AMF-R Tubules Concentrate in a Pericentriolar Microtubule Domain After MSV Transformation of Epithelial MDCK Cells

Abstract

Autocrine motility factor receptor (AMF-R) is localized to an intracellular microtubule-associated membranous organelle, the AMF-R tubule. In well-spread untrans-formed MDCK epithelial cells, the microtubules originate from a broad perinuclear region and AMF-R tubules extend throughout the cytoplasm of the cells. In Moloney sarcoma virus (mos)-transformed MDCK (MSV-MDCK) cells, microtubules accumulate around the centrosome, forming a microtubule domain rich in stabilized detyrosinated microtubules. AMF-R tubules are quantitatively associated with this pericentriolar microtubule domain and the rough endoplasmic reticulum and lysosomes also co-distribute with the pericentriolar mass of microtubules. The Golgi apparatus is closely associated with the microtubule organizing center (MTOC) within the juxtanuclear mass of AMF-R tubules, and no co-localization of AMF-R tubules with the Golgi marker β-COP could be detected by confocal microscopy. After nocodazole treatment and washout, microtubule nucleation occurs exclusively at the centrosome of MSV-MDCK cells, and only after microtubule extension to the cell periphery does the microtubule cytoskeleton reorganize to generate the pericentriolar microtubule domain after 30–60 min. AMF-R tubules dispersed by nocodazole treatment concentrate in the pericentriolar region in parallel with the reorganization of the microtubule cytoskeleton. MSV transformation of epithelial MDCK cells results in the stabilization of a pericentriolar microtubule domain responsible for the concentration and polarized distribution of AMF-R tubules.

Keywords

autocrine motility factor receptor epithelial transformation microtubule cytoskeleton membrane tubule Madin–Darby canine kidney

THE POLARIZATION of the microtubule cytoskeleton, combined with directional microtubule-associated motors, regulates the cytoplasmic distribution and interactions of the membranous organelles of the cell (Kelly 1990; Kreis 1990; Cole and Lippincott–Schwartz 1995). Microtubule-associated tubular organelles include the tubular endosomes, tubular lysosomes, and the endoplasmic reticulum (ER) (Terasaki et al. 1986; Swanson et al. 1987; Tooze and Hollinshead 1992; Robinson et al. 1996). The receptor for autocrine motility factor (AMF-R) has been localized to a distinct microtubule-associated intracellular tubular organelle that does not contain the lysosomal marker LAMP-2 and does not co-localize with the calnexin-labeled rough ER. By electron microscopy, the majority of intracellular AMF-R labeling is localized to smooth membranous tubules. However, ribosome-studded tubules and half rough/half smooth tubules are also labeled, suggesting that AMF-R tubules are associated with the ER (Benlimame et al. 1995). As described for ER tubules labeled with the fluorescent dye DiOC6 and tubular lysosomes labeled by fluid-phase endocytosis (Terasaki et al. 1986; Swanson et al. 1987), disruption of the microtubule cytoskeleton resulted in the loss of the peripheral orientation of AMF-R tubules and the formation of short unoriented tubules and vesicles (Benlimame et al. 1995). After disruption of the actin cytoskeleton, AMF-R tubules extend along microtubules to the periphery of MDCK cells (Benlimame et al. 1995).

AMF-R mediates transduction of the AMF motility signal via a pertussis toxin-sensitive G-protein, inositol phosphate production, protein kinase C activation, and production of the lipoxygenase metabolite 12-HETE (Silletti and Raz 1996). In human bladder and colorectal tumors, AMF-R expression correlates with tumor progression and can therefore serve as a marker for tumor malignancy (Nakamori et al. 1994; Otto et al. 1994). Our characterization of the AMF-R tubule in the polarized MDCK epithelial cell (Benlimame et al. 1995) led us to study AMF-R expression after epithelial transformation in vitro in the highly invasive Moloney sarcoma virus-transformed MDCK (MSVMDCK) cell line (Behrens et al. 1989). Disruption of the polarized MDCK epithelial phenotype, loss of E-cadherin expression, and increased cellular motility are associated with significantly increased (eightfold) AMF-R expression (Simard and Nabi 1996). The inverse relationship of AMF-R and E-cadherin in MSVMDCK transformants in vitro concords with that observed in vivo in human bladder carcinomas (Otto et al. 1994). MSV transformation is due to the expression of the v-mos oncogene product, a serine/threonine kinase which associates with and phosphorylates tubulin (Zhou et al. 1991). Mos fusion protein injected into PtK1 cells localizes to the kinetochore and blocks mitosis. The effects of Mos were dependent on its kinase activity and resulted in abnormal spindle morphology (Wang et al. 1994).

The microtubule cytoskeleton of fibroblasts radiates from the centrally located microtubule organizing center (MTOC) to the cell periphery. The minus ends of the microtubules are associated with the centrosome, the site of microtubule nucleation, and the plus (growing) ends of the microtubules are located at the periphery of the cell. In polarized epithelial cells, microtubules are oriented vertically and extend from the apical (minus end) pole to the basal (plus end) pole of the cells (Gorbsky and Borisy 1985; Tucker et al. 1986; Achler et al. 1989; Bacallao et al. 1989). The vertically oriented microtubules of polarized MDCK cells are not organized by the apically located centrioles (Bré et al. 1987; Buendia et al. 1990) and the microtubules of spreading MDCK cells do not originate from a unique MTOC but rather from multiple microtubule nucleating sites (Bré et al. 1990). Microtubules of epithelial cells are further distinguished from those of fibroblasts by their increased stability (Pepperkok et al. 1990; Wadsworth and McGrail 1990; Shelden and Wadsworth 1993), and stable noncentrosomal microtubules of polarized MDCK cells contain detyrosinated tubulin (Bré et al. 1987). We show here that after MSV transformation of epithelial MDCK cells, the microtubule cytoskeleton reorganizes to generate a polarized pericentriolar microtubule domain around the centriole. The microtubule distribution in these transformed invasive epithelial cells is responsible for the pericentriolar concentration of AMF-R tubules.

Materials and Methods

Cells and Antibodies

MSV-MDCK (DoCl₁ cells; from ATCC, Rockville, MD) and MDCK II cells were grown in Dulbecco's minimum essential medium (DMEM) supplemented with 10% fetal calf serum (FCS), glutamine, nonessential amino acids, vitamins, penicillin, and streptomycin (Gibco; Burlington, Ontario, Canada). Cells were cultured at 37C in a humidified atmosphere with 5% CO₂.

Monoclonal antibodies against AMF-R (Nabi et al. 1990) and LAMP-2 (Nabi et al. 1991; Nabi and Rodriguez–Boulan 1993) were used in the form of concentrated hybridoma supernatant. The following rabbit polyclonal antibodies were kindly provided as indicated: to detyrosinated (Glu) tubulin (Gundersen et al. 1984) by Dr. Gregg Gundersen (Department of Anatomy and Cell Biology, Columbia University, New York, NY); to pericentrin (Doxsey et al. 1994) by Dr. Steve Doxsey (University of Massachussetts, Worcester, MA); to calnexin by Dr. John Bergeron (Department of Anatomy and Cell Biology, McGill University, Montreal, Quebec, Canada); and to β-COP (Donaldson et al. 1990) by Dr. Jennifer Lippincott–Schwartz (NICHD, Bethesda, MD). Mouse monoclonal antibody to α-tubulin was from ICN (Mississauga, Ontario, Canada). Secondary antibodies conjugated to either fluorescein or Texas Red were purchased from Jackson Immunoresearch Laboratories (West Grove, PA). The fluorescent antibodies were designated for use in multiple labeling studies and no interspecies crossreactivity was detected. To detect anti-AMF-R, secondary antibodies specific for the μ-chain of rat IgM were used. Except where otherwise indicated, all chemicals were purchased from ICN.

Immunofluorescence

Intracellular immunofluorescent labeling was essentially as previously described (Benlimame et al. 1995). Briefly, cells plated on glass coverslips were fixed by addition of pre-cooled (-80C) methanol/acetone (80%/20% v/v) directly to the coverslip and then placed at −20C for 15 min. The cells were extensively washed with PBS, pH 7.4, supplemented with 0.1 mM Ca²⁺ and 1 mM Mg²⁺ (PBS/CM), and then incubated in PBS/CM supplemented with 0.2% BSA as a blocking agent (PBS/CM/BSA). All other washings and incubations were done with PBS/CM/BSA. After labeling, the coverslips were mounted in Airvol (Air Products and Chemicals; Allentown, PA) and observed with a Zeiss x63 Plan Apochromat objective. For double labeling, species-specific secondary antibodies coupled to Texas Red and FITC were used and visualized with selective filters. No crossreactivity could be detected.

Confocal Microscopy

Confocal microscopy was performed with the x60 Nikon Plan Apochromat objective of a dual-channel BioRad 600 laser scanning confocal microscope equipped with a krypton/argon laser and the corresponding dichroic reflectors to distinguish fluorescein and Texas Red labeling. To generate a composite image of the complete cell depth for cells double labeled for AMF-R (Texas Red) and tubulin (fluorescein), Z-series of optical sections collected at 1-μ steps encompassing the major portion of the immunofluorescent label were projected using BioRad COMOS software, in which the most intense value for each pixel is presented. Merging of the fluorescein and Texas Red channels generated the green-red double labeling of AMF-R and tubulin. Confocal images were printed using a Polaroid TX 1500 video printer.

The degree of concentration of microtubule and AMF-R labeling within the pericentriolar region was quantified from images collected in one session from MSV-MDCK and MDCK cells double immunofluorescently labeled in parallel for α-tubulin and AMF-R. Cells were imaged under conditions of equivalent pinhole and black level settings, and the gain was adjusted for each image to ensure that the pixel values in each section were not maximal and therefore not saturating. Z-series of optical sections at 1-μ steps encompassing the majority of the immunofluorescently labeled portion of the cell (from two or three sections) were projected using the Comos summation function, in which intensity values for each pixel are averaged over the selected sections. The region of intense juxtanuclear microtubule labeling (about 10–20% of the total cell area) was circumscribed, as was the equivalent juxtanuclear region in the AMF-R image, and average pixel intensity values for the circumscribed region and the total cell area obtained. To discount the contribution of the nonspecific nuclear AMF-R labeling (Benlimame et al. 1995), the fluorescent signal and area of the nuclear region were subtracted from those of the total cell area. The ratio of average pixel intensity in the juxtanuclear region relative to the corresponding value for the total cell area is indicative of the proportion of total cellular microtubules and AMF-R tubules located within the defined juxtanuclear region. The juxtanuclear proportion of AMF-R relative to that of tubulin generates a semiquantitative indicator of the concentration of AMF-R tubules within this subdomain of the microtubule cytoskeleton in MDCK and MSV-MDCK cells. At least 25 cells were sampled per cell type.

Nocodazole Washout

MSV-MDCK cells were treated with 20 μ nocodazole in culture medium for 30 min at 37C, rinsed five times with warm DMEM, and then reincubated with prewarmed culture medium for the indicated periods of time at 37C. To visualize microtubule nucleation after nocodazole treatment, cells were fixed immediately after the five DMEM washes.

Results

Juxtanuclear Concentration of AMF-R Tubules in MSV-MDCK Cells

The relationship between AMF-R tubule distribution and the microtubule cytoskeleton in MDCK and MSVMDCK cells was assessed by double immunofluorescent microscopy for AMF-R (the red signal) and tubulin (the green signal) (Figure 1). The microtubules of MDCK cells have been previously described to extend to the cell periphery from a broad perinuclear region (Bré et al. 1987; Bacallao et al. 1989; Bré et al. 1990), similar to that shown here, and AMF-R tubules are located throughout the cytoplasm of MDCK cells (Figure 1A). In the MSV-MDCK cells, both AMF-R and tubulin labeling concentrate in the juxtanuclear region generating the distinctive yellow overlapping region adjacent to the red nucleus (Figure 1B). The punctate nonspecific labeling of the nucleus (in red) by anti-AMF-R antibodies has been previously described (Benlimame et al. 1995).

Quantification of the juxtanuclear AMF-R tubule distribution in MSV-MDCK cells was performed on composite images representing the average pixel intensity of confocal optical sections encompassing the complete cell height. Average pixel intensity of both tubulin and AMF-R labeling was determined in the juxtanuclear region defined by increased microtubule labeling, which ranged from 10 to 20% of the total cell area in both MDCK and MSV-MDCK cells, and in the complete cell. These values generate an indicator of the proportion of AMF-R labeling located in the juxtanuclear region of the cell relative to the distribution of tubulin (see Materials and Methods). Interpretation of data obtained by this semiquantitative approach must be performed with caution, and conclusions as to the differential microtubule concentration in the juxtanuclear region of MDCK or MSV-MDCK cells cannot be inferred. The distribution of microtubules in the two cell types does, however, serve as an internal control, and the proportion of AMF-R labeling in this juxtanuclear region, relative to that of tubulin, was significantly increased in MSV-MDCK cells relative to MDCK cells (Figure 1C). These results clearly demonstrate an increased density of AMF-R tubules associated with the juxtanuclear microtubule dense region of MSV-MDCK cells compared to MDCK cells.

To assess whether the formation of the pericentriolar microtubule domain is associated with microtubule stabilization, we labeled MDCK and MSV-MDCK cells with antibodies to detyrosinated tubulin (Gundersen et al. 1984). As previously described (Bré et al. 1987), stabilized detyrosinated noncentrosomal microtubules of MDCK cells extend throughout the cell (Figures 2A and 2B). In MSV-MDCK cells, detyrosinated microtubules are more concentrated in the pericentriolar region and do not extend to the cell periphery to the same extent as microtubules labeled for α-tubulin (Figures 2C and 2D). The pericentriolar microtubule domain of MSV-MDCK cells therefore contains a distinct population of microtubules, including an increased number of stable microtubules. Detyrosinated microtubules co-distribute with AMF-R tubules in the same pericentriolar region (Figures 2E and 2F). We were unable to identify co-alignment between AMF-R tubules and detyrosinated microtubules in the cell periphery.

Figure 1

Concentration of AMF-R tubules in the pericentriolar region in MSV-MDCK cells. MDCK (A) and MSV-MDCK (B) cells were double immunofluorescently labeled with primary antibodies to AMF-R and α-tubulin and Texas Red anti-rat IgM and FITC anti-mouse secondary antibodies and confocal sections collected at 1-μ steps. Z-series of optical sections encompassing the total cell height were projected to generate a composite image of the entire cell. Superposition of the Texas Red and FITC images generated the red-green images in which the red represents AMF-R labeling, the green tubulin labeling, and the yellow overlap of the two signals. Bar = 20 μ. (C) Quantification of the pericentriolar concentration of AMF-R in MDCK and MSV-MDCK cells. Using composite confocal images representing the total height of the cell, the ratio of average pixel intensity of AMF-R in the pericentriolar region compared to the total cell area relative to that of tubulin was determined in MDCK (white) and MSV-MDCK cells (black). Data are presented as the mean ± SEM (p<0.0001) and represent the analysis of more than 25 cells. For further details see Materials and Methods.

Optical sectioning by confocal microscopy of MSV-MDCK cells double immunofluorescently labeled for AMF-R and tubulin revealed the presence of a distinct focal MTOC within the juxtanuclear tubulin label of MSV-MDCK cells (Figure 3). Sections near the substrate revealed the extension of microtubules to the cell periphery from a central site adjacent to the cell nucleus (Figure 3B). Sections at a further distance of 0.8 μ from the substrate clearly show the site of origin of the microtubule fibers or the MTOC (Figure 3D). Confocal sections permit the visualization of distinct AMF-R labeled structures within the dense juxtanuclear AMF-R label which exhibit neither the linear elongation nor the peripheral orientation of AMF-R tubules located in the cell periphery (Figures 3A and 3C). In sections close to the substrate, AMF-R tubules are also visualized in peripheral regions of the cell, where their orientation towards the cell periphery is similar to that of the microtubules (Figures 3A and 3B). At increasing distances from the substrate, the AMF-R label is restricted to a region around the focal MTOC (Figure 3C). The dense juxtanuclear AMF-R labeling in MSV-MDCK cells (Figure 1B) is therefore a three-dimensional accumulation of AMF-R-labeled tubules and vesicles that co-distributes with microtubules to one side of the nucleus around the site of a focal MTOC in MSV-MDCK cells.

The Pericentriolar Microtubule Domain and the Distribution of Other Membrane Organelles

The localization of the Golgi apparatus at the MTOC is microtubule-dependent (Ho et al. 1989; Corthésy–Theulaz et al. 1992), and we addressed the relationship of the Golgi apparatus to AMF-R tubules localized in the pericentriolar region by double immunofluores-cent labeling and confocal microscopy. In MDCK cells, AMF-R tubules extend throughout the cell, in contrast to the β-COP-labeled Golgi, which is limited to the perinuclear region. By immunoelectron microscopy, specific labeling of the Golgi apparatus with anti-AMF-R antibodies was not detected (Benlimame et al. 1995). In MSV-MDCK cells, the Golgi is located within the same juxtanuclear cellular space as the AMF-R tubule concentration, but no overlap of the β-COP-labeled Golgi is observed with AMF-R-labeled tubules by confocal microscopy (Figures 4A and 4B). The area occupied by the juxtanuclear concentration of AMF-R tubules is larger than that occupied by the Golgi apparatus, and the two organelles intercalate such that they can coexist within the same cytoplasmic space.

Double immunofluorescent labeling of MSV-MDCK labels with antibodies to LAMP-2, a marker for lysosomes and late endosomes (Kornfeld and Mellman 1989), and to detyrosinated tubulin (Figures 5A and 5B) or with antibodies to calnexin, a marker for the rough ER (Hochstenbach et al. 1992), and to tubulin (Figures 5C and 5D) reveals that lysosomes and the rough ER also cluster within the pericentriolar microtubule domain. Lysosomes are also distributed to peripheral cellular regions and cellular extensions. The similar high degree of pericentriolar concentration of AMF-R tubules and rough ER in MSV-MDCK cells is consistent with the identification of AMF-R tubules as smooth tubules associated with the ER (Benlimame et al. 1995).

Figure 2

Detyrosinated microtubules codistribute with AMF-R tubules in the pericentriolar region. MDCK (A, B) and MSV-MDCK cells (C–F) were double immunofluorescently labeled with antibodies to detyrosinated tubulin (A, C, E) and antibodies to α-tubulin (B, D) or AMF-R (F). Bar = 20 μ.

Figure 3

Confocal microscopy of AMF-R tubules and microtubules in MSV-MDCK cells. MSV-MDCK cells were double labeled for AMF-R (A, C) and tubulin (B, D). Confocal optical sections of a labeled cell are presented near the substrate (A, B) and towards the apical pole of the cell (C, D). The central site of the MTOC from which microtubules radiate to the cell periphery is indicated by arrows (B, D). Distinct AMF-R vesicles and tubules are observed in the densely labeled pericentriolar region. Bar = 20 μ.

Concomitant Reorganization of the Microtubule Cytoskeleton and Redistribution of AMF-R Tubules to the Pericentriolar Region

Microtubules of MDCK cells are nucleated at acentriolar sites in spreading isolated cells (Bré et al. 1990), and we assessed whether microtubule nucleation occurs at the centrosome of MSV-MDCK cells. Reformation of the microtubule cytoskeleton was followed after microtubule disruption with 20 μ nocodazole and subsequent washout of the drug. After a 30-min treatment with 20 μ nocodazole at 37C, complete disruption of the microtubule cytoskeleton is observed (Figures 6A and 6B), but the centrosome, labeled with antibodies to pericentrin (Doxsey et al. 1994), remains intact. Immediately after washing (five washes with DMEM over a period of 3 min) to remove the nocodazole, budding microtubule asters are observed at the site of the centrosome (Figures 6C and 6D). Similar experiments in MDCK cells showed multiple noncentrosomal sites of microtubule tubulation after nocodazole washout (Bré et al. 1987), which were not observed here in MSV-MDCK cells. Continued washout of the nocodazole-treated cells for 5 min at 37C results in the extension of microtubules throughout the cytoplasm (Figures 6E and 6F). The centrosome can still be identified as the site of origin of microtubules, and the dense pericentriolar concentration of microtubules of untreated cells is not yet observed. Incubation at 37C for 15 min results in the reorganization of the microtubule cytoskeleton to generate dense microtubule bundles around the nucleus (Figures 6G and 6H). At 30–60 min after nocodazole washout, tubulin labeling is concentrated around the centrioles and microtubule labeling of fibers extending to the periphery is diminished, suggestive of the localized central stabilization of microtubules (Figures 6I and 6J). The dense juxtanuclear labeling of tubulin observed in MSV-MDCK cells is therefore due to the postnucleation reorganization of microtubules that concentrate in the pericentriolar region.

Figure 4

The paranuclear AMF-R tubule concentration does not colocalize with the Golgi apparatus. MSV-MDCK cells (A, B) were double immunofluorescently labeled for AMF-R (red) and β-COP (green). Confocal images are presented near the substrate (A) and at the apical pole of the cell at a distance of 1.6 μ (B). The β-COP-labeled Golgi apparatus appears to intercalate among the surrounding AMF-R tubules. Identical structures labeled for both AMF-R and β-COP were not observed. Bar = 20 μ.

Treatment of MSV-MDCK cells with nocodazole and loss of microtubule integrity resulted in the dispersion of AMF-R-labeled tubules throughout the cell (Figures 7A and 7B). Similar to the effect of nocodazole disruption on MDCK cells (Benlimame et al. 1995), in the absence of microtubules, AMF-R tubules exhibited a loss of linear extension and peripheral orientation (Figures 7A and 7B). After formation of a microtubule network 5 min after nocodazole washout, AMF-R tubules are oriented along microtubules towards the cell periphery but remain distributed throughout the cell (Figures 7C and 7D). Coordinately with the juxtanuclear reorganization of microtubules, 30 min after washout of nocodazole essentially all AMF-R tubules had concentrated at the pericentriolar site of microtubule concentration (Figures 7E and 7F). At 120 min after nocodazole washout, the microtubule cytoskeleton had acquired a distribution resembling that of control cells in which the majority of AMF-R tubules co-distributed with a dense microtubule concentration in the pericentriolar region (Figures 7G and 7H). Distribution of AMF-R tubules to the pericentriolar region is therefore dependent on the reorganization of the microtubule cytoskeleton.

Discussion

Epithelial to mesenchymal cell transitions occur during normal development but also following epithelial cell transformation to generate invasive, motile carcinoma cells (Birchmeier and Birchmeier 1993). Loss of E-cadherin expression or function is a major regulator of the loss of epithelial polarization and acquisition of invasive capability (Birchmeier et al. 1993). Decreased E-cadherin expression and loss of polarized epithelial cytoarchitecture in MSV-MDCK transformants (Simard and Nabi 1996) are shown here to be associated with the reorganization of the microtubule cytoskeleton and the concentration of the microtubule-associated organelle, the AMF-R tubule, in the pericentriolar region.

Microtubules in epithelial cells do not radiate from a central MTOC, as in motile fibroblasts, and micro-tubule nucleation occurs at sites other than the centrioles (Bacallao et al. 1989; Bré et al. 1990). In MSV-MDCK cells, a dense mass of microtubules accumulates to one side of the nucleus around the centrosome. By confocal microscopy, microtubules radiating from the MTOC can be visualized within the dense pericentriolar mass of tubulin labeling of MSV-MDCK cells. The pericentriolar microtubule concentration is a consequence of the dynamic reorganization of the microtubule cytoskeleton that occurs after microtubule nucleation and extension throughout the cytoplasm. The microtubules of epithelial cells are stabilized relative to fibroblasts (Pepperkok et al. 1990; Wadsworth and McGrail 1990; Shelden and Wadsworth 1993), and retention of mechanisms that stabilize microtubules after epithelial transformation may be responsible for the formation of the pericentriolar microtubule domain in MSV-MDCK cells. Although we cannot exclude a role for acentriolar microtubule nucleation in the formation and maintenance of the pericentriolar microtubule domain of MSV-MDCK cells, the concentration of stabilized detyrosinated microtubules clearly indicates that it constitutes a stabilized subdomain of the microtubule cytoskeleton of these cells.

Figure 5

Concentration of rough endoplasmic reticulum and lysosomes in the pericentriolar domain of MSV-MDCK cells. MSV-MDCK cells were double immunofluorescently labeled with mouse anti-LAMP-2 (A) and rabbit anti-detyrosinated tubulin (B) or with rabbit anti-calnexin (C) and mouse anti-tubulin (D). Bar = 20 μ.

Figure 6

Nucleation of microtubules at the centrosome after nocodazole washout. MSV-MDCK cells were treated with 20 μ nocodazole for 30 min (A, B), rinsed five times with DMEM over a period of 3 min (C, D), and then incubated in DMEM with FCS at 37C for 5 (E, F), 15 (G, H), or 60 (I, J) min. The cells were fixed and double immunofluorescently labeled for tubulin (A, C, E) and pericentrin (B, D, F). Microtubule asters form exclusively at the centrosomes. Bar = 20 μ.

In MDCK cells, AMF-R tubules are distributed throughout the cell, and the extension and peripheral orientation of AMF-R tubules are dependent on microtubule integrity (Benlimame et al. 1995). In MSV-MDCK cells, AMF-R tubules concentrate and co-distribute with the microtubules of the pericentriolar region. The pericentriolar concentration of AMF-R tubules is a consequence of microtubule reorganization, because after nocodazole treatment AMF-R tubules disperse and concentrate only in coordination with the redistribution of microtubules to the pericentriolar region. The juxtanuclear accumulation of microtubules in MSV-MDCK cells is associated with the increased concentration of a microtubule-associated organelle, the AMF-R tubule, in that same pericentriolar region. Quantification of confocally generated images reveals that the proportion of cellular AMF-R labeling localized to the microtubule dense juxtanuclear region is significantly greater in MSV-MDCK cells than in MDCK cells. The pericentriolar concentration of the AMF-R tubule in MSV-MDCK is therefore not simply a consequence of the reorganization of the microtubule cytoskeleton but instead reflects its specific association with the microtubules of the pericentriolar domain.

Other cellular organelles, including the Golgi apparatus, endosomes, and lysosomes, cluster in the pericentriolar region of the cell in a microtubule-dependent manner (Kreis 1990; Cole and Lippincott–Schwartz 1995). AMF-R and β-COP do not co-localize in confocal sections distinguishing AMF-R-labeled tubules and vesicles from the Golgi apparatus as well as from the ER–Golgi intermediate compartment (ERGIC), also labeled by β-COP in exocrine pancreas (Donaldson et al. 1990; Duden et al. 1991; Oprins et al. 1993). In contrast to the AMF-R tubule, the Golgi apparatus exhibits a perinuclear distribution in both MDCK and MSV-MDCK cells, and in MSV-MDCK cells the Golgi apparatus does not co-distribute with the pericentriolar microtubule domain but is situated within the pericentriolar mass of AMF-R tubules. The distribution of the Golgi apparatus and AMF-R tubules in the pericentriolar region is perhaps best visualized in the confocal image in Figures 3C and 3D. The location of the Golgi apparatus can be identified by the horseshoe-shaped absence of staining amid the pericentriolar concentration of AMF-R tubules (Figure 3C) and appears to circumscribe the focal site of the MTOC (Figure 3D). AMF-R tubules are distributed throughout the dense microtubule-labeled juxtanuclear region and cede space to allow the Golgi apparatus proximity to the focal MTOC.

The pericentriolar localization of the Golgi apparatus is due to the dynein-mediated movement of Golgi fragments to the minus ends of microtubules (Ho et al. 1989; Corthésy–Theulaz et al. 1992). The bidirectional movement of AMF-R tubules along microtubules between the cell center and periphery can be induced by modulation of cytoplasmic pH (Nabi et al. 1992), as described for tubule lysosomes in macrophages and fibroblasts and late endosomes in polarized MDCK cells (Heuser 1989; Parton et al. 1991). AMF-R tubules could therefore move along microtubules to concentrate in the pericentriolar region. However, the closer association of the Golgi apparatus of MSV-MDCK cells with the focal MTOC implies either a differential avidity for minus-end targeting or a differential basis for the association of these two organelles with microtubules in the pericentriolar region.

The microtubule association of AMF-R tubules resembles that of the ER and tubular lysosomes, whose extension to the cell periphery is microtubule-dependent (Terasaki et al. 1986; Swanson et al. 1987) and, in the case of the tubular lysosome, shown to be mediated by the plus-end-directed motor kinesin (Hollen-beck and Swanson 1990). In a similar fashion to AMF-R tubules, the rough ER labeled with calnexin extends throughout the cytoplasm of MDCK cells (Benlimame et al. 1995) and concentrates within the pericentriolar microtubule domain of MSV-MDCK cells. The distribution of both AMF-R tubules and the rough ER to this subdomain of the microtubule cytoskeleton is consistent with the described association of smooth AMF-R tubules with ribosome-studded elements of the ER (Benlimame et al. 1995). Lysosomes also cluster within the pericentriolar subdomain of MSV-MDCK cells, although not to the same qualitative extent as either AMF-R tubules or the rough ER.

The increased density of AMF-R labeling relative to that of tubulin in a defined pericentriolar region after MSV transformation of MDCK cells is indicative of the selective association of AMF-R tubules with a sub-population of microtubules in MSV-MDCK cells. Microtubule dynamics have been shown to drive the formation of ER membranes into tubular membrane networks (Waterman–Storer et al. 1995). Increased stabilization of microtubules of the pericentriolar region of MSV-MDCK cells, as reflected in the concentration of detyrosinated microtubules in this region, might be responsible for the polarized accumulation of AMF-R tubules and the rough ER in this defined cytoplasmic space.

Selective stabilization of localized regions of microtubules has been proposed to be implicated in cell polarization and morphogenesis (Kirschner and Mitchison 1986). Stabilized detyrosinated microtubules are oriented towards the direction of cell movement (Gundersen and Bulinski 1988). Microtubules are stabilized in vivo by interaction with MAPs (Takemura et al. 1992; Umeyama et al. 1993; Baas et al. 1994). EMAP-115 has been identified as an epithelial cell-specific MAP, and its redistribution after epithelial transformation might conceivably be implicated in the phenomena described here (Masson and Kreis 1993). The ability of the v-mos oncogene product to associate with and phosphorylate tubulin (Zhou et al. 1991) and to localize to the kinetochore and block mitosis (Wang et al. 1994) supports a role for Mos-mediated phosphorylation of tubulin or associated factors in the reorganization of the microtubule cytoskeleton and the specific association of AMF-R tubules with a select domain of microtubules in MSV-transformed MDCK cells.

Figure 7

Kinetics of AMF-R tubule association with the MTOC after nocodazole washout. MSV-MDCK-E cells were treated with 20 μ no-codazole for 30 min (A, B) and then rinsed extensively and incubated for 5 (C, D), 30 (E, F), or 120 (G, H) min at 37C in the absence of nocodazole. The cells were double immunofluorescently labeled for AMF-R (A, C, E, G) and α-tubulin (B, D, F, H). N, nucleus. Bar = 20 μ.

AMF-R mediates motility stimulation of tumor cells by the cytokine AMF (Liotta et al. 1986; Silletti et al. 1991), which stimulates directed pseudopodial extension (Guirguis et al. 1987) and integrin relocation to the cell surface (Timar et al. 1996). The presence of the AMF receptor within a distinct microtubule-associated organelle implies a motility-specific function for this organelle. A role for microtubule-directed vesicular targeting in pseudopodial extension and polarization of the motile cell (Singer and Kupfer 1986) requires the specific association of membrane vesicles with microtubule tracks targeted to the site of de novo lamellipod formation. The specific redistribution of the AMF-R tubule to a pericentriolar microtubule domain after epithelial transformation is indicative of the select association of AMF-R tubules with a subpopulation of microtubules in the motile, invasive MSV-MDCK cells.

Footnotes

Acknowledgements

Supported by grants from the National Cancer Institute of Canada with funds from the Canadian Cancer Society and from the Medical Research Council of Canada, and by an establishment award from the Fonds de la Recherche en Santé du Québec. DS was the recipient of a studentship from FCAR.

We thank Jennifer Lippincott–Schwartz, Steve Doxsey, Gregg Gundersen, and John Bergeron for kindly providing antibodies. We are particularly grateful to Nicole Leclerc for helpful suggestions concerning both the experimental work and the manuscript. We also thank Diane Gingras and Dale Laird for assistance with the confocal microscopy. The photographic reproductions were the work of Jean Léveillé.

References

Achler

Filmer

Merte

Drenckhahn

(1989) Role of microtubules in polarized delivery of apical membrane proteins to the brush border of the intestinal epithelium. J Cell Biol 109:179–189

Baas

Pienkowski

Cimbalnik

Toyama

Bakalis

Ahmad

Kosik

(1994) Tau confers drug stability but not cold stability to microtubules in living cells. J Cell Sci 107:135–143

Bacallao

Antony

Dotti

Karsenti

Stelzer

Simons

(1989) The subcellular organization of Madin-Darby canine kidney cells during the formation of a polarized epithelium. J Cell Biol 109:2817–2832

Behrens

Mareel

Van Roy

Birchmeier

(1989) Dissecting tumor cell invasion: epithelial cells acquire invasive properties after the loss of uvomorulin-mediated cell-cell adhesion. J Cell Biol 108:2435–2447

Benlimame

Simard

Nabi

(1995) Autocrine motility factor receptor is a marker for a distinct tubular membrane organelle. J Cell Biol 129:459–471

Birchmeier

(1993) Molecular aspects of mesenchymal-epithelial interactions. Annu Rev Cell Biol 9:511–540

Birchmeier

Weidner

Behrens

(1993) Molecular mechanisms leading to loss of differentiation and gain of invasiveness in epithelial cells. J Cell Sci 17(Suppl):159–164

Bré

M-H

Kreis

Karsenti

(1987) Control of microtubule nucleation and stability in Madin-Darby canine kidney cells: the occurence of noncentrosomal, stable detyrosinated microtubules. J Cell Biol 105:1283–1296

Bré

M-H

Pepperkok

Hill

Levilliers

Ansorge

Stelzer

Karsenti

(1990) Regulation of microtubule dynamics and nucleation during polarization in MDCK II cells. J Cell Biol 111:3013–3021

10.

Buendia

Bré

M-H

Griffiths

Karsenti

(1990) Cytoskeletal control of centrioles movement during the establishment of polarity in Madin-Darby canine kidney cells. J Cell Biol 110:1123–1135

11.

Cole

Lippincott–Schwartz

(1995) Organization of organelles and membrane traffic by microtubules. Curr Opin Cell Biol 7:55–64

12.

Corthésy-Theulaz

Pauloin

Pfeffer

(1992) Cytoplasmic dynein participates in the centrosomal localization of the Golgi complex. J Cell Biol 118:1333–1345

13.

Donaldson

Lippincott–Schwartz

Bloom

Kreis

Klausner

(1990) Dissociation of a 110-kD peripheral membrane protein from the Golgi apparatus is an early event in brefeldin A action. J Cell Biol 111:2295–2306

14.

Doxsey

Stein

Evans

Calarco

Kirschner

(1994) Pericentrin, a highly conserved centrosome protein involved in microtubule organization. Cell 76:639–650

15.

Duden

Griffiths

Frank

Argos

Kreis

(1991) β-COP, a 110-kD protein associated with non-clathrin-coated vesicles and the Golgi complex, shows homology to β-adaptin. Cell 64:649–665

16.

Gorbsky

Borisy

(1985) Microtubule distribution in cultured cells and intact tissues: improved immunolabeling through the use of reversible embedment cytochemistry. Proc Natl Acad Sci USA 82:6889–6893

17.

Guirguis

Margulies

Taraboletti

Schiffmann

Liotta

(1987) Cytokine-induced pseudopodial protrusion is coupled to tumour cell migration. Nature 329:261–263

18.

Gundersen

Bulinski

(1988) Selective stabilization of microtubules oriented toward the direction of cell migration. Proc Natl Acad Sci USA 85:5946–5940

19.

Gundersen

Kalnoski

Bulinski

(1984) Distinct populations of microtubules: tyrosinated and nontyrosinated alpha tubulin are distributed differently in vivo. Cell 38:779–789

20.

Heuser

(1989) Changes in lysosome shape and distribution correlated with changes in cytoplasmic pH. J Cell Biol 108:855–864

21.

Allan

van Meer

Berger

Kreis

(1989) Reclustering of scattered Golgi elements occurs along microtubules. Eur J Cell Biol 48:250–263

22.

Hochstenbach

David

Watkins

Brenner

(1992) Endoplasmic reticulum resident protein of 90 kilodaltons associates with the T- and B-cell antigen receptors and major histocompatibility complex antigens during their assembly. Proc Natl Acad Sci USA 89:4734–4738

23.

Hollenbeck

Swanson

(1990) Radial extension of macrophage tubular lysosomes supported by kinesin. Nature 346:864–866

24.

Kelly

(1990) Microtubules, membrane traffic, and cell organization. Cell 61:5–7

25.

Kirschner

Mitchison

(1986) Beyond self-assembly: from microtubules to morphogenesis. Cell 45:329–342

26.

Kornfeld

Mellman

(1989) The biogenesis of lysosomes. Annu Rev Cell Biol 5:483–525

27.

Kreis

(1990) Role of microtubules in the organisation of the Golgi apparatus. Cell Motil Cytoskel 15:67–70

28.

Liotta

Mandler

Murano

Katz

Gordon

Chiang

Schiffman

(1986) Tumor cell autocrine motility factor. Proc Natl Acad Sci USA 83:3302–3306

29.

Masson

Kreis

(1993) Identification and molecular characterization of E-MAP-115, a novel microtubule-associated protein predominantly expressed in epithelial cells. J Cell Biol 123:357–371

30.

Nabi

Le Bivic

Fambrough

Rodriguez–Boulan

(1991) An endogenous MDCK lysosomal membrane glycoprotein is targeted basolaterally before delivery to lysosomes. J Cell Biol 115:1573–1584

31.

Nabi

Rodriguez–Boulan

(1993) Increased LAMP-2 polylactosamine glycosylation is associated with its slower Golgi transit during establishment of a polarized MDCK epithelial monolayer. Mol Biol Cell 4:627–635

32.

Nabi

Watanabe

Raz

(1990) Identification of B16-F1 melanoma autocrine motility-like factor receptor. Cancer Res 50: 409–414

33.

Nabi

Watanabe

Raz

(1992) Autocrine motility factor and its receptor: role in cell locomotion and metastasis. Cancer Metast Rev 11:5–20

34.

Nakamori

Watanabe

Kameyama

Imaoka

Furukawa

Ishikawa

Sasaki

Kabuto

Raz

(1994) Expression of autocrine motility factor receptor in colorectal cancer as a predictor for disease recurrence. Cancer 74:1855–1862

35.

Oprins

Duden

Kreis

Geuze

Slot

(1993) β-COP localizes mainly to the cis-Golgi side in exocrine pancreas. J Cell Biol 121:49–59

36.

Otto

Birchmeier

Schmidt

Hinke

Schipper

Rübben

Raz

(1994) Inverse relation of E-cadherin and autocrine motility factor receptor expression as a prognostic factor in patients with bladder carcinomas. Cancer Res 54:3120–3123

37.

Parton

Dotti

Bacallao

Kurtz

Simons

Prydz

(1991) pH-induced microtubule-dependent redistribution of late endosomes in neuronal and epithelial cells. J Cell Biol 113:261–274

38.

Pepperkok

Bré

Davoust

Kreis

(1990) Microtubules are stabilized in confluent epithelial cells but not in fibroblasts. J Cell Biol 111:3003–3012

39.

Robinson

Chiplonkar

Luo

(1996) A method for co-localization of tubular lysosomes and microtubules in macrophages: fluorescence microscopy of individual cells. J Histochem Cy-tochem 44:1109–1114

40.

Shelden

Wadsworth

(1993) Observation and quantification of individual microtubule behavior in vivo: microtubule dynamics are cell-type specific. J Cell Biol 120:935–945

41.

Silletti

Raz

(1996) Regulation of autocrine motility factor receptor expression in tumor cell locomotion and metastasis. Curr Top Microbiol Immunol 213/II:137–169

42.

Silletti

Watanabe

Hogan

Nabi

Raz

(1991) Purification of B16-F1 melanoma autocrine motility factor and its receptor. Cancer Res 51:3301–3311

43.

Simard

Nabi

(1996) Inverse relation of autocrine motility factor receptor and E-cadherin expression following transformation of epithelial MDCK cells. Biochem Biophys Res Commun 219:122–127

44.

Singer

Kupfer

(1986) The directed migration of eukaryotic cells. Annu Rev Cell Biol 2:337–365

45.

Swanson

Bushnell

Silverstein

(1987) Tubular lysosome morphology and distribution within macrophages depends on the integrity of cytoplasmic microtubules. Proc Natl Acad Sci USA 84:1921–1925

46.

Takemura

Okabe

Umeyama

Kanai

Cowan

Hirokawa

(1992) Increased microtubule stability and alpha tubulin acetylation in cells transfected with microtubule-associated proteins MAP1B, MAP2 or tau. J Cell Sci 103:953–964

47.

Terasaki

Chen

Fujiwara

(1986) Microtubules and the endoplasmic reticulum are highly interdependent structures. J Cell Biol 103:1557–1568

48.

Timar

Trikha

Szekeres

Baza

Tovari

Silletti

Raz

Honn

(1996) Autocrine motility factor signals integrin-mediated metastatic melanoma cell adhesion and invasion. Cancer Res 56:1902–1908

49.

Tooze

Hollinshead

(1992) In AtT20 and HeLa cells brefeldin A induces the fusion of tubular endosomes and changes their distribution and some of their endocytic properties. J Cell Biol 118:813–830

50.

Tucker

Milner

Currie

Muir

Forrest

Spencer

M-J

(1986) Centrosomal microtubule-organizing centres and a switch in the control of protofilament number for cell surface-associated microtubules during Drosophila wing morphogenesis. Eur J Cell Biol 41:279–289

51.

Umeyama

Okabe

Kanai

Hirokawa

(1993) Dynamics of microtubules bundled by microtubule associated protein 2C (MAP2C). J Cell Biol 120:451–465

52.

Wadsworth

McGrail

(1990) Interphase microtubule dynamics are cell type-specific. J Cell Sci 95:23–32

53.

Wang

Yew

Peloquin

Vande Woude

Borisy

(1994) Mos oncogene product associates with kinetochores in mammalian cells and disrupts mitotic progression. Proc Natl Acad Sci USA 91:8329–8333

54.

Waterman–Storer

Gregory

Parsons

Salmon

(1995) Membrane/microtubule tip attachment complexes (TACs) allow the assembly dynamics of plus ends to push and pull membranes into tubulovesicular networks in interphase Xenopus egg extracts. J Cell Biol 130:1161–1169

55.

Zhou

Oskarsson

Paules

Schulz

Cleveland

Vande Woude

(1991) Ability of the c-mos product to associate with and phosphorylate tubulin. Science 251:671–675