On The Physiology of Amoeboid Movement: IV.—The action of magnesium

It was shown previously that whereas cytolysis occurred rapidly in pure isotonic NaCl or KCI (at about pH 7), amoebae remained alive for a longer time in isotonic CaCl₂ and MgCl₂-Movement was inhibited in these solutions, but reversibly, for, provided immersion had been brief, recovery ofigccurred on return to natural sea-water. In mixtures of two salts, it was found that movement only occurred if Ca were present. On the other hand, the amoeba lived almost as well in mixtures of (NaCl + MgCl₂) as in mixtures of (NaCl + CaCl₂), for although no movement took place in the Mg mixture yet within certain limits of concentration movement was ultimately resumed on transference of the amoeba to sea-water, even after some hours immersion.

The same relation was found between solutions of (KCl + MgCl₂) and solutions of (KCl + CaCl₂). But in mixtures containing only (NaCl + KCl) the permeability of the cell seemed to increase, and cytolysis occurred rapidly as in pure NaCl or KCl.

These considerations lead to the conclusion that although Mg cannot replace Ca with respect to the mechanism of movement, yet either Mg or Ca can ensure stability of the cell-surface. Sr and Ba can also do this. This action seems to be connected with a reduction of the permeability of the cell which tends, as is well known, to be brought about not only by Ca but by Mg or indeed any divalent metal (Osterhout, 1922; Lillie, 1923).

If mixtures of NaCl+MgCl₂ or of KCl+MgCl₂ are made alkaline (pH 8 to 8.5), a very feeble movement is occasionally seen for a short time, when the molecular ratio of alkali-metal to magnesium lies between 5 and 10. This might be expected, for although there is no Ca in the medium the Mg will prevent loss of Ca from the cell by lowering the permeability. The increased alkalinity will also tend to produce movement by preventing loss of Ca from the actual contractile mechanism itself, just as it does in solutions of (NaCl + CaCl₂) deficient in Ca (Pantin, 1926).

It might be suggested that if the Mg acts by reducing permeability an increase in alkalinity of the external medium could not affect the alkalinity of the cell-interior. But Mg alone cannot stabilise the cell-surface completely, and the fast OH-ion may be able to penetrate to some extent. A consideration of more importance is that owing to the continuous production of CO₂ in the cell there is a dynamic diffusion gradient of C_H across the cell-membrane. Raising the C_OH of the external medium is thus bound to raise the C_OH just below the cell-membrane, in order to maintain the steepness of the gradient. Because of buffer action it is unlikely that there will be any effect on the protoplasm except immediately below the cell-membrane, at which point the gradient is presumably steepest; but it is just this region that appears to be most intimately concerned with the mechanism of amoeboid movement.

If amoebae are placed in a series of solutions of (NaCl + MgCl₂), it is seen that, apart from the absence of movement, the behaviour in the Mg - deficient solutions resembles that which occurs in mixtures of (NaCl + CaCl₂) where calcium is deficient. It is found that in both cases the adverse changes which lead to death occur successively, commencing in the mixtures where the concentration of the divalent metal is least. Where the molecular ratio is greater than Na/Mg = 500, swelling and cytolysis occur as rapidly as in pure NaCl. From solutions containing Na/Mg= 128 down-wards the increasing Mg concentration begins to have a definite stabilising influence on the cell-surface. The effect increases more and more rapidly as the Mg concentration is raised till Na/Mg = 10. The stability is indicated by the longer time the amoeba remains alive in the solution and the inhibition of the swelling which occurs in pure NaCl. Osterhout (1922) has shown parallel changes in the viability of Laminaria in the presence of increasing concentrations of divalent metals. In low concentrations of Mg the amoeba assumes an irregular spherical form : as the concentration increases the amoeba tends to assume a “proteus” form (fig. 1) and appears more normal.

The fact that these changes are closely similar to those occurring in mixtures of (NaCl + CaCl₂), although in the Ca mixtures they are accompanied by movement, supports the suggestion made-in the previous paper (Pantin, 1926) that the reduction of movement which occurs in Ca-deficiency is directly related to an increase in permeability, probably accompanied by loss of Ca from the cell: this inhibits movement because Ca is specifically necessary for the mechanism to function.

The similarity of the action of Mg and Ca ceases when the molecular ratio falls below Na/Mg = 10; In the case of Ca, an Na/Ca ratio below this value rapidly inhibits movement and appears to produce a characteristic gelation of the ectoplasm. But in solutions of NaCl + MgCl₂ the “proteus “condition seen in lower Mg concentrations is maintained, to some extent, right up to (Na/Mg = 1.5) and even, for a short time, in pure MgCl₂. The viscosity of the protoplasm does not seem to increase enormously as it does in excess of Ca ; but this is difficult to judge because of the absence of movement. It may be noted that in experiments conducted by Professor Chambers and the writer on these amoebae the cell-surface was torn in excess of Ca and of Mg and the medium freely admitted to the protoplasm : in both cases coagulation followed. But whereas the coagulum formed with Ca was exceedingly tough, that formed with Mg was much looser and “slimy.”

In high concentrations of Mg the “proteus” condition becomes enormously exaggerated. In such a solution an amoeba may become reduced to a small central mass of protoplasm with a dozen or more irregular pseudopodia radiating from it (fig. 1), in which condition it remains. The significance of this is not obvious, though it seems to indicate that the excitor mechanism becomes inco-ordinated.

Like Mg, Ba is able to maintain the stability of the cell-surface but cannot support amoeboid movement. The range over which stabilisation occurs is approximately the same as it is with Ca and Mg. Yet the action of Ba seems to be more closely allied to that of Ca and Sr than to Mg, for like Ca excess of Ba markedly increases the viscosity of the ectoplasm. In solutions containing a large proportion of Ba (Na/Ba = 10 or less) the amoeba assumes a form very like that which occurs in excess Ca and quite unlike the Mg “proteus” form (fig. 1). Very occasionally feeble activity was seen for a short time in solutions of about the composition Na/Ba = 20, and in these cases the amoebae approached the unipodal limax type seen in Ca.

It is possible that Ba not only stabilises the cell-surface but also acts, like Ca, directly on the contractile mechanism itself; but although it acts at the same site as Ca the resulting system does not sufficiently approach the normal conditions to allow movement.

Although Ba is definitely toxic to amoeba, yet inhibition in Ba is for a short time reversible where Na/Ba>10.

Mines (1912) suggested that the action of Mg on various contractile tissues was essentially connected with the fact that it is a divalent cation. He showed that, as one might expect, the far more powerful trivalent cations are able to perform the function of Mg when they were present in very low concentration. Osterhout (1922) and Gray (1916) have also shown that, like Mg, Ce can reduce the permeability of the cell and is effective in great dilutions. As in solutions of NaCl + MgCl₂, amoebae did not move in a solution of isotonic NaCl in the presence of any concentration of CeCl₈ The solutions were rigidly buffered at pH 6.8 to 7.6 in different experiments, so this effect is not due to the acidity consequent on hydrolysis of CeCl₈ In a narrow range of low concentration (10⁻⁵ M), cerium is able to maintain the stability of the cell. Moreover, there is some approach to the “proteus “form seen in the presence of Mg (fig. I). At concentrations below this (e.g. CeCl₈ 10⁻⁶ M) is seen the typical swelling and cytolysis which occurs in pure NaCl.

Raising the concentration of CeCl₈ causes rapid coagulation of the ectoplasm of many of the amoebae, and at a concentration of 10⁻⁴ M almost all are rapidly killed and coagulated. The ectoplasm often ruptures at the posterior end of the amoeba and the endoplasm disperses into the medium where it coagulates, leaving a hollow “glove” of coagulated ectoplasm. It is significant that cerium is only able to maintain the stability of the cell at a concentration immediately below that at which irreversible coagulation of the ectoplasm occurs.

Although the action of Ce resembles that of Mg the resemblance is incomplete. Cerium is far less effective than Mg in maintaining the normal condition of the cell. However carefully the concentration was adjusted it was found impossible to bring about a condition similar to that found in optimum mixtures of NaCl + MgCl₂. Again, a slight excess of cerium causes complete coagulation and the coagulum is tough, whereas even in pure isotonic MgCl₂ no coagulation takes place for a long time, and the coagulum which is formed even when Mg is allowed to penetrate the cell is decidedly not tough, but loose and slimy.

Perhaps the most remarkable observation made throughout the whole series of experiments was the extraordinarily complete antagonism found to exist between Mg and Ca. Cases of antagonism between Mg and Ca are not rare (e.g. Osterhout, 1922 ; Loeb, 1906a; Hogben, 1925), but the effect is usually small and only to be seen when large amounts of other ions are present. In these amoebae, Mg and Ca exert a complete antagonism when no other cation is present. Good movement takes place in the presence of (MgCl₂ + CaCl₂) alone, and the range of concentrations of Ca over which the effect is observed is enormously greater even than the range over which mixtures of NaCl + CaCl₂ permit movement (fig. 2). Almost the only comparable case among contractile mechanisms which appears to approach the condition seen here is that recorded by Loeb (1906b) in the cilia of certain sea-urchin larvæ.

These experiments lend the strongest support to the idea that Mg acts essentially by reducing the permeability of the cell. In view of the marked inhibiting effect of an excess of Ca on the one hand, and of its deficiency on the other, in the absence of Mg, there seems to be no alternative but to suppose that Mg prevents the entrance of Ca into the cell when Ca is present in excess, and prevents its exit when Ca is deficient in the surrounding medium. This implies that the point at which Ca acts upon the mechanism of amoeboid movement is not at the very surface of the cell, but that it lies beneath the surface film which governs permeability.

We again have evidence in these experiments that the cessation of amoeboid movement in mixtures of (NaCl + CaCl₂) where Ca is in excess is not due to the action of Ca as a divalent cation increasing impermeability, but is the result of its direct specific action on the contractile mechanism. If this inhibition resulted from decreased permeability, movement could not take place in (MgCl₂ + CaC1₂), for both cations apparently reduce permeability. It is probably the very impermeability induced by the Mg which, by preventing entrance of excess Ca, allows movement to continue.

The enormous range of antagonism of Mg and Ca is only maintained for a few hours, and even in the central part of the range movement is much diminished after some sixteen hours (fig. 2). It follows that the monovalent cations of sea-water are necessary for completely normal conditions of the cell.

In view of the preceding experiments it is not surprising that Mg exerts a marked influence on movement in the presence of Na and Ca. When a relatively small amount of Mg is added to mixtures of (NaCl + CaC1₂) (fig. 3) the effect is to increase slightly the range of concentrations of Ca over which movement can take place, especially for those weaker in Ca. The maximum velocity is also increased-slightly and movement is maintained for a longer period of time. These effects are greatly increased when the Mg concentration is raised to that of the Mg in sea-water (fig. 3, curve 2). The increase in the maximum velocity is in particular enormous, and it now corresponds to a ratio Na/Ca ≑ 45 (instead of 20, as in the absence of Mg), which is the same as that of sea-water. This shows that in the presence of Mg less Ca is required for optimum movement.

On raising the concentration of Mg still further the range of Ca concentrations over which movement takes place increases for the higher Ca concentrations. For the lower Ca concentrations the range tends to end abruptly (fig. 3, curve 1). The behaviour approaches that found in simple mixtures of (MgCl₂ +CaCl₂) after the lapse of some hours, when the range has contracted. In fig. 7 the absolute concentration of calcium is not the same for corresponding values of the Na/Ca ratio of the three curves. The added Mg necessarily dilutes the mixture so that an Na/Ca ratio of 32 in curves (2) and (3) corresponds in absolute Ca concentration to an Na/Ca ratio of about 16 in curve (1).

The sudden drop in curve (3) where a large amount of Mg is present is due to the fact that although the velocity of individual amoebae tends to increase slightly as the Ca concentration falls, yet the number of amoebae actually moving in each dish becomes rapidly smaller. The result is that in one dish a very few amoebae may be moving well, with a consequent high average velocity, while in the next dish none may be moving at all. Had we been dealing with a tissue such as muscle we could only have observed the statistical effect of all the cells, and the resulting curve would have been quite different in character.

In the presence of K as well as Na and Ca, the effects of adding Mg are still more marked but otherwise similar to those described above. Fig. 4 shows the effect of adding Mg to solutions containing Na, K, and Ca, where the ratio Na/K is the same as in sea-water. When all four ions are present, a considerable degree of acclimatisation of the amoeba can take place towards the solution, so that where the proportion of the salts differs moderately from that of sea-water the velocity may actually increase for the first twenty-four hours.

Viability is very great in solutions which approach the composition of sea-water. When the salts are actually in the same proportions as in sea-water, the velocity of the amoeba is the same as in the natural medium within the limits of variation, and the amoebae will live almost as long as in clean natural sea-water without food.

We have just seen that the addition of Mg produces a profound improvement in the movement of amoebae in mixtures of (NaCl+CaCl₂). If cerium is added to mixtures of Na and Ca, there is perhaps a slight, but certainly not a marked, improvement, however carefully the Ce concentration is adjusted. Here again cerium proves to be an imperfect substitute for magnesium.

We have seen that Ca in small quantities is specifically necessary for amoeboid movement, and only Sr can replace it. But movement is only maintained for long if the cell-surface be stabilised. The addition of divalent metals such as Mg or excess Ca will do this, though Ca directly inhibits the actual mechanism of movement before the concentration is reached at which the cell-surface approaches complete stability. Stabilisation is probably connected with reduction of permeability, which prevents both penetration of the medium into the cell and loss from the cell of certain necessary substances, particularly calcium.

However, the cell is not completely stabilised when a single divalent metal is added to a solution of NaCl. Viability is greatly prolonged if Mg is added to a solution of NaCl which already contains CaCl_2, and still further if a small amount of K is added. Only when all four cations are present in the proportions of sea-water does the cell remain normal indefinitely.

The action of K in increasing viability is specific and only occurs near the Na/K ratio of sea-water. The action in amoeba is relatively small and is not obviously related to the varied cases of a specific action of K in many tissues. But the outstanding features both in these cases and in amoeboid movement are that only a small amount of K is required to produce its effect, and that this action of K is shared only by Rb and, to a less extent, by Cs. Probably K affects similarly some mechanism common to all cells, such as the cell-surface: whether the observed reaction of the cell relates mainly to viability, as it does in amoeba, and in developing eggs (Herbst, 1901; Loeb, 1921), or to the marked effects seen in contractile tissues (Clark, 1921 ; Hogben, 1925), probably depends on secondary changes peculiar to each type of cell following initial derangement of this mechanism.

Zwaardemaker (1919) suggests that this action of K is due to its radioactivity. But Clark (1921) shows this to be unlikely; in particular, there is ample K within the cell to provide any necessary radiation, even though K be removed from the external medium. It is more probable that the specific action of K is related to the high K concentration frequently found within cells (Bayliss, 1924). And it is noteworthy that Mitchell and Wilson (1921) have shown that Rb and Cs can partially replace the K normally present in muscle cells: these are just the two elements which replace K in its “specific” functions. Possibly the action of K is related to the maintenance of a potential difference across some surface, as Mines (1912) suggested, especially since, as Loeb (1900 b) pointed out, the ionic velocities of K, Rb, and Cs are almost equal to one another, but considerably greater than those of Li or Na. Such a potential difference might well be necessary for the maintenance of the normal permeability of the cell.

Both Mg and K seem to act on the cell-surface, but do not seem to affect amoeboid movement solely by controlling the permeability conditions of the cell. Their addition to a solution of (NaCl + CaCl₂) not only affects viability but also greatly increases the absolute velocity of movement. Perhaps we have here a single molecular complex in the cell-surface, with two distinct functions. It may control the permeability of the cell : this will affect amoeboid movement indirectly by preventing loss of the normal cell-constituents. But this same complex may also be a part of the whole mechanism of amoeboid movement itself, and the relation here is direct. The necessity of Ca for movement involves a separate complex from the one just suggested, but it also enters into the complete mechanism of amoeboid movement.

The cell-surface is evidently an important part of the mechanism of movement, for direct observation of limax amoebae indicates that movement depends on the formation of gelated ectoplasm from fluid endoplasm at and immediately below this surface (Pantin, 1923). But the point at which Ca is specifically required for movement appears to lie below the actual cell surface: for when the permeability of the surface is reduced by excess of Mg variation of the Ca concentration of the medium scarcely affects, the impermeability of the surface retarding gain or loss of Ca to or from the mechanism. The Ca is probably required actually in the ectoplasm itself in order that this may function normally. The direct action of excess Ca on the viscosity of the ectoplasm confirms this.

Clark (1913) has suggested that in the frog’s heart Ca combines with some lipoid substance to form a soap essential for the contractile mechanism. Since the action of Ca on amoeba resembles its action on the heart, it may well be that some Ca-soap in the ectoplasm is necessary for normal contractility. The antagonism of the alkali metals and Ca points in the same direction, for monovalent and divalent soaps are markedly antagonistic (Clowes, 1916). Moreover, the properties of alkali metal soaps show a regular gradation with the atomic weight (Harkins, 1924).

In an earlier paper (Pantin, 1923) certain evidence led to the idea that a sensitive lipoid surface might control amoeboid movement. It was suggested that the surface of the cell was the one in question, but the experiments described above point to some interface actually within the ectoplasm. Owing to the continual interchange of ectoplasm and endoplasm in amoeboid movement this interface cannot be a structure differentiated from the rest of the protoplasm. But we have seen that other factors besides Ca affect movement, so such a Ca-lipoid surface cannot comprise the whole mechanism. These other factors are related to the action of Mg and K on the cell-surface. Proteins probably play an important part here. The tendency to form membranes by concentrating at surfaces is seen both in proteins and protoplasm. The fact that cerium, whose action somewhat resembles that of Mg (Mines, 1913) appears in amoeba only to support the cell membrane in concentrations where there is incipient coagulation of the ectoplasm points in the same direction. Moreover, the ready changes of state, sol⇆gel, of the ectoplasm during amoeboid movement are characteristic of proteins rather than lipoids.

These considerations indicate that both lipoids and proteins are involved in the protoplasmic structure on which amoeboid movement depends, especially near the cell surface. Now Loeb (1922) points out that when a protein passes towards the gel state the protein molecules become held together by the affinity of the aliphatic, water-insoluble portions of the molecules, and the ionised terminal groups project into the surrounding watery medium. If lipoids are present it is reasonable to suppose that just as two protein molecules are bound together by their aliphatic parts, so will the aliphatic end of the lipoid molecule tend to bind it to the protein molecule. In view of the work of Langmuir (1922) we may be justified in considering the gelated surface layers of protoplasm to consist of a network of protein molecules resulting from the mutual attraction of their aliphatic radicles on to which is adsorbed a monomolecular layer of lipoid molecules. The same conditions will obtain, though to a less extent, throughout the protoplasm.

The advantages of this conception are great. We are at once provided with two distinct sites at which cations and other substances can act on the protoplasm, the terminal (carboxyl) groups of the proteins on the one hand, and of lipoids on the other. At the same time the two parts of the complete mechanism are intimately connected, as they appear to be in amoeboid movement. The action of ions on the lipoid site may well depend on calcium-soap formation in the manner suggested earlier.

Cations will also affect the system by combining to form proteinates (Loeb, 1922). There is reason to suppose that a large proportion of undissociated metal proteinate may be formed, as has been shown to occur in gelatine by Northrop and Kunitz (1924). It may be well to point out that even the highly ionised NaCl molecules behave as though 25 per cent. were undissociated, when isotonic with sea-water (Jones, 1912).

The formation of undissociated proteinates will profoundly affect the cell. The properties of protoplasm prominent in amoeboid movement (e.g. swelling, etc.) are seen in pure proteins in the neighbourhood of the isoelectric point : yet this point is usually far more acid than the cell interior (cf. Needham and Needham, 1925). It is probably significant that Northrop and Kunitz (1924) have shown that undissociated metal proteinates show effects typical of the Donnan equilibrium as the concentration of the metallic ion is varied, quite apart from the effect of the H-ion.

If this hypothesis resembles the truth we may expect that the primary action of a certain cation will depend upon its valency electrons. This is practically Mines’ hypothesis that certain ions act by their electric charge. But the action will not depend solely upon the valency but also upon the successive dissociation constants corresponding to each valency electron (Lewis; 1923). It 1s hoped to discuss these ideas fully elsewhere.

It therefore seems that in amoeboid movement we must assume the essential mechanism to be more complex than a simple lipoid surface, and it is probably significant that in muscle also Hill (1925) has shown that such a surface cannot alone account for muscular contraction.

It is very remarkable that the physiology of amoeboid movement should show such a strong general resemblance to the physiology of highly differentiated contractile tissues. It indicates that the properties of the latter depend far more on the general nature of undifferentiated protoplasm than upon the highly specialised structure which they have evolved.

1. Although movement only occurs if Ca or Sr is present in the medium, yet Mg and Ba as well are able to prevent the increased permeability and cytolysis seen in pure NaCl. Cerium also has a similar action at very low concentrations, but it is very much less effective than Mg.

2. Excess of Mg never causes the marked increase in viscosity seen in the ectoplasm when Ca is in excess. For this and other reasons it seems that inhibition of movement in excess Ca is due to direct action on the contractile mechanism, and is not simply the result of decreased permeability.

3. This is borne out by the fact that good movement occurs in mixtures of (MgCl₂ + CaCl₂) alone, and over a far greater range of concentrations than occurs in mixtures of (NaCl + CaCl₂). This can be readily explained by assuming that Mg reduces the permeability of the cell so far that Ca can neither penetrate nor leave the cell and thereby derange the contractile mechanism.

4. Even in any mixture of (NaCl + CaCl₂), or of (MgCl₂ + CaCl₂), movement gradually falls off. Only when all four cations of sea-water are present is movement normal indefinitely: fully normal permeability is maintained only under these circumstances.

5. Since the addition of Mg and K to a solution of (NaCl+CaCl₂) not only establishes the normal degree of impermeability but also enormously increases the absolute velocity of movement, it seems probable that the same mechanism which controls permeability is also a part of the whole mechanism of amoeboid movement.

6. The relation of the action of ions to the chemical structure of protoplasm is discussed.