Figure 1.
GXP (GTP/GDP/GMPCPP) islands promote microtubule rescue.
(A) Schematic of the total internal reflection fluorescence (TIRF) microscopy setup for incorporating rhodamine-labeled GXP islands into alexa-488-labeled microtubules. (B) Kymograph of microtubule growth and depolymerization taken from time-lapse microscopy (Movie S1). Three types of rescue events were observed at GXP islands: (top right) rescue at the end of the GXP island, in which islands exhibited no detectable depolymerization; (middle right) rescue in the middle of the GXP island, in which islands partially depolymerized prior to re-growth; (bottom right) catastrophe through the island, in which the GXP island completely depolymerized.
Figure 2.
Island size and composition affect rescue efficiency.
(A) The probability of rescue at either the middle or end of an island increased with the size of the island (see Table S1 for details). (B) Islands with a higher GMPCPP content (74% GMPCPP, island length ≤1 µm; 24% no rescue, 20% rescue in the middle of the island, 56% rescue at the end of the island; N = 41) had a higher probability of rescue than islands with a lower GMPCPP content (50% GMPCPP, island length ≤1 µm; 43% no rescue, 21% rescue in the middle of the island, 36% rescue at the end of the island; N = 14). See Table S1 for islands of all sizes.
Figure 3.
Island dynamics are distinct from GDP-tubulin extensions and GMPCPP seeds.
(A) Shortening rates of GDP-tubulin extensions (white solid line) and GXP islands (white dashed line) and the time to microtubule regrowth (yellow double arrow measured from extrapolated yellow dotted line) were measured from kymographs of depolymerizing microtubules. (B) GXP islands shorten more slowly than GDP-tubulin extensions. Distribution of shortening rates measured for extensions and islands (N = 30). The inset shows a box-plot of the data, which has a median shortening rate of 0.49 µm/s for extensions and 0.16 µm/s for islands (p<0.00001, Wilcoxon rank-sum test). Only cases where extension and island shortening could both be measured in the same depolymerization event were analyzed. (C) Microtubules grow more readily from islands than from GMPCPP seeds. Time to regrowth was measured from kymographs of dynamic microtubules and was on average shorter from islands than from GMPCPP seeds. Median time to regrowth was 6 seconds for GXP islands (N = 60) and 27 seconds for GMPCPP seeds (N = 60) (inset, p<0.00001, Wilcoxon rank-sum test; outliers >100 are not shown).