Abstract
Primary pulmonary hypertension (PPH) is a rare but severe and progressive disease characterised by obstructive lesions of small pulmonary arteries. Patients with PPH often have mutations in the bone morphogenetic protein receptor type II (BMPR2) gene, whereas some carry mutations in the activin receptor-like kinase 1 (ALK‐1) gene, generally associated with hereditary haemorrhagic telangiectasia (HHT) type 2, a vascular dysplasia affecting multiple organs. The aim of this study was to determine whether members of families with PPH and confirmed or probable HHT had ALK‐1 mutations.
ALK‐1 and BMPR2 mutation analysis was performed on deoxyribonucleic acid from affected members of four families with PPH and confirmed or suspected HHT.
ALK‐1 mutations were identified in all four families and three novel mutations found in exon 10, leading to truncated proteins. In the fourth family, a missense mutation, previously reported in four independent HHT families, was detected in exon 8. Analysis of the BMPR2 gene revealed no exonic mutations in the probands with both PPH and HHT.
The present data bring to 10 the number of reported families with primary pulmonary hypertension and hereditary haemorrhagic telangiectasia type 2, representing 16% of the 61 families with known activin receptor-like kinase 1 mutations. Such mutations might predispose to primary pulmonary hypertension, and specialists should be aware of the potential link between these two disorders.
This study was supported by grant HHT-FY02-226 (M. Letarte) from the March of Dimes (New York, NY, USA), and by grants NIH-HL60056 (J.H. Morse) from Columbia University (New York, NY, USA) and NIH-HL49171 (D.A. Marchuk) from Duke University (Durham, NC, USA).
Primary pulmonary hypertension (PPH) is a rare disease with an estimated incidence of 1–2 cases per million population 1. The presenting symptoms usually include fatigue, anorexia and shortness of breath, which, if left untreated, lead to a progressive increase in pulmonary arterial pressure, right ventricular failure and death 2. The affected small pulmonary arteries and arterioles are characterised by intimal proliferation, medial hypertrophy, concentric fibrosis and the presence of plexiform lesions composed of both vascular smooth muscle cells and endothelial cells 3. Monoclonal endothelial cell proliferation is found in the plexiform lesions of PPH but not in secondary pulmonary hypertension 4.
Recently, PPH has been found to be caused by mutations in either of two genes: the bone morphogenetic protein receptor type II gene (BMPR2) 5–7 and the activin receptor-like kinase 1 gene (ACVRL1 or ALK‐1) 8, both members of the transforming growth factor‐β (TGF‐β) receptor superfamily. A recent abstract reported an endoglin gene (ENG) mutation in a patient with HHT and dexfenfluramine-associated PPH 9. ALK‐1 and ENG are the two genes associated with hereditary haemorrhagic telangiectasia (HHT). Mutations in ALK‐1 lead to HHT type 2 (HHT2) 10, 11, whereas mutations in ENG are responsible for HHT type 1 (HHT1) 12. HHT is an autosomal dominant vascular disorder that occurs at an incidence of >1 in 10,000 population. It is characterised by vascular dysplasia with the formation of mucocutaneous telangiectases and arteriovenous malformations (AVMs) in the lung, brain and liver 13.
To date, 46 unique BMPR2 mutations have been reported in patients with a family history of PPH 5–7, as well as in some patients with spontaneous PPH 14. These mutations include missense, nonsense and frameshift mutations, as well as splice site mutations, occurring in the ligand-binding, transmembrane, kinase and cytoplasmic tail domains of BMPR2. ALK‐1 mutations associated with PPH are also of varying types and found throughout the gene 8. Hence, the identification of mutations in both BMPR2 and ALK‐1 genes underlines the importance of the TGF‐β superfamily members in the maintenance of vascular integrity 15. It also suggests that mutations in these related signalling pathways may lead to an imbalance in the regulation of TGF‐β/bone morphogenetic protein (BMP)‐mediated signals in endothelium that are manifest as vascular dilatation in HHT and vascular obliteration/obstruction in PPH. Dysregulation of these same pathways was recently reported for various forms of nonfamilial PPH 16.
In the present study, ALK‐1 and BMPR2 mutation analysis was performed in four families with PPH and either known or suspected HHT. The present article reports ALK‐1 mutations in all four PPH families and discusses the significance of these findings.
Materials and methods
Study subjects
All studies and procedures were reviewed and approved by the Columbia Presbyterian Medical Center Institutional Review Board (Columbia University, New York, NY, USA), the Research Ethics Board of the Research Institute of the Hospital for Sick Children (Toronto, Canada) and Duke University Medical Center Institutional Review Board (Duke University Medical Center, Durham, NC, USA).
The original study group consisted of a cohort of 104 families with either the familial form of PPH (two or more affected family members) or seemingly sporadic cases, referred for genetic evaluation from 1998 until January 2003. The diagnosis of PPH was made using a combination of results from echocardiography, right heart catheterisation and, when available, histological studies of the lung. The evaluation and work-up excluded other causes of PPH, such as human immunodeficiency virus infection, connective tissue diseases and the use of appetite suppressant drugs. In four of the families (families 60, 82, 91 and 100), a definite diagnosis of HHT was made based on the Curaçao criteria 17. Probands from the remaining 100 families did not fulfil the criteria for HHT. Careful family histories of these families failed to find any other members with HHT.
The results of cardiac catheterisation confirmed the diagnosis of PPH in probands from the four families (families 60, 91 and 100 on site, 82 off site). Their pedigrees were prepared from medical records and interviews with family members, and are illustrated in figure 1⇓. The clinical features, results of right heart catheterisation and mutation analyses for the probands are illustrated in table 1⇓.
Mutation analysis of activin receptor-like kinase 1 and bone morphogenetic protein receptor type II genes
Deoxyribonucleic acid (DNA) samples from the patients were analysed for ALK‐1 mutations in two centres. In two of the families (families 60 and 82), coding exons of the ALK‐1 gene were analysed by sequencing using the Open Gene Automated DNA Sequencing System II (Visible Genetics, Inc., Toronto, Canada), as described previously 18. DNA samples from the other two families (families 91 and 100) were analysed using the BigDye Terminator Cycle Sequencing Ready Reaction (Applied Biosystems, Foster City, CA, USA) and run on an ABI Prism 3100 sequencer (Perkin Elmer, Wellesley, MA, USA). A Sequencher (version 4.1.4; Gene Codes Corp., Ann Arbor, MI, USA) was used to analyse the data. Samples with sequence changes were reamplified and resequenced for verification.
The 13 BMPR2 exons of the four probands were sequenced as previously described 5.
Results
Figure 1⇑ illustrates the presence of HHT and PPH in the families analysed. Table 1⇑ illustrates the clinical features, results of right heart catheterisation and mutations found in the four probands. ALK‐1 mutations were identified in four families with both PPH and a confirmed or suspected diagnosis of HHT. The three novel mutations are predicted to lead to truncations in exon 10, whereas the fourth mutation is a previously reported missense mutation in exon 8 (table 1⇑). None of the four probands showed exonic mutations in the BMPR2 gene.
Family 60
In this family, a novel complex mutation (cytosine (C) to thymine (T) substitution at complementary DNA base 1450 with insertion of guanine (G) between bases 1450 and 1451 (1450C>T, 1450_1451insG)) was identified in exon 10 of ALK‐1. It causes a substitution at amino acid 484 followed by a frameshift with truncation at residue 493 in the kinase domain (table 1⇑). This mutation was identified in the 11‐yr‐old proband, who is still alive and received epoprostenol therapy for PPH. This patient also had reactive airway disease, bruised easily and experienced frequent spontaneous haemorrhages from the nose, gums and skin for several years. The unaffected sibling did not carry the familial mutation. Their father, who had a history of very frequent epistaxis, was never diagnosed with HHT, but had PPH and died aged 36 yrs after lung transplant rejection (fig. 1⇑). The lung surgical pathology report described histological features of PPH with intimal proliferation, medial hypertrophy, and concentric intimal fibrosis and plexiform lesions (Heath and Edwards grade V/VI). There was also dilatation of small pulmonary arteries and arterioles compatible with HHT. None of the father's five siblings showed symptoms of PPH or HHT nor did they carry the ALK‐1 mutation. The familial mutation was not found in the proband's asymptomatic grandmother. DNA from the grandfather, who died of emphysema, was not available for analysis, and two of the grandfather's siblings died of pancreatic carcinoma.
Family 82
The nonsense mutation found in this PPH family was a 1435C>T (table 1⇑). It was detected in exon 10 of ALK‐1 and leads to a stop codon at arginine 479. The male proband, followed in Europe on epoprostenol therapy, succumbed to PPH aged 29 yrs. The proband's father and grandmother also carried this mutation and all three patients had a confirmed diagnosis of HHT (fig. 1⇑). The proband's sister also has HHT and no symptoms of PPH, but the sister's DNA was not available for testing.
Family 91
The missense mutation identified in exon 8 of ALK‐1 in this family (1120C>T) leads to substitution of arginine 374 (table 1⇑). This mutation has been previously described as the causative mutation for HHT in four unrelated families with no reported signs of PPH 11, 19, 20. The proband had a history of HHT and died of PPH aged 51 yrs. The proband had severe epistaxis and had undergone nasal septal dermatoplasty; coumadin anticoagulation, however, had to be discontinued due to the severity of the epistaxis. Abdominal ultrasonography results were consistent with multiple hepatic AVMs. Liver biopsy revealed vascular abnormalities consistent with HHT and perivascular congestion and fibrosis consistent with right heart failure. Head CT results were normal. The proband's mother also has HHT, but a detailed family history is unavailable.
Family 100
The nonsense mutation found in exon 10 in this family is a 1385C>G causing truncation at serine 462 (table 1⇑). The female proband had nosebleeds and telangiectases consistent with HHT. The proband also had multiple pulmonary AVMs: a large spontaneously thrombosed right upper lobe AVM, and AVMs in the left upper and lower lobes of the lung. Nonenhancing areas in the right lobe of the liver were compatible with thrombosed AVMs. The proband died of PPH aged 20 yrs, and was not a candidate for embolisation of the pulmonary AVMs. The proband's sister exhibited no manifestations of HHT and did not carry the mutation in the ALK‐1 gene. The familial mutation was detected in the proband's affected mother but not in the mother's three siblings, who showed no symptoms of either PPH or HHT (fig. 1⇑). All of the other family members affected with HHT had a history of severe nosebleeds and prominent telangiectases.
Discussion
From a cohort of 104 PPH families initially referred for BMPR2 mutation analysis, four probands with HHT were identified. They fulfilled three of the four Curaçao diagnostic criteria for HHT 16 with telangiectases, severe epistaxis and a family history of HHT. In addition, one proband (family 100) had known pulmonary and hepatic AVMs and another (family 91) hepatic AVMs. All four probands showed pulmonary arterial hypertension, the severity of which was diagnosed at cardiac catheterisation and required epoprostenol therapy.
These patients had mutations in the ALK‐1 gene, confirming the diagnosis of HHT2. The missense mutation in exon 8 has been previously described in several HHT families 11, 19, 20, in which none of the patients had PPH. The three mutations in exon 10 were novel and comprised two nonsense mutations and one complex mutation that would be predicted to result in truncated proteins. It is interesting to note that 10 (19%) of the 53 reported mutations in the ALK‐1 gene (present in a total of 61 families) were identified in patients with PPH and clinical manifestations of HHT. In the first report of PPH in HHT families, six ALK‐1 mutations were found, one each in exons 2, 3, 6 and 8 and two in exon 10 8. The new total of five exon 10 mutations in the 10 known PPH/HHT probands suggests that this region of the protein is functionally important since its absence might predispose to pulmonary hypertension.
Two of the mutations in exon 10 fall within the conserved carboxyl-terminal region of ALK‐1, which comprises residues 479–489, and is referred to as the nonactivating nondownregulating box (NANDOR BOX) 20, 21. The substitution 1435C>T results in a termination codon at position 479, thus deleting the NANDOR BOX. The complex mutation (which modifies residue 484 and subsequent residues and causes termination at residue 493) would also alter the sequence and structure of the NANDOR BOX. The third exon 10 mutation (1385C>G) also leads to a truncated protein lacking this region. Data obtained with similar truncation mutants of the related transforming growth factor‐β (TGF‐β) type I receptor, ALK‐5, have shown that this domain is important in TGF‐ β‐induced receptor signalling downregulation 21. Signals inhibited in the ALK‐5 mutants included induction of plasminogen-activator inhibitor‐1 and fibronectin and phosphorylation of mothers against decapentaplegic, homolog 2 (Drosophila) (Smad2) 21. Thus mutations leading to truncation or deletion of the NANDOR BOX of ALK‐1 probably impair its signalling activity, yielding a nonfunctional protein associated with HHT2 and PPH 20.
No mutations were detected in the BMPR2 gene in the four PPH probands with HHT. However, it is interesting to compare the potential effects of ALK‐1 and BMPR2 mutations in inducing PPH. Both ALK‐1, a type I receptor, and BMPR2, a type II receptor, are serine/threonine kinases that belong to the TGF‐β superfamily. Activation of either of these receptors through TGF‐β and BMP2/4 respectively leads to phosphorylation of Smad1 and Smad5, resulting in proliferation and migration of endothelial cells 22. These data suggest that regulation of TGF‐β/BMP‐mediated endothelial pathways is critical in sustaining the vascular integrity of the pulmonary circulation.
It is difficult to differentiate clinically between the two types of HHT. Pulmonary AVMs are more frequent in patients with HHT1, although they have also been reported in patients with HHT2 23–25. Several reports have suggested an increased prevalence of liver involvement in families with HHT2 26–28. HHT2 is also associated with lower penetrance, milder phenotype and later disease onset. However, recent evidence for the association of PPH, a severely debilitating and fatal disorder, in patients with HHT2 stresses the importance of performing molecular analysis for HHT, particularly HHT2, in PPH patients without BMPR2 mutations to identify a genetic cause for the disorder. Including the present study, there are now 10 known families with ALK‐1 mutations and both PPH and HHT2, in contrast to the one ENG mutation in the HHT1 patient with appetite suppressant PPH 9. They represent 16% (10 of 61) of the total number of families reported with ALK‐1 mutations and HHT2, a significant proportion.
HHT can be difficult to detect clinically, particularly in young children, as illustrated by the 11 yr‐old child with PPH from family 60. This child showed frequent epistaxis but had not developed telangiectases, which often become evident between the ages of 20 and 40 yrs 24. In addition, the family lacked a history of HHT. A diagnosis of HHT was not entertained in her 36‐yr‐old father ante mortem despite PPH, a history of severe epistaxis, and histological findings compatible with both PPH and HHT, with pulmonary plexiform lesions and dilatation of the small arterioles. In such cases, awareness of HHT symptoms in PPH patients becomes important in terms of identifying a familial condition and its potential risks. Reciprocally, clinical screening of patients known to have an ALK‐1 mutation and their first-degree relatives may have the potential advantage of identifying markers of PPH and possibly allow earlier disease detection and intervention. Based on overall risk/benefit considerations, estimation of pulmonary arterial systolic pressure by Doppler echocardiography is currently the most useful screening tool for the detection of asymptomatic pulmonary hypertension and should be considered in HHT2 families.
Additional factors, whether genetic or environmental, may be required for the onset of PPH in subjects with mutations in either ALK‐1 or BMPR2 genes. Both diseases show autosomal dominant inheritance, but PPH families with BMPR2 mutations show incomplete penetrance, and very few HHT patients with ALK‐1 mutations develop PPH. Appetite suppressant drugs and human immunodeficiency virus infection are known causes of PPH 1, and mutations in BMPR2 have been found in fenfluramine-associated PPH 29. Hence, appetite suppressants may become a more widely appreciated initiator of PPH over time in both HHT1 and HHT2 9. Voelkel et al. 30 hypothesised that PPH, like cancer, requires “two hits”. Mutations in ether ALK‐1 or BMPR2 genes would represent the first “hit” and predispose to vascular changes.
Future studies should determine the factors contributing to the development of primary pulmonary hypertension in patients with activin receptor-like kinase 1 or bone morphogenetic protein receptor type II mutations.
Acknowledgments
The authors wish to thank the patients, their families and the physicians who helped with the clinical aspects of this study.
- Received July 23, 2003.
- Accepted October 9, 2003.
- © ERS Journals Ltd