A variant in the carboxyl-terminus of connexin 40 alters GAP junctions and increases risk for tetralogy of Fallot


GJA5 gene (MIM no. 121013), localized at 1q21.1, encodes for the cardiac gap junction protein connexin 40. In humans, copy number variants of chromosome 1q21.1 have been associated with variable phenotypes comprising congenital heart disease (CHD), including isolated TOF. In mice, the deletion of Gja5 can cause a variety of complex CHDs, in particular of the cardiac outflow tract, corresponding to TOF in many cases. In the present study, we screened for mutations in the GJA5 gene 178 unrelated probands with isolated TOF. A heterozygous nucleotide change (c.793C>T) in exon 2 of the gene leading to the p.Pro265Ser variant at the carboxyl-terminus of the protein was found in two unrelated sporadic patients, one with classic anatomy and one with pulmonary atresia. This GJA5 missense substitution was not observed in 1568 ethnically-matched control chromosomes. Immunofluorescent staining and confocal microscopy revealed that cells expressing the mutant protein form sparse or no visible gap-junction plaques in the region of cell–cell contact. Moreover, analysis of the transfer of the gap junction permanent tracer lucifer yellow showed that cells expressing the mutant protein have a reduced rate of dye transfer compared with wild-type cells. Finally, use of a zebrafish model revealed that microinjection of the GJA5-p.Pro265Ser mutant disrupts overall morphology of the heart tube in the 37% (22/60) of embryos, compared with the 6% (4/66) of the GJA5 wild-type-injected embryos. These findings implicate GJA5 gene as a novel susceptibility gene for TOF.


Tetralogy of Fallot (TOF, MIM no. 187500) is the most common cyanotic heart defect (MIM no. 217095); it is estimated to occur in 2.5–3.5 per 10 000 live births, and accounts for 6.8% of all congenital heart disease (CHD).1 TOF can arise as the consequence of prenatal infections, maternal illnesses, and maternal therapeutic and non-therapeutic drug exposures. A proportion of TOF cases is associated with chromosome imbalances; 15% are heterozygous for chromosome 22q11.2 deletion, and nearly 7% have trisomy 21 (Down syndrome).2, 3 Non-syndromic TOF can be caused by dominant mutations in transcriptional regulators (NKX2.5,4, 5 GATA4,6, 7 ZFMP2/FOG2,8 FOXH19) and in genes involved in signal transduction (NOTCH1,10 JAG1,11, 12) CFC1,9 TDGF1,9 GDF1,13 and NODAL14). However, mutations in these genes have been related also to other CHDs.8, 9, 15, 16, 17 In addition, large de novo copy number variants have been implicated as a cause of non-syndromic TOF.18 Among these, a copy number variant at chromosome 1q21.1 was found in 1% of non-syndromic sporadic TOF cases.18 Recently, duplications at chromosome 1q21.1 have been associated with a 30-fold increase in the risk of TOF.19 GJA5 gene (MIM no. 121013), localized at 1q21.1, encodes for the cardiac gap junction protein connexin 40 (Cx40), and is highly expressed in the human outflow tract.18 A systematic study of the anatomy and histology of Cx40-deleted mice has revealed a variety of complex CHDs in these animals, with the most common malformations being of the conotruncal type, corresponding to TOF in many cases.20 In humans, abnormalities in the GJA5 gene have been associated with atrial fibrillation.21, 22, 23

In the present study, we evaluated whether GJA5 mutations might be pathogenic in isolated TOF, by gene screening in a series of non-syndromic TOF patients.

Materials and methods

Study subjects

The study cohort included 178 patients with non-syndromic TOF. One hundred and forty-seven cases had TOF with classic anatomy and 31 had TOF with pulmonary atresia. All patients were sporadic with the exception of four familial cases with TOF with classic anatomy. All patients underwent clinical assessment, electrocardiogram, echocardiogram, and chest-X-ray studies. Diagnosis of TOF was obtained by echocardiogram in all children. All patients showed viscero-atrial situs solitus, d-ventricular loop, and normally related great arteries. Patients showed neither major nor minor extracardiac features of DiGeorge syndrome/velo-cardio-facial-syndrome, neither deletion of 22q11 and 10p13 DiGeorge syndrome regions, which were excluded using standard molecular protocols. Moreover, these patients had been previously analyzed for mutations in the GATA4, NKX2.5, ZFPM2/FOG2, GDF1, ISLET1, and JAG1 genes by denaturing high performance liquid chromatography (DHPLC).8, 12 The study cohort also included 784 unrelated normal control individuals obtained as ethnically matched anonymous samples from the Mendel Institute (Italians of Caucasian origin). Ethical approval for this study was obtained from the Ethics Committees of the participating institutions, and informed consent for the genetic analyses was obtained from all patients or their legal guardians.

Molecular studies

To search for mutations in the GJA5 gene, the entire coding region of the gene, together with exon–intron boundaries, was PCR amplified from genomic DNA. Primers were designed based on the cDNA in GenBank (NM_005266.5) and the corresponding genomic regions. PCR products were screened by DHPLC by use of the Wave 3500 HT System (Transgenomic, Omaha, NE, USA) at column temperatures recommended by Navigator software, version (Transgenomic). Amplimers with abnormal elution profiles were purified (Microcon PCR (Millipore, Billerica, MA, USA)) and then sequenced bidirectionally using the ABI BigDye Terminator Sequencing Kit v.3.1 (Applied Biosystems, Foster City, CA, USA) and ABI Prism 3130xl Genetic Analyzers (Applied Biosystems). Primer pair sequences, as well as PCR and DHPLC analysis settings, are available upon request. The HomoloGene tool of NCBI (http://www.ncbi.nlm.nih.gov/nuccore?Db=homologene&Cmd=Retrieve&list_uids=3856) was used to analyze the level of conservation of sequence variants in orthologous genes. Numbering for the mutation started at the adenine nucleotide (A) in the ATG initiation codon. Genotyping of GJA5 c.793C>T variant in healthy controls was performed on genomic DNA using a custom TaqMan genotyping assay (Forward primer 5′-IndexTermCCCCTCTGTGGGCATAGTC-3′; reverse primer 5′-IndexTermGCCATTCTCCAGGCACTGATTA-3′; Applied Biosystems), according to manufacturer’s protocol on a 7900HT real-time PCR platform (Applied Biosystems). The reactions were cycled with standard TaqMan conditions (2 min at 50°C, 10 min at 95°C, 40 cycles with 15 s at 95°C, and annealing/extension at 60°C for 1 min). The genotypes were called with the SDS 2.2.2 software package (Applied Biosystems).

Statistical analyses

The occurrence of a mutation in a subject was assumed to follow a binomial distribution, for which the power to detect at least one mutation was equal to one minus the probability to observe zero events. Study power for different scenarios according to a set of hypothesized true mutation prevalence was provided (ie, from 2 to 10 per 1000 subjects). According to these scenarios a sample size of 784 subjects has a power >99% to detect at least one mutation with a type I error (α) equal to 0.05, assuming a mutation prevalence of 1% (Supplementary Figure S1).

The prevalence difference of mutation between patients and controls was tested using the Fisher’s exact test. A P-value <0.05 was considered as statistically significant.

The Fisher’s exact test was used to compare the frequency distributions for the morphology of the heart in zebrafish experiments. The significant threshold was set at P=0.05.

Plasmids constructs

The single-base change resulting in the p.Pro265Ser amino-acid substitution was introduced into a full-length cDNA of the wild-type human GJA5 cDNA cloned into a p-yellow fluorescence protein (YEFP)-N1 tagged plasmid, kindly provided by MH Gollob (Departement of Medicine, University of Ottawa Heart Institute, Ottawa, Canada), by site-directed mutagenesis with the use of the QuickChange Site II-direct mutagenesis kit (Stratagene, La Jolla, CA, USA). Both wild-type and mutated human GJA5 cDNAs were subsequently subcloned at the C-terminus of pCDNA6/V5HIS plasmid by using primers Cx40 forward (5′-IndexTermCACACAGGATTCATGGGCGATTGGAGCTTCCTG-3′) and Cx40 reverse (5′-IndexTermCACACACTCGAGCACTGATAGGTCATCTGACCT-3′), including restriction enzyme site for cloning.

Localization studies

For indirect immunofluorescent staining, murine neuroblastoma (N2A) cells, a gap-junction-deficient cell line, were transfected with GJA5-pYEFP-N1 wild-type or mutated plasmid using lipofectamine 2000 (Invitrogen Life Technologies, Carlsbad, CA, USA), as indicated by the manufacturer protocol. Cells were blocked and fixed in 4% paraformaldeide at 24–48 h after transfection. Nucleus was highlighted with Hoechst and the cells were examined using a fluorescence microscope (PCM Ecplise TE300, Nikon Instruments, Tokyo, Japan).

For confocal laser scanning microscopy, 3 × 103 N2A cells were seeded on glass coverslips, and transiently transfected with either wild-type or mutated GJA5- pCDNA6/V5HIS plasmid, as described below. Cells were fixed after 24–48 h with 4% paraformaldehyde, permeabilized with 0.5% Triton X-100, and stained with mouse monoclonal anti-V5 antibody (Life Technologies, Paisley, UK), followed by green-fluorescent Alexa Fluor-488 goat anti-mouse antibody (Molecular Probes, Eugene, OR, USA); actin cytoskeleton was detected by red-fluorescent Alexa Fluor-594 phalloidin; (Molecular Probes, Eugene, OR). The extensively rinsed cover glass was then mounted on the microscope slide with Vectashield Mounting Medium with DAPI (Vector Laboratories, Burlingame, CA, USA). Imaging was performed on a Leica TCS SP2 AOBS confocal microscope (Leica Microsystems, Mannheim, Germany), using excitation spectral laser lines at 405, 488, and 594 nm, tuned with an acousto-optical tunable filter. Signals from different fluorescent probes were taken in sequential scanning mode.

Dye transfer studies

Dye transfer through gap junction channels was investigated using transiently transfected N2A cells; experiments were performed 24 h after transfection with wild-type or mutated constructs. We patch-clamped only adjacent cells with YEFP-positive gap junctions localized at the cell membrane level in both experimental groups. Cells plated on glass coverslips were superfused at room temperature with Tyrode solution containing (mM): 154 NaCl, 4 KCl, 2 CaCl2, 1 MgCl2, 5 HEPES-NaOH, 5.5 D-glucose, adjusted to pH 7.35. Lucifer Yellow (LY) CH (2 mM) (Molecular Probes, Eugene, OR, USA) was dissolved in the pipette solution containing (mM): 110 K+-aspartate, 23 KCl, 0.4 CaCl2 (calculated free Ca2+=10–7 M), 3 MgCl2, 5 HEPES KOH, 1 EDTA KOH, 5 ATP-Na salt, pH 7.3. Large patch electrodes (about 1.5 MΩ) were used to facilitate the diffusion of LY in the cytosol of the patched cell after achievement of the whole-cell configuration, detected by monitoring membrane capacitance (Multiclamp 700B, Axon Instruments, Inverurie, Scotland). Measurements were performed with a Nikon TE200 inverted microscope at × 63 magnification (Nikon Instruments). Cells were excited with a xenon lamp (100 W) through suitable bandpass filters and a gray filter to limit dye bleaching. Cell images were acquired once in bright field mode and every minute in fluorescence mode by a MOTIC 2300 camera. Image acquisition and analysis were respectively controlled and performed by ImageJ software (National Institutes of Health, Bethesda, MD, USA); exposure time and gain factor were maintained constant throughout all measurements. Dye permeation through gap-junction channels was investigated using N2A cell pairs. Fluorescence in the injected and recipient cell was monitored at 1 min intervals for 15 min after patch rupture; cell autofluorescence was eliminated by digital subtraction of images acquired before dye dialysis (cell attached configuration). The time course of dye transfer to injected cell and, secondarily, to the recipient one was obtained by measuring fluorescence changes over time. The fluorescence at each time point was expressed as percent of the fluorescence measured at 15 min from the injected cell. Injected cell fluorescence at 15 min, taken as a reference, was assumed to result from equilibration with pipette solution and therefore to represent a LY concentration of 2 mM. LY concentration at each time-point was calculated from the fluorescence normalized to the reference one. The number of LY molecules (MNLY) corresponding to each LY concentration is: MNLY=Vol × C × N, where Vol indicates cell volume (assumed 0.75 pL), C is LY concentration and N is Avogadro’s number. LY flow rate is expressed as the change over time of MNLY present in the recipient cell. Means were compared by two-way ANOVA. Statistical significance was defined as P<0.05. Data are expressed as mean±SE of independent determinations.

Zebrafish and in situ hybridization experiments

Zebrafish were kept and the embryos were staged as described before.24 All procedures involving experimental animals were performed in compliance with local animal welfare laws, guidelines, and policies. In situ hybridizations were done essentially as described25 using probe specific for myocardial marker cardiac myosin light chain 2.26 Embryos were cleared in methanol and mounted in 100% glycerol before pictures were taken. Phenylthiourea was added to suppress pigmentation in developing embryos at 20 hpf.

RNA and injections

pCS2+−Gja5 constructs were linearized with NotI enzyme and capped mRNA prepared with the Message Machine kit (Ambion, Austin, TX, USA). mRNA was diluted in nuclease-free water, and 1 nl per embryo was injected at the 1- to 2-cell stage.


Mutation analysis

One hundred and seventy-eight unrelated TOF probands (either sporadic (n=174) or familial (n=4)) with isolated TOF with (n=31) or without (n=147) pulmonary atresia were screened for mutations in GJA5 gene by DHPLC. Amplicons bearing heteroduplexes were then analyzed by DNA sequencing. A heterozygous nucleotide change (c.793C>T) in exon 2 of the gene leading to the p.Pro265Ser variant was found in two unrelated sporadic patients with TOF, one with classic anatomy (TF131) and one with pulmonary atresia (AP18) (Figures 1a and b). None of them disclosed any conduction defect at electrocardiogram. The p.Pro265Ser substitution was found to alter an amino-acid residue that was highly conserved across multiple species, including chimpanzee, dog, cows, mouse, rat, chicken, and zebrafish (GenBank accession numbers: NP_001071490.1, NP_032147.1, NP_062153.1, NP_990835.1, NP_001007214.1, http://blast.ncbi.nlm.nih.gov/Blast.cgi) (Figure 1c). Family members were unable to be contacted to verify if p.Pro265Ser had de novo occurrence or was transmitted by parents. The p.Pro265Ser missense change was not observed in 784 unaffected, unrelated, ethnically matched control subjects (Italians of Caucasian origin) (1568 control chromosomes), a sample with a power >99% to detect a variant with 1% prevalence such as p.Pro265Ser substitution. Significantly, the prevalence of this substitution in patients was statistically different from that found in controls (2/178 (1.12) versus 0/784 (0.00), P=0.03). In addition to p.Pro265Ser variant, mutation analysis disclosed four previously reported GJA5 polymorphisms (c.-188C>T (rs1775479); c.-185A>G (rs1692140); c.531C>T, p.Tyr123Tyr (rs2232191); c.*53A>G (rs1692141)).

Figure 1

DHPLC elution profiles (a) and sequence electropherograms (b) of the novel GJA5 variant p.Pro265Ser. The position of the mutated nucleotide is indicated by an arrow. (c) Phylogenetic conservation analysis of GJA5 p.Pro265Ser variant. The mutant amino-acid residue 265 is highlighted in yellow.

Protein localization studies

To determine if the p.Pro265Ser variant had an effect on protein localization, N2A cells deficient in gap junctional communication were transiently transfected with either the wild-type or the p.Pro265Ser mutant constructs directly ‘tagged’ with YEFP at the carboxyl-terminus (CT) end of the protein and visualized using both indirect immunofluorescent staining and confocal microscopy. Following 24 h incubation, cells expressing wild-type Cx40 showed cytoplasmic punctate staining and aggregation at the plasma membrane of large gap-junction plaques in the region of cell–cell contact (Figure 2a). A different cellular distribution was observed for the p.Pro265Ser variant. Cells expressing this variant formed sparse or no visible gap-junction plaques and had intracellular retention of Cx40 within a wide array of punctuate structures (Figure 2b). Similar staining patterns were obtained using confocal microscopy (Figures 2c and d).

Figure 2

Differential localization of the wild-type and GJA5 p.Pro265Ser mutant proteins in N2A cells. N2A cells were transfected with plasmids encoding either wild-type GJA5 or the GJA5 mutant p.Pro265Ser, tagged with YEFP at the CT. In cells expressing YEFP-tagged GJA5 wild-type gap-junction-like structures were readily seen at cell–cell interfaces (white arrows) (a). However, cells expressing YEFP-tagged p.Pro265Ser mutant failed to assemble gap junctions or formed sparse gap junctions (b). Confocal microscopic analysis of the wild-type and GJA5- p.Pro265Ser mutant proteins in N2A cells. N2A cells expressing wild-type GJA5 formed large gap-junction plaques between adjacent cells (c). In contrast, N2A cells transfected with the p.Pro265Ser mutant showed formation of punctuate gap-junction-like structures at cell–cell interfaces (d).

Dye transfer studies

In order to explore whether functional intercellular channels are formed, in particular by the mutant proteins that traffic to the plasma membrane, the transfer of the gap junction permeant tracer LY between cells transfected either with the wild-type or mutant construct was investigated. Dye transfer efficiency was quantified by measuring the rate of dye transfer between adjoining cells. As shown in Figure 3a, LY dialyzed into a cell transfected with wild-type GJA5 cDNAs diffused easily to the adjacent cell reaching equilibration in about 13–15 min. As LY does not diffuse through the N2A cell membrane, the intercellular diffusion of the dye suggests the presence of permeable gap junctions, in this case formed by Cx40. The rate of fluorescence increment in injected cells was similar between cells transfected with mutant and wild-type (n=14 for both) constructs thus their results were pooled. On the other hand, the rate of fluorescence increment in the recipient cell was lower steep in the presence of mutated gap junctions (P<0.05). For instance, at 13 min the relative fluorescence dye intensity was 18.9±3.4% (corresponding to 2.9 × 105 dye molecules per second) and 9.2±2.7% (corresponding to 1.3 × 105 dye molecules per second) with wild-type and mutated constructs, respectively, (P<0.05, Figure 3b).

Figure 3

Dye transfer (LY) through gap-junction channels in N2A cell pairs transfected with either wild-type (WT), or mutated (MUT) constructs. (a) Examples of the intercellular diffusion of the dye at different time points after patch rupture (time 0); the bright field image and the fluorescence one at 15 min after patch rupture are superimposed. (b) Average time course of absolute (left panel) and relative (right panel) fluorescence; in each panel the upper curve refers to the injected cell (data from WT and MUT were pooled) and the lower ones to WT and MUT recipient cells, respectively.

Zebrafish experiments

Previous data have shown that Cx40 has an important role in cardiomyocytes of the atria, cardiac conduction system, and endothelial cells of large arteries. It has already been demonstrated that Cx40 knockout mice exhibit cardiac electrophysiological and morphological phenotypes.27 We used the zebrafish embryo model system to elucidate the role of mutant GJA5 in heart patterning in vivo. To this end, we injected synthetic mRNA encoding mutant GJA5-p.Pro265Ser or as a control wild-type GJA5 into zebrafish embryos at the 1- to 2-cell stage and assessed the morphology of the heart. Aberrant heart morphology was apparent in embryos-expressing mutant GJA5 at 24 h post fertilization. In situ hybridization analysis for cardiac myosin light chain 2, a cardiac myosin marker, showed that the overall morphology of the heart tube was disrupted in the 37% (22/60) of GJA5- p.Pro265Ser-injected embryos, compared with the 6% (4/66) of GJA5-wild-type-injected embryos (Fisher’s exact test, P<0.0001) (Figure 4). Noteworthy, the mutant GJA5-injected embryos did not display gross developmental defects and had normal body size compared with wild-type GJA5-injected embryos. These data demonstrate that mutant GJA5-p.Pro265Ser has a dominant effect over endogenous wild-type GJA5 and disrupts normal heart morphology.

Figure 4

Expression of mutant GJA5-p.Pro265Ser disrupts zebrafish heart morphology in zebrafish in vivo. In situ hybridization staining for cardiac myosin (cardiac myosin light chain 2) in 48 hpf zebrafish embryos that were injected at the 1- to 2-cell stage with wild-type GJA5 (left) or mutant GJA5-p.Pro265Ser (right). The overall morphology of the heart tube is disrupted in GJA5-p.Pro265Ser-injected embryos compared with GJA5-wild-type-injected embryos. Dorsal views are shown with anterior to the top. A, atrium; v, ventricle.


Deficient or improper gap junction channel function caused by mutations in the genes encoding connexin proteins has been associated with a variety of diseases, such as peripheral neuropathy, oculo-dento-digital dysplasia, sensorineural deafness, skin disorders, and cataracts.28 This is the first report of GJA5 variant in TOF patients and in general in CHDs. Various GJA5 defects other than p.Pro265Ser were previously associated with familial and sporadic atrial fibrillation.21, 22, 23 Similarly, different mutations of cardiac transcription factor NKX2.5 have been identified in kindreds with conduction abnormalities (atrioventricular block) and concurrent congenital heart malformations, primarily secundum atrial septal defect, as well as in patients with sporadic TOF.4, 15, 29 In the present study, variant p.Pro265Ser was found in 1% of TOF patients. The low incidence of p.Pro265Ser variant in the cohort we studied may be consistent with the high degree of heterogeneity of this CHD. On analysis of 230 patients with both syndromic and non syndromic TOF, genetic abnormalities were identified in the 18% of patients.30 The most common genetic alteration was 22q11.2 deletion associated with DiGeorge/VCFS syndrome. Single-gene mutations in JAG1, NKX2.5, and TBX1 were found in 1.3%, 0.9%, and 0.4% of the cohort, respectively. In a parallel study, we found JAG1, NKX2.5, and ZPFM2/FOG2 mutations in 2.7, 1.1, and 0.6% of sporadic patients with isolated TOF.8, 12 Clearly, TOF is a heterogeneous CHD, with small contributions from different loci. Present results further expand the genetic heterogeneity of this defect.

GJA5 gene mutated in TOF maps to the 1q21.1 chromosomal region. Microdeletion 1q21.1 (del 1q21.1) and the reciprocal microduplication 1q21.1 (dup 1q21.1) are newly recognized genomic disorders, characterized by developmental delay, neuropsychiatric abnormalities, dysmorphic features, and a variety of congenital malformations, including CHDs.31, 32, 33 In addition, both del 1q21.1 and dup 1q21.1 have been found in patients with non-syndromic CHD.18, 34, 35 A recent study of 512 patients with apparently isolated sporadic TOF using an array designed to detect deletions or duplications revealed that 5/512 individuals had a deletion or duplication of chromosome 1q21.1, encompassing GJA5 gene.18 A more recent study further defined the relationship between 1q21.1 rearrangements and isolated TOF showing that 1q21.1 duplications confer an odds ratio for TOF of 30.19 Furthermore, within the 1q21.1 region, the same authors identified three smaller overlapping duplications in three patients with TOF in which GJA5 gene was the only gene in common. Moreover, they found a GJA5 triplication in one patient with pulmonary atresia.19 Similarly, we identified GJA5 p.Pro265Ser variant in one patient with TOF and one with TOF and pulmonary atresia, which is considered the extreme end of the anatomic spectrum of TOF and is a cardiac defect with a very high genetic involvement.36, 37 Taken together, present and previous findings strongly suggest that GJA5 is the key gene in the etiology of cardiovascular diseases found in patients with 1q21.1 chromosome imbalances.

In agreement with a pathogenic role of Cx40 mutations in CHDs, previous studies on animal models have demonstrated that this gene has a major role in the cardiac conduction system and in heart morphogenesis. Cx40 gene is strongly expressed in both atria and ventricles in early stages of development.38 Kirchhoff et al39 have shown that Cx40-deficient mice may have cardiac hypertrophy in conjunction with common atrioventricular junction or a ventricular septal defect. In a subsequent study, Gu et al20 have shown that both Cx40 heterozygous and homozygous-null mice exhibit complex cardiac malformations, such as endocardial cushion defects and conotruncal malformations. In particular, over one third of the heart from homozygous-null mice showed either DORV or TOF. Consistent with results obtained in mice, we demonstrated that expression of Cx40-p.Pro265Ser mutant in zebrafish disrupts the formation of heart tube in over one third of the cases. Taken together, present and previous results provide reliable evidence that Cx40 participates in early cardiac morphogenesis and its absence, altered expression, or abnormal functioning increases the probability of cardiac malformations.

Present data have shown that overexpression of Cx40-p.Pro265Ser mutant in N2A cells, which are gap junction- and connexin-deficient, reduces dye transfer through gap junctions. Previous studies in mouse aortic endothelium deficient in Cx40 showed a similar reduction in dye transfer rate compared with wild-type aortas. On the basis of protein localization results, we hypothesize that reduced intercellular communication in cells expressing p.Pro265Ser variant may be the consequence of a reduction in membrane associated Cx40, and the fact that we do see some punctuate staining and even plaque formation in these cells might explain residual dye transfer capacity of mutant-expressing cells. Perhaps, Cx40-p.Pro265Ser variant is a hypomorphic allele with reduced level of protein expression, or more likely, an allele acting in a dominant-negative fashion, as this latter hypothesis would better explain the dominant effect over the endogenous wild-type protein observed in zebrafish embryos microinjected with the mutant transcript. It might be possible that the mutant protein forms nonfunctional chimeric connexons with wild-type Cx40, which are not properly inserted into the cell membrane.

The GJA5-p.Pro265Ser variant identified here affects a highly conserved proline located in the CT-domain. Connexins-CT usually function as vital regulatory elements of channel gating, and may have influences on the rates of connexin trafficking and the synthesis, assembly, and degradation of gap-junction channels. Nuclear magnetic resonance (NMR) data have shown that the Cx40CT domain is intrinsically disordered in structure.40 In general, intrinsically disordered proteins have been identified as having important roles in molecular recognition, molecular assembly, protein modifications, and entropic chains.41 In the case of membrane proteins, like Cx40, intrinsically disordered domains seem to have an important role in cell signaling events, acting as components of macromolecular processes and substrates for regulatory events such as phosphorylation.42 In this regard it is intriguing that phylogenetically conserved proline265 is contained within a stretch of 11 amino acids (VQGLTPPPDFN) that is highly affected upon binding of the c-Src SH3 domain. Therefore, it might be that an SH3-binding site is lost as a result of the p.Pro265Ser mutation, leading to dysregulation of Cx40. c-Src is a plasma membrane-associated tyrosine kinase, and c-Src induced phosphorylation of Cx43 has been correlated with an inhibition of gap junction intercellular communication. It was also shown that Src activation may lead to phosphorylation of tyrosine residues as well as indirectly serine phosphorylation (perhaps MAPK-mediated),43 Moreover, it was seen that tyrosine phosphorylation by Src occur through binding to proline-rich-residues.44 The Cx40CT sequence does not contain a consensus PXXP SH3-binding motif, but contains two peptides rich in prolines and/or positively charged residues.45 One of these proline-rich domain is the Cx40CT domain affected by SH3 corresponding to the peptide between Val258 to Asn268, and containing Pro265 residue mutated in our TOF patients. NMR analyses have also shown that binding by Src may have an important role in regulating interactions at the CT with other molecular partners of Cx40, in example zonula occludens-1 (ZO-1).46, 47 This might be relevant as interaction with ZO-1 was found to influence the formation47 and size of the Cx43 gap-junction plaque,48 as well as the internalization and remodeling of Cx43 upon changes in the intracellular environment.46, 49

Future studies should be designed to investigate the role of p.Pro265Ser in regulating interactions of Cx40-CT with its molecular partners and biological consequences.

Accession codes




  1. 1

    Ferencz C, Rubin JD, McCarter RJ et al. Congenital heart disease: prevalence at livebirth. The Baltimore-Washington Infant Study. Am J Epidemiol 1985; 121: 31–36.

    CAS  Article  PubMed  Google Scholar 

  2. 2

    Amati F, Mari A, Digilio MC et al. 22q11 deletions in isolated and syndromic patients with tetralogy of Fallot. Hum Genet 1995; 95: 479–482.

    CAS  Article  PubMed  Google Scholar 

  3. 3

    Goldmuntz E, Clark BJ, Mitchell LE et al. Frequency of 22q11 deletions in patients with conotruncal defects. J Am Coll Cardiol 1998; 32: 492–498.

    CAS  Article  PubMed  Google Scholar 

  4. 4

    Goldmuntz E, Geiger E, Benson DW : NKX2.5 mutations in patients with tetralogy of fallot. Circulation 2001; 104: 2565–2568.

    CAS  Article  PubMed  Google Scholar 

  5. 5

    McElhinney DB, Geiger E, Blinder J, Benson DW, Goldmuntz E : NKX2.5 mutations in patients with congenital heart disease. J Am Coll Cardiol 2003; 42: 1650–1655.

    CAS  Article  PubMed  Google Scholar 

  6. 6

    Nemer G, Fadlalah F, Usta J et al. A novel mutation in the GATA4 gene in patients with Tetralogy of Fallot. Hum Mutat 2006; 27: 293–294.

    Article  PubMed  Google Scholar 

  7. 7

    Tomita-Mitchell A, Maslen CL, Morris CD, Garg V, Goldmuntz E : GATA4 sequence variants in patients with congenital heart disease. J Med Genet 2007; 44: 779–783.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  8. 8

    De Luca A, Sarkozy A, Ferese R et al. New mutations in ZFPM2/FOG2 gene in tetralogy of Fallot and double outlet right ventricle. Clin Genet 2011; 80: 184–190.

    CAS  Article  PubMed  Google Scholar 

  9. 9

    Roessler E, Ouspenskaia MV, Karkera JD et al. Reduced NODAL signaling strength via mutation of several pathway members including FOXH1 is linked to human heart defects and holoprosencephaly. Am J Hum Genet 2008; 83: 18–29.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  10. 10

    Garg V, Muth AN, Ransom JF et al. Mutations in NOTCH1 cause aortic valve disease. Nature 2005; 437: 270–274.

    CAS  Article  PubMed  Google Scholar 

  11. 11

    Eldadah ZA, Hamosh A, Biery NJ et al. Familial Tetralogy of Fallot caused by mutation in the jagged1 gene. Hum Mol Genet 2001; 10: 163–169.

    CAS  Article  PubMed  Google Scholar 

  12. 12

    Guida V, Chiappe F, Ferese R et al. Novel and recurrent JAG1 mutations in patients with tetralogy of Fallot. Clin Genet 2011; 80: 591–594.

    CAS  Article  PubMed  Google Scholar 

  13. 13

    Karkera JD, Lee JS, Roessler E et al. Loss-of-function mutations in growth differentiation factor-1 (GDF1) are associated with congenital heart defects in humans. Am J Hum Genet 2007; 81: 987–994.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  14. 14

    Roessler E, Pei W, Ouspenskaia MV et al. Cumulative ligand activity of NODAL mutations and modifiers are linked to human heart defects and holoprosencephaly. Mol Genet Metab 2009; 98: 225–234.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  15. 15

    Schott JJ, Benson DW, Basson CT et al. Congenital heart disease caused by mutations in the transcription factor NKX2-5. Science 1998; 281: 108–111.

    CAS  Article  PubMed  Google Scholar 

  16. 16

    Garg V, Kathiriya IS, Barnes R et al. GATA4 mutations cause human congenital heart defects and reveal an interaction with TBX5. Nature 2003; 424: 443–447.

    CAS  Article  PubMed  Google Scholar 

  17. 17

    De Luca A, Sarkozy A, Consoli F et al. Familial transposition of the great arteries caused by multiple mutations in laterality genes. Heart 2010; 96: 673–677.

    Article  PubMed  Google Scholar 

  18. 18

    Greenway SC, Pereira AC, Lin JC et al. De novo copy number variants identify new genes and loci in isolated sporadic tetralogy of Fallot. Nat Genet 2009; 41: 931–935.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  19. 19

    Soemedi R, Topf A, Wilson IJ et al. Phenotype-specific effect of chromosome 1q21.1 rearrangements and GJA5 duplications in 2436 congenital heart disease patients and 6760 controls. Hum Mol Genet 2012; 21: 1513–1520.

    CAS  Article  PubMed  Google Scholar 

  20. 20

    Gu H, Smith FC, Taffet SM, Delmar M : High incidence of cardiac malformations in connexin40-deficient mice. Circ Res 2003; 93: 201–206.

    CAS  Article  PubMed  Google Scholar 

  21. 21

    Gollob MH, Jones DL, Krahn AD et al. Somatic mutations in the connexin 40 gene (GJA5) in atrial fibrillation. N Engl J Med 2006; 354: 2677–2688.

    CAS  Article  PubMed  Google Scholar 

  22. 22

    Yang YQ, Liu X, Zhang XL et al. Novel connexin40 missense mutations in patients with familial atrial fibrillation. Europace 2010; 12: 1421–1427.

    Article  PubMed  Google Scholar 

  23. 23

    Yang YQ, Zhang XL, Wang XH et al. Connexin40 nonsense mutation in familial atrial fibrillation. Int J Mol Med 2010; 26: 605–610.

    CAS  PubMed  Google Scholar 

  24. 24

    Westerfield M : The Zebrafish Book. A Guide for the Laboratory Use of Zebrafish (Danio rerio). University of Oregon Press: Eugene, OR, 1995.

    Google Scholar 

  25. 25

    Thisse C, Thisse B, Schilling TF, Postlethwait JH : Structure of the zebrafish snail1 gene and its expression in wild-type, spadetail and no tail mutant embryos. Development 1993; 119: 1203–1215.

    CAS  PubMed  Google Scholar 

  26. 26

    Yelon D, Horne SA, Stainier DY : Restricted expression of cardiac myosin genes reveals regulated aspects of heart tube assembly in zebrafish. Dev Biol 1999; 214: 23–37.

    CAS  Article  PubMed  Google Scholar 

  27. 27

    Gros D, Dupays L, Alcolea S, Meysen S, Miquerol L, Theveniau-Ruissy M : Genetically modified mice: tools to decode the functions of connexins in the heart-new models for cardiovascular research. Cardiovasc Res 2004; 62: 299–308.

    CAS  Article  PubMed  Google Scholar 

  28. 28

    Pfenniger A, Wohlwend A, Kwak BR : Mutations in connexin genes and disease. Eur J Clin Invest 2011; 41: 103–116.

    CAS  Article  PubMed  Google Scholar 

  29. 29

    Sarkozy A, Conti E, Neri C et al. Spectrum of atrial septal defects associated with mutations of NKX2.5 and GATA4 transcription factors. J Med Genet 2005; 42: e16.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  30. 30

    Rauch R, Hofbeck M, Zweier C et al. Comprehensive genotype-phenotype analysis in 230 patients with tetralogy of Fallot. J Med Genet 2010; 47: 321–331.

    CAS  Article  PubMed  Google Scholar 

  31. 31

    Brunetti-Pierri N, Berg JS, Scaglia F et al. Recurrent reciprocal 1q21.1 deletions and duplications associated with microcephaly or macrocephaly and developmental and behavioral abnormalities. Nat Genet 2008; 40: 1466–1471.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  32. 32

    Mefford HC, Eichler EE : Duplication hotspots, rare genomic disorders, and common disease. Curr Opin Genet Dev 2009; 19: 196–204.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  33. 33

    Mefford HC, Sharp AJ, Baker C et al. Recurrent rearrangements of chromosome 1q21.1 and variable pediatric phenotypes. N Engl J Med 2008; 359: 1685–1699.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  34. 34

    Christiansen J, Dyck JD, Elyas BG et al. Chromosome 1q21.1 contiguous gene deletion is associated with congenital heart disease. Circ Res 2004; 94: 1429–1435.

    CAS  Article  PubMed  Google Scholar 

  35. 35

    Erdogan F, Larsen LA, Zhang L et al. High frequency of submicroscopic genomic aberrations detected by tiling path array comparative genome hybridisation in patients with isolated congenital heart disease. J Med Genet 2008; 45: 704–709.

    CAS  Article  PubMed  Google Scholar 

  36. 36

    Digilio MC, Marino B, Grazioli S, Agostino D, Giannotti A, Dallapiccola B : Comparison of occurrence of genetic syndromes in ventricular septal defect with pulmonic stenosis (classic tetralogy of Fallot) versus ventricular septal defect with pulmonic atresia. Am J Cardiol 1996; 77: 1375–1376.

    CAS  Article  PubMed  Google Scholar 

  37. 37

    Marino B, Digilio MC, Toscano A et al. Anatomic patterns of conotruncal defects associated with deletion 22q11. Genet Med 2001; 3: 45–48.

    CAS  Article  PubMed  Google Scholar 

  38. 38

    van Veen AA, van Rijen HV, Opthof T : Cardiac gap junction channels: modulation of expression and channel properties. Cardiovasc Res 2001; 51: 217–229.

    CAS  Article  PubMed  Google Scholar 

  39. 39

    Kirchhoff S, Kim JS, Hagendorff A et al. Abnormal cardiac conduction and morphogenesis in connexin40 and connexin43 double-deficient mice. Circ Res 2000; 87: 399–405.

    CAS  Article  PubMed  Google Scholar 

  40. 40

    Bouvier D, Kieken F, Kellezi A, Sorgen PL : Structural changes in the carboxyl terminus of the gap junction protein connexin 40 caused by the interaction with c-Src and zonula occludens-1. Cell Commun Adhes 2008; 15: 107–118.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  41. 41

    Dunker AK, Brown CJ, Lawson JD, Iakoucheva LM, Obradovic Z : Intrinsic disorder and protein function. Biochemistry 2002; 41: 6573–6582.

    CAS  Article  PubMed  Google Scholar 

  42. 42

    Dunker AK, Obradovic Z : The protein trinity – linking function and disorder. Nat Biotechnol 2001; 19: 805–806.

    CAS  Article  PubMed  Google Scholar 

  43. 43

    Zhou L, Kasperek EM, Nicholson BJ : Dissection of the molecular basis of pp60(v-src) induced gating of connexin 43 gap junction channels. J Cell Biol 1999; 144: 1033–1045.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  44. 44

    Lin R, Warn-Cramer BJ, Kurata WE, Lau AF : v-Src phosphorylation of connexin 43 on Tyr247 and Tyr265 disrupts gap junctional communication. J Cell Biol 2001; 154: 815–827.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  45. 45

    Jia CY, Nie J, Wu C, Li C, Li SS : Novel Src homology 3 domain-binding motifs identified from proteomic screen of a Pro-rich region. Mol Cell Proteomics 2005; 4: 1155–1166.

    CAS  Article  PubMed  Google Scholar 

  46. 46

    Duffy HS, Ashton AW, O’Donnell P et al. Regulation of connexin43 protein complexes by intracellular acidification. Circ Res 2004; 94: 215–222.

    CAS  Article  PubMed  Google Scholar 

  47. 47

    Toyofuku T, Yabuki M, Otsu K, Kuzuya T, Hori M, Tada M : Direct association of the gap junction protein connexin-43 with ZO-1 in cardiac myocytes. J Biol Chem 1998; 273: 12725–12731.

    CAS  Article  PubMed  Google Scholar 

  48. 48

    Hunter AW, Barker RJ, Zhu C, Gourdie RG : Zonula occludens-1 alters connexin43 gap junction size and organization by influencing channel accretion. Mol Biol Cell 2005; 16: 5686–5698.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  49. 49

    Barker RJ, Price RL, Gourdie RG : Increased association of ZO-1 with connexin43 during remodeling of cardiac gap junctions. Circ Res 2002; 90: 317–324.

    CAS  Article  PubMed  Google Scholar 

Download references


The authors wish to thank the patients who participated in this research. We especially thank Dr MH Gollob for kindly providing the full-length human wild-type GJA5 cDNA. This work was supported in part by the Italian Ministry of Health grant RC2011.

Author information



Corresponding author

Correspondence to Alessandro De Luca.

Ethics declarations

Competing interests

The authors declare no conflict of interest.

Additional information

Supplementary Information accompanies the paper on European Journal of Human Genetics website

Supplementary information

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Guida, V., Ferese, R., Rocchetti, M. et al. A variant in the carboxyl-terminus of connexin 40 alters GAP junctions and increases risk for tetralogy of Fallot. Eur J Hum Genet 21, 69–75 (2013). https://doi.org/10.1038/ejhg.2012.109

Download citation


  • congenital heart disease
  • tetralogy of Fallot
  • 1q21.1
  • GJA5
  • connexin 40

Further reading