Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://cellmotility.genebee.msu.ru/html/articles/vorobjev2002CBI.pdf Äàòà èçìåíåíèÿ: Mon Oct 7 16:08:04 2002 Äàòà èíäåêñèðîâàíèÿ: Mon Oct 1 20:13:56 2012 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: m 8

Cell Biology International 2002, Vol. 26, No. 9, 791­799 doi:10.1006/cbir.2002.0946, available online at http://www.idealibrary.com on

CENTROSOME-DEPENDENT ANISOTROPIC RANDOM WALK OF CYTOPLASMIC VESICLES
IVAN V. MALY* and IVAN A. VOROBJEV Laboratory of Cell Motility, A. N. Belozersky Institute of Physico-Chemical Biology, Moscow State University, Moscow 119899, Russia Received 27 September 2001; accepted 28 June 2002

We approach the problem of an apparently random movement of small cytoplasmic vesicles and its relationship to centrosome functioning. Motion of small vesicles in the cytoplasm of BSC-1 cells was quantified using computer-assisted microscopy. The vesicles move across the cytoplasm frequently changing their directions with negligible net displacement. The autocorrelation function for consecutive velocities of individual vesicles becomes indistinguishable from zero in 10 s. Variance in the displacement is proportional to time. The motion of vesicles is anisotropic: It has diffusivity along the radii drawn from the centrosome several times higher than the tangential diffusivity. This anisotropy is abolished by ultraviolet microbeam irradiation of the centrosome when the microtubule array loses radial structure. We conclude that the motion of the vesicles in the cytoplasm can be described as diffusion-like random walk with centrosomedependent anisotropy. The present analysis quantitatively corroborates the `trial and error' 2002 Elsevier Science Ltd. All rights reserved. model of vesicular transport.
KEYWORDS: vesicle movement; diffusion; microtubules; microirradiation.

INTRODUCTION In early studies of intracellular motility it was noted that motion of small vesicles differed from diffusion, or the Brownian motion, in that vesicles could displace at a speed impossible for a Brownian particle of their size in such a viscous medium as the cytoplasm (Rebhun, 1972). It was exactly this difference that led to the investigation of the mechanism of vesicle movement that was finally established to be the action of molecular motors that dragged the vesicles along microtubules or actin filaments (Goodson et al., 1997; Bray, 2001). However, recent studies have demonstrated that even movement of specialized transport vesicles, being the result of the active drag along microtubules and serving well its transport function, can nevertheless appear chaotic (Wacker et al., 1997; Rodionov et al., 1998; Suomalainen et al., 1999).
*To whom correspondence should be addressed; Present address: Department of Cell and Molecular Biology, Northwestern University Medical School, 303 E. Chicago Ave., Chicago, IL 60611, U.S.A. Tel.: (312) 503-2854; Fax: (312) 503-7912; E-mail: i-maly@ northwestern.edu 1065­6995/02/$-see front matter

The frequent change of direction in the saltatory movement of a vesicle is attributed to a limited processivity of molecular motors, association of different motors with the same particle, and availability of differently oriented cytoskeletal fibres for the motor reassociation (Wacker et al., 1997; Rogers et al., 1997) or existence of a specific switch mechanism (Welte et al., 1998; Gross et al., 2000). But is there a biologically meaningful regularity behind the apparently random movement of small vesicles? About 100 years ago Theodor Boveri characterized the centrosome as centre of the cell dynamics (see Wilson, 1925). Radial orientation of vesicle movement in regions close to the cell centre was noted early (Rebhun, 1967), however, even when a quantitative description of the vesicle movement was introduced, the related to the centrosome organization of the movement was only stated qualitatively (Freed and Lebowitz, 1970). In certain specific cell types such as melanophores the radial-symmetric structure of the cell and, concurrently, the preferred motility along cell radii are
2002 Elsevier Science Ltd. All rights reserved.


792

Cell Biology International, Vol. 26, No. 9, 2002

evident (Rogers et al., 1997), but in the more common cell architecture the vesicle movement seems apparently irregular (Wacker et al., 1997). We hypothesize that the regularity in the vesicle motion is merely masked by its random features, so quantitative analysis is necessary to reliably reveal it. For experimental testing the role of the centrosome in the organization of the vesicle movement ultraviolet microirradiation can be chosen as a tool of precise intervention into functioning of a living cell (Berns et al., 1977; Uzbekov et al., 1995; Khodyakov et al., 1997). We found cells of strain BSC-1 especially suitable for such experiments because the location of their centrosome can be clearly seen under phase contrast microscope (Maniotis and Schliwa, 1991; Hinchcliffe et al., 2001) without using dyes that could otherwise interfere with the experiment. A quantitative framework was developed that allowed to integrate the random and regular in the movement of vesicles and to reveal the role of the centrosome in its organization. We show that centrosome is involved in the establishment of the anisotropic (preferentially radial) pattern of the vesicle motility.

Planfluor 100, 1.30 NA objective. The intensity of incident light was reduced by using neutral density and orange filters. 16-bit deep, 0.068 m/ pixel images were acquired into memory of a PC computer with MicroMAX digital imaging system based on 1317 1035 cooled CCD camera (Princeton Instruments, Princeton, NJ, U.S.A.) under control of WinView32 software (Princeton Instruments, Princeton, NJ, U.S.A.). The images were acquired at 1-s intervals with typical series of 100 images. The temperature on the microscope stage (37 C) was maintained by the air curtain incubator (Nicolson Precision Instruments). For double immunostaining cells were fixed with glutaraldehyde, then permeabilized with Triton X-100, and processed as described elsewhere (Vorobjev et al., 1997). Primary antibodies were monoclonal mouse anti -tubulin (Amersham) and polyclonal rabbit anti -tubulin (Vorobjev et al., 2000). Secondary antibodies were conjugated with Oregon Green (Molecular Probes) and lysaminerhodamin (Jackson Inc.) respectively. Microirradiation The microirradiation apparatus used was described previously (Uzbekov et al., 1995). The light source was mercury arc lamp HBO100, 100 W (Osram). More than 95% of radiation power that passed band-pass interference filter (Carl Zeiss) was in the wavelength range 260­310 nm. The image of a diaphragm 0.16 mm in diameter that was placed in the focus of custom made quartz collector (NA = 0.45) was transferred into the region of the centrosome by quartz objective Ultrafluar 100/1.25 (Opton). The diameter of the irradiated area in the focal plane can be estimated as 2 m. Radiation power density in the focal plane was therefore 4 nJ m 2 s 1. Position of the beam in the field of view was determined by observing glass containing uranium oxide as an object and fixed in the cross hair on the ocular lens. For microirradiation of cells, the objective prewarmed to 37 C was immersed into the culture dish and focused through the culture medium. Exposure to the ultraviolet microbeam was 15 s, which corresponded to irradiation dose 180 nJ with the energy density 60 nJ m 2. Quantitative analysis of intracellular motility Position of vesicles in the digital frames of the time-lapse sequences was traced with ImagePC software (National Institutes of Health, Bethesda, MD, U.S.A.). In each cell 30 phase-contrast

MATERIALS AND METHODS Cell culture Cells of strain BSC-1 (American Type Culture Collection, Rockville, MD, U.S.A.) that descends from kidney epithelia of green monkey were grown in DMEM-F12 medium (Sigma Chemical, St Louis, MO, U.S.A.) supplemented with 10% foetal bovine serum (Hyclone Laboratories, Logan, UT, U.S.A.) under 5% CO2 in 37 C. The cells were plated onto coverslips attached with Sylgard silicone elastomer (Dow Corning, Midland, MI, U.S.A.) over a hole drilled in a 35 mm culture dish. Microscopy Experiments began on the third day after the passage when the culture reached confluence. We made an experimental wound in the monolayer of cells across the coverslip and allowed 4 h for recovery. By that time damaged cells shrank completely whereas cells from inside the monolayer extended large, flat lamellae into the wound. Movement of vesicles inside such lamellae was monitored by means of digital phase contrast videomicroscopy. Cells were observed on Nikon Eclipse TE-200 inverted microscope equipped with phase contrast


Cell Biology International, Vol. 26, No. 9, 2002

793

vesicles about 0.75 m in size were selected in the lamella. Position of the optical centroid of the image of each vesicle was then measured in every fifth frame. The 1-s intervals between frames allowed tracking the individual vesicles unambiguously between the points separated by 5 s. The spatial resolution of the system was estimated as approximately 0.2 m, or 3 pixels of the digital image. The vector of displacement between consecutive time points was decomposed into two orthogonal components. The radial component was the projection of the vector onto the radius drawn from the middle of the centrosphere through the middle point of the displacement. The radial component was taken positive when directed away from the centrosphere. The tangential component was the projection of the vector onto the normal to the radius in the middle point of the displacement. The tangential component was taken positive when directed clockwise around the centrosphere. To get the resultant radial displacement of a vesicle, the radial components of the constituent displacements measured as above were summed, and the resultant tangential displacement was obtained analogously. Further data analysis was performed with Mathcad software (Mathsoft, Inc., Cambridge, MA, U.S.A.). RESULTS For observations we selected cells at the edge of experimental wound with the extended lamellae and clearly visible centrosphere (Fig. 1). In the lamellae of BSC-1 cells numerous phase-contrast vesicles of approximately 0.75 m size (Fig. 1) were in continuous motion. The lamellae were flat, so that the vesicles never left the focus plane during their motion. Since the depth of focus of the setup used was about the size of the vesicles, their motion could be considered two-dimensional. By means of computer-assisted digital microscopy we measured position of an individual vesicle in consecutive video frames at 5 s intervals during 1 min. Intact cells Movement of the vesicles was quantified in 17 cells, 30 vesicles in each. Representative trajectories of vesicles are shown in Figure 2 together with their orientation with respect to the cell centre. The centre of the cell was defined to be in the middle of the phase-contrast centrosphere near the nucleus. Although preferential movement to and from the cell centre can be noted, the trajectories of the

Fig. 1. Small phase-contrast vesicles in the lamella of a BSC-1 cell. A and B--two consecutive optical sections. In A, two radii are drawn from the centrosphere into the lamella. In B, the centrosphere is shown by the arrowhead. Scale bar--10 m.

vesicles appear essentially chaotic, with frequent changes in direction and speed. As seen from the trajectories in Figure 2, the displacement in 5-s intervals often constitutes several micrometers. The average displacement of vesicles during 1 min was, however, statistically indistinguishable from zero in


794

Cell Biology International, Vol. 26, No. 9, 2002

Fig. 2. Representative trajectories of vesicles. Points are positions of the vesicle centroid recorded during 1 min at 5-s intervals. Initial position is in the origin of the co-ordinates. Direction toward the cell centre is indicated with an arrow.

11 of the 17 cells (at significance level 0.05). In 6 other cells, the velocity averaged over 1 min did not exceed 2.5 m/min. The low or negligible average velocity comes out of the lack of persistency in displacement of an individual vesicle and is not a result of averaging of velocities of different vesicles. The persistency of the vesicle movement can be characterized by an angle between the consecutive displacements of an individual vesicle as vectors (Freed and Lebowitz, 1970). This angle will be 0 if the direction of the vesicle movement is perfectly conserved; it will be 180 if the direction is reversed. If the direction of every displacement of the vesicle is chosen at random, or if forward and reverse movement along a line randomly alternate, the average angle will be 90 , which is the average of uniform distribution between 0 and 180 . The average angle is therefore a measure of the autocorrelation in the direction of the vesicle movement. It is convenient to map the interval (0, 180 ) linearly onto ( 1, 1) so that the value 1 of the so-introduced autocorrelation function corresponds to zero angle between displacements reflecting full correlation in standard way, the value ­1 corresponds to 180 reflecting the full negative correlation, and the value 0 reflects absence of any correlation. We averaged the angle between displacements of a vesicle in 5-s intervals and plotted the corresponding value of the autocorrelation function versus the time interval between the intervals in which the displacements took place (Fig. 3). The directions of displacements in adjacent 5-s intervals are positively correlated indicating that the movement is processive.

However, the correlation of the directions of displacements separated by another 5-s interval is indistinguishable from zero (P< 0.05). The correlation remains statistically equal to zero as the time interval between the displacements increases further. This quantitatively indicates that the movement of a vesicle is rapidly losing its direction, and explains the low or negligible average velocity of

Fig. 3. Autocorrelation function for the vesicle wandering. Dashed lines show 95% confidence interval for the zero correlation. Directions of segments of vesicle trajectories which are separated in time by > 10 s have insignificant correlation. Dots represent values of the empiric autocorrelation function computed for 30 vesicles traced during 1 min with 5-s intervals in a representative cell. Solid line is exponential approximation of the empiric function with the decrement 0.33 s 1. The value of the autocorrelation function at time 0 is assigned to 1 to stipulate that the momentary velocity of an vesicle has a definite value.


Cell Biology International, Vol. 26, No. 9, 2002

795

Table 1. Apparent diffusion constants of vesicles in BSC-1 cells and the minimal significance level, P, at which the hypothesis of their equality can be rejected against the alternative that the radial one is larger
Cell Radial, m2/min, SE 12.2 19.4 7.1 9.1 15.7 11.1 14.8 5.4 4.9 12.8 12.9 7.4 28.1 13.2 12.8 15.9 11.6 3.1 4.8 1.9 2.3 4.1 2.2 3.7 1.3 1.2 3.4 3.4 1.8 3.9 3.7 3.5 4.1 3.0 Tangential, m2/min, SE 3.4 1.9 1.3 2.0 2.5 1.3 3.8 1.0 0.5 1.9 0.8 0.5 0.4 2.4 0.5 8.1 1.7 0.8 0.4 0.3 0.5 0.6 0.3 0.9 0.2 0.1 0.5 0.2 0.1 0.1 0.7 0.1 2.1 0.4 4 1 2 2 8 2 2 8 2 8 2 1 3 2 3 P

Fig. 4. Variance in radial (closed circles) and tangential (open circles) components of vesicle displacement vs. duration of the displacement. Also shown are linear regression lines for the radial (solid) and for tangential (long dash) components, and 70% confidence intervals for the radial (short dash) and for tangential (dot) components. The data in the plot are for 30 vesicles.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

· 10 · 10 · 10 · 10 · 10 · 10 · 10 · 10 · 10 · 10 · 10 · 10 · 10 · 10 · 10 0.048 7 · 10

3 3 3 3 4 4 3 4 4 4 4 4 5 3 4

4

SE, standard error of the estimate.

the vesicles as time average for individual vesicles rather than population average. Thus, the average velocity is not an adequate characteristic of the observed movement. In a suitable characteristic of the intensity of the movement, the displacements in opposite direction must not average out but add up. Mean square displacement is such a characteristic employed earlier to describe the locomotion of fibroblasts (Gail and Boone, 1970) or bacteria (Berg, 1993), where it displays linear growth with time. Since the vesicles at times exhibit statistically significant mean velocity we compute variance in the vesicle displacement instead of the mean square displacement and plot it versus time (Fig. 4). The variance in both radial and tangential components increased in proportion to time (Fig. 4). This property is characteristic to diffusion processes and Brownian random walk, and, formally, one-half of the slope of the plot of variance versus time is the diffusion constant, or diffusivity (Berg, 1993). We use further the word `pseudodiffusion' or `apparent diffusion' to indicate that the fundamental nature of the vesicle motion is known to be different from conventional thermal diffusion. In each cell, the apparent diffusion constants were

different for the radial and tangential components of displacement, which quantitatively revealed anisotropy in the vesicle motion (Table 1). The radial component of the motion had higher apparent diffusion constant, in the range 4.9­28.1 m2/min. The tangential component had lower apparent diffusion constant, between 0.4 and 8.1 m2/min. In each investigated cell radial motility prevailed as reflected by the ratio of the radial to tangential apparent diffusivity in the range 2 to 56 with the mean 11. The difference between the apparent diffusion constants of the two components was significant at significance level P< 0.005 (except cell #16, where still P< 0.05). The preferred motion along cell radii corresponded to the radial orientation of microtubules stretching from the centrosome towards cell periphery (Fig. 5). UV microirradiation Pattern of motility of the granules demonstrates visible changes 20 min after UV microirradiation of the centrosome. Radial displacements for the long distances become infrequent making radial and tangential components of motion apparently similar. This pattern remains the same for at least 2 h after irradiation. No visible changes were observed in the cells where the part of cytoplasm without centrosome had been irradiated (data not shown).


796

Cell Biology International, Vol. 26, No. 9, 2002

Fig. 6. Variance in radial (closed circles) and tangential (open circles) components of vesicle displacement vs duration of the displacement, after UV microirradiation of the centrosome. Also shown are linear regression lines for the radial (solid) and for tangential (long dash) components, and 70% confidence intervals for the radial (short dash) and for tangential (dot) components. The data in the plot are for 30 vesicles.

Table 2. Apparent diffusion constants of vesicles in BSC-1 cells 1 h after UV microirradiation of the centrosome, and the minimal significance level, P, at which the hypothesis of their equality can be rejected against the alternative that the radial one is larger
Cell Radial, m2/min, SE 1.35 3.63 1.87 0.84 3.83 5.56 2.3 0.20 4.93 0.50 0.21 1.08 1.45 1.12 Tangential, m2/min, SE 0.54 0.94 1.18 0.79 5.33 4.28 1.00 0.13 1.28 0.31 0.20 0.80 1.11 0.48 P

1 2 3 4 5 6 7 Fig. 5. Microtubules ( -tubulin, red) diverging radially from the centrosome ( -tubulin, green, but superposition makes the spot appear yellow, (A) and phase-contrast image of the same cell (B). The preferential orientation of vesicle motion (white double-sided arrow) to and from the phase-contrast centrosphere (arrowhead) coincides with the orientation of the microtubules. Scale bar--10 m.

0.125 0.101 0.299 0.444 0.230 0.243 0.144

SE, standard error of the estimate.

In 1 h after irradiation of the centrosome by ultraviolet microbeam, the vesicle motion continued, yet became isotropic. Isotropic vesicle wandering is reflected by the radial apparent diffusivity

being indistinguishable from the tangential one at the significance level of at least 0.1 (Fig. 6, Table 2). At the same time, the radial arrangement of the microtubules in the lamellum was perturbed, and no microtubule convergence centre was evident. Instead numerous randomly oriented microtubules ran throughout the cytoplasm (Fig. 7).


Cell Biology International, Vol. 26, No. 9, 2002

797

DISCUSSION We have presented a quantitative analysis of the motion of small phase-contrast vesicles in cultured mammalian cells. General applicability of our pseudodiffusion framework for the vesicle motion follows from the fact that the observed trajectories of the vesicles look practically indistinguishable from trajectories reported for other cell and vesicle types. So, secretory vesicles that transport a specific protein in Vero fibroblast cells move rapidly, but with random tracks (Wacker et al., 1997). Motion of pigment granules that accounts for their dispersion in the cytoplasm of melanophore cells (Rodionov et al., 1998), of adenovirus particles on their way to the nucleus (Suomalainen et al., 1999) and of lipid droplets redistributing in Drosophila embryos (Welte et al., 1998; Gross et al., 2000) is also random. Despite these particles frequently exhibiting long processive runs, their resultant average velocity is by an order of magnitude smaller and often statistically close to zero. Fast and processive displacements whose direction alternated randomly were earlier reported for a number of vesicle types in a number of cell lineages (Freed and Lebowitz, 1970). The random movement was shown to be microtubule-dependent (Freed and Lebowitz, 1970; Wacker et al., 1997) and could be affected by genetic modifications of dynein (Gross et al., 2000) or expression of p50/dynamitin (Suomalainen et al., 1999), or mutations in another regulator of motility klar (Welte et al., 1998). In other cases, it depended on actin filaments and could be stopped by sodium azide (Rodionov et al., 1998). In our experiments, disarrangement of microtubules was accompanied by the corresponding change of the vesicle motion. We conclude that the intense random motion of vesicles results from the active drag by molecular motors along cytoskeletal fibres. The random character of the motion is evidently due to the processivity of a motor being smaller than necessary to move across the cell. That the next motor in action and the filament along which it will move are not necessarily the same as during the previous run will cause the transported vesicle to change direction of its movement (Wacker et al., 1997). A mechanism that specifically switches direction of transport was also proposed (Welte et al., 1998; Gross et al., 2000). The limited processivity of vesicle movement was quantified here as autocorrelation of the direction of vesicle displacements that rapidly reached zero as time passed. In applied studies on different aspects of intracellular transport, the vesicle motion has been

Fig. 7. Microtubules are randomly distributed in the cell with centrosome being ablated by the UV microirradiation. Scale bar--10 m.

Table 3. Apparent diffusion constants of vesicles in BSC-1 cells 1 h after UV microirradiation of the part of cytoplasm not containing the centrosome, and the minimal significance level, P, at which the hypothesis of their equality can be rejected against the alternative that the radial one is larger
Cell Radial, m2/min, SE 3.73 1.61 1.62 5.22 3.46 19.23 0.54 0.41 0.71 1.34 0.88 5.15 Tangential, m2/min, SE 0.94 0.18 0.43 0.95 0.31 1.22 0.14 0.04 0.19 0.23 0.08 0.34 P

1 2 3 4 5 6

8 · 10 2 · 10 0.025 8 · 10 2 · 10 2 · 10

3 4

4 4 5

SE, standard error of the estimate.

The reduction of the difference between the radial and tangential apparent diffusivities of vesicles was achieved mainly by decrease of the radial apparent diffusivity. In some cells we observed simultaneous decrease of the radial diffusivity and increase of the tangential one. The ratio of the radial diffusivity to tangential fell down and was in the range 1 to 4 with the mean 2. In the intact cells after 2 h of observation the radial and tangential apparent remained statistically distinct with their ratio being in the range 1 to 38 with the mean 13 (4 cells). The diffusivities also differed significantly in the cells that had been irradiated in a region of the cytoplasm close to the centrosome, but not in the centrosome itself (Table 3). Two hours after the irradiation, the ratio of the radial and tangential diffusivities remained high in this control group of cells, in the range 4 to 15 with the mean 9.


798

Cell Biology International, Vol. 26, No. 9, 2002

typically characterized with a formal measure, such as average distance between the most distant points on a trajectory of a vesicle (Wacker et al., 1997). A fundamental analysis of details in a complex motion of vesicles led to a conclusion that a clear connection between its numerous parameters was absent (Breuer et al., 1988). We report that a major property of vesicle motion is linear gain of variance in displacement with time, that makes vesicle motion formally identical to the random walk (Berg, 1993). This analogy makes the complex vesicle movement easily comprehensible and allows characterizing it with a single informative quantity, the apparent diffusion constant. The word `apparent' is important here, because the active mechanism of vesicle movement makes it incomparably more intense than possible for the thermal Brownian motion. For the real, thermal diffusion of a vesicle of the radius r = 0.75 m in the cytoplasm with viscosity = 30 poise (Mogilner and Oster, 1996) the diffusion constant D = kT/3 r, where kT = 4.1 ·10 21 N·m (Berg, 1993), should be 11.6 · 10 3 m2/min. We have measured that the motion of vesicles along the cell radii has the apparent diffusivity 4.9­28.1 m2/min. So, the intensity of the stochastic, but active vesicle movement exceeds the intensity of thermal motion by three orders of magnitude. The apparent diffusivity of the vesicles is higher for the displacements along cell radii than in the tangential direction. This quantitatively reveals the anisotropy of the intracellular transport that was qualitatively characterized before (Rebhun, 1967; Freed and Lebowitz, 1970; Rogers et al., 1997). The radial direction along which the vesicle movement is the most intense corresponds to the radial arrangement of microtubules that emanate from the centrosome. That the anisotropy of the intracellular transport depends on the centrosome is demonstrated by the vesicle motion becoming isotropic after the microtubule-organizing function of the centrosome is abolished by ultraviolet microirradiation. Random movement of transport vesicles suggested the `trial and error' model for vesicular transport (Wacker et al., 1997). According to this concept, random walk of a transport vesicle is an efficient strategy to search for the target membrane where the successful molecular recognition and membrane fusion occur. This idea is consistent with that the vesicle's molecular address code specifies only the target membrane but not the route by which the latter is reached (Rothman, 1994). Pseudodiffusion description proposed in this study provides an adequate quantitative framework for

this theory. The testable, quantitative prediction for vesicular transport can now be made based on the general theory of diffusion processes (Berg, 1993). The mean time to reach the target membrane should be s2/2D, if the distance between the source and target membranes is s and the apparent diffusivity of the vesicle is D. This formula is exact for the one-dimensional unrestricted random walk. Since the radial apparent diffusivity of the small vesicles in BSC-1 cells greatly exceeds the tangential one, their motion can be regarded onedimensional (along the cell radius) in the first approximation. If the typical D is 12.5 m2/min, then, for example, a secretory vesicle budding off the Golgi apparatus will reach the plasma membrane in the lamellipodium 20 m away in 16 min. Such a travel time is physiologically meaningful, moreover, this estimate is close to the characteristic time of secretion of a viral protein in Vero cells (Wacker et al., 1997). In a more precise model of intracellular transport, its anisotropic and at least two-dimensional (in a flat cell region) nature can be taken into account by using the apparent diffusion tensor (de Groot and Mazur, 1984) whose components are the radial and tangential apparent diffusion constants measured in this work and solving the diffusion equation with relevant boundary conditions. Experimental testing of the quantitative predictions will confirm or refute the `trial and error' theory. The vesicles monitored in our experiment were not apparently specialized for transport. Nevertheless, a random but rapid motion of such vesicles throughout the cytoplasm seems to be an efficient strategy for the large animal cell to enhance and integrate its metabolism. Such motion, analogous to stirring in a retort, should increase rates of metabolite exchange between the vesicles and the cytosol. Even the continuous random drag of metabolically passive but large vesicles (storage compartments, etc.) will lead to an intensive stirring of the viscous cytosol itself. Thus the active pseudodiffusion of vesicles can be considered as the animal homologue of the cytoplasmic streaming in plant cells. In a typical plant cell, the vesicles circulate around the cell primarily under the action of myosin motors that move along actin fibres. This is believed to speed up and integrate metabolism in large plant cells (Alberts et al., 1989). Diffusion-like motion to and from the cell centre along radially oriented microtubules (Kellogg et al., 1994) under the action of motors of the dynein and kinesin families (Goodson et al., 1997), appears to be a realization of the same function in the structural and biochemical


Cell Biology International, Vol. 26, No. 9, 2002

799

framework of a typical large animal cell. In the context of animal morphogenesis, the anisotropic motion between the cell centre and periphery may be particularly important for effective cell spreading, formation of long processes and maintenance of an extended, polarized cell form. ACKNOWLEDGEMENTS This work was supported by grant 02-04-48839a from Russian Basic Science Foundation to IAV. REFERENCES
ALBERTS B, BRAY D, LEWIS J, RAff M, ROBERTS K, WATSON JD, 1989. Molecular Biology of the Cell. New York, Garland. 1294 pp. BERG HC, 1993. Random Walks in Biology. Princeton, NJ, Princeton University Press. 152 pp. BERNS MW, RATTNER JB, BRENNER S, MEREDITH S, 1977. The role of the centriolar region in animal cell mitosis. A laser microbeam study. J Cell Biol 72: 351­367. BRAY D, 2001. Cell Movements: from Molecules to Motility. New York, Garland. 372 pp. BREUER AC, EAGLES PA, LYNN MP, ATKINSON MB, GILBERT SP, WEBER L, LEATHERMAN J, HOPKINS JM, 1988. Long-term analysis of vesicle translocation in isolated axoplasm of Myxicola infundibulum. Cell Motil Cytoskel 10: 391­399. FREED JJ, LEBOWITZ MM, 1970. The association of a class of saltatory movements with microtubules in cultured cells. J Cell Biol 45: 334­354. GAIL MH, BOONE CW, 1970. The locomotion of mouse fibroblasts in tissue culture. Biophys J 10: 980­993. GOODSON HV, VALETTI C, KREIS TE, 1997. Motors and membrane traffic. Curr Opin Cell Biol 9: 18­28. GROOT DE SR, MAZUR P, 1984. Non-equilibrium thermodynamics. New York, Dover. 510 pp. GROSS SP, WELTE MA, BLOCK SM, WIESCHAUS EF, 2000. Dynein-mediated cargo transport in vivo: a switch controls travel distance. J Cell Biol 148: 945-955. HINCHCLIFFE EH, MILLER FJ, CHAM M, KHODJAKOV A, SLUDER G, 2001. Requirement of a centrosomal activity for cell cycle progression through G1 into S phase. Science 291: 1547-1550.

KELLOGG DR, MORITZ M, ALBERTS BM, 1994. Centrosome and cellular organization. Annu Rev Biochem 63: 639­ 674. KHODJAKOV A, COLE RW, RIEDER CL, 1997. A synergy of technologies: combining laser microsurgery with green fluorescent protein tagging. Cell Motil Cytoskel 38: 311­317. MANIOTIS A, SCHLIWA M, 1991. Microsurgical removal of centrosomes blocks cell reproduction and centriole generation in BSC-1 cells. Cell 67: 495­504. MOGILNER A, OSTER G, 1996. Cell motility driven by actin polymerization. Biophys J 71: 3501­3510. REBHUN L, 1967. Saltatory particle movements and their relation to the mitotic apparatus. In: Levine L, ed. The Cell in Mitosis. New York, Academic Press. 67­106. REBHUN L, 1972. Polarized intracellular particle transport: saltatory movements and cytoplasmic streaming. Int Rev Cytology 32: 93­167. RODIONOV VI, HOPE AJ, SVITKINA TM, BORISY GG, 1998. Functional coordination of microtubule and actin based motility in melanophores. Curr Biol 8: 165­168. ROGERS SL, TINT IS, FANAPOUR PC, GELFAND VI, 1997. Regulated bidirectional motility of melanophore pigment granules along microtubules in vitro. Proc Natl Acad Sci USA 94: 3720­3725. ROTHMAN JE, 1994. Mechanisms of intracellular protein transport. Nature 372: 55­63. SUOMALAINEN M, NAKANO MY, KELLER S, BOUCKE K, STIDWILL RP, 1999. Microtubule-dependent plus- and minus end-directed motilities are competing processes for nuclear targeting of adenovirus. J Cell Biol 144: 657-672. UZBEKOV RE, VOTCHAL MS, VOROBJEV IA, 1995. Role of the centrosome in mitosis: UV micro-irradiation study. J Photochem Photobiol B 29: 163­170. VOROBJEV IA, SVITKINA TM, BORISY GG, 1997. Cytoplasmic assembly of microtubules in cultured cells. J Cell Sci 110: 2635­2645. VOROBJEV IA, UZBEKOV RE, KOMAROVA YA, ALIEVA IB, 2000. -tubulin distribution in interphase and mitotic cells upon stabilization and depolymerization of microtubules. Membr and Cell Biol 14: 219­235. WACKER I, KAETHER C, KROMER A, MIGALA A, ALMERS W, GERDES H-H, 1997. Microtubule-dependent transport of secretory vesicles visualized in real time with a GFP-tagged secretory protein. J Cell Sci 110: 1453­1463. WELTE MA, GROSS SP, POSTNER M, BLOCK SM, WIESCHAUS EF, 1998. Developmental regulation of vesicle transport in Drosophila embryos: forces and kinetics. Cell 92: 547-557. WILSON EB, 1925. The Cell in Development and Heredity. New York, Macmillan. 1232 pp.