Urine Formation in A Pulmonate Land Snail, Achatina Fulica

The morphology of gastropod kidneys has been investigated in great detail and many inferences about the mechanism of action of the related parts have been drawn. A recent excellent review of morphology and physiology, intended to cover British prosobranch molluscs (Fretter & Graham, 1962), serves also as a review of most of the current literature on molluscan excretion. Harrison’s work (1962) on excretory processes in an archaeogastropod, the abalone (Haliotis rufescens), appeared too late for inclusion in the review. Many of the experiments on the abalone are comparable to those reported here. Considerable attention has been paid to the form in which nitrogen is excreted, Needham (1935) having demonstrated a uricotelic metabolism for those pulmonates included in his study.

We do not know of previous investigations on pulmonates in which urine and blood samples were taken systematically after the introduction of selected substances into the haemocoel. The results described below were obtained during two brief sessions at the Hawaii Marine Biological Laboratory. They cover only the well-hydrated state of the experimental animal and include no studies on the handling of normal excretory products. What is intended in this paper is an analysis of the sequential processes of urine formation in a pulmonate gastropod to determine the overall mechanism of urine formation.

Within the phylum Mollusca it is difficult to find species from which sufficient volumes of blood and urine may be obtained to permit easy, serial quantitative analyses. Achatina fulica, the giant African snail, proved to be such a species. Large individuals were hardy and had so thin a shell that the tissues could be exposed without damaging the animal. A wide range of sizes was available up to about 200 g. total weight. An effort was made to select the largest animals available but so many were used that the mean total weight did not much exceed 130 g. Of this weight the shell constituted about 37 % with a variability of from 17 to 41 %. The large animals were about 4 in. in shell length with seven turns in the spire. The animals were collected weekly in the Kaneohe valley of Oahu, Hawaii, where they have established themselves as a pest. In the laboratory the animals were kept on sod in aquaria supplied with vessels of fresh water and some form of fruit, usually avocado or papaya which were the only foods sought eagerly by the animals while in captivity.

Suitable anaesthesia could be obtained with ether. A wad of cotton in the bottom of a wide-mouthed jar was wetted with ether and the snail rested on the top of the bottle with the shell aperture down. As the snail became increasingly affected the musculature relaxed until the body of the snail drooped from the shell into the bottle. From 30 min. to 1 hr. was required for complete relaxation and the anaesthesia persisted for the 10 or 15 min. ordinarily required for the surgical preparations.

Two shallow hack-saw cuts each in. in length were made on the external surface of the shell parallel to the whorl and one full turn from the aperture. A screw-driver tip was inserted into one of the grooves thus made and the intervening shell was broken out with a twisting motion. The opening was enlarged with small bone forceps until all the necessary kidney tissue was exposed. The heart lies along the kidney but occupies a more central position. When it was necessary to make the pericardial sac accessible, the shell of the inner surface, representing a portion of the whorl another full turn toward the spire, was cut away. Space was thus provided in which the tissues could be manipulated. Unless these operations were accomplished without evidence of bleeding the snail was discarded. For the purpose of making injections and taking blood samples it was found satisfactory to introduce a long, multiply perforated plastic catheter into the haemocoel of the foot. This was done through an anterior stab wound which was closed, with the catheter in place, by a single ligature sewn through the body wall. Similar plastic catheters could be tied into the pericardial sac or anywhere along the ureter in the long length from the kidney to the pneumostome. The path of the duct can be well understood by consulting Bullough’s (1950) figure of the similar structure in Helix. We were not successful in inserting the catheters through the nephridial pore just inside the pneumostome. The kidney of Achatina is a thin structure with a compressed lumen much flattened in one dimension. The internal surface is greatly increased by the presence of folds running across the lumen. A plastic tube could be tied anywhere in the kidney lumen and this was done for some of the infusion experiments. The tissues exposed by operation were ordinarily covered with cotton wool moistened with Ringer solution. Adhesive tape prevented excessive bulging through the artificial opening in the shell when an animal retracted strongly. Upon recovery from anaesthesia a snail would ordinarily crawl about. In some experiments the animals were allowed to move about freely but in most cases the animal was taped to a board so that urine could be collected quantitatively.

Analyses were made for chloride, glucose, inulin, p-aminohippuric acid, phenol red and urea by colorimetric methods in which the optical density was finally read on a Beckman DU spectrophotometer for comparison with appropriate standards. Ordinarily identical volumes of normal blood were added to the blood standards before deproteinization. The blood protein of A. fulica is low but deproteinization was routinely performed by the Somogyi (1945) method. Aliquot parts of the resulting solutions were then used for the following analyses.

Glucose was measured either with Benedict’s (1931) quantitative reagent or by the anthrone method of Young & Raisz (1952).

Inulin was purified to remove alkali-labile compounds according to the method of Weil (1952), so that glucose and inulin could be determined simultaneously. An aliquot treated with anthrone reagent gave both inulin and glucose. Another aliquot, digested with sodium hydroxide to destroy the glucose, gave the inulin. Glucose was determined by difference.

p-Aminohippuric acid was determined by the method of Bratton & Marshall (1939). Urea was determined by the method of Archibald (1945).

Blood for the chloride determinations was deproteinized with phospho-tungstic acid. An aliquot of the supernatant solution was then analysed by Sendroy’s method (Hawk, Oser & Summerson, 1954).

Phenol red was measured spectrophotometrically without deproteinization but after addition of sea water to the samples. The concentration in blood samples was calculated by comparison with known dilutions of dye to which equivalent volumes of normal blood had been added. The concentrations in urine samples were measured by comparison with known dilutions of dye in sea water.

Equipment for the controlled infusion of test substances into the haemocoel was not available. Test substances were therefore injected at intervals calculated to produce a reasonably constant blood concentration. For filtered or reabsorbed substances quite satisfactory results were obtained by taking blood samples and injecting the test material at the mid-point of the urine collection period. In a few cases excellent results were obtained after a single priming dose. When the phenol red concentrations in the blood were intermediate or high the intermittent additions of phenol red did not alter the blood concentration markedly and urine/blood (U/B) ratios may be regarded as reliable. Phenol red is actively excreted at low blood concentrations at which time there is no stable reservoir of the dye. Therefore measurements made between intermittent injections are subject to error in determination of the actual mean blood concentration. It will be noted hereafter that some of the very high U/B ratios obtained are discounted on this basis.

Strohl (1913–14) has reviewed the older anatomical and observational evidence that urine formation in many molluscs begins with the production of a dilute fluid in the pericardium or pericardial glands of molluscs. The modem evidence may be said to have begun with Picken’s (1937) demonstration of massive filtration into the pericardium of Anodonta. More recently, Harrison (1962) has shown that filtration occurs into the pericardium of the abalone (Haliotis rufescens).

In the present investigations we were not able to demonstrate in any way that filtration into the pericardial sac is the initial process in urine formation. A reno-pericardial canal connects the pericardium with the lumen of the kidney, and since dye could be injected into the pericardium and collected from the kidney the canal is patent. A small amount of pericardial fluid was normally present in the sacs of those animals examined. When a catheter was tied in place with its tip in the sac the fluid entered the catheter and rose and fell with each heart beat; however, no significant quantity of fluid was ever collected. Efforts to ligate the duct or to exert pressure on the kidney over the reno-pericardial canal did not in any way decrease the rate of urine formation. It is recognized that this evidence against the formation of significant amounts of pericardial fluid is negative. It is convincing because the rate of formation of urine in Achatina is so high that the demonstration should have been an easy one.

Filtration is nevertheless the initial process in urine formation in Achatina. The argument rests upon the results of chemical analyses of substances injected into the blood stream of experimental animals. There is an insignificant number of blood cells in a sample of blood drawn during an experiment. The results of analysis may therefore be expressed directly as a U/B ratio. If no water reabsorption occurs in the process of urine formation a U/B ratio of one indicates excretion of a substance by simple filtration. A substance filtered but reabsorbed, such as glucose, gives a ratio less than one, and a substance actively secreted by the kidney cells gives a ratio greater than one. It would be confirmatory evidence for a process of filtration if such substances reach a U/B ratio of one in the presence of specific poisons which prevent active transport in either direction. Finally it should be possible to affect the filtration rate by simple back pressure.

The results of an experiment with inulin in which the variability was somewhat greater than usual are presented in Fig. 1. It is important to understand the sources of variability because the U/B ratios presented throughout this paper are calculated from the actual momentary analytical data, not from smoothed curves. Variability in mixing in the blood, lag in urinary sampling, and simple analytical errors may be exaggerated by the computation of ratios. Too much weight must not be placed on a single point on any of the following graphs ; it is the trends and the means which will allow proper interpretation.

It may be noted in Fig. 1 that the correspondence between the concentrations of inulin in blood and urine is on the whole quite good, and may be taken to indicate that there is an overall U/B ratio of one. 2,4-Dinitrophenol, which poisons most active transport, was given during the later parts of the experiment and made no significant difference to the excretion of inulin.

The results of twelve experiments with inulin are plotted in Fig. 2. Blood and urine analyses are here combined as U/B ratios and plotted against blood concentration. It may be seen that between blood concentrations of 0·4 and 1·2 mg./ml. the increase in blood concentration simply increased the amount of inulin filtered and hence the mean U/B ratios remained close to one.

At the lower end of the range a number of points are well above one. Such a result would be obtained if water were absorbed from the urine during its formation. The relationship might appear most clearly if the U/B ratios were plotted against urine volume per unit of time. In Fig. 3 such a plot may be examined. The curve shows clearly that high U/B ratios are indeed correlated with a low rate of urine formation. It is suggested that this is due to the filtration of inulin at blood concentration and its later concentration by water reabsorption as the urine flows slowly through the kidney and part of the ureter.

Computations of clearance, defined generally by C = UV/B, where U and B are the concentrations in the urine and in the blood respectively, have proven useful in vertebrate renal studies. It may already be seen that the U/B ratios obtained with inulin in Achatina are so close to one that the clearance will vary directly with the volume of urine. That this is the case is shown in Fig. 4 where clearance is plotted against urine volume. Clearance of filtered substances is known not to be much affected by blood concentration and this proved to be the case.

We have been interested to note whether any special mechanism exists in a pulmonate for the excretion of urea in this presumably uricotelic animal. When urea is injected into the blood it is excreted in the same way as inulin. Evidence for this view is presented in Fig. 5 where there are plotted the U/B ratios against the blood concentrations. The ratio remains close to 1 over a considerable range of blood concentration.

When the U/B ratios are plotted against urine volume per unit of time (Fig. 6) it may be noted that no very high ratios are obtained at the low rates of urine flow as was true with inulin. It is probable that this results from the lower molecular size of urea. When water is reabsorbed undoubtedly urea also diffuses in the same direction and at a considerable rate. Although at low rates of urine flow there is a preponderance of U/B ratios higher than one, the effect is not a striking one. Clearance is again proportional to urine volume as in the case of inulin.

Experiments with both urea and inulin have now been seen to involve rates of urine flow as high as 10·2 ml./hr. In terms of an individual animal the data plotted in Fig. 1 may be totalled over the experimental time period for illustration. In the course of the 6 hr. experiment injections totalling 33·3 ml. were made in order to maintain a reasonably constant concentration of inulin in the blood. During the same time urine to a total volume of 27·6 ml. was produced. Since the estimated blood volume of this 112·5 g. snail was only about 25 ml. (Martin, Harrison, Huston & Stewart, 1958) a volume equal to the blood volume was filtered into the kidney.

In contrast, consistently low rates of urine flow were often encountered, although in about half such instances regular injections into the haemocoel were made. Flow rates of the order of 0·6−1·2 ml./hr. are illustrated in Figs. 6, 9 and 16. High rates of urine flow were not continued without adequate fluid injection but moderate rates could be maintained for considerable periods. In the experiment reported in Table 3, for example, no injections were made for 135 min. The volume of urine formed remained relatively close to a mean of 2·88 ml./hr.

An attempt was made to determine the rate of urine production in unoperated animals. Four specimens were placed in covered finger-bowls, given food and water ad lib., and the urine was collected at intervals. After 24 hr. the urine production averaged 0·44 ml./100 g./hr. Another collection was made 48 hr. later, at which time urine production averaged 0·31 ml./100 g./hr. The experiment was unsatisfactory in one sense. The animals spent most of the time in the finger-bowls in an inactive state. In this condition the formation of less urine would be expected. The snails live by preference in very humid environments, hence while the highest rates of urine formation obtained in this work would rarely be encountered, undoubtedly the lower rates would quite commonly occur in nature.

Most of the experiments reported here were carried out with an unrestrained flow of urine. In a few experiments the urinary catheter was provided with a bent metal tip which could be hooked over the collecting tube at any given distance above the snail kidney. Back pressure of a known amount was thus imposed on the urinary system. A typical experiment is graphed in Fig. 7 in which it may be seen that the volume of urine drops off considerably as the back pressure is raised. Results of three such experiments are summarized in Fig. 8, where the volume of urine per unit of time and the inulin clearance are both plotted against the ureteral back pressure. This plot reveals a clear relationship between back pressure and volume of urine formed, and defines two important limits. The first of these is the back pressure at which urine formation is stopped. This amounts to 12 cm. of water. The other limit is the maximum rate of urine formation, which comes close to 12 ml./hr. or 288 ml./day.

Back pressure might have two different effects. Filtration might continue at a relatively high rate, to be followed by much more water reabsorption when the back pressure-was high. Or the back pressure might be transmitted all the way to the primary filtration site and greatly reduce the filtration rate. A choice between these alternatives is possible on the basis of the inulin clearance. If filtration had continued at a high rate in spite of the back pressure and been followed by much water reabsorption to give a low urine volume one would expect a high clearance of inulin. This was not the case; clearance depended upon urine volume. The U/B ratios did rise significantly above one at low flow rates but were in excess of two only in a few cases. At higher back pressures, in a few cases, about twice as much filtrate was formed as appeared as urine, but usually the filtration mechanism was very sensitive to back pressure.

Ordinarily when a process of filtration initiates urine formation there is conservation of a useful part of the filtration fraction. The analysis of this process in Achatina has been only begun in this work. The reabsorption of two substances has been studied, that of glucose in some detail, that of chloride in a preliminary manner.

The concentration of glucose in the blood of Achatina is extremely variable. Values for thirty-one snails from 0·6 to 244 mg. % (mean 55·7 mg.%) were obtained in the course of this work. Anaesthesia apparently made no difference to the blood glucose. Evidence that glucose is filtered will be presented below. That glucose can be absorbed from the kidney was readily demonstrated. In a series of experiments, illustrated in Table 1, glucose solution was infused into the kidney lumen. Urine, including the infusion fluid and any filtrate formed, was collected from a catheter in the ureter. The blood glucose of snail number 1 was quite low. This snail reabsorbed a large part of the exogenous and the filtered glucose. Snail 2 had intermediate blood glucose and absorbed a substantial amount of the glucose. A third snail with high blood glucose absorbed considerably less of the glucose. The glucose gradient in all three cases was from the kidney lumen into the blood. That active absorption of glucose was involved in each case, however, is shown by the results of phlorizin treatment. A dramatic reduction in the amount of glucose reabsorbed occurred in each of the three experiments.

Many experiments were carried out in which blood glucose and urine glucose were measured simultaneously. In Fig. 9A the U/B ratios for glucose are plotted against blood glucose concentration. It may be noted that the ratio increases toward a maximum of one with increasing blood concentration as would be expected. Two factors are involved in presenting the kidney with more glucose than it can reabsorb, first the blood concentration and second the rate of filtration. In Fig. 9B the U/B ratios are plotted against rate of urine flow and again there is a trend toward a ratio of one at high rates. To make it possible to evaluate the relative importance of the two factors the data are replotted in Figs. 9C and D. In C are plotted the U/B ratios for all those experiments in which the blood glucose concentrations are below 1 mg./ml. Now it may be seen that the U/B ratios did not exceed 0·5 even at the higher rates of flow observed. In D are plotted the ratios for all experiments in which the blood glucose was above 1 mg./ml. Here the ratio quickly climbs to and above a value of 0·75 and approaches 1 in a few cases. A high blood glucose is therefore of primary importance in causing a loss of glucose from the body.

The effect of phlorizin in arresting glucose reabsorption has already been demonstrated in the data of Table 1. The effect is worth illustrating in a more specific manner. Phlorizin had a prompt action, increasing the clearance of glucose to that of inulin as is shown in Fig. 10. Preliminary experiments showed that doses of at least 1 mg./100g. tissue were needed to arrest glucose transport and that recovery from a single such dose might occur quickly.

From the results of the various experiments with glucose it seemed that it might be possible to determine the maximum capacity of the kidney to transport glucose. In Fig. 11 are plotted the results of ten experiments with glucose in which it was possible to compute the number of milligrams of glucose reabsorbed per unit of time. These rates were plotted against blood glucose concentration in order to determine whether an upper limit was reached. Low and high rates of urine formation are represented and are well distributed over the range plotted. The rate of urine formation does not seem to affect the reabsorption rate for glucose. The maximum rates of between 3 and 4·2 mg./hr./ioo g. of animal weight occurred at blood concentrations above 140 mg. %.

Terrestrial animals often suffer from salt deficiencies. It was not possible to analyse blood and urine for cations and consequently only a few analyses of chloride were made as a possible indicator of salt conservation. In all cases the urine chloride was lower than the blood chloride. Results for a series of normal animals are shown in Table 2.

The rate of urine formation did not appear to make very much difference in the reabsorption of chloride. In the experiment reported in Table 3 urine chlorides were followed at about half-hour intervals to establish a base period. The volume of urine formed was then reduced by injection of benemid, which has no known effect on chloride transport, and the chlorides were followed for another 3 hr. Although the urine flow was reduced to about one-fifth of its original volume the urine chloride level remained the same.

The injection of extra-strength saline into the snail appears to require some caution in interpretation. The Helix Ringer described by Bernard & Bonnet (1930) did not interfere with the continued action of the heart when injected into the pericardium and was therefore used in some of the experiments described here. When an effort was made to increase the chloride ion content of the urine by injections of triple-strength Helix Ringer the animal died about an hour after the first injection. The results of the experiment are shown in Table 4 to illustrate the fact that the urine chloride rose only slowly after the blood chloride had been forced to a considerable increase over normal. Even then the early death of the animal leaves some question about the reason for the increased urine chloride.

The presence of secretory mechanisms in pulmonate kidneys has been known for some time (Strohl, 1913-14) as has the accumulation and excretion of indigo sulphonate, the progenitor of phenol red. Rowntree & Geraghty (1910) selected phenol red from a considerable number of dyes because it has far less propensity than indigo sulphonate for accumulating (staining) in the kidney cells and is thus both far less toxic and more rapidly excreted. Even so it will be remarked below that not all of the injected phenol red could be accounted for in the course of these experiments.

The results of ten experiments in which phenol red was injected into the blood are summarized in Fig. 12. It may be noted first that the U/B ratio was always greater than one. This fact is not due to water reabsorption. Six experiments were performed in which phenol red and inulin clearances were determined simultaneously. The ratio of their clearances never fell below 1·5 and commonly reached a value of 7·5. Phenol red must therefore be secreted by the snail kidney. As the blood concentrations fell the U/B ratios sometimes reached levels which we consider unreliable. Other very high values have been reported in molluscs but high ratios at low blood concentrations may result from sampling difficulties as discussed previously under methods. The curve at the two extremes is therefore represented by broken lines. At the high blood concentrations so much dye is presented to the epithelial cells that the transfer maximum is exceeded. The U/B ratio falls even though the total amount transported may remain high. The last portion of the curve probably represents the phenomenon called by Smith (1951) self-depression. The kidney cells are actually inhibited by the high concentrations of dye, although the inhibition is reversible.

In Fig. 13 the clearance of phenol red has been plotted against the blood concentration. The interpretation of such a curve, if protein binding of the dye is not a significant factor, is that the highest reliable clearances give an estimate of the total blood flow through the kidney. The solid line represents a region of the curve in which we feel some confidence. The upper end indicates a clearance of about 16 ml. of blood in 10 min. or a flow rate of 1·6 ml./min.

An alternative method of calculating the blood flow through the kidney offers a comparison with the previous estimate and has the additional advantage of providing an evaluation of the filtration fraction. Calculation of the millimoles of phenol red secreted per unit of time indicated that secretion rates were maximal at blood concentrations of 0·08 mM/1. (3 mg. %) (cf. Fig. 16A). This is somewhat higher than the critical concentration of about 0·042 mM/1. (1·5 mg. %) in man. Up to this critical point 3·58 ml. of blood must be flowing through the kidney every 10 min. to provide the quantity of phenol red secreted. In addition to this quantity of blood cleared by secretion there is a volume necessary to provide the phenol red excreted by filtration. The filtration rates are quite variable, the volume of blood flow and the filtration fraction may therefore also be expected to vary over a considerable range. Filtration rates (taken as urine flow rates) were tabulated for each of the points below the critical concentration level on the curve of Fig. 16 A. The minimum value was 0·3 ml./10 min. and the maximum was 1·5 ml./10 min. According to this estimate, therefore, the rate of blood flow varies from 3·88 ml./10 min. (3·58 + 0·30) to 5·08 ml./10 min. (3·58 + 1·50).

The above data yield filtration fractions of 7·75% (0·30/3·88) and 29·5% (1·50/5·00). These calculations of filtration fraction may be compared with values obtained by taking the reciprocals of the phenol red/inulin clearance ratios which are plotted in Fig. 14. When the ratios were grouped by similarity of blood flow and filtration rate a minimum average filtration fraction of 11·7 % at a blood flow of 0·41 ml./min. was obtained. The maximum average filtration fraction of 38% occurred in one animal with a blood flow through the kidney of 0·37 ml./min. and the very high filtration rate of 0·139 ml/min.

Secretory processes in the vertebrate kidney are poisoned by dinitrophenol. In Fig. 15 the result of a typical experiment is plotted in which phenol red transport has been poisoned in this way. The U/B ratio had reached its maximum when the first injection of dinitrophenol solution was made into the blood. Within 20 min. a dramatic inhibition of phenol red secretion had occurred. Because the kidney was no longer putting out a large quantity of phenol red the continued injections raised the concentration of phenol red in the blood. The falling urine concentration and the rising blood concentration quickly met and became essentially identical, thus providing additional evidence for the process of filtration.

There is much evidence that the vertebrate transport mechanism which secretes phenol red also handles p-aminohippuric acid (PAH). The excretion of PAH and the probability that the mechanism is the same will be examined in this and the next section. U/B ratios for PAH did not reach such high levels as those for phenol red, and at low blood concentrations there was not the tremendous affinity for PAH that there was for phenol red. The data collected appear to be best represented, as they are in Fig. 16, by plotting the clearance of PAH against the rate of urine flow. It may be noted that the clearances are an order of magnitude greater than those of inulin (cf. Fig. 4). Simultaneous clearances for inulin and PAH confirm this difference and indicate clearly that PAH is actively secreted by the snail kidney. The clearance is moderately increased by an increase in urine volume per unit of time. The rate of urine flow in the experiments shown in Fig. 16 reached the maximum of which this snail is capable and the curve appeared to have levelled off at a clearance of about 3·0 ml./10 min., but the evidence is not decisive. This is about the same maximum obtained as when clearance is plotted against blood concentration. Clearance of PAH as a measure of blood flow through the kidney gives a distinctly lower value than that of phenol red.

The secretion of PAH is inhibited by dinitrophenol. The evidence is presented in Fig. 15 where dinitrophenol was administered after the establishment of a substantial difference in concentration of PAH between blood and urine. It is of considerable interest to note that the administration of phenol red at a rather low concentration increased the rate of excretion of PAH, even though there was no signify-increase in the rate of urine production.

If the transport mechanism for phenol red is also that for PAH the total amount of each transported per unit of time should be of the same magnitude. In Fig. 17 the rates of excretion of both phenol red and PAH are plotted against the respective blood concentrations. Values are given in micromoles and millimoles per litre to permit direct comparison. It may be noted that the rates attained are certainly of the same order of magnitude, although that for PAH rises a little higher than the maximum rate for phenol red. The shape of the phenol red curve reminds us, however, of the inhibition which was noted in the transport of this dye and suggests that the inhibitory effects of high phenol red concentrations may be responsible for the difference in rate.

Another form of evidence that the transport mechanism is the same comes from experiments on competition. We were not successful in showing that the administration of PAH depresses phenol red transport, perhaps because of the greater affinity of the secretory cells for phenol red already described. The results presented in Fig. 18 demonstrate quite clearly that the transport of PAH is depressed by the simultaneous excretion of phenol red. The maximal rate of transport of PAH reached only half of that shown in Fig. 17.

When there is competition for the transport sites in the cells the secretion of one compound or the other may be depressed but the total amount transported should remain at about the same level. That this is the case is illustrated in Fig. 19 where the total amounts of phenol red and PAH transported simultaneously are plotted for comparison with the transport rates of phenol red and PAH shown in Figs. 17 A and B.

Molluscs have long been characterized as possessing ‘kidneys of accumulation’ because concrements of various kinds have been described in the excretory system, some of which appear to remain for long periods of time. The phenomenon is particularly striking after a period of desiccation. Phenol red, PAH and inulin are not thought to be held in vertebrate kidney cells except perhaps under abnormal ionic conditions. In a study of molluscan kidneys it is important to evaluate the retention of the test substances, and to try to differentiate accumulation in kidney cells from accumulation in other tissues of the animal.

Experiments were taken at random from the group in which triple clearances had been recorded, for computation on the injected, excreted and residual amounts of phenol red, PAH and inulin. The results of the computations are set forth in Table 5. In snail number 1 small amounts of each substance were injected and it was possible at the end of the experimental period to account for a substantial portion of each substance. If excretion had been followed without further injection of test substances the snail would have eliminated most of them within a few hours. In snails 2 and 3 larger amounts were injected and the fraction of each substance which could be accounted for was lower. The large and relatively inert molecule of inulin was best recovered in every case, as would be expected. A surprisingly large amount of PAH disappeared from the blood without appearing in the urine. Phenol red does not stain tissues very effectively, but at high concentrations it is not surprising that it should, affect many cells of the body. We could detect a reddish colour due to phenol red in the mucous tracks left by treated snails which crawled about in the laboratory. The small fraction of the injected phenol red accounted for in snail 3 after more than 20 hr. indicates the importance, transiently at least, of the mechanism of cellular accumulation.

It is difficult at this time to evaluate the role of the kidney cells in the accumulation process. The type of evidence cited by vertebrate renal physiologists (cf. Smith, 1951) to show that substances are not accumulated in the renal cells includes the correspondence of clearances obtained on rising blood concentrations with those obtained on the same, but falling, blood concentrations. This was not true in Achatina, where higher clearances were obtained on falling blood concentrations than on rising blood concentrations. It might be possible to determine the extent of phenol red accumulation from this difference in clearance. Results of calculations for two animals indicated a storage of 1·06 and 1·32 mg. These amounts are very similar despite wide differences in amounts injected, 117 mg. during 560 min. and 39 mg. during 260 min. respectively. However, blood concentrations were high enough in both to indicate that the secretory cells were working at maximum capacity, in fact the analyses showed that in one animal while 1·06 mg. was stored in the kidney there were about 20 mg. continuously present in the blood, and the kidney had excreted 14 mg. of phenol red. The other snail stored 1·32 mg. while holding about 5 mg. in the blood and excreting 5 mg. Since kidney tissue makes up only a fraction of the total tissue weight of the animal the participation of athrocytes, and indeed of many other tissues, appears to be very likely.

The process of filtration in gastropods has received little experimental attention. The evidence presented by Harrison (1962) from experiments with the abalone Haliotis rufescens clearly involved the pericardium. The compounds appearing in samples of pericardial fluid, including inulin, were at the same concentration as in blood. Harrison was not able to collect a sufficient quantity of pericardial fluid to account for all of the urine formed but this seems most likely to be due to the technique employed. Filtrate was allowed to leave the pericardium not only through a small plastic catheter, but through the right and left reno-pericardial canals. Unless the normal routes of escape had been tied off there seems to be no reason why most of the filtrate should flow through the catheter. There is therefore no necessity to postulate any other source of filtrate in this marine gastropod.

In contrast, no evidence could be found that the structures within the pericardium of Achatina fulica are the site of filtration. Identification of the site has not been completed but it seems possible that the direct arterial blood supply to the kidney provides the filtration pressure and that a specialized area of the kidney may be involved. Reports (Cúenot, 1914) in the literature indicate a variation in histological appearance in different parts of gastropod kidneys, and we noted the external colour differences in two parts of the kidney mentioned for other gastropods by Fretter & Graham (1962). An interpretation of this change in location of the site of filtration is premature, but the volume of filtrate might be better regulated if the site of filtration were moved away from the heart. Back pressure, for example, will greatly reduce filtration, as has been demonstrated in this paper, but back pressure into the pericardium is not a useful device because of the risk of pericardial restraint and a consequent failure of the circulation. Vigorous circulation of blood would be compatible with a low filtration rate in Achatina but not in Haliotis.

When the rate of filtration in Achatina, corrected to a unit of body weight as it is in Table 6, is compared with that of a number of other animals including the two molluscs Haliotis rufescens and Octopus dofleini, it proves to be surprisingly high. It should be recalled that the filtration rate cited for the snail is the maximum one, and that it may be reduced to zero when the snail is inactive. The relatively large volume of filtrate might be due to a very large cardiac output or to a high filtration fraction. The second seems the more probable until further measurements can be made. Picken (1937) demonstrated a rate of filtration in the fresh water Anodonta cygnea as high as 3·26 ml./kg./min. Thus, as would be expected of a land form, the rate for Achatina lies between those for the marine and freshwater molluscs. Values for a filtration fraction have been obtained for Achatina and range all the way from 7·7 to 38%. These values too may be compared with those obtained for other animals as is done in column IV of Table 6. Although there is shown to be a considerable degree of overlap with the vertebrate values the snail gives the highest fraction of any of the animals included, while the marine animals, dogfish, Haliotis and Octopus had the lowest. It may be assumed that the tissues at the site of filtration are remarkably permeable in Achatina. Filtration stopped at a back pressure of 12 cm. of water leaving the impression that the blood pressure must reach at least this minimal value in the healthy snail; however, it is not to be expected that blood pressures would greatly exceed this level. The blood protein is low and the filtration pressure meets little opposition, especially when as in these experiments the urine is drained away as rapidly as it is formed.

The similarities in blood glucose concentration between Achatina and vertebrates, and the observations on filtration rate, suggest that the load thrown on the kidneys by glucose reabsorption might be similar in these animals. In columns V and VI of Table 6 there are shown comparisons of the transfer maxima for glucose, and of the blood concentration required to produce maximal transfer. The snail now falls somewhat short of the very high rates achieved by mammals but compares favourably with a frog. The kidney cells are saturated at a lower concentration so it is not surprising that glucose was usually present in snail urine under the conditions of these experiments.

Finally, in comparing Achatina with other animals, the rates of excretion of phenol red are of some interest. The comparison on a unit weight basis produces a wide diversity of values. In Column VII of Table 6 it may be noted that even on this basis Achatina occupies an intermediate position, its kidney being able to transfer a very considerable amount of phenol red, and much more than either of the other molluscs. Following Smith’s lead in correcting to some extent for the differences in filtration rate, column VIII of Table 6 may offer a somewhat better comparison among the different animals represented. On this basis the snail shows the lowest figure for phenol red transfer. However, the choice of a very high filtration rate instead of a modest one results in this lower rate. The value obtained for Achatina is still of the same order of magnitude as is observed with all the animals except the bird.

In spite of the tremendous morphological differences which exist the processes involve in urine formation in these gastropod molluscs are shared with vertebrate animals. The correspondence is at several levels, from cellular mechanisms of transfer, and susceptibility in common to specific poisons, to details of circulatory physiology.

We wish to acknowledge support of this work through Grant N-onr-52011 of the Office of Naval Research and Grant HE 02557 of the National Heart Institute, U.S. Public Health Service. This is contribution no. 210 from the Hawaii Marine Biological Laboratory.