Cell Motility and the Cytoskeleton 34:230-245 (1996) Microtubules Can Modulate Pseudopod Activity From a Distance Inside Macrophages Gustavo R. Rosania and Joel A. Swanson Department of Cell Biology and the Program in Biological and Biomedical Sciences, Harvard Medical School, Boston, Massachusetts Microtubules are thought to influence cell shape as structural components of an integrated cytoskeletal matrix. Here we show that microtubules can affect the dynamics of macrophage pseudopodia without being integrated into their structure. Macrophages landing on glass surfaces spread within 15 min into flattened circular cells with radial symmetry, and the radial distribution of microtubules reflected this symmetry. Depolymerization of microtubules using nocodazole, colchicine, or vinblastine did not inhibit macrophage spreading or the early establishment of radial symmetry. Shortly after spreading, however, macrophages without microtubules gradually became asymmetric, assuming irregular, lobed profiles. The asymmetry resulted from exaggerated protrusion and retraction of pseudopodia, with net retraction overall. This loss of radial symmetry could be inhibited by treatment of initially spread cells with cytochalasin D, indicating that the change in cell shape was mediated by the actin cytoskeleton. Intact microtubules suppressed the exaggerated pseudopod movements, even when they were separated by a distance from the cell margin. In cells treated with taxol, microtubules remained clustered near the cell center after spreading, yet the dynamics of pseudopodia at the cell margin were reduced and cells maintained a circular profile. Similarly, in cells treated with low concentrations of nocodazole, a much reduced microtubule cytoskeleton nonetheless suppressed pseudopod dynamics. We propose that microtubules act to stabilize cell shape at a distance from the cell edge via a biochemical intermediate that affects the structure or function of the microfilament System. 0 1996 Wiley-Liss, Inc. Key words: microtubules, macrophage pseudopodia, cell shape INTRODUCTION Microtubules have long been thought to play a role in regulating cellular morphogenetic processes, including the assembly and positioning of the cleavage furrow [Fishkind and Wang, 19951, the extension and guidance of growth cones in neurons [Tanaka et al., 199.51, and the establishment or maintenance of cell shape and polarity [Dugina et al., 1995; Glasgow and Danielle, 1994; Pletjushkina et al., 1994; Schliwa and Honer, 1993; Tomasek and Hay, 1984; Vasiliev et al., 19701. In spreading polarized fibroblasts, microtubule ends lie in the vicinity of cell attachment sites, suggesting that microtubules stabilize focal contacts at the sites of cellsubstrate adhesion and thereby organize actin filaments 0 1996 Wiley-Liss, Inc. [Rinnerthaler et al., 19881. Microtubules can also affect contractile activity at the cell cortex, suggesting that they also control the function of myosin [Bornens et al., 19891. Nevertheless, the mechanism by which microtubules influence the organization, dynamics, or contractile activity of the microfilament system is unknown. The effects of microtubules on microfilaments may be studied by looking at how the two systems interact to maintain cell shape. In most cells, shape reflects both the Received January 25, 1996; accepted April 4, 1996. Address reprint requests to Joel A. Swanson, Department of Anatomy and Cell Biology, University of Michigan Medical School, Ann Arbor, MI 48109. Microtubules Modulate Pseudopod Activity viscoelastic properties [Janmey, 19911 and the dynamics of the cytoskeleton. In spread cells, pseudopod extensions occur continuously at sites of actin polymerization at the periphery of the cell [Okabe and Hirokawa, 1989; Symons and Mitchison, 19911. Cytochalasins, which block actin polymerization, inhibit cell spreading, phagocytosis, ruffling, and pseudopod extensions at the cell edge [Axline and Reaven, 19741. Counteracting the continuous growth of actin polymer at the edge of the cell is a continuous centripetal flow of actin, thought to be mediated in part by myosin [Bray and White, 1988; Fischer et al., 1988; Heath and Hollifield, 1993; Lee et al., 19931. The depolymerization of monomers from the “pointed” ends of actin filaments and the polymerization of monomers at the “barbed” ends of the filaments lead to the treadmilling of subunits along the polymer [Wang, 19851, which may also contribute to centripetal flow. Pseudopod extension or retraction may result when the cortical flow of actin is unbalanced by polymerization of new actin filaments at the cell edge. Microtubules could influence cell shape by affecting any of these processes. Studies using a variety of microtubule-specific drugs indicate a role for microtubules in the maintenance of cell shape. When elongated fibroblasts [Gail and Boone, 1971; Middleton et al., 1988; Vasiliev et al., 19701, PC12 cells [Joshi et al., 19851, or other polarized cell types are incubated with colchicine or nocodazole, microtubules depolymerize and cell shape changes from polarized to symmetric. When drug removal allows microtubules to re-polymerize, asymmetric shape is restored [Vasiliev et al. , 19701. Fibroblasts spreading in the presence of microtubule inhibitors form discoidal shapes, indicating that microtubules are required for both the establishment and maintenance of the asymmetry [Vasiliev et al., 19701. Fibroblasts treated with taxol, a drug that stabilizes microtubules [de Brabander et al., 1981; Schiff and Honvitz, 19801 and disrupts the arrangement of microtubules in the cell [Pletjushkina et al., 19941, also acquire circular, symmetric shapes [Pletjushkina et al., 19941. In experiments with primary fibroblasts [Middleton et al., 19891 and cells that are normally circular or unpolarized [Domnina et al., 1985; Middleton et al., 19881, depolymerization of microtubules showed little effect on cell shape, indicating that the effects of microtubules are context dependent. It has been proposed that microtubules stabilize cell shape by acting as rigid, compression-resistant struts, linked to the actomyosin network and mechanically counteracting the contractile force generated by that network [Ingber, 1993; Joshi et al., 19851. Consistent with this hypothesis, microtubule depolymerization in PC 12 cells increases tension in neurites [Dernell et al., 19881, leading to their retraction [Joshi et al., 1985; Solomon 231 and Magendantz, 19811. Simultaneous incubation in nocodazole and cytochalasin inhibited retraction [Solomon and Magendantz, 19811, indicating that microfilaments are somehow linked to the microtubules and mediate the retraction. Microtubules have also been proposed to stabilize cell shape by serving as tracks that direct components for protrusive activity toward the cell edge [Bershadsky and Vasiliev, 19931. In fibroblasts, disrupting vesicle trafficking with antibodies that block the function of the microtubule-dependent mechanoenzyme kinesin [Rodionov et al., 19931, leads to changes in pseudopod activity and cell shape similar to those observed when microtubules are depolymerized. This is consistent with the idea that polarized membrane insertion at the leading edge is responsible for increased protrusive activity [Kupfer et al., 1987; Singer and Kupfer, 19861. Finally, microtubules could act as intracellular signal transducers [de Brabander et al., 19771, controlling cell shape by affecting the levels of some diffusible molecule or ion that in turn affects the dynamics of the actin cytoskeleton. Although there is no clear evidence for a diffusible molecule linking microtubule depolymerization with changes in actin dynamics, a number of signaling molecules influencing protrusive activity and cytoskeletal organization [Hall, 19941 could serve as intermediates. In this study, time-lapse video microscopy and quantitative morphometric image analysis were used to examine the contribution of microtubules to the establishment and maintenance of cell shape. In macrophages, we report that microtubules are essential for maintaining cell symmetry, yet their effects are independent of microtubule distribution inside the cell. Microtubules modulate pseudopod dynamics at the cell margin without necessarily extending all the way to that margin. We conclude that macrophage microtubules do not stabilize cell shape by mechanically supporting the cell edge, nor are they necessary as tracks that extend to the cell margin. These observations support the hypothesis that microtubules affect cell shape globally through a chemical intermediate influencing the structure and function of the actin cytoskeleton. MATERIALS AND METHODS Materials Fluorescent probes were obtained from Molecular Probes (Eugene, OR). Monoclonal antibodies against vimentin (H5) and tubulin (E7) were developed by Joshua Sanes and Michael Klymkowsky, and obtained from the Developmental Studies Hybridoma Bank (Department of Pharmacology and Molecular Sciences, The Johns Hopkins University School of Medicine, Baltimore, MD 21205, and the Department of Biology, University of Iowa, Iowa City, IA 52242). Secondary antibodies were 232 Rosania and Swanson from Vector Laboratories (Burlingame, CA). Other reagents were obtained from Sigma Chemical Co. (St. Louis, MO). Cell Culture Bone marrow-derived macrophages were obtained from femurs of C3H/HeJ mice (Jackson Laboratory, Bar Harbor, ME) and cultured in vitro [Cannon and Swanson, 19921. Five-day macrophages were suspended in cold PD (137 mM NaC1, 3 mM KC1, 7 mM phosphate buffer, pH 7.4) and 2 X lo6 cells were plated in 100 mm Lab-Tek polystyrene Petri dishes in Dulbecco-Modified Eagle’s Medium plus 10% fetal calf serum, overnight at 37°C in 5% CO,. BSA-Coated Cover Slips Thirteen millimeter, circular, no. 1 cover slips were cleaned for 2 h in concentrated sulfuric acid, washed overnight under continuous flow of ddH20, allowed to dry, and stored in a clean, dust-free container. On the day of the experiment, cover slips were placed in 24-well plates and incubated in 1 ml of 0.1 m g / d polyL-lysine in PD for 15 rnin with continuous agitation. Afterwards, they were washed 3 X 5 min in PD, and incubated with 10 mg/ml BSA in PD for at least 30 rnin in continuous agitation. Immediately prior to addition of cells, cover slips were washed 2 x 5 rnin in Ringer’s buffer (155 mM NaC1, 5 mM KC1, 2 mM CaCl,, 1 mM MgCl,, 2 mM NaH2P0,, 10 mM HEPES, and 10 mM glucose) without agitation. were incubated under permeabilization buffer for 15 rnin at room temperature prior to fixation. Subsequent incubations were in cytoskeletal buffer. For morphometry , each preparation was stained for tubulin (fluorescein immunofluorescence) and for total cellular protein using Texas Red [Coates et al., 19921. For immunofluorescent labeling, permeabilized cells were incubated with blocking buffer (cytoskeletal buffer plus 2% heat-inactivated goat serum) for 15 min, then for 1 h in hybridoma cell supernatant (E7 for tubulin or H5 for vimentin), diluted 1:4 in cytoskeletal buffer. After incubation with primary antibody, cells were washed 3 X 15 min in blocking buffer. Cells were stained with secondary antibody (fluorescein, horse antimouse IgG, Vector Laboratories, Burlingame, CA) in blocking buffer plus 25 pg/ml Texas Red sulphonyl chloride. For F-actin staining, cells were incubated with 20 U/ml of NBD-phallicidin (400 U/ml stock, in MeOH). For micrographs and cytoskeletal visualization, cells were observed with a fluorescein filter set on a Zeiss Axioskop epifluorescence microscope (Zeiss, Thornwood, NY). General cell morphology was observed with a Texas Red filter set. For quantitative measurements, video images of cells were collected on a Zeiss IM35 inverted microscope, with a multichannel plate intensifier (Videoscope International, Washington, DC) and a Nuvicon video camera connected to a Metamorph image processing system (Universal Imaging Corporation, West Chester, PA). Synchronized Cell Spreading Image Processing and Cell Morphometry Cells were removed from dishes using cold Ringer’s buffer, and were suspended in Ringer’s buffer containing either taxol, nocodazole (both from 10 mM stock in DMSO), colchicine or vinblastine for 45 min at 4°C. Cells (2-4 x lo4) were dropped onto BSA-coated cover slips in 24-well plates, and left to settle for 5-15 min at 4°C. Under these conditions, cells attached to the BSAcoated surface, but they did not spread. After cells had attached, the buffer was removed and replaced with fresh buffer containing the different drugs. The plate was then transferred to 37°C water bath, and cell spreading began immediately. After various intervals, cells were fixed and processed for immunof luorescence or image analysis. DMSO (0.2%) did not affect cell morphology. To outline cell shape, background-subtracted images of cells labeled with Texas Red were collected using a X63 Zeiss planapo oil immersion lens. A low pass filter was applied to the images to remove pixelation from the cell edge. This improved image signal-to-noise as judged by comparing processed video images with the cells seen through the eyepiece. Images were then thresholded to make binary masks, and morphometric measurements were applied to these masks using built-in measurement routines of the Metamorph imaging system. Four shape parameters were useful: area, which measures the surface covered by the cell; length, which measures the longest chord through the cell; radial ratio, which is the ratio of the distance from the geometric center to the closest point on the perimeter of the cell, divided by the distance from that center point to the furthest point on the perimeter; and shape factor, which is a ratio equal to 47r (Area)/Perimete?. Shape factor and radial ratio measured how closely the cell profile resembled a circle. For each data point, at least 50 cells were measured in two or three different experiments, unless otherwise indicated. lmmunofluorescence Cells were fixed for 30 min in 2% paraformaldehyde in PD at room temperature, then were washed 3 x 5 rnin in PD, and permeabilized in cytoskeletal buffer (30 mM HEPES, 10 mM EGTA, 0.5 mM EDTA, 5 mM MgSO,, 33 mM potassium acetate, pH 7.4, and 7% PEG 400) plus 0.5% Triton X-100. For extracted cells, cells Microtubules Modulate Pseudopod Activity Videomicroscopy For time-lapse movies and motion analysis, 25 mm, circular, no. 1 BSA-coated cover slips were assembled on Leiden chambers in a temperature-controlledmicroscope stage (Medical Systems Corp., Greenvale, NY). One milliliter of Ringer's buffer, with or without drugs, was immediately added into the chamber, covered with silicon oil to prevent evaporation, and warmed to 37°C. Fifty microliters of cell suspension was then added to the chamber. Individual cells were located as they were falling onto the cover slip. Time-lapse movies of cell spreading were recorded using a video camera (DAGE-MTI, Inc. , Michigan City, IN) connected to the Metamorph Imaging System. Data were stored on a Panasonic optical disc recorder. Kinetic Analysis of Pseudopod Dynamics Cells were allowed to spread on BSA-coated cover slips in Leiden chambers, with or without drugs, at 37°C. To measure pseudopod extension and retraction, the areas of displacement during a 5 min interval were determined. Two phase-contrast video images of individual cells were collected 5 min apart, 30 to 60 min after the cells had begun spreading. Binary images of the cell profile were generated from the video frames. Subtracting the early binary from the latter one produced a binary image of the areas of extension, and vice versa. Kinetic Analysis of Cell Spreading Time lapse video images of spreading cells were analyzed by measuring cell areas in each frame. Phase contrast, background-subtractedimages of at least 8 cells were collected at 15 sec intervals for 15 min. An edgedetection routine was applied to the stack using a Sobel filter to outline the silhouette of the cell. This silhouette was then thresholded, binarized, and eroded such that the processed image could be satisfactorily superimposed on the shape of the cell in the original phase contrast video image. The area of the objects in this image-processed stack were then measured. RESULTS Effects of Microtubules on Cell Shape Macrophages treated with nocodazole, vinblastine, or colchicine spread into highly irregular and asymmetric forms, while control cells spread into circles (Fig. 1). By immunofluorescence microscopy, microtubules in control cells were arranged as spokes radiating from the centrosome to the cell margin. Longer microtubules bent at the margin. Cells that were incubated with microtubule inhibitors and extracted before fixation were completely devoid of filamentous microtubule staining. In 233 unextracted cells, free tubulin labeled cytoplasm diffusely. Although nocodazole, colchicine, and vinblastine are different chemically, their effects on cell shape were similar. Thus, their effects on cell shape were most likely a consequence of their common effect on microtubule polymerization. All depolymerizing drugs significantly decreased cell area, length, radial ratio (P < .001, t-test) and shape factor (P < .05, t-test) as determined by measuring at least 50 cells in three separate experiments. To test whether microtubules were affecting the establishment of cell shape, the morphologies of cells were measured at various times during spreading. Spreading was synchronized by allowing attachment to cover slips at 4"C, followed by warming. Cells fixed at various times after warming were analyzed morphometrically (Fig. 2). Within 5 minutes, 100% of control cells had microtubules while 100% of nocodazole-treated cells lacked microtubules, as determined by anti-tubulin immunofluorescence (n = 100, data not shown). Both control and nocodazole-treated cells spread synchronously, symmetrically, and at similar rates during the first 5 min. Both cell populations reached one-half of their maximum area within the first 5 min and about 90% of their maximum area within the first 15 min. There were no significant differences in the length, shape factor, and radial ratio of cells spreading for the first 5 min. Therefore, symmetrical spreading was microtubule-independent. Maintenance of circular profiles required microtubules. The loss of symmetry in nocodazole-treated cells followed the initially symmetric spreading response. Although cell length remained relatively constant, the areas of the nocodazole-treated cells decreased relative to control cells. The radial ratio and shape factors also dropped significantly below control values after 15 min. Since nocodazole-treated cells decreased in area but not length, it indicated that the cells were retracting or extending incompletely in the absence of microtubules, and that these differences might be responsible for the changes in morphology. To examine the role of microtubule distribution in the maintenance of cell shape, we analyzed spreading of taxol-treated cells (Fig. 3). By immunofluorescence, microtubules in taxol-treated cells were shorter, detached from the centrosome, and excluded from the cell margin. Nonetheless, macrophages in taxol maintained their circular profiles. When cells spread in taxol plus nocodazole, microtubules were depolymerized and spreading was irregular. This indicated that the effect of taxol on cell symmetry was related to its effect on microtubule stability rather than to some other biochemical effect of taxol. Quantitative measurements of cell shape supported these impressions (Fig. 4). The symmetry of cells in 234 Rosania and Swanson TEXAS RED UNEXTRACTED IgG Fig. 1 . Microtubule depolymerization leads to changes in cell shape. Cells were suspended in Ringer’s (A,B,C) or Ringer’s plus different microtubule inhibitors: 20 pM colchicine (D,E,F); 20 pM nocodazole (G,H,I); 200 nM vinblastine (J,K,L). They were then allowed to spread for 60 min before fixing and staining with Texas Red to outline EXTRACTED IgG cell shape (A,D,G,J) and anti-tubulin immunofluorescence (fluorescein signal, B,C,E,F,H,I,K,L). Cells were either fixed without extraction, or were extracted before fixation, as indicated. Scale bar = 10 p m . Microtubules Modulate Pseudopod Activity 50- 1200- v 1OOOh P i - E E 800- 2 3. 4 235 600- 5bD 400- 4 5 40: 30 2010- 200- 0 0 20 40 60 Time (min) 80 0 1 100 0 20 40 60 Time (min) 80 0 100 I I 0.80.8- 0.75- 0.750.70 .+ 2 0.65- 8 0.65- Y Z .d -0 2 0.60.550.50.45 taxol indicated that microtubules could stabilize cell morphology without extending to the cell margin. The geometry of the microtubule system was also manipulated with nocodazole, vinblastine, or colchicine at concentrations that depolymerized microtubules incompletely. Under these conditions, microtubules remained attached to the centrosome but often failed to extend all the way to the cell margin. As in taxol, the cell extended beyond the region of cytoplasm containing microtubules (data not shown). Changes in cell shape caused by partial microtubule depolymerization were less pronounced than when microtubules were completely depolymerized from the cell. The effect of low doses of nocodazole and other inhibitors on cell spreading and shape corroborates the idea that the extension of microtubules to the edge of the cell was not necessary to maintain a circular profile. 0.7- U LL 0.6(D 5 0.550.50.45 I 1 I Effect of Microtubules on Pseudopod Dynamics We next measured the effects of microtubule inhibitors on pseudopod dynamics. Cells spreading in nocodazole or taxol were examined by time-lapse video microscopy. Within 5 min of attachment, control cells spread symmetrically into discoidal shapes (Fig. 5). Once cells spread to their maximum area, they generally held their position and did not move much from their original landing point. Instead, cell margins began actively extending and retracting, such that the overall dynamics were balanced and cells maintained a relatively constant area and shape. Despite this difference, spreading was similar in control and nocodazole-treated cells. The rate of spreading at half the maximal area, and the total extent of spreading, calculated for five control cells and five nocodazole-treated cells, were not significantly different (data not shown). 236 Rosania and Swanson Fig. 3. Changes in cell shape in taxol are independent of microtubule distribution. Cells were pre-treated with drugs, then were allowed to spread for 60 min in buffer (A,B), 10 pM taxol (C,D), 20 pM no- codazole (E,F), or 20 pM nocodazole plus 10 pM taxol (G,H). Cells were stained with Texas Red (A,C,E,G) and with anti-tubulin immunofluorescence (B,D,F,H). Scale bar = 10 pm. Microtubules Modulate Pseudopod Activity 237 B 1 4 4T ' Con ' T Tax ' Noc 'Noc+Tai ' Con Tax ' Noc ' 'Noc+Tax D C 0.657 0.7.- 2 0.65- U 2 0.6- -.!2 0.55- 2 0.5- -0 0.6 Y 4-8 2 ~ 0.55 0) a 2 v, 0.5 0.45 0.4 0.45- Con Tax ' Con Noc 'Noc+Tax ' Tax ' Noc 'Noc+Tax Fig. 4. In the presence of microtubules, the morphology of taxoltreated cells closely resembled that of control cells. Cells were pretreated with drugs, then were allowed to spread in 10 p M taxol (Tax), 20 pM nocodazole (Noc), or 20 pM nocodazole plus 10 pM taxol (Noc Tax). "Con" indicates cells in buffer only. Morphological measurements indicate cell area (A), length (B),radial ratio (C), and shape factor (D) after 60 min of spreading. Each data point is cumulative from two experiments, n > 50. Differences between control and nocodazole taxol are significant (P < .001, t-test). Loss of radial symmetry in nocodazole began with a retraction at one or more sites along the cell margin (Fig. 6). As these retractions occurred, a new pseudopod extended in a different direction. With this exaggerated retraction and extension, cell shape became progressively asymmetric. Frequently, parts of the retracting cell remained attached, with thin strips of cytoplasm connecting the attachment sites to the cell body. Sometimes they broke off, leaving fragments of cytoplasm on the cover slip. Cell spreading in taxol was very similar to control cells (Fig. 7). Once the cells spread to their maximum size, pseudopod activity remained confined to the cell margin. With time, taxol-treated cells drifted from their original landing point, indicating that although extension and retraction seemed to be quantitatively balanced, they might be uncoordinated. Moreover, while control cells ruffled mostly at their margins, taxol-treated cells ruffled mostly in the center of the cell. These observations indicate that the spatial arrangement of the microtubules was not critical for maintaining cell shape, although their arrangement may have determined where ruffling occurred. We next examined the dynamics at the cell margin by measuring the areas covered by extension and retraction. Two cell profiles, taken 5 min apart, were compared morphometrically (Fig. 8). Relative to controls, the areas covered by extension and retraction increased in cells treated with nocodazole. Taxol, on the other hand, did not affect extension or retraction, or the net balance of pseudopod dynamics. In low concentrations of nocodazole, extension and retraction were intermediate between controls and 20 pM nocodazoletreated cells. However, the net change of cell area was similar to controls, indicating that whenever cells contained microtubules, extension matched retraction. Thus, microtubules suppressed and balanced extension and retraction such that cell shape and area remained constant. + + Role of Other Cytoskeletal Elements To examine F-actin, spread cells were fixed and stained with NBD-phallicidin, a fluorescent probe for F-actin (Fig. 9). As previously reported [Amato et al., 19831, F-actin in control cells accumulated in discrete, 238 Rosania and Swanson Fig. 5 . Time-lapse sequence of spreading in control cells. The initial spreading was symmetrical. Thereafter, pseudopod extension and retraction were restricted to the cell margin, and were balanced such that cell shape remained roughly circular. Labels indicate minutes after the beginning of spreading. Scale bar = 10 pm. Fig. 6. Time-lapse sequence of a nocodazole-treated cell. The initial spreading was radially symmetric, but later the cell margin extended and retracted at several points. During retraction, fragments of cyto- plasm still attached to the cover slip were left behind. Active extension eventually led to changes in cell shape. Labels indicate minutes after the beginning of spreading. Scale bar = 10 pm. punctate foci. F-actin in taxol-treated cells resembled controls, with slightly more labeling in the center of the cell. In nocodazole-treated cells, most of the F-actin was concentrated in the cell body, and actin-rich foci were absent from the cell margin. We tested whether the effect of nocodazole on cell shape could be blocked by cytochalasin D (Fig. 10). Control and nocodazole-treated cells were allowed to attach to coverslips at 4°C and then warmed to 37°C for 5 min. During this interval, cells spread symmetrically to about 50% of their maximum area. Cytochalasin D was then added and cells were left for 60 min. Cytochalasin Microtubules Modulate Pseudopod Activity 239 Fig. 7. Time-lapse sequence of taxol-treated cell. The initial spreading was symmetrical and protrusive activity along the cell margin is visibly reduced. After some time, the cytoplasmic mass tended to drift. Labels indicate minutes after the beginning of spreading. Scale bar = 10 pm. I: I -400 Con I , , 20 plvl Noc 20 plvl Tax 0.2 pM Noc Fig. 8. The effect of microtubule-specific drugs on pseudopod dynamics. Bars represent the average area covered by extension (hatched bars) or retraction (black bars) by individual cells in a 5-min time interval. Untreated cells (Con) or cells pre-treated with different drugs (20 pM nocodazole, 20 pM Noc; 20 p M taxol, 20 pM Tax; 0.2 pM nocodazole, 0.2 p M Noc) were allowed to spread in the different conditions for 30 min before the images were collected. Cumulative data from two experiments, n = 8; error bars represent standard error of the mean. D inhibited further spreading by control cells, but did not lead to changes in the shape factor or radial ratio. In nocodazole-treated cells, cytochalasin D inhibited further cell spreading and the nocodazole-related shape changes: cells looked radially symmetric, and their shape factor and radial ratio reflected this (Fig. 10). The organization of F-actin was also compared in control or nocodazole-treated cells that were left to spread for 5 min and then incubated with cytochalasin D for 60 min. Under these conditions, the F-actin staining of nocodazole-treated cells resembled that of control cells (Fig. 9). Thus, changes in cell shape induced by microtubule depolymerization correlated with changes in the pattern of F-actin staining, consistent with the idea that microtubules influenced cell shape by affecting the dynamics or spatial organization of the actin cytoskeleton. The inhibition of shape changes by cytochalasin D indicated that the nocodazole-induced cell shape changes depended on the organization and dynamics of the actin cytoskeleton. We also examined the distribution of the intermediate filament protein vimentin. Cells were allowed to land and spread on BSA-coated cover slips for 60 min, then were fixed, permeabilized, and stained. In all cells, vimentin filaments were restricted into a tightly tangled perinuclear mass (Fig. 11). Although filaments in nocodazole were more tightly packed than in taxol or controls, the periphery of the cell was completely free of filamentous vimentin in every condition examined. This architecture of intermediate filaments seemed to be peculiar to cells under the conditions used in this experiment, since macrophages plated overnight on cover slips display filamentous vimentin staining throughout the cytoplasm [Swanson et al., 19921. Because of their restricted distribution in the present circumstances, intermediate filaments did not appear to contribute directly to the maintenance of cell shape. 240 Rosania and Swanson Fig. 9. The effect of different drugs on F-actin organization in macrophages. Cells spread for 60 min in Ringer’s (A) or Ringer’s plus 20 pM nocodazole (B), 20 p M taxol (C), or in Ringer’s for 5 min and then in 10 pM cytochalasin D for 60 minutes (D).Cells were stained for F-actin with NBD-phallicidin. Scale bar = 10 pm DISCUSSION Although microtubules are not necessary for macrophages to spread into radially symmetric circles, they do contribute to the maintenance of that shape. Without microtubules, cells spread into circles, then exaggerated extension and retraction change their shape. Microtubule depolymerization stimulates pseudopod activity at the cell margin. Conversely, the presence of intact microtubules, anywhere in the cell, inhibits that activity. Here, microtubules do not serve as part of a structural matrix that maintains cell shape. Instead, these observations are consistent with a model in which microtubules modulate the level of a chemical intermediate that regulates the actin cytoskeleton. Spatial Arrangement of Microtubules in Relation to Pseudopod Dynamics and Cell Shape in Macrophages The stabilizing effect of microtubules on cell shape and pseudopod dynamics is independent of microtubule distribution. As in fibroblasts [de Brabander et al., 1981; Schiff and Honvitz, 19801, microtubules of taxol-treated macrophages polymerized independently of the centrosome, such that the radial geometry of the microtubule array was disrupted. Detached microtubules formed bundles around the nucleus and were generally removed from the periphery of the cell. In other respects, however, taxol-treated macrophages resembled control cells. Macrophages allowed to spread in nocodazole plus taxol acquired irregular shapes, indicating that taxol did not stabilize cell shape by a microtubule-independent mechanism. Accordingly, the inhibitory effect of microtubules on cell shape and pseudopod dynamics in macrophages depended on the polymerization state of microtubules, but not on the exact position of microtubules in the cell. Thus, microtubules appear to act at a distance from the cell edge, inhibiting movements without necessarily extending out to the edge. Interaction of Microtubules With Other Cytoskeletal Systems The effects of microtubules on cell shape were mediated by the actin cytoskeleton. In untreated and taxoltreated macrophages , F-actin accumulated in discrete, Microtubules Modulate Pseudopod Activity A B 14007 1300 N 1200 1100 ; lo00 45 1 h E 5 5. 6 6 900 800 700 J D C ‘- :-.3 2 35 30 600 .0 40 v 9, 2 241 0.8 0.75 0.7 0.65 F. 5 2 % m 0.6 0.55 0.5 0.45 5 0.8 0.75 0.7 0.65 0.6 0.55 0.5 0.45 Con Cyt Noc Cyt+Noc Fig. 10. Cell spreading and nocodazole-induced changes in cell shape are dependent on the actin cytoskeleton. Cells were suspended in Ringer’s (Con, Cyt) or Ringer’s plus 20 JLMnocodazole (Noc, Cyt + Noc) and allowed to spread for 5 min. Cytochalasin D was then added to cells in Ringer’s (Cyt) or nocodazole (Noc) and cells were allowed to spread for 60 min. Morphological measurements were made to assay changes in cell area (A), length (B), radial ratio (C), and shape factor (D). Each data point is cumulative from two experiments, n > 50. Differences in radial ratio and shape factor between Noc and Cyt Noc are significant ( P < ,001; t-test). symmetrically distributed foci adjacent to the plasma membrane, consistent with previous reports of microfilament structure in these cells [Amato et al., 1983; Yin and Hartwig, 19881. In nocodazole-treated cells, F-actin staining in the pseudopod was generally diffuse and asymmetric, with most of the F-actin accumulating in the cell body. In addition, cytochalasin D, which blocks the polymerization of actin by capping actin filaments, inhibited cell spreading, as expected, but did not cause cells to lose their circular profiles. Nevertheless, cytochalasin D inhibited the changes in cell shape induced by nocodazole, indicating that the effect of microtubule depolymerization on cell shape was mediated by the actin cytoskeleton. Actin-rich foci associated with the macrophage plasma membrane are likely to be involved in pseudopod extension, attachment, and the establishment of cell shape, by analogy with other systems [Burridge et al., 1988; Clark and Brugge, 1995; Okabe and Hirokawa, 1989; Symons and Mitchison, 19911. Under the conditions used in this study, vimentin filaments were absent from the cell margin, making it unlikely that they are directly involved in the changes in protrusive activity, cell area, or shape. Thus, microfilaments are responsible for the changes in pseudopod dynamics and cell morphology observed when microtubules are depolymerized . + Microtubules as Regulators of lntracellular Signaling Based on the inhibitory effects of microtubules on pseudopod dynamics, it is possible that microtubules suppress some intracellular signal which would otherwise lead to exaggerated protrusive activity. In macrophages, as well as in other cell types, pseudopod dynamics are under the control of signal transduction pathways [Hall, 19941. Ligation of the M-CSF receptor triggers a signaling pathway leading to pronounced ruffling at the cell margins [Racoosin and Swanson, 19921. Microinjection of ras protein into fibroblasts [Bar Sagi and Feramisco, 19861 or activation of protein kinase C in macrophages [Swanson, 19891 also increases pseudopod dynamics. Microtubule depolymerization can activate protein kinases [Manie et al., 1993; Shinohara-Gotoh et a]., 19911 and transcription factors [Rosette and Karin, 19951; it can affect mRNA levels in the cell and stimulate DNA synthesis [Pavlath et al., 19931. How could microtubules affect signal transduction 242 Rosania and Swanson Fig. 11. Intermediate filaments are not present in the pseudopod and are restricted to the cell body. Cells spread for 60 min in Ringer’s (A,B) or Ringer’s plus 20 p M nocodazole (C,D) or 20 p M taxol (E,F) were stained with Texas Red (A,C,E) and with anti-vimentin immunofluorescence (B,D,F). Scale bar = 10 Ilm. pathways? One possibility is that intact microtubules, but not soluble tubulin, bind signaling proteins like protein kinase A [Glantz et al., 19931 or protein kinase C [Kiley and Parker, 19951, and allosterically affect their enzymatic activity. One candidate second messenger, known to affect both actin dynamics and pseudopod extension, is calcium. Calcium stimulates myosin-dependent contractile activity and affects a variety of actin-associated proteins including gelsolin, cap Z, L-plastin, and fodrin [Maxfield, 19933. Treatments that raise intracellular calcium interfere with cell shape, cell migration, and chemotaxis [Gilbert et al., 19941. Interestingly, mitochondria and the endoplasmic reticulum both serve as intracellular calcium stores and are also typically associated with microtubules [Vale, 19871. Microtubule depolymerization could affect the signal transduction mechanisms regulating intracellular calcium levels or the ability of cellular organelles like mitochondria to sequester intracellular calcium. Another way microtubules might influence signal transduction pathways controlling the actin cytoskeleton is by affecting focal contacts [Rinnerthaler et al., 19881. Microtubules Modulate Pseudopod Activity 243 Focal contacts are complexes of proteins at sites of cell normal segregation of the cell margin into active and attachment [Bumdge et al., 19881, and are thought to inactive zones [Gail and Boone, 1971; Vasiliev et al., transduce signals that organize actin [Clark and Brugge, 19701. In fibroblasts, microtubules are necessary to sus19951. Focal contact proteins include a variety of phos- tain high levels of protrusive activity seen at the leading phoproteins that interact with the actin directly, such as edge of polarized cells [Bershadsky et al., 19911, alfimbrin, protein kinases, and other proteins [Clark and though as in macrophages, microtubule depolymerizaBrugge, 1995; Messier et al., 1993; Miyamoto et al., tion in fibroblasts induces all margins to become active 19951. The association of actin filaments with focal con- [Gail and Boone, 1971; Vasiliev et al., 19701. Taxoltacts stabilizes cell adhesions and permits continued treatment affects cell shape in fibroblasts, causing cells pseudopod extension [Miyamoto et al., 19951. Changes to lose protrusive activity at the leading edge and to in cytoskeletal tension induced by microtubule depoly- increase protrusive activity at the cell margin [Pletjushmerization might destabilize cell attachments and focal kina et al., 19941. The presence of actin stress fibers in contacts, and influence the associated signaling events fibroblasts and the absence of intermediate filaments in affecting the recruitment of actin filaments to the sites of the macrophage pseudopod may be responsible for difadhesion. ferences between these two cell types. Microtubules as Orchestrators of Localized Functions In addition to the global inhibitory effect on protrusive activity described in this study, microtubules have been suggested to stimulate pseudopod activity locally by transiently interacting and stabilizing structures whose assembly is rate-limiting for translocation, by polarizing the Golgi near sites of protrusive activity, and by serving as tracks mediating the net transport of vesicles towards the leading edge of the cell [Bershadsky and Vasiliev, 19931. Recent evidence indicates that localized stimulatory activity is dependent on kinesin-based transport [Rodionov et al., 19931 and microtubule dynamics [Tanaka et al., 1995; Liao et al., 19951. Nevertheless, microtubules could also send diffusible signals that localize cellular morphogenetic processes to specific sites in the cell. For example, to position the cleavage furrow [Fishkind and Wang, 19951, microtubules may provide a localized positive signal that stimulates the contractile activity along the cleavage plane of the cell, and a diffuse negative signal that prevents cortical contraction from occumng elsewhere in the cell. Microtubules might also influence localized cellular functions by mechanically exerting anisotropic effects on cytoskeletal tension. For example, to promote axonal growth in a particular direction [Tanaka et al., 19951, microtubules could bundle in parallel, such that they are better able to resist compressive forces in one direction. Mechanically, microtubules could guide the recruitment and polymerization of actin filaments in the growth cone by preferentially stabilizing focal contacts in the direction of alignment. Cell Type-Specific Differences of Microtubule Function In contrast to their role in macrophages, microtubules in fibroblasts help establish and maintain cell asymmetry. Microtubule depolymerization induces elongated cells to become more circular, and inhibits the ACKNOWLEDGMENTS The authors thank Paul Janmey, Claude Lechene, Rong Li, and Yu-Li Wang for insightful discussions, and George McNamara and Melissa Johnson for help with the image processing system. This study was supported by NIH grant A135950 to J.S. REFERENCES Amato, P.A., Unanue, E.R., and Taylor, D.L. (1983): Distribution of actin in spreading macrophages: A comparative study of living and fixed cells. J. Cell Biol. 96:750-761. Axline, S.G., and Reaven, E.P. (1974): Inhibition of phagocytosis and plasma membrane mobility of the cultivated macrophage by cytochalasin B. J. Cell Biol. 62547-659. Bar Sagi, D., and Feramisco, J.R. (1986): Induction of membrane ruffling and fluid-phase pinocytosis in quiescent fibroblasts by ras proteins. Science 233:1061-1068. Bershadsky, A.D., and Vasiliev, J.M. (1993): Mechanisms of regulation of pseudopodial activity by the microtubule system. In Jones, G . , Wigley, C., Warn, R. (eds.) “Cell Behaviour: Adhesion and Motility.” Cambridge: The Company of Biologists Limited. Bershadsky, A.D., Vaisberg, E.A., and Vasiliev, J.M. (1991): Pseudopodial activity at the active edge of migrating fibroblasts is decreased after drug induced microtubule depolymerization. Cell Motil. Cytoskeleton 19:152-158. Bornens, M., Paintrand, M., and Celati, C. (1989): The cortical microfilament system of lymphoblasts displays periodic oscillatory activity in the absence of microtubules: Implications for cell polarity. J. Cell Biol. 109:1071-1083. Bray, D., and White, J.G. (1988): Cortical flow in animal cells. Science 239:883-888. Burridge, K., Fath, K., Kelly, T., Nuckolls, G . , and Turner, C. (1988): Focal adhesions: Transmembranejunctions between the extracellular matrix and the cytoskeleton. Annu. Rev. Cell Biol. 4:487-525. Cannon, G.J., and Swanson, J.A. (1992): The macrophage capacity for phagocytosis. J. Cell Sci. 101:907-913. Clark, E.A., and Brugge, J.S. (1995): Integrins and signal transduction pathways: The road taken. Science 268:233-239. Coates, T.D., Watts, R.G., Hartman, R., and Howard, T.H. (1992): 244 Rosania and Swanson Relationship of F-actin distribution to development of polar shape in human polymorphonuclear neutrophils. J. Cell Biol. 117:765-774. de Brabander, M., De Mey, J., Van de Veire, R., Aerts, F., and Geuns, G. (1977): Microtubules in mammalian cell shape and surface modulation: An alternative hypothesis. Cell Biol. Int. Rep. 1:453-461. de Brabander, M., Gevens, G., Nuydens, R., Willbrords, R., and De Mey, J. (1981): Taxol induces the assembly of free microtubules in living cells and blocks the organizing capacity of centrosomes and kinetochores. Proc. Natl. Acad. Sci. U.S.A. 78: 5608-5612. Dernell, T.J., Joshi, H.C., Steel, V.L., Buxbaum, R.E. and, Heidemann, S.R. (1988): Tension and compression in the cytoskeleton of PC-12 neurites 11: Quantitative measurements. J. Cell. Biol. 107:666-674. Domnina, L.V., Rovensky, J.A., Vasiliev, J.M., and Gelfand, I.M. (1985): The effect of microtubule destroying drugs on the spreading and shape of cultured epithelial cells. J. Cell Sci. 74:267-282. Dugina, V.B ., Alexandrova, A.Y., Lane, K., Bulanova, E., and Vasiliev, J.M. (1995): The role of the microtubular system in the cell response to HGF/SF. J. Cell Sci. 108:1659-1667. Fischer, G.W., Conrad, P.A., Dehiasio, R.L., and Taylor, D.L. (1988): Centripetal transport of cytoplasm, actin and the cell surface in lamellipodia in fibroblasts. Cell Motil. Cytoskeleton 11:235-247. Fishkind, D.J., and Wang, Y.-L. (1995): New horizons for cytokinesis. Cun. Opin. Cell Biol. 7:23-31. Gail, M.H., and Boone, C.W. (1971): Effect of colcemid on fibroblast motility. Exp. Cell Res. 65:221-227. Gilbert, S.H., Perry, K., and Fay, F.S. (1994): Mediation of chemoattractant-induced changes in [CaZ+Iiand cell shape, polarity and locomotion by InsP,, DAG, and protein kinase C in newt eosinophils. J. Cell Biol. 127:489-503. Glantz, S.B., Li, Y., and Rubin, C.S. (1993): Characterization of a distinct tethering and intracellular targeting domains in AKAP75, a protein that links CAMP-dependent protein kinase I1 beta to the cytoskeleton. J . Biol. Chem. 268:12796-12804. Glasgow, J.E., and Danielle, R.P. (1994): Role of microtubules in random cell migration: stabilization of cell polarity. Cell Motil. Cytoskeleton 27:88-96. Hall, A. (1994): Small GTP-binding proteins and the regulation of the actin cytoskeleton. Annu. Rev. Cell Biol. 10:31-54. Heath, J.P., and Hollifield, B.F. (1993): On the mechanism of cortical actin flow and its role in cytoskeletal organisation of fibroblasts. In Jones, G., Wigley, C., Warn, R. (eds.): “Cell Behaviour: Adhesion and Motility. ” Cambridge. The Company of Biologists Limited, pp. 35-56. Ingber, D.E. (1993): Cellular tensegrity: Defining new rules of hiological design that govern the cytoskeleton. J. Cell Sci. 104: 6 13-627. Janmey, P.A. (1991): Mechanical properties of cytoskeletal polymers. CUR. Opin. Cell Biol. 2:4-11. Joshi, H.C., Chu, D., Buxbaum, R.E., and Heidemann, S.R. (1985): Tension and compression in the cytoskeleton of PC-12 neurites. J. Cell Biol. 101:697-705. Kiley, S.C., and Parker, P.J. (1995): Differential localization of protein kinase C isozymes in U937 cells: evidence for distinct isozyme functions during monocyte differentiation. J. Cell Sci. 108:1003-1 016. Kupfer, A., Kronebusch, P. J., Rose, J.K., and Singer, S.J. (1987): A critical role for the polarization of membrane recycling in cell motility. Cell Motil. Cytoskeleton 8: 182-189. Liao, G., Nagasaki, T., and Gundersen, G.G. (1995): Low concentrations of nocodazole interfere with fibroblast locomotion without significantly affecting microtubule level: Implications for the role of dynamic microtubules in cell locomotion. J. Cell Sci. 108:3473-3483. Manie, S . , Schmid-Alliana, A,, Kubar, J., Ferva, B., and Rossi, B. (1993): Disruption of the microtubule network in human monocytes induces expression of interleukin-1 but not of interleukin-6 nor tumor necrosis factor-alpha. Involvement of protein kinase A stimulation. J. Biol. Chem. 268:13675-13681. Maxfield, F.R. (1993): Regulation of leukocyte locomotion by C a Z f . Trends Cell Biol. 3:386-391. Messier, J.M., Shaw, L.M., Chafel, M., Matsudaira, P., and Mercurio, A.M. (1993): Fimbrin localized to an insoluble cytoskeleta1 fraction is constitutively phosphorylated on its headpiece domain in adherent macrophages. Cell Motil. Cytoskeleton 25: 223-233. Middleton, C.A., Brown, A.F., Brown, R.M., and Roberts, D.J.H. (1988): The shape of cultured epithelial cells does not depend on the integrity of their microtubules. J. Cell Sci. 91:337-345. Middleton, C.A., Brown, A.F., Brown, R.M., Karavanova I.D., Roberts, D.J.H., and Vasiliev, J.M. (1989): The polarization of fibroblasts in early primary cultures is independent of microtubule integrity. J. Cell Sci. 94:25-32. Miyamoto, S . , Teramoto, H., Coso, O.A., Gutkind, J.S., Burbelo, P.D., Akiyama, S.K., and Yamada, K.M. (1995): Integrin function: molecular hierarchies of cytoskeletal and signaling molecules. J. Cell Biol. 131:791-805. Okabe, S . , and Hirokawa, N. (1989): Incorporation and turnover of biotin-labeled actin microinjected into fibroblastic cells: An immunoelectron microscopic study. J. Cell Biol. 109:158I1595. Pavlath, G.K., Shimizu, Y., and Shimizu, N. (1993): Cytoskeletal active drugs modulate signal transduction in the protein kinase C pathway. Cell Struct. Funct. 18:151-160. Pletjushkina, O.J., Ivanova, O.J., Kaverian, I.N., and Vasiliev, J.M. (1 994): Taxol-treated fibroblasts acquire an epithelioid shape and a circular pattern of actin bundles. Exp. Cell Res. 212: 201-208. Racoosin, E.L., and Swanson, J.A. (1992): M-CSF-induced macropinocytosis increases solute endocytosis but not receptor-mediated endocytosis in mouse macrophages. J. Cell Sci. 102: 867-880. Rinnerthaler, G., Geiger, B., and Small, J.V. (1988): Contact formation during fibroblast locomotion: Involvement of membrane ruffles and microtubules. J. Cell Biol. 106:747-760. Rodionov, V.I., Gyoeva, F.K., Tanaka, E., Bershadsky, A.D., Vasiliev, J.M., and Gelfand, V.1. (1993): Microtubule-dependent control of cell shape and pseudopodial activity is inhibited by the antibody to kinesin motor domain. I. Cell. Biol. 123: 18111820. Rosette, C., and Karin, M. (1995): Cytoskeletal control of gene expression: Depolymerization of microtubules activates NF-kB. J. Cell Biol. 128:1111-1119. Schiff, P.B., and Horwitz, S.B. (1980): Taxol stabilizes microtubules in mouse fibroblast cells. Proc. Natl. Acad. Sci. U.S.A. 77: 1561-1565. Schliwa, M., and Honer, B. (1993): Microtubules, centrosomes and intermediate filaments in directed cell movement. Trends Cell Biol. 3:377-380. Shinohara-Gotoh, Y., Nishida, E., Hoshi, M., and Sakai, H. (1991): Activation of microtubule-associated protein kinase by microtubule disruption in quiescent rat 3YI cells. Exp. Cell Res. 193:161-166. Microtubules Modulate Pseudopod Activity Singer, S.J., and Kupfer, A. (1986): The directed migration of eukaryotic cells. Annu. Rev. Cell Biol. 2:337-365. Solomon, F., and Magendantz, M. (1981): Cytochalasin separates microtubule disassembly from loss of asymmetric morphology. J. Cell Biol. 89:157-161. Swanson, J . 1989. Phorbol esters stimulate macropinocytosis and solute flow through macrophages. J. Cell Sci. 94:135-142. Swanson, J.A., Locke, A., Ansel, P., and Hollenbeck, P.J. (1992): Radial movement of lysosomes along microtubule in permeabilized macrophages. J . Cell Sci. 103:201-209. Symons, M.H., and Mitchison, T.J. (1991): Control of actin polymerization in live and perrneabilized fibroblasts. J. Cell Biol. 114503-513. Tanaka, E., Ho, T., and Kirschner, M.W. (1995): The role of microtubule dynamics in growth cone motility and axonal growth. J . Cell Biol. 128:139-155. 245 Tomasek, J . J . , and Hay, E.D. (1984): Analysis of the role of microfilaments and microtubules on the acquisition of bipolarity and elongation of fibroblasts in hydrated collagen gels. J . Cell Biol. 99~536-549. Vale, R.D. (1987): Intracellular transport using microtubule-based motors. Annu. Rev. Cell Biol. 3:347-378. Vasiliev, J.M., Gelfand, I.M., Domnina, L.V., Ivanova, 0.Y., Komm, S.G. and Olshevskaya. L.V. (1970): Effect of colcemid on the locomotory behavior of fibroblasts. J . Embryol. Exp. Morphol. 24625-640. Wang, Y.-L. (1985): Exchange of actin subunits at the leading edge of living fibroblasts: Possible role of treadmilling. J. Cell Biol. 101:597-602. Yin, H.L., and Hartwig, J.H. (1988): The structure of the macrophage actin cytoskeleton. J . Cell Sci. Suppl. 9:169-184.