ABSTRACT
The kinetics of 36C1 fluxes across cAMP-stimulated, short-circuited locust rectum were studied. Raising external K+ from 0 to 100 mM increased both Kt and Vmax for net Cl transport by four- to six-fold. Hill plots of indicated non-cooperative Cl interactions. The sequence for cation stimulation of was K > Rb > Cs > Na > NH4. Low levels of K were stimulatory only when added to the mucosal side. Cyclic AMP (cAMP) caused a small active absorption of K, although this was minor compared to the fourfold increase in transepithelial K diffusion (PK ). Neither cAMP stimulation of nor of PK was sensitive to Cl removal, suggesting that K-stimulated Cl absorption and K transport are not mediated by the same co-transport mechanism. Potassium is the counter-ion for electrogenic Cl transport because was less than 10% of the during cAMP exposure under Isc conditions, but equalled at open-circuit.
INTRODUCTION
The rectum of the desert locust Schistocerca gregaria actively absorbs water and solutes from the lumen of the hindgut (Phillips, 1964a, b). Most of the fluid reaching the rectal lumen originates at the Malpighian tubules, which secrete an isosmotic ‘primary urine’ which typically contains much higher levels of K (140 mM) and considerably less Na (20–47 mM) than are present in the haemolymph (Phillips, 1964a, b, c, 1981; Hanrahan, 1982).
We found evidence that Cl is actively transported across this epithelium by an unusual, K-stimulated mechanism (Hanrahan & Phillips, 1982). The purpose of the present work is to examine the relationship between K and Cl transport in more detail ; specifically, (i) we use steady-state tracer fluxes to establish whether K alters the maximal rate of transport , the apparent affinity of the transport mechanism for Cl , or both; (ii) we quantify the K requirements of Cl transport and determine the sidedness and cation selectivity of the stimulation and (iii) we test whether the active components of net Cl and and K absorptions are interdependent by measuring the effects of cAMP and Cl omission on transepithelial 42K fluxes. The results indicate that only low concentrations of K are required, that stimulation involves increases in and , and that K acts only from the mucosal side. There is a small active component of K absorption, but most K transport is passive and electrically coupled to Cl absorption. Some of these conclusions have been reported in preliminary form as symposia proceedings (Hanrahan & Phillips, 1980a,b, 1982, 1983).
MATERIALS AND METHODS
Adult female locusts, Schistocerca gregaria, were obtained from a colony maintained at U.B.C. The dissection and chambers were those of Williams, Phillips, Prince & Meredith (1978). The rectum was cut longitudinally to produce a sheet and attached across a collar-shaped opening with fine tungsten pins and a rubber ‘O’-ring. Five ml of solution were placed in both half-chambers and stirred with 95 % O2/5 % CO2 when HCO3 was present, or with 100% O2 without HCO3. Normal saline contained (mequiv l−1) : 110 Na, 10 K, 20 Mg, 10 Ca, 110 Cl, 10 HCO3, 10 glucose and 100 sucrose. Nominally Cl-free saline was prepared by replacing all Cl with methylsulphate. K-free saline was prepared by isosmolal replacement of K with sucrose. All salines contained the following amino acids as substrates (mmol l−1) : 2·9 alanine, 1·0 arginine, 1·3 asparagine, 5·0 glutamine, 11·4 glycine, 1·4 histidine, 1·4 lysine, 13·1 proline, 1·5 serine, 1·9 tryosine, 1·8 valine. Saline pH was 7·2 in HCO3-containing salines and 7·0–7·4 under HCO3-free conditions. Experiments were performed at 22°C.
The voltage clamp used in these studies has been described in detail elsewhere (Hanrahan, 1982; Hanrahan, Meredith, Phillips & Brandys, 1983). Short-circuit current (Isc) and transepithelial potential (Vt) were recorded on a strip chart recorder. Corrections were made automatically for asymmetries between the voltage-sensing KC1 agar bridges and for resistance of the bathing saline by the method of Rothe, Quay & Armstrong (1969). These corrections were especially important during large increases in [NaCl] and [K methylsulphate].
Kinetics of transepithelial 36Cl fluxes
Tissues were equilibrated in normal saline under Isc conditions for 3 h, and then the external medium was replaced bilaterally with Cl-free saline (methylsulphate substitution). Isc was near zero under these conditions. Approximately 0·5 h later, the chambers were rinsed three times with fresh Cl-free saline and cAMP was added to the serosal side to a final concentration of 1 mm. Small aliquots of 2 M-NaCl were then added to both sides in order to raise [Cl] stepwise from 0 to 2, 4, 10, 40 and 114 mm. 36C1 fluxes were measured during each step as described previously (Williams et al. 1978; Hanrahan, Phillips & Steeves, 1983). 36C1 (New England Nuclear, carrier-free, 5·9 mCig−1 Cl) was added as H36C1 to the ‘hot’ side in amounts that were too small to alter [Cl] significantly. One ml samples were taken from the ‘cold’ side at 15-min intervals and replaced with fresh saline. Samples were placed in 10 ml of scintillation fluid (ACS Amersham Corp., Oakville, Ontario) and were counted at constant quench using a scintillation counter (Isocap, Nuclear Chicago). Appropriate corrections were made for dilution during sampling. No correction was necessary for tracer backflux since 36C1 activity on the cold side reached only 0·5 % of that on the hot side.
Cyclic AMP was present when measuring transepithelial flux kinetics in order to stimulate active Cl transport: cAMP was equally effective in stimulating Cl transport when added immediately after dissection or after equilibration in vitro for several hours in normal saline. After switching to nominally Cl-free solution (cAMP still present), attained a new steady-state value 15 min after each step increase in [Cl]. Data from the second flux period were used in all calculations. The dependence of on [Cl] was determined in salines containing 0, 10 and 100mm-K in separate experiments: [K] was adjusted with K methylsulphate.
Measurement of K-dependence of Cl transport
The increase in Isc was used as a measure of while adding K-methylsulphate bilaterally. The equivalence of Isc and during cAMP-stimulation has been established over the range 0 ·140mm-K (see Table 2 and Fig. 1). In other words, elevating [K] does not cause electrogenic transport of ions other than Cl.
After mounting and short-circuiting the tissues, the chambers were rinsed repeatedly in K-free saline for 3 – 4 h. Cyclic AMP was added to the serosal side, causing Isc to increase to a new steady-state level within 1 h. Aliquots of K-methylsulphate were then added to both sides at 0’5 h intervals to yield sequential concentrations of 0, 2, 4, 10, 40, 100, 140 and 200mM-K. An identical protocol was followed during control experiments, except that Na-methylsulphate was added. Transepithelial potential and resistance were measured at 15 min intervals.
Transepithelial 42Kfluxes
42K ( New England Nuclear Corp., 0 ·13 – 0·15 Ci g-1 ) was added as 42KC1 to normal saline or as 42K2CO3 to Cl-free saline. Samples were taken at 15 min intervals and counted using an automatic gamma counter (1085, Nuclear Chicago). Initial radioactivity of the labelled side served as a reference in order to correct for tracer decay during experiments (total experimental and counting time = 11 h).
Calculations and statistics
In order to compare the instantaneous with tracer fluxes measured at intervals, ISC recordings were integrated using a planimeter (model L30M, Lasico, Los Angeles, California). Values are means ±standard errors unless stated otherwise ; N = number of recta. Where appropriate, significant difference were determined using paired or unpaired t-tests.
When kinetics of were studied on individual rectal preparations, was estimated by subtracting the mean backflux for other preparations at the same [Cl]. This was justified because was very much smaller and very much less variable than (i.e. contributed on average only 5%, maximum 17%, of the total variance in net flux). Calculating the kinetic parameters, Kt and Jmax, for Cl transport on individual rectal preparations in this way (i.e. ignoring the small variance in Jsm) gave an indication of kinetic variation that could not have been obtained from a single Woolf plot using calculated from mean and values.
RESULTS
Kinetics of Cl absorption
ISC and increased hyperbolically as [Cl] was elevated on both sides of the rectum (Fig. 1). In contrast, increased linearly and showed no evidence of saturation. As expected, and ISC were nearly identical at all Cl concentrations.
Fig. 2 shows that forward fluxes of 36C1 were significantly lower in K-free saline (P<<0·01) and higher in l00mm-K saline (P<<0·01) when compared to normal saline (l0mm-K). Since is identical when tissues are bathed in saline containing 1 or 140mm-K (Hanrahan, 1982) and increases linearly with chloride concentration (r2 = 0·998; Fig. 1), presumably occurs by passive diffusion. was calculated as at each [K] using the measured in normal saline. Data from individual preparations were fitted by linear regression to the Michealis-Menten equation using the Woolf transformation (Haldane, 1957), because this method is considered to be least sensitive to measurement errors (Blunck & Mommsen, 1978). Weighting procedures were not applied because the type of error (absolute vs relative) was unknown. Woolf plots were linear, as indicated in Fig. 3. For ten preparations used to calculate Ktand at 10 mm-K, 98·2 ±0·65 % of the variation in was attributable to this linear relationship with [Cl].
As shown in Table 1, both/Gand increased significantly (P< 0·001) as the [K] was raised from 0 to 100 mM. Thus the stimulation of Cl absorption by K did not result from a simple increase in affinity of the transporter for Cl.
Cooperative interactions between Cl-binding sites have been demonstrated for transporting epithelia of prawn intestine (Ahearn, 1978) and the mosquito posterior rectum (Bradley & Phillips, 1977; Phillips, Bradley & Maddrell, 1978). Fig. 4 shows that Hill plots of mean had slopes near one at all concentrations of K (e.g. 1·09 at 0mM, 0·91 at 10mM, and 0·99 at 100mM-K). Using the slope as a measure of the Hill constant or number of interacting sites (Segel, 1975), these results suggest that the rate-limiting step in transrectal Cl absorption does not involve cooperative interactions between Cl-binding sites.
Isc and 36Cl fluxes in ‘high K’ saline
All previous studies of transport across insect hindgut have employed high-Na, low- K salines, even though the rectum normally contains a K-rich, low-Na fluid secreted by the Malpighian tubules. Therefore, transepithelial 36C1 fluxes were measured in saline containing 140 mM-K (i) to determine if ISC still equals , as it does in normal saline, (ii) to measure the possible effects of high [K] on and (iii) to examine whether responsiveness to cAMP is altered by this physiological level of K.
Both and ISC increased during exposure to 1 mM-cAMP in high K saline (Table.2). Surprisingly, Rt was 40Ωcm2higher during cAMP stimulation in high-K saline (140 mm) than in normal saline (10mm-K) during cAMP stimulation (Hanrahan, 1982; P << 0·01). As discussed in a later section, this high resistance is explained by the fact that K permeability of the epithelium varies inversely with saline [K].
It is of interest to compare , in high-K saline with , in normal saline. From Fig. 1, /tequivcm−2h−1 at [Cl] = 50mm, [K] = l0mM, in close agreement with observed when [K] = 140 mM and [Cl] = 50 mM (0·9 – 1·3 μequivcm−2h−1 ; Table 2). Chloride permeability (PCl) is apparently not affected by [K], at least over this range. In summary, after stimulation ΔISC equals when external [K] =0mM (Hanrahan, 1982), 10mM (Fig. 1) or 140mM (Table 2). Active Cl transport is the major electrogenic process in stimulated locust rectum under all conditions studied.
Apparent K activation constant (Ka) of Cl transport
Since ΔISC is a good measure of active Cl transport regardless of saline [K], the apparent Ka of K stimulation was estimated from the effects of K addition on ΔISC in the presence of cAMP.
Fig. 5A shows that addition of cAMP alone caused ISC to increase from 0·86 μequiv cm−2 h−1 to l ·57 μequiv cm−2 h−1, in close agreement with obtained in a free saline. When external [K] was then elevated from 0 mm to 100 mm by stepwise condition of KCH3SO4 to both sides, ISC increased similarly to a maximum rate of l2 ·68 μequiv cm−2 h−1. Above 100mm-K, Isc decreased reversibly (Fig. 5A). Vt increased from 7 ·2 to 16 ·8 mV when 1 mm-cAMP was added to K-free saline (data not shown). Rt remained constant when [K] was raised from 10 to 200mm (Fig. 5B), a surprising result considering the normally high K permeability of this epithelium. To control for possible artefacts which might result from increases in osmotic pressure and ionic strength, Na-methylsulphate was added in parallel experiments under identical conditions (Fig. 5C,D). Na addition did not produce large step-lIsce increases in ISC (Fig. 5C). Also, high [Na] (>100 mm) did not inhibit ISC, in marked contrast to the effects of elevated K levels. Rt declined in a predictable manner when saline [Na] was increased above 10 mM (Fig. 5D), in contrast to the relatively constant Rt observed following K addition over the same concentration range (Fig. 5B).
The difference between mean ISC obtained during Na and K additions was used as a measure of K-dependent Cl transport, since Δ ISC equals at all K concentrations. Fig. 6 shows a Woolf plot of the K-dependent ISC stimulation. A linear relationship was obtained between ( [K] /K-dependent ISC) vs [K] when [K] was greater than 2 mM (r2 = 0 · 9984). The Ka was 3·2mM-K and the maximum K-dependent ISC was 7·8 μequiv cm−2 h−1
The effect of 1 mM-cAMP on ISC was also measured in two tissues when choline was the only monovalent cation added to the saline. After 3 h in the absence of both Na and K, cAMP increased ISC by 4·0 and 4·5 μ equiv cm−2h−1, identical to values observed with 200mm-Na present. Stimulations of this magnitude were also observed when both Na and K were replaced by tetramethyl ammonium (TMA). Cyclic AMP caused ISC to increase from 1·33 to 2·62/iequiv cm−2 h−1 and from 0·91 to 2·66 μequiv cm−2 h−1 in two preparations exposed to choline saline lacking Na, K, Ca and Mg. This suggests that the K-independent component of cAMP-stimulated Δ ISC is either independent of these cations, or choline and TMA can substitute for them. In the next section we examine the relative ability of other cations to stimulate Cl transport.
Selectivity of K stimulation of Cl transport
Tissues were equilibrated for 2–4 h under ISC conditions in K-free saline and then exposed to 1 mM-cAMP. After 2–3 h, various test cations were added bilaterally to a final concentration of 40 mM. Fig. 7 shows the selectivity sequence which was estimated by comparing the cation-stimulated ISC after 1 h. Arranged in order of decreasing potency, the sequence was: 1·0K>0·58Rb>0·49Cs>0·08NFU (and 0·2 Na from Fig. 5). The series K > Rb > Cs > Na is sequence I of Eisenman’ (1961), corresponding to a selectivity site having moderately weak field strength. This contrasts with the high selectivity of the Cl site described previously (Hanrahan & phillips, 1980b).
Sidedness of K stimulation
ISC was measured during stepwise addition of K-methylsulphate to either the mucosal or serosal side to determine whether K ‘activation’ of Cl transport occurred specifically at one side of the epithelium. Recta were equilibrated under ISC conditions in K-free saline for 3–4 h and then exposed to 1 mM-CAMP. After ISC reached a new steady-state, aliquots of K-methylsulphate were added to the mucosal or serosal side (final concentration; 2–10mM). Only low [K] was used in order to minimize the K diffusion current caused by a transepithelial K gradient, and to reduce contamination of the K-free side. To estimate K diffusion current, ‘ISC’ was recorded during asymmetrical K additions when Cl transport was abolished by (i) adding 1 mM-azide to normal saline and stirring with Nz, and (ii) by replacing Cl with methylsulphate. Corrections for K diffusion ranged from 0–21 % in the presence of a 10mm:0mm (mucosa: serosa) gradient. The mean K diffusion currents measured in this way were subtracted from the Isc measured in unpoisoned tissues in order to calculate true Cl-dependent ISC with K gradients present. Fig. 8 shows the effects of adding K-methylsulphate to one side of the epithelium. After corrections, ISC attributable to active Cl transport increased from 1·55 to 6·85 μequivcm−2 h−1 when 10mM-K was added to the mucosa. In contrast, ISC was not changed significantly by serosal addition of K (P>>0·2), suggesting that the K activation site is accessible only from the mucosal side.
Transepithelial 42Kfluxes under ISC conditions
Normal saline
Active K transport has been reported across locust rectum in vivo (Phillips, 1964b, c) and in vitro (Williams et al. 1978); however, the relative magnitudes of active and passive components are not known nor are the ionic requirements of K absorption. Considering the dependence of active Cl transport on external K, it was of some interest to study the properties of active K transport, particularly the effect of cAMP, and to test the possibility that there are reciprocal ionic requirements for K and Cl absorption.
Fig. 9 shows the effects of 1 mM-CAMP on (a) unidirectional fluxes of 42K and (b) across recta bathed in normal saline (114mM-Cl, 10mM-K). ,and increased from about 0·35μequivcm−2 h−1 initially to 2·08 and l·56μequivcm−2 h−1 respectively after adding cAMP. A small but significant was observed after 45 min of cAMP stimulation (0·63 ±0·26; P<0·05) but it was less than 7% of the measured under these conditions.
The four-fold stimulation of both and suggests that cAMP increases the aparent transepithelial K permeability () from 0·98 ±0·14 to 4·0 ±0·62 × 10−5cms−1 (x̄±S.E., N = 6). This method of calculating permeability may lead to some error because it assumes that the tissue acts as a single barrier (Schultz & Frizzell, 1976) whereas 42K must penetrate two membrane barriers because the locust rectum is a tight epithelium (Hanrahan et al. 1982). This is discussed in a later section; however, the stimulations of both unidirectional fluxes do indicate that PK increases during cAMP stimulation.
In summary, cAMP has two important effects on transepithelial K movements under 1sc conditions: (i) it produces a large (four-fold) increase in transepithelial K diffusion, and (ii) it induces a small active net K absorption.
Cl-free saline
Table 3A shows the effects of 1 HIM-CAMP on 42K fluxes in Cl-free saline. After 1 h, both unidirectional fluxes increased by about 400%. The significance of was marginal, using a conservative statistical criterion. When Cl was restored to normal levels on both sides (114mM), there was no change in or (data not shown; P>0·2, paired t-test), although ISC increased six-fold to values typical of cAMP stimulation in normal saline (11·4 ±0·74μequivcm−2 h−1 ; N = 12).
These results show that neither the increased K permeability nor the small produced by cAMP is affected by omitting Cl from the saline. In view of this independence, it seems unlikely that there is strict chemical coupling between Cl and K movements at either the apical or basal membrane. This conclusion is further supported by the finding that is less than 8 % of under ISC conditions (Fig. 9) and also by the previous observation that 35 % of is cation-independent.
42Kfluxes in high-K saline
The apical membrane of this epithelium is usually bathed in K-rich (140 mm) Malpighian tubule fluid in vivo. The small observed under ISC conditions during cAMP stimulation (Fig. 9B) might result from the low concentration of K in the saline. Table 3B shows the effects of 1 mM-cAMP on 42K fluxes under ISC conditions when recta were bathed bilaterally in high-K saline (140mM-K, 50mM-Cl). Unidirectional and net 42K fluxes were similar to those measured in normal saline containing only 10mM-K (i.e. Fig. 9). As before, increased to 0· 66μequiv cm−2 h−1during cAMP exposure although this was not significant as judged by the overlap of 90 % confidence intervals. Although increased steadily over the course of the experiment, addition of 1 mM-cAMP did not increase (P>>0·2, paired t-test), in marked contrast to the stimulation observed in normal saline.
Several characteristics of K transfer should be noted. First, was 0·6–0·8 Jtequivcm−2 h−1 under ISC conditions whether saline [K] was 10 or 140 mm. This indicates that active K transport must saturate at low external levels of K (<10 mM). Second, PK decreases several-fold at high [K] because does not increase proportionally when [K] is raised to high levels. This is shown by comparing in Fig. 9 and Table 3 B. PK is three- to four-fold higher when tissues are bathed in normal saline (10mm-K, 114mM-Cl) than in ‘high-K’ saline (140mM-K, 50mM-Cl). Although high-K saline contains less Cl than does normal saline, lower [Cl] could not explain
The reduced PK under high-K conditions, because PK in Cl-free saline (with 10 mM-K) is also four-fold higher than in high-K saline. Finally, cAMP does not increase PK by four-fold in tissues bathed with high-K saline, in marked contrast to those bathed in normal or Cl-free salines containing only 10mm-K. The simplest explanation for these results and the finding that Rt does not decrease when [K] is increased from 10mm to 200mM (Fig. 5B), is that K permeability declines at high external K concentrations.
Fig. 10 shows the concentration dependence of PK. K-methylsulphate was added bilaterally to give concentrations between 2 and 200 mM. was measured during two 15-min flux intervals at each [K] and data from the second period were used in calculations of PK. PK declined from 8 × 10−5 to 1 × 10−5 cms−1 when K concentration of the saline was increased from 2mm to >100mM.
Transepithelial 42 K fluxes under open-circuit conditions
Chloride absorption across locust rectum is electrogenic and must, under opencircuit conditions, be matched by a similar flow of cations from mucosa to serosa or a flow of anions in the opposite direction. To determine whether K diffuses trans-epithelially to maintain electroneutrality during Cl transport, open-circuit 42K fluxes are measured (i.e. under control conditions), then during cAMP-stimulation (10mm-K bilaterally), and finally, when mucosal [K] was raised to 100 mM in the presence of 1 mM-cAMP to mimic normal in vivo K gradients (10:1) across locust rectum.
Vt ranged between 8–10 mV in normal saline (10mm-K) before adding cAMP (Fig. 11A), in agreement with the previous results. Both forward and back fluxes of 42K were less than 1 μequiv cm−2 h−1 (Fig. 11B). Serosal addition of cAMP (ImM) increased Vt from 8 to 28 mV, enhanced by 500%, and produced a small but significant increase in (P<0·01). The resulting ranged from 4·5 to 5·0 μequiv cm−2 h−1. It is noteworthy that equalled at open-circuit both before and after cAMP addition. This result indicates that K is the main counter-ion during active Cl transport even when the [Na] is 11-fold higher than [K] in the external salfl (Hanrahan & Phillips, 1983).
The ratio of unidirectional 42K fluxes at open-circuit is higher in normal saline than that predicted from the Ussing flux ratio equation. Under control conditions (normal saline, no cAMP) the 42K flux ratios were between 4 and 6, as compared to a predicted ratio of 1·4. It is unlikely that this discrepancy could result from active transport since no was observed under ISC conditions in the absence of cAMP (Fig. 10). An alternative explanation for high flux ratios is that transmural K movements are not independent. The flux ratio increased further after addition of cAMP (Fig. 11), which is consistent with the appearance of small under ISC conditions (Fig. 9). However, a step increase in mucosal [K] from 10 to 100 mM elevated the steady-state flux ratio to >100: 1, rather than the predicted value of 10·4. This larger discrepancy is probably not due to enhanced active transport, since is similar whether [K] is 10 mM or 140 mM, but it is again consistent with non-independence between transmural 42K fluxes, as has been shown in the basolateral membrane of the turtle colon (Kirk & Dawson, 1981). Increasing luminal [K] to 140 mM in the presence of cAMP to mimic the situation in vivo (Fig. 11) led to a large increase in and corresponding dramatic decrease in the Vt opposing Cl transport.
DISCUSSION
The results in this paper suggest that K is required on the mucosal side of locust rectum for Cl to be actively transported at a maximal rate. Since this dependence is seen under ISC conditions and ISC equals , K apparently does not stimulate transepithelial Cl absorption simply by acting as a counter-ion. Furthermore, , and are much smaller than under ISC conditions and are Cl-independent. We conclude that 80 % of net K absorption under open-circuit conditions is electrically coupled to transepithelial Cl absorption (Hanrahan & Phillips, 1983). This would constitute a feedback loop because the concentration of potassium in the lumen would indirectly control passive K reabsorption by modulating the rate of electrogenic Cl transport.
Effects of K on active Cl transport
The relationship between steady-state across locust rectum and external [Cl] is satisfactorily described by the Michaelis-Menten equation. K addition increases both Kt and , but do these changes directly reflect the properties of the Cl ‘pump’? The active step for transepithelial Cl absorption has been localized at the apical membrane using Cl-sensitive microelectrodes under identical conditions to those during 36C1 fluxes (Hanrahan & Phillips, 1980b, 1983). The net electrochemical gradient opposing Cl entry varies directly with changes in the rate of Cl transport following K addition, whereas the gradient favouring Cl exit across the basal membrane remains constant. These observations imply that the active entry step is ratelimiting and that steady-state flux kinetics will be largely determined by the apical membrane ‘pump’.
Localized electrical coupling across the apical membrane might be the basis for the K-dependence of Cl transport; i.e. K might depolarize the apical membrane thereby reducing the electrochemical potential against which the Cl pump must work, and juch a mechanism might not be obvious from measurements of transepithelial fluxes, However, in order to explain the 10-fold difference between and under ISC conditions, tight electrical coupling during Cl entry across the apical membrance would require a very large active return of K from cell to mucosal side (i.e. recycling), because the net electrochemical gradient for K across the apical membrane as measured using ion-sensitive microelectrodes is 0 mV under ISC conditions (Hanrahan, 1982). K secretion has not been observed across this tissue under the wide variety of conditions investigated.
KC1 co-entry at a rate equal to the rate of K-dependent Cl transport would contribute significantly to PK. The observation that PK is not changed by Cl removal argues against chemical coupling between potassium and chloride. Moreover, when equivalent electromotive forces (e.m.f.) across the apical (Ea) and basal (Eb) membranes were calculated under open-circuit conditions using membrane potentials and resistances obtained by flat-sheet cable analysis (Hanrahan, 1982), Ea and Eb were −55·7 and −52·5 mV before adding cAMP, and −67·9 and −39·9 mV after adding cAMP (cell negative). This increase in apical membrane e.m.f. could be explained by an electrogenic Cl pump, but not by a model which involves parallel electroneutral KC1 co-entry and K back-diffusion to the mucosal side, because the measured K gradient would generate an e.m.f. of 12·4 mV in the wrong direction under those conditions.
The simplest model for K stimulation of active Cl transport, consistent with all our data, is one in which K enhances active Cl entry in a manner analogous with enzyme activation. This model might also apply to other insect epithelia where K-stimulated Cl transport has been reported (Cooper, Eaton & Jungreis, 1980).
The K-insensitive component of cAMP-stimulated observed in Fig. 5C may result from a single population of Cl pump sites functioning at a low rate under K-free conditions and capable of a graded response to K. Alternatively, two populations of Cl pump sites may exist; one which operates without K and another which is only functional when [K] is elevated.
Passive K transport
is largely passive under open-circuit conditions and electrically coupled to active Cl transport. In support of this view, is only 8 % of when locust rectum is short-circuited whereas equals under open-circuit conditions. K acts as the counter-ion for electrogenic Cl transport even when much higher concentrations of Na are present in the mucosal solution (114 mM-Na DS 10 mM-K). The predominance of K as the counter-ion is ensured in vivo because (i) natural K levels (140 min) are much higher than Na levels (20–40 mM) in the rectal lumen, and (ii) cAMP (which mediates the actions of CTSH) elevates PK by about 400%. In contrast, PNa is unaffected by cAMP (Spring & Phillips, 1980).
When K was added to the mucosal side under open-circuit conditions to mimic the in vivo K gradient, we expected the mucosal side to become negative with respect to the serosal side (despite the stimulatory effect of K on active Cl transport), because epithelial K conductance is normally high. However, no reversal of Vt was observed when mucosal [K] was elevated to 140 mM. Also, Rt did not change when [K] was elevated from 10 to 200 mm. Both these observations are explained if PK declined as [K] was increased. This decline in permeability was confirmed by measuring 42K backflux as a function of bilateral [K] under ISC conditions. The exact mechanism this concentration-dependence has not been studied in detail, but it may be analogous to the inverse relationship between mucosal [Na] and the rate of Na entry at the apical membrane reported in frog skin [Biber & Curran, 1970; Fuchs, Hviid Larsen & Lindemann, 1977; Moreno et al. 1973; Rick et al. 1975; Rotunno, Vilallongs, Fernandez & Cereijido, 1970; Van Driessche & Lindemann, 1979).
Calculating PK from under ISC conditions might result in an over-estimate of transepithelial permeability because it assumes that the epithelium is a single barrier to tracers when in fact PK depends on membrane potentials and intracellular and extracellular K activities (Schultz & Frizzell, 1976). Intracellular potential has been measured under ISC conditions as a function of external potassium concentration (Hanrahan, 1982). Using equation 14 of Schultz & Frizzell (1976), we calculated that 40 % of the apparent decline in K permeability is due to membrane depolarizations and changes in intracellular K activity, but the remaining 60 % of the decline in PK must be due to a real reduction in K permeability. Also, errors due to the simplifying assumptions used in calculating PK do not explain why cAMP stimulates by fourfold in normal saline (l0mm-K; Fig. 9) but not in high-K saline (140 mm-K; Table 3).
Hormonal regulation of salt reabsorption in locust rectum appears to be highly efficient because electrogenic Cl transport and counter-ion permeability (PK) are stimulated simultaneously. What advantages might arise from a K-inhibitable PK? When the hindgut contains unmodified (high-K) Malpighian tubule fluid, a K-sensitive PK would prevent Vt from reversing to negative values and drawing Na from the haemolymph into the gut lumen. The maximum transepithelial electrochemical gradient for Na developed across the rectum in situ is smaller than for Cl or K (Phillips, 1964b,c), and active Na transport during cAMP stimulation is weak compared to Cl absorption (20 %; Spring & Phillips, 1980; Williamset al. 1978). Reducing the loss of Na in this manner may be important for an insect feeding on fresh plant matter that is low in Na (14 mu) compared to K (114mm-lettuce; Hanrahan, 1982). Finally, salt-loaded locusts can produce a strongly hypertonic urine in order to conserve body water (Phillips, 1964a,b,c). A decline in potassium permeability might prevent excess K reabsorption under these conditions.
Properties of active K transport
Net flux of 42K from mucosa-to-serosa was measured under ISC conditions during cAMP stimulation. The presence of a small active absorption of K is consistent with earlier findings that K is maintained far below electrochemical equilibrium in recta of salt-depleted (hydrated) locusts (Phillips, 1964b,c). In the present study, there was no net flux of K until cAMP was added. This differs from the very low rate of K absorption observed by Williams et al. (1978) using a different saline (Berridge, 1966) and a voltage clamp which did not correct for series resistance. We did not measure the Ki of active K absorption in this study; however, it is presumably less than 10 mM-K since was identical when the bathing saline contained either 10 or 140mM-K. This high-affinity, low capacity system for K absorption could be responsible for reducing [K] in the rectal fluid to the low levels reported in salt-depleted locusts (0·5 mM; Phillips, 1964c). In summary, K is absorbed transepithelially by electrical coupling to Cl transport under open-circuit conditions and also by an active system which transports at a low rate but with a high affinity for K. The fact that there was no reduction in the cAMP-stimulated when Cl was omitted from the saline supports the notion that KC1 co-transport is not involved in transepithelial K absorption.
Acknowledgement
This work was supported by NSERC, Canada.