Advertisement

Hepatitis D virus-induced interferon response and administered interferons control cell division-mediated virus spread

  • Zhenfeng Zhang
    Affiliations
    Department of Infectious Diseases, Molecular Virology, Heidelberg University, Heidelberg, Germany
    Search for articles by this author
  • Yi Ni
    Affiliations
    Department of Infectious Diseases, Molecular Virology, Heidelberg University, Heidelberg, Germany
    Search for articles by this author
  • Florian A. Lempp
    Affiliations
    Department of Infectious Diseases, Molecular Virology, Heidelberg University, Heidelberg, Germany
    Search for articles by this author
  • Lisa Walter
    Affiliations
    Department of Infectious Diseases, Molecular Virology, Heidelberg University, Heidelberg, Germany
    Search for articles by this author
  • Pascal Mutz
    Affiliations
    Department of Infectious Diseases, Molecular Virology, Heidelberg University, Heidelberg, Germany

    Division of Virus-Associated Carcinogenesis, German Cancer Research Center (DKFZ), Heidelberg, Germany
    Search for articles by this author
  • Ralf Bartenschlager
    Affiliations
    Department of Infectious Diseases, Molecular Virology, Heidelberg University, Heidelberg, Germany

    German Center for Infection Research (DZIF) - Heidelberg Partner Site, Heidelberg, Germany

    Division of Virus-Associated Carcinogenesis, German Cancer Research Center (DKFZ), Heidelberg, Germany
    Search for articles by this author
  • Stephan Urban
    Correspondence
    Corresponding author. Address: Department of Infectious Diseases, Molecular Virology, University Hospital Heidelberg, Im Neuenheimer Feld 344, D-69120 Heidelberg, Germany; Tel.: +49-6221-564902, fax: +49-6221-561946.
    Affiliations
    Department of Infectious Diseases, Molecular Virology, Heidelberg University, Heidelberg, Germany

    German Center for Infection Research (DZIF) - Heidelberg Partner Site, Heidelberg, Germany
    Search for articles by this author
Open AccessPublished:May 27, 2022DOI:https://doi.org/10.1016/j.jhep.2022.05.023

      Highlights

      • Besides HBV-dependent de novo infection, HDV also propagates through cell division.
      • IFN response efficiently suppresses cell division-mediated HDV spread.
      • The effect of IFN treatment is more efficient when HDV-induced IFN response is low.
      • IFN response destabilizes HDV RNA during cell division.

      Background & aims

      Besides HBV-dependent de novo infection, cell division-mediated spread contributes to HDV persistence and dampens the effect of antivirals that abrogate de novo infection. Nonetheless, the combination of these antivirals with interferons (IFNs) showed strong synergism in recent clinical trials, implying a complementary mode-of-action of IFNs. Therefore, we investigated the effect of IFN response on cell division-mediated HDV spread.

      Methods

      Cells infected with HDV were passaged to undergo cell division. The effect of the IFN response was evaluated by blocking HDV-induced IFN activation, by applying different IFN treatment regimens, and by adjusting HDV infection doses.

      Results

      Cell division-mediated HDV spread was highly efficient following infection of HuH7NTCP cells (defective in IFN production), but profoundly restricted in infected IFN-competent HepaRGNTCP cells. Treatment with IFN-α/-λ1 inhibited HDV spread in dividing HuH7NTCP cells, but exhibited a marginal effect on HDV replication in resting cells. Blocking the HDV-induced IFN response with the JAK1/2 inhibitor ruxolitinib or knocking down MDA5 augmented HDV spread in dividing HepaRGNTCP cells. The virus-induced IFN response also destabilized HDV RNA in dividing cells. Moreover, the effect of exogenous IFNs on cell division-mediated HDV spread was more pronounced at low multiplicities of infection with weak virus-induced IFN responses.

      Conclusions

      Both HDV-induced IFN response and exogenous IFN treatment suppress cell division-mediated HDV spread, presumably through acceleration of HDV RNA decay. Our findings demonstrate a novel mode-of-action of IFN, explain the more pronounced effect of IFN therapy in patients with lower HDV serum RNA levels, and provide insights for the development of combination therapies.

      Lay summary

      Chronic hepatitis D is a major health problem. The causative pathogen hepatitis D virus (HDV) can propagate through viral particle-mediated infection and the division of infected cells. Although viral particle-dependent infection can be blocked by recently developed drugs, therapies addressing the cell division route have not been reported. Taking advantage of relevant cell culture models, we demonstrate that the widely used immune modulator interferon can efficiently suppress HDV spread through cell division. This work unveils a new function of interferon and sheds light on potentially curative combination therapies.

      Graphical abstract

      Keywords

      Linked Article

      • Inhibiting cell-to-cell transmission to reach HDV cure: The importance of IFN-α
        Journal of HepatologyVol. 77Issue 4
        • Preview
          While HBV persists in cells through the establishment of stable episomal DNA,1 HDV is thought to be maintained in infected organisms by viral spreading. This hypothesis arose from our knowledge of the HDV life cycle that includes the production of an edited genome and a viral protein (L-HDAg) necessary for HDV morphogenesis that inhibits its replication.2 This is now enforced by Zhang and colleagues reporting a short half-life of HDV RNAs3 that would not allow HDV maintenance in non-dividing cells.
        • Full-Text
        • PDF
      See Editorial, pages 903–905

      Introduction

      Chronic hepatitis D (CHD) manifests as the most severe form of viral hepatitis with fast progression to liver fibrosis, cirrhosis, and hepatocellular carcinoma.
      • Farci P.
      • Niro G.A.
      Clinical features of hepatitis D.
      • Niro G.A.
      • Smedile A.
      • Ippolito A.M.
      • Ciancio A.
      • Fontana R.
      • Olivero A.
      • et al.
      Outcome of chronic delta hepatitis in Italy: a long-term cohort study.
      • Rizzetto M.
      • Hamid S.
      • Negro F.
      The changing context of hepatitis D.
      Different meta-analyses estimated that at least 12 million individuals suffer from CHD.
      • Stockdale A.J.
      • Kreuels B.
      • Henrion M.Y.R.
      • Giorgi E.
      • Kyomuhangi I.
      • de Martel C.
      • et al.
      The global prevalence of hepatitis D virus infection: systematic review and meta-analysis.
      ,
      • Miao Z.
      • Zhang S.
      • Ou X.
      • Li S.
      • Ma Z.
      • Wang W.
      • et al.
      Estimating the global prevalence, disease progression, and clinical outcome of hepatitis delta virus infection.
      CHD is established as a result of either simultaneous co-infection by HBV and HDV (chronification rate <5%) or HDV super-infection of patients who are already persistently infected with HBV (chronification rate ∼90%).
      • Farci P.
      • Niro G.A.
      Clinical features of hepatitis D.
      Unlike the DNA virus HBV, which replicates via reverse transcription and encodes its own reverse transcriptase/DNA polymerase, HDV is a single-stranded RNA virus with negative polarity. Its small (∼1,700 nucleotides) highly self-complementary circular genome relies on the host DNA-dependent RNA polymerases for replication.
      • Lucifora J.
      • Delphin M.
      Current knowledge on hepatitis delta virus replication.
      HDV encodes only one viral protein that is expressed in 2 isoforms, the small (S-) and large hepatitis delta antigen (L-HDAg). HDV exploits HBV-encoded envelope proteins for virus particle formation and thus, entry into hepatocytes through the hepatocyte-specific receptor sodium taurocholate co-transporting polypeptide (NTCP). At least in vitro, several envelope proteins from other viruses have been reported to be able to package HDV RNA and infect cells through respective receptors.
      • Perez-Vargas J.
      • Amirache F.
      • Boson B.
      • Mialon C.
      • Freitas N.
      • Sureau C.
      • et al.
      Enveloped viruses distinct from HBV induce dissemination of hepatitis D virus in vivo.
      Following entry, disassembly of the particle, and transport of the ribonucleoprotein complex to the nucleus, HDV replicates its RNA via a double rolling circle mechanism.
      • Lucifora J.
      • Delphin M.
      Current knowledge on hepatitis delta virus replication.
      The newly-produced genomes form ribonucleoproteins with S-HDAg and L-HDAg and bind, via L-HDAg, to the self-assembly competent HBV envelope proteins for virion assembly and egress. Remarkably, besides de novo infection, HDV can also efficiently spread without envelopment through the proliferation of infected cells.
      • Giersch K.
      • Bhadra O.D.
      • Volz T.
      • Allweiss L.
      • Riecken K.
      • Fehse B.
      • et al.
      Hepatitis delta virus persists during liver regeneration and is amplified through cell division both in vitro and in vivo.
      Accordingly, the HDV-infected hepatoma cell line HepG2NTCP allows the maintenance of HDV RNA by cell division-mediated HDV spread for >100 days.
      • Giersch K.
      • Bhadra O.D.
      • Volz T.
      • Allweiss L.
      • Riecken K.
      • Fehse B.
      • et al.
      Hepatitis delta virus persists during liver regeneration and is amplified through cell division both in vitro and in vivo.
      In HDV-infected patients who underwent liver transplantation and were under anti-HBsAg-therapy with immunoglobulins, it has been observed that HDAg-expressing hepatocytes could persist for >1 year.
      • Samuel D.
      • Zignego A.L.
      • Reynes M.
      • Feray C.
      • Arulnaden J.L.
      • David M.F.
      • et al.
      Long-term clinical and virological outcome after liver transplantation for cirrhosis caused by chronic delta hepatitis.
      ,
      • Mederacke I.
      • Filmann N.
      • Yurdaydin C.
      • Bremer B.
      • Puls F.
      • Zacher B.J.
      • et al.
      Rapid early HDV RNA decline in the peripheral blood but prolonged intrahepatic hepatitis delta antigen persistence after liver transplantation.
      Furthermore, investigations of HDV survival in mono-infected mice with transplanted primary human hepatocytes (PHHs) indicate that HDV can persist in the absence of HBV for >6 weeks.
      • Giersch K.
      • Helbig M.
      • Volz T.
      • Allweiss L.
      • Mancke L.V.
      • Lohse A.W.
      • et al.
      Persistent hepatitis D virus mono-infection in humanized mice is efficiently converted by hepatitis B virus to a productive co-infection.
      Taken together, these in vitro and in vivo observations highlight the importance of cell division-mediated spread for HDV persistence.
      Both bulevirtide (Hepcludex, formerly Myrcludex B), an EMA-approved entry inhibitor of HDV/HBV, and lonafarnib, an inhibitor under clinical development that interferes with HDV assembly, address replication steps that are crucial for de novo HDV propagation. Recent clinical trials demonstrated promising results of both drugs; however, long-term treatment is anticipated since HDV rebound was frequently observed following treatment withdrawal after 24 weeks.
      • Koh C.
      • Da B.L.
      • Surana P.
      • Huang A.
      • Kapuria D.
      • Rotman Y.
      • et al.
      A phase 2 study of lonafarnib, ritonavir and peginterferon lambda for 24 weeks: interim end-of-treatment results from the LIFT HDV study.
      ,
      • Wedemeyer H.
      • Schöneweis K.
      • Bogomolov P.
      • Voronkova N.
      • Chulanov V.
      • Stepanova T.
      • et al.
      Final results of a multicenter, open-label phase 2 clinical trial (MYR203) to assess safety and efficacy of myrcludex B in cwith PEG-interferon Alpha 2a in patients with chronic HBV/HDV co-infection.
      The failure to induce HDV negativation by both drugs within a short treatment time may be linked to their inability to block cell division-mediated HDV spread.
      Interferon alpha (IFN-α), as an off-label drug, has been used for CHD treatment since the 1980s. It can decrease HDV loads in most eligible patients; however, the effect is usually moderate and many patients relapse after drug withdrawal.
      • Heidrich B.
      • Yurdaydin C.
      • Kabacam G.
      • Ratsch B.A.
      • Zachou K.
      • Bremer B.
      • et al.
      Late HDV RNA relapse after peginterferon alpha-based therapy of chronic hepatitis delta.
      ,
      • Wedemeyer H.
      • Yurdaydin C.
      • Hardtke S.
      • Caruntu F.A.
      • Curescu M.G.
      • Yalcin K.
      • et al.
      Peginterferon alfa-2a plus tenofovir disoproxil fumarate for hepatitis D (HIDIT-II): a randomised, placebo controlled, phase 2 trial.
      Sustained virological responses occur only rarely as indicated by HBsAg negativation and seroconversion to anti-HBsAg.
      • Abdrakhman A.
      • Ashimkhanova A.
      • Almawi W.Y.
      Effectiveness of pegylated interferon monotherapy in the treatment of chronic hepatitis D virus infection: a meta-analysis.
      Recent clinical studies implementing combination therapies of pegylated IFN-α (or IFN-λ1) either with bulevirtide or lonafarnib revealed strong synergistic anti-HDV effects manifesting in faster and more profound reductions of serum HDV RNA loads during treatment.
      • Koh C.
      • Da B.L.
      • Surana P.
      • Huang A.
      • Kapuria D.
      • Rotman Y.
      • et al.
      A phase 2 study of lonafarnib, ritonavir and peginterferon lambda for 24 weeks: interim end-of-treatment results from the LIFT HDV study.
      ,
      • Wedemeyer H.
      • Schöneweis K.
      • Bogomolov P.
      • Voronkova N.
      • Chulanov V.
      • Stepanova T.
      • et al.
      Final results of a multicenter, open-label phase 2 clinical trial (MYR203) to assess safety and efficacy of myrcludex B in cwith PEG-interferon Alpha 2a in patients with chronic HBV/HDV co-infection.
      This implies that both IFNs possess a yet-unknown anti-HDV function, which synergistically augments the activity of inhibitors that interfere with entry or assembly.
      HDV replication in hepatocytes is sensed by the pattern recognition receptor melanoma differentiation antigen 5 (MDA5) and activates profound IFN-β/λ responses.
      • Zhang Z.
      • Filzmayer C.
      • Ni Y.
      • Sultmann H.
      • Mutz P.
      • Hiet M.S.
      • et al.
      Hepatitis D virus replication is sensed by MDA5 and induces IFN-beta/lambda responses in hepatocytes.
      ,
      • Wang W.
      • Lempp F.A.
      • Schlund F.
      • Walter L.
      • Decker C.C.
      • Zhang Z.
      • et al.
      Assembly and infection efficacy of hepatitis B virus surface protein exchanges in 8 hepatitis D virus genotype isolates.
      Thus, HDV infection per se already results in the release of IFNs, which may inhibit HDV replication to some extent. However, both virus-induced and exogenous IFNs have little effect on HDV replication in already infected and resting cells.
      • Zhang Z.
      • Filzmayer C.
      • Ni Y.
      • Sultmann H.
      • Mutz P.
      • Hiet M.S.
      • et al.
      Hepatitis D virus replication is sensed by MDA5 and induces IFN-beta/lambda responses in hepatocytes.
      ,
      • Lempp F.A.
      • Schlund F.
      • Rieble L.
      • Nussbaum L.
      • Link C.
      • Zhang Z.
      • et al.
      Recapitulation of HDV infection in a fully permissive hepatoma cell line allows efficient drug evaluation.
      So far it is unclear how the endogenously induced IFN response and therapeutically applied IFN-α affect HDV propagation and how the limited therapeutic outcome of IFN mono-treatment can be explained.
      Herein, we characterized cell division-mediated HDV spread in different in vitro infection models and evaluated the impact of both the HDV-(self)-induced IFN response and therapeutic IFN treatment on this HBV-independent dissemination of HDV. Moreover, we investigated the interplay between both types of IFN responses in controlling HDV and the effect of IFN response on HDV RNA stability.

      Materials and methods

      Cells, plasmids, and viruses

      NTCP-overexpressing HuH7NTCP, HepaRGNTCP, and HepaRGNTCP-derived cell lines expressing short-harpin RNAs (shRNAs) targeting MDA5 (HepaRGNTCP-shMDA5) and non-targeting shRNA control (HepaRGNTCP-shNT) were described previously.
      • Zhang Z.
      • Filzmayer C.
      • Ni Y.
      • Sultmann H.
      • Mutz P.
      • Hiet M.S.
      • et al.
      Hepatitis D virus replication is sensed by MDA5 and induces IFN-beta/lambda responses in hepatocytes.
      All cell lines used in this study were free of mycoplasma as determined by the PCR mycoplasma test kit (PromoCell, Germany).
      The 1.1-fold HDV antigenomic RNA encoding plasmid pJC126 (genotype 1) was provided by Dr. John Taylor (Fox Chase Cancer Center, USA). Similar constructs were generated by chemical synthesis (Eurofins Genomics, Ebersberg, Germany) to produce a different HDV genotype 1 (isolate Ethiopia, Genbank accession No. U81989) and genotype 3 (Genbank accession No. L22063).
      • Wang W.
      • Lempp F.A.
      • Schlund F.
      • Walter L.
      • Decker C.C.
      • Zhang Z.
      • et al.
      Assembly and infection efficacy of hepatitis B virus surface protein exchanges in 8 hepatitis D virus genotype isolates.
      Construct pT7HB2.7 encoding HBV (genotype D) envelope proteins under transcriptional control of their authentic promoters was a gift from Camille Sureau (INTS, France).
      • Sureau C.
      • Fournier-Wirth C.
      • Maurel P.
      Role of N glycosylation of hepatitis B virus envelope proteins in morphogenesis and infectivity of hepatitis delta virus.
      HDV stocks were produced by co-transfection of HuH7 cells with the HDV-encoding constructs and pT7HB2.7.
      • Lempp F.A.
      • Mutz P.
      • Lipps C.
      • Wirth D.
      • Bartenschlager R.
      • Urban S.
      Evidence that hepatitis B virus replication in mouse cells is limited by the lack of a host cell dependency factor.
      Viruses in the cell culture medium were purified using heparin affinity chromatography as described.
      • Lempp F.A.
      • Mutz P.
      • Lipps C.
      • Wirth D.
      • Bartenschlager R.
      • Urban S.
      Evidence that hepatitis B virus replication in mouse cells is limited by the lack of a host cell dependency factor.

      Cell division-mediated HDV spread

      HDV-infected cells were trypsinized at day 5 post infection (p.i.) and passaged at the indicated dilution factors to study cell division-mediated HDV spread.
      Further details related to the materials and methods are described in supplementary material and CTAT table.

      Results

      Cell division-mediated HDV spread in different hepatic cell lines

      To investigate cell division-mediated spread, susceptible cells were mono-infected by HDV (no HBV helper) and passaged after infection to undergo cell division. Due to the absence of HBV, no infectious progeny HDV could be produced in this system. Two hepatic cell lines, HuH7NTCP and HepaRGNTCP, were used to evaluate the effect of IFN response. IFN production is defective in HuH7NTCP cells, but they support downstream IFN signaling cascades upon exogenous IFN treatment. HepaRGNTCP cells are competent for both IFN production and IFN signaling. Our previous studies revealed that HDV replication and HDV-induced IFN activation are fully established at d5 p.i.
      • Zhang Z.
      • Filzmayer C.
      • Ni Y.
      • Sultmann H.
      • Mutz P.
      • Hiet M.S.
      • et al.
      Hepatitis D virus replication is sensed by MDA5 and induces IFN-beta/lambda responses in hepatocytes.
      Accordingly, we chose this point in time for passage of HDV-infected cells (1:6 dilution) and performed a second passage 5 days later (Fig. 1A). Quantification of viral replication was achieved by counting HDAg-positive cells and measuring intracellular HDV RNA at d5 p.i. and d5 post passage. As shown in Fig. 1B-C, the percentage of HDV-positive cells and intracellular HDV RNA levels only slightly (<33%) decreased after each cell passage in HuH7NTCP cells (Fig. 1C) indicating that HDV RNA replication persisted although the cells proliferated. The formation of clusters of HDV-positive cells (Fig. 1B upper images) indicates the clonal expansion of single infected cells after passage.
      Figure thumbnail gr1
      Fig. 1Cell division-mediated HDV spread is efficient in IFN production-deficient HuH7NTCP cells but suppressed in IFN-competent HepaRGNTCP cells.
      (A) Schematic of the experimental setting. (B–C) HuH7 NTCP cells and HepaRGNTCP cells were infected with HDV and passaged (1:6 dilution) every 5 days for 2 passages. At d5 p.i. (P0) and d5 post each passage, HDV-positive cells were visualized by IF staining of HDAg. Representative images are depicted in (B) and the quantification result of HDAg-positive cells using ImageJ is shown in (C, left panel) (mean ± SD, n = 6 images). Intracellular HDV RNA was quantified by RT-qPCR (C, right panel) (mean ± SD, n = 3 independent replicates). Note that HDV RNA copy numbers at single-cell level differ in the 2 cell lines used. (D-E) HDV-infected HuH7NTCP and HepaRGNTCP cells were passaged (1:800 dilution) at d5 p.i. to allow for clonal expansion. At d7 post-seeding, HDV-positive cells were visualized by IF of HDAg (D) and quantified using ImageJ (E) (mean ± SD, n = 12 images). Scale bars in (B): 200 μm. Scale bars in (D): 400 μm. n.s., not significant; ∗p <0.05; ∗∗∗p <0.001 by one-way ANOVA with Bonferroni test (C) or by Student’s t test (E). IF, immunofluorescence; IFN, interferon; LoD, limit of detection; p.i. post-infection; RT-qPCR, reverse-transcription quantitative PCR.
      We performed the same experiment using HepaRGNTCP cells and found that HDV-positive cells declined dramatically after the first cell passage: 21.2-fold (95.3%) reduction. Remarkably, the reduction was less profound after the second passage: 3.6-fold (72.2%) reduction (Fig. 1C). In contrast to HuH7NTCP cells, we could not observe extended cluster formation of HDAg-positive HepaRGNTCP cells after cell passage (Fig. 1B lower images). To verify this effect at higher cell dilutions, infected cells from both cell lines were diluted 1:800 and clonally expanded for 8 days (Fig. 1A). While in passaged HuH7NTCP cells, HDV-positive cells formed large clusters (>60 cells), only a few scattered HDAg-positive cells were observed in the expanded HepaRGNTCP cell culture (Fig. 1D-E). This indicates that cell division-mediated HDV spread is highly efficient in IFN production-deficient HuH7NTCP cells, but impaired in IFN-competent HepaRGNTCP cells.
      To monitor the HDV-induced IFN response upon HDV infection and cell passage, intracellular mRNAs of IFN-β, IFN-λ1, and radical S-adenosyl methionine domain containing 2 (RSAD2: an IFN-stimulated gene [ISG]) were quantified by reverse-transcription quantitative PCR (RT-qPCR). Consistent with previous observations,
      • Zhang Z.
      • Filzmayer C.
      • Ni Y.
      • Sultmann H.
      • Mutz P.
      • Hiet M.S.
      • et al.
      Hepatitis D virus replication is sensed by MDA5 and induces IFN-beta/lambda responses in hepatocytes.
      HepaRGNTCP, but not HuH7 NTCP cells showed profound IFN responses at d5 post HDV infection (Fig. S1). Remarkably, the IFN response in HepaRGNTCP cells declined significantly after passage, which is probably due to the overall reduction of HDV RNA replication in these cells (Fig. 1C).

      Both IFN treatment and the virus-induced IFN response restrict cell division-mediated HDV spread

      The inverse correlation between the efficiency of cell division-mediated HDV spread and the level of IFN response in HuH7NTCP and HepaRGNTCP cells implies a possible suppression of this spread by the IFN-induced antiviral state. We evaluated this hypothesis and analyzed the effect of IFN treatment on HDV-propagating HuH7NTCP cells. HDV-infected HuH7NTCP cells were passaged (1:32 dilution) at d5 p.i. and different doses of IFN-α or IFN-λ1 were applied immediately after re-seeding for 6 days (Fig. 2A). As depicted in Fig. 2B, both IFNs blocked the expansion of HDV-positive cells in a dose-dependent manner (4.2-fold [76.2%] reduction with 100 IU/ml IFN-α and 6.7-fold [85.1%] reduction with 10 ng/ml IFN-λ1). In addition to the total number of HDV-positive cells, we also observed that the overall sizes of HDV-positive cell clusters under IFN treatment were significantly smaller than those of untreated cells (Fig. 2C).
      Figure thumbnail gr2
      Fig. 2IFN treatment inhibits cell division-mediated HDV spread in HuH7NTCP cells.
      (A) Schematic of the experimental setting. HuH7NTCP cells were infected with HDV and passaged (1:32 dilution) at d5 p.i. Thereafter, cells were treated with different doses of IFN-α and IFN-λ1 for 5 days. (B) HDV-positive cells were visualized by IF staining of HDAg and quantified using ImageJ (mean ± SD, n = 12 images). (C) Representative IF images of untreated cells and cells treated with 100 IU/ml IFN-α or 10 ng/ml IFN-λ1 are shown. Scale bars: 400 μm. ∗∗∗p <0.001 by one-way ANOVA with Bonferroni test. IF, immunofluorescence; IFN, interferon.
      To elucidate whether the suppression of cell division-mediated HDV spread in HepaRGNTCP cells was due to the virus-induced endogenous IFN response, we followed 2 strategies: (i) application of ruxolitinib to suppress the JAK/STAT signaling pathway (Fig. 3A) and (ii) shRNA-mediated stable knockdown of the innate immune sensor MDA5 that selectively senses HDV RNA replication (Fig. 3E and Fig. S2).
      • Zhang Z.
      • Filzmayer C.
      • Ni Y.
      • Sultmann H.
      • Mutz P.
      • Hiet M.S.
      • et al.
      Hepatitis D virus replication is sensed by MDA5 and induces IFN-beta/lambda responses in hepatocytes.
      As shown in Fig. 3B and Fig. 3F, treatment with 5 μM of ruxolitinib and knockdown of MDA5 abolished the activation of IFNs and ISGs. Regarding their effect on cell division-mediated HDV spread, both ruxolitinib treatment (Fig. 3C-D) and MDA5 knockdown (Fig. 3G-H) efficiently restored HDV spread: 11.9-fold increase with 5 μM ruxolitinib and 13.7-fold increase in HepaRGNTCP-shMDA5 cells.
      Figure thumbnail gr3
      Fig. 3Virus-induced IFN response triggers the suppression of cell division-mediated HDV spread in HepaRGNTCP cells.
      (A-D) HepaRGNTCP cells were infected with HDV and passaged (1:6 dilution) at d5 p.i. The JAK inhibitor ruxolitinib (5 μM) was added to the medium from d1 p.i. onwards to block the activation of IFN response. (E-H) HepaRGNTCP cells expressing a shRNA against MDA5 (shMDA5), or a non-targeting shRNA control (shNT), were infected with HDV and passaged (1:6 dilution) at d5 p.i. Intracellular mRNAs of IFN-β, IFN-λ1, and RSAD2 at d5 p.i. were quantified by RT-qPCR (B, F) (mean ± SD, n = 3 independent replicates). HDV-positive cells were visualized by IF staining of HDAg (C, G) and quantified using ImageJ (D, H) (mean ± SD, n = 10 images). Scale bars: 200 μm. ∗∗p <0.01; ∗∗∗p <0.001 by Student’s t test. IF, immunofluorescence; IFN, interferon; p.i. post-infection; RT-qPCR, reverse-transcription quantitative PCR; shRNA, short-hairpin RNA.
      To exclude that IFN-induced reduction of HDAg-positive cells was caused by reduced proliferation of HDV-replicating cells, we quantified the proliferation capacity of HDV-positive and HDV-negative HepaRGNTCP cells by i) staining newly synthesized cellular DNA with BrdU, and ii) tracking cell division using the Celltrace Violet dye and flow cytometry. Even though high amounts of inocula are used, HDV can productively replicate in only a subpopulation (<30%) of the cells, so each infected well is a co-culture of HDV-positive cells and HDV-negative cells. In both assays, HDV-positive and HDV-negative cells from the HDV-infected wells showed comparable proliferation capacity, although the proliferation was slightly less than naïve cells from uninfected wells (Fig. S3).
      Taken together, these results demonstrate that both virus-induced IFN response and exogenous IFN treatment efficiently suppress cell division-mediated HDV spread. Suppression of HDV is probably mediated by clearance of HDV RNA during cell division.

      Bulevirtide and lonafarnib do not impair cell division-mediated HDV spread

      The combination of IFN with bulevirtide and lonafarnib showed a strong synergistic anti-HDV effect in clinical trials. This might be due to their different modes of action, so we next tested whether bulevirtide and lonafarnib, besides their inhibition of de novo infection, have any effect on cell division-mediated HDV spread. We passaged (1:32 dilution) HDV-infected HuH7NTCP cells at d5 p.i. and treated the cells with 0.5 μM bulevirtide or 1 μM lonafarnib for 6 days (Fig. S4A). Although the concentrations of the antivirals applied were >30x their IC50 as an entry inhibitor (bulevirtide) or a secretion inhibitor (lonafarnib),
      • Lempp F.A.
      • Schlund F.
      • Rieble L.
      • Nussbaum L.
      • Link C.
      • Zhang Z.
      • et al.
      Recapitulation of HDV infection in a fully permissive hepatoma cell line allows efficient drug evaluation.
      neither antiviral affected the efficiency of cell division-mediated HDV spread (Fig. S4B-C).

      Antiviral effect of the IFN response against HDV in dividing cells is stronger than that in resting cells

      We next compared the inhibitory activity of the IFN response in resting and dividing cells by passaging HDV-infected HepaRGNTCP cells at different dilutions (1:1, 1:2, 1:4, or 1:8). At d5 post splitting, all the cells reached confluency with similar density. However, the higher the dilution, the more division events the cells needed to reach confluency. (Fig. 4A). The number of HDAg-positive cells inversely correlated with the dilution factors. This implies that the more rounds of cell division the HDV-infected HepaRGNTCP cells undergo, the higher the loss rate of HDV-positive cells (Fig. 4B and Fig. S5B). In HuH7NTCP cells, we compared the inhibitory effect of IFN treatment on HDV replication in resting cells and dividing cells in parallel. The confluent HDV-infected HuH7NTCP cells were either kept without passaging or passaged (1:32 dilution) at d5 p.i. followed by a 6-day treatment with IFN-α or IFN-λ1 (Fig. 4C). Similarly, IFN treatment caused marginal (<10%) loss of HDV-positive cells in the case of resting cells, whereas under conditions of proliferation the number of HDV-positive cells was reduced significantly: 4.1∼6.5-fold (75.6∼84.6%) reduction (Fig. 4D and Fig. S5D).
      Figure thumbnail gr4
      Fig. 4Inhibition of HDV replication by IFN response is profound in dividing cells but marginal in resting cells.
      (A-B) HepaRGNTCP cells were infected with HDV and passaged at given dilutions at d5 p.i. (C-D) HuH7NTCP cells were infected with HDV. At d5 p.i., cells were either passaged at a 1:32 dilution or kept without passaging (resting), before being treated with 100 IU/ml IFN-α or 10 ng/ml IFN-λ1. At d5 post passage (A-B) or d6 post IFN treatment (C-D), HDV-positive cells were visualized by IF of HDAg and quantified using ImageJ. Values shown are mean ± SD (n = 8 images). Fold reduction compared to P0 (B) or untreated cells (D) are depicted on top of bars. n.s., not significant; ∗, p <0.05; ∗∗∗, p <0.001 by one-way ANOVA with Bonferroni test. IF, immunofluorescence; IFN, interferon; p.i. post-infection.
      We also validated the anti-HDV effect of IFN using patient-derived HDV genotype 1 (isolate Ethiopia) and genotype 3. IFN treatment of infected HuH7NTCP cells showed a similar effect: replication of both strains was sensitive to IFN in dividing cells but insensitive in resting cells (Fig. S6). To clarify whether stronger suppression of HDV replication in dividing cells was due to a higher level of IFN response in these cells, we compared the mRNA levels of 2 ISGs RSAD2 and MxA (also known as MX1) in dividing and resting cells after IFN treatment. As shown in Fig. S7, the mRNA levels of the ISGs upon IFN treatment in dividing cells were similar to or even lower than those in resting cells. Taken together, these results demonstrate that both virus-induced IFN response and exogenous IFN treatment exert much stronger inhibitory activity against HDV replication in dividing cells than in resting cells.

      HDV-induced IFN response accelerates HDV RNA decay during cell division

      As shown in Fig. 4, HDV replication in dividing cells is strongly affected by the IFN response. HDV replicates in the nucleus where HDV RNA might be protected from degradation. During cell division, the nuclear membrane is resolved, and HDV RNA is destabilized. To test this hypothesis, we investigated HDV RNA decay in resting and dividing HepaRGNTCP and HepaRGNTCP-shMDA5 cells. We labeled the nascent HDV RNA using 4-thiouridine (4sU) for 6 h at d5 p.i. Then one subpopulation of the cells was passaged to allow for cell division, while the other remained in the resting state (Fig. 5A). The 4sU-labeled HDV RNA was isolated at 0 h, 24 h, or 48 h post labeling, biotinylated by HPDP-biotin, and further purified using streptavidin beads. Total and 4sU-labeled HDV RNA were quantified using RT-qPCR. The result indicated that the decay of HDV RNA in HepaRGNTCP cells was significantly faster during cell division than in the resting status (10.2-fold vs. 3.5-fold reduction of 4sU-labeled HDV RNA after 48 h) (Fig. 5B). The decay of HDV RNA in both resting and dividing HepaRGNTCP-shMDA5 cells was slower (Fig. 5B). Consistently, total HDV RNA in dividing HepaRGNTCP cells also declined faster (Fig. 5C). Notably, at 0 h post 4sU labeling, the total HDV RNA in HepaRGNTCP-shMDA5 cells was only 2.1-fold of that in HepaRGNTCP cells. However, the 4sU-labeled HDV RNA in HepaRGNTCP-shMDA5 cells was 5.3-fold of that in HepaRGNTCP cells, indicating faster HDV RNA synthesis in MDA5-depleted cells. Taken together, these results demonstrate an acceleration of HDV RNA decay and an inhibition of HDV RNA synthesis by the virus-induced IFN response.
      Figure thumbnail gr5
      Fig. 5HDV-induced IFN response in HepaRGNTCP cells destabilizes viral RNA during cell division.
      (A) Schematic of the experimental setting. HepaRGNTCP and HepaRGNTCP cells with stable knockdown of MDA5 (shMDA5) were infected with HDV. At d5 p.i., nascent HDV RNA was labeled by 4sU for 6 h. The cells were either passaged (1:8 dilution, dividing cells) or kept without passage (resting cells). Total RNA was extracted immediately after 4sU labeling (0 h), or 24 h, or 48 h later. 4sU-labeled RNAs were biotinylated by HPDP-biotin and further purified using streptavidin magnetic beads. As a carrier RNA and a reference, the same amount of 4sU-labeled mouse cellular RNA was mixed with each HDV RNA sample before biotinylation and pulldown. The HDV RNA and mouse GAPDH RNA bound to the beads were quantified using RT-qPCR. The pulldown efficiency was normalized using the values of mouse GAPDH RNA. (C) Total HDV RNA at the same time points was quantified using RT-qPCR. Values shown are mean ± SD (n = 5 independent replicates). n.s. not significant; ∗p <0.05; ∗∗p <0.01; ∗∗∗p <0.001 by one-way ANOVA with Bonferroni test for comparison of groups with the same condition (resting or dividing) or by 2-way ANOVA with Bonferroni test for comparison of different conditions. 4sU, 4-thiouridine; IFN, interferon; p.i. post-infection; RT-qPCR, reverse-transcription quantitative PCR; shRNA, short-hairpin RNA.
      L-HDAg, an inhibitor of HDV replication, is produced by adenosine deaminase RNA specific 1 (ADAR1)-mediated editing of the stop codon of the HDV open reading frame. Both the constitutive form (p110) and the IFN-inducible form (p150) of ADAR1 can mediate editing of HDV RNA. We further measured the levels of ADAR1 and the ratio of L-HDAg to S-HDAg in resting and dividing HepaRGNTCP-shNT and HepaRGNTCP-shMDA5 cells. As shown in Fig. S8, the level of p150 activated by HDV-induced IFN was much lower than the constitutive p110, and the ratio of L-HDAg to S-HADg was similar in both cell lines regardless of their proliferation status. Therefore, the IFN-inducible form of ADAR1 might not be critical for suppressing HDV during cell division.

      The efficacy of cell division-mediated HDV spread in HepaRGNTCP cells inversely correlates with the level of initial infection

      Considering that the viral loads and the level of infection (defined as the percentage of actively HDV-replicating hepatocytes) vary considerably in different patients, we analyzed HDV-induced IFN response and the efficiency of cell division-mediated HDV spread at different multiplicities of infection (MOI). HepaRGNTCP and HepaRGNTCP-shMDA5 cells were infected using MOI of 1, 0.25, 0.0625, and 0.0156 IU/cell and passaged (1:8 dilution) at d5 p.i. (Fig. 6A). As expected, RT-qPCR analyses of mRNAs of IFN-β, IFN-λ1, and RSAD2 at d5 p.i. indicated an HDV dose-dependent activation of the IFN response: the level of these mRNAs in cells infected with 1 IU/cell of HDV was 11-180-fold higher than that in cells infected with 0.0156 IU/cell (Fig. 6B). However, the reduction of HDV-positive cells after passage was significantly alleviated when a low MOI was applied: 2.7-fold (63.0%) reduction for 0.0156 IU/cell vs. 11.2-fold (91.1%) reduction for 1 IU/cell (Fig. 6C). Considering the 1:8 dilution during the passage, HDV-positive cells amplified 3-fold (8/2.7) for the infection with 0.0156 IU/cell. In contrast to HepaRGNTCP cells, the reduction of HDV-positive cells after passage of HDV-infected HepaRGNTCP-shMDA5 was always around 2.7-fold regardless of the MOI (Fig. 6D). The fold change was comparable to the HepaRGNTCP cells infected with 0.0156 IU/cell of HDV. Taken together, these data indicate that cell division-mediated HDV spread is more pronounced when the levels of HDV infection and virus-induced IFN response are low.
      Figure thumbnail gr6
      Fig. 6Cell division-mediated HDV spread in HepaRGNTCP cells is more efficient at low MOI of HDV.
      (A) Schematic of the experimental setting. HepaRGNTCP cells and HepaRGNTCP cells with stable knockdown of MDA5 (shMDA5) were infected with different MOIs of HDV and passaged (1:8 dilution) at d5 p.i. (B) Levels of IFN-β, IFN-λ1, and RSAD2 mRNAs in HepaRGNTCP cells were determined at d5 p.i. using RT-qPCR. Values shown are mean ± SD (n = 3 independent replicates). (C, E) HDV-positive cells at d5 p.i. (P0) and d5 post passage (P1) were visualized by IF staining of HDAg and quantified using ImageJ. Values shown are mean ± SD (n = 8 images). (D, F) Representative images of high MOI (1 IU/cell) and low MOI (0.0156 IU/cell) HDV infection are depicted. Scale bar: 400 μm. ∗∗∗p <0.001 by Student’s t test. IF, immunofluorescence; IFN, interferon; MOI, multiplicity of infection; p.i. post-infection; RT-qPCR, reverse-transcription quantitative PCR.

      Combined effect of virus-induced IFN response and exogenous IFN treatment on suppression of cell division-mediated HDV spread

      To simulate the interplay between exogenous IFN treatment and the HDV-induced endogenous IFN response (a situation occurring most likely during IFN therapy in patients with CHD), we infected IFN-competent HepaRGNTCP cells with HDV at high (1 IU/cell) or low (0.0156 IU/cell) MOI. At day 5 p.i., cells were passaged (1:8 dilution) and either left untreated or treated with different doses of IFN-α or IFN-λ1 for another 5 days (Fig. 7A). As depicted in Fig. 7B, the effect of IFN treatment on cell division-mediated HDV spread inversely correlated with the applied MOI of HDV. IFN treatment induced much stronger suppression of cell division-mediated HDV spread when low MOI was applied during initial infection. This confirms our hypothesis that exogenous IFN treatment boosts the antiviral state to suppress cell division-mediated HDV spread. This effect is more pronounced at low MOIs that induce low IFN responses. In the case of high MOIs, HDV-induced IFN responses might have saturated the signaling pathways, and therefore the effect of IFN treatment is less pronounced. The clinical implication of these findings is that exogenous IFN treatment might be more efficient in patients with low levels of HDV infection in the liver.
      Figure thumbnail gr7
      Fig. 7Combined effect of virus-induced IFN response and exogenous IFN treatment on suppression of cell division-mediated HDV spread.
      (A) Schematic of the experimental set-up. (B) HepaRGNTCP cells were infected with either high MOI (1 IU/cell) or low MOI (0.0156 IU/cell) of HDV and passaged (1:8 dilution) at d5 p.i. Cells were either untreated or treated with different doses of IFN-α and IFN-λ1 after passage for 5 days. HDV-positive cells were visualized by IF staining of HDAg and quantified using ImageJ. Values shown are mean ± SD (n = 8 images). ∗∗∗p <0.001 by one-way ANOVA with Bonferroni test. IF, immunofluorescence; IFN, interferon; MOI, multiplicity of infection; p.i. post-infection.

      Discussion

      Cell division-mediated HDV spread was recently reported in a PHH-transplanted mouse model and in the hepatoma cell line HepG2NTCP.
      • Giersch K.
      • Bhadra O.D.
      • Volz T.
      • Allweiss L.
      • Riecken K.
      • Fehse B.
      • et al.
      Hepatitis delta virus persists during liver regeneration and is amplified through cell division both in vitro and in vivo.
      Here, we used mono-infection of HuH7NTCP and HepaRGNTCP (when IFN response is blocked) cells with HDV and confirmed this observation. Moreover, we evaluated the contribution of virus-induced and exogenously stimulated IFN signaling pathways on HDV spread by cell division. In HuH7NTCP cells, which are deficient in IFN production following HDV infection, profound cell division-mediated propagation of HDV was observed. Remarkably, this propagation was dramatically restricted in IFN-competent HepaRGNTCP cells (Fig. 1). This HBV-independent HDV spread is probably crucial for maintaining HDV in the liver, especially when de novo infection is abolished. Our findings may explain the long-term HDV persistence in patients (>1 year) with liver transplantation and anti-HBsAg therapy,
      • Samuel D.
      • Zignego A.L.
      • Reynes M.
      • Feray C.
      • Arulnaden J.L.
      • David M.F.
      • et al.
      Long-term clinical and virological outcome after liver transplantation for cirrhosis caused by chronic delta hepatitis.
      ,
      • Mederacke I.
      • Filmann N.
      • Yurdaydin C.
      • Bremer B.
      • Puls F.
      • Zacher B.J.
      • et al.
      Rapid early HDV RNA decline in the peripheral blood but prolonged intrahepatic hepatitis delta antigen persistence after liver transplantation.
      and in HDV mono-infected mice (>6 weeks) with transplanted PHHs.
      • Giersch K.
      • Helbig M.
      • Volz T.
      • Allweiss L.
      • Mancke L.V.
      • Lohse A.W.
      • et al.
      Persistent hepatitis D virus mono-infection in humanized mice is efficiently converted by hepatitis B virus to a productive co-infection.
      Moreover, a number of HDV-like agents were discovered recently from vertebrates and invertebrates (reviewed in
      • Perez-Vargas J.
      • Pereira de Oliveira R.
      • Jacquet S.
      • Pontier D.
      • Cosset F.L.
      • Freitas N.
      HDV-like viruses.
      ). Most of these agents are not associated with hepadnaviruses and do not encode L-HDAg homologs with prenylation moieties. These HDV-like agents presumably replicate in different organs but the liver and may use different helper envelope proteins,
      • Hetzel U.
      • Szirovicza L.
      • Smura T.
      • Prahauser B.
      • Vapalahti O.
      • Kipar A.
      • et al.
      Identification of a novel deltavirus in Boa constrictors.
      ,
      • Szirovicza L.
      • Hetzel U.
      • Kipar A.
      • Martinez-Sobrido L.
      • Vapalahti O.
      • Hepojoki J.
      Snake deltavirus utilizes envelope proteins of different viruses to generate infectious particles.
      which is supported by a recent in vitro study showing the usage of non-hepadnaviral envelope proteins by HDV.
      • Perez-Vargas J.
      • Amirache F.
      • Boson B.
      • Mialon C.
      • Freitas N.
      • Sureau C.
      • et al.
      Enveloped viruses distinct from HBV induce dissemination of hepatitis D virus in vivo.
      Alternatively, these agents may propagate through division of infected cells.
      • Paraskevopoulou S.
      • Pirzer F.
      • Goldmann N.
      • Schmid J.
      • Corman V.M.
      • Gottula L.T.
      • et al.
      Mammalian deltavirus without hepadnavirus coinfection in the neotropical rodent Proechimys semispinosus.
      This implies that cell division-mediated spread is likely an evolutionarily conserved default persistence strategy of delta-like agents, including HDV.
      Here, we demonstrate the suppression of cell division-mediated HDV spread by IFN. The inhibitory activity is much stronger compared to the inhibition of intracellular HDV RNA replication in resting cells (Fig. 4) or the inhibition observed during de novo infection.
      • Zhang Z.
      • Filzmayer C.
      • Ni Y.
      • Sultmann H.
      • Mutz P.
      • Hiet M.S.
      • et al.
      Hepatitis D virus replication is sensed by MDA5 and induces IFN-beta/lambda responses in hepatocytes.
      ,
      • Lempp F.A.
      • Schlund F.
      • Rieble L.
      • Nussbaum L.
      • Link C.
      • Zhang Z.
      • et al.
      Recapitulation of HDV infection in a fully permissive hepatoma cell line allows efficient drug evaluation.
      This broadens our understanding of the multi-functional antiviral activities of IFN. Nonetheless, the mechanism underlying the IFN-mediated anti-HDV effect is still unclear. Besides modulating the function of immune cells, e.g. natural killer cells and T cells, which is not the scope of this study, direct anti-HDV effects induced by IFNs were observed previously in cell culture models,
      • Zhang Z.
      • Filzmayer C.
      • Ni Y.
      • Sultmann H.
      • Mutz P.
      • Hiet M.S.
      • et al.
      Hepatitis D virus replication is sensed by MDA5 and induces IFN-beta/lambda responses in hepatocytes.
      ,
      • Han Z.
      • Nogusa S.
      • Nicolas E.
      • Balachandran S.
      • Taylor J.
      Interferon impedes an early step of hepatitis delta virus infection.
      mouse models
      • Giersch K.
      • Homs M.
      • Volz T.
      • Helbig M.
      • Allweiss L.
      • Lohse A.W.
      • et al.
      Both interferon alpha and lambda can reduce all intrahepatic HDV infection markers in HBV/HDV infected humanized mice.
      ,
      • Suarez-Amaran L.
      • Usai C.
      • Di Scala M.
      • Godoy C.
      • Ni Y.
      • Hommel M.
      • et al.
      A new HDV mouse model identifies mitochondrial antiviral signaling protein (MAVS) as a key player in IFN-beta induction.
      and patients.
      • Mentha N.
      • Clement S.
      • Negro F.
      • Alfaiate D.
      A review on hepatitis D: from virology to new therapies.
      ,
      • Alavian S.M.
      • Tabatabaei S.V.
      • Behnava B.
      • Rizzetto M.
      Standard and pegylated interferon therapy of HDV infection: a systematic review and meta- analysis.
      IFN response was found to inhibit early steps of HDV infection.
      • Zhang Z.
      • Filzmayer C.
      • Ni Y.
      • Sultmann H.
      • Mutz P.
      • Hiet M.S.
      • et al.
      Hepatitis D virus replication is sensed by MDA5 and induces IFN-beta/lambda responses in hepatocytes.
      ,
      • Han Z.
      • Nogusa S.
      • Nicolas E.
      • Balachandran S.
      • Taylor J.
      Interferon impedes an early step of hepatitis delta virus infection.
      However, high IFN doses are needed to achieve such an effect.
      • Han Z.
      • Nogusa S.
      • Nicolas E.
      • Balachandran S.
      • Taylor J.
      Interferon impedes an early step of hepatitis delta virus infection.
      After the establishment of replication, both the HDV-induced IFN response and exogenous IFN treatment only moderately inhibit HDV replication, and long-term treatment is needed to achieve significant inhibition.
      • Zhang Z.
      • Filzmayer C.
      • Ni Y.
      • Sultmann H.
      • Mutz P.
      • Hiet M.S.
      • et al.
      Hepatitis D virus replication is sensed by MDA5 and induces IFN-beta/lambda responses in hepatocytes.
      IFN response in dividing cells is not yet well characterized. Fig. S3 shows that the proliferative capacity of HDV-replicating cells is comparable to that of HDV-negative cells, indicating that the suppression is likely mediated by clearance of HDV markers during mitosis. Studies have demonstrated that IFN and ISG production is downregulated during mitosis due to the condensation of chromosomes and consequently suppression of cellular transcription in the G2/M phase.
      • Bressy C.
      • Droby G.N.
      • Maldonado B.D.
      • Steuerwald N.
      • Grdzelishvili V.Z.
      Cell cycle arrest in G2/M phase enhances replication of interferon-sensitive cytoplasmic RNA viruses via inhibition of antiviral gene expression.
      One possible explanation for the strong suppression of HDV by IFN response during cell division is that the nucleus-resident HDV replication intermediates become accessible to cellular nucleases including the ones induced by IFNs during mitosis. HDV RNA stability assays in HepaRGNTCP and HepaRGNTCP -shMDA5 cells confirmed that the virus-induced IFN response destabilizes HDV RNA and likely also inhibits HDV RNA synthesis during cell division (Fig. 5), while the effect is much smaller in resting cells. Further investigations are needed to identify the responsible ISG(s) and the mechanism of action.
      Using IFN-competent HepaRGNTCP cells, we found that the effect of IFN treatment on cell division-mediated HDV spread is more pronounced when the applied MOI and HDV-induced response are low (Fig. 6). Therefore, it is, reasonable to hypothesize that similar conditions apply to patients, i.e. the level of HDV-induced IFN response correlates with the HDV load. Indeed, clinical trials have shown that the reduction of serum HDV RNA levels under IFN treatment was generally faster in patients with CHD and lower viral titers at baseline.
      • Heidrich B.
      • Yurdaydin C.
      • Kabacam G.
      • Ratsch B.A.
      • Zachou K.
      • Bremer B.
      • et al.
      Late HDV RNA relapse after peginterferon alpha-based therapy of chronic hepatitis delta.
      ,
      • Wedemeyer H.
      • Yurdaydin C.
      • Hardtke S.
      • Caruntu F.A.
      • Curescu M.G.
      • Yalcin K.
      • et al.
      Peginterferon alfa-2a plus tenofovir disoproxil fumarate for hepatitis D (HIDIT-II): a randomised, placebo controlled, phase 2 trial.
      ,
      • Lutterkort G.L.
      • Wranke A.
      • Hengst J.
      • Yurdaydin C.
      • Stift J.
      • Bremer B.
      • et al.
      Viral dominance patterns in chronic hepatitis delta determine early response to interferon alpha therapy.
      In individual patients, the reduction was less significant in the first course of treatment (e.g. 24 weeks) when the viral loads were high, while it became more significant in later courses of treatment when the viral loads had decreased.
      • Heidrich B.
      • Yurdaydin C.
      • Kabacam G.
      • Ratsch B.A.
      • Zachou K.
      • Bremer B.
      • et al.
      Late HDV RNA relapse after peginterferon alpha-based therapy of chronic hepatitis delta.
      These in vitro and in vivo observations stress the importance of taking viral loads and the levels of virus-induced IFN response into consideration when using IFN therapy.
      Recent combination therapies like bulevirtide plus peg-IFN-α (Myr203) or lonafarnib plus peg-IFN-λ1 (LIFT) demonstrated strong synergistic anti-HDV effects in terms of faster and more profound reduction of serum HDV RNA load.
      • Koh C.
      • Da B.L.
      • Surana P.
      • Huang A.
      • Kapuria D.
      • Rotman Y.
      • et al.
      A phase 2 study of lonafarnib, ritonavir and peginterferon lambda for 24 weeks: interim end-of-treatment results from the LIFT HDV study.
      ,
      • Wedemeyer H.
      • Schöneweis K.
      • Bogomolov P.
      • Voronkova N.
      • Chulanov V.
      • Stepanova T.
      • et al.
      Final results of a multicenter, open-label phase 2 clinical trial (MYR203) to assess safety and efficacy of myrcludex B in cwith PEG-interferon Alpha 2a in patients with chronic HBV/HDV co-infection.
      Here, we reveal that both bulevirtide and lonafarnib do not affect the cell division-mediated spread (Fig. S4). Therefore, the synergism observed in the clinical combination treatments might be due to inhibition of both extracellular spread (targeted by bulevirtide and lonafarnib) and cell division-mediated HDV spread (exerted by peg-IFN-α and peg-IFN-λ1).
      Taken together, we demonstrate that: i) both the HDV-induced IFN response and exogenous IFN treatment efficiently suppress cell division-mediated HDV spread, but only moderately impair intracellular HDV replication in resting cells; ii) the effect of IFN treatment inversely correlates with the amount of HDV-infected hepatocytes and the level of virus-induced IFN response; and iii) virus-induced IFN response accelerates HDV RNA decay during cell division. These findings unveil a novel mode-of-action of IFN against HDV, help to explain the clinical observations with IFN treatment, and provide insights into the development of novel combination therapies.

      Abbreviations

      4sU, 4-thiouridine; ADAR1, adenosine deaminase RNA specific 1; CHD, chronic hepatitis D; HBcAg, hepatitis B virus core antigen; HDAg, hepatitis delta antigen; IFN, interferon; ISG, IFN-stimulated gene; MDA5, melanoma differentiation antigen 5; MOI, multiplicity of infection; NTCP, human sodium taurocholate co-transporting polypeptide; PHHs, primary human hepatocytes; p.i., post infection; RSAD2, radical S-adenosyl methionine domain containing 2; RT-qPCR, reverse-transcription quantitative PCR; shRNA, short-hairpin RNA.

      Financial support

      This work was funded by German Center for Infectious Research (DZIF), TTU Hepatitis Projects 5.709 and 5.822, and by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation)– Projektnummer 272983813 – TRR 179 (TP15 and TP09).

      Authors’ contributions

      Conception and design of the study: Z.Z., Y.N., and S.U.; generation, collection, assembly, analysis and interpretation of data: Z.Z., Y.N., FA.L., L.W., P.M., R.B., and S.U.; drafting of the manuscript: Z.Z, and S.U.; Critical revision and approval of the final version of the manuscript: Z.Z., Y.N., FA.L., L.W., P.M., R.B., and S.U.

      Data availability statement

      All data, analytic methods, and study materials supporting the findings of this study are available in the article along with supplementary material and CTAT table.

      Conflict of interest

      Prof. Dr. Stephan Urban holds patents on bulevirtide/Hepcludex. Other authors who have taken part in this study declared that they do not have anything to disclose regarding funding or conflict of interest with respect to this manuscript.
      Please refer to the accompanying ICMJE disclosure forms for further details.

      Acknowledgments

      We thank Wenshi Wang, Talisa Richardt, Anna-Clara Franke, Christa Kuhn, Nadine Gillich, Marco Binder, Sandra Wüst, and Ombretta Colasanti for providing materials and help for the experiments. We thank Franziska Schlund for lab organization and cell culture support. We are especially indebted to Volker Lohmann, Alessia Ruggieri, and Macro Binder for continuous intellectual input and help with materials.

      Supplementary data

      The following are the supplementary data to this article:

      References

        • Farci P.
        • Niro G.A.
        Clinical features of hepatitis D.
        Semin Liver Dis. 2012; 32: 228-236
        • Niro G.A.
        • Smedile A.
        • Ippolito A.M.
        • Ciancio A.
        • Fontana R.
        • Olivero A.
        • et al.
        Outcome of chronic delta hepatitis in Italy: a long-term cohort study.
        J Hepatol. 2010; 53: 834-840
        • Rizzetto M.
        • Hamid S.
        • Negro F.
        The changing context of hepatitis D.
        J Hepatol. 2021; 74: 1200-1211
        • Stockdale A.J.
        • Kreuels B.
        • Henrion M.Y.R.
        • Giorgi E.
        • Kyomuhangi I.
        • de Martel C.
        • et al.
        The global prevalence of hepatitis D virus infection: systematic review and meta-analysis.
        J Hepatol. 2020; 73: 523-532
        • Miao Z.
        • Zhang S.
        • Ou X.
        • Li S.
        • Ma Z.
        • Wang W.
        • et al.
        Estimating the global prevalence, disease progression, and clinical outcome of hepatitis delta virus infection.
        J Infect Dis. 2020; 221: 1677-1687
        • Lucifora J.
        • Delphin M.
        Current knowledge on hepatitis delta virus replication.
        Antivir Res. 2020; 179104812
        • Perez-Vargas J.
        • Amirache F.
        • Boson B.
        • Mialon C.
        • Freitas N.
        • Sureau C.
        • et al.
        Enveloped viruses distinct from HBV induce dissemination of hepatitis D virus in vivo.
        Nat Commun. 2019; 10: 2098
        • Giersch K.
        • Bhadra O.D.
        • Volz T.
        • Allweiss L.
        • Riecken K.
        • Fehse B.
        • et al.
        Hepatitis delta virus persists during liver regeneration and is amplified through cell division both in vitro and in vivo.
        Gut. 2019; 68: 150-157
        • Samuel D.
        • Zignego A.L.
        • Reynes M.
        • Feray C.
        • Arulnaden J.L.
        • David M.F.
        • et al.
        Long-term clinical and virological outcome after liver transplantation for cirrhosis caused by chronic delta hepatitis.
        Hepatology. 1995; 21: 333-339
        • Mederacke I.
        • Filmann N.
        • Yurdaydin C.
        • Bremer B.
        • Puls F.
        • Zacher B.J.
        • et al.
        Rapid early HDV RNA decline in the peripheral blood but prolonged intrahepatic hepatitis delta antigen persistence after liver transplantation.
        J Hepatol. 2012; 56: 115-122
        • Giersch K.
        • Helbig M.
        • Volz T.
        • Allweiss L.
        • Mancke L.V.
        • Lohse A.W.
        • et al.
        Persistent hepatitis D virus mono-infection in humanized mice is efficiently converted by hepatitis B virus to a productive co-infection.
        J Hepatol. 2014; 60: 538-544
        • Koh C.
        • Da B.L.
        • Surana P.
        • Huang A.
        • Kapuria D.
        • Rotman Y.
        • et al.
        A phase 2 study of lonafarnib, ritonavir and peginterferon lambda for 24 weeks: interim end-of-treatment results from the LIFT HDV study.
        AASLD. 2019;
        • Wedemeyer H.
        • Schöneweis K.
        • Bogomolov P.
        • Voronkova N.
        • Chulanov V.
        • Stepanova T.
        • et al.
        Final results of a multicenter, open-label phase 2 clinical trial (MYR203) to assess safety and efficacy of myrcludex B in cwith PEG-interferon Alpha 2a in patients with chronic HBV/HDV co-infection.
        J Hepatol. 2019; 70: e81
        • Heidrich B.
        • Yurdaydin C.
        • Kabacam G.
        • Ratsch B.A.
        • Zachou K.
        • Bremer B.
        • et al.
        Late HDV RNA relapse after peginterferon alpha-based therapy of chronic hepatitis delta.
        Hepatology. 2014; 60: 87-97
        • Wedemeyer H.
        • Yurdaydin C.
        • Hardtke S.
        • Caruntu F.A.
        • Curescu M.G.
        • Yalcin K.
        • et al.
        Peginterferon alfa-2a plus tenofovir disoproxil fumarate for hepatitis D (HIDIT-II): a randomised, placebo controlled, phase 2 trial.
        Lancet Infect Dis. 2019; 19: 275-286
        • Abdrakhman A.
        • Ashimkhanova A.
        • Almawi W.Y.
        Effectiveness of pegylated interferon monotherapy in the treatment of chronic hepatitis D virus infection: a meta-analysis.
        Antivir Res. 2021; 185104995
        • Zhang Z.
        • Filzmayer C.
        • Ni Y.
        • Sultmann H.
        • Mutz P.
        • Hiet M.S.
        • et al.
        Hepatitis D virus replication is sensed by MDA5 and induces IFN-beta/lambda responses in hepatocytes.
        J Hepatol. 2018; 69: 25-35
        • Wang W.
        • Lempp F.A.
        • Schlund F.
        • Walter L.
        • Decker C.C.
        • Zhang Z.
        • et al.
        Assembly and infection efficacy of hepatitis B virus surface protein exchanges in 8 hepatitis D virus genotype isolates.
        J Hepatol. 2021;
        • Lempp F.A.
        • Schlund F.
        • Rieble L.
        • Nussbaum L.
        • Link C.
        • Zhang Z.
        • et al.
        Recapitulation of HDV infection in a fully permissive hepatoma cell line allows efficient drug evaluation.
        Nat Commun. 2019; 10: 2265
        • Sureau C.
        • Fournier-Wirth C.
        • Maurel P.
        Role of N glycosylation of hepatitis B virus envelope proteins in morphogenesis and infectivity of hepatitis delta virus.
        J Virol. 2003; 77: 5519-5523
        • Lempp F.A.
        • Mutz P.
        • Lipps C.
        • Wirth D.
        • Bartenschlager R.
        • Urban S.
        Evidence that hepatitis B virus replication in mouse cells is limited by the lack of a host cell dependency factor.
        J Hepatol. 2016; 64: 556-564
        • Perez-Vargas J.
        • Pereira de Oliveira R.
        • Jacquet S.
        • Pontier D.
        • Cosset F.L.
        • Freitas N.
        HDV-like viruses.
        Viruses. 2021; 13
        • Hetzel U.
        • Szirovicza L.
        • Smura T.
        • Prahauser B.
        • Vapalahti O.
        • Kipar A.
        • et al.
        Identification of a novel deltavirus in Boa constrictors.
        mBio. 2019; 10: e00014-e19
        • Szirovicza L.
        • Hetzel U.
        • Kipar A.
        • Martinez-Sobrido L.
        • Vapalahti O.
        • Hepojoki J.
        Snake deltavirus utilizes envelope proteins of different viruses to generate infectious particles.
        mBio. 2020; 11
        • Paraskevopoulou S.
        • Pirzer F.
        • Goldmann N.
        • Schmid J.
        • Corman V.M.
        • Gottula L.T.
        • et al.
        Mammalian deltavirus without hepadnavirus coinfection in the neotropical rodent Proechimys semispinosus.
        Proc Natl Acad Sci U S A. 2020; 117: 17977-17983
        • Han Z.
        • Nogusa S.
        • Nicolas E.
        • Balachandran S.
        • Taylor J.
        Interferon impedes an early step of hepatitis delta virus infection.
        PLoS One. 2011; 6e22415
        • Giersch K.
        • Homs M.
        • Volz T.
        • Helbig M.
        • Allweiss L.
        • Lohse A.W.
        • et al.
        Both interferon alpha and lambda can reduce all intrahepatic HDV infection markers in HBV/HDV infected humanized mice.
        Sci Rep. 2017; 7: 3757
        • Suarez-Amaran L.
        • Usai C.
        • Di Scala M.
        • Godoy C.
        • Ni Y.
        • Hommel M.
        • et al.
        A new HDV mouse model identifies mitochondrial antiviral signaling protein (MAVS) as a key player in IFN-beta induction.
        J Hepatol. 2017; 67: 669-679
        • Mentha N.
        • Clement S.
        • Negro F.
        • Alfaiate D.
        A review on hepatitis D: from virology to new therapies.
        J Adv Res. 2019; 17: 3-15
        • Alavian S.M.
        • Tabatabaei S.V.
        • Behnava B.
        • Rizzetto M.
        Standard and pegylated interferon therapy of HDV infection: a systematic review and meta- analysis.
        J Res Med Sci. 2012; 17: 967-974
        • Bressy C.
        • Droby G.N.
        • Maldonado B.D.
        • Steuerwald N.
        • Grdzelishvili V.Z.
        Cell cycle arrest in G2/M phase enhances replication of interferon-sensitive cytoplasmic RNA viruses via inhibition of antiviral gene expression.
        J Virol. 2019; 93 (e01885-1818)
        • Lutterkort G.L.
        • Wranke A.
        • Hengst J.
        • Yurdaydin C.
        • Stift J.
        • Bremer B.
        • et al.
        Viral dominance patterns in chronic hepatitis delta determine early response to interferon alpha therapy.
        J Viral Hepat. 2018; 25: 1384-1394