1.3
TITLE: TNF MICROSATELLITE POLYMORPHISMS IN ANCESTRAL HAPLOTYPES IN RELATION TO OTHER MHC POLYMORPHISMS

M. Tevfik Dorak,¹ Charlotte Hilliard,¹ Malak Kotb.<sup>1

1</sup>Dept of Surgery, University of Tennessee Health Sciences Center, Memphis, Tennessee

TNF microsatellite polymorphisms have been used in numerous disease association studies but have not been evaluated systematically using the latest typing systems with fluorescent–labeled primers. We have typed a panel of 60 IHW cell lines representing 25 ancestral haplotypes for all five microsatellite loci on ABI 3100 DNA Analyzer using HEX or FAM–labeled primers. The same panel has already been typed at HSP70 and NOTCH4 by fluorescent–labeled primers. In addition to the cell line panel, we also typed a sample of Caucasian population (n = 206) at the most polymorphic of the TNF microsatellites, TNFa, and several other class III loci (Factor B [BF] TaqI RFLP, LTA NcoI RFLP, and HSP70 pentanucleotide polymorphism). We verified most of the typing results presented in the classic paper of Udalova et al (1993) with two exceptions: MT14B and EMJ are not heterozygous but homozygous for a11b4 and a4b5, respectively. Two pairs of homozygous cell lines (LWAGS/PMG075 and EMJ/SLE005) were identical at HLA–B, –Cw and –DRB1 but not at class III polymorphisms. The cell lines representing the ancestral haplotypes 7.1, 54.1 and 60.1 all had the TNF a11b4c1d3e3 haplotype, and also carried the same LTA (allele 2), NOTCH4 (R10) and HSP70 (183bp) alleles. The population data were used to estimate linkage disequilibrium (LD) and further examine the class III haplotypes. The most frequent TNFa alleles were: a2 (30.3%), a11 (16.2%) and a6 (13.4%). All six TNF a11 homozygote samples were also homozygote for BF*S, LTA*2 and HSP70*183 but only one was homozygote at DRB1 (for *15). This sample showed positive LD (P < 0.005) between TNFa7 and BF*FA, TNFa7 and DRB1*07; TNFa6 and LTA*1; TNFa2 and DRB1*03, TNFa11 and DRB1*15 (&Dgr;values varying between 0.0212 and 0.0411). These results suggested conservation of certain class III haplotypes among diverse HLA haplotypes and the possibility of heterogeneity in the class III region despite being identical at the classical HLA loci. These findings reemphasized that the unit of inheritance is the whole haplotype for MHC, and haplotypes rather than alleles of individual loci should be examined in disease association and transplantation studies with particular attention to linkage disequilibrium.