Protoplasmic Action and Nervous Action

The question of the condition of the salts in the living protoplasm arises here; and this question is of considerable general importance, since it has been held by certain investigators that these salts are present chiefly or entirely in a combined or adsorbed state and are hence not free to act as conductors. As we shall see later in dealing with the phenomena of stimulation and transmission, the electrical conductivity of the internal protoplasm appears to be a necessary factor in its normal activity; and any evidence that this conductivity is what we should expect it to be from the known saltcontent of protoplasm is of interest. The contention that the salts in protoplasm are non-ionized, or that the ions are in some manner rendered immobile, is inconsistent with the physico-chemical observations relating to the behavior of salts in the presence of proteins, lipoids, or other colloids. According to Bugarszky and Liebermann, the addition of even large quantities of protein to salt solutions affects the ionic concentration only slightly;^ Michaelis and Rona^ have shown by '^com-

pensation dialysis" that the salts in serum (containing 10 per cent protein) are freely dialyzable. Pauli and Samec have also shown that alkali salts are not more soluble in serum than in water.^ There is the possibility of the formation of combinations with the amino-acids present in the proteins; but the salts of such weak acids would theoretically be almost completely hydrolyzed; i.e., a stable combination with protein in which the salt is completely and firmly bound is scarcely conceivable. Any protein-salt combinations thus formed would hydrolyze, and the products of hydrolysis would diffuse out through the membrane if the latter were permeable; and further hydrolysis would proceed until an equilibrium was reached in which a large proportion of salt was present in the free dissolved state."^

General chemical theory thus indicates that only a very small proportion of the salt in the cell can be in a state of permanent combination with protein; hence the characteristic difference between the salt-content of the cells and that of the medium cannot be thus explained. It is probable, therefore, that the semi-permeability of the plasma membrane is an essential factor in making possible this difference. If it were possible to measure the electrical conductivity of the cell interior apart from that of the plasma membrane, the question could be answered at once. According to the view presented above, the chief

2 In other words, the proteins and other compounds present in the cell could not hold more than a very small proportion of the salts in a combined and indiffusible form. barrier to the movement of ions is at the semi-permeable plasma membrane. The increased conductivity observed at death, or after cytolysis by saponin or other compounds, favors this view, since semi-permeability is then lost; i.e., there is a general parallelism between the ability of salts, sugars, and other soluble compounds to pass the plasma membranes and the electrical conductivity

— a fact indicating that the chief condition rendering living cells such poor conductors of electricity is the impermeability of the membranes to ions. From these considerations we should conclude that the ions within the cell are free to diffuse within the space inclosed by the plasma membrane. But the case requires to be tested by experiment, for it might be held, in spite of the foregoing considerations, that the electrolytes are in some manner chemically combined in living protoplasm and are set free only at death.

Hober has attempted to measure the conductivity of the cell interior, using in part methods suggested to him by Nernst. A device of the plan shown in Fig. i was first employed.' A rapidly alternating current is induced in the circuit containing the two balanced condensers Fig. I. — I, inductorium; C, condenser; G, spark-gap; S, coil (self-induction) of primary circuit; 53, coil of secondary circuit; A, adjustable condenser; B, condenser with space between plates for insertion of vessel (F) containing suspension of cells; B, detector in bridge; Ri, Ri, resistances (from Hober, loc. cit.).

A and B. When the capacities of the condensers are the same and the resistances of the two halves of the circuit (on either side of the bridge) are the same (or current passes through the detector; but if the capacity of one condenser (A) is changed, the current passes until the capacity of the other (B) is made the same. Now the capacity of a condenser is increased if a conducting layer is introduced into the dielectric between the plates, and to a degree which is proportional to the conductivity of the layer. Hober's method, therefore, consists in introducing between the plates of one of the condensers (B), after the system has been brought into a balanced condition, a glass vessel containing a suspension of living cells (blood corpuscles) in isotonic sugar solution. Such a suspension does not conduct electricity (in the usual Kohlrausch sense), yet it increases the capacity of the condenser and so allows a current to flow across the bridge. This result shows that the suspension contains an electrically conducting fluid; this can only be in the interior of the cell, since the suspension-medium is sugar solution, which is a non-conductor. The addition of saponin (which destroys the membrane and increases the Kohlrausch conductivity) was found not to change the capacity, indicating that the conductivity of the cell-contents is the same whether the semipermeable membrane is present or not. If a glass vessel containing a salt solution is placed between the condenser plates, a similar increase of capacity is shown; and by comparing the effects produced by salt solutions of known conductivity with those produced by suspensions

of cells, the internal conductivity of the latter can be estimated. The measurements cannot be made very exact; they indicate, however, that the internal conductivity of the corpuscle is of the order of that of a o.i n KCl solution (between o.i n and o.oi n). Hober also experimented with another method somewhat different in principle.^ If high-frequency oscillations are induced (lo^ per second) in a circuit containing a condenser and a detector (spark-gap), and the beaker {B, Fig. 2), encircled by a coil of wire forming part of

Fig. 2. — C, condenser; V, beaker containing suspension of corpuscles; G, sparkgap; S, coil in which current is induced (Hober, loc. cit.). the circuit, is filled with a conducting solution, the oscillations are damped. Hober again found that cells suspended in sugar solution produce this effect, and to the same degree whether they are intact or cytolyzed with saponin, although the Kohlrausch conductivity is, of course, much greater in the latter case. The low Kohlrausch conductivity of intact cells is thus apparently due to the inclosing plasma membranes; the internal conductivity of the protoplasm is high. Again by comparing the effects produced by cellsuspensions with those produced by salt solutions of

known conductivity, Hober found that the suspension gave the same effect as a NaCl solution of a concentration similar to that of blood plasma (i.e., between o.i and 0.4 per cent). Further accuracy is not possible. Hober also used the method of measuring the conductivity of cells by means of rapidly oscillating currents.^ According to theory, we should expect that the higher the oscillatory frequency the less difference would the presence of the membrane make. Under the conditions, frog's muscle, which had been washed thoroughly in sugar solution, behaved as if it had a conductivity between o.i and 0.2 NaCl; the greater the frequency of alternation the greater the conductivity. This is . readily understood if we reflect that when the carriers of the current (ions) have to pass through only very short distances, they are not impeded in their movements by the membrane.

From the results of the experiments above Hober reaches the following conclusion:^ ''It can be regarded as certain that the blood corpuscles possess a very considerable internal conductivity, even when the external conductivity is minimal; i.e., that free ions are present in their interiors, and that these are prevented from reaching the outside only by the presence of a barrier to diffusion"; i.e., the plasma membrane. The observations cited by Kite^ and others, as indicating the impermeability of the internal protoplasm to diffusing substances, are probably to be explained as indicating the rapidity with which a semi-permeable

2 Hober, Physikalische Chemie der Zelle und der Gewebe (i9i4),p. 385. barrier is formed at any free surface of living protoplasm.' This property of forming fresh semi-permeable surface films at exposed protoplasmic surfaces has long been known. Nageli^ describes experiments with the root hairs of Hydrocharis; by crushing these structures under a cover-glass the protoplasm may be expressed in the form of separate rounded vacuolated masses. Each of these shows the same osmotic properties as the original entire protoplast; i.e., shrinks in hypertonic solutions, resists the entrance of dyes (during the living state), and in general behaves as if it were surrounded with a semi-permeable membrane. Such an experiment might also be regarded as indicating the impermeability of all parts of the protoplast to dissolved substances. But it can be simulated by the simple experiment of breaking or cutting up a large drop of an insoluble liquid, such as chloroform, in a protein-containing medium; each resulting droplet remains separate and exhibits the same properties as the original droplet, and the effect can be shown to depend on the rapid formation of a thin protein adsorption-film at the surface of each newly formed droplet. In a somewhat similar manner, although the conditions are more complex, living protoplasm forms films at surfaces which are freshly exposed by cutting or other injury; this property is shown only during life and is presumably a manifestation of the normal property of construction and repair, which is dependent on metabolic synthesis, as we have seen. This is indicated by an experiment of Pfeffer's,^ in which root hairs are placed in

a weak solution of acid, which kills the cells. If then they are placed in a weakly hypotonic solution containing dye, the latter immediately enters as soon as the cell swells; i.e., the continuity of the plasma membrane is then no longer automatically maintained (as in living cells) when the membrane is stretched or ruptured. Presumably in living protoplasm both metabolic and purely physical factors (adsorption) take part in the formation of new surface-films. There is evidence that the protoplasmic surface-films undergo a change in their physical properties soon after they are formed; this is well shown in the freshly cut surfaces of sea-urchin eggs. Kiister found that the protoplasm of plant cells could be broken into fragments by strong plasmolysis; these fragments at first readily unite or cohere, but later lose this property.^ In the coalescence of fragments of inert inorganic material a similar behavior has been observed; freshly formed surfaces reunite readily, but not older surfaces ; apparently the progressive deposition or adsorption of foreign materials at the surfaces alters their properties and prevents fusion.^ There is also evidence that the formation of new surface-films plays an essential part in the normal return of irritable cells to the resting state after stimulation; during the period of recovery of excitability (refractory period) the cell-surface is apparently the seat of progressive changes of this kind (see below under refractory period).

Hober cites the case of myxomycetes possessing a clear surface hyaloplasm and a granular interior ;3 when ' For an account of these phenomena cf. Bancroft's Applied Colloid Chemistry, chap, v, on coalescence. the cell swells in hypotonic solution the thickness of the hyaloplasm remains constant; apparently this thickness is regulatively maintained by the transformation of material derived from the internal protoplasm. Chambers' observations on marine eggs show many interesting instances of the formation of films at cut surfaces or at the surfaces of protoplasmic fragments; regions where the protoplasm is broken down by mechanical or other injury soon become delimited by films bounding them from the adjoining unaltered protoplasm.^ In the formation of artificial vacuoles by the injection of solutions into egg cells through micro-pipettes, films with semi-permeable properties are formed about the introduced droplets.^ The composition of the salt solution is an important factor in the formation of such vacuoles; Chambers has recently shown that pure solutions of NaCl diffuse into the protoplasm without forming films, while if sufficient CaCla is present, each droplet of solution surrounds itself with a definite film and forms a vacuole. A recent study by Seifriz^ of the physical properties of protoplasm, as exhibited under micro-dissection, gives many interesting details on filmformation by living protoplasm under various conditions.

De Vries showed in 1885'' that the normal vacuoles of plant cells are surrounded by membranes having the same osmotic properties as the plasma membranes (those inclosing the entire protoplast). In his experimerits the cell was placed in a 10 per cent solution of KNO3 containing some eosin to serve as indicator of permeability; the outer protoplasm dies in one or two hours and becomes colored, but the vacuole remains at first clear and uncolored, showing impermeability of its membrane to the dye. Later the membrane becomes permeable to the dye, presumably as a result of deathchanges, and the vacuole contents become colored. Isolated vacuoles show osmotic properties similar to those of the whole protoplast. The observations of Kite and Chambers on artificial vacuoles in sea-urchin eggs illustrate the same phenomenon, the limiting surfacefilm of the vacuole having apparently the same properties as the surface-film of the entire cell.

The evidence just cited shows the error of regarding the properties of continuous protoplasm as identical with those of partitioned protoplasm. At boundary surfaces of whatever kind the living substance exhibits special properties, in particular a high resistance to the diffusion of water-soluble substances including ions. On the other hand, the internal protoplasm appears (at least in many cases) to be freely penetrable to diffusing substances of low molecular weight and to ions; hence it possesses a considerable electrical conductivity. This latter conclusion is of great importance for the theory of stimulation and conduction, as will be seen below.

As additional evidence for the permeability of continuous or unpartitioned protoplasm to water-soluble substances, we may cite the type of distribution shown by many diffusible substances in cells. Many such substances are concentrated in special regions or structures, in which they appear to be held in chemical or other combination; a good illustration is the water-soluble red pigment of Arbacia eggs, echinochrome, which is contained chiefly in minute round chromatophores scattered throughout the protoplasm. That this material is free to diffuse is shown, however, by the fact that cytolytic action, even temporary, causes its rapid diffusion into the surroundings; and. complete loss of pigment may thus result. The localization or condensation of the pigment in the chromatophores indicates apparently that it is present there in some form of loose chemical combination or adsorption; but in any such case an equilibrium between the adsorbed and the dissolved compound must exist (Cads = KCs{f). Presumably the dissolved portion of the pigment is homogeneously distributed throughout the protoplasm, hence the chromatophores adsorb (or combine) equal quantities and are similar in appearance. But any local decrease in the concentration of dissolved pigment, due to diffusion through the altered plasma membrane, disturbs the adsorption equilibrium and leads to the liberation of the adsorbed pigment, which may thus be completely lost from the cell. The conditions are similar in any other case of reversible adsorption.

Hermann's observations on the difference between the transverse and the longitudinal conductivity of nerve fibers also indicate the higher conductivity of nonpartitioned protoplasm as compared with partitioned. He found that the current encounters several times the resistance when the fibers are placed transversely between the electrodes, as compared with fibers arranged lengthwise, with equal distances between the electrodes and equal sectional areas of tissue. The same is true for

muscle.^ This result indicates again that the interposed plasma membranes offer a high resistance to the movement of ions, while the resistance of the continuous protoplasm is relatively slight.^ ♦ ^Hermann, Arch. ges. Physiol., V (1872), 223. The character of the intercellular partitions and their arrangement will, of course, affect the inner conductivity of the cell, but regarding these conditions we have little information at present. Overton's and later researches have shown that the permeabiHty of plasma membranes to lipoid-solvent substances is a universal property, that the permeability to water-soluble substances of low molecular weight, not soluble in organic solvents, is relatively very slight, and that the ability of substances to penetrate most forms of protoplasm has a close dependence on their relative solubility in the two classes of solvents; i.e., on their partition-ratios between organic (oil-like) solvents and water. The permeability for the two classes of substances, water-soluble and '^organo-soluble/' thus depends on different conditions. It is further significant that the permeabiHty for hpoid-soluble substances is m_uch less variable than that for the lipoid-insoluble substances and water; the latter form of permeability shows wide variations in the same cell under different physiological conditions, while the former appears to undergo little change. The characteristic semi-permeability of the living plasma membranes thus relates to the lipoidinsoluble group of substances; this fact, when considered in connection with the universal permeabiHty to the Hpoid-soluble group, suggests that the normal semi-permeabiHty depends on the presence in the membranes of water-insoluble compounds possessing the solvent properties of organic solvents. The characteristic water-insolubiHty of the surface-

layer of protoplasm is also intelKgible on this point of view. When we say that a cell or tissue is permeable to a given substance, we state merely the ability of the substance to penetrate; the method of penetration is not stated. Thus a dissolved substance may penetrate a partition by diffusing through the interspaces, or by dissolving in or combining chemically with the substance of the partition. In the case of ions there may be an apparent penetration resulting from a change in the electrical conditions at the surface of the partition. For example, when a current is passed by zinc electrodes through a bath of a zinc salt solution divided into two compartments by a zinc partition, it is not strictly correct to say that the zinc ions penetrate the partition. The effect is the same as if they did, but in reahty zinc ions are deionized and deposited as zinc on one face, while from the other face the metal passes into solution as zinc ions. There is apparently a penetration in such a case, but not in reality; and it is possible that under certain conditions the passage of a current through a plasma membrane may similarly depend on the chemical combination of ions on the one face simultaneously with their release from the other. The physical and chemical conditions of these various types of permeability require careful examination.

The factors determining the penetration of dissolved substances through living plasma membranes are various, and for the purpose of the present discussion they may be grouped under the following heads: (i) structural conditions, including thickness, density (water-content), state of dispersion of structural colloids, size of interstices or pores, (2) chemical conditions, including special composition and solvent properties of constituents, (3) physical conditions of other kinds, such as electrical polarization at the faces of the partition or other surfaces, electrical endosmose effects, filtration effects, and (4) factors dependent on the Hving condition of the membrane, including variation of permeabiHty due to metabolic or other changes; the last-named factors are apparently the ones chiefly responsible for the active types of material transport in and through cells in absorption and secretion (factors of ''physiological

A complete discussion of these various factors need not be attempted here, especially since the whole subject of permeability has recently been carefully reviewed in the well-known textbooks of Bayliss and Hober. Certain aspects of the problem of permeability require special consideration, however, especially the dependence of permeabiHty on physical conditions, and the nature of the factors controlling the normal or physiological variations of permeability. The most characteristic pecuHarity of the Kving protoplasmic membranes, especially of irritable cells, is that their properties vary with both the external and the internal conditions, and are subject to regulative control of a highly definite kind. Such properties cannot be derived from any simple static type of structure, or explained on the basis of the pecuUarities of the colloidal compounds composing the membrane. We must recognize that the distinctively vital factors, those depending on the metaboHc activity of the cell, are probably the most important of all, and they are the least understood at present. The simpler

physico-chemical or nonvital factors can, however, be shown in some cases to have definite relations to certain characteristic physiological effects; e.g., in the action of salts and lipoid-solvent compounds on Hving cells. These are the substances through whose action the physiological properties and activities of cells may most readily be altered in a reversible manner; and this action is referable in many cases to a direct alteration of the plasma membranes.

Semi-permeability requires a certain closeness of physical texture; i.e., density of structural material; apparently also the structure-forming substances must be water-insoluble. In all semi-permeable artificial membranes water-insolubility is essential; the most perfect examples are the precipitation-membranes of the ferrocyanides and other insoluble salts of heavy metals. That the chief compounds composing the plasma membranes are also water-insoluble is shown by the characteristic insolubihty of Hving cells in their normal aqueous media.

The relation between structural density and permeability in artificial membranes is well shown in the ''ultrafilters" devised by Bechhold for the^ separation of various colloids from their suspension-media.^ These filters consist of disks of gelatine hardened in formalin. Bechhold found that the higher the colloid-content of these partitions, the more numerous were the colloidal sub- ^ Bechhold, Colloids in Biology and Medicine, EngUsh translation, or third Gennan edition, 1920.

stances which were held back in filtration experiments; for example, disks of 10 per cent content prevented the passage of nearly all the colloids employed. From such facts we should expect that a still higher degree of impermeability (to crystalloids) would require a higher degree of structural density; i.e., lower water-content. Experience with artificial semi-permeable membranes of copper ferrocyanide bears out this expectation; this is well seen in Morse's studies of these membranes, described in his book, Aqueous Solutions.^ He found that for the formation of good semi-permeable membranes an extremely fine porosity in the supporting porcelain cells was necessary; '^ excessive fineness of texture is absolutely indispensable to the correct measurement of osmotic pressure" (p. 15). A good semi-permeable membrane is thus essentially a fine-textured structure of water-insoluble material; hence it resists the passage of water as well as of dissolved substances. Morse also considers that the colloidal character of the precipitate is an important factor in semi-permeabihty. Even under the best conditions a high degree of semi-permeability is difficult to obtain with precipitation-membranes, and this property is subject to change; it is affected by temperature (cf. pp. 85-86) and especially by electrolytes (cf. pp. 91-92) which cause rapid deterioration in the membrane. LiCl was the least harmful of the electrolytes investigated (cf. p. 214) and Morse gives determinations of the osmotic pressure of this salt.

Since fineness of texture is favorable to semi-permeabihty (i.e., to the production of osmotic effects), one might expect that any insoluble porous partition would ^ Morse, Aqueous Solutions, Carnegie Institute Publications (1914). show osmotic action (semi-permeability) if only its pores were sufficiently minute. In fact, Thomas Graham observed many years ago (1854) that certain fine-grained porcelains had definite osmotic action. Recently the pore-diameters of osmotically acting porcelain disks have been investigated by Bigelow and Bartell/ and well-marked osmotic effects were found when the pores had a diameter of the order o. 2-0.3 5/x.^

In general we may conclude from such facts that in living semi-permeable membranes the degree of porosity is low; i.e., the interspaces between the water-insoluble colloidal particles forming the structural material are at least as small as in porcelain and probably smaller, since Bartell's membranes are not completely semipermeable. The physical conditions are probably closely comparable with those existing in well-supported precipitation-membranes of copper ferrocyanide, such as Morse employed. In plasma membranes, however, variations in the subdivision of the colloidal constituents may cause variations in the size of the interspaces, and hence in the permeability. The progressive deterioration to which artificial colloidal membranes are subject in the presence of electrolytes is apparently prevented in Hving membranes by compensatory factors dependent on metaboKsm; presumably any interruptions of continuity are at once automatically repaired by the formation or deposition of new structural material (see below under stimulation) .

Such finetextured membranes must resist the passage of water as well as of dissolved substance. In Morse's experiments many days were often required to reach osmotic equihbrium. There is also evidence that many living plasma membranes offer a high resistance to the passage of water: i.e., are relatively water-impermeable. This statement may seem surprising in view of the fact that the passage of water into and out of living cells in anisotonic solutions is usually rapid; but the membranes are extremely thin and the ratio of the surface to the inclosed volume is very large, so that relatively rapid entrance of water is quite consistent with a very low specific permeability to the Hquid.

The loss of semi-permeability at death is associated with an increase of permeabihty to water as well as to dissolve substances. This is shown in some experiments of Bernstein who, with another problem in mind (the conditions under which water is held in cells) , determined the relative rates of evaporation of water from living and dead tissues.^ Bernstein used the method (introduced by Liebig) of measuring the ''force of imbibition" of membranes. A tube widened at one end (thistle tube) is fixed vertically with its narrow end dipping into mercury; the tube is completely filled with water and the upper expanded end is closed by the membrane under examination. As water evaporates through the membrane the mercury rises in the tube, showing the development of a pressure; the rate of evaporation through the membrane is thus indicated. Bernstein used fine tubes ending above in equally sized funnels which were closed (A) with living membrane and (B) with dead membrane