GATA 4 Deletions Associated with Congenital Heart Diseases in South Brazil

Abstract

The normal development of the heart comprises a highly regulated machinery of genetic events, involving transcriptional factors. Congenital heart disease (CHD), have been associated with chromosomal abnormalities and copy number variants (CNVs). Our goal was to investigate through the multiplex ligation-dependent probe amplification (MLPA) technique, the presence of CNVs in reference genes for normal cardiac development in patients with CHD. GATA4 , NKX2–5 , TBX5 , BMP4 , and CRELD1 genes and 22q11.2 chromosome region were analyzed in 207 children with CHD admitted for the first time in a cardiac intensive care unit from a pediatric hospital. CNVs were detected in seven patients (3.4%): four had a 22q11.2 deletion (22q11DS) (1.9%), two had a GATA4 deletion (1%) and one had a 22q11.2 duplication (0.5%). No patients with CNVs in the NKX2–5 , TBX5 , BMP4 , and CRELD1 genes were identified. GATA4 deletions appear to be present in a significant number of CHD patients, especially those with septal defects, persistent left superior vena cava, pulmonary artery abnormalities, and extracardiac findings. GATA4 screening seems to be more effective when directed to these CHDs. The investigation of CNVs in GATA4 and 22q11 chromosome region in patients with CHD is important to anticipating the diagnosis, and to contributing to family planning.

Introduction

Congenital heart diseases (CHDs) consists of structural changes in the heart and large vessels and are recognized as the leading cause of neonatal mortality, affecting approximately 1% of newborns. ^{1 2} Regarding CHDs etiology, 20% are attributed to chromosomal and/or single-gene alterations; the remaining resulting from a combination of genetic, epigenetic, and environmental factors. Chromosomal abnormalities and copy number variants (CNVs) contribute to the risk of CHD. ³

CNVs in GATA4 , NKX2–5 , and TBX5 genes have been identified among mechanisms that may explain some CHDs, since all these are transcription factors (TFs) strongly involved in cardiogenesis. The altered expression of GATA4 (OMIM: 600576) can results in common CHDs, such as atrial septal defects (ASDs), ventricular septal defects (VSDs), and pulmonary stenosis (PS). Facial dysmorphisms and mental retardation may also be present. ^{4 5 6} It is believed that, all TFs acts together by regulating the cardiac septum formation, whereas its haploinsufficiency results in developmental heart disorders. ⁷ Several clinical syndromes associated with CHDs have also been related to different CNVs. The 22q11 deletion syndrome (22q11DS; OMIM: 188400) was the first that became known and it is one the most common human microdeletion syndromes. ⁸

Multiplex ligation-dependent probe amplification (MLPA) is a variation of the multiplex polymerase chain reaction method. It is used to identify CNVs, including deletions and duplications in CHDs predisposition genes. ⁹ CNVs searching through MLPA in patients with CHD can help either in the early diagnosis or in the identification of the specific cause of the disease. Thus, we used the MLPA technique to investigate the presence of 22q11DS and other CNVs related to genes involved in normal cardiac development in patients with CHDs at a referral hospital from Southern Brazil.

Materials and Methods

All participants were ascertainment from the cardiac intensive care unit (ICU) at the Hospital da Criança Santo Antônio (HCSA), Porto Alegre, RS, Brazil, during a period of 1 year. HCSA belongs to the Santa Casa of Misericórdia of Porto Alegre (SCMPA), and is a reference center for the evaluation and treatment of patients with CHD.

Genomic DNA was extracted from peripheral blood by a standard protocol. MLPA assay was performed using the SALSA MLPA Probemix P311 CHD kit (MRC-Holland, Amsterdam, the Netherlands) for CNVs screening on GATA4 , NKX2–5 , TBX5 , BMP4 , and CRELD1 genes, and on 22q11.2 chromosome region following the manufacturer's recommendations. The P311-A1 kit included probes are described in Table 1 . The MLPA products were studied on an ABI3130 sequencer (ThermoFischer Scientific), and were analyzed with Coffalyser software (MRC-Holland) that normalizes the signals from all probes and compare them with reference samples. For each comparative analysis, three normal controls were used.

Table 1. Probe sequences by SALSA MLPA Probemix P311 congenital heart disease.

Gene	Exon	Partial sequence (24 nt adjacent to ligation site)
GATA4	Upstream	CATGCTCAAGAT-AGGCACTGGAGC
	Upstream	GAGGTTCTTCTT-TAAAATCCATTC
	Exon 1	TTTCTTCCCTTT-CTTTGCTCCTTC
	Exon 3	CTCAGTAGATAT-GTTTGACGACTT
	Exon 4	CTACATGAAGCT-CCACGGGGTACG
	Exon 5	AAGAACCTGAAT-AAATCTAAGACA
	Exon 6	CAACTCCAGCAA-CGCCACCACCAG
	Exon 7	CACAAGGCTATG-CGTCTCCCGTCA
	CTSB gene	AAGTGTAGCAAG-ATCTGTGAGCCT
NKX2–5	Exon 1	CCGGCCAAGTGT-GCGTCTGCCTTT
	Exon 1	AGGTGAGGAGGA-AACACAGGCCCC
	Exon 2	CGCTCCAGCTCA-TAGACCTGCGCC
	Exon 2	CGGGATTCCGCA-GAGCAACTCGGG
TBX5	Exon 1	GACGTTGGAAGA-AGACCTGGCCTA
	Intron 1	CTATTCTGGGTA-AGCAGTAAACCC
	Exon 2	GCCTGACGCAAA-AGACCTGCCCTG
	Exon 3	AATCAAAGTGTT-TCTCCATGAAAG
	Exon 5	TCCTTCCAGAAA-CTCAAGCTCACC
	Exon 6	TACCAGCCTAGA-TTACACATCGTG
	Exon 8	GTGAGGCAAAAA-GTGGCCTCCAAC
	Exon 8	CCATTGTACCAA-GAGGAAAGGTGA
	Exon 9	AAGAAGATTCCT-TCTACCGCTCT
	Exon 9	TTTGCTTTGGTT-TTGTCCTGCCTT
BMP4	Exon 1	TGCAGGGACCTA-TGGTGAGCAAGG
	Intron 1	CGCAGGCCGAAA-GCTGTTCACCGT
	Exon 3	TGGTAACCGAAT-GCTGATGGTCGT
	Exon 4	AACATCTGGAGA-ACATCCCAGGGA
CRELD1	Exon 4	CAAGTCAGACTT-CGAGTGCCACCG
	Exon 11	GCATTCCCCATC-TTAACTGATTTA
22q11 region (genes)	CDC45	ATGTTCGTGTCC-GATTTCCGCAAA
	GP1BB	CACAACCGAGCT-GGTGCTGACCGG
	DGCR8	GACTCAGCGACT-GCACCAGTGGCA

Abbreviation: MLPA, multiplex ligation dependent probe amplification.

Some patients of this study were subject of a previous report by Rosa et al, ¹⁰ where a high-resolution GTG-banding karyotype, and a fluorescence in situ hybridization (FISH) for 22q11 microdeletion were performed, using the DiGeorge/VCFS Region Probe (TUPLE1; Vysis, Abbott Laboratories, AbbottPark, Illinois, United States). Rosa et al ¹⁰ found chromosomal anomalies through karyotype in 29 patients (14%), and a 22qq11 microdeletion by FISH in four patients (1.9%).

All patients were subject of a dysmorphological examination by a clinical geneticist to classify as a syndromic or nonsyndromic CHDs case. CHDs were described based on the results of echocardiography, cardiac catheterization, and/or surgical description, using the classification of Botto et al ¹¹ . In each participant, family history of CHDs was also register. The study was approved by the institutional Ethics Committee in May 10, 2017 (number: 2.315.917). Written informed consent was obtained from patient's parents who participated in this study.

Results

A total of 210 patients were hospitalized in the Cardiac ICU of the HCSA during the period of the study. From these, three individuals were excluded due to lack of DNA sample for the MLPA analysis. Thus, final sample consisted of 207 patients, 110 males (52.4%), and 107 females (47.6%), aged between 1 day and 13.5 years (average of 220 days). By ethnicity, 79% were Caucasians and 21% of Afro-descendants. Included patients were hospitalized either for cardiac surgery (74.8%), cardiologic evaluation (14.1%), or cardiac catheterization (9.2%). The most frequent CHDs observed were VSDs (17.4%), followed by ASDs (15.9%), the tetralogy of Fallot (11.1%), coarctation of the aorta (10.6%), and atrioventricular septal defects (9.7%). Sixty-four patients (30.9%) were classified as syndromic CHDs and 69.1% as nonsyndromic CHDs. Familial recurrence of CHD was observed in 15.7% of the studied sample.

MLPA assay detected CNVs in seven patients (3.4%). Four had a 22q11.2 deletion. Clinical phenotype observed in the four patients with 22q11DS has already been described by Rosa et al ¹⁰ in a previous work. In addition, we identified heterozygous deletions in all seven exons of GATA4 gene (del GATA 4) in two patients (0.96%; Fig. 1A ). Additional details of exons and sequences involved in the del GATA 4 are included in Table 2 . Clinical findings presented in these two male patients are reported on Table 3 . One child had a VSD, and the other a PS. Notably, both had an associated persistence of left superior vena cava (PLSVC), and an ostium secundum-type ASD. They both had also a membranous-type VSD, and an abnormality of the pulmonary artery (bicuspid pulmonary valve, and PS, respectively). Through the physical exam, patient 2 was considered as with a syndromic CHD. Noteworthy, this patient had a previous clinical suspicion of Williams–Beuren syndrome (WBS), due to its facial dysmorphism ( Fig. 1A ). However, additional analysis using MLPA for abnormalities in 7q11.23 (WBS region) rule outed this diagnostic possibility. In patient 1, complementary evaluations disclosed the presence of an asymptomatic ectopic kidney. Thus, although both presented associated extracardiac findings, their phenotypes seem to be different among them. This could be explained by differences in the size of GATA4 deletion implicated in each patient, since MLPA evaluated only the involvement of this gene. None of them had a family history of CHDs. The frequency of delGATA4 among the 36 patients with VSDs was 2.8%, and among the nine patients with PS or pulmonary atresia was 11.1%. From 14 patients with PLSVC who account for 6.8% of all studied sample, two had delGATA4 (14.3%). Noteworthy, from four patients with PLSVC associated to abnormality of the pulmonary artery, two presented delGATA4 (50%).

Fig. 1.

Multiplex ligation dependent probe amplification (MLPA) P311-B1 CHD analysis. MLPA analysis ( A ) positive for deletion of the GATA4 (patients 1 and 2); ( B ) positive for 22q11.2DS (patients 3, 4, 5, and 6), and ( C ) positive for 22q11DupS (patient 7). Reference values: DQ, dosage quotient; Normal, between red and blue lines; 0.80 < DQ < 1.2; heterozygous deletion (red dots) 0.40 < DQ < 0.65; heterozygous duplication (blue dots) 1.30 < DQ < 1.65.

Table 2. delGATA4 information.

Length (nt)	GATA4 exon	Ligation site NM_002052.5	Partial sequence (24 nt adjacent to ligation site)
362	Upstream	8.4 kb upstream of exon 1	CATGCTCAAGAT-AGGCACTGGAGC
229	Upstream	6.9 kb upstream of exon 1	GAGGTTCTTCTT-TAAAATCCATTC
	Start codon	612–614 (exon 2)
148	Exon 1	22 nt after exon 1	TTTCTTCCCTTT-CTTTGCTCCTTC
337	Exon 3	1231–1232	CTCAGTAGATAT-GTTTGACGACTT
355	Exon 4	1513–1514	CTACATGAAGCT-CCACGGGGTACG
142	Exon 5	1589–1590	AAGAACCTGAAT-AAATCTAAGACA
238	Exon 6	1666–1667	CAACTCCAGCAA-CGCCACCACCAG
202	Exon 7	1842–1843	CACAAGGCTATG-CGTCTCCCGTCA
	Stop codon	1938–1940 (exon 7)
130	CTSB gene	89.3 kb downstream	AAGTGTAGCAAG-ATCTGTGAGCCT

Table 3. CNVs detected by MLPA and clinical phenotype of CHD patients.

Clinical and genetic features	Patient
	1	2	7
CNV	delGATA4	delGATA4	dup22q11
Syndromic	N	Y	N
Age	year 1 mo 17 d	year 10 mo 17 d	3 y 3 mo 13 d
Growth retardation			+
Microcephaly	+
Craniofacial features
High forehead		+
Broad forehead			+
Telecanthus			+
Up-slanting palpebral fissures
Epicanthic folds		+	+
Hypoplastic nares	+
Long philtrum	+	+
Thin upper lip	+
High arched palate		+
Micrognathia	+	+
Over-fold helix		+
Preauricular pits			+
CHD
ASD	+	+
VSD	+	+
PLSVC	+	+
PVS		+
Subvalvular aortic ring			+
Renal anomalies
Ectopic kidney	+
Mental retardation			+

Abbreviations: ASD, atrial septal defect; CHD, congenital heart disease; CNV, copy number variation MLPA, multiplex ligation dependent probe amplification; N, no; PLSVC, persistent left superior vena cava; PVS, pulmonary valve stenosis; VSD, ventricular septal defect; Y, Yes.

In addition, heterozygous deletion in the 22q11.2 chromosome region identified in four patients (1.9%) was the most prevalent CNV ( Fig. 1B ). These results agree with previous findings of Rosa et al ¹⁰ who used only FISH technique. However, one patient had a heterozygous duplication of 22q11.2 region (22q11DupS; 0.5%; Fig. 1C ), which was not detected through FISH analysis by Rosa et al. ¹⁰ Duplication occurred in a region close to FISH probes hybridization site. Finally, no alterations were identified in the NKX2–5 , TBX5 , BMP4 , and CRELD1 genes.

Discussion

With the advent of molecular cytogenetic techniques, such as MLPA, CNVs, detection has been substantially improved. Panels of nucleotide variants and CNVs when combined have demonstrated efficiency as first line diagnostic test in constitutional imbalances, with a detection potential of 15 to 30% for CHD-causing variants. ^{12 13} Here, we detected CNVs in 3.4% of the patients tested, very similar to the result found by Sørensen et al ¹⁴ (3.2%), but lower than frequencies described by Monteiro et al ¹⁵ (23.4%), and by Campos et al ¹⁶ (17.9%). The authors believe that such variations probably were caused by differences in criteria for patient selection.

It is known that CNVs involving GATA4 , located at 8p23.1, are associated with CHDs, where mechanism of pathogenicity leads to gene haploinsufficiency. The most frequent defects are VSDs, and ASDs. Our study identified del GATA4 (∼151 kb) in two patients (0.96%; Fig. 1A ). PS, a CHD also associated to del GATA4 , was observed in one of our patients. Other complex defects, as the tetralogy of Fallot, and double-outlet right ventricle, have been also described in individuals with del GATA4 . Facial dysmorphisms and mental retardation may be present, especially in patients with larger 8p deletions involving GATA4 ^{5 6} (our two patients had facial dysmorphisms, in special patient 2). Our data agree with the report of El Malti et al ¹³ who detected one case among 154 patients (0.7%). Moreover, recent studies have not detected patients with GATA4 alterations, despite the similarity of the clinical phenotype of the patients. ^{17 18 19}

GATA 4 deletions have been described in familial cases of CHDs, noted overlapping between phenotypes and cardiac septal defects. In these cases, interactions among GATA4 , TBX5 , and NKX2–5 is interrupted ^{4 5} that lead in to proliferation defects of cardiomyocytes during embryogenesis. ^{22 23} In our sample, however, none of our patients with del GATA4 had family history of CHD.

Among the 207 patients evaluated, four cases presented 22q11DS (∼630kb) [22q11:17,847,478–18,477,850 (hg18)], covering CDC45 , GP1BB , and DGCR8 genes ( Fig. 1B ). In addition, one case presented 22q11DupS ( Fig. 1C ; Table 3 ). 22q11DS has a very heterogeneous phenotype including CHD, hypocalcemia, immunodeficiency, facial dysmorphic, and neurodevelopment disorders, ⁸ corroborating with the clinical findings of this study. 22q11DS is among the main cause of CHDs, ¹⁰ and it was the most frequent alteration found in our study (1.9%). This finding agrees with the previous reports, where detection rate range from 0.4 to 8.5%. ^{11 13 14 15 18} 22q11DupS identification is reported less commonly, ranging from 0.7 to 2.5%, ^{13 14 16} very similar to our own (0.5%). Identification of 22q11DupS has been increased due to screening techniques currently used. However, its prevalence, and definitive phenotype remain unknown. ²¹

Conclusion

Thus, del GATA4 seems to be present in a significant number of patients with CHD-causing variants, especially among those with septal defects, PLSVC, abnormality of the pulmonary artery, and extracardiac findings (perhaps, manifestations reminiscent of WBS, and renal abnormalities). Thus, testing for del GATA4 could be directed to patients with these findings. The frequency of abnormalities involving GATA4 can be even higher and depends on the technique used for its detection. Based on all this, we believe that GATA4 gene should be considered in the screening of patients with CHDs.

Investigation for GATA4 mutations, and of 22q11.2 deletion syndrome in patients with CHDs constitute an important issue in prenatal diagnosis, contributing to family planning through appropriate genetic counseling and management of extracardiac symptoms, reducing the consequences of the disease. Understanding the etiology of CHDs will help to incorporate therapeutic and preventive interventions using molecular approaches as screening method, increasing the role of genetics in clinical care.

Funding Statement

Funding This study was supported by FAPERGS (17/2551–0001063–9), Research Support Foundation of the State of Rio Grande do Sul; PROEXT, Support Program for University Extension of MEC - Ministry of Education and Culture; and CAPES, Coordination of Superior Level Staff Improvement (001). Research productivity fellowship Brazil (CNPq) (301834/2016–4).