Many genes have been implicated in syndromal and/or non-syndromal cleft lip with or without palate (CL/P). The majority of these candidate genes show expression in the facial ectoderm, including MSX1, BMP4, PVRL1, MID1 and p63, although some (eg. MSX1 and BMP4) also have key roles in the mesenchyme 1. In addition to coordinating mesenchymal outgrowth, the facial ectoderm also plays a number of other pivotal roles in facial morphogenesis, including facilitating the initial contact of the converging processes and the subsequent elimination of the epithelial seam in a manner that is likely analogous to that which occurs during formation of the more studied secondary palate. Recently, mutations in the IRF6 (Interferon Regulatory Factor 6) gene were shown to cause the allelic disorders, Van der Woude and Popliteal pterygium syndromes, both of which have CL/P as a major clinical feature 23. Significantly, common polymorphisms in IRF6 have also been found to account for up to 12% of the contribution to the high incidence of non-syndromal CL/P, highlighting it as one of the most significant CL/P loci identified to date 4. Although the exact physiological function of this gene is not known, preliminary findings in mice found that Irf6 is expressed in the medial edge epithelia of the fusing secondary palatal shelves, tooth buds, hair follicles and skin 5. Surprisingly, however, the expression of IRF6 has not been reported during development and closure of the primary palate, an event which is distinct both in terms of embryological timing and underlying genetics from that of the secondary palate.

Several secreted factors emanating from the facial ectoderm, including fibroblast growth factor 8 (FGF8) and bone morphogenic protein 4 (BMP4) induce mesenchymal expression of genes such as Msx1 and Msx2, that promote mesenchymal cell proliferation and prominence outgrowth 67. Evidence suggests that the mesenchymal LIM-Homeodomain (HD) encoding gene, Lhx7 (also referred to as Lhx8 8 and L3 9), is similarly under the control of ectodermal-derived signals, namely Fgf8 10111213. In the mouse, Lhx7 and its close homologue, Lhx6, were reported to be expressed only in the maxillary and rostral mandibular processes, palatal shelves and basal forebrain 91014. In Lhx7-knockout mice, an isolated secondary palatal cleft was the only reported feature: the secondary palatal shelves formed and elevated normally but failed to properly contact and fuse 15. Of note however, is that most inbred mouse strains rarely display lateral facial clefts analogous to CL/P in humans. This is probably due in part to the altered growth rates that give rise to the elongated facial morphology although differences in sensitivity to gene dosage or redundancies between related genes may also play a role 1. Consistent with this, mutations in MSX1 in humans are associated with clefts of the primary palate whereas knockout of both Msx1 and Msx2 are required to produce a primary palate cleft in mice 16. In contrast, the chick in some ways provides a more suitable model system for studies on primary palatal clefting as this species, like humans, shows greater susceptibility to this anomaly. Here, we have isolated chick cDNAs orthologous to human IRF6, LHX7 and LHX6 and investigated their profile of expression during morphogenesis of the midface with an emphasis on the period around formation of the primary palate.

Here we have isolated the chick orthologues of LHX7, LHX6 and IRF6 and shown a high degree of sequence conservation with their mouse and human counterparts, suggesting evolutionary conserved functions for these proteins. It should be noted that despite repeated attempts at amplification of mRNA/cDNA across the equivalents of exons 1 to 3 and analysis of available chick genomic sequence, cLHX7 did not contain the equivalent of human and mouse exon 2. As this exon does not encode a known functional domain and its absence maintains the reading frame, it is likely that cLHX7 also encodes a functional LIM-HD transcription factor. Interestingly, compared to the mouse, chick and human LHX7 had an additional 4 bp in the last coding exon producing a C-terminus with an additional nine amino acids. The validity of the mouse Lhx7 sequence over this region was confirmed by sequencing murine (Swiss) genomic DNA. That the additional 4 bp is also evident in rat LHX7 indicates this 4 bp deletion likely represents a recent evolutionary event that may be restricted to the murine lineage. The biological significance of this must await determination.

In order to determine whether the chick would provide an appropriate model with which to investigate the roles of LHX7, LHX6 and IRF6 in craniofacial development and CL/P, their respective expression patterns around the time of primary palate morphogenesis were determined. Similarly to the mouse, cLHX7 and cLHX6 were expressed in ectomesenchyme of the maxillary and mandibular primordia 910 although in contrast to the mouse, the mandibular expression of cLHX7 was not prominent. Differential expression of cLHX7 and cLHX6 in the mandibular primordia was also evident at the anterior tips of the palatal shelves suggesting thee two genes are under distinct regulatory control which is consistent with results from Cre-mediated Fgf8-knockout mice 11 in which Lhx7 but not Lhx6 expression was lost.

Of particular interest, our expression studies in the chick have identified unique LHX7 and LHX6 expression domains. We detected strong cLHX7 and cLHX6 expression in the mesenchyme immediately underlying the pre-fusion epithelia of the medial nasal process, from around HH22, which remained in the mesenchymal bridge post fusion. This expression has not previously been reported for the mouse and importantly suggests a role for LHX7 and LHX6 in outgrowth/survival of the medial nasal process during formation of the primary palate. cLHX7 and cLHX6 expression was also detected in the otic vesicle (from HH25 to HH30) a site of expression also not been reported in mice or any other species and therefore may be unique to the chick.

The strong expression in maxillary and medial nasal mesenchyme subjacent to the pre- and post fusion ectoderm indicate that LHX7 and LHX6 would be good candidate genes for craniofacial anomalies, in particular CL/P despite the isolated secondary palate cleft in Lhx7 knock-out mice. In this regard, hLHX7 localizes to chromosome 1p31-4 (and not 4q as previously suggested 20) and is found less than 1.4 Mb from marker D1S1665 which showed the most significant linkage in one cohort of non-syndromal CL/P cases 20. In fact, this same region produced the only positive lod score for an individual Finnish family presenting with Van der Woude syndrome-like features 21. These data and the results reported herein put forward a case for screening patients with non-syndromal CL/P or IRF6 mutation negative Van der Woude syndrome for mutations in LHX7.

IRF6 is mutated in Van der Woude and Popliteal pterygium syndromes 23 and has recently been identified as one of the most significant non-syndromal CL/P loci to date 4. This report is the first to describe IRF6 expression in the facial primordia prior to and during morphogenesis of the primary palate and supports the notion of a primary ectodermal defect in patients harboring mutations in IRF6. In concordance with the later embryonic stages assessed in the mouse 5, cIRF6 was similarly detected in the leading edge ectoderm of the palatal shelves and the ridges of the primitive oral cavity in the chick. Interestingly, we also detected some unique expression domains of IRF6, which may be specific to the chick. Strong IRF6 expression was detected in the egg tooth primordium indicating it as an excellent marker for this structure. Like other genes that are expressed in the ectoderm of the developing face such as SHH and BMP4 2223, high IRF6 expression was also detected in the apical ectodermal ridge of the limb buds, which is consistent with the presence of limb hypoplasia or agenesis of digits, syndactyly, as well as valgus or varus deformities of the feet seen in Popliteal pterygium syndrome 24.

The data presented herein shows both highly conserved and unique temporal and spatial expression of LHX7, LHX6 and IRF6 in the chick, particularly in the facial primordia around the time of their fusion to form the primary palate. The manipulability, minimal cost and susceptibility of chicks to CL/P will enable more detailed investigations into the functions of these genes in midfacial development and their role in contributing to common orofacial clefts.

BJW participated in the design of the study, carried out all described experiments and drafted the manuscript. TCC conceived, designed and coordinated the study, advised on protein alignments, and helped draft the manuscript. Both authors read and approved the final manuscript.

During review of our article, a study by Inoue et al (Inoue M, Kawakami M, Tatsumi K, Manabe T, Makinodan M, Matsuyoshi H, Kirita T, Wanaka A. Expression and regulation of the LIM homeodomain gene L3/Lhx8 suggests a role in upper lip development of the chick embryo. Anat Embryol. 2006 Epub ahead) was published that similarly reported restricted mesenchymal expression of chick LHX7 and demonstrated that it, like its murine counterpart, appears to be under the regulatory control of epithelial signals including FGF8.