The involvement of collagen in bone biomineralization is commonly admitted, yet its role remains unclear. Here we show that type I collagen in vitro can initiate and orientate the growth of carbonated apatite mineral in the absence of any other vertebrate extracellular matrix molecules of calcifying tissues. We also show that the collagen matrix influences the structural characteristics on the atomic scale, and controls the size and the three-dimensional distribution of apatite at larger length scales. These results call into question recent consensus in the literature on the need for Ca-rich non-collagenous proteins for collagen mineralization to occur in vivo. Our model is based on a collagen/apatite self-assembly process that combines the ability to mimic the in vivo extracellular fluid with three major features inherent to living bone tissue, that is, high fibrillar density, monodispersed fibrils and long-range hierarchical organization.

Increasing interest in bio-interfaces for medical, diagnostic, or biotechnology applications has highlighted the critical scientific challenge behind both the understanding and control of protein-solid surface interactions. In this context, this Account focuses on the molecular-level characterization of the interactions of peptides, ranging in size from a few amino acids to long sequences, with metal and oxide surfaces. In this Account, we attempt to fill the gap between the well-known basic studies of the interaction of a single amino acid with well-defined metal surfaces and the investigations aimed at controlling biocompatibility or biofilm growth processes. We gather studies performed with surface science tools and macroscopic characterization techniques along with those that use modeling methods, and note the trends that emerge. Sulfur drives the interaction of cysteine-containing peptides with metal surfaces, particularly gold. Moreover, intermolecular interactions, such as hydrogen bonds may induce surface self assembly and chiral arrangements of the peptide layer. Depending on the solvent pH and composition, carboxylates or amino groups may also interact with the surface, which could involve conformational changes in the adsorbed peptide. On oxide surfaces such as titania or silica, researchers have identified carboxylate groups as the preferential peptide binding groups because of their strong electrostatic interactions with the charged surface. In high molecular weight peptides, systematic studies of their interaction with various oxide surfaces point to the preferential interaction of certain peptide sequences: basic residues such as arginine assume a special role. Researchers have successfully used these observations to synthesize adhesive sequences and initiate biomineralization. Studies of the interaction of peptides with nanoparticles have revealed similar binding trends. Sulfur-containing peptides adhere preferentially to gold nanoparticles. Peptides containing aromatic nitrogen also display a high affinity for various inorganic nanoparticles. Finally, we describe a novel class of peptides, genetically engineered peptides for inorganics (GEPIs), which are selected from a phage display protocol for their high binding affinity for inorganic surfaces. Extended investigations have focused on the mechanisms of the molecular binding of these peptides to solid surfaces, in particular the high binding affinity of some sulfur-free sequences of GEPIs to gold or platinum surfaces. We expect that this clearer view of the possible preferential interactions between peptides and inorganic surfaces will facilitate the development of new, more focused research in various fields of biotechnology, such as biocompatibility, biomimetics, or tissue engineering.