Idiopathic pulmonary arterial hypertension (IPAH; formerly known as primary pulmonary hypertension, PPH) is a rare disorder characterised by pulmonary vascular proliferation and remodelling resulting in loss of patency of the pulmonary arteries. This leads to sustained elevation of pulmonary artery pressures with subsequent progressive right heart failure and often death.^1–3 While the majority of cases (>90%) have no known family history of the disease, ∼6% of the cases are known to be familial and are inherited in an autosomal dominant manner with markedly reduced penetrance.⁴ In previous studies, linkage analysis identified a familial IPAH locus on chromosome 2q33.^5,6 Subsequent heterozygous germline mutations in the bone morphogenetic protein receptor, type II (BMPR II) gene (BMPR2) were identified in at least 50% of familial cases.^7–9 BMPR-II is a member of the transforming growth factor-β (TGF-β) receptor superfamily signalling pathway, which is critical in both cell differentiation and cell growth mediated through transcriptional regulation.^10,11 Interestingly, germline mutations were also identified in 26–40% of sporadic cases.^12,13 This study was performed to establish the frequency of BMPR2 mutations in a different population of patients with sporadic IPAH.

METHODS

Patients

Between March 2002 and January 2004, 99 consecutive non-related patients >20 years of age with IPAH and negative family history were evaluated. In all cases, diagnosis of IPAH was made by experienced and specialised cardiologists or pulmonologists according to WHO criteria.^1,14 Other causes of pulmonary hypertension such as illicit drug abuse or anorectic drug use were excluded by careful medical history evaluation. Underlying heart and lung diseases (recurrent pulmonary embolism, connective tissue disease, obstructive or restrictive lung disease) and HIV infection were excluded by a cascade of clinical examinations including laboratory tests, arterial blood gas analysis, chest x ray, pulmonary function tests, echocardiography, ventilation perfusion scanning, chest computed tomography, pulmonary angiography (when indicated), and cardiac catheterisation (table 1). EDTA blood samples were collected for genetic analysis in all patients. The study protocol was approved by the Ethics Committee of the Medical Faculty of the University of Heidelberg and written informed consent was given by each patient.

BMPR2 mutation analysis

Mutation analysis of the BMPR2 coding sequence and intron/exon boundaries was performed in two independent laboratories using two different detection methods. In one laboratory (RK, BJ) the entire BMPR2 coding sequence and all intron/exon boundaries were analysed by dHPLC (WAVE, Transgenomic, Omaha, NE, USA). Primer sequences used for dHPLC are listed in table 2. Genomic DNA samples of all 99 patients were analysed using at least two different dHPLC protocols per PCR fragment according to the manufacturer’s instructions. Any fragments demonstrating aberrant properties were subjected to DNA sequence analysis. In the second laboratory (MWP, WCN) the entire coding region including intron/exon boundaries of the BMPR2 gene was investigated by direct sequencing (BigDye Terminator version 2.0, Applied Biosystems) using an ABI 3700 DNA Analyzer (Applied Biosystems, Foster City, CA, USA). The PCR conditions and primer data are available at http://www.med.uni-heidelberg.de/humangen/ger/humgen/pph-scl.htm. The analyses were conducted concurrently in a blinded fashion and the results were compared only after analyses were completed by each laboratory.

RT-PCR

cDNA analysis was performed by RT-PCR to determine the significance of the potential splice-site mutation. RNA was prepared from peripheral blood leukocytes using TRIzol reagent, as described by the manufacturer (Life Technologies, Carlsbad, CA, USA). Total RNA (1 µg) was reverse transcribed using random hexamer primers (Pharmacia Biotech Piscataway, NJ, USA) and 200 units of MMLV reverse transcriptase (SuperScript, Life Technologies) for 1 h at 42°C. RT-PCR products were obtained using region specific primers (sequences available on request) and subsequently analysed by dye terminator sequencing.

RESULTS

Heterozygous germline BMPR2 mutations were identified in 11 of the 99 (11.1%; 95% confidence interval: 5.7 to 19%, calculated by exact approximation) sporadic IPAH patients (table 3). We identified five novel nonsense mutations (E31X, R321X, Q450X, W466X, Q495X), one known missense mutation (R491W),⁷ and one known nonsense mutation (R584X).⁸ Additionally, we identified two novel splice site mutations. The genomic DNA of patient K3771A had a substitution at the first nucleotide of intron 8 (IVS8+1G→T). No RNA was available to test the consequence of this splice mutation on the protein level. The genomic DNA of patient MHH10 demonstrated a change at the third nucleotide of intron 10 (IVS10+3A→T), which leads to a deletion of exon 10. Deletion of exon 10 results in the frameshift G426fs(+27) (fig 1).

In two cases we identified novel frameshift mutations, both predicted to result in premature termination of the BMPR-II peptide. In the first case, K4690A, we found a 4 bp deletion (c.1313–1316 del CAGA) in exon 10 that generates a stop codon 35 amino acids downstream of the mutation. In patient MHH52, we detected a 1 bp insertion (c.1388 insA), which results in a stop codon six amino acids downstream of the mutated codon E464.

We detected an unclassified variant (UV) in patient MHH07 which results in the substitution of glutamic acid for glycine at codon 386 (fig 2). It seems likely that the substitution of glycine for glutamic acid in the kinase domain of BMPR-II, a change from an acidic to an aliphatic amino acid, has functional relevance. In the absence of functional data, however, the substitution E386G has been characterised as a UV.

Analysis of parental DNA samples for the E31X mutation detected in patient K4518A confirmed the mutation in the non-symptomatic mother. No parental DNA was available for any of the other cases in which BMPR2 mutations were identified. All results were confirmed by both laboratories; each mutation was detected by both of the methods described above.

DISCUSSION

Our findings of 11 heterozygous germline BMPR2 mutations in 99 sporadic IPAH patients (11%) are significantly different from those of previously published studies. Thomson et al determined the frequency of BMPR2 mutations in a cohort of 50 sporadic IPAH patients and found germline mutations in 13 of the patients (26%).¹² Their conclusion was that germline mutations predispose to sporadic IPAH as they do in the familial form of the disease. Morisaki and colleagues found 12 mutations in a group of 30 Japanese patients (40%).¹³ They hypothesised that some of these cases might actually be familial in nature, since most of their samples were from adults for whom no parental DNA was available. Our findings of only an 11% mutation rate in our cohort of 99 IPAH patients is significantly different from both the studies by Thompson et al (p = 0.0196) and Morisaki et al (p = 0.0003). There are several potential explanations for this discrepancy. The first is population differences between the three cohorts. Regarding the study by Thomson et al, both our cohort and theirs are of Northern European descent. Despite this seemingly common ancestry, it is possible that different environmental factors and lifestyles could affect the penetrance and fitness of BMPR2 mutations as well as the phenocopy rate. In comparison to the study by Morisaki et al, the population difference is more pronounced in that their cohort is of Japanese ancestry. Therefore, genetic differences affecting penetrance of BMPR2 mutations between the two populations may also play a role. Second, the smaller sample size of both the Thomson et al and the Morisaki et al studies introduces the potential for overestimation of the prevalence of BMPR2 mutations in sporadic IPAH due to sampling error. Third, patient recruitment differences may have occurred between the three studies. Any other causes of pulmonary hypertension were thoroughly excluded in our cohort. Finally, to ensure that no mutations were missed in our patients due to low grade mosaicism or sequencing artefacts, two different methods were used in two independent laboratories: direct sequencing and dHLPC. However, it cannot be ruled out in any of the three studies that undetected gene rearrangements, large deletions, insertions, or intronic mutations in the BMPR2 gene are responsible for the clinical phenotype in those patients for whom no mutations were identified. Thus, we conclude that the disparity between the estimated frequencies of mutations in the three studies is not due to differences in methodology.

Recent studies have tried to determine the role of BMPR2 in the pathogenesis of pulmonary arterial hypertension (PAH). Atkinson and coworkers reported a reduced expression of BMPR2 in lung tissue in all patients with PAH.¹⁵ Most remarkable was the reduction in expression to far below the expected 50% in patients with a BMPR2 mutation. Neither haploinsufficiency nor a dominant negative effect, both previously suggested to explain the pathogenic outcome of dominant BMPR2 mutations,⁴ can explain this phenomenon. Geraci and colleagues have used microarray analysis to characterise the gene expression pattern in IPAH patients.¹⁶ Interestingly, comparison of expression patterns could be used to distinguish between sporadic IPAH and familial IPAH. They conclude that sporadic and familial IPAH are mechanistically distinct. Taken together, both the published expression data and our mutation data suggest that while reduction of BMPR2 expression is common in pulmonary hypertension patients, additional factors other than germline BMPR2 mutations may lead to the development of sporadic IPAH.

Although BMPR2 plays an important role in the pathogenesis of IPAH, our data indicate that in most cases of sporadic IPAH BMPR2 mutations are not the primary cause of the disease. However, identification of individuals with BMPR2 mutations could enable more accurate risk assessment in other family members. Therefore, despite the low mutation frequency in sporadic IPAH, genetic screening for germline BMPR2 mutations may be warranted.