The evidence supports a viral aetiology for primary biliary cirrhosis

Several lines of evidence implicate an infectious process in patients with primary biliary cirrhosis (PBC). For example, the disease clusters in specific geographic areas and children emigrating from a low incidence area develop the relative incidence of their new home [[1]Mason A. Zhang G. Linking human betaretrovirus infection with primary biliary cirrhosis.Clin Res Hepatol Gastroenterol. 2010; 34: 359-366Google Scholar]. Also, unrelated members in the same household, such as spouses and care providers, have been reported to develop PBC. The disease reoccurs following liver transplantation, and more potent regimens using tacrolimus have resulted in earlier and more severe recurrence. In contrast, cyclosporine protects against the development of recurrent PBC, which is interesting because it is a weaker immunosuppressant that has antiviral activity against several viral agents, including betaretroviruses [[2]Montano-Loza A.J. Wasilenko S. Bintner J. Mason A.L. Cyclosporine A inhibits in vitro replication of betaretrovirus associated with primary biliary cirrhosis.Liver Int. 2010; 30: 871-877Crossref PubMed Scopus (23) Google Scholar]. In co-cultivation studies, there are data showing that a transmissible agent from PBC patients’ lymph nodes can trigger a PBC specific phenotype with aberrant expression of mitochondrial proteins on the surface of normal biliary epithelial cells in vitro [[3]Sadamoto T. Joplin R. Keogh A. Mason A. Carman W. Neuberger J. Expression of pyruvate-dehydrogenase complex PDC-E2 on biliary epithelial cells induced by lymph nodes from primary biliary cirrhosis.Lancet. 1998; 352: 1595-1596Abstract Full Text Full Text PDF PubMed Scopus (54) Google Scholar]. A few clues emerged to suggest that a virus might be implicated in PBC. With the help of Bob Garry, my laboratory found evidence of antibody reactivity to retroviruses in PBC serum [[4]Mason A. Xu L. Guo L. Munoz S. Jaspen J.B. Bryer-Ash M. et al.Detection of retroviral antibodies in primary biliary cirrhosis and other idiopathic biliary disorders.Lancet. 1998; 351: 1620-1624Abstract Full Text Full Text PDF PubMed Scopus (138) Google Scholar] and James Neuberger’s laboratory found electron microscopy evidence of viral particles in PBC biliary epithelial cells (BEC) in collaboration with Bill Carmen [[5]Xu L. Shen Z. Guo L. Fodera B. Keogh A. Joplin R. et al.Does a betaretrovirus infection trigger primary biliary cirrhosis?.Proc Natl Acad Sci USA. 2003; 100: 84548459Crossref Scopus (204) Google Scholar]. We then used a PCR method capable of amplifying any retrovirus reverse transcriptase gene and cloned a human betaretrovirus (HBRV) sequence with 95% nucleotide homology to the mouse mammary tumor virus (MMTV) from a PBC biliary epithelium cDNA library [[5]Xu L. Shen Z. Guo L. Fodera B. Keogh A. Joplin R. et al.Does a betaretrovirus infection trigger primary biliary cirrhosis?.Proc Natl Acad Sci USA. 2003; 100: 84548459Crossref Scopus (204) Google Scholar]. We found that HBRV was detected in ∼75% PBC patient’s lymph nodes by RT-PCR and immunochemistry. However, the virus was difficult to detect in liver samples and was only observed in ∼30% of PBC samples by RT-PCR [[5]Xu L. Shen Z. Guo L. Fodera B. Keogh A. Joplin R. et al.Does a betaretrovirus infection trigger primary biliary cirrhosis?.Proc Natl Acad Sci USA. 2003; 100: 84548459Crossref Scopus (204) Google Scholar]. Indeed, the low viral burden has led to considerable controversy, as others have been unable to link HBRV with PBC. Specifically, Selmi and colleagues found no proviral DNA sequences in any patients’ livers [[6]Selmi C. Ross S.R. Ansari A.A. Invernizzi P. Podda M. Coppel R.L. et al.Lack of immunological or molecular evidence for a role of mouse mammary tumor retrovirus in primary biliary cirrhosis.Gastroenterology. 2004; 127: 493-501Abstract Full Text Full Text PDF PubMed Scopus (106) Google Scholar] and Johal found that a low frequency of patients without PBC bore proviral DNA sequences [[7]Johal H. Scott G.M. Jones R. Camaris C. Riordan S. Rawlinson W.D. Mouse mammary tumour virus-like virus (MMTV-LV) is present within the liver in a wide range of hepatic disorders and unrelated to nuclear p53 expression or hepatocarcinogenesis.J Hepatol. 2009; 50: 548-554Abstract Full Text Full Text PDF PubMed Scopus (42) Google Scholar], which is to be expected as there is general agreement that both genetic susceptibility and an environmental factor are required to trigger disease [[1]Mason A. Zhang G. Linking human betaretrovirus infection with primary biliary cirrhosis.Clin Res Hepatol Gastroenterol. 2010; 34: 359-366Google Scholar]. On close inspection all these PCR studies are concordant but interpretation varies considerably depending on the author. As betaretrovirus replication is similar to HTLV (which usually infects 1 in 1000 PBMC) it is not surprising that Selmi found no infection using a single round of PCR; whereas Johal and our laboratory found proviral DNA in a small proportion of patients using the nested PCR technique. Importantly, we used RT-PCR to detect viral RNA in the liver, which is more sensitive than PCR for detecting retroviruses and neither Selmi nor Johal employed this technique. Also no other laboratory has studied lymph nodes that are the main reservoir of HBRV infection (Fig. 1) [[5]Xu L. Shen Z. Guo L. Fodera B. Keogh A. Joplin R. et al.Does a betaretrovirus infection trigger primary biliary cirrhosis?.Proc Natl Acad Sci USA. 2003; 100: 84548459Crossref Scopus (204) Google Scholar]. In contrast, the MMTV Western blot studies were not concordant, which may be attributable to reagents and methods used. For example, we showed numerous blots where PBC patients had antibody reactivity to the betaretrovirus gp52 surface (envelope) protein [[8]Mason A. Xu L. Shen Z. Fodera B. Joplin R. Neuberger J. et al.Patients with primary biliary cirrhosis make anti-viral and anti-mitochondrial antibodies to mouse mammary tumor virus.Gastroenterology. 2004; 127 ([author reply 1864–1865]): 1863-1864Abstract Full Text Full Text PDF PubMed Scopus (24) Google Scholar]. We also demonstrated that the antiviral reactivity was specific for viral proteins by blocking anti-mitochondrial antibodies (AMA) reactivity. Selmi et al. reported that they saw AMA reactivity and dismissed other bands as background reactivity to MMTV blots. However, they only showed 1 blot using PBC serum and did not block AMA to assess whether there was serological reactivity to viral envelope proteins [[6]Selmi C. Ross S.R. Ansari A.A. Invernizzi P. Podda M. Coppel R.L. et al.Lack of immunological or molecular evidence for a role of mouse mammary tumor retrovirus in primary biliary cirrhosis.Gastroenterology. 2004; 127: 493-501Abstract Full Text Full Text PDF PubMed Scopus (106) Google Scholar]. Since then, we have taken a direct virological approach to resolve the issue of whether there is infection in patients with PBC. The human betaretrovirus has been isolated by co-culturing PBC lymph nodes with Hs578T cells in vitro. Evidence for viral isolation has been confirmed by electron microscopy, RT-PCR and reverse transcriptase activity in Hs578T supernatants [[9]Indik S. Mojiri A. Wong I. Zhang G. Wasilenko S. Mason A. Isolation of a human betaretrovirus resembling mouse mammary tumor virus from patients with primary biliary cirrhosis [Abstract].Retrovirology. 2009; 6: 55Crossref PubMed Scopus (3) Google Scholar]. We have also used linker mediated PCR to identify integration sites in vivo and in vitro, which is considered the gold standard for demonstrating retroviral infection. Accordingly, we have studied lymph node, liver and BEC samples and found that PBC patients have HBRV integration sites in ∼70% of both lymph nodes and biliary epithelium [[9]Indik S. Mojiri A. Wong I. Zhang G. Wasilenko S. Mason A. Isolation of a human betaretrovirus resembling mouse mammary tumor virus from patients with primary biliary cirrhosis [Abstract].Retrovirology. 2009; 6: 55Crossref PubMed Scopus (3) Google Scholar]. Collectively, these studies show that PBC patients harbor a transmissible retrovirus that can be localized to the site of disease in the majority of patients. Arguably, any causative agent linked to PBC should be associated with the aberrant expression of pyruvate dehydrogenase (PDC)-E2 on the cell surface of biliary epithelium and in lymphoid tissue, a highly specific PBC phenotype that is thought to lead to the formation of AMA [[10]Wasilenko S. Mason G. Mason A. Primary biliary cirrhosis, bacteria and molecular mimicry: what’s the molecule and where’s the mimic?.Liver International. 2009; 29: 779-782Crossref PubMed Scopus (17) Google Scholar]. In vivo, HBRV is detected in PBC patient’s cells with aberrant PDC-E2 expression. In vitro, homogenized PBC patients’ lymph nodes, the conditioned supernatants containing HBRV and even pure MMTV have all been shown to trigger the mitochondrial phenotype in healthy biliary epithelium, whereas control lymph node homogenates and other viruses do not [[5]Xu L. Shen Z. Guo L. Fodera B. Keogh A. Joplin R. et al.Does a betaretrovirus infection trigger primary biliary cirrhosis?.Proc Natl Acad Sci USA. 2003; 100: 84548459Crossref Scopus (204) Google Scholar]. Importantly, no in vivo patient data exist to link the mitochondrial phenotype with either bacteria or xenobiotics; indeed the idea of molecular mimicry has been circulating for over 50 years and never proven [[10]Wasilenko S. Mason G. Mason A. Primary biliary cirrhosis, bacteria and molecular mimicry: what’s the molecule and where’s the mimic?.Liver International. 2009; 29: 779-782Crossref PubMed Scopus (17) Google Scholar]. Of interest, betaretroviral infection has also been linked to the mitochondrial phenotype in several immunodeficient mouse models that spontaneously express AMA. For example, MMTV p27 capsid and gp52 envelope proteins have been detected in lymphoid tissues and biliary epithelium that also express aberrant PDC-E2. Furthermore, we have found that the development of AMA mirrors anti-MMTV production [[1]Mason A. Zhang G. Linking human betaretrovirus infection with primary biliary cirrhosis.Clin Res Hepatol Gastroenterol. 2010; 34: 359-366Google Scholar]. Indeed, MMTV has been shown to be central in triggering viral cholangitis in the NOD.c3c4 mouse model of PBC, as highly active antiretroviral therapy and MMTV neutralizing antibodies abrogate cholangitis [[1]Mason A. Zhang G. Linking human betaretrovirus infection with primary biliary cirrhosis.Clin Res Hepatol Gastroenterol. 2010; 34: 359-366Google Scholar]. Of interest, NOD.c3c4 mice treated with lamivudine and zidovudine (Combivir) develop viral resistance with mutations in the YMDD region of the reverse transcriptase gene, similar to mutations found in hepatitis B virus or HIV occurring as a result of antiviral therapy. The autoimmune model has not been that helpful for managing PBC as immunosuppression has limited efficacy and utility. Similarly, our initial attempts to treat PBC patients with the antiviral combination of Combivir have had mixed success. While initial pilot studies showed significant biochemical improvements and histological benefit in cholangitis, ductopenia and necroinflammatory score [[11]Mason A. Farr G. Xu L. Neuberger J. Pilot studies of single and combination antiretroviral therapy in patients with primary biliary cirrhosis.Am J Gastroenterol. 2004; 99: 1-8Google Scholar], ultimately we found that Combivir lacked efficacy. Indeed, some patients developed evidence of resistance to therapy with biochemical rebound [[12]Mason A.L. Wasilenko S.T. Other potential medical therapies: the use of antiviral agents to investigate and treat primary ciliary cirrhosis.Clin Liver Dis. 2008; 12 (xi): 445-460Abstract Full Text Full Text PDF PubMed Scopus (15) Google Scholar]. Also, stringent endpoints were not achieved in a randomized controlled trial with Combivir, even though patients experienced a significant improvement in hepatic biochemistry comparable to that observed with the FXR agonist, Obeticholic acid [[1]Mason A. Zhang G. Linking human betaretrovirus infection with primary biliary cirrhosis.Clin Res Hepatol Gastroenterol. 2010; 34: 359-366Google Scholar]. Using the NOD.c3c4 mouse model with MMTV infection, however, we have found that highly active antiretroviral therapy with Truvada and Kaletra is efficacious without the development of resistance. Recently, the same combination has been reported to normalize hepatic biochemistry in a PBC patient with HIV and HBRV co-infection [[13]Schembri G. Schober P. Killing two birds with one stone.Lancet. 2011; 377: 96Abstract Full Text Full Text PDF PubMed Scopus (13) Google Scholar]. Accordingly, a randomized controlled trial with Truvada and Kaletra is planned to treat patients with PBC who are unresponsive to ursodeoxycholicacid (UDCA). Indeed, it is notable that clinical trials ultimately led to the recognition that H. pylori infection caused peptic ulcer disease and the proof that a viral association with PBC may be resolved in a similar fashion. In summary, there are converging data to suggest a mechanistic link of betaretrovirus infection with the mitochondrial phenotype of PBC in co-culture studies and in a mouse model. However, we still lack firm patient data linking virus with disease. Accordingly, before we can endorse the argument that the evidence supports a viral aetiology for primary biliary cirrhosis, further studies will be required to definitively demonstrate integrations sites in diseased biliary epithelium and the serological reactivity to HBRV in the majority of patients with PBC. The suggestion by Dr. Selmi that the human betaretrovirus is a human endogenous retrovirus is incorrect; if this were the case all human DNA would be PCR-positive for the provirus, whereas Dr. Selmi found all his hepatic DNA samples negative. However, the closely related mouse mammary tumor virus is both an endogenous and exogenous virus of mice. The author declared that he does not have anything to disclose regarding funding or conflict of interest with respect to this manuscript.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.