Development of a bacterial screen for novel hypoxanthine-guanine phosphoribosyltransferase substrates

The lack of de novo purine biosynthesis in many parasitic protozoans makes the enzymes in the salvage of purines attractive chemotherapeutic targets. Hypoxanthine-guanine phosphoribosyltransferase (HGPRT) is a key enzyme for purine salvage and bacterial complementation screens for HGPRT inhibitors are known. The low KMS for purine bases makes purine analogs unattractive as competitive inhibitors for this enzyme. Despite the availability of many crystal structures of HGPRTs, it is only recently that selective inhibitors of the enzyme have been developed. Therefore, novel purine analogs which act as substrates for the HGPRT reaction and thereby inhibit downstream enzymes or get incorporated into the nucleotide pool are an attractive altenative for drug design. We have used a combination of two E. coli strains Sphi606 (ara, deltapro-gpt-lac, thi, hpt) and Sphi609 (ara, deltapro-gpt-lac, thi, hpt, pup, purH,J, strA) to identify inhibitors and substrates of HGPRT. E. coli Sphi609 is deficient in both de novo synthesis as well as salvage enzymes of purine nucleotide synthesis, while E. coli Sphi606 is deficient in salvage enzymes only. Hence, expression of functional HGPRTs in E. coli Sphi606 grown in minimal medium makes it susceptible to HGPRT substrates, which inhibit downstream processes. Growth of E. coli Sphi609 in minimal medium can be made conditional for the expression of a functional HGPRT and this growth would be susceptible to both HGPRT substrate analogs and inhibitors. A substance that strictly acts as an inhibitor will affect growth of transformed E. coli Sphi609 only. For this purpose, we compared the human and P. falciparum enzymes with known HGPRT substrate analogs. Our data with 6-mercaptopurine, 6-thioguanine and allopurinol show that these compounds act by being substrates for HGPRT. Our results with allopurinol suggest that it is a better substrate for P. falciparum HGXPRT than the human enzyme. Therefore, species-specific substrates can be tested out successfully in E. coli Sphi606. The formation of products from substrates like allopurinol lacking a labile proton at N7 raises the possibility that the deprotonation of substrates might occur at N9 rather than at N7 or a purine anion might be the true substrate for the reaction.