Role of high-level S1PR3 in Prognostic significance and tumor microenvironment of Low-grade Glioma

doi:10.21203/rs.3.rs-1251483/v2

PDF

Research Article

Role of high-level S1PR3 in Prognostic significance and tumor microenvironment of Low-grade Glioma

https://doi.org/10.21203/rs.3.rs-1251483/v2

This work is licensed under a CC BY 4.0 License

Version 2

posted 23 May, 2022

You are reading this latest preprint version

The S1P/S1PR3 axis has been demonstrated to be associated with the development of a variety of human cancers. Still, the potential role of this axis in the biological mechanisms of glioma, particularly TME cell infiltration, remains unknown. We analyzed gene expression data of gliomas and corresponding clinicopathological information to assess overall survival by univariate and multivariate Cox regression. We also examined the signaling pathways and genes associated with S1PR3 overexpression by Gene Set Enrichment Analysis, and analyzed the characteristics of oncogenic pathways (OP), biological pathways, and TME immune cell infiltration in glioma samples with different S1PR3 expressions, and validated the results with PCR, cell proliferation, and migration assays. Cox regression analysis confirmed that S1PR3 overexpression was significantly correlated with overall survival and clinicopathological features. Nomogram for prognostic prediction was successfully constructed using S1PR3 and related genes, and Gene Set Enrichment Analysis revealed the association of S1PR3 overexpression with cell cycle, DNA replication, MAPK, p53 and TGF-β signaling pathway. Meanwhile, S1PR3 was significantly associated with many immune checkpoints and immune cell infiltration. In conclusion, S1PR3 overexpression is a potential prognostic molecular marker for poor prognosis in glioma. In addition, S1PR3 plays an important role in immune cell infiltration in TME. And identification of S1PR3-regulated signaling pathways and immune checkpoints enhances our understanding of TME infiltration characteristics, and contributes to the development of immunotherapeutic strategies.

S1P

S1PR3

glioma

immunity

TME

Sphingosine 1-phosphate (S1P) is a lipid mediator formed by sphingolipid metabolism, which exists in the extracellular environment and is involved in various physiological and pathological processes, with the main function of regulating cell-matrix and cell-cell adhesion and influencing cell proliferation, migration, and differentiation [1]. S1P has five specific cell surface G protein (G12/13, Gi/o and Gq) coupled receptors, called S1PR1-5 [2]. Several reports have demonstrated that the S1P/S1PR3 axis was closely related to proliferation, migration and angiogenesis in a variety of human cancer cells, such as ependymomas, ovarian cancer cells, nasopharyngeal carcinoma and breast cancer [3–6]. S1PR3 can be coupled to Gi/o, Gq, G12/13 families, causing activation of small GTPases such as Rho, Rac and Ras7. In addition, S1PR3 is widely expressed in humans, with high expression in the central nervous system (CNS), immune system, and cardiovascular system [7].

Low grade gliomas (LGG), including WHO grade II and III, are common primary tumors in the CNS, accounting for approximately 20-29% of primary CNS tumors [8, 9]. LGG has the potential to evolve into glioblastoma (GBM), but LGG is less aggressive than GBM, with a median survival time of more than 7 years, compared to GBM, which has a 5-year survival rate of 4-5% [9, 10]. The main treatment method for LGG remains surgical resection, but surgery does not completely stop the recurrence of LGG due to the presence of glioma stem cells and immune cell infiltration of the tumor microenvironment (TME) [11, 12]. Moreover, glioma recurrence caused by tumor remnants has been reported to occur shortly after the operation [11]. Although a combination of surgical resection, temozolomide (TMZ) chemotherapy and radiotherapy is currently the main treatment strategy, strategies tailored to the innate immune and stem cell properties of the glioma may ultimately achieve a survival benefit. Notably, as a current biological research hotspot with great potential, S1PR3 is attracting attention to discover its novel role in cancer, and research on S1PR3 provides new targets for improving cancer therapy [13–15]. Previous studies have shown that S1PR3 expression may affect TME formation and tumor differentiation in cancer cells [16]. In fact, gliomas are more heterogeneous than other types of cancer. The immune signature of the tumor and TME immune cell infiltration may influence novel immunotherapy and developmental plasticity of glioma cells.

The discovery of lymphatic vessels in the CNS has provided support for immunotherapy to penetrate the blood-brain barrier, making immunotherapy a strategy with great potential for the treatment of CNS tumors [17]. Recently, several studies have shown cancer treatment strategies using S1PR3 as a therapeutic target and their associated mechanisms. One study pointed out that significant upregulation of S1PR3 was found in the GBM TME flow region [16]. The high expression of S1PR3 inhibited YAP phosphorylation. It promoted translocation of YAP nuclear, thereby promoting the YAP-c-MYC complex formation and increasing the transcription of PGAM1, an important glycolytic enzyme, which affects energy metabolism in cancer cells [13]. S1PR3 was associated with the TGF-β/SMAD3 signaling pathway and together promoted the progression of human lung adenocarcinoma [18]. Cancer formation involves the activation of proto-oncogenes and the inactivation of oncogenes. As a promising immune target, the role of S1PR3 in glioma has not been comprehensively studied by scholars, and our big picture understanding of the impact of S1PR3 expression on glioma progression and its mechanisms is limited. Thus, a comprehensive analysis of S1PR3 expression in LGG is strongly needed. Further study of cancer samples grouped according to S1PR3 expression will provide new strategies for the treatment of LGG.

In our study, the genomic information of 518 LGG samples from the TCGA database was used to assess S1PR3 expression comprehensively and to determine the prognostic and immunological characteristics of LGG and other cancer cells with different S1PR3 expression. Based on S1PR3 expression in LGG, we identified two distinct immune phenotypes, the immune desert dedifferentiation phenotype and the immune activation differentiation phenotype. We also found that S1PR3 expression was associated with DNA damage and some classical OP.

Cell culture

We used human astrocytes HA-1800, M2 macrophages, oligodendroglioma Hs683 cell line, and U251 glioma cell line obtained from the American Typical Culture Collection (ATCC, Manassas, VA, USA) for in vitro experiments. All cells were maintained in DMEM supplemented with 10% FCS, 2 mM non-essential amino acids and 2 mM glutamine at 37°C, 5% CO2 and 95% humidity. In S1P (Sigma-Aldrich, Germany) and S1PR3 inhibitor (CAY10444, CAYMAN Chemicals, Michigan, USA) stimulation experiments, glioma cells were cultured in DMEM containing 0.05% FCS.

Quantitative real‐time PCR (qRT‐PCR)

We extracted total RNA from cultured normal and tumor cells using Trizol reagent (Life Technologies Corporation). We used a High Capacity cDNA Reverse Transcription Kit (Fermentas) to synthesize the first cDNA strand. Expression was analyzed by the following on-demand gene expression assays from Applied Biosystems: S1PR3 (Hs01019574_m1); and eukaryotic 18S rRNA endogenous control (4310893E). The quantitative real-time PCR was conducted in a 7900 HT Fast Real-Time PCR system. Finally, every mRNA level was normalized to 18S rRNA and using the relative quantification (2-ΔΔCt) method to calculate the fold change.

Cell viability analysis

Glioma cells were inoculated into 96-well plates at 15,000 cells per well. After incubation for 24h, the medium was removed and the cells were incubated for 48h using fresh medium containing CAY10444, S1P or M2 cells. When the incubation time was reached, the medium in the multi-well plate was removed and replaced with fresh medium containing 10% resazurine. The plate was placed back into the incubator until the medium changed from blue to light pink.

Scratch wound healing assay

Glioma cells are inoculated into 12-well plates with a total of 400,000 cells per well. When the cells reached confluence, use a 200 ul pipette tip to set a defined wound scratch into the middle of the cell layer. Wounds were imaged using the PALM Robo software of the AxioVision HXP120C microscope (Carl Zeiss Microscopy, Jena, Germany). The exact location of the image was saved to analyze the same area after 24h. Treated cells with CAY10444 and/or S1P, M2 cells for 24 h and then analyzed for wound width (software AxioVision SE64 Rel. 4.9, Carl Zeiss Microscopy).

Data Acquisition and Bioinformatic Analysis

All the data on gliomas collected in our study were derived from the Cancer Genome Atlas dataset (TCGA, https://portal.gdc.cancer.gov/). The clinical data and gene expression profiles of 518 low-grade glioma cases were identified for further analysis. We compared normalized RNA-seq data with S1PR3 gene expression data using the R programming language. The genes associated with S1PR3 were identified and their correlation with S1PR3 expression was analyzed by Pearson analysis. The differential expression of S1PR3 in various tissues and tumors was also determined by integrating datasets from Genotype Tissue Expression (GTEx) and TCGA. All of our data were analyzed with the R Bioconductor packages and R (version 3.6.1).

Gene set variation analysis and Gene Ontology annotation

The limma package and DESeq2 package in R were used to differentially analyze each gene in the S1PR3 low and high expression group, and the intersection of the results of the two packages was used as differentially expressed genes (DEGs). The Gene Set Variation Analysis (GSVA) analysis and the GSVA package in R were used to investigate the variation in biological processes (BP) between different S1PR3 expression groups [19]. All biological features have been defined from the Kyoto Encyclopedia of Genes and Genomes (KEGG) gene set, with relevant data from the MSigDB database v7.1 [20]. Gene ontology (GO) annotation of S1PR3-related genes was conducted in the clusterProfiler package in R with a cutoff value of FDR < 0.05.

Assessment of independent prognostic factors

We performed a nomogram‐based model to visualize the association between survival rates and individual predictors using the RMS package in R. We performed univariate and multivariate cox regression analysis for OS analysis, as well as progression-free survival (PFS) analysis to verify the feasibility of our model as an independent prognostic factor. We further used the "survival ROC" package in R to assess prognostic ability by area under the curve (AUC) and receiver operating characteristic curve (ROC) analysis.

Estimation of TME cell infiltration and assessment of the correlation of S1PR3 gene features with other relevant BP

In order to quantify the relative abundance of every cell infiltrate in the glioma TME, we used the single sample gene set enrichment analysis (ssGSEA) algorithm. The genomes used to mark each TME-infiltrating immune cell type were derived from Charoentong's research [21]. We classified immune cells into pro-tumor immune cells (TAM, imDC, pDC, CD56dimNK, Th2, MDSC, Neutrophil, and Treg) and anti-tumor immune cells (TcmCD4, TcmCD8, TemCD4, TemCD8, NKT, ActCD4, ActCD8, ActDC, CD56briNK, Th1, Th17, and NK) based on their cellular functions. Furthermore, we used 29 immune signatures to detect characteristics between different S1PR3 expression groups [22]. Analysis of the abundance of different leukocyte subsets with glioma gene expression profiles by the deconvolution method CIBERSORT [23]. To study the connection between S1PR3 gene characteristics and some associated biological pathways, we collected gene sets that store genes related to many BP [24, 25] (Table S1). In addition, 10 groups of OP genomes were included in our study in order to better explore the mechanisms of S1PR3 gene signatures in response to different treatments [26] (Table S2).

The landscape of genetic variation of S1PR3 in glioma

To examine the expression level of S1PR3 in glioma cells, we first confirmed S1PR3 expression in normal brain tissue in the GTEx database. S1PR3 was widely expressed in brain tissue (Figure S1A). Next, we confirmed S1PR3 expression in LGG and GBM by qRT-PCR. We observed a significant increase in S1PR3 expression in HS683 and U251 cells compared with HA-1800, and the expression level of S1PR3 was higher in U251 cells than in HS683 cells (Figure 1A). The S1PR3 expression pattern in gliomas was further validated by investigating S1PR3 expression levels in normal, LGG and GBM tissues. The results showed that S1PR3 expression was significantly increased in gliomas. The difference was more pronounced in GBM than in LGG tissues (Figure 1B-C , S1B). Meanwhile, S1PR3 expression in different histological subtypes of LGG also showed differences, and the highest expression level was found in astrocytoma (Figure 1D).

To investigate the correlation between S1PR3 expression and patient survival, we compared OS and PFS (Figure 1E, S1C) in patients with low and high S1PR3 expression. Patients in the low group had a significant survival advantage over those in the high group. To further assess the clinical significance of S1PR3 overexpression in LGG, we compared S1PR3 expression differences between molecular subtypes. Our results showed that S1PR3 expression was significantly associated with grade, IDH mutation status, MGMT promoter status, x1p19q codeletion status, and primary/recurrence type (Figure 1F-J). We further investigated the effect of S1PR3 on OS by univariate and multivariate Cox regression analysis, and S1PR3 expression could be an independent factor for LGG prognosis (Figure 1K-L). We further studied the differences in somatic mutation distribution between the high and low S1PR3 expression groups (Figure 1M-N). The IDH1 mutation was significantly higher in the low expression group than in the high expression group, 86% and 65%, respectively, which is consistent with a better prognosis of LGG with IDH1 mutations [27]. The tumor TP53 mutation load was higher in the high expression group than in the low expression group, 59% and 34%, respectively, which is consistent with a poorer prognosis for glioma patients with high TP53 expression [28].

Establishment of a prognostic prediction nomogram for glioma associated with S1PR3 expression

To further assess the prognostic ability of S1PR3 expression in gliomas, we performed ROC analysis on 518 samples. The 1-year survival AUC associated with S1PR3 expression was 0.650 (Figure 2A), the 3-year survival AUC was 0.684 (Figure 2B), and the 5-year AUC was 0.730 (Figure 2C). The results of the ROC study reaffirmed that S1PR3 expression could be an independent prognostic factor in LGG patients. Based on the results of the ROC study, we built a nomogram integrating S1PR3 expression and clinicopathological variables, which can be used as a quantitative method to provide a prediction of LGG prognostic risk (Figure 2D). To verify the accuracy of the nomogram, we used the nomogram to re-predict 1-, 3-, and 5-year survival rates (Figure 2E-G) and used the C-index to assess predictive abilities. The final resulting C index was 0.87361, indicating that the predictive accuracy of our nomogram in terms of survival predictability was nearly high. Our nomogram could be not only used for a comprehensive analysis of S1PR3 expression and other clinical parameters in LGG patients but also served as a prediction of patient survival.

Annotation of classification functions determined by consensus clustering analysis

To further understand the mechanism by which S1PR3 overexpression affects glioma progression, we investigated the DEGs between the high and low S1PR3 expression groups. Then we performed KEGG and GO analysis on DEGs in S1PR3 high expression group. According to KEGG enrichment analysis of DEGs, S1PR3 overexpression may be associated with recognized oncogenic and immune-related pathways such as neuroactive ligand, ECM, cytokine-cytokine receptor interaction, PI3K-Akt, cAMP, NF-B and others (Figure 3A). Interestingly, further GO enrichment analysis of DEGs revealed that S1PR3 overexpression may be associated with regulating immune cell activation, proliferation and intercellular interactions, such as T cell activation, positive regulation of cell activation, regulation of T cell activation, regulation of cell-cell adhesion, leukocyte cell-cell adhesion, et al. (Figure 3B). The top 15 GO terms for LGG indicated that DEGs were enriched in gene expression, cellular biosynthetic process, regulation of gene expression, cellular aromatic compound metabolic process, nucleic acid metabolic process, immune response, RNA metabolic process, regulation of RNA metabolic process, regulation of transcription DNA-templated, regulation of nucleic acid-templated transcription, regulation of RNA biosynthetic process, transcription DNA-templated, et al. (Figure 3C).

Meanwhile, we performed GSEA analysis on LGG tissues with high S1PR3 expression. We displayed representative BP (Figure 3D-N), including cell cycle, cytokine-cytokine receptor interaction, DNA replication, JAK-STAT signaling pathway, MAPK signaling pathway, mTOR signaling pathway, p53 signaling pathway, TGF-β signaling pathway, et al. S1PR3 expression in LGG was associated with tumor immune and OP. High expression was involved in the malignant progression of gliomas.

Identification of S1PR3-related genes and S1PR3 inhibitors inhibited proliferation and migration of glioma cells

To better understand the mechanisms by which S1PR3 expression affects glioma progression, we examined S1PR3-related genes (Table S3). We selected five of the most positively (Figure 4A-E) and five of the most negatively associated genes (Figure 4F-J) for further study. The five most positively associated genes were IQGAP1, ELF4, ECM2, ANQ6, VIM. The five most negatively associated genes were RHBDL1, KCNIP2, CKMT1B, CKMT1A, RUNDC3A. In addition, we selected the top 20 LGG somatic mutation genes for the Pearson correlation study with S1PR3. The results showed that 12 of them were correlated with S1PR3 expression, including ARID1A, ARID1B, EGFR, HMCN1, IDH1, IDH2, NF1, NIPBL, PIK3CA, PTEN, RYR2, TP53 (Figure S1D-O). We further selected the top 20 differentially expressed genes in LGG, the results indicated that 12 of them were associated with S1PR3 expression (Figure S2).

The results of these genetic correlations suggested that the positively associated genes might have a synergistic effect with S1PR3 high expression, promoting the malignant progression of LGG, and the negatively associated genes have an antagonistic effect, providing a direction for further research.

We next validated the above study with in vitro experiments. Cell viability decreased significantly after treating HS683 cells with 10 and 20 µM CAY10444 for 48 h (Figure 4K). Interestingly, 20 µM CAY10444 showed a stronger inhibitory effect on HS683 cells than TMZ. We got similar results in the study of U251 cells (Figure 4L). When another specific inhibitor of S1PR3, TY-52156, was used to block the S1PR3 pathway in HS683 and U251 glioma cells, we observed similar results to CAY10444 (Figure 4 M-N). When U251 and HS683 cells were stimulated with S1P, cell viability appeared mildly elevated to 113.35% and 114.35%, respectively. When U251 and HS683 cells were co-treated with 10 µM and 20 µM of CAY10444 and S1P, respectively, we found a significant decrease in cell viability compared to the S1P alone application group (Figure 4 O-P). Given these findings, it could be deduced that CAY10444 could affect S1P mediated proliferation on glioma cells. We further investigated the migration of HS683 cells using a scratch wound-healing assay. The results showed that CAY10444 could affect the role of S1P in promoting HS683 cell migration (Figure 4Q).

Characterization of TME immune cell infiltration in gliomas with different S1PR3 expression

To investigate the effect of S1PR3 expression on LGG TME immune cell infiltration, we first investigated the enrichment of 23 immune cells in high and low S1PR3 expression groups. We found that the high expression group showed an enrichment advantage for approximately all immune cells compared to the low expression group (Figure 5A, Table S4). Furthermore, in the analysis of immune signatures, the high expression group was upregulated in almost all immune signatures compared to the low group (Figure S3A, Table S5). Interestingly, the high expression group was significantly upregulated in both anti-tumor and pro-tumor immune cells and signatures, indicating that the effect of S1PR3 expression on tumor immunity is complex, which needs further study. Previous studies have also demonstrated that S1PR3 promotes immune cell recruitment by driving leukocytes to roll on endothelial cells[29]. To investigate the mechanisms involved, we compared the correlation between pro- and anti-tumor immune cells in different S1PR3 expression groups. We investigated the correlation between pro- and anti-tumor immune cells in the whole sample, S1PR3 high and low expression groups (Figure 5B-D), respectively. We found that the low expression group had a higher anti-tumor ability, while the high expression group showed more pro-tumor ability.

To better understand the effect of S1PR3 expression on TME, we further analyzed the expression differences of typical immune-related genes in high and low expression groups. The expression of checkpoint inhibitor-associated genes did not show significant differences between the high and low groups (Figure 5E). The expression of stimulator-associated genes was significantly increased in the high group (Figure 5F), while the expression of major histocompatibility complex (MHC) (Figure 5G) was significantly increased in the low group. Based on that, we suggested that different S1PR3 expression involving different DNA damage-related phenotypes may inhibit or promote the anti-tumor ability of immune cells. Given this, we examined both the high and low groups in which BP were enriched for DNA damage and showed that Angiogenesis, Antigen processing Machinery, CD8 T effector, EMT, FGFR3-related genes, Immune checkpoint, Mismatched repair and Pan-F-Target were significantly increased in the high group compared to the low group. (Figure 5H). We further analyzed the enrichment of 10 OP in the S1PR3 high and low expression groups and found that nearly all OP were significantly enriched in the high group (Figure 5I). We next performed an OS analysis of immune cells infiltrating TME affected by S1PR3 expression. In the immune cell high enrichment group compared with the low enrichment group, OS was significantly different, including CD4+ T, CD8+ T, Macrophages, Neutrophils, T cell regulatory tregs (Figure S3 B-F). The prognosis was poorer in the high immune cell enrichment group, which was consistent with the prognosis of the high S1PR3 group, which further validated our results. We calculated the ESTIMATEScore, immuneScore and stromalScore for LGG and GBM using the ESTIMATE algorithm in order to further identify the potential mechanisms associated with S1PR3 expression. The results showed that S1PR3 expression was correlated with all three scoring systems in LGG (Figure 5J-L) and GBM (Figure S3G-I), respectively. In addition, we examined the correlation between BP and the correlation between OP in different S1PR3 expression groups, respectively. We investigated the correlation between BP in the whole sample, S1PR3 high and low expression groups (Figure S3J-L), respectively. It was further demonstrated that the high expression group was enriched for DNA damage. Then we investigated the correlation between OP in the whole sample, S1PR3 high and low expression groups (Figure S3M-O), respectively. It was found that although OP was enriched in the high expression group, the correlation between OP was higher in the low expression group than in the high group, which needs further study.

We further validated the effects of CAY10444 and macrophages on glioma cells. After 48h of treatment, M2 cells could reverse the inhibitory proliferative effect of CAY10444 on U251 cells, increasing from 17.69% to 28.14% (Figure 6A). Similar results were observed in HS683 cells, where M2 cells could reverse the inhibitory proliferative effect of CAY10444, increasing from 16.75% to 30.02% (Figure 6B). The scratch wound healing assay showed that CAY10444 significantly reversed the effect of M2 cells to promote HS683 cell migration from 26.40% to 19.97% (Figure 6C). By examining the immune cell infiltration characteristics of TME in the S1PR3 high and low expression groups, we found significant differences in immune cell infiltration, tumor differentiation and clinical features. Regarding the immune characteristics, the high expression group was the immune desert dedifferentiated phenotype showing promotion of tumor immune infiltration and immunosuppression. The low expression group was an immune activating differentiated phenotype that showed anti-tumor immune infiltration and immune activation. Moreover, the high expression group exhibited DNA damage and dedifferentiation compared to the low expression group.

Expression of S1PR3 in pan-cancer and the effect of S1PR3 expression on immune cell infiltration across tumor types

We first analyzed the expression levels of S1PR3 in a pan-cancer cohort and showed that S1PR3 was expressed in a variety of cancer tissues (Figure S4A). In addition, data from the GTEx database confirmed that while S1PR3 expression was significantly higher in multiple tumor tissues than in normal tissues, there were still some cancers that showed opposite changes, suggesting that S1PR3 expression is not regulated in the same way across cancers (Figure 6D). Since the immune infiltration characteristics of each cancer TME are different, different immune cells and immune signatures may be involved. We analyzed the correlation between S1PR3 expression and immune cells (Figure 6E) and immune signatures (Figure S4B) for each cancer type. We performed ssGSEA analysis on samples for each cell population in the pan-cancer cohort. In most cancers, most immune cells are associated with S1PR3 expression, which includes both anti-tumor and pro-tumor immune cells, indicating that the effect of S1PR3 expression on immune function may not be consistent. Checkpoint blockade immunotherapy is currently a hot topic in glioma treatment, and the key clinically validated biomarkers reflecting treatment response include TME cell infiltration, tumor mutation burden (TMB), and Microsatellite instability (MSI) [30]. Radar plots of labeled TMB and MSI showed a significant correlation between S1PR3 expression and TMB and MSI in multiple cancers (Figure 6F-G). We also analyzed the PD-L1 (CD274) and PD-1 (CD8A) expression values in correlation with S1PR3 expression. A significant correlation was discovered between CD274 and S1PR3 expression in a variety of cancers (Figure 6H). Similar results were observed for studies in CD8A (Figure 6I). Throughout the checkpoint blockade immunotherapy marker analysis, inconsistent trends were observed across cancers with positive and negative correlations between markers and S1PR3 expression.

Despite aggressive multimodal therapy, short-term recurrence and malignant progression of LGG currently remain intractable, and examining the signaling pathways that lead to rapid recurrence and malignant prognosis is critical to optimizing current treatment strategies. The effects of S1PR3 inhibitors on the proliferation and migration of GBM cells have been reported [16]. However, results regarding the effects of the S1P/S1PR3 signaling system on GBM cell proliferation and migration were partially inconsistent in their conclusions. In present study, we performed a comprehensive analysis of S1PR3 expression in glioblastoma to investigate the impact of diverse S1PR3 expression patterns on the malignant progression and prognosis of glioma. Furthermore, we constructed a nomogram for predicting survival.

By analyzing the S1PR3 expression values of 518 LGG patients in the TCGA database, we identified two patterns of high and low expression. We verified that low expression of S1PR3 has a significant survival advantage, and we demonstrated the predictive ability of S1PR3 expression by conducting ROC analysis. Through GO, KEGG, and ssGSEA analysis, we have a better understanding of classical cellular pathways with different expression patterns. We also performed immune cell infiltration, BP and OP gene analysis to identify the potential mechanisms by which S1PR3 expression promotes tumorigenesis. Our study finally identified two immune phenotypes associated with S1PR3 expression: the low S1PR3 expression group presented an immune activation differentiation phenotype; the high expression group presented an immune desert dedifferentiation phenotype. The results of this study support the utility of S1PR3 expression as a potential predictor of prognosis in LGG patients, which acts through multiple mechanisms to promote tumor progression and immune evasion. The results of this study support the utility of S1PR3 expression as a potential predictor of prognosis in LGG patients, which acts through multiple mechanisms to promote tumor progression and immune evasion.

Our PCR results confirmed that S1PR3 expression was significantly higher in glioma cells than in normal cells. The analysis of glioma samples from the TCGA database confirmed that S1PR3 expression was considerably higher in glioma tissue than in normal cells adjacent to normal tissue, which supports previous related studies [31]. High S1PR3 expression in LGG patients was significantly associated with higher WHO grade, histological grade and poorer OS. Although S1PR3 expression levels were correlated with many clinical parameters [32], the results of our COX and correlation analyses suggest that S1PR3 expression can not only influence the prognosis of LGG together with other parameters, but also serve as an independent predictor of OS in LGG patients. We confirmed the difference in the distribution of IDH1 mutations and TP53 mutations between the high and low S1PR3 expression groups by somatic mutational analysis, with the high expression group having lower IDH1 mutations and higher TP53 mutations. While lower IDH1 mutations [33] and higher TP53 mutations [34] were associated with poorer survival in glioma patients, this further confirmed the impact of S1PR3 expression on OS in LGG patients. We constructed a risk scoring system that can predict the prognosis of LGG patients and further validated the performance of the system in predicting 1-, 3-, and 5-year survival rates. The results demonstrated the high accuracy of our system. Our GSEA analysis showed that high S1PR3 expression was associated with cell cycle, DNA replication, JAK-STAT signaling pathway, MAPK signaling pathway, mTOR signaling pathway, p53 signaling pathway, TGF-β signaling pathway, PD-L1 expression and PD-1 checkpoint pathway, and RNA degradation signaling pathway. Aberrant activation of the essential signaling pathways JAK-STAT, MAPK, mTOR, p53, and TGF-β showed association with the development of glioma [35–39]. We investigated and validated the relationship between S1PR3 and related genes based on the TCGA dataset and identified the five most positively associated genes as ANO6, ECM2, ELF4, IQGAP and VIM. In comparison, the five most negatively associated genes were CKMT1A, CKMT1B, KCNIP2, RHBDL1 and RUNDC3. ANO6, ECM2, ELF4, IQGAP and VIM have been shown to be positively associated with malignant progression of gliomas [40–44]. We further confirmed with cellular experiments that blocking S1PR3 expression inhibits the proliferation of glioma cells and that inhibitors of S1PR3 reversed the promotion of S1P in glioma cell migration.

We revealed two distinct immune patterns based on S1PR3 expression, which have significantly different TME cell infiltration characteristics. The high expression group showed an immune desert dedifferentiation phenotype that is immunosuppressive, while the low expression group showed an immune activation differentiation phenotype that corresponds to immune and stroma activation. The immune desert phenotype was associated with immune tolerance and lack of initiation and activation of T cells [45]. In contrast to the immuno-desert phenotype, the immune activation phenotype does not show a large enrichment of immune cells. In fact, the immune activation phenotype also showed the presence of large numbers of immune cells, but the difference was that the immune cells were kept in the stroma surrounding the tumor cell nest instead of penetrating the stroma. And the stroma may be confined within the tumor envelope or may penetrate the tumor itself, making it appear that the immune cells are actually inside the tumor [46, 47]. S1PR3 has been shown to play a multifactorial role in immunity by influencing macrophage chemotaxis and killing, dendritic cell maturation, eosinophil and neutrophil recruitment while contributing to immune cell recruitment by driving leukocyte rolling on endothelial cells[29]. EMT has been reported as a classical process of mesenchymal acquisition by epithelial cells, which is associated with tumorigenesis, metastasis, invasion and drug resistance in many cancers [48]. Moreover, EMT can drive the tumorigenic and stemness characteristics of cells and induce stem cell-like properties in tumor cells. In our study on BP, EMT was highly enriched in the S1PR3 high expression group. This feature perhaps have led to alkylator TMZ resistance. Higher levels of TGF-β expression were significantly associated with poor prognosis of glioma, and TGF-β expression promoted the proliferation and migration of glioma cells, as well as induced an aggressive glioma phenotype [39]. Moreover, TGF-β-induced EMT in hepatocellular carcinoma cells was associated with changes in the expression of stem markers [49]. Our study on OP also showed significant TGF-β enrichment in the S1PR3 high expression group, implying that this association may also apply to LGG. Our in vitro experiments also confirmed the reversal of M2 macrophage-induced glioma cell proliferation and migration after blocking the S1PR3 pathway.

Applying S1PR3 expression patterns and the associated immune infiltration patterns to other cancers, it was found that cancers did not show significant concordance with each other. This suggested that S1PR3 expression was heterogeneous among different cancers. However, S1PR3 expression showed significant differences between almost all cancers and normal tissues. In cancer, the correlation between the immune cell infiltration characteristics of TME and S1PR3 expression may lead to uncontrolled immune dysregulation and dedifferentiation. Study of TMB, MSI, CD274 and CD8A suggested that S1PR3 may be an effective target for immunotherapy. The association of S1PR3 expression with multiple tumor distinct phenotypes in the pan-cancer cohort study may reflect TME immune infiltration specificity, differential immune checkpoint expression, and diversity of BP.

Overall, our results suggest that S1PR3 expression may be an important factor affecting the prognosis of glioma patients. Our study provides preliminary evidence for a regulatory mechanism of S1PR3 expression on LGG and other tumor TME immune cell infiltration and BP. For S1PR3 expression alone or in combination with other clinical parameters to predict LGG prognosis, our nomogram revealed satisfactory predictive ability. We also identified signaling pathways and genes associated with S1PR3 expression in LGG. These provide theoretical support for further research. Our study will contribute to the development of prediction tools for glioma prognostic, quantification of S1PR3 expression in individual cancers, and potential future application in predicting cancer recurrence, choice of treatment modality.

Data Availability

Additional Supporting Information may be found online in the supporting information tab for this article.

Conflicts of Interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Funding Statement

This research was funded by the Fund of Tang Du hospital (No.2021YFJH005, No.2021SHRC033).

Acknowledgments

None.

1. Pan Y, Gao F, Zhao S, Han J, Chen F. Role of the SphK- S1P- S1PRs pathway in invasion of the nervous system by SARS- CoV- 2 infection. Clin Exp Pharmacol Physiol. 2021;48(5) May:637–50.

2. Kumar A, Nagar S. Regulation of Immune Cell Migration by Sphingosine-1- Phosphate. Cell Mol Biol. 2018;61:121.

3. Dai L, Liu Y, Xie L, Wu X, Qiu L, Di W. Sphingosine kinase 1/sphingosine-1-phosphate (S1P)/S1P receptor axis is involved in ovarian cancer angiogenesis. Oncotarget. 2017;8:74947–61.

4. Kim E, Kim J, Kim SG, Hwang S, Lee CH, Moon A. Sphingosine 1-phosphate regulates matrix metalloproteinase-9 expression and breast cell invasion through S1P3-Gαq coupling. J Cell Sci. 2011;124 Pt 13:2220–30.

5. Magrassi L, Marziliano N, Inzani F, Cassini P, Chiaranda I, Skrap M, et al. EDG3 and SHC3 on chromosome 9q22 are co-amplified in human ependymomas. Cancer Lett. 2010;290:36–42.

6. Lee HM, Lo K, Wei W, Tsao SW, Tin G, Chung Y, et al. Oncogenic S1P signalling in EBV-associated nasopharyngeal carcinoma activates AKT and promotes cell migration through S1P receptor 3. J Pathol. 2017;242:62–72.

7. Fan X, Zhong D, Li G. Recent advances of the function odatf sphingosine 1 ‐ phosphate ( S1P ) receptor S1P3. J Cell Physiol. 2020;236:1564–78.

8. Wang TJC, Mehta MP. Low-Grade Glioma Radiotherapy Treatment and Trials. Neurosurg Clin NA. 2019;30:111–8.

9. Yu Y, Yang B, Yu J, Zhao G, Chen F. Dequalinium chloride inhibits the growth of human glioma cells in vitro and vivo: a study on molecular mechanism and potential targeted agents. Acta Neurochir (Wien). 2020;162:1683–90.

10. Pan Y, Zhao S, Chen F. The potential value of dequalinium chloride in the treatment of cancer: Focus on malignant glioma. Clin Exp Pharmacol Physiol. 2021;48:445–54.

11. Weller M, Wick W, Aldape K, Brada M, Berger M, Nishikawa R, et al. Glioma. Nat Rev Dis Prim. 2015;16;1 July:15017.

12. Lim M, Xia Y, Bettegowda C, Weller M. Current state of immunotherapy for glioblastoma. Nat Rev Clin Oncol. 2018;15:422–42.

13. Shen Y, Zhao S, Wang S, Pan X, Zhang Y, Xu J, et al. S1P/S1PR3 axis promotes aerobic glycolysis by YAP/c-MYC/PGAM1 axis in osteosarcoma. EBioMedicine. 2019;40:210–23.

14. Toyomoto M, Inoue A, Iida K, Denawa M, Kii I, Marie F, et al. Article S1PR3 – G 12 -biased agonist ALESIA targets cancer metabolism and promotes glucose starvation Article S1PR3 – G 12 -biased agonist ALESIA targets cancer metabolism and promotes glucose starvation. Cell Chem Biol. 2021;:1–13.

15. Physiology C. Intracellular Sphingosine-1-Phosphate Receptor 3 Contributes to Lung Tumor Cell Proliferation. 2021;:539–52.

16. Tate K, Chase Cornelison R, Bhargava S MJ. TMIC-33. INVESTIGATING S1PR3-MEDIATED GLIOBLASTOMA INVASION MECHANISMS USING 3D HYDROGEL - TISSUE CULTURE INSERT MODEL. Neuro Oncol. 2019;21:254–5.

17. Xu S, Tang L, Li X, Fan F, Liu Z. Immunotherapy for glioma : Current management and future application. Cancer Lett. 2020;476 February:1–12.

18. Chen S, Ahn Y, Lonardo F, Lee M. TGF-β/SMAD3 Pathway Stimulates Sphingosine-1 Phosphate Receptor 3 Expression: IMPLICATION OF SPHINGOSINE-1 PHOSPHATE RECEPTOR 3 IN LUNG ADENOCARCINOMA PROGRESSION. J Biol Chem. 2016;3:27343–53.

19. Sun X, Dong B, Yin L, Zhang R, Du W, Liu D, et al. PMTED : a plant microRNA target expression database. BMC Bioinformatics. 2013;3:174.

20. Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA, et al. Gene set enrichment analysis: A knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A. 2005;102:15545–50.

21. Charoentong P, Angelova M, Charoentong P, Finotello F, Angelova M, Mayer C, et al. Pan-cancer Immunogenomic Analyses Reveal Genotype-Immunophenotype Relationships and Predictors of Response to Checkpoint Blockade Graphical. CellReports. 2017;18:248–62.

22. He Y, Jiang Z, Chen C, Wang X. Classification of triple-negative breast cancers based on Immunogenomic profiling. J Exp Clin Cancer Res. 2018;37:1–13.

23. Newman AM, Liu CL, Green MR, Gentles AJ, Feng W, Xu Y, et al. Robust enumeration of cell subsets from tissue expression profiles. Nat Methods. 2015;12:453–7.

24. Mariathasan S, Turley SJ, Nickles D, Castiglioni A, Yuen K, Wang Y, et al. TGFβ attenuates tumour response to PD-L1 blockade by contributing to exclusion of T cells. Nat Publ Gr. 2018;22 554(7693):544–8.

25. Yasin Ş, Gejman RS, Winer AG, Liu M, Allen EM Van, Velasco G De, et al. Tumor immune microenvironment characterization in clear cell renal cell carcinoma identifies prognostic and immunotherapeutically relevant messenger RNA signatures. Genome Biol. 2016;17:1–25.

26. Sanchez-Vega F, Mina M, Armenia. Pathways, Oncogenic Signaling Cancer, The Atlas, Genome. 2019;173:321–37.

27. Karpel G, Trang M, Enyuan TTN, Markus S. Novel IDH1 ‑ Targeted Glioma Therapies. CNS Drugs. 2019;33:1155–66.

28. Zhang Y, Dube C, Jr MG, Cruickshanks N, Wang B, Coughlan M, et al. The p53 Pathway in Glioblastoma. Cancers (Basel). 2019;10:297.

29. Bryan AM, Del Poeta M. Sphingosine‐1‐phosphate receptors and innate immunity. Cell Microbiol. 2018;20:e12836.

30. Razvan C, Robin M, Mark A, Andrew A, Erin M, Jennifer Y, et al. Pan-tumor genomic biomarkers for PD-1 checkpoint blockade–based immunotherapy. Science (80- ). 2018;362:6411.

31. James Ross, Monica Chau, Brandon Miller, Subhas Mukherjee, Changming Zhang, Jun Kong, Emily Kaissi, Austin Newsam, Darius Mahboubi, Jameson Berry, Carol Tucker-Burden, Daniel BratJames Ross, Monica Chau, Brandon Miller, Subhas Mukherjee, Changming Zhang, DB. TMIC-14. HYPOXIA INDUCIBLE OLIG2 MEDIATES GLIOMA STEM CELL MIGRATION. Neuro Oncol. 2018;19 suppl_6:246.

32. Li Q, Li Y, Lei C, Tan Y, Yi G. Sphingosine-1-phosphate receptor 3 signaling. Clin Chim Acta. 2021;519 November 2020:32–9.

33. Arita H, Narita Y, Yoshida A, Hashimoto N, Yoshimine T IK. IDH1/2 mutation detection in gliomas. Brain Tumor Pathol. 2014;32:79–89.

34. Cancer Genome Atlas Research Network, Brat DJ, Verhaak RG, Aldape KD, Yung WK, Salama SR, Cooper LA, Rheinbay E, Miller CR, Vitucci M, Morozova O, Robertson AG, Noushmehr H, Laird PW, Cherniack AD, Akbani R, Huse JT, Ciriello G, Poisson LM, Barnholtz-Sloa ZJ. Comprehensive, Integrative Genomic Analysis of Diffuse Lower-Grade Gliomas. N Engl J Med. 2015;372:2481–98.

35. Ciechomska I, Jayaprakash C, Maleszewska M, Kaminska B. Histone Modifying Enzymes and Chromatin Modifiers in Glioma Pathobiology and Therapy Responses. 2020.

36. Perreault S, Larouche V, Tabori U, Hawkin C, Lippé S, Ellezam B, et al. A phase 2 study of trametinib for patients with pediatric glioma or plexiform neurofibroma with refractory tumor and activation of the MAPK/ERK pathway: TRAM-01. BMC Cancer. 2019;19:1250.

37. Li X, Wu C, Chen N, Gu H, Yen A, Cao L. PI3K / Akt / mTOR signaling pathway and targeted therapy for glioblastoma. Oncotarget. 2016;7:33440–50.

38. Sheng J, He X, Yu W, Chen Y, Long Y, Wang K, et al. p53-targeted lncRNA ST7-AS1 acts as a tumour suppressor by interacting with PTBP1 to suppress the Wnt / β -catenin signalling pathway in glioma. Cancer Lett. 2021;503 November 2020:54–68.

39. Zhang C, Zhang X, Xu R, Huang B, Chen A, Li C, et al. TGF- β 2 initiates autophagy via Smad and non-Smad pathway to promote glioma cells ’ invasion. J Exp Clin Cancer Res. 2017;36:162.

40. Xuan Z, Wang Y. ANO6 promotes cell proliferation and invasion in glioma through regulating the ERK signaling pathway. Onco Targets Ther. 2019;20:6721–31.

41. Bernstock JD, Mooney JH, Ilyas A, Chagoya G, Estevez-Ordonez D, Ibrahim A NI. Molecular and cellular intratumoral heterogeneity in primary glioblastoma: clinical and translational implications. J Neurosurg. 2019;23 August:1–9.

42. Gong R, Li Z, Fu K, Ma C, Wang W, Chen J. Long Noncoding RNA PVT1 Promotes Stemness and Temozolomide Resistance through miR-365 / ELF4 / SOX2 Axis in Glioma. Exp Neurobiol. 2021;30:244–55.

43. Gao C, Liang C, Nie Z, Liu Y, Wang J, Zhang D. Alkannin inhibits growth and invasion of glioma cells C6 through IQGAP / mTOR signal pathway. 2015;8:5287–94.

44. Herrera-oropeza GE, Herrera-oropeza GE, Angulo-rojo C, Hernández-rosales M, Aviña-padilla K, Hernández-rosales M. Glioblastoma multiforme: a multi-omics analysis of driver genes and tumour heterogeneity. Interface Focus. 2021;11:20200072.

45. Kim JM CD. Immune escape to PD-L1/PD-1 blockade: seven steps to success (or failure). Ann Oncol. 2016;27:1492–504.

46. Gajewski TF. The next hurdle in cancer immunotherapy: overcoming the non–T-cell–inflamed tumor microenvironment. In: Seminars in oncology. Elsevier; 2015. p. 663–71.

47. Salmon H, Franciszkiewicz K, Damotte D, Dieu-Nosjean M-C, Validire P, Trautmann A, et al. Matrix architecture defines the preferential localization and migration of T cells into the stroma of human lung tumors. J Clin Invest. 2012;122:899–910.

48. Du B, Shim JS. Targeting Epithelial–Mesenchymal Transition (EMT) to Overcome Drug Resistance in Cancer. Molecules. 2016;21:965.

49. Dituri F, Mancarella S, Giannelli G, Chieti A. TGF- β as Multifaceted Orchestrator in HCC Progression : Signaling , EMT , Immune Microenvironment , and Novel Therapeutic Perspectives. Semin Liver Dis. 2019;39:53–69.

No competing interests reported.

PDF

Version 2

posted 23 May, 2022

You are reading this latest preprint version

Role of high-level S1PR3 in Prognostic significance and tumor microenvironment of Low-grade Glioma

Status:

Version 2

Abstract

Figures

Introduction

Materials And Methods

Cell culture

Quantitative real‐time PCR (qRT‐PCR)

Cell viability analysis

Scratch wound healing assay

Data Acquisition and Bioinformatic Analysis

Gene set variation analysis and Gene Ontology annotation

Assessment of independent prognostic factors

Estimation of TME cell infiltration and assessment of the correlation of S1PR3 gene features with other relevant BP

Results

The landscape of genetic variation of S1PR3 in glioma

Establishment of a prognostic prediction nomogram for glioma associated with S1PR3 expression

Annotation of classification functions determined by consensus clustering analysis

Identification of S1PR3-related genes and S1PR3 inhibitors inhibited proliferation and migration of glioma cells

Characterization of TME immune cell infiltration in gliomas with different S1PR3 expression

Expression of S1PR3 in pan-cancer and the effect of S1PR3 expression on immune cell infiltration across tumor types

Discussion

Conclusions

Declarations

Data Availability

Conflicts of Interest

Funding Statement

Acknowledgments

References

Competing Interests

Supplementary Files

Status:

Version 2