Studies on the Exchange and Regulation of Sodium in the Larva of Aedes Aegypti (L.): I. The Steady-State Exchange

The larva of Aëdes aegypti (L.) lives naturally in fresh waters in which it keeps the concentration of salts in its haemolymph far above that in the medium. It does this by secreting salts from the medium into the haemolymph against a steep concentration gradient by means of the anal papillae, and by resorbing salts from the urine through the walls of the rectum; the Malpighian tubules also contribute to the work of salt retention by secreting a urine which is slightly hypotonic to the haemolymph (Ramsay, 1950, 1951, 1953). The larvae can keep the levels of sodium, potassium and chloride in the haemolymph constant despite wide variations in their concentrations in the medium (Wigglesworth, 1938; Ramsay, 1953). The anal papillae are responsible for most of the permeability of the larva to salts and water (Wigglesworth, 1933 a, b;Treherne, 1954).

The steady-state exchange of sodium between the starved larva and the medium has been studied by Treherne (1954). The exchange appears to be a first-order process and it occurs mainly through the anal papillae. It is independent of the potassium concentration in the medium over the range 0·159-4·o mM/1., and of the sodium concentration over the range 4-8 mM/1.

In this paper I shall describe experiments on the steady-state exchange of sodium between the haemolymph (or whole larva) and the medium. ²²Na has been used to study the exchange. The sodium in the haemolymph has been measured by equilibrating or rearing larvae in labelled media, and by flame photometry using an ‘Eel’ flame photometer. Experiments on the regulation of the sodium level in the haemolymph will be described in a further paper. The aim throughout the work was to try to throw more light on the processes of sodium exchange and regulation by altering suitably the conditions to which the larvae were exposed.

Larvae were reared from the eggs in lots of 150-300 in 500 ml. of the following medium:

The pH was adjusted to about 6·3; 0·3 g. of a mixture of dog biscuit and ‘Bemax’ was added and the resulting infusion provided a comfortable excess of food up to the 4th instar and beyond. Larvae reared in this way moulted into the 4th instar within a few hours of one another. The addition of the foodstuff to the medium caused an increase in the sodium, potassium and calcium contents of about 1, 13 and 5 % respectively.

Both fed and starved 4th instar larvae were used in the experiments. The fed larvae were always given an excess of food, and the starved larvae were kept in a medium similar to the one they were reared in but without food for at least 60 hr. before the start of an experiment. I have found that larvae which have recently moulted into the 4th instar can withstand 9-12 days of rigorous starvation before any die. All the experiments were done at 28° C. and the media were aerated with compressed air.

To study the exchange of sodium between the larvae and the medium the larvae were reared in unlabelled or labelled media, starved if necessary, and put (after appropriate washing) into 500 ml. of labelled or unlabelled media. The alteration in radioactivity in the haemolymph was noted at appropriate intervals. In the case of fed larvae the experimental media contained o·1 g. of foodstuff. When the uptake of radioactivity was followed the total sodium content was found by flame photometry. In some cases the exchange of sodium in single larvae was studied. For this the larvae were reared in labelled media and put into unlabelled media. They were removed from these at appropriate intervals, dried with filter paper, and put on a marked area on a special counting disk. The radioactivity was measured for 2-4 min. and the larvae were returned to the media. It was assumed that self-absorption of radioactivity stayed constant. For measurements of radioactivity haemolymph was collected (under a binocular microscope) from each of a group of five larvae into silicone-lined glass braking-pipettes 0·1-0·2 μl. in volume. The larvae had previously been dried with filter paper. The haemolymph was then placed, with the washings, on separate counting disks and after it had dried it was assayed for radioactivity. For each group the mean radioactivity per μl. and its standard deviation were calculated. Radioactivity adhering to the surface of the larvae was, if necessary, removed before collecting the haemolymph by placing them in a slow stream of tap water for 1 min. The radioactivity was measured with G.E.C. GM4 endwindow counters and appropriate scaling equipment. Because of the long half-life of ²²Na (2·6 years) it was not necessary to correct the measurements for radioactive decay. The accuracy of the measurements was as follows:

(a) Radioactivity of the whole larvae; the standard deviation varied from 2-6% at the start of an experiment to 10 % at the end when the radioactivity was very low.

(b) Radioactivity of haemolymph samples; for weak samples the standard deviation was between 4·8 and 7·1%, for active samples it was between 1-6 and 3-2 %. The pipettes were calibrated with a solution of ²²Na of known radioactivity. The standard deviation of their volumes was 1·4%. Several pipettes were used during each experiment as a deposit from the haemolymph slowly forms on the silicone lining. ²²Na was obtained as a practically weightless solution of the chloride from the Radiochemical Centre, Amersham. The specific activities of the labelled media were between 1·5 and 5·4 μg. “Na/g. Na.

For measurement of the sodium level in the haemolymph by flame photometry 0·1—0·2 μ1. of haemolymph was collected from each of a group of five larvae and placed, with the washings, in a small polythene tube containing 5 ml. of distilled water. The larvae had previously been washed in a large volume of distilled water and dried on filter paper. The tube was then stoppered and shaken and the sample analysed on the ‘Eel ‘flame photometer using standard solutions containing up to 1·7 parts per million of sodium (as chloride). At least three estimations were made on each sample and the mean and its standard deviation were calculated. In this way the sodium in solutions containing roughly 0·1-0·5 μg-Na/ml. could be analysed with an accuracy of ± 1-4 % (standard deviation).

For a many-celled animal like a mosquito larva the permeability constants describe only the overall process of permeation. The curves described by equations (1) and (2) will from now on be called ‘exchange uptake’ and ‘exchange washout’ curves. Plots of the logarithmic expressions of these equations against time give straight lines the slopes of which are equal to Kout-The error in determining the points increases as the exchange nears completion (Solomon, 1952) so it is not legitimate to calculate regression lines in order to find K_out. Following Treherne (1954) is here found as the mean of the slopes of the lines joining the individual points to the origin.

In what follows the term ‘normal larvae’ refers to larvae either fed or starved which have not been operated on.

It has been found that the sodium level in the haemolymph of normal larvae is not reduced by starvation for periods greater than 200 hr., but remains steady at about 100 mM/1. The sodium level in fed larvae also remains steady at about this level up to the time of pupation.

Results for the exchange of sodium in fed and starved normal larvae are shown in Figs. 1 and 2. It is clear that the exchange proceeds much more rapidly in fed than in starved larvae, and that all the sodium in the haemolymph (and whole larva) is exchangeable with that in the medium. When the values for

During the course of an experiment with starved larvae the pH of the medium rises slowly from 6 · 3 to about 7·2, but this does not appear to affect the exchange of sodium which proceeds exponentially throughout the experiment. In the case of fed larvae the pH of the medium is made to change in a complex manner by the addition of the foodstuff and the subsequent growth of microbes. However this change may be ignored as a change of pH from 5·2 to 7·2 (a range greater than that encountered) was found not to affect the exchange. The small amounts of potassium and calcium which were added to the medium with the foodstuff were shown not to alter the rate of exchange in starved larvae. This is to be expected from Treherne’s observations.

The information provided by the experiments on normal larvae is given in Table 1. Data obtained from single larvae are not included as these do not adequately represent the exchange in a population of larvae. The figures marked with an asterisk will not be considered further as they were obtained from larvae put by accident into a medium containing a slight growth of mould.

Table 1 shows that feeding reduces $T12$ by a factor of 6·33-7·0.

For starved larvae the following figures are available:

It is clear that while K_out and [Na_in] remain constant, K_in∝ (1/[Na_ont]). This means that identical amounts of sodium exchange in the three different media.

The sodium in the haemolymph may exchange with that in the medium through the anal papillae, the gut and the general body surface. As far as the permeation of sodium is concerned the larva may be regarded, as a first approximation, as three resistances in parallel, in which case the sum of the permeabilities of the three tissues should equal the permeability of the intact larva. An estimate of the exchange occurring through the papillae can be obtained by using larvae which have had the papillae destroyed by 5 % NaCl (Wigglesworth, 19336). This treatment makes the sodium concentration in the haemolymph of both fed and starved larvae drop to about 80 mM/1. (see Koch, 1938) but it then remains steady at this level. The exchange uptake of sodium in fed and starved papilla-less larvae is shown in Figs. 2 and 3 and the results are summarized in Table 2. Feeding reduces $T12$ by a factor of 3·14—6·6. A comparison of Tables 1 and 2 shows that in the case of fed larvae destruction of the papillae reduces K_out to 7·29-14·6% and K_in to 4 · 15-11·7% of the normal value. In the case of starved larvae K_out is reduced to 14-16·2% and K_in to 11-13·7% of the normal value. The discrepancy between the alteration in K_OTjt and is due to the drop in haemolymph sodium caused by the destruction of the anal papillae. These results show that roughly 90 % of the exchange of sodium occurs through the papillae. This estimate is supported (in the case of starved larvae) by data for the exchange in larvae with the gut blocked by means of a drop of Whitehead’s varnish put on the mouthparts. Although these larvae were adversely affected by the varnish the rate of exchange was reduced only by a relatively small amount (Fig. 2). Treherne (1954) showed that the exchange through the general body surface was very small.

To summarize:

1. Although feeding to satiety does not cause an appreciable alteration in the sodium level in the haemolymph it increases the rate of exchange by a factor of 6·33-7

2. Roughly 90 % of the exchange in both fed and starved larvae occurs through the anal papillae and the rest occurs through the gut and general body surface. Therefore roughly 90% of the increase in exchange rate due to feeding occurs through the anal papillae.

The outflux of labelled sodium from the haemolymph of starved larvae placed in flowing distilled water is shown in Fig. 1. The larvae are able to retain the sodium to a very considerable extent. At about 140 hr. a rapid drop in the sodium level occurs and the larvae start to die. This experiment confirms and extends information given by Treherne (1954).

Wigglesworth (1933 a) gives a detailed description of the anal papillae of larvae reared and presumably starved in fresh water. I have found that considerable cytological changes follow upon feeding and starvation. The papillae studied were fixed in Carnoy’s fluid and stained with iron haematoxylin and eosin. Papillae of fed larvae are illustrated in Fig. 4A and those of starved larvae in Fig. 4B. The cytoplasm of fed papillae is much thicker than that of starved ones. The nuclei of fed papillae are usually rather irregular in shape and are found inside large vesicles. This suggests that they have shrunk considerably during preparation and that when living they are much larger than those of starved papillae. As far as can be judged by studying the surface of the papillae the density per unit area of the striations is about the same in the fed and starved papillae.

In larvae reared and kept in the medium the distribution of Na, K and Cl between the haemolymph and the medium is as follows:

Clearly the larva can bring about a very considerable accumulation of these elements. It can also keep the levels of these elements in the haemolymph remarkably constant when their levels in the medium are varied between < 1 and 80100 mM/1. (Wigglesworth, 1938; Ramsay, 1953). Furthermore larvae made deficient in these elements can restore them to their normal levels in the face of a steep concentration gradient (Wigglesworth, 1938; Ramsay, 1953; Stobbart, unpublished). Because of these facts, and by comparison with the situation in other tissues, it seems likely that Na is actively pumped into the haemolymph. In what follows I shall assume that this is so. I shall also assume, as a first approximation, that the larva consists of only the haemolymph compartment (see Treheme, 1954).

About 90% of the steady-state exchange occurs through the anal papillae and the rest occurs mainly through the gut. It will be shown in a further paper that almost all the net transport by sodium-deficient larvae occurs through the papillae, so it would be simplest to suppose that the steady-state exchange results from an active influx through the papillae which balances a passive outflux through these and other tissues. But by feeding the larvae the fluxes and the exchange rate may be increased by a factor of about 6·5. Consider first the fluxes through the papillae. The relative sizes of the fluxes in fed and starved larvae are shown in Fig. 5 A. If the outfluxes are due to passive diffusion through leaks in some semi-permeable membrane in the papillae, the fed larvae would have to move 6·5 times more sodium inwards to keep the sodium in the haemolymph at the same level. This seems most improbable. It is unlikely that feeding has a bad effect on the semipermeable membrane increasing its in → out permeability by a factor of 6·5; it is also unlikely that the electrochemical activity of the sodium inside fed larvae is 6·5 times greater than in starved larvae when its concentration in both is the same. It is more likely that both influx and outflux are mediated by a carrier mechanism which is linked to the metabolism of the larva, and that when the larvae are fed the carrier is altered so that the fluxes are increased. The carrier is assumed to be confined to some osmotic barrier in the papillae between the haemolymph and the medium. The barrier itself is assumed to be largely impermeable to sodium. According to this view, in fed larvae almost all the influx and at least 85% (= 5·5/6·5) of the outflux are carrier-mediated. In starved larvae almost all the influx is presumably carrier-mediated but there is no evidence as to whether the outflux is carrier-mediated or whether it results from passive diffusion, but presumably some of it results from passive diffusion. This interpretation is shown diagrammatically in Fig. 5-B; it has been assumed arbitrarily that in starved larvae half the outflux is due to passive diflfusion. Treherne (1954) showed that potassium did not compete with sodium for the exchange process, and it has been shown that the exchange is independent of [Na_out] over the range 2-8 mM/1. These observations were only made on starved larvae, but it seems probable they would apply to fed larvae as well. The carrier mechanism therefore must be specific to sodium and saturated with it at the external concentration.

The results are most satisfactorily explained in terms of a sodium pump (that is, a source of free energy) working in conjunction with something similar to an exchange diffusion mechanism (Ussing, 1948) which is confined to the osmotic barrier. The pump is supposed to split off at the inner surface of the barrier sufficient sodium from the ‘exchange diffusion mechanism’ (EDM) to balance any passive loss. The EDM then takes up some sodium from the medium. When the larvae are fed the exchange, but not the pumping of sodium, is increased (Fig. 5B). The exchange component of the fluxes may be regarded as an exchange diffusion in that no osmotic work is associated with it. The increase in exchange could be brought about by (i) an increased rate of movement of the EDMs, (ii) an increase in the amount of sodium carried by each EDM, (iii) a synthesis of more EDMs, (iv) an increase in any possible enzymic catalysis of exchange of sodium between EDM and medium or haemolymph (see Mitchell, 1954 a, b). A similar explanation was offered by Mitchell & Moyle (1953) for the exchange and net transport of phosphate in Micrococcus pyogenes. On the other hand Croghan (1958 a, b) found that in Artemia salina an ionic pump occurs in the gut and an exchange diffusion mechanism in the general body surface.

It has so far been assumed that influx and outflux through the anal papillae are equal. However when the papillae are removed the sodium level in the haemolymph drops from about 100 to 80 mM/1., and larvae which were reared in the medium are unable to bring back the sodium to 100 mM/1. This suggests that the influx through the papillae is normally somewhat greater than the outflux. This is to be expected as the anal papillae are the principal organs for the collection of salts, but it does not invalidate any of the arguments which have been developed.

The exchange occurring through the gut accounts for roughly 10 % of the total exchange in both fed and starved larvae, so clearly feeding greatly increases the exchange through the gut. The very small exchange through the general body surface may be neglected (Treheme, 1954). Although some of the outflux through the gut no doubt occurs as a loss with the urine, it seems probable that most of the outflux occurs through the midgut at any rate in starved larvae, as blocking the mouth stops the medium entering the gut, and it greatly reduces the outflux from starved papilla-less larvae (Treheme, 1954). Arguments similar to the ones developed for the anal papillae may be applied to the gut, and it is possible that a similar sort of Na pump may occur in it. It is, however, not advisable to make too close a comparison between the two tissues at present as it is not yet known what proportion of outflux in papilla-less larvae occurs as losses with the urine and what effect feeding has on these losses. Furthermore, although the gut must be able to do osmotic work (as papilla-less larvae can keep the haemolymph sodium at about 80 mM/1. indefinitely despite any loss with the urine) it will be shown in a further paper that net transport inwards due to the gut occurs only in fed papilla-less sodium-deficient larvae which have been reared in distilled water. This net transport would appear to occur only after a considerable time lapse. However the gut has not been studied in much detail here and a more detailed study might well resolve these uncertainties.

Treheme’s observations on the outflux of sodium from the haemolymph of starved larvae placed in flowing distilled water have been confirmed and extended. The outflux under these conditions is much smaller than it is when the larvae are in the medium. These results are compatible with the sort of sodium carrier and pump suggested earlier. It would be interesting to know how much of this outflux occurs through the anal papillae, the gut, the general body surface, and in the urine.

A considerable difference has been shown to exist between the anal papillae of fed and starved larvae. The exact significance of this difference is not at present apparent, but as there is good evidence that the chief function of the anal papillae is to transport salts, it seems likely that the difference is correlated with the difference in the rates of exchange of sodium. For this reason the cytology of the anal papillae is worthy of further study, preferably by means of fixation with OsO₄ and electron microscopy.

In conclusion it should be pointed out that while for the sake of simplicity it has been assumed that the carrier mechanism occurs in some one barrier (in the anal papillae) between the haemolymph and the medium, the sodium has in fact to move through the internal and external cytoplasmic boundaries and through the cytoplasm itself. The implications of this sort of system are discussed in detail by Ussing (1948). The carrier models that have been suggested here are meant to describe only the overall fluxes of Na. It would be of interest to know the value of any potential differences between the haemolymph and the medium, and to test the effect upon the exchange of a wider range of sodium concentrations in the medium, and of a variety of metabolic inhibitors.

I should like to thank Dr J. A. Kitching under whose supervision this work was done for guidance and encouragement, and Professor J. E. Harris, F.R.S. for his active interest in the problem. I am also grateful to Dr K. Zerahn for some helpful discussion and to Dr K. M. Smith, F.R.S. for letting me use the photographic facilities of the A.R.C. Virus Research Unit, Cambridge.

The isotopes were bought out of a Royal Society grant to Dr Kitching. This work was done while I held a Maintenance Allowance from the Department of Scientifi.c and Industrial Research.