Six to eight-week-old male BALB/c and C57BL/6 (C57BL/6J substrain was used in this study) mice were purchased from Hunan SJA Laboratory Animal Co., Ltd. (Hunan, China) and were housed in a specific pathogen-free facility at Huazhong University of Science and Technology. In the new facility, a 12-hour light/12-hour dark cycle was used, temperature of 20-26°C and 40%-60% humidity were kept. The mice food was composed of 12% fats, 20.6 proteins, and 67.4% carbohydrates.

This study was carried out according to the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health and was approved by the Institutional Animal Care and Use Committee at Tongji Medical College, Huazhong University of Science and Technology (permit number: S814).

To establish substantial and sustained microbiota engraftment from donor mice into recipient mice, BALB/c mice were subjected to a streamlined ABX conditioning regimen, as described previously (Staley et al., 2017). Two antibiotic cocktails were used; cocktail 1 consisted of three highly absorptive antibiotics: ampicillin, cefoperazone sodium salt, and clindamycin hydrochloride, and cocktail 2 comprised three lower-degree absorptive antibiotics: ertapenem (Invanz), neomycin sulfate, and vancomycin hydrochloride. The antibiotics were dissolved in drinking water at a concentration of 1 mg/mL. BALB/c mice were treated with antibiotic cocktail 1 on days 0 to 7, with cocktail 2 on days 9 to 18, and then switched back to cocktail 1 during days 18 to 25 (Figure 3A).

To reconstitute the gut microbiota after ABX treatment, FTM was carried out as described previously (Lei et al., 2016; Villarino et al., 2016; Rosshart et al., 2017). Fresh fecal pellets from naïve C57BL/6 or BALB/c mice were collected and resuspended in 1 mL of sterile PBS. After filtration through a 70-μm cell strainer, the fecal material was transplanted into recipient BALB/c mice using daily oral gavages on days 27 to 34 (Figure 3A).

A previously described (Huang et al., 2006; Zhu et al., 2019; Guo et al., 2021), well-established HBV infection mouse model was used in this study. Briefly, 10 μg of HBV plasmid, pAAV/HBV1.2, was diluted in PBS and injected at 0.1 mL/g of mouse body weight via the tail vein within 5–8 s. At the indicated time points, blood samples were collected from the mice, and sera were diluted 1:10 in PBS. HBsAg, HBsAb, HBeAg, HBeAb, and HBcAb levels were measured using an ELISA kit (Kehua, Shanghai, China). Serum HBV DNA titers were assessed using a qPCR kit (Sigma-Aldrich, St. Louis, MO, USA).

Liver samples were fixed in 4% paraformaldehyde and embedded in paraffin, as described previously (Wang et al., 2014). A rabbit anti-HBV core antigen (HBcAg) polyclonal antibody (Dako, Copenhagen, Denmark) was used for IHC staining, and the sections were visualized using a DAKO EnVision™ Detection System (Dako).

Fecal pellets were harvested and stored at –80°C. Bacterial DNA was extracted using an E.Z.N.A. Soil DNA Kit (Omega Bio-tek, Norcross, GA, USA). 16S rRNA gene sequencing was performed as described previously (Zhu et al., 2019; Guo et al., 2021), using primers to amplify the V3-V4 hypervariable regions of the bacterial 16S rRNA gene (338F, 5′-ACTCCTACGGGAGGCAGCAG-3′ and 806R 5′-GGACTACHVGGGTWTCTAAT-3′). Amplicons were purified and paired-end sequenced (2 × 300) using an Illumina MiSeq System (Illumina, San Diego, CA, USA). The sequencing data have been deposited into the NCBI Sequence Read Archive database (accession No.: SRP174629).

Mouse livers were perfused with 10 mL of PBS and homogenized in Hanks’ buffer. The homogenate was resuspended in 36.5% Percoll (Sigma-Aldrich), and lymphocytes were isolated by density gradient centrifugation. The lymphocytes were cultured in RPMI 1640 medium and stimulated with CD8+ T-cell epitope (L^d-HBV S28–39 epitope, IPQSLDSWWTSL, 10 μg/mL) as described previously (Schirmbeck et al., 2001; Sette et al., 2001). The following antibodies were used for cell-surface staining and intracellular cytokine staining: BV421-anti-CD8, APC-Cy7-anti-CD4, PE-anti-CTLA4, and PE-Cy7-anti-PD1 (eBioscience, San Diego, CA, USA). For intracellular cytokine staining, APC-anti-IFN-γ, PerCP-Cy5.5-anti-IL-10, and FITC-anti-TNF-α (BioLegend, San Diego, USA) were used. Dead cells were excluded by staining with Fixable Viability Dye eFluor 506 (eBioscience). Samples were analyzed using a FACSCanto II flow cytometer (BD Biosciences, San Jose, CA, USA). Data were analyzed using FlowJo (version 10.0; TreeStar, Ashland, OR, USA).

Data are presented as the mean ± SD. For the comparison analyzing between two groups, two-tailed unpaired Student t-test was used for data with normal distribution, Mann-Whitney test were used for data with non-normal distribution. Means of multiple groups were compared using one-way ANOVA followed by Tukey’s multiple comparison tests. Kaplan–Meier survival analysis was used to compare the frequency of HBsAg-positive mice. P < 0.05 was considered significant. Statistical analyses were performed using SPSS (version 12.0; SPSS Inc., Chicago, IL, USA).

To confirm that the genetic background of the host influences the HBV infection outcome, BALB/c and C57BL/6 mouse HBV infection models were established. Serum levels of HBV surface antigen (HBsAg) in BALB/c mice were lower than those in C57BL/6 mice at 21 days post infection (dpi) and 28 dpi (Figure 1A). HBsAg was cleared from the serum within 21 days in BALB/c mice but persisted for more than 91 days in 20% of the C57BL/6 mice (Figure 1B). HBeAg and HBV DNA were detected in sera at 7 dpi but were undetectable after 21 dpi in both groups (Figures 1C, D). HBV surface antibody (HBsAb) was detectable in sera after 28 dpi, and HBsAb levels were slightly, albeit not significantly, higher in BALB/c mice than in C57BL/6 mice (Figure 1E). HBV e antibody (HBeAb) was undetectable in sera of both BALB/c and C57BL/6 mice (Figure 1F). HBV core antibody (HBcAb) was detected at 7 dpi and gradually decreased in both BALB/c and C57BL/6 mice, with no significant differences (Figure 1G). At 56 dpi, HBV core antigen (HBcAg) was detected in the livers of C57BL/6 mice but not in those of BALB/c mice (Figure 1H). These results indicated that HBV infection outcomes differ depending on the host genetic background. HBV infection tends to be transient in BALB/c mice but persistent in C57BL/6 mice.

Given that the gut microbiota affects HBV infection outcomes in mice (Chou et al., 2015), we comparatively investigated the gut microbiota composition in naïve adult BALB/c and C57BL/6 mice. Fecal pellets were collected and analyzed by 16S rRNA sequencing. We identified 494 operational taxonomic units (OTUs); 250 OTUs were shared by both BALB/c mice and C57BL/6 mice, 110 OTUs were unique to BALB/c mice, and 134 OTUs were unique to C57BL/6 mice (Figure 2A).

The gut microbiota composition in BALB/c and C57BL/6 mice was assessed at the phylum, family, and genus levels (Figure 2B). At the phylum level, the abundance of Firmicutes was higher, whereas those of Bacteroidetes and Verrucomicrobia were lower in BALB/c mice than in C57BL/6 mice. At the family level, the abundances of Lachnospiraceae, Rikenellaceae, Bacteroidaceae, Ruminococcaceae, and Porphyromonadaceae were higher, whereas those of Bacteroidales_S24-7_group, Erysipelotrichaceae, and Prevotellaceae were lower in BALB/c mice than in C57BL/6 mice. At the genus level, the abundances of Alistipes, norank_f_Lachnospiraceae, Bacteroides, and Odoribacter were higher, whereas those of norank_f_Bacteroidales_S24-7_group, Candidatus Saccharimonas, Rikenella, and Parabacteroides were lower in BALB/c mice than in C57BL/6 mice.

As for diversity, no significant differences were observed in OTU counts, Chao1 community richness, and Shannon–Weaver index values, indicating that the α-diversity was comparable between BALB/c mice and C57BL/6 mice (Figure 2C). However, the β-diversity differed significantly between BALB/c mice and C57BL/6 mice as determined by Bray-Curtis dissimilarity (Figure 2D). The Firmicutes/Bacteroidetes ratio was significantly higher in BALB/c mice than in C57BL/6 mice (Figure 2E). The ratio of Faecalibacterium prausnitzii/Escherichia coli at species level was comparable (data not shown). These results demonstrated that the gut microbiota composition is influenced by the genetic background in mice.

To confirm that HBV infection outcomes in BALB/c and C57BL/6 mice are influenced by the gut microbiota composition, we depleted the gut microbiota in BALB/c mice using ABX and then reconstituted it with fecal microbiota from naïve adult BALB/c or C57BL/6 mice (Figure 3A). After ABX treatment, fecal pellets were collected and analyzed by 16S rRNA sequencing on day 25. The OTU counts, Chao1 community richness, and Shannon–Weaver index value were significantly lower in ABX-treated mice than in control mice (Figure 3B), and Bray-Curtis dissimilarity indicated that the microbiota composition was significantly shifted after ABX treatment (Figure 3C; ANOSIM, r = 0.5778, P = 0.001; PERMANOVA, pseudo-F = 5.87, P = 0.001). On days 27 to 34, fecal microbiota from naïve adult BALB/c or C57BL/6 mice was transplanted into recipient mice (Figure 3A). On day 35, OTU counts, Chao1 community richness, and Shannon–Weaver index values in mice of the ABX+BALB/c FMT and ABX+C57BL/6 FMT groups were restored to the levels in the control mice and were significantly higher than those in mice of the ABX group (Figure 4A). The Bray-Curtis dissimilarity was also restored after FMT in both the ABX+BALB/c FMT and ABX+C57BL/6 FMT groups (Figure 4B). These results indicated that ABX-depleted gut microbiota in BALB/c mice can be reconstituted by transplantation of commensal bacteria from BALB/c or C57BL/6 mice.

To investigate the effect of FTM on the gut microbiota composition, gut microbiota profiles were assessed at the phylum, family, and genus levels at day 35. In mice of each treatment group, the gut microbiota comprised five phyla, including Bacteroidetes, Firmicutes, Proteobacteria, Saccharibacteria, and Verrucomicrobia (Figure 5A). The Firmicutes/Bacteroidetes ratio was higher, albeit not significantly, in the control and ABX+BALB/c FMT groups than in the ABX and ABX+C57BL/6 FMT groups (Figure S1). The abundances of Proteobacteria and Saccharibacteria were comparable among the control, ABX+BALB/c FMT, and ABX+C57BL/6 FMT groups but significantly higher than those in the ABX group. The abundance of Deferribacteres was significantly lower in the ABX and ABX+C57BL/6 FMT groups than in the control group. However, in the ABX+BALB/c FMT group, the abundance of Deferribacteres was even higher than that in the control group. The abundance of Tenericutes was significantly decreased in the ABX group but was restored in the ABX+BALB/c FMT and ABX+C57BL/6 FMT groups (Figure 5B).

At the family level, the gut microbiota profiles were shifted in the ABX, ABX+BALB/c FMT, and ABX+C57BL/6 FMT groups compared with the control group (Figure 6A). ABX treatment reduced the abundance of most families, including Lactobacillaceae, Rikenellaceae, Ruminococcaceae, and norank_o_Mollicutes_RF9. FMT restored the abundance of these families, but patterns differed between the ABX+BALB/c FMT and ABX+C57BL/6 FMT groups. In both the ABX+BALB/c FMT and ABX+C57BL/6 FMT groups, the abundance of Lactobacillaceae was restored to a level comparable to that in the control group. However, the abundance of Rikenellaceae remained significantly lower than that in the control group after FMT. The abundance of Ruminococcaceae was higher in the ABX+BALB/c FMT group than in the ABX+C57BL/6 FMT group. The abundance of norank_o_Mollicutes_RF9 in the ABX+C57BL/6 FMT group was even higher than that in the control group (Figure 6B).

At the genus level, the gut microbiota profiles were shifted even more obviously after ABX treatment and FMT (Figure 7A). The shared and unique OTUs was shown in a 4-way Venn diagram (Figure 7B). We detected five distinct response patterns to ABX treatment and FMT at the genus level (Figure 7C). Genera in the 1^st pattern, including Ruminiclostridium and Enterococcus, showed no significant shift after ABX treatment and FMT, suggesting that the colonization of these commensal bacteria is very stable and not affected by ABX treatment and FMT. Genera exhibiting the 2^nd pattern were ablated by ABX treatment and restored by FMT, with no significant difference between the ABX+BALB/c FMT and ABX+C57BL/6 FMT groups. The 2^nd pattern included Lactobacillus, Odoribacter, Candidatus Saccharimonas, Desulfovibrio, Rikenella, norank_Molicutes_RF9, and Ruminococcaceae_UCG-010. These commensal bacteria were sensitive to ABX treatment and abundant in the fecal microbiota of both BALB/c mice and C57BL/6 mice. In the 3^rd pattern, genera were ablated by ABX treatment and restored by FMT, but their abundances were significantly higher in the ABX+BALB/c FMT group than in the ABX+C57BL/6 FMT group. The 3^rd pattern included Alistipes, Ruminiclostridium, Mucispirillum, Alloprevotella, and unclassified_f_Erysipelotrichaceae. In the 4^th pattern, genera were ablated by ABX treatment and restored by FMT, but their abundances were significantly lower in the ABX+BALB/c FMT group than in the ABX+C57BL/6 FMT group. The 4^th pattern included Alloprevotella and unclassified_f_Erysipelotrichaceae. Genera exhibiting the 5^th pattern were ablated by ABX treatment, but FMT had only a limited restorative effect, with no significant difference between the ABX+BALB/c FMT and ABX+C57BL/6 FMT groups. The 5^th pattern included Ruminococcaceae_UCG-013, Anaerotruncus, and Faecalitalea. These data showed that commensal bacteria from fecal microbiota do not completely, but selectively colonize the guts of recipient mice.

To evaluate the role of microbiota depletion and reconstitution during HBV infection, mice were hydrodynamically injected with the HBV plasmid pAAV/HBV1.2 on day 48 (Figure 3A). At 7 dpi, serum HBsAg levels were comparable among the control, ABX, and ABX+BALB/c FMT groups but were significantly higher in the ABX+C57BL/6 FMT group than in the ABX+BALB/c FMT group. At 14 dpi, serum HBsAg levels in the ABX, ABX+BALB/c FMT, and ABX+C57BL/6 FMT groups were significantly higher than those in the control group. However, after 21 dpi, only the HBsAg level remained higher in the ABX+C57BL/6 FMT group than in the control group. HBsAg was cleared within 21 days in the control group but persisted for 35, 42, and >63 days in the ABX+BALB/c FMT, ABX, and ABX+C57BL/6 FMT groups, respectively (Figures 8A, B).

Serum HBeAg levels were comparable among the four groups at 7 dpi and rapidly decreased after 14 dpi in all mice (Figure 8C). HBV DNA was detected in all mice at 7 dpi and disappeared after 14 dpi, with no significant differences among the four groups (Figure 8D). HBsAb was detected after 21 dpi, and the HBsAb level was significantly higher in the control group than in the three other groups (Figure 8E). HBeAb and HBcAb levels were comparable among the four groups (Figures 8F, G). At 56 dpi, HBcAg was undetectable in livers of mice from the control, ABX, and ABX+BALB/C FMT groups but was still detectable in livers of mice from the ABX+C57BL/6 FMT group (Figure 8H).

These results indicated that gut microbiota depletion can extend the duration of HBV infection in mice. Prolonged HBV infection due to gut microbiota depletion could not be reversed by BALB/c FMT and was even further prolonged by C57BL/6 FMT.

Specific T-cell responses in the liver were assessed at 42 dpi. After stimulation with CD8+ T cell epitope, the proportion of CD8+IFNγ+ T cells was decreased in the ABX, ABX+BALB/c FMT, and ABX+C57BL/6 FMT groups. FMT from BALB/c mice upregulated the proportion of CD8+IFNγ+ T cells after ABX; however, the effect was not significant. The proportion of CD8+IFNγ+ T cells was significantly higher in the ABX+BALB/c FMT group than in the ABX+C57BL/6 FMT group (Figure 9A). The proportion of CD8+TNFα+ T cells was decreased in the ABX, ABX+BALB/c FMT, and ABX+C57BL/6 FMT groups compared to the control group. BALB/c FMT had no effect on the production of TNFα in CD8+ T cells, but FMT from C57BL/6 mice significantly decreased TNFα production in CD8+ T cells compared to ABX and ABX+BALB/c FMT (Figure 9B). The expression of IL-17 and granzyme B in CD8+T cells was comparable among the four groups (Figure S2).

Finally, we measured the expression of inhibitory immune molecules on T cells in the liver. We observed no differences in the expression of IL-10 and CTLA-4 on CD4+ and CD8+ T cells among the four groups. PD-1 expression on CD4+ and CD8+ T cells was enhanced in the ABX and ABX+C57BL/6 FMT groups (Figure 9C). These data indicated that gut microbiota depletion impairs specific T-cell responses in the liver and that these responses can be restored partially by BALB/c FMT, but are further inhibited by C57BL/6 FMT.