Bioi Cell (1991) 71, 191-200

© Elsevier, Paris

191

O r i g i n a l article

An helicoidal structure surrounding the cilium axoneme: visualization by the monoclonai antibody CC-248 G u i l l e r m i n a B a u t i s t a - H a r r i s * , C a t h e r i n e K l o t z * * , Nicole B o r d e s * * , D a n i e l S a n d o z

Centre de Biologie Cellulaire, CNRS, 67, rue Maurice-Giinsbourg, 94205 Ivry-sur-Seine, France (Received 17 December 1990; accepted 18 March 1991)

Summary - Monoclonal antibody CC-248 labels cilia differentially on Triton X-100 permeabilizedciliated epithelium of quail oviduct by indirect immunofluorescence. On isolated ciliated cells, a punctuated staining is seen at the distal region over the bend of cilia. Electron micrographs of immunoperoxidase and immunogold techniques showed that the punctuated fluorescence corresponds to a helical disposition of CC-248 antigenic sites. This labeling was arranged on the axonemal distal region either as a simple or a double helix externally disposed around the nine microtubular doublets. These results suggest the existence of a detergent insoluble structure in the ciliary matrix that might concern the ciliary skeleton, probably acting as an elastic recoil that keeps the structural integrity of the axoneme during bending. The cross-reactivity of CC-248 MAb with the intermediate filament cytoske!eton of ciliated and smooth muscle cells indicates that this structure might be related to the intermediate filament family. axoneme

I dilated cells I differential labeling I monoclonal antibody

Introduction Cilia are motile cell processes with a complex structure. They are composed of the ciliary framework, the axoneme, which is embedded in a matrix and surrounded by a differentiation of the plasma membrane [11]. As a result of many studies with the electron microscope, freeze.etch techniques and biochemistry, the fine structure of the axoneme has become well known. Nine peripheral doublet microtubules, two central singlet microtubules, dynein arms and a variety of other appendages compose it. Some of these are thought to maintain the structural integrity of the cilium. The nexin links bridging the adjacent doublet tubules, seem to act as a restricting elastic force converting the sliding of doublets into a bent of the cilium [34]. Radial spokes are projected from the doublets towards the two central singlets. The interaction of radial spokes with the central sheath wich surrounds the two cen-

*Correspondence and reprints: Universidad de Las Palmas de Gran Canaria, Dpto de Morfologfa, Unidad de Histologia, Apartado 550, Las Palmas 35080, Spain **Present address: Centre de G~n~tique Mol~culaire, CNRS, 91190 Gif-sur-Yvette, France

Abbreviations: BSA, bovine serum albumin; DAB, 3,Y-diamine-

benzidine; HEPES, 4 (2-hydroxyethyl)-l-piperazine-ethanesulfonicacid; Ig, immunoglobulin; MAb, monoclonal antibody; NC, nitrocellulose; PAb, polyclonal antibody; PBS, phosphate buffer saline (1.5 mM KH2POo 2.7 mM KCI, 6.4 mM Na2HPO4.2H20, 137 mM NaCI, pH 7.4); PMSF, phenylmethyl-sulfonyl fluoride; SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis; TBS, tris buffer saline (I0 mM Tris, pH 7.4, 0.15 M NaCI); TBS-Tw, tris buffer salineTween 20 (I0 mM Tris pH 7.4, 0.15 NaCI, 0.1% Tween 20).

tral microtubules, is thought to be involved in the coordination of the sliding, forming the main part of the mechanism that converts active interdoublet sliding into local bending [24, 33]. Furthermore, the existence of a fibrous skeleton in cilia has recently been suggested. Nearly 80% of the tubulin is solubilized when axonemes of isolated gill cilia from the bay scallop Aequipecten irradians are heated at 45°C in a Tris-EDTA solution at pH 8, remaining mostly minor proteins in a nine-fold symmetrical configuration. This fibrous skeleton is composed largely of ciliary tektins, nexin linkages and other structural proteins [28]. However, knowledge of the ciliary membrane and axoneme embedding matrix is fragmentary by lack of morphologic and biochemical data. Ciliary membrane is in continuity with the plasma membrane with some specific differences. In addition to the lipidic components, glycolipids, cholesterol and phos. pholipids, it appears that tubulin is the major protein component which differs from axonemal tubulin [26]. The relative amounts of the membrane tubulin proteins seem to vary from organism to organism [9]. Stephens et a! [27] have suggested that membrane tubulin and its associated proteins constitute an integral protein skeleton in ciliary membrane as evidenced on gill cilia from A irradians. This membrane skeleton appears to be closely apposed to the axoneme and is attached to the outer doublets by fine radial bridges. In regard to periaxonemal ciliary constitution, it remains largely unknown. A Y-shaped fibrillar material [5, 12], linking the peripheral doublets of the axoneme to the transmembrane particles of ciliary membrane, has been evidenced in cilia from quail oviduct [5] and seminal vesicles of Discoglossus [6]. This Y-shaped structures are essentially concentrated in the necklace region and in the basal portion of the cilia, while they are more dispersed on the rest of the ciliary shaft [6, 23]. At the distal end of cilia,

192

G Bautista-Harris et at

microtubules are attached to the m e m b r a n e by microtubule-capping structures [9, 31]. It has been shown that cifia contain a high level o f calmodulin [13, 26, 29] either free within the periaxonemal space in absence of calcium or associated with the axonemal dynein arms, spoke heads and central sheath, when calcium is present and calmodulin has all the Ca 2+ binding sites filled [13, 29]. Tubulin also appears to be a soluble component o f ciliary matrix [1]. In addition, actin and myosin [7, 22] have been detected by immunocytochemistry at the proximal region of cilia. Actin has also been detected in flagella by immunoblot [21]. Immunochemical detection of ciliary constituents, which has not been evidenced by other techniques, is increasing. By the use o f a MAb elicited against a crude suspension o f quail oviduct basal bodies, we evidence by immunocytochemistry a Triton-insoluble structure, disposed helicoidally along the distal region of the axonemal cilium that seems to belong to the ciliary skeleton. Materials and methods Antibodies Monocional antibody CC-248 was obtained as described by Klotz et al [17]. Basal bodies from laying quail (Coturnix coturnix japonica) oviduct were semipurified and used as immunogen to produce monoclonal antibodies. The clone CC-248 secretes an immunoglobulin of IgM type as determined by Outcherlony's double diffusion test against antibodies specific for various classes of mouse heavy chains immunoglobulins [17]. Culture supernatant of CC-248 clone was used either undiluted or in a 1:2-1:5 dilution in PBS. Polyclonal antitubulin antibody raised against Paramecium ciliary tubulin, which recognizes ciliary tubulin of quail oviduct, was kindly provided by A Adoutte (Universit6 Paris Sud, Orsay/France), it was used at a dilution of I: 200. Anti alpha-tubulin and anti.desmin monoclonal antibodies were purchased from Amersham International (Buckingham. shire/UK) and they were used at a dilution of I : 300 and 1:20 respectively. Fluorescein conjugated lgs goat antimouse (Pasteur production, Paris/France), diluted 1 : 100 in PBS, was preadsorbed before incubation using scratched air-dried frozen sections. Horseradish peroxidase-conjugated lgs goat antimouse (Cappel laboratories, USA) diluted 1:40 in PBS, was preadsorbed with a suspension of isolated ciliated cells before use. Gold conjugated (5 nm) IgM goat antimouse was purchased from Jansen EM Grade (Belgium) and used diluted 1 : 5 in PBS. nSl-conjugated lgs sheep antimouse (Amersham International Buckinghamshire/UK) was used at a dilution of 1:200. Frozen sections Oviduct was removed from decapitated quail and the infundibulum segment was quickly frozen without previous fixation. Sections of 4 #m thick were obtained with a cryostat. Isolation of ciliated cell~ Ciliated cells were prepared accordingly to the procedure I described by Anderson [2]. Oviduct segments were filled with the extraction medium (20 mM Hepes buffer, pH 7.5, 1 mM EDTA, 25 mM KCI, 250 mM sucrose and 0.05% Triton X-100), clamped at both ends and placed in the same medium without Triton X-100. They were stirred for 2 min on a vortex mixer every 5 min for 30 min at 4°C. After collecting the content of the segments they were centrifuged at 1500 g for 10 min. Pellets, composed of isolated ciliated cells, were washed twice and laid into observation chambers by centrifugation at 2000 g as described by Sandoz et al [22].

Isolation o f smooth muscle cells Isolated smooth muscle cells were obtained by mechanical dissociation following the Bagby and Pepe [3] method with personal modifications. Fresh gizzards from adult quails were removed, cartilage and connective tissue were trimmed away and small longitudinal pieces of smooth muscle were torn with tweezers. They were immersed in a 1 : 1 (v/v) mixture of glycerol and standard salt solution (100 mM KCI, 10 mM KH2PO4, 1 mM MgCl2, 1 mM PMSF and 10 mM NaN 3, pH 7.0) and kept for some hours at 4°C until the muscle pieces took a whitish colour and the solution attained a redish colour. Then, fresh standard salt solutionglycerol was substituted and strips of smooth muscle were thinly teased.with fine needles and stored at - 20°C for a minimum o f ~ : o ~iays. Muscle strips were then suspended m standard salt ~-olution-glycerolin a proportion 1:4 (v/v) and Triton X-100 was added at a 0.1% concentration. Several steps of vigorous stir followed by a rest at 0°C, 5 min each, were consecutively performed until a pure suspension of isolated smooth muscle cells was obtained (about 30 min). Purity of the suspension was checked at each step by phase contrast microscopy. Tissue debris was pelleted at 300 g for I min and smooth muscle cells in the supernate were subsequently collected by centrifugation at 1500 g for 10 min. After two washes with standard salt solution, they were processed for immunocytochemistry like ciliated cells. Immunocytochemistry Immunofluorescence, immunoperoxidase and immunogold labeling were carried out over frozen sections and/or isolated ciliated cells in a damp chamber at room temperature. No fixation was performed whereas it decreased the affinity of the antibody. lmmunofluorescence was achieved as follows: frozen sections were treated with 0.05% Triton X-100 in PBS during 7 min with the aim to permeabilize membranes. After two washes in PBS to take away the detergent, they were preincubated with 3% BSA in PBS, then incubated for 45 min with primary antibody and washed 3 times in PBS. Fiuorescein-conjugated secondary antibody was then applied for 45 min followed by three washes in PBS. Sections were mounted with Mowiol (Mowio1488, Chimidis, Paris, France). Isolated ciliated cells were treated using the same procedure except for the detergent as they were permeabilized during their obtention process. Control samples were incubated with the second antibody only. lmmunofluorescence samples were observed with a Leitz Dialux Microscope equipped for UV epi-illumination. lmmunoperoxidase and immunogold labeling was performed over isolated ciliated cells following the same protocol as for immunofluorescence. Peroxidase activity was revealed using DAB and hydrogen peroxide as described by Graham and Karnowsky [15]. Light microscopic samples were mounted with Eukitt and observed with a Zeiss Ultraphot Microscope. lmmunoperoxidase and immunogold samples for electron microscopy were processed as follows. Electron microscopy After immunocytochemistry technique, samples for electron microscopy were fixed in 3% glutaraldehyde in PBS, postfixed with 1% OsO4 in cacodylate buffer, dehydrated in ethanol and embedded in araldite. Thin sections were stained with uranyl acetate and lead citrate. Alternatively, a set of sections were unstained to enhance peroxidase and gold granule contrast. Sect!ons were observed in an electron microscope Philips EM 300 at ~0 kV. lmmunoblot SDS-PAGE was performed according to Laemmli [19]. Samples from pellets of ciliated cells and smooth muscle cells were solubilized in denaturing buffer by sonication at 30 W for I min by pulses of I s and thereafter boiled for 5 min. Samples were run on a 10% SDS-PAGE, stained with Coomassie Brillant Blue

Helicoidal structure surrounding the cilium axoneme R-250 and destainec~ in 5% acetic acid with 2.5% glycerol and 20% methanol. The two following transfer procedures were used: 1) electrotransfer was achieved in 192 mM glycine, 25 mM TrisHCi, 0.1% SDS at 50 mA for 40 min in a semiphor semi-dry transfer unit; 2) simple diffusion transfer was accomplished in the same buffer for 24-36 h at 4°C with 1 kg weight [32]. Visualization of proteins was carried out on the NC sheets with Ponceau Red. For immunostaining, NC sheets were incubated for I h in non-fat powdered milk 5% (w/v) in TBS-Tw at 30°C. Then they were incubated in the primary antibody for the same time at room temperature and thereafter nsI- or horseradish peroxidase-conjugated second antibody was used diluted in TBSTw. Peroxidase was revealed with DAB and hydrogen peroxide adding imidazol to enhance the reaction. Results

Immunofluorescence on frozen sections, isolated ciliated cells and smooth muscle cells Before the immune reaction on sections, a treatment with Triton X-100 was performed in order to permeabilize cells. No staining was seen unless permeabilization was accomplished. Infundibular epithelium of laying quail oviduct is constituted by almost exclusively ciliated cells with a few nonciliated cells interspersed. Figures la, c, d provide the immunofluorescence labeling on infundibulum quail oviduct with CC-248 MAb, which takes place as two fluorescent

193

strands. On a careful comparison with the phase contrast image (fig lb) we can see that the most apical fluorescent strand corresponds to a part of cilia while the strand below is at the terminal web level. A faint fluorescence is observed in the rest of the cytoplasm. Smooth muscle cells from muscular layer and from blood vessels are also labeled. On these cells staining takes place on the cytoplasm leaving the nucleus area unstained (fig la). Immunofluorescence on isolated ciliated cells gave similar results. Cilia were stained differentially (fig 2a, b, c), a strong fluorescence taking place at their distal region while proximal region and basal bodies exhibited no labeling. This staining seemed to occur in a punctuated form as shown in figure 2a. To determine whether this type of labeling is specific to CC-248 MAb, a polyclonal antitubulin antibody and a monoclonal anti alpha-tubulin antibody were used. In both cases cilia were stained uniformly on its whole length as manifest figures 2d and 2e. In addition, cells that are sufficiently flattened against the slide showed a faint fluorescence at the terminal web level (fig 2b). Cytoplasm of isolated smooth muscle cells from gizzard were intensely stained with CC-248 MAb. Labeling appeared to be on longitudinally arranged filaments (fig 2f, g). Immunofluorescence control samples, either frozensections or isolated ciliated cells and smooth muscle cells were uniformly negative.

Fig 1. Immunofiuorescence staining of quail oviduct frozen sections with CC-248 MAb. a, b. Fluorescent and phase contrast images of the same field. Fluorescence is found in epithelial (E) and muscular (M) layers. In the epithelium, staining is mainly in the apical part of ciliated cells (between arrowheads). Cell body of smooth muscle cells from muscular layer (M) and from vessels (V) are stained. e, d. In the epithelium, a strong fluorescence is associated with the apical part of cilia (arrow) and terminal web (arrowhead). The unstained strand between them corresponds to the proximal part of cilia and basal body region. A weak fluorescence stains the cytoplasm. Bar = 5 ~m.

194

G Bautista-Harris et ai

Immunoperoxidase on isolated ciliated cells The differential labeling along the ciliary axoneme is more clearly evidenced with immunoperoxidase staining as shown on figure 3a. The intensity and length of the stain on cilia are qualitatively different on each ciliated cell and even among them, appearing to be related with the ciliary bending in such a way that the axonemal region stained is the straight distal region over the bend. Electron microscopy immunoperoxidase technique was performed in order to determine the precise localization of the CC-248 antigen. Results strengthened the foregoing data. As provided on figures 3b and 3c, cilia are stained at their distal part in a gradual form. On higher magnification of longitudinal sections of cilia, it was seen that peroxidase reaction appears as a simple or double "curl" around the axoneme (fig 4b, c). It was usual to observe a peroxidase positive reaction clustered as external bouquets placed in zig-zag at both sides of ciliary axonemes (fig 4a). Ciliated cells from quail oviduct that had been isolated with Triton X-100 kept their usual shape with the ciliary apparatus and the cytoskeleton well preserved. The cytoskeleton appeared as a loose network where CC-248 labeling occurs without any exact location, seeming to be in preference on some cross points between the cytokeratin intermediate filaments. Terminal web displayed a sprinkled peroxidase labeling on the cytoskeleton under the basal body area (fig 3d).

Control immunoperoxidase samples were uniformly negative (fig 3e). Colloidal gold labeling on isolated ciliated cells and smooth muscle cells Same results as above were obtained with immunoelectron colloidal gold labeling. Gold granules appeared in groups disposed alternatively at both sides of the distal part of longitudinally sectioned ciliary axonemes. Proximal region of cilia and basal bodies zone were not stained. At the cytoskeletal level, gold was in association with the intermediate filaments mainly decorating the fibrillar lattice preferably on some crossing points (fig 5a, b). Gold disposition was easily appreciated if samples were without conventional contrasting uranyl acetate and lead citrate stain (fig 5a and 6a, b). On these samples the helical aspect of the labeling was more evident, specially in those where the section was tangential to the ciliary axoneme; simple or double helicoidal loops arranged around the apical extremity of axoneme (fig 6a, b, c) were observed as it was seen on immunoperoxidase samples. Figure 6d shows that cilia sectioned transversally have a labeling clustered around the axoneme and that those in oblique sections have it disposed in loops. On isolated and permeabilized smooth muscle cells, the cytoskeleton appeared on a loose pattern and gold beads

Fig 2. lmmunofluorescence staining of isolated ciliated cells from quail oviduct (a-e) and smooth muscle cells from gizzard if-g) with CC248 MAh (a, b, f), anti alpha-tubulin MAb (d) and antitubulin PAb (e). a, b, c. Fluorescent (a, h) and phase contrast (¢) images of ciliated cells with CC248 MAb. Cilia are strongly stained at their distal region. In ciliated cells with spread cilia, a punctuated fluorescence is seen (arrow). Comparison of fluorescent image (h) with that in phase contrast (c) shows that staining is located at the apical region of cilia. Proximal region and basal bodies are not stained (braquet in c). A faint fluorescence takes place at the terminal web region, d, e. Anti alpha-tubulin MAb (d) and antitubulin PAb (e) stain cilia at their whole length. Double arrow points to cilia that have been broken from cells and have fallen at the bottom of the observation chamber. They are stained at their whole length with antitubulin antibodies (d, e) but only a part of them with CC-248 MAb (a). f, g. Fluorescent (0 and phase contrast (g) images of isolated smooth muscle cells with CC-248 MAb. Staining occurs on longitudinal filaments that occupy most of the cytoplasm. Bar = 5 pm.

Helicoidal structure surrounding the cilium axoneme

195

+++++ +

1

+~+.+

+'~

__

+

++ + j:~.~',i+.+..+. ++m~.+ ,.~+ +,,,

~+~I~

"'+-~+

:+'.++++-~+X,~++~.. ?: +~

~

. ~. ..,.?

•

+

'+

~,+?+~

"'b.

.

+

~

~%

+~

~i~

,,

+®® + ,,,++ .,+ +

"

,,. , g ~ $ . " +~.,>~.+

•. , , , , , , . . , +~

"

.

++, +..++' !+'+ i +~i ~?~ :+.

.... r'J,,

m

.,

+ ++++ :++ + +.+++O,~+,.,.+,~++.+~,,.+~,~

• ",+',,~:'~ ~ +~+ ~"~i! +'~!'~`+ . ,C ~,~,+.. ~ +'~W+, .......

Fig 3. Light and electron microscopy immunoperoxidase staining of isolated ciliated cells with CC-248 MAb. a. Light microscopy immunoperoxidase staining. Labeling appears to be on the straight region over the bend (arrows). b, e. Electron microscopy immunoperoxidase staining. Longitudinal (b) and transversal (e) sections of cilia show that a strong peroxidase re~.tion, located at the distal part, appears to be related with axonemal components. Proximal region of cilia and basal bodies are clearly less stained. d. Electron microscopy immunoperoxidase of the terminal web region. A sprinkled peroxidase labeling takes place on the cytoskeleton under the basal body area (arrows). e. in control staining no incubated with primary antibody, no peroxidase reaction is detectable. Bar in a -- 5/am, in b, e, d, e = 1/~m.

b

+.

t::

/

i Fig 4. Electron microscopy immunoperoxidase staining at the distal region of cilia, a. In longitudinal section of cilia peroxidase, reaction appears clustered as external bouquets (arrows) at the lateral sides of the cilium, b. In a broken cilium peroxidase, reaction appears on an untwisted curl-like structure (arrow). e. Occasionally, peroxidase staining seems to be on a double curl (arrow). Bar = 100 nm.

196

G Bautista-Harris et al

were linked to intermediate filaments (fig 6e). Dense plaques exhibited no labeling. As for the foregoing techniques, immunogold controls were negative.

Immunoblot Figure 7 shows immunoblot o f ciliated and smooth muscle cells samples.

8 .-

:?

.~,o~,

•

.

~!

ill: ~

',

!

Fig $. Electron microscopy colloidal gold staining with CC-248 MAb on isolated ciliated cells. Unstained (a) and stained (b) sections with uranyl acetate and lead cytrate, a. Longitudinal section of cilia. Gold granules appear at the distal part of ciliary axonemes. b. Terminal we is labeled at the cytoskeleton fibrillar latice (arrow). Basal body zones are not stained. On the left corner the inset shows a high magnification of the labeling. Bar = 200 nm, inset = 20 rim.

Helicoidal s t r u c t u r e s u r r o u n d i n g t h e c i l i u m a x o n e m e

With CC-248 MAb a band of 55 kDa molecular weight is strongly stained in the muscle fraction while a faint staining is seen at approximately the same level in the ciliated cells fraction; smooth muscle myosin exhibits a labeling with a similar intensity (fig 7a). Anti-desmin MAb recognizes the 55 kDa band on smooth muscle cells and oviduct fractions, while no staining appears in ciliated cells samples.

Discussion

Using immunofluorescence, our results demonstrate a differential labeling of cilia from epithelium quail oviduct. CC-248 MAb stains them in approximately their distal part in a punctuated form. The possibility that remnants of the ciliary membrane at the proximal region of cilia could constrain the accessibility of the antibody, was excluded, because ciliary axoneme~ .vere completely demembranated as it was seen on the electron micrographs. The total

,',..-~

c

~

197

absence of label at this region could be caused by the exclusive location of CC-248 antigenic sites or by a masking of them by other ciliary elements which location is restricted to this part. Differential labeling of ciliary axonemes has been described in the literature so far. Sandoz et ai [22] have shown by immunocytochemical techniques on demembrahated ciliated cells of quail oviduct, that the presence of actin and myosin are restricted to the proximal region of cilia. Antibody directed against actin stains axonemes only on the proximal-half portion, seeming to be associated with the external part of the peripheral doublets, as had been suggested by Piperno and Luck [21] for Chlamidomonas flagella, while antimyosin antibody stains a small area of axonemes just above the transitional region. Chailley et al [7] supported these findings on isolated cilia from quail oviduct, suggesting an heterogeneous distribution of actin whithin the axonemal length as well as a special organization of axonemes at the lower level of cilia.

.....

•

pq

!

i

~!:!iii¸~¸!¸"¸ i:!i~.

~:i¸

'i % 1 ~ , ~e,

,,'~,.

Fig 6. Electron microscopy colloidal gold labeling on cilia and smooth muscle cells, a, b. Distal region of cilia sectioned tangentially. Sections are not contrasted with uranyl acetate and lead citrate. Gold granules are clustered° forming a spiral around the ciliary axoneme. e. In cilia with the microtubular doublets detached, labeling is found disposed in a double helix (arrow). d. On transversal sections, labeling is external to the axoneme (arrow). Oblique sections show gold granules arranged in loops (arrowhead). e. Isolated smooth muscle cell. Gold appears in clusters beside cytoskeleton intermediate filaments and external to the dense plaques (P).

198

G

Bautista-Harris et al

nal boundary of the axoneme suggesting that they are on a fibrillar network. Since on KCI extracted axonemes this helical labeling is mine clearly evidenced (manuscript in preparation), we think that in unextracted axonemes their image could be overshadowed by the most electron dense matrix on which it is embedded. CC-248 MAb binds also to cytoplasmic filaments at the cytoskeletal level. Cytoskeleton of vertebrate ciliated cells is exceptionally developed [I0]. As shown by Gounon et al [14] on thick sections and immunocytochemical technique, a well developed meshwork of cytokeratin intermediate filaments extends throughout the ciliated cell cytoplasm and especially the apical region. This meshwork is composed of interwoven f'daments which constitute tight and separate bundles just beneath the basal body area; a lateral bundle, stretching down to the basal pole of the cell, connects the apical array of intermediate filaments to the basal part of the cell. Ciliated cells from quail oviduct that have been isolated with Triton X-100 keep their usual shape and the cytoskeleton appears as a loose network where CC-248 binding takes place irregularly on the intermediate filament cross-points. Smooth muscle cells are also recognized by CC-248 MAb. Labeling seems to occur at the cytoskeletal level as provided by immunofluorescence and immunogold technique. On isolated and permeabilized smooth muscle cells, the cytoskeleton network appears on a loose pattern and gold particles are linked to intermediate filaments. Cytoskeletcn of smooth muscle cells is constituted by intermediate filaments arranged in a dense network of mainly longitudinal fibers throughout most of the cell, the major component being desmin, vimentin or both, as has been demonstrated by immunochemical techniques [4, 16,

Our results are in agreement with the foregoing in regard to the heterogeneity of the distribution of some elements within the cilium as well as to a different organization. It seems very probable, therefore, that beside the detergent insoluble axoneme, which has a known morphology almost invariable throughout its length, cifia possess other components concerning presumably the ciliary matrix. A heterogeneous distribution of some of them into the cilia could produce a different spatial organization among the various regions along it. Light microscopy immunoperoxidase on isolated ciliated cells showed that CC-248 MAb binds to cilia at the straight distal region over the bend. As provided by immunoelectron microscopy with peroxidase and colloidal gold, decoration occurs with a helicallike disposition, seeming to be around the periphery of microtubule doublets. A similar structure has not been described yet. Mesland et al [20] show a drawing cilium with a spiral on the ciliary shaft, representing a structure that, the authors believe, encircles the axoneme, but they do not make any commentary about it, neither indicating why they suppose its existence nor how they have evidenced it. Taking into consideration that the size of the antibodies and the gold particles restrict resolution, the possibility that the location of the CC-248 antigenic sites might be directly in the microtubules has been discarded since antitubulin antibodies, either polyclonal or monoclonal, label cilia uniformly throughout its whole length. In the same way, we do not think that CC-248 MAb recognizes the nexin links, because on sea-urchin flagellar axonemes treated with 0.6 M KCI and dialysed for 24 h at 4°C procedure which extracts the outer and inner dynein arms and usually opens the axoneme profile resulting in arrays of nine or fewer doublets held together by nexin links - , there is no labeling like it does on axonemes that conserve their closed round profile (data not shown). We do not

18, 3o1.

Molecular characteristics of the CC-248 antigen remain to be determined. Our results are not conclusive. Whereas a 55 kDa band is clearly recognized on smooth muscle s a m pies, the labeling on ciliated cells samples is not so evident. As shows figure 7a the faint staining observed on ciliated cells at 55 kDa ]eve] is exhibited also by smooth muscle myosin. Although we intuitively think that this is the molecular weight of CC-248 antigen, we still have no

exclude the possibility that the molecule, in which the CC-248 epitope is located, experiments a conformational change conceivably related with bending, allowing the accessibility of the antibody to the epitope. On high magnification of electron micrographs, perox. idase and gold labelings appear overall around the exter-

C III

c m

m ov

c

•

220

Kd

•

, ,

,

L??,, !ili ....iili ¸ 55 Kd

•

~

I

¸ i!il ili~!ii~i~i !i~l

a

a

b

]Fig 7. lmmunoblot of SDS-PAGE smooth muscle (m) cUiated cells (c) and whole oviduct (ov) fractions, a. CC-248 MAb-12Sl conjugated Ig. a': control. In the smooth muscle fraction, a 55 kDa band is well stained, whereas a faint reaction, apparently not specific, is observed on myosin and two more bands. In the ciliated cell fraction, a faint reaction is also observed at approximately 55 kDa level• b. Anti-desmin MAb - peroxidase conjugated Ig. The 55 kDa corresponding to desmin is well stained in smooth muscle and whole oviduct samples. There is no labeling in the ciliated cell sample.

Helicoidalstructure surrounding the ciliumaxoneme definitive conclusions because results are not repetitive. This could be due to the fact that the quantity of antigen is not sufficient to give a signal or that a denaturation of the antigenic site occurs during the electrophoretic process. Cross-reactivity between antibodies against ciliary components, like tektin and intermediate filaments, like cytokeratin, vimentin and desmin, has been reported [8, 25] and the similarities that exist between these two groups of proteins have grown [25]. Since our immunocytochemistry results suggest that the helical ciliary structure detected by the MAb CC-248 is related to intermediate f'daments, a MAb anti-desmin has been tested. As it was expected, this antibody recognized the desmin of smooth muscle cells by immunocytochemistry and immunoblot; however, ciliated cells did not exhibit the same staining as CC-248, and cilia were not labeled at all (not shown). Immunoblots were also negative. Other anti-intermediate filament antibodies and anti-tektin antibodies remain to be tested, the latter with the aim of determining if this structure is tektinrelated. The results exposed above suggest that the MAb CC-248 cross-reacts with intermediate filaments on immunocytochemical samples whereas the negative staining on immunoblot is due to the denaturation of the antigenic sites during the electrophoretic process. However, more experiments are needed to substantiate this conclusion. Although we were not able yet to characterize the antigen recognized by this MAb, CC-248 might be an Ab elicited against an epitope shared by different intermediate filament proteins. In any case, this MAb allowed us to evidence a ciliary structure, not yet visualized, helicoidally arranged around the nine peripheral doublets and that appears to be related to the bend of cilia.

Acknowledgments We thank B Chailley for her critical reading of the manuscript. Warm thanks are also due to MC Lain6, who instructed us in the immunocytochemical techniques and gave us helpful advice. We are grateful to A Adoutte for the generous gift of the antitubulin polycional antibody. This work was supported by a 8rant from Las Pesmas Universitary'Foundation (Canary Islands, Spain) and by a grant from the FPI Education and Science Spanish Ministry.

References I Adoutte A, Ramanathan R, Lewis RM, Dute RR, Ling YY, Kung C, Nelson DL (1980) Biochemical studies of the excitable membrane of Paramecium tetraurelia. III. Proteins of cilia and ciliary membranes. J Cell Biol 84, 717-738 2 Anderson RGW (1974) Isolation of ciliated or unciliated baseS bodies from the rabbit oviduct. J Cell Bio160, 393-404 3 Bagby RM, Pepe FA (1978) Striated myofibrils in antimyosin stained, isolated chicken gizzard smooth muscle ceils. Histochem 58, 219-235 4 Baghy R (1986) Toward a comprehensive three-dimensiones model of the contractile system of vertebrate smooth muscle cells. Int Rev Cytol 105, 67-128 5 Boisvieux-Ulrich E, Sandoz D, Chesiley B (1977) A freezefracture and thin section study of the ciliary necklace in quail oviduct. Biol Cell 30, 245-252 6 CheSlleyB, WDiaye A, Boisvieux-UlrichE, Sandoz D (1981) Comparative study of the distribution of fuzzy coat, lectin receptors, and intramembrane particles of the ciliary membrane. Eur J Cell Biol 25, 300-307 7 Chailley B, Bork K, Gounon P, Sandoz D (1986) Immunological detection of actin in isolated cilia from quail oviduct. Biol Cell 58, 43-52

199

8 Chang X J, Piperno G (1987) Cross-reactivity of antibodies specific for flagellar tektin and intermediate filament subunits. J Cell Biol 104, 1563-1568 9 Dentler WL (1981) Microtubule-membrane interactions in cifia and flagella. Int Rev Cytol 72, 1-47 10 Franke WW, Appelhaus B, Schmid E, Freudenstein C, Osborn M, Weber K (1979) Identification and characterization of epithelial cells in mammalian tissues by immunofluorescence microscopy using antibodies to prekeratin. Differentiation 15, 7-25 11 Gibbons IR (1967) The organization of cifia and flagella. In: Molecular organization and Biological Function (Allen JM, ed). Harper and Row Publishers, New York, 211-236 12 Gilula biB, Satir P (1972) The ciliary necklace. A cifiary membrane specialization. J Cell Biol 53, 494-509 13 Gordon RE, WilliamsKB, Puszkin S (1982) Immune localization of calmodulin in the ciliated cells of hamster tracheal epithelium. J Cell Biol 95, 57-63 14 Gounon P, Lain~ MC, Sandoz D (1987) Cytokeratin filament organization in the cifiated cells of the quail oviduct. Fur J Cell Biol 44, 229-237 15 Graham RC, Karnowsky MJ (1966) The early stages of absorption of injected horseradish peroxidase in the proximal tubules of mouse kidney: ultrastructural cytochemistry by a new technique. J Histochem Cytochem 14, 291-305 16 Huiatt TW, Robson RM, Arakawa N, Stromer MH (1980) Desmin from avian smooth muscle. J Biol Chem 255, 6981-6989 17 Klotz C, Bordes N, Lesn~MC, Sandoz D, Bornens M (1986) A protein of 175 000 daltons associated with striated rootlets in ciliated epithelia, as revealed by a monoclonal antibody. Cell Motil C.vtoskeleton 6, 56-67 18 Lazarides E, Hubbard BD (19J6) Immunological characterization of the subunit of 100A filaments from muscle cells. Proc Natl Acad Sci USA 73, 4344-4348 19 Laemmli UK (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature (Lond) 227, 680-685 20 Mesland DA, Hoffman JL, Cesigor E, Goodenough UW (1980) Flagellar tip activation stimulated by membrane adhesions in Chlamydomonas gametes. J Cell Bio184, 599-617 21 Piperno G, Luck DJL (1979) An actin-like protein is a component of axonemes from Chlamidomonas flagella. J Biol Chem 254, 2187-2190 22 Sandoz D, Oounon P, garsenti E, Sauron ME (1982) Immunocytochemicai localization of tubulin, actin, and myosin in axonemes of ciliated cells from quail oviduct. Proc Natl Acad $ci USA 79, 3198-3202 23 Sandoz D, Chailley B, Boisvieux-Ulrich E, Lemuliois M, Lesn6 MC, Banti~ta-Harris G (1988) Organization and functions of cytoskeleton in metazoan ciliated cells. Biol Cell 63, 183-193 24 Satir P (1989) The role of axonemal components in ciliary motility. Comp Biochem Phyziol 94A, 351-357 25 Steffen W, Linck RW (1989) Tektins in ciliary and flagellar microtubules and their association with other cytoskeletal systems. In: Cell movement, vol. 2: kinesin, d.vnein and microtubule dynamics. (Warner FD and Mclntosh JR, eds) New York, Allan R Liss, 67-81 26 Stephens RE (1983) Reconstitution of ciliary membranes containing tubulin. J Cell Biol 96, 68-75 27 Stephens RE, Oleszko-szuts S, Good MJ (1987) Evidence that tubulin forms and integral membrane skeleton in molluscan gill cilia. J Cell 5ci 88, 527-535 28 Stephens RE, Oleszko-szuts S, Linck RW (1989) Retention of ciliary ninefold structure after removal of microtubules. J Cell $ci 92, 391-402 29 Stommel EW, Stephens RE, Masure HR, Head JF (1982) Specific localization of scallop gill epithelial calmodulin in cilia. J Cell Biol 92, 622-628 30 Stromer MH, Bendayan M (1988) Arrangement of desmin intermediate filaments in smooth muscle cells as shown by high-resolution immunocytochemistry. Cell Motil Cytoskeleton 11, 117-125 31 Suprenant KA, Dentler WL (1988) Release of intact

200

G Bautista-Harriset al

microtubule-capping structures from Tetrahymena cilia. J Cell Biol 107, 2259-2269 32 Ternynck Th, Avrameas S (1987) Techniques immunoenzymatiques. Editions INSERM, Paris

33 Warner FD, Satir P (1974) The structural basis of ciliary bend formation. J Cell Biol 63, 35-63 34 Warner FD (1983) Organization of interdoublet links in Tetrahymena cilia. Cell Motil 3, 321-332