A novel signature of palmitoylation for predicting prognosis and therapeutic response of hepatocellular carcinoma

Hepatocellular carcinoma (HCC) represents an extremely complex and heterogeneous malignant tumor. Protein palmitoylation, a highly conserved and pivotal form of post-translational protein modification, is extensively implicated in cancer progression and exerts a regulatory role in immune responses. Nevertheless, the prognosis significance and therapeutic potential of palmitoylation-related genes (PRGs) in HCC remain incompletely explored. In the present study, we systematically analyzed multiple transcriptome cohort samples from the TCGA, ICGC and GEO databases, including the TCGA-LIHC, ICGC-LIRI, GSE16757, GSE54236, GSE14520, GSE45267 and GSE36376 cohorts. Subsequently, within the TCGA training cohort, a PRGsSig comprising three hub PRGs, namely HSP90AA1, CTHRC1, and PTDSS2, was constructed via machine learning algorithms. Then, the efficacy of this PRGsSig on prognosis prediction was assessed in the training and validation cohorts. Further analysis of immunotherapy response indicated that patients with low PRGsSig scores benefited more from treatment. Additionally, remarkable disparities were observed between patients in different signature score groups in terms of clinical characteristics, tumor mutation burden, tumor microenvironment, and potential drugs. Furthermore, among the three hub PRGs, PTDSS2 was significantly upregulated in HCC cells and its knockdown significantly inhibited the proliferation and metastasis of HCC cells. In conclusion, we established a robust PRGsSig that offers valuable insights for prognostic prediction and informs treatment strategies in HCC.