Identification and mapping of nuclear matrix-attachment regions in a one megabase locus of human chromosome 19q13.12: long-range correlation of S/MARs and gene positions.

The first draft human genome sequence now available allowed the identification of an enormous number of gene coding areas of the genomic DNA. However, a great number of regulatory elements such as enhancers, promoters, transcription terminators, or replication origins can not be identified unequivocally by their nucleotide sequences in complex eukaryotic genomes. One important subclass of these type of sequences is scaffold/matrix attachment regions (S/MARs) that were hypothesized to anchor chromatin loops or domains to the nuclear matrix and/or chromosome scaffold. We developed an experimental selection procedure to identify S/MARs within a completely sequenced one megabase (1 Mb) long gene-rich D19S208-COX7A1 locus of human chromosome 19. A library of S/MAR elements from the locus was prepared and shown to contain -20 independent S/MARs. Sixteen of them were isolated, sequenced, and assigned to certain positions within the locus. A majority of the S/MARs identified (11 out of 16) lie in intergenic regions, suggesting their structural role, i.e., delimitation of chromatin domains. These 11 S/MARs subdivide the locus into 10 domains ranging from 6 to 272 kb with an average domain size of 88 kb. The remaining five S/MARs were found within intronic sequences of APLP1, HSPOX1, MAG, and NPHS1 genes, and can be tentatively characterized as regulatory S/MARs. The correspondence of the chromatin domains defined by the S/MARs to functional characteristics of the genes therein is discussed. The approach described can be a prototype of a similar search of long sequenced genomic stretches and/or whole chromosomes for various regulatory elements. Copyright 2001 Wiley-Liss, Inc.