The cohesin-like RecN protein stimulates RecA-mediated recombinational repair of DNA double-strand breaks

RecN is a cohesin-like protein involved in DNA double-strand break repair in bacteria. The RecA recombinase functions to mediate repair via homologous DNA strand invasion to form D-loops. Here we provide evidence that the RecN protein stimulates the DNA strand invasion step of RecA-mediated recombinational DNA repair. The intermolecular DNA tethering activity of RecN protein described previously cannot fully explain this novel activity since stimulation of RecA function is species-specific and requires RecN ATP hydrolysis. Further, DNA-bound RecA protein increases the rate of ATP hydrolysis catalysed by RecN during the DNA pairing reaction. DNA-dependent RecN ATPase kinetics are affected by RecA protein in a manner suggesting a specific order of protein–DNA assembly, with RecN acting after RecA binds DNA. We present a model for RecN function that includes presynaptic stimulation of the bacterial repair pathway perhaps by contributing to the RecA homology search before ternary complex formation.


Supplementary Figure 2 -RecN K67A-dependent tethering of linear duplex DNA molecules. The RecN and
RecN K67A proteins stimulate the ligase-dependent intermolecular ligation of linear duplex DNA. Ligation reactions were carried out in dilute conditions (as described in the methods below) such that the primary product of ligation in the absence of added RecN protein is self-ligated circular plasmid with various topoisomers of the intramolecular, covalently closed plasmid product. The wild-type RecN (1M, panel A) or RecN K67A mutant protein (1 M, panel A or the indicated concentration, panel B) was incubated with 2.4 M nt linear, 2.4 kb plasmid DNA substrate (marker, lds M) before the addition of D. radiodurans DNA ligase (to 4 nM). The ligated DNA products were purified from the dilute reaction mixture and analyzed in a 0.7% agarose gel. The migration of linear multimeric products relative to the BstE II digested lambda DNA marker standard (M) was used to identify the intermolecular ligation products as dimers (4.8 kb), trimers (7.2 kb), etc.
The addition of RecN or RecN K67A mutant stimulates the intermolecular ligation of linear duplex DNA (2.4 kbp) to form dimers, trimers, etc. As has been noted for the wild-type RecN protein [1], increasing concentrations of the RecN K67A protein inhibits the generation of large, higher-ordered, multimers (panel B). This is likely due to protein aggregation. However, DNA bridging is occurring at the concentration used in the Dloop formation assay of figure 2B.
DNA ligation assays were carried out as previously described [1]. Briefly, ligation reactions were carried out in a final reaction volume of 100 L in buffer N and an ATP regeneration system (10 units mL -1 pyruvate kinase and 2.5 mM phosphoenolpyruvate). Wild-type RecN or RecN K67A mutant protein (final concentrations on figure) were incubated with 2.4 M nt linearized pEAW3 DNA for 30 min at 37°C. After incubation, each reaction was provided with 12 L of 10X DNA ligase reaction buffer (300 mM Tris-Cl (pH 8), 40 mM MgCl2, 1 mM DTT, 260 M NAD + , 500 g mL -1 BSA) and 4 nM D. radiodurans Ligase A and incubated at 30°C for 30 min. Ligation reactions were stopped by addition of 80 L of termination buffer (20 mM Tris-HCl ( 80% +), 20 mM EDTA, 0.5% SDS) and deproteinized with 100 g predigested proteinase K followed by 20 min incubation at 37°C. DNA from each reaction was recovered by phenol:chloroform:iso-amyl-alcohol solution (25:24:1) and precipitated with ethanol. The precipitated DNA was resuspended in 10 L TE plus 5 L of a solution containing 60 mM EDTA, 6% SDS, 25% (weight per volume) glycerol, and 0.2% bromphenol blue and electrophoresed in 0.7% agarose gels in 0.5X TBE buffer. DNA was visualized by ethidium bromide staining and inverted images were obtained using a Fotodyne image system equipped with a digital CCD camera and Foto/Analyst PC image software v10.21. The steady-state rates were measured and graphed relative to the rate of RecN ATP hydrolysis in the absence of added RecA protein (RecN only). The error bars represent the standard deviation of the relative rate (the standard deviation of the average rate divided by the average rate) of six or seven independent experiments. The experimental procedure was the same as for figure 4, reaction 5 except for the concentration of RecA used. The RecA∆C17 was purified as described [2].

Supplementary Figure 3 -The RecN ATPase activity is stimulated by D. radiodurans (Dr) RecA but not E. coli (Ec
A low concentration (0.5 M) RecA protein was used so that the rate of ATP hydrolysis measured in the absence of RecN was minimized. When the DrRecA or DrRecA K83R mutant protein is included, the RecN ATP hydrolysis rate increases ~2-fold under these lower (than in the experiments of Fig. 4) RecA concentration conditions. EcRecA protein nucleates slowly onto duplex DNA. Therefore, we also included an experiment with the EcRecA∆C17 mutant protein that has been shown to nucleate rapidly onto duplex DNA [2]. When the EcRecA or EcRecA∆C17 mutant protein is included, the measured rate of RecN ATP hydrolysis is similar to that of RecN alone.
The experiments of panel A and figure 4 are carried out under the conditions previously optimized for RecN activity [1]. However, the DNA strand exchange and D-loop formation assays of figure 1 and 2 are carried out under different solution conditions. The major difference in conditions (see methods for figure 4) is that the RecN-optimized solution (Buffer N) contains 1% polyethylene glycol (PEG) and 17.5 mM magnesium acetate (panel A and Fig. 4) and the RecA-optimized solution (Buffer A) contains 10 mM magnesium acetate and no PEG. We therefore compared the RecN ATP hydrolysis rate in the presence and absence of 0.5 M RecA proteins under the RecA-optimized solution conditions (panel B). Under the RecA-optimized conditions we observe a very low rate (~ 2 M min -1 ) of RecN ATP hydrolysis in the absence of a RecA protein. When the DrRecA or the DrRecA K83R proteins are included, we observe a 15 to 20-fold stimulation of the rate of RecN ATP hydrolysis. This is consistent with the stimulation observed in figure 5. However, we observe an only additive effect when the EcRecA or EcRecA∆C17 mutant protein is included suggesting the E. coli proteins do not stimulate RecN ATP hydrolysis.