Active Potassium Transport by the Isolated Lepidopteran Larval Midgut: Stimulation of Net Potassium Flux and Elimination of the Slower Phase Decline of the Short-Circuit Current

The larval midgut of the tobacco hornworm (Manduca sexta) contains a mechanism for the active electrogenic transport of potassium or other alkali metal ions from the blood side to the lumen side of the epithelium (Blankemeyer, 1976; Wolfersber ger, Harvey & Cioffi, 1982). The midgut potassium pump appears to be identical with the electrogenic cation pump in cells of several other insect epithelia, but quite distinct from the Na⁺/K⁺ pump of nerve, muscle, and vertebrate epithelial cells (Harvey, 1982). The insect electrogenic cation pump provides the primary driving force for fluid secretion by insect salivary glands and Malpighian tubules (Maddrell, 1977). Fluid secretion by these tissues is stimulated by diuretic hormone (Maddrell, 1962), serotonin (Berridge & Patel, 1968; Maddrell, Pilcher & Gardiner, 1969), and other related substances. The signals generated by interaction of these substances with their receptors at the cell membrane are transduced intracellularly by cyclic AMP and calcium (Berridge, 1980). A large amount of indirect evidence is consistent with the hypothesis that stimulation of fluid secretion by insect epithelia involves a direct effect of cyclic nucleotide, and possibly calcium ion, on the electrogenic cation pump (Maddrell, 1977). If the electrogenic cation pump in the midgut is the same as that in the salivary glands and Malpighian tubules, then one might expect midgut potassium transport to be stimulated by serotonin, diuretic hormone, or related substances. However, there are only two preliminary communications reporting stimulation of active potassium transport in isolated midgut and in both cases the reported stimulants were lipophylic derivatives of cyclic AMP (Smith & Moffett, 1980; Wol-fersberger & Giangiacomo, 1980).

The net flux of potassium ion from the blood side to the lumen side of the midgut epithelium can be conveniently and continuously measured as the short-circuit current (SCC). With isolated midgut, it is consistently observed that after a very brief increase the SCC decreases with time as the sum of two exponential terms (Wood & Moreton, 1978). In the central region of M. sexta midgut, the rapid phase of the decline in SCC has a half time of 7 min. Therefore, after approximately 25 min the rapid phase of the decline in the SCC is more than 90 % completed and the tissue is in a pseudo-steady-state during which the SCC decreases at a much slower rate (t_1/2 = 174 min) and there is very close agreement between and SCC (Cioffi & Harvey, 1981). Based on morphological evidence, Schultz & Jungreis (1977) concluded that the rapid phase of the decline in SCC was due to massive release of matrix plugs from midgut goblet cells and that the slower phase of the decline in SCC was due to further plug loss plus cell death. These authors proposed that massive plug release is caused by the procedures usually used in mounting the midgut on a chamber and that cell death in vitro is due to inadequacies in the usual bathing solution which promote tissue histolysis. However, they were unable to control either phase of the decline in the SCC and did not even speculate on what changes in experimental procedures or bathing solution composition might prevent the decline in SCC. Since Cioffi (1980) has presented very convincing evidence that the goblet cell matrix plugs of Schultz & Jungreis (1977) are actually apocrine secretion droplets from midgut columnar cells, one is left with the hypothesis that loss of cell potassium, preferentially into the lumen side bathing solution by unknown mechanisms, is the most likely cause for the rapid phase of the decline of the SCC (Wood, 1972) and that the slow phase of the decline in SCC is due to tissue degeneration.

In this communication we show that dibutyryl-cyclic AMP not only stimulates the pseudo-steady-state SCC of isolatedM. sexta midgut in a dose-dependent manner but also increases ⁠. We also show that blood from fourth or fifth instar M. sexta larvae contains a dialysable factor that increases the pseudo-steady-state SCC of isolated fifth instar midgut. Furthermore, it is shown that dialysable substances in fourth instar larval blood are able to eliminate the slower phase of the decay of the SCC of isolated fifth instar midgut. Some of the results in this paper have previously been presented in abstract form (Wolfersberger & Giangiacomo, 1980).

Fifth instar larvae of Manduca sexta (Lepidoptera, Sphingidae), weighing between S g and 7 g, were used for all electrophysiological experiments. These animals were purchased as third instar larvae from Carolina Biological Supply Co. and reared at 26 °C with a 16 h light/8 h dark cycle on a commercial artificial diet. The midgut was isolated as described by Cioffi & Harvey (1981) and mounted on the chamber described by Wood & Moreton (1978), so that only the central region of the epithelium covered the aperture. The average wet weight of the 0· 5 cm² portion of midgut exposed to bathing solution was 17· 8 mg. The bathing solution was composed of 32mm-KCl, 1 mm-CaCl₂, 1mm-MgCl₂, 166mm-sucrose and 5mm-TrisHCl, pH 8· 3. The bathing solution in each half of the chamber was circulated during all experiments by gas lift pumps operated by oxygen at a flow rate of approximately 5 ml min^{− 1}.

The transepithelial potential difference (PD) and short-circuit current (SCC) were measured using the three calomel electrode system described by Wood & Moreton (1978). After measurement of initial PD, each midgut was automatically kept short-circuited by means of a solid state voltage clamp and the SCC was recorded on a potentiometric chart recorder. Unidirectional fluxes of potassium, from the blood side to the lumen side, ⁠, and in the reverse direction, ⁠, were measured using ⁴²K by the methods of Cioffi & Harvey (1981)..

All chemicals were highest purity commercial products. Nucleotides were purchased from Sigma Chemical Company and P-L Biochemicals, Inc. Caffeine, cholera toxin, 5-hydroxytryptamine hydrochloride (5-HT), and theophylline were purchased from Sigma Chemical Co., ⁴²KC1 was purchased from New England Nuclear. In order to collect blood from feeding fourth or fifth instar hornworms, the larvae were suspended by their heads, their horns amputated, and the blood was allowed to drain into a small beaker stored on ice. Aliquots of the freshly collected blood were transferred to dialysis tubing sacs (MW cutoff 12 000, A. H. Thomas Co., No. 3787-D12) and dialysed against 100 volumes of distilled deionized water for 24 h at 4 °C. The solution against which the blood had been dialysed was frozen, and the water was removed by freeze drying. The freeze-dried extract was then reconstituted to the same volume as that of the original aliquot of blood using distilled deionized water.

The results of our tests of some prospective effectors of transepithelial potassium transport on the SCC of isolated hornworm midgut are summarized in Table 1. The only commercial compound tested that stimulated the SCC of isolated midgut was 1⁶,0²’ -dibutyryladenosine 3’ :5’-cyclic monophosphate (dbcAMP). Fig. 1 shows that when dbcAMP was added to the chamber, during pseudo-steady-state, there was an immediate increase in the SCC. The SCC increased to a maximum value, usually within 10 min, and then resumed a slow exponential decline. The rate of decline of the SCC in dbcAMP-stimulated midguts was the same as that in untreated, control midguts (Table 3). Essentially identical results were obtained whether dbcAMP was added to both sides or only to the blood side bathing solution. However, addition of dbcAMP only to the lumen side was without effect on pseudo-steady-state SCC, and dbcAMP had no discernible effect on SCC if it was added to the bathing solution immediately after mounting the midgut on the chamber. Fig. 2 shows that the maximum amount of stimulation of pseudo-steady-state SCC increased in a linear fashion with the logarithm of dbcAMP concentration for concentrations between 1 mm and 10mm. 0· 5 mm-dbcAMP consistently caused a small increase in SCC, but the average SCC, following addition of dbcAMP to a final concentration of 30 MM, was the same as that predicted in the absence of this compound. dbcAMP caused not only an increase in pseudo-steady-state SCC but also a corresponding increase in (Fig. 3) and a small increase in (Fig. 4). These flux experiments showed that dbcAMP stimulates proportional increases in both unidirectional fluxes so that the flux ratio remains unchanged while increases (Table 2).

Various concentrations of several other prospective effectors of insect K⁺ transport had no significant effect on the pseudo-steady-state SCC of isolated hornworm midgut (Table 1). Our first sample of cyclic AMP from P-L Biochemicals consistently inhibited the decline in short-circuit current during pseudo-steady-state. However, other samples of cyclic AMP from P-L as well as all cyclic AMP samples from Sigma had no effect on SCC. Neither cholera toxin nor Antibiotic A23187 affected the time course of the SCC. If caffeine or theophylline was added to the chamber as an aliquot from a concentrated solution which had been carefully adjusted to pH 8· 3, neither of these had a significant effect on SCC. However, direct addition of either of these compounds to the bathing solution in the chamber decreased the pH with a proportional decrease in SCC. For example, addition of theophylline to a final concentration of 20 mm caused the bathing solution pH to drop from 8· 3 to 7· 7 which was accompanied by a 20% decrease in the SCC. The effects of adding caffeine or theophylline in combination with cyclic AMP or dbcAMP were no different from the effects of addition of either cyclic nucleotide alone.

The addition of freshly collected larval blood to the solution bathing an isolated and short-circuited midgut resulted in an immediate increase in the SCC. It also caused the stirred bathing solutions to foam vigorously, which led to an imbalance in hydrostatic pressure and damage to the midgut. In order to investigate the effects of components of larval blood on midgut K⁺ transport, we prepared protein-free blood extracts by the gentle procedure described in Materials and Methods. Fig. 5 shows that addition of an extract containing the dialysable components of fifth instar larval blood to an isolated and short-circuited midgut caused an immediate increase in the pseudo-steady-state SCC. After reaching a maximum stimulated value, within 10 min after blood extract addition, the SCC declined at the same rate as that in untreated control midguts Table 3. Addition of an extract containing the dialysable components of fourth instar larval blood to an isolated and short-circuited midgut caused an immediate increase in the pseudo-steady-state SCC (Fig. 6). However, the magnitude of the increase and the time course of the decrease in SCC differed from that observed following addition of fifth instar blood extract. The average maximum increase in SCC following addition of fourth instar larval blood extract was less than 50 % of that caused by the same concentration of fifth instar larval blood extract, but the rate of decline in SCC during a 50 min interval following addition of fourth instar larval blood extract was zero (Table 3).

The results presented above clearly show that dbcAMP causes an increase in the rate of pseudo-steady-state potassium transport across the central region of the isolated tobacco hornworm midgut. The cyclic nucleotide causes proportional increases in both unidirectional fluxes of potassium ion so that while the net flux increases the flux ratio remains the same as in untreated midgut. The result that the SCC declines at the same rate in both untreated and dbcAMP treated midguts is consistent with this nucleotide stimulating potassium transport by acting directly on the pump (Berridge, 1980), rather than by stimulating a variety of metabolic processes which might be expected to promote general tissue maintenance and eliminate the slow phase of SCC decline in addition to increasing the rate of potassium transport. The failure of the divalent cation ionophore A23187 (Reed & Lardy, 1972) to stimulate the SCC of isolated midgut (Table 1) argues against the effects of dbcAMP on potassium transport being mediated by calcium ions. The observation that dbcAMP has no effect on potassium transport when added only to the lumen side of the tissue suggests that the apical plasma membrane of midgut cells is either less permeable or contains more phosphodiesterase activity than the basal membrane or both. Finally the result that dbcAMP had no discernible effect on the rapid phase of the decline of the SCC is another argument in favour of a major portion of this current being due to a process other than active transepithelial potassium transport (Cioffi & Harvey, 1981; Wood & Moreton, 1978; Wood, 1972).

The effects of dbcAMP on the midgut are similar to the effects of this compound on the salivary gland (Berridge, 1970, 1977, 1980), even to the extent that the linear portion of the dose versus response curve occurs over the same concentration range in both tissues. However, most of our results with other prospective effectors of active potassium transport are quite different from those obtained with other inset epithelia. Cyclic AMP, which stimulates the potassium pump in salivary gland and Malpighian tubules (Maddrell, 1980), as well as cholera toxin, which is reported to increase intracellular cyclic AMP concentration (Finkelstein, 1975), had no effect on the potassium pump in the midgut. Methyl xanthines, which are reported to increase intracellular cyclic AMP concentration by inhibiting phosphodiesterase (Butcher & Sutherland, 1962), stimulate salivary gland fluid secretion both alone and synergistically with cyclic nucleotides (Berridge, 1970) but have no effect on midgut potassium transport. In this regard it is interesting to note that theophylline does not stimulate secretion by Malpighian tubules oiRhodnius (Maddrell, Pilcher & Gardiner, 1971). Very low concentrations of serotonin stimulate secretion by salivary gland and Malpighian tubules but even millimolar concentrations of this hormone have no effect on midgut potassium transport. All of these negative results are most simply explained by making two reasonable assumptions. (1) The midgut is essentially impermeable to cyclic AMP and methyl xanthines. (2) The plasma membranes of midgut cells lack receptors for 5-hydroxytryptamine and cholera toxin. If these assumptions are correct, there is no reason to deviate from the basic hypothesis, with which our dbcAMP results are completely compatible, that the electrogenic cation pump in the midgut is the same as that responsible for massive Na⁺ or K⁺ movement in other insect epithelia.

One result of our studies which appears difficult to reconcile with the results of others is our finding that caffeine had no effect on SCC. Smith & Moffett (1980) found that caffeine, at concentrations of 1 mm to 10 mm, inhibited the PD and SCC of isolated tobacco hornworm midgut. Since we assume that these workers were careful to avoid caffeine-induced changes in bathing solution pH, we are left with the possibility that the effects of caffeine on the SCC of the isolated posterior region of hornworm midgut, the region used by Smith & Moffett (D. F. Moffett, personal communication), are different from the effects of caffeine on the central region of hornworm midgut. This explanation is plausible because not only are the potassium transporting goblet cells structurally quite different in these two midgut regions but these two regions of midgut also differ in transport parameters such as potassium pool size, flux ratio, and SCC decay profile (Cioffi & Harvey, 1981).

Comparison of Fig. 5 with Fig. 1 reveals that the effect on pseudo-steady-state SCC of the addition of a dialysable extract of fifth instar larval blood is similar to the effect of addition of dibutyryl-cyclic AMP. In both cases the SCC increases for about 10 min and then declines with the usual exponential time course (Table 3). The final concentration of dialysable blood components required to produce this effect is so low that one can eliminate various ions and metabolites in the extract as causing the effect. Therefore one is led to conclude that larval blood contains a water-soluble transportstimulating factor. Since attempts to isolate the active substance from blood extracts have been unsuccessful, our knowledge of the chemistry of this factor is quite limited. On the basis of its passing through the dialysis membrane one may conclude that it is a water soluble compound with a molecular weight of less than 12000 and, based on the failure of 5-HT to stimulate midgut potassium transport, one may conclude that this factor is not closely related in structure to serotonin. The similarity of the tissue response to blood extract and dbcAMP suggests that intracellular transmission of the signal, generated by interaction of this factor with its receptor on the basal plasma membrane to the potassium pump on the apical plasma membrane, involves cAMP.

Comparison of Fig. 6 with Figs 1 and 5 reveals both similarities and differences between the effects of fourth instar larval blood extract and fifth instar larval blood extract or dbcAMP on the pseudo-steady-state SCC of isolated midgut. In all three cases, the SCC begins to increase immediately after addition of stimulant and reaches a maximum within approximately 10 min. However, the SCC of midguts that have been stimulated by dbcAMP or fifth instar blood extract declines in the usual fashion after reaching the new stimulated level, while the SCC of midguts which have been stimulated by fourth instar blood extract remains steady at the stimulated level and does not decline for at least 50min (Table 3). These results demonstrate that fourth instar blood contains not only transport-stimulating factors present in fifth instar blood but also other factors not present in fifth instar blood which delay or prevent degeneration of the isolated midgut. Addition of these factors to the bathing medium will allow one to work with an isolated midgut in a true steady-state which will be extremely advantageous for a variety of studies concerning active ion transport by this tissue.

We are grateful to Prof. W. R. Harvey for his critical reading of the manuscript and helpful suggestions. This research was supported in part by a research grant (AI-09503) from the National Institute of Allergy and Infectious Diseases, U.S. Public Health Service.