Abstract
We examined the histochemical localization of carbonic anhydrase (CA) in Bowman's glands by light and electron microscopy. Neither CAI nor CAII was detected immunohistochemically in the duct cells. However, by enzyme histochemistry the duct cells revealed electron-dense precipitates demonstrative of CA in the microvilli and intercellular digitations. The reaction product was also noted in small vesicles in the cytoplasm of duct cells. In cells of the acini, the well-developed short microvilli, basolateral cell membrane, and mitochondria along the basolateral membrane showed strong deposits indicating CA activity. Dense reaction product of CA was also detected in a small core within the electron-lucent granules of the secretory cells, although CAI and CAII were not detected by immunostaining in the secretory granules. Although the functional significance of CA in Bowman's glands is obscure, the enzyme may play a role in regulation of pH and ion balance in the mucous layer covering the olfactory epithelium. The presence of CA activity in the ducts suggests that these structures are not simple tubes serving as a conduit for secretory substances but participate in modifying the luminal content by secreting CA.
Keywords
Materials and Methods
Immunohistochemistry
The experimental protocol was approved by the Animal Care and Experimentation Committee of our institutes. Under general urethane anesthesia (sodium pentobarbital, 1.5 g/kg body weight), female albino guinea pigs (approximately 300 g body weight) were perfused with physiological saline followed by perfusion with 0.5% zinc-10% formalin. After perfusion, the rostrum containing the nose was quickly removed and immersed in the same fixative for another 2 hr. The specimens were decalcified with 5% ethylenediamine tetraacetic acid (EDTA) (pH 7.3) for 14 days at 4C. The dehydrated specimens were embedded in paraffin and cut at 20 μm. After dissolving the paraffin, the sections were incubated with 1% H2O2, 5% normal horse serum (NHS), and then incubated overnight at room temperature with one of the primary antibodies. The primary antibody against anti-CAI (anti-human CAI-IgG, Code PC046; Binding Site, Birmingham, UK) was used at a dilution of 1:100, whereas CAII (anti-human CAII-IgG, Code PC076; Binding Site) was used at a 1:3000 dilution with 1% NHS-0.01 M phosphate buffer, pH 7.2, containing 0.9% NaCl (PBS). The sections were rinsed in PBS and flooded for 1 hr with a 1:400 dilution of biotinylated anti-sheep IgG and incubated in Vectastain ABC reagent for 1 hr. Sites of bound primary antibodies were visualized by development in 3,3′-diaminobenzidine-H2O2 substrate medium. We also tested the specificity of the immunohistochemical reaction with red blood cells in the nasal capillary, which are known to possess CAI and CAII.
Enzyme Histochemistry
Anesthetized guinea pigs were perfused with physiological saline followed by perfusion with 4% paraformaldehyde-3% glutaraldehyde. After perfusion, the rostrum containing the nose was quickly removed and immersed in the same fixative for another 2 hr. The specimens were decalcified with 5% EDTA (pH 7.3) for 14 days at 4C and were frozen after treatment with a series of solutions of graded sucrose. The frozen tissues were cut into 20 μm sections and reacted in Hansson's medium (1967) for CA activity. We also tested the specificity of the enzyme histochemical reaction using acetazolamide.
Electron Microscopic Histochemistry
For exact localization of CA activity in Bowman's gland, tissue sections showing distinct reaction products in the acini and ducts by light microscopy were selected for electron microscopy. The selected sections were rinsed in 0.01 M phosphate buffer (pH 9.4) and further fixed with 1% OsO4 buffered with 0.1 M phosphate buffer (pH 5.0) and treated with 1% uranyl acetate in 0.1 M maleate buffer (pH 5.2), based on the method of Sugai and Ito (1980). The dehydrated sections were mounted in Epon and cut into ultrathin sections for the electron microscopy.
Measurement of CA Activity in the Olfactory Mucus
The rostrum of anesthetized guinea pig nose as above mentioned was cut down and the olfactory epithelium was opened. The olfactory mucous layer was washed with distilled water and the washing solution was examined for CA activity (Maren 1960) and hemoglobin concentration (Crosby and Furth 1956).
Results
Bowman's gland of guinea pig consisted of an intraepithelial portion in which the ducts were located and an extraepithelial portion in which mainly branched tubuloalveolar secretory acini were located (Figure 1a).
The immunohistochemical reactions using CAI or CAII antibodies showed different results. CAII was detected in Bowman's glands but CAI was not (Figure 1b and 1c). In particular, the cytoplasm near the basolateral membrane of the acinar cells showed localization of CAII, but the supranuclear region and ducts revealed no immunohistochemical reaction. CAI and CAII were not present in three types of olfactory epithelial cells, consisting of the receptor, sustentacular, and basal cells (Figure 1b and 1c) except for some epithelial cells on the endoturbinate showing CAII (Figure 1d). Both CAI and CAII were detected on red blood cells in the nasal capillaries.
Immunohistochemistry of CAI and CAII antibodies in the olfactory epithelium of guinea pig. (a) Stained with hematoxylin. (b) Immunohistochemically stained with anti-CAI. (c,d) Stained with anti-CAII. (a-c) The roof of the nasal cavity; (d) the region of the endoturbinate. (a) Bowman's gland of guinea pig consists of an intraepithelial portion in which the ducts are located (arrows) and an extraepithelial portion in which mainly branched tubuloalveolar secretory acini are located. (b) Bowman's glands show no CAI activity. (c) CAII is the localized near the basolateral membrane of acinar cells but not in the supranuclear region of these cells or their ducts (arrows). Note also lack of immunoreactivity in the receptor, sustentacular, and basal cells of the olfactory epithelium. (d) An infrequent epithelial cell (arrow) typically located in the endoturbinate shows CAII activity. The acinar cells of Bowman's gland also showed CAII activity. OE, olfactory epithelium; N, nerve bundle; V, blood vessel; BG, Bowman's gland. Bar = 25 μm.
Light microscopic examination of enzyme histochemistry showed the localization of CA in Bowman's glands, passing through the olfactory epithelium and opening into the lumen (Figure 2). No CA was present in the olfactory epithelium, including the receptor, sustentacular, and basal cells, except for some epithelial cells on the endoturbinate, as mentioned above (Okamura et al. 1996). These reactions for CA were completely inhibited when the tissue sections were incubated in normal Hansson's medium containing 10−5 M acetazolamide.
The electron microscopic examination defined the fine structures and showed definite electron-dense precipitates as reaction products for CA. The duct was composed of a simple cuboidal epithelium with electron-lucent and amorphous cytosol and with well-developed microvilli on the luminal surface (Figure 3). The cytoplasmic processes on the lateral cell membrane of duct cells showed several interdigitations. Flocculent, fine electron-dense reaction precipitates were detected as deposits covering the luminal surface of the microvilli, and reaction products were localized between apposed lateral cell membranes (Figure 3 and 4). CA was also detected in small vesicles scattered throughout the cytoplasm (Figure 4).
In the acini, well-developed short microvilli protruded through the secretory cell surface into the gland lumen, ovoid and flattened nuclei were located towards the basal side of secretory cells, and quite numerous rod-shaped mitochondria were present along the basolateral cell membrane. The apposed lateral cell membrane with very well-developed cytoplasmic processes showed intercellular digitations. The basal cell membrane showed well-developed basal invaginations. Dense reaction product of CA was detected in the cytoplasm of the basolateral cell membranes, showing interdigitations, and was found in mitochondria along the basolateral membrane (Figure 5 and 6). Reaction precipitate was also detected on the luminal surface of microvilli. The cytoplasm of acinar cell was filled with abundant, densely distributed secretory granules (Figure 5). These granules were classified into two groups according to the texture and electron density of the secretory granule matrix. The first group consisted of electron-dense granules with compact matrix, and the second group consisted of electron-lucent granules with loose matrix. The reaction product was detected as electron-dense particles scattered on the loose matrix of the electron-lucent secretory granules (Figure 7).
Light microscopic features of olfactory epithelium at the region of the roof of a guinea pig nasal cavity, enzyme histochemically reacted for CA. Note the acini and duct (arrows) passing through the olfactory epithelium, shown as black precipitates representing reaction products for CA. Note also the lack of such precipitates in three types of ordinary olfactory epithelial cells, i.e., olfactory receptor cells, sustentacular, and basal cells, except for some epithelial cells showing CA activity (not seen in this region). Tissue sections showing distinctive reaction products in Bowman's glands were selected for electron microscopic examination. The reactions for CA were completely inhibited when the tissue sections were incubated in normal Hansson's medium containing 10−5 M acetazolamide. OE, olfactory epithelium; BG, Bowman's gland. Bar = 25 μm.
Electron microscopic examination of the duct cells of Bowman's gland. The duct of Bowman's gland is composed of simple cuboidal epithelium. The components consist of electron-lucent and amorphous cytosol, and the microvilli are well-developed on the luminal surface. Note the electron-dense precipitates for CA activity in these microvilli and the cytoplasmic processes on the lateral cell membrane. L, lumen; N, nucleus; MV, microvilli; LM, lateral cell membrane. Bar = 2 μm.
High magnification of microvilli and lateral cell membrane of the duct cell of Bowman's gland. Note the well-developed microvilli on the luminal surface and the presence of electron-dense precipitates as reaction products for CA activity in these villi. The cytoplasmic processes on the lateral cell membrane of the ductal epithelial cells, showing intracellular interdigitations, also indicate CA activity (arrowheads). Note also CA activity in intracellularly-located small vesicles (arrows). MV, microvilli. Bar = 4 μm.
Electron microscopic examination of an acinar cell of Bowman's gland. Several well-developed and short microvilli with strong CA activity have migrated from the secretory cells into the glandular lumen. Note also that the round, flat nuclei of acinar cells are close to the basal side of the cell, and many rod-shaped mitochondria are distributed across the basolateral cell membrane. Dense reaction product of CA was detected in the cytoplasm of the basolateral cell membranes, showing interdigitations. The cytoplasm of the acinar cell is filled with densely distributed secretory granules. L, lumen; N, nucleus; MV, microvilli; Mt, mitochondria. Bar = 1 μm.
High magnification of the basolateral cell surfaces of an acinar cell. Note the strong CA activity in the apposed basolateral cell surfaces, with well-developed cytoplasmic processes that show intercellular interdigitations, and in the mitochondria along the basolateral membrane. Mt, mitochondria; BLM, basolateral cell membrane. Bar = 4 μ m.
CA activity and hemoglobin concentration from the solution washed out of the olfactory mucous layer showed 3.91 U/ml and 0.003 mg/ml, respectively. Hemoglobin concentration was very low and near zero, demonstrating that this solution was not affected by the blood serum. These results showed that the olfactory mucous layer included CA activity.
Discussion
Only a few studies have provided a detailed analysis of the histological features of Bowman's gland (Frisch 1967; Andres 1969; Yamamoto 1976). We recently described the presence of CA activity in Bowman's gland and indicated that the duct was not merely a conduit for secretions from the acinar cells (Okamura et al. 1996). The present analysis offers light and electron microscopic extension of the above study and demonstrates the exact histochemical localization of CA in Bowman's gland.
CA was detected in the cytoplasm, basal cell processes and lateral cell membrane of the acinar epithelial cells and was found in mitochondria along the basolateral membrane. The CA reaction produced in the cytoplasm can be attributed to CAII because mitochondrial CA is known to be CAV, and intense immunohistochemical reactivity for cytoplasmic CAII was detected in the cytoplasm near the basolateral membrane of acinar cell. Although the significance of CA enzyme in the cytoplasm is not clear, it is possible that the enzyme plays a role in ion transport across the basal cell membrane by supplying bicarbonate or hydrogen ions for exchange. In this regard, the cytoplasm of the acinar cells contains many mitochondria adjacent to the basolateral cell membrane. The mitochondria supply CO2 and H2O for generating H+ and HCO3 −, which can be exchanged for K+ and Cl−, respectively, as in gastric parietal cells (Spicer et al. 1984). Unlike the acini, CA activity in the duct was detected on apposed lateral cell membranes but not in the cytoplasm. The cytoplasm of duct cells consisted of amorphous cytosol and very few mitochondria. This finding suggests that the duct epithelial cells engage in different ion transport compared with secretory acini. CA on the lateral cell membrane of ducts may be CAIV because it is known that membrane-bound CA is Type IV (Whitney and Briggle 1982).
Two types of secretory granules have been detected in acinar cells, including secretory granules with an electron-dense matrix and others with an electron-lucent matrix. Yamamoto (1976) showed that the secretory granules undergo maturational changes by demonstrating that the dense granules were younger or immature granules, whereas the less dense granules were mature secretory granules. In this case, the mature secretory granules show the localization of CA. CA was detected as electron-dense punctutes scattered in electron-lucent secretory granules. This CA apparently is secreted into the lumen through a merocrine secretory mechanism. CA has been observed in secretory granules of serous acinar cells in submandibular (Spicer et al. 1982), tracheobronchial (Spicer et al. 1982), parotid, and exorbital lacrimal glands (Hennigar et al. 1983), and presumably is secreted by these cells. Parkkila et al. (1990) have suggested that both CAII and CAVI are localized in the secretory granules in human parotid and submandibular glands but that only the CAVI is secreted. We detected no CAII in secretory granules by immunostaining, although the granules showed CA activity with enzyme histochemistry. The CA in the secretory granules of acini may therefore be CAVI.
High magnification of the secretory granules present in an acinar cell. The secretory granules that fill the cytoplasm of these cells are of two types, based on the matrix texture: electron-dense granules with compact matrix (∗) and electron-lucent granules with loose matrix. CA activity is detected as electron-dense dots scattered among the electron-lucent secretory granules (arrows). Bar = 0.2 μ m.
Results of several studies have suggested that CA is involved in a number of biological functions, such as pH regulation, ion balance, and stabilization of macromolecules in secretions in the oral cavity, airway ducts, and conjunctiva (Spicer et al. 1984). However, the role of CA enzyme in Bowman's gland is still an enigma. The mucus covering the surface of the olfactory epithelium with the ciliary surface of olfactory receptor cells serves to keep the olfactory mucosa moist and furnishes the necessary solvents for primary contact with odor stimuli (Bloom and Fawcett 1975). Frings et al. (1991) reported that the concentration of Na+, K+, Ca2+ in the mucous layer over the olfactory epithelium affected the sensitivity of odor detection. It is known that this mucus is secreted from both Bowman's gland and the olfactory epithelial cells, although the majority is derived from Bowman's gland (Andres 1966; Frisch 1967; Getchell et al. 1984). We detected CA activity in the solution obtained from washing the olfactory mucous layer. Because CA was not observed histochemically in three types of ordinary olfactory epithelial cells in tissue sections, the enzyme in the wash solution presumably derived from Bowman's glands. The immunoreactivity seen in sparse sensory cells in the epithelium could not account for the enzyme at the surface of the epithelium. Therefore, we speculate that CA secreted from the Bowman's glands may be involved in the regulation of physical properties of the mucous layer.
Our results also showed accumulation of the enzyme histochemical reaction precipitates of CA on the luminal side of the microvilli at the apical surface of duct cells. These precipitates probably originate from one of two possible sources. First, the surface cell membrane of the duct cells contains the CA enzyme and shows histochemical reactivity on the luminal side of the plasmalemma. Second, enzyme secreted by acinar cells adheres to the microvilli at the surface of the ducts. Our results provide evidence for the latter explanation because stain deposits in duct cells appeared too sparse to account for the heavy deposits on the duct surface. Instead, the small enzyme-positive foci shown in acinar granules could provide the source of the CA at the duct surface. These findings suggest that the mature secretory granules contain CA enzyme, which is secreted into the gland lumen through a regulatory secretion mechanism.
Ion-transporting epithelia that show CA activity are present in several organs, including the pancreas (Kumpulainen and Jalovaara 1981; Spicer et al. 1982; Buanes et al. 1986), liver (Spicer et al. 1982; Carter et al. 1989), salivary glands (Hennigar et al. 1983), and sweat glands (Briggman et al. 1983). Pancreatic ductal CA correlates with and is therefore implicated in mediating the high bicarbonate content of pancreatic fluid. CA in the ducts of sweat glands facilitates Na+ and H2O resorption from the lumen to produce a hypotonic absorbate. Although the precise relation of CA to ion transport in the ducts of Bowman's gland has not been determined, we speculate that CA in ducts may also function in bicarbonate secretion or Na+ absorption, similar to its function in the ducts of other organs. This means that the duct of Bowman's gland is not a simple tube serving as a conduit for substances secreted from acini nor simply a lining cell layer but rather that it plays a special physiological function by regulating secretions from acini.
Using routine neutral or alkaline-buffered OsO4 for postfixation after enzyme histochemistry for CA results in dissolving of the cobalt sulfide reaction product and its loss from the section (Yokota 1969; Sugai and Ito 1980). To avoid such loss, Sugai and Ito (1980) used acid-buffered OsO4 for postfixation. In this study, we used the latter method for electron microscopic examination of the localization of CA in cryostat sections decalcified with EDTA. The use of this technique enabled us to obtain fine details of all cellular structures, including the electron-dense precipitates representing reaction products for CA, without dissolution of reaction products. Therefore, this method is recommended for examination of cryostat sections decalcified with EDTA at the electron microscopic level.
Finally, we found that infrequent nasal epithelial cells possess CAII. We speculated that these cells represent specialized chemoreceptors, e.g., detecting changes of pH or CO2 concentration (Okamura et al. 1996). The known physiological activity of CA and the distribution of these cells in the nose appear consistent and this interpretation. Recently, Coates et al. (1998) reported on electrophysiological responses to pulses of CO2 and considered this to be related to cells shown by enzyme histochemistry to contain CA in frog nasal cavity. CAII may play a role in the hydration of CO2 in these cells.
Footnotes
Acknowledgements
We should like to thank Dr Samuel S. Spicer (MUSC) for kindly reading the manuscript and for his useful suggestions. This manuscript was also prepared to extend our congratulations on the tenth anniversary of Prof Naonori Sugai as Chairman of the Department of Anatomy and Histology.