Tissues for Assessing Histochemical Methods for Calcium

With one plate (fig. 1)

IT is desirable that histochemical methods be assessed critically. Histo-chemical methods should be reproducible in a test-tube and follow known chemical criteria, and should produce consistent results in sections of animal or plant tissues. For test purposes such consistent results should be related to the known presence of the indicated substance in the tissue in reasonable quantity, as shown by chemical or biochemical analysis. If possible the distri-bution of the substance in the tissue should be indicated by criteria other than the results of the histochemical method alone, if that tissue is to be used as a true ‘test’ of the method. In the tissue the substance should be present in a condition neither ‘masked’ nor ‘bound’, so that its reactions parallel the reactions which it gives as a single substance in a test-tube. These require-ments are hard to meet for many of the substances for which histochemical methods have been suggested. It is a peculiar difficulty of histochemical methods that they have to be carried out on substances which are in close asso-ciation with a meshwork of precipitated proteins, and other substances. The spurious results which may result from this association are well known. However, it is because of this difficulty that critiques of methods should be carried out upon tissue sections as well as models, if possible.

The conditions outlined can be met for some substances, and this paper gives an account of three tissues from invertebrate animals which can be used satisfactorily as standard objects when trying out or criticizing methods said to reveal calcium salts in tissues. These tissues are preferable to the bone in young vertebrates, which has been used by some workers (e.g. Stock, 1949), as the matrix of bone has special staining properties which may give false differential colourings unrelated to the calcium present. The tissues may be used in conjunction with models of the type used by Cameron (1930) and give parity of results, and allow one to form a better idea of the precision of the cytological localization.

The three tissues are the digestive gland of the snail Helix (either Helix aspersa or H. pomatia), the gut diverticulum of the crab Carcinus maenas, and the left colleterial gland of the cockroach Periplaneta americana. The snails were obtained in the neighbourhood of Oxford, and were kept in the laboratory on a diet of cabbage. The crabs were obtained fresh from Plymouth. The cockroachs were from a stock maintained by Dr. P. C. J. Brunet at Oxford.

The main requirement of the investigation of these three tissues as suitable test objects was to establish as clearly as possible the presence of calcium deposits in them and the nature of the calcium salts present.

Three alternative techniques were available to identify calcium in the tissues : microchemical methods applied to histological slides ; microincineration of histological sections and application of microchemical tests to the ash ; differential centrifugation of the homogenized tissue and application of chemical tests to the fractions, if microscopically identifiable as corresponding to elements in the tissue. All these methods were employed.

Lison (1936, 1953) states that the microchemical method of Schujeninoff is the only completely specific method for the detection of calcium. The section is treated with dilute sulphuric acid mixed with alcohol on the coverslip. In the presence of large quantities of calcium salts characteristic rosettes of gypsum crystals are formed. I found that this reaction does not occur strikingly with the three tissues in section, before incineration. However, after microincineration, the nature of the white ash is in the first instance a good indication of the metal present, and if dilute sulphuric acid is applied to the ash, the rapid formation of gypsum crystals is certain evidence of the presence of calcium.

In all three of the test tissues, microincineration produced white ash at the supposed sites of calcium. Application of dilute sulphuric acid resulted in the rapid formation of gypsum crystals, further confirming the presence of

Histochemical Methods for Calcium 3 calcium. The localization of the ash is badly disturbed by the application of the acid, and there is a tendency for the gypsum crystals to develop at the edges of the acid drop rather than in any relation to the previous distribution of the ash, as noted by Manigault (1936−7). Nevertheless, the sites of calcium are sufficiently clearly indicated (fig. 1, B).

The detailed histology of the digestive gland is complex. However, the most easily identifiable cell type is the calcium cell, Kalkzelle, or ‘lime cell’. The last term is a misnomer, unfortunately used in the literature, for it is not likely that the animal isolates calcium oxide as such.

The calcium cells occur throughout the epithelium of the digestive tubules, in between the digestive cells. They are usually about 50 μ across at the base where the cells are attached to the basement membrane and connective tissue sheath. They are sometimes roughly triangular in outline in sections, being shorter than the adjacent digestive cells, which are usually about 100 μ in height and appear to squeeze the calcium cells slightly. Calcium cells in teased preparations of living tissue become spherical, with a diameter of about 50 to 60 μ, and are therefore about as wide as they are tall when in place in the epithelium. The large oval nucleus is central in the cell. The long axis of the nucleus is about 10 to 20 μ, and the nucleus possesses one, two, or three well-marked nucleoli and an abundance of karyosomes (fig. 1, A). The surrounding calcium spherules may give the nucleus an indented appearance, which may have led Cretin (1923) to describe it as resembling a chestnut. The calcium spherules pack the cytoplasm of the calcium cells. They are usually about 4 to 5 μ in diameter and are perfectly spherical. However, they may vary in size down to 1 μ or less. They are highly réfringent and show in some cases internal structure in addition to the external protein coat which they all possess. The differentially centrifuged spherules may be decalcified with dilute acid under the coverslip, and the external protein coats remain spherical and undistorted, but are only visible by phase-contrast microscopy. In their characters the calcium spherules perfectly parallel the spherules figured by Waterhouse (1950), and their form is typical of the calcosphèrites which are found in a variety of invertebrates (Keilin, 1921).

Review of the literature quickly establishes that there are two schools of thought on the nature of the calcium salt composing the spherules in the calcium cells, as has recently been pointed out by Fretter (1952). Robertson (1941) in his extensive review of calcium in invertebrates identifies the spherules as composed of calcium phosphate, in accordance with such authors as Sioli (1935) and Krijgsman (1928). More recently Wagge (1948) and Heaysman (1951) found that the spherules were principally composed of calcium carbonate, and this view was previously held by Barfurth (1880) and others. The accumulation of radioactive phosphorus in the calcium cells has been established by Fretter (1952), and this author suggested that there was probably a balance between phosphate and carbonate in them. This amiable compromise resolves the controversy, and it is certainly a likely condition. However, it is difficult with the radioautographic technique used by Fretter to determine whether the phosphorus is accumulated actually within the spherules, or whether it is in the cytoplasm immediately surrounding them, which is another possibility.

The spherules showed the following properties :

(a) a weak birefringence between crossed polaroids, with the form of a black polarization cross, indicating a radial arrangement of micellae within the spherules ;

(b) effervescence with dilute acids such as hydrochloric acid. The gas was identified as carbon dioxide ;

(c) after fixation and embedding the spherules were very noticeable objects in sections, because of their refringency, and under a Watson 2 mm fluorite objective they appeared to have a violet-pink colour, probably as a consequence of their optical properties.

Tests with ammonium molybdate solution, and with silver nitrate, did not produce any significant results.

The weakness of the anisotropy of the spherules is surprising if the calcium is present as carbonate, unless it is in the form termed ‘amorphous’ (Prenant, 1927). However, the amount of effervescence is certainly sufficient to suggest the presence of a high quantity of carbonate. Heaysman (1951) goes so far as to identify the salt as aragonite upon the basis of Meigen’s reaction. In a cold solution of cobalt nitrate calcite does not colour, and colours very little in a warm solution, whereas aragonite, especially with the hotter solution, is coloured lilac or violet (Lison, 1953). My tests showed that the colour of the calcium spherules was indistinguishable between control sections deparaffinized in xylene and mounted immediately in balsam, and sections which had been brought to distilled water and treated with warm cobalt nitrate solution. Definitive identification of the crystalline form of the spherules must wait upon crystallographic studies of the spherules, after separation from the tissue by centrifugation.

The positive identification of the spherules as calcium phosphate by authors such as Krijgsman (1928) is puzzling, unless the proportion of carbonate and phosphate varies considerably from one locality to another, perhaps in accordance with diet. An alternative explanation is that the methods of identification used were insufficiently precise. I have been unable to confirm Krijgsman’s observation that the addition of ammonium molybdate solution, slightly acidified with nitric acid, to tissue sections, accompanied by gentle warming, produces a yellow precipitate with the spherules. He also found that the spherules dissolve in dilute acids without effervescence. I do not agree. His most demonstrative proof of the apparent phosphate content of the spherules was obtained by boiling the fresh gland for 2 minutes with concentrated nitric acid, filtering, and adding ammonium molybdate solution. A yellow precipitate insoluble in acids but soluble in ammonia was produced. The hydroquinone / molybdate test confirmed the presence of phosphate by a deep blue coloration. The whole tissue, if treated in this way, undoubtedly can be shown to contain phosphate. The radioautographic studies of Fretter (1952) clearly show that the digestive gland is a storage organ for phosphate. Nevertheless, the properties of the snail spherules themselves, and their contrast with the properties of the spherules found in Carcinus, convince me that the calcium salt is predominantly the carbonate in the snails which I have studied.

It is often possible to find areas of calcareous deposits immediately below the mantle covering the digestive gland, which differ from those in the cells of the digestive epithelium. The deposits are intracellular, in the form of calcosphèrites which are usually smaller than those of the calcium cells of the digestive gland (fig. 1, c). The structure of the cells containing them is often difficult to make out in fixed preparations, but the cells appear to be the same as the large round cells which are a characteristic component of the connective tissue of molluscs. It is often more convenient to dissect out the whole tip of the visceral hump without removing the overlying mantle, when fixing the digestive gland ; in sections of such material the submantle deposits are found. They would appear to correspond to the deposits figured by Manigault in the connective tissue of Helix (Manigault, 1939, fig. 47).

The ovotestis lies within the loop of the digestive gland which occupies the tip of the visceral hump. In sections of such material the ova within the ovotestis may show a further site of calcium in the form of small granules within the egg cytoplasm (fig. i, D).

In fixed preparations the epithelium of the gut diverticulum is found to be composed principally of two types of cell: a columnar cell with vacuolated cytoplasm, throughout the basal region of which a number of calcium spherules is scattered, and a somewhat narrower columnar cell with an intensely basiphil cytoplasm (fig. 1, E). The two types are easily distinguishable in preparations stained with toluidine blue, in consequence of this basiphilia. The cells are approximately 60 μ in height and have a basal nucleus of diameter about 10 μ, which has a well-marked nucleolus and very little chromatin. Further subdivision of the cell types of this epithelium is irrelevant to the present purpose. The cells are extremely fragile, and in teased preparations, mounted under the cover-slip, it is usual to find large numbers of calcium spherules released from shattered cells. In fixed preparations it is noticeable that the number of calcium spherules to each cell is smaller than in the snail tissue, where the entire volume of the cell is packed with spherules. It would appear from the basal distribution of the calcium spherules in the crab cells, and the considerable vacuolation of the distal cytoplasm, that these cells have additional functions other than calcium storage.

The nature of the calcium stores of the ‘liver’ or digestive gland of the crab does not appear to have been in dispute. Robertson (1937) states that the spherules are largely composed of calcium phosphate, and my findings are in agreement with this statement.

The spherules are completely isotropic. Mounted immediately after removal of the paraffin from sections, they do not show the apparently tinted and réfringent appearance of the snail spherules, but are colourless. With dilute mineral acids they do not produce effervescence, but dissolve rapidly. The ammonium molybdate reagent does not produce a noticeable yellow precipitate when applied to sections, but a very heavy positive coloration for phosphate is formed when it is applied to spherules separated out by centrifuging. Application of silver nitrate solution to sections and to the separated spherules results in the rapid formation of a yellow precipitate on and around the spherules ; this can only be the iodide or the phosphate. When dilute acid is applied to a very large aggregation of differentially separated spherules, there is rapid dissolution, and occasionally a very slight amount of effervescence, which may indicate that a small quantity of carbonate is present. With dilute sulphuric acid, gypsum crystals are quickly formed. There can be little doubt that the spherules contain a preponderance of calcium phosphate.

Brunet (1952) has reviewed the literature concerning the identification of the crystals in the lumen of the colleterial tubules (fig. 1, F). There is no disagreement over the identification of them as calcium oxalate by Hallez (1909). My investigations confirm the presence of calcium in these crystals, and their high insolubility suggests the oxalate salt. Microchemical tests accord with this identification.

In order to study the distribution of calcium in a tissue, fixation must be carried out in non-acid fixatives which do not contain metal salts likely to interfere with the histochemical methods used.

A mixture of absolute alcohol with concentrated (40%) formaldehyde solution (50: 50) was found to be preferable to alcohol alone, for the fixation of the tissues discussed. Alcohol alone produced considerable distortions. The pH of the mixture was measured as somewhat acid (pH 47) with a Marconi glass electrode pH meter, but this was not found to disturb the localizations of the calcium deposits mentioned above. Fixation was usually carried out overnight. It is reasonable to adjust the pH of this fixative to neutrality with alkali, if this is felt to be necessary, although it results in no detectable change in the histochemical distribution of the calcium deposits.

A number of carefully controlled comparisons of the results obtained by fixing pieces of the same tissue in alcohol, formaldehyde-alcohol (50: 50), and formaldehyde-calcium (Baker, 1945) showed that the use of the last fixative did not lead to false histochemical distributions of calcium, and that fixation was reasonably good. However, it is very possible that in other tissues there are likely to be sites where calcium might be taken up from this fixative, and the additional controls necessary when using it militate against it.

The tissues may be used most satisfactorily if embedded in 58° or 60° C paraffin wax and sectioned at 6 μ. Thin sections are advisable. The tissues may also be embedded successfully in celloidin, ester-wax, ‘Aquax’ (G. T. Gurr Ltd.), or gelatin. The practice of hardening gelatin blocks in formaldehyde / alum solution was found to be inadvisable, as it tended to remove the calcium salts present. Satisfactory hardening was obtained with 70% alcohol.

My thanks are given to Professor A. C. Hardy, F.R.S., for the facilities he afforded me in his Department, and to Dr. J. R. Baker and Dr. P. C. J. Brunet for much encouragement. I am grateful for the scholarship awarded me by the Christopher Welch Trustees, and for the additional financial assistance of the Department of Scientific and Industrial Research.