Visualization of microtubule growth in living platelets reveals a dynamic marginal band with multiple microtubules

Abstract

The marginal band of microtubules maintains the discoid shape of resting blood platelets. Although studies of platelet microtubule coil structure conclude that it is composed of a single microtubule, no investigations of its dynamics exist. In contrast to previous studies, permeabilized platelets incubated with GTP-rhodamine-tubulin revealed tubulin incorporation at 7.9 (+/- 1.9) points throughout the coil, and anti-EB1 antibodies stained 8.7 (+/- 2.0) sites, indicative of multiple free microtubules. To pursue this result, we expressed the microtubule plus-end marker EB3-GFP in megakaryocytes and examined its behavior in living platelets released from these cells. Time-lapse microscopy of EB3-GFP in resting platelets revealed multiple assembly sites within the coil and a bidirectional pattern of assembly. Consistent with these findings, tyrosinated tubulin, a marker of newly assembled microtubules, localized to resting platelet microtubule coils. These results suggest that the resting platelet marginal band contains multiple highly dynamic microtubules of mixed polarity. Analysis of microtubule coil diameters in newly formed resting platelets indicates that microtubule coil shrinkage occurs with aging. In addition, activated EB3-GFP-expressing platelets exhibited a dramatic increase in polymerizing microtubules, which travel outward and into filopodia. Thus, the dynamic microtubules associated with the marginal band likely function during both resting and activated platelet states.

Figure 1

Localization of microtubule plus ends in the platelet marginal band. (A) Methanol-fixed resting mouse platelets labeled with anti-EB1 antibodies. EB1 “comets” () are labeled along the marginal band. An average of 8.66 (± 2.11; n = 26) comets were observed along the marginal band. (B) Histogram showing EB1 signals observed along the microtubule coil of each mouse platelet. The number of signals ranged from 4 to 12, with a median of 8. (C) Anti-EB1 staining in resting human platelets. EB1 comets () are identified. An average of 8.9 (± 1.5; n = 34) comets are seen along each coil. (D) Histogram of EB1 signals along the microtubule coil of human platelets. The number of signals ranged from 6 to 13, with a median of 9.

Figure 2

Localization of rhodamine-tubulin in platelets. (A) Mouse platelets permeabilized in 0.4% octyl-β-d-glucopyranoside (OG) were incubated with rhodamine-labeled tubulin in 0.1 mM GTP, washed, and then formaldehyde-fixed. Rhodamine-labeled tubulin incorporation at multiple points around the marginal band indicates that free microtubule plus ends exist within the microtubule coil. An average of 7.9 (± 1.9; n = 48) signals of rhodamine-labeled tubulin were observed along the periphery of each platelet. (B) Histogram of fluorescent tubulin signals along the marginal band. The number of signals ranged from 4 to 12, with a median of 8. (C) Fluorescence micrograph of purified mouse platelets that had been permeabilized in 0.5% Triton X-100, incubated in rhodamine-labeled tubulin, and then washed and fixed in formaldehyde. Under these permeabilization conditions, rhodamine-labeled tubulin incorporation occurred at a single site () along the platelet marginal band.

Figure 3

Multiple microtubules polymerize along the marginal band in living platelets. (A) Fluorescence microscopy of platelets released from cultured mouse megakaryocytes directed to express EB3-GFP showing the movements of multiple EB3-GFP comets, an indicator of polymerizing microtubules, along the periphery of the platelet. The time course follows a platelet with EB3-GFP comets moving both clockwise () and counterclockwise () over the course of 32 seconds. The final panel (76 seconds) shows the same cell with several EB3-GFP comets and foci (). (B) Life history plots of 8 individual EB3-GFP comets within a resting platelet show distance traveled over time. (C) Western blot detection of endogenously expressed EB3 in platelet lysates.

Figure 4

The microtubule coil contains both stable and dynamic microtubules. (A) Immunofluorescence analysis of mouse platelets using antibodies to Tyr-tubulin, Glu-tubulin, and acetylated tubulin indicates that all 3 modified tubulin forms exist in the marginal band. Cells were double-labeled with an antibody to β1-tubulin. (B) Immunofluorescence analysis of mouse platelets using antibodies to Tyr-tubulin (top panel) and acetylated tubulin (middle panel). Note that both isoforms accumulated within platelet marginal bands, although some platelets appear to label more strongly with one antibody than with the other (bottom panel). (C) Western blot of mouse platelet lysates; equal protein quantities were loaded and probed with antibodies to Glu-tubulin (glu), acetylated tubulin (ace), and Tyr-tubulin (tyr).

Figure 5

Microtubule coil diameters decrease in aging platelets and inhibition of microtubule dynamics prevents microtubule coil shrinkage. (A) Immunofluorescence image of platelets from mice treated with RAMPS (right) and control platelets (left) stained with antibodies to β1-tubulin show that young platelets have larger microtubule coil diameters than the microtubule coils of normal circulating platelets. (B) RAMPS treatment alters the average platelet diameter. RAMPS-treated platelets averaged 3.56 (± 0.5) μm (n = 31) in diameter, compared with 2.88 (± 0.3) μm (n = 31) for control platelets. Error bars indicate SD. (C) 5-FU treatment alters the average diameter of platelets. 5-FU–treated platelets averaged 3.54 (± 0.7) μm compared with 2.94 (± 0.4) μm (n = 85) in control platelets. After an additional 20 days, diameters of platelets from 5-FU–treated mice return to control levels. (D) Average microtubule coil diameters of human platelets treated with paclitaxel and nontreated control cells at day 0 and day 3 (n = 150 for each sample). Cells treated with paclitaxel for the 3-day incubation period maintain coil sizes similar to day-0 control cells, whereas day-3 cells incubated without paclitaxel possess smaller microtubule coils. Error bars indicate SD. (E) Immunofluorescence images of human platelets visualized with tubulin antibody labeling. Comparison of day-0 and day-3 platelets plus or minus paclitaxel, as shown, monitors the effect of paclitaxel on the reduction of microtubule coil size.

Figure 6

Visualization of microtubule polymerization in living platelets during activation. (Ai) Thrombin-activated human platelets fixed and stained with anti-EB1 antibodies. An average of 25.5 (± 8.6) sites of EB1 localization are seen. (Aii) Thrombin-activated human platelets fixed and double-labeled with anti-EB1 (green) and antidetyrosinated tubulin (red) antibodies. (B) A platelet released from a megakaryocyte infected with EB3-GFP. Note the microtubule growth (indicated by the fluorescent comets) that occurred during activation. (C) In vivo labeling of extended filopodia with multiple EB3-GFP comets. Note (arrowheads) the movement of a single EB3-GFP comet, which appears to move away from the cell body. (D) Kymograph of EB3-GFP comet movements; platelet cell body faces up in this figure. Arrowheads track the movements of a single EB3-GFP comet in consecutive frames.

Figure 7

Localization of γ-tubulin in resting and activated platelets. (A) Purified mouse platelets labeled with anti–γ-tubulin antibody show the localization of microtubule minus ends along the microtubule coil. γ-Tubulin foci averaged 9.06 (± 1.61) signals along the microtubule coil of each platelet (N = 31). (B) Histogram of γ-tubulin signals along the marginal band of mouse platelets. The number of signals ranged from 6 to 12, with a median of 9. (C) Anti–γ-tubulin–stained human platelets show localization at multiple points along the microtubule coil. An average of 9.2 (± 1.7; n = 33) γ-tubulin signals are seen along the coil. (D) Histogram of γ-tubulin signals on the marginal band of human platelets. The number of signals per platelets ranged from 6 to 13, with a median of 9. (E) Western blot detection of γ-tubulin in mouse and human platelet lysates, and HEK 293 cell lysate. Samples of equal protein concentrations were examined. The single protein band detected in each lane shows a relative mobility of approximately 48 kDa, corresponding to the molecular weight of γ-tubulin. (F) Purified human platelets activated with thrombin were formaldehyde-fixed and incubated in anti–γ-tubulin antibody (green) and anti–αβ-tubulin antibody (red). Note the γ-tubulin labeling within the central portion of the activated platelets, while the αβ-tubulin extends outward within filopodia.

Figure 8

Model of the resting platelet marginal band. The model illustrates multiple microtubule coils forming from a single stable microtubule (blue), while several dynamic microtubules (green arrows) polymerize in both directions around the marginal band of the resting platelet.