酰基甘油激酶在肝癌中的表达水平、空间定位及其与患者预后的相关性分析

Objective To investigate the expression, distribution and cellular localization of acylglycerol kinase (AGK) in liver cancer tissues, and the correlation of AGK expression with the prognosis of liver cancer patients. Methods AGK mRNA expression data and clinical information of hepatocellular carcinoma (HCC) patients were downloaded from The Cancer Genome Atlas (TCGA) database in January 2024. The expression differences of AGK mRNA between HCC tissues and paracancerous tissues were compared, and the high and low expressions of AGK were judged by using the median expression of AGK mRNA in 369 HCC tissues as a cut-off value. A univariate logistic regression model was used to analyze the relationship between clinical pathological characteristics and high expression of AGK. The mRNA expressions in HCC tissues and paracancerous tissues of 9 datasets from the Hepatocellular Carcinoma Molecular Landscape Database (HCCDB) 2.0 were compared. The spatial distribution and cellular localization of AGK were analyzed based on multidimensional data from Bulk transcriptome sequencing (RNA-seq), single-cell sequencing and spatial RNA-seq. The expression of AGK protein in liver cancer tissues was analyzed using the Human Protein Atlas (HPA) database. Kaplan-Meier method and log-rank test were employed to compare the differences in overall survival (OS) among patients with different AGK mRNA expressions in HCCDB25 dataset of HCCDB 2.0 and HPA database. The correlation between expressions of AGK and hepatic stem cell-related markers was analyzed by using Spearman rank test based on Tumor Immune Estimation Resource (TIMER) 2.0 database. Results Data from both the TCGA and 9 datasets of HCCDB 2.0 showed that AGK mRNA expression in HCC tissues was higher than that in paracancerous non-tumorous tissues and normal liver tissues, and the difference was statistically significant (all P < 0.001). HPA database immunohistochemical testing revealed that AGK protein was primarily localized in the cytoplasm, with positive or strong positive expression in HCC tissues and negative or weak positive expression in normal liver tissues; mass spectrometry data showed that it was upregulated in tumor samples (165 cases) compared to normal liver tissues (165 cases) (P < 0.001). Univariate logistic regression analysis indicated that tumor family history and tumor pathological differentiation in HCC patients from TCGA database were associated with high AGK expression in tumor tissues (P values were 0.028 and 0.050), while other factors such as age, gender, body mass index, alpha fetoprotein level, Child-Pugh classification, inflammation degree in paracancerous tissues, Ishak fibrosis score, pathological TNM staging, tumor clinical staging, and tumor vascular infiltration had no impact on AGK expression level in tumor tissues (all P > 0.05). One hundred and fifty-eight patients were divided into high and low KGK mRNA expression groups based on the median expression of AGK mRNA in tissues of HCCDB25 dataset, analysis showed that patients in low AGK mRNA expression group (79 cases) had better overall survival (OS) compared to the high expression group (79 cases) in tumor tissues, and the difference was statistically significant (P = 0.038), while there was no significant difference in OS between high (79 cases) and low (79 cases) expression groups in paracancerous tissues (P = 0.760). In HPA database, patients were divided into high and low AGK mRNA expression groups based on AGK mRNA values in liver cancer tissues corresponding to the lowest P value during OS analysis by Kaplan-Meier method; in all stages of HCC patients, low AGK mRNA expression group (279 cases) had better OS than the high expression group (76 cases), and the difference was statistically significant (P = 0.022). The OS of high AGK mRNA expression group in patients with stages Ⅱ-Ⅲ was worse than that of low expression group, and the difference was statistically significant (P = 0.007). The UMAP plot obtained through dimensionality reduction and cell clustering analysis based on single-cell sequencing data in HCCDB 2.0 revealed that AGK gene expression in liver cancer tissues was primarily distributed in tumor cells, NK/T cells, stromal cells, and myeloid cells. Spatial transcriptomic analysis of tissue samples from 5 HCC patients using HCCDB 2.0 online tools showed that AGK expression varied across different liver cancer tissue regions (non-tumorous tissue, paracancerous tissue, tumor junction, tumor focus, and portal vein tumor thrombus), with 3 cases showing AGK expression enrichment in tumor cells of the tumor junction, tumor focus and portal vein tumor thrombus, while lower in normal hepatocytes, stromal cells and immune cells. In 2 cases, AGK expression was more widespread. Analysis of 3 patients with significant AGK enrichment showed that in HCC samples with complete fibrous capsule, AGK was mainly localized in tumor cells of the tumor focus and junction areas, with weaker expression in paracancerous normal tissues; while in samples with incomplete capsule, high AGK expression was primarily in tumor cells of the tumor junction, tumor focus and portal vein tumor thrombus. TIMER 2.0 database assessment showed that AGK gene expression in 371 patients of TCGA database was positively correlated with the expressions of liver cancer stem cell-related marker genes, including PROM1 (rho = 0.250), TYH1 (rho = 0.188), CD44 (rho = 0.268), ANPEP (rho = 0.171), CD47 (rho = 0.435), EPCAM (rho = 0.246), KRT19 (rho = 0.203), TGFB1 (rho = 0.285), and SOX9 (rho = 0.328) (all P < 0.001). Conclusions AGK expression is significantly upregulated at both mRNA and protein levels in tumor tissues of HCC patients, it predominantly localizes in the tumor tissues and the cytoplasm of tumor cells within the junction areas, and its high expression closely associates with poor prognosis of patients. Its expression is positively correlated with the expression of liver cancer stem cell-related markers.