Cleidocranial Dysplasia

A number sign (#) is used with this entry because of evidence that cleidocranial dysplasia (CCD) is caused by heterozygous loss-of-function mutation in the RUNX2 gene (600211), encoding transcription factor CBFA1, on chromosome 6p21.

Heterozygous duplication in RUNX2 resulting in a gain of function causes metaphyseal dysplasia and maxillary hypoplasia with or without brachydactyly (MDMHB; 156510).

Description

The main clinical features of CCD include persistently open skull sutures with bulging calvaria, hypoplasia or aplasia of the clavicles permitting abnormal facility in apposing the shoulders, wide pubic symphysis, short middle phalanx of the fifth fingers, dental anomalies, and often vertebral malformation.

See 168550 for a discussion of the combination of cleidocranial dysplasia and parietal foramina. Pycnodysostosis (265800) and mandibuloacral dysplasia (248370) are disorders to be considered in the differential diagnosis of cleidocranial dysplasia. Acroosteolysis and bone sclerosis with tendency to fracture are differentiating features of pycnodysostosis.

Mundlos (1999) provided a review of the clinical features of cleidocranial dysplasia and the molecular basis of this disorder.

Clinical Features

One of the most colorful families was described by Jackson (1951). The condition occurred in many descendants of a Chinese man named Arnold who embraced the Mohammedan religion and 7 wives in South Africa. Jackson (1951) was able to trace 356 descendants, of whom 70 were affected by the 'Arnold head.' For translation of original description by Marie and Sainton (1898), see Bick (1968). Ramesar et al. (1996) estimated that more than 1,000 descendants of the first progenitor now have the disorder.

A family with delayed eruption of deciduous and permanent teeth reported by Arvystas (1976) probably had cleidocranial dysplasia.

Dore et al. (1987) described a 34-year-old woman with cleidocranial dysostosis and scoliosis diagnosed at age 13 years. The scoliosis continued to progress after skeletal maturation. Syringomyelia was diagnosed at the age of 34. The authors noted reports of 2 previous patients with cleidocranial dysostosis and syringomyelia and suggested that this association may be a more common problem than generally recognized.

Jensen (1990) studied development in 7 males and 10 females, aged 5 to 46 years, with CCD; 11 were followed longitudinally. Height and radius length were decreased, especially in females. Longitudinal data showed growth retardation and slightly retarded skeletal maturation throughout childhood. The metacarpophalangeal pattern profile demonstrated great variation in bone length, presumably resulting from extra epiphyses in metacarpals II and V and from multiple cone-shaped epiphyses. Jensen (1990) concluded that CCD is a generalized skeletal dysplasia. Chitayat et al. (1992) described the range of variability in affected members in 3 generations of a family. The propositus presented with respiratory distress due to a narrow thorax. The clavicles were hypoplastic with discontinuity in the central portions. A 17-year-old aunt of the proposita showed large fontanels and multiple wormian bones as well as a wide symphysis pubis with hypoplasia of the iliac bones. The 25-year-old mother of the proposita showed typical hand abnormalities by x-ray: thin metacarpal and metatarsal diaphyses of digits 2 to 5 and short middle phalanx of fingers 2 and 5. The grandmother likewise showed wormian bones. On the basis of a review of 13 patients, Reed and Houston (1993) concluded that underossification of the hyoid bone could be added to the delayed ossification that affects the skull, teeth, pelvis, and extremities in CCD.

Other Features

Cooper et al. (2001) assembled a series of 90 CCD individuals and 56 relative controls ascertained from genetic and dental practices in the United States, Canada, Europe, and Australia. A number of previously unrecognized complications were significantly increased, including: genua valga, scoliosis, pes planus, sinus infections, upper respiratory complications, recurrent otitis media, and hearing loss. Primary cesarean section rate was significantly increased compared to relative controls and the general population rate. Dental abnormalities, including supernumerary teeth, failure of exfoliation of the primary dentition, and malocclusion, were found to be serious and complex problems that required intervention. Several clinical recommendations based on the findings of this study were presented: a hearing evaluation at birth and during early childhood; a team approach to the management of dental abnormalities on a long-term basis with the overall goal to provide an esthetic facial appearance and functioning occlusion by late adolescence or early adulthood; medical and surgical evaluation for the consequences of delayed craniofacial development (obstructive sleep apnea and sinusitis and otitis); and evaluation for submucous cleft palate.

Morava et al. (2002) reported a mother and daughter with CCD who also had biochemical signs of hypophosphatasia (see 241500; 146300), including decreased levels of alkaline phosphatase (171760). Both patients had a heterozygous mutation in the RUNX2 gene (600211.0012), and the authors concluded that the mutation caused secondary features of hypophosphatasia. Unger et al. (2002) reported a similar patient with CCD and osteopenia with decreased serum alkaline phosphatase. Radiographically, she had a Bowdler spur of the right fibula and was initially diagnosed with hypophosphatasia. Reevaluation at age 11 years revealed findings of classic CCD, although no mutations were identified in the RUNX2 gene. Unger et al. (2002) concluded that osteopenia, osteoporosis, and decreased alkaline phosphatase may be underemphasized findings in CCD, but likely occur only in a minority of patients.

Cogulu et al. (2004) described horseshoe kidney, hypospadias, and undescended testis in a patient with CCD.

Inheritance

Cleidocranial dysplasia is an autosomal dominant disorder.

However, Goodman et al. (1975) reported a family in which 2 brothers with cleidocranial dysplasia were born to unaffected first-cousin parents; he also reported a case born from a niece/uncle union. Several older reports of affected sibs with presumably normal parents were reviewed by Lasker (1946). Goodman et al. (1975) supported recessive inheritance of CCD in the cases he studied.

Zackai et al. (1997) presented a family with 2 affected sisters born to nonconsanguineous and unaffected parents. They suggested germline mosaicism as the most likely mechanism.

Pal et al. (2007) reported 2 brothers and a maternal half-brother with CCD confirmed by genetic analysis. Initial DNA testing in the unaffected mother did not detect the mutation, but further testing using heteroduplex analysis applying high-resolution melting analysis, followed by subcloning, detected low-level somatic mosaicism in maternal blood and buccal swab. The findings indicated germline mosaicism in the mother as the inheritance mechanism for CCD in this family.

Cytogenetics

Nienhaus et al. (1993) proposed that the CCD gene is located on either the long or the short arm of chromosome 6. They observed a male patient with a pericentric inversion of chromosome 6 and classic CCD together with mild to moderate mental retardation, hearing deficiency, and unusual facial appearance.

Narahara et al. (1995) observed CCD in association with a t(6;18)(p12;q24) translocation.

Mapping

In 2 kindreds with typical features of CCD, Mundlos et al. (1995) used the candidate gene approach to map the disorder to chromosome 6p. Linkage was established between CCD and 4 loci--D6S426, D6S451, D6S459, and TCTE1 (186975)--that span a region of 10 cM on 6p. One highly polymorphic microsatellite from this region, D6S459, showed allelic loss in all affected members of 1 family with 2 different sets of primers. The presence of a deletion in this area was confirmed by Southern blot analysis using a probe derived from the amplification product of the D6S459 marker. Thus, the CCD gene was assigned to 6p21.

Feldman et al. (1995) performed linkage studies in 5 families with CCD, including 24 affected and 20 unaffected individuals, using microsatellite markers spanning 2 candidate regions on chromosomes 8q and 6. The strongest support for linkage was with the 6p marker D6S282, with a 2-point lod score of 4.84 at theta = 0.03. The multipoint lod score was 5.70 for location in the 19-cM interval between D6S282 and D6S291. Feldman et al. (1995) pointed out that the gene for bone morphogenetic protein-6 (BMP6; 112266) is located on chromosome 6 and that comparative mapping based on mouse-human homology (Copeland et al., 1993) would place BMP6 on human 6p, thus making BMP6 a candidate gene for CCD.

Gelb et al. (1995) confirmed linkage of CCD to 6p21. Based on their data and those described by Mundlos et al. (1995), they further refined the localization of CCD to a 6-cM region of 6p21 that includes a microdeletion at D6S459. Ramesar et al. (1996) investigated the original family from South Africa and also showed linkage to 6p21.3-p21.1. The maximum lod score was 7.14 at theta = 0.00 with marker D6S459. Using their own and previous mapping data, they refined the localization of the CCD gene to a 4- to 5-cM region between D6S451 and D6S465.

Molecular Genetics

Mundlos et al. (1997) found the linkage to 6p21 in studies of 3 additional large families with 39 affected members. The region in which the refined localization placed the gene was covered by 14 yeast artificial chromosomes (YACs). Three known genes were identified within the contig: TCTE1 (186975), MUT (609058), and CBFA1 (600211). CBFA1 was a reasonable candidate gene for CCD because a member of the 'runt' family had previously been described as a bone-specific nuclear-matrix-binding transcription factor. By fluorescence in situ hybridization to YACs, they confirmed the presence of a deletion on 6p in 1 family and enabled them to narrow the region to approximately 1.5 Mb. They also studied the patient with a pericentric inversion involving 6p21-q16 previously documented by Nienhaus et al. (1993) and found results supporting the assignment. Thus, in some families, the phenotype segregated with deletions, resulting in heterozygous loss of CBFA1.

In other families, Mundlos et al. (1997) found insertion, deletion, and missense mutations leading to translational stop codons in the DNA-binding domain or in the C-terminal transactivating region of the CBFA1 protein (see, e.g., 600211.0001; 600211.0003). In-frame expansion of a polyalanine stretch segregated in an affected family with brachydactyly and minor clinical findings of CCD; see 600211.0003. Heterozygous loss of function of CBFA1 appeared to be sufficient to produce CCD.

In 29 patients with CCD from 19 unrelated families, Baumert et al. (2005) sequenced the RUNX2 gene and identified 12 different RUNX2 mutations. They examined phenotypic data using homogeneity analysis and observed mild to full-blown expression of the CCD phenotype, with intrafamilial clinical variability. Baumert et al. (2005) commented that homogeneity analysis simplified grouping the patients into distinct entities, but noted that the analysis separated individuals with the same mutation, emphasizing the clinical variability within the patient cohort.

El-Gharbawy et al. (2010) studied a 7-year-old boy with CCD complicated by severe progressive kyphoscoliosis, who also displayed features of hypophosphatasia (see 241500), including Bowdler spurs, severe osteopenia, and low alkaline phosphatase. After no RUNX2 mutation was found by sequencing, the authors performed array CGH and identified a 50- to 70-kb deletion that predicted a disruption of the C terminus of RUNX2, encompassing the coding sequence for amino acids 327 to 521 and involving the SMAD 1,2,3,5 binding sites and the nuclear matrix targeting signal (NMTS) regions. El-Gharbawy et al. (2010) emphasized the need to search for deletions when sequencing of the target gene is normal, and noted that the C-terminal region of RUNX2 appears to play an integral role in human osteogenesis and osteoblast differentiation.

Genotype/Phenotype Correlations

To correlate CBFA1 mutations in different functional domains with the CCD clinical spectrum, Zhou et al. (1999) studied 26 independent cases of CCD, and a total of 16 new mutations were identified in 17 families. Most mutations were de novo missense mutations that affected conserved residues in the runt domain and completely abolished both DNA binding and transactivation of a reporter gene. These, and mutations that resulted in premature termination in the runt domain, produced a classic CCD phenotype by abolishing transactivation of the mutant protein with consequent haploinsufficiency. Zhou et al. (1999) further identified 3 putative hypomorphic mutations that resulted in a clinical spectrum including classic and mild CCD, as well as an isolated dental phenotype characterized by delayed eruption of permanent teeth (600211.0010). Functional studies showed that 2 of the 3 mutations were hypomorphic in nature and 2 were associated with significant intrafamilial variability in expressivity, including isolated dental anomalies without the skeletal features of CCD. Together these data showed that variable loss of function due to alterations in the runt and C-terminal proline/serine/threonine-rich (PST) activation domains of CBFA1 may give rise to clinical variability, including classic CCD, mild CCD, and isolated primary dental anomalies.

Pathogenesis

Zheng et al. (2005) observed growth plate abnormalities in a patient with a 1-bp insertion (600211.0013) in the RUNX2 gene. Histologic analysis of the rib and long-bone cartilages showed a markedly diminished zone of hypertrophy; analysis of limb cartilage RNA revealed a 5- to 10-fold decrease in the hypertrophic chondrocyte molecular markers VEGF (192240), MMP13 (600108), and COL10A1 (120110). Zheng et al. (2005) concluded that humans with CCD have altered endochondral ossification due to altered RUNX2 regulation of hypertrophic chondrocyte-specific genes during chondrocyte maturation.

Nomenclature

Sillence et al. (1987) proposed the gene symbol CCD for the mutation in both mouse and man; this symbol has also been used for 'central core disease' (117000).

Animal Model

Sillence et al. (1987) described cleidocranial dysplasia in mice. The change was radiation-induced and inherited as an autosomal dominant with variable expressivity but almost complete penetrance. The homozygous state was lethal in utero. The features were variable clavicular hypoplasia, delayed closure of cranial fontanels and sutures, and variable hypoplasia of pelvic bones, in particular, ischiopubic rami. Selby et al. (1993) investigated the interactions between 2 unlinked genes causing a semidominant skeletal dysplasia in mice: cleidocranial dysplasia (Ccd) and 'short digits' (Dsh). Each mutant is a homozygous lethal. The Ccd mutation was reported by Selby and Selby (1978). Selby et al. (1993) found that mice who were heterozygous for both mutations showed 7 different synergistic interactions, including one that yielded an entirely new abnormality not predicted from any abnormalities found in either of the single homozygotes. Although Selby et al. (1993) did not expect to find antagonistic interactions, they in fact found 3 in the double heterozygote. In all cases, the effects of Dsh were either partly or completely suppressed by Ccd. A classic example of comb shape in chickens in which interaction of 2 mutations at different loci led to a completely new phenotype was cited.

Lou et al. (2009) generated a mouse model of CCD using a hypomorphic Runx2-mutant allele (neo7), in which only part of the transcript is processed to full-length Runx2. Runx2 neo7/neo7 mice expressed a reduced level of wildtype transcript (55 to 70%) and protein and had grossly normal skeletons with no abnormalities observed in the growth plate, but exhibited developmental defects in calvaria and clavicles that persisted through postnatal growth. Clavicle defects were caused by disrupted endochondral bone formation during embryogenesis. These hypomorphic mice had altered calvarial bone volume, as observed by histology and micro-CT imaging, and decreased expression of osteoblast marker genes. Runx2 neo7/+ mice had 79 to 84% of wildtype transcript and exhibited a normal bone phenotype. Lou et al. (2009) concluded that there is a critical gene dosage requirement of Runx2 for the formation of intramembranous bone tissues during embryogenesis and that a decrease to 70% of wildtype Runx2 levels results in the CCD phenotype, whereas levels above 79% produce a normal skeleton, suggesting that the range of bone phenotypes in CCD patients is attributable to quantitative reduction in the functional activity of RUNX2.

History

In a discussion of genetic skeletal dysplasias in the Museum of Pathological Anatomy in Vienna, Beighton et al. (1993) pictured the skeleton of a 25-year-old man with cleidocranial dysplasia who died in 1909 of tuberculous pneumonia. The skeleton showed the characteristic hypoplasia of the clavicles in association with a large, patent anterior fontanel. Other minor features were bilateral genu valgum and slight medial bowing of the tibia and fibula.

Brueton et al. (1992) presented 3 patients (mother, daughter, and an unrelated patient) with congenital clavicular hypoplasia or agenesis thought to have cleidocranial dysplasia due to chromosome 8q22 rearrangements and manifesting micrognathia, exophthalmos, and lack of a generalized skeletal dysplasia. Because 8q22 was disrupted in all 3 patients, Brueton et al. (1992) suggested that the gene(s) responsible for the cleidocranial dysplasia phenotype might be located in that region. However, Mundlos et al. (1995) demonstrated no linkage to the 8q22 region in 2 families with classic cleidocranial dysplasia.