An´ alisis de Se˜ nales

Although HFM is a complex disorder, there is a wealth of evidence to suggest a major genetic determinant in some cases of the condition. A low empiric recurrence risk in first degree relatives of affected individuals o f 2-3% has been suggested (Grabb 1965). However, a study of 97 propositi documented that 45% had a family history of some features of the disorder, with first degree relatives being most frequently affected (Rollnick and Kaye 1983). In some instances affected relatives displayed only a ‘microform’ of the disorder, for example mild grade microtia or a preauricular tag, providing further evidence that these isolated features represent the mildest end of the phenotypic spectrum (Rollnick and Kaye 1983; Bennun et al.

1985; Rollnick et al. 1987; Gorlin 1990). Segregation analysis performed on a cohort of 74 probands with affected relatives favours an autosomal dominant mode of inheritance in those families studied (Kaye et al. 1992).

Most affected individuals are cytogenetically normal however a number of chromosomal abnormalities have been observed in association with HFM. These

include del (5p) (Neu et al. 1982), monosomy (6q) (Greenberg et al. 1988), del (8q)

(Townes and White 1978), trisomy 18 (Greenberg et al. 1988), ring chromosome 21 (Greenberg et al. 1988), and 47 XXY (Poonawalla et al. 1980). Furthermore, a

number of abnormalities affecting chromosome 2 2 have been observed including del

(22q) (Greenberg et al. 1988), dup (22q) (Herman et al. 1988) and trisomy 22

(Kobrynski et al. 1993). In addition there are also reports o f chromosomal mosaicism including trisomy 7 (mosaic) (Hodes et al. 1981), trisomy 9 (mosaic) (Wilson and Barr 1983), and trisomy 22 (mosaic) (de Ravel et al. 2001) which may account for localised features and low recurrence risk observed in HFM.

Twin studies are more difficult to interpret, as there are an excess in the number of incidences of discordance in monozygotic twins, although rare cases of concordance with variable expression have been reported (Burck 1983; Ryan et al. 1988). This includes an unusual case of concordant monozygotic twins exhibiting a mirror image of the condition (Satoh et al. 1995). Low concordance rates may be explained in part by the postulation that the process o f monozygotic twinning itself may result in an early malformation complex or vascular disruption, as a result of a monochorionic placenta, to one of the resultant twins (Schinzel et al. 1979).

The best evidence for the possibility of a single gene having a major effect comes from rare familial cases in which the condition appears to segregate in a dominant manner, albeit with reduced penetrance and variable phenotypic expression within and between families (Regenbogen et al. 1982; Stoll et al. 1998; Singer et al.

with a Goldenhar-like syndrome to 8ql l-8ql3, the region harbouring the EYAl gene

underlying branchio-oto-renal (BOR) syndrome (Abdelhak et al. 1997a). A familial case o f hemifacial microsomia in association with acro-osteolysis exhibiting

autosomal recessive inheritance has also been reported (Brady et al. 1999).

1.5.2.1.1 Mouse models

Evidence from the mouse lends support to a genetic involvement in HFM. The

recessive lethal mouse first arch (far) mutant gives rise to a dominantly transmitted

hemifacial maxillary malformation, with reduced penetrance, when bred into a different genetic background (Juriloff et al. 1987). Additionally, a transgenic mouse

model caused by an insertional mutation on mouse chromosome 1 0 produces a

phenotype in hemizygous mice resembling HFM in humans (Naora et al. 1994). Interestingly, the phenotype in these mice including microtia, low set ears and an abnormal bite resulting from hypoplasia of the second branchial arch, is accompanied by local haemorrhage. This suggests an interaction between genetic and

environmental factors that may involve the same molecular or physiological pathways in development.

In addition there are a number of knockout mouse models with phenotypes reminiscent of those seen in HFM in humans. The orthologous regions o f these genes in humans represent potential candidate loci for human HFM.

Goosecoid

Goosecoid {GSC) is a homeobox transcription factor expressed during mid-

embryogenesis in developing structures of the first and second branchial arches and

first branchial cleft (Gaunt et al. 1993). Mice with a homozygous disruption of Gsc

exhibit numerous craniofacial and ear abnormalities (Yamada et al. 1995; Rivera- Perez et al. 1995). In particular, mandibular hypoplasia, underdevelopment of facial musculature and aplasia of the external acoustic meatus and tympanic bone.

Additionally, rib fusions were observed in a percentage of mice that were either bilateral or unilateral in some cases, demonstrating variability in the phenotype.

GSC is situated on mouse chromosome 12 and chromosome 14q32 in humans.

Fibroblast Growth Factor 8

Members of the fibroblast growth factor family {FGF), particularly FGF8,

have been implicated in the regulation of branchial arch development by acting as

epithelial signalling molecules. Fgf8 null mice die during gastrulation (Sun et al.

1999), however, mice with a conditional inactivation of the gene in the first branchial arch only exhibit severe craniofacial defects. Cartilaginous elements developing from the maxillary and mandibular processes fail to develop, together with absence of a number of bones derived from first arch mesenchyme (Trumpp et al. 1999).

Furthermore, a percentage of mice exhibited a less severe phenotype, affecting only one side of the head.

The FGF8 gene is situated on mouse chromosome 19, with the human homologue on

Endothelin-1

Endothelin-1 (EDNl) acts as a vasoconstrictor and mitogen on vascular

smooth muscle cells. Mice with a homozygous disruption of the gene have abnormalities of first branchial arch derived structures (Kurihara et al. 1994). Including a poorly developed mandible, with retarded mandibular bones and incomplete midline fusion. The ears were described as hypoplastic with absence of the auditory ossicles and tympanic ring. The zygomatic and temporal bones were underdeveloped and cleft palate was a common finding (Kurihara et al. 1994).

EDNJ is located on chromosome 13 in the mouse and human chromosome 6p24.

Transcription factor AP-2 a

The transcription factor AP-2a {TFAP2A) is a retinoic acid responsive gene

expressed in neural crest cell lineages which provide the patterning for craniofacial

morphogenesis. Tfap2a null mice exhibit severe craniofacial defects and failure of

cranial neural tube closure (Schorle et al. 1996; Zhang et al. 1996). Analysis of

chimeric mice reveals a direct role for Tfap2a in neural tube closure, eye

development and craniofacial morphogenesis. Craniofacial abnormalities occur independently of other defects and include CL/P, and mandibular and maxillary dysmorphogenesis consistent with expression in development of the facial processes

(Nottoli et al. 1998). TFAP2A is also situated on mouse chromosome 13 and '

Distal-less homeobox 2

DLX-2 is a vertebrate homologue of the Drosophila Distal-less homeobox

gene. Murine Dlx-2 is expressed in embryogenesis during critical times of

craniofacial morphogenesis in all four branchial arches. Expression is particularly evident in the mesenchyme of the maxillary and mandibular processes o f the first branchial arch, as well as the ectoderm of the frontonasal prominence and otic pit

(Bulfone et al. 1993). Homozygous loss of Dlx-2 results in abnormal morphogenesis

of bone and cartilage elements derived from the dorsal components of the first and second branchial arches (Qiu et al. 1995). Although the morphogenesis o f particular structures was consistently affected, detailed aspects of the abnormalities observed

varied between individual mice. The DLX2 gene is located on mouse chromosome 2,

the human homologue is located on chromosome 2q32.