Osteogenesis imperfecta (OI) is a genetically and clinically heterogeneous disorder characterized by bone fragility and reduced bone mass generally caused by defects in type I collagen structure or defects in proteins interacting with collagen processing. We identified a homozygous missense mutation in SEC16B in a child with vertebral fractures, leg bowing, short stature, muscular hypotonia, and bone densitometric and histomorphometric features in keeping with OI with distinct ultrastructural features. In line with the putative function of SEC16B as a regulator of trafficking between the ER and the Golgi complex, we showed that patient fibroblasts accumulated type I procollagen in the ER and exhibited a general trafficking defect at the level of the ER. Consequently, patient fibroblasts exhibited ER stress, enhanced autophagosome formation, and higher levels of apoptosis. Transfection of wild-type SEC16B into patient cells rescued the collagen trafficking. Mechanistically, we show that the defect is a consequence of reduced SEC16B expression, rather than due to alterations in protein function. These data suggest SEC16B as a recessive candidate gene for OI.

Osteogenesis imperfecta (OI) is an inherited genetic disorder characterized by frequent bone fractures and reduced bone mass. Most cases of OI are caused by dominantly inherited heterozygous mutations in one of the two genes encoding type I collagen, COL1A1 and COL1A2 . Here we describe a five-year-old boy with typical clinical, radiological and bone ultrastructural features of OI type I. Establishing the molecular genetic cause of his condition proved difficult since clinical exome and whole exome analysis was repeatedly reported negative. Finally, manual analysis of exome data revealed a silent COL1A2 variant c.3597 T > A (NM_000089.4), which we demonstrate activates a cryptic splice site. The newly generated splice acceptor in exon 50 is much more accessible than the wild-type splice-site between the junction of exon 49 and 50, and results in an in-frame deletion of 24 amino acids of the C-terminal propeptide. In vitro collagen expression studies confirmed cellular accumulation and decreased COL1A2 secretion to 45%. This is the first report of a cryptic splice site within the coding region of COL1A2. which results in abnormal splicing causing OI. The experience from this case demonstrates that routine diagnostic approaches may miss cryptic splicing mutations in causative genes due to the lack of universally applicable algorithms for splice-site prediction. In exome-negative cases, in-depth analysis of common causative genes should be conducted and trio-exome analysis is recommended.