DNA binding and bridging by human CtIP in the healthy and diseased states

Abstract The human DNA repair factor CtIP helps to initiate the resection of double-stranded DNA breaks for repair by homologous recombination, in part through its ability to bind and bridge DNA molecules. However, CtIP is a natively disordered protein that bears no apparent similarity to other DNA-binding proteins and so the structural basis for these activities remains unclear. In this work, we have used bulk DNA binding, single molecule tracking, and DNA bridging assays to study wild-type and variant CtIP proteins to better define the DNA binding domains and the effects of mutations associated with inherited human disease. Our work identifies a monomeric DNA-binding domain in the C-terminal region of CtIP. CtIP binds non-specifically to DNA and can diffuse over thousands of nucleotides. CtIP-mediated bridging of distant DNA segments is observed in single-molecule magnetic tweezers experiments. However, we show that binding alone is insufficient for DNA bridging, which also requires tetramerization via the N-terminal domain. Variant CtIP proteins associated with Seckel and Jawad syndromes display impaired DNA binding and bridging activities. The significance of these findings in the context of facilitating DNA break repair is discussed.

The linear dsDNA molecule was prepared as follows.The central part was obtained by digestion of a large homemade plasmid with BssHII (New England Biolabs) produced following published protocols (1).Without further purification, the fragment was ligated to highly biotinylated handles of ~1 kbp ending in MluI.Handles for C-trap constructs were prepared by PCR (see Supplementary Table 1 for primers) including 200 µM final concentration of each dNTP (dGTP, dCTP, dATP), 140 µM dTTP and 66 µM Bio-16-dUTP (Roche) using the plasmid pSP73-JY0 as template (2) followed by digestion with the restriction enzyme MluI (New England Biolabs).Labelled handles were ligated with the central part with T4 DNA Ligase (New England Biolabs) for 15 h at 16°C followed by 1 h at 37°C before heat inactivation, in the presence of BssHII enzyme to avoid tandem (doublelength) tethers.These handles were highly biotinylated to facilitate the capture of DNA molecules in C-Trap experiments.The sample was ready for use without further purification.
DNAs were never exposed to intercalating dyes or UV radiation during their production and were stored at 4˚C with 1 mM EDTA pH 8.0 to final concentration to preserve them (see Supplementary Table 2 for sequence).
The loop-containing DNA was prepared as follows.Two equal long dsDNA fragments of 7728 bp labelled with biotins in one end and containing two single-stranded regions in the opposite end, were annealed into the gaps created in a large homemade plasmid.This large plasmid is the same than the one employed above to fabricate the linear dsDNA substrate, and contains two regions of five spaced BbvCI restriction sites (see Supplementary Table 2 for sequence).These poly-BbvCI regions were extracted from the pNLrep plasmid (kindly gifted by Prof. Dr. Ralf Seidel) and they are the same than the ones present in the plasmid employed to fabricate the magnetic tweezers DNA substrate, but separated by 6459 bp.The large plasmid was digested with the restriction enzyme Nt.BbvCI (New Englands Biolabs).This produced two sets of five nicks on one of the strands leading to two 63-nt gaps during heat inactivation of the nicking enzyme (3).This plasmid with two gaps was employed in a later step without further purification.The 7728 bp-long dsDNA fragments were produced by PCR amplification with Phusion High-Fidelity DNA Polymerase (Thermo Scientific) using pSP73-JY0 as DNA template (2) (see Supplementary Table 1 for primers) followed by purification (QIAGEN).The PCR fragment of 7688 bp contains a BbvCI restriction site in each end.Therefore, it was digested with Nt.BbvCI restriction enzyme creating a 5´-overhang of 13 nt in one end and a 3´-overhang of 12 nt in the opposite end.This simpler strategy to generate both 3´ and 5´-overhangs by using nicking enzymes has been previously described (3,4).Without further purification, the 5´-overhang was filled in by DNA polymerase (Klenow Fragment [3'-5' exo-], New England Biolabs) employing dGTP, dATP, biotin-16-dUTP and biotin-14-dCTP (Thermo Fisher Scientific) for 1 h at 37˚C.Reaction was followed by heat inactivation of the enzyme for 20 min at 75˚C.After that, the 3´-overhang of 12 nt was annealed with a small branched connector.This small branched connector was produced by annealing two partially complementary oligonucleotides (Supplementary Table 1) by heating at 95°C for 5 minutes and cooling down to 20°C at a 1°C minute -1 rate in hybridization buffer (10 mM Tris-HCl pH 8.0, 1 mM EDTA, 200 mM NaCl, 5 mM MgCl2).This small branched connector contains a 40-bp duplex stem ending in a 3´-overhang of 12 nt complementary to the 3´-overhang of the long dsDNA fragment, and two single-stranded regions complementary to the gaps created by Nt.BbvCI in the large plasmid.A 25X excess of the small branched connector was annealed into the 3´-overhang of the long dsDNA fragment by heating 10 min at 72°C, and slowly cooling down to 42°C at a 0.1°C every 25 sec rate in annealing buffer (10 mM Tris-HCl pH 7.5, 1 mM MgCl2) followed by overnight with T4 DNA Ligase (New England Biolabs).The complete long dsDNA fragments already labelled with biotins in one end and with two single-stranded regions in the opposite end were gel extracted and purified (QIAGEN).Finally, a 5.4X excess of these long dsDNA fragments was hybridized into the gaps previously created in the large plasmid by heating 5 min at 80°C, and slowly cooling down to 30°C at a 0.5°C min -1 rate (but increasing 10 sec every minute) in annealing buffer (50 mM Tris-HCl pH 8.0, 1 mM EDTA, 100 mM NaCl) as described in (5).A final step of overnight ligation with T4 DNA Ligase was performed to seal all the nicks.The sample was ready for use without further purification.As usual, DNAs were never exposed to intercalating dyes or UV radiation during their production and were stored at 4˚C with EDTA pH 8.0 to 1 mM final concentration to preserve them.
For the binding of labelled short DNAs in trans, three types of molecules were prepared: two fork DNA molecules labelled with ATTO488 or Cy5, and a dsDNA small fragment of 139 bp labelled with ATTO488.
ATTO488-Fork and Cy5-Fork DNA molecules were prepared by annealing two partially complementary oligonucleotides (see Supplementary Table 1) by heating at 95°C for 5 minutes and cooling down to 20°C at a 1°C minute -1 rate in hybridization buffer (10 mM Tris-HCl pH 8.0, 1 mM EDTA, 200 mM NaCl, 5 mM MgCl2).ATTO488-Fork molecule features two 37 nt-long polydTs with an ATTO488 label in the 5´-end of one of the branches, linked to a 26 bp-long dsDNA stem.Cy5-Fork molecule features two single-stranded regions of 31 and 32 nucleotides linked to a 60 bp-long dsDNA stem labelled with Cy5 in the 5´-end.Without further purification, samples were stored at -20°C.