ABSTRACT
The normal healthy eggs of the trout are impermeable both to water and to intracellular electrolytes. The barrier to diffusion of both water and salts lies in the vitelline membrane, whose static properties maintain the normal hypertonicity of the yolk without expenditure of energy.
When eggs are exposed to conditions which destroy the normal impermeability of the vitelline membrane, exosmosis occurs suddenly at the end of a latent period, and once it has begun exosmosis is usually rapidly completed. Exosmosis is a sign of an unhealthy cell.
The osmotic stability of the vitelline membrane varies with four factors: (a) the age of the eggs, (b) the degree of mechanical disturbance caused by the environment, (c) the presence or absence of calciuip in the environment, (d) the nature and concentration of the exosmotic reagent used.
Loss of the impermeable properties of the vitelline membrane is accompanied by a drastic change in the distribution of the material of the membrane and in the organisation of the egg cell.
The egg cells of animals appear to fall into two groups in respect to the equilibrium between themselves and their external osmotic environment. The eggs of marine invertebrates contain a concentration of osmotic substances approximately equal to that of their surrounding environment ; if the concentration of this medium be altered, water enters or leaves the cells, whose surfaces appear to act as semipermeable membranes—being permeable to water but not to dissolved solutes. The eggs of fresh-water organisms and of birds are, however, different. In such eggs the concentration of the osmotically active substances inside the cell is very much higher than that found in the external medium, and this difference is maintained for very long periods of time (see Gray, 1920). The problems presented by cells of this type have recently been re-investigated by Straub (1929) and by Needham (1931), both of whom used the unfertilised eggs of the hen. In the newly laid hen’s egg the yolk, with a freezing-point of — 0.6° C., is separated by a relatively thin vitelline membrane, from the egg-white, which has a freezingpoint of — 0.45° C. : there is thus a difference of about two atmospheres between the osmotic pressures of the yolk and of the egg-white. Straub concluded that this asymmetrical distribution of osmotically active substances was maintained by virtue of the powers possessed by the vitelline membrane to do active osmotic work. When normal egg yolks were placed in diluted egg-white (Δ = — 0.21° C.) the freezing-point of the yolks rose from — 0.58° to — 0-38° C.; when eggs subjected to such treatment were transferred from the diluted egg albumen to normal albumen (A = 0.46° C.) the freezing-point of the yolks fell from — 0.38° to — 0.51° C., a value considerably below that of the external medium to which the yolks had been exposed. If, in every case, the osmotic substances were uniformly distributed throughout the yolk, the egg must be capable of concentrating these substances from a diluted environment and must expend energy in the process. The validity of Straub’s conclusion has, however, been questioned by Smith and Shepherd (1931), who found that normal hen’s eggs immersed in water or in diluted egg-white exhibit marked heterogeneity in respect to the distribution of osmotic substances. These authors have shown that under such circumstances the peripheral regions of the yolk have a much lower osmotic pressure than that which exists at the centre of the egg. On transfer to normal egg-white, the peripheral regions of the “diluted” yolk are hypotonic to the external environment, whereas the core of the egg is still hypertonic to normal egg-white. Thus in normal egg-white the average osmotic pressure of such yolks can rise to a value above that of the external medium, by a process of equilibration which is largely if not entirely confined to the periphery of the yolk. Although there may not be a sharp boundary between the diluted periphery and the more concentrated core, the following analysis of Straub’s experiment is of interest in view of the facts to be considered later in this paper. Let the eggs be immersed in water of freezing-point — 0.21° C., and left until a peripheral volume of yolk equal to 0.54 of the whole egg has a freezing-point equal to the external medium whilst the central core of the egg remains at its normal value of 0’58° C.—then the average freezing-point of the whole yolk will be — 0.38° C. ; if such eggs are now transferred to a medium having a freezing-point of — 0-46° C. and left there until the peripheral region has adjusted its freezing-point to this value by diffusion—then the average freezingpoint of the yolks will rise to — 0.516° C. In order that the peripheral zone of the yolk should occupy half the volume of the whole egg, the junction between the central core and the diluted periphery must lie on a surface which is o-i4r below the surface of the egg (r being the radius of the whole egg). Even if there is no sharp boundary between the periphery and the core of the egg, the time required for a uniform distribution of the electrolytes throughout the whole egg would be very great in view of the size of the egg: the freezing-point of the yolk would tend to become uniform at a rate which is dependent on the concentration gradient across the vitelline membrane and not on the difference of the average concentrations of the whole yolk and of the outside medium. If Smith and Shepherd’s observations account adequately for the marked increase in the average concentration of yolk transferred from diluted to normal egg-white, the main basis of Straub’s suggestion is no longer tenable. Needham (1931) concludes that the maintenance of the steady state of a normal hen’s egg “is not determined by the vitelline membrane, nor the yolk, nor the white alone, but is produced by some collaboration of the three phases. It may be due (a) to the performance of thermodynamic work at the membrane, the glycolytic mechanism of the yolk providing the energy and the membrane utilising it, or (b) to the possession of osmotic properties by the vitelline membrane which are instantly lost when it is removed from the egg and brought into contact with salt solutions, or (c) to some feature of the physical structure of the yolk and white which retards the attainment of equilibrium and which is not wholly destroyed by artificial mixing.” Needham concludes that “the second of these possibilities is unlikely. Grave objections have been raised to the first, and there is as yet no positive evidence in favour of the third” (Needham, I931, P-343).
The present paper deals with the eggs of the trout (S.fario) which are analogous to those of the hen in that the osmotic concentration (Δ = – 0.48° C.) of the yolk is much higher than that of the surrounding medium of tap water (Δ = – 0.02° C.). As in the hen’s egg, the difference between the osmotic pressure of the yolk and its medium is maintained in the unfertilised egg for very long periods of time, although in each case the oxygen consumption of the eggs is too small to be measured by ordinary methods. Before presenting the data on which can be based a relatively simple conception of the manner in which the yolk of the trout maintains its hypertonicity to the external environment, it is perhaps useful to consider those circumstances under which it would appear to be necessary to postulate an active process concerned with the maintenance of the normal osmotic equilibrium. If osmotic substances are free to move within the yolk, and are not free to move across the normal cell membrane, and if both yolk and membrane are permeable to water, then water must enter the cell from a hypotonic medium unless (a) the membrane can exert a sufficiently great hydrostatic pressure in an opposite direction, or (b) there exists an active process which prevents water from entering the cell by an expenditure of energy. If, on the other hand, the membrane is permeable to osmotic substances as well as to water, a steady state of hypertonicity of the yolk can only be maintained by an active process involving an expenditure of energy, and under such circumstances the egg cell would swell unless there is an opposing hydrostatic pressure or an active excretion of water. It is obvious that the distribution of osmotic substances and the distribution of water must be considered together.
There is, however, one possibility which appears to have been overlooked, namely, how far a normal egg cell has been shown to be permeable either to salts or to water. Clearly, if any one of the cell membranes were impermeable to both salts and water, the maintenance of a high state of internal hypertonicity presents no difficulty. In the present paper data will be given to show that neither water nor salts pass across the inner membrane of the normal trout’s egg in measurabh quantities, and that it is not impossible that this state of impermeability also characterises the normal healthy egg of the hen.
I. THE STRUCTURE AND OSMOTIC PROPERTIES OF THE NORMAL EGG
Structure of the unfertilised egg
The eggs of the brown trout (Salmo fario) are approximately 4-5 mm. in diameter. Each egg consists of a very thin membrane—only a few microns in thickness—which encloses a large central vacuole filled with fluid yolk. This surface membrane is thickened at one point to form the blastodisc, and throughout its whole extent are found oil drops of varying sizes firmly fixed in position ; the oil drops are larger and more numerous in the blastodisc than elsewhere. The vitelline membrane together with the yolk constitutes the egg proper, which is itself enclosed within an elastic but tough “chorion”1. It is significant to note that the egg is not attached to the chorion at any point, but is separated from it by a perivitelline space filled with an aqueous medium. Owing to the high specific gravity of the yolk and the accumulation of oil drops within the blastodisc, the egg is in equilibrium with gravity when it is resting on the bottom of the “chorion” with the blastodisc at its upper pole (see Fig. I A). The existence of the perivitelline space can be readily demonstrated by viewing the egg in a horizontal plane, through a horizontal microscope, although, when the egg is in gravitational equilibrium, the perivitelline space can only be recognised at the upper pole of the egg.
In view of the comparison to be made between the osmotic properties of the trout’s egg and that of the hen, it is perhaps well to stress the fact that in the former case the yolk is separated by a membrane from the fluid in the perivitelline space, and that this membrane is continuous with the blastodisc, since both contain enclosed oil globules which are absent from the yolk itself. In physical properties, the “vitelline membrane “of the hen’s egg resembles the “chorion “of the trout. They may be homologous—since both are of ovarian origin. Needham is inclined to deny the existence of a membrane between the yolk and the vitelline membrane of the hen’s egg (Needham and Smith, 1931, p. 290), but the evidence is not altogether conclusive. It is therefore important to realise that the ‘‘chorion “and the ‘‘vitelline membrane” in the trout’s egg can be clearly recognised as separate structures.
As shed by the fish, the eggs are translucent and retain their normal condition for many days and in some cases for many weeks. Eventually the eggs die and become white and opaque. As shown some years ago (Gray, 1920), the loss of translucency is due to the exosmosis of electrolytes from the yolk to the surrounding medium of tap water; owing to the loss of salts the abundant globulins in the yolk are precipitated; at all times the chorion remains impermeable to globulin.
The present investigations are primarily concerned with the nature of the membrane which separates the yolk from the perivitelline space, but in order to make a comparison of the egg with other types of cells, it is convenient first to consider the nature of the yolk itself.
Composition and properties of the yolk
By puncture of the eggs with a sharp pipette it is possible to obtain samples of the yolk in considerable quantities. When obtained in this way, the yolk is pale yellow in colour and perfectly transparent: it is free from oil drops unless these have been displaced from the egg membrane during manipulation. The yolk shows no signs of internal structure even under the highest magnifications. Its viscosity is 15.20 times that of water, and its freezing-point is about — 0.48° C. (Gray, 1920). If the yolk is dialysed against distilled water across a parchment membrane, electrolytes rapidly diffuse outward and to an extent which is equivalent (in units of electrical conductivity) to 7.3 gm. of sodium chloride per litre of yolk. This is equivalent to a 0.125 M solution of sodium chloride with a freezing-point of — 0.47° C., from which it may be inferred that practically the whole of the osmotically active substances present in the yolk consist of freely diffusible electrolytes.
As the normal yolk appears to be optically homogeneous and has a relatively high electrical conductivity, it hardly seems possible that it can, by itself, exercise any restraint upon the diffusion of intracellular ions apart from its viscosity.
The nature of the “chorion “of the egg
As already stated, the yolk is separated from the external environment by (a) the vitelline membrane, (ó) the perivitelline space, (c) the “chorion” of the egg. The latter structure is about 90 µ1 thick and is of a fibrous nature (Bogucki, 1930). In the newly laid egg, the chorion is a relatively delicate structure, but, after immersion in water for some hours, it hardens and becomes tough and elastic, so that extremely high pressures can be applied to the eggs without rupturing the chorion. In the case of fresh eggs, the chorion will withstand pressures of seven to eight atmospheres, although at such pressures the yolk is forced through the chorion. Eggs which have remained for some weeks in the incubator rupture at much lower pressures.
An attempt to measure the tensile strength of this membrane was made as follows. A glass plate AB (Fig. 2) was hinged firmly but lightly at one end A to a slab of smooth hard wood W; to the end B was attached a balance scale (D). An egg (E) was placed between the glass plate and the wood, and weights were added to the scale pan until the egg ruptured. It was found that all fresh eggs readily withstood the strain exerted by the weight of I kg., whilst the breaking-point was usually reached between 1.5 and 2.0 kg.
Prior to breaking, the area of contact between the egg and the glass was about 30 sq. nun. which makes a breaking pressure of 2.3 kg. equivalent to a force of 7.7 kg. per sq. cm., or about 7.8 atmospheres.
Although the tensile strength of the chorion of a healthy egg could balance the difference in osmotic pressure between the yolk and the environment, there can be little doubt that such a force is, in fact, not generated in fife. Firstly, the osmotic pressures of the yolks of freshly laid eggs and of aged eggs are approximately the same as that of normal eggs and yet none of these types of eggs swell up and burst in tap water. Secondly, the perivitelline space between the egg surface and the chorion is filled with liquid of the same osmotic pressure as the external medium (see Svetlov, 1929, and p. 283 of this paper). It will later be shown that the chorion of the normal egg is freely permeable to water and to electrolytes, and it can therefore be regarded as an inert structure playing no essential rôle in the maintenance of the osmotic equilibrium of the egg itself.
Normal impermeability of the egg surface to intracellular electrolytes
The impermeability of the egg surface to the salts in the yolk has already been described (Gray, 1920). In view of the great importance of this fact, however, an additional series of experiments was performed to determine whether or not a very slow state of exosmosis characterises the normal egg of the trout as appears to be the case in the hen’s egg (Smith and Shepherd, 1931). For this purpose, individual eggs were transferred to 3 c.c. of tap water in an aerated conductivity cell. After 6 hours no loss of electrolytes had taken place from the eggs. Even when the external medium was replaced by distilled water, no exosmosis occurred except in those cases where the egg subsequently showed signs of opacity. Many experiments of this type were performed, and their negative results show quite clearly that an egg in its normal state of translucency shows no measurable loss of electrolytes until a brief period before its translucency is replaced by partial or total opacity. Even when an egg is exposed to an exosmotic reagent of reasonable strength (see Fig. 3) there is nearly always a well-defined latent period during which no exosmosis occurs ; as soon as exosmosis starts, however, it reaches its maximum rate within a short time—a fact which is in harmony with the observations of Brown (1905) on the eggs of Fundulus. It may therefore be concluded that a normal egg is impermeable to the salts which are contained in the yolk, whilst the observations of Pumphrey (1931) show that it is essentially an impermeability to anions which is responsible for this effect.
This impermeability of the egg surface to osmotic substances in the yolk is fully confirmed by the observations of Svetlov (1929), who found no diminution in the freezing-point of the yolk after the eggs were exposed for a period of 50 days to stream water. That the barrier to the diffusion of salts is situated in the vitelline membrane and not in the chorion is shown by Svetlov’s observation that the freezing-point of the normal perivitelline fluid is the same as that of the external medium; if the osmotic pressure of the outside medium is changed, the osmotic pressure of the whole egg (egg + perivitelline fluid) changes by an amount which is exactly equivalent to a replacement of the perivitelline fluid by external solution, with no change in the osmotic pressure of the yolk itself.
These conclusions can be very readily confirmed qualitatively by immersing the eggs in a solution of electrolytes of varying strength and specific gravity, and viewing the egg through a horizontal microscope. As long as the egg is surrounded by an electrolyte solution whose specific gravity is less than that of the egg itself, the latter rests on the bottom of the container and the ovum is orientated in its normal manner. As soon as the egg is immersed in a Ringer solution whose specific gravity is greater than that of the egg (about 1-09), the latter floats to the surface and the ovum now rests with the blastodisc in contact with the upper surface of the chorion and remains there for a very prolonged period (see Fig. 1 B). In other words, the fluid in the perivitelline space adjusts its concentration of electrolytes very quickly to the same level as that in the external medium. The evidence seems to point very definitely to the conclusion that the vitelline membrane of the normal egg is highly impermeable to the osmotic substances in the yolk.
The osmotic properties of the chorion and of the egg membrane of the trout are well illustrated during an experiment analogous to that of Straub with the hen’s egg. Svetlov (1929) found that in a normal environment of stream water the average1 freezing-point of trout eggs was — 0.455° C., on transferring the eggs to 0.12 M sodium chloride solution, the average freezing-point of the eggs fell to — 0.485° C., although the freezing-point of the external solution was only — 0.39° C. Clearly this result is analogous to that obtained by Straub. In the case of the trout, however, there can be no doubt that the explanation is to be found on the lines indicated by Smith and Shepherd, since Svetlov showed quite conclusively that the average fall of the freezing-point of the eggs (in the sodium chloride solution) is entirely due to a passive replacement of the water in the perivitelline space by the sodium chloride solution.
The impermeability of the normal egg to water
When trout eggs are immersed in comparatively highly concentrated solutions of electrolytes no visible change occurs in the size of the perivitelline space. This fact in conjunction with the impermeability of the ovum to electrolytes suggests that the surface of the egg is also impermeable to water. This conclusion was confirmed by the fact that no measurable decrease in weight occurs when healthy eggs are immersed for 6 hours in any balanced solution of electrolytes up to and including those which are 32 times as concentrated as the intracellular electrolytes (see Fig. 4). In the stronger of these solutions the eggs float on the surface. Eventually they sink to the bottom, but when this occurs the eggs have invariably undergone a process of irreversible increase in permeability (see p. 292 seq.). Eggs which float remain translucent when transferred to tap water, eggs which have sunk invariably turn opaque. Most eggs will not survive for more than 8 hours in Ringer solution which is more than eight times as concentrated as the intracellular electrolytes. It is, however, possible to expose eggs for longer periods to solutions which are not so concentrated but which are at the same time very strongly hypertonic to the normal external environment (= 3.3 per cent. Ringer) of the eggs. Table I shows the percentage loss of water after 24 hours’ immersion in Ringer solutions.
The external environment can be concentrated one hundredfold and the weight of the egg changes by less than 1 per cent, in 24 hours. The very slight changes in weight which were observed in the live eggs may have been due to changes in the size of the perivitelline space, but they are too small for verification by visual observation. The fact that the weight of the egg does not fall in hypertonic solution is, by itself, no evidence that the vitelline membrane is impermeable to water, for the weight of the whole egg includes the weight of the yolk and that of the perivitelline fluid. No detectable increase occurs in the size of the vitelline space and no change occurs in the freezing-point of the yolk (Svetlov), and we may therefore conclude that the vitelline membrane is highly impermeable to water as well as to salts. In these properties the egg of the trout closely resembles the fertilised eggs of Fundulus which were investigated by Loeb (1912), and differs fundamentally from those of sea-urchins and other marine invertebrates. In the case of the trout, there is no need to postulate an active mechanism for maintaining internal hypertonicity as has been suggested for the egg of the hen ; the trout’s egg is apparently completely insulated both in respect to salts and to water—although an interchange of cations may occur with those in the environment.
The whole of the available data suggest that under normal conditions the surface membrane of the unfertilised egg of the trout is impermeable to water and impermeable to intracellular electrolytes and proteins. For this reason the egg is in osmotic isolation from its environment and no energy need be expended for keeping salts inside the cell or for preventing water from entering the cell. As already stated, the only parallel appears to be offered by the eggs of Fundulus which Loeb (1912) declared to be impermeable to salts and water as long as they were immersed in a physiologically balanced solution. In the case of Fundulus, there seems some doubt whether the isolation of the cell is due to the properties of the protoplasmic surface or to the tough vitelline membrane (chorion) round the eggs, but in the case of the trout the latter membrane has been shown to be freely permeable and that the isolation of the yolk is effected by the membrane lying on the surface of the yolk in immediate contact with the perivitelline fluid. At first sight the impermeable qualities (in respect to water) of the trout’s egg seem to distinguish it sharply from the surface of an echinoderm egg, which allows water to pass in or out of a cell with comparative freedom. It should be noted, however, that the difference between these two types of cell is only one of degree. The surface of the sea-urchin’s egg is, in absolute units, highly impermeable to water: Northrop (1927) has shown that it is, in fact, 1/1000 times less permeable than a collodion membrane of the same thickness. It does not, therefore, seem a large step to reach the more absolute state of impermeability to water characteristic of the egg of the trout. It would be of interest to know whether the ovarian egg of the fish when in contact with body fluids of approximately the same freezingpoint as its own yolk, is also impermeable to water, or whether the impermeability of the eggs only sets in when the egg is laid in water. It is significant perhaps that the gills of the eel are apparently impermeable when the animal is in fresh water, but permeable to water and to chlorides when the animal is in salt water (Keys, 1931).
At various points in this paper comparison has been made between the unfertilised eggs of the trout and those of the hen. As morphological systems there is obviously a very close similarity between the two cells, although Needham and Smith (1931) are inclined to doubt whether there is any structural layer (and, of course, perivitelline space) between the yolk and the fibrous “vitelline membrane” of the hen’s egg. According to Needham, the weight of the available evidence is against the suggestion that the barrier to diffusion from the hen’s egg is located solely in a structural membrane. His opinion is based on the fact that when the vitelline membrane is dissected from the egg of a hen and interposed between two salt solutions, relatively rapid equilibration occurs across the membrane. Were such an experiment carried out with the egg of the trout, there can be little doubt that the mechanical disturbance involved would completely destroy the normal state of impermeability of the protoplasmic membrane. Such an experiment has in fact been made by Bogucki (1930), who found that the “membranes “of the trout or salmon’s egg when attached to the base of a glass tube readily allowed salts to pass through them1.
The extent to which mechanical disturbance destroys the normal properties of the egg surface of the fish depends on the age of the eggs, and on the degree to which the egg is being subjected to disturbances of its chemical environment (see p. 288 seq.). Thus, a young egg of the trout can be punctured by a needle and show no loss of salts—since the injury remains local and is quickly healed. If the egg is older or if the disturbance is more violent the more readily does localised injury involve an irreversible state of permeability. Similarly, the more the external environment of the egg differs from the normal the more likely is the spread of the effects of localised injuries. If we apply these conclusions to the hen’s egg a reasonable explanation appears to be available for the fact that the impermeability of the vitelline membrane is least affected by mechanical disturbance when it is washed in physiologically balanced Ringer solution and interposed between normal yolk and normal egg white (Needham, 1931). As pointed out by Needham and others, the normal rate of diffusion of salts across the surface of the hen’s egg must be extremely slow, and it may perhaps be questioned how far diffusion ever occurs across the normal egg surface in the natural life history of the egg. Under these circumstances it is of importance to consider how far the hen’s egg is actually permeable to water. Obviously, the degree of permeability to water must be low, otherwise the cell would gain water from the egg-white more rapidly than is in fact the case. There remains the possibility that in its natural condition the egg surface— like that of the trout—is quite impermeable to water. If the normal membrane of the hen’s egg were permeable to water—even to a small extent—the rate of uptake or loss of water from diluted or concentrated egg-white should vary with the osmotic gradient across the membrane. The data of Smith and Shepherd (1931) reveal the remarkable fact that such is not the case. Even markedly hypertonic solutions fail to withdraw water from the yolk, nor is there an uptake from hypotonic media which bears any relationship to the osmotic gradient across the egg membranes. Until the hypotonicity of an external medium has been raised to a freezing-point of half that of normal egg-white, the weight of the yolk remains substantially unchanged, and even when water has entered from still more hypotonic solutions it cannot be drawn out of the yolk again by concentrating the external medium. Smith and Shepherd remark that “There is a broad zone of concentration on either side of the point of isotony (to normal egg white) where the uptake of water is small2 (although quite definite) and almost independent of the concentration.” These changes are very different from those which would be expected to occur if the hen’s egg, like the muscle cells investigated by Hill (1930), were permeable to water when the egg is under normal conditions of environment. If these conclusions are justified, one is tempted to assume that under normal conditions the eggs of the trout and those of the hen are surrounded by surfaces which are impermeable to salts and to water. As the eggs of the trout increase in age, so they become more and more susceptible to mechanical injury and to such disturbances in the environment which are capable of destroying irreversibly the normal impermeable surfaces (see below). There is, however, one difference to be noted between the two types of egg. In the case of the trout, the breakdown of the essential surface occurs comparatively suddenly : for many days an egg may remain in an environment of distilled water without loss of salts—then relatively suddenly the surface breaks down and exosmosis is complete within two or three hours. In the case of the hen, however, the process of equilibration (under normal environmental conditions) is a much slower process, for Smith and Shepherd (1931) have shown that a slow but steady equilibration is constantly occurring in normal unfertilised eggs. This difference may be partly due to the larger size of the hen’s egg, but the situation is complicated in the latter case by the spontaneous generation of osmotically active substances in the two phases, a process which makes it difficult to say just how rapid is the equilibration being effected by diffusion across the vitelline membrane.
Whether or not this comparison between the osmotic properties of the egg of the trout and the hen can be justified is perhaps uncertain, but in the fish’s egg any process of osmotic equilibration between the interior of the cell and its external environment is a sign of a moribund unhealthy cell: as long as the cell is healthy it remains impermeable to intracellular electrolytes and to water, and this state is due to the static properties of the vitelline membrane.
II. DISTURBANCE OF THE NORMAL STATE OF IMPERMEABILITY
Factors influencing osmotic stability
Since the egg surface appears to act as a static membrane, it is of some interest to consider those disturbances in the environment which are capable of destroying its normal impermeable properties. At an early stage it became clear that the ability to resist the effect of an adverse environment decreases with the age of the eggs. This is very obvious if the eggs are subjected to mechanical agitation in their normal environment of tap water. As received from the hatchery (24-36 hours old) the unfertilised eggs are remarkably resistant to mechanical agitation and can be subjected to relatively violent disturbances without exhibiting any sign of osmotic instability. For example, they can be placed in a test-tube of water and violently agitated by air bubbles or by shaking for several hours without showing any sign of instability. After a sojourn of some weeks in the hatchery, however, very slight mechanical disturbance induces an exosmosis of salts within a very few minutes. The process of exosmosis can only be followed quantitatively when the egg is under observation in a conductivity cell, but a large number of such observations has shown that a detectable loss of electrolytes is very quickly followed by a localised or general loss of translucency at the surface of the egg (see p. 292); this visible change forms a convenient indication that exosmosis has occurred. If eggs are allowed to remain undisturbed in the hatchery for 3 to 4 weeks, and are then carefully transferred to the surface of a wad of cotton-wool in a Petri dish, a loss of electrolytes can readily be induced by gentle pressure applied to the chorion by the point of a blunt needle. Within a short time, a variable area of the egg surface surrounding the point of pressure becomes opaque, whereas in the case of fresh eggs, the point of a needle can be driven through both the chorion and the vitelline membrane, and then withdrawn without inducing any exosmosis of salts or subsequent loss of transparency.
In any attempt to assess the effect of a changed environment on the osmotic stability of the vitelline membrane, two factors must be maintained at a constant level : (a) the age of the eggs, (b) the degree of mechanical agitation to which the eggs are subjected.
In the case of fresh healthy eggs, the normal state of impermeability of the surface can be destroyed in a variety of ways, since almost any change in the normal external environment will sooner or later induce osmotic instability with a resultant exosmosis of salts. It is surprising to note, however, that (1) variations in concentration of balanced solutions of electrolytes, (2) physiological variations in hydrogen-ion concentration, (3) M/100 KCN, and (4) the presence of reasonable concentrations (5 per cent.) of the lower alcohols or ether, produce relatively little effect on the stability of the eggs if the latter are absolutely fresh and not subjected to mechanical disturbance. The most potent agents for producing exosmosis have been found to be aniline, pyridine, and heptyl alcohol—although probably many similar compounds would be found to be equally effective. The period of time which elapses before eggs begin to exhibit osmotic instability depends very definitely on four distinct factors : (a) the age of the eggs, (a) the type and concentration of exosmotic reagent employed, (c) the degree of mechanical agitation to which the eggs are subjected in the solution, (d) the presence or absence of calcium. The operation of these factors can be seen in the following experiments with heptyl alcohol. In Table II, column 5, the times given are those which elapsed before fifteen out of a total of twenty eggs exhibited visible signs of exosmosis.
In Table II, an absence of mechanical disturbance indicates that the eggs were completely undisturbed during the whole period of the experiment; when subjected to mechanical disturbance (+) they were placed in test-tubes with a bubble of enclosed air and rotated on a vertical wheel about 25 times per minute—a degree of agitation which does not influence the stability of young eggs in their normal environment. The presence or absence of calcium indicates distilled water with or without an amount of calcium equivalent to that in Cambridge tap water (about 0-002 N).
The greatest degree of osmotic stability is exhibited by fresh eggs in the presence of calcium and with an absence of mechanical disturbance ; the minimum of resistance is shown by old eggs in the absence of calcium during exposure to mechanical agitation. It is important to note that in no case were eggs subjected to violent agitation—but only to disturbances which they will survive for several hours when they are young and healthy. The effect of mechanical disturbance is very marked on eggs which are immersed in pure distilled water. Young healthy eggs, when lying undisturbed in distilled water, will retain their normal state for several days— but when rotated on a wheel they begin to show signs of exosmosis after a few hours or even minutes. It is of some interest to note that the onset of exosmosis from fresh eggs lying undisturbed in a solution of heptyl alcohol is marked by visible phenomena which are precisely similar to those seen in older eggs when these are subjected to slight mechanical disturbance only (see p. 292). The identity of the two phenomena is seen by subjecting eggs in heptyl alcohol to localised mechanical disturbance. Thus, in one batch of eggs, spontaneous instability occurred in saturated heptyl alcohol after 1 hour’s immersion. If, however, after 10 min. immersion, the surface of an egg was firmly pressed by the point of a blunt needle, a typical zone of permeability formed round the point of pressure and spread over the surface of the egg. This observation was made on many occasions and shows fairly clearly that the spontaneous loss of impermeability which occurs in solution of heptyl alcohol is due to localised injury.
It should be noted that when spontaneous osmotic breakdown of the cell surface takes place the process starts suddenly and does not necessarily originate at one point—for more than one area of disturbance can be seen simultaneously on the surface of a single egg, particularly if the exosmotic solution is present in relatively low concentration.
The results of these experiments leave the impression that the effect produced on the vitelline membrane by a change in the external environment of the egg is best defined in terms of the increased susceptibility which is produced in the vitelline membrane to localised mechanical injury. Such injury once started leads to a rapid loss of impermeability to electrolytes, proteins and water. Why heptyl alcohol, aniline and pyridine, should be highly active sensitisers is not known, and it is probable that many other substances might be found to be equally effective. The stabilising effect of calcium can also be illustrated by observing its effect on eggs immersed in solutions of sodium chloride. Table III illustrates this point.
The effect of calcium on the maintenance of the osmotic stability of eggs is in keeping with the older observations of Loeb (1912) on Fundulus eggs1, and with the work of Osterhout (1922) on Laminaria tissue. Phenomena of this kind have been explained by the assumption that the impermeable cell surface consists of a calcium compound which is unstable in salt solutions unless free calcium ions are present in the external medium, whereas the stability of the cell surface is maintained in distilled water as long as the cell does not substitute hydrogen ions for calcium ions (see Gray, 1926). Clearly a physical dispersion of the cell membrane could lead to a loss of impermeability, but in view of the precise way in which the loss of impermeability affects the whole egg (see p. 292) it is doubtful whether physical dispersion of the cell surface gives a complete picture of the effects produced by a lack of calcium ions in the surrounding medium. According to Pantin (1931) the presence of calcium ions depresses the rate and the extent to which salts leave the cells of Gunda in a hypotonic medium, but in the case of the trout, neither the rate nor the amount of exosmosis is affected by the presence of calcium once the process of exosmosis has begun. After allowing for the effect of an exosmosis of hydrogen ions from the egg, the amount of electrolytes diffusing into tap water was found by electrolytic methods to be the same as that which diffused into distilled water. The primary effect of calcium on the egg is to prolong the latent period during which no exosmosis occurs.
If it be assumed that the essential difference between tap water and distilled water is due to the presence or absence of calcium ions, the observations of Pumphrey2(1931) are of interest. The potential difference across the egg surface is normally from 13-18 millivolts when the external medium was tap water; on replacing the tap water by distilled water the potential difference rose to 60-100 millivolts. Osterhout and Hill (1930) have shown that the ability of a Nitella cell to transmit a propagated wave from a localised area of injury depends on the potential difference across the protoplasmic surface—the higher this potential difference the larger is the injury current and the more readily does it propagate itself over the cell surface. On the other hand, the microscopic changes (see p. 292 seq.) which accompany the loss of the normal impermeable properties of the vitelline membrane suggest that the presence of calcium may alter the physical state of the membrane in such a way as to delay its migration from a point of injury. If calcium is present, the disturbance required to effect a localised rupture of the membrane may be greatly increased, whereas in the absence of calcium the material of the membrane loses its normal physical stability and is capable of extensive movements in the region of localised injury (see p. 294).
Visible changes which accompany the loss of normal impermeability to electrolytes
It has already been noted that the vitelline membrane of the egg with its included oil drops is normally translucent, although it is not completely transparent. If an egg be exposed to a strong exosmotic reagent such as 5 per cent, aniline solution, the whole of the egg surface quickly becomes less translucent and eventually opaque owing to the precipitation of the globulins in the yolk. In its opaque state, the interior of the egg cannot be examined microscopically, but by exposing such an egg to Ringer solution the globulins are re-dispersed, and portions at least of the vitelline membrane can usually be seen in their normal position closely applied to the outer chorion.
On the other hand, the loss of impermeability which is induced by mechanical agitation (or by reagents such as heptyl alcohol) is attended by visible phenomena of quite a different character. When aged1 eggs (or young eggs immersed in a saturated solution of heptyl alcohol) are pressed with the point of a blunt needle—a small opaque ring (Fig. 5 A) quickly forms round the point of contact with the needle. The region of the egg surface lying within the opaque ring remains translucent, and the ring itself gradually enlarges in size, although the width of the opaque periphery remains practically constant (Fig. 5 A-D). In many cases such a ring may pass over the whole of the egg surface with the exception of the blastodisc, although in a few cases it may become stationary before this condition is reached: after about 1 min. the peripheral opaque edge of the ring thickens on its inner side until the whole area enclosed by the ring has become opaque (Fig. 5 F). Phenomena of this type have invariably been observed to occur when eggs have been subjected to localised mechanical injury, and after a clearly defined latent period they occur spontaneously when osmotic instability is induced by heptyl alcohol. When first observed it seemed possible that this curious phenomenon could be regarded as evidence of a wave of irreversible permeability passing over the surface of the vitelline membrane from localised areas of injury. Such waves are known to occur in other tissues (Osterhout and Hill, 1930, 1931), and are known to be propagated with greater ease if the E JW.F. across the protoplasm is increased ; the E.M.F. across a trout egg surface is much higher in a medium of distilled water than in a medium of tap water (Pumphrey, 1931), and a reasonable explanation would be thus forthcoming for the greater ease with which the rings form in solutions of heptyl alcohol in distilled water than in similar solutions in tap water. Further investigation revealed, however, another explanation. As already mentioned, the opacity which rapidly follows the formation of a ring makes it impossible to follow the events which are occurring inside an egg which is immersed in water. If localised injury is induced in an egg immersed in Ringer solution, the subsequent precipitation of globulins can be prevented and a well-defined series of morphological changes can be observed if the egg is examined through a horizontal microscope. Under these conditions, no visible ring is formed—but the vitelline membrane withdraws from the surface of the chorion leaving a lens of fluid (Fig. 6) between the two membranes. In some cases, this fluid lens remains unchanged for a considerable period, but in most cases the whole egg cell gradually shrinks in volume, leaving a larger and larger perivitelline space (see Fig. 7). Simultaneously, a portion of the blastodisc rises through the perivitelline space at the upper pole of the egg (Figs. 6B and 7 B) and becomes adherent to the chorion: the vitelline membrane shrinks still further and is eventually aggregated as an irregular mass containing oil globules, which is attached to the chorion (Fig. 7 H). As soon as the vitelline membrane begins to shrink away from the chorion, the fluid in the perivitelline space can be shown to contain globulin in solution, for if the outside medium be changed to tap water, rapid coagulation occurs within the perivitelline space. There can be little doubt that the “rings” which form on eggs when injured locally in a medium of tap water are due to a rapid loss of yolk through the vitelline membrane at and near the region of injury : the opaque edge of the ring, being thinner than the rest, exhibits precipitation of globulins until these are redispersed by the electrolytes of the yolk which continues to flow out from within the vitelline membrane.
These observations illustrate two important points: (1) as soon as the vitelline membrane becomes permeable to electrolytes and water, it also becomes permeable to proteins and is capable of contracting in such a way as to force the material of the yolk into the perivitelline space, (2) the changes which accompany the normal loss of impermeability occur quite suddenly and end in a complete destruction of the whole structure of the egg cell. Except in a few cases, where the retraction of the vitelline membrane from the chorion is localised, the process is rapidly completed when once it has begun. Under normal conditions, the vitelline membrane is probably in a state of elastic tension as in the case of other protoplasmic films (see Plowe, 1931)-In young healthy cells, however, the membrane is highly extensible and it is not easily ruptured. Once local rupture has occurred, however, the elasticity of the membrane forces yolk through the opening until the membrane is aggregated as a compact mass or until the rupture has been healed. The whole process is very similar to the phenomena described by Plowe (1931) in plant cells. Owing to the included oil drops, the contracted vitelline membrane and blastodisc are in equilibrium when resting against the upper surface of the chorion. It is, however, possible that the state of elastic tension is itself the effect of localised injury.
The series of phenomena which accompany the breakdown of the normal properties of the vitelline membrane appear to be somewhat analogous to the process of haemolysis of a red blood corpuscle (Parpart, 1931), and suggest very strongly that exosmosis from cells of this type is an infallible sign of an injured cell.
In a limited number of cases, the extrusion of yolk into the perivitelline space has been found to cease at stages varying from Fig. 7 D-F, and in this condition it is obvious that there may be a relatively stable and sharp-cut boundary between the material within the vitelline membrane and the fluid in the enlarged perivitelline space; in such cases the average osmotic pressure of the entire egg is no guide to the osmotic pressure of the material inside the vitelline membrane. The enlarged perivitelline space of such an egg would thus represent the “blisters” which appear on the surface of a hen’s egg when these eggs are exposed to hypotonic media (see Smith and Shepherd, 1931).
The visual phenomena which attend a loss of electrolytes from an egg suggest that exosmosis is usually due to a mechanical rupture of the vitelline membrane at one or more places. If the egg is young and healthy, it is not easy to effect a rupture of the membrane, but in other cases very slight disturbance is sufficient to break the elastic membrane, and thereby causes the whole egg cell to collapse much like a pricked bubble.
Note on the quantitative estimation of exosmosis
The rate of exosmosis of salts from a cell or group of cells can readily be followed by the conductivity method described elsewhere (Gray, 1920), but it is by no means easy to interpret its real meaning. In the case of relatively large structures the rate of exosmosis may be expected to follow Fick’s law for a considerable period as long as the whole surface or a definite fraction of it is permeable to electrolytes. If, however, the cell is very small, Fick’s law will break down at a comparatively early stage owing to a fall in the concentration of salts at the centre of the egg. In actual practice the observed course of exosmosis from a permeable cell will at first conform to Fick’s law and, passing through a phase which it is not easy to analyse, will then approach more and more to an exponential form where the rate of exosmosis becomes proportional to the total amount of electrolytes left in the cell (assuming a large volume of external medium). The size of the trout’s egg is such that the transitional period is reached when approximately half the internal electrolytes have left the cell. This type of curve is obtained when a dead egg, artificially impregnated with electrolytes, is transferred to distilled water. Under such circumstances the electrolytes diffuse outwards at the same rate as from an agar block of the same size (see fig. 8, Curve A).
The exosmosis from a living cell follows quite a different course, since the curves are invariably sigmoidal (see fig. 9). The sigmoidal nature is clearly in conformity with the conclusion that the normal barrier to diffusion is of a heterogeneous nature in the sense that permeability does not set in simultaneously at all points on the egg’s surface. It is impossible to draw from such curves any reliable picture of what is happening at the egg-surface, they simply present an empirical expression of those processes of diffusion which attend the total disruption of the normal cell-surface.
REFERENCES
If the tough membrane enclosing the teleost egg is derived from the follicular cells of the ovary, the term “chorion” is justified, otherwise the term “vitelline membrane” is preferable. Many authors have applied the term “chorion “to the tough membrane of the trout’s egg, and have applied the term “vitelline membrane” to the delicate membrane which separates the yolk from the perivitelline fluid. Throughout this paper the following nomenclature will be followed :
Trout egg. Tough outer membrane = chorion. Membrane between yolk and perivitelline space = vitelline membrane.
Hen’s egg. Visible membrane between yolk and white = vitelline membrane.
By “average” freezing-point is meant the freezing-point of the yolk + perivitelline fluid.
That profound changes in permeability to ions result from mechanical disturbances is evident from the recent work of Osterhout (1931) and of Blinks (1930).
It does not seem clear that this small uptake of water is not due to causes quite independent of changes in the concentration of the external medium, since changes of some magnitude occur when the egg is left undisturbed in the shell. Probably, as Smith and Shepherd point out, such changes are influenced by events which occur independently in the yolk and in the white, and which are not due to any simple process of equilibration between the two phases.
Loeb also noted the deleterious effect of the higher alcohols on the impermeability of these eggs.
This author also noted the stabilising effect of calcium ions when the normal egg is punctured by a micro-electrode. In the absence of calcium ions an egg was much more hable to exhibit exosmosis and die.
“Aged” eggs are unfertilised eggs which have been in the hatchery for 3-4 weeks.