Comprehensive analysis of copy number variations in congenital heart defects Tunisian patients: chromosomal microarray analysis insights

Abstract

Background

Congenital heart defects represent a major global health burden, affecting nearly one million newborns annually. Identifying the underlying genetic causes is essential for improved diagnosis, patient management, and genetic counseling.

Patients and methods

We conducted a cytogenetic study integrating conventional karyotyping, fluorescence in situ hybridization (FISH), and chromosomal microarray analysis (CMA 44 K) in 20 Tunisian patients presenting syndromic CHDs and referred to our Genetics Department.

Results and discussion

CMA identified pathogenic copy number variations in four patients. These included an inherited 11 Mb deletion at 9p24.2 together with a 10 Mb duplication of 20pter; a de novo 1.2 Mb deletion at 15q26.2 with an 11 Mb duplication at 2q36.3; a de novo 113 kb deletion at 17q21.32; and a de novo 48 Mb duplication at 8q22. Several CNVs overlapped known deletion/duplication syndromes, some with previously infrequent cardiac involvement. Genotype–phenotype correlations enabled prioritization of CHD relevant genes including DOCK8, HTR2B, KANSL1, ZFPM2, and TRPS1, whose dosage sensitivity and interactions with cardiac developmental pathways may contribute to the observed phenotypes.

Conclusion

This study reinforces the clinical utility of CMA in detecting cryptic chromosomal abnormalities in syndromic CHD. The identified CNVs and gene candidates offer new insights into CHD genetic architecture and support CMA as a first-tier diagnostic tool. These findings highlight the contribution of rare, pathogenic CNVs in syndromic cases and suggest their integration into refined diagnostic and counseling strategies. Further functional studies are necessary to elucidate the roles of these candidates in cardiogenesis.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13039-026-00749-4.

Introduction

Cardiac morphogenesis requires tightly regulated and sequential developmental events. Disruption at any stage can result in congenital heart defects, among the most common birth anomalies, with a global incidence of approximately 9 per 1,000 live births [1] and contributing significantly to infant mortality [2]. While early embryonic disruptions lead to more severe malformations, the genetic mechanisms driving these defects remain insufficiently characterized; only ~ 20% of genetic causes are currently known [3]. CHD etiology is widely considered multifactorial, involving interactions between genetic and environmental factors.

Chromosomal aberrations contribute significantly to CHD pathogenesis. Conventional karyotyping identifies anomalies in 3–18% of affected patients [4], with higher detection rates when using technologies such as FISH and chromosomal microarray analysis. CMA enables genome-wide detection of CNVs at resolutions far exceeding standard karyotyping and has become a recommended first-line test in syndromic CHDs [5]. Pathogenic CNVs are detected in ~ 3–25% of syndromic CHDs and 3–10% of non-syndromic forms [6]. CNV characterization contributes to understanding CHD biology and improves clinical management and genetic counseling.

Single-gene disorders account for up to 34% of CHD cases, often as part of Mendelian syndromes [7]. Numerous genes, including NKX2-5, GATA4, TBX1, and TBX5, play critical roles in heart development, although the precise mechanisms remain incompletely understood.

This study aimed to evaluate chromosomal abnormalities in Tunisian infants with syndromic CHD using conventional and molecular cytogenetics, identify candidate genes within pathogenic CNVs, and contribute to expanding CHD-related gene panels relevant for diagnosis and precision medicine, particularly for North African populations where genomic tools remain limited.

Materials and methods

Patients

From January 2014 to the present, 250 patients with syndromic or isolated CHDs were referred to our department for cytogenetic evaluation. Ages ranged from the neonatal period to 41 years. For this study, we selected 20 patients with syndromic CHD and complex phenotypes for CMA analysis. All underwent karyotype, FISH, and CMA when indicated. Parental studies were performed whenever possible to determine inheritance.

Karyotype

For all patients included in this study, a conventional R-band karyotype at a 450-band resolution was performed. In the case of anomalies seen on karyotypes or detected further by array CGH. A parental karyotype was performed to complete the inheritance form. According to a standard protocol, metaphase chromosome spreads were prepared from phytohemagglutinin-stimulated peripheral blood lymphocytes. Cell cultures were incubated For 72 h. At least 25 mitoses were studied for each sample using Cytovision^® karyotyping software version 4.0.

FISH analysis

FISH was carried out on metaphase spread chromosomes with commercial probes. The patient’s DNA and the probe were denatured at 75 °C, which will allow in a second step, the specific hybridization of the probe with its target sequence. After overnight hybridization at 37 °C and washing, chromosomes were counterstained with a 4,6-diamino-2-phenylindole (DAPI) and hybridizations were observed using an Axioskop Zeiss^® fluorescent microscope.

Chromosomal microarray analysis CMA

CMA was performed using Agilent^® CGH 44 K arrays according to manufacturer instructions (Feature Extraction 9.1, CGH Analytics 4.5). CNVs were classified according to the ACMG/ClinGen 2019 guidelines using a quantitative, point-based scoring system. Evidence including gene content, dosage sensitivity, overlap with known pathogenic or benign regions, inheritance, and CNV size was assigned numerical scores. The final classification was based on the cumulative score, defining CNVs as pathogenic (≥ 0.99), likely pathogenic (0.90–0.98), variants of uncertain significance (− 0.89 to 0.89), likely benign (− 0.90 to − 0.98), or benign (≤ − 0.99).

Bioinformatic interpretation used the UCSC Genome Browser (https://genome.ucsc.edu/) (hg18), DGV (DGV: http://dgv.tcag.ca/dgv/app/home), OMIM (OMIM: https://omim.org/), DECIPHER, and Franklin by Genoox (https://franklin.genoox.com/clinical-db/home). Candidate genes were prioritized based on known CHD involvement, pLI/HI scores, dosage sensitivity, and expression during cardiogenesis. CNV classification followed the semi-quantitative scoring system published in 2014 [8].

Results

The study identified infants with CHDs, representing a birth incidence of 2.2% of all cases diagnosed in the laboratory during the study period (9 years). This percentage reflects the proportion of CHD cases among all patients referred to genetic testing, rather than a birth incidence population.

Among the 250 CHD patients, 88.1% had syndromic associations. The most common cardiac defects were atrioventricular septal defects (90%), Tetralogy of Fallot (7.3%), and complex malformations (2.7%). Developmental delay was present in 70% and dysmorphic features or growth delay in 10%.

Twenty syndromic CHD patients (Table S1) were analyzed by chromosomal microarray analysis (Table 1), and pathogenic CNVs were identified in four patients (20%). Patient 1 carried an unbalanced rearrangement resulting from a maternal reciprocal translocation, with a deletion on chromosome 9 and a duplication on chromosome 20 corresponding to the inherited derivatives der(9) and der(20) (Fig. 1). This patient presented atrial and ventricular septal defects (ASD, VSD), severe psychomotor delay with absent sitting acquisition, axial hypotonia, fetal macrosomia, cleft lip and palate, and facial dysmorphism. CNVs encompassed genes implicated in cardiac development and neurodevelopment mainly DMRT 1,2 and 3, KANK1, VLDLR, and DOCK8.

Table 1.

Clinical and genetic characteristics of patients with detected copy number variations by chromosomal microarray analysis

Patients	CHD	Clinical features	Other congenital malformations	Karyotype and Fish Results	Chromosomal abnormality	Region	Genomic positions	De novo/inherited	Number of genes in the region	Candidates genes
P1	ASD VSD	Psychomotor retardation (sitting position not acquired) Axial hypotonia and fetal macrosomia	cleft lip and palate, Facial dysmorphia	46, XX, ish (22q11.2) (Tuple1+) ish der (9) (IKBI-2009B+) der20(PPBI-402366++) mat	Deletion associated with duplication	9p24.2 (11Mb) 20pter (10 MB)	204,367 − 11,449,399 74,057 − 9,770,641	Maternal inherited	49 genes with 37 coding genes 137 genes with 99 coding genes	DOCK8
P2	PDA	Psychomotor retardation, Growth delay	Facial dysmorphia	46, XX ish del (15) (Subtel 15qx1) dup (2)(Subtel 2qx3)	Deletion associated with duplication	15q26.2 (1,2 MB) 2q36.3 (11 MB)	98,886,631 − 100,729,914 231,022,582 − 242,900,037	de novo	16 genes with 9 coding genes 167 genes with 119 coding genes	HTR2B
P3	TOF	Psychomotor retardation, hypocalcemia, hypothyroidism	Facial dysmorphia, convulsive seizures, non-palpable testes	46, XY.ish dup (8) (q22.13) (WCP8+)	Duplication	8q32.2 (48 Mb)	94,619,246 − 143,333,332	De novo	282 genes with 160 coding genes	TRPS1
P4	ASD VSD	Absence of language, autism	Facial dysmorphia,	46 XY; t (1;2), (p31; p22) pat	Deletion	17q21.32 (113kb)	41,587,723 − 41,700,815	De novo	One gene region	KANSL1

Bold values indicate the size of the CNV in Megabases

ASD: atrial Septal Defect, VSD: Ventricular Septal Defect, TOF: Tetralogy of Fallot, PDA: Persistence of the ductus arteriosus

Fig. 1.

A : CGH array profile of patient 1(P1) showing a deletion of 11 MB between oligonucleotides 204,367 and 11,449,399 bp in 9p24.2, B: CGH array profile of patient N°1 (P1) showing a duplication of the 10 MB between oligonucleotides 74,057 and 9,770,641 bp in 20pter, C: FISH image showing 20p duplication in P1, D: FISH image showing maternal translocation between chromosomes 9 and 20

Patient 2 exhibited a deletion at 15q26 associated with a duplication at 2q3 (Fig. 2). This CNV occurred de novo and was associated with psychomotor retardation, growth delay, and facial dysmorphism. The deleted region at 15q26 includes genes previously linked to growth impairment and congenital anomalies.

Fig. 2.

A : CGH array profile of patient N°2 (P2) showing a deletion of almost 1.2 MB between oligonucleotides 98,886,631 and 100,729,914 bp at 15q26.2, B : CGH array profile of patient N°2 showing a duplication of 11 Mb oligonucleotides 231,022,582 and 242,900,037 bp in 2q36.3, C: FISH image for confirmation of 15q26.2 deletion, D: FISH image for confirmation of 2q36.2 duplication

Patient 3 harbored an isolated de novo duplication at 8q22 (Fig. 3). Clinically, this patient showed psychomotor retardation, hypocalcemia, hypothyroidism, facial dysmorphism, convulsive seizures, and non-palpable tests. The duplicated region contains genes involved in neurodevelopmental and endocrine pathways such as COH1, TRPS1 and EXT1.

Fig. 3.

A: CGH array profile of patient N°3 (P3) showing an 8q duplication of approximately 48 MB between oligonucleotides 94,619,246 and 143,333,332, B: FISH image showing duplication of chromosome 8 by the WCP8 paint probe

Patient 4 carried a de novo deletion at 17q12.2 (Fig. 4), presenting with ASD, VSD, absence of language development, autism spectrum disorder, and facial dysmorphism. This region includes dosage-sensitive genes known to be associated with neurodevelopmental disorders and congenital malformations.

Fig. 4.

CGH array profile of patient N°4 showing a deletion at 17q21.32 of approximately 113 KB between oligonucleotides 41,587,723 and 41,700,815

Whenever probes were available, CNVs were confirmed by FISH analysis; however, the 17q12.2 deletion could not be validated due to the unavailability of specific probes for this region.

Discussion

Normal fetal heart development is a critical process that involves a complex series of events. Any disruption in this process could result in congenital heart defects. Depending on their type and location, these malformations can be mild or severe. The severity of heart defects and associated symptoms can vary considerably depending on the underlying cause. Importantly, up to 25% to 30% of the CHD cases have other associated extracardiac manifestations, leading to a complicated diagnosis and management [9].

Understanding heart embryology and the development of congenital heart defects remain crucial for managing affected patients’ efficiency. This knowledge enables accurate diagnosis, appropriate treatment, and genetic counseling. Indeed, the establishment of genotype-phenotype correlation through the chromosomal anomalies established in this study has allowed the identification of CHD candidate genes. In the first case, the deletion in the 9p24.3 region reported here has been associated with 20 pter duplication. The 9p24.3 deletion encompasses about 11 MB and harbors several 37 coding genes among the 50 genes included. Interestingly, the DOCK8 (DEDICATOR OF CYTOKINESIS 8) [OMIM: 9p24.3] gene is known to be associated with autosomal recessive recurrent infection syndrome (Hyper-IgE Syndrome) [10]. Indeed, the DOCK8 gene is expressed in the placenta, lung, kidney, and pancreas, and at low levels in heart and brain tissues, and seems to be involved in the actin cytoskeleton regulation [11, 12] indicating that its effect on cardiac development may be indirect or mediated through regulatory mechanisms. The limited direct evidence linking DOCK8 mutations specifically to congenital heart defects seems to be remaining but its recurrent incidence in 9p24 rearrangement segregating mainly in a family with congenital heart defects, including pulmonary and aortic valvular and pulmonary artery stenosis through five generations emphasize its direct implication or involvement in its regulation [13] either by the DOCK8 gene or its coding or non-coding transcripts [14]. Although DOCK8 is not considered a haploinsufficient gene and heterozygous loss of function is unlikely to be sufficient to cause disease on its own, its recurrent involvement in 9p24.3 deletions reported in patients with congenital heart defects suggests a potential contributory role Alternatively, the cardiac phenotype observed in 9p24 rearrangements may result from the combined effects of regulatory disruption, involvement of non-coding transcripts, or dosage alteration of neighboring haploinsufficient genes, such as SMARCA2 or DMRT1, which have been implicated in developmental processes. Taken together, these findings support a broader pathogenic role of the 9p24 region in congenital heart defects rather than implicating DOCK8 as a single causative gene.

Otherwise, the combined 15q26.2 deletion and 2q36.3 duplication revealed no known CHD-associated genes in the deleted region. However, duplications involving the 2q36 region have been repeatedly associated with variable congenital anomalies, including cardiac malformations, with phenotypic variability largely attributed to differences in duplication size and breakpoint position. Interestingly, it has been previously [15] hypothesized that the critical region of the duplication responsible for the cardiac phenotype is the proximal areas of the 2q3 region [14]. The duplication detected here at 2q36.3: 242,900,037(Hg18: build 36) is close to the critical region previously reported. It includes the HTR2B gene, (5-HYDROXYTRYPTAMINE RECEPTOR 2B) [OMIM: 601122], encodes a serotonin receptor (SEROTONIN 5-HT-2B RECEPTOR), and proposed as a candidate gene for heart defects since it has been expressed in artery aorta and coronary (Franklin by Genoox database) and heart atrial appendage. However, its specific role in CHDs is not well understood. The HTR2B gene is by its central and peripheral physiological functions of serotonin, the 5HT-2 acts on blood vessel contraction and platelet shape changes, including the heart [16]. In addition, the inactivation of the HTR2B gene in mice leads to embryonic and neonatal death caused by cardiac malformations [17]. The mutant embryos showed a specific reduction in the expression levels of a receptor tyrosine kinase, ERBB2, leading to fetal lethality. ERBB2 is a key regulator of cardiomyocyte proliferation, differentiation, and cardiac morphogenesis, and its dysregulation has been implicated in congenital heart defects. These findings suggest that the 5-HT2B receptor exerts its effects during heart development at least partly through the ERBB2 signaling pathway, highlighting ERBB2 as a significant downstream effector in HTR2B-associated cardiac phenotypes. Moreover, an interstitial 48 MB duplication in 8q22.13 was identified. Even if interstitial 8q duplications are very rare, and only a few cases are reported and characterized by CMA. Most of them are only investigated by standard karyotype or FISH. Among the 280 genes involved in this region, 160 genes are protein coding. The TRPS1 (ZINC FINGER TRANSCRIPTION FACTOR) [OMIM: 604386] at 8q23.3 is a transcription factor acting in skeletal development pathways. Curiously, its mutations have been associated with CHDs. Indeed, TRPS1 mutations are linked to the autosomal dominant tricho-rhino-phalangeal syndrome, which is characterized by skeletal and craniofacial malformations [18] [19]. and in more recent studies, congenital cardiac abnormalities, including persistent foramen oval, persistent ductus arteriosus, aortic stenosis, and left heart failure [20]. However, the mechanisms by which the mutation leads to cardiac phenotype are still poorly understood. The same region also encompasses ZFPM2 8Zinc Finger Protein Member 2) [OMIM:603693], a gene previously associated with Tetralogy of Fallot (TOF), which corresponds to the patient’s cardiac defect in the present case. Interestingly, ZFMP2 knockout mice showed conotruncal defects, including abnormal cardiac looping and septation [21], and its mutations in humans and canines have been linked to TOF and tricuspid atresia which supports its role in cardiac development [22]. These findings, together with the presence of other CNVs in the present cohort affecting multiple genes, highlight that complex phenotypes may result from the combined effect of several dosage-sensitive genes within the affected regions.

The last case concerns a 113Kb deletion in 17q21.32, englobing the gene KANSL1 (AT8 REGULATORY NSL COMPLEX, SUBUNIT 1) [OMIM: 612452], a gene involved in chromatin remodeling.Indeed, this gene encodes a protein that belongs to the histone acetyltransferase (HAT) family and plays a role in chromatin modification [23]. It is expressed early in fetal life in most tissues and organs of the body, leading to a serious pleotropic effect by the alteration of the development and functioning of many organs [24]. Indeed, the loss of one copy of this gene affects the development, inducing several abnormalities and is responsible for the known Koolen de Vries syndrome [OMIM: 610443], which is characterized by congenital malformations, developmental delay/intellectual disability, neonatal/childhood hypotonia, epilepsy, dysmorphisms, and behavioral features. Psychomotor developmental delay is noted in all individuals from an early age. Most people with KdVS have a mild to moderate intellectual disability. Other findings include speech and language delay (100%), epilepsy (~ 33%), congenital heart defects (25%−50%), renal and urologic anomalies (25%−50%), and cryptorchidism. The behavior in most cases is described as friendly, amiable, and cooperative [25].

Finally, the implementation of Chromosomal Microarray analysis provided valuable insights into cryptic chromosomal abnormalities. CMA is a powerful tool for studying the effect of alterations in several genes simultaneously, identifying genes associated with specific phenotypes, and facilitating the diagnosis of syndromic congenital heart defects.

Although chromosomal abnormalities were identified in only 20% of cases, leaving approximately 80% unexplained. However, further characterization using other high-throughput techniques remains valuable in both syndromic and isolated cases of CHD.

Given the complex interplay of genetic factors in syndromic congenital heart defects, it is conceivable that the candidate genes identified in our study, including HTR2B, DOCK8, KANSL1,ZFPM2 and TRPS1, may participate in interactions not only during cardiogenesis but also with genes implicated in various other organ development. These complex interactions may contribute to the phenotypic manifestations observed in syndromic congenital heart defects, which affect not only the heart but also other organs.

Nevertheless, despite the rapid expansion of next-generation sequencing technologies, including whole-genome sequencing capable of detecting both SNVs and CNVs, chromosomal microarray analysis remains a valuable first-line cytogenomic tool. In many regions, particularly across North Africa, the limited availability of high-throughput sequencing platforms, economic constraints, and the need for rapid and straightforward data interpretation make CMA a practical and cost-effective approach. Its robust performance for clinically relevant CNV detection continues to provide essential diagnostic yield, especially in settings where access to comprehensive NGS-based pipelines is still developing.

Conclusion

Chromosomal microarray analysis enabled the identification of clinically relevant CNVs in a subset of Tunisian patients with congenital heart defects, revealing dosage alterations in genes involved in cardiac and multisystem development. These findings underscore the genetic heterogeneity of CHDs and highlight the importance of genome-wide CNV screening for precise diagnosis and clinical management.

Although many cases remain unexplained, CMA provided essential diagnostic information and remains particularly valuable in regions with limited access to high-throughput sequencing technologies. In North Africa, its affordability, rapidity, and ease of interpretation support its continued use as a first-line cytogenomic tool. Integrating CMA with advanced sequencing approaches in the future will further increase diagnostic yield and contribute to a deeper understanding of the complex genetic basis of congenital heart defects.

Author contributions

Soumaya Mougou Zerelli: Conceptualization, Supervision, review & editing, medical consultation; Rim Khelifi: Experimental work, Analysis, Interpretation of data, Writing – original draft; Houcemeddine Othmane: helped in editing and review, Ayda Bennour, Leila Dardour, Wafa Slimani, Khouloud Rjiba: Experimentation, Partial analysis; Houda Ajmi, Amira Benzarti, Hamza Hadj Abdallah, Ahmed Rassass, Rim Kooli, Oussama Mghirbi, Molka Kammoun, Hela Ben Khelifa, Farouk Bahri, Mkaddem Hayet, Aouina Ammar, Sahbi Ghanmi, Jihene Mathlouthi, Hayet Ben Hamida, Zakia Habboul, Tarek Kemis, Haib Kharrat, Hechmi Ben Hammouda, Chaieb Jalel, Neila Hachicha, Mohamed Khelif, Nadia Hassine, Tej, Manel bellalah, Fatma Chouikh, Mejaouel Houssine, Abdallah Mahdhaoui, Chokri Kortas, Aida Guith, Faouzi Maatouk, Elyes Naffeti, Habib Soua: Patient referral, Clinical assessment; Moez Gribaa, Ali Saad: Resources. All authors read and approved the final version of the manuscript for publication.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Data availability

Data and materials are available from the corresponding author and are available upon request.

Declarations

We confirm that the study protocol is performed according to the relevant guidelines and regulations.

Ethics approval and consent to participate

The local Ethics Board of the University Teaching Hospital Farhat Hached approved the present study (CER:14-2022) and informed that written consent was obtained from the parents of all the children. Consent is an informed choice by the parent or legal guardian (either in writing or verbally). A separate consent form was completed for each child participant.

Consent for publication

The local Ethics Board approved the present study (CER:14-2022) and informed written and verbal consent was taken from the parents of all children for publication.

Competing interests

The authors declare no competing interests.