Identifying tumor immunity-associated molecular features in liver hepatocellular carcinoma by multi-omics analysis

Abstract

Background: Although current immunotherapies have achieved some successes for hepatocellular carcinoma (HCC) patients, their benefits are limited for most HCC patients. Therefore, the identification of biomarkers for promoting immunotherapeutic responses in HCC is urgently needed. Methods: Using the TCGA HCC cohort, we investigated correlations of various molecular features with antitumor immune signatures (CD8⁺ T cell infiltration and cytolytic activity) and an immunosuppressive signature (PD-L1 expression) in HCC. These molecular features included mRNAs, microRNAs (miRNAs), long non-coding RNAs (lncRNAs), proteins, and pathways. Results: We found that the mutations of several oncogenes and tumor suppressor genes significantly correlated with reduced antitumor immune signatures, including TTN, CTNNB1, RB1, ZFHX4, and TP53. It indicates that these genes' mutations may inhibit antitumor immune responses in HCC. Four proteins (Syk, Lck, STAT5, and Caspase-7) had significant positive expression correlations with CD8⁺ T cell enrichment, cytolytic activity, and PD-L1 expression in HCC. It suggests that these proteins' expression could be useful biomarkers for the response to immune checkpoint inhibitors Similiarly, we identified other types of biomarkers potentially useful for predicting the response to ICIs, including miRNAs (hsa-miR-511-5p, 150-3p, 342-3p, 181a-3p, 625-5p, 4772-3p, 155-3p, 142-5p, 142-3p, 155-5p, 625-3p, 1976, 7702), many lncRNAs, and pathways (apoptosis, cytokine-cytokine receptor interaction, Jak-STAT signaling, MAPK signaling, PI3K-AKT signaling, HIF-1 signaling, ECM receptor interaction, focal adhesion, and estrogen signaling). Further, tumor mutation burden showed no significant correlation with antitumor immunity, while tumor aneuploidy levels showed a significant negative correlation with antitumor immunity. Conclusion: The molecular features significantly associated with HCC immunity could be predictive biomarkers for immunotherapeutic responses in HCC patients. They could also be potential intervention targets for boosting antitumor immunity and immunotherapeutic responses in HCC.

Keywords: antitumor immunity; biomarker; cancer immunotherapy; hepatocellular carcinoma; multi-omics analysis.

FIGURE 1

Four genes whose mutations have significant correlations with reduced antitumor immune responses in HCC. (A) Four genes whose mutations are correlated with reduced CD8⁺ T cell enrichment levels or cytolytic activity in HCC. (B) Kaplan-Meier curves showing two of the four genes whose mutations are correlated with better overall survival in HCC.

FIGURE 2

(Continued).

FIGURE 3

Proteins having significant expression correlations with CD8⁺ T cell enrichment, cytolytic activity, or PD-L1 expression in HCC (|r| > 0.3). The Pearson correlation coefficients (r) are shown.

FIGURE 4

MicroRNAs (miRNAs) having significant expression correlations with CD8⁺ T cell enrichment, cytolytic activity, or PD-L1 expression in HCC (|r| > 0.3). The Pearson correlation coefficients (r) are shown.

FIGURE 5

Long non-coding RNAs (lncRNAs) having significant expression correlations with CD8⁺ T cell enrichment, cytolytic activity, or PD-L1 expression in HCC (|r| > 0.3). (A) LINC01871 showing the strongest expression correlations with both CD8⁺ T cell enrichment and cytolytic activity in HCC. The Pearson correlation coefficients (r) and p-values are shown. (B) Heatmap showing 85 lncRNAs with significant positive expression correlations with CD8⁺ T cell enrichment, cytolytic activity, and PD-L1 expression in HCC (r > 0.3).

FIGURE 6

Cancer-associated pathways whose enrichment is significantly correlated with CD8⁺ T cell enrichment, cytolytic activity, or PD-L1 expression in HCC. The cancer-associated pathways having significant positive correlations of their enrichment with CD8⁺ T cell enrichment (A), cytolytic activity (B), and PD-L1 expression (C) in HCC (r > 0.3).

FIGURE 7

Correlations of tumor mutation burden (TMB) and tumor aneuploidy levels (TALs) with CD8⁺ T cell enrichment, cytolytic activity, and PD-L1 expression in HCC. Heatmap showing that TALs have significant negative correlations with CD8⁺ T cell enrichment, cytolytic activity, and PD-L1 expression, and TALs have a significant negative correlation with PD-L1 expression in HCC. The Spearman correlation coefficients (p) and p-values are shown.