Article Text
PDF
Abstract
Backgrounds Chimeric antigen receptor (CAR)-modified T cells (CAR-T) are limited in solid tumors due to the hostile tumor microenvironment (TME). Combination therapy could be a promising approach to overcome this obstacle. Recent studies have shown that sphingosine 1-phosphate receptor (S1PR)3 has tremendous potential in regulating the immune environment. However, the functional significance of S1PR3 in T-cell-based immunotherapies and the molecular mechanisms have not been fully understood.
Methods Here, we studied the combination of EpCAM-specific CAR T-cell therapy with pharmacological blockade of S1PR3 against solid tumor. We have applied RNA sequencing, flow cytometry, ELISA, cellular/molecular immunological technology, and mouse models of solid cancers.
Results Our study provided evidence that S1PR3 high expression is positively associated with resistance to programmed cell death protein-1 (PD-1)-based immunotherapy and increased T-cell exhaustion. In addition, pharmacological inhibition of S1PR3 improves the efficacy of anti-PD-1 therapy. Next, we explored the possible combination of S1PR3 antagonist with murine EpCAM-targeted CAR-T cells in immunocompetent mouse models of breast cancer and colon cancer. The results indicated that the S1PR3 antagonist could significantly enhance the efficacy of murine EpCAM CAR-T cells in vitro and in vivo. Mechanistically, the S1PR3 antagonist improved CAR-T cell activation, regulated the central memory phenotype, and reduced CAR-T cell exhaustion in vitro. Targeting S1PR3 was shown to remodel the TME through the recruitment of proinflammatory macrophages by promoting macrophage activation and proinflammatory phenotype polarization, resulting in improved CAR-T cell infiltration and amplified recruitment of CD8+T cells.
Conclusions This work demonstrated targeting S1PR3 could increase the antitumor activities of CAR-T cell therapy at least partially by inhibiting T-cell exhaustion and remodeling the TME through the recruitment of proinflammatory macrophages. These findings provided additional rationale for combining S1PR3 inhibitor with CAR-T cells for the treatment of solid tumor.
- tumor microenvironment
- receptors, chimeric antigen
- macrophages
- lymphocytes, tumor-infiltrating
- immunotherapy
Data availability statement
All data relevant to the study are included in the article or uploaded as supplementary information.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
- tumor microenvironment
- receptors, chimeric antigen
- macrophages
- lymphocytes, tumor-infiltrating
- immunotherapy
WHAT IS ALREADY KNOWN ON THIS TOPIC
Chimeric antigen receptor (CAR)-modified T cells (CAR-T) are limited in solid tumor due to the hostile tumor microenvironment. Recent studies have shown that sphingosine 1-phosphate receptor (S1PR)3 has tremendous potential in regulating the immune environment. However, the functional significance of S1PR3 in T-cell-based immunotherapies and the molecular mechanisms have not been fully understood.
WHAT THIS STUDY ADDS
In this study, we show that S1PR3 could increase the antitumor activities of CAR-T cell therapy at least partially by inhibiting T-cell exhaustion and remodeling the tumor microenvironment through the recruitment of proinflammatory macrophages.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
Our study demonstrated S1PR3 as a promising target to change the hostile tumor microenvironment. Accordingly, our findings supported the development of new antitumoral treatments by combining S1PR3 inhibitor and CAR-T cells against solid tumor.
Introduction
Chimeric antigen receptor (CAR) T-cell immunotherapy (CAR-T) has revolutionized adoptive cell therapy (ACT) in hematological malignancies.1 2 Clinical efficacy in solid tumors, however, faces great challenges and has been repeatedly hindered. One of the major obstacles is the hostile tumor microenvironment (TME) of solid tumors; immunosuppressive cells such as M2 macrophages, regulatory T cells (Tregs), and myeloid-derived suppressor cells (MDSCs) in the TME and immunosuppressive ligands, such as programmed cell death 1 ligand 1 (PD-L1), present in the TME might all induce intrinsic antitumor immune responses as well as CAR-T cell responses.3–6 Recent strategies to improve CAR-T cell therapy mainly involve engineering the CAR protein, T cells, and the interaction between T cells and other components in the TME.7 It is still difficult for CAR-T cell therapy alone to obtain good clinical efficacy in solid tumors. Combination therapy with immune-modulatory agents may be a potential way to overcome this obstacle.
Sphingolipid sphingosine-1-phosphate (S1P), a bioactive lipid molecule, plays an important role in cancer and disorders of the immune system.8 9 S1P exerts extracellular functions by binding to five specific G-protein-coupled receptors (GPCRs), S1P receptors (S1PRs) 1–5. For the most part, binding of S1PR3 has been shown to promote the occurrence and development of tumors by activating downstream signaling pathways, including phosphatidylinositol 3-kinase (PI3K), mitogen-activated protein kinase (MAPK), and phospholipase C (PLC).10–12 Some reports have also established a negative role of S1PR3 in immune-modulatory effects. On the one hand, several studies have shown that S1PR3 exerts immunosuppressive effects by directly promoting tumor angiogenesis.10 13 On the other hand, it has also been shown that S1PR3 can modulate the macrophage cytokine phenotype toward an immunosuppressive profile,14 15 which inhibits the functions of immune effector cells. As an important inflammatory mediator, S1P binding to the S1PR3 receptor may also be involved in regulating the release of various anti-inflammatory cytokines in the TME, promoting cancer with inflammation.8 These studies suggested that S1PR3 is a promising candidate target for cancer immunotherapies. However, the functional significance of S1PR3 in T-cell-based immunotherapies and the molecular mechanisms have not been well established.
Here, our database screening of differentially expressed genes between patients who responded to programmed cell death protein 1 (PD-1) immunotherapy and those who did not identified S1PR3 as a crucial genetic alteration that correlated with T-cell exhaustion and the efficacy of PD-1 immunotherapy. Importantly, S1PR3 inhibition reduces T-cell exhaustion and might regulate the recruitment of proinflammatory macrophages to normalize the TME, improving CAR-T immunotherapy in mouse breast and colon cancer models. Thus, S1PR3 may be a druggable target to improve T-cell-based cancer immunotherapy.
Results
High S1PR3 expression is positively associated with poor prognosis and T-cell exhaustion
We previously performed a wide genetic screen to initially identify 25 functional genes that potentially induce nivolumab resistance in renal clear cell carcinoma (RCC) (GSE67501) and melanoma (GSE79691) from the Gene Expression Omnibus (GEO) database (online supplemental figure 1A–C). Among them, S1PR3 was the top candidate in the list (figure 1A and online supplemental figure 1D). Furthermore, S1PR3 was highly expressed in a variety of malignant tumors (online supplemental figure 1E and F) and differed significantly in patients with effective nivolumab treatment compared with those who did not respond, which was not observed in the other four S1P receptor family members (online supplemental table 1). To further test whether S1PR3 is associated with other inhibitory receptors in T cells, we examined public cancer databases from The Cancer Genome Atlas (TCGA) melanoma sets and found that S1PR3 shared a strong correlation with the expression of negative immune receptor genes (eg, IDO2, IL-10, FOXP3, TGFB1, PDCD1, TIM3, TIGIT and CTLA4) (figure 1B). Moreover, S1PR3 was positively correlated with the T-cell exhaustion signature (TIM3, TIGIT, LAG3, PDCD1, CXCL13 and LAYN) in different patients with cancer, which was not obviously observed in the other four S1P receptor family members (figure 1C and online supplemental figure 2A). Consistent with the above observations, we found that high expression of S1PR3 was significantly associated with worse overall survival (OS) outcomes, disease-free survival (DFS) and poor tumor stage (figure 1D and online supplemental figure 1G and H) in multiple cancers, which further supported the potential clinical relevance of S1PR3.
Supplemental material
Supplemental material
Supplemental material
Supplemental material
High S1PR3 expression is positively correlated with T-cell exhaustion. (A) Eleven common differential expressed genes (DEGs) in 11 patients with renal cell carcinoma (RCC) (GSE67501) and 9 patients with melanoma (GSE79691) received nivolumab treatment, response versus no response. DEGs were defined with |logFC|≥1.0 and adjusted p<0.05. (B) Correlation analysis between S1PR3 expression with IL10, IDO2, PDCD1, TIM3, CTLA4, TIGIT, FOXP3 and TGFB1 in colorectal cancer. (C) Correlation analysis between S1PR3 expression with T-cell exhaustion signature (eg, HAVCR2, TIGIT, LAG3, PDCD1, CXCL13 and LAYN) in colon adenocarcinoma (COAD), pancreatic adenocarcinoma (PAAD), melanoma, kidney renal clear cell carcinoma (KIRC)/kidney renal papillary cell carcinoma (KIRP), prostate adenocarcinoma (PRAD) and stomach adenocarcinoma (STAD). (D) Kaplan-Meier survival analysis comparing overall survival (OS) between patients with high and low S1PR3 expression in COAD, lung adenocarcinoma (LUAD), KIRP and uveal melanoma (UVM). Log-rank test was used for statistical analysis. (E) Flow-cytometry showing the correlation between S1PR3 high expression with the inhibitory receptors (PD-1, TIM-3, and LAG-3) in Balb/c mouse lymph nodes derived CD3+T cells (n=3). Data were analyzed using unpaired Student’s t-test. (F) Schematic of the treatment schedule for the TY-52156. Balb/c mice were injected with CT26 cells (1×106 cells) in the right flank, when the tumor volume came to 80~100 mm3 on day 7, mice were treated with PBS or S1PR3 antagonist TY-52156 at 10 mg/kg orally every other day. ig, intragastric administration; s.c. subcutaneous administration. (G) CT26 tumor volume was measured and recorded by the caliper every 2 days. Tumor growth was plotted starting from the day before initial dose of TY-52156 (n=8–10 mice pooled). Data were from one of two independent experiments. Mixed-effects analysis. (H) Survival curves of each cohort shown in (F) (n=5–7). Log-rank test was used for statistical analysis. (I) Representative flow cytometry plot showing the fraction of CD3+T cells, CD8+T cells and PD-1+CD8+, LAG-3+CD8+, TIM-3+CD8+exhausted T cells in tumor tissue of mice bearing CT26 tumors on day 20 after treatment (gated on live cells). (J) Quantification showing the individual values of identifying CD3+T cells, CD8+T cells and PD-1+CD8+,LAG-3+CD8+, TIM-3+CD8+exhausted T cells in tumor tissue of mice bearing CT26 tumors on day 20 after treatment. n=3 mice per group. Unpaired Student’s t-test. (K) Promoted tumor regression by treatment with combination of S1PR3 antagonist and anti-PD-1. Balb/c mice were injected with CT26 cells (5×105 cells) in the right flank. When the tumor volume came to 70–80 mm3 on day 7, mice were treated with PBS or S1PR3 antagonist TY-52156 at 10 mg/kg orally every other day and injected with anti-PD-1 200 µg intraperitoneally. Tumor volumes were monitored and recorded by the caliper every 2 days. Compound dosing was discontinued on day 20. Tumor growth was plotted starting from the day before initial treatment (n=8–10 mice pooled). Mixed-effects analysis. Statistical significance was determined at the levels of *p<0.05; **p<0.01; ***p<0.005 and ****p<0.001. LAG-3, lymphocyte-activation gene 3; PBS, phosphate buffer saline; PD-1, programmed cell death protein 1; S1PR, sphingosine 1-phosphate receptor; TIM-3,T cell immunoglobulin domain and mucin domain-3; TPM,transcripts per million.
To study the functional role of S1PR3 in cancer and T cells, we first detected the expression of S1PR3 in mouse breast cancer cells, colon cancer cells and mouse T cells (online supplemental figure 3A). We then chose a potent and selective S1PR3 antagonist, TY-52156, to explore the impact of pharmacological S1PR3 inhibition on different cells. According to the in vitro analysis, TY-52156 could effectively bind and reduce S1PR3 on tumor cells and T cells when the concentration reached 10 µM, and less S1P was produced in the supernatant caused by TY-52156 (online supplemental figure 3A and B). Next, we measured the protein expression of inhibitory receptors and S1PR3 in mouse T cells. High S1PR3 expression was positively correlated with the presence of an “exhausted” phenotype in T cells (figure 1E). We then studied its functional role in cytolytic T cells. We took advantage of a transplantable mouse CT26 colon cancer model (figure 1F) and found that the growth rates of CT26 tumors were significantly decreased and survival increased in mice treated with the selective S1PR3 inhibitor TY-52156 compared with the PBS control (figure 1G and H). Next, we wondered whether S1PR3 affects the antitumor function of tumor-infiltrating T cells (TILs) by regulating T-cell exhaustion. To test whether pharmacological S1PR3 inhibition diminishes T-cell exhaustion, we sacrificed animals from the CT26 tumor cohort and analyzed TIL and T-cell exhaustion markers (PD-1, TIM-3 and LAG-3) by flow cytometry on day 20. We identified a much smaller number of CD8+ T cells in the S1PR3 inhibitor-treated group expressing PD-1, TIM-3, and LAG-3 than in the PBS group (figure 1I and J). TY-52156 also sensitized B16-F10 and CT26 tumors to PD-1 checkpoint inhibitor therapy (figure 1K and online supplemental figure 4A and B). Taking these data together, we suggest that S1PR3 may mediate T-cell exhaustion and inhibition of antitumor immune responses. Targeting S1PR3 provides a rational immune-regulating agent combination with T-cell-based immunotherapies such as anti-PD-1 or CAR-T-cell therapy.
Supplemental material
Supplemental material
Pharmacological S1PR3 inhibition of solid tumor cell direct activity and cytotoxicity of murine EpCAM-CAR T cells in vitro
Previous work from our group developed third-generation murine CAR constructs targeting EpCAM (figure 2A). EpCAM-specific CAR-T cells were transduced and expanded with interleukin (IL)-2. Forty-eight hours after retrovirus transduction, 60–80% of T cells stably expressed the CAR structure on their surface (figure 2B). Both the EpCAM-positive solid tumor cell lines 4T1 and MC38 were sorted successfully, making these cell lines ideal target cells (figure 2C). However, a recent study by our group showed that 5×106 high-dose EpCAM CAR-T cells reinfusion may cause pulmonary toxicity and lack long-term tumor control.16 A hostile TME is one of the major obstacles. Given the short-term efficacy of murine EpCAM CAR-T cells in solid tumors in vivo, we sought an approach that would further enhance the migration and persistence of CAR-T cells. Thus, we next investigated the effects of pharmacological S1PR3 inhibition on CAR-T cell immunotherapy.
S1PR3 inhibition directs function and cytotoxicity of mEpCAM-CAR-T cells in vitro. (A) Schematic diagram of the modular composition of EpCAM-targeting CAR used in this study. The mouse third generation EpCAM-specific CAR is composed of a mouse CD8a signal peptide and antibody derived single-chain (heavy and light) variable fragment (scFv), a CD8a hinge and transmembrane (TM) domain and murine CD28, 4-1BB, and CD3ζ signaling domains. (B) Modified T cells expressing the CAR on the surface were determined by flow cytometry 48 hours after transduction. Percentages show the numbers of CAR positive cells. Untransduced T cells were used as control T. (C) Flow cytometry of EpCAM expression on the indicated target cell lines. B16-F10 cells do not express EpCAM antigen, while both 4T1 (BALB/c breast cancer cell line) and MC38 (C57BL/6J colon cancer cell line) cells are EpCAM antigen positive. (D) ELISA analysis of IFN-γ, IL-2 and TNF-α production in supernatants collected from CAR-T or control T cells co-cultured with 4T1 or MC38 cells in 96-well plates at increasing effector to target cell (E:T) ratios in the presence or absence of TY-52156 (10uM). Culture supernatant was collected 24 hours after incubation. (E) EpCAM CAR-T cells were cocultured with 4T1 cells in the presence or absence of TY-52156 (10 µM) at a 4:1 E:T ratio for 24 hours. Representative intracellular staining of IFN-γ and granzyme B in CD8+CAR-T cells (left) and summarized data (right) are shown. Cytokine secretion was blocked by brefeldin A and monensin for 6 hours before IFN-γ and granzyme B antibody staining. (F,G) Flow cytometry-based cell surface expression of (F) central memory T cell phenotype (CD44+CD62L+), effector memory T cell phenotype (CD44+CD62L−) and (G) T-cell activation phenotype (CD25+ and CD69+). EpCAM CAR-T cells were cocultured with 4T1 cells in the presence or absence of TY-52156 (10μM) at a 4:1 E:T ratio for 48 hours. (H) Percentage of CD4+ or CD8+ CAR-T cells positive for cell surface markers of T cell exhaustion (PD-1, LAG-3, and TIM-3) on 48 hours after TY-52156 stimulation (10 µM) with 4T1 cells (4:1 E:T ratio). (I) Significant enriched KEGG pathway terms of differentially expressed genes (DEGs) between EpCAM CAR-T cells + TY-52156 group and EpCAM CAR-T cells group at 72 hours post-stimulation with 4T1 cells (|logFC|≥1, Q≤0.05). (J) Heatmap of selected DEGs in comparisons between CD3+ T cells from EpCAM CAR-T cells + TY-52156 group and EpCAM CAR-T cells group at 72 hours post-stimulation with 4T1 cells. (K) Gene-set enrichment analysis (GSEA) of glycosphingolipid metabolism (up) and GTPases activate NADPH oxidases (down) signaling pathways was performed on all gene sets. Data are presented as mean±SD and are one result of three independent experiments. Statistical significance was determined by two-tailed unpaired Student’s t-test. *p<0.05; **p<0.01; ***p<0.005 and ****p<0.001. CAR, chimeric antigen receptor; CAR-T, CAR-modified T cells; IFN, interferon; IL, interleukin; KEGG, Kyoto Encyclopedia of Genes and Genomes; LAG-3, lymphocyte-activation gene 3; PD-1, programmed cell death protein 1; PD-L1, programmed cell death 1 ligand 1; S1PR, sphingosine 1-phosphate receptor; TIM-3, T cell immunoglobulin domain and mucin domain-3; TNF, tumor necrosis factor.
The cytotoxic ability of CAR-T cells directly determines antitumor activity. We first detected S1PR3 protein expression on EpCAM CAR-T cells by flow cytometry, which was similar to that on untransduced T cells, indicating that the introduction of CAR did not affect S1PR3 protein expression. Moreover, a potent and selective S1PR3 inhibitor effectively blocked S1PR3 on CAR-T cells (online supplemental figure 3A). To test the impact of pharmacological S1PR3 inhibition on the cytotoxicity of CAR-T cells, murine EpCAM CAR-T cells were stimulated with 4T1 or MC38 cells in vitro for 24 hours. Cytokine secretion by CAR-T cells in response to target antigen was measured in the presence or absence of TY-52156. As shown in figure 2D, treatment with TY-52156 significantly increased CAR-T cell cytokine release of interferon (IFN)-γ, IL-2 and tumor necrosis factor (TNF)-α after coculture with target cells 4T1 or MC38. Cytokine secretion by CAR-T cells in response to antigen stimulation was also enhanced in the presence of different concentrations of TY-52156 (online supplemental figure 5A). Similar promotion trends of antigen-induced expression of the intracellular cytotoxic effector molecules IFN-γ and granzyme B were also increased in EpCAM CAR-T cells when a pharmacologic dose of TY-52156 was used (figure 2E). To verify the enhancement effect of S1PR3 inhibition on the killing function of CAR-T cells, we selected CAY10444, another selective inhibitor targeting S1PR3, and found that the cytokine release level of CAR-T cells was also greatly enhanced (online supplemental figure 6A). Further confirmed that S1PR3 blockage directly affects T cells, thus leading to T cells functioning better. We precultured CAR-T cells with S1PR3 antagonist for 24 hours and then killed target cells in the absence of the inhibitor, and we found that the CAR-T cells precultured with S1PR3 inhibitor released higher levels of cytokines and had stronger killing function (online supplemental figure 5B). Interestingly, S1PR3 inhibitor can also reduce the levels of IL-6 and transforming growth factor(TGF)-β secreted by tumor cells, which is conducive to enhancing the function of T cells (online supplemental figure 9). These results suggested that pharmacologic inhibition of S1PR3 not only directly increased the cytotoxic capacity of CAR-T cells under conditions of antigen exposure, but also led to tumor cells secretome change, thus indirectly affecting the T-cell function.
Supplemental material
Supplemental material
Supplemental material
S1PR3 inhibition preserves the Tcm phenotype and reduces CAR-T cell exhaustion
To determine whether pharmacologic inhibition of S1PR3 altered the distribution of T-cell activation and memory phenotypes, the phenotype of CAR-T cells in response to antigen stimulation was then evaluated. Studies have demonstrated that CAR-T cells with a central memory phenotype have higher antitumor activity than naive or effector memory CAR-T cells.17 Given the role of S1P in T-cell migration, we determined whether S1PR3 antagonist-stimulated CAR-T cells exhibit a memory phenotype. After long-term antigen stimulation, cell surface marker expression indicative of the CD44+CD62L+ central memory T-cell (Tcm) phenotype, which normally decreases with proliferation, was found to be maintained in S1PR3 antagonist-treated CD8+ CAR-T cells (figure 2F), while cell surface marker expression indicative of the effector T cell (Tem) phenotype was not enhanced (figure 2F). In addition, upregulation of the activation markers CD69 and CD25 (figure 2G and online supplemental figure 6B) was also observed in S1PR3 antagonist-treated EpCAM CAR-T cells on target antigen stimulation for 48 hours. The continuous activation of CAR-T cells and the maintenance of the central memory phenotype are conducive to the antitumor activity of CAR-T cells, improving cell persistence in vivo.
Inducing T-cell exhaustion is a key immune escape mechanism for tumor cells. As S1PR3 shared a strong correlation with T-cell exhaustion, we then tested whether S1PR3 inhibition reversed the exhaustion status of CAR-T cells. We examined the abundance of the exhaustion-related cell surface markers programmed cell death protein 1 (PD-1), T cell immunoglobulin domain and mucin domain-3 (TIM-3), and lymphocyte-activation gene 3 (LAG-3). After target antigen stimulation by incubation with 4T1 cells for 48 hours and pharmacological S1PR3 inhibition, the proportion of cells positive for PD-1 and LAG-3 was significantly lower for both CD4+ and CD8+ EpCAM CAR-T cells treated with TY-52156 (figure 2H). The proportion of cells positive for TIM-3 in CD8+ EpCAM CAR-T cells treated with TY-52156 was also downregulated (figure 2H). Another potent S1PR3 antagonist, CAY10444, also reduced CAR-T cell exhaustion in vitro (online supplemental figure 6C). The reduction in these immune checkpoint inhibitor proteins could contribute to higher numbers of functional T cells within a solid TME, suggesting the possible role of TY-52156 in improving CAR-T-cell persistence in vivo.
S1PR3 inhibition upregulates T-cell activation, cytotoxicity, and energy metabolism-related gene expression and downregulates T-cell exhaustion-related gene expression
To better understand whether S1PR3 inhibition changes the transcriptional profiles of CAR-T cells, we stimulated EpCAM CAR-T cells with an S1PR3 antagonist and prolonged 4T1 tumor cell antigen stimulation at three time points: 24 hours, 48 hours, and 72 hours. Finally, we sorted CAR-T cells for RNA sequencing (RNA-seq) analysis at 72 hours (online supplemental figure 8A). A total of 5230 differentially expressed genes (DEGs) were identified (|logFC|≥1, Q≤0.05), including 2457 upregulated genes and 2773 downregulated genes (online supplemental figure 8B). Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment overview analysis found that the majority of the DEGs were enriched in signal transduction and the immune system (online supplemental figure 8C), specifically in the PD-1/PD-L1 pathway, T-cell receptor (TCR) signaling pathway, TNF signaling pathway and nuclear factor kappa-B (NF-κB) signaling pathway (figure 2I and online supplemental figure 8C and D). Notably, the S1PR3 antagonist caused downregulation of coinhibitory receptors and increased the expression of genes related to effector function; that is, messenger RNAs (mRNAs) encoding effector proteins and TCR signaling-associated genes (eg, IL-2Rα, IFN-γ, TNF, granzyme, CD28, CD25, CD69 and TCR adaptor proteins) were upregulated, and inhibitory surface receptors typically upregulated in hyporesponsive T cells (eg, PDCD1, HAVCR2, CD274, TIGIT, CTLA4 and LAG3) were downregulated (figure 2J). Gene set enrichment analysis (GSEA) further confirmed that T-cell activation-related cytokine gene expression, including IL-2, IFN-γ and T-cell proliferation and TCR signaling genes, was significantly upregulated in S1PR3 antagonist-treated EpCAM CAR-T cells (online supplemental figure 8F–I). Studies have suggested the involvement of a broad number of transcription factors (TFs), including NFAT, Eomes, T-bet, TOX, BATF, and FoxO1, a master regulator that initiates and controls the process of T-cell dysfunction. To determine whether S1PR3 regulates their expression, we analyzed RNA-seq data and demonstrated decreased expression levels of most TFs (figure 2J). Among them, the TF Eomes showed the most significant change. Meanwhile, KEGG pathway analysis showed an interaction between SP1R3 and lipid metabolism at the transcriptome level (online supplemental figure 8C). In addition, GSEA performed on T-cell energy metabolism-related gene sets further confirmed the upregulation of glycosphingolipid metabolism and that GTPases activate NADPH oxidase signaling-associated genes, indicating that inhibiting S1PR3 enhances T-cell energy metabolism (figure 2K). Taken together, we concluded that under prolonged antigen exposure, CAR-T cells treated with an S1PR3 inhibitor still showed stronger activation and cytotoxicity. The continuous enhancement of TCR signaling, the enhanced expression of cytokines such as IL-2, IL-21 and IFN-γ that antagonize T-cell exhaustion, and the enhanced lipid metabolism and oxidative phosphorylation metabolism eventually contribute to the reduction in T-cell exhaustion.
Supplemental material
S1PR3 inhibition improves CAR-T-cell mediated tumor control and survival
To further investigate whether S1PR3 antagonist could enhance the antitumor activities of CAR-T cells, mice bearing EpCAM-positive MC38 colon cancer and 4T1 breast tumor xenografts were established (figure 3A). Each mouse was intraperitoneally injected with cyclophosphamide (CTX) for lymphodepletion on day 0 and infused EpCAM CAR-T cells or control T cells on days 1 and 3, followed by TY-52156 or CAY10444 treatment at a dose of 10 mg/kg once every other day for 1 week (figure 3A and online supplemental figure 6D). Tumor growth and survival were observed for 2 months. In a syngeneic tumor model using C57BL/6J mice bearing subcutaneous MC38 tumors, our data showed that EpCAM CAR-T cells significantly reduced tumor growth compared with control T cells for the first 2 weeks of treatment (figure 3B, right). All tumors, however, eventually progressed, and survival was not significantly improved by this treatment regimen (figure 3C). This observation was more obvious in the immunosuppressive 4T1 breast cancer tumor models. Neither EpCAM CAR-T therapy nor TY-52156 treatment alone inhibited tumor growth or improved mouse survival (figure 3F,G).
Antitumor efficacy of combination therapy of S1PR3 inhibitor and EpCAM CAR-T cells in immunocompetent murine tumor model. (A) Schematic of the treatment schedule for the TY-52156 and EpCAM CAR-T cells combination therapy. C57BL/6J mice were subcutaneously injected with MC38 cells (1×106 cells) in the right flank and treated with CTX (2.5 mg/kg) when the tumor volume came to ~100 mm3 on day 0. Balb/c mice were injected with 4T1 cells (5×105 cells) in situ into the fourth inguinal mammary fat pads and treated with CTX (2.5 mg/kg) when the tumor volume came to~80 mm3 on day 0. On day 1 and 3, 1×106 EpCAM CAR-T cells or control T cells were intravenously injected into tumor-bearing mice. On day 0, 2, 4, 6 and 8, PBS or S1PR3 antagonist TY-52156 at 10 mg/kg was administered orally for four times. CTX, cyclophosphamide. i.p. intraperitoneal injection. i.v. intravenous injection. ig. intragastric administration. (B) MC38 tumor volume was measured and recorded by the caliper every 2 days. The left side is the tumor volume data of all groups, and the right side is the line graph of the tumor volume in the EpCAM CAR-T cells (green) and EpCAM CAR-T cells combination with the TY-52156 group (red). Tumor growth was plotted starting from the day before initial dose of CTX (n=7–10 mice pooled). Data were from one of two independent experiments. Mixed-effects analysis. (C) Survival curves of each cohort shown in (B) (n=5). Log-rank test was used for statistical analysis. (D,E) Tumor weight (n=3) (D) and body weight (n=7–10) (E) of MC38 tumors in mice given different treatment. Results were determined by two-tailed unpaired Student’s t-test. (F) 4T1 tumor volume was measured and recorded by the caliper every 2 days. The left side is the tumor volume data of all groups, and the right side is the line graph of the tumor volume in the EpCAM CAR-T cells (green) and EpCAM CAR-T cells combination with the TY-52156 group (red). Tumor growth was plotted starting from the day before initial dose of CTX (n=7–10 mice pooled). Data were from one of the two independent experiments. Mixed-effects analysis. (G) Survival curves of each cohort shown in (F) (n=5). Log-rank test was used for statistical analysis. (H,I) Tumor weight(n=3) (H) and body weight (n=7–10) (I) of 4T1 tumors in mice given different treatment. Results were determined by two-tailed unpaired Student’s t-test. All data are presented as the mean values±SD. Statistical significance was determined at the levels of *p<0.05; **p<0.01; ***p<0.005 and ****p<0.001. Data are pooled from representatives of two experiments. CAR, chimeric antigen receptor; CAR-T, CAR-modified T cells; PBS, phosphate buffer saline; S1PR, sphingosine 1-phosphate receptor.
Compared with other treatments, the combination of TY-52156 and EpCAM CAR T cells resulted in marked tumor regression (figure 3B) and longer survival times (figure 3C). Combination therapy produced complete regression and a curative response in almost all treated MC38 bearing-mice. Notably, 100% of the mice in the combination therapy group still survived even when all the mice in the other groups had died. Similar therapeutic benefits of this combination approach were observed in 4T1 breast tumor xenografts. Combination therapy with EpCAM CAR-T therapy and TY-52156 delayed tumor growth and increased mouse survival (figure 3F and G). Tumor weights were measured and were in accordance with tumor volume (figure 3D and H). The body weight of mice showed no significant difference in all groups (figure 3E and I). No histology and morphological abnormality appeared in the H&E staining of the tissue sections from heart, liver, spleen, lung, and kidney (online supplemental figure 9A and B), suggesting no severe toxicity caused by CAR-T and S1PR3 antagonist treatment in vivo. To check whether enhancement of EpCAM CAR-T-cell antitumor efficacy via inhibition S1PR3 is similar between different types of antagonists, we then investigated the therapeutic effects of CAY10444 on the 4T1 tumor model. Tumor size assessments revealed that, compared with EpCAM CAR-T cells, mice treated with the CAY10444 and EpCAM CAR-T cells had significantly suppressed tumor growth (online supplemental figure 6E and F).Considering these results, we showed S1PR3 antagonist sensitized solid tumor to CAR-T immunotherapy and the broad applicability of our combination approach in solid tumors.
Supplemental material
S1PR3 inhibition induces antitumor immunity and promotes tumor infiltration of CAR-T cells
We have seen that the S1PR3 antagonist has gained attention for its ability to induce T-cell infiltration and antitumor immunity (figure 1H). Therefore, we assessed the infiltration of T cells into the tumor after the S1PR3 antagonist in combination with mEpCAM CAR-T cells in the subcutaneous MC38 and 4T1 tumor model. Twenty days post CAR-T cells administration, we enumerated the total T cells in the tumor, blood, spleen and tumor-draining lymph nodes (TDLNs) via flow cytometry. The most noticeable difference in the TME in mice receiving S1PR3 antagonist was a substantial increase in T cells compared with animals given EpCAM CAR-T cells without S1PR3 antagonist. TY-52156 induced recruitment of CD3+CD8+ cytotoxic TILs to tumors, and this response was amplified in both solid tumors after the combination of TY-52156 and mEpCAM CAR-T cells (figure 4A and B). In addition, TY-52156 enhanced accumulation of T cells with an activated phenotype. Compared with single CAR-T cells treatment, TY-52156 in combination with CAR-T cells markedly increased the population of granzyme B positive CD8+ T cells in the MC38 tumor region (figure 4A). To assess whether S1PR3 inhibition changes T-cell infiltration to non-tumor tissues, we also analyzed changes of T cells in blood, spleen and TDLNs. Interestingly, TY-52156 treatment also enhanced the accumulation of CD8+ cytotoxic T cells in blood, spleen and TDLNs (online supplemental figure 10A–C). Similarly, another S1PR3 inhibitor, CAY10444 also increased CD8+ tumor infiltration T cells (online supplemental figure 6G and H), indicating that S1PR3 inhibition promotes systemic changes of T cells in immune microenvironment, the recruitment of T cells may be the result of local proliferation of these cells.
Supplemental material
S1PR3 inhibition promotes tumor infiltration of both endogenous T cells and adoptively transferred mEpCAM-CAR T cells in TME. (A) Left, representative flow cytometry plot identifying CD3+CD8+granzyme B+ endogenous TILs of mice bearing MC38 tumors on day 20 after treatment (gated on live cells). Right, quantification showing the individual values of CD3+TILs, CD8+TILs, and CD8+granzyme B+TILs. n=3 mice per group. (B) Left, representative flow cytometry plot of CD4+ and CD8+ endogenous TILs of mice bearing 4T1 tumors on day 20 after treatment (gated on CD3+ T cells). Right, quantification showing the individual values of CD3+TILs, CD8+TILs. n=3 mice per group. (C) Schematic of the treatment schedule for the TY-52156 and EpCAM CAR-T cells combination therapy. Balb/c mice were injected with 4T1 cells (5×105 cells) in situ into the fourth inguinal mammary fat pads and treated with CTX (2.5 mg/kg) when the tumor volume came to ~80 mm3 on day 0. On day 1, 1×106 EpCAM CAR-T cells or control T cells were intravenously injected into tumor-bearing mice. From day 0, S1PR3 antagonist TY-52156 at 10 mg/kg was administered orally every other day for four times. CTX, cyclophosphamide. i.p. intraperitoneal injection. i.v. intravenous injection. ig. intragastric administration. (D) IFN-γ and IL-2 cytokine production in blood of 4T1 tumor bearing mice on day 7. n=3 mice per group. (E) Left, representative flow cytometry plot of CAR+ T cells in tumor and blood of mice bearing 4T1 tumors on day 7 after treatment (gated on CD3+ T cells). Right, quantification showing the individual values of CAR+T cells in tumor and blood. n=3 mice per group. (F) Left, representative flow cytometry plot of PD-1,TIM-3 and LAG-3 on CAR+ T cells in tumor of mice bearing 4T1 tumors on day 7 after treatment (gated on CAR+ T cells). Right, quantification showing the individual values of PD-1,TIM-3 and LAG-3+CAR+T cells in tumor and blood. n=5 mice per group. All data are presented as the mean values±SD. Statistical significance was determined by two-tailed unpaired Student’s t-test at the levels of *p<0.05; **p<0.01; ***p<0.005 and ****p<0.001. Data are pooled from representatives of two experiments. CAR, chimeric antigen receptor; CAR-T, CAR-modified T cells; IFN, interferon; IL, interleukin; LAG-3,lymphocyte activation gene-3; PBS, phosphate buffer saline; PD-1, programmed cell death protein 1; S1PR, sphingosine 1-phosphate receptor; TILs, tumor-infiltrating T cells; TIM-3, T cell immunoglobulin domain and mucin domain-3; TME, tumor microenvironment.
To confirm that the increase in T cells in mice receiving S1PR3 antagonist and CAR-T combination therapy was most likely due to the expansion of CAR-T cells, flow cytometry was performed to detect CAR-T cells in the TME. We administered a single dose of TY-52156 at various time points prior to or after injection of EpCAM CAR-T cells (figure 4C). Like the findings with CD8+ T cells, there was a remarkable increase in probable EpCAM CAR-T cells in both TME and blood in 4T1 bearing mice receiving TY-52156 treatment (figure 4E). We also observed that a significant decreased fraction of exhausted CAR-T cells expressed PD-1, LAG-3 and TIM-3 in combination treatment group (figure 4F).In addition, we found increased IFN-γ and IL-2 cytokine production in the blood of 4T1 tumor bearing mice receiving TY-52156 and CAR-T combination therapy compared with single CAR-T cells treatment (figure 4D). Together, these data indicate that S1PR3 antagonist promoted antitumor immunity and recruitment of EpCAM CAR-T cells. Considering the significant impact of S1PR3 on the IL-2 signaling pathway (online supplemental figure 8G) and T-cell proliferation pathway(online supplemental figure 8H) observed in the in vitro RNA-seq analysis, we speculate that the S1PR3 inhibitor directly enhances local proliferation of CAR-T cells.
S1PR3 inhibition induces macrophage activation and proinflammatory phenotype alteration, reprogramming the TME
S1P is an important inflammatory mediator, and S1P-S1PR3 signaling is involved in the regulation of the immune inflammatory response.8 Thus, we assessed the immune responses in the TME of combination S1PR3 antagonist and EpCAM CAR-T cell therapy. Following the same ACT strategy as demonstrated above (figure 5A), we characterized the TME immediately on day 20 after EpCAM CAR-T cell infusion with TY-52156. Differences were found in the presence of macrophages, dendritic cells, Treg cells and MDSCs in mice treated with or without TY-52156. There was a marked increase in the presence of macrophages in the TME after treatment with TY-52156 in both MC38-bearing and 4T1-bearing mice (figure 5A and C). Similar trends were also observed in the bone marrow and spleen of tumor-bearing mice (online supplemental figure 11A and B). In addition, a marked decrease in the presence of MDSCs was observed in 4T1 tumors after TY-52156 and CAR-T-cell treatment compared with single CAR-T cell therapy (figure 5D). There was a significant increase in dendritic cells (DCs) and a decrease in the presence of Treg cells in MC38 tumors (figure 5E). Thus, the enhanced function of EpCAM CAR-T cells plus TY-52156 was strongly associated with the re-establishment of an immunosuppressive TME, turning immunologically “cold” tumors into “hot” tumors.
Supplemental material
S1PR3 inhibition increased immune cells recruitment in tumor tissue of tumor-bearing mice, reprogram tumor microenvironment. (A) Left, representative flow cytometry plot. Right, assessment of the frequency of CD45+immune cells, CD11b+F4/80+ tumor-infiltrating macrophages, CD206+ M2-like tumor-infiltrating microphages and I-A/I-E (MHC II) + M1-like tumor-infiltrating macrophages of mice bearing MC38 tumors on day 20 after treatment (n=3 each group). (B) Immune molecule mRNA level measured by q-PCR in TME of MC38 bearing mice (n=3 each group). Results are represented relative to the PBS group. (C) Left, representative flow cytometry plot (gated on live cells). Right, assessment of the frequency of CD11b+F4/80+tumor-infiltrating macrophages (% of live tumor cells) in 4T1 tumors of all treatment groups on day 20 after treatment (n=3 each group). (D) Assessment of the frequency of CD11b+Gr-1+ MDSCs (% of live tumor cells) in mice bearing 4T1 tumors on day 20 after treatment (n=3 each group). (E) Assessment of the frequency of CD11b+Gr-1+ MDSCs, CD4+CD25+Foxp3+ Treg cells and CD11b+CD11c+DC cells in MC38 tumors on day 20 after treatment (n=3 each group). All data are presented as the mean values±SD. Statistical significance was determined by two-tailed unpaired Student’s t-test at the levels of *p<0.05; **p<0.01; ***p<0.005 and ****p<0.001. Data are pooled from representatives of two experiments. Arg-1, arginase-1; CAR, chimeric antigen receptor; CAR-T, CAR-modified T cells; DCs, dendritic cells; IL, interleukin; iNOS,inducible no synthase; MCP, monocyte chemotactic protein; MDSCs, myeloid-derived suppressor cells; mRNA, messenger RNA; PBS, phosphate buffer saline; q-PCR, quantitative PCR; S1PR, sphingosine 1-phosphate receptor; TGF, transforming growth factor; TME, tumor microenvironment; Treg, regulatory T cell.
Tumor associated macrophages (TAMs) may exert either proinflammatory or proangiogenic functions during tumor progression, depending on whether they are polarized toward the M1 or M2 population, respectively. We analyzed the markers generally associated with M1 (major histocompatibility complex II (MHC II), inducible no synthase (iNOS), monocyte chemotactic protein (MCP-1)) or M2 (CD206, arginase-1 (Arg-1)) macrophage populations. In all samples analyzed, there were lower levels of tumor-residing M2-like (CD11b+F4/80+CD206+) macrophages in TY-52156 and CAR-T cells combination treated mice (figure 5A). mRNA expression of M2-like macrophages related to marker Arg-1 and CD206 were also significantly decreased in the TME after treatment with TY-52156 (figure 5B). This was associated with increased expression of markers associated with M1-like macrophages, including MHC II(I-A/I-E), iNOS and MCP-1 (figure 5A and B). MCP-1 is a chemokine for macrophages, indicating that TY-52156 enhanced the chemotaxis of intratumorally M1 macrophages. Both the activation of M1 macrophages and lower levels of M2 macrophages may contribute to tumor control. These results suggested that S1PR3 inhibition improving the antitumor effects of CAR-T cells was attributed to an influx of TAMs, M1-like polarization and reduction of M2 macrophages recruitment in tumor tissue of tumor-bearing mice.
We next investigated how the immune-regulatory molecule S1PR3 might influence macrophage generation and function. C57BL/6 mice bone marrow-derived macrophages (BMDMs) were differentiated in the presence or absence of TY-52156, followed by the addition of IFN-γ/lipopolysaccharides (LPS) or IL-4/IL-13 to cell cultures before functional analyses (figure 6D). Macrophage changes were first apparent after 48 hours of treatment, S1PR3 antagonist added to macrophages differentiated under neutral conditions (M-0), proinflammatory and anti-inflammatory conditions increased inflammatory cytokines levels such as TNF-α, IL-6 and IL-1β, which characterizes proinflammatory M1-like macrophages (figure 6E). S1PR3 antagonist treatment also significantly upregulated expression of MHC II (I-A/I-E) and downregulated expression of CD206, consistent with macrophage activation and M2 polarization inhibition (figure 6F). Real-time PCR of anti-inflammatory cytokines, cytokine receptors were downregulated by TY-52156 treatment (figure 6I), while IL-6, IL-1β, iNOS and MCP-1 transcripts were significantly upregulated (figure 6H). Together, these results indicate that S1PR3 inhibition facilitates differentiation, activation, and survival of macrophages with proinflammatory properties, in agreement with the increase in TAMs differentiation and higher M1-like macrophage infiltration identified in TY-52156 treated tumors. Thus, S1PR3 inhibition reduces the accumulation of suppressive macrophages and enhances the trafficking of T cells in the TME, promoting antitumor T-cell responses.
S1PR3 inhibition induces macrophage activation and phenotype altering, differentiates proinflammatory macrophages. (A) Scheme of C57BL/6 mice bone marrow cells differentiation in the presence of M-CSF for 6 days, followed by 48 hours polarization with cell culture medium (matured M0 macrophages), IFN-γ/LPS (M1-like macrophages) and IL-4/IL-13 (M1-like macrophages) in the presence or absence of TY-52156. q-PCR, ELISA, and flow cytometry analysis were performed for TY-52156 conditioned macrophages with different polarization and its control group (n=3). BMDMs, bone marrow derived macrophages. (B) Concentration of proinflammatory cytokines in the culture supernatants were measured by ELISA from different polarization groups with or without TY-52156 (n=3). (C) Expression of CD206 and I-A/I-E (MHC II) in M0-treated, IFN-γ/LPS and IL-4/IL-13-treated BMDMs in the presence or absence of TY-52156 (n=3). (D) Real-time PCR analyses showed that transcriptions of IL-1β, interleukin-6 (IL-6), iNOS and monocyte chemoattractant protein-1 (MCP-1) in IFN-γ/LPS-stimulated M1-like BMDMs were significantly enhanced by co-culturing with TY-52156. Results are represented relative to matured BMDMs (M0) (n = 3). (E) Real-time PCR analyses showed that the transcription levels of Arg-1, CD206, IL-10 and IL-4R in IL-4/IL-13-stimulated M2-like BMDMs were inhibited by co-culturing with TY-52156. Results are represented relative to matured BMDMs (M0) (n = 3). All data are presented as the mean values±SD. Statistical significance was determined by two-tailed unpaired Student’s t-test at the levels of *p<0.05; **p<0.01; ***p<0.005 and ****p<0.001. Data are pooled from representatives of three times technical replicates (E–H). Arg-1, arginase-1; IFN, interferon; iNOS, inducible no synthase; LPS, lipopolysaccharides; M-CSF, macrophage colony stimulating factor; mRNA, messenger RNA; q-PCR, quantitative PCR; S1PR, sphingosine 1-phosphate receptor.
Discussion
Compared with blood malignancies, solid tumors encounter major challenges to treatment with CAR-T cells, largely due to limited T-cell infiltration and a hostile immunosuppressive TME, particularly for immunologically cold tumors such as triple-negative breast cancer.18 Likewise, normalization of the hostile TME represents a promising strategy for improving immunotherapy. Recent insights into the possible role of S1PR3 as a therapeutic target have attracted enormous attention. As a potential modulator of the TME, S1P binds and activates S1PRs and can control various aspects of angiogenesis, inflammation, and immune cell functions, such as T cells, Tregs, MDSCs and macrophages. Our study further establishes that S1PR3, as an important extracellular GPCR of S1P, drives tumor-infiltrating T-cell exhaustion and that pharmacological inhibition of S1PR3 improves the efficacy of CAR-T cells in preclinical mouse models of solid tumors. Mechanistically, an S1PR3 antagonist could regulate the recruitment of proinflammatory macrophages to the TME and suppress the secretion of inhibitory cytokines such as IL-10, IL-6 and TGF-β, which are produced by and can polarize or attract suppressor cells. Furthermore, the S1PR3 antagonist enhanced the trafficking of T cells in the TME, eventually strengthening the antitumor function of infiltrating CAR-T cells (figure 7). As a result, S1PR3 may be an attractive drug target for boosting T-cell-based immune responses in patients with cancer.
S1PR3 inhibition enhances the anti-tumor ability of CAR-T cells by increasing the infiltration of CAR-T cells and remodeling TME. Left: The solid TME consists of the extracellular matrix, cancer-associated fibroblasts (CAFs), dysregulated tumor vasculature, various inhibitory immune cells, soluble molecules, such as tumor associated macrophages (TAMs), myeloid-derived suppressor cells (MDSCs) and regulatory cells (Tregs), to form a “cold” TME, which further suppresses the infiltration of CAR-T cells and limits treatment efficacy against solid tumor. Right: S1PR3 antagonist could regulates recruitment of proinflammatory macrophages to the TME and suppress the secretion of inhibitory cytokines such as IL-10, IL-6 and TGF- β, which are produced by and can polarize or attract suppressor cells. Further, S1PR3 antagonist could enhance the trafficking of T cells in the TME, eventually strengthening the antitumor function of infiltrating CAR T cells. Arg-1, arginase-1; CAR, chimeric antigen receptor; CAR-T, CAR-modified T cells; IFN, interferon; IL, interleukin; ROS, reactive oxygenspecies; S1PR, sphingosine 1-phosphate receptor; TGF, transforming growth factor; TME, tumor microenvironment; VEGF, vascular endothelial growth factor.
One of the hallmarks of cancer is that excessive suppressor determinants in the TME cause functional impairment of T cells, leading to T-cell exhaustion. A previous study found that intrinsic S1P signaling could engage the lipid transcription factor PPARγ to attenuate lipolysis capacity in T cells, driving T-cell dysfunction in an S1PR-independent manner.19 In addition, S1P can also affect the differentiation,20 21 proliferation22 and function of T cells21 23 by binding to its different receptors in a cell-intrinsic and cell-extrinsic manner. In our study, we found that the S1PR3 axis may drive T-cell exhaustion and dysfunction. High expression of S1PR3 was correlated with the T-cell exhaustion signature (TIM3, TIGIT, LAG3, PDCD1, CXCL13 and LAYN) and negative immune receptor genes (eg, IDO2, IL-10, FOXP3, TGFB1, PDCD1, TIM3, TIGIT and CTLA4), suggesting that S1PR3 inhibition may help T cells block endogenous inhibitory signals and extend their effector function. Consistent with exploratory database analysis, high S1PR3 expression was positively correlated with the presence of an “exhausted” phenotype in T cells. S1PR3 inhibition prevented T-cell exhaustion by downregulating the expression of PD-1, TIM-3, and LAG-3 on CAR-T cells, and attenuation of exhausted activity endowed these CAR-T cells with enhanced activation ability (enhanced CD25 and CD69 expression). S1PR3 may also have an important upstream role in the generation of memory T cells. S1PR3 inhibition increased the Tcm phenotype of CAR-T cells in vivo, potentially contributing to the robust recall responses of CAR-T cell therapy,17 leading to improved antitumor activity of CAR-T cells in vitro and in vivo. S1PR3 inhibition also promoted CAR-T cell recruitment and persistence at the tumor site. TY-52156 treatment was associated with a marked increase in the number of CAR-T cells through day 7 after therapy. Interestingly, these T cells also had an activated phenotype, as they expressed GZMB+ and generated more IFN-γ and IL-2 after therapy with an S1PR3 antagonist.
Our discovery of the S1PR3 axis that may drive T-cell exhaustion and dysfunction substantially extends the current knowledge of the molecular mechanism underlying this immunological process. Unfortunately, we have not yet ascertained the mechanistic link. It has been well demonstrated that S1P is an important inflammatory mediator that can regulate the release of various inflammatory cytokines, including IL-6, IL-10, TGF-β and all by mediating different receptors, such as S1PR3.8 24 IL-10 is a STAT3-induced cytokine, and the IL-10/IL-10R signaling pathway can reprogram T-cell metabolic profiles and restrict the function of exhausted T cells.25–27 TGF-β also plays an important role in mediating T-cell exhaustion. TGF-β upregulates CD70 expression on effector T cells, while it preferentially induces Foxp3 expression in naive T cells.28 TGF-β signaling could upregulate the expression of SPHK1 on T cells and further inhibit T-cell activation by mediating the SPHK1-S1P-PPARγ signaling axis.19 Activation of S1PR3 via the TGF-β/SMAD3 signaling pathway could also directly induce lung cancer progression and metastasis, indicating some connection between S1PR3 and TGF-β.11 In our study, we demonstrated that the expression of S1PR3 in solid tumors was positively correlated with the immunosuppressive cytokines IL-10 and TGF-β. Consistent with the above exploratory data set findings, S1PR3 antagonist was found to significantly inhibit tumor cell secretion of TGF-β in vitro and in vivo. Thus, our present hypothesis is a mechanistic link between S1PR3 and TGF-β in mediating T-cell exhaustion. S1PR3 deficiency may mediate inactivation of TGF-β signaling, thus overcoming endogenous CD8+ T-cell exhaustion and rendering CAR-T cells resistant to exhaustion. However, crosstalk between S1PR3 and TGF-β signaling may be more complicated, and additional studies will need to be explored. In addition, compared with effector T cells, exhausted T cells are now recognized to have a unique transcriptional program. Gene expression changes include transcription factors that control TCR signaling pathways, co-stimulation, cytokine signaling, as well as cellular metabolism.29 RNA-seq data suggest that S1PR3 inhibitors enhance TCR signal transduction of CAR-T cells, participate in transcriptional regulation, and enhance lipid metabolism and oxidative phosphorylation. Developing exhausted T cells often exhibit metabolic dysregulation, including suppressed cellular respiration, decreased glucose uptake, and mitochondrial energy imbalance. By blocking S1PR3, T cell energy metabolism can be restored, meeting the energy demands of activated effector T cells and thus reducing T-cell exhaustion.
Enhancing CAR-T cell trafficking and persistence without altering the immunosuppressive TME is unlikely to lead to significant sustained antitumor effects in solid tumors.30 The immunocompetent mouse model we developed allowed us to assess how the S1PR3 antagonist affected the TME and improved antitumor activity of CAR-T cells. We demonstrate that the persistence of EpCAM CAR-T cells in the TME was associated with enhanced M1-like macrophages (highly expression of murine MHC-II, iNOS, MCP-1) and a marked loss of M2-like macrophages (lower expression of CD206, Arg-1), which can inhibit antigen presentation and adaptive immune responses.31 32 Moreover, a rapid influx of I-A/I-E+CD206−iNOS high expression macrophages caused by S1PR3 antagonist within the TME was also observed and associated with enhanced control of tumor growth. S1PR3 also induces alterations in the balance of immune-stimulatory and suppressive macrophages. In our study, S1PR3 inhibition in BMDMs reversed IL-4/IL-13 induced M2 macrophages polarization with increased production of inflammatory cytokines, such as IL-6, TNF-α, MCP-1 and decreased expression of CD206, Arg-1. Recent work has highlighted the critical role of S1PR3 in the recruitment of anti-inflammatory monocytes that can enhance healing outcomes, tissue regeneration and sepsis recover.14 15 33 However, the mechanism of S1PR3-mediated macrophages polarization remains unclear. It has been reported that S1PR3 regulates myeloid differentiation and activates inflammatory programs by signaling pathways of the TNF-α–NF-κB axis.34 35 Recent studies demonstrated that S1P-induced anti-inflammatory macrophages are mediated through inhibition of the stimulator of interferon genes (STING) pathway.36 As a “Find me” signal of apoptotic cells, S1P can also activate erythropoietin (EPO) signal of macrophages, promote phagocytosis of apoptotic cells by activating PPARγ, inhibit inflammatory response, and lead to immune tolerance.37 Further investigations were needed to explain the role of S1PR3 in macrophage polarization.
Therefore, our work uncovers S1PR3 functions as a negative regulator of the function of T cells and macrophages, targeting S1PR3 may provide a new opportunity to remodel the TME and promote antitumor responses of CAR-T cell-based immunotherapy for treating solid tumors.
Materials and methods
Cell lines and cell culture
The murine breast cancer cell line 4T1, melanoma cell line B16-F10, and colon cancer cell line CT26 were obtained from American Type Culture Collection (ATCC). 4T1 cells and B16-F10 cells were cultured in high glucose Roswell Park Memorial Institute (RPMI) 1640 (Gibco) supplemented with 10% fetal bovine serum (FBS) (PAN, German) and 1% penicillin/streptomycin (HyClone, USA). The colon adenocarcinoma cell line MC38 was purchased from the National Institutes of Health (NIH) and cultured in Dulbecco's modified eagle medium (DMEM) (Gibco, USA) with 10% FBS (PAN, German) and 100 U/mL penicillin/streptomycin (HyClone, USA). 293T packaging cells (obtained from ATCC) were cultured in DMEM (Gibco, USA) with 10% FBS (PAN, German). All cells were routinely maintained at 37°C in a 5% CO2 atmosphere incubator and tested bimonthly for the absence of mycoplasma.
Vector construct and virus production
CAR structures and plasmid constructs have been described.16 CAR plasmids, along with a pCL-Eco helper plasmid (Addgene #12371), were transfected into 293 T cells using HighGene transfection reagent (ABclonal) to generate retrovirus. Forty-eight hours after transfection, supernatants were collected and filtered using a 0.45 µm filter and stored at −80°C.
CAR-T cell generation and transduction
For EpCAM CAR-T cells, T cells were harvested from the lymph nodes of 6–8 weeks old female C57BL/6 J and BALB/c mice. T cells were cultured in RPMI-1640 supplemented with 10% FBS (PAN), 1% penicillin/streptomycin (HyClone), 2 mM L-glutamine (Gibco), 1 mM sodium pyruvate (Gibco), 0.1 mM non-essential amino acids (Gibco), 0.05 mM β-mercaptoethanol (Gibco), and 50 U/mL mouse IL-2 (Novoprotein). Anti-mouse CD3 antibody (BioLegend, clone 17A2) and anti-mouse CD28 antibody (BioLegend, clone E18) (2 µg/mL) were precoated on 6-well plates overnight at 4°C to stimulate T cells. Forty-eight hours later, T cells were transduced with retrovirus and polybrene (6 µg/mL, Sigma) at 2500 rpm and 32°C for 90 min and cultured for an additional 48 hours with IL-2.
Murine tumor models
For in vivo efficacy experiments, 6–8 weeks old female C57BL/6 J and BALB/c mice were purchased from HFKbio (Beijing). All animal experiments were performed according to protocols approved by the Experimental Animal Ethics Committee of West China Hospital (no.20220217010). 4T1 cells (5×105) were inoculated in situ into the fourth inguinal mammary fat pads of BALB/c mice. When the tumor volume reached 80 mm3 (day 7), tumor-bearing mice were randomly grouped (n=8–10). To establish the MC38 tumor model, 1×106 tumor cells were injected into the right flank of C57BL/6J Mice, and when the tumor volume reached 100 mm3 (day 7), tumor-bearing mice were randomly grouped (n=8–10). Then, 2.5 mg/kg CTX was intraperitoneally injected on day 0 for lymphodepletion. On days 1 and 3, 1×106 EpCAM CAR T cells or control T cells were intravenously injected into tumor-bearing mice. On days 0, 2, 4, 6 and 8, PBS or the S1PR3 antagonist TY-52156 (MCE)/CAY10444 (MCE) at 10 mg/kg was administered orally four times. Tumor growth was measured by a caliper every 2 days. Tumor volumes were calculated as length×width2/2. Lack of survival was defined as death or tumor size >2000 mm3. For CAR T-cell infiltration analysis, a 4T1 tumor model was established. EpCAM CAR T cells (1×106) or control T cells were transferred after CTX lymphodepletion, and the S1PR3 antagonist TY-52156 (MCE) at 10 mg/kg was administered orally every other day four times. Tumor samples and blood were collected on day 7 post CAR T-cell transfer.
BMDM culture
Bone marrow cells were isolated from C57BL/6 J or BALB/c mouse femurs. Red blood cells were removed with lysis buffer (NH4Cl 8.29 g 0.15 M, KHCO3 1.0 g 10 mM, EDTA 37.3 mg 0.1 mm, ddH2O added to 1 L, adjusted to pH 7.2–7.4, sterilized with a 0.2 µm filter). Total bone marrow cells were suspended in complete RPMI-1640 supplemented with 10% FBS (PAN) and 1% penicillin/streptomycin (HyClone) with 10 ng/mL macrophage colony stimulating factor (M-CSF)(Novoprotein) and seeded on 6-well or 12-well plates. On day 3, non-adherent cells were discarded, and adherent cells were further cultured for three more days with fresh medium supplemented with 10 ng mL/1 murine M-CSF. CD11b+F4/80+ adherent cells were considered to be mature BMDMs (M0).
For M1 polarization, M0 BMDMs were stimulated with 20 ng/mL murine IFN-γ and 100 ng/mL LPS for 48 hours. For M2 polarization, M0 BMDMs were stimulated with 20 ng/mL IL-4 and 20 ng/mL IL-13 for 48 hours. For TY-52156 conditioning, 10 µM TY-52156 was added from the start of the cultures. Macrophages were harvested after 48 hours and analyzed by flow cytometry or quantitative reverse-transcriptase PCR for the expression of M2 or M1 activation. Cell culture supernatants were collected for analysis of cytokine release.
ELISA
To test the level of S1P present in the culture supernatant, CAR-T cells or untransduced T cells (Control T) were cocultured with 1×104 4T1 cells at an E:T ratio of 4:1 for 24 hours. TY-52156 (10 µM) was added to the medium. The release of S1P in cell culture supernatants was determined by using a Mouse S1P ELISA Kit (ZCIBIO Technology, Shanghai) according to the manufacturer’s instructions.
To assess the function of the S1PR3 antagonist on CAR-T cells on antigen-specific stimulation, CAR-T cells or untransduced T cells (Control T) were cocultured with 1×104 EpCAM-positive tumor cells at increasing E:T ratios (1:1, 2:1, 4:1, 8:1) for 24 hours. The medium of the coculture of effector and target cells was added with or without TY-52156 (10 µM)/CAY10444 (10 µM). The release of IFN-γ, IL-2, and TNF-α in cell culture supernatants was determined by using ELISA MAX Deluxe sets (BioLegend) according to the manufacturer’s instructions. For animal studies, mouse serum was collected from the ophthalmic vein on day 7 after T-cell infusion. Then, 1:10 diluted serum was tested by mouse IFN-γ and IL-2 ELISA MAX Deluxe sets (BioLegend) according to the manufacturer’s protocol.
BMDMs were differentiated as described above. Forty-eight hours after polarization, cell culture supernatants were collected. The release of IL-6, IL-1β, and TNF-α was determined by using ELISA MAX Deluxe sets (BioLegend) according to the manufacturer’s instructions.
Flow cytometry
To evaluate S1PR3 expression on tumor cells, mouse T cells and EpCAM CAR-T cells, cells were stained with mouse anti-S1PR3 rabbit polyclonal antibody (species: mouse) (HUA BIO) for 30 min at room temperature. Then, the secondary antibody goat anti-rabbit IgG-APC (HUA BIO) was used for 40 min at room temperature.
To evaluate CAR expression, transduced T cells were incubated for 30 mins with recombinant EpCAM protein tagged with the human Fc domain (Novoprotein), and then the EpCAM protein was detected using PE-Cy7-conjugated mouse anti-human IgG Fc antibody (BioLegend, clone M1310G05) for another 30 mins.
To analyze the in vitro activation and phenotype of CAR-T cells, EpCAM CAR-T cells were cocultured with 4T1 cells in the presence or absence of TY-52156 (10 µM) at a 1:1 E:T ratio for 48 hours. T-cell activation was assessed by APC CD25 (BioLegend, clone 3C7) and PE-Cy7 CD69 (BioLegend, clone H1.2F3), and the phenotype was assessed by cell-surface staining of PerCP/Cy5.5 CD44 (BioLegend, clone IM7), APC/Cy7 CD62L (BioLegend, clone MEL-14), APC PD-1 (BioLegend, clone 29F.1A12), APC TIM-3 (BioLegend, clone RMT3-23), and APC LAG-3 (BioLegend, clone C9B7 W) followed by flow cytometric analysis.
In the ex vivo analysis of immune cells derived from the spleens, tumors, blood, and bone morrows. Cells were stained with live/dead dye (Invitrogen) to discriminate dead cells and then stained with cell surface antibodies as described bellow: PerCP CD45 (BioLegend, clone 18009F), BV421 CD3 (BioLegend, clone 17A2), PE-Cy7 CD4 (BioLegend, clone GK1.5), FITC CD8 (BioLegend, clone 53–5.8), APC CD11b (BioLegend, clone M1/70), PE F4/80 (BioLegend, clone BM8), FITC I-A/I-E (BioLegend, clone M5/114.15.2), PE Gr-1 (BioLegend, clone RB6-8C5), FITC CD11c (BioLegend, clone N418) and APC CD25 (BioLegend, clone 3C7). For intracellular cytokine staining of IFN-γ and granzyme B, Golgi-blocking monensin and brefeldin A (eBioscience) were added 4–6 hours before antibody staining. Cells were stained with live/dead dye and surface-staining antibodies as described above. Then, the cells were fixed and permeabilized using the Intracellular Staining Fix/Perm Buffer set (BioLegend) according to the manufacturer’s protocol. Cells were then stained with APC anti-mouse IFN-γ antibody (BioLegend, clone XMG1.2) and PE granzyme B (BioLegend, clone GB11). For intracellular staining of CD206, cells were fixed and permeabilized using the Intracellular Staining Fix/Perm Buffer set (BioLegend) and then stained with BV421 anti-mouse CD206 antibody (BioLegend, clone C068C2). For intracellular staining of Foxp3, cells were fixed and permeabilized using the True-Nuclear Transcription Factor Buffer Set (BioLegend) and then stained with PE Foxp3 (BioLegend, clone MF-14).
Flow cytometry acquisition was performed using an Agilent NovoCyte analyzer (Agilent Technologies, USA), and NovoExpress software was used for data analysis.
Quantitative reverse transcriptase PCR
Total RNA from tumor cells and BMDMs was purified using the TIANGEN RNAsimple Total RNA Kit (TIANGEN, China). Genomic DNA (gDNA) was cleaned, and complementary DNA (cDNA) synthesis was performed using an Evo-MLV RT Kit with gDNA clean for q-PCR II (Accurate Biology, China). Real-time PCR was then run using SsoAdvanced Universal SYBR Green Supermix (BIO-RAD), and the results were analyzed with a BIO-RAD CFX Connect Real-Time PCR System (BIO-RAD). The relative expression levels of mRNAs were normalized to the expression of mouse β-actin and are represented relative to control cells. The following primers were used. β-actin:F, 5’-GTACTCTGTGTGGATCGGTGG-3’;R, 5’-AACGCAGCTCAGT AACAGTCC-3’; CD163:F, 5’-GTGGCTGTGAGCTCACTTCT-3’;R, 5’-TGCCCCACTTGTCATGG ATC-3’;iNOS:F, 5’-CTATGGCCGCTTTGATGTGC-3’;R, 5’-TTGGGATGCTCCA TGGTCAC-3’;Arg-1: F, 5’-CGTGTACATTGGCTTGCGAG-3’;R, 5’-TCCATCAC CTTGCCAATCCC-3’;CD206: F, 5’-AAATGGCTTCCTGGAGAGCC-3’;R, 5’-AC CCTCCGGTACTACAGCAT-3’;IL-1β: F, 5’-GGGCTGCTTCCAAACCTTTG-3’;R, 5’-AAGACACAGGTAGCTGCCAC-3’;IL-4R: F, 5’- ATGTCACCTGAGAACAGC GG-3’;R, 5’-ATGACTCTGCCAAGGCTGAC-3’;IL-10 :F, 5’- GGGTGAGAAGCT GAAGACCC-3’;R, 5’-TGGCCTTGTAGACACCTTGG-3’;MCP-1: F, 5’CAGCCAG ATGCAGTTAACGC-3’;R, 5’-GCTGCTGGTGATCCTCTTGT-3’;IL-6: F, 5’-TCCA GTTGCCTTCTTGGGAC-3’;R, 5’-GACAGGTCTGTTGGGAGTGG-3’;TGF-β: F, 5’- AGCTGGGCAAGTGGTTACAG-3’;R, 5’-CAAAAGGGA GCCCCATCCTT-3’.
Data processing and differential expression analysis
Sequencing data generated during this study are available at the GEO, and the accession numbers for the RNA sequences reported in this paper are GSE67501 and GSE79691. The DEGs between patients with two different outcomes of progression and remission after nivolumab treatment were screened out using GEO2R. |LogFC|>1 and false discovery rate (FDR)<0.05 were selected for statistical significance. For hub gene validation, the transcriptome profiling count data of multiple cancers were downloaded from TCGA database. OS analysis of the S1PR3 gene in multiple cancers using the Gene Expression Profiling Interactive Analysis (GEPIA) tool. The Kaplan-Meier method and log-rank test were used for survival analyses (p<0.05) of each cancer type.
RNA-seq analysis
EpCAM CAR-T cells were stimulated for 24 hours, 48 hours, and 72 hours with 4T1 cells (E:T=4:1) in RPMI 1640 (Gibco). Then, CD3+ T cells were isolated by CD3 MicroBeads (BioLegend), and the extracted RNA was sent to the Microarray Core Facility of Beijing Genomics Institute (BGI) for RNA-seq analysis. Briefly, the mRNA library was constructed using an Illumina mRNA Library kit and sequenced by BGISEQ-500. RNA-seq data were aligned with HISAT (V.2.0.4). The R-DESeq2 package was used to conduct DEG analysis. The DEGs with |logFC |≥1 and Q value≤0.05 are presented by volcano maps. Genes with fold change(FC)≥2 and Q value≤0.001 were selected to perform enrichment analysis using the Dr Tom network platform of BGI (http://biosys.bgi.com). Protein–protein interaction information was parsed from the STRING database (STRING, V.11). All heatmaps were presented by the log2(TPM+1) normalization method, and GSEA was performed on data sets using the MsigDB database: MsigDB MH (Hallmark) and MsigDB M2 Reactome.
Statistical analysis
All data are presented as the mean±SD. Two-tailed unpaired Student’s t-test was used for two-sample comparisons. Tumor growth data were analyzed with two-way analysis of variance. For the survival studies, a log-rank test was used. All statistical analyses were performed using GraphPad Prism V.8.0 software. *p<0.05; **p<0.01; ***p<0.005 and ****p<0.001 were considered statistically significant.
Highlights
High expression of sphingosine 1-phosphate receptor (S1PR)3 is correlated with poor prognosis and increased T-cell exhaustion.
S1PR3 inhibition improves chimeric antigen receptor (CAR)-modified T cell (CAR-T) activation, reduces exhaustion and increases the central memory phenotype in vitro.
S1PR3 inhibition reprograms the tumor microenvironment to amplify the recruitment of CD8+ T cells and CAR-T cells.
S1PR3 inhibition activates tumor macrophages toward the proinflammatory M1 phenotype.
Targeting S1PR3 could enhance CAR-T cell antitumor efficacy in solid tumors.
Data availability statement
All data relevant to the study are included in the article or uploaded as supplementary information.
Ethics statements
Patient consent for publication
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
GG, WL and PS contributed equally.
Contributors GG designed and performed experiments, analyzed data, and wrote the manuscript. WL and PS performed the RNA sequencing analysis of the TCGA and GEO databases and assisted in the analysis. DQ built the CAR construct. XH, QM and BZ provided suggestions on experimental design and data analysis. YW conceived and provided funding support for the project, edited the manuscript and is responsible for the overall content as guarantor. GG, WL and PS have full access to all the data in the study and take responsibility for the accuracy of the data analysis.
Funding This work was supported by the National Natural Science Foundation of China (no. 81872489 to YW) and Sichuan Science and Techonology Program (2023YFS0003 to YW).
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.