Quantitative Analysis of Resting Membrane Electrogenesis in Insect (Diptera) Skeletal Muscle : II. Testing of A Model Involving Contributions from Potassium and Sodium Ions, and the Anomalous Effect of Reducing Extracellular Sodium

In the preceding paper (Dawson & Djamgoz, 1988), we showed that the K⁺ equilibrium potential (E_K) of prepupal Calliphora erythrocephala skeletal muscle was significantly more negative than the prevailing membrane potential (E_M). We predicted, therefore, that at rest these muscles must be permeable to at least one ion other than K⁺ with an equilibrium potential more positive than E_M. Furthermore, Na⁺ was shown to meet this requirement, and Cl⁻ was found not to be involved in resting membrane electrogenesis (Dawson & Djamgoz, 1988). In analogous muscles of Drosophila melanogaster, Jan & Jan (1976) suggested that the resting membrane has a significant permeability to Na⁺, the ratio (α) of the partial permeabilities to Na⁺ and K⁺ (p_Na and p_K, respectively) being 0-23. To test this proposal directly, the effect of reducing extracellular Na⁺ concentration on E_M was tested using Tris⁺ as the substitute. As extracellular Na⁺ was decreased, the membrane potential initially hyperpolarized, as predicted. However, in salines containing up to one-half as much Na⁺ as that found in normal saline, E_M values surprisingly repolarized to levels more positive than normal. Comparable results were obtained by substituting sucrose or MgCl₂ for NaCl (Jan & Jan, 1976). The latter experiments were carried out by ‘soaking’ the muscles in modified salines, and impaling several cells in each preparation. In earlier experiments also, the dependence of membrane potentials on extracellular ion concentrations was tested in a similar way, sometimes by soaking or ‘conditioning’ muscles in modified salines for several hours (Wood, 1963, 1965; Huddart, 1966; Piek, 1975). It is not clear from these studies whether sufficient controls were carried out to ensure that the changes in membrane potentials seen were reversible or whether timedependentloss of membrane potentials occurred. Importantly, Wareham, Duncan & Bowler (1973) demonstrated in cockroach coxal muscles that membrane potentials gradually decay if bicarbonate-free Ringer’s solutions are used (possibly due to deterioration of intracellular pH; see Thomas, 1984, for a review of mechanisms of intracellular pH regulation). Most early Ringer’s solutions (including that used by Jan & Jan, 1976) lacked bicarbonate. In the present study, we have used a modified Ringer and attempted to monitor additional membrane parameters during solution changes, as well as ensuring that the effects observed were reversible. Thus, the physiological integrity of the preparations and possible secondary effects of solution changes have been assessed. Preliminary reports of these findings have been given earlier (Dawson & Djamgoz, 1984a,b).

Experiments were carried out on prepupal Calliphora erythrocephala Meig (Diptera) at room temperature (18–20°C). Details of the preparations and ionsensitive microelectrodes were as given in the preceding paper (Dawson & Djamgoz, 1988). Other details are as follows.

Microelectrodes

Conventional microelectrodes were drawn from borosilicate glass capillaries (1 mm o.d., 0·6mm i.d.) on a vertical puller. They were filled with 2·5 mol l⁻¹ KC1 and had tip resistances (d.c.) in the range 10–20 MΩ. The tip potentials of these electrodes were <5 mV, as determined by breaking the tips and noting the resulting difference in the potential reading in the bath (Adrian, 1956). The microelectrode was connected to a microprobe system (WPIM701) which had an inbuilt bridge circuit for current injection for measurement of membrane input resistance. Ion-sensitive microelectrodes were constructed and used as described previously (Djamgoz & Dawson, 1986; Dawson & Djamgoz, 1988).

Modified solutions were based on Rice’s saline and were of constant ionic strength and osmolarity. In the K⁺ substitution solutions, Na⁺ was used as a substitute for K⁺. Na⁺ substitution experiments mostly involved a 100-fold reduction of Na⁺ activity using a variety of substitutes for NaCl: choline chloride, glucose, tétraméthylammonium chloride, magnesium chloride or N-methyl D-glucamine. The osmolarity of each solution was measured using a mini-osmometer (Roebling) and adjusted to normal (402mosmoll⁻¹) using additional glucose, if necessary.

A micropipette was pulled from 1·5 mm o.d. glass capillary and its tip broken to 100–200 μm diameter. It was attached to a microelectrode with sealing wax such that the tip of the broken pipette was around 0·5 mm behind the tip of the microelectrode, whilst at right angles to their axes the two were separated by about 1mm. The broken pipette was filled with Ringer containing 97·2 mmol l⁻¹ K⁺ (substituted for Na⁺). Constant-sized (around 4 μ1) droplets could be ejected from the pipette by applying air pulses of Is duration from an air cylinder set to 10–30 Pa via a solenoid-activated valve driven by a physiological stimulator. The assembly of pipettes was positioned in the recording chamber such that the broken pipette was ‘upstream’ of the recording microelectrode to ensure that the ejected pulses of the high-K⁺ Ringer were washed onto the impaled cell and away from it.

A typical recording of the changes in the resting potential (E_M) resulting from varying the extracellular K⁺ activity (aK_o) between 0·97 and 97·2, using Na⁺ as the substitute, is shown in Fig. 1. During the solution changes, the cell was twice returned to normal Rice’s saline to ensure that the E_M changes recorded were reversible, and that the cell remained healthy throughout the experiment, which could last up to 4 h. Sustained changes in E_M were found to result from alterations in aK_o and these were fastest when aK_o was reduced (see Fig. 1). For example, when aK_o was reduced from 9·7 to 3·6 mmol l⁻¹, the E_M reached a new, steady level within 15 min; when aK_o was increased from 36 to 97·2 mmol l⁻¹, the E_M did not reach a steady value for more than 30 min. Such effects were not related to the age of the preparation, since ‘slow’ responses were observed even at the start of an experiment. Data similar to Fig. 1 were obtained from 11 cells (each from a different preparation). The E_M depended on the logarithm of aK_o more or less linearly, with the maximum slope of the line giving only 31 mV per 10-fold change in aK_o, however (not illustrated). This is considerably less than the 58 mV per 10fold change in aK_o expected for a membrane that is permeable only to K⁺, and agrees with our conclusion in the preceding paper that K⁺ is not likely to be the only ion involved in resting membrane electrogenesis in these muscles.

To eliminate the possibility that there might be some time-dependent deterioration of the preparations, a number of shorter substitution experiments involving only three solution changes were also performed. A cell was impaled in normal Ringer (aK_o = 9·7 mmol l⁻¹) and a steady E_M recorded. Low-K⁺ Ringer (aK_o = 0·97 mmol l⁻¹) was then introduced and the membrane potential allowed to equilibrate. This was followed by high-K⁺ Ringer (aK_o = 97·2 mmol l⁻¹), and finally the cell was returned to normal Ringer to ensure that the original E_M was obtained. Using this protocol, a maximum change of 30mV was recorded for aKo values between 9·7 and 97 (cf. 31 mV for the same step in full substitution experiments), and a maximum change of 24mV was recorded for aK_o values between 0·97 and 9·7 mmol l⁻¹ (cf. 23 mV from the full substitution experiments). Thus, the slopes obtained for a given aKo range using the two types of protocol were not significantly different, providing further evidence that the membrane properties of the cells did not change during the full substitution experiments.

In the rest of the paper, experiments designed to test this model of resting membrane electrogenesis and critically determine the best value of α are described and discussed.

These experiments were carried out to test directly whether Na⁺ had any effect on E_M-Fig. 3 shows a typical recording of the effect on E_M of reducing extracellular Na⁺ activity from 124·6 to 1·25 mmol l⁻¹ (using Mg²⁺ as the substitute). Immediately after the introduction of the low-Na⁺ Ringer into the recording chamber, E_M showed an initial hyperpolarization of some 8 mV from its resting level of -50 mV. This was followed after less than lmin by a sustained depolarization of some 18 mV to a steady E_M level of −40 mV which was reached 10 min after the introduction of the low-Na⁺ Ringer. This effect reversed completely within 5 min of the return to normal Ringer.

To test that this result was not in some way due to the unsuitability of Mg²⁺ as a Na⁺ substitute, this type of experiment was repeated using a variety of other substitutes: choline, N-methyl D-glucamine, tetramethylammonium and glucose. Typical traces are shown in Fig. 4. In all cases, when low-Na⁺ was substituted for normal Ringer, an initial hyperpolarization of 2–7mV was seen, and this was followed by a sustained phase leading to E_M levels 2–8 mV more positive than the original resting potentials.

Two possible causes of the unexpected depolarization of the E_M in low-Na⁺ Ringers were considered: (a) a decrease in aK_i and a corresponding positive shift in E_K, which would account for the change in E_M, since K⁺ has been shown to be involved in membrane electrogenesis (Fig. 1); and (b) a decrease in the partial permeability of the membrane to K⁺, which would again lead to a depolarization of E_M, without a change in E_K. The following experiments were carried out to distinguish between these two possibilities.

A typical chart recording of an experiment in which the aK_i and E_M of a cell were measured using a potassium-sensitive microelectrode, and the effect of a low-Na⁺ Ringer on these parameters, is shown in Fig. 5. The upper trace gives the membrane potential of the cell (initially −38 mV) and the lower one gives aK_i (= 80 mmol l⁻¹). When low-Na⁺ Ringer was applied to the preparation, the E_M changed, as noted earlier. However, there was no significant change in aK₁. We concluded, therefore, that the sustained depolarization of E_M seen in low-Na⁺ Ringer’s solutions could not be due to a change in the E_K.

Any change in the relative permeability of the cell membrane to K⁺ and Na⁺ might be reflected as a net change in the input resistance of the membrane. In this group of experiments, the effect of lowering extracellular Na⁺ activity on the input resistances of the cell membranes was investigated using intracellular current injection by bridge circuitry. A typical recording from one such experiment is shown in Fig. 6. The upper trace shows the membrane potential (resting level, −42 mV) and its response to the application of the current pulses. The lower trace is the current monitor. Clearly, the delayed depolarizing phase of the E_M change is accompanied by a sustained increase in the input impedance of the cell membrane.

The value of the input resistance in normal Ringer was 96±9k Ω, and this increased by an average of 26 % in low-Na⁺ Ringer to give 121 ± 10 k Ω (N = 7).

To examine whether the membrane input resistance increase observed when the cell was exposed to low-Na⁺ Ringer was due to a reduction in the K⁺ permeability of the membrane, K⁺ pulsation experiments were carried out. This type of experiment is illustrated in Fig. 7A. The uppermost trace gives membrane potential, including the transient depolarizing responses resulting from high-K⁺ pulses, the applications of which are denoted by short bars beneath the E_M trace. In normal Ringer (125 mmol l⁻¹ Na⁺) the pulses depolarized the E_M by some 15 mV. In low-Na⁺ Ringer, however, the membrane responses were reduced to around 3 mV. The effect was completely reversible. The reduction in membrane responses started during the initial hyperpolarization of the membrane and persisted into the delayed depolarization phase, indicating a delay between the apparent reduction of partial K⁺ permeability and membrane response, similar to the situation in K⁺ substitution experiments (see Fig. 1). Average data are shown in Fig. 7B. Clearly, in the low-Na⁺ Ringer, the responsiveness of the cell membrane to high-K⁺ pulses is significantly decreased, a 100-fold reduction in uNa_o leading to an average reduction in membrane response to 4 % of its original value. Thus, the partial permeability of the cell membrane to K⁺ decreases substantially in a low-Na⁺ medium, giving rise to the anomalous effect of reduced extracellular Na⁺ on membrane potential. The value of α is not fixed, therefore, and this might explain the slight deviation of the data points from the straight line in Fig. 2.

Table 2. Resulting data are illustrated in Fig. 8, showing the relationships between predicted and measured values of E_M, and also expressed in Table 2. Clearly, the best prediction for the data is obtained for the value of α (0·036) derived from K⁺ substitution experiments involving solution changes in which Na⁺ activity was kept equal to or greater than the normal value.

The purpose of this set of experiments was to test whether the most realistic value of α (0·036) deduced in the first part of this paper would explain membrane electrogenesis in a restricted range of ionic concentrations around normal. The product was kept constant so as to prevent passive movements of intracellular Cl⁻. Substitution solutions were constructed as described in Materials and methods. A typical result is shown in Fig. 9. As expected, all E_M changes were quick and reversible, occurring within less than Imin of the completion of the solution change (cf. Fig. 1). The changes in E_M with changes in (aK_o+0·036aNa_o) are represented graphically in Fig. 10. The latter illustrates the linear relationship between E_M and the logarithm of (aK_o+0·036aNa_o). The slope of the line is 58 mV per 10-fold change in aK_o+0·036aNa_o with an extremely high correlation coefficient (0·996). Thus, this confirms that the value of 0·036 for α accurately predicts the prevailing resting membrane potential in a cell.

Measurements of the intracellular K⁺ activities (aK₁) and calculations of the K⁺ equilibrium potentials (E_K) presented in the preceding paper indicated strongly that in addition to K⁺, the resting membrane of prepupal Calliphora skeletal muscle was permeable to an ion with an equilibrium potential more positive than the E_M. This has been confirmed in the present study. Thus, the generation of the E_M has been shown to be explicable in terms of K⁺ and Na⁺ permeabilities. From experiments involving combined K⁺ and Na⁺ substitutions, the value of a was initially deduced to be 0·086. Jan & Jan (1976) reached a qualitatively similar conclusion in analogous muscles of Drosophila, although they estimated α to be 0·23. In the skeletal muscles of Locusta, Hoyle (1953) originally showed that a 10fold change in the extracellular K⁺ concentration resulted in an E_M change of only 50 mV, thereby indicating that K⁺ was not the only ion contributing to E_M. A comparable result was obtained by Janiszewski, Olszewska & Borkowska (1976) in Acheta. The situation in Orthoptera has been clarified by Leech (1986), who has demonstrated that, as in Diptera, resting membrane electrogenesis in the skeletal muscles of Schistocerca is due to K⁺ and Na⁺ permeabilities, α having a value of 0·016.

Attempts to test the Calliphora model by performing Na⁺ substitution experiments to determine directly the effect of Na⁺ on E_M were only partially successful, however. When the extracellular Na⁺ concentration was reduced, the EM hyperpolarized, as expected if Na⁺ were involved in generating the E_M-This initial hyperpolarization was followed by an unexpected sustained depolarizing phase, during which the E_M became more positive than the resting level. Broadly similar effects have been reported in a variety of other insects. Jan & Jan (1976) obtained a comparable result in Drosophila. The role of Na⁺ in resting membrane electrogenesis in insect skeletal muscles was studied in Periplaneta, Carausius and Locusta by Wood (1957, 1961, 1963) and in Schistocerca by Lea & Usherwood (1973), and consistent results were obtained in all four species: E_M values were found to be relatively depolarized in salines containing steadily decreasing Na⁺ levels. A similar effect was also shown for leech neuropile glial cells (Walz & Schlue, 1982). Wood (1963) found that as the extracellular Na⁺ concentration was reduced, the intracellular Na⁺ concentration (measured by flame photometry) decreased in parallel, thereby indicating again a finite permeability of the membranes to Na⁺. When aNa_i drops, one might expect that the activity of an electrogenic Na⁺/K⁺ pump would be slowed (Thomas, 1972) and that this might indirectly affect the E_M. This possibility can be ruled out, however, since such a secondary effect would also affect the intracellular K⁺ activity, which is clearly not the case (Fig. 4). Djamgoz (1986, 1987) has presented evidence to suggest that in insects with ‘conventional’ haemolymph ion levels the Na⁺/K⁺ pump does not directly generate a part of the E_M.

The response of Calliphora skeletal muscles to lowered extracellular Na⁺ involves a significant increase in the membrane input impedance (Fig. 5) and a fall in the partial permeability to K⁺ (p_K) (Figs 6, 7). The mechanism of the reduction in PK is not clear. It has been shown that skeletal muscles of Diptera possess three types of K⁺ conductance mechanism (Salkoff & Wyman, 1983). (i) The classical delayed, rectifying K⁺ conductance of the kind described in detail in the squid giant axon by Hodgkin & Huxley (1952). Lowering the extracellular Na⁺ concentration has been shown to have no effect on this conductance, however (Hodgkin & Huxley, 1952). (ii) A fast, transient K⁺ conductance, the time course of which could not explain the sustained change in E_M. (iii) A Ca²⁺-activated K⁺ conductance (gKCa) of the type found in snail neurones (Meech & Standen, 1975) and a variety of other cell types (see Meech, 1978 for a review). Reducing extracellular Na⁺ concentration might be expected to have an indirect effect on gKCa- For example, in sheep heart Purkinje fibres, a reduction in aNa_o causes an increase in the intracellular Ca²⁺ activity (aCa_i) by inhibiting the Na⁺/Ca²⁺ exchange mechanism (Bers & Ellis, 1982), and similarly in snail neurones (Partridge & Thomas, 1976). Such a rise in aCa_i would activate g_KCa and lead to an increase in p_K and a hyperpolarization in E_M, which is in fact observed. This scheme does not, however, explain the sustained depolarizing change in E_M. The depolarization might, however, be explained if reducing aNao led to a reduction in aCa_i. Such an effect has been reported in leech sensory neurones (Deitmer & Schlue, 1984), but it is not known if it also exists in Calliphora muscles. Interestingly, Jan & Jan (1976) found that the anomalous effect of reducing aNa_o on E_M could be rectified by increasing the external Ca²⁺ concentration. Thus, it appears that there is a link between the transmembrane Na⁺ gradient and/or aCa_i (not involving conventional Na⁺/Ca²⁺ exchange) and the K⁺ permeability of the membranes, but further work is required to clarify the situation.

In conclusion, the available data suggest that the value of a obtained from the high-Na⁺ end of the K⁺ substitution experiment (Fig. 2; Table 1), i.e. α = 0·036, is a realistic estimate. At rest therefore, nearly 97 % of the membrane permeability is to K⁺ and some 3 % to Na⁺. These figures have been confirmed by experiments on resting muscles involving no change of bathing media. Further evidence for the correctness of this value of α was obtained from experiments in which the effect of varying aK_o and aNa_o in a narrow range on E_M whilst keeping (aK_o+ 0·036aNa_o) × αCl_o constant was tested. The latter experiments also showed that the E_M changes induced by these solutions were rapid, equilibrating within a few minutes, as opposed to several tens of minutes experienced in the earlier K⁺ substitution experiments (Fig. 1), thereby also providing additional evidence for the passive distribution of Cl⁻ in these muscles.

This study was supported by a SERC-CASE studentship to JD (ICI pic as collaborating body). We thank Drs J. Hardie and S. Irving and Professor W. R. Schlue for many useful discussions.