Search

It is known that celiac disease is strongly associated with an HLA class II component and that most patients carry the dimer DQA1*0501, DQB1*0201. We show in this study that the risk for a carrier of this heterodimer is independent from the number of possible heterodimers, from whether DQA1*0501 and DQB1*0201 are in cis or trans position and from the number of DQA1*0501 (one or two) but strongly depends on the number of DQB1*0201. In the Tunisian population we studied, the risk of developing celiac disease is estimated to be 6.8 times greater for those having a double dose of DQB1*0201 than for other dimer carriers. We replicated this result in published data of four other populations (Italy, Czekoslovakia, United Kingdom, Norway).

In order to investigate the HLA component involved in rheumatoid arthritis (RA), we tested genetic models by the marker association-segregation chi 2 (MASC) method, using the HLA genotypic distribution observed in a sample of 97 RA patients. First we tested models assuming the involvement of a susceptibility gene linked to the DR locus. We showed that the present data are compatible with a simple model assuming the effect of a recessive allele of a biallelic locus linked to the DR locus and without any assumption of synergistic effect. Then we considered models assuming the direct involvement of the DR allele products, and we tested the unifying-shared-epitope hypothesis, which has been proposed. Under this hypothesis the DR alleles are assumed to be directly involved in the susceptibility to the disease because of the presence of similar or identical amino acid sequences in position 70-74 of the third hypervariable region of the DRBI molecules, shared by the RA-associated DR alleles DR4Dw4, DR4Dw14, and DR1. This hypothesis was strongly rejected with the present data. In the case of the direct involvement of the DR alleles, hypotheses more complex than the unifying-shared-epitope hypothesis would have to be considered.

The marker association segregation chi-square (MASC) method was applied to a sample of 416 Caucasians affected with insulin-dependent diabetes mellitus (IDDM), for which information on the parental and sibship status was available, as well as HLA typing. We show that the model which best explains all the observations assumes a cis or trans complementation of two tightly linked genes within the HLA region, an additional maternal effect, and other familial factors. The HLA molecule corresponding to the complementation of Arg52(+) and Asp57(-) has been recently proposed as explaining susceptibility to IDDM. However, this hypothesis does not account for the overall observations made on the HLA marker in IDDM patients and their relatives. The MASC method may also be applied to evaluate the risk for relatives of an affected individual (the "index"). For example, the risk for a sib depends not only on the parental status and on the number of HLA haplotypes he shares with the index, but also on which haplotype the index himself inherited from his mother and father.

It has been shown that genetic factors within the HLA region are involved in the etiology of several diseases. For some of these, the existence of another genetic factor has been suggested, although not proven. A possible way to give evidence for another locus (G) is to show that the disease and an unlinked HLA-marker locus (M) do not segregate independently. The usual lod-score method, which assumes monogenic inheritance, is inappropriate for this test. We propose a correction of this method for performing a linkage analysis between the G and M loci, taking into account the role of HLA. A very simple way of using the HLA information is by modifying, for each individual of a pedigree, the penetrance values at the G locus according to the number of HLA haplotypes shared with the index case. These penetrance values are inferred from the observed IBD (identity-by-descent) distribution of HLA haplotypes in a sample of affected sib-pairs. The advantage of using this empirical distribution is that it is not based on any assumptions concerning the mode of inheritance at the HLA-linked locus. This correction method was established using a two-locus model with restrictive assumptions. Its value is discussed for various sets of parameters in more general and realistic two-locus models using simulations.

One of the current issues in genetic epidemiology is detecting susceptibility genes on the genome. It is common now to undertake systematic screening of the genome using approaches based on a measure of the haplotype sharing in sib pairs. Here, we compare the efficiency of two statistics, the maximum likelihood score (MLS) and the nonparametric linkage score (NPLa) on the simulated data provided for GAW11. A question often raised is whether it is better to perform a single-step or a two-step strategy. For the simulated model, and whatever the strategy used, we show here that the answer is not unequivocal. In both cases, the power to detect susceptibility genes in a single replicate with MLS or NPL is extremely low. With two replicates, only one of the four simulated loci could be detected with reasonable power. When gametic disequilibrium is suspected, methods testing for both linkage and association might be more powerful.

Our recent segregation analysis, carried out on 27 large pedigrees from a Caribbean island (Desirade), has shown the presence of recessive major gene(s) controlling susceptibility to leprosy per se and nonlepromatous leprosy, respectively. Linkage analysis was performed between each of these two detected genes and each of five markers typed in the Desirade population: HLA, ABO, Rhesus, Gm and Km. No positive significant lod score was observed. However, for leprosy per se close linkage was excluded with Rhesus and Gm (and also with ABO and HLA, considering a lower value for the frequency of the gene controlling susceptibility to leprosy per se). The highest lod score, although not significant, was obtained between the gene for nonlepromatous leprosy and ABO. Our overall results, joined with previous studies and experimental data, suggest that the gene controlling susceptibility to leprosy per se and that controlling susceptibility to nonlepromatous leprosy might be different, acting at successive stages of the immune response to infection with Mycobacterium leprae.

The power to detect linkage by the LOD-score method is investigated here for diseases that depend on the effects of two genes. The classical strategy is, first, to detect a major-gene (MG) effect by segregation analysis and, second, to seek for linkage with genetic markers by the LOD-score method using the MG parameters. We already showed that segregation analysis can lead to evidence for a MG effect for many two-locus models, with the estimates of the MG parameters being very different from those of the two genes involved in the disease. We show here that use of these MG parameter estimates in the LOD-score analysis may lead to a failure to detect linkage for some two-locus models. For these models, use of the sib-pair method gives a non-negligible increase of power to detect linkage. The linkage-homogeneity test among subsamples differing for the familial disease distribution provides evidence of parameter misspecification, when the MG parameters are used. Moreover, for most of the models, use of the MG parameters in LOD-score analysis leads to a large bias in estimation of the recombination fraction and sometimes also to a rejection of linkage for the true recombination fraction. A final important point is that a strong evidence of an MG effect, obtained by segregation analysis, does not necessarily imply that linkage will be detected for at least one of the two genes, even with the true parameters and with a close informative marker.

For elucidating the genetic component of multifactorial diseases, it is important to investigate the effect of several factors and the possible interaction between them. In particular, for many diseases it is interesting to study the interactive effect of two genes. In this context, the marker-association-segregation chi 2 method (MASC), initially proposed to detect the involvement of a candidate gene in multifactorial diseases, is developed here to investigate the involvement of two candidate genes and to model the joint effect of these two genes. In particular, it is possible to precisely determine whether the joint effect of both genes is multiplicative. This extension simultaneously uses information on two markers, one for each candidate gene, at both the population and the familial segregation level. We show here that there can be an important gai of power to detect the effect of a second gene in a disease when information is used simultaneously on two markers instead of studying each marker separately. This extension of MASC is then applied on a sample of insulin-dependent diabetes (IDD) families typed for the markers of two candidate regions: HLA and that of the insulin gene (INS). This analysis allows us to confirm the involvement of INS in IDD, and the best-fitting model is a multiplicative (noninteractive) effect of HLA and INS, with a biallelic locus for INS and a complementation model for HLA.

Corrigendum to Season of birth and not vitamin D receptor promoter polymorphisms is a risk factor for multiple sclerosis by DA Fernandes de Abreu, MC Babron I, C Rebeix, J Fontenille, D Yaouanq, B Brassat, F Fontaine, F Clerget-Darpoux, F Jehan and Feron. Mult Scler 2009; 15: 1146-1152 [DOI: 10.1177/1352458509106780].