Endoplasmic reticulum stress promotes hepatocellular carcinoma by modulating immunity: a study based on artificial neural networks and single-cell sequencing

Abstract

Introduction: Hepatocellular carcinoma (HCC) is characterized by the complex pathogenesis, limited therapeutic methods, and poor prognosis. Endoplasmic reticulum stress (ERS) plays an important role in the development of HCC, therefore, we still need further study of molecular mechanism of HCC and ERS for early diagnosis and promising treatment targets.

Method: The GEO datasets (GSE25097, GSE62232, and GSE65372) were integrated to identify differentially expressed genes related to HCC (ERSRGs). Random Forest (RF) and Support Vector Machine (SVM) machine learning techniques were applied to screen ERSRGs associated with endoplasmic reticulum stress, and an artificial neural network (ANN) diagnostic prediction model was constructed. The ESTIMATE algorithm was utilized to analyze the correlation between ERSRGs and the immune microenvironment. The potential therapeutic agents for ERSRGs were explored using the Drug Signature Database (DSigDB). The immunological landscape of the ERSRGs central gene PPP1R16A was assessed through single-cell sequencing and cell communication, and its biological function was validated using cytological experiments.

Results: An ANN related to the ERS model was constructed based on SRPX, THBS4, CTH, PPP1R16A, CLGN, and THBS1. The area under the curve (AUC) of the model in the training set was 0.979, and the AUC values in three validation sets were 0.958, 0.936, and 0.970, respectively, indicating high reliability and effectiveness. Spearman correlation analysis suggests that the expression levels of ERSRGs are significantly correlated with immune cell infiltration and immune-related pathways, indicating their potential as important targets for immunotherapy. Mometasone was predicted to be the most promising treatment drug based on its highest binding score. Among the six ERSRGs, PPP1R16A had the highest mutation rate, predominantly copy number mutations, which may be the core gene of the ERSRGs model. Single-cell analysis and cell communication indicated that PPP1R16A is predominantly distributed in liver malignant parenchymal cells and may reshape the tumor microenvironment by enhancing macrophage migration inhibitory factor (MIF)/CD74 + CXCR4 signaling pathways. Functional experiments revealed that after siRNA knockdown, the expression of PPP1R16A was downregulated, which inhibited the proliferation, migration, and invasion capabilities of HCCLM3 and Hep3B cells in vitro.

Conclusion: The consensus of various machine learning algorithms and artificial intelligence neural networks has established a novel predictive model for the diagnosis of liver cancer associated with ERS. This study offers a new direction for the diagnosis and treatment of HCC.

Keywords: Artificial neural network; Endoplasmic reticulum stress; Hepatocellular carcinoma; Immune infiltration; Prognosis.

Fig. 1

The overall flow of this study

Fig. 2

Differential gene expression analysis between liver cancer and normal tissues.(A) Principal component analysis (PCA) of genes without batch removal for datasets including GSE25097, GSE62232, and GSE65372. (B) PCA of genes with batch removal for datasets including GSE25097, GSE62232, and GSE65372. (C) A volcano plot representing 763 differentially expressed genes (DEGs) between liver cancer tissues and normal tissues. (D) A heatmap showing the 763 DEGs between HCC and normal tissues

Fig. 3

Functional enrichment of DEGs between liver cancer tissues and normal tissues.(A-C) Gene ontology (GO) analysis of DEGs, including molecular function (MF), cellular component (CC), and biological process. (D) Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis of DEGs. (E) Gene Set Variation Analysis (GSVA) of DEGs.

Fig. 4

Identification of ERS-related differentially expressed genes. (A)ERS-related genes (n = 312) were cross-referenced with DEGs (n = 763) to identify ERS-related differentially expressed genes (ERSRGs) (n = 11). (B) Residual plot for the selection of ERSRGs using the Random Forest (RF) algorithm. (C) The 11 selected ERSRGs are arranged in descending order according to the Gini coefficient. (D) Selection of ERSRGs using the Support Vector Machine (SVM) algorithm. (E) Identification of 6 ERSRGs through the intersection of genes selected by RF and SVM.

Fig. 5

Construction and Validation of the Artificial Neural Network (ANN) Prediction Model for Liver Cancer. (A) The ANN comprises 6 neurons as the input layer, 2 neurons as the output layer, and 5 neurons as the hidden layer. (B) The AUC value of the ANN prediction model in the training set. (C-E) The AUC values of the ANN prediction model in the validation groups

Fig. 6

Expression and Immune infiltration Analysis of Six ERSRGs. (A) The expression of six ERSRGs between HCC and normal tissues. (B) The Co-mutation analysis of six ERSRGs. (C) The relationship of six ERSRGs expression levels with the immune cell infiltration and immune-related pathway. (D)Correlation between the expression levels of 6 ERSRGs and immune-related molecules

Fig. 7

Predict the top 20 candidate drugs for endoplasmic reticulum stress-related genes based on PubChem. (A) Predict the top 20 most significant candidate compounds for ERSRGs using the DSigDB database. (B-G) Molecular docking between THBS1 and Mometasone

Fig. 8

Survival, mutation and methylation Analysis Related to ERSRGs. (A-F) Kaplan-Meier curves representing the differences in overall survival between groups with high and low expression levels of six ERSRGs. (G) Comparison of gene mutations and copy number variations among the six ERSRGs. (H-M) Calculation of methylation levels of six ERSRGs genes

Fig. 9

Single-cell RNA-seq Data Analysis (GSE149613). (A) UMAP plot of annotated cell types. (B) Bubble chart showing markers corresponding to different cell types. (C) UMAP diagram shows the expression of ppp1R16A (D) Go enrichment analysis, including BP, CC, MF. (E)KEGG enrichment analysis; The color close to blue indicates a smaller p value, and the larger bubble table indicates that more differential genes are enriched in this pathway.(F)Analysis of downstream signaling pathways of GSEA at the single-cell level

Fig. 10

Inference of cell–cell communications in TME. (A-B) A cell–cell communications between the identified cell types. (C) The incoming and outgoing signaling pathways of each cell type. (D-G) The hierarchical diagram displays the specific interaction between MIF and COLLAGE pathways

Fig. 11

Validation of PPP1R16A Expression Levels and Knockout of PPP1R16A Significantly Inhibits Proliferation, Invasion, and Migration Capabilities of HCC (A)Expression levels of PPP1R16A mRNA in HCC cell lines. (B-C) Knockout efficiency of PPP1R16A mRNA in Hep3B and HCCLM3 cells. (D) CCK8 assays indicate that the knockdown of PPP1R16A inhibits the proliferation abilities of Hep3B and HCCLM3 cells. (E-F) Knockout of PPP1R16A inhibits the migration and invasion capabilities of Hep3B and HCCLM3 cells (*p < 0.05, **p < 0.01, ***p < 0.001, ns: not significant)