No abstract

BackgroundEditing deaminases have a pivotal role in cellular physiology. A notable member of this superfamily, APOBEC3G (A3G), restricts retroviruses, and Activation Induced Deaminase (AID) generates antibody diversity by localized deamination of cytosines in DNA. Unconstrained deaminase activity can cause genome-wide mutagenesis and cancer. The mechanisms that protect the genomic DNA from the undesired action of deaminases are unknown. Using the in vitro deamination assays and expression of A3G in yeast, we show that replication protein A (RPA), the eukaryotic single-stranded DNA (ssDNA) binding protein, severely inhibits the deamination activity and processivity of A3G.Principal Findings/MethodologyWe found that mutations induced by A3G in the yeast genomic reporter are changes of a single nucleotide. This is unexpected because of the known property of A3G to catalyze multiple deaminations upon one substrate encounter event in vitro. The addition of recombinant RPA to the oligonucleotide deamination assay severely inhibited A3G activity. Additionally, we reveal the inverse correlation between RPA concentration and the number of deaminations induced by A3G in vitro on long ssDNA regions. This resembles the “hit and run” single base substitution events observed in yeast.SignificanceOur data suggest that RPA is a plausible antimutator factor limiting the activity and processivity of editing deaminases in the model yeast system. Because of the similar antagonism of yeast RPA and human RPA with A3G in vitro, we propose that RPA plays a role in the protection of the human genome cell from A3G and other deaminases when they are inadvertently diverged from their natural targets. We propose a model where RPA serves as one of the guardians of the genome that protects ssDNA from the destructive processive activity of deaminases by non-specific steric hindrance.

Mammalian Pol ι has an unusual combination of properties: it is stimulated by Mn2+ ions, can bypass some DNA lesions and misincorporates “G” opposite template “T” more frequently than incorporates the correct “A.” We recently proposed a method of detection of Pol ι activity in animal cell extracts, based on primer extension opposite the template T with a high concentration of only two nucleotides, dGTP and dATP (incorporation of “G” versus “A” method of Gening, abbreviated as “misGvA”). We provide unambiguous proof of the “misGvA” approach concept and extend the applicability of the method for the studies of variants of Pol ι in the yeast model system with different cation cofactors. We produced human Pol ι in baker's yeast, which do not have a POLI ortholog. The “misGvA” activity is absent in cell extracts containing an empty vector, or producing catalytically dead Pol ι, or Pol ι lacking exon 2, but is robust in the strain producing wild-type Pol ι or its catalytic core, or protein with the active center L62I mutant. The signature pattern of primer extension products resulting from inaccurate DNA synthesis by extracts of cells producing either Pol ι or human Pol η is different. The DNA sequence of the template is critical for the detection of the infidelity of DNA synthesis attributed to DNA Pol ι. The primer/template and composition of the exogenous DNA precursor pool can be adapted to monitor replication fidelity in cell extracts expressing various error-prone Pols or mutator variants of accurate Pols. Finally, we demonstrate that the mutation rates in yeast strains producing human DNA Pols ι and η are not elevated over the control strain, despite highly inaccurate DNA synthesis by their extracts.

Inosine triphosphate pyrophosphatase (ITPA), a key enzyme involved in maintaining the purity of cellular nucleoside triphosphate pools, specifically recognizes inosine triphosphate and xanthosine triphosphate (including the deoxyribose forms) and detoxifies them by catalyzing the hydrolysis of a phosphoanhydride bond, releasing pyrophosphate. This prevents their inappropriate use as substrates in enzymatic reactions utilizing (d)ATP or (d)GTP. A human genetic polymorphism leads to the substitution of Thr for Pro32 (P32T) and causes ITPA deficiency in erythrocytes, with heterozygotes having on average 22.5% residual activity, and homozygotes having undetectable activity. This polymorphism has been implicated in modulating patients’ response to mercaptopurines and ribavirin. Human fibroblasts containing this variant have elevated genomic instability upon treatment with base analogs. We find that the wild-type and P32T forms are dimeric in solution and in the crystal structure. This abolishes the previous speculation that the P32T change disrupts dimerization as a mechanism of inactivation. The only difference in structure from the wild-type protein is that the area surrounding Thr32 is disrupted. Phe31 is flipped from the hydrophobic core out into the solvent, leaving a hole in the hydrophobic core of the protein which likely accounts for the reduced thermal stability of P32T ITPA and ultimately leads to its susceptibility to degradation in human cells. Circular dichroism and thermal denaturation studies confirm these structural results. We propose that the dimer of P32T variant subunit with wild-type subunit is degraded in cells similarly to the P32T homodimer explaining the level of loss of ITPA activity in heterozygotes.

No abstract