Type I collagen is a long, triple helical protein, which forms the matrix of bone, skin and other tissues. During the last year, we continued characterization of structural consequences of Osteogenesis Imperfecta (OI) mutations in the type I triple helix. We expanded the map of changes in the triple helix stability (melting temperature), which now includes over four dozen mutations at different sites along the helix. To relate this map to peptide-based stability predictions, we proposed a model for extracting the activation energy of local helix unfolding from the reported peptide data. We tested the model and determined its parameters by measuring the H-D exchange rate for glycine NH residues involved in inter-chain hydrogen bonds in collagen. From comparison of the peptide-based and expanded mutation-based stability maps, we refined our model of regional variations in the triple helix structure. The refined regions align with regions important for collagen fibril assembly and ligand binding, and they appear to contribute to at least some of the observed regional variations in OI phenotype.? ? Disruption of collagen interactions with ligands, particularly other extracellular matrix proteins and proteoglycans, is one of possible mechanisms relating structural defects in collagen triple helix to functional abnormalities in OI. In the last year, we completed development of a novel confocal microscopy assay for visualizing and quantitatively measuring binding of various matrix proteins to individual collagen fibrils based on differential fluorescent labeling of collagen and the ligand. With this assay, e.g., we found that mammalian collagenases attack damaged and poorly assembled collagen fibers via preferential binding to microfibrils, which become more exposed at fiber defects. Another study revealed that decorin and biglycan bind to surfaces of collagen fibrils of different size with the same affinity, suggesting that they regulate the fibril size by favoring larger surface area rather than a specific fibril diameter. These measurements also demonstrated similar binding of decorin and biglycan to fibrils reconstituted from collagen with truncated terminal peptides (telopeptides), suggesting that the telopeptides do not contain or block important binding sites on fibril surfaces. We are now focusing on measurement of decorin and biglycan interactions with fibers containing mutant collagen molecules that cause OI and other connective tissue disorders.? ? The overwhelming majority of severe OI cases are caused by single amino acid substitutions. However, several recessive OI and EDS cases were described, in which all type I collagen was synthesized in the form of alpha-1 homotrimer rather than the normal heterotrimer of two alpha-1 and one alpha-2 chains. Formation of type I homotrimers was also observed in cultures of breast cancer cells and in cultures of normal bone cells from individuals predisposed to common, age-related osteoporosis. Our previous studies revealed altered regional stability of the homotrimer triple helix and a small increase in the denaturation temperature of the whole molecule, but they did not offer clues to potential mechanisms of pathology. Over the last two years, we made a potential breakthrough. We discovered that while hetero- and homotrimers coassemble within the same fibrils, they segregate at a subfibrillar level, potentially forming separate microfibrils. Our measurements revealed a significant reduction in the homotrimer cleavage rate by major tissue collagenases, MMP-1 and MMP-13. More detailed examination suggested that the initial MMP binding to collagen is not affected by the lack of the alpha-2 chain, but that this chain is essential for the next step of triple helix unwinding and opening, preceding the cleavage. In tissues with mixed collagen composition, e.g., containing type I homo- and heterotrimers or type I homotrimers and type III collagen, the abnormal cleavage rate of one component will alter the normal remodeling process. In preliminary experiments, live imaging of the degradation of mixed hetero/homotrimer fibrils by confocal microscopy with differential fluorescence labeling provided first direct evidence of such abnormal remodeling. We believe that it may play an important role in various pathologies associated with homotrimer synthesis and hope that better understanding of the underlying molecular mechanism may help to develop new treatment strategies.? ? Most but not all cases with clinical symptoms of OI are caused by mutations in type I collagen. In collaboration with clinical researchers from Bone and Extracellular Matrix Branch of NICHD, we reported that severe/lethal skeletal pathology reminiscent of OI is also caused by recessive null mutations in the cartilage associated protein (CRTAP) or prolyl-3-hydroxylase (P3H1). CRTAP and P3H1 form a tight three-protein complex with cyclophilin B in the Endoplasmic Reticulum (ER). We found that the disruption of this complex significantly delays type I procollagen folding, resulting in overhydroxylation and overglycosylation of Lys residues, which is also observed in many OI cases. Further measurements conducted during the last year revealed abnormally high rate of synthesis of the collagen chains and delayed secretion of the mature protein by dermal fibroblasts from patients lacking P3H1. The combination of faster chain synthesis, slower triple helix folding, and delayed secretion causes accumulation of up to five times the normal amount of collagen within cells. The resulting ER swelling likely leads to ER stress contributing to the severe/lethal outcome in the patients.
Showing the most recent 10 out of 31 publications
