Postweaning Craniofacial Growth in the OIM Mouse Model of Osteogenesis Imperfecta

Osteogenesis imperfecta (OI) type III is a severe genetic disorder of type I collagen (Col1) resulting in bone fragility, reduced stature, and impaired craniofacial growth resulting in midface hypoplasia, dental malocclusions, and macrocephaly. While the adult OI murine (OIM) mouse model exhibits craniofacial phenotypes similar to patients with OI, little is known about the developmental trajectories of these phenotypes. To investigate the mechanisms by which Col1 mutations alter postnatal craniofacial growth, we analyzed the phenotype of the OIM mouse from the age of weaning until adulthood (skeletal maturity). OIM and wild-type (WT) littermates were scanned in-vivo with a Skyscan 1176 micro-CT system at 4 weeks (weaning) and 16 weeks (adulthood). 3D landmarks were collected using 3D Slicer software. Centroid size (a proxy for craniomandibular size) was compared using Mann-Whitney U tests. Morphological analysis for shape variation, including Generalized Procrustes analysis (GPA) and principal component analyses (PCA), were performed using the "geomorph" package in R. Procrustes ANOVAs were used to test for significant differences in craniomandibular shape between the genotypes. Morphological disparity was estimated as the Procrustes variance and statistically compared using the morphol.disparity function in R. Craniomandibular centroid sizes were significantly smaller in the OIM mice than the WT mice at both weeks 4 and 16 (p< 0.010). When the effects of size were accounted for by the GPA, significant shape differences were present (p< 0.002) throughout growth. For their size, both juvenile and adult OIM mice had shortened midfaces and increased cranial vault dimensions (relative macrocephaly) compared to WT littermates. Morphological differences were seen around the rostrum, temporal crests, and zygomatic arches. Marked vertical expansion of the neurocranium at the sagittal and coronal sutures presented with a concomitant basicranial shortening in the juvenile OIM mice only. Compared to WT, both juvenile and adult OIM mice had mandibles that were longitudinally shortened and mediolaterally wide. Morphological differences were seen around the incisal ramus and angular process at both ages, and in the coronoid process of adult OIM mice. Significantly different (p=0.001) morphological disparity between the genotypes existed only at the juvenile stage, demonstrating an overall decrease in variance throughout postnatal development period. This suggests that mice with greater variances in cranial shape experienced higher rates of attrition. This is potentially related to increased craniofacial fracture rates seen in the OIM mice. These results suggest that while the OIM craniofacial phenotype differs significantly from the WT throughout postweaning growth, functional constraints (such as feeding performance) may limit the degree of potential phenotypic divergence. Furthermore, while previous work in human patients with OI has linked the development of macrocephaly to basicranial shortening during early skull development, our results provide limited support for this hypothesis due to the absence of basicranial changes in adult OIM mice. Future investigations will examine perinatal and pre-weaning growth in the OIM mouse model to better understand the development of these craniofacial dysmorphologies, and to identify optimal growth windows during which interventions might recover bone quality and growth trajectories in patients with OI.