South Parks Road, Oxford OXI 3QT

Transcription

1 J. Physiol. (1974), 238, pp With 6 text-figures Printed in Great Britain AN ESTIMATE OF SODIUM/POTASSIUM PUMP ACTIVITY AND THE NUMBER OF PUMP SITES IN THE SMOOTH MUSCLE OF THE GUINEA-PIG TAENIA COLI, USING [3H]OUABAIN BY ALISON F. BRADING AND J. H. WIDDICOMBE From the University Department of Pharmacology, South Parks Road, Oxford OXI 3QT (Received 13 June 1973) SUMMARY 1. Binding of tritiated ouabain to the smooth muscle of the guinea-pig taenia coli showed two components, one saturable at lower glycoside concentrations and the other linear with increasing concentrations. 2. The saturable component alone was affected by extracellular potassium concentrations. This component seems to be bound to sodium pumping sites, and when completely saturated binds 1 1 x 111 molecules per mg fresh wt. of tissue, or 25-3 molecules per square micron of membrane, assuming a volume: surface area ratio of 1.5 /im. 3. Only a fraction of 42K uptake by the cells can be blocked by ouabain at maximal concentrations. In normal Krebs solution two thirds can be blocked. The remaining one third is presumably passive uptake. The fraction blocked is reduced as the extracellular potassium concentration, and thus passive uptake, is increased. 4. The amount of potassium pumped into the cells at various concentrations of extracellular potassium has been calculated. In normal Krebs solution the amount pumped in 45 min was 2- m-mole/kg fresh wt., and this was increased at higher potassium concentrations. 5. On the assumption of a stoichiometry of 3Na: 2K, the pump sites in normal Krebs solution have a turnover rate of 132 min'. 6. Indirect calculations of sodium movements suggest that the sodium permeability may be about -9 x 18 cm sec-1 and the pump may generate a current of -9 x 1-7 A cm-2. This crossing an Ohmic membrane resistance of 3-6 kq cm2 would be equivalent to a potential difference of 3-5 mv. INTRODUCTION Estimates of the passive permeability of smooth muscle cell membranes to sodium ions have proved very difficult to make, because of the rapid movement of this ion species across the cell membrane, and the complexity

2 236 ALISON F. BRADING AND J. H. WIDDICOMBE of the mechanisms by which this movement is achieved (Brading, 1973; Casteels, Droogmans & Hendrickx, 1973). Estimates of PNa for taenia coli in the literature range from about 7 x 1- cm sec-1 (Brading, 1971) to 1*8 x 1-8 cm sec-1 (Casteels, 1969). It is probable that neither of these estimates gives a reasonable value, and the true permeability may be intermediate. Membrane potentials calculated by means of the Goldman constant field equation with the lower value for the permeability are of the same order as the measured values, but the figure used by Brading is the minimum justifiable one, and probably an underestimate. The potentials calculated with the higher PNa are substantially lower than the experimentally determined values. Thus Casteels (1969) gives a predicted potential of -37 mv as compared with a measured value of -57 mv. This has led to the postulate that there is a significant contribution of electrogenic sodium pumping to the membrane potential. Further calculations by Casteels, Droogmans & Hendrickx (1971 b) based on the flux estimates of Casteels, have shown that an electrogenic component of up to -22 mv could be generated if all the ions actively pumped in this tissue are uncoupled and generate a current across the cell membrane. It is possible that Casteels has over-estimated the sodium permeability, and that it is not necessary to have to postulate such a large electrogenic component. There is however other evidence which supports some electrogenic contribution to the resting membrane potential, and that is the depolarization that occurs following the removal of potassium ions or the addition of ouabain (Tomita & Yamamoto, 1971; Casteels et al. 1971a). These procedures produce depolarizations of the order of 5 mv. Because it has not proved possible to get direct estimates of sodium permeability that are at all convincing, this paper is concerned with the use of indirect methods to give some indication as to what the true passive sodium permeability may be. Recent work with 3H-labelled ouabain has enabled estimates to be made of the number of sodium pumping sites in various tissues (see Baker & Willis, 1969, 1972a, b; Landowne & Ritchie, 197). There seems to be a relationship between the passive permeability of the membrane to sodium ions and the number of pump sites. The present paper describes an attempt to label pump sites in the taenia coli, and makes estimates of the pump activity from inhibition by ouabain of 42K uptake. These estimates of the density of the pump sites, and activity of pumping suggest that the passive sodium permeability is rather low, and probably intermediate between the two values quoted above. Furthermore, our results suggest that any electrogenic contribution is small, being of the order of 5 mv.

3 Na PUMP SITES IN SMOOTH MUSCLE 237 METHODS White guinea-pigs were killed by a blow on the head and bled out. Pieces of taenia were dissected from the caecum, and the fresh weight recorded. The tissue samples, weighing about 2 mg, were mounted on stainless-steel holders, and placed in Krebs solution for at least 1 hr before being transferred into an experimental solution. The Krebs solution used had the following composition (mm): Na , K+ 5 9, Ca2+ 2-5, Mg2+ 1-2, Cl , HCO3-15-5, H2P4-1-2 and glucose 11-5, equilibrated with a gas mixture of 3 % CO2 and 97 % 2 at 36 C. Alterations in the potassium concentration were achieved by substitution with sodium. Measurement of binding Tritiated ouabain was obtained from New England Nuclear Corporation. The sample (specific activity 13 c/m-mole) was supplied in an ethanol: benzene 9:1 solution, which was evaporated to dryness in a stream of nitrogen. The tracer was then diluted with normal saline containing 1-8 M-non-active ouabain, and stored in a refrigerator. For measurements of ouabain uptake, a preliminary experiment was carried out (see Fig. 1) to establish the time course of uptake, and to investigate the effect of different washing techniques. In one set of tissues each tissue was rinsed for 2 sec in an individual test tube of tracer free Krebs solution at room temperature before preparation of the tissue for counting. In the other set, after a 2 sec rinse, tissues were transferred to an organ bath of Krebs solution at 5 C for 5 min, and then to a second bath of Krebs solution at 5 C for a further 5 min. This was designed to give adequate time for loss of extracellular tracer, whilst slowing down the release of bound ouabain. The results show that only the tracer that would be expected in the extracellular fluid was lost, and this procedure was adopted for the rest of the experiments. An exposure time of 1 hr was chosen, since uptake seemed to be relatively complete after this time. After exposure of the tissues to the [3H]ouabain-containing solutions at various concentrations of ouabain, they were rinsed as described, blotted and weighed. They were then placed in glass counting vials, dissolved in 1 ml. NEN protocol (5 C overnight), neutralized with glacial acetic acid and counted in a liquid scintillation counter using Brays scintillation fluid. Small aliquots of the radioactive loading medium were then added to the same vials, and they were recounted and the counts in the medium were obtained by subtraction. Measurement of 42K uptake Tissues were equilibrated for 1 hr in Krebs solution containing the desired potassium concentration, and then pre-incubated for half an hour in a similar solution to which the appropriate ouabain concentration had been added. The tissues were then placed in a 42K solution containing the same ouabain and potassium concentrations. Forty-five minutes later they were removed and washed for 2 min in non-active solution (removing an estimated 75 % of the extracellular counts). They were then blotted and weighed, and counted in a gamma scintillation counter, to determine the uptake of potassium by the tissue. Aliquots of the bathing solution were also counted, and analysed for potassium content, and the uptake of the tissues was expressed as m-mole K+/kg fresh wt of tissue. II PHY 238

4 238 ALISON F. BRADING AND J. H. WIDDICOMBE 18 3r -` 14 L. Wbo E 13-1 ki -v C C D Time (min) Fig. 1. Time course of uptake of [3H]ouabain. The upper points are from tissues rinsed for 2 sec before preparation for counting, and the lower points from tissues washed for 1 min in two changes of Krebs solution at 5 C. The difference between the two lines is equal to the counts that would be present in 42 % of the tissue volume. RESULTS Binding of 3H ouabain Fig. 1 shows the time course of tracer uptake at 36 C, obtained with the two different rinsing procedures described in the Methods. The curves can be seen to be separated by a constant amount, equivalent to 66 cpm/ mg fresh wt. of tissue. These counts would have been dissolved in a volume equivalent to 42 % of the tissue weight, a figure reasonably close to the extracellular space, indicating that not a lot of bound ouabain was released during the 1 min exposure to cold Krebs solution. The curves also indicate that the uptake is nearly complete after an exposure to the tracer of 1 hr. The uptake of radioactive ouabain from solutions of differing ouabain concentration was measured in normal Krebs solution, and in solutions containing different amounts of potassium. Fig. 2 shows the results obtained from several different experiments on different days. Each point is the mean of about six tissues. It is clear from the Figure that the uptake consists of at least two components, one at lower ouabain con-

5 Na PUMP SITES IN SMOOTH MUSCLE 239 centrations which is saturable, and dependent on the extracellular potassium concentration and a second at higher concentrations which appears linear, and is unaffected by potassium. At low concentrations of ouabain potassium ions appear to reduce the uptake of labelled ouabain, which is consistent with this fraction being bound to the pump sites. 114 _ -c, to.e 1'12 O Cd~~~~~ 11 '-' 19 A 18 l I I I I ' Ouabain concentration (m) Fig. 2. Binding of [3H]ouabain to tissues at different glycoside concentrations, and in Krebs solution containing zero potassium (), one tenth of the normal potassium (.6 mm *), and the normal potassium (6 mm, A). Tissues were pre-treated for 1 hr in Krebs solution with the desired potassium at 36 C before exposure to a similar solution containing [3H]ouabain for a further 1 hr. The results are plotted with logarithmic axes. Fig. 3 is an analysis of one particular experiment in which the uptake of tritiated ouabain was measured in a zero potassium solution. The results are plotted on linear co-ordinates. The straight line has the slope of the linear regression of all the individual points at ouabain concentrations of 1O- and 1-5 M (off scale on Fig. 3). It is assumed that at these high ouabain concentrations the saturable binding is negligible in comparison with the linear component, and so the line is drawn with a zero 11-2

6 24 ALISON F. BRADING AND J. H. WIDDICOMBE intercept. The bar represents plus to minus the standard deviation of the slope at that concentration of ouabain. The triangles are the points of the total uptake minus the linear component, and the curve is a Langmuir binding curve with a Ke of 8 x 18 M and approaching complete saturation at 1.1 x 111 molecules of ouabain per mg fresh wt. 1-c _15x111 bo E E C 51' X Ouabain concentration (M) Fig. 3. Binding of [3H]ouabain in zero potassium Krebs solution. Results plotted with linear axes. The upper points () are the total ouabain bound. The straight line has been drawn to pass through the origin, and has the slope of the regression line calculated through the individual values at 1-71 x 1O-5 and 1-71 x 14M ouabain. The bar is calculated using the standard deviation of the slope at that concentration. The lower points (A) are calculated by subtracting the linear component from the total binding. The curve is a Langmuir plot with a K of 8 x 1- m. 42K uptake Experiments were carried out to measure the total uptake of 42K by the cells of tissues exposed for 45 min to varying concentrations of ouabain and potassium. The majority of extracellular counts were removed by a 2 min rinse, which was assumed to remove 75 % of the extracellular counts (a to of extracellular exchange of 1 min). Ouabain progressively decreased the uptake at each potassium concentration until at a certain level no further depression occurred on increasing the ouabain concentration. At these levels the active uptake of potassium is presumably completely blocked, and the total uptake is now passive, and varies with

7 Na PUMP SITES IN SMOOTH MUSCLE 241 the potassium concentration in the external medium. In Fig. 4 the figures are plotted out against the logarithm of the ouabain concentration. The passive component has been subtracted from the total uptake in each case, and the results scaled with 1 % as the uninhibited uptake. The same sigmoid curve has been fitted by eye to the points in '8 mm-k+, 6 mm-k+ and the points obtained in 3 and 6 mm-k+ (these last two not being clearly differentiable). The parallel shifts of the curve are to be expected if potassium competes for the same sites as the ouabain molecules. 1 A 8 6- C Z 4- c Ouabain concentration (M) Fig. 4. The effect of potassium on the log dose-response curve of inhibition ofpotassium uptake by ouabain. Ordinate: % inhibition of active potassium uptake. Abscissa: logarithm of the ouabain concentration. Increasing potassium up to 3 mm shifts the curve to the right., -8 nm-k+; *, 6 mm-k+; A, 3 mm-k+; A, 6 mm-k+. Comparison of binding and pump inhibition Since it is not possible with this type of experiment to obtain a curve of 42K uptake in zero potassium solutions, Fig. 5 compares the inhibition of active uptake with the saturable component of [3H]ouabain binding in the presence of one tenth the normal potassium concentration. The continuous line drawn through the binding points is a Langmuir binding curve with a Ke of 1-2 x 17 M, with 99 % saturation at 1-21 x 111 molecules bound per mg fresh wt. The increased value of Ke necessary to fit the points in the presence of potassium is consistent with a competitive interaction between ouabain and potassium molecules. The active uptake of 42K approaches complete inhibition at concentrations similar to those that achieve nearly complete saturation of the binding sites, and

8 242 ALISON F. BRADING AND J. H. WIDDICOMBE the two curves intersect fairly near to 1P2 x 1-7 M and 6 x 11 molecules which is the point where they should theoretically intercept if 5 % inhibition is achieved with 5 % of the pump sites bound to ouabain molecules. This provides further evidence that the saturable binding sites and the sodium pumping sites are the same. jo (~~~~~~~~~~~~~~~~~~~ C 11 I ~~~~~~~~~~~~~ X Ouabain concentration (M) Fig. 5. A comparison of the saturable uptake of [3H]ouabain (A) with the percentage inhibition of potassium uptake () in solutions with one tenth the normal potassium concentration (.6--8 mm). The left-hand ordinate gives the molecules of ouabain bound per mg fresh wt. of tissue, and the right-hand ordinate the percentage inhibition of potassium uptake. The curve through the ouabain binding points is a Langmuir plot with a Ke of 1-2 x 1-7 M. Each point is the mean of six determinations. Estimates of pump rate (a) from uptake measurements The difference between the total cellular uptake of 42K and the passive uptake in tissues totally blocked by ouabain is illustrated in Fig. 6. The uptake is calculated from the counts in the tissue after 45 min exposure to the isotope, and will underestimate the real influx, since tracer will have entered and left the cells during this time. Allowance can be made for this if it is assumed that the exchange of potassium follows a simple exponential time course. Then the amount of labelled potassium, K*, taken up after any time is given by d-t k [K* K*], (1) where k is the rate constant for potassium loss from the tissue, K* is the final steady-state amount of labelled potassium taken up and K* is the actual amount taken up. kk* is constant and is in fact equal to

9 Na PUMP SITES IN SMOOTH MUSCLE 243 the initial rate of uptake of labelled potassium. Integration of (1) leads to the equation K* = K*(1e-ekt). (2) Table 1 contains values of k calculated from the uptake of labelled potassium after 45 min (K45) and also values of K* for various extracellular potassium concentrations. The amount of labelled potassium which would have entered and remained in the cells at time t if there 6 e, - Total E S. 4 2 IS Pasiv K concentration (mm) Fig. 6. Cellular uptake of 42K during 45 min exposure to solutions of various [K].t. The upper points are the uptake in uninhibited tissues. The lower points are the uptake in tissues where active pumping is completely blocked by ouabain. had been no efflux is given by kk*t and the values which would have been observed after 45 min have been calculated and are given in Table 1. These figures are the total influx that has occurred and can be divided into passive and active uptake. The percentage of this influx that is active can be calculated from the results shown in Fig. 6. In normal Krebs solution the measured uptake after 45 min was 26-6 m-mole/kg fresh wt., and the corrected total influx was 33-2 m-mole/kg fresh wt. The maximum suppression of the uptake by ouabain suggests that 6-2 % of the influx is actively pumped, which means that 2- m-mole/kg fresh wt., were pumped in 45 min. This corresponds to 1-86 p-mole cm-2 sec' if the volume: surface area ratio is 1-5 jm.

10 244 ALISON F. BRADING AND J. H. WIDDICOMBE I:4 C)t cs E- Q- C oo~~~~c e 4a -- O su -Sno _- - C X,>CS ~~~~~

11 Na PUMP SITES IN SMOOTH MUSCLE 245 (b) From efflux measurements It is also possible to estimate the amount of potassium pumped entirely from previous published values. Casteels (1969) gives a rate constant of '18 min', and Brading (1971) gives values of -17 or -112 mindepending on the assumptions made. These values are very similar to those calculated from the 45 min uptake in this paper, which gives a constant of 13 min- in normal Krebs solution. From these rate constants, and the intracellular potassium concentration, the efflux can be calculated and from this, on the basis of the constant field theory, PRK If the tissue is in a steady state, and values have been calculated for the permeability of the membrane to potassium, the constant field theory can be used to estimated the passive influx since Mi KR [K X VFIRT 1-exp (-VVF/RT) where PR is the permeability, M, the influx, [K]o the extracellular concentration (in M. cm-3) and the rest of the equation is equal to for influx of a +ve ionic species, with a resting potential of 5 mv at 37 C. Published estimates of PK range from 6'71 x 1-8 cm sec-1 (Brading, 1971) to 11 x 1-8 cm sec-' (Casteels, 1969). The corresponding total fluxes are 2-6 and 4 p-mole cm-2 sec-1. For an extracellular concentration of 6 mm potassium the passive influx comes to p-mole cm-2 sec-1 and, on the assumption that the efflux and influx are equal in the steady state, the pumped influx will be 1P p-mole cm-2 sec-1 or 18* m mole/kg fresh wt. in 45 min. The ratio of pumped to passive influx from the above figures is 1P9: 1 or 1-74: 1, similar to the figures for uptake in this paper, which indicate a ratio of 1P51: 1 pumped to passive potassium uptake. DISCUSSTON The over-all binding of tritiated ouabain to the cells of the taenia coli in the guinea-pig can be described as the sum of two processes, a saturable component which is affected by the external potassium concentration, and probably reflects the binding of ouabain to the sodium/potassium pump sites, and a second linear component which is not sensitive to the extracellular potassium concentration, and is similar to the 'non-specific' binding described in other tissues. In zero external potassium, the saturable component has an equilibrium constant of about 8 x1-8 M

12 246 ALISON F. BRADING AND J. H. WIDDICOMBE ouabain, and at a ouabain concentration of 1P7 x 1-6 M (96 % occupancy) binds about 111 molecules of ouabain per mg fresh weight. With potassium present in the medium, the amount bound at this concentration is less, but saturation of the binding sites is probably significantly less than complete. Freeman-Narrod & Goodford (1962) have calculated that the volume to surface area ratio in the taenia coli is 1-5 /im. If one assumes an extracellular space of 4 % (Brading & Jones, 1969), then each milligram of fresh tissue will have a membrane area of 4 cm2 (this can be compared with the mammalian C fibres which have a surface area of 6 cm2/mg, Keynes & Ritchie, 1965). For 1% binding this leads to an estimate of 275 molecules bound to the saturable sites for each square micron of membrane. Consideration of the errors involved suggests that there are between 25 and 3 sites/,tm2, which can be compared with an estimate of 75 for mammalian C fibres (Lindowne & Ritchie, 197). Baker & Willis (1972a) in their paper on ouabain binding to intact cells, conclude that the bulk of cells so far examined have between 5 and 15 pumping sites/fcm2 membrane, with the exception of the red blood cell which probably has less than one per square micron. Since mammalian C fibres have a volume to surface area ratio similar to the cells of the taenia coli, it is interesting that the smooth muscle cells appear to have only one third of the number of pump sites. If the inward current in the action potential of the C fibres is carried by sodium ions, as in other nerves, and if all or part of the inward current in the taenia coli action potentials is carried by calcium ions, as has been suggested (Brading, Bulbring & Tomita, 1969), this could be one explanation for the smaller number of pump sites in the smooth muscle. The other obvious possibility is that the passive membrane permeability of the C fibres to sodium is greater than is that of smooth muscle but there is no evidence on this point. Since it is very difficult to make direct estimates of the amount of sodium pumped under normal conditions by the taenia coli, the pump rate can only be measured satisfactorily from the potassium uptake. This requires assumptions as to the coupling of the two pumped ion movements. The experiments on 42K uptake described in this paper, and the effect of ouabain on this uptake suggest that about one third of the steadystate influx of potassium is passive, and two thirds active. The pumped influx in normal Krebs solution was calculated to be 1-86 p-mole cm2 sec-1. This agrees well with the indirect estimates made from efflux studies, which suggest that between 1-71 and 2-54 p-mole cm-2 secpotassium is pumped across the cell membrane in the steady state. The uptake results with varying extracellular potassium concentrations

13 Na PUMP SITES IN SMOOTH MUSCLE 247 suggest that in normal Krebs solution the extracellular ionic environment is not quite optimal for maximum pump activation, since more potassium is pumped when the extracellular potassium is increased at the expense of sodium ions. In the high potassium solutions the intracellular sodium is probably reduced or unchanged, and so the increased activation is probably due to a more complete potassium ion activation of the mechanism, or to a reduction of the possible inhibitory effect of extracellular sodium on the pump (Casteels et al. 1973). Estimates of the pump density and the potassium influx suggest that each site moves about 44 potassium ions per second, or 264 potassium ions per minute. If the pump moves potassium and sodium ions with a stoichiometry of 2K: 3Na as suggested for many tissues (see Thomas, 1972) and this movement is achieved for each activation of a pump site, then the turnover rate would be about 132 min-. This is presumably not a maximum since it is probable that with increased intracellular sodium higher pumping rates are achieved. One situation in which the intracellular sodium is probably elevated is during exposure to high concentrations of acetylcholine. Bolton (1973) has studied this situation in the smooth muscle of the guinea-pig ileum, and has estimated from the contribution of sodium pumping to the potential achieved during exposure to acetylcholine, that the maximal pump current is probably 16 A cm-2. If this current is carried by one third of the sodium pumped, this figure is equivalent to 312 p-mole sodium pumped per cm2 per second, and a turnover rate of 15,66 min-. Maximum turnover rates in other tissues so far studied (Baker & Willis, 1972a) seem to lie between about 35 and 15, min', which compares well with this figure for smooth muscle. In the unstimulated taenia coli with a 3:2 Na: K coupling the total pumped sodium would be about 2-79 p-mole cm-2 sec1, and the current carried by one third of this, 93x 1-7 A cm2. This current flowing through an ohmic membrane resistance of 3-6 kq cm2 (Tomita, 197) would be equivalent to a potential difference of between 2*7 and 5-4 mv. This is substantially less than the figure of 22 mv referred to earlier, but that estimate was based on a passive sodium influx of 7 p-mole cm-2 sec-l, and on the assumption that all the sodium ions pumped contribute to the current. If the passive efflux of sodium is considered negligible compared to the pumped efflux, then it is possible to calculate indirectly the steadystate Na permeability. This comes to about -92 x 1-8 cm sec-1, a value intermediate between the two extremes of 7 x 1- (Brading, 1971) and 1-81 x 1- cm sec-1 (Casteels, 1969). The numerical values presented in the above discussion rely heavily

14 248 ALISON F. BRADING AND J. H. WIDDICOMBE on the estimate of the volume:surface area ratio used. The published data available, which result in the figure of 1-5,tm (Freeman-Narrod & Goodford, 1962) were obtained from measurements using light microscopy, where the small unevennesses of the surface area are hard to measure. Recent unpublished work (Gillian Wootton, personal communication) from electronmicrographs of carefully prepared taenia coli muscle suggests that a more realistic figure for the ratio would be about -6,gm or less. If this figure is correct, the estimates of pump density and membrane permeability given above will be over-estimated by a factor of nearly three. This work was supported by grants from the Medical Research Council. REFERENCES BAKER, P. F. & WILLIS, J. S. (1969). Number of sodium pumping sites in cell membranes. Biochim. biophys. Acta 183, BAKER, P. F. & WILLIS, J. S. (1972a). Binding of the cardiac glycoside ouabain to intact cells. J. Physiol. 224, BAKER, P. F. & WILLIS, J. S. (1972b). Inhibition of the sodium pump in squid giant axons by cardiac glycosides: dependence on extracellular ions and metabolism. J. Physiol. 224, BOLTON, T. B. (1973). Effects of electrogenic sodium pumping on the membrane potential of longitudinal smooth muscle from terminal ileum of guinea-pig. J. Physiol. 228, BRADING, A. F. (1971). Analysis of the effluxes of sodium, potassium and chloride ions from smooth muscle in normal and hypertonic solutions. J. Physiol. 214, BRADING, A. F. (1973). Ion distribution and ion movements in smooth muscle. Phil. Trans. R. Soc. B 265, BRADING, A. F., BULBRING, E. & TOMITA, T. (1969). The effect of sodium and calcium on the action potential of the smooth muscle of the guinea-pig taenia coli. J. Physiol. 2, BRADING, A. F. & JONES, A. W. (1969). Distribution and kinetics of CoEDTA in smooth muscle, and its use as an extracellular marker. J. Physiol. 2, CASTEELS, R. (1969). Calculation of the membrane potential in smooth muscle cells of the guinea-pig's taenia coli by the Goldman equation. J. Physiol. 25, CASTEELS, R., DROOGMANS, G. & HENDRICKX, H. (1971 a). Membrane potential of smooth muscle cells in K-free solution. J. Physiol. 217, CASTEELS, R., DROOGMANS, G. & HENDRICKX, H. (1971 b). Electrogenic sodium pump in smooth muscle cells of the guinea-pig's taenia coli. J. Physiol. 217, CASTEELS, R., DROOGMANS, G. & HENDRIC1KX, H. (1973). Effect of sodium and sodium-substitutes on the active ion transport and on the membrane potential of smooth muscle cells. J. Physiol. 228, FREEMAN-NARROD, M. & GOODFORD, P. J. (1962). Sodium and potassium content of the smooth muscle of the guinea-pig taenia coli at different temperatures and tensions. J. Physiol. 163,

15 Na PUMP SITES IN SMOOTH MUSCLE 249 KEYNES, R. D. & RITCHIE, J. M. (1965). The movements of labelled ions in mammalian non-myelinated nerve fibres. J. Physiol. 179, LANDOWNE, D. & RITCHIE, J. M. (197). The binding of tritiated ouabain to non. myelinated nerve fibres. J. Physiol. 27, THOMAS, R. C. (1972). Electrogenic sodium pump in nerve and muscle cells. Physiol. Rev. 52, TOMITA, T. (197). Electrical properties of mammalian smooth muscle. In Smooth Muscle, ed. BULBRING, E., BRADING, A. F., JONES, A. W. & TOMITA, T. London: E. Arnold. TOMITA, T. & YAMAMATO, T. (1971). Effects of removing the external potassium on the smooth muscle of guinea-pig taenia coli. J. Physiol. 212,