Human DNA primase synthesizes short RNA primers that DNA polymerase α then elongates during the initiation of all new DNA strands. Even though primase misincorporates NTPs at a relatively high frequency, this likely does not impact the final DNA product since the RNA primer is replaced with DNA. We used an extensive series of purine and pyrimidine analogues to provide further insights into the mechanism by which primase chooses whether or not to polymerize a NTP. Primase readily polymerized a size-expanded cytosine analogue, 1, 3-diaza-2-oxo-phenothiazine-NTP, across from a templating G but not across from A. The enzyme did not efficiently polymerize NTPs incapable of forming two Watson-Crick hydrogen bonds with the templating base with the exception of UTP opposite purine deoxyribonucleoside. Likewise, primase did not generate base-pairs between two nucleotides with altered Watson-Crick hydrogen bonding patterns. Examining the mechanism of NTP polymerization revealed that human primase can misincorporate NTPs via both template misreading and a primer-template slippage mechanism. Together, these data demonstrate that human primase strongly depends on Watson-Crick hydrogen bonds for efficient nucleotide polymerization, much more so than the mechanistically related herpes primase, and provide insights into the potential roles of primer-template stability and base tautomerization during misincorporation.
KeywordsFidelity; misincorporation; hydrogen bonds; Watson-Crick; DNA polymerase; kinetics DNA polymerases lack the capacity to initiate the synthesis of new DNA strands de novo on single-stranded DNA templates. Rather, they require the 3' end of a pre-existing primer bound to the template in order to replicate the template. In most replicative systems, DNA primases solve this problem by synthesizing short RNA primers that a replicative DNA polymerase then elongates (1,2). In eukaryotes, primase synthesizes a short RNA primer (8 -12 nucleotides long) that DNA polymerase α (pol α) elongates by another about 20 nucleotides to generate a DNA primer(3). Pol α dissociates and either pol δ or pol ε finishes the bulk of DNA replication on the leading and lagging strands (4,5). After completion of Okazaki fragment synthesis on the lagging strand, the RNA primer is removed and replaced with DNA. Consequently, errors during primer synthesis do not become part of the genome and, therefore, primase need not † This work was supported by NIH grant GM54194 to R.D.K. and grants from the Academy of Sciences of the Czech Republic (Z4 055 905), the Ministry of Education (LC 512) and the Grant Agency of the ASCR (IAA400550902) to M. H.