Hedgehog Signaling in Animal Development and Human Disease 221
Identification of PTCHl mutations in both sporadic and familial
cases of basal cell carcinoma and medulloblastoma, while compelling in
its consistency, is nonetheless a correlative argument. Targeted deletion
of mouse pte 1 and a number of transgenic experiments have provided
convincing functional data that aberrant Hh signaling is a cause and not
a consequence of tumorigenesis. Mice heterozygous for a ptc1 deletion,
like humans, suffer a variety of congenital malformations including
polydactyly and rib anomalies (Goodrich et al. 1997; Hahn et al. 1998).
Rhabdomyosarcomas and medulloblastomas, both seen in BCNS, are
also seen in these mice at a relatively high frequency.
Interestingly, pte 1 heterozygous mice do not develop basal cell carcinomas. However, mice overexpressing Shh in the skin do develop a
condition strikingly similar to BCC in humans (Fan et al. 1997; Oro et
al. 1997). Excess Shh ligand appears to be sufficient for tumorigenesis,
presumably by saturating and inactivating Ptc proteins. Mice overexpressing a mutated form of Smo in the skin also develop a condition
similar to BCC (Xie et al. 1998). These missense mutations, originally
identified in sporadic BCCs, are proposed to make Smo resistant to Ptc
inhibition and hence constitutively active. Why does unrestrained Shh or
Smo activity in mouse skin give rise to BCC while inactivation of ptc1
does not? Perhaps a second pte gene, possibly pte2, controls proliferation in the skin in parallel with, or in place of, ptc1. It is intriguing that
this role of ptc1 might not be conserved between mice and humans.
12.4.2 GLI3
Components at the distal end of the Hh signaling pathway, the Gli
family of transcription factors, have also been implicated in human
disorders. Glii was originally identified as a putative oncogene that is
amplified in human gliomas (Kinzler et al. 1987), and only later was
linked to Hh signaling (Orenic et al. 1990). Gli2 has important effects on
the formation of the esophagus and trachea. Gli2 deficient mice that are
heterozygous for Gli3, develop tracheo-esophageal fistula (Motoyama
et al. 1998). While this defect is a congenital abnormality of unknown
etiology in humans, Gli2 might be a candidate gene.
GLl3 has been directly implicated in three distinct human autosomal
dominant disorders. Greig cephalopolysyndactyly (GCPS), Pallister-
Identification of PTCHl mutations in both sporadic and familial
cases of basal cell carcinoma and medulloblastoma, while compelling in
its consistency, is nonetheless a correlative argument. Targeted deletion
of mouse pte 1 and a number of transgenic experiments have provided
convincing functional data that aberrant Hh signaling is a cause and not
a consequence of tumorigenesis. Mice heterozygous for a ptc1 deletion,
like humans, suffer a variety of congenital malformations including
polydactyly and rib anomalies (Goodrich et al. 1997; Hahn et al. 1998).
Rhabdomyosarcomas and medulloblastomas, both seen in BCNS, are
also seen in these mice at a relatively high frequency.
Interestingly, pte 1 heterozygous mice do not develop basal cell carcinomas. However, mice overexpressing Shh in the skin do develop a
condition strikingly similar to BCC in humans (Fan et al. 1997; Oro et
al. 1997). Excess Shh ligand appears to be sufficient for tumorigenesis,
presumably by saturating and inactivating Ptc proteins. Mice overexpressing a mutated form of Smo in the skin also develop a condition
similar to BCC (Xie et al. 1998). These missense mutations, originally
identified in sporadic BCCs, are proposed to make Smo resistant to Ptc
inhibition and hence constitutively active. Why does unrestrained Shh or
Smo activity in mouse skin give rise to BCC while inactivation of ptc1
does not? Perhaps a second pte gene, possibly pte2, controls proliferation in the skin in parallel with, or in place of, ptc1. It is intriguing that
this role of ptc1 might not be conserved between mice and humans.
12.4.2 GLI3
Components at the distal end of the Hh signaling pathway, the Gli
family of transcription factors, have also been implicated in human
disorders. Glii was originally identified as a putative oncogene that is
amplified in human gliomas (Kinzler et al. 1987), and only later was
linked to Hh signaling (Orenic et al. 1990). Gli2 has important effects on
the formation of the esophagus and trachea. Gli2 deficient mice that are
heterozygous for Gli3, develop tracheo-esophageal fistula (Motoyama
et al. 1998). While this defect is a congenital abnormality of unknown
etiology in humans, Gli2 might be a candidate gene.
GLl3 has been directly implicated in three distinct human autosomal
dominant disorders. Greig cephalopolysyndactyly (GCPS), Pallister-