The Search For Novel Nonmhc Susceptibility Genes

Recent genomewide searches for T1DM susceptibility provided preliminary evidence for the existence of at least 18 loci associated with T1DM (164). Because of the large number of markers tested, many of these putative regions suggestive of T1DM susceptibility linkage may have occurred by chance. Saturating the putative locus with many more informative markers is a must to demonstrate or rule out the existence of significant linkage. Further complicating the issue of suggestive linkage from genomewide scans is the broad range of selection criteria used for inclusion of families in the various datasets. Variations in the age of onset of T1DM can make the detection of linkage difficult. Confounding matters even more is the interaction of the various disease susceptibility loci that is the hallmark of polygenic diseases like T1DM, in that this interaction (additive, multiplicative, or epistatic) adds an additional level of difficulty in determining the significance of a logarithm of odds (LOD) score. Once suggestive linkage to a region has been determined, other powerful association-based tests that take advantage of linkage disequilibrium between a marker and the actual susceptibility alleles can be used to formally implicate a locus and a particular allele with susceptibility (70).

The initial genome scan by Davies et al. (164) suggested the existence of 18 susceptibility loci. To date, suggestive linkage to 16 of those 18 susceptibility loci (164) (termed IDDM, followed by a number) does not necessarily imply the importance of that locus to susceptibility. The demonstration of significant linkage to some of these sites has been replicated in independent datasets, in some cases implicating a candidate gene. Linkage to other loci originally detected in the scan by Davies et al. awaits replication. It is important to note that some of the loci that showed suggestive, albeit weak, linkage in the scan by Davies et al. (164) could not be confirmed in other populations. This indicates the importance of replication in other datasets as well as the selection of a dense set of markers to cover as much of the genome as possible. Recent studies have revealed the genetic interaction of certain of these loci and their mode of interaction. For example, IDDM1 (which is the class II HLA locus) and IDDM2 (which is the VNTR upstream of the INS promoter) interact epistatically (i.e., the genes are functionally related, either in a biochemical or a physiological pathway), whereas IDDM1 and IDDM4 may act independently (165). In addition, IDDM4 and IDDM5 appear to have a greater frequency of shared alleles, identical-by-descent subgroups of datasets, which have HLA-DR3 in common. The linkage of IDDM7 is stronger in subgroups lacking HLA-DR3 and, finally, IDDM7 appears to show stronger evidence of linkage in subgroups homozygous for class I VNTR alleles at IDDM2 (70). All of these results will most likely depend on the nature of the population under study as there are already-established differences among populations in T1DM susceptibility at particular loci. The relative risk of haplo-types at IDDM2 is dependent on the presence of HLA-DR4 in a French population (152) but not in American, Belgian, Finnish, or British datasets (166).

There is continuous debate, however, as to what criteria should be used when attempting to map non-HLA loci. Historically, in a monogenic disease, the generally accepted standard for significant linkage is considered to be a LOD score greater than 3 (p < 10-4). Thomson (167,168) suggested a statistical level of p = 0.001 in considering presumable linkage in genomewide searches, followed by confirmation of the results in other populations to confirm linkage. Lander and Kruglyak (45), however, proposed more stringent criteria of a LOD score of 2.2, indicating suggestive linkage (p < 7 x 10-4) and a LOD score of 3.6 (p < 2 x 10-5) to achieve significant linkage. Even with a LOD score of 3.6, there is still a possibility of a false-positive rate of 5% in genomewide scans. Moreover, Lander and Kruglyak propose that further replication studies in different populations are necessary in order to verify significant linkage. Very few of the loci indicated in Table 2 will stand the criteria of Lander and Kruglyak for significant linkage. To circumvent this, some investigators propose that the HLA exerts a very powerful effect on overall susceptibility and, as a consequence, some loci will be missed although they may play some role in other biochemical or physiological pathways involved in susceptibility. Indeed, one means by which some loci were found has been to stratify based on the number of HLA alleles shared in sib-pair-based mapping approaches. This method was used to describe IDDM13 (169). Another approach stratifies based on the presence or absence of autoantibodies (170). To date, there has not been any rigorous statistical analysis on the validity of these stratifications, and although they may be justified, the recent failure to find linkage at all described loci except IDDM1, IDDM2, IDDM5, and IDDM8 in more than 500 sib-pairs using stringent criteria (Polychronakos, personal communication) may, in part, be explained by the application of such stratifications.

Fig. 9. Results of an initial genomewide (309 markers) multipoint NPL analysis using GENE-HUNTER (171). The maximum NPL value (0.002) corresponds to marker D10S1237 at chromosome 10q25. The evidence for linkage to this region was strengthened with the analysis of additional microsatellites and additional family members. (From ref. 46.)

Fig. 9. Results of an initial genomewide (309 markers) multipoint NPL analysis using GENE-HUNTER (171). The maximum NPL value (0.002) corresponds to marker D10S1237 at chromosome 10q25. The evidence for linkage to this region was strengthened with the analysis of additional microsatellites and additional family members. (From ref. 46.)

Finally, the last approach was to evaluate empirical power and efficiency of mapping complex disorders, such as autoimmune diabetes, studying large multiplex families from genetically and culturally homogeneous populations, such as a Bedouin family from Israel (46). The extended pedigree of this multiplex Bedouin family included 248 individuals, along with 19 affected individuals in 3 generations. Results from genome scans for linkage indicate a predominant peak by nonparametric linkage (NPL) analysis that is seen for the long arm of chromosome 10 (10q25, IDDM17), with the maximum NPL occurring at D10S1237 (p < 0.002) (see Fig. 9). A high-resolution map of the candidate region on 10q was constructed using the Centre d'Etude du Polymorphisme Humain (CEPH) genotyping database and the Massachusetts Institute of Technology (MIT) physical mapping database to identify polymorphic markers and yeast artificial chromosome (YAC) clones (46). With the higher density of markers, the evidence of linkage increased substantially (p = 0.00004) (46). Furthermore, preliminary evidence indicates that the high-risk haplotype of IDDM17 in the Bedouin Arab family may be present in as many as 5% of US families (Eisenbarth, personal communication).

In sum, genomewide searches are considered only an initial stage in discovering novel susceptibility genes in type 1 diabetes and in other polygenic diseases. First, linkage must be proven, which is not an easy task. Second, a large number of families from independent populations are required to confirm linkage and understanding the interactions of putative susceptibility genes.


Recently, it has been described that the position of provisional loci found in T1DM colocalize or overlap with loci found in different autoimmune/inflammatory diseases (48,172). This is consistent with the hypothesis that, like the MHC, some of these provisional loci may involve common susceptibility genes or biochemical pathways that are central to normal immune function. Concannon et al. (148) identified a novel locus for T1DM (MLS=3.31) (Table 2) at human chromosome 1q at marker D1S1617. This locus colocalizes with loci for systemic lupus erythematosus (SLE) (49) and ankylosing spondylitis (50). In human SLE, this locus is linked to high serum levels of antichromatin antibody, and in mouse SLE. This locus is linked to both anti-chromatin and anti-DNA antibody production (173,174). This colocalization of suggestive genetic linkage for three autoimmune diseases suggests that genes at this locus may be involved in a pathway that might affect the quantitative regulation of antibody levels and that this may ultimately contribute to the development of the disease phenotype. Therefore, particular as yet undiscovered genes or pathways may contribute to immune dysregulation, a phenomenon detected in many different autoimmune diseases, possibly prior to the onset of overt clinical symptoms (175).

Ten centimorgans is the approximate limit of resolution of a typical first-stage genomewide scan. For example, IDDM2 found at 1lpl5.5 (164) is located at the exact position as loci for SLEk (176), ankylosing spondylitis (50), asthma (177), and multiple sclerosis (178). All four disease loci have been defined by the same polymorphic marker, D11S922, at the 0.323-cM position on human chromosome 11 (179). A candidate gene at 1lpl5.5 is the insulin gene itself. VNTR polymorphisms in the 5' end of the insulin gene have been associated and linked to IDDM (180). One interpretation of this genetic linkage to insulin as a candidate gene is that there might be an involvement of an imprinted gene (157,181-183), which may be under the same transcriptional effects of the VNTR as is the insulin gene. Colocalization of multiple autoimmune diseases at this location suggests that whatever the exact gene found at IDDM2 (180), it may play a broader role in autoimmune development.

Although this general pattern of locus colocalization appears not to be found in human nonautoimmune disease (172), it is possible that a pattern of colocalization of autoimmune disease may be the result of common biological pathways shared among related autoimmune/inflammatory abnormalities in coexisting human autoimmune disorders (172).

Was this article helpful?

0 0
Natural Vertigo And Dizziness Relief

Natural Vertigo And Dizziness Relief

Are you sick of feeling like the whole world Is spinning out of control. Do You Feel Weak Helpless Nauseous? Are You Scared to Move More Than a Few Inches From The Safety of Your Bed! Then you really need to read this page. You see, I know exactly what you are going through right now, believe me, I understand because I have been there & experienced vertigo at it's worst!

Get My Free Ebook

Post a comment