Cell Motility and the Cytoskeleton 40:368–378 (1998) Disassembly of Actin Filaments Leads to Increased Rate and Frequency of Mitochondrial Movement Along Microtubules Mira Krendel, George Sgourdas, and Edward M. Bonder* Program in Cellular and Molecular Biodynamics, Department of Biological Sciences, Rutgers University–Newark, Boyden Hall, Newark, New Jersey In activated sea urchin coelomocytes, cytoplasmic organelles move along distinct actin and microtubule dependent pathways, actin-based motility is driven by an unconventional myosin, and microtubule disassembly does not effect actindependent organelle motility [D’Andrea et al., 1994: J. Cell Sci. 107:2081–2094]. Given the growing evidence for potential interactions between components of the actin and microtubule cytoskeletons, we examined the effect of actin filament disassembly on the movement of mitochondria along microtubules in activated coelomocytes. Coelomocytes treated with cytochalasin B (CB), to disrupt actin filaments, exhibited a thinning of the cytoplasm, enhanced lateral undulation of microtubules, and ceased centripetal cortical flow of actin. Interestingly, the loss of actin filaments resulted in a ,1.5-fold increase in the average velocity of outward and inward moving mitochondria and increased the frequency of centripetal movement. To test if enhanced motility along microtubules was a consequence of decreased actin-myosin interaction, coelomocytes were treated with 2,3butanedione monoxime (BDM), a potent inhibitor of myosin activity [Cramer and Mitchison, 1995: J. Cell Biol. 131:179–189]. BDM inhibited all types of actin-based motility observed in these cells including retrograde cortical flow, protrusion and retraction of the cell edge, and movement of intracellular organelles. Surprisingly, BDM treatment stopped the movement of mitochondria in CB-exposed cells, suggesting that BDM can also act as an inhibitor of organelle movement along microtubules. Collectively, these data demonstrated that microtubule-dependent mitochondrial motility and microtubule movement were sensitive to changes in the assembly state of the actin cytoskeleton. Cell Motil. Cytoskeleton 40:368–378, 1998. r 1998 Wiley-Liss, Inc. Key words: actin filaments; motility; organelle velocity; motor proteins INTRODUCTION Motility of intracellular organelles, a phenomenon observed in all eukaryotic cells, plays an important role in many physiological processes including, but not limited to, transport of secretory organelles, axonal transport, and cellular polarization [Sheetz et al., 1989; Kelly, 1990; Hirokawa, 1993]. Experimental evidence indicates that organelles may be transported along microtubules by the dynein and kinesin motor protein families and along actin filaments by the myosin family of proteins [reviewed in Langford, 1995]. Kinesins translocate organelles toward the plus-end of microtubules while dyneins drive move- r 1998 Wiley-Liss, Inc. ment toward the minus-end [Kelly, 1990; Skoufias and Scholey, 1993]. From a cellular perspective, kinesins move cargo outward toward the periphery of the cell *Correspondence to: Edward M. Bonder, Department of Biological Sciences, Rutgers University–Newark, University Heights, 101 Warren Street, Newark, NJ 07102. E-mail: firstname.lastname@example.org Contract grant sponsor: American Heart Association, New Jersey Affiliate; Contract grant sponsor: Charles and Johanna Busch Fund at Rutgers University. Received 30 January 1998; accepted 13 April 1998 Actin Modulates Microtubule-Based Motility while dyneins escort cargo inward or centripetally [see Vale, 1987; Hirokawa, 1996]. The actin-dependent motor protein, myosin-II, has long been known to participate in cytoplasmic streaming in plants where cytoplasmic organelles are transported along actin cables [Shimmen and Tazawa, 1982; Kachar, 1985]. Additionally, organelle motility in squid axoplasm [Langford et al., 1994; Bearer et al., 1996], in neuronal growth cones [Evans and Bridgman, 1995] and in axons and sea urchin coelomocytes depleted of microtubules [Morris and Hollenbeck, 1995; D’Andrea et al., 1994] has been linked to various members of the unconventional myosin family. Individual organelles may also be associated with multiple motor proteins. For example, evidence suggests the presence of both microtubule- and actin-based motors on the same subset of golgi-derived organelles in epithelial cells [Fath et al., 1994] and individual axoplasmic organelles have been documented to switch between actin and microtubular tracks [Kuznetsov et al., 1992]. Furthermore, dynactin, the dynein activator protein complex, may be associated with the actin binding protein spectrin [Holleran et al., 1996], establishing a potential molecular linkage between the actin cytoskeleton and microtubuledependent motor proteins. Consequently, intracellular organelle motility is likely to be the result of coordinated interplay between the two transport systems leading to physiologically appropriate organelle motility. Sea urchin coelomocytes contain abundant membranous organelles concentrated in the perinuclear region of the cell that upon stimulation are induced to undergo outward directed translocation [D’Andrea et al., 1994]. Organelles translocating along actin filaments move slower than those moving along microtubules. Unconventional myosin classes I, VI, VII, and IX have been identified in coelomocytes using biochemical, immunological, or molecular biological techniques [D’Andrea et al., 1994; Sirotkin et al., 1996]. The microtubule motor protein, kinesin, is localized to endosomes and ER in non-activated coelomocytes [Henson et al., 1992]. Since coelomocytes possess functionally distinct organelle transport pathways, an understanding of organelle motility in activated coelomocytes might provide insight into coordinated regulation of microtubule- and actin-dependent organelle transport. D’Andrea et al.  demonstrated that disruption of microtubules did not result in a detectable effect on the velocity of organelles moving along actin filaments. However, since there appears to be reasonable evidence suggesting an association between microtubule motors and the actin cytoskeleton [Kuznetsov et al., 1992; Holleran et al., 1996], disruption of actin cytoskeletal integrity might have profound effects on organelles translocating along microtubules. To investigate this possibility, two approaches were used to modify actin cytoskeletal function: 369 depolymerization of actin filaments using cytochalasin B and inhibition of myosin motor activity using the pharmacological agent BDM. Disassembly of actin filaments resulted in accelerated movement of mitochondria along microtubules as well as enhanced lateral movement of microtubules. Treatment of coelomocytes with BDM, previously shown to specifically inhibit the ATPase activity of myosins II and V without affecting kinesin ATPase [Cramer and Mitchison, 1995], resulted in the cessation of actin- and microtubule-dependent mitochondrial motility. These results suggest that the mechanisms regulating microtubule-dependent organelle translocation and microtubule dynamics are responsive to changes in the actin cytoskeleton. MATERIALS AND METHODS Quantitative Analysis of Organelle Motility Strongylocentrotus purpuratus coelomocytes were plated onto coverslips either directly from coelomic fluid or after isolation by sucrose gradient centrifugation [D’Andrea et al., 1994]. Cells were washed with isotonic buffer (0.5 M NaCl, 5 mM MgCl2, 20 mM HEPES, pH 7.4) and activated by perfusion of hypotonic buffer (0.3 M NaCl, 5 mM MgCl2, 20 mM HEPES, pH 7.4). BDM (Sigma, St. Louis, MO) was used from a 200 mM stock in water prepared on the day of experimentation and cytochalasin B (Sigma) was diluted from a 0.5 mg/ml stock in DMSO. Coelomocyte activation and organelle motility were monitored using a Zeiss (Thornwood, NY) Axiophot light microscope equipped with DIC optics and a Plan-Neo x100 (NA 1.3) objective lens. Images were collected using a Hamamatsu (Bellevue, WA) Newvicon camera, frame averaged and digitally enhanced using a Quantex image processor prior to time-lapse video documentation. To track individual organelle movements, videotapes were played back through a Macintosh computer and movie sequences were created using NIH Image software (Version 1.44). X and Y coordinates of organelles in a 20 µm by 20 µm region of a cell were determined over 120–150 sec, using NIH Image, and plotted using Delta Graph software. The positions of individual organelles, undergoing a single continuous excursion, were mapped at three-sec intervals and the resulting coordinates were used to calculate the average velocity of moving organelles. Fluorescence Labeling of Coelomocyte Organelles Cells were exposed to Rhodamine-123 (1.66 µg/ ml), acridine orange (0.5 mg/ml), or neutral red (0.5 µg/ml) in either isotonic or hypotonic buffer for 10–15 min. Labeled cells were observed by DIC and epifluorescence microscopy using a Zeiss Axiophot microscope 370 Krendel et al. coupled to a Dage SIT camera. Images were averaged and recorded onto S-VHS video tape. To disrupt mitochondrial membrane potential, cells were exposed to 1 µM FCCP before, during, or after labeling with fluorescent vital dyes. Immunofluorescent Labeling Coelomocytes were fixed as described by Edds  and stained with fluorescent phalloidin (Molecular Probes, Eugene, OR) to label filamentous actin. For indirect immunofluorescent staining of microtubules, cells were fixed with methanol at -20°C, rinsed with phosphate-buffered saline (PBS), blocked with PBS containing 3% BSA, and incubated with DM1a monoclonal antibodies against a-tubulin (Sigma). Primary antibody staining was visualized using FITC-labeled goat antimouse secondary antibodies. Immunofluorescence images were collected using a Bio-Rad (Gaithersburg, MD) MRC-1024 laser scanning confocal microscope. RESULTS Mitochondrial Motility in Coelomocytes In non-activated sea urchin coelomocytes, the majority of intracellular organelles were located centrally, surrounding the cell nucleus where they appeared to undergo a slow churning movement (Fig. 1A) [see also D’Andrea et al., 1994]. Cell activation, by exposure to hypotonic buffer, resulted in rapid induction of movement of slightly elongated and highly refractile organelles outward into the periphery of the cell (see Table I and Fig.1). By phase contrast microscopy, the outward moving organelles appeared phase dense. Organelle movement was saltatory, and individual organelles were capable of changing their direction of movement. Organelles moved along linear radial tracks and in many cases several organelles would follow the same path, as if moving along the same microtubule or actin microfilament bundle (Fig. 1B) [D’Andrea et al., 1994]. Actin filaments formed radially aligned bundles throughout the lamella and dense meshworks in the perinuclear and lamellar regions (Fig. 2A). Microtubules formed a loose basket around the nucleus with radial microtubules extending into the lamellae but seldom reaching out to the edge of the cell (Fig. 2B). To determine the identity of the outward translocating organelles, we labeled coelomocytes with a number of cell-permeant, fluorescent vital dyes; rhodamine-123 to stain mitochondria and acridine orange and neutral red to label acidic compartments. Rhodamine 123 labeled most of the peripherally translocating organelles in cells activated with hypotonic buffer (see Fig. 1D), whereas acridine orange and neutral red stained only perinuclear organelles (data not shown). Comparisons of DIC and fluorescence images established that most of the organelles labeled with rhodamine 123 were small, slightly elongated, and highly refractile by DIC and phase dense by phase contrast microscopy. By contrast, perinuclear acidic vesicles were often large, round, and phase-lucent (data not shown). To verify the specificity of rhodamine 123 labeling, coelomocytes were treated with carbonyl cyanide p-(trifluoromethoxy)phenylhydrazone (FCCP), an uncoupler of mitochondrial oxidative phosphorylation, which disrupts the membrane potential and eliminates uptake of rhodamine 123. In all cases, cells exposed to FCCP prior to, during, or after exposure to rhodamine 123 did not show uptake of dye into the translocating organelles. Staining of the large, round perinuclear organelles with either acridine orange or neutral red was not effected by exposure to FCCP. These observations establish that one hallmark of coelomocyte activation is the rapid induction of mitochondrial motility. Effect of Cytochalasin B Treatment on Mitochondrial Motility Cytochalasin B is commonly used in studies of actin-dependent motility since it disrupts the actin cytoskeleton in vivo, presumably by slowing plus-end assembly and destabilizing filaments [Bonder and Mooseker, 1986; Sampath and Pollard, 1991]. Coelomocytes exposed to CB in isotonic buffer showed a thickening of the outer edge of the lamellae, a loss of centripetal cortical flow, inhibition of protrusion/retraction activity of the lamellae, ‘‘thinning’’ of the cytoplasm, and apparent brownian motion of perinuclear organelles (data not shown) [see also Edds, 1993]. In agreement with Edds , CB resulted in mass depolymerization of actin filament bundles and meshworks while leaving behind a rim of residual actin around the edge of the cell and occasional cytoplasmic actin aggregates (Fig 2C). When coelomocytes were first treated with hypotonic buffer to induce outward translocation of mitochondria and then exposed to CB, the initial response was a ‘‘snapping’’ back of the cytoplasm along with many, but not all, mitochondria into the perinuclear region, as well as inhibition of cortical flow and protrusive activity. Mitochondria that remained in the lamellae continued to move along linear tracks (Fig. 3) of radially aligned individual microtubules or bundles of microtubules present within the lamellar cytoplasm that was depleted of actin filaments (Figs. 2D and 3). All the effects of CB were fully reversible upon incubation in hypotonic buffer lacking CB. Treatment with CB resulted in several noteworthy qualitative effects on mitochondrial motility, microtubule organization, and microtubule motility. In control cells, we never observed mitochondria that completely trans- Actin Modulates Microtubule-Based Motility 371 Fig. 1. Identification of mitochondrial motility in coelomocytes. A: A video-DIC micrograph of a single coelomocyte showing the presence of numerous organelles in the perinuclear region of the cell. B: The same cells as shown in A after treatment with hypotonic buffer. Numerous refractile organelles migrated outward into the broad, thin lamellar region, often along straight linear tracks. C: Video-DIC micrograph of a coelomocyte after treatment with hypotonic buffer containing the mitochondria specific vital dye, rhodamine-123. Numerous organelles have migrated out into the periphery of the cell; note the two discrete groups of refractile organelles (arrows). D: Fluorescence micrograph of the cell presented in C. Rhodamine-123 labeling is well correlated with the motile refractile organelles observed using DIC microscopy, in particular compare the organelles identified by arrows, see C and D. located out to the very edge of the cell whereas in CB-treated cells mitochondria often traveled to the edge of the lamellae where they appeared ‘‘stranded’’ or moved tangentially along the edge of the cell on microtubule tracks visible by DIC. Additionally, mitochondria appeared to move inward with greater frequency and movement along microtubules appeared faster (see be- low). In the absence of an intact actin cytoskeleton, the microtubule network tended to collapse forming parallel bundles or loose ‘‘knots’’ of microtubules (see Figs. 2D and 3). Interestingly, individual microtubules or bundles of microtubules started to undergo wild lateral undulations through the thinned lamellar cytoplasm (Fig. 3, see arrows). 372 Krendel et al. TABLE I. Effect of CB on Coelomocyte Mitochondrial Translocation Control Average velocity (µm/sec 6 SEM) Fraction of time spent moving (%) CB Inward Outward Inward Outward 0.098 6 0.005 7.2 0.102 6 0.003 22.5 0.145 6 0.005 17.9 0.149 6 0.004 29.5 Quantitative Analysis of Mitochondrial Motility Effect of BDM on Mitochondrial Motility Individual mitochondria present within 20 by 20 µm square sectors of lamellae were tracked over a 2-min time span to determine organelle velocity and relative frequency of inward and outward movements. The position of each mitochondrion was mapped at 3-sec intervals and the average velocities were calculated for those mitochondria that changed position over the 3 sec. Previously, D’Andrea et al.  identified mitochondria undergoing a continuous excursion and used the information gathered to calculate average velocity. The two methods for determining average velocity of mitochondria resulted in essentially identical measurements (data not shown). An added advantage of tracking individual moving and stationary mitochondria within the 20 by 20 µm area is the ability to determine both the frequency of inward versus outward movement and the percentage of time mitochondria spent moving. The velocities observed in control, activated coelomocytes ranged from 0.01 µm/sec to 0.48 µm/sec, with an average outward velocity of 0.102 µm/sec and average inward velocity of 0.098 µm/sec (Table I). Individual mitochondria were stationary approximately 70% of the time and the frequency for outward movement was 3 times higher than for inward movement (Table I). Average velocity of both outward and inward organelle movement increased approximately 1.5-fold in cells treated with CB (Table I). This 1.5-fold increase in average velocity was a characteristic feature of CB treatment as reflected in individual cell-by-cell comparisons before and after exposure to CB (Fig. 4). Comparison of the velocities in control vs. CB-treated coelomocytes by an unpaired t-test showed that the calculated difference was statistically significant with a 99.9% confidence level. In addition to increased velocity, mitochondria in CB-treated cells spent a greater proportion of time moving as compared to control cells (50 vs. 30% respectively; see Table I). Further, greater than 70% of the organelles in CB-treated cells moved at rates of 0.1 µm/sec or higher whereas in control cells less than 50% of the organelles moved with velocities of 0.1 mm/sec or higher. Another striking change was the doubling in frequency of inward motility following disassembly of the actin cytoskeleton (17.9% in CB-treated vs. 7.2% in control cells; see Table I). The observed increase in mitochondrial motility after CB treatment could have resulted from the loss of the actin filament network and/or the consequent loss of functional actin-myosin interaction. In each case, the presence of an intact actin cytoskeleton contributes a ‘‘braking’’ effect that would impede the progress of organelles moving along microtubules. To examine if myosin motor activity could influence microtubule-based organelle transport, coelomocytes were treated with BDM, previously shown to be a specific inhibitor of myosin-II and V activity [Cramer and Mitchison, 1995]. Treatment of non-activated cells with 5–40 mM BDM resulted in a dose-dependent attenuation of cortical flow and protrusion/ retraction of lamellae. In the presence of 20 or 40 mM BDM, cortical flow and lamellar activity were completely stopped without any readily detectable destructive changes in overall cell morphology, presence of actin filaments, or presence of microtubules (see Fig. 2). One qualitative effect of BDM was that actin filaments and microtubules appeared ‘‘collapsed,’’ giving the impression that more bundles were present in treated cells vs. non-treated cells (compare Fig. 2E, F with Fig. 2A, B). When coelomocytes were first treated with isotonic buffer containing 20–40 mM BDM and then activated by hypotonic buffer containing BDM, there was little or no mitochondrial movement observed (data not shown); the inhibitory effect was fully reversible upon incubation of cells with hypotonic buffer lacking BDM. To assess whether the observed effect of BDM on mitochondrial dispersion was due to an effect on coelomocyte activation, coelomocytes were first treated with hypotonic buffer to initiate mitochondrial translocation and then exposed to hypotonic buffer containing BDM. Mitochondria that were translocating after treatment with hypotonic buffer would stop moving after 3–5 min of exposure to hypotonic buffer containing BDM and mitochondria resumed translocation after the preparations were washed free of BDM (Fig. 5). Interestingly, mitochondria that ceased moving after BDM treatment appeared to be loosely tethered to the location where they stopped (Fig. 5B). Occasionally, some mitochondria would undergo short excursions in either the outward or inward direction that were slower and lasted for shorter time periods than typically observed in non-treated cells (Fig. 5). Actin Modulates Microtubule-Based Motility 373 Fig. 2. Effect of CB and BDM on the organization of actin filaments and microtubules. A & B: Laser scanning confocal micrographs of individual, non-activated coelomocytes labeled with rhodaminephalloidin to detect actin filaments (A) and by indirect immunofluorescence to detect microtubules (B). C & D: Laser scanning confocal micrographs of coelomocytes stained for actin filaments (C) and microtubules (D) after treatment with isotonic buffer containing 1 µg/ml of CB. Exposure to CB resulted in depolymerization of the majority of actin filaments, leaving behind a narrow, peripheral ridge of phallodin stained actin (Panel C). CB treatment led to the rearrange- ment of microtubules into a ‘‘collapsed’’ network consisting of large bundles of microtubules (compare panels B and D). E and F: Laser scanning confocal micrographs of coelomocytes stained for actin (E) and microtubules (F) after treatment with isotonic buffer containing BDM. There appears to be no detrimental effect on the quantity of actin filaments or microtubules after exposure to BDM. However, qualitatively, treatment with BDM appeared to have an effect on the organization of microtubules since the microtubules in treated cells tended to form bundles. Bar - 10 µm. Effect of BDM on Mitochondrial Motility in CB Treated Coelomocytes ity could be a consequence of mitochondria being ‘‘stuck’’ to an actin filament by a poisoned myosin consequently preventing movement along microtubules. If inhibition of microtubule motility is a secondary effect of BDM, Since BDM is reported to be a specific inhibitor of myosin function, the observed lack of microtubule motil- 374 Krendel et al. Fig. 3. Mitochondrial and microtubule motility in CB-treated cells. Four frames from a two-minute time-lapse video sequence of a single cell after exposure to CB. In this sequence, outward and inward movement of mitochondria (arrowheads) as well as lateral undulation of a microtubule (see arrow) can be detailed. Bar - 10 µm. disruption of the actin-myosin interaction should release the mitochondria and allow movement along a microtubule until another poisoned actin-myosin interaction forms. This attachment and release could account for the occasional, short excursions that mitochondria take in the presence of BDM. To investigate this possibility, coelomocytes were activated using hypotonic buffer containing CB, followed by exposure to both CB and BDM. Prior to BDM treatment, mitochondria were observed moving along microtubular tracks and movement was ceased within 5 min after exposure to BDM (Fig. 6) and the majority of mitochondria remained stationary. Some mitochondria occasionally performed short outward and inward excursions along microtubular tracks but they never translocated for distances comparable to those measured in non-BDM-treated cells (Fig. 6). After washout of BDM using hypotonic buffer containing CB, mitochondrial motility rapidly returned to the levels observed prior to BDM treatment (Fig. 6). DISCUSSION Intracellular organelle transport is driven by distinct families of motor proteins, the myosins that move cargo Actin Modulates Microtubule-Based Motility 375 that the dynamic properties, distribution, and length of microtubules are modulated by the actin cytoskeleton. Furthermore, quantitative analysis of live cells established that the direction and velocity of mitochondria moving along microtubules are linked to the assembly state of cytoplasmic actin filaments. Thus, our observations indicate that microtubule dynamics and microtubuledependent organelle motility are effected by the surrounding actin cytoskeleton. Actin Cytoskeleton Moderates the Velocity of Mitochondria Moving Along Microtubules Fig. 4. Quantitive analysis of mitochondrial motility in coelomocytes. Three individual cells were analyzed for outward and inward mitochondrial velocities prior to treatment with CB (j, h) and after treatment with CB (e, B). In each case, treatment with CB resulted in an ,50% increase in the average mitochondrial velocity. Error bars represent SEM. along actin filaments and the kinesins and dyneins that carry cargo along microtubules [Skoufias and Scholey, 1993; Fath and Burgess, 1994; Langford, 1995]. While the two transport systems consist of distinct components performing different functions, they must be closely linked in order to provide coordinated organelle transport. The existence of a coordinated interaction is plausible given that organelles can possess both actin- and microtubule-dependent motors [Fath et al., 1994], can switch between actin and microtubule tracks [Kuznetsov et al., 1992], and the actin binding protein spectrin is associated with dynactin, the dynein activator complex [Holleran et al., 1996]. Previously, we demonstrated that coelomocytes possess both actin- and microtubule-dependent organelle motility that could be distinguished by the velocity of organelle translocation [D’Andrea et al., 1994]. Fast organelle motility in coelomocytes occurs primarily along microtubules with an average velocity of ,0.1 µm/sec that can be eliminated by disassembling microtubules with nocodazole. Slow organelle motility is likely to be driven by an unconventional myosin [D’Andrea et al., 1994; Sirotkin et al., 1996] having a velocity less than 0.05 µm per sec. Depolymerization of microtubules has no detectable effect on the velocity of organelles moving along actin filaments [D’Andrea et al., 1994]. In the present study, we used organelle motility in sea urchin coelomocytes to investigate the consequences of actin cytoskeletal integrity and myosin motor protein activity on organelle translocation along microtubules. We report Correlative analysis of fluorescence and DIC images of rhodamine 123 labeled cells led to the conclusion that the small, oblong, highly refractile organelles imaged by video-DIC were mitochondria and that all or almost all motile organelles in activated coelomocytes were mitochondria. These data provide documentation that, similar to axonal transport of mitochondria [Morris and Hollenbeck, 1995], coelomocyte mitochondria are transported along both actin filaments and microtubules. Previously, Henson and co-authors , using anti-sea urchin egg kinesin antibodies, demonstrated that some populations of coelomocyte organelles, but not mitochondria, are associated with the microtubule-dependent motor protein kinesin. Lee and Hollenbeck  found that in neurons increased kinesin phosphorylation promoted kinesin association with membranes and correlated with an increase in axonal transport of organelles. Thus, if coelomocyte activation, in addition to being calcium dependent [Henson and Schatten, 1983; Hyatt et al., 1984], was also accompanied by an increase in kinase activity there could result elevated kinesin phosphorylation leading to mitochondrial association and motility. In support of this idea are our unpublished observations that stimulation of protein kinase C by the phorbol ester, TPA, can induce organelle motility in resting coelomocytes. Interestingly, when the actin filament network in coelomocytes was disrupted with CB the average inward and outward velocity of mitochondria moving along microtubules increased. Previously, Morris and Hollenbeck  found that the average velocity of mitochondria moving in axons increased after treatment with cytochalasin. Can the observed increase in average mitochondrial velocity after CB treatment be explained by a simple ‘‘statistical effect’’ due to the loss of slow, actin-dependent movement? Elimination of the slow moving contingent of mitochondria would result in a statistically elevated average velocity, without any increase in the actual rate of translocation of organelles moving along microtubules. In coelomocytes, unlike axons, it is possible to distinguish between the velocities for actin-dependent and microtubule-dependent mitochondrial movement. With this information it is possible to eliminate slow, 0.05 µm/sec or less, actin-dependent 376 Krendel et al. Fig. 5. Analysis of mitochondrial movements after BDM treatment. Trajectories of individual organelles were traced in a 20-by-20 µm area within a single lamella after activation with hypotonic buffer (Panel A), after incubation in hypotonic buffer containing 20 mM BDM (Panel B), and after wash-out of BDM with hypotonic buffer (Panel C). Trajectories were generated by mapping the position of 10-15 mitochondria every 3 seconds over 120 seconds using NIH Image software. Treatment with BDM resulted in a dramatic decrease in the movement of mitochondria and in many cases the mitochondria appeared to remain essentially stationary. Removal of BDM restored typical mitochondria motility (Panel C). Fig. 6. Analysis of mitochondrial movement after treatment with BDM and CB. Coelomocytes were initally activated with hypotonic buffer to induce organelle motility followed by subsequent treatment with hypotonic buffer containing CB (Panel A), hypotonic buffer containing CB and BDM (Panel B), and finally hypotonic buffer with CB (Panel C). Trajectories were generated as described in Figure 5. Interestingly, treatment with BDM greatly slowed or completely inhibited the movement of mitochondria along microtubules (Panel B). The inhibitory effect of BDM on microtubule-based motor activity was fully reversable upon incubation with hypotonic buffer lacking BDM (Panel C). movements from the calculations to obtain a corrected microtubule driven velocity. Consequently, if the observed increase in the calculated average velocity is a statistical effect, the corrected velocity before CB treatment should equal the velocity measured after CB treatment. Using this analysis, as predicted from the data in D’Andrea et al. , mitochondria moved along microtubules with a velocity of ,0.1 µm/sec in control cells. Interestingly, disruption of actin filaments resulted in an ,35% increase in the average velocity for both outward and inward mitochondrial movement on microtubules. The calculated difference in velocities was statistically significant (99.9% confidence level) as judged by an unpaired t-test indicating that mitochondria do indeed move faster along microtubules after actin filaments were disassembled in the lamellae of coelomocytes. Whether the newly observed acceleration of organelles moving along microtubules occurs in other systems (e.g., axons) after disruption of actin filaments remains unknown. One possible explanation for the observed increase in mitochondrial velocity is that actin filaments provide mechanical resistance countering the propulsive forces of the microtubule motor proteins present on the organelle surface. In vitro measurements of viscoelasticity suggest that actin filaments have a twofold higher viscosity than equivalent solutions of nonfilamentous actin [Sato et al., 1985]. Since, actin filament depolymerization by CB lowers viscosity [MacLean-Fletcher and Pollard, 1980] it Actin Modulates Microtubule-Based Motility is reasonable to predict that the viscosity of the coelomocyte cytoplasm decreased as a consequence of CB treatment. If treatment with CB resulted in a 2-fold decrease in the viscosity of the cytoplasm, then based on Stokes’ Law, there should be an acccompanying 2-fold increase in organelle velocity. Interestingly, the measured ,1.5-fold increase in outward and inward mitochondrial velocity is consistent with the above viscosity-based effect of actin filaments on organelles moving along microtubules. Alternatively, since mitochondria in activated coelomocytes may be associated both with actin- and microtubule-dependent motors, binding interactions between myosin-type motors and actin filaments may slow down movements along microtubules. In this case, CB treatment eliminates myosin-actin filament interactions and the observed velocities are solely dependent upon microtubule motor protein activity. In order to test such a ‘‘myosin-braking’’ model, cells were treated with BDM to inhibit myosin motor activity. BDM, a specific inhibitor of myosins, has been used in studies of muscle contraction [Higuchi and Takemori, 1989; McKillop et al., 1994; Siegman et al., 1994], postmitotic cell spreading [Cramer and Mitchison, 1995], formation of focal contacts [Chrzanowska-Wodnicka and Burridge, 1996], and neuronal growth cone motility [Lin et al., 1996]. Similar to other cells [Lin et al., 1996; Waterman-Storer and Salmon, 1997], BDM inhibited myosin-driven cortical flow in coelomocytes; however, unexpectedly, BDM inhibited all organelle motility in coelomocytes including organelle motility occurring along microtubules. To date, BDM is proposed to directly inhibit the myosin II and V ATPase [Herrmann et al., 1992; McKillop et al., 1994; Cramer and Mitchison, 1995]. While typically used as a myosin specific inhibitor, BDM can potentially effect other cellular activities including dephosphorylation of acetylcholinesterase [Hobbiger, 1963], ion channel conductance [Phillips and Altschuld, 1996], and/or reduce the level of myosin II regulatory light chain phosphorylation [Siegman et al., 1994; ChrzanowskaWodnicka and Burridge, 1996]. Consequently, the effects of BDM may be less specific than originally thought and results obtained in in vivo experiments utilizing BDM should be interpreted cautiously. In addition to accelerated velocities, mitochondria spent greater periods of moving after CB-treatment. This effect was especially pronounced for inward mitochondrial movement where there was a greater than twofold increase in the frequency of movement. While it is tempting to speculate on the potential mechanism of this observation, the absence of available information on the detailed polarity of microtubules and the presence of dynein motor proteins preclude such discussion. 377 Organization of Microtubules Is Modulated by Actin Filaments In addition to the observed effect on mitochondrial velocities, depolymerization of actin filaments altered the intracellular distribution of microtubules and microtubuleassociated organelles. After cytochalasin treatment, it appeared that the microtubules formed tangled bundles that often extended out to the edge of the cell. In the presence of cytochalasin, many mitochondria accumulated at the edge of the cell while in control cells the outermost regions of the lamellae did not contain mitochondria. Strikingly, the time-lapse video images documented the presence of wild, lateral undulations of microtubules or microtubule bundles giving an impression that these movements were actively driven. These observations suggest that actin filaments, by forming a cytoplasmic matrix, can limit lateral motility, bundle formation, and plus-end growth of microtubules at the cell periphery. Our findings complement and extend the recent report of Waterman-Storer and Salmon  who elegantly demonstrated that actin-dependent cortical flow profoundly effects microtubule bending, elongation, and breakage and support the proposal by Edds  that actin filaments may limit microtubule distribution in coelomocytes. Our findings show that mitochondrial transport and redistribution in activated coelomocytes are a result of coordinated activity of actin- and microtubule-based transport systems. Interestingly, the structural integrity of the actin cytoskeleton exerted profound effects on the motility characteristics of mitochondria and microtubules. This result raises the possibility that localized assembly/disassembly of actin filaments may participate in spatial regulation of microtubule dynamics, microtubule-dependent motor activity, and organelle positioning. ACKNOWLEDGMENTS We are grateful to Rob Harrell, Volodia Sirotkin, and Tim Mure for their help and assistance during the course of experimentation and manuscript preparation. REFERENCES Bearer, E.L., DeGiorgis, J.A., Jaffe, H., Medeiros, N.A., and Reese, T.S. (1996): An axoplasmic myosin with a calmodulin-like light chain. Proc. Natl. Acad. Sci. U.S.A. 93:6064–6068. Bonder, E.M., and Mooseker, M.S. (1986): Cytochalasin B slows but does not prevent monomer addition at the barbed end of the actin filament. J. Cell. Biol. 102:282–288. Chrzanowska-Wodnicka, M., and Burridge, K. (1996): Rho-stimulated contractility drives the formation of stress fibers and focal adhesions. J. Cell. Biol. 133:1403–1415. Cramer, L.P., and Mitchison, T.J. (1995): Myosin in involved in postmitotic cell spreading. J. Cell Biol. 131:179–189. 378 Krendel et al. D’Andrea, L., Danon, M., Sgourdas G.P., and Bonder, E.M. (1994): Identification of coelomocyte unconventional myosin and its association with in vivo particle/vesicle motility. J. Cell Sci. 107: 2081–2094. Edds, K.T. (1984): Differential distribution and function of microtubules and microfilaments in sea urchin coelomocytes. Cell Motil. Cytoskeleton 4:269–281. Edds, K.T. (1993): Effects of cytochalasin and colcemid on cortical flow in coelomocytes. Cell Motil. Cytoskeleton 26:262–273. Evans, L.L., and Bridgman, P.C. (1995): Particles move along actin filament bundles in nerve growth cones. Proc. Natl. Acad. Sci. U.S.A. 92:10954–10958. Fath, K.R., and Burgess, D.R. (1994): Membrane motility mediated by unconventional myosin. Curr. Opin. Cell Biol. 6:131–135. Fath, K.R., Trimbur, G.M., and Burgess, D.R. (1994): Molecular motors are differentially distributed on Golgi membranes from polarized epithelial cells. J. Cell. Biol. 126:661–675. Henson, J.H., and G. Schatten. (1983): Calcium regulation of the actin-mediated cytoskeletal transformation of sea urchin coelomocytes. Cell Motil. Cytoskeleton 3:525–534 Henson, J., Nesbitt, D., Wright, B.D., and Scholey, J.M. (1992): Immunolocalization of kinesin in sea urchin coelomocytes. J. Cell Sci. 103:309–320. Herrmann, C., J. Wray, F. Travers, and T. Barman. (1992): Effect of 2,3-butanedione monoxime on myosin and myofribrillar ATPases. An example of an uncompetitive inhibitor. Biochemistry 31:12227–12232. Higuchi, H., and Takemori, S. (1989): Butanedione monoxime suppresses contraction and ATPase activity of rabbit skeletal muscle. J. Biochem. 105:638–643. Hirokawa N. (1993): Axonal transport and the cytoskeleton. Curr. Opin. Neurobiol. 3:724–731. Hirokawa, N. (1996): Organelle transport along microtubules: The role of KIFs. Trends Cell Biol. 6:135–141. Hobbiger, F. (1963): Reactivation of phosphorylated acetylcholinesterase. In Koelle, G.B. (ed.): ‘‘Cholinesterases and Anticholinesterase Agents. Handbuch der Experimentellen Pharmacologie.’’ Springer-Verlag, Berlin: Springer-Verlag, pp. 921–988. Holleran, E.A., Tokito, M.K., Karki, S., and Holzbaur, L.F. (1996): Centractin (ARP1) associates with spectrin revealing a potential mechanism to link dynactin to intracellular organelles. J. Cell Biol. 135: 1815–1829. Hyatt, H.A., Shure, M.S., and Begg, D.A. (1984): Induction of shape transformation in sea urchin coelomocytes by the calcium ionophore A23187. Cell Motil. Cytoskeleton 4:57–71. Kachar, B. (1985): Direct visualization of organelle movement along actin filaments dissociated from Characean algae. Science. 227:1355–1357. Kelly R.B. (1990): Microtubules, membrane traffic, and cell organization. Cell. 61:5–7. Kuznetsov, S.A., Langford, G.M., and Weiss, D.G. (1992): Actindependent organelle movement in squid axoplasm. Nature. 356:722–725. Langford, G.M. (1995): Actin- and microtubule-dependent organelle motors: interrelationships between the two motility systems. Curr. Opin. Cell Biol. 7:82–88. Langford, G.M., Kuznetsov, S.A., Johnson, D., Cohen, D.L., Weiss, D.G. (1994): Movement of axoplasmic organelles on actin filaments assembled on acrosomal processes: Evidence for a barbed-end-directed organelle motor. J. Cell. Sci. 107:2291– 2298. Lee, K.-D. and Hollenbeck, P.J. (1995): Phosphorylation of kinesin in vivo correlates with organelle association and neurite outgrowth. J. Biol. Chem. 270:5600–5605. Lin, C.H., Espreafico, E.M., Mooseker, M.S., and Forscher, P. (1996): Myosin drives retrograde F-actin flow in neuronal growth cones. Neuron 16:769–782. MacLean-Fletcher, S., and Pollard, T.D. (1980): Mechanism of action of cytochalasin B on actin. Cell. 20:329–341. McKillop, D.F.A., Fortune, N.S., Ranatunga, K.W., and Geeves, M.A. (1994): The influence of 2,3-butanedione 2-monoxime (BDM) on the interaction between actin and myosin in solution and in skinned muscle fibers. J. Muscle Res. Cell Motil. 15:309–318. Morris, R.L., and Hollenbeck, P.J. (1995): Axonal transport of mitochondria along microtubules and F-actin in living vertebrate neurons. J. Cell Biol. 131:1315–1326. Phillips, R.M., and Altschuld, R.A. (1996): 2,3-butanedione 2-monoxime (BDM) induces calcium release from canine cardiac sarcoplasmic reticulum. Biochem. Biophys. Res. Commun. 229:154–157. Sampath, P., and Pollard, T.D. (1991): Effects of cytochalasin, phalloidin and pH on the elongation of actin filaments. Biochemistry 30:1973–1980. Sato, M., Leimbach, G., Schwarz, W.H., and Pollard, T.D. (1985): Mechanical properties of actin. J. Biol. Chem. 260: 8585–8592. Sheetz, M.P., Steuer, E.R., and Schroer, T.A. (1989): The mechanism and regulation of fast axonal transport. Trends Neurosci. 12: 474–478. Shimmen, T. and Tazawa M. (1982): Reconstitution of cytoplasmic streaming in Characeae. Protoplasma. 113:127–131. Siegman, M.J., Mooers, S.U., Warren, T.B., Warshaw, D.M., Ikebe, M., and Butler, T.M. (1994): Comparison of the effects of 2,3butanedione monoxime on force production, myosin light chain phosphorylation and chemical energy usage in intact and permeabilized smooth and skeletal muscles. J. Muscle Res. Cell. Motil. 15:457–472. Sirotkin, V., Krendel, M., and Bonder, E.M. (1996): Sea urchin eggs and coelomocytes express multiple myosins. Mol. Biol. Cell 7s:37a. Skoufias, D.A. and Scholey, J.M. (1993): Cytoplasmic microtubulebased motor proteins. Curr. Opin. Cell Biol. 5:95–104. Vale, R.D. (1987): Intracellular transport using microtubule-based motors. Ann. Rev. Cell Biol. 3:347–378. Waterman-Storer, C.M., and Salmon, E.D. (1997): Actomyosin-based retrograde flow of microtubules in lamella of migrating epithelial cells influences microtubule dynamic instability and turnover and is associated with microtubule breakage and treadmilling. J. Cell Biol. 139:417–434.