Prediction of peptide-induced silica formation under a wide pH range by molecular descriptors

Silica nanoparticles have demonstrated broad applicability from wastewater treatment to medicine. By increasing intermolecular interaction of silica precursor, peptides can induce the formation of biosilica usually under neutral and basic but not acidic conditions, and here, they are defined as silica-forming peptides (SFPs). In addition, the silica-forming capability of most putative SFPs has not been evaluated. To address both questions, we developed molecular descriptors to simulate and predict the silica-forming capabilities of SFPs under a wide range of pH values. We used van der Waals and electrostatic interactions as molecular descriptors to probe the physisorption between amino acid residues and proto-silicate. We then performed multiple-dimensional simulations using the canonical R5 peptide and three novel SFPs from the brown alga Sargassum fusiforme. Finally, we experimentally validated the silica-forming capability of the four SFPs at pH values from 4 to 8. Our results showed that the molecular descriptors provide accurate predictions on the adsorption of proto-silicate by SFPs. Thus, the method can be used to screen SFPs that are capable of inducing the formation of silica in a wide pH range more importantly including acidic conditions.