Mineral-associated proteins have been proposed to regulate many aspects of biomineralization, including the location, type, orientation, shape, and texture of crystals. To understand how proteins achieve this exquisite level of control, we are studying the interaction between the phosphoprotein osteopontin (OPN) and the biomineral calcium oxalate monohydrate (COM). In the present study, we have synthesized peptides corresponding to amino acids 220-235 of rat bone OPN (pSHEpSTEQSDAIDpSAEK), one of several highly phosphorylated, aspartic-, and glutamic acid-rich sequences found in the protein. To investigate the role of phosphorylation in interaction with crystals, peptides containing no (P0), one (P1), or all three (P3) phosphates were prepared. Using a novel combination of confocal microscopy and scanning electron microscopy, we show that these peptides adsorb preferentially to {100} faces of COM and inhibit growth of these faces in a phosphorylation-dependent manner. To characterize the mechanism of adsorption of OPN peptides to COM, we have performed the first atomic-scale molecular-dynamics simulation of a protein-crystal interaction. P3 adsorbs to the {100} face much more rapidly than P1, which in turn adsorbs more rapidly than P0. In all cases, aspartic and glutamic acid, not phosphoserine, are the amino acids in closest contact with the crystal surface. These studies have identified a COM face-specific adsorption motif in OPN and delineated separate roles for carboxylate and phosphate groups in inhibition of crystal growth by mineral-associated phosphoproteins. We propose that the formation of close-range, stable, and face-specific interactions is a key factor in the ability of phosphoproteins to regulate biomineralization processes.

Biomineralization is characterized by a high degree of control over the location, nature, size, shape, and orientation of the crystals formed. For many years, it has been widely believed that the exquisitely precise nature of crystal formation in biological tissues is the result of stereochemically specific interactions between growing crystals and extracellular matrix proteins. That is, the ability of many mineralized tissue proteins to adsorb to particular faces of biominerals has been attributed to a steric and electrical complementarity between periodic regions of the polypeptide chain and arrays of ions on the crystal face. In recent years, however, evidence has accumulated that many mineral-associated proteins lack periodic structure even when adsorbed to crystals. It also appears that protein-crystal interactions involve a general electrostatic attraction rather than arrays of complementary charges. In the present work, we review these studies and present some relevant new findings involving the mineral-modulating phosphoprotein osteopontin. Using molecular dynamics simulations, we show that the adsorption of osteopontin peptides to hydroxyapatite crystals does not involve a unique conformation of the peptide molecule, and that the adsorbed peptides are not aligned with rows of Ca(2+) ions on the crystal face. Further, we show that the interface between osteopontin peptides and calcium oxalate monohydrate crystals consists of peptide regions of high electronegativity and crystal faces of high electropositivity. Collectively, the above-mentioned studies suggest that interactions between mineral-modulating proteins and biologically relevant crystals are primarily electrostatic in nature, and that molecular disorder assists these proteins in forming multiple bonds with cations of the crystal face.

In vitro studies have shown that the phosphoprotein osteopontin (OPN) inhibits the nucleation and growth of hydroxyapatite (HA) and other biominerals. In vivo, OPN is believed to prevent the calcification of soft tissues. However, the nature of the interaction between OPN and HA is not understood. In the computational part of the present study, we used molecular dynamics simulations to predict the adsorption of 19 peptides, each 16 amino acids long and collectively covering the entire sequence of OPN, to the {100} face of HA. This analysis showed that there is an inverse relationship between predicted strength of adsorption and peptide isoelectric point (P<0.0001). Analysis of the OPN sequence by PONDR (Predictor of Naturally Disordered Regions) indicated that OPN sequences predicted to adsorb well to HA are highly disordered. In the experimental part of the study, we synthesized phosphorylated and non-phosphorylated peptides corresponding to OPN sequences 65–80 (pSHDHMDDDDDDDDDGD) and 220–235 (pSHEpSTEQSDAIDpSAEK). In agreement with the PONDR analysis, these were shown by circular dichroism spectroscopy to be largely disordered. A constant-composition/seeded growth assay was used to assess the HA-inhibiting potencies of the synthetic peptides. The phosphorylated versions of OPN65-80 (IC50 = 1.93 µg/ml) and OPN220-235 (IC50 = 1.48 µg/ml) are potent inhibitors of HA growth, as is the nonphosphorylated version of OPN65-80 (IC50 = 2.97 µg/ml); the nonphosphorylated version of OPN220-235 has no measurable inhibitory activity. These findings suggest that the adsorption of acidic proteins to Ca2+-rich crystal faces of biominerals is governed by electrostatics and is facilitated by conformational flexibility of the polypeptide chain.

Osteopontin (OPN) inhibits the growth of calcium oxalate monohydrate (COM) and other crystal phases in a phosphorylation-dependent manner. In the present study, the role of OPN phosphate groups in adsorption to, incorporation into and inhibition of COM crystals was studied by comparing OPN isoforms differing in phosphorylation. OPN isoforms purified from rat bone (bOPN), which contains 10 phosphates, and cow milk (mOPN), which contains 25 phosphates, were compared with rat recombinant OPN (rOPN), which is not phosphorylated. Using fluorescence-labeled proteins and confocal microscopy, we show that mOPN and rOPN, like bOPN, adsorb preferentially to the edges between {100} and {121} faces of preformed COM crystals, and to a lesser extent to the {100} and {121} faces. Using scanning electron microscopy, we show that growth of COM in the presence of bOPN or mOPN results in a ‘dumbbell’ morphology, whereas crystals grown with rOPN are only slightly affected. COM crystals grown in the presence of low concentrations of fluorescence-labeled bOPN incorporate the protein into the crystal lattice. In crystals imaged in the {010} plane, incorporation of bOPN results in a cross-shaped pattern of fluorescence, consistent with preferential adsorption to {100}/{121} edges throughout the growth process.

Mice lacking the gene encoding matrix gla protein (MGP) exhibit massive mineral deposition in blood vessels and die soon after birth. We hypothesize that MGP prevents arterial calcification by adsorbing to growing hydroxyapatite (HA) crystals. To test this, we have used a combined experimental-computational approach. We synthesized peptides covering the entire sequence of human MGP, which contains three sites of serine phosphorylation and five sites of γ-carboxylation, and studied their effects on HA crystal growth using a constant-composition autotitration assay. In parallel studies, the interactions of these sequences with the {100} and {001} faces of HA were analyzed using atomistic molecular dynamics (MD) simulations. YGlapS (amino acids 1-14 of human MGP) and SK-Gla (MGP43-56) adsorbed rapidly to the {100} and {001} faces and strongly inhibited HA growth (IC(50) = 2.96 μg/mL and 4.96 μg/mL, respectively). QR-Gla (MGP29-42) adsorbed more slowly and was a moderate growth inhibitor, while the remaining three (nonpost-translationally modified) peptides had little or no effect in either analysis. Substitution of gla with glutamic acid reduced the adsorption and inhibition activities of SK-Gla and (to a lesser extent) QR-Gla but not YGlapS; substitution of phosphoserine with serine reduced the inhibitory potency of YGlapS. These studies suggest that MGP prevents arterial calcification by a direct interaction with HA crystals that involves both phosphate groups and gla residues of the protein. The strong correlation between simulated adsorption and measured growth inhibition indicates that MD provides a powerful tool to predict the effects of proteins and peptides on crystal formation.