Single molecule tracking reveals spatio-temporal dynamics of bacterial DNA repair centres

Abstract

Single-particle (molecule) tracking (SPT/SMT) is a powerful method to study dynamic processes in living bacterial cells at high spatial and temporal resolution. We have performed single-molecule imaging of early DNA double-strand break (DSB) repair events during homologous recombination in the model bacterium Bacillus subtilis. Our findings reveal that DNA repair centres arise at all sites on the chromosome and that RecN, RecO and RecJ perform fast, enzyme-like functions during detection and procession of DNA double strand breaks, respectively. Interestingly, RecN changes its diffusion behavior upon induction of DNA damage, from a largely diffusive to a DNA-scanning mode, which increases efficiency of finding all sites of DNA breaks within a frame of few seconds. RecJ continues being bound to replication forks, but also assembles at many sites on the nucleoid upon DNA damage induction. RecO shows a similar change in its mobility as RecN, and also remains bound to sites of damage for few hundred milliseconds. Like RecN, it enters the nucleoid in damaged cells. Our data show that presynaptic preparation of DSBs including loading of RecA onto ssDNA is highly rapid and dynamic, and occurs throughout the chromosome, and not only at replication forks or only at distinct sites where many breaks are processes in analogy to eukaryotic DNA repair centres.

Figure 1

Model for early steps during the repair of double strand breaks (DSBs) via homologous recombination (“presynapsis”). RecN visibly assembles at sites of DSBs (indicated by the flash) at a very early time point, and binds to 3′ ends. RecJ endonuclease (or alternatively the AddAB complex) resects one strand, helped by helicase RecQ or RecS. ssDNA is bound single strand-binding protein SsbA (SSB in E. coli), which needs to be displaced by RecO and RecR to load RecA onto ssDNA. The RecA nucleoprotein filament is now ready to search for a homologous DNA duplex on the sister chromosome for the formation of a crossover.

Figure 2

Single-molecule tracking (SMT) of RecN-YFP. (a) Example of a single RecN-YFP molecule acquired with an integration time of 15 ms. (b,f) Normalized signal intensity at the site of localization. The signal was measured starting 5 frames before the molecule appears and ends 5 frames after the molecule disappears in a single step. The grey shaded area highlights the time of appearance of the molecule. The intensity drawn in (b) corresponds to the trajectory shown in the montage in (a). (c,g) Projection of all recorded tracks into the coordinate system of the cell. The trajectories shown in (b,f) are highlighted and color-coded according to its length. (d,h) Displacement of the trajectories over time. Shown is the Euclidian distance of the molecules to the site of appearance (black line) and its sequential displacements between consecutive frames (green line). The dotted red line represents an upper threshold setting the limit for immobile displacements. (e) and (i) Color-coded displacements of the trajectories over time. The radius of the circle corresponds to the limit for a track to be considered as immobile.

Figure 3

Comparative analysis of diffusive behavior. (a) Simultaneous comparison of frame-to-frame displacements for RecN-YFP before and after treatment of the cells with Mitomycin C (MMC). The dotted and dashed lines indicate the distributions of the immobile and mobile fraction. The red line represents the fit deriving from the sum of the two normal distributions and the green line depicts the fit assuming a single normal distribution. (b) Bubble plot showing the diffusion coefficients (y-axis) and fraction size (size of the bubble, also stated in percentage above the bubbles) of the fluorescent fusion proteins. Blue bubbles represent untreated samples and red bubbles show samples after treatment with 50 ng/ml MMC. R-square test shows highly significant changes (p < 0.01) between all pairs of +/−MMC.

Figure 4

Residence time and spatial distributions of the molecules. (a) Cumulative distribution of residence times of RecN-YFP before and after treatment with MMC. Histograms show events of resting fitted either by a one- or two-component exponential function (red and green line). (b) Residence lifetime of fluorescent fusions of RecN, RecJ, RecO, PfkA and TetR. Times of resting before treatment with MMC are shown in blue and resting times of molecules after treatment with MMC are shown in red. Results of the 1- (single bubble) and 2 - (two bubbles) component fits are shown side-to-side, the size of fractions for the two populations are stated as percentage above the bubbles. All changes between exponentially growing and MMC-treated cells were found to be highly significant, R-square test (p < 0.01). (c) Spatial distribution of RecN-YFP molecules in a standardized cell before and after induction of DNA damage. Mobile molecules are plotted in blue and molecules that remained inside a circle with radius r = 120 nm over their complete lifetime are plotted in red (see track as example in Fig. 1e). (d) Heat map of all RecN-YFP molecules before and after induction of DNA damage.

Figure 5

Spatial distribution of single RecN-YFP molecules classified as immobile and mobile. 1D Histograms of tracks binned to the long (a) and short axis of the cell (b). Left and right panel show distributions in absence and presence of DNA damage, respectively.

Figure 6

Spatial distribution of classified single molecules trajectories (based on SMT). (a) Localization of mobile and immobile trajectories within a model cell with standardized geometry. (b) 1D histogram of the classified trajectories binned to the long axis of the cell. (c) 1D histogram of the classified trajectories binned to the short axis of the cell. We analyzed >195 trajectories for each condition (see Supplementary Table S3).

Figure 7

Time-lapse epifluorescence experiments, (A) with exponentially growing RecJ-CFP cells 45 min after induction of DSBs with 100 ng/ml MMC. Pictures were taken every minute. White box indicates area enlarged on the right. White arrowheads indicate localization of static RecJ-CFP foci, grey arrowheads dynamically appearing and disappearing foci. (B–C) Stream acquisition (100 ms intervals) showing the assembly and disappearance of RecN-YFP centers. Note that foci arise from several molecules, because experiments are done by epifluorescence, not by SMT! Outlines of cells are indicated by dashed lines. (B) Disappearance of single RecN-YFP assembly, (C) appearance of a RecN-YFP focus at a different site in the cell following the disappearance of an earlier focus. Scale bars 2 µm.

Figure 8

Mobility of RecN assemblies in the presence of DNA damage (A). Representative two-dimensional trajectories of RecN-YFP acquired at 1 s time intervals. The color-code corresponds to the time-scale indicated by the color bar. (B) Mean-squared displacement analysis of time-lapse microscopy data acquired at different time-intervals for RecN-YFP and the oriC-locus following treatment with 50 ng/ml MMC. Shown are time-ensemble-averaged MSDs for the first ten time-lags of the trajectories acquired at each time interval. Only trajectories with a length ≥10 frames were included. To cover the full range of the MSDs curves the data were plotted in a log-log scale.

Figure 9

Results of variational bayes single particle tracking (vbSPT) analysis. (a) Left: PfkA-YFP; Right: TetR-YFP. (b) RecN-YFP in the absence (left) and presence of DNA damage (right). (c) Strains deleted for addAB and recJ expressing RecN-YFP in the absence (left) and presence of DNA damage (right). (d) RecJ-YFP in the absence (left) and presence of DNA damage (right). (e). RecO-YFP in the absence (left) and presence of DNA damage (right). The vbSPT analysis provides information about the number of diffusive states, the diffusion coefficients, the mean dwell time of each state and the transition probabilities between the states. The circles represent the individual states and the area of the circle indicates its relative occupancy. Arrows and associated numbers in grey show the probabilities that a molecule transitions from one state to the other in one time step. Dotted arrows indicate transitions with very low probabilities.