Autophagy-related gene P4HB: a novel diagnosis and prognosis marker for kidney renal clear cell carcinoma

Abstract

Autophagy can protect cells and organisms from stressors such as nutrient deprivation, and is involved in many pathological processes including human cancer. Therefore, it is necessary to investigate the role of autophagy-related genes (ARGs) in cancer. In this study, we investigated the gene expression of 222 ARGs in 1048 Kidney Renal Clear Cell Carcinoma (KIRC) cases, from 5 independent cohorts. The gene expression of ARGs were first evaluated in the The Cancer Genome Atlas (TCGA) by Recevier Operating Characteristic (ROC) analysis to select potential biomarkers with extremely high ability in KIRC detection (AUC≥0.85 and p<0.0001). Then in silico procedure progressively leads to the selection of two genes in a three rounds of validation performed in four human KIRC-patients datasets including two independent Gene Expression Omnibus (GEO) datasets, Oncomine dataset and Human Protein Atlas dataset. Finally, only P4HB (Prolyl 4-hydroxylase, beta polypeptide) gene was experimentally validated by RT-PCR between control kidney cells and cancer cells. Following univariate and multivariate analyses of TCGA-KIRC clinical data showed that P4HB expression is an independent prognostic indicator of unfavorable overall survival (OS) for KIRC patients. Based on these findings, we proposed that P4HB might be one potential novel KIRC diagnostic and prognostic biomarker at both mRNA and protein levels.

Keywords: KIRC; P4HB; autophagy; diagnostic biomarker; prognostic biomarker.

Figure 1

Procedure for the selection and validation of the diagnosis and prognosis biomarkers in KIRC.

Figure 2

P4HB protein expression was significantly higher in KIRC tissues in comparison with normal tissues, while GABRAPL1 protein expression was significantly lower. Representive IHC images of P4HB (A) and GABRAPL1 (B) in normal (left) and KIRC (middle) tissues. Images were downloaded from HPA Database. Statistical analyses of the protein expression levels of P4HB and GABRAPL1 according to the information of normal and KIRC tissues (right). * p<0.05; ** p<0.01.

Figure 3

ROC analysis of P4HB and GABRAPL1 never related to KIRC diagnosis showing a very high ability to discriminate controls from KIRC samples validated in TCGA (A and B). The AUC is plotted as sensitivity% vs 100-specifificity%. The calculated AUC is reported in each case. The p value is <0.0001 in all cases; Detection of P4HB and GABARAPL1 mRNA expression level in H293T cell (normal kidney), OS-RC-2 and Caki-1 cell (cancer cells) by RT-PCR (C). GAPDH gene was used as the internal control. P4HB expression in cancer cells is clearly higher than normal kidney cell, while GABARAPL1 expression in cancer cells is clearly lower than normal kidney cell.

Figure 4

Kaplan-Meier survival analyses on differential P4HB expression groups with OS in the included 532 KIRC patients (A). The patients were stratified into high and low P4HB groups by quarter (25% upper vs 75% lower). Compared with low mRNA expression of P4HB, high P4HB expressions were significantly correlated with poor OS (p < 0.0001); The interaction network of P4HB protein with other proteins (B). CALR, HSP90B1, HSPA5, PPIB, MTTP, P4HA3, PDIA6, PRKCSH, and CALU physically/functionally connect P4HB. Note: The interaction network was obtained from STRING database.