Approaches to the affinity labelling of E. coli DNA-dependent RNA polymerase

Repair of the mutagenic DNA oxidation product, 5-formyluracil

The oxidation of the thymine methyl group can generate 5-formyluracil (FoU). Template FoU residues are known to miscode, generating base substitution mutations. The repair of the FoU lesion is therefore important in minimizing mutations induced by DNA oxidation. We have studied the repair of FoU in synthetic oligonucleotides when paired with A and G. In E. coli cell extract, the repair of FoU is four orders of magnitude lower than the repair of U and is similar for both FoU:A and FoU:G base pairs. In HeLa nuclear extract, the repair of FoU:A is similarly four orders of magnitude lower than the repair of uracil, although the FoU:G lesion is repaired 10 times more efficiently than FoU:A. The FoU:G lesion is shown to be repaired by E. coli mismatch uracil DNA glycosylase (Mug), thermophile mismatch thymine DNA glycosylase (Tdg), mouse mismatch thymine DNA glycosylase (mTDG) and human methyl-CpG-binding thymine DNA glycosylase (MBD4), whereas the FoU:A lesion is repaired only by Mug and mTDG. The repair of FoU relative to the other pyrimidines examined here in human cell extract differs from the substrate preferences of the known glycosylases, suggesting that additional, and as yet unidentified glycosylases exist in human cells to repair the FoU lesion. Indeed, as observed in HeLa nuclear extract, the repair of mispaired FoU derived from misincorporation of dGMP across from template FoU could promote rather than minimize mutagenesis. The pathways by which this important lesion is repaired in human cells are as yet unexplained, and are likely to be complex.

The functional and fluorescence properties of Escherichia coli RNA polymerase reacted with fluorescamine

1. Fluorescamine (4-phenylspiro[furan-2,(3)1'-phthalan]-3,3'-dione) reacts rapidly with Escherichia coli RNA polymerase and produces a fluorescent derivative which is inactivated to an extent dependent upon reagent concentration. Excess fluorescamine is rapidly hydrolysed. Reaction is with xi-amino gruops of lysine residues in all subunits as revealed by gel electrophoresis and fluorescence scanning. 2. The extent of inactivation and fluorescence yield are diminished in the presence of added template, a finding which provides evidence for the existence of reactive and essential amino groups which can be at least partially shielded by DNA in the binary complexes. The relative decrease of fluorescence is greatest in the betabeta' subunits. Holoenzyme and core enzyme show essentially the same behavior. 3. The inactivation of activity by fluorescamine is primarily at the level of initiation. Template binding and chain propagation are less affected. 4. The enzyme derivatized by fluorescamine shows an intense fluorescence with a peak at 490 nm and an excitation maximum at 390 nm. The fluorescence lifetime is in the range of 3-8 ns and the emission is highly polarized. In reactions carried out at high ionic strength the fluorescence yield is approximately double that at low ionic strength and insensitive to the presence of template. 5. Energy transfer is observed between the derivatized enzyme as donor and ethidium bromide as acceptor in the presence of template to which both the enzyme and intercalating dye are bound. The transfer efficiency is a function of the relative concentrations and of the conditions of reaction with fluorescamine. An average transfer distance of approx. 4-5 nm has been calculated suggesting a close proximity between bound polymerase and helical regions of the template.

Affinity labeling of Escherichia coli DNA-dependent RNA polymerase with 5-formyl-l-(alpha-D-ribofuranosyl)uracil 5'-triphosphate

5-Formyl-1-(alpha-D-ribofuranosyl)uracil 5'=triphosphate has been used to affinity label E. coli DNA-dependent RNA polymerase. It is a noncompetitive inhibitor of the enzyme with Ki=0.54 mM. A short preincubation of the enzyme and alpha-fo5UTP is required to achieve maximum inhibition, and the entent of the inhibition is dependent upon the alpha-fo5UTP concentration. When a preincubation mixture of alpha-fo5UTP/enzyme is diluted, the enzyme regains activity with time showing that the inhibition is reversible, presumably occurring by Schiff base formation between an amino group on the enzyme and the formyl group. Upon sodium borohydride reduction of an enzyme/alpha-fo5UTP preincubation mixture the enzyme is irreversibly inhibited. alpha-fo5UTP is more effective in inhibiting the enzyme than alpha-fo5U, and the inhibition is decreased by the presence of ATP, UTP, or GTP in the preincubation mixture, suggesting that inhibition is occurring at a triphosphate binding site. The stoichiometry of binding of alpha-fo5UTP to the enzyme was determined using the gamma-32P-labeled derivative. After a 20-s preincubation of enzyme/alpha-fo5UTP followed by NaBH4 reduction the stoichiometry of binding was 1.1:1 (alpha-fo5UTP bound: inactivated enzyme), and this rose to 2.42:1 after a 10-min preincubation. After a 20-s preincubation the [gamma-32P]-alpha-fo5UTP was shown to be located on the beta subunit of RNA polymerase by cellulose acetate electrophoresis in 6 M urea.