CD8 T-cell gene scores after the anti-VEGFA treatment(Fig. 1D and E). Collectively, anti-VEGF therapy is associated with thetumor-immunosuppressive genetic landscape in a dose-dependentmanner.

Enrichment of Tregs in spleens and tumors post anti-VEGFtreatment

We treated the same tumor-bearing mice with another VEGFblockade agent (due to supply-discontinuation of the anti-VEGFA),anti-VEGFR2 (800 mg/dose, i.p. every 3 days until the endpoint,which is equivalent to 100 mg/dose of the anti-VEGFA regardingmedian survival, Supplementary Fig. S2A and S2B). Tumors were,respectively, harvested on D10, D17, and D25 after continuous anti-VEGFR2 administration (Fig. 2A). The tumor size was found to beequivalent to the tumor size in the nontreated (NT) group on D17.Flow cytometry analysis of TILs was then performed, and thegating strategy is shown (Fig. 2B). The frequencies of Tregs(CD4þFoxp3þ), Tconv (CD4þFoxp3�), CD8þ, and other cells(CD3þCD4�CD8� T cells), with or without the treatment in spleenand tumor are shown in pie charts (mean) with statistical signif-icance indicated. On D17 posttumor implantation, the percentageof Tregs in the spleens was approximately 1.7 times greater in aVEGFR2-treated animals compared with controls, but no similarchange was seen in spleens of healthy mice with the same treatmentschedule and dose (Fig. 2C; Supplementary Fig. S3A–S3E). Intumors, the Tregs fraction was decreased by approximately 25%at D17, but then almost tripled by D25 to reach 35.5% (Fig. 2D).Because of the prolonged survival of mice in the anti-VEGF–treatedgroup, the significant increase in the tumor-infiltrating Tregs at alater time might be a result of a cumulative effect. To address thisissue we then included a control group (anti-PD-L1) that provides asimilar length of survival as the anti-VEGFR2 treatment (Fig. 2E).A significant increase in Treg frequency, Tregs/CD8þ T-cell ratio,and Tregs/Tconv ratio were detected in the anti-VEGFR2–treated

Anti-VEGF Therapy Promotes Treg Function

AACRJournals.org Cancer Res; 80(3) February 1, 2020 501

on September 7, 2021. © 2020 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst November 13, 2019; DOI: 10.1158/0008-5472.CAN-19-1577

tumors on D25, compared with the anti-PD-L1–treated tumorssampled at the same time point. No significant increase wasobserved in these parameters between the NT (D17, the endpoint)and the anti–PD-L1 (D25) groups (Fig. 2F). These data led usto predict that these differences between anti-VEGFR2 and anti-PD-L1 groups (�>2-fold higher) were the consequence ofanti-VEGFR2 treatment. To replicate the finding we observed inthe GL-261 tumormodel, a second gliomamodel was tested. KR-158Bis an invasive murine glioma line that was derived from spontaneousgliomas in Trp53/Nf1 double-mutant mice in the C57BL/6 back-ground. Similar results were found in this model (SupplementaryFig. S4A–S4G). These results suggest that the tumor plays a criticalrole in the anti-VEGF–associated Treg accumulation.

Blocking Tregs using anti-CD25 antibody before treatment withVEGF therapy restores antitumor activity mediated by anti-VEGFR2 treatment

To confirm whether Tregs are the predominant factor limiting theefficacy of VEGF blockade, we sought to eliminate Tregs throughCD25-targeted inhibition before anti-VEGFR2 treatment. Two dosesof anti-CD25 antibody were administered on day 10 and 12, followedby the anti-VEGFR2 treatments every 3 days until the treatmentendpoint (Fig. 3A). The animals showed a moderate improvementin survival using anti-VEGFR2 and anti-CD25 antibody as a mono-therapy (compared with untreated animals). A significant benefit wasachieved in the combination group (Fig. 3B).We then evaluated CD25expression on CD4þ T cells and found that the anti-CD25 antibody

0

15

30

Ent

ities

foun

d Immune-related pathwaysOther pathways

200 μg/0 μg

Upregulated525 genes

0 µg

50 µg

100 µg

200 µg Tum

or re

sect

ion

and

gene

ana

lysi

s

VEG

FA m

Ab

ever

y 3

days

10 d

ays

NBs0 µg50 µg100 µg200 µg

PC1 (93.5%)

PC

2 (4

.3%

)

A

B

C D E

Tumor i.c.

-4,000

-2,000

0

2,000

4,000

Sig

natu

re s

core

Treg

0 μ

g

50

μg

10

0 μ

g

20

0 μ

g

NB

s

0 μ

g5

0 μ

g1

00

μg

20

0 μ

g

NB

s

0 μ

g5

0 μ

g1

00

μg

20

0 μ

g

NB

s

−2,000

0

2,000

Treg signature score

-4,000

-2,000

0

2,000

4,000

CD

8E/M

sco

re

P < 0.001P < 0.001

Figure 1.

Anti-VEGFA treatment alters the tumor-immune genetic landscape and enhances Treg signature genes. A, Treatment schedule of anti-VEGFA treatment. C57BL/6micewere intracranially inoculatedwithGL-261-LuciferaseGBMcells, and after 10days, the tumor engraftmentwas confirmedwith IVIS imagingof luminescence, andonly tumor-bearingmicewere randomly divided into 4 groups (9 per group). The animalswere then treatedwith the anti–VEGFA antibody (0, 50, 100, or 200 mg perdose, respectively) every 3 days until they reached the endpoint. B, Pathway enrichment was performed for all genes (total 525 genes) that showed a significantelevation between 0 and 200 mg per dose groups (fold increase >2; P < 0.05) using the Reactome pathway database (murine) based on the ranking of the entitiesfound (https://reactome.org/). Blue, immune-related pathways. C, The RNA-seq results from all 5 groups were also enriched against a 750 immune (CounterImmunology Panel for Mouse)-related geneset. Clusters were circled for different treatment groups, including the normal brain (NB) group. D, Increase in Tregsignature genes after the anti-VEGFA treatment. TheRNA-seqdatawere enrichedusing theTreg toTconv (GSE7852_treg_vs_tconv_up)gene set providedbyGSEA(https://software.broadinstitute.org/software/cprg/). A score was given by GSEA for each tumor sample RNA-seq. A heat map (left) and the comparisons of thescores (means� SEM) among the groups (right) are shown. E, A dose-dependent increase in exhausted to memory (E/M) CD8 signature gene scores using an E/MCD8 (GSE9650_ exhausted_vs_memory _cd8_tcell_up) geneset. An one-way ANOVA was used to determine the significant differences between the means.

Long et al.

Cancer Res; 80(3) February 1, 2020 CANCER RESEARCH502

on September 7, 2021. © 2020 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst November 13, 2019; DOI: 10.1158/0008-5472.CAN-19-1577

eliminates CD25 expression on TILs at D25, which potentially was dueto the competition of the blocking and detecting antibodies, as theefficiency of the CD25 blockade is 56% to 77% (26). Nevertheless, theseresults indicate efficient effects of controlling CD25 on CD4þ T cells(no CD8þCD25þ T cells were detected) by the blockade in vivo(Fig. 3C). IFNg production was thenmeasured fromCD4þ and CD8þ

TILs after a rechallenge with anti-CD3/CD28 beads (Fig. 3D). Inter-estingly, both anti-CD25 (green dots) and anti-VEGFR2 (red dots),which were used individually, significantly increased IFNg production

from both CD4þ and CD8þ T cells that may be responsible for themoderate survival benefit shown in Fig. 3B. The anti-VEGFR2treatment alone produced relatively less cytokine from both T-cellsubsets compared with the anti-CD25. When a combinatory strategywas employed, the IFNg production was only restored (an additiveeffect, red dots vs. blue dots) in the CD8þ cells, not the CD4þ cells. Thedata also suggest that the prolonged survival in the combinationtreatment may rely on enhanced CD8þ T-cell function. In summary,Tregs play a critical role in suppressing the antitumor effects of

Figure 2.

An enrichment of Tregs in spleen and tumors by anti-VEGFA or anti-VEGFR2 treatment. A, The animal experiment designed to test the tumor-infiltrating Tregs andother T-cell subsets using anti-VEGFR2 antibody in vivo. Six- to 8-week-oldC57BL/6mice (7–8mice/group)were intracranially inoculatedwith 4� 104 ofGL-261-Luctumor cells on day 0 (D0); the IVIS images were taken on day 10 (D10) to confirm the tumor engraftment. The tumor-burdenedmice were then randomly separatedinto twogroups, followedby the administration of either saline only (nontreated group, NT) or anti-VEGFR2 (800mg/dose) antibody every 3 days until the treatmentendpoint. Tumors were, respectively, harvested on day 10, day 17 (the NT group endpoint), and day 25 (the treated group endpoint) post-intracranial administration,and the sample collection and analysiswere finishedwithin 3 days after these dates.B, The gating strategy.C andD, Posttreatment kinetic analysis of the percentageof Tregs, Tconv, CD8þ T, and others (CD3þCD4�CD8�) cells in the spleen (C) and tumor (D) summarized in the pie (mean) chart, with statistical analysis. E,Experimental design. Three groups of GL-261-Luc tumor-burdened mice (10 mice/group) were treated on day 10 after the intracranial administration with eithersaline, anti-VEGFR2 (800mg/dose), or anti-PD-L1 (200mg/dose), every 3days until the treatment endpoints. The tumorswere resected andanalyzed at day 17 for theNT groups, and day 25 for both anti-VEGFR2 and anti-PD-L1 groups. The survival curves and proportions of indicated T-cell subsets in those resected tumors areshown. F, Intratumoral percentage of Tregs, Tregs/CD8þ T cells, or Tregs/Tconv ratio is shown. The data represent mean � SEM, the survivals are shown usingKaplan–Meier survival curves, and the log-rank test was used to determine the significance. The data are representative of 2–3 independent experiments. A Mann–Whitney U test was used to determine the significant difference between the two groups. � , P < 0.05; �� , P < 0.01; ��� , P < 0.001.

Anti-VEGF Therapy Promotes Treg Function

AACRJournals.org Cancer Res; 80(3) February 1, 2020 503

on September 7, 2021. © 2020 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst November 13, 2019; DOI: 10.1158/0008-5472.CAN-19-1577

anti-VEGF treatment, which can be overcome through targetedinhibition of Tregs.

Elevation of SLC7A11 gene expression after anti-VEGFtreatment is associated with increased intratumoral Tregs

When looking for the increased expression of genes that overlapamong the three doses of anti-VEGFA–treated tumors, three geneswere revealed: SLC7A11, Gpr141, and Cth (Fig. 4A and B). UsingqPCR, we confirmed that the gene expression of SLC7A11 and thexCT-binding partner CD98 (SLC3A2) were increased with anti-VEGFR2 treatment (Fig. 4C and D). The protein expression of xCTwas higher in treated tumors than nontreated tumors (Fig. 4E).Because anti-VEGF therapy leads to a more hypoxic tumor micro-environment (8), we evaluated the expression of the hypoxia-relatedgene, HIF1a. A dose-dependent increase in HIF1a gene expressionwas observed (Fig. 4F). We then evaluated whether conditionalhypoxia in vitro could alter the xCT expression using cobalt(II)chloride hexahydrate (CoCl2; ref. 27). The result demonstrates thatCoCl2 increases tumor xCT expression and glutamate production.However, no direct effect of anti-VEGFR2 on xCT expression was seenwhen the tumor cells were cultured with the anti-VEGF (Fig. 4G–I),suggesting an indirect influence instead. Finally, a positive correlationbetween SLC7A11 gene expression and Treg signature scores was

determined in these treatment animals and the mesenchymal sub-group of GBM (Fig. 4J; Supplementary Fig. S5), which has beendemonstrated to have more immune infiltrations and higher levelsof the endothelial markers CD31 and VEGFR2 than othersubgroups (28–30). These data imply that a dysfunctional gluta-mate/cystine antiporter in tumorsmay influence the function of Tregs.

Glutamate promotes Tregs function in vitroPrevious studies show that glutamate plays a central role in the

malignant progression of glioma via numerous mechanisms (31). Wefound in this study that anti-VEGF can enhance the expression ofglutamate/cystine antiporter, which leads to more glutamate produc-tion by the tumors. Thus, we performed a Treg-suppression assay inthe presence of L-glutamate in vitro. The results demonstrate a dose-dependent inhibition of Tconv cell proliferation (Fig. 5A–C). Asignificant decrease in the activation of Tconv cells, measured byCD69 expression on the population, was observed (Fig. 5D). Mean-while, phenotypic analysis of Tregs revealed that adding glutamateincreases the surface expression of early and late T-cell activationmarkers CD69 and CD154, and Ki67 in a dose-dependent manner(Fig. 5E-G). When glutamate was added to Treg- or Tconv-onlycultures, the total cell counts of Tregs increased, and no change wasobserved in the Tconv cells (Fig. 5H and I), suggesting that the activity

10 15 20 25 30 35 400

50

100

Days a�er tumor implanta�onPerc

enta

ge o

f sur

viva

l

NTaCD25aVEGFR2aVEGFR2/CD25

***

**Day 0 Endpoint

A

CD8

4.1

22.4

17.1

35.6

D

B

% C

D4+C

D25+

NT

αCD2

5 αV

EGFR

2αV

EGFR

2/CD

25

NT

αCD2

5αV

EGFR

2αV

EGFR

2/CD

25

0

20

40

60

80

*********

CD4 CD

25

38.6

0.7

56.6

1.8

C

IFN

γ

% C

D4+

IFγ+

0

5

10

15

20

25

****

***

**

CD8+ IF

Nγ+

0

20

40

60**

*

*

CD4 6.4 IF

Nγ

10.8

19.8

2.6 Tumor i.c.

Figure 3.

Blocking Tregs using an anti-CD25 antibody during the early phase of anti-VEGF therapy restores some of the antitumor reactivity mediated by the anti-VEGFR2.A, Treatment strategy. Four groups (n¼ 7) of 6-to 8-week-old C57BL/6micewere intracranially inoculatedwith GL-261-Luc, and 10 days after the tumor intracranial,two doses of the anti-CD25 antibody (200 mg/dose) were given intraperitoneally to the tumor-burdenedmice on D10 and D12. The anti-VEGFR2 (800 mg/dose) wasstarted on day 11 (D11) and continued every 3 days until the treatment endpoint. B, The survival of the differentially treated mice was plotted using the Kaplan–Meiercurve, and the log-rank test determined a significant difference. C, The frequency of CD4þCD25þ T cells in the TILs of mice treated with or without the anti-VEGFR2.The TILs were collected from tumor samples at the endpoints, and phenotypical analysis was carried out by sequentially gating live cells, CD45þ, and CD3þ T cells.A representative flow cytometry analysis (left) and summary (right) are shown. D, Adding anti-CD25 therapy before the continuous treatment of anti-VEGFR2restores the IFNg production of T cells. The tumorswere resected from themice at each endpoint, and TILswere isolated and challenged using anti CD3/CD28 beads(cell to bead ratio¼ 1:1). The percentages of CD4þ IFNgþ and CD8þIFNgþ T cells in TILs weremeasured using flow cytometry, and the representative flow cytometrydata (left) and calculated results (right) are shown. Each color represents the corresponding group (NT, black; anti-CD25, green; anti-VEGFR2, red; and combination,blue). The experiments were repeated twice. A Mann–Whitney U test was used to determine the significant difference between the two groups. The graphs show asum of two repeated experiments. � , P < 0.05; �� , P < 0.01; ��� , P < 0.001.

Long et al.

Cancer Res; 80(3) February 1, 2020 CANCER RESEARCH504

on September 7, 2021. © 2020 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst November 13, 2019; DOI: 10.1158/0008-5472.CAN-19-1577

of glutamate in T cells is limited to the Treg compartment. To testwhether these two cell populations expressed glutamate receptors,protein expression of metabotropic and ionotropic glutamate recep-tors (mGluR1, 4, and AMPA3; ref. 32) on Tregs and Tconv cells werethen evaluated. Only mGluR1 was detected, with a dose-dependentenhancement of this receptor corresponding to glutamate exposure inboth cell types, and with the Tregs showing a more dramatic enhance-ment compared with the Tconv cells (Fig. 5J). These results suggestthat glutamate promotes an activated Treg phenotype and function.

DiscussionMounting evidence has shown that tumor angiogenesis and immu-

nosuppression are key features in refractory malignancies. VEGFsignaling has been involved in fostering immunosuppression (bothsystemically and intratumorally), which may be unlocked through theuse of antiangiogenic agents (10, 33–36). Blocking the VEGF/VEGFRpathway has been demonstrated to increase survival and decreasetumor immunosuppressive status in animal models, includingGBM (37, 38). However, VEGF blockade with bevacizumab prolongsonly progression-free survival (PFS), without significant enhancementof overall survival benefit in patients with GBM (4, 5). These findings

are likely multifactorial. Our data suggest that immune suppressivemechanisms contribute to anti-VEGF failure. Our recently publishedstudy on human gliomas showed that the level of Foxp3þ T cellslocated in the perivascular tumor niche correlates with tumor vascu-lature and is an independent predictor of shortened PFS (not overallsurvival) in astrocytic gliomas (12). Moreover, VEGFR2 has beenfound to be selectively expressed by Foxp3high CD4þ Tregs and can beincreased by TGFb (39), suggesting that the Treg population may bepreferentially affected by anti-VEGF treatment. These data suggestinterconnections between glioma angiogenesis and inhibitory immuneinfiltration. In this report, we provide an in-depth understanding ofanti-VEGF therapy resistance and a strategy for improving the ther-apeutic effect of the antiangiogenic agents in GBM.

We observed that at the beginning of the treatment, anti-VEGFshowed a moderate reduction of Tregs (4% lower than NT) andenhancement of antitumor response (increased INFg production fromboth CD4þ and CD8þ T cells, and more prolonged survival than NT),which reproduced the previous reports that anti-VEGF can help torestore the tumor immune microenvironment (37, 40, 41). However,the response could not be sustained despite continued administrationof the drug. Ultimately, the treatment failed and the mice succumbedto the tumor, similar to outcomes from human clinical trials. Thus, we

Figure 4.

A dose-dependent upregulation of the SLC7A11 gene and its associateswith Tregs after the anti-VEGFR2 treatment.A, Identification of genes that were significantlyupregulated by anti-VEGFA treatment. The RNA-seq data from the anti-VEGFA treatment experiment shown in Fig. 1Bwere analyzed. Overlapping genes that wereenriched (fold increase >2; P < 0.05), respectively, using the 0-mg group as a reference, from the 50, 100, or 200 mg group are presented using a Venn diagram, andoverlapped genes from the three comparisons are shown. B, A dose-dependent increase in SLC7A11 gene expression determined by FPKM was observed after theanti-VEGFA treatment. C and D, SLC7A11 and SLC3A2 gene expression was increased using anti-VEGFR2 therapy. The total RNA was isolated from tumor tissues atthe endpoints of 4 tumor-burdened mice in each group (the same treatment schedule as is described in Fig. 2A), and RT-qPCRs were performed. E, Anti-VEGFR2increases xCT protein expression in vivo. The same tumors described inC andDwere also evaluated by fluorescent IHC. The in vivo experiments were repeated threetimes. F,Adose-dependent increase appeared in HIF1a after the anti-VEGFA treatment. The HIF1a expression in a different dose of anti-VEGFA treatment describedin Fig. 1Bwas graphed. G, Hypoxia increases xCT expression on the tumor cells. The GL-261 cells were cultured under hypoxic conditions using 100 or 150 mmol/L ofCoCl2 for 48 hours in vitro, and the xCT expression on the cells was measured using flow cytometry. H, Anti-VEGFR2 does not change the xCT expression of tumorcells in vitro. The GL-261 cellswere cultured inmedium containing 20 mg/mL of the anti-VEGFR2 antibody for 48 hours (13), and the xCT expression on these cellswasassessed using flowcytometry. I,Glutamatewas induced using conditional hypoxia. TheGL-261 cellswere cultured inmediumwith orwithout 100mmol/L of CoCl2 for48 hours, and glutamate concentration in the culture supernatant was assessed using three individual experiments. J, SLC7A11 gene expressionwas highly correlatedwith Treg signature scores. An one-wayANOVAwasused todetermine the significance inB,F,G, andH. AMann–WhitneyU testwas used to determine the significantdifference in C, D, and I. The association between SLC7A11 and the Treg score was determined using Spearman rank correlation coefficient.

Anti-VEGF Therapy Promotes Treg Function

AACRJournals.org Cancer Res; 80(3) February 1, 2020 505

on September 7, 2021. © 2020 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst November 13, 2019; DOI: 10.1158/0008-5472.CAN-19-1577

sought to determine the immunologic manifestations of treatmentfailure and observed a link between treatment failure and Tregenrichment. We found a dose-dependent increase in Treg signaturegenes in the tumor after the anti-VEGF treatment and confirmed theoverrepresentation of these cells, compared with CD8þ T cells orCD4þ Tconv cells in spleens and tumors sequentially. These resultsimply that anti-VEGF treatment leads to a more immunosuppressivephenotype. Interestingly, the anti-VEGF had no similar impact onTregs in lymphoid tissues of healthy animals, suggesting that the tumorplays a critical role in the immunosuppression induced by anti-VEGFtherapy, that is, factors produced by glioma cellsmay be involved in the

promotion of Treg function. Previous animal studies have shown thatanti-VEGF inhibits Treg accumulation in mice 2 weeks after thetreatment (equivalent to the results we found on D17), and we provideevidence that this effect was short-lived, and enhanced intratumoralTregs were found in the tumors of failed animals. The VEGF blockadespecifically acts on the CD4 compartment by stimulating Tregs andsuppressing Tconv functions, which can be overcome through CD25blockade, a clinically available agent that inhibits Tregs but hasdemonstrated only modest therapeutic activity against establishedtumors (42, 43). Because of the potential suppression of activatedeffector cells by the CD25 blockade, we only administered two doses

0 200 4000.8

1.2

1.6

Ki67

0 200 400

2,200Tr

egs

CD15

4(M

FI)

L-glut (μmol/L) L-glut (μmol/L)

L-glut (μmol/L)L-glut (μmol/L)

L-glut (μmol/L)

(μmol/L)

L-glut (μmol/L)L-glut (μmol/L)

Tregs66.8%

Tconv.33.1 %

54.0 %

42.6 %

32.2 %1 μmol/L

1 μmol/L

175 μmol/L

350 μmol/L

175 μmol/L

350 μmol/L

Tregs

A

B

F

H83.3 %

84.9 %

84.5 %

0 200 40030

35

40

45

50

55

%Su

pres

sion

0 200 400600

650

700

CD69

MFI

0 200 400

P < 0.05 P < 0.01

P < 0.01 P < 0.01

P < 0.001

600

800

1,000

Treg

s CD6

9 (M

FI)

0 200 4001.0

1.5

2.0

2.54.0

5.5

Abso

lute

cou

nt(1

05 )

P < 0.01

TregsTconv

D

E

I

C

G

Cell trace violet

Cell

coun

ts

0 200 4000

50

100

% D

ivid

ed T

conv

JmGlutR1

β-Ac�n

β-Ac�n

1 175 350

mGlutR1

Tregs

Tconv

Figure 5.

Adose-dependent enhancement of Tregproliferation, activation, and suppressive function byglutamate in vitro.A,A representative flowcytometry analysis of Tregsfunctional analysis in the presence of L-glutamate in vitro. The Tregs (CD4þCD25þ) and Tconv cells (CD4þCD25�) were freshly isolated from the spleens (n¼ 5) of 6-to 8-week-old C57BL/6mice. CellTrace Violet–labeled Tconv cells were coculturedwith Tregs and stimulatedwith anti-CD3/CD28 beads in the presence of 1, 175 and350 mmol/L of L-glutamate, respectively. B and C, Glutamate dose-dependently enhances Tregs’ suppressive function. On day 3, the cocultured cells were analyzedusing flow cytometry, and the Treg suppressionwas assessed bymeasuring the nondivided population of the Tconv cells.D,Glutamate dose-dependently decreasesthe CD69 expression on the Tconv cells. The cocultured Tconv cells were also evaluated for the early activation marker, CD69, using flow cytometry. E and F,Glutamate enhances Tregs’ activation by upregulating the surface expression of CD69 and CD154. A flow cytometry analysis was carried out to assess the markerexpressions. G, Glutamate enhances Tregs’ proliferation. The Tregs (5 � 105/mL) were cultured in media containing different concentrations of L-glutamate in thepresence of anti-CD3/CD28 beads (n¼ 8); the cells were lysed, and qRT-PCRwas performed to determine the Ki67 gene expression.H,Glutamate does not directlyinfluence Tconv cell proliferation. The CellTrace Violet–labeled Tconv cells (1.5� 105/mL; n¼ 5) were stimulated with anti-CD3/CD28 beads and cultured in differentdoses of glutamate (in the absence of Tregs) for 3 days. The divided cell population was gated (left) and summarized data were graphed (right). I, Glutamate onlyincreases cell counts of Tregs, not of theTconv cells. The anti-CD3/CD28bead–stimulated Tregs andTconv cells (5� 105/mL)were individually culturedwith differentconcentrations of the L-glutamate, and the cell counts were carried out after 3 days (n ¼ 8). All the data are representative of three independent experiments, andvalues are means � SEM. An one-way ANOVA was used to determine the significance in each graph. J, Enhanced protein expression of metabotropic glutamatereceptor 1 (mGlutR1) on activated Tregs by L-glutamate in vitro. Tregs and Tconv cells were, respectively, isolated and activated by anti-CD3/CD28 beads in thepresence of an indicated amount of L-glutamate for 3 days. Western blot analysis was performed to assess the expression of mGlutR1, and b-actin was used as aloading control. The result was representative of two independent experiments.

Long et al.

Cancer Res; 80(3) February 1, 2020 CANCER RESEARCH506

on September 7, 2021. © 2020 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst November 13, 2019; DOI: 10.1158/0008-5472.CAN-19-1577

before anti-VEGF treatment. IFNg-producing CD4þ T cells wereenhanced by anti-CD25 monotherapy, but this enhancement wasdiminished by combination with anti-VEGFR2 (to a similar level asthe anti-VEGFR2 monotherapy). These data confirm our finding thatthe anti-VEGFR2 treatment mostly impacts CD4þ T cells, offsettingthe benefit of Treg depletion in this compartment. In contrast, IFNg-producing CD8þ T cells were enhanced by the combination, whichmay be responsible for the prolonged survival.

Next, the identification of SLC7A11 gene overexpression with anti-VEGF treatment led us to link between anti-VEGF therapy and itsimpact on CD4þ T cells. It is well known that this gene encodes acrucial glutamate/cystine antiporter, xCT. Increased production of thisantiporter has been implicated in enhanced glutamate production inother cancer types (44, 45). In the central nervous system (CNS), xCTcontrols the exchange between glutamate and cystine. Overexpressionof xCT elevates glutamate in the extracellular space, promoting tumorprogression, and drug resistance. The knockdown of xCT has beenshown to decrease glutamate production in cancer cells (44, 46, 47).Under physiologic conditions, glutamate is an excitatory neurotrans-mitter that is pivotal for the proper functioning of the CNS (48);however, under pathophysiologic conditions (e.g., autoimmune dis-eases, neurodegenerative diseases, and cancers), this amino acid hasbeen found to be excessively produced with leakage into the extra-cellular matrix, where it can act as a neurotoxin that damages neuralcells (49, 50). Under hypoxic conditions, gliomas have been shown to

overexpress xCT (51), and pathophysiologic glutamate has beendetected in the extracellular space. A result of an analysis of 9 patientswith glioma indicates that glutamate concentrations were 100-foldhigher in the peritumoral versus nontumoral zones, which can drivemalignancy by promoting proliferation and invasion (52). Highglutamate can also cause glioma-associated seizures (53, 54). Theanti-VEGF treatment magnifies this effect by further elevating gluta-mate production, which potentially leads to increased tumor cellinvasion (8), and fatal seizures. We found that the effect of anti-VEGF on xCT overexpression was not direct, and intermediaryfactor(s) may be involved. Because it has been demonstrated thatanti-VEGF induces hypoxia, which elicits HIFdependent expressionof SLC7A11/ xCT (55), we hypothesize that the hypoxic tumormicroenvironment induced by the anti-VEGF therapy may be oneof the pathways involved in the dysfunction of glutamate/cystineantiporter. As a consequence, glutamate was overproduced by thetumor, promoting the suppressive functions of Tregs and resulting intreatment failure (Fig. 6). More studies focusing on the mechanismproposed in this study are necessary.

Excessive glutamate can act as a neurotoxin and can triggerimmunologic involvement in the CNS, as a “protective autoim-munity mechanism” (56, 57). This mechanism has been describedas a beneficial action for the CNS to protect astrocytes fromneurotoxin attacks under pathophysiologic conditions (57, 58).Gliomas may adopt this mechanism to protect themselves and in

Figure 6.

Schema: anti-VEGF therapy enhancesTregs presentation and activity in gli-oma, dampening the antitumor effica-cy of the treatment. Dysfunctionalglutamate/cystine antiporter, SLC7A11/xCT, induced by the therapy, enhancesthe suppressive functionof theTregsviaelevated tumor-derived glutamate.

Anti-VEGF Therapy Promotes Treg Function

AACRJournals.org Cancer Res; 80(3) February 1, 2020 507

on September 7, 2021. © 2020 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst November 13, 2019; DOI: 10.1158/0008-5472.CAN-19-1577

doing so, turn the neurotoxin into an immunosuppressive inducer(i.e., the elevated glutamate levels in vivo may act in a paracrinemanner to activate glutamate receptors on the Tregs and enhancetheir immunosuppressive function). A potential solution might beto remove the excessive glutamate using an xCT inhibitor orthrough glutamate starvation. It is known that cystine starvationcan potentially induce xCT expression as an adaptiveresponse (59, 60). However, to our knowledge, there have beenno prior studies examining the effect of glutamate starvation onxCT expression. Investigating whether xCT loss of functionrestores susceptibility to anti-VEGF therapy will be an importantstudy for further determination. Nevertheless, because glutamateplays a crucial role in the excitatory network of the CNS, blockingthis axis may impair the function of normal brains. An FDA-approved drug, sulfasalazine that blocks xCT and inhibits gluta-mate release, has been shown to have no clinical effect but inducedgrade 4 toxicity in phase I/II trial in recurrent WHO grade 3 and 4astrocytic gliomas in adults (61). In summary, because Tregs act asdownstream of the xCT/glutamate axis in anti-VEGF treatmentresistance, combining VEGF blockade with early Treg depletionrepresents a potentially safe and effective therapeutic strategy forpatients with GBM.

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

Authors’ ContributionsConception and design: Y. Long, Z. Lin, J. HuangDevelopment of methodology: Y. Long, A. Karachi, L. Jin, A.Y. Hou, L.P. Deleyrolle,D.A. MitchellAcquisition of data (provided animals, acquired and managed patients, providedfacilities, etc.): Y. Long, H. Tao, A. Karachi, L. Jin, L.P. DeleyrolleAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): Y. Long, A. Karachi, Y.E. Chang, W. Zhang, A.Y. Hou,E.J. Sayour, Z. Lin, J. HuangWriting, review, and/or revision of the manuscript: Y. Long, A. Karachi,A.J. Grippin, L. Jin, K.A. Dyson, M. Na, L.P. Deleyrolle, E.J. Sayour, M. Rahman,Z. Lin, J. HuangAdministrative, technical, or material support (i.e., reporting or organizing data,constructing databases): Y. Long, E.J. SayourStudy supervision: J. Huang

AcknowledgmentsThis work was supported by the Wells Endowment (00107592), Heilongjiang for

Application Technology Research and Development (GA15C108), The FloridaCenter for Brain Tumor Research (FCBTR) and Accelerate Brain Cancer Cure(ABC2), and UFHCC/ Wells BTC Pilot Award.

The costs of publication of this article were defrayed in part by the payment of pagecharges. This article must therefore be hereby marked advertisement in accordancewith 18 U.S.C. Section 1734 solely to indicate this fact.

Received May 20, 2019; revised August 15, 2019; accepted November 8, 2019;published first November 13, 2019.

References1. Carmeliet P. Mechanisms of angiogenesis and arteriogenesis. Nat Med 2000;

6:389.2. Hicklin DJ, Ellis LM. Role of the vascular endothelial growth factor pathway in

tumor growth and angiogenesis. J Clin Oncol 2005;23:1011–27.3. Norden AD, Drappatz J, Wen PY. Antiangiogenic therapies for high-grade

glioma. Nat Rev Neurol 2009;5:610.4. Gilbert MR, Dignam JJ, Armstrong TS, Wefel JS, Blumenthal DT, Vogelbaum

MA, et al. A randomized trial of bevacizumab for newly diagnosed glioblastoma.N Engl J Med 2014;370:699–708.

5. Chinot OL, Wick W, Mason W, Henriksson R, Saran F, Nishikawa R, et al.Bevacizumab plus radiotherapy–temozolomide for newly diagnosed glioblas-toma. N Engl J Med 2014;370:709–22.

6. McIntyre A, Harris AL. Metabolic and hypoxic adaptation to antiangio-genic therapy: a target for induced essentiality. EMBO Mol Med 2015;7:368–79.

7. Conley SJ, Gheordunescu E, Kakarala P, Newman B, KorkayaH,Heath AN, et al.Antiangiogenic agents increase breast cancer stem cells via the generation oftumor hypoxia. Proc Natl Acad Sci U S A 2012;109:2784–9.

8. Keunen O, Johansson M, Oudin A, Sanzey M, Rahim SAA, Fack F, et al. Anti-VEGF treatment reduces blood supply and increases tumor cell invasion inglioblastoma. Proc Natl Acad Sci U S A 2011;108:3749–54.

9. Patel AP, Tirosh I, Trombetta JJ, Shalek AK, Gillespie SM, Wakimoto H, et al.Single-cell RNA-seq highlights intratumoral heterogeneity in primary glioblas-toma. Science 2014;344:1396–401.

10. Motz GT, Coukos G.The parallel lives of angiogenesis and immunosuppression:cancer and other tales. Nat Rev Immunol 2011;11:702–11.

11. Das S, Marsden PA.Angiogenesis in glioblastoma. N Engl J Med 2013;369:1561–3.

12. Mu L, Yang C, Gao Q, Long Y, Ge H, DeLeon G, et al. CD4þ and PerivascularFoxp3þ T cells in glioma correlate with angiogenesis and tumor progression.Front Immunol 2017;8:1451.

13. Kozin SV, Boucher Y, Hicklin DJ, Bohlen P, Jain RK, Suit HD. Vascularendothelial growth factor receptor-2-blocking antibody potentiates radiation-induced long-term control of human tumor xenografts. Cancer Res 2001;61:39–44.

14. Kloepper J, Riedemann L, Amoozgar Z, Seano G, Susek K, Yu V, et al. Ang-2/VEGF bispecific antibody reprograms macrophages and resident microglia to

anti-tumor phenotype and prolongs glioblastoma survival. Proc Natl Acad Sci US A 2016;113:4476–81.

15. Deligne C, Metidji A, Fridman WH, Teillaud JL. Anti-CD20 therapy induces amemory Th1 response through the IFN-g/IL-12 axis and prevents protumorregulatory T-cell expansion in mice. Leukemia 2015;29:947–57.

16. Deng L, Liang H, Burnette B, Beckett M, Darga T, Weichselbaum RR, et al.Irradiation and anti–PD-L1 treatment synergistically promote antitumor immu-nity in mice. J Clin Invest 2014;124:687–95.

17. ArceVargas F, Furness AJS, Solomon I, Joshi K,Mekkaoui L, LeskoMH, et al. Fc-Optimized anti-CD25 depletes tumor-infiltrating regulatory T Cells and syner-gizes with PD-1 blockade to eradicate established tumors. Immunity 2017;46:577–86.

18. Baker GJ, Castro MG, Lowenstein PR. Isolation and flow cytometric analysis ofglioma-infiltrating peripheral blood mononuclear cells. J Vis Exp 2015;105:53676.

19. Edgar R, Domrachev M, Lash AE. Gene Expression Omnibus: NCBI geneexpression and hybridization array data repository. Nucleic Acids Res 2002;30:207–10.

20. FeuererM,Herrero L, Cipolletta D, Naaz A,Wong J, Nayer A, et al. Lean, but notobese, fat is enriched for a unique population of regulatory T cells that affectmetabolic parameters. Nat Med 2009;15:930–9.

21. Rubenstein JL, Kim J, Ozawa T, Zhang M, Westphal M, Deen DF, et al. Anti-VEGF antibody treatment of glioblastoma prolongs survival but results inincreased vascular cooption. Neoplasia 2000;2:306–14.

22. Phillips HS, Kharbanda S, Chen R, Forrest WF, Soriano RH, Wu TD, et al.Molecular subclasses of high-grade glioma predict prognosis, delineate a patternof disease progression, and resemble stages in neurogenesis. Cancer Cell 2006;9:157–73.

23. DeLay M, Jahangiri A, Carbonell WS, Hu Y-L, Tsao S, Tom MW, et al.Microarray analysis verifies two distinct phenotypes of glioblastomas resistantto antiangiogenic therapy. Clin Cancer Res 2012;18:2930–42.

24. Jahangiri A, De Lay M, Miller LM, Carbonell WS, Hu Y-L, Lu K, et al. Geneexpression profile identifies tyrosine kinase c-Met as a targetable mediator ofantiangiogenic therapy resistance. Clin Cancer Res 2013;19:1773–83.

25. Peterson TE, Kirkpatrick ND, Huang Y, Farrar CT, Marijt KA, Kloepper J, et al.Dual inhibition of Ang-2 andVEGF receptors normalizes tumor vasculature andprolongs survival in glioblastoma by alteringmacrophages. ProcNatl Acad Sci US A 2016;113:4470–5.

Long et al.

Cancer Res; 80(3) February 1, 2020 CANCER RESEARCH508

on September 7, 2021. © 2020 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst November 13, 2019; DOI: 10.1158/0008-5472.CAN-19-1577

26. Rech AJ, Mick R, Martin S, Recio A, Aqui NA, Powell DJ, et al. CD25 blockadedepletes and selectively reprograms regulatory T cells in concert with immu-notherapy in cancer patients. Sci Transl Med 2012;4:134ra62.

27. WuD, Yotnda P. Induction and testing of hypoxia in cell culture. J Vis Exp 2011.DOI: 10.3791/2899.

28. Ge H, Mu L, Jin L, Yang C, Chang YE, Long Y, et al. Tumor associated CD70expression is involved in promoting tumor migration and macrophage infiltra-tion in GBM. Int J Cancer 2017;141:1434–44.

29. Wang Q, Hu B, Hu X, Kim H, Squatrito M, Scarpace L, et al. Tumor evolution ofglioma-intrinsic gene expression subtypes associates with immunologicalchanges in the microenvironment. Cancer Cell 2017;32:42–56.

30. Brennan Cameron W, Verhaak Roel GW, McKenna A, Campos B, NoushmehrH, Salama Sofie R, et al. The somatic genomic landscape of glioblastoma. Cell2013;155:462–77.

31. de Groot J, Sontheimer H. Glutamate and the biology of gliomas. Glia 2011;59:1181–9.

32. Niswender CM, Conn PJ. Metabotropic glutamate receptors: physiology, phar-macology, and disease. Annu Rev Pharmacol Toxicol 2010;50:295–322.

33. Gabrilovich D, Ishida T, Oyama T, Ran S, Kravtsov V, Nadaf S, et al. Vascularendothelial growth factor inhibits the development of dendritic cells anddramatically affects the differentiation of multiple hematopoietic lineagesin vivo. Blood 1998;92:4150–66.

34. Shrimali RK, Yu Z, Theoret MR, Chinnasamy D, Restifo NP, Rosenberg SA.Antiangiogenic agents can increase lymphocyte infiltration into tumor andenhance the effectiveness of adoptive immunotherapy of cancer. Cancer Res2010;70:6171–80.

35. Hamzah J, Jugold M, Kiessling F, Rigby P, Manzur M, Marti HH, et al. Vascularnormalization in Rgs5-deficient tumours promotes immune destruction. Nature2008;453:410.

36. Munn LL, Jain RK.Vascular regulation of antitumor immunity. Science 2019;365:544–5.

37. VoronT, Colussi O,Marcheteau E, Pernot S,NizardM, Pointet A-L, et al. VEGF-A modulates expression of inhibitory checkpoints on CD8þ T cells in tumors.J Exp Med 2015;212:139–48.

38. Renner DN, Malo CS, Jin F, Parney IF, Pavelko KD, Johnson AJ. Improvedtreatment efficacy of antiangiogenic therapy when combined with picornavirusvaccination in the GL261 glioma model. Neurotherapeutics 2016;13:226–36.

39. Suzuki H, Onishi H,Wada J, Yamasaki A, Tanaka H, Nakano K, et al. VEGFR2 isselectively expressed by FOXP3high CD4þ Treg. Eur J Immunol 2010;40:197–203.

40. Terme M, Pernot S, Marcheteau E, Sandoval F, Benhamouda N, Colussi O, et al.VEGFA-VEGFR pathway blockade inhibits tumor-induced regulatory T-cellproliferation in colorectal cancer. Cancer Res 2013;73:539–49.

41. Yang J, Yan J, Liu B. Targeting VEGF/VEGFR tomodulate antitumor immunity.Front Immunol 2018;9:978.

42. Rech AJ, Mick R, Martin S, Recio A, Aqui NA, Powell DJ, et al. CD25 blockadedepletes and selectively reprograms regulatory T cells in concert with immu-notherapy in cancer patients. Sci Transl Med 2012;4:134ra62–ra62.

43. Quezada SA, Peggs KS, Simpson TR, Shen Y, Littman DR, Allison JP. Limitedtumor infiltration by activated T effector cells restricts the therapeutic activity ofregulatory T cell depletion against established melanoma. J Exp Med 2008;205:2125–38.

44. Lo M, Ling V, Wang YZ, Gout PW. The xc- cystine/glutamate antiporter: amediator of pancreatic cancer growth with a role in drug resistance. Br J Cancer2008;99:464–72.

45. Herner A, Sauliunaite D,Michalski CW, ErkanM, DeOliveira T, Abiatari I, et al.Glutamate increases pancreatic cancer cell invasion and migration via AMPAreceptor activation and Kras-MAPK signaling. Int J Cancer 2011;129:2349–59.

46. Ji X, Qian J, Rahman SMJ, Siska PJ, Zou Y, Harris BK, et al. xCT (SLC7A11)-mediated metabolic reprogramming promotes non-small cell lung cancerprogression. Oncogene 2018;37:5007–19.

47. Koppula P, Zhang Y, Shi J, Li W, Gan B. The glutamate/cystine antiporterSLC7A11/xCT enhances cancer cell dependency on glucose by exportingglutamate. J Biol Chem 2017;292:14240–9.

48. Nakanishi S. Molecular diversity of glutamate receptors and implications forbrain function. Science 1992;258:597–603.

49. Ankarcrona M, Dypbukt JM, Bonfoco E, Zhivotovsky B, Orrenius S, Lipton SA,et al. Glutamate-induced neuronal death: a succession of necrosis or apoptosisdepending on mitochondrial function. Neuron 1995;15:961–73.

50. Vaarmann A, Kovac S, Holmstr€om KM, Gandhi S, Abramov AY. Dopamineprotects neurons against glutamate-induced excitotoxicity. Cell Death Dis 2013;4:e455.

51. Chen D, Fan Z, Rauh M, Buchfelder M, Eyupoglu IY, Savaskan N. ATF4promotes angiogenesis and neuronal cell death and confers ferroptosis in axCT-dependent manner. Oncogene 2017;36:5593.

52. Marcus HJ, Carpenter KLH, Price SJ, Hutchinson PJ. In vivo assessment of high-grade glioma biochemistry using microdialysis: a study of energy-related mole-cules, growth factors and cytokines. J Neurooncol 2010;97:11–23.

53. Buckingham SC, Campbell SL, Haas BR, Montana V, Robel S, Ogunrinu T, et al.Glutamate release by primary brain tumors induces epileptic activity. Nat Med2011;17:1269.

54. Huberfeld G, Vecht CJ. Seizures and gliomas—towards a single therapeuticapproach. Nat Rev Neurol 2016;12:204.

55. Lu H, Samanta D, Xiang L, Zhang H, Hu H, Chen I, et al. Chemotherapy triggersHIF-1–dependent glutathione synthesis and copper chelation that induces thebreast cancer stem cell phenotype. Proc Natl Acad Sci U SA 2015;112:E4600–E9.

56. Moalem G, Leibowitz–Amit R, Yoles E, Mor F, Cohen IR, Schwartz M.Autoimmune T cells protect neurons from secondary degeneration after centralnervous system axotomy. Nat Med 1999;5:49.

57. Schwartz M, Shechter R. Protective autoimmunity functions by intracranialimmunosurveillance to support the mind: the missing link between health anddisease. Mol Psychiatry 2010;15:342.

58. Schwartz M, Raposo C. Protective autoimmunity: a unifying model for theimmune network involved in CNS repair. Neuroscientist 2014;20:343–58.

59. Zhang Y, Shi J, Liu X, Feng L, Gong Z, Koppula P, et al. BAP1 links metabolicregulation of ferroptosis to tumour suppression. Nat Cell Biol 2018;20:1181–92.

60. Koppula P, Zhang Y, Zhuang L, Gan B. Amino acid transporter SLC7A11/xCTat the crossroads of regulating redox homeostasis and nutrient dependency ofcancer. Cancer Commun 2018;38:12.

61. Robe PA, Martin DH, Nguyen-Khac MT, Artesi M, Deprez M, Albert A, et al.Early termination of ISRCTN45828668, a phase 1/2 prospective, randomizedstudy of Sulfasalazine for the treatment of progressing malignant gliomas inadults. BMC Cancer 2009;9:372.

AACRJournals.org Cancer Res; 80(3) February 1, 2020 509

Anti-VEGF Therapy Promotes Treg Function

on September 7, 2021. © 2020 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst November 13, 2019; DOI: 10.1158/0008-5472.CAN-19-1577

2020;80:499-509. Published OnlineFirst November 13, 2019.Cancer Res   Yu Long, Haipeng Tao, Aida Karachi, et al.   That Promotes VEGF Blockade Resistance in GlioblastomaDysregulation of Glutamate Transport Enhances Treg Function

  Updated version

  10.1158/0008-5472.CAN-19-1577doi:

Access the most recent version of this article at:

  Material

Supplementary

  http://cancerres.aacrjournals.org/content/suppl/2019/11/13/0008-5472.CAN-19-1577.DC1

Access the most recent supplemental material at:

   

   

  Cited articles

  http://cancerres.aacrjournals.org/content/80/3/499.full#ref-list-1

This article cites 60 articles, 21 of which you can access for free at:

  Citing articles

  http://cancerres.aacrjournals.org/content/80/3/499.full#related-urls

This article has been cited by 1 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

  .pubs@aacr.org

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/80/3/499To request permission to re-use all or part of this article, use this link

on September 7, 2021. © 2020 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst November 13, 2019; DOI: 10.1158/0008-5472.CAN-19-1577