Distribution of KHRI 3 epitopes in the inner ear

Hearing Research, 66 (1993) 245-252 0 1993 Elsevier Science Publishers B.V. All rights reserved 0378-5955/93/$06.00

HEARES 01897

245

Distribution of KHRI 3 epitopes in the inner ear

Martin Ptok a, Thankam Nair b, Thomas E. Carey b and Richard A. Altschuler ’ ‘I HNO-Hearing Research Laboratories, Department of Otolaryngology, University of Tiibingen, Tiibingen, FRG and I’ Kresge Hearing Research

Institute, Ann Arbor, Michigan USA

(Received 20 April 1992; Revision received 6 November 1992; Accepted 13 December 1992)

Murine monoclonal antibodies against cochlear structures were previously generated to obtain probes for elucidating the function of cochlear

cell subsets. Preliminary immunocytochemical characterization showed that the monoclonal antibody KHRI 3 binds to supporting cells but not

sensory cells in the guinea pig cochlea. We have now investigated KHRI 3 epitopes in other species and other parts of the inner ear. The KHRI 3

epitope appears to be species-specific since no immunolabeling was seen in rat inner ear nor in chick inner ear. In immunocytochemical assays in

the guinea pig vestibular tissues KHRI 3 stained saccular wall cells and transitional epithelial cells in the utricle and ampules as well as clusters

of cells in the endolymphatic sac. In Western blots KHRI 3 stained a broad 70-75 kDa band in lanes loaded with guinea pig cochlea

homogenates - as seen previously - as well as in lanes loaded with vestibular tissue homogenates. The immunolabeling patterns suggest that

KHRI 3 epitopes are cell membrane components or related to membrane structures. Thus the monoclonal antibody KHRI 3 appears to define a

nonsensory cell subset in the guinea pig inner ear that can be identified by expression of KHRI 3 epitopes.

Monoclonal antibody; Inner ear; Supporting cells; Immunocytochemistry; Western blotting

Introduction

To evaluate if inner ear structures possess unique proteins we previously generated murine monoclonal antibodies (mAbs) against guinea pig cochlear epithe- lium (Zajic et al., 1991). Preliminary characterization showed that the mAb KHRI 3 identifies an antigen that is specifically expressed by supporting cells in the guinea pig cochlea. In Western blots of detergent ex- tracts from guinea pig cochleae KHRI 3 stained a tissue-specific broad band of 70-75 kDa. The KHRI 3 epitope is also expressed in peripheral nerves. How- ever this epitope may be expressed on a different protein than that in the cochlea since KHRI 3 stained a M, 68-70 band in lanes loaded with non-auditory tissues containing peripheral nerves. The purpose of this study was to evaluate the expression of KHRI 3 epitopes in other inner ear tissues including the guinea pig saccule, utricule, ampule and endolymphatic sac as well as inner ear tissues of other rodents and birds using immunocytochemistry and Western blots.

Correspondence co: Martin Ptok, HNO-Hearing Research Laborato-

ries, Department of Otolaryngology, University of Tiibingen, Silcher-

strasse 5, W 7400 Tiibingen, FRG. Fax: (49) 7071 293-311.

Methods

Animals

Pigmented guinea pigs (200-300 g) obtained from Murphy’s Breeding Lab, Plainfield, IN, or from Fa. Schmid, Holzgerlingen, FRG, chicken 3-7 days after hatching obtained from Omega Chicks, Hazlett, Ml, and mature rats (Long-Evans rats, Charles River, Portage, MI) were used. All animals were kept under the surveillance of the Unit of Animal Medicine, Uni- versity of Michigan or under the surveillance of the veterinarians of the University of Tiibingen.

Immunocytochemistry

Preparation of tissues Guinea pigs and rats were deeply anesthetized with

UrethaneR and perfused through the heart with PBS followed by 2% paraformaldehyde in PBS, pH 7.4 at 4°C. Temporal bones including cochleae were re- moved, fixed locally with the same fixative through the round window and kept in 2% paraformaldehyde for 30 min. For immunofluorescence and immunoperoxi- dase procedures, bony shell, vascular stria, tectorial membrane and Reissner’s membrane were removed leaving spirals with the organ of Corti exposed which were rinsed in PBS at least 1 h, described in detail

246

previously (Fex and Altschuler, 1981, 1986; Altschuler et al., 1985a,b; Fex et al., 198.5). From the remaining temporal bone the ampules, saccule and utricle were dissected (as described earlier in Ptok et al., 1991). The saccular wall and the roofs of the ampules and the utricle were opened. With a capsule knife otoconia, gelatin layer, subcupular meshwork and cupula were removed to allow antibody access to cell surfaces. The vestibular organs were left in fixative for 30 min.

Chicken were deeply anesthetized with Urethane@ and killed by CO, inhalation. After decapitation the head was sectioned in the mid-sagital plane, the basilar papilla identified and opened immediately to allow access for the fixative. Fixation time was 30 min. The inner ear organs were dissected as described previously (Ptok et al., 1991).

Immunojluorescence procedures Prior to each incubation the specimens were washed

three times for 3 min in PBS. Blocking of unspecific second antibody binding sites was achieved by incuba- tion with 3% horse serum for 10 min. In assays using chick inner ear organs blocking was performed by incubation with avidin and biotin using an avidin/bio- tin blocking kit (Vectastain ABC Kit; Vector Laborato- ries, Burlingame, CA) according to the manufacturer’s specification: Specimen were incubated with avidin D for 15 min, rinsed briefly with buffer and incubated with biotin blocking solution for 15 min. These steps preceded the addition of primary antibody. Excess horse serum was removed from free floating whole mount preparations or blotted from slides and the tissue was incubated for 30 min with primary antibody (diluted either 1 : 1, 1: 5, 1: 10, 1: 20, 1: 50, 1: 100, 1: 200 with PBS containing 0.03% Triton X-100). Con- trols were performed by replacing IU-IRI 3 either with PBS or with the supernatant from the partner myeloma line SP2/0 that was used for the fusion. The SP2/0 supematant was chosen as a control to detect possible artifacts due to the culture medium or metabolic prod- ucts. The biotinylated second antibody (anti mouse IgG) was diluted 1: 200 with PBS. All incubations were carried out for 30 min at room temperature.

Slides were counterstained with Hematoxylin (Sigma, St. Louis, MO), dehydrated in ethanol (70% for 5 min, 100% twice for 5 min, dipped in xylene) and mounted wet in Krystallon (EM Diagnostic Systems, Gibbstown, NJ).

Immunoperoxidase procedures Labeling was performed following the protocol de-

scribed for immunofluorescence, except that incuba- tion with avidin conjugated rhodamine was substituted by incubation with ABC complex (Vector Laboratories, CA): The specimens were incubated with the Vectas- tain ABC reagent (Vectastain ABC Kit; Vector Labo-

ratories, Burlingame, CA) (Hsu et al., 1981) for 45 min, washed and stained for 5 min in peroxidase substrate solution (Diaminobenzidine 10 mg diluted in 20 ml

PBS containing 70~1 H,O, 3%). Some of the surface preparations were dehydrated

in alcohol and embedded in EmBED 812 (Polyscien- ces, Warrington, PA). Then these specimens were sec- tioned at 4 Frn and mounted in Krystallon.

Documentation Peroxidase (DAB) stained slides were examined and

photographed on a Leitz Dialux microscope with a Kodak T-max 100 film. Fluorescence slides were exam- ined using a Leitz Orthoplan fluorescence microscope and photographed with a Kodak T-max 400 film.

Immunoblotting

To determine the molecular weight of the molecule expressing the epitopes of the antibodies described here, one-dimensional polyacrylamide gel elec- trophoresis in sodium dodecyl sulfate (SDS-PAGE) was performed according to the method of Laemmli (1970) under both reducing and non-reducing condi- tions.

The tissues (either cochlea, saccule plus utricle or endolymphatic sac) were solubilized on ice for 30 min in lysis buffer (1% NP-40 in PBS, protease inhibitors including 1 nM phenyl methyl sulfonyl fluoride-PMSF, leupeptin lpg/ml, antipain 2pg/ml, benzamidine 10 pg/ml, aprotinin 10 kp/ml, chymostatin lpg/ml, pepstatin lpg/ml) and clarified by centrifugation (11,000 x g) for 5 min at 4°C. For reduced gels tissue homogenates were boiled for 5 min in sample buffer containing 2-mercaptoethanol. Homogenates of tissues and a molecular weight standard (Rainbow-standard, Amersham, Arlington Heights, IL) were mixed with loading buffer (Tris-HCl, pH 6.8, 2% SDS, 10% glyc- erol, 0.005% bromophenol blue) to a final concentra- tion of 1OOpg protein/sample/lane and loaded onto a 3% stacking and 7% separating SDS gel. The final protein concentration corresponded to approximately four cochleae, 2 saccules plus 2 utricles plus 2 ampules or 8 to 10 endolymphatic sacs. The electrophoretically separated proteins were transferred onto nitrocellulose paper (Millipore Corporation, Bedford, MA) elec- trophoretically (Towbin et al., 1979). An immunoblot assay kit (Bio-Rad Lab., Richmond, CA) was used following the manufactures instructions: after elec- trophoretic transfer the membrane was cut into strips and then immersed in Tris-buffered saline (TBS) pH 7.5 for 10 min. Blocking of unspecific binding sites was performed by incubating the nitrocellulose filter in TBS containing 1% gelatin for 30 min. After incuba- tion with primary antibody diluted in Tween ZO/TBS 1: 20 and subsequent washes the nitrocellulose was

incubated with gold conjugated goat-anti-mouse sec- ondary antibody diluted 1: 25 in dilution buffer con- taining TBS with 0.05% Tween 20 and 0.4% gelatin. After red bands appeared the membrane was washed with ddH,O and a gold enhancement procedure car- ried out: after washes in 0.2 M citrate buffer, pH 3.7, the enhancement solution (0.85 g of hydroquinone in 90 ml citrate buffer) mixed with 0.11 g of silver lactate in 10 ml ddH,O was applied for 5-15 min and finally fixed with solution provided with the kit.

The remaining gel was stained with Coomassie-blue to check if protein transfer was sufficient.

Results

Immunocytochemistry

Results in guinea pig cochleae were as we have previously described (Ptok et al., 1989a; Zajic et al., 1991) (Fig. 1): The most striking immunolabeling was found proximal to the IHC corresponding to the apical surface of inner border cells and at the phalangeal processes of outer pillar cells (‘wine-glass’ pattern), primarily at the borders of these cells. Inner sulcus cells and the phalangeal processes of Deiters’ cells were stained in a dotted pattern, but less prominent. KHRI 3 immunolabeling was also found in the guinea pig vestibular system. In the saccule KHRI 3 immunos- tained saccular wall cells. In surface preparations KHRI

247

3 binding ‘outlined’ the saccular wall cells resulting in a honeycomb-like immunostaining (Fig. 2A, left panel). No immunobinding was observed in the macula.

A similar honeycomb-like immunostaining was seen in surface preparations of the utricle (Fig. 2B, left panel) and all three ampules (Fig. 2C, left panel). In these organs a band of cells between the area of sensory cells and the area of dark cells were im- munopositive. No immunolabeling was found in the hair cell area (macula utriculi or crista ampullaris) nor in the roof of the utricle or the planum semilunatum. When these tissues were incubated with either PBS (results not shown here) or SP 2/O supernatant no significant labeling was seen (Fig. 2, right panels).

In the guinea pig endolymphatic sac clusters of cells were immunolabeled (Fig. 3 left panel): In surface preparations cell borders were immunolabeled. Addi- tionally immunolabeled punctae at the apical surface of these cells were seen. At the light microscopic level, immunolabeled cells did not exhibit anatomical fea- tures which would permit classification of these im- munolabeled cells as either light or dark cells. No labeling was observed when the primary antibody was replaced by either PBS or SP2/0 supernatant (Fig. 3 right panel).

In cross-section of all vestibular organs including the endolymphatic sac immunobinding appeared to be not only at the luminal side of the cells but also distributed all around the cell border in a pattern resembling the cell membrane. Fig. 4 (left panel) shows such a typical

Fig.

four

1. KHRI 3 immunolabeling in the cochlea: Surface preparation of a guinea pig cochlea, middle turn. The most striking immunolabeli

Id proximal to the IHC and at the phalangeal processes of outer pillar cells (‘wine-glass’ pattern). Inner sulcus cells and phalangeal plal

Deiters’cells are stained in a dotted pattern, but less prominent (original magnification 400 X ).

Ing is :es of

immunolabeling pattern in a cross-section of the guinea pig utricle (Fig. 4 right panel: control section incubated with SP2/0 supernatant).

endogeneous avidin activity with avidin D/biotin co pletely abolished staining (Fig. 5B).

KHRI 3 immunolabeling was not found in either rat or chick inner ear. An initially observed labeling of supporting cells in the basilar papilla was found to be an unspecific effect due to insufficient blocking of endogeneous avidin activity (Fig. 5A). Blocking of the

Western blotting

Proteins either from guinea pig cochlea, from s, cule, utricle and ampules or from the endolympha sac were separated by SDS-PAGE and transferred

lrn-

ac- itic to

Fig. 2. KHRI 3 immunolabeling in the guinea pig vestibular organs (original magnification 400X): In saccuie surface preparations (A), utr surface preparation (B) and ampule surface preparation (C) a ‘honeycomb-like immunolabeling was found (left panel). In the saccule only t of the saccular wall were immunolabeled. In the utricle and in the ampules transitional epithelial cells (cells between the sensory epithelium

the region of dark cells were immunolabeled. Fig. (A-C) right panel: corresponding controls incubated with SP 2/O supernatant.

ricle cells and

Fig. 2 (continued).

nitrocellulose. Under non-reducing conditions im- munostaining of nitrocellulose strips loaded with pro- teins from the tissues mentioned above revealed M, 70-75 kDa bands in the cochlea as shown previously (Ptok et al., 1991) (Fig. 6, lane A) and in lanes loaded with saccule, utricle, and ampule homogenates (Fig. 6, Lane 0. No immunostaining was seen in the endolym- phatic sac homogenate lanes (Fig. 6, lane E) or when the primary antibody was replaced by either PBS (re- sults not shown) or SP 2/O supernatant (Fig. 6, lane B,D,F).

Under reducing conditions no immunolabeling was seen (results not shown).

Discussion

As described previously polyclonal (Ptok et al., 1989a; Orozco et al., 1990) or monoclonal antibodies (Ptok et al., 1989a, 1991; Zajic et al., 1991) can be generated to cochlear tissue. One of the mAbs binding to cochlear structures - KHRI 3 - appears to bind to

Fig. 3. KHRI 3 immuno~abeling in the endoiymphatic sac: In the end(~lymphatic sac clusters of celis were found to be immunopositive. In surface

preparations the cell borders of immunopositive cells were labeled intensely. Additionally a punctate staining at the apical surface of

immunopositive cells was observed left panel). (Surface preparation, original magnification 400X ). Right panel: Control incubated with SP 2/O

supernatant.

250

Fig. 4. Cross-section of a guinea pig utricle, immunolabe~ed with KMRI 3: ~mmunolabeling appeared to be around the whole ceil body, possibly associated with a membrane structure. Shown here is a cross section of the macula uticuli (left panel) (corresponding control right panel).

(Immunoperoxidase labeling, connterstaining with hematoxilin; original magnification 1000X ).

Fig. 5. Chick basilar papilla, immunostained with KHRI 3: Supporting cells exhibited labeling (A). However, biocking the endogeneous avidin activity with biotin (B) completely abolished ‘immunolabei~ng’, thus indicating that labeling was due to binding of endogeneous avidin to the

biotinylated second antibody. (Original magnification 1000 x ).

251

Fig. 5 (continued)

ACE BDF

46-

29-

Ftg. 6. Western blot (non-reducing conditions) of cochlear, vestibular

and endolymphatic sac homogenates immunostained with KHRI 3

and SP 2/O. A tissue specific band corresponding to 70-75 kD was

detected in lanes loaded with either cochlea (A) or utricle, saccule

and ampules CC’). No staining was observed in lanes loaded with

endolymphatic sac homogenates (E) or when tissues were incubated

with SP 2/O supernatant (B,D,F).

an antigen specific to the cochlea. Previous experi- ments also showed immunobinding of KHRI 3 in pe- ripheral nerves. However the cochlear antigens might be carried on a different protein than the antigen in the peripheral nerves since the two antigens behave differently when separated with SDS-PAGE. Alterna-

tively the epitope might be on the same protein modi- fied e.g. by different levels of glycosylation in either the cochlea or the peripheral nerve. Furthermore the

cochlear epitope proved to be stable to paraformalde- hyde fixation whereas in nerves immunolabeling was abolished by paraformaldehyde fixation. While these results indicate that KHRI 3 specifically binds to an epitope in the cochlea it remained to be determined if this epitope is expressed in other parts of the guinea pig inner ear, and secondly if this epitope is expressed in inner ear of other species.

KHRI 3 binding was found in the saccule, utricle, ampules and in the endolymphatic sac using immuno- cytochemical assays. Yet immunolabeling in vestibular organs differed from immunolabeling in the cochlea. In the cochlea the most prominent immunolabeling was found in supporting cells in close approximation to sensory cells, namely in inner border cells, outer pillar cells and Deiters’ cells. Supporting cells close to or

252

between vestibular hair cells lacked immunolabeling, instead a population of supporting cells bordering the neuroepithelial sheet was immunopositive.

In the cochlea immunolabeling was found at cell borders facing a lumen - either an extracellular space (at the reticular lamina) or a broad intercellular space (tunnel of Corti). In the vestibular system including the endolymphatic sac immunolabeling was found all ‘around’ the labeled cells including cell borders not facing a lumen.

In the cochlea KHRI 3 showed a punctate staining pattern. Though in the vestibular system the immuno- labeling was also present as a punctate staining the most prominent immunolabeling was found in a homo- geneous distribution along the borders of transitional epithelial cells in the utricle and ampules, at borders of saccular wall cells and at borders of a subset of en- dolymphatic sac cells.

While the punctate staining pattern may correspond to binding to microvilli as suggested by our previous findings (Ptok et al., 1989b) ultrastructural analysis will be necessary to determine which structure of the cell border is carrying the KHRI 3 epitope (thus indicating molecular relationship with microvilli). Ultrastructural investigations will also enable us to determine if the KHRI 3 epitope is an intra- or extracellular compo- nent.

Despite different morphological localization of the antigens the cochlear epitope and the vestibular epi- tope may be comprised of the same molecular struc- ture, since the electrophoretic mobility is identical. However, based on these results the absence of im- munopositive bands in lanes loaded with endolym- phatic sac homogenates cannot be explained.

While sensory cells have long been in the focus of inner ear research, less is known about the supporting cells in the inner ear, although it can be assumed that their presence is important for normal hearing. This study adds new insight into the molecular composition of cochlear and vestibular supporting cells. As indi- cated by immunolabeling with KHRI 3 cochlear inner border cell, pillar cells and vestibular transitional ep- ithelial cells express a common molecular structure. The mAb KHRI 3 appears to identify a subset of supporting cells in the guinea pig inner ear. This is the first report classifying supporting cells of the inner ear by their molecular characteristics. Further experiments will assess whether the expression of KHRI 3 epitopes common to cells of this subset is paralleled by a spe- cialized cell function. If so it remains to be determined

if this cell function is unique to the guinea pig inner ear.

Acknowledgements

Supported by DFG (Postdoctoral Fellowship for M.P.) and NIH (Program Project Grant NS-0.5785).

References

Altschuler, R.A., Hoffman, D.W., Reeks, K.A. and Fex, J. (1985a) Localization of dynorphin B-like and alpha-neoendorphin-like immunoreactivities in the guinea pig organ of Corti. Hear. Res. 17, 249-258.

Altschuler, R.A., Reeks, K.A., Marangos, P.J. and Fex, J. (1985b) Neuron-specific enolase-like immunoreactivity in inner hair cells but not outer hair cells in the guinea pig organ of Corti. Brain. Res. 327, 379-384.

Fex, J. and Altschuler, R.A. (1981) Enkephalin-like immunoreactivity of olivocochlear nerve fibers in cochlea of guinea pig and cat. Proc. Nat]. Acad. Sci. USA 78, 1255-1259.

Fex, J. and Altschuler, R.A. (1986) Neurotransmitter-related im- munocytochemistry of the organ of Corti. Hear. Res. 22,249-263.

Fex, J., Kachar, B., Rubio, J.A., Parakkal, M.H. and Altschuler, R.A. (1985) Glutaminase-like immunoreactivity in the organ of Corti of guinea pig. Hear. Res. 17, 101-113.

Hsu, S.M., Raine, L. and Fanger, H. (1981) Use of avidin-biotin- complex (ABC) in immunoperoxidase techniques: a comparison between ABC and unlabeled antibody (PAP) procedures. J. His- tochem. Cytochem. 29, 577.

Laemmli, U.K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680.

Orozco, C.R., Niparko, J.P.. Richardson, B.C., Ptok, M. and Altschuler, R.A. (1990) Experimental model of immune mediated hearing loss employing cross-species immunization. Laryngoscope 100, 941-947.

Ptok, M., Orozco, CR., Zajic, G., Rajan, T., Schacht, J., Carey, TX. and Altschuler, R.A. (1989a) A&k&per gegen Innenohrstruk- turen: VorlIufige Ergebnisse tierexperimenteller Untersuchun- gen. Arch. Otorhinolaryngol. (Suppl. II), 73.

Ptok, M., Raphael, Y., Carey, T.E., Schacht, J. and Altschuler, R.A. (1989b) Immunocytochemical characterization of changes in the reticular lamina after inner ear trauma. 26. Workshop on Inner ear Biology, Paris, France, P 50.

Ptok, M., Nair, T.S., Schacht, J., Altschuler, R.A. and Carey, T.E. (1991) Monoclonal antibodies to inner ear antigens: II. Antigens expressed in sensory cells. Hear. Res. 57, 79-90.

Towbin, H., Staehlin, T. and Gordon, J. (1979) Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets. Procedure and some applications. Proc. Natl. Acad. Sci. USA 73, 2599.

Zajic, G., Nair, T.S., Ptok, M., Van Waes, C., Altschuler, R.A., Schacht, J. and Carey, T.E. (1991) Monocional antibodies to inner ear antigens: I. Antigens expressed by supporting cells of the guinea pig cochlea. Hear. Res. 52, 59-72.