Editor—Congenital heart disease occurs commonly. One form, heterotaxy, accounts for approximately 3-4% of the total incidence and has a mortality rate approaching 45%.1Given that the diagnosis is based on the discordance of the left-right (LR) sidedness between the abdominal viscera and atria,2heterotaxy describes a group of malformations arising from the abnormal development of LR asymmetry.3

In familial cases one can find subjects with complete, mirror image reversal of normal LR anatomy (situs inversus), and others who manifest the hallmark visceroatrial discordance as well as other laterality malformations (sometimes collectively called situs ambiguus). Moreover other family members with normal LR anatomy (situs solitus) are obligate disease gene carriers by virtue of their pedigree position.

Many genes have been implicated in normal and abnormal LR axis development among non-human vertebrates.4 Knowledge remains sparse, however, regarding the molecular genetics of human LR malformations. Positional cloning identified a gene,ZIC3, on chromosome Xq24-27.1, in which mutations have been found among one sporadic and six familial cases of LR axis malformations.5 A few mutations have also been found in LEFTYA and in the activin receptor type IIB gene (ACVR2B), identified on the basis of their homology to the corresponding genes known to cause laterality defects in the mouse.6 7

Here we describe a family in which LR malformations segregate across five generations. Although male to male transmission has not occurred, males and females appear to be affected similarly, and linkage analysis has excluded a disease locus on the X chromosome (see below). Both situs inversus and situs ambiguus are found in seven affected subjects and pedigree position implicates four apparently normal subjects as obligate gene carriers. These observations strongly support a model of autosomal dominant inheritance with reduced penetrance. The pedigree comprising 36 subjects is illustrated in fig 1

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Pedigree of family with laterality defects and 6p21 marker haplotypes. Disease haplotype is boxed. Symbols for situs inversus, situs ambiguus, isolated cardiac defects, and unaffected disease gene carriers are as indicated in symbol definitions. Alleles in parentheses in missing subjects were inferred from their children's marker genotypes.

Seven subjects in five generations manifest laterality defects of multiple organs (fig 1). Of these, four are situs inversus (II.2, III.7, III.10, and IV.6), and three are situs ambiguus (IV.8, V.1, and V.4). There is considerable variability of expression in the situs ambiguus group. IV.8 has mirror image reversal of the heart and of the colon but normal position of the liver, stomach, and spleen, while complex heart malformations were identified in the other two, leading either to prenatal termination (V.1) or surgery (V.4). II.4, III.1, III.6, and IV.4 are obligate disease gene carriers by virtue of their pedigree position but without apparent LR abnormalities. III.9 and V.2 have isolated cardiac defects without any other LR abnormality. The malformation observed in III.9, ventricular inversion in combination with transposition of the great arteries, is usually classified with heterotaxy under the common aetiology of “abnormal looping” defects. Therefore, III.9 was scored as affected in the linkage analysis, while V.2, who showed hypoplastic left heart syndrome (HLHS), which has not been linked embryologically to the cardiac looping defects, was scored as having an unknown disease status. All subjects manifesting laterality defects but III.10, who was unavailable, were included in the linkage analysis and scored as affected. In all subjects, disease status phenotype was assigned before marker genotyping.

Informed consent was obtained from patients participating in this study, which was approved by the Institutional Board at Baylor College of Medicine. Genomic DNA was extracted from whole blood or cell lines (lymphoblast or fibroblast) with the Puregene DNA Isolation Kit (Gentra Systems) according to the manufacturer's protocol. DNA from paraffin embedded tissue was extracted as previously described.8Amplifications were performed on HYBAID Omnigene thermocyclers under standard conditions.

The initial screening was performed at the Center for Medical Genetics in Marshfield, WI, using marker screening set 6, consisting of short tandem repeat markers with an average heterozygosity of 0.76 and an average sex equal spacing between markers of 10.0 cM.

A total of 346 markers (333 autosomal and 13 X linked) were used for the genome wide screening. Conventional two point and multipoint linkage analysis between the trait locus and the polymorphic markers was carried out using VITESSE version 1.0.9 For the initial screening, a two point lod score analysis was performed using the affecteds only approach with an autosomal or X linked dominant mode of inheritance. Subjects with confirmed LR abnormalities were scored as affected, married in as unaffected, and the remaining subjects were considered as having an unknown disease phenotype. All marker allele frequencies were equalised and the disease gene frequency was set to 0.00005.

For fine mapping using additional markers, linkage analysis was performed using disease gene penetrance and marker allele frequencies estimated on the basis of our family data using the FASTLINK 3.0 version of ILINK.10 11 Penetrance was estimated on the basis of disease phenotype status only without including any marker genotype information, and found to be equal to 0.5 for heterozygotes. All unaffected subjects except V.2 were designated as such in the fine mapping linkage analysis. Multipoint linkage analysis was carried out by means of the sliding window technique,12 using three markers at a time, with genetic distances obtained from the maps developed at the Center for Medical Genetics in Marshfield (WI).

Results of the genome wide screen showed eight markers, one on chromosome 5 and seven on chromosome 6, with a two point lod score between 1 and 2, and two on chromosome 6 with a lod score higher than 2. Three of these 10 markers, two on chromosome 6 and one on chromosome 5, were located in regions that were excluded by negative lod scores obtained from fine mapping after typing additional markers (data not shown). Of the remaining seven markers, five were located on chromosome 6p and two on chromosome 6q (table 1).

As shown in table 1, fine mapping using additional markers in these regions and a reduced penetrance model continued to yield positive lod scores, which increased from 2.33 to 2.67 for D6S426 on 6p21 but decreased from 2.27 to 1.81 for D6S305 on 6q25. Furthermore, multipoint linkage analysis yielded maximum multipoint lod scores of 2.95 at theta=0 from D6S426 on 6p21 (fig 2) and of 2.05 at theta=0 from D6S305 on 6q25 (data not shown). The use of different penetrance values did not significantly modify these results and consistently resulted in increased lod scores for 6p21 and reduced lod scores for 6q25 with respect to the two point analysis.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Multipoint linkage analysis for chromosome 6p21 markers.

Among the 6p21 markers, obligate recombination events were observed for D6S105 in V.1 and for D6S1960 in IV.6 and IV.8 (fig 1). These two markers, which are located 32 cM apart, thus define the limits of the critical interval for the disease locus.

A genome wide search for linkage in this family has thus tentatively identified a new locus for LR axis malformations on chromosome 6p21. The highest two point lod score obtained was 2.67 at theta=0 with marker D6S426, and a maximum lod score of 2.95 was obtained at the same location by multipoint analysis. The first flanking recombinant markers telomeric and centromeric to this region, respectively, are D6S105 and D6S1960.

Results from simulation analysis carried out with the SLINK program13 using the same disease model of the actual linkage analysis and one marker with four equally frequent alleles at a genetic distance of 1 cM from the disease locus predicted that a maximum lod score of 3.12 could be obtained in this pedigree. However, this theoretical maximum would be attainable only in the event that none of the unaffected subjects carries the susceptibility haplotype. In contrast, six out of 13 at risk subjects (IV.9, IV.10, IV.11, IV.13, V.5, and V.6) have inherited the disease haplotype and thus appear to be non-penetrant carriers (fig 1). This is in agreement with the penetrance of 50% estimated in our pedigree before marker genotyping on the basis of disease status information only.

Marker D6S305 on chromosome 6q25 gave a lod score of 2.27 in the affecteds only two point analysis. Multipoint analysis of the entire family using a reduced penetrance model consistently decreased the lod scores for this region with a maximum of 2.05. Although a negative lod score of less than −2 is generally deemed necessary formally to exclude a region in linkage analysis, these findings significantly reduced the odds of linkage to 6q25. False positive lod scores of this order of magnitude are not uncommon in genome wide screens. However, while we conclude that there is a high likelihood of one or more genes on chromosome 6p21 influencing LR axis development, we cannot formally exclude that a second disease susceptibility gene is located on 6q25. Investigation of these regions in other independent families with the same phenotype will be important in order to provide a definite confirmation of our findings.

The present study cannot determine the precise mechanism by which this locus contributes to the phenotype, particularly given the presence of several non-penetrant carriers. Chance alone may be at work, or the locus may be necessary but not sufficient to cause disease in the absence of modifying loci elsewhere in the genome. It is possible also that there are two or more linked susceptibility loci within this large interval in 6p.

Several genes that affect LR axis development have been identified in the mouse.4 Where human homologues have been identified, none maps to 6p, and none of the remaining murine genes maps to regions homologous to 6p.14 A search for candidate genes among those transcripts localised to the critical region shows several that might be involved, but no immediately compelling candidates. Identification of the susceptibility gene (or genes) in this interval awaits further understanding of the genetic pathway in model systems or the future discovery of suitable candidate genes in this region of the human genome.