Hearing Research, 20 (1985) 207-214

Elsevier

207

HRR 00663

Immunocytochemical localization of contractile and contraction associated proteins in the spiral ligament of the cochlea

M.M. Henson ‘, K. Burridge *, D. Fitzpatrick 2, D.B. Jenkins 3, H.C. Pillsbury ’

and O.W. Henson, Jr. * ‘, Dicwion of Otolaryngologv, Department of Surge<v, and ’ Department of Anatomy, University of North Carolina, Chapel HIII,

NC 27514; and ’ Department of Biomedical Sciences, School of Dental Medrcme. Southern Illinois State Universrty,

2800 College Ave., Alton, IL 62002, U.S.A.

(Received 17 May 1985; accepted 21 August 1985)

Most of the extracellular fibers of the spiral ligament are associated with a distinct band of ‘anchoring’ cells which occur at the

boundary between the spiral ligament and the otic capsule. These cells are characterized by parallel arrays of intracellular filaments

which, along with the extracellular fibers, insert into electron dense, conical adhesion plaques. The intracellular filaments show a close

morphological resemblance to the ‘stress fibers’ of cultured fibroblasts (Henson et al., 1984). In the present study we have

demonstrated by immunofluorescence techniques that the anchoring cells, unlike adjacent cells of the spiral ligament, contain a

complement of proteins that is typically associated with stress fibers and with contractile systems. In addition to actin, the cells

contain myosin, tropomyosin, a-actinin and tahn. These results lend further support to the hypothesis that the anchoring cells have

the capacity to create and/or maintain tension on the spiral ligament-basilar membrane complex and to influence the mechanical

properties of the basilar membrane.

cochlea. spiral ligament, immunocytochemistry, bat

Introduction

Recent studies on the ears of mice and bats have shown that the outer margin of the spiral ligament contains a specialized group of cells which anchor some of the extracellular fibrous matrix to

the bony otic capsule (Henson et al., 1984). These ‘anchoring’ cells lie in a distinct band (Fig. 1) and

they contain parallel arrays of intracellular fila-

ments which insert into electron dense, conical adhesion plaques. As shown in Fig. 2 small bun-

dles of the extracellular fibrous matrix also insert into these plaques such that the ends of the in-

tracellular and extracellular fibers are in close association. In the horseshoe bat, Rhinolophus, the anchoring cells provide the sole mode of attach- ment of most of the spiral ligament to the otic capsule. Henson et al. (1984) have noted that the intracellular fibers show a close morphological re- semblance to, and are similar in diameter to, the stress fibers which form in cultured fibroblasts. On

this basis it was suggested that they may be capa- ble of creating or maintaining radial tension on

the spiral ligament-basilar membrane complex. Stress fibers are composed of actin filaments

and they are typically associated with ‘contractile related proteins’ (myosin, tropomyosin and (Y- actinin) (Burridge, 1981; Byers et al., 1984). In addition the protein talin has recently been identi-

fied in adhesion plaques and may play a role in the organization of actin near the membrane (Bur- ridge and Connell, 1983a; Burridge and Connell, 1983b). The purpose of this report is to show that the anchoring cells have the necessary actin, myosin

and accessory proteins to provide a firm biochem- ical basis for the contention that the intracellular filaments are capable of contraction.

Methods

The animals used in this study were mustache bats, Pteronotus p. parnellii, from Jamaica, WI.

‘at,

a@ E). Xls tri-

I

and white laboratory mice (ICR). The animals were killed by decapitation, the heads cut in the midsagittal plane and the cochleae rapidly re- moved and placed in fixative. Tissue to be pre- pared for light or transmission electron micro- scopy was processed according to the protocol previously described (Henson et al., 1984). Cochleae to be used for immunocytochemistry were fixed in a solution of 60% ethanol, 30% chloroform and 10% acetic acid for 45 min. The membrane covering the round window was pierced and the apex opened to allow the fixative to penetrate throughout the cochleae. The tissue was then washed briefly with 0.1 M phosphate buffer, pH 7.3, and decalcified in 0.1 M EDTA in phosphate buffer. The EDTA solution was changed each day. After decalcification was complete (approximately 4 days) the cochleae were washed in buffer, de- hydrated through an ethanol series (30%, 50%. 70%, 95% 100%; two changes, 10 min each) to xylene. After two changes of xylene, 10 min each, the tissue was placed in a 1 : 1 xylene-paraffin mixture for 45 min and then embedded in paraf- fin. Sections were cut at 3 and 5 pm and mounted on small (12 mm diameter) coverslips. Just prior to exposure of the tissue to the antibody, the paraffin was removed through two changes of xylene, 10 min each, and rehydrated through a graded ethanol series consisting of two changes, 10 min each, in 1008, 90%, 708, 50% and 30% ethanol. The tissue was then placed in a buffer consisting of 150 mM NaCl, 50 mM Tris-HCl and 0.1% sodium azide, pH 7.6. The coverslips were drained and overlaid with the antibodies; they were incubated for 1 h at 37°C in a humidified chamber. Coverslips were drained and then washed for 10 min in the buffer described above. After draining the coverslips were overlaid with a second antibody, FITC-labelled goat anti-rabbit IgG (Cappel Laboratories Inc., Cochranville, PA). They were again incubated for 1 h at 37°C in a humidified chamber, washed and treated to a final rinse in distilled water for 1 min.

Fig. 2. Transmission electron micrograph of an anchoring cell

in the bat, Preronotus p. pamellii. Note the insertion of ex-

tracellular filaments (EF) into conical adhesion plaques which

indent the cell surface. The intracellular filaments (IF) can be

seen to traverse the cell and to insert at both ends into the adhesion plaques. x 11000.

Finally the coverslips were drained, mounted in

gelvatol on glass slides and examined with a Leitz Orthoplan microscope equipped for epifluores- cence. Antibodies used have been characterized elsewhere: rabbit anti-beef cardiac muscle (Y-

actinin (Burridge and McCullough, 1980); rabbit anti-chicken gizzard talin (Burridge and Connell

1983a. 1983b); rabbit anti-chicken gizzard actin (Burridge, 1976). The myosin antibody was raised

in rabbits against purified chicken gizzard myosin. The staining properties on cultured fibroblasts and

its cross-reaction with purified smooth muscle or

fibroblast myosin were very similar to the anti-giz- zard myosin described by Gordon (1978; Burridge, unpublished results). The antibody against tropomyosin was raised in rabbits against purified chicken skeletal muscle tropomyosin. This anti-

body cross-reacts with smooth muscle and non- muscle tropomyosin (Burridge, unpublished re- sults).

Several antibodies were used which gave no

staining of the tissue above background. These

209

antisera included a rabbit antibody raised against

chicken gizzard vinculin, an autoimmune rabbit antiserum that reacted with vimentin (Gordon et

al., 1978) and an antibody against beef brain spectrin (fodrin) (Burridge et al., 1982). The lack

of staining observed with the anti-vinculin may

have been due to a lack of species cross-reactivity since this antibody reacted strongly with chicken

vinculin, but only very weakly with various mam- malian tissues (Burridge, unpublished results).

Results

As shown in Fig. 1, the thin band of anchoring

cells is closely applied to the internal surface of the otic capsule. These cells do not occur throughout

the entire extent of the spiral ligament-otic cap- sule junction but are concentrated in the central region of attachment. In areas where anchoring cells do not occur the extracellular fibers of the

spiral ligament insert directly into the bone and

there are generally few cells of any kind at the

Fig. 3. Phase contrast (A) and fluorescence (B) micrographs of the same field of a section through the otic capsule (OC) and spiral ligament of Pteronorus. This section has been treated with antibodies against tropomyosin. The same structures are marked at

identical points in each micrograph: AC. anchoring cell; EF, extracellular fiber; rbc, red blood cell in blood vessel.

210

junction. Thus, in sectioned material the position occupied by anchoring cells could easily be de- termined. Antibody-treated sections were first ex- amined with phase contrast microscopy and an area ~ontai~ng anchoring cells was identified and photographed; the selected field was then photo- graphed using fluorescence microscopy. Fig. 3 is an example of a pair of micrographs in which

sections of the cochlea of Ptermotus had been Created with antibodies against tropomyosin. It is clear that tropomyosin is primarily localized in the anchoring cells; adjacent cells of the spiral liga- ment are only weakly labelled and the large ex- tracellular fibers appear as dark holes or bands. The only other structures of the spiral ligament which are prominently labelled are blood vessels

Fig. 4. lmmunofluorescent labeling of the anchoring cells in Prerono~us after treating sections with antibodies against myosin (A) and a-actinin (B). Each illustration is a montage of five micrographs so that almost the entire extent of the band of anchoring ceils can be

seen.

211

and red blood cells. The distribution of tro] sin in the mouse is identical to that observec

bat.

?omyo-

j in the

In order to determine the distribution of the proteins over broader areas, montages of micro-

Fig. 5. Distribution of talin in the spiral ligament of tk

Note that the prominent fluorescence is limited to the

anchoring cells.

le mouse. : band of

Fig. 6. The spiral ligament shown in this montage has been

treated with antibodies against actin. Note that actin has a wide

distribution in the cells of the spiral ligament (bat)

graphs were assembled to include almost all of the

band of anchoring cells. Fig. 4 shows the fluores- cence images that were produced when the tissue

was treated with antibodies against cY-actinin and

myosin. In both cases intensely labelled cells are restricted to the zone where the only cellular ele-

ments are anchoring cells. The pattern of labelling with tissue treated with antibodies against

tropomyosin is identical to that of cy-actinin and

myosin in both the mouse and the bat.

In the mouse the fluorescence pattern obtained after the tissue was treated with antibodies against

actin and talin also conformed to the pattern of distribution of the other three proteins studied;

that is, they were restricted to the region of anchoring cells as shown in Fig. 5. In Pteronotus,

however, fluorescence micrographs of sections treated with antibodies against actin and talin

showed that these two proteins are much more widely distributed than the others and that their

distribution is almost identical. The extensive dis- tribution of actin in the bat is shown in Fig. 6; this can be compared with the band-like distribution of

myosin, tropomyosin and a-actinin (Figs. 3 and

4). Fluorescence patterns of cells treated with anti-

bodies against talin (Fig. 7) revealed numerous

spine-like processes and the general shape of the

images resembled those of cells seen in transmis- sion electron micrographs (Figs. 2 and 7). Since

the spine-like processes of the anchoring cells are

largely composed of adhesion plaques, it appears that talin may be more localized in this region than the other proteins.

Discussion

The main purpose of this study was to demon-

strate that the anchoring cells of the spiral liga- ment have the same complement of proteins as other cells which have the ability to contract or

generate tension. The results clearly show that the

anchoring cells, unlike adjacent cells of the spiral

ligament, contain abundant actin, a-actinin, myosin, tropomyosin and talin.

Many of these proteins are also found in other cells in locations where they may not be related to a contractile system. In particular, actin may have

a structural function as in microvilli and in the stereocilia and supporting cells in the ear (Flock et al., 1982; Flock and Cheung, 1977; Slepecky and Chamberlain, 1982, 1983; Tilney et al., 1980). In such locations actin is not associated with myosin. When it is associated with myosin, together with proteins such as tropomyosin and a-actinin, it is

usually involved in the generation of some form of

Fig. 7. TEM and fluorescence micrographs of anchoring cells. Spine-like processes are evident in tissue which has been treated with

antibodies against talin. Note that the shapes of individual cells are outlined. A, X 18000.

213

contraction or tension. This may lead to cell motil-

ity or the movement of organelles, but in some cases such as in the stress fibers of cultured

fibroblasts, shortening is restricted by tight ad- hesions of the cell to an extracellular matrix and

thus tension is generated. The generation of con- tractile or tensile forces is most efficiently pro-

duced by organized bundles of filaments such as those found in muscle or cultured fibroblasts. This appears to be the situation found in the anchoring

cells. In another study (Fitzpatrick and Henson, unpublished results) it has been shown that the intracellular fibers of the anchoring cells are com-

posed of actin. Triton-X 100 was used to disrupt the cell membranes and the filaments of the

anchoring cells were exposed to heavy meromyo-

sin; under these circumstances the filaments formed arrowhead complexes characteristic of actin

filaments. The extracellular filaments, by contrast, were not decorated. Our observations on the other

proteins, however, did not provide any informa-

tion about their precise localization in relation to the intracellular bundles. By comparison with stress

fibers in other cells, we suggest that these proteins are closely associated with the bundles of actin filaments.

The fact that talin is also a protein component of the anchoring cells and has a distribution iden- tical to that of actin is of some interest. Talin has

been identified in adhesion plaques and it is implicated in the attachment of bundles of actin filaments to the membrane (Burridge and Connell, 1983a. 1983b). In the anchoring cells it may be

that this is the mode of attachment of the actin

filaments to the cell membrane since talin was

found to extend into the spine-like adhesion plaque regions of the cells (Fig. 7). It cannot be stated, however, that this is the only protein involved in the attachment of actin filaments to the cell mem- brane. In other systems vinculin is also involved and our failure to detect this protein may have been due to a lack of immunological cross-reactiv- ity with these cells. The results reported here in relation to actin and talin in the anchoring cells are consistent with the hypothesis that actin fila- ments and talin are functionally linked.

We think that the intracellular filaments in anchoring cells are identical to, or closely resem- ble, the stress fibers which occur in cultured

fibroblasts. The most striking structural similari-

ties are in the diameter of the filaments (ca. 8 nm), their organization into parallel arrays and their

insertion into adhesion plaques. In addition, we

have shown that the same contractile associated proteins that occur in stress fibers also occur in the

anchoring cells. Although it was not possible to localize all of the proteins within the cell, it would

appear that the morphological and biochemical criteria for calling intracellular filaments ‘stress

fibers’ have been satisfied (Byers et al., 1984). Stress fibers in cultured fibroblasts appear to de- velop in response to tension (Burridge, 1981; Byers

et al., 1984) and they are able to apply tension to

the substrate (Harris et al., 1980, 1981). The con- traction of stress fibers is isometric and not under

nervous control; this is very similar to the condi- tions under which the anchoring cells probably

exist. There are two morphological indications that

the intracellular fibers in the anchoring cells apply

tension: (1) the extracellular fibers are often

sharply bent where they encounter the band of

anchoring cells (Henson et al., 1984); and (2) the conical shape of the adhesion plaques suggests that there is a tension applied to the internal surface of the plaques.

Although we would like to call attention to the similar nature of intracellular fibers of anchoring

cells and stress fibers of cultured fibroblasts, we do not intend to imply that the anchoring cells are fibroblasts or fibrocytes that have become at- tached to parts of the extracellular matrix in an

area where tension exists. The cells that we have

called anchoring cells have been called fibrocytes by other authors (Morera et al., 1980; Takahashi and Kimura, 1970) but the shape and ultrastruct-

ural features of the anchoring cells are not similar to fibrocytes in other connective tissues. One of the primary differences is that the extracellular fibers associated with the anchoring cells actually insert into the cells (Figs. 2 and 7). In this respect it is interesting to note that the conical adhesion

plaques show a morphological resemblance to myotendinous junctions (Tidball, 1984; Trotter et al., 1983). In muscle, the interface between the cell and the tendon is increased in area by the shape of the junction, and the greatest part of the interface is almost parallel to the direction of force. It is thought that the increase in area reduces stress on

the membrane, while the parahe1 direction of the interface causes the junction to be loaded in shear more than in tension (Tidball, 1983). These same properties seem to exist in anchoring cells. One must certainly question whether anchoring cells represent a new class of cells that are speciahzed for maintaining tension on the fibers of an ex- traceilular matrix.

The association of the i~tra~~lu~~ f~~arn~nts with extracellular filaments via adhesion plaques provides the basis by which contractions of the intracellular filaments can exert a force on the extraceflufar fibrous matrix of the spiral ligament. The functional implications in terms of radial ten- sion on the basilar membrane have been discussed in our previous paper (Henson et al., 1984). To summarize: (1) radial tension may be created in the spiral ligament-basilar membrane complex by the fibers within anchoring ceils; (2) radial tension has the capacity to affect the vibratory properties of the basilar membrane; and (3) changes in the mechanical properties of the basilar membrane that have been observed to occur at death or under anesthesia might be att~b~t~~ at least in part, to changes in the isometric contraction of fibers within the a~~ho~ng cells of the spiral ligament.

We would like to thank Randolph Wynne and Gferm Nu~koIls for expert ~hoto~aF~~ assis- tance. This work was supported by USFHS grants NS 12445, NS 19031 and GM 29860.

Burridge, K. (X976): Changes in c&.tiar glycoproteins after transformation: ~dentifi~tio~ of specific ~y~prot~~s and antigens in s&&n d~~l~fate geis. Proc. Natt Acad. Sci. 73,4457-4461.

Burridge, K. (1981): Are stress fibers contractiI& Nature (London) 294,691-692.

Burridge, K. and Cannel& L. (t983a): A new protein of ad- hesion plaques and ruftling membranes. f. CeiI RioI. 97, 359-367.

Burridge, K. and Canneli, L. (X983b): Talin: A cytwk&tal component concentrated in adhesion plaques and other sites of actin-membrane interaction, Ceil Motility 3, 405-417.

Burridge, K. and M~Cu~~~u~~ L. jl9g0): The asstxtation of u-actinin with the plasma membrane. J. Supramnl. Struct. 13, 53-65.

Burridge, K., Kelly, T. and Mangeat, P. (1982): Noner~~hro~yt~ spectrins: act~n-membrane attachment proteins murring in many ceil types. J. Cell Bioi. 95, 47X-486.

Byers, H.R., White, GE. and Fujiwara, K, (1984): Grganiza- tion and function of stress fibers in ceils in vitro and in situ. Celi Muscle Motitity 5, 83-l 37.

Fiock, A. and Cheung H,C. (1973): Actin f~~~rn~~ts in sensory hairs of inner ear receptor celis. J. CetI Bioi. 75, 339-343.

Flock, A., Bretscher, A. and Weber, K. (1982): lmmunohis- toehemicat tocatization of several cytaskeletal proteins in inner ear sensory and supporting cells. Hearing Res. 6, X-89.

Gordon, W.E. (1978): fmmunofluorescent and uhrastructurat studies of “sarcomeric”’ units in stress fibers of cultured non-muscle cells. Exp, Cell Res. 117, 253-268.

Gordon, WE., 3~shnel~, A. and Burridge, K. (f978): Char- acterization of intermediate (10 nm] filaments of cultured ceils using an autoimmune rabbit antiserum. Celi 13, 249-261.

Harris, AX., Witd, P. and Stopak, D. (1980): Siiiccme rubber substrata: a new wrinkle in the study of c&t focomotion. Science 208, 177-179.

Harris, A.K., Stopak, D. and Wild, P. (1982): Fibroblast trac- tion as a mechanism for collagen morphogenesis. Nature (London) 290,249-251”

Hensan, MM., Henson, 0.W. Jr. and Jenkins, D.B. (1984): The attachment of the spiral ligament to the cochlear wall: Anchoring celfs and the creation of tension. Hearing Res. 16,231-242.

Morera, C., dal Sasso, A, and furato, S. (1980): Submicroscopic structure of the spiral ligament in man. Rev. Laryngol. 101, 73-85.

Slepecky, N. and Chamberlain, S.C. (1982): Distribution and plant? of actin in the sensory hair c&Is of the chinchilhr cochlea. Cell Tissue Res. 224, 15-24.

Slapecky, N. and Chamberlain, SC. (1983); Distribution and polarity of actin in inner ear supporting cells. Hearing Res. IO, 359-370.

T~~~~sh~, T. and Kimura, R.S. (1970): The ultrastructure of the spiral ligament in the Rfiesus monkey. Acta Oto- Laryngol. 69, 46-60.

Tidball, J.G. (1983): The geometry of actin f~larn~t-m~mbr~e associations can modify adhesive strength of the myoten- dinous junction. Gel1 Motiiity 3,439-447.

Tidbah, J.G. (1984): ~~ote~dinous junction: ~o~holo~cai changes and mechanical failure associated with muscle deli

atrophy. Exp. Mot. Pathot. 40. I-12. Titney, L.G., Derosier, Df. and Mulroy, M.J. j1980): The

~r~~i~t~on of actin filaments in the stereocilia of cochlear hair cells. J. Cell Biol. 86, 244-259.

Trotter, J.A., Eberhard, S. and Samora, A, (1983): Structural connections of the muscle-tendon jUrtcfiOn. Cd Motihty 3, 431-438.