The media player is loading...
Movie S7. Two-photon imaging of the confinement of PAGFP-actin fluorescence after repetitive photoactivation and uncaging of MNI-glutamate (60 pulses of 0.6 ms duration at 1 Hz) at a point (square) distal to the apex of the spine shown in Figure 5D. The 2D images were acquired every 10 s for 7 min. Photoactivation was induced at the moment when the white square turns red. Scale bar, 1 μm.
The media player is loading...
Movie S8. Two-photon imaging of PAGFP-actin fluorescence after repetitive photoactivation and uncaging of MNI-glutamate (60 pulses of 0.6 ms duration at 1 Hz) at a point (square) distal to the apex of the spine shown in Figure 5G in the presence of Lat A (0.1 μM). The 2D images were acquired every 15 s for 10 min. Photoactivation was induced at the moment when the white square turns red. Scale bar, 1 μm.
A new pool of F-actin formed during the enlargement of the spine with this uncaged-glutamate protocol (Figures 5D–5F and Movie S7). Unlike the two pools of F-actin labeled by point activation in resting spines (Figures 2 and 5A), the “enlargement pool” was distributed throughout the spine head (Figures 5D and 5E). During the growth of the spine, we often detected synchronized ruffling of the spine head and the enlargement pool (Figure 5D, yellow arrows, and Movie S7), as if the polymerization of new actin fibers were driving the enlargement. The PAGFP-actin fluorescence in the head was relatively stable, decaying to 50% of the maximal value more than 4 min after photoactivation/glutamate uncaging in 56% (24/43) of spines (Figure 5F). In the rest, the decay reached 50% within 1.5–4 min, as described below. The formation of this new, stable pool is consistent with a previous observation that the turnover of F-actin in spines went down after cultured neurons were stimulated (Star et al., 2002). The flow of actin in the dynamic pool was not readily apparent during spine enlargement in most experiments (Figures 5D and 5E), possibly because fluorescence signals from the enlargement pool dominated those from the dynamic pool.
The growth of the spine and formation of the enlargement pool were abrogated in the presence of inhibitors of actin polymerization or calmodulin. Indeed, 0.1 μM Lat A caused actin dynamics to look like controls during the glutamate-uncaging experiments (Figures 5G–5I and Movie S8). The spines did not enlarge (Matsuzaki et al., 2004) (Figure 5I). This low concentration of Lat A did not appear to affect the dynamic and stable pools of filaments because the flow of actin was unchanged (Figures 5G and 5H), and the time course of decay was similar to that of controls (Figure 5I). Two inhibitors of calmodulin, W7 (20 μM, n = 14) and calmidazolium (20 μM, n = 10) (data not shown), also blocked the formation of the enlargement pool and growth of the spine (Matsuzaki et al., 2004). Thus, the pool of F-actin that forms in tandem with spine enlargement was distinct from the other filament pools in terms of stimulus dependence, distribution, stability, and pharmacological properties.
Fluorescence of activated PAGFP-actin in the enlargement pool decayed within 4 min in 44% of spines (19/43). In these cases, we often detected an outflow or release of this entire pool of fibers from the spine head into the dendritic shaft (Figure 6A and Movie S9). In 42% of these cases, a small proportion of the activated PAGFP-actin within those fibers remained in the shaft after release (Figure 6A, red arrow, and Movie S9).