Functional Mutations in 5′UTR of the BMPR2 Gene Identified in Chinese Families with Pulmonary Arterial Hypertension

Abstract

Pulmonary arterial hypertension (PAH) is a progressive pulmonary vasculopathy with significant morbidity and mortality. Bone morphogenetic protein receptor type 2 (BMPR2) has been well recognized as the principal gene responsible for heritable and sporadic PAH. Four unrelated Chinese patients with PAH and their family members, both symptomatic and asymptomatic, were genetically evaluated by sequencing all exons and the flanking regions of BMPR2. Functionality of the aberrant mutations at the 5′ untranslated region (UTR) of BMPR2 in the families with PAH was determined by site mutation, transient transfection, and promoter-reporter assays. Four individual mutations in the BMPR2 gene were identified in the 4 families, respectively: 10-GGC repeats, 13-GGC repeats, 4-AGC repeats in 5–UTR, and a novel missense mutation in exon 7 (c.961C>T; p.Arg321X). Moreover, we demonstrated that (1) these 5′UTR mutations decreased the transcription of BMPR2 and (2) the GGC repeats and AGC repeats in BMPR2 5′UTR bore functional binding sites of EGR-1 and MYF5, respectively. This is the first report demonstrating the presence of functional BMPR2 5′UTR mutations in familial patients with PAH and further indicating that EGR-1 and MYF5 are potential targets for correcting these genetic abnormalities for PAH therapy.

Pulmonary arterial hypertension (PAH) is physiologically defined by a mean pulmonary arterial pressure (PAP) of 25 mmHg or greater at rest and is characterized by pathologic pulmonary vascular remodeling, including smooth muscle hypertrophy and intimal thickening. The prevalence of PAH is about 15 per million adults. ¹ The average age of onset is in the third decade of life, but there is wide variation within families, and childhood onset is also common. At least 6% of PAH cases have a recognized family history, in which the disease segregates as an autosomal dominant trait with incomplete penetrance and an estimated lifetime risk of 10%–20%.^2–5 Different subtypes of PAH are recognized, including idiopathic (IPAH) and heritable (HPAH) forms, as well as secondary PAH. HPAH is diagnosed in patients with proven germline mutations in genes associated with pulmonary hypertension as well as in familial cases with or without identified germline mutations. ⁶ The mutation of the bone morphogenetic protein type 2 receptor (BMPR2) gene has been well studied during the past 2 decades and recognized as one of the most prevalent genetic abnormalities in HPAH and IPAH. Approximately 25% of all patients with PAH have a BMPR2 mutation.^7–10

BMPR2 is a member of the transforming growth factor β (TGF-β)–receptor superfamily of the transmembrane serine/threonine kinase receptor, which is expressed at high levels in the normal pulmonary microvascular endothelial cells of humans and mice. Signaling transduction through BMPR2 is essential for embryonic development, pulmonary vascular cell growth and differentiation, angiogenesis, organogenesis, endothelial cells, and smooth muscle cell junction and interaction.^8–12 BMPR2-mediated signaling has been implicated in vascular instability. Patients with PAH who carry the BMPR2 mutation may have an increased risk of vascular dissection or rupture and subsequent hemoptysis due to this instability (PMID: 24194909). This finding may help to explain why BMPR2 mutation carriers are often young at diagnosis and at death, with more severely compromised hemodynamic status.

In this study, we comprehensively studied and analyzed the genetics of the 4 families with PAH in order to establish the relationship between PAH and BMPR2 in the Chinese population with PAH. Notably, it is the first report describing the missense mutation in exon 7 and novel mutations in the 5′ untranslated region (5′UTR) of the BMPR2 gene in Chinese patients from 4 unrelated families with IPAH.

METHODS

Study subjects

Subjects were chosen from patients hospitalized in the ward of the First Affiliated Hospital of Guangzhou Medical University. The study was approved by the local ethics committee. All subjects were diagnosed according to consensus standards.^13,14 Inclusion criteria for IPAH and HPAH included the following: (1) PAP > 25 mmHg, with a pulmonary capillary wedge pressure ≤15 mmHg measured by right heart catheterization at rest, and (2) the absence of other disorders known to cause pulmonary hypertension; for example, ventilation and perfusion lung scans were used to exclude pulmonary embolism, and contrast echocardiography and/or measurements of oxygen saturation during cardiac catheterization were used to exclude intracardiac shunting and echocardiography and cardiac catheterization to exclude left heart diseases. HPAH was diagnosed by more than 1 confirmed case in first- to third-degree relatives. ¹⁵

Mutation detection

Blood samples were collected in vacuum blood collection tubes containing ethylenediaminetetraacetic acid. Genomic DNA was isolated from 500 μL peripheral blood by the FlexlGene DNA kit (Qiagen), and all DNA samples were quantified by a NANODROP 2000c spectrophotometer (Thermo) and stored at −70°C until use. Genomic fragments were amplified by polymerase chain reaction (PCR). Primers were designed using the program PRIMER 3 (http://primer3.ut.ee/) to amplify the coding regions and the intron/exon boundaries of the BMPR2 gene. PCR was sequenced by Sangong Bioth (Shanghai).

Plasmid construction

The 5′-flanking region and promoter sequence of the human BMPR2 gene, located at chromosome 2q33, was identified from GeneBank (NM_001204.6). According to this sequence information, BMPR2 5′UTR normally contains 3 AGC (3-AGC) repeats and 12 GGC (12-GGC) repeats located at −919 to −927 and −928 to −963 upstream from the start codon, respectively. We constructed 4 promoter-reporter plasmids to determine whether these polymorphisms had effects on BMPR2 gene expression. The PCR products of 2850 bp amplified using genomic DNA from a healthy donor as template and 5′-GGTACCCTAGAAGAAGCAAGCAGAGCA-3′ (forward) and 5′-AAGCTTAAATGTCAAGATACCACA-3′ (reverse) as primers were ligated into pGL4-basic vector upstream from the firefly luciferase gene sequence (Promega) to create a BMPR2-pGL4-wild reporter gene (carrying 3-AGC repeats at −919 to −927 and 12-GGC repeats at −928 to −963 upstream from the start codon). The following 3 plasmid constructs were created in a similar manner using genomic DNA from patients with PAH as PCR templates: (1) BMPR2-pGL4-10GGC reporter gene carrying 3-AGC repeats at −919 to −927 and 10-GGC repeats at −928 to −957 upstream from the start codon, (2) BMPR2-pGL4-13GGC reporter gene carrying 3-AGC repeats at −919 to −927 and 13-GGC repeats at −928 to −966 upstream from the start codon, and (3) BMPR2-pGL4-4AGC reporter gene carrying 4-AGC repeats at −919 to −930 and 12-GGC repeats at −928 to −963 upstream from the start codon. We verified the promoter-reporter constructs by sequencing. The pGL4 basic vector containing no promoter element was used as a negative control.

Plasmid transfection and luciferase assay

HEK293T cells were cultured in Dulbecco's modified Eagle's medium (DMEM), supplemented with 10% fetal bovine serum (Gibco) at 37°C in 5% CO₂. An A549 cell line was plated onto 96-well plates at 1 × 10⁴/well. After 24 hours of culture, the cells were transfected transiently with 100 ng of the pGL4-BMPR2-wild or pGL4-BMPR2-mut vectors using Lipofectamine 3000 (Invitrogen), plus 2 ng of the phRL-SV40 internal control plasmid encoding renilla luciferase (Promega). After 48 hours of incubation, the activities of the reporter vector with firefly luciferase and the internal standard with renilla luciferase were quantified by a dual-luciferase reporter assay system (Promega). Each transfection was carried out in sextuplicate and repeated 3 times.

RNA interference

The small interference RNA (siRNA) targeted to EGR-1, MYF5, and nontargeting control siRNA was synthesized by GenePharma. HEK293T cells were cultured in DMEM, supplemented with 10% fetal bovine serum (Gibco) at 37°C and 5% CO₂. The cells were plated onto 96-well plates at 1 × 10⁴/well. After 24 hours of plating, cells were cotransfected with 100 ng of pGL4-BMPR2 plasmids and 10 nM EGR-1, MYF5, or control siRNA using Lipofectamine 3000 (Invitrogen). Luciferase activity was determined using the method described above. Each transfection was carried out in sextuplicate and repeated 3 times. The efficacy of knockdown was assessed by quantitative real-time (RT) PCR using EGR-1-specific primers, MYF5-specific primers, and control 18S primers.

Statistical analysis

Data were presented as mean ± standard error of the mean. Differences among groups were assessed using one-way ANOVA for unpaired samples and considered significant at P < 0.05. GraphPad Prism 5 was used for all analyses.

RESULTS

Clinical features and genetic analysis

PCR amplicons of whole exons and their flanking regions of BMPR2 from genomic DNA of patients with PAH and their family members were sequenced and analyzed. Clinical features and mutations of patients were summarized. Pedigree of the families with PAH is presented in Figure 1.

Figure 1.

Pedigree and genetic analysis of 4 families of pulmonary arterial hypertension (PAH). A, Pedigree of PAH families carrying BMPR2 genetic mutations. Filled symbols: affected individuals; open symbols: unaffected individuals; triangles: spontaneous abortion. Slashed symbols indicate that the individual is deceased; arrows indicate that the individual is the proband. B, DNA sequence analysis of 4 representative family members with PAH. Family 1: −933_-928delGGC; family 2: c.-930_-928dupGGC and c.-919_-917dupAGC; family 3: c.-919_-917dupAGC; family 4: c.961C>T (p.Arg321X).

Family 1. The proband was a 9-year-old boy who first presented with dyspnea on exertion at the age of 4. As his symptoms worsened, he sought medical attention in early 2015. Echocardiography demonstrated hypertrophy of both the right atrium and right ventricle. Significant tricuspid regurgitation was also observed. Right heart catheterization (RHC) was performed, and the main pulmonary arterial pressure (MPAP) was measured to be 93/53 (70) mmHg. Unfortunately, the patient died of irreversible right heart failure resulting from heritable PAH 5 months after being diagnosed definitively, despite oral ambrisentan 5 mg daily. The patient's mother had a history of abortion before giving birth to this child, and moreover, the mother's third fetus was also aborted spontaneously at the gestational age of 14 weeks after the mother gave birth to the patient. After analysis, no mutation was observed in the coding region, but a substitution mutation of 10-GGC repeats was found in the 5′UTR of BMPR2 at −928 to −957 upstream from the start codon. Interestingly, the same mutation was also found in his mother (obligate carrier) and grandmother.

Family 2. In family 2, the proband was a 17-year-old male (II-1) with progressive dyspnea on exertion without an obvious etiology and much lower exercise tolerance compared to children at the same age, for a decade. The patient was admitted to the hospital for severe hemoptysis. Angiography revealed obviously enlarged bilateral pulmonary arteries and arteriostenosis in the branches of the right upper pulmonary artery. RHC revealed right pulmonary arterial pressure (RPAP) to be 87/44 (61) mmHg, left pulmonary arterial pressure (LPAP) 87/45 (62) mmHg, and MPAP 91/42 (61) mmHg. We found the patient had a mutation, the 13-GGC repeat at −928 to −966 upstream from the start codon in 5′UTR of BMPR2. The same mutation was also found in his father and in his family members. As we went back to the published literature, a similar mutation was described by Limsuwan et al. ¹⁶ in a case of pulmonary hypertension associated with congenital heart disease (aPH/CHD) in children. Here, the proband's aunt (II-2) was found carrying the same mutation as well as a G-930A mutation; in other words, she also carried the 4-AGC repeats at −919 to −930 upstream from the start codon.

Family 3. The proband was a 16-year-old female who started to feel dyspnea and chest tightness without an obvious etiology after moderate level of physical activity 3 years ago. Chest CT imaging revealed changes consistent with pulmonary hypertension accompanied with distal pulmonary embolism, enlarged right atria, right ventricle with pericardial effusion, thinned thoracic aorta, and bilateral effusion. RHC was performed, and the MPAP was measured to be 74/44 (54) mmHg, and right ventricle pressure was 76/26 (42.67) mmHg. As we went back to the family history of family 3, we surprisingly found that the patient bore a mutation similar to that of the proband's aunt from family 2 (II-2). The mutation included 4-AGC repeats at −919 to −930 upstream from the start codon of BMPR2. Furthermore, the same mutation was also found in the patient's father and his family members.

Family 4. The proband was a 21-year-old female who was diagnosed with PAH at the age of 19 with dyspnea, chest pain, and mild hemoptysis. Echocardiography revealed an enlarged right atrium and right ventricle with hypertrophied free wall. RHC revealed RPAP to be 121/53 (77) mmHg, LPAP 121/32 (74) mmHg, and MPAP 104/51 (70) mmHg. The proband died of PAH in 2015. We identified a nonsense mutation at the nucleotide position 961 (R321X) in exon 7, which creates a stop codon and therefore a truncated protein (Fig. 1), a novel heterozygous exonic mutation.

The 10-GGC, 13-GGC, 4-AGC repeat polymorphisms in 5′UTR of BMPR2 altered the transcriptional activity of the gene

Luciferase reporter vectors that contain various numbers of GGC and AGC repeats of BMPR2 5′UTR were constructed to determine the effects of these polymorphisms on the transcriptional activity of BMPR2. HEK293T cells were transiently transfected with the vectors, and luciferase activities were measured thereafter. The results revealed that the vectors carrying either 10- or 13-GGC repeats had decreased transcriptional activity compared with that carrying the 12-GGC repeats, and the one carrying 4-AGC repeats showed lower transcriptional activity than that with 3-AGC repeats (Fig. 2).

Figure 2.

Transcriptional activity analysis of bone morphogenetic protein receptor type 2 (BMPR2) 5′ untranslated region. A, Luciferase expression of constructs containing 10-GGC repeats, 13-GGC repeats, or 12-GGC repeats (wild) in HEK-293T cells (13-GGC repeats vs. 12-GGC repeats, P = 0.002; 10-GGC repeats vs. 12-GGC repeats, P = 0.0013). B, Luciferase expression of constructs containing 4-AGC repeats or 3-AGC repeats (wild) in HEK293T cells (4-AGC repeats vs. 3-AGC repeats, P = 0.00015). The firefly luciferase activity of each construct was normalized against the internal control of renilla luciferase (mean ± SEM, n = 3; asterisk: P < 0.05).

Knockdown of EGR-1 and MYF5 decreased the transcription activity of BMPR2 5′UTR

We analyzed the GGC and AGC repeat regions of BMPR2 5′UTR using the MAT Inspector program (Genomatix) and discovered that these regions contain putative binding elements for transcription factors EGR-1 and MYF5. As indicated in Figure 3, the 12- and 13-GGC repeat domains contain 6 putative binding sites for EGR-1, while the 10-GGC repeat region contains 5. The 3- and 4-AGC repeat regions have 2 putative binding sites for MYF5.

Figure 3.

Identifying the transcription factor binding sites in 5′ untranslated region of bone morphogenetic protein receptor type 2 (BMPR2) using MAT Inspector (Genomatix). A, The 10-GGC repeats contain 5 putative binding sites for EGR-1. B, The 12-GGC repeats contain 6 putative binding sites for EGR-1. C, The 13-GGC repeats contain 6 putative binding sites for EGR-1. D, The 3-AGC repeats contain 2 putative binding sites for MYF5. E, The 4-AGC repeats contain 2 putative binding sites for MYF5.

Based on the above program analysis, we next examined whether the putative binding sites for EGR-1 and MYF5 in BMPR2 5′UTR were functional and whether the transcription factors EGR-1 and MYF5 regulate BMPR2 gene transcription. Specific siRNA against either EGR-1 or MYF5 was designed and synthesized, and quantitative RT-PCR revealed that the knockdown efficiency of EGR-1 (Fig. 4B) and MYF5 (Fig. 4D) was more than 70% on messenger RNA (mRNA) level. Luciferase reporter assay indicated that either EGR-1 or MYF5 knockdown remarkably decreased the transcriptional activity of BMPR2 5′UTR (Fig. 4A, 4C), suggesting the presence of functional EGR-1 and MYF5 binding sites in the BMPR2 5′UTR region. The 10-GGC, 13-GGC, and 4-AGC repeat mutations decreased BMPR2 transcriptional activity, likely by altering the binding capacity BMPR2 promoter to EGR-1 and MYF5.

Figure 4.

Small interference RNA (siRNA) knockdown of EGR-1 and MYF5 inhibited bone morphogenetic protein receptor type 2 (BMPR2) transcription activity. A, Luciferase expression of constructs containing 10-GGC repeats, 13-GGC repeats, or 12-GGC repeats (wild) in HEK-293T cells with control siRNA or EGR-1 siRNA (siEGR-1). C, Luciferase expression of constructs containing 4-AGC repeats or 3-AGC repeats (wild) in HEK293T cells with nontargeting control siRNA (NC) or MYF5 siRNA (siMYF5). B, D, Expression of EGR-1 and MYF5 mRNA was measured in HEK-293T cells by quantitative real-time polymerase chain reaction at 24 hours after transfection with siEGR-1, siMYF5, and NC (mean ± SEM, n = 3; asterisk: P < 0.05).

DISCUSSION

Previous literature reported that BMPR2 gene mutation in patients with PAH was mostly located in exons, introns, or intron/exon boundaries.^2,3,17–19 Mutations in 5′UTR are rarely reported. In a BMPR2 mutation study, Wang et al. ²⁰ reported a promoter mutation, G-669A, in Chinese patients with FPAH. Limsuwan et al. ¹⁶ demonstrated that 9 out of 30 child patients with PH/CHD harbor GGC and AGC repeats, as well as 9-bp duplication mutations. In this study, by genetic analyses of 4 unrelated Chinese families with PAH, we identified 4 novel mutations in the well-studied HPAH-related BMPR2 gene. Notably, as far as we know, the BMPR2 c.961C>T mutation in family 4, the GGC repeat variant in families 1 and 2, and the AGC repeat variant in family 3 are reported for the first time. Although the AGC mutation carrier in family 2 did not suffer from PAH, patients with PAH carrying AGC repeat mutations were also found in unrelated family 3, indicating that this type of mutation might be important to the BMPR2 gene regulation. Then, through in vitro luciferase reporter assay, we validated that the GGC repeat variant in the 5′UTR of BMPR2 is the disease-causing mutation in families 1, 2, and 3. In family 4, the R321X mutation (premature termination in exon 7) was predicted to encode a nonfunctional BMPR2 receptor, due to lack of the kinase domain region. The lack of kinase activity would be consistent with a disease model of haploinsufficiency. Alternatively, the prematurely terminated products could act as dominant negatives, which have been observed for endoglin and TβR-II.^21,22 At the cell surface, given that BMPR2 as a dimer is likely to be present, it might be possible that only 25% of such complexes are functional. ²³

The BMPR2 promoter and 5′UTR is highly rich in GC but is rare in TATA. ²⁴ Analyses of the sequence predicted that this region contains putative binding sites for a number of transcription factors such as EGR-1, MYF5, AP1, SP1, and so on. The EGR-1 binding sequence in the BMPR2 promoter is GGCG, which is involved in GGC repeat regain. The MYF5 binding sequence in the BMPR2 promoter is CAGC, which is involved in AGC repeat regain. Hu et al. ²⁵ reported a 2.5-fold increase in transcriptional activity after inclusion of a sequence that includes the EGR-1 and MYF5 binding sites.

EGR-1 is a zinc-finger transcription factor, which is expressed and activated on mitogenic stimulation. ²⁶ EGR-1 stimulates transcription of human TGF-β receptor type II through binding to proximal promoter sequences. ²⁷ MYF5, a basic helix-loop-helix transcription factor involved in myogenic differentiation and control of early somitogenesis, affects more than one cell type derived from this important embryological compartment.^28–30 Gaddipati et al. ³¹ demonstrated that EGR-1 is essential for transcriptional regulation of BMPR2 and acted as both a positive and negative transcriptional modulator. In our study, we found that 10-GGC and 13-GGC repeat polymorphism variants in the 5′UTR of BMPR2 decreased transcriptional activity and could downregulate the expression of BMPR2 compared with 12-GGC repeats. This finding might be caused by mutations of the EGR-1 binding sites located at this region. Specific knockdown of EGR-1 significantly abolished the enhanced transcription activity, suggesting that EGR-1 binding sites located at the GGC repeats are functional, while the patients with PAH harboring BMPR2 mutations at this position exhibited disrupted EGR-1 binding sites and subsequently abolished EGR-1-induced BMPR2 transcription activity. Similar relationships were also discovered and confirmed between functional MYF5 binding sites and the mutations occurring at the AGC repeats within BMPR2 5′UTR. However, the relative expression levels and activity of these two transcription factors in patients with PAH and their mechanistic actions to BMPR2 gene transcription still remain largely unknown and require further investigation.

We found that the patients with PAH with 5′UTR GGC repeat mutations in the BMPR2 gene exhibit more severe degrees of disease, with a significantly younger age at diagnosis, a shorter survival time from diagnosis to death, and poorer physical health since birth. For example, in family 1, the proband who first presented with dyspnea on exertion at the age of 4 was 9 when he died of PAH. We hypothesize that the role of EGR-1 and MYF5 in the regulation of BMPR2 may link developmental and gene-environment interaction. EGR-1 and MYF5 are expressed in and regulate the process of early development. BMP signaling is involved in the regulation of proliferation, differentiation, and apoptosis and plays an important role in early development.

It is well known that nonsense, missense, and splice-site mutations and small insertions or deletions in the genes can lead to the expression of defective protein products, which are often responsible for a hereditary disease. Overall, in this study, 4 unrelated Chinese families with PAH were analyzed clinically and genetically. We uncovered PAH-causing BMPR2 mutations in all of these families. Our data suggest that the occurrence of GGC repeat variants and AGC repeat variants at the 5′UTR may decrease the transcription activity of BMPR2, due to the disrupted functional binding sites of transcription factors EGR-1 and MYF5 in this region. Our data also underscore the importance of including exons encoding 5′UTR for BMPR2 mutation screening. Moreover, our results pose interesting questions for further follow-up, such as whether PAH can be caused by reduced levels of mRNA expression. Further specific studies documenting altered binding of BMPR2 promoter or 5′UTR to various transcription factors will also be required to fully answer these questions.