Fig. 3. Determinants of DSB substrate specificity by Polλ.

a Structural superposition of 1nt SSB crystal structures of Polλ (PDB ID code 1XSN¹¹, protein in gray and DNA in transparent khaki surface) and Polμ (PDB ID code 4M04¹², protein in light blue). Loop1 and the ‘thumb loop’ regions of Polλ are shown in cyan and magenta, respectively. b Structure-based sequence alignment of Loop1 (cyan) in Pols β, λ, and μ (structurally-equivalent regions highlighted in yellow). The Polλ ΔL1 variant was generated by deleting Ser463-Gln471 (underlined) and replacing them with Lys462-Thr465 from Polβ¹³. c Sequence alignment of ‘thumb loop’ (TL, magenta) of Pols β, λ and μ. The Polλ ΔTL variant was generated by deleting Val537-Val545 of Polλ (underlined) and fusing the ends of β-strands 7 and 8 with a single glycine residue. Thumb loop mutations used in this study are marked in green. d Interactions of the TL with an SSB substrate (PDB ID code 1XSN¹¹, gray). Hydrogen bonds with the DNA template strand are shown as black dashed lines. e DNA fragments with partly complementary ends (GCG3ʹ overhangs) were introduced into Polm^−/−Poll^−/− cells (Polml^−/−). f Representative gel electrophoresis (top) identification of the fraction of NHEJ that is AfeI-sensitive and polymerase-dependent. Quantification of the average fraction of polymerase-dependent repair using GCG3ʹ DSB substrates from triplicate electroporations (bottom; error bars represent the standard deviation). g As in (e), except using noncomplementary blunt and GC3ʹ overhangs. h As in (f), except using PvuII sensitivity to identify polymerase-dependent repair.