An E. coli over-expression system for multiply-phosphorylated proteins and its use in a study of calcium phosphate sequestration by novel recombinant phosphopeptides

Abstract

Phosphoproteins and phosphopeptides were expressed by E. coli to give yields of 30–200 mg of purified protein per litre with an average degree of phosphorylation at multiple sites of 61–83%. The method employed two compatible cohabiting plasmids having low and high copy number, expressing a protein kinase and, more abundantly, the substrate (poly)peptide, respectively. It was used to phosphorylate recombinant β-casein or osteopontin at multiple casein kinase-2 sites. Two constructs were designed to produce shorter peptides containing one or more clusters of phosphorylation sites resembling the phosphate centres of caseins. In the first, a 53-residue 6-His tagged phosphopeptide was expressed at a 5-fold higher molar yield. The second had multiple tandem repeats of a tryptic phosphopeptide sequence to give a similar increase in efficiency. Each recombinant phosphopeptide was purified (30–100 mg) and small-angle X-ray scattering measurements showed that they, like certain casein and osteopontin phosphopeptides, sequester amorphous calcium phosphate to form calcium phosphate nanoclusters. In principle, the method can provide novel phosphopeptides for the control of biocalcification or be adapted for use with other kinases and cognate proteins or peptides to study the effect of specific phosphorylations on protein structure. Moreover, the insertion of a phosphate centre sequence, possibly with a linker peptide, may allow thermodynamically stable, biocompatible nanoparticles to be made from virtually any sequence.