RPP30 is a novel diagnostic and prognostic biomarker for gastric cancer

Abstract

Objective: This study aimed to identify the hub gene in gastric cancer (GC) tumorigenesis. A biomarker prediction model was constructed and analyzed, and protein expression in histopathological samples was verified in a validation cohort. Methods: Differentially expressed genes (DEGs) were identified from GC projects in The Cancer Genome Atlas (TCGA) database. Functional enrichment analysis of DEGs was performed between the high- and low- Ribonuclease P protein subunit p30 (RPP30) expression groups. ROC analysis was performed to assess RPP30 expression to discriminate GC from normal tissues. Functional enrichment pathways and immune infiltration of DEGs were analyzed using GSEA and ssGSEA. Survival analysis and nomogram construction were performed to predict patient survival. Immunohistochemical staining of GC tissues was performed to validate RPP30 expression in GC and paracancerous samples. Results: Gene expression data and clinical information of 380 cases (375 GC samples and 32 para-cancerous tissues) were collected from TCGA database. The AUC for RPP30 expression was found to be 0.785. The G alpha S signaling pathway was the most significantly enriched signaling pathway. Primary therapy outcome (p < 0.001, HR = 0.243, 95% CI = 0.156-0.379), age (p = 0.012, HR = 1.748, 95% CI = 1.133-2.698), and RPP30 expression (p < 0.001, HR = 2.069, 95% CI = 1.346-3.181) were identified as independent prognostic factors. As a quantitative approach, a nomogram constructed based on RPP30 expression, age, and primary therapy outcome performed well in predicting patient survival. Nineteen of the 25 tissue samples from the validation cohort showed positive RPP30 expression in GC tissues, whereas 16 cases showed negative RPP30 staining in normal tissues. The difference between the two was statistically significant. Conclusion: High RPP30 expression was significantly correlated with disease progression and poor survival in GC, promoting tumorigenesis and angiogenesis via tRNA dysregulation. This study provides new and promising insights into the molecular pathogenesis of tRNA in GC.

Keywords: RPP30; bioinformatics analysis; early diagnosis; gastric cancer; prognosis.

FIGURE 1

Identification of DEGs between RPP30^high and RPP30^low groups. (A) Volcano plot of DEG profiles between RPP30^high and RPP30^low groups. A total of 233 DEGs were obtained, of which 151 were upregulated and 82 were downregulated. (B) Heatmap of GO analysis showing the co-expression of differential gene profiles in TCGA between RPP30^high and RPP30^low groups. Red indicates upregulated genes; blue indicates downregulated genes; each row indicates each gene expression in different samples, whereas each column indicates the expression of all genes in each sample.

FIGURE 2

Functional enrichment analysis of DEGs between RPP30^high and RPP30^low groups of GC in TCGA. (A) Enriched GO terms in the biological process category. The x-axis represents the proportion of DEGs, and the y-axis represents different categories. In this category, skin development (GO: 0043588), epidermis development (GO: 0008544), epidermal cell differentiation (GO:0009913), and keratinocyte differentiation (GO: 0030216) were primarily enriched. (B) GSEA results showed that the G alpha S signaling pathway was the most enriched. (NES, normalized enriched score; P. adj, adjusted p-value; FDR, false discovery rate; gene sets with P. adj <0.05, FDR q-value < 0.25, and |NES| > 1 are considered as significant).

FIGURE 3

RPP30 expression level was associated with immune infiltration in the GC microenvironment. (A) Correlation between the marker gene of 24 immune cells and RPP30 expression level is shown in the lollipop chart. The size of the dots shows the absolute value of Spearman’s correlation coefficient (r). Larger dots indicate higher correlation coefficients. RPP30 expression was positively correlated with the abundances of immunocytes (Th2 cells, activated dendritic cells, Th1 cells, and helper T cells) and negatively correlated with the presence of Th17 cells. (B) Difference between RPP30^high and RPP30^low groups in terms of Th2 cell infiltration.

FIGURE 4

RPP30 expression between normal and GC tumor samples in TCGA database. (A) RPP30 was significantly upregulated in GC (p < 0.001). (B) ROC analysis of RPP30 expression showed high diagnostic efficiency in discriminating between tumor and normal tissues. The AUC of RPP30 in the diagnosis of GC was 0.785.

FIGURE 5

Relationship between RPP30 expression and T-stage of TCGA in GC. The results showed that RPP30 expression was significantly correlated with T-stage (p = 0.033). Bonferroni correction between each group also showed significance.

FIGURE 6

Validation of RPP30-based nomogram for GC patients. (A) Effect of RPP30 expression on the OS of GC patients in TCGA cohort. The results showed that higher RPP30 expression was associated with poor OS (HR = 1.53 [1.10–2.14], p = 0.012). (B) Nomogram for the prediction of 1-, 3-, and 5-years OS of GC patients. The C-index was 0.704 (95% CI = 0.680–0.728). (C) The calibration plot of the nomogram indicated a good agreement between the prediction and the ideal line.

FIGURE 7

Subgroup survival analysis of clinical characteristics of GC. (A) High RPP30 expression was correlated with worse OS in the T1 and T2 stages of GC, but not in the T3 and T4 stages. In the N stages, high RPP30 expression was associated with worse OS in the N0, but not in the N1 - N3 stages of GC. (B) The K-M plot of OS showed that high RPP30 expression had higher HR value (HR = 2.3, p = 0.033) in subgroups of T1 and T2 stages. (C) The hazard ratio in the high-RPP30 expression group was 0.57 times higher than that in the low expression group (p = 0.009).

FIGURE 8

RPP30 protein expression in GC and para-cancerous tissues assessed via immunohistochemical staining. (A) RPP30 expression was found in GC specimens, especially in the nucleus of the glandular epithelium. (B) RPP30 protein expression was negative in para-cancerous tissues.