Membrane Potentials in the Muscles of the South American Frog, Leptodactylus Ocelatus

The potentials were measured and recorded by using glass microelectrodes (tip diameter 1–2μ) filled with 3 M-KCl (Ling & Gerard, 1949) and connected to a high impedance cathode follower unit and a direct coupled amplifier. The preamplifier and input stages were similar to those described by Bishop & Harris (1950). An oscilloscope was used when it was desired to record action potentials (A.P.); the resting potentials (R.P.) were read off a meter connected across the pre-amplifier output. A source in series with the input permitted calibration to be carried out with the electrode in situ. Readings quoted are the mean of at least six penetrations, the S.D. is given or indicated on the figure.

The sartorius muscles of L. ocelatus were dissected from small specimens, muscles weighing over 200 mg. being unsuitable because they soon deteriorate, probably on account of inadequate oxygenation. The sartorius was attached to a piece of thread at each end, and to one end was tied a short piece of rubber thread, this enabling the muscle to be held by means of clips with slight stretch across a Petri dish containing the saline solution.

The basic solution employed was of the following composition: Na 120, K 2, Ca 1, Mg 1, Cl 96, HCO₃ 30, HPO₄ 1, mM./l. When the K concentration increased this was effected by adding KC1 to a constant total volume because this reduces or prevents swelling (Boyle & Conway, 1941). Other solutions are specified in the text. The solution was stirred by bubbling with oxygen + 5 % CO₂.

The muscles when freshly dissected were put in the solution and observed microscopically in order to see the penetration of the micro-electrode. It was found that on the inner surface of the Sartorius there was a layer of transparent gelatinous material, a thread of which could sometimes be drawn up by the microelectrode out of the solution. The thickness of the layer seemed to increase with time, and it is likely that it swells in the saline solution, thus contributing an increasing fraction to the extracellular volume of the muscle. The presence of this coat probably accounts for the fact that after the composition of the solution was changed more than 10 min. was required for the potential to become stable even though the microelectrodes were put into the outer layer of fibres to which the solution has most ready access. The mean diameter of the fibres in the living muscle was about 80 μ; this was based on a count of the number of fibres seen side by side across the calibration of a 2 mm. graduated eyepiece. The fibres appeared to run the whole length of the muscle (2–3 cm.).

A series of measurements was made to compare the potential when the muscle was in the standard solution with the potential in a solution containing more (or less) than the standard concentration of K. Fig. 1 shows the results obtained for three muscles whose R.P. in 2 m-equiv./l. solution was too mV. within 2 mV. Other muscles showed R.P.’S in this solution varying from specimen to specimen, though muscles from a particular frog were very similar. The range of values from various frogs in the 2 mM. solution was from 85 to 103 mV. Readings were taken at 23–27° C. In an experiment in which the temperature was reduced to 6° C. the reduction of potential was less than the standard deviation. In one specimen the potential rose as the potassium concentration was reduced below 1 mM., but in three others it remained nearly the same as it had been in the 1 mM. solution. The form of the membrane potential/K concentration curve is similar to the one reported by Jenerick (1953) for the muscles of Rana pipiens. It is noteworthy that in a saline mixture having 147 mM. of K there is still a potential of about. 16 mV.

A series of measurements was made using mixtures of 0·1 M-Na and K phosphates (each one having a composition intermediate between mono- and di-basic salts). The muscles twitched in these mixtures if less than 10 mM. K was present. Potentials obtained were always lower than those found in the same muscle in a saline mixture of K concentration equal to that of the phosphate mixture (>99% significant by t test). In pure K phosphate mixture (K= 150 mM.) the potential was 4–6 mV., and in the dibasic salt (K = 200 mm.) it was difficult to tell when a penetration had been made ; the membrane potential was certainly less than 4 mV.

In pure KC1 the potential was lower than in a saline mixture containing the same K concentration, in 147 mM.KCl the value was 8 ± 1 mV. ; in 300 mM. KCl it was between o and 4 mV. In pure KHCO₃ (170 mM.) a value of 8 ± 2 mV. was obtained.

A comparison was made of the R.P. in three different chloride-bicarbonate mixtures (Table 1). The K concentration was 2 mM. in each. The figures suggest that the R.P. may increase with the proportion of HCO₃ and the consequent increase of pH. More alkaline mixtures could not be used because the muscle twitched spontaneously.

The potentials in the phosphate mixture, as already stated, were lower than in the saline mixture of the same K concentration.

Attempts were made to replace Cl by other anions. Bromide could be used giving a R.P. = 88·5 ± 2 mV., the control in the usual chloride mixture was 90 ± 5 mV. The bromide solution did not cause continuous spontaneous activity. It was noticed that on penetrating a fibre in the bromide solution there would be a local contracture which disappeared in a few minutes. This could not be attributed to gross damage since the same electrode would later give high and steady R.P.’S. Stimulation in the bromide solution caused long-lasting contractions.

Both nitrate and sulphate if used to replace all the Cl (but leaving the HCO₃) led to spontaneous twitching. Potential readings were therefore taken in mixtures in which half the Cl was replaced by its equivalent of either nitrate or sulphate. The results of two muscles (a) and (b) were, in normal saline, (a) = 87 ± 3 mV., (b) 96 ± 1·7 mV ; in half nitrate, (a) was 83 ± 5·5 mV. (b) was 82 ± 1 mV., and in half sulphate, (a) was 87 ± 4·5 mV. Thus sulphate did not appear to reduce the R.P. while nitrate did so. This latter effect was reversible.

When all the Na was replaced by equivalent tetraethylammonium there was no immediate reduction of the resting potential, but this was almost certainly due to the slow loss of Na from the surroundings of the fibres. Action potentials observed during the equilibration period (i.e. before all the Na can have been removed) lasted 10 times as long as the normals. If the muscle was kept in the solution for more than 15–20 min. the R.P. began slowly to fall. It was reduced by about 10 mV. in the next 20 min. These observations lead one to the conclusion that if time is given for most of the muscle Na to be carried away by the Na-free solution the R.P. is reduced. Probably its eventual rate of fall will be determined by the rate of K leakage from the muscle which was not determined in this solution. Replacement of all the Na by guanidine also caused the R.P. to fall ; the rate of fall was greater than that caused by the tetraethylammonium.

Tubocurarine in concentration 6 × 10⁻⁶ stopped nerve stimulation being effective after a delay of 10 min. It did not affect the R.P., and indirect excitability could readily be restored by washing in the ordinary saline mixture.

The effect of two metabolic poisons was investigated. Addition of dinitrophenol at a concentration of 25 mg./l. only led to an insignificant fall of R.P. in $112$ hr. from 87 ± 2·5 to 84·5 ± 2 mV. Replacing 1/10 of the chloride by fluoride prevented indirect excitation but did not affect the R.P.

Observations were made of the A.P. resulting from indirect stimulation. The amplitude of the ‘overshoot’ was 35–50 mV. Fig. 2 illustrates some responses. In Ringer’s solution the greater the reverse membrane potential the faster was the rate of fall of the spike. In nitrate mixture (half Cl replaced by nitrate) the rate of fall was much reduced even when the maximum of the spike was high, while the rate of rise was not visibly affected. In bromide the action potential appeared to be sometimes lengthened as in nitrate, but in addition there was an indication of repetitive firing. These effects are being further investigated in the species Rana temporaria. Added in proof: In R. temporaria the slow rate of fall has sometimes been seen in ordinary Ringer, so it cannot be regarded as a specific result of nitrate treatment.)

Samples of plasma were analysed for Na, K and Cl. Values obtained were (in μ-equiv./g. plasma): Na 122, K 4·8 (mean of 6, range 2·1–6·7), Cl 90·5 (mean of 3). The variability of K content may be due to variations in the time since the last meal or to the presence of various parasites.

Na and K contents, dry weight, and Cl and SO₄ spaces were estimated on the muscles. The dry weight was 19·2 ± 0·9% of the wet weight. The K content of freshly dissected, unsoaked muscles was rather variable, ranging from 99 to 85μ-equiv./g., and the Na content ranged from 37 to 32μ-equiv./g. The Cl in freshly dissected muscle corresponded to a Cl space of 32·3%, which exceeds the Na space (26·2–30·3% based on the above plasma and muscle Na values). The excess Cl may be explained either by presence of intracellular Cl or by combination of Cl with basic extracellular material. The latter is favoured by the fact that the SO₄ space of soaked muscles was also rather high. If we take for calculation a dry weight of 19·2% and an extracellular space of 30% the internal K concentrations would be between 167 and 195μ-equiv./g. cell water in the fresh muscle.

Measurements of the space accessible to Cl and SO₄ were made in muscles exposed to either the normal saline mixture, or to one having half the Cl replaced by SO₄ for periods of 1 hr. The Cl space was 33·2%, i.e. a little greater than the fresh muscle value which was obtained on specimens taken from the same frog. The sulphate space, determined by chemical estimation of the SO₄ liberated from a previously soaked muscle, ranged from 28 to 37% (mean of 11 = 32·8%). It is unlikely that SO₄ enters the fibres because muscles do not swell in isotonic potassium sulphate solution.

The K-content of the soaked muscles was a little lower than that of the fresh muscles, it ranged from 72 to 92μ-equiv./g. (mean of 14=83·9μ-equiv./g.). The corresponding internal K concentration comes to 176μ-equiv./g.

Experiments, made in parallel with the potential measurements, on the kinetics of the loss of radio Na from previously treated muscles also permitted estimation of the space occupied by the fast-moving component of the Na according to the method used by Harris & Bum (1949) as well as of the internal, or more slowly moving Na. The extracellular space by this method was 27–35 % and the internal Na concentration came to about 7μ-equiv./g. cell water.

In a pure KC1 solution and before there had been time for appreciable penetration of K along with Cl the application of the equation is simplified because Na_e is zero and r₁Na_i proves to be negligible. To give 8 mV. in 147 HIM. KC1 r₂ comes to 0·17, the value is little affected whether we take [Cl_i] as 1 or 10 mM. Turning now to the saline mixtures a good agreement with experiment can be obtained if one takes r₁ as 0·015 ; the most doubtful figure in making the calculation is the Cl_i, which has been assumed constant at 1 mM. The magnitude of this latter figure somewhat affects r₁ and it has to be supposed that [CLJ does not increase during the time necessary for the K ions of the external solution to equilibrate with the solution in the immediate vicinity of the fibres. Jenerick (1953) made this same assumption. The value for r₁ fits the results obtained in the phosphate mixtures, in which the Cl ions do not present any problem. It is worth remarking that the differences of R.P. found between various muscles in a given solution must arise from differences of r₁rather than of Na_i and K_i since the value of the fraction is much more sensitive to changes in r₁ than to changes in Na_i and K_i.

The values of r₁ r₂ found for Leptodactylus do not differ greatly from those found by Jenerick (1953) for Rana pipiens (he found r₁ = 0 ·027 and r₂ = 0 ·23).

The effect of the anions on r₁, r₂ is of interest. Can the permeability ratios be changed by changes of solution composition? Nitrate does appear to reduce the resting potential, which might be explicable by a changed r₁ the reduction of the [Cl_e] term being insufficient to account for the observed change. Br and SO₄ ions do not appear to alter r₁ in the resting state.

The prolongation of the A.P. seen in nitrate solution could be explicable by a change of the permeability ratios applying in the active state, thus if the K permeability were reduced it would take longer for the polarization to be restored by an outward K current. Present preliminary results of permeability studies do not indicate that the resting state permeabilities are affected. Here it may be noted that the recent observation of Carey & Conway (1954) that there is less entry of K from a sulphate mixture than from a chloride mixture into R. temporaria muscles does not show clearly that sulphate depresses K permeability. The observations shown in their fig. 5A and B are confined to a period of 100 min. and only indicate that the initial rapid entry of K amounts to exchange of about twice as much alkali ion when in the chloride as it does when in the sulphate. Tangents at 80 min. do not show greatly different slopes.

An influence of Br and NO₃ ions in lengthening the duration of the active state of muscle contraction has been described by Sandow & Mauriello (1953) and by Hill & Macpherson (1954). Nitrate has also been investigated by Ritchie (1954). While our observation of the slowing effect of nitrate on the recovery from the action potential may be related to the effect on the mechanical response, it does not suffice to account for the very much longer duration of the mechanical response in the nitrate medium.

1. The resting potential of the fibres of the South American frog Leptodactylus ocelatus has been measured using microelectrodes.

2. In 2 mM. potassium Ringer solution values of 85–103 mV. were obtained. The value for a given muscle depends upon the external K concentration and the relation between resting potential and the concentrations of K, Na and Cl in the medium can be fitted to the constant field solution of the diffusion equation.

3. Use of phosphate solutions in place of chloride solutions led to lower resting potentials at a given K concentration, which is understandable since the contribution of Cl ions to the resting potential is not present in the phosphate solution.

4. The action potential was observed. Its rate of fall was less when the positive overshoot was low than when the overshoot was high.

5. Analyses of muscles and plasma were made.

This work was carried out in the tenure of a Technical Assistance Contract from U.N.E.S.C.O. by one of us (E. J. H.). We have to thank Miss K. Nishie for help in the analyses. We also wish to record our thanks to Prof. B. Katz, F.R.S., for his suggestions made in the preparation of the paper.