The natural history of primary biliary cirrhosis: of genes and cooperation
Article Outline
Primary biliary cirrhosis (PBC) is considered a model of organ specific autoimmune disease based on the classic serological findings of antimitochondrial antibodies (AMA), the focal T cell infiltrates, and the selective destruction of epithelial cells in the liver and often the salivary glands. Over the past two decades, investigators have attempted to determine the genetic basis of PBC by seeking associations with the major histocompatibility complex (MHC). Although such studies have been very useful in other autoimmune diseases, they have been disappointing in PBC. In fact, as discussed below, even though the familial risk of developing PBC can be 50–100-fold higher than that in the general population, there have been no obvious MHC associations. More recently, research has turned the attention to determining what genetic factors are involved, not in the etiology, but rather in the severity of disease in a given patient. In the paper by Corpechot et al. [1], apolipoprotein A polymorphism has been added to a growing list of genes which may influence the natural history of PBC [2]. The work presented herein is intriguing and prompts a discussion of not only severity genes, but also the larger question of etiological genes and the role of a shotgun genomic analysis.
PBC is a chronic autoimmune liver disease characterized by immune-mediated damage to the intrahepatic biliary epithelial cells, chronic cholestasis, fibrosis, and eventually cirrhosis and liver failure [3]. The presence of AMA have been known for over 35 years, but it was not until 1987 that they were identified as members of the E2 subunit of 2-oxo-acid dehydrogenase complex; over 90% of PBC patients have AMA directed against the E2 subunit of the pyruvate dehydrogenase complex (PDC-E2). Although the etiology of PBC remains unknown, there are a number of clues [3]. First, the disease is not reported in children and predominantly affects women. Second, the disease appears to be relatively more common in Westernized nations and there is some suggestion that it may be increasing. Third, the autoreactivity to mitochondrial antigens is highly conserved and PBC patient sera react even with plant chloroplasts. Fourth, the early biliary infiltrate may include eosinophilia and, in approximately 20–40% of cases, granulomas. Fifth, some patients with PBC develop antinuclear antibodies (ANAs), often these ANAs are found in the AMA negative population. Sixth, AMAs appear for many years before clinical disease is manifest. Seventh, it is often impossible, during the early phases of clinical disease, to predict the natural history or the severity in a given patient. Eighth, familial clustering of PBC is well recognized and the prevalence of PBC in first degree relatives has been determined to be between 4 and 6% [4], significantly higher than that in the general population.
Support for a genetic component in PBC also comes from the observation that family members of PBC patients are characterized by immunological abnormalities [5]. In particular, siblings are positive for other autoantibodies and have an incidence of autoimmune disease more frequent than matched controls [6], [7]. Although there is no documented available data on twins, six sets of twins with PBC are known to us and, in five of the six sets, PBC is shared. In this latter set of twins, although the disease appears discordant, the twins are reported to be an identical set. Further data on the concordance of PBC in twins would be a very important link in determining the relative roles of genetics versus environment.
In fact, most investigators have focused their attention on genetic susceptibility features, particularly associations with the MHC. For example, several studies have addressed the association of MHC in PBC in different population groups. Morling et al. studied a Danish population of patients with PBC by restriction fragment length polymorphism and found no association with HLA-DPB1 alleles in PBC, but a weak association with DR7 [8]. Likewise, a study performed in North American Caucasoid PBC patients showed no association between the HLA-DPB-1 genotype and disease [9]. In contrast, an association of DPB1*0301 and PBC was reported in both German and American Caucasoids [9], [10]. Seki et al. from Japan reported that 84% of PBC patients in their study were DPB1*0501 positive, compared with 55% of controls, and suggested that DPB1*0501 can be a primary susceptibility allele in PBC [11]. However, this data has not been reproduced by another study [12]. A very high association with HLA DR8 (79% in patients compared with 23% in the control group) was reported in Japanese PBC patients [13] as well as in various Caucasoid populations [14], and significant associations with PBC have also been described with the haplotype DR8-DQB1*0402 [15]. Strong linkage disequilibrium among HLA DR/DQ alleles may partly explain the association of the haplotype HLA DR8-DQB1*0402 with susceptibility to PBC. Associations with PBC have also been described in the MHC class III allotype C4B 2 [16] and C4A*Q0 allele [17]. Although the effect of human leukocyte antigen (HLA) types on disease susceptibility, rather than disease progression, has not been extensively studied, Gregory et al. [14] in 1993 suggested that DR typing may identify a clinical subgroup with a worse prognosis. Interestingly, a recent study has shown that in patients with HIV infection, a single amino acid change in HLA-B35, has a significant influence on the progression of AIDS [18].
In addition to HLA genes, several other genes located in the MHC, such as the genes encoding TAP I and TAP II and tumor necrosis factor-α (TNF-α), have been investigated for their possible contribution to PBC susceptibility, but the data is either incomplete or inconclusive. In addition, work has been done on natural resistance associated macrophage protein 1 (Nramp1). Nramp1 participates in the upregulation of MHC class II genes and the expression of certain cytokines which include TNF-α and interleukin (IL)-1β. Since all of these processes are implicated in autoimmune diseases, recent research has attempted to elucidate the role of the Nramp1 gene, the human homologue of the murine macrophage resistance gene, formerly lty/Lsh/Bcg, in autoimmune pathogenesis. A preliminary study of the influence of the Nramp1 promoter region polymorphism in PBC detected a higher frequency of allele 5 in PBC patients compared with normal controls [19]. Due to the small sample size and the overall rarity of this particular allele, it remains unclear at present whether the association is biologically significant. There have also been efforts placed on identifying associations with IL-10 microsatellites, cytotoxic T lymphocyte antigen 4 (CTLA-4) polymorphisms and a variety of other cytokine and chemokine genes, however, in most cases, the group sizes have been small and further work needs to be done [20]. Researchers in PBC seem to be following on the trail of work in rheumatoid arthritis and systemic lupus erythematosus in studying genes that have shown an association in these other autoimmune processes [21]. Clearly, much works needs to be done.
The advent of the genomic era, and the subsequent use of proteinomics, will allow definitive study of genes that determine susceptibility. There are two principal methods for mapping susceptibility loci. The first is based on markers which co-segregate with the disease within families (linkage) and the second relies on markers with allele frequencies that differ in individuals with the disease and those without disease (association). Linkage studies have an advantage over association studies, in that they extend over a greater genetic distance and may therefore be detected using fewer markers. Linkage may be sought within extended pedigrees by counting recombinant and non-recombinant meiosis (classical linkage analysis) or within simpler pedigrees by searching for regions of the genome shared by affected individuals (e.g. affected pair analysis).
Genome-wide scans involve the analysis of polymorphic markers, usually microsatellites (di- or trinucleotide repeat polymorphic sequences) distributed throughout the whole genome. These markers are used to determine the inheritance of different chromosomal segments in affected and unaffected members of a family. Such analysis depends on: (1), a sizable and well validated clinical resource; (2), a map of highly polymorphic markers covering the entire genome; and (3), the technology to complete the large number of genotyping required. The discovery of short sequence repeats, and subsequently, cytosine–adenosine (CAn) repeat microsatellites [22], together with fluorescent labeling, sizing [23] and mapping within the human genome [24], allow screening.
The association method uses case–control studies to compare the frequency of a particular allele in patients and healthy controls. The association between a locus and a disease may reflect a direct relationship between the marker allele and a phenotype, or alternatively, a linkage disequilibrium between the marker allele and a susceptibility locus. Thus, the strongest disease associations are with combinations of alleles at multiple loci rather than with an individual gene. The frequency of genetic polymorphism may vary in different populations and/or ethnic groups. This issue can be circumvented by two methods that use internal controls in family based studies, namely the haplotype relative risk test (HRR) [25] and the transmission disequilibrium test (TDT) [26].
Sibling pair linkage analysis is widely used in the analysis of complex traits with an unknown mode of inheritance. It requires a minimum of two affected sibling pairs who have the disease. In addition, the availability of the parents, as well as the sample size, also influence the power to map and resolve the location of susceptibility genes. Genotype can be tested for similarities by either identity by state (i.b.s.) or identity by descent (i.b.d.). In i.b.s., similarities may or may not be the result of inheritance of the marker allele, and in i.b.d., parental samples are used to prove that the markers are part of the same haplotype as that which includes the putative susceptibility gene [27].
The major problems with association tests occur because there is an infinite number of associations between disease and haplotypes that can be evaluated; controlling the proportion of false positive results requires decisions about the number and types of tests to be performed. A second problem with the application of association tests for genetically modulated disorders is the variation of ethnically diverse study participants.
The above observations should be placed in perspective with the data reported herein by Corpechot et al. [1]. Clearly, while all data contribute to our understanding of PBC, the current observation is but a stepping stone. Essentially, Corpechot et al. [1], have shown that the Apo-E allele and genotype distributions were the same when patients with PBC were compared with a control population. However, when a subgroup analysis was performed on the 72 French patients studied herein, it was noted that the ε4 allele carriers were younger and had significantly worse liver function tests than either the ε2 homozygous+heterozygous group or ε3 homozygous allele carriers. Interestingly, response to ursodeoxycholic acid (UDCA) therapy also correlated with the apo-E alleles. However, the mechanism by which the ε4 allele identifies patients with severe disease is unclear. Clearly, PBC is a cholestatic disease, and liver damage may be secondary not to the immune system, but rather to the influence of stasis and bile acids. Corpechot et al. [1] speculate that the ε4 allele leads to the accumulation of endogenous hepatic bile acids, and thus promotes injury. If this observation can be confirmed in other ethnic groups, then the results remind us that the natural history of PBC can be influenced by more than a primary immune attack. In fact, if one could fully understand all the efferent pathways involved in target cell injury, then more specific therapies could be designed. Indeed, it may be possible to develop better treatment programs without ever knowing the etiology of disease.
These data should also be placed in the context of the larger picture of what causes PBC. Obviously, PBC is a relatively uncommon disorder. Further, as studies of other alleles in PBC have demonstrated, there is often ethnic variability [28]. There have already been cooperating centers for the study of drug efficacy in PBC, including methotrexate and UDCA. If we are to succeed in defining the genetic basis of PBC, additional and similar international cooperation is needed to identify sib–sibs with PBC, twins both concordant and discordant with PBC and trios (trios include a patient with PBC and two living parents). With the remarkable advances in the human genome and growing sophistication in many of the techniques described above, it should be possible to develop a genetic map which will help toward understanding specifically not only the etiology and natural history of PBC, but also in determining generic genes which may determine loss of tolerance. Such data will auger in a great day for researchers, and especially hope for patients with PBC.
References
- . Apolipoprotein E polymorphism, a marker of disease severity in primary biliary cirrhosis?. J Hepatol. 2001;35:324–328
- Single nucleotide polymorphisms of the mannose-binding lectin are associated with susceptibility to primary biliary cirrhosis. J Autoimmun. 2001;in press
- Primary biliary cirrhosis: an orchestrated immune response against epithelial cells. Immunol Rev. 2000;174:210–225
- . The geoepidemiology of primary biliary cirrhosis: contrasts and comparisons with the spectrum of autoimmune diseases. Clin Immunol. 1999;91:206–218
- . Primary biliary cirrhosis. N Engl J Med. 1996;21:1570–1580
- Antimitochondrial antibodies in kindreds of patients with primary biliary cirrhosis: antimitochondrial antibodies are unique to clinical disease and are absent in asymptomatic family members. Hepatology. 1992;16:899–905
- . Demonstration of an intracellular copper-binding protein by orcein staining in long-standing cholestatic liver diseases. Gut. 1976;17:787–790
- DNA polymorphism of HLA class II genes in primary biliary cirrhosis. Immunogenetics. 1992;35:112–116
- . Genes within the HLA class II region confer both predisposition and resistance to primary biliary cirrhosis. Tissue Antigens. 1994;43:71–77
- . Association of primary biliary cirrhosis with the allele HLA-DPB1*0301 in a German population. Hepatology. 1995;21:398–402
- Association of primary biliary cirrhosis with human leukocyte antigen DPB1*0501 in Japanese patients. Hepatology. 1993;18:73–78
- . HLA DPB polymorphism in primary sclerosing cholangitis and primary biliary cirrhosis. Hepatology. 1995;21:959–962
- . Do primary sclerosing cholangitis (PSC) patients who express HLA-DR4 haplotype have a more rapidly progressive disease?. Abstract Hepatology. 1994;20:154A
- Primary biliary cirrhosis: contribution of HLA class II allele DR8. Q J Med. 1993;86:393–399
- . Susceptibility to primary biliary cirrhosis is associated with the HLA-DR8-DQB1*0402 haplotype. Hepatology. 1992;16:1404–1408
- . A major histocompatibility complex class III allotype (C4B 2) associated with primary biliary cirrhosis (PBC). Tissue Antigens. 1987;29:141–145
- HLA DRw8 and complement C4 deficiency as risk factors in primary biliary cirrhosis. Gastroenterology. 1991;101:1367–1373
- Effect of a single amino acid change in MHC class I molecules on the rate of progression to AIDS. N Engl J Med. 2001;344:1668–1675
- . Identification of novel alleles at a polymorphic microsatellite repeat region in the human NRAMP1 gene promoter: analysis of allele frequencies in primary biliary cirrhosis. J Med Genet. 2000;37:150–152
- . Genetic and familial considerations of primary biliary cirrhosis. Am J Gastroenterol. 2001;96:8–15
- IL10.G microsatellites mark promoter haplotypes associated with protection against the development of reactive arthritis in Finnish patients. Arthritis Rheum. 2001;44:1209–1214
- . Abundant class of human DNA polymorphisms which can be typed using the polymerase chain reaction. Am J Hum Genet. 1989;44:388–396
- Application of automated DNA sizing technology for genotyping microsatellite loci. Genomics. 1992;14:1026–1031
- A comprehensive genetic map of the human genome based on 5,264 microsatellites. See comments Nature. 1996;380:152–154
- . Case–parental control method in the search for disease-susceptibility genes. Letter; comment Am J Hum Genet. 1994;55:414–415
- . The TDT and other family-based tests for linkage disequilibrium and association. Editorial Am J Hum Genet. 1996;59:983–989
- . Asymptotic properties of affected-sib-pair linkage analysis. Am J Hum Genet. 1993;52:362–374
- . TNF polymorphisms in primary biliary cirrhosis: a critical appraisal. J Hepatol. 1999;31:366–368
PII: S0168-8278(01)00176-3
doi:10.1016/S0168-8278(01)00176-3
© 2001 European Association for the Study of the Liver. Published by Elsevier Inc. All rights reserved.
