Congenital heart diseases (CHDs), which may have a significant impact on cardiac structure, function, or both, are the most common defects in liveborn children (1). Affecting the heart and/or great vessels, CHDs can be isolated or identified within a syndromic presentation; they are highly diverse, ranging from asymptomatic to fatal. Multiple genetic factors have been implicated in playing a role in CHDs' complex etiology (2).

Most of the known CHD causative genes encode transcription factors (TF) that assist in regulating the cardiac embryogenesis, such as NKX2-5, GATA5, TBX20, TBX1, GATA4, GATA6, TBX5, MEF2C, HAND1, NR2F2, and HAND2 (3). Among the identified TF genes are the T-box gene family, which encodes proteins notably harboring a well-conserved T-box domain which assists in DNA binding (4). One of the T-box genes is TBX5 (OMIM ID: 601620) that is predominantly expressed in the heart and the forelimbs (5). Pathogenic variants in TBX5 have been associated with Holt-Oram syndrome (HOS; OMIM ID: 142900), which is characterized by congenital anomalies in both the heart and the upper limbs, in line with the TBX5 protein expression (6). Interestingly, variants in TBX5 have also been reported to cause non-syndromic CHD (7,8). Previous observations showed that the malformations of HOS were distinctively heterogenous even among individuals sharing the same genetic variant (9–13). A wide phenotypic spectrum of septal defects, conduction abnormalities, and tetralogy of Fallot has been frequently linked to the bulk of the TBX5-associated heart disorders (14). However, the clinical association between HOS and total anomalous pulmonary venous return (TAPVR) has rarely been described (15).

TAPVR makes up 1–3% of the total CHD cases, with an incidence of around 7 per 100,000 live births (16). TAPVR describes the improper connection of the pulmonary veins to the right atrium or to the systemic venous system rather than the normal connection to the left atrium (17). It can also be anatomically sub-divided based on the level of the venous connection anomaly into 4 subtypes (supracardiac, infracardiac, cardiac, and mixed types) (18). To our knowledge, all the studies that claimed the association between HOS and TAPVR were only based on clinical diagnosis-no genetic testing was performed in the previous literature to confirm that TBX5 was the implicated gene (15,19,20). Furthermore, to date, no case of mixed-type TAPVR has been reported in patients with HOS.

Over the past few years, whole-exome sequencing (WES) has been successfully implemented to uncover the underlying molecular etiology of multiple diseases. including CHD (21,22). Consequently, an accurate diagnosis of certain cases was achieved based on the identified genetic findings, which may inform families and influence patient care (23).

We present a multi-generation family affected by autosomal dominant (AD)-CHD with intrafamilial variability of specific manifestations and severity. Interestingly, the proband had a rare lethal mixed-type TAPVR. Our genetic investigation identified a hereditary disease-causing variant (DCV) in the T-box domain of TBX5. Consequently, the identified genetic findings prompted additional phenotyping which modified the diagnosis from non-syndromic CHD to HOS. Noteworthy, this is the first time TBX5 has been associated with TAPVR, specifically, mixed-type TAPVR, in patients with or without HOS. Also, we aimed to augment the previously reported corpus of genetic knowledge related to null-variants in the T-box domain of TBX5 to underpin the phenotypic intra- and interfamilial variable expressivity. Finally, we intended to illustrate the impact of harboring the variant on the protein's structure and function.

The TBX5 gene encodes for a transcriptional factor protein belonging to the T-box protein family and plays a major role in regulating the early cardiac and upper limb developmental processes (34). Interestingly, TBX5 interacts with other ‘cardiac-critical’ transcriptional factors, such as NKX2-5 and GATA4, which both also help in the early stages of the heart development pathways (35). Thus, genetic variations in the TBX5 can lead to both cardiac and/or skeletal defects (36). Variants in TBX5 cause HOS, also known as the heart-hand syndrome, which is inherited as an AD trait (6).

In this study, we identified the underlying genetic etiology behind a family with AD-CHDs in which the index case presented with mixed-type TAPVR while her mother had ASD. After conducting a trio-based WES analysis, we identified a maternally inherited pathogenic DCV in TBX5. Notably, prior to our investigation, the proband's mother (III-2), who suffered from ASD along with triphalangeal thumb, had never been diagnosed with HOS. The genetic finding of the DCV in TBX5 triggered us to perform a thorough review of her medical records and consequently reached her proper diagnosis with HOS.

LoF variants in TBX5, such as the pathogenic nonsense variant (c.577G>T; p.Gly193*) identified here, are established to be disease-causing in HOS. The resultant protein is predicted to harbor a PTC and probably would yield an aberrantly truncated protein (only 193 out of 518 TBX5′s amino acid residues might be translated, representing nearly 35% of the total protein's length). The transcribed TBX5 mRNA harboring the PTC might instead undergo degradation through the NMD process, resulting in TBX5 haploinsufficiency (37). Previous functional in vitro analysis for this DCV showed that the mutated TBX5 protein failed to activate its targeted genes (MYH6, and NPPA), and no synergistic transactivation between the mutated protein and other transcription factors (NKX2-5, and GATA4) was detected. The designated pathways, which are considered essential for cardiac development, could be nullified in embryos harboring this DCV and probably lead to the consequent heart defect (32). By utilizing protein modeling tools, we studied the impact of harboring this variant (p.Gly193*) on the TBX5′s structure and function. The resultant TBX5 structure formed a destabilized dimer due to a significant reduction in the number of the non-covalent bonds, and its ability to bind DNA might be subsequently deteriorated.

Our findings illustrate that at the intrafamilial level, the elucidated CHD ranged from a simple cardiac manifestation (ASD) to a complex, ultimately lethal one (mixed-type TAPVR). Furthermore, the same DCV (c.577G>T; p.Gly193*) has been previously reported in a Chinese Han family variably affected by ASD and/or ventricular septal defect (VSD), bicuspid aortic valve (BA), and atrial fibrillation (AF) (32). Hence, variable expressivity of the cardiac defects accompanied by this specific DCV can be observed at both the intrafamilial and interfamilial levels, and it is widely ranging in severity, structural deficits, and functional compromise.

The nonsense DCV (c.577G>T; p.Gly193*) is located in the TBX5′s T-box domain. Although pathogenic DCVs have been identified throughout the entire gene, they appear to cluster within this domain in particular (4). The T-box domain is a highly conserved domain across various species; it is also essential for interactions with other transcription factors and DNA binding (38). Therefore, we reviewed the reported clinical effects of the T-box's null variants on the expressivity of cardiac and skeletal malformations (Table II). We noticed that the cardiac defects were mainly associated with limb defects (and thus HOS) and ranged from isolated cases of septal defects (most common) to complex cases presenting with a combination of multiple cardiac manifestations, left ventricular noncompaction cardiomyopathy, conductive heart failure, and even valve anomalies. Nevertheless, this is the first time TBX5 has been discovered to cause mixed-type TAPVR.

The associated skeletal deformities accompanied by the CHDs in HOS were mainly described in the hand and predominantly in the preaxial radial ray (Table II). Similarly, both the deceased proband (IV-1) and her mother (III-2) presented with a triphalangeal thumb, which has been frequently reported in the literature as a manifestation of HOS (Table II). Other skeletal malformations were reported to a lesser extent in the postaxial region and the other upper extremities (forearms, arms, and shoulder complex), but were rarely ascribed to the lower extremities (feet, hip, etc.), as shown in Table II.

While HOS typically features both cardiac and limb deformities, in certain instances, the severity of the accompanying malformations tends to be more pronounced either in the heart or in the limbs of the same patient. Moreover, a wide range of accompanying anomalies (whether they were cardiac or skeletal) seem to be variably expressed even among patients harboring the same DCV (Table II). Several models propose that the disease-causing mechanism and the severity levels can be attributed to disturbances in TBX5′s binding to downstream protein-partners or recognizing targeted motifs that are vital to the embryo's heart and/or skeletal development, such as NKX2-5, GATA4, GATA6, TBX20, and MEF2C (38).

Although Table II shows a spectrum of the composite cardiac defects, we did not notice any correlation between the type or the location of the null-variant within the T-box domain and the consequent phenotypic severity. Broadly speaking, early studies suggested that type (missense vs. null variants) and/or the location of the DCVs could influence the severity of HOS's cardiac and skeletal manifestations (39). Nonetheless, these observations were based on a small number of cases, and inconsistencies were observed in the later-described ones (36,40). A recent study by Vanlerberghe et al, conducted a phenotype-genotype correlation on the largest molecular investigation of 78 HOS cases and found out that the null-variants were associated with less-severe cardiac manifestations when compared with the missense variants (9). Additionally, no relation was established between the type of the molecular variation (null-variants vs. missense) and the severity of the skeletal anomalies (9). Ultimately, as demonstrated in the family we report, variability is such to prevent predictions of phenotypic consequences based on genotype (40).

The TAPVR subtype identified here revealed both supracardiac and infracardiac venous anomalies, resulting in a mixed-typed classification of the TAPVR phenotype. The 3D CT scan of the proband's heart (IV-) showed that both the RPVC and LPVC failed to connect to the left atrium and instead drained directly into the SVC and IVC, respectively. The mixed-type TAPVR is rare amongst TAPVR cases-accounting for around 5% of the total occurrences (41). Cardiac, supracardiac, and infracardiac subtypes of TAPVR have been previously reported with HOS (15,19,20,42). To the best of our knowledge, mixed-type TAPVR has never been reported in HOS. Hence, this is the first genetic investigation to identify an HOS case with mixed-type TAPVR.

In summary, we utilized the trio-based WES approach to identify a pathogenic nonsense DCV in a family with hereditary CHDs. Also, we showed the importance of genetic testing in reaching the proper diagnosis-our findings aided the correct diagnosis of the proband's and her mother's cases with HOS. We explored the broad phenotypic spectrum of the reported null-variants within the T-box domain of TBX5. Importantly, we reported a patient presenting with mixed-type TAPVR; to our knowledge, this is the first study documenting the connection between the mixed-type TAPVR and HOS. Additionally, this is the first molecular investigation to ever associate TAPVR with HOS.