Histochemical localization of adenylate cyclase activity in some mammalian taste papillae

Transcription

1 Chemical Senses Volume 7 Number Histochemical localization of adenylate cyclase activity in some mammalian taste papillae Hiromichi Nomura and Naokazu Asanuma Department of Oral Physiology, Matsumoto Dental College, Shiojiri, , Japan (Received May 1981; revised February 1982; accepted June 1982) Abstract. The localization of adenylate cyclase activity in the fungiform, foliate and circumvallate papillae of rats, rabbits, cats and dogs was determined histochemically using an incubation medium with a high ph. Light-microscopic study showed that adenylate cyclase activity is localized not only at the apex of taste buds but also in other tissues, such as the von Ebner's glands and the blood vessels or capillaries. The adenylate cyclase activity at the apex of taste buds was detectable in all the taste papillae of rats, rabbits, cats and dogs except for the fungiform papillae of rabbits, though the amount of reaction product varied in different papillae. Electron-microscopic study showed that the number and density, as well as the size, of small round-shaped electron-dense granules caused by the precipitation of lead with imidodiphosphate at the apex of taste buds are low in the circumvallate papillae of cats compared with those in the foliate papillae of rabbits. This may explain the result that the amount of reaction product varied in different papillae. Introduction An increasing accumulation of evidence suggests that cyclic nucleotides play an important role in the gustatory organs. There are biochemical studies showing that homogenates of bovine taste papillae exhibit high adenylate cyclase and phosphodiesterase activities (Kurihara and Koyama, 1972; Kurihara, 1972; Price, 1973). Cagan (1976) showed that the labelling of cyclic AMP of bovine taste papillae prelabelled with [8-14 C]adenine was increased slightly in the presence of the taste stimulus sucrose, though the stimulatory effects were not statistically significant. Cyclic nucleotides perfused to the taste organs via the ligual artery were shown to bring about an enhancement of the gustatory responses in rats and frogs (Kasahara and Shimotahira, 1978; Kurihara et al., 1980). In the previous histochemical study, one of the authors (Nomura, 1978) found that adenylate cyclase and cyclic AMP phosphodiesterase activities are localized at the apex of the taste buds in the foliate papillae of rabbits. A similar observation on the localization of adenylate cyclase activity has been carried out in the circumvallate papillae of hamsters (Yamamoto and Ozawa, 1977). There is, however, no observation on the localization of adenylate cyclase activity in other mammals. In our tentative study, no adenylate cyclase activity was detected in the taste papillae of cats and dogs when the methods of Howell and Whitfield (1972) was used. Maguire and Gilman (1974) showed that adenylate cyclase in rat cerebral cortex exhibits a broad optimum ph with a peak below ph 8.0 when ATP is the substrate, but an optimum at ph is obtained with adenylylimidodiphosphate (AMP-PNP). Kempen etal. (1978) also showed that the optimum ph of rat IRL Press Limited, Oxford, England X/82/ S2.00/0 71

2 H. Nomura and N. Asanuma pancreatic adenylate cyclase is 8.9 when AMP-PNP is the substrate. Thus, by the use of an incubation medium with a high ph, we re-examined adenylate cyclase activities in different taste papillae in cats, dogs, rats as well as in rabbits. Materials and Methods Kitten, pups, rats and rabbits were used. Parts of the tongue containing fungiform, foliate or circumvallate papillae were removed from the animals which had been sacrificed by i.v. injection of amobarbital sodium (~ 100 mg/kg body weight). After rinsing with physiological saline, some of the tissues were fixed with 2% glutaraldehyde in 0.05 M cacodylate-nitrate buffer (ph 7.4) containing 0.25 M sucrose for 30 min at 5 C and then washed overnight with the same buffer at 5 C. After embedding in warm-over gelatin (10%), the tissues were frozen and cut into sections, 6 ^m thick, with a cryostat. The other tissues were embedded in O.C.T. compound (Lab Tak Products, Naperville, Illinois, USA) without fixation, frozen in cooled acetone (-60 C) and then cut into sections. In the light-microscopic study, the localization of adenylate cyclase activity was determined by the use of a newly devised incubation medium, the composition being as follows; 70 mm Tris-HCl (ph 8.9), 2 mm MgSO 4, 10 mm NaF, 2 mm AMP-PNP, 0.2 M sucrose, 0.5 mm l-/7-bromotetramisole, 2 mm theophylline and 1.8 mm lead citrate. The tissue sections were incubated for min at 30 C. After washing with distilled water three times, the tissue sections were developed in dilute yellow ammonium sulfide. Control incubations were carried out utilizing the same incubation media and time without the substrate, lead citrate or one of other chemicals in the medium. Another incubation medium was devised according to the observation by Kempen et al. (1978), who showed that barium ion permits a high adenylate cyclase activity and gives a high yield of precipitation with imidodiphosphate (PNPi) without causing non-enzymatic hydrolysis of AMP-PNP. The composition of the incubation medium was as follows; 80 mm Tris-HCl (ph 8.9), 5 mm MgCl 2, 1 mm AMP-PNP, 200 mm sucrose or urea, 0.5 mm l-/?-bromotetramisole, 10 mm theophylline and 4 mm BaCl 2. The tissue sections were usually incubated for 2 3 h at room temperature. After the incubation, the tissue sections were washed briefly with a Tris-HCl buffer (ph 8.2) and then immersed in 0.1 M AgNO 3 for several hours under day-light. In the electron-microscopic study, the localization of the adenylate cyclase activity was determined by the procedure of Howell and Whitfield (1972). The tissue sections were preincubated in a medium containing apyrase (20 mg/ml) for 30 min at 30 C to remove endogenous ATP, and then incubated in an incubation medium for min at 30 C. The composition of the incubation medium was as follows; 80 mm Tris-malate (ph 7.4), 450 mm glucose, 4 mm MgSO 4, 4 mm Pb(NC>3) 2, 0.5 mm AMP-PNP, 2 mm theophylline and 1 mm 1-p-bromotetramisole. After incubation, the tissue sections were washed briefly in 0.1 M cacodylate-nitrate buffer (ph 7.4) containing 250 mm glucose, cut into small blocks with a razor blade, postfixed in 1% osmium tetraoxide dissolved in the same buffer used for washing, dehydrated in ethanol and embedded in Epon 812. The tissues were then cut by an LKB 4800 microtome, stained with uranyl acetate 72

3 Localization of adenylate cyclase activity in taste papillae and lead acetate, and examined in a JEM electron-microscope. Results Comparison of adenylate cyclase activities at ph 7.4 and 8.9 When adenylate cyclase activity in the foliate papillae of rabbits was determined by the use of incubation medium as used in the previous study (Nomura, 1978), the incubation time needed for forming an adequate amount of reaction product was 1.5 h at room temperature (23 25 C). By the use of the incubation medium with a high ph described above, the incubation time was substantially reduced to min at 30 C, which is only 1/4-1/3 of the incubation time needed in the previous study. The previous study (Nomura, 1978) showed that the reaction product for adenylate cyclase activity was formed in the lamina proprio, the connective tissue core of the papillae, the taste bud itself as well as at the apex of taste buds in unfixed papillae, but was merely formed at the apex of taste buds in sufficiently fixed papillae. In the present study, the reaction product was formed in the connective tissue core of the papillae and the von Ebner's glands as well as at the apex of taste buds even in sufficiently fixed papillae. This may also be due to an increase in the sensitivity of histochemical reactions by the incubation medium with a high ph. However, the reaction product was not detectable in the lamina proprio and the taste bud itself even in unfixed papillae. This may be due to an artifact caused by the contamination of some nucleotides in the chemical used, as discussed in the previous paper (Asanuma and Nomura, 1982). Adenylate cyclase activities in the fungiform, foliate and circumvallate papillae of rats, rabbits, cats and dogs When adenylate cyclase activities were determined at the apex of taste buds in the fungiform, foliate and circumvallate papillae of rats, rabbits, cats and dogs by the incubation medium as used in the previous study, a high activity was exhibited only in the foliate papillae of rabbits. A low activity was exhibited in the foliate papillae of rats and the circumvallate papillae of rats and rabbits, but no activity was exhibited in all other papillae. In contrast, by the use of the incubation medium at ph 8.9, pronounced activities were exhibited in all the papillae except for the fungiform papillae of rabbits (Figures 1 3). These results suggest that adenylate cyclase activity is generally present at the apex of taste buds in mammals, though the amount of the reaction products varies in different papillae. Based on the incubation time needed and the degree of deposition of the reaction products, the order of the adenylate cyclase activity at the apex of taste buds in different papillae could be determined as follows: Foliate papilla of rabbit > foliate papilla of rat and circumvallate papillae of rat and rabbit > fungiform papilla of rat > foliate papilla of dog and circumvallate papillae of cat and dog > fungiform papillae of cat and dog > fungiform papilla of rabbit. As seen in Figures 1A, IB, 2A and 3A, reaction products for adenylate cyclase activity were also formed in the subepithelial layer of the papillae as 73

4 H. Nomura and N. Asanuma s B D V - / Fig. 1. Localization of adenylate cyclase activity in different papillae. Activity occurs not only at the apices of taste buds but also in the subepithelial layers in rat and rabbit. Fig. 1 is fungiform papillae, Fig. 2 foliate papillae, and Fig. 3 circumvallate papillae. A: rat, B: rabbit, C: dog, and D: cat. Scale bars represent 100 jim in all the photographs except for Fig. 1A, 3A and 3C, in which the bars represent 50 /im. Incubation time is 45 min in Fig. 2B, 1 h in Fig. 1A and 2A, 2.5 h in Fig. IB and 1C, and 1.5 h in other figures. sporadically scattered rods or lines in rats and rabbits. These products were probably caused by the adenylate cyclase in the endothelial cells of blood vessels or capillaries. Effects of some sodium and chloride salts on adenylate cyclase activity The previous study (Nomura, 1978) showed that an addition of 0.4 M NaCl to the incubation medium completely depressed the formation of reaction products for adenylate cyclase activity in spite of the absence of influence on the formation of reaction products for ATPase and cyclic AMP phosphodiesterase activities. In order to decide whether or not this depressant effect of NaCl is due to a real effect of salty substances on the adenylate cyclase activity that is associated with the sensory transduction process, the effects of some sodium and chloride salts on the adenylate cyclase activity at the apex of taste buds were examined. When 0.4 M NaCl, KC1, Na benzoate, choline chloride, Na acetate or urea was added to the incubation medium, the formation of reaction product for the adenylate cyclase activity was completely depressed by NaCl, KC1 and Na benzoate, and largely by choline chloride and Na acetate, but not by urea. Since 0.4 M Na benzoate tastes less salty, the depressant effect of salts on the adenylate cyclase activity does not seem to be related to the reception of salty stimulus. The 74

5 Localization of adenylate cyclase activity in taste papillae % c / B Fig. 2. Localization of adenylate cyclase activity in different papillae. Activity occurs not only at the apices of taste buds but also in the subepithelial layers in rat and rabbit. Fig. 1 is fungiform papillae, Fig. 2 foliate papillae, and Fig. 3 circumvallate papillae. A: rat, B: rabbit, C: dog, and D: cat. Scale bars represent 100 jim in all the photographs except for Fig. 1A, 3A and 3C, in which the bars represent 50 /im. Incubation time is 45 min in Fig. 2B, 1 h in Fig. 1A and 2A, 2.5 h in Fig. IB and 1C, and 1,5 h in other figures. depressant effect of salts on adenylate cyclase activity was also observed in the von Ebner's glands. This result also gives evidence against the view that adenylate cyclase is related to the reception of salty substances in mammalian taste receptors. Experiments on the validity of the histochemical methods for adenylate cyclase activity The validity of the histochemical methods for adenylate cyclase activity using lead salts has been questioned by some investigators (LeMay and Jarett, 1975; Kempen etal., 1978). Thus, the following experiments were carried out in the present study. Kempen etal. (1978) showed that 4 mm lead nitrate increases cyclic AMP formation on heat-treated pancreatic tissue fragments in rats. In the present study, however, no precipitation of lead was detectable at the apex of taste buds in the foliate papillae of rabbits after the tissue sections were treated with heat (100 C, 2 min) or with a protein precipitant (10% trichloroacetic acid, 5 min). This indicates that the precipitation of lead salts at the apex of taste buds is caused by an enzymatic reaction. Kempen et al. (1978) showed that barium ion permits a high adenylate cyclase 75

6 H. Nomura and N. Asanuma i Fig. 3. Localization of adenylate cyclase activity in different papillae. Activity occurs not only at the apices of taste buds but also in the subepithelial layers in rat and rabbit. Fig. 1 is fungiform papillae, Fig. 2 foliate papillae, and Fig. 3 circumvallate papillae. A: rat, B: rabbit, C: dog, and D: cat. Scale bars represent 100 /ira in all the photographs except for Fig. 1A, 3A and 3C, in which the bars represent 50 (im. Incubation time is 45 min in Fig. 2B, 1 h in Fig. 1A and 2A, 2.5 h in Fig. IB and 1C, and 1.5 h in other figures. activity and gives a high yield of precipitation of PNPi without causing nonenzymatic hydrolysis of AMP-PNP. Thus, we examined adenylate cyclase activity by the use of an incubation medium, to which 4 ram BaCl 2 was added, and from which lead and NaF were omitted. By the use of this incubation medium, we were able to detect adenylate cyclase activity at the apex of taste buds in the foliate papillae of rabbits, though the precipitation was not remarkable (Figure 4). This result indicates that the deposits of lead salts at the apex of taste buds obtained in the previous and present studies are indicative of a real enzymatic reaction of adenylate cyclase activity. Comparison of adenylate cyclase activity between different taste papillae As described above, the amount of reaction products for adenylate cyclase activity at the apex of taste buds varied with the papillae. This may be due to either the difference of the number of density of the adenylate cyclase molecules or the difference of the enzymatic activity of each adenylate cyclase molecule between the papillae. In order to solve this problem, the apices of taste buds of the foliate papillae of rabbits, where the highest activity was exhibited, and of the circumvallate papillae of cats, where a low activity was exhibited, were compared by electron microscopic histochemistry. The small round-shaped electro-dense 76

7 Localization of adenylate cyclase activity in taste papillae Fig. 4. Localization of adenylate cyclase activity in the foliate papilla of rabbit. Activity was determined by a newly devised incubation medium, to which 4 mm BaCl 2 was added and from which lead salt and NaF were omitted. The bar represents 100 /an. granules caused by the precipitation of lead with PNPi appear to be localized similarly on the slender microvilli of the type I taste bud cells in both the papillae, but the number and density as well as the size of the granules are different in both the papillae. Therefore, the variation of the amount of reaction product for adenylate cyclase activity in different papillae may arise from the number and density of adenylate cyclase molecules and also from the enzymatic activity of each molecule. Discussion There is accumulation of evidence suggesting that cyclic nucleotides may play important roles in the gustatory organs as set out in the Introduction. The previous studies (Yamanoto and Ozawa, 1977; Asanuma and Nomura, 1982) showed that an adenylate cyclase activity is localized on the microvillous cell membrane of the type I taste bud cells, and the present study showed that the adenylate cyclase activity at the apex of taste buds is generally observed in all the mammals examined. These facts indicate that the high adenylate cyclase activity in homogenates of bovine taste papillae found in a biochemical study (Kurihara and Koyama, 1972) is attributed to the enzyme localized in the microvilli of the type I taste bud cells, though some of the activity may be due to the enzyme activity in von Ebner's glands and blood vessels or capillaries. It is well known that hormone-sensitive adenylate cyclase activities are affected 77

8 Fig. S. Localization of adenylate cyclase activity at the apices of taste buds in the circumvallate papilla of cat (A) and in the foliate papilla of rabbit (B). Activity occurs on the microvilli of type I taste bud cells as small round-shaped electron-dense granules similarly in both the papillae. Note that microvilli are poor in the circumvallate papilla of cat. 1 2 o at Pennsylvania State University on May 10, O.

9 Localization of adenylate cyclase activity in taste papillae by the hormones concerned. Reik et al. (1970) showed that some of the hormonesensitive adenylate cyclase activity in rat liver survives glutaraldehyde fixation and can be stimulated by the hormones, isoproterenol and glucagon. Accordingly, if the adenylate cyclase on the microvillous cell membrane of the type I taste bud cells is associated with the initial process of sensory transduction in mammalian gustatory cells, the enzyme activity should be affected by taste stimulus. In the previous histochemical study, one of the authors (Nomura, 1978) examined the effects of taste stimulus on the ATPase, the adenylate cyclase and the cyclic AMP phosphodiesterase activities at the apex of taste buds in the foliate papillae of rabbits and found that 0.4 M NaCl added to the incubation medium completely depressed the formation of reaction product of the adenylate cyclase activity in spite of the absence of influence on the ATPase and the cyclic AMP phosphodiesterase activities. Gustatory responses in rats, rabbits and cats were compared in the chorda tympani nerve by Pfaffmann (1955) and in the glossopharyngeal nerve by Yamada (1965). According to them, in the chorda tympani nerve responses NaCl is most effective in rat and least effective in rabbits, while in the glossopharyngeal nerve responses it is most effective in rabbits and least effective in cats. Since fungiform papillae are innervated by the chorda tympani nerve and foliate and circumvallate papillae by the glossopharyngeal nerve, if the adenylate cyclase at the apex of taste buds is related to the reception of NaCl, the enzyme activity must be higher in the fungiform papillae of rats and the foliate and circumvallate papillae of rabbits than in the fungiform papillae of rabbits and the circumvallate papillae of cats. The order of the adenylate cyclase activity at the apex of taste buds determined in the present study agreed with this prediction. This might be indicative of a prominent role of cyclic nucleotides in the reception of NaCl in mammalian taste organs. However, it was shown in the present study that 0.4 M Na benzoate, which was less salty, depressed the adenylate cyclase activity at the apex of taste buds. This indicates that the depression of adenylate cyclase activity by 0.4 M NaCl may be due to an effect of NaCl on a histochemical reaction, such as salting out effect, but not to an influence of NaCl on the enzyme activity. In order to interpret this ambiguous result, the following possibilities may be considered. A possibility is that the type I taste bud cells and, consequently, the adenylate cyclase in the microvilli of this type of cells is not associated with the transduction process of gustatory organs. According to Murray (1971), typical synaptic contacts with nerves exist only in type III taste bud cells. This suggests that the transduction process of mammalian gustatory organs is brought about merely in the type III, but not the type I, taste bud cells. In this case, the adenylate cyclase activity in the microvilli may not be altered by taste stimulation. Another possibility is that the modification of the adenylate cyclase activity by taste stimulus is so small that one cannot detect the change of the enzyme activity caused by taste stimulus with histochemical methods. Cagan (1976) studied the effects of taste stimulus on the formation of cyclic AMP in bovine taste papillae and found that only a slight increase in cyclic AMP formation was brought about by taste stimulus. This result supports this possibility. In order to reveal the role of the adenylate cyclase in the microvilli of the type I 79

10 H. Nomura and N. Asanuma taste bud cells, further studies must be carried out. References Asanuma.N. and Nomura,H.: 1982, 'Histochemical localization of adenylate cyclase and phosphodiesterase activities in the foliate papillae of the rabbit', Chem. Senses, 7. Cagan.R.H.: 1976, 'Biochemical studies of taste sensation: II. Labeling of cyclic AMP of bovine taste papillae in response to sweet and bitter stimuli', J. Neurosci. Res., 2, Howell.S.L. and Whitfield.M.: 1972, 'Cytochemical localization of adenyl cyclase activity in rat islets of Langerhans', J. Histochem. Cytochem., 20, Kasahara.Y. and Shimotahira.K.: 1978, 'Taste reception and cyclic nucleotide', J. Physiol. Soc. Japan, 40, 291. Kern pen.h.j.m., de Pont.J.J.H.H.M., Bonting.S.L. and Stadhouders.A.M.: 1978, 'The cytochemical localization of adenylate cyclase: Fact or artifact?', J. Histochem. Cytochem., 26, Kurihara.K.: 1972, 'Inhibition of cyclic 3',5'-nucleotide phosphodiesterase in bovine taste papillae by bitter taste stimuli', FEBS Lett., 27, Kurihara.K. and Koyama.N.: 1972, 'High activity of adenyl cyclase in olfactory and gustatory organs', Biochem. Biophys. Res. Commun., 48, Kurihara.K., Nagahama.S. and Kashiwayanagi.M.: 1980, 'A molecular mechanism of taste transduction in the frog', in van der Starre.H. (ed.), Ol/action and Taste VII, IRL Press, London & Washington DC, pp LeMay,A. and Jarett.L.: 1975, 'Pitfalls in the use of lead nitrate for the histochemical demonstration of adenylate cyclase activity', J. Cell Bioi, 65, Maguire.M.E. and Gilman.A.G.: 1974, 'Adenylate cyclase assay with adenylyl imidodiphosphate and product detection by competitive protein binding', Biochim. Biophys. Acta, 358, Murray.R.G.: 1971, 'Ultrastructure of taste receptors', in L.M.Beidler (ed.), Handbook of Sensory Physiology IV, Springer-Verlag, Berlin, pp Nomura.H.: 1978, 'Histochemical localization of adenylate cyclase and phosphodiesterase activities in the foliate papillae of rabbits. I. Light microscopic observations', Chem. Senses and Flavour, 3, Pfaffmann.C: 1955, 'Gustatory nerve impulses in rat, cat and rabbit', J. Neurophysiol., 18, Price.S.: 1973, 'Phosphodiesterase in tongue epithelium: activation by bitter taste stimuli', Nature, 241, Reik.L., Petzold.G.L., Haggins,J.A., Greengard.P. and Barrnett.R.J.: 1970, 'Hormone-sensitive adenyl cyclase: Cytochemical localization in rat lever', Science (Wash.), 168, Yamada.K.: 1965, 'The glossopharyngeal nerve response to taste and thermal stimuli in the rat, rabbit and cat', Kumamoto Med. J., 18, Yamamoto.T. and Ozawa.H.: 1977, 'Adenylate cyclase in paraneurons. A histochemical study', Arch. Histol. Jap., 40, Suppl.,