The action of the excretory apparatus of Calliphora vomitoria in handling injected sugar solution

The Malpighian tubules of insects were once considered freely and passively permeable to organic solutes, so that such molecules would be passively excreted, providing there was adequate fluid secretion and that they were not reabsorbed (Ramsay, 1958). The free permeability of the Malpighian tubules of some insects to organic solutes has recently been confirmed (Maddrell & Gardiner, 1974). Recovery of glycine which is thought to diffuse into the tubule fluid has been demonstrated in the recta of locust and cockroach (Balshin & Phillips, 1971 ; Wall & Oschman, 1970). However, evidence is accumulating that the Malpighian tubules of insects can preferentially transport molecules whose presence could either be deleterious, e.g. dyes and acylamides (Maddrell et al. 1974), or beneficial, e.g. sugars (Knowles, 1975).

Recently, it has been suggested that the isolated Malpighian tubules of C. vomitoria reduce the excretion of trehalose and glucose (Knowles, 1975). It was proposed that the tubule hydrolyses trehalose to glucose derivatives and that the excretion of glucose is specifically restricted. This report describes the excretory pattern of a range of sugars after they have been injected into the haemolymph of the intact animal.

All experiments were performed on male flies between 48 and 60 h after pupal emergence. They were fed on a mixture of sugar and dried milk, and water was given ad libitum. Larvae were purchased from commercial suppliers.

Each fly, with its ventral side uppermost, was restrained on a slide covered with beeswax. A standard volume of Ringer (4·7 µl) was injected through the mesopleuron. The injection of this volume, about one third of the haemolymph volume, induced a diuresis in the fly. Gentle prodding of the abdomen caused the fly to release urine, which was collected at regular intervals. Haemolymph was taken from the site of injection, but was sampled less frequently than the urine and not taken more than three times from any one fly, thus preventing too great a loss of heamolymph. Sample volumes were calculated from the measured diameter of the droplets when immersed in paraffin oil. Sugars were universally labelled with ¹⁴C and purchased from the Radiochemical Centre, Amersham. Each batch of 50 µCi of ¹⁴C was dissolved in either 1·5 ml or 0·75 ml of Ringer. The higher specific activity was preferred for sugars which were considered metabolically active, namely glucose and trehalose. The concentration of sugars injected into each fly was 50 mM, this would be reduced by approximately two-thirds by haemolymph volume. In some cases sugars were separated by thin layer chromatography (T.L.C.). The procedure has been described elsewhere (Knowles, 1975).

The rates of induced urine excretion resulting from fluid injection are shown in Fig. 1. It is apparent that diuresis is stimulated within 5 min. Excretion is fast at first, and then gradually reduced, reaching a low level after about an hour and a half.

If the monosaccharide sorbose (molecular weight 180) is injected into the haemolymph, then considerable amounts of radioactivity are found in the urine (Fig. 2). Sorbose is not thought to be metabolized by insects (Fraenkel, 1940), so in all probability the radioactivity levels reflect chemical concentrations. There are several interpretations of these observations. Firstly, injected sorbose may have been shunted directly from the site of injection, in the thorax, into the alimentary canal. This is unlikely, since, when haemolymph is sampled at both the abdomen and thorax, the specific activities of both samples are approximately one-third of the injected fluid; which is to be expected if the haemolymph and injected fluid mix freely (Fig. 2). Secondly, the Malpighian tubules could be secreting fluid containing sorbose at a concentration above that of their surrounding haemolymph. This seems improbable because in vitro experiments show that the isolated tubules of Calliphora do not secrete sorbose at high concentrations (Knowles, 1975). Indeed, it can be shown that at increased rates of tubule fluid production (which are likely to occur during diuresis), the tubule fluid concentration of diffusing solutes such as sorbose is decreased (Maddrell & Gardiner, 1974). Thirdly, the fluid leaving the Malpighian tubules could contain sorbose at concentrations below that of their bathing haemolymph; subsequent water reabsorption by either the ileum or rectum could elevate the final sorbose concentration in the urine. This third interpretation seems the most likely, since if xylose is injected similar results are again obtained (Fig. 3). Thus the non-metabolizable sugars are promptly removed from circulation, if this is coupled to high rates of urine excretion, then the combined effect would be to deplete the haemolymph of these sugars.

To investigate the fate of useful sugars, preliminary experiments were performed to compare the amount of radioactivity excreted following the injection of labelled sorbose with that following the injection of labelled glucose. Less radioactivity is excreted following glucose injection than following sorbose injection (Table 1). However, it is well known that insects rapidly convert glucose to trehalose (Wyatt, 1967). Experiments were performed which verified this : in all cases there was a marked transference of radioactivity from glucose to trehalose (Fig. 4). The urine radioactivity detected after glucose injection may partly reflect the excretion of isotope derived from both haemolymph trehalose and residual glucose. Chromatography of urine samples did not reveal peaks of radioactivity in either the glucose or trehalose zones (Fig. 5) ; instead, there was a broad zone of increased radioactivity about 2·5 in. along the T.L.C. plate. This may indicate that several compounds which are present in the urine have become labelled. The levels of glucose and trehalose excreted are therefore much lower than is suggested by the measurements of radioactivity given in Table 1.

Since the haemolymph concentration of trehalose is about tenfold greater than that of glucose (Wyatt, 1967; Normann, 1975), the major threat to wasteful sugar losses may be the excretion of trehalose. The excretion of trehalose was compared with that of sucrose (Table 2). The sucrose experiments were performed with a Ringer of low specific activity and for comparison the sucrose values ought to be doubled: alternatively, the results may be expressed as the urine/haemolymph ratios. Whichever expression is favoured, it is clear that less radioactivity is excreted when trehalose is injected than when an equal amount of sucrose is injected. Chromatography of haemolymph samples taken 45 min after injection of labelled trehalose shows that the majority of the isotope is recoverable at the trehalose zone (Fig. 6). However, chromatography of the urine sample failed to distinguish any radioactivity peaks (Fig. 7), suggesting that general metabolism may have distributed the injected radioactivity among several species of molecules with a consequent lowering of specific activity for each substance : without preliminary fractionation there would be insufficient separation of several solutes by one dimensional chromatography.

It is noticeable from Table 2 and Fig. 8 that, after sucrose injection, the levels of radioactivity in the urine are well below those found in the haemolymph. Some preliminary experiments with inulin have yielded similar results (Knowles, 1974). This is in contrast to the experiments with sorbose and xylose where these solutes were concentrated in the urine. This could mean that the Malpighian tubule walls are much less permeable to disaccharides than monosaccharides.

Although the radioactive content of the urine and haemolymph samples were measured at intervals, it is important to relate the radioactivity of the urine sample with the radioactivity of the haemolymph from which that urine was secreted. Therefore, 100 nl of the dye amaranth were injected 5 min after diuresis had been stimulated by the injection of 4·7 µl of Ringer. Coloured urine gradually appeared after 3-4 min, and the colour intensity was fully developed by 5 min. Dissection of the excretory apparatus revealed that dye first appeared around the junction of the Malpighian tubule common collecting ducts with the gut. It is concluded that the lag period between initial secretion of tubule fluid and voidance of urine is around 5 min.

Evidence has been presented which suggests that the fly, even during diuresis, will reabsorb water. If we assume that it is only reabsorption which raises the concentration of sorbose in urine, it is possible to estimate the magnitude of this reabsorption.

A fivefold difference in concentration would necessitate at least a fivefold reduction in tubule fluid volume, and therefore the volume of fluid initially secreted by the tubules would have been fivefold greater than the voided volume. The difference in these volumes will be the extent of water reabsorption. The rate at which urine is voided has been obtained from the diuresis profile of Fig. 1, and the calculated rates of water reabsorption per time interval are shown in the last column of Table 3. Although it must be emphasized that these are crude estimates of water fluxes across the tissues, one is surprised by the very high rates of fluid flow which are indicated, especially by those of the Malpighian tubules. If all four tubules secrete fluid at equal rates, then the maximal rate of fluid produced by each tubule is about 40 nl/min (Table 3, line 4). This rate of fluid secretion has never been achieved from isolated Malpighian tubules of Callphora. However, the proposed rate of fluid production per unit length of tubule is comparable to that achieved by stimulated Rhodnius tubules. In Calliphora, each tubule is approximately 35 mm long, so the rate of secretion would be 1·1 nl/ min/mm: the comparable value for Rhodnius is 2·0 nl/min/mm (Maddrell, 1969).

The suggested rates of water reabsorption are also higher than the rates previously recorded from in situ, isolated rectal preparations of the blowfly. Phillips (1969) obtained values for rectal water reabsorption of around 20 nl/min/rectum. Perhaps other tissues of Calliphora, such as the lengthy ileum, provide additional water reabsorption. It is possible that the ileum absorbs large amounts of isomotic fluid, which would markedly reduce the tubule fluid volume, and that the rectum specifically regulates the osmotic pressure of the urine in accordance with the needs of the insect.

Water reabsorption may appear contradictory to the immediate physiological needs of a diuresing insect in that such reabsorption would tend to extend the period of time before the fly has excreted the desired volume of fluid. However, during this period of prolonged diuresis the fly may circulate a far greater volume of haemolymph through the excretory apparatus than would occur if the tubule fluid was more directly excreted. The data from Table 3 show that the total volume of haemolymph secreted by the Malpighian tubules is probably at least 7·0 µl, well above the 3·0 µl of urine which is actually collected. This continuous recycling of haemolymph may facilitate solute excretion by increasing the volume of haemolymph filtered of unwanted substances.

The urine/haemolymph radioactivity ratios of solutes may be taken as approximate indices of the efficiency with which the excretory apparatus removes those solutes For sorbose and xylose, the urine/haemolymph ratios which occur 45 min after injection of isotope are 3·7 and 4·3 respectively; but for sucrose, the comparable value is only o·1, about a fortyfold difference. In comparison an investigation utilizing the isolated Malpighian tubules of Calliphora did not reveal such a large difference between the permeabilities of sucrose and sorbose. The ratios of radioactivity appearing in the tubule fluid/bathing medium from isolated tubules were: sucrose, 0·25; sorbose, 0·63; and xylose, 0·74 (Knowles, 1975). In addition the solute permeability constants for the total surface area of individual tubules were: sucrose, o·6; sorbose, 2·1 ; and xylose, 4·5 nl/min (Knowles, 1974). This suggests that when the Malpighian tubules are functioning in vivo, they are less permeable to disaccharides than are isolated tubules. However, it is possible that the low sucrose clearance by the intact fly represents nothing more than recovery of the constituent monosaccharides of sucrose, following sucrose digestion in the ileum. Further evidence is required to resolve this argument, but if subsequent work shows that the in vivo tubule is significantly less permeable to molecules larger than monosaccharides, then the secretiondiffusion theory of insect excretion (Ramsay, 1958) will require amendment.

It is my pleasure to thank Professor J. Shaw for his encouragement and supervision of this work. My thanks are also owing to the Science Research Council for their financial support.