REEP5 as a prognostic factor for overall survival in ESCC patientsTo evaluate the prognostic significance of REEP5 in esophageal squamous cell carcinoma, we conducted Kaplan-Meier and COX proportional hazards model analyses using data from the GEO and TCGA databases. Our findings indicate that patients with high REEP5 expression experienced significantly shorter overall survival compared to those with low REEP5 expression (median survival: 1261 days) in the GEO database. This prognostic observation was consistently validated in the TCGA database (Fig. 1H, N). Furthermore, in COX analysis for staging phase, gender, and age, REEP5 emerged as a significant prognostic factor (HR = 1.639; 95% CI: 1.071–2.510). Moreover, REEP5 was identified as an independent prognostic indicator, unaffected by age and gender, in the TCGA dataset (Tables 1 and 2).Figure 1 Relationship between REEP5 and Clinicopathological Features in Esophageal Squamous Cell Carcinoma (ESCC). A, Association of REEP5 with ESCC clinicopathological features in the Gene Expression Omnibus (GEO) database. B, Correlation of REEP5 with ESCC clinicopathological features in The Cancer Genome Atlas (TCGA) database. C and I,, There is a trends indicate an enrichment of REEP5 in high-grade ESCC across both GEO and TCGA databases. Differences were statistically analyzed using one-way ANOVA. D and J, Analysis reveals that REEP5 expression does not correlate with the tumor site in ESCC within both GEO and TCGA databases. Differences were statistically analyzed using an unpaired t-test. E and K, Analysis show no significant association between REEP5 expression and alcohol consumption in ESCC patients in both databases. Differences were statistically analyzed using an unpaired t-test. F and L, Analysis show no link between REEP5 expression and smoke consumption among ESCC patients in the GEO and TCGA databases. Differences were statistically analyzed using an unpaired t-test. G and M, Analysis indicate that REEP5 expression does not vary significantly with gender among ESCC patients in both GEO and TCGA databases. Differences were statistically analyzed using an unpaired t-test. H and N, Kaplan-Meier analysis of REEP5 expression using GEO and TCGA databases, the cohort was divided into two groups, based on the optimal cutoff point of REEP5 risk scores. The significance of prognostic value was assessed using the log-rank test.Table 1 Univariate and multivariate analysis of overall survival (OS) parameters in the GEO database.Table 2 Univariate and multivariate analysis of overall survival (OS) parameters in the TCGA database.Association of REEP5 with clinicopathological factorsDifferent levels of REEP5 expression in patients revealed distinct clinicopathological patterns in the GEO and TCGA datasets. TNM staging, tumor location in the esophagus (Fig. 1D and J), along with drinking and smoking habits (Fig. 1E, K, F and L), also including gender (Fig. 1G and M), showed an asymmetric distribution correlated with increased expression of REEP5. Patients with higher expression of REEP5 tended to have shorter survival times compared to those with lower expression (Fig. 1A, B). The comparative analysis of different TNM stages in the GEO database reveals that the enrichment of REEP5 in stage III is showing an upward trend compared to stage II (Fig. 1C). However, the enrichment level of REEP5 in stage I is consistently higher than that observed in stages II and III. This trend was also observed in the TCGA database (Fig. 1I), although the differences were not statistically significant. The findings indicate that ESCC with elevated malignancy levels may demonstrate an enrichment of REEP5. Furthermore, for patients diagnosed with ESCC, high expression levels of REEP5 are associated with a detrimental effect on overall survival prognosis.Role of REEP5 in mRNA splicing and protein stabilizationTo elucidate the biological functions of REEP5, we employed Pearson correlation analysis (R > 0.5, P < 0.05) in the GEO and TCGA databases to identify genes highly related to REEP5. Subsequent GO and KEGG analyses of these genes indicated that the biological processes(BP) associated with REEP5 in the GEO dataset include cell apoptosis, protein transport, mRNA splicing, DNA repair, and protein stabilization (Fig. 2A). The cellular components(CC) and molecular functions(MF) most related to REEP5 were nucleoplasm and protein binding, respectively (Fig. 2B, C). The KEGG most related to REEP5 was Huntington (Fig. 2D). Similar biological functions, including mRNA splicing and protein stabilization, and molecular function involving protein binding were observed in the TCGA database (Fig. 2E-H). These analyses suggest that REEP5 likely plays a crucial role in transcriptional and translational processes in esophageal squamous cell carcinoma.Figure 2 Relationship between REEP5 and mRNA Splicing and Protein Stabilization in Esophageal Squamous Cell Carcinoma. A-C, Biological Processes (BP), Cellular Components (CC), and Molecular Functions (MF) associated with REEP5 in the Gene Expression Omnibus (GEO) database. D, KEGG pathway analysis of REEP5 in the GEO database. E-G, Biological Processes (BP), Cellular Components (CC), and Molecular Functions (MF) associated with REEP5 in The Cancer Genome Atlas (TCGA) database. H, KEGG pathway analysis of REEP5 in the TCGA database.REEP5 positively correlates with mRNA spliceosome assembly and disassemblyInvestigations using the Amigo2 gene set revealed that mRNA splicing mechanisms include cis and trans splicing, spliceosome assembly and disassembly, and snRNP assembly among others. To further examine the specific splicing processes involved and the impact of increased expression of REEP5 in esophageal squamous cell carcinoma, we employed Gene Set Enrichment Analysis (GSEA) and Gene Set Variation Analysis (GSVA) across the GEO and TCGA datasets. The GSEA results indicated that in groups with high REEP5 expression, there was significant activity in snRNP complex and spliceosome assembly (Fig. 3A, B). Additionally, GSVA analysis in the GEO dataset showed a positive correlation between increased expression of REEP5 and the catalysis of the spliceosome, tri-snRNP complex assembly, and spliceosome complex disassembly, supporting the involvement of REEP5 in these critical splicing steps. Similar trends were observed in the TCGA database, particularly for spliceosome complex disassembly (Fig. 3C, D). These findings suggest a positive correlation between REEP5 and the accelerated assembly and disassembly of mRNA splicing complexes in esophageal squamous cell carcinoma.Figure 3 Correlation between REEP5 and mRNA splicing processes was investigated using data from the GEO and TCGA databases. A and B, Gene Set Enrichment Analysis (GSEA) revealed strong associations of REEP5 with the spliceosome snRNP complex and splicing complex assembly in esophageal squamous cell carcinoma (ESCC). C and D, Heatmaps depicting REEP5 expression and mRNA splicing functional richness scores across patients in the GEO and TCGA cohorts are presented, sorted by descending REEP5 expression levels. The accompanying bar and line graphs display the Pearson correlation analysis results (R and P values) .REEP5 positively correlates with Cancer-related immunosuppressive checkpointsRecent studies have identified a role for mRNA splicing in modulating tumor immunosuppression23. Consequently, we explored the relationship between REEP5 and various immunosuppressive checkpoints including TIM-3, PD-1, PD-L2, CD200R1, CTLA-4, CD47, HVEM, and TIGIT24. In the GEO dataset, positive correlations were found between REEP5 and CTLA-4, TIM-3, PD-1, TIGIT, and HVEM. These associations were also confirmed in the TCGA dataset, particularly for CTLA-4, TIM-3, and HVEM (Fig. 4A). The results support the hypothesis that REEP5 and other immunosuppressive checkpoints might be elevated due to interactions between tumor cells and T cells, leading to T cell infiltration and subsequent exhaustion. This creates an immunosuppressive microenvironment in esophageal squamous cell carcinoma, potentially adversely affecting the prognosis of immune therapies.Figure 4 Relationship between REEP5 expression and cancer-related immunosuppressive checkpoints and immune cell infiltration in GEO and TCGA databases. A, The Pearson correlation coefficient between REEP5 expression and cancer-related immunosuppressive checkpoints is depicted. The width of each band indicates the magnitude of the R-value, while the color denotes the p-value, with statistical significance assessed through Pearson’s method. B and C, The disparities in immune cell infiltration between cohorts with high versus low expression of REEP5 are examined. The statistical relevance of these differences is determined using the Wilcoxon rank-sum test.Relationship between REEP5 expression and T cell infiltration in esophageal squamous cell carcinomaTo explore the relationship between REEP5 expression levels and immune cells in esophageal squamous cell carcinoma (ESCC), we conducted cellular fractionation using public single-cell sequencing datasets (GSE196756). We found that T cells and B cells in ESCC significantly increased by over 20% compared to adjacent normal tissue, while mast cell numbers significantly decreased (Fig. 5A, D). Additionally, REEP5 expression in ESCC immune cells was similar to that of inhibitory immune checkpoints such as CTLA-4, TIM-3, and HVEM, and was notably enriched in mast cells (Fig. 5B, C, E, F).Additionally, CIBERSORT analysis using both GEO and TCGA databases was performed to investigate the association between REEP5 expression and immune cell infiltration in ESCC. Results from the GEO database indicated significant correlations between REEP5 expression and infiltration levels of plasma cells, CD8 + T cells, CD4 + memory T cells, resting memory CD4 T cells, resting NK cells, activated NK cells, and M2-type macrophages. Specifically, significant differences in CD8 + T cell infiltration were observed between high and low-expression groups of REEP5, with the high-expression group showing notably higher levels of CD8 + T cell infiltration, a finding corroborated by analyses from the TCGA database (Fig. 4B, C; P < 0.05). These findings underscore a strong correlation between REEP5 expression and the extent of CD8 + T cell infiltration in ESCC, highlighting REEP5 as a potential modulator of immune responses in this context.Figure 5 Results of cell clustering in cancerous and paracancerous tissues in GSE196756. A, Cell clustering patterns in cancerous tissues from GSE196756. B and C, REEP5 expression across various cell types within cancerous tissues from GSE196756. D, Cell clustering in paracancerous normal tissues specifically from GSE196756. E and F, REEP5 expression in different cell types within the paracancerous tissues of GSE196756.Elevated expression of REEP5 in esophageal squamous cell carcinoma samplesDue to the lack of control tissue samples in the TCGA-ESCA dataset, we sourced normal esophageal samples from the GTEx database to evaluate the differential expression levels of the REEP5 gene between esophageal cancer and normal esophageal tissues. Additionally, as a primary research objective of this study, we aimed to investigate the differential expression of REEP5 between ESCC and adjacent normal tissue, which necessitates experimental validation. To achieve this, we conducted qPCR and Western blot analyses to compare REEP5 expression levels in these two tissue types. The results indicated that in the TCGA-GTEx database consortium, REEP5 was expressed at higher levels in ESCA than in normal esophageal samples (Fig. 6A, p < 0.001). In experiments comparing ESCC with adjacent normal tissue, the expression of REEP5 was significantly elevated (Fig. 6B, C,D). Specifically, the mean relative expression level of REEP5 mRNA in cancerous tissue was 1.48 times that of adjacent normal tissue (P = 0.0006), and the relative expression level of REEP5 protein was 2.38 times that of adjacent normal tissue (p < 0.0001).Figure 6 REEP5 expression is elevated in cancer tissues of ESCC patients. A, Comparative analysis of REEP5 utilizing data from TCGA and the GTEx databases(unpaired t-test, p  < 0.05). B and C, Western blot analysis of REEP5 in cancer and adjacent paracancerous tissues. D, RT-qPCR analysis of REEP5 expression in cancer and adjacent paracancerous tissues. Data are presented as mean ± standard error of the mean (SEM). (*mean under paired t-test, p <0.05, **mean under paired t-test, p <0.01, ***mean under paired t-test, p <0.001, ****mean under paired t-test, p <0.0001).Significant increase in REEP5 protein levels in esophageal squamous cell carcinoma samplesTo confirm the qPCR and Western blot findings, we further analyzed pathological sections of 10 pairs of ESCC and adjacent non-cancer tissues using immunohistochemistry (IHC), capturing images at 200× and 400× magnifications (Fig. 7A-J and a-j). Using Image-Pro Plus software, images captured at 200× magnification were analyzed for total area and integrated optical density to calculate the average optical density (average optical density = integrated optical density / total area), serving as a measure of REEP5 protein density in specific samples.The results indicated a significant increase in REEP5 protein expression levels in cancer tissues. Specifically, the average density of REEP5 protein in cancer tissues was 0.0150067, compared to 0.006807 in adjacent normal tissues, indicating an 2.2-fold increase in REEP5 protein expression in cancer tissues relative to adjacent normal tissues (Fig. 7K). This difference was statistically significant (p = 0.037). Additionally, analysis of IHC staining patterns revealed that REEP5 protein in cancer tissues tended to aggregate more in the nucleus and cytoplasm compared to normal tissues.Figure 7 REEP5 is enriched in the nucleus and cytoplasm of ESCC cells. A-J, Immunohistochemistry(IHC)staining of REEP5 in cancer tissues. a-j, Immunohistochemistry(IHC) staining of REEP5 in adjacent paracancerous tissues. K, the mean optical density values of REEP5 proteins in patients with ESCC (= integrated optical density/total area of the image). Data are presented as mean ± standard error of the mean (SEM). (* mean under paired t-test, p <0.05).
