Even these “incredibly simple systems,” says van Oijen, “have transient intermediates with no product, for example these pauses, so you need single-molecule studies in real time.” But even simple DNA polymerase systems are multimeric complexes, whose in vitro assembly yields a few percent of productive complexes.
Rather than do hundreds of individual laser trap experiments, the Boston team stuck one end of the DNA to glass and put a bead on the other end. Fluid flow stretched out hundreds of the DNA molecules in a single field of view.
With only a leading strand polymerase bound and active, the stretched DNA shortened as the lagging strand was converted into highly coiled single-stranded DNA. But when ribonucleotides were added, there were transient pauses in this shortening as primase periodically kicked into action. When a lagging strand polymerase was allowed to join in, the researchers also saw lagging strand loop formation, but still saw the primase-related pauses.
The pauses took several seconds—long enough for an entire primer to be made. Thus, primer synthesis may not start until the lagging strand finishes making its Okazaki fragment. The contortions of active primase may freeze the linked helicase and thus halt the leading strand.
To rationalize how the leading and lagging strands stay together, “people have stressed minimizing the dead time on the lagging strand,” says van Oijen. “But they have not really been explicitly addressing the possibility of the leading strand actively being modulated by the lagging strand.” Having established this fact, van Oijen wants to add a second bead to the remaining free DNA end, so that lagging and leading strand speeds can be compared. Just like the current work, he says, “it's a cute little experiment.”