Lidocaine and Barium Distinguish Separate Routes of Transbasal K⁺ Uptake in the Posterior Midgut of the Tobacco Hornworm (Manduca Sexta)

The posterior midgut of lepidopteran larvae consists of two major cell types, columnar and goblet, which show restricted coupling with intracellular Lucifer Yellow dye but are apparently well coupled with respect to small ions and electrical current (Moffett et al. 1982; Moffett and Koch, 1988a). The isolated, short-circuited tissue transports K⁺ from hemolymph to lumen at a rate as high as 1 μmolcm^-2min^-l. Active K⁺ transport involves two transmembrane steps: entry across the basal membrane followed by extrusion across the apical membrane. Studies using X-ray microanalysis (Dow et al. 1984) and K⁺-selective microelectrodes (Moffett and Koch, 1988b) show that K⁺ is actively transported across the goblet cell apical membrane against a large electrochemical gradient. Under typical experimental conditions the transbasal electrochemical gradient is favorable for K⁺ entry (Moffett et al. 1982). However, the transbasal gradient becomes unfavorable when extracellular K⁺ concentration is low and under hypoxia (Chao et al. 1990). The continuation of some net K⁺ transport under these conditions is evidence that transbasal K⁺ movement involves an active component that is masked under typical conditions in vitro.

Evidence for multiple parallel pathways for K⁺ entry across the basal membrane was provided by the effects of Ba²⁺. Although generally regarded as a competitive inhibitor of K⁺ passage through membrane channels, Ba²⁺ partly inhibits net K⁺ transport by the midgut in standard saline, but stimulates it in high-K⁺ saline. This finding was interpreted as evidence that Ba²⁺ inhibits one route of transbasal K⁺ entry while stimulating another; about one-third of transbasal K⁺ influx appears to be Ba²⁺-insensitive (Moffett and Koch, 1985). In standard saline, in which the inhibitory effect dominates, application of Ba²⁺ results in hyperpolarization of the basal membrane and an increase in transbasal resistance (Moffett and Koch, 1988a), effects that are consistent with blockage of a current-carrying process.

The present studies characterize the effect of lidocaine on net K⁺ transport by the posterior midgut. Lidocaine exerts its local anesthetic effect by blocking the voltage-gated Na⁺ and Ca²⁺ channels of excitable cells, but it has also been reported to block K⁺ channels in the basolateral membrane of amphibian urinary bladder (Van Driessche, 1986) and turtle colon (Dawson et al. 1988). In the midgut, lidocaine, like Ba²⁺, inhibits the short-circuit current (I_sc) and increases basal membrane resistance. However, the effect of each inhibitor is almost unaffected by the presence of the other, and the two inhibitors show different forms of dependence on extracellular K⁺ concentration. These results suggest that lidocaine and barium distinguish separate routes of transbasal K⁺ entry.

A preliminary account of these studies has been presented (Koch and Moffett, 1990).

Larvae of Manduca sexta were reared on an artificial diet (Yamamoto, 1969). The morphologically distinct posterior midgut (Cioffi, 1979) was removed from cold-anesthetized fifth-instar larvae and immediately transferred to oxygenated ice-cold bathing solution. The Malpighian tubules and adhering fat body were removed and the tissue was mounted as described below.

The standard bathing solution, designated ‘32KS’, contained, in mmoll^-1: 32 KC1, 5 CaCl₂, 5 MgCl₂, 5 Tris-HCl (pH8.0), 166 sucrose. In some experiments, the concentration of KC1 was varied as indicated by the solution designation; for example ‘70KS’ contains 70 mmol I^-1 KC1. In other experiments the solutions were also made nominally divalent-cation-free by omission of CaCl₂ and MgCl₂. Except for the experiments on side specificity, all experiments were conducted with identical solutions on both sides of the tissue. Unless indicated otherwise, bathing solutions were oxygenated with 100 % O₂.

Pure lidocaine (Sigma) and its quaternary amine congeners QX-222 and QX-314 (kindly provided by Dr Bertil Takman of Astra Pharmaceutical) were added to bathing solutions to give the concentrations indicated. At the pH of the bathing solutions used in this study (8.0) the maximum concentration of lidocaine that can be attained is 5 mmoll^-1. Barium chloride was added to the half-chambers as a 1 moll^-1 solution in distilled water to give the final concentrations indicated.

For studies of inhibitor effects on I_sc, the tissue was mounted in a modified Ussing chamber as in previous reports (Moffett and Koch, 1985). Except for the experiments on side specificity, all experiments were conducted with identical solutions on both sides of the tissue. Tissues were continuously short-circuited and I_sc was recorded using a strip chart recorder. Solution resistance was compensated using predetermined values.

The results of inhibitor experiments are expressed in terms of the ratio of I_scafter inhibition to the control value of I_sc measured just before addition of inhibitor (I/I_o). Recovery of I_sc is expressed as a percentage of the current recorded immediately before inhibitor was added. Although I_sc decays spontaneously with time, with our present techniques the rate of decay is less than 10% h^-1 after initial equilibration. Since the effects of Ba²⁺ and lidocaine are rapid and exposures to the inhibitors were brief, no corrections for timedependent decay of I_sc were made.

Microelectrodes were pulled from microcapillary glass (Fredrich Haer 1011) in a vertical puller (David Kopf) and filled with 1 moll^-1 KC1. Useful electrodes had tip resistances of 25–30MΩ when measured in 32KS. The midgut was mounted in a chamber modified from the design of Thompson et al. (1982) and penetrated from the hemolymphal side with electrical reference to the hemolymphal solution. All experiments were conducted under short-circuit conditions with identical solutions on both sides. The microelectrode voltage was amplified by a Keithley electrometer and recorded on one channel of a Gould Brush four-channel pen recorder, with I_sc being simultaneously recorded on another channel.

The microelectrode was advanced at right angles to the tissue by a hydraulic drive (Friedrich Haer). Penetration of a cell was indicated by a sharp negative deflection of between –20 and –40 mV (Moffett et al. 1982); this is the transbasal potential (V_b). The change in transbasal resistance was measured as the change in the magnitude of deflection of V_b in response to the passage of a pulse of constant current. This is a relative measure; the absolute value of transbasal resistance cannot be measured readily by this method because of the large series resistance component presented by the bathing solution.

All data are given ±the standard errors of their means. For each tissue in the microelectrode experiments, several impalements were made and the averages of the values obtained were taken as best estimates for that tissue. The means presented are the averages of these best estimates.

Bilateral addition of lidocaine partly inhibits I_SC (Fig. 1). In six tissues I/I_o and the percentage reversibility of inhibition were measured as a function of the duration of exposure to lidocaine for durations of 5, 10, 20 and 40 min. In three of the tissues the order of exposure durations was ascending; in the other three the order was descending. Results from the two series were not distinguishable, and pooled results are shown in Fig. 2. The inset to Fig. 2 shows the correlation between I/I₀ and fractional recovery.

On the basis of time course and reversibility, two phases of lidocaine inhibition can be distinguished in these results. The initial phase consists of a rapid drop in I_scduring the first 10min. The inhibition that occurs during this phase is almost completely reversible (Fig. 2). If exposure to lidocaine continues beyond 10 min, a slower drop in I_sc is seen. This phase takes up to l h to complete and is accompanied by a progressive decrease in reversibility (Fig. 2). To maximize reversibility, data were collected at 7.5 min after addition of lidocaine in all subsequent experiments.

The limited solubility of lidocaine in bathing solutions at pH 8.0 appears to preclude the maximal effect of the inhibitor. At a concentration of 5 mmol l^-1, the inhibitory effect was almost exactly twice that at 2.5 mmol l^-1 (Fig. 3).

The rapid effect of lidocaine is on the hemolymphal side (Fig. 4). In four experiments, 75±3.1% of the inhibition in effect by 7.5 min after bilateral application was attributable to the hemolymphal lidocaine, regardless of which side was exposed first.

The effects of lidocaine on transbasal voltage (V_b) and transbasal resistance (R_b) were measured in microelectrode experiments. Transbasal penetrations were made in tissues superfused with 32KS oxygenated with 100% O₂. The mean V_bwas –36.5±0.84 mV in seven experiments. The superfusate was then changed to oxygenated 32KS with 5 mmol l^-1 lidocaine and changes in V_b, R_b and I_sc were recorded. Fig. 5A shows typical results. The V_b hyperpolarized synchronously with the decrease in I_SC; the mean change in V_b was –6.35±1.63 mV. The increase in the deflection of V_b in response to the transepithelial constant-current pulse indicates increased R_b; the mean change in basal resistance was 22±6%. The effects of lidocaine on V_b, R_b and I_sc are comparable to those of 6 mmol l^-1 Ba²⁺ (Fig. 5B). As in previous studies (Moffett and Koch, 1988a), Ba²⁺ hyperpolarized the basal membrane with a concomitant increase in R_b and decrease in I_sc.

Although in previous studies we were usually able to maintain impalements through a number of solution changes, in these experiments re-introduction of lidocaine-free solution usually caused the microelectrode to be dislodged. This observation could have been due to contraction of the intestinal muscularis following washout of lidocaine.

Two quaternary amine analogues of lidocaine, QX-222 and QX-314, were used to determine whether lidocaine must permeate the cell membrane to be effective. The pKa of lidocaine is 8.1, so at pH8.0 about half of the dissolved lidocaine is in the uncharged, lipid-soluble form. In contrast, the analogues are charged in solution and thus only weakly lipid-soluble. In two trials, neither analogue had a measurable effect on I_sc after 30 min of exposure to 10 mmol I^-1, the highest concentration tested.

These experiments determined the effect of lidocaine on the relationship between bathing solution K⁺ concentration ([K⁺]_o) and I_sc. Each tissue was presented with stepwise changes in concentration over the series 10, 20, 32, 50 and 70 mmol l^-1 K⁺. In three of the six experiments, the tissues were equilibrated in 10KS and presented with ascending [K⁺]_o; the other three were equilibrated in 70KS and presented with descending [K⁺]_o. The timing of the experiments was similar to that shown in Fig. 3 of Moffett and Koch (1985), except that in the present experiments percentage recovery was also measured. Briefly, tissues were equilibrated in the initial solution until I_SC became stable (usually about 45 min). Then the corresponding lidocaine-containing solution was substituted. After 7.5 min of lidocaine exposure, the original solution was restored for measurement of recovery, which typically required 10–15 min. The bathing solution was then replaced with the next solution in the series, and the sequence of lidocaine exposure and recovery repeated.

The mean values of I_sc for each [K⁺]_o differed between the two protocols (Table 1), and I/I₀ was consistently lower for the tissues equilibrated at high [K⁺]_o than for those equilibrated at low [K⁺]_o. Nevertheless, within each series the lidocaine-sensitive fraction of the total current was almost constant. The reversibility of the inhibition was independent of the K⁺ concentration of the equilibration solution and test solutions. The mean reversibility for the series was 0.72±0.03.

The effect of decreased O₂ tension on lidocaine inhibition of I_SC was measured in 10 experiments in which the tissues were equilibrated in 32KS gassed with 100 % O₂, then switched to 32KS gassed with air (21 % O₂) and then to 32KS gassed with 5 % O₂. A 7.5 min exposure to 5 mmol l^-1 lidocaine was interposed during each of the three O₂ tensions.

In contrast to the constancy of I/I₀ when I_sc was altered by changing [K⁺]_o, reduction of O₂ tension greatly enhanced the inhibitory effect of lidocaine (Table 2). However, decreased O₂ tension reduced the reversibility of the inhibition. In 21 % O₂ the recovery was only 0.60±0.03 as compared to 0.73 in 100 % O₂, and currents were so low in the 5 % O₂ experiments that percentage recovery was not computed.

The experiments reported above on the dose-response relationship, the effect of varying [K⁺]_o and the effect of low O₂ tension were also carried out in nominally divalent-cation-free solutions. As previously reported (Moffett and Koch, 1985), initial values of I_SC were higher in these solutions; the mean I_SC in 32K was 543±μAcm^-2 for four experiments compared to the combined mean of 376μAcm^-2 for the six experiments in the present report (Table 1). The relationship between I_SC and [K⁺]_o became quasi-saturating, as expected from previous studies (Moffett and Koch, 1985). The dose-response curve for lidocaine in divalent-cation-free solutions was similar to that in standard solution: on average, 2.5 mmol l^-1 lidocaine inhibited 19% of the current while 5 mmol l^-1 lidocaine inhibited 44%. As in standard solutions, I/I₀ for 5 mmol I^-1 lidocaine was constant when [K⁺]_o was varied: the overall mean for two experiments using ascending [K⁺]_o was 0.68±0.02 and for two with descending [K⁺]_o it was 0.45±5.1. Enhancement of lidocaine inhibition by decreased O₂ tension in divalent-cation-free solution was similar to that observed in standard solutions; in two experiments, I/I_o in 5 mmol l^-1 lidocaine was 0.67±0.03 in 100% O₂ but 0.06±0.02 in 5% O₂.

The extent of competition between Ba²⁺ and lidocaine for inhibition of I_sc was determined by comparing the I/I₀ of each inhibitor given singly to the 1/I_O of that inhibitor in the presence of the other. In nine of eighteen experiments 6 mmol I^-1 Ba²⁺ was given first, followed by 5 mmol I^-1 lidocaine; in the other nine experiments the order of inhibitors was reversed. Fig. 6 shows the average results in the form of two diagrams of changes in I/I_o, one for experiments in which Ba²⁺ was given first and one for those in which lidocaine was given first. When given first, Ba²⁺ inhibited 57% of the current (I/I was 0.43), but when given in the presence of lidocaine, Ba²⁺ inhibited 69 % of the remaining current. Similarly, when given first, lidocaine inhibited 44% of the current, but it inhibited 68% of the current remaining after Ba²⁺ blockade.

Previous studies suggested that the transbasal electrical potential of midgut consists of at least two components; a Nernstian component generated mainly by the transmembrane gradient of K⁺, and a contribution due to the current of net K⁺ entry through conductive pathways (Chao et al. 1990). Cation channels in the basal membrane detected by patch-clamp (Moffett and Lewis, 1990) and noise analysis (Zeiske et al. 1986) are one possible route of conductive K⁺ entry. These would be expected to mediate passive K⁺ entry in tissues bathed in well-oxygenated solution containing normal to high K⁺ concentrations, conditions under which the transbasal K⁺ electrochemical gradient favors K⁺ entry (Moffett and Koch, 1988a; Chao et al. 1990). An active component of K⁺ uptake was inferred from observations that a short-circuit current may be maintained under low extracellular [K⁺] and/or hypoxia, while under these conditions the electrochemical gradient is unfavorable for K⁺ entry (Chao et al. 1990). If it were electrogenic, an active pathway would also contribute to the entry-related component of the transbasal potential.

As previously reported (Moffett and Koch, 1988a), hemolymphal Ba²⁺ hyperpolarizes the basal membrane and increases transbasal resistance. In view of the reputation of Ba²⁺ as a competitive inhibitor of K⁺ channels, this effect was interpreted as the consequence of blockage of one route of K⁺ entry, with diversion of some K⁺ entry current to a second conductive pathway (Moffett and Koch, 1985). According to this interpretation, the increase in resistance to the K⁺ entry current caused by blockage of the Ba²⁺-sensitive pathway increases the magnitude of the entry-related component of V_b, hyperpolarizing the basal membrane. The decrease in I_sc that occurs at low and normal K⁺ concentrations can be regarded as the result of the increased transapical potential against which the apical pump must work (Moffett and Koch, 1988a). In contrast, direct inhibition of the apical pump by hypoxia decreases the entry-related voltage. component, depolarizing the basal membrane (Moffett and Koch, 1988a; Chao et al. 1990).

The present studies show that the effects of lidocaine on V_b and R_b are similar to those of Ba²⁺, arguing that both inhibitors affect the entry-related component of V_b. The fractional inhibition of either inhibitor, defined in terms of the current remaining after the first inhibitor has taken effect, is increased in the presence of the other (Fig. 6). This is evidence against competition between the two inhibitors, for such competition would have decreased the fractional inhibition of lidocaine after Ba²⁺, rather than increasing it. However, the absolute decrease in current due to either of the inhibitors is less when it is added second than when it is added first (Fig. 6). The decrease in the amount of current abolished by either inhibitor when it is second in order of addition is consistent with an overlap between the two inhibitors of about 18 % of the total transbasal current. That is to say, if the basal membrane contained a number of channels all of equal conductance, about 18% of the channels would be susceptible to blockage by either inhibitor. The 17 % of I_SC that remains in the presence of both inhibitors may be carried by an uptake pathway insensitive to either inhibitor, but could well be due to the fact that, because of its low solubility in 32KS solution, the concentration of lidocaine was suboptimal.

The concentrations of lidocaine used in these studies (2.5–5 mmol l^-1) are an order of magnitude higher than the range of about 200–800 μmol l^-1 in which effects on Na⁺ and K⁺ channels are detectable and in which the drug is effective as a local anesthetic. The possibility that such a high dose might still be suboptimal could be explained by the likelihood that in this system, as in others, the binding of lidocaine to its active site is reversible. To achieve maximal inhibition of I_sc, the inhibitor must be concentrated enough to block the susceptible channels essentially all the time. In contrast, blockage of neuronal conduction requires only that some critical fraction of the voltage-gated channels be blocked at any instant.

The relationship between I_sc and [K⁺]_o has been analyzed as the sum of a saturating and a linear component (Moffett and Koch, 1983, 1985). These components are believed to reflect different concentration dependences of parallel transbasal uptake processes. According to this analysis, under standard conditions about two-thirds of the transported K⁺ enters the cells through the saturating component. The relative contribution of the saturating component to total K⁺ uptake is greatest at low [K⁺]_o and decreases as [K⁺]_o increases.

The control values of I_sc measured in the experiments in which a series of [K⁺]_o was presented (Table 1) are not comparable to results obtained in our previous studies of the K⁺ concentration dependence of I_sc (Moffett and Koch, 1985), which used a similar protocol. The difference is assumed to reflect the accumulation of irreversible lidocaine block over the course of the repeated exposures to lidocaine in the present experiments, making control values of I_sc measured later in each experiment progressively lower than would be expected from the previous studies. Nevertheless, the fractional inhibition of the remaining I_SC by lidocaine was remarkably constant over the range of extracellular K⁺ concentrations, regardless of whether the concentrations were presented in ascending or descending order (Table 1). This result indicates that, unlike the situation for the divalent-cation-sensitive uptake process, the relative contribution of the lidocaine-sensitive uptake process to total uptake is independent of [K⁺]_o.

The values of I/I_o for the ascending series of [K⁺]_o are significantly greater than those of the descending series (Table 1). Although no definite conclusion can be drawn, the difference could reflect an effect of the equilibration [K⁺]_o on intracellular [K⁺]. Previous studies gave values of about 70 mmol I^-1 for tissues superfused in 10KS, but greater than 100 mmol l^-1 for tissues superfused in 70KS (Moffett et al. 1982). Although any difference in the cytoplasmic K⁺ concentration ([K⁺]_i of the tissues at the start of the two experiments undoubtedly diminished as the experiments progressed, the greater degree of inhibition occurred in the descending concentration series in which tissues could be expected to have higher [K⁺]i values. The result implies that the fraction of total uptake that is lidocaine-inhibitable is more sensitive to [K⁺]_i than to [K⁺]_o. Although higher [K⁺]_i is associated with more negative values of V_b, voltage-sensitivity of lidocaine block can be ruled out because I / I₀ is constant over a range of [K⁺]_o previously shown to result in large changes in V_b (Moffett and Koch, 1988a).

Deletion of Ca²⁺ and Mg²⁺ from bathing solution favors the saturating component of K⁺ uptake, while addition of Ba²⁺ favors the linear component at the expense of the saturating component, particularly if the bathing solutions are also Ca²⁺-and Mg²⁺-free (Moffett and Koch, 1983, 1985). In Ca²⁺,Mg²⁺-free solution, Ba²⁺ stimulates I_sc when [K⁺]_o is 50 mmol l^-1 or higher. Lidocaine inhibition, unlike that of Ba²⁺, was not affected by removal of Ca²⁺ and Mg²⁺, and lidocaine inhibited I_SC at high as well as normal and low [K⁺]_o. The insensitivity of the lidocaine effect to [K⁺]_o and to deletion of alkaline earth ions is additional evidence that lidocaine affects a component of transbasal K⁺ uptake insensitive both to Ba²⁺ and to the presence or absence of Ca²⁺ and Mg²⁺, unlike the pathway identified by Moffett and Koch (1985).

Both Ba²⁺ and lidocaine are regarded as blockers of membrane channels. Such blockers of passive movement would be expected to inhibit I_SC only under conditions in which the transbasal K⁺ electrochemical gradient favors K⁺ entry, and to have no (or perhaps a stimulatory) effect when the transbasal gradient is unfavorable for K⁺ entry. In 32KS gassed with 5% O₂, the transbasal K⁺ electrochemical gradient rapidly becomes unfavorable for K⁺ entry. Under these conditions the inhibitory effect of Ba²⁺ actually increased. This result was taken to suggest that the entry process inhibited by Ba²⁺ is thermodynamically active (Chao et al. 1990). The similar increase in effectiveness of lidocaine in hypoxia found in the present studies suggests that lidocaine, like Ba²⁺, inhibits a thermodynamically active K⁺ uptake process.

Quaternary analogues of lidocaine are effective when applied to the cytoplasmic side of axons but not when applied to the outside (Frazier et al. 1970). The failure of extracellular quaternary analogues to inhibit the I_SC of midgut suggests that lidocaine must dissolve in the membrane lipid to reach its site of action. Lidocaine competes favorably with K⁺ for detection by resin-based K⁺-selective microelectrodes, and was detected in the cytoplasm with such microelectrodes after application to the hemolymphal side (K. F. Moffett and A. Koch, unpublished observations). However, the block of I_sc probably does not come primarily from free intracellular lidocaine, since the drug is markedly more effective when applied to the hemolymphal side of the tissue, although lidocaine would be expected to enter the cytoplasm with equal ease from either side of the tissue. Thus, lidocaine may reach its blocking site from within the lipid component of the basal cell membrane, as has been suggested for lidocaine block of Na⁺ channels in frog muscle (Schwarz et al. 1977).

The effects of lidocaine on transepithelial K⁺ transport described here add to the evidence that K⁺ enters the intracellular transport pool of posterior midgut by multiple pathways, rather than by a single readily characterizable mechanism. The substantial effect on I_SC of partial inhibition of K⁺ uptake by Ba²⁺ and lidocaine suggests that regulation of the uptake pathways could be a potent mechanism for physiological control of transport of K⁺.

These studies were supported by NSF DCB 8315739. We thank Robin Woods and Anna-Lise Dragoy for technical assistance.