- Email: [email protected]
Establishment of inner ear epithelial cell culture: Isolation, growth and characterization
211
Henring Research, 38 (1989) 277-288
Elsevier
HRR 01187
Establishment of inner ear epithelial cell culture: Isolation, growth and characterization ICE. Rarey and K. Patterson Departments
of Anatomy
and CeN Biology, and Surgery, College of Medicine, University of Florida, Gainesville, Florida, U.S.A.
(Received 13 June 1988; accepted 21 November 1988)
Select epithelial regions of the bovine inner ear were established and maintained in cell culture. Mkginal cells from the stria vascularis and dark cells from the posterior wall of the utricle were isolated, dissociated and placed in culture medium. Within 24 h, cellular islands of hexagonal-shaped, epithelial-like cells from both the stria vascularis and posterior utricular wall were readily identifiable by inverted light microscopy. Ultrastructural examination of both the cultured stria marginal cells and utricular dark cells revealed that both cell types had numerous microvilli on their apical surfaces and interdigitating infoldings of their basolateral surfaces. Apical tight junctional complexes were present between apposing cells. These findings demonstrate that inner ear bovine epithelial cells can be successfully isolated and maintained in culture, and that such cells retain certain of their in viva morphological characteristics Inner ear; Bovine; Stria vascularis; Dark cell; Epithelial; Cell culture
Introduction Understanding the individual roles of certain inner ear cells has been the focus of studies which examine cellular mechanisms associated with inner ear microhomeostasis. To better define the individual roles of certain inner ear tissues and/or cells, various techniques have been employed for isolation and characterization. These techniques have included cultures of the otic vesicle (Lawrence and Merchant, 1953), cochlear and vestibular tissues (Ann&o et al., 1979; Anniko and Sobin, 1983; Okano and Iwai, 1975; Okano et al., 1975; Russell and Richardson, 1987; Van De Water and Ruben, 1974; Yamashita and Vosteen, 1975), temporal bone cells (Maurizi et al., 1983), and stria marginal cells (Lim and Flock, 1985). Although these techniques have provided insight into the in vitro response of the various cells, a disadvantage
has been the limited viability of these tissues or cells. Primary culture of the endothelial cells from stria vascularis and spiral ligament tissues has been reported (Bowman et al., 1985). These cultures provide the opportunity for long term study of the cellular responses of the inner ear endothelial cells as they relate to the structural and functional integrity of the blood-labyrinthine barrier. To determine if isolated nonsensory inner ear epithelial cells, which form the perilymph-endolymph barrier, could be maintained in vitro, select epithelial regions were dissociated and placed in culture. Results indicate that cultures of bovine stria marginal cells and utricular dark cells can be established in vitro and that these cultured epithelial cells retain several of their in vivo morphological characteristics. Materials and Methods
K.E. Rarey, Box J-235, J. Hillis Miller Health Science Center, University of Florida, Gainesville, FL 32610, U.S.A. Correspondence to:
0378-5955/89/$03.50
Bovine temporal bone specimens were harvested with an oscillating bone-plug cutter and placed on
0 1989 Elsevier Science Publishers B.V. (Biomedical Division)
278
ice within 30 min of sacrifice. Each temporal bone was drilled with a dental engine under a dissecting microscope until the inner ear bony labyrinth was identified. Cold phosphate buffered saline (PBS) solution was dripped onto each temporal bone while drilling to remove bone dust and to reduce possible damage to the tissues from the heat discharge of the drill bits. Each turn of the lateral cochlear wall was identified and dissected from adjacent cochlear tissues and placed in Alpha Minimal Essential Medium (MEM) (Gibco) with ribonucleosides, deoxyribonucleosides and the following antibiotics: gentamicin (40 pg/ml), penicillin (50 units/ml), streptomycin (50 pg/ml), and Fungizone (2.5 pg/ml) (Gibco). Stria vascularis tissue from all turns of the cochlea was gently dissected from the spiral ligament with fine forceps and placed in fresh medium. In addition, the bony vestibule was opened and the posterior wall of the utricle was removed. Samples of stria vascularis and posterior utricular wall tissues from four bovine temporal bones were pooled respectively. Five to 10 mg (wet weight) of stria vascularis and 1-2 mg (wet weight) of posterior utricular wall tissue were routinely obtained. Epithelial cells from the two tissue samples were dissociated either by treatment with 1% Collagenase/Dispase (Boehringer, Mannheim) or 0.125% trypsin (Gibco) in MEM Alpha for 1.5-3 h at 37” C. Following dissociation, the samples were centrifuged at 300 X g, rinsed and plated onto 12-well polystyrene culture plates (Costar). The medium used to rinse and culture was Alpha MEM with ribonucleosides and deoxyribonucleosides, supplemented with 10% fetal calf serum, gentamicin (5 pg/ml), penicillin (25 units/ml), streptomycin (25 pg/ml), Fungizone (5 pg/ml), and 30 mM Hepes buffer (Gibco) at pH 7.4. The cultured epithelial cells were grown at 37 o C, with maximal humidity in 5% CO,/95% air. Cultures were examined daily by inverted microscopy. The culture medium was changed weekly. In order to subculture the two epithelial cell types, monolayers of each cell type were rinsed with a 0.05% trypsin solution containing 0.02% EDTA, 6 mM NaOH, 122 mM NaCl, 5.1 mM KCl, 0.8 mM Na,HPO,, 0.2 mM KH,PO,, 0.1 mM anhydrous dextrose, 0.03% NaHCO, and 30 mM Hepes buffer. After 30 s, the solution was
discarded and approximately two additional milliliters were added to the well. After 4 min unattached cells were removed with a sterile pipette, gently aspirated, and distributed into new wells. Fresh culture medium was then added to each well. Transmission electron microscopy Cultures were rinsed in three changes of PBS and fixed for 2 h in a 2.5% glutaraldehyde phosphate buffered (pH 7.4) solution. The epithelial monolayers were then rinsed in buffer and postfixed in a 1% osmium tetroxide phosphate buffered (pH 7.4) solution. After dehydration in graded concentrations of ethanol, the monolayers were embedded in Epon/Araldite. The embedded monolayers were cut from the culture wells with a dremmel saw and ultra-thin sections of the cultured stria and dark cells were made. Sections were stained with uranyl acetate and lead citrate and examined in either a Philips 400 electron microscope or a JEOL 100s electron microscope operated at 60 kV. Freeze-fracture To determine whether the cultured epithelial cells developed tight junctions, some epithelial monolayers were grown on polystyrene coverslips and fixed with a 2.5% glutaraldehyde phosphate buffered (pH 7.4) solution for 1 h, and then placed in 20% glycerol phosphate buffered saline as described by Pauli et al. (1977). Areas for freeze-fracture were selected under an inverted microscope and punched from the plastic coverslip with a 3 mm punch. The 3 mm disks were placed on specimen carriers and rapidly frozen in liquid nitrogen. The monolayers were fractured with a double replicating device in a Balzers high vacuum freeze-fracture unit at - 115O C. Platinum-shadowed carbon replicas of the fractured faces were placed in Chlorox bleach for 24 h. The replicas were rinsed in double distilled water, placed on grids and examined in a Philips 400 or JEOL 100s electron microscope operated at 60 kV. Results Digestion of stria vascularis and posterior utricular wall tissues in 1% collagenase-dispase for 3 h
279
Fig. 1. Photomicrograph Primary
of an island of stria marginal culture, 5 days. X 2140.
cells.
at 37O C under agitation normally produced 3-5 individual islands of epithelial-like cells within 72 h (Figs. 1 and 2). Generally these epithelial islands were quickly encircled and overgrown by fibroblast-like cells. On occasion, when the growth of the fibroblast-like cells sufficiently hindered the expansion of the epithelial colonies, the monolayers were treated with cold (4O C) 0.05% trypsin for 1 min. Because the epithelial cells adhered more strongly to the culture wells than the fibroblast-like cells, most fibroblasts could be removed
Fig. 3. Electron
micrograph
of stria marginal
Fig. 2. Photomicrograph Primary
of an island of utricular culture, 3 days. x 1875.
dark
cells.
without disrupting the epithelial cells. Following three brief rinses (15 s each) with sterile water, the epithelial monolayers were again incubated in growth medium. A modified digestion of inner ear tissues, using 0.125% trypsin for 1.5 h instead of 3 h, produced from 5-20 colonies, consisting of 5-10 epithelial cells in less than 18 h. By using this modified digestion, the epithelial cells initiated mitotic activity much sooner than did cells digested by other methods. As a result epithelial islands grew larger
cells after 7 days in primary culture. Several microvilli (A) surfaces of the epithelial cells. X 12 880.
can be identified
on the apical
280
Fig. J. Electron micrograph of an apical tight junctional complex (arrow) between two adjacent stria marginal cells. Observe the ““m erous lateral infoldings of the plasma membranes directly beneath the junctional complex. A = apical surface. Primary cult ure, 21 days. x 60 350.
and at a more rapid rate before being encircled by fibroblasts. In addition, it was easier to remove unwanted fibroblast-like cells from the culture wells with this modified digestion protocol. Insulin (5 pg/ml), transferrin (5 pg/ml), prostaglandin E, (25 ng/ml), hydrocortisone (50 nM) and cholera toxin (1 ng/cc) were added singly and in combination to the growth media to enhance the growth of the isolated cells. The addition of such agents did not increase the cell growth of the inner ear cells. The inner ear epithelial cells were readily identifiable in primary culture by their typical hexagonal cellular appearance under inverted phase-contrast microscopy. Utricular epithelial cells grew
more quickly than the marginal cells of the stria vascularis. Nonepithelial cells appeared fibroblastlike with no indication of cell-to-cell confluency. Electron microscopy of the stria and utricular epithelial cells revealed characteristic properties which are present in vivo. Microvilli were seen on the apical surfaces of the cultured epithelial cells (Fig. 3). Interdigitating infoldings were observed on the basolateral surfaces of both the stria and utricular cells (Figs. 3, 4 and 5). In addition, zonula occludentes were observed between apposing epithelial cells (Figs. 4 and 5). Below the tight junctional complexes, desmosomes were seen (Fig. 5). When both cell types were processed for freeze-fracture, strands of particles were observed
281
Fig. 5. Electron micrograph of another representative junctional complex (opened arrow) between two stria marginal Desmosomes (solid arrow) can be observed beneath the complex. A = apical surface. Primary culture, 21 days. X 92 300.
in the region of the zonula occludentes just beneath the apical surfaces of the cultured utricular and stria epithelial cells (Fig. 6). Subculture of primary epithelial monolayers was routinely performed with cold (4°C) 0.05% trypsin (Figs. 7a-e). Stria and utricular epithelial cells began readhering within 30 min after transfer (Fig. 7b). By 24 h (Fig. 7e) several hundred islands
of up to 100 cells each were typically observed. These subcultured cells appeared to retain their primary morphological characteristics through several passages. A foamy or bubble-l&e appearance (Figs. 8 and 9) was typical of the passaged stria and utricular cells. As in primary culture, the utricular epithelial cells grew at a faster rate than those of the stria vascularis, reaching
Fig. 6. F:reeze-fracture
replica of the organization of a tight junctional complex (zonula occludens) between two vestibular Parallel strands of particles on the P face (PF) can be seen. Primary culture. 21 days. x 46750.
confluence (380 mm2) in about 10 days. Subcultured stria epithelial cells grew to confluence in approximately 15 days. Discussion This study reports the isolation, growth and characterization of cultured bovine epithelial cells from the stria vascularis and posterior wall of the utricle. Both epithelial cell types were established in primary culture for 8-10 weeks. Because of the viability of such cells, they were passaged and observed for an additional 22-26 weeks. Contrary to the apparent reported ease of culturing epithelial cells from other tissues, there were several inherent obstacles in establishing cul-
dark
tures of inner ear epithelial cells. These obstacles were the small amount of tissues that could be harvested and prepared for cell culture, and the species from which the inner ear tissues were obtained. Due to the small quantity of stria and posterior utricular wall tissues, specimens had to be pooled from more than one animal. In addition, it was found that bovine inner ear epithelial cells responded better in culture than those from other species, e.g.. guinea pig (unpublished results). Initially a 1% collagenase-dispase solution was used to digest the inner ear tissues for 3 h. However, cells of the stria vascularis and posterior utricular wall did not rapidly dissociate. Therefore, a 0.125% trypsin solution was used to digest
283
Fig. 7. Time-lapse photomicrographs min al ‘ter 0.05% trypsin treatment.
of utricular epithelial cells being subcultured. (a) 15 nun after 0.05% trypsin treatment. (b) 20 (c) 30 min after 0.05% trypsin treatment. (d) 30 min after 0.05% trypsin treatment. (e) 24 h after 0.05% trypsin treatment. (a-d) X 560; (e) X 520.
284
Fig. 8. Photomicrograph of subcultured stria epithelial cells. These epithelial ceils had been maintained in primary culture and subcultured for 14 days. Typically, the cells developed a foamy-like appearance (arrows). X 880.
the tissues. This solution appeared to increase the dissociation of the stria and utricular cells in a shorter period of time (1.5 h). Because the cells were more readily dissociated, they could be placed in growth media more quickly. This step was thought to be crucial in that the cellular viability was considered better with a shorter dissociation period. The epithelial cells from both the bovine stria and utricular tissues grew better in Alpha MEM containing ribonucleosides and deoxyribonucleosides than in Alpha MEM without these additives. The epithelial cells grew equally well in 10% fetal calf serum as in 20% fetal calf serum. However, when the fetal calf serum was excluded from the growth media, the cultured epithelial cells were less viable. In order to control for unwanted microorganisms, gentamicin, penicillin, streptomycin, and Fungizone were included in the growth media. At the stated concentrations, these antibiotics appeared to have no adverse effects on the cultured cells and eliminated contamination by bacteria. Unlike the earlier studies of culturing
for 49 days
inner ear endothelial cells on fibronectin-coated wells (Bowman et al., 1985) the epithelial cells grew as well or better when plated directly onto the polystyrene wells without the use of a substrate, like fibronectin. In addition, different agents (e.g., insulin, transferrin, prostaglandin E,, hydrocortisone, and cholera toxin), reported to enhance epithehal cell growth in vitro (Ha~ond et al., 1984; Lechner, 1984; Stampfer, 1982; Taub et al., 1979) were incorporated into the growth media in order to facilitate culturing of the inner ear epithelial cells. Although no quantitative methods were employed, there was no observable increase of the number or the size of the epithelial cells. Another major problem in culturing the inner ear epithelial cells was suppressing the growth of unwanted cells and purifying the epithelial monolayers. Because the epithelial cells grew at a slower pace than other cell types, fibroblast-like cells would have the tendency to encircle small epithelial islands of cells and overgrow the islands. It was demonstrated that when the monolayers were
285
Fig. 9. Photomicrograph subsequently subcultured
of passaged utricular dark cells. These vestibular cells were maintained in primary culture for 10 days and for 6 days. Like the strial epithelial cells, the subcultured utricular cells developed a bubble-like appearance (arrows). X 1000.
treated with cold (4OC) 0.05% trypsin for 1 min, most fibroblast-like cells were removed without disruption of the epithelial cells. This technique allowed a greater purity of the epithelial cell culture. In comparison to the growth of the stria marginal cells, the utricular dark cells grew more rapidly, i.e., there was more mitotic division in the dark cells, resulting in many more confluent islands. Since the marginal cells of the stria are highly differentiated (i.e., less primitive) in vivo, a possibility exists that they are less metabolically active in vitro. Cellular characteristics of primary epithelial cells of the stria and posterior wall of the utricle were similar to those observed in vivo as observed by light and electron microscopy. The cultured cells appeared cuboidal in shape, and had apical and basolateral specializations as described for other epithelial cells grown in culture (BelloRheuss and Weber, 1987; Cereijido et al., 1980; Simmons, 1982). Microvilli were seen on the api-
cal surfaces of the cells and interdigitating infoldings were observed basolaterally. In this regard, the stria marginal cells appeared to have less basolateral infoldings in culture than in vivo. Junctional complexes were observed. Tight junctions (zonulae occludens) and desmosomes (maculae adherens) were observed between apposing epithelial cells by electron microscopy and freezefracture techniques. Such junctions appeared similar to those observed in vivo (Rarey and Ross, 1982; Reale et al., 1975; Anniko and BaggerSjoback, 1984). As the dissociated inner ear epithelial cells plated onto the culture wells, they became polarized morphometrically. Microvilli were evident on their apical surfaces, and infoldings of their basolateral surfaces occurred. Such polarization would thus indicate that the cultured cells were attempting to become polarized functionally. Other transporting cells in vitro also demonstrated similar polarization (Cereijido et al., 1980; Sim-
286
mons, 1982). If there is indeed a coupling of structural and functional polarization of cultured inner
ear
epithelial
histochemical whether their
cells,
then
assessments
cultured
in vivo
epithelial
enzymatic
studies
demonstrated
ATPase
activity
utricular
epithelial
rometric
microassay
biochemical
could cells
retained
properties. the
in both
any of
Preliminary
presence
of
the cultured
Na-K-
stria
and
cells in vitro as shown by fluo(unpublished
results).
As the cultured epithelial cells tured, they became characteristically like or developed
or
determine
blister-like
were subculmore foamy-
appearances
as ob-
served by light microscopy. Such an appearance of these cells is similar to that reported for other ion-transporting and secretory epithelial cells vivo and in vitro (Birek et al., 1982; Cereijido al., 1980;
Fujimoto
al., 1984). mations
by the cultured
fluid (Birek
et al.,
Sugahara
of these blister-like
has been hypothesized
an attempt port
and Ogawa, 1983;
The formation
in et et for-
to be the result of
epithelial 1982).
This
cell to transobservation
may further strengthen the hypothesis that the inner ear epithelial cells become functionally polarized
in vitro.
The observed cellular characteristics of the cultured inner ear epithelia emphasize that primary culture
offers
response
a convenient
means
of select cells independent
ing tissues
and of the effects
to examine
the
of surround-
of humoral
agents.
Therefore, the specific roles of such cells may become more evident. Culture of inner ear cells represents
a more economical
inner ear cells, e.g., reducing animals. Confluent thelial cells ideally
means to study select the use of laboratory
monolayers of inner ear epican provide a means to ex-
amine in vitro mechanisms
of electrolyte
transepi-
thelial transport and metabolic characteristics of such cells under various experimental conditions.
Acknowledgements The authors wish to acknowledge Dr. Philip Bowman for co-pioneering the earlier study of primary culture of inner ear endothelial cells, which provided the basis to examine the feasibility of culturing select inner ear epithelial cells, and Ms. Linda Mobley for her assistance in editing.
References Anniko, M. and Bagger-Sjoback, D. (1984) The Stria Vascularis. In: I. Friedmann Ballantyre. J. (Eds.), Ultrastructural Atlas of the Inner Ear, Butterworths, London, England, pp. 184-208. Anniko, M. and Sobin, A. (1983) Organ culture of the crista ampullaris of the embryonic guinea pig. Arch. Otolaryngol. 103, 262-264. Anniko, M., Van De Water, T.R. and Nordemar, H. (1979) Embryogenesis of the inner ear. I. Development and differentiation of the mammalian crista ampullaris in viva and in vitro. Arch. Oto-Rhino-Laryngol. 224, 2855299. Bello-Rheuss, E. and Weber, M.R. (1987) Electrophysiological studies of primary cultures of rabbit distal tubule cells. Am. J. Physiol. 252, F899-F909. Birek, C., Aubin, J.E., Bhargava, U., Brunette. D.M. and Melcher, A.H. (1982) Dome formation by oral epithelia in vitro. In Vitro 18, 3822392. Bowman. P.D., Rarey, K.E., Rogers, C. and Goldstein, G.W. (1985) Primary culture of capillary endothelial cells from the spiral ligament and stria vascularis of bovine inner ear. Retention of several endothelial cell properties in vitro. Cell Tissue Res. 241, 479-486. Cereijido, M., Ehrenfeld, J., Meza, 1. and Martinez-Palomo, A. (1980) Structural and functional membrane polarity in cultured monolayers of MDCK cells. J. Membr. Biol. 52, 147-159. FuJimoto. T. and Ogawa, K. (1983) Cell membrane polarity m dissociated frog urinary bladder epithelial cells. J. Histochem. Cytochem. 31, 131-138. Hammond. S.L.. Ham, R.G. and Stampfer. M.R. (1984) Serum-free growth of human mammary epithelial cells: rapid clonal growth in defined medium and extended serial passage with pituitary extract. Cell Biol. 81, 543555439. Lawrence, M. and Merchant. D.J. (1953) Tissue culture techniques for the study of the isolated otic vesicle. Ann. Otol. Rhino]. Laryngol. 62, 770-785. Lechner. J.F. (1984) Interdependent regulation of eptthelial cell replication by nutrients, hormones, growth factors and cell density. Fed. Proc. 43, 116-120. Lim, D.J. and Flock, A. (1985) Dissociated single cells from the inner ear: stria cells. Am. J. Otol. 6, 153-167. Maurizi, M., Binaglia, L.. Donti. E.. Ottaviani, F., Paludetti, G. and Venti Donti, G. (1983) Morphological and functional characteristics of human temporal bone cell cultures. Cell. Tissue. Res. 229. 5055513. Okano. Y. and Iwai, H. (1975) Effect of the high potassium medium on cultured cochlear epithelial cells. Arch. OtoRhino-Laryngol. 209. 121-125. Okano. Y.. Yamashita, T. and Iwai. H. (1975) In vitro morphological study of cochlear epithelium. Arch. Oto-RhinoLaryngol. 290, 151-158. Pauli, B.U.. Weinstein, R.S.. Soble. L.W. and Alroy, J. (1977) Freeze-fracture of monolayer cultures. J. Cell Biol. 72. 763-769. Rarey, K.E. and Ross, M.D. (1982) A survey of the effects of loop diuretics on the xonula occludentes of the perilymph-
287 endolymph barrier by freeze fracture. Acta Otolaryngol. 94, 307-316. Reale, E., Luciano, L., Franke, K., Pannese, E., Wermbter, G. and Iurato, S. (1975) Intercellular junctions in the vascular stria and spiral ligament. J. Ultrastruct. Res. 53, 284-297. Russell, I.J. and Richardson, G.P. (1987) The morphology and physiology of hair cells in organotypic cultures of the mouse cochlea. Hear. Res. 31, 9-24. Stampfer, M.R. (1982) Cholera toxin stimulation of human mammary epithelial cells in culture. In Vitro 18, 531-537. Simmons, N.L. (1982) Cultured monolayers of MDCK cells: a novel model system for the study of epithelial development and function. Gen. Pharm. 13, 287-291.
Sugahara, K., Caldwell, J.H. and Mason, R.J. (1984) Electrical currents flow out of domes formed by cultured epithelial cells. J. Cell Biol. 99, 1541-1544. Taub, M., Chauman, L., Saier, M.H., Jr. and Sato, G. (1979) Growth of Madin-Darby canine kidney epithelial cell MDCK line in hormone-supplemented, serum-free medium. Cell Biol. 76, 3338-3342. Van De Water, T.R. and Ruben, R.J. (1974) Growth of the inner ear in organ culture. Ann. Otol. Rbinol. Laryngol. 83 (suppl 14), 1-16. Yamashita, T. and Vosteen, K.H. (1975) Tissue culture of the organ of Corti and isolated hair cells from the newborn guinea pig. Acta Otolaryngol. (suppl 330) 77-90.
Henring Research, 38 (1989) 277-288
Elsevier
HRR 01187
Establishment of inner ear epithelial cell culture: Isolation, growth and characterization ICE. Rarey and K. Patterson Departments
of Anatomy
and CeN Biology, and Surgery, College of Medicine, University of Florida, Gainesville, Florida, U.S.A.
(Received 13 June 1988; accepted 21 November 1988)
Select epithelial regions of the bovine inner ear were established and maintained in cell culture. Mkginal cells from the stria vascularis and dark cells from the posterior wall of the utricle were isolated, dissociated and placed in culture medium. Within 24 h, cellular islands of hexagonal-shaped, epithelial-like cells from both the stria vascularis and posterior utricular wall were readily identifiable by inverted light microscopy. Ultrastructural examination of both the cultured stria marginal cells and utricular dark cells revealed that both cell types had numerous microvilli on their apical surfaces and interdigitating infoldings of their basolateral surfaces. Apical tight junctional complexes were present between apposing cells. These findings demonstrate that inner ear bovine epithelial cells can be successfully isolated and maintained in culture, and that such cells retain certain of their in viva morphological characteristics Inner ear; Bovine; Stria vascularis; Dark cell; Epithelial; Cell culture
Introduction Understanding the individual roles of certain inner ear cells has been the focus of studies which examine cellular mechanisms associated with inner ear microhomeostasis. To better define the individual roles of certain inner ear tissues and/or cells, various techniques have been employed for isolation and characterization. These techniques have included cultures of the otic vesicle (Lawrence and Merchant, 1953), cochlear and vestibular tissues (Ann&o et al., 1979; Anniko and Sobin, 1983; Okano and Iwai, 1975; Okano et al., 1975; Russell and Richardson, 1987; Van De Water and Ruben, 1974; Yamashita and Vosteen, 1975), temporal bone cells (Maurizi et al., 1983), and stria marginal cells (Lim and Flock, 1985). Although these techniques have provided insight into the in vitro response of the various cells, a disadvantage
has been the limited viability of these tissues or cells. Primary culture of the endothelial cells from stria vascularis and spiral ligament tissues has been reported (Bowman et al., 1985). These cultures provide the opportunity for long term study of the cellular responses of the inner ear endothelial cells as they relate to the structural and functional integrity of the blood-labyrinthine barrier. To determine if isolated nonsensory inner ear epithelial cells, which form the perilymph-endolymph barrier, could be maintained in vitro, select epithelial regions were dissociated and placed in culture. Results indicate that cultures of bovine stria marginal cells and utricular dark cells can be established in vitro and that these cultured epithelial cells retain several of their in vivo morphological characteristics. Materials and Methods
K.E. Rarey, Box J-235, J. Hillis Miller Health Science Center, University of Florida, Gainesville, FL 32610, U.S.A. Correspondence to:
0378-5955/89/$03.50
Bovine temporal bone specimens were harvested with an oscillating bone-plug cutter and placed on
0 1989 Elsevier Science Publishers B.V. (Biomedical Division)
278
ice within 30 min of sacrifice. Each temporal bone was drilled with a dental engine under a dissecting microscope until the inner ear bony labyrinth was identified. Cold phosphate buffered saline (PBS) solution was dripped onto each temporal bone while drilling to remove bone dust and to reduce possible damage to the tissues from the heat discharge of the drill bits. Each turn of the lateral cochlear wall was identified and dissected from adjacent cochlear tissues and placed in Alpha Minimal Essential Medium (MEM) (Gibco) with ribonucleosides, deoxyribonucleosides and the following antibiotics: gentamicin (40 pg/ml), penicillin (50 units/ml), streptomycin (50 pg/ml), and Fungizone (2.5 pg/ml) (Gibco). Stria vascularis tissue from all turns of the cochlea was gently dissected from the spiral ligament with fine forceps and placed in fresh medium. In addition, the bony vestibule was opened and the posterior wall of the utricle was removed. Samples of stria vascularis and posterior utricular wall tissues from four bovine temporal bones were pooled respectively. Five to 10 mg (wet weight) of stria vascularis and 1-2 mg (wet weight) of posterior utricular wall tissue were routinely obtained. Epithelial cells from the two tissue samples were dissociated either by treatment with 1% Collagenase/Dispase (Boehringer, Mannheim) or 0.125% trypsin (Gibco) in MEM Alpha for 1.5-3 h at 37” C. Following dissociation, the samples were centrifuged at 300 X g, rinsed and plated onto 12-well polystyrene culture plates (Costar). The medium used to rinse and culture was Alpha MEM with ribonucleosides and deoxyribonucleosides, supplemented with 10% fetal calf serum, gentamicin (5 pg/ml), penicillin (25 units/ml), streptomycin (25 pg/ml), Fungizone (5 pg/ml), and 30 mM Hepes buffer (Gibco) at pH 7.4. The cultured epithelial cells were grown at 37 o C, with maximal humidity in 5% CO,/95% air. Cultures were examined daily by inverted microscopy. The culture medium was changed weekly. In order to subculture the two epithelial cell types, monolayers of each cell type were rinsed with a 0.05% trypsin solution containing 0.02% EDTA, 6 mM NaOH, 122 mM NaCl, 5.1 mM KCl, 0.8 mM Na,HPO,, 0.2 mM KH,PO,, 0.1 mM anhydrous dextrose, 0.03% NaHCO, and 30 mM Hepes buffer. After 30 s, the solution was
discarded and approximately two additional milliliters were added to the well. After 4 min unattached cells were removed with a sterile pipette, gently aspirated, and distributed into new wells. Fresh culture medium was then added to each well. Transmission electron microscopy Cultures were rinsed in three changes of PBS and fixed for 2 h in a 2.5% glutaraldehyde phosphate buffered (pH 7.4) solution. The epithelial monolayers were then rinsed in buffer and postfixed in a 1% osmium tetroxide phosphate buffered (pH 7.4) solution. After dehydration in graded concentrations of ethanol, the monolayers were embedded in Epon/Araldite. The embedded monolayers were cut from the culture wells with a dremmel saw and ultra-thin sections of the cultured stria and dark cells were made. Sections were stained with uranyl acetate and lead citrate and examined in either a Philips 400 electron microscope or a JEOL 100s electron microscope operated at 60 kV. Freeze-fracture To determine whether the cultured epithelial cells developed tight junctions, some epithelial monolayers were grown on polystyrene coverslips and fixed with a 2.5% glutaraldehyde phosphate buffered (pH 7.4) solution for 1 h, and then placed in 20% glycerol phosphate buffered saline as described by Pauli et al. (1977). Areas for freeze-fracture were selected under an inverted microscope and punched from the plastic coverslip with a 3 mm punch. The 3 mm disks were placed on specimen carriers and rapidly frozen in liquid nitrogen. The monolayers were fractured with a double replicating device in a Balzers high vacuum freeze-fracture unit at - 115O C. Platinum-shadowed carbon replicas of the fractured faces were placed in Chlorox bleach for 24 h. The replicas were rinsed in double distilled water, placed on grids and examined in a Philips 400 or JEOL 100s electron microscope operated at 60 kV. Results Digestion of stria vascularis and posterior utricular wall tissues in 1% collagenase-dispase for 3 h
279
Fig. 1. Photomicrograph Primary
of an island of stria marginal culture, 5 days. X 2140.
cells.
at 37O C under agitation normally produced 3-5 individual islands of epithelial-like cells within 72 h (Figs. 1 and 2). Generally these epithelial islands were quickly encircled and overgrown by fibroblast-like cells. On occasion, when the growth of the fibroblast-like cells sufficiently hindered the expansion of the epithelial colonies, the monolayers were treated with cold (4O C) 0.05% trypsin for 1 min. Because the epithelial cells adhered more strongly to the culture wells than the fibroblast-like cells, most fibroblasts could be removed
Fig. 3. Electron
micrograph
of stria marginal
Fig. 2. Photomicrograph Primary
of an island of utricular culture, 3 days. x 1875.
dark
cells.
without disrupting the epithelial cells. Following three brief rinses (15 s each) with sterile water, the epithelial monolayers were again incubated in growth medium. A modified digestion of inner ear tissues, using 0.125% trypsin for 1.5 h instead of 3 h, produced from 5-20 colonies, consisting of 5-10 epithelial cells in less than 18 h. By using this modified digestion, the epithelial cells initiated mitotic activity much sooner than did cells digested by other methods. As a result epithelial islands grew larger
cells after 7 days in primary culture. Several microvilli (A) surfaces of the epithelial cells. X 12 880.
can be identified
on the apical
280
Fig. J. Electron micrograph of an apical tight junctional complex (arrow) between two adjacent stria marginal cells. Observe the ““m erous lateral infoldings of the plasma membranes directly beneath the junctional complex. A = apical surface. Primary cult ure, 21 days. x 60 350.
and at a more rapid rate before being encircled by fibroblasts. In addition, it was easier to remove unwanted fibroblast-like cells from the culture wells with this modified digestion protocol. Insulin (5 pg/ml), transferrin (5 pg/ml), prostaglandin E, (25 ng/ml), hydrocortisone (50 nM) and cholera toxin (1 ng/cc) were added singly and in combination to the growth media to enhance the growth of the isolated cells. The addition of such agents did not increase the cell growth of the inner ear cells. The inner ear epithelial cells were readily identifiable in primary culture by their typical hexagonal cellular appearance under inverted phase-contrast microscopy. Utricular epithelial cells grew
more quickly than the marginal cells of the stria vascularis. Nonepithelial cells appeared fibroblastlike with no indication of cell-to-cell confluency. Electron microscopy of the stria and utricular epithelial cells revealed characteristic properties which are present in vivo. Microvilli were seen on the apical surfaces of the cultured epithelial cells (Fig. 3). Interdigitating infoldings were observed on the basolateral surfaces of both the stria and utricular cells (Figs. 3, 4 and 5). In addition, zonula occludentes were observed between apposing epithelial cells (Figs. 4 and 5). Below the tight junctional complexes, desmosomes were seen (Fig. 5). When both cell types were processed for freeze-fracture, strands of particles were observed
281
Fig. 5. Electron micrograph of another representative junctional complex (opened arrow) between two stria marginal Desmosomes (solid arrow) can be observed beneath the complex. A = apical surface. Primary culture, 21 days. X 92 300.
in the region of the zonula occludentes just beneath the apical surfaces of the cultured utricular and stria epithelial cells (Fig. 6). Subculture of primary epithelial monolayers was routinely performed with cold (4°C) 0.05% trypsin (Figs. 7a-e). Stria and utricular epithelial cells began readhering within 30 min after transfer (Fig. 7b). By 24 h (Fig. 7e) several hundred islands
of up to 100 cells each were typically observed. These subcultured cells appeared to retain their primary morphological characteristics through several passages. A foamy or bubble-l&e appearance (Figs. 8 and 9) was typical of the passaged stria and utricular cells. As in primary culture, the utricular epithelial cells grew at a faster rate than those of the stria vascularis, reaching
Fig. 6. F:reeze-fracture
replica of the organization of a tight junctional complex (zonula occludens) between two vestibular Parallel strands of particles on the P face (PF) can be seen. Primary culture. 21 days. x 46750.
confluence (380 mm2) in about 10 days. Subcultured stria epithelial cells grew to confluence in approximately 15 days. Discussion This study reports the isolation, growth and characterization of cultured bovine epithelial cells from the stria vascularis and posterior wall of the utricle. Both epithelial cell types were established in primary culture for 8-10 weeks. Because of the viability of such cells, they were passaged and observed for an additional 22-26 weeks. Contrary to the apparent reported ease of culturing epithelial cells from other tissues, there were several inherent obstacles in establishing cul-
dark
tures of inner ear epithelial cells. These obstacles were the small amount of tissues that could be harvested and prepared for cell culture, and the species from which the inner ear tissues were obtained. Due to the small quantity of stria and posterior utricular wall tissues, specimens had to be pooled from more than one animal. In addition, it was found that bovine inner ear epithelial cells responded better in culture than those from other species, e.g.. guinea pig (unpublished results). Initially a 1% collagenase-dispase solution was used to digest the inner ear tissues for 3 h. However, cells of the stria vascularis and posterior utricular wall did not rapidly dissociate. Therefore, a 0.125% trypsin solution was used to digest
283
Fig. 7. Time-lapse photomicrographs min al ‘ter 0.05% trypsin treatment.
of utricular epithelial cells being subcultured. (a) 15 nun after 0.05% trypsin treatment. (b) 20 (c) 30 min after 0.05% trypsin treatment. (d) 30 min after 0.05% trypsin treatment. (e) 24 h after 0.05% trypsin treatment. (a-d) X 560; (e) X 520.
284
Fig. 8. Photomicrograph of subcultured stria epithelial cells. These epithelial ceils had been maintained in primary culture and subcultured for 14 days. Typically, the cells developed a foamy-like appearance (arrows). X 880.
the tissues. This solution appeared to increase the dissociation of the stria and utricular cells in a shorter period of time (1.5 h). Because the cells were more readily dissociated, they could be placed in growth media more quickly. This step was thought to be crucial in that the cellular viability was considered better with a shorter dissociation period. The epithelial cells from both the bovine stria and utricular tissues grew better in Alpha MEM containing ribonucleosides and deoxyribonucleosides than in Alpha MEM without these additives. The epithelial cells grew equally well in 10% fetal calf serum as in 20% fetal calf serum. However, when the fetal calf serum was excluded from the growth media, the cultured epithelial cells were less viable. In order to control for unwanted microorganisms, gentamicin, penicillin, streptomycin, and Fungizone were included in the growth media. At the stated concentrations, these antibiotics appeared to have no adverse effects on the cultured cells and eliminated contamination by bacteria. Unlike the earlier studies of culturing
for 49 days
inner ear endothelial cells on fibronectin-coated wells (Bowman et al., 1985) the epithelial cells grew as well or better when plated directly onto the polystyrene wells without the use of a substrate, like fibronectin. In addition, different agents (e.g., insulin, transferrin, prostaglandin E,, hydrocortisone, and cholera toxin), reported to enhance epithehal cell growth in vitro (Ha~ond et al., 1984; Lechner, 1984; Stampfer, 1982; Taub et al., 1979) were incorporated into the growth media in order to facilitate culturing of the inner ear epithelial cells. Although no quantitative methods were employed, there was no observable increase of the number or the size of the epithelial cells. Another major problem in culturing the inner ear epithelial cells was suppressing the growth of unwanted cells and purifying the epithelial monolayers. Because the epithelial cells grew at a slower pace than other cell types, fibroblast-like cells would have the tendency to encircle small epithelial islands of cells and overgrow the islands. It was demonstrated that when the monolayers were
285
Fig. 9. Photomicrograph subsequently subcultured
of passaged utricular dark cells. These vestibular cells were maintained in primary culture for 10 days and for 6 days. Like the strial epithelial cells, the subcultured utricular cells developed a bubble-like appearance (arrows). X 1000.
treated with cold (4OC) 0.05% trypsin for 1 min, most fibroblast-like cells were removed without disruption of the epithelial cells. This technique allowed a greater purity of the epithelial cell culture. In comparison to the growth of the stria marginal cells, the utricular dark cells grew more rapidly, i.e., there was more mitotic division in the dark cells, resulting in many more confluent islands. Since the marginal cells of the stria are highly differentiated (i.e., less primitive) in vivo, a possibility exists that they are less metabolically active in vitro. Cellular characteristics of primary epithelial cells of the stria and posterior wall of the utricle were similar to those observed in vivo as observed by light and electron microscopy. The cultured cells appeared cuboidal in shape, and had apical and basolateral specializations as described for other epithelial cells grown in culture (BelloRheuss and Weber, 1987; Cereijido et al., 1980; Simmons, 1982). Microvilli were seen on the api-
cal surfaces of the cells and interdigitating infoldings were observed basolaterally. In this regard, the stria marginal cells appeared to have less basolateral infoldings in culture than in vivo. Junctional complexes were observed. Tight junctions (zonulae occludens) and desmosomes (maculae adherens) were observed between apposing epithelial cells by electron microscopy and freezefracture techniques. Such junctions appeared similar to those observed in vivo (Rarey and Ross, 1982; Reale et al., 1975; Anniko and BaggerSjoback, 1984). As the dissociated inner ear epithelial cells plated onto the culture wells, they became polarized morphometrically. Microvilli were evident on their apical surfaces, and infoldings of their basolateral surfaces occurred. Such polarization would thus indicate that the cultured cells were attempting to become polarized functionally. Other transporting cells in vitro also demonstrated similar polarization (Cereijido et al., 1980; Sim-
286
mons, 1982). If there is indeed a coupling of structural and functional polarization of cultured inner
ear
epithelial
histochemical whether their
cells,
then
assessments
cultured
in vivo
epithelial
enzymatic
studies
demonstrated
ATPase
activity
utricular
epithelial
rometric
microassay
biochemical
could cells
retained
properties. the
in both
any of
Preliminary
presence
of
the cultured
Na-K-
stria
and
cells in vitro as shown by fluo(unpublished
results).
As the cultured epithelial cells tured, they became characteristically like or developed
or
determine
blister-like
were subculmore foamy-
appearances
as ob-
served by light microscopy. Such an appearance of these cells is similar to that reported for other ion-transporting and secretory epithelial cells vivo and in vitro (Birek et al., 1982; Cereijido al., 1980;
Fujimoto
al., 1984). mations
by the cultured
fluid (Birek
et al.,
Sugahara
of these blister-like
has been hypothesized
an attempt port
and Ogawa, 1983;
The formation
in et et for-
to be the result of
epithelial 1982).
This
cell to transobservation
may further strengthen the hypothesis that the inner ear epithelial cells become functionally polarized
in vitro.
The observed cellular characteristics of the cultured inner ear epithelia emphasize that primary culture
offers
response
a convenient
means
of select cells independent
ing tissues
and of the effects
to examine
the
of surround-
of humoral
agents.
Therefore, the specific roles of such cells may become more evident. Culture of inner ear cells represents
a more economical
inner ear cells, e.g., reducing animals. Confluent thelial cells ideally
means to study select the use of laboratory
monolayers of inner ear epican provide a means to ex-
amine in vitro mechanisms
of electrolyte
transepi-
thelial transport and metabolic characteristics of such cells under various experimental conditions.
Acknowledgements The authors wish to acknowledge Dr. Philip Bowman for co-pioneering the earlier study of primary culture of inner ear endothelial cells, which provided the basis to examine the feasibility of culturing select inner ear epithelial cells, and Ms. Linda Mobley for her assistance in editing.
References Anniko, M. and Bagger-Sjoback, D. (1984) The Stria Vascularis. In: I. Friedmann Ballantyre. J. (Eds.), Ultrastructural Atlas of the Inner Ear, Butterworths, London, England, pp. 184-208. Anniko, M. and Sobin, A. (1983) Organ culture of the crista ampullaris of the embryonic guinea pig. Arch. Otolaryngol. 103, 262-264. Anniko, M., Van De Water, T.R. and Nordemar, H. (1979) Embryogenesis of the inner ear. I. Development and differentiation of the mammalian crista ampullaris in viva and in vitro. Arch. Oto-Rhino-Laryngol. 224, 2855299. Bello-Rheuss, E. and Weber, M.R. (1987) Electrophysiological studies of primary cultures of rabbit distal tubule cells. Am. J. Physiol. 252, F899-F909. Birek, C., Aubin, J.E., Bhargava, U., Brunette. D.M. and Melcher, A.H. (1982) Dome formation by oral epithelia in vitro. In Vitro 18, 3822392. Bowman. P.D., Rarey, K.E., Rogers, C. and Goldstein, G.W. (1985) Primary culture of capillary endothelial cells from the spiral ligament and stria vascularis of bovine inner ear. Retention of several endothelial cell properties in vitro. Cell Tissue Res. 241, 479-486. Cereijido, M., Ehrenfeld, J., Meza, 1. and Martinez-Palomo, A. (1980) Structural and functional membrane polarity in cultured monolayers of MDCK cells. J. Membr. Biol. 52, 147-159. FuJimoto. T. and Ogawa, K. (1983) Cell membrane polarity m dissociated frog urinary bladder epithelial cells. J. Histochem. Cytochem. 31, 131-138. Hammond. S.L.. Ham, R.G. and Stampfer. M.R. (1984) Serum-free growth of human mammary epithelial cells: rapid clonal growth in defined medium and extended serial passage with pituitary extract. Cell Biol. 81, 543555439. Lawrence, M. and Merchant. D.J. (1953) Tissue culture techniques for the study of the isolated otic vesicle. Ann. Otol. Rhino]. Laryngol. 62, 770-785. Lechner. J.F. (1984) Interdependent regulation of eptthelial cell replication by nutrients, hormones, growth factors and cell density. Fed. Proc. 43, 116-120. Lim, D.J. and Flock, A. (1985) Dissociated single cells from the inner ear: stria cells. Am. J. Otol. 6, 153-167. Maurizi, M., Binaglia, L.. Donti. E.. Ottaviani, F., Paludetti, G. and Venti Donti, G. (1983) Morphological and functional characteristics of human temporal bone cell cultures. Cell. Tissue. Res. 229. 5055513. Okano. Y. and Iwai, H. (1975) Effect of the high potassium medium on cultured cochlear epithelial cells. Arch. OtoRhino-Laryngol. 209. 121-125. Okano. Y.. Yamashita, T. and Iwai. H. (1975) In vitro morphological study of cochlear epithelium. Arch. Oto-RhinoLaryngol. 290, 151-158. Pauli, B.U.. Weinstein, R.S.. Soble. L.W. and Alroy, J. (1977) Freeze-fracture of monolayer cultures. J. Cell Biol. 72. 763-769. Rarey, K.E. and Ross, M.D. (1982) A survey of the effects of loop diuretics on the xonula occludentes of the perilymph-
287 endolymph barrier by freeze fracture. Acta Otolaryngol. 94, 307-316. Reale, E., Luciano, L., Franke, K., Pannese, E., Wermbter, G. and Iurato, S. (1975) Intercellular junctions in the vascular stria and spiral ligament. J. Ultrastruct. Res. 53, 284-297. Russell, I.J. and Richardson, G.P. (1987) The morphology and physiology of hair cells in organotypic cultures of the mouse cochlea. Hear. Res. 31, 9-24. Stampfer, M.R. (1982) Cholera toxin stimulation of human mammary epithelial cells in culture. In Vitro 18, 531-537. Simmons, N.L. (1982) Cultured monolayers of MDCK cells: a novel model system for the study of epithelial development and function. Gen. Pharm. 13, 287-291.
Sugahara, K., Caldwell, J.H. and Mason, R.J. (1984) Electrical currents flow out of domes formed by cultured epithelial cells. J. Cell Biol. 99, 1541-1544. Taub, M., Chauman, L., Saier, M.H., Jr. and Sato, G. (1979) Growth of Madin-Darby canine kidney epithelial cell MDCK line in hormone-supplemented, serum-free medium. Cell Biol. 76, 3338-3342. Van De Water, T.R. and Ruben, R.J. (1974) Growth of the inner ear in organ culture. Ann. Otol. Rbinol. Laryngol. 83 (suppl 14), 1-16. Yamashita, T. and Vosteen, K.H. (1975) Tissue culture of the organ of Corti and isolated hair cells from the newborn guinea pig. Acta Otolaryngol. (suppl 330) 77-90.