«THE RUBIDIUM AND POTASSIUM PERMEABILITY OF FROG MUSCLE MEMBRANE BY R. H. ADRIAN From the Physiological Laboratory, University of Cambridge (Received ...»
The reason for the poor survival is unknown. Because of the dubious state of the muscles soaked for long periods in the potassium solution, freshly dissected muscles have been compared with the 4-5 day isolated muscles 140 R. H. ADRIAN containing rubidium. Where such comparisons are made, they include any effects that can be attributed to long isolation by itself.
Figure 2 also shows the rate at which potassium replaces internal rubidium when muscles soaked for 48 hr in the 100 m-equiv/l. Rb + sulphate solution are transferred to the 100 m-equiv/l. K+ sulphate solution.
Potassium from external solution replaces internal rubidium at about 4 times the rate that internal potassium is replaced by rubidium when muscles are first put into the rubidium solution. Muscles soaked for 5 days in the 100 m-equiv/l. Rb+ sulphate solution, when transferred to the 100 m-equiv/l. K + sulphate solution have an internal potential of + 8 + 0 4 mV ( S.E. of mean; 12 fibres, 3 muscles).
0 20 40 60 80 100 Hours Fig. 2. The filled symbols show the time course of the replacement of potassium by rubidium, for muscles which had previously been soaked for 48 hr in a sulphate solution containing 100 m-equiv/l. K. The open symbols show the time course of the replacement of internal rubidium by potassium for muscles which had been previously soaked for 48 hr in a sulphate solution containing 100 m-equiv/l. Rb.
Triangles: internal rubidium concentrations. Circles: internal potassium concentrations. The concentrations are given as in-mole/kg fibre water, and the internal sodium concentration of none of the muscles was greater than 20 rn-mole/kg fibre water. The mean internal sodium concentration of all the muscles in this figure was 12 rn-mole/kg fibre water.
Recovery of muscles containing rubidium. When muscles which have been in the 100 m-equiv/l. Rb+ sulphate solution for up to 5 days, are put into a normal Ringer's solution (A) or a Ringer's solution made with 2a5 m-mole/l. RbCl (B) they recover the ability to twitch within 2 min.
Rb AND K IN MUSCLE 141 Resting and action potentials were measured within 30 min of the transfer to the Rb-Ringer's solution. The mean resting potential was - 88-6 + 1-2 mV and the mean size of the action potential was 131-5 + 1-5 mV (+ S.E.
of mean; 47 fibres, 9 muscles). The average intracellular cation concentrations of eight of these nine muscles were: K+, 22 m-mole/kg fibre water;
Rb+, 126 m-mole/kg fibre water; Na+, 10 m-mole/kg fibre water. Freshly
dissected muscles for comparison gave mean values for the resting potential of - 93-0 + 1-6 mV, and for the action potential of 123 + 1-1 mV ( ± S.E.
of mean; 35 fibres, 4 muscles). The average intracellular cation concentrations of these four muscles were: K+, 146 m-mole/kg fibre water; Na+, 16 m-mole/kg fibre water. Figure 3 shows action potentials from two freshly dissected muscles and two muscles with a high internal rubidium content.
After about an hour in the Rb-Ringer's solution muscles with a high internal rubidium concentration become inexcitable. Fibres in this B. H. ADRIAN 142 condition have an internal potential of between -60 and -30 mV. Even within the first hour in the Rb-Ringer's solution there is some decline in the size of the action potential and the resting potential. Resting and action potentials were measured in four muscles during the second halfhour in the Rb-Ringer's solution. Action potentials with a mean value of 96-8 + 4-2 mV were recorded from ten fibres whose mean resting potential was - 75-4 + 2-2 mV; seventeen fibres could not be excitedtheir mean resting potential was -66 + 1F6 mV (S.E. of mean). The
average intracellular cation concentrations of these four muscles were:
Rb+, 130 m-mole/kg fibre water; K+, 21 m-mole/kg fibre water; Na+, 11 m-mole/kg fibre water. The falls in resting potential and inexcitability do not appear to be associated with marked changes in the intracellular cation concentrations.
When muscles, after 5 days in the 100 m-equiv/l. Rb + sulphate solution, are transferred to a sulphate solution with a rubidium concentration of
2.5 m-equiv/l., only a small repolarization takes place. Such a muscle, with an internal potential of -10 mV in the 100 m-equiv/l. Rb+ sulphate solution, was transferred to solution C (Table 1). After 15 min the internal potential of most of the fibres was about -35 mV. There was an unusually large scatter in the values; the largest recorded was -53 mV. The muscle was then put into the Rb-Ringer's solution (B, Table 1). Measurements were made on four fibres within the first 5 min in the Rb-Ringer's solution.
The average resting potential was -90 mV. All four fibres were excitable;
the average action potential was 124 mV. It appears that a muscle fibre with rubidium on both sides of the membrane cannot achieve a resting potential of -90 mV in the 2-5 m-equiv/l. Rb + sulphate solution, nor can it maintain such a potential in the Rb-Ringer's solution for more than about half an hour. If the rubidium permeability were only about 10 times the sodium permeability, the temporary presence of a large resting potential in the Rb-Ringer's solution would be due to the low internal chloride concentration which results from soaking in a sulphate solution.
The rate of repolarization of the action potential. Figure 3 does not show any conspicuous differences between the action potentials from freshly dissected muscles and muscles containing a high concentration ofrubidium.
In particular the rate of change of potential during repolarization is not obviously different. Baker, Hodgkin & Shaw (1962) have shown that the repolarization of the action potential of a perfused squid axon is markedly slowed when the internal isotonic K2SO4 is replaced by isotonic Rb2SO4.
Frog muscle appears to behave differently in this respect, though it should be remembered that while the replacement of potassium by rubidium was very close to 100 % in the squid axon, the mean replacement in the muscle was only about 85 %.
Rb AND K IN MUSCLE 143 Measurement of the rate of repolarization in Rb-containing muscles suggests that the repolarization is, in fact, somewhat slower than in freshly dissected muscles. Rates of repolarization were measured by electrical differentiation of the action potential. The average rate of repolarization of freshly dissected muscles was 84 + 0 9 V/sec (26 fibres, 2 muscles at 200 C, S.E. of mean: cf. 86 + 2 V/sec at 17.40 C, Nastuk & Hodgkin, 1950).
The rate of repolarization of the Rb-containing muscles was 60 + 1.0 V/sec (17 fibres, 2 muscles at 200 C). It is difficult to attach precise significance to the difference because neither the resting potential nor the action potential were exactly the same in the freshly dissected and Rb-containing muscles (see p. 141). Despite the uncertainties the results suggest that, if the repolarization of the action potential is normally brought about 'by an outward movement of potassium, the mechanism does not discriminate to any great degree betwen potassium and rubidium.
Conductance changes with hyperpolarization. When a constant inward current is passed across the membrane of a muscle fibre in Ringer's fluid, the electrotonic potential does not reach a steady level for about a second.
This behaviour is shown in the photographic record of Fig. 4A. This record was obtained by the method described by Adrian & Freygang (1962a). The lower trace represents the change in membrane potential produced by a current which is, to a reasonable approximation, proportional to the deflexion of the upper trace. The membrane conductance at each instant is therefore proportional to the ratio of the deflexions of the two traces at that time. The conductance was measured from the records 100 msec from the beginning, and at the end of each of a series of hyperpolarizations produced by current pulses of increasing size. In Fig. 4A the conductance measurements at 100 msec are plotted against the internal potential at 100 msec, and the conductance measurements at the end of the pulse of current are plotted against the internal potential at the end of the pulse. Depolarizing pulses were not used because of the presence of sodium in the external solution. The membrane conductance increases with increasing hyperpolarization both at short and at long times, though during each current pulse the membrane conductance falls to a final steady level. This behaviour has been described in detail by Adrian & Freygang (1962a). They concluded that the conductance changes with voltage and time are due mainly to changes in the potassium conductance, and that the chloride conductance, which makes up about 70 % of the resting conductance in Ringer's solution, falls slightly with hyperpolarization. The changes of conductance with time were tentatively explained in terms of a model with two membranes in series; the outer permeable to sodium and potassium, the inner permeable to potassium only and able to pass large inward but only very small outward currents.
144 R H. A:DRIAN The photographic record and graph of conductance against internal potential in Fig. 4B were obtained by the same method as Fig. 4A. The difference between the experiments in Fig. 4A and Fig. 4B is that the potassium inside and outside the muscle fibre has been replaced by rubidium. In Fig. 4B the muscle is in Rb-Ringer's solution (solution B, Table 1) and has an internal rubidium concentration of 126 m-mole/kg fibre water and an internal potassium concentration of only 21 m-mole/kg fibre water. Though the muscle twitched and propagated an apparently
-10 B A I sec tw
normal action potential, the membrane conductance fell slightly with increasing hyperpolarization and rose rather than fell in the course of each current pulse. Adrian & Freygang (1962a) calculated the contribution of the chloride conductance to AV/V %. Their experimental arrangements were identical to those of the present experiments. For an 80,u fibre with a chloride permeability of 1.9 x 10-6 cm/sec, and no other ion able to carry current, the assumptions of the constant field equation suggest that AV/V % would be 2-8 % for a hyperpolarization of 10 mV, and 2-2 % for Rb AND K IN MUSCLE 145 a hyperpolarization of 40 mV. These values are very close to those shown in Fig. 4B. It is possible that when rubidium replaces potassium on both sides of the membrane, very little current is carried by rubidium and almost all the current is carried by chloride. The results of the experiment shown in Fig. 4 B are therefore consistent with the suggestion put forward to explain the failure of a Rb-containing muscle to maintain a high resting potential in a Rb-Ringer's solution (p. 140).
3 2 1
Figure 5 shows the conductance plotted against the internal potential for a freshly dissected muscle in a Rb-Ringer's solution. For inward currents, which will be carried by an outward movement of chloride and an inward movement of rubidium, the total conductance declines slightly with increasing hyperpolarization. In the course of each current pulse the conductance rises as it does in Fig. 4 B. The cause of this rise is not known;
it is not due to chloride from the current-passing electrode because it is still present when the current-passing electrode is filled with 2 M potassium Physiol. 175 10 R. H. ADRIAN 146 citrate. It has been argued above that the rubidium-filled fibre fails to maintain a high resting potential in Rb-Ringer's solution because the rubidium permeability of the membrane is very low. If this is the case the greater part of the conductance of the fibres shown in Figs. 4B and 5 must be due to the movement of chloride ions.
It is known that the chloride transport number of the membrane at the resting potential is between 0*6 and 0 7 (Hodgkin & Horowicz, 1959;
Hutter & Noble, 1960). An attempt was therefore made to measure the
membrane conductance of the same fibre (containing potassium) successively in K-Ringer's solution, Rb-Ringer's solution, and again in KRinger's solution. It did not prove possible to do this experiment without a slow decline in the resting potential during the time that the electrodes were in the fibre (13 min). Figure 6 plots the conductance as a function of the internal potential, measured in K-Ringer's solution initially (open circles), in Rb-Ringer's solution (open triangles), and then once more in K-Ringer's solution (filled circles). The mean potential in each solution Rb AND K IN MUJSCLE 147 in the absence of externally applied current is marked by a short vertical line. In both sets of measurements in K-Ringer's solution, the conductance, measured early in the course of the current, rose with hyperpolarization. But in the Rb-Ringer's solution the conductance remained constant.
At an internal potential of -85 mV (the estimated initial resting potential of this fibre) the conductance in Rb-Ringer's solution is about two-thirds of the conductance in the K-Ringer's solution. This result strongly suggests 6(/% ~~~~~5 0
that in Rb-Ringer's solution the inward movement of rubidium at any particular membrane potential is much less than the inward movement of potassium from a K-Ringer's solution at the same membrane potential.
Figure 7 shows the characteristic behaviour of the membrane conductance of fibres, filled with rubidium, but in a K-Ringer's solution. The conductance does rise with increasing hyperpolarizing currents, in these circumstances carried by an inward movement of potassium and an outward movement of chloride, but it rises only a small amount compared to 10-2 148 R. H. ADRIAN the conductance rise in K-Ringer's solution when the internal cation is potassium (cf. Figs. 6, 4A). The electrotonic potential also falls slightly in the course of a long current pulse.
Membrane current in sulphate solutions with 100 m-equiv/l. K+ or 100 m-equiv/l. Rb + The ability of a muscle fibre to pass large inward, but only small outward currents is clearly seen when the conductance is measured in an isotonic K2S04 solution (Katz, 1949) or an isotonic sulphate solution with a potassium concentration of 100 m-equiv/l. The curve relating membrane