Comparative analysis of taxol-derived fluorescent probes to assess microtubule networks in a complex live three-dimensional tissue

Abstract

Drosophila oogenesis is an excellent in vivo model for investigating cytoskeletal dynamics because of the rapid cytoskeletal remodeling that occurs at the end of stage 10; however, there are few robust tools for detecting microtubules in live complex tissues. The recent development of membrane permeable taxol-based fluorescent probes to label microtubules is significant technical progress, but the effectiveness of these probes and the potential stabilizing effects of the taxol derivative have not been well characterized in vivo. Here, we compared three commercially available taxol-derived microtubule labels to determine their efficacy and potential artifacts. We found that all three probes labeled microtubules with differences in permeability, brightness, and signal to noise ratio. Like taxol, however, all of the probes disrupted the F-actin cytoskeleton at higher concentrations. We also found that the efflux pump inhibitor, verapamil, increased the intensity of the label and modestly increased the severity of the F-actin defects. Of the three probes, Tubulin Tracker (ThermoScientific) was the most permeable and was brightest, with the highest signal to noise ratio. Furthermore, washing out the probe after a 30-min incubation significantly reduced the F-actin artifacts without compromising signal brightness.

Keywords: F-actin; cytoskeleton; drosophila; microtubules; oogenesis; taxol.

FIGURE 1

Relative fluorescent intensity, signal to noise ratio, and permeability of taxol-derived fluorescent probes in Drosophila nurse cells. (a–a”) Schematic of egg chambers during stages 10A through 11 shows that egg chambers are composed of an oocyte and nurse cells connected by ring canals; F-actin cable arrays develop during stage 10B; and nurse cells contract during stage 11 in a process called “dumping.” (b) Microtubules can be visualized, but with very poor resolution, by expressing UAS-α-tubulin-GFP using a germline Gal4 driver (MTD-Gal4). (c–e) Brightness corrected images of maximum projections show labeling of nurse cell microtubules after 90-min incubations with Tubulin Tracker, Viafluor, and SiR-Tubulin, respectively. Yellow boxes represent the high magnification insets (c’–e’). (c”–e”) Raw images of c–e show relative signal intensity of the three probes. (f) The maximum pixel intensity along the midline of the raw images like c”–e” (yellow dashed line in c”, normalized to intensity at t0) increased over time, with Tubulin Tracker being significantly brighter. (g) The signal to noise ratio is highest with Tubulin Tracker. (h) Early timepoints during live imaging show microtubules (arrowheads) only at the peripheral edge of the egg chamber. (i) Microtubules at the peripheral edge of the egg chambers appear significantly earlier in Tubulin Tracker treated egg chambers compared to Viafluor and SiR-Tubulin. (j) Late timepoints during live imaging show microtubules (arrowheads) throughout the egg chamber. (k) Microtubules at the center of the egg chambers appear significantly earlier in Tubulin Tracker treated egg chambers compared to Viafluor and SiR-Tubulin. Scale Bar = 25 μm. *One-way ANOVA, Tukey's Multiple Comparison, N = 6 egg chambers/treatment

FIGURE 2

Reducing the concentration of Tubulin Tracker significantly decreased brightness, signal to noise ratio, and permeability. (a–c) Brightness corrected images show that microtubules are visible in egg chambers treated with Tubulin Tracker at three dilutions 1:1,000 (a), 1:5,000 (b), and 1:10,000 (c), although microtubules become less visible with each dilution. Yellow boxes represent the high magnification insets (a’–c’). (a”–c”) Raw images of (a)–(c) show relative signal intensity of the three dilutions. (d) The maximum pixel intensity along the midline of egg chambers increased over time, and has lower intensity with each dilution. (e) The signal to noise ratio is significantly lower in egg chambers treated with 1:5,000 and 1:10,000 Tubulin Tracker. (f,g) Microtubules at the peripheral edge and at the center of the egg chambers appear later with lower concentrations of Tubulin Tracker. Scale bar = 25 μm. *One-way ANOVA, Tukey's Multiple Comparison, N = 6 egg chambers/concentration

FIGURE 3

The effects of taxol and taxol-derived probes on F-actin cables. (a) The paradigm for determining the effect of taxol and taxol-derived probes was to collect whole ovaries, incubate them in differing doses of taxol or taxol-drived probes for 90 min, followed by fixing and staining with phalloidin to observe the effect of the probes on the F-actin cytoskeleton. (b–d) Treating egg chambers with taxol resulted in two concentration dependent F-actin cable defects including a complete loss of cables (b). This was often accompanied by the presence of clumps of phalloidin positive material near the membrane (arrowheads). (c) We also observed F-actin cable detachment from the cortex (arrowheads) in a fraction of taxol-treated egg chambers. (d) Normal F-actin cables were also observed. (e) Quantification of taxol dose-dependent F-actin cable defects. DMSO is the vehicle control. (f and f”) Quantification of dose dependent F-actin cable defects by taxol-derived probes. (g) For the analysis of dose-dependent effects of the fluorescent microtubule probes on F-actin cables, we combined the egg chambers without cables and those with detached cables into a single category of “abnormal cables.” The effects of the three probes were indistinguishable and dose dependent. (ANOVA Tukey's Multiple Comparison, N = 3 replicates, over five egg chambers/replicate). Scale bar = 25 μm

FIGURE 4

Effect of verapamil on the labeling efficacy of SiR-Tubulin and associated F-actin cable defects. (a and b) Brightness corrected images show microtubule labeling of nurse cells after 90-min incubations with SiR-Tubulin and verapamil (a) or SiR-Tubulin and DMSO (b). Yellow boxes represent the high magnification insets (a’ and b’). (a” and b”) Raw images of (a) and (b) show relative signal intensity of SiR-Tubulin with verapamil versus DMSO. (c) The maximum pixel intensity along the midline of egg chambers is five-fold higher in egg chambers treated with SiR-Tubulin and verapamil compared to SiR-Tubulin and DMSO. (d) Egg chambers treated with SiR-Tubulin and verapamil have a slightly higher signal to noise ratio than egg chambers treated with SiR-Tubulin and DMSO. (e and f) Microtubules at the peripheral edge of the egg chambers (e) and at the center of the egg chambers (f) appeared earlier when treated with verapamil and SiR-Tubulin. (g) Treating egg chambers with SiR-Tubulin and verapamil significantly increased the percent of abnormal egg chambers compared to SiR-Tubulin and DMSO (Two-way ANOVA, N = 3 replicates, over five egg chambers/replicate). Scale Bar = 25 μm. *Two-way t test, N = 6 egg chambers/treatment

FIGURE 5

The effect of Tubulin Tracker washout on microtubule labeling and F-actin cytoskeleton disruption. (a) For the washout experiments, we treated egg chambers for 30 min with Tubulin Tracker (Washout) or DMSO (Control) followed by three rinses in live imaging media and 90 min of imaging in live imaging media with DMSO (Washout) or Tubulin Tracker (Control). (b and c) Brightness corrected images show labeling efficiency of nurse cells in the washout group treated with 1:1,000 Tubulin Tracker (b) or the control (c) at the end of the 90-min imaging period. Yellow boxes represent the high magnification insets (b’ and c’). (b” and c”) Raw images of (b) and (c) show relative signal intensity of the washout versus control. (d) The maximum pixel intensity along the midline of egg chambers remains relatively constant over the imaging period for the washout and increases in the control. (e) Washout treatment increased the signal to noise ratio compared to control. (f and g) Microtubules are immediately visible at the peripheral edge and in the center of the egg chambers after washout. (h) Washout of Tubulin Tracker significantly decreased the frequency of F-actin cable defects on the F-actin cytoskeleton at the highest concentration of Tubulin Tracker (Two-way ANOVA, Bonferroni's Multiple Comparison; N = 3 replicates, over five egg cambers per replicate). Scale bar = 25 μm. *Two-way t test, N = 6 egg chambers/treatment