1. The time-course of the fluxes causing net sodium uptake through the anal papillae, and the control of this uptake, have been studied in fed sodium-deficient fourth-instar larvae of the mosquito Aëdes aegypti in 2 mM/1 NaCl.

  2. Influx, outflux and uptake rate all decline exponentially with time, suggesting that effective negative feedback operates continually upon the transporting system.

  3. Uptake is impaired or eventually stopped when certain tissues are destroyed, or ligatured or clamped out of the circulation. The effect is not explicable in terms of impaired circulation of haemolymph through the papillae, and is taken to indicate that a hormone is involved in sodium uptake. Interruption of the nerve cord suggests that an abdominal monitoring centre conveys information forwards along the nerve cord to a control centre in the thoracic ganglia to inhibit or increase hormone production.

  4. The hormone is apparently produced by the thoracic ganglia and the retro-cerebral complex (corpora allata and homologues of the prothoracic glands combined). Evidence that a second hormone, produced in the head, is also involved in sodium uptake is at present inconclusive.

  5. A possible mechanism of hormone action is suggested and discussed in relation to what is known of the functioning of the ionic carriers in the anal papillae.

As a result of numerous investigations during the last 36 years we now have a fairly detailed knowledge of the processes of ionic and osmotic regulation in freshwater mosquito larvae—mainly of Aëdes aegypti (Wigglesworth, 1933 ac, 1938; Koch, 1938; Ramsay, 1950, 1951, 1953; Treherne, 1954; Stobbart, 1959, 1960, 1965, 1967, 1970a, b). It has been shown that the larvae are capable of maintaining ionic balance in very dilute solutions by means of sodium and chloride ‘pumps’ situated in their anal papillae. These pumps actively secrete Na+ and Cl− into the haemolymph and have high specificity and affinity for Na+ and H+, and Cl and (also OH) respectively. They can take up Na+ and Cl jointly or independently, and certainly in the latter, and probably in the former case Na+ is exchanged for H+ and Cl is exchanged for (and possibly OH). The pumps are characterized by high exchange components in the Na+ and Cl fluxes which they generate, and to a first approximation exhibit typical enzyme-substrate kinetics as described by the Michaelis-Menten equation. The passive losses of ions which occur through the general body surface are small. It has proved possible to correlate changes in the activity of the pumps with changes in the cytology of the anal papillae both at the light microscope and the electron microscope level, and such correlations suggest that the pumps are integral parts of membrane-mitochondria associations as seems to be characteristic of secretory cells in general (Stobbart, 1959, 1960; Copeland, 1964; Sohal & Copeland, 1966; Stobbart, 1967).

However, practically nothing has so far been discovered about the control of salt secretion apart from a demonstration that the control is both effective and precise (Stobbart, 1960) and so I undertook the work now to be described in an attempt to throw some light upon the processes responsible for this control in the larva of A. aegypti. In what follows the terms ‘influx’ and ‘outflux’ describe ionic fluxes found with radioactive tracers, and ‘net loss’, ‘loss’, ‘net uptake’ and ‘uptake’ describe net movements of ions.

In view of the recent demonstrations of hormonal control of salt transport by the Malpighian tubules and possibly the recta of various insects (Maddrell, 1963; Berridge, 1966; Mills, 1967; Wall, 1966; Vietinghoff, 1966,1967; Pilcher, 1970) and the involvement of neurohypophysial and other hormones in the processes of ionic transport and balance in amphibia and fish (see Shaw, 1963; Crabbé, 1964; Fanestil & Edelman, 1966; Bentley, 1969; Motáis & Maetz, 1964; Maetz, Bourguet & Lahlouh, 1964; Stanley & Fleming, 1966; Schreibman & Kallman, 1969; etc.) hormonal (or neurohormonal) control of salt transport by the anal papillae of Aëdes larvae was here adopted as a guiding hypothesis. In view of the precision of the control (Stobbart, 1960) it is probably necessary to postulate in addition some sort of centre(s) in the larva which can monitor the salt concentration, and control the output of hormone appropriately. Such a control centre could, in fact, be the hormone-producing tissue itself if it were stimulated to produce by low salt levels and inhibited by high ones. Attempts to identify hormone-producing regions have been made by ligaturing or clamping the larvae at various places and by destroying various tissues (a) by pinching the larvae with fine forceps, (b) by freezing localized regions of the larvae briefly with solid CO2. More extensive surgery was not practicable because of the hydrostatic pressure of the haemolymph, the need for fairly large numbers of operated larvae, and the need for maintenance of the impermeability of the general body surface.

The experiments were all performed at 28°C and, apart from some preliminary work and an investigation into the effects of the surgical procedures upon the permeability of the body wall, used fed fourth-instar larvae (stock L, Stobbart, 1967) reared under salt-deficient conditions as described earlier (Stobbart, 1960). These larvae were actually starving during the experiments which lasted 5–15 h, but this is of no account as it takes at least 20 h for the effect of starvation on salt movements to become apparent (Stobbart, 1958). Fed larvae instead of starved larvae were used almost entirely because (a) the time course of salt uptake in fed larvae is much quicker and therefore technically more convenient, (b) fed larvae terminate net uptake more precisely when an appropriate internal sodium level has been achieved than do starved ones (Stobbart, 1960) which suggests that the control processes are working more effectively in the fed larvae.

In the main body of the work the general design of the experiments was as follows. The salt-deficient larvae were subjected to surgery (where appropriate) and were then placed in 2 mM/1 NaCl for suitable times. They were then removed from this and placed for a very short time in NaCl labelled with 22Na at high specific activity. They were then sampled to follow sodium uptake and any changes in sodium fluxes which occurred as the uptake proceeded. In order to reduce the work to manageable proportions only the uptake and fluxes of sodium were measured, though the larvae must also have been taking up chloride as well (Stobbart, 1967).

In the preliminary work starved larvae of stock OSU (Stobbart, 1967) were used. These larvae were reared normally (Stobbart, 1959) starved for 72 h and then slightly depleted of salts by placing them in de-ionized water at a density of 1 larva/2 ml for various times. They were then placed in either 0·05 HIM/1 Na2SO4 or 0·1 mM/1 KCl, and at appropriate times they and/or the media in which they had been kept were sampled to follow the salt loss sustained by the larvae and the resultant alterations in sodium and chloride fluxes.

To investigate the effect of the surgical procedures on permeability of the body wall larvae of stock L were used which had been reared normally and then starved for 6 d. Operated and control larvae were placed in de-ionized water at a density of 1 larva/2 ml and the loss of sodium to the de-ionized water was followed for suitable times. Two ml of water were removed for each analysis and were replaced with 2 ml of water of the same NaCl concentration. Over the first few hours the loss of sodium will be due almost entirely to passive losses through the anal papillae, the excretory system, and the body wall. The aim of using severely starved, normally reared larvae was to minimize the error in the measurements of loss due to the progressively increased transport inwards of salts by the anal papillae as the external salt concentration builds up.

Measurement of fluxes

(a) Na influx

The larvae were removed, either singly or in groups of five, from the solution of 2 mM/1 NaCl or from the very dilute rearing medium, and placed in 2 miw/l NaCl labelled with 22Na (*NaCl) at high specific activity ( ∼ 1·0 Ci/M Na). They were left in this for 30–60 s during which time sufficient labelled Na entered the larvae to be readily measurable while the amount of net uptake which occurred was negligible. The larvae were now removed from the labelled solution and dried. If the larvae were operated ones their weight had been taken prior to operation; if not, they were now weighed. Next they were ashed (450°C for 5–6 h) on platinum, converted to chlorides, and transferred to planchettes for measurement of radioactivity, 9 μM of dextrose being added to the planchettes as a spreader. The specific activity of the labelled solution was similarly measured and the flux was calculated and expressed as mμM Na/mg wet wt/h; 1 mμM Na/mg, it may be noted, is equivalent to 1 mM/kg; however, as a larva weighs roughly 1 mg (if reared under sodium deficient conditions, 1·66 mg if reared normally) expression of the fluxes as mμM/mg/h gives one a rough estimate of the amount of sodium entering or leaving a larva in this time. In some experiments five larvae were combined into a single group which constituted a sample, in others the five larvae of a sample were analysed separately, in which case the mean for the sample and its standard deviation (S.D.) were calculated. After measurement of radioactivity the samples were transferred back to platinum and the dextrose was ashed off (450°C for 12 h); the samples were then reconverted back to chloride and analysed for sodium. In the preliminary work sodium influx from 0·05 mM/1 Na2SO4 was measured in starved larvae. Here five groups of ten larvae constituted a sample (for which the mean and its S.D. were found), the specific activity of the labelled Na2SO4 was lower, and the influx was measured over a period of 10–30 min; otherwise the techniques used were similar.

(b) Na outflux

For these measurements the larvae were reared in a very dilute medium containing 22Na at high specific activity. The specific activity was not accurately known because inactive Na was added at intervals to the medium with the foodstuff. Upon reaching the fourth instar the larvae were placed in 2 mM/1 *NaCl (specific activity ∼ 1·0 Ci/M Na). The larvae were removed singly from the 2 mM/1 *NaCl, or from the very dilute rearing medium labelled with 22Na, and were placed in 2 ml of unlabelled 2 mM/1 NaCl in a small Polythene tube. They were left in this solution for 30–60 s, during which time a small but measurable amount of labelled Na left the larvae while the amount of net uptake which occurred was negligible. The larvae were then removed, dried and weighed (if this had not already been done—see above) and the radioactivity and the Na they contained were measured as described (in part) above. The lots of 2 ml of unlabelled 2 mM/1 NaCl were now dried down with 9 μM of dextrose on to planchettes for measurement of radioactivity. From a knowledge of the radioactivity and total Na content of the larvae the specific activity of the Na in an individual larva can now be calculated, and from a knowledge of this and the amount of radioactivity which has reached the unlabelled external solution the outflux can be found. Five larvae constituted a sample and for each group the mean outflux and its S.D. were found and expressed as mμM Na/mg/h. In the preliminary work the outflux into 0·05 mM/I Na2SO4 was measured in a similar way, but here the larvae were labelled with 22Na by rearing and then starving them in the presence of *Na of constant and accurately known specific activity, and so it was not necessary to measure directly the specific activity of their sodium. A group of 40 larvae constituted a sample, and was placed for 10-30 min in 80 ml of unlabelled 0·05 mM/1 Na2SO4. After removal and weighing of the larvae the Na2SO4 solution was dried down on to a planchette for measurement of radioactivity, and from a knowledge of this and the specific activity of the sodium in the larvae, the outflux was found and expressed as above. In one case a sample was split into four groups of 10 larvae which were each placed in 20 ml of unlabelled Na2SO4, and then treated as above—this allowed the mean and its S.D. to be found for the sample.

(c) Cl influx from 0·1 mM/1 KCl, Cl outflux into 0·1 mM/1 KCl

These fluxes were only measured in the preliminary work, and the techniques were the same as those used for the sodium fluxes except that: (i) the larvae were labelled with 36C1; (ii) 0·1 mM/1 K *C1 replaced 0·05 mM/1 *NaCl; (iii) measurements of radioactivity in the larvae were made on macerates of ten larvae, five such macerates constituting a sample (see Stobbart, 1967).

Some mention must be made of the accuracy of the flux measurements. In the influx measurements the labelled ion was placed in the outside solution and measurements were made of radioactivity in the larvae. Here it is desirable for the number of ions in the outside solution to be much larger than the number in the larvae, for this minimizes changes in external specific activity due to unlabelled ions emerging from the larvae. This condition was readily met, and as measurements were made over a short period of time ‘back movement’ of labelled ion and hence the error in flux measurement must have been negligible. In the case of the outflux measurements in the main body of the work the situation is less satisfactory. Here the labelled ion was initially in the larvae and measurements of radioactivity were made on the outside solution, and so ideally the number of ions in the outside solution should have been much smaller than that in the larvae in order to minimize changes in specific activity in the larvae. Although here also the flux measurements were made over short periods of time, the ratio no. internal ions/no. external ions was approximately 1/130 (for a 1 mg larva with internal sodium concentration of 30 mM/kg placed in 2 ml of 2 mM/1 NaCl). Furthermore the situation is worsened by the fact that some of the outfluxes measured were high (see Fig. 18), of the order of 300 mμM/mg/h. It is therefore necessary to make some estimate of the errors incurred. Let us take the worst possible case of a 1 mg larva with internal sodium concentration of 30 mM/kg and an outflux of 500 mμM/mg/h, and let us assume that there is no influx in this larva (which in fact is not the case). We can calculate that over a period of 1 min (the duration of the flux measurement) such a larva will have lost less than 3% of its sodium. Over this time therefore we can regard it as being in a steady state, and we may apply to the problem the equations derived by Shaw (1963, table 5.4), which describe the appearance of labelled ion in the external solution for the situations (a) where the sizes of the internal and external ion pools are known, (b) where the internal pool (size known) is very much greater than the external pool (unknown). We find in fact that over a period of 1 min the appearance of ions in our experimental situation (a) is 90% of that in the ideal situation (b). The errors therefore are −10% or less. In the case of the outflux measurements of the preliminary work the errors are negligible at − 1% or less.

Lastly we must note that the flux measurements in the main body of the work will be less accurate than the estimates of internal sodium concentration because of the difficulty of placing the larvae in labelled solutions for accurately measured small intervals of time. Also additional errors will have occurred in the outflux measurements because of the need to estimate the specific activity of the sodium in each larva. It is difficult to estimate the size of these errors but they are likely to be less than 10% in the worst cases.

Measurements of radioactivity

These were made as described earlier (Stobbart, 1965) except that: (i) in some outflux measurements less than 1000 pulses (300–800) were counted where the radio-activity was very low; (ii) in experiments where the treatment of samples took a long time a small correction had to be applied to the data to compensate for the decay of 22Na.

Measurements of sodium and potassium

These measurements were made with the E.E.L. flame photometer. For the measurements of the sodium content of the larvae the larvae was ashed and converted to chlorides before analysis (see above). A correction was applied for the small amount of sodium added with the HC1 during the conversion to chlorides and, where appropriate, for that added with the dextrose which served as a spreader during the measurements of radioactivity. The results were expressed as mM/kg and where the samples consisted of five groups of ten larvae, or five single larvae the means and their standard deviations were found.

Measurements of the chloride content of larvae

These were made with Aminco–Cotlove chloridometer using macerates of larvae (Stobbart, 1965,1967); where possible the standard deviations were found (see above).

Application of ligatures and clamps, destruction of tissues

Larvae were ligatured at various places along the body using fine threads (of negligible sodium content) teased from twist silk, and the simple knots shown in Fig. 1 a. The knots had to be tightened just the right amount otherwise the delicate cuticle of the larvae was cut. Such ligatures, of course, are not releasable—releasable ligatures were applied by the method shown in Fig. 1 b using long human hairs to tie the larvae on to Perspex rods. The rods were then stood upright in 2 mM/1 NaCl by placing one of their ends in a hole in a slab of Perspex on the bottom of a beaker. About 50% of these ligatures could be released without causing obvious tearing of the cuticle. Larvae were clamped by placing them on beds of beeswax resin cement (Krogh & Weis-Fogh, 1951) on microscope slides, and laying over them a fine copper wire (40 gauge) which was then pressed down into the cement on either side of the larva. Fig. 2 illustrates the way in which the technique was used to clamp off the brain, the brain and retrocerebral complex (see Fig. 3) and the brain and thoracic ganglia, from the rest of the body. The retrocerebral complex (RCC) in culicine larvae appears to be known only morphologically (Clements, 1963). It consists of (i) the single corpus paracardiacum situated in the head and less well developed in the larva than in the pupa (see Cazal, 1948 on Culex) and presumably represented in Aëdes by the recurrent ganglion of Christophers (1960); (ii) the paired corpora allata situated in the prothorax (Fig. 3) and surrounded by tissue which is probably homologous with the peritracheal glands of Chironomus (Possompès, 1953) and the prothoracic and pericardial glands of other insects, as these glands are all histologically similar and atrophy after metamorphosis (Bodenstein, 1944; Novák, 1966). The clamping technique used here separates the corpus paracardiacum from the corpora allata and peritracheal glands, and so it is convenient in what follows to designate only the latter two (prothoracic) glands by the term RCC. Thus defined the RCC lies beneath the cuticle of the prothorax and is to some extent accessible to freezing and cautery, and application of these two techniques to it suggests that it is indeed involved in the control of metamorphosis (Stobbart, unpublished). After clamping, the larvae were covered with drops of 2 mM/1 NaCl and placed in a humid atmosphere. The drops of 2 HIM/1 NaCl could easily be replaced with fresh drops if the Na concentration was in danger of being lowered appreciably by uptake, or by drops of 2 mM/1 *NaCl when measurements of influx were required. As Cu stands below Na in the electrochemical series there was no danger of the larvae being poisoned by copper ions.

Fig. 1.

(A) Knots used for tying ligatures at neck and between thorax and abdomen. (B) Method used for applying releasable ligature between thorax and abdomen, x, Perspex rod; 3, long human hair; 3, rubber band.

Fig. 1.

(A) Knots used for tying ligatures at neck and between thorax and abdomen. (B) Method used for applying releasable ligature between thorax and abdomen, x, Perspex rod; 3, long human hair; 3, rubber band.

Fig. 2.

Method for clamping brain and thoracic ganglia or brain and retrocerebral complex (RCC) out of the circulation, 1, brain; a, RCC; 3, thoracic ganglia; 4, abdominal nerve cord; 5, thin copper wire pressed down into bed of beeswax-resin mixture.

Fig. 2.

Method for clamping brain and thoracic ganglia or brain and retrocerebral complex (RCC) out of the circulation, 1, brain; a, RCC; 3, thoracic ganglia; 4, abdominal nerve cord; 5, thin copper wire pressed down into bed of beeswax-resin mixture.

Fig. 3.

Drawing of a fourth instar Aëdes larva from the dorsal side to show the retrocerebral complex and associated structures (modified from Christophers, 1960). 1, brain; 2, recurrent ganglion; 2a, recurrent nerve; 2b, frontal ganglion; 3, nerve to corpus allatum; 4, corpus allatum; 5, peritracheal gland; 6, aorta; 7, neck diaphragm; 8, rudiment of pupal respiratory trumpet; 9, wing rudiments; 10, tracheae; 11, ocelli; 12, rudiment of compound eye; 13, rudiment of antenna; 14, retractor muscles of flabellum; 15, pharynx. The recurrent ganglion lies on the ventral side of the aorta beneath the brain. See also Cazal (1948) for an account of these structures in the pupa of Culex, and Bodenstein (1944) for an account of the corpora allata.

Fig. 3.

Drawing of a fourth instar Aëdes larva from the dorsal side to show the retrocerebral complex and associated structures (modified from Christophers, 1960). 1, brain; 2, recurrent ganglion; 2a, recurrent nerve; 2b, frontal ganglion; 3, nerve to corpus allatum; 4, corpus allatum; 5, peritracheal gland; 6, aorta; 7, neck diaphragm; 8, rudiment of pupal respiratory trumpet; 9, wing rudiments; 10, tracheae; 11, ocelli; 12, rudiment of compound eye; 13, rudiment of antenna; 14, retractor muscles of flabellum; 15, pharynx. The recurrent ganglion lies on the ventral side of the aorta beneath the brain. See also Cazal (1948) for an account of these structures in the pupa of Culex, and Bodenstein (1944) for an account of the corpora allata.

Various tisses were destroyed (or at any rate their functions were severely impaired) by pinching the larvae with fine forceps. The larvae were placed on microscope slides and using forceps that were not too fine (to avoid tearing the cuticle) a fold of epidermis containing the desired tissue (nervous tissue, fat body and limb anlagen) was taken up and crushed. When the thoracic ganglia or the nerve cord in the first abdominal segment were treated in this way the larvae were unable to swim with their characteristic flexing movements—this loss of ability was used as the criterion for successful crushing of nervous tissue. The larvae could still swim for the first hour or so by means of their mouthparts, but later on they became too swollen and turgid to move. They remained alive for at least 24 h, however, as judged by the continued beating of the heart. Fat body and limb anlagen at one side of the thorax were treated in the same way—here it was possible to observe visually some destruction of the tissue. The larvae, however, soon recovered from the operation and swam about normally.

Localized tissue destruction by freezing was accomplished in the following way. Larvae were placed on microscope slides, a piece of solid CO2 was broken and a suitable splinter was grasped with fine forceps whose ends were coated with small sleeves of foam rubber (to allow CO2 to move away freely and thus prevent a build-up of gas pressure between the solid CO2 and the forceps). The solid CO2 was touched briefly against the surface of the larva causing the underlying tissues to freeze in a fraction of a second. The frozen tissue was easily distinguished from normal tissue. Thawing occurred in about the same time upon removal of the solid CO2. As before the thoracic ganglia, the nerve cord in the first abdominal segment and the fat body and limb analagen at one side of the thorax were destroyed, with gross effects on the larvae identical to those described above.

It is of course essential to have adequate controls to establish the possible effects of these rather drastic operations upon the permeability of the body wall and upon sodium movements. Sham-operated larvae were used to achieve this, and in Table 1 is given a summary of the various experimental and sham operations used in this work.

Table 1.

Summary of the experimental and sham operations performed on the larvae

Summary of the experimental and sham operations performed on the larvae
Summary of the experimental and sham operations performed on the larvae

The larvae were caused to swell by most operations which involved the destruction of the thoracic ganglia or the nerve cord in the first abdominal segment, or the isolation of the brain, or brain and thoracic ganglia, from the rest of the body. This swelling will be dealt with in a later paper; here it suffices to state that the swelling is independent of sodium uptake, and causes a weight gain of about 20% in 5 h. Because of this swelling the weight immediately prior to operation was used in calculating sodium fluxes and concentrations in operated larvae.

All the operations described under this heading were performed beneath a binocular microscope. With careful adjustment of the illumination the various tissues could be seen quite easily beneath the cuticle.

Miscellaneous techniques

Larvae were washed with de-ionized water before transference to the various solutions and before analyses were made on them. Drying of the larvae was carried out as described earlier (Stobbart, 1965) and they were then weighed to an accuracy of better than 1% in groups, or singly, on torsion balances of 100 or 5 mg capacity.

The lines through the points in the figures were drawn in by eye unless a statement is made to the contrary, and the usual convention of significance was used in the statistical tests.

(1) Preliminary work; the time-course of activation of the sodium and chloride pumps (starved larvae)

The aim here was to examine the activation process for independent uptake of sodium and chloride. Activation was achieved by placing the larvae for various times in de-ionized water thus depleting them of ions. The resultant fluxes found when the larvae were placed in 0·05 mM/1 Na2SO4 or 0·1 mM/1 KCl are shown in Figs. 4 and 5. Figs. 4A and 5A show the fluxes in relation to time spent in de-ionized water, Figs. 4B and 5B show them in relation to the percentage of the total sodium or chloride lost to de-ionized water. In the case of the chloride data the chloride lost to de-ionized water was not measured directly because of technical difficulties. However, as the sodium and potassium losses are together almost exactly equivalent to the chloride loss (Stobbart, 1967) they were measured instead, and regarded as the chloride loss.

Fig. 4.

The relationship between Na fluxes and Na depletion. ⃝ - - - - ⃝, Influx; ● — ●, outflux. Each point is the mean of 50 larvae (40 in the case of outflux). Here and in other figures the vertical extents of each of the lines emerging from a symbol indicate a standard deviation of the mean (S.D.). The S.D. was found for only one outflux point.

Fig. 4.

The relationship between Na fluxes and Na depletion. ⃝ - - - - ⃝, Influx; ● — ●, outflux. Each point is the mean of 50 larvae (40 in the case of outflux). Here and in other figures the vertical extents of each of the lines emerging from a symbol indicate a standard deviation of the mean (S.D.). The S.D. was found for only one outflux point.

Fig. 5.

The relationship between Cl fluxes and Cl depletion. ⃝ - - - - ⃝, Influx; ● — ●, outflux. Each point is the mean of 50 larvae (40 in the case of outflux). The S.D. was found for only one outflux point.

Fig. 5.

The relationship between Cl fluxes and Cl depletion. ⃝ - - - - ⃝, Influx; ● — ●, outflux. Each point is the mean of 50 larvae (40 in the case of outflux). The S.D. was found for only one outflux point.

It is clear that the activation of the chloride pump is much the most sluggish. In the case of sodium, influx and outflux approach equality after some 4 h of treatment with de-ionized water and the loss of some 2% of the sodium content; the corresponding figures for the chloride fluxes are 17 h and 30%. This is in keeping with the much lower rates of chloride uptake than sodium uptake demonstrated earlier; when activation is achieved the flux measurements are in agreement with those of the earlier work (Stobbart, 1965, 1967).

(2) The effect of the surgical procedures upon the permeability of the body wall

In Fig. 6 are shown the rates at which sodium is lost from severely starved larvae, some of which have been operated upon, when placed at a density of 1 larva/2 ml in de-ionized water. Clearly there is no systematic change in rate of sodium loss over the first 5 h due to the operations, and we may conclude that by themselves they do not appreciably alter the rate of sodium loss through the body wall. Indirect evidence leading to the same conclusion is presented later. There is evidence of considerable sodium loss, however, after 45 h in larvae with the thoracic ganglia pinched. None of the groups achieved a true sodium balance (cf. Stobbart, 1965, fig. 1) and the external sodium concentrations rose to levels higher than those noted earlier, which is to be expected in view of the more rigorous starvation to which the present larvae were subjected.

Fig. 6.

The effect of various operations upon the rate of sodium loss from severely starved larvae to de-ionized water. The larvae were at a density of 1/2 ml. (A) ⃞—⃞, ligatured between thorax and abdomen (10); ⃝—⃝, ligatured at neck (10); ⃝—⃝, sham ligature (round respiratory siphon) (10). (B) ⃝- -⃝, sham-pinching in first abdominal segment (5); ⃞- ⋅ -⃞ nerve cord in first abdominal segment pinched (5); ●—●, unoperated (25); ⃟⋅⋅⋅⃟, thoracic ganglia pinched (15); ◆–⋅–◆ sham pinching at side of thorax (16). The figures in brackets show the numbers of larvae used.

Fig. 6.

The effect of various operations upon the rate of sodium loss from severely starved larvae to de-ionized water. The larvae were at a density of 1/2 ml. (A) ⃞—⃞, ligatured between thorax and abdomen (10); ⃝—⃝, ligatured at neck (10); ⃝—⃝, sham ligature (round respiratory siphon) (10). (B) ⃝- -⃝, sham-pinching in first abdominal segment (5); ⃞- ⋅ -⃞ nerve cord in first abdominal segment pinched (5); ●—●, unoperated (25); ⃟⋅⋅⋅⃟, thoracic ganglia pinched (15); ◆–⋅–◆ sham pinching at side of thorax (16). The figures in brackets show the numbers of larvae used.

(3) The time course of sodium uptake and contemporaneous flux changes in fed sodium-deficient larvae in 2 mM/l NaCl

The time course of sodium uptake has been determined on several occasions (Fig. 7A–F), and the results have always been very similar. The initial sodium concentration in the larvae is about 30 mM/kg and it appears to rise asymptotically to 70–80 mM/kg. About 80% of the uptake is completed in 2 h. The results suggest strongly that an exponential relationship is involved and that at any time the rate of uptake is proportional to the difference between the internal sodium concentration at that time and some final value. The data of Fig. 7D and E (series I, II) are the most extensive and here simultaneous measurements of influx were also made. The relationship between internal sodium concentration, influx, and time, may be shown by assigning a pair of points for each simultaneous measurement of sodium concentration and influx (Fig. 8). The results for influx suggest that here as well there is an exponential relationship between influx and time with the rate of influx falling from an initial value to some final normal value.

Fig. 7.

The time course of sodium uptake from 2 mM/1 NaCl in fed sodium-deficient larvae. Each point is the mean of five larvae; in A D and E the S.D.S were not found. In C and F the different symbols refer to different experiments.

Fig. 7.

The time course of sodium uptake from 2 mM/1 NaCl in fed sodium-deficient larvae. Each point is the mean of five larvae; in A D and E the S.D.S were not found. In C and F the different symbols refer to different experiments.

Fig. 8.

The relationship between internal sodium concentration, sodium influx, and time, in fed sodium-deficient larvae placed in 2 mM/1 NaCl—data of Fig. 7D and E plus simultaneous influx measurements (series I, II). Each pair of points represents mean values for the same group of five larvae, S.D.S were not found. ●—●, Influx (series I); ∎—∎, outflux (series I); ⃝ - - - ⃝, influx (series II); ⃞ - - - ⃞, outflux (series II). In the case of the data for internal sodium concentration the lines drawn through the points are the least squares fits of equation (3), and in the case of the influx data they are the least squares fits of equation (7) (see text).

Fig. 8.

The relationship between internal sodium concentration, sodium influx, and time, in fed sodium-deficient larvae placed in 2 mM/1 NaCl—data of Fig. 7D and E plus simultaneous influx measurements (series I, II). Each pair of points represents mean values for the same group of five larvae, S.D.S were not found. ●—●, Influx (series I); ∎—∎, outflux (series I); ⃝ - - - ⃝, influx (series II); ⃞ - - - ⃞, outflux (series II). In the case of the data for internal sodium concentration the lines drawn through the points are the least squares fits of equation (3), and in the case of the influx data they are the least squares fits of equation (7) (see text).

In the case of internal sodium concentration we apparently have
where Na = internal sodium concentration at t = ∞, Na = internal sodium concentration at any time, k = constant; or
On integration
where Na0 = internal sodium concentration at t = o. Rearranging and taking logs
In the case of influx we apparently have
where I = influx rate at time = ∞, I = influx rate at any time, k = constant; or,
On integration
where Io = influx rate at t = o. Rearranging and taking logs we have
From (4) and (8) we see that plots of
against time provide a method of verifying the assumptions in (1) and (5) as the plots should yield straight lines and estimates of k (slopes of the lines). Unfortunately both Naro and Na and I and Io, must be estimated from the data, so that only a rough test can be performed. If we combine the two sets of data (the validity of this is discussed later) and make reasonable estimates of the above parameters (Na = 71·3, Nao = 29·5 mM/kg; Z = 7·5, Io = 61·5 mμM/mg/h) we may obtain the semilogarithmic plots shown in Fig. 9. Over the first 2 h the data both for internal sodium concentration and influx lie satisfactorily on a straight line; but as time progresses the errors in estimating
Fig. 9.

(A) The relationship between log10[(.II)/(II) and time. (B) The relationship between log10[(Na − Na)/( Na − Na0)] and time. The slopes of the lines × 2·303 equal the (identical) time constants of equations (3) and (7); see text. Data of series I and II combined.

Fig. 9.

(A) The relationship between log10[(.II)/(II) and time. (B) The relationship between log10[(Na − Na)/( Na − Na0)] and time. The slopes of the lines × 2·303 equal the (identical) time constants of equations (3) and (7); see text. Data of series I and II combined.

become larger as Na →Na, and II, and this is reflected in increasing scatter in the data. Because of this increasing error in the logarithmic terms Solomon’s (1952) method was used to estimate the slopes of the lines over the first 2 h. Clearly the slopes are very similar−0·90 and 0·94 h−1 respectively for the influx and sodium concentration data. This is to be expected, for if, as seems to be the case
Then
where R = d(Na)/di at any time and Ro = d(Na)/dt at t = o. Now at any time
i.e.
As both the first and second terms of this equation are simple exponential expressions it follows that the term for outflux must also be a simple exponential expression and that the time constants must be the same in all three terms (see below).

It would appear from Fig. 9 that the assumptions made in (1) and (5) are justifiable and that as a first approximation the relationships of (3) and (7) hold. The lines drawn through the points in Fig. 10A were calculated from equations (3) and (7) using the estimates above, of Na, Na0 and I and Io, and a value of 0·92 h−1 for the time constant; clearly the curves give a very reasonable fit to the data. From equation (10)

Fig. 10.

(A) The data for internal sodium concentration and influx of Fig. 8 (series I and II combined) in relation to time. ●—●, Internal sodium concentration; ⃝—⃝, influx;- - -, outflux. The curves drawn through the data for internal sodium concentration and influx were calculated from equations (3) and (7) using a time constant estimated from Fig. 9; the outflux curve was calculated from equation (11) using the same time constant (see text). ⧫, Internal sodium concentration in groups of five larvae whose respiratory siphons had been denied access to the air by keeping them in small cages of Nylon gauze (S.D.S not found). (B) The relationships between sodium fluxes and internal sodium concentration, and between rate of net sodium uptake and internal sodium concentration. —, I = I0 − [(I0I)/( Na − Na0)] Na −Na,; - - -,0 = I0R0 − [(I0IR0)/ ( Na − Na0)] Na − Na0; - - - - dNa/dt = 7?,− R0 − [(R0)/( Na − Na0)]Na − Na0; or dNa/dt = kNakNa0. These lines were calculated using the estimated parameters used in calculating the curves in Fig. 10A (see text).

Fig. 10.

(A) The data for internal sodium concentration and influx of Fig. 8 (series I and II combined) in relation to time. ●—●, Internal sodium concentration; ⃝—⃝, influx;- - -, outflux. The curves drawn through the data for internal sodium concentration and influx were calculated from equations (3) and (7) using a time constant estimated from Fig. 9; the outflux curve was calculated from equation (11) using the same time constant (see text). ⧫, Internal sodium concentration in groups of five larvae whose respiratory siphons had been denied access to the air by keeping them in small cages of Nylon gauze (S.D.S not found). (B) The relationships between sodium fluxes and internal sodium concentration, and between rate of net sodium uptake and internal sodium concentration. —, I = I0 − [(I0I)/( Na − Na0)] Na −Na,; - - -,0 = I0R0 − [(I0IR0)/ ( Na − Na0)] Na − Na0; - - - - dNa/dt = 7?,− R0 − [(R0)/( Na − Na0)]Na − Na0; or dNa/dt = kNakNa0. These lines were calculated using the estimated parameters used in calculating the curves in Fig. 10A (see text).

and the curve calculated from this equation is also shown in Fig. 10 A. The difference between this curve and the influx curve yields of course the exponentially declining rate of net uptake. The outflux has been measured directly during the period of net uptake and does in fact decline as the internal sodium concentration increases (Fig. 18), a fact known also from earlier work (Stobbart 1960) in which outflux was measured before and after net uptake had occurred.

The foregoing manipulation of the data for internal sodium concentration and influx is useful in establishing the exponential relationship with time, but the curves so obtained are not the best possible fits to the data. Curves of best (least squares) fit may be obtained by the method of Stevens (1951) which has been applied to the two sets of data shown in Fig. 8. The four curves shown in this figure represent the best fits to the data. The analysis of variance is set out in Table 2 and leads to the following conclusions:

Table 2.

Analysis of variance for exponential regression of internal sodium concentration (Na) and influx (I) on time, data of series I and II (see text)

Analysis of variance for exponential regression of internal sodium concentration (Na) and influx (I) on time, data of series I and II (see text)
Analysis of variance for exponential regression of internal sodium concentration (Na) and influx (I) on time, data of series I and II (see text)

(1) There is no significant difference between the two lines of fit to the data for internal sodium concentration, and the joint line is therefore a good fit to the data.

(2) There is a significant difference in the two lines of fit to the influx data, the influx rate dropping more rapidly in series I and following an exponential curve more closely than in series II.

The greater variability in the influx data is presumably due to the difficulty of immersing the larvae in labelled NaCl solution for accurately measured short intervals of time. The difference between series I and II is therefore likely to be due to systematic errors incurred on the two occasions. In view of this I have thought it acceptable to combine the influx data of series I and II in order to get the best overall estimate of the behaviour of influx. Lines of best fit calculated for the combined data of series I and II (see Table 2) are shown in Fig. 11, and resemble closely the lines found by the approximate method (Fig. 10). Note that the lines of best fit for the combined data yield similar though not identical time constants. Unfortunately the data are not extensive enough to allow a test for the significance of the difference between these constants. For the sake of simplicity therefore, and also because it is a reasonable approximation, I shall assume in what follows that they are identical.

Fig. 11.

The data for internal sodium concentration and influx of Fig. 8 (series I and II combined) in relation to time. ●—●, Internal sodium concentration; ⃝—⃝, influx. In the case of the data for internal sodium concentration the line is the least squares fit of equation (3) in the case of the outflux data the line is the least squares fit of equation (7) (see text).

Fig. 11.

The data for internal sodium concentration and influx of Fig. 8 (series I and II combined) in relation to time. ●—●, Internal sodium concentration; ⃝—⃝, influx. In the case of the data for internal sodium concentration the line is the least squares fit of equation (3) in the case of the outflux data the line is the least squares fit of equation (7) (see text).

Assuming that
and
it is useful to consider the relationship between I and Na, and o and Na. Na can vary between Nao and Na, and for any value of Na − Na there is a corresponding value of I. If we consider the changes which have occurred in I and Na during a given period of time clearly
and
as the time constants are identical; on integration therefore
As Io, I, Na and Na0 are supposed to be constant the relationship between I and Na –Na0 should be rectilinear. Similarly we can show that
Lines calculated from (14) and (15) are shown in Fig. 10B using the estimated parameters which were used in calculating the curves in Fig. 10A. The difference between these lines (also shown in Fig. 10 B) of course shows the rectilinear relationship between rate of net Na uptake and Na concentration, and is defined by
(from (14) and (15)): this line is also defined by
hence
This line extrapolated to zero Na intercepts the ordinate at k Na, and the ordinate at Na0(= 29·5 mM/kg) at R0. Note that variations of k in (3) and (7) will cause departures from linearity in (14) and (15), as will variations in the parameters of Io, I Na0 and Na in the case of (14), and in these plus in the case of (15). Other things being equal, however (16) will not be affected by changes in Io and I (i.e. changes in influx and outflux).

All the influx data collected from unoperated larvae of this work are shown plotted against their corresponding values for internal sodium concentration in Figs. 19 and 20. Clearly a reasonable rectilinear relationship prevails over much of the range of internal sodium concentration in agreement with (14). Six larvae, however, with low internal sodium yielded unexpectedly high influx values. The regression line for the data is given in the figures and the correlation coefficient is highly significant (P < 0·001). Fig. 19 also shows the I/Na–Na0 line of Fig. 10B which, as expected, is seen to resemble the regression line closely both in position and slope. The few outflux data available are shown similarly plotted in Fig. 22. Here there is considerable scatter in the data which one could not expect to be as accurate as those for influx since here the internal specific activity had to be found for each larva. So far as they go, the data suggest a rectilinear relationship (cf. (15)) and the correlation coefficient is significant (P = 0·02–0·01). Both the regression line for the data and the o/Na —Na0 line of Fig. 10 B are shown in the figure and, again as expected, they are similar both in slope and position. Details of the regression lines in Figs. 19, 20 and 22 are given in Table 4. From the foregoing considerations it seems reasonable to suppose as a first approximation that k is constant and identical in (3) and (7) and that the constants and coefficients of (14) and (15) likewise do not alter. We have seen from equations (16) and (1) that the difference between the coefficients of (14) and (15) provides an alternative estimate of k; and we deduce from Table 4 that this difference, which is highly significant (Table 5) amounts to 1·0875 h−1 with a S.D. of 0·1212. This estimate is in reasonable agreement with the estimates of k found by means of exponential regression from the combined data of series I and II (Table 2). It would in fact be preferable to estimate k as the difference between these regression coefficients if one could obtain accurate estimates of influx and outflux on the same larvae using 24Na and 22Na. However this, though possible, would be for various technical reasons excessively difficult.

(4) The effect of the surgical procedures upon the net uptake of sodium from 2 mM/l NaCl

When fed sodium-deficient larvae are ligatured between the thorax and abdomen, their ability to take up sodium is greatly impaired (Figs. 12, 17). Under these conditions the internal sodium concentration increases from 20 to 30 to about 40 mM/kg in the first half-hour or so and thereafter remains roughly constant. In the unoperated larvae the rate of uptake is initially more rapid and the internal concentration continues to rise to about 75 mM/kg. The effect is not merely due to ligaturing, for a sham ligature (around the respiratory siphon, Fig. 12) and denial to the respiratory siphon of access to the air (Fig. 10) have no effect. Now the head and thorax constitute 34·8% ± 0·8 (S.D.) of the weight of the larva, and as the general body surface is relatively impermeable to ions (Treherne, 1954) isolation of these regions by the ligature converts them into a dead space which has nevertheless been included in the foregoing analyses. It is useful therefore to examine the situation in isolated abdomens, and a comparison was made between (1) abdomens ligatured off from the rest of the body just before analysis and (2) those ligatured off before the start of net uptake Fig. 13). In both cases the head and thorax were dissected away before the analyses were made. In (1) the time course of net uptake is very similar to that in unoperated larvae (cf. Fig. 7) but the sodium concentrations are some 10 mM/kg higher, which is perhaps not surprising as the abdomens contain the transporting tissue. In (2) the uptake is identical for the first 2 h and thereafter the sodium concentration drops to the initial value. After the first 2 h these abdomens have taken up about 30 mM/kg but as compared to whole larvae this has gone into only about 65% of the body, it would therefore appear as an increase of 30 × 0·65 = 19·5 mM/kg in whole larvae, a figure in reasonable agreement with the increases given in Figs. 12 and 17 for whole larvae ligatured in the same way. According to Fig. 13 the sodium concentration in whole ligatured larvae should drop to the initial value after about 10 h. There is some evidence for a drop in Fig. 12, though not in Fig. 17, but we must note that in all three cases the scatter in the data has increased after 10 h. The behaviour of isolated abdomens in fact is just what we would expect after removing the dead space from whole ligatured larvae.

Fig. 12.

The time course of sodium uptake from 2 mM/1 NaCl in fed sodium-deficient larvae. Each point is the mean of five larvae. ⃝—⃝, Unoperated larvae; ⃞⋅⋅⋅⋅⋅⋅⃞, larvae ligatured at neck; ◑, sham ligatured around respiratory siphon (controls); ⧫ and ●, larvae ligatured at thorax (⧫, S.D.s not found).

Fig. 12.

The time course of sodium uptake from 2 mM/1 NaCl in fed sodium-deficient larvae. Each point is the mean of five larvae. ⃝—⃝, Unoperated larvae; ⃞⋅⋅⋅⋅⋅⋅⃞, larvae ligatured at neck; ◑, sham ligatured around respiratory siphon (controls); ⧫ and ●, larvae ligatured at thorax (⧫, S.D.s not found).

Fig. 13.

The time course of sodium uptake from 2 mM/1 NaCl into the abdomens of fed sodium-deficient larvae. Each point is the mean of five larvae. ⃞—⃞, Abdomens ligatured off and isolated from the rest of the body just before analysis; ⋏- - -⋏, abdomens ligatured off just before the start of uptake and isolated from the rest of the body before analysis.

Fig. 13.

The time course of sodium uptake from 2 mM/1 NaCl into the abdomens of fed sodium-deficient larvae. Each point is the mean of five larvae. ⃞—⃞, Abdomens ligatured off and isolated from the rest of the body just before analysis; ⋏- - -⋏, abdomens ligatured off just before the start of uptake and isolated from the rest of the body before analysis.

Several attempts were made to demonstrate uptake of sodium in larvae after removal of a releasable ligature between thorax and abdomen. The difficulty here is that although there may be no obvious tearing of the cuticle after the ligature has been released the permeability of the body wall may have been greatly increased, and it is probably for this reason that most of the attempts were not successful. Eventually the following procedure was adopted. Ten experimental larvae were weighed and releasable ligatures were applied to them (Fig. 1 B). They were then placed in 2 mM/1 NaCl for about 0-75 h after which time their sodium concentration would have reached roughly the highest level it could, and would have been comparable to that of unoperated larvae (Figs. 12, 17). The ligatures were now released and the larvae were placed separately in 5 ml. de-ionized water at room temperature (22°C) for about 1 h. This water was kept and later analysed for sodium, while the larvae were returned to 2 mM/1 NaCl (at 28°C) for a further 9 h. At the end of this time they were analysed for sodium.

These larvae were compared with ten unoperated larvae which were treated in a similar manner. The analyses of the larvae therefore show if uptake has occurred, and the analyses of the water show if the ligatures have altered the permeability of the body surface. The results are summarized in Table 3. From this we see that unligatured larvae increased their sodium concentration by some 49 mM/kg during the 10 h stay in NaCl solution, and that in all but one of the larvae whose ligatures were successfully released there was a very considerable increase (about × 7) in the permeability of the body wall. Only two of these larvae were able to take up sodium and these were the ones which had sustained the smallest increases in permeability. The average rate of loss to de-ionized water by larvae 2–5 is 25·8 mμM/mg/h which may be compared with an initial rate of net uptake in unligatured larvae of about 37·5 m/tM/mg/h (Fig. 10); as larvae with the ligature released are comparable with larvae with the nerve cord in the first abdominal segment pinched, and which show a depressed rate of net uptake (see below and Fig. 14), it is not surprising that they were unable to take up sodium. In fact they lost it; and as it was noted at the time of sampling that their cuticles had become blackened where they had been compressed by the ligature, it is very likely that they were unable to repair the cuticle and so suffered further increases in permeability. Larvae 6 and 10 took up sodium to levels comparable with, though lower than, those achieved by larvae with the nerve cord pinched (Fig. 14). This might be due to further increases in permeability, or to practically complete transection of the nerve cord (which was easily seen) as compared to more pinching. Overall we may conclude that larvae with the ligature released are capable of a certain amount of net uptake, but that this is readily obscured by increases in the permeability of the body wall due to the ligaturing.

Table 3.

Na concentration of unligatured depleted larvae 26·8 mM/k ± 0·90 (5) (mean ± S.D. (no. of observations))

Na concentration of unligatured depleted larvae 26·8 mM/k ± 0·90 (5) (mean ± S.D. (no. of observations))
Na concentration of unligatured depleted larvae 26·8 mM/k ± 0·90 (5) (mean ± S.D. (no. of observations))
Table 4.

Regression data for influx on internai sodium concentration (I–VII) and for outflux on internai sodium concentration

Regression data for influx on internai sodium concentration (I–VII) and for outflux on internai sodium concentration
Regression data for influx on internai sodium concentration (I–VII) and for outflux on internai sodium concentration
Table 5.

The significance (P) of the differences between (a) the regression coefficients for influx on internal sodium concentration for unoperated larvae and some operated ones, (b) the regression coefficients for influx and outflux on internal sodium concentration in unoperated larvae, (c) the regression coefficients for outflux on internal sodium concentration in unoperated larvae and some operated ones

The significance (P) of the differences between (a) the regression coefficients for influx on internal sodium concentration for unoperated larvae and some operated ones, (b) the regression coefficients for influx and outflux on internal sodium concentration in unoperated larvae, (c) the regression coefficients for outflux on internal sodium concentration in unoperated larvae and some operated ones
The significance (P) of the differences between (a) the regression coefficients for influx on internal sodium concentration for unoperated larvae and some operated ones, (b) the regression coefficients for influx and outflux on internal sodium concentration in unoperated larvae, (c) the regression coefficients for outflux on internal sodium concentration in unoperated larvae and some operated ones
Fig. 14.

The time course of sodium uptake from 2 mM/1 NaCl in fed sodium-deficient larvae. Each point is the mean of five larvae. ⃝—⃝, Unoperated larvae; ◑, larvae sham-pinched at side of thorax (controls); ◆, larvae with thoracic ganglia pinched; ●- - - ●, larvae with nerve cord in first abdominal segment pinched.

Fig. 14.

The time course of sodium uptake from 2 mM/1 NaCl in fed sodium-deficient larvae. Each point is the mean of five larvae. ⃝—⃝, Unoperated larvae; ◑, larvae sham-pinched at side of thorax (controls); ◆, larvae with thoracic ganglia pinched; ●- - - ●, larvae with nerve cord in first abdominal segment pinched.

We consider below sodium uptake by larvae ligatured or clamped at the neck, and by larvae clamped diagonally across the thorax (Fig. 2). As the head contributes 14 ± 1·3% and the thorax 20·8 ± 1·5% of the total weight, the dead spaces incurred will be 14% in the first case and about 24·4% in the second case (head +12 thorax). Two possibilities are open in comparing the performance of operated and unoperated larvae: (i) the sodium level in the body posterior to the clamp or ligature is compared with that in the corresponding region of the unoperated larvae, the region being isolated just before analysis; (ii) the performances on a whole-larvae basis are compared. The second alternative has been chosen (a) to economize on time and effort, (b) to accommodate those treatments which do not involve the isolation of parts of the body. As the whole-body sodium content of sodium-deficient larvae is 30 mM/kg (Fig. 7) while that of isolated abdomens is slightly more than 40 mM/kg (Fig. 13) we can calculate that about 90% of the total body sodium is in the abdomen, so the sodium dead spaces incurred are acceptably small.

When sodium-deficient larvae are ligatured or clamped at the neck the effect is much smaller than that of a ligature between thorax and abdomen, and is apparent after the first h as a lower rate of net uptake (Figs. 12, 17), but by the 5th h the internal sodium concentration is the same as in unoperated larvae (Figs. 12, 16,17).

When the thoracic ganglia or the nerve cord in the first abdominal segment are pinched, roughly equal appreciable depressions of the rate of net uptake result (Fig. 14). These depressions are roughly comparable to that caused by ligaturing at the neck, but become apparent earlier, and recovery (so far as sodium uptake is concerned) seems to be achieved some 6–7 h later at about 12 h (cf. Figs. 12, 17). Sham pinching has no effect at all.

These results and those obtained from ligatured larvae would appear to implicate the nerve cord and the thoracic ganglia in the process of sodium uptake which was therefore examined in larvae in which these tissues had been destroyed by freezing. The results are shown in Fig. 15, from which we see that sham freezing has no effect.

Fig. 15.

The time course of sodium uptake and contemporaneous changes in sodium influx in fed sodium-deficient larvae in 2 mM/1 NaCl. Each point is the mean of five larvae. Measurements of internal sodium concentration. ⃝—⃝, Unoperated larvae, used also for influx measurements; ●—●, unoperated larvae, used also for outflux measurements; ▄, sham-frozen larvae (side of thorax) used also for influx measurements (controls); □, sham-frozen larvae (side of thorax) used also for outflux measurements (controls); ◊, larvae with nerve cord in first abdominal segment frozen, used also for influx measurements; ◑, larvae with nerve cord in first abdominal segment frozen, used also for outflux measurements; ⧫, larvae with thoracic ganglia frozen, used also for influx measurements; ●, larvae with thoracic ganglia frozen, used also for outflux measurements. Measurements of influx in some of the foregoing larvae. ○, Unoperated larvae; ×, sham-frozen larvae (controls); ▾, nerve cord in first abdominal segment frozen; △, thoracic ganglia frozen. For measurements of outflux see Fig. 18.

Fig. 15.

The time course of sodium uptake and contemporaneous changes in sodium influx in fed sodium-deficient larvae in 2 mM/1 NaCl. Each point is the mean of five larvae. Measurements of internal sodium concentration. ⃝—⃝, Unoperated larvae, used also for influx measurements; ●—●, unoperated larvae, used also for outflux measurements; ▄, sham-frozen larvae (side of thorax) used also for influx measurements (controls); □, sham-frozen larvae (side of thorax) used also for outflux measurements (controls); ◊, larvae with nerve cord in first abdominal segment frozen, used also for influx measurements; ◑, larvae with nerve cord in first abdominal segment frozen, used also for outflux measurements; ⧫, larvae with thoracic ganglia frozen, used also for influx measurements; ●, larvae with thoracic ganglia frozen, used also for outflux measurements. Measurements of influx in some of the foregoing larvae. ○, Unoperated larvae; ×, sham-frozen larvae (controls); ▾, nerve cord in first abdominal segment frozen; △, thoracic ganglia frozen. For measurements of outflux see Fig. 18.

Freezing the nervous tissue, however, greatly increases the variability of the results, and may depress the rate of sodium uptake and cause drastic losses of sodium after about 15 h.

An attempt was now made to discover if the retrocerebral complex (RCC) is concerned with sodium uptake. As mentioned earlier the RCC (Fig. 3) consists of the corpora allata and tissues thought to be homologous to the prothoracic glands of other insects. Accordingly, sodium uptake was measured in (i) larvae with the RCC and brain clamped out of the circulation, (ii) larvae with the thoracic ganglia and brain clamped out, (iii) larvae with the brain alone clamped out, and (iv) normal larvae. The results are shown in Fig. 16, and show that clamping out of the brain and RCC, and of brain and thoracic ganglia, cause roughly equal and noticeably greater depressions in the rate of uptake than does clamping out of the brain alone (Figs. 12, 17), and that recovery from these operations is slower than in the latter case, for after 10 h the sodium in the operated larvae is still significantly lower than in normal larvae and those with the brain alone clamped out (Fig. 16).

Fig. 16.

The time course of sodium uptake and contemporaneous changes in sodium influx in fed sodium-deficient larvae in a mM/1 NaCl. Each point is the mean of four or five larvae. ⃝—⃝, Unoperated larvae (internal sodium concentration); □, larvae clamped at neck-controls (internal sodium concentration); ●- - -●, larvea with thoracic ganglia and brain clamped out (internal sodium concentration); ◆ larvae with retrocerebral complex and brain clamped out (internal sodium concentration); ◐⋅-⋅-⋅-⋅◐, unoperated larvae (influx); ▄, larvae clamped at neck-controls (influx); ×, larvae with thoracic ganglia and brain clamped out (influx); ___, larvae with retrocerebral complex and brain clamped out (influx).—, The influx expected from operated larvae (●, △) assuming that the relationship between sodium concentration and influx is the same as in normal larvae.

Fig. 16.

The time course of sodium uptake and contemporaneous changes in sodium influx in fed sodium-deficient larvae in a mM/1 NaCl. Each point is the mean of four or five larvae. ⃝—⃝, Unoperated larvae (internal sodium concentration); □, larvae clamped at neck-controls (internal sodium concentration); ●- - -●, larvea with thoracic ganglia and brain clamped out (internal sodium concentration); ◆ larvae with retrocerebral complex and brain clamped out (internal sodium concentration); ◐⋅-⋅-⋅-⋅◐, unoperated larvae (influx); ▄, larvae clamped at neck-controls (influx); ×, larvae with thoracic ganglia and brain clamped out (influx); ___, larvae with retrocerebral complex and brain clamped out (influx).—, The influx expected from operated larvae (●, △) assuming that the relationship between sodium concentration and influx is the same as in normal larvae.

Fig. 17.

The time course of sodium uptake and contemporaneous changes in sodium influx in fed sodium-deficient larvae in a mM/1 NaCl. Each point is the mean of four or five larvae. ●—●, Unoperated larvae (internal sodium concentration); ◆- - -◆, larvae ligatured at neck (internal sodium concentration); ▲⋅-⋅-⋅-⋅▲, larvae ligatured at thorax (internal sodium concentration); ◝-⋅⋅-◝ unoperated larvae (influx); ◊, latvae ligatured at neck (influx); △, larvae ligatured at thorax (influx). ___,_,__ the influxes expected from larvae ligatured at the thorax and neck respectively assuming that the relationship between sodium concentration and influx is the same as in normal larvae.

Fig. 17.

The time course of sodium uptake and contemporaneous changes in sodium influx in fed sodium-deficient larvae in a mM/1 NaCl. Each point is the mean of four or five larvae. ●—●, Unoperated larvae (internal sodium concentration); ◆- - -◆, larvae ligatured at neck (internal sodium concentration); ▲⋅-⋅-⋅-⋅▲, larvae ligatured at thorax (internal sodium concentration); ◝-⋅⋅-◝ unoperated larvae (influx); ◊, latvae ligatured at neck (influx); △, larvae ligatured at thorax (influx). ___,_,__ the influxes expected from larvae ligatured at the thorax and neck respectively assuming that the relationship between sodium concentration and influx is the same as in normal larvae.

(5) The effect of the surgical procedures upon the fluxes contemporaneous with sodium uptake from 2 mM/I NaCl

We have so far considered only the alterations in net uptake caused by surgery. In fact, however, in all cases of surgery, except for the pinching of nervous tissue, measurements of influx were made simultaneously with the measurements of internal sodium concentration and on the same groups of larvae. In the case of larvae with nervous tissue frozen a similar series of outflux measurements was made in addition. The results of these measurements have been incorporated into Figs. 1518.

Fig. 18.

The time course of changes in sodium outflux contemporaneous with sodium uptake. The larvae used were the ones whose internal sodium concentrations are illustrated in Fig. 15. Each point is the mean of four to five larvae. The logarithmic scale on the ordinate is used merely for convenience. ◝—◝, Unoperated larvae; □, sham-frozen larvae (controls); ●-⋅-⋅-●, nerve cord in abdominal segment I frozen; ◆---◆, thoracic ganglia frozen.

Fig. 18.

The time course of changes in sodium outflux contemporaneous with sodium uptake. The larvae used were the ones whose internal sodium concentrations are illustrated in Fig. 15. Each point is the mean of four to five larvae. The logarithmic scale on the ordinate is used merely for convenience. ◝—◝, Unoperated larvae; □, sham-frozen larvae (controls); ●-⋅-⋅-●, nerve cord in abdominal segment I frozen; ◆---◆, thoracic ganglia frozen.

If we consider the data for net uptake and influx in Figs. 1517 we can see that it would in principle be possible, for any of the treatments, to find the time course for changes in outflux. We should need to differentiate graphically the curves for net uptake at various times and to subtract the net uptake rates so obtained from the appropriate influx rates. Needless to say the data are not of sufficient quality to warrant this, but we can nevertheless arrive at some semiquantitative information about the outfluxes.

The first point to note concerns the data for sham-frozen larvae illustrated in Figs. 15 and 18. Consider first the larvae used for measurements of internal sodium concentration and influx which are symbolized by □ and × in Fig. 15. We see that the results for both influx and internal sodium concentration in sham-frozen larvae are extremely similar to the corresponding results from unoperated larvae (o and ○ in Fig. 15) with regard to both mean values and scatter. This can only mean that for both groups the influx and outflux values must be extremely similar—i.e. the freezing process does not appreciably alter the permeability of the body wall. Similar considerations apply to the larvae used for measurements of internal sodium concentration (▄ of Fig. 15) and outflux (□ of Fig. 18). Here the outflux data are slightly higher than those from the unoperated larvae. However, as the accuracy of the outflux measurements is less than that of the influx measurements and as the discrepancy is in any case small, we may reasonably conclude overall that sham freezing has a negligible effect on the permeability of the body wall.

We may now consider the effects upon the fluxes of freezing nervous tissue in the thorax and abdomen. The effects of these two treatments are broadly similar. The net uptake of sodium may be depressed and drastic losses of sodium may occur after 15 h, and in general the variability of the data is dramatically increased. This is all in marked contrast to the behaviour of influx in the operated larvae which from 5 h onwards is practically identical to its behaviour in unoperated larvae and the sham-frozen controls (Fig. 15). This can only mean that the operations cause large and erratic increases in outflux as is shown to be the case in Fig. 18 (note that the high values of outflux shown here will be some 10% too low).

Turning now to the case of clamped larvae we see from Fig. 16 that the behaviour of influx in the experimental larvae is apparently identical with that in the unoperated larvae and in the controls clamped at the neck. The internal sodium concentrations in the experimental larvae (thoracic ganglia and brain clamped out, RCC and brain clamped out), however, are significantly lower than those of unoperated larvae and controls and in many cases show more scatter. We must conclude therefore that in the experimental animals outflux did not decline so rapidly from its initial value as in the unoperated larvae and the controls. Note, however, that after 4 h the experimental larvae showed a slow net uptake of about 3 mμM/mg/h, so that on average the outflux in these larvae was at this time 3 mμM/mg/h lower than the influx.

A similar situation is shown in Fig. 17 for larvae ligatured at the thorax. The behaviour of influx in the unoperated larvae and the controls (ligatured at the neck) is practically identical, and the behaviour of influx in the experimental larvae is similar though it probably declined more rapidly (as we would expect from the lower rate of net uptake in these larvae). After 1 h the internal sodium concentration of the experimental larvae remained practically constant, so did the influx and, so it follows, did the outflux. Between and 2 h the net uptake of the control larvae temporarily halted, and then started up again being for the period 2–4 h about 8 mμM/mg/h. The influx data suggest that during this period the outflux must have been reduced to a very low level, but more observations would be needed to establish this definitely.

We may summarize the results presented so far in this section as follows: (i) In unoperated larvae the rate of net uptake decfines exponentially as a consequence of the exponential declines in influx and outflux which have been described earlier, (ii) Surgical procedures seem to have little effect on the behaviour of influx except in the case of larvae ligatured at the thorax (in which the rate of decline is initially more rapid than normal). This apparent lack of effect on influx could in part be due to the fact that readings were not taken sufficiently soon after surgery to reveal possible differences between normal and experimental larvae. It is, however, certain that the circumstances relating to the decline of influx in the two categories of larvae are different as the relationship between internal sodium concentration and influx is altered by the surgery. This is shown in Figs. 16 and 17, in which the thick lines show the decline in influx with time to be expected from operated larvae assuming that the sodium concentration/influx relationship has not been altered. Clearly the observed values of influx are much lower than the expected ones. This point is also illustrated though in a different way by the influx/sodium concentration plots of Figs. 19-21. (iii) The recovery after approximately 1 h in the rate of sodium uptake in clamped larvae and larvae ligatured at the neck is due to reductions in the outflux and not to increases in the influx which by this time is approaching its minimum value, (iv) The variability in the internal sodium concentrations of larvae with nervous tissue frozen is due to large and erratic increases in outflux. (v) The stoppage of net uptake after about i h in larvae ligatured at the thorax is due to the outflux at this time approximating to the influx (which is similar to that of unoperated larvae). Large erratic increases in outflux apparently do not occur (Fig. 17) though later at about 3 h outflux may exceed influx (Figs. 12, 13).

Fig. 19.

The relationship between influx and internal sodium concentration. The points are for individual larvae except for 3, which are the means of 5. Circular symbols — (regression line), measurements made on unoperated larvae on all the different occasions (◑, series I and II combined),………., I/(Na–Na,) line of Fig. 10B; ■- - -■, larvae with nerve cord frozen in abdominal segment I (▲ ◊ of Fig. 15); ☐- - -☐, larvae with thoracic ganglia frozen (△ ⧫ of Fig. 15). The broken line represents the mean influx value for the combined data from larvae with nervous tissue frozen. ⧫ Sham-frozen larvae-controls ( ×, □ of Fig. 15).

Fig. 19.

The relationship between influx and internal sodium concentration. The points are for individual larvae except for 3, which are the means of 5. Circular symbols — (regression line), measurements made on unoperated larvae on all the different occasions (◑, series I and II combined),………., I/(Na–Na,) line of Fig. 10B; ■- - -■, larvae with nerve cord frozen in abdominal segment I (▲ ◊ of Fig. 15); ☐- - -☐, larvae with thoracic ganglia frozen (△ ⧫ of Fig. 15). The broken line represents the mean influx value for the combined data from larvae with nervous tissue frozen. ⧫ Sham-frozen larvae-controls ( ×, □ of Fig. 15).

Fig. 20.

The relationship between influx and internal sodium concentration. The points are for individual larvae. Circular symbols —, as in Fig. 19. ■- - -■, larvae with thoracic ganglia and brain clamped out, the line represents the mean influx value (×, ● of Fig. 16); ☐-⋅-⋅-☐, larvae with retrocerebral complex and brain clamped out (regression line) Ũ, ⧫ of Fig. 16. ⧫ larvae clamped at neck-controls (▄,□ of Fig. 16).

Fig. 20.

The relationship between influx and internal sodium concentration. The points are for individual larvae. Circular symbols —, as in Fig. 19. ■- - -■, larvae with thoracic ganglia and brain clamped out, the line represents the mean influx value (×, ● of Fig. 16); ☐-⋅-⋅-☐, larvae with retrocerebral complex and brain clamped out (regression line) Ũ, ⧫ of Fig. 16. ⧫ larvae clamped at neck-controls (▄,□ of Fig. 16).

We must now consider the relationship between these influx (and outflux) data and internal sodium concentration in the various categories of larvae. We have seen that in unoperated larvae there are reasons for expecting a rectilinear relationship as indeed seems to be the case (Figs. 19, 20, 22). All the sets of data are illustrated in Figs. 1922 and appear to be suitable for analysis by means of linear regression. We see from some of these figures that the various sham-operated or control larvae do not appear to differ significantly from the unoperated larvae, but it would clearly have been desirable to make influx measurements over a wider range of internal sodium concentrations in the case of the shams and controls. The results of the regression analyses are given in Table 4. Calculation of the regression of the fluxes on internal sodium concentration is justifiable on the grounds that the larvae are clearly responding to losses of sodium from their bodies. We see from the Table that the correlation coefficient and the regression coefficient for the influx data of unoperated larvae (category I) are highly significant, and that larvae ligatured at the neck (II), or with the RCC and brain clamped out (III), also yield a significant correlation between influx and internal sodium concentration. In contrast all those treatments (IV-VII) fail to yield a significant correlation which involve destroying the thoracic ganglia, or the nerve cord in the 1st abdominal segment, or clamping or ligaturing the thoracic ganglia out of the circulation. So far as the data in categories IV–VII are concerned, therefore, influx is not related to internal sodium concentration. It is therefore justifiable to calculate the means and their standard deviations for the influx values of each of these categories in order to see whether significant differences exist between them in this respect; in fact none do, and we can conclude that the sets of data do not differ significantly. It is still possible, however, that the combined data of these categories may yield a small though significant correlation between influx and internal sodium concentration, but Table 4 shows that in fact no such correlation exists. The lines drawn in Figs. 1921 through the points of categories IV–VII therefore merely indicate the mean influx values.

The significance of the differences between the regression coefficients of I–III are shown in Table 5. As might be expected, the coefficient for the unoperated larvae differs significantly from those of the two operated categories which do not, however, differ significantly between themselves.

The relationship between outflux and internal sodium concentration in unoperated larvae is shown in Fig. 22. There is, as mentioned earlier, considerable scatter in the data, but they are at any rate compatible with a rectilinear relationship and they yield a significant correlation coefficient and regression coefficient (Table 4). The data from larvae with the thoracic ganglia frozen show a great deal of scatter and some very high values for outflux (Fig. 22). Here there is no significant correlation between outflux and internal sodium concentration. However, the data from larvae with the nerve cord in the first abdominal segment frozen apparently yield a significant correlation (Table 4), and the regression coefficient for these data differs significantly from that obtained from unoperated larvae (Table 5). Inspection of the data reveals that the significance of the correlation depends entirely on the inclusion of the single pair of observations which contains the very high value for outflux, and in view of the highly erratic nature of the increases caused in outflux by freezing nervous tissue (Figs. 15, 18, 22) this apparent significance must be considered fortuitous—at any rate there is no significant correlation with internal sodium concentration over the range 60–105 mM/kg. Assuming no correlation, it is not surprising, in view of the scatter, to find no significant difference between the mean influx values of the two sets of data from frozen larvae.

Fig. 21.

The relationship between influx and internal sodium concentration. The points are for individual larvae. —, Regression line for unoperated larvae; symbols omitted for clarity, … indicates area occupied by symbols (see Fig. 19); ⃟,Ũ, larvae ligatured at neck (regression line) ◊, ⧫ of Fig. 17; ∎- -∎, larvae ligatured at thorax, the line represents the mean influx value (△, ▲ of Fig. 17).

Fig. 21.

The relationship between influx and internal sodium concentration. The points are for individual larvae. —, Regression line for unoperated larvae; symbols omitted for clarity, … indicates area occupied by symbols (see Fig. 19); ⃟,Ũ, larvae ligatured at neck (regression line) ◊, ⧫ of Fig. 17; ∎- -∎, larvae ligatured at thorax, the line represents the mean influx value (△, ▲ of Fig. 17).

Fig. 22.

The relationship between outflux and internai sodium concentration. The points are for individual larvae. ●—●, Normal larvae (regression line), ○ of Fig. 18, ● of Fig. 15; …., outflux/(Na–Nao) line of Fig. 10B; ▄, larvae with thoracic ganglia frozen, ⧫ of Fig. 18, ● of Fig. 15; □, larvae with nerve cord in first abdominal segment frozen, ● of Fig. 18, ◑ of Fig. 15; ○, sham-frozen larvae (controls), □ of Fig. 18, □ of Fig. 15.

Fig. 22.

The relationship between outflux and internai sodium concentration. The points are for individual larvae. ●—●, Normal larvae (regression line), ○ of Fig. 18, ● of Fig. 15; …., outflux/(Na–Nao) line of Fig. 10B; ▄, larvae with thoracic ganglia frozen, ⧫ of Fig. 18, ● of Fig. 15; □, larvae with nerve cord in first abdominal segment frozen, ● of Fig. 18, ◑ of Fig. 15; ○, sham-frozen larvae (controls), □ of Fig. 18, □ of Fig. 15.

Figs. 15, 18 and 22 pose an apparent paradox, in that they show that larvae with nervous tissue frozen can have roughly normal internal concentrations of sodium while exhibiting very high values of outflux. As influx in such larvae apparently behaves normally (Fig. 15) we must conclude that the increases in outflux start erratically several hours after the freezing and that the measurements were made before internal sodium concentrations had dropped appreciably.

In order to make consideration of the data easier the main findings of this work have been summarized briefly in Table 6.

Table 6.

Summary of the behaviour of uptake, influx, and outflux in normal and operated larvae

Summary of the behaviour of uptake, influx, and outflux in normal and operated larvae
Summary of the behaviour of uptake, influx, and outflux in normal and operated larvae

When expressed on a whole-larva basis the rate of sodium uptake declines exponentially to zero. Earlier results (Stobbart, 1960), in which uptake was expressed in terms of haemolymph sodium concentration, gave a somewhat different relationship between uptake and time which approximated to a fast uptake rate, followed abruptly after 12 h by a slow uptake rate. In both cases, however, uptake is practically complete after 5 h; it therefore seems that sodium is preferentially secreted into the haemolymph initially, in all probability to remove the stress imposed on nerve-muscle physiology by the low haemolymph concentration.

(A) The effects of ligatures upon the circulation

The ability of larvae to take up sodium is greatly impaired when they are ligatured between thorax and abdomen, and, as is also the case in Chironomus larvae (D. A. Wright, unpublished), uptake is possible after release of the ligature. As a ligature at the neck appears to have a relatively small effect upon uptake, it would appear that some factor originating principally in the thorax and transported (partly at least) in the haemolymph is essential for uptake. However, before accepting this conclusion the effect of the ligatures upon the circulation of the haemolymph through the anal papillae must be considered, as interruption of the circulation would presumably reduce the supply to the transporting tissue of the papillae of any hormone which might be produced in the abdomen. The heart of the larva possesses in the anterior half of each of the abdominal segments I–VII a pair of laterally placed ostia. The haemolymph is passed forwards along the heart to its modified anterior region, the aorta, which conducts the haemolymph forward through the thorax to the head capsule. The aorta has no ostia (Christophers, 1960; Jones, 1954). After entering the head capsule the haemolymph flows through the neck to return to the haemocoel in the body. It is clear therefore that so far as the circulation of haemolymph is concerned a ligature at the neck is practically the same as a ligature at the thorax and both these ligatures do in fact reduce to a similar extent the speed at which the haemolymph circulates, though their effects upon net uptake are very different. Circulation in ligatured larvae is presumably made possible by some haemolymph leaking out through the anterior ostia during each contraction cycle of the heart. The haemolymph circulates through the papillae in an orderly way, up one side and down the other, and this circulation is readily observed after pinching the thorax of a larva; this causes fat body cells to be dislodged into the haemolymph, and their progression round the body and through the papillae may be followed easily. In the case of larvae ligatured between thorax and abdomen the application of the ligature alone often dislodges the cells. The circulation of haemolymph through the papillae appears not to be a continuous process—the papillae seem to be closed off from the haemocoel for appreciable periods, an observation which perhaps receives support from the higher concentration of sodium in the abdomen than in the whole body (Fig. 13). The foregoing considerations rule out interruption of circulation as an explanation for the effects of a ligature between thorax and abdomen, and the observations on uptake of sodium by isolated abdomens (Fig. 13) also argue against such an interpretation. If the effect of the ligature were merely to stop or reduce the flow of haemolymph through the papillae we should expect uptake to proceed until the concentration of sodium in the papillae reaches a level at which its out-wardly directed electrochemical gradient reduces or stops further uptake. Instead we find a certain amount of net uptake in the isolated abdomens followed by a drop back to roughly the initial sodium concentration, so clearly other factors are at work. It is relevant to note here that the simplest way to explain the observed asymptotic increase of internal sodium is to suppose that sodium is transported inwards against a consequentially increased electrochemical gradient until further uptake is impossible. However, such a model would predict a maximum outflux at maximum internal sodium concentration and exactly the opposite is observed; furthermore, the few available measurements of potential (Stobbart, 1965) show that the electrochemical gradient for sodium, which is always outwardly directed, is smallest at the start of uptake when the outflux is largest (present work) and largest at the conclusion when the outflux is smallest. We may therefore reasonably conclude that some factor (hormone) originating principally in the thorax and transported (partly at least) in the haemolymph is in fact essential for sodium uptake, and that an explanation of the hormonal effects upon uptake and fluxes of sodium must be sought in terms of the type of carrier mechanism proposed earlier (Stobbart, 195967). In this type of mechanism ion-carrier molecules were supposed to diffuse between the inner and outer surfaces of, and be confined to, an ion-impermeable osmotic barrier located in the inpushings of the outer plasma membrane of the anal papillae. The carriers have a high affinity for both H+ and Na+, but under normal circumstances would be occupied almost entirely by Na+ because of the very high concentrations of Na+ (relative to H+) in the haemolymph and the external medium. Under these conditions the carriers mediated 1: 1 exchanges of sodium which would clearly be very similar to exchange diffusion (Ussing, 1947, 1948). Net uptake was supposed to occur when some energy-dependent process located at the inner surface of the barrier temporarily reduced the affinity of the carrier for Na+ while it was at the inner surface allowing H+ from the haemolymph to replace Na+. The carrier thus transported H+ to the outer surface of the barrier where the H+ was lost from the carrier and replaced by Na+ according to the law of mass action and the high ratio of Na+/H+ in the external medium. Upon returning to the inner surface of the barrier the carrier was again converted to the H+-combining form and yielded up its sodium to the haemolymph. The net result was therefore an uptake of Na+ in exchange for H+. Any carrier molecules which did not have their affinity for sodium reduced could mediate sodium exchanges as before (Stobbart, 1967, fig. 12 *). A similar model causing an exchange of Cl for was also put forward at the same time.

(B) The control of sodium uptake

We shall consider the effects of the surgical procedures upon sodium movements later, and deal first with the control of sodium uptake in unoperated larvae. Any scheme proposed must be compatible with the results of this and earlier work (Stobbart, 1959, 1960, 1965, 1967, 1970a), and the main points to be considered are summarized below, (i) A factor originating from the thorax appears to be necessary for uptake, (ii) A time-lag is found in the stimulation of fluxes and uptake, (iii) The steady-state fluxes can be increased independently of uptake by feeding the larvae, and the uptake rate is also increased by feeding, (iv) Net uptake is normally the result of an increased influx and outflux and is therefore apparently linked to an exchange system, (v) Influx, outflux and rate of uptake are proportional to sodium deficiency (equations (14), (15) and (16) or (1)—these conditions arise as we saw earlier, from the exponential decline of uptake rate and influx with time, and imply that effective negative feedback is operating continually on the transporting system, (vi) A Michaelis-Menten relationship exists, for a fixed degree of sodium-deficiency, between influx, outflux, rate of uptake and sodium concentration, (vii) The sodium taken up is accompanied by chloride or is exchanged largely against H+, and the affinities of ions for the cationic carrier are in the order H+ > Na+ > K+.

A possible explanation of these observations would be to suppose that a hormone, produced in the thorax in proportion to the degree of sodium deficiency and in greater amounts in fed than in starved larvae, in some way mediates both the synthesis of carrier molecules (which combine with sodium according to the law of mass action) and their conversion to cyclical Na+/H+ carriers. In normal (i.e. sodium-replete) larvae production of hormone would presumably continue at a low level in order to ensure production of sufficient carriers to balance the relatively small passive losses of sodium through the integument and in the urine. Negative feedback could be generated by reduced sodium deficiency acting upon the site of hormone production (the hormone being continually degraded in the body) or, if the hormone were persistent, by its being inactivated by the process of sodium uptake. Support for the suggestion that the hormone is persistent comes from a consideration of uptake in larvae ligatured or clamped at the neck (Figs. 12, 16, 17). In normal larvae body volume is under an effective nervous control (Ramsay, 1953; Stobbart, 1970b), but application of a ligature or clamp inhibits urine production so that the larvae swell by osmosis and show an increase in weight of approximately 20% which is complete by 5 h (Stobbart, 1970b). Now the sodium concentrations in the ligatured or clamped larvae have been calculated in terms of the initial (unswollen) weight, but in spite of this the final sodium concentration achieved is the same as in normal unswollen larvae. It is, therefore, as though a certain degree of sodium deficiency elicits an appropriate amount of hormone which then mediates a corresponding amount of uptake irrespective of the simultaneous dilution of the body fluids, and this implies persistence of the hormone and its inactivation in proportion to the amount of sodium taken up. Clearly if the larvae had been responding continually to sodium concentration, the final values would have been approximately 20% higher than normal when calculated in terms of the initial weight. The hormone would appear to be produced by the thoracic ganglia and the RCC (Fig. 16; Table 6).

Such a scheme might, however, be too simple, for Figs. 12 and 17 and Table 6 show that ligaturing the larvae at the neck also affects both the uptake of sodium and the fluxes, and it may be that a factor originating in the head is necessary for maximal rates of uptake, though it could be argued that these effects are caused by the reduced rate of circulation of haemolymph through the papillae.

The nerve cord appears to be implicated in the uptake of sodium as the uptake rate is slowed down and can become very erratic if the thoracic ganglia or the nerve cord in abdominal segment I are injured by pinching or freezing, and the significant (negative) correlation between influx and internal sodium concentration is abolished in larvae so frozen (Fig. 19; Table 6). These observations suggest the presence of a monitoring centre in the abdomen and the conveyance of information from it by way of the nerve cord to a thoracic control centre and thence to the site(s) of hormone production. The correlation between influx and sodium concentration is found in larvae with the brain and RCC clamped out of the circulation and in those ligatured at the neck, but the regression and correlation coefficients are lower than normal (Table 4; Figs. 20, 21) probably because of secondary effects of the operations.

There is some evidence that outflux is increased in larvae ligatured between thorax and abdomen, as the internal sodium concentration may drop after uptake has ceased —especially in isolated abdomens (Figs. 12, 13)—and conceivably this could be an effect of hormone withdrawal. The very large and erratic increases in outflux found after nervous tissue has been frozen, however, are unexpected (Figs. 15, 18). These increases are not likely to be due to losses in the urine, as urine output is much curtailed in the operated larvae (Stobbart, 1970b). Freezing does not affect the permeability of the general body surface, so presumably the increases in outflux occur through the anal papillae. Perhaps the increases are caused by the release from damaged nervous tissue of various molecules which interfere with the transporting system, and it may be worth noting in this respect that small molecules can have a marked effect upon ion-transporting systems in insects (Maddrell, Pilcher & Gardiner, 1969).

Direct demonstration of hormonal control of sodium uptake in Aëdes, and the internal factor(s) to which sodium-deficient larvae respond, must await further study. However, neurosecretory control of potassium transport by Malpighian tubules is well established (Maddrell, 1963, 1964, 1966; Berridge, 1966), as are hormonal effects upon sodium transport in fish and amphibia, amongst the former of which interesting phyletic differences are coming to light (references cited earlier). It seems probable therefore that hormonal control of electrolyte content will prove to be of general occurrence, and in this respect the aquatic invertebrates would seem to offer a rich field for investigation.

I am indebted to Mr J. Upton of the Department of Mathematical Statistics of the University of Newcastle upon Tyne for a computer analysis of exponential regression and for statistical advice, to Professor J. Shaw for criticizing the manuscript of this paper, and to Messrs P. H. Cobbold and D. A. Wright for permission to quote unpublished results.

Bentley
,
P. J.
(
1969
).
Neurohypophyseal hormones in amphibia: a comparison of their actions and storage
.
Gen. comp. Endocr
.
13
,
39
44
.
Berridge
,
M. J.
(
1966
).
The physiology of excretion in the cotton stainer, Dysdercus fascia tut, Signoret. IV. Hormonal control of excretion
.
J. exp. Biol
.
44
,
553
66
.
Bodenstein
,
D.
(
1944
).
The corpora allata of mosquitoes
.
68th Rep. Conn, agrie. Exp. Stn. Conn. St. Ent. 44th Rep. Bull
.
488
,
396
405
.
Cazal
,
P.
(
1948
).
Les glandes endocrines rétro-cérébrales des insectes (étude morphologique)
.
Bull, biol. Fr. Belg
.
Suppl
.
32
.
Christophers
,
S. R.
(
1960
).
Aëdes aegypti (L.) the Yellow Fever Mosquito. Its Life History, Bionomics and Structure
.
739
pp.
Cambridge University Press
.
Clements
,
A. N.
(
1963
).
The Physiology of Mosquitoes
.
393
pp.
London
:
Pergamon Press
.
Copeland
,
E.
(
1964
).
A mitochondrial pump in the cells of the anal papillae of mosquito larvae
.
J. Cell. Biol
.
23
,
253
63
.
Crabbé
,
J.
(
1964
).
Stimulation by aldosterone of active sodium transport across the isolated ventral skin of amphibia
.
Endocrinology
75
,
809
11
.
Fanestil
,
D. D.
&
Edelman
,
I. S.
(
1966
).
On the mechanism of action of aldosterone on the sodium transport: effects of inhibitors of R.N.A. and protein synthesis
.
Fedn. Proc
,
23
,
912
16
.
Jones
,
J. C.
(
1954
).
The heart and associated tissues of Anopheles quadrimaculatus Say. (Diptera: Culicidae)
.
J. Morph
.
94
,
71
123
.
Koch
,
H. J.
(
1938
).
Absorption of chloride ions by the anal and papillae of diptera larvae
.
J. exp. Biol
.
15
,
152
60
.
Krogh
,
A.
&
Weis-Fooh
,
T.
(
1951
).
The respiratory exchange of the desert locust (Schistocerca gregaria) before during and after flight
.
J. exp. Biol
.
28
,
344
57
.
Maddrell
,
S. H. P.
(
1963
).
Excretion in the blood-sucking bug, Rhodnius prolixus Stàl. I. The control of diuresis
.
J. exp. Biol
.
40
,
247
56
.
Maddrell
,
S. H. P.
(
1964
).
Excretion in the blood-sucking bug, Rhodniusprolixus Stàl. III. The control of the release of diuretic hormone
.
J. exp. Biol
.
41
,
459
72
.
Maddrell
,
S. H. P.
(
1966
).
The site of the release of the diuretic hormone in Rhodnius—a new neuro-haemal system in insects
.
J. exp. Biol
.
45
,
499
508
.
Maddrell
,
S. H. P.
,
Pilcher
,
D. E. M.
&
Gardiner
,
B. O. C.
(
1969
).
Stimulatory effect of 5-hydroxy-tryptamine (serotonin) on secretion by Malpighian tubules of insects
.
Nature, Lond
.
222
,
784
5
.
Maetz
,
J.
,
Bourcuet
,
J.
&
Lahlouh
,
B.
(
1964
).
Urophyse et osmorégulation chez Caurassius auratus
.
Gen. comp. Endocr
.
4
,
401
14
.
Mills
,
R.
(
1967
).
Hormonal control of excretion in the American cockroach. I. Release of a diuretic hormone from the terminal abdominal ganglion
.
J. exp. Biol
.
46
,
35
41
.
Motáis
,
R.
&
Maetz
,
J.
(
1964
).
Action des hormones neurohypophysaires sur les échanges de sodium (mesurés à l’aide du radio-sodium Na”) chez un téléostéen euryhalin: Platichthys flesus (L
.).
Gen. comp. Endocr
.
4
,
310
24
.
Novák
,
V. J. A.
(
1966
).
Insect Hormones
.
478
pp.
London
:
Methuen and Co. Ltd
.
Pilcher
,
D. E. M.
(
1970
).
Hormonal control of the Malpighian tubules of the stick insect, Carausius morosus
.
J. Exp. Biol
.
53
,
633
65
.
Possompès
,
B.
(
1953
).
Recherches expérimentales sur le déterminisme de la métamorphose de Calliphora erythrocephala Meig
.
Archs. Zool. exp. gen
.
89
,
203
364
.
Ramsay
,
J. A.
(
1950
).
Osmotic regulation in mosquito larvae
.
J. exp. Biol
.
27
,
145
57
.
Ramsay
,
J. A.
(
1951
).
Osmotic regulation in mosquito larvae: the role of the malpighian tubules
.
J. exp. Biol
.
28
,
62
73
.
Ramsay
,
J. A.
(
1953
).
Exchanges of sodium and potassium in mosquito larvae
.
J. exp. Biol
.
30
,
78
89
.
Schreibman
,
M. P.
&
Kallman
,
K. D.
(
1969
).
The effect of hypophysectomy on fresh water survival in teleosts of the order Atheriniformes
.
Gen. comp. Endocr
.
13
,
27
38
.
Shaw
,
J.
(
1963
).
Kinetic aspects of ion regulation in aquatic animals
.
Viewpoints in Biology
,
2
,
163
201
.
Sohal
,
R. S.
&
Copeland
,
E.
(
1966
).
Ultrastructural variations in the anal papillae of Aides aegypti (L.) at different environmental salinities
.
J. Insect Physiol
.
12
,
429
39
.
Solomon
,
A. K.
(
1953
).
The permeability of the human erythrocyte to sodium and potassium
.
J. gen. Physiol
.
36
,
57
110
.
Stanley
,
J. G.
&
Fleming
,
W. R.
(
1966
).
The effect of hypophysectomy on sodium metabolism of the gill and kidney of Fundulus kansae
.
Biol. Bull
.
131
,
155
65
.
Stevens
,
W. L.
(
1951
).
Asymptotic regression
.
Biometrics
7
,
247
67
.
Stobbart
,
R. H.
(
1958
).
The exchange and regulation of sodium in the larva of Aides aegypti (L
.).
Ph.D. Thesis
,
University of Bristol
.
Stobbart
,
R. H.
(
1959
).
Studies on the exchange and regulation of sodium in the larva of Aides aegypti (L.). I. The steady-stage exchange
.
J. exp. Biol
.
36
,
641
53
.
Stobbart
,
R. H.
(
1960
).
Studies on the exchange and regulation of sodium in the larva of Aides aegypti (L.). II. The net uptake and the fluxes associated with it
.
J. exp. Biol
.
37
,
594
608
.
Stobbart
,
R. H.
(
1965
).
The effect of some anions and cations upon the fluxes and net uptake of sodium in the larva of Aides aegypti (L
.).
J. exp. Biol
.
42
,
39
43
.
Stobbart
,
R. H.
(
1967
).
The effect of some anions and cations upon the fluxes and net uptake of chloride in the larva of Aides aegypti (L.), and the nature of the uptake mechanisms for sodium and chloride
.
J. exp. Biol
.
47
,
35
57
.
Stobbart
,
R. H.
(
1971
).
Evidence for Na+/H+ and CL/HCO3, exchanges during independent sodium and chloride uptake by the larva of the mosquito. Aides aegypti (L
.).
J. exp. Biol
.
54
,
19
27
.
Stobbart
,
R. H.
(
1971
).
Factors affecting the control of body volume in the larvae of the mosquitoes Aides aegypti (L.) and Aides detritus Edw
.
J. exp. Biol
.
54
,
67
83
.
Treherne
,
J. E.
(
1954
).
The exchange of labelled sodium in the larva of Aides aegypti (L
.).
J. exp. Biol
.
31
,
386
401
.
Ussing
,
H. H.
(
1947
).
Interpretation of the exchange of radio-sodium in isolated muscle
.
Nature, Lond
.
160
,
263
.
Ussing
,
H. H.
(
1948
).
The use of tracers in the study of active ion transport across animal membranes
.
Cold Spring Harb. Symp. quant. Biol
.
13
,
193
200
.
Vietinghoff
,
U.
(
1966
).
Einfluss der Neurohormone C1 und D1 auf die Absorptionleistung der Rectal-drilsen der Stabheuschrecke (Carausius morosus Br
.)
Naturwissenschaften
53
,
162
3
.
Vietinghoff
,
U.
(
1967
).
Neurohormonal control of ‘renal function’ in Carausius morosus Br
.
Gen. comp. Endocr
.
9
,
503
.
Wall
,
B. J.
(
1966
).
Evidence for antidiuretic control of rectal water absorption in the cockroach Periplaneta americana L
.
J. Insect Physiol
.
13
,
565
78
.
Wigglesworth
,
V. B.
(
1933a
).
The effect of salts on the anal papillae of mosquito larvae
.
J. exp. Biol
.
10
,
1
14
.
Wigglesworth
,
V. B.
(
1933b
).
The function of the anal gills of mosquito larvae
.
J. exp. Biol
.
10
,
16
26
.
Wigglesworth
,
V. B.
(
1933c
).
The adaptation of mosquito larvae to salt water
.
J. exp. Biol
.
10
,
37
37
.
Wigglesworth
,
V. B.
(
1938
).
The regulation of osmotic pressure and chloride concentration in the haemolymph of mosquito larvae
.
J. exp. Biol
.
15
,
335
47
.
*

There is an error in this figure. For ‘Energy for reduction of carrier affinity for Cl− or H+’ read ‘Energy for reduction of carrier affinity for Cl or Na+’.