TISSUE AND CELL, 1992 24 (3) 347-354 0 1992 Longman Group UK Ltd.
KLAUS WERNER WOLF and KATHARINA SPANEL-BOROWSKI*
THE INTERPHASE MICROTUBULE CYTOSKELETON OF FIVE DIFFERENT PHENOTYPES OF MICROVESSEL ENDOTHELIAL CELL CULTURES DERIVED FROM BOVINE CORPUS LUTEUM Keywords: Microvessel endothelial cells. cytoskeleton.
microtubules. cell culture
ABSTRACT. The interphase microtubule cytoskeleton of five different mlcrovesscl endothehal cell cultures. recently established from bovine corpus luteum, was analysed using anti-tuhulin immunofluorescence. An antibody against acetylated microtubules detected four cell type5 each of which possessed a single cilia. The length of the cilia were up to 10 pm for cell type, I and 2. Ciliary stubs had a length of up to 0.37,um in cell types 4 and 5. Cilia were missing in cell type 3. Long and short cilia were located in the perinuclear region from where cqtoplasmic microtubules radiated. Cell type 3 displayed straight microtubules rather than the wavy path seen in the other cell types. The amount of tyrosinated microtubules visuahzed by a spwfic antibody was consistently higher than that of posttranslationally acetylated microruhulcs. The latter were more apparent in cell types 4 and 5 than in the other cell types. WC conclude: Differences in the cytoplasmic microtubule inventory of each microvcssel cndothelial ccl1 type points at individual functions maintained in culture.
Introduction
different appearance of cell organelles. the development of cytoplasmic and axonemal microtubules (MTs) obviously varies among the five cell types. Cell type 1 displays a single cilium which is not detected in the other cell types. Cytoplasmic MTs are readily observed in cell types 3 to 5 whereas such MTs are not found in cell types 1 and 2. Since conventional ultrastructural analysis gives poor insight into the true architecture of MTs, immunofluorescence microscopy was applied to monolayers of the above described five different microvessel endothelial cell types. Monoclonal antibodies against two tubulin isotypes have been applied. These are (i) the YL l/2 antibody elicited against tyrosinated n+tubulin (Kilmartin et al., 1982) and (ii) the 6-llB-1 antibody raised against post-translationally acetylated actubulin (Piperno and Fuller, 1985). The YL l/2 antibody is believed to detect dynamic MTs, i.e. those with a high rate of exchange with the pool of tubulin subunits (Kreis, 1987). In contrast, MTs labeled with the 6-llB-1 antibody are less dynamic. In all likelihood,
Five different phenotypes of microvessel endothelial cells have been isolated from bovine corpus luteum and cultured for several months (Spanel-Borowski and von der Bosch, 1990). Cells are characterized by use of phase contrast microscopy, scanning electron microscopy as well as by immunocytochemistry. Cells of type 1 are arranged into an isomorphic monolayer of so-called cobble stone pattern apparent to a lesser degree in the polymorphic monolayer of cell type 2. Monolayers of cell types 3 to 5 are flat. These results have been recently confirmed and extended by ultrastructural findings (Spanel-Borowski, 1991). In addition to lnstitut fiir Biologic der Medizinischen Universitat zu Ltiheck. Ratzeburger Allee 160, D-2400 Liibeck, Germany. *Anatomischcs Institut der Umversit&, Pestalozistr. 20, CH-4056 Basel, Switzerland. Correspondence to Dr. Katharina Spanel-Borowski. Received 21 October 1991 Revised 11 December 1991 347
WOLF
348
acetylation is not the cause which renders MTs stable, but the consequence of slow MT turnover (Piperno et al., 1987; Webster and Borisy, 1989). Therefore, acetylation serves as a marker of stable MTs such as axonemal MTs of cilia. When density of acetylated cytoplasmic MTs is moderate, the 6-llB-1 antibody allows detection of cilia. Material and Methods Post-confluent cultures of microvessel endothelial cells maintained in 24-well culture plates were used. Phenotypes 1 to 5 were derived from bovine corpus luteum, isolated and cultured as described (Spanel-Borowski and van der Bosch, 1990). Cultures were subpassaged at a split ratio of 1:4 by use of 0.02% trypsin (cell types 1, 3 to 5) or 0.02% trypsin-0.02% EDTA (cell type 2), respectively. Cells were grown on cover slips coated with 1% collagen solution (Vitrogen@ Collagen Corporation, Palo Alto, California). At the time of processing, cell types 1 and 2 were still subconfluent, while cell types 3 to 5 appeared confluent due to higher cell proliferation. For anti-tubulin immunofluorescence, cover slips with cells adhering were treated with MT-stabilizing buffer (100 mM PIPES, 1 mM MgS04, 1 mM EGTA, pH 6.8) containing 1% Triton X-100. Details are described by Wolf and Bastmeyer (1991). After fixation with 2% paraformaldehyde and 0.25% glutaraldehyde in MT stabilizing buffer cytoskeletons were immersed in 100% methanol at -20°C. We used either rat monoclonal YL l/2 antibody (gift from Dr J. V. Kilmartin, MRC, Lab. Molec. Biol., Cambridge, UK), or mouse monoclonal 6llB-1 antibody (gift from Dr G. Piperno, Rockefeller Univ., NY, USA) as primary antibodies. Both antibodies were diluted 1:lOO with phosphate-buffered saline (PBS). As second antibodies, we applied biotin-conjugated anti-rat or anti-mouse antibodies (Sigma, Deisenhofen, Germany), diluted 1:50 with PBS, at room temperature for 1 hr. Specimens were further incubated with 1 mg polylysine (Serva, Heidelberg, Germany)/ ml PBS for 10 min, and with rhodaminecoupled avidine (Sigma), diluted 1:50 with PBS for 1 hr. In one experiment, the cytoplasmic MTs of cell type 5 cells were depolymerized with
AND SPANEL-BOROWSKI
a modified ice-cold lysis buffer (100 mM PIPES, pH 6.8,50 yM CaCIZ) containing 1% Triton X-100 (Kochanski and Borisy, 1990). For the detection of the remaining MTs, a monoclonal mouse anti-P tubulin antibody (Sigma) was diluted 1:lOO in PBS, and the specimens were incubated for 1 hr. All specimens were mounted in citrate buffer (100 mM citric acid, 200 mM Na2HP04, pH 7.0) containing 10 pg/ml 4’,6-diamidino-2phenylindole-2HC1, DAPI (Serva) in order to display chromatin arrangement. The preparations were documented using a Zeiss universal photomicroscope equipped with epifluorescence. Images were recorded on Tmax 400 (Kodak) developed in Rodinal (Agfa). Results Cell type 1 possessed an inconspicuous MT cytoskeleton. Tyrosinated MTs were few in number (Fig. la). As a rule, they were wavy and concentrated within the juxtanuclear area. Acetylated MTs were hardly to be discerned. Faintly fluorescent threads were detected by the 6-llB-1 antibody (Fig. lb). One intensely labeled thread 610pm in length projected from the surface of each cell. The threads represented single cilia having their base close to the nucleus. A few cells of larger size were observed in the monolayer of cell type 1. These cells were occasionally binucleate and are referred to as aging cells in the present text. Their MTs were less wavy than those in proliferating cells (Fig. lc, d). Tyrosinated MTs of aging cells occurred around the nucleus and a few of them extended towards the cell periphery. Acetylated MTs were lower in amount than tyrosinated MTs, yet acetylated MTs of aging cells were higher in number compared to acetylated MTs of proliferating cell type 1. Proliferating cell type 2 showed a denser MT network than cell type 1 (Fig. 2a, b). There was high MT density around the nuclei in addition to a large number of wavy MTs observed in the cell periphery. Single cilia showed a length of about 6 pm. In the vicinity of the ciliary base some weakly fluorescent threads of acetylated MTs were detected. Aging cells displayed a large number of tyrosinated MTs of apparently curved appearance (Fig. 2c, d). They radiated out from the perinuclear area and bended close to the
INTERPHASE
MICROTUBULE
Figs
l-5.
CYTOSKELETON
Microvessel endothelial cell types 1 to S derived from
are treated
with
Tyrosioated
a microtubule-stabilizing
microtubules
Fig.
la.
Fig.
b. Cell
type
(arrows).
Ic. d. Agmg
cell
single cilium
labeled
with
bovine
corpus
anti-tub&
la. c. 2a. c. 3a. 4a. Sa. whereas
luteum
antibodles.
Figs lb,
d, 2b. d.
1. Single
cilia
arc wiblc
(b)
using
I contains
acctylated
MTs
(d).
an antibody
against
acetylated
x 801).
type
cilium
cell type 2 Note
is indicated
and
microtuhulcs.
Fig. ?a. b. Cell type 2. A smglc Fig. 2c. d. Aging
buffer
arc seen in Figs
3h. 4b. Sb. c show acctylated
microtubule\
349
(arrow).
is marked
the large
x 300
(arrw).
size compared
x X00. x X00 to ccl1 type
Im
Fig. Za, b. The
INTERPHASE
MICROTUBULE
CYTOSKELETON
151
cell margin. Acetylated MTs were weakly developed in the cytoplasm. In cell type 3, tyrosinated MTs were abundant both in the vicinity of the nuclei as well as in the peripheral cytoplasm (Fig. 3a, b). The MTs were more or less of straight appearance and oriented parallel to the long axis of the cell. The density of acetylated MTs was low. Cilia were missing, Cell type 4 showed abundant tyrosinated MTs radiating from the perinuclear area (Fig. 4a, b). Though less dense, acetylated MTs were displaying a similar array as tyrosinated MTs. Due to the perinuclear abundancy of acetylated MTs, intensely fluorescent threads of approximately 0.3 pm length, presumably ciliary stubs, could be noticed only when the stubs projected over the perinuclear area. Cell type 5 possessed the highest density of tyrosinated and acetylated MTs of the 5 cell types (Fig. 5a-c). Tyrosinated MTs were slightly more concentrated in the juxtanuclear area, though a considerable amount of wavy MTs was also apparent in other parts of the cytoplasm. Highly fluorescent threads, probably ciliary stubs, could be seen only when the density of acetylated MTs was low in the surrounding cytoplasm. We had to verify in cell types 4 and 5 that the fluorescent threads were true cilia, thus, ruling out the presence of unusually long centrioles. Endothelial cells of type 5 were
treated with a MT destabilizing buffer prior to immunolabeling. The procedure caused depolymerization of cytoplasmic and ciliary MTs while centrioles were expected to remain stable (Kochanski and Borisy, 1990). There was a diffuse background, and the cytoplasmic inventory was lost. One pair of fluorescent spots remained apparent in interphase cells (Fig. 5d, e). These spots. smaller than the fluorescent threads of cells immunolabeled under MT stabilizing conditions, probably represented centrioles. Prophase cells showed two comparable spots being distant from one another (Fig. 5f, g). Therefore, cell type 5 and most probably also cell type 4, develop ciliary stubs. A semiquantitative evaluation of the appearance and amount of MTs is given in Table 1.
Discussion Single cilia
Using an antibody against acetylated MTs and comparing five different newly established microvessel endothelial cell cultures, the presence of single cilia or of ciliary stubs is revealed. The technique shows the superiority in detection of cilia compared to conventional technique by use of transmission electron microscope. By the latter, single cilia were found in cell type 1 only
Fig. 3a, b. Cell type 3. Tyrosinated microtubules are arranged parallel to the long axis of the cells (a). The density of acetylated microtubules is low in interphase cells (asterisks). but a mitotic spindle shows bright immunofluorescence (arrow-head). x 800. Fig. 4a, b. Cell type 4. In cells with low density of acetylated microtubules. detectable (arrow). x 800.
single cilia arc
Fig. 5a-c. Cell type 5. The cell shows the highest density of tyrosinated (a) and acetylated (b. c) microtubules. A ciliary stub is indicated (arrow). x 800. Fig. 5d. Cell type 5 of interphase cells. Prior to labeling with an antibody against Ptubulin. cells have been treated with a microtubule-destabilizing buffer. The cytoplasmic microtubules are depolymerized. Only small fluorescent spots, most probably centrioles (arrows) have remained. x 1250. Fig. Se. Same cells as in Fig. 5d. The nuclei are stained with DAPI, a DNA-specific fluorescent dye. Large fluorescent spots occur in the nuclear matrix. x 12.50. Fig. 5f. Cell type 5 is treated with a microtubule destabilizing buffer and labeled with an antibody against j%tubulin. Two fluorescent spots probably indicate centrioles (arrow). x 1250. Fig. 5g. Same cell as in Fig. 5f. The chromosomes
are condensed. DAPI staining. x 1250.
WOLF AND SPANEL-BOROWSKI
352 Table
1.
Cell type
Cytoplasmic tyrosinated
Microtubules acetylated
Shape
Cilia (length pm)
1
+,++
+
wavy
6-10
2 3 4 5
+++ +++ ++,+++ +++
+ + +,++ ++
wavy straight wavy wavy
about 6 absent 0.25-0.37 0.25-0.37
Semiquantitative evaluation: + not much, ++ moderate, +++ abundant
(Spanel-Borowski, 1991). The previous ultrastructural analysis obviously failed in visualizing short cilia of cell types 4 and 5, both growing as flat monolayers. Long cilia serve as a morphological marker of cell type 1. The length of them is comparable to that of PTKl cells (Jensen et al., 1979). Short cilia are found in cell type 5. As is commonly known, cilia show 9 microtubular doublets and 1 central pair of MT forming the axoneme. Exceptions of the 9+2 pattern have been reported. For example, in somatic cells of multicellular organism, centrioles elongate into ciliary stubs lacking a central pair of MTs. Although not tested by an experiment under MT destabilizing conditions, it is likely that also cell type 4 develops single cilia. The loss of cilia at the onset of mitosis may indicate their role in the regulation of the cell cycle (Tucker and Pardee, 1982). However, these authors point also at findings not compatible with this view. Our results question the assumed role in cell cycle regulation because 4 out of 5 proliferating microvessel endothelial cell have developed cilia. Thus, the significance of single cilia is still an open question in accordance, with the opinion of Beertsen et al. (1975). Heterogeneity of the interphase microtubute cytoskeleton
Microtubules are usually attached at one end to an organizing element, a microtubuleorganizing centre (MTOC) according to the terminology of Pickett-Heaps (1969). The MTOCs of mammalian interphase cells consist of a centriole pair embedded in a dense matrix. The centrioles may be contiguous with cilia (Wheatley, 1982). The MTOCs appear to determine the number of MTs
within the cell, their direction (Tucker, 1984), and probably their protofilament number (Evans et&. ,1985). Both, the tyrosinated and the acetylated variety of cytoplasmic MTs presently examined in five types of cdtured microvessel endothelial cells occur in high density around juxtanuclear centrosomes and are possibly organized by them. Thus, the distribution of cytoplasmic MTs within the examined set of endothelial cells of bovine origin is conventional. It has been repeatedly documented that endothelial cells are heterogeneous with respect to their biochemical and structural properties (reviewed by Gerritsen, 1987; Fajardo, 1989). However, heterogeneity in the microtubular cytoskeleton of cultured microvessel endothelial cells appears to be a new finding for endothelial cell cytoskeleton (reviewed by Blose, 1984). It adds another morphological element to the list showing differences among the five types of bovine microvessel endothelial cells derived from bovine corpus luteum. The list demonstrates variations in phase contrast morphology, immunocytochemical parultrastructure, ameters, and a mitosis-associated feature, i.e. the aster size of telophase spindles (Spanel-Borowski, 1991; Spanel-Borowski and van der Bosch, 1990; Wolf and SpanelBorowski, 1991). In the light of the diverse functions of MTs such as in cell motility, intracellular transport, and cellular organization, differences in the MT inventory in addition to above mentioned differences in morphology - indicate distinct functional properties of each microvessel endothelial cell type in culture. Whether the observed differences are typical for the in vivo situation, remains an open question. Cell types 1 and 2 have still been subconfluent at the onset of this study. Since a
INTERPHASE
MICROTUBULE
3%
CYTOSKELETON
low amount of cytoplasmic MTs appears in cell type 1, whereas the amount is high in cell type 2, an effect of cell density on the quantity of MTs is unlikely, yet it cannot be definitely ruled out. The appearance of cytoplasmic MTs is wavy in all cell types with the exception of spindle-shaped cell type 3. The rather regular path of cytoplasmic MTs in cell type 3 is miniscent of the MT arrangement of fibrocytes which are reported to have a regular path that is not found in epithelial cells (BrC et al.. 1987; Wadsworth and McGrail, 1990). All cell types examined contain some acetylated MTs. Cell type 5, however, and to a lower extent cell type 4 also, are characterized by the presence of a high amount of acetylated MTs. The ultrastructural analysis of cell types 3 to 5 shows abundant micropinocytotic vesicles as well as dilated rough endoplasmic reticulum (Spanel-Borowski, 1991). Since tyrosinated MTs are coordinately abundant in these cell types, MT mediated secretion may occur in the cells. Acetylation identifies MTs of enhanced stability (Piperno et al., 1987; Webster and Borisy. 1989), and acetylated MTs are, prob-
ably, those MTs which have been seen by us at the ultrastructural level in cell types 3 to 5. Acetylated MTs may contribute to reinforce the cytoskeleton of cell type 5 which is of larger size than the other cell types. Under this premise, the formation of acetylated MTs in large-sized aging cells of cell type 1 may be explained. The presence of acetylated MTs may as well be correlated to cell polarity which is known to develop and to be maintained in cultured endothelial cells. The role of acetylated MTs in the occurrence of polarized cells has been recently reported (Lim et al., 1989; Houliston and Maro, 1989: Jasmin et al., 1990).
Acknowledgements
We thank Dr J. Piperno (Rockefeller Univ., New York, USA) for the generous sample of 6-llB-1 antibody, Dr J. V. Kilmartin (MRC, Lab. Molec. Biol., Cambridge, UK) for the YL l/2 antibody, and Mrs Kyburg for expert technical assistance. This study was supported by the Deutsche Forschungsgemeinschaft (Sp 232/4-l).
References
Bcertscn. W., Ever&. V. and Houtkooper, J. M. 1975. Frequency of occurrence and position of cilia in fibroblasts of the peridontal ligament of the mouse incisor. Cell Tiss. Res., 163, 415-431. Blase. S. H. 1984. The endothelial cytoskeleton. In Biology of Endorhelial Cells (ed. E. A. Jaffc). pp. 141~~154. Martinus Nijhoff Publishers, Boston. The Hague, Dodrecht. BrC. M.-H., Kreis, T. E. and Karsenti, E. 1987. Control of microtubule nucleation and stability in Madin-Darby canine kidney cells: The occurrence of noncentrosomal. stable detyrosinated microtubules. J. Cell Biol.. 105, 1283-1296. Evans, L.. Mitchison, T. and Kirschner, M. 1985. Influence of the centrosome on the structure of nucleated microtubules. J. Cell Biol., 100, 1185-1191. Fa)ardo. L. F. 1989. The complexity of endothelial cells. Am. J. Ch. Pathol.. 92, 241-250. Gerritsen, M. E. 1987. Functional heterogeneity of vascular endothelial cells. Biochem. Pharmacol.. 36, 2701-2711. Houliston, E. and Mare, B. 1989. Posttranslational modification of distinct microtubule subpopulations during cell polarization and differentiation in the mouse preimplantation embryo. J. Cell Eiol.. 108, 54%551. Jasmin. B. J., Changeux, J.-P. and Cartaud, J. 1990. Compartmentalization of cold-stable and acetylated microtubules in the subsynaptic domain of chick skeletal muscle fibre. Narure, 344, 673-675. Jensen, C. G., Jensen, L. C. W. and Rieder. C. L. 1979. The occurrence and structure of primary cilia in a subline of Potorous tridactylus. Exp. Cell Res.. 123, 444-448. Kilmartin, J. V., Wright, B. and Milstein. C. 1982. Rat monoclonal antitubulin antibodies derived by using a new secreting rat cell line. J. Cell Biol.. 93, 576-582. Kochanski. R. S. and Borisy. G. G. 1990. Mode of centriole duplication and distribution. J. Cell &o/.. 110, 159% 1605. Krcis, T. E. 1987. Microtubules containing detyrosinated tubulin are less dynamic. EMBO J., 6, 2597-2606. Lim, S.-S., Sammak. P. J. and Borisy, B. B. 1989. Progressive and spatially differentiated stability of microtubules in developing neuronal cells, J. Cell Biol., 109, 253-263. Pickett-Heaps, J. D. 1969. The evolution of the spindle apparatus. An attempt at comparative ultrastructural cytology in dividing plant cells. Cyrobios, 1, 257-280.
WOLF AND SPANEL-BOROWSKI
354
Piperno, G. and Fuller, M. T. 1985. Monoclonal antibodies specific for an acetylated form of o-tubulin recognize the antigen in cilia and flagella from a variety of organisms. .I. Cell Biol., 101, 2085-2094. Piperno, G., LeDizet, M. and Chang, X. 1987. Microtubules containing acetylated cu-tubulin in mammalian cells in culture. J. CeN Biol., 104, 2089-302. Spanel-Borowski, K. 1991. Diversity of ultrastructure in different phenotypes of cultured microvessel endothelial cells isolated from bovine corpus luteum. Cell Tiss. Res.. 266, 3749. Spanel-Borowski, K. and van der Bosch, J. 1990. Different phenotypes of cultured microvessel endothelial cells obtained from bovine corpus luteum. Study by light microscopy and by scanning electron microscopy (SEM). Cell Tim. Res., 261, 35-47. Tucker, R. W. and Pardee, A. B. 1982. Primary cilia and their role in the regulation of DNA replication and mitosis. In Cell Growth (ed. C. Nicolini), pp. 365-376. Plenum Publ. Corp., New York. Tucker, J. B. 1984. Spatial distribution of microtubule-organizing centers and microtubules. J. Cell Bid., 99, S5s-62s. Wadsworth, P. and McGrail, M. 1990. Interphase microtubule dynamics are cell-type specific. J. Cell Sci., 95, 57-65. Webster, D. R. and Borisy, G. G. 1989. Microtubules are acetylated in domains that turn over slowly. J. Cell Sci., 92, 57-65. Wheatley, D. N. 1982. The Centride, a Central Enigma of Cell Biology. Elsevier Biomed. Press, Amsterdam. Wolf, K. W. and Bastmeyer, W. 1991. Cytology of Lepidoptera V. The microtubule cytoskeleton in eupyrene spermatocytes of Ephestia kuehniella (Pyralidae), Inachis io (Nymphalidae), and Orgyia antiqua (Lymantriidae). Eur. Cell Bid.,
55, 225-237.
Wolf, K. W. and Spanel-Borowski, K. 1992. Heterogeneity of spindle structure in different microvessel endothelial cell types derived from bovine corpus luteum. Protoplasma 166,42-48.
KLAUS WERNER WOLF and KATHARINA SPANEL-BOROWSKI*
THE INTERPHASE MICROTUBULE CYTOSKELETON OF FIVE DIFFERENT PHENOTYPES OF MICROVESSEL ENDOTHELIAL CELL CULTURES DERIVED FROM BOVINE CORPUS LUTEUM Keywords: Microvessel endothelial cells. cytoskeleton.
microtubules. cell culture
ABSTRACT. The interphase microtubule cytoskeleton of five different mlcrovesscl endothehal cell cultures. recently established from bovine corpus luteum, was analysed using anti-tuhulin immunofluorescence. An antibody against acetylated microtubules detected four cell type5 each of which possessed a single cilia. The length of the cilia were up to 10 pm for cell type, I and 2. Ciliary stubs had a length of up to 0.37,um in cell types 4 and 5. Cilia were missing in cell type 3. Long and short cilia were located in the perinuclear region from where cqtoplasmic microtubules radiated. Cell type 3 displayed straight microtubules rather than the wavy path seen in the other cell types. The amount of tyrosinated microtubules visuahzed by a spwfic antibody was consistently higher than that of posttranslationally acetylated microruhulcs. The latter were more apparent in cell types 4 and 5 than in the other cell types. WC conclude: Differences in the cytoplasmic microtubule inventory of each microvcssel cndothelial ccl1 type points at individual functions maintained in culture.
Introduction
different appearance of cell organelles. the development of cytoplasmic and axonemal microtubules (MTs) obviously varies among the five cell types. Cell type 1 displays a single cilium which is not detected in the other cell types. Cytoplasmic MTs are readily observed in cell types 3 to 5 whereas such MTs are not found in cell types 1 and 2. Since conventional ultrastructural analysis gives poor insight into the true architecture of MTs, immunofluorescence microscopy was applied to monolayers of the above described five different microvessel endothelial cell types. Monoclonal antibodies against two tubulin isotypes have been applied. These are (i) the YL l/2 antibody elicited against tyrosinated n+tubulin (Kilmartin et al., 1982) and (ii) the 6-llB-1 antibody raised against post-translationally acetylated actubulin (Piperno and Fuller, 1985). The YL l/2 antibody is believed to detect dynamic MTs, i.e. those with a high rate of exchange with the pool of tubulin subunits (Kreis, 1987). In contrast, MTs labeled with the 6-llB-1 antibody are less dynamic. In all likelihood,
Five different phenotypes of microvessel endothelial cells have been isolated from bovine corpus luteum and cultured for several months (Spanel-Borowski and von der Bosch, 1990). Cells are characterized by use of phase contrast microscopy, scanning electron microscopy as well as by immunocytochemistry. Cells of type 1 are arranged into an isomorphic monolayer of so-called cobble stone pattern apparent to a lesser degree in the polymorphic monolayer of cell type 2. Monolayers of cell types 3 to 5 are flat. These results have been recently confirmed and extended by ultrastructural findings (Spanel-Borowski, 1991). In addition to lnstitut fiir Biologic der Medizinischen Universitat zu Ltiheck. Ratzeburger Allee 160, D-2400 Liibeck, Germany. *Anatomischcs Institut der Umversit&, Pestalozistr. 20, CH-4056 Basel, Switzerland. Correspondence to Dr. Katharina Spanel-Borowski. Received 21 October 1991 Revised 11 December 1991 347
WOLF
348
acetylation is not the cause which renders MTs stable, but the consequence of slow MT turnover (Piperno et al., 1987; Webster and Borisy, 1989). Therefore, acetylation serves as a marker of stable MTs such as axonemal MTs of cilia. When density of acetylated cytoplasmic MTs is moderate, the 6-llB-1 antibody allows detection of cilia. Material and Methods Post-confluent cultures of microvessel endothelial cells maintained in 24-well culture plates were used. Phenotypes 1 to 5 were derived from bovine corpus luteum, isolated and cultured as described (Spanel-Borowski and van der Bosch, 1990). Cultures were subpassaged at a split ratio of 1:4 by use of 0.02% trypsin (cell types 1, 3 to 5) or 0.02% trypsin-0.02% EDTA (cell type 2), respectively. Cells were grown on cover slips coated with 1% collagen solution (Vitrogen@ Collagen Corporation, Palo Alto, California). At the time of processing, cell types 1 and 2 were still subconfluent, while cell types 3 to 5 appeared confluent due to higher cell proliferation. For anti-tubulin immunofluorescence, cover slips with cells adhering were treated with MT-stabilizing buffer (100 mM PIPES, 1 mM MgS04, 1 mM EGTA, pH 6.8) containing 1% Triton X-100. Details are described by Wolf and Bastmeyer (1991). After fixation with 2% paraformaldehyde and 0.25% glutaraldehyde in MT stabilizing buffer cytoskeletons were immersed in 100% methanol at -20°C. We used either rat monoclonal YL l/2 antibody (gift from Dr J. V. Kilmartin, MRC, Lab. Molec. Biol., Cambridge, UK), or mouse monoclonal 6llB-1 antibody (gift from Dr G. Piperno, Rockefeller Univ., NY, USA) as primary antibodies. Both antibodies were diluted 1:lOO with phosphate-buffered saline (PBS). As second antibodies, we applied biotin-conjugated anti-rat or anti-mouse antibodies (Sigma, Deisenhofen, Germany), diluted 1:50 with PBS, at room temperature for 1 hr. Specimens were further incubated with 1 mg polylysine (Serva, Heidelberg, Germany)/ ml PBS for 10 min, and with rhodaminecoupled avidine (Sigma), diluted 1:50 with PBS for 1 hr. In one experiment, the cytoplasmic MTs of cell type 5 cells were depolymerized with
AND SPANEL-BOROWSKI
a modified ice-cold lysis buffer (100 mM PIPES, pH 6.8,50 yM CaCIZ) containing 1% Triton X-100 (Kochanski and Borisy, 1990). For the detection of the remaining MTs, a monoclonal mouse anti-P tubulin antibody (Sigma) was diluted 1:lOO in PBS, and the specimens were incubated for 1 hr. All specimens were mounted in citrate buffer (100 mM citric acid, 200 mM Na2HP04, pH 7.0) containing 10 pg/ml 4’,6-diamidino-2phenylindole-2HC1, DAPI (Serva) in order to display chromatin arrangement. The preparations were documented using a Zeiss universal photomicroscope equipped with epifluorescence. Images were recorded on Tmax 400 (Kodak) developed in Rodinal (Agfa). Results Cell type 1 possessed an inconspicuous MT cytoskeleton. Tyrosinated MTs were few in number (Fig. la). As a rule, they were wavy and concentrated within the juxtanuclear area. Acetylated MTs were hardly to be discerned. Faintly fluorescent threads were detected by the 6-llB-1 antibody (Fig. lb). One intensely labeled thread 610pm in length projected from the surface of each cell. The threads represented single cilia having their base close to the nucleus. A few cells of larger size were observed in the monolayer of cell type 1. These cells were occasionally binucleate and are referred to as aging cells in the present text. Their MTs were less wavy than those in proliferating cells (Fig. lc, d). Tyrosinated MTs of aging cells occurred around the nucleus and a few of them extended towards the cell periphery. Acetylated MTs were lower in amount than tyrosinated MTs, yet acetylated MTs of aging cells were higher in number compared to acetylated MTs of proliferating cell type 1. Proliferating cell type 2 showed a denser MT network than cell type 1 (Fig. 2a, b). There was high MT density around the nuclei in addition to a large number of wavy MTs observed in the cell periphery. Single cilia showed a length of about 6 pm. In the vicinity of the ciliary base some weakly fluorescent threads of acetylated MTs were detected. Aging cells displayed a large number of tyrosinated MTs of apparently curved appearance (Fig. 2c, d). They radiated out from the perinuclear area and bended close to the
INTERPHASE
MICROTUBULE
Figs
l-5.
CYTOSKELETON
Microvessel endothelial cell types 1 to S derived from
are treated
with
Tyrosioated
a microtubule-stabilizing
microtubules
Fig.
la.
Fig.
b. Cell
type
(arrows).
Ic. d. Agmg
cell
single cilium
labeled
with
bovine
corpus
anti-tub&
la. c. 2a. c. 3a. 4a. Sa. whereas
luteum
antibodles.
Figs lb,
d, 2b. d.
1. Single
cilia
arc wiblc
(b)
using
I contains
acctylated
MTs
(d).
an antibody
against
acetylated
x 801).
type
cilium
cell type 2 Note
is indicated
and
microtuhulcs.
Fig. ?a. b. Cell type 2. A smglc Fig. 2c. d. Aging
buffer
arc seen in Figs
3h. 4b. Sb. c show acctylated
microtubule\
349
(arrow).
is marked
the large
x 300
(arrw).
size compared
x X00. x X00 to ccl1 type
Im
Fig. Za, b. The
INTERPHASE
MICROTUBULE
CYTOSKELETON
151
cell margin. Acetylated MTs were weakly developed in the cytoplasm. In cell type 3, tyrosinated MTs were abundant both in the vicinity of the nuclei as well as in the peripheral cytoplasm (Fig. 3a, b). The MTs were more or less of straight appearance and oriented parallel to the long axis of the cell. The density of acetylated MTs was low. Cilia were missing, Cell type 4 showed abundant tyrosinated MTs radiating from the perinuclear area (Fig. 4a, b). Though less dense, acetylated MTs were displaying a similar array as tyrosinated MTs. Due to the perinuclear abundancy of acetylated MTs, intensely fluorescent threads of approximately 0.3 pm length, presumably ciliary stubs, could be noticed only when the stubs projected over the perinuclear area. Cell type 5 possessed the highest density of tyrosinated and acetylated MTs of the 5 cell types (Fig. 5a-c). Tyrosinated MTs were slightly more concentrated in the juxtanuclear area, though a considerable amount of wavy MTs was also apparent in other parts of the cytoplasm. Highly fluorescent threads, probably ciliary stubs, could be seen only when the density of acetylated MTs was low in the surrounding cytoplasm. We had to verify in cell types 4 and 5 that the fluorescent threads were true cilia, thus, ruling out the presence of unusually long centrioles. Endothelial cells of type 5 were
treated with a MT destabilizing buffer prior to immunolabeling. The procedure caused depolymerization of cytoplasmic and ciliary MTs while centrioles were expected to remain stable (Kochanski and Borisy, 1990). There was a diffuse background, and the cytoplasmic inventory was lost. One pair of fluorescent spots remained apparent in interphase cells (Fig. 5d, e). These spots. smaller than the fluorescent threads of cells immunolabeled under MT stabilizing conditions, probably represented centrioles. Prophase cells showed two comparable spots being distant from one another (Fig. 5f, g). Therefore, cell type 5 and most probably also cell type 4, develop ciliary stubs. A semiquantitative evaluation of the appearance and amount of MTs is given in Table 1.
Discussion Single cilia
Using an antibody against acetylated MTs and comparing five different newly established microvessel endothelial cell cultures, the presence of single cilia or of ciliary stubs is revealed. The technique shows the superiority in detection of cilia compared to conventional technique by use of transmission electron microscope. By the latter, single cilia were found in cell type 1 only
Fig. 3a, b. Cell type 3. Tyrosinated microtubules are arranged parallel to the long axis of the cells (a). The density of acetylated microtubules is low in interphase cells (asterisks). but a mitotic spindle shows bright immunofluorescence (arrow-head). x 800. Fig. 4a, b. Cell type 4. In cells with low density of acetylated microtubules. detectable (arrow). x 800.
single cilia arc
Fig. 5a-c. Cell type 5. The cell shows the highest density of tyrosinated (a) and acetylated (b. c) microtubules. A ciliary stub is indicated (arrow). x 800. Fig. 5d. Cell type 5 of interphase cells. Prior to labeling with an antibody against Ptubulin. cells have been treated with a microtubule-destabilizing buffer. The cytoplasmic microtubules are depolymerized. Only small fluorescent spots, most probably centrioles (arrows) have remained. x 1250. Fig. Se. Same cells as in Fig. 5d. The nuclei are stained with DAPI, a DNA-specific fluorescent dye. Large fluorescent spots occur in the nuclear matrix. x 12.50. Fig. 5f. Cell type 5 is treated with a microtubule destabilizing buffer and labeled with an antibody against j%tubulin. Two fluorescent spots probably indicate centrioles (arrow). x 1250. Fig. 5g. Same cell as in Fig. 5f. The chromosomes
are condensed. DAPI staining. x 1250.
WOLF AND SPANEL-BOROWSKI
352 Table
1.
Cell type
Cytoplasmic tyrosinated
Microtubules acetylated
Shape
Cilia (length pm)
1
+,++
+
wavy
6-10
2 3 4 5
+++ +++ ++,+++ +++
+ + +,++ ++
wavy straight wavy wavy
about 6 absent 0.25-0.37 0.25-0.37
Semiquantitative evaluation: + not much, ++ moderate, +++ abundant
(Spanel-Borowski, 1991). The previous ultrastructural analysis obviously failed in visualizing short cilia of cell types 4 and 5, both growing as flat monolayers. Long cilia serve as a morphological marker of cell type 1. The length of them is comparable to that of PTKl cells (Jensen et al., 1979). Short cilia are found in cell type 5. As is commonly known, cilia show 9 microtubular doublets and 1 central pair of MT forming the axoneme. Exceptions of the 9+2 pattern have been reported. For example, in somatic cells of multicellular organism, centrioles elongate into ciliary stubs lacking a central pair of MTs. Although not tested by an experiment under MT destabilizing conditions, it is likely that also cell type 4 develops single cilia. The loss of cilia at the onset of mitosis may indicate their role in the regulation of the cell cycle (Tucker and Pardee, 1982). However, these authors point also at findings not compatible with this view. Our results question the assumed role in cell cycle regulation because 4 out of 5 proliferating microvessel endothelial cell have developed cilia. Thus, the significance of single cilia is still an open question in accordance, with the opinion of Beertsen et al. (1975). Heterogeneity of the interphase microtubute cytoskeleton
Microtubules are usually attached at one end to an organizing element, a microtubuleorganizing centre (MTOC) according to the terminology of Pickett-Heaps (1969). The MTOCs of mammalian interphase cells consist of a centriole pair embedded in a dense matrix. The centrioles may be contiguous with cilia (Wheatley, 1982). The MTOCs appear to determine the number of MTs
within the cell, their direction (Tucker, 1984), and probably their protofilament number (Evans et&. ,1985). Both, the tyrosinated and the acetylated variety of cytoplasmic MTs presently examined in five types of cdtured microvessel endothelial cells occur in high density around juxtanuclear centrosomes and are possibly organized by them. Thus, the distribution of cytoplasmic MTs within the examined set of endothelial cells of bovine origin is conventional. It has been repeatedly documented that endothelial cells are heterogeneous with respect to their biochemical and structural properties (reviewed by Gerritsen, 1987; Fajardo, 1989). However, heterogeneity in the microtubular cytoskeleton of cultured microvessel endothelial cells appears to be a new finding for endothelial cell cytoskeleton (reviewed by Blose, 1984). It adds another morphological element to the list showing differences among the five types of bovine microvessel endothelial cells derived from bovine corpus luteum. The list demonstrates variations in phase contrast morphology, immunocytochemical parultrastructure, ameters, and a mitosis-associated feature, i.e. the aster size of telophase spindles (Spanel-Borowski, 1991; Spanel-Borowski and van der Bosch, 1990; Wolf and SpanelBorowski, 1991). In the light of the diverse functions of MTs such as in cell motility, intracellular transport, and cellular organization, differences in the MT inventory in addition to above mentioned differences in morphology - indicate distinct functional properties of each microvessel endothelial cell type in culture. Whether the observed differences are typical for the in vivo situation, remains an open question. Cell types 1 and 2 have still been subconfluent at the onset of this study. Since a
INTERPHASE
MICROTUBULE
3%
CYTOSKELETON
low amount of cytoplasmic MTs appears in cell type 1, whereas the amount is high in cell type 2, an effect of cell density on the quantity of MTs is unlikely, yet it cannot be definitely ruled out. The appearance of cytoplasmic MTs is wavy in all cell types with the exception of spindle-shaped cell type 3. The rather regular path of cytoplasmic MTs in cell type 3 is miniscent of the MT arrangement of fibrocytes which are reported to have a regular path that is not found in epithelial cells (BrC et al.. 1987; Wadsworth and McGrail, 1990). All cell types examined contain some acetylated MTs. Cell type 5, however, and to a lower extent cell type 4 also, are characterized by the presence of a high amount of acetylated MTs. The ultrastructural analysis of cell types 3 to 5 shows abundant micropinocytotic vesicles as well as dilated rough endoplasmic reticulum (Spanel-Borowski, 1991). Since tyrosinated MTs are coordinately abundant in these cell types, MT mediated secretion may occur in the cells. Acetylation identifies MTs of enhanced stability (Piperno et al., 1987; Webster and Borisy. 1989), and acetylated MTs are, prob-
ably, those MTs which have been seen by us at the ultrastructural level in cell types 3 to 5. Acetylated MTs may contribute to reinforce the cytoskeleton of cell type 5 which is of larger size than the other cell types. Under this premise, the formation of acetylated MTs in large-sized aging cells of cell type 1 may be explained. The presence of acetylated MTs may as well be correlated to cell polarity which is known to develop and to be maintained in cultured endothelial cells. The role of acetylated MTs in the occurrence of polarized cells has been recently reported (Lim et al., 1989; Houliston and Maro, 1989: Jasmin et al., 1990).
Acknowledgements
We thank Dr J. Piperno (Rockefeller Univ., New York, USA) for the generous sample of 6-llB-1 antibody, Dr J. V. Kilmartin (MRC, Lab. Molec. Biol., Cambridge, UK) for the YL l/2 antibody, and Mrs Kyburg for expert technical assistance. This study was supported by the Deutsche Forschungsgemeinschaft (Sp 232/4-l).
References
Bcertscn. W., Ever&. V. and Houtkooper, J. M. 1975. Frequency of occurrence and position of cilia in fibroblasts of the peridontal ligament of the mouse incisor. Cell Tiss. Res., 163, 415-431. Blase. S. H. 1984. The endothelial cytoskeleton. In Biology of Endorhelial Cells (ed. E. A. Jaffc). pp. 141~~154. Martinus Nijhoff Publishers, Boston. The Hague, Dodrecht. BrC. M.-H., Kreis, T. E. and Karsenti, E. 1987. Control of microtubule nucleation and stability in Madin-Darby canine kidney cells: The occurrence of noncentrosomal. stable detyrosinated microtubules. J. Cell Biol.. 105, 1283-1296. Evans, L.. Mitchison, T. and Kirschner, M. 1985. Influence of the centrosome on the structure of nucleated microtubules. J. Cell Biol., 100, 1185-1191. Fa)ardo. L. F. 1989. The complexity of endothelial cells. Am. J. Ch. Pathol.. 92, 241-250. Gerritsen, M. E. 1987. Functional heterogeneity of vascular endothelial cells. Biochem. Pharmacol.. 36, 2701-2711. Houliston, E. and Mare, B. 1989. Posttranslational modification of distinct microtubule subpopulations during cell polarization and differentiation in the mouse preimplantation embryo. J. Cell Eiol.. 108, 54%551. Jasmin. B. J., Changeux, J.-P. and Cartaud, J. 1990. Compartmentalization of cold-stable and acetylated microtubules in the subsynaptic domain of chick skeletal muscle fibre. Narure, 344, 673-675. Jensen, C. G., Jensen, L. C. W. and Rieder. C. L. 1979. The occurrence and structure of primary cilia in a subline of Potorous tridactylus. Exp. Cell Res.. 123, 444-448. Kilmartin, J. V., Wright, B. and Milstein. C. 1982. Rat monoclonal antitubulin antibodies derived by using a new secreting rat cell line. J. Cell Biol.. 93, 576-582. Kochanski. R. S. and Borisy. G. G. 1990. Mode of centriole duplication and distribution. J. Cell &o/.. 110, 159% 1605. Krcis, T. E. 1987. Microtubules containing detyrosinated tubulin are less dynamic. EMBO J., 6, 2597-2606. Lim, S.-S., Sammak. P. J. and Borisy, B. B. 1989. Progressive and spatially differentiated stability of microtubules in developing neuronal cells, J. Cell Biol., 109, 253-263. Pickett-Heaps, J. D. 1969. The evolution of the spindle apparatus. An attempt at comparative ultrastructural cytology in dividing plant cells. Cyrobios, 1, 257-280.
WOLF AND SPANEL-BOROWSKI
354
Piperno, G. and Fuller, M. T. 1985. Monoclonal antibodies specific for an acetylated form of o-tubulin recognize the antigen in cilia and flagella from a variety of organisms. .I. Cell Biol., 101, 2085-2094. Piperno, G., LeDizet, M. and Chang, X. 1987. Microtubules containing acetylated cu-tubulin in mammalian cells in culture. J. CeN Biol., 104, 2089-302. Spanel-Borowski, K. 1991. Diversity of ultrastructure in different phenotypes of cultured microvessel endothelial cells isolated from bovine corpus luteum. Cell Tiss. Res.. 266, 3749. Spanel-Borowski, K. and van der Bosch, J. 1990. Different phenotypes of cultured microvessel endothelial cells obtained from bovine corpus luteum. Study by light microscopy and by scanning electron microscopy (SEM). Cell Tim. Res., 261, 35-47. Tucker, R. W. and Pardee, A. B. 1982. Primary cilia and their role in the regulation of DNA replication and mitosis. In Cell Growth (ed. C. Nicolini), pp. 365-376. Plenum Publ. Corp., New York. Tucker, J. B. 1984. Spatial distribution of microtubule-organizing centers and microtubules. J. Cell Bid., 99, S5s-62s. Wadsworth, P. and McGrail, M. 1990. Interphase microtubule dynamics are cell-type specific. J. Cell Sci., 95, 57-65. Webster, D. R. and Borisy, G. G. 1989. Microtubules are acetylated in domains that turn over slowly. J. Cell Sci., 92, 57-65. Wheatley, D. N. 1982. The Centride, a Central Enigma of Cell Biology. Elsevier Biomed. Press, Amsterdam. Wolf, K. W. and Bastmeyer, W. 1991. Cytology of Lepidoptera V. The microtubule cytoskeleton in eupyrene spermatocytes of Ephestia kuehniella (Pyralidae), Inachis io (Nymphalidae), and Orgyia antiqua (Lymantriidae). Eur. Cell Bid.,
55, 225-237.
Wolf, K. W. and Spanel-Borowski, K. 1992. Heterogeneity of spindle structure in different microvessel endothelial cell types derived from bovine corpus luteum. Protoplasma 166,42-48.
