The Cytoplasmic Inclusions of the Neurones of Patella vulgata

Living specimens of P. vulgata were obtained from Plymouth. The pedal ganglia (‘nerve-cords’) were dissected out in sea-water. Neurones are obtain-able from any part of the long ganglia. Occasionally the cerebral or pleural ganglia were used instead of the pedal, but their neurones were not fully investigated. For studies by phase-contrast microscopy the ganglia were teased in sea-water, slightly compressed by the coverslip, and examined without further treatment. Positive phase-contrast was used. The following vital dyes were tried: neutral red, Janus green, brilliant cresyl blue, methylene blue, Nile blue, and dahlia. Each dye was dissolved at 0-5% in distilled water. Two drops of this stock solution were added to 2-5 ml. of sea-water, and the ganglion was gently teased in this. Dyeing was allowed to proceed for 10 to 20 min, and the fragments of tissue were then placed on a slide in the same fluid and slightly compressed by the coverslip.

The histochemical tests applied are listed in the Appendix (p. 299).

The Golgi techniques of Aoyama (1929), Weigl (Mann-Kopsch) (1910), and Kolatchev (1916) were used. Routine microscopical preparations were made.

The cytoplasmic inclusions seen in neurones of P. vulgata are mitochondria and two kinds of lipid globules.

The mitochondria are mostly filamentous or rod-shaped. They are well seen by phase-contrast microscopy. The best fixatives are Altmann, Champy, and Helly; the best methods of staining are those of Metzner (Metzner and Krause, 1928) and Hirschler (1927). They appear thicker by Hirschler’s than by Metzner’s technique. Some of them can be blackened by the Kolatchev technique.

The two kinds of lipid globules are readily distinguished in life, one kind being yellow, the other devoid of colour.

The yellow globules vary in diameter from about f p. to They are generally not quite spherical, especially the bigger ones.

The yellow globules are vitally coloured red by neutral red ; they are coloured green by methylene blue, brilliant cresyl blue, and Nile blue, by mixture of their own colour with that of the dye.

The yellow pigment is evenly distributed throughout the globules. On treatment with concentrated sulphuric acid the globules become blue-black. Their colour is therefore presumably due to a carotenoid pigment. Such pigments occur commonly in globules contained in the neurones of gastropods (Cain, 1948; Chou, 1957).

Cerebroside is probably present in large quantity in the yellow globules, since in sections of ganglia fixed in cold acetone they give a strongly positive reaction with Sudan black. No cholesterol or fatty acid has been demonstrated in them; the response to Cain’s test for plasmalogen (Cain, 1948) is weak and uncertain. Phospholipid appears to be restricted to their surfaces, but this may not have been its distribution during life.

Although the yellow globules are essentially lipid, yet there are other constituents. They react quite strongly to the PAS test, but feebly after the action of saliva. This is suggestive of the presence of carbohydrate. The residual colour after the use of saliva is presumably due to the presence of cerebroside.

The globules appear to contain tyrosine and histidine. The evidence for the presence of the latter is a positive reaction to the coupled tetrazonium test, controlled by blocking reactions (see Appendix). It is therefore to be presumed that the globules contain protein. They are strongly basiphil, and become orange with pyronine/methyl green (PMG); Feulgen’s test is negative.

There is a feebly positive reaction to Gomori’s test for alkaline phosphatase.

The globules that are without natural colour are easily dyed blue in life by methylene blue, and they will therefore be called ‘blue’ globules ; neutral red will also colour them. They are spherical and generally less than 1 /z in diameter ; most of them are smaller than the smaller yellow globules. Although most of them are coloured blue in life by brilliant cresyl blue, the smallest of them show the metachromatic colour of this dye. The globules are easily seen in life by phase-contrast microscopy. They appear to have a dark rim when examined by positive phase-contrast; this is presumably an optical artifact (‘optical membrane’ of Oettlé, 1950).

Unlike the yellow globules, the ‘blue’ ones are almost entirely lipid in composition. Carotenoid is absent; the two chief lipids are cerebroside and phospholipid. The former constituent is particularly easily shown by the method of Casselman and Baker (1955). Uncertain or feebly positive reactions are given to the PAS test and to the tests for amino-acids.

Both the yellow and ‘blue’ globules become grey with the Weigl technique, but the ‘blue’ ones become darker than the yellow. The Kolatchev technique is unrealiable with these globules: sometimes it darkens some of them. Aoyama’s method does not darken them. None of the ‘Golgi’ methods shows a typical ‘dictyosome’ picture. The ‘blue’ globules are coloured red by Metzner’s method for mitochondria (as lipid globules often are). When a fixative containing osmium tetroxide is used, the grey caused by the osmium mingles with the red of the Metzner dye and darkens it.

Four kinds of cells are found in the pedal ganglia. Three of these, which I shall call large, small, and yellow, are unipolar ; there are also a few bipolar cells.

Large cells (figs. 1, A and 2, A). Although these are larger than the others, yet they are much smaller than the neurones of many molluscs. Their diameter, transverse to the direction of the axon, is about 13 p. The cell is pear-shaped, tapering gradually to the axon.

The yellow globules of various sizes are distributed in the periphery of the cell. The ‘blue’ globules are almost confined to the centre.

Small cells (figs. 1, B and 2, B). The shape is similar to that of the large cells. The diameter tranverse to the direction of the axon is about 9 /z. These cells are chiefly distinguished by the absence or rarity of yellow globules. The mitochondria are similar to those of the large cells.

Yellow cells (figs. 1, c and 2, c). These are markedly different from the large and small cells. They are nearly spherical instead of pear-shaped. The axon is difficult to see in the living cell unless it happens to lie at right angles to the optical axis of the microscope ; but it can be seen in certain fixed preparations (for instance, in ganglia treated by Holmes’s method), and there can be no doubt that these are nerve-cells. Their most striking feature is that they are almost completely filled with yellow globules. The latter resemble those in the large cell in all respects except that there are hardly any small ones : very few of them are less than 1 μ, in diameter. The globules are so numerous that the nucleus cannot be distinguished in the living cell (fig. i, c), but it is easily seen in fixed preparations. It is remarkably small (fig. 2, c).

The dense packing of the yellow globules in these cells makes the study of mitochondria difficult. In preparations fixed for mitochondria (Altmann, or Helly with postchroming) and dyed by Metzner’s method, small red globules are seen between the yellow ones (fig. 2, c).

Bipolar cell (fig. 2, D). These are too infrequent for full study and I have seldom been able to distinguish them in life. The mitochondria seen in fixed preparations are shorter thain in the large and small cells, and not numerous. They are best shown by Hirschler’s method. The rather scanty evidence suggests that the globules are of the same nature as the ‘blue’ globules of the large and small cells, but they are smaller.

The neurones of Helix are not so diverse as those of Patella. In the former animal I have not seen cells corresponding to the yellow and bipolar neurones of the limpet. The neurones of Helix are variable in size. Since the larger ones contain yellow globules while the smaller contain few or none, there is a general resemblance to the large and small cells of Patella, but there are no cells of intermediate size in Patella.

The yellow globules of Patella show a general similarity to those of Helix. They are in both cases histochemically complex, since the evidence suggests that they contain lipid, protein, and carbohydrate. The lipids of the globules are, however, simpler in Patella : I have only been able to demonstrate carotenoid, phospholipid, and cerebroside (the latter in great quantity). The yellow globules of Helix contain also cholesterol and its esters, and plasmalogen; there is much less cerebroside than in Patella.

The ‘blue’ globules of both animals contain phospholipid, but little or no protein of carbohydrate. Those of Patella contain cerebroside in addition to phospholipid.

It may be recollected that the yellow and ‘blue’ globules of Limnaea are similar to those of Helix (Chou, 1957). Thus the globule system of these three gastropods is essentially similar. The highly refractile ‘colourless’ globules of the neurones of Helix are_; however, not represented in Patella.

All the cytoplasmic inclusions seen in the various neurones of Helix, Limnaea, and Patella are represented diagrammatically in fig. 3. No neurone of these three animals actually contains all the kinds of cytoplasmic inclusion shown in the diagram. Most neurones, however, contain filamentous mitochondria and ‘blue’ (wholly lipid) globules.

The cells of Patella studied by Lacy and Rogers (1956) and Lacy and Horne (1956) were evidently the large cells. They saw the pigmented or ‘lipochrome’ globules, and noticed that the globules in the centre of the cell are smaller than the others. They noticed that the long threads in the centre of the cell can be blackened by post-osmication in the Kolatchev technique, but did not remark that they can also be coloured by mitochondrial staining methods. The fact that certain mitochondria in a cell may differ from others in their capacity to reduce osmium tetroxide was shown recently by Meyer (1957) in his study of the neurones of Hirudo.

It is important to recognize that different authors have applied the term ‘Golgi’ to entirely different constituents of the gastropod neurone. Moussa (1950) and Boyle (1937) give the name of ‘Golgi’ to deposits of osmium or silver on the surface of the yellow globules. I have shown that in Helix the curved rods or ‘dictyosomes’ seen in Weigl (Mann-Kopsch) preparations are artifacts caused by the modification of the ‘blue’ globules. Lacy and Rogers (1956) and Lacy and Horne (1956), on the contrary, regard the long threads of the central region of the neurone of Patella as the ‘Golgi elements’, presumably because they are easily osmicated. They brought forward no histochemical evidence, however, of any chemical similarity to any object in vertebrate neurones described by Golgi. Blackening by post-osmication gives no histochemical information, since a very wide variety of chemically unrelated substances are capable of reducing osmium tetroxide (Bahr, 1954).

It seems best to describe objectively the distribution, size, shape, and structure of cytoplasmic inclusions, their reactions to vital dyes, and their composition as revealed by reliable histochemical tests. It would not appear that anything is gained by using the name of Golgi in describing these structures.

Several authors have described secretory processes in these cells, but what they say is hypothetical. I have no concrete evidence of the time-sequence of the supposed stages.

1. The cytoplasmic inclusions of the neurones of the pedal ganglia are filamentous mitochondria and lipid globules.

2. Two kinds of lipid globules occur. One is yellow, the other devoid of natural colour.

3. The yellow globules owe their colour to the presence of carotenoid. They are chemically complex, since they contain not only cerebroside and phospholipid, but also a certain amount of amino-acids (presumably as protein) and carbohydrate.

4. The other globules consist almost entirely of cerebroside and phospholipid.

5. Four kinds of neurones are present, of which three are unipolar and one bipolar. The three kinds of unipolar neurones are easily distinguishable by their morphological characters and by their differing content of lipid globules.

I have great pleasure in acknowledging my debt to Dr. J. R. Baker for suggesting and supervising this investigation, and to Professor A. C. Hardy, F.R.S., for providing me with facilities for working in his Department.

The work was carried out during tenure of an Inter-University Council Fellowship through the Carnegie Corporation of New York, and study leave from the Department of Zoology, University of Hong Kong.