Is there a Common Genetic Basis for Autoimmune Diseases?

Autoimmune diseases (ADs) represent a diverse collection of diseases in terms of their demographic profile and primary clinical manifestations. The commonality between them however, is the damage to tissues and organs that arises from the response to self-antigens. The presence of shared pathophysiological mechanisms within ADs has stimulated searches for common genetic roots to these diseases. Two approaches have been undertaken to sustain the “common genetic origin” theory of ADs. Firstly, a clinical genetic analysis showed that autoimmunity aggregates within families of probands diagnosed with primary Sjögren's (pSS) syndrome or type 1 diabetes mellitus (T1D). A literature review supported the establishment of a familiar cluster of ADs depending upon the proband's disease phenotype. Secondly, in a same and well-defined population, a large genetic association study indicated that a number of polymorphic genes (i.e. HLA-DRB1, TNF and PTPN22) influence the susceptibility for acquiring different ADs. Likewise, association and linkage studies in different populations have revealed that several susceptibility loci overlap in ADs, and clinical studies have shown that frequent clustering of several ADs occurs. Thus, the genetic factors for ADs consist of two types: those which are common to many ADs (acting in epistatic pleitropy) and those that are specific to a given disorder. Their identification and functional characterization will allow us to predict their effect as well as to indicate potential new therapeutic interventions. Both autoimmunity family history and the co-occurrence of ADs in affected probands should be considered when performing genetic association and linkage studies.


Introduction
Autoimmune diseases (ADs) are chronic conditions initiated by a loss of immunological tolerance to selfantigens. The chronic nature of such diseases has a significant impact in terms of the utilization of medical care, direct and indirect economic costs and quality of life. The estimated incidence of ADs is about 90 cases per 100,000 person-year and their prevalence is about 3% of the population (Cooper and Stroehla 2003). Almost all ADs disproportionately affect middle agewomen and are among the leading causes of death for this group of patients. The older the patient, the lower the male:female ratio becomes (Cooper and Stroehla 2003).
Although the etiology of ADs is unknown, these diseases are known to feature genetic and environmental factors in their development Vyse and Todd 1996). The impact of genetic predisposition on susceptibility to ADs was first identified by the analysis of disease concordance rates in monozygotic twins. The monozygotic disease concordance rate ranges from about 15% for rheumatoid arthritis (RA) (Silman et al. 1993) to a fairly robust 57% for systemic lupus erythematosus (SLE) (Winchester 1992) climate. Comparisons of these high concordance rates with disease incidence in the general population predict that genetic predisposition is the dominant factor in AD susceptibility. The dramatic decrease in the concordance rate of siblings compared with that of monozygotic twins supports the presence of multiple genes contributing to the genetic predisposition. Finally, the estimation of familial aggregation, or recurrent risk ratio (l R ) for ADs (which is the ratio of the risk of disease prevalence among the specific-relative pairs ( R ) of affected probands to the prevalence of the disease in the general population), supports a fundamental role for genetic predisposition in disease susceptibility . However, population studies have established that each population holds a mutational pool, in which most mutations (i.e. polymorphisms) have mild or even undetectable effects individually, but in combination with other alleles may promote or protect from, autoimmune phenomena. Such interplay between genetic variants will generate a change in the measurable risk of developing an autoimmune phenotype. This characteristic is the main reason why ADs are not inherited in a simple, classical Mendelian way, but instead have a complex or an as yet unknown mode of inheritance (Vyse and Todd 1996).
As a group, ADs represent a diverse collection of diseases in terms of their demographic profile and primary clinical manifestations. The commonality between them, however, is the damage to tissues and organs that arises from the response to self-antigens (loss of tolerance). The vast majority of research generally focuses strongly on individual clinical diseases even though these autoimmune phenotypes could represent pleiotropic outcomes of specific, common disease genes that underlie similar immunogenetic mechanisms . Therefore, our group is investigating in a well-defined population whether or not clinically different ADs share the same susceptibility risk factors (i.e. genetic variants). This hypothesis of a common genetic origin for diverse ADs has been sustained by two important findings. The first consisted of a clinical analysis in which aggregation of ADs was observed in families of patients with primary Sjö gren's syndrome (pSS) and type 1 diabetes mellitus (T1D). The second was a large genetic association study carried out in three different ADs namely RA, SLE and pSS. Herein, we summarize the main results of these and other similar studies that reveal new insights concerning the inheritance pattern and the genetic risk factors for the development of ADs.

Genetic history of the sample population
Historical and genetic evidence suggests that the population of Antioquia is useful for the genetic mapping of complex traits. The state of Antioquia (capital, Medellín) is geographically located in the north-western part of Colombia between the Central and Western branches of the Andean Mountains and is inhabited by the "Paisa" community, a description of which has already been published (Bravo et al. 1996;Jimenez et al. 1996). Anthropological and historical studies describe this population as the most clearly defined in Colombia. Its ethno-historical origin stems most probably from the Spaniards, Jews (Christianized Sephardim or Marranos), and Basques. The admixture between Paisa and African or Amerindian populations has been historically documented as low (Parsons 1968). Several lines of genetic evidence suggest that the Paisa community exhibits the features of a genetically isolated community. Firstly, the identity coefficient method has estimated the ancestral ethnic components as 85% Caucasian and 15% Amerindian (Bravo et al. 1996). The African contribution to the Paisa community was estimated as being not significantly different from 0 (Jimenez et al. 1996). Secondly, strong admixture distortions in the gender vectors of racial blending in this community were found, with more than 96% of the chromosomes being of Caucasian origin and most of the mitochondrial component of Amerindian origin (Carvajal-Carmona et al. 2000).

Familial aggregation of autoimmunity
Primary SS is a late-onset AD characterized by a lymphocytic and plasma cell infiltration of the exocrine glands, as well as by the production of autoantibodies leading to dryness of mucosa, mainly oral and lachrymal (Anaya et al. 2001). Contrary to pSS, T1D appears early in life as a consequence of autoimmune damage to the insulin-producing pancreatic b cells (Expert Committee on the Diagnosis and Classification of Diabetes Mellitus 1997). There is strong evidence indicating familial clustering of various ADs in patients with several ADs including pSS and T1D (Anaya et al. 2006a, c) (Table I).
Bloch and Bunin first suggested a shared immunopathological mechanism for SS, SLE, systemic sclerosis (SSc) and autoimmune thyroid diseases (AITD), as well as a possible familial aggregation of these diseases in pSS patients (Bloch and Bunim 1963). They described the clinical and immunological characteristics of 57 cases of SS and some immunological abnormalities among their relatives. However, most of those patients (70%) had secondary SS (Bloch and Bunim 1963). In fact, SS may coexist with other ADs such as RA (Moutsopoulos et al. 1979), SLE (Steinberg and Talal 1971), AITD (Foster et al. 1993), and SSc (Alarcon-Segovia et al. 1974). Later, Reveille et al (1984) examined the presence of ADs in family members of 51 patients with pSS. They investigated the relationships of HLA genes and heavy chain immunoglobulin (Gm) haplotypes to disease and autoantibody expression in six large kindreds, each having one or more members with pSS. Segregation analyses suggested a Mendelian dominant genetic effect common among the many ADs and serologic reactions that were not linked to HLA or Gm (Reveille et al. 1984). Recently, we performed a familial aggregation study of ADs in 101 families of pSS patients and 124 control families (Anaya et al. 2006c). In the families of pSS patients, 38 (37.6%) had at least one first degree relative (FDR) with an AD, compared with 27 (21.8%) of control families (OR 2.2, 95% CI: 1.2 -3.9, p ¼ 0.01). Similarly to Reveille et al, we found that AITD, RA, and SLE were the most common ADs among relatives of the pSS patients (Reveille et al. 1984;Anaya et al. 2006c).
Multiple sclerosis and SSc were also observed, although with lower frequency. Moreover, the risk of familial autoimmunity increased with the number of affected FDR (Anaya et al. 2006c). Familial aggregation of AITD, SLE, RA, T1D, vitiligo (VIT), and all ADs taken together as a single trait, was observed (Anaya et al. 2006c).
Familial autoimmunity was also investigated in 98 T1D families and 113 families of healthy control individuals all from the same Colombian population (Anaya et al. 2006a). Our results showed that in the families of T1D patients, 25 (25.5%) presented at least one FDR having an AD compared with 9 (8%) in control families (OR: 3.96, 95% CI: 1.75 -9; p ¼ 0.0006), supporting the findings of Tait et al (2004) who observed a similar increase of ADs in family members of British patients with T1D. In our cohort, familial aggregation was observed for AITD, T1D and all ADs taken together as a single trait (Anaya et al. 2006a). Of particular interest, AITD (mainly hypothyroidism) is the most common AD encountered among FDR of pSS and T1D patients; this has also been reported in familial studies in multiple sclerosis (Broadley et al. 2000), VIT (Alkhateeb et al. 2003), juvenile RA (JRA) (Prahalad et al. 2002), and SLE (Priori et al. 2003). The next most common diagnosed ADs in FDR of pSS patients are RA and SLE, while in T1D SLE fills this position (Anaya et al. 2006a, c). All these diseases, as we will discuss, share similar susceptibility gene polymorphisms including HLA and non-HLA variants (Becker et al. 1998;Wandstrat and Wakeland 2001;Correa et al. 2005;Gomez et al. 2005), which may account for aggregation. In fact, shared genetic factors are the most likely cause of familial aggregation; however, it is important to keep in mind that shared environmental factors can also contribute to such aggregation. For a specified relative type, a l R . 1.0 suggests familial aggregation of the disease, but does not identify whether genetic or environmental factors are aggregating (Laird and Cuenco 2003). Thus, familial aggregation studies should be control studies in which population matched proband families lacking the trait of interest should be included.

Maternal transmission of the autoimmunity trait
A predominant inheritance of the autoimmunity trait from mothers has been observed in patients with pSS (Reveille et al. 1984;Anaya et al. 2006c), SLE (Priori et al. 2003) and T1D (Tait et al. 2004), indicating a plausible and preferential transmission of susceptibility alleles from mothers to offspring. Maternal transmission of autoimmunity could be influenced by the high preponderance of ADs in females compared to the general population. However, this higher than expected frequency of the autoimmunity trait maternal transmission would warrant further studies of mtDNA, genomic imprinting, maternal-offspring compatibility, and indirect genetic effects.
Collectively, results indicate that in pSS, T1D as well as in other ADs, pathologic autoimmunity aggregates as a trait (Table I), and emphasize the importance of the autoimmunity family history as a substantial risk factor for the development of any AD. This summary also points out the magnitude of the autoimmunity family history when carrying out both linkage and association genetic studies (i.e. transmission disequilibrium test).

Population genetic evidence for the common origin of ADs
The autoimmune phenotype and genotype might vary among populations. Diverse populations would behave genetically different when referring to its genetic variants, depending on the population natural and epidemiological history (Mori et al. 2005). In addition, the effects of genotype on phenotype in any given population may depend upon environment and the length of exposure to an undefined etiological insult. Therefore, there is a need to explore genetic associations in diverse populations. Confirming the results within different populations favors a more complete and homogeneous comprehension of the pathogenic mechanisms of ADs.
The "common variants/multiple disease (CV/MD)" hypothesis implies that "complex phenotypes are not unique entities but are mosaics of common disease specific alleles and non-disease specific modifying alleles in the population, influenced by a vast array of environmental factors" (Becker 2004). We surmise that if the CV/MD hypothesis was to be validated in an autoimmune setting, it would provide a convincing framework for the possibility that a common set of alleles might contribute to dissimilar clinical phenotypes.

Conclusion and perspectives
Herein, we have uncovered two approaches supporting the hypothesis of a common genetic origin for diverse ADs. A familial core for ADs depending upon the proband phenotype is illustrated as a graphic contingency table in Figure 2, and the chromosomal regions overlapping among the most frequent ADs are depicted in Figure 3. The precise mechanisms by which possession of the genetic variants affect diverse ADs are not completely understood and are beyond the scope of this review. However, genetic variation plays an important role in the determination of individual changes in protein expression as discussed elsewhere (Price et al. 1999;Camargo et al. 2004;Knight 2005;Castiblanco and Anaya 2006;Serrano et al. 2006). Since variation in gene expression is heritable and can be mapped as a quantitative trait, haplotype structures and functional assays should be considered in such genetic studies.
The co-occurrence of ADs is a well-known phenomenon. Epidemiological studies have described an increased statistical susceptibility of people with one AD to develop other ADs (Sloka 2002). This Figure 2. Familial autoimmunity clusters. The ordinate represents the proband's disease, while the axis despites the correspondent aggregated AD. Circle size represents the weighted analyzed reports and does not necessarily indicate the aggregation's magnitude. This figure was generated from Table I   situation is well illustrated by the type 2 autoimmune polyglandular syndrome, originally described by Schmidt (1926), and by the "multiple autoimmune syndrome", described by Humbert and Dupond (1988). In these scenarios a single genotype is responsible for diverse phenotypes, making these cases the "magna prove" for a common origin of ADs. Recently, Namjou et al. (2005) after stratification of multiplex SLE pedigrees by the presence of AITD, observed that 5q14.3 -15 is a region linked to both diseases. Thus, besides autoimmunity family history, the co-occurrence of ADs should be also considered when performing association and linkage studies.
The common genetic background evidence for ADs has grown and allows us to infer that AD phenotypes have a clinical behaviour that can be independent from their genetic causes. The heterogeneity of ADs could be due to a collection of diverse disorders based on epidemiology pathology or diagnostic results but in fact the underlying immunogenetic mechanism might be similar. Identification of such common genetic causes will enhance our understanding of the common mechanisms of these complex, frequent and sometimes devastating diseases and will permit us to predict them as well as to discover new therapeutic interventions.