| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
1 Physiological Laboratory, Department of Anatomy, 2 Multi-Imaging Centre, Department of Anatomy and 3 Department of Pharmacology, University of Cambridge, Cambridge, UK
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
2.0) and intracellular [Na+], [K+] and [Cl-]. Increased extracellular osmolarities produced hyperpolarizing shifts in Em in fibres studied in Cl--free Ringer solution consistent with the Goldman-Hodgkin-Katz (GHK) equation. In contrast, fibres exposed to hypertonic Ringer solutions of normal ionic composition showed no such Em shifts, suggesting a Cl--dependent stabilization of membrane potential. This stabilization of Em was abolished by withdrawing extracellular Na+ or by the combined presence of the Na+–Cl- cotransporter (NCC) inhibitor chlorothiazide (10 µM) and the Na+-K+-2Cl- cotransporter (NKCC) inhibitor bumetanide (10 µM), or the Na+-K+-ATPase inhibitor ouabain (1 or 10 µM) during alterations in extracellular osmolarity. Application of such agents after such increases in tonicity only produced a hyperpolarization after a time delay, as expected for passive Cl- equilibration. These findings suggest a model that implicates the NCC and/or NKCC in fluxes that maintain [Cl-]i above its electrochemical equilibrium. Such splinting of [Cl-]i in combination with the high PCl/PK of skeletal muscle stabilizes Em despite volume changes produced by extracellular hypertonicity, but at the expense of a cellular capacity for regulatory volume increases (RVIs). In situations where PCl/PK is low, the same cotransporters would instead permit RVIs but at the expense of a capacity to stabilize Em.
(Received 28 November 2003;
accepted after revision 16 December 2003;
first published online 23 December 2003)
Corresponding author C. L.-H. Huang: Physiological Laboratory, University of Cambridge, Cambridge, UK. Email: clh11{at}cam.ac.uk
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The present experiments extended the above reports. First, they introduced an assessment of cell volume (V) using confocal microscope scanning of amphibian skeletal muscle fibres in the xz-plane over time. This offered significant advantages over earlier measurements that only assessed steady-state fibre diameter in response to extracellular osmotic change (cf. Dydynska & Wilkie, 1963; Blinks, 1965). The cylindrical geometry of skeletal muscle fibres made it possible to simplify earlier confocal microscopy measurements of cell volumes in cardiac myocytes that required three-dimensional image reconstructions from serial xy-scans in successive z-planes (Satoh et al. 1996). Secondly, the volume changes were compared with electrophysiological measurements of cell resting potentials in fibres studied not only in Cl--free but also in Cl--containing and Na+-free extracellular Ringer solutions. Thirdly, these results were mapped onto predictions from a simple Goldman-Hodgkin-Katz (GHK) analysis of the resting potential (Goldman, 1943; Hodgkin & Katz, 1949). This suggested that the resting potential was stabilized by an elevation of [Cl-]i/[Cl-]o above its expected equilibrium level despite hypertonic cell shrinkage in normal, but not Cl--free, Ringer solutions. Fourthly, the consequences both of a withdrawal of extracellular [Na+] and pharmacological manoeuvres using cation-Cl- cotransport inhibitors implicated Na+-Cl- and Na+-K+-2Cl- cotransporters (NCC and NKCC) in this stabilization (Gamba et al. 1993).
The findings reconciled the apparent inability of skeletal muscle to perform RVIs in response to increased extracellular tonicity with reports of bumetanide-sensitive Na+ influxes under similar conditions (Clausen et al. 1979; Dorup & Clausen, 1996) and bumetanide effects upon Em in both isotonic and hypertonic solutions (van Mil et al. 1997; Geukes Foppen et al. 2002). They suggested a model in which ion fluxes through the NCC and the NKCC elevate [Cl-]i despite its tendency to passively dissipate down its electrochemical gradient (Hodgkin & Horowicz, 1959) under conditions of extracellular hypertonicity
This splinting of [Cl-]i by cation-Cl- cotransporters would stabilize Em at its resting value in skeletal muscle in hypertonic solutions due to a high membrane PCl/PK, thereby countering the tendency of an increase in [K+]i to result in membrane hyperpolarization. However, this would restrict the ability of muscle to perform RVIs. In contrast, other cell types with a low PCl/PK, such as erythrocytes (Tosteson & Hoffman, 1960), would use such cotransporters to perform RVIs but not stabilize Em.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Solution A (chloride (Cl-)-free, sulphate (SO42-) Ringer solution): 75 Na2SO4, 1.25 K2SO4, 8 CaSO4, 3.0 Hepes
Solution B (normal Ringer solution): 115 NaCl, 2.5 KCl, 1.8 CaCl2, 3.0 Hepes.
Solution C (Na+-free, choline Ringer solution): 115 choline chloride, 2.5 KCl, 1.8 CaCl2, 3.0 Hepes.
Studies of cell volume used whole cutaneous pectoris muscles from cold-adapted winter frogs, Rana temporaria, killed by concussion followed by pithing (Schedule 1: Animal Procedures Act, Home Office, U.K), dissected and pinned out in isotonic Ringer solution B. Preliminary experiments confirmed that the properties of cutaneous pectoris fibres were similar to those of the amphibian sartorius muscle used on earlier occasions (e.g. Adrian, 1956), including relatively large diameters and resting potentials, and a capacity for regenerative action potential activity. The muscles were mounted, ventral side uppermost, in a 0.5 ml volume chamber at sarcomere lengths of between 2.5 and 2.8 µm on a coverslip that formed the floor of the chamber. This arrangement permitted free flow of fluid around the ventral aspect of the muscle, while its dorsal aspect remained in contact with the coverslip, permitting imaging using an inverted confocal microscope. The muscles were studied in the presence of the fluorescent membrane-impermeant dye sulforhodamine B (Lissamine rhodamine B200: 75%, Aldrich, UK) added to the bathing solution at a concentration of 62.5 µg ml-1 (Fraser et al. 1998). This remains in the extracellular space and does not affect membrane electrophysiological properties (Gallagher & Huang, 1997), providing a continuous vital stain for the fibre margins throughout imaging in the course of the osmotic stress procedures.
Fibre volume could then be determined from cross-sectional areas measured using xz confocal scanning of amphibian cutaneous pectoris muscle fibres. The muscles were scanned every 10–30 s in the xz-plane using a Zeiss LSM-510 laser scanning confocal microscope (incorporating a Axiovert 100M inverted microscope) using the x40 oil immersion objective, to give a frame size of 240 µm along the x-axis and 120 µm along the z-axis. Lissamine rhodamine fluorescence was activated using a 543 nm wavelength laser line. Fluorescence emission was captured at > 560 nm using a long pass filter set. Initial scanning in the xy-plane ensured that the x-axis ran perpendicular to the fibre axis. Cutaneous pectoris is an extremely thin muscle, one to three fibres thick, that has been employed previously in the study of end-plates (e.g. Ceccarelli & Hurlbut, 1975) and so solution changes applied from the ventral aspect of the muscle would reach its dorsal aspect placed against the viewing coverslip. Preliminary confocal microscopy in the xz-plane showed that the cross-sectional areas of the dorsal fibres varied from approximately 300 to nearly 1500 µm2, with transverse diameters from 20 to 160 µm. The majority of fibres run parallel to one another. Staining of the extracellular compartment by Lissamine rhodamine permitted imaging of transverse ‘virtual slices’ of groups of several muscle fibres in which the fibre cross-sections appeared dark against a fluorescent extracellular space. This made it possible to obtain sequential images of fibre cross-sections. As muscle lengths were held constant throughout the experiments, this provided a robust and continuous indication of fibre volume over the course of the osmotic manoeuvres.
The cutaneous pectoris muscle fibres were subjected to hypertonic Cl--free, normal and Na+-free Ringer solutions (solutions A, B and C, respectively). A single muscle was successively washed with and left for at least 30 min at each of the five concentrations of hypertonic sucrose Ringer solution (mM): 10, 20, 50, 100, 200, corresponding to external osmotic pressures of 260, 270, 300, 350, 450 mosmol 1-1, respectively. Fibre volume fully recovered on replacement of isotonic Ringer, even after exposure to 100 mM sucrose.
In-house image analysis software was then used to calculate fibre cross-sectional areas. The images were first corrected for intensity variation in the z-direction, as fluorescence signal attenuation was greater in the deeper areas of the fibre than in those areas more superficial. The recordings from the deepest areas were corrected by applying a x4 multiplication factor to the brightness, progressing linearly to unity correction in the most superficial areas. A simple noise filter was then applied. Finally, a threshold intensity was set, and image pixels lighter than the threshold were transformed to white and those darker to black. The software then calculated the area of each individual fibre in each frame. This method enabled consistent detection of volume changes as small as 2–5% corresponding to diameter changes of <2%, and was a significant improvement on the accuracy of previous methods (Dydynska & Wilkie, 1963; Blinks, 1965).
The latter computations incorporated results from control experiments that calibrated distances determined by xz-scanning and the Zeiss LSM software. First, experiments to correct for length measurements orthogonal to the z-direction used 25 µm diameter fluorescent microspheres (Calbiochem, UK) within the same experimental chamber as those where muscles were examined. These were attached to a micropipette, in turn mounted on a Luigs & Neumann (Ratingen, Germany) micromanipulator with a calibrated z-axis excursion. Each microsphere was moved along the z-axis in 12 µm increments from a position in which the microsphere was in contact with the coverslip and therefore closest to the inverted microscope objective, to a distance that more than covered the z-range including the greater part of the muscle cross-sections. Their apparent maximum diameters as a function of z-position were then measured and plotted. Figure 1A (
) demonstrates that the measured maximum diameter varied little with z-position; it was approximated by a line given by diameter (µm) =-0.026z+ 27, corresponding to a consistent offset plus a <10% change over a full-frame excursion of over 100 µm. In our experiments fibres were in contact with the coverslip and thus were unable to move significantly apart from the small changes in diameter caused by the cellular volume changes. Thus the maximum z-excursion observed was approximately 10 µm.
|
) demonstrates that the estimated z-distance was 1.31 times the actual z-distance but followed a linear relationship that permitted a consistent z-correction throughout the distances over which our measurements were made. Thirdly, viable muscle fibres under the experimental solutions adopted above were moved along the z-axis from their normal position against the viewing coverslip, and their normalized cross-sectional areas as determined by confocal xy-scanning determined under conditions of constant osmolarity. Figure 1B demonstrates that at such excursions this did not significantly alter the measurements of fibre cross-sectional area.
The above observations thus led us to consider that at least over the range of conditions under which our experiments were taking place and for our specific microscope, the relationship between measured and actual distances along the z-axis was consistent and the error in x with changing z position was small. Nevertheless, in the present experiments all fibre cross-sectional areas were normalized to a control value obtained in the same fibre examined in the isotonic solution before sucrose was introduced. The analysis here was primarily concerned with (a) changes in volume relative to the volume obtained in isotonic solution and (b) the presence or absence of time-dependent regulatory volume adjustments that might take place after an initial volume change rather than their absolute magnitude. Figure 2A and B shows typical images, prior to intensity corrections and image processing for measurements of cross-sectional areas, of fibre profiles before and after an imposed osmotic manoeuvre, and in which the viewing coverslip forms the upper edge of each image.
|
1.5 times their in situ length as described on earlier occasions for electrophysiological studies (Koutsis et al. 1995). The muscle was mounted in the bath in isotonic Ringer solution to give centre sarcomere lengths of 2.5–2.8 µm, similar to those used for the experiments using cutaneous pectoris, as measured using an eyepiece graticule through a x40 water immersion objective. Bath temperature was controlled at 5–10°C by circulating cooled water through a glass coil placed in the chamber using a Minipuls 3 peristaltic pump (Gilson, France). A digital thermometer (J. Bibby Science Products, UK) was used to monitor the temperature near the muscle. Measured volumes of isotonic and hypertonic solutions were added to or withdrawn from the bath using syringes mounted at opposite ends of the bath.
The initial Em of successive adjacent fibres was measured using standard glass capillary microelectrodes (3 M KCl; resistance 5–20 M
; tip potentials <5 mV: Adrian, 1956) against the bath voltage as followed by reference electrodes; both were connected to the recording electronics via bridges containing normal (Cl--containing) Ringer solution in contact with balanced Ag–AgCl reference junctions. Depending on the variability of the values, the Em values of up to 15 fibres were recorded. At the end of a 30 min period in hypertonic Ringer, the steady-state Em values in up to 30 adjacent muscle fibres were recorded. This protocol was performed using the same concentrations of hypertonic sucrose Ringer as used for the cutaneous pectoris muscle in the volume studies. A different muscle was used for each sucrose concentration. Em values were obtained from the means and S.E.M. values of single impalements made in a large number of fibres before and after each solution change, as it was difficult to sustain prolonged (>60 min) impalements of single fibres, particularly over such solution changes. In the latter event, Em values then tended to deteriorate with the alterations in fibre volume produced by these solution changes. The protocol was repeated in Na+-free and Cl- -free Ringer solutions.
As the electrophysiological experiments were performed in sartorius muscle fibres as originally adopted by Adrian (1956), in which it was possible to obtain means and standard errors of resting potential values from large numbers of fibres within a single preparation, control volume measurements were also made in sartorius muscle. These provided steady-state results in agreement with those obtained using cutaneous pectoris and the earlier reports (data not shown, but see also Dydynska & Wilkie, 1963; Martin et al. 2003). However, access to the tissue of solution changes in the sartorius preparation was significantly less rapid than was achieved from both sides of the muscle in the thinner (1–3 fibres thick) cutaneous pectoris and was also impeded for surface sartorius fibres viewed in an inverted confocal configuration. These features slowed the time course of the osmotic responses of sartorius fibres following solution changes. In addition, whereas the fine layer of connective tissue in cutaneous pectoris resulted in the muscle fibres being separated by a z-distance of <10 µm from the coverslip, the thicker layer of connective tissue in sartorius resulted in a separation of 
40 µm.
Together, the confocal microscope and electrophysiological experiments made it possible to compare V and 
Em following osmotic manipulations. Pharmacological experiments used the ion cotransport inhibitors bumetanide (10 µM) and chlorothiazide (10 µM), and the Na+–K+-ATPase inhibitor ouabain (1 and 10 µM) (Sigma Chemical Co., Poole, Dorset). Bumetanide and chlorothiazide were dissolved in DMSO at concentrations of 10 mM and 100 mM, respectively. Controls in which the relevant quantity (always <0.1% v/v) of DMSO vehicle alone was added to the bathing solutions did not significantly alter membrane potential (P>> 0.05).
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The experiments demonstrated that fibre volume changes closely agreed with both earlier work and expectations from simple osmotic behaviour. They extended those earlier findings to fibres studied in a range of ionic extracellular and pharmacological conditions (Dydynska & Wilkie, 1963; Blinks, 1965). The method of determining fibre volumes introduced here offered several advantages over earlier measurements that assessed steady-state fibre diameter (cf. Dydynska & Wilkie, 1963; Blinks, 1965). First, the xz-scans revealed that muscle fibre cross-sections assumed a wide variety of unpredictable triangular, square or even crescentic shapes (Fig. 2; see also Blinks, 1965). Secondly, volume reductions often resulted in rounded profiles of the fibre cross-section (Fig. 2A) simply becoming more angular (Fig. 2B): simple measurements of distances between easily defined points in the fibre profiles, as exemplified by the vertical cursors in Fig. 2 would then have underestimated the true changes in volume. Thirdly, full xz-scans could be obtained rapidly (in <1 s) in an experimental chamber that permitted rapid solution changes and simultaneous imaging. Accordingly, volume changes in response to altered osmotic conditions could easily be followed in serial scans with a time resolution as brief as 2 s, hitherto not achievable in earlier studies. It was thus possible to identify a given fibre and track its change in cross-sectional area through time. Fourthly, cutaneous pectoris muscle consists of a single sheet of fibres, one to three fibres thick, which permits clear visualization of fibre cross-sectional areas. Cutaneous pectoris was also used in previous volume studies with which the present findings could be compared (Dydynska & Wilkie, 1963). Finally, all the muscle fibres in the sample were subject to an identical osmotic procedure so it was possible to superimpose their traces and calculate mean changes in volume.
Figure 3A shows typical changes in cross-sectional area in six individual fibres following perfusion with a range of hypertonic normal Ringer solutions of progressively increased osmolarity and additionally confirms the ability of the muscle to recover its initial volume following return to isotonic Ringer. The solutions were added in the following order (Fig. 3A: arrows): 10 mM, 20 mM, 50 mM and 100 mM sucrose, wash-out with isotonic Ringer, second wash-out with isotonic Ringer, 200 mM sucrose. Figure 3B shows that, despite their wide range of initial cross-sectional areas that accordingly traversed different z-axis distances from the coverslip, cross-sectional areas for individual fibres varied linearly with extracellular [1/osmolarity].
|
200 s (see averaged traces to a larger timescale: Fig. 4B). Fibre volumes did demonstrate small (<3%) though often statistically insignificant evidence of subsequent regulatory volume increases (RVIs). Thus introduction of 100 mM sucrose Ringer resulted in an initial fall in normalized volume to 0.681 ± 0.0097, but this then rose slightly by 3% to its final steady-state value of 0.715 ± 0.0101 (n= 6; significant to a level of P < 0.01). Full recovery on return to isosmotic conditions took place, even after almost 3 h in hyperosmotic solution.
|
Results of the kind shown in Fig. 4 made it possible to derive plots of mean steady-state normalized volume against 1/osmolarity as adapted from Dydynska & Wilkie (1963), Blinks (1965) and Martin et al. (2003) on earlier occasions, in fibres studied in the different solutions. Figure 5 summarizes mean (±S.E.M.) results for muscles in Cl--free, normal and Na+-free Ringer solutions in the presence of different sucrose concentrations. The filled circle (•) represents the volume on return to isotonic normal Ringer solution. It confirms that the volume changes reversed on return to the isotonic solution to a normalized volume of 0.97 ± 0.005 (n= 6). The lack of extracellular Cl- in SO42- Ringer depletes the muscle of intracellular Cl- (Hodgkin & Horowicz, 1959), its major permeant anion, thereby preventing any major anion or cation movement across-the membrane. This would be expected to cause the muscle membrane to behave as a K+ electrode (Adrian, 1956). In contrast, replacement of extracellular Na+ ions in choline Ringer solution would limit movement of cations other than K+ across the membrane.
|
Solution A (Cl--free Ringer solution); 
: slope = 0.23 ± 0.023 osmol l-1; P= 0.0021; r= 0.985;
Solution B (normal Ringer solution); 
: slope = 0.25 ± 0.026 osmol l-1; P= 0.0007; r= 0.978.
Solution C (Na+-free Ringer solution); 
: slope = 0.23 ± 0.017 osmol l-1; P= 0.0009; r= 0.992.
Finally, further points obtained from fibres studied in 100 mM sucrose-containing normal Ringer solution gave values of mean normalized volume (given as means and S.E.M.s below) that suggested that muscle fibres continued to conform to such simple osmotic predictions following treatment with ouabain (0.71 ± 0.018, n= 3), as well as bumetanide, whether in the absence (0.68 ± 0.014, n= 6) or presence of chlorothiazide (0.71 ± 0.026, n= 4) (all at 10 µM; all represented by the filled square (), Fig. 3; to be compared with 0.71 ± 0.011, n= 6 in the absence of such agents).
Similar Em values in different isotonic Ringer solutions
The steady-state fibre volumes measured here thus showed simple variations with extracellular hypertonicity and relatively little evidence of appreciable time-dependent regulatory volume increases (RVIs), in contrast to the situation in some other, often non-excitable, cells, in agreement with earlier reports (see Discussion). This would predict corresponding variations in the concentrations of those intracellular ions that are conserved through such volume changes and a hyperpolarization of a cell membrane potential that is primarily determined by its K+ permeability, PK, at least where Cl- was either absent or always at electrochemical equilibrium. The experiments that follow were prompted by and extend observations from Cl--depleted systems (Adrian, 1956). They first examined steady-state resting potentials of muscle fibres studied in a range of isotonic Ringer solutions.
Em values obtained under isotonic conditions were statistically indistinguishable from each other (P > 0.05) in such Cl--free (-76.7 ± 2.75 mV; n= 14), normal (-74.2 ± 1.22 mV; n= 102) and Na+-free (-72.2 ± 1.33 mV, n= 71) Ringer solutions. They were slightly hyperpolarized (-81.5 ± 2.71 mV; n= 17; P < 0.05) in a reduced (80 mM) Na+ Ringer solution in which the NaCl was replaced by an isosmotic sucrose concentration. Comparisons of the Em values in normal and Cl--free Ringer solution were compatible with established PNa/PK values of 0.02–0.03 (Adrian, 1956; Hodgkin & Horowicz, 1959; Hodgkin & Horowicz, 1959; Hutter & Noble, 1960; Harris, 1965) in the Goldman-Hodgkin-Katz (GHK) equation, assuming Cl- to be at electrochemical equilibrium and values for intracellular (i) and extracellular (o) ion concentrations of: [Na+]i= 10 mM (Marunaka, 1988), [Na+]o= 115 mM (see Methods: solution B), [K+]i= 139 mM (Adrian, 1956; Hodgkin & Horowicz, 1959); [K+]o= 2.5 mM, [Cl-]i= 5 mM (based on a range of 3–7 mM: Hodgkin & Horowicz, 1959; Vaughan-Jones, 1982; Barry & Dulhunty, 1984) and [Cl-]o= 120 mM. Additionally, Em values obtained in normal, Na+-reduced and Na+-free Ringer solutions, allowing for previous reports of choline permeation through Na+ channels (Renkin, 1961; Mullins & Noda, 1963; Spindler et al. 1998), were compatible with previously reported PCl/PK ratios in the region of 2.
The dependence of Em on extracellular osmolarity in Ringer solutions of different ionic composition
Muscle fibres were then exposed to 30 min incubations with Ringer solutions (A–C) to which were added the following sucrose concentrations (mM): 10, 20, 50, 100 and 200; this gave extracellular osmolarities of 260, 270, 300, 350 and 450 mosmol l-1, respectively. The measured Em at each osmolarity was calculated as the difference between the mean steady state Em in isotonic Ringer and the corresponding value after 30 min in the test solution, and the results were plotted against 1/osmolarity ( in Fig. 6). They were first superimposed upon predicted Em values that assume that intracellular K+ content is conserved but Cl- is either absent (cf. Adrian, 1956) or passively distributed across the plasma membrane (Hodgkin & Horowicz, 1959)( and dashed lines) through the changes in fibre volume produced by osmotic manipulation (cf. Fig. 5). In the absence of other regulatory mechanisms for K+ and Cl- transfer, and PNa/PK values of 0.02–0.03 (see above), fibre membrane potentials in isotonic and hypertonic (h) Ringer solutions close to (RT/F)(ln([K+]o/[K+]i)) and (RT/F)(ln([K+]o/[K+]ih)), respectively, would give Em= (RT/F)(ln([K+]i/[K+]ih)) = (RT/F)(ln(Vhypertonic/Visotonic)), where V denotes the cell volumes (means ±S.E.M.) under those corresponding conditions and R, T and F have their usual physical meanings. The calculated points (means: S.E.M.s smaller than the dimensions of the data points) indicated that increasing osmolarity would hyperpolarize Em along a linear relationship with a positive slope of 7.79 ± 0.569 mV osmol l-1 (r= 0.990, P value for slope <0.0001; dashed lines in Fig. 6).
|
Em expected from the GHK equation following fibre exposure to the highest tonicities using 200 mM sucrose Ringer solution. This assumes that all the major intracellular ions (Na+, K+ and Cl-) are conserved and so all their concentrations would proportionally increase with increasing extracellular tonicity. It explored PCl/PK values between 0.1 and 10 (Fig. 7, abscissa) for PNa/PK (; from top to bottom) values between 0 and 0.1. This is in accordance with earlier published PNa/PK (0–0.03: Adrian, 1956; Hodgkin & Horowicz, 1959) and PCl/PK (2) values (Hodgkin & Horowicz, 1959; Hutter & Noble, 1960; Harris, 1965), and the accepted values of intracellular ionic concentrations for fibres initially in isotonic solutions, as indicated above.
|
Em against [1/osmolarity] in muscle fibres in Cl--free Ringer solution showed that Em hyperpolarized with increasing extracellular osmolarity. The regression line through Fig. 6A had a slope of 6.62 ± 1.99 mV osmol l-1 (regression coefficient r= 0.88 and P < 0.05 for a significant slope). The experimental values showed no significant differences from predicted Em values (P > 0.05 at all osmolarities examined; Fig. 6A) assuming Cl- is absent or in electrochemical equilibrium. Similarly, the mapping of the Em value resulting from an extracellular 200 mM sucrose Ringer solution onto Fig. 7 was consistent with the GHK predictions for a fibre showing a markedly reduced PCl/PK(<0.1: box a) and compatible with the values expected in Cl--free Ringer solutions, and confirmed earlier reports (Adrian, 1956).
Fibres in normal Ringer solution (Fig. 6B) gave contrasting results. The regression line through the experimental Em values had an insignificant slope (3.67 ± 4.15 mV mosmol l-1; regression coefficient r= 0.40; P= 0.43). Each of the steady-state Em values after 30 min were significantly more depolarized than the predictions arising from Cl- being at electrochemical equilibrium (two-tailed t test significance levels: 10 mM: P < 0.01, 20 mM: P < 0.001, 50 mM: P < 0.001, 100 mM: P < 0.001, 200 mM sucrose: P < 0.01). However, the small (3 mV) mean value of Em obtained in 200 mM sucrose Ringer solution did map onto the plot of Em against(PCl/PK) (Fig. 7) in a region compatible with accepted PCl/PK values (1.5–2.5: see above; box b). This suggests that cellular Cl- content was conserved despite cell shrinkage; this would result in an increase in [Cl-]i/[Cl-]o above expectations from electrochemical equilibrium. The resulting change in the Cl- Nernst potential would stabilize Em through the high membrane PCl/PK.
Figure 6C plots results obtained in a Na+-free Ringer solution with otherwise unaltered [K+]o and [Cl-]o. Yet, in contrast to Fig. 6B, the relationship between Em and 1/osmolarity did show a significant slope of 7.86 ± 1.43 mV osmol l-1 (significantly different from zero to P < 0.01; regression coefficient r= 0.94). Experimental values showed no significant differences from the Em values expected (P > 0.05 at all osmolarities examined; Fig. 6C) if Cl- were in electrochemical equilibrium. Similarly, the mapping of the Em (-15 mV) value obtained in the presence of Na+-free 200 mM sucrose Ringer solution onto Fig. 7, assuming a PCl/PK(1.5–2.5) as in the previous case, was now incompatible with a conserved intracellular Cl- content even for PNa/PK= 0 (box c). Rather, it suggested that Cl- had re-equilibrated in muscle fibres studied in Na+-free Ringer solutions.
Ion transport mechanisms involved in stabilizing Em
The above results were compatible with the hypothesis in which [Cl-]i/[Cl-]o was conserved to values that were above expectations from electrochemical equilibrium in fibres exposed to hypertonic, normal Ringer solutions. This would stabilize Em despite fibre shrinkage owing to the relatively high PCl/PK known to exist in skeletal muscle. The results suggested that this process depended upon a Cl- transport mechanism that was itself dependent upon extracellular Na+, for there are two such possible processes in skeletal muscle membrane. First, the thiazide-sensitive Na+-Cl- cotransporter (NCC) (Hebert et al. 1996; Monroy et al. 2000) permits large transcellular NaCl fluxes in renal tubular epithelia, and may also function in skeletal muscle (Dorup & Clausen, 1996). Secondly, bumetanide-sensitive Na+-K+-2Cl- cotransporters (NKCC) occur in cultured myocytes (Liedke, 1992) and rat skeletal muscle (Wong et al. 1999; Lindinger et al. 2002) and have been implicated in amphibian skeletal muscle vacuolation following osmotic stress (Khan et al. 2000). Hypertonicity is known to stimulate a bumetanide-sensitive 22Na+ influx in rat skeletal muscle (Clausen et al. 1979; Chinet, 1993). Both these systems may rely upon ionic gradients (Lindinger et al. 2002) that are in turn dependent upon Na+–K+-ATPase, whose activity is reduced by cardiac glycosides such as ouabain (Schwartz et al. 1975).
Our experiments demonstrated that addition of a combination of chlorothiazide and bumetanide, or of ouabain, prior to increases in extracellular tonicity abolished the stabilization of Em, implicating both NCC and NKCC. Figure 8A shows that the addition of (from left to right) 10 µM chlorothiazide (agent1), 10 µM bumetanide (agent2) and a combination of both of these (agent3) produced no significant (P > 0.05) membrane potential changes in fibres studied in control isotonic solutions. Figure 8B summarizes results obtained 30 min following exposure to hypertonic (200 mM) sucrose Ringer solution conducted in the presence of these agents. There were no shifts in Em from values obtained in isotonic solutions in the presence of either chlorothiazide or bumetanide. However, the inclusion of both diuretics resulted in significant (P < 0.001) hyperpolarization. This suggests that both the NCC and NKCC can contribute to stabilization of Em in hypertonic solutions. Finally, Fig. 8C shows results of adding the diuretic agents 30 min after the onset of the hypertonic challenge. This had no immediate effect on Em such that the findings initially paralleled those in isotonic Ringer. However, membrane potentials were also obtained after a further delay of 30 min, the required time interval expected for Cl- equilibration (Hodgkin & Horowicz, 1959). The group treated with the bumetanide–chlorothiazide combination then demonstrated a significant (P < 0.01) hyperpolarization of -8.11 ± 2.73 mV (n= 42 fibres).
|
10 mV depolarization in isotonic Ringer solution (10 µM: Fig. 8A; agent 4), as also reported by Marunaka (1988), it permitted a significant (P < 0.01) hyperpolarization following hypertonicity (10 µM: Fig. 8B; agent 4). This is compatible with suggestions (Lindinger et al. 2002) that NKCC and possibly NCC activity is sensitive to Na+ gradients established by Na+-K+-ATPase. Similarly, the capacity of cells to stabilize Em was impaired when extracellular NaCl was replaced isotonically with sucrose to produce a [Na+]o of 80 mM. A hypertonic challenge then produced a significant (P < 0.05), 8.7 ± 4.0 mV (n= 36 fibres), hyperpolarization of Em. A simple model of the effects of Na+-driven Cl-cotransport systems
Figure 9 illustrates the results of a finite difference modelling of changes in V(%), Em(mV), [Na+]i, [K+]i and [Cl-]i (mM) (ordinates) brought about by NKCC activity following the initial passive changes in these variables induced by changing from isotonic to hypertonic extracellular solution. This model sought to assess specifically whether NKCC activity in combination with a particular cell membrane PCl/PK might account for the steady-state relationship between Em and V following such an osmotic manipulation. It assumed that (a) the normalized fibre volume follows simple osmotic predictions for a cell whose membrane is freely permeable to water whilst conserving its intracellular contents of Na+, K+, and Cl-, (b) Em can be predicted from values of intracellular and extracellular [Na+], [K+], and [Cl-] using the GHK equation, and (c) fluxes through the NKCC markedly exceed passive Cl- effluxes, thereby enabling NKCC activity to maintain [Cl-]i/[Cl-]o above expected equilibrium values. No assumptions were made about the rate constants for NKCC operation apart from the condition that each cycle of NKCC activity produces increments in [Na+]i, [K+]i and [Cl-]i in the ratio 1: 1: 2 and remaining transmembrane ion movements were at least one order of magnitude slower.
|
438 mosmol l-1) challenge imposed from the starting point of an isotonic (238 mosmol l-1) extracellular solution. These subsequent adjustments were simulated in a stepwise fashion, with each iteration cycle progressively increasing [Na+]i, and [K+]i each by 10 µM in unit fibre volume; the consequences of successive cycles are represented in mM along the abscissae of Fig. 9. Alterations in step size over a range from 1 µM to 1 mM made no difference to the overall results of the computational procedure. The relative change in fibre volume necessary to maintain transmembrane isotonicity was calculated and then used to adjust the concentration of each intracellular ion at each step. The predicted value of Em was also calculated using the GHK equation and the iteration then repeated. Figure 9 plots the changes in volume, Em and [Na+]i, [K+]i and [Cl-]i as determined over successive cycles. It shows a progressive increase in [Na+]i and [Cl-]i but a paradoxical decrease in [K+]i owing to dilution by accompanying water entry, as modelled by the increase in cell volume. The changes in ionic concentrations cause a net depolarization largely accounted for by alterations in [Cl-]i/[Cl-]o, as the percentage changes in [Cl-]i are much greater than those of [K+]i in view of the different absolute intracellular concentrations of these ions.
However, the relationship between changes in Em and V varied strikingly with PCl/PK. Figure 9A shows that when PCl/PK is 2, a full correction of Em by 10–12 mV is only accompanied by a 2% increase in V, and therefore can entirely account for the present experimental results. In contrast, when PCl/PK is reduced to 0.2 (Fig. 9B), the NKCC activity that produces the same increase in volume (2%) does so with a profoundly smaller effect (<1 mV) on Em. Similar modelling for NCC activity, which would increase [Na+]i and [Cl-]i in the ratio 1: 1, revealed qualitatively similar effects concerning the relationship between ionic concentrations, fibre volume and Em, since this transporter also increases [Na+]i, [Cl-]i and volume while reducing [K+]i. Thus, if NKCC/NCC activity were to be restrained by Em, the high PCl/PK that occurs in skeletal muscle would stabilize Em but permit little adjustment in volume. In contrast, when PCl/PK is low such as in smooth muscle or erythrocytes (Tosteson & Hoffman, 1960; review: Chipperfield & Harper, 2000), the minimal changes in Em would not limit NKCC activity, thereby allowing large RVIs.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The experiments subjected amphibian skeletal muscle fibres to extracellular hypertonicity in Cl--free, normal and Na+-free Ringer solutions. They introduced confocal microscopy xz-scanning techniques for measurements of fibre cross-sectional areas. The volume determinations were calibrated against control measurements of reference distances that were made both orthogonal to and along the z-direction. The latter established consistent and linear refractive corrections for measurements in the xz-plane, through the required z-distances. Such techniques thus provided accurate and rapid determinations of fibre cross-sectional areas and therefore V, over the time course of the osmotic manoeuvres. All values were normalized to control readings made in isotonic solutions as the present analysis primarily concerned fractional changes rather than absolute values of V to determine the presence or absence of appreciable time-dependent adjustments following the initial responses of V to the osmotic manoeuvres.
The measurements confirmed that in amphibian skeletal muscle V monotonically decreased with increased external tonicity and that there was minimal or no evidence of the regulatory volume increases (RVIs) observed in other cell types (Tosteson & Hoffman, 1960; Lang et al. 1998a,b; Chipperfield & Harper, 2000). Such simple osmotic behaviour was consistently observed in fibres studied in normal, Cl--free or Na+-free Ringer solutions or the presence of the Na+-K+-2Cl- (NKCC) and Na+-Cl- (NCC) transport inhibitors bumetanide and chlorothiazide and the Na+-K+-ATPase inhibitor ouabain.
The corresponding electrophysiological measurements of cell resting potentials from muscle fibres in Cl--free solutions first confirmed previous reports (Adrian, 1956) that the relationship between V and Em closely follows simple predictions of the Goldman-Hodgkin-Katz (GHK) equation provided that all intracellular ions are conserved following volume change, and that cellular volume reduction proportionally increases the concentrations of all intracellular ions.
The investigation was then extended to fibres in more physiological, Cl--containing, normal Ringer solutions to permit passive Cl- redistributions (Hodgkin & Horowicz, 1959) following exposures to extracellular hypertonicity over the time course of the experiments (>15 min). Cell shrinkage would then be expected to decrease [Cl-]i owing to the more negative equilibrium potential expected from an increased [K+]i. Instead, Em values obtained in hypertonic solutions continued to follow a GHK relationship in which all intracellular ions were conserved assuming previously established baseline intracellular [Na+], [K+], and [Cl-] and PCl/PNa and PCl/PK values for fibres in isotonic solutions (Hodgkin & Horowicz, 1959; Hutter & Noble, 1960; Harris, 1965). The latter were corroborated by Em measurements in fibres studied in isotonic, control conditions. The findings thus suggested that a significant and maintained increase in [Cl-]i/[Cl-]o and a consequent divergence of the equilibrium potentials for Cl- and K+ had accompanied cellular shrinkage.
A combination of an elevated [Cl-]i/[Cl-]o and the high PCl/PK found in skeletal muscle could potentially splint Em despite an initial (ms s) hypertonic cell shrinkage. Normally, this would not persist over longer time intervals (min h) as a passive Cl- efflux would then restore [Cl-]i/[Cl-]o towards electrochemical equilibrium. Nevertheless, the present experiments suggested that Na+-dependent Cl- influxes can maintain such an elevated [Cl-]i/[Cl-]o. Thus, both extracellular Na+ withdrawal and pharmacological blockers of NKCC and NCC (Gamba et al. 1993; Lang et al. 1998a,b) abolished the steady-state splinting of Em. Measured Em values under such conditions then hyperpolarized in hypertonic solutions following a GHK prediction in which Cl- equilibrium potentials approximated those of [K+] and suggesting that Cl- had now passively redistributed over the time course of our experiments.
These findings prompted further investigations for a possible role of NKCC and NCC rather than for Na+-K+-ATPase activity in splinting Em. Na+-K+-ATPase activity should be capable of influencing Em, whether through its action on [K+]i or [Na+]i, or its electrogenic effect, only in fibres exposed to Na+-containing extracellular solutions. Yet increases in extracellular osmolarity hyperpolarized Em not only in fibres studied in Cl--free Ringer but also in Na+-free extracellular solutions whilst it spared Em in (Na+-containing) normal Ringer solution. Furthermore, increased external osmolarity hyperpolarized rather than depolarized Em in ouabain-treated fibres contrary to expectations from a straightforward block of Na+-K+-ATPase.
Of possible Na+ cotransporters, the Na+-Cl- cotransporter (NCC) permits large NaCl fluxes across renal tubular epithelial membranes (Sun et al. 1991) but no direct molecular evidence currently exists for its expression in skeletal muscle. Nevertheless, Dorup & Clausen (1996) demonstrated that increased extracellular osmolarity stimulated a bumetanide-sensitive 22Na+ influx in rat skeletal muscle abolished by removal of extracellular Cl- but not of extracellular K+. A second, bumetanide-sensitive, Na+-K+-2Cl-, cotransporter (NKCC) is implicated in net sodium reabsorption in a variety of epithelia and regulatory volume increases (RVIs) in many cell types (see, e.g. Tosteson & Hoffman, 1960). It may be activated by cell shrinkage produced by hypertonic external solutions and thereby promotes RVIs through osmotic effects of net increases in intracellular Na+, K+ and Cl- (Lang et al. 1998a,b). The NKCC has been demonstrated in cultured myocytes (Liedke, 1992) and recent molecular and functional evidence suggests that it may also be expressed in rat skeletal muscle (Wong et al. 1999). It has been implicated in amphibian skeletal muscle vacuolation following osmotic stress (Khan et al. 2000). Finally, hypertonicity does stimulate a bumetanide-sensitive 22Na+ influx in rat skeletal muscle (Clausen et al. 1979; Chinet, 1993).
The final experiments directly support roles for both NCC and NKCC systems in splinting Em during exposures to extracellular hypertonicity. Thus fibre pretreatment with a combination of the NCC blocker chlorothiazide and the NKCC blocker bumetanide abolished this Em stabilization. Ouabain partially (10 mV) depolarized fibres studied in isotonic Ringer solution, in agreement with reports by Marunaka (1988). However it also abolished Em stabilization in hypertonic solutions, permitting a net hyperpolarization of Em, directly fitting suggestions (Lindinger et al. 2002) that Na+ gradients established by Na+-K+-ATPase activity (Schwartz et al. 1975) influence NKCC and NCC activity. Partial reductions in [Na+]o produced similar effects. Finally, applications of bumetanide and chlorothiazide in combination after introduction of hypertonic Ringer solution exerted no immediate effect on Em but produced a delayed hyperpolarization after a time interval during which passive Cl- redistribution would be expected to have occurred. All these results thus suggest that persistent NKCC and NCC activity can maintain [Cl-]i/[Cl-]o above its electrochemical equilibrium and thereby stabilize Em following fibre shrinkage.
A simple model then demonstrated that NKCC (and/or NCC) activity together with the relatively high membrane PCl/PK in skeletal muscle, quantitatively accounted for the stabilization of steady-state Em in fibres subjected to osmotically imposed volume change. This assumed that the transporter influxes exceeded passive Cl- effluxes and therefore could elevate [Cl(]i/[Cl-]o above electrochemical equilibrium levels. It adopted established values of transporter stoichiometries, membrane permeabilities and baseline values of intracellular electrolyte concentrations for fibres in isotonic conditions (see above; Hodgkin & Horowicz, 1959; Hutter & Noble, 1960; Harris, 1965). Successive finite difference iterations of transporter activity then calculated the corresponding adjustments in V and Em and the coupled electrolyte concentration changes following an initial cell shrinkage produced by a hypertonic challenge.
This simulation predicted that the relationships between corrections in Em and in V varied strikingly with PCl/PK. At the relatively high PCl/PK(2) in skeletal muscle, Em could be fully corrected in the absence of significant volume corrections. Volume therefore alters passively with changes in extracellular osmolarity in the absence of significant RVI, giving a situation that successfully reproduces the present experimental observations in skeletal muscle studied in normal, Cl--containing Ringer solutions.
The effects of PCl/PK have potentially broader implications for the differing relationships between V and Em shown by cell types other than skeletal muscle. For cells with a low PCl/PK such as smooth muscle (review: Chipperfield & Harper, 2000) or erythrocytes (Tosteson & Hoffman, 1960) similar simulations demonstrated that transporter activity could produce large corrections in V but little correction of Em, which then more closely approximated the K+ equilibrium potential. This could permit large RVIs but without significant stabilization of Em.
These schemes involving interplay of NKCC/NCC transporter activity and PCl/PK do not permit immediate, simultaneous control of both V and Em. However, one may speculate that normal function in central nervous system neurones requires precise control of both V, to preserve a highly specialized neuronal morphology, and Em, which is critical to cellular excitability. Nevertheless, these cells exist within a unique environment protected from significant tonicity changes by the glial cell population (Rose & Ransom, 1996; review: Hertz et al. 2000) and so need not intrinsically depend upon mechanisms that necessarily compromise regulation of either V or Em (Rose & Ransom, 1997; cf. Yan et al. 2003).
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionMethodsResults |
|---|
) and a 0–200 mM (
