Fig 2: Additional SLX4 segments resolved in the structure of the XPF-ERCC1-SLX4IP-SLX4330-555-DNA complex.a Left: Cryo-EM map of the XPF-ERCC1-SLX4IP-SLX4330-555-DNA complex. Three SLX4 segments (orange) are visualised and labelled. Right: Cryo-EM map of the DNA-free complex from the same data collection shown in the same orientation. Only SLX4 residues 526-552 are visualised, in line with the higher-resolution reconstructions shown in Fig. 1. Maps are shown without any b-factor sharpening applied to facilitate visualisation of weaker densities. b Schematic depiction of the sequence elements and domains contained in the SLX4 constructs used for in vitro endonuclease assays of XPF-ERCC1-SLX4IP-SLX4 complex variants. c In-vitro endonuclease assay comparing the activity of XPF-ERCC1 to the activities of different variants of the XPF-ERCC1-SLX4IP-SLX4330-555 complex. Conversion of the input substrate (see Fig. 3a) into 23-nt product was monitored by detection of Cy3 fluorescence. To facilitate visualisation, 10 min time points are marked with a red dot. The removal of the UBZ-2 domain and the addition of the BTB domain do not have an appreciable effect on endonuclease activity. Removal of the sequence elements that are solely visualised in the DNA-bound XPF-ERCC1-SLX4IP-SLX4330-555 complex strongly reduces cleavage activity. Two additional experimental repeats and all protein sample processing controls are provided in Supplementary Fig. 7a–e. Source Data are provided as a Source data file.

Fig 3: XPF-ERCC1 endonuclease activity in the presence of SLX4IP and SLX4330-555.a Y-fork DNA substrate used to assay activity of XPF-ERCC1 complexes. The preferred cleavage site of XPF-ERCC1 is denoted by an arrow and liberates a 23-nt DNA fragment with a 3’-Cy3 fluorophore used for detection of the product. b Comparison of nuclease activity of XPF-ERCC1, XPF-ERCC1-SLX4IP, XPF-ERCC1-SLX4IP-SLX4330-555, and XPF-ERCC1-SLX4330-555. Conversion of uncleaved input substrate (a) into product (23-nt fragment) was monitored by detection of Cy3 fluorescence. SLX4IP has no impact on activity, while the presence of SLX4 in the complex increases cleavage, as evidenced by the more rapid disappearance of the band corresponding to the intact substrate. To facilitate visualisation, the 5 min time points are marked with red dots. c Coomassie stained SDS-PAGE gel of protein sample processing control (before final 10x dilution into the nuclease reaction) to confirm that equivalent amounts of endonuclease were present in all samples. Two additional repeats of the experiment shown in panels b and c are provided in Supplementary Fig. 5. d Y-fork DNA substrate used to assemble an XPF-ERCC1-SLX4IP-SLX4330-555-DNA complex for cryo-EM. Bonds protected against endonucleolytic cleavage by phosphorothioate linkages are indicated by asterisks (*). Source Data are provided as a Source data file.

Fig 4: Structure of a DNA-bound XPF-ERCC1-SLX4IP-SLX4330-555 complex.a Cryo-EM map of the XPF-ERCC1-SLX4IP-SLX4330-555-DNA complex. XPF is shown in cyan, ERCC1 in blue, SLX4 in orange, SLX4IP in yellow, the scissile DNA strand in light green, and the non-scissile strand in dark green. b Atomic model of the XPF-ERCC1-SLX4IP-SLX4330-555-DNA complex (colours as in b). c Surface view of the proteins in the complex with residues in proximity to DNA (<3.5 Å) shown in pink. DNA is not shown. d Close-up view of the DNA contacts with the ERCC1 HhH2 domain. e Close-up view of the DNA contacts with the XPF nuclease and RecA1 domains. Manganese ions bound in the active site shown as spheres. f View of the XPF active site with two metal ions coordinated by negatively charged residues (shown in cyan) and a DNA backbone phosphate.

Fig 5: Characterisation of the SLX4-XPF interaction.a Structures of the SLX4 fragment 526-552, with residues interacting with XPF shown as yellow sticks (left), and structure of XPF, with residues interacting with SLX4 shown as red sticks (right). The area of XPF interacting with the SLX4 helix 529-534 is marked by a dashed line. b Small-scale in-vitro co-sedimentation assay of XPF-ERCC1 with Strep-tagged SLX4330-555 wild type and 5 A mutant. c Small-scale in-vitro co-sedimentation assay of MBP-fused SLX4523-555 wild type and 5 A mutant. d Co-IP of wild type and 5A-mutant SLX4 from HEK293TN cells. 1% of input lysate to the IP was analysed as input. Source data are provided as a Source data file.