INHIBITION OF MIDGUT ION TRANSPORT BY ALLATOTROPIN (Mas-AT) AND MANDUCA FLRFamides IN THE TOBACCO Hornworm MANDUCA SEXTA

The midgut is an important site for both nutrient absorption and ionic regulation in lepidopteran larvae (Dow, 1986). The midgut epithelium of the tobacco hornworm Manduca sexta is composed of two major cell types, columnar and goblet cells (Cioffi, 1979). It is thought that the goblet cells are responsible for ion transport whereas the columnar cells are responsible for nutrient absorption (for reviews, see Klein et al. 1996; Sacchi and Wolfersberger, 1996; Turunen and Crailsheim, 1996). As larvae grow and develop, there are dramatic changes in the rate of active ion transport across the midgut. For example, during the molt between the fourth and fifth instars, there is a reduction in short-circuit current (I_sc), a measure of in vitro active ion transport (Chamberlin et al. 1997). After the molt, I_sc rises and remains high until the insect reaches the wandering stage. At this point in the insect’s development, the larva is committed to pupation and there is an abrupt decline in midgut active ion transport (Chamberlin, 1994; Chamberlin and King, 1998). These developmental fluctuations in midgut ion transport could be controlled by the nervous system or by hormonal stimuli, but this aspect of lepidopteran midgut physiology has not been explored. It is clear, however, that many neuropeptides, which are present in the neurosecretory cells of the central nervous system, have also been found in midgut endocrine cells or in the gut innervations of the enteric nervous system (for a review, see Sehnal and Zitnan, 1996). Among those neuropeptides, allatotropin (Mas-AT) immunoreactivity and mRNA have been identified in the nervous tissue during all stages of development and in the midgut of the larval stage of the tobacco hornworm (Zitnan et al. 1993, 1995b; Veenstra and Hagedorn, 1993; Veenstra et al. 1994; Taylor et al. 1996; Bhatt et al. 1997). Nevertheless, the role of Mas-AT during the larval stage is unclear because the only physiological roles of Mas-AT that have been identified are in the pharate adult and adult stages. These are the stimulation of juvenile hormone (JH) biosynthesis by the corpora allata (CA) of pharate adults and adult females (Kataoka et al. 1989; Hebda et al. 1994) and the stimulation of heart rate in the adult moth (Veenstra et al. 1994).

In addition to Mas-AT, three Manduca FLRFamides (F10, F7G and F7D; see Table 1 for amino acid sequences) have been identified in larval nervous tissue (Kingan et al. 1996) and larval midgut (Kingan et al. 1997). These peptides have been shown to affect muscle activity in the adult lepidopteran gut (Kingan et al. 1996; Fujisawa et al. 1993). Despite the presence of FLRFamides and Mas-AT-immunoreactive substances in the larval lepidopteran midgut, the role of these, or any other peptides, in modulating larval midgut function has not been explored. Peptide hormones have been shown to modulate ion transport in a variety of insect epithelia (for a review, see Gäde et al. 1997) and, Mas-AT or the FLRFamides may therefore have a similar function in the larval lepidopteran midgut. The present study examines, for the first time, the effects of several peptides (Mas-AT, F7G, F7D, F10 and others) on in vitro ion transport in the midgut of lepidopteran larvae.

Manduca sexta L., originally derived from a colony of USDA (North Dakota) stock, were reared and staged as described by Chamberlin et al. (1997). Larvae were reared individually on a commercial diet (no. 9783, BioServ, Frenchtown, NJ, USA) at 25 °C on a 16 h:8 h light:dark cycle. The feeding fourth instars used in this study weighed between 0.9 and 1.4 g and showed no signs of preparation for molting (e.g. gut purging, head capsule slippage). Pharate fifth instars were selected as stage F larvae (black mandibles of the fifth instar visible through a clear slipped head capsule; Baldwin and Hakim, 1991). Stage F larvae were set aside 4–7 h before lights off. These larvae molted within a few hours and were designated day 0 in the next light cycle. For this study, feeding fifth instars, weighing between 3.5 and 5 g, were used on day 2. Wandering larvae were collected on day 5 from Gate II larvae (weighing less than 5 g on day 2; Chamberlin, 1994).

Giant silkworms (Hyalophora cecropia) were raised from eggs that had been collected in the field. The hatched larvae were reared in leaves of black cherry (Prunus serotina) at room temperature. Silkworm (Bombyx mori) larvae were provided by Dr Donald H. Dean (The Ohio State University).

Short-circuit current (I_sc) and transepithelial potential difference (PD) were measured using the method described by Chamberlin et al. (1997). Midguts were dissected and mounted in modified Ussing chambers that had a 0.0636 cm²opening. Each half of the chamber contained 7 ml of saline (Chamberlin, 1994), which was continuously bubbled with 100 % oxygen. The PD was monitored using calomel electrodes connected to the Ussing chambers via agar bridges. The I_sc was passed through silver electrodes. The I_sc and PD were monitored using a World Precision Instruments DVC 1000 epithelial voltage clamp, and data were recorded with a computer-based acquisition program, Datacan V (Sable Systems, Inc., Version 1.0).

Synthetic Mas-AT and FMRFamide was purchased from Sigma Chemical Company (St Louis, MO, USA). Manduca allatostatin (Mas-AST) and cockroach allatostatin (Dip-AST7) were provided by Dr S. S. Tobe (University of Toronto, Canada). F10 was provided by Dr T. G. Kingan (University of California, Riverside, CA, USA) and F7G and F7D was supplied by Dr J. L. Witten (University of Wisconsin, Milwaukee, USA). Mas-AT was dissolved in a solution containing 50 mmol l⁻¹ imidazole buffer and 1 mmol l⁻¹ methionine (pH 7.2). Other peptides were dissolved in 50 % methanol with 2 mmol l⁻¹ HCl. Stock solutions of peptides were diluted prior to use in saline.

Posterior midguts from day 2 fifth instars were mounted in Ussing chambers and the I_sc was monitored for approximately 1 h. The voltage clamp was then turned off and the PD recorded 5 min later. After the PD measurement, the voltage clamp was turned on again, and 5 min later 10 μl of the peptide solution was added to the hemolymph side of the tissue. The I_sc was then monitored for 5 min, after which the voltage clamp was turned off again and the PD recorded. Resistance was calculated from the PD and I_sc measured before and after the addition of peptide. In some experiments, two different peptides were added sequentially to the hemolymph side of the posterior midgut to determine whether the different peptides exhibited additive or synergistic effects on the posterior midgut. To determine whether there were regional differences in the responses to these peptides, the effects of Mas-AT, F7G and F7D on the I_sc of the anterior and middle midgut sections were tested.

Mas-AT was diluted 50-fold with its primary antiserum (provided by Dr S. J. Kramer, Sandoz Agro Inc., USA) in imidazole buffer. The mixture was incubated for 3 days at 5 °C while being continuously stirred. Mas-AT, without antiserum (control), was incubated under the same conditions. The effect of preabsorbed Mas-AT or the control solution on posterior midgut I_sc was determined as described above.

To confirm that the Mas-AT employed in this study was biologically active and that this activity could be blocked by preabsorption by its antiserum, stimulation of corpora allata (CA) juvenile hormone (JH) biosynthesis by this peptide was measured using the in vitro radiochemical assay of Pratt and Tobe (1974), as modified by Feyereisen and Tobe (1981). A pair of CA was dissected from adult females and transferred into 100 μl of medium-199 (Gibco) with 20 mmol l⁻¹ Hepes and 2 % Ficoll (Sigma, St Louis, MO, USA) containing L-[³H-methyl]methionine (NEN, Boston, MA, USA) at a final concentration of 0.1 mmol l⁻¹ (final specific activity of 3700 mBq mmol⁻¹). Corpora allata were incubated for 3 h at 30 °C with or without Mas-AT (20 nmol l⁻¹) or Mas-AT preabsorbed to antiserum. After incubation, 250 μl of iso-octane was mixed with the incubation medium and the mixture was centrifuged at 2500 gfor 5 min. A 200 μl sample of the iso-octane (top) phase was then removed, mixed with 1 ml of distilled water and centrifuged at 2500 gfor 5 min. A 150 μl sample of the top iso-octane phase was removed and its radioactivity determined by liquid scintillation counting (EcoLume, ICN, USA).

The effects of Mas-AT, F7G and F7D on the I_sc of the posterior midgut of feeding fourth instars, pharate fifth instars, day 2 fifth instars and wandering (day 5) fifth instars were compared. In addition, the posterior midguts of H. cecropia and B. mori were mounted in Ussing chambers and bathed in Manduca saline. The effects of Mas-AT, F7G and F7D on the I_sc of these epithelia were determined as described above.

Values are expressed as means ± standard error of the mean (S.E.M.) with N indicating the number of midguts measured. Statistical analyses were conducted using an analysis of variance (ANOVA) and a Tukey’s post-hoc test. When appropriate, paired analyses were conducted. Data presented as percentages were log-transformed before statistical analyses. Pe:0.05 was considered to represent a significant difference.

Figs 1 and 2 show some responses of the posterior midgut to Mas-AT and FLRFamides when the I_sc was not interrupted for PD measurements. Addition of Mas-AT (50 nmol l⁻¹) to the hemolymph side of the tissue promptly inhibited the I_sc, and maximal inhibition was reached within 5 min. The I_sc then recovered slowly (Fig. 1A). The inhibition of I_sc by Mas-AT was quickly reversed by rinsing the chambers with peptide-free saline (Fig. 1B). Addition of Manduca FLRFamide peptides, F7G and F7D (100 nmol l⁻¹), to the hemolymph side of the tissue also rapidly inhibited the I_sc (Fig. 2A,B), but the inhibition of I_sc by these two peptides was maintained for more than 1 h after treatment. Addition of all of these peptides to the lumen side of the midgut did not affect the I_sc (data not shown).

Inhibition of I_sc by Mas-AT, F7G and F7D was dose-dependent (Fig. 3), and all three peptides inhibited the I_sc at doses higher than 1 nmol l⁻¹. Maximal inhibition of I_sc by Mas-AT was achieved at 50 nmol l⁻¹, whereas F7G and F7D inhibited the I_sc maximally at 10 nmol l⁻¹. At 100 nmol l⁻¹, Mas-AT produced a significantly greater inhibition of I_sc than did F7D or F7G (Fig. 3). All three peptides significantly increased the resistance of the epithelium (Table 2). Addition of another Manduca FLRFamide, F10, or the molluscan tetrapeptide FMRFamide only slightly inhibited I_sc (7.9±1.3 %, N=7; 3.4±0.6 %, N=7, respectively, at 1 μmol l⁻¹; Fig. 3). Addition of other peptides, Manduca allatostatin (Mas-AST) and a cockroach (Diploptera punctata) allatostatin (Dip-AST7) did not change the I_sc when added at concentrations of 10–100 nmol l⁻¹ (data not shown).

Mas-AT that had been preabsorbed with its antiserum failed to inhibit the I_sc (Fig. 4A). This was not due to non-specific damage to Mas-AT during the preabsorption process because Mas-AT that had been incubated in the absence of antiserum did inhibit the I_sc (Fig. 4B). Confirmation that preabsorption of Mas-AT with its antiserum abolishes its biological activity was obtained using an in vitro radiochemical assay of JH biosynthesis by CA. Total JH production by CA from day 3 adult females was 4.3±0.3 pmol h⁻¹ CA pair⁻¹ (N=4). When Mas-AT (20 nmol l⁻¹) was added to the incubation medium, this rate increased significantly to 22.7±3.6 pmol h⁻¹ CA pair⁻¹ (N=4). When preabsorbed Mas-AT was added, the rate of production of JH was only 4.5±1.7 pmol h⁻¹ CA pair⁻¹ (N=4). Thus, preabsorption with antiserum abolished the effects of Mas-AT in both the midgut and the CA.

After the I_sc had been reduced by a saturating concentration (200 nmol l⁻¹) of Mas-AT, subsequent addition of Mas-AT did not significantly reduce the I_sc (Fig. 5A). Addition of F7G or F7D, however, significantly inhibited the I_sc (Table 3; Fig. 5A,C). Reversing the order of peptide exposure also produced additive inhibition of the I_sc (Fig. 5B,D; Table 3). The effects of F7G and F7D did not appear to be additive (Fig. 5E; Table 3).

The effects of Mas-AT, F7G and F7D on I_sc were determined in different midgut regions of day 2 fifth instars. Under control conditions, the I_sc of the anterior region (149.6±11.0 μequiv cm⁻² h⁻¹, N=7) was significantly higher than that of the middle (99.1±9.8 μequiv cm⁻² h⁻¹, N=7) and posterior (Table 2) regions. The greatest peptide-induced inhibition of the I_sc, however, occurred in the posterior region (Fig. 6A). Of the three peptides tested, Mas-AT produced a significantly greater inhibition of I_sc than F7G and F7D in all three regions of the midgut. The effects of these three peptides on posterior midguts of larvae at different stages of development were also determined (Fig. 6B). The control I_sc was highest in midguts dissected from the feeding stages of fourth (95.4±16.4 μequiv cm⁻² h⁻¹, N=7) and fifth (Table 2) instars and lowest in those of molting larvae (23.8±5.6 μequiv cm⁻² h⁻¹, N=6) and wandering larvae (12.6±2.0 μequiv cm⁻² h⁻¹, N=6). Addition of Mas-AT, F7G or F7D caused the greatest inhibition of I_sc in feeding day 2 fifth instars, but minimal inhibition was observed in pharate fifth instars or wandering larvae. In the presence of Mas-AT, the lowest I_sc of the posterior midgut of feeding larvae was 47.3±5.9 μequiv cm⁻² h⁻¹ (N=6). This value is still significantly higher than the control I_sc of molting or wandering larvae.

When the posterior midguts of two other lepidopteran species were exposed to Mas-AT, F7G and F7D, the effects on I_sc were quite different from those observed in Manduca sexta. Although the control I_sc values of H. cecropia (96.5±4.9 μequiv cm⁻² h⁻¹, N=7) and B. mori (74.3±5.4 μequiv cm⁻² h⁻¹, N=7) were quite high when bathed in Manduca saline, Mas-AT only inhibited the I_sc of H. cecropia and B. mori midguts by 5.4±1.3 % (250 nmol l⁻¹) and 3.9±0.8 % (100 nmol l⁻¹), respectively. Furthermore, addition of 100 nmol l⁻¹ F7G and F7D did not affect the I_sc of the midguts of these two species.

Allatotropin (Mas-AT) has previously been shown to stimulate JH biosynthesis by the CA (Kataoka et al. 1989) and to accelerate heart activity (Veenstra et al. 1994). Both these effects are restricted to the pharate adult and adult stages of development. No effect of this peptide has been detected in larvae (Kataoka et al. 1989; Unni et al. 1991; Hebda et al. 1994; Veenstra et al. 1994). Nonetheless, Mas-AT mRNA and immunoreactivity have been detected in both the larval and adult stages (Zitnan et al. 1993, 1995b; Veenstra et al. 1994; Taylor et al. 1996; Bhatt and Horodyski, 1998), suggesting that Mas-AT may have a biological role in larval insects that differs from its role in the adult. In the present investigation, it was demonstrated that the posterior midgut of feeding M. sexta larvae is a target for Mas-AT action. The peptide rapidly inhibits in vitro active ion transport across the midgut epithelium. Furthermore, two Manduca FLRFamides, F7G and F7D, elicit a similar rapid inhibition of ion transport in the same tissue. Both F7G and F7D are members of a larger family of RFamide peptides characterized by a common C terminus (Kingan et al. 1996). Although F10 is similar to F7G and F7D and shares a common C terminus with these peptides, it did not inhibit midgut I_sc. This difference in biological activity may be due to differences in the structures of the peptides in saline. In solution, F7G and F7D do not exhibit a secondary structure, whereas F10 does (T. G. Kingan, personal communication). This altered conformation may be responsible for the lack of activity of F10 in the present study. In adult M. sexta, FLRFamides exhibit myostimulatory activities on the skeletal flight muscles (F10; Kingan et al. 1990) and on the ileum (F7G and F7D; Kingan et al. 1996). Although F10 was shown to have myosuppressive activity on the adult midgut of a related sphingid moth, Agrius convolvuli (Fujisawa et al. 1993), the larval midgut represents a novel target of FLRFamides (Zitnan et al. 1993, 1995a; Kingan et al. 1997). It is not known whether Mas-AT or FLRFamides exert myostimulatory or myosuppressive actions on the larval midgut of M. sexta, but it is clear from the present study and other investigations that Mas-AT, F7G and F7D have multiple physiological activities in different tissues during distinct developmental stages. The inhibition of I_sc by both Mas-AT and the FLRFamides was rapid, and the inhibition due to Mas-AT was reversible. The response of the midgut to these peptides was dose-dependent, and the peptide concentrations required for activity are similar to those required in other target tissues (Kataoka et al. 1989; Kingan et al. 1996; Veenstra et al. 1994).

Although Mas-AT and the FLRFamides F7G and F7D inhibited I_sc in the posterior midgut, the only structural feature these peptides have in common is the Famide in the C terminus. It is unlikely that this structure is the sole determinant of biological activity, since F10, which also shares this structure (see Table 1), is physiologically inactive on the larval posterior midgut. In addition, the actions of Mas-AT and the FLRFamides are additive, indicating that these peptides act on different cells within the midgut and/or through different receptors on the same cells. The specific cellular target(s) of these peptides is not known. Nevertheless, the sensitivity of the posterior midgut to these peptides points to a target or targets specific to this region of the midgut. The columnar and goblet cells of the posterior midgut have a different morphology from those in the anterior and middle regions (Cioffi, 1979) and, therefore, either cell type could be the target of these peptides. It seems likely, however, that the inhibition of I_sc by these peptides is due to a direct action on the goblet cells because it is this cell type that is thought to be responsible for transepithelial active ion transport (Klein et al. 1996). It is also possible that F7G or F7D acts primarily on midgut muscle cells since muscle cells have been shown to be targets of these peptides (Kingan et al. 1996). Modulation of muscle contraction may then have a secondary effect on the ion-transporting epithelial cells by changing the shape or surface area of the epithelium. Although the exact target(s) of Mas-AT, F7D and F7G has not been identified, it is also possible that the additive effects of these peptides is due to binding to distinct receptors on the basolateral membranes of the same cells. Support for the existence of distinct receptors for these peptides is provided by the observation that, in the adult CA, Mas-AT stimulates JH biosynthesis, whereas F7G and F7D do not (K.-Y. Lee, F. M. Horodyski and M. E. Chamberlin, unpublished observation).

The inhibitory actions of Mas-AT and FLRFamides appear to be species-specific. These peptides exert little or no inhibition of active ion transport across the posterior midgut of the silkmoths H. cecropia and B. mori. In contrast, it has been suggested that Mas-AT stimulates JH biosynthesis by the CA of Lepidoptera because it is effective in both M. sexta and Heliothis virescens (Kataoka et al. 1989). Additional studies on more lepidopteran species will need to be conducted before it can be concluded that the activity of Mas-AT on the midgut is truly restricted to M. sexta or whether its effect on the CA is widespread in the order.

It is not known whether the peptides that affect midgut active ion transport act as hormones or paracrines. The distribution of these peptides within the insect can provide clues to the source and the means by which they act on the target tissues. In the larva, Mas-AT mRNA is present at highest levels in the frontal ganglion and terminal abdominal ganglion (Taylor et al. 1996; Bhatt and Horodyski, 1998). Because the frontal ganglion cells project their axons down the recurrent nerve and innervate muscles at the foregut–midgut boundary (Copenhaver and Taghert, 1991), these cells may be the source of Mas-AT that acts as a paracrine on the posterior midgut. The Mas-AT synthesized in the other portions of the central nervous system may be released into the hemolymph and act as a hormone on the midgut. Because the gene for F7G and F7D has not been isolated, the cells in the nervous system that contain mRNA encoding these peptides have not been identified.

In addition to the cells that contain Mas-AT mRNA, other cells of the enteric nervous system and the midgut exhibit immunoreactivity against antiserum to Mas-AT (Zitnan et al. 1993). These include intrinsic neurons of the enteric plexus and the midgut endocrine cells. The presence of Mas-AT mRNA in these cells has not yet been investigated, so it is not possible to conclude whether these cells contain Mas-AT or a peptide that cross reacts with the antiserum. If Mas-AT is indeed present in these cells, then they could be a source of peptide for a paracrine action on nearby cells, which may include the ion-transporting cells of the posterior midgut. Direct support for this hypothesis will require the demonstration that the immunoreactive material is identical to Mas-AT. Recently Kingan et al. (1997) showed that F10, F7G and F7D are present in the larval midgut, and this finding indicates a potential paracrine role for FLRFamides.

During larval growth and development, the rate of midgut active ion transport changes dramatically. Previous studies (Chamberlin, 1994; Chamberlin et al. 1997; Chamberlin and King, 1998) have shown that in vitro active ion transport activity is high during the feeding stages of the fourth and fifth instars, but declines during the molting or wandering stages. The present study confirms these earlier findings and shows that Mas-AT and two FLRFamides greatly inhibited the I_sc in the midguts of feeding fourth and fifth instars, but that the I_sc of midguts dissected from molting or wandering larvae was only minimally affected by these peptides. Although these peptides inhibit the I_sc of midguts from feeding larvae, the inhibited I_sc is not as low as the control I_sc observed in molting or wandering larvae. Therefore, even if these peptides are involved in the suppression of active ion transport during molting and metamorphosis, it is clear that other processes are involved in reducing midgut ion transport during larval–larval molting and wandering. The relative insensitivity of the midguts of molting and wandering larvae to these peptides may reflect a loss of receptors during these periods. It is also possible that during larval–larval molting, when new midgut cells are formed (Baldwin and Hakim, 1991), the epithelium contains a large population of cells that have yet to express the receptor(s).

We are grateful to Drs T. D. Kingan, S. S. Tobe and J. L. Witten for providing synthetic peptides and to Dr S. J. Kramer for the gift of antiserum for this study. We thank Drs D. H. Dean and K. S. Johnson for providing insects and Dr D. A. Schooley for technical comments. The technical assistance of Simona Aizicovici is greatly appreciated. This study was supported by the National Science Foundation (IBN-9307051 to F.M.H. and IBN-9407313 to M.E.C.) and an Ohio University Postdoctoral Fellowship.