Integration of genomics and transcriptomics highlights the crucial role of chromosome 5 open reading frame 34 in various human malignancies

C5orf34 expression analysis

ONCOMINE (http://www.oncomine.org/) is a robust genome-wide expression analysis service provided by a translational bioinformatics service [8]. The information was gathered to probe into how the expression level of C5orf34 differed among cancer types. The significance levels in this investigation were established at a p-value of 0.05, a gene rank in the top 10%, and a fold change of 2.

The Cancer Genome Atlas (TCGA) was used for the purpose of collecting profile data of expression from related cancers and adjoining normal samples, together with information outlining related clinicopathological features, from a database that is of landmark cancer genomic comprising huge cancer samples information covering 33 cancer types [9]. The abbreviations of the TGCA pan-cancer names were shown in Supplementary Table 1. GTEx [10], a large public database, was used to enrich tissue-specific normal samples in addition to those obtained from TCGA. For the purpose of performing pan-cancer analysis, we used software to extract and combine C5orf34 expression levels. The “wilcox.test” algorithm in R was used to investigate the differences in mRNA expression between cancer types in depth [11]. We also integrated TCGA database and GTEx database to further study the differences between normal tissues and cancer tissues using the R-package “ggpubr”.

Immunohistochemical (IHC) analysis

At the Ningde Municipal Hospital of Ningde Normal University in China, we collected paracancerous tissues and tumor tissues from CESC, UCS, UCEC, and OV, and 10 tumor type pairs in each were collected to validate the difference of expression of C5orf34. The Human Ethics Committee at Ningde Municipal Hospital of Ningde Normal University, China, granted the experimental procedure, and the approval reference number is 20220203. After deparaffinization, epitope retrieval, and hydration, 3% hydrogen peroxide was used to inhibit endogenous peroxidase activity in the slices for 15 minutes. The slides were then placed in a humidity box and treated overnight at 4°C with the primary antibody C5orf34 (1:400, Novus, USA), followed by placing in a secondary antibody. Finally, diaminobenzidine was used to view the slides, which were then counterstained with hematoxylin. An Olympus BX63 microscope was utilized to examine immunohistochemical sections, and Image J software was employed for quantitative analysis of slides. The mean and standard deviation were adopted to represent the data. The SPSS 21.0 statistical program was used to conduct the statistical analysis.

C5orf34 expression correlation to clinical phenotype and prognosis

Similar to previous study [12], we conducted an analysis using survival data sourced from the TCGA database to explore the relationship between clinical outcomes and the expression of C5orf34. The association between mRNA expression levels and patient survival rates was evaluated using disease-free interval (DFI), disease-specific survival (DSS), overall survival (OS), and progression-free interval (PFI). Kaplan-Meier and Cox analyses were performed for the purpose of investigating the correlation between C5orf34 expression and survival outcomes in patients with various cancer types. The R software (version: 4.0.3) was used to plot KM curves and forest plots. Following that, we performed a correlation study of clinicopathological parameters such as tumor grade, tumor stage, gender, age, race, and tumor status with the aid of R’s “limma” and “ggpubr” packages. The R software package “survminer” and “survival” were used to create the survival curves. The significance level was set at p < 0.05.

The C5orf34 methylation and prognosis correlation

DNA methylation is widely established to play a significant role in a variety of malignancies. The relationship between patient prognosis and C5orf34 methylation in different malignancies was investigated in this study. The methylation data were acquired in its raw form from the TCGA database. The log-rank test and Kaplan-Meier tests were conducted to examine the relationship between survival outcome and C5orf34 methylation in patients with various cancer types.

C5orf34 mutation analysis in a variety of cancers

In clinical practice, gene mutations, including gene single-nucleotide variant (SNV) and copy number variation (CNV), are common in almost all cancers. Thus, we performed C5orf34 mutation analysis in an open platform that visualizes and analyses gene copy number variations and mutations in a variety of cancers known as CBio Cancer Genome Portal [13]. Then, with the aid of the R packages “survminer” and “survival,” the correlation between patient prognosis and CNV was explored by log-rank tests and Kaplan-Meier analysis. The SNV is a single nucleotide at a specific chromosomal location. It is the most prevalent sort of genetic variation, and it has been proven in a variety of genomic locations for a long time [14]. Another study on the correlation between prognosis and SNV was conducted using a similar manner.

Association of immunity with C5orf34 expression in pan-cancers

To begin, Sangerbox website [15], a useful online platform for TCGA data analysis, was used to investigate the relationship between immune-related genes and C5orf34 expressions, such as immune inhibitors, chemokine ligands, chemokine receptors, immunostimulators, and genes encoding major histocompatibility complex genes. The “limma” package in R was used to perform the co-expression analysis of C5orf34 and immune-related genes, and the results were visualized using the “RColorBrewer” and “reshape2” packages in R software. Following that, the CIBERSORT-ABS, QUANTISEQ, TIMER, EPIC, MCPCOUNTER, CIBERSORT, and XCELL algorithms were used to determine the relative proportion of certain immune cells [16]. The close association of immune cells infiltration with C5orf34 expression was then investigated thoroughly.

Tumor mutation burden (TMB), microsatellite instability (MSI), and C5orf34 gene expression correlation analysis

To further investigate the immune-predictor significance of C5orf34 in pan-cancer, we investigated two new immunotherapy parameters: MSI and TMB. We hypothesized that C5orf34 expression in relation to TMB and MSI would indicate that a greater degree of genomic instability, as described by the MSI and TMB, may result in heightened immune-surveillance possibilities.

Pan-cancer stemness indices and tumor microenvironment

Tumor cells, extracellular matrix, and mesenchymal cells make up the majority of the tumor microenvironment. These play a crucial role in the onset, progression, invasion, and metastasis of tumors [17]. The correlation between the proportion of stromal cells and immune cells and C5orf34 expression in TCGA tumor samples cells was investigated by means of the ESTIMATE method. The ESTIMATE score is based on gene expression features, which can indicate tumor purity; it also shows high accuracy in the estimate [18]. An analysis of the Spearman correlation was conducted between the C5orf34 expression level and the matrix score with the aid of the estimate and limma packages.

Melta used an approach known as the one-class logistic regression (OCLR) machine learning to identify epigenetic and transcriptomic feature sets from non-transformed pluripotent stem cells together with their differentiated posterity in a paper published in the journal “Cell” in 2018 [19]. The RNA stemness scores (RNAss) and DNA stemness scores (DNAss) were calculated based on epigenetic signatures and OCLR-based transcriptomic markers on all pan-cancer 33 TCGA cohorts, which were also highly correlated with immune checkpoint activation, immunocyte infiltration, and tumor heterogeneity. As a result, we investigated the correlation between DNAss/RNAss and C5orf34 expression in a comprehensive manner.

Prediction of targeted miRNAs of C5orf34 and single-cell analysis of C5orf34 in female reproductive system tumors

The miRNet [20], miRTarBase [21], and StarBase [22] databases were employed to predict the miRNAs closely related to C5orf34. TISCH2 [23], short for the Tumor Immune Single Cell Hub 2, is a valuable repository of single-cell RNA sequencing data obtained from both human and mouse tumors. This resource facilitates a thorough analysis of gene expression within the tumor microenvironment (TME) across a wide spectrum of cancer types. GSE118828 [24] provided single-cell data of OV, GSE139555 [25] provided single-cell data of UCEC, and GSE168652 provided single-cell data of CESC. In the TISCH2 platform, we conducted single-cell type analysis for OV, UCEC, and CESC. More importantly, we analyzed the single-cell distribution characteristics of C5orf34 in the aforementioned female reproductive system tumors.

C5orf34 expression in human cancers at differential levels

We used the Oncomine database for the purpose of examining C5orf34 expression in cancer tissues and normal tissues and discovered that its expression was notably higher in cancers than in healthy tissues (Figure 1A). C5orf34, on the other hand, was shown to be less expressed in several malignancies, including leukemia, prostate cancer, lymphoma, breast cancer, ovarian cancer, and central nervous system cancer (Figure 1A).

Figure 1. Comparison of C5orf34 expression level in several malignancies and its matching normal tissues based on (A) Oncomine platform, (B) TCGA cohort, (C) TCGA and GTEx cohorts. (D) Representative images and quantification analysis of C5orf34 immunohistochemistry in gynecologic cancers (^***P < 0.001, ^**P < 0.01, ^*P < 0.05).

We then utilized the TCGA database to validate our findings. Similarly, C5orf34 mRNA level was considerably elevated in most malignancies such as CHOL, LIHC, ESCA, BRCA, READ, COAD, LUSC, STAD, LUAD, HNSC, BLCA, KIRC, KIRP, GBM, and UCEC in contrast with matched adjacent normal tissues (Figure 1B). In addition to THCA and PRAD, C5orf34 expression was found to be significantly elevated in the majority of cancers.

Due to the limited amount of paracancerous samples in the TCGA database, we combined data from TCGA and GTEx platforms to investigate the differential expression of C5orf34 in normal and cancerous tissues. In most cancers, the count of C5orf34 in tumor tissues was much greater than in normal tissues, which was in line with our earlier findings. C5orf34 expression level was also considerably lower in PRAD than in normal prostate tissues; however, there was no major difference between the expression of C5orf34 in normal thyroid tissues and THCA. Furthermore, in the case when normal tissue samples were added to LAML populations, the expression level of C5orf34 was lowered (Figure 1C). IHC results confirmed that OV, UCEC, CESC, and UCS tissues had higher C5orf34 expression levels than normal tissues, as shown in Figure 1D.

C5orf34’s multifaceted prognostic value in human cancers

With the aid of the data from TCGA database, the link between C5orf34 expression level and pan-cancer prognostic value was then explored. High mRNA expression was determined to be highly essential factor by the univariate Cox regression analysis of C5orf34-related survival in LGG, CESC, KIRP, LUAD, KIRC, THCA, LIHC, SARC, UCEC, ACC, KICH, THYM, PCPG, MESO and UVM (Figures 2A, 3A, 4A, 5A). Of note, high expression levels of C5orf34 led to a prolonged survival duration of patients with THYM (OS: p = 0.028), which was in line with the earlier findings from a Cox regression analysis (Figure 2A).

Figure 2. Relationship of C5orf34 expression with the overall survival in pan-cancer. (A) The forest plots of C5orf34 in pan-cancer’s overall survival; (B–L) C5orf34’s survival curves as regards overall survival of pan-cancer samples.

Figure 3. Relationship of C5orf34 expression with disease-specific survival in pan-cancer. (A) The forest plots of C5orf34 in pan-cancer’s disease-specific survival; (B–K) C5orf34’s survival curves as regards disease-specific survival of pan-cancer samples.

Figure 4. Relationship of C5orf34 expression with disease-free interval in pan-cancer. (A) The forest plots of C5orf34 in pan-cancer’s disease-free interval; (B–F) C5orf34’s survival curves as regards disease-free interval of pan-cancer samples.

Figure 5. Relationship of C5orf34 expression and progression-free interval in pan-cancer. (A) The forest plots of C5orf34 in pan-cancer’s progression-free interval; (B–K) C5orf34’s survival curves as regards progression-free interval of pan-cancer samples.

The Kaplan-Meier cumulative curves indicated elevated C5orf34 expression was significantly related to the worse OS, DSS and PFI of UVM, LIHC, UCEC, LGG, KIRC, KICH, and ACC (Figures 2B–2L; 3B–3K; 5B–5K). In addition, elevated C5orf34 expression was significantly related to the worse DFI of UCEC, THCA, SARC, LUAD, and LIHC (Figure 4B–4F).

Relationship between clinical characteristics and C5orf34 expression in several cancers

Subsequently, we studied the correlation between mRNA expression and clinical features in multiple malignancies. LIHC, BRCA, THYM, HNSC, LUAD, CHOL, ESCA, and UCEC all exhibited substantial differences in C5orf34 expression in various age groups. In LIHC, BRCA, HNSC, LUAD, CHOL, ESCA, and THYM, decreased C5orf34 expression level was observed in patients more than 60 years old (Supplementary Figure 1). Furthermore, in the following six malignancies, the expression level of C5orf34 differed considerably by gender. In MESO, THCA, READ, KIRC, and HNSC, C5orf34 expression level was greater in the male group, but C5orf34 expression level was lower in the SARC male group (Supplementary Figure 1). Furthermore, C5orf34 expression differed by race in 5 cancers types including LIHC, UCEC, BRCA, ESCA, and ACC. Except for ACC, non-white patients had higher expression levels of C5orf34 (Supplementary Figure 1).

In terms of the differences in C5orf34 expression by grade, it was discovered that C5orf34 was expressed more in G3-G4 patients in HNSC, ESCA, LIHC, LGG, UCEC, and CESC, and lower in patients of G3-G4 in STAD (Figure 6). In HNSC, UCS, LIHC, THCA, UCEC, and ACC, C5orf34 expression level appears to be greater in stage III-IV patients than in stage I and II patients, whereas C5orf34 expression level appears to be higher in stage I-II patients in THYM (Figure 6). Moreover, recurrent patients had a higher expression level of C5orf34 in KICH, LGG, PCPG, ACC, KIRC, UCEC, THCA, and UVM, as shown in Figure 6.

Figure 6. Relationship of C5orf34 expression with tumor stage, tumor grade, and tumor status.

Association between prognosis and C5orf34 methylation in several cancers

We explored the potential relationship between C5orf34 methylation and numerous human malignancies after collecting data on C5orf34 methylation from the TCGA. With the exception of OV and PAAD, the majority of cancer tissues had lower C5orf34 methylation levels than normal tissues (Figure 7A). Furthermore, increased C5orf34 methylation levels were correlated with a favorable patient prognosis in several cancer patients. Patients with higher C5orf34 methylation levels had better DFI, PFS, and DSS in CHOL (Figure 7B–7D); on the other hand, patients with higher C5orf34 methylation levels had better DSS and OS in GBM (Figure 7E, 7F); patients with higher C5orf34 methylation levels had better DSS and OS in LIHC (Figure 7G, 7H); Patients with elevated C5orf34 methylation levels had prolonged PFS and OS duration in MESO (Figure 7I, 7J);, patients with elevated C5orf34 methylation levels had better DFI status in PRAD (Figure 7K);, patients with higher C5orf34 methylation had better PFS, DSS, and OS in UCEC (Figure 7L–7N); Patients with elevated C5orf34 methylation levels in UVM had prolonged OS, PFS, and DSS duration (Figure 7O–7Q).

Figure 7. C5orf34 methylation landscape on TCGA cancer types. (A) The methylation level of C5orf34 differs dramatically between normal and malignant tissues; (B–Q) In UCEC, LIHC, PRAD, MESO, CHOL, GBM, and UVM, there was a correlation between C5orf34 methylation and patient prognosis.

Copy number variation in C5orf34 in a variety of cancers

CNV of critical genes has long been thought to play a role in the onset, progression, and patient prognosis of many malignancies. With the aid of the cBioPortal database, we explored the copy number variations of C5orf34. We discovered that the frequency of alteration was greater in esophageal adenocarcinoma and lung squamous cell carcinoma than in other malignancies. Amplification was the most common form of variation, followed by deep deletion and mutation (Figure 8A). The C5orf34 mutation count in a variety of malignancies was also presented in Figure 8B. Subsequently, the C5orf34 mutation types, locations, and case numbers were explored using the cBioPortal database. C5orf34 has 29 mutation locations (1 truncating and 28 missense) between amino acids 0 and 638 (Figure 8C).

Figure 8. C5orf34 mutations and copy number alterations. (A) The cBioPortal database shows the mutation level of C5orf34; (B) The cBioPortal database shows the incidence of C5orf34 mutations in several TCGA pan-cancer investigations; (C) From the cBioPortal database, a mutation diagram of C5orf34 in several cancer types throughout protein domains.

Relationship between C5orf34 variants and prognosis in several cancers

The potential relationship between pan-cancer prognosis and C5orf34 variation was then further investigated. As depicted in Figure 9, ACC patients with amplification of C5orf34 exhibited a more favorable DFI status compared to those with the wild-type C5orf34. Additionally, they also demonstrated a superior PFS when contrasted with patients having C5orf34 deletion mutations or the wild-type variant. BLCA patients with deletion mutation had a shortened PFS and DFI duration in comparison to those with C5orf34 wide type or amplification. In comparison to patients with the wild-type C5orf34, individuals suffering from CESC and UCEC, who harbored C5orf34 deletion or amplification mutations, experienced a significantly reduced DFI duration. Patients with KICH who had C5orf34 deletion mutations exhibited an extended duration of PFS, OS, and DSS. In contrast, those who underwent C5orf34 amplification showed a poorer prognosis. In contrast to KIRP patients with C5orf34 amplification or the wild-type, those with C5orf34 deletion mutations experienced a significantly reduced duration of PFS, OS, and DSS. Patients with LGG who exhibited amplification of C5orf34 experienced a more unfavorable OS and DSS outcome. In LIHC patients, those with C5orf34 amplification, as well as those with deletion mutations, showed similar DSS and OS, both of which were worse when compared to individuals with the wild-type C5orf34. In the case of LUAD patients, those who had deletion mutations and amplification of C5orf34 experienced significantly improved PFS compared to individuals with the wild-type C5orf34. THCA patients with amplification of C5orf34 exhibited a more unfavorable prognosis, characterized by worse PFS, DSS, and DFI outcomes. UCEC patients with C5orf34 amplification or deletion mutations exhibited poorer survival outcomes compared to those with the wild-type C5orf34. For SARC, STAD, PAAD, PRAD, and THYM, patients who experienced amplification together with deletion mutation had worse prognosis in comparison to those with C5orf34 that is of wild type.

Figure 9. Relationship of C5orf34 copy number variation with prognosis in ACC, BLCA, CESC, KICH, KIRP, LGG, LIHC, LUAD, PAAD, PRAD, SARC, THYM, UCEC, STAD, and THCA.

Furthermore, our study revealed the relevance of C5orf34 SNV to the clinical outcomes of several malignancies (Supplementary Figure 2). In the case of UCEC, mutant C5orf34 led to a longer OS and PFS duration compared to the wild-type C5orf34. Conversely, in CESC, mutant C5orf34 resulted in a poorer OS, DSS, and PFS status when compared to the wild-type C5orf34.

Co-expression of C5orf34 with immune-related genes across various cancer types

The immune microenvironment within tumors is widely recognized for its complexity, and immune-related genes play a pivotal role as key regulators of this environment. In order to elucidate C5orf34’s immunomodulatory influence in the tumor microenvironment, we investigate the co-expression connection between immune-related genes and C5orf34 (including chemokine ligands, immune inhibitors, MHC, and chemokine receptors, and immunostimulators) in pan-cancer. Most immune-related genes were related to the expression of C5orf34 in the majority of malignancies, as illustrated in Figure 10. In the vast majority of malignancies, there was a positive correlation between the immune-related genes and C5orf34 expression. Notably, the expression level of C5orf34 exhibited a negative correlation with immune-related genes in THYM, TGCT, and GBM.

Figure 10. Relationship of C5orf34 expression with immune-related genes (including immunostimulators, chemokine receptors, chemokine ligands, immunoinhibitors, and genes encoding major histocompatibility complex).

Relationship between C5orf34 and immune cells infiltration in pan-cancers

In addition to the aforementioned immune-related genes, the infiltration of immune cells plays a pivotal role in the development and progression of malignant cancers. As a result, we investigated the interaction between infiltration of immune cells and expression of C5orf34 (Figure 11A). C5orf34 was shown to have a positive correlation with Th2 cells positively in a majority of malignancies, including SKCM, KICH, BLCA, ACC, and LIHC.

Figure 11. Relationship of C5orf34 expression wtih (A) immunocyte infiltration, (B) tumor mutation burden, and (C) microsatellite instability in pan-cancer.

In THYM, C5orf34 exhibited a significant positive correlation with the majority of CD8+ T cells and CD4+ T cells, while showing a strong negative correlation with macrophages, including both M1 and M2 subtypes. In many cancers, including THYM, STAD, UVM, and OV, T cell follicular helper cells displayed a positive correlation with C5orf34, except in the case of DLBC, where a negative correlation was observed. In THYM and LIHC, there was a favorable correlation between C5orf34 and T cell regulatory (Tregs). In BRCA, regarding the correlation with Tregs, a positive association was noted in BRCA-LumA, while a negative correlation was observed in BRCA-LumB and BRCA-Her2.

The expression of C5orf34 was positively correlated with TMB in SARC, ACC, STAD, KICH, PAAD, BRCA, THCA, LUAD, LGG, and BLCA. Nonetheless, an inverse association with TMB was observed in the THYM cohorts (Figure 11B). Among the READ, THCA, STAD, SARC, ACC, HNSC, UCEC, and BLCA cohorts, tumors with a high Microsatellite Instability (MSI-High) status exhibited higher C5orf34 expression levels compared to those with genetic stability. However, the inverse trend was only observed in the COAD cohorts (Figure 11C). Despite the strength of these associations, the correlation coefficients between TMB and C5orf34, together with MSI, were below 0.6 in nearly all malignancies, suggesting that C5orf34 was not expected to affect tumorigenesis by playing a role in the genetic alterations process and that it was not adequate to predict the patient’s response to immune-checkpoint blockade (ICB) either by its own.

Tumor microenvironment and stemness indices in pan-cancers

To thoroughly comprehend the relationship between C5orf34 and pan-cancers stemness, researchers employed DNAss and RNAss. C5orf34 exhibited a high positive correlation with both RNAss and DNAss in GBM, BRAC, LGG, SKCM, LUSC, STAD, LUAD, and TGCT (all p < 0.01). Notably, C5orf34 was found to have a negative correlation with RNAss in PRAD and KIRP. Additionally, it displayed a negative correlation with DNAss in THYM. (Figure 12). The ESTIMATE technique was used to calculate ESTIMATEScore, ImmuneScore, and StromalScore for the purpose of performing further investigations of the tumor microenvironment. Figure 12 shows the relationship between the ratio of stromal and immune cells and C5orf34 expression in TCGA tumor samples. C5orf34 exhibited an adverse association with the estimated score, immune score, and stromal score in STAD, PCPG, TCGT, SARC, SKCM, LUAD, LUSC, COAD, GBM, KIRP, ACC, and UCEC. C5orf34 showed a favorable relationship with the estimated score, immune score, and stromal score in PRAD and THCA.

Figure 12. Relationship of C5orf34 expression with stromal scores, RNA stemness scores, DNA stemness scores, immune scores, and ESTIMATE scores in pan-cancer.

Single-cell analysis of C5orf34 in female reproductive system tumors

With the help of miRNet, miRTarBase, and StarBase databases, a total of 47 miRNAs were found to be closely related to C5orf34, which established the framework for future study (Supplementary Tables 2–4). The single-cell results of GSE118828 showed that OV micro-environment was composed of conventional CD4+ T cells, endothelial, fibroblasts, malignant cells, monocyte/macrophage, and myofibroblasts (Supplementary Figure 3A). The expression of C5orf34 is predominantly observed in certain immune cells of OV, such as conventional CD4 T cells and monocyte/macrophage (Supplementary Figure 3A). The single-cell results of UCEC based on GSE139555 revealed that UCEC micro-environment was composed of conventional CD4+ T cells, CD8+ T cells, exhausted CD8+ T cells, fibroblasts, Tprolif, and Treg cells (Supplementary Figure 3B). C5orf34 expression is relatively uniform across various types of immune cells in UCEC (Supplementary Figure 3B). The single-cell results of GSE168652 showed that CESC micro-environment was composed of CD8+ T cells, stromal cells, endothelial, fibroblasts, malignant cells, monocyte/macrophage, and smooth muscle cell (Supplementary Figure 3C). Similar to UCEC, various cell types in CESC exhibit similar C5orf34 expression patterns (Supplementary Figure 3C).