The role of KPNA2 as a monotonically changing differentially expressed gene in the diagnosis, risk stratification, and chemotherapy sensitivity of chronic hepatitis B-liver cirrhosis-hepatocellular carcinoma

Abstract

Purpose: Chronic hepatitis B-liver cirrhosis-hepatocellular carcinoma (CLH), commonly called the "liver cancer trilogy", is a crucial evolutionary phase in the emergence of hepatocellular carcinoma (HCC) in China. Previous studies on early diagnostic biomarkers of HCC were limited to the end-stage of HCC and did not focus on the evolutionary process of CLH.

Methods: 11 monotonically changing differentially expressed genes (MCDEGs) highly correlated with CLH were screened through bioinformatic analysis and KPNA2 was identified for further research. The serum KPNA2 expression in different CLH states was detected by Enzyme linked immunosorbent assay (ELISA). A nomogram model was constructed using univariate and multivariate Cox regression methods.

Results: The single-cell RNA-seq and bulk RNA-seq revealed that KPNA2 related to immune infiltration in HCC and may participate in cell cycle pathways in HCC. The serum KPNA2 expression was monotonically upregulated in CLH and was valuable for diagnosing different CLH states. Besides, chronic hepatitis B(CHB) patients, liver cirrhosis (LC) patients, and HCC patients were classified into subgroups with distinct serum KPNA2 expressions. Accordingly, patients with different serum KPNA2 expressions displayed various clinicopathological features. The AUC value of the nomogram model was 0.959 in predicting the likelihood of developing HCC in CHB patients or LC patients. Finally, we found that KPNA2 expression was negatively correlated with the IC₅₀ of four chemotherapeutic drugs in HCC.

Conclusion: KPNA2 was a novel serum biomarker for diagnosing different CLH states, monitoring the dynamic evolution of CLH, and a new therapeutic target for intervening in the progression of CLH.

Keywords: Biomarker; Diagnosis; Hepatocellular carcinoma; Karyopherin subunit alpha 2; Risk stratification.

Fig. 1

A-F Initial screening for MCDEGs highly correlated with CLH. A The PCA plot based on mRNA expression in GSE542348 cohort. B, C Volcano plot showing DEGs between chronic hepatitis and liver cirrhosis, liver cirrhosis and hepatocellular carcinoma, respectively. D, E Common upregulated and downregulated DEGs in GSE54238. F Heatmap displaying differential expression of the common genes in GSE54238

Fig. 2

Second screening for MCDEGs highly correlated with CLH. A–M The K–M curves of thirteen candidate genes in TCGA dataset. N Common candidate genes identified by initial screening were associated with the prognosis of HCC patients in TCGA dataset. O Heatmap displaying differential expression of the common genes in TCGA dataset

Fig. 3

Third screening for MCDEGs highly correlated with CLH. A–H The box plot displaying differential expression of eight genes in GSE121248. I Heatmap displaying differential expression of eight genes in GSE54238. J–L Chord diagram showing the correlation of eight genes in chronic hepatitis, liver cirrhosis and hepatocellular carcinoma, respectively

Fig. 4

Bioinformatic analysis of KPNA2 in HCC. A Differential expression between normal liver tissue and liver cancer tissue based on HPA database(immunohistochemistry). B The Sankey diagram showing relationship between KPNA2 expression and clinical features. C–H The GSEA of DEGs between KPNA2 high and low expression groups. I PPI network of KPNA2 generated by STRING. J Association between KPNA2 expression and immune cell infiltration levels in LIHC. K–N Association between KPNA2 expression and immune checkpoint genes (PDCD1, HAVCR2, CTLA4, CD274) in LIHC. O The box plot presenting the distributions of each immune subset at each copy number status in LIHC. LIHC, Liver Hepatocellular Carcinoma

Fig. 5

Analysis of single cell sequence data from multiregional sample in HCC. A Quality control assessment and data filtering of single cell sequence data. B Highly variable genes used for clustering and cell identification. C Top 20 principal components were identified based on P value < 0.05. D Different cell clusters visualized by UMAP reduction. E Corresponding annotation of the cell clusters. F KPNA2 expression between different immune cell clusters. G The pathways highly enriched in different immune cell clusters

Fig. 6

Diagnostic value of KPNA2 compared with AFP for distinguishing CLH states. Expression level of serum KPNA2 (A) and serum AFP (B) in different CLH states. For CHB vs LC, ROC curve analysis of KPNA2 (C) and AFP (D). For CHB vs HCC, ROC curve analysis of KPNA2 (E), AFP (F) and KPNA2 + AFP (G). For LC vs HCC, ROC curve analysis of KPNA2 (H), AFP (I) and KPNA2 + AFP (J). (* p < 0.05; ** p < 0.01; *** p < 0.001)

Fig. 7

Differential expression of KPNA2 in HCC tissues of different histologic grades, microvascular invasion, and Ki-67 positive rate. Left column(KPNA2 high expression group):KPNA2 high expression/high-grade/MVI:M2/Ki-67( +)50%; Right column(KPNA2 low expression group): KPNA2 low expression/low-grade/MVI:M0/Ki-67( +)10%. MVI: microvascular invasion

Fig. 8

A predictive model was established for CLH. A The nomogram combining four independent risk factors was developed for predicting the occurrence of HCC. The predicive ability of nomogram were tested by ROC analysis (B), Calibration curve (C) and Decision curve analysis (D)

Fig. 9

Drug sensitivity analysis for HCC patients with different KPNA2 expression. Association between KPNA2 expression and IC₅₀ values of 5-fluorouracil (A), doxorubicin (B), gemcitabine (C), sorafenib (D), respectively. Molecular docking simulation outcomes of KPNA2-5-fluorouracil complex (E), KPNA2-doxorubicin complex (F), KPNA2-gemcitabine complex (G), KPNA2-sorafenib complex (H)