Complete recovery takes mi. depolarization. These results are consistent with depolarization serving

Transcription

1 J. Physiol. (1973), 231, pp With 12 text-figures Printed in Great Britain CALCIUM ENTRY IN RESPONSE TO MAINTAINED DEPOLARIZATION OF SQUID AXONS BY P. F. BAKER, H. MEVES AND E. B. RIDGWAY* From the Laboratory of the Marine Biological Association, Plymouth, and the Physiological Laboratory, University of Cambridge, Cambridge CB2 3EG (Received 17 January 1973) SUMMARY 1. Intracellular aequorin was used to monitor changes in Ca entry in response to maintained depolarization either produced electrically or by exposure to K-rich solutions. 2. External K concentrations greater than 50 mm produce a phasic light response. The light rises to a peak in a few sec and then falls in 0-S- 5 min to a new steady level that is always greater than the level in the absence of K. 3. The phasic light response does not result from depletion of available aequorin at the periphery of the axon, but rather seems to reflect a phasic entry of Ca in response to depolarization. 4. Similar phasic responses are produced by prolonged electrical depolarization. These results are consistent with depolarization serving both to activate and also to inactivate Ca entry. 5. Following inactivation and after return to normal sea-water, there is an appreciable relative refractory period during which the response both to K-rich sea-water and electrical depolarization is reduced in size. Complete recovery takes mi. 6. The response to 410 mm-kcl is dependent on the previous treatment of the preparation. Pre-treatment with 100 or 200 mm-kcl reduced the response to 410 mm-kcl. The potential for half inactivation was about -25 mv in 112 mm-ca and -40 mv in 20 mm-ca. 7. The rate of onset of inactivation is potential dependent and is faster for depolarizations to zero potential than for smaller ones. 8. The phasic Ca entry produced by K-rich solutions is insensitive to external tetrodotoxin and internal tetraethylammonium ions, but is blocked by external Mn2+, Co2+ and Ni2+ ions and by the drugs D-600 and iproveratril. This suggests that the phasic Ca entry involves the late Ca channel. * Present address: Department of Physiology, Medical College of Virginia, Richmond, Virginia 23298, U.S.A.

2 528 P. F. BAKER, H. MEVES AND E. B. RIDGWA Y 9. Recovery of the outward K current after a long depolarization is much faster than recovery of the late Ca entry system. This provides further support for the view that the late Ca channel and the K channel are distinct. IIn-FODUCTION The Ca entry that follows depolarization of squid axons is made up of two components (Baker, Hodgkin & Ridgway, 1971). An early component that is blocked by tetrodotoxin and seems to reflect Ca entering through the Na-permeability channels and a late component that is insensitive to external tetrodotoxin and internal tetraethylammonium ions and seems, on pharmacological grounds, to be quite distinct from the K-permeability system (Baker, Meves & Ridgway, 1973). This late Ca channel is blocked by externally applied Mn2+, Co2+ and Ni2+ ions and by the organic Ca antagonists D-600 and iproveratril. The component of Ca entry that is sensitive to tetrodotoxin only persists for a very short time - less than 500 jsec for a depolarization of 80 mv from the resting potential at 220 C (see Fig. 18, Baker, Hodgkin & Ridgway, 1971), presumably because the early Ca entry is blocked by inactivation of the Na channel (Hodgkin & Huxley, 1952). This contrasts with the behaviour of the late Ca channel which, once opened, continues to allow Ca ions to enter at a steady rate for at least 100 msec at 22 C (see Fig. 20, Baker, Hodgkin & Ridgway, 1971). The purpose of this paper is to explore the behaviour of the late Ca channel in response to much longer depolarizations, either produced electrically or by exposure to KCl. The main conclusion is that prolonged depolarization ultimately leads to closure or inactivation of the late Ca channel. Part of this work has already been described briefly (Baker, Meves & Ridgway, 1971). METHODS Material. Giant axons gum in diameter were dissected from mantles of Loligo forbesi. In general refrigerated mantles were used, but in a few experiments the axons were obtained from freshly killed squid. Aequorin procedures. The techniques used were essentially those described by Baker, Hodgkin & Ridgway (1971) and Baker et al. (1973). The flow technique for changing the external solution while still monitoring the light output from the axon was modified to allow solution changes to be made fairly rapidly around axons impaled with an internal electrode. The flow cell was a 5-5 ml. cuvette. Solution was supplied to the bottom of the cuvette through a light-tight seal from a syringe outside the box housing the axon and photomultiplier, and excess solution was sucked off from the top of the cuvette. Axons were impaled in the normal way and transferred to the flow cell. A complete change of external solution was achieved by flowing 50 ml. through the chamber at a flow rate of about 2 ml./sec. The light output

3 INACTIVATION OF Ca ENTRY 529 measured in terms of photomultiplier anode current was recorded on a pen writer that had a response time of 0-5 sec for a full scale deflexion. Solutions. The solutions used were mainly high-ca artificial sea-waters (ASW) of composition: 112 mm-cacl2, 0 mm-mgcl2, 2-5 mm-nahco3, 10 mi-kcl and 400 mm- XC1 where X is the major external cation, Na, K, Li or choline. Sea-waters of different K contents were made by mixing high-ca K-ASW with high-ca Na or choline-asw. Unless otherwise stated, periods of exposure to K-rich sea-water were separated by recovery periods of at least 10 min in choline or Na-ASW (the reason for this is discussed on page 539). The inhibitors Mn2+ and D-600 were prepared as described by Baker, Meves & Ridgway (1973). Electrical measurement.. The resting potential was determined by inserting a fine glass capillary ( jtm diameter) containing 0-6 M-KC1 into the fibre through the cannula. The potential difference was measured between this capillary and an indifferent Ag-AgCl electrode in a cup with 0-6 M-KCl which was connected to the external bath by a bridge filled with an agar gel of ASW. The resting potentials were corrected for the liquid junction potentials at the ASW bridge, but not for the potential between 0-6 M-KCl and axoplasm. The latter has been estimated as 6 mv (Curtis & Cole, 1942), i.e. the true resting potential is 6 mv more negative than the measured potential. In the experiments with electrical depolarization we used a side hole electrode (Lecar, Ehrenstein, Binstock & Taylor, 1967) with a current wire. A small hole (less than 50 /tm in diameter) was made into the wall of a glass capillary, about 10 mm from its closed end, by means of a high frequency spark. A 50 gtm Pt wire was attached to the outside of the capillary; the wire was insulated except for the distal 35 mm which was carefully cleaned and platinized. The glass capillary with the side hole was filled with 0-6 M-KC1 and served to measure the membrane potential; the platinized Pt wire was used for passing current through the membrane to a Pt electrode in the external bath. The experiment illustrated in Fig. 11 was done with the voltage clamp method described in the preceding papers (Baker, Hodgkin & Ridgway, 1971;Bakeretal. 1973). RESULTS General observations with K-rich sea-waters Fig. 1 compares the light output in response to a change from Na-ASW to Na-free ASW in which the Na has been replaced by either choline or K. In the presence of choline the light output rises to a new steady level that is reached in about 2-4 min. Similar results are obtained when Na is replaced by Li (see Baker, Hodgkin & Ridgway, 1971). The response to K is quite different. The light output rises rapidly and then falls to a new steady level that is higher than in Na-ASW. The time between starting the solution change and the peak of the K response depended on the rate at which the solution was changed but the interval was normally about 10 sec. The new steady light output was reached in min. Further flows of K-ASW did not change the steady light output. These experiments show that complete replacement of external Na by choline, Li or K results in a maintained increase in the light output. Much of this increase in light presumably reflects the increase in Ca influx and

4 530 P. F. BAKER, H. MEVES AND E. B. RIDGWAY decrease in Ca efflux that is observed in the absence of external Na ions (see Baker, Blaustein, Hodgkin & Steinhardt, 1969; Blaustein & Hodgkin, 1969; Baker, Hodgkin & Ridgway, 1971; Baker, 1972). In addition replacement of Na by K also produces a fast transient increase in light. The properties of this transient light suggest that it reflects activation of the 'late Ca channel' (Baker, Meves & Ridgway, 1971, 1973). It is insensitive 0* ge8 A 0.1 F-T min Fig. 1. Comparison of the changes in light emission that follow complete replacement of external Na by either K (A.) or choline (B). Ordinate: light output (pta); abscissa: time (min). All solutions were Mg-free artificial sea-waters containing 112 mm-ca. The solutions were changed at the arrow. Axon diameter 750 /um. Temp. 20 C. It should be noted that complete replacement of external Na by Li gave results similar to those observed with choline. to external tetrodotoxin and internal tetraethylammonium ions, but is reduced by external Co2+, Mn2+ and Ni2+ ions (see Fig. 6) and by the organic drugs D-600 and iproveratril (see Fig. 5) all of which block the 'late Ca channel' (Baker et al. 1973). The maintained increase in light output in the presence of K is always greater than that in the presence of Li or choline. Addition of Mn2+, Co2+, Ni2+ or D-600 all reduce the maintained light production in high-k and it is of interest that simultaneous exposure to both Mn2+ (50 mm) and 410 mm-kcl results in a much reduced K- response, suggesting that the site of action of Mn2+ must be very superficial.

5 INACTIVATION OF Ca ENTRY 531 The response to different K concentrations To avoid simultaneous changes in the concentrations of both external Na and K, the preparation was normally kept in Na-free choline sea-water and choline was replaced by K to give the required K concentrations. The response to a range of K concentrations is shown in Fig. 2. Apart from 50 mm-k, which only produced a small rise in the steady light output, high K concentrations had a phasic effect on the light emission: after an initial peak the light declined to a steady value that was higher than the E +20 r 0... f--i so mm-kci _ I 30 I 40 I 50 I 60 min Fig. 2. Effect of different K concentrations on resting potential (above) and light emission (below) recorded simultaneously on two inkwriters. 112 mm-ca choline ASW containing 50, 100, 200 and 410 mm-kcl was applied for the times indicated; between tests the fibre was in 112 mm-ca choline ASW with 10 mr-kcl. Light curve interrupted for control of zero line before 200 mm-kcl test. Temp. 200 C. resting light. The size and rate of rise of the initial peak was greater for high K concentrations and its time course roughly followed the rate of depolarization. Fig. 2 also shows that the decline in light output was not accompanied by a change in potential. The time for the light response to fall to half decreased as the K concentration was increased and averaged 67-1 sec for 100 mm-kcl, 22-3 sec for 200 mm-kcl and 8-7 sec for 410 mm- KCl. The final steady light level was much less sensitive to the K concentration (Fig. 3). After returning to normal ASW (containing 10 mm-kcl) the light output rapidly fell to the resting level, sometimes, as in Fig. 2, undershooting its original value.

6 532 P. F. BAKER, H. MEVES AND E. B. BIDG WA Y The aequorin responses to KCl often increased during the course of an experiment in much the same way as Baker, Hodgkin & Ridgway (1971) observed with electrical stimulation or voltage clamp pulses. Very large responses to KCl were sometimes obtained from axons that had been stored at 40 C for about 20 hr before injecting aequorin. Fig. 4a shows the light response to 410 mm-kcl from such an extremely sensitive fibre, recorded 1 hr 48 min after injection of aequorin. The experiment was done in a smaller flow cell and on a faster time base than in Fig. 2. The resting glow in 10 mm-kcl was 300 na and is barely visible because of the small amplification employed. The response to 410 mm-kcl started 6 sec after the beginning of the fluid change, probably the time required to fill the dead space of the flow system. The mv L l, I, mm-kci Fig. 3. Collected results from five axons exposed to different K concentration in Mg-free choline sea-water containing 112 mm-ca. Ordinate: the increase in light output both for the peak (0) and final steady level (0, measured 2-5 min after the solution change) expressed relative to the final steady level observed in choline sea-water containing 200 mm-kcl; abscissa: the measured internal potential (mv). The K concentration used to depolarize the axons is also shown on the abscissa. Temp ' C. The vertical bars represent 2 x s.e. of the mean. light emission rose steeply, reached a peak value of 64-5 #ta (corresponding to 215 times the resting glow) and then declined in a matter of seconds to a much smaller constant level (about 6 times the resting glow). It stayed at this level for a further 70 sec (not shown in Fig. 4a) until the fluid was changed back to 10 mm-kcl. Even larger light responses with peak values up to 258,sA were found when the 410 m - KCl test was repeated during the following 3 hr. One of these large responses was obtained in the presence of 3-2 #M tetrodotoxin, indicating that the aequorin response to K depolarization is not due to activation of the tetrodotoxin-sensitive component of Ca entry.

7 INACTIVATION OF Ca ENTRY 533 The most striking feature of these experiments with K-rich sea-water is the phasic light response. In a single experiment a phasic response was also seen with sea-water containing 400 mm-rbcl. In some experiments with 410 mm-kcl, simultaneous measurements of the membrane potential showed that the light response began to decline before the membrane potential had reached its new steady level. Two general kinds of explanation for the phasic light output in response to high K seem most likely: either (1) the high rate of Ca entry leads to temporary depletion of the available aequorin near the surface membrane, or (2) the high rate of Ca entry is not maintained. The available experimental evidence favours the second alternative a 6k- C mm-kci I I I I I b sec >1_ mm-kci I I I I I d sec 0...' mm-kci I I I I I sec mm-kci sec Fig. 4. Aequorin response to high K concentration in sea-water with 112 mm-ca, 0 mm-mg (a) and in sea-water with 11 mrm-ca, 55 mm-mg (b). Record (a) in 112 mm-ca choline Cl ASW with 10 and 410 mm-kcl. Record (b), taken 133 min later on a 5 times higher gain, in 11 mn -Ca, 55 mm-mg ASW with 0 mm-kci (460 mm choline Cl) and 460 mm-kcl (0 mm choline Cl). The flow cell used for this experiment was a tube, 3 mm i.d. This reduced the dead space during solution changes. Note that the times to peak response and final steady state were much shorter than normal. In (c) and (d) the data of (a) and (b) are replotted as 4(light output) against time. Temp. 200 C. Evidence for inactivation of Ca entry If the phasic response results from temporary depletion of the aequorin available at the surface of the axon, the shape of the response should depend on the absolute amount of Ca that enters. This does not seem to be

8 534 P. F. BAKER, H. MEVES AND E. B. BIDG WA Y mv rest. pot I9JJL~110 jj- jfj110mm -KCI 0-8 D < 02t.4_1 > min Fig. 5. Effect of D-600 on the aequorin response to 200 mm-kc1. Inkwriter record of light emission in 112 mm-ca choline Cl ASW with 10, 100 or 200 mm-kcl as indicated. 81 min after the start of the experiment D-600 (10-4 g/ml) was added to the external solution; it was removed 20 mi later. The resting potential is given on top of the Figure. Note two interruptions in time scale. Axon diameter 715,um. Temp. 200C i1 K h0 i0 V110 jp 7l~ j mm-kci 6 a C 4 2 I 0 b 5 min Fig. 6. Reduction of the aequorin response to 410 mm-kcl by 50 mm-mncl2 (record b) and by a preceding response (record d). Inkwriter record of light emission in 112 mm-ca choline Cl ASW with 10 or 410 mm-kcl as indicated. Record a shows normal response to 410 mm-kcl. 13 min before record b 50 m M-nCl2 was added to the external solution. Records c and d were obtained in the absence of Mn ions. In record d 410 m -KCl was applied 75 sec after the end of the previous test. Axon diameter 700 /sm. Temp. 190 C.

9 INACTIVATION OF Ca ENTRY 535 the case. KCl responses of much decreased size, but essentially similar shape to the normal response, are illustrated in Figs. 4, 5 and 6. Fig. 4 compares the response to K-rich ASW in the presence of high and low external Ca concentrations. The size of the response was markedly reduced in the low Ca solution, but the light emission still followed the same general time course. Closer inspection of record b reveals that the peak is decreased by a factor of 8 whereas the maximum rates of rise and fall are 13 and 30 times smaller than in record a. In the experiment of Fig. 5, a large reduction of the KCl response was obtained by adding the drug D-600 to the external solution (see p. 530). In the absence of D-600 the light responses to 100 and 200 mm-kcl resemble those in Fig. 2. D-600 (10- g/ml.) reduced the resting light from 305 to 205 na. The drug also markedly decreased the peak of the light response to 200 mm-kcl without, however, altering its general time course. The peak of the reduced response to 200 mm-kcl exactly equalled the peak of the 100 mm-kcl response obtained earlier in the experiment, but despite the reduction in size, the rapid falling phase characteristic of exposure to 200 mm-kcl was fully preserved. Similar results were obtained with the,8-blocking drug propranolol (0.1 mg/ml.). Fig. 6 illustrates two other experiments in which the response to high K is reduced in size, but not in general shape. In Fig. 6b, 50mM-MnCl2 was included in the external solution. The response to 410 mm-kcl was greatly reduced in size but remained phasic. It should be noted in particular that even the peak of the phasic light response seen in the presence of M-n never exceeded the steady level maintained in high K sea-water in the absence of the inhibitor. Essentially similar results were obtained in the presence of 25 mm-coso4. The experiment shown in Fig. 6c and d illustrates a quite different means of reducing the size of the response to high K. If two exposures to high-k are given within a few minutes of each other, the second response is smaller than the first. Despite the reduction in size, the shape of the second response is almost identical to that of the first. The interpretation of this and similar experiments is discussed on p To conclude, the size of the response to high-k can be reduced in a number of different ways, but in no instance is its phasic form changed. These observations suggest that the phasic light response to high KCl concentrations is unlikely to result from utilization of aequorin near the periphery of the axon, but rather reflects a phasic entry of Ca. This might be produced if depolarization has a dual action serving both to activate Ca entry and also, more slowly, to inactivate it. An alternative explanation might be that maintained depolarization accelerates the rate of Ca removal but this seems unlikely because replacement of external Na by K reduces the rate of Ca extrusion (Blaustein & Hodgkin, 1969).

10 536 P. F. BAKER, H. MEVES AND E. B. BIDGWA Y Phasic light responses following electrical depolarization The experiments described so far provide evidence for a phasic entry of Ca during a maintained depolarization with high external concentrations ofk or Rb ions. In view of the possibility that the response may be specific for K or K-like ions, it is important to know whether similar behaviour can be obtained with prolonged electrical depolarization. This was exama b C #A 140 #A 140 #A 25 sec Fig. 7. Aequorin response to strong outward currents. Inkwriter record of three light responses to a depolarizing current of 140 PsA applied for 9 5 see (record a), 25 see (record b) and 8 see (record c). The interval between records a and b was 10 min; record c started 1 min 50 see after the end of record b. The current was passed from an internal current wire attached to a micro-electrode with side hole (see Methods). The potential records showed a rapid depolarization from mv (= resting potential) to + 23 mv, followed by a further much slower potential change; 2-5 and 15 see after the beginning ofcurrent flow the potential was + 23 and + 27 mv respectively. External solution: 112 mm-ca NaCl ASW with 1-6 /m tetrodotoxin. The injected aequorin was overlapped with an injection of 2-5 M- TEA. Axon diameter 850 4sm. Temp. 200 C. ined by following the light responses to outward currents of several seconds duration (Fig. 7). Current was passed from a 35 mm long platinized Pt wire attached to the internal micro-electrode. The micro-electrode had a side hole, about 10 mm from its closed tip, which made possible measurement of the membrane potential inside the region of uniform current flow (see Methods). The fibres were usually treated with external tetrodotoxin

11 INACTIVATION OF Ca ENTRY 537 and internal TEA to minimize the decrease of membrane resistance during depolarization. Currents which reduced the membrane potential to a value between -40 and -25 mv reversibly raised the light emission to a new steady level, an effect similar to that of 50 mm-kcl (see Fig. 2). To mimic the phasic effect of200 or410mm-kclthe membrane potential had to be reduced to about 0 mv which required outward currents of more than 100 pta. Currents of this size, flowing for several seconds, are likely to cause polarization at the electrodes. The potential record usually showed a rapid decrease in membrane potential followed bya further much slower potential change (see legend of Fig. 7) which was probably mainly due to polarization. Repeated application of such strong currents for several seconds often damaged the fibres; consequently, only a small number of measurements could be made on each fibre. In spite of these limitations, the experiments clearly established that the essential features of the light response to 200 or 410 mm-kcl can be imitated by depolarizing currents of sufficient strength. Record a in Fig. 7 shows the phasic light response to an outward current of 140 1tA which reduced the membrane potential to about +23 mv. The light emission rose rapidly to a peak and decreased again, decaying to about half its peak value within 8 sec. The break of the current led to a faster drop in light, followed by a slow decrease towards the resting level. In record b, taken 10 min later, the current was applied for a longer time. During the last 10 sec of current flow the light is seen to decrease less steeply, probably approaching a steady state as in the KCl experiments. The third test (record c) started 1 min 50 see after the end of b and elicited a much smaller light response than in a or b, indicating that the light response to depolarizing currents is followed by a relative refractory period very similar to that seen with K-rich solutions. It should be noted that this method of depolarizing the membrane gave the standard, bell-shaped relation between the rate of rise of the light response and the magnitude of the depolarization (Baker, Hodgkin & Ridgway, 1971; Baker et al. 1973). It is unlikely that the effect of depolarizing currents is due to K ions accumulating immediately external to the axon membrane. The time constant associated with concentration changes in this space is msec (Frankenhaeuser & Hodgkin, 1956) and the concentration of K at the end of the 8-25 see pulses in Fig. 7 can be estimated from the steady-state equation I = FPbA[K+], where F is the Faraday, Pb is the effective permeability of the barrier separating the space from the external solution and A[K+] is the change of K concentration in the space. Assuming a value of 6 x 10- cm/sec for Pb (Frankenhaeuser & Hodgkin, 1956), an axon diameter of 850,tm (see legend of Fig. 7) and an internal current wire of 35 mm length, the current of 140 /za (flowing through at, least 0-94 cm2 of membrane) should lead to a maximum A[K+] of 25 m r. This concentration change is much too

12 538 P. F. BAKER, H. MEVES AND E. B. BIDG WA Y small to produce a phasic light response; it would be further reduced in a TEAinjected axon where only part of the outward current is carried by K+ ions. Time course of recovery from inactivation Figs. 6d and 7c show that following inactivation either by high-k or by strong outward currents there is a relative refractory period during which the response to a second depolarization is much reduced in size. In Fig. 6d, after a recovery period of 75 sec, the peak of the second response to 410 mm-kcl was only 25 % of the first. Complete recovery took min. In Fig. 7 c, a second electrical depolarization given 110 see after the first elicited about half of the original response. Complete recovery took about 10 min. 100 A 100 I AB * 60 - * / I I I I I I I I Q Fig. 8. Recovery of light response to K-rich sea-water after a preceding response. A. Ordinate: peak of second light response as % of the first response. Abscissa: time (min) after end of preceding KCl application. Each symbol represents a separate experiment. The test solution used for the first and second light response was 410 mm-kcl (empty and half-filled symbols) or 200 mr-kcl (filled symbols). Between tests the fibre was in 112 mm-ca choline Cl ASW with 0 or 10 mm-kcl, except in * where 112 m -Ca NaCl ASW was used. The first test lasted 2-10 min. The curve was drawn from the equation y = 1- exp(- t/ir) with T = 5 min. Temp o C. B. The data of Fig. 8A replotted as V(test response) - V(resting light) x 100 V(control response) - V(resting light) against time. The curve was drawn from y = 1- exp(- t/t) with X = 3-5 mi. 1?he time course of the slow recovery from inactivation was investigated using K depolarizations. Fully quantitative data were rather difficult to obtain both because of the long and somewhat variable relative refractory period and also because of the tendency for the light response to change progressively (usually increase) during an experiment. To allow for these changes, the following sequence of tests was adopted: (1) 410 mm-kcl followed by a recovery period of min, (2) 410 mm-kcl followed by a variable recovery period before a second application of the same solution, and (3) a recovery period of min followed by a final exposure to

13 INACTIVATION OF Ca ENTRY mm-kcl. Each exposure to 410 mm-kcl was normally for 3-5 min. The results of seven experiments are collected in Fig. 8. In Fig. 8A the second response to high-k is plotted as a percentage of the first response, after allowing for any progressive changes in the high-k response. The values show considerable scatter, but are fitted roughly by an exponential function with a time constant of 5 min. 80 and 100 % recovery from the effect of a previous depolarization took 8 and 20 min respectively. There was some evidence that the recovery time was dependent on the duration of the initial exposure to high K. The recovery was somewhat faster after a short exposure (less than 2 min) and slower after a long exposure (more than 10 min), but this effect was not investigated further. The light produced by aequorin in vitro is proportional to the square of the Ca concentration (Shimomura, Johnson & Saiga, 1963; Ashley, 1970; Baker, Hodgkin & Ridgway, 1971), yet a major problem in interpreting data obtained from intracellular aequorin is to decide in each case whether the square relation holds under these conditions. One way to test the relationship in the squid axon is to stimulate the nerve repetitively at two different frequencies. If light is proportional to [Ca]2 and aiming that Ca entry is linearly related to the number of impulses per second, doubling the frequency of stimulation should quadruple the rate of light production. This is found in some axons, but in others light production is directly proportional to the frequency of stimulation. Baker, Hodgkin & Ridgway (1971) have suggested that this difference in behaviour may depend on whether the Ca entry is large by comparison with the resting [Ca] in the axon. Thus, if the light production is proportional to [CaR + ACa]2, where Ca, is the resting Ca and ACa the extra Ca entry, the response will be proportional to (CaR' + 2CaRACa + ACa2) and a linear response will be seen when ACa << Ca. and a square relation when ACa > CaR. If light production in all axons were proportional to [Ca]2, the square root of the light response would give a better measure of the true changes in Ca entry. This was not, however, verified for all the axons used in Fig. 8, and the square-rooting procedure shown in Fig. 8B may not be valid. Fig. 8B shows (V(Peak light) - V(Resting light)) for the second response as a percentage of the first. The data are again fitted by an exponential, but with a shorter time constant (3 5 min). A further source of error is that the size of the peak is dependent to some extent on the rate at which the external solution is changed. Although every attempt was made to keep this constant, some variation could not be avoided. As the K channel is known also to inactivate slowly in response to maintained depolarization (Ehrenstein & Gilbert, 1966), it was of interest to compare the time course of inactivation of the K channel with that of the 'late Ca channel'. In practice, we were only able to compare the rates of recovery from inactivation. The recovery of the light response after inactivation with 210 mm-kcl was compared with the recovery of the potassium outward current measured with 80 mv pulses of 3 msec duration. In sea-water containing 112 mm-ca, the K-outward current which had dropped during a 3 min application of 210 mm-kcl to about a quarter of its value in 10 mm-kcl, recovered rapidly after returning to sea-water

14 540 P. F. BAKER, H. MEVES AND E. B. RIDG WA Y with 10 mm-kcl reaching 77-5 and 100 % of its final value in 35 and 103 sec respectively. It seems likely that this recovery rate was largely determined by the velocity of the concentration change. The recovery of the K current was much faster than the recovery of the light response. 3 min 45 sec after return to sea-water the light response was only 48 % of its final response. This experiment shows that the long-lasting refractory state of the late light is not accompanied by a comparable depression of the K current and gives further support to the notion that the late Ca entry does not take place through the K channel (Baker et al. 1973). 4 A B C #A 21 'X mm-kc Fig. 9. Effect of pre-treatment with 100 mm-kcl on the light response to 410 mx-kcl. (A) and (C) without pre-treatment; (B) with pre-treatment. Ordinate: light output (#A); abscissa: time. The time marks under the records represent 10 see intervals. The recorder was slowed by a factor of six during the pre-treatment period in B which lasted 5.5 min. Between tests the axon was in Mg-free choline ASW containing 112 mm-ca. The records were obtained in the order shown with a gap of about 15 min between each determination. Axon diameter 700 /m. Temp O C. The small reduction in light associated with each solution change was not seen in all experiments and may be an artifact resulting from movement of the axon. Steady-state inactivation The extent of inactivation under steady-state conditions was investigated by pretreating axons with external K concentrations between 50 and 200 mm and then subjecting the fibre to a standard depolarization, usually exposure to 410 mm-kcl, but in a few experiments electrical depolarization was used. A set of experimental records using K-depolarization is shown in Fig. 9. Pretreatment with 100 mm-kcl clearly reduced the size of the light response to 410 mm-kcl and, in addition, the time course of the response was slowed down. This procedure was repeated pretreating with different K concentrations. In general, pretreatment with 140 or 200 mm-kcl led to a larger reduction of the test response than pretreatment with 100 mm-kcl. These findings are qualitatively consistent with the inactivation hypothesis. There was, however, considerable variation in the

15 INACTI VATION OF Ca ENTR Y extent of inactivation produced by a given K concentration and in two out of three experiments with 50 mm-kcl there was a small increase in the response to 410 mm-kcl. In an attempt to relate light to Ca entry (see previous section), the following ratio was calculated: <(conditioned light) - <(resting light) <(unconditioned light) - <(resting light) where 'conditioned light' and 'unconditioned light' refer to the peak values with and without pretreatment with KCl and 'resting light' is the light emission in 0 K-choline-ASW. In Fig. 10 the ratio is plotted as a function of the resting potential during _ mv mm-kci Fig. 10. Effect of pre-treatment with different K concentrations on the light response to 410 (or 300) mm-kc1. Ordinate: light response expressed as V(conditioned response) - V(resting light) 100 /(unconditioned response) - V(resting light) and abscissa: resting potentials during pre-treatment period (averaged from 3 other axons). The KC1 concentration used for pre-treatment is shown underneath. The measurements were made in the sequence *, 112 mm-ca; *, 20 mm-ca; 0, 112 mm-ca. 0 and 0, Mg-free choline ASW containing 112 mm-ca with 410 nm-kcl as the test solution. *, Mg-free choline ASW containing 20 mm-ca with 300 mm-kcl as the test solution. The pre-treatment period lasted 4-5 min. Axon diameter 700 jim. Temp. 200 C. The curves have been drawn through the points by eye and the increase following pre-treatment with 50 mm-kcl in 112 Ca has been ignored.

16 542 P. F. BAKER, H. MEVES AND E. B. RIDGWAY the pre-treatment period for two different external Ca concentrations. In this particular fibre there was very little inactivation following pre-treatment with 100 mm-kcl in sea-water containing 112 mm-ca. In sea-water containing 20 mm-ca the reduction of the test response after pre-treatment with 100 mm-kcl was much more pronounced than in 112 mm-ca and 50 mm-kcl also produced some inactivation. These findings are compatible with a shift of the inactivation curve towards more negative internal potentials, analogous to the shift of the Na inactivation curve in low [Ca] <~~~~~~~~~~ 0 mv Fig. I11. Light response to single electrical depolarizing pulses of duration 33 msec before exposure tokg (0), in the presence of 210 mm KCI (-), 15 min after removal of the KCI (v) and 68 min later (Q@). Ordinate: 4v(1ight output (na)); abscissa: depolarization from the resting potential (mv). In 210 rmm-kc1 the axons were depolarized on average by 43-5 mv and the electrical depolarizations have been shifted along the abscissa by this amount. Axon diameter 675 #'m. Temp C. described by Frankenhaeuser &r Hodgkin (1957). The potential for half inactivation in the experiment shown in Fig. 10 is -17 mv in 112 mm-ca and-34 mv in 20 mm-ca. In another experiment the potential for half inactivation was-20 mvin I12 mm-ca. These values referto the measured internal potential. If allowance is made for the junction potential between 06 m-kc1iand axroplasm (see p. 529), the potential for half inactivation becomes -b23 to -e26 mv in 112 mm-ca and -p40 mv in 20 mm-ca. An alternative explanation for the apparent steady-state inactivation produced by pre-treatment with K-rich solutions is that K ions may shift the relation between depolarization and activation of Ca entry to larger depolarizations. This was examined by measuring the light response to voltage clamp pulses of different amplitude before, during and after treat-

17 INACTIVATION OF Ca ENTRY 543 ment with a K-rich solution. The results are shown in Fig. 11. Varying the depolarizing pulse amplitude between 0 and 200 mv gave the usual bellshaped voltage response curve (see Baker, Hodgkin & Ridgway, 1971; Baker et al. 1973) with a maximum light response at a depolarization of about 100 mv. In ASW containing 210 mm-kcl the light response was reduced to zero over the whole potential range. After 7 min the fibre was returned to sea-water containing 10 mm-kcl. The bell-shaped voltage response curve slowly reappeared and reached 65 % of its original size in 15 min. These results make it unlikely that pre-treatment with K produces a shift in the potential dependence of Ca entry. Another way to examine the inactivation produced by exposure to different K concentrations is to compare the final steady light level in the presence of K with the peak response. This can be made reasonably quantitative by calculating, for each experiment, the ratio /(steady light in K) - V(resting light),/ (peak light in K) - J (resting light) The values obtained for this ratio were in 50 mm-k; in 100 m - K; in 200 mus-k and in 410 mm-k. If it is assumed that all the extra light in the presence of K is due to activation of a single Ca entry mechanism (see p. 546), these calculations suggest that the extent of inactivation in the steady state is increased as the K concentration is raised, an observation that is consistent with inactivation being a voltage-dependent process. The rate of onset of inactivation This proved very difficult to study quantitatively largely because the decay of the light emission follows the reduction in Ca entry with a considerable delay that is determined by the rather slow intracellular buffering of the Ca ions that have entered. The mitochondria seem to be largely responsible for returning the intracellular Ca concentration to its resting level and this process has a time constant of about 10 sec (Baker, Hodgkin & Ridgway, 1971). It follows that the time course of the phasic records in Figs. 1, 2, 4, 6, 7 and 9 should not be taken as the time course of inactivation, although the slower fall in 100 mm-kcl as compared with 410 mm-kcl may be indicative of a slower rate of inactivation in the presence of the lower K concentration (see p. 544). Two approaches were used to obtain information on the rate of inactivation. The first is illustrated in Fig. 12. Two identical depolarizing pulses were given in close succession. Electrical depolarization has the advantage over KCl that the period of depolarization can be kept short. In record a of Fig. 12, the first pulse was terminated at the peak of the light response while in record b it was switched off before the peak was reached. In both cases the second pulse, starting 10 see after the end of the first pulse, elicited a light response of subnormal size. Its peak was 66 and 89 % of the normal peak in records a and b respectively. The experiment seems to 22 PHY 231

18 544 P. F. BAKER, H. MEVES AND E. B. RIDGWAY show that inactivation of Ca entry begins during the rapid rising phase of the phasic light response, i.e. before the light emission actually begins to fall. In addition, Fig. 12 shows that a depolarization to about 0 mv must last more than 5 sec ifa phasic time course of the light response is to be obtained. The second approach involved pre-treating the axon with K-rich solutions for different times before exposure to 410 mm-kcl. In two experiments, the reduction in the 410 mm-kcl response was half complete after 40 and 60 sec pre-treatment with 100 mm-kcl and after less than 10 sec pre-treatment with 210 mm-kcl. These results using KCl depolarizations are consistent with those illustrated in Fig. 12 where large electrical depolarizations were used. In conclusion, the available data suggest that the rate of inactivation is dependent on potential, the rate increasing as the resting potential is reduced to zero. The effect of larger depolarizations was not examined. a 4- b 2-0 _ tj t t t t t 1701A 170 ua 1704A 1701A 30 sec 30 sec Fig. 12. Inkwriter records of light response to a pair of depolarizing pulses. Record a: first pulse 170,A for 5 sec, 10 sec interval, second pulse 170 #tsa for 10 sec. Record b: same as a, but first pulse shortened to 2-5 sec. Record b 22 min after record a. The potential records showed a rapid depolarization from the resting potential (-55 mv in a, -53 mv in b) to + 8 mv, followed by a further much slower potential change; 2-5 and 5 sec after the beginning of current flow the potential was + 8 and + 10 mv respectively. External solution: 112 mm-ca NaCl ASW. The axon was injected with aequorin and 2 5 M TEA. Axon diameter 920 /m. Temp. 17 C. DISCUSSION The most interesting feature of the work described in this paper is the observation that Ca entry in response to a maintained depolarization is phasic. The pharmacological properties of this phasic response are consistent with the entry of Ca occurring through the late Ca-channel (Baker et al. 1973). The transient form of the response seems to result from a dual action of depolarization on this channel: depolarization serving first to open and more slowly to close (inactivate) the channel.

19 INACTIVATION OF Ca ENTRY 545 The inactivation hypothesis is supported by the observation that pretreatment of the fibre with mm-kcl reduces the light response to a test depolarization with 410 mm-kcl. These experiments also give information about the extent of inactivation in the steady state, indicating that the potential for half inactivation is about -25 mv in 112 mm-ca and -40 mv in 20 mm-ca. The interpretation of these experiments is, however, complicated by a number of factors. (a) Pre-treatment of the fibre with a low KCl concentration (50 mm-kcl) in two out of three experiments led to an increase rather than to a decrease of the test response (p. 541). (b) There was considerable variation in the extent of inactivation produced by a particular K concentration. (c) The experiment in Fig. 12 suggests that inactivation develops during the rising phase of the light response; if the rate of inactivation during the rising phase is affected by the pretreatment, the peak of the test response would not be a correct measure of the inactivation reached at the end of the pre-treatment period. (d) Different quantitative results are obtained depending on whether the light or the square root of the light is used (p. 539). In view of these uncertainties we consider the quantitative data for the steady-state inactivation of the late Ca permeability as tentative. A further complication might arise if Ca entry leads to accumulation of Ca ions in the vicinity of the membrane. This effect complicates the analysis of Ca currents in cardiac muscle fibres (Beeler & Reuter, 1970). In an 800 jsm axon a steady inflow of 10 p-mole/cm2 sec would raise the calcium concentration at the surface by 14 /Lm above its normal level which is about 1 /IM for an external solution with 112 mm-ca (see Fig. 22 of Baker, Hodgkin & Ridgway, 1971). In 112 mm-ca, an extra Ca inflow of 10 p-mole/cm2 sec occurs during stimulation at 125/sec. Possibly, the Ca entry during the initial phase of the light response to K-rich solutions is much larger than during repetitive stimulation with this frequency. Some support for this was found in experiments where the light output during repetitive stimulation was compared with that produced by exposure to 410 mm-kcl. The peak response in high K was always appreciably greater (in some experiments twenty times higher) than the steady level achieved during stimulation at 200/sec. One could argue that the reduction of the light response to 410 mm-kcl after pre-treatment with mm-kcl or during the relative refractory period is also due to a local increase in ionized Ca. The observation that the light emission is raised during the pre-treatment period, indicating a rise in surface Ca, is not inconsistent with this argument. Against it, however, is the observation that inactivation is essentially unchanged when the Ca entry is markedly reduced, for instance by Mn. Furthermore, there is no indication of an increased surface Ca or resting light during the refractory period. It seems unlikely, therefore, that the effect of pre-treatment or the phenomenon of refractoriness are wholly caused by an accumulation of Ca ions in the surface layer. Yet the possibility cannot be excluded that the quantitative results are affected by local changes in ionized Ca. The rate of inactivation seems to be potential dependent, but even for depolarizations to zero membrane potential is quite slow, taking many seconds to reach a steady value. The steady light output in K-rich sea- 22-2

20 546 P. F. BAKER, H. MEVES AND E. B. RIDGWA Y waters is always higher than in sea-waters containing 0 or 10 mm-kcl and is rather insensitive to external K between 100 and 410 mm (Figs. 2 and 3). An insensitivity of this kind could be observed if the increase of steadystate inactivation is compensated by the larger activation of the Ca entry mechanism at low membrane potentials. However, it cannot be excluded that the steady light output in K-rich sea-water is at least partly determined by the activation of an additional process, for instance Na-Ca counter transport (Baker et al. 1969). At present there is no experimental evidence that enables the participation of Na-Ca counter transport to be evaluated. It is known that external K activates Cadependent Na efflux and presumably Ca influx and replacement of external Na by K reduces Ca efflux (Baker et al. 1969; Blaustein & Hodgkin, 1969), both of which should lead to a rise in intracellular ionized Ca. If [K]0 is promoting Ca entry in exchange for internal Na, the steady light output should be dependent on the internal Na concentration, but this has not been examined experimentally. In view of the slow onset of inactivation, it seems unlikely that inactivation will affect Ca entry through the late Ca channel during a single action potential, but as the extent of inactivation in the steady state is dependent on potential, it is quite possible that slow or maintained changes in membrane potential may modify the number of late Ca channels available for activation. It is of interest that in a number of nerves the release of nervous transmitter substances seems to be dependent on the previous potential of the presynaptic terminal (del Castillo & Katz, 1954; Hagiwara & Tasaki, 1958; Takeuchi & Takeuchi, 1962; Miledi & Slater, 1966; Bloedel, Gage, Llinas & Quastel, 1967; Hubbard, 1970): depolarization clearly reduces and hyperpolarization may increase transmitter release in response to a given stimulus. It seems possible that the inactivation process described in this paper may contribute to these phenomena. Katz & Miledi (1971) have examined the response of the squid giant synapse to prolonged electrical depolarization of the presynaptic terminal and conclude that if inactivation of Ca entry does occur, it only develops very slowly - over a period of seconds. As Katz & Miledi point out there is no contradiction between these findings and those described in this paper. It should, however, be noted that the conditions of Katz & Miledi's experiments were somewhat different from ours. In response to large presynaptic depolarizations of 200 mv (i.e. close to, or beyond Eca), transmitter release is suppressed and repolarization is associated with a surge of transmitter release. Assuming that release is dependent on Ca entry, Katz & Miledi used the surge of release as a measure of the degree of inactivation of Ca entry. They found that even after depolarizations lasting 4 sec, the off-response was 70 % of its maximum value. This is rather difficult to compare quantitatively with our results for two reasons: (1) Katz & Miledi were working at 110 C, whereas all our experiments were done at about 200 C and (2) we did not examine the effects of depolarizations as large as those used by Katz &

21 INACTIVATION OF Ca ENTRY 547 Miledi. In our experiments, axons were depolarized to internal potentials close to zero and at most to + 20 mv, whereas the potential used by Katz & Miledi to suppress release probably made the inside of the terminal more positive than mv. We have no information on the rate of inactivation at such large positive internal potentials. There is also a strong resemblance between the phasic Ca entry observed in nerve and the phasic contractile response of many muscles to depolarization produced either electrically or by K-rich solutions (Hodgkin & Horowicz, 1960; Littgau, 1963; Frankenhaeuser & Lannergren, 1967; Lannergren, 1967; Heistracher & Hunt, 1969), which suggests that there may be similarities between Ca entry in nerve and the mechanism of Ca release in muscle. It should be stressed that the present results rely almost entirely on the use of aequorin as an indicator of Ca movements. Hodgkin & Keynes (1957) showed that the Ca entry in response to a prolonged K-depolarization (20-30 min) is much less than that produced by repetitive electrical stimulation, a difference that might be explicable in terms of the inactivation of Ca entry that is produced by maintained depolarization. But, apart from these tracer data, there is no other independent evidence for inactivation of Ca entry. So far, no Ca currents have been detected, even in some old axons where Ca entry was apparently very large, and the possibility must be considered that the late phase of Ca entry might reflect mainly release of intracellular Ca or some sort of electroneutral exchange. We wish to thank the Director and staff of the Laboratory of the Marine Biological Association, Plymouth, for their help in providing facilities and an excellent supply of experimental material. E. B. R. was supported by a post-doctoral fellowship from the U.S. Public Health Service and P. F. B. was in receipt of a grant from the Browne Fund of the Royal Society. REFERENCES ASHLEY, C. C. (1970). An estimate of calcium concentration changes during the contraction of single muscle fibres. J. Phy8iol. 210, P. BAKER, P. F. (1972). Transport and metabolism of calcium ions in nerve. Prog. Biophy8. molec. Biol. 24, BAKER, P. F., BLAUSTEIN, M. P., HODGKIN, A. L. & STEINHARDT, R. A. (1969). The influence of calcium on sodium efflux in squid axons. J. Physiol. 200, BAKER, P. F., HODGKIN, A. L. & RIDGWAY, E. B. (1971). Depolarization and calcium entry in squid giant axons. J. Physiol. 218, BAKER, P. F., MEVES, H. & RIDGWAY, E. B. (1971). Phasic entry of calcium in response to depolarization of giant axons of Loligo forbesi. J. Physiol. 216, 70P. BAKER, P. F., MEvEs, H. & RIDGWAY, E. B. (1973). Effects of manganese and other agents on the calcium uptake that follows depolarization of squid axons. J. Physiol. 231, BEELER, G. W. & REUTER, H. (1970). Membrane calcium current in ventricular myocardial fibres. J. Physiol. 207,

22 548 P. F. BAKER, H. MEVES AND E. B. RIDGWA Y BLAusTEiN, M. P. & HODGKIN, A. L. (1969). The effect of cyanide on the efflux of calcium from squid axons. J. Physiol. 200, BLOEDEL, J., GAGE, P. W., LLINAS, R. & QUASTEL, D. M. J. (1967). Transmission across the squid giant synapse in the presence of tetrodotoxin. J. Physiol. 188, 52-53P. CURTis, H. J. & COLE, K. C. (1942). Membrane resting and action potentials from the squid giant axon. J. cell. comp. Physiol. 19, DEL CASTILLO, J. & KATZ, B. (1954). Changes in end-plate activity produced by presynaptic polarization. J. Physiol. 124, EHRENSTEIN, G. & GILBERT, D. L. (1966). Slow changes of potassium permeability in the squid giant axon. Biophys. J. 6, FRANKENXAEUSER, B. & HODGKIN, A. L. (1956). The after-effects of impulses in the giant nerve fibres of Loligo. J. Physiol. 131, FRANKENHAEUSER, B. & HODGKIN, A. L. (1957). The action of calcium on the electrical properties of squid axons. J. Physiol. 137, FRANKENHAEUSER, B. & LANNERGREN, J. (1967). The effect of calcium on the mechanical response of single twitch muscle fibres of Xenopu8 laevi8. Acta physiol. 8cand. 69, HAGIWARA, S. & TASAKi, I. (1958). A study on the mechanism of impulse transmission across the giant synapse of the squid. J. Physiol. 143, HEEISTRACHER, P. & HUNT, C. C. (1969). The relation of membrane changes to contraction in twitch muscle fibres. J. Physiol. 201, HODGKIN, A. L. & HOROWICZ, P. (1960). Potassium contractures in single muscle fibres. J. Physiol. 153, HODGKIN, A. L. & HUXLEY, A. F. (1952). The dual effect of membrane potential on sodium conductance in the giant axon of Loligo. J. Physiol. 116, HODGKIN, A. L. & KEYNEs, R. D. (1957). Movements of labelled calcium in squid giant axons. J. Physiol. 138, HUBBARD, J. I. (1970). Mechanism of transmitter release. Prog. Biophys. molec. Biol. 21, KATZ, B. & MILEDI, R. (1971). The effect of prolonged depolarization on synaptic transfer in the stellate ganglion of the squid. J. Physiol. 216, LANNERGREN, J. (1967). Contractures of single slow muscle fibres of Xenopus laevi8 elicited by potassium, acetylcholine or choline. Acta physiol. scand. 69, LECAR, H., EHRENSTEIN, G., BINSTOCK, L. & TAYLOR, R. C. (1967). Removal of potassium negative resistance in perfused squid giant axons. J. yen. Physiol. 50, LUTTGAU, H. C. (1963). The action of calcium ions on potassium contractures of single muscle fibres. J. Physiol. 168, MILEDI, R. & SLATER, C. R. (1966). The action of calcium on neuronal synapses in the squid. J. Physiol. 184, SHIMOMURA, O., JOHNSON, F. H. & SAIGA, Y. (1963). Microdetermination of calcium by aequorin luminescence. Science, N.Y. 140, TAKEUCHI, A. & TAKEUCHI, N. (1962). Electrical changes in pre- and post-synaptic axons of the giant synapse of Loligo. J. yen. Physiol. 45,