Figure 4: Time-lapse TIRF microscopy images.

These images are timepoints of TIRF movies (see Supplementary Movies 1–3). The bright circular spots correspond to fluorescent beads used for position and focus reference; the number of bright spots increase through the course of the experiments because beads continue to deposit on the surface of the slide. For the top two rows, the initial reagents present were [protected tile]=200 nM, [F]=440 nM, [S]=220 nM, and [sink]=40 nM; 20 nM catalyst was introduced at time t≈0. For the bottom row, only [protected tile]=200 nM was initially present; 220 nM deprotector was introduced at time t≈0. No DNA nanotubes form within the time frame of the experiment when no catalyst was present. When a small amount (20 nM) of catalyst was added, DNA nanotubes formed slowly, initially limited by nucleation due to low active tile concentration. From this, it can be inferred that the upstream catalyst system possesses multiple turnover, because the catalyst concentration is lower than the sink concentration. This multiple turnover is consistent with fluorimeter assay of the upstream catalyst (see Supplementary Fig. S3). When a super-stoichiometric quantity of trigger was added, DNA nanotubes formed immediately.