Identification and functional characterization of the human EXT1 promoter region

Background

Mutations in Exostosin-1 (EXT1) or Exostosin-2 (EXT2) cause the autosomal dominant disorder multiple osteochondromas (MO). This disease is mainly characterized by the appearance of multiple cartilage-capped protuberances arising from children's metaphyses and is known to display clinical inter- and intrafamilial variations. EXT1 and EXT2 are both tumor suppressor genes encoding proteins that function as glycosyltransferases, catalyzing the biosynthesis of heparan sulfate. At present, however, very little is known about the regulation of these genes. Two of the most intriguing questions concerning the pathogenesis of MO are how disruption of a ubiquitously expressed gene causes this cartilage-specific disease and how the clinical intrafamilial variation can be explained. Since mutations in the EXT1 gene are responsible for ~ 65% of the MO families with known causal mutation, our aim was to isolate and characterize the EXT1 promoter region to elucidate the transcriptional regulation of this tumor suppressor gene.

Methods

In the present study, luciferase reporter gene assays were used to experimentally confirm the in silico predicted EXT1 core promoter region. Subsequently, we evaluated the effect of single nucleotide polymorphisms (SNP's) on EXT1 promoter activity and transcription factor binding using luciferase assays, electrophoretic mobility shift assays (EMSA), and enzyme-linked immunosorbent assays (ELISA). Finally, a genotype-phenotype study was performed with the aim to identify one or more genetic modifiers influencing the clinical expression of MO.

Results

Transient transfection of HEK293 cells with a series of luciferase reporter constructs mapped the EXT1 core promoter at approximately − 917 bp upstream of the EXT1 start codon, within a 123 bp region. This region is conserved in mammals and located within a CpG-island containing a CAAT- and a GT-box. A polymorphic G/C-SNP at − 1158 bp (rs34016643) was demonstrated to be located in a USF1 transcription factor binding site, which is lost with the presence of the C-allele resulting in a ~ 56% increase in EXT1 promoter activity. A genotype-phenotype study was suggestive for association of the C-allele with shorter stature, but also with a smaller number of osteochondromas.

Conclusions

We provide for the first time insight into the molecular regulation of EXT1. Although a larger patient population will be necessary for statistical significance, our data suggest the polymorphism rs34016643, in close proximity of the EXT1 promoter, to be a potential regulatory SNP, which could be a primary modifier that might explain part of the clinical variation observed in MO patients.