Main

Terminal deletions of the long arm of chromosome 4 are associated with a recognizable phenotype consisting of cleft palate, dysmorphic facial features (flat round face, micrognathia, short nose, apparent hypertelorism), upper extremity malformations, and growth and mental retardation. Over the past 30 years, 50 cases have been reported, 23 (41%) of which have had a cardiovascular malformation (CVM).114 Compared with the 10% frequency in the Baltimore-Washington Infant Study, a case-control study of liveborn infants with a confirmed CVM in 1981–1989,14 right ventricular outflow tract obstructive defects are more common (35%) in patients with 4q deletion.

Unlike interstitial deletions of 4q, which exhibit a more variable phenotype, terminal 4q deletions demonstrate a correlation between clinical phenotype severity and chromosomal breakpoint.1,14,15 In particular, there appears to be a karyotype-phenotype correlation because CVM has been noted in almost two thirds of patients with the terminal deletions at 4q31, whereas only about a quarter of patients with the more distal deletions at 4q34 or q35 have CVM. It has been postulated that genes distal to 4q34 may play a critical role in producing a CVM,14 though the molecular basis is unknown.

Recently, the dHAND gene, a basic helix-loop-helix transcription factor, was mapped to 4q33.16dHAND is expressed in the heart, aortic sac, and neural crest–derived tissue. A targeted deletion of dHAND in mouse embryos resulted in embryonic lethality due to right-sided heart failure and aortic sac malformation.17,18 The expression pattern and phenotypes in mutant dHAND mice are consistent with the high frequency of CVMs in patients with terminal 4q deletions. The overrepresentation of right-sided CVMs is intriguing, which suggested that dHAND deficiency might cause the heart anomalies in 4q− syndrome and prompted us to investigate the correlation of dHAND deletion with CVMs in patients with terminal deletions involving 4q33-q34.

MATERIALS AND METHODS

We studied five patients (three new, and two reported previously14,19 with structural rearrangements of the long arm of chromosome 4 in which the breakpoint involved q33-q34) (Table 1). Routine cytogenetic analysis was performed on cultured lymphocytes treated with phytohemagglutinin as a mitogen. A BAC clone containing the whole human dHAND gene was isolated by standard procedures.20 The presence of all exons of dHAND in the BAC clone was confirmed by polymerase chain reaction (PCR) amplification and sequencing. Briefly, exons 1 and 2 of dHAND were amplified by PCR. The following primer pairs were used:

  • Pair 1, forward (ex1F11): ACGCTGGGGCGCGTGGAG

  •  Reverse (R10): GGCCAGCAGGTCCATGAGGTAGG

  • Pair 2, forward (F1): ACGTACCCGCCGACACCAAACTCT

  •  Reverse (RP1): CCACCGCCTGCCGCCCCCTGGTA

  • Pair 3, forward (ex2F1): CCTCCCCGCCGGCTAGGGTAGC

  •  Reverse (ex2R1): GCGCCTTGGCCCCTGCTCACTC

Table 1 Summary of patients with terminal deletion

PCR products were purified with QIAquick columns (Qiagen, Inc., Valencia, CA) following the manufacturer's manuals. Each exon was sequenced by using an ABI 377 automated sequencer (Applied Biosystems) with primer RP1, ex2F1, and the following five primers:

  • FP1: TGAAGCGCCGAGGCACCGCCAACC

  • FP3: CGCGGAGGGCGAAATGAGTCTGGT

  • EX1R11: GCCGCTGGCATACTCGGGGCTGTA

  • R9: TGGCCAGGCGCAGGGTCTTGATTT

  • EX2F2: GGGGCTGGGACTCCCCCATCGTAT.

Sequence comparison from the PCR fragments with the DNAstar and BLAST program indicated that the sequence was identical with the human dHAND sequence in GenBank (accession number: NM_021973). The BAC DNA containing the entire dHAND coding sequence was labeled with digoxigenin-11-dUTP as described by Zhao et al.21 Fluorescence in situ hybridization (FISH) was performed as described previously.22 Digoxigenin labeled probe was detected by using the Oncor Kit (Oncor, Gaithersburg, MD) according to the manufacturer's instructions. A minimum of 10 metaphase cells were examined for each specimen.

RESULTS

Metaphase karyotypes and FISH results are shown in Table 1. Patient 1, who inherited a der(4)t(4;13)(q33;q12) from her mother, was monosomic for 4q33→qter and monosomic for 13pter→q12. The remaining four patients were monosomic for a segment of chromosome 4 due to terminal deletion of 4q. FISH analysis revealed deletion of dHAND in Patients 1, 2, and 3 (Fig. 1). Valvar pulmonic stenosis was diagnosed in Patients 1 and 2. Patient 1 also had partial anomalous pulmonary venous return. No CVM was present in Patient 3. Although the breakpoint in Patient 2, described by Tsai el al,14 was assigned to 4q34.2 by GTG-banding, FISH revealed deletion of dHAND gene indicating cryptic rearrangement of 4q in this patient including the dHAND locus at 4q33. Further FISH studies are warranted to better characterize this rearrangement.

Fig. 1
figure 1

FISH analysis of Patients 1–3. BAC DNA containing the human dHAND gene was labeled with digoxigenin-11-dUTP.21 Hybridization signals were detected with antidigoxingenin coupled with rhodamine supplied in the Oncor Kit according to the manufacturer's instructions. Hybridization signal specific for dHAND is present on only one chromosome, number 4 (10 metaphase cells were examined for each specimen).

dHAND was not deleted in Patients 4 and 5 (data not shown), who were a father and daughter, respectively, both of whom had an apparently similar deletion of 4q with a breakpoint assigned to 4q34.2 by GTG-banding, distal to the dHAND gene (Fig. 2). Patient 5 was referred for prenatal diagnosis because of a positive serum profile for Down syndrome and a family history of two previous abnormal pregnancies. The previous abnormalities included a fetus with a diaphragmatic hernia and a fetus with anencephaly. No cytogenetic studies were preformed on the first infant; the fetus with anencephaly had normal chromosomes. Analysis using standard cytogenetic methods revealed a small, apparently terminal deletion of chromosome 4 [46,XX,del(4)(q34.2)]. The pregnancy continued and was complicated by the development of intrauterine atrial flutter that was treated prenatally with procainamide and digoxin. At birth the female infant had no dysmorphic features. An echocardiogram revealed a patent ductus arteriosus with no other structural defects. The postnatal course was complicated by the persistent atrial flutter that required cardioversion twice. The infant required increasing respiratory support and she became asystolic and unresponsive to resuscitation. The immediate cause of death was cardiac tamponade due to pneumopericardium post mechanical ventilation. An autopsy detected a small atrial septal defect (1.0 × 0.5 cm) and no other unexpected findings.

Fig. 2
figure 2

GTG-banded chromosome 4 homologs of Patients 4 and 5 are shown in A and C, respectively. B and D: FISH on metaphase chromosomes of Patients 4 and 5, respectively, with CEP4 (at centromere) and TelVysion 4q (at 4q35) probes (Vysis Inc., Downers Grove, IL) performed according to the manufacturer's recommendations. Hybridization signals specific to the CEP4 probe are present on both chromosome 4 homologs. Hybridization signal specific to the TelVysion 4q probe is present on only one chromosome, number 4 (10 metaphase cells were examined per specimen).

The father is physically and intellectually normal and has had a normal echocardiogram. Surprisingly, his chromosome studies revealed that he carried a chromosome 4 that appeared identical with that found in his daughter. In search of a cryptic translocation, FISH was performed with whole chromosome painting probe specific for chromosome 4. Both chromosome 4 homologs painted along their entire length and no other region of hybridization were detected. The terminal deletion was further confirmed by FISH with a 4q telomeric probe (Fig. 2)C.

DISCUSSION

CVMs are very common in 4q− syndrome with a high frequency of right ventricular outflow tract obstruction defects. The dHAND gene is highly expressed in the right side of a developing heart,18 suggesting that deletions of dHAND gene might cause CVM in this syndrome. To examine the association of dHAND haploinsufficiency with congenital heart malformation in 4q− syndrome, we performed FISH with a human dHAND genomic probe on five patients with partial monosomy 4q with or without a CVM. Deletion of the dHAND gene was revealed by FISH in three patients. Among these three patients, two (Patients 1 and 2) have a CVM and one (Patient 3) has no CVM as reported previously.6,22,23 Of the two patients (Patients 4 and 5) without deletion of the dHAND gene, one (Patient 5) had an atrial septal defect found at autopsy. These results indicate that dHAND is not the only factor causing congenital cardiac defects and that other genes distant to the dHAND gene are responsible for CVM, or factor(s) other than dHAND gene deletion might contribute to the congenital cardiac defects.

dHAND null-mutant mice show heart failure, absence of right-sided heart development, dilated aortic sac, and failure to form normal aortic arteries,17,18,23 suggesting that dHAND is required for formation of the right ventricle of the heart. However, no CVMs were found in mice carrying a hemizygous mutation, indicating that haploinsufficiency of dHAND is not necessarily a cause for CVMs in 4q− syndrome.

Our data cannot exclude the possibility of an incomplete penetrance in dHAND deletion in CVM and the genetic background in this condition. Searching for the mutation of the dHAND gene in the right ventricular outflow defect will provide direct evidence. Thus far, we have found no mutation of the dHAND gene in patients with nonsyndromic right ventricular outflow trunk defects.

Another interesting finding in this study is that Patient 4, with a terminal deletion of 4q by karyotype analysis, confirmed by FISH using telomeric probe, is physically and intellectually normal and has had a normal echocardiogram. He did not receive medical attention until his daughter was found to have an abnormal karyotype. Cryptic translocation was ruled out by FISH with whole chromosome painting probe specific for chromosome 4. Both chromosome 4 homologs painted along their entire length and no other region of hybridization was detected.

Finally, the breakpoint in Patient 2 was assigned by GTG-banding to 4q34.2, distal to the dHAND locus; however, FISH revealed deletion of the dHAND gene, indicating a more complex rearrangement of 4q in this patient. Therefore, FISH proves to be an invaluable technique for elucidating cryptic or complex chromosome rearrangements.