The Fate of Calcium in The Diet of Rhodnius Prolixus: Storage in Concretion Bodies in the Malpighian Tubules

Young (non-adult) stages of Rhodnius prolixus take very large meals of blood of about 10–12 times their body weight (Buxton, 1930). Since the mammalian or avian blood on which they usually feed contains at least 1.5 mmol l⁻¹ calcium (Altman and Dittmer, 1974), fed Rhodnius seem likely to face a dietary excess of calcium considerably greater than that faced by most other insects and, indeed, most other animals. This paper examines the question of how Rhodnius deals with the excess calcium. It turns out that the largest part is deposited in the cells of the upper Malpighian tubules, some is retained in the haemolymph and relatively little is eliminated from the body. Conceivably, the calcium deposits may act as a store and be drawn upon when calcium is in short supply, but their normal function is more likely to be to immobilise excess calcium and so prevent it from interfering with such vital functions as water conservation.

The insects used were Rhodnius prolixus Stål (Hemiptera) kept in laboratory culture. They came from three different stocks held (1) in the Department of Zoology, Cambridge, UK, (2) at Instituto Venezolano de Investigaciones Cientificas (IVIC), Caracas, Venezuela, and (3) at the Centro de Investigaciones Biomédicas, Universidad de Carabobo, (Nucleo Aragua) at Maracay, Venezuela. Experiments were performed with fourth- and fifth-stage (juvenile) insects from the Cambridge stock, except where mentioned.

Insects were fed artificially on saline containing ⁴⁵Ca²⁺, with 2 mmol l⁻¹ ATP as a phagostimulant (Friend and Smith, 1975), through a stretched Parafilm membrane with the saline warmed to 38°C in a water bath. The meal size was assessed by weighing the insects before and after feeding.

To inject ⁴⁵Ca²⁺ in saline into the haemolymph of an insect, one of its metathoracic legs was cut half way along the femur. A 10 μl Hamilton syringe containing ⁴⁵Ca²⁺ in saline was pushed into the cut end of the leg and waxed in position with a beeswax-resin mixture applied with an electrically heated wire loop. After the requisite amount of fluid had been injected, the leg was held tightly at its base with fine forceps while the needle was withdrawn and the cut end of the leg sealed with wax. Haemolymph samples were taken later by cutting a metathoracic and/or prothoracic leg and collecting the haemolymph that emerged in a drawn-out Pasteur pipette while gently squeezing the insect. The haemolymph was blown out under liquid paraffin to measure its volume from the diameter of the spherical drop. The cut end of the leg was then resealed with wax.

⁴⁵Ca²⁺ accumulated in Malpighian tubules was routinely determined after washing them for at least 10 min in tracer-free 5-HT-containing saline, so as to remove ⁴⁵Ca²⁺ from the surface of the tubule and from the lumen. Each tubule was then osmotically disrupted in a 10 μl drop of distilled water. The tubule plus drop was then transferred to liquid scintillant (Aquasol) in a vial for counting.

X-ray diffraction was performed in the Molecular Structure Laboratory at IVIC using a rotating anode X-ray generator.

For electron microscopy, tubules were immersed overnight at 4°C in a fixative solution consisting of 2.4% glutaraldehyde in 0.1 mol l⁻¹ phosphate buffer, pH7.4, plus 6% sucrose. Following washing in buffer solution the tissues were post-fixed in 0.5 % aqueous osmium tetroxide solution for 30 min. Specimens were then briefly washed in buffer followed by a further three washings in distilled water, en bloc stained with 2 % aqueous uranyl acetate for 1 h, dehydrated through an ascending series of ethanols, 5min each, to 100% ethanol, 3×20min, and thence to propylene oxide, 3x20min. The tubules were finally embedded in Araldite CY212. Ultrathin sections, 50–70 nm thick, were cut on a Reichert Ultracut microtome, mounted on copper grids, stained with uranyl acetate and lead citrate and examined in a Philips EM 420 at 80 kV.

Analyses of calcium in Malpighian tubules or in samples of haemolymph were made using flame photometry after sonicating the tubules or drops of haemolymph in 25 μ1 of distilled water in small plastic vials. 50μl of 0.5 mol l⁻¹ HC1 was then added, and the volume made up to 150 μl with distilled water. Calcium measurements were made by diluting each vial’s contents for use with an atomic absorption flame spectrophotometer in the Chemistry Department at IVIC. The results are given as mmol of calcium per litre of fluid for samples of haemolymph and as mmol of calcium per litre of volume of the cells for analyses of the calcium content of tubules.

The activity of calcium in small samples of haemolymph (15–20 μl) was measured using a calcium-selective membrane electrode, built using a membrane obtained from W. Müller, Zurich, Switzerland (Affolter and Sigel, 1979).

All measurements are given as mean± s.E. (N). Experiments were performed at room temperature, 22–24°C.

Absorption of the products of digestion in insects usually occurs largely by diffusion through the permeable epithelia of the midgut; this process is accelerated by uptake of fluid, which concentrates the solutes in the lumen of the gut (Treherne, 1967). As a result, even unwelcome or toxic materials, such as sulphate ions, magnesium ions and nicotine, from the diet all readily cross into the haemolymph. The influence of such materials is reduced both by the protective role of the perineurium, the epithelial blood-brain barrier of insects (Treherne and Pichon, 1972), and by the ability of the Malpighian tubules to remove those substances from the haemolymph by active transport (Maddrell and Phillips, 1975a; Phillips and Maddrell, 1974; Maddrell and Gardiner, 1976). It was natural, then, first to examine the ability of the Malpighian tubules of Rhodnius to excrete calcium ions.

The upper, fluid-secreting parts of the Malpighian tubules of Rhodnius have a low permeability to substances that passively cross the wall via the paracellular route and do not enter the cells of the tubule (Maddrell, 1980; O’Donnell et al. 1984; Whittembury et al. 1986). To determine whether calcium ions cross the epithelium other than by passive means, we measured ⁴⁵Ca²⁺ levels in the fluid secreted by upper Malpighian tubules isolated from fifth-stage insects and bathed in standard saline containing 10⁻⁵moll⁻¹ 5-HT and ⁴⁵Ca²⁺ at 25 000ctsmin⁻¹μl^_1. Because even slight damage can greatly increase tubule permeability, we first similarly measured the permeability of the tubules to [¹⁴C]sucrose; sucrose is known to cross the tubule wall entirely paracellularly (O’Donnell et al. 1984; Whittembury et al. 1986). The 10 least permeable tubules of 50 tubules tested had a mean permeability to [¹⁴C]sucrose of 0.033 ±0.004nl min⁻¹mm⁻² (5.5×10⁻⁸cms⁻¹) but a mean permeability to ⁴⁵Ca²⁺ of only 0.020±0.002nlmin⁻¹mm⁻² (3.3×10⁻⁸cm s⁻¹). This means that fluid secreted by a tubule at, say, 70nlmin⁻¹ contained on average only 0.2% of the concentration of calcium in the bathing fluid. Clearly, calcium does not receive an accelerated passage across the tubule wall. Indeed, the tubules’ permeability to [¹⁴C]sucrose is consistently 1.6 times as high as the permeability to⁴⁵Ca²⁺ (Fig. 1).

Although ⁴⁵Ca²⁺ only slowly crosses the tubule wall into the lumen, it is, however, unlike sucrose, rapidly accumulated in the tubule cells. For example, four tubules from two fifth-stage insects taken 7 days after moulting were exposed to [¹⁴C]sucrose solution for 2h then washed twice for 5-min in tracer-free saline. The tubules then contained radioactivity equivalent to only 1.74±0.20nl of the original solution. The other four tubules from the same insects exposed instead to ⁴⁵Ca²⁺-containing solution for 2h contained radioactivity equivalent to 51.0±3.6nl of the original bathing saline, even after two 5-min washes in tracer- free saline. As described below, tubules exposed for longer periods accumulate proportionately more calcium.

Properties of Rhodnius Malpighian tubules are often different in the upper, fluid-secreting part of the tubule and in the lower part of the tubule, which is concerned with active transport of uric acid and reabsorption of KC1 (O’Donnell et al. 1983; Maddrell and Phillips, 1975b). Active transport of nicotine and other alkaloids, however, goes on equally in both upper and lower segments of the tubule (Maddrell and Gardiner, 1976). We investigated whether ⁴⁵Ca²⁺ is accumulated in the lower tubule as it is in the upper tubule. We exposed entire Malpighian tubules (i.e. upper and lower regions) from fifth-stage insects taken 10 days after moulting to saline containing ⁴⁵Ca²⁺ and 5-HT for periods of up to 7 h. The tubules were then washed for 20 min in non-radioactive saline with 5-HT. The lower tubules were cut away from the upper parts and the separate pieces counted. The results showed that the lower tubules contained very few counts, equivalent on average to 0.6±0.1nl (N=15) of bathing fluid, while the upper tubules contained radioactivity equivalent to, on average, 9.9±1.2nl (N=4) after 25 min, 21.5±2.3nl (N=4) after 60min, 34.4±1.9nl (N=4) after 140min and 76.2±3.9nl (N=4) after 420 min. Even more dramatic differences came from experiments with insects injected with ⁴⁵Ca²⁺. Tubules taken 8–12 days later contained, on average, ⁴⁵Ca²⁺ equivalent to 5563±594 nl (N=15) of the haemolymph in the cells of the upper tubule but ⁴⁵Ca²⁺ equivalent to only 1.2±0.3nl (N =8) of the haemolymph in the cells of the lower tubule, more than 4000 times less.

Evidently, significant accumulation of ⁴⁵Ca²⁺ is entirely confined to the upper Malpighian tubules. Their content of ⁴⁵Ca²⁺ appears to increase steadily with time. All further experiments were carried out using the upper tubules.

A total of 87 tubules from fifth-stage insects taken 10 days after moulting were incubated in ⁴⁵Ca²⁺-containing saline for periods of up to 400 min and their ⁴⁵Ca²⁺ content then determined. The results are shown as the lower line in Fig. 2. They show a continuous accumulation of ⁴⁵Ca²⁺, though at a steadily declining rate. Preliminary experiments had suggested that tubules taken from fed insects accumulated ⁴⁵Ca²⁺ more rapidly; many transport properties of the Malpighian tubules of Rhodnius are stimulated in the days after a meal (Maddrell and Gardiner, 1975; O’Donnell et al. 1983). So we repeated the experiments with fourth-stage insects fed 7 days previously (i.e. they were taken 2–3 days before moulting). Although these tubules are smaller than those from fifth-stage insects (Maddrell et al. 1988), they accumulate ⁴⁵Ca²⁺ nearly twice as fast (upper line in Fig. 2).

To test accumulation beyond 7h, recently fed fifth-instar insects were injected with 3 × 10⁶cts min^{−1 4S}Ca²⁺ in 5 μl of saline through a metathoracic leg. After the required period, samples of haemolymph were taken, and the upper tubules were removed and washed for 20 min in tracer-free 5-HT-containing saline. The ⁴⁵Ca²⁺ contents of the haemolymph samples and tubules were then determined. The results are given in Fig. 3 and show that ⁴⁵Ca²⁺ is taken up at a more or less uniform rate for at least 12 days. At the end of this time, each tubule contained ⁴⁵Ca²⁺ equivalent to that found in 6300 nl of haemolymph and more than half of the injected ⁴⁵Ca²⁺ had been sequestered in the cells of the upper tubules.

The total volume of the cells (excluding the extracellular spaces) in a single upper tubule was determined to be 100 nl by measurement of the overall tubule dimensions under a binocular microscope fitted with an ocular micrometer and from measurements made on a series of electron micrographs of tubules cut in cross section (courtesy of Dr Helen Skaer). The concentration of calcium in Rhodnius haemolymph measured by flame photometry was 8.13±0.20 mmol I⁻¹ of fluid (N =8). It follows that, 12 days after injection of ⁴⁵Ca²⁺, the calcium concentration in the cells is at least 60 times higher than that in the haemolymph, that is close to 500 mmol l⁻¹ of cells. As a direct test of this, we measured by flame photometry the calcium content of 30 tubules from 11 unfed insects. They contained an average of 410±53mmoll⁻¹ of cells (N=11).

To investigate how feeding affects the ability of tubules to accumulate calcium, we exposed tubules from fed and control (unfed) fifth-stage insects for 90 min to saline containing ⁴⁵Ca²⁺ and 5-HT. The tubules were washed for 20 min in tracer- free 5-HT-containing saline, before their radioactive content was determined. The results are shown in Fig. 4. By the second day after feeding, the rate of ⁴⁵Ca²⁺ accumulation was more than three times higher than in the unfed controls; a significant difference persisted over at least the next 20 days (data for the period beyond 9 days are not shown).

We have seen that calcium levels in upper Malpighian tubules may reach 0.4 mol l⁻¹. Thus, it is possible that the greater part of the calcium in the diet may be retained in the tubule cells. On reaching the fifth stage, a Rhodnius has in its lifetime consumed a maximum of 300 μl of blood (Gardiner and Maddrell, 1972). Since mammalian blood contains 1.5mmoll⁻¹ Ca²⁺, this represents, at most, 450 nmol of calcium. The haemolymph volume of a fifth-stage Rhodnius is about 15 μl (Maddrell and Gardiner, 1980); at 8 mmol l⁻¹ its calcium content is thus 120 nmol. Four upper tubules, each of 100 nl and with 0.4 mol l⁻¹ calcium in them together contain at least 160nmol. So, tubules and haemolymph may together hold at least 60% of the lifetime calcium load. The figure may be considerably higher than this as very few insects succeed in taking a full meal at every possible opportunity.

To test the possibility that calcium is retained in the body and not eliminated, we measured ⁴⁵Ca²⁺ elimination from five fifth-stage insects fed 2 days previously and injected through a leg with 3.3 ×10⁶ ctsmin^{−1 45}Ca²⁺ in 5 μl of saline. Each insect was kept in an individual glass vial on a small piece of filter paper. Excreted material was absorbed by the paper and was removed in a daily check and its ⁴⁵Ca²⁺ content measured. The results (Fig. 5) showed that ⁴⁵Ca²⁺ is steadily eliminated, but at a very slow rate. By 15 days after injection, only 0.54±0.09% (N=5) of the injected ⁴⁵Ca²⁺ had been excreted.

As a further test, we fed fourth-stage insects on saline containing 10⁵ ctsmin⁻¹μl^{−1 45}Ca²⁺ until they had ingested about 15 μl, when they were allowed to feed to completion on rabbit blood (a further 100μl of fluid). The elimination of ⁴⁵Ca²⁺ was followed, this time for a period of 21 days (Fig. 6). The average amount eliminated in this time was 73404±14921 ctsmin⁻¹ (N=7), about 4% of that ingested.

There is thus a noticeably higher rate of elimination of ⁴⁵Ca²⁺ ingested by fourth-stage insects than from fed fifth-stage insects injected with ⁴⁵Ca²⁺. It is instructive to consider why this might be so. The fifth-stage insects were each injected with 3.3 × 10⁶ ctsmin⁻¹⁴⁵Ca²⁺, while the fourth-stage insects were each fed a meal containing about 1.5 × 10⁶ctsmin⁻¹⁴⁵Ca²⁺; it follows that both had ⁴⁵Ca²⁺ at an overall specific activity of about 5000ctsmin⁻¹ nmol⁻¹ calcium. However, ⁴⁵Ca²⁺ injected directly into the haemolymph of fifth-stage insects would initially be present there at a specific activity some five times higher than this. In addition, the Malpighian tubules of these insects would be capable of uptake of ⁴⁵Ca²⁺ at a rate about four times higher than at the time of feeding 2 days earlier (Fig. 4). Because of these two factors, ⁴⁵Ca²⁺ uptake into the Malpighian tubules of this group of insects would initially be very rapid, leaving less available to be eliminated from the body. In contrast, ⁴⁵Ca²⁺ ingested by fourth-stage insects would mix with calcium in both meal and haemolymph before reaching the tubules and would initially be taken up only slowly by them, allowing a more rapid loss of ⁴⁵Ca²⁺ by excretion.

Both sets of results agree in showing that much the greater part of the calcium in the diet is retained in the body.

The question naturally arises as to what might be the function of the calcium deposits in the Malpighian tubules. We shall see below that changes in diet can greatly affect the quantity of calcium in the tubules. The deposited calcium might, therefore, act as a store through which the haemolymph level can be maintained.

To investigate possible unloading of ⁴⁵Ca²⁺ from upper Malpighian tubules, we took tubules from fifth-instar insects injected with ⁴⁵Ca²⁺ 9 days earlier and, after dissection (when they are immersed in saline for 5 min), they were washed for 10 min in standard saline containing 5-HT, and then were each put in a separate 100 μl drop of fresh 5-HT-containing saline. The secreted fluid was collected at 20 min intervals for 80 min, the tubule was removed and the numbers of counts remaining in it, in the bathing drop and appearing in the four secreted drops were measured. The results showed that of an average of 230000±10200ctsmin⁻¹ (N=8) initially present in each tubule, 17.5±1.1 % was lost from the tubule in 80min, 15.7±0.9 % to the bathing drop but only 1.6±0.1 % to the secreted fluid. The percentages of counts lost in the secreted fluid in the four successive collection periods were, respectively, 0.47±0.05%, 0.41 ±0.03%, 0.38±0.03% and 0.37±0.03 %. It is clear that counts are continuously lost to the secreted fluid, but that many more counts are lost through the basal cell membrane than through the luminal one.

⁴³Ca²⁺ injected into the animal for periods shorter than 9 days, and therefore in contact with the tubule walls for periods shorter than 9 days, is lost more rapidly. Up to 48h following injection, 24.9±1.0% (N=48) of the intracellular ⁴⁵Ca²⁺ is lost (to both bath and secreted fluid) in l h; at 72 h after injection, 18.3±1.1 % (N=8) is lost in 1 h and, as we have seen, by 9 days after injection, 17.5 % is lost in 80 min, equivalent to perhaps 13 % in 1 h. In a different set of experiments, ⁴⁵Ca²⁺ was also incorporated into the tubules (by injection of ⁴⁵Ca²⁺ into the haemolymph) and the percentage of the injected ⁴⁵Ca²⁺ appearing in the secreted fluid was measured. 24h after injection, 1.39±0.11 % (N=16) was lost into the secreted fluid in 20min; after 52–76h of loading, 0.36±0.02% (N=7) was lost in 20min; and 12 days after injection, only 0.18±0.04 % (N= 10) was lost to the secreted fluid in 20 min. Both sets of experiments show that ⁴⁵Ca²⁺ in the tubule walls can be unloaded, that unloading is preferentially to the haemolymph and they also suggest that ⁴⁵Ca²⁺ becomes somewhat less accessible to the bathing saline with longer times of contact between haemolymph and tubule.

The upper Malpighian tubules in Rhodnius, unlike the lower regions, contain so-called concretion bodies (Wigglesworth and Salpeter, 1962), (sometimes known as spherical bodies, concretions, granular bodies, dense bodies or mineralized granules), well-known in a range of invertebrates (Brown, 1982) and which, in other insects, are thought to contain metals such as zinc, iron and calcium (Sohal, 1974). They are thus promising candidates for the intracellular deposits of calcium in Rhodnius. Confirmation that this is so comes from two sources.

Rhodnius prolixus in the IVIC stock are routinely fed through a thin membrane on heparinised sheep blood from a stock of sheep whose red cells contain an unusually low concentration of potassium: 20 mmol l⁻¹ in whole blood compared with 60 mmol l⁻¹ present in rabbit whole blood. On dissection of fifth-stage IVIC insects, it is at once clear that the upper tubules in these insects are almost completely transparent, unlike the greyish-white opaque appearance of this part of the tubule in insects from the Cambridge stock. Under the electron microscope, the tubules from these two stocks appear very different. The IVIC insects still have numerous concretion bodies, but all of them are very small compared with the array of large concretion bodies in the Cambridge insects (Fig. 7). The calcium content of tubules from these insects, determined by flame photometry, was only 21±3mmoll⁻¹ (N= 11). By a fortunate chance, a few insects were available from another Venezuelan stock of Rhodnius (from Maracay). This stock of insects is fed on chicken blood [which is richer in calcium, 4.5 mmol l⁻¹ in laying hens (Altman and Dittmer, 1974)] and have cells in the upper tubule that are densely white. Here the density of concretion bodies is several times higher than in the Cambridge stock (Fig. 7). In two tubules the average calcium content was 1190 mmol l⁻¹.

We injected several IVIC and Cambridge fifth-stage insects with ⁴⁵Ca²⁺ in saline as before. After 12 h, four upper tubules from a Cambridge insect contained an average of 40 200 ± 660cts min⁻¹ tubule⁻¹, while four upper tubules from an insect from the IVIC stock had accumulated only 9±3cts min⁻tubule⁻¹, more than 4000 times less. By 6 days, upper tubules from two Cambridge insects contained 142300±9500 and 102200±19900ctsmin⁻¹ tubule⁻¹, while upper tubules from two IVIC insects contained 2050±340 and 66± 18 cts min⁻¹ tubule⁻¹, again very many times less.

There is evidently a strong correlation between the presence of a large number of sizeable concretion bodies and the calcium content of the tubules or the number of counts of ⁴⁵Ca²⁺ taken up by the tubules from the haemolymph. We believe that this is strong evidence that the concretion bodies are the sites of calcium accumulation.

Since calcium is present at such high concentration in the Malpighian tubule cells, the question arises as to the form in which it is deposited. We collected together 80 of the most opaque Malpighian tubules dissected from eight fifth-stage and twelve adult insects reared on rabbits. The tubules were mounted in the slot of an aluminium sample holder and were allowed to dry. They were then examined by X-ray diffraction to find out whether significant amounts of the calcium deposits were crystalline. Fig. 8 shows the results of the scan together with a scan of a crystalline calcium salt [hydroxyapatite, 3Ca₃(PO₄)₂.Ca(OH)₂]. Clearly, no reflections arise from crystalline material in the tubules. Calcium must be deposited in an amorphous form in them.

We fed in vitro heparinised low-K⁺ sheep blood to 20 fifth-stage insects previously fed in vivo on rabbit blood (batch A) and allowed 20 fifth-stage insects that had been reared in vitro on heparinised low-K⁺ sheep blood to feed on rabbit blood (batch B). At daily intervals thereafter, we examined the tubules from insects from each group. Within 48 h, tubules from insects in batch B were significantly more opaque than tubules from control insects reared on heparinised blood alone. By 8 days the tubules appeared as opaque as those from insects reared on rabbit blood since hatching. By 13 days, the tubules had reached an average calcium concentration of 239±27 mmol l⁻¹ (N=4) compared with a value of 410±53mmoll⁻¹ (N=11) obtained in the animals fed on rabbit blood in vivo. Tubules from insects in batch A did not appear any less opaque for the first 8 days after feeding. Then they lost opacity; at least some of the tubules, taken in the period 8–12 days after feeding with heparinised blood, were not much less transparent than those from insects reared on heparinised blood. By 13 days, the average calcium content of 16 tubules from four insects in batch A had declined from the 410 mmol l⁻¹ of control insects to 202 ±61 mmol I⁻¹.

To test the idea that insects fed heparinised low-K⁺ sheep blood might not be able to take up calcium from the diet we fed fifth-stage insects about 40 μl each of ⁴⁵Ca²⁺ in saline at 135 000 cts min⁻¹μl⁻¹. Four insects (batch 1) previously reared on heparinised blood were then immediately allowed to feed to completion on heparinised low-K⁺ sheep blood (a further 300 μl). Three further insects reared on rabbit blood in vivo (batch 2) were fed to completion on heparinised low-K⁺ sheep blood immediately after ingesting ⁴⁵Ca²⁺ in saline. Finally, three insects reared on rabbit blood were fed to completion on rabbit blood straight after taking ⁴⁵Ca²⁺ in saline (batch 3). The concentrations of calcium in these different bloods are the same, so there were no consequent differences in specific activity of ⁴⁵Ca²⁺.

We took haemolymph samples from the insects on days 3, 4 and 5 after feeding. The concentration of ⁴⁵Ca²⁺ in the haemolymph declined in all the insects by about 25 % during this period. The concentrations of ⁴⁵Ca²⁺ on day 5 were 57909±13850ctsmin⁻¹μl⁻¹ (batch 1),78180±1890ctsmin⁻¹μl⁻¹ (batch 2) and 64870 ±20380ctsmin⁻¹μl⁻¹ (batch 3). These results indicate that the insects fed heparinised blood (batches 1 and 2) had absorbed as much ⁴⁵Ca²⁺ as those fed on rabbit blood (batch 3). The variability in the figures was probably due to the fact that all animals did not ingest exactly the same amount of ⁴⁵Ca²⁺ in the saline ingested. To normalize for this, we removed the intact midgut containing the blood meal from each insect on day 5 and disrupted it by sonication in 3 cm³ of 0.3 mol l⁻¹ KOH and determined its content of ⁴⁵Ca²⁺ from a 3μl sample. We could then compare the ⁴⁵Ca²⁺ concentration in the haemolymph with the ⁴⁵Ca²⁺ still present in the blood meal. Each microlitre of haemolymph from insects in batches 1, 2 and 3 contained 1.87±0.15%, 2.11±0.30% and 1.07±0.20%, respectively, of the ⁴⁵Ca²⁺ still in the gut. On this criterion, insects fed heparinised low-K⁺ sheep blood (batches 1 and 2) had significantly higher levels of ⁴⁵Ca²⁺ in the haemolymph than had insects fed on rabbit blood (batch 3). Therefore, there is no evidence of any suppression in ⁴⁵Ca²⁺ absorption from heparinised blood, indeed the figures suggest that more is absorbed, not less.

We also took all four upper Malpighian tubules from each insect of the three groups and counted their ⁴⁵Ca²⁺ contents without washing them further than occurred during the 5 min dissection period under saline. We found that the tubules from the insects fed heparinised low-K⁺ sheep blood had taken up considerable amounts of ⁴⁵Ca²⁺, equivalent to that in 927±98nl of haemolymph (batch 1) and 1175 ± 153 nl of haemolymph (batch 2). However, tubules from insects fed on rabbit blood (batch 3) had taken up significantly more ⁴⁵Ca²⁺, equivalent to 2625± 665 nl of haemolymph.

To estimate the total ⁴⁵Ca²⁺ absorbed by each insect, we added the ⁴⁵Ca²⁺ found in the tubules to that in the haemolymph, assuming from the quantities of haemolymph obtained in bleeding the insects that the haemolymph volume was 25 μl (unfed fifth-stage insects have 15 μl of haemolymph). By comparing this total with the sum of that in the gut and the amount absorbed, we could estimate the ⁴⁵Ca²⁺ absorbed from the gut as a fraction of the ⁴⁵Ca²⁺ ingested. Our results showed that the three groups of insects did not differ in the degree to which ⁴⁵Ca²⁺ was absorbed; insects in batch 1 absorbed 34.7±1.7% of the ⁴⁵Ca²⁺ in the meal, while those in batches 2 and 3 absorbed 34.8±7.7 % and 38.2±3.8 % of the ⁴⁵Ca²⁺ in the meal, respectively.

In these experiments, much more ⁴⁵Ca²⁺ was apparently taken up into the tubules of insects fed heparinised blood than in our earlier experiments (see above). We believe that removing tubules straight into scintillation fluid, after exposing them only to the dissection fluid, without first washing them for 20 min in 5-HT-containing saline may well explain the difference.

Particularly in insects from the Maracay stock, so opaque are the upper tubules that it is possible to see regional differences in them. The lowermost portion (nearest the transparent lower tubule) appeared decidedly more opaque than did the uppermost, tip end of the upper tubule. We removed four tubules from a fifthstage Cambridge insect and bathed them for 135 min in a ⁴⁵Ca²⁺-containing saline with 5-HT. The tubules were then washed for 10 min in saline containing 5-HT alone. Each upper tubule was cut into upper and lower halves and the ⁴⁵Ca²⁺ content of each was measured. The uppermost portions contained an average of 2063 ± 123 cts min⁻¹ but the lower halves contained on average 4315±373 ctsmin⁻¹ each, more than twice as much. As before, there is a correlation between the density of concretion bodies, estimated from the turbidity (opacity) of the tubule segment, and the ability to take up ⁴⁵Ca²⁺.

We measured the activity of calcium in three samples of haemolymph. The first two were volumes of 20–30 μl each taken from five fifth-instar insects from the Cambridge stock. The third was 15 μl taken from 13 fourth-instar insects from the IVIC stock fed on heparinised low-K⁺ sheep blood. The activities of calcium in the first two samples were 2.75 and 3.39 mmol l⁻¹, respectively, and that in the third was 2.24 mmol l⁻¹. These figures can be compared with total haemolymph calcium concentrations of 8.5±1.2mmoH⁻¹ (rabbit-fed insects) and 8.0±1.8mmoll⁻¹ (insects fed on heparinised low-K⁺ sheep blood).

Our findings provide evidence that most of the calcium in the blood consumed by Rhodnius is not eliminated from the body but is deposited at very high concentration in an amorphous form in concretion bodies in the cells of the upper Malpighian tubules. The average concentration of calcium in tubule cells from rabbit-fed Rhodnius was 410 mmol l⁻¹. From the relatively trivial amounts of calcium measured in cells with reduced concretion bodies, we can assume that calcium is mostly confined to the concretion bodies. Since the bodies occupy only a fraction of the cell volume, the calcium concentration in them must be much higher than in the cell as a whole, perhaps of the order of 5–10 mol I⁻¹. Since the calcium content of Malpighian tubules can be depleted, for example by feeding Rhodnius on heparinised low-K⁺ sheep blood, it is conceivable that calcium can be withdrawn from the deposits in the Malpighian tubules to maintain haemolymph levels when the insect feeds on blood deficient in calcium. However, this is most unlikely to be the normal function of the deposits, as all bloods likely to be ingested by Rhodnius contain significant amounts of calcium. It is more probable that the concretion bodies are the means whereby calcium is effectively removed from the metabolic pool - so-called deposit excretion (Maddrell, 1971). Calcium ions are relatively small, so that, during rectal concentration of urine by reabsorption of water (Ramsay, 1952), they (and other small solutes) would tend to cross passively back into the haemolymph by diffusion through the rectal wall and so be excreted only relatively ineffectually (Maddrell, 1980). More important still, high concentrations of calcium ions in the rectum might well interfere with the ability of the rectal cells to take up water. Water recovery is so vital for most insects (Edney, 1977; Phillips et al. 1986; Maddrell, 1980, 1987) that preventing interference by metal ions, such as calcium, may well be the major role of the ability of concretion bodies to sequester such potentially interfering substances. The need for them in insects would be greater than in other animals because the concentrating action of the rectum is more pronounced than that in the homologous parts of the excretory system in other animals (Phillips and Dockrill, 1968). One might argue that free calcium levels in the rectum would be lowered by binding to macromolecules in the faecal material. However, calcium binding sites on macromolecules entering the rectum from upstream parts of the alimentary canal are likely already to be saturated with non-absorbed calcium. So, any calcium entering the urine would tend to remain in the ionic form as the rectal contents are concentrated; no macromolecules accompany calcium in urine formation as they are retained by the filtering action of the Malpighian tubule wall (Coast, 1960).

Under normal circumstances, when the insect takes full meals in each instar, the amount of calcium deposited in the tubules increases steadily. The opacity of the tubules also increases at each succeeding instar. The number of concretion bodies in the tubules probably increases; their size certainly does. It seems likely that calcium is deposited in insoluble form in the concretion bodies so that the earliest deposited calcium lies in the centre. Our data show that ⁴⁵Ca²⁺ deposited over periods of the order of 10–12 days is released less rapidly than ⁴⁵Ca²⁺ incorporated for shorter periods. In electron micrographs where their contents have been preserved, the concretion bodies appear as layered structures suggestive of a series of phases of deposition, occurring perhaps after every meal when calcium uptake rates are higher. The post-feeding increase in calcium uptake is very reminiscent, especially in its timing, of the increases in transport of para-aminohippuric acid and of uric acid that occur after a meal (Maddrell and Gardiner, 1975; O’Donnell et al. 1983).

Insects from the Maracay stock whose tubules had very high densities of concretion bodies had been routinely fed on chicken blood. At least in laying hens, the blood calcium concentration is more than twice as high as in rabbits (Altman and Dittmer, 1974). Since hens are in lay for at least 80% of the year, it is very likely that, by maturity, these Rhodnius had ingested twice as much calcium as insects fed on rabbit blood. If so, and if only the usual amounts of calcium are excreted, contained in the haemolymph and used elsewhere in the body, then more than twice as much is left to be deposited in the Malpighian tubules. This may be the explanation of the observation that the density of the concretion bodies in the upper tubules of fifth-stage Maracay insects was so much higher than in the Cambridge insects. Two tubules from the Maracay stock had an average calcium content of 1190 mmol l⁻¹, nearly three times the average of tubules from rabbit-fed insects.

We have found that ⁴⁵Ca²⁺ recently incorporated into the concretion bodies is steadily released into saline in vitro at a rate of about 25 % per hour. A meal of heparinised sheep blood also promotes calcium release from the tubules. Because cytosolic levels of free calcium are so low (about 10⁻⁷moll⁻¹, Campbell, 1973), lower than those in saturated solutions of such sparingly soluble salts of calcium as carbonate, sulphate, oxalate and phosphate, energy would have to be continuously expended in maintaining the calcium content of the concretion bodies if it is present there as any of these salts. So, it is not surprising to find that ⁴⁵Ca²⁺ steadily passes into the bathing solution from tubules in vitro; presumably this represents the steady-state efflux from the concretion bodies. The effects of heparinised low-K⁺ sheep blood could take effect at the level of the calcium concentrating mechanism of the concretion bodies. This could explain why tubules from insects fed heparinised low-K⁺ sheep blood accumulate significant amounts of ⁴⁵Ca²⁺ but nearly all this is lost if the tubules are washed for 20 min in tracer-free saline. If the concretion bodies are to prevent calcium from reaching the rectum in significant amounts, it is of course important that any calcium released from them passes to the haemolymph rather than to the lumen of the tubule. In support of this, we have found that about 10 times as much ⁴⁵Ca²⁺ passes to the haemolymph side in unit time as enters the lumen in vitro. This is in spite of the observation that the area of cell membrane facing the lumen is at least twice that in direct contact with the haemolymph (Maddrell, 1980). Presumably, membrane-sited calcium-exporting pumps are predominantly found on the haemolymph-facing cell membrane. Of course, if calcium levels in the urine are to be kept low, the tubule walls should be relatively impermeable to calcium. We have found that tubules are consistently 60% more permeable to [¹⁴C]sucrose than to ⁴⁵Ca²⁺, although calcium ions are smaller than sucrose molecules. Sucrose molecules have a radius of 0.45 nm while calcium ions have a hydrodynamic radius of 0.31 nm and a hydrated radius of 0.41 nm. We found earlier (Skaer et al. 1987) that the intercellular junctions are more permeable to anions than to cations, so it is not surprising to find a low transepithelial permeability for ⁴⁵Ca²⁺.

One can try to calculate the loss of ⁴⁵Ca²⁺ through the tubules as a fraction of that taken up into the cells. In recently fed fifth-stage insects, the amount of ⁴⁵Ca²⁺ taken up in 1 h from the haemolymph by one upper tubule is equal to that contained in 35 nl of haemolymph (based on the uptake rate over the first 24 h, Fig. 3). Assuming that the rate of secretion by each of the tubules is 0.5 nl min⁻¹ and that the permeability of the tubule to ⁴⁵Ca²⁺ is 0.02 nl min⁻¹ mm⁻², it can be calculated, using Ramsay’s formula (equation 1), that the ⁴⁵Ca²⁺ loss through each tubule in 1 h is equal to that contained in about 8 nl of haemolymph, equivalent to 22% of the amount taken up by the cells. However, not all the calcium in the haemolymph is free; its activity is about 3 mmol l⁻¹ and its concentration 8 mmol l⁻¹. So ⁴⁵Ca²⁺ loss passively through the tubule wall in 1 h will only be equivalent to that contained in 3nl of haemolymph, i.e. calcium is taken up and deposited in the tubules 12 times as fast as it is lost into the primary urine. No doubt this is a major part of the explanation of how Rhodnius retains in its body most of the calcium from its meals of blood.

We are very grateful to Dr Ricardo Montoreano of the Centro de Investigaciones Biomédicas (BIOMED), Universidad de Carabobo, Nucleo Aragua, Apdo 2188, Maracay, Venezuela, for the gift of Rhodnius from his colony; to Dr Zulli Benzo and Rebeca Gomes of Centro de Quimica at IVIC, for the calcium analyses; to Dr Leonardo Mateu of the Molecular Structure Laboratory at IVIC for the X-ray diffraction study; to Dr P. Sepulveda of the Department of Cell Biology, AFRC, Babraham, for help with determinations of calcium activity, and to Dr David Herbst for his help with our preliminary experiments. We also want to thank Mrs D. Otero for help with the drawings. Finally, we wish to thank the British Council for their assistance which made possible the collaboration between our two laboratories.