Immune-checkpoint proteins VISTA and PD-1nonredundantly regulate murine T-cell responsesJun Liua,b, Ying Yuana,1, Wenna Chena, Juan Putrac, Arief A. Suriawinatac, Austin D. Schenkd, Halli E. Millera,Indira Guleriae, Richard J. Barthd, Yina H. Huangc, and Li Wanga,2

aDepartment of Microbiology and Molecular Genetics, Medical College of Wisconsin, Milwaukee, WI 53226; bJiangsu Center for Drug Screening, ChinaPharmaceutical University, Nanjing 210009, People’s Republic of China; cDepartment of Pathology and dDepartment of Surgery, Geisel School of Medicineat Dartmouth, Norris Cotton Cancer Center, Lebanon, NH 03756; and eBoston Children’s Hospital and Brigham and Women’s Hospital, Renal Division,Harvard Medical School, Boston, MA 02115

Edited by Michael J. Bevan, University of Washington, Seattle, WA, and approved April 2, 2015 (received for review October 23, 2014)

V-domain immunoglobulin suppressor of T-cell activation (VISTA)is a negative immune-checkpoint protein that suppresses T-cellresponses. To determine whether VISTA synergizes with anotherimmune-checkpoint, programmed death 1 (PD-1), this study charac-terizes the immune responses in VISTA-deficient, PD-1-deficient (KO)mice and VISTA/PD-1 double KO mice. Chronic inflammation andspontaneous activation of T cells were observed in both single KOmice, demonstrating their nonredundancy. However, the VISTA/PD-1double KO mice exhibited significantly higher levels of these pheno-types than the single KO mice. When bred onto the 2D2 T-cellreceptor transgenic mice, which are predisposed to developmentof inflammatory autoimmune disease in the CNS, the level ofdisease penetrance was significantly enhanced in the double KOmice compared with in the single KO mice. Consistently, the magni-tude of T-cell response toward foreign antigens was synergisticallyhigher in the VISTA/PD-1 double KO mice. A combinatorial blockadeusing monoclonal antibodies specific for VISTA and PD-L1 achievedoptimal tumor-clearing therapeutic efficacy. In conclusion, our studydemonstrates the nonredundant role of VISTA that is distinct fromthe PD-1/PD-L1 pathway in controlling T-cell activation. These find-ings provide the rationale to concurrently target VISTA and PD-1pathways for treating T-cell-regulated diseases such as cancer.

immune-checkpoint | autoimmunity | tumor immunity |cancer immunotherapy | T-cell activation

T-cell activation requires engagement with antigen-presentingcells (APCs) that present the cognate peptide on MHC mol-

ecules. Antigen recognition is regulated by cosignaling ligands andreceptors, whose integrated signaling determines the outcome ofT-cell activation, differentiation, and function (1). The B7 familyof coreceptors belongs to the Ig superfamily, consisting of stimu-latory receptors such as CD28 and inducible T-cell costimulator(ICOS), cytotoxic T-lymphocyte–associated protein 4 (CTLA-4) andprogrammed death 1 (PD-1) coinhibitory receptors, and B7-H3and B7-H4 ligands, whose receptors have yet to be identified (1, 2).Together with additional inhibitory molecules such as T-cell im-munoglobulin domain and mucin domain 3 (Tim-3), lymphocyte-activation gene 3 (LAG3), and B- and T-lymphocyte attenuator(BTLA), these immune-checkpoint proteins play critical roles inmaintaining peripheral tolerance and controlling autoimmunity.VISTA is a newly identified Ig domain-containing immune-

checkpoint molecule that directly suppresses T-cell activation invitro and in vivo (3, 4). The human and murine VISTA proteinsshare 90% identity and display similar expression patterns (5).VISTA is constitutively expressed within the hematopoietic com-partment, with the highest expression level on myeloid cellsand a lower level on CD4+, CD8+ T cells, and Foxp3+CD4+ reg-ulatory T cells. A soluble VISTA-Ig protein that contains theextracellular domain fused with Ig-crystalizable fragment (Fc) orfull-length VISTA expressed on APCs acts as a ligand to suppressT-cell proliferation and cytokine production (3). In addition,VISTA might function as an inhibitory receptor on T cells tosuppress their activation (6). VISTA-specific monoclonal antibody

(mAb) enhances disease severity in the experimental autoimmuneencephalomyelitis (EAE) model, as well as boosts antitumor im-munity in multiple murine tumor models (7).The regulatory functions of immune-checkpoints have been

demonstrated using KO mice, which typically manifest loss of pe-ripheral tolerance and heightened T-cell responses. For example,CTLA-4 KO mice die of young age because of overwhelming lym-phoproliferative disease and inflammation (8, 9). PD-1 KO micedisplay genetic background-dependent late-onset autoimmune dis-ease of either dilated cardiomyopathy or arthritis (10, 11). PD-1binds to two ligands, PD-ligand 1 (L1) and PD-L2 (12, 13). Althoughneither PD-L1 nor PD-L2 KO mice develop overt organ-specificautoimmune disease, PD-L1 genetic deficiency exacerbates diseasein the EAEmodel, as well as impairs fetomaternal tolerance (14, 15).These data suggest a critical role for PD-L1 in maintaining T-cellperipheral tolerance at tissue sites. In contrast, genetic disruption ofother immune checkpoint ligands such as B7-H4 fail to display anysignificant alterations of T-cell responses in vivo, suggesting there is ahierarchy and potential redundancy among various immune-check-point regulators (16). In this context, it is important to note that theexpression pattern of VISTA overlaps with multiple other immune-checkpoint regulators (3). It is therefore critical to determinewhether VISTA exerts nonredundant immune regulatory functions.Studies have shown that several immune-checkpoints function-

ally synergize with each other (17–19). For example, the combined

Significance

Multiple immune-checkpoint proteins, such as programmeddeath 1 (PD-1), LAG3, and TIM3, are coexpressed on immunecells and functionally synergize with each other. V-domainimmunoglobulin suppressor of T-cell activation (VISTA) is arecently identified immune-checkpoint molecule that sup-presses T-cell activation. This study establishes that VISTA andPD-1 exert nonredundant immune regulatory functions andsynergistically regulate T-cell responses. Combinatorial treat-ment using VISTA- and PD-ligand 1-specific monoclonal anti-bodies achieved synergistic therapeutic efficacy in murinetumor models. This study critically advances our knowledge ofthe immune regulatory function of VISTA and provides a ra-tionale for targeting both VISTA and PD-1 to more effectivelytreat T-cell-regulated diseases such as cancer.

Author contributions: L.W. designed research; J.L., Y.Y., W.C., J.P., A.D.S., H.E.M., and L.W.performed research; J.L., I.G., Y.H.H., and L.W. contributed new reagents/analytic tools;J.L., Y.Y., W.C., J.P., A.A.S., A.D.S., R.J.B., Y.H.H., and L.W. analyzed data; and L.W. wrotethe paper.

Conflict of interest statement: L.W. is involved with the commercial development ofVISTA with ImmuNext Inc Corporation and received research support, salary, and/orconsulting fees.

This article is a PNAS Direct Submission.1Present address: College of Pharmacy, Shanghai University of Traditional Chinese Med-icine, Shanghai 201203, People’s Republic of China.

2To whom correspondence should be addressed. Email: lilywang@mcw.edu.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1420370112/-/DCSupplemental.

6682–6687 | PNAS | May 26, 2015 | vol. 112 | no. 21 www.pnas.org/cgi/doi/10.1073/pnas.1420370112

Dow

nloa

ded

by g

uest

on

June

7, 2

020

disruption of LAG3 and PD-1 genes results in lethal autoimmunediseases, whereas loss of either gene alone leads to subtle pheno-types (17). In addition, PD-1 functionally synergizes with TIM3(18, 19). A combinatorial antibody-mediated blockade of bothPD-1 and TIM3 results in optimal T-cell responses against cancer,as well as during chronic viral infection (18, 19). In the current study,we use KO mice to address whether VISTA synergizes with PD-1 inregulating T-cell responses. Our data establish that VISTA and thePD-1/PD-L1 pathways nonredundantly regulate T-cell activationand demonstrate the feasibility of concurrently targeting these twoimmune-checkpoints to enhance tumor-specific immune responses.

ResultsLoss of T-Cell Peripheral Tolerance on Combined Genetic Disruption ofVISTA and PD-1 or PD-L1. To determine whether VISTA and PD-1regulate immune responses in a redundant or independent/syn-ergistic manner, Vista−/−Pdcd1−/− mice (herein referred to asVISTA/PD-1 double KO mice) were generated on the C57BL/6background and characterized. The double KO mice were bornfertile and produced normal litter sizes. Normal thymic de-velopment and lymphocyte populations (T and B lymphocytes,natural killer cells, and natural killer T cells) in the bone marrow,spleen, and lymph nodes were observed in 6–8-wk-old KO mice.Comprehensive multiorgan histological analyses were per-

formed in 12-mo-old WT, VISTA KO, PD-1 KO, and VISTA/PD-1 double KO mice (Fig. 1). Hematoxylin and eosin (H&E)stained sections from heart, lung, liver, kidney, pancreas, salivarygland, small and large intestines, and brain were examined.Several organs, including lung, liver, and pancreas in the doubleKO mice, were heavily infiltrated with leukocytes (Fig. 1A, Top)and showed significant tissue necrosis, presumably as a result ofimmune cell-mediated destruction (Fig. 1A, Bottom). The levelsof leukocyte infiltration and tissue necrosis in the KO mice were

blindly quantified on the basis of a semiquantitative scoringmethod, and the VISTA/PD-1 double KO mice showed thehighest scores compared with WT and single KO mice (Fig. 1B).Despite the significant accumulation of activated T cells, thedouble KO mice did not develop overt autoimmune disease.Serum levels of IgM and IgA were moderately elevated in ageddouble KO mice (Fig. S1). Our cohoused PD-1 KO mice alsodeveloped accumulation of spontaneously activated T cells andchronic inflammation in multiple organs but failed to developarthritis or other previously reported autoimmune phenotypes(10). This discrepancy might be a result of the different housingconditions or the age of the mice analyzed.Compared with VISTA KO and PD-1 KO mice, VISTA/PD-1

double KO mice at the age of 6–7 mo showed significantly in-creased frequencies of CD44hiCD62Llo CD8+ and CD4+ T cells,which is indicative of an activated or memory phenotype (Fig. 2A and C). On in vitro restimulation, the double KO T cellsproduced significantly higher levels of cytokines, such as IFNγ,TNFα, and IL-17A, than WT and single KO cells (Fig. 2 B, D–F).PD-1 binds to ligands PD-L1 and PD-L2 (20). To corroborate

the results seen in the VISTA/PD-1 double KO mice, we bredVISTA KO onto the previously described PD-L1 KO (15) andgenerated the VISTA/PD-L1 double KO mice. Our data dem-onstrated spontaneous activation of peripheral CD4+ and CD8+T cells in the VISTA/PD-L1 double KO mice, which was com-parable to that of VISTA/PD-1 double KO mice (Fig. S2 A andB). Together, these data support the conclusion that VISTA andthe PD-1/PD-L1 pathway nonredundantly control the peripheraltolerance of T cells.The phenotype of spontaneous T-cell activation in the double

KO mice (Fig. 2 and Fig. S2) indicates that both VISTA and PD-1regulate the threshold of TCR activation to autoantigens. Wehypothesize that disruption of both pathways will increase

Fig. 1. Histologic analysis of aged VISTA KO, PD-1 KO, and VISTA/PD-1 double KO mice. Necropsy was performed on 12-mo-old WT (n = 16), VISTA KO (n = 15),PD-1 KO (n = 28), and VISTA/PD-1 double KO (n = 25) mice. Organs were fixed, paraffin embedded, sectioned, and stained with H&E. Two representative H&Esections from lung, liver, and pancreas of the VISTA/PD-1 double KO mice are shown in A. Clusters of tissue-infiltrating leukocytes were marked with blackarrows. (Top) Areas of necrotic tissues were marked with white arrows. (Bottom). All images are of 200×magnification. (Scale bar: 50 μ.) The inflammatory stateof the tissues was evaluated on the basis of a semiquantitative method that scores the level of the leukocyte infiltration and tissue necrosis (B).

Liu et al. PNAS | May 26, 2015 | vol. 112 | no. 21 | 6683

IMMUNOLO

GYAND

INFLAMMATION

Dow

nloa

ded

by g

uest

on

June

7, 2

020

predisposition to autoimmune disease on a susceptible back-ground. To test this hypothesis, the VISTA/PD-1 double KOmice were bred with the 2D2 mice, which express a TCR trans-gene specific for the self-antigen, myelin oligodendrocyte gly-coprotein (MOG35–55) (21). Previous studies reported that 4%of 2D2 mice developed spontaneous EAE between the ages of 3and 5 mo (21). A similar incidence of spontaneous EAE (1/30,∼3%) was observed in our colony of 2D2-WT mice younger than6 mo (Fig. 3A and Table S1). 2D2-PD-1 KO mice showed similarincidence of spontaneous disease as WT mice (2/40, 5%). Incontrast, genetic deficiency of VISTA accelerated disease onset,such that ∼60% (25/42) of the 2D2-VISTA KO mice rapidlysuccumbed to complete hind limb paralysis within 3 mo. Combineddeficiency of VISTA and PD-1 further increased disease in-cidence to ∼90% (35/37) (Fig. 3 A and B). Analysis of CNS tissuefrom paralyzed 2D2-VISTA/PD-1 double KO mice confirmedthe accumulation of inflammatory cells and demyelination (Fig.3 C and D). The disease onset age ranged between 4 and 16 wk in

the 2D2-VISTA/PD-1 double KO mice, which was similar to theages observed in the 2D2-VISTA KO mice (5–16 wk) (Fig. S3).Only one WT mouse developed disease around 12 wk of age. Asmall percentage of diseased mice from the 2D2-VISTA KO (2/40)and 2D2-VISTA/PD-1 double KO (3/37) mice developed atypicalEAE, manifested as unilateral paralysis rather than bilateral pa-ralysis. Detailed information regarding disease onset, severity,penetrance, and mortality is presented in Table S1.

Combined Disruption of VISTA and PD-1 Synergistically AugmentsT-Cell Responses on Antigen Challenge. Spontaneous T-cell activa-tion and enhanced autoimmunity in aged VISTA/PD-1 doubleKO mice indicate that VISTA and PD-1 control T-cell tolerancetoward autoantigens. We hypothesize that these pathways alsocritically regulate T-cell responses toward foreign antigens. Toaddress this question, mice were immunized with soluble anti-genic peptides together with poly (I:C) (TLR3 agonist) as adju-vant. 2W1S, an MHC class II-restricted peptide, and OVA257-264

,an MHC class I-restricted peptide, were used (22). On day +7postimmunization, splenic T cells were isolated and restimulatedex vivo with the respective peptides. Significantly higher numbersof IFNγ-producing T cells were present in the VISTA/PD-1double KO mice compared with in WT or single KO mice, in-dicating that VISTA and PD-1 nonredundantly control T-cellresponses (Fig. 4 A and B).We have previously reported that VISTA functions as a ligand

that engages an unknown receptor on T cells and suppressesT-cell activation (3). Because both VISTA and PD-L1 are highlyexpressed on CD11b+ myeloid APCs (2, 3), we hypothesize thata combined deficiency of VISTA and PD-L1 on APCs maximallyenhances T-cell activation. To test this hypothesis, CD11b+

myeloid DCs were isolated from WT, VISTA KO, PD-L1 KO,and the VISTA/PD-L1 double KO mice (15, 23) and were usedto stimulate naive CD4+ OTII TCR transgenic T cells in thepresence of cognate peptide OVA323–339. Our data show that thecombined deficiency of VISTA and PD-L1 on myeloid APCssynergistically enhanced T-cell proliferation and IFNγ pro-duction (Fig. 4 C and D).

A B

DC

E

CD8 + T cells

F

CD4 + T cells

CD4 + T cells

IL-1

7A+ %

wt

PD-KO

V-KO

double-KO wt

PD-KO

V-KO

double-KO

0.0

0.5

1.0

1.5

2.0

2.5

Spleen Blood

nsp< 0.0001

p< 0.0001

nsp: 0.001

p< 0.0001

IFNγ+ T

NFa

+ %

wt

PD-KO

V-KO

double-KO wt

PD-KO

V-KO

double-

KO0

5

10

15

20p: 0.002

p: 0.02p: 0.01

p: 0.003

nsns

Spleen Blood

IFNγ+ %

wt

PD-KO

V-KO

double-KO wt

PD-KO

V-KO

double-KO

0

5

10

15

20p< 0.0001

nsp: 0.02

p: 0.01

nsns

Spleen Blood

CD

44hi

CD

62Llo

w%

wt

PD-KO

V-KO

double-KO wt

PD-KO

V-KO

double-KO

0

20

40

60

80p< 0.0001 p< 0.0001

p:0.0002p:0.002

nsp:0.01

Spleen Blood

CD

44hi

CD

62Llo

w%

wt

PD-KO

V-KO

double-KO wt

PD-KO

V-KO

double-

KO0

20

40

60

80

100p< 0.0001

p< 0.0001p: 0.01

p< 0.0001

p< 0.0001p: 0.01

Spleen Blood

IFNγ+ %

wt

PD-KO

V-KO

double-KO wt

PD-KO

V-KO

double-KO

0

10

20

30p: 0.0007

p: 0.001p: 0.008

p: 0.001

p: 0.007p: 0.04

Spleen Blood

Fig. 2. Spontaneous T-cell activation in the VISTA KO, PD-1 KO, and VISTA/PD-1 double KO mice. Splenic T cells were collected from age- and sex-matched 6–7-mo-old WT (n = 6), VISTA KO (n = 4), PD-1 KO (n = 6), andVISTA/PD-1 double KO (n = 8) mice. The percentages of CD8+ and CD4+ Tcells with activated phenotype (CD44hi CD62Llo) were quantified by flowcytometry. T cells were stimulated ex vivo overnight with soluble anti-CD3/CD28 mAbs, and their cytokine production (i.e., IFNγ, TNFα, and IL-17A) wasexamined by intracellular staining. CD8+ T-cell phenotypes were shown in Aand B. CD4+ T-cell phenotypes were shown in C–F. Representative results ofat least three independent experiments were shown.

Fig. 3. Combined genetic deficiency of VISTA and PD-1 exacerbated auto-immune disease on the susceptible background. The CNS disease incidence(A) and mortality (B) were monitored in 2D2 TCR transgenic mice that werebred onto the VISTA KO, PD-1 KO, and double KO genetic background.Representative H&E stained spinal cord section from paralyzed double KOmice is shown (C). Enlarged images show areas of extensive lymphocyteinfiltration. Luxol fast blue staining of spinal cord sections confirmed ex-tensive demyelination (D). 2D2-WT (n = 30), 2D2-VISTA KO (n = 42), 2D2-PD-1KO (n = 40), 2D2-VISTA/PD-1 double KO (n = 37). Only one 2D2-WT mousedeveloped disease.

6684 | www.pnas.org/cgi/doi/10.1073/pnas.1420370112 Liu et al.

Dow

nloa

ded

by g

uest

on

June

7, 2

020

Although the receptor for VISTA (VISTA-R) is unknown, wespeculate that the engagement of VISTA-R on T cells suppressesTCR signaling independent of PD-1. In addition, we hypothesizethat the coengagement of VISTA-R and PD-1 on T cells syn-ergistically impairs TCR signaling. To test these hypotheses,proximal TCR signaling events were examined using immobi-lized fusion proteins VISTA-Ig and PD-L1-Ig, both of whichsuppressed T-cell proliferation and cytokine production in vitro(3, 12). LAT is a proximal signaling adaptor that is phosphory-lated on TCR stimulation and forms a complex with multiple sig-naling molecules including SH2 domain containing leukocyteprotein of 76kDa (SLP76) and phospholipase C (PLC)-γ1 (24). Todetermine whether VISTA functions by interfering with thephosphorylation of LAT, a solid phase immunoprecipitation assaywas performed, in which the proximal signaling complexes couldbe recovered as the bound fraction to the plastic surface (25, 26).Our data show that in the presence of coimmobilized control-Ig orVISTA-Ig proteins, plate-bound anti-CD3e mAb pulled-downcomparable amounts of CD3ζ. This result excluded the possibilitythat the immobilized VISTA-Ig displaced anti-CD3e mAb orimpaired the binding of anti-CD3emAb to the TCR/CD3 complex(Fig. 5A). Despite the similar engagement of CD3ζ, immobilizedVISTA-Ig significantly reduced the amount of LAT recruited tothe CD3 complex, as well as its phosphorylation (Fig. 5A). Whentotal cell lysates were examined, the phosphorylation of severalproximal and downstream signaling molecules such as SLP76,PLC-γ1, Akt, and Erk1/2 were also impaired (Fig. 5B).Next, the effect of coengaging VISTA-Ig and PD-L1-Ig was

examined (Fig. 5 C and D). Preactivated T cells were analyzed,

as they expressed a high level of PD-1 (27). Consistent withour hypothesis, the combined engagement of VISTA-Ig and PD-L1-Ig maximally reduced the phosphorylation of LAT and itsrecruitment to the CD3 complex (Fig. 5C). Furthermore, VISTA-Igand PD-L1-Ig coengagement maximally reduced the phosphoryla-tion of SLP76, PLC-γ1, Akt, and Erk1/2 in total cell lysates (Fig. 5D).Together, these results support the conclusion that VISTA-Rand PD-1 each impairs early TCR signaling and results in themost robust suppression when combined.Enhanced T-cell responses in the VISTA/PD-1 double KO

mice suggest that these two immune checkpoints could be tar-geted concurrently to optimize antitumor immunity. This hy-pothesis was tested in transplantable murine tumor models (Fig. 6).

BA

DC

IFNγ

(pg/

ml)

WT

VISTA-KO

PD-LI-KO

VISTA/PD-L1-KO

0

20

40

60

80

100p = 0.01nsns

Spot

s / 5

00k

Sple

en c

ells

WT

VISTA-KO

PD1-KO

VISTA/PD1-KO

0

150

300

450

600

750

WTVISTA-KOPD1-KOVISTA/PD1-KO

p = 0.04

p = 0.005

p < 0.0001

OVA257-264

Spot

s / 5

00k

Sple

en c

ells

WT

VISTA-KO

PD1-KO

VISTA/PD1-KO

0

100

200

300

400

500

600

p = 0.06

p = 0.02

p = 0.022W1S

Ratio (APC:T cells)

CPM

1:3 1:60

50000

100000

150000

200000nsp = 0.003

nsnsp = 0.003

ns

Fig. 4. VISTA and the PD-1 collaboratively controlled antigen-specific T-cellresponses. Six- to 7-wk-old WT (n = 8), VISTA KO (n = 9), PD-1 KO (n = 7), andVISTA/PD-1 double KO (n = 6) mice were immunized with 50 μg solublepeptides OVA257-264 (A) or 2W1S (B), together with TLR3 agonist poly (I:C)(100 μg) as adjuvant. Splenocytes were harvested on day +7 postimmunizationand restimulated with the respective peptides. IFNγ-producing cells wereenumerated by the ELISPot assay. To stimulate T cells in vitro, CD11b+ CD11c+

DCs were sorted from WT, VISTA KO, PD-L1 KO, and VISTA/PD-L1 double KOmice and incubated with naive CD4+ OTII TCR transgenic T cells in the presenceof cognate peptides OVA323–339 (10 ng/mL). [3H]-Thymidine was added to theculture for the last 8 h of the 72-h culture period for measuring T-cell pro-liferation (C). The production of IFNγ was quantified from the culture super-natants by ELISA (D).

A

p-LAT

Total-LAT

p-SLP76

Total-SLP76

p-PLCγ-1

Total-PLCγ-1

p-Akt

Total-Akt

p-Erk1/2

Total-Erk1/2

Control-Ig

VISTA-Ig

Rested

B

p-LAT

Total-LAT

CD3zeta

Control-Ig

VISTA-Ig

C

D

p-LAT

Total-LAT

CD3zeta

Control-Ig

VISTA-Ig

PD-L1-Ig

combo

no-stim

p-LAT

Total-LAT

p-SLP76

Total-SLP76

p-Akt

Total-Akt

p-Erk1/2

Total-Erk1/2

Control-Ig

VISTA-Ig

PD-L1-Ig

combo

no-stim

p-PLCγ-1

Total-PLCγ-1

DO11.10 cells DO11.10 cells

CD4+CD25- T cells CD4+CD25- T cells

Fig. 5. Engagement of both VISTA and PD-L1 during TCR activation maxi-mally suppressed TCR signaling. To determine whether VISTA engagementimpairs the recruitment of signaling adaptor protein LAT, DO11.10 hybrid-oma cells (100 × 106) were stimulated with plate-bound anti-CD3 mAb (2C11,3 μg/mL), together with coimmobilized control-Ig (8 μg/mL) or VISTA-Ig fu-sion protein (8 μg/mL) for 10 min at 37 °C and lysed in situ. After removingthe unbound cell lysates, plate-bound protein was eluted off the plate andexamined by Western blotting (A). To examine the effect of VISTA on thephosphorylation of TCR signaling molecules, CD25−CD4+ T cells were puri-fied from naive splenocytes and stimulated with plate-bound 2C11 (3 μg/mL),together with control-Ig (8 μg/mL) or VISTA-Ig (8 μg/mL) for 5 min at 37 °C.Total cell lysates were prepared, and the phosphorylation status of LAT,SLP76, PLC-γ1, Akt, and Erk1/2 was examined (B). To determine whethercoengagement of both VISTA and PD-L1 maximally suppresses LAT activa-tion, DO11.10 cells were stimulated with plate-bound 2C11 (2.5 μg/mL), to-gether with control-Ig (10 μg/mL), VISTA-Ig (5 μg/mL), PD-L1-Ig (5 μg/mL), orboth Ig fusion proteins. Cells were lysed after 10 min stimulation, and plate-bound proteins were recovered and examined as described earlier (C). Todetermine the synergistic effects of engaging both VISTA and PD-L1, pre-activated splenic CD4+ T cells were stimulated with plate-bound 2C11(2.5 μg/mL), together with control-Ig (9 μg/mL), VISTA-Ig (3 μg/mL), PD-L1-Ig (6 μg/mL), or both Ig fusion proteins for 10 min at 37 °C. Total celllysates were harvested for Western blotting analysis (D). Representativeresults from two to three independent experiments were shown.

Liu et al. PNAS | May 26, 2015 | vol. 112 | no. 21 | 6685

IMMUNOLO

GYAND

INFLAMMATION

Dow

nloa

ded

by g

uest

on

June

7, 2

020

Our data show that in the CT26 colon cancer model, thecombinatorial treatment of VISTA and PD-L1 mAbs on day +2after tumor inoculation led to tumor regression and long-termsurvival (8/8), whereas either mAb alone was less effective (1/8 in

the VISTA mAb-treated group and 3/8 in the PD-L1 mAb-treatedgroup rejected tumor) (Fig. 6A). Analysis of tumor-specific T-cellactivation showed synergistically enhanced cytokine production(IFNγ and TNFα) and granzyme B production by tumor-specificCD8+ T cells from tumor-draining LN (Fig. 6B). Next, combi-natorial treatment was tested on larger established tumors. Whentreatment was initiated on day +5 after CT26 inoculation, theaverage tumor size reached ∼4–5 mm in diameter (Fig. 6C). Tofacilitate tumor-specific T-cell activation in this therapeutic set-ting, mice were treated with anti-CD25 mAb on day +5 andday + 20 to transiently deplete CD25+Foxp3+CD4+ regulatoryT cells. Synergistic efficacy was achieved with the combinedtreatment of VISTA and PD-L1 mAbs, which led to ∼80% tumorregression (8/10), whereas single mAb treatment failed to showany significant effect (Fig. 6C).We next evaluated the therapeutic efficacy of the combina-

torial treatment in a less immunogenic tumor model, such as theB16BL6 melanoma model, which is responsive to therapeuticblockade of CTLA-4 and PD-1 in previous studies (28–31). AGM-CSF secreting cellular vaccine (GVAX) (28) was applied onday +3, +6, +9, and +12 after tumor inoculation to boost tumor-specific T-cell responses. A sublethal whole-body irradiation(250 rads) was applied on day +3, which was shown to facilitateT-cell-mediated antitumor immune responses (32). Consistentwith our hypothesis, treatment with both VISTA and PD-L1mAbs significantly suppressed tumor growth and conferred sur-vival advantage, whereas single mAb treatment was largely in-effective (Fig. 6D). Together, these data indicate that VISTAand PD-L1/PD-1 pathways independently control tumor-specificT-cell responses, and combined therapeutic blockade synergis-tically enhances antitumor immunity.

DiscussionVISTA and PD-1 both function as immune checkpoint proteinsthat suppress T-cell activation. They share overlapping expressionpatterns within the hematopoietic compartment. It is thereforeimportant to define their independent immune-regulatory roles.Evidence based on studies of the KO mice indicates that VISTA

and PD-1 nonredundantly regulate immune responses. Geneticdisruption of VISTA accelerated autoimmune disease on a sus-ceptible background, as well as resulted in multiorgan chronic in-flammation because of spontaneous T-cell activation (22). AgedPD-1 KO mice were reported to develop late-onset autoimmunityin the C57BL/6 background (10). Our study of PD-1 KO micealso showed accumulation of spontaneously activated T cellsand chronic inflammation in multiple organs. Furthermore, in thecurrent study of the VISTA/PD-1 double KO mice, we providedstrong evidence for independent control of T-cell responses bythese two checkpoints. Synergistic or additive T-cell activationwas observed from aged double KO mice compared with thesingle KO mice, which might reflect the lack of “brakes” in TCRsignaling, resulting in lower threshold of T-cell activation and lossof peripheral tolerance to autoantigens. Similarly, synergisticT-cell responses were observed in double KO mice in response toforeign antigens.We hypothesize that VISTA and PD-1 both function as brakes

for T-cell activation. Because APCs lacking both VISTA andPD-L1 stimulated T cells better than single KO APCs or WTAPCs in vitro, these data indicate that VISTA and PD-L1 en-gage independent inhibitory receptors on T cells. To determinethe effects of VISTA-R and PD-1 on TCR signaling, immobi-lized VISTA-Ig and PD-L1-Ig were used to engage VISTA-Rand PD-1, respectively. Our data show that VISTA-Ig and PD-L1-Ig fusion proteins impaired the activation of LAT, as well asthe phosphorylation of proximal signaling molecules (SLP76 andPLC-γ1) and downstream molecules (Akt and Erk1/2). Thecoengagement of both Ig fusion proteins to their receptorsresulted in additive effects. On the basis of these data, we con-clude that similar to PD-1, VISTA-R impairs early TCR sig-naling. PD-1 has been shown to accumulate at the immunesynapse and recruits Src homology region 2 domain-containing

A

B

C

D

IFNγ+ T

NFα

+ % C

D8+

contro

l-Ig

aVISTA

aPD-L1

combo

0

5

10

15

20p = 0.04

ns

ns

IFNγ+ G

ranz

B+ %

CD

8+

control-Ig

aVISTA

aPD-L1

combo

0

5

10

15

20

25p = 0.018

p = 0.018

ns

IFNγ+ %

CD8

+ce

lls

contro

l-Ig

aVISTA

aPD-L1

combo

0

10

20

30

40

50p = 0.01

p = 0.02

P = 0.05

Time (Days)

Tum

or s

ize

( mm

2 )

10 20 30 400

50

100

150

200

250

control-Ig (n=8)anti-VISTA (n=8)anti-PD-L1 (n=8)combo (n=8)

****

****

******

**Time (Days)

Perc

ent s

urvi

val

0 20 40 60 80 1000

25

50

75

100control-Iganti-VISTAanti-PD-L1combo

Time (Days)

tum

or s

ize

(mm

2 )

9 12 15 18 210

50

100

150

200

notreatGvax/RT + control-IgGvax/RT + anti-PD-L1Gvax/RT + anti-VISTAGvax/RT + combo

ns *****

**

****

(n=10/group)

Time (Days)

tum

or s

ize

(mm

2 )

0 4 8 12 16 20 240

50

100

150

200

250

control-Ig (n= 9)aPD-L1 (n= 10)aVISTA (n= 10)combo (n= 10)

*********

Time (Days)

Perc

ent s

urvi

val

0 20 40 60 800

20

40

60

80

100

control-IgaPD-L1aVISTAcombo

Time (Days)

Perc

ent s

urvi

val

0 20 40 600

25

50

75

100

notreatGvax/RT + control-IgGvax/RT+ anti-PD-L1Gvax/RT + anti-VISTAGvax/RT + combo

Fig. 6. Optimal therapeutic efficacy on combined blockade of VISTA andPD-L1 in murine tumor models. CT26 colon carcinoma cells (100,000) wereinoculated on the flank of naive mice on day 0 (n = 8). Day +3 after tumorinoculation, mice were treated with control Ig (300 μg), anti-VISTA mAb(300 μg), anti-PD-L1 mAb (200 μg), or combined anti-VISTA and anti-PD-L1mab, every 2–3 d continuously for 3 wk. Tumor size was measured by a caliperand recorded as area (mm2) (A). The rate of tumor-free survival was alsoshown (A). To examine tumor-specific T-cell responses, lymphocytes wereharvested from tumor-draining lymph nodes on day +14 after tumor in-oculation (1 × 105), when average tumor size reached ∼8–10mm. Expression ofIFNγ, TNFα, and granzyme B by CD8+ T cells on stimulation with irradiated tumorcells was detected by flow cytometry (B). To evaluate the efficacy of antibodytreatment on larger established tumors, mice were inoculated with a higherdose of CT26 cells (2.5 × 105). On day +5 after tumor inoculation, when tumorsize reached ∼4–5 mm, mice were treated with VISTA or PD-L1 mAb or both, asdescribed earlier. Mice were also treated with anti-CD25 specific antibody onday +5 and day +20 to transiently deplete Foxp3+CD4+ Tregs. Tumor growthand survival of the CT26 tumor-bearing mice were monitored and shown (C).A less immunogenic B16BL6 melanoma tumor model was examined to furthervalidate the efficacy of the combinatorial treatment. Mice were inoculated withB16BL6 (25,000) cells. On day +3 after tumor inoculation, mice were conditionedwith low-dose irradiation (250 rads) and treated with four doses of GVAX beforethe treatment with either VISTA or PD-L1 mAb or both. Tumor growth andsurvival of tumor-bearing mice were monitored and shown (D). Representativeresults from two to three independent experiments were shown.

6686 | www.pnas.org/cgi/doi/10.1073/pnas.1420370112 Liu et al.

Dow

nloa

ded

by g

uest

on

June

7, 2

020

phosphatase (SHP)-1/2 to down-regulate TCR signaling (33, 34).Whether or not similar phosphatases might be involved inVISTA-R-mediated effects remains to be determined.We speculate that multiple mechanisms underlie the syner-

gistic T-cell activation when both VISTA and PD-1 areblocked in vivo. In addition to being a ligand, VISTA acts as areceptor that transduces inhibitory signals during T-cell acti-vation (6). This function likely contributes to the synergisticT-cell activation seen in the VISTA/PD-1 double KO mice.Furthermore, VISTA might exert T-cell extrinsic functions.VISTA is highly expressed on myeloid cells such as Cd11b+DCs and macrophages (3). Our future studies will dissectlineage-specific roles of VISTA in controlling both innate andadaptive immune responses.Additional immune regulatory molecules, such as LAG3 and

Tim3, have been shown to synergize with PD-1 to control T-cellresponses (19, 35). Our current study shows that VISTA andPD-1 synergistically regulate T-cell responses against self- andforeign antigens, and concurrently targeting both moleculesleads to optimal therapeutic efficacy in murine tumor models.The absence of overt autoimmune disease in the double KOmice suggests that the combinatorial blockade of VISTA andPD-1 might achieve optimal therapeutic efficacy with less severeimmune-related adverse events, and therefore might be moreamenable for the treatment of cancer.

Materials and MethodsMice. C57BL/6 mice were purchased from Charles River Laboratories. VISTAKO mice on a fully backcrossed C57BL/6 background were obtained from theMutant Mouse Regional Resource Centers (University of California–Davis)(36). 2D2 TCR transgenic mice were purchased from the Jackson Laboratory.PD-1 KOmice were provided by Dr. Honjo (10). PD-L1 KOmice were as described(15). All animals were maintained in a pathogen-free facility at the Medical

College of Wisconsin. All animal protocols were approved by the InstitutionalAnimal Care and Use Committee of the Medical College of Wisconsin.

Mice Necropsy and Semiquantitative Pathological Analysis.Age- and sex-matchedWT and VISTA KOmice were killed by CO2 asphyxiation. Organswere harvestedand fixed in 10% (vol/vol) buffered formalin. H&E stain was performed ontissue sections. Tissue inflammatory status was scored in a blind manner by apathologist using the following semiquantitative scoring criteria: 0 = normal;1 = mild/small foci of dense lymphocytic infiltrate; 2 = moderate/multiple fociof dense/large activated lymphocytic infiltrate with/without germinal center;and 3 = marked reactive/activated or atypical lymphocytic infiltrate.

Flow Cytometry and Data Analysis. Flow cytometry analysis was performedeither on FACScalibur or LSRII, using CellQuest software (BD Bioscience). Dataanalyses were performed using FlowJo software (Treestar).

Graphs and Statistical Analysis. All graphs and statistical analysis were generatedusing Prism 4 (GraphPad Software, Inc.). Student’s t test (two tailed) or two-wayANOVA were used for data analyses. ***P < 0.005, **P < 0.025, *P < 0.05.

Additional materials and methods are provided in the SI Materialsand Methods.

ACKNOWLEDGMENTS. We thank Dr. Miyuki Azuma of Tokyo Medical andDental University for providing MIH5 hybridoma (anti-PD-L1 mAb),Dr. Tasuku Honjo (Kyoto University) for PD-1 KO mice, and Dr. PhilippaMarrack (National Jewish Health) for DO11.10 T-cell hybridoma. Weappreciate the experimental protocols, discussions, and manuscript editingprovided by Dr. Subramaniam Malarkannan during manuscript preparation.This study was supported by research funding from National Cancer InstituteGrant CA164225 (to L.W.), NIH Grant R01 AI089805 (to Y.H.H.), the Advancinga Healthier Wisconsin Research and Education Program fund (L.W.), and theMelanoma Research Foundation Career Development Award (to L.W.). Thiswork was also supported by the Office of the Assistant Secretary of Defensefor Health Affairs through the Peer Reviewed Cancer Research Program underAward No. W81XWH-14-1-0587 (to L.W.).

1. Chen L, Flies DB (2013) Molecular mechanisms of T cell co-stimulation and co-inhibition. Nat Rev Immunol 13(4):227–242.

2. Greenwald RJ, Freeman GJ, Sharpe AH (2005) The B7 family revisited. Annu Rev Im-munol 23:515–548.

3. Wang L, et al. (2011) VISTA, a novel mouse Ig superfamily ligand that negativelyregulates T cell responses. J Exp Med 208(3):577–592.

4. Flies DB, Wang S, Xu H, Chen L (2011) Cutting edge: A monoclonal antibody specificfor the programmed death-1 homolog prevents graft-versus-host disease in mousemodels. J Immunol 187(4):1537–1541.

5. Lines JL, et al. (2014) VISTA is an immune checkpoint molecule for human T cells.Cancer Res 74(7):1924–1932.

6. Flies DB, et al. (2014) Coinhibitory receptor PD-1H preferentially suppresses CD4⁺ Tcell-mediated immunity. J Clin Invest 124(5):1966–1975.

7. Le Mercier I, et al. (2014) VISTA Regulates the Development of Protective AntitumorImmunity. Cancer Res 74(7):1933–1944.

8. Tivol EA, et al. (1995) Loss of CTLA-4 leads to massive lymphoproliferation and fatalmultiorgan tissue destruction, revealing a critical negative regulatory role of CTLA-4.Immunity 3(5):541–547.

9. Waterhouse P, et al. (1995) Lymphoproliferative disorders with early lethality in micedeficient in Ctla-4. Science 270(5238):985–988.

10. Nishimura H, Nose M, Hiai H, Minato N, Honjo T (1999) Development of lupus-likeautoimmune diseases by disruption of the PD-1 gene encoding an ITIM motif-carryingimmunoreceptor. Immunity 11(2):141–151.

11. Nishimura H, et al. (2001) Autoimmune dilated cardiomyopathy in PD-1 receptor-deficient mice. Science 291(5502):319–322.

12. Freeman GJ, et al. (2000) Engagement of the PD-1 immunoinhibitory receptor by anovel B7 family member leads to negative regulation of lymphocyte activation. J ExpMed 192(7):1027–1034.

13. Latchman Y, et al. (2001) PD-L2 is a second ligand for PD-1 and inhibits T cell acti-vation. Nat Immunol 2(3):261–268.

14. Keir ME, et al. (2006) Tissue expression of PD-L1 mediates peripheral T cell tolerance.J Exp Med 203(4):883–895.

15. Guleria I, et al. (2005) A critical role for the programmed death ligand 1 in fetoma-ternal tolerance. J Exp Med 202(2):231–237.

16. Zhu G, et al. (2009) B7-H4-deficient mice display augmented neutrophil-mediatedinnate immunity. Blood 113(8):1759–1767.

17. Okazaki T, et al. (2011) PD-1 and LAG-3 inhibitory co-receptors act synergistically toprevent autoimmunity in mice. J Exp Med 208(2):395–407.

18. Jin HT, et al. (2010) Cooperation of Tim-3 and PD-1 in CD8 T-cell exhaustion duringchronic viral infection. Proc Natl Acad Sci USA 107(33):14733–14738.

19. Sakuishi K, et al. (2010) Targeting Tim-3 and PD-1 pathways to reverse T cell ex-haustion and restore anti-tumor immunity. J Exp Med 207(10):2187–2194.

20. Keir ME, Butte MJ, Freeman GJ, Sharpe AH (2008) PD-1 and its ligands in toleranceand immunity. Annu Rev Immunol 26:677–704.

21. Bettelli E, et al. (2003) Myelin oligodendrocyte glycoprotein-specific T cell receptortransgenic mice develop spontaneous autoimmune optic neuritis. J Exp Med 197(9):1073–1081.

22. Wang L, et al. (2014) Disruption of the immune-checkpoint VISTA gene imparts aproinflammatory phenotype with predisposition to the development of autoimmu-nity. Proc Natl Acad Sci USA 111(41):14846–14851.

23. Wang L, et al. (2008) Programmed death 1 ligand signaling regulates the gener-ation of adaptive Foxp3+CD4+ regulatory T cells. Proc Natl Acad Sci USA 105(27):9331–9336.

24. Dustin ML, Depoil D (2011) New insights into the T cell synapse from single moleculetechniques. Nat Rev Immunol 11(10):672–684.

25. Lysechko TL, Ostergaard HL (2005) Differential Src family kinase activity requirementsfor CD3 zeta phosphorylation/ZAP70 recruitment and CD3 epsilon phosphorylation.J Immunol 174(12):7807–7814.

26. Muhammad A, et al. (2009) Sequential cooperation of CD2 and CD48 in the buildupof the early TCR signalosome. J Immunol 182(12):7672–7680.

27. Pentcheva-Hoang T, Chen L, Pardoll DM, Allison JP (2007) Programmed death-1concentration at the immunological synapse is determined by ligand affinity andavailability. Proc Natl Acad Sci USA 104(45):17765–17770.

28. Leach DR, Krummel MF, Allison JP (1996) Enhancement of antitumor immunity byCTLA-4 blockade. Science 271(5256):1734–1736.

29. Holmgaard RB, Zamarin D, Munn DH, Wolchok JD, Allison JP (2013) Indoleamine 2,3-dioxygenase is a critical resistance mechanism in antitumor T cell immunotherapytargeting CTLA-4. J Exp Med 210(7):1389–1402.

30. Spranger S, et al. (2014) Mechanism of tumor rejection with doublets of CTLA-4, PD-1/PD-L1, or IDO blockade involves restored IL-2 production and proliferation of CD8(+)T cells directly within the tumor microenvironment. J Immunother Cancer 2:3.

31. Curran MA, Montalvo W, Yagita H, Allison JP (2010) PD-1 and CTLA-4 combinationblockade expands infiltrating T cells and reduces regulatory T and myeloid cellswithin B16 melanoma tumors. Proc Natl Acad Sci USA 107(9):4275–4280.

32. Klebanoff CA, Khong HT, Antony PA, Palmer DC, Restifo NP (2005) Sinks, suppressorsand antigen presenters: How lymphodepletion enhances T cell-mediated tumor im-munotherapy. Trends Immunol 26(2):111–117.

33. Sheppard KA, et al. (2004) PD-1 inhibits T-cell receptor induced phosphorylation ofthe ZAP70/CD3zeta signalosome and downstream signaling to PKCtheta. FEBS Lett574(1-3):37–41.

34. Chemnitz JM, Parry RV, Nichols KE, June CH, Riley JL (2004) SHP-1 and SHP-2 associatewith immunoreceptor tyrosine-based switch motif of programmed death 1 uponprimary human T cell stimulation, but only receptor ligation prevents T cell activation.J Immunol 173(2):945–954.

35. Woo SR, et al. (2012) Immune inhibitory molecules LAG-3 and PD-1 synergistically reg-ulate T-cell function to promote tumoral immune escape. Cancer Res 72(4):917–927.

36. Tang T, et al. (2010) A mouse knockout library for secreted and transmembraneproteins. Nat Biotechnol 28(7):749–755.

Liu et al. PNAS | May 26, 2015 | vol. 112 | no. 21 | 6687

IMMUNOLO

GYAND

INFLAMMATION

Dow

nloa

ded

by g

uest

on

June

7, 2

020