The Smc5/6 Complex Restricts HBV when Localized to ND10 without Inducing an Innate Immune Response and Is Counteracted by the HBV X Protein Shortly after Infection

Abstract

The structural maintenance of chromosome 5/6 complex (Smc5/6) is a restriction factor that represses hepatitis B virus (HBV) transcription. HBV counters this restriction by expressing HBV X protein (HBx), which targets Smc5/6 for degradation. However, the mechanism by which Smc5/6 suppresses HBV transcription and how HBx is initially expressed is not known. In this study we characterized viral kinetics and the host response during HBV infection of primary human hepatocytes (PHH) to address these unresolved questions. We determined that Smc5/6 localizes with Nuclear Domain 10 (ND10) in PHH. Co-localization has functional implications since depletion of ND10 structural components alters the nuclear distribution of Smc6 and induces HBV gene expression in the absence of HBx. We also found that HBV infection and replication does not induce a prominent global host transcriptional response in PHH, either shortly after infection when Smc5/6 is present, or at later times post-infection when Smc5/6 has been degraded. Notably, HBV and an HBx-negative virus establish high level infection in PHH without inducing expression of interferon-stimulated genes or production of interferons or other cytokines. Our study also revealed that Smc5/6 is degraded in the majority of infected PHH by the time cccDNA transcription could be detected and that HBx RNA is present in cell culture-derived virus preparations as well as HBV patient plasma. Collectively, these data indicate that Smc5/6 is an intrinsic antiviral restriction factor that suppresses HBV transcription when localized to ND10 without inducing a detectable innate immune response. Our data also suggest that HBx protein may be initially expressed by delivery of extracellular HBx RNA into HBV-infected cells.

Fig 1. Smc6 co-localizes with PML in uninfected human hepatocytes in vitro and in vivo.

(a) Uninfected PHH were stained for Smc6 (green) and PML (red). Nuclei were stained with DAPI (blue). (b, c) PHH were infected with HBV for 13 days. PHH were stained as for (a), and were also stained for HBV core or HBV S antigen (HBsAg) where indicated. Yellow arrows indicate HBV-infected PHH (i.e. HBV core-positive, HBsAg-positive or Smc6-negative), white arrows indicate uninfected cells (i.e. HBV core-negative, HBsAg-negative or Smc6-positive). (a)-(c) are representative images from at least n = 3 independent experiments performed with two independent PHH donors. (d) Liver tissue from one representative uninfected (top row) and one HBV-infected (bottom row) humanized mouse were stained as for (a). In the infected animals, the vast majority of human hepatocytes were positive for HBsAg and negative for Smc6 [13]. Comparable images were obtained with liver tissue from additional uninfected (n = 1) and HBV-infected (n = 4) mice. All scale bars represent 10 μm. Nuclei are outlined by white dotted lines in all images.

Fig 2. Impact of PML and Sp100 depletion on the nuclear distribution of Smc6 and the HBV replication cycle.

(a) Uninfected PHH were transfected with siRNA to the indicated gene(s), a non-targeting control siRNA (Ctrl) or were mock-transfected (mock), and incubated for 13 days. A representative Western blot is displayed on the top left. The blot has been cropped for ease of presentation. All gels were run under the same experimental conditions and full-length blots are presented in S16 Fig. Quantitation of Smc6, PML and Sp100 protein levels (all relative to GAPDH) from n = 4 independent experiments in two independent PHH donors are shown; bar height indicates mean levels expressed as a percentage of mock-transfected cells and the errors bars represent the standard error of the mean (s.e.m.). Statistical significance relative to mock is displayed. (b) Same set-up as (a). Cells were stained for Smc6 (green) and either PML or Sp100 (red). Nuclei were stained with DAPI and outlined by white dotted lines in all images. Representative confocal microscopy images from at least n = 2 independent experiments are shown. Scale bar represents 10 μm. Quantitation of the number of Smc6, PML and Sp100 nuclear foci per PHH expressed as a percentage of the control siRNA is displayed to the right of the representative images. The bar height indicates the mean and the errors bars represent the standard error of the mean; at least n = 44 nuclei were analyzed per siRNA. Statistical significance relative to the control siRNA is displayed. (c) PHH were transfected with siRNA to the indicated gene(s), a non-targeting control siRNA (Ctrl) or were mock-transfected (mock), and incubated for 6 days before infection with wild-type HBV or HBVΔX. HBeAg levels were measured on day 14 post-infection. Mean HBeAg levels ± s.e.m. of n = 6 independent experiments in two independent PHH donors are expressed as a percentage of mock-transfected (wild-type HBV) or siSmc6-transfected (HBVΔX) cells. The mean HBeAg level for mock-transfected wild-type HBV-infected PHH was 1457 ng/mL. Mean HBeAg levels for mock-transfected and siSmc6-transfected HBVΔX-infected PHH were 3 ng/mL and 2199 ng/mL, respectively. Statistical significance relative to mock is displayed. Statistical significance was tested by two-tailed t-test (PML and Sp100 foci) or one-way ANOVA with Dunnett's multiple comparison correction (all other comparisons).*p<0.05, **p<0.01 and ***p<0.001, ns: not statistically significant (p>0.05).

Fig 3. Temporal changes in HBV RNA, HBV core and Smc6 levels after HBV infection.

(a) HBV cccDNA and (b) HBV RNA levels were analyzed on the indicated days post HBV-infection by Southern blot and Northern blot, respectively. These blots have been cropped for ease of presentation; full-length blots are presented in S16 Fig. HBeAg levels are shown directly below the corresponding lane in each blot. Cell culture media was changed on days 1, 4, 7 and 10 (Southern blot experiment) and days 1, 3, 6 and 10 (Northern blot experiment). ND: not determined. (c) PHH were mock-infected (left plot) or infected with HBV (right plot). Nuclear-localized HBV core and Smc6 mean fluorescence intensity levels were measured on the indicated days post-infection by confocal microscopy. PHH were stained as in Fig 1B (upper row). Single cell quantitation was performed and each cell defined by nuclear Smc6 and HBV core status. The plots represent the number of cells in each population, expressed as a percentage of the total nuclei analyzed (n≥237 per time-point). The HBV core background was calculated as the mean + 2 x standard deviations of the red channel background signal in mock-infected PHH. The Smc6 background was calculated as the mean—2 x standard deviations of the Smc6 level (green channel) in mock-infected PHH. Values above these limits were considered positive and values below were considered negative. HBV core and Smc6 status of HBV-infected PHH was determined using background levels calculated from time-matched mock-infected controls.

Fig 4. HBx RNA is detected early after HBV infection of PHH.

RNA-Seq reads at various time-points after HBV-infection or mock-infection (-) of PHH (donor 1). The study design is outlined in Fig 6A. The x-axis denotes the HBV genome position by reference to the schematic above and the y-axis displays the sequencing coverage of the HBV genome, with the range in parentheses (note the different scale for different time-points). The time-point post-infection is shown adjacent to the y-axis. The green box denotes the HBx transcript region using the canonical transcription start site. h: hour, d: day, bp: base-pair.

Fig 5. HBx RNA is detected in cell culture-derived virus preparations and CHB patient plasma.

RNA-Seq reads in a cell-free virus stocks from HepAD38 and HepG2-H1.3x- cells prepared by PEG precipitation (top) and plasma samples from donors chronically infected with HBV genotype A or D (bottom). Note that HepG2-H1.3x- cells can produce HBVΔX virus in the absence of HBx protein because the HBV genome is integrated into the host genome and is not episomal [7]. HBV DNA levels in the patient isolates were ≥1x10⁹ copies/mL. Of note, RNA-Seq analysis of plasma samples from two additional CHB patients also identified viral RNA reads enriched within the HBx region; however, this data was not included because the maximum coverage of the HBV genome was less than 100. The x-axis denotes the HBV genome position by reference to the schematic above and the y-axis displays the sequencing coverage of the HBV genome, with the range in parentheses (note the different scale for different viruses). The green box denotes the HBx transcript region using the canonical transcription start site. bp: base-pair.

Fig 6. HBV infection does not result in a prominent global transcriptional response in PHH.

(a) Schematic summarizing the design of the study with two independent PHH donors (donors 1 and 2). Duplicate samples for each donor were analyzed on day 13 post-infection, with single samples per donor at other time-points. Cell culture media was changed on days 1, 3, 6, 8 and 10. tx: treatment, d: day, h: hour. (b) Immunofluorescence staining of HBV core (red) and nuclei (blue) at 13 days (d) post-infection with HBV (+) or mock (-). (c) Number of differentially expressed genes (DEGs) relative to time-matched mock-infected control for PHH infected with HBV (left panel) or treated with IFN-α or poly(I:C) (right panel) (red, over-expressed; blue, under-expressed). A total of 20,396 genes were analyzed. All DEGs had an absolute mean fold-change >2 and a false discovery rate (FDR) <0.05, and passed a low-expression filter.

Fig 7. HBV infection does not induce ISG expression in PHH.

Expression of top IFN-induced genes in PHH infected with HBV or treated with either IFN-α or poly(I:C) plotted by mean log₂ fold-change of two independent donors. Heatmap columns represent comparisons relative to time-matched mock-infected control. Rows represent individual genes; over-expression (red) and under-expression (blue) indicated by scale bar for log₂ fold change values. All genes (n = 100) had a mean fold-change >2 and a false discovery rate (FDR) <0.05 at 8 hours post-treatment with IFN-α relative to the time-matched mock-infected control, and passed a low-expression filter. None of these genes were identified as DEGs at any time post-infection with HBV (S2 Table).

Fig 8. HBV infection does not induce IFNs or other cytokines in PHH.

Levels of secreted (a) HBeAg and HBsAg and (b) cytokines in two independent PHH donors infected with HBV or mock-infected (left panels), or treated with either IFN-α or poly(I:C) (right panels). IP-10 (CXCL10) levels at 8 hours post-treatment with poly(I:C) and both 8 and 24 hours post-treatment with IFN-α were greater than the upper limit of quantitation (ULOQ; 1347 pg/mL), and were extrapolated from the standard curve.

Fig 9. HBVΔX infection does not induce IFNs or ISG expression in PHH.

(a) Schematic summarizing the design of the study. PHH were seeded at day (d) -4. Triplicate samples were analyzed at each time-point; samples were pooled prior to luminex analysis. Cell culture media was changed on days 1, 3, 7 and 10. Levels of (b) HBeAg and HBsAg, (c) ISGs and (d) cytokines from PHH infected with HBVΔX or mock-infected (left panels, circles) or in uninfected PHH treated with 100 IU/mL IFN-α (right panels, grey squares). For (b) and (c), the circles and squares indicate the mean, and the errors bars represent the standard error of the mean. The HBVΔX and mock-infected PHH were mock-transfected (none) or transfected with the indicated siRNA, as shown in the legend in (a). RNA levels were expressed as fold-change relative to the time-matched, mock-infected control with no siRNA. HBeAg and HBsAg levels were not measured at 4, 8, 12 or 24 hours (h) post-infection for HBVΔX-infected, mock-infected or IFN-treated PHH. IP-10 levels at 8 and 24 hours post-treatment with IFN-α were >ULOQ (1317 pg/mL), and were extrapolated from the standard curve.