Microdissection Experiments on Sea-Urchin Eggs At Cleavage

This paper describes a series of microdissection experiments on sea-urchin eggs which were done in the course of investigations into the mechanism of cell division. It was hoped that these experiments would throw some light on the validity of a theory of cleavage which M. M. Swann and I have developed and which we have called the ‘expanding membrane’ theory. This theory has been outlined earlier (Mitchison, 1952), and it is only necessary here to describe a few of the salient points which are relevant to the microdissection experiments. We believe that cleavage is started by the diffusion of a substance from the two groups of daughter chromosomes at the end of anaphase. This substance reaches the poles of the cell and starts a wave of expansion in the cell membrane or cortex which gradually spreads round to the equator. The furrow is formed initially by the passive contraction of the elastic cortex at the equator, but its subsequent advance is due to an active inwards expansion of the cortical material in the furrow wall when the diffusing substance reaches it. In order that this mechanism should work, there must be no tension in the cell membrane during the later stages of cleavage and therefore no internal pressure in the cell. The asters are not thought to play an active part in the process, but only to act as passive guides for the advancing furrow and thus ensure an orderly cleavage.

There are two other main theories of cleavage. In one, cleavage is thought to be due to the contraction of a ring of gelated material at the equator, e.g. Marsland (1951). In the other, an astral mechanism is invoked, with the active process being either an expansion of the spindle remnant (Dan, 1943), or a growth of the asters (Gray, 1924).

Some of the experiments to be described are not original and were done for confirmation and, to some extent, clarification of results of Chambers, who has done the pioneer work in this field. Some of the experiments on tearing blastomeres have already been described in Mitchison (1952).

The sea-urchin eggs used were those of Psammechinus millaris, except in the experiments on tearing blastomeres where the eggs of Arbacia lixula were also used. It was necessary to remove the fertilization membrane and any traces of the egg jelly. This could be done quite easily with the Psammechinus eggs at Millport by passing them through fine mesh bolting silk about 2 min. after fertilization. In the case of the Arbacia eggs at Naples, the procedure was more complicated and produced a larger number of broken eggs. The jelly was removed by treating the unfertilized eggs for a short time with acidified sea water. The fertilization membranes were removed by placing the eggs in isotonic urea about 2 min. after fertilization, stripping off the swollen membranes by passing through bolting silk, and then washing off the urea as soon as possible.

The calcium-free (Ca-free) sea water used to remove the hyaline layer in some of the experiments was made up of NaCl, 60 pts. ; MgSO₄, 13 ; KC1, 1·5 ; NaHCO_s, 1 ; adjusted to isotonicity with distilled water. The calcium and magnesium-free (Ca and Mg-free) sea water used in the blastomere puncturing experiment was made up of NaCl, 65 pts.; KCl 1·75; NaHCO₃, 1; adjusted to isotonicity.

All the experiments were done with a Zeiss Peterfi micromanipulator under a 16 mm. objective, and the microdissection instruments were made with a de Fonbrune microforge. Room temperatures varied from 17 to 21° C.

Chambers (1938) described a microdissection experiment in which he tore open one blastomere at a dumb-bell stage in Arbacia punctulata eggs which were cleaving in KC1 isotonic with sea water. The damaged blastomere disintegrated, and he said that ‘the absence of divalent cations in the medium permitted the continued outflow… of the interior of the egg…’. If the cleavage furrow was far advanced, it continued its course after the operation, and closed off the intact blastomere. If, on the other hand, it was not far advanced ‘… the contents of the untorn blastomere flowed through the connecting stalk while the blastomere shrank in size much like a gradually deflating balloon’. In this case both the intact and the tom blastomere disintegrated. Sichel & Burton (1936) assumed from this experiment that there was a positive tension in the membrane, and calculated it to be 0-09 dyne/cm. It is an essential point in the ‘expanding membrane’ theory that there should not be a positive tension in the intact membrane during the later stages of cleavage, so it was decided to repeat these experiments and see whether the interpretation seemed correct. It also seemed necessary to try this experiment in other media since Chambers appeared to have found differing results in them. In his paper of 1938, he says that the major part of one of the blastomeres of a cleaving egg can be cut away without disturbing the course of the cleavage furrow, and that the presence of the hyaline layer is not essential. This would imply that the deflating balloon effect does not occur either in ordinary or in Ca-free sea water. On the other hand, he describes later (Chambers, 1951) the deflating balloon experiment as being done in Ca-free sea water. In another paper (Chambers, Chambers & Leonard, 1949), he describes a contraction of the egg and an outflow of the contents when punctured at the astral stage in a medium containing 9 parts NaCl to 1 part KCl. This medium is free of calcium and magnesium, but it has over four times as much potassium as is normally present in sea water.

Preliminary experiments were done on Psammechinus miliaris eggs in ordinary, and in Ca-free sea water. When the dividing eggs had reached the dumb-bell stage, they were gripped over the furrow by a holding needle (shaped like a question mark), and one of the blastomeres was torn in two by a slicing cut with a straight needle. In ordinary sea water, the cut sealed at once and the egg continued to cleave with the tom blastomere represented by a small and usually irregular shaped fragment. In Ca-free sea water, the reaction which followed the puncturing depended on the diameter of the furrow neck. If the furrow was less than about 30μ, the furrow continued to advance and sealed off the intact blastomere (Pl. 14, figs. 1, 2). There was no flow through the neck, and the intact blastomere was unaffected. The tom blastomere, however, showed a striking injury reaction which spread through it in about 10 sec. and which appeared as a granulation and slight swelling of cytoplasm, a roughening of the surface, and the assumption of a brownish colour in transmitted light (this is probably the same as the ‘black cytolysis’ of some authors). The tom blastomere did not dissolve away. If, on the other hand, the furrow diameter was greater than about 30 μ, the injury reaction spread through the furrow neck to the intact blastomere which swelled a little and turned the same brownish colour as the tom blastomere (Pl. 14, figs. 3-5). This was accompanied by a partial retraction of the furrow and by a momentary movement of the cytoplasm in the furrow neck towards the torn blastomere, but these reactions appeared to be due rather to the swelling of the cytoplasm in the previously intact blastomere than to the contraction of its membrane. There were no signs of the regular continuous movement, or the contraction of the previously intact blastomere, which were found by Chambers.

The main experiments were done in a number of different media on the eggs of Arbacia lixula, since this sea urchin is the nearest equivalent in Europe to the American A. punctulata used by Chambers. In ordinary sea water, in Ca-free sea water, and in Ca and Mg-free sea water, the Arbacia eggs reacted to tearing in the same way as the Psammechinus eggs in ordinary sea water. It was impossible to cause cytolysis or outflow from the eggs since the cuts sealed up at once, and cleavage continued with large parts of one blastomere removed. This is shown in Pl. 14, figs. 6-10; fig. 6 shows a cleaving egg in Ca-free sea water under the holding needle ; fig. 7 shows it after a part of one blastomere has been removed by a cut, and figs. 8 and 9 show the completion of cleavage; fig. 10 is another egg which has nearly completed cleavage with only a very small part left of the torn blastomere. These photographs emphasize two points which are relevant to theories of cleavage. First, that cleavage can continue with one blastomere in an irregular shape. If there was any substantial internal pressure in the cleaving egg, this blastomere should round up at once. Secondly, the fact that nearly all of one blastomere, together with its astral apparatus, can be removed without stopping cleavage, argues against an astral mechanism of cell division.

If Arbacia eggs were allowed to develop after fertilization in either ordinary or Ca-free sea water and then placed in isotonic KC1 when they were cleaving, they reacted to cutting in exactly the same way as the Psammechinus eggs in Ca-free sea water. If the furrow was nearly closed, only the torn blastomere cytolysed, while if the furrow had only advanced half way through or less both blastomeres cytolysed (Pl. 14, fig. 11). There was no outflow and no contraction of the previously intact blastomere. The only difference between Arbacia and Psammechinus was that in the former the injury reaction was less conspicuous because of the dark colour of the eggs. The conditions of this experiment are as nearly as possible identical with those in the deflating balloon experiment of Chambers, so the finding of a different result seems to be due to the use of another species of Arbacia.

It is important at this point to stress the fact that isotonic KCl is a most unsatisfactory medium to use for microdissection experiments on developing eggs. The most obvious objection to it is that A. lixula eggs would neither develop nor cleave in it unless they were placed there while they were actually cleaving. Even then, only those eggs which were already in the dumb-bell stage completed cleavage, though very slowly and often irregularly. Eggs which were in earlier stages reverted back either to spheres or, more often, to irregular shapes. There was often a curious stratification of the egg contents so that one end of an egg appeared much redder than the other. These changes are shown in Pl. 14, fig. 12, which is a photograph of a field of Arbacia eggs placed in isotonic KCl at cleavage. It should be added that eggs develop and cleave normally both in Ca-free and in Ca and Mg-free sea water.

It is apparent from these experiments that the result of puncturing one blastomere of a cleaving sea-urchin egg depends to some extent on the species of sea urchin. The eggs of Psammechinus miliaris are unaffected in ordinary sea water but cytolyse in Ca-free sea water. The eggs of Arbacia lixula are more resistant; they are unaffected in ordinary, Ca-free and Ca and Mg-free sea water, and only cytolyse in KC1. The main difference between the eggs of Arbacia lixula and those of A. punctulata appears to be that the latter show the deflating balloon effect in KC1 while the former do not.

There are, however, definite points of similarity between the reactions of these eggs. In the first place, Psammechinus and Arbacia lixula show the same kind of cytolysis. Secondly, it appears that the two species of Arbacia only cytolyse in abnormally high concentrations of KC1. This statement implies the following interpretation of Chambers’s work on A. punctulata : that cleaving eggs cytolyse in isotonic KC1 but not in ordinary or Ca-free sea water (Chambers 1938); that the cytolysis in the mixture of 9 parts NaCl and 1 part KC1 (Chambers et al., 1949) is due to the high proportion of KC1; and that when Chambers (1951) describes the deflating balloon effect in Ca-free sea water he is, in fact, describing his original experiment which was done in isotonic KC1. Thirdly, although the deflating balloon effect and eventual dissolution occur only in A. punctulata, the cytolytic breakdowns in all three eggs show certain common features. There is an injury reaction which spreads over the egg (this reaction is described by Chambers in his paper of 1951, though not in his paper of 1938), this reaction spreads to the intact blastomere only if the furrow neck is still comparatively large, and there is flow or indications of movement in the furrow neck only when the injury reaction has spread to the intact blastomere and it is in the process of cytolysing.

In view of the fact that A. lixula eggs are unaffected by cutting in Ca and Mg-free sea water, it seems more probable that the cytolysis and continued outflow of A. pimctulata eggs in isotonic KCl is due to the high concentration of K+ ions rather than to the absence of divalent ions as suggested by Chambers.

In summary, it seems that no legitimate conclusions about the normal state of the cell membrane can be made from the deflating balloon effect for the following reasons: (1) the effect in A. punctulata only occurs in isotonic KC1 (or, possibly, in abnormally high concentrations of K+ ions). In the nearly related species A. lixula the effect does not occur, and, in any case, isotonic KC1 appears to be a very unsatisfactory medium since it inhibits division unless the eggs are placed in it during the actual process of cleavage. (2) The effect only occurs after an injury reaction has spread through the intact blastomere and while the whole egg is in the process of cytolysis.

We can conclude, therefore, that Sichel & Burton’s deduction of a positive membrane tension apply only to one species of sea-urchin egg, in an abnormal medium and after it has suffered irreparable damage. The fact that eggs will continue to cleave with one blastomere in an irregular shape indicates that, on the contrary, there is no membrane tension and no internal pressure. These are the conditions demanded by the expanding membrane theory of cleavage.

These experiments were done in order to see whether the cleavage furrow could pass through an obstruction placed in its path—in this case, a microneedle. A cleaving egg in ordinary sea water was held between two microloops (about 50 μ diameter, and placed vertically) and a microneedle was then inserted into the egg and through the loops. The needle was inserted off the central (polar) axis of the egg and was either parallel to the central axis, i.e. paraxial (Pl. 14, figs. 13, 14), or parallel to the equator and piercing the bottom of the furrow at two places (Pl. 14, figs. 15-17). In the first case the advancing furrow passed through the needle without any tendency to gape and eventually left the needle transfixing the two blastomeres. In the second case the furrow also passed through the needle but left it free of the egg.

These results do not appear to be consistent with the idea of a simple contracting ring at the bottom of the furrow being responsible for cleavage. If there was such a ring, it is most improbable that it could pass through a paraxial needle without a serious distortion. Its continuity would be broken, and, since it must be under tension, it should gape. A more likely effect with a contracting ring would be that as soon as the ring reached the needle it would tend to displace the egg on the needle so that the final effect would be the needle running along the central axis of the egg with the nearly complete furrow disposed symmetrically round it.

These difficulties do not arise with the expanding membrane theory since the furrow is being pushed inwards by the expansion of its own walls and the rest of the cell membrane, rather than being pulled inwards by a contracting ring. As a result a wound in the furrow would tend to close rather than to gape, and, since the motive force is provided by the expansion of a large area of membrane, a small break in it should have little effect.

Chambers (1938) made the important observation that if the asters in a cleaving egg were destroyed by stirring with a microneedle, cleavage nevertheless continued (though if this was done at an early amphiaster stage, cleavage was inhibited). Since this is a crucial piece of evidence against any astral mechanism of cleavage, it was decided to repeat this experiment.

The same instruments were used in this experiment as in the previous one, i.e. two microloops for holding the egg, and a microneedle. When the needle had been inserted into the egg (in ordinary sea water), it was moved round so that the contents of the egg were stirred up. The asters were initially distorted and then, after about a quarter of a minute, they vanished. If this was done in metaphase or early anaphase the eggs did not cleave, but if it was done when the egg had already started to change its shape preparatory to division, cleavage continued. This is shown in Pl. 15, figs 1-5 : fig. 1 is the egg in the ‘wall-sided’ stage just before the operation; fig. 2 shows the egg being stirred; figs. 3-5 show the egg completing cleavage. Unfortunately, the photographs are not good enough to show the presence or absence of the asters, but in the experiment the asters had vanished in the stirred egg by the time fig. 3 was taken, whereas at this stage in a normal egg the asters could still be seen.

This experiment entirely confirms the results of Chambers and was done with a different sea urchin (Psammechinus miliaris rather than Arbacia punctulata). An independent confirmation of the fact that cleavage can take place without asters has also been provided by a study of the effects of colchicine on sea-urchin eggs (Swann & Mitchison, 1953).

There is one further point about this experiment that is worth making. The expanding membrane theory suggests that although the asters are not essential for cleavage, they do act as passive guides for the advancing furrow. One would therefore expect that the cleavage shapes should become somewhat irregular if the asters were destroyed. This, in fact, occurred in the stirring experiments ; a common feature being the fact that the two sides of the furrow did not lie opposite each other (Pl. 15, fig. 3). An exactly similar appearance was shown by a number of cleaving eggs in which the asters had been destroyed by the action of colchicine (Swann & Mitchison, 1953).

A number of experiments were carried out on the effect of compression on cleaving eggs in order to see how far the results might be consistent with the various theories of cleavage. A certain amount of work has already been published on the effect of compressing sea-urchin eggs. Zeuthen (1951) compressed Psammechinus eggs under a cover-slip until their thickness was about 20 μ. (one-fifth of their original diameter) and found that this inhibited cleavage although nuclear cycles continued. Danielli (1952) obtained similar results, though he states that the eggs would eventually divide if they were not too much compressed. Chambers et al. (1949) also compressed Arbacia eggs under a cover-slip fixed to a microneedle until their diameter was 2-3 times their original diameter and it appears from their paper that the eggs did not cleave. This experiment, however, was carried out in the high K⁺ mixture mentioned earlier (p. 517). The general conclusion from this earlier work seems to be that cleavage is only inhibited if there is large amount of flattening.

The compression experiments were done on cleaving eggs in ordinary and in Ca-free sea water. The two microloops mentioned earlier were used to flatten the eggs, and the flattening was carried out both parallel to and at right angles to the polar axis.

If the eggs were flattened at right angles to the polar axis in ordinary sea water cleavage continued normally (Pl. 15, fig. 6). In Ca-free sea water, however, although cleavage continued it was somewhat delayed, in some cases up to 5 min. This is shown in Pl. 15, figs. 7 and 8. In fig. 7 the ‘wall-sided’ egg on the left is at an earlier stage of cleavage than the egg about to be compressed, while in fig. 8, 5 min. later, it has completed cleavage though the compressed egg has its furrow only half way through. It can be seen from these photographs that the difficulty of compressing eggs in this direction was that as soon as the furrow had formed, the loops tended to slip on to one blastomere. The resulting flattening was usually less than when the eggs were compressed along the polar axis, and the flattening was only effective on one blastomere.

If the eggs were flattened along the polar axis in ordinary sea water, cleavage continued but was delayed up to 5 min. compared to the surrounding eggs. In some cases the furrow reverted back before finally completing the cleavage. This is shown in Pl. 15, figs. 9-13. One side of the furrow has reverted in fig. 11. The other egg in the field started by being at the same stage in cleavage as the compressed egg (fig. 9), but completed cleavage when the compressed egg had its furrow only about one third through (fig. 12).

If the eggs were flattened along the polar axis in Ca-free sea water, cleavage was inhibited in all cases, unless the furrow was nearly complete. The furrow might show several attempts to advance but it finally reverted back. This is shown in Pl. 2, figs. 14-18. In this case the furrow reverted back twice. There was no sign of the coalescence between the two walls of the furrow which has been described by Chambers (1946); the furrow simply moved backwards.

The main conclusion to be drawn from these experiments is a confirmation of earlier work; that eggs continue to cleave when compressed unless the degree of flattening is considerable. Presumably the difference between the results of the flattenings in the two directions is due to the fact mentioned above; that the average degree of compression is less when the eggs are flattened at right angles to the polar axis. There are two further points to be drawn from these results. First, cleavage is delayed by compression before it is finally stopped. Secondly, eggs in Ca-free sea water are more susceptible to compression than eggs in ordinary sea water. Presumably this is due to the hyaline layer in ordinary sea water strengthening the cell membrane by acting as an elastic cover, and so lessening the degree of stretch of the membrane.

These results are probably compatible with all the main theories of cleavage so it is only worth pointing out here that they are quite consistent with the expanding membrane theory. Since the membrane is elastic, flattening the egg will increase the membrane tension, and it will therefore take longer for the tension to fall to zero as the membrane expands. This would account for the delay induced by compression. If the flattening is considerable the capacity for expansion of the membrane will not be sufficient to counteract the stretch of the membrane; the membrane will remain with a positive tension, and cleavage will not be possible. The reversions of the furrow might be due to a number of causes, but the most likely one seems to be stickiness between the microloops and the cell surface. This would allow the tension of the membrane within the loops to be higher than that of the furrow region. The furrow could therefore form, but when the tension in the membrane round it had fallen to a low value, the membrane would suddenly start to slide under the loop towards the poles, the tension would equalize all over, and the furrow would be pulled out.

It might be argued that the fact that eggs are able to cleave when compressed under a cover-glass (e.g. Danielli, 1952) is an argument against the expanding membrane theory. They are under a constant compressing force (unlike the eggs in the experiments above, which are under a constant deformation) and it might be expected that the membrane tension could never fall to zero. There are, however, two objections to this argument. First, the eggs may be supported by the remains of their asters. Secondly, a cover-slip pressing on an egg will only maintain a uniform tension over the membrane if the membrane can slip freely when it is against the glass. If the egg is flattened to any extent this probably does not occur.

It is well known that the cell membrane of sea-urchin eggs is highly elastic and has a considerable capacity to seal itself after damage even in Ca-free sea water (e.g. Chambers, 1924), but it is worth publishing a pair of photographs which show these properties in a striking fashion. Pl. 15, fig. 19, shows an egg which had just cleaved in Ca-free sea water and in which one blastomere had been cytolysed by a slicing cut. The other blastomere had been rested against the microloop and the needle had been carefully pushed through it. The needle had not punctured the membrane but had pushed it ahead so that the membrane finally joined up with itself on the far side of the egg. The needle was then pushed right through, so that the final result was a ring-shaped cell with a hole through the middle in which the needle lay. It was like a doughnut on a stick, and the egg could be rotated freely on the needle.

Pl. 15, fig. 20, shows that the cell membrane can be stretched considerably before it breaks. This was a cleaving egg in Ca-free sea water. When it was at the dumb-bell stage, it was placed against the microloop and pierced by the needle. The photograph shows the needle pushing out the cell membrane of one blastomere from the inside. The membrane had just given way at the tip of the needle when the photograph was taken, but before that it had remained stretched out for more than a minute. This shows that the material of the membrane must have an elastic limit (% stretch at the breaking point) of several hundred per cent.

It is a pleasure to record my thanks to the Director and Staff of the Marine Station, Millport, and of the Stazione Zoologica, Naples, for their help in this work.

All eggs are of Psammechinus miliaris except in Pl. 14, figs. 6-12, which are of Arbacia lixula. OSW = ordinary sea water; CFSW = Ca-free sea water.