Evolutionary relationships between HLA-B alleles as indicated by an analysis of intron sequences

The HLA-B locus is the most polymorphic of the class I genes encoded within the human major histocompatibility complex. This poly morphism is mainly l ocated in exons 2 and 3, which code for the molecule's alpha 1 and alpha 2 domains and includes thr antigenic peptide binding site. However informatio n about adjacent non-coding regions (introns 1 and 2) has not been extensiv rly reported but could be very important in establishing an understanding o f the evolutionary mechanisms involved in the polymorphism generation oi KL A-B and the Mhc loci. In the present work, inh-uns 1 and 2 of 14 HLA-B alle les are studied and their significance is discussed; 10 have been sequenced in our own laboratory and the other 3 have been previously reported by oth ers. Different serological families share the complete intron 1 sequence; a t this region, 12 out oi 14 HLA-B alleles could be included in four groups with the same intron 1 sequence: a) B*0702, B*4201, B*4801; b) B*27052, B*4 002, B*4011; c) B*40012, B*4101, including B*4501, B*5001 (these latter two alleles have specific characteristics in both introns 1 and 2, which may r eflect a common evolutionary pathway); and d) B*44031, B*44032. The other a lleles, B*1402, and B*1801, do not have identical intron 1 sequences compar ed to any of the described groups, but share many similarities with them. T he B*1801 evolutionary pathway seems to be very specific since it branches separately from other alleles both in intron 1 and intron 2 dendrograms. On the other hand, HLA-B allelic group distribution and similarities accordin g to intron 1 sequences were not confirmed when using intron 2, especially in the cases oi B*4002, B*4101 and B*4801. This would suggest that both poi nt mutations fixed by genetic drift and gene conversion events are involved in HLA-B diversification. The latter events could be supported by the stro ng homology between intron 1 and, to a lesser extent, intron 2, and also th e CG content within them. Finally, the precise knowledge oi these nun-codin g regions could be important for developing DNA base typing strategies for the HLA-B alleles.