Protoplasmic Action and Nervous Action

(e.g., killed by heat). Results of the following kind were obtained with frog skin: It is clear that the rate of evaporation is much greater through the dead membrane. Similar experiments with plant tissues (leaves) gave similar results; von Mohl (in 1847) and Naegeli (in 1861) had already called attention to the fact that frozen plant tissues (leaves, fruits) dried more rapidly than Hving.^ Bernstein experimented also with thin sheets of muscles (abdominal obhque of frog) , both living and killed with chloroform, and found that the rate of evaporation through the living muscle was about half of that through the dead; the difference was greatest during the first hour, when the living tissue was relatively normal.

Bernstein's conclusion was that living tissue has greater water-binding power than dead, and he regarded the electrically polarized condition of the plasma membranes as the chief factor hindering the outward passage of water. Apparently a potential difference of the value of ca. 0.1 volt exists normally between the outer and the inner surfaces of the plasma membrane; this gradient is positive externally and negative internally. Resistance is thus offered to the passage of the positively charged water outwardly across the membrane. This explana-

tion, while ingenious, does not seem sufficient; in muscle and other living cells immersed in anisotonic solutions water seems to pass with equal readiness in either direction through the membrane; for example, Arhacia eggs in hyper- and hypotonic sea water shrink and swell (respectively) at about the same rates/ Hober refers the difference observed by Bernstein to the greater turgor of the living cells ;^ this explanation also seems doubtful, since turgor is either absent or slight in vertebrate tissues. The simplest as well as most probable explanation is that the permeabiUty to water increases with death, along with the permeability to other substances. As the plasma membrane loses its semipermeability, w^ith the associated fineness of texture, it also loses its relative impermeability to water.

Permeability to water is one of the little studied properties of cells. Yet it is an important property which appears to be constant for a particular cell under definite conditions. There is evidence that it varies with the physiological state and activity of the cell, and in certain cases, especially in gland cells, the indications are that it is under nervous control. Thus when the chorda t^onpani is stimulated, the submaxillary gland secretes a copious watery saliva; similarly the sweatglands and the kidney secrete actively under certain conditions, but not under others. In most cases the secretory substances leave the cell in aqueous solution; and although the factors are complex, there seems to be little doubt that the transport of the secretion across the cell-boundary is associated with an increased permea-

bility to water as well as to dissolved substances. The character of the bioelectric variation accompanying secretion also indicates this. Antropoff^ has studied the influence of the waterpermeability of osmometer membranes on the rate of osmotic transfer of water. He reaches the general formulation : i.e., the rate of passage of water into an osmometer -5^ j at any time is proportional to the permeability of the membrane to water (a) and to the osmotic pressure of the solution (P) less the hydrostatic or other pressure (Pi) resisting the transfer. The rate of osmotic transfer of water is thus proportional to the product of the effective osmotic pressure into the permeability to water. This formulation defines the conditions for perfect membranes; i.e., those which are permeable to water and impermeable to solute.

In the Arbacia egg the permeability to water is altered in a remarkable manner as a result of fertilization. A quantitative measure of this permeability may be obtained by measuring the rate at which water enters or leaves the egg in dilute or concentrated sea water. This rate is found to be increased about fourfold after fertilization, indicating that this process is associated with a marked increase in permeability to water. If mixed fertilized and unfertilized eggs are placed in dilute or concentrated sea water, the former undergo change of volume much more rapidly than the latter, and

within a minute or less the two can easily be distinguished by their difference in size. The semi-permeability of the plasma membrane is not impaired, although the permeability to water is increased. The conditions of this phenomenon are probably complex, and determined by metabolic factors of unknown nature. The change of permeability is progressive and occupies a considerable time, some 20 minutes elapsing (at 20°) before it nears its final stage; it is arrested by anaesthetics (chloral hydrate, ure thane, alcohols) and by cyanide in higher concentrations.'

If the impermeability to water in this cell is due to the presence of water-insoluble substances (lipoids or cholesterol) in the surface-film, the change above would indicate that these are altered or removed in part; i.e., a change in the chemical composition of the surface layer is to be assumed. The nature of this change is unknown ; but Lyon's observation that the iodine-combining power of the egg is decreased after fertilization'' may indicate a decrease in certain unsaturated compounds which would otherwise take up the iodine — possibly cholesterol or unsaturated lipins (lecithin). The permeabiHty of fertilized eggs to water can be artificially modified in certain ways; it is decreased by anaesthetics (chloral, ure thane, alcohols),^ a fact also indicating a dependence on the lipoid-content of the protoplasmic surface-film.

Chambers finds certain differences in the behavior of fertilized and unfertilized echinoderm eggs in microdissection;^ the surface of the unfertilized egg is less easily cut or torn with a needle, but exhibits less power of repair than that of the fertilized egg. This latter difference is probably to be correlated with the increased rate of metabolism following fertilization; oxygenconsumption, COz-evolution and heat-production are then greater,^ also the susceptibility to KCN poisoning and other forms of chemical injury. Facts of this kind illustrate the close correlation existing between the physical state of the plasma membrane and the physiological properties and metabolic activity of the cell. Other examples of this correlation will be described later, especially in relation to stimulation-processes, which are intimately associated with changes in the membranes.

It was first pointed out by Nernst in 1890 that the solvent properties of the substances composing a membrane may determine the latter's permeability to dissolved substances, and hence the production of osmotic effects. Thus when benzol is separated from ether by a partition consisting of bladder membrane soaked in water (in effect a layer of water), the ether, because of its greater solubility in water, passes the partition more rapidly than the benzol, and a pressure is set up on the benzol side of the partition.^ The membrane is impermeable to

^ Chambers, Proceedings of the Society of Experimental Biology and Medicine, XVII (i9i9),4i. benzol, which is ahnost insoluble in water, while ether is readily soluble (to the extent of about 12 volumes per cent) and penetrates the partition. Experiments illustrating the same principle may be performed with gases; thus if a tube, containing air, closed below with a water-soaked membrane and connected above with a manometer, is placed in a vessel of ammonia or hydrochloric acid gas, the pressure rapidly rises in the tube because of the penetration of the water-soluble gas through the water-layer. Ramsay's well-known experiment with a tube filled with nitrogen and separated from a hydrogen atmosphere by a palladium partition (at 350° to 400°) illustrates the same general phenomenon; hydrogen but not nitrogen penetrates the partition, hence the pressure rises within the tube.'

In all of these cases the permeability of the partition depends on its consisting (in part) of some material which acts as a solvent to the penetrating substance. Nernst called attention to the possibility that similar conditions might exist in living cells; i.e., that the solubility of substances in constituents of the protoplasmic membranes might be the condition of their selective absorption by cells; and this form of explanation was later adopted by Overton who reached the conclusion, as a result of his studies on permeability already described, that the protoplasmic surface layer consisted chiefly of compounds with solvent properties similar to those of typical organic solvents.

Later, chiefly because of experiments indicating a parallelism between the penetration of various acid and basic dyes into living cells and the solubility of the same dyes in solutions of lecithin and cholesterol,^ Overton referred the characteristic permeability of the plasma membrane specifically to the presence of lipoids. This evidence is in itself scarcely conclusive, especially since other factors are now known to be of importance in the penetration of dyes into cells, and many exceptions have been found to the rule that lipoid-solubility connotes ready penetration. The state of colloidal subdivision has been shown to be an important factor,^ and the negative electrification of the cell-surface must also play a part by favoring the adsorption and penetration of the positive particles of the basic dyes, which Overton found to penetrate most readily. But other evidence of various kinds supports Overton's conclusion in the main;^ and there can be little doubt that the lipoids form essential constituents of plasma membranes, even although other compounds, especially proteins, may be of equal importance; e.g., as furnishing a structural support to the lipoids. The influence of lipoid-solvents on the permeability and activity of cells, the characteristic cytolytic action of these compounds, the influence of salt-solutions on oil-water emulsions, the peculiar relations of cholesterol to the mechanical properties of the plasma membranes, the properties of artificial lipoidimpregnated membranes, and many other facts all indicate the important role of lipoids in determining the properties of plasma membranes. It is not to be con-

3 Especially the results with strong and weak acids and bases, above cited. Cf. the discussion in chap, viii, pp. 428 £f ., of Hober's Physikalische Chemie der Zelle. eluded that these membranes are simple continuous sheets of lipoid; the very fact that their properties are complex and vary from species to species is inconsistent with any such simple view, and the existence of specific cytolysins may indeed be taken as proof that proteins form an essential part of their composition. In fact, as already pointed out, the plasma membrane is not to be regarded as a simple passive layer of colloidal or other material, but rather as a special living structure with a characteristic metabolism of its own, and with both its physical and chemical properties modified in correspondence with its situation at the cell-boundary. All that can safely be maintained is that its properties are intimately dependent on the properties of its lipoid constituents, hence vary with changes in the physical and chemical state of the latter. It will be unnecessary to review in further detail the large body of experimental fact indicating the presence of lipoids in the surface-films of cells; this evidence is discussed at length with full references to the literature in Hober's and Bayliss' textbooks.

Of late years electrical conditions have been shown to be of great importance in determining the permeability of artificial partitions (parchment, porcelain, and other substances), and from general principles it seems certain that such factors must play a corresponding part in the protoplasmic membranes. In artificial membranes two factors have been shown to be of importance: (i) the potential-difference between the solutions in contact with the opposite faces of the partition, and (2) the potential-difference between the soKd or colloidal material composing the partition and the fluid occupying its interstices or pores/ The influence of these electrical factors is shown with especial clearness in the phenomena of electrical endosmose, and in the related phenomenon of negative or anomalous osmosis. In both phenomena an electrified layer of fluid (occupying the pores of the partition, hence in contact with the surface of the structural material) is situated within the electrical field between the two surfaces of the partition; the fluid is accordingly transported across the partition by electrostatic attraction in one or the other direction, so

work involved in the transport is electrical; in negative osmosis the electrical field is maintained by the unequal diffusion-rates of the ions of the solution bathing the partition; in electrical endosmose by the external current traversing the partition. It will be evident that under these conditions the permeability of the partition to electrified material, whether water, colloidal particles, or ions, will be different in the two directions. Water, for example, will diffuse outward in an osmometer, i.e., from the more to the less concentrated solution, when the electrical diffusion field is oriented in such a way that the charged layer of fluid in the pores is attracted outwardly with sufficient force to overcompensate the purely osmotic effect. This condition is met when a stronger solution of HCl is

separated from a weaker solution by a porcelain partition; the solution on the more dilute side is then positive, because of the diffusion field; the solid substance of the partition is positive and the water in the pores is negative; accordingly the latter is drawn toward the more dilute solution. Anomalous osmosis of this kind has long been known; Graham (1854) observed, for example, that K2SO4 showed positive osmosis in alkaline solution and negative in acid when a bladder membrane was used; NaCl, on the contrary, showed positive osmosis in acid solution and was indifferent in neutral solution.^ These changes in the direction of transport are now recognized as depending on the influence of the ions on the charged condition of the structural surfaces. Many surfaces are rendered positive by H ions and polyvalent cations; the adjacent water-layer, being then negative, moves in a corresponding direction in the potentialgradient between the two surfaces of the membrane. Water may thus move in one or the other direction according to the electrolyte content. The influence of ions on the direction of transport is shown with especial clearness in the experiments of Perrin^ and others on electrical endosmose. The influence of ions on anomalous osmosis has recently been studied by Loeb, using gelatinepermeated membranes of collodion; the effect varies with the Ph of the solution and becomes minimal at the isoelectric point of gelatine. Isoelectric membranes can,

however, become charged and exhibit anomalous osmosis under the influence of trivalent ions.^ The transport of the ions of salts, and consequently the permeability of a given membrane to a diffusing salt, are similarly affected by the electrical state of the partition, and this influence has recently been studied by Girard, using bladder membranes. If the polarization of the membrane (P.D. between its opposite faces) has a certain orientation, the penetration of a salt like MgCli or Na2S04 in the one direction is facilitated, in the other direction hindered.^

The application of the principles above to the case of the plasma membrane is somewhat uncertain, since the structure of the latter is in many respects different from that of fixed porous membranes of macroscopic dimensions. Yet variations in the P.D . across the plasma membrane must have a corresponding effect on the permeabiHty; this effect, however, is probably accompanied in living protoplasm by other effects, such as chemical effects depending on the electrode-Hke action of the membrane, and effects on colloidal dispersion. There is no doubt that the permeabihty of many plasma membranes, especially of irritable cells, is very sensitive to changes of electrical condition; thus the turgor motors of plants {Mimosa, Dioncea.) depend for their action upon variations of permeability, which are readily induced by the electric current. The passage of a current through a muscle or nerve, in such a way as to decrease locally the resting polarization of the cell-

surface (i.e., cause depolarization), causes stimulation, which, as will be shown below, is attended with an increase of permeability. We must conclude that electrical factors are of great importance in determining the properties of living plasma membranes; variations in the external electrical conditions occasion corresponding variations in the membrane, and with these are connected various physiological effects. The polar disintegration shown by various cells through which currents are passed also indicates a direct action on the membrane; the character of this action is determined by the direction of the current. The law of polar stimulation is the expression of similar conditions in irritable tissues.^ The physiological effects of salts and electrolytes are undoubtedly to be referred largely to changes in electrical conditions, as shown by the numerous parallels between the action of salts in purely physical phenomena like electrical endosmose or change of colloidal aggregation, and their action on vital activities of various kinds.

The relation of the inorganic salts present in the external medium to the properties and activities of living cells has been the subject of much investigation since Ringer's time, and the present brief account will be confined to the more general and fundamental relations of this kind, especially those indicating that the chief basis of the physiological action of the salts of the medium is their action upon the protoplasmic surface layers. Since much of the pioneer work in this field is due to Overton, and since many of the results of his studies are applicable to living protoplasm in general, I shall first give a somewhat detailed summary of his earlier experiments on the action of salt solutions on the muscle and nerve of frogs. ^

Overton first examined carefully the well-known effect, reversible loss of irritability in isotonic solutions of indifferent non-electrolytes such as sugar, using small muscles; e.g., sartorius, cutaneus pectoris, and foot muscles. He found that all indifferent non-electrolytes, independently of their special composition (dextrose, sucrose, lactose, erythrite, mannite, alanin, taurin, and asparagin), produce this effect. The action is not poisonous, since the addition of a small proportion of an isotonic solution of sodium chloride to the non-electrolyte

solution prevents the loss of irritability; the inference is therefore justified that the effect depends upon a withdrawal of electrolytes, especially sodium chloride, from the tissue. That the electrolytes thus removed come almost entirely from the interstitial spaces of the tissue, and not from the interior of the cells, was later proved by Urano and Fahr.^ The important conclusion follows that the presence of electrolytes (salts) in the external medium is essential to the normal irritability of the cell.

Overton discusses the question whether the removal of sodium salts acts by preventing the conduction of stimulation or by deranging the contractile mechanism of the cell. Biedermann had previously shown that after incorpoiation of sufficient water in hypotonic salt solution, a muscle may lose the power of contraction without losing that of conducting stimuli.^ By immersing a portion of a sartorius in isotonic sugar solution, Overton showed that stimuli are not transmitted through the salt-free muscle. It is known that a muscle deprived of irritability in sugar solution will shorten in solutions of chloroform or other cytolytic substances; presumably, therefore, the contractile mechanism is structurally intact, but fails to act in the absence of electrolytes because of the failure of conduction. Without the power of transmitting stimuli, the muscle is unable to contract as a whole.

The least concentration of NaCl required for the maintenance of irritability was determined by using mixtures of isotonic sugar solution (6 per cent) and NaCl solution (0.7 per cent). No significant decline in irritability was found until the salt-content fell below 0.15 per cent. At o.i per cent irritability was distinctly less than normal, and, with further decrease in concentration, it declined rapidly, becoming zero at about 0.07 per cent. In mixtures containing less than 0.07 per cent NaCl irritability was lost as rapidly as in pure sugar solution.^ The time required for complete loss of irritability thus represents the time required for diffusion to reduce the NaCl of the intercellular spaces to this concentration.

Sugar-treated muscles rapidly regain irritability in solutions of all Na salts. The nature of the anion associated with the Na was found to be indifferent, provided it was not too toxic; with most salts the minimal concentration for the maintenance of irritability was about the same as with NaCl. Apparently, therefore, it is the Na ion, and not the undissociated molecule or the anion, which is responsible for the maintenance of irritability. A definite function, that of preserving the normal irritabihty of muscle and nerve, is thus to be ascribed to the Na salts in blood plasma; their role is not merely osmotic, as formerly supposed.^ This relation of Na ions to irritability and contractility is a specific peculiarity of muscle cells; the conditions in other contractile forms of protoplasm are often widely different; thus the contraction of cilia, spermatozoa

^ This effect apparently varies with oxygen tension. Pond has recently shown that in solutions saturated with oxygen the NaCl content may be reduced to less than .05 per cent without loss of irritability: Jour. Gen. Physiol. ^ III (192 1), 807. 2 Cf. also J. Loeb, "On the Production of Rhythmical Contractions in Muscles by Ions," Festschrift fur Pick (1899), p. loi. and protozoan structures like the stalk of Vorticella is independent of Na ions in the medium ; cilia, in fact, often exhibit prolonged and vigorous activity in media, like isotonic KCl or MgCl? which rapidly and completely abolish irritability in muscle and nerve. ^

Since analysis shows little or no sodium in the interior of the muscle cells, the conclusion follows that the essential action of the Na ions is exerted upon the semi-permeable surface layer or plasma membrane and that penetration into the internal protoplasm is unnecessary. Overton, however, points out that it is not necessary to assume impermeability to these ions under all conditions; it is possible that under some conditions, e.g., stimulation, there may be an exchange between the Na ions in the medium and other cations (e.g., K) present in the muscle cell. This suggests that impermeability to Na ions is a characteristic of the muscle during the resting state only, a view implying that changes of permeability are an essential factor in stimulation. Similar views were expressed by Bernstein and Brtinings at about the same time.^

Experiments on the substitution of other cations for Na showed that lithium salts were the only ones capable of maintaining irritability in the same manner as Na salts. Pure isotonic solutions of LiCl (0.435 P^^ cent) are injurious to muscle, but a mixture of this solution with an equal volume of isotonic sugar solution has indifferent properties like those of an NaCl solution. The least concentration of LiCl required for maintaining ^ The cilia of many marine animals, e.g., Arenicola and Mitylus, remain active for hours in pure solutions of K and Mg salts.

irritability is ca 0.05 per cent, a value comparable with that found for NaCl. Lithium chloride, nitrate, sulphate, phosphate, and acetate all showed similar action. All salts of the other alkali cations (NH4, K, Rb, and Cs), in concentrations equivalent to 0.07 per cent NaCl, promptly destroyed irritability. NaCl solutions containing a little K and Ca (Ringer's solution) were more effective in reviving irritabiHty than pure NaCl solutions; irritability, however, was not sustained by solutions containing Ca and K in the same proportions, but sugar in place of NaCl. Na and Ca salts, especially when both are present in the solution, antagonize the toxic action of potassium salts; this is readily shown by using concentrations of KCl, which, acting in pure solution, rapidly destroy irritabiHty.