Figure 4

The growth mechanism in VASP anchored actin networks is dependent on the presence of NMMII. (a) Time resolved TIRF images of a VASP anchored actin (magenta) network at 200 s and 600 s after start of actin polymerization. (b) Time frames were subtracted from each other to visualize freshly incorporated actin over time. The subtracted frames are color coded with a temporal lookup table from red over white to blue to show that filament growth occurs alongside previously formed bundles. This is apparently due to barbed-end elongation of filaments that are guided and captured by VASP within the bundles. (c) Time resolved images of the network in the presence of NMMII. Here, clusters appeared that grew without visible contractions towards the clusters. (d) Subtracted time frames of the networks with NMMII. Here, barbed end elongation along previously formed filaments could not be observed. Monomers were incorporated over the entire area of the cluster. Partly, the intensity in the subtracted frames was affected by the fluctuations caused by the activity of NMMII. Time is in seconds, scale bars = 5 µm. (e) The width of arbitrarily chosen bundles and clusters over time. On average, in the presence of NMMII bundles grew much wider (up to 2 µm in width). The increase in width without NMMII depends on the barbed-end elongation of filaments and thus occurs on a much slower time scale. At least 10 individual bundles or clusters were measured. (f) A fraction of 20% labeled NMMII (cyan) was used to visualize the localization of the motor in the network. Where motors were present, clusters slowly appeared in the network which indicated a strong correlation between the motor activity and the emergence of actin clusters. A line is drawn at one of the bundles emerging from a cluster with labeled active motors to identify the directionality of filaments growing along the bundle. Scale bars = 3 µm. (g) A Kymograph analysis. The bottom kymograph has dashed, yellow lines to better visualize individual filaments growing alongside the cluster while the unedited kymograph is shown on top. Here, it is shown that the bundles emerging from the clusters display directional growth. Most filaments that grew on the bundles emerging from clusters were elongated away from the cluster but not towards the cluster. (h) A flow analysis of the network growth with VASP and actin only reveals that the direction of actin polymerization (red arrows) is mostly alongside bundle orientation (white lines). Scale bar = 1 µm. (i) A flow analysis in the presence of NMMII reveals how clusters are formed in a swelling manner, as the actin mass flow (red arrows) grows radially from the inside towards the outside of the cluster, regardless of the local bundle orientation (white lines). Scale bar = 1 µm. (j) The mean absolute value <|cos(Θ)|> of the cosine of the angle Θ between the local actin’s flow and the local bundle orientation shows how uniformly the vectors are distributed in both cases. For a random collection of angles, the expected value is 2/π ~ 0.64, while 1 would represent a network where all vectors point towards the same direction. In the presence of only VASP, the alignment between bundles and actin flow is relatively high (0.8) due to the growth of filaments alongside previously formed bundles. In the presence of VASP and NMMII, the mean angle is initially very close (0.75) but slowly decays (0.65 after 10 min) due to the altered growth mechanism by cluster swelling. The thick lines are a 3 points average.