Mutations in the cardiac transcription factor NKX2.5 affect diverse cardiac developmental pathways

Heterozygous mutations in NKX2.5, a homeobox transcription factor, were reported to cause secundum atrial septal defects and result in atrioventricular (AV) conduction block during postnatal life. To further characterize the role of NKX2.5 in cardiac morphogenesis, we sought additional mutations in groups of probands with cardiac anomalies and first-degree AV block, idiopathic AV block, or tetralogy of Fallot. We identified 7 novel mutations by sequence analysis of the NKX2.5-coding region in 26 individuals. Associated phenotypes included AV block, which was the primary manifestation of cardiac disease in nearly a quarter of affected individuals, as well as atrial septal defect and ventricular septal defect. Ventricular septal defect was associated with tetralogy of Fallot or double-outlet right ventricle in 3 individuals. Ebstein’s anomaly and other tricuspid valve abnormalities were also present. Mutations in human NKX2.5 cause a variety of cardiac anomalies and may account for a clinically significant portion of tetralogy of Fallot and idiopathic AV block. The coinheritance of NKX2.5 mutations with various congenital heart defects suggests that this transcription factor contributes to diverse cardiac developmental pathways.

J. Clin. Invest.104:1567–1573 (1999).

Transcription factors are increasingly recognized as having key roles in the complex biological processes governing cardiac development (1–3). The homeobox transcription factor, tinman, a Drosophila NK-gene, has been of particular interest because it is expressed in the dorsal vessel, an insect equivalent of the vertebrate heart. Targeted disruption of tinman results in the absence of the dorsal vessel (4).

A murine Nkx2.5 homeobox gene (5), also independently described as a cardiac-specific homeobox gene (CSX; see ref. 6), is a vertebrate homologue of tinman. Nkx2.5 is expressed in early cardiac mesoderm and in heart muscle lineage throughout life (5–7); targeted disruption is lethal to embryos and arrests cardiac development at the linear heart tube stage (8, 9). Molecular dissection of Nkx2.5 revealed the participation of highly conserved regions of Nkx2.5 in DNA binding, protein-protein interactions, nuclear translocation, and regulation of other transcription factors (9–18).

We recently described 3 heterozygote mutations in NKX2.5 that resulted in atrioventricular (AV) conduction block. Many genotype-positive individuals also had a secundum atrial septal defect (ASD) (19). In kindreds with NKX2.5 mutation, there was a history of other congenital heart malformations in some family members including ventricular septal defect and tetralogy of Fallot (TOF). To further characterize the role of NKX2.5 in human cardiac morphogenesis, we have identified additional mutations and defined the clinical phenotypes produced by these mutations.

In this report, 7 additional NKX2.5 mutations are described. As reported previously, AV conduction abnormalities and ASD occurred commonly (19). However, we also found evidence of several different types of ventricular septal defects (VSD) that were sometimes associated with double-outlet right ventricle and tetralogy of Fallot. In addition, tricuspid valve abnormalities, including Ebstein’s anomaly, were noted in some individuals. These observations illustrate an essential role for NKX2.5 in atrial, ventricular, and conotruncal septation, AV valve formation, and maintenance of AV conduction.

Mutation analysis.

As summarized in Table 2, 7 NKX2.5 mutations were identified in 7 probands. Numbering for all the mutations starts at the adenine nucleotide (A) in the ATG initiation codon (8). Five mutations changed a restriction enzyme site allowing independent confirmation; 2 mutations were confirmed by allele-specific oligonucleotide hybridization (23). One missense mutation is predicted to result in codon change in the 5′-coding region (Arg25Cys), and 3 missense mutations are predicted to result in codon changes in the homeodomain (Asn188Lys, Arg189Gly, and Tyr191Cys). Two single nucleotide changes are predicted to produce a termination codon in the homeodomain (Gln149ter) or in the 3′-coding region (Tyr259ter). One mutation changed the first nucleotide of the intron (the splice-donor site) (Int 1DSG+1T).

Table 2

Genotype/Phenotype summary

These NKX2.5 sequence variants were considered mutations based on their cosegregation with cardiovascular disease, the significant change each was predicted to cause in the protein structure of NKX2.5, and their absence in more than 100 chromosomes derived from unrelated normal subjects.

In addition to the 7 mutations, an A172G polymorphism or silent mutation (Glu21Glu) was observed in 52 control samples such that 33% were AA, 36% were AG, and 31% were GG. The polymorphism creates a BpmI restriction enzyme site allowing independent confirmation.

Clinical features of genotype-positive individuals

Group 1: Cardiovascular malformations and first-degree AV block. Six members from 3 generations of Family BEA had a Gln149ter mutation (Figure 1). Three individuals had an ASD (Table 2). I-2 was diagnosed with ASD and advanced second-degree AV block in adulthood and underwent surgical closure and pacemaker implantation at age 25 years. II-2 underwent ASD closure at 6 years of age, and at age 30 years II-2 has marked first degree AV block. III-2 underwent ASD closure at age 10 months for failure to thrive, a muscular VSD that spontaneously closed had been noted earlier in life (Figure 2). II-4 was diagnosed with tetralogy of Fallot and ASD in infancy. II-4 was treated with a Blalock-Taussig shunt at 18 months and definitive surgery at age 6 years. A pacemaker was implanted at age 8 years. III-3 had ECG evidence of first-degree AV block at age 5 years. Her sister, III-4, had echocardiographic documentation of a midmuscular VSD that spontaneously closed in infancy and a normal ECG at age 3 years.

Figure 2

Echocardiogram/Doppler showing secundum ASD and posterior muscular VSD (BEA III-2). Subcostal view shows large secundum ASD (left). The parasternal short axis view shows posterior muscular VSD with left ventricle to right ventricle blood flow on color Doppler interrogation (right). The VSD spontaneously closed during the first 6 months of life; the ASD was surgically closed at 10 months of age. LA, left atrium; LV, left ventricle; RA, right atrium; RV, right ventricle.

Five members from 3 generations of Family BEF had an Arg189Gly mutation (Figure 1). Based on echocardiographic evidence of reduced ventricular function and rhythm disturbances (including complex ventricular ectopy, atrial flutter, atrial fibrillation, and/or advanced second-degree AV block), II-2, III-4, and III-5 were considered to have a familial cardiomyopathy; all were treated with a pacemaker. Transesophageal echocardiography revealed a small ASD in II-2 and III-4. IV-5 underwent surgical closure of ASD in 1998 at age 7 years. Her brother, IV-3, had echocardiographic evidence of a large ASD with a little left to right shunt. Findings at cardiac catheterization and surgery demonstrated a small tricuspid valve. Both IV-3 and IV-5 have prolonged PR interval. I-1 was not genotyped, but he had a history of recurrent syncope and died suddenly in 1947 at age 52 years.

Seven members of Family BEJ had a Tyr259ter mutation (Figure 1). II-4 and II-5 had undergone ASD closure and pacemaker implantation at age 27 and 19 years, respectively. II-1 had advanced second-degree AV block and ventricular dysrhythmias. III-2 underwent ASD closure at age 6 years; a perimembranous VSD was spontaneously occluded by tricuspid valve tissue. III-5 underwent ASD closure at age 13 years. At 15 months of age, III-6 underwent closure of an ASD and closure of a VSD associated with double-outlet right ventricle by baffling the left ventricular outflow through the VSD into the aorta. IV-1 was diagnosed in infancy with ASD and “Swiss cheese” VSD. Clinical information was available on 2 individuals who were not genotyped. III-1 died of pneumonia complicated by severe heart failure resulting from large ASD and VSD in 1967, at 22 months of age. III-8 died in 1979 at 17 days of age after surgery for coarctation of the aorta; he also had ASD and VSD. All affected family members had ECG evidence of prolonged AV conduction.

BEP II-3 had a Tyr191Cys mutation. The father, mother, and siblings did not carry this mutation, indicating that it is de novo (pedigree not shown). First-degree AV block, ASD, and VSD were diagnosed in infancy. Surgical closure of ASD and VSD was performed at 6 months of age because of medically refractory heart failure.

Five members of family BET had an Asn188Lys mutation (Figure 1). I-1 was diagnosed with ASD at age 37 years, and he underwent ASD closure and pacemaker implantation at the age of 53 years. Medically refractory heart failure led to heart transplantation at age 60 years. II-2 was followed for many years with a diagnosis of Ebstein’s anomaly of the tricuspid valve; she underwent ASD closure at age 37 years. II-3 underwent ASD closure at age 8 years. III-2 has Ebstein’s anomaly and underwent ASD closure at age 2 years (Figure 3). III-5 underwent ASD closure at age 7 years and was noted to have a tethered tricuspid valve. All affected family members have evidence of abnormal AV conduction. In addition, III-2 and III-5 have right bundle branch block, commonly seen in Ebstein’s anomaly. Clinical data are available for several individuals who could not be genotyped. II-5 was a “blue baby” who died in 1956 at 13 months of age. III-1 and III-4 were spontaneous first trimester abortions. III-3 was delivered prematurely after a 36-week gestation by emergency Cesarean section for placenta abruption; death occurred shortly after birth.

Figure 3

Echocardiogram and ECG characteristics of Ebstein’s anomaly (BET III-2). (a) Four chamber echocardiogram during diastole (left) and systole (right); arrows indicate location of tricuspid and mitral valve (left and right). A sail-like anterior tricuspid valve leaflet is evident in diastole; inferior displacement of the septal leaflet is evident in systole. RA, right atrium; RV, right ventricle; LV, left ventricle. (b) ECG shows first-degree AV block and right bundle branch block (bottom).

Group 2: Idiopathic second- or third-degree AV block. CHB3 has a splice site mutation in exon 1 (Int 1DSG+1T, Figure 4); his was the only NKX2.5 mutation among the 10 probands tested. No phenotypic feature of this individual distinguished him from others in this group. He presented at 12 years of age with a 1-year history of recurrent syncope. Advanced second-degree AV block was identified on ECG (not shown); no other cardiac abnormalities were noted. Pacemaker implantation was performed. Evaluation of his mother and 2 younger brothers was normal, and none carried the mutation (pedigree not shown). His father (not genotyped) had died suddenly — presumably because of arrhythmias — in 1984 at 29 years of age. His heart weight was 300 g, and no major abnormalities were noted at autopsy.

Figure 4

Diagram of the NKX2.5 gene showing 10 heterozygote mutations. Three mutations were reported previously (Gln170ter, Thr178Met, Gln198ter) (19). The region enclosed by boxes indicates exon 1 (left) and exon 2 (right); the horizontal line indicates the intron. Shaded boxes indicate the 5′ and 3′ untranslated regions (UTR), respectively. The white box denotes the coding region, and the black boxes indicate the conserved TN domain, homeodomain (HD), and NK domain, respectively, within the coding region. Nonsense mutations are indicated above and missense mutations below the gene diagram.

Group 3: Tetralogy of Fallot negative for del22q11. TOF7 has an Arg25Cys mutation. Hers was the only NKX2.5 mutation of 20 probands tested. There was no phenotypic feature of this individual that distinguished her from other group members, but unlike BEA II-4, who also had tetralogy of Fallot, she did not have AV block or an ASD. She underwent surgery at 12 months of age for typical tetralogy of Fallot and 2 small muscular VSDs. She had small pulmonary arteries with areas of peripheral narrowing, and she had undergone balloon dilation on 2 occasions. There was no family history of other heart disease; we were not able to genotype other family members.

Observations in this study provide additional evidence for a role of NKX2.5 beyond early cardiac progenitor commitment; the coinheritance of heterozygous NKX2.5 mutations with various congenital heart defects suggests that this transcription factor contributes to diverse cardiac developmental pathways including atrial, ventricular, and conotruncal septation, AV conduction, and AV valve formation. As described previously, AV block and ASD are common in kindreds in whom congenital heart disease is due to NKX2.5 mutation (19). Additionally, NKX2.5 mutation causes AV block without associated congenital heart defects, and, as demonstrated in the patients we studied, AV block was the principal clinical finding in 23% of genotype-positive individuals. Further, based on results of the present study, NKX2.5 mutation may account for a clinically significant portion of idiopathic AV block.

However, cardiovascular defects resulting from NKX2.5 mutations extend beyond AV block and ASD. For example, VSD was present in 31% of genotype-positive individuals, including conoventricular VSD associated with tetralogy of Fallot and double-outlet, right ventricle and muscular VSD that closed spontaneously in infancy (24). Results of this study suggest that in individuals without a del22q11, NKX2.5 mutations may account for a significant portion of conotruncal defects (25). Abnormalities of the tricuspid valve were also present (15%), and for the first time, we have identified a genetic cause of Ebstein’s anomaly. Based on the diversity of cardiac phenotypes, we hypothesize that many additional forms of congenital heart disease may result from NKX2.5 mutations. The small sample size we studied does not allow estimation of a portion of cardiac anomalies resulting from NKX2.5; our study is further limited by concentration on mutations in coding sequences only.

The 7 NKX2.5 mutations described in this study, and the 3 reported previously, are most often located in the region encoding the homeodomain and less often in the 5′ and 3′ ends of the gene (Figure 4). Of the 10 total mutations, 5 are predicted to result in truncated proteins that have the potential for dominant-negative or deleterious gain of function effects. The other 5 missense mutations are predicted to alter single, highly conserved amino acids in the homeodomain. The mechanism by which these mutations produce diverse cardiac malformations remains to be determined. Pauli et al. (26) reported a patient with haploinsufficiency of NKX2.5 (CSX) resulting from a distal chromosome 5q deletion that manifested ASD, AV block, and ventricular noncompaction. Because of the large number of our patients who had ASD and AV block, many of the mutations we have identified may function as null alleles. However, some specific genotype/phenotype correlation can be speculated. For example, the 4 individuals with tricuspid valve abnormalities, including Ebstein’s anomaly, come from 2 families where missense mutations in adjacent codons (Asn188Lys and Arg189Gly) were identified.

Schott et al. (19) described 2 families with an identical NKX2.5 mutation and an identical haplotype indicating common ancestry. In subjects we evaluated, each proband had a novel mutation. Out of the 7 mutations, 3 occurred in CpG islands, which have an elevated mutation rate compared with other dinucleotides (27). The mutation in BEP II-11 is a sporadic occurrence as neither parent carries the Tyr191Cys mutation. TOF7 and CHB3 may be examples of sporadic occurrence, but we could not confirm this possibility because both parents could not be genotyped.

Previous studies have identified Nkx2.5 as an upstream regulator and/or transcriptional activator of other genes expressed during cardiac development including Hand1 (eHand) (9, 11), myocyte enhancer factor-2 (MEF2) (9, 12), myosin light chain 2V (MLC2V) (9), atrial natriuretic factor gene (9, 13), brain natriuretic peptide (BNP) (9), α-cardiac actin gene (14), cardiac ankyrin repeat protein gene (15), N-myc (9), and MSX2 (9). Nkx2.5 binding to target DNA may occur in conjunction with other factors including GATA-4 (16, 18) and serum response factor (14). The demonstration of synergistic transcriptional activation mediated by Nkx2.5 and GATA-4 (28), and the elucidation of the HoxB1-Pbx1 and Ultrabithorax-Extradenticle structure of heteromeric complexes (29, 30), suggests ways in which mutations may alter the specificity and affinity of DNA binding, while also disrupting the tightly regulated spatial and temporal gene expression that is essential during cardiac development.

The severe phenotypes associated with heterozygous NKX2.5 mutations in humans are surprising, given the phenotypes of heterozygous mutations modeled in other organisms. For example, ablation of Nkx2.5 in mice was embryonic lethal, but heart defects were not observed in heterozygous mutant mice (8, 9). The differences between mice and humans may reflect a bias resulting from observing only severely affected humans. Alternatively, these differences may reflect different cardiac development in mice and humans, genetic redundancy in humans, and/or unrecognized phenotypes in heterozygous mutant mice.

Diverse cardiac malformations are recognized in other monogenic human disorders, e.g., heterozygous TBX5 mutations in Holt-Oram syndrome (31–33), as well as gene-targeted mice, e.g., heterozygous mutations in retinoic X receptor–deficient mice (34). Although variable expressivity is not a feature predicted by classic embryological models of cardiac development, heritable monogenic mutations can clearly cause pleiotropic cardiovascular defects. Based on results of the present studies, NKX2.5 appears to be a likely candidate gene for a number of forms of cardiovascular disease in the young.

We are indebted to the members of these families for their participation. These studies would not have been possible without the skillful assistance of Linda Johnson, Barbara Roberts, Marlene Brabham, Ping Lu, and Kris Houston. This work was supported in part by National Institutes of Health grant 1 P50 HL61006-01 (to D.W. Benson) and Howard Hughes Medical Institute (to J.G. Seidman and C.E. Seidman).

1. Olson, EN, Srivastava, D. Molecular pathways controlling heart development. Science 1996. 272:671-676.
   View this article via: PubMed CrossRef Google Scholar
2. Harvey, RD. Nk-2 homeobox genes and heart development. Dev Biol 1996. 178:203-216.
   View this article via: PubMed CrossRef Google Scholar
3. Fishman, MC, Olson, EN. Parsing the heart: genetic modules for organ assembly. Cell 1997. 91:153-156.
   View this article via: PubMed CrossRef Google Scholar
4. Bodmer, R. The gene, tinman, is required for specification of the heart and visceral muscles in Drosophilia. Development 1993. 118:719-729.
   View this article via: PubMed Google Scholar
5. Lints, TJ, Parsons, LM, Hartley, L, Lyons, I, Harvey, RP. Nkx 2-5: a novel murine homeobox gene expressed in early heart progenitor cells and their myogenic descendents. Development 1993. 119:419-431.
   View this article via: PubMed Google Scholar
6. Komuro, I, Izumo, S. Csx: a murine homeobox–containing gene specifically expressed in the developing heart. Proc Natl Acad Sci USA 1993. 90:8145-8149.
   View this article via: PubMed CrossRef Google Scholar
7. Shiojima, I, et al. Molecular cloning and characterization of human cardiac homeobox gene CSX1. Circ Res 1996. 79:920-929.
   View this article via: PubMed Google Scholar
8. Lyons, I, et al. Myogenic and morphogenetic defects in the heart tubes of murine embryos lacking the homeo box gene Nkx2-5. Genes Dev 1995. 9:1654-1666.
   View this article via: PubMed CrossRef Google Scholar
9. Tanaka, M, Chen, Z, Bartunkova, S, Yamasaki, N, Izumo, S. The cardiac homeobox gene Csx/Nkx2.5 lies genetically upstream of multiple genes essential for heart development. Development 1999. 126:1269-1280.
   View this article via: PubMed Google Scholar
10. Chen, CY, Schwartz, RJ. Identification of novel DNA binding targets and regulatory domains of a murine Tinman homeodomain factor, nkx2-5. J Biol Chem 1995. 270:15628-15633.
    View this article via: PubMed CrossRef Google Scholar
11. Biben, C, Harvey, RP. Homeodomain factor Nkx2-5 controls left/right asymmetric expression of bHLH gene eHand during murine heart development. Genes Dev 1997. 11:1357-1369.
    View this article via: PubMed CrossRef Google Scholar
12. Gajewski, K, Kim, Y, Lee, YM, Olson, EN, Schulz, RA. D-mef2 is a target for Tinman activation during Drosophilia heart development. EMBO J 1997. 16:515-522.
    View this article via: PubMed CrossRef Google Scholar
13. Durocher, D, Chen, C-Y, Ardati, A, Schwartz, RJ, Nemer, M. The atrial natriuretic factor promoter is a downstream target for Nkx2-5 in the myocardium. Mol Cell Biol 1996. 16:4648-4655.
    View this article via: PubMed Google Scholar
14. Chen, CY, et al. Activation of the cardiac alpha-actin promoter depends upon serum response factor, Tinman homologue, Nkx2-5, and intact serum response elements. Dev Genet 1996. 19:119-130.
    View this article via: PubMed CrossRef Google Scholar
15. Zou, Y, et al. CARP, a cardiac ankyrin repeat protein is downstream in the Nkx2-5 homeobox gene pathway. Development 1997. 124:793-804.
    View this article via: PubMed Google Scholar
16. Durocher, D, Charron, F, Warren, R, Schwartz, RJ, Nemer, M. The cardiac transcription factors Nkx2-5 and GATA-4 are mutual cofactors. EMBO J 1997. 16:5687-5696.
    View this article via: PubMed CrossRef Google Scholar
17. Kasahara, H, Izumo, S. Identification of the in vivo casein kinase II phosphorylation site within the homeodomain of the cardiac tissue-specifying homeobox gene product Csx/Nkx2-5. Mol Cell Biol 1999. 19:526-536.
    View this article via: PubMed Google Scholar
18. Lee, Y, et al. The cardiac tissue-restricted homeobox protein Csx/ Nkx2-5 physically associates with the zinc finger protein GATA4 and cooperatively activates atrial natriuretic factor gene expression. Mol Cell Biol 1998. 18:3120-3129.
    View this article via: PubMed Google Scholar
19. Schott, J-J, et al. Congenital heart disease caused by mutations in the transcription factor Nkx2-5. Science 1998. 281:108-111.
    View this article via: PubMed CrossRef Google Scholar
20. Buyon, JP, et al. Autoimmune-associated congenital heart block: demographics, mortality, morbidity and recurrence rates obtained from a national neonatal lupus registry. J Am Coll Cardiol 1998. 31:1658-1666.
    View this article via: PubMed CrossRef Google Scholar
21. Johnson, MC, Hing, A, Wood, MK, Watson, MS. Chromosome abnormalities in congenital heart disease. Am J Med Genet 1997. 70:292-298.
    View this article via: PubMed CrossRef Google Scholar
22. Turbay, D, Wechsler, SB, Blanchard, KM, Izumo, S. Molecular cloning, chromosomal mapping, and characterization of the human cardiac-specific homeobox gene hCsx. Mol Med 1996. 2:86-96.
    View this article via: PubMed Google Scholar
23. Benson, DW, et al. A missense mutation in the pore region of HERG causes familial long QT syndrome. Circulation 1996. 93:1791-1795.
    View this article via: PubMed Google Scholar
24. Van Praagh, R, Geva, T, Kreutzer, J. Ventricular septal defects: how shall we describe, name and classify them? J Am Coll Cardiol 1989. 14:1298-1299.
    View this article via: PubMed CrossRef Google Scholar
25. Goldmuntz, E, et al. Frequency of 22q11 deletions in patients with conotruncal defects. J Am Coll Cardiol 1998. 32:492-498.
    View this article via: PubMed CrossRef Google Scholar
26. Pauli, RM, Scheib-Wixted, S, Cripe, L, Izumo, S, Sekhon, GS. Ventricular noncompaction and distal chromosome 5q deletion. Am J Med Genet 1999. 85:419-423.
    View this article via: PubMed CrossRef Google Scholar
27. Ketterling, RP, Vielhaber, E, Sommer, S. The rates of G:C-T:A and G:C-C:G transversions at CpG dinucleotides in the human Factor IX gene. Am J Hum Genet 1994. 54:831-835.
    View this article via: PubMed Google Scholar
28. Shiojima, I, et al. Context-dependent transcriptional cooperation mediated by cardiac transcription factors Csx/Nkx2.5 and GATA-4. J Biol Chem 1999. 274:8231-8239.
    View this article via: PubMed CrossRef Google Scholar
29. Piper, DE, Batchelor, AH, Chang, C-P, Cleary, ML, Wolberger, C. Structure of a HoxB1-Pbx1 heterodimer bound to DNA: role of the hexapeptide and a fourth homeodomain helix in complex formation. Cell 1999. 96:587-597.
    View this article via: PubMed CrossRef Google Scholar
30. Passner, JM, Ryoo, HD, Shen, L, Mann, RS, Aggarwal, AK. Structure of a DNA-bound Ultrabithorax-Extradenticle homeodomain complex. Nature 1999. 397:714-719.
    View this article via: PubMed CrossRef Google Scholar
31. Li, QY, et al. Holt-Oram syndrome is caused by mutations in TBX5 a member of the Brachyury (T) gene family. Nat Genet 1997. 15:21-29.
    View this article via: PubMed CrossRef Google Scholar
32. Basson, CT, et al. Mutations in human TBX5 cause limb and cardiac malformation in Holt-Oram syndrome. Nat Genet 1997. 15:30-35.
    View this article via: PubMed CrossRef Google Scholar
33. Basson, CT, et al. Different TBX5 interactions in heart and limb defined by Holt-Oram syndrome mutations. Proc Natl Acad Sci USA 1999. 96:2919-2924.
    View this article via: PubMed CrossRef Google Scholar
34. Gruber, PJ, et al. RXRα deficiency confers genetic susceptibility for aortic sac, conotruncal, atrioventricular cushion, and ventricular muscle defects in mice. J Clin Invest 1996. 98:1332-1343.
    View this article via: JCI PubMed Google Scholar