Bone Abstracts

ASARM peptide (acidic, serine- and aspartate-rich motif) and osteopontin (OPN) fragments accumulate in X-linked hypophosphatemia patients and/or in the Hyp mouse model and, when phosphorylated, potently inhibit mineralization in osteoblast cultures. To investigate this inhibition, we modeled the binding to hydroxyapatite of the human OPN-ASARM peptide (DDSHQSDESHHSDESDEL) using RosettaSurface computational simulations. Peptide binding to hydroxyapatite atomic planes constructed to have different chemical terminations was computed using a structure-prediction algorithm for peptide–solid surface interactions. {100}, {001} and {010} monoclinic hydroxyapatite planar surfaces were built having different calcium-to-phosphate ratios. Miller indices (hkl) planes (surfaces) were created with mixed–charge to reflect surfaces likely occurring during crystal growth, and leaving intact interfacial phosphate and hydroxyl ions since P-O and O-H bonds are strong and their breaking is energetically unfavourable. Binding affinities, specificities and structure were determined for ASARM-Sp0 (without phosphoserine) and two phosphorylated forms of ASARM (ASARM-Sp3 and ASARM-Sp5, with 3 or 5 phosphoserines). Energy-minimized peptide conformations in solution and adsorbed to mineral were predicted by RosettaSurface. Adsorption data revealed highly significant, phosphorylation-dependent differences in binding energies for the peptides. All peptide conformers were generally unstructured both in solution and upon adsorption. Adsorbed peptides showed a degree of crystal lattice matching via the phosphate and carboxylate groups coordinating with surface calcium; binding to the (100) and (010) terminations showed the highest binding energies. In conclusion, peptide–mineral binding modeling has provided mechanistic data on how OPN and its phosphorylated peptides act as potent inhibitors of mineralization.