Results

ANALYSIS OF THE BMPR2GENE

Sequencing of genomic DNA of the panel of sporadic PPH subjects showed a total of 11 novel heterozygous mutations of theBMPR2 gene in 13 of the 50 patients studied. The clinical features of these 13 patients are described in table 1. For patients 1 and 2, the nucleotide sequence showed substitutions of guanine for adenine in exons 2 and 3 respectively (table 2). These change the sequences of codon 60 from TGC to TAC (patient 1) and codon 117 from TGT to TAT (patient 2), both changing a highly conserved encoded amino acid from cysteine to tyrosine (table 2). As these mutations do not result in either gain or loss of a restriction site, genomic sequencing of parental samples was used to show the presence of the mutation in the fathers of both patients.

Genomic sequencing of patient 8 identified a substitution of thymine for cytosine in exon 11, changing the sequence of codon 483 from TGT to CGT and the encoded amino acid from cysteine to arginine (table 2). No additional family members were available for study. In patients 6 and 10, deletion of an adenine in exon 9 and a guanine in exon 12 both lead to a change in the coding reading frame and predict premature truncation of the 1038 amino acid protein at codon positions 423 and 803, respectively. Analysis of genomic DNA samples from unaffected parents, either by direct sequencing (patient 6) or restriction enzyme analysis with Fnu4HI (patient 10) of the appropriate exonic PCR fragment, showed the absence of the mutation, confirming these patients had spontaneous mutations of theBMPR2 gene (table 2, fig 1). The possibility of incorrect paternity was excluded by the analysis of informative markers. The mutation observed in patient 10 was also detected in two further sporadic patients, 11 and 12, ascertained independently. No parental samples were available and the possibility of the patients having inherited the mutation from a common ancestor was excluded through the examination of genotypes from microsatellite markers from within and surrounding the BMPR2 gene on chromosome 2.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Restriction fragment analysis and segregation of mutations of the BMPR2 gene. (A) The mutation in patient 5 (R322X), detected by loss of a TaqI restriction site, was also present in the patient's asymptomatic father as indicated by the presence of the 237 base pair (bp) fragment. (B) The mutation in patient 10 (2386delG), detected by loss of a Fnu4H I restriction site producing an additional 114 bp fragment, is a spontaneous (de novo) mutation since it is not present in either parent.

In three additional patients, insertions of nucleotide residues in the genomic sequence occurred. In patient 4, an additional thymine was detected in exon 6 at position 787 (table 2). In patient 7 both a guanine and adenine were inserted at nucleotide position 1247-8 of exon 9, while in patient 9 an adenine was inserted in exon 12 at position 1969 and confirmed through restriction digest analysis withMseI (table 2). Each of the mutations predicts premature truncation of the BMPR-II protein through shifts of the reading frame (table 2). Parental material was not available for analysis for these patients.

For the remaining three patients (3, 5, and 13) the substitution of cytosine for thymine at CpG dinucleotides occurred in exons 6, 8, and 12, respectively (table 2). In all three patients the mutations result in the change of the encoded amino acid arginine CGA to the stop codon TGA. These sequence changes were confirmed by restriction digest analysis of exonic PCR products with eitherHaeIII (exons 6 and 12) orTaqI (exon 8). While parental material was not available for patients 3 and 13, restriction analysis showed the mutation to be of paternal origin in patient 5. None of these sequence changes were detected in the analysis of a panel of at least 150 chromosomes from unrelated normal subjects, indicating that these mutations are not polymorphisms. Mutations of the entire protein coding sequence of the BMPR2 gene were also excluded, by direct sequencing, in the remaining 37 patients diagnosed with sporadic PPH.

Discussion

The World Health Organization maintains the broad classification of PPH into familial and sporadic forms (http://www.who.int/ncd/cvd/pph.html). However, several lines of evidence point to the possibility of a common genetic basis for both the sporadic and familial forms of the disease. Firstly, the disease gene, although inherited as an autosomal dominant trait, acts with low penetrance, so in the absence of detailed genealogical data, familial cases may be easily overlooked.15 Secondly, both forms of the disease share the same histopathological features and follow a similar clinical course.1 The present study provides direct evidence that germline mutations distributed throughout theBMPR2 gene predispose to sporadic PPH in addition to the familial form of the disease (table 1, fig2).11

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Diagram of the TGF-β signalling pathway and position of BMPR-II mutations. Following ligand binding, BMPR-II forms a heteromeric complex with a type I receptor, resulting in activation of the type I receptor kinase domain. This initiates phosphorylation of cytoplasmic signalling proteins, termed Smads, leading to signal transduction.12 16 The identified heterozygous germline BMPR2 mutations, illustrated on the BMPR2-II molecule (upper right), are likely to perturb the normal cellular effects of pathway activation, including growth inhibition, and extracellular matrix deposition leading to the characteristic histological features seen in the pulmonary vasculature in PPH patients.•=missense mutation,•=nonsense mutation, ✚= insertion, ▴=deletion. The corresponding patient number is noted to the right of the symbol.

We screened the entire protein coding sequence of theBMPR2 gene by direct sequencing and identified heterozygous mutations in approximately one quarter of the study cohort. This strategy, while capable of detecting splice site and coding region mutations, would fail to pick up non-coding region substitutions in either the promoter or introns as well as larger gene deletions and rearrangements. In addition, the direct sequencing methods used in the study may not detect all heterozygous mutations. Hence, the proportion of patients with sporadic PPH harbouring germline BMPR2mutations is likely to be greater than the 26% detected in this study.

To examine the frequency of spontaneous mutation of theBMPR2 gene in patients with sporadic PPH, we looked for the presence or absence of the disease allele in parental samples. DNA was available from both parents for five of the 13 probands studied. In two patients, the mutation was not detected in either parent and had clearly arisen spontaneously (table 2). In each of the remaining three families, paternal transmission was shown yet detailed pedigree analysis failed to indicate any suggestion of preceding disease (table 2). The present study shows germline mutations of BMPR2 to be present in at least 26% of patients without a family history of the disease, yet the familial form of the condition appears to be uncommon.5 A similar range and type of mutation was detected in the sporadic patients as compared to those reported in a recent series of familial PPH patients.11

One possible explanation for the markedly reduced penetrance for this disease gene is that the vascular abnormalities characteristic of PPH are triggered by the acquisition of a second somatic mutation within the BMPR-II signalling pathway. This situation would be analogous to that found in many tumours where somatic mutations in TGF-β growth inhibitory pathways (for example, TGF-βRI, TGF-βRII, and Smad 3 and 4) lead to uncontrolled proliferation.16 In PPH, endothelial cells comprising plexiform lesions show monoclonality, in contrast to a polyclonal origin in secondary pulmonary hypertension.17 Importantly, the only other reported human germline mutation of a type II receptor is a deleterious kinase domain amino acid substitution of TGF-βRII.18 This mutation predisposes to colorectal cancer, with the loss of the wild type allele in the tumour tissue, as predicted by the Knudson “two hit” hypothesis of tumorigenesis. Alternatively, the presence of a heterozygous mutation in the BMPR-II signalling pathway, either constitutional or somatic, together with exposure to an environmental insult could trigger abnormal vascular remodelling. For example, pulmonary hypertension indistinguishable from PPH may present in association with HIV infection or following ingestion of appetite suppressants, most recently of the fenfluramine/dexfenfluramine group.19-21 Additionally, an epidemic of cases was reported following the ingestion of adulterated rapeseed oil in northern Spain.22 Finally, the markedly reduced penetrance of the disorder could arise from the effects of one or more modifier genes, either autosomal or sex linked in nature.

The identification of inactivating heterozygous mutations of theBMPR2 gene in familial and sporadic PPH underlines the importance of the TGF-β super family in the regulation of vascular development and integrity (fig 2). Mutations in the gene encoding endoglin (a TGF-β receptor complex accessory protein) and the putative TGF-β type 1 receptor, ALK-1, have been identified in hereditary haemorrhagic telangiectasia, a condition characterised by arteriovenous malformations.23 The main effects of TGF-β on vascular cells are growth inhibition, cell differentiation, and stimulation of collagen synthesis.12 16 The detailed molecular mechanisms associated with the pathogenesis of PPH will require further investigation, particularly focusing on identifying the key ligands, interacting receptors, and downstream mediated pathways.

PPH is a diagnosis of exclusion and the current absence of specific diagnostic tools frequently contributes to the delay in securing the diagnosis. Molecular genetic analysis, based upon the detection of germline mutations of the BMPR2 gene must now be considered as a potential diagnostic aid for this group of patients. Because the risk for disease in first degree relatives of a newly diagnosed patient with PPH is low, uncertainty has persisted regarding appropriate strategies for clinical screening of at risk family members. It will be of considerable importance to ensure that the use of genetic analysis in the diagnosis of PPH be supported by appropriate genetic counselling, as part of an integrated programme of family based screening. The detection of germline mutations of theBMPR2 gene will enable appropriate targeting of screening and intervention.