EB1 stabilizes microtubules in Xenopus egg extracts. (a) Determination of EB1 concentration in extracts. Anti-EB1 Western blot shows a titration of known input amount of bacterially expressed EB1 vs. a known volume of extract. We estimated the concentration of EB1 in extracts to be ∼8 ng/μl. (b) EB1 can be immunodepleted from extracts. Anti-EB1 Western blot of CSF and interphase extracts mock immunodepleted using mouse IgG or immunodepleted of EB1 show >95% removal of EB1 protein. (c) Coomassie-stained gel of proteins eluted from the EB1 immunoprecipitation. Identities of the eluted proteins were determined by LC/MS of bands cut from the gel and were independently established by Western blotting. Dynein heavy chain (DHC), p150^glued (glued), dynein IC (DIC), p50^dynamitin (p50), arp1/centractin (arp1), and EB1 were present in the EB1 but not the mock immunoprecipitation. The immunoprecipitating antibody heavy chains (HC) and light chains (LC) are indicated. (d) EB1 and mock immunoprecipitations probed with biotinylated EB1 show direct binding of EB1 to the p150^glued dynactin component. The band marked “glued” was identified as biotinylated EB1 bound to 150^glued because it exactly comigrated with a p150^glued band from the EB1 immunoprecipitation, run on a different portion of the same gel and directly stained with Coomassie as in c. The lower intensity bands near the p150^glued band have not been characterized. (e–h) Immunodepletion of EB1 results in shorter microtubules in CSF but not interphase, extracts. Representative centrosome-nucleated asters were visualized by rhodamine-tubulin fluorescence in mock-immunodepleted and EB1-immunodepleted crude extracts (see MATERIALS AND METHODS). (e and f) CSF extracts. (g and h) Interphase extracts. (e and g) Mock immunodepletion. (f and h) EB1 depletion. The CSF and interphase experiments are separate experiments and are not directly comparable. See text for quantitation. (i and l) Readdition of EB1 to depleted CSF extracts reveals the importance of EB1 for microtubule stability. Examples of centrosome-nucleated asters visualized by rhodamine-tubulin fluorescence are shown; separate experiments show add-back of EB1 to mock depleted and EB1-depleted extracts. These experiments are separate from the experiments in e and f and are not directly comparable because different extract preparations varied slightly in their abililty to form asters. (i) Mock depletion, no EB1 add-back. (j) EB1 depletion, no EB1 add-back. (k) Mock depletion, EB1 add-back. (l) EB1 depletion, EB1 add-back. See text for quantitation. Bars, 10 μm.

EB1 localization varies with cell cycle and microtubule polymerization status. (a) EB1 localizes uniformly along the lengths of microtubules in interphase extracts. Tubulin fluorescence and EB1 fluorescence on a centrosome-nucleated aster assembled in an interphase extract are shown, with the corresponding line scans of the microtubule indicated by the arrows (tubulin in red; EB1 in green). (b) EB1 localizes both along the sides and as a comet at the tips of polymerizing microtubules in CSF-arrested extracts. Tubulin fluorescence and EB1 fluorescence on a centrosome-nucleated aster assembled in a CSF extract are shown, with the corresponding line scans of the microtubule indicated by the arrows (tubulin in red; EB1 in green). For this line scan, the y-axes for tubulin fluorescence (left) and EB1 fluorescence (right) were plotted separately to emphasize the approximately fourfold difference in EB1 intensity between the microtubule tip and the wall. In both a and b, the line scan extends beyond the microtubule end, and EB1 fluorescence drops to background level. (c) Correlation between microtubule dynamic instability and comet appearance. One example of EB1 fluorescence on microtubules nucleated from a centrosome in CSF extracts is shown from a time lapse taken at 1.5-s intervals. Microtubule polymerization, pausing, and depolymerization are indicated. A different microtubule undergoing polymerization and depolymerization is illustrated in a kymograph. The position of the centrosome and microtubule end is indicated. Note the appearance of the EB1 comet during polymerization and its disappearance during pausing and depolymerization. The images are scaled and contrast enhanced to show the low level EB1 fluorescence along the microtubule length. Bars, 10 μm.

EB1 fluorescence recovery after FRAP reveals rapid turnover of EB1 on microtubule sides. Alexa⁴⁸⁸-EB1 was added to the extract and laser photobleaching of the microtubule (bleach site marked by an arrowhead) was done during a digital stream acquisition of 1.2 s/frame. (a) Fluorescence recovery on a microtubule in CSF extract. The bleach mark is visible in the third frame. (b) Fluorescence recovery on a microtubule in an interphase extract. The bleach mark is visible in the second frame. In both montages, the 50 frames shown represent 60 s. (c) Graphic depiction of the recovery half-lives for microtubules in CSF (green bars, n = 21) and interphase (blue bars, n = 21) extracts, with percentage of microtubules per recovery time shown for both conditions. The mean recovery half-life was 3.6 ± 2.4 s for CSF extracts and 12.0 ± 6.5 s for interphase extracts.

EB1 fluorescence intensity on polymerizing microtubule ends decays exponentially with time, consistent with a first order dissociation reaction. (a) Example of the fluorescence decay of Alexa⁴⁸⁸-EB1 at a single point on the microtubule over time. The image above the example shows the measurement axis (arrow indicates direction of measurement). The mean calculated decay half-time was 2.6 ± 1.5 s (n = 34). (b) Example of the fluorescence distribution of Alexa⁴⁸⁸-EB1 along the length of the microtubule at a single point in time. The microtubule is oriented with the distal ends facing toward the right (image above the example shows measurement axis, with arrow in direction of measurement). The mean calculated half-length of the comet was 0.7 ± 0.3 μm (n = 32).

EB1 fluorescence reveals filamentous extensions on pausing microtubules. (a) Montage of filamentous extensions at the ends of microtubules in Xenopus egg extracts, visualized by Alexa⁴⁸⁸-EB1 fluorescence. The typical radius of curvature is ∼1 μm. (b) The extensions probably contain tubulin. For the same microtubule imaged by Alexa⁴⁸⁸-EB1 and rhodamine-tubulin fluorescence, the filamentous extension is barely visible by tubulin fluorescence. (c) Rarely, extensions bifurcated and/or elongated for several microns. Frames from a time lapse of EB1 fluorescence (interval 5 s) reveal a bifurcating extension that grew several microns. (d) Typically, presence of the extensions correlated with microtubule pausing, and loss of the extension by breakage was followed by the abrupt onset of microtubule polymerization. Frames from a time lapse of EB1 fluorescence (interval 5 s) show a microtubule that on two separate occasions paused with a protofilament extension (brackets). Loss of the protofilament extension was associated with polymerization, and during the second polymerization burst a piece of the detached extension can be seen beside the polymerizing microtubule. Bars, 2 μm.

EB1 interacts with pure microtubules. Dissociation constant for EB1 binding to taxol-stabilized microtubules was calculated by best fit line to the data from an in vitro EB1-microtubule–copelleting experiment (see text for details).

EB1 binding to microtubule walls and plus ends by separate mechanisms, with constant dissociation, would produce the observed localization pattern. Microtubule wall binding (vertical arrows) is in a steady state with dissociation, producing a faint uniform localization pattern. Near the polymerizing plus end, an additional targeting mechanism occurs (horizontal arrows), either by EB1 copolymerization with tubulin or by its recognition of a structural or chemical property of the end, which allows accumulation of EB1 in excess. Dissociation shapes the tail of the comet. The excess accumulation of EB1 as a comet shape predominates during CSF arrest.