Positional dissociation between the genetic mutation responsible for pseudohypoparathyroidism type Ib and the associated methylation defect at exon A/B: evidence for a long-range regulatory element within the imprinted GNAS1 locus

Abstract

Pseudohypoparathyroidism type Ib (PHP-Ib) is a paternally imprinted disorder which maps to a region on chromosome 20q13.3 that comprises GNAS1 at its telomeric boundary. Exon A/B of this gene was recently shown to display a loss of methylation in several PHP-Ib patients. In nine unrelated PHP-Ib kindreds, in whom haplotype analysis and mode of inheritance provided no evidence against linkage to this chromosomal region, we confirmed lack of exon A/B methylation for affected individuals, while unaffected carriers showed no epigenetic abnormality at this locus. However, affected individuals in one kindred (Y2) displayed additional methylation defects involving exons NESP55, AS and XL, and unaffected carriers in this family showed an abnormal methylation at exon NESP55, but not at other exons. Taken together, current evidence thus suggests that distinct mutations within or close to GNAS1 can lead to PHP-Ib and the associated epigenetic changes. To further delineate the telomeric boundary of the PHP-Ib locus, the previously reported kindred F, in which patient F-V/51 is recombinant within GNAS1, was investigated with several new markers and direct nucleotide sequence analysis. These studies revealed that F-V/51 remains recombinant at a single nucleotide polymorphism (SNP) located 1.2 kb upstream of XL. No heterozygous mutation was identified between exon XL and an SNP approximately 8 kb upstream of NESP55, where this affected individual becomes linked, suggesting that the genetic defect responsible for parathyroid hormone resistance in kindred F, and probably other PHP-Ib patients, is located >or=56 kb centromeric of the abnormally methylated exon A/B. A region upstream of the known coding exons of GNAS1 is therefore predicted to exert, presumably through imprinting of exon A/B, long-range effects on G(s)alpha expression.