Scientific Papers

Systematic analysis of integrated bioinformatics to identify upregulated THBS2 expression in colorectal cancer cells inhibiting tumour immunity through the HIF1A/Lactic Acid/GPR132 pathway | Cancer Cell International


Biological characteristics of THBS2 in pan-cancer

Expression of THBS2 in pan-cancer

The expression of THBS2 in 22 tissue samples in the HPA and GTEx cohorts was examined using the HPA database. The results indicated that THBS2 was significantly upregulated in various tissues, including cervical, endometrial and skin tissues (Fig. 1A). Furthermore, the expression of THBS2 was examined in various cell types using the TISCH database. Among the cells exhibiting more significant THBS2 expression (log[TPM/10 + 1] > 1), fibroblasts had remarkably high expression of THBS2 in most cancers (Fig. 1B). Gene expression data extracted from the CCLE database indicated that the expression of THBS2 was higher in MESO, SKCM, GBM, THCA and LGG cells (Fig. 1C). A box plot was generated to demonstrate the differential expression of THBS2 between tumour and normal tissues in 16 solid tumour cohorts. In particular, THBS2 was considerably downregulated in KICH and UCEC and upregulated in other tumour types including COAD (Fig. 1D).

Fig. 1
figure 1

Expression of THBS2 in pan-cancer. A Histogram of THBS2 expression in the HPA and GTEx cohorts. Horizontal coordinates represent TPM values; the right side of 0 represents the HPA cohort; the left side represents the GTEx cohort and vertical coordinates represent tissue types; B Heat map demonstrating THBS2 expression in each cell type in pan-cancer single-cell datasets; C Box plot demonstrating THBS2 expression in different cancer cell lines in the CCLE database; D: Box plots demonstrating dysregulation of the THBS2 gene in 16 cancer types; red represents tumour tissues; blue represents normal tissues and * represents significant differences

THBS2 is associated with the immune microenvironment in pan-cancer

Four types of immune-related scores were calculated using the ESTIMATE algorithm. Subsequently, the relationship between THBS2 expression and these immune-related scores was assessed. The results revealed that THBS2 expression was positively correlated with ESTIMATEScore, ImmuneScore and StromalScore. THBS2 expression was strongly correlated with tumour purity in THCA, COAD, READ, LICH and KICH, whereas it was negatively correlated with tumour purity in THCA, DLBC, LGG and GBM (Fig. 2A–D). A heat map was generated to demonstrate the correlation between THBS2 expression and the proportion of 28 types of tumour-infiltrating immune cells. In particular, THBS2 expression was positively associated with the proportion of memory T cells, NK cells and regulatory T cells in pan-cancer and negatively associated with the proportion of activated CD8 T cells in CHOL, DLBC, MESO and TGCT (Fig. 2E). Radar plots was generated to demonstrate the correlation between IPSs and THBS2 expression in pan-cancer. The plot showed that THBS2 expression was positively associated with IPSs in THCA, KICH and LIHC and negatively associated with IPSs in COAD, READ and SKCM (Fig. 2F–I). In addition, THBS2 expression was positively associated with angiogenesis and epithelial–mesenchymal transition and negatively associated with DNA repair, cell cycle and DNA replication in most cancer types (Fig. 2J).

Fig. 2
figure 2

Correlation between THBS2 expression and the immune microenvironment in pan-cancer. AD Radar plot demonstrating the correlation between THBS2 expression and four types of immune-related scores in pan-cancer, with increasing correlation coefficients shown in ascending order; E Heat map demonstrating the correlation between THBS2 expression and the proportion of 28 types of tumour-infiltrating immune cells in pan-cancer, with the colour of the dots indicating a high correlation and positive or negative correlation and * indicating significance; FI Radar plot demonstrating the correlation between IPSs and THBS2 expression in pan-cancer; J Heat map demonstrating the correlation between multiple cell functions and THBS2 expression in pan-cancer

THBS2 is associated with prognosis in pan-cancer

Univariate Cox analysis was performed to examine the relationship between THBS2 expression and prognosis in pan-cancer based on OS, DSS, DFI and PFI data. DFI data were not available in the LAML, SKCM, UVM, GBM and THYM cohorts, whereas DSS and PFI data were not available in the LAML cohort. Forest plots were generated to visualise the results of univariate Cox analysis. In particular, THBS2 expression was remarkably associated with OS in UVM, LGG, GBM, PAAD, KIRP, MESO, ACC, STAD and KIRC (Fig. 3A) and with DSS in UVM, LGG, GBM, PAAD, KIRP, MESO, ACC, STAD and KIRC (Fig. 3B). Furthermore, THBS2 expression was significantly associated with DFI in ACC, LIHC, MESO and PAAD (Fig. 3C) and with PFI in PRAD, UVM, DLBC, BRCA, KICH, MESO, COAD, KIRC and PAAD (Fig. 3D).

Fig. 3
figure 3

THBS2 expression is associated with prognosis in pan-cancer. A Forest plot demonstrating the relationship between THBS2 expression and OS as assessed via univariate Cox analysis; B Forest plot demonstrating the relationship between THBS2 expression and DSS as assessed via univariate Cox analysis; C Forest plot demonstrating the relationship between THBS2 expression and DFI as assessed via univariate Cox analysis; D Forest plot demonstrating the relationship between THBS2 expression and PFI as assessed via univariate Cox analysis

Genomic instability of THBS2 in pan-cancer

Mutations in genetic loci can lead to transcriptional errors and genomic instability, affecting tumorigenesis. The cBioPortal database was used to analyse THBS2 mutations and evaluate the frequency of various mutation types in pan-cancer. The results revealed that THBS2 had a higher frequency of mutation in UCEC, SKCM, LUSC, LUAD and COAD and a higher frequency of copy number amplification in STAD, ACC, SARC, UCS and MESO (Fig. 4A). High-frequency mutations significantly associated with THBS2 mutations in pan-cancer are shown in Fig. 4B. The mutation frequency of the MFSD138 gene was remarkably higher in the THBS2-mutation group and lower in the non-THBS2-mutation group. Based on OS, DSS, DFI and PFI data, survival was compared between the THBS2-mutation and non-THBS2-mutation groups in pan-cancer. KM curves indicated that survival was significantly better in the THBS2-mutation group (Fig. 4C–E). However, OS was not significantly different between the two groups. Furthermore, the correlation of THBS2 expression with CIN, LOH, MSI and TMB was assessed in pan-cancer. The results revealed a positive correlation between MSI and THBS2 expression in COAD and a negative correlation between CIN and THBS2 expression in KIHC and LUSC. In addition, THBS2 expression was positively correlated with LOH in UVM, PRAD and THYM. TMB was negatively associated with THBS2 expression in LIHC and UVM (Fig. 4F).

Fig. 4
figure 4

Genomic instability of THBS2 in pan-cancer. A Histogram demonstrating the distribution of mutation types of THBS2 in pan-cancer; B Gene mutations significantly associated with THBS2 mutations in pan-cancer; CE KM curves demonstrating the relationship between THBS2 mutations and survival based on DSS, DFI and PFI data in pan-cancer; F Histogram demonstrating the correlation between THBS2 expression and CIN, LOH, MSI or TMB in pan-cancer

THBS2 methylation in pan-cancer

Methylation has been associated with the development, mutagenesis and immune tolerance of various cancers. The methylation levels of THBS2 were compared between tumour and normal tissues using the UALCAN database (https://ualcan.path.uab.edu/). The results showed that THBS2 methylation levels were remarkably lower in BLCA, BRCA, COAD, LIHC and LUSC tissues than in normal tissues (Fig. 5A–E). Further analysis of the correlation between the cg21652958 methylation site in the 5′-UTR of THBS2 and THBS2 expression showed that the methylation levels of cg21652958 were significantly positively correlated with THBS2 expression in DLBC, KICH and LIHC (Fig. 5F). KM curves were generated to compare OS between the low- and high-cg21652958-methylation groups in ACC, LUAD and OV. The results revealed that OS was better in the low-cg21652958-methylation group in ACC and in the high-cg21652958-methylation group in LUAD and OV (Fig. 5G–I). Further analysis of the relationship between THBS2 expression and the cg21732383 methylation site in TSS1500 (1500-bp upstream of the transcription initiation site) showed that the methylation levels of cg21732383 were most strongly correlated with THBS2 expression in LIHC (Fig. 5J). KM curves showed that OS was better in the high-cg21732383-methylation group in LGG and in the low-cg21732383-methylation group in SARC and UCEC (Fig. 5K–M).

Fig. 5
figure 5

THBS2 methylation in pan-cancer. AE Box plots demonstrating differences in THBS2 methylation levels between normal and tumour tissues in pan-cancer; red represents tumour tissues; blue represents normal tissues; p-values represent significant differences; F Radar plot demonstrating the correlation between methylation levels and THBS2 expression at cg21652958 locus in pan-cancer; GI KM curves demonstrating OS in the high- and low-cg21652958-methylation groups; orange represents the high-cg21652958-methylation group; green represents the low-cg21652958-methylation group; p-values represent significant differences; J Radar plot demonstrating the correlation between the methylation levels of cg21732383 and THBS2 expression in pan-cancer; KM Radar plot demonstrating differences in OS between the high- and low-cg21732383-methylation groups based on KM curves

Prognostic characterisation and role of THBS2 in COAD

Expression of THBS2 in COAD

The abovementioned results suggest that the expression, genomic variations and methylation of THBS2 are closely related to tumour prognosis.

To investigate the impact of THBS2 on colon cancer, its expression pattern was examined in TCGA-COAD cohort and its relationship with the enrichment scores of the HALLMARK gene set was assessed. The results showed that THBS2 expression was positively associated with epithelial–mesenchymal transition and angiogenesis and negatively associated with DNA repair, oxidative phosphorylation, spermatogenesis and peroxisome (Fig. 6A). The protein–protein interaction network of THBS2 in COAD is shown in Fig. 6B. Based on this network, the expression of THBS2 was positively associated with that of ADAMTS12, MMP2 and ADAMTS2 (Fig. 6C). Based on the median expression of THBS2 (3.92), TCGA-COAD cohort was divided into high- and low-THBS2-expression groups. A total of 1052 differentially expressed genes were identified between the two groups, which included 4 copper death-related genes and 28 iron death-related genes. As shown in the heat map in Fig. 6D, LCN2 and NOS2 were included in the THBS2 group, whereas the remaining genes were upregulated. Furthermore, the 992 upregulated genes in the high-THBS2-expression group were subjected to KEGG and GO functional enrichment analyses. KEGG analysis revealed that the genes were primarily enriched in pathways associated with rheumatoid arthritis, cell adhesion molecules, phagosomes, viral protein interactions with cytokines and cytokine receptors, ECM–receptor interactions and proteoglycans in cancer. GO analysis revealed that the genes were enriched in biological processes such as cell structural organisation, cell migration and leukocyte chemotaxis; cellular components such as the extracellular matrix and endoplasmic reticulum lumen and molecular functions such as growth factor binding and cytokine binding (Fig. 6E–H). In addition, 60 genes were down-regulated in the high THBS2-expression group. These down-regulated genes were further used for enrichment analyses. As shown in the figure, there are no significant enrichment results obtained in KEGG pathway analysis (Additional file 1: Fig. S1). The possible reason for this is due to the low number of genes involved in the enrichment analysis. Therefore, we were more concerned regarding the expression profiles of the upregulated genes in the high-THBS2-expression group.

Fig. 6
figure 6

Expression of THBS2 in COAD. A Stick plot demonstrating the correlation between THBS2 expression and the enrichment scores of the HALLMARK gene set; the size of the dots represents a strong correlation; the right side represents a positive correlation; the left side represents a negative correlation; the colour represents p-values; the smaller the p-value, the more significant the correlation; B Protein–protein interaction network of THBS2 constructed using the STRING database; C Stick plots demonstrating the correlation between the expression of THBS2 and that of other proteins in the network; D Heat map of differential expression of copper death- and iron death-related genes between the high- and low-THBS2-expression groups; E Bubble plots of KEGG, GO-BP (biological process), GO-MF (molecular function) and GO-CC (cellular component) enrichment analyses of marker genes in the high-THBS2-expression group; the size of the dots represents the number of enriched differentially expressed genes; the colour represents the significance of the enrichment

Differences in immunological characteristics between high- and low-THBS2-expression groups

KM curves revealed that PFI was worse in the high-THBS2-expression group, indicating that THBS2 expression was remarkably associated with prognosis in COAD (Fig. 7A). To investigate the mechanisms through which THBS2 affects the prognosis of COAD through immune-related factors, the expression of immunomodulators was examined in the two THBS2 groups. The results indicated that the expression of immunomodulators was higher and that of CCL28, CXCL1, CXCL2 and CXCL3 chemokines was lower in the high-THBS2-expression group (Fig. 7B). Furthermore, the proportion of 28 types of tumour-infiltrating immune cells was compared between the two THBS2 groups. As shown in the box plot in Fig. 7C, the degree of immune cell infiltration was lower in the low-THBS2-expression group than in the high-THBS2-expression group. The ESTIMATE algorithm was used to calculate stromal, immune and ESTIMATE scores in the two groups. The three types of scores were considerably lower in the low-THBS2-expression group than in the high-THBS2-expression group (Fig. 7D–G). Similarity p-values were calculated in the two THBS2 groups and three immunotherapy cohorts using SubMap. As shown in the heat map in Fig. 7H, the low-THBS2-expression group was more responsive to PD-L1 immunotherapy, whereas the high-THBS2-expression group was more responsive to CTLA4 immunotherapy. In addition, all four IPSs were considerably higher in the low-THBS2-expression group than in the high-THBS2-expression group (Fig. 7I–L). These results indicated that patients in the low-THBS2-expression group were more likely to benefit from immunotherapy.

Fig. 7
figure 7

Immunological differences between high- and low-THBS2-expression groups. A KM curves demonstrating PFI in the high- and low-THBS2-expression groups; B: Heat map demonstrating the differential expression of immunomodulators between the high- and low-THBS2-expression groups; C Box plot demonstrating differences in the proportion of 28 types of tumour-infiltrating immune cells between the two THBS2 groups; DG Box plots demonstrating differences in the four types of scores between the two THBS2 groups as evaluated using the ESTIMATE algorithm; H Heat map of similarity between the high- and low-THBS2-expression groups and immunotherapy response groups (with and without response subgroups) obtained from SubMap analysis; I, J Violin plots demonstrating differences in IPSs between the high- and low-THBS2-expression groups

THBS2 is associated with genomic mutations in COAD

Gene mutations can promote, cause or orchestrate the malignant progression of tumours, and the study of mutations at the genomic level is crucial for developing tumour-targeting drugs and novel therapies for treating cancer. A waterfall plot integrating clinical information (age, stage and sex) and the top 20 mutated genes in the two THBS2 groups was plotted. The results showed that the SNV frequency of APC, TP53 and TTN was remarkably high in colon cancer (Fig. 8A). Subsequently, heat maps demonstrating the co-linearity and mutual exclusivity of the top 10 mutated genes in the two groups were plotted. The results revealed that co-linearity and mutual exclusivity were higher in the high-THBS2-expression group than in the low-THBS2-expression group (Fig. 8B, C). Furthermore, forest plots of the top 100 mutated genes with significant differences between the two groups were plotted to demonstrate their p-values and upper and lower confidence intervals (Fig. 8D). All samples in TCGA-COAD cohort were divided into two groups, mutant-THBS2 (MT) and wildtype-THBS2 (WT), based on the SNV status of the THBS2 gene. Both TMB and TNB were found to be significantly higher in the MT group than in the WT group (Fig. 8E, F). The mutation frequency of the top 10 genes was highly co-linear and mutually exclusive in the MT group (Fig. 8G).

Fig. 8
figure 8

Genomic mutations associated with THBS2 in COAD. A Waterfall plot demonstrating the SNV frequency of top 20 genes in TCGA-COAD cohort; B Heat map demonstrating the co-linearity and mutual exclusivity of top 10 mutated genes in the high-THBS2-expression group; C Heat map demonstrating the co-linearity and mutual exclusivity of top 10 mutated genes in the low-THBS2-expression group; D Forest plot demonstrating differences in the mutation frequency of top 100 genes between the high- and low-THBS2-expression groups, with p-values representing significant differences; E, F Box plots demonstrating differences in TMB and TNB between the mutant-THBS2 (MT) and wildtype-THBS2 (WT) groups; G Heat map demonstrating the co-linearity and mutual exclusivity of top 10 mutated genes in the MT group

THBS2 affects lactate metabolism in CT26 cells through HIF1

To investigate the relationship between THBS2 and HIF1, THBS2 was overexpressed in CT26 cells (Fig. 9A). Immunofluorescence analysis and western blotting revealed increased nuclear translocation of HIF1 in THBS2-overexpressing cells (Fig. 9B, C). As shown in Fig. 9D, THBS2 overexpression led to an increase in ECAR and a decrease in OCR in CT26 cells, indicating a shift toward anaerobic metabolism instead of oxidative phosphorylation for energy production. Consistently, THBS2 overexpression increased lactate production in CT26 cells (Fig. 9E). These results suggest that THBS2-overexpressing CT26 cells tend to derive energy through anaerobic glycolysis and produce lactate.

Fig. 9
figure 9

THBS2 influences lactate metabolism in CT26 cells through HIF1. A Representative images from three independent western blotting experiments showing the overexpression of THBS2 in CT26 cells. B Immunofluorescence staining of HIF1 (green) and nuclei (blue) in CT26 cells overexpressing THBS2. Representative images from three independent experiments are shown. C Detection of HIF1 expression in the nucleus and cytoplasm of CT26 cells overexpressing THBS2. D Measurement of extracellular acidification rate (ECAR) and oxygen consumption rate (OCR) in CT26 cells overexpressing THBS2 on a Seahorse analyser. E Assessment of lactate metabolism in CT26 cells overexpressing THBS2 (n = 3; *** indicates p < 0.001 compared with the vector group). F Detection of downstream transcriptional genes of HIF1 via PCR (n = 3; *** indicates p < 0.001 compared with the vector group). G Representative images of western blotting showing the knockdown of HIF1 in THBS2-overexpressing CT26 cells. Representative images from three independent experiments are shown. H Evaluation of lactate metabolism after THBS2 overexpression and HIF1 knockdown in CT26 cells (n = 3; ns indicates no significant difference compared with the vector group)

HIF1 regulates several genes involved in oxygen supply and metabolic adaptation. When stabilised and translocated to the nucleus, HIF1 binds to specific DNA sequences and initiates the transcription of target genes [29,30,31]. As shown in Fig. 9F, THBS2 overexpression enhanced the transcription of downstream genes regulated by HIF1. To rescue THBS2 overexpression-induced anaerobic glycolysis, HIF1 was knocked down in THBS2-overexpressing CT26 cells (Fig. 9G). The results revealed that HIF1 knockdown reversed the THBS2 overexpression-induced increase in lactate levels (Fig. 9H). Altogether, these results suggest that THBS2 enhances anaerobic glycolysis in tumour cells by modulating the nuclear translocation of HIF1.

THBS2 influences M2 polarisation of macrophages through GPR132

We have previously reported that upregulation of THBS2 in CT26 cells leads to a shift toward anaerobic glycolysis and lactate production. Lactate can interact with GPR132 to activate downstream signalling pathways and impact macrophage function and polarisation [32]. Particularly, in the context of inflammation and the tumour microenvironment, high levels of lactate can promote the polarisation of macrophages toward the M2 phenotype. Therefore, in this study, we investigated whether upregulation of THBS2 in tumour cells exerts similar effects on macrophages.

RAW264.7 macrophages were co-cultured with the conditioned medium from THBS2-overexpressing CT26 cells. Flow cytometry revealed that the co-cultured cells exhibited an M2-like phenotype (Fig. 10A). The conditioned medium from CT26 cells was centrifuged to isolate high-molecular-weight proteins. These proteins failed to replicate the M2 polarisation-promoting effect (Fig. 10B). However, the addition of lactate to the culture medium of RAW264.7 cells facilitated their differentiation (Fig. 10C), suggesting that lactate is crucial for macrophage polarisation. To verify this hypothesis, GPR132-knockout (KO) mouse models were established, and peripheral blood monocytes were phenotypically characterised using flow cytometry (Fig. 10D). Macrophages isolated from GPR132-KO mice were co-cultured, demonstrating that GPR132-deficient macrophages did not respond to lactate induction (Fig. 10E). Altogether, these results indicate that THBS2 influences M2 polarisation of macrophages by interacting with GPR132.

Fig. 10
figure 10

THBS2 influences M2 polarisation of macrophages through GPR132. A Collection of conditioned medium from THBS2-overexpressing CT26 cells and its addition to RAW264.7 cells. Flow cytometric analysis of CD163 and CD206 expression in cells after 24 h; B Centrifugation of conditioned medium from CT26 cells to isolate high-molecular-weight proteins and the addition of the medium to RAW264.7 cells. Flow cytometric analysis of CD163 and CD206 expression in cells after 24 h; C Addition of lactate to RAW264.7 cell culture medium. Flow cytometric analysis of CD163 and CD206 expression in cells after 24 h. D Flow cytometric analysis of GPR132 expression in peripheral blood mononuclear cells from GPR132-knockout (KO) mice; E Collection of conditioned medium from THBS2-overexpressing CT26 cells and its addition to macrophages from GPR132-KO mice. Flow cytometric analysis of CD163 and CD206 expression in cells after 24 h

THBS2 overexpression in tumour cells mediates M2 polarisation of macrophages, resulting in the suppression of proliferation and cytotoxicity of CD8 + T cells

To assess the role of THBS2 overexpression in the tumour microenvironment, its impact on T-cell immune responses and anti-tumour immunity was examined in OT-1 mice. The transgenic T-cell receptor (TCR) in OT-1 mice recognises and responds to ovalbumin (OVA) antigen with high specificity. As shown in Fig. 11A, CD8 + T cells from OT-1 mice were co-cultured with M2-polarised macrophages influenced by THBS2-overexpressing CT26 cells from previous experiments. CFSE assay revealed that the proliferation of CD8 + T cells was significantly suppressed in response to OVA antigen when the cells were co-cultured with THBS2-CT26-influenced (M2-polarised) macrophages (Fig. 11B). Additionally, the release of IFN-γ was reduced (Fig. 11C). The cytotoxicity assay against CT26-OVA cells (CT26 cells overexpressing OVA antigen) demonstrated a significant killing effect of CD8 + T cells from OT-1 mice, whereas cytotoxicity was significantly diminished when influenced by M2-polarised macrophages (Fig. 11D). Furthermore, HIF1 was knocked down in CT26-THBS2 cells (CT26-THBS2-siHIF1 cells), and conditioned media derived from CT26, CT26-THBS2 and CT26-THBS2-siHIF1 cells were used to induce RAW264.7 polarisation. The macrophages were subsequently co-cultured with CD8 + T cells from OT-1 mice in the presence of OVA antigen. As shown in Fig. 11E, the release of IFN-γ decreased after THBS2 was overexpressed in CT26 cells and increased after HIF-1 was knocked down in THBS2-overexpressing CT26 cells. Altogether, these results indicated that M2 polarisation induced by THBS2-overexpressing CT26 cells significantly impaired the proliferation and cytotoxic potential of CD8 + T cells.

Fig. 11
figure 11

M2 macrophages suppress the proliferation and cytotoxicity of CD8 + T cells. A Schematic representation of co-culture of OT-1 cells with macrophages; B Flow cytometric analysis of CFSE in CD8 + T cells; C Flow cytometric analysis of IFN in CD8 + T cells; D Cytotoxic effects of CD8 + T cells on CT26 cells overexpressing OVA. Flow cytometric analysis of 7-AAD in CT26 cells. E IFN-γ released from OT-1 CD8 T cells

THBS2 influences the infiltration levels, proliferation and cytotoxicity of T cells in the tumour microenvironment

To examine the effects of THBS2 on the tumour microenvironment, CT26 cells and THBS2-overexpressing CT26 cells (CT26-THBS2) were injected into Balb/c mice. As shown in Fig. 12A, CT26-THBS2 tumour-bearing mice had significantly high tumour volume. In addition, these mice had a high abundance of CD3 + /CD8 + cells, indicating high infiltration levels of cytotoxic T lymphocytes (Fig. 12B). The increased abundance of CD3 + /CD4 + and Foxp3 + cells suggested increased infiltration levels of helper T lymphocytes and regulatory T cells, respectively (Fig. 12C). Additionally, CT26-THBS2 tumour-bearing mice had high macrophage infiltration, with a higher proportion of M2-polarised macrophages (Fig. 12D). After CT26 cells overexpressing OVA antigen (CT26-OVA) and CT26 cells overexpressing both OVA antigen and THBS2 (CT26-OVA-THBS2) were implanted into Balb/c mice, the mice were intravenously injected with CD8 + T cells from OT-1 mice. Treatment with CD8 + T cells significantly reduced tumour volume, whereas the impact on THBS2-overexpressing tumour cells was minimal (Fig. 12E). CFSE assay revealed that the proliferation of CD8 + T cells was considerably reduced in the CT26-OVA-THBS2 group (Fig. 12F), indicating that THBS2 inhibited the proliferation of CD8 + T cells in vivo.

Fig. 12
figure 12

THBS2 influences the infiltration levels, proliferation and cytotoxicity of T cells in the tumour microenvironment. A Establishment of tumour-bearing Balb/c mouse models using CT26 and CT26-THBS2 cells. Tumour size was measured after 30 days; B Flow cytometric analysis of CD3 + /CD8 + cell abundance in tumour tissues of different groups; C Flow cytometric analysis of CD3 + /CD4 + and Foxp3 + cell abundance in tumour tissues of different groups; D Flow cytometric analysis of F4/80 + /CD11b + , CD163 + and CD206 + cell abundance in tumour tissues of different groups; E Implantation of CT26-OVA and CT26-OVA-THBS2 cells into Balb/c mice, followed by intravenous injection of CD8 + T cells from OT-1 mice for treatment. Tumour size was measured on day 0; F: Flow cytometric analysis of CFSE in tumour tissues

Immunosuppressive phenotype induced by high expression of THBS2 depends on GPR132 expression

To investigate the relationship between THBS2 and GPR132 in vivo, GPR132-KO Balb/c mouse models were established and inoculated with CT26 and CT26-THBS2 cells to induce tumour formation (Fig. 13A). The results showed that overexpression of THBS2 did not significantly alter tumour volume in GPR132-KO mice (Fig. 13B). In addition, the abundance of tumour-infiltrating CD8 + cells or TIM3 + /PD1 + CD8 + T cells was not substantially different across various groups of mice, indicating that the functional exhaustion of cytotoxic T lymphocytes in the tumour was not significantly affected (Fig. 13C). Similarly, no significant differences were observed in macrophage infiltration, M2 polarisation and CFSE dilution levels of CD8 + T cells across groups (Fig. 13D, E). These results suggested that GPR132 knockout blocked tumour progression and immunosuppression induced by THBS2-overexpressing CT26 cells.

Fig. 13
figure 13

Immunosuppressive phenotype induced by high expression of THBS2 depends on GPR132 expression. A Establishment of GPR132-KO and tumour-bearing Balb/c mouse models; B Assessment of tumour size 30 days after the injection of CT26 and CT26-THBS2 cells in mice; C Flow cytometric analysis of CD3 + /CD8 + and TIM3 + /PD1 + cell abundance in tumour tissues of different groups. D Flow cytometric analysis of F4/80 + /CD11b + , CD163 + and CD206 + cell abundance in tumour tissues of different groups; E Flow cytometric analysis of CFSE in CD8 + T cells



Source link