Expression levels of HAVCR2 and PDCD1 are correlated on Treg cells in advanced melanoma
Analysis from the TCGA database demonstrated that the genes HAVCR2 encoding Tim-3 was enriched in differentially expressed genes between PD-1low and PD-1high cohorts (Figure 1A). To evaluate whether HAVCR2 expression was associated with PD-1 expression in melanoma, we performed mRNA expression of PDCD1 encoding PD-1 and HAVCR2 in tumor biopsies and TCGA databases. Correlation results demonstrated that HAVCR2 showed positive interaction with PDCD1 in the TME (Figure 1B-C). Exploring the clinical significance of PDCD1 and HAVCR2 in melanoma patients, we analyzed the gene expression levels in primary and metastasis melanoma. Expectedly, both PDCD1 and HAVCR2 expression were significantly higher in advanced melanoma compared to primary melanoma in TCGA datasets (Figure 1D-E). Otherwise, we also found that tumor infiltrated Treg cells were increased in metastasis melanoma (Figure 1F). To investigate the dynamic changes of PDCD1 and HAVCR2 on Treg cells in the tumor progression, we analyzed the gene expression on tumor infiltrated Treg cells in TCGA datasets. Interestingly, the checkpoint molecules PDCD1and HAVCR2 were upregulated in advanced melanoma patients (Figure 1H-I). Further corroborating these observations, a positive correlation > 0.4 between Tim-3 and PD-1 expression on human Treg cells was discovered (Figure 1G). Overall, these data suggested that Tim-3 is engaged by PD-1 on Treg cells toward the progression of melanoma.
Tim-3 is upregulated Treg cells upon anti-PD-1 treatment
Although PD-1-blocking antibodies have presented excellent effects in melanoma therapy, only a few patients experience ideal therapeutic effects [4, 5]. To elucidate the role of immune cells after anti-PD-1 therapy, we assessed the proportion changes of Treg cells in the PBMCs. Results showed a dramatic rise in Treg cell expression, which reflecting immune regulation changes after checkpoint blockade therapy (Figure 2A-B). To further identify functional differences of Treg cells, we then investigate the expression of Tim-3 on CD4+CD25+Foxp3+ Treg cells from PBMCs or purified Treg cells upon anti-PD-1 antibody. The dynamic expression of Tim-3 on Treg cells was checked for confirming the immune suppression after incubated with an anti-PD-1 antibody in the standard medium for 48 h. As shown in Figure 2C, the percentage of Tim-3+ cells were significantly upregulated in Treg cells after PD-1 blockade (p < 0.01). We also found that immune-suppressive cytokines production IL-10 and TGF-β were highly increased in Treg cells after anti-PD-1 treatment (Figure 2 D-E; p <0.05). Besides, purified Treg cells from PBMCs were further sorted to understand the characteristics of Tim-3 expression on Treg cells completely. Similarly, flow cytometric data demonstrated that blockade PD-1 pathway markedly increased the level of Tim-3 expression on purified Treg cells, suggesting an upregulated immune suppression (Figure 2F, p < 0.001). Further corroborating these observations, the cytokines production IL-10 on purified Treg cells were significantly increased after anti-PD-1 treatment (Figure 2G, p < 0.01). Moreover, melanoma samples from patients were obtained and analyzed to investigate the levels of Tim-3 and immunosuppressive factors. In melanoma tissues, we found a significant upregulation of Tim-3 on Treg cells after anti-PD-1 therapy in vitro (Figure 2H, p < 0.05), which was similar to PBMCs. Thus, these findings suggest that blocking the PD-1 pathway significantly upregulates Tim-3 expression on Treg cells. Consistent with our previous study, we also found that Tim-3 downregulated with decreasing the proportion of TGF-β secretion after PD-1 pathway blockade both in PBMCs and melanoma tissues, reflecting the excellent efficacy of anti-PD-1 therapy (Figure S1A-D). These results support the hypothesis that Tim-3 expression can be a potential biomarker for predicting tumor-progression and resistance to anti-PD-1 immunotherapy in melanoma patients.
For verification, that Treg cells were in a functional state, the proliferation of Treg cells had been analyzed after different treatments. It was hoped that Treg cells in the anti-PD-1 group displayed a lower proliferation than did those in the control group (Figure S2A-B), which suggested a suppressive state of Treg cells after anti-PD-1 therapy. Brought jointly, these results confirmed that Tim-3 expression represents an adaptive response for keeping the suppressive status of Treg cells in response to PD-1 blockade.
To confirm that the upregulation of the Tim-3 level after anti-PD-1 immunotherapy, we analyzed the immunohistochemistry (IHC) staining using samples from anti-PD-1 nonresponding melanoma patients. As expected, the checkpoint molecules Tim-3 was increased upon PD-1 immunotherapy in advanced melanoma patients (Figure 3A-B). These results were corroborated with the aforementioned data. To identify the role of Tim-3 expression on tumor growth, we established an immunocompetent skin melanoma model using B16-F10 cells. Our data identified that the PD-1 pathway blockade upregulated the proportion of Tim-3 expression on melanoma-associated Treg cells (Figure 3C). Moreover, we observed a slight rise in the secretion of IL-10 and TGF-β with anti-PD-1 therapy in vivo compared to control melanoma mice, implying an enhanced immune suppression (Figure 3D-E). These results were similar to the previous in vitro data. Taken together, these results confirmed that Tim-3 expression represents an adaptive response to maintain the suppressive status of Treg cells in response to PD-1 blockade.
To investigate the molecular mechanism of Tim-3 upregulation upon an anti-PD-1 treatment, we evaluated the candidate differentially expressed genes between PD-1low and PD-1high cohorts using the online website of DAVID. As shown in Figure 3F, the top 30 related functional and signaling pathways were identified. These genes were enriched in pathways related to cancers, PI3K-Akt signaling, cytokine-cytokine receptor interaction, Jak-STAT signaling pathway, and so on (Figure 3F). The STAT3 pathway has been identified that presented an essential role in the inflammatory response of adaptive immune cells. To determine the clinical significance of the STAT3 gene, we performed the clinical correlative outcomes in TCGA melanoma datasets (Figure 3G). It was not unexpected for STAT3 to be strongly associated with OS in the melanoma patients (Figure 3G). Above all, these data proposed that STAT3 could mediate the tumor immunity upon PD-1 blockade.
Downregulation of STAT3 decreases Tim-3 expression on Treg cells
The STAT3 pathway performs the core part of the inflammatory response of adaptive immune cells. And STAT3 also acts both as a co-transcription pathway for Foxp3 and a mediator for IL-10 in Treg cells [19]. To further address the molecular mechanisms triggered in Treg cells upon Tim-3 upregulation, we evaluated a crucial mediator of carcinogenesis STAT3 through immunosuppression in melanoma. To identified the relationship between tumor infiltrated levels of Treg cells and STAT3, the melanoma cohort of the TCGA dataset was filtered and analyzed by Spearman correlation analysis. As shown in Figure 4 A, STAT3 showed a high correlation with Treg cells in melanoma datasets. Moreover, PDCD1 and HAVCR2 exhibited a positive correlation with STAT3, but the association in PDCD1 is weak (rpearson= 0.195; 0.381, respectively) (Figure 4 B-C). These findings support that STAT3 could act as a crucial mediator of Treg cells associated checkpoint receptor expression in TME. Besides, we also found that the expression of the level of p-STAT3 was increased in the anti-PD-1 group with increased Tim-3 expression both in vitro or in vivo (Figure 4D-E). According to the above results, we hypothesized that Tim-3 was upregulated through the STAT3 pathway in Treg cells after the PD-1 blockade.
A small-molecule inhibitor of p-STAT3 (Stattic) was used to confirm whether Tim-3 was regulated after anti-PD-1 therapy. Cells were incubated with the inhibitor for 48 h after anti-CD3/28 stimulation. Flow analysis indicated that Stattic reduced the level of Tim-3 compared with the vehicle controls after TCR stimulation (Figure 4F). Moreover, we also performed a flow cytometric analysis of Tim-3 on Treg cells upon depletion of STAT3 in melanoma tissues (Figure 4G).Similar to the results, we also found a slight decrease in the Tim-3 expression on melanoma infiltrated Treg cells (Figure 4G), suggesting that inhibited STAT3 could abrogate the Tim-3 expression on Treg cells. However, we also found that PD-1 expression upregulated upon exposure to Stattic (Figure 4H). To determine the impact of STAT3 downregulation on Treg cells function, we used Treg cells to analyze the production of TGF-β by stimulating with brefeldin A in vitro (Figure 4I). The percentage of cytokines TGF-β in Treg cells was found to decrease after STAT3 depletion (Figure 4I, p <0.0001). Melanoma infiltrated Treg cells exhibited similar in vitro Tim-3 increasing trend in STAT3 pathways blockade. These results strongly suggested that STAT-3 downregulation damped the level of Tim-3 expression in Treg cells, leading a decreased immunosuppression. Furthermore, to confirm the therapeutic efficacy of combining STAT3 downregulation with anti-PD-1, we treated Treg cells with Stattic after anti-PD-1 treatment for 48 hours in vitro (Figure 4J-K). To our expectations, the STAT3 downregulation with anti-PD-1 therapy-induced to decreased expression of Tim-3 on Treg cells (Figure 4J). The immunosuppressive cytokines of TGF-β were also analyzed in Treg cells (Figure 4K). A slight decrease of TGF-β production was found in Treg cells with PD-1 and STAT3 pathways blockade (Figure 4K). Altogether, these findings confirmed that targeting STAT3 could abrogate the immunosuppression of Treg cells, providing a potential target in melanoma treatment.
STAT3 inhibitor enhances the efficacy of anti-PD-1 in melanoma
For the identification part of Tim-3 expression on tumor expansion, we established an immunocompetent that had been found for the skin melanoma model using B16-F10 cells. The generated melanoma mouse models were obtained with four different treatments: IgG for control, a PD-1 antibody, Stattic, and anti-PD-1 combination with Stattic (Figure 5A). First, we found that mice bodyweight in the anti-PD-1 treatment group was increased during the tumor progression, while STAT3 downregulated did not alter mice’s weight (Figure 5B). Expectedly, in the PD-1 blockade group, anti-PD-1 treatment did not alter the tumor growth in the melanoma model (Figure 5C). Interestingly, in the combination group, we also found a marked reduction in tumor growth in the B16-F10 melanoma mouse model (Figure 5D, p < 0.0001). Under these conditions, STAT3 downregulation significantly increased the anti-PD-1 efficacy in melanoma. In vivo depletion experiments also demonstrated that the STAT3 pathway blockaded were essential in inhibiting tumor progression, suggesting a potential target therapy.
To confirm the efficacy of combining STAT3 blockade with anti-PD-1 in the C57BL/6 animal model, we performed H&E and TUNEL staining of tumors. Tumor metastasis and apoptosis were evaluated by H&E and TUNEL staining experiments. A significant increase for tumor cell necrosis on melanoma was observed in the STAT3 inhibitor group with anti-PD-1 therapy, indicating excellent therapy efficacy (Figure 5E). Moreover, H&E staining showed no evident metastasis of spleens in the Stattic and anti-PD-1 co-treatment group (Figure 5E). TUNEL staining results confirmed that the frequency of apoptosis cells was increased after STAT3 depletion with blockade (Figure 5E). Therefore, these data suggest that Stattic with PD-1 blockade inhibits tumor growth and metastasis, providing potential therapy for melanoma treatment.
To determine the impact of STAT3 downregulation on anti-melanoma immune response, we analyzed the cell phenotype and cytokine expression pattern in spleen tissue using flow cytometry (Figure 5F). We found a gradient of low down to elevated degrees of immune cells in the spleen (Figure 5F). As expected, the flow cytometric study elaborates that CD4+ and CD8+ T cells had been exceedingly concentrated within the cell clusters (Figure 5F). Next, we explored CD8+ T cell-driven cytokines and multi-inhibitory receptors for disclosing the fundamental downstream mechanisms. Analysis of CD8 subsets, assessed by the manifestation of PD-1, Tim-3, IL-10, TGF-β, granzyme B, and the levels of p-STAT3, indicated that Tim-3 and TGF-β were differentially expressed in these subsets with a distinct transcriptional signature compared to that of other subsets (Figure 5G). These results suggest a potential mechanism of Tim-3 regulation in anti-melanoma immunity.
Blockade STAT3 pathway promotes anti-melanoma immune response
To examine the influence of STAT3 downregulation upon the anti-tumor immune response, we analyzed the tumor infiltrated lymphocytes that had been studied in melanoma mice. Interestingly, STAT3 depletion and anti-PD-1 therapy decreased the percentages of CD8+ and CD4+T cells, leading to an increased CD8/CD4 ratio in STAT3 downregulation and PD-1 blockade combinational treatment in melanoma according for monitoring mice or PD-1 blockade group (Figure 6A-C). Similar results were observed in STAT3 downregulation when compared to the control group (Figure 6B-C). Moreover, in vivo experiments confirmed a marked reduction in the percentages of Treg cells upon to STAT3 inhibitor with anti-PD-1 antibody (Figure 6D). The findings implied that Treg cells were essential in the inhibition of tumor growth (Figure 6D, p < 0.05). To understand how the STAT3 pathway regulates the tumor infiltrated Treg cells, we evaluated the expression of Tim-3 and immunosuppressive cytokines. Our data identified that the PD-1 pathway blockade upregulated the proportion of Tim-3 expression on melanoma-associated Treg cells (Figure 6E). While combining STAT3 downregulation and anti-PD-1 decreased the manifestation of Tim-3 on Treg cells during the anti-tumor response (Figure 6E). Moreover, we observed a slight rise in the secretion of TGF-β and IL-10 in the anti-PD-1 group compared to the IgG2a group (Figure 6F-G), which was similar to in vitro data. Furthermore, flow cytometric analysis demonstrated that STAT3 inhibitor with PD-1 treatment weakened the secretion of TGF-β and IL-10 on Treg cells compared to anti-PD-1 treatment melanoma mice, implying an enhanced anti-tumor immunity (Figure 6 F-G). These data suggested combination therapy with anti-PD-1 and STAT3 pathway inhibitor could promote the anti-tumor immunity and suppressed the tumor progression. Thus, the above data suggested that Tim-3 expression on Treg cells in the TME is STAT3-dependent, providing further support for STAT3 as a target and enhancing the immunotherapy for patients suffering from melanoma.
To confirm the role of combination STAT3 downregulation and anti-PD-1 treatment on CD8+ T cells, we analyzed the cell phenotype and cytokine expression pattern of CD8+ T cells in melanoma (Figure S3A-C). Interestingly, there was a reverse trend in CD8 T cells compared with Treg cells after different treatments (Figure S3A-C). STAT3 downregulation with and without anti-PD-1 therapy was related to an increased expression of Tim-3 on CD8+ T cells as an activation towards CD8+T cells (Figure S3A). Moreover, the STAT3 blockade induced secretion of IL-10 and TGF-β in CD8+ T cells compared that in IgG-treated mice (Figure S3B-C). Our finding indicated that STAT3 presents a reverse event in the crosstalk between Treg cells and CD8+ T cells. Altogether, these results strongly identified that STAT3 downregulation damped Treg function in melanoma, enhanced the anti-tumor immunity. Based on the above data, our work demonstrates that Tim-3 was upregulated through the STAT3 pathway in checkpoint receptors inhibitor therapy.
Furthermore, to investigate the systemic toxicity after different treatments, we conducted the main organs of the H&E staining experiment, including the heart, liver, lungs, kidneys, and brains. There was no apparent toxicity in the main organs after the end of treatment (Figure 6H). Based on these data, our work raises safety therapy by combination using p-STAT3 inhibitor Stattic and anti-PD-1 to regulate tumor immune therapy.