Comprehensive pan-cancer analysis of heat shock protein 110, 90, 70 and 60 families

Background: Here we carried out a panoramic analysis of the expression and prognosis of HSP110, HSP90, HSP70, and HSP60 families in 33 types of cancer, with the aim of deepening the systematic understanding of heat shock proteins (HSPs) in cancer. Materials and Methods: Next-generation sequencing data of multiple tumors were downloaded from TCGA, CCLE and Oncomine databases. RStudio 3.6.1 was used to analyze HSP110, HSP90, HSP70 and HSP60 families based on their expression in 33 types of cancer. The validations in vivo (stomach adenocarcinoma and colon adenocarcinoma tissues) were performed by qRT-PCR. Results: HSPs were differentially expressed in different cancers. The results revealed mainly positive correlations among the expressions of HSPs in different cancers. Expressions of HSP family members were generally associated with poor prognosis in respiratory, digestive, urinary and reproductive system tumors and associated with good prognosis in cholangiocarcinoma, pheochromocytoma and paraganglioma. TCGA mutation analysis showed that HSP gene mutation rate in cancers was 0-23%. CCLE mutation analysis indicated that HSP gene mutation rate in 828 cell lines from 15 tumors was 0-17%. CNV analysis revealed that HSPs have different degrees of gene amplifications and deletions in cancers. Gene mutations of 15 HSPs influenced their protein expressions in different cancers. Copy number amplifications and deletions of 22 HSPs also impacted protein expression levels in pan-cancer. HSP gene mutation was generally a poor prognosis factor in cancers, except for uterine corpus endometrial carcinoma. CNVs in 14 HSPs showed varying influences on survival status in different cancers. HSPs may be involved in the activation and inhibition of multiple cancer-related pathways. HSP expressions were closely correlated with 22 immune cell infiltrations in different cancers. The qRT-PCR validation results in vivo showed that HSPA2 was down-regulated in stomach adenocarcinoma and colon adenocarcinoma; HSPA7 and HSPA1A also were down-regulated in colon adenocarcinoma. HSPA2-HSPA7 (r = 0.031, p = 0.009) and HSPA1A-HSPA7 (r = 0.516, p < 0.001) were positive correlation in colon adenocarcinoma. Conclusion: These analysis and validation results show that HSP families play an important role in the occurrence and development of various tumors and are potential tumor diagnostic and prognostic biomarkers as well as anti-cancer therapeutic targets.