Frontiers in Oncology (Aug 2023)

Pan-cancer analysis of kinesin family members with potential implications in prognosis and immunological role in human cancer

  • Ming Zhong,
  • Lian Gong,
  • Lian Gong,
  • Na Li,
  • Na Li,
  • Hui Guan,
  • Kai Gong,
  • Yong Zhong,
  • Enyi Zhu,
  • Xiaohua Wang,
  • Shan Jiang,
  • Jinhong Li,
  • Yan Lei,
  • Yu Liu,
  • Jiasi Chen,
  • Zhihua Zheng

DOI
https://doi.org/10.3389/fonc.2023.1179897
Journal volume & issue
Vol. 13

Abstract

Read online

BackgroundKinesin is a molecular motor for transporting “goods” within cells and plays a key role in many types of tumors. The multi-angle study of kinesin at the pan-cancer level is conducive to understanding its role in tumorigenesis and development and clinical treatment potential.MethodsWe evaluated the expression of KIF genes, performed differential analysis by using the R package limma, and explored the pan-cancer prognosis of KIF genes by univariate Cox regression analysis. To evaluate the pan-cancer role of KIF genes as a whole, we defined the KIFscore with the help of gene set variation analysis (GSVA) and explored the KIFscores across normal tissues, tumor cell lines, and 33 tumor types in TCGA. Next, we used spearman correlation analysis to extensively study the correlation between the KIFscore and tumor prognosis and be-tween the KIFscore and clinical indicators. We also identified the relationship between the KIFscore and genomic variation and immune molecular signatures by multiplatform analysis. Finally, we identified the key genes in clear cell renal cell carcinoma (ccRCC) through machine learning algorithms and verified the candidate genes by CCK8, wound healing assay, Transwell assay, and flow cytometry.ResultsIn most cancers, KIFscores are high and they act as a risk factor for cancer. The KIFscore was significantly associated with copy number variation (CNV), tumor mutation burden (TMB), immune subtypes, DNA repair deficiency, and tumor stemness indexes. Moreover, in almost all cancer species, the KIFscore was positively correlated with T cell CD4+ TH2, the common lymphoid pro-genitor, and the T cell follicular helper. In addition, it was negatively correlated with CXCL16, CCL14, TNFSF13, and TNFRSF14 and positively correlated with ULBP1, MICB, and CD276. Machine learning helped us to identify four hub-genes in ccRCC. The suitable gene, KIF14, is highly expressed in ccRCC and promotes tumor cell proliferation, migration, and invasion.ConclusionOur study shows that the KIF genes play an important pan-cancer role and may become a potential new target for a variety of tumor treatments in the future. Furthermore, KIF14, a key molecule in the KIF genes, can provide a new idea for the ccRCC treatment.

Keywords