Postnatal maturation of the dog stria vascularis— an immunohistochemical study
Article first published online: 19 DEC 2002
Copyright © 2003 Wiley-Liss, Inc.
The Anatomical Record Part A: Discoveries in Molecular, Cellular, and Evolutionary Biology
Volume 270A, Issue 1, pages 82–92, January 2003
How to Cite
Coppens, A. G., Salmon, I., Heizmann, C. W., Kiss, R. and Poncelet, L. (2003), Postnatal maturation of the dog stria vascularis— an immunohistochemical study. Anat. Rec., 270A: 82–92. doi: 10.1002/ar.a.10009
- Issue published online: 19 DEC 2002
- Article first published online: 19 DEC 2002
- Manuscript Accepted: 31 AUG 2002
- Manuscript Received: 26 MAR 2002
- Fond National de Recherche Scientifique
- stria vascularis;
- postnatal maturation;
The lateral wall of the dog cochlear duct was investigated by classical staining and immunohistochemistry for NaK/ATPase β2 isoform, cytokeratins (Cks), vimentin, nestin, and S100A6 during the postnatal cochlear maturation, i.e., from birth to postnatal day 110. The dog stria vascularis was immature at birth. Fine melanin granules were evident in the stria from the second week of life, and melanin concentration increased drastically beyond the first month. The marginal cells were NaK/ATPase- and Ck-positive; intermediate cells were either nestin- and S100A6-positive or vimentin-positive; the basal cells were vimentin-positive; the capillary endothelium showed vimentin and nestin labeling; the cell layer underlying the stria was nestin-positive. The fibrocytes of the spiral ligament and spiral prominence expressed nestin and vimentin. The epithelial cells overlaying the spiral prominence and the external sulcus were Ck-positive, and transiently nestin- and vimentin-positive. Double immunolabeling, for S100A6 and either nestin, vimentin, or NaK/ATPase, and for nestin and vimentin suggested the presence of two distinct intermediate cell types. The results enabled us to differentiate the cell types forming the lateral wall of the dog cochlear duct, and to follow their postnatal maturation. This study may form a basis for future investigations about spontaneous cochleosaccular degeneration in dogs. This species is an important companion animal, and a possible model for the study of comparable diseases in humans. Anat Rec Part A 270A:82–92, 2003. © 2003 Wiley-Liss, Inc.
The lateral wall includes the stria vascularis covering the spiral ligament from the anchoring of the Reissner's membrane to the spiral prominence. The stria vascularis is composed of three different cell layers: marginal, basal, and intermediate. The marginal cells are epithelial in origin and form a continuous layer at the endolymphatic surface (Kuijpers et al., 1991; Schucknecht, 1993). The basal cells derive from mesoderm and form continuous layers abutting the underlying spiral ligament (Carlisle et al., 1990; Schucknecht, 1993). The intermediate cells are scattered between the former two layers. The intermediate cells include mostly melanocyte-like cells. Whether other types of intermediary cells exist is still a subject of debate (Cable and Steel, 1991; Conlee et al., 1994a, b; Motohashi et al., 1994; Souter and Forge, 1998).
The spiral ligament and spiral prominence are connective tissues that consist mainly of fibrocytes. Four different types of fibrocytes have been described based on their location, ultrastructure, and immunohistochemical status (Kikuchi et al., 1995). The spiral prominence and external sulcus are lined by epithelial cells; the root cells are located under the cells bordering the external sulcus, extending their basal processes deeply into the spiral ligament (Kikuchi et al., 1995).
The cochlear duct is filled with endolymph, a fluid with a high potassium (K+) concentration and a positive endocochlear potential (Ryan and Woolf, 1992; Takeuchi et al., 2000). The stria vascularis and other cells of the lateral cochlear wall are essential for the production and maintainenance of these electrical and ionic gradients. Recent studies of K+ channel knock-out mice have increased our understanding of the role of the different cell types (Marcus et al., 2002). Basal cells form a barrier between the intrastrial space and the perilymph. Marginal cells secrete K+ into the endolymph and keep the intrastrial space K+ concentration low. This low intrastrial K+ concentration and the high intermediate cell cytosolic K+ concentration generate the intrastrial potential through the KCNJ10 K+ channel. The intrastrial potential actually equals the endolymphatic potential, since the marginal cell layer has nearly no transepithelial voltage of its own (Marcus et al., 2002). The external sulcus cells and the fibrocytes of the spiral ligament and spiral prominence contribute in circulating K+ toward the stria vascularis (Kikuchi et al., 1995).
In cochleosaccular degeneration, the primary alteration lies in the stria vascularis. In this defect, the cochlear duct is collapsed, the stria vascularis and organ of Corti are degenerated, the tectorial membrane is abnormal, and the sacculus is usually collapsed. This type of deafness is reported in several mammalian species, including dogs and humans (Steel and Bock, 1983; Steel and Harvey, 1992; Schucknecht, 1993; Strain, 1996; Hardisty et al., 1998). In humans, the best documented cochleosaccular entities are the Waardenburg's syndromes (Cable et al., 1994; Hardisty et al., 1998), which are associated with pigmentation defects, as in dogs. In Jervell and Lange-Nielsen syndrome (in humans), cochleosaccular degeneration is associated with cardiac anomalies (Hardisty et al., 1998). Cochleosaccular degeneration may also be observed in nonsyndromic hearing impairment (Lalwany et al., 1997; Steel, 1999). The dog cochlea is immature at birth, and histopathological reports of spontaneous congenital cochleosaccular degeneration in this species suggest that the lesions appear during the postnatal maturation of the cochlea and stria vascularis (Johnsson et al., 1973; Pujol and Hidling, 1973; Steel and Bock, 1983; Strain, 1996), providing a model for the observation of the evolving degeneration. A description of normal postnatal maturation of the dog stria vascularis can thus aid our understanding of the mechanism and timing of cochleosaccular degeneration.
In the current study we aimed to examine the normal postnatal maturation of the dog stria vascularis. In several mammalalian species, different cell types forming the stria vascularis have been distinguished by the use of specific antibodies (Schulte and Adams, 1989; Kuijpers et al., 1991, 1992; Agrup et al., 1996). In the present work, antibodies against NaK/ATPase β2 isoform was used because this enzyme has been found mainly in marginal strial cells of several mammalian species (Ichimiya et al., 1994; ten Cate et al., 1994; Nakazawa et al., 1995; Erichsen et al., 1996). Intermediate filament proteins cytokeratins 5/8, vimentin, and nestin antibodies were used because the expression of these proteins may be related to the cell embryological origin (Regauer et al., 1985; Kuijpers et al., 1992; Tohyama et al., 1992), and thus may allow specific cell staining.
In an attempt to characterize the intermediate cell population, double immunolabelings for anti-nestin, vimentin, and NaK/ATPase with anti-S100A6 (previously described in dog intermediate cells (Coppens et al., 2001)) and for anti-nestin with anti-vimentin were also undertaken on dog cochleae.
MATERIAL AND METHODS
The inner ears of 19 beagle puppies (males and females) from the breeding colony of our institution were collected for this study. The dogs were cared for according to the principles of the NIH Guide for the Care and Use of Laboratory Animals. The experimental protocol was approved by the University Ethics and Animal Care Committee (reference LA 1230324/4). The investigations were performed in accordance with the principles of the Declaration of Helsinki. The puppies, aged from birth to postnatal day (PD) 95, were randomly selected from seven different litters. The ages for histological examination were: birth, 1 PD (two puppies), 3 PD, 5 PD (two puppies), 8 PD, 10 PD, 12 PD (two puppies), 14 PD, 18 PD (two puppies), 24 PD, 25 PD, 27 PD, 33 PD, 91 PD, and 95 PD. An adult cochlea from our personal collection (110 PD) was added. Other puppies from the litters were assessed for hearing by electrophysiological tests. No hearing defects were detected in the colony.
The puppies were deeply anesthetized with 4 mg/kg intramuscular xylazine (Rompun, Bayer Belgium), followed by intraperitoneal or intravenous pentobarbital, 30 mg/kg (Nembutal; Sanofi Belgium). They received 7,500 IU/ kg intrasplenic or intravenous heparin (Léo, Belgium). They were transcardially perfused with 0.1 molar (M) phosphate-buffered saline (PBS) followed by 4% buffered paraformaldehyde in 0.1 M PBS. The end of the thoracic aorta of puppies >1 month old was clamped before perfusion. The temporal bones were immediately dissected, the bullae were opened, and the samples were immersed in the same fixative at 4°C for 24–48 hr. The temporal bones were decalcified for 17–120 days in 0.12 M ethylenediaminetetraacetate (EDTA) (Tritriplex III, Merck-Belgolabo, Leuven, Belgium) in 0.01 M PBS, pH 7.3. The left temporal bones of the 95 PD and 110 PD dogs were decalcified in 5% formic acid for 25 and 40 days, respectively. The decalcified samples were embedded in paraffin and serially cut (6–9 μm thick) parallel to the modiolus. For each puppy, five to 10 sections through the modiolus were dewaxed, hydrated, and stained with either hematoxylin-eosin, cresyl violet, or Fontana staining for classical histology.
Dewaxed sections were treated for 30 min with methanol containing 1% peroxide solution to inhibit endogeneous peroxidase activity, and then washed in 0.05 M Tris-buffered saline (TBS) pH 7.4. The sections were preincubated in a humid chamber for 30 min at 37°C with 1/20 normal horse serum (NHS) in TBS. The sections were covered for 24 hr at 4°C with antibodies diluted in TBS containing 2% bovine serum albumin (BSA) (Sigma; Bornem, Belgium). Polyclonal rabbit NaK/ATPase (β2 subunit) antiserum (U.B.I.; Lake Placid, NY), 1/4000; monoclonal mouse cytokeratins (RCK102) antiserum (Monosan, by Sanvetech; Santa Cruz, CA) against Cks 5 and 8, 1/10; monoclonal mouse vimentin (V6630) antiserum (Sigma; Belgium), 1/1200 and polyclonal rabbit nestin antiserum, 1/3,000, were used. The specificity and characteristics of the nestin antiserum have been described elsewhere (Tohyama et al., 1992). Dewaxed sections were pretreated for 3 min at room temperature with proteinase K (Ready-to-use; Dako; Glostrup, The Netherlands) before they were labeled with cytokeratins.
The primary immune antiserum was revealed by the avidin-biotin procedure as follows: 30 min incubation in biotinylated anti-rabbit, anti-mouse, or anti-goat gamma globulin (Vector Labs,) diluted to 1/100 in TBS with BSA 2%, followed by 30 min incubation with avidin-biotin-peroxidase complex (ABC) standard kit reagent (Vector Labs; Burlingame, CA). The intermediate washings were done with TBS. The staining was visualized by diaminobenzidine (DAB)/H2O2 solution (liquid DAB substrate pack from Biogenex; San Ramon, CA). For every antibody, 10–20 sections (for each developmental age) were labeled. Immunolabeled sections were not counterstained. Control sections, with the primary antibody omitted, remained unlabeled.
NaKATPase-S100A6, nestin-S100A6, S100A6-vimentin, and nestin-vimentin associations were tested. The procedure for the S100A6 antiserum was described previously (Coppens et al., 2001). Combinations with S100A6 antibody were undertaken on sections from puppies aged 24 PD, 33 PD, and 91 PD. For double immunolabeling, S100A6 was diluted at 1/2500 and NaK/ATPase at 1/2400. The dilutions of antibodies, reagents, washing solutions, and procedures were the same as for simple labeling unless otherwise specified.
Deparaffinized and rehydrated sections were treated for blocking endogenous peroxidase, and were then incubated with NHS. Sections were incubated with the first primary antibodies for 6 hr. Sections were rinsed in TBS, and incubated with biotinylated secondary antibody. Immunostaining was revealed by the ABC procedure and visualized by DAB/H2O2 solution (liquid DAB substrate pack from Biogenex,) (brown staining). Sections were rinsed in running tap water for 10 min and washed in TBS for 5 min, in TBS with Triton 2% for 10 min, and in TBS for 5 min (two washes). Sections were incubated with NHS followed by the second primary antibodies for 18 hr at 4°. Sections were rinsed in TBS for 5 min (three washes) and incubated with biotinylated secondary antibody. Immunostaining was revealed by the ABC procedure and visualized by 0.5 mg/ml DAB (Sigma, St. Louis, MO) dissolved in 0.1 M Tris nickel buffer (2 g/L nickel ammonium sulfate) with 0.01% H2O2 (dark blue staining).
Each possible sequence in the order of primary antibodies was tested and compared to determine the optimal combination. Nonspecific immunolabeling was excluded by exchanging the secondary antibody (anti-goat, or anti-rabbit, or anti-mouse) in single labeling tests, and by replacing the first secondary biotinylated antibody by NHS in all the possible combinations of double labeling.
Classical Histology and Fontana Staining
In newborn puppies (0 PD), the multilayered stria vascularis was recognizable from the base to the apex of the cochlea (Fig. 1A). The limit between the stria and underlying connective tissue could be recognized by the presence of elongated nuclei (just beneath the stria) in the lighter-stained connective tissue (Fig. 1B). Nuclei ordered in one row close to the endocochlear surface were those of the marginal cells (Fig. 1B). Less ordered rows of rounded nuclei scattered underneath the marginal cells layer represented the basal and intermediate cell layers. Intermediate or immature basal cells were not distinguishable at this stage (Fig. 1B). The spiral prominence and external sulcus were covered with cuboidal cells (Fig. 1A). Capillaries in transversal sections were observed in the mid-portion of the stria thickness, and a large capillary lumen was systematically observed in the spiral prominence (Fig. 1A and B). Fontana staining failed to reveal melanin in the structure at birth (Fig. 1C). Changes in the different strial cell populations were difficult to observe with classical staining. During the first weeks of life, the marginal cells increased in size and the basal cells became progressively flattened. Fontana staining revealed few, scarcely distinguishable granules of melanin that appeared in some intermediate cells during the third week (Fig. 1D). Melanin granules were not observed in every section, or in every puppy.
The morphological aspect of the stria vascularis did not change from the third week, with classical staining. The main criterion used to classify the different cell types in the stria was their position in the multilayered structure. The monocellular layer of marginal cells lined the endolymphatic space, with the nucleus located in the apical part of the cell, close to the luminal surface (Fig. 1E). In the basal area of the stria, the nuclei of the basal cells were flat and elongated. The basal cells were organized lying on the spiral ligament. In the middle area of the stria, presumptive intermediate cells were randomly scattered; some cells exhibited a stellate appearance. The limits of the different cell cytoplasms remained difficult to delineate. Numerous transversal sections of capillaries were observed. Capillaries in longitudinal sections ran alongside and just beneath the stria in the underlying connective tissue. Fontana staining revealed small melanin pigment granules, irregularly distributed in the stria from the basal to the apical turns.
The concentration of melanin continued to increase drastically beyond the first month. Fontana staining in 3-month-old puppies showed large amounts of melanin, as either finely grained substance or large dense aggregates (Fig. 1F). Because of the poorly delineated cytoplasm limits, it was difficult to associate melanin with specific cell types: Finely grained pigments sometimes appeared to be localized in the intermediate cells, but melanin was distributed in the entire thickness of the stria. In some instances, melanin appeared to be closely associated with the capillaries (Fig. 1F). Melanin distribution was identical among the various cochlear turns.
Classical staining did not reveal any modification in the overall pattern of the spiral ligament during postnatal maturation. Fontana staining failed to reveal melanin pigment in the spiral ligament from birth to adult stages.
At birth, a weak NaK/ATPase β2 isoform labeling revealed the lateral boundaries of the marginal cells of the stria vascularis in all cochlear turns (Fig. 2A).
During the first 2 weeks of life, the intensity of labeling progressively increased, and the lateral and basal boundaries of the marginal cells were stained. During the first 3 weeks, an increasing number of immunostained processes penetrated progressively deeper in the stria toward the more basal cell layers. The observation of labeled cytoplasmic continuity suggests that NaK/ATPase expression is associated with marginal cell processes.
At the end of the first month, the labeling closely surrounded the immunonegative capillaries and immunonegative cells scattered in the middle area of the stria (Fig. 2B). The luminal surface of the marginal cells remained immunonegative or lightly marked. Throughout maturation the basal cells layers remained immunonegative.
The external sulcus cells, the spiral ligament, and the spiral prominence remained immunonegative during the postnatal period (Fig. 2C).
From birth onward, Cks were expressed in the marginal cells of the stria vascularis, and in the cells covering the spiral prominence and external sulcus from the base to the apex of the cochlea (Fig. 3A). Intense labeling was found in the apical part of these cells, forming a continuous, thin immunomarked network extending along the cochlear duct lumen. In the external sulcus cells, the apical and basal parts were labeled (Fig. 3B). The Ck labeling pattern remained unchanged during postnatal maturation (Fig. 3A and B).
At birth a strong vimentin immunoreactivity was observed in the basal mid-portion of the stria thickness, including the immature basal cells, some cells in the middle area of the stria, and capillaries (Fig. 4A). The marginal cells, forming the luminal side of the stria, were unstained. The fibrocytes of the spiral ligament and spiral prominence, and the cells covering the spiral prominence and external sulcus expressed vimentin (Fig. 4B). Underneath the external sulcus cells, the long and slender root cell processes were labeled and penetrated deeply into the spiral ligament (Fig. 4B).
During the first 3 postnatal weeks, vimentin immunostaining receded toward the basal area of the stria (Fig. 4C). Labeled processes extending between capillaries and between the cells in the middle area of the stria could be observed from the second week onward (Fig. 4C). A heterogeneous cytoplasmic immunolabeling surrounded some nuclei of the intermediate layer. The cells overlying the spiral prominence and the external sulcus were positive at birth, but they became negative during the third week.
At the mature stage the flattened vimentin-labeled basal cells were recognizable and localized in the more basal part of the stria. The endothelium of the capillaries was labeled. Some rare vimentin-labeled cells were scattered in the middle area of the stria, representing a vimentin-positive subpopulation of intermediate cells (Fig. 4D). The marginal cells remained immunonegative throughout postnatal maturation. The fibrocytes of the spiral ligament and spiral prominence remained vimentin-labeled (Fig. 4E), while the cells covering the spiral prominence and external sulcus were unlabeled.
From birth onward, numerous cells in the middle area of the stria and capillary endothelial cells were nestin-immunopositive from the base to the apex of the cochlea (Fig. 5A–C). These intermediate cells were scattered between the marginal and basal cell layers. A layer of elongated, strongly positive cells underlined the border between the stria and connective tissue (Fig. 5A) and was in continuity with the Reissner's membrane, which was also positive (Fig. 5B).
Based on the classification of Kikushi et al. (1995), type I and II fibrocytes in the spiral ligament and spiral prominence were poorly stained at birth (Fig. 5A) and the long and slender root cell processes were labeled. From the third week, the type I and II fibrocytes became strongly immunoreactive, whereas types III and IV were more lightly stained (Fig. 5C and D). The cells overlying the spiral prominence and the external sulcus were strongly labeled at birth (Fig. 5A), but immunonegative at the adult stage (Fig. 5D).
NaK/ATPase and S100A6.
The S100A6 antiserum stained the intermediate cells from 24 PD onward, as described elsewhere (Coppens et al., 2001).
From 24 PD onward, the NaK/ATPase-S100A6 double labeling revealed three different cell types in the stria: the luminal layer of NaK/ATPase-labeled marginal cells (brown), and the S100A6-labeled intermediate cells (dark blue) scattered in the middle area and the unlabeled basal cells (Fig 6A and B). The NaK/ATPase-labeled infoldings from the marginal cell penetrated deeply into the stria forming compartments, and closely surrounded the S100A6-labeled cells and the unlabeled strial capillaries (Fig. 6A).
S100A6 and vimentin.
From 24 PD onward, numerous cells in the intermediate area were S100A6-positive (brown) but vimentin-negative. A few small cells scattered in the intermediate area were vimentin-positive (dark blue) (Fig. 6C). The vimentin-immunomarked basal cells sent vimentin-labeled processes penetrating between the S100A6-labeled intermediate cells (Fig. 6C).
Nestin and S100A6.
From 24 PD onward, most of the intermediate cells contained a dark blue and brown staining and thus co-expressed nestin and S100A6 from the base to the apex of the cochlea (Fig. 6D). Both proteins were cytoplasmic. Several marked cells were associated with capillaries. Few intermediate cells remained nestin- and S100A6-negative.
Nestin and vimentin.
At birth, the basal mid-portion of the stria thickness was vimentin-labeled (Fig. 6E). From birth, nestin-labeled cells (brown) were scattered between the unlabeled marginal cell layer and the vimentin-labeled cell layer (dark blue). These nestin-positive cells were in contact with the vimentin-positive area. The unlabeled area including the marginal cells bordering the cochlear duct was narrow. The endothelium of the capillaries was vimentin- and nestin-labeled.
During the first postnatal weeks, the vimentin-labeled area progressively receded to become confined close to the underlying spiral ligament (Fig. 6F). Some scattered intermediate cells retained vimentin expression. During this stage of maturation, the unlabeled marginal cell area became broader, penetrating progressively deeper between the capillaries and the intermediate cells toward the basal area of the stria. The nestin-labeled cells became enclosed in the unlabeled middle area of the stria: the basal processes of the marginal cells. The contact surface between the vimentin-labeled basal cells and the nestin-labeled intermediate cells decreased. The vimentin-labeled processes of the basal cells extended between the vessels and the intermediate cells.
From the end of the third week, the overall pattern of immunodistribution did not change further up to the mature stage (Fig. 6G). No colocalization was observed in the intermediate cells; the intermediate cells were either nestin- or, rarely, vimentin-labeled.
In altricious species, inner ear morphology, endolymph ionic composition, endocochlear potential, and auditory function evolve over a period of few postnatal weeks (Ryan and Woolf, 1992; Steel and Harvey, 1992). In these related processes, the lateral wall, and mainly the stria vascularis appear to play a key role, and each cell type may be important (Souter and Forge, 1998). In the present morphological study, a panel of immunomarkers, including NaK/ATPase β2 isoform, cytokeratins, vimentin, nestin, and S100A6 enabled us to discriminate the different cell types forming the normal lateral wall of the dog cochlear duct, and to follow their normal postnatal maturation. Because the stria vascularis is involved in cochleosaccular degeneration, this study may be a basis for further investigations into the mechanisms leading to cochleosaccular deafness in dogs.
The NaK/ATPase β2 subunit immunostaining enabled us to highlight strial marginal cells (mostly their basolateral membrane) and to follow the growth of the processes they develop toward the basal cell layer during the postnatal period. These marginal cell maturational changes have been reported in other laboratory mammals through light and electron microscopy (EM) studies (Steel and Barklay, 1989; Souter and Forge, 1998). NaK/ATPase is a membrane-bound enzyme composed of α (the transporting) and β (the stabilizing) subunits that participate in the active transport of sodium and potassium through cell membranes. The β2 isoform has been found in rat (ten Cate et al., 1994), gerbil (Nakazawa et al., 1995), and guinea pig (Ichimiya et al., 1994) marginal cells. The β2 isoform was observed from postnatal day 10 in mouse lateral wall, and concentrations of the β subunits increased in a similar manner as the α subunits (Erichsen et al., 1996). A close correlation has been observed between the NaK/ATPase expression and EP establishment in the mouse and rat (Erichsen et al., 1996). A similar correlation could be hypothesized in dog species.
The current study shows that β2 isoform expression can be useful for cell-specific morphological descriptions of marginal cells in dogs. Its expression may also be of potential use in the study of functional aspect, in degenerating stria, since NaK/ATPase is essential for K+ secretion in the endolymph, and indirectly for EP generation and maintenance (Marcus et al., 2002).
In this study, Cks were found to be expressed in marginal cells of dogs from birth, forming a thin network along the luminal border of the lateral wall. The monoclonal antibody RCK102 used in the present work exclusively labels epithelial-derived tissues (Regauer et al., 1985; Anniko et al., 1990; Kuijpers et al., 1991). Ck expression in the marginal cells of dog stria thus confirmed their ectodermic origin. Ck subclasses 5 and 8 have been identified in the stria of man and rat (Anniko et al., 1990; Kuijpers et al., 1991). The strong expression of Cks in marginal cells from birth should allow the identification of this cell type even in a degenerating stria.
The present study revealed, from birth onward, scattered nestin-immunolabeled cells in the intermediary area of the stria from the base to the apex of the dog cochlea that can be interpreted as intermediate cells. To our knowledge, this is the first report about nestin-intermediate filament distribution in the cochlea. Among the IFP family, nestin is a newly described protein, first reported by Tohyama et al. (1992) in embryonic rat neuroepithelium. Nestin has been described in several mammals; it is abundantly expressed in neuroepithelial stem cells of the central nervous system during embryogenesis. Its expression, however, may not be limited to neuroepithelial cells: Previous works have reported nestin expression in various tissues, including mammalian sensory organs, but to date no data are available as regards the inner ear (Tohyama et al., 1992; Pixley, 1996; Namiky and Tator, 1999; Ahmad et al., 2000).
Double immunolabeling in mature dog cochlea revealed co-expression of S100A6 and nestin in most intermediate cells (nestin-S100A6 double immunolabeling), while rare intermediate cells were vimentin-positive; the latter cells failed to react with nestin or S100A6 antibody (S100A6-vimentin and nestin-vimentin double immunolabeling). These results strongly indicate the presence of two intermediate cell populations in the mature dog stria.
It is well documented in laboratory mammals that most intermediate cells are melanocyte-like cells of neuroepithelial origin (Steel and Barklay, 1989; Conlee et al., 1994a, b). However, several morphological studies have described two types of intermediate cells in normal adult laboratory mammals. It has been hypothesized that the two forms of intermediate cells represent either two distinct cell types, or different maturational stages of the same melanocyte-like cell population (Cable and Steel, 1991; Conlee et al., 1994a, b; Motohashi et al., 1994). The numerous nestin/S100A6-positive intermediate cells observed in our material were interpreted as melanocyte-like cells. The nature of the rare vimentin-labeled intermediate cells in the mature dog stria remains undetermined. The vimentin expression suggests a different, mesenchymal origin for these cells (Steel and Barkway, 1989; Carlisle et al., 1990; Motohashi et al., 1994). However, transient vimentin expression in melanocytes has been reported (Schrott et al., 1988). Consequently, the present results cannot favor either one of the two current hypotheses about the nature of this potentially important intermediate cell subpopulation.
Nestin expression indicated that melanocyte-like intermediate cells were present in the dog stria from birth, but underwent postnatal changes. A previous developmental study in dogs reported a progressive increase in S100A6 expression in intermediate cells during the first postnatal month (Coppens et al., 2001). In the present study, we found that this S100A6 labeling occurred the nestin-positive intermediate cells that were present from birth. Consequently, S100A6 labeling may be interpreted as a sign of functional maturity of these cells.
On the other hand, small, rare melanin granules also appeared from the third week, while S100A6 became largely expressed. The onset of melanin synthesis may also indicate a functional maturity of the melanocyte-like intermediate cells. Surprisingly, melanin density, and aggregate formation and spreading through the stria thickness increased through the whole period of observation (until 110 days), which is largely beyond the age at which auditory function has reached maturity in dogs (Poncelet et al., 2000). Although melanocyte-like cells are essential to normal maturation and function of the stria vascularis, the definitive functional or maturational significance of melanin density and distribution remains hypothetical (Steel and Barclay, 1989; Cable et al., 1994; Steel and Harvey, 1992; Conlee et al., 1994a, b). S100A6, nestin, and vimentin immunolabeling, and melanin production may be tools with which to identify intermediary cells and/or detect their maturation failure in degenerating stria.
In cochleosaccular degeneration, the primary abnormality lies in the stria vascularis, which is responsible for generating the EP (which is essential to hearing) (Souter and Forge, 1998; Takeuchi et al., 2000). Intermediate cells (melanocyte-like cells) are involved in this process. The relationship between skin melanocyte defect and strial degeneration is clear in some mouse models with a cochleosaccular type of deafness. The defect results from a failure in melanoblast migration or survival, and several implicated genes have now been identified in this species (Cable et al., 1992; Hardisty et al., 1998). Recently, mutated genes have been identified in human Waardenburg syndromes causing cochleosaccular deafness associated with pigment defects (Hardisty et al., 1998, Steel, 1999).
However, not all forms of cochleosaccular degeneration result from melanocyte-like cell defects in the stria. Indeed, two genes involved in forming ion channels of the marginal cells have been identified, and mutations have been shown to cause the Jervell and Lange-Nielsen syndrome (Hardisty et al., 1998; Steel, 1999).
Although cochleosaccular deafness is also associated with pigment defects, it remains to be determined whether melanocyte-like cell defects are the primary cause of the cochleosaccular degeneration in this species.
Our finding of vimentin expression in the basal cells of dogs throughout postnatal development and at the adult stage argue in favor of their mesenchymal origin, as it is well documented in mammals (Schucknecht, 1993).
Vimentin labeling revealed profound morphological changes of the basal area of the dog stria vascularis during postnatal maturation: basal cells became flattened and sent growing labeled processes toward the luminal surface of the stria. These results parallel those reported in rodents (Steel and Barkway, 1989). It took the dog basal cells 1 month to acquire their mature shape. This observation may complement recent studies that also considered a role of basal cell maturation in the onset and rise of the EP in laboratory species establishing a barrier between perilymph and interstitial space (Souter and Forge, 1998; Marcus et al., 2002). The basal cell maturation is in the same time that the onset and rise of the EP.
Other Cells of the Lateral Wall
The external sulcus cells and the cells covering the spiral prominence were Ck-positive from birth to adulthood, and were transiently vimentin-positive. Co-expression of vimentin and Cks in the epithelial lining of mammalian cochlea has been reported previously (Schulte and Adams, 1989, Bauwens et al., 1991; Kuijpers et al., 1992). The loss of vimentin and nestin immunolabeling in these structures during the first week of life could be interpreted as postnatal maturation. It has been suggested that the epithelial cells overlying the external sulcus contribute to the permeability change of the scala media boundaries during postnatal maturation, and hence to EP production and maintenance (Steel and Harvey, 1992).
Because the spontaneous cochleosaccular lesions described in dogs and humans are comparable, it has been proposed that dogs can serve as an animal model for the study of this type of deafness (Johnsson et al., 1973; Steel and Bock, 1983; Cable et al., 1994; Lalwany et al., 1997). The immunomarkers used in the present study may shed some light on the development of cochleosaccular degeneration through the study of spontaneous cases in dogs. Questions regarding the strial cell type primarily involved, the relationships between skin pigmentation defects and cochleosaccular degeneration, and possible genetic influences on some developmental spots may be addressed through the methods reported herein.
We are greatly indebted to Dr. Michael Nguyen (Laboratory of Molecular Biology, NINDS, Bethesda, MD) for providing us with the anti-nestin antibody. We also thank Doctor A. Résibois for reviewing the manuscript.
- 1996. Morphological and functional characteristics of the different cell types in the stria vascularis: a comparison between cells obtain from fresh tissue preparation and cells cultured in vitro. Hear Res 102: 155–166. , , , , .
- 2000. Identification of neural progenitors in the adult mammalian eye. Biochem Biophys Res Communi 270: 517–521. , , .
- 1990. Regional variations in the expression of cytokeratin proteins in the adult human cochlea. Eur Arch Otorhinolaryngol 247: 182–188. , , , , , .
- 1991. Expression of intermediate filament proteins in the adult human cochlea. Ann Otol Rhinol Laryngol 100: 211–218. , , , , .
- 1994. Effects of mutations at the W locus (c-kit) on inner ear pigmentation and function in the mouse. Pigment Cell Res 7: 17–32. , , , .
- 1991. Identification of two types of melanocyte within the stria vascularis of the mouse inner ear. Pigment Cell Res 4: 87–101. , .
- 1990. Endocochlear potential generation is associated with intercellular communication in the stria vascularis: structural analysis in the viable dominant spotting mouse mutant. Cell Tissue Res 262: 329–337. , , .
- 1994a. Ongoing proliferation of melanocytes in the stria vascularis of adult guinea pigs. Hear Res 79: 115–122. , , .
- 1994b. Pigment-dependent differences in the stria vascularis of albino and pigmented guinea pigs and rats. Hear Res 72: 108–124. , , , .
- 2001. Immunolocalization of the calcium binding S100A1, S100A5 and S100A6 proteins in the dog cochlea during postnatal development. Dev Brain Res 126: 191–199. , , , , .
- 1996. NaK-ATPase α- and β-isoforms in the developing cochlea of the mouse. Hear Res 100: 143–149. , , , , .
- 1998. The molecular genetics of inherited deafness—current knowledge and recent advances. J Laryngol Otol 112: 432–437. , , .
- 1994. Immunolocalization of Na+, K+-ATPase, Ca++-ATPase, calcium-binding proteins, and carbonic anhydrase in the guinea pig inner ear. Acta Otolaryngol (Stockh) 114: 167–176. , , .
- 1973. Vascular anatomy and pathology of the cochlea in Dalmatian dogs. In: de LorenzoAJD, editor. Vascular disorders and hearing defects. Baltimore: University Park Press. 249–295. , , , .
- 1995. Gap junction in the rat cochlea: immunohistochemichal and ultrastructural analysis. Ant Embryol 191: 101–118. , , , .
- 1992. Developmentally-regulated coexpression of vimentin and cytokeratins in the rat inner ear. Hear Res 62: 1–10. , , , .
- 1991. Expression of cytokeratin polypeptides during development of the rat inner ear. Histochemistry 96: 511–521. , , , .
- 1997. A five-generation family with late-onset progressive hereditary hearing impairment due to cochleosaccular degeneration. Audiol Neurootol 2: 139–154. , , , , , , , , , , .
- 2002. KCNJ10 (Kir4.1) potassium channel knockout abolishes endocochlear potential. Am J Physiol Cell Physiol 282: C403–C407. , , , .
- 1994. Dysgenesis of melanocytes and cochlear dysfunction in mutant microphtalmia (mi) mice. Hear Res 80: 10–20. , , , , .
- 1995. Ultrastructural localization of NaK/ATPase in the gerbil cochlea. J. Histochem Cytochem 43: 981–995. , , .
- 1999. Cell proliferation and nestin expression in the ependyma of the adult rat spinal cord after injury. J Neuropathol Exp Neuro 58: 489–498. , .
- 1996. Characterisation of olfactory receptor neurons and other cell types in dissociated rat olfactory cell cultures. Int J Dev Neurosci 14: 823–839. .
- 2000. Maturation of the auditory system in normal dogs as reflected by BAEP wave V latency-intensity curve and rarefaction-condensation differential potentials. Am J Vet Res 61: 1343–1348. , , , .
- 1973. Anatomy and physiology of the onset of auditory function. Acta Otolaryngol 76: 1–10. , .
- 1985. Intermediate filament cytoskeleton of amnion epithelium and cultured amnion epithelial cells: expression of epidermal cytokeratins in cells of a simple epithelium. J Cell Biol 100: 997–1009. , , .
- 1992. Development of the lower auditory system in the gerbil. In: Romand, editor. Development of the auditory and vestibular system. Vol. II. Amsterdam: Elsevier Science Publisher. p 243–271. , .
- 1988. Intermediate filaments in the cochleas of normal and mutant (W/Wv, sl/sld) mice. Arch Otolaryngol 245: 250–254. , , .
- 1993. Pathology of the ear. Philadelphia: Lea and Febiger. p 1–75 .
- 1989. Immunohistochemical localization of vimentin in the gerbil inner ear. J Histochem Cytochem 37: 1787–1797. , .
- 1998. Intercellular junctional maturation in the stria vascularis: possible association with onset and rise of endocochlear potential. Hear Res 119: 81–95. , .
- 1989. Another role for melanocytes: their importance for normal stria vascularis development in the mammalian inner ear. Development 107: 453–463. , .
- 1983. Hereditary inner ear abnormalities in animals: relationship with human abnormalities. Arch Otolaryngol 109: 22–29. , .
- 1992. Development of auditory function in mutant mice. In: Romand, editor Development of the auditory and vestibular system. Vol. II. Amsterdam: Elsevier Science Publisher. p 221–243. , .
- 1999. Deafness genes, expression of surprise. Trends Genet 15: 207–210. .
- 1996. Aetiology, prevalence and diagnostic of deafness in dogs and cats. Br Vet J 152: 17–36. .
- 2000. Mechanism of endocochlear potential: role played by intermediate cells in the stria vascularis. Biophys J 7: 2572–2582. , , .
- 1994. NaK/ATPase and subunit isoform distribution in the rat cochlear and vestibular tissues. Hear Res 75: 151–160. , , .
- 1992. Nestin expression in embryonic human neuroepithelium and in human neuroepithelial tumor cells. Lab Invest 66: 303–313. , , , , , .