Gramicidin Switches Transport in Insect Epithelia From Potassium to Sodium

Insect epithelia such as the salivary glands of blowflies and the Malpighian tubules of herbivorous insects secrete fluids containing KCl as the predominant salt (Berridge et al. 1976; Ramsay, 1953). In contrast, the Malpighian tubules of most blood feeders, including mosquitoes and tsetse flies, secrete fluids containing predominantly NaCl (Gee, 1976; Williams and Beyenbach, 1983). As a further variant, the upper Malpighian tubules of Rhodnius prolixus secrete a fluid containing approximately equimolar NaCl and KCl after the insect has taken a blood meal (Maddrell, 1969).

These apparently diverse patterns of ion transport can be accommodated in a single model in which an apical V-ATPase plus cation/H⁺ antiporter(s) can drive transport of Na (Cl) and/or K (Cl) into the lumen. Which of Na⁺ or K⁺ is actively transported depends on the nature of the passive ion-permeable channels and transporters on the basolateral plasma membranes (Maddrell and O’Donnell, 1992). If this model is correct, it ought to be possible to change substantially the relative concentrations of Na⁺ and K⁺ in the fluid secreted by insect epithelia by altering the relative permeability of the basolateral membranes.

In the present paper, we describe the dramatic effects of treating two well-studied insect transporting epithelia with gramicidin D, an ionophore known to confer on membranes very high permeabilities to Na⁺ and K⁺ (see Wills, 1981). Such treatment effectively allows these ions equal ease of access to the intracellular compartment. We chose to investigate the effects of gramicidin D on the salivary glands of Calliphora erythrocephala, which essentially transport iso-osmotic KCl solution (Berridge et al. 1976) and the upper Malpighian tubules of Rhodnius prolixus, whose secreted fluid contains nearly equal concentrations of Na⁺ and K⁺ (Maddrell, 1969). In Calliphora salivary glands, the basal plasma membrane preferentially allows entry of K⁺ and Cl⁻ (Berridge et al. 1976), whereas in Rhodnius, basal entry of ions depends instead on the furosemide-sensitive Na⁺/K⁺/2Cl⁻ cotransporter (O’Donnell and Maddrell, 1984).

Tubules and salivary glands were isolated as described previously (Maddrell et al. 1988) and bathed either in standard saline (Maddrell et al. 1993) containing 146mmol l⁻¹ Na⁺ and 8.6mmol l⁻¹ K⁺, or in a saline containing 77mmol l⁻¹ Na⁺ and 77mmol l⁻¹ K⁺. In all cases stimulation of fluid transport was achieved by including 10⁻⁶ moll⁻¹ 5-hydroxytryptamine (5-HT) in the saline (Berridge and Patel, 1968; Maddrell et al. 1971, 1991).

Fig. 1 shows the effects of gramicidin D on the rate of fluid secretion and [K⁺] of the secreted fluid of an upper Malpighian tubule of a fifth-instar Rhodnius. The tubule was bathed in standard saline containing 146mmol l⁻¹ Na⁺ and 8.6mmol l⁻¹ K⁺ (Maddrell et al. 1993) to which had been added 3.6 μg ml⁻¹ gramicidin D. The transient stimulation of fluid secretion is the result of increased oxygen availability and stirring of the bathing droplet which occurs whenever a solution is added to the bathing droplet (Maddrell et al. 1988). The potassium concentration of the secreted fluid rapidly dropped from 96mmol l⁻¹ to 5mmol l⁻¹ in response to 3.6 μg ml⁻¹ gramicidin D (Fig. 1A). The response is even more pronounced with higher concentrations; on exposure to 125 μg ml⁻¹ gramicidin D, for example, the potassium concentration of the secreted fluid dropped from 88mmol l⁻¹ to 0.2mmol l⁻¹. For gramicidin concentrations of 1.2 μg ml⁻¹ and above, the K⁺ concentration dropped on average to 3.3±0.9mmol l⁻¹ (mean ± S.E., N=5; see Fig. 3). These results show that addition of gramicidin converts Rhodnius’ Malpighian tubule, an epithelium that normally secretes equimolar NaCl and KCl, into one which no longer secretes significant amounts of potassium.

Experiments with isolated salivary glands of Calliphora indicate that this epithelium, which normally secretes potassium, can be converted into a sodium-secreting epithelium by addition of gramicidin. Fig. 2 shows the effects adding 22.5 μg ml⁻¹ gramicidin D to the saline bathing an isolated salivary gland; the potassium concentration of the secreted fluid dropped from 125 mmoll⁻¹ to 18 mmol l⁻¹. On average, the potassium concentration of the secreted fluid changed from 150.1±10.9 mmol l⁻¹ to 23.3±1.4 mmol l⁻¹ (means ± S.E., N=7) after addition of 22.5 μg ml⁻¹ gramicidin. The sodium concentration in the secreted fluid after treatment was 152.5±3.1 mmol l⁻¹ (mean ± S.E., N=4).

Both the Malpighian tubules of Rhodnius and the salivary glands of Calliphora are extremely sensitive to gramicidin D. Although the dose of the drug required for maximal effect on mammalian epithelia such as the descending colon and the urinary bladder of rabbits is 0.13mgml⁻¹ (Wills, 1981), potassium concentrations in the fluid secreted by Malpighian tubules are decreased substantially by concentrations more than 1000-fold lower (Fig. 3).

These results show that insect epithelia that normally secrete fluid containing high concentrations of potassium can be converted to epithelia that secrete sodium. This conversion depends on the fact that the treatment with gramicidin allows sodium and potassium more nearly equal access to the intracellular environment (see Wills, 1981). However, in all the experiments described above, the bathing saline was essentially Na⁺-rich, reflecting the normal ionic composition of the haemolymph, so that the apical transporters might in these cases have been presented with fluid containing more Na⁺ than K⁺ (depending on the relative effectiveness of the gramicidin pores and the permeability of the plasma membrane).

It has been argued that insect epithelia such as Malpighian tubules can regulate their intracellular environment to be K⁺-rich and Na⁺-poor and at the same time secrete fluid containing a predominance either of K⁺ or of Na⁺ solely through the activity of the apical ion transporter(s) (Maddrell, 1978). This ability was thought to depend on an ability actively to transport both cations from cell to lumen but with a preference for Na⁺. Rhodnius Malpighian tubules secrete Na⁺ more rapidly than K⁺ at all intracellular concentrations of Na⁺ higher than 20mmol l⁻¹, when [K⁺]i is more than 100mmol l⁻¹ (Maddrell, 1978). If this hypothesis is correct and applies to other insect epithelia, then treatment with gramicidin in saline containing equal concentrations of Na⁺ and K⁺ should lead to the secretion of fluid containing higher concentrations of Na⁺ than K⁺. In fact, addition of gramicidin D at 20.8 μg ml⁻¹ to saline containing equal concentrations of Na⁺ and K⁺ (77mmol l⁻¹) reduced the K⁺ concentration of the fluid secreted by Rhodnius tubules from 91.5±2.8mmol l⁻¹ to 37.6±1.9mmol l⁻¹ (mean ± S.E., N=4). The K⁺ concentration of the fluid secreted by Calliphora salivary glands dropped from 123.4±5.4mmol l⁻¹, before addition of 45 μg ml⁻¹ gramicidin D, to 76.7±7.6mmol l⁻¹ (mean ± S.E., N=7) by 25–35min after addition of the drug.

What these results suggest, then, is that indeed the apical ion transporters [V-type H+-ATPase plus cation/H+ antiporter(s)] of Rhodnius tubules show a preference for Na⁺ over K⁺, but that those of Calliphora salivary glands show an ability to transport Na⁺ and K⁺ with equal effectiveness. In both epithelia, nonetheless, transport under all normal conditions will maintain the cell interior as K⁺-rich and Na⁺-poor while allowing transepithelial transport of either Na⁺-or K⁺-rich fluids. In Rhodnius this result seems to require that the apical cation/H⁺ antiporter has an absolute preference for Na⁺ over K⁺, presumably to match the rapid basal entry of Na⁺via the Na⁺/K⁺/2Cl⁻ cotransporter. In Calliphora, Na⁺ enters the cell across the basal side more slowly than K⁺ (Berridge et al. 1976) so that the equal ability of the apical cation/H⁺ antiporters to transfer Na⁺ or K⁺ into the lumen will still result in intracellular Na⁺ levels being kept low. It will be of great interest to discover whether in these epithelia the difference in apical transport of Na⁺ and K⁺ is achieved by one cation/H⁺ antiporter with variable Na⁺/K⁺ preference or by two distinct Na⁺/H⁺ and K⁺/H⁺ antiporters.

These considerations go a long way to explain how it is that ouabain, the potent inhibitor of Na⁺/K⁺ ATPase, only rarely affects the rate or composition of fluid secretion by insect epithelia (Anstee and Bell, 1975; Maddrell and Overton, 1988). Epithelial fluid transport in insects is often very rapid (in extreme cases each cell can transfer its own volume of fluid in 15s, Maddrell, 1991). It makes good sense that the apical ion transporters driving fluid secretion automatically regulate the intracellular cation levels rather than regulation being provided by the activity of the basal Na⁺/K⁺ ATPase, as this would need to be present at very high concentrations and, in Na⁺-transporting epithelia, would transfer important ions in a direction opposite to that required.