Multiple negative regulators restrict the ability of T cells to attack tumors. This work demonstrates the role of PI3K-interacting protein 1 (Pik3ip1) in restraining T-cell responses and antitumor immunity.
An anti-Pik3ip1 mAb was generated to identify the Pik3ip1 expression pattern of hematopoietic cells. Pik3ip1−/− mice and a Pik3ip1 fusion protein were generated to investigate the effect of Pik3ip1 on T-cell–mediated antitumor immunity in MC38 and B16-F10 tumor models. Immunoblotting and confocal microscopy were used to identify inhibitory effects of Pik3ip1 on T-cell receptor (TCR) signaling. Pik3ip1 expression was quantified, and its impact on T-cell function in human tumors was measured.
We demonstrated that Pik3ip1 was predominantly expressed on T cells and served as an essential rheostat for T-cell–mediated immunity. A Pik3ip1 genetic deficiency led to enhanced T-cell responsiveness upon immunization with a neoantigen. Pik3ip1−/− mice exhibited a marked increase in antitumor immunity and were resistant to tumor growth. Furthermore, Pik3ip1 extracellular domain fusion protein enhanced MC38 tumor growth was observed. Mechanistically, we found that Pik3ip1 inhibited TCR signaling by mediating the degradation of SLP76 through Pik3ip1 oligomerization via its extracellular region. Consistent with the results from the mouse models, PIK3IP1 expression correlated with T-cell dysfunction in human tumors.
Our data reveal a critical role for Pik3ip1 as a novel inhibitory immune regulator of T-cell responses and provide a potential molecular target for cancer immunotherapy.
The identification of novel negative checkpoint regulator pathways should have important therapeutic implications. Here we identified a candidate gene named PI3K-interacting protein 1 (Pik3ip1) as a novel check point regulator in tumor immunology. Pik3ip1 genetic deficiency leads to enhanced T-cell responsiveness. Pik3ip1−/− mice exhibit a marked increase in antitumor immunity and are resistant to tumor growth. Consistent with results from mouse model, Pik3ip1 expression correlates with T cells dysfunction in tumor patients. This study provides compelling evidence to uncover a critical role for Pik3ip1 as a novel immune inhibitory regulator of T-cell responses against antigen-specific immune challenges and tumor, thus provide a potential molecular target for cancer immunotherapy.
In recent years, the field of cancer immunotherapy has made substantial progress in improving cancer treatment. It has been well recognized that multiple negative checkpoint regulators (NCRs) could restrict the ability of T cells to effectively attack tumors. Releasing these brakes through antibody blockade, first with an anti-CTLA-4 antibody and then with anti-PD-1 and anti-PD-L1 antibodies, has emerged as an exciting strategy for cancer therapy (1–3). Numerous studies have also shown the critical functions in maintaining peripheral tolerance and controlling antitumor immune responses by several other inhibitory molecules such as T-cell immunoglobulin domain and mucin domain 3 (Tim-3; ref. 4), lymphocyte activation gene 3 (LAG-3; ref. 5), and B- and T-lymphocyte attenuator (BTLA; ref. 6). Thus, the identification of additional NCR pathways could have important therapeutic implications.
PI3K/Akt/mTOR pathway is one of the key signaling pathways involved in T-cell activation and function (7). During T-cell activation, Src homology 2 domain-containing leukocyte protein of 76 kDa (SLP76) associates with PI3K subunit p85, leading to the downstream Akt activation (8). Inhibition of PI3K isoform promoted robust CD8+ T-cell activation, improved antitumor immunity, and host survival by enhancing the cytokines production of cytotoxic lymphocytes (9, 10). PI3K/Akt/mTOR complex has also been shown to be involved in the differentiation of various immunosuppressive T-cell subsets including Foxp3+ Treg cell and tumor-associated macrophages (11, 12). Thus, specific PI3K/Akt/mTOR-targeted therapies could regulate cellular metabolism and immune cell function in cancer and other diseases (13).
PI3K-interacting protein 1 (Pik3ip1), a cell-surface protein with an intracellular motif homologous to the PI3K regulatory subunit p85, was reported as a negative regulator of the PI3K pathway (14). It is reported that the overexpression of Pik3ip1 in mouse hepatocytes leads to a reduction in PI3K signaling and the suppression of hepatocyte carcinoma development (15). The LP Kane group first reported that Pik3ip1 plays an inhibitory role in T-cell activation. Ectopic expression of Pik3ip1 in Jurkat or D10 T cell lines inhibited the activity of an NFAT/AP-1 transcriptional reporter (16, 17). In addition, both the kringle and p85-like domain were shown to be important for the inhibitory function of Pik3ip1 in activated T cells. Using an inducible knockout (KO) mouse model, the authors showed that Pik3ip1−/− T cells exhibit more robust activation and faster clearance of Listeria monocytogenes infection than T cells from wild-type (WT) mice through the inhibition of the PI3K/Akt pathway (16, 17). However, whether Pik3ip1 contributes to antitumor T-cell immunity is largely unknown.
In this study, we found that Pik3ip1−/− mice were resistant to MC38 and B16-F10 tumor growth compared with littermates. The ectodomain Pik3ip1 fusion protein promoted MC38 tumor growth in vivo. Mechanistic studies further suggested that in addition to inhibiting the PI3K/Akt pathway, Pik3ip1 could also form oligomers upon TCR stimulation to degrade SLP76, thus globally inhibiting the downstream pathways of TCR signaling, including the ERK1/2 and p38 MAPK pathways. Furthermore, PIK3IP1 expression in tumor infiltrating T cells was correlated with impaired functionality in patients with hepatocellular carcinoma (HCC) and oral squamous cell carcinoma (OSCC). Taken together, our results identified Pik3ip1 as a novel negative immune regulator that might be a potential target for future cancer immunotherapy.
Materials and Methods
C57BL/6, Rag-1−/− and OT-1 mice were purchased from the Model Animal Research Center of Nanjing University (Nanjing, China). Pik3ip1−/− mice was created in C57BL/6 mice by microinjection of TALENs in fertilized eggs from Cyagen Biosciences Inc. The mouse Pik3ip1 gene (GenBank accession number: NM_178149.4; Ensembl: ENSMUSG00000034614) is located on mouse chromosome 11. Exon 4 was selected as target site. TALEN mRNAs (listed in Table S1 in Supplementary Material and Method) generated by in vitro transcription were then injected into fertilized eggs for KO mouse productions. The founders were genotyped by PCR followed DNA sequencing analysis. The positive founders were breeding to the next generation. Pik3ip1 homozygous (+/+) C57BL/6 mice (littermates, LT) generated from Pik3ip1 heterozygotes were bred and maintained under conditions identical to those of the Pik3ip1−/− mice and were used as controls for experiments. All animal experiments were approved by the Institutional Animal Care and Use Committee of the Sun Yat-sen university and performed following local rules.
The MC38 murine colon adenocarcinoma cell line and B16-F10 melanoma cell line used in the tumor model were purchased from the ATCC. Cell were passaged a maximum of 4 times, tested for mycoplasma contamination, and identified by STR before the experiments. MC38 tumor cells (1,000,000) or B16-F10 tumor cells (1,000,000) were inoculated on the right flank of Pik3ip1−/− mice and littermates. For protein treatment MC38 tumor model, 200 μg Pik3ip1 fusion protein or mouse IgG protein were injected peritoneally starting from day 0 and every 3 days throughout the whole experiment. To generate chimerism, Rag-1−/− mice were lethally irradiated with a total dose of 10 Gy radiation followed by bone marrow (BM) reconstitution with 2 × 107 cells from either Pik3ip1−/− mice and littermates via tail vein approximately 8 hours later. Recipient mice were bled at 7 weeks to analyze the T-cell constitution. At 8 weeks, MC38 (1,000,000) was inoculated on the right flank of the recipient mice. All tumors were measured with calipers every 2 days. The tumor volume was estimated by the formula: (L × W2)/2. The mice were euthanized right before the tumors were dissected.
Patients and specimens
Fresh paired samples of blood, tumor tissue, adjacent normal tissues were obtained from patients with oral squamous carcinoma (OSCC) and hepatic cellular cancer (HCC) who underwent curative resection at the Stomatological Hospital and Cancer Center of Sun Yat-sen University. None of the patients had received anticancer therapy before sampling. All samples were anonymously coded with local ethical guidelines. The study was approved by the institutional review board of Stomatological Hospital and Cancer Center of Sun Yat-sen University and was conducted in agreement with the Helsinki Declaration. Written informed consent was provided by all participants at baseline and during follow-up.
Production of the Pik3ip1–Ig fusion protein
Pik3ip1–mIg was prepared by fusing the coding region of extracellular domain of Pik3ip1 to the Fc constant region of mouse IgG2a. The construct was transfected into 293T cells by the polyethyleneimines (Invitrogen) transfection method and cultured in serum-free DMEM. The supernatant was collected at 5 days and the fusion protein was purified by HiTrap Protein A HP (GE Healthcare) and dialyzed in LPS-free PBS. The purity and expected molecular size of fusion protein were confirmed by electrophoresis on polyacrylamide gels and by MS analysis.
Generation of anti-Pik3ip1 mAbs
Wistar rats were immunized with purified mouse Pik3ip1–mIg mixed with complete Freud's adjuvant (Sigma Aldrich) and boosted three times with mouse Pik3ip1-mIgG2a in incomplete Freud's adjuvant. Sera from the mice were collected, and their specific binding to Pik3ip1 was determined by ELISA and by FACS analysis on Pik3ip1/CHO cells. The splenocytes from mice with highest titer of antisera were fused with YB2/0 myeloma cells to produce hybridoma cells. After several rounds of selection by ELISA and FACS, 5 clones, 17C9, 19G3, 22A11, 23G3, 25H3, which consistently stain Pik3ip1/CHO cells, were selected. The isotype of these clones is rat IgG2a. The culture supernatant of hybridoma was concentrated and purified by a protein G-Sepharose column (Pierce) and dialyzed in LPS-free PBS.
OVA257–264: SIINFEKL and OVA323-339: ISQAVHAAHAEINEA peptides were synthesized by GenScript. Peptides (100 μg) were dissolved in PBS, mixed with the TLR3 agonist polyI:C (100 μg i.p.; Sigma-Aldrich). Splenocytes were harvested on day 7 postimmunization and restimulated with the relevant peptides in complete IMDM. IFNγ and TNFα producing CD4+ or CD8+ T cells were analyzed by FACS. IFNγ and TNFα production was analyzed by ELISA (Invitrogen).
Isolation of mononuclear cells from blood and tissues
PBMCs were isolated by Ficoll density gradient centrifugation as manufacturer's instruction (TBD). Tumor-infiltrating leukocytes (TILs) were obtained as manufacturer's instruction (Miltenyi). Briefly, fresh tumor tissue were digested in 200 μg/mL collagenase IV and 50 μg/mL DNase (Sigma-Aldrich) solution and disassociated in C tube using GentleMACS dissociator (Miltenyi). Cell suspension were performed by a density gradient centrifugation step to remove erythrocytes and dead cells.
Plasmids for HA- or FLAG or GFP-tagged SLP76, LAT, mPik3ip1 and their fragment were cloned into the pcDNA3.1 vector for transient expression, and transfected using Lipofectamine 2000 according to the manufacturer's instructions (Invitrogen).
In vitro and in vivo T-cell proliferation
For in vitro studies, purified CD3+ T cells (100,000 cells per well) were cultured in 96-well flat-bottom plates coated with 1.25 μg/mL anti-CD3 and 1.25 μg/mL anti-CD28 antibody. Cultures were analyzed on day 3 for Carboxyfluorescein succinimidyl ester (CFSE; Invitrogen) profiles or according to a time course as indicated. For in vivo studies, 3 × 106 splenocytes were separated from OT-1 transgenic mice, labeled with 10 μmol/L CFSE, were injected to female C57BL/6 mice and Pik3ip1−/− mice via tail vein injection (day 0). On day 1, host mice were immunized with OVA257–264 peptides (100 μg) emulsified in the right flank. On day 2, draining lymph node (dLN) cells were harvested and the percentages of donor T cells were determined by congenic marker (Vα2 and Vβ5.1/5.2). Donor CD8+ T-cell proliferation were analyzed by flow cytometry.
Staining cell surface or intracellular targets was described previously (18). Briefly, single-cell preparations were stained for dead cells using a Zombie NIR Fixable Viability Kit (Biolegend). For intracellular staining, cells were stimulated with 50 ng/mL PMA (Sigma-Aldrich) and 5 μg/mL Ionomycin (Sigma-Aldrich) in the presence of GolgiPlug (BD Bioscience). After 4 hours, cells were stained for dead cells and surface markers then fixed prior to intracellular staining for cytokines. Data were analyzed using a FACSVerse flow cytometer (BD Biosciences) and using FlowJo software (TreeStar). All antibodies for FACS were listed in Table S1 in Supplementary Materials and Methods.
For confocal analysis, purified mouse CD3+ T cell were stained with rabbit anti-p-SLP76, SLP76, p-LAT, LAT plus rat antimouse Pik3ip1, followed by Alexa Fluor 594–conjugated anti-rat IgG or Alexa Fluor 488–conjugated anti-rabbit IgG. For oligomerization analysis, purified CD3+ T cell from KO or LT were stained with rat antimouse monoclonal Pik3ip1 antibody, followed by Alexa Fluor 594–conjugated anti-rat IgG. Then the T cells were incubated with either Pik3ip1 ecto-domain fusion protein or control protein tagged with biotin, followed by Alexa Fluor 488–conjugated streptavidin. Positive cells were quantified using Zen black software or detected by LSM 780 confocal microscopy (Zeiss). All antibodies for immunofluorescence and IHC staining were listed in Table S1 and Supplementary Materials and Methods.
Immunoprecipitation and immunoblot analysis.
For immunoblot analysis, cell extracts were prepared after transfection or stimulation with appropriate ligands and resolved by SDS-PAGE. For immunoprecipitation, whole-cell extracts were prepared after transfection, followed by incubation overnight with anti-FLAG beads (Sigma-Aldrich). Beads were washed five times with low-salt lysis buffer, and immunoprecipitates were eluted with 3× SDS Loading Buffer (Cell Signaling Technology) and resolved by SDS-PAGE. Proteins were transferred to polyvinylidene fluoride (PVDF) membranes (Bio-Rad) and further incubated with indicated antibodies. The LumiGlo Chemiluminescent Substrate System (KPL) was used for protein detection. The densitometry was measured by Image J (NIH) software.
All statistical analyses were performed by GraphPad Prism 7.0. All tests were two-sided and a P value of less than 0.05 was considered significant and represented as *, P < 0.05; **, P < 0.01; ***, P < 0.001, and ****, P < 0.0001.
Pik3ip1 was predominantly expressed on T cells with dynamic changes in vitro and in vivo
To study protein expression, a mouse Pik3ip1-specific mAb was produced. Antibody specificity was confirmed by staining CHO cells stably transfected with a plasmid encoding full-length Pik3ip1 and by loss of staining when Pik3ip1–Ig blocked the antibody (Supplementary Fig. S1).
Using this mAb, Pik3ip1 expression on hematopoietic cells was analyzed in four central and peripheral mouse hematopoietic organs (Supplementary Fig. S2). In BM, spleen, lymph node, and peripheral blood cells, relatively higher expression of Pik3ip1 was observed on CD4+ and CD8+ T cells, but markedly lower expression was detected on natural killer (NK) cells, B cells, CD11b+ monocytes, and CD11c+ dendritic cells (DC; Fig. 1A). Moreover, in spleen, the surface expression of Pik3ip1 on naïve and central memory CD4+ and CD8+ T-cell populations was far more obvious than that on Tregs and CD62Llo CD44hi effector memory CD4+ and CD8+ T cells (Fig. 1B).
In addition to assessing freshly isolated cells, we also determined the Pik3ip1 expression on splenic CD4+ and CD8+ T cells, NK cells, B cells, CD11b+ monocytes, and CD11c+ DCs upon in vitro culture with or without activation. Splenocytes were cultured with medium, anti-CD3 (for activating T cells), ConA (for activating B cells), or IFNγ and lipopolysaccharide (LPS; for activating CD11b+ monocytes, CD11c+ DCs, and NK cells). Strikingly, Pik3ip1 expression was quickly lost on all CD4+ and CD8+ T cells, including CD62Lhi CD44lo naïve T cells, CD62Lhi CD44hi central memory T cells, CD62Llo CD44hi effector memory T cells and Tregs, upon in vitro stimulation (Fig. 1B). Pik3ip1 expression did not fluctuate in B cells, NK cells, CD11b+ monocytes, or CD11c+ DCs with stimulation (Fig. 1B).
This phenomenon was further confirmed in vivo. OT-1 TCR transgenic mice were immunized with the cognate antigen OVA257–264 peptide emulsified in complete Freund's adjuvant (CFA). At 24 hours after immunization, cells from the dLN were analyzed for Pik3ip1 expression. Pik3ip1 expression on CD4+ and CD8+ T cells decreased drastically upon immunization with antigen (CFA/OVA) but not immunization with adjuvant alone (Fig. 1C).
Because Pik3ip1 expression was downregulated after stimulation, we wondered how the expression levels of Pik3ip1 were altered under continuous tumor antigen stimulation conditions. We compared the expression levels of Pik3ip1 on CD4+ and CD8+ T cells in samples of leukocytes freshly isolated from the spleen, dLN, nondraining lymph node (nLN), and tumor of a xenograft tumor model. Pik3ip1 was expressed in significantly higher proportions of CD4+ and CD8+ T cells in the spleen, peripheral blood mononuclear cell (PBMC) population and nLN, whereas the mean percentages of Pik3ip1-expressing T cells in the dLN and TIL population were significantly lower. The expression pattern of Pik3ip1 on CD4+ and CD8+ T cells followed the pattern of PBMCs, spleen, and nLN > dLN > TILs (Fig. 1D).
Collectively, these data strongly suggest that Pik3ip1 is differentially expressed on T cells with a dynamic feature that might contribute to the control of antitumor T-cell responses.
Pik3ip1 deficiency enhanced T-cell responses to antigen-specific immune challenges
We first evaluated whether Pik3ip1 deficiency affected T-cell responses to an acute immune challenge with a neoantigen. We compared the cell-mediated immune responses of littermates and Pik3ip1−/− mice to two soluble peptides (100 μg), OVA257–264 (MHC class I restricted) or OVA323–339 (MHC class II restricted), mixed with the toll-like receptor (TLR) agonist polyI:C as an adjuvant. On day 7 after immunization, splenocytes were examined for the production of IFNγ and TNFα upon restimulation with the peptides. Significantly increased levels of IFNγ and TNFα cytokine secretion were observed in cells derived from the Pik3ip1−/− mice in response to OVA257–264 and OVA323–339 peptide immunization (Fig. 2A).
Furthermore, T-cell proliferation in response to peptide immunization was also analyzed in Pik3ip1−/− mice in vivo. CFSE-labeled naïve splenocytes from OT-1 TCR transgenic mice were adoptively transferred into Pik3ip1−/− mice and littermates and subsequently activated with the OVA257–264 peptide. We observed that the percentages of CD8+ OT-1 cells were significantly increased in the Pik3ip1−/− mice compared with the littermates (Fig. 2B).
T-cell proliferation in response to tumor-associated antigens was then measured in littermates and Pik3ip1−/− mice. Interestingly, the proliferation of the Pik3ip1−/− T cells was substantially higher than Pik3ip+/+ T cells in the presence of the anti-CD3/anti-CD28 mAbs after tumor vaccination. This enhanced response was also sustained, as indicated by time-course studies. Cultured supernatants from Pik3ip1−/− T cells contained increased levels of IFNγ and TNFα from 24 to 72 hours after stimulation (Fig. 2C). Therefore, Pik3ip1 expressed on T cells appears to function as an inhibitor of antigen-specific polyclonal TCR stimulation.
Pik3ip1 deficiency promoted antitumor T-cell functions to inhibit tumor growth in vivo
To address whether T-cell dysfunction can be influenced by Pik3ip1 in the setting of antitumor immunity, MC38 tumor cells were injected subcutaneously into littermates or Pik3ip1−/− mice, and their growth was monitored. There was better tumor control in the Pik3ip1−/− mice than in the littermates, as indicated by slower growth in the Pik3ip1−/− mice. Additionally, 40% of the Pik3ip1−/− mice were tumor free, whereas all littermates had tumors (Fig. 3A and B). Tumor-infiltrating CD8+ T cells and their activation evidenced by Ki67 expression were significantly increased in the Pik3ip1−/− mice compared with the littermates (Fig. 3C and D; Supplementary Fig. S3). We also observed significantly greater frequencies of multifunctional CD8+ T cells that produced IFNγ and TNFα or expressed CD107a on the cell surface in the Pik3ip1−/− mice than in the littermates (Fig. 3E). To check the cytotoxic T-cell (CTL) activity of the Pik3ip1−/− T cells, purified CD8+ T cells isolated from tumor-bearing mouse splenocytes were used in a CTL assay against MC38 tumor cells. As shown in Fig. 3F, MC38-specific killing was significantly promoted in the Pik3ip1−/− mice. Although there was a trend towards a decrease in Treg frequency and an increase in myeloid-derived suppressor cell (MDSC) and IL17+ Th17 frequency in the tumors of the Pik3ip1−/− mice, there was significant within-group heterogeneity, and differences were not statistically significant (Fig. 3G). In contrast to CD8+ T cell, we were unable to detect differences in the activation status and effector molecule production of the tumor-infiltrating CD4+ T cells from the Pik3ip1−/− and littermate mice (Fig. 3C–E). Moreover, there were no differences in the CD4+ and CD8+ T-cell populations (Supplementary Fig. S4A), activation phenotype (Supplementary Fig. S4B), or cytokine production (Supplementary Fig. S4C) in the spleen, dLN, or peripheral blood from the Pik3ip1−/− and littermate tumor-bearing mice. Moreover, we observed a similar antitumor T-cell effect in Pik3ip1−/− mice using a B16-F10 tumor model (Supplementary Fig. S5A and S5B). More robust T-cell effector functions were observed in Pik3ip1−/− mice than in littermates (Supplementary Fig. S5C and S5D). Collectively, these data suggest that Pik3ip1 deficiency enhances antitumor responses by promoting T-cell activation and effector functions.
Pik3ip1 deficiency in hematopoietic cells was sufficient to control tumor growth in vivo
Given that Pik3ip1 is a critical regulator of PI3K, a major family of intracellular signal-transducing enzymes involved in a multitude of cellular processes (19), although Pik3ip1 expression was observed on CD45+ host immune cells and not on CD45− cells, including tumor cells and fibroblasts (Supplementary Fig. S6), it is still necessary to carefully identify the role of Pik3ip1 in immune regulation to exclude the possibility that a deficiency in Pik3ip1 expression in endothelial or epithelial cells can impair T-cell trafficking. To evaluate the contributions of Pik3ip1 to hematopoietic and nonhematopoietic cells in MC38 tumor growth, we used BM chimeric models in which lethally irradiated Rag-1−/− mice (recipients) were reconstituted with BM from either Pik3ip1−/− (BM−/−) or littermate (BM+/+) donors. Complete myeloablation is accomplished by giving a lethal dose of radiation followed by the rescue of the immune system with BM from the donor mice to prevent lethality and replacement of the immune system. In this model, Pik3ip1 expression was deficient in the hematopoietic cells of the mice reconstituted with BM−/−, whereas the somatic cells and radiation-resistant cells remained Pik3ip1 sufficient. After the BM transfer, the mice were allowed to reconstitute for 8 weeks. The mice were bled at 7 weeks, and the BM−/− and BM+/+ mice were found to have similar and reasonable levels of CD4+ and CD8+ T cells. At 8 weeks after reconstitution, we injected MC38 tumor cells subcutaneously into the BM−/− and BM+/+ mice and monitored tumor growth (Fig. 4A). We found better tumor control in the mice reconstituted with BM−/− than in those reconstituted with BM+/+, as indicated by the slower growth in the BM−/− mice, which is consistent with our data from the conventional Pik3ip1−/− mouse model (Fig. 4B and C). The percentages of tumor-infiltrating CD4+ and CD8+ T cells and their effector phenotype (CD44hi CD62Llo) were significantly increased in the BM−/− mice compared with the BM+/+ mice (Fig. 4D and E). CD8+ TILs displayed higher baseline surface expression of CD107a and produced more IFNγ and TNFα upon PMA and ionomycin stimulation than BM+/+ cells (Fig. 4F). Together, these results confirm that the observed deceleration of tumor progression by Pik3ip1 deficiency is solely an immune cell-associated effect.
Pik3ip1 inhibited T-cell activation by mediating the degradation of SLP76 through Pik3ip1 oligomerization via its extracellular region
Two studies previously demonstrated that Pik3ip1 inhibits T-cell activation by dampening distal NFAT/AP-1 and Akt signaling (16, 17). From our results, we found that Pik3ip1 could not only inhibit Akt signaling but also inhibit other downstream TCR signaling pathways including ERK1/2 and p38 MAPK (Fig. 5A and B). To determine whether and how Pik3ip1 could impact proximal upstream TCR signaling in primary T cells, we purified splenic T cells from Pik3ip1−/− and littermate tumor-bearing mice, stimulated these cells with anti-CD3/anti-CD28 mAbs and examined the upstream adaptors LAT and SLP76. LAT is a proximal signaling adaptor that is phosphorylated upon TCR stimulation and forms a complex with multiple signaling molecules, including SLP76 and phospholipase C (PLC)-γ1 (20). Compared with that in the Pik3ip1+/+ T cells, the phosphorylation of SLP76 in the Pik3ip1−/− T cells was markedly enhanced, which was maximally apparent at 5 minutes after stimulation (Fig. 5A and B).
And we also noticed that the phosphorylation and the protein amount of total SLP76 but not LAT increased considerably in the Pik3ip1−/− T cells upon stimulation with anti-CD3/anti-CD28 mAbs (Fig. 5A and B). We next observed that Pik3ip1 interacted with both LAT and SLP76 in mouse T cells through confocal microscopy analysis (Fig. 5C). These results indicated that Pik3ip1 might inhibit TCR signaling by mediating the degradation of SLP76. To test this hypothesis, we overexpressed SLP76 or LAT with increasing amounts of Pik3ip1 and found that Pik3ip1 specifically degraded SLP76 but not LAT (Fig. 5D; Supplementary Fig. S7A). We next investigated the molecular mechanisms underlying Pik3ip1-mediated SLP76 degradation. Two major systems that eukaryotic cells use for protein clearance are the ubiquitin–proteasome and autophagy–lysosome pathways (21, 22). To identify which degradation system dominantly regulates the degradation of SLP76, we examined the protein stability of SLP76 using pharmacologic approaches. We observed that the proteasome inhibitor MG132 but not the autophagic-sequestration inhibitor 3-methyladenine (3-MA) or the lysosomal-acidification inhibitor chloroquine (CQ) enhanced the protein level of SLP76 (Fig. 5E; Supplementary Fig. S7B), indicating that the protein stability of SLP76 was controlled by Pik3ip1 through a proteasome-dependent pathway.
Interestingly, we found that the extracellular region of Pik3ip1, not its intracellular region, led to the degradation of SLP76 (Fig. 5F; Supplementary Fig. S7C). Because many transmembrane proteins, such as TLR4, and some membrane proteins undergo oligomerization to achieve their active form (23–25), we wondered whether Pik3ip1 also needs to form oligomers via its extracellular region to degrade SLP76. Indeed, we found that Pik3ip1 extracellular region could form oligomers (Fig. 5G). In addition, we observed that Pik3ip1 extracellular regions could induce the oligomerization of Pik3ip1 (Fig. 5H), suggesting that the interaction between Pik3ip1 extracellular regions plays a critical role in affecting the activation and function of Pik3ip1.
A Pik3ip1 ecto-domain fusion protein altered the cellular composition of the tumor-immune microenvironment and enhanced MC38 tumor growth
Based on our mechanistic studies, we hypothesized that a Pik3ip1 ecto-domain fusion protein (Pik3ip1 extracellular region) could be an antagonist of T-cell activity. Therefore, we assessed the antitumor response after generating a Pik3ip1 fusion protein consisting of the extracellular domain of Pik3ip1 fused with the Fc region of a mouse IgG molecule. MC38 tumor cells were injected subcutaneously into littermates, and tumor growth was monitored. The tumor microenvironment (TME) plays a crucial role in suppressing tumor-specific T-cell responses (26, 27). Pik3ip1 fusion protein treatment significantly promoted tumor growth in the MC38 model (Fig. 6A and B) and altered the TIL composition. Pik3ip1-Ig administration decreased the percentage of tumor-infiltrating CD4+ T cells (Fig. 6C; Supplementary Fig. S8). Within the CD4+ TIL population, decreased frequencies of Foxp3+ Tregs and IL-17+ Th17 cells were found in the Pik3ip1–Ig treatment group (Fig. 6D). The development of activated phenotypes by tumor-infiltrating CD8+ T cells and the production of effector molecules by both CD4+ and CD8+ tumor-infiltrating T cells were strongly suppressed by Pik3ip1–Ig, as evidenced by the decreased numbers of CD44hi CD62Llo and IFNγ+ TNFα+ effector T cells in the Pik3ip1–Ig treatment group (Fig. 6E and F). Our data indicate that the Pik3ip1–Ig fusion protein can alter the cellular composition of the TME and function as an antagonist of antitumor immunity by decreasing T-cell activation and effector molecule production. As T-cell exhaustion is linked to the expression of multiple inhibitory molecules (28–30), we examined PD-1, and LAG-3 expression on T cells in tumor tissue. PD-1 and LAG-3 expression was significantly increased in the Pik3ip1–Ig group compared with the control group (Fig. 6G). Although the percentages of both Th17 cells and Tregs in the CD4+ T-cell population were found to be decreased in the TME in the Pik3ip1–Ig treatment group, more studies are needed to determine the roles of these cells in the antitumor T-cell response.
PIK3IP1 marked a dysfunctional T-cell phenotype in patients with OSCC and HCC
To further define the phenotypic and functional features of the phenotype of T cells with differentially expressed PIK3IP1, PBMCs and paired TILs from patients with OSCC and HCC were stained with a panel of activation markers. We observed that PIK3IP1hi T cells exhibited a PD-1hi HLA-DRhi CD62Lhi CD28hi phenotype (Fig. 7A and B; Supplementary Fig. S9). Furthermore, when stimulating these T cells from the peripheral blood and tumor tissue, the PIK3IP1hi T cells were obviously incapable of producing TNFα and IFNγ (Fig. 7C). Moreover, the PIK3IP1hi T cells from the peripheral blood showed a lower proliferative capacity in both sets of patients (Fig. 7D). Together, these data indicate that PIK3IP1 positivity identifies a less activated T-cell phenotype and is correlated with impaired antitumor T-cell effector functions and proliferation in patients with cancer.
The results presented here are the first study ever to describe the function and expression of Pik3ip1 as a negative regulator of antitumor T-cell response. Pik3ip1 exerts immunosuppressive effects on T cells in vitro and in vivo and could be an important mediator in controlling T-cell responses against acute challenges and cancer. Thus, we propose that Pik3ip1 is a promising new target for cancer immunotherapy.
Pik3ip1 was predominantly expressed on all T-cell subpopulations, especially naïve and central memory CD4+ and CD8+ T cells, indicating that Pik3ip1 appears to track with the maturity of T cells (Fig. 1). The expression pattern of Pik3ip1 distinguishes it from other NCR molecules, such as PD-1, PD-L1, and PD-1H (31–33). The steady-state expression of Pik3ip1 was largely restricted to T cells, whereas that of other NCRs, such as PD-1H, is largely restricted to hematopoietic cells and is highest on both antigen-presenting cells (APCs; macrophages and myeloid DCs) and CD4+ T lymphocytes (33). In contrast to CTLA-4 and PD-1, Pik3ip1 exhibited downregulated expression on resting T cells after short-term in vitro culture. Our observation agrees with a previous study by the LP Kane group (16), which showed that Pik3ip1 surface expression is downmodulated in T cells to allow T-cell activation in an in vitro D10 T-cell activation model. Thus, we further explored the expression pattern of Pik3ip1 in different T-cell subsets in an in vivo antigen-specific activation model as well as a mouse tumor model, and checked the expression statuses of other hematopoietic cell compartments. The antigen-dependent decreasing expression pattern by Pik3ip1 might reflect a necessary role for the lymphoid tissue microenvironment in maintaining or regulating Pik3ip1 expression, but not an outcome of a wide-spectrum T-cell activation. However, when continuous tumor antigen stimulation was encountered, a pattern of Pik3ip1 expression on T cells following decrease from the peripheral lymph organs to the TME (Fig. 1D), which could reflect the necessity and complexity of the regulation by transcription factors or cytokines in maintaining Pik3ip1 expression in the TME.
Although it is shown that the infiltration of a large number of Tregs into a tumor is often associated with a poor prognosis. However, the role of Tregs in the tumor-immune microenvironment is associated with their subgroups. It has been reported that the presence of CD45RA− Foxp3lo CD25lo Treg cells may indicate better prognosis in tumors (34). In Fig. 6D, although we observed a decreased number of Tregs in Pik3ip1 fusion protein treatment group, we also reported more effector T cells and a more activated T-cell phenotype. We speculate whether Pik3ip1 affecting the specific Tregs differentiation. To elucidate the function of Pik3ip1 in those Tregs, future studies using chimeric mice, Foxp3-cre mice, or a Treg adoptive transfer model are required.
PI3K/Akt/mTOR pathway has been shown to regulate tumor immunosuppressive environment through tumor-intrinsic and immune-intrinsic effects. Within the TME, PI3K isoform inhibitors have been shown to promote CTL infiltration through limiting immunosuppressive cytokines/chemokines or immune checkpoint ligands expression by tumor cell (35, 36). On the other hand, PI3K inhibition could attenuate naïve T-cell differentiation into Th effector lineages or promote Treg expansion. Inhibitors of PI3K reduced tumor burden through limiting recruitment of immunosuppressive components including MDSCs and Treg cells in tumors (11, 37, 38). In this study, we revealed that Pik3ip1, a critical regulator of PI3K, serves as an essential rheostat for T-cell–mediated immunity, which sets the threshold for T-cell responses under homeostatic conditions and neoantigen exposure, especially tumor antigen stimulation. A similar subtle mechanism has been reported for other molecules such as Itk and TIPE2, which serve as modulators to critically fine tune the T-cell response (39, 40). Furthermore, given that PI3K is a major family of intracellular signal-transducing enzymes involved in a multitude of cellular processes, BM chimeric models were used to carefully select the role of Pik3ip1 in immune regulation. These results confirmed that Pik3ip1 deficiency decelerated tumor progression via solely an immune cell-associated effect, which could provide a new target for immunotherapeutic strategies.
A previous study by the LP Kane group showed enhanced activation of Akt and pS6 in Pik3ip1−/− mice compared with littermates following TCR activation (16). pS6 is downstream of the PI3K/Akt/mTOR pathway, which is often related to cell proliferation and survival. Interestingly, we found several other downstream pathways, such as the ERK1/2 and p38 MAPK pathways, in addition to the Akt signaling pathway could inhibited by Pi3kip1. Suggesting that Pik3ip1 may have a global effect on the regulation of TCR signaling by interacting with certain upstream proximal adaptors including SLP76. SLP76 has been reported to have a remarkable function in the development or function of T cells. Sixty percent of SLP76-deficient mice die during the perinatal period (41), and transgenic expression of SLP76 in the T-cell compartment rescues T-cell development (42). Our data demonstrated that Pik3ip1 can directly interact with SLP76 to mediate SLP76 degradation, thus acting as a brake for TCR signaling. Intriguingly, we found that the inhibitory function of Pik3ip1 required Pik3ip1 oligomerization. Upon interacting through its extracellular regions, Pik3ip1 formed oligomers to degrade SLP76. This working model further validated that a Pik3ip1 extracellular domain–IgG fusion protein can modulate the cellular composition of the tumor-immune microenvironment and enhance MC38 tumor growth. Because Pik3ip1 can be expressed in many types of tumor cells and tumor-infiltrating T cells, tumor cells might inhibit T-cell functions through a physical interaction among Pik3ip1 molecules on both cell surfaces. Thus, developing an anti-Pik3ip1 antibody might be another potential method for future cancer treatment. Correspondingly, PIK3IP1 expression also correlated with T-cell dysfunction in human tumors.
Collectively, our findings reveal a critical negative regulatory role for Pik3ip1 in antitumor T-cell responses by degrading SLP76 via oligomerization. Our findings identify Pik3ip1 as a potential molecular target for cancer diagnosis or immunotherapy.
Disclosure of Potential Conflicts of Interest
L. Chen reports receiving commercial research grants from and holds ownership interest (including patents) in NextCure and Tayu Biotech, is a consultant/advisory board member for NextCure, Vcanbio, GenomiCare, and Zai Lab. No potential conflicts of interest were disclosed by the other authors.
Conception and design: J. Wang, J. Cui, L. Chen, B. Cheng, Z. Wang
Development of methodology: Y. Chen, J. Wang, J. Song, J. Fang, X. Liu, Q. Li, S. Wen, J. Xia, L. Luo
Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): Y. Chen, J. Song, J. Fang, X. Liu, T. Liu
Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): Y. Chen, J. Wang, X. Wang, X. Li, J. Song, J. Fang, D. Wang, D. Ma, L. Chen
Writing, review, and/or revision of the manuscript: J. Wang, J. Cui, B. Cheng, Z. Wang
Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): J. Wang, X. Wang, Q. Li, S. Wen, S.G. Zheng, J. Cui
Study supervision: J. Wang,
Other (performed the experiments and analyzed data): X. Wang
Other (performed survival data analysis): X. Liu
Other (contributed resources): G. Zeng
This project was supported by grants from National Natural Science Foundations of China (Nos. 81772896, 81630025, 81602383) and Science and Technology Planning Project of Guangzhou City of China (No. 2017004020102).
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.