Fig 1.
Schematic of Tpm1.1 binding to both sides of actin filament.
(A) Growth of Tpm1.1 strands on both sides on actin filament in opposite directions. The meeting point where two intensities of fluorescently labelled Tpm1.1 start to overlap is indicated with the blue line. This type of Tpm1.1 association with actin leads to fully saturated filament. (B) Initial growth of Tpm1.1 strands on the same side of actin can lead to a gap formation or the fully saturated region. A single Tpm1.1 isomer stretches over seven actin monomers; therefore gaps may be formed by short stretches of undecorated actin filament with a length of 1 to 6 actin monomers.
Fig 2.
Association of tropomyosin Tpm1.1 to actin filaments.
(A) Schematic of Tpm1.1 molecules binding to actin filament. (B) Actin filaments decorated with AF647-Tpm1.1 under the flow in the microfluidic channel. Green and magenta arrowheads indicate barbed end and pointed end of the actin filament, respectively. (C) Snapshots of a single actin filament from a two colour TIRF time lapse image stack show binding of Tpm1.1 to phalloidin-AF488 labelled pre-formed actin filament. Data shown in Fig 2B and C were for Tpm1.1 at 62.5nM at 92nm/pixel resolution and 1Hz acquisition speed. Scale bar, 5 μm.
Fig 3.
Intensity profile and kymographs of tropomyosin Tpm1.1 at different concentrations.
(A) The total intensity of the region of interest (ROI) as a function of time during the association phase. The ROI is selected by the user in the image, and intensity thresholding is then used to define the boundaries of individual filaments. Note the rapid rise in intensity at approximately 30 seconds. The increase is due to the Tpm1.1 which was bound to the actin filament. The intensity plateaus as the entire actin filament is coated with Tpm1.1. (B) The total intensity of the ROI as a function of time during the dissociation of Tpm1.1 from the same actin filament. The intensity profiles in panels A. and B. correspond to the kymograph shown in the panel C. The tropomyosin was loaded into the microfluidics channel at (C) 62.5nM, (D) 250nM, and (E) 500nM during the association phase. After full saturation of the actin filament the channel was washed with buffer only (dissociation phase). The vertical axes of each kymograph show the length of the filament where an individual pixel represents 92nm. The frames (pixels) shown in the horizontal axis were typically recorded at 1Hz for the 62.5nM association and 10Hz at the other concentrations. The magenta dashed line shows the boundary between the association and dissociation phases. The scale bars show 20 pixels (1.84 μm) in length. The yellow circle at approximately 64 seconds, at pixel 10 of the filament in panel C indicates the initial nucleation of binding.
Fig 4.
Two distinct Tpm1.1 intensity levels on an actin filament.
(A) Kymograph, with pixel 30 indicated by the dotted line, and (B) intensity profile at pixel 30 of the filament shown in Fig 3D. The red line shows the fit to the intensity profile of pixel 30 of a function with identical magnitude intensity transitions. The transition magnitudes were found to be consistent across all the pixels in the kymograph. The fit optimised both the magnitude and the timing, with the results showing the transitions aligned with the single and double bound levels.
Fig 5.
(A) Boundaries as determined by intensity thresholding. The outer boundary occurs at the earlier time and lower intensity threshold (magenta) and the inner boundary is the later time, higher intensity threshold (green). (B) Boundaries after user editing. The boundaries have been cut and joined to show the likely cross-over–each boundary then indicating binding on different sides of the actin filament. (C) Nucleation points (light blue circles) and meeting points on the same groove (yellow circles) can be automatically detected at the extrema in each time-pixel boundary. Other points of interest can also be marked by the user to allow for other analyses.
Fig 6.
Identification and editing of crossing boundaries in complex kymographs.
Processing of kymographs containing multiple nucleation points required the following general steps: (A) Identification of the crossing points; (B) detection of edges by intensity thresholding; and (C) the final editing generating, in this example, crossing points between boundaries.
Fig 7.
Estimation of elongation rates from edges of kymograph.
(A) Kymograph of the AF647-Tpm1.1 nucleation and elongation on single actin filament. Two levels of intensities represent separate Tpm binding on two sides of the filament. (B) Edges of the kymograph were detected as described in the text. Green and magenta edges show individual Tpm1.1 elongation stretches towards barbed and pointed end of the filament, respectively. The yellow arrow demonstrates the average Tpm1.1 elongation rate extracted by linear fits to the kymograph boundary data. (C) Individual rates for each Tpm1.1 elongation segment on both grooves of actin filament. The weighted average elongation rate towards the barbed end was 136 nm s-1 and towards to pointed end 104 nm s-1 for the filament edges in Fig 3B. (D) Tpm1.1 elongation on actin towards opposite ends of the filament as a function of the feed concentration for all filaments (S1 Dataset). The lines show linear least squares fits constrained to go through the origin. These indicate that the elongation increases with the feed concentration at 1.1 nm s-1 nM-1 towards the pointed end and at 0.88 nm s-1 nM-1 towards the barbed end. The shaded area between 0 and 150 nM represents data points with slower protein delivery into the channel, which may have altered the actual concentration at which the association took place.
Fig 8.
Inferred rate versus actual rate for simulated data.
Movies of filaments elongating at different rates, 10−4 to 0.1 pixels/frame, were simulated. Kymographs were extracted, edges detected from the simulated data, and the inferred elongation rate determined in the same manner as the experimental kymographs. The data (points) was fitted to Eq (1) (line). For the example shown, the rate was inferred from segments of the boundary 10 temporal pixels in length, and the movie was simulated at a 0.1 Hz frame rate with 100 nm/pixel.
Fig 9.
Example of the Kaplan-Meier estimate of nucleation time.
The estimator data, S(t), for the filament are shown with black dots. The best fit exponential decay, , is shown by the dashed line, from which a nucleation time tnuc ≈ 1370 seconds was determined. The Kaplan-Meier analysis here was done on the filament shown in Fig 3C.