Ultrastructural evidence of cell communication between epithelial dark cells and melanocytes in vestibular organs of the human inner ear.код для вставкиСкачать
THE ANATOMICAL RECORD 242:267-277 (1995) Ultrastructural Evidence of Cell Communication Between Epithelial Dark Cells and Melanocytes in Vestibular Organs of the Human Inner Ear MASAZUMI MASUDA, KAZUTO YAMAZAKI, JIN KANZAKI, AND YASUHIRO HOSODA Departments of Otorhinolaryngology (M.M., J.K.) and Pathology (K.Y., Y.H.), Keio University School of Medicine, Tokyo, Japan ABSTRACT Background: The possibility of interaction between epithelial dark cells and melanocytes in the mammalian inner ear has been pointed out because of their morphological and biochemical characteristics, although very few studies have dealt directly with communication between these two types of cells. We investigated the dark cell area of human vestibular organs in order to clarify the ultrastructual evidence for cell interaction between epithelial dark cells and melanocytes. Methods: All of the material was obtained from vestibular schwannoma operations. Paraffin sections were stained with hematoxylin and eosin (H&E) and by the Fontana-Massontechnique. Other paraffin sections were also stained immunohistochemically for S-100 protein. Glutaraldehyde fixed specimens were investigated by scanning electron microscopy (SEM) and transmission electron microscopy (TEM). Results: Light microscopy revealed melanin pigment granules in the cytoplasm of epithelial dark cells. Melanocytes in the subepithelial layer stained positively for 5-100 protein. The presence of intraepithelial melanocytes was confirmed by the presence of cell profiles with a large number of melanin pigment granules and S-100protein in the cytoplasm. SEM showed that the dark cells had a pentagonal surface with microvilli on the apical surface edge. They had complicated structures at the basal portion of their cytoplasm. Melanocytes extending cytoplasmic processes to adjacent areas were observed under the dark cells. TEM showed that the dark cells were tightly linked by junctional complexes in the upper lateral portion of their cytoplasmic membrane and interdigitated by lateral infoldings. Compound melanosomes (phagosomes or secondary lysosomes) found in the cytoplasm of the dark cells contained poorly pigmented melanosomes with a periodic internal structure. Gap junctions were clearly showed between adjacent melanocytes in the subepithelial layer. Conclusions: The characteristic substructures of dark cells and melanocytes suggested the presence of intimate cell interaction between these two types of cells in the vestibular organs of the human inner ear, although it is not clear at this stage whether such cell interaction is specific only for patients with vestibular schwannoma. Dark cells and melanocytes form a cell community that serves to maintain homeostasis in vestibular organs through communication in which cell information obtained by both dark cells and melanocytes serves to facilitate the system. o 1995 Wiley-Liss, Inc. Key words: Human inner ear, Dark cell, Melanocyte, Eumelanosome, Pheomelanosome, Gap junction, Membranous apposition, 5-100 protein A number Of studies have been conducted On the dark and melanocytes Of the inner ear using such experimental mammals as guinea pigs, gerbils, and mice (Quevedo and Fleischmann, 1980; Yoshihara et al., 0 1995 WILEY-LISS, INC. Received October 14, 1994; accepted January 18, 1995. Address reprint requests to Masazumi Masuda, M.D., Department of Otorhinolaryngology, Keio University School of Medicine, 35 Shinanomachi, Shinjuku-ku, Tokyo 160, Japan. 268 M. MASUDA ET AL. TABLE 1. Clinical details of patients Age Case no. and ID 1 M. H. 2 T. 0. 3 K . 0. 4 I. 0. 5 F. I. 6M.T. 7 T. 0. 8M.U. 9 T. K. 10 T. I. llM.N. 12 K.A. 13 C. S. (years) and sex 46,F 43, M 55,F 48; M 52. F 47;F 65, F 31,M 59, M 57, M 42.F 57;F 36, M Side and size' of neurinoma (0. cm) Right (1.0) Right (1.5) Left (2.5) Right (3.0) Left (3.0) Left (2.0) Left (1.0) Left (1.5) Right (1.0) Right (2.0) Left (1.5) Left (1.2) Left (1.5) Chief comdaint Rt. hearing loss Rt. hearing loss Lt. hearing loss Rt. hearing loss, dizziness Lt. hearing loss, facial palsy Lt. hearing loss Lt. hearing loss Lt. hearing loss Rt. hearing loss Rt. tinnitus Lt. hearing loss. dizziness Lt. hearing loss' Lt. hearing loss, tinnitus 'All cases were neurinoma in pathological investigation. 'EMCF = extended middle cranial fossa approach. 3LSC = lateral semicircular canal, SSC = superior semicircular canal, PSC 41HC = immunohistochemically stained with S-100 protein. 1987; Carlisle et al., 1990; Cable and Steel, 1991), but there have been few studies in humans due to the difficulty of obtaining human material (Ishida et al., 1986; Igarashi et al., 1989, Masuda et al., 1994). The results of the studies in experimental mammals have suggested that dark cells play some role in the absorption and secretion of endolymph and perilymph, because they possess well-developed cytoplasmic infoldings that are quite similar to those of renal tubule cells, choroid plexus cells, and the marginal cells of the stria vascularis (Rhodin, 1958; Hinojosa and Rodriguez, 19661, and because of the high concentrations of Na-KATPase (Yoshihara and Igarashi, 19871, Ca-ATPase (Yoshihara et al., 1987), and carbonic anhydrase (Watanabe and Ogawa, 1984) in cytoplasmic infoldings that contain a large number of mitochondria. Several reasons suggest that melanocytes have some essential role in the metabolism of the inner ear: (1) melanosomes themselves have several characteristic biochemical functions, such as being a biological reservoir for divalent ions, including M 2 + and Ca2+,and for other chemical substances essential to enzymatic activation or cell membrane regulation (Sealy et al., 1980; Meyer zum Gottesburge-Orsulakova and Kaufmann, 1986); (2) melanosomes are scavengers and producers of toxic-free radicals (Barr, 1983; Schllreuter and Wood, 1989), and (3) they possess semiconductive properties for various stimuli, phonic, acoustic, and electrical (Mcginnes et al., 1974; Barr, 1983). Intimate interaction between these two types of cells, i.e., epithelial dark cells and melanocytes, can be assumed because of the characteristic physiological function and close location of these cells in the inner ear. In the latest studies, several researchers have pointed out that both melanocytes and their dendritic processes enter the epithelial dark cell layer across the basal lamina under certain conditions, such as experimental endolymphatic hydrops (Meyer zum Gottesburge-Orsulakova, 1988) and exposure to high intensity noise (Kawaguchi, 1992). In the white spotting mouse mutant (Steel and Barkway, 1989; Carlisle et al., 1990),the essential role of melanocytes for the normal development of stria vascularis, especially for normal process of interdigitation = 0peration EMCF~ 1 or 2 2 1 1 2 2 1 1 2 1 2 1 1 1 Surgical materials3 LSC LSC ssc, LSC LSC PSC ssc, LSC Ut, LSC, PSC, ssc LSC, PSC, ssc LSC, PSC, ssc LSC, PSC, ssc PSC. ssc Ut, Lsc PSC, ssc posterior semicircular canal, Ut = Methods TEM TEM TEM TEM TEM TEM TEM TEM SEM TEM IHC~ TEM IHC~ utricle. between marginal and basal cell processes results in generation of the endocochlear potential. The purpose of the present study is to clarify the ultrastructural evidence of cell interaction between dark cells and melanocytes using human material. According to former studies (Ishida et al., 1986; Igarashi et al., 19891, some variations may exist in the morphology of melanosomes in the human inner ear. Melanosomes with internal structure resembling that of eumelanosome have been found in the human fetus (Igarashi et al., 19891, but the melanocytes in the adult human inner ear possess melanosomes with internal structure resembling that of pheomelanosome (Ishida et al., 1986).Therefore, another purpose of this study is not only to identify the cellular phenotype of the melanocytes of the human inner ear by immunohistochemical techniques, but to determine the detailed morphology of the melanosomes in the human inner ear, MATERIALS AND METHODS Cases Thirteen surgical specimens obtained from patients with vestibular schwannoma were studied, all taken from Japanese adults (see Table 1for detailed clinical information). Six of the 13 patients were male, and seven were female. The age of the patients ranged from 31 to 65 years, and their average age was 49 years. Five specimens were from the right ear, and the other eight were from the left ear. Surgery consisted of an extended middle cranial fossa approach (EMCF)type 1 or type 2 (Kanzaki et al., 1991). Material was obtained from the lateral semicircular canal (10 cases), superior semicircular canal (8 cases), posterior semicircular canal (7 cases), and utricle (2 cases). It was immediately processed for light and electron microscopic examination. Light Microscopy The excised specimens were immediately fixed in 20% formalin and processed according to a conventional histologic technique for subsequent hematoxylin and eosin (H&E) staining. We confirmed that there was no marked invasion of the schwannoma and no CELL INTERACTION I N HUMAN INNER EAR other pathological changes. To determine the exact distribution of both melanin pigment granules and melanin pigment containing cells, specimens adjacent to those submitted to light microscopy were stained by the Fontana-Masson technique. Light Microscopic lmmunohistochemistry To make clear phenotypes of cells, especially cells in the epithelial layer, we applied immunohistochemical staining method using S-100 protein for the excised specimen adjacent to those stained with haematoxylin and eosin, and by the Fontana-Masson technique. To localize the immunohistochemical marker S-100 protein, melanin pigment bleeching was performed before the immunohistochemical procedure. Deparaffinized tissue sections obtained from two patients were incubated in a 10% solution of H,Oz in 0.05M Tris-HC1 buffer (pH 7.6) for 5 hours. The immunohistochemical staining reaction was performed after it was confirmed that the color of melanin had completely disappeared through the light microscope. An indirect immunostaining method was applied to these bleeched sections to demonstrate S-100 protein. In brief, sections were exposed to a X 200 dilution of primary antibody for 3 hours a t room temperature. Polyclonal rabbit antihuman S-100protein was purchased from DAKO Co. (CA). Scanning Electron Microscopic Examination (SEM) Specimens were fixed in 2.5% glutaraldehyde for 2 hours at 4°C. The material was washed three times in phosphate buffer, postfixed for 2 hours in 1%similarly buffered OsO4, and dehydrated in graded alcohol. The alcohol that the specimens contained was then replaced by acetaticisoamyl for 10 minutes following a critical point drying procedure. The dried specimens were coated with 8 nm thick platinum and examined with a Hitachi S-4000 scanning electron microscope. Transmission Electron Microscopy (TEM) The material was immediately fixed in 2.5% glutaraldehyde in 0.1 M phosphate buffer (pH 7.4) for 4 hours, washed three times in phosphate buffer, postfixed for 2 hours in 1%similarly buffered OsO4, dehydrated in graded alcohol and acetone, and embedded in Epoxy resin. Thick sections, in the 1 pm range, were stained with toluidine blue. Suitable blocks were thinsectioned using an LKB ultratome with a diamond knife. Sections in the gray to silver range were collected on 150-mesh grids, stained with uranyl acetate and lead citrate, and examined with a Joel-1200EX electron microscope. Observations So far, no apparent differences could be seen between materials taken from different parts of the vestibular organ in a patient through light microscopic and electron microscopic observation. There have been no apparent differences between patients neither. Light Microscopic Investigation Dark cells comprised part of the epithelial lining of vestibular organs. Brown pigment granules were present in both epithelial dark cell layer and subepithelial layer (Fig. la). We employed the Fontana-Masson technique to make clear the localization of these 269 pigment granules. In addition to the dark cells, which had pigment granules stained positively by the Fontana-Masson technique in their cytoplasm, there was another kind of cell with a number of pigment granules in the intraepithelial layer (Fig. lb). With this technique, it is possible that those pigment granules were in fact in the cytoplasmic processes of subepithelial melanocytes interdigitating into the epithelial layer and not in the cytoplasm of dark cells or another kind of intraepithelial cells. Melanocytes that had well-developed dendritic processes, fibroblasts with spindleshape cell bodies, and small blood vessels could be discerned in the subepithelial layer (Fig. la,b). Light Microscopic lmmunohistochemistry After applying the bleeching technique, we could distinguish that the melanocytes in the subepithelial layer stained immunohistochemically for S-100 protein. It was also suggested that the existence of two types of cells in the epithelial dark cell layer, i.e., cells that were positively stained for S-100 protein (intraepithelial melanocytes) and cells that were not positively stained for S-100 protein (epithelial dark cells) (Fig. lc). Scanning Electron Microscopic Investigation Viewed from the endolymphatic lumen, the dark cells usually appeared pentagonal or polygonal (Fig. 2a). There was a specialized surface structure of rich microvilli along the border between the adjacent dark cells. The diameter of these dark cells was 2-4 pm. We could not see any melanocytes on the luminal surface of these dark cells. Viewed from the fractured side, we could see the border between adjacent epithelial dark cells; however, there was no space between neighboring cells (Fig. 2b). These dark cells were 7 pm tall and 3 pm wide in the fractured plane. There were complicated structures at the basal portion of the dark cells. Melanocytes were present immediately below the epithelial dark cells and extended cytoplasmic processes from their cell bodies. Transmission Electron Microscopic lnvestigation Dark cells. Dark cells had irregularly shaped nuclei and cuboidal cell bodies with relatively high electrondense cytoplasm (Fig. 3a). These dark cells ranged from 9 to 12 pm high and from 4 t o 6 pm wide. There were short microvilli in the apical portion of their cytoplasm and well-developed cytoplasmic infoldings in the basolateral portion of their cytoplasm. We observed rough endoplasmic reticulum and numerous mitochondria in their cytoplasm. These dark cells were tightly linked with each other through a junctional complex consisting of tight junctions, zonulae adherentes, and desmosomes in the apical portion of their lateral surface (Fig. 3b). They also had interdigitated lateral cytoplasmic infoldings, and there were no obvious wide spaces around the basolateral cytoplasmic infoldings in relatively well fixed specimens (Fig. 3c). Phagocytosed compound melanosomes (phagosomes or secondary lysosomes) were frequently (1compound melanosomes per every 10-13 dark cells) found in the cytoplasm of dark cells (Fig. 4a). These compound melanosomes enclosed by an unit-limiting- membrane con- 270 M. MASUDA ET AL. Fig. 1. Light micrographs of posterior semicircular canal form case 11. (a) H&E stain. Epithelial dark cells (E), melanocytes (M), fibroblasts(F), and intraepithelial melanocytes (arrowheads) with a number of melanin granules. x 330 (b) Fontana-Masson stain. Epithelial dark cells (E) with melanin granules. Im: intraepithelial melanocytes. M: melanocytes in the subepithelial layer. 13:blood vessel. x 330 (c) lmmunohistochemicalstain with S-100 protein. Some of the intraepithelial cells (arrowheads) as well as melanocytes (M) in the subepithelial layer are positively stained for S-100protein. x 330 CELL INTERACTION IN HUMAN INNER EAR Fig. 2. SEM of superior semicircular canal from case 9. (a) Endolymphatic luminal surface view of epithelial dark cells. Dark cells (E) exhibiting a pentagonal or polygonal surface. Microvilli (arrowheads). ~ 4 0 0 0(b) Fractured side view of the vestibular dark cell area. Lateral cell borders (small arrowheads) between two adjacent dark cells. Complicated structures (large arrowheads) a t the basal portion of the dark cells. Subepithelial dendritic melanocyte (M). x 4000 271 Fig. 3. CELL INTERACTION IN HUMAN INNER EAR 273 Fig. 4. TEM of lateral semicircular canal from case 10(a), and case 3(b,c). (a) Liberated melanin granules (P) in the endolymphatic lumen. Compound melanosomes (C) in the cytoplasm of dark cells (E). Melanosomes (arrowheads) in the dendritic processes of subepithelial melanocytes. x 9000 (b,c) High magnification views of poorly pigmented melanosomes in the compound melanosomes. An unit-limiting membrane (arrowheads) enclosing melanosomes could be seen. x 115000 Fig. 3.TEM of lateral semicircular canal from case 10(a,c)and case l(b,d). (a)Melanin granules (P) in the endolymphatic lumen. Epithelial dark cells (E) with microvilli on their luminal surface. Melanosomes (arrows) in the cytoplasm of dark cells. Melanosomes (arrowheads) in both intraepithelial and subepithelial melanocytes. Mi: a number of mitochondria. Im: intraepithelial melanocytes. M: subepithelial melanocytes. x 4300 (b)Junctional complex of epithelial dark cells. Tight junction (arrowheads), zonula adherens (A), and desmosome (D). X 52000 ( c ) Lateral infoldings of epithelial dark cells. E: epithelial dark cells. X 20000 (d)High-magnification view of melanosomes in the cytoplasm of subepithelial melanocytes. Poorly pigmented melanosomes (Stage I1 or I11 pheomelanosomes)that have the spherical granular form of internal structure range from 200 to 400 nm in diameter. x 55000 274 M. MASUDA ET AL. tained several poorly pigmented melanosomes. They had the ellipsoidal lamellar form of internal structure compatible with that of eumelanosome (Jimbow et al., 1979; Quevedo and Fleischmann 1980; Jimbow e t al., 1983) (Fig. 4b,c). These melanosomes ranged from 200 to 500 nm long. We also occasionally found melanin granulelike material in the endolymphatic lumen (Figs. 3a, 4a). lntraepithelial rnelanocytes. In the present study we frequently (one melanocyte per every 7-9 dark cells) found melanocytes within the epitheleial dark cell layer. They had ellipsoidal nuclei with fairly evenly dispersed chromatin and relatively clear cytoplasm (Fig. 3a). The shape and size of these intraepithelial melanocyte cell bodies varied slightly, but their mean size was 6 x 10 pm. In the cytoplasm there were spherical melanosomes and cell organelles, such as mitochondria and rough endoplasmic reticulum. Poorly pigmented melanosomes revealed the spherical granular form of internal structure. The morphological characteristics of these melanosomes were compatible with those containing pheomelanin (Stage I1 or I11 pheomelanosomes) (Jimbow et al., 1979; Quevedo and Fleischmann 1980; Jimbow et al., 1983) and their average size was 300 x 400 nm (Fig. 3a). There were no marked differences between the shape and size of the melanosomes in the epithelial melanocytes and in the subepithelial melanocytes (Fig. 3d). We also observed two parallel opposing membranes between intraepithelial melanocytes and the dendritic processes of other melanocytes (Fig. 5a). The gap between the outer leaflets was 24 nm wide and the gap between inner leaflets was 6 nm wide (Fig. 5b). Gap junctions. In the subepithelial layer, gap junctions were clearly recognized between the plasma membranes of adjacent melanocytes as well as between melanocyte cell bodies and their own cytoplasmic processes (Fig. 6a,b) (Staehelin, 1974). The gaps between the two outer leaflets of the plasma membranes of these gap junctions were 12 nm, and gaps between the inner leaflets of these gap junctions were 2 nm (Fig. 6b). Periodic substructures with relatively constant repeating lattices were seen in the interspace between the opposed plasma membranes of gap junctions. We were able to observe a t most 3-4 gap junctions around a single subepithelial melanocyte. DISCUSSION lntraepithelial Melanocytes In the present study, examination of H&E, FontanaMasson, and immunohistochemically stained specimens suggested the existence of intraepithelial melanocytes. Detailed TEM observations clearly showed the presence of both intraepithelial melanocytes and subepithelial melanocytes. At the same time, these melanocytes were positively stained for S-100 protein, demonstrating that they originated in neural crest during development (Clemmensen and Fenger, 1991). Cells that do not originate in neural crest can possess melanin pigment granules. This has been clearly shown in such spotting mouse mutants as the S1/Sld mouse and the W/W” mouse (Mayer and Green, 1968). Fig. 5. TEM of posterior semicircular canal from case 7. (a) Membranous appositions with narrow gaps (small arrowheads) between the intraepithelial melanocyte and another melanocyte dendrite. Small vesicle (large arrowhead) in the dendritic process of a melanocyte. E: nucleus of a n epithelial dark cell. N: nucleus of a n intraepithelial melanocyte. x 18000 (b) High-magnification view of the same membranous appositions(arrowhead). Small vesicle (V) with a round configuration and short spikelike projections. x 37000 There are no melanocytes in inner ear in these animals, and they have black eyes whose melanin pigment is produced by retinal epithelial cells embryonically derived from the optic cup. In the present study the melanocytes in the vestibular dark cell area were found to have the same origin as those in other parts of the human body, e.g., in the skin, mucous membranes, lentigo, pia mater, brain, and choroid membrane (Jimbow e t al., 1976). Melanosomes in Human Inner Ears In the present study we occasionally found compound melanosomes (phagosomes or secondary lysosomes) in CELL INTERACTION I N HUMAN INNER EAR 275 Fig. 6. TEM of lateral semicircular canal from case 1. (a)Gap junction (arrowhead) between a subepithelial melanocyte cell body and a dendritic process. N nucli of subepithelial melanocytes. M: melanosomes. Mi: mitochondria. F: fine filamentous material around the melanocyte and its dendritic processes. x 18000 (b) High-magnification view of the same gap junction. G: gap junction. x 100000 the cytoplasm of dark cells. Some of them contained poorly pigmented melanosomes. They had the ellipsoidal lamellar form of internal structure compatible with that of eumelanosome (Jimbow et al., 1979; Quevedo and Fleischmann 1980; Jimbow et al., 1983). Igarashi et al. (1989) showed that melanosomes in the inner ear display an internal structure resembling that of eumelanosome using human fetal material obtained when the inner ear structure had attained developmental maturity (Igarashi and Ishii, 1980).Ishida et al. (19861, in contrast, said melanosomes exhibit an internal structure resembling that of pheomelanosome in the adult human inner ear and that these melanosomes have the biochemical characteristics of eumelanin, although there is a discrepancy in the form of their internal structure. The melanosomes in our studies of human adults, mainly displayed an internal structure resembling that of pheomelanosome. Although there is evidence that many natural melanins are copolymers of eumelanin and pheomelanin and that hardly any pure eu- melanin or pheomelanin exist in mammals (Prota, 1980), it is generally accepted that the morphological characteristics of melanosomes, i.e., ellipsoidal lamellar eumelanosome and spherical granular pheomelanosome correspond well to the biochemical characteristics of each melanin content, i.e., eumelanin and pheomelanin according to the studies on melanogenesis in normal skin and hair follicles (Jimbow et al., 1983). Thus based on the fact that there are differences between the shape of human fetal and adult melanosomes and that a few melanosomes with an internal structure resembling that of eumelanosome are actually present in specimens from human adults, we can speculate that the switch from eumelanogenesis to pheomelanogenesis occurs around the time of birth. In a study on human red hair, two different populations of melanocytes, generating eumelanin and pheomelanin, respectively, were found to be present in the same hair follicle (Jimbow et al., 1983). In agouti mice, in contrast, eumelanosomes and pheomelanosomes coexist in the same melanocytes because of a switch from eumelanogenesis 276 M. MASUDA ET AL. to pheomelanogenesis during the normal growth of hair follicles (Sakurai et al., 1975). It is said that generally the genetic factors of not only melanocytes but of the tissue around them in various organs have a powerful effect on the type of melanogenesis in melanocytes (Quevedo and Fleischmann, 1980). Now we suspect that these changes in eumelanogenesis and pheomelanogenesis in the human inner ear are caused by differences in the environment before and after the birth, such as sound stimuli being directly delivered to the ear, not via the maternal body, and exposure to ultraviolet rays that are capable of stimulating melanocytes in other parts of the body (Jimbow and Fitzpatrick, 19751, or by the effects of aging alone. Cable et al. (1991) suggested in their study of melanocytes, using the inbred CBNCa strain and normal littermates of the viable dominant spotting mouse mutant, that aging could be one of the causes of the morphological changes in the melanocytes of the inner ear. To the best of our knowledge, no biochemical analysis of melanogenesis in the human inner ear has been performed; however, if this kind of change in melanogenesis really does occur in the human inner ear and is characteristic only of the melanocytes of the human inner ear, it will support the concept of melanocytes playing an important role in inner ear metabolism. Melanin Transmission Our light microscopic and TEM observations frequently showed that melanosomes are actually present in the cytoplasm of epithelial dark cells (Masuda et al., 1994). This suggests the possibility of melanosome transmission from both intra- and subepithelial melanocytes to dark cells. Earlier researchers have noted the disappearance of the plasma membranes of both the basal infoldings of dark cells and melanocyte dendrites in the subepithelial layer. Fusion between the cytoplasmic processes of these two cells has been described (Meyer zum Gottesburge-Orsulakova, 1988; Igarashi et al., 19891,but this disappearance of plasma membranes is caused by the cutting angle of the specimen, and phagocytosis seems to be a rational explanation when we consider the process of melanosome transmission between dark cells and melanocytes. Moreover, we sometimes found that melanosomes were present in the endolymphatic lumen just beyond epithelial dark cells containing compound melanosomes. We hypothesize two possibilities based on this findings. One is that this only represents a waste disposal process, i.e., that melanocytes that died during the normal cell cycle were being transferred into the epithelial layer and then into the endolymphatic lumen. The other is that it represents a process of melanosome secretion from epithelial dark cells into the endolymph, i.e., that when melanocytes are in an activated metabolic state, the melanocytes in the subepithelial layer synthesize a number of melanosomes (Masuda et al., 1994) and that these melanosomes are transferred into the epithelial dark cells for secretion into the endolymph. The possibility of phagocytic uptake of melanin pigment granules by epithelial dark cells from the endolymphatic lumen still remains, although how and why the granules occur in the lumen pose a problem. Gap Junctions In the previous report we demonstrated the presence of gap junctions in the melanocytes of the human inner ear by means of detailed TEM examination (Masuda et al., 1994). Thus far, gap junctions have been found between the plasma membranes of adjacent melanocytes and between melanocyte cell bodies and their own dendritic processes in the subepithelial layer, but no gap junctions have been detected connecting the apposed plasma membranes of melanocytes and other kinds of cells. Currently the most general and important function of gap junctions is believed to be the transmission of regulatory and signaling substances that might play an inductive or repressive role in controlling the rate of division and differentiation or response to hormones in both developing and adult tissues. Moreover, gap junctions can provide cells with a way to obtain cellular information in a cell community (Staehelin, 1974; Alberts et al., 1983). In the vestibular organs of the human inner ear, melanocytes can construct a cell community by transmitting cell information or other kinds of signals through these gap junctions. Cell Communication in Human Vestibular Organs In our earlier studies we frequently observed isolated cilia in subepithelial melanocytes (Masuda et al., 1994). Although there is still controversy about the physiological functions of isolated cilia at this stage (Masuda et al., 19941,the following hypothesis is based on the physiological functions of appropriate melanocyte substructures, such as isolated cilia and gap junctions in the dark cell area of human vestibular organs. Melanocytes in the subepithelial layer may monitor environmental changes around them via isolated cilia. This information is then transmitted from one melanocyte to another at a suitable speed and integrated via gap junctions. Based on this cellular information, stimulated melanocytes try actively to generate melanosomes (Masuda et al., 1994). In earlier studies it was speculated that exposure to noise activates the metabolic state in the tissues of the organ of Corti and stria vascularis (Ryan et al., 1982) and that this results in increased levels of toxic free radicals and unpaired electrons. Melanosomes have characteristic functions as powerful scavengers of toxic free radicals (Ban-, 1983; Schllreuter and Wood, 1989) and ability to bind these compounds (Sealy et al., 1980; Meyer zum Gottesburge-Orsulakova and Kaufmann, 1986). Melanocytes monitoring various environmental changes in the tissues of the subepithelial layer may produce melanosomes to achieve homeostasis in the vestibular organs. Exposure of the inner ear to noise also leads to increased levels of Ca2+ in the endolymph (Ikeda et al. 1988). Characteristic dark cell enzymes, such as Ca2+' ATPase, would be of help in maintaining Ca2+concentrations in the endolymph (Yoshihara et al., 1987). Moreover, it appears that dark cells utilize melanosomes and may excrete melanosomes into the endolymph, because melanosomes have the ability to serve as divalent ion reservoirs (Sealy et al., 1980; Meyer zum Gottesburge-Orsulakova and Kaufmann, 1986). The observation that melanocytes in the vestibular organs of guinea pigs are activated to produce melanosomes because of imbalanced calcium homeo- CELL INTERACTION I1X HUMAN INNER EAR stasis in experimental endolymphatic hydrops corroborates this suspicion (Meyer zum Gottesburge-Orsulakova, 1988). Based on transcellular information about environmental changes around them, some subepithelial melanocytes transfer their dendritic processes into the epithelial layer for effective transmission of melanosomes to dark cells. For the same purpose, melanocyte cell bodies themselves sometimes move in between epithelial dark cells and turn into intraepithelial melanocytes, although some intraepithelial melanocytes may be present between epithelial dark cell layers from the outset. Based on changes in dark cell metabolism, those melanosome generation and transmission may be regulated by signal transfer between dark cells and melanocytes via junctional specializations such as gap junctions. If dark cells themselves really have some perceptive role through their apical membrane specializations, e.g., microvilli, dark cells could monitor the various changes in endolymph directly, more effectively, and more rapidly than monitoring by melanocytes in the subepithelial layer. This would suggest a communication system facilitated by transcellular signals resulting from the perceptive function of both epithelial dark cells and melanocytes. Based on the ultrastructural morphological evidence in the present study, which suggests the possibility of cell communication, we hypothesize that the cell community, which mainly consists of dark cells and melanocytes, works as a whole to maintain the homeostasis of vestibular organs by transmission of intercellular signals or information via gap junctions and transmission of essential materials by melanosome transmission, although we cannot completely deny the possibility that such relatively active cell communication between melanocytes and dark cells is specific only for the patients with vestibular scwannoma. ACKNOWLEDGMENTS We express gratitude t o Dr. Y. Inoue, Dr. Y. Satoh, Dr. S. Yoshihara, and Dr.T. Matsunaga for kindly providing specimens, and to Dr. Y. Toyama (Dept. of Anatomy, Chiba Univ.) for valuable discussion. The technical assistence of Y. Enomoto, N. Yagi, K. Takeichi, and T. Fujiwara is gratefully acknowledged. LITERATURE CITED Alberts, B., D. Bray, J . Lewis, M. Raff, K. Roberts, and J.D. Watson 1994 Cell junctions. In: Molecular Biology of the Cell, 3rd ed. Alberts, B. et al. (eds) Garland, New York, pp. 949-1009. Barr, F. E. 1983 Melanin: The organizing molecule. Med. Hypoth., 11:l-140. Cable, J., and K. P. Steel 1991 Identification of two types of melanocyte within the stria vascularis of the mouse inner ear. Pigment Cell Res., 4(2):87-101. Carlisle, L., K. P. Steel, and A. Forge 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(2):329-337. Clemmensen, 0.J., and C. Fenger 1991 Melanocytes in the anal canal epithelium. Histopathology, 18(3):237-241. Hinojosa, R., and E. E. Rodriguez 1966 The fine structure of the stria vascularis of the cat inner ear. Am. J. Anat., 118(2):631-663. Igarashi, Y., and T. Ishii 1980 Embryonic development of human organ of Corti: Electron microscopic study. Int. J . Ped. Otolaryngol., 2 5 - 6 2 , Igarashi, Y., I. Takeyama, and I. Takahashi 1989 Melanocytes in 277 vestibular dark cell areas in human fetuses. Acta Otolaryngol. (Stockh), 108(1-2):9-18. Ikeda, K., J . Kusakari, and T. Takasaka 1988 Ionic changes in cochlear endolymph of the gunea pig induced by acoustic injury. Hear Res., 32.103-110. Ishida, O., T. Ishii, and K. Jimbow 1986 Electron microscopic study of melanocytes and melanosomes of human inner ear. Nippon Hifuka Gakkai Zasshi, 96(8):829-834. Jimbow, K., and T. B. Fitzpatrick 1975 Changes in distribution pattern of cytoplasmic filaments in human melanocytes during ultraviolet-mediated melanin pigmentation: The role of the 100Angstrom filaments in the elongation of melanocytic dendrites and in the movement and transfer of melanosomes. J . Cell Biol., 65(2):481-488. Jimbow, K., 0. Ishida, S. Ito, Y. Hori, C. J . Witkop, and R. A. King 1983 Combined chemical and electron microscopic studies of pbeomelanosomes in human red hair. J. Invest. Dermatol., 81(6): 506-511. Jimbow, K., 0. Oikawa, S. Sugiyama, and T. Takeuchi 1979 Comparison of eumelanogenesis and pheomelanogenesis in retinal and follicular melanocytes: Role of vesiculo-globular bodies in melanosome differentiation. J . Invest. Dermatol., 73(4)4):278-284. Jimbow, K., W. J . Quevedo, T. B. Fitzpatrick, and G. Szabo 1976 Some aspects of melanin biology: 1950-1975. J . Invest. Dermatol., 67(1):72-89. Kanzaki, J., R. Shiobara, and S. Toya 1991 Classification of the Extended Middle Cranial Fossa Approach. Acta Otolaryngol. (Stockh) Suppl., 487.6-16. Kawaguchi, K. 1992 Susceptibility of organ of corti with or without melanin to acoustic overstimulation. J . Otolaryngol. Jpn., 95(4): 556-566. Masuda, M., K. Yamazaki, J. Kanzaki, and Y. Hosoda 1994 Ultrastructure of melanocytes in the dark cell area of human vestibular organs-Functional implications of gap junction, isolated cilia, and annulate lamellae. Anat. Rec., 240.481-491. Mayer, T. C., and M. C. Green 1968 An experimental analysis of the pigment defect caused by mutations a t the W and S1 loci in mice. Dev. Biol., 18(1):62-75. Mcginnes, J. E., P. M. Corry, and P. H. Proctor 1974 Amorphus semiconductor switching in melanins. Science, 183.853-855. Meyer zum Gottesburge-Orsulakova, A.-M. 1988 Physiology and pathophysiology of inner ear melanin. Pigment Cell Res., 1(4): 238-249. Meyer zum Gottesburge-Orsulakova, A.-M., and R. Kaufmann 1986 Is a n imbalanced calcium-homeostasis responsible for the experimental induced endolymphatic hydrops? Acta Otolaryngol. (Stockh), 102:93-98. Prota, G. 1980 Recent advances in the chemistry of melanogenesis in mammals. J. Invest. Dermatol., 75.122-127. Quevedo, W. J., and R. D. Fleischmann 1980 Developmental biology of mammalian melanocytes. J . Invest. Dermatol., 75(1):116-120. Rhodin, J. 1958 Anatomy of kidney tubules. Int. Rev. Cytol., 7.485534. Ryan, A. F., P. Goodwin, N. K. Woolf, and F. Sharp 1982 Auditory stimulation alters the pattern of 2-deoxyglucose uptake in the inner ear. Brain Res., 234(2):213-225. Sakurai, T., H. Ochiai, and T. Takeuchi 1975 Ultrastructural change of melanosomes associated with agouti pattern formation in mouse hair. Dev. Biol., 47(2):466-471. Schllreuter, K. U., and J . M. Wood 1989 Free radical reduction in the human epidermis. Free Radic. Biol. Med., 6.519-532. Sealy, R. C. C., C. C. Felix, J . S. Hyde, and H. M. Scwarts 1980 Structure and reactivity of melanins: influence of free radicals and metal ions. In: Free radicals in biology, Vol. 4. Pryor, W. A., ed. Academic Press, New York, pp. 209-259. Staehelin, L. A. 1974 Structure and function of intercellular junctions. Int. Rev. Cytol., 39(191):191-283. Steel, K.P., and C. Barkway 1989 Another role for melanocytes: their importance for normal stria vascularis development in the mammalian inner ear. Development, 107(3):453-463. Watanabe, K., and A. Ogawa 1984 Carbonic anhydrase activity in stria vascularis and dark cells in vestibular labyrinth. Ann. Otol. Rhino]. Laryngol., 93(3 Pt 1):262-266. Yoshihara, T., and M. Igarashi 1987 Cytochemical study of K+-pnitrophenyl phosphatase (K + -NPPase) in planum semilunatum and dark cells of the squirrel monkey. Acta Otolaryngol. (Stockh), 104(1-2):22-28. Yoshihara, T., M. Igarashi, M. Usami, and T. Kanda 1987 Cytochemical studies of Ca + + -ATPase activity in the vestibular epithelia of the guinea pig. Arch. Otorhinolaryngol., 243(6):417-423.