Effect of Manduca Sexta Diuretic Hormone and Related Peptides on Isolated Malpighian Tubules of the House Cricket Acheta Domesticus (L.)

Kataoka et al. (1989) identified a 41-residue diuretic peptide (Manduca-diuretic hormone; Manduca-DH) from Manduca sexta, which resembles vertebrate neuropeptides of the corticotropin releasing factor (CRF) family. Synthetic Manduca-DH stimulated fluid excretion in vivo in head-ligated post-eclosion Pieris rapae adults and a pronounced loss of fluid through the gut and the epidermis in pre-wandering post-feeding Manduca sexta larvae. Malpighian tubules are generally the primary target for diuretic hormones, but Kataoka et al. (1989) found that Manduca-DH did not have a direct effect on isolated tubules of Manduca sexta larvae, and suggested that the peptide may release a ‘true’ diuretic hormone. This is an interesting suggestion, given that CRF stimulates the release of adrenocorticotropic hormone (ACTH) from the vertebrate pituitary, and Rafaeli et al. (1986) found that ACTH stimulates cyclic AMP production and fluid secretion by isolated Malpighian tubules of Locusta migratoria.

Malpighian tubules from adult virgin female Acheta domesticus (Linnaeus) 1–2 weeks old were used in the majority of experiments. Crickets were maintained and reared as previously described (Clifford et al. 1977; Coast, 1988) and were fed on a diet of turkey starter crumbs with water provided ad libitum. In one set of experiments, Malpighian tubules from 1–12 h post-eclosion Pieris rapae adults were used. Pieris rapae pupae were obtained from Professor L. M. Schoonhoven (The Netherlands).

Manduca-DH and a chemically more stable analogue in which methionine at residues 2 and 11 was replaced by norleucine ([Nle_2,11]-Manduca-DH) were synthesized in Texas with tBoc solid-phase peptide chemistry using methods similar to those employed by Kataoka et al. (1989). Final cleavage-deprotection of Manduca-DH was by low-high hydrofluoric acid (HF) cleavage (Tam et al. 1983). The cleavage of both peptides was with the scavenger-HF mixture (1.5 ml of anisole: 500 μl of ethanedithiol: 10 ml of liquid HF per gram of resin) for 2h at –15°C. The temperature was raised to 0°C for 45 min to ensure removal of tosyl groups from arginine side chains. Analytical HPLC immediately after preparative HPLC showed the chosen fractions to be homogeneous. Peptides were quantified by PicoTag amino acid analysis (Cohen et al. 1984) and the amino acid composition and mass, determined by fast atom bombardment mass spectrometry, were consistent with the target sequences. Samples were dried by vacuum centrifugation for transatlantic shipment.

In London, the peptides were redissolved in methanol at 10⁻⁵ mol I⁻¹ and stored at – 20°C. Prior to assay, samples were transferred to polypropylene centrifuge tubes containing 12.5 μg of gamma globulin and taken to dryness in a vacuum centrifuge. Test material was resuspended in Acheta domesticus saline (Coast, 1988) and assayed for its ability to stimulate fluid secretion and cyclic AMP production by isolated Malpighian tubules of Acheta domesticus. These bioassays have been described in detail elsewhere (Coast, 1988; Coast et al. 1991; Kay et al. 1991). Fluid secretion was measured over 40 min periods before and after the addition of assay material. Diuretic activity was determined as the increase in rate of secretion compared with controls (Ringer substitution) and expressed as a percentage of the response to a supramaximal amount of corpora cardiaca extract (5 gland pairs per 50μl; Coast and Wheeler, 1990). Cyclic AMP production was determined in groups of 4–5 tubules incubated at room temperature in Ringer containing 10⁻⁴moll⁻¹ 3-isobutyl-l-methyl xanthine (IBMX) to inhibit phosphodiesterase activity. Reactions were stopped after 30 min by the addition of ice-cold methanol and the cyclic AMP released by sonication was measured using a competitive protein binding assay (Coast et al. 1991). Additionally, Manduca-DH was assayed for its ability to stimulate fluid secretion and cyclic AMP production by isolated Malpighian tubules of Pieris rapae using identical methods to those employed with cricket tubules, but with the saline of Nicolson (1976a).

Synthetic bovine CRF, sucker fish urotensin-I, sauvagine, bovine gamma globulin and IBMX were obtained from Sigma.

All data are presented as mean±standard error (S.E.). Student’s t-test was used to determine significant differences in fluid secretion and cyclic AMP production caused by peptide addition. Values for t and degrees of freedom (d.f.) are given, with P<0.05 accepted as significant. Dose-response curves were analysed using the Maximum Likelihood Program (Numerical Algorithms Group Ltd).

Within 2 min of the addition of 10⁻⁶moll⁻¹Manduca-DH there was a significant increase in fluid secretion (t=3.011; d.f. = 13; P<0.01), and the response peaked after 15 min (Fig. 1A). The dose-response curve for the effect of Manduca-DH on fluid secretion (Fig. 2A) appeared truncated and saturated at about 60% of the maximum response achievable with extracts of corpora cardiaca. Analysis of the dose-response curve gave an EC₅₀ of 1.3×10⁻⁶±0.2×10⁻⁶moll⁻¹. In Manduca sexta, an EC₅₀ of 0.02pmol of injected peptide was cited by Kataoka et al. (1989) which, with a haemolymph volume of approximately 1ml (S. E. Reynolds, personal communication), is equivalent to about 2×10⁻¹¹ moll⁻¹. Thus, the peptide is about five orders of magnitude more potent in Manduca sexta than in Acheta domesticus. Addition of 10⁻⁴moll⁻¹ IBMX displaced the dose-response curve to the left (Fig. 2A; EC₅₀= 5.4×10⁻⁷±0.6×10⁻⁷moll⁻¹), and the diuretic activity of 10⁻⁶moll⁻¹Manduca-DH did not differ significantly (t=0.799; d.f. = 15; P=0.442) from the maximum (IBMX alone increased fluid secretion by 16±2.4 % of maximum). This suggested that the peptide stimulated adenylyl cyclase. 10⁻⁶ mol I⁻¹Manduca-DH significantly (t=19.117; d.f. = 10; P⪡0.001) increased the production of cyclic AMP by isolated Malpighian tubules (Fig. 3).

Some CRF-related peptides (Table 1) were also assayed. Kataoka et al. (1989) showed that injection of 500pmol of CRF, urotensin-I or sauvagine had no effect on fluid excretion by head-ligated post-eclosion Pieris rapae adults. At 5×10⁻⁶moll⁻¹, urotensin-I brought about a significant stimulation of fluid secretion by Acheta domesticus tubules after 2min (t=2.550; d.f. = 11; P<0.05), but this was not sustained (Fig. 1B). Urotensin-I, sauvagine and CRF all stimulated fluid secretion in a dose-dependent fashion (results not shown), but the responses did not saturate. At 10⁻⁵moll⁻¹, all three peptides increased fluid secretion by about 20% of the maximum. Ovine CRF stimulates adenylyl cyclase in the pituitary (Labrie et al. 1983) and 10⁻⁶moll⁻¹ urotensin-I, sauvagine and CRF all elicited small increases in cyclic AMP production in cricket tubules (Fig. 3). The CRF antagonist α-helical-CRF_[9–41] (Rivier et al. 1984) had no significant effect either on cyclic AMP production (Fig. 3) or on fluid secretion (t=0.241; d.f. = 11; P=0.192). Manduca-DH (10⁻⁶moll⁻¹) was assayed alone and together with CRF antagonist. Manduca-DH alone increased fluid secretion by 114±19.2 pl mm⁻¹ min⁻¹ (N=7). At 10⁻⁴moll⁻¹, α-helical-CRF_[9–41] blocked the response, the difference between the control and experimental rates being –47±53.8 pl mm⁻¹ min⁻¹(N=6). Lower concentrations of the antagonist were without effect.

The [Nle_2,11]-Manduca-DH analogue stimulated cyclic AMP production to a greater extent than did Manduca-DH (Fig. 3) and was more active in the fluid secretion assay, giving a maximum response without the addition of IBMX (EC5₀=2.8×10⁻⁷±0.8×10⁻⁷moll⁻¹; Fig. 2B). Methionine residues in Manduca- DH may, therefore, be prone to oxidation. A sample of synthetic Manduca-DH from Peninsula Laboratories Inc. gave a truncated dose-response curve in the fluid secretion assay and saturated at 69% of maximum, somewhat higher than that for the peptide synthesized by us (60 % of maximum), but lower than that for [Nle_2,11]-Manduca-DH. The material from Peninsula Laboratories was also about 10 times more potent (EC₅₀=1.6×10⁻⁷±0.4×10⁻⁷mol I⁻¹). Such differences may reflect the extent to which oxidation had occurred.

Because Manduca-DH acts directly on cricket tubules it is unlikely to be a releasing factor in the host species. In Pieris brassicae, post-eclosion diuresis is stimulated by a peptide from the brain and the retrocerebral complex working via cyclic AMP to increase tubule fluid secretion (Nicolson, 1976a,b). Manduca-DH stimulates fluid excretion in post-eclosion Pieris rapae adults and, when assayed at 10⁻⁶moll⁻¹ on Malpighian tubules taken from adult butterflies within 1–12 h of eclosion, it increased fluid secretion by 196±26.1% (t=5.793; d.f. = 11; P<0.001) and cyclic AMP production from 1.6±0.22 to 4.3±0.37 pmol per tubule (t=6.309; d.f.=8; P<0.001). Thus, a Manduca-like peptide is probably the hormone controlling post-eclosion diuresis in butterflies. In Manduca sexta, weight loss occurs largely during the 48 h preceding pupation (Baker et al. 1987). However, at this time, Manduca sexta tubules lose their ability to excrete dyes (Nijhout, 1975), and fluid secretion may be switched off prior to the end of the stadium (Nicolson, 1976b; Ryerse, 1978; Chung and Keeley, 1989). Reabsorption of fluid from the rectum of Manduca sexta larvae might be subject to neural or hormonal control (Reynolds and Bellward, 1989). Thus, Manduca-DH could act in vivo to inhibit fluid uptake from the rectum in the period before tubule secretion is switched off, so increasing faecal water loss.

Manduca-DH belongs to a family of insect diuretic peptides that are themselves members of a superfamily of peptides including CRF, sauvagine and urotensin-I (G. M. Coast, G. J. Goldsworthy, M. S. Johnson, I. Kay and C. H. Wheeler, in preparation). These peptides are involved, either directly or indirectly, in the regulation of hydromineral balance, and all appear to act via cyclic AMP. They are sufficiently similar for there to be discernible cross-reactivity in Acheta domesticus tubules. As in vertebrates, α-helical CRF_[9–41] appears to be an antagonist, and the amino terminal region of the peptides may be important for signal transduction. This could explain differences in activity between Manduca-DH and [Nle_2,11]-Manduca-DH.

The work at Birkbeck was supported by a grant from the AFRC, and at Texas A & M by an NSF grant (DCB-8918438) and the Tamu Biotechnology Support Laboratory which provided the instrumentation. We thank Alan Tyler for his technical assistance and Graham Goldsworthy for comments on a final draft of this manuscript.