Use of site-directed mutagenesis to enhance the epitope-shielding effect of covalent modification of proteins with polyethylene glycol

Abstract

Modification by covalent attachment of polyethylene glycol (PEG) can reduce the immunogenicity and prolong the circulating life of proteins, but the utility of this approach for any protein is restricted by the number and distribution of PEG attachment sites (e.g., epsilon-amino groups of lysine residues). We have developed a strategy for introducing additional sites for PEG attachment by using site-directed mutagenesis to selectively replace arginine with lysine codons and tested it with purine nucleoside phosphorylase (PNP) from Escherichia coli, an extremely stable but immunogenic enzyme, that could potentially be used to treat an inherited deficiency of PNP. A triple mutant, RK3, possessing three Arg----Lys substitutions was constructed that increased the number of lysines per PNP subunit from 14 to 17, providing an additional 18 potential PEG attachment sites per hexameric enzyme molecule. The wild-type and RK3 enzymes had similar catalytic activity, antigenicity, and immunogenicity. After PEG modification, both enzymes retained catalytic activity, the plasma half-life of both enzymes in mice increased from approximately 4 hr to 4 days, and the binding of both enzymes by antisera raised against each unmodified enzyme was markedly diminished. However, antibody raised against wild-type PEG-PNP did not bind the PEG-RK3 enzyme. PEG-RK3 PNP was also substantially less immunogenic than wild-type PEG-PNP. Accelerated antibody-mediated clearance of PEG-PNP occurred in 2 of 12 mice treated with PEG-RK3 PNP, compared with 10 of 16 mice treated with the modified wild-type enzyme. This combined use of directed mutagenesis and PEG modification is aimed at permitting the widest choice of proteins, including products of genetic and chemical "engineering," to be used for therapy of inherited and acquired disorders.