Mutation affecting the proximal promoter of Endoglin as the origin of hereditary hemorrhagic telangiectasia type 1

Abstract

Background: Hereditary hemorrhagic telangiectasia (HHT) is a vascular multi-organ system disorder. Its diagnostic criteria include epistaxis, telangiectases in mucocutaneous sites, arteriovenous malformations (AVMs), and familial inheritance. HHT is transmitted as an autosomal dominant condition, caused in 85% of cases by mutations in either Endoglin (ENG) or Activin receptor-like kinase (ACVRL1/ACVRL1/ALK1) genes. Pathogenic mutations have been described in exons, splice junctions and, in a few cases with ENG mutations, in the proximal promoter, which creates a new ATG start site. However, no mutations affecting transcription regulation have been described to date in HHT, and this type of mutation is rarely identified in the literature on rare diseases.

Methods: Sequencing data from a family with HHT lead to single nucleotide change, c.-58G > A. The functionality and pathogenicity of this change was analyzed by in vitro mutagenesis, quantitative PCR and Gel shift assay. Student t test was used for statistical significance.

Results: A single nucleotide change, c.-58G > A, in the proximal ENG promoter co-segregated with HHT clinical features in an HHT family. This mutation was present in the proband and in 2 other symptomatic members, whereas 2 asymptomatic relatives did not harbor the mutation. Analysis of RNA from activated monocytes from the probands and the healthy brother revealed reduced ENG mRNA expression in the HHT patient (p = 0.005). Site-directed mutagenesis of the ENG promoter resulted in a three-fold decrease in luciferase activity of the mutant c.-58A allele compared to wild type (p = 0.005). Finally, gel shift assay identified a DNA-protein specific complex.

Conclusions: The novel ENG c.-58G > A substitution in the ENG promoter co-segregates with HHT symptoms in a family and appears to affect the transcriptional regulation of the gene, resulting in reduced ENG expression. ENG c.-58G > A may therefore be a pathogenic HHT mutation leading to haploinsufficiency of Endoglin and HHT symptoms. To the best of our knowledge, this is the first report of a pathogenic mutation in HHT involving the binding site for a transcription factor in the promoter of ENG.

Keywords: Endoglin promoter; Hereditary hemorrhagic telangiectasia (HHT); Rare disease; Transcription regulation.

Fig. 1

a Schematic representation of the family tree with the distribution of HHT and normal relatives. b Clinical diagnostic criteria of the affected relatives. c Mutated sequence corresponding to this particular HHT family, showing the nucleotide substitution c.-58 G > A at the proximal promoter of Endoglin. Arrows point to the two unaffected individuals, who were also sequenced

Fig. 2

Histograms representing the relative luciferase activity of wild type and mutated Endoglin promoter reporters after transfection into endothelial HMEC-1 cells. Site-directed mutagenesis was performed on the wild type promoter/reporter plasmid (350/+350 CD105 pXP2) of Endoglin. The effect of TGF-β treatment is also shown

Fig. 3

Results of RT quantitative PCR showing the RNA transcription levels of Endoglin and ACVRL1/ALK1 genes of activated monocytes from a HHT patient compared with his control brother. The results are compared to the endogenous control 18S. The qPCR was repeated three times, and the results shown are representative, with triplicates for each sample

Fig. 4

Schematic representation of the proximal promoter region encompassing the −58 (G > A) site and the transcription factors binding to it, according to the MatInspector program (Genomatix)

Fig. 5

Gel shift assay of nuclear extract from endothelial cells showing a retarded band of protein-DNA, within the proximal promoter of Endoglin. Nuclear extract from endothelial cells (HMEC-1 cell line) was incubated with the double stranded biotinylated oligonucleotide, and in the presence of unspecific competitor (poli dI-dC), a retarded band could be detected (lane 2). However, the sample was specifically treated with 100× excess of the unlabeled double-stranded nucleotide (cold probe) (lane 3); when the mutated oligonucleotide was added, the binding efficiency of the probe was not diminished (lane 4). The experiment was repeated 4 times. This image is representative of the obtained results