Association of a tumor necrosis factor α polymorphism with MS susceptibility

Methods.

We studied the putative association of this variant in case-control analysis. Cases consisted of 121 patients, who constituted 73% of a prevalence cohort of patients with MS in Olmsted County, MN. Controls consisted of 243 patients who were rigorously matched for age, sex, and ethnicity, as described elsewhere.4 We used allele-specific PCR to genotype patients for this polymorphism. Two allele-specific reactions were developed using “mutant” and “wild-type” allele-specific primers. The specific primer for the mutant allele was downstream 5′GGC ATC CTG TCT GGA AA, and the specific primer for wild-type allele was downstream 5′ CCC ATC CTG TCT GGA AG. The nonspecific upstream primer was 5′ CGG ATC ATG CTT TCA GTG; allele-specific nucleotides are italicized. The reactions were optimized to obtain the appropriate product size and were subsequently optimized to achieve specificity using controls whose genotype was determined by sequencing. The annealing temperature was 54 °C for the reaction to determine the mutant allele and 58 °C to determine the wild-type allele. A second downstream primer was incorporated into the reactions to yield a nonspecific band to prove the integrity of the reaction when the specific allele that is being tested is absent.

Results.

Our results and those reported by Fernandez–Arquero et al.1 are summarized and contrasted in the table. Our results do not show a significant association with MS. In fact, different from the Spanish study, carriers of the A allele were more frequent in the controls than in cases.

Discussion.

Fernandez–Arquero et al.1 reported linkage dysequilibrium of the TNFA-376A allele with DRB1*0301, DQA1*0501, DQB1*0201 haplotype, but found that the TNFA-376A allele was associated with MS in patients positive and negative for this major histocompatibility complex haplotype. There was an apparent synergistic effect with DRB1*1501, which is associated with predisposition to MS worldwide. Too few individuals in our population carried this allele to permit an evaluation of linkage dysequilibrium or synergistic effects.

The lower allele frequency of the TNFA-376A allele in our population, in patients with MS and in controls, reduced the power to detect an association. In this North American sample composed primarily of people of German, Scandinavian, and British descent, there is no apparent excess of TNFA-376A allele polymorphism in patients with MS compared with controls rigorously matched for ethnicity. Assuming that our inability to replicate the results is not merely an issue of power, the association described by Fernandez–Arquero et al.1 is a population-specific association or a chance (false-positive) result.