Apolipoprotein H induces sex-specific steatohepatitis and gut dysbiosis during chronic hepatitis B infection

Summary Apolipoprotein H (APOH) is involved in lipid metabolism and functions as an acute-phase protein during hepatitis B virus (HBV) infection. Herein, we explored whether APOH acts on the development of fatty liver upon chronic HBV infection. Serum APOH level was significantly lower in cirrhosis patients than in healthy controls or patients with chronic infection. It showed sex bias, with elevated levels in female patients with chronic infection. Also, serum APOH levels were negatively correlated with HBV surface antigen (HBsAg) but positively correlated with albumin and triglyceride levels. In In vitro HBV infection model, HBV upregulated APOH expression in a non-temporal manner, and HBsAg levels were elevated by silencing APOH. RNA sequencing (RNA-seq) demonstrated bidirectional expression of APOH, which impacted the immunoregulation upon infection or the metabolic regulation in HepG2.2.15 cells. Then, ApoH−/− mice with persistent HBV replication displayed steatohepatitis and gut microbiota dysbiosis with synergistic sex differences. Our study deciphers the roles of APOH in chronic liver diseases.


INTRODUCTION
Chronic hepatitis B (CHB) virus infection remains an unresolved global public health problem, with more than 400 million people infected worldwide. [1][2][3][4][5] Recently, multiple studies have revealed that approximately a quarter of CHB patients have concurrent steatosis. The potential consequences of developing non-alcoholic fatty liver disease (NAFLD) in patients with CHB infection include active hepatitis, progression of chronic liver disease, and the development of hepatocellular carcinoma. [6][7][8][9] It has been reported that the gut microbiota acts as a major determinant of this relationship in the onset and clinical course of liver diseases and that sex is significantly associated with viral infection, lipid metabolism, and gut microbiota composition. [10][11][12] It remains to be seen what mechanisms underlie the relationship between steatosis and chronic hepatitis B virus (HBV) infection.
It is well known that apolipoprotein H (APOH, also known as beta2-glycoprotein I) is an abundant plasma apolipoprotein primarily produced in the liver 13,14 and is closely associated with lipid metabolism. [15][16][17][18][19][20] For instance, APOH binds to lipoproteins and activates the lipoprotein lipase during triglyceride metabolism. Moreover, the distinct APOH protein isoforms alter the expression of apolipoprotein B (APOB), apolipoprotein A (APOA), high-density lipoprotein cholesterol (HDL-C), triglyceride (TG), and total cholesterol (TC). In addition, APOH is reportedly an acute-phase protein during HBV infection, and the HBV surface antigen (HBsAg) binds to APOH with high affinity. [21][22][23] We previously found that HBV and their large surface antigens directly upregulate APOH expression, which inhibits HBsAg secretion and further induces hepatocellular ER stress. 24,25 Therefore, we hypothesize that APOH might be a point of convergence between hepatocyte steatosis and HBV-related chronic liver diseases.
This study analyzed the dynamics of serum APOH levels, by which we aim to understand the association between APOH and liver function in different disease conditions in patients with chronic HBV infection. Furthermore, we analyzed RNA sequencing (RNA-seq) data from HepG2.2.15 cells silencing or overexpressing APOH. We further constructed an ApoH gene-knockout mouse model and a persistent HBV replication mouse model to investigate the associations between APOH levels and hepatocyte steatosis, gut HBV stimulation upregulates APOH in NTCP-reconstituted HepG2 cells To confirm whether the stage of HBV infection impacts APOH expression levels, we used an in vitro HBV infection system, i.e., sodium taurocholate co-transporting polypeptide (NTCP)-reconstituted HepG2 cells. We confirmed that HBV temporally upregulated APOH mRNA expression in the early stage of HBV infection (Figure 2A). Figure 2B showed the levels of HBV S mRNA in NTCP-reconstituted HepG2 cells after HBV infection. Further experimentation indicated that APOH gene knockdown ( Figure 2C) upregulated HBV S mRNA expression ( Figure 2D).
Variable APOH expression reveals changes in enriched signaling pathways and alters the expression of metabolic regulatory genes APOH was expressed at the mRNA level in different groups ( Figure 3A). We consulted the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways to analyze which signaling pathways were affected by iScience Article APOH expression. In cells overexpression APOH, the enriched signaling pathways were mainly cytokinecytokine receptor interactions and Toll-like receptor signaling pathway ( Figure 3B). However, in APOHsilenced cells, metabolic pathways were enriched based on KEGG pathway analysis ( Figure 3C). A list of 19 differentially expressed genes (DEGs) between the two groups is shown in Figure 3D. Moreover, the analyzed metabolic regulatory pathways are illustrated in the bubble diagram ( Figure 3E).
Notably, cytochrome P450 (CYP) enzyme-related genes were over-represented in the 19 DEGs, including CYP3A4, CYP3A7, CYP7A1, CYP11A1, CYP21A2, and CYP26A1. CYP450 enzymes have been well documented to act as a terminal oxidase in the multi-function oxidase system to metabolize different endogenous substrates and xenobiotics (Table S2). 26,27 The isoenzymes CYP7A1, CYP11A1, and CYP21A2 primarily participate in steroid biosynthesis and metabolism. Of the nine DEGs examined ( Figure 3F), CYP7A1, CYP11A1, and NT5E showed significantly increased expression in the APOH-silencing group (p < 0.05), while CYP3A4 was sharply decreased in the same group (p < 0.05). Therefore, we conclude that APOH production is decreased in the background of chronic HBV infection, which activates metabolism pathways and mainly affects steroid hormone biosynthesis.
APOH-knockout mice exhibit signs of liver damage and steatohepatitis Next, we constructed an ApoH-knockout mouse model to investigate the potential function of APOH in the liver. The 6-week ApoH À/À mice were used for genotype identification. A part of the sequencing data is shown in Figure 4A, and the PCR products were resolved on an agarose gel ( Figure 4B). ApoH mRNA expression was highly reduced, confirmming the knockdown efficiency ( Figure 4C). The serum alanine aminotransferase (ALT) and aspartate aminotransferase (AST) levels were significantly higher in ApoH À/À mice than in wild-type (WT) mice (p < 0.05) ( Figures 4D and 4E), but no significant differences were observed for TG and TC levels (p > 0.05) ( Figures 4F and 4G). We also found apparent steatosis in the liver sections of ApoH À/À mice stained with H&E or Oil Red O ( Figure 4H). The results suggested that the downregulation of ApoH expression induced mouse spontaneous steatohepatitis. iScience Article ApoH À/À mice display altered microbiota diversity and differentially abundant bacterial species We explored the gut community diversity and the difference in bacterial species abundance between ApoH À/À and WT mice. The 16S rRNA sequencing data indicated no significant difference in the alpha and beta diversities between the two mice groups (p > 0.05) (alpha diversity: Figures 5A-5C, beta diversity: Figures 5D and 5E). We further analyzed the bacterial composition at the phylum and genus levels ( Figures 5F-5H). The total bacterial composition and abundance were reduced in ApoH À/À mice compared to those in the WT mice. In ApoH À/À mice, at the phylum level, the abundance of Bacteroidetes was increased, while that of Firmicutes and Actinobacteria was sharply decreased. At the genus level, the abundance of Lachnospiraceae, Parabacteroides, Staphylococcus, and Acetatifactor increased, in contrast to the decrease in abundance for Lactobacillus, Bifidobacterium, ClostridiumXIVa, Eubacterium, and Barnesiella.
Next, to investigate potential metabolic differences between ApoH À/À and WT mice, we performed the predictive functional profiling of microbial communities using the PICRUSt MetaCyc pathway ( Figure 5I). The activity of phosphopantothenate biosynthesis I (PANTO-PWY) pathway was higher in ApoH À/À mice (p < 0.05). The downregulated pathways in ApoH À/À mice included the following: superpathways of polyamine biosynthesis II, 2,3-butanediol biosynthesis, de novo pyrimidine deoxyribonucleotides biosynthesis, and polyamine biosynthesis I, along with de novo pyrimidine deoxyribonucleotides biosynthesis III, sucrose degradation IV (sucrose phosphorylase), and coenzyme M biosynthesis I.
ApoH À/À mice with persistent HBV infections exhibit increased levels of HBsAg secretion and liver steatosis The HBV replication mouse model was generated using six-to seven-week-old ApoH À/À mice and C57BL/6 WT mice as controls ( Figure 6A). The mouse model was used to study the effects of APOH iScience Article in the context of persistent HBV infection and immune tolerance. Serum HBsAg levels were higher in ApoH À/À mice than in WT mice ( Figure 6B). We also found that the HBsAg levels in female mice were higher than those in male mice after seven weeks of HBV infection ( Figure S3). Serum ALT and AST levels in male ApoH À/À mice were significantly lower than those in male WT mice (p < 0.05) ( Figures 6C and  6D). The serum TG and TC levels were sharply increased in male mice compared with those in females, independent of APOH expression (p < 0.05) ( Figures 6E and 6F). Furthermore, significant levels of hepatocyte steatosis were observed in ApoH À/À mice with a persistent HBV infection ( Figure 6G). Therefore, ApoH downregulation predominately promoted HBsAg secretion and induced steatohepatitis in female ApoH À/À mice.
Mice with persistent HBV replication have altered gut microbiota community diversity and differentially abundant bacterial species We also explored the gut community diversity and bacterial species abundance in mice with persistently replicating HBV. The Venn diagram shows the number of common and unique species in different groups ( Figure 7A). Figure 7B illustrates the bacterial composition at the phylum level in the four groups. We found that the abundance of Verrucomicrobiota sharply decreased in ApoH À/À mice, and sex differences synergistically affected gut microbiota composition. The abundance of Firmicutes, Actinobacteria, and Patescibacteria had increased in female ApoH À/À mice compared with that in female WT mice. Concurrently, the abundance of Deferribacterota and Bacteroidota had significantly decreased in female ApoH À/À mice. However, among male mice, the abundance of Deferribacterota and Bacteroidota had observably increased in Apoh À/À mice compared with that in WT mice, while the abundances of Actinobacteria and Patescibacteria had sharply decreased in Apoh À/À mice. We further analyzed the bacterial composition at the genus level among the four groups, and Figure 7C shows the detailed gut microbiota composition in ApoH À/À female mice. A significant difference was observed in the alpha diversity between WT and ApoH À/À mice and also between female and male mice in the ApoH À/À group (p < 0.05) ( Figure 7D). Next, the beta diversity assessment performed by principal coordinate analysis (PCoA) at the phylum level revealed a significant difference between the WT and ApoH À/À groups (p < 0.05) ( Figure 7E).
Finally, we used PICRUSt software to perform predictive functional profiling of microbial communities via KEGG enrichment analysis ( Figure 7F). The gut microbiota mainly performed metabolic functions in ApoH À/À and WT mice of different groups, but no significant difference was observed between groups (p > 0.05).

DISCUSSION
In the present study, we analyzed the serum APOH levels in patients with HBV-related chronic liver diseases and utilized the HepG2.2.15 cell line and an ApoH À/À plus HBV replication mouse model to further investigate the effect of APOH on hepatocyte steatosis and gut microbiota dysbiosis. The graphical abstract summarizes the key findings.
It has been known that APOH is an acute-phase protein in viral infection. 28 In our previous study, we found that APOH is highly expressed in HepG2.2.15 cells, with HBV and the large surface antigen directly upregulating APOH expression. 24 Therefore, we were intrigued to explore the role of APOH during the course of HBV infection. Firstly, we analyzed APOH levels in the sera of patients with chronic HBV infection and iScience Article correlated with other biomarkers in HBV-related liver diseases. We found that the serum APOH levels were significantly decreased in patients with cirrhosis, and the reduction was more obvious in females than in males during the chronic infection phase. Further analysis indicated that the serum APOH levels were negatively correlated with the HBsAg levels and positively associated with the TG and albumin levels in patients with HBV-related chronic liver diseases. Taken together, we propose that APOH might play different roles in acute versus chronic HBV infection, and the potential sex-synergistic effect might be another hinge point between APOH and viral infection.
Next, on one hand, we confirmed that HBV significantly upregulated APOH expression in a non-temporal manner. We considered that APOH performs additional functions during chronic HBV infection, besides acting as an acute-phase protein in viral infection. On the other hand, we found that, during the course of HBV infection, APOH is switched from performing an early regulatory role in the immune function to a later regulatory role in metabolism. Further analysis indicated that some significant DEGs in APOH-silencing group, including CYP21A2, CYP3A4, CYP11A1, CYP3A7, and AKR1810, are associated with steroid biosynthesis and metabolism. 26 Therefore, we speculate that the regulatory effect of APOH is associated with metabolic functions of liver in a gender-dependent manner with or without HBV infection.
In addition, patients suffering from chronic HBV infection with concomitant steatosis are at an increased risk of hepatic fibrosis, cirrhosis, and hepatocarcinoma. 8,9 This motivates researchers to understand the iScience Article molecular complexities associated with the diseases. Based on the current literature on the association of APOH and lipid metabolism, [15][16][17][18][19][20] we believe that APOH protein polymorphism underlie variations in plasma lipid levels, as shown in Table S1. Moreover, we are interested in verifying the effect of APOH on lipid metabolism during chronic HBV infection in a future study.
In vivo, we also found that ApoH À/À mice exhibited spontaneous steatohepatitis. Considering that gut-liver axis regulation and gut microbiota dysbiosis have been highlighted as crucial pathogenetic factors of NAFLD, 29-31 here we further investigated the gut community diversity and the difference in bacterial species abundance in ApoH À/À mice. The total bacterial composition and abundance at the phylum and genus levels had significantly decreased in ApoH À/À mice, consistent with previous reports. 32 Thus, it is likely that ApoH deletion in mice might induces dysbiosis in metabolism regulation and gut microbiota. We should have a better understanding of its role by constructing the non-alcoholic fatty liver mouse model in the background of ApoH À/À mice.
Next, we constructed a persistent HBV replication mouse model in the background of ApoH À/À mice. During the periodic detection of serum HBsAg titers, we found that the serum HBsAg levels were higher in ApoH À/À mice than in WT mice, which supports our previous finding from in vitro cell experiments, wherein high expression of APOH inhibited the HBsAg secretion. 25 However, we found that the serum HBsAg, ALT, TG, and TC levels and the degree of fatty liver appeared to show sex-synergistic differences. It has been well documented that sex is a key factor in shaping the immune responses, contributing to differences in the pathogenesis of infectious diseases, between male and female patients. 33,34 Moreover, metabolic homeostasis is differentially regulated between men and women, and there are sex-specific effects on lipid and cholesterol metabolism. Briefly, sexual differences exist in ApoH deletion-regulated lipid metabolic dysbiosis, HBsAg secretion, and the development of liver injury in chronic HBV infection. iScience Article Another factor in the development of fatty liver disease is the gut microbiota. Then, we detected the community diversity and the difference in bacterial species abundance in WT and ApoH À/À mice on the background of persistent HBV replication. We found that ApoH deletion altered the community and diversity of the gut microbiota. The abundance of Verrucomicrobiota, an intestinal commensal microbiota in healthy individuals, had sharply decreased in ApoH À/À mice. Second, sex differences affected the community and diversity of the gut microbiota. In the abundance of Actinobacteria and Patescibacteria versus Deferribacterota and Bacteroidota, an opposite trend was observed in ApoH À/À female and male mice compared with that in WT mice. Third, the microbial communities were mainly associated with metabolic functions in different groups. Sex and gut microbiota have been reported to regulate host metabolism and viral infection. Figure S4 indicated the significant metabolic alterations in patients with chronic HBV infection. However, lipid metabolic dysbiosis and HBV-related liver injury also influence gut microbiota hemostasis. 10,35,36 In addition, physiological processes are differentially regulated between men and women. Sex-specific ApoH deletion in mice promoted HBsAg secretion, disturbed lipid metabolism, and aggravated hepatocyte steatosis. Our on-going study should enhance our understanding of the regulatory mechanism.
Researchers have been trying to understand the molecular mechanisms underlying the development of CHB infection and NAFLD, and this study attempts to elucidate the pleiotropic effect of APOH in HBVrelated chronic liver diseases. It provides a new perspective on sex-specific lipid metabolic dysbiosis by combining gut microbiota and HBV infection. This study provides a valuable framework to decipher the roles of APOH in chronic liver diseases. Obviously, in-depth studies on the mechanism of chronic liver diseases are crucial for clarifying the role of APOH in liver diseases, where animal models that can adequately reflect the clinical settings should be constructed.

Limitations of the study
In our study, we had the unique and original opportunity to investigate the potential regulatory role of APOH during chronic HBV infection; this study, however, had some limitations. Firstly, the cross-sectional iScience Article design and low sample size may limit the statistical power of APOH levels and correlation analysis. Then, we analyzed some open-source data deriving from the GEO databases to validate our findings (Figures S1 and S2). Next, in this study, we found ApoH deletion in mice might induce dysbiosis in metabolism regulation and gut microbiota. We should have a better understanding of its role by constructing the non-alcoholic fatty liver mouse model in the background of ApoH À/À mice to pursue the further study on the association of ApoH-induced abnormality in metabolism and development of fatty liver and gut microbiota dysbiosis.
In conclusion, we report unique effects of APOH in HBV-related chronic liver diseases. It provides a new perspective on sex-specific lipid metabolic dysbiosis by combining gut microbiota and HBV infection.

STAR+METHODS
Detailed methods are provided in the online version of this paper and include the following: Code statement: This paper does not report original code. The source and identifier of analysis code used in the paper can be found in the 'key resources table' and/or 'quantification and statistical analysis' section.
Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

EXPERIMENTAL MODEL AND SUBJECT DETAILS
Animal care and generation of ApoH À/À mice All experiments on mice were approved by the Institutional Animal Care and Use Committee of Xiamen University (XMULAC20210119) and were conducted in accordance with institutionally approved protocols and guidelines for animal care and use. The detail protocol for the generation of the ApoH -/mouse model was described in our previous publication. 37

Mouse model of persistent HBV replication
The HBV replication mouse model was constructed using six-to seven-week-old female and male mice that were either ApoH -/or WT controls. First, HBV plasmids were purchased from the Beijing Five-Plus Institute of Molecular Medicine Co. LTD. The HBV plasmid was derived from a recombinant adeno-associated virus type 8 with 1.2 times the HBV genome (rAAV8-1.3HBV, subtype ayw). Theres plasmids were then injected into the tail vein of the mice by hydrodynamic injection (5 3 10 10 mg/200 mL per mouse). After injuection, orbital venous blood was collected weekly for 9 weeks, and the titer of HBsAg was measured using enzyme-linked immunosorbent assay (ELISA).

Patient samples
Serum samples from 53 patients with chronic HBV infection and 18 healthy controls were used in this study. The samples were collected from the First Hospital of Jilin University in Changchun, China. Patients with chronic infection, chronic hepatitis, and cirrhosis were diagnosed 38 and treated with antiviral drugs in an outpatient setting. All procedures in this study were performed in accordance with the ethical standards of the medical ethics committee of the Zhongshan Hospital Xiamen University, and the First Hospital of Jilin University.

ELISA for APOH in human serum
Serum samples were drawn from individuals using sterile syringes and stored individually in blood collection tubes to avoid cross-contamination. The serum samples were kept at À80 C until used for analysis. The serum APOH levels (40,000-fold dilution) were quantified using a commercial ELISA kit (Uscn Life Science Inc, Wuhan, China) according to the manufacturer's instructions.

HBV infection
The detail procedure for HBV infection has been previously described. 39 Briefly, HepG2-NTCP-tet cells were cultured with 4 mg/mL doxycycline (DOX) for 4 days to induce NTCP expression. The cells were ll OPEN ACCESS