Cytokine-like Molecule CCDC134 Contributes to CD8⁺ T-cell Effector Functions in Cancer Immunotherapy

The eradication of tumors by the immune system depends on the generation of antigen-specific T cells that can migrate to tumor sites and perform effector functions. CD8⁺ T cells, also called CTLs, play a critical role in the antitumor immune response (1, 2). Upon encountering antigens, CD8⁺ T cells undergo a program of activation, clonal expansion, and differentiation into effector cells (3, 4). These CTLs are capable of killing tumor cells via the granule exocytosis pathway, which involves the release of effector molecules, such as granzyme B and perforin, and the secretion of proinflammatory cytokines, such as IFNγ and TNFα (5). It is believed that the perforin/granzyme-mediated pathway of cytotoxicity is the primary mechanism for tumor cell killing by CD8⁺ T cells (6), although many recent studies have shown that IFNγ appears to be the most critical molecule for CD8⁺ T-cell–mediated tumor immunity (7–12). Furthermore, tumor cells themselves are major targets of IFNγ via the antiproliferative and antimetabolic effects of this cytokine as well as the ability of this cytokine to inhibit angiogenesis (13–15).

For processes ranging from activation by antigen to differentiation into effector cells, CD8⁺ T cells require support from cytokines produced by CD8⁺ T cells themselves or other immune cells such as CD4⁺ T cells, B cells, or dendritic cells. Many studies have demonstrated that certain cytokines exert overlapping and distinct regulation over CD8⁺ T-cell responses to tumor antigens. The cytokines belonging to the γ_c cytokine receptor family are particularly important for these CD8⁺ T-cell responses. This cytokine family consists of IL2, IL4, IL7, IL9, IL15, and IL21 and is so named because the receptors for these cytokines share the γ_c receptor chain. The cytokines of the γ_c family all signal through the JAK–STAT pathway. Interestingly, all of these cytokines activate JAK1 and JAK3, and it is believed that JAK3 selectively associates with γ_c but not with other cytokine receptors (16–19). Furthermore, IL2, IL7, IL9, and IL15 primarily activate STAT5, whereas IL4 primarily activates STAT6 and IL21 primarily activates STAT3. Among these cytokines, IL2, the first cytokine to be extensively tested in the clinic for immunotherapy, is a potent growth factor for T lymphocytes and has mainly displayed antitumor effects in the generation and maintenance of CTL responses to tumors. IL15 is the primary cytokine known to stimulate the maintenance of memory CD8⁺ T cells, and IL7 can act in synergy with IL15 to support the survival and expansion of memory CD8⁺ T cells. Many studies have demonstrated that some γ_c cytokines, including IL2, IL7, IL15, and IL21, have the potential to expand and prolong CTL responses at different stages of the immune response by promoting the activation, proliferation, and effector function of CTLs (18, 20). Therefore, these cytokines have potential roles in adoptive transfer therapy using CTLs (21, 22).

However, the molecular and cellular nature of the tumor immune microenvironment influences the antitumor function of CD8⁺ T cells and the disease outcome by altering the balance of suppressive and cytotoxic responses in the vicinity of the tumor. To inhibit the antitumor immune response, local tumor cells often produce inhibitory factors, such as TGFβ, IL10, and galectins (23, 24), or induce a population of suppressive cells, such as regulatory T cells, tumor-associated macrophages, or tolerogenic dendritic cells, to inhibit the activation and cytotoxic responses of CD8⁺ T cells (25, 26). Therefore, the naturally occurring T-cell responses against malignancies are often not sufficient to cause the regression of tumors. However, in recent years, there has been increasing evidence indicating that adoptive cell transfer of cells cultured in the presence of a proinflammatory cytokines, particularly γ_c family-cultured tumor-infiltrating CD8⁺ T cells isolated from tumor tissues, has the potential to expand antitumor CD8⁺ T cells and enhance CTL killing activity in vitro and in vivo (27, 28).

CCDC134 was first identified in our lab. CCDC134 is conserved across different species, including human, rat, mouse, and dog, but shares no obvious homology with any known proteins. We previously reported that CCDC134 plays an important role in ELK1 transcription regulation and MAPK signal transduction through the Raf-1/MEK/ERK and JNK/SAPK pathways (29).

In the present study, we report that CCDC134 demonstrates functions similar to those of the γ_c cytokine IL2 by directly promoting activation, proliferation, and antigen-specific cytotoxicity of CD8⁺ T cells via the JAK3–STAT5 signaling pathway in vitro. Moreover, CCDC134 displayed a strong antitumor effect in a CD8⁺ T-cell–dependent manner in vivo, which was correlated with increased lymphocyte infiltration into the tumor and enhanced cytolytic activity of CD8⁺ T cells. These results clearly indicate that CCDC134 is a novel stimulator of CD8⁺ T cells, with the potential to be applied therapeutically during cancer immunotherapy.

Six-week-old female C57BL/6 and SCID-beige mice from Vital River Laboratories were used for in vivo experiments. TCR transgenic OT-I mice (H-2^b) recognize SIINFEKL from chicken ovalbumin (OVA_257–264) and IFNγ^−/− mice (B6.129S7-Ifng^tm1Ts/J) were kindly provided by Dr. Zhang MH (Tsinghua University, Beijing, China) and Dr. Yin YH (Peking University, Beijing, China), respectively.

For generation of CCDC134 transgenic mice, the pcDB-CCDC134 construct was linearized to release the human CCDC134 gene with a CMV promoter, and then the CMV-CCDC134 was injected into B6 mouse zygotes using standard techniques. The zygotes were implanted into pseudopregnant recipient mice and genotyping founder mice were gotten. After potential founder mice were screened by PCR, the human CCDC134 expression was further identified by in situ hybridization and ELISA. After repetitious passage and identification, two inbred strains of CCDC134 transgenic mice were obtained.

The cDNA for human or mouse CCDC134 was subcloned into the pMH expression vector with C-terminal 6xHis tag. Proteins were produced via transient expression in HEK293E cells. For experiments requiring pure proteins, purification was performed by affinity chromatography using with a Ni-Sepharose 6 Fast Flow column (GE Healthcare). Peak fractions of >95% pure CCDC134 were quantified by high-performance liquid chromatography (HPLC), SDS-PAGE, and Coomassie blue staining with BSA as a standard. Endotoxin levels (0.023 eu/mg protein) were determined (BioWhittaker Limulus Amebocyte Lysate QCL-1000 pyrogen testing).

For intracellular cytokine staining, CD8⁺ T cells were maintained in media with brefeldin A (10 μg/mL; Sigma-Aldrich) for 5 hours. Cells were first stained with surface markers, fixed with 4% formaldehyde for 20 minutes, and permeabilized with 0.1% Triton X-100 (eBioscience). Intracellular cytokine staining was performed using specific antibodies. Flow cytometry analysis was performed using FACSCalibur and analyzed with FlowJo software (BD Biosciences).

Peripheral blood lymphocytes (PBL) or isolated CD8⁺ T cells at a concentration of 0.5–50 × 10⁶/mL were labeled with 2.5 μmol/L carboxyfluorescein succinimidyl ester (CFSE; Invitrogen) in 5% FBS/PBS at 37°C for 5 minutes. Cells were washed twice, and CFSE-labeled cells were incubated in RPMI-1640 with 10% FBS and cell division was detected.

OT-I CD8⁺ T cells were stimulated with anti-CD3/anti-CD28 mAbs in vitro for 4 days as effector cells. The EG7 target cells (1 × 10⁴ cells/200 μL/well) and serial dilutions of effector cells were incubated in a 96-well U-bottomed plate at 37°C for 8 hours. Reactions were performed in triplicate at the indicated effector:target ratios. The target cells or effector cells alone were used as spontaneous lactate dehydrogenase (LDH) release control. For the maximum target LDH release, lysis solution was added to the target cells. The supernatants were assayed for LDH release using a CytoTox 96 Non-Radioactive Cytotoxicity Assay Kit (Promega). The percentage of cell-mediated cytotoxicity was determined by the following equation: cytotoxicity (%) = (experimental−effector spontaneous−target spontaneous)/(target maximum − target spontaneous) × 100.

Mice were injected s.c. with 2 × 10⁵ B16 cells in the axilla, and rhCCDC134 or PBS as control were injected intraperitoneally (i.p.) once each day when tumor were already established (mean diameter is about 3 mm or average volume of tumors to 16 mm³, 6–10 days). Tumor volumes were measured along two orthogonal axes (a = length, b = width) and calculated as tumor volume = ab²/2. Tumors were excised and weighed after sacrificing animals.

The individual tumors were homogenized through 200 mesh strainers to obtain single-cell suspensions. Red blood cells were lysed using an ammonium chloride lysis buffer. Tumor-infiltrating lymphocytes (TIL) were isolated from tumor suspensions using Percoll (GE Healthcare). The collected cells were stained with different antibodies and flow cytometry analysis was performed as described previously. The total number of different immune cells per 10⁷ of tumor cells was calculated by multiplying the percentage of positive cell population.

In vivo depletion of CD8⁺ and CD4⁺ T cell was performed using blocking monoclonal antibodies (mAb). Each mouse was injected i.p. with 0.5 mg rat anti-CD8 (clone 2.43), anti-CD4 (clone GK1.5) mAbs, or normal rat IgG (Beijing Zsbio. Co.), as control antibody in 200 μL PBS on day 1 before rhCCDC134 administration (twice per week). And, the depletion of CD4⁺ and CD8⁺ T cells was consistently greater than 95% as determined by flow cytometry.

Statistical significance of the tumor-free survival plots was determined using the log rank test. A Student t test and one-way ANOVA was applied to compare statistical significance between treatment groups. All analysis was performed using Prism 4 software (GraphPad Software).

We first evaluated whether CCDC134 is expressed in human primary immune cells using RT-PCR. We found that CCDC134 transcripts were present in T cells and dendritic cells (Fig. 1A). Moreover, CCDC134 expression and secretion was induced following the activation of T cells with anti-CD3/anti-CD28 mAbs or phytohemagglutinin (Fig. 1B and C). To further compare the expression levels of CCDC134 mRNA during the activation of CD4⁺ and CD8⁺ T cells, we sorted CD4⁺ and CD8⁺ T cells from peripheral blood mononuclear cells (PBMC) of healthy donors and then stimulated these cells with anti-CD3/anti-CD28 mAbs. Increased expression of CCDC134 was detected both in activated CD8⁺ and CD4⁺ T cells. Notably, the expression of CCDC134 was more than 5-fold higher in CD8⁺ T cells than CD4⁺ T cells. These results indicate that CCDC134 expression may be related to the activation of T cells, especially CD8⁺ T cells (Fig. 1D).

To address the effects of CCDC134 on T cells, we obtained recombinant human CCDC134 protein (rhCCDC134) from engineered HEK293E cells. We isolated PBLs from human healthy donors and added rhCCDC134 to these cells stimulated with anti-CD3/anti-CD28 mAbs, and monitored the proliferation of T cells. The results revealed that CCDC134 promoted the proliferation of CD8⁺ and CD4⁺ T cells in response to anti-CD3/anti-CD28 mAb stimulation (Fig. 1E).

To further investigate whether CCDC134 directly promotes the proliferation of CD4⁺ or CD8⁺ T cells, we isolated CD8⁺ and CD4⁺ T cells from human PBL and stimulated them with anti-CD3/anti-CD28 mAbs in the presence or absence of rhCCDC134 (1, 10, or 100 ng/mL). We found that CCDC134 significantly induced the upregulation of CD25 and CD69, two early and transient T-cell activation markers (30–31), in activated CD8⁺ T cells (Fig. 1F), but not activated CD4⁺ T cells (data not shown). Furthermore, the expression of effector molecules, including IFNγ, perforin, and TNFα, was enhanced in these activated CD8⁺ T cells based on FACS and ELISA analysis. Maximal stimulation was observed with 10 ng/mL of rhCCDC134 (Fig. 1F and G).

In addition, the CCDC134 cDNA sequences from both human and mouse encode 229-amino acid-long polypeptides, and the homology between their CCDC134 amino acid sequences without a signal peptide is as high as 90%, suggesting that mouse CCDC134 may have the same function as human CCDC134 (Supplementary Fig. S1). To address whether mouse CCDC134 mediates the specific functions in CD8⁺ T cells, we purified recombinant mouse CCDC134 protein (rmCCDC134). As expected, we found that rmCCDC134 augmented the expression of CD69, CD25, and IFNγ in mouse CD8⁺ but not CD4⁺ T cells, following costimulation with anti-CD3/anti-CD28 mAbs (data not shown). Furthermore, rhCCDC134 demonstrated similar effects on mouse splenic CD8⁺ T cells. The rhCCDC134 stimulation resulted in increase of CD25 and CD69, as well as TNFα, granzyme B, IFNγ, and perforin expression in CD8⁺ but not CD4⁺ T cells of mouse spleen (Fig. 2 and Supplementary Fig. S2A and S2B). Maximal effects were also observed with 10 ng/mL of rhCCDC134. These results suggest that CCDC134 directly promotes the activation of CD8⁺ but not CD4⁺ T cells costimulated with anti-CD3/anti-CD28 mAbs in both humans and mice.

To investigate whether CCDC134 directly promote the expansion of CD8⁺ T cells in a manner independent of other immune cells, purified CD8⁺ or CD4⁺ T cells from mouse spleen were labeled with CFSE and then stimulated with anti-CD3/anti-CD28 mAbs plus rhCCDC134. Like the IL2 stimulation, the CCDC134 significantly promoted the proliferation of CD8⁺ T cells in a dose-dependent manner (Fig. 3A), but not CD4⁺ T cells (Supplementary Fig. S2C).

In addition, we isolated CD8⁺ T cells from OT-1 mice and stimulated these cells with anti-CD3/anti-CD28 mAbs plus rhCCDC134. Our results revealed that CCDC134 significantly promoted the killing activity of CTLs against EG7 targets and increased the cytotoxicity to levels well above the limit of detection at E:T ratios of 5:1 or higher in a dose-dependent and effector–target ratio-dependent manner (Fig. 3B). Moreover, CCDC134 significantly enhanced the levels of TNFα, granzyme B, and IFNγ (Fig. 3C). These results suggested that CCDC134 directly promotes the proliferation and the cytotoxicity of CD8⁺ T cells.

To further understand the activity of CCDC134 in vivo, we established three tumor graft models using the B16 mouse melanoma. Initially, CCDC134-transgenic mice were established, in which human CCDC134 expression was identified by in situ hybridization and ELISA assay (Supplementary Fig. S3A and B). Then B16 tumor cells were subcutaneously (s.c.) inoculated into CCDC134 transgenic mice, and the tumor cell growth and survival rates of tumor-bearing mice were evaluated. Significant retardation of established B16 tumor growth was detected and the survival time was markedly prolonged in CCDC134 transgenic mice (Fig. 4A and B).

Next, we established B16 graft tumors in C57BL/6 mice. Treatment with rhCCDC134 resulted in a significant suppression of tumor growth in a dose-dependent manner (Fig. 4C). In addition, CCDC134 also demonstrated strong antitumor activity against EG7 thymoma and Lewis lung adenocarcinoma (LLC-1) cells (Supplementary Fig. S4A–S4C).To further examine whether the antitumor activity of CCDC134 was dependent on lymphocytes, we established B16 graft tumors in C.B.17 SCID-beige mice. We found that CCDC134 was unable to suppress tumor growth in SCID-beige mice (Fig. 4D), suggesting that the antitumor activity of CCDC134 was dependent on lymphocytes.

We first analyzed the absolute number of immune cells, particularly CD8⁺ T cells, at tumor sites after CCDC134 treatment. As expected, CCDC134 administration significantly enhanced migration of CD8⁺ T cells into tumors. Interestingly, greater numbers of CD4⁺ T and B cells, but not NK cells, were also recruited into the tumors after treatment with CCDC134 (Fig. 5A). Enhanced expression of the IFNγ, granzyme B, perforin, and TNFα was also observed in CD8⁺ T cells from CCDC134-treated mice (Fig. 5B and C). Similarly, we also found that CCDC134 could promote the expression of granzyme B and perforin, but not IFNγ, in CD4⁺ T cells (Supplementary Fig. S5A and S5B).

Then, we treated C57BL/6 mice with depleting anti-CD8 or anti-CD4 mAbs and monitored CCDC134-induced growth suppression to B16 tumors. Our results significantly revealed that anti-CD8 mAbs completely blocked the antitumor activity of CCDC134. However, much less efficient residual tumor growth was observed in mice treated by anti-CD4 mAbs (Fig. 5D and E). Surprisingly, after treatment with the anti-CD8 mAb, not only the CCDC134-mediated recruitment of CD8⁺ T cells, but also the CD4⁺ T cells and B cells, to the tumors was also diminished (Fig. 5F). However, after CD4⁺ T cells were depleted, there was no decrease in the CCDC134-recruited CD8⁺ T cells or B cells to the tumor sites, suggesting that the infiltration of CCDC134-recruited CD8⁺, but not the CD4⁺ T cells, into the tumor is essential for the antitumor activity of CCDC134.

IFNγ is the most critical molecule for CD8⁺ T-cell–mediated tumor immunity (8, 9, 15). Therefore, we investigate the role of IFNγ in the CCDC134-induced antitumor effect. Surprisingly, we found that the therapeutic benefit of CCDC134 was absent in IFNγ^−/− mice (Fig. 6A). However, we found that IFNγ was not responsible for the CCDC134-induced recruitment of CD8⁺ T, CD4⁺ T, or B cells to the tumor (Fig. 6B), although depletion of IFNγ abrogated the CCDC134-induced augmentation of lytic activity in CD8⁺ T cells via blocking the production of granzyme B, perforin, and TNFα (Fig. 6C and D). In addition, the absence of IFNγ also blocked the CCDC134-induced production of granzyme B and perforin in CD4⁺ T cells (Supplementary Fig. S5C and S5D). These results suggest that IFNγ is essential for CCDC134-induced CD8⁺ T-cell–mediated tumor immunity.

As described above, we found that CCDC134 has similar activity to CD8⁺ T cells with IL2, so we surmise whether CCDC134 shares a coreceptor with IL2. We first confirmed the fact that CCDC134 can significantly bind to activated CD8⁺ T cells by reacting ¹²⁵I-CCDC134 (Fig. 7A), and the binding could be competitively inhibited by increasing concentrations of unlabeled CCDC134, IL2, or IL15 (both IL2 and IL15 can bind IL2R), but not the IL12 as negative control (Fig. 7B).

To investigate whether CCDC134 activate the intracellular signals of IL2R, we determined the phosphorylation of JAK3, STAT3, and STAT5 in CCDC134-activated CD8⁺ T cells. We observed that like IL2, CCDC134 significantly promoted STAT5 phosphorylation. In contrast, unlike IL6 (a positive control for p-STAT3), there was no significant change in the STAT3 phosphorylation after CCDC134 treatment. Furthermore, CCDC134 selectively induced an increase in the phosphorylation of JAK3 (Fig. 7C).

Moreover, JAK3 inhibitor (JAK3-I) completely blocked the CCDC134-mediated increase of p-STAT5 but not p-STAT3 (Fig. 7D). However, we found that JAK3-I blocked the IL2-mediated increase in both p-STAT5 and p-STAT3, as it is believed that IL2-induced activation of CD8⁺ T cells is dependent on either the JAK3–STAT5 or JAK3–STAT3 pathway (32).

Next, we assessed the role of JAK3–STAT5 signaling in mediating the proliferative effects of CCDC134 on CD8⁺ T cells. As expected, the inhibitors of JAK3 (JAK3-I) or STAT5, but not the inhibitor of STAT3, completely blocked CCDC134-induced proliferation of CD8⁺ T cells (Fig. 7E), suggesting that activation of JAK3–STAT5 is required to stimulate the proliferation of CD8⁺ T cells mediated by CCDC134.

In the present study, we identified a previously unknown γ_c-related protein, CCDC134, which significantly augments the activation, proliferation, and cytotoxicity of CD8⁺ T cells in vitro and exerts the ability to suppress tumor growth in a CD8⁺ T-cell–dependent manner in vivo.

So far, the physiologic bioactivity of CCDC134 as a classical secreted protein remains unclear. We previously identified that CCDC134 could inhibit MAPK-mediated ELK1 transcriptional activity in HeLa cells (29). CCDC134 also serve as a regulator of hADA2a (human alteration/deficiency in activation 2a) in acetyltransferase activity, DNA damage-induced apoptosis, and cell-cycle arrest (33). Recently, it was found that CCDC134 inhibit the migration and invasion of tumor cells via the MAPK pathway (34).

As CCDC134 is expressed by many lineages of immune cells and can particularly be upregulated and secreted by activated T cells, we hypothesized that the function of CCDC134 might involve T cells. Therefore, we selectively addressed whether CCDC134 was involved in the function of CD4⁺ or CD8⁺ T cells. Initially, our results clearly demonstrated that CCDC134 was capable of promoting TCR signaling-induced proliferation of CD4⁺ and CD8⁺ T cells in a PBMC culture system. Notably, the effect of CCDC134 on promoting the proliferation of CD8⁺ T cells was stronger than the effect on CD4⁺ T cells. Subsequently, we isolated CD4⁺ or CD8⁺ T-cell populations and investigated the direct bioactivity of CCDC134 on them. These results revealed that CCDC134 significantly promoted activation and proliferation of CD8⁺ T cells but not CD4⁺ T cells. More importantly, we showed that CCDC134 directly promoted the expansion and antigen-specific cytotoxicity of CD8⁺ T cells by augmenting the expression of IFNγ, TNFα, granzyme B, and perforin in vitro.

We have explored the antitumor role of CCDC134 in vivo. The results revealed that CCDC134 administration led to strong inhibition of tumor growth and prolonged survival in CCDC134 transgenic mice and C57BL/6 mice, but not in SCID mice. These results suggested that the antitumor activity of CCDC134 is immune system-dependent. To explore the possible antitumor mechanism of CCDC134, we evaluated the number and activation state of tumor-infiltrating CD8⁺ T, CD4⁺ T, B, and NK cells. Our results showed that CCDC134 not only significantly promoted the infiltration of activated CD8⁺ T cells, but also promoted infiltration of CD4⁺ T and B, but not NK cells, into the tumor. However, only the depletion of CD8⁺ T cell, but not the CD4⁺ T cell, could completely blocked the antitumor activity of CCDC134. These data suggest that the antitumor activity of CCDC134 is dependent on CD8⁺ but not CD4⁺ T cells. Surprisingly, the intratumoral accumulation of CD4⁺ T and B cells following CCDC134 treatment was also reduced with anti-CD8 mAb treatment, suggesting that recruitment of CD4⁺ T and B cells induced by CCDC134 may be dependent on CD8⁺ T cells. Richard and colleagues previously demonstrated that CD8⁺ T cells are capable of recruiting other immune cells into a tumor and proposed the theory that in addition to the contact-mediated lytic mechanism, the control of tumor growth by CD8⁺ T lymphocytes is partially mediated by the recruitment of other host effector cells to the tumor site (35–36).

IFNγ is a cytokine known to play an essential role in coordinating tumor immune responses by regulating the migration of T cells into tumor tissue (37). The IFNγ produced by tumor-infiltrating T cells plays two distinct roles in antitumor activity: the activation of antitumor T cells and the generation of direct tumoricidal activity via the production of inducible nitric oxide synthetase (15, 38). In this study, we found that CCDC134 significantly promoted the expression of IFNγ in tumor-infiltrating CD8⁺ but not CD4⁺ T cells. Furthermore, upregulation of granzyme B, perforin, and TNFα in tumor-infiltrating CD8⁺ or CD4⁺ T cells by CCDC134 was abolished, and CCDC134 antitumor activity was also blocked in IFNγ^−/− mice. On the basis of these findings, it is likely that the CCDC134-driven antitumor activity of CD8⁺ T cells is mainly mediated by IFNγ.

In addition, we used the IL2, a member of the γ_c family reported to function as a CD8⁺ T-cell growth factor (39), as positive control, our results clearly revealed that many similar activities were showed between CCDC134 and IL2, such as they directly promoted the activation and proliferation of CD8⁺ T cells, and enhanced the cytotoxicity of tumor antigen-specific CD8⁺ T cells in vitro and in vivo. Moreover, it has been shown that both CCDC134 and IL2 could activate CD8⁺ T cells through the JAK3–STAT5 signaling pathway, the downstream signaling pathway of IL2R γ_c (40). Importantly, two members of the γ_c cytokine family, IL2 and IL15, could significantly block the CCDC134 binding to activated CD8⁺ T cells. On the basis of these data, CCDC134 may signal through the γ_c chain as a potential member of the γ_c cytokine family. However, the receptor for CCDC134 remains to be confirmed.

In summary, our results suggest that CCDC134 represents a potential member of the γ_c cytokine family, as it demonstrated critical effects on the activation, expansion, and effector functions of CD8⁺ T lymphocytes. Moreover, CCDC134 may serve as an attractive candidate to promote antitumor activity in cancer immunotherapy.

No potential conflicts of interest were disclosed.

Conception and design: J. Huang, L. Xiao, D. Ma, X. Qiu

Development of methodology: L. Xiao, W. Shao

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): J. Huang, L. Xiao, X. Gong, Y. Yin

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): J. Huang, L. Xiao, X. Gong, X. Qiu

Writing, review, and/or revision of the manuscript: J. Huang, L. Xiao, X. Gong, X. Qiu

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): J. Huang, L. Xiao, Q. Liao, Y. Meng, Y. Zhang

Study supervision: J. Huang, X. Qiu

This work was supported by the National Science and Technology Major Projects of New Drugs (2012ZX09103301-032) and the National Natural Sciences Foundation of China (81372254 and 30973389).

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.