Introduction
Atrial fibrillation (AF), typically characteristic of rapid and chaotic electrical activity in the atria with subsequent desynchronized atrial contractions, constitutes the most common type of cardiac arrhythmia in the setting of clinical practice, accounting for approximately one-third of hospitalizations for irregular heart rhythms (1). The prevalence of AF is approximately 1% in the general population, and increases markedly with advancing age, from approximately 0.5% of individuals in their fifties to nearly 10% of the octogenarian population (2). According to the Framingham Heart Study, the lifetime risk for development of AF is projected to be 25% for persons over the age of 40 (3). Currently, 2.5 million Americans are suffering from AF, but with the aging population and improved cardiovascular survival, that number is expected to exceed 16 million by the year 2050 (4). AF is responsible for substantially increased cardiovascular morbidity and mortality. Patients with AF have a 5-fold increased risk of stroke, and it is estimated that 15–20% of all strokes are attributable to AF (5). The risk of cerebrovascular thromboembolism ascribed to AF also increases strikingly with advancing age, rising from 1.5% at 50–59 years of age to 23.5% at 80–89 years of age (5). AF independently increases the risk of congestive heart failure and increases the risk of mortality 2-fold (6). Additionally, AF may contribute to complications such as adverse hemodynamics, reduced exercise tolerance, degraded quality of life, impaired cognition or dementia, and tachycardia-induced cardiomyopathy (7). Therefore, AF has become an immense and growing public health burden. Only in the United States, the costs attributable to the care of individuals with nonvalvular AF are in excess of $6.4 billion/year (8). Despite the high prevalence and clinical significance, the molecular mechanism underlying AF remains poorly understood.
Traditionally, AF has been perceived as a condition that occurs in the context of atrial electrical and structural remodeling that can result from miscellaneous cardiac and systemic disorders, including hypertension, coronary artery disease, rheumatic heart disease, congenital heart disease, chronic pulmonary heart disease, cardiomyopathy, cardiac surgery, diabetes mellitus, hyperthyroidism and electrolyte imbalance (1). However, in 30–45% of AF patients, no obvious risk factors can be identified by routine medical examination, and such AF is termed ‘idiopathic’ or ‘lone’ (1).; up to 15% of these cases have a positive family history, and are thus defined as familial AF (9). Increasing evidence demonstrates the familial aggregation of AF and an enhanced vulnerability to AF in the close relatives of patients with AF, suggesting that genetic defects may be involved in the pathogenesis of AF in a subset of patients (10–16). Genome-wide genetic linkage analysis with highly polymorphic microsatellite markers mapped susceptibility loci for AF on human chromosomes 10q22, 6q14-16, 11p15.5, 5p13, 10p11-q21 and 5p15, of which AF-causing mutations in two genes, KCNQ1 on chromosome 11p15.5 and NUP155 on chromosome 5p13, were identified and functionally characterized (17–23). In addition, genetic screening of candidate genes revealed an increasing number of AF associated mutations in genes encoding potassium channel subunits (KCNH2, KCNA5, KCNJ2, KCNJ8 and KCNE1-5), sodium channel subunits (SCN5A and SCN1B-3B), signaling peptide (NPPA), gap junctions (GJA1 and GJA5), and others (24–44). Nevertheless, these causative mutations appear to be relatively rare causes of AF, and the genetic determinants for AF in an overwhelming majority of patients remain elusive.
Emerging evidence indicates that abnormal embryological development of the cardiovascular system, particularly the pulmonary venous myocardium, is a major anatomic substrate for AF (45). Developmental biology studies substantiate the key role for several transcription factors, including NKX2.5, GATA4, GATA5 and GATA6, in the normal cardiovascular morphogenesis (46–48), and multiple mutations in GATA4, GATA5 and GATA6 have been causally associated with AF (49–57). NKX2.5 is a member of the NK2-family of transcription factors and its expression and functions overlap with those of the GATA family during cardiovascular development, particularly in synergistic regulation of target gene expressions cooperatively with GATA4 (58), which justifies NKX2.5 as a prime candidate gene for AF.
To evaluate the prevalence of NKX2.5 mutations in patients with familial AF and to explore the mechanism by which mutated NKX2.5 results in or predisposes to AF, the coding exons and exon/intron boundaries of NKX2.5 were sequenced in patients with familial AF in contrast to control individuals, and the functional characteristics of the mutant NKX2.5 were assessed in comparison with its wild-type counterpart using a dual-luciferase reporter assay system.
Materials and methods
Study subjects
A cohort of 110 unrelated index patients with familial AF was recruited from the Han Chinese population in China. The available relatives of the probands were enrolled. A total of 200 ethnically-matched unrelated healthy individuals were enlisted as controls. Peripheral venous blood samples were prepared and clinical data including medical records, electrocardiogram and echocardiography reports were collected. The study subjects were clinically classified using a consistently applied set of definitions (9,57). Briefly, AF was diagnosed by a standard 12-lead electrocardiogram demonstrating no P-waves and irregular R-R intervals irrespective of clinical symptoms. Lone AF was defined as AF occurring in individuals <60 years of age without other cardiac or systemic diseases by physical examination, electrocardiogram, transthoracic echocardiogram, and extensive laboratory tests. Familial AF was referred to as the presence of documented lone AF in additional two or more first- or second-degree relatives. Relatives with AF occurring at any age in the setting of structural heart disease (hypertensive, ischemic, myocardial or valvular) were classified as ‘undetermined’ for having an inherited form of AF. The ‘undetermined’ classification was also used if documentation of AF on an electrocardiogram tracing was lacking in relatives with symptoms consistent with AF (palpitation, dyspnea and light-headedness), or if a screening electrocardiogram and echocardiogram were not performed, regardless of the symptoms. Relatives were classified as ‘unaffected’ if they were asymptomatic and had a normal electrocardiogram. Paroxysmal AF was defined as AF lasting >30 sec that terminated spontaneously. Persistent AF was defined as AF lasting more than seven days and requiring either pharmacologic therapy or electrical cardioversion for termination. AF that was refractory to cardioversion or that was allowed to continue was classified as permanent. The study protocol was reviewed and approved by the local institutional Ethics Committee and written informed consent was obtained from all participants.
Genetic screening
Genomic DNA from all participants was isolated from blood lymphocytes with the Wizard Genomic DNA Extraction Kit (Promega, Madison, WI, USA), according to the manufacturer’s instructions. Initially, the whole coding region and splice junctions of the NKX2.5 gene were sequenced in 110 unrelated index patients with familial AF. Subsequently, genotyping NKX2.5 in the available relatives of the mutation carrier and 200 ethnically-matched unrelated healthy individuals who served as controls, was performed. The referential genomic DNA sequence of NKX2.5 was derived from GenBank (accession no. NT_023133), a gene sequence database at the National Center for Biotechnical Information (NCBI; http://www.ncbi.nlm.nih.gov/). Using the online Primer 3 software (http://frodo.wi.mit.edu/), the primer pairs used to amplify the coding exons (exon 1–2) and intron-exon boundaries of NKX2.5 by polymerase chain reaction (PCR) were designed as follows: primer 1, forward, 5′-CAC GAT GCA GGG AAG CTG-3′ and reverse, 5′-AGT TTC TTG GGG ACG AAA GC-3′ (the PCR product was 477 base pairs in size); primer 2, forward, 5′-CCT CCA CGA GGA TCC CTT AC-3′ and reverse, 5′-CGA GTC CCC TAG GCA TGG-3′ (the product was 463 base pairs); primer 3, forward, 5′-AGA ACC GGC GCT ACA AGT G-3′ and reverse, 5′-GAG TCA GGG AGC TGT TGA GG-3′ (the product was 473 base pairs). The PCR was performed using HotStar TaqDNA Polymerase (Qiagen, Hilden, Germany) on a Veriti Thermal Cycler (Applied Biosystems, Foster, CA, USA) with standard conditions and concentrations of reagents. Amplified products were purified with the QIAquick Gel Extraction Kit (Qiagen). Both strands of each PCR product were sequenced with a BigDye® Terminator v3.1 Cycle Sequencing Kit on an ABI-PRISM 3130xl DNA Analyzer (both from Applied Biosystems). The sequencing primers were the same as mentioned above for the specific region amplifications. DNA sequences were analyzed with the DNA Sequencing Analysis Software v5.1 (Applied Biosystems). The variant was corroborated by re-sequencing of an independent PCR-generated amplicon from the subject and met the quality control threshold with a call rate exceeding 99%. For an identified sequence variant, the Exome Variant Server (EVS; http://evs.gs.washington.edu/EVS) and NCBI’s single nucleotide polymorphism (SNP; http://www.ncbi.nlm.nih.gov/SNP) online databases were queried to confirm its novelty.
Multiple alignments of NKX2.5 protein sequences across species
Conservation of the amino acid altered by missense mutation was appraised by aligning human NKX2.5 to chimpanzee, monkey, dog, cow, mouse, rat, chicken and zebrafish NKX2.5 using the HomoloGene and Show Multiple Alignment links on the NCBI’s website (http://www.ncbi.nlm.nih.gov/homologene).
Prediction of the disease-causing potential of an NKX2.5 sequence variation
The disease-causing potential of an NKX2.5 sequence variation was predicted by an online program, MutationTaster (http://www.mutationtaster.org), automatically giving a probability for an alteration to be either a pathogenic mutation or a benign polymorphism. Notably, the P-value given here is the probability of the prediction rather than the probability of error as used in t-test statistics, i.e., a value close to 1 indicates a high ‘security’ of the prediction.
Expression plasmids and site-directed mutagenesis
The recombinant expression plasmid NKX2.5-pEFSA and the atrial natriuretic factor (ANF)-luciferase reporter plasmid, which contains the 2600-bp 5′-flanking region of the ANF gene, i.e., ANF(−2600)-Luc, were kindly provided by Dr Ichiro Shiojima, from the Department of Cardiovascular Science and Medicine, Chiba University Graduate School of Medicine, Chuo-ku, Chiba, Japan. The identified mutation was introduced into the wild-type NKX2.5 using a QuickChange II XL Site-Directed Mutagenesis Kit (Stratagene) with a complementary pair of primers. The mutant was sequenced to confirm the desired mutation and to exclude any other sequence variations.
Luciferase reporter gene assays
COS-7 cells were cultured in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal calf serum. The ANF(−2600)-Luc reporter construct and an internal control reporter plasmid pGL4.75 (hRluc/CMV; Promega) were used in transient transfection assays to examine the transcriptional activation function of the NKX2.5 mutant. COS-7 cells were transfected with 0.4 μg of wild-type or mutant NKX2.5-pEFSA expression vector, 1.0 μg of ANF(−2600)-Luc reporter construct, and 0.04 μg of pGL4.75 control reporter vector using PolyFect Transfection Reagent (Qiagen). For co-transfection experiments, 0.2 μg of wild-type NKX2.5-pEFSA, 0.2 μg of mutant NKX2.5-pEFSA or empty vector pEFSA, 1.0 μg of ANF(−2600)-Luc, and 0.04 μg of pGL4.75 were used. Firefly luciferase and Renilla luciferase activities were measured with the Dual-Glo Luciferase Assay System (Promega) 48 h after transfection. The activity of the ANF promoter was presented as fold activation of Firefly luciferase relative to Renilla luciferase. Three independent experiments were performed at minimum for wild-type and mutant NKX2.5.
Statistical analysis
Data are expressed as the means ± SD. Continuous variables were tested for normality of distribution and student’s unpaired t-test was used for comparison of numeric variables between two groups. Comparison of the categorical variables between two groups was performed using Pearson’s χ2 test or Fisher’s exact test when appropriate. A two-tailed P-value of <0.05 was considered to indicate a statistically significant difference.
Results
Characteristics of the study subjects
A cohort of 110 unrelated index patients with familial AF and a total of 200 ethnically-matched unrelated healthy individuals used as controls were enrolled and clinically evaluated. None of the participants had apparent traditional risk factors for AF. There was no significant difference between patient and control groups in baseline characteristics including age, gender, body mass index, blood pressure, fasting blood glucose, serum lipid, left atrial dimension, left ventricular ejection fraction, heart rate at rest, as well as life style (data not shown). In the present study, four patients were also diagnosed with hypertension in accordance with the criterion that the average systolic or diastolic blood pressure (2 readings made after 5 min of rest in the sitting position) is 140 or 90 mm Hg, respectively, but at the time of initial diagnosis of AF, their blood pressures were normal. The baseline clinical characteristics of the 110 index patients with familial AF are summarized in Table I.
NKX2.5 mutation
Direct sequencing of the coding exons and flanking introns of the NKX2.5 gene was carried out following PCR amplification of genomic DNA from the 110 index patients with familial AF. A heterozygous NKX2.5 mutation was identified in 1 out of 110 unrelated index patients, with a mutational prevalence of ~0.91% based on the patient cohort. In particular, a substitution of cytosine (C) for thymine (T) in the second nucleotide of codon 145 (c.434T>C), predicting the transition of phenylalanine (F) into serine (S) at amino acid 145 (p.F145S) was identified in an index patient from family 1. The sequence chromatograms showing the detected heterozygous NKX2.5 mutation of c.434T>C compared with corresponding control sequence are shown in Fig. 1. A schematic diagram of NKX2.5 delineating the putative structural domains and location of the mutation identified in AF patients is presented in Fig. 2. The missense mutation was not found in the control population nor was it reported in the EVS’s and NCBI’s SNP databases. Genetic scan of the available family members of the mutation carrier showed that the mutation was present in all affected family members alive, but was absent in unaffected family members examined. Analysis of the pedigree showed that the mutation cosegregated with AF transmitted as an autosomal dominant trait in the family with complete penetrance. The pedigree structure of the family is presented in Fig. 3. The phenotypic characteristics and results of genetic screening of the affected family members are listed in Table II.
Notably, congenital atrial septal defect was confirmed by the early echocardiogram in patient I-1 and patient II-2 from family 1. In addition, two previously reported NKX2.5 sequence polymorphisms, including c.63A>G and c.606C>G, were observed in both AF patients and control individuals. However, there was no significant difference in either of the two allele frequencies between the AF patient and healthy control groups. All the sequence variants and their allele frequencies are listed in Table III.
Alignment of multiple NKX2.5 protein sequences
A cross-species alignment of NKX2.5 protein sequences displayed that the altered amino acid was completely conserved evolutionarily, underscoring its functional importance (Fig. 4).
Disease-causing potential of the NKX2.5 variation
The sequence variation of c.434T>C detected in NKX2.5 was automatically predicted to be a disease-causing mutation with a P-value of 0.999997. No SNP in the altered region was found in the MutationTaster database.
Transcriptional activity of the NKX2.5 mutant
The transcriptional activation function of the mutated NKX2.5 in COS-7 cells was characterized using a luciferase reporter, which was driven by the promoter of ANP, one of NKX2.5-directed cardiac target genes. The activity of the ANP promoter was expressed as fold activation of Firefly luciferase relative to Renilla luciferase. The same amounts of wild-type (0.4 μg) and F145S-mutant NKX2.5 (0.4 μg) activated the ANP promoter by ~12- and ~3-fold, respectively. When the same amount of wild-type NKX2.5 (0.2 μg) was cotransfected with F145S-mutant NKX2.5 (0.2 μg), the induced activation of the ANP promoter was ~5-fold. These results suggest that the NKX2.5 mutation results in a significantly reduced transcriptional activation compared with its wild-type counterpart (Fig. 5).
Discussion
In the present study, a novel heterozygous NKX2.5 mutation, p.F145S, was identified in a family with familial AF. This missense mutation was present in all the affected family members examined but was absent in the unaffected family members available and in the 400 normal chromosomes from a matched control population. A cross-species alignment of multiple NKX2.5 protein sequences exhibited that the altered amino acid was completely conserved evolutionarily. Functional analysis demonstrated that the mutant NKX2.5 was associated with a significantly decreased transcriptional activity. Therefore, it is highly likely that functionally impaired NKX2.5 is involved in the pathogenesis of AF in this family. To our knowledge, this is the first report on the relationship between NKX2.5 loss-of-function mutation and enhanced susceptibility to AF.
NKX2.5 was first discovered as a homologue of tinman, a Drosophila cardiac transcription factor. It shows high expression in early embryonic heart progenitor cells and persists through adulthood, indispensable for proper cardiovascular development and maturation (46). The human NKX2.5 gene maps to chromosome 5q34 and consists of two exons encoding a protein of 324 amino acids (46). The NKX2.5 protein contains a homeodomain (HD), an evolutionarily conserved domain that recognizes and binds to a consensus DNA motif, AAGTG. In addition to the HD, NKX2.5 contains N- and C-terminal regulatory domains. The HD is centrally located at amino acid positions 138–197 and is involved in nuclear translocation and interaction with other transcription factors as well as DNA binding (59). The NKX2.5 mutation of p.F145S identified in this study is located in HD, and may thus be expected to exert influence on the transcriptional activity of NKX2.5 by interfering with its nuclear distribution or DNA-binding ability.
Previous investigations revealed that NKX2.5 is an upstream regulator of several genes expressed during embryogenesis including the genes that encode ANF, brain natriuretic peptide and α-cardiac actin (60). As a well-known NKX2.5 downstream target molecule, ANF contains several NKX2.5 binding sites in its proximal promoter region, including −242 from the transcription start site, which has been confirmed as an in vivo binding site of NKX2.5 (61). Therefore, the functional characteristics of the NKX2.5 mutation can be explored by analysis of the transcriptional activity of the ANF promoter in cells transfected with NKX2.5 mutant in contrast to its wild-type counterpart. In this study, the functional role of the novel p.F145S mutation of NKX2.5 identified in familial AF patients was analyzed by transcriptional activity assays and the results demonstrated a significantly decreased transcriptional activity on a downstream gene. These findings indicate that NKX2.5 haploinsufficiency caused by mutation is potentially an alternative pathophysiological mechanism of AF.
The finding that functionally compromised NKX2.5 predisposes to AF may be partially due to the abnormally developed pulmonary vein myocardium. The pulmonary venous vessel is ensheathed by a layer of myocardium referred to as pulmonary myocardial sleeve, which has been verified to be responsible for the initiation and perpetuation of AF by several potential arrhythmogenic mechanisms including enhanced intrinsic pacemaker activity and liability to reentrance (62–64). Genetic-labeling lineage tracing studies have validated that NKX2.5 is expressed in the atria as well as in the pulmonary myocardium and is essential for the localized formation of the sinoatrial node during embryogenesis. NKX2.5 may functionally serve as a suppressor of the sinoatrial node lineage gene program, which limits pacemaker activity to the sinoatrial and atrioventricular nodes. When the level of NKX2.5 protein decreased in a hypomorphic model, the pulmonary cardiomyocytes shifted to connexin40-negative, HCN4-positive cells, a nodal-like phenotype with pacemaker activity (63). In NKX2.5-null mouse embryos, HCN4 was activated along the entire embryonic heart tube, whereas connexin40 expression was inhibited, ectopic pacemaker cells were observed throughout the heart tube (64). Hence, NKX2.5 loss-of-function mutation presumably contributes to formation of the pulmonary myocardium sleeve and switch of the pulmonary myocardium to a sinoatrial node-like phenotype, creating an atrial electrophysiological matrix in favor of AF.
There are some downstream genes transactivated by NKX2.5, and mutations in several target genes have been linked to AF, including the genes encoding ANF and gap junction protein connexin40 (38,39,41–43). Therefore, it is probable that mutated NKX2.5 confers susceptibility to AF by decreasing expressions of target genes.
It is of note that congenital atrial septal defect was documented in 2 AF patients carrying p.F145S mutation of NKX2.5. Similarly, congenital cardiovascular malformations were previously confirmed in AF patients carrying GATA4, GATA5 or GATA6 mutations (49–57). Markedly, a long list of mutations in these genes has been implicated in a wide variety of congenital cardiovascular anomalies (65–75). These observational results indicate that AF may share a common genetic origin with congenital heart disease.
In conclusion, the findings of the present study provide novel insights into the molecular mechanism of AF, suggesting potential implications for early prophylaxis and allele-specific therapy of this common arrhythmia.