Genetics of pulmonary hypertension: from bench to bedside

The normal adult pulmonary circulation is a highly distensible, low resistance, low pressure circuit with little or no resting vascular tone. The thin walls of pulmonary vessels, the lack of well-developed arterioles, and the crescentic shape of the thin-walled right ventricle reflect the low ambient haemodynamic stress. Pulmonary arterial hypertension (PAH) refers to a spectrum of diseases in which the pulmonary artery pressure is persistently elevated (mean pulmonary artery pressure >25 mmHg), in the absence of a raised pulmonary artery wedge pressure (<12 mmHg), excluding a post capillary process (mostly represented by left heart disease) 1, 2. This severe disorder leads to progressive right heart hypertrophy, dilatation and failure. PAH typically results from chronic obstruction of small pulmonary arteries due at least in part to endothelial and vascular smooth muscle cell dysfunction and proliferation 2, 3.

A revised diagnostic classification of PAH has been recently proposed 2, 4. This classification reflects recent advances in the understanding of pulmonary vascular diseases, and recognises the pathophysiological and therapeutic similarities between “unexplained” pulmonary hypertension (primary or idiopathic pulmonary hypertension) and PAH coexisting with other disorders such as collagen vascular diseases (scleroderma spectrum, systemic lupus erythematosus, mixed connective tissue disease), human immunodeficiency (HIV) infection, portal hypertension, congenital systemic to pulmonary shunts, and anorexigen exposure (table 1⇓). In addition, this classification identifies conditions resulting from other mechanisms for which specific therapies exist, including chronic thromboembolic disease (obstructive pulmonary hypertension which can be cured by thromboendarteriectomy), chronic hypoxic lung disease (vasoconstrictive pulmonary hypertension which may benefit from long-term oxygen therapy) and left-sided cardiac disease (passive postcapillary venous pulmonary hypertension, which can deteriorate with prostanoids with a risk of pulmonary oedema).

Primary pulmonary hypertension (PPH) is an uncommon disorder, with an estimated yearly incidence of 2 per million people 5. PPH is twice as common in females than males and symptoms develop typically in the third and fourth decades of life, although the disease may occur at any age 6. The mainstay of treatment is anticoagulants and pulmonary vasodilators including calcium channel antagonists, prostaglandins (epoprostenol and other prostacyclin analogues), and endothelin receptor antagonists. When medical therapy is ineffective or unavailable, lung or heart-lung transplantation is currently the only alternative therapy. Until recently, the cellular and molecular basis of this devastating disease has remained elusive.

Familial primary pulmonary hypertension

PPH can be either sporadic or clustered in families. The first detailed description of familial PPH by Dresdale et al. 7 in 1954 included a description of related subjects with severe pulmonary vascular disease of unknown aetiology. Later, it became apparent that this familial condition was less rare than initially believed, in part because of patient and physician unawareness of the familial occurrence of the disease, but also because of the markedly reduced penetrance of the genetic defects 2. A pioneering report by Loyd et al. 8 in 1984, clearly supported the concept that a proportion of cases of so-called sporadic PPH were in fact mislabelled as nonfamilial when family histories were incomplete, suggesting that the incidence of familial PPH is higher than previously reported. Because of the lack of data, the true incidence of familial PPH is unknown. The best estimate has been provided by the National Institute of Health Registry for PPH reported by Rich et al. 9 in 1987. This prospective analysis of 187 PPH patients from 32 North American medical centres identified 12 cases with familial history affecting a first-blood relative. Consequently, it has been widely accepted that ≥6% of individuals diagnosed with PPH have a family history of the disorder, and thereby identifying first-degree relatives as being at additional risk of developing the disorder. Familial PPH segregates as an autosomal dominant trait but with markedly reduced penetrance 2, 8. Recognition of an inherited form of this disorder enabled chromosomal localisation using linkage analysis, the locus designated PPH1 on chromosome 2q31–33 10. The completion of the first draft of the Human Genome Project and the contribution of additional patients and families contributed to the discovery of the PPH1 gene in the year 2000, using positional cloning. The bone morphogenetic protein receptor type II (BMPR-II) gene (BMPR2), coding for a receptor member of the transforming growth factor (TGF)-β family, has been identified as causative of familial PPH 11, 12. BMPR-II is a ubiquitously expressed receptor for a family of secreted growth factors named bone morphogenetic proteins (BMPs) 13 (fig. 1⇓). These proteins are part of the TGF-β superfamily, which include the three mammalian TGF-β isoforms, the activins and inhibins, together with >30 members of the BMP subfamily 13. BMPs were originally identified as proteins regulating growth and differentiation of bone and cartilage but more recent studies have revealed that these multifunctional cytokines regulate growth, differentiation and apoptosis of various cell types. Members of the TGF-β family signal by binding to type I and II cell surface receptors and activating serine/threonine kinases by formation of heteromeric complexes 14, 15.

By direct sequence analysis of the protein-encoding portion of the BMPR2 gene in PPH patients, germline heterozygous missense, nonsense splice-site and frameshift mutations have been detected in more than half of the families analysed 16. These mutations are predicted to interrupt the BMP-mediated signalling pathways, resulting in a predisposition to proliferation rather than apoptosis of cells within small pulmonary arteries. A number of possible explanations exist to account for the PPH families in which BMPR2 gene mutations have not been identified, including mutations in portions of the BMPR2 gene that are not sequenced in these studies (including promoter intronic and 3′-untranslated regions).

The majority of BMPR2 pathogenic mutations described by Machado et al. 16 (59%) were predicted to lead to a premature truncation of the BMPR-II transcript, most of which will be lost through a process known as nonsense-mediated ribonucleic acid decay. In 36% of mutations nonconservative amino acid substitutions occurred at highly conserved positions. Using either computer modelling of the structural features of the ligand binding domain or in vitro-expression analysis, these missense mutations were predicted to lead to loss of BMPR-II function. These data strongly support the hypothesis that the predominant molecular mechanism implicated in the dominant inheritance of PPH is one of haploinsufficiency or reduction in the available level of BMPR-II protein 16. However, in the absence of more detailed analyses an inhibitory effect of the defective proteins upon the normal actions of the BMPR-II complexes, a process referred to as a dominant/negative action, cannot be excluded.

These molecular genetic results suggest that the target cell(s) within the pulmonary arterial wall are sensitive to BMPR2 gene dosage and that the TGF-β pathway, mediated through BMPR-II, is critical for the maintenance and/or normal response to injury of the pulmonary vasculature. However, the age of onset of disease both within families and between subjects carrying identical but recurrent mutations is remarkably variable. These findings, together with the previously reported large number of individuals who harbour a deleterious mutation but show no clinical evidence of disease (nonpenetrance), points to the likely requirement for additional factors, either environmental or genetic, in the pathogenesis of the disease. Environmental factors may include drug exposure, as recently suggested in sisters harbouring BMPR2 germline mutations and developing severe PPH after exposure to fenfluramine derivatives 17. The acquisition of additional genetic factor defects, as in neoplasia, may involve other members of the TGF-β pathway. Recent microdissection studies of PPH lesions have suggested that proliferating endothelial cells harbour genetic alterations associated with microsatellite instability and concomitant perturbation of growth and apoptosis gene expression akin to neoplasia 18.

Finally, the marked sex bias for presentation of the disease may implicate either a hormonal component or a role for an X-linked locus in disease predisposition 6. An alternative suggestion for the observed female predominance was that the genetic defect(s) characteristic of the disease might lead to male foetus loss. This was supported by Loyd et al. 8 who showed that more females than males were born to subjects who were obligate carriers for the gene defect, suggesting selective wastage of male foetuses or a distorted primary sex ratio. An abnormal sex ratio of progeny was the first clinical finding which pointed to the possibility that the PPH gene may have a role in embryonic development. A significant trend for the disease to develop at a younger age in members of subsequent generations has also been reported 8. This manifestation, known as genetic anticipation, is reminiscent of the phenomena associated with increasing instability of trinucleotide repeat amplification, which forms the basis for several neurological diseases, including fragile X syndrome, myotonic dystrophy, and Huntington's disease. Anticipation in familial PPH was confirmed by Loyd et al. 8, showing that the age (±SD) at death was 45.6±14.5 versus 36.3±12.6, and 24.2+11 yrs in successive generations. To date no molecular genetic evidence to explain anticipation in PPH has been presented. It remains possible that an alternative explanation for this phenomenon in PPH, might reflect increased environmental exposure, rather than a genetic mechanism.

Arterial remodelling in PPH involves hypertrophy and proliferation of endothelial, smooth muscle cells, as well as fibroblasts, but how defects in BMPR-II contribute to these changes is currently unknown. However, BMP2 has been shown to inhibit growth of cultured aortic vascular smooth muscle cells indicating they have an intact BMP signalling pathway and that defects in BMP-mediated signalling may predispose to proliferation under certain environmental conditions. Recently, Morrell et al. 19 have demonstrated functional abnormalities of pulmonary artery smooth muscle cells isolated from PPH patients, including resistance to the antiproliferative effects of TGF-β1 and BMPs. The TGF-β superfamily plays a critical role in the development and maintenance of the integrity of the normal vasculature. Thus, the altered cellular responses may contribute to the formation of lesions that lead to the vascular obliteration observed in the pulmonary arteries of patients with PPH.

Sporadic primary pulmonary hypertension

In both familial and sporadic PPH patients, histological examination of pulmonary arteries shows increased muscularisation of pulmonary arteries, intimal hyperplasia, endothelial cell proliferation, and in situ thrombosis. These findings, together with the similar clinical features and progression of the disease and the determination of previously unknown co-ancestry between so-called sporadic PPH patients, suggest that a proportion of so-called sporadic PPH patients may indeed have the same underlying molecular genetic defect as seen in familial PPH. This is further supported by the fact that the disease gene, although inherited as an autosomal dominant trait, acts with low penetrance, hence in the absence of detailed genealogical data, familial cases may be easily overlooked.

Two recent studies have clearly demonstrated that so-called sporadic PPH may indeed represent unidentified familial cases. First, a search on several American families has uncovered a kindred affected by familial PPH that spans seven generations and involves five subfamilies initially not known to be related 20. This kindred included 12 affected members who were initially thought to have sporadic PPH and seven affected members whose conditions were first misdiagnosed as other cardiopulmonary diseases. Mutational analysis of the BMPR2 gene in six affected members revealed that all had the same missense mutation in exon 6. Further genealogical investigations revealed that all these families were in fact related, with common ancestors born in the nineteenth century. The discovery of this very large kindred affected by familial PPH supports the hypothesis that cases of presumed sporadic PPH are in fact inherited 20. As resolution of this question has important implications for patient diagnosis, genetic counselling and clinical screening, the BMPR2 gene has been investigated in a cohort of 50 unrelated patients with a clinical diagnosis of PPH and no identifiable family history of PPH, using direct sequencing of the entire coding region and intron/exon boundaries of the BMPR2 gene 21. In addition, deoxyribonucleic acid (DNA) from available parent pairs was used to assess the occurrence of spontaneous (de novo) mutations contributing to sporadic PPH. A total of 11 different heterozygous germline mutations of the BMPR2 gene were identified in 13 of the 50 PPH patients studied, including missense (n=3) nonsense (n=3), and frameshift (n=5) mutations each predicted to alter the cell signalling response to specific ligands and hence of functional significance. Parental analysis revealed three occurrences of paternal transmission and two of de novo (spontaneous) mutation of the BMPR2 gene in sporadic PPH (fig. 2⇓). Therefore, the sporadic form of PPH is associated with germline mutations of the gene encoding the receptor protein BMPR-II in ≥26% of cases 21. This suggests that molecular description of PPH patients, based upon the presence or absence of BMPR2 mutations may have important implications for screening of relatives.

Acquired genetic defects in PPH

Lee et al. 22 have shown that endothelial cells within PPH plexiform lesions demonstrate monoclonality, whereas PAH lesions develop via polyclonal expansion of endothelial cells in scleroderma or congenital systemic to pulmonary shunts. Microdissection studies by Yeager et al. 18 recently demonstrated that endothelial cells in PPH lesions are genetically unstable with 50% of lesions demonstrating microsatellite instability. Endothelial cells in PPH lesions harbour mutations permissive for clonal cell growth, including mutation of at least one allele of TGF-β receptor type II in one third of the lesions from sporadic PPH individuals. These results suggest that in some sporadic PPH patients proliferating endothelial cells have acquired (somatic) genetic alterations.

Other forms of pulmonary arterial hypertension

Exposure to appetite suppressants aminorex, fenfluramine and dexfenfluramine is a definite risk factor for PAH 2, 4, 5. Increased risk for PAH was first reported in the late sixties with exposure to aminorex and in the mid-nineties with fenfluramine or dexfenfluramine. However, the exact mechanism(s) by which fenfluramine derivatives promote PAH has not been established 23. Individual susceptibility factors are likely to play an important role since the absolute risk in the general population is low (around 1 case for 10,000 individuals exposed to fenfluramine derivatives). BMPR2 mutations were determined in patients with sporadic PAH who had taken fenfluramine derivative 17. BMPR2 mutations that predict changes in the primary structure of the BMPR-II protein were found in ≥9% of PAH patients with a history of fenfluramine derivatives intake 17. This differed significantly from normal healthy French controls, in whom no BMPR2 mutations could be identified. Moreover, the patients with mutations of the BMPR2 gene had a significantly shorter duration of exposure to fenfluramines as compared to patients with no evidence of mutations 17. Therefore, a proportion of patients with fenfluramine derivative-associated PAH have BMPR2 mutations. These BMPR2 defects do not rule out a role for additional mechanisms triggering the occurrence of the disease, including drug-induced alteration of the serotonin pathway. The possible relevance of serotonin in fenfluramine-induced PAH is supported by the fact that fenfluramine derivatives induce valvular heart disease very similar to carcinoid syndrome and that a decrease in platelet serotonin storage with enhanced blood concentration of free serotonin has been reported in sporadic cases of PPH 23. Further investigations should determine whether abnormal TGF-β signalling alone, or in combination with other factors, plays a role in the occurrence of PAH associated with fenfluramine derivative intake.

As discussed above, the current classification of PAH recognises the clinical, pathophysiological and therapeutical similarities between PPH and PAH associated with, for example, collagen vascular diseases, HIV infection, portal hypertension, congenital systemic to pulmonary shunts, and anorexigen exposure. The present authors and others have systematically assessed patients with well-characterised conditions associated with PAH. The only patient population in which BMPR2 mutations could be identified was that of fenfluramine-exposed individuals. In conclusion, despite striking similarities between PAH variants, germline BMPR2 genetic defects have only been identified to date in inherited cases of PPH and fenfluramine-associated cases.

Hereditary haemorrhagic telangiectasia (Osler-Weber-Rendu syndrome)

Other components of the TGF-β receptor pathway have been implicated in hereditary haemorrhagic telangiectasia. Hereditary haemorrhagic telangiectasia (Osler-Weber-Rendu syndrome) is an autosomal dominant vascular dysplasia characterised by mucocutaneous telangiectases affecting the nasal and gastrointestinal mucosa, lips, tongue, fingertips, and macroscopic arteriovenous malformations 24. These lead to recurrent epistaxis and gastrointestinal blood loss, as well as arteriovenous malformations, particularly in the pulmonary, hepatic, and cerebral circulation. Pulmonary artriovenous malformations create clinically significant right-to-left shunts, causing hypoxaemia, orthodeoxia, platypnoea, chest murmurs, paradoxical embolism, stroke, and cerebral abscesses. In this setting the occurrence of pulmonary hypertension is usually due to the hyperkinetic state promoting “high cardiac output heart failure” (venous “postcapillary” pulmonary hypertension) 24. However, PAH that is clinically and histologically undistinguishable from PPH may occur in this patient population 25.

Mutations in two genes encoding family members of the TGF-β receptor pathway underlie hereditary haemorrhagic telangiectasia 15, 24. These genes code for a type I receptor (activin-receptor-like kinase 1 (ALK-1), located on chromosome 12) and a type III receptor (accessory) (endoglin, located on chromosome 9), respectively. Clinical analysis of five kindreds plus one individual patient with hereditary haemorrhagic telangiectasia led to the identification of 10 PAH cases 25. These subjects suffered with “precapillary” pulmonary hypertension confirmed by right-heart catheterisation demonstrating elevated mean pulmonary artery pressure, low pulmonary artery wedge (“capillary”) pressure, low cardiac output and markedly elevated pulmonary vascular resistances. Family members spanning the whole spectrum of both conditions could be studied (i.e. unaffected family members, and family members with PAH, hereditary haemorrhagic telangiectasia, or both) (fig. 3⇓). Analysis of genes encoding TGF-β receptor proteins, including ALK-1, endoglin, and BMPR2, allowed the demonstration that PAH in association with hereditary haemorrhagic telangiectasia can involve mutations in ALK1 25 (fig. 3⇓). Lastly, immunohistochemical analysis of paraffin-embedded lung sections from a subject with plexiform PAH and mutations in ALK1 demonstrated the presence of ALK-1 protein product in diseased pulmonary vascular endothelium 25. This later finding lends support for the suggestion that endothelium cell dysfunction is crucial to the pathogenesis of both hereditary haemorrhagic telangiectasia and PAH. TGF-β signalling seems to affect both vascular differentiation and proliferation. The pleiotropic nature of TGF-β as a growth factor offers a potential explanation for the variable complications of hereditary haemorrhagic telangiectasia. The net effect of ALK1 dysfunction may depend on local vascular interactions and other environmental or genetic factors.

The serotonin hypothesis

“Unexplained” pulmonary hypertension in patients with rare diseases provides valuable tools to analyse a possible relevant pathobiological basis of PAH. Development of severe PAH with no identifiable cause in a patient with a rare familial platelet storage deficiency allowed Hervé et al. 26 to hypothesise that serotonin, a pulmonary vasoconstrictor and growth factor for vascular smooth muscle cells stored in platelets, could play a critical role in PAH. This observation subsequently led to the description of altered serotonin turnover leading to an increased availability of free serotonin in patients with PPH 27.

Recent studies have extended the hypothesis that serotonin may play a critical role in PPH. Hyperplasia of pulmonary artery smooth muscle cells is a pathological hallmark of PPH. Eddahibi et al. 28 recently showed that pulmonary artery smooth muscle cells from PPH patients grow faster than pulmonary artery smooth muscle cells from controls when stimulated by serotonin or serum. These effects were due to an increased expression of the serotonin transporter (5HTT) which mediates internalisation of indoleamine 28. In the presence of 5HTT inhibitors the growth stimulatory effects of serum and serotonin were markedly reduced and a difference between the growth of pulmonary artery smooth muscle cells from patients and controls was no longer observed. As compared with controls the expression of 5HTT was increased in cultured pulmonary artery smooth muscle cells as well as in platelets and lungs from patients with PPH, where it predominated in the media of thickened pulmonary arteries and in onion-bulb lesions. The L-allelic variant of the gene promoter, which is associated with 5HTT overexpression and increased pulmonary artery smooth muscle cell growth was present in 65% of PPH patients but in only 27% of controls 28. Therefore, 5HTT may play a significant role in the pathogenesis of pulmonary artery smooth muscle cells proliferation in PPH, and a 5HTT polymorphism could confer susceptibility to PPH.

Type Ia glycogen storage disease

The present authors have recently described in detail patients who developed PAH in the context of other rare diseases, allowing the analysis of novel pathophysiological pathways to be extended. For instance, hereditary haemorrhagic telangiectasia supported and extended the concept of altered TGF-β signalling in the context of severe precapillary pulmonary hypertension 25. More recently, patients displaying severe PAH and type I glycogen storage disease allowed further investigation of the possible promoting role of metabolic disturbances in PAH patients 29.

Type Ia glycogen storage disease is an autosomal recessive disorder caused by a deficiency of glucose-6-phosphatase (type Ia, von Gierke's disease) or glucose-6-phosphate translocase (type Ib). Its estimated incidence is 1 per 100,000. Since the pioneering description of Pizzo 30 in 1980, seven cases of severe pulmonary hypertension have been described in patients with type I glycogen storage disease 29, 30. In all but one case this condition was extremely severe with a fatal outcome within months of cardiopulmonary arrest and/or untractable congestive right heart failure. When available, post-mortem examination of the lungs demonstrated pulmonary hypertensive arteriopathy, as previously described in PPH. The only case occurring before the age of 10 was in a young girl with atrial septal defect of the secundum type 31. Interestingly, neither atrial septal defect nor type Ia glycogen storage disease typically lead to severe pulmonary hypertension in the first decade, suggesting that these conditions could have cumulative negative effects with a faster progression of the pulmonary vascular disease. The six remaining cases of severe PAH complicating the course of type I glycogen storage disease occurred in the second or third decade, suggesting that long-term metabolic abnormalities may contribute to the long latency of pulmonary vascular disease 29. In his initial report Pizzo 30 suggested that the occurrence of pulmonary arterial hypertension in type I glycogen storage disease could be due to an abnormal production of vasoconstrictive amines such as serotonin. The present authors recently investigated a case of type I glycogen storage disease with severe PAH in the absence of BMPR2 germline mutations, and a series of 13 patients with type I glycogen storage disease and no evidence of pulmonary vascular disease 29. Elevated plasma serotonin concentrations were found in patients with severe PAH and/or type I glycogen storage disease, as compared with controls. The present authors hypothesised that type I glycogen storage disease could represent another condition where abnormal handling of serotonin is one event in a multistep process leading to PAH. Interestingly, no 5HTT polymorphism putatively conferring susceptibility to PPH could be identified in patients with elevated plasma serotonin concentrations and type I glycogen storage disease (J-M Launay, Dept of Biochemistry, Hôpital Lariboisière, Assistance Publique, Hôpitaux de Paris, France, personal communication).

The future

Conclusion

The last couple of years have witnessed an outstanding number of major genetic discoveries allowing a better understanding of the pathophysiology of this devastating orphan disease. Much of this novel information emphasises the relevance of the transforming growth factor-β signalling pathway in pulmonary arterial hypertension 15. While this achievement has generated extreme interest, the pathobiology of severe pulmonary arterial hypertension remains unclear and genomic approaches to pulmonary hypertension research may identify additional contributory molecular pathways, implicating, for example, vasoactive agents, cytokines, chemokines and growth factors 3. Future research strategies should aim primarily at identifying novel therapeutic targets aimed at improving clinical care in this condition. Lastly, there is an urgent need to develop validated guidelines for genetic counselling to help families and pulmonary vascular specialists to cope with this difficult disorder.