Debaenst, S.; Sahd, L.; Debaene, C.; Jarayseh, T.; De Saffel, H.; Bek, J. W.; Steyaert, W.; Van Haver, D.; Dufour, S.; Impens, F.; Coucke, P.; De Clercq, A.; Willaert, A.

Phenotypic variability is common in human diseases, even when the same genes are affected. In this study, three zebrafish models of Osteogenesis Imperfecta (OI) with dominant glycine substitutions in type I collagen genes (col1a1amh13/+, col1a1adc124/+, and col1a2mh15/+) were characterized for phenotypic severity and variability, using a newly developed standardized scoring system. Comprehensive analyses of the vertebral columns in these models revealed histological and ultrastructural differences that corresponded with phenotypic severity. Increasing skeletal severity correlated with a higher incidence of skeletal deformities and abnormalities. This, in turn, was associated with thinner bones and increased disorganization of collagen fibrils, fiber accumulation and mineralization, elastin deposits, and increased cell proliferation in the notochord and intervertebral ligament (IVL). Additionally, osteoblast function and bone regenerative capacity were increasingly compromised. These characteristics, combined with genetic information, have the potential to predict the severity of phenotypic outcomes in dominant forms of OI, caused by mutations in type I collagen. A remarkable intra-familial phenotypic variability in the col1a2mh15/+ mutant holds potential for future approaches that could help in understanding the underlying mechanisms of this variability and the identification of modifier genes. Finally, through proteomics analysis three potential protein biomarkers (HSP47, Col8a1, and Bcan) were identified, that could serve as indicators of disease severity. These biomarkers not only have diagnostic value, but will allow stratification by OI type, have predictive value towards progression of the clinical presentation and will play a role in treatment guidance. Validation in human tissue samples will further reveal their clinical relevance.