DNA sequence variation within and around human minisatellites.

Hypervariable minisatellites have found increasing use in a wide range of applications including linkage analysis, relationship testing and forensic medicine. Conventional length analysis via restriction digestion, agarose gel electrophoresis and Southern blot hybridization produces at best, less than 100 resolvable alleles and involves inherently error prone length estimations. Forensic analyses based on such a system have recently been heavily criticised and the population assumptions used to calculate match frequencies questioned. However, allele length is not the only criterion by which minisatellite variability may be assessed. All well characterised hypervariable human minisatellites also show variation in the dispersion patterns of minisatellite variant repeats. At the most well characterised human minisatellite MS32, MVR variation has been assayed by both restriction analyses and a more direct MVR-PCR approach. Both procedures have shown that allele length analysis seriously underestimates the true variability at minisatellite loci. Many individuals incorrectly typed as homozygous by length analysis have been demonstrated to be heterozygous by internal MVR mapping. MVR allelic analysis has been used to directly show the existence of many hundreds of different alleles, each one absolutely defined, from which we can infer the existence of many thousands of alleles. With a theoretical capacity to define many millions of alleles MVR analysis adds a new dimension to the study of variation in human DNA. The high observed allelic variability is directly reflected in the very high individual specificity obtained when MVR-PCR is applied to total genomic DNA to obtain a diploid code of the two superimposed alleles. MVR diploid code analysis produces a highly portable digital output which provides unambiguous match criteria. Diploid code analysis may be used to obtain highly discriminatory information from mixed DNA samples and, via the use of allele specific flanking primers in knockout MVR-PCR, from admixtures far lower than currently approachable. MVR analysis of allelic variation has revealed the existence of a terminal mutation hotspot which has been confirmed in studies of de novo mutation events in pedigrees. This work has revealed the complex nature of minisatellite mutation, with, for the first time, strong evidence for a role for unequal interallelic exchange and preliminary evidence for a size increase bias.