Effect of single-strand break on branch migration and folding dynamics of Holliday junctions

Abstract

The Holliday junction (HJ), or four-way junction, is a central intermediate state of DNA for homologous genetic recombination and other genetic processes such as replication and repair. Branch migration is the process by which the exchange of homologous DNA regions occurs, and it can be spontaneous or driven by proteins. Unfolding of the HJ is required for branch migration. Our previous single-molecule fluorescence studies led to a model according to which branch migration is a stepwise process consisting of consecutive migration and folding steps. Folding of the HJ in one of the folded conformations terminates the branch migration phase. At the same time, in the unfolded state HJ rapidly migrates over entire homology region of the HJ in one hop. This process can be affected by irregularities in the DNA double helical structure, so mismatches almost terminate a spontaneous branch migration. Single-stranded breaks or nicks are the most ubiquitous defects in the DNA helix; however, to date, their effect on the HJ branch migration has not been studied. In addition, although nicked HJs are specific substrates for a number of enzymes involved in DNA recombination and repair, the role of this substrate specificity remains unclear. Our main goal in this work was to study the effect of nicks on the efficiency of HJ branch migration and the dynamics of the HJ. To accomplish this goal, we applied two single-molecule methods: atomic force microscopy and fluorescence resonance energy transfer. The atomic force microscopy data show that the nick does not prevent branch migration, but it does decrease the probability that the HJ will pass the DNA lesion. The single-molecule fluorescence resonance energy transfer approaches were instrumental in detailing the effects of nicks. These studies reveal a dramatic change of the HJ dynamics. The nick changes the structure and conformational dynamics of the junctions, leading to conformations with geometries that are different from those for the intact HJ. On the basis of these data, we propose a model of branch migration in which the propensity of the junction to unfold decreases the lifetimes of folded states, thereby increasing the frequency of junction fluctuations between the folded states.

Figure 1

Schematics for assembly of the HJ construct for AFM experiments. (a) Components for the construct. Vertical double-lined bars correspond to the 20 bp nonhomologous sections of synthetic duplexes. (b) Hemijunctions L and R were obtained by ligation of synthetic central duplexes with a HindIII-digested 318 bp DNA fragment. Phosphorylation of one oligonucleotide was omitted to obtain HJ with SSB in one arm at the position 47 bp from the end of the short arm (27 bp from the initial position of the branching point), indicated by an asterisk. (c) The two hemijunctions were annealed in TNM buffer to form the HJ.

Figure 2

Schematics of the HJ designs used for the smFRET experiments. Mobile HJs were nicked at the branch point and labeled with fluorescent dyes on the opposite arms. The homology region is shown in boldface. In addition to an intact version, the immobile HJs had four nicked versions with SSBs at different distances from the branch point (respective sites shown with arrows) and were labeled on the perpendicular arms.

Figure 3

AFM images of HJs (a) with and (b) without SSBs.

Figure 4

Time trajectories of fluorescence intensities in donor (green line) and acceptor (red line) channels (a) and their recalculations as FRET efficiencies (b) and dye-to-dye distances in basepairs (c) obtained from intact and nicked mobile HJs (left and right panel, respectively). ATATA design, 10 mM MgCl₂. The dashed line at 14 bp in panel c (right) corresponds to the HJ configuration when the SSB is in its vertex.

Figure 5

Statistical parameters of branch migration in mobile HJs. ATATA design, 10 mM MgCl₂. Data were averaged over 25 molecules in the case of intact HJs, and 29 in the case of nicked ones. (a) Normalized histograms of FRET efficiency plateaus. Vertical dashed lines show the range of FRET efficiencies corresponding to the donor-to-acceptor distance of 14 bp (nick in the HJ's center). (b) Residence times at different migration steps. Error bars show the standard error. (c) Normalized distributions of step sizes are in numbers of basepairs.

Figure 6

Examples of time trajectories of FRET efficiency obtained on the intact (a) and nicked immobile HJs with the nick located at the junction's vertex (b) and shifted 1 bp (c), 2 bp (d), and 3 bp (e) from it (the position of the nick is shown schematically on the right of each panel). The imaging buffer contained 50 mM MgCl₂.

Figure 7

Normalized histograms of FRET efficiencies obtained on the intact (a) and nicked (b–e) immobile HJs. Time trajectories were smoothed using the same algorithm that was used for plateau assignment in the branch migration experiments. The imaging buffer contained 50 mM MgCl₂. The dashed line corresponds to the maximum FRET efficiency histogram obtained on the intact immobile HJ in the absence of Mg²⁺ (see Fig. S5). Numbers above the peaks show maximal values obtained from the Gaussian fits.