Introduction: Craniosynostosis, premature cranial suture fusion, affects 1:2000 children. We are using murine models to study the molecular mechanisms mediating premature cranial suture fusion. In these models, the posterior frontal (PF) cranial suture fuses while all other sutures, including the sagittal (SAG) and coronal (COR) remain patent. Previous work from our laboratory indicates that suture-specific dura mater initiates PF suture fusion via multiple paracrine (e.g. TGF-b1 and FGF-2) signaling pathways. Numerous fate-based studies have demonstrated that cranial suture fusion is controlled by these pro-osteogenic paracrine factors and, moreover, we have found that these same pro-osteogenic cytokines are absent in the patent SAG and COR dura mater. We had long suspected that cranial suture patency was a passive process (i.e. insufficient levels of osteogenic cytokines), however, in this study, we demonstrated that Noggin, a BMP antagonist, actively maintained SAG and COR suture patency. Using a model of COR craniosynostosis, we investigated how premature suture fusion affected the "brake" (i.e. Noggin) on suture fusion. Furthermore, we investigated the effects of Apert and Crouzon gain-of-function FGFR2 mutations on Noggin expression in chimeric dura mater and osteoblasts. Finally, we examined the effects of Noggin misexpression on cranial suture fate.

Methods: In situ hybridization screening was performed on mouse cranial sutures (days 15, 25, 35, and 45) for the BMP antagonists: DCR4, DCR5, DCR6, DCR7, Gremlin, Cerberus, and Noggin. The calvaria of transgenic beta-galactosidase/ Noggin mice were harvested at similar time points, treated with an X-gal solution and histologically examined. PF, SAG and COR dural cell cultures were established and Western blots were performed. The rat COR craniosynostosis model was based a previous published overexpression FGF-2 adenoviral model from our laboratory. Chimeric gain-of-function FGFR2 mutated dura mater and osteoblasts created using retroviral vectors. A Noggin over-expression adenovirus was injected into the PF suture of 22 day-old calvaria in vitro or 3 day-old mice in vivo.

Results: In situ hybridization demonstrated that Noggin is the only BMP antagonist present in murine cranial sutures. X-gal staining of transgenic mice demonstrated no Noggin expression in the PF suture complex before, during or after suture fusion. In contrast, SAG and COR sutures expressed high levels of Noggin throughout the period of predicted suture fusion. PF dural cultures did not produce Noggin. In contrast, SAG and COR dural cultures constitutively produced high levels of Noggin. In our model of COR craniosynostosis, we found that FGF-2 overexpression dramatically suppressed COR Noggin expression. Apert and Crouzon gain-of-function FGFR2 mutations suppressed constitutive SAG dura mater Noggin expression and BMP-4 induced Noggin expression in chimeric osteoblasts. Finally, Noggin misexpression maintained PF suture patency in vitro and in vivo.

Discussion: Since normal cranial suture patency depends on BMP antagonism, some forms of craniosynostoses may result from inappropriate FGF-mediated down-regulation of Noggin expression. Ongoing studies are investigating the effects of a Pfeiffer (Pro250Arg) FGFR1 transgenic mutation on Noggin expression and cranial suture fate in vivo.