Ionic Selectivity of the Sodium Channel of Frog Skeletal Muscle D O N A L D T. C A M P B E L L From the Department of Physiology and Biophysics, University of Washington School of Medicine, Seattle, Washington 98195. Dr. Campbell's present address is the Department of Physiology, Yale University School of Medicine, New Haven, Connecticut 06510.
ABSTRACT The ionic selectivity of the Na channel to a variety of metal and organic cations is studied in frog semitendinosus muscle. Na channel currents are measured under voltage clamp conditions in fibers bathed in solutions with all Na + replaced by a test ion. Permeability ratios are calculated from measured reversal potentials using the Goldman-Hodgkin-Katz equation. The permeability sequence was Na + ~- Li + --~ hydroxylammonium > hydrazinium > ammonium > guanidinium > K÷ > aminoguanidinium in the ratios 1:0.96:0.94:0.31:0.11:0.093: 0.048:0.031. No inward currents were observed for Ca +÷, methylammonium, methylguanidinium, tetraethylammonium, and tetramethylammonium. The results are consistent with the Hille model of the Na channel selectivity filter of the node of Ranvier and suggest that the selectivity filter of the two channels is the same. INTRODUCTION T h e ionic selectivity o f axon sodium channels has been well described (Chandler a n d Meves, 1965; Hille, 1971, 1972, 1975b, c; Meves a n d Vogel, 1973). This p a p e r extends the study o f Na channel permeabilities to skeletal muscle using the muscle voltage clamp m e t h o d described in the first o f this series o f papers. In the past such experiments have been difficult to p e r f o r m on muscle for the lack o f a voltage clamp technique sufficiently fast to provide g o o d resolution o f Na channel currents that also permits the many necessary solution changes without disturbing the preparation. In nerve the Na channel is permeable to many small cations (Chandler a n d Meves, 1965; Tasaki et al., 1966; Hille, 1971, 1972). In particular, Hille was able to characterize the ionic selectivity at the frog n o d e o f Ranvier by systematically testing the permeability to an exhaustive list o f small organic and metal cations. H e p r o p o s e d that a 3.1 x 5.1-A oxygen-lined pore would explain the exclusion of all i m p e r m e a n t m o n o v a l e n t cations tested on simple geometric and chemical g r o u n d s (Hille, 1971, 1972, 1975 c). T h e observed selectivity sequence o f perm e a n t cations and deviations f r o m i n d e p e n d e n c e have been successfully described by a f o u r - b a r r i e r Eyring rate t h e o r y model o f the sodium channel that includes a negatively c h a r g e d site at this selectivity filter (Hille, 1975 b, c). As is shown here Na channel permeabilities observed for frog muscle are nearly identical to those observed for f r o g n o d e a n d are thus also consistent with this model. This work has a p p e a r e d in preliminary form (Campbell, 1974, 1975). THE JOURNAL
OF GENERAL
PHYSIOLOGY
" VOLUME
67, 1976 " pages 295-307
295
296
THE JOURNAL OF GENERAL PHYSIOLOGY" VOLUME 6 7 ' 1976 METHODS
Voltage Clamp Preparation The voltage clamp method is that described in the previous paper (Hille and Campbell, 1976) except experiments were performed before all the improvements described there had been made. Specifically, compensation for the attenuating effects of the impedance ZEB were not available. The abbreviations and nomenclature in Hille and Campbell (1976) are used throughout. A piece of a single muscle fiber is removed from the dorsal head of a frog semitendinosus muscle and placed across partitions separating the four pools of the recording chamber. Pool A, containing the area of m e m b r a n e u n d e r voltage clamp control, is filled either with normal Ringer or Ringer with all the sodium replaced by a test ion. The other three pools contain 120 mM CsF. Cutting the fiber ends in CsF eliminates mechanical activity and contributes to the electrical fidelity of the voltage clamp (see the preceding paper). Of importance to this study, cutting in CsF also raises the Na channel reversal potential, presumably by replacing some of the Na + and K ÷ ions inside the fiber with impermeant Cs + ions. The importance of a high reversal potential for measuring the permeability to relatively i m p e r m e a n t ions will be discussed later. Two feedback amplifiers are used. One is required for potentiometric recording of the m e m b r a n e potential which appears as --EM in pool A. T h e other amplifier supplies the end of the fiber in pool E with the m e m b r a n e current required to hold this potential to the voltage clamp command potential. Membrane current is not measured explicitly, but is assumed to be proportional to the voltage in the current-injecting pool (VE). The reservations in this assumption are discussed in the previous paper and in a later section on errors. All records were corrected electronically for leakage and capacity currents. The frequency response of the amplifiers used in these experiments is lower than shown in the first paper but still adequate for rapid voltage clamp control. Compensation tot capacity and leakage currents permits all other current to be measured starting about 80 tzs after the voltage step.
Recording and Analysis T e n to twenty current traces, in response to voltage clamp steps spaced about 7.5 mV apart, are recorded on a storage oscilloscope and then on film. Peak Na channel currents (measured as VE) are graphed to determine the reversal potential. Reversal potentials from bracketing runs in normal Ringer (ENa) are averaged and subtracted from the reversal potential measured in the test solution (Es). The resulting change in reversal potential is used to calculate the desired permeability ratio from the Goldman-HodgkinKatz voltage equation (Goldman, 1943; Hodgkin and Katz, 1949). The change in reversal potential between the two solutions is given by (Hille, 1971): Es - EN, = 2.303
(RT/F) log10 (Ps[S]/P~a[Na]),
where 2.303 RT/F is 55.2 mV at 5°C, [S] is the activity of test ion in test Ringer, [Na] is the activity of Na + in the control Ringer, and Ps/PNa is the ratio of channel permeabilities to the test and Na + ions. This way of determining Ps/PNa is used because unlike methods based on flux or current relations it does not assume independence and does not depend on the n u m b e r of channels open (Hille, 1971, 1975 c). It does assume that the selectivity of the channel is i n d e p e n d e n t of m e m b r a n e voltage and the solution bathing the fiber. These assumptions may not be true. For some models of the Na channel selectivity filter even the Goidman-Hodgkin-Katz voltage equation does not hold strictly true and permeability ratios may depend on other experimental conditions (Hille, 1975 b). Nevertheless,
297
DONALD T. CAMPBELL Muscle Na Channel Selectivity
the above method seems to be the most useful way to d et er m i n e selectivity o f a channel for comparison with other results as long as these reservations are kept in mind.
Activities For 0.1 mol/kg solutions at 25°C the activity coefficients are: LiCI, 0.790; NaCI, 0.778; KC1, 0.770 (Robinson and Stokes, 1965). For 0.07 M CaC12, the activity coefficient is 0.55. T h e Guggenheim convention (Butler, 1968; Shatkay, 1968) defines the single ion activity coefficient for Ca ++ as the square of the activity coefficient o f the salt, in this case 0.30. For the metal cations the activity coefficients are in the ratio 1.02:1.00:0.99:0.39 for Li +, Na +, K +, and Ca ++. These ratios are used with the cation concentrations to compute the activities required by the above equation. Since activity data are not available for the organic cations tested, their activity coefficients were assumed equal to that o f 0.1 mol/kg NaC1 for the calculation of permeability ratios.
Solutions T h e CsF solution used in the end pools contains 120 mM CsF with 1 mM imidazole or tris(hydroxymethyl)aminomethane buffer, pH 7.4. T h e composition of the test solutions is given in Table I. Test solutions generally contain 110 mM NaC1 or an osmotically equivalent concentration of a Na substitute, 2 mM CaCI~, and 4 mM tris(hydroxymethyl)aminomethane buffer, pH 7.4. For simplicity the solutions are named "Na Ringer," "Li Ringer," and so on according to the test ion. Hydrazine and hydroxylamine have pKa's below 8 and the pH of these solutions is adjusted to 5.97 and 5.79 to obtain a higher concentration o f the ionized species. T h e hydrazinium and h y d r o x y l a m m o n i u m Ringer also have slightly different concentrations of other ions as a result of dilution that occurred when adjusting them to their p H . T h e osmolarity o f the Na Ringer is 203 mosM, h y d r o x y l a m m o n i u m Ringer 154 mosM. All others tested had osmolarities within 5% of the Na Ringer value. TABLEI
C O M P O S I T I O N OF T E S T S O L U T I O N S Ringer
Anion
IS+]
Na 1/8 Na
CICI-, Br-
110 13.8
Li K
CICI-
110 110
Ca Hydroxylammonium
C1CI-
89.5 52.9
Hydrazinium
CI-
68.9
Ammonium Guanidinium Aminoguanidinium Methylammonium Methylguanidinium Tetramethylammonium Tetraethylammonium
CICINOf CISO4 BrBr-
Comment
Osmolarity maintained with 96.2 mM TMA Br. 4 mM CsCI added for some experiments. pH = pH pH = pH
5.79; [S+] calculated from and [S]totat; [Ca ++] = 1.6. 5.97; [S+] calculated from and [S]totaJ; [Ca ++] = 1.3.
110 110 110 110
73.3 110 110
Except as noted, all solutions also contain 2 mM CaCI2, 4 mM tris(hydroxymethyl)-aminomethane buffer and have a pH between 7.3 and 7.4.
298
THE JOURNAL OF GENERAL PHYSIOLOGY'VOLUME 6 7 ' 1976
Errors The previous paper describes several c u r r e n t - d e p e n d e n t voltage errors. At maximum sodium currents these errors may total 20 mV, but since the net m e m b r a n e current at the Na channel reversal potential is roughly 1/10 the peak value an error of only about 2 mV is expected. In addition, the method described below for measuring attenuation artifact includes this error, therefore the attenuation correction simultaneously reduces currentdependent errors in reversal potential measurement. Another error might be expected to arise from contributions of Na channels in the membrane of the transverse tubular system which presumably is u n d e r poor voltage clamp control. T h e voltage clamp calculations illustrated in Fig. 12 of the first paper show that at least for the model chosen there is a negligible effect of tubular sodium conductance on the reversal potential. The current signal in these experiments was not corrected for the capacitative component of the current-injecting pathway ZED as described in the first paper. Because this error is also proportional to the m e m b r a n e current it should not greatly affect the measurement of reversal potential. In fact, all of the dynamic errors taken together apparently have little effect on measured reversal potential changes. For instance, when the A pool solution is changed from normal Na Ringer to Ringer with 7/8 of the NaCI replaced by the i m p e r m e a n t salt tetramethylammonium (TMA) bromide, the average change in reversal potential measured in 16 experiments is -48.1 mV at 5°C while the theoretical change predicted by the Nernst equation is -49.8 mV. This is the method used to determine the a m o u n t of "attenuation artifact," a voltage error found when this voltage clamp method is used to study the node of Ranvier (Dodge and Frankenhaeuser, 1958; 1959; Hille, 1971). In that preparation the attenuation may be 10-20%. For the muscle preparation the attenuation averaged 3.5%, and the average reversal potential changes used to calculate permeability ratios were corrected by this amount. This 1.7-mV deviation from the Nernst prediction, if artifactual, implies that the combination of all errors has only a very small effect on the reversal potential change measured between these two solutions. The reversal potential changes are corrected for the differences in junction potential that the control and test Ringer solutions make with the agar bridge in pool A. These potentials are measured with respect to a Beckman 38402 ceramic junction saturated KC1 reference electrode (Beckman Instruments, Inc., Fullerton, Calif.). T h e measured junction potentials are given in Table II. For most solutions the correction is less than 2 inV. RESULTS
T h e o b s e r v e d c h a n g e s in r e v e r s a l p o t e n t i a l are listed in T a b l e I I . F o r p e r m e a n t ions the a v e r a g e r e v e r s a l p o t e n t i a l c h a n g e s are c o r r e c t e d for a t t e n u a t i o n a n d j u n c t i o n p o t e n t i a l s a n d t h e n u s e d to calculate p e r m e a b i l i t y ratios. R e v e r s a l p o t e n t i a l c h a n g e s a r e e x p r e s s e d as m e a n + SEM. I n s o l u t i o n s s h o w i n g n o i n w a r d c u r r e n t s , n o r e v e r s a l p o t e n t i a l c a n be m e a s u r e d . F o r these i m p e r m e a n t ions the p o t e n t i a l t h a t first gives a m e a s u r a b l e t r a n s i e n t o u t w a r d c u r r e n t is t a k e n as the u p p e r limit o f the reversal p o t e n t i a l a n d u s e d to calculate a limit to the p e r m e a b i l i t y ratio. I n the case o f i m p e r m e a n t ions the lowest limit o n p e r m e a b i l ity is r e p o r t e d r a t h e r t h a n a n a v e r a g e .
Lithium S o d i u m c h a n n e l s o f m u s c l e a r e a b o u t as p e r m e a b l e to l i t h i u m as they a r e to s o d i u m . I n Fig. 1, the voltage c l a m p series in Li R i n g e r is n e a r l y i d e n t i c a l to the
1
46.7 0.84 + 1.4
48 48 49 48 42 46 49 46 37 43 49 49 48 48 48 49 0.0 0.55 + 0.5 0.5 0.96
1.5 1.5 - 1.5 0.0 0.0 - 1.5
73. I 2.96 -2.4 73.2 0.048
72* 75* 78*
81"
72 57
77
K
Ca
AND
+4.2 >84.2 77
>61 >68 >75
>66 >77 >54
CHANGES
19.3 1.7 - 0.8 19.1:[: 0.94
22 18
22 15
37.0 1.3 + 1.0 39.3~ 0.31
34 41 38
31 38 37 40
H~drazinium
PERMEABILITY Hydroxylammonium
54 55
49
51
NH4
54.6 2.2 + 0.5 57.0 0.093
69 48 52 57 71 63 57 49 45 60 57 57 48 51 55 55
34
CHANNEL
OF
80.5 1.4 -0.2 83.1 0.031
84 77 81 80
>115
>103 >98 >115 >89 >112
>104 >101 >112 >103 > 109 >92
TMA
TEA
>109
>109 >92 >109 >94
MUSCLE
-4.5 +0.9 +2.3 -O.9 >111.9 >114.4 >116.9 >115.2 97 >109
>100
>95
> 109
AminoMethylMethylguanidinium, ammonium guanidinium
IN SODIUM
Guanidinium
RATIOS
52.3 1.4 - 1.8 52,3 0.11
II
* T h e K Ringer in these experiments also contained 4 mM CsCl as explained in the text. Corrected also for concentration o f ionized species. For i m p e r m e a n t ions the largest limit on reversal potential change was used to c o m p u t e the limit for the permeability ratio. Potentials are in millivohs.
MEAN SEM ]unction potential After corrections 0s/PNa
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
1/8 Na
Fiber
Li
POTENTIAL
REVERSAL
TABLE
t~
0 Z
300
THE JOURNAL OF GENERAL PHYSIOLOGY" VOLUME 6 7 ' 1976
1 [mY) - 100
-50
,
0
ms
I
2
5
5 I0 ~ E [m V)
°t 0
-2oo
-
"---~ ~
~;(/ m'x,,., X ~ / "
/ Lithium
- 400L-
FIGURE I. Lithium currents in the muscle Na channel. (Left) Peak current-voltage relations in Na and Li Ringer. The smooth curves were drawn by hand. Although the reversal potential in the two solutions is the same, the maximum currents in Li Ringer are about 20% less than in Na Ringer. (Right) Voltage clamp series from the same experiment showing currents for voltage steps spaced 15 mV apart. Fiber ends were cut in 120 mM CsF. Fiber 5, 5°C. control series in Na Ringer. T h e c o r r e s p o n d i n g current-voltage relation shows almost no change in reversal potential u p o n substituting the Li Ringer. In six experiments Li Ringer lowered the reversal potential 0.5 ± 0.6 mV, for a permeability ratio PLi/Pya o f 0.96. A l t h o u g h the muscle Na channel is a b o u t as permeable to Li + ions as it is to Na + ions, the current-voltage relation o f Fig. 1 shows that the currents in Li Ringer are 16-20% less than in Na Ringer. Evidently Li + ions block Na channels in addition to passing t h r o u g h them. In other words Li + ions violate the i n d e p e n d e n c e principle (Hodgkin and Huxley, 1952; Hille, 1975 c) in the muscle Na channel. As is shown below, some o t h e r p e r m e a n t ions show similar deviations from i n d e p e n d e n c e .
Potassium Potassium is the only other metal ion tested f o u n d to be measurably p e r m e a n t in the Na channel. It is also one o f the most difficult to test. After a minute or two in K Ringer with C1- as the anion, the muscle swells and the m e m b r a n e often develops large irreversible leaks destroying the p r e p a r a t i o n . In later experiments preparations bathed in isotonic K2SO4 survived well. In addition, the high external concentration o f potassium increases the potassium permeability o f other channels giving rise to large rectifying b a c k g r o u n d currents that must be subtracted to determine currents in Na channels. For this reason four experiments were p e r f o r m e d in K Ringer with 4 mM CsC1 a d d e d to help keep these extra potassium conductances blocked. In Table II the reversal potential changes f r o m these experiments are noted with an asterisk (*). In seven experiments the reversal potential in K Ringer was 73.2 ± 3.0 mV lower than in Na Ringer, c o r r e s p o n d i n g to a permeability ratio PK/Pya o f 0.048. Sodium is about 21 times more p e r m e a n t than potassium in the Na channel o f muscle.
DONALD T . CAMPBELL
301
Muscle Na Channel Selectivity
Hydroxylammonium and Hydrazinium T w o o f the most p e r m e a n t sodium substitutes tested are h y d r o x y l a m m o n i u m and hydrazinium. Voltage clamp series and c u r r e n t voltage relations f r o m a fiber bathed in these solutions are shown in Fig. 2. In four e x p e r i m e n t s , changing the bathing m e d i u m to h y d r o x y l a m m o n i u m Ringer with 52.9 mM h y d r o x y l a m m o n i u m ions gave an average reversal potential change o f - 1 9 . 1 + 1.7 inV. After correcting for concentration the permeability ratio is 0.94. Alt h o u g h the sodium channel is almost as permeable to h y d r o x y l a m m o n i u m as it is to sodium, the currents in h y d r o x y l a m m o n i u m Ringer are only half the size expected f r o m the same concentration o f Na + ions. T h e currents are e x p e c t e d to be 22% smaller just f r o m the low p H (5.79) o f the solution (Hille, 1968; Woodhull, 1973; Campbell and Hille, 1976). In addition the h y d r o x y l a m m o n i u m ions may also block channels directly. T h e h y d r o x y l a m m o n i u m Ringer was toxic to the fiber and m e a s u r e m e n t s in this solution had to be completed within a minute to avoid irreversible damage. In seven e x p e r i m e n t s hydrazinium Ringer lowered the reversal potential by 38.0 ± 1.3 inV. After c o m p e n s a t i n g for the concentration hydrazinium ions (68.9 raM) the permeability ratio is 0.31. Currents in hydrazinium are about 35% less than predicted by i n d e p e n d e n c e , while at p H 5.97, the blockage expected f r o m H + ions is only 16%. Fig. 2 also shows a positive shift in the depolarization r e q u i r e d to o p e n Na channels in both h y d r a z i n i u m and h y d r o x y l a m m o n i u m Ringer. This calcium-like effect is due in part to the low p H o f these solutions, although guanidinium Ringer shows a similar but smaller effect at p H 7.4.
Other Permeant Organic Cations A m m o n i u m , guanidinium, and a m i n o g u a n i d i n i u m are also p e r m e a n t to the muscle Na channel. A m m o n i u m Ringer r e d u c e d the reversal potential by 52.3 ± t.4 mV for a permeability ratio o f 0.11. Fig. 3 shows a series o f voltage clamp currents for a fiber in guanidinium Ringer. T h e middle trace is near the reversal O
I ms
2
5
I (mV)
-2OO c / (
¢ -40
mY )
. . . .
"~ '.
20[ S"'
"
:.C---.. 'V~ • / "
o ~-8 Sodium ~,Hydrozinium o Hydroxyrommonium
s:o,u Hydroxylommonium
,o[ b", ~,. "'
ABSTRACT The ionic selectivity of the Na channel to a variety of metal and organic cations is studied in frog semitendinosus muscle. Na channel currents are measured under voltage clamp conditions in fibers bathed in solutions with all Na + replaced by a test ion. Permeability ratios are calculated from measured reversal potentials using the Goldman-Hodgkin-Katz equation. The permeability sequence was Na + ~- Li + --~ hydroxylammonium > hydrazinium > ammonium > guanidinium > K÷ > aminoguanidinium in the ratios 1:0.96:0.94:0.31:0.11:0.093: 0.048:0.031. No inward currents were observed for Ca +÷, methylammonium, methylguanidinium, tetraethylammonium, and tetramethylammonium. The results are consistent with the Hille model of the Na channel selectivity filter of the node of Ranvier and suggest that the selectivity filter of the two channels is the same. INTRODUCTION T h e ionic selectivity o f axon sodium channels has been well described (Chandler a n d Meves, 1965; Hille, 1971, 1972, 1975b, c; Meves a n d Vogel, 1973). This p a p e r extends the study o f Na channel permeabilities to skeletal muscle using the muscle voltage clamp m e t h o d described in the first o f this series o f papers. In the past such experiments have been difficult to p e r f o r m on muscle for the lack o f a voltage clamp technique sufficiently fast to provide g o o d resolution o f Na channel currents that also permits the many necessary solution changes without disturbing the preparation. In nerve the Na channel is permeable to many small cations (Chandler a n d Meves, 1965; Tasaki et al., 1966; Hille, 1971, 1972). In particular, Hille was able to characterize the ionic selectivity at the frog n o d e o f Ranvier by systematically testing the permeability to an exhaustive list o f small organic and metal cations. H e p r o p o s e d that a 3.1 x 5.1-A oxygen-lined pore would explain the exclusion of all i m p e r m e a n t m o n o v a l e n t cations tested on simple geometric and chemical g r o u n d s (Hille, 1971, 1972, 1975 c). T h e observed selectivity sequence o f perm e a n t cations and deviations f r o m i n d e p e n d e n c e have been successfully described by a f o u r - b a r r i e r Eyring rate t h e o r y model o f the sodium channel that includes a negatively c h a r g e d site at this selectivity filter (Hille, 1975 b, c). As is shown here Na channel permeabilities observed for frog muscle are nearly identical to those observed for f r o g n o d e a n d are thus also consistent with this model. This work has a p p e a r e d in preliminary form (Campbell, 1974, 1975). THE JOURNAL
OF GENERAL
PHYSIOLOGY
" VOLUME
67, 1976 " pages 295-307
295
296
THE JOURNAL OF GENERAL PHYSIOLOGY" VOLUME 6 7 ' 1976 METHODS
Voltage Clamp Preparation The voltage clamp method is that described in the previous paper (Hille and Campbell, 1976) except experiments were performed before all the improvements described there had been made. Specifically, compensation for the attenuating effects of the impedance ZEB were not available. The abbreviations and nomenclature in Hille and Campbell (1976) are used throughout. A piece of a single muscle fiber is removed from the dorsal head of a frog semitendinosus muscle and placed across partitions separating the four pools of the recording chamber. Pool A, containing the area of m e m b r a n e u n d e r voltage clamp control, is filled either with normal Ringer or Ringer with all the sodium replaced by a test ion. The other three pools contain 120 mM CsF. Cutting the fiber ends in CsF eliminates mechanical activity and contributes to the electrical fidelity of the voltage clamp (see the preceding paper). Of importance to this study, cutting in CsF also raises the Na channel reversal potential, presumably by replacing some of the Na + and K ÷ ions inside the fiber with impermeant Cs + ions. The importance of a high reversal potential for measuring the permeability to relatively i m p e r m e a n t ions will be discussed later. Two feedback amplifiers are used. One is required for potentiometric recording of the m e m b r a n e potential which appears as --EM in pool A. T h e other amplifier supplies the end of the fiber in pool E with the m e m b r a n e current required to hold this potential to the voltage clamp command potential. Membrane current is not measured explicitly, but is assumed to be proportional to the voltage in the current-injecting pool (VE). The reservations in this assumption are discussed in the previous paper and in a later section on errors. All records were corrected electronically for leakage and capacity currents. The frequency response of the amplifiers used in these experiments is lower than shown in the first paper but still adequate for rapid voltage clamp control. Compensation tot capacity and leakage currents permits all other current to be measured starting about 80 tzs after the voltage step.
Recording and Analysis T e n to twenty current traces, in response to voltage clamp steps spaced about 7.5 mV apart, are recorded on a storage oscilloscope and then on film. Peak Na channel currents (measured as VE) are graphed to determine the reversal potential. Reversal potentials from bracketing runs in normal Ringer (ENa) are averaged and subtracted from the reversal potential measured in the test solution (Es). The resulting change in reversal potential is used to calculate the desired permeability ratio from the Goldman-HodgkinKatz voltage equation (Goldman, 1943; Hodgkin and Katz, 1949). The change in reversal potential between the two solutions is given by (Hille, 1971): Es - EN, = 2.303
(RT/F) log10 (Ps[S]/P~a[Na]),
where 2.303 RT/F is 55.2 mV at 5°C, [S] is the activity of test ion in test Ringer, [Na] is the activity of Na + in the control Ringer, and Ps/PNa is the ratio of channel permeabilities to the test and Na + ions. This way of determining Ps/PNa is used because unlike methods based on flux or current relations it does not assume independence and does not depend on the n u m b e r of channels open (Hille, 1971, 1975 c). It does assume that the selectivity of the channel is i n d e p e n d e n t of m e m b r a n e voltage and the solution bathing the fiber. These assumptions may not be true. For some models of the Na channel selectivity filter even the Goidman-Hodgkin-Katz voltage equation does not hold strictly true and permeability ratios may depend on other experimental conditions (Hille, 1975 b). Nevertheless,
297
DONALD T. CAMPBELL Muscle Na Channel Selectivity
the above method seems to be the most useful way to d et er m i n e selectivity o f a channel for comparison with other results as long as these reservations are kept in mind.
Activities For 0.1 mol/kg solutions at 25°C the activity coefficients are: LiCI, 0.790; NaCI, 0.778; KC1, 0.770 (Robinson and Stokes, 1965). For 0.07 M CaC12, the activity coefficient is 0.55. T h e Guggenheim convention (Butler, 1968; Shatkay, 1968) defines the single ion activity coefficient for Ca ++ as the square of the activity coefficient o f the salt, in this case 0.30. For the metal cations the activity coefficients are in the ratio 1.02:1.00:0.99:0.39 for Li +, Na +, K +, and Ca ++. These ratios are used with the cation concentrations to compute the activities required by the above equation. Since activity data are not available for the organic cations tested, their activity coefficients were assumed equal to that o f 0.1 mol/kg NaC1 for the calculation of permeability ratios.
Solutions T h e CsF solution used in the end pools contains 120 mM CsF with 1 mM imidazole or tris(hydroxymethyl)aminomethane buffer, pH 7.4. T h e composition of the test solutions is given in Table I. Test solutions generally contain 110 mM NaC1 or an osmotically equivalent concentration of a Na substitute, 2 mM CaCI~, and 4 mM tris(hydroxymethyl)aminomethane buffer, pH 7.4. For simplicity the solutions are named "Na Ringer," "Li Ringer," and so on according to the test ion. Hydrazine and hydroxylamine have pKa's below 8 and the pH of these solutions is adjusted to 5.97 and 5.79 to obtain a higher concentration o f the ionized species. T h e hydrazinium and h y d r o x y l a m m o n i u m Ringer also have slightly different concentrations of other ions as a result of dilution that occurred when adjusting them to their p H . T h e osmolarity o f the Na Ringer is 203 mosM, h y d r o x y l a m m o n i u m Ringer 154 mosM. All others tested had osmolarities within 5% of the Na Ringer value. TABLEI
C O M P O S I T I O N OF T E S T S O L U T I O N S Ringer
Anion
IS+]
Na 1/8 Na
CICI-, Br-
110 13.8
Li K
CICI-
110 110
Ca Hydroxylammonium
C1CI-
89.5 52.9
Hydrazinium
CI-
68.9
Ammonium Guanidinium Aminoguanidinium Methylammonium Methylguanidinium Tetramethylammonium Tetraethylammonium
CICINOf CISO4 BrBr-
Comment
Osmolarity maintained with 96.2 mM TMA Br. 4 mM CsCI added for some experiments. pH = pH pH = pH
5.79; [S+] calculated from and [S]totat; [Ca ++] = 1.6. 5.97; [S+] calculated from and [S]totaJ; [Ca ++] = 1.3.
110 110 110 110
73.3 110 110
Except as noted, all solutions also contain 2 mM CaCI2, 4 mM tris(hydroxymethyl)-aminomethane buffer and have a pH between 7.3 and 7.4.
298
THE JOURNAL OF GENERAL PHYSIOLOGY'VOLUME 6 7 ' 1976
Errors The previous paper describes several c u r r e n t - d e p e n d e n t voltage errors. At maximum sodium currents these errors may total 20 mV, but since the net m e m b r a n e current at the Na channel reversal potential is roughly 1/10 the peak value an error of only about 2 mV is expected. In addition, the method described below for measuring attenuation artifact includes this error, therefore the attenuation correction simultaneously reduces currentdependent errors in reversal potential measurement. Another error might be expected to arise from contributions of Na channels in the membrane of the transverse tubular system which presumably is u n d e r poor voltage clamp control. T h e voltage clamp calculations illustrated in Fig. 12 of the first paper show that at least for the model chosen there is a negligible effect of tubular sodium conductance on the reversal potential. The current signal in these experiments was not corrected for the capacitative component of the current-injecting pathway ZED as described in the first paper. Because this error is also proportional to the m e m b r a n e current it should not greatly affect the measurement of reversal potential. In fact, all of the dynamic errors taken together apparently have little effect on measured reversal potential changes. For instance, when the A pool solution is changed from normal Na Ringer to Ringer with 7/8 of the NaCI replaced by the i m p e r m e a n t salt tetramethylammonium (TMA) bromide, the average change in reversal potential measured in 16 experiments is -48.1 mV at 5°C while the theoretical change predicted by the Nernst equation is -49.8 mV. This is the method used to determine the a m o u n t of "attenuation artifact," a voltage error found when this voltage clamp method is used to study the node of Ranvier (Dodge and Frankenhaeuser, 1958; 1959; Hille, 1971). In that preparation the attenuation may be 10-20%. For the muscle preparation the attenuation averaged 3.5%, and the average reversal potential changes used to calculate permeability ratios were corrected by this amount. This 1.7-mV deviation from the Nernst prediction, if artifactual, implies that the combination of all errors has only a very small effect on the reversal potential change measured between these two solutions. The reversal potential changes are corrected for the differences in junction potential that the control and test Ringer solutions make with the agar bridge in pool A. These potentials are measured with respect to a Beckman 38402 ceramic junction saturated KC1 reference electrode (Beckman Instruments, Inc., Fullerton, Calif.). T h e measured junction potentials are given in Table II. For most solutions the correction is less than 2 inV. RESULTS
T h e o b s e r v e d c h a n g e s in r e v e r s a l p o t e n t i a l are listed in T a b l e I I . F o r p e r m e a n t ions the a v e r a g e r e v e r s a l p o t e n t i a l c h a n g e s are c o r r e c t e d for a t t e n u a t i o n a n d j u n c t i o n p o t e n t i a l s a n d t h e n u s e d to calculate p e r m e a b i l i t y ratios. R e v e r s a l p o t e n t i a l c h a n g e s a r e e x p r e s s e d as m e a n + SEM. I n s o l u t i o n s s h o w i n g n o i n w a r d c u r r e n t s , n o r e v e r s a l p o t e n t i a l c a n be m e a s u r e d . F o r these i m p e r m e a n t ions the p o t e n t i a l t h a t first gives a m e a s u r a b l e t r a n s i e n t o u t w a r d c u r r e n t is t a k e n as the u p p e r limit o f the reversal p o t e n t i a l a n d u s e d to calculate a limit to the p e r m e a b i l i t y ratio. I n the case o f i m p e r m e a n t ions the lowest limit o n p e r m e a b i l ity is r e p o r t e d r a t h e r t h a n a n a v e r a g e .
Lithium S o d i u m c h a n n e l s o f m u s c l e a r e a b o u t as p e r m e a b l e to l i t h i u m as they a r e to s o d i u m . I n Fig. 1, the voltage c l a m p series in Li R i n g e r is n e a r l y i d e n t i c a l to the
1
46.7 0.84 + 1.4
48 48 49 48 42 46 49 46 37 43 49 49 48 48 48 49 0.0 0.55 + 0.5 0.5 0.96
1.5 1.5 - 1.5 0.0 0.0 - 1.5
73. I 2.96 -2.4 73.2 0.048
72* 75* 78*
81"
72 57
77
K
Ca
AND
+4.2 >84.2 77
>61 >68 >75
>66 >77 >54
CHANGES
19.3 1.7 - 0.8 19.1:[: 0.94
22 18
22 15
37.0 1.3 + 1.0 39.3~ 0.31
34 41 38
31 38 37 40
H~drazinium
PERMEABILITY Hydroxylammonium
54 55
49
51
NH4
54.6 2.2 + 0.5 57.0 0.093
69 48 52 57 71 63 57 49 45 60 57 57 48 51 55 55
34
CHANNEL
OF
80.5 1.4 -0.2 83.1 0.031
84 77 81 80
>115
>103 >98 >115 >89 >112
>104 >101 >112 >103 > 109 >92
TMA
TEA
>109
>109 >92 >109 >94
MUSCLE
-4.5 +0.9 +2.3 -O.9 >111.9 >114.4 >116.9 >115.2 97 >109
>100
>95
> 109
AminoMethylMethylguanidinium, ammonium guanidinium
IN SODIUM
Guanidinium
RATIOS
52.3 1.4 - 1.8 52,3 0.11
II
* T h e K Ringer in these experiments also contained 4 mM CsCl as explained in the text. Corrected also for concentration o f ionized species. For i m p e r m e a n t ions the largest limit on reversal potential change was used to c o m p u t e the limit for the permeability ratio. Potentials are in millivohs.
MEAN SEM ]unction potential After corrections 0s/PNa
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
1/8 Na
Fiber
Li
POTENTIAL
REVERSAL
TABLE
t~
0 Z
300
THE JOURNAL OF GENERAL PHYSIOLOGY" VOLUME 6 7 ' 1976
1 [mY) - 100
-50
,
0
ms
I
2
5
5 I0 ~ E [m V)
°t 0
-2oo
-
"---~ ~
~;(/ m'x,,., X ~ / "
/ Lithium
- 400L-
FIGURE I. Lithium currents in the muscle Na channel. (Left) Peak current-voltage relations in Na and Li Ringer. The smooth curves were drawn by hand. Although the reversal potential in the two solutions is the same, the maximum currents in Li Ringer are about 20% less than in Na Ringer. (Right) Voltage clamp series from the same experiment showing currents for voltage steps spaced 15 mV apart. Fiber ends were cut in 120 mM CsF. Fiber 5, 5°C. control series in Na Ringer. T h e c o r r e s p o n d i n g current-voltage relation shows almost no change in reversal potential u p o n substituting the Li Ringer. In six experiments Li Ringer lowered the reversal potential 0.5 ± 0.6 mV, for a permeability ratio PLi/Pya o f 0.96. A l t h o u g h the muscle Na channel is a b o u t as permeable to Li + ions as it is to Na + ions, the current-voltage relation o f Fig. 1 shows that the currents in Li Ringer are 16-20% less than in Na Ringer. Evidently Li + ions block Na channels in addition to passing t h r o u g h them. In other words Li + ions violate the i n d e p e n d e n c e principle (Hodgkin and Huxley, 1952; Hille, 1975 c) in the muscle Na channel. As is shown below, some o t h e r p e r m e a n t ions show similar deviations from i n d e p e n d e n c e .
Potassium Potassium is the only other metal ion tested f o u n d to be measurably p e r m e a n t in the Na channel. It is also one o f the most difficult to test. After a minute or two in K Ringer with C1- as the anion, the muscle swells and the m e m b r a n e often develops large irreversible leaks destroying the p r e p a r a t i o n . In later experiments preparations bathed in isotonic K2SO4 survived well. In addition, the high external concentration o f potassium increases the potassium permeability o f other channels giving rise to large rectifying b a c k g r o u n d currents that must be subtracted to determine currents in Na channels. For this reason four experiments were p e r f o r m e d in K Ringer with 4 mM CsC1 a d d e d to help keep these extra potassium conductances blocked. In Table II the reversal potential changes f r o m these experiments are noted with an asterisk (*). In seven experiments the reversal potential in K Ringer was 73.2 ± 3.0 mV lower than in Na Ringer, c o r r e s p o n d i n g to a permeability ratio PK/Pya o f 0.048. Sodium is about 21 times more p e r m e a n t than potassium in the Na channel o f muscle.
DONALD T . CAMPBELL
301
Muscle Na Channel Selectivity
Hydroxylammonium and Hydrazinium T w o o f the most p e r m e a n t sodium substitutes tested are h y d r o x y l a m m o n i u m and hydrazinium. Voltage clamp series and c u r r e n t voltage relations f r o m a fiber bathed in these solutions are shown in Fig. 2. In four e x p e r i m e n t s , changing the bathing m e d i u m to h y d r o x y l a m m o n i u m Ringer with 52.9 mM h y d r o x y l a m m o n i u m ions gave an average reversal potential change o f - 1 9 . 1 + 1.7 inV. After correcting for concentration the permeability ratio is 0.94. Alt h o u g h the sodium channel is almost as permeable to h y d r o x y l a m m o n i u m as it is to sodium, the currents in h y d r o x y l a m m o n i u m Ringer are only half the size expected f r o m the same concentration o f Na + ions. T h e currents are e x p e c t e d to be 22% smaller just f r o m the low p H (5.79) o f the solution (Hille, 1968; Woodhull, 1973; Campbell and Hille, 1976). In addition the h y d r o x y l a m m o n i u m ions may also block channels directly. T h e h y d r o x y l a m m o n i u m Ringer was toxic to the fiber and m e a s u r e m e n t s in this solution had to be completed within a minute to avoid irreversible damage. In seven e x p e r i m e n t s hydrazinium Ringer lowered the reversal potential by 38.0 ± 1.3 inV. After c o m p e n s a t i n g for the concentration hydrazinium ions (68.9 raM) the permeability ratio is 0.31. Currents in hydrazinium are about 35% less than predicted by i n d e p e n d e n c e , while at p H 5.97, the blockage expected f r o m H + ions is only 16%. Fig. 2 also shows a positive shift in the depolarization r e q u i r e d to o p e n Na channels in both h y d r a z i n i u m and h y d r o x y l a m m o n i u m Ringer. This calcium-like effect is due in part to the low p H o f these solutions, although guanidinium Ringer shows a similar but smaller effect at p H 7.4.
Other Permeant Organic Cations A m m o n i u m , guanidinium, and a m i n o g u a n i d i n i u m are also p e r m e a n t to the muscle Na channel. A m m o n i u m Ringer r e d u c e d the reversal potential by 52.3 ± t.4 mV for a permeability ratio o f 0.11. Fig. 3 shows a series o f voltage clamp currents for a fiber in guanidinium Ringer. T h e middle trace is near the reversal O
I ms
2
5
I (mV)
-2OO c / (
¢ -40
mY )
. . . .
"~ '.
20[ S"'
"
:.C---.. 'V~ • / "
o ~-8 Sodium ~,Hydrozinium o Hydroxyrommonium
s:o,u Hydroxylommonium
,o[ b", ~,. "'
