THE ANATOMICAL RECORD 251:114–121 (1998) Ultrastructure of the Binary Parotid Glands in the Free-tailed Bat, Tadarida thersites. I. Principal Parotid Gland TOSHIKAZU NAGATO,1* BERNARD TANDLER,2 AND CARLETON J. PHILLIPS3 1Second Department of Oral Anatomy, Fukuoka Dental College, Fukuoka 814-0193, Japan 2Department of Oral Anatomy II, Kyushu Dental College, Kitakyushu 803-8580, Japan 3Department of Biological Sciences, Illinois State University, Normal, Illinois 61790-4120 ABSTRACT Background: Many species of bats have two sets of submandibular glands, principal and accessory. The accessory gland may resemble the principal one but more often shows wide morphological divergence. The free-tailed bat, Tadarida thersites, is very unusual in that it has two sets of parotid glands rather than binary submandibular glands. We studied the ultrastructure of the principal parotid gland to establish a baseline for comparison with the accessory parotid. Methods: Two specimens of adult free-tailed bats, one male and one female, were live-trapped in western Kenya. Parotid glands were fixed for electron microscopy using a protocol expressly designed for field fixation and then embedded by conventional means. Results: Histologically, the principal parotid is a typical serous gland. The secretory granules of the endpiece cells have an unusual substructure in that they contain variable numbers of lucent halos and one or several spherules. Intercalated duct cells contain a significant number of dense, serous-like granules. Striated ducts have the usual basal configuration of mitochondria and folded plasma membranes, but the supranuclear cytoplasm contains many small, dense granules, so that these ducts resemble the granular convoluted tubules found in the submandibular glands of many families of rodents. The apices of the duct cells have a peculiar contour—the luminal surfaces obliquely invaginate into the apical cytoplasm, so that in thin section the luminal membranes appear to be underlaid by a layer of vacuoles. Conclusion: Although the principal parotid gland of the free-tailed bat shows some distinctive, species-specific ultrastructural features, it basically is similar to the parotid gland in two other molossid bats, Tadarida brasiliensis and Molossus molossus. The distinctive features in the principal parotid gland of T. thersites might relate to its feeding on hard-bodied insects and perhaps to the production of lysozyme. Anat. Rec. 251:114–121, 1998. r 1998 Wiley-Liss, Inc. Key words: principal parotid gland; salivary glands; granular ducts; chiropterans; bats; Tadarida thersites In his comprehensive review of chiropteran salivary glands, Robin (1881) noted the presence in some species of two sets of submandibular salivary glands, which he labeled principal and accessory. His designations were based solely on gross anatomical considerations. In our large-scale, comparative ultrastructural study of bat salir 1998 WILEY-LISS, INC. *Correspondence to: Dr. Toshikazu Nagato, 2nd Department of Oral Anatomy, Fukuoka Dental College, 2-15-1 Tamura, Sawaraku, Fukuoka 814-0193, Japan Received 23 December 1997; Accepted 20 January 1998 PRINCIPAL PAROTID GLAND IN TADARIDA vary glands, we have come across such binary salivary glands in a number of species. Rather than using gross anatomical criteria as the basis for labeling, we have based identification of principal and accessory glands entirely on microscopic features (Tandler et al., 1996a). The usual situation is that one of the sister glands is a close likeness of conventional glands (this we designate as principal), whereas the second gland may exhibit morphological— and sometimes histochemical—differences from typical glands (this we designate as accessory). Although binary submandibular glands are not uncommon either in bats or in other kinds of mammals, accessory parotid glands are extremely rare. In fact, the only species in which they are reported to occur are humans (reviewed by Toh et al., 1993) and several other primates (Huntington, 1913); rats, in which they are called anterior buccal glands (Redman, 1972); sheep (Rodriguez-Veiga and Zúñiga, 1985); and possibly laboratory mice (Domon and Shiga, 1986). We have found that the free-tailed bat, Tadarida thersites, also has an accessory parotid gland. Both principal and accessory glands have a number of unusual features, although the principal one is closer in structure to conventional chiropteran parotid glands. The ultrastructure of the principal parotid gland is detailed in the present report; that of the accessory parotid gland is described in the accompanying article (Tandler et al., 1998). MATERIALS AND METHODS Two adult specimens (one of each sex) of the free-tailed bat, T. thersites, were collected by mist-netting in the Kakamega Forest of western Kenya. Voucher specimens (CJP 4889 and 4790) were deposited in the research collection of the Carnegie Museum of Natural History, Pittsburgh, Pennsylvania. The bats were killed in the field with T-61 Euthanasia Solution; their major salivary glands were immediately extirpated, immersed in fixative, and minced. The initial fixative was cacodylate-buffered, halfstrength Karnovsky’s (1965) solution in 2.5% dimethyl sulfoxide and contained 0.01 M LiOH and 0.01 M CaCl2. After 20 hr in this mixture, tissues were washed and stored in 0.5 M cacodylate buffer at ambient temperature. Once refrigeration became available, tissue samples were transferred to fresh 3% glutaraldehyde and stored at 4°C (Phillips, 1985). In the laboratory, the parotid gland samples were extensively washed in phosphate-buffered sucrose and then postfixed for 2 hr in 2% osmium tetroxide in the same buffer. After rinsing in distilled water, specimens were soaked overnight in acidified uranyl acetate (Tandler, 1990). Another aqueous rinse was followed by dehydration in ascending concentrations of ethanol, passage through propylene oxide, and embedment in Epon: Maraglas (Tandler and Walter, 1977). Thin sections were serially stained with acidified uranyl acetate (Tandler, 1990) and lead citrate (Venable and Coggeshall, 1965) and examined in a JEOL 1200EX electron microscope (JEOL, Tokyo, Japan). RESULTS At low magnification, the principal parotid gland of T. thersites conforms in structure to conventional parotid glands (Fig. 1). Secretory endpieces consist of wedgeshaped cells that contain numerous serous granules (classification based on ultrastructural morphology; cf. Young and van Lennep, 1978; Tandler and Phillips, 1993) (Fig. 2). 115 The granules are rather large, measuring up to 1.4 µm in diameter. Some contain a single (often several) eccentrically placed, dense spherule(s); these almost always are encircled by a thin lucent halo. Small lucent haloes—some unrelated to spherules, some containing material that matches the spherules in density—may be scattered throughout the granule matrix; these display a proclivity for lining up against the inner aspect of the limiting membrane (Fig. 3). Like other serous cells, those in the principal parotid of T. thersites have a well-developed rough endoplasmic reticulum (RER) and a prominent Golgi apparatus. Lipid droplets tend to be abundant in some cells, almost always in an infranuclear position. The serous cells surround a more or less central lumen; intracellular canaliculi originating from the endpiece lumen pervade the endpieces. The serous cells lack basal folds. Very flat myoepithelial cells are present at the endpiece periphery; their perikarya usually are located near the endpiece-intercalated duct junction, and their processes are joined to the adjacent serous cells by desmosomes. Intercalated ducts consist of simple cuboidal epithelial cells (Fig. 1) that often are edged by myoepithelial cell processes (Fig. 4). Although almost every duct cell contains a number of serous-type granules in its supranuclear cytoplasm, some cells have more than others (Fig. 5). These granules are smaller (,0.45 µm) than those in the endpiece cells and have a homogeneously dense matrix. In keeping with their secretory function, the duct cells have an extensive Golgi apparatus. Where the Golgi saccules are flat and parallel, a moderately dense material may be present between adjacent saccules. Myoepithelial cell processes are particularly numerous on these ducts. Consisting of pseudostratified epithelium, the major intralobular ducts show a number of departures from conventional striated ducts. Even at very low magnification (Fig. 6) it is apparent that although they exhibit basal striations, they contain secretory granules in their apical cytoplasm and that their luminal surface has an erose contour. Figure 7 shows the overall cytology of these duct cells—the infranuclear portion of the cell is typical in structure, whereas the supranuclear cytoplasm is suffused with granules. At higher magnification (Fig. 8), the basal striations are revealed as consisting of highly plicated plasma membranes interspersed with mitochondria. The most striking feature of the tall cells is their content of numerous serous-type granules (Figs. 6, 7, 9). These are homogeneously dense and measure approximately 1.0 µm in diameter. They largely are confined to a subluminal zone; between the granules and the nucleus is a region with a considerable amount of RER. A relatively small Golgi apparatus sometimes caps the nucleus. The luminal surface of the tall duct cells bears many short, stubby microvilli. This surface is highly indented, so that a section passing through an appropriate plane gives the impression that the luminal plasmalemma is underFig. 1. (overleaf.) Survey electron micrograph of the principal parotid gland of Tadarida thersites. The gland consists mainly of serous acini composed of heavily granular cells. An intercalated duct is indicated by the arrow; the light-appearing cell in its wall is a mononuclear wandering cell. 31,150. Fig. 2. (overleaf.) A transverse section through an entire endpiece showing the general morphology of the serous cells and their abundance of serous granules. 33,900. 116 NAGATO ET AL. Figs. 1 and 2. PRINCIPAL PAROTID GLAND IN TADARIDA 117 Fig. 3. Higher magnification view of endpiece serous granules. The patterns exhibited by these granules depends on just a few substructural elements: dense spherules of varying size, each of which is delimited by a light halo, and haloes that encompass matrix material rather than spherules. The haloes tend to be aligned just within the limiting membrane of the granules. 336,000. laid by a series of lacunae (Fig. 9). Because the lining membrane of these pseudolacunae is, in fact, the luminal membrane, it also bears microvilli. These indentations bear many V-shaped invaginations, indicating that they are a preferred site for exocytosis of duct granules to take place. The sparse basal cells have a triangular shape, with the triangle base resting on the basement membrane, to which they are attached by hemidesmosomes. They also are joined to their taller neighbors by desmosomes. Their few mitochondria are considerably smaller than those in the tall cells. in the latter species, the principal parotid gland in T. thersites has ‘‘conventional’’ secretory endpieces and as is typical (but not uniformly so) of insectivorous bats, the acinar cells can be described as serous (Phillips et al., 1987, 1993). Serous granules in the acinar cells of T. thersites add another type of substructure to the already extensive gallery of serous granule morphology (Phillips et al., 1987; Tandler and Phillips, 1993). In addition to the spherules, aliquots of matrix material are delimited by light halos. DISCUSSION Fig. 4. (overleaf.) Transverse section of an intercalated duct. The duct cells contain a variable number of small, dense secretory granules. A myoepithelial cell process (Mec) borders the duct on the right. The light cells in the left half of the duct wall are mononuclear wandering cells that have invaded the ductular epithelium. 36,500. Fig. 5. (overleaf.) The apical cytoplasm of an intercalated duct cell that is provided with many serous-type secretory granules. These granules lack the substructure of their acinar cell counterparts. 328,800. The molossid free-tailed bat, T. thersites, is a member of the subgenus Mops, which is characterized by insectivorous bats specialized to feed on beetles (Freeman, 1979). In general histology and ultrastructure, the principal parotid gland of T. thersites is somewhat similar to a geographically wide-ranging New World species, T. brasiliensis. As 118 NAGATO ET AL. Figs. 4 and 5. PRINCIPAL PAROTID GLAND IN TADARIDA 119 Fig. 6. Survey electron micrograph of an obliquely sectioned striated duct. Even at this low magnification, it is evident that the duct cells contain dense secretory granules and that their apical surfaces have a peculiar configuration. 3780. Fig. 7. Several striated duct cells whose supranuclear cytoplasm contains an abundance of dense secretory granules. 33,700. Although a similar pattern of halo-delineated matrix material is present in the serous granules of the canine parotid gland (Nagato and Tandler, 1986), those in the dog do not have a propensity for alignment on the inner aspect of the granule membrane. Secretory granules with multiple spherule-like inclusions are present in submandibular acinar cells of the nine-banded armadillo; each of these small inclusions is enclosed by a relatively thick, lucent halo (Ruby and Canning, 1978). Serous granules in the principal parotid gland of T. thersites have a radically different substructure from those in the parotid gland in another beetle-eating molossid bat, Eumops glaucinus, that contain a single, large spherule with a distinctive substructure. In contrast, the serous granules in T. thersites bear some resemblance to the great majority of those in the parotid gland of T. brasiliensis; the latter granules have numerous small spherules that are scattered throughout the granule matrix. In some but not all of these granules, the small spherules are surrounded by a light halo. None of the granules in T. brasiliensis contains large spherules (Tandler, unpublished observations). Occasional granules in T. brasiliensis have an ornate substructure (Phillips et al., 1987). Generally speaking, serous granules such as those in the acinar cells of the principal parotid gland of T. thersites are thought to be enzyme-rich, whereas seromucous granules in frugiverous bats are considered to be low in enzyme content (Phillips et al., 1993). It seems likely that lysozyme is a component of the granules in T. thersites, because immunohistochemical studies have revealed lysozymelike reactivity in the parotid gland acinar cell serous granules in the related bat, T. brasiliensis (Phillips et al., 1998) Intercalated ducts in the principal parotid gland of T. thersites are noteworthy because they are different from those in T. brasiliensis. In thersites the duct cells typically contain numerous electron-dense serous-type secretory granules, whereas the homologous cells in brasiliensis generally lack secretory granules (Tandler et al., 1998). This particular interspecific difference is interesting because lysozyme-like immunoreactivity is associated with parotid gland intercalated duct secretory granules in some species of bats (Phillips et al., 1998). From comparative studies, it appears that insectivorous bats that feed on hard-bodied, chitinous insects such as beetles are far more likely to have secretory intercalated ducts than are bats that feed primarily on moths or other soft-bodied insects (Tandler et al., 1998). Moreover, there appears to be a correlation (observational rather than statistical) between diet and the number and variety of parotid and submandibular gland cells that might produce lysozyme (Phillips et al., 1998), Based on such data, we hypothesized that 120 NAGATO ET AL. Fig. 8. The infranuclear portion of striated duct cells. Although these cells clearly are involved in elaboration of an organic secretory product, they possess the classical lineaments of striated ducts—basal compartments that are delineated by highly folded plasma membranes and that contain rod-shaped, vertically oriented mitochondria. 315,200. lysozyme might serve as a salivary chitinase in some species (Phillips et al., 1998). Although both T. thersites and T. brasiliensis eat beetles, large coleopterans compose the bulk of the diet for thersites and less than half of the diet for brasiliensis (Freeman, 1979). Although the secretory endpieces of the principal parotid gland in T. thersites are typical in structure, their striated ducts are modified. The tall cells are informed by numerous secretory granules in their apical cytoplasm. In keeping with their secretory function, many of the duct cells have much more RER than is the usual case in these gland components. Although striated duct cells in many species may contain a modicum of secretory granules, these usually are quite small (Tandler et al., 1990; Tandler, 1993). The only other chiropteran salivary gland with numerous, fairly sizeable secretory granules in the striated ducts is the parotid gland of the vampire bat (Tandler et al., 1997). It may not be fair to call the ducts in the vampire parotid ‘‘striated’’ because they have lost all traces of basal striations; however, by virtue of their position relative to intercalated and excretory ducts, they certainly are homologous to striated ducts. These granular ducts in the vampire bat clearly are wholly given over to secretion. A similar situation exists in both the parotid and submandibular glands of the slow loris, a primate (Tandler et al., 1996). In these glands, like in the vampire bat parotid, ducts homologous to striated ducts have lost their basal striations and elaborate one or several giant (up to 11 µm in diameter) secretory granules per constituent cell. Because the ducts in T. thersites still possess basal striations, however reduced, it is highly likely that they still function in ion transport despite their obvious commitment to elaboration of organic products. In this respect, they resemble the granular convoluted tubules of rodent submandibular glands (Tamarin and Sreebny, 1965). In rodents, granular convoluted tubules lie between the intercalated and striated ducts and are marked by the presence of abundant serous type secretory granules in the apical cytoplasm of their constituent cells. In some rodents, such as the laboratory mouse, these ducts are sexually dimorphic. The structure and manifold functions of granular convoluted tubules have been comprehensively reviewed by Gresik (1994). Developmentally (and presumably evolutionarily), these ducts are direct lineal descendants of striated ducts (Gresik, 1994). As such, they retain a reduced level of basal striations. Whole-cell patch clamp studies of granular convoluted tubules from the submandibular gland of the male mouse have shown that these ducts engage in Na1 conductance (Dinudom et al., 1993a); this conductance is sensitive to amiloride (Dinudom et al., 1993b). These data show that granular convoluted tubules still retain the ability to transport electrolytes. In contrast to rodent granular convoluted tubules, the striated ducts of the submandibular gland of the European hedgehog have a normal array of basal striations while producing numerous secretory granules (Tandler and MacCallum, 1974). These granules must have an unusual composition because they are better preserved by fixation solely with osmium tetroxide than they are by initial fixation in aldehyde mixtures followed by osmium. This finding is contrary to all other secretory granules in any salivary gland in any species. The luminal surface of the granule-bearing duct cells in T. thersites is quite irregular, appearing to have multiple indentations. The only other species of bat in which we have observed a comparable configuration of duct cell apical surfaces is the mastiff bat, Eumops glaucinus, which is a member of the same family (Molossidae) (Tandler, unpublished observations). In this respect, these surfaces resemble that of granular convoluted tubule cells in the male mouse, which has been described as having intracellular canaliculi with a modest number of microvilli (Caramia, 1966). These canaliculi, which can penetrate quite deeply into the granular convoluted tubule cells, become much more evident after castration and show images that strongly suggest that granule exocytosis takes place at their lining membrane. Such canaliculi are absent from granular convoluted tubule cells in the female mouse (Caramia, 1966). In contrast to the mouse, surface indentations of duct cells were present in the female free-tailed bat. LITERATURE CITED Caramia F. Ultrastructure of mouse submaxillary gland. II. Effect of castration in the male. J. Ultrastruct. Res. 1966;16:524–536. Dinudom A, Young JA, Cook DI. Na1 and Cl- conductances are controlled PRINCIPAL PAROTID GLAND IN TADARIDA 121 Fig. 9. The erose apical surface of several striated duct cells. Inpouchings of the surface are cut crosswise, so that this surface appears to be underlaid by vacuoles. Note that these putative vacuoles have microvilli. 39,200. by cytosolic Cl- concentration in the intralobular duct cells of mouse mandibular glands. J. Membrane Biol. 1993a;135:289–295. Dinudom A, Young JA, Cook DI. Amiloride-sensitive Na1 current in the granular duct cells of mouse mandibular glands. Pflügers Arch. 1993b;423:164–166. Domon M, Shiga Y. Sialographic investigations on the anatomy of mouse parotid glands. Anat. Rec. 1986;214:165–167. Freeman PW. Specialized insectivory: Beetle-eating and moth-eating molossid bats. J. Mammal 1979;60:467–479. Greeik EW. The granular convoluted tubule (GCT) of rodent submandibular glands. Microsc. Res. Tech. 1994;27:1–24. Huntington GS. The microscopic anatomy of the salivary glands in the lower primates. In: Huntington GS, von W. Schulte H, eds. Studies in Cancer and Allied Subjects. Vol. 4. New York: Columbia University Press, 1913;73–113. Karnovsky MJ. A formaldehyde-glutaraldehyde fixative of high osmolality for use in electron microscopy. J. Cell Biol. 196;27:137–138. Nagato T, Tandler B. Ultrastructure of dog parotid gland. J. Submicrosc. Cytol. 1986;18:249–260. Phillips CJ, Nagato T, Tandler B. Comparative ultrastructure and evolutionary patterns of acinar secretory product of parotid salivary glands in Neotropical bats. In: Patterson BD, Timm RM, eds. Studies in Neotropical Mammalogy: Essays in Honor of Philip Hershkovitz. Chicago: Field Museum of Natural History, 1987: 213–229. Phillips CJ, Weiss A, Tandler B. Plasticity and patterns of evolution in mammalian salivary glands: Comparative immunohistochemistry of lysozyme in bats. Eur. J. Morphol. 1998;in press. Phillips J, Tandler B, Nagato T. Evolutionary diversity of salivary gland scinar cells: a format for understanding molecular evolution. In: Dobrosielski-Vergona K, ed. Biology of the Salivary Glands. Boca Raton: CRC Press, 1993:41–80. Redman RS. The anterior buccal gland of the rat: A mucous salivary gland which develops as a branch of Stensen’s duct. Anat. Rec. 1972;172:167–178. Robin H. Recherches anatomiques les mammiferes de l’Order des chiroptéres. Ann. Sci. Nat. Zool. 1881;12:1–180. Rodriguez-Veiga E, Zúñiga J. Compared anatomy of the parotid gland in the small ruminants (sheep and goat). Proceedings of the XII Anatomy Congress, London. 1985;abstract 604. Ruby JR, Canning HB. Ultrastructure of the acinar cells in the submandibular gland of the nine-banded armadillo. J. Morphol. 1978;155:1–18. Tamarin A, Sreebay LM. The rat submaxillary gland. A correlative study by light and electron microscopy. J. Morphol. 1965;117: 295–352. Tandler B, Phillips CJ, Nagato T. Histological convergent evolution of the accessory submandibular gland in four species of frog-eating bats. Eur. J. Morphol. 1996a;34:163–168. Tandler B. Improved uranyl acetate staining for electron microscopy. J. Electron Microsc. Tech. 1990;16:81–82. Tandler B. Structure of the duct system in mammalian salivary glands. Microsc. Res. Tech. 1993;26:57–74. Tandler B, MacCallum DK. Ultrastructure and histochemistry of the submandibular gland of the European hedgehog, Erinaceous europaeus L. II. Intercalated ducts and granular striated ducts. J. Anat. 1974;117:117–131. Tandler B, Phillips CJ. Structure of serous cells in salivary glands. Microsc. Res. Tech. 1993;26:32–48. Tandler B, Phillips CJ, Toyoshima K, Nagato T. Comparative studies of the striated ducts of mammalian salivary glands. In: Motta PM, ed. Cells and Tissues: A Three Dimensional Approach by Modern Techniques in Microscopy. New York: Alan R. Liss, 1989: 243–248. Tandler B, Toyoshima K, Phillips CJ. Ultrastructure of the principal and accessory submandibular glands of the common vampire bat. Am. J. Anat. 1990;189:303–315. Tandler B, Pinkstaff CA, Nagato T, Phillips CJ. Giant secretory granules in the ducts of the parotid and submandibular glands of the slow loris. Tissue Cell 1996b:28:321–329. Tandler B, Nagato T, Phillips CJ. Ultrastructure of the binary parotid glands in the free-tailed bat, Tadarida thersites. II. Accessory parotid gland. This volume. Tandler B, Nagato T, Toyoshima K, Phillips CJ. Ultrastructure of the parotid gland in the common vampire bat, Desmodus rotundus, with special emphasis on oncocytes. J. Submicrosc. Cytol. Pathol. 1997;29: 37–49. Tandler B, Nagato T, Toyoshima K, Phillips CJ. Comparative ultrastructure of intercalated ducts in mammalian salivary glands: A review. Anat. Rec. 1998;in press. Toh H, Kodama J, Fukuda J, Rittman B, Mackenzie I. Incidence and histology of human accessory parotid glands. Anat. Rec. 1993;236: 586–590. Venable JH, Coggeshall R. A simplified lead citrate stain for use in electron microscopy. J. Cell Biol. 1965;25:407–408. Young JA, van Lennep E. The Morphology of Salivary Glands. London: Academic Press, 1978.