The mutational spectrum of the Sonic Hedgehog ... - Semantic Scholar

© 1999 Oxford University Press

Human Molecular Genetics, 1999, Vol. 8, No. 13 2479–2488

The mutational spectrum of the Sonic Hedgehog gene in holoprosencephaly: SHH mutations cause a significant proportion of autosomal dominant holoprosencephaly Luisa Nanni1, Jeffrey E. Ming1, Maureen Bocian3, Kathryn Steinhaus3, Diana W. Bianchi4, Christine de Die-Smulders5, Aldo Giannotti6, Kiyoshi Imaizumi7, Kenneth L. Jones8, Miguel Del Campo8, Rick A. Martin9, Peter Meinecke10, Mary Ella M. Pierpont11, Nathaniel H. Robin12, Ian D. Young13, Erich Roessler2 and Maximilian Muenke1,2,+ 1Departments

of Pediatrics and Genetics, The Children’s Hospital of Philadelphia, University of Pennsylvania School of Medicine, 34th and Civic Center Boulevard, Philadelphia, PA 19104-4399, USA, 2Medical Genetics Branch, National Human Genome Research Institute, National Institutes of Health, Bethesda, MD 20892-1852, USA, 3University of California Irvine, Orange, CA, USA, 4Tufts University School of Medicine, Boston, MA, USA, 5University Hospital Maastricht, The Netherlands, 6Bambino Gesu’ Hospital, Rome, Italy, 7Kanagawa Children’s Medical Center, Yokohama City, Japan, 8University of California, San Diego, CA, USA, 9St Christopher’s Hospital for Children, Philadelphia, PA, USA, 10Altonaer Kinder-Krankenhaus, Hamburg, Germany, 11University of Minnesota Hospital, Minneapolis, MN, USA, 12Case Western Reserve University, Cleveland, OH, USA and 13Nottingham City Hospital, Nottingham, UK Received July 26, 1999; Revised and Accepted September 28, 1999

Holoprosencephaly (HPE) is a common developmental anomaly of the human forebrain and midface where the cerebral hemispheres fail to separate into distinct left and right halves. We have previously reported haploinsufficiency for Sonic Hedgehog (SHH) as a cause for HPE. We have now performed mutational analysis of the complete coding region and intron–exon junctions of the SHH gene in 344 unrelated affected individuals. Herein, we describe 13 additional unrelated affected individuals with SHH mutations, including nonsense and missense mutations, deletions and an insertion. These mutations occur throughout the extent of the gene. No specific genotype–phenotype association is evident based on the correlation of the type or position of the mutations. In conjunction with our previous studies, we have identified a total of 23 mutations in 344 unrelated cases of HPE. They account for 14 cases of familial HPE and nine cases of sporadic HPE. Mutations in SHH were detected in 10 of 27 (37%) families showing autosomal dominant transmission of the HPE spectrum, based on structural anomalies. Interestingly, three of the patients with an SHH mutation also had abnormalities in another gene that is

expressed during forebrain development. We suggest that the interactions of multiple gene products and/or environmental elements may determine the final phenotypic outcome for a given individual and that variations among these factors may cause the wide variability in the clinical features seen in HPE. INTRODUCTION Holoprosencephaly (HPE) is a clinically variable and genetically heterogeneous malformation in which the developing forebrain fails to correctly separate into right and left hemispheres. Clinical expression of HPE is highly variable even within a single family (1). In its most severe form (alobar HPE), there is no interhemispheric fissure, a single brain ventricle is present and there may be cyclopia and/or a proboscis-like nasal structure. Less severe facial dysmorphisms (considered HPE microforms) include microcephaly, ocular hypotelorism, impaired sense of smell or single central upper incisor (2,3). The majority of HPE cases are apparently sporadic, although clear examples of autosomal dominant (AD) inheritance have been described. Interestingly, up to 30% of obligate carriers of an HPE gene in AD pedigrees are clinically unaffected (4).

+To whom correspondence should be addressed at: Medical Genetics Branch, National Human Genome Research Institute, National Institutes of Health, 10 Center Drive, MSC 1852, Building 10, 10C101, Bethesda, MD 20892-1852, USA. Tel: +1 301 402 8167; Fax: +1 301 496 7157; Email: [email protected]

2480 Human Molecular Genetics, 1999, Vol. 8, No. 13

Based on recurrent cytogenetic abnormalities, at least 12 chromosomal regions are proposed to contain genes involved in HPE (5). The gene at the locus designated HPE3 (6) was identified as Sonic Hedgehog (SHH) and we proposed that haploinsufficiency for SHH causes HPE (7,8). Recently, mutations in two other genes have been identified in association with HPE: (i) ZIC2, which maps to an HPE critical region on 13q32 (9,10); and (ii) SIX3, which maps to the HPE2 locus on 2p21 (11). Shh plays a critical role in early forebrain and central nervous system development (12,13). Mice homozygous for a disrupted Shh gene show defects in the development of midline neural structures, lack ventral cells in the brain and display craniofacial anomalies, including cyclopia and a proboscis-like nasal structure (14). This phenotype is consistent with the defects seen in human HPE. The Shh protein is a secreted intercellular signaling molecule which is synthesized as a precursor that undergoes autocatalytic cleavage into a highly conserved N-terminal domain (Shh-N) and a more divergent C-terminal domain (Shh-C) (15). During the autoprocessing reaction, a cholesterol moiety is covalently attached to the C-terminus of Shh-N (16–18). Shh-N contains all the known signaling activities (16,19–21). In contrast, Shh-C mediates both the enzymatic cleavage and cholesterol modification of the protein (17,18,22). This modification is crucial for proper patterning activity. We had previously performed mutational analysis of the first and second exons of SHH in 30 familial cases, as well as the third exon in 41 familial and 184 sporadic cases of HPE (8,23). Ten different mutations consistent with a loss of function of the mutated allele were found, accounting for nine cases of familial HPE and one case of sporadic HPE. Here we report mutational analysis of the entire coding region and exon– intron boundaries of SHH in 344 unrelated HPE patients. We report 13 additional affected individuals with SHH mutations and provide further evidence that both SHH-N and SHH-C are equally affected by mutations.

Table 1. Summary of sequence variations in the SHH gene Patients

Sequence change

Expected effect

Reference

Exon 1 Pedigree I

160–161 insGCTG 3–4 ins

Pedigree E

189–196 del

ADHPE 10

GGG→AGG

Gly31→Arg

8

Pedigree A

GAT→GTT

Asp88→Val

Present study

ADHPE 6

CAG→TAG

Gln100→Stop

8

Gln100→His

24

Lys105→Stop

8

Sporadic HPE CAG→CAC

Present study

13–15 del with frameshift Present study

Exon 2 ADHPE 14

AAG→TAG

Pedigree B

AAC→AAA

Asn115→Lys

Present study

ADHPE 15

TGG→GGG

Trp117→Gly

8

ADHPE 2

TGG→CGG

Trp117→Arg

8

Tyr158→Stop

24

Glu188→Gln

24

Familial HPE TAC→TAG Exon 3 Familial HPE GAG→CAG CAG→TAG

Gln209→Stop

Present study

Familial HPE GAC→AAC

Patient M

Asp222→Asn

24

ADHPE 11

GTG→GAG

Val224→Glu

23

ADHPE 43

GCG→ACG

Ala226→Thr

23

Pedigree C

AGC→AGA

Ser236→Arg

Present study

Pedigree H

GAG→TAG

Glu256→stop

Present study

Patient L

939–959 del

263–269 del

Present study

ADHPE 4

939–959 del

263–269 del

23

ADHPE 3

GAG→TAG

Glu284→Stop

23

Patient J

GGC→GAC

Gly290→Asp

Present study

Pedigree F

1283–1291 del

Sporadic HPE GCG→ACG

378–380 del

Present study

Ala383→Thr

23

Pedigree G

1361–1375 del

404–408 del

Present study

RESULTS

Pedigree D

CCG→GCG

Pro424→Ala

Present study

We screened 344 HPE cases (78 familial and 266 clinical sporadic) for the entire coding region and exon–intron boundaries of SHH using SSCP. This study includes the same 41 families and 184 sporadic cases that had portions of the SHH gene previously analyzed and confirms the original findings. We detected 13 previously unreported probands with a mutation in SHH (Table 1). Of these mutations, the sequence change was present in more than one family member in eight of the nine kindreds depicted in Figure 1. Interestingly, one of these mutations (the 21 bp deletion) was also present in a family previously reported (23). Taking into account the 10 mutations described previously by our group (8,23) and four additional mutations reported by another group (24), a total of 27 probands have been noted to have SHH mutations (Table 1 and Fig. 2A and B).

Patient K

TCG→TTG

Ser436→Leu

Present study

Missense mutations Six missense mutations were noted. The first was observed in two affected children and their clinically unaffected mother. Her three sisters also carried the mutation: one is clinically

unaffected with a child who also showed the mutation and has microcephaly and moderate learning disabilities; the other two have microcephaly and a high arched palate (Figs 1A and 3J). The GAT→GTT (Asp88→Val) sequence change occurs in the N-terminal signaling domain at an invariant position in the hedgehog family of proteins (Fig. 2B). This mutation was confirmed by the loss of an MboI restriction site (data not shown). The second missense mutation was detected in a child with HPE and his clinically unaffected mother (Fig. 1B). The AAC→AAA (Asn115→Lys) change occurs at an invariant position in the hedgehog proteins, in the N-terminal domain (Fig. 2B). The third missense mutation was detected in a child with HPE as well as his mother and his maternal grandmother, both of whom have a single central maxillary incisor, his great grandfather, who has ocular hypotelorism, and in a great uncle and his daughter, both clinically unaffected (Figs 1C and 3E and F). The AGC→AGA (Ser236→Arg) sequence change also

Human Molecular Genetics, 1999, Vol. 8, No. 13 2481

Figure 1. Pedigrees for the nine families carrying SHH mutations. Solid symbols, HPE phenotype; half-solid symbols, MC, OH, CLP, DD or SCI; symbols enclosing a dot, clinically unaffected mutation carriers; *, tested individuals with normal SHH sequence. The following individuals were not available for the study: III.7 in pedigree A; II.2 and III.4 in pedigree C; II.1 in pedigree 0; II.2, II.6 and II.7 in pedigree I. For clinical details and abbreviations, see Table 2.

predicts a change at an invariant position in the hedgehog proteins, in the C-terminal domain (Fig. 2B). A GGC→GAC (Gly290→Asp) sequence change was observed in a 20-year-old woman with HPE (Fig. 3C and D and Table 2, patient J) and is predicted to occur in the processing domain. This mutation was confirmed by loss of a

BanII restriction site (data not shown). Her parents were not available for study. Of note, a mutation in the ZIC2 gene, predicting an alanine repeat expansion, was also found in this patient (9). The fifth change was seen in an affected child and her clinically unaffected mother (Figs 1D and 3B). The CCG→GCG (Pro424→Ala) sequence change occurs 39 amino


Figure 2. (A) Location of mutations in SHH in individuals with HPE. The mutations are distributed evenly between the coding regions. SHH-N, signaling domain; SHH-C, autocatalytic cleavage and cholesterol transferase domain; #, insertion; star, frameshift deletion; solid circle, missense mutations; solid diamond, nonsense mutations; *, deletions; SP, signal peptide, removed after translation; ↓, cleavage/cholesterol addition. (B) Amino acid sequence alignment for human Sonic (hSHH), mouse Sonic (mShh), chicken Sonic (cShh), zebrafish Sonic (zShh), and Drosophila hedgehog (dhh). The locations of the predicted amino acid changes in HPE patients are shown in blue (8,23,24; present study). Amino acids conserved between species are shown in red. The positions of the splice sites of exons 1, 2 and 3 are indicated by arrows and a gap is introduced in the protein sequence to mark the position of the autocatalytic cleavage between Gly197 and Cys198 of the human protein. The residues essential for hedgehog autoprocessing activity in Drosophila are in green (see text).

acids from the end of the protein. In addition, this patient’s karyotype showed loss of the terminal portion of 18p, derived from a maternal balanced cryptic translocation [t(1;18)

(q43;p11.3)]. The del(18p) chromosome is deleted for TGinteracting factor (TGIF), a candidate gene for the HPE4 locus at 18p (25). The last missense mutation was detected in a


Figure 3. Holoprosencephaly (HPE) in individuals with Sonic Hedgehog (SHH) mutations. (A) Microcephaly, absence of nasal bones, midline cleft lip and palate and semilobar HPE in patient II.1 of pedigree F with SHH and TGIF mutations. (B) Semilobar HPE, premaxillary agenesis and midline cleft lip in patient II.1 in pedigree D with an SHH mutation and a TGIF deletion. (C) Patient J with severe semilobar HPE on brain scan (D) and relatively mild facial findings with SHH and ZIC2 mutations. (E) Patient IV.1 of pedigree C with semilobar RPE on MRI (F), microcephaly, prominent globes, premaxillary agenesis and cleft lip. (G) Microcephaly, ocular hypotelorism, flat nose with no palpable cartilage, midface and philtrum hypoplasia and normal intelligence in individual II.3 of pedigree I with an SHH mutation. (H) Normal brain magnetic resonance imaging (MRI) of the same individual. (I) One of the twins (II.6/II.7) of family I with alobar HPE and severe facial findings (synophthalmia and proboscis above the eye). (J) Microcephaly, ocular hypotelorism, bilateral inferior iris colobomata, repaired rightsided cleft lip and palate and a normal brain MRI in patient III.4 of pedigree A. (K) Microcephaly; philtrum hypoplasia and a normal brain MRI in individual II.1 of pedigree E. (L) Patient III.2 of family E with lobar HPE, microcephaly and philtrum hypoplasia.

patient with semilobar HPE (Table 2, patient K). This TCG→TTG (Ser436→Leu) change predicts an amino acid substitution 27 amino acids from the end of the protein. The parents were unavailable for analysis. Deletions Four different deletions were observed. An 8 bp deletion leading to a frameshift (bases 189–196) was detected in a fetus with HPE, as well as in a sibling with microcephaly and severe developmental delay and their mother, who has a single central incisor and anosmia (Figs 1E and 3K and L). A 21 bp deletion in the C-terminal domain at codons 263– 269 was detected in a sporadic patient (Table 2, patient L). This predicts a deletion of seven amino acids (RLLLTAA). This deletion immediately precedes a critical His residue (His270, which corresponds to His329 in the Drosophila hedgehog protein) required in the autoprocessing cleavage reaction (26). The identical deletion was previously found to

segregate with the affected members of an AD HPE family (23). A 9 bp deletion was detected in an affected child and her clinically normal mother (Figs 1F and 3A). It occurs at codons 378–380 and predicts the loss of three amino acids (APF). Of note, this patient also showed a missense mutation in TGIF, a candidate gene for the HPE4 locus at 18p (25). A 15 bp deletion (codons 404–408) was detected in a clinically unaffected mother of an HPE fetus (Fig. 1G). The fetus was not available for study. This deletion is predicted to result in omission of the amino acids GDRGG in a minimally conserved region of SHH-C. Nonsense mutations Two nonsense mutations were detected. The first mutation, a de novo CAG→TAG change, was detected in a sporadic HPE patient (Table 2, patient M) and predicts premature termination of the protein at position 209, 12 amino acids after the cleavage


Table 2. Phenotype evaluation of HPE patients with an SHH mutation Patients Familial Pedigree A

Individual

Clinical findings

Additional information

I.2 II.1 II.2 II.3 II.4 II.5 III.4

Clinically normal Clinically normal Clinically normal Clinically normal MC, high arched palate LD, MC, high arched palate MC, OH, bilateral inferior iris colobomata, repaired right side CL, fused lower central incisors HPE, OH, PA, single nostril, bilateral CL, absent olfactory nerves, fused thalami Mild MC, moderate LD Mild MC, moderate LD Clinically normal Clinically normal HPE, MC, flat nose, bilateral CLP OH Clinically normal SCI, mild OH Died at 2 years old, ‘open skull’, CL Clinically normal SCI, mild OH DD Clinically normal Clinically normal HPE, MC, sloped forehead, OH, prominent globes, bilateral optic nerve colobomata, PA, CL Clinically normal HPE, PA, midline CL LD Clinically normal MC, SCI, anosmia, absent frenulum Mild DD, CHD Severe DD, MC, philtrum hypoplasia HPE, proboscis Clinically normal HPE, MC, OH, absent nasal bones, midline CLP Clinically normal Clinically normal HPE, MC, moderate hydrocephalus, OH, absent proboscis and nares, CLP Clinically normal Clinically normal MC, SCI, choanal stenosis HPE, MC, PA, single nasal cavity, bilateral CL and absent anterior palate Clinically normal Mild OH HPE, MC, OH, single nostril, CP MC, OH, flat nose, no palpable nasal cartilage, midfacial and philtrum hypoplasia HPE, OH, cyclopia, proboscis HPE, cyclopia with double orbitis, proboscis

Normal SHH Normal SHH SHH mutation carrier SHH mutation carrier

III.5

Pedigree B

Pedigree C

Pedigree D Pedigree E

Pedigree F Pedigree G

Pedigree H

Pedigree I

III.6 III.7 I.1 I.2 II.2 I.1 1.2 11.1 11.2 11.4 111.2 111.4 111.5 111.6 IV.1 I.2 II.1 I.1 I.2 II.1 III.1 III.2 III.3 I.2 II.l I.1 I.2 II.1 I.1 I.2 II.1 II.2 I.1 I.2 II.2 II.3 II.6/II.7 II.10

Sporadic Patient J

HPE, MC

Patient K Patient L Patient M

HPE, MC, single nostril, midfacial hypoplasia, CL HPE, corpus callosum agenesis, pachygyria, ventricular dilatation HPE, hydrocephalus, occipital encephalocele, CLP

Normal brain (CT) Normal brain (MRI); DD; bilateral hearing loss; 46,XY Alobar HPE; fetus 46,XY

Not available for study Normal SHH SHH mutation carrier Lobar HPE; 46,XY Normal SHH Not available for study SHH mutation carrier Not available for study Normal SHH SHH mutation carrier Semilobar HPE; severe micropenis; 46,XY SHH mutation carrier; cryptic 46,XX, t(1;18)(q43;p11.3) Semilobar HPE; DI; l8p– Not available for study Normal SHH Normal brain (MRI) Normal SHH; normal brain (MRI) Lobar HPE SHH mutation carrier Semilobar HPE; TGIF mutation; 46,XX Normal SHH SHH mutation carrier Semilobar HPE; fetus 46,XX; not available for study Normal SHH Normal SHH Lobar HPE; fetus 46,XX Normal SHH Normal SHH Alobar HPE; ambiguous genitalia; 46,XY; not available for study Normal IQ; SHH mutation Alobar HPE; 46,XY; not available for study Alobar HPE; SHH mutation Semilobar HPE; DI; progressive spastic paraplegia; SS; ZIC2 mutation; 46,XX; 20 years old; parents not available Semilobar HPE; optic nerve hypoplasia; parents not available MCA; 46,XX; parents normal SHH Semilobar HPE; 46,XX; parents normal SHH

CHD, congenital heart defect; CL, cleft lip; CLP, cleft lip and palate; DD, developmental delay; DI, diabetes insipidus; LD, learning disabilities; MC, microcephaly; MCA, multiple congenital anomalies; OH, ocular hypotelorism; PA, premaxillary agenesis; SCI, upper single central incisor; SS, short stature.


site. Such a protein would be predicted to fail to undergo the autocatalytic cleavage reaction and would not be expected to be modified by cholesterol. The second nonsense mutation was found in a female fetus with HPE and her brother with a single central incisor and microcephaly (Fig. 1H). The GAG→TAG change was confirmed by the gain of an XbaI restriction site (data not shown) and predicts termination of the protein at position 256. Neither parent carried the mutation. Paternity testing showed that the father was in fact the biological father of both of the two siblings. This pedigree is consistent with germline mosaicism in one of the parents. Insertion A 4 bp insertion (GCTG) after the third codon of the gene was found in a large autosomal dominant HPE family (Figs 1I and 3G–I). Two affected members in the kindred were available for the study and both showed the 4 bp insertion. The predicted resulting peptide would terminate at codon 62. Neither parent was a mutation carrier and haplotyping confirmed that they were indeed the biological parents. This pedigree is consistent with germline mosaicism in one of the parents. Predicted polymorphisms Five nucleotide changes that would not be predicted to cause a change in the SHH protein were detected: two putative polymorphisms in exon 3 (Ala275→Ala and Val335→Val); three in intronic sequences (49 bases upstream of exon 2, 71 bases downstream of exon 2 and 92 bases downstream of exon 2). DISCUSSION In order to refine our estimate of the frequency of mutations in SHH that cause HPE and to attempt to establish a genotype– phenotype correlation, we performed an extensive mutational analysis of the entire coding region and exon–intron boundaries of the SHH gene. We identified 13 additional patients with SHH gene mutations: five were in familial HPE cases and eight were in sporadic cases. ‘Familial’ versus ‘sporadic’ cases were categorized based on clinical and not molecular findings. We had previously identified SHH mutations in 10 HPE probands. Combining all our data, we estimate that mutations in SHH account for 14/78 (18%) of the clinical familial HPE cases and for 9/266 (3.4%) of the sporadic cases. Therefore, SHH mutations are present in 23/344 (6.7%) of our total cohort of HPE patients. Using broad criteria for inclusion as an AD pedigree, including features that are present to a significant degree in the general population (e.g. microcephaly, hypotelorism, mental retardation), the frequency of SHH mutations in AD pedigrees was 24% (11/46). When we limited criteria for inclusion in an AD pedigree to structural anomalies (e.g. anosmia, corpus callosum agenesis in association with HPE, cleft lip and palate, single central incisor), the frequency of SHH mutations was 10 out of 27 (37%). The finding of SHH mutations in only a minority of the total cohort of HPE patients underscores the significant etiological heterogeneity of this condition. These figures should be considered estimates. Seven of the 23 changes would predict an obvious loss of function due to a nonsense mutation or frameshift. However, the functional effect of many of the missense mutations is unknown. Thus,

the calculations of SHH mutation frequency could change, depending on the results of functional studies. Interestingly, as illustrated by four of our pedigrees (Fig. 1B, D, F and G), mutational analysis of the clinically unaffected parents showed evidence for clinically unaffected mutation carriers. This emphasizes that clinically normal parents of a child with HPE have some risk of carrying an SHH gene mutation, making counselling for recurrence risk more challenging. Furthermore, we previously reported a family with a nonsense mutation in SHH in which a clinically normal family member also carried the mutation (8). It has been previously estimated that 30% of obligate carriers of HPE are clinically normal (4). In addition, in two kindreds (pedigrees H and I) two siblings had an SHH mutation, but neither parent carried the mutation. Paternity testing showed that the father was in fact the biological father of affected sibs in both cases. This is strongly suggestive of germline mosaicism and would be the first such cases for the SHH gene. However, we should note that the effect of these mutations on SHH function are unknown. Although these alterations were not present in >200 normal control chromosomes, the changes could represent rare polymorphisms which do not affect SHH activity. Functional studies are in progress to determine whether the activity of the abnormal protein is decreased. No specific genotype–phenotype correlation according to type or location of the mutations in SHH is observed. In addition, there were no clinical features unique to individuals with an SHH mutation compared with those without a detected mutation. The phenotype of carriers of an SHH mutation within a single family can vary from alobar HPE to clinically normal individuals (1). As another example, a 21 bp deletion was detected in a sporadic case (Table 2, patient L) who had pachygyria, which is not typically associated with HPE, and additional congenital anomalies including a two-vessel umbilical cord, ventricular septal defect of the heart, malrotation of the large bowel and bicornuate uterus. The same deletion was also detected in an AD HPE family (23). Affected members in this family showed only features frequently associated with HPE, such as microcephaly, single central incisor and cleft lip/ palate. Thus, the same deletion can give rise to somewhat different phenotypes. Based on known domains in the hedgehog family of proteins, we can speculate about the effects of the mutations we observe in our HPE patients. The association of chromosomal deletions including the SHH gene with HPE is consistent with a loss of SHH function (27). Mutations could affect any of the three principal functional regions of Shh: (i) changes in the SHH-N signaling region; (ii) interference with the autocatalytic cleavage reaction in SHH-C, since an intact precursor molecule has no patterning activity (21); and (iii) interference with the cholesterol transferase activity. We encounter likely examples of all three mechanisms in our sample of patients. Combining the present data with previously published information (8,23,24), eight mutations predict premature termination of the protein: five cause truncation of SHH-N and three truncate within SHH-C (Fig. 2) and may lead to abnormal processing (26). A total of seven missense mutations are observed in the signaling domain. In addition, certain of the residues were abnormal in two unrelated individuals. Both a nonsense and a


missense (Q100H) mutation have been reported for Gln100 (8,24). Two different missense mutations (Trp117→Gly and Trp117→Arg) have been reported for nucleotide 500 (8). Trp117 is a residue located just distal to the first α-helix motif in the murine protein (28) that is invariant between species studied and mutations at this position may destabilize the SHH-N fragment. The 263–269 deletion was found in two probands and occurs just before a key His residue (His270) that is essential for hedgehog autoprocessing activity in Drosophila (26). Furthermore, this deletion encompasses a Thr residue (Thr267) that is required for thioester formation in the Drosophila hedgehog protein, which is the first step in the autoprocessing reaction. The 263–269 deletion has also been shown to inhibit autoprocessing of the protein in vitro (E. Roessler et al., unpublished data). Given the great intrafamilial clinical variability in kindreds carrying an SHH mutation, we speculate that other genes acting in the same or different developmental pathways might act as modifiers for expression of the HPE spectrum. We identified three cases in which individuals with an SHH mutation also had a base change in a second gene which acts in brain development. The Gly290→Asp change in SHH was present in a patient (Table 2, patient J) who also had a mutation predicting an expansion of an alanine repeat in exon 2 of ZIC2. This alanine repeat extension has been described in other HPE patients (9,10). Her parents were not available for analysis. Since functional effects of these mutations are not known, we do not know whether the effects of one or the other mutation is causing HPE. Although there is no known interaction between the ZIC2 and SHH pathways, it is possible that their biological functions converge on a common pathway. Therefore, reduced amounts or activity of one protein might negatively affect the expression or function of the other. The relative roles of ZIC2 and SHH in the developing brain and face may be deduced from the frequency of associated craniofacial malformations in individuals with a mutation in one of the genes. As with most cases of HPE, affected individuals with SHH mutations often have facial anomalies. However, craniofacial anomalies appear to be less severe and less frequent in individuals with ZIC2 mutations (10). Second, a Pro424→Ala change in SHH was noted in both a child who was deleted for 18pter and the putative HPE4 gene, TGIF (25), and her mother, who carried a balanced translocation involving chromosome 18 (Fig. 1D). The third example is a 9 bp deletion in SHH (378–380del) in a child with HPE (Fig. 1F) who also showed a missense mutation (Thr151Ala) in TGIF (25). TGIF maps to the minimal critical region of HPE4 on chromosome 18p. This gene is expressed during early brain development in mice (29) and four missense mutations have been identified in patients with HPE (25). TGIF codes for a transcription factor which competitively inhibits binding of the retinoic acid receptor (RXR) to a retinoid-responsive promoter. Therefore, decreased TGIF levels may theoretically lead to enhanced binding of RXR, mimicking increased retinoic acid (RA) levels. Interestingly, prenatal RA exposure is associated with HPE-like malformations in animal models (30). In addition, high doses of RA down-regulate Shh expression in craniofacial primordia in the chicken (31,32). Thus,

decreased TGIF could result in decreased SHH expression, which could accentuate the effects of an allele of SHH with reduced or no activity. Of note, only 10% of patients carrying a TGIF deletion show HPE. It is possible that either maternal RA levels or altered activity in another protein could modify the effect of TGIF. In addition to potential modifier genes, environmental factors could also modulate the clinical expression of HPE (33). The importance of cholesterol in normal SHH function (18) raises the possibility that cholesterol levels in utero may contribute to the variable phenotype of HPE caused by SHH mutations. Maternal hypocholesterolemia in early gestation causes an HPE-like phenotype in rodent embryos (reviewed in ref. 5). Moreover, HPE is associated with severe cases of Smith–Lemli–Opitz syndrome, which is due to a defect in 7dehydrocholesterol reductase, the final step in the cholesterol biosynthetic pathway (34,35). These speculated interactions between genes might help to explain the clinical variability seen in HPE. Functional studies of the mutations detected in SHH, TGIF and ZIC2 will be essential in determining the significance of this hypothesis. As additional genes related to HPE are identified and the interaction between their products is investigated, a more complete understanding of the numerous genetic and environmental factors which contribute to normal brain development and HPE will be elucidated. MATERIALS AND METHODS Patient samples All of the patients were evaluated by a clinical geneticist prior to enrollment in this study and extensive pedigrees were taken. All had normal chromosomes [except for del(18p) in II.1 pedigree D]. The genomic DNA from HPE patients and family members (when available) was extracted from lymphocytes or established lymphoblastoid cell lines by routine methods. All samples were obtained by informed consent according to the guidelines of our institutional review board. PCR methods, single strand conformational polymorphism (SSCP) analysis and sequencing Human SHH is composed of three exons. We used six pairs of primers (exon 2 and exon 3 were split into two and three overlapping amplicons, respectively) that encompassed the intron– exon boundaries and coding regions of the entire SHH gene. Five primer pairs are published elsewhere (8,23). Amplification of amplicons 3a and 3b was performed using the primer pairs 3F1a/11 (3a) and 3F2/3R1 (3b). Because of the high GC content of the 3′-end of exon 3 (3c), we had been unable to satisfactorily complete the mutation analysis of this exon (23). For the remainder of the 3′-end of the gene we modified our original strategy (36) as follows: 3F3 (5′-GCCCGGGCCAGCGCGTGTACGTGG-3′) and 3R3 (5′-CCCCTCCCCCGGCCCCCCGGCTTC-3′) were used to amplify 3c under PCR conditions of 94°C for 5 min, 94°C for 45 s, 60°C for 30 s, 72°C for 45 s for 35 cycles, followed by 72°C for 6 min. This reaction yielded a 483 bp product that on digestion with BssHII generated two fragments of 249 and 234 bp. Amplification of 3c was performed in a 15 µl reaction volume, using 60–100 ng

Human Molecular Genetics, 1999, Vol. 8, No. 13 2487 DNA template, 200 µM each dATP, dGTP and dTTP, 125 µM dCTP, 3.5 µCi [α-32P]dCTP (800 Ci/mmol, 10 mCi/ml), 30 pmol each primer, 1.5 µl 10× PCR buffer (Gibco, Gaithersburg, MD), 1.25 µl 10× PCR Enhancer (Gibco), 1.5 mM MgSO4 (Gibco) and 1 U AmpliTaq polymerase (Perkin Elmer, Foster City, CA). All of the PCR reactions were performed in a PTC-100 thermal cycler (MJ Research, Waltham, MA). SSCP analysis was performed as described elsewhere (37). Sequencing of the amplicons demonstrating SSCP band shifts was performed by the Protein and DNA Core Facility of the Children’s Hospital of Philadelphia on an ABI Prism 377 analyzer. None of the 13 detected SSCP alterations was found in >200 control chromosomes from unrelated normal Caucasian individuals or in >200 chromosomes from unrelated normal Hispanic individuals. Subcloning to confirm deletion sizes For the four deletions detected, we subcloned the PCR products from the affected individuals using a TA Cloning kit (Invitrogen, Carlsbad, CA). Plasmid DNA containing the mutated allele was extracted (spin miniprep; Qiagen, Valencia, CA) and sequenced. Paternity testing In two kindreds (Fig. 1H and I) paternity testing was performed to confirm the identity of the biological father, using 10 polymorphic microsatellite markers from 10 different chromosomes. After amplification, the samples were run on a denaturing (8.3 M urea) polyacrylamide gel and analyzed by autoradiography. ACKNOWLEDGEMENTS We are grateful to all of the families for their participation in this study. This work was supported in part by the Catholic University of Rome, Italy (L.N.), by NIH grant HD01218 (J.E.M.), by NIH grants HD28732 and HD29862 and by the Division of Intramural Research, NHGRI, NIH (M.M.). REFERENCES 1. Ming, J.E. and Muenke, M. (1998) Holoprosencephaly: from Homer to Hedgehog. Clin. Genet., 53, 155–163. 2. DeMyer, W., Zeman, W. and Palmer, C.G. (1964) The face predicts the brain: diagnostic significance of median facial anomalies for holoprosencephaly (arhinencephaly). Pediatrics, 34, 256–263. 3. Muenke, M. (1994) Holoprosencephaly as a genetic model for normal craniofacial development. Dev. Biol., 5, 293–301. 4. Cohen, M.M.Jr (1989) Perspectives on holoprosencephaly: Part I. Epidemiology, genetics, and syndromology. Teratology, 40, 211–235. 5. Roessler, E. and Muenke, M. (1998) Holoprosencephaly: a paradigm for the complex genetics of brain development. J. Inherit. Metab. Dis., 21, 481–497. 6. Frézal, J. and Schinzel, A. (1991) Report of the committee on clinical disorders, chromosome aberration, and uniparental disomy. Cytogenet. Cell Genet., 58, 986–1052. 7. Belloni, E., Muenke, M., Roessler, E., Traverso, G., Siegel-Bartelt, J., Frumkin, A., Mitchell, H.F., Donis-Keller, H., Helms, C., Hing, A.V., Heng, H.H.Q., Koop, B., Martindale, D., Rommens, J.M., Tsui, L.C. and Scherer, S.W. (1996) Identification of Sonic Hedgehog as a candidate gene responsible for holoprosencephaly. Nature Genet., 14, 353–356. 8. Roessler, E., Belloni, E., Gaudenz, K., Jay, P., Berta, P., Scherer, S.W., Tsui, L.-C. and Muenke, M. (1996) Mutations in the human Sonic Hedgehog gene cause holoprosencephaly. Nature Genet., 14, 357–360.

9. Brown, S.A., Brown, L.Y., Yu, C.C., Warburton, D. and Muenke, M. (1998) ZIC2, a human homologue of odd-paired, is in the 13q32 critical deletion region and mutations are associated with holoprosencephaly. Am. J. Hum. Genet., 63, A3 (abstract). 10. Brown, S.A., Warburton, D., Brown, L.Y., Yu, C., Roeder, E.R., StengelRutkowski, S., Hennekam, R.C.M. and Muenke, M. (1998) Holoprosencephaly due to mutations in ZIC2, a homologue of Drosophila odd-paired. Nature Genet., 20, 180–183. 11. Wallis, D.E., Roessler, E., Hehr, U., Nanni, L., Wiltshire, T., RichieriCosta, A., Gillessen-Kaesbach, G., Zackai, E.H., Rommens, J. and Muenke, M. (1999) Mutations in the homeodomain of the human SIX3 gene cause holoprosencephaly. Nature Genet., 22, 196–198. 12. Echelard, Y., Epstein, D.J., St-Jacques, B., Shen, L., Mohler, J., McMahon, J.A. and McMahon, A.P. (1993) Sonic hedgehog, a member of a family of putative signaling molecules, is implicated in the regulation of CNS polarity. Cell, 75, 1417–1430. 13. Roelink, H., Augsburger, A., Heemskerk, J., Korzh, V., Norlin, S., Ruiz i Altaba, A., Tanabe, Y., Placzek, M., Edlund, T., Jessell, T.M. and Dodd, J. (1994) Floor plate and motor neuron induction by vhh-1, a vertebrate homolog of hedgehog expressed by the notochord. Cell, 76, 761–775. 14. Chiang, C., Litingtung, Y., Lee, E., Young, K.E., Corden, J.L., Westphal, H. and Beachy, P.A. (1996) Cyclopia and defective axial patterning in mice lacking Sonic hedgehog gene function. Nature, 383, 407–413. 15. Lee, J.J., Ekker, S.C., von Kessler, D.P., Porter, J.A., Sun, B.I. and Beachy, P.A. (1994) Autoproteolysis in hedgehog protein biogenesis. Science, 266, 1528–1537. 16. Porter, J.A., von Kessler, D.P., Ekker, S.C., Young, K.E., Lee, J.J., Moses, K. and Beachy, P.A. (1995) The product of hedgehog autoproteolytic cleavage is active in local and long-range signalling. Nature, 374, 363–366. 17. Porter, J.A., Ekker, S.C., Park, W.-J., von Kessler, D.P., Young, K.E., Chen, C.-H., Ma, Y., Woods, A.S., Cotter, R.J., Koonin, E.V. and Beachy, P.A. (1996) Hedgehog patterning activity: role of lipophilic modification mediated by the carboxy-terminal autoprocessing domain. Cell, 86, 21–34. 18. Porter, J.A., Young, K.E. and Beachy, P.A. (1996) Cholesterol modification of hedgehog signaling proteins in animal development. Science, 274, 255–259. 19. Fan, C.-M., Porter, J.A., Chiang, C., Chang, D.T., Beachy, P.A. and Tessier-Lavigne, M. (1995) Long-range sclerotome induction by Sonic hedgehog: direct role of the amino-terminal cleavage product and modulation by the cyclic AMP signalling pathway. Cell, 81, 457–465. 20. Fietz, M.J., Jacinto, A., Taylor, A.M., Alexandre, C. and Ingham, P.W. (1995) Secretion of the amino-terminal fragment of the Hedgehog protein is necessary and sufficient for hedgehog signalling in Drosophila. Curr. Biol., 5, 643–649. 21. Marti, E., Bumcrot, D.A., Takada, R. and McMahon, A.P. (1995) Requirement of the 19K form of sonic hedgehog for induction of distinct ventral cell types in CNS explants. Nature, 375, 322–325. 22. Martin, G. (1996) Pass the butter. Science, 274, 203–204. 23. Roessler, E., Belloni, E., Gaudenz, K., Vargas, F., Scherer, S.W., Tsui, L.-C. and Muenke, M. (1997) Mutations in the C-terminal domain of Sonic Hedgehog cause holoprosencephaly. Hum. Mol. Genet., 6, 1847–1853. 24. Odent, S., Attié Bitach, T., Blayau, M., Mathieu, M., Martin, A., Augé, J., Delezoïde, A.L., LeGall, J.Y., LeMarec, B., David, V. and Vekemans, M. (1998) SHH mutations associated with variable clinical presentation in four families and its expression during early human development. Am. J. Hum. Genet., 63, A169 (abstract). 25. Gripp, K.W., Edwards, M.C., Mowat, D., Meinecke, P., Richieri-Costa, A., Zackai, E.H., Elledge, S. and Muenke, M. (1998) Mutations in the transcription factor TGIF in holoprosencephaly. Am. J. Hum. Genet., 63, A32 (abstract). 26. Hall, T.M.T., Porter, J.A., Young, K.E., Koonin, E.V., Beachy, P.A. and Leahy, D.J. (1997) Crystal structure of a hedgehog autoprocessing domain: homology between hedgehog and self-splicing proteins. Cell, 91, 85–97. 27. Roessler, E., Ward, D.E., Gaudenz, K., Belloni, E., Scherer, S.W., Donnai, D., Siegel-Bartelt, J., Tsui, L.-C. and Muenke, M. (1997) Cytogenetic rearrangements involving the loss of the Sonic Hedgehog gene at 7q36 cause holoprosencephaly. Hum. Genet., 100, 172–181. 28. Hall, T.M.T., Porter, J.A., Beachy, P.A. and Leahy, D.J. (1995) A potential catalytic site revealed by the 1.7-Å crystal structure of the amino-terminal signaling domain of Sonic hedgehog. Nature, 378, 212–216. 29. Bertolino, E., Wildt, S., Richards, G. and Clerc, R.G. (1996) Expression of a novel murine homeobox gene in the developing cerebellar external granular layer during its proliferation. Dev. Dyn., 205, 410–420.


30. Sulik, K.K., Dehart, D.B., Rogers, J.M. and Chernoff, N. (1995) Teratogenicity of low doses of all-trans retinoic acid in presomite mouse embryos. Teratology, 51, 398–403. 31. Chang, B.-E., Blader, P., Fisher, N., Ingham, P.W. and Strähle, U. (1997) Axial (HNF3β) and retinoic acid receptors are regulators of the zebrafish sonic hedgehog promoter. EMBO J., 16, 3955–3964. 32. Helms, J.A., Kim, C.H., Hu, D., Minkoff, R., Thaller, C. and Eichele, G. (1997) Sonic hedgehog participates in craniofacial morphogenesis and is down-regulated by teratogenic doses of retinoic acid. Dev. Biol., 187, 25–35. 33. Ming, J.E., Roessler, E. and Muenke, M. (1998) Human developmental disorders and the Sonic hedgehog pathway. Mol. Med. Today, 8, 343–349. 34. Kelley, R.L., Roessler, E., Hennekam, R.C., Feldman, G.L., Kosaki, K., Jones, M.C., Palumbos, J.C. and Muenke, M. (1996) Holoprosencephaly in RSH/Smith–Lemli–Opitz syndrome: does abnormal cholesterol metabolism affect the function of Sonic hedgehog? Am. J. Med. Genet., 66, 478–484.

35. Fitzky, B.U., Witsch-Baumgartner, M., Erdel, M., Lee, J.N., Paik, Y.K., Glossmann, H., Utermann, G. and Moebius, F.F. (1998) Mutations in the delta-7-sterol reductase gene in patients with the Smith–Lemli–Opitz syndrome. Proc. Natl Acad. Sci. USA, 95, 8181–8186. 36. Vargas, F.R., Roessler, E., Gaudenz, K., Belloni, E., Whitehead, A.S., Kirke, P.N., Mills, J.L., Hooper, G., Stevenson, R.E., Cordeiro, I., Correia, P., Felix, T., Gereige, R., Cunningham, M.L., Canún, S., Antonarakis, S.E., Strachan, T., Tsui, L.-C., Scherer, S.W. and Muenke, M. (1998) Analysis of the human Sonic Hedgehog coding and promotor regions in sacral agenesis, triphalangeal thumb, and mirror polydactyly. Hum. Genet., 102, 387–392. 37. Muenke, M., Schell, U., Hehr, A., Robin, N.H., Losken, H.W., Schinzel, A., Pulleyn, L.J., Rutland, P., Teardon, W., Malcolm, S. and Winter, R. (1994) A common mutation in the fibroblast growth factor receptor 1 gene in Pfeiffer syndrome. Nature Genet., 8, 269–274.