Ink injections showing pharyngeal arch arteries (PAAs) in wild-type (top), Tbx1+/- (middle) and Tbx1-/- (bottom) mutant mouse embryos. Wild-type embryos have three pairs of PAAs (3-6), Tbx1+/- embryos have small or absent 4th PAAs, and in Tbx1-/- mutants, PAAs 3-6 do not develop. Credit: Courtesy of E. Lindsay, Baylor College of Medicine, USA.

DiGeorge syndrome (DGS), a complex congenital disorder, is partly characterized by cardiovascular (CVS) defects, lack of parathyroid and thymus glands, and facial abnormalities. Despite the common embryological origins of affected tissues — many derive from the pharyngeal arches — elucidating the genetic basis of DGS has proved a challenge. This is because most DGS patients are heterozygous for a 3-Mb, chromosome 22 deletion, in 22q11, that encompasses many genes. The race to discover the genes that underlie DGS has been intense, as shown by four recent publications, three of which report that mutations in the T-box gene, Tbx1 , cause DGS-like CVS defects in mice. The fourth paper implicates another 22q11 gene, CRKL , in the disease.

Deletions that had been made in the region of mouse chromosome 16 (MMU16) that is homologous to the DGS region on 22q11 set the stage for the recent findings. In particular, two independently generated, partially overlapping deletions, Df(16)1 and Idd–Arvcf , defined a candidate region for the CVS defects. In their latest study, Lindsay and colleagues used Cre-loxP technology to engineer additional MMU16-deletion and -duplication alleles. They found that loss of a critical interval, encompassing ∼700 kb, caused fully penetrant CVS defects. A duplication of this interval rescued the CVS phenotype — this ruled out a role for the action of long-range regulatory elements in this phenotype and indicated that it was caused by a dosage-sensitive gene in this region. The authors then identified a 140-kb PAC that rescued the CVS phenotype caused by the Df(16)1 deletion. Of the four genes on this PAC, only Tbx1 was strongly expressed in mouse pharyngeal arches, and so Lindsay et al. knocked it out. The Tbx1+/− mutants had an identical DGS-like CVS phenotype to Df(16)1/+ mutants, and Tbx1−/Df(16)1 compound mutants fully reiterated the Tbx1−/− phenotype.

These and other findings strongly implicated TBX1 deficiency as the cause of CVS defects in DGS. Merscher et al. also generated a large MMU16 deletion, which caused DGS-like CVS defects in 50% of heterozygous mutants and parathryroid abnormalities in some animals. The failure of a duplication of the Idd–Arvcf interval to rescue this deletion phenotype further narrowed down the candidate interval, and a human BAC from this region partially rescued the deletion phenotype. This BAC contained TBX1. Confirmation of TBX1's role came when the authors knocked out Tbx1 in mice and generated mutants with DGS-like CVS defects. Jerome and Papaioannou also knocked out Tbx1 in a candidate-gene approach to studying DGS. They report that their Tbx1−/− mutants develop CVS defects and other DGS features, such as thymus, parathyroid and facial abnormalities. It is not known whether TBX1 mutations in humans underlie the whole DGS phenotype, but these abnormalities in Tbx1 nulls — as also reported by Lindsay et al. — indicate this should not be excluded.

But what of the other genes in the region? Guris et al. report that loss of another 22q11 homologue in mice, Crkol , causes post-migratory defects in neural crest cells, which contribute to many of the tissues affected in DGS — null mutants have CVS, parathyroid and thymus, and facial defects. As Crkol lies outside the Df(16)1 interval, this study indicates that independent loci might contribute to the DGS phenotype — a possibility as TBX1 mutations in DGS patients have not been found. However, Lindsay et al. suggest that DGS might be caused — not by neural-crest-cell defects — but by loss of a pharyngeal-arch-segmentation function of Tbx1. Only time and additional studies will tell.