Table 3.

Error specificity during G and 8-oxoG bypass

RPA/RFC/PCNA	Yeast						Human	Mouse
	δexo+		δexo−		η		ηa	η
	−	+	−	+	−	+	−	–
3′-CGATT
Sequencedb	23	12	24	23	52c	72	32	28
G → C	1.4	<0.7	8.5	4.8	17	12	34	<23
G → T	1.4	0.7	2.1	2.9	<5.6	4.0	52	68
3′-CGoATT
Sequenced	23	22	24	24	49	70	32	31
G → C	1300	<220	1000	<160	41	76	300	<20
G → T	3100	4900	2900	3800	260	150	4400	4500

^aAdditional experiments using a 45-mer template (see ‘Materials and Methods’ section) with slight sequence difference gave error rates of: (3′- … CG_oATC …) G → C = 610 × 10⁻⁴, G → T = 4600 × 10⁻⁴; (3′- … CAG_oTC …) G → C = <230 × 10⁻⁴, G → T = 4500 × 10⁻⁴; (3′- … CAG_oT T …) G → C = 350 × 10⁻⁴ = 4000 × 10⁻⁴.

^bDNA from dark blue plaques described in Table 2 was sequenced and error rates (×10⁻⁴) were calculated as previously described (42). Only changes at the G/8-oxoG site are given (the TAG stop codon is in italics in the left-most column). Note that G → A changes are not detectable by color screening as the amber stop codon (TAG) changes instead to an ochre stop codon (TAA) in the lacZα sequence.

^cThese sequenced samples are the same as those used in a published report (50). The specific error rates shown here have not previously been reported.