Tumor-Intrinsic PD-L1 Signals Regulate Cell Growth ... · Growth, Pathogenesis, and Autophagy in...

Therapeutics, Targets, and Chemical Biology

Tumor-Intrinsic PD-L1 Signals Regulate CellGrowth, Pathogenesis, and Autophagy inOvarian Cancer and MelanomaCurtis A. Clark1,2, Harshita B. Gupta2, Gangadhara Sareddy3,4, Srilakshmi Pandeswara2,Shunhua Lao2, Bin Yuan2,5, Justin M. Drerup1,2, Alvaro Padron2, Jos�e Conejo-Garcia6,Kruthi Murthy1,2,Yang Liu2,7, Mary Jo Turk8, Kathrin Thedieck9,Vincent Hurez2, Rong Li1,3,5,Ratna Vadlamudi1,3,4, and Tyler J. Curiel1,2,3

Abstract

PD-L1 antibodies produce efficacious clinical responses indiverse human cancers, but the basis for their effects remainsunclear, leaving a gap in the understanding of how to rationallyleverage therapeutic activity. PD-L1 is widely expressed intumor cells, but its contributions to tumor pathogenicity areincompletely understood. In this study, we evaluated thehypothesis that PD-L1 exerts tumor cell–intrinsic signals thatare critical for pathogenesis. Using RNAi methodology, weattenuated PD-L1 in the murine ovarian cell line ID8agg andthe melanoma cell line B16 (termed PD-L1lo cells), whichexpress basal PD-L1. We observed that PD-L1lo cells prolifer-ated more weakly than control cells in vitro. As expected, PD-L1lo cells formed tumors in immunocompetent mice relativelymore slowly, but unexpectedly, they also formed tumors moreslowly in immunodeficient NSG mice. RNA sequencing anal-ysis identified a number of genes involved in autophagy and

mTOR signaling that were affected by PD-L1 expression. Insupport of a functional role, PD-L1 attenuation augmentedautophagy and blunted the ability of autophagy inhibitors tolimit proliferation in vitro and in vivo in NSG mice. PD-L1attenuation also reduced mTORC1 activity and augmented theantiproliferative effects of the mTORC1 inhibitor rapamycin.PD-L1lo cells were also relatively deficient in metastasis to thelung, and we found that anti-PD-L1 administration could blocktumor cell growth and metastasis in NSGmice. This therapeuticeffect was observed with B16 cells but not ID8agg cells, illus-trating tumor- or compartmental-specific effects in the thera-peutic setting. Overall, our findings extend understanding ofPD-L1 functions, illustrate nonimmune effects of anti-PD-L1immunotherapy, and suggest broader uses for PD-L1 as abiomarker for assessing cancer therapeutic responses. CancerRes; 76(23); 6964–74. �2016 AACR.

IntroductionPD-L1 (B7-H1, CD274), an immune co-signaling molecule in

the B7-homology (B7-H) family (1), negatively regulates T-cellfunctions through PD-1 and CD80 interactions (2) and is immu-nopathogenic in diverse cancers (3). Anti–PD-L1 monoclonalantibodies (aPD-L1) are clinically efficacious against distinctcancers where the principal mechanism of action is thought tobe protecting PD-1–expressing antitumor T cells from inhibitionby tumor PD-L1 (3–7).aPD-L1 is FDA-approved for bladder andlung cancers. aPD-1 was recently FDA-approved to treat mela-noma, renal cell carcinoma, non–small cell lung cancer, and headand neck cancers. It is also proposed to work by protecting PD-1–expressing antitumor T cells from inhibition by tumor surface–expressed PD-L1, among other potential mechanisms (7, 8).

Because of the promise ofaPD-L1 andaPD-1 immunotherapy,much attention has focused on effects of the tumor PD-L1/T cellPD-1 axis in cancer immunotherapy and immunopathogenesis.Following the initial report that tumor PD-L1 kills antitumor Tcells (4), most work on PD-L1 signals in cancer has focused ontumor-extrinsic PD-L1 effects, particularly on T cells. However,recent work shows that tumor PD-L1 prevents tumor apoptosis(9), reduces chemotherapy-mediated killing by altering mitogen-activated protein kinase signals (10), and regulates tumor glucosemetabolism in sarcomas (11). Thus, tumor PD-L1 has important

1The Graduate School of Biomedical Sciences, University of Texas HealthScience Center, San Antonio, Texas. 2Department of Medicine, University ofTexas Health Science Center, San Antonio, Texas. 3Cancer Therapy &Research Center, University of Texas Health Science Center, San Antonio,Texas. 4Department of Obstetrics and Gynecology, University of TexasHealth Science Center, San Antonio, Texas. 5Department of Molecular Med-icine, University of Texas Health Science Center, San Antonio, Texas. 6TumorMicroenvironment and Metastasis Program, The Wistar Institute, Philadel-phia, Pennsylvania. 7Xiangya School of Medicine, Central South University,Changsha, Hunan, PR China. 8Department of Microbiology and Immunology,Geisel School of Medicine at Dartmouth, Hanover, New Hampshire. 9Depart-ment of Pediatrics, Center for Liver, Digestive and Metabolic Diseases,University of Groningen, University Medical Center Groningen, Groningen,the Netherlands and Department for Neuroscience, School of Medicine andHealth Sciences, University Oldenburg, Oldenburg, Germany.

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

Corresponding Author: Tyler J. Curiel, Department of Medicine, University ofTexas Health Science Center at San Antonio, STRF MC 8252, 8403 Floyd CurlDrive, San Antonio, TX 78229. Phone: 210-562-4083; Fax: 210-450-1234; E-mail:[email protected]

doi: 10.1158/0008-5472.CAN-16-0258

�2016 American Association for Cancer Research.

CancerResearch

Cancer Res; 76(23) December 1, 20166964

on June 3, 2020. © 2016 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst September 26, 2016; DOI: 10.1158/0008-5472.CAN-16-0258





http://cancerres.aacrjournals.org/



tumor-intrinsic signaling and survival effects, but these are littlestudied.

In our studies of aPD-L1 immunotherapy, we used RNAitechnology to engineer low PD-L1–expressing tumors cells(termed PD-L1lo) from parental PD-L1–expressing cell lines andmade the unexpected observation that PD-L1lo tumor cells pro-liferated more slowly than control cells in vitro, prompting us tostudy tumor-intrinsic PD-L1 signaling effects in more detail. Wefound that tumor PD-L1 had profound effects on expression ofmany genes that play important roles in tumor signaling andmetabolism, including mTOR and autophagy pathways. Autop-hagy is a catabolicmechanismutilized by tumor cells tomodulatecellular stress and metabolism and is targeted to treat somecancers (12). We demonstrated that aPD-L1 antibody used asimmunotherapy had unexpected, important tumor-specific,immune-independent effects on tumor growth and metastaticspread and that tumor PD-L1 also affected sensitivity to pharma-cologic autophagy and mTOR inhibitors and cytokines. Thesedata shift the current paradigm regarding the role of PD-L1 effectson tumor immunopathogenesis and response to immune check-point blockade inhibitory antibodies to include tumor-intrinsicsignaling consequences, indicate approaches to treatment on thebasis of tumor PD-L1 status in novel ways, and suggest that PD-L1could be a biomarker for treatments other thanaPD-L1 oraPD-1.

PD-1, PD-L1, and B7-H3 are immunoglobulin superfamilymembers (13), and all regulate tumor mTOR (and our new datahere; refs. 14, 15). Thus, this superfamily appears tohave commontumor cell–intrinsic effects on important cell signaling and func-tional outcomes that require much additional study. Our workdemonstrates additional tumor-intrinsic effects in this superfam-ily, with a focus on PD-L1.

Materials and MethodsMice

Wild-type (WT) C57BL/6J (BL6), bd TCR knockout (KO), andNOD.Cg-PrkdcscidIl2rgtm1Wj1/SzJ [nonobese diabetic/severe com-bined immunodeficiency (NOD/SCID)/IL2Rg KO, NSG] micewere purchased from Jackson Laboratory. PD-L1 KO mice werea kind gift from Lieping Chen (16). All mice were syngeneic BL6andmaintained under specific pathogen-free conditions and givenfood and water ad libitum. Age- and sex-matched mice that were atleast 8 weeks of age were used for all experiments. Only femaleswere used for ID8agg ovarian cancer studies. All animal studieswere approved by The University of Texas Health Science Center atSan Antonio Institutional Animal Care and Use Committee.

In vivo tumor challenges, treatments, and assessmentsTumor challenge into all mice except NSG was injection of 5�

105 B16 cells subcutaneously as described (16) or 4� 106 ID8aggcells intraperitoneally. For NSG mice, 1.5 � 105 B16 or 1 � 106

ID8agg cell challenges were used. B16 growthwasmeasured everyother daywithVernier calipers and volume calculated as (length�width2)/2. Survival was determined as tumor >1,500 mm3 ordistress (16). ID8agg tumor burden was determined by in vivoluciferase imaging, and survival was assessed by ascites formation,weight gain, or distress (17).

aPD-L1 (10F.9G2) and aPD-1 (RMP1-14) antibodies orrespective isotype controlswere injected intraperitoneally. InNSGmouse challenges, we gave 200 mg/mouse every other day starting1 day before tumor challenge. In WT challenge, we gave 100 mg/

mouse aPD-L1 or isotype control every 5 days for B16 or ID8aggcell starting on day 7 after tumor challenge for times indicated.

For pharmacologic autophagy inhibition in vivo, mice wereinjected intraperitoneally with 60 mg/kg chloroquine every otherday (18, 19) or 25 mg/kg 3-methyladenine every 5 days (20)versus PBS control starting on day 7 following tumor challenge,regimens shown to inhibit tumor cell autophagy effectively in vivo.

In vivo luciferase detection was performed on an IVIS Lumina(Perkin-Elmer) with isoflurane-anesthetized mice 15 minutesafter intraperitoneal injection with 200 mL luciferin potassium(15 mg/mL, Gold Biotechnology) with a 30-second exposure. Anidentical region of interest was drawn over each abdomen, andtumor burden was quantified as average radiance (photons/sec-ond/cm2/sr).

Cell lines and transfectionsMouse B16-F10 melanoma (herein "B16" for simplicity) and

human ES2 ovarian cancer cells were purchased from the ATCC;ID8 were a kind gift from George Coukos (University of Penn-sylvania, Philadelphia, PA). Cells were not revalidated for thiswork. We generated an aggressive ID8 line, ID8agg, by serialpassage through WT hosts (Supplementary Fig. S1). Mouse cellswere used in passages less than 5 and ES2 was passaged for lessthan 6 months. All cells were maintained in 5% FBS-containingRPMI-1640 medium supplemented with 1% penicillin/strepto-mycin, 1% L-glutamate, and 1% HEPES buffer. For serum star-vation experiments, 0.1% FBS replaced 5.0%.

Stable PD-L1 knockdown in B16 and ID8agg lines was gener-ated using lentivirus transduction particles containing validatedPD-L1 shRNA (Sigma, TRCN0000068001) against murinePdcd1Ig1 in pLKO.1-puro vector using puromycin selection permanufacturer's protocol (Supplementary Fig. S2C). ScrambledPD-L1 shRNA control sequence (Sigma, SHC216V) was used togenerate control lines. Individual PD-L1lo clones were selected in2 mg/mL puromycin. For ES2, PD-L1 shRNA (Sigma,TRCN0000056914) against human Pdcd1Ig1 and controls wereused as described above. Individual PD-L1lo clones were selectedin 4 mg/mL puromycin. The PD-L1–overexpressing (PD-L1hi) B16polyclonal line was generated by transfection of pCMV6-PD-L1-GFP plasmid (OriGene, MG203953) using Turbofect (Thermo-Fisher) per the manufacturer's protocol (Supplementary Fig.S2D). Pooled PD-L1hi clones stably expressing PD-L1-GFP wereobtained using G418 selection. Luciferase expression was engi-neered into ID8agg cells by transfection with pGL4.51 [luc2/CMV/neo] plasmid DNA (Promega) and Attractene transfectionreagent (Promega) according to manufacturer's instructions, andindividual clones were selected with G418.

Flow cytometryCells were stained and sorted as previously described (21),

using LSR II and FACSAria hardware and analyzed by FACSDiva(BD Bioscience) and FlowJo software (FlowJo, LLC). Anti-mousePD-L1 (10F.9G2), anti-mouse PD-1 (29F.1A12), and matchedisotype control antibodies were purchased from BioLegend. Cellswere treated with recombinant mouse IFNg (0.1 ng/mL, R&DSystems) for 48 hours as noted to induce PD-L1.

In vitro cell proliferationCells (6 � 103) were plated in 96-well plastic culture plates in

medium and treated 12 hours later with chloroquine (50 mmol/L),

Tumor-Intrinsic PD-L1 Signals in Cancer Pathogenesis

www.aacrjournals.org Cancer Res; 76(23) December 1, 2016 6965




3-methyladenine(5mmol/L), rapamycin(5nmol/L; Sigma for all),TNFa (10 ng/mL; R&D Systems), cisplatin (Sigma; 1–10 mmol/L),or paclitaxel (Sigma; 1–30 nmol/L). aPD-L1 (10F.9G2) or aPD-1(RMP1-14) antibodies or respective isotype controls were used at50 mg/mL for the entire culture period. Data shown are fromoptimized drug concentrations in preliminary work not shown.Cell proliferation rates were determined using MTT (5 mg/mL),whichmeasuresmetabolic activity, 72 hours after treatment. Absor-bance wasmeasured at 540 nmol/L using a BioTek Synergy 2Multi-Mode Plate Reader. Data are presented as mean absorbance versusmedium control� SEM. Proliferation was assessed in triplicate andcompiled from at least 3 separate experiments. Cell viability wasdetermined using a Vi-Cell XR (Beckman Coulter).

RNA sequencingTotal RNA was isolated from control ID8agg and ID8agg PD-

L1lo (clone 3) cells using RNeasy (Qiagen), and RNA quality wasensured on an Agilent Bioanalyzer. Fifty basepair single readsequencing was performed using an Illumina HiSeq 2000 systemat TheUTHSCSAGenome Sequencing Facility. Data are presentedas gene sequence fold change in PD-L1lo versus control. RNA-seqdata were also analyzed by DAVID Bioinformatics and correctedfor multiple comparisons. Genomic data will be deposited in theGene Expression Omnibus database.

ImmunoblottingCell lysates were prepared in RIPA buffer (20mmol/L Tris-HCl,

pH 8.0, 150mmol/L NaCl, 1mmol/L disodium EDTA, 1mmol/LEGTA, 2.5 mmol/L sodium pyrophosphate, 1 mmol/L b-glycer-ophosphate, 1% triton-X100) plus 1 mmol/L phenylmethylsul-phonyl fluoride andHalt protease/phosphatase inhibitor cocktail(Thermo Scientific). Protein concentration was measured byBradfordmethod (ThermoScientific). Fiftymicrogramsof proteinwas separated by 4% to 15% SDS-PAGE (BioRad Hercules),transferred to polyvinylidene fluoride (PVDF) membranes (GEWater and Process Technologies), blocked in TBS (pH 7.4) plus0.1% Tween-20 and 5% skim milk, and incubated overnight at4�C with 1:1,000 diluted phospho- and/or total antibodiesagainst indicated proteins (Cell Signaling) plus anti-mouseb-actin (Santa Cruz Biotechnology). Membranes were incubatedwith horseradish peroxide–conjugated antibodies for 1 hour.Proteins were detected by enhanced chemiluminescence (Pierce).Band quantification and normalization to total protein were byImageJ software (22). Data showmeans of 3 individual blots withcomparisons only made between like blots from the same gels.

MicroscopyCells were plated at low confluence in 6-well plates (50,000

cells/well). On day 2, cells were exposed to serum starvation (0%FBS), normal medium (10% FBS), or chloroquine (50 mmol/L)for 24 hours. Medium was removed, cells were washed with PBSand treated with 4% paraformaldehyde/PBS for 20 minutes atroom temperature, washed, and then permeabilized with 0.1%Triton X-100 for 10 minutes. Cells were then blocked with 5%normal goat serum (Cell Signaling Technology) containing 0.3%Triton X-100 in PBS for 60 minutes. Diluted primary antibody,anti-mouse LC3 A/B (Cell Signaling Technology), was applied inblocking buffer overnight at 4�C. Alexa Fluor-555 secondaryantibody diluted in 1% normal goat serum in PBS were addedfor 1 hour at ambient temperature. Cells were fixed using Vecta-shield hard set mounting medium containing DAPI dye (Vector

Laboratories). Images were acquired using confocal microscopy(Olympus FV-1000) and overlaid using ImageJ (22).

Quantitative RT-PCRTotal RNA was isolated from homogenized whole lung tissue

using RNeasy (Qiagen). cDNA was synthesized with 1 mg of totalRNAusing the ImPromII Reverse Transcription System (Promega)and random primers. Quantitative PCR (qPCR) was conductedusing the 7900HT Real-Time PCR System (Applied Biosystems),amplified with transcript-specific primers with SYBR Green(Thermo Scientific), according to manufacturer's instructions.Mouse primers were: Tyr, 50-CTCTGGGCTTAGCAGTAGGC-30

and 50-GCAAGCTGTGGTAGTCGTCT-30; gp100, 50-ACATTTCAT-CACCAGCAGGGTGCC-30 and 50-AACAAGTGGGTGCTGGCC-30; Trp-2, 50-GTCCTCCACTCTTTTACAGACG-30 and 50-ATTCGG-TTGTGACCAATGGGT; Trp-1, 50 CCCCTAGCCTATATCTCCC-TTTT-30 and 50-TACCATCGTGGGGATAATGGC-30; and Gapdh,50-AACGACCCCTTCATTGAC-30 and 50-TCCACGACATACTCAG-CAC-30 as the internal control (23).

Statistical analysisStatistical analyses were conducted with Prism software

(GraphPad). Data in bar graphs are means � SEM. For tumorgrowth, we used 2-way ANOVA plus Bonferroni post-tests tocompare replicate means. Kaplan–Meier estimates and the log-rank test were used to analyze statistical differences in survival. Forall other single measurement assays, we used an unpaired t test.P < 0.05 was considered significant.

ResultsID8agg and B16 express inducible PD-L1

The ID8 ovarian cancer cell line produces tumors that replicateimportant aspects of human disease, including local spread andascites after intraperitoneal injection into syngeneic BL6mice (24)but disease is not manifested for 10 to 15 weeks (16). Wegenerated an aggressive ID8 line, ID8agg, by serial passagethrough WT hosts that produces rapid disease (SupplementaryFig. S1). Flow cytometry showed that ID8agg cells express PD-L1that is upregulated by IFNg (Fig. 1A) and immunoblot andconfocal microscopy confirmed PD-L1 expression (Supplemen-tary Fig. S2A). B16melanoma cells also express basal PD-L1 that isfurther upregulated by IFNg (Fig. 1B, Supplementary Fig. S2B).

Tumor-intrinsic PD-L1 regulates ID8agg and B16 cellproliferation in vitro and in vivo

We used plasmids stably expressing shRNA to reduce basal andIFNg-inducible PD-L1 expression (PD-L1lo) in ID8agg and B16cells and a constitutively active vector to increase PD-L1 expres-sion in B16 (PD-L1hi, Fig. 1A and B, Supplementary Fig. S2). PD-L1lo B16 cells proliferated significantly slower than control B16cells and PD-L1hi cells exhibited the highest rate of proliferation invitro (Fig. 1C). Subcutaneous WT mouse challenge elicited asimilar trend, with PD-L1hi B16 cells producing the fastest tumorgrowth and PD-L1lo cells the slowest (not shown). Tumor PD-L1can inhibit antitumor immunity by impeding antitumor CD8þ T-cell function (4). Thus, to dissect tumor cell–intrinsic effectsversus immune effects in vivo, we challenged severely immuno-deficient NSG mice with distinct B16 cell constructs subcutane-ously. Consistent with in vitro and WT challenge data, in theabsence of tumor-specific immunity, PD-L1hi B16 tumors grew

Clark et al.

Cancer Res; 76(23) December 1, 2016 Cancer Research6966




fastest and PD-L1lo tumors grew the slowest (Fig. 1D, left), whichcorrelated with reduced overall survival (Fig. 1D, right). Further-more, PD-L1hi cells generated the greatest lung metastases fol-lowing subcutaneous challenge, whereas PD-L1lo cells producedthe fewest as detected by PCR for tumor-specific gene products(Fig. 1E). In support, PD-L1lo ID8agg cells grew more slowly invitro (Fig. 1F) and in vivo compared with control ID8agg afterintraperitoneal challenge in WT (not shown) and NSGmice (Fig.1G). Together, these data clearly demonstrate that tumor-intrinsicPD-L1 controls tumor growth in vitro and in vivo in distinct tumorsand anatomic compartments. PD-L1–mediated growth and pro-motion of bona fidemetastatic tumor spread were all independentof antitumor immunity.

aPD-L1 retards B16, but not ID8agg, tumor growth inNSGmiceTo determine whether tumor cell proliferation rate is influ-

enced by PD-L1 or PD-1 surface expression, we first showed thatB16 also expresses PD-1 as reported (15) and found that ID8aggalso expresses PD-1 (Fig. 2A). Proliferation of control but not PD-L1lo B16 cells in vitro was significantly reduced by both aPD-L1and aPD-1 (Fig. 2B). CD80 was not detected (not shown). Toassess whether aPD-L1 reduces tumor growth in mice lackingantitumor immunity, we challenged NSG mice with control B16cells and treated with aPD-L1, which reduced control B16 (Fig.

2C) and PD-L1hi B16 growth (Supplementary Fig. S3A), consis-tentwith in vitrodata.We confirmed thataPD-1 slows control B16growth in NSG mice (not shown) as reported (15). To determinethe role of tumor cell–intrinsic PD-L1 in metastatic propensity,primary lung metastases in NSGmice challenged subcutaneouslywith B16 were assessed as mRNA of melanoma-specific genes inlung homogenates as described (23). aPD-L1 or aPD-1 eachreduced B16 lung metastases significantly and similarly (Fig.2D, Supplementary Fig. S3B). Similar to B16, aPD-L1 andaPD-1 each reduced rates of control ID8agg cell proliferation invitro but without significant effect on PD-L1lo ID8agg (Fig. 2E). Incontrast, aPD-L1 had negligible treatment effects on peritonealID8agg cell challenge into WT mice (Supplementary Fig. S4) anddid not slow tumor growth or improve survival in NSG challenge(Fig. 2F and G). Thus, whereas certain PD-L1 signaling outcomesare similar in distinct tumors (and see data to follow), aPD-L1effects in vivo can be dissociated from those in vitro, as well as fromPD-L1 effects in distinct tumors as also seen in lack of aPD-L1influence onPD-L1–expressing sarcoma growth inNSGmice (11).

Tumor PD-L1 is the target of aPD-L1–mediated B16 growthinhibition

It is possible that aPD-L1 acts on host PD-L1 to mediate tumorgrowth inhibition. We first showed that aPD-L1 treated control

100A B

DC

E F G

80

60

40

120

100

80

60

5100

80

60

40

20

0Ctrl

PD-L1lo Days post ID8agg challenge

Days post B16 challengeDays post B16 challenge

Clone 3 Clone 6 0

08

3,000

2,000

1,000

0

Tum

or v

olum

e (m

m3 )

10 12 14 16 180

25

50

75

100

10 20 30

20 40 600

25

50

75 CtrlPD-L1lo Clone 3

PD-L1hi

PD-L1loCtrl

P = 0.0002

P = 0.0092

P = 0.0087

100

Perc

ent s

urvi

val

Perc

ent s

urvi

val

43

1.5

1.0

0.5

0.0Trp1 Trp2 Gp100 Tyr

40

20

0PD-L1hi

PD-L1lo

PD-L1lo

PD-L1hi P = 0.009

Ctrl

Ctrl Clone 4 Clone 10

20

0

100

80

60

40

20

0101 102 103 103102 104

PD-L1 ExpressionP = 0.006

P = 0.011

PD-L1 Expression

Nor

mal

ized

cel

l cou

nts

Prol

ifera

tion

(% c

trl)

Rel

ativ

e m

RN

A le

vels

(fold

cha

nge)

Prol

ifera

tion

(% c

trl)

Nor

mal

ized

cel

l cou

nts 100

80

60

40

20

0103102 1050 104

PD-L1 Expression

Nor

mal

ized

cel

l cou

ntsIsotype

PD-L1lo Clone 3Ctrl

BasalIFNγ

IsotypePD-L1lo Clone 4

PD-L1lo Clone 4

PD-L1lo Clone 10

PD-L1hi

Ctrl

CtrlBasalIFNγ

IsotypePD-L1lo Clone 4CtrlPD-L1hi

Figure 1.

Tumor-intrinsic PD-L1 controls immune-independent growth andmetastatic spread. Flow cytometry for PD-L1 expression of in vitro cultured cells. IFNg (0.1 ng/mL)was added for 48 hours as indicated for ID8agg (A) or B16 (B). C, Proliferation in vitro of B16 cells determined by MTT versus control (ctrl, set at 100%).P value, unpaired t test. D, NSG mice challenged subcutaneously with indicated B16 cells. P values for tumor size by two-way ANOVA and for survival by log-ranktest. E, NSG mice challenged with indicated B16 cells and sacrificed on day 18. Genes in whole lung lysates by qPCR. P value, unpaired t test. �, P < 0.05;�� , P < 0.01. F and G, Proliferation in vitro (F) of ID8agg cells as in C and survival in vivo (G) as in D after intraperitoneal ID8agg challenge.






B16, but not PD-L1lo B16 inWTmice (Supplementary Fig. S5A andS5B). To assess host PD-L1 effects specifically,we showed thataPD-L1 treated parental B16 challenge into PD-L1KOmice as effectivelyas in WT mice, whereas aPD-L1 had no effect on PD-L1lo B16challenge in PD-L1 KO mice (Supplementary Fig. S5C and S5D).Therefore, tumor PD-L1 expression is necessary and sufficient foraPD-L1 effects on B16 growth in vivo, and host PD-L1 is notrequired, consistent with antiproliferative effects in vitro (Fig. 2B).

PD-L1 alters tumor-intrinsic signalingData thus far clearly indicate a major cell-intrinsic, immune-

independent role for tumor PD-L1 in both ovarian cancer andmelanoma cells. We next investigated PD-L1–dependent signal-ing pathways. RNA-seq of control versus PD-L1lo ID8agg cellsidentified 1,269 differentially expressed basal genes using a 2-fold

change cutoff (Fig. 3A and data not shown), including significantgene expression differences in canonical andnoncanonical autop-hagy pathways (Table 1A).

Tumor PD-L1 regulates autophagyWe assessed PD-L1 effects on autophagic flux as conversion of

LC3-I to LC3-II in parental ID8agg or PD-L1lo cells. Serumstarvation for 24 hours augmented autophagy in control cells asexpected. Autophagic flux was significantly higher in basal PD-L1lo versus control ID8agg cells as assessed by LC3-II/LC3-I ratio(25), but serum starvation did not increase LC3-II further in PD-L1lo cells (Fig. 3B), suggesting that tumor PD-L1 regulates basaland starvation-induced autophagy in ID8agg. In support of PD-L1–mediated autophagy suppression, LC3 foci formation waslower in control versus PD-L1lo ID8agg (Fig. 3C). PD-L1 also

105 1050.029 0.037 9.41

54.236.4

5.22

93.90.89

B16A B

DC

E

F G

ID8aggPD

-1

PD-L1

4,000Isotype

P = 0.0001P = 0.0047

Ctrl

Medium

120

100

80

60

40

20

0Prol

ifera

tion

(% m

ediu

m a

lone

)

Anti-PD-L1 Anti-PD-1 Days post

Avg

radi

ance

(p/s

/cm

2 /sr)

Perc

ent s

urvi

val

5.0¥106

100

75

50

25

00 10 20 30 40 50

4.0¥106

3.0¥106

2.0¥106

1.0¥106

07 13

P = 0.2028P = 0.2914

ID8agg-luciferase challenge Days post ID8agg-luciferase challenge

PD-L1lo Clone 3PD-L1lo Clone 6

IsotypeAnti-PD-L1

IsotypeAnti-PD-L1

P = 0.048

Anti-PD-L13,000

2,000

1,000

010 12 14

Trp1 Trp2 Gp100 Tyr

16 18 20 220.0

0.5

1.0

1.5

120 CtrlPD-L1lo Clone 4PD-L1lo Clone 10

IsotypeAnti-PD-1Anti-PD-L1

P = 0.0006P = 0.0001

100

80

60

40

Prol

ifera

tion

(% m

ediu

m a

lone

)

20

0Medium Anti-PD-L1 Anti-PD-1

Days post B16 challenge

Tum

or v

olum

e (m

m3 )

Rel

ativ

e m

RN

A le

vels

(fold

cha

nge)

105 105

104 104

104 104

103 103

103

-103 -103

-103

0 0

0 0

Figure 2.

aPD-L1 reduces B16 growth and metastatic spread in NSG mice. A, PD-1 and PD-L1 expression in B16 melanoma and ID8agg ovarian cells measured by flowcytometry. B, Proliferation in vitro of B16 cells � aPD-L1 or aPD-1 (50 mg/mL each) determined by MTT versus control (ctrl, set at 100%). P value, unpaired ttest. C, NSG mice challenged with indicated B16 cells and treated with 200 mg of aPD-L1 every other day starting one day following challenge. P value, two-wayANOVA. D, qPCR for indicated genes from whole lung lysates from mouse challenged as in C, Given 200 mg of aPD-L1 or aPD-1 every other day starting onday following challenge, day 18. Unpaired t test. � , P < 0.05; �� , P < 0.01. E, Proliferation in vitro of ID8agg cells treated as in B. NSG mice challenged withID8agg-luciferase and treated with 200 mg of aPD-L1 every other day starting one day following challenge. P values for average luciferase radiance (F) bytwo-way ANOVA and for survival (G) by log-rank test.

Clark et al.





blunted basal autophagic flux in B16, but in contrast to ID8agg,autophagy was significantly induced during serum starvation inPD-L1lo B16 (Fig. 3D), the appropriate physiologic response.Consistent with Western blot data, there were more LC3 foci inPD-L1lo B16 cells indicative of higher autophagosome formationversus control B16 under basal conditions and serum starvation(Fig. 3E), consistent with tumor cell PD-L1–mediated suppres-sion of autophagy in B16.

PD-L1 sensitizes tumor cells to autophagy inhibitors in vitroBecause PD-L1 depletion deregulates autophagy, including

cell-specific effects (Fig. 3B–E), we tested effects of the pharma-cologic autophagy inhibitor chloroquine (18, 19). In contrast tobasal PD-L1-dependent proliferation effects, PD-L1lo B16 cellswere the most resistant to in vitro chloroquine-mediated prolif-eration inhibition, whereas PD-L1hi cells were the most sensitive

(Fig. 4A). Consistent with B16 cell data, PD-L1lo ID8agg cells weresignificantly more resistant to chloroquine-mediated prolifera-tion inhibition versus control ID8agg cells (Fig. 4A). Using theautophagy inhibitor 3-methyladenine (20), similar, but less pro-nounced, proliferation effects were seen in both B16 and ID8aggcells in vitro (data not shown).

PD-L1 sensitizes B16, but not ID8agg, cells to pharmacologicautophagy inhibitors in vivo

Consistentwith in vitrodata, chloroquine and3-methyladeninewere significantly more effective in reducing control versus PD-L1lo B16 challenge into WTmice (Fig. 4B–D). Because autophagyinhibitors could also affect antitumor immunity, we assessedeffects in T-cell–deficient bd TCR KO mice, where chloroquineretained its clinical activity against control B16 (Fig. 4E) andremained ineffective against PD-L1lo (Fig. 4F). Similarly, in NSGmice, chloroquine was significantly effective against challengewith control B16 cells (Fig. 4G) but was ineffective in reducingtumor growth inmice challengedwithPD-L1lo B16 cells (Fig. 4H).Together, these data support the concept that tumor PD-L1sensitizes B16 melanoma to autophagy inhibitors independentof T-cell immunity and that tumor PD-L1 expression predictsautophagy-dependent growth. Consistent with PD-L1-mediateddependence on autophagy that modulates cellular stress, serumstarvation, which inhibits mTOR signaling and constrains celldependence on autophagy, significantly reduced PD-L1hi andcontrol B16 cell viability in vitro under serum starvation condi-tions (with greatest effect in PD-L1hi cells). We found reduced butsignificant effects in parental B16 and negligible effects in PD-L1lo

cells (Fig. 4I), suggesting tumor PD-L1 hinders autophagy thatnormally compensates effects of serum starvation or other cyto-toxic conditions. In contrast, neither chloroquine nor 3-methy-ladenine improved survival against control or PD-L1lo ID8aggchallenge in vivo (Fig. 4J). These data further establish commonand cell-specific PD-L1 signaling effects.

Tumor PD-L1 regulates tumor mTOR signalsmTOR is a serine/threonine kinase regulating cellular growth

and metabolism that is elevated in many cancers (26, 27) and

A B

DCtrl PD-L1lo

PD-L1loCtrl

PD-L1loCtrl

med+

med− med−

med+R R+C R+C

β-Actin

LC3-ILC3-II

β-Actin

LC3-ILC3-II

Serum + -

Serum+ -

Serum+ -

Serumβ-Actin

LC3-ILC3-II

Serum

4 P = 0.045

P = 0.023

3

2

1

0LC3-

II (f

old

incr

ease

)

Ctrl PD-L1lo

Ctrl CtrlPD-L1lo PD-L1lo

PD

-L1l

oC

trl

+ -

++ --

Ctrl0

10

20

30

P = 0.003

P = 0.007

P = 0.044

CtrlPD-L1lo

LC3-

II (f

old

incr

ease

)

PD

-L1lo

Ctr

l

PD-L1lo

RC C

R R+C R+CRC C

E

CFigure 3.

PD-L1 regulates tumor autophagy genesand functions. A, RNA was isolated fromin vitro–cultured control or PD-L1lo

ID8agg cells, and global genes wereassessed using DAVID Bioinformatics.B,Western blot analysis of lysed ID8aggcells from basal (þ) or serum-starved(�; 24 hours) conditions (left). Right,summary data of three independentexperiments. P values, unpaired t test.C, Confocal images of autophagosomeformation by LC-3 aggregation (red) incontrol versus PD-L1lo ID8agg underbasal or serum-starved (24 hours)conditions. Blue, DAPI for nuclei. D,Analyses of B16 cells as inB, (medþ or�for basal and serum-starved conditions,respectively) treated with rapamycin (R)for 16 hours, chloroquine (C) for6 hours, or both. E, Confocal images ofautophagosome formation forcontrol versus PD-L1lo B16 as in C.

Table 1. Selected genes from RNA-seq analyses of control and PD-L1lo ID8agg,as ratio of gene expression in PD-L1lo versus control cells

Gene Fold change (PD-L1lo/Ctrl) P

A. Autophagya

Ire1a 2.58 5.8E�10Atg9b �2.46 7.7E�08Atg2a �1.79 6.5E�04Perk �1.65 2.6E�03Atg7 �1.69 6.7E�03Atf4 1.41 3.9E�02Atg12 1.43 6.1E�02Atg13 �1.42 7.5E�02

B. mTORb

Prkg2 2.69 3.0E�10Pik3cd �3.27 3.3E�09Irs1 2.81 1.4E�08Map3k6 �2.75 1.9E�08Rhebl1 �2.95 5.4E�07Rps6ka2 �3.32 8.0E�07Atp5k �2.05 2.1E�05Rps6kl1 �8.81 5.0E�05Pdk4 �2.05 1.4E�03Map2k6 �2.35 1.4E�03

aAutophagy pathway genes.bmTOR pathway signaling genes.






recently shown to be affected by tumor-intrinsic PD-1 and PD-L1(11, 15). As mTOR regulates important cellular processes includ-ing autophagy through distinct mTORC1 and mTORC2 com-plexes with distinct functions (28), we assessed additional mTORsignaling details. ID8agg RNA-seq identified potential mTORsignaling effects of PD-L1 (Table 1B). For example, the mTORC1activatorRheb1was significantly lower in PD-L1lo cells (P¼5.35�10�7), consistent with reducedmTORC1 signaling. Prps6, a trans-lation regulator controlled by mTORC1, was also significantlyreduced (P ¼ 7.97� 10�7). mTORC2 activity is defined by targetphosphorylation and cannot be assessed this way.

Tumor PD-L1 regulates mTOR distinctly during serumstarvation and treatment and in distinct tumors

We used immunoblots to validate that tumor PD-L1 pro-motes basal mTORC1 signaling as assessed by P70S6KT389

phosphorylation (Fig. 5A and B). PD-L1lo ID8agg cells para-doxically increased mTORC1 during serum starvation whereasPD-L1lo B16 cells did not (Fig. 5A and B). Rapamycin effectivelysuppressed mTORC1 in control B16 and ID8agg cells asexpected and equally suppressed mTORC1 in PD-L1lo B16 andID8agg cells (Fig. 5D, Supplementary Fig. S6). PD-L1lo B16 andID8agg cells were more sensitive to rapamycin-mediated sup-pression of proliferation in vitro versus control cells, by MTTassay (Fig. 5C), suggesting that PD-L1 regulation of metabolicactivity is mTORC1-dependent, although mTORC2 effects can-not be excluded (29, 30). Under basal conditions, PD-L1inhibited phosphorylation of the mTORC2 substrate AktS473

in B16 and ID8agg (Fig. 5A and B), consistent with reduced

mTORC2 activity (31–33). Rapamycin differentially affectedAktS473 phosphorylation in B16 versus ID8agg in a PD-L1–dependent manner (Fig. 5D). mTORC1 and phosphorylatedP70S6KT389 can mediate a negative feedback loop on PI3K/Akt(34, 35), which is derepressed by rapamycin. Hence, rapamycineffects on AktS473 phosphorylation could be mTORC2-independent.

We next treated cells with chloroquine to inhibit autophagyand found that it inhibited mTORC1 as expected (36, 37) incontrol ID8agg but paradoxically increased mTORC1 in controlB16 cells. Chloroquine had no effect on mTORC1 in PD-L1lo

ID8agg cells but significantly increased mTORC1 in PD-L1lo B16cells (Supplementary Fig. S6). Nonetheless, no observed mTOReffects readily explain how PD-L1 sensitizes cells to chloroquine-mediated cell proliferation inhibition. Thus, PD-L1–mediatedmTOR and autophagy effects could be independent. Other effectsof chloroquine or rapamycin or their combination varied betweenthe 2 tumors on the basis of basal versus serum starvationconditions and PD-L1 status (Fig. 5D, Supplementary Fig. S6 andmuch additional data not shown), likely reflecting distinct tumormutational landscapes in ovarian cancer (38) versus melanoma(39) and further supporting common versus cell-specific PD-L1signaling effects. As mTOR and autophagy regulate responses tocytotoxic insults including chemotherapy and cytokines (40, 41),we explored additional effects. Tumor PD-L1 rendered B16 cellsresistant to TNFa in vitro, whereas ID8agg cells were intrinsicallymore resistant and PD-L1 had no effect on TNF-a–mediatedcytotoxicity (Supplementary Fig. S7A). Surprisingly, tumor PD-L1 did not significantly or consistently alter sensitivity of B16 or

PD-L1lo

P = 0.0004

P = 0.0002

P = 0.015

P = 0.014

P = 0.06P = 0.53

P = 0.14

P = 0.89

P = 0.04

P = 0.008

P = 0.02

P = 0.39

P = 0.038

PD-L1lo Clone 10

A B C D

HGFE

I J

PD-L1lo Clone 4

PD-L1hi

0

3,000ctrl + PBSctrl + CQ

ctrl + PBSctrl + CQ

ctrl + PBSctrl + CQ

ctrl + 3MA

Replete

PD-L1hi PD-L1lo

clone 4

PD-L1lo Clone 10 + PBSPD-L1lo Clone 10 + CQ

PD-L1lo Clone 4 + PBSPD-L1lo Clone 4 + CQPD-L1lo Clone 4 + 3MA

PD-L1lo Clone 10 + PBSPD-L1lo Clone 10 + CQPD-L1lo Clone 10 + 3MA

PD-L1lo Clone 4 + PBSPD-L1lo Clone 4 + CQ

PD-L1lo Clone 6PD-L1lo Clone 6 + CQPD-L1lo Clone 6 + 3MA

PD-L1lo Clone 3PD-L1lo Clone 3 + CQPD-L1lo Clone 3 + 3MA

PD-L1lo

clone 10ctrl PD-L1hi PD-L1lo

clone 4PD-L1lo

clone 10 Days post ID8agg challenge0

0

20

40

60

80 ctrlctrl + CQctrl + 3MA

Perc

ent s

urvi

val 100

20 40Days post ID8agg challenge

00

20

40

60

80

Perc

ent s

urvi

val 100

20 40 60 80 100Days post ID8agg challenge

00

20

40

60

80

Perc

ent s

urvi

val 100

20 40 60 80

Days post B16 challenge6

0

500

1,000

1,500

Tum

or v

olum

e (m

m3 )

Tum

or v

olum

e (m

m3 ) 2,000

0

500

1,000

1,500

2,000

8 10 12 1614Days post B16 challenge

6 8 10 12 1614Tum

or v

olum

e (m

m3 )

0

1,000

2,000

3,000

Days post B16 challenge8 10 12 16 18 2014

Tum

or v

olum

e (m

m3 )

0

500

1,000

1,500

2,000

Days post B16 challenge86 10 12 16 1814 Tu

mor

vol

ume

(mm

3 )

0

500

1,000

1,500

Days post B16 challenge86 10 12 16 1814 Tu

mor

vol

ume

(mm

3 )

0

500

1,000

1,500

Days post B16 challenge86 10 12 16 1814

ctrl

DepleteSerum

2,000

1,000

0

125

100

Viab

ility

(nor

mal

ized

to m

ediu

m a

lone

)

75

50

25

0

8 10 12 14 16 18 20Days post B16 challenge

Tum

or v

olum

e (m

m3 )

20 40 60 80 100% Proliferation inhibition

ctrl

ctrl

ID8

agg

B16

Figure 4.

Tumor PD-L1 regulates response to pharmacologic autophagy inhibitors. A, Indicated cells were cultured with 50 mmol/L chloroquine and proliferation inhibition(100%-% proliferation by MTT, with untreated set at 0%) assessed 72 hours later. P values from unpaired t test. B–D, WT mice challenged with indicated cellsand treated with chloroquine (CQ) or 3-methyladenine (3MA) as described in Materials and Methods. P values, two-way ANOVA. E and F, bd TCR KO micechallenged and treated as in B–D. G and H, NSG mice challenged and treated as in B–D. I, Indicated B16 cells from basal (serum replete) or serum-starved (serumdeplete; 24 hours) conditions and cell viability normalized to basal controls assessed on a Vi-Cell. P values from unpaired t test. J, WT females challengedwith indicated ID8agg, 4 � 106 cells intraperitoneally, and treated with chloroquine or 3MA as in B–D. No differences are significant.

Clark et al.





ID8agg to the cytotoxic agents cisplatin or paclitaxel (Supplemen-tary Fig. S7B and S7C). Thus, tumor PD-L1 is a prosurvival agentbut effects depend on the specific toxic insult.

Cell-intrinsic PD-L1 regulates proliferation, mTOR signaling,and autophagy in human ovarian cancer cells

To test human relevance, we first showed that the human ES2ovarian cancer cell line expressed basal PD-L1 and then engi-neered PD-L1lo ES2 cells (Fig. 6A and B). Some ES2 cells alsoexpress basal PD-1 (Fig. 6B). Consistent with B16 and ID8aggdata, PD-L1lo ES2 cells proliferated significantly slower thancontrol ES2 (Fig. 6C). LC3-II versus LC3-I was increased in PD-L1lo but not in control cells (Fig. 6D), consistent with increasedautophagic flux when PD-L1 was reduced. In support, LC3 fociwere reduced in control versus PD-L1lo cells indicative of PD-L1–mediated decrease in autophagosome formation (Fig. 6E). Alto-gether these data are consistent with PD-L1–dependent suppres-sion of autophagic flux in human cancer cells, similar tomouse cell data. Further consistent with B16 and ID8agg data,PD-L1 in ES2 cells augmented mTORC1 signals (P70S6KT389

phosphorylation). In contrast to B16 and ID8agg data, PD-L1knockdown eliminated ES2 mTORC2 activation (AktS473

phosphorylation; Fig. 6F).Validating our observations in B16 and ID8agg, control ES2

cells were significantly more sensitive to proliferation inhibitionby chloroquine comparedwithPD-L1lo ES2 cells (Fig. 6G), furthersuggesting that PD-L1 regulation of cancer cell autophagy andautophagy dependence could be a common mechanism in PD-L1–expressing mouse and human cancer cells.

DiscussionTumor-expressed PD-L1 alters tumor immunopathogenesis by

delivering negative signals to PD-1–expressing antitumor T cells

(3–7), but emerging evidence shows that PD-L1 and PD-1 alsohave tumor-intrinsic functions (11, 15). Our work now clearlyestablishes that tumor PD-L1 has additional, important tumorintrinsic effects. We found that tumor PD-L1 promoted cell-intrinsic growth in 2 distinct tumor types (melanoma andovarian cancer), in distinct anatomic compartments (skin andperitoneum) and in an immune-independent fashion for both.Thus, these properties are not specific to one single cell type oranatomic location.

aPD-L1 can alter PD-L1–mediated cell-intrinsic growth signals,as a-PD-L1 slowed B16 tumor growth in NSG mice incapable ofmediating tumor-specific immunity or antibody-dependent cel-lular cytotoxicity. Melanoma PD-L1 promoted immune-indepen-dent bona fidemetastases from subcutaneous melanoma, a morestringent model than intravenous cell challenge. Increased tumorgrowth and proliferation likely contribute to increased metasta-ses, but additional factors could also contribute. Further work isneeded to understand applicability to additional cancers anddefine mechanisms.

aPD-1 and aPD-L1 comparably inhibited B16 growth in vitroand comparably reduced B16 metastases in vivo. These data areconsistent with tumor PD-L1/tumor PD-1 cooperation, as onlyabout 5% to 10% of cells expressed PD-1. In vivo, aPD-L1 effectsappears to be directed to tumor and not other PD-L1–expressingcells, as aPD-L1 (i) had no significant effect on PD-L1lo B16melanoma growth inWTmice and (ii) was fully protective in PD-L1 KO mice challenged with parental B16 cells. However, thistreatment effect could be tumor- or compartment-specific, asaPD-L1 did not slow tumor growth or improve survival inintraperitoneal ID8agg ovarian cancer challenged NSG mice,despite reducing in vitroproliferation. In support of tumor-specificeffects, a recent report showed that aPD-L1 did not slow PD-L1þ

sarcoma growth in NSG mice (11). Specific tumors might berefractory to PD-L1–driven growth or survival signals owing to

D PD-L1lo

PD-L1lo PD-L1lo

PD-L1lo PD-L1lo B16 ctrl PD-L1lo

p-P70S6KT389

P70S6K

Akt

β-Actin

p-AktS473

ID8agg ctrl

Serum +

0.0

0.5

1.0

1.5

1.0

0.3

0.2

0.1

0.0

1.5

1.0

0.5

0.0

1.5 150

125

100

75

50

25

0

1.0

0.5

0.0

1.5P = 0.019

P = 0.003

P = 0.004P = 0.035

P = 0.024

P = 0.028

P = 0.036P = 0.031

P = 0.037

P = 0.002

P = 0.035

P = 0.006

P = 0.001

B16

CBA B16ID8aggctrl ID8aggPD-L1lo

ctrlPD-L1lo

ctrl ctrl PD-L1lo PD-L1loctrl ctrl PD-L1lo PD-L1loctrl ctrl PD-L1lo PD-L1loctrl ctrl

p-P

70S

6KT

389

p-P

70S

6KT

389

p-A

ktS

473

p-A

ktS

473

Pro

lifer

atio

n (

% m

ediu

m a

lon

e)

Serum - Serum + Serum - Serum + Serum - Serum + Serum - Medium Rapamycin

ID8agg ctrlB16 ctrl

+ +R R+C R+CRC C + +R R+C R+CRC C −− R R+C R+CRC C − R+CR C − R+CR C

Figure 5.

Tumor PD-L1 regulates mTOR distinctly and blunts rapamycin proliferation inhibition. Summary data of Western blot analyses for P70S6KT389 and AktS473

phosphorylation as ratios of phospho-protein to total protein under basal or serum-starved (24 hours) conditions for ID8agg (A) and B16 (B) cells. Statistical analysesfrom average of three independent experiments. C, Indicated cells were cultured with 5 nmol/L rapamycin and proliferation by MTT, with control set at 100%,assessed 72 hours later. D, Representative Western blot analyses for treatments with rapamycin (R) for 16 hours, chloroquine (C) for 6 hours, or both (Rþ C) underbasal (þ) or serum-starved (�) conditions for ID8agg and B16. Summary data for these blots are in Supplementary Fig. S6.






tumor-specific mutations (38, 39), degree of PD-L1 expression,antibody affinity and delivery differences, or other factors.

Tumor PD-L1 suppressed tumor autophagy in melanoma andovarian cancer, suggesting that autophagy disruption could be arelatively general tumor PD-L1 effect. Tumor PD-L1 regulatesglucose metabolism in sarcomas (11), suggesting that PD-L1could have a variety of important cell-intrinsic metabolic effects.Furthermore, B7-H3 regulates glucose metabolism in breast can-cer cells (42), suggesting that the B7-H superfamily, to which PD-L1 andB7-H3belong (43) or the immunoglobulin superfamily towhich these and PD-1 all belong (13), could have importanttumor metabolic effects, which is an area meriting additionalinvestigations.Our RNA-seqdata demonstrated numerous poten-tially PD-L1–regulated genes. Ingenuity Pathway Analysis deter-mined that many critical pathways were affected, includingimmune cell trafficking (chemokines and cytokines), inflamma-tion, TGFb signaling, metabolism, and cancer stem cell genesamong others (our unpublished data), in addition to effectsshown here. Thus, much additional, important PD-L1–mediatedtumor cell signaling likely awaits definition.

We found that PD-L1–replete B16 melanoma and ID8aggovarian cancer cells exhibit low basal autophagy, with highmTORC1activity, andwere disproportionately sensitive to autop-hagy inhibitor–mediated growth reduction comparedwith cancercells with low PD-L1 expression. We postulate that further autop-hagy reduction in cancer cells with elevatedmTORC1 activity andlow autophagic activity (directly related to PD-L1 expression) iscatastrophic. In support, autophagy levels can be fine-tuned to

allow cells tomeet metabolic demands (44) without compromis-ing other functions. In vivo, this PD-L1–mediated balance aug-ments clinical response to autophagy inhibitors. These datasuggest that autophagy inhibitors could boost aPD-L1 treatmentof PD-L1–positive cancers, yet other tumor- and/or immune-mediated effects may contribute. Despite high levels of autopha-gic flux, PD-L1lo tumor cells might have a reduced autophagyrequirement due to lower metabolic demand and/or cellularstress, as supported by reduced sensitivity to autophagy inhibi-tors. Alternatively, pharmacologic inhibitors could incompletelyinhibit this high autophagy for clinical effect. Genetic approachesto autophagy manipulation are required to explore mechanismsfurther, but our data support the thesis that tumor PD-L1 expres-sion and basal autophagy levels could be a biomarker for effectiveclinical autophagy inhibitor use. Our data also suggest thatautophagy inhibition plus aPD-L1 merits clinical investigation.

As autophagy is a tumor survival mechanism, we expected thattumor PD-L1 could affect cell viability after specific insults andfound differences in survival in serum starvation, mTOR, andautophagy inhibitors and TNF-a. Surprisingly, we did not detecta robust difference in sensitivity to cis-platinum or paclitaxelin vitro, despite potential indications for a role of tumor-intrinsicPD-L1 in chemoresistance. Anabstract suggested that tumorPD-L1protects from cytotoxic chemotherapy (10). That work used PD-L1overexpression and did not specify cell lines or cytotoxic agents,making usunable to comment onpotential differences infindings.

PD-L1 regulates tumor mTOR signals in melanoma and sar-coma (11, 15), butmechanistic details are lacking and unreported

100Isotype

Isotype controlsBA

D E F G

C0.22105 105

104 104

103103

-103 -103

-103 103 104 1050 -103 103 104 1050

0 0

0.09

1.39

0.15 2.58

94.62.67

PD-L1

ctrl Clone 1 Clone 2ctrl Clone 1 Clone 2

P = 0.0012

P = 0.0233

ctrl0

10

20

30

40

50

Clone 1

ctrl

120P = 0.0001

P = 0.0002P = 0.0001

100

80

60

4020

0

Prol

ifera

tion

(%ct

rl)

Pooled Clone 1 Clone 2

Clone 2

p-P70S6KT389

% P

rolif

erat

ion

inhi

bitio

n

P70S6K

p-AktS473

Akt

b-Actin

LC3-I

+ Serum -

LC3-II

ctrl

PD-L

1lo

b-Actin

PD-1

98.3

PD-L1lo Clone 1PD-L1lo Clone 2

PD-L1lo

PD-L1lo

PD-L1lo

PD-L1lo

ctrl

80

60

40

20

0-103 103 104 1050

PD-L1 Expression

Nor

mal

ized

cel

l cou

nts

Figure 6.

Tumor cell–intrinsic PD-L1 regulates proliferation, mTOR signaling, and autophagy in human ovarian cancer cells. A, Flow cytometry for PD-L1 expression of in vitro–cultured ES2 human ovarian cancer cells showing PD-L1 knockdown by shRNA. B, PD-1 and PD-L1 expression by flow cytometry. C, Proliferation in vitro ofES2 cells determined by MTT versus control (ctrl, set at 100%). P value, unpaired t test. D,Western blot analysis for LC3I/II in ES2 cell lysates from basal conditions.E, Confocal images of autophagosome formation by LC-3 aggregation (red) in control versus PD-L1lo ES2 under basal or serum-starved (24 hours) conditions.Blue, DAPI for nuclei. F, Western blot analysis for P70S6KT389 and AktS473 phosphorylation in ES2 cells under basal conditions. G, Control and PD-L1lo ES2cells were cultured with 50 mmol/L chloroquine and proliferation inhibition (100%-% proliferation by MTT, with untreated set at 0%), assessed 72 hours later.P values, unpaired t test.

Clark et al.





in ovarian cancer. We show that low PD-L1 in melanoma andovarian cancer cells results in reduced phosphorylation of themTORC1 substrate P70S6KT389. PD-L1lo cells proliferated moreslowly than control lines, suggesting that lower mTORC1activity in PD-L1lo cells compromises their growth. Prolifera-tion inhibition by the (predominantly) mTORC1 inhibitorrapamycin abolished P70S6KT389 phosphorylation in controland PD-L1lo B16 and ID8agg cells in vitro as expected, yet PD-L1lo cells were more rapamycin-sensitive than control cells,suggesting that proliferation of PD-L1lo tumors is more depen-dent on basal mTORC1 activity compared with PD-L1–repletetumor cells. PD-L1–mediated mTORC1 effects could differ intumor cell metabolism, survival, and therapeutic outcomes.Nonetheless, these data support the concept that tumor PD-L1–dependent mTORC1 activity drives proliferation. Thus, PD-L1expression could predict mTORC1 activity and rapamycin-sensitive tumor growth, which is clinically exploitable. Forinstance, these data suggest that aPD-L1 and/or aPD-1 immu-notherapies combined with mTOR inhibitors in treating PD-1/PD-L1–replete tumors or mTOR inhibitors alone in treatingPD-1lo/PD-L1lo tumors are rational approaches.

Other mTORC1 substrates, mTORC2, or other rapamycineffects could also mediate specific PD-L1–dependent effects.For example, mTORC2 and its substrate Akt drive tumor cellgrowth and survival (45). We found that phosphorylation ofthe mTORC2 substrate AktS473 was induced in PD-L1lo B16 andID8agg cells. In contrast to parental ID8agg cells, rapamycin didnot further induce pAktS473 in PD-L1lo cells, demonstrating aPD-L1–dependent effect, perhaps from low basal mTORC1 andP70S6K activities, in which case negative feedback on Akt couldbe inactive and thus unaffected by rapamycin. We speculatethat PD-L1–driven mTORC1 activation accounts for reducedautophagy, but genetic and additional approaches are requiredfor definitive mechanistic insights. Additional study of PD-L1–mediated mTOR signaling and outcomes of mTOR inhibition(e.g., metabolic activity), and cell-specific effects could helpoptimize mTOR inhibition strategies to treat selected cancers.As PD-L1, PD-1, and B7-H3 all regulate tumor mTOR signaling(11, 15, 42), this could be another general feature of selectedimmunoglobulin superfamily members and requires addition-al study.

Finally, we demonstrated human relevance by showing thatcell-intrinsic PD-L1 in the human ES2 ovarian cancer cell linecontrols proliferation, mTORC1, autophagic flux, and sensitivityto autophagy inhibitors. PD-L1 in human melanoma cells con-trols PD-1 effects on melanoma growth and mTOR signals (15).Thus, human studies merit much additional attention.

In summary, we show that tumor cell–intrinsic PD-L1 signalsmediate effects common to melanoma and ovarian cancer cellsin mouse lines and effects extend to human ovarian cancer cells.

PD-L1 promotes tumor cell proliferation and immune-indepen-dent growth inmelanoma and ovarian cancer cells andmetastaticmelanoma spread in vivo. Tumor cell–intrinsic PD-L1 alteredautophagy inhibitor and mTOR inhibitor efficacy. PD-L1 pro-moted basal mTORC1 activation in all cells tested and inhibitedphosphorylation of themTORC2 substrate Akt inmouse lines butnot the human line tested. aPD-L1 inhibited in vivo melanomagrowthbut not ovarian cancer growth in an immune-independentfashion and melanoma but not ovarian cancer cells were slowedby autophagy inhibitors in vivo. Thus, tumor PD-L1 could be abiomarker for response to mTOR or autophagy inhibitors inselected cancers. Tumor PD-L1 appeared to cooperate with tumorPD-1 for selected effects. As our understanding of tumor cell–intrinsic PD-L1 effects increases, our ability to predict treatmentresponses to various agents and combine them in rationale waysfor more effective clinical use will improve. Given the extraordi-nary effect that PD-L1 appears to have on diverse, critical cellularprocesses, much additional work in this area is merited.

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

Authors' ContributionsConception and design: C.A. Clark, T.J. CurielDevelopmentofmethodology:C.A.Clark, S. Pandeswara, B. Yuan, J.M.Drerup,R.K. Vadlamudi, T.J. CurielAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): C.A. Clark, H. Gupta, G.R. Sareddy, S. Lao, J.M.Drerup, A. Padron, Y. Liu, V. HurezAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): C.A. Clark, H. Gupta, G.R. Sareddy, B. Yuan, J.M.Drerup, K. Murthy, K. Thedieck, R. Li, R.K. Vadlamudi, T.J. CurielWriting, review, and/or revision of the manuscript: C.A. Clark, H. Gupta, J.M.Drerup, J.R. Conejo-Garcia, M.J. Turk, K. Thedieck, V. Hurez, R. Li, T.J. CurielAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases):C.A. Clark, S. Pandeswara, A. Padron, K.Murthy,T.J. CurielStudy supervision: K. Murthy, T.J. Curiel

Grant SupportWork was supported by grants from the NIH, European governments and

private foundations. T.J. Curiel received grants CA170491, CA54174,CA054174, CDMRP, The Holly Beach Public Library, and The Owens Foun-dation and the Skinner endowment. K. Thedieck received grants from RosalindFranklin Fellowship, BMBF e:Med Young investigator network GlioPATH(01ZX1402B), BMBF e:Med Demonstrator project MAPTor-NET (31P9013A),Ubbo Emmius Funds, Stichting Michelle). M.J. Turk received grant CA120777.C.A.Clark received grantsCA206398 andUTHSCSAMD/PhDProgram support.

The costs of publication of this articlewere defrayed inpart by the payment ofpage charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Received January 28, 2016; revised September 16, 2016; accepted September16, 2016; published OnlineFirst September 26, 2016.

References1. Zou W, Chen L. Inhibitory B7-family molecules in the tumour microen-

vironment. Nat Rev Immunol 2008;8:467–77.2. Paterson AM, Brown KE, Keir ME, Vanguri VK, Riella LV, Chandraker A,

et al. The programmed death-1 ligand 1:b7-1 pathway restrains diabeto-genic effector T cellsin vivo. J Immunol 2011;187:1097–105.

3. Topalian SL, Drake CG, Pardoll DM. Targeting the PD-1/B7-H1(PD-L1)pathway to activate anti-tumor immunity. Curr Opin Immunol 2012;24:207–12.

4. Dong H, Strome SE, Salomao DR, Tamura H, Hirano F, Flies DB, et al.Tumor-associated B7-H1 promotes T-cell apoptosis: a potential mecha-nism of immune evasion. Nat Med 2002;8:793–800.

5. Brahmer JR, Tykodi SS, ChowLQ,HwuWJ, Topalian SL,HwuP, et al. Safetyand activity of anti-PD-L1 antibody in patients with advanced cancer.N Engl J Med 2012;366:2455–65.

6. Taube JM, Anders RA, Young GD, Xu H, Sharma R, McMiller TL, et al.Colocalization of inflammatory response with B7-h1 expression in human






melanocytic lesions supports an adaptive resistance mechanism ofimmune escape. Sci Transl Med 2012;4:127ra37.

7. Pardoll DM. The blockade of immune checkpoints in cancer immuno-therapy. Nat Rev Cancer 2012;12:252–64.

8. Topalian SL,Hodi FS, Brahmer JR,Gettinger SN, SmithDC,McDermottDF,et al. Safety, activity, and immune correlates of anti-PD-1 antibody incancer. N Engl J Med 2012;366:2443–54.

9. Azuma T, Yao S, Zhu G, Flies AS, Flies SJ, Chen L. B7-H1 is a ubiquitousantiapoptotic receptor on cancer cells. Blood 2008;111:3635–43.

10. Orzechowski A, Dong H. B7-H1 confers tumor chemoresistance by regu-lating MAPK/ERK activation. Cancer Res 2014;74:abstract 5026.

11. Chang CH, Qiu J, O'Sullivan D, Buck MD, Noguchi T, Curtis JD, et al.Metabolic competition in the tumormicroenvironment is a driver of cancerprogression. Cell 2015;162:1229–41.

12. MahalingamD,MitaM, Sarantopoulos J,Wood L,Amaravadi RK,Davis LE,et al. Combined autophagy and HDAC inhibition: a phase I safety,tolerability, pharmacokinetic, and pharmacodynamic analysis of hydro-xychloroquine in combination with the HDAC inhibitor vorinostat inpatients with advanced solid tumors. Autophagy 2014;10:1403–14.

13. Wang S, Bajorath J, Flies DB, Dong H, Honjo T, Chen L. MolecularModeling and functional mapping of B7-H1 and B7-DC uncouple costi-mulatory function from PD-1 interaction. J Exp Med 2003;197:1083–91.

14. Jiang B, Liu F, Liu Z, Zhang T, HuaD. B7-H3 increases thymidylate synthaseexpression via the PI3k-Akt pathway. Tumour Biol 2016;37:9465–72.

15. Kleffel S, Posch C, Barthel SR, Mueller H, Schlapbach C, Guenova E, et al.Melanoma cell-intrinsic PD-1 receptor functions promote tumor growth.Cell 2015;162:1242–56.

16. Lin PY, Sun L, Thibodeaux SR, Ludwig SM, Vadlamudi RK, Hurez VJ, et al.B7-H1-dependent sex-related differences in tumor immunity and immu-notherapy responses. J Immunol 2010;185:2747–53.

17. Facciabene A, Peng X, Hagemann IS, Balint K, Barchetti A, Wang LP, et al.Tumour hypoxia promotes tolerance and angiogenesis via CCL28 and T(reg) cells. Nature 2011;475:226–30.

18. Xie X, White EP, Mehnert JM. Coordinate autophagy and mTOR pathwayinhibition enhances cell death in melanoma. PLoS One 2013;8:e55096.

19. Maes H, Kuchnio A, Peric A, Moens S, Nys K, De Bock K, et al. Tumor vesselnormalization by chloroquine independent of autophagy. Cancer Cell2014;26:190–206.

20. Li J, Yang D, Wang W, Piao S, Zhou J, Saiyin W, et al. Inhibition ofautophagy by 3-MA enhances IL-24-induced apoptosis in human oralsquamous cell carcinoma cells. J Exp Clin Cancer Res 2015;34:97.

21. Curiel TJ, Wei S, DongH, Alvarez X, Cheng P, Mottram P, et al. Blockade ofB7-H1 improves myeloid dendritic cell-mediated antitumor immunity.Nat Med 2003;9:562–7.

22. Schneider CA, Rasband WS, Eliceiri KW. NIH Image to ImageJ: 25 years ofimage analysis. Nat Methods 2012;9:671–5.

23. Pedersen SR, Sorensen MR, Buus S, Christensen JP, Thomsen AR. Com-parison of vaccine-induced effector CD8 T cell responses directed againstself- and non-self-tumor antigens: implications for cancer immunothera-py. J Immunol 2013;191:3955–67.

24. Roby KF, Taylor CC, Sweetwood JP, Cheng Y, Pace JL, Tawfik O, et al.Development of a syngeneic mouse model for events related to ovariancancer. Carcinogenesis 2000;21:585–91.

25. Pampliega O, Orhon I, Patel B, Sridhar S, Diaz-Carretero A, Beau I, et al.Functional interaction between autophagy and ciliogenesis. Nature 2013;502:194–200.

26. Riaz H, Riaz T, Hussain SA. mTOR inhibitors: a novel class of anti-canceragents. Infect Agent Cancer 2012;7:1.

27. GuertinDA, SabatiniDM.Defining the role ofmTOR in cancer. Cancer Cell2007;12:9–22.

28. Dao V, Pandeswara S, Liu Y, Hurez V, Dodds S, Callaway D, et al.Prevention of carcinogen and inflammation-induced dermal cancer byoral rapamycin includes reducing genetic damage. Cancer Prev Res (Phila)2015;8:400–9.

29. Sarbassov DD, Ali SM, Sengupta S, Sheen JH, Hsu PP, Bagley AF, et al.Prolonged rapamycin treatment inhibits mTORC2 assembly and Akt/PKB.Mol Cell 2006;22:159–68.

30. Lamming DW, Ye L, Katajisto P, Goncalves MD, Saitoh M, Stevens DM,et al. Rapamycin-induced insulin resistance is mediated by mTORC2 lossand uncoupled from longevity. Science 2012;335:1638–43.

31. Sarbassov DD, Guertin DA, Ali SM, Sabatini DM. Phosphorylation andregulation of Akt/PKB by the rictor-mTOR complex. Science 2005;307:1098–101.

32. Hresko RC, Mueckler M. mTOR.RICTOR is the Ser473 kinase for Akt/protein kinase B in 3T3-L1 adipocytes. J Biol Chem 2005;280:40406–16.

33. Jacinto E, Facchinetti V, Liu D, Soto N, Wei S, Jung SY, et al. SIN1/MIP1maintains rictor-mTOR complex integrity and regulates Akt phosphoryla-tion and substrate specificity. Cell 2006;127:125–37.

34. HsuPP, Kang SA, Rameseder J, Zhang Y,OttinaKA, LimD, et al. ThemTOR-regulated phosphoproteome reveals a mechanism of mTORC1-mediatedinhibition of growth factor signaling. Science 2011;332:1317–22.

35. Yu Y, Yoon SO, Poulogiannis G, Yang Q, Ma XM, Villen J, et al. Phospho-proteomic analysis identifies Grb10 as an mTORC1 substrate that nega-tively regulates insulin signaling. Science 2011;332:1322–6.

36. Settembre C, Zoncu R, Medina DL, Vetrini F, Erdin S, Huynh T, et al. Alysosome-to-nucleus signalling mechanism senses and regulates the lyso-some via mTOR and TFEB. EMBO J 2012;31:1095–108.

37. Zoncu R, Bar-Peled L, Efeyan A, Wang S, Sancak Y, Sabatini DM. mTORC1senses lysosomal amino acids through an inside-out mechanism thatrequires the vacuolar H(þ)-ATPase. Science 2011;334:678–83.

38. Westcott PM,Halliwill KD, ToMD, RashidM, Rust AG, Keane TM, et al. Themutational landscapes of genetic and chemical models of Kras-driven lungcancer. Nature 2015;517:489–92.

39. Spranger S, BaoR,Gajewski TF.Melanoma-intrinsic beta-catenin signallingprevents anti-tumour immunity. Nature 2015;523:231–5.

40. Zoncu R, Efeyan A, Sabatini DM.mTOR: from growth signal integration tocancer, diabetes and ageing. Nat Rev Mol Cell Biol 2010;12:21–35.

41. Baginska J, Viry E, Berchem G, Poli A, Noman MZ, van Moer K, et al.Granzyme B degradation by autophagy decreases tumor cell susceptibilityto natural killer-mediated lysis under hypoxia. Proc Natl Acad Sci U S A2013;110:17450–5.

42. Nunes-Xavier CE, Karlsen KF, Tekle C, Pedersen C, Oyjord T, Hongisto V,et al. Decreased expression of B7-H3 reduces the glycolytic capacity andsensitizes breast cancer cells to AKT/mTOR inhibitors. Oncotarget2016;7:6891–901.

43. Chapoval AI, Ni J, Lau JS, Wilcox RA, Flies DB, Liu D, et al. B7-H3: acostimulatory molecule for T cell activation and IFN-gamma production.Nat Immunol 2001;2:269–74.

44. Kaur J, Debnath J. Autophagy at the crossroads of catabolism and anab-olism. Nat Rev Mol Cell Biol 2015;16:461–72.

45. Laplante M, Sabatini DM. mTOR signaling in growth control and disease.Cell 2012;149:274–93.


Clark et al.




Correction

Correction: Tumor-Intrinsic PD-L1 SignalsRegulate Cell Growth, Pathogenesis, andAutophagy in Ovarian Cancer and Melanoma

In this article (Cancer Res 2016;76:6964–74), which was published in the December1, 2016, issue of Cancer Research (1), the authors regret that the legend for Supple-mentary Fig. S3 incorrectly noted "[. . .]as in panels for Fig. 2B and C." It should read,"[. . .]as in panels for Fig. 2C and D."

In addition, the authors regret that the legend for Supplementary Fig. S4 incorrectlystated, "PD-L1 KO mice were challenged with either control (B) or PD-L1lo (C) B16cells and treated with anti-PD-L1 (100 mg/mouse) every 5 days starting day 7following B16 challenge." This statement has been deleted.

The online Supplementary Figure legends have been replaced.

Reference1. Clark CA, Gupta HB, Sareddy G, Pandeswara S, Lao S, Yuan B, et al. Tumor-intrinsic PD-L1 signals

regulate cell growth, pathogenesis, and autophagy in ovarian cancer and melanoma. Cancer Res2016;76:6964–74.

Published online May 15, 2017.doi: 10.1158/0008-5472.CAN-17-0351�2017 American Association for Cancer Research.

CancerResearch

Cancer Res; 77(10) May 15, 20172770

http://crossmark.crossref.org/dialog/?doi=10.1158/0008-5472.CAN-17-0351&domain=pdf&date_stamp=2017-4-29

2016;76:6964-6974. Published OnlineFirst September 26, 2016.Cancer Res Curtis A. Clark, Harshita B. Gupta, Gangadhara Sareddy, et al. and Autophagy in Ovarian Cancer and MelanomaTumor-Intrinsic PD-L1 Signals Regulate Cell Growth, Pathogenesis,

Updated version

10.1158/0008-5472.CAN-16-0258doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerres.aacrjournals.org/content/suppl/2016/09/24/0008-5472.CAN-16-0258.DC1

Access the most recent supplemental material at:

Cited articles

http://cancerres.aacrjournals.org/content/76/23/6964.full#ref-list-1

This article cites 44 articles, 15 of which you can access for free at:

Citing articles

http://cancerres.aacrjournals.org/content/76/23/6964.full#related-urls

This article has been cited by 13 HighWire-hosted articles. Access the articles at:

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

Subscriptions

Reprints and

[email protected]

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

Permissions

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

.http://cancerres.aacrjournals.org/content/76/23/6964To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/lookup/doi/10.1158/0008-5472.CAN-16-0258

http://cancerres.aacrjournals.org/content/suppl/2016/09/24/0008-5472.CAN-16-0258.DC1

http://cancerres.aacrjournals.org/content/76/23/6964.full#ref-list-1

http://cancerres.aacrjournals.org/content/76/23/6964.full#related-urls

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

mailto:[email protected]

http://cancerres.aacrjournals.org/content/76/23/6964


Tumor-Intrinsic PD-L1 Signals Regulate Cell Growth ... · Growth, Pathogenesis, and Autophagy in...

Documents

Transcript of Tumor-Intrinsic PD-L1 Signals Regulate Cell Growth ... · Growth, Pathogenesis, and Autophagy in...