The Fate of Thorium Dioxide Injected Into the Pedal Sinus of Bullia (Gastropoda: Prosobranchiata)

The fate of foreign particulate matter in the haemolymph of gastropods is a subject which has not received a great deal of attention, despite its importance to many different aspects of gastropod physiology, haematology and parasitology. Indeed, apart from a short paper by Tripp (1961) on the snail Australorbis, the most closely related work of any consequence appears to be that performed on lamellibranch molluscs, notably species of oyster (Takatsuki, 1934; Ranson, 1936; Stauber, 1950; Tripp, 1958). Cuénot (1914) carried out experiments on a variety of molluscs but was concerned only with the immediate removal of foreign particles from the circulation, paying no attention to their ultimate disposal. The foreign particles injected into the blood-stream by these and other workers have most commonly been indian ink or suspensions of carmine. Such substances have disadvantages in that, first, the aggregations which these particles form in sea water or invertebrate blood are almost certainly much larger than the animals are ever called on to dispose of under natural conditions and, secondly, the animal must be killed and sectioned before even the gross distribution of the injected particles can be assessed. A radio-opaque dye in the form of fine, colloidally suspended particles would overcome both these limitations providing it was non-toxic.

A dye which meets these requirements is Thorotrast, a colloidal suspension of thorium dioxide which is very opaque to X-rays and has consequently been used extensively in clinical arterio-radiography and in studies of the blood-vascular systems of various vertebrate animals (Veal & McFetridge, 1944; Barclay, Franklin & Prichard, 1944; Barclay, 1951; Foxon, Griffith & Price, 1956; Foxon, 1961a,b). It has also been used in studying blood movements in the sandy-beach snail, Bullia (Brown, 1964b). However, its original use was in clinical hepatolienography (Radt, 1929,1930; Volicer, 1931), for the thorium particles are taken up from the vertebrate blood stream by the macrophage and reticulo-endothelial system, giving intense shadows in the liver and spleen some days after injection. Thorium dioxide is chemically unreactive and is not excreted to any appreciable extent by the vertebrate body; indeed it is this latter fact that is largely responsible for discontinuing the clinical use of Thorotrast, for it accumulates in the tissues, where its radio-activity causes serious diseases (Schmidt & Herzog, 1950; Prezyna, Ayres & Mulry, 1953). The properties of Thorotrast have, nevertheless, made it an invaluable tool in research on the reticulo-endothelial system in vertebrates other than man (Irwin, 1932; Tripoli, 1934; Foxon & Rowson, 1956; Foxon, 1961 a). Such studies do not appear to have been carried out on invertebrate animals.

Two species of the sandy-beach snail, Bullia, were used in our experiments. They were B. laevissima (Gmelin) and B. digitalis Meuschen, both species being collected from Hout Bay, on the coast of the Cape Peninsula, South Africa. The animals were kept in the laboratory in a large tank of aerated sea water with a substratum of beach sand, at 15° C (± 0·5) but were removed to plastic buckets before injecting, X-raying, etc., and sometimes for feeding. Snails were fed at weekly intervals throughout the series of experiments. While the temperature at which they were kept is consistent with the temperatures the animals experience when submerged in their natural environment, the amount of food they received was quite arbitrary. Their opportunities to feed in their natural habitat are highly irregular and often vary enormously from season to season (Brown, 1961, 1964a). In the laboratory they will live for at least 6 weeks without food. These facts are mentioned as it appears likely that the rate of phagocytic uptake of thorium may be influenced both by temperature and by nutritional state. Gordon & Katsh (1952) have shown that the activity of cells of the endothelial system in rats is closely correlated with the nutritional state of the animals.

After 7 days in the laboratory snails were removed from the water and Thorotrast was injected into the cephalopedal sinus under conditions which were made as sterile as possible. They were then X-rayed at intervals, eventually being killed and removed from their shells at appropriate stages. They were X-rayed again after removal from their shells as this allowed the exposure to be greatly reduced, so increasing the degree of contrast in the resulting radiographs. Some of the animals were dissected and the various tissues and organs were X-rayed separately, while others were cut into pieces for embedding in wax and sectioning. Sections were viewed both by dark-ground illumination and by the method of combined dark-ground, phase-contrast and oil immersion suggested by Baxter (1960). Micrographs were prepared from some of the sections and enlarged for more detailed study. Blood was extracted from living snails from time to time in order to study the possible uptake of thorium by the haemocytes.

In a preliminary series of experiments the amount of Thorotrast injected was 0·25 c.c. in the case of Bullia laevissima and 0·1 c.c. in the case of the smaller B. digitalis. This is proportionately much more than was used by Foxon and his co-workers in their-studies of Amphibia and is greatly in excess of the dose of 0·8 c.c. per kilogram of body weight suggested for man and other mammals by Tripoli (1934). Nevertheless, it proved insufficient and in subsequent experiments the dose was increased to 0·5 c.c. for B. laevissima and 0·25 c.c. for B. digitalis. These large doses were necessary for two reasons. First, the blood volumes of these animals are very much greater than in vertebrates (Brown, 1964b) and the blood system is an open one; consequently the Thorotrast becomes greatly diluted shortly after injection so that its effective radioopacity is much diminished. Secondly, it was necessary to use an exposure of X-rays which would penetrate the shell of the animal as well as its living tissues, the optimum exposure being 0·03 sec. at 200 mA. and 66 kV., at a focus-film distance of approximately 100 cm. At these values low concentrations of the dye cast no visible shadow on the X-ray film.

Following investigations on the uptake and disposal of thorium dioxide particles, a brief study was made of the effect of such foreign matter on the haemocyte population. We were able to revert to lower doses of the dye for this study. Cell counts were made using the haemocytometer techniques usually employed for vertebrate animals, but with the modifications introduced by Yeager & Tauber (1935) for marine invertebrates. The diluting fluid consisted of 20 c.c. boiled and filtered sea water with the addition of three drops of 10% acetic acid and 3 c.c. of 0·01% gentian violet. The dilution pipette was calibrated so as to dilute 1 mm.³ of haemolymph 100 times. In addition to total cell counts mitotic indices were calculated, again following the methods of Yeager and Tauber (1933, 1935). The mitotic index is found by counting 2000 cells, taken at random, and noting the number of mitotic figures; the result is then expressed as the number of mitotic figures per thousand haemocytes.

Such counts were made immediately after capture of the snails, after 1 week in the laboratory, and each day after the injection of Thorotrast. Ten marked individuals of Bullia laevissima were used, and five control snails of the same species were kept in the same tank. Saline made up in triple-distilled water was injected into these control snails in place of Thorotrast. For this series of tests all haemolymph samples were drawn from the buccal sinus system, though isolated samples were taken from other (untreated) snails from different sites including the heart, the pedal sinus, the anterior aorta and the efferent branchial vein.

Radiographs show that within 5−10 min. after injection the Thorotrast has been distributed throughout the body of the animal. There is no occlusion of the vessels such as is found after injection of indian ink into lamellibranchs (Stauber, 1950). Comments on the passage of haemolymph from the pedal sinus to other parts of the circulatory system of Bullia have been made elsewhere (Brown, 1964b). No change can be seen in radiographs taken within 3−4 days of injection, but after that period it is apparent that the distribution of thorium is no longer uniform. Both the posterior part of the mantle and the kidney tissues usually cast a more distinct shadow than other parts of the body and often the gill can be made out, but the most striking change is that the outline of the heart, both auricle and ventricle, can be clearly seen (Pl. 1, fig. 1). Sometimes short lengths of the vessels entering and leaving the heart are also outlined by the dye. These shadows become more intense as time proceeds and after 6 or 7 days it can be seen that the pericardial cavity has also become heavily involved (Pl. 1, fig. 2). While the shadows cast by the heart and pericardial cavity darken, the gill shadow normally fades while the somewhat diffuse shadow cast by the kidney intensifies, its lumen soon becoming as deeply involved as the pericardial cavity (Pl. 1, fig. 3). In a few cases this has been noted sooner than a week after injection of the dye but in others nearly a fortnight elapsed before the lumen of the kidney cast a shadow of comparable intensity. From this stage on the Thorotrast shadows gradually fade, the pericardial shadow being the first to disappear, after which the outline of the heart can again be clearly seen. By this time the shadow cast by the lumen of the kidney has also grown faint, though the diffuse shadow cast by its tissues remains. Fading continues until no shadows appear on the radiographs, the whole process from injection of the dye taking from 4 to 6 weeks under the conditions described.

There appears to be a considerable variation in the uptake of thorium particles by the tissues in different snails. The posterior part of the mantle can be seen in some snails but not in others; in some radiographs the gill can be seen as a faint, though distinct, shadow, while in others, taken at the same time after injection, none of the pallial organs can be made out—this in spite of the uniformity of exposures, focus-film distance and subsequent processing. Care was also taken to maintain all the snails under identical conditions. We can only conclude that the differences seen in the radiographs represent a real variability in the uptake of thorium in the animals themselves. Foxon & Rowson (1956) have found similar differences in the frog, Rana temporaria, and state that ‘there is therefore some degree of variation in what may take place when the Thorotrast is removed from the blood. Such variation may be correlated with variation in the activity of the cells of the reticulo-endothelial system.’ We have not been able to relate the variability in Bullia to the sex or the size of individuals. In spite of these differences the most intense shadows always occur in the wall of the heart, the pericardial cavity and the lumen of the kidney, in that time sequence. The period taken for these to develop and regress varies, however, from snail to snail.

In some radiographs, particularly those taken after full development of the kidney shadow, the floor of the pallium could be seen to cast a shadow, especially posteriorly where it is joined by the mantle. Washing out the mantle cavity with sea water introduced through an arterial catheter pushed through the siphon reduced this shadow considerably and in some cases virtually eliminated it. The shadow is largely caused, then, by thorium which has left the body of the snail and accumulated on the floor of the pallium.

Sections cut through various regions of snails killed 4, 8 and 21 days after injection of Thorotrast show the changing distribution of thorium dioxide particles in the tissues. Viewed by dark-ground illumination or by the method of Baxter (i960) aggregations of thorium particles in Bullia have the same appearance as those studied in other (vertebrate) animals (see Foxon, 1961a). Any possible chance that we were viewing particles of some other substance was eliminated by making a 3-week autoradiograph of one of the sections ; a-tracks are apparent on the processed microfilm, proving the radio-activity of the particles and showing that they are α-emitters.

In all the sections examined thorium particles were seen only inside amoebocytes, no other type of cell having taken them up ; nor could any free particles of thorium be found, even in the snail killed 4 days after injection of the dye. The thorium-containing amoebocytes, however, were found in most of the tissues in one or other of the sections, as well as in the vessels and sinuses. The exceptions were the tissues of the digestive gland and the reproductive organs.

In the 4-day sections virtually all the particle-laden cells were clearly haemocytes which had not yet left the circulatory system, though we gained the impression that most of them were attached to the walls of the vessels or to the sides of the sinuses. Many of those apparently attached, however, may have been free in the living animal. The greatest number of haemocytes was to be found in the heart, clustered against the muscles and filling the small spaces between them ; nearly all these haemocytes contained particles of thorium. The gill vessels were also found to be crowded with similar particle-laden cells, as were the blood spaces of the kidney. No migration of these cells through the epithelia or other tissues was, however, encountered.

In sections of the 8-day snail the position was rather different and invasion of the tissues was in evidence. The kidney tissue had been invaded to a considerable extent and some amoebocytes could be seen in the mantle epithelium. However, the most marked invasion was of the heart itself and a very large number of particle-laden cells could be seen between its muscles, in the epithelium and in the pericardial cavity. A few cells were noted in and attached to the pericardial wall—a negligible number compared with those migrating through the heart. No amoeboid cells could be found in the reproductive organs or the digestive gland, though some were present in the associated blood sinuses. Thorium-laden haemocytes were still present in the gill, but no migration of these through the gill epithelium was in evidence.

The 21-day animal gave similar, though not identical, results as far as most of the tissues were concerned, and migrations of haemocytes through the heart wall and mantle epithelia were still apparent. However, very few cells containing thorium particles were seen in the circulatory system; they had disappeared from the gill vessels and from the anterior sinuses and only the arteries and the heart itself still had a few laden cells attached to them. On the other hand very large numbers of haemocytes which did not contain any Thorotrast were seen throughout the vascular system. Both the pericardial cavity and the lumen of the kidney appeared to be nearly choked with a mass containing so much thorium that the entire area showed up a brilliant, opaque white under dark-ground illumination. Under these circumstances it was not possible to be certain that the thorium particles were really confined to amoebocytes.

A few sections cut through an individual of Bullia laevissima killed 6 weeks after injection of the dye showed that even at this stage thorium particles were still present in some of the tissues. The circulatory system, pericardial cavity and the lumen of the kidney were, however, quite free of them; nor could any be detected in the mantle epithelia.

Counts of the blood cells in freshly collected snails give an indication of the variability of the haemocyte population under natural conditions, a variability found also in other invertebrates (see Yeager & Tauber, 1935; Nicol, 1960). Differences between individuals are far less marked after one week in captivity under identical conditions, as shown in Table 1. In addition to this variability between individuals there is also a considerable difference between samples of haemolymph taken from different sites ; moreover, these differences are just as marked in snails kept for weeks or months in the laboratory as in those which have been brought straight from the beach. In general, samples from the heart and arteries are richer in haemocytes than those taken from the veins, while the sinuses show a still lower haemocyte population. Samples of haemolymph taken in connexion with the present series were invariably drawn from the buccal sinus.

We were able to distinguish clearly only two types of haemocyte, though many intermediate forms were found to occur. Using the terminology of George & Ferguson (1950) these may be referred to as lymphoid cells and granular macrophages, the latter Being more numerous and larger than the lymphoid cells. We were not able to distinguish ‘eosinophilic granular amoebocytes’ as described by George & Ferguson for a variety of gastropods. It was, however, possible to confirm the observations of Kollman (1908) that mitotic divisions may be seen in haemocytes but are confined to the smaller lymphocytes. In samples from freshly collected untreated snails, mitotic figures are seen less frequently than once in 2000 cells (total count), but the mitotic index increases after the injection of Thorotrast. The increase is most marked in the first 3 days after injection, indices of 11 to 14 being common by the fourth day. The index continues to increase, but more slowly, until maxima of 16 to 18 are reached after about a week. The number of mitotic figures then begins to decrease but so slowly that it has not been possible to follow its return to normal. Mitotic indices of 4 to 7 may still be encountered months after injection of the dye.

The increase of the haemocyte population following injection of Thorotrast is reflected in Table 1. It can be seen that the apparent increase is far from uniform and that in some cases there appears to be an actual decrease in the number of haemocytes during the first day or two. A sharp increase is only found on the fourth or fifth day after injection. Few of the cells extracted at any stage contained thorium dioxide particles and the lymphocytes were invariably free of them.

Unlike vertebrate animals, Bullia is able to eliminate from its body foreign particles which cannot be metabolized and which are completely insoluble. That such particles are phagocytosed by the haemocytes, which then migrate out of the body, is in keeping with the findings of previous workers on the Mollusca. Both radiographs and sections show that several routes may be followed by such laden haemocytes in reaching the outside of the animal. The chief of these involves the heart, the pericardial cavity, the lumen of the kidney and the mantle cavity, in that sequence; though impossible to observe directly in our experiments, it is logical to suppose that the renopericardial canal and the nephropore are also involved. The chief pathway for migrating haemocytes would thus appear to be from the haemolymph through the wall of the heart (both auricle and ventricle) into the pericardial cavity, then through the renopericardial canal into the lumen of the kidney and out through the nephropore into the mantle cavity. This contrasts with the findings of Tripp (1961) with respect to the gastropod Australorbis, in which the haemocytes are said to migrate mainly through the mantle epithelia and adjacent surfaces.

Minor routes in Bullia are through the pericardial wall from the surrounding tissue, through the mantle itself from the pallial system of blood spaces, and through the tissues of the kidney—presumably from the associated sinuses. If migration does occur through the branchial epithelia it must be very slight indeed. The absence of migration through the digestive gland and gut is in marked contrast to what occurs in the oyster (Stauber, 1950) though in neither animal is the reproductive system involved. Migratory pathways in Bullia are shown graphically in Text-fig. 1.

The migration of haemocytes through the heart wall is probably not as laborious as might at first be supposed. As in other gastropods, there is no endothelium and the muscles are everywhere invaded by blood spaces. In section the entire heart wall has a spongy appearance and it is clear that haemocytes may penetrate almost to the epithelium dividing the heart from the pericardial cavity without being obstructed by the tissues. That a filtration of fluid may occur across this epithelium in some molluscs has been suggested (see Nicol, 1960, etc.). Once in the pericardial cavity the only way out of the system is through the renopericardial canal, whose cilia, beating always away from the heart, may well help to transport the haemocytes into the lumen of the kidney.

Particles of Thorotrast in sea water or invertebrate blood tend to form small, fairly uniform aggregations of a size conveniently phagocytosed. This is doubtless the reason why no fibroblast formation was witnessed, for in Australorbis fibroblasts are only formed where the foreign particles are too large for phagocytosis (Tripp, 1961). The size of the particles also probably accounts for the fact that the initial distribution of the dye takes place rapidly, there being no occlusion of the blood vessels as witnessed by Stauber (1950) after injecting indian ink into the oyster.

Though it is apparent that Bullia is able to deal with foreign, non-metabolizable particles efficiently and with a minimum of inconvenience or disruption of its normal mode of life, it has been decided not to attempt any comparison with other animals as far as rate of disposal is concerned. As already stated, this rate almost certainly depends on the prevailing temperature and the nutritional state of the animal. It may also depend on the size and nature of the particles and probably varies with the quantity injected. The relationship between this latter factor and rate of disposal is likely to be very complex for it has been shown that the haemocyte count increases after the injection of foreign material; there is no indication that this increase is proportional to the amount of material injected and there must, in any case, be a limit to the possible rate of cytogenesis. A large number of variables is thus involved in the rate of disposal of Thorotrast or other foreign material.

The ‘latent period’ of 3 or 4 days between injection of the dye and increased haemocyte counts may be explained on the grounds that macrophages laden with foreign particles tend to adhere to surfaces. This is a reasonable assumption in view of their subsequent migrations and is supported by a study of sections. It is also supported by the fact that the majority of haemocytes seen in extracted blood did not contain thorium particles, for if they were attached to the surfaces bounding the blood spaces and vessels they would not be available for extraction with the blood. If this is true then the increase in haemocyte count after 3 or 4 days actually represents the final elimination of free particles of thorium from the circulating fluid and implies that the haemocyte population begins to increase before then.

The origin of the haemocytes has interested several workers on the Mollusca, though little concrete evidence has been presented. Cuénot (1896) reported that the blood cells of gastropods divide by mitosis, mitotic figures sometimes being present in considerable numbers. Kollmann (1908) found division limited to the lymphocytes and considered that these cells give rise to the macrophages and other haemocytes, a view in direct opposition to the earlier theory of Cuénot (1891). The findings of George & Ferguson (1950) tended to support the views of Kollmann. Tanaka, Takasugi & Maoka (1961) working on the oyster, Gryphea, obtained a type of cell from the ‘liver ‘from which the five types of haemocyte present in this animal appear to develop.

Our findings on Bullia definitely support the theory of Kollmann, postulating the development of the larger macrophages from lymphocytes. The lymphocytes divide while the macrophages do not appear to do so, and intermediate forms occur. More striking evidence is provided by the fact that, although only the macrophages take up thorium particles, the injection of such material causes an increase in the lymphocytic mitotic index as well as in the total cell count. It would be surprising indeed if these events were unrelated. It may be asked why, if the lymphocytic mitotic index increases by more than 1700%, the total haemocyte count increases only four or five times. The answer is quite obviously that enormous numbers of cells are leaving the system, laden with thorium dioxide, so that effectively the rate of mortality of the macrophages has been very greatly increased by the injection of Thorotrast. Increased cytogenesis may thus be seen as necessary not only to increase the number of cells which can remove the foreign particles from the blood but also to replace the cells which have been lost to the body in performing this function.

The first author (A.C.B.) takes full responsibility for the planning of this work and the interpretation of the results, the second author being concerned entirely with the radiographic techniques. We are indebted to Dr Spong, of the University of Cape Town, for his advice on the handling, storage and disposal of Thorotrast and to Dr G. J. Broekhuysen for his comments on the preparation of positive radiographs and micrographs. The expenses involved were largely met by a research grant made available by the University of Cape Town.

Plate 1 shows three positive prints made from radiographs of Bullia laevittima which had been killed and removed from their shells at different times after the injection of Thorotrast. The exposure in each case was 0-03 sec. at 100 mA. and 55 kV. at a focus-film distance of 100 cm. The figures are slightly larger than natural size.