Identification of a low frequency missense mutation in MUC6 contributing to pulmonary artery hypertension by whole-exome sequencing.

Pulmonary arterial hypertension (PAH) represents a progressive disease characterized by abnormally high blood pressure in the pulmonary artery. Although mutations in the bone morphogenetic receptor 2 (BMPR2) are found in 80% of heritable, their low penetrance suggests that other unidentified genetic modifiers are required for this disease. In this report, whole-exome sequencing (WES) and a linkage analysis were performed on genomic DNA isolated from four affected relatives and one non-affected relative in two PAH families. By focusing on meaningful variants which were presented in the four affected family members, but not presented in the non-affected individual, 49 SNP and eight indel variants in 39 genes were identified as candidates. Further high-throughput multiplex genotyping and Sanger sequencing were carried out to confirm the putative causal mutations in 150 individuals (30 idiopathic PAH [IPAH] patients, 30 chronic thromboembolic pulmonary hypertension [CTEPH] patients, and 90 normal controls). A heterozygous and deleterious mutation in the gene MUC6 (p.Pro1716Ser) was confirmed in the IPAH group (20/30, 67%) and CTEPH group (1/30, 3.33%); no variant was detected in the 90 normal controls. MUC6, which is short for mucin 6, encodes high molecular weight glycoprotein produced by many epithelial tissues and forms an insoluble mucous barrier that protects the lumens. We re-confirmed this low frequency mutation with the 1000 Genomes database across all species; no population or frequency data of this allele were acquired. We also found that this mutation site was highly conserved in different species and predicted MUC6 has the protection function of the airway and pneumoangiogram based on genomic sequence data. The compound heterozygous MUC6 gene mutation (p.Pro1716Ser) suggests a novel disease mechanism leading to PAH.

Background

Pulmonary hypertension (PH) presents as a serious and life-threatening disease characterized by abnormally high blood pressure (hypertension) in the pulmonary circulation, usually leading to death with elevated pulmonary vascular resistance (PVR) and right ventricular failure. It is also very common in patients with collagen vascular disease, heart failure, chronic obstructive pulmonary disease (COPD), and sleep apnea. According to the most recent classification, PH was divided into five types, including arterial, venous, hypoxic, thromboembolic, and miscellaneous. PAH and chronic thromboembolic pulmonary hypertension (CTEPH) are two main types of PH, share the same features of vasoconstriction, smooth-muscle cell and endothelial-cell proliferation and thrombosis.

Even if they receive treatment early, patients with PH have a very poor prognosis and high mortality. In particular, idiopathic PAH (IPAH) is the most severe type of PH, which has an untreated median survival of 2–3 years from the time of diagnosis. However, the pathogenesis and hereditary natures of PH have not been fully defined.

Despite intensive efforts in human genome scanning during the last decade, to date, only a few genes have been reported to be associated with PAH, and they are strikingly present in the same genetic and functional pathways. For example, mutations in BMPR2 associate with at least 50–70% of the familial cases,[1,2] as well as approximately 15% of IPAH cases.[3,4] The mutations of the genes encoding activin-like kinase-type 1 (ALK1) [5] and endoglin (ENG), two other members of the transforming growth factor-beta superfamily, were also found among patients with hereditary hemorrhagic telangiectasia (HHT) and co-existent PAH. However, 25% of heritable PAH and 85% of IPAH cases cannot be accounted for by identified mutations in the known PAH genes.[6] Therefore, it is crucial to analyze genetic profiles of patients for candidate genes and identify novel genes causing this complex disease.

Recently, next-generation sequencing (NGS), particularly whole-exome sequencing (WES; also known as targeted exome capture) has become a highly efficient strategy to identify novel causative genes for complex disorders, especially sequencing and filtering multiple affected individuals within a family pedigree to identify the disease-causing mutations.[7] It provides a powerful approach to identify new susceptibility genes of Mendelian disorders.

In the present study, we used WES to screen the genomes of four affected relatives and one healthy relative from two PAH families with the goal of discovering genetic variants predicted to contribute to PAH. Exome sequence data from PAH families were further confirmed in sporadic patients. We identified MUC6 gene mutation (p.Pro1716Ser) co-segregated with PAH. The mutation was completely new, which is not found in the single nucleotide polymorphism (SNP) database, the 1000 Genomes Project, and matching normal controls. This mutation site is also highly conserved in different species and MUC6 may have the protection function of the respiratory tract epithelium and pneumoangiogram against infectious agents. These findings provide new insights into the molecular mechanism by which MUC6 acts as an important factor in PAH development. We therefore believe that WES can help to identify novel genetic variants in patients with PAH and can assist future efforts to select individualized therapeutic approaches of PAH.

Materials and methods

Individuals (patients)

The study was approved by the Ethics Committee of Beijing Hospital. The participants provided their written informed consent to participate in this study. PAH was diagnosed as an increase in mean pulmonary arterial pressure (mPAP) ≥ 25 mmHg at rest as assessed by right heart catheterization based on European Society of Cardiology guidelines.[8] Blood samples from all available family members, sporadic idiopathic pulmonary hypertension (IPH) patients, and CTEPH patients were generously provided by Dr. Yang from the Chaoyang Hospital in Beijing, PR China. Non-PAH control DNA samples were conserved in our laboratory from the Beijing Hospital (90 age- and sex-matched controls). The blood samples were obtained after receiving written informed consent from all research participants. Total genomic DNA was extracted from all samples with extraction kit (Biochain, Beijing, PR China) following the protocol provided by the manufacturer, then quantified with NanoDrop spectrophotometer (Thermo Fisher Scientific Inc., Waltham, MA, USA).

Whole-exome sequencing

The exome analysis was performed at Beijing Genomics Institution (BGI) in Shenzhen. The DNA (3 µg) of four affected relatives and one non-affected relative in the two PAH families was fragmented to approximately 200 bp following sonication for single- or paired-end library preparation. Exonic sequences were enriched via the NimbleGen (Madison, WI, USA) 2.1 M exon capture array, targeting 34 Mb of sequence from exons and flanking regions in nearly 20,000 known genes. Following exome capture, high-throughput sequencing was performed with the Genome Analyzer II platform (Illumina Inc., San Diego, CA, USA). The average depth of sequencing reads per base-pair (bp) was 70 × . Paired sequencing reads were aligned to the reference human genome (UCSC, hg19) by using Short Oligonucleotide Analysis Package (SOAP) [9]. Nucleotides in exons and two flanking regions (10 bp flanking each exon) were further analyzed.

Candidate variant identification

Once all variants were obtained, variant filtration was performed. The detected variants were filtered according to the following criteria: DNA variants present in the Single Nucleotide Polymorphism database v132 (dbSNP132), the 1000 Genomes Project database, Hapmap database, and YH database were all removed with a minor allele frequency ≥ 0.005; variants within intergenic, intronic, and UTR regions and synonymous mutations were excluded from further analyses. Finally, functionally deleterious variants which have an impact on protein function were focused in further study and predicted with protein prediction tools Gen2Phen.

Variant validation using the MassARRAY™ system and Sanger sequencing

Genotyping using genomic DNA was carried out with the MassARRAY™ System (Sequenom, Inc., San Diego, CA, USA), which was based on the matrix-assisted laser desorption ionization-time of flight mass spectrometry platform (MALDI-TOF MS). Specific primers for the respective polymorphic sites were designed for the PCR amplification. After purification of the amplification product, a single extension primer annealed to the amplification product and generated an allele-specific product. The allele-specific extension product was detected by mass spectrometry with distinct masses.[10] Approximately 10% of the SNP mutations found in WES could not be validated by the Sequenom’s MassARRAY Designer software because of its special structure, including high GC content, copy number variation (CNV), and repeat region. These SNPs were confirmed by Sanger sequencing according to the standard protocol. Sequencing reads for each sample were assembled and analyzed with Variant Reporter software. Genotype frequencies for all the SNPs were in Hardy–Weinberg equilibrium, both in controls and in patients with P values > 0.05. The percentage of successful genotyping was > 90%.

Statistical analyses

Statistical analyses were performed with SPSS software (version 18, SPSS Inc., Chicago, IL, USA). The statistical significance of associations between categorical variables was based upon the one-tailed Fisher’s exact test; a P value < 0.05 was considered to be significant.

Accession codes

The following GenBank reference sequence was used for multiple sequence alignment: MUC6, NM_005961.

Results

Clinical presentation of the two PAH families

We analyzed two PAH families of three generations including four affected members and seven non-affected members (Fig. 1). We sequenced the exomes of five diagnosed members from these two families: a female patient and her sister diagnosed at 31 and 35 years, respectively, her non-affected mother, a male patient diagnosed at 46 years, and his affected daughter diagnosed at 15 years. Patient AI-2 had died because of cardiovascular disease; patient AI-1 died at a young age with a self-reported history of angina, syncope, and breathlessness, which were the symptoms of PAH. Physical examination of all the other members revealed completely normal results.

Fig. 1.

Pedigree of two PAH families. Squares indicate male; circles indicate female; white figures represent normal individuals; black figures represent affected individuals with PAH. The two families included four PAH patients of both sexes and some potential carriers (AII-2, AII-3, BI-1, BII-2, AI-1, and AI-2). WES was performed in four patients and one non-affected individual indicated by arrows (AII-2, AII-3, BI-1, BII-2, and AI-3).

Mutation detection and analysis by whole-exome sequencing

We sequenced exomic regions of the patients AII-2, AII-3, BI-1, BII-2, and AI-3 and obtained 4.4–6.4 million bases of sequence per sample, achieving a mean depth of target region between 69.08 × and 77.20 × , respectively (Table 1). The mean percentage of covered bases was in the range of 99.08–99.19% from the five relatives, which is sufficient for analyzing the SNPs, short insertions, or deletions.

Table 1.

Total variants identified through the WES analysis in patients with PAH.

Members	A-I3 (control)	A-II2	A-II3	B-I1	B-II2
Target region (bp)	44,131,461	44,234,141	44,131,461	44,234,141	44,131,461
Reads mapped to genome	61,341,026	59,875,249	64,170,438	58,639,948	62,215,793
Mean depth of target region (X)	71.80	71.74	77.20	69.08	73.80
Coverage of target region (%)	99.19	99.08	99.12	99.13	99.11
Total SNPs (n)	94,598	92,141	93,742	93,123	94,400
Missense	12,022	12,110	12,283	11,829	12,088
Total number of indels	7085	6989	7050	6915	6928
Frameshift variants	245	233	228	241	233

The mean rates of nucleotide mismatch were 0.37%; more specifically, 93,600 SNPs and 6993 indels per sample were found. We focused only on variants on non-synonymous and splice-site variants, gains of stop codon, or frame-shift mutations, which are possibly responsible for the protein sequences and functions. We also excluded variants that reported in the Single Nucleotide Polymorphism database v132 (dbSNP132), the 1000 Genomes Project database, Hapmap database, and YH database.

After filtration of variants, 600 SNPs and 88 indels still remained. Subsequently, we compared the variants presented in the four affected relatives (AII-2, AII-3, BI-1, and BII-2; Fig. 1) but not in the non-affected family member (AI-3; Fig. 1). Of these, 39 genes with 57 common homozygous or heterozygous variants in the four affected patients were identified. Therefore, it was probable that these 57 variants were predicted to be the causal mutations for PAH.

In order to exclude mutations in the genes associated with PAH, we focused on the variant result of PAH-related genes BMPR2, ALK1, and ENG in the patients. No mutation was found in these genes, further suggesting that the PAH development requires other factors.

Variants detected in IPH and CTEPH patients

To verify the candidate mutations, the genomes of four affected relatives (AII-2, AII-3, BI-1, and BII-2; Fig. 1) and the non-affected family member (AI-3) were Sanger sequenced for all 57 variants. The mutations were homozygous or heterozygous in two or all four affected individuals and absent in the unaffected individual; no other mutation was found within these genes.

Finally, in order to investigate if these candidate variants are the causal mutations for PAH, we performed the MassARRAY system and Sanger sequencing assay before mutation analysis. We analyzed the 57 variants of interest in 30 IPH patients and 90 age- and sex-matched controls, who presented with normal PAP. By using the MassARRAY™ System and Sanger sequencing assay (mutations with special structures that could not be validated by the MassARRAY™ system were confirmed by Sanger sequencing), the G-to-A mutation at nucleotide 5146 (heterozygous mutation, p.Pro1716Ser) was present in the 30 IPH with a percentage of 67%. All the other 90 control individuals did not carry this mutation, which is located in the MUC6 gene (NM_005961).

Different clinical PH groups have different pathophysiological features.[11,12] In order to determine whether the MUC6 gene mutation (p.Pro1716Ser) is the causal mutation of CTEPH, we sequenced MUC6 gene in an additional 30 CTEPH and 90 age- and sex-matched normal controls. We therefore observed a total of 1/30 (3.33%) CTEPH with MUC6 variants, while no variant was detected in the normal samples (P < 0.001). The result indicated that MUC6 is rarely mutated in CTEPH patients.

Figure 2 indicates the heterozygous G.A transition in exon 31 of MUC6 gene in proband, sporadic IPH patients, CTEPH patients, and normal controls. This point mutation changes the Proline in position 1716 with a Serine. Table 2 shows basic information of IPH patients, CTEPH patients, and healthy controls used for validation.

Fig. 2.

Sequence analyses of the MUC6 mutations. Heterozygous mutation of the MUC6 gene (the green arrow indicates the mutation site) in proband, sporadic IPH patient, CTEPH patient, and normal control.

Table 2.

Basic information of IPH patients, CTEPH patients, and healthy controls used for validation.

	IPH (n = 30)	CTEPH (n = 30)	Controls (n = 90)
Age (mean ± SD)	44.03 ± 10.77	56.80 ± 8.96	51.00 ± 8.03
Gender
Male	8	19	40
Female	22	11	50
BMI (kg/m2)	24.02 ± 2.98	24.17 ± 2.79	23.95 ± 3.06
mPAP (mmHg)	50.11 ± 12.53	47.42 ± 12.94	–

SD, standard deviation; BMI, body mass index; mPAP, mean pulmonary arterial pressure at rest as assessed by right heart catheterization.

Importance of the MUC6 mutation

The 1000 Genomes Project describes the variants and genotypes of the samples from 26 different populations around the world. In order to reconfirm this rare mutation in all the population or some particular population, we selected the data from 1000 genomes and found that no population or frequency data of this allele was acquired. In the meantime, we summarized all the variation consequences of MUC6 gene (Table 3).

Table 3.

Summary of variation consequences of MUC6 gene.

Type	Number	Description
Downstream gene variant	1699	A sequence variant located 3′ of a gene
Upstream gene variant	1621	A sequence variant located 5′ of a gene
Intron variant	1102	A transcript variant occurring within an intron
3 prime UTR variant	23	A UTR variant of the 3′ UTR
5 prime UTR variant	2	A UTR variant of the 5′ UTR
Coding sequence variant	2	A sequence variant that changes the coding sequence
Synonymous variant	415	A sequence variant where there is no resulting change to the encoded amino acid
Splice region variant	44	A sequence variant in which a change has occurred within the region of the splice site, either within 1–3 bases of the exon or 3–8 bases of the intron
Missense variant	850	A sequence variant, that changes one or more bases, resulting in a different amino acid sequence but where the length is preserved
Inframe deletion	11	An inframe non-synonymous variant that deletes bases from the coding sequence
Inframe insertion	4	An inframe non-synonymous variant that inserts bases into in the coding sequence
Frameshift variant	51	A sequence variant which causes a disruption of the translational reading frame, because the number of nucleotides inserted or deleted is not a multiple of three
Stop gained	23	A sequence variant whereby at least one base of a codon is changed, resulting in a premature stop codon, leading to a shortened transcript
Splice donor variant	6	A splice variant that changes the 2 base region at the 5′ end of an intron
Splice acceptor variant	3	A splice variant that changes the 2 base region at the 3′ end of an intron
All	5812	All variations

We then compared the human sequence for interspecies homology and found that the mutation site of the MUC6 gene is highly conserved among different species (Fig. 3a). We also used the Atlas of Genetics and Cytogenetics in Oncology and Haematology database (http://atlasgeneticsoncology.org/index.html) to analyze the peptide structure of MUC6. The schematic diagram indicated that MUC6 contains D1, D2, D′, and D3 domains at the N-terminal region, the TR domain at the central region, the STP domain, and the CK domain at the C-terminal region. The MUC6 gene has a highly similar 5′ region with other mucin genes; however, it shows no similarity in the 3′ region which contains long stretches of tandem repeats that encode serine- and threonine-rich domains (Fig. 3b). Pro1716Ser is located in the tandem repeat domain, which exhibits a high level of polymorphism that can result in differences in the length of the protein backbone. Therefore, this polymorphism may affect the structural rearrangement of the VNTR domain that will lead to reduction of the protection of the respiratory tract epithelium against infectious agents.

Fig. 3.

Identification of MUC6 gene mutation. (a) Heterozygous mutation site of MUC6 gene is highly conserved in different species. (b) Schematic representation of MUC6 peptide structure.

Discussion

WES is widely performed to identify causal variants in rare heritable disorders. Accompanied by genetic linkage studies, it is usually used in large sick families with multiple affected and non-affected individuals to identify the genes that carry the disease-causing mutations. Thus, other populations including sporadic patients and normal individuals can be screened for mutations in those candidate genes. Finally, case-control studies can be performed when a gene was deemed to play an important role in the pathogenesis of the disease.[13]

In this study, we reported two PAH families and our WES result of novel causative gene identification in PAH. Family A lived in Beijing, which is a northern city in China. Family B lived in the southern Chinese provinces. We found that the compound heterozygous MUC6 gene mutation (p.Pro1716Ser) on chromosome 11p15 at position 1,017,655 (GRCh37/hg19) was significantly associated with PAH. The presence of MUC6 variants in PAH family members and nearly 70% of IPHs suggests that MUC6 plays an important role in the genesis of this disease in Chinese people. Furthermore, the increase in PVR among PAH patients can be explained by vasoconstriction, proliferative and obstructive remodeling of the pulmonary vessel wall, inflammation, and thrombosis. However, CTEPH is the consequence of mechanical obstruction of the pulmonary arteries, which is the most important pathophysiological process caused by non-resolution of acute thrombotic masses that later undergo fibrosis. Our study demonstrated the MUC6 gene mutation (p.Pro1716Ser) was not the causal mutation of CTEPH due to the different pathophysiological features.

The MUC6 gene is a member of the mucins family, which has a high molecular weight, with heavily glycosylated proteins (glycoconjugates) produced by epithelial tissues.[14] Members of the mucins family show very little sequence homology, but all of them contain long stretches of tandem repeats of sequence that encode serine- and threonine-rich domains. These repeated regions in MUC6 are composed of 507 bp and comprise 50% of the polypeptide.[15]

The human MUC6 gene is located at chromosome position 11p15, a recombination-rich region which also contains other mucin genes (MUC2, MUC5AC, and MUC5B). The MUC6 gene extends approximately 24 kb and the mRNA length is approximately 8 kb.[16] It encodes secretory-type gastric mucin, a high molecular weight secreted glycoprotein with a molecular mass of about 105 kDa, highly expressed in the gastric mucosa, duodenal Brunner’s glands, pancreas, gall bladder, endocervix, seminal vesicle, and the common bile duct. It plays an essential role in the cytoprotection of epithelial surfaces from acid, proteases, mechanical trauma, and pathogenic microorganisms in a variety of epithelial tissues, especially in the gastrointestinal tract. Airway mucins are major components of the soluble layer that protect the lungs from any potentially pathogens and environmental toxins.[17]

In the present study, the overexpression of MUC6 was an early event in the carcinogenesis of the pancreas, bile duct, and endocervix. Alterations in the expression of MUC6 were also correlated significantly with dedifferentiation of bronchial epithelium. Epigenetic modification in the promoter regions, the upstream transcription factors, and signaling pathways is involved in the regulation of expression of MUC6 gene.[18]

Recently, some studies have suggested mucins show a high degree of inherited genetic variability because of a variable number of tandem repeats. MUC1 gene polymorphism contributes to the individual susceptibility to gastric cancer development; some common variations in MUC5AC genes are involved in the development of stomach cancer.[19,20] The other researchers addressed the association of genetic variability in MUC2 and evolution of gastric cancer precursor lesions (GCPLs), especially in H. pylori-infected patients, suggesting an important role of the secreted mucin in the predisposition to gastric carcinogenesis.[21] MUC6 gene polymorphism contributes to an elevated risk in gastric cancer patients from northern Portugal.[15] Some data suggest that short MUC6 alleles are associated with H. pylori infection.[22] However, the biological significance of mucin gene polymorphism is less well understood. A previous study suggests the polymorphisms of these variable number tandem repeat of MUC6, MUC2, and MUC5AC are predicted to induce the difference in the length of the encoded mucin polypeptides.[23]

As we know, mucins are overproduced in the patients with airway obstruction, which greatly contribute to lung pathophysiology in asthma, COPD, or cystic fibrosis.[24-27] Based on the information, we can deduce a conclusion that the compound heterozygous mutations of MUC6 gene (p.Pro1716Ser) may impair the protective function of airways and pneumoangiograms. Finally, the decrease of pulmonary vasodilation can lead to PH.

Conclusions

In summary, in our current study we identified a PAH-related novel MUC6 gene mutation (p.Pro1716Ser) by WES. The major challenge for mutation analysis is to determine whether the variants are potentially causative factors for the disease. Since the role(s) of the candidate gene is not well understood in this paper, further studies are needed to elucidate the molecular mechanisms of the phenotype and genotype relationships in the coming years.