Synthesis of nucleotide analogues that potently and selectively inhibit human DNA primase.

DNA primase synthesizes short RNA oligonucleotides that DNA polymerase alpha further elongates in order to initiate the synthesis of all new DNA strands during eukaryotic DNA replication. To develop potent and specific primase inhibitors, we combined 2'-modified sugars with bases incapable of normal Watson-Crick hydrogen bonding. The presence of a 2'-hydroxyl in either the ara or ribo configuration greatly enhances the ability of primase to polymerize a nucleotide. Further modifying the 2'-position by including both a hydroxyl and methyl group at this position greatly reduced the ability of primase to polymerize the resulting nucleotides. Replacing the base of the NTP with analogues incapable of normal Watson-Crick hydrogen bonding (benzimidazole, nitrobenzimidazole, and dichlorobenzimidazole) resulted in compounds that inhibited primase quite well and with similar potency. We synthesized arabinofuranosylbenzimidazole triphosphate (araBTP) and found that this sugar change increased inhibition by 2-4-fold relative to the ribofuranosyl analogue. AraBTP inhibited polymerization of both purines and pyrimidines, although primase polymerized only small amounts of the compound. Interestingly, even though araBTP was not readily polymerized by primase, it inhibited primase almost as potently as araATP, a compound that primase polymerizes extremely rapidly and that results in very strong chain termination. Importantly, this compound was a very weak inhibitor of and only slowly polymerized by DNA polymerase alpha, indicating that it is a specific primase inhibitor. The potential utility and mechanistic implications of these inhibitors are discussed.