Table 5.

RAD51 is involved in gap-lesion repair by both TLS and HDR in human cells

Treatment	DNA substrate mix	E. coli transformantsa		Plasmid repair (%)	TLSb (%)	HDRb (%)
		KanR	CmR
siRAD51	GP-abasic + hDNA	129	751	13.1 ± 5.1	6 ± 2.3	6.2 ± 2.4
siControl	GP-abasic + hDNA	187	460	29.6 ± 5.8	14.5 ± 2.8	14.5 ± 2.8

	Number of isolates (%)
Treatment	siRAD51	siControl
Plasmid type	GP-abasic + hDNA	GP-abasic + hDNA
Event type
Base substitutions
A	7	18
C	5	5
G	1	2
T	–	–
(−1) deletion	17	4
(−2) deletion	–	–
Complex TLS eventsc		4
Total TLS	30 (46%)	33 (49%)
HDR product	31 (47%)	33 (49%)
Other eventsd	5	2
Total number of isolates	66	68

H1299 cells were transfected with siRNA directed against RAD51 or control siRNA. Once knock down was established, the DNA mixture containing the gap plasmid bearing an abasic site (kan^R) and the control gap-plasmid (cm^R) in the presence of hDNA was introduced into the cells. Following 8 h incubation, the DNA was extracted and used to transform an E. coli indicator strain. Plasmid repair was calculated based on the ratio of kan^R/cm^R colonies. Each result represents the average of at least four experiments. Results of single gap-filling events were obtained by sequence analysis of plasmid DNA extracted from single kan^R colonies. Depicted is the sequence at the position across from the lesion.

^aThe number of transformants obtained in a typical assay with 100 μl of transformation mixture.

^bThe repair by HDR or TLS was calculated by multiplying total plasmid repair levels by the fraction of HDR or TLS events out of the total sequences analyzed.

^cComplex TLS events involved mutation in the adjacent nucleotides. These were ACC, CAG, CCT and CCCC instead of CXC, when X represents the position across from the damage.

^dOther events depict several base pair deletions or insertion of non-homologous sequence.