SEC61G is overexpressed in head and neck squamous cell carcinoma
Using Oncomine data, we observed that SEC61G was upregulated in almost all cancer types, compared to normal tissues (p<0.001, |log2 fold change|>1.5 in all gene ranks), including breast cancer, kidney cancer, brain and central nervous system cancer, and head and neck cancer (Figure 1A). Then, we explored SEC61G expression in the TCGA dataset using TIMER. SEC61G was highly expressed in pan-cancers compared to normal tissues, including HNSCC (Figure 1B). Further, SEC61G expression in HPV-negative HNSCC was higher than expression in HPV-positive HNSCC (Figure 1B). The analyses of the four Oncomine datasets (Pyeon Multi-Cancer, Estilo Head-Neck, Ye Head-Neck, Peng Head-Neck) also showed that SEC61G was overexpressed in HNSCC (Figure 1C).
Furthermore, the TCGA dataset analysis showed that regardless of the HPV infection status, the expression of SEC61G was higher in HNSCC than in normal tissues (all p<0.001, Kruskal-Wallis test). Our results also showed the SEC61G expression correlated positively with the clinical stage (p=0.047), where later disease-stage patients tended to express more SEC61G. In contrast, SEC61G expression did not correlate with pathological grade (p=0.77) (Figure 1D).
SEC61G overexpression correlates with poor overall survival in HNSCC
To examine the relationship between SEC61G expression and overall survival (OS), we divided patients into high and low expression groups based on the median SEC61G expression in HNSCC-TCGA. Kaplan-Meier survival analysis showed that the OS of high SEC61G-expressing patients was significantly poorer than that of patients with low SEC61G expression (p<0.001) (Figure 2A). Similar results were observed in HPV-negative and HPV-positive subgroups (p<0.001 and p=0.001, respectively) (Figure 2B, 2C). Multivariate analysis also showed that SEC61G expression was an independent prognostic factor for HNSCC [hazard ratio (HR)= 1.80, 95% CI: 1.35-2.39, p<0.001] (Table 2).
To further verify the prognostic value of SEC61G expression in HNSCC, we analyzed the GSE65858 data set (including survival data). Detailed clinicopathologic features are listed in Table 1. Kaplan-Meier survival analysis indicated that patients with high SEC61G expression had an inferior OS than did patients with low SEC61G expression (p=0.019). Subgroup analyses also showed that patients with high SEC61G expression had a worse OS than did patients with low SEC61G expression in the HPV-negative and -positive subgroups (p=0.048 and p=0.042, respectively) (Figure 2D-F). The MAV confirmed that SEC61G was an independent prognostic factor for OS in HNSCC (HR=1.87, 95% CI: 1.14-3.07, p=0.013) (Table 2).
SEC61G demethylation and DNA amplification in HNSCC
To investigate the mechanism of SEC61G upregulation in HNSCC, we analyzed SEC61G methylation and CNV using cBioPortal and the UALCAN web platform. The results showed that SEC61G expression was negatively correlated with methylation (R=0.258, p<0.001) in HNSCC (Figure 3A). SEC61G promoter methylation in tumor tissues from the TCGA-HNSCC dataset was significantly lower than methylation in normal tissues adjacent to tumors (p<0.001) (Figure 3B).
CNV data showed that SEC61G DNA amplification was present in 9.66% (51/528) patients. Further, SEC61G expression in the DNA amplification group was significantly higher than in the other groups (Deletion, Diploid, and Gian groups) (all p<0.001) (Figure 3C). After grouping patients with amplified SEC61G into the altered group and the other patients into the unaltered group, Kaplan-Meier analysis showed that the OS of the altered group was lower than the OS of the unaltered group (p=0.034) (Figure 3D).
Functional enrichment analyses by GSEA and GO
To explore the potential biological functions of SEC61G that promote tumor progression, we divided patients into high- and low-expression groups based on the median SEC61G expression. GSEA analyses showed that high SEC61G expression positively upregulated the signal pathways involving oxidative-phosphorylation, protein export and proteasomes (Figure 4A-C). However, Fc-gamma R-mediated phagocytosis, T cell receptor signaling pathway, B cell receptor signaling pathway, natural killer cell-mediated cytotoxicity, chemokine signaling pathway, and leukocyte trans-endothelial migration were down-regulated (Figure 4D-I).
To further elucidate the biological functions of SEC61G, we analyzed the differentially expressed genes (DEGs) between the DNA amplification and non-amplification groups (Figure 4J). GO analysis showed that 16 biological processes (BP) and 4 molecular functions (MF) were enriched (Figure 4K. Among the 16 BP terms, 10 were associated with immune responses, including "regulation of innate immune response," "negative regulation of immune system process," "natural killer cell-mediated immunity," "regulation of cell activation," "negative regulation of innate immune response," "T cell receptor signaling pathway," "positive regulation of IκB kinase/NF-κB signaling," "regulation of response to cytokine stimulus," "regulation of response to interferon-gamma," and "response to interferon-beta." The remaining BP terms were "defense response to virus," "regulation of viral process," "cellular defense response," "regulation of tissue remodeling," "cytolysis," and "viral entry into host cell." The 4 MF terms were "MHC protein binding," "MHC class I protein complex binding," "endopeptidase activity," and "GTP binding" (Figure 4K).
SEC61G expression negatively correlates with immune infiltration in HNSCC
Considering that the GSEA and GO analyses indicated that genes and terms associated with the immune system were enriched in HNSCC, we further analyzed the correlation between SEC61G expression, CNV, and immune cell infiltration in HNSCC. We observed that SEC61G expression negatively correlated with B cell, CD8+ T cell, CD4+ T cell, macrophage, neutrophil, and dendritic cell infiltration (all p<0.001) (Figure 5A). SEC61G DNA amplification was also significantly negatively correlated with B cell, CD8+ T cell, CD4+ T cell, macrophage, neutrophil, and dendritic cell infiltration (all p<0.05) (Figure 5B).
At the same time, it is considered that SEC61G may participate in endopeptidase activity and MHC class I protein complex binding in GO analyses. Besides, antigen peptide transporter (TAP) is to transport the peptide to the endoplasmic reticulum, and subsequent peptide loading by MHC class I molecules. So, we analyzed the association between TAP1 and TAP2 expression with SEC61G. Results showed that TAP 1 and TAP2 expression was negatively correlated with SEC61G expression in the HNSCC-TCGA datasets (Figure 5C, D).
SEC61G predicts the efficacy of immune checkpoint inhibitors (ICIs)
PD-L1 (CD274) expression plays a vital role in tumor immune escape and is also a predictive marker for therapeutic efficacy of ICIs. The result showed that PD-L1 expression was also negatively correlated with SEC61G expression in the HNSCC-TCGA datasets (Figure 5E). Considering that SEC61G was negatively correlated with immune cell infiltration and PD-L1 expression, we analyzed SEC61G expression during immune checkpoint therapy. A dataset (IMvigor 210 data) generated from metastatic urothelial cancer patients treated with atezolizumab was downloaded. Kaplan-Meier survival analysis showed that patients with high SEC61G expression had a lower OS than patients with low SEC61G expression (p=0.006, Breslow test) (Figure 5F).
SEC61G is upregulated and correlated with adverse outcome in oropharyngeal cancer
To verify the difference of SEC61G expression and its prognostic value in HNSCC, we used 91 OPC cases and 56 normal adjacent tissue samples from our center to detect SEC61G protein expression by IHC. According to the staining intensity, SEC61G expression was divided into negative (Figure 6A and C), weakly (Figure 6B and D), moderately (Figure 6E), and strongly positive (Figure 6F), as shown in figure 6. Among the normal tissue samples, 55 cases (98.2%) were negative, and only 1 case (1.8%) was weakly positive. Among the OPC 91 cases, 44 cases were weakly positive, 9 were moderately positive, and 5 were strongly positive. The overall SEC61G-positive rate was 63.7%, and only 33 cases (36.3%) were SEC61G-negative (Figure 6E). Rank-sum tests showed that SEC61G expression in OPC tissues was higher than expression in normal adjacent tissues (p<0.001, Figure 6E-F). The median follow-up time was 60 months (range 3-131months). Detailed demographic and clinical characteristics are listed in Table 5. Based on SEC61G expression in the tumor tissues, OPC patients were divided into negative and positive expression groups. SEC61G expression correlated to the degree of pathological keratinization, i.e., the proportion of SEC61G-positive samples patients with keratinization was higher than in patients with non-keratinization (p=0.009) (Table 3). Kaplan-Meier survival analyses showed that patients with high SEC61G expression had a lower OS (55.4% vs. 32.4%, p=0.030) and PFS (53.2% vs. 24.1%, p=0.003) than did patients with low SEC61G expression (Figure 6G and H). MAV confirmed that high SEC61G expression was an independent inferior factor for OS (HR=2.49, 95% CI: 1.15-5.30, p=0.020) and PFS (HR=2.82, 95% CI: 1.36-5.84, p=0.005) when adjusted for sex, age, clinical stage, and chemotherapy cycles (Table 4).
The expression and prognostic value of the other subunits of the SEC61 complex
The complex has three subunits, namely α, β, and γ. The expression and prognostic value of SEC61G have been effectively analyzed and verified. We also tried to analyze other subunits of SEC61. The results showed that the expressions of SEC61A1, A2, and B in HNSCC were higher than those in normal control tissues by UALCAN online tool, respectively (Figure 7 A, C, and E, all p<0.001). However, Kaplan-Meier survival analysis showed that these subunits did not significantly affect the prognosis of HNSCC in TCGA data (Figure 7 B, D, and F, all p>0.05).