Integrated analysis of histone lysine lactylation (Kla)-specific genes suggests that NR6A1, OSBP2 and UNC119B are novel therapeutic targets for hepatocellular carcinoma

Abstract

Histone lysine lactylation (Kla) plays a vital role in the tumorigenesis of hepatocellular carcinoma (HCC). Hence, we focused on Kla-specific genes to select novel therapeutic targets. Differentially expressed Kla-specific genes (DEKlaGs) were identified from TCGA with the cut-off criteria |log₂(FlodChange (FC))| > 2, p-value < 0.05, following investigating the prognostic value. The correlation between lactate accumulation and prognostic DEKlaGs expression was further investigated. On the other hand, we explored the roles of Kla activation in the immune microenvironment, immunotherapy, and drug resistance. We conducted gene set enrichment analysis (GSEA) to predict the pathways influenced by Kla. The predictive power of Cox model was further identified in ICGC and GEO databases. A total of 129 DEKlaGs were identified, and 32 molecules might be potential prognostic biomarkers. A Cox model including ARHGEF37, MTFR2, NR6A1, NT5DC2, OSBP2, RNASEH2A, SFN, and UNC119B was constructed, which suggested unfavorable overall survival in high-risk score group, and risk score could serve as an indicator for large tumor size, poor pathological grade and advanced stage. NR6A1, OSBP2 and UNC119B could inhibit NK cell as well as TIL cell infiltration, and impair Type-I and II IFN responses in HCC, thereby contributing to unsatisfactory prognosis and immunotherapy resistance. OSBP2 and UNC119B were identified to be related to chemotherapy resistance. GSEA showed that WNT, MTOR, MAPK and NOTCH signaling pathways were activated, indicating that these pathways might play a crucial role during the Kla process. On the other hand, we showed that NR6A1 and OSBP2 were overexpressed in GEO. OSBP2 and UNC119B contributed to poor survival and advanced stage in ICGC. In summary, histone Kla was related to HCC prognosis and might serve as an independent biomarker. NR6A1, OSBP2 and UNC119B were associated with the prognosis, immunotherapy, and chemotherapy resistance, suggesting that NR6A1, OSBP2 and UNC119B might be novel candidate therapeutic targets for HCC.

Figure 1

Histone lysine lactylation (Kla) was related to prognosis of HCC. (A) Volcano map of DEKlaGs in HCC with the cut-off criteria |log₂FC| > 2, p-value < 0.05. (B) Cox regression model formula in line with DEKlaGs. (C) Survival analysis showed lower overall survival in high-risk score group than in low-risk score group. (D) Risk score in accordance with DEKlaGs might be an Indicator for large tumor size, poor pathological grade, and advanced clinical stage. (E) Univariate (left) and multivariate (right) Cox regression analysis based on risk score and clinical information, suggesting that risk score might be an independent biomarker for HCC. (F) Survival analysis along with pathological grade and T-stage. DEKlaGs differentially expressed Kla-specific genes, FC Foldchange, HCC hepatocellular carcinoma, T T-stage in TNM system.

Figure 2

Kla-related genes were related to unfavorable prognosis. (A–D) Lactate accumulation-related genes including EP300, HIF1A, LDHA, and LDHB were overexpressed in HCC tissues. (E) High expression of Kla-related genes was associated with poor overall survival. Although there was no significance in survival analysis, patients with high expression of Kla activation-related genes tended to have an unsatisfactory prognosis. (F) Activation of Kla could induce prognosis-associated molecule expression.

Figure 3

Kla was related to HCC immune infiltration. (A) High risk score was positively related to cancer progression-associated immune cells, especially T cells regulatory (Tregs). (B) Activation of Kla inhibited immune processes, including Type-I-IFN-Response and Type-II-IFN-Response. (C) Heatmap of the correlation between prognosis-associated DEKlaGs and infiltration level of each immune cell.

Figure 4

NR6A1, OSBP2 and UNC119B promoted immunotherapy resistance of HCC. (A–C) NR6A1, OSBP2 and UNC119B were negatively associated with immune cell proportion score (IPS) including ips-ctla4-neg-pd1-neg, ips-ctla4-neg-pd1-pos, ips-ctla4-pos-pd1-neg, and ips-ctla4-pos-pd1-pos, suggesting that NR6A1, OSBP2 and UNC119B contributed to immunotherapy resistance of HCC patients. (D) Expression of NR6A1, OSBP2 and UNC119B was positively related to various immune checkpoints, such as CD40, CD276, CALT4, TNFSF4. These observations showed that NR6A1, OSBP2 and UNC119B might be crucial immunotherapeutic targets for HCC.

Figure 5

OSBP2 AND UNC119B enhanced chemotherapy resistance of HCC. (A) NR6A1 was positively related to drug susceptibility. (B) OSBP2 promoted Selumetinib, Vemurafenib, and Dabrafenib resistance of HCC patients. (C) UNC119B decreased the chemotherapy response to AP-26113, Alectinib and Carfilzomib.

Figure 6

Kla induced activation of WNT, MAPK, NOTCH, MTOR signaling. (A) Overexpression of NR6A1 was related to activation of WNT, MAPK and MTOR signaling pathways. (B) OSBP2 induced the activation of WNT, MAPK and NOTCH. (C) Upregulated UNC119B promoted WNT, MAPK, MTOR signaling activity. The horizontal axis indicates a series of genes, while the vertical axis indicates the corresponding running ES. The peak of the curve is the ES. The black vertical lines represent the target genes in the gene set. The red meant bigger logFC, while blue is opposite. ES enrichment score.

Figure 7

NR6A1, OSBP2 and UNC119B promoted HCC progression. (A) High NR6A1, OSBP2 and UNC119B expression contributed to HCC proliferation. (B) NR6A1 and UNC119B were related to the poor pathological grade of HCC. (C) OSBP2 and UNC119B might be two advanced-stage biomarkers for HCC. These observations indicating that NR6A1, OSBP2 and UNC119B played a crucial role in HCC progression, which shows promise to be three therapeutic targets.

Figure 8

Identification of prognosis-associated DEKlaGs and Cox model in GEO and ICGC. (A) NR6A1 and OSBP2 expression in GSE25097 and GSE54236. UNC119B failed to search from GEO series, which showed that NR6A1 and OSBP2 were overexpressed in HCC cases. (B) Identification of Cox model in ICGC by survival and stage correlation analysis, suggesting that the risk formular according to prognosis-associated DEKlaGs could predict the prognosis of HCC patients accurately. (C,D) OSBP2 and UNC119B were related to poor survival and advanced stage in ICGC. (E) NR6A1 showed no difference in ICGC database. These findings indicated that OSBP2 and UNC119B might serve as two potential prognostic biomarkers for HCC.