- Email: [email protected]
Triterpenoid saponins stimulate the sugar taste receptor cell through a G protein-mediated mechanism in the blowfly, Phormia regina
Journal of Insect Physiology 48 (2002) 367–374 www.elsevier.com/locate/jinsphys
Triterpenoid saponins stimulate the sugar taste receptor cell through a G protein-mediated mechanism in the blowfly, Phormia regina Arifa Ahamed a, Seiji Tsurumi b, Taisaku Amakawa a,c,∗ a
Graduate School of Science and Technology, Kobe University, Kobe 657-8501, Japan b Radioisotope Research Center, Kobe University, Kobe 657-8501, Japan c Faculty of Human Development, Kobe University, Kobe 657-8501, Japan Received 29 August 2001; accepted 17 January 2002
Abstract The blowfly has taste chemosensilla on the labellum. The sensory receptor cells in the chemosensillum are highly specialized for the tastes of sugar, salt and water, respectively. Previously we introduced chromosaponin I (CSI) and glycyrrhizin (GL), as sweet substances for the blowfly, Phormia regina. Application of these triterpenoid saponins induced feeding responses as well as impulses of the sugar taste receptor cell in the LL-type sensillum at a much lower concentration than that of sucrose. In the present paper, we show the involvement of G protein-mediated cascade in the CSI- and GL-responses as well as in sugar responses. CSI activates the sugar signal transduction cascade after penetrating through the membrane. On the other hand, GL exerts dual effects to stimulate the sugar signal transduction possibly by activating it inside the cell and also by interacting with the pyranose sugar receptor site. A non hydrolyzable G protein inhibitor guanosine 5⬘-O-(2-thiodiphosphate), GDPβS, markedly decreased the responses of the sugar receptor cell to the two triterpenoid saponins as well as the response to sucrose and fructose. These results suggest that CSI and GL are direct activators of G protein. 2002 Elsevier Science Ltd. All rights reserved. Keywords: Blowfly; G protein; Sugar taste; Transduction; Triterpenoid saponins
1. Introduction The taste organ of the fly is a sensillum showing a much less complicated structure than the vertebrate taste bud. A sensillum contains four sensory processes elongated from four functionally differentiated taste cells which respond with unique impulses to sugars, salts, water and large anions like caprylate, respectively. Thus, it is easy to observe the electrophysiological response from a single taste cell by giving adequate stimulus to the tip of the sensillum. Furthermore, it is also easy to examine the pharmacological effects of reagents on a single taste cell by applying reagents to a sensillum for checking the response. One of the characteristic phenomena of taste reception
Corresponding author. Tel.: +81-78-803-7747; fax: +81-78-8037761. E-mail address: [email protected] (T. Amakawa). ∗
in insects is its very fast response: the latency is as short as within several milliseconds for sugar or salt stimulation. Because of this fast response, it is widely accepted that the primary mechanism of taste reception includes the ionotropic type ion channels rather than the cascade type such as G protein-mediated, which take longer times to respond. In fact, Murakami and Kijima (2000) recently succeeded in recording sugar activated ion channels in the chemosensillum of fleshfly pupae. They further showed using GDPβS that no contribution by G protein was seen. Amakawa et al. (1990, 1992), on the other hand, reported that dibutyryl cyclic GMP (dbcGMP), a membrane permeable analogue of cyclic GMP, elicits the response of the sugar receptor cell of the blowfly with poor adaptation. They suggested that dbcGMP penetrated the cell membrane and opened the cyclic nucleotide gated channel as a mimic of the second messenger. In the fleshfly, Koganezawa and Shimada (1997) showed
0022-1910/02/$ - see front matter 2002 Elsevier Science Ltd. All rights reserved. PII: S 0 0 2 2 - 1 9 1 0 ( 0 2 ) 0 0 0 5 6 - 2
368
A. Ahamed et al. / Journal of Insect Physiology 48 (2002) 367–374
that the proven G protein inhibitor guanosine 5⬘-O-(2thiodiphosphate), GDPβS, markedly depressed the sugar responses of the labellar sugar receptor cell using sucrose, fructose and valine as stimulants. This indicates that the response of the sugar receptor cell is elicited through a G protein-mediated second messenger cascade. Another piece of evidence supporting G protein inclusion in the reception process comes from molecular genetic research: Ueno et al. (2001) recently identified a taste receptor gene, Tre, that controls the taste sensitivity to trehalose in Drosophila melanogaster and hypothesized that TRE belongs to a novel family of G protein-linked transmembrane receptors that may operate as taste receptors. Previously, as a part of our electrophysiological studies on triterpenoid saponins, we have shown that chromosaponin I (CSI), a γ-pyronyl-triterpenoid saponin isolated from pea and other leguminous plants (Tsurumi et al., 1992), stimulates the sugar receptor cells of the blowfly without affecting the sugar receptor sites on the membrane (Ahamed et al., 2000). Another triterpenoid saponin, glycyrrhizin (GL), found in the root of licorice (Glycyrrhizia glabra), is used extensively as a sweetener as well as a medicine for humans. We showed that GL, a structural analogue of CSI, induced full proboscis extension in the blowfly as does CSI and that it exerted more intense signals of the sugar receptor cell than CSI (Ahamed et al., 2001). It is known that the sugar receptor cells of the blowfly, Phormia regina, possess at least two types of receptor sites that are distinguished by selective competitive inhibition with polysaccharides, starch for the pyranose site (the P site) and levan for the furanose site (the F site) (Hara, 1983). The GL-induced stimulation of the sugar receptor cells was separated into two parts: activation via the P site on the membrane and an additional mechanism, possibly direct activation of the sugar signal transduction cascade inside the cells not mediated by the P site (Ahamed et al., 2001). In the present paper, we show that the responses to CSI and GL as well as to sugars are greatly reduced by GDPβS, indicating the involvement of a G proteinmediated second messenger cascade in these responses of sugar taste receptor cells in blowflies. We furthermore studied in detail the effect of these triterpenoid saponins on the taste receptor cell, using the G protein inhibitor, GDPβS.
2. Materials and methods 2.1. Fly The blowflies P. regina were reared in the laboratory at 24±1 °C and fed with 0.1 M sucrose and water. Seven
to nine day old adult blowflies were used for our experiments. 2.2. Electrophysiological procedure The impulses were recorded by the tip recording method of Hodgson et al. (1955) and the responses were recorded from the labellar LL-type sensilla as described before (Ahamed et al., 2001). Except for elongated stimuli, the duration time of stimulus was 30 s. The impulses during the initial period (to 0.15 s after the beginning of stimulation) were ignored because the impulse frequency was not proportional to the receptor potential during the initial period (Morita, 1969, 1972; Ozaki and Amakawa, 1992). To compare the magnitude of the sugar and GL responses, we counted the number of impulses during a period of 0.2 s starting 0.15 s after the beginning of stimulation. In the case of CSI, due to the variable latency, we selected the maximum number of impulse during a period of 0.2 s (Ahamed et al., 2000). The experiments were carried out at 24±1 °C, relative humidity being kept at 70–80%. 2.3. Chemicals CSI was isolated from pea and purified as described previously (Tsurumi et al., 1992). GL (ammonium salt, purity approx. 75%), GDPβS and GTP were purchased from Sigma Chemical Co. (St Louis, MO, USA). Sucrose, fructose, sodium deoxycholate (DOC) and other chemicals were purchased from Wako Pure Chemicals Industries Ltd (Osaka, Japan). 2.4. Test solutions CSI and GL were dissolved in 20 mM MOPS buffer (pH 6.6). Sucrose and fructose solutions contained 10 mM NaCl for electrical conductance. Since the cell membrane is virtually impermeable to the nonhydrolyzable nucleotide, GDPβS, it is necessary to aid the incorporation of the reagent into the cell. Direct microinjection was not possible, since the sensory process of the fly taste receptor cell is too slender. Amakawa and Ozaki (1989) suggested that 0.03% DOC temporarily increases the permeability of the receptor membrane without membrane lysis. The sugar receptor cell preserved its responsiveness after such mild detergent treatment. So DOC was used to introduce GDPβS or GTP into the sugar taste receptor cells. The GDPβS or GTP was dissolved in 0.03% DOC in 67 mM phosphate buffer, pH 7.2, as described before (Amakawa and Ozaki, 1989). After treatment for 2 min with one of these DOC solutions, the electrophysiological responses to sugars or saponins were recorded 5 min after the DOC treatment. Treatment with 0.03% DOC alone was used as the control. To see the effect of GDPβS on the response to CSI
A. Ahamed et al. / Journal of Insect Physiology 48 (2002) 367–374
369
or GL, GDPβS was added to saponin solution because these saponins introduced GDPβS inside the sugar receptor cells without an aid of DOC. The CSI specific antibody was purified in our laboratory (Ahamed et al., in preparation). The polyclonal antibodies raised against CSI were purified with soyasaponin I- and CSI-affinity chromatography. The triterpenoid and sugar moieties of soyasaponin I were the same as those of CSI, but the former does not have the γ-pyronyl moiety. CSI-specific antibody was obtained from the fraction which binds to the CSI-affinity column but does not bind to the soyasaponin I-affinity column. The specificity of the purified antibody was confirmed by a ligand-molecule interaction analysis with BIACORE 3000. The antibody concentration used for the experiments was 200 µg/ml, dissolved in a neutral buffer. The incubation of CSI with the antibody was carried out at 24±1 °C. 3. Results 3.1. Effect of GDPbS on sugar response Figs. 1A and B show the time course of the responses of the sugar receptor in the blowfly to 50 mM sucrose and 100 mM fructose, respectively, after treatment with 0.03% DOC (control) or with 20 mM GDPβS in 0.03% DOC for 2 min. DOC treatment alone appeared to slightly accelerate adaptation (Amakawa and Ozaki, 1989) and GDPβS alone did not induce any impulse (data not shown). The responses of the sugar receptor cell to both sucrose and fructose were apparently reduced after GDPβS treatment, compared with control. However, we did not find complete inhibition by GDPβS treatment. We observed no change in the sugar response when GDPβS was added without DOC, indicating that this nonhydrolyzable G protein inhibitor must be introduced inside the sugar receptor cell to exert its action. No remaining effects of GDPβS were observed 10 min after removing the reagent from the labellar hair. These results were almost the same as those by Koganezawa and Shimada (1997) for the fleshfly. Treatment with 20 mM GTP in 0.03% DOC for 2 min did not induce any change in impulse frequency to the sugar (data not shown). 3.2. Dose dependency of GDPbS on saponin-induced response of the sugar taste receptor cell To reveal the effects of GDPβS, we examined the effects of various concentrations of GDPβS on CSI- and GL-induced response of the sugar taste receptor cell. Previously we showed that the optimal response of the sugar taste receptor cell was obtained at 0.1 mM CSI and 3.0 mM GL (Ahamed et al., 2000, 2001). We selected these optimal concentrations of CSI and GL for the
Fig. 1. Time course of electrophysiological responses from the LLtype sensillum to 50 mM sucrose (A) and 100 mM fructose (B) after 0.03% DOC treatment with (䊉) and without (䊊) 20 mM GDPβS for 2 min. Data are the average (±SE) of the number of impulses generated during a period of 0.2 s (n ⫽ 12).
following experiments. Figs. 2A and B show the dosedependent depression by GDPβS in the responses to CSI and GL, respectively. Interestingly the effect of GDPβS was observed by simply adding it to CSI or GL solution without DOC treatment, suggesting that CSI and GL mimic the function of DOC in introducing GDPβS inside the sugar receptor cell. Although GDPβS (5–40 mM) greatly depressed the CSI response in a concentrationdependent manner, the decrease in impulse frequency to GL was observed only in the 20 and 40 mM GDPβS treatments. We used 20 mM GDPβS for further experiments. 3.3. Effects of GDPbS on the response to CSI and GL Fig. 3A shows the dose–response curve of CSI in the absence and presence of 20 mM GDPβS. Addition of GDPβS greatly decreased the impulse frequency to CSI
370
A. Ahamed et al. / Journal of Insect Physiology 48 (2002) 367–374
Fig. 2. Effect of various concentrations (5, 10, 15, 20 and 40 mM) of GDPβS on 0.1 mM CSI (A) and 3.0 mM GL (B) responses. GDPβS was added to the test solutions of CSI and GL. The open and closed bars indicate the impulse frequency in the absence and presence of GDPβS, respectively. Data are the average (±SE) of the maximum number of impulses generated during a period of 0.2 s for CSI (n ⫽ 8) and the number of impulses generated during a period of 0.2 s starting 0.15 s after the beginning of stimulation for GL (n ⫽ 10).
over concentrations from 0.03 to 1.0 mM (Fig. 3A). Fig. 3B shows the dose–response curve for GL in the absence and presence of 20 mM GDPβS. Compared to CSI, the inhibitory effect of GDPβS on the impulse frequency to GL was less. In the next experiment, we examined the time course of GDPβS-induced inhibition in the impulse frequency to CSI and GL (Figs. 4A and B). The CSI response was decreased by GDPβS immediately after beginning (Fig. 4A). In contrast, the impulse frequency to GL was not inhibited initially by GDPβS treatment, but it was greatly decreased after 8 s (Fig. 4B). GL exerts dual effects
Fig. 3. Electrophysiological responses from the LL-type sensillum to various concentrations (0.01, 0.03, 0.1, 0.3 and 1.0 mM) of CSI (A) and (0.1, 0.3, 1.0, 3.0 and 10.0 mM) GL (B) in the absence (䊊) and presence (䊉) of 20 mM GDPβS. GDPβS was added to the test solutions of CSI and GL. Data are the average (±SE) of the number of impulses generated during a period of 0.2 s (n = 10 – (20) as in Fig. 2.
to induce the sugar response in blowfly: first, by binding to the pyranose receptor site on the sugar receptor cells as a stimulant, and second, by another mechanism after penetrating the cell membrane (Ahamed et al., 2001). We suspect that when GL and GDPβS are added together, GDPβS cannot inhibit the receptor-mediated sugar signal exerted by GL which begins rapidly before
A. Ahamed et al. / Journal of Insect Physiology 48 (2002) 367–374
371
Fig. 5. The effect of GDPβS on the response to 3.0 mM GL stimulus. Response to GL was recorded after DOC treatment for 2 min with (䊉) and without 20 mM GDPβS (䊊). Data are the average (±SE) of the number of impulses generated during a period of 0.2 s (n ⫽ 10).
Fig. 4. Time course of the response of sugar receptor cell to 0.1 mM CSI (A) and 3.0 mM GL (B) in the absence (䊊) and presence (䊉) of 20 mM GDPβS. GDPβS was added to the test solutions of CSI and GL. Data are the average (±SE) of the number of impulses generated during a period of 0.2 s (n ⫽ 8).
the penetration of GDPβS into the cells. The receptormediated response is absent in the case of CSI, as CSI does not interact with the receptor molecule on the membrane (Ahamed et al., 2000). To see the effects of GDPβS on the receptor-mediated response, we treated the chemosensilla with GDPβS before the GL stimulus. Fig. 5 shows the time course of the GL response after treating with 20 mM GDPβS in 0.03% DOC or with only DOC (control) for 2 min. Here the impulse frequency to GL greatly decreased from the beginning, suggesting that the G protein-mediated cascade is involved in the receptor-mediated response induced by GL as well as by sugars. The sugar response to CSI was almost completely depressed after the treatment with 20 mM GDPβS in 0.03% DOC, while not with only 0.03% DOC (data not shown). When the GL stimulus is repeated with an interval of 3–10 min, the impulse frequency to the second stimulus is higher than that to the first one (Fig. 6, Ahamed et al., 2001). We also examined the effects of GDPβS on the second GL response. GDPβS completely diminished the enhanced response to the second GL stimulus. No change in response was observed in the presence of GTP (data not shown).
Fig. 6. Effect of GDPβS on the response to second GL stimulus. Response to the second GL stimulus was recorded after DOC treatment with (䊉) or without 20 mM GDPβS (䊊) followed by the first and second 3.0 mM GL stimuli. Time intervals between the two GL stimuli are shown in the abscissa. The duration time for GL stimuli was 30 s. A broken line indicates the impulse frequency to the first GL stimulus. Data are the average (±SE) of the number of impulses generated during a period of 0.2 s starting 0.15 s after the beginning of stimulation (n ⫽ 6).
3.4. Remaining impulses induced by GL and CSI Sucrose and fructose did not induce a remaining effect after removing their stimuli (Amakawa et al., 1990, 1992), but both GL and CSI showed remaining effects. When the electrode including a buffer solution was applied to a chemosensillum 1 min after 0.3 mM GL
372
A. Ahamed et al. / Journal of Insect Physiology 48 (2002) 367–374
stimulus for 30 s, the response to the buffer was 13 impulses/0.2 s, which is the remaining effect of the GL stimulus (Fig. 7). This frequency value is greater than that observed under continuous application of GL (Fig. 4B). When the GL stimulus was given to a chemosensillum for 2 min, the response to GL decreased gradually to 4 impulses/0.2 s at the end of the stimulus because of adaptation. This difference clearly shows that the GL molecule incorporated inside the sugar receptor cell can stimulate the sugar signaling cascade. Furthermore, the great increase in the frequency after removing the GL stimulus is consistent with the ideas that GL molecules in a test solution induce the adaptation cascade by interacting with the P site and that, in the case of buffer without GL, activation of the adaptation cascade does not occur (Figs. 4B and 7). When the buffer solution was given 1 min after 0.1 mM CSI stimulus, the response to the buffer was 7 impulses/0.2 s (Fig. 7), which is the remaining effect of CSI stimulus. In contrast to GL, this frequency was the same as when the CSI stimulus was applied continuously (Fig. 4A). The absence of increase in frequency after removing the CSI stimulus is in accordance with the ideas that CSI neither interacts with sugar receptor sites nor induces adaptation and that the remaining response of CSI is due to the CSI-induced activation of sugar signaling inside the cells.
Fig. 8. The effect of the CSI-specific antibody on the impulse frequency of CSI response in the sugar taste receptor cell. The open bar shows the response to 0.1 mM CSI in the absence of antibody (control) and the shaded bars show the response to CSI solution pre-incubated with the antibody for 1 min, 30 min, 1 h and 2 h, respectively. Data are the average (±SE) of the maximum number of impulses generated during a period of 0.2 s (n ⫽ 7).
The response to the CSI solution decreased gradually with increased time of incubation with the antibody.
3.5. Effect of CSI-specific antibody on CSI response When 0.1 mM CSI was mixed with the CSI-specific antibody, the impulse frequency to the CSI solution decreased (Fig. 8). The antibody was incubated with 0.1 mM CSI before the electrophysiological measurement.
Fig. 7. Time course of the remaining impulse of sugar receptor cell to 20 mM MOPS buffer after treating with 0.1 mM CSI (䊊) or 3.0 mM GL (䊉) for 30 s followed by a 1 min interval. Data are the average (±SE) of the number of impulses generated during a period of 0.2 s (n=2–5).
4. Discussion 4.1. Responses to sugars are mediated by G protein in the sugar taste receptor cell of blowfly It is established that a nonhydrolyzable GDP analogue, GDPβS, strongly binds to G protein and resists the replacement by GTP, leading to G protein inactivation (Eckstein et al., 1979). Koganezawa and Shimada (1997) first electrophysiologically showed the involvement of G protein in the sugar response induced by several sites on the sugar receptor cell in the fleshfly. We found that GDPβS also greatly depressed both the sucrose and fructose responses in the blowfly when applied with DOC before the sugar stimuli (Figs. 1A and B). The effect of the inhibitor incorporated into the cells was completely reversible and disappeared about 10 min after treating with the inhibitor in a DOC solution. We could not find any effect of GTP, which we used as a control. Previously, Koganezawa and Shimada (1997) suggested that the intracellular concentration of GDPβS remaining in the taste receptor cells after DOC treatment is about 1/1000–1/100 of the concentration in the solution applied externally. Probably for this reason, in the present study, the depressing effect of GDPβS was
A. Ahamed et al. / Journal of Insect Physiology 48 (2002) 367–374
observed at a millimolar concentration in the test solutions. Murakami and Kijima (2000) proposed that a receptor-channel complex be activated directly by sucrose without mediation by a second messenger or G protein. They used the labellar taste cells of a pharate adult of the fleshfly (12–24 h before emergence). We cannot, at present, exclude the possibility that both the ionotropic type and the signal cascade type exist in the sugar taste receptor cell and the ionotropic receptor complex works in parallel to the G protein-mediated receptor-coupled signal transduction cascade. The fact that the treatment by GDPβS did not completely inhibit the induced responses supports this idea. 4.2. Responses to CSI and GL are also mediated by G protein It has already been shown that CSI and GL induce the response of the sugar taste receptor cell in the blowfly, P. regina (Ahamed et al., 2000, 2001). In the present paper, we showed that GDPβS greatly decreased the responses to both CSI and GL, implying that CSI- and GL-induced sugar signal transduction involves a G protein-mediated cascade. It is noteworthy that, in contrast to sucrose and fructose, GDPβS inhibited the response to CSI when applied simultaneously with CSI without the aid of DOC. Since GDPβS can show its inhibitory effect only inside the cells, CSI is likely to induce incorporation of the GDPβS molecule inside the cells. In a previous work we showed that the impulse induced by CSI appeared after a relatively long latency (0.1–1.0 s) (Ahamed et al., 2000). We assume that this latent period is enough for GDPβS to enter inside the membrane and inhibit the signaling cascade. We also proposed that CSI penetrates into the sugar receptor cells before activating the sugar signaling cascade because of the latency in the CSIinduced impulse (Ahamed et al., 2000). CSI is likely to make the membrane permeable for GDPβS as well as for itself. Effect of GDPβS on the response to GL was different from that to CSI. The effects of GDPβS were separated into two parts. When GDPβS was applied with GL at the same time, GDPβS-induced inhibition was observed only in the later phase of the GL response (later than 8 s from the beginning of stimulus) and little inhibition was found within the initial 2 s (Fig. 4B). When GDPβS was incorporated into the cells with the aid of DOC before GL stimulus, the response was greatly decreased from the beginning (Fig. 5). These observations indicate that the GL response is composed of two parts, a rapid and a slow response. This is consistent with our previous ideas including dual effects of GL: interaction with the P site on the membrane and direct activation of the sugar signaling cascade inside the cells (Ahamed et al., 2001). The later response is delayed because of the requirement
373
of penetration of the GL molecule into the sugar receptor cells. We suspect that the early phase of GL response mediated by the P site is very rapid compared with the penetration of GDPβS and then only the later phase was inhibited by GDPβS when they were applied simultaneously. The fact that the simultaneous application of GDPβS with GL inhibited only the later phase of the GL response is consistent with the finding that GDPβS greatly inhibited the whole response to CSI when mixed with CSI (Figs. 3A and 4A) because of the absence of a receptor site-mediated response for CSI. 4.3. Remaining impulses induced by CSI and GL molecules inside the sugar receptor cells Remaining impulses to CSI and GL were clearly observed when the electrode including buffer was given to the chemosensillum 1 min after removing their stimuli (Fig. 7). The frequencies of the remaining impulses to CSI and GL were almost constant for 8 s but decreased gradually in the following several minutes. The presence of remaining effects strongly suggests that both CSI and GL molecules incorporated into the sugar receptor cells can stimulate the sugar signaling cascade inside the cells. Furthermore, the remaining effect of GL showed an interesting characteristic: the frequency of remaining impulses to GL (13 impulses/0.2 s) was greater than the frequency (4 impulses/0.2 s) under continuous GL stimulus. This fact suggests that under continuous GL stimulus the impulse generation is suppressed by the adaptation cascade which is induced by the interaction of GL with the P site on the membrane and that the adaptation cascade is not induced when the GL molecule inside the cell directly activates the sugar signaling cascade. This idea is further consistent with the finding that the frequency of remaining impulses to CSI (7 impulses/0.2 s) was the same as the frequency under a continuous CSI stimulus, because CSI neither interacts with sugar receptor sites nor induces the adaptation cascade. We also examined the effect of antibody against CSI. The antibody decreased greatly the CSI action to induce impulses in sugar receptor cells (Fig. 8), suggesting that when CSI makes a big complex with the antibody, it can no longer penetrate the membrane. In the previous paper, when 3.0 mM GL stimulus was repeated with an interval of 3–10 min, the impulse frequency to the second stimulus was higher than that to the first one and doubled with a 6 min interval (Fig. 6, Ahamed et al., 2001). However, repeated application of CSI stimuli did not induce such a synergistic effect. The enhancing effect of the second stimulus was observed only with the combination of GL stimuli. Although the mechanism for this synergistic effect is not yet known, involvement of the G protein-mediated cascade for this phenomenon is suggested because of a complete disap-
374
A. Ahamed et al. / Journal of Insect Physiology 48 (2002) 367–374
pearance of the enhancing effect of the second GL stimulus following the treatment with GDPβS. 4.4. Constant frequency of impulse in CSI response and in the later phase of GL response Impulse frequency to sucrose decreases rapidly due to an adaptation mechanism (Figs. 1A and B). Amakawa and Ozaki (1989) and Ozaki and Amakawa (1992) showed that protein kinase C, IP3, Ca2+ and phorbol ester promote adaptation on the sugar taste receptor cell of the blowfly, which implies a possible involvement of a G protein related feedback mechanism in the sugar taste receptor cell. In contrast to sucrose, impulse generation by CSI stimulus showed a constant frequency under the continuous stimulus (Fig. 4A) and the impulse frequency in the later phase of the GL response was also constant when the stimulus was given continuously (Fig. 4B). To explain the constant frequency, two possibilities are raised. First, CSI and GL may inhibit the adaptation cascade. Second, CSI and GL incorporated in the sugar receptor cells may activate the sugar-signaling cascade continuously. In the previous section, to explain the difference in the remaining effects between CSI and GL, we introduced two ideas that the adaptation cascade is induced by the interaction of the GL molecule with the P site on the membrane but not for CSI and that the adaptation cascade is not induced when both CSI and GL directly activate the sugar signaling cascade inside the cell. These ideas are consistent with the second possibility but not with the first one. Therefore, the simplest explanation for the constant frequency in both CSI and GL responses is due to the constant stimulatory effects of these saponins on the sugar transduction cascade, which does not induce adaptation. 4.5. Conclusion Although CSI is tasteless for humans, the other triterpenoid saponin GL induces sugar responses both in humans and insects. Naim et al. (1994) reported very interesting evidence regarding some amphipathic taste substances (saccharin, sodium cyclamate, aspartame, neohesperidin dihydrochalcone, naringin etc.) that are direct activators of G protein in vertebrates. Since both CSI and GL are amphipathic in nature and also induce sugar impulses in the blowfly, it raises a possibility that these compounds may be the direct activators of G protein in inducing the sugar response. Our results with GDPβS strongly support the above idea, since in both cases we found a marked inhibition in the response of
the sugar taste receptor cell (Figs. 4A, 4B and 5). We believe that our present work will help to open new concepts in the mechanism of action of triterpenoid saponins in the invertebrate chemosensillum and will help with further investigations about their effects in vertebrates. References Ahamed, A., Tsurumi, S., Ozaki, M., Amakawa, T., 2000. Chromosaponin I stimulates the sugar taste receptor cells of the blowfly, Phormia regina. Comparative Biochemistry and Physiology Part A 125, 343–349. Ahamed, A., Tsurumi, S., Ozaki, M., Amakawa, T., 2001. An artificial sweetener stimulates the sweet taste in insect: dual effects of glycyrrhizin in Phormia regina. Chemical Senses 26, 507–515. Amakawa, T., Kawata, K., Ozaki, M., 1992. Nucleotide receptor-site on the labellar sugar receptor cell of the blowfly Phormia regina. Journal of Insect Physiology 38, 365–371. Amakawa, T., Ozaki, M., 1989. Protein kinase C-promoted adaptation of the sugar receptor cell of the blowfly Phormia regina. Journal of Insect Physiology 35, 233–237. Amakawa, T., Ozaki, M., Kawata, K., 1990. Effects of cyclic GMP on the sugar taste receptor cell of the fly Phormia regina. Journal of Insect Physiology 36, 281–286. Eckstein, F., Cassel, D., Lekoviyz, H., Lowe, M., Selinger, Z., 1979. Guanosine 5⬘-O-(2-thiophosphate): an inhibitor of adenylate cyclase stimulation by guanine nucleotides and fluoride ions. Journal of Biological Chemistry 254, 9829–9834. Hara, M., 1983. Competition of polysaccharides with sugar for the pyranose and furanose site in the labellar sugar receptor cell of the blowfly, Phormia regina. Journal of Insect Physiology 29, 113– 118. Hodgson, E.S., Lettvin, J.Y., Roeder, K.D., 1955. Physiology of primary chemoreceptor unit. Science 122, 417–418. Koganezawa, M., Shimada, I., 1997. The effects of G protein modulators on the labellar taste receptor cells of the fleshfly (Boettcherisca peregrina). Journal of Insect Physiology 43, 225– 233. Morita, H., 1969. Electric signs of taste receptor activity. In: Pfaffmann, C. (Ed.), Olfaction and Taste, Vol. III. Pergamon Press, Oxford, pp. 370–381. Morita, H., 1972. Primary processes of insect chemoreception. In: Kotani, M. (Ed.), Advances in Biophysics III. University of Tokyo Press, Tokyo, pp. 161–198. Murakami, M., Kijima, H., 2000. Transduction ion channels directly gated by sugars on the insect taste cell. Journal of General Physiology 115, 455–466. Naim, M., Seifert, P.R., Nurnberg, B., Grunbaum, L., Schultz, G., 1994. Some taste substances are direct activators of G-proteins. Biochemical Journal 297, 451–454. Ozaki, M., Amakawa, T., 1992. Adaptation-promoting effects of IP3, Ca2+, and phorbol ester on the sugar taste receptor cell of the blowfly, Phormia regina. Journal of General Physiology 100, 867–879. Tsurumi, S., Takagi, T., Hashimoto, T., 1992. A γ-pyronyl-triterpenoid saponin from Pisum sativum. Phytochemistry 31, 2435–2438. Ueno, K., Ohta, M., Morita, H., Mikuni, Y., Nakajima, S., Yamamoto, K., Isono, K., 2001. Trehalose sensitivity in Drosophila correlates with mutations in and expression of the gustatory receptor gene Gr5a. Current Biology 11, 1451–1455.
Triterpenoid saponins stimulate the sugar taste receptor cell through a G protein-mediated mechanism in the blowfly, Phormia regina Arifa Ahamed a, Seiji Tsurumi b, Taisaku Amakawa a,c,∗ a
Graduate School of Science and Technology, Kobe University, Kobe 657-8501, Japan b Radioisotope Research Center, Kobe University, Kobe 657-8501, Japan c Faculty of Human Development, Kobe University, Kobe 657-8501, Japan Received 29 August 2001; accepted 17 January 2002
Abstract The blowfly has taste chemosensilla on the labellum. The sensory receptor cells in the chemosensillum are highly specialized for the tastes of sugar, salt and water, respectively. Previously we introduced chromosaponin I (CSI) and glycyrrhizin (GL), as sweet substances for the blowfly, Phormia regina. Application of these triterpenoid saponins induced feeding responses as well as impulses of the sugar taste receptor cell in the LL-type sensillum at a much lower concentration than that of sucrose. In the present paper, we show the involvement of G protein-mediated cascade in the CSI- and GL-responses as well as in sugar responses. CSI activates the sugar signal transduction cascade after penetrating through the membrane. On the other hand, GL exerts dual effects to stimulate the sugar signal transduction possibly by activating it inside the cell and also by interacting with the pyranose sugar receptor site. A non hydrolyzable G protein inhibitor guanosine 5⬘-O-(2-thiodiphosphate), GDPβS, markedly decreased the responses of the sugar receptor cell to the two triterpenoid saponins as well as the response to sucrose and fructose. These results suggest that CSI and GL are direct activators of G protein. 2002 Elsevier Science Ltd. All rights reserved. Keywords: Blowfly; G protein; Sugar taste; Transduction; Triterpenoid saponins
1. Introduction The taste organ of the fly is a sensillum showing a much less complicated structure than the vertebrate taste bud. A sensillum contains four sensory processes elongated from four functionally differentiated taste cells which respond with unique impulses to sugars, salts, water and large anions like caprylate, respectively. Thus, it is easy to observe the electrophysiological response from a single taste cell by giving adequate stimulus to the tip of the sensillum. Furthermore, it is also easy to examine the pharmacological effects of reagents on a single taste cell by applying reagents to a sensillum for checking the response. One of the characteristic phenomena of taste reception
Corresponding author. Tel.: +81-78-803-7747; fax: +81-78-8037761. E-mail address: [email protected] (T. Amakawa). ∗
in insects is its very fast response: the latency is as short as within several milliseconds for sugar or salt stimulation. Because of this fast response, it is widely accepted that the primary mechanism of taste reception includes the ionotropic type ion channels rather than the cascade type such as G protein-mediated, which take longer times to respond. In fact, Murakami and Kijima (2000) recently succeeded in recording sugar activated ion channels in the chemosensillum of fleshfly pupae. They further showed using GDPβS that no contribution by G protein was seen. Amakawa et al. (1990, 1992), on the other hand, reported that dibutyryl cyclic GMP (dbcGMP), a membrane permeable analogue of cyclic GMP, elicits the response of the sugar receptor cell of the blowfly with poor adaptation. They suggested that dbcGMP penetrated the cell membrane and opened the cyclic nucleotide gated channel as a mimic of the second messenger. In the fleshfly, Koganezawa and Shimada (1997) showed
0022-1910/02/$ - see front matter 2002 Elsevier Science Ltd. All rights reserved. PII: S 0 0 2 2 - 1 9 1 0 ( 0 2 ) 0 0 0 5 6 - 2
368
A. Ahamed et al. / Journal of Insect Physiology 48 (2002) 367–374
that the proven G protein inhibitor guanosine 5⬘-O-(2thiodiphosphate), GDPβS, markedly depressed the sugar responses of the labellar sugar receptor cell using sucrose, fructose and valine as stimulants. This indicates that the response of the sugar receptor cell is elicited through a G protein-mediated second messenger cascade. Another piece of evidence supporting G protein inclusion in the reception process comes from molecular genetic research: Ueno et al. (2001) recently identified a taste receptor gene, Tre, that controls the taste sensitivity to trehalose in Drosophila melanogaster and hypothesized that TRE belongs to a novel family of G protein-linked transmembrane receptors that may operate as taste receptors. Previously, as a part of our electrophysiological studies on triterpenoid saponins, we have shown that chromosaponin I (CSI), a γ-pyronyl-triterpenoid saponin isolated from pea and other leguminous plants (Tsurumi et al., 1992), stimulates the sugar receptor cells of the blowfly without affecting the sugar receptor sites on the membrane (Ahamed et al., 2000). Another triterpenoid saponin, glycyrrhizin (GL), found in the root of licorice (Glycyrrhizia glabra), is used extensively as a sweetener as well as a medicine for humans. We showed that GL, a structural analogue of CSI, induced full proboscis extension in the blowfly as does CSI and that it exerted more intense signals of the sugar receptor cell than CSI (Ahamed et al., 2001). It is known that the sugar receptor cells of the blowfly, Phormia regina, possess at least two types of receptor sites that are distinguished by selective competitive inhibition with polysaccharides, starch for the pyranose site (the P site) and levan for the furanose site (the F site) (Hara, 1983). The GL-induced stimulation of the sugar receptor cells was separated into two parts: activation via the P site on the membrane and an additional mechanism, possibly direct activation of the sugar signal transduction cascade inside the cells not mediated by the P site (Ahamed et al., 2001). In the present paper, we show that the responses to CSI and GL as well as to sugars are greatly reduced by GDPβS, indicating the involvement of a G proteinmediated second messenger cascade in these responses of sugar taste receptor cells in blowflies. We furthermore studied in detail the effect of these triterpenoid saponins on the taste receptor cell, using the G protein inhibitor, GDPβS.
2. Materials and methods 2.1. Fly The blowflies P. regina were reared in the laboratory at 24±1 °C and fed with 0.1 M sucrose and water. Seven
to nine day old adult blowflies were used for our experiments. 2.2. Electrophysiological procedure The impulses were recorded by the tip recording method of Hodgson et al. (1955) and the responses were recorded from the labellar LL-type sensilla as described before (Ahamed et al., 2001). Except for elongated stimuli, the duration time of stimulus was 30 s. The impulses during the initial period (to 0.15 s after the beginning of stimulation) were ignored because the impulse frequency was not proportional to the receptor potential during the initial period (Morita, 1969, 1972; Ozaki and Amakawa, 1992). To compare the magnitude of the sugar and GL responses, we counted the number of impulses during a period of 0.2 s starting 0.15 s after the beginning of stimulation. In the case of CSI, due to the variable latency, we selected the maximum number of impulse during a period of 0.2 s (Ahamed et al., 2000). The experiments were carried out at 24±1 °C, relative humidity being kept at 70–80%. 2.3. Chemicals CSI was isolated from pea and purified as described previously (Tsurumi et al., 1992). GL (ammonium salt, purity approx. 75%), GDPβS and GTP were purchased from Sigma Chemical Co. (St Louis, MO, USA). Sucrose, fructose, sodium deoxycholate (DOC) and other chemicals were purchased from Wako Pure Chemicals Industries Ltd (Osaka, Japan). 2.4. Test solutions CSI and GL were dissolved in 20 mM MOPS buffer (pH 6.6). Sucrose and fructose solutions contained 10 mM NaCl for electrical conductance. Since the cell membrane is virtually impermeable to the nonhydrolyzable nucleotide, GDPβS, it is necessary to aid the incorporation of the reagent into the cell. Direct microinjection was not possible, since the sensory process of the fly taste receptor cell is too slender. Amakawa and Ozaki (1989) suggested that 0.03% DOC temporarily increases the permeability of the receptor membrane without membrane lysis. The sugar receptor cell preserved its responsiveness after such mild detergent treatment. So DOC was used to introduce GDPβS or GTP into the sugar taste receptor cells. The GDPβS or GTP was dissolved in 0.03% DOC in 67 mM phosphate buffer, pH 7.2, as described before (Amakawa and Ozaki, 1989). After treatment for 2 min with one of these DOC solutions, the electrophysiological responses to sugars or saponins were recorded 5 min after the DOC treatment. Treatment with 0.03% DOC alone was used as the control. To see the effect of GDPβS on the response to CSI
A. Ahamed et al. / Journal of Insect Physiology 48 (2002) 367–374
369
or GL, GDPβS was added to saponin solution because these saponins introduced GDPβS inside the sugar receptor cells without an aid of DOC. The CSI specific antibody was purified in our laboratory (Ahamed et al., in preparation). The polyclonal antibodies raised against CSI were purified with soyasaponin I- and CSI-affinity chromatography. The triterpenoid and sugar moieties of soyasaponin I were the same as those of CSI, but the former does not have the γ-pyronyl moiety. CSI-specific antibody was obtained from the fraction which binds to the CSI-affinity column but does not bind to the soyasaponin I-affinity column. The specificity of the purified antibody was confirmed by a ligand-molecule interaction analysis with BIACORE 3000. The antibody concentration used for the experiments was 200 µg/ml, dissolved in a neutral buffer. The incubation of CSI with the antibody was carried out at 24±1 °C. 3. Results 3.1. Effect of GDPbS on sugar response Figs. 1A and B show the time course of the responses of the sugar receptor in the blowfly to 50 mM sucrose and 100 mM fructose, respectively, after treatment with 0.03% DOC (control) or with 20 mM GDPβS in 0.03% DOC for 2 min. DOC treatment alone appeared to slightly accelerate adaptation (Amakawa and Ozaki, 1989) and GDPβS alone did not induce any impulse (data not shown). The responses of the sugar receptor cell to both sucrose and fructose were apparently reduced after GDPβS treatment, compared with control. However, we did not find complete inhibition by GDPβS treatment. We observed no change in the sugar response when GDPβS was added without DOC, indicating that this nonhydrolyzable G protein inhibitor must be introduced inside the sugar receptor cell to exert its action. No remaining effects of GDPβS were observed 10 min after removing the reagent from the labellar hair. These results were almost the same as those by Koganezawa and Shimada (1997) for the fleshfly. Treatment with 20 mM GTP in 0.03% DOC for 2 min did not induce any change in impulse frequency to the sugar (data not shown). 3.2. Dose dependency of GDPbS on saponin-induced response of the sugar taste receptor cell To reveal the effects of GDPβS, we examined the effects of various concentrations of GDPβS on CSI- and GL-induced response of the sugar taste receptor cell. Previously we showed that the optimal response of the sugar taste receptor cell was obtained at 0.1 mM CSI and 3.0 mM GL (Ahamed et al., 2000, 2001). We selected these optimal concentrations of CSI and GL for the
Fig. 1. Time course of electrophysiological responses from the LLtype sensillum to 50 mM sucrose (A) and 100 mM fructose (B) after 0.03% DOC treatment with (䊉) and without (䊊) 20 mM GDPβS for 2 min. Data are the average (±SE) of the number of impulses generated during a period of 0.2 s (n ⫽ 12).
following experiments. Figs. 2A and B show the dosedependent depression by GDPβS in the responses to CSI and GL, respectively. Interestingly the effect of GDPβS was observed by simply adding it to CSI or GL solution without DOC treatment, suggesting that CSI and GL mimic the function of DOC in introducing GDPβS inside the sugar receptor cell. Although GDPβS (5–40 mM) greatly depressed the CSI response in a concentrationdependent manner, the decrease in impulse frequency to GL was observed only in the 20 and 40 mM GDPβS treatments. We used 20 mM GDPβS for further experiments. 3.3. Effects of GDPbS on the response to CSI and GL Fig. 3A shows the dose–response curve of CSI in the absence and presence of 20 mM GDPβS. Addition of GDPβS greatly decreased the impulse frequency to CSI
370
A. Ahamed et al. / Journal of Insect Physiology 48 (2002) 367–374
Fig. 2. Effect of various concentrations (5, 10, 15, 20 and 40 mM) of GDPβS on 0.1 mM CSI (A) and 3.0 mM GL (B) responses. GDPβS was added to the test solutions of CSI and GL. The open and closed bars indicate the impulse frequency in the absence and presence of GDPβS, respectively. Data are the average (±SE) of the maximum number of impulses generated during a period of 0.2 s for CSI (n ⫽ 8) and the number of impulses generated during a period of 0.2 s starting 0.15 s after the beginning of stimulation for GL (n ⫽ 10).
over concentrations from 0.03 to 1.0 mM (Fig. 3A). Fig. 3B shows the dose–response curve for GL in the absence and presence of 20 mM GDPβS. Compared to CSI, the inhibitory effect of GDPβS on the impulse frequency to GL was less. In the next experiment, we examined the time course of GDPβS-induced inhibition in the impulse frequency to CSI and GL (Figs. 4A and B). The CSI response was decreased by GDPβS immediately after beginning (Fig. 4A). In contrast, the impulse frequency to GL was not inhibited initially by GDPβS treatment, but it was greatly decreased after 8 s (Fig. 4B). GL exerts dual effects
Fig. 3. Electrophysiological responses from the LL-type sensillum to various concentrations (0.01, 0.03, 0.1, 0.3 and 1.0 mM) of CSI (A) and (0.1, 0.3, 1.0, 3.0 and 10.0 mM) GL (B) in the absence (䊊) and presence (䊉) of 20 mM GDPβS. GDPβS was added to the test solutions of CSI and GL. Data are the average (±SE) of the number of impulses generated during a period of 0.2 s (n = 10 – (20) as in Fig. 2.
to induce the sugar response in blowfly: first, by binding to the pyranose receptor site on the sugar receptor cells as a stimulant, and second, by another mechanism after penetrating the cell membrane (Ahamed et al., 2001). We suspect that when GL and GDPβS are added together, GDPβS cannot inhibit the receptor-mediated sugar signal exerted by GL which begins rapidly before
A. Ahamed et al. / Journal of Insect Physiology 48 (2002) 367–374
371
Fig. 5. The effect of GDPβS on the response to 3.0 mM GL stimulus. Response to GL was recorded after DOC treatment for 2 min with (䊉) and without 20 mM GDPβS (䊊). Data are the average (±SE) of the number of impulses generated during a period of 0.2 s (n ⫽ 10).
Fig. 4. Time course of the response of sugar receptor cell to 0.1 mM CSI (A) and 3.0 mM GL (B) in the absence (䊊) and presence (䊉) of 20 mM GDPβS. GDPβS was added to the test solutions of CSI and GL. Data are the average (±SE) of the number of impulses generated during a period of 0.2 s (n ⫽ 8).
the penetration of GDPβS into the cells. The receptormediated response is absent in the case of CSI, as CSI does not interact with the receptor molecule on the membrane (Ahamed et al., 2000). To see the effects of GDPβS on the receptor-mediated response, we treated the chemosensilla with GDPβS before the GL stimulus. Fig. 5 shows the time course of the GL response after treating with 20 mM GDPβS in 0.03% DOC or with only DOC (control) for 2 min. Here the impulse frequency to GL greatly decreased from the beginning, suggesting that the G protein-mediated cascade is involved in the receptor-mediated response induced by GL as well as by sugars. The sugar response to CSI was almost completely depressed after the treatment with 20 mM GDPβS in 0.03% DOC, while not with only 0.03% DOC (data not shown). When the GL stimulus is repeated with an interval of 3–10 min, the impulse frequency to the second stimulus is higher than that to the first one (Fig. 6, Ahamed et al., 2001). We also examined the effects of GDPβS on the second GL response. GDPβS completely diminished the enhanced response to the second GL stimulus. No change in response was observed in the presence of GTP (data not shown).
Fig. 6. Effect of GDPβS on the response to second GL stimulus. Response to the second GL stimulus was recorded after DOC treatment with (䊉) or without 20 mM GDPβS (䊊) followed by the first and second 3.0 mM GL stimuli. Time intervals between the two GL stimuli are shown in the abscissa. The duration time for GL stimuli was 30 s. A broken line indicates the impulse frequency to the first GL stimulus. Data are the average (±SE) of the number of impulses generated during a period of 0.2 s starting 0.15 s after the beginning of stimulation (n ⫽ 6).
3.4. Remaining impulses induced by GL and CSI Sucrose and fructose did not induce a remaining effect after removing their stimuli (Amakawa et al., 1990, 1992), but both GL and CSI showed remaining effects. When the electrode including a buffer solution was applied to a chemosensillum 1 min after 0.3 mM GL
372
A. Ahamed et al. / Journal of Insect Physiology 48 (2002) 367–374
stimulus for 30 s, the response to the buffer was 13 impulses/0.2 s, which is the remaining effect of the GL stimulus (Fig. 7). This frequency value is greater than that observed under continuous application of GL (Fig. 4B). When the GL stimulus was given to a chemosensillum for 2 min, the response to GL decreased gradually to 4 impulses/0.2 s at the end of the stimulus because of adaptation. This difference clearly shows that the GL molecule incorporated inside the sugar receptor cell can stimulate the sugar signaling cascade. Furthermore, the great increase in the frequency after removing the GL stimulus is consistent with the ideas that GL molecules in a test solution induce the adaptation cascade by interacting with the P site and that, in the case of buffer without GL, activation of the adaptation cascade does not occur (Figs. 4B and 7). When the buffer solution was given 1 min after 0.1 mM CSI stimulus, the response to the buffer was 7 impulses/0.2 s (Fig. 7), which is the remaining effect of CSI stimulus. In contrast to GL, this frequency was the same as when the CSI stimulus was applied continuously (Fig. 4A). The absence of increase in frequency after removing the CSI stimulus is in accordance with the ideas that CSI neither interacts with sugar receptor sites nor induces adaptation and that the remaining response of CSI is due to the CSI-induced activation of sugar signaling inside the cells.
Fig. 8. The effect of the CSI-specific antibody on the impulse frequency of CSI response in the sugar taste receptor cell. The open bar shows the response to 0.1 mM CSI in the absence of antibody (control) and the shaded bars show the response to CSI solution pre-incubated with the antibody for 1 min, 30 min, 1 h and 2 h, respectively. Data are the average (±SE) of the maximum number of impulses generated during a period of 0.2 s (n ⫽ 7).
The response to the CSI solution decreased gradually with increased time of incubation with the antibody.
3.5. Effect of CSI-specific antibody on CSI response When 0.1 mM CSI was mixed with the CSI-specific antibody, the impulse frequency to the CSI solution decreased (Fig. 8). The antibody was incubated with 0.1 mM CSI before the electrophysiological measurement.
Fig. 7. Time course of the remaining impulse of sugar receptor cell to 20 mM MOPS buffer after treating with 0.1 mM CSI (䊊) or 3.0 mM GL (䊉) for 30 s followed by a 1 min interval. Data are the average (±SE) of the number of impulses generated during a period of 0.2 s (n=2–5).
4. Discussion 4.1. Responses to sugars are mediated by G protein in the sugar taste receptor cell of blowfly It is established that a nonhydrolyzable GDP analogue, GDPβS, strongly binds to G protein and resists the replacement by GTP, leading to G protein inactivation (Eckstein et al., 1979). Koganezawa and Shimada (1997) first electrophysiologically showed the involvement of G protein in the sugar response induced by several sites on the sugar receptor cell in the fleshfly. We found that GDPβS also greatly depressed both the sucrose and fructose responses in the blowfly when applied with DOC before the sugar stimuli (Figs. 1A and B). The effect of the inhibitor incorporated into the cells was completely reversible and disappeared about 10 min after treating with the inhibitor in a DOC solution. We could not find any effect of GTP, which we used as a control. Previously, Koganezawa and Shimada (1997) suggested that the intracellular concentration of GDPβS remaining in the taste receptor cells after DOC treatment is about 1/1000–1/100 of the concentration in the solution applied externally. Probably for this reason, in the present study, the depressing effect of GDPβS was
A. Ahamed et al. / Journal of Insect Physiology 48 (2002) 367–374
observed at a millimolar concentration in the test solutions. Murakami and Kijima (2000) proposed that a receptor-channel complex be activated directly by sucrose without mediation by a second messenger or G protein. They used the labellar taste cells of a pharate adult of the fleshfly (12–24 h before emergence). We cannot, at present, exclude the possibility that both the ionotropic type and the signal cascade type exist in the sugar taste receptor cell and the ionotropic receptor complex works in parallel to the G protein-mediated receptor-coupled signal transduction cascade. The fact that the treatment by GDPβS did not completely inhibit the induced responses supports this idea. 4.2. Responses to CSI and GL are also mediated by G protein It has already been shown that CSI and GL induce the response of the sugar taste receptor cell in the blowfly, P. regina (Ahamed et al., 2000, 2001). In the present paper, we showed that GDPβS greatly decreased the responses to both CSI and GL, implying that CSI- and GL-induced sugar signal transduction involves a G protein-mediated cascade. It is noteworthy that, in contrast to sucrose and fructose, GDPβS inhibited the response to CSI when applied simultaneously with CSI without the aid of DOC. Since GDPβS can show its inhibitory effect only inside the cells, CSI is likely to induce incorporation of the GDPβS molecule inside the cells. In a previous work we showed that the impulse induced by CSI appeared after a relatively long latency (0.1–1.0 s) (Ahamed et al., 2000). We assume that this latent period is enough for GDPβS to enter inside the membrane and inhibit the signaling cascade. We also proposed that CSI penetrates into the sugar receptor cells before activating the sugar signaling cascade because of the latency in the CSIinduced impulse (Ahamed et al., 2000). CSI is likely to make the membrane permeable for GDPβS as well as for itself. Effect of GDPβS on the response to GL was different from that to CSI. The effects of GDPβS were separated into two parts. When GDPβS was applied with GL at the same time, GDPβS-induced inhibition was observed only in the later phase of the GL response (later than 8 s from the beginning of stimulus) and little inhibition was found within the initial 2 s (Fig. 4B). When GDPβS was incorporated into the cells with the aid of DOC before GL stimulus, the response was greatly decreased from the beginning (Fig. 5). These observations indicate that the GL response is composed of two parts, a rapid and a slow response. This is consistent with our previous ideas including dual effects of GL: interaction with the P site on the membrane and direct activation of the sugar signaling cascade inside the cells (Ahamed et al., 2001). The later response is delayed because of the requirement
373
of penetration of the GL molecule into the sugar receptor cells. We suspect that the early phase of GL response mediated by the P site is very rapid compared with the penetration of GDPβS and then only the later phase was inhibited by GDPβS when they were applied simultaneously. The fact that the simultaneous application of GDPβS with GL inhibited only the later phase of the GL response is consistent with the finding that GDPβS greatly inhibited the whole response to CSI when mixed with CSI (Figs. 3A and 4A) because of the absence of a receptor site-mediated response for CSI. 4.3. Remaining impulses induced by CSI and GL molecules inside the sugar receptor cells Remaining impulses to CSI and GL were clearly observed when the electrode including buffer was given to the chemosensillum 1 min after removing their stimuli (Fig. 7). The frequencies of the remaining impulses to CSI and GL were almost constant for 8 s but decreased gradually in the following several minutes. The presence of remaining effects strongly suggests that both CSI and GL molecules incorporated into the sugar receptor cells can stimulate the sugar signaling cascade inside the cells. Furthermore, the remaining effect of GL showed an interesting characteristic: the frequency of remaining impulses to GL (13 impulses/0.2 s) was greater than the frequency (4 impulses/0.2 s) under continuous GL stimulus. This fact suggests that under continuous GL stimulus the impulse generation is suppressed by the adaptation cascade which is induced by the interaction of GL with the P site on the membrane and that the adaptation cascade is not induced when the GL molecule inside the cell directly activates the sugar signaling cascade. This idea is further consistent with the finding that the frequency of remaining impulses to CSI (7 impulses/0.2 s) was the same as the frequency under a continuous CSI stimulus, because CSI neither interacts with sugar receptor sites nor induces the adaptation cascade. We also examined the effect of antibody against CSI. The antibody decreased greatly the CSI action to induce impulses in sugar receptor cells (Fig. 8), suggesting that when CSI makes a big complex with the antibody, it can no longer penetrate the membrane. In the previous paper, when 3.0 mM GL stimulus was repeated with an interval of 3–10 min, the impulse frequency to the second stimulus was higher than that to the first one and doubled with a 6 min interval (Fig. 6, Ahamed et al., 2001). However, repeated application of CSI stimuli did not induce such a synergistic effect. The enhancing effect of the second stimulus was observed only with the combination of GL stimuli. Although the mechanism for this synergistic effect is not yet known, involvement of the G protein-mediated cascade for this phenomenon is suggested because of a complete disap-
374
A. Ahamed et al. / Journal of Insect Physiology 48 (2002) 367–374
pearance of the enhancing effect of the second GL stimulus following the treatment with GDPβS. 4.4. Constant frequency of impulse in CSI response and in the later phase of GL response Impulse frequency to sucrose decreases rapidly due to an adaptation mechanism (Figs. 1A and B). Amakawa and Ozaki (1989) and Ozaki and Amakawa (1992) showed that protein kinase C, IP3, Ca2+ and phorbol ester promote adaptation on the sugar taste receptor cell of the blowfly, which implies a possible involvement of a G protein related feedback mechanism in the sugar taste receptor cell. In contrast to sucrose, impulse generation by CSI stimulus showed a constant frequency under the continuous stimulus (Fig. 4A) and the impulse frequency in the later phase of the GL response was also constant when the stimulus was given continuously (Fig. 4B). To explain the constant frequency, two possibilities are raised. First, CSI and GL may inhibit the adaptation cascade. Second, CSI and GL incorporated in the sugar receptor cells may activate the sugar-signaling cascade continuously. In the previous section, to explain the difference in the remaining effects between CSI and GL, we introduced two ideas that the adaptation cascade is induced by the interaction of the GL molecule with the P site on the membrane but not for CSI and that the adaptation cascade is not induced when both CSI and GL directly activate the sugar signaling cascade inside the cell. These ideas are consistent with the second possibility but not with the first one. Therefore, the simplest explanation for the constant frequency in both CSI and GL responses is due to the constant stimulatory effects of these saponins on the sugar transduction cascade, which does not induce adaptation. 4.5. Conclusion Although CSI is tasteless for humans, the other triterpenoid saponin GL induces sugar responses both in humans and insects. Naim et al. (1994) reported very interesting evidence regarding some amphipathic taste substances (saccharin, sodium cyclamate, aspartame, neohesperidin dihydrochalcone, naringin etc.) that are direct activators of G protein in vertebrates. Since both CSI and GL are amphipathic in nature and also induce sugar impulses in the blowfly, it raises a possibility that these compounds may be the direct activators of G protein in inducing the sugar response. Our results with GDPβS strongly support the above idea, since in both cases we found a marked inhibition in the response of
the sugar taste receptor cell (Figs. 4A, 4B and 5). We believe that our present work will help to open new concepts in the mechanism of action of triterpenoid saponins in the invertebrate chemosensillum and will help with further investigations about their effects in vertebrates. References Ahamed, A., Tsurumi, S., Ozaki, M., Amakawa, T., 2000. Chromosaponin I stimulates the sugar taste receptor cells of the blowfly, Phormia regina. Comparative Biochemistry and Physiology Part A 125, 343–349. Ahamed, A., Tsurumi, S., Ozaki, M., Amakawa, T., 2001. An artificial sweetener stimulates the sweet taste in insect: dual effects of glycyrrhizin in Phormia regina. Chemical Senses 26, 507–515. Amakawa, T., Kawata, K., Ozaki, M., 1992. Nucleotide receptor-site on the labellar sugar receptor cell of the blowfly Phormia regina. Journal of Insect Physiology 38, 365–371. Amakawa, T., Ozaki, M., 1989. Protein kinase C-promoted adaptation of the sugar receptor cell of the blowfly Phormia regina. Journal of Insect Physiology 35, 233–237. Amakawa, T., Ozaki, M., Kawata, K., 1990. Effects of cyclic GMP on the sugar taste receptor cell of the fly Phormia regina. Journal of Insect Physiology 36, 281–286. Eckstein, F., Cassel, D., Lekoviyz, H., Lowe, M., Selinger, Z., 1979. Guanosine 5⬘-O-(2-thiophosphate): an inhibitor of adenylate cyclase stimulation by guanine nucleotides and fluoride ions. Journal of Biological Chemistry 254, 9829–9834. Hara, M., 1983. Competition of polysaccharides with sugar for the pyranose and furanose site in the labellar sugar receptor cell of the blowfly, Phormia regina. Journal of Insect Physiology 29, 113– 118. Hodgson, E.S., Lettvin, J.Y., Roeder, K.D., 1955. Physiology of primary chemoreceptor unit. Science 122, 417–418. Koganezawa, M., Shimada, I., 1997. The effects of G protein modulators on the labellar taste receptor cells of the fleshfly (Boettcherisca peregrina). Journal of Insect Physiology 43, 225– 233. Morita, H., 1969. Electric signs of taste receptor activity. In: Pfaffmann, C. (Ed.), Olfaction and Taste, Vol. III. Pergamon Press, Oxford, pp. 370–381. Morita, H., 1972. Primary processes of insect chemoreception. In: Kotani, M. (Ed.), Advances in Biophysics III. University of Tokyo Press, Tokyo, pp. 161–198. Murakami, M., Kijima, H., 2000. Transduction ion channels directly gated by sugars on the insect taste cell. Journal of General Physiology 115, 455–466. Naim, M., Seifert, P.R., Nurnberg, B., Grunbaum, L., Schultz, G., 1994. Some taste substances are direct activators of G-proteins. Biochemical Journal 297, 451–454. Ozaki, M., Amakawa, T., 1992. Adaptation-promoting effects of IP3, Ca2+, and phorbol ester on the sugar taste receptor cell of the blowfly, Phormia regina. Journal of General Physiology 100, 867–879. Tsurumi, S., Takagi, T., Hashimoto, T., 1992. A γ-pyronyl-triterpenoid saponin from Pisum sativum. Phytochemistry 31, 2435–2438. Ueno, K., Ohta, M., Morita, H., Mikuni, Y., Nakajima, S., Yamamoto, K., Isono, K., 2001. Trehalose sensitivity in Drosophila correlates with mutations in and expression of the gustatory receptor gene Gr5a. Current Biology 11, 1451–1455.