Genetics, Genomics, Gene Expression Profiling, and Epigenetics in Sjögren’s Syndrome

1

Historical Background

Henrik Sjögren, a Swedish ophthalmologist, described the condition that bears his name in 1933. He coined the term keratoconjunctivitis sicca to distinguish the ocular surface features from those seen in vitamin-A deficiency (xerophthalmia). The term xerostomia is, however, used to describe oral dryness. Henri Gougerot, a French dermatologist, had already described three patients with sicca syndrome and salivary gland atrophy in 1925. Jan Mikulicz-Radecki, an Austro-Polish surgeon described the histological features in 1892.

The distinction between primary Sjögren’s syndrome (pSS) and secondary Sjögren’s syndrome (SS) was set out in the 1960s. Also in the 1960s, the link with mucosa-associated lymphoid tissue (MALT) B-cell lymphoma was reported, Chisholm and Mason described their scoring system for the histological features of salivary gland biopsies in pSS, and the anti-Ro/SSA and anti-La/SSB antibodies were first identified. Subsequently these antibodies were shown to be associated with HLA-DR3 and other human leukocyte antigen (HLA) haplotypes and the neonatal lupus syndrome.

The glandular features and management of pSS and secondary SS are generally regarded as being similar, although fibrosis, for example, is a more typical feature in scleroderma-related secondary SS. Unless otherwise stated, this chapter will focus on pSS.

2

Epidemiology, Prevalence, and Classification Criteria

Sjögren’s syndrome is a worldwide disease with a strong female bias that traditionally reported as 9:1 but is possibly as high as13:1. Typically, pSS presents in the fifth or sixth decade but can present at any age including, rarely, in childhood.

Initial research into the prevalence of pSS came up with widely differing estimates as low as 0.08% (1 in 1250) using the San Diego (CA) criteria, or as high as 3% of the adult female population in a community-based study in the United Kingdom. One explanation for this variation was the use of different, more permissive classification criteria in the latter study.

The most widely used classification criteria for pSS are the American-European Consensus Group (AECG) criteria. These require a combination of oral and/or ocular dryness symptoms/signs, and at least one positive test for anti-Ro/SSA and anti-La/SSB antibodies, and/or a minor labial salivary gland biopsy with features of focal periductal lymphocytic sialadenitis. More recent studies using the AECG criteria have estimated the community prevalence at 0.1% to 0.4% and 0.04% to 0.05% in the hospital setting.

More recently, a preliminary American College of Rheumatology criteria set requires two out of three components, including positive autoantibodies (also including antinuclear antibody [ANA] ≥1:320 and a positive rheumatoid factor), a positive minor labial salivary gland biopsy, and/or an abnormal Ocular Staining Score ≥3. In interpreting genetic and other research studies, it is essential to have agreed classification criteria so that there is confidence that participants in a study have the specified condition. At the present time an international group of experts is bringing these criteria together to produce American College of Rheumatology–European League Against Rheumatism consensus criteria.

3

Immunopathology of Sjögren’s Syndrome

A detailed description of the immunopathology of pSS is beyond the scope of this review. In order to place the genetic data in context it is helpful, however, to be aware of the broad themes underpinning potential pathogenetic mechanisms. The triggering factor for pSS is unknown. One dominant hypothesis is that of “autoimmune epitheliitis.” In this hypothesis, an unknown trigger upregulates HLA class II expression and costimulatory factors on the ductal epithelial cells in secretory glands along with stimulating production of proinflammatory cytokines and other molecules. The HLA molecules present an unknown set of antigens to T cells, thus further upregulating the immune response, which then becomes chronic through the development of focal lymphocyte aggregation and the production of chemokines and other molecules that maintain these structures, resulting in persistent nonresolving inflammation. In some patients these lymphoid aggregates take on features similar to those seen in secondary lymphoid organs, with marked B-cell expansion and fully formed germinal centers (GCs) with light and dark zones thought capable of supporting the development of high-affinity, somatically mutated B cells. The presence of GCs may indicate a greater likelihood of subsequent B-cell lymphoma development, likely resulting from the genetic instability associated with DNA hypermutation (see above). Systemic B-cell activation with autoantibody production and hypergammaglobulinemia is also a hallmark of pSS (see above). The role of both T- and B-cell immunity would place genes encoding molecules involved in a broad range of adaptive immune responses, including HLA, but also antigen processing, cytokines, chemokines, B-cell and T-cell regulation, and receptor molecules as potential candidates for susceptibility to pSS.

Along with T cells and B cells, another key theme is of upregulation of the innate immune system. This “first line of defense” is triggered by, for example, common bacterial lipopolysaccharides and other highly conserved pathogen-associated molecular patterns and molecules released from damaged cells (danger-associated molecular patterns, also referred to as alarmins ) to quickly upregulate the immune system in a non-HLA, non–antigen-specific manner while waiting for the adaptive antigen-driven immune response, which may take days or weeks to mature. Pattern recognition receptors such as the Toll-like receptors (TLRs) recognize these molecules, leading to upregulation of type I and type II interferons that have a broad and rapid stimulatory effect on the immune system affecting T-cells, B-cells, and also natural killer cells. An upregulated “interferon (IFN) signature” has been recognized for some time in conditions such as pSS, systemic lupus erythematosus (SLE) and other connective tissue diseases. We will review how studies of genetics and epigenetics, transcriptomics, proteomics, and metabolomics have informed, or have the potential to inform, our understanding of pSS. Hypothetical relationships between these “-omics” and the etiology of complex traits are illustrated in Fig. 8.1 .

4

“Traditional” Genetics: Human Leukocyte Antigens

In the late 1950s and 1960s the HLAs were identified as key components of transplant rejection. To summarize two decades of research that established our knowledge of the basic components of the immune system, Class I HLAs are found on the surface of almost all cells, whereas class II HLAs are found predominantly on cells of the immune system. The genes encoding the human HLA proteins are found on chromosome 6. There are three main HLA Class I gene loci ( HLA-A , HLA-B , and HLA-C ), and three main Class II HLA gene loci ( HLA-DR , HLA-DQ , and HLA-DP ), but there are over 200 other genes in this region of chromosome 6. The proximity of these loci to each other mean that certain combinations such as the HLA-A1 , B8 , DR3 , DQ2 combination (haplotype) are typically inherited as a “block” together (linkage disequilibrium). Class I molecules typically present antigens derived from inside cells to cytotoxic (killer) T cells. In virally infected cells, this allows the cytotoxic T cells to recognize infected cells and destroy them. Class II molecules are typically found on antigen-presenting cells and present antigens originally from outside of cells (although they are then internalized and processed before presentation on the cell surface by HLA class II molecules), such as bacterial antigens, to helper T cells that then stimulate B cells to support antibody production that recognizes the bacterial molecules and facilitates their removal by the immune system.

The HLA proteins were found to have multiple variants at each locus (eg, HLA-DR1 , DR2 , DR3 , etc.). This was originally characterized by antibody reactivity analysis (serotyping) but subsequently, after the discovery of the polymerase chain reaction (PCR), by genotyping. Although the genotyping approach is more precise, serotyping can identify antigenic specificities that are common to several genotypes.

The identification of the HLA antigens was followed by the discovery that certain autoimmune diseases were linked to particular HLA types. Rheumatoid arthritis (RA) was found to be associated with HLA-DR4 in Caucasian populations and HLA-DR1 in other ethnic groups, leading to the idea of the “shared epitope” hypothesis. These findings in turn led to a number of other key developments for our understanding of autoimmune disease.

One of these is the development of statistical methods that examine the concordance of disease in identical and nonidentical twins or between parents and children or siblings to calculate how much of the risk of developing an autoimmune disease such as RA is attributed to genes and how much to the environment, as well as understanding how much of the genetic element was caused by HLA region or other non-HLA region genes. In addition, these findings of HLA associations with disease or particular autoantibodies associated with disease have generated a dominant model of autoimmunity in which a particular antigen or set of antigens presented by particular HLA molecule(s) play a critical role in triggering the condition.

There have been no large-scale monozygotic versus dizygotic twin studies, or large multiplex family studies in pSS. Based on case reports and small studies, the estimated concordance rate for SS is low and the sibling prevalence likewise, suggesting that the heritability of SS is low and environmental factors play a greater role. In practical terms, this would suggest that in the absence of a preexisting family history, the likelihood of the child of a patient with pSS developing pSS as well is likely to be only modestly greater than the general population. In contrast, however, a recent study by Kuo et al. researching the Taiwanese National Health Insurance Research Database suggests that half the phenotypic variance in pSS can be explained by familial factors.

PSS has been closely linked to the presence of particular genes of the human major histocompatibility complex (MHC) that encodes HLA proteins. A meta-analysis found that the HLA class II alleles DRB1∗03:01 , DQA1∗05:01 , and DQB1∗02:01 were associated with an increased risk, and DQA1∗02:01 , DQA1∗03:01 , and DQB1∗05:01 with a lower risk, although it should be noted that there was some heterogeneity in the classification criteria used for pSS. These links are principally between the HLA types and the presence of anti-Ro/SSA and anti-La/SSB autoantibodies rather than with the disease per se. Patients with high levels of both anti-Ro/SSA and anti-La/SSB antibodies have a very high (∼90%) likelihood of being HLA DR3 DQ2 positive (typically associated with the DRB1∗03-DQB1∗02-DQA1∗0501 extended haplotype), whereas pSS patients who have high levels of anti-Ro/SSA antibody only and are negative for anti-La/SSB antibodies have an increased frequency of DR2(15) and DQ6 (typically associated with the DRB1∗1501-DQA1∗0102-DQB1∗0602 extended haplotype). Conversely, secondary SS in patients with RA is associated with HLA-DR4 , emphasizing that the clinical and histopathological similarities between pSS and secondary SS do not extend into identical genetic backgrounds.

Another issue to consider is that genetic susceptibility can differ between disease subsets within a single disease, or overlap between different autoimmune diseases with common features, both clinical and serological. RA is an example of a disease where serology (anti-cyclic citrullinated peptide [CCP] antibodies), the environment (smoking status), and genetics ( HLA and PTPN22 ) interact in a complex manner ( Fig. 8.2 ).

Celiac disease, pSS, and autoimmune thyroid disease are recognized disease associations. Celiac disease is an HLA-associated condition with DQA1∗05 and DQB1∗02 either on the same or separate chromosomes found in 90% of patients and these also form part of the extended HLA-DR3 DQ2 haplotype associated with pSS, perhaps at least in part explaining this association. It is not clear whether having two copies of a susceptible HLA gene is associated with particular clinical features such as lymphoma.

Another observation from HLA studies is that the susceptibility alleles vary in different population groups. Women with anti-Ro/SSA and anti-La/SSB antibodies in Japan (irrespective of current disease status with regard to SLE or pSS) are more likely to have the DRB1∗08032 allele or the associated haplotype rather than the DRB1∗03 in Caucasian populations. DRB1∗0803 has also been associated with pSS in Chinese populations.

Despite several decades of research into HLA associations with disease the precise reasons for these associations remains largely speculative. There was hope, for example, that if the precise details of the HLA association were known, including the key HLA peptide binding pockets that this could be used to identify the disease-specific antigens or otherwise explain the nature of the HLA association. Alternatively, it was proposed that candidate antigens could be identified by their ability to stimulate patient T cells when presented on appropriately HLA-restricted antigen-presenting cells. Although these studies have yielded results of interest, to date they have not identified specific antigens associated with pSS.

5

Candidate Gene Analysis

Candidate gene analysis is a powerful tool to examine genetic contribution to disease. Using PCR technology, the incidence of a particular gene variant in a patient population can be compared to the incidence among population controls. It does require a hypothesis to decide which genes to examine.

Lessard et al. have summarized in detail the results of a large number of candidate gene studies in pSS. Not surprisingly, the themes underpinning the search for genes of interest outside of classical HLA molecules derive from our understanding of immunopathology, namely genes coding for molecules involved in the innate immune system and IFN-related genes (eg, IRF5 and STAT4 ); cytokine genes (eg, for interleukin (IL)-10, IL-1 family, IL-6); MHC-related genes such as TAP , lymphotoxin β and TNF, and the Ro/La autoantigens themselves; and B-cell related genes (eg, BLK , EBF1 , and BAFF ).

Gorr et al. have performed a literature search in PubMed of genes and proteins that have been associated with pSS (excluding gene expression and proteomic studies) as a useful and updated resource for researchers.