TIGIT-Fc Promotes Antitumor Immunity

T-cell immunoreceptor with Ig and ITIM domains (TIGIT, also known as WUCAM, Vstm3, or VSIG9), an inhibitory receptor expressed on lymphocytes belonging to the receptor of the Ig superfamily regulatory network that involves multiple players (e.g., CD96/TACTILE, CD112R/PVRIG), is a major emerging target for cancer immunotherapy (1). As an inhibitor of antitumor responses, TIGIT has the capacity to hinder multiple steps of the cancer immunity cycle, and therefore, it is of interest in the development of a first-in-class checkpoint-blocking drug (2); new clinical data have shown that combination therapy with the anti-TIGIT antibody tiragolumab combined with atezolizumab appears safe and effective against non–small cell lung cancer (3). Several mechanisms of action have been proposed for TIGIT-mediated inhibition of effector T cells and natural killer (NK) cells and the suppression of tumor-specific immunity. TIGIT is reported to interfere with the costimulation effect of DNAM-1 (4) or to directly deliver inhibitory signals to effector cells (5). TIGIT is also reported to enhance the suppressive functions of regulatory T cells (Treg) and hence, has the potential to inhibit a wide range of immune cells (6, 7).

TIGIT-Fc is an Fc fusion protein in which the extracellular domain of TIGIT is fused genetically to the Ig Fc domain. This genetic method enables Fc fusion proteins to have some antibody-like properties, such as long serum half-life and easy expression and purification, making them an attractive platform for research agents and therapeutic drugs (8). In a previous report, TIGIT-Fc showed an immunosuppressive effect by inhibiting T-cell activation in vitro in a dendritic cell (DC)–dependent manner and inhibited delayed-type hypersensitivity reactions in mice (9). We previously also showed that TIGIT-Fc demonstrates a therapeutic effect in a mouse model of lupus (10). Interestingly, TIGIT-Fc is also shown to promote NK-cell activation (11) and has no effect on cytokine production by tumor-specific CD8⁺ T cells in vitro (12), indicating that TIGIT-Fc may be a multifaceted immunomodulator and that its function may be dependent on the immune microenvironment. In particular, the impact of TIGIT-Fc on antitumor immunity is currently unknown.

Here, we found that TIGIT-Fc reduced the growth of human tumors in a xenograft model containing coimplanted human T cells by supporting a stimulatory immunologic mechanism of action for TIGIT-Fc. TIGIT-Fc treatment alone was sufficient to delay tumor growth in vivo and reverse the exhaustion of antitumor T cells and NK cells in multiple tumor models. We further demonstrated that TIGIT-Fc not only directly subverted the exhaustion of tumor-infiltrating NK cells, similar to the effect of blockade of TIGIT with antibodies, but also sustained tumor-specific T-cell function in a CD4⁺ T cell–dependent manner. An enhanced effect was achieved by the combined therapy of TIGIT-Fc and PD-L1 blockade. TIGIT-Fc also potentially mediated the antibody-dependent cell-mediated cytotoxicity (ADCC) lysis of tumor cells in vitro. Hence, these findings demonstrate that TIGIT-Fc may coordinate both the reversal of NK-cell exhaustion and the acceleration of the activation of T cells for tumor control and support the clinical development of TIGIT-Fc for cancer immunotherapy.

The human cell lines A375, A431, SK-BR-3, SK-OV-3, MCF-7, Daudi, and H2126 were purchased from the ATCC. The identities of the cell lines were verified by short tandem repeat analysis. CT26 cells were purchased from the Chinese Academy of Sciences (Shanghai, P.R. China). MC38 cells purchased from Kerafast, Inc. All cell lines were confirmed to be Mycoplasma free. The cells were maintained in DMEM with 10% FBS. A375, A431, SK-BR-3, SK-OV-3, and MCF-7 cells were obtained and cultured since February 2016. H2126 cells were obtained and cultured since April 2016. CT26 and MC38 cells were obtained and cultured since March 2018. H2126-CD155 cells were generated by retroviral infection containing CD155 cDNA (pBABEpuro). Stable colonies were selected in the presence of puromycin (100 ng/mL; Sigma-Aldrich) for pBABEpuro vector. After selection and propagation, stable transfectants were cultured in DMEM containing 10% FBS. Cell lines were typically passaged two to three times between thawing and experimental use.

Human peripheral blood mononuclear cells (PBMC) were isolated from 40 healthy volunteer donors by layering over Ficoll-Paque Plus (GE Healthcare) as per the manufacturer' instructions and were enriched for CD4⁺ or CD8⁺ T cells using RosetteSep (STEMCELL Technologies) T-cell enrichment kit as per the manufacturer's instructions. All procedures involving human subjects were conducted in accordance with the Declaration of Helsinki. All specimens were collected under an approved protocol by the Second Military Medical University Review Board, and written informed consent was obtained from each donor.

TIGIT fusion proteins were prepared as described previously (10). Briefly, a recombinant plasmid was constructed by fusing the Fc segment of human IgG1 or murine IG2a, encoding the hinge-CH2-CH3 segment, to the C-termini of the extracellular domains of human and murine TIGIT, respectively (Supplementary Table S1). All fusion proteins were obtained via the FreeStyle 293 Expression System (Invitrogen) and subsequently purified using protein A-sepharose from the harvested cell culture supernatant. The purity of the fusion proteins was determined by polyacrylamide gel electrophoresis. The protein concentration was measured according to the UV absorbance at a wavelength of 280 nm.

Mice were housed in a specific pathogen–free barrier facility. All the in vivo experiments were approved by the Institutional Animal Care and Use Committee of Second Military Medical University (Shanghai, P.R. Chima). To generate tumor-reactive human T cells for mouse models, A375 cells were suspended at 1 × 10⁶/mL in DMEM + 1% FCS and mitotically inactivated with mitomycin C (Sigma-Aldrich) added to 100 μg/mL. Cells were treated at 37°C for 1.5 hours and then thoroughly washed before use. T cells were then cultured with mitomycin C–treated A375 cells (at 1:1 ratio) for 10 days in RPMI1640 supplemented with 10% FBS and IL2 (100 IU/mL, ImmunoTools).

Tumor-reactive T cells were mixed with A375 cells at a 1:5 ratio, and the cell mixtures were implanted by subcutaneous injection of 5 × 10⁶ cells into the flank of NSG mice (Shanghai Model Organisms Center, Inc., Shanghai, P.R. China). In some experiments, tumor-reactive CD8⁺ T cells were further isolated using CD8⁺ T-cell isolation kit (Miltenyi Biotec). Control IgG (10 mg/kg, Selleck Chemicals), atezolizumab (10 mg/kg, Selleck Chemicals), etigilimab (10 mg/kg, Mereo BioPharma), tiragolumab (10 mg/kg, Roche), TIGIT-Fc (10 mg/kg), CD155-Fc (10 mg/kg, BioLegend), atezolizumab plus TIGIT-Fc (5 mg/kg each), atezolizumab plus etigilimab (5 mg/kg each), and atezolizumab plus tiragolumab (5 mg/kg each) were administered intraperitoneally 1 hour immediately after implantation. Some groups of mice received intraperitoneally either 100 μg anti-CD4 (GK1.5, BioLegend) or 100 μg anti-CD8β (53.5.8, InVivoMAb) to deplete T-cell subsets or 100 μg anti-asGM1 (eBioscience) to deplete NK cells every week from day 0 of tumor growth. For syngeneic tumor models, MC38 (5 × 10⁶) or CT26 cells (5 × 10⁶) were subcutaneously inoculated into C57BL/6 mice or BALB/c mice (Shanghai Experimental Animal Centre of Chinese Academy of Sciences). For tumor xenograft models, CT26, A431, SK-BR-3, SK-OV-3, H2126, MCF-7, or Daudi cells (5 × 10⁶ for all cell lines) were subcutaneously inoculated into BALB/c nude mice (Shanghai Experimental Animal Center of Chinese Academy of Sciences) or NSG mice. For MCF-7 models, female BALB/c nude mice were implanted with 0.72 mg 60-day release 17β-estradiol pellets (Innovative Research of America) in the mammary fat pads. When tumor volumes reached an average of approximately 50 mm³, the mice were randomly divided into groups of 8 mice each. Multiple dose studies consisted of 3 weeks of intraperitoneal treatment with murine TIGIT-Fc (mTIGIT-Fc; 10 mg/kg) or control IgG (10 mg/kg), with a 2× loading dose on the first dose (day of randomization). For tumor rechallenge model, all the animals survived for 80 days and experienced a complete response among the MC38 tumor-bearing mice or CT26 tumor-bearing mice that were initially treated with mTIGIT-Fc, were subsequently given a second implantation of MC38 cells or CT26 cells, respectively (5 × 10⁶). For all in vivo experiments, the tumors were measured every day with digital calipers, and tumor volumes were calculated by the following formula: volume = length × (width)²/2. Mice were euthanized, and tumors were harvested for further analysis. The experiments were ended if the volume of the any tumor reached 1,300 mm².

Tumor tissues were minced into approximately 1 to 3 mm³ fragments and then digested using media containing collagenase I (0.1% w/v, Sigma) and DNAse (0.005% w/v, Sigma), and a GentleMACS Dissociator (Miltenyi Biotec, 130-093-235) for 1 hour at 37°C, cells were then strained with 70 and 40 μm cell strainers prior for further analysis, and tumor-infiltrating lymphocytes (TIL) were isolated on a discontinuous Percoll gradient (GE Healthcare). Isolated cells were washed with complete RPMI media twice and then used in various cell function assays.

Cell surface staining was performed at 4°C for 30 minutes and was analyzed using a FACSCalibur flow cytometer (BD Biosciences) and CellQuest Software (BD Biosciences). A minimum of 1 × 10⁴ cells were examined. FcR Blocking Reagent (human cell and mouse cell, Miltenyi Biotec) was used according to the manufacturer's instructions. Antibodies to mouse CD3ϵ (145-2C11, BD Biosciences), CD107a (1D4B, BD Biosciences), IFNγ (XMG1.2, BioLegend), TNF (MP6-XT22, BioLegend), CD155 (4.24.1, BioLegend), CD8α (53-6.7, BioLegend) and antibodies to human CD3 (HIT3a, BioLegend), CD45 (2D1, BioLegend), CD4 (RPA-T4, BD Biosciences), PD-1 (EH12, BD Biosciences), TIGIT (MBSA43, eBioscience), and CD8 (RPA-T8, BD Biosciences) were used in this study. For TIL experiments, the cells were stimulated with phorbol 12-myristate 13-acetate (30 ng/mL; Sigma) and 1 μmol/L ionomycin (Sigma) in the presence of monensin (2.5 μg/mL; eBioscience) for 4 hours, after which they were stained for surface markers and then fixed and permeabilized with eBioscience FoxP3 fixation buffer according to the manufacturer's instructions. Fixed cells were stained with antibodies to IFNγ (XMG1.2; BioLegend) and TNF (MP6-XT22, BioLegend), and Rat IgG1 antibody was used as an isotype control (RTK2071, BioLegend).

Radioactive ⁵¹Cr (50 μCi, PerkinElmer) was used to label 1 × 10⁶ target cells (A431, SK-BR-3, SK-OV-3, H2126, and H2126_CD155 cells)for 1 hour at 37°C. Labeled target cells (n = 5,000) were plated in each well of a 96-well plate (Corning) in a volume of one hundred microliters. Assays were performed for 4 hours in RPMI medium supplemented with FBS. Effector PBMCs (Lonza) were added at a volume of 100 μL at different effector-to-target (E:T) ratios (from 1:50 to 20:1) with TIGIT-Fc or control IgG (40 μg/mL). The cells were incubated together for 4 hours at 37°C. Supernatant (30 μL) from each well was collected and transferred to the filter of a Luma Plate (PerkinElmer Life Sciences), and the filter was allowed to dry overnight. Radioactivity released into the culture medium was measured using a β-emission-reading liquid scintillation counter (PerkinElmer Life Sciences). The percentage of specific lysis was calculated as follows: (sample counts − spontaneous counts)/(maximum counts − spontaneous counts) × 100.

For proliferation assays, carboxyfluorescein diacetate succinimidyl ester (CFSE; Thermo Fisher Scientific)-labeled PBMCs were stimulated for 48 hours with plate-bound anti-CD3/CD28 (InVivoMAb, 1 μg/mL) and either plate-bound control IgG or TIGIT-Fc (20 mg/mL). CFSE dilution was measured by flow cytometry. For the detection of IFNγ, TNF, and IL2, the supernatant of cells was harvested and measured according to the manufacturer's instructions (BioLegend).

For suppression assays, PBMCs were cultured in DMEM with 10% FBS. CD4⁺CD25⁺ Tregs and CD4⁺CD25⁻ responder cells were isolated using the Regulatory T-cell Isolation Kit (Miltenyi Biotec). CD4⁺CD25⁻ responder cells (2 × 10⁴/well) and CD4⁺CD25⁺ Tregs were cultured in triplicate in the presence of soluble anti-CD3/CD28 (1 μg/mL). After 54 hours, cells were pulsed with 1 μCi [³H]thymidine (PerkinElmers) for an additional 18 hours and harvested. [³H] thymidine incorporation was analyzed to assess proliferation. Percentage of suppression = 100 − C.P.M. of well with the indicated ratio of effector: Tregs/mean C.P.M. of wells with effectors alone.

For cell viability assays, SK-BR-3 and SK-OV-3 cells were plated in triplicate at 5 × 10³ per well in 96-well plates overnight in DMEM with 10% FBS. After plating, the cell culture medium was replaced with assay medium containing 0.1% FBS. TIGIT-Fc and control IgG were added at multiple concentrations (from 0.001 to 100 μg/mL), and 72 hours later, cell viability was measured using the CellTiter-Glo Luminescent Cell Viability Assay (Promega).

For migration assays, cell mobility was assessed with a modified two-chamber migration assay (8-mm pore size, BD Biosciences) according to the manufacturer's instructions and as described previously (13) Briefly, approximately 2 × 10⁴ SK-BR-3 or SK-OV-3 cells were plated into the upper chamber for 18 to 24 hours along with different treatments (TIGIT-Fc or control IgG, 1 mg/mL) and allowed to migrate into the lower chamber (DMEM with 20% FBS at 37 °C in 5% CO₂). The cells at the bottom of the membrane were fixed and stained with 20% methanol, 0.2% crystal violet, whereas the cells in the upper chamber were removed using cotton swabs.

Invasion assays were performed as previously described using transwell cell culture chambers (8-μmol/L pore size polycarbonate membrane, Costar; ref. 14). Briefly, 100 μL of cell suspension (SK-BR-3 and SK-OV-3 cells) at 1 × 10⁶ cells/mL with different treatments (TIGIT-Fc or control IgG, 1 mg/mL) in DMEM supplemented with 0.5% FBS was loaded into the upper chamber, and the lower chamber was loaded with 600 μL of DMEM with 10% FBS. The membranes were precoated with Matrigel (BD Pharmingen). The chamber plates were incubated at 37°C for 24 hours, and then the filter was fixed in 4% paraformaldehyde and stained with hematoxylin (Sigma). The cells on the upper side of the filter were removed with a cotton swab, and the cells that passed through the membrane were counted in 10 randomly selected microscopic fields, and all cells stained were counted.

For colony formation assays, approximately 500 SK-BR-3 and SK-OV-3 cells were seeded onto 6-well plates or 3.5-cm dishes along with different treatments (TIGIT-Fc or control IgG, 1 mg/mL). Colonies were allowed to form in an incubator at 37°C and 5% CO₂ for 10 days. At the end of the incubation period, the clones were fixed and stained with 0.5% crystal violet, and colonies larger than 50 μmol/L in diameter were counted.

Data were analyzed using IBM SPSS for Macintosh, Version 22.0. Unless otherwise specified, the Student t test was used to evaluate the significance of differences between two groups, and ANOVA was used to evaluate differences among three or more groups. Differences between samples were considered significant when P < 0.05.

Targeting coinhibitory receptors is highly relevant in cancer where therapeutic effects are being exploited clinically. PD-1/PD-L1 blockers are efficacious in the treatment of cancer, but resistance to this therapy is increasing, and the responses to anti–PD-1 or anti–CTLA-4 immunotherapy are relatively minimal in some tumors (such as colorectal carcinoma). Reports show that antibodies targeting TIGIT have synergistic effects with PD-1/PD-L1 blockade in cancer, both in preclinical (15) and clinical studies (3). Because we and others previously have shown that TIGIT-Fc is an immune-suppressive agent (9, 10), we initially aimed to test whether TIGIT-Fc had the capacity to promote resistance to anti-PD-L1 therapy in xenograft models of the human melanoma cell line A375. Both TIGIT-Fc and atezolizumab were confirmed to bind to A375 cells (Fig. 1A). To generate allogenic T-cell lines with specificity to A375 cells (16), primary human T cells were expanded in culture and implanted subcutaneously in NSG mice together with tumor cells. Allogenic T cells expressed both PD-1 and TIGIT before implantation (Supplementary Fig. S1). Our data showed that the administration of the anti-PD-L1 antibody atezolizumab led to a growth inhibition of the A375 tumors in the presence of human T cells, consistent with a previous report (16). TIGIT-Fc alone significantly inhibited the tumor growth of A375 xenografts compared with an isotype-matched control antibody. The combined therapy of TIGIT-Fc and atezolizumab therapy showed a superior tumor growth inhibition reaching nearly 100% (Fig. 1B). On the basis of the Ig-like structure of TIGIT-Fc, we hypothesized that TIGIT-Fc may have a two-sided biological effect, in that it is a potent agonist of CD155 and a membrane-bound TIGIT antagonist. We tested the antitumor effect of the TIGIT antibodies etigilimab or tiragolumab in a T cell–based A375 tumor model. Etigilimab and tiragolumab alone showed no inhibitory effect on the growth of the A375 tumor xenograft, and combined therapy with etigilimab/atezolizumab or tiragolumab/atezolizumab showed no synergistic effect compared with atezolizumab monotherapy (Fig. 1C and D), suggesting that the antitumor efficacy of TIGIT-Fc was not due to TIGIT blockade in A375 xenografts.

A previous report shows that blockade of TIGIT prevents NK-cell exhaustion (17). To assess whether TIGIT-Fc could block TIGIT on NK cells similar to that of anti-TIGIT in tumor-bearing mice, mTIGIT-Fc was first tested against subcutaneous MC38 colon adenocarcinoma tumors because this model is considered a standard and has been demonstrated by many laboratories as anti–PD-L1–sensitive and immunogenic (18). In the MC38 tumor model (Fig. 2A), we found that tumor growth was inhibited by the administration of mTIGIT-Fc (Fig. 2A) and improved overall survival of mice (Fig. 2B). Three of 8 mice treated with mTIGIT-Fc showed complete tumor regression, suggesting that TIGIT-Fc had the capacity to induce antitumor immunity. Administration of mTIGIT-Fc resulted in a higher frequency of CD107a⁺, TNF⁺, IFNγ⁺, granzyme B⁺, and perforin⁺ tumor-infiltrating NK cells than mice treated with control IgG (Fig. 2C; Supplementary Fig. S2A), indicating that mTIGIT-Fc reversed the exhaustion of the tumor-infiltrating NK cells. A significantly higher frequency of tumor-infiltrating CD8⁺ T cells with surface expression of CD107a, TNF, IFNγ, as well as granzyme B and perforin, was also observed with mTIGIT-Fc treatment (Fig. 2D). Another colorectal cancer mouse model, CT26, was also employed, in which it was observed that mTIGIT-Fc delayed tumor growth and prolonged overall mouse survival (Fig. 2E and F), and 2 of 8 mice showed complete tumor regression with mTIGIT-Fc treatment. Consistent with the observation in MC38 tumors, mTIGIT-Fc alleviated NK-cell exhaustion, as a higher frequency of tumor-infiltrating NK cells expressing CD107a, TNF, IFNγ, granzyme B, and perforin was observed (Fig. 2G; Supplementary Fig. S2B). mTIGIT-Fc–treated tumors also showed a higher frequency of CD8⁺ T cells expressing TNF and IFNγ (Fig. 2H). In both MC38 and CT26 models, mTIGIT-Fc significantly decreased the proportion of intratumoral Tregs and increased the CD8⁺ T-cell:Treg ratio (Supplementary Fig. S2C and S2D). Thus, these results indicated that TIGIT-Fc showed a similar TIGIT blockade effect as anti-TIGIT to prevent the exhaustion of NK cells in tumor-bearing mice, with a beneficial effect on T cells.

We then determined whether the antitumor effect of TIGIT-Fc could occur in nude mice, which lack T cells, in the CT26 tumor model. We found that mTIGIT-Fc treatment inhibited tumor growth (Fig. 3A) and that tumor-infiltrating NK cells showed improved function, as indicated by the increased frequency of cells expressing CD107a, TNF, or IFNγ (Fig. 3B). The therapeutic value of the recombinant human TIGIT-Fc–targeting regimen was further assessed in mice xenografted with human cancer cells. Vehicle-treated A431 tumors progressed rapidly and reached volumes of greater than 1,000 mm³ in less than 30 days (Fig. 3C). Conversely, TIGIT-Fc treatment effectively delayed tumor growth to this volume for approximately 40 days. TIGIT-Fc treatment also efficiently inhibited tumor growth and delayed tumor growth for 45 and 30 days in SK-BR-3 and SK-OV-3 tumors, respectively (Fig. 3C).

Because we detected high CD155 expression in the mouse tumor cell line CT26 and in the human tumor cell lines A431, SK-BR-3, and SK-OV-3 (Fig. 3D), we therefore used H2126 cells with very low detectable CD155 expression to further evaluate the antitumor role of TIGIT-Fc. We observed that the therapeutic benefit of TIGIT-Fc treatment on tumor growth in the H2126 model was minimal (Fig. 3C). Because H2126 cells also expressed a low level of TIGIT ligands CD112 and Nectin-4 (Supplementary Fig. S3), we used MCF-7 cells to further assess the role of TIGIT ligand in antitumor effect. MCF-7 cells expressed a low level of CD155 but high level of CD112 and nectin-4 and did respond to TIGIT-Fc treatment, whereas in mice xenografted with Daudi cells, TIGIT-Fc showed no detectable antitumor effect. Further supporting that tumor ligand expression contributes to the antitumor effect of TIGIT-Fc, we found that in a tumor model using H2126 cells engineered to express CD155 TIGIT-Fc treatment had an antitumor effect (Fig. 3C). In all the tumor models using nude mice, despite delayed tumor growth and a survival advantage, mice treated with TIGIT-Fc eventually succumbed to disease, and no mice rejected tumor cells, suggesting that the antitumor effect of TIGIT-Fc could be weakened by T-cell dysfunction.

To further explore the mechanism that underlies the antitumor effect of TIGIT-Fc, mice were depleted of NK cells, CD4⁺ T cells, and CD8⁺ T cells in combination with mTIGIT-Fc treatment in syngeneic immune-competent mouse models. In the CT26 tumor model, we found that the depletion of CD4⁺ T cells, CD8⁺ T cells, or NK cells significantly inhibited the therapeutic effect of mTIGIT-Fc (Fig. 4A and B). A previous report shows that depletion of NK cells abolishes the therapeutic effect of anti-TIGIT (17). Here, 2 of 8 mice rejected tumors after treatment with TIGIT-Fc combined with anti-asialoGM1 (anti-asGM1), suggesting that TIGIT-Fc has NK cell–independent antitumor mechanisms.

The tumor growth effect of TIGIT-Fc, which was significantly weakened with CD4⁺ T-cell depletion, indicated a role of CD4⁺ T cells in TIGIT-Fc treatment. To further assess the impact of T cells on the therapeutic effect of TIGIT-Fc, CD8⁺ T cells only or CD4⁺ and CD8⁺ T cells were provided by coimplanting human cells into NSG mice bearing A375 tumors. Reduced tumor growth was only observed in mice treated with TIGIT-Fc together with both CD4⁺ and CD8⁺ T cells, whereas the impact of the TIGIT-Fc therapy with CD8⁺ T cells alone was minimal (Fig. 4C). Additional treatment with recombinant CD155 abolished the antitumor effect of TIGIT-Fc, suggesting that the effect of TIGIT-Fc was dependent on CD155. Functional assessment of TIGIT-Fc in vitro in anti-CD3/anti-CD28-stimulated PBMCs from healthy donors demonstrated the fusion protein enhances CD4⁺ and CD8⁺ T-cell proliferation (Fig. 4D). TIGIT-Fc also significantly increased Th1 cytokine production (IFNγ, TNFα, and IL2) in these cultures compared with control IgG (Fig. 4E). Although TIGIT-Fc did not affect the CD25^hiFoxp3⁺ Treg frequencies (Supplementary Fig. S4A), treatment of Tregs with TIGIT-Fc showed a decreased ability to suppress T-cell receptor (TCR)–stimulated proliferation of effector T cells in vitro (Supplementary Fig. S4B).

Our data thus far showed that TIGIT-Fc has the capacity to evoke antitumor immunity involving both T cells and NK cells. We then tested whether the mice that survived with TIGIT-Fc treatment generated memory responses using a tumor rechallenge model. MC38 tumor-bearing mice that were initially treated with mTIGIT-Fc and experienced a complete response were subsequently given a second implantation of MC38 cells in the absence of any further treatment (Fig. 5A). Mice that previously cleared tumors with treatment were protected, whereas naïve mice showed tumor engrafting. Similar results were obtained in CT26 tumor-bearing mice (Fig. 5B) with mTIGIT-Fc treatment, where a rechallenge with tumor cells showed no sign of growth. These results demonstrated that TIGIT-Fc monotherapy could elicit antitumor immune memory.

The combination of TIGIT-Fc and atezolizumab showed an antitumor effect in the human T cell/A375 model. We therefore investigated whether the combination of mTIGIT-Fc and anti–PD-L1 in mice with established syngeneic tumors. In both MC38 and CT26 colorectal adenocarcinoma (Fig. 5C and D), treatment with mTIGIT-Fc or anti–PD-L1 monotherapy resulted in reduced tumor growth, with 2 of 8 and 3 of 8 mice tumor free after 35 days (25%–37.5%), respectively. Following combinatorial TIGIT-Fc/anti–PD-L1 immunotherapy, 75% and 62.5% of the MC38- and CT26-inoculated mice, respectively, were tumor free after 35 days. Thus, these data showed that the combined therapy of TIGIT-Fc and anti–PD-L1 induces potent antitumor effects, leading to improved tumor control.

TIGIT-Fc itself may have potent ADCC effect because it is an Fc-fused protein, and therefore we evaluated the ability of TIGIT-Fc to induce ADCC lysis of human tumor cell targets expressing CD155 in vitro utilizing PBMC-dependent ADCC assays. Our data show that the CD155 high–expressing cell lines A431, SK-BR-3, and SK-OV-3 were sensitive to ADCC-mediated lysis by TIGIT-Fc (Fig. 6A) over a range of E:T ratios, whereas the CD155-negative cell line H2126 was resistant to lysis. We next investigated whether TIGIT-Fc had an intrinsic effect on tumor cells by performing in vitro cell viability, migration, and invasion assays using CD155 high–expressing tumor cells. However, no effect on proliferation of both SK-BR-3 and SK-OV-3 cells was observed after TIGIT-Fc treatment compared with control IgG (Fig. 6B). In migration, invasion, and colony formation assays, TIGIT-Fc treatment also showed a negligible effect on tumor cells (Fig. 6C and D). To confirm these findings, the effects of TIGIT-Fc on the in vivo tumor growth was further examined in NSG mice (to exclude the impact of immune cells). Our data showed TIGIT-Fc treatment had no detectable effect of tumor growth of both A431 cells and SK-BR-3 cells (Fig. 6E).

TIGIT-Fc was used to simulate membrane-bound TIGIT function in an earlier study (9). Upon binding with TIGIT-Fc, the activation of CD155 in human monocyte-derived DCs (MDDC) leads to decreased secretion of the proinflammatory cytokine IL12 and increased secretion of IL10. It should be noted that in addition to TIGIT-Fc, CD226-Fc also has been shown to modify DC cytokine production, suggesting that this effect is mainly caused by CD155 signaling in MDDCs (9). We previously showed that murine TIGIT-Fc shows therapeutic effect in a mouse model of lupus (10), in which TIGIT-Fc delayed the onset of proteinuria, decreased the serum concentrations of autoantibodies, and increased survival rate. Although the mechanism of such observations is not clear, we speculate the main immune suppression effect of TIGIT-Fc is based on the antigen-presenting cells (APC), as the percentage of T lymphocytes was not altered by the fusion protein among TIGIT-Fc–treated mice. Interestingly, subsequent studies report that agonistic anti-TIGIT inhibit anti-CD3/anti-CD28–mediated T-cell proliferation and cytokine production in the absence of APCs in humans and mice (19–21). A study has further found that the IFNγ production of CD8⁺ T cells is suppressed by melanoma cells expressing a truncated version of CD155 in a similar manner as cells expressing wild-type CD155 (22). Taken together, these studies suggest that the TIGIT-CD155 interaction can inhibit T-cell functions without downstream signaling via CD155, highlighting the cell-intrinsic mechanisms of TIGIT.

The observation that inhibitory signals can also be directly transmitted via the cytoplasmic tail of TIGIT further support its cell-intrinsic role. Two publications established that the ITIM motif is essential for human TIGIT signaling, whereas mouse TIGIT inhibition can be mediated by either the ITIM motif or the ITT motif alone (5, 23). Another group suggests an important role for the intracellular ITT-like motif in human TIGIT and highlight two different signaling pathways interfering with NK-cell cytotoxicity or IFNγ production (24, 25). Whole-genome microarray analysis shows that mouse T-cell activation is suppressed by TIGIT engagement by downregulating TCR expression, together with several other molecules involved in TCR and CD28 signaling (21).

Structurally, TIGIT-Fc is an immune adhesin, which is an Ig-like chimeric protein comprised of target-binding regions fused to the Fc-hinge region of Ig and is designed to have a long half-life and antibody-like properties. Mechanistically, TIGIT-Fc has the capacity to block TIGIT-associated interactions, including the binding of CD155–TIGIT. In a xenograft model system to investigate the T cell–mediated tumor cell killing, administration of TIGIT-Fc in A375 xenografts resulted in tumor growth inhibition. The antitumor effect of TIGIT-Fc was entirely dependent upon the presence of the tumor-reactive human T cells in this model. These data demonstrated the ability of TIGIT-Fc to increase tumor cell elimination by T cells and support the further investigation of TIGIT-Fc for the treatment of cancer. TIGIT-Fc treatment enhanced NK cell–mediated antitumor immunity both in syngeneic tumor models and tumor xenograft models in nude mice, which is very similar to the effect of anti-TIGIT (17), demonstrating that blockade of TIGIT signaling in tumor-infiltrating NK cells by TIGIT-Fc can be achieved. Unlike anti-TIGIT, which promotes tumor-specific T-cell immunity in an NK cell–dependent manner, TIGIT-Fc exhibited a direct activation effect on tumor-specific T cells. We further found that CD4⁺ T cells were critical for the therapeutic effects of TIGIT-Fc and that TIGIT-Fc treatment augmented proliferation of T cells and mediated Th1 development from naïve CD4⁺ T cells in vitro. These data are consistent with a previous study that shows that agonist anti-CD155 induce Th1 development in both humans and mice, as evidenced by production of IFNγ and upregulation of Tbx21 transcription (26). We observed here that T cells derived from different healthy donors show different responses to TIGIT-Fc, indicating a heterogeneity in CD155 signaling-based T-cell function. Our results also showed that TIGIT-Fc treatment decreased the proportion of intratumoral Tregs and increased the CD8⁺ T-cell:Treg ratio in mouse models, and TIGIT-Fc directly inhibited the ability of Tregs to suppress TCR-stimulated proliferation of effector T cells in vitro. These data are consistent with a previously report that TIGIT⁺ Tregs inhibit the generation of effector T-cell responses and suggests that regulation of Tregs may underline the antitumor effect of TIGIT-Fc.

CD155 is a member of the poliovirus receptor–related (PRR) family of adhesion molecules, along with CD111 (nectin-1/PRR-1), CD112 (nectin-2/PRR-2), nectin-3 (PRR-3), and nectin-4 (PRR-4). CD155 was initially identified as a receptor for poliovirus in humans (27) and binds to another PRR family member, nectin-3, as well as to the matrix protein vitronectin, thereby mediating cell–cell or cell–matrix adhesion, respectively, and cell migration (28). CD155 is a ligand for CD226, CD96, and TIGIT on T cells and NK cells. The expression of CD155 in normal tissue, such as epithelial and endothelial cells, is very low, but most tumor cells express high levels of CD155 (28). Dysregulation of CD155 expression promotes tumor cell invasion, migration, and proliferation and is associated with a poor prognosis and enhanced tumor progression (28–30), but the cell-intrinsic and cell-extrinsic effects of CD155 has not been fully explored. In the current study, we showed that targeting CD155 with TIGIT-Fc did not have an effect on the development of tumor cells in vitro, indicating that at least CD155 did not directly participate in oncogenic signaling. TIGIT-Fc showed a potent ADCC effect, which is well established as a major mode of tumor-cell killing with immune cells, paving the way for TIGIT-Fc as a possible therapeutic candidate for CD155-overexpressing tumors.

Our study has several limitations. We provide evidence that promoting tumor-specific immunity with TIGIT-Fc can be achieved, but our in vivo efficacy models may not fully recapitulate human tumors, and the data are from a small number of animals. In our current study, the TIGIT-Fc has complement binding and fixation, as well as Fcγ-dependent, antibody-dependent, and cell-mediated cytotoxicity; therefore, future studies should use the non–FcR-binding versions of TIGIT-Fc (e.g., LALA-PG Fc variants) to better understand the relative contributions of TIGIT itself versus FcR engagement. The mechanisms responsible for these therapeutic effects of TIGIT-Fc are currently not well characterized. Because some other cells in the tumor microenvironment also express CD155, for example, tumor-infiltrating myeloid cells, the mechanisms of TIGIT-Fc may be more complicated. Hence, these findings will need further validation. To sum up, our study reveals that TIGIT-Fc can induce the antitumor immune response of NK cells and T cells. On the basis of the wide application of T cell– and NK cell–based treatment strategies in clinical practice, targeting agents that intervene in the CD155–TIGIT axis hold translational potential. Our results support the conclusion that treatment with TIGIT-Fc alone or in combination with other immune checkpoint inhibitors is a promising strategy to improve the outcomes of tumor immunotherapy.

X. Shen reports grants from National Natural Science Foundation of China during the conduct of the study. W. Fu reports grants from National Natural Science Foundation of China, Shanghai Sailing Program, and Open Project Grant from Engineering Research Center of Cell & Therapeutic Antibody, Ministry of Education, of Shanghai Jiao Tong University during the conduct of the study, as well as a patent for 202011000490.7 pending. Y. Wei reports grants from National Natural Science Foundation of China during the conduct of the study. J. Zhu reports grants from National Natural Science Foundation of China during the conduct of the study. Y. Yu reports grants from National Natural Science Foundation of China during the conduct of the study. C. Lei reports grants from National Natural Science Foundation of China during the conduct of the study. J. Zhao reports grants from National Natural Science Foundation of China during the conduct of the study and is a shareholder of KOCHKOR Biotech, Inc., Shanghai. S. Hu reports grants from National Natural Science Foundation of China, Shanghai Rising-Star Program, Shanghai Chenguang Program, and Shanghai Biomedical Technology Support Project during the conduct of the study, as well as a patent for 202011000490.7 pending.

X. Shen: Conceptualization, formal analysis, investigation, methodology, writing–review and editing. W. Fu: Conceptualization, resources, data curation, software, formal analysis, funding acquisition, validation, investigation, visualization, methodology, writing–original draft, project administration, writing–review and editing. Y. Wei: Conceptualization, formal analysis, investigation, methodology, writing–review and editing. J. Zhu: Conceptualization, investigation. Y. Yu: Conceptualization, investigation. C. Lei: Conceptualization, formal analysis, methodology, writing–original draft, project administration, writing–review and editing. J. Zhao: Conceptualization, investigation, project administration, writing–review and editing. S. Hu: Conceptualization, resources, data curation, software, formal analysis, supervision, funding acquisition, validation, investigation, visualization, methodology, writing–original draft, project administration, writing–review and editing.

This study was supported by the National Natural Science Foundation of China (grant numbers 81773261, 31970882, 81903140, 81670573, 81972252, and 81602690); the Shanghai Rising-Star Program (grant number 19QA1411400); the Shanghai Sailing Program (19YF1438600); the Shanghai Chenguang Program (grant number 17CG35); and the Shanghai Biomedical Technology Support Project (20S11906600) and the Open Project Grant from Engineering Research Center of Cell & Therapeutic Antibody, Ministry of Education, of Shanghai Jiao Tong University.

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