In the gel shown in Figure
In the gel shown in Figure 8, at early time-points, 14% of the input dsDNA is converted to these intermediate species. These persist in the gel to later times, with approximately 50% of these being unwound. This unwinding occurs at a considerably slower rate and is due to that fraction of the population of azidothymidine that completely unwound their respective DNA molecules, or possibly dissociated prematurely from their substrate molecules. These enzyme molecules unwind the intermediates by entering the substrate from the end opposite to which a sequestered enzyme is positioned. The encounter then displaces the previously trapped enzyme molecule, which does not appear to have been altered. To demonstrate that the enzyme was not being altered permanently, unmodified dsDNA was added to unwinding reactions using adozelesin-modified DNA that had been allowed to proceed to completion. This second sample of DNA was unwound at a rate comparable to that of control reactions (data not shown). This indicates that the modification of DNA by adozelesin sequesters but does not permanently affect the RecBCD enzyme. In the second model to explain the biphasic time-courses, modification by adozelesin results in a modification of the RecBCD enzyme similar to that induced by the recombination hotspot χ. During translocation and unwinding of χ-containing DNA at 37 °C, RecBCD unwinds at 900 bp s−1 before χ. It pauses at χ for 5(±0.5) s, becomes modified, and the resulting unwinding rate following the pause is reduced to 143 bp/s−1. It was proposed that before χ, both the RecB and D motors are active. During the pause, RecD is inactivated by an unknown mechanism and the resulting reduced rate following χ-recognition is presumed to be due to translocation by the RecB motor. Here, the sites of modification by adozelesin may induce the translocating RecBCD enzyme to pause, as it would when encountering a correctly oriented and positioned χ-sequence. During this pause, the RecD subunit becomes inactivated. When translocation and unwinding resume, the altered complex, which has only one motor operating, translocates more slowly. This would result in biphasic time-courses. The 41% reduction in unwinding rate to 61(±28) bp s−1 in the early phase of these time-courses suggests that dual motors in RecBCD may be able to unwind a limited amount of adozelesin-modified DNA, whereas the single motor complex with only RecB active has great difficulty in unwinding adozelesin-modified DNA, as indicated by the 82% reduction in unwinding rate to 18(±8) bp s−1. If pseudo χ modification of RecBCD by adozelesin is occurring, then apparent χ-specific fragments should be visible in agarose gels. Consistent with this, during unwinding of both χ°- and χ+-DNA, adozelesin induced the formation of apparent χ-specific fragments. These are indicated as apparent, as we do not attribute their formation to an enzyme reaching χ and eliciting the anticipated response, i.e. the production of the anticipated and correct-sized χ-specific fragments. Instead, they result from adozelesin-induced modification of RecBCD. The DNA substrates used in χ0 and χ+ experiments were identical in sequence, and were constructed using EcoRI and BamHI, restriction enzymes, respectively (see Material and Methods). Since distinct sized apparent χ-specific fragments were produced using each of these substrates, this implies that RecBCD pausing leading to modification occurred at the sites in each substrate, likely determined by the sites of modification by adozelesin. The fragment sizes are different, as the substrates were cleaved with different restriction enzymes. The two models may not be mutually exclusive, since it is conceivable that both could be occurring. Additional work will be required to map precisely the sites of trapping and/or pseudo-χ recognition, to ascertain whether they are the same. The ability of adozelesin to trap a translocating enzyme may have important ramifications in vivo. Here, only one end of dsDNA may be exposed, so that once an enzyme such as RecBCD binds and initiates translocation and DNA unwinding, it may become entrapped permanently. This will arise as no free end is available to permit entry of a second RecBCD molecule to facilitate displacement of the trapped enzyme.