The composite structure of the mammalian skull, which forms predominantly via intramembranous ossification, requires precise pre- and post-natal growth regulation of individual calvarial elements. Disturbances of this process frequently cause severe clinical manifestations in humans. Enhanced DNA binding by a mutant MSX2 homeodomain results in a gain of function and produces craniosynostosis in humans. Here we show that Msx2-deficient mice have defects of skull ossification and persistent calvarial foramen. This phenotype results from defective proliferation of osteoprogenitors at the osteogenic front during calvarial morphogenesis, and closely resembles that associated with human MSX2 haploinsufficiency in parietal foramina (PFM). Msx2-/- mice also have defects in endochondral bone formation. In the axial and appendicular skeleton, post-natal deficits in Pth/Pthrp receptor (Pthr) signalling and in expression of marker genes for bone differentiation indicate that Msx2 is required for both chondrogenesis and osteogenesis. Consistent with phenotypes associated with PFM, Msx2-mutant mice also display defective tooth, hair follicle and mammary gland development, and seizures, the latter accompanied by abnormal development of the cerebellum. Most Msx2-mutant phenotypes, including calvarial defects, are enhanced by genetic combination with Msx1 loss of function, indicating that Msx gene dosage can modify expression of the PFM phenotype. Our results provide a developmental basis for PFM and demonstrate that Msx2 is essential at multiple sites during organogenesis.

The genetic analysis of congenital skull malformations provides insight into normal mechanisms of calvarial osteogenesis. Enlarged parietal foramina (PFM) are oval defects of the parietal bones caused by deficient ossification around the parietal notch, which is normally obliterated during the fifth fetal month. PFM are usually asymptomatic, but may be associated with headache, scalp defects and structural or vascular malformations of the brain. Inheritance is frequently autosomal dominant, but no causative mutations have been identified in non-syndromic cases. We describe here heterozygous mutations of the homeobox gene MSX2 (located on 5q34-q35) in three unrelated families with PFM. One is a deletion of approximately 206 kb including the entire gene and the others are intragenic mutations of the DNA-binding homeodomain (RK159-160del and R172H) that predict disruption of critical intramolecular and DNA contacts. Mouse Msx2 protein with either of the homeodomain mutations exhibited more than 85% reduction in binding to an optimal Msx2 DNA-binding site. Our findings contrast with the only described MSX2 homeodomain mutation (P148H), associated with craniosynostosis, that binds with enhanced affinity to the same target. This demonstrates that MSX2 dosage is critical for human skull development and suggests that PFM and craniosynostosis result, respectively, from loss and gain of activity in an MSX2-mediated pathway of calvarial osteogenic differentiation.

Throughout its complex morphogenesis, the vertebrate skull must at once protect the brain and expand to accommodate its growth. A key structural adaptation that allows this dual role is the separation of the bony plates of the skull with sutures, fibrous joints that serve as growth centers and allow the calvarial bones to expand as the brain enlarges. Craniosynostosis, the premature fusion of one or more calvarial bones with consequent abnormalities in skull shape, is a common developmental anomaly that disrupts this process. We found previously that a single amino acid substitution in the homeodomain of the human MSX2 gene is associated with the autosomal dominant disorder craniosynostosis, Boston type. This mutation enhances the affinity of Msx2 for its target sequence, suggesting that the mutation acts by a dominant positive mechanism. Consistent with this prediction, we showed that general overexpression of Msx2 under the control of the broadly expressed CMV promoter causes the calvarial bones to invade the sagittal suture. Here we use tissue-specific overexpression of Msx2 within the calvarial sutures to address the developmental mechanisms of craniosynostosis and skull morphogenesis. We demonstrate that a segment of the Msx2 promoter directs reporter gene expression to subsets of cells within the sutures. In late embryonic and neonatal stages, this promoter is expressed in undifferentiated mesenchymal cells medial to the growing bone. By P4, promoter activity is reduced in the suture, exhibiting a punctate pattern in undifferentiated osteoblastic cells in the outer margin of the osteogenic front. Overexpression of Msx2 under the control of this promoter is sufficient to enhance parietal bone growth into the sagittal suture by P6. This phenotype is preceded by an increase in both the number and the BrdU labeling of osteoblastic cells in the osteogenic fronts of the calvarial bones. These findings suggest that an important early event in MSX2-mediated craniosynostosis in humans is a transient retardation of osteogenic cell differentiation in the suture and a consequent increase in the pool of osteogenic cells.

Mutations in the FGFR1-FGFR3 and TWIST genes are known to cause craniosynostosis, the former by constitutive activation and the latter by haploinsufficiency. Although clinically achieving the same end result, the premature fusion of the calvarial bones, it is not known whether these genes lie in the same or independent pathways during calvarial bone development and later in suture closure. We have previously shown that Fgfr2c is expressed at the osteogenic fronts of the developing calvarial bones and that, when FGF is applied via beads to the osteogenic fronts, suture closure is accelerated (Kim, H.-J., Rice, D. P. C., Kettunen, P. J. and Thesleff, I. (1998) Development 125, 1241–1251). In order to investigate further the role of FGF signalling during mouse calvarial bone and suture development, we have performed detailed expression analysis of the splicing variants of Fgfr1-Fgfr3 and Fgfr4, as well as their potential ligand Fgf2. The IIIc splice variants of Fgfr1-Fgfr3 as well as the IIIb variant of Fgfr2 being expressed by differentiating osteoblasts at the osteogenic fronts (E15). In comparison to Fgf9, Fgf2 showed a more restricted expression pattern being primarily expressed in the sutural mesenchyme between the osteogenic fronts. We also carried out a detailed expression analysis of the helix-loop-helix factors (HLH) Twist and Id1 during calvaria and suture development (E10-P6). Twist and Id1 were expressed by early preosteoblasts, in patterns that overlapped those of the FGF ligands, but as these cells differentiated their expression dramatically decreased. Signalling pathways were further studied in vitro, in E15 mouse calvarial explants. Beads soaked in FGF2 induced Twist and inhibited Bsp, a marker of functioning osteoblasts. Meanwhile, BMP2 upregulated Id1. Id1 is a dominant negative HLH thought to inhibit basic HLH such as Twist. In Drosophila, the FGF receptor FR1 is known to be downstream of Twist. We demonstrated that in Twist(+/)(−) mice, FGFR2 protein expression was altered. We propose a model of osteoblast differentiation integrating Twist and FGF in the same pathway, in which FGF acts both at early and late stages. Disruption of this pathway may lead to craniosynostosis.