The Regulation of Haemolymph Potassium Activity During Initiation and Maintenance of Diuresis in Fed Rhodnius Prolixus

Haematophagous insects differ from herbivores in that the immediate excretory pressure on them is more to eliminate excess dietary sodium than to eliminate potassium. The blood-sucking bug Rhodnius prolixus ingests at each meal a volume of blood equal to as much as ten times its unfed weight (Buxton, 1930). Immediately after the meal, it produces a sodium-rich urine at high rates in response to two synergistic diuretic hormones, 5-HT and a peptide (DH), whose release is stimulated by feeding (Maddrell et al. 1993). During steady diuresis, fluid is transported from the midgut into the haemolymph, whence it is removed at an equal rate by the activity of the four Malpighian tubules. Potassium levels in the ingested plasma are low, but almost regardless of the level of K⁺ in the midgut, the fluid transported into the haemolymph from the midgut contains only about 3 mmol l⁻¹ potassium (Farmer et al. 1981). Unlike most other blood-feeding insects, the iso-osmotic fluid secreted by the upper lengths of the Malpighian tubules contains as much potassium as sodium (Ramsay, 1952; Maddrell, 1969), so that, by itself, this process would rapidly deplete the haemolymph of potassium. Potassium (chloride), but not sodium (chloride) or water, is reabsorbed from the fluid leaving the upper tubules by the activity of the lowermost lengths of the lower Malpighian tubules, resulting in the elimination of hypo-osmotic urine (Maddrell and Phillips, 1975). The blood ingested is hypo-osmotic to that of the insect, so the secretion and subsequent reabsorption of potassium (chloride) serves to maintain both ionic and osmotic balance.

There are two obvious ways in which the regulation of haemolymph potassium concentration may be compromised by the activity of the epithelia involved in the removal of surplus fluid from the blood meal.

The first is that during initial stimulation of the Malpighian tubules, were the upper tubules to become active before the lower tubules, they would rapidly lower the potassium concentration in the haemolymph ([K⁺]_h). [K⁺]_h is about 4mmol l⁻¹ (Ramsay, 1952; Maddrell and Phillips, 1975) and, in saline containing 5mmol l⁻¹ potassium, stimulated upper tubules secrete a fluid containing about 80mmol l⁻¹ potassium (Maddrell, 1969). It follows that, if this happened in vivo and could be maintained, it would remove all potassium from the haemolymph within a minute. There is an obvious need to activate the lower tubules rapidly, so as to prevent any such loss. This might be achieved in one of two ways. The lower tubules might respond to lower concentrations of the hormones appearing in the haemolymph or they may be no more sensitive but respond more rapidly. In the present paper, we describe both the dose–response relationships and the speeds of response for the activation of upper and lower Malpighian tubules resulting from stimulation by 5-hydroxytryptamine (5-HT) and by the natural stimulant, haemolymph from recently fed Rhodnius. Such haemolymph is known to contain both 5-HT and the peptide diuretic hormone, which collaborate synergistically to stimulate activity by the upper Malpighian tubules (Maddrell et al. 1993).

The second potential source of changes in [K⁺]_h is any discrepancy between the input of potassium ions to the haemolymph by the midgut epithelium and the removal of potassium ions from the haemolymph by the combined activities of the upper and lower Malpighian tubules. In other words, the activities of all three epithelia during sustained diuresis must be coordinated. We describe the ways in which the transport activities of both the upper and lower tubules are affected by changes in potassium activity in the bathing medium. We can then estimate how successfully the combined effect of the two segments of the tubules contributes to the stability of [K⁺]_h.

The insects used were Rhodnius prolixus Stål (Hemiptera) kept at 27°C in laboratory culture at the Department of Zoology, Cambridge, UK. Experiments were performed with third-and fifth-stage (juvenile) insects.

Malpighian tubules were dissected from insects under saline using a binocular microscope and isolated in drops of saline under liquid paraffin as described earlier (Maddrell et al. 1988). The standard saline had the following composition (mmol l⁻¹): NaCl, 129; KCl, 8.6; CaCl₂, 2.0; MgCl₂, 8.5; NaHCO₃, 10.2; NaH₂PO₄, 4.3; Hepes, 8.6; glucose, 20; pH adjusted to 7.0 with NaOH. The saline contains bicarbonate to avoid the difficulty, emphasised by Thomas (1989), that cells without available bicarbonate are not able to regulate their internal pH.

The secretory activity of the upper Malpighian tubules was determined as described in detail in Maddrell et al. (1988). In this technique, isolated tubules, usually from fifth-instar insects, are placed in drops of saline under liquid paraffin. Drops of saline larger than 100 μl are required to ensure sufficient supply of oxygen for rapidly secreting tubules from fifth-instar Rhodnius (B. J. Holloway, unpublished results). The cut end of the tubule is then pulled out of the saline and wrapped around a fine glass rod pushed into the wax base of a small Petri dish. To stimulate rapid fluid secretion by the tubules, 5-hydroxytryptamine (5-HT; obtained from Sigma) was included in the saline at the appropriate concentration (Maddrell et al. 1971). A second method of stimulation was to add haemolymph taken from fifth-instar Rhodnius that had been fed within the preceding 2 h. When a range of concentrations was required, the haemolymph was appropriately diluted with saline. Not more than 5–10 μl of haemolymph can conveniently be collected from fed fifth-instar insects. So, for experiments that involved stimulation by active haemolymph, we used the smaller tubules from third-instar Rhodnius and put them in 30 μl bathing drops. The dose–response curves for tubules from third-instar Rhodnius to stimulants are quantitatively similar to those of tubules from fifth instars (Maddrell et al. 1991, 1993). The transport activity of the lower Malpighian tubules was followed using the techniques developed by Maddrell and Phillips (1975), in which the upper and lower tubule segments are bathed in separate drops of saline or haemolymph, so that they can each receive different stimulating regimes.

Potassium concentrations in drops of haemolymph or tubule fluid were measured with potassium-selective microelectrodes. Unfilamented borosilicate glass capillary tubing (GC100-15, Clark Electromedical Instruments, Reading, UK ) was washed in nitric acid for 5min, rinsed 3–5 times in deionized water, and dried for a minimum of 20min at 200°C on a hot plate. Micropipettes were pulled from the tubing using a vertical micropipette puller (Narishige, Tokyo, Japan), and replaced on the hot plate for 10min before silanization. The latter process makes the glass surface hydrophobic and facilitates filling with, and retention of, the hydrophobic neutral carrier cocktail. A drop of dimethyldichlorosilane was pipetted onto the inside of a 150mm diameter Pyrex Petri dish, which was then inverted over the micropipettes which had been placed on the hot plate. Micropipettes were removed after a minimum of 20min exposure to the silane vapour, and could be stored over silica gel for up to 2 weeks before filling.

Micropipette tips were filled with a short length of a potassium-selective neutral carrier ionophore cocktail based on valinomycin (Potassium ionophore 1, cocktail B, Fluka, Ronkonkoma, NY). The cocktail was taken up in the tip of a plastic tuberculin syringe pulled out over a low flame to a fine tip (Thomas, 1978), then injected into the shank of the micropipette. The cocktail ran to the tip of the micropipette by capillarity, and any trapped air bubbles were removed with a fine glass fibre. The micropipette was then backfilled with a 0.5mol l⁻¹ KCl solution injected through a second tuberculin syringe pulled to a fine tip. Reference electrodes were fabricated from 1mm o.d. filamented glass tubing. The tip and shank were filled with 1mol l⁻¹ sodium acetate, and the rest of the barrel was then backfilled with 1mol l⁻¹ KCl.

Electrodes were connected through chlorided silver wires to an electrometer of high input impedance (>10¹⁴ Ω; Keithley Instruments 600B) and were calibrated in solutions containing the following concentrations (in mmol l⁻¹) of KCl and NaCl respectively; 2 and 178, 8 and 142, 20 and 130, and 80 and 70. Although ion-selective electrodes measure ion activity and not concentration, data can be expressed in terms of concentrations if it is assumed that the ion activity coefficient is the same in calibration and experimental solutions. Expression of data in terms of concentrations simplifies comparisons with previous studies in which ion concentrations were measured by techniques such as flame photometry. Potassium concentrations in drops of haemolymph or secreted fluid were measured under paraffin oil by positioning potassium-selective and reference electrodes in the drop and measuring the potential change relative to that in drops of calibration solutions.

Lower Malpighian tubules in a feeding Rhodnius might be activated earlier than the upper tubules after feeding if they were more sensitive to the hormones that appear in circulation. During feeding in Rhodnius, 5-HT increases rapidly in concentration from the low values found in the resting insect (Lange et al. 1989). It reaches a concentration of about 5X10⁻⁸ mol l⁻¹ after 5min (Maddrell et al. 1991) and then steadily declines to a lower level, 1X10⁻⁸–2X10⁻⁸ mol l⁻¹, by 1h after the start of feeding (Lange et al. 1989; Maddrell et al. 1991). This suggests that stimulation by 5-HT might be important in the early stages of diuresis after feeding begins. However, 5-HT is not the only hormone that appears in circulation during the first few minutes after feeding begins, and its concentration reaches no more than 5X10⁻⁸ mol l⁻¹. This level is insufficient to cause more than partial stimulation of the upper tubules (Maddrell et al. 1971, 1993) and both 5-HT and the peptide diuretic hormone (at least) are required for full stimulation (Maddrell et al. 1993). In addition, ketanserin, which causes very effective blockade of stimulation of upper Malpighian tubules by 5-HT alone (Maddrell et al. 1991), does not affect fluid secretion by upper tubules bathed in samples of haemolymph taken at any time during the period after feeding begins (S. H. P. Maddrell, unpublished observations). So, in the experiments described below we have investigated not only the dose–response relationship for stimulation of upper and lower tubules by 5-HT alone but also the relative sensitivities of the two segments of the tubule to samples of active haemolymph taken from fed Rhodnius.

The relationship between the potassium concentration of fluid after passage through the lower Malpighian tubule and the concentration of 5-HT in a bathing drop is shown in Fig. 1. A concentration of a little above 10⁻⁸ mol l⁻¹ is sufficient to produce about 50% of the maximum reabsorption. Previous studies suggest that an appreciably higher concentration, 3X10⁻⁸–4X10⁻⁸ mol l⁻¹ (Maddrell et al. 1993), is required to produce 50% of the maximum fluid secretion rate by the upper tubule (Maddrell et al. 1993). The lower tubule is evidently more sensitive to 5-HT than is the upper tubule.

The effects of a relatively narrow range of different dilutions of haemolymph collected from recently fed fifth-instar Rhodnius on fluid secretion rates by upper tubules and potassium reabsorption by lower tubules are shown in Fig. 2A. It is important to point out that the results shown do not provide a full dose–response curve. The experimental method was to test a range of dilutions of active haemolymph on upper tubules and then treat lower tubules only with those dilutions that led to partial activation of the upper tubules. As Fig. 2A shows, in every case these dilutions also gave partial activation of the lower tubule. These results demonstrate that there were no dramatic differences in the sensitivity of upper compared to lower tubules in response to haemolymph containing diuretic hormones.

Fig. 2B shows the result of an experiment to test this finding directly. It records the responses of an upper and a lower tubule from a third-instar Rhodnius to two successively higher concentrations of active haemolymph and to a final saturating concentration of 5-HT. Both upper and lower tubules showed threshold stimulation by 10% haemolymph, partial stimulation by 14% haemolymph and maximal stimulation by 2.3X10⁻⁶ moll⁻¹5-HT. Seven other similar experiments gave the same type of result.

If the lower tubules are not activated more quickly than the upper tubules by being more sensitive to the diuretic hormones in the haemolymph, perhaps they respond more quickly to particular doses of these hormones. Stimulation by 5-HT of K⁺ reabsorption by tubules from fifth-instar insects (Fig. 3A,B) was apparent within 1min, and maximal stimulation was achieved in as little as 3min. The mean time for 50% stimulation of the lower tubule was 2.7±0.4min (N=3). In contrast, upper tubules did not begin to secrete fluid at significant rates until about 3min after addition of 5-HT to the bathing drop, and the mean time for secretion to reach 50% of the maximum rate was 4.1±0.2min (N=9). For third-instar Rhodnius, the mean times for 50% response of lower and upper tubules to 5-HT were 2.0±0.1min (N=4) and 3.9min (N=2), respectively

The initiation of fluid secretion, before drops of measurable size emerged from the open end of the tubule, could be detected by the movement of crystals of uric acid in the tubule lumen. In experiments with tubules from fifth-instar insects, uric acid movements were detected on average 2.2±0.1min (N=8) after addition of 5-HT, about 1.9min before fluid secretion rates had reached 50% of the maximum value.

Samples of active haemolymph (from recently fed fifth instars) also stimulated lower tubules more rapidly than upper tubules from third-instar insects (Fig. 4). The mean times for 50% stimulation of K⁺ reabsorption and fluid secretion were 1.6±0.1min (N=5) and 4.2±0.3min (N=5), respectively.

These results show that potassium reabsorption by the lower tubule was more rapidly stimulated by a single dose of either 5-HT or haemolymph than was fluid secretion by the upper tubule.

During post-prandial diuresis, the potassium concentration of the insect’s haemolymph depends on the rate of potassium absorption into the haemolymph from the midgut and its rate of removal from the haemolymph by the Malpighian tubules. To find out how well potassium is regulated, we measured the potassium concentration in the urine ([K⁺]_u) and haemolymph ([K⁺]_h) in insects during diuresis after the meal. We found no systematic changes with time after the meal in either of these variables. The mean values of [K⁺]_u and [K⁺]_h were 2.5±0.3mmol l⁻¹ (N=8) and 3.6±0.2mmol l⁻¹ (N=19), respectively. These figures are similar to those determined by Ramsay (1952).

Previous studies (Maddrell, 1969) have shown that fluid secretion rates are relatively unaffected by a range of bathing saline potassium concentrations ([K⁺]_b). These early studies, however, mostly used unphysiologically high potassium concentrations to examine the stability of fluid secretion rates over a wide range of conditions. Experiments with potassium concentrations closer to those found in vivo show that both the rate of fluid secretion (Fig. 5A) and the potassium concentration of the secreted fluid ([K⁺]_sf, Fig. 5B) decline considerably with decreases in potassium concentration of the bath. The decline in secretion rate was especially noticeable about 20min after the reduction in [K⁺]_b (Fig. 6). These changes in fluid secretion rate and [K⁺]_sf are additive in that the net rate of potassium movement from the upper tubule to the lower is the product of [K⁺]_sf and the rate of fluid secretion.

It was known that the bathing saline potassium levels also affect the potassium concentration of the fluid emerging from the end of the lower tubules (Maddrell and Phillips, 1976). We have now investigated this in a more systematic way. Fig. 7 shows a typical experiment and Fig. 8 shows the aggregated results of a set of experiments to assess the effects of reductions in [K⁺]_b on the reduction of potassium levels in the secreted fluid. Clearly, K⁺ reabsorption is considerably more effective when the potassium level in the bathing medium is lower.

The results demonstrate that the diuretic hormones and autonomous homeostatic control systems work in concert to regulate haemolymph potassium levels closely.

The dose–response characteristics and the time course of the responses of the upper and lower tubules to 5-HT solutions in saline and to haemolymph samples containing 5-HT and the peptide DH help to explain how it is that potassium reabsorption in the lower tubule is fully stimulated before significant rates of ion and fluid secretion by the upper Malpighian tubule are established during the initiation of diuresis. It is not clear what the effective stimulus for the tubules is in the initial stages (0–20min) because the concentrations of 5-HT are then relatively higher and rapidly changing relative to the steady state later on (20–240min; Lange et al. 1989). Since the lower tubules are sensitive to lower concentrations of 5-HT than are the upper tubules (Fig. 1), it could well be that the activation of the lower tubules before the upper tubules in diuresis owes something to their greater sensitivity to the 5-HT appearing in the haemolymph at that time. Even if this is not important – the lower tubules are no more sensitive than the upper tubules to the concentrations of 5-HT and peptide diuretic hormone that occur in the haemolymph at a later stage (Fig. 2A) – then the swifter response of the lower tubule to particular concentrations of 5-HT or active haemolymph (Figs 3, 4) will ensure that the lower tubules will be activated before the upper ones at the onset of diuresis.

These initial, hormonally controlled, events cannot regulate the haemolymph levels of potassium once diuresis is fully under way, because the haemolymph then contains concentrations of diuretic hormones significantly higher than that required to activate fully both the upper and the lower Malpighian tubules (see Fig. 2A and Maddrell et al. 1993). Instead, autonomous mechanisms to minimise potassium loss now come into play. These include decreases in upper tubule potassium and fluid secretion rates and an accompanying increase in lower tubule potassium reabsorption when bathing fluid potassium concentrations are reduced. In short, potassium homeostasis now depends upon the direct responses of both upper and lower tubules to any changes in haemolymph potassium concentration.

When K⁺ levels in the haemolymph fall, the upper tubule secretes potassium more slowly because both the fluid secretion rate and the potassium concentration in the secreted fluid decline. This adjustment occurs in two stages. Once the bathing medium K⁺ concentration falls below 12mmol l⁻¹, the potassium levels in the secreted fluid decline steadily (Fig. 5B). But when the medium K⁺ concentration falls further, to below about 6 mmol l⁻¹, the rate of fluid secretion is also reduced (Fig. 5A). So, at the normal haemolymph levels of K⁺ (3–4mmol l⁻¹, see above), both effects occur and the rate of change of potassium concentration in the secreted fluid will be a sensitive function of haemolymph potassium levels. Not only does the upper tubule secrete less potassium in response to a lower [K⁺]_b, but the lower tubule reabsorbs potassium more effectively (Figs 7, 8). The concerted actions of the upper and lower Malpighian tubules constitute an elegant negative feedback system that very effectively regulates the potassium concentration of the haemolymph.