Hemodynamic and Clinical Profiles of Pulmonary Arterial Hypertension Patients with GDF2 and BMPR2 Variants

Asians have a higher carrier rate of pulmonary arterial hypertension (PAH)-related genetic variants than Caucasians do. This study aimed to identify PAH-related genetic variants using whole exome sequencing (WES) in Asian idiopathic and heritable PAH cohorts. A WES library was constructed, and candidate variants were further validated by polymerase chain reaction and Sanger sequencing in the PAH cohort. In a total of 69 patients, the highest incidence of variants was found in the BMPR2, ATP13A3, and GDF2 genes. Regarding the BMPR2 gene variants, there were two nonsense variants (c.994C>T, p. Arg332*; c.1750C>T, p. Arg584*), one missense variant (c.1478C>T, p. Thr493Ile), and one novel in-frame deletion variant (c.877_888del, p. Leu293_Ser296del). Regarding the GDF2 variants, there was one likely pathogenic nonsense variant (c.259C>T, p. Gln87*) and two missense variants (c.1207G>A, p. Val403Ile; c.38T>C, p. Leu13Pro). The BMPR2 and GDF2 variant subgroups had worse hemodynamics. Moreover, the GDF2 variant patients were younger and had a significantly lower GDF2 value (135.6 ± 36.2 pg/mL, p = 0.002) in comparison to the value in the non-BMPR2/non-GDF2 mutant group (267.8 ± 185.8 pg/mL). The BMPR2 variant carriers had worse hemodynamics compared to the patients with the non-BMPR2/non-GDF2 mutant group. Moreover, there was a significantly lower GDF2 value in the GDF2 variant carriers compared to the control group. GDF2 may be a protective or corrected modifier in certain genetic backgrounds.


Introduction
Pulmonary hypertension (PH) is defined as a mean pulmonary artery pressure (mPAP) > 20 mmHg based on the 2022 European Society of Cardiology (ESC)/European Respiratory Society (ERS) guidelines [1,2].It can be divided into five groups based on different mechanisms [1,2].As PH progresses, right ventricular heart failure develops once the right ventricle fails to supply adequate cardiac output, and the death rate ranges from 4.5 to 12.3 per million population [3].There are similar sex and age distributions of PH patients in the Han population compared to Western population-based studies; however, a lower 1-year survival rate for Chinese PH patients was reported in a previous registry, which could be attributed to the unique genetic variants of Asian PH patients [4,5].
Around 25-30% of idiopathic pulmonary arterial hypertension (PAH) patients have an underlying genetic cause and could be classified as heritable PAH patients.The 6th World Symposium on Pulmonary Hypertension disclosed a list of genes that have different levels of connection to PAH [6].Among those PAH-related genes, heterozygous germline variants in the bone morphogenetic protein receptor type-2 gene (BMPR2) account for approximately 70-80% of heritable PAH and 10-20% of idiopathic PAH cases [7].The transforming growth factor-β (TGF-β) superfamily includes two major branches, including the TGF-β-activinnodal branch and the bone morphogenetic protein (BMP)-growth differentiation factor (GDF) branch [8].Loss or dysfunction in the balance between the TGF-β-activin-nodal branch and the BMP-GDF branch have played a critical role in predisposition and disease progression in PAH [8].
Several studies have indicated that Asian PAH has a higher carrier rate of PAHrelated genetic variants than Caucasian PAH [9][10][11].However, reports using whole exome sequencing (WES) for identifying genetic variants from Southeastern Asian idiopathic and heritable PAH cohorts are still rare.A better understanding of the mechanisms by which BMPR2 variants define a subclass of patients with more severe disease is critical to improve our knowledge of PAH.This study aimed to use WES to identify PAH-related genetic variants and investigate their hemodynamic and clinical profiles in a Taiwanese idiopathic and heritable PAH cohort.

Results
A total of 69 patients were enrolled in this study.A flowchart of the study is illustrated in Figure 1.The basic characteristics of the patients are reported in Table 1.There was a female predominance, and the average age at diagnosis was 50 ± 20 years.The average six-minute walking distance (6MWD) was 332 ± 127 m, mPAP was 41 ± 16 mmHg, pulmonary arterial wedge pressure (PAWP) was 14 ± 4 mmHg, and pulmonary vascular resistance (PVR) was 8 ± 7 Wood units.The peak tricuspid regurgitation peak gradient was 52 ± 30 mmHg, and the CI was 2.7 ± 1.1 L/min/m 2 .Pulmonary artery saturation was 66 ± 12%.The N-terminal prohormone of BNP (NT-proBNP) was 1869 ± 2988 ng/L.Most patients were in WHO Fc III (63.8%), with 49.3% patients presenting slow progression and 13% patients reporting a rapid progression of symptoms.
Around 25-30% of idiopathic pulmonary arterial hypertension (PAH) patients have an underlying genetic cause and could be classified as heritable PAH patients.The 6th World Symposium on Pulmonary Hypertension disclosed a list of genes that have different levels of connection to PAH [6].Among those PAH-related genes, heterozygous germline variants in the bone morphogenetic protein receptor type-2 gene (BMPR2) account for approximately 70-80% of heritable PAH and 10-20% of idiopathic PAH cases [7].The transforming growth factor-β (TGF-β) superfamily includes two major branches including the TGF-β-activin-nodal branch and the bone morphogenetic protein (BMP)growth differentiation factor (GDF) branch [8].Loss or dysfunction in the balance between the TGF-β-activin-nodal branch and the BMP-GDF branch have played a critical role in predisposition and disease progression in PAH [8].
Several studies have indicated that Asian PAH has a higher carrier rate of PAH-related genetic variants than Caucasian PAH [9][10][11].However, reports using whole exome sequencing (WES) for identifying genetic variants from Southeastern Asian idiopathic and heritable PAH cohorts are still rare.A better understanding of the mechanisms by which BMPR2 variants define a subclass of patients with more severe disease is critical to improve our knowledge of PAH.This study aimed to use WES to identify PAH-related genetic variants and investigate their hemodynamic and clinical profiles in a Taiwanese idiopathic and heritable PAH cohort.

Results
A total of 69 patients were enrolled in this study.A flowchart of the study is illustrated in Figure 1.The basic characteristics of the patients are reported in Table 1.There was a female predominance, and the average age at diagnosis was 50 ± 20 years.The average six-minute walking distance (6MWD) was 332 ± 127 m, mPAP was 41 ± 16 mmHg pulmonary arterial wedge pressure (PAWP) was 14 ± 4 mmHg, and pulmonary vascular resistance (PVR) was 8 ± 7 Wood units.The peak tricuspid regurgitation peak gradient was 52 ± 30 mmHg, and the CI was 2.7 ± 1.1 L/min/m 2 .Pulmonary artery saturation was 66 ± 12%.The N-terminal prohormone of BNP (NT-proBNP) was 1869 ± 2988 ng/L.Most patients were in WHO Fc III (63.8%), with 49.3% patients presenting slow progression and 13% patients reporting a rapid progression of symptoms.Of the 69 idiopathic PAH patients, 36 distinct rare variants of 20 genes from 28 patients (28/69, 40.6%;EIF2AK4 was not counted due to the obscured heterozygous contribution of EIF2AK4 in PAH) were identified (Figure 1).Six rare variants were classified as likely pathogenic (LP) or pathogenic (P) according to the ACMG 2015 guidelines, and the remaining thirty variants were classified as variants of uncertain significance (VUS).Among all the candidate genes, BMPR2, ATP13A3, and GDF2 had the highest incidence of variants (Table 2).Details of the VUS of the EIF2AK4 gene and other likely benign variants are listed in Table 2 and Supplemental Table S1, respectively.In addition, a pie chart (Figure 2) was illustrated to explain the distribution of genetic variants in the idiopathic PAH cohort.We found four BMPR2 variants in five female PAH patients (5/69, 7.2%) in this study (Table 2).There were two nonsense variants (c.994C>T, p. Arg332*; c.1750C>T, p. Arg584*) and one missense variant (c.1478C>T, p.Thr493Ile) that has been reported previously [14][15][16], and there was also one novel in-frame deletion variant (c.877_888del, p.Leu293_Ser296del).However, we did not find any gross insertion, deletion, or copy number changes (CNVs) on BMPR2.Among the study subjects, two patients (A100593 and A100719) were mother and daughter, and they had family aggregates of heritable  We found four BMPR2 variants in five female PAH patients (5/69, 7.2%) in this study (Table 2).There were two nonsense variants (c.994C>T, p. Arg332*; c.1750C>T, p. Arg584*) and one missense variant (c.1478C>T, p.Thr493Ile) that has been reported previously [14][15][16], and there was also one novel in-frame deletion variant (c.877_888del, p.Leu293_Ser296del).However, we did not find any gross insertion, deletion, or copy number changes (CNVs) on BMPR2.Among the study subjects, two patients (A100593 and A100719) were mother and daughter, and they had family aggregates of heritable PAH carrying a BMPR2 variant (c.1750C>T, p. Arg584*) and an ATP13A3 splicing variant (c.970+1 G>A).The onset time of symptoms was 58 years for the mother and 39 years for the daughter.The mPAP values at diagnosis were 56 and 60 mmHg for the mother and daughter, respectively, and their PVR values at diagnosis were 35.0 and 10.2 Wood units, respectively.Both of them were at the intermediate risk level based on the 2022 ESC/ERS guidelines [2].According to ACMG 2015 guidelines, all four BMPR2 variants were classified as likely pathogenic or pathogenic, and the ATP13A3 variant was classified as likely pathogenic.There were three patients carrying novel GDF2 variants in the study cohort (3/69, 4.3%).One rare GDF2 variant (c.259C>T, p. Gln87*) was a nonsense-type variant and classified as likely pathogenic (LP).Another two rare GDF2 variants (c.1207G>A, p. Val403Ile; c.38T>C, p. Leu13Pro) belonged to the missense type and were classified as VUS.The allelic frequencies of all three variants were absent in the normal East Asian population database (gnomAD).
Our function analysis was established by reprogrammed peripheral blood mononuclear cells of one of the GDF2 variant carriers, and we differentiated them into endothelial cells to carry out the migration and angiogenesis functional assay.Stronger migration (Figure 4A) and angiogenesis functions (Figure 4B) were observed.

Discussion
PAH patients with different genetic factors present a variety of clinical manifestations and outcomes.In the present study, the average age at diagnosis was 50 ± 20 years, with a female predominance.Compared with the patients in the N-gene subgroup, the BMPR2 variant carriers had higher mPAP, higher PVR, worse CI, and worse pulmonary artery saturation.Moreover, younger ages, higher mPAP values, and higher PVR values have been reported in GDF2 variant carriers.

Clinical Manifestations of PAH Patients with BMPR2 Variants
It has been established that genetic variation in the transforming growth factor β (TGF-β) signaling pathway plays an important role in the pathogenesis of idiopathic and heritable PAH, such as BMPR2, activin receptor-like kinase 1 (ACVRL1), and endoglin (ENG) [6, [17][18][19].Abnormalities in BMPR2 result in the aberrant activation of the TGF-β signaling pathway and the regulation of cell growth or apoptosis of pulmonary artery vascular smooth muscle cells [20].Patients bearing BMPR2 genetic variants are usually younger and have higher mPAP or PVR at diagnosis [7,21,22].The present study reported five female patients with BMPR2 variants.Although male patients have been reported to have a higher BMPR2 mutation rate compared to female patients [23], the female patients had higher penetrance than the males, with 42% and 14%, respectively [4,24,25].Moreover, it has been shown that being female is the single most definitive modifier for the penetrance of BMPR2 mutations in PAH [6,25].The present study reported 5 females with BMPR2 variants (Supplemental Table S2), which could be explained by the superior penetrance of BMPR2 mutations in female patients compared to male patients.Furthermore, genotype-phenotype correlation analyses have demonstrated that BMPR2 variant patients were younger at diagnosis (27.2 vs. 31.6years, p = 0.0003) compared with those without BMPR2 mutations [26].However, the present study reported that a younger age was insignificant in the BMPR2 mutation subgroup in comparison to the N-gene subgroup (43 ± 11 vs. 51 ± 20 years, p = 0.626, Table 3).This result was limited by the small sample size, comprising only five patients.Furthermore, BMPR2 variant carriers had a significantly shorter time to lung transplantation (approximately 10 years earlier than non-carriers) [7,9,11,22].This study demonstrated that BMPR2 variant carriers had a higher mPAP, higher PVR, worse CI, and worse pulmonary artery saturation (Table 1), which is consistent with the poor outcomes of patients with BMPR2 mutations in previous studies.

Potential Mechanisms of the GDF2 Variant and PAH
GDF2 (BMP9) is a ligand of the BMPR2 receptor.There are rare coding variants of GDF2, which occurred in 6.7% of cases in this study and has been implicated as a causative gene for the development of PAH [10].The first case report of a 5-year-old boy with homozygous GDF2 nonsense variants had severe PAH and right heart failure at 3 years of age [27].More studies have indicated that the involvement of GDF2 in PAH is facilitated through autosomal dominant behavior because there are heterozygous variants in GDF2 [10,19,[28][29][30].In addition to the BMPR2 and GDF2 variants, ACVRL1 and ENG variants also encode receptors belonging to the TGF-β superfamily and affect vascular proliferation in a manner similar to BMPR2 [19,31,32].Furthermore, more downstream genes of the TGF-β/BMP pathway, associated with the BMP axis, have been reported, such as SMAD1/5/8, SMAD4, and BMPR1B [33].
Unlike other GDFs, the GDF2 ligand binds to the endothelium-specific BMPR2/activin receptor-like kinase 1 (ALK1) heterotetrametric complex, activates intracellular phosphorylation relay through the BMP-specific SMAD1/5/8 signaling route to ensure the antiproliferative ability, and downregulates the proliferative effect via the activin receptor type IIA (ActRIIA)-SMAD2/3 pathway [34,35].In patients with the GDF2 variant, a decreased expression of the GDF2 ligand might prompt GDF2 to bind to ActR-IIA/IIB instead of BMPR2, which drives a similar impact to the BMPR2 variant on endothelial cell proliferation [36].Sotatercept is a novel drug for PAH therapy that has the ability to sequester excess ActRIIA ligands and thereby rebalance between two counterbalancing signaling pathways [34].Also, an in vivo study demonstrated that GDF2 suppressed endothelial cell proliferation in healthy control subjects but increased it in PAH patients with BMPR2 variants, which supported evidence of switching to the ActR-IIA/IIB rather than the BMPR2 downstream signaling pathway in PAH [36].
A previous study demonstrated that the BMPR2 variant can lead to endothelial dysfunction and PAH in the absence of protective modifiers [37].For example, the overexpression of endoglin increases the GDF2 response, which is opposite to the result of silencing BMPR2 or ActRIIA expression [35].Moreover, another study indicated that daily GDF2 administration had no significant effect on mice bearing a heterozygous endothelial BMPR2 deletion (BMPR2 EC+/− ) but led to excessive angiogenesis in BMPR2 EC−/− mice, which implied that the impact of the GDF2 variant on PAH was associated with the BMPR2 variant [36].The present study demonstrated an extremely low serum GDF2 (BMP9) concentration in a GDF2 nonsense variant (c.259C>T, p.Gln87*), which supports the mechanism of PAH caused by the loss of normal signaling via the BMP9-BMPR2-SMAD1/5/8 signaling route.However, whether GDF2 is a protective or corrected modifier in certain genetic backgrounds [36,[38][39][40] or whether the GDF2 variant leads to PAH by overexpressing the ActR-IIA/IIB-SMAD2/3 signaling route warrants further investigation [36].
All three GDF2 variant carriers in the present study were male.Regarding the GDF2 mutation, a investigation into gender disparity was obscured.There were similar hemodynamic appearances between the male GDF2 variant carriers and general PAH cohort in our subanalysis (Supplemental Table S3), but the influence of younger age at onset in the GDF2 variant carriers was absent due to our small sample size.However, gender's contribution in GDF2 mutation-associated PAH requires further investigation.The subanalysis of the present study illustrated that gender's contribution was not remarkable in both BMPR2 and GDF2 variant carriers (Supplemental Tables S2 and S3).
With regard to the interaction between GDF2 (BMP9) and BMP10, circulating BMP9 was encoded by the GDF2 gene [41], and one study indicated that despite there being no differences in plasma BMP10 levels in male PAH patients compared to a control group, the plasma BMP10 levels in female PAH patients were significantly lower than the control females [42].In our study, BMP10 had a counterbalancing-like elevated level compared to the reduced GDF2 (BMP9) level (Figure 3B).Whether there was counterbalancing protein interaction between the two productions of GDF2 encoding and whether the character was contributed to a certain GDF2 variant or the whole picture of the GDF2 mutation warrants further investigation.

Genetic Variants Other Than BMPR2 and GDF2 in PAH Patients
Due to inadequate knowledge of the entire PAH gene background in Asian populations, the present study did not include patients who were not specifically targeted for "screened variants" as a control group.Instead, the present study took patients without BMPR2 and GDF2 (N-gene subgroup) variants as a more convincing control group.
In addition to genes directly affecting the TGF-β/BMP pathway, this study reported four NOTCH variant carriers.NOTCH receptor signaling controls smooth muscle cell proliferation and maintains smooth muscle cells in an undifferentiated state, which has been reported to be associated with the pathogenesis of PAH and is involved in pulmonary vascular morphogenesis [43].NOTCH3 KO mice demonstrated a protective effect against hypoxia-induced PAH, and a γ-secretase inhibitor that inhibits the NOTCH pathway could be applied in the treatment of PAH in a mouse model [44].However, only a few genetic variants of the NOTCH pathway in PAH patients have been reported [45,46].
ATP13A3 was also identified in the present study.ATP13A3 encodes a P-type ATPase, a cation channel transporter, and has been reported to be associated with PAH [6, 19,47].The monoallelic ATP13A3 variant has been associated with adult-onset PAH [2,4,5,20], and Machado et al. demonstrated that biallelic ATP13A3 variants in childhood-onset PAH are characterized by extreme morbidity and mortality [48].The PAH patients with or without ATP13A3 genetic variants are reported in Supplemental Table S4, which illustrates that there were higher mPAP values (52 ± 7 vs. 38 ± 13 mmHg, p = 0.043) in the ATP13A3 variant subgroup compared to the non-ATP13A3 variant carriers.In addition, two out of the four ATP13A3 variant carriers also carried a nonsense mutation in the BMPR2 gene (A100593 and A100719), which demonstrated the speculative contribution of the two ATP13A3 variants to PAH instead of the observed effect despite the small sample size.The two patients (A100593 and A100719) in the present study were mother and daughter, and they had family aggregates of heritable PAH, carrying a BMPR2 variant (c.1750C>T, p. Arg584*) and an ATP13A3 splicing variant (c.970+1 G>A).In this case, heterozygous variants in ATP13A3 could act as a second hit for the development of PAH; thus, they require further investigation.Due to the study design adopting the process of "retrospective enrollment" and "prospective screening", the two subjects were diagnosed as idiopathic PAH at initial diagnosis.However, we realized they should have been classified as heritable PAH after a genetic investigation.
The detection of biallelic pathogenic EIF2AK4 mutations is able to make precise and accurate molecular diagnoses of autosomal recessively inherited pulmonary capillary hemangiomatosis (PCH) and pulmonary veno-occlusive disease (PVOD), and this was documented in the 6th World Symposium on Pulmonary Hypertension [6].A carrier of biallelic EIF2AK4 counts for up to 25% of patients with sporadic PVOD or PCH [6].However, EIF2AK4 can also play a "second hit" role, contributing to autosomal dominantly inherited hereditary PAH [49].The present study separated the variants into VUS and likely benign variants to facilitate further investigations on the heterozygous contribution between EIF2AK4 and PAH in the future.Furthermore, the PAH patients with or without EIF2AK4 genetic variants are reported in Supplemental Table S5.However, there was no significant difference between the EIF2AK4 variant and control subgroups.
Higher carrier rates of some PAH-related genes in Asian populations compared to Caucasian populations have been reported.In this study, BMPR2 variants accounted for around 15% in idiopathic PAH and more than 50% in heritable PAH.The present study reports five patients with BMPR2 variants (7.2%).Moreover, the GDF2 variants are relatively common (6.7%) in Chinese PAH patients [10], and three patients bearing GDF2 variants (4.3%) were reported in this study.Furthermore, the PTGIS gene was reported as a causative gene specifically in Chinese patients in [50], and four PAH patients (5.8%) in this study had variants of this gene.The findings regarding variants that were obtained in the present study align with those of previous publications.However, the carrier rates are inconsistent with previous studies due to the small sample size of this study.

Limitations and Strengths
Previous studies have investigated gene variants in Asian PAH patients, including in Chinese pediatric cohorts and Korean cohorts [9,11].A recent study in central Taiwan identified 14 potential PAH-related genetic variants by using WES analysis to study 45 PAH patients [51].However, these studies lacked data on GDF2 variants, and there is no functional assay to validate the contribution of GDF2 variants in Asian cohorts.In BMPR2 has been relatively well studied compared to other gene variants, and a key novel aspect of this study was the fact that it placed more focus on GDF2 and compared the clinical appearances of GDF2 and BMPR2 variants and carried out the migration and angiogenesis functional assay (Figure 4).
The purpose of this study was to investigate PAH-related genetic variants in Asian idiopathic and heritable PAH cohorts using sequencing and filtering strategies such as the two-tiered approach.The use of exploratory data for this study's Asian cohort was limited by the inadequate genetic database for the PAH cohorts.In spite of this limitation, the present study could still provide insights into the puzzle of how ethnic differences and genetic background affect PAH.In addition, the threshold of 0.0001 in all control/healthy individuals with no higher allele frequency in the subgroups is a more conservative threshold and in line with PAH incidence.However, the greater the number of novel variants with lower minor allele frequency that are present in a sample population, the more difficult it will be to home in on the causal gene [52].Furthermore, A MAF cutoff of 0.001 has previously been recommended for filtering variants responsible for dominant Mendelian disorders [52].Due to the inadequate sample size of the present study and the limited genetic background data in the gnomAD database and Taiwan Biobank, a minor allele frequency of 0.001 was used to define variants as very rare variants, and a cutoff of 0.001 was adopted in the present study [53].
In spite of its small sample size, this study used WES to identify PAH-related genetic variants and to investigate their hemodynamics, clinical profiles, and functional assay results in Asian idiopathic and heritable PAH cohorts.Novel genetic variants were detected in an Asian database of PAH cohorts.Further validation is required to confirm the pathogenic variants likely associated with PAH.

Materials and Methods
The cohort of idiopathic PAH patients was "retrospectively enrolled", with the earliest diagnostic date of this cohort being 4th March 2001 and the most recent diagnostic date being 28 March 2022.However, for our investigation into genetic background, patients were "prospectively screened", starting from 1 August 2021 to 31 December 2023.In addition, the demographics, clinical presentations, hemodynamics, and multi-parameters of risk assessment of patients diagnosed with PAH were documented for analysis.However, the lack of available survival data represents a limitation of this study.The Institutional Review Board of Kaohsiung Veterans General Hospital approved this study (KSVGH21-CT8-11).Written informed consent was obtained from all the participants.

Serum GDF2 (BMP9) and BMP10 of PAH Patients
Serum GDF2 concentrations were measured using the Human BMP-9 DuoSet ® RRELISA kits (R&D Systems, Inc., Catalog number: DY3209, Minneapolis, MN, USA) according to the manufacturer's instructions.GDF2 levels were analyzed in patients with BMPR2 and GDF2 or non-BMPR2 and non-GDF2 variants.The concentrations of BMP10 in the culture medium or patient serum was measured using enzyme-linked immunosorbent assay kits (DuoSet R&D Systems, Catalog number: DY2926-05) according to the manufacturer's instructions.

Hemodynamics and Cardiopulmonary Function Tests
Pulmonary hypertension was defined as a mean pulmonary arterial pressure >20 mmHg at rest, as assessed by right heart catheterization according to international guidelines [2].The World Health Organization (WHO) functional class (Fc), 6MWD, peak oxygen consumption, ventilatory equivalents for carbon dioxide (VE/VCO2), NT-proBNP, heart rate, right atrial pressure, cardiac output (CO), cardiac index (CI), pulmonary artery saturation, pressure, mPAP, PAWP, PVR, and peak tricuspid regurgitation peak gradient were analyzed using hemodynamic and cardiopulmonary function tests.

Validation of Variants by Polymerase Chain Reaction and Sanger Sequencing
The candidate variants were validated by polymerase chain reaction (PCR) and Sanger sequencing.Briefly, specific PCR primers were designed manually, and PCR was conducted using PCRBIO Ultra Polymerase (PCR Biosystems, London, UK) and SensoQuest Labcycler (SensoQuest, Göttingen, Germany).Purified PCR products were sequenced using an ABI PRISM Terminator Cycle Sequencing kit v3.1 on an ABI 3730 DNA sequencer (Applied Biosystems, Bedford, MA, USA).Validation data for candidate variants, obtained using PCR and Sanger sequencing, are shown in Supplemental Figure S1.

Generation of PAH-Specific iPS Cell Lines and Assessment of Function Analysis
Seven individuals were recruited into this study, including three patients diagnosed with idiopathic PAH, one GDF2 variant carrier (A100537), and three healthy individuals with no history of cardiovascular disease (Ctrl).There was no significant bias in gender or age in the three subgroups.The primary cultures of peripheral blood mononuclear cells were established from all individuals and were used for reprogramming at passages 2-4.Sendai virus delivery of OCT4, KLF4, c-Myc, and SOX2 was used to generate 5-8 independent iPSC lines per individual.These iPSC cell lines were thoroughly characterized and shown to be fully reprogrammed to pluripotency, as evaluated by colony morphology, growth dynamics, sustained long-term passaging (>20 passages), alkaline phosphatase (AP) staining (Figure 5A), the expression of pluripotency-associated transcription factors (OCT4, SOX2, and NANOG) and surface markers (SSEA4 and TRA1-60) (Figure 5B), the silencing of Sendai virus transgenes (Figure 5C), in vitro pluripotent differentiation ability and the generation of teratomas comprising derivatives of the three main embryo germ layers (Figure 5D), and karyotype stability (Figure 5E).Moreover, sequence analysis of the GDF2 gene in the iGDF2, iPAH, and control-iPS cell clones confirmed the heterozygous GDF2 mutation in the iGDF2-iPS cell clone (Figure 5F).Three replicates with the same patient were established in each subgroup.The peripheral blood mononuclear cells of one of the GDF2 variant carriers were reprogrammed and differentiated into endothelial cells.Further migration and angiogenesis functional assays were established by using endothelial cells (Figure 4).A number of studies have shown that CD34+ circulating progenitor cells (CPCs) and several subsets of CD34+ cells (such as CD133+/KDR+) can participate in vascular repair and growth and may be associated with vascular endothelial function.Therefore, these cells may reflect vascular integrity and have been used as biomarkers of vascular repair.CD34+ CPCs are a diverse group of progenitors, consisting of both hematopoietic and non-hematopoietic CPCs, with CD133 and KDR often used as more definitive antigen markers for endothelial progenitor cells (EPCs) [28].Friedrich et al. demonstrated that CD34−/CD133+/VEGFR-2+ progenitor cells are precursors of CD34+/CD133+/VEGFR-2+ EPCs and functionally more potent than CD34+/CD133+/VEGFR-2+ EPCs with respect to homing and vascular repair [19].In this study, we have established a procedure for the direct differentiation of human iPSCs to mesoderm-lineage cells via a chemically defined monolayer differentiation protocol.As shown in Figure 6A, human iPSCs can be differentiated to endothelial progenitors by 3 days of culture in an optimized defined medium supplemented with GSK3 inhibitors and another 7 days in StemPro-34 medium.On day 10 of differentiation, the differentiated cells developed typical EPC appearance, with an initial spindle shape and expressed EPC markers (CD133 and KDR).They could form a typical cobblestone-like shape of endothelial cells and express endothelial markers, including CD31 and VE-cadherin (CD144), through to day 21 post-differentiation.On day 10 of induction differentiation, FACS analysis showed that around 95% and 75% of the cell population expressed the EPC makers CD133 and KDR, respectively, and 1% of cells expressed CD34, indicating that the differentiated cells on day 10 were apparently precursors of "classical" CD34+/133+ EPCs.On day 21 of differentiation, around 75% and 99% of the cells stained positive for the EC maker CD31 and CD144, respectively, and 49% of cells expressed CD34, indicating that the CD34−/133+/KDR+ cells on day 21 could give rise to an endothelial progeny (Figure 6B).No differences were observed in the differentiation efficiency of iPSCs from the control, iPAH, and iGDF2 individuals.The functional features of the endothelial cells from the control, iPAH, and iGDF2 individuals were also confirmed through the uptake of acetylated LDL and lectin binding by direct fluorescent staining (Figure 6C), indicating that the iPSC-ECs from all subjects showed similar EC cobblestone morphology, acetylated LDL uptake, and UEA-1 lectin binding.Panel (B): FACS analysis on D10 showed that the EPC maker of CD133 was expressed in 95% of the cell population, and KDR was expressed in 75% of the cell population.In addition, CD34 was expressed in 1% of the cells.Both markers indicated that the differentiated cells on D10 are apparently precursors of "classical" CD34+/133+ EPCs.On D21 of differentiation, 75% and 99% of the cells stained positive for the EC makers CD31 and CD144, respectively, and 49% of cells expressed CD34, indicating that the CD34−/133+/KDR+ cells on day 21 could give rise to an endothelial progeny.Panel (C): There were no differences in the differentiation efficiency of the iPSCs between the control, iPAH, and iGDF2 individuals.The functional features of the endothelial cells from the control, iPAH, and iGDF2 individuals were also confirmed through the uptake of acetylated LDL and lectin binding by direct fluorescent staining, indicating that iPSC-ECs from all subjects showed similar EC cobblestone morphology, acetylated LDL uptake, and UEA-1 lectin binding.

Statistical Analysis
In this study, all continuous variables are expressed as mean ± standard deviation (SD), and categorical variables are expressed as numbers and percentages.The Kolmogorov-Smirnov test was applied to check the normality of normal distributions.For clinical analysis, the 69 PAH patients were categorized into three groups: BMPR2 genetic variant carriers, GDF2 genetic variant carriers, and a N-gene subgroup.Continuous variables were compared using a one-way ANOVA with Tukey's post hoc test.Categorical variables were examined using the chi-square test and Fisher's exact test.The serum GDF2 with an initial spindle shape and expressed EPC markers (CD133 and KDR).On D21, the ECs with typical cobblestone-like shapes expressed endothelial markers, including CD31 and VE-cadherin (CD144).Panel (B): FACS analysis on D10 showed that the EPC maker of CD133 was expressed in 95% of the cell population, and KDR was expressed in 75% of the cell population.In addition, CD34 was expressed in 1% of the cells.Both markers indicated that the differentiated cells on D10 are apparently precursors of "classical" CD34+/133+ EPCs.On D21 of differentiation, 75% and 99% of the cells stained positive for the EC makers CD31 and CD144, respectively, and 49% of cells expressed CD34, indicating that the CD34−/133+/KDR+ cells on day 21 could give rise to an endothelial progeny.Panel (C): There were no differences in the differentiation efficiency of the iPSCs between the control, iPAH, and iGDF2 individuals.The functional features of the endothelial cells from the control, iPAH, and iGDF2 individuals were also confirmed through the uptake of acetylated LDL and lectin binding by direct fluorescent staining, indicating that iPSC-ECs from all subjects showed similar EC cobblestone morphology, acetylated LDL uptake, and UEA-1 lectin binding.

Statistical Analysis
In this study, all continuous variables are expressed as mean ± standard deviation (SD), and categorical variables are expressed as numbers and percentages.The Kolmogorov-Smirnov test was applied to check the normality of normal distributions.For clinical analysis, the 69 PAH patients were categorized into three groups: BMPR2 genetic variant carriers, GDF2 genetic variant carriers, and a N-gene subgroup.Continuous variables were compared using a one-way ANOVA with Tukey's post hoc test.Categorical variables were examined using the chi-square test and Fisher's exact test.The serum GDF2 concentrations between the subgroups of non-BMPR2/non-GDF2, BMPR2, and GDF2 were analyzed using the Kruskal-Wallis test.Statistical significance was set at a two-tailed p-value < 0.05.The SPSS 19 statistical software package (SPSS, Inc., Chicago, IL, USA) was used for all calculations.

Conclusions
This study investigated WES-based genetic variants in an Asian PAH cohort.The most common variants were BMPR2 (c.994C>T, p. Arg332*; c.1750C>T, p. Arg584*; c.1478C>T, p. Thr493Ile; c.877_888del, p. Leu293_Ser296del), ATP13A3, and GDF2 (c.259C>T, p. Gln87*; c.1207G>A, p. Val403Ile; and c.38T>C, p. Leu13Pro).Compared with the patients in the N-gene subgroup, the BMPR2 variant carriers had higher mPAP, higher PVR, worse CI, and worse pulmonary artery saturation.The GDF2 variant carriers in the PAH cohort had worse clinical and hemodynamic performances and were younger in age.Moreover, there was a significantly lower GDF2 value in the subgroup of GDF2 variants than in the non-mutant subgroup.Whether GDF2 is a protective or a corrective modifier in certain genetic backgrounds requires further investigation.

Institutional Review Board Statement:
The study was conducted in accordance with the Declaration of Helsinki, and approved by the Institutional Review Board of Kaohsiung Veterans General Hospital (protocol code KSVGH21-CT8-11, approved on 8 November 2022).
Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.

Figure 2 .
Figure 2. Pie chart of the distribution of the genetic variants of the idiopathic PAH cohort.

Figure 2 .
Figure 2. Pie chart of the distribution of the genetic variants of the idiopathic PAH cohort.

21 Figure 3 .
Figure 3. Serum GDF2 (BMP9) and BMP10 concentrations in subgroups of BMPR2 variant carriers, GDF2 variant carriers, and the N-gene subgroup.Values are mean +/− SD; n = number of patients in subgroup.Panel (A): insignificantly lower circulating levels of BMP9 were demonstrated in the BMPR2 variant carriers, and significantly lower BMP9 levels were reported for the GDF2 variant carriers compared to the GDF2 concentrations in the N-gene subgroup.Panel (B): serum BMP10 levels were significantly higher than in the GDF2 variant carriers compared to the BMPR2 variant carriers and the N-gene subgroup.(* means statistical significance).

Figure 3 .
Figure 3. Serum GDF2 (BMP9) and BMP10 concentrations in subgroups of BMPR2 variant carriers, GDF2 variant carriers, and the N-gene subgroup.Values are mean +/− SD; n = number of patients in subgroup.Panel (A): insignificantly lower circulating levels of BMP9 were demonstrated in the BMPR2 variant carriers, and significantly lower BMP9 levels were reported for the GDF2 variant carriers compared to the GDF2 concentrations in the N-gene subgroup.Panel (B): serum BMP10 levels were significantly higher than in the GDF2 variant carriers compared to the BMPR2 variant carriers and the N-gene subgroup.(* means statistical significance).

Figure 4 .
Figure 4. Function analysis was carried out by reprogramming the peripheral blood mononuclear cells of one of the GDF2 variant carriers and differentiat into endothelial cells to conduct the migration and angiogenesis functional assay.Stronger migration (A) and angiogenesis functions (B) were observed.

Figure 4 .
Figure 4. Function analysis was carried out by reprogramming the peripheral blood mononuclear cells of one of the GDF2 variant carriers and differentiating them into endothelial cells to conduct the migration and angiogenesis functional assay.Stronger migration (A) and angiogenesis functions (B) were observed.

Figure 5 .
Figure 5. Characterizations of cells fully reprogrammed to pluripotency in three iPSC cell lines.Panel (A): colony morphology, growth dynamics, sustained long-term passaging (>20 passages), alkaline phosphatase (AP) staining.Panel (B): expression of pluripotency-associated transcription factors (OCT4, SOX2, and NANOG) and surface markers (SSEA4 and TRA1-60).Panel (C-E): silencing of Sendai virus transgenes (C), in vitro pluripotent differentiation ability and generation of teratomas comprising derivatives of the three main embryo germ layers (D), and karyotype stability (E).Panel (F): sequence analysis of the GDF2 gene in the iGDF2, iPAH, and control-iPS cell clones confirmed the heterozygous GDF2 mutation in the iGDF2-iPS cell clone.

Figure 6 .
Figure 6.Generation and characterization of iPSC-derived patient-specific endothelial progenitor cells (EPCs) and endothelial cells (ECs).Panel (A): Human iPSCs differentiated into mesoderm on D3 by being cultured in an optimized defined medium supplemented with GSK3 inhibitors.It took another 7 days in StemPro-34 medium for the cells to develop a typical EPC appearance, along with an initial spindle shape and expressed EPC markers (CD133 and KDR).On D21, the ECs with typical cobblestone-like shapes expressed endothelial markers, including CD31 and VE-cadherin (CD144).Panel (B): FACS analysis on D10 showed that the EPC maker of CD133 was expressed in 95% of the cell population, and KDR was expressed in 75% of the cell population.In addition, CD34 was expressed in 1% of the cells.Both markers indicated that the differentiated cells on D10 are apparently precursors of "classical" CD34+/133+ EPCs.On D21 of differentiation, 75% and 99% of the cells stained positive for the EC makers CD31 and CD144, respectively, and 49% of cells expressed CD34, indicating that the CD34−/133+/KDR+ cells on day 21 could give rise to an endothelial progeny.Panel (C): There were no differences in the differentiation efficiency of the iPSCs between the control, iPAH, and iGDF2 individuals.The functional features of the endothelial cells from the control, iPAH, and iGDF2 individuals were also confirmed through the uptake of acetylated LDL and lectin binding by direct fluorescent staining, indicating that iPSC-ECs from all subjects showed similar EC cobblestone morphology, acetylated LDL uptake, and UEA-1 lectin binding.

Figure 6 .
Figure 6.Generation and characterization of iPSC-derived patient-specific endothelial progenitor cells (EPCs) and endothelial cells (ECs).Panel (A): Human iPSCs differentiated into mesoderm on D3 by being cultured in an optimized defined medium supplemented with GSK3 inhibitors.It took another 7 days in StemPro-34 medium for the cells to develop a typical EPC appearance, along with an initial spindle shape and expressed EPC markers (CD133 and KDR).On D21, the ECs with typical cobblestone-like shapes expressed endothelial markers, including CD31 and VE-cadherin (CD144).Panel (B): FACS analysis on D10 showed that the EPC maker of CD133 was expressed in 95% of the cell population, and KDR was expressed in 75% of the cell population.In addition, CD34 was expressed in 1% of the cells.Both markers indicated that the differentiated cells on D10 are apparently precursors of "classical" CD34+/133+ EPCs.On D21 of differentiation, 75% and 99% of the cells stained positive for the EC makers CD31 and CD144, respectively, and 49% of cells expressed CD34, indicating that the CD34−/133+/KDR+ cells on day 21 could give rise to an endothelial progeny.Panel (C): There were no differences in the differentiation efficiency of the iPSCs between the control, iPAH, and iGDF2 individuals.The functional features of the endothelial cells from the control, iPAH, and iGDF2 individuals were also confirmed through the uptake of acetylated LDL and lectin binding by direct fluorescent staining, indicating that iPSC-ECs from all subjects showed similar EC cobblestone morphology, acetylated LDL uptake, and UEA-1 lectin binding.
Author Contributions: M.-T.W. wrote the manuscript and interpreted the data.S.-K.C. analyzed and interpreted the data.W.-C.H., K.-P.W. and M.-T.W. recruited patients and critically reviewed the manuscript for important intellectual content.L.-W.C. and M.-T.W. designed the study.All authors have read and agreed to the published version of the manuscript.

Table 1 .
Basic characteristics of pulmonary arterial hypertension patients in Kaohsiung VeteransGeneral Hospital (n = 69) †.Data of continuous variables are expressed as mean ± SD; changes in categorical variables were analyzed by chi-square tests and are expressed as n (%).

Table 2 .
Details of genetic variants in PAH-related genes.

Table 3 .
Pulmonary arterial hypertension patients with or without the genetic variants bone morphogenetic protein receptor type 2 (BMPR2) and growth differentiation factor 2 (GDF2), as well as clinical and hemodynamic presentations at initial diagnosis (n = 69) †.