Decatenation of DNA braids by E. coli topoisomerases I and III. (A) Sketch of two DNA molecules attached to a paramagnetic bead: two intact dsDNAs attached by only one strand (left), two nicked dsDNAs (middle) and two nicked dsDNA with a 27-bp bulge (right) on each DNA molecule. The three substrates were braided by rotating the magnet and were used in E. coli topoisomerases I and III decatenation experiments. These DNA braids mimic catenated DNA structures that arise during chromosomal replication and catenated dimers of nicked circular DNA. (B) Diagram illustrating the braiding procedure and the crossover angle (β) between two braided strands. The crossover angle is obtained by braiding two parallel DNA molecules (top) by one turn (ΔCa) (bottom left). A more catenated DNA substrate is formed by braiding two DNA molecules by several turns (ΔCa = n) (bottom right). (C) Example of an extension versus catenation number for a nicked DNA braid with 27-bp bulges at different forces (: 0.3 ± 0.1 pN; : 0.5 ± 0.15 pN; : 1.5 ± 0.2 pN; : 2.0 ± 0.3 pN). (D–I) Plots showing typical decatenation runs for different substrates. Each run is defined as a series of decatenation events in which pausing cannot be observed. (D) Decatenation of an intact dsDNA braid by topoisomerase I. No activity was observed for this type of substrate. (E) Decatenation of a nicked dsDNA braid by topoisomerase I. Poor activity was observed with this substrate, confirming that the presence of the nicks is not sufficient to recapitulate robust decatenation activity. (F) Decatenation of a nicked dsDNA braid with a 27-bp bulge by topoisomerase I. (G) Decatenation of an intact dsDNA braid by topoisomerase III. (H) Decatenation of a nicked dsDNA braid by topoisomerase III. (I) Decatenation of a nicked dsDNA braid with a 27-bp bulge by topoisomerase III. In all cases, the length of the DNA is plotted against time. Manual introduction or removal of the catenanes resulted in shortening or elongation, respectively, of the DNA braid, (green arrows), whereas decatenation by an enzyme resulted in elongation. The gray trace corresponds to the measured events, whereas the red trace corresponds to an unweighted running average over 50 events.

Topoisomerases I and III have similar number of decatenation events and decatenation rates per run. (A) The number of catenanes per run resolved by topoisomerases I and III on nicked DNA braids with 27-bp bulges. Histogram of the mean number of catenanes resolved in a run (ΔCa). The two enzymes remove similar numbers of catenanes in a run. The inset shows the distribution for topoisomerase I (top) and topoisomerase III (bottom). The distributions do not fit to a simple exponential decay and hence the average number of turns removed was calculated. (B) Decatenation rate per run by topoisomerases I and III on nicked DNA braids with 27-bp bulges. The histogram shows the distribution of the decatenation rate per run. Topoisomerase III has a similar decatenation rate per run as topoisomerase I. The inset shows the distribution for topoisomerase I (top) and topoisomerase III (bottom). The solid curve corresponds to a fit of an exponential decay to the distribution. In both panels, shaded bars correspond to topoisomerase I and white ones to topoisomerase III and error bars correspond to the standard error of the mean. Details on the number of events used for each histogram are found in Supplementary Table S1.

Time lags before a decatenation run for topoisomerases I and III. (A) Initial and secondary time lags for topoisomerases I and III on nicked DNA braids with 27-bp bulges. Histogram of initial time lag before decatenation indicates that topoisomerases I and III are comparable. In contrast, histogram of the secondary time lag indicates a larger difference between the two enzymes (P-value < 0.0001). The inset shows the distribution of secondary time lags for topoisomerase I (top) and topoisomerase III (bottom). (B) Effect of crossover geometry of the DNA braid on the secondary time lag by topoisomerases I and III. Histogram showing the distribution of secondary time lag for two sets of DNA crossover angles for topoisomerase III. Topoisomerase III exhibits shorter secondary time lags in the case of large crossover angles (P-value < 0.0001), whereas topoisomerase I has similar secondary time lags for all crossover angles (not shown). Shaded bars correspond to small (β = 24°) and white to large (β = 47°) angles (see the text for a definition of the crossover angle). The inset shows the distribution of secondary time lags for topoisomerase III activity for small (top) and large (bottom) crossover angles. In all histograms, the error bars correspond to the standard error of the mean. The solid curve corresponds to a fit of an exponential decay to the distribution. Details on the number of events used for each histogram are found in Supplementary Table S1.

Topoisomerase III is a faster decatenating enzyme. (A) Total decatenation rate by topoisomerases I and III on a nicked DNA braid with 27-bp bulges. Histogram shows the distribution of the total decatenation rate. Topoisomerase III has a significant (P-value < 0.0001) faster total decatenation rate than topoisomerase I, in agreement with bulk experiments. The inset shows the distribution of total decatenation rates for topoisomerases I (top) and III (bottom). (B) Effect of braid geometry on the total decatenation rate by topoisomerases I and III. Histogram showing the distribution of the total decatenation rate for two DNA crossover angles. It indicates that topoisomerase III has a larger total decatenation rate for large crossover angle, whereas the decatenation activity of topoisomerase I is not strongly dependent on the crossover angle. Shaded bars correspond to small (β = 24°) and white to large (β = 47°) angles (see the text for a definition of the crossover angle). The histogram of the total decatenation rate indicates a significant difference between the two crossover angles in the case of topoisomerase III (P-value < 0.0001). The inset shows the distribution of total decatenation rate for topoisomerase III for small (β = 24°) (top) and large (β = 47°) (bottom) crossover angles. In all histograms, the error bars correspond to the standard error of the mean. The solid curve corresponds to a fit of an exponential decay to the distribution. Details on the number of events used for each histogram are found in Supplementary Table S1.