Autoimmune conditions, including the family of autoimmune liver diseases (AILD), are multifactorial in origin, reflecting a complex interplay of environmental and genetic factors that evoke loss of self-tolerance and pathological autoimmune responses (1). The specific aetiological pathways engendering such diseases are poorly understood but microbial pathogens are likely contributors, potentially triggering disease in susceptible hosts either directly or indirectly via molecular mimicry (2). In contrast to conventional ‘infectious’ conditions in which the specific pathways linking pathogen to disease can be tracked and thus directly treated, the possible microbial triggers for AILD histological phenotypes, such as chronic destructive cholangitis or plasma cell-rich hepatitis, remain obscure and thus at the current time are not specifically targeted. Striking advances in available genomic technology is facilitating a different approach to this problem, such that this long-standing impasse to improved understanding and treatment of the AILDs may soon be resolved. Newly emergent genetic capabilities enable rapid characterization of collective microbial genomes within a given milieu, the so-called ‘metagenomics’ (3, 4). This new ‘omic’ is a broad scientific pursuit with an immense power to probe, in a hypothesis-free manner, the infectious aetiology of autoimmune disease. Such analyses have already been successfully applied to the characterization of viral populations in human diseases (5, 6) and the diverse bacterial populations constituting the intestinal microflora (7). The work from Andrew Mason's team in this month's journal (8) therefore stands out as an example of how broadly speaking such genome-scale approaches promise to vastly assist our understanding of the microbial contributions to complex disease pathogenesis.
In their ongoing pursuit of the infectious aetiologies of primary biliary cirrhosis (PBC), in particular, Xu et al. (8) undertook a search for foreign (pathogen-related) genomic sequences in PBC liver tissue using representational difference analysis (RDA), a strategy involving the subtractive hybridization of two polymerase chain reaction-amplified genomes. Their comparisons of liver and skin genomic DNA sequences from a patient with PBC led to the identification of an ancestral repeat sequence, MER115, which appears to be increased in copy number in patients with PBC compared with the control liver tissue. The sequence maps to an intergenic region and shows no transcriptional activity and thus its biological significance is unclear. Furthermore, the study power is limited by sample size and a lack of independent replication data. Nevertheless, the possible discovery of a PBC-associated copy number variant through the use of RDA illustrates the potential for unbiased genome-wide technologies to reveal unknown and/or unpredicted connections between molecular pathways and disease aetiopathogenesis.
While their RDA-based study was inspired in part by the authors' prior work suggesting betaretrovirus relevance to PBC (9), no microbial sequences were identified in the genomic DNA from the liver of a patient with PBC. This result may reflect the choice of skin as the competitor or ‘driver’ DNA, poor amplification or hybridization of exogenous microbial relative to host DNA as well as other technical issues potentially related to the use of RDA. Additionally, a priming infectious trigger may reside outside of the liver, e.g. lymph nodes. Thus, the failure to detect a pathogen-derived sequence in this study does not exclude the possibility that microbial elements occur in PBC patient genomes and more studies of this issue are required. Indeed, among the differentially expressed clones isolated in this same study from a liver affected with primary sclerosing cholangitis, four contained Escherichia coli one Mycoplasma hyorhinis and one hepatitis B viral sequences. Additional studies are thus warranted to continue the search for microbes that are associated with, and potentially causally relevant to, PBC pathophysiology as well as other AILDs.
Representational difference analysis is just one of the many newly emergent technologies targeted at molecular dissection of human disease. Among these new tools, a particularly promising advance is the ultrafast sequencing capacity provided by new DNA sequencing systems (10). Specifically designed to enable whole genome sequencing, the next-generation sequencing instruments provide remarkable rapidity and cost-effectiveness by massive parallelization of the sequencing process. The power of this technology is well illustrated by its recent use to resequence the tumour and skin cell-classed genomes from an acute myeloid leukaemia patient and thereby identify eight from over two million single nucleotide polymorphisms detected to be tumour specific. The identification, so precisely and elegantly, of relevant variants is clearly potentially important to disease pathogenesis and hopefully the development of new treatments (11). Ultrafast sequencing is also being widely applied to the resequencing of chromosomal regions thought to contain disease susceptibility alleles and has already enabled the sequencing/resequencing of many microbial genomes. While not yet practical for genome resequencing of large human populations, ultrafast sequencing capability has very rapidly been translated into multiple biological applications, including, e.g. genome-wide transcript profiling, chromosome immunoprecipitation and methylation analyses, all now achievable using high-throughput sequencing methods.
These sequencing-based technologies, which allow the genome and its expression to be interrogated in a single experiment, provide an unprecedented opportunity to probe and characterize the microbial populations associated with disease across specific cells and tissues. The work reported by Xu et al. (8) exemplifies the potential of increasingly sophisticated genomic technologies to unravel the genetic determinants of disease. As we rapidly approach the era of routine whole human genome sequencing, the broad application of these technologies in metagenomic studies will almost certainly alter the landscape of autoimmune disease genetics, providing the long sought-after understanding of the complex relationships between microbes and the human genome that probably drive the aberrant immune responses underpinning autoimmunity. Such knowledge will in turn enable the design of the more efficacious, target-based therapies required to improve clinical outcomes in AILD.