2391
Journal of Cell Science 110, 2391-2401 (1997)
Printed in Great Britain © The Company of Biologists Limited 1997
JCS8164
Non-centrosomal microtubule formation and measurement of minus end
microtubule dynamics in A498 cells
Anne-Marie C. Yvon1 and Patricia Wadsworth1,2,*
1Program
in Molecular and Cellular Biology and 2Department of Biology, University of Massachusetts, Amherst, MA 01003, USA
*Author for correspondence (e-mail: [email protected])
SUMMARY
Experiments performed on a cell line (A498) derived from
a human kidney carcinoma revealed non-centrosomal
microtubules in the peripheral lamella of many cells. These
short microtubules were observed in glutaraldehyde-fixed
cells by indirect immunofluorescence, and in live cells
injected with rhodamine-labeled tubulin. The non-centrosomal microtubules were observed to form de novo in living
cells, and their complete disassembly was also observed.
Low-light-level fluorescence microscopy, coupled to imaging
software, was utilized to record and measure the dynamic
behavior of both ends of the non-centrosomal microtubules
in these cells. For each, the plus end was differentiated from
the minus end using the ratio of their transition frequencies
and by measuring total assembly at each end. For comparative purposes, dynamics of the plus ends of centrosomally
nucleated microtubules were also analyzed in this cell line.
Our data reveal several striking differences between the
plus and minus ends. The average pause duration was
nearly 4-fold higher at the minus ends; the percentage of
time spent in pause was 92% at the minus ends, compared
to 55% at plus ends. Dynamicity was decreased 4-fold at the
minus ends, and the average number of events per minute
was reduced from 7.0 at the plus end to 1.5 at the minus
ends. The minus ends also showed a 6-fold decrease in
frequency of catastrophe over the plus ends. These data
demonstrate that in living cells, microtubules can form at
sites distant from the perinuclear microtubule organizing
center, and once formed, non-centrosomal microtubules can
persist for relatively long periods.
INTRODUCTION
sites other than the nuclear surface, consistent with the idea of
a dispersed or flexible centrosome in these cells (Mazia, 1987).
In other cases, the MTOC may function to nucleate microtubules, but the microtubules are subsequently released and
rearranged. Perhaps the best example of this is in neuronal cells,
where recent work indicates that axonal and dendritic microtubules are nucleated and assembled at the cell body’s MTOC,
and then transported into the developing processes (reviewed by
Baas, 1997; Sharp et al., 1995). This results in numerous microtubules in the axon that are no longer associated with the MTOC,
and thus have two ‘free’ ends. In another example, that of
nucleated erythrocytes, the microtubule array consists of a
peripheral, hoop-like bundle of microtubules called the marginal
band. These microtubules originate in association with a conventional MTOC and are subsequently released and rearranged
into the characteristic peripheral bundle (Cohen, 1991).
Individual microtubule behavior is described by the model of
dynamic instability, whereby stochastic transitions between
assembly and rapid disassembly occur (Caplow, 1992; Erickson
and O’Brien, 1992; Mitchison and Kirschner, 1984a). Although
it is rarely observed with purified tubulin, an attenuated or
paused state, where no loss or addition of tubulin subunits can
be detected, has been observed in vivo (Shelden and
Wadsworth, 1993; Walker et al., 1988). Dynamics of microtubule plus ends have been measured in vitro and in vivo by
Microtubules are cytoskeletal polymers composed of α and β
tubulin heterodimers; this arrangement produces an inherent
polarity in their structure, resulting in differences in their ends.
In most cultured interphase animal cells, the minus ends of the
microtubules are embedded in the perinuclear microtubule organizing center (MTOC) or centrosome, while the plus ends extend
out in a radial array toward the periphery of the cell. The MTOC,
or centrosome, is a major site of microtubule nucleation in cells,
and isolated MTOCs can nucleate microtubule assembly under
conditions where spontaneous microtubule formation does not
occur (Mitchison and Kirschner, 1984b). In many cell types,
however, a perinuclear MTOC and a radial arrangement of
microtubules is not observed. For example, in polarized epithelial cells, the microtubule organizing material is dispersed
across the apical region of the cell and an array of parallel microtubules extends from it toward the basal end of the cell (Bacallao
et al., 1989; Bre et al., 1987). A reorganization of the microtubule nucleating material, from a conventional pericentriolar
localization to one on the nuclear surface, is seen in multinucleate myotubes following myoblast fusion (Tassin et al., 1985). In
higher plant cells, a well defined MTOC is absent; instead, the
surface of the nucleus appears to be a major site of nucleation
(Lambert, 1993). Plant cell microtubules can also assemble at
Key words: Microtubule dynamics, Minus end, Non-centrosomal
microtubule
2392 A. C. Yvon and P. Wadsworth
various groups (Cassimeris et al., 1988; Sammak and Borisy,
1988; Schulze and Kirschner, 1986; Shelden and Wadsworth,
1993; Walker et al., 1988); the data clearly demonstrate that the
dynamic behavior of microtubule plus ends is regulated
throughout the cell cycle and is cell type specific (Belmont et
al., 1990; Shelden and Wadsworth, 1993). Although the
dynamic behavior of microtubule minus ends has been characterized in in vitro experiments (Kowalski and Williams, 1993b;
Panda et al., 1996; Vasquez et al., 1994; Walker et al., 1989),
the usual inaccessible position of minus ends in the centrosome
has largely prevented previous measurement of dynamic instability parameters in vivo (Rodionov and Borisy, 1997).
In addition to dynamic instability behavior, microtubules also
exhibit opposite-end assembly-disassembly behavior, or treadmilling (Margolis and Wilson, 1981). Recently, treadmilling of
individual microtubules has been observed in experimentally
produced cytoplasmic fragments which lack a conventional
MTOC (Rodionov and Borisy, 1997). In addition, photoactivation experiments show the poleward movement of marked spots
on spindle microtubules, a phenomenon termed poleward flux
(Mitchison, 1989). For flux to occur, the loss and gain of tubulin
subunits at opposite microtubule ends must be coupled to
translocation of the entire microtubule lattice (Mitchison and
Sawin, 1990). The demonstration of treadmilling in vivo and
the observation of flux, both in vitro (Sawin and Mitchison,
1991) and in vivo (Mitchison, 1989; Mitchison and Salmon,
1992), directly demonstrate that subunits exchange at microtubule minus ends under some circumstances.
In the experiments reported here, we have observed the
formation of individual, non-centrosomal microtubules in the
peripheral lamellae of a human kidney carcinoma cell line. Our
results demonstrate that individual microtubules can form at
sites distant from the MTOC and that once formed, these
microtubules can persist for many minutes. Since these microtubules are observed in an area of the cell that does not have a
dense microtubule population, we were able to directly
measure the behavior of both ends of these microtubules. Our
data demonstrate that one end of the non-centrosomal microtubules was active, undergoing the process of dynamic instability. The other end was extremely stable, spending an average
of 94% of the time in a state where no changes in length could
be detected. The centrosomal and non-centrosomal microtubule plus ends in these cells were less dynamic than microtubules previously measured in various cells, supporting the
view that microtubule stability may contribute to the persistence of non-centrosomal microtubules in these cells.
MATERIALS AND METHODS
Materials
All materials for cell culture were obtained from Gibco-BRL, Gaithersburg, MD, with the exception of fetal calf serum, which was obtained
from Hyclone Laboratories, Inc., Logan, UT. A498 cells were the kind
gift of Drs Mary Ann Jordan and Leslie Wilson, University of California, Santa Barbara. Unless otherwise noted, all other chemicals were
obtained from Sigma Chemical Company, St Louis, MO.
Cell culture and microinjection
All cells were grown in 10% fetal calf serum and antibiotics, with 5%
CO2, at 37°C. A498 and PtK-1 cells were grown in 90% MEM, supplemented with 1.0 mM sodium pyruvate; A431 and BSC-1 cells were
grown in 90% DME; CaOV-3 cells were grown in 90% DME, sup-
plemented with 4.5 g/l glucose; CHO cells were grown in 90% Ham’s
F-12. For use in experiments, cells were plated on etched glass coverslips (Bellco Glass Co., Vineland, NJ) 24-72 hours before use. For
microinjection and subsequent observation, cells were transferred to
medium containing 10 mM Hepes buffer, pH 7.3, lacking bicarbonate. Pressure microinjection was performed on a Zeiss IM-35 inverted
microscope using a ×32 phase objective lens, an Eppendorf 5242
microinjector and a Narishige micromanipulator. Needles were pulled
from Omega Dot capillary glass tubes (Friedrich and Dimmock,
Millville, NJ) on a Brown-Flaming P-80 micropipette puller.
Rhodamine tubulin was microinjected at a needle concentration of 2.2
mg/ml; for co-injection experiments, fluorescent dextran (Molecular
Probes, Eugene, Oregon) and tubulin were injected at needle concentrations of 0.1 mg/ml and 3.3 mg/ml, respectively. Solutions to be
microinjected were centrifuged in an Eppendorf 5415C centrifuge at
14,000 rpm for 15 minutes immediately prior to injection, and the
supernatant was transferred to a clean tube. Following injection, the
cells were returned to a 37°C incubator for at least 2 hours to allow
incorporation of the labeled tubulin into microtubules.
Immunocytochemistry
Immunofluorescence localization of microtubules was performed on
cells fixed in 0.25% glutaraldehyde in PBS for 1 minute, followed by
brief lysis (1 minute) in a solution containing 0.5% Triton X-100, 1 mM
MgSO4, 5 mM EGTA, 80 mM Pipes, pH 6.8, with additional fixation in
0.25% glutaraldehyde for 5 minutes. Cells were rinsed in PBS containing 0.1% Tween-20 and 0.02% Azide (PBS-Tw-Az) and then incubated
for 1 hour at 37°C in a monoclonal anti-α-tubulin antibody (DM1a;
1:100 dilution in PBS-Tw-Az with 1% BSA). Cells were then rinsed in
PBS-Tw-Az and incubated in affinity purified, fluorescein-conjugated,
goat anti-mouse secondary antibodies (Organon Teknika, Durham, NC;
1:50 dilution in PBS-Tw-Az with 1% BSA) for 30-45 minutes at room
temperature. Stained cells were mounted in 90% glycerol, containing
0.1% p-phenylenediamine, and sealed with nail polish. Immunofluorescent preparations were observed and photographed on a Zeiss Axiovert
microscope using a ×63 neofluor objective lens and TMax 400 film. To
obtain images for reproduction, 35mm negatives were digitized using a
Sprint Scan-35 negative scanner and Adobe Photoshop software running
on a Macintosh 8100 and printed using a Tektronix Phaser 440 printer.
Preparation of rhodamine-labeled tubulin
Rhodamine-labeled tubulin was prepared as described previously
(Shelden and Wadsworth, 1993). Briefly, phosphocellulose purified
porcine brain tubulin (Sloboda et al., 1976) was assembled in a buffer
containing 100 mM Pipes, 10 mM MgSO4, 2 mM EGTA, 4 M
glycerol, and 1 mM GTP, pH 6.9, and then incubated for 10 minutes
at 37°C with a 40-fold molar excess of 5,6-carboxy succinimidyl ester
of rhodamine (Molecular Probes, Eugene, OR). The fluorescent
microtubules were collected by centrifugation (100,000 g for 2 hours)
through a 40% sucrose cushion. The resulting pellet was resuspended
in a buffer containing 1 M glutamate, 1 mM EGTA and 0.5 mM
MgSO4, at pH 6.9, and disassembled on ice for 30 minutes. The
protein was centrifuged for 45 minutes at 100,000 g, and the supernatant was retained. The supernatant was then carried through 2 cycles
of temperature-dependent assembly and disassembly. The final pellet
was resuspended in injection buffer (20 mM glutamate, pH 7.0, 0.5
mM MgSO4, and 1 mM EGTA) and stored at −80°C in small aliquots.
Low-light level fluorescence microscopy and data analysis
Images of rhodamine-labeled microtubules in the injected cells were
obtained as previously described, using a ×100 1.3 NA apochromat
lens (Shelden and Wadsworth, 1993). A Dage ISIT video camera,
operated at maximum gain, was used to collect images, which were
digitized using a Perceptics PixelPipeline card in a Macintosh Quadra
950 running BDS Image (Oncor Inc., Gaithersburg, MD) software.
Thirty-two frame averages were collected at 2 second intervals, stored
on a Perceptics PixelBuffer store card, and subsequently transferred
Non-centrosomal microtubule dynamics 2393
to optical discs. For coverslips to be immunostained for total tubulin
following live cell imaging, only one fluorescence image was
collected. In order to facilitate re-location of the same cell following
staining, a phase image was also collected using a Dage SIT video
camera. Coverslips were fixed within 2 minutes of fluorescence image
collection and stained as described above under Immunocytochemistry. The previously imaged cell was identified and a second set of
phase and fluorescence images, to verify the location and existence of
the non-centrosomal microtubule, was collected as described above.
Quantification of microtubule dynamic behavior was performed using
laboratory-written software (Shelden and Wadsworth, 1993). The
position of an individual microtubule in a sequence of images was
marked using a mouse-driven cursor; the position information was used
to create a microtubule ‘life history’ plot using the Cricket Graph
program (Cricket Software, Malvern, PA). Phases of growth, shortening
and pause were marked on the plots; additional in-house software was
used to determine the duration, distance, and rate of growth and shortening events and the duration of pause events. Due to the detection limits
of this method, only changes of greater than 0.5 µm were considered
growth or shortening events; we estimate that the initial bright dot during
non-centrosomal microtubule formation must be approximately 0.5 µm
(or 800 tubulin dimers) to be detected. Data for each microtubule was
entered into a spreadsheet (Kaleidagraph, Synergy Software, Reading,
PA) for statistical analysis. The frequency of catastrophe was determined
by dividing the sum of growth to shortening and pause to shortening
transitions, by either the sum of the distance grown or the sum of the
time in growth and pause. The frequency of rescue was determined by
dividing the sum of shortening to growth and shortening to pause transitions, by the time spent shortening or the distance shortened. Dynamicity was calculated by dividing the sum of the total length grown and
shortened, by the time of observation of the particular microtubule.
To correlate the amount of unpolymerized fluorescent tubulin with
cell thickness, images of cells co-injected with fluorescein dextran and
rhodamine tubulin were collected. The thicker edge and the flatter area
adjacent to it were compared using the ratios of pixel intensity per
unit area between the two regions. These were determined by
selecting a portion of both regions of the cell with the cursor and
measuring the fluorescence intensity in the same regions of both the
fluorescein and rhodamine images. The ratio of the rhodamine to fluorescein intensity was then determined and compared.
RESULTS
Morphology of A498 cells and formation of noncentrosomal microtubules
The morphology of A498 cells is shown in Fig. 1. Many cells
have an unusual morphology in the form of phase dense
regions in the peripheral lamellae (Fig. 1A), which are
transient, forming and dispersing over time. A498 cells have
what appears to be a conventional MTOC, as determined by
immunofluorescence staining for pericentrin (data not shown),
and a normal, though dense, radial array of microtubules. We
have observed non-centrosomal microtubules of varying
lengths both in the periphery of living cells which had been
microinjected with fluorescently labeled tubulin (Fig. 1C-E)
and in uninjected, glutaraldehyde-fixed cells by indirect
immunofluorescence (Fig. 1B), thus eliminating the possibility that their formation was an artifact of tubulin microinjection. In both preparations, a rim of soluble tubulin fluorescence
was frequently observed at the cell periphery (Fig. 1E). To
determine whether this was caused simply by an increase in
cell thickness in that region, or by some specific compartmentalization phenomenon, fluorescein-labeled dextran was coinjected with rhodamine-labeled tubulin. The dextran diffuses
throughout the cytosol, providing a marker for cytoplasmic
volume. The ratios between the intensities of dextran and
tubulin in the thick lamella and in the area proximal to it were
compared. Since the ratio was not statistically different
between the two regions, the increased lamellar brightness
must be due to increased volume in that area (data not shown).
Non-centrosomal microtubules are often observed in these
peripheral regions.
We have observed the formation of non-centrosomal microtubules in living cells injected with rhodamine-labeled tubulin.
Forming non-centrosomal microtubules (Fig. 2) usually appear
as bright spots, and elongation occurs unidirectionally from
that point. It is interesting to note that non-centrosomal microtubule disassembly (Fig. 2A) occurs in that manner also, from
one end back toward the other, rather than from both ends
toward the center. The possibility that a segment of a longer
microtubule coursing in and out of the plane of focus was
mistaken for non-centrosomal microtubule formation was
eliminated by through-focusing in the region where de novo
microtubule formation was observed. Because some cells
contain stable populations of microtubules that turn over very
slowly (Webster and Borisy, 1989), it was necessary to confirm
that these are truly non-centrosomal microtubules, and not just
fluorescent segments assembled onto the ends of microtubules
that had not incorporated rhodamine-tubulin. In order to do
this, live cells were microinjected with rhodamine-labeled
tubulin and allowed to recover for at least 2.5 hours before an
Fig. 1. Morphology of A498 cells. (A) Phase-contrast light microscopy of live cells; many cells have phase-dense peripheral lamella.
(B) Immunofluorescence observation of glutaraldehyde-fixed cells stained with an antibody to α-tubulin. (C-E) Low light level fluorescence
microscopy of live cells injected with rhodamine tubulin. In many cells, the microtubule array does not extend all the way to the cell edge. Noncentrosomal microtubules of varying lengths can be observed at the cell periphery of both fixed and living cells (B-E). The thicker edge in (E) is
brighter due to an increased amount of soluble fluorescent tubulin. Bars: 50 µm (A); 20 µm (B); 2 µm (C-E).
2394 A. C. Yvon and P. Wadsworth
Fig. 2. Formation (A and B) and disassembly (A) of free microtubules. Images were collected at 2 second intervals; elapsed time, in seconds, is
indicated in the corner of each panel. (A) Arrowheads indicate a microtubule that completely disassembles; the asterisks indicate a forming
microtubule. Both are placed at the inactive end of the microtubules, at the same pixel location in each panel. (B) Two microtubules form, as
indicated by the arrowheads and asterisks, which are located at the same pixel position in each panel. Another non-centrosomal microtubule is
observed at the cell’s edge, and is evidence that the same plane of focus is maintained in this thin lamella throughout the course of data
collection. Bar, 2 µm.
image of a non-centrosomal microtubule was collected.
Following fixation and immunofluorescence staining for
tubulin, the same microtubule was located and imaged again.
The results are shown in Fig. 3. In no case was a ‘free’ microtubule in a live cell proven, by immunocytochemistry, to be
part of a longer microtubule (n=6).
Non-centrosomal microtubule formation was relatively
infrequent: in 135 minutes of observation, 19 instances were
observed, a frequency of one formation per 7.1 minutes. Interestingly, complete disassembly of non-centrosomal microtubules was twice as rare as their formation. It occurred only
9 times, a frequency of one disassembly per 15 minutes of
observation. So once formed, non-centrosomal microtubules
are persistent; we have seen them persist for as long as 17
minutes (data not shown). This is not surprising, in view of the
fact that they are present in A498 cells more often than not,
but we rarely observe their formation. During observation of
this cell line for this and other studies, formation of non-centrosomal microtubules by severing of longer microtubules has
been observed only once. Therefore, it is unlikely that this is
the primary mechanism for their formation. Finally, non-centrosomal microtubules could arise by transport, to the cell
periphery, of microtubules which were nucleated at and
released from the MTOC. We did not observe either centrifugal or centripetal motion of non-centrosomal microtubules in
these experiments, making transport unlikely as the mode by
which these microtubules arise.
We estimate that these short non-centrosomal microtubules
exist in 65% of A498 cells at any given time. This value is
likely an underestimate because we cannot determine if noncentrosomal microtubules are present in the central regions of
the cell, where the microtubule density is high. We have also
examined other cell lines for the presence of non-centrosomal
microtubules using preparations of glutaraldehyde-fixed cells
Fig. 3. Visualization of the same non-centrosomal microtubule by
live fluorescence analog cytochemistry (A) and
immunocytochemistry (B). This microtubule was imaged following
microinjection of fluorescently labeled tubulin, fixed 1.5 minutes
after collection of the fluorescent image, and imaged again following
immunofluorescence staining with an antibody to α-tubulin.
Arrowheads denote both free ends in each panel. The asterisk in B
indicates a non-centrosomal microtubule which formed in the time
between live cell imaging and fixation. Bar, 2 µm.
Non-centrosomal microtubule dynamics 2395
stained with anti-tubulin antibodies and living cells injected
with rhodamine-labeled tubulin. Non-centrosomal microtubules have been observed in A431, BSC-1, CaOV-3, CHO
and PtK1 cells at varying frequencies, but in all cases, less
commonly than in A498 cells (data not shown). The occurrence
and certainly the observation of non-centrosomal microtubules
may be related to cell morphology. They are more frequently
observed in cells characterized by a spread, flattened periphery
containing a sparse microtubule array.
Dynamics of non-centrosomal microtubules in vivo
Since both ends of non-centrosomal microtubules were accessible, we were able to study dynamic behavior at the plus and
Fig. 4. Dynamic behavior of individual noncentrosomal microtubules. Images were collected at
2 second intervals; elapsed time, in seconds, is
indicated in the corner of each panel. The asterisks
were placed at the same pixel location in each panel
to show the stability of one end of the noncentrosomal microtubule. The other end undergoes
dynamic changes: in both A and B, it grows between
the first and second panels, then shortens between
the second and third panels, and grows again
between the third and last panels. Partial closure of
the field diaphragm is responsible for the dark areas
on the left side of the panels in B. Bar, 2 µm.
minus ends of the same microtubule, which allowed a distinction to be made between the two so that the populations could
be compared. The position of each end of non-centrosomal
microtubules was tracked using a mouse-driven cursor, and
plotted versus time (Figs 4, 5). In 75% of the microtubules
analyzed, one end had no growth or shortening events at all.
(An event is defined as an excursion greater than or equal to
0.5 µm.) In the remainder of the cases, one end had significantly fewer transitions between growth and shortening or vice
versa, than the other. For each microtubule, the ratio of the transition frequencies was used to differentiate the active end from
the inactive end. We also determined the total growth at each
end of the non-centrosomal microtubules: in each case, the end
2396 A. C. Yvon and P. Wadsworth
Fig. 6. Comparison of the dynamic behavior of the plus end of
centrosomal microtubules (A) with the active end of non-centrosomal
microtubules (B). The magnitude of events and number of transitions
is similar between the two groups. The plots are offset on the y-axis
for clarity.
Fig. 5. Life history plots of both ends of individual non-centrosomal
microtubules. The graphs represent the excursions made by each end
of individual microtubules during a 76 second observation period.
One end (represented by squares) is consistently more dynamic, as
measured by the frequency of transitions, than the other end
(represented by solid diamonds). The microtubule in A underwent 9
transitions at its active end and 0 transitions at its inactive end, that in
B underwent 11 at its active end and 6 at its inactive end, and that in
C underwent 7 at its active end and 0 at its inactive end. The plots are
offset on the y-axis for clarity.
that showed greater total growth was also the active end, as
defined above. Over a 33-minute period of measurement, the
average growth per microtubule at the active end was 1.5 µm,
while that at the inactive end was 0.2 µm. Thus, on the basis
of transition frequencies and net growth, we designated the
active ends the plus ends, and the inactive ends the minus ends
(Bergen and Borisy, 1980; Horio and Hotani, 1986). It should
be noted that, in some cases, our limit of resolution may
preclude distinction between minus end growth or shortening
and slight translocation of the microtubule as a whole.
Dynamics of the plus ends of radially extending microtubules,
which were most likely centrosomally nucleated, were also
analyzed in this cell line and compared with the dynamics of
the plus ends of non-centrosomal microtubules. The two sets
of data correspond closely, reinforcing the correctness of our
original classification of the plus and minus ends (Fig. 6 and
Table 1). It should be noted that in most cases, the non-centrosomal microtubules used for dynamic analysis were present
at the start of image acquisition, and their formation was not
observed.
Fig. 5 is a representation of the types of behavior observed
at both ends of non-centrosomal microtubules. In many cases,
the plus end was very active, while the minus end was completely inactive (Fig. 5A). Some minus ends did undergo events
(Fig. 5B), but could still be distinguished from the plus end
based on the frequency of transitions. Occasionally, a plus end
displayed relatively little dynamic behavior, however, its minus
end experienced no transitions at all (Fig. 5C). The low number
of events observed at one end of each non-centrosomal microtubule is highlighted in Fig. 7. The total number of growth
events at the plus ends was 52, compared with 7 at the minus
ends; there were only 9 shortening events at the minus ends,
as opposed to 52 at the plus ends.
As Table 1 indicates, dynamic analysis revealed striking differences between the plus and minus ends in several parameters. The pause duration was nearly 4-fold higher at the minus
ends; and consequently, the percentage of time spent in pause
Non-centrosomal microtubule dynamics 2397
Table 1. Microtubule dynamic instability parameters in
A498 cells*
Centrosomal
plus ends
Noncentrosomal
plus ends
Noncentrosomal
minus ends
Pause duration (seconds)
% Time shortening
% Time growing
% Time pausing
17.3
17.3
20.2
62.5
11.2
21.6
20.0
58.4
44.0
3.30
2.30
94.4
Average events/MT
Dynamicity (dimers/second)
Rescue frequency (/second)
Catastrophe frequency (/second)
6.33
78.8
0.112
0.024
6.43
76.2
0.114
0.031
1.11
9.50
0.136
0.0047
43
15
28
15
28
15
# of MTs analyzed
# of cells analyzed
*Standard deviations are not reported because these values were calculated
for a population of microtubules, rather than on an individual basis.
increased to 94% at the minus ends, compared to 58% at the
plus ends. Also, the dynamicity was decreased by a factor of
8 at the minus ends; and, the average number of events per
microtubule was reduced from 6.4 at the plus ends to 1.1 at the
minus ends. These parameters reflect the very stable nature of
the minus ends. Perhaps the most intriguing result was the
nearly 7-fold decrease in catastrophe frequency at the minus
ends, compared with the plus ends. Although greatly suppressed at the minus ends, when a catastrophe did occur, it was
rescued with essentially the same frequency as at the plus ends.
It should be noted that the rescue frequency at the minus ends
was calculated on the basis of a small number of events since
catastrophe was so rare.
Predictably, there were no statistically significant differences in the rates of growth and shortening between the centrosomal and non-centrosomal microtubule plus ends in A498
cells (Table 2 and Fig. 6). The non-centrosomal microtubule
plus and minus ends also showed no significant rate differences
(Table 2 and Fig. 7), a result corroborated by previous in vitro
studies which showed notable overlap in the growth and shortening rate distributions of plus and minus ends (Kowalski and
Williams, 1993b). Comparison of the growth and shortening
rates of A498 microtubules with those of other cell lines (Table
2) highlights the variability in the parameters of microtubule
dynamic instability from one cell type to another. The microtubule growth rate in A498 cells falls between that in untreated
BSC-1 cells and that in MAP2-injected BSC-1 cells, and is
slower than in PtK1 or CHO cells. The rate of shortening
Table 2. Microtubule growth and shortening rates
Growth rate
(µm/minute)
Shortening
rate (µm/minute)
A498 centrosomal plus ends
A498 non-centrosomal plus ends
A498 non-centrosomal minus ends
6.17 (±3.5)
6.90 (±2.5)
6.85 (±2.7)
−6.56 (±5.2)
−8.40 (±5.5)
−6.56 (±3.2)
PtK1 plus ends*
CHO plus ends*
BSC-1 plus ends†
BSC-1 + 1.0 mg/ml MAP2†
11.9 (±6.5)
19.7 (±8.1)
7.6 (±3.6)
5.2 (±3.0)
−19.8 (±10.8)
−32.2 (±17.7)
−15.0 (±12.5)
−10.3 (±8.9)
*From Shelden and Wadsworth (1993).
†From Dhamodharan and Wadsworth (1995).
Fig. 7. Histograms of growth and shortening rates at both ends of
non-centrosomal microtubules. Although the distribution of rates is
similar at both ends of the microtubules, significantly fewer events
occurred at the minus ends.
measured in A498 cells is slower than in any of the other cell
types, even those microinjected with MAP2.
DISCUSSION
Our data clearly demonstrate that non-centrosomal microtubules can form in the peripheral lamella of A498 cells. Once
formed, these microtubules persist, allowing the dynamic
behavior of each end of the microtubule to be analyzed. We
find that the minus end of these microtubules is extraordinarily stable: in the majority of cases, it did not grow or shorten
at all; the catastrophe frequency was remarkably low; and,
every catastrophe we observed was rescued.
Comparison of minus end dynamic Instability in
vitro and in vivo
The dynamic instability behavior of microtubule plus and
minus ends has been measured previously in vitro, using
purified brain tubulin (Panda et al., 1996; Vasquez et al., 1994;
Walker et al., 1988). The data reveal differences in the dynamic
behavior at the two microtubule ends: at the minus end, the
association and dissociation rate constants are lower and the
rescue frequency is higher than at the plus end, at all concentrations of tubulin examined. The frequency of catastrophe is
greater at the plus end at low tubulin concentrations, but this
difference is negligible at higher tubulin levels (Walker et al.,
1988). Because these experiments are performed using purified
tubulin, the differences between the plus and minus ends
cannot be attributed to regulatory molecules in the preparation.
Rather, these differences must result from the orientation of
tubulin dimers in the microtubule lattice (Mandelkow and
Mandelkow, 1989). Tubulin dimers contain one non-exchangeable GTP on the α subunit, and an exchangeable GTP, which
is hydrolyzed subsequent to incorporation into the microtubule
lattice, on the β subunit (reviewed by Caplow, 1992). Recent
2398 A. C. Yvon and P. Wadsworth
immunofluorescence studies, using an antibody to the Nterminus of α-tubulin (Fan et al., 1996), confirm previous
results showing that β-tubulin, with its exchangeable GTP, is
exposed at the plus end of the microtubule (Mitchison, 1993).
Given this orientation of the dimer within the microtubule
lattice, the enhanced stability of minus ends, relative to plus
ends, could be the result of having a non-hydrolyzable GTP in
the terminal position. The hypothesis that differences in the
structure of the two microtubule ends contribute to differences
in their behavior is supported by experiments in which new
minus and plus ends are created by severing microtubules with
a u.v. microbeam. Newly created minus ends are initially
stable, and then grow slowly; newly created plus ends, which
lack a GTP cap, rapidly disassemble (Walker et al., 1989).
The behavior of the minus ends we have measured in vivo
can be compared with previous observations of minus end
behavior in vitro. In A498 cells, the catastrophe frequency is
6-fold higher at the plus end than at the minus end, a difference greater than that measured using purified tubulin in vitro
(Panda et al., 1996; Vasquez et al., 1994; Walker et al., 1988).
Interestingly, the catastrophe frequency of the minus ends in
vivo, while greatly suppressed relative to that of the plus ends,
is actually greater than that reported for minus ends in several
in vitro studies (Panda et al., 1996; Vasquez et al., 1994; Walker
et al., 1988). Regardless of the exact values, the minus ends in
this study are highly stable, especially when compared with the
plus ends. This stability probably results from the inherent
structural differences, as discussed above, which could directly
modulate the kinetic behavior of the two ends. Differences in
the structure of the two ends of the microtubule could also
affect microtubule behavior indirectly, by modulating the interactions of regulatory molecules with microtubule ends. In
support of this view, several molecules are now known to differentially affect the two microtubule ends: XMAP stimulates
microtubule turnover by increasing the association and dissociation rate constants at the plus end (Vasquez et al., 1994),
XKCM1 stimulates the frequency of catastrophe at the plus end
(Walczak et al., 1996), and Kar3 stimulates disassembly at the
minus end (Endow et al., 1994; Saunders et al., 1997). Recent
studies also demonstrate that the drug vinblastine differentially
affects the two ends, stabilizing the plus end while increasing
turnover at the minus end (Panda et al., 1996). Cellular
molecules that stimulate catastrophe transitions at the plus end
could contribute to the differences that we observe in vivo in
the behavior of the plus and minus ends. Such a scenario would
be consistent with the observation that the rescue frequency is
comparable at the two microtubule ends, while the catastrophe
frequency is different.
Although minus end events were uncommon, when non-centrosomal microtubules underwent events, those events were
statistically indistinguishable in rate, duration and distance
(data not shown) at the plus and minus ends. The release or
destabilization of a minus end dimer, possibly caused by a
Kar3-like factor or loss of a cap, would result in dynamics
which would then presumably be regulated by the MAPs
bound along the microtubule wall, the same molecules which
influence plus end dynamics.
Comparison of minus end behavior with previous in
vivo observations
Historically, the behavior of minus ends has been difficult to
study in vivo. Typically, a high density of minus ends is associated with the MTOC, making imaging of individual minus
ends in living cells extremely difficult. Electron microscopic
analysis has shown that minus ends are sometimes found at
sites distant from the MTOC: in neuronal cells, minus ends are
found along the length of the axon (Yu et al., 1993) and in
mammalian spindles, the minus ends of interpolar microtubules are located throughout the spindle region (Mastronarde
et al., 1993). There have also been reports in cultured cells of
microtubule minus ends not associated with the centrosome in
the interphase array (Bre et al., 1987). In all of these cases, the
density of microtubules precludes the analysis of individual
minus end dynamic behavior in vivo. Recent work in cultured
mammalian cells demonstrated the release of microtubule
minus ends from the MTOC, and their subsequent behavior in
that context (Keating, 1997).
One approach to analyzing microtubule minus ends in vivo
has been to create new ends using u.v. irradiation or mechanical cutting (Nicklas et al., 1989; Spurck et al., 1990; Tao et
al., 1988; Wilson and Forer, 1988). In most cases, the newly
created minus ends are stable, neither growing nor shortening
detectably. However, in the case of crane fly spermatocyte
meiotic spindles, poleward motion of the irradiated region has
been observed; this could result from minus end growth
(Wilson and Forer, 1988). One explanation for the behavior of
minus ends in this and other studies is that, due to the structure
of the minus end, the transition to rapid shortening is kinetically unfavorable; therefore, these ends either grow slowly or
remain in an attenuated state, during which no growth or shortening is detectable (Toso et al., 1993). Alternatively, new minus
ends might be rapidly and efficiently capped by molecules that
reduce the likelihood of catastrophe.
Additional information regarding the dynamic behavior of
minus ends in living cells comes from analysis of flux and treadmilling of microtubules (Mitchison, 1989; Rodionov and Borisy,
1997; Sawin and Mitchison, 1991, 1994; Keating, 1997). Fluorescent marks, created on spindle microtubules by photoactivation of caged tubulin, are observed to move poleward at a rate
of about 0.5 um/minute; this concurrent plus end subunit
assembly, minus end subunit disassembly, and translocation
poleward of the entire microtubule lattice has been termed
poleward flux (Mitchison, 1989). In fragments of interphase cells
lacking centrosomes, microtubules with two free ends are
observed; photobleaching experiments demonstrate that the
apparent cytoplasmic translocation of these microtubules results
from the coordinated gain and loss of subunits at the two ends
(Rodionov and Borisy, 1997). Experiments in which a laserbleached spot was created on microtubules released from the
MTOC revealed stable or shortening minus ends, in conjunction
with microtubule translocation (Keating, 1997). In all of these
cases, net loss of subunits occurs at the minus end. Specific regulatory components, for example Kar3-like factors, might be
required to facilitate subunit turnover at minus ends (Saunders
et al., 1997). Alternatively, minus end capping factors could be
released, resulting in minus end turnover. We have observed no
translocation of the non-centrosomal microtubules through the
cytoplasm, as might be expected for treadmilling, and the
dynamic behavior of the inactive end was not coordinated with
the opposite, active end. Thus, if treadmilling is modulated by
specific non-tubulin factors, either these are not active in A498
cells, or the ends are inaccessible to them.
Non-centrosomal microtubule dynamics 2399
Stability of microtubules in A498 cells and
persistence of non-centrosomal microtubules
The data reported here show that the plus ends of both noncentrosomal and centrosomal microtubules in A498 cells are
considerably more stable than the plus ends of microtubules in
other cell lines. For example, the rescue frequency is 58%
higher and the catastrophe frequency is 56% lower at the plus
ends in A498 cells than in a non-cancerous kidney cell line,
BSC-1 (Dhamodharan and Wadsworth, 1995). In addition, the
elongation rate of microtubules in A498 cells is slower than in
any other cell line studied, with the exception of MAP2injected BSC-1 cells, and the shortening rate is the slowest
overall (Table 2). One possible explanation for the remarkable
stability of microtubules in A498 cells is that high levels of
stabilizing MAPs contribute to the low level of microtubule
dynamics in these cells. Neuronal MAPs, MAP-2 and tau,
suppress microtubule dynamic instability by promoting rescue
and decreasing catastrophe transitions (Drechsel et al., 1992;
Kowalski and Williams, 1993a; Pryer et al., 1992). Recent in
vitro experiments show that the major non-neuronal MAP,
MAP-4, increases rescue transitions (Ookata et al., 1995).
Alternatively, it has been shown that tubulin isotype composition plays a role in microtubule dynamics (Panda et al.,
1994). Perhaps A498 cells express an isotype of α or β tubulin
which confers unusual stability on their microtubules. Regardless of its cause, the low level of microtubule dynamic behavior
in these cells almost certainly contributes to the persistence
of non-centrosomal microtubules. For example, following a
catastrophe, the disassembly rate is slow and the rescue
frequency is comparatively high, so more often than not, the
microtubule is only partially disassembled.
Origin of non-centrosomal microtubules
A compelling question to emerge from this study is how and why
do non-centrosomal microtubules form in A498 cells? It seems
likely that their formation is the result of an alteration in some
aspect(s) of the usual regulation of microtubule nucleation and/or
dynamics. One possible explanation of non-centrosomal microtubule formation in A498 cells is that centrosomal components
with nucleating activity exist outside of the MTOC. Exciting new
work has identified a microtubule nucleation complex, termed γtubulin ring complex (γTuRC), which contains γ-tubulin, α and
β tubulin, and at least four other proteins (Moritz et al., 1995;
Zheng et al., 1995). An intriguing hypothesis is that A498 cells
have active γTuRCs which are dissociated from the centrosome
and could act to nucleate non-centrosomal microtubules. Two
recent pieces of work support this hypothesis. First, Shu et al.
(1995) report a transient association of γ-tubulin with the minus
ends of midbody microtubules, and suggest that γ-tubulin acts to
nucleate microtubules at sites distant from the centrosome.
Second, in extracts of two human cell lines, most of the γ-tubulin
exists in the cytosolic, rather than centrosomal, fractions
(Moudjou et al., 1996; Stearns et al., 1991; Zheng et al., 1991).
However, much of this soluble pool is associated with the
chaperone Tcp1, and there is no indication that these γ-tubulincontaining particles are actually functional nucleators (Moudjou
et al., 1996). Finally, it should be noted that a nucleating complex,
such as γTuRC, could potentially act as a minus end cap which,
when associated with the microtubule end, could act to suppress
addition or loss of subunits.
An alternative explanation for the formation of non-centrosomal microtubules is that tubulin or assembly-promoting
MAPs are over-expressed in these cells, decreasing the critical
concentration for spontaneous microtubule assembly (Murphy
et al., 1977). If this were the case, centrosomal polymerization
would still be favored, due to the presence of nucleation sites
in the pericentriolar material, but non-centrosomal polymerization could occur. Non-centrosomal microtubule formation
has been previously observed in cases when the tubulin concentration is above that which is critical for spontaneous polymerization: following release from drug induced disassembly
(Bre et al., 1987), and in cytoplasts lacking an MTOC (Karsenti
et al., 1984). Additional data on the soluble tubulin concentration, the ratio of unpolymerized to polymerized tubulin, and/or
the concentration of MAPs in these cells is necessary to draw
any conclusions in this regard. Finally, the possible lack of
certain regulatory molecules could also contribute to spontaneous microtubule assembly. Recent work in sea urchin eggs
has identified an activity which blocks assembly at microtubule
minus ends in vitro (Spittle and Cassimeris, 1996). The authors
hypothesize that a tubulin dimer-binding protein is present in
vivo and functions to prevent spontaneous polymerization. If
cells were deficient in such a factor, non-centrosomal microtubules could form freely; this might be the case with A498
cells.
The extremely slow dynamic turnover of microtubules in
A498 cells suggests another interpretation of why they are
found in these cells. In this scenario, the inherent stability of
microtubules in these cells might contribute to spontaneous
formation of non-centrosomal microtubules. This is consistent
with the observations that taxol, which suppresses the
dynamics of microtubules, induces the formation of non-centrosomal microtubules in cells (De Brabander et al., 1981).
Once formed, the low level of dynamic turnover would
increase the probability that they persist. In fact, non-centrosomal microtubules appear to arise, at some frequency, in
diverse cells; in most cases however, dynamic instability
probably ensures their rapid disappearance, resulting in less
persistent and fewer detectable non-centrosomal microtubules.
Thus, the parameters of dynamic instability in A498 cells
might increase both the probability that a stable nucleus of a
microtubule might form and that it might persist in the
cytoplasm.
Regardless of the mechanism by which the non-centrosomal
microtubules form, their formation does not induce a
detectable disruption of the nucleated array and both the noncentrosomal and centrosomal microtubules co-exist in the same
cytoplasm. Our data, and recent observations of microtubule
treadmilling in cytoplasmic fragments which lack an MTOC
(Rodionov and Borisy, 1997), strongly suggest that the type of
microtubule dynamic turnover which predominates in a given
array can be modulated by alterations in the quantity or activity
of nucleating factors, of tubulin, and of regulatory factors such
as MAPs.
Conclusion
In conclusion, we have observed and quantitated the dynamic
behavior of microtubule minus ends in vivo: the data directly
demonstrate that individual minus ends in vivo are remarkably
stable and suggest that their stability is intrinsic to their
structure, possibly due to a ubiquitous GTP cap. Further, we
2400 A. C. Yvon and P. Wadsworth
have observed the de novo formation of microtubules at sites
distant from the centrosome. This work has served to challenge
the traditional doctrine that microtubules are nucleated solely
at MTOCs and offers an example of a case in which that is not
so.
We thank Dr S. Doxsey (University of Massachusetts Medical
School, Worcester, MA) for providing an antibody to pericentrin, Dr
T. Rahmani for technical assistance, and George Drake for computer
programming and support. This work was supported by NSF (MCB
9204646).
REFERENCES
Baas, P. W. (1997). Microtubules and axonal growth. Curr. Opin. Cell Biol. 9,
29-36.
Bacallao, R., Antony, C., Dotti, C., Karsenti, E., Stelzer, E. H. K. and
Simons, K. (1989). The subcellular organization of madin-darby canine
kidney cells during the formation of a polarized epithelium. J. Cell Biol. 109,
2817-2832.
Belmont, L. D., Hyman, A. A., Sawin, K. E. and Mitchison, T. J. (1990).
Real-time visualization of cell cycle-dependent changes in microtubule
dynamics in cytoplasmic extracts. Cell 62, 579-589.
Bergen, L. G. and Borisy, G. G. (1980). Head-to-tail polymerization of
microtubules in vitro. J. Cell Biol. 84, 141-150.
Bre, M.-H., Kreis, T. E. and Karsenti, E. (1987). Control of microtubule
nucleation and stability in madin-darby canine kidney cells: the occurrence
of noncentrosomal, stable detyrosinated microtubules. J. Cell Biol. 105,
1283-1296.
Caplow, M. (1992). Microtubule dynamics. Curr. Opin. Cell Biol. 4, 58-65.
Cassimeris, L., Pryer, N. K. and Salmon, E. D. (1988). Real-time
observations of microtubule dynamic instability in living cells. J. Cell Biol.
107, 2223-2231.
Cohen, W. D. (1991). The cytoskeletal system of nucleated erythrocytes. Int.
Rev. Cytol. 130, 37-84.
De Brabander, M., Geuens, G., Nuydens, R., Willebrords, R. and De Mey,
J. (1981). Taxol induces the assembly of free microtubules in living cells and
blocks the organizing capacity of the centrosomes and kinetochores. Proc.
Nat. Acad. Sci. USA 78, 5608-5612.
Dhamodharan, R. and Wadsworth, P. (1995). Modulation of microtubule
dynamic instability in vivo by brain microtubule associated proteins. J. Cell
Sci. 108, 1679-1689.
Drechsel, D. N., Hyman, A. A., Cobb, M. H. and Kirschner, M. W. (1992).
Modulation of the dynamic instability of tubulin assembly by the
microtubule-associated protein tau. Mol. Biol. Cell 3, 1141-1154.
Endow, S. A., Kang, S. J., Satterwhite, L. L., Rose, M. D., Skeen, V. P. and
Salmon, E. D. (1994). Yeast Kar3 is a minus-end microtubule motor protein
that destabilizes microtubules preferentially at the minus ends. EMBO J. 13,
2708-2713.
Erickson, H. P. and O’Brien, E. T. (1992). Microtubule dynamic instability
and GTP hydrolysis. Annu. Rev. Biophys. Biomol. Struct. 21, 145-166.
Fan, J., Griffiths, A. D., Lockhart, A., Cross, R. A. and Amos, L. A. (1996).
Microtubule minus ends can be labelled with a phage display antibody
specific to alpha-tubulin. J. Mol. Biol. 259, 325-330.
Horio, T. and Hotani, H. (1986). Visualization of the dynamic instability of
individual microtubules by dark field microscopy. Nature 321, 605-607.
Karsenti, E., Kobayashi, S., Mitchison, T. and Kirschner, M. (1984). Role of
the centrosome in organizing the interphase microtubule array: properties of
cytoplasts containing or lacking centrosomes. J. Cell Biol. 98, 1763-1776.
Keating, T. J., Peloquin, J. G., Rodionov, V. I., Momcilovic, D. and Borisy,
G. G. (1997). Microtubule release from the centrosome. Proc. Nat. Acad. Sci.
USA 94, 5078-5083.
Kowalski, R. J. and Williams, R. C. Jr (1993a). Map-2 alters the dynamic
properties of microtubule assembly and disassembly. J. Biol. Chem. 268,
9847-9855.
Kowalski, R. J. and Williams, R. C. Jr (1993b). Unambiguous classification
of microtubule-ends in vitro: dynamic properties of the plus- and minusends. Cell Motil. Cytoskel. 26, 282-290.
Lambert, A.-M. (1993). Microtubule-organizing centers in higher plants. Curr.
Opin. Cell Biol. 5, 116-122.
Mandelkow, E. and Mandelkow, E.-M. (1989). Tubulin, microtubules and
oligomers: molecular structure and implications for assembly. In Cell
Movement, vol. 2 (ed. F. D. Warner and J. R. McIntosh), pp. 23-45. Alan R.
Liss, New York.
Margolis, R. L. and Wilson, L. (1981). Microtubule treadmills – possible
molecular machinery. Nature 293, 705-711.
Mastronarde, D. N., McDonald, K. L., Ding, R. and McIntosh, J. R. (1993).
Interpolar spindle microtubules in PTK cells. J. Cell Biol. 123, 1475-1489.
Mazia, D. (1987). The chromosome cycle and the centrosome cycle in the
mitotic cycle. Int. Rev. Cytol. 100, 49-92.
Mitchison, T. J. and Kirschner, M. W. (1984a). Dynamic instability of
microtubule growth. Nature 312, 237-242.
Mitchison, T. J. and Kirschner, M. W. (1984b). Microtubule assembly
nucleated by isolated centrosomes. Nature 312, 232-237.
Mitchison, T. J. (1989). Polewards microtubule flux in the mitotic spindle:
evidence from photoactivation of fluorescence. J. Cell Biol. 109, 637-652.
Mitchison, T. J. and Sawin, K. E. (1990). Tubulin flux in the mitotic spindle:
where does it come from, where is it going? Cell Motil. Cytoskel. 16, 93-98.
Mitchison, T. J. and Salmon, E. D. (1992). Poleward kinetochore fiber
movement occurs during both metaphase and anaphase-A in newt lung cell
mitosis. J. Cell Biol. 119, 569-582.
Mitchison, T. J. (1993). Localization of an exchangeable GTP-binding site at
the plus ends of microtubules. Science 261, 1044-1047.
Moritz, M., Braunfeld, M. B., Sedat, J. W., Alberts, B. and Agard, D. A.
(1995). Microtubule nucleation by γ-tubulin-containing rings in the
centrosome. Nature 378, 638-640.
Moudjou, M., Bordes, N., Paintrand, M. and Bornens, M. (1996). γ-Tubulin
in mammalian cells: the centrosomal and the cytosolic forms. J. Cell Sci. 109,
875-887.
Murphy, D. B., Johnson, K. A. and Borisy, G. G. (1977). Role of tubulin
associated proteins in microtubule nucleation and elongation. J. Mol. Biol.
117, 33-52.
Nicklas, R. B., Lee, G. M., Rieder, C. L. and Rupp, G. (1989). Mechanically
cut mitotic spindles: clean cuts and stable microtubules. J. Cell Sci. 94, 415423.
Ookata, K., Hisanaga, S.-i., Bulinski, J. C., Murofushi, H., Aizawa, H.,
Itoh, T. J., Hotani, H., Okumura, E., Tachibana, K. and Kishimoto, T.
(1995). Cyclin B interaction with microtubule-associated protein 4 (MAP4)
targets p34cdc2 kinase to microtubules and is a potential regulator of M-phase
microtubule dynamics. J. Cell Biol. 128, 849-862.
Panda, D., Miller, H. P., Banerjee, A., Luduena, R. F. and Wilson, L. (1994).
Microtubule dynamics in vitro are regulated by the tubulin isotype
composition. Proc. Nat. Acad. Sci. USA 91, 11358-11362.
Panda, D., Jordan, M. A., Chu, K. C. and Wilson, L. (1996). Differential
effects of vinblastine on polymerization and dynamics at opposite
microtubule ends. J. Biol. Chem. 271, 29807-29812.
Pryer, N. K., Walker, R. A., Skeen, V. P., Bourns, B. D., Soboeiro, M. F. and
Salmon, E. D. (1992). Brain microtubule-associated proteins modulate
microtubule dynamic instability in vitro: Real time observations using video
microscopy. J. Cell Sci. 103, 965-976.
Rodionov, V. I. and Borisy, G. G. (1997). Microtubule treadmilling in vivo.
Science 275, 215-218.
Sammak, P. J. and Borisy, G. G. (1988). Direct observation of microtubule
dynamics in living cells. Nature 332, 724-726.
Saunders, W., Hornack, D., Lengyel, V. and Deng, C. (1997). The
Saccharomyces cerevisiae kinesin-related motor Kar3p acts at preanaphase
spindle poles to limit the number and length of cytoplasmic microtubules. J.
Cell Biol. 137, 417-431.
Sawin, K. E. and Mitchison, T. J. (1991). Poleward microtubule flux in mitotic
spindles assembled in vitro. J. Cell Biol. 112, 941-954.
Sawin, K. E. and Mitchison, T. J. (1994). Microtubule flux in mitosis is
independent of chromosomes, centrosomes, and antiparallel microtubules.
Mol. Biol. Cell 5, 217-226.
Schulze, E. and Kirschner, M. (1986). Microtubule dynamics in interphase
cells. J. Cell Biol. 102, 1020-1031.
Sharp, D. J., Yu, W. and Baas, P. W. (1995). Transport of dendritic
microtubules establishes their nonuniform polarity orientation. J. Cell Biol.
130, 93-103.
Shelden, E. and Wadsworth, P. (1993). Observation and quantification of
individual microtubule behavior in vivo: Microtubule dynamics are cell-type
specific. J. Cell Biol. 120, 935-945.
Shu, H.-B., Li, Z., Palacios, M. J., Li, Q. and Joshi, H. C. (1995). A transient
association of γ-tubulin at the midbody is required for the completion of
cytokinesis during the mammalian cell division. J. Cell Sci. 108, 2955-2962.
Sloboda, R. D., Dentler, W. L. and Rosenbaum, J. L. (1976). Microtubule
Non-centrosomal microtubule dynamics 2401
associated proteins and the stimulation of tubulin assembly in vitro.
Biochemistry 15, 4497-4505.
Spittle, C. S. and Cassimeris, L. (1996). Mechanisms blocking microtubule
minus end assembly: evidence for a tubulin dimer-binding protein. Cell
Motil. Cytoskel. 34, 324-335.
Spurck, T. P., Stonington, O. G., Snyder, J. A., Pickett-Heaps, J. D., Bajer,
A. and Mole-Bajer, J. (1990). UV microbeam irradiations of the mitotic
spindle. II. Spindle fiber dynamics and force production. J. Cell Biol. 111,
1505-1518.
Stearns, T., Evans, L. and Kirschner, M. W. (1991). γ Tubulin is a highly
conserved component of the centrosome. Cell 65, 825-836.
Tao, W., Walter, R. J. and Berns, M. W. (1988). Laser-transected
microtubules exhibit individuality of regrowth, however most free new ends
of the microtubules are stable. J. Cell Biol. 107, 1025-1035.
Tassin, A.-M., Maro, B. and Bornens, M. (1985). Fate of microtubuleorganizing centers during myogenesis in vitro. J. Cell Biol. 100, 35-46.
Toso, R. J., Jordan, M. A., Farrell, K. W., Matsumoto, B. and Wilson, L.
(1993). Kinetic stabilization of microtubule dynamic instability in vitro by
vinblastine. Biochemistry 32, 1285-1293.
Vasquez, R. J., Gard, D. L. and Cassimeris, L. (1994). XMAP from Xenopus
eggs promotes rapid plus end assembly of microtubules and rapid
microtubule polymer turnover. J. Cell Biol. 127, 985-993.
Walczak, C. E., Mitchison, T. J. and Desai, A. (1996). XKCM1: A Xenopus
kinesin-related protein that regulates microtubule dynamics during mitotic
spindle assembly. Cell 84, 37-47.
Walker, R. A., O’Brien, E. T., Pryer, N. K., Soboeiro, M. F., Voter, W. A.,
Erickson, H. P. and Salmon, E. D. (1988). Dynamic instability of individual
microtubules analyzed by video light microscopy: Rate constants and
transition frequencies. J. Cell Biol. 107, 1437-1448.
Walker, R. A., Inoue, S. and Salmon, E. D. (1989). Asymmetric behavior of
severed microtubule ends after ultraviolet-microbeam irradiation of
individual microtubules in vitro. J. Cell Biol. 108, 931-937.
Webster, D. R. and Borisy, G. G. (1989). Microtubules are acetylated in
domains that turn over slowly. J. Cell Sci. 92, 57-65.
Wilson, P. J. and Forer, A. (1988). Ultraviolet microbeam irradiation of
chromosomal spindle fibres shears microtubules and permits study of the
new free ends in vivo. J. Cell Sci. 91, 455-468.
Yu, W., Centonze, V. E., Ahmad, F. J. and Baas, P. W. (1993). Microtubule
nucleation and release from the neuronal centrosome. J. Cell Biol. 122, 349359.
Zheng, Y., Jung, M. K. and Oakley, B. R. (1991). γ-Tubulin is present in
Drosophila melanogaster and Homo sapiens and is associated with the
centrosome. Cell 65, 817-823.
Zheng, Y., Wong, M. L., Alberts, B. and Mitchison, T. (1995). Nucleation of
microtubule assembly by a γ-tubulin-containing ring complex. Nature 378,
578-583.
(Received 19 February 1997 – Accepted 18 July 1997)