The Passive Permeability of Insect Malpighian Tubules to Organic Solutes

The excretory system of insects is essentially a secretion/reabsorption system ; the Malpighian tubules secrete a fluid which is passed to the hind gut where selective reabsorption occurs (Ramsay, 1958; Maddrell, 1971). As Ramsay (1958) discovered, Malpighian tubules are highly permeable to such organic solutes as amino acids and sugars. These useful small molecules are reabsorbed in the hind gut, particularly in the rectum (Ramsay, 1958), where the uptake of amino acids has recently been shown to be an active process (Balshin & Phillips, 1971). As Ramsay (1958) was the first to realize, this apparently pointless committal of useful substances to the excretory flow and their later reabsorption is but one aspect of what from the insects point of view is a particularly sane arrangement. New toxic molecules or elevated concentrations of other useless or toxic molecules are automatically excreted simply by there being no mechanism for their reabsorption.

To repeat, the reabsorption of useful molecules from the excretory fluid goes on mainly in the hind gut. The hind gut of insects is lined with cuticle and, in the locust, this cuticular lining of the rectum has a rather low permeability to organic solutes as small as disaccharides (Phillips & Dockrill, 1968). Yet the Malpighian tubules of the stick insect are highly permeable to disaccharides ; for example, the fluid they secrete contains sucrose at 45% of the concentration that it occurs in the bathing fluid (Ramsay, 1958). Were the locust to operate in a similar fashion it would seem that the excretory system would rapidly drain the haemolymph of substances of the order of the size of trehalose, because such molecules would cross the walls of the Malpighian tubules but would be too large to cross the cuticular lining of the rectum and so could not be reabsorbed there.

As a first step to investigate this type of paradox this paper describes experiments to determine the passive permeability of a range of Malpighian tubules to a range of organic solutes of varying size. To anticipate the results it appears that Malpighian tubules have a considerable permeability even to molecules as large as inulin. The consequences of this finding are discussed on p. 648.

Malpighian tubules of the following insects were examined: Calliphora erythro-cephala (Diptera) adult insects; Schistocerca gregaria (Orthoptera) adult and larval insects; Rhodnius prolixus (Hemiptera) 5th-stage larvae; Triatoma phyUosoma (Hemiptera) 5th-stage larvae; Manduca sexta (Lepidoptera) newly emerged adults.

All the insects were taken from laboratory cultures maintained in the Department of Zoology, Cambridge. Triatoma phyllosoma is a very large blood-sucking insect closely related to Rhodnius and included for comparison with Rhodnius.

Preparations of isolated Malpighian tubules from the above species were made as described elsewhere (Calliphora erythrocephala (Berridge, 1966); Schistocerca gregaria (Maddrell & Klunsuwan, 1973); Rhodnius prolixus (Maddrell, 1969); Manduca sexta and Triatoma phyllosoma as for Rhodnius). Basically tubules are dissected from the insect and placed in an appropriate Ringer ‘s solution under liquid paraffin. Secreted fluid is collected at the cut end of the tubule held away from the bathing solution on a glass rod stuck into the wax base of the experimental chamber. Secretion by Malpighian tubules of Rhodnius and Triatoma was stimulated with solutions containing 5-hydroxy-tryptamine (Maddrell, Pilcher & Gardiner, 1971); the other tubules secrete without the need for an added stimulant. Only the upper, secretory parts of the Malpighian tubules of Rhodnius and Triatoma were used.

The following substances were used in this investigation: urea, benzoic acid, L-valine, L-aspartic acid, L-glutamic acid, L-glutamine, D-xylose, L-lysine, L-3-phenylalanine, L-fucose, uric acid, L-tyrosine, maltose, sucrose, lactose, maltotriose and inulin. These compounds, labelled either with carbon-14 or tritium, were obtained from the Radiochemical Centre, Amersham.

In a typical experiment a solution containing the test substance at a suitable concentration was made up, and individual Malpighian tubules or sets of tubules were set up in it. After sufficient fluid had been secreted to flush out the fluid initially in the lumen of the tubules further drops of secreted fluid were collected for assay of their contained radioactivity, and the rate at which the fluid was secreted was measured. After the volume of such drops had been noted (from measurements of their diameter (Maddrell, 1969)) they were absorbed on to the tips of small triangular pieces of Whatman no. 3 chromatography paper which were carefully dried and then transferred for counting to glass vials containing the scintillation mixture (0·3 % PPO in xylene). Alternatively, the drops of fluid were transferred by fine pipette straight into scintillation vials containing Bray ‘s fluid. Samples were counted using an Intertechnique ABAC SL40 scintillation counter. As a check that a particular test substance was not metabolized in crossing the walls of the Malpighian tubules, samples of the secreted fluid and bathing fluid were subjected to paper chromatography in three different solvent systems and scanned radiometrically with a Berthold Scanner Model LB 2722. Of the substances used, very nearly all were found not to be metabolized to a measurable extent by Malpighian tubules. The concentration of a test substance in the secreted fluid was expressed as a fraction of that in the bathing medium, henceforth called the S/M ratio.

Values for a were calculated from measurements of the dimensions and the secretory rates of several Malpighian tubules from each species of insect used. The measurements taken as typical are set out in Table 1.

Ramsay himself was not able extensively to use the third criterion in his work on Malpighian tubules isolated from Carausius because of the difficulty of altering the rate of secretion of fluid. He was able, however, to perfuse tubules at differing rates and in the one experiment he carried out, using urea as the test substance, he was able to conclude that this substance did cross the tubule wall in a passive fashion.

The fullest set of tests was carried out using 5 HT-stimulated tubules of Rhodnius. The permeabilities of the wall of this tubule to the various test substances are set out in Table 2, together with their molecular weights and the results of tests relevant to the manner in which the substances enter the tubule lumen. In Fig. 1 results for four test substances are plotted to show the dependence of the M/S ratio on the rate of fluid secretion; the linear relationships are what one would expect for substances crossing the tubule wall passively. Fig. 2 shows how the S/M ratio depends on the molecular weight of test substances. Not unexpectedly, the larger compounds penetrate the tubule wall more slowly, but even they penetrate the wall surprisingly easily so that inulin, for example, reaches a concentration close to 1 % of that of the bathing solution in a tubule secreting at 36 nl.min⁻¹. There is also an apparent tendency for the more highly charged molecules to penetrate more slowly than the more electrically neutral compounds. Specific examples are L-valine and D-xylose which appear at 7 and 35 % respectively in fluid secreted at 36 nl.min⁻¹ and even more striking, maltose and sucrose which appear in such fluid at about 1 and 7% respectively, though their molecular weights are the same. As a result, of course, the line drawn through the points in Fig. 2 should be taken only as an indication of the way in which molecules of a particular size might behave. Clearly other factors such as the charge on the molecule have a large effect on the rate of penetration of any particular substance.

The results of exactly similar experiments using different Malpighian tubules are presented in Table 3 and Fig. 3. In Table 3 have been included, for comparison, figures for the Malpighian tubules of Carausius morosus calculated from those of Ramsay (1958) and Ramsay & Riegel (1961). Again, what is striking is the permeability of the tubule walls even to compounds as large as inulin. In individual cases where tubules secrete fluid particularly slowly, this high permeability results in the concentration of inulin in the secreted fluid reaching nearly 50 % of its concentration in the bathing fluid.

As before, the essentially passive nature of the movements tested is shown by the facts that the concentration of test substances was never higher in the secreted fluid than it was in the bathing fluid and that, where it was measured, the S/M ratio was not affected by changes in M.

Apart from the tubules of Calliphora, which have distinctly higher permeabilities to the substances tested, the Malpighian tubules of the different insects studied are closely similar in their permeability properties.

It is a point of some concern to establish whether the experiments so far described can be used in an extrapolation to the situation in vivo. To investigate this point we injected non-metabolized tracer substances into insects and followed the subsequent course of their excretion. For these experiments we chose inulin and L-fucose as markers and injected them into adult Schistocerca and fed 5th-stage larvae of Rhodnius, respectively. While inulin is not metabolized by the locust, it does become sequestered in the fat body and gonads (Loughton & Tobe, 1969). We therefore followed the concentration of inulin in the haemolymph of locusts over a period of up to 10 days after injection. The results are set out in Fig. 4 from which it is clear that to a first approximation the inulin concentration declines exponentially.

The faeces collected from eight injected locusts was found to contain 4·5 % of the inulin injected 5 days previously. From Fig. 4, the haemolymph at this time contains 61 % of the inulin injected initially. On the bold assumption that the fat body and gonads sequester inulin from the haemolymph at a rate proportional to its concentration therein, one can calculate from these figures the time constant for the removal of inulin by the Malpighian tubules as 126370 min. Using the inulin dilution method of Loughton & Tobe (1969), we measured the haemolymph volume of ten adult locusts and found this to be 608 ± 41 μl (mean + S.E.M.). The Malpighian tubules of adult locusts secrete fluid into the hindgut at a rate of 8 41 μl.h⁻¹ (Phillips, 1964). From these figures one can calculate that the in vivo U/P ratio for inulin is 0·036 and that, from the formula U/P = b[(a + b), the permeability of the Malpighian tubules to inulin, b, is 0 ·019 nl.mm⁻².min⁻¹ which is in excellent agreement with the figure of 0·022 nl. Mm⁻². Min⁻¹ found in our in vitro experiments (Table 3).

As a further test of the relevance of our in vitro results to the in vivo situation, we injected radioactive L-fucose into the haemolymph of recently fed Rhodnius and followed its appearance in the urine for about 30 min before taking a haemolymph sample. This gave simultaneous measurements of the concentration of L-fucose in the haemolymph and urine. Using the formula U/P = b/(a + b) as before, the results for five insects gave a value of b for L-fucose of o·66 nl. Mm⁻². Min⁻¹ which is in good agreement with the figure of o·6o nl. mm⁻². min⁻¹ from our in vitro experiments (Table 2).

What emerges from this survey of insect Malpighian tubules is that they are permeable to a wide variety of organic substances. This of course makes good sense in an excretory organ, where as pointed out earlier, automatic excretion of toxic molecules is ensured by a relatively unselective committal of materials to the primary excretory fluid followed by a selective reabsorption of useful substances. Thus passive excretion of relatively large molecules requires that the Malpighian tubule be permeable to them ; from results of the present investigation it appears that this is the case.

What is at first sight surprising is that if one compares the permeability of the locust Malpighian tubules with that of the cuticular lining of its rectum (Phillips & Dockrill, 1968), they appear to be poorly matched. For example, from the present results one would expect the fluid leaving the Malpighian tubules to contain disaccharides at concentrations of about 5–10% of the levels occurring in the haemolymph. Yet the rectal cuticle has a very low permeability to substances as large as this. If such materials are not to be lost, then it follows that they must be reabsorbed elsewhere. It is entirely possible that the hind gut anterior to the rectum (through which the primary excretory fluid must in any case pass en route to the rectum) is lined with more permeable cuticle and that reabsorption of useful molecules other than the very smallest can go on here. As has been emphasized before (Maddrell, 1971), the study of the anterior hind gut (often called the ileum and/or the colon) seems to be very worthwhile.

With the benefit of hind sight it is possible to make sense of the apparently greater permeability of Malpighian tubules compared with the rectal wall. Absorption of water in the rectum must lead of course to a great concentration of substances dissolved in the rectal fluid. Were the rectal wall to be permeable to materials intended for excretion there would be a strong tendency for them to re-enter the haemolymph passively, which would be counter-productive. In addition, as Phillips & Dockrill (1968) have pointed out, the rectal cells, on whose controlling activity the insect vitally depends, would be bathed by high concentrations of substances some of which are toxic. Again the impermeability of the rectal cuticle makes sense.

From the point of view of compounds that an insect might wish to circulate in its haemolymph, the permeability of the Malpighian tubules forces a choice on the insect. If a substance of low molecular weight is to be circulated, then the insect must be prepared either continually to reabsorb the substance from the excretory fluid or countenance its steady loss from the haemolymph. Alternatively, materials must be circulated in a form in which they are attached to some large carrier substance, the complex then being sufficiently large to escape excretion. In this connexion it is worth noting that lipids, which many insects use as the fuel for flight, are mainly transported as diglyceride-protein conjugates and it is possible that unesterified fatty acids are also bound to protein (Gilbert, 1967). In addition, it is possible that insect juvenile hormone is circulated as a hormone-protein complex (Pratt, 1972).

So large are some of the substances to which Malpighian tubules are permeable that one can scarcely doubt that they must cross the tubule wall in the lateral spaces between the cells. These spaces are long, thin and narrow and are ‘closed ‘by septate desmosomes and ‘tight ‘junctions. However, such an arrangement may be permeable as has been emphasized by Frömter & Diamond (1972). Fluid secretion by Malpighian tubules is thought to be achieved by an osmotic coupling of water fluxes to a primary active transport of ions (Maddrell, 1971). To be efficient, such a mechanism demands that ion transport occurs across a membrane which is osmotically relatively tight. If the intercellular route is leaky, then it follows that ion and water movements probably use an intracellular route. One envisages then that Malpighian tubules produce an iso-osmotic flow of fluid through the cells and that the excretorily useful permeability is achieved by a parallel leaky intercellular route (Fig. 5). The continuous fluid flow into the lumen ensures that a concentration gradient always exists between the haemolymph and the primary excretory fluid and this in turn ensures the passive appearance in this fluid of substances able to diffuse across the tubule wall. Just how important this function is emerges clearly from the fact that fluid is secreted by the Malpighian tubules far faster than it is allowed to leave the insect. A similar disparity arises in mammalian kidneys (Smith, 1951) for exactly the same reason.

Although Malpighian tubules are permeable structures this is not to deny the existence in them of specific active transport systems for materials which the insect has a constant need to excrete. As will be described in a later paper (Maddrell et al. 1974) at least two such mechanisms exist in insect Malpighian tubules.