Holoprosencephaly 3

A number sign (#) is used with this entry because of evidence that holoprosencephaly-3 (HPE3) is caused by heterozygous mutation in the SHH gene (600725), which encodes the human Sonic hedgehog homolog, on chromosome 7q36.

For a phenotypic description and a discussion of genetic heterogeneity of holoprosencephaly, see HPE1 (236100).

Clinical Features

Berry et al. (1984) and Johnson (1989) provided information on a family (family 2 in Johnson, 1989) in which holoprosencephaly occurred in 2 sibs and their first cousin, who were offspring of parents with a single central maxillary incisor. Johnson (1989) reported a second patient (family 1) with full-blown holoprosencephaly whose mother and sister had only a single central maxillary incisor. Johnson (1989) suggested that holoprosencephaly is a developmental field defect of which the mild forms can be single median incisor, hypotelorism, bifid uvula, or pituitary deficiency.

Ardinger and Bartley (1988) reported a family in which 3 individuals in 3 successive generations had severe brain anomalies and 12 individuals had minor manifestations, mainly microcephaly. Other findings in the family included single central incisor and hypotelorism, which have been suggested as mild manifestations of autosomal dominant familial holoprosencephaly.

Nanni et al. (1999) presented a panel of 12 photographs illustrating the range of severity in holoprosencephaly resulting from mutation in the SHH gene.

Marini et al. (2003) studied a family, previously described by Camera et al. (1992), in which the mother presented with a single central maxillary incisor and mild hypotelorism and her daughter and 2 fetuses were diagnosed with HPE. Sequencing of DNA in this family identified a nonsense mutation in the SHH gene (600725.0019).

By detailed ophthalmologic examination of 5 patients with genetically confirmed HPE3, Pineda-Alvarez et al. (2011) found several subtle abnormalities, including refractory errors, small corneal diameter, coloboma, foveal hypoplasia, blepharoptosis, hyperopia, strabismus, and astigmatism. These findings occurred without brain malformations; the patients had single central incisors, microcephaly, hypotelorism, and depressed nasal bridge; 1 had hypoplasia of the left frontal lobe. The patients were part of a larger cohort of 10 patients with genetically confirmed HPE. All had at least 2 ophthalmologic anomalies, including refractive errors, microcornea, microphthalmia, blepharoptosis, exotropia, and coloboma. The findings contributed to the understanding of the phenotypic variability of the HPE spectrum and showed that subtle intraocular abnormalities can occur in HPE.


Pfitzer and Muntefering (1968) observed 4 affected children whose mothers were relatives and had the same anomalous karyotype thought to represent balanced translocation between chromosome 3 and a chromosome of the C group. With the introduction of G-banding techniques, Pfitzer et al. (1982) demonstrated that this reciprocal 7/C-translocation was a balanced rearrangement between the short arm of chromosome 3 and the distal part of the long arm of chromosome 7--t(3;7)(p23;q36). Burrig et al. (1989) demonstrated a fifth case of cyclopia in this family, detected prenatally, and showed an unbalanced karyotype attributable to the above-mentioned balanced translocation. They could find no reports of cyclopia associated with similar chromosome abnormalities.

Lurie et al. (1990) pointed out that at least 9 cases of HPE have occurred in patients with confirmed loss of 7q34-q36. They reported balanced rearrangements involving 7q in 2 mothers examined after the birth of their nonkaryotyped infants with HPE and hydronephrosis. They suggested that in both infants del(7q) was the most probable cause of HPE. Cyclopia and cebocephaly were conspicuous features in the cases of del(7q). Sporadic cases of cyclopia have been observed in association with trisomy 13, ring 13, and other chromosomal abnormalities and many have had normal karyotypes. Masuno and Orii (1990) also pointed to reports of holoprosencephaly in association with terminal 7q deletion.

Kleczkowska et al. (1990) described the case of a female fetus with hemilobar holoprosencephaly and 46,XX,der(7)t(7;8)(q36.1;p12)mat karyotype. The holoprosencephaly sequence was considered to be related to the distal 7(q36.1-qter) deficiency. Hatziioannou et al. (1991) reviewed the evidence suggesting that a locus for holoprosencephaly resides at or near 7q36.

Gurrieri et al. (1993) characterized the 7q deletions in 13 HPE patients and constructed a high resolution physical map of 7q32-qter. They defined the HPE minimal critical region in 7q36 between D7S292 and D7S392. They pictured one of the patients with the characteristic facies of the severe form of HPE which included a single fused eye (cyclopia) and a nose-like structure (proboscis) above the eye. Midline structures of the forebrain were absent, consistent with alobar HPE.

Belloni et al. (1996) refined the position of the HPE3 locus by detailed characterization of HPE3 patients with rearrangements involving chromosome 7q36. They also established a contig of genomic clones in this region. Belloni et al. (1996) demonstrated that a cDNA for SHH, the human Sonic hedgehog homolog, showed specific hybridization to the contig which spanned the translocation breakpoint. Further analysis revealed that SHH mapped approximately 250 and 15 kb centromeric of T1 and T2, respectively (T1 and T2 represent the translocation breakpoints in 2 unrelated patients with a mild form of HPE3). Belloni et al. (1996) proposed that the chromosomal rearrangements remove distal cis-acting regulatory elements or exert long-range position effects causing aberrant expression of the gene. They noted that HPE patients exhibiting deletions of the SHH region are generally more severely affected than are the translocation patients. The mild HPE phenotype displayed by the patient with the T2 balanced translocation included premaxillary aplasia with midline cleft lip, hypotelorism, sensorineural hearing loss, lack of tooth eruption, and cervical cord compression due to stenosis.

Benzacken et al. (1997) reported 4 new cases of holoprosencephaly in fetuses with abnormal karyotypes. Three of these had terminal deletions of 7q, confirming the importance of 7q36 in holoprosencephaly. The fourth fetus had an apparently balanced de novo translocation, t(7;13)(q21.2;q33), without any visible loss of the distal part of chromosome 7q. Benzacken et al. (1997) proposed either a long range positional effect or the existence of genes involved in prosencephalon development at 7q21.2 or 13q33 as an explanation for this.

Nowaczyk et al. (2000) reported an infant with holoprosencephaly, sacral anomalies, and situs ambiguus associated with partial monosomy 7q/trisomy 2p, der(7)t(2;7)(p23.2;q36.1), as a result of an adjacent-1 segregation of a t(2;7) in the father. The chromosomal abnormality was diagnosed prenatally after sonographic detection of HPE in the fetus. The baby was born at 37 weeks' gestation and died in the neonatal period; he had dysmorphic features consistent with the HPE sequence. Postmortem examination showed semilobar HPE, abdominal situs ambiguus, multiple segments of bowel atresia, dilatation of the ureters, and bony sacral anomalies. Molecular analysis confirmed hemizygosity for the SHH and HLXB9 (142994) genes, which were thought to be responsible for the HPE and sacral phenotypes, respectively. Immunohistochemical studies showed intact dopaminergic pathways in the mesencephalon, suggesting that midbrain dopamine neuron induction requires only one functioning SHH allele.


Muenke et al. (1993) performed linkage studies in 10 families with autosomal dominant HPE. The phenotypic features in affected individuals varied from the most severe forms with single brain ventricle and cyclopia to milder forms with ocular hypotelorism and midface hypoplasia to clinically unaffected carriers. Under the most conservative model-free analysis, linkage between HPE and D7S22 showed a combined lod score of 7.2 at theta = 0.0, with 1 family independently presenting a lod score of 3.0 at theta = 0.0. Muenke et al. (1993) concluded that autosomal dominant HPE is at the locus that has been designated HPE3 and mapped to 7q36.

Muenke et al. (1994) suggested that mutations in the HPE3 gene are responsible for both sporadic HPE and a majority of families with autosomal dominant HPE. Clinical evaluation of the affected individuals in the 9 families in the report of Muenke et al. (1994) confirmed the previously reported phenotypic variability of autosomal dominant HPE. In each family, one or more obligate gene carriers had classic (alobar, semilobar, or lobar) HPE, many of whom died during early infancy. Others had HPE microforms such as microcephaly, mental retardation, microphthalmia, ocular coloboma, ocular hypotelorism, midface hypoplasia, single central upper incisor, cleft lip, and cleft lip and palate. Some obligate gene carriers had normal phenotypes, including normal intellect. In 1 of the 9 families, linkage to D7S22 and other markers on chromosome 7q was excluded, thus indicating genetic heterogeneity. The clinical manifestations, including the HPE microforms, did not differ between individuals in the unlinked kindred and those in the other 8 kindreds linked to 7q36.

Molecular Genetics

Roessler et al. (1996) identified SHH as the gene responsible for HPE3. They analyzed 30 autosomal dominant HPE families and found 5 families that segregated different heterozygous SHH mutations. Two of these mutations predict premature termination of SHH protein (600725.0002 and 600725.0003). The remaining 3 mutations altered highly conserved residues in the vicinity of the alpha helix-1 motif (600725.0004 and 600725.0005) or the signal cleavage site (600725.0001). Roessler et al. (1996) noted that in humans loss of one SHH allele is sufficient to cause HPE, whereas in the mouse both alleles need to be lost to produce a similar CNS phenotype. They observed that haploinsufficiency for SHH in human is sufficient to disturb ventral midline neurogenesis but is insufficient to cause ventralization defects of sclerotome or limb abnormalities.

In 30 unrelated children with holoprosencephaly, Orioli et al. (2001) analyzed for mutations in the SIX3 (603714), SHH, TGIF (602630), and ZIC2 (603073) genes. They identified 3 novel mutations, 2 in the SHH gene and 1 in the ZIC2 gene. Their results explained 8% (2 of 26 newborn samples) of the HPE cases in the South American population studied.

Among 94 fetuses with HPE and a normal karyotype, Bendavid et al. (2006) used quantitative multiplex PCR of short fluorescent fragments (QMPSF) to screen for microdeletions in the 4 major HPE genes, SHH, SIX3, ZIC2, and TGIF. Microdeletions were identified in 8 (8.5%) fetuses: 2 in SHH, 2 in SIX3, 3 in ZIC2, and 1 in TGIF. Further analysis showed that the entire gene was missing in each case. Point mutations in 1 of the 4 genes were identified in 13 of the fetuses. Combining the instances of point mutations and microdeletions for the 94 cases yielded the following percentages: SHH (6.3%), ZIC2 (8.5%), SIX3 (5.3%), and TGIF (2%). Bendavid et al. (2006) reported the use of 2 complementary assays for HPE-associated submicroscopic deletions: a multicolor fluorescence in situ hybridization (FISH) assay using probes for the 4 major HPE genes and 2 candidate genes (DISP1, 607502 and FOXA2, 600288) followed by quantitative PCR to selected samples. Microdeletions for SHH, ZIC2, SIX3, or TGIF were found in 16 of 339 severe HPE cases (i.e., with CNF findings; 4.7%). In contrast, no deletions were found in 85 patients at the mildest end of the HPE spectrum. Based on their data, Bendavid et al. (2006) suggested that microdeletion testing should be considered as part of an evaluation of holoprosencephaly, especially in severe HPE cases.

Modifier Genes

Martinelli and Fan (2009) demonstrated that a constructed mouse Shh N116K mutant, which corresponds to the HPE3-associated SHH mutation N115K (600725.0020), caused markedly decreased binding to Gas1 (139185), resulting in decreased Shh signaling. These findings indicated that HPE due to the N115K mutation results from an inability of mutant SHH to bind to GAS1 normally, thus interrupting the positive regulatory effect of GAS1. Martinelli and Fan (2009) suggested that mutations in GAS1 may act as possible modifiers of HPE.

In 4 Brazilian patients with HPE or HPE-like phenotype, Ribeiro et al. (2010) identified 4 different heterozygous nonsynonymous variants in the GAS1 gene that were predicted to be damaging. Two of 4 patients also carried heterozygous missense mutations in the SHH gene. The authors suggested that mutations in the GAS1 gene may confer susceptibility to the development of HPE or may act as a modifier locus for HPE in conjunction with mutations in other genes.

Genotype/Phenotype Correlations

Among 34 patients with holoprosencephaly, Dubourg et al. (2004) observed that mutations in the SHH gene were associated with choanal stenosis and ophthalmologic malformations.

Mercier et al. (2011) reported the clinical and molecular features of a large European series of 645 HPE probands (51% fetuses) and 699 relatives in order to examine genotype/phenotype correlations. The facial features were assigned to 4 categories: categories 1 and 2 had severe facial defects, whereas microforms were listed as 3 and 4. SHH mutations were found in 67 (10.4%) probands. The patients had alobar (28%), semilobar (34%), lobar (4%), or microform (34%) HPE, but the 4 categories of facial defects were evenly distributed, although the proportion of coloboma was relatively high (15% for the series as a whole). Extracranial defects were found in 24%, mostly visceral or renal/urinary. The mutations showed high heritability (73%), and 23% of parents with mutations had a microform. Statistical analysis for the whole study showed a positive correlation between the severity of the brain malformation and facial features for those with mutations in the SHH gene, and that microforms were associated with SHH mutations. Based on these results, Mercier et al. (2011) proposed an algorithm for molecular analysis in HPE.


Ardinger and Bartley (1988) reported a family in which the transmission pattern of holoprosencephaly appeared to be autosomal dominant.

Odent et al. (1998) reviewed 258 HPE records involving at least 1 affected child and found 97 cases in 79 families with nonsyndromic, nonchromosomal HPE. A high degree of familial aggregation was found in 29% of families. By segregation analysis, Odent et al. (1998) concluded that autosomal dominant inheritance with incomplete penetrance (82% for major and 88% for major and minor) was the most likely mode of inheritance. Sporadic cases accounted for 68%, and the recurrence risk after an isolated case was predicted to be 13 to 14%.

In familial holoprosencephaly pedigrees, Suthers et al. (1999) reported a skewed sex ratio among transmitting parents with SHH gene mutations, 14 of 16 being mothers (p = 0.002). Suthers et al. (1999) also found that of 16 reported cases of single maxillary central incisor with no other congenital malformation, 13 were female (p = 0.0085). Suthers et al. (1999) concluded that boys with SHH gene mutations may be at greater risk of major malformations outside the central nervous system, thus reducing their reproductive fitness, explaining the observed skewed sex ratio.

Population Genetics

In a targeted screening study of 4 genes in 86 Dutch patients with holoprosencephaly, Paulussen et al. (2010) found that 21 (24%) had heterozygous mutations in 1 of 3 of the genes. Three (3.5%) had mutations in the SHH gene, 9 (10.5%) had mutations in the ZIC2 gene (603073), and 9 (10.5%) had mutations in the SIX3 gene (603714). None had mutations in the TGIF gene (602630). Two deletions were detected, 1 encompassing the ZIC2 gene and another encompassing the SIX3 gene. About half of the mutations were de novo; 1 was germline mosaic. There was marked clinical variability, but those with ZIC2 mutations tended to have less severe facial malformations. Five of 7 parental carriers were asymptomatic, and 2 had minor HPE signs.