Identification of SATB2 as the cleft palate gene on 2q32-q33

Cytogenetic evidence, in the form of deletions and balanced translocations, points to the existence of a locus on 2q32-q33, for which haploinsufficiency results in isolated cleft palate (CPO). Here we show by high-resolution FISH mapping of two de novo CPO-associated translocations involving 2q32-q33 that one breakpoint interrupts the transcription unit of the gene encoding the DNA-binding protein SATB2 (formerly KIAA1034). The breakpoint in the other translocation is located 130 kb 3' to the SATB2 polyadenylation signal, within a conserved region of non-coding DNA. The SATB2 gene is transcribed in a telomeric to centromeric direction and lies in a gene-poor region of 2q32-q33; the nearest confirmed gene is 1.26 Mb centromeric to the SATB2 polyadenylation signal. SATB2-encoding transcripts are assembled from 11 exons that span 191 kb of genomic DNA. They encode a protein of 733 amino acids that has two CUT domains and a homeodomain and shows a remarkable degree of evolutionary conservation, with only three amino acid substitutions between mouse and human. This protein belongs to the same family as SATB1, a nuclear matrix-attachment region binding protein implicated in transcriptional control and control of chromatin remodelling. There are also sequence similarities to the Drosophila protein DVE. Whole mount in situ hybridization to mouse embryos shows site- and stage-specific expression of SATB2 in the developing palate. Despite the strong evidence supporting an important role for SATB2 in palate development, mutation analysis of 70 unrelated patients with CPO did not reveal any coding region variants.