Long-Term Hepatitis B Virus Infection Induces Cytopathic Effects in Primary Human Hepatocytes, and Can Be Partially Reversed by Antiviral Therapy

Abstract

Chronic infection of hepatitis B virus (HBV) remains a major health burden worldwide. While the immune response has been recognized to play crucial roles in HBV pathogenesis, the direct cytopathic effects of HBV infection and replication on host hepatocytes and the HBV-host interactions are only partially defined due to limited culture systems. Here, based on our recently developed 5 chemical-cultured primary human hepatocytes (5C-PHHs) model that supports long-term HBV infection, we performed multiplexed quantitative analysis of temporal changes of host proteome and transcriptome on PHHs infected by HBV for up to 4 weeks. We showed that metabolic-, complement-, cytoskeleton-, mitochondrial-, and oxidation-related pathways were modulated at transcriptional or posttranscriptional levels during long-term HBV infection, which led to cytopathic effects and could be partially rescued by early, rather than late, nucleot(s)ide analog (NA) administration and could be significantly relieved by blocking viral antigens with RNA interference (RNAi). Overexpression screening of the dysregulated proteins identified a series of host factors that may contribute to pro- or anti-HBV responses of the infected hepatocytes. In conclusion, our results suggest that long-term HBV infection in primary human hepatocytes leads to cytopathic effects through remodeling the proteome and transcriptome and early antiviral treatment may reduce the extent of such effects, indicating a role of virological factors in HBV pathogenesis and a potential benefit of early administration of antiviral treatment. IMPORTANCE Global temporal quantitative proteomic and transcriptomic analysis using long-term hepatitis B virus (HBV)-infected primary human hepatocytes uncovered extensive remodeling of the host proteome and transcriptome and revealed cytopathic effects of long-term viral replication. Metabolic-, complement-, cytoskeleton-, mitochondrial-, and oxidation-related pathways were modulated at transcriptional or posttranscriptional levels, which could be partially rescued by early, rather than late, NA therapy and could be relieved by blocking viral antigens with RNAi. Overexpression screening identified a series of pro- or anti-HBV host factors. These data have deepened the understanding of the mechanisms of viral pathogenesis and HBV-host interactions in hepatocytes, with implications for therapeutic intervention.

Keywords: cytopathic effects; long-term HBV infection; nucleot(s)ide analogues; quantitative temporal proteomics; viral-host interaction.

FIG 1

Temporal profiling of long-term HBV-infected hepatocytes. (A) 5C-PHHs were infected with HBV at an MOI of 200. (B) Extracellular HBsAg and HBeAg were evaluated every 3 days and analyzed by qPCR with specific primers. (C) Intracellular HBV pgRNA and cccDNA were extracted at indicated time points and analyzed by qPCR with specific primers. (D) Distribution of HBV transcripts along the HBV genome were quantified by nucleotide mapping. Read density for each track was represented by height on the y axis. (E) Mock- and HBV-infected cells were collected at 2, 7, and 28 dpi, and then proteomic analysis was applied. Scatterplots displayed pairwise comparisons between HBV- and mock-infected cells; each point represents a single protein. Benjamini-Hochberg corrected significance was utilized to estimate P values. Orange dots, P < 0.01; red dots, top and bottom deregulated proteins. (F) Parallel samples were applied to RNA-seq analysis. Heatmaps displaying top 50 dysregulated genes (left panel, 2 dpi; middle panel, 7 dpi; right panel, 28 dpi).

FIG 2

Cytopathic effects caused by long-term HBV infection. (A) 5C-PHHs were infected with HBV (MOI of 200), and samples were collected at 28 dpi and proteomics analysis applied. DAVID analysis of upregulated (left) and downregulated (right) proteins (>1.5-fold change and P < 0.05) against a background of all proteins quantified in the proteomics. Components of all clusters are shown in Table S2 in the supplemental material (B) Integrity toxic analysis of all dysregulated proteins by IPA analysis. (C) 5C-PHHs were infected with HBV at an MOI of 200 for 28 days. Representative TEM photographs showing abnormal mitochondria ultrastructure after long-term HBV infection in hepatocytes. (D) Immunofluorescent analysis of Hoechst 33342 (blue), MitoTracker Green (green), and MitoSOX (red) stain in mock- and HBV-infected 5C-PHH cells. (E) Confocal photographs of Hoechst 33342 (blue) and JC-1 (monomers, green; aggregates, red) stain in mock- and HBV-infected 5C-PHH cells, as indicated.

FIG 3

Transcriptional and posttranscriptional regulation of protein expression. (A) Schematics of experiment workflow. (B) K-means-based hierarchical cluster analysis of proteins and transcripts to identify global mechanisms of protein regulation by HBV infection. Right panels show examples of each cluster. (C) Numbers of proteins from protein degradation, transcriptional upregulation, and transcriptional downregulation shortlists according to relative criteria appearing in each cluster. (D) Proteins from protein degradation appearing in cluster 5 and cluster 8 are displayed.

FIG 4

Identification of proteins dysregulated by HBV while recovered by nucleot(s)ide analogues. (A) 5C-PHH cells infected with HBV (MOI of 200) were treated with or without entecavir (100 nM) at 0 dpi for 28 days. (B and C) Supernatants were collected every 3 days, and HBV DNA (B) and HBsAg and HBeAg analyses (C) via commercial kits were applied. (D) Scatterplots of protein changes in comparison between values of ETV-treated and the corresponding HBV-infected cells (left, 7 dpi; right, 28 dpi). P values were estimated using Benjamini-Hochberg-corrected significance. Orange dots, P < 0.01; red dots, top and bottom deregulated proteins. (E) Overlap of proteins identified in ETV rescue screen (stringent criteria; see Fig. S5A in the supplemental material) and RNA and protein screen. STRING was applied to analyze protein correlation that was degraded by HBV infection but rescued by early ETV treatment. (F) STRING analysis of protein-degraded targets that could not be rescued by ETV treatment.

FIG 5

Late administration of nucleot(s)ide analogues had limited effects on relieving HBV cytotoxicity. (A) 5C-PHH cells were infected with HBV (MOI of 200) and then treated with or without ETV (1 μM) or TDF (5 μM) at 10 dpi. (B and C) Supernatants were collected every 3 days and applied to HBV DNA (B) and HBsAg and HBeAg (C) analysis. (D) Intracellular cccDNA was extracted and detected by qPCR with indicated primers. (E) Comparison of ETV-treated or untreated HBV-infected cells with mock-infected cells for 28 days. Values were log₂ transformed. (F) Comparison of TDF-treated or untreated HBV-infected cells with relative mock-infected cells. (G) Volcano plots displaying pairwise comparisons between early ETV-treated and untreated HBV-infected cells (left, 7 dpi; right, 28 dpi). (H) Gene set enrichment analysis (GSEA) pathway terms enriched among genes rescued by ETV therapy at 28 dpi are displayed.

FIG 6

Blocking HBV antigen by RNAi treatment significantly relieved HBV-induced host remodeling. (A) 5C-PHH cells infected with HBV (MOI of 200) were treated with (siHBV) or without (HBV) a combination of two siRNAs targeting all HBV-derived RNAs (100 nM) at 7 dpi for 1 week. (B) Supernatants were collected every 3 days, and HBsAg and HBeAg analysis was applied. (C) Distribution of HBV transcripts along the HBV genome was quantified by nucleotide mapping. Read density for each track was represented by height on the y axis. (D) KEGG pathway enrichment terms enriched among genes rescued by RNAi therapy (left, cells downregulated by HBV while they were rescued by RNAi; right, cells upregulated by HBV while they were relieved by RNAi) are displayed. (E) Immunofluorescent analysis of Hoechst 33342 (blue), MitoTracker Green (green), and MitoSOX (red) stain in mock- and HBV-infected 5C-PHH cells with or without RNAi treatment, as indicated.

FIG 7

Overexpression screening of dysregulated proteins identified in combined analysis. (A) Schematics of experiment workflow. Summary of sensitive and stringent criteria used in this study; this facilitated identification of a shortlist of hits dysregulated by HBV infection with high confidence. (B) HepG2 cells were cotransfected with overexpressed plasmids along with prcccDNA and pCMV-Cre plasmids. Dot plots showing the Z-score normalized average HBsAg and HBeAg values from two independent experiments. Top and bottom proteins are indicated by red dots. (C and D) Confirmation analysis for selected pro- and anti-HBV genes. MyD88 was included as a positive control. HBV antigen levels (C) and HBV RNA levels (D) were normalized to the enhanced green fluorescent protein (EGFP) control. (E and F) Selected antiviral (E) and proviral (F) gene expression plasmids were cotransfected with pHBV1.3 plasmid into HepG2 cells. Cells were harvested at 5 days posttransfection. Core particle DNA was detected by Southern blot analysis.

FIG 8

Schematics illustrating the correlations between HBV and host cells. Global temporal quantitative proteomics analysis of long-term HBV-infected primary human hepatocytes accompanied by parallel transcriptomic analysis uncovered extensive remodeling of host proteome and transcriptome, as well as revealed cytopathic effects of long-term viral replication. Metabolic-, complement-, cytoskeleton-, mitochondrial-, and oxidation- related pathways were modulated at transcriptional or posttranscriptional levels, which could be partially rescued by early, but not late, nucleot(s)ide analogs therapy and could be relieved by RNAi. Thirteen proteins, many without characterized functions in promoting or inhibiting HBV replication, were identified by our unbiased analysis and further overexpression screen.