Background: The roles of mitochondria and the endoplasmic reticulum (ER) in the progression of hepatocellular carcinoma (HCC) are well established. However, a special domain that regulates the close contact between the ER and mitochondria, known as the mitochondria-associated endoplasmic reticulum membrane (MAM), has not yet been investigated in detail in HCC. Methods: The TCGA-LIHC dataset was only used as a training set. In addition, the ICGC and several GEO datasets were used for validation. Consensus clustering was applied to test the prognostic value of the MAM-associated genes. Then, the MAM score was constructed using the lasso algorithm. In addition, uncertainty of clustering in single-cell RNA-seq data using a gene co-expression network (AUCell) was used for the detection of the MAM scores in various cell types. Then, CellChat analysis was applied for comparing the interaction strength between the different MAM score groups. Further, the tumor microenvironment score (TME score) was calculated to compare the prognostic values, the correlation with the other HCC subtypes, tumor immune infiltration landscape, genomic mutations, and copy number variations (CNV) of different subgroups. Finally, the response to immune therapy and sensitivity to chemotherapy were also determined. Results: First, it was observed that the MAM-associated genes could differentiate the survival rates of HCC. Then, the MAM score was constructed and validated using the TCGA and ICGC datasets, respectively. The AUCell analysis indicated that the MAM score was higher in the malignant cells. In addition, enrichment analysis demonstrated that malignant cells with a high MAM score were positively correlated with energy metabolism pathways. Furthermore, the CellChat analysis indicated that the interaction strength was reinforced between the high-MAM-score malignant cells and T cells. Finally, the TME score was constructed, which demonstrated that the HCC patients with high MAM scores/low TME scores tend to have a worse prognosis and high frequency of genomic mutations, while those with low MAM scores/high TME scores were more likely to have a better response to immune therapy. Conclusions: MAM score is a promising index for determining the need for chemotherapy, which reflects the energy metabolic pathways. A combination of the MAM score and TME score could be a better indicator to predict prognosis and response to immune therapy.
Background: Hepatocellular carcinoma (HCC) is the second most common cancer-associated cause of death globally. It is thus vital that the novel diagnostic and prognostic biomarkers associated with early-stage HCC be identified. Keratin 17 (KRT17) has previously been reported to be associated with certain cancer types. However, its relationship with HCC remains to be defined. Methods: The expression of KRT17 in the TCGA LIHC database and in 44 pairs of samples collected from patients with HCC was assessed using qRT-PCR, WB, and IHC. The prognostic relevance of KRT17 was assessed using Kaplan–Meir curves. The important cancer- and KRT17-related biological processes were defined through gene set enrichment analysis (GSEA). The functional link between KRT17 expression and tumor cell proliferation/survival was assessed using flow cytometry, colony formation assay, CCK-8 assay, and subcutaneous tumor model approaches. Protein–protein interaction (PPI) networks and analyses of immune cell infiltration were also employed to define the signaling pathways associated with KRT17 expression in HCC. Results: HCC tissue samples exhibited increased KRT17 mRNA and protein expression that was predictive of poorer patient survival (P <0.001). GSEA and functional experiments revealed that KRT17 functioned as a regulator of HCC. PPI network analyses also revealed that KRT17 expression was linked to immune cell infiltration and activation in patients with HCC. Conclusions: We found that increased KRT17 levels were associated with poorer survival, more aggressive disease, and altered immune cell infiltration in patients suffering from HCC. KRT17 may function as an oncogene and a prognostic biomarker in this cancer type.