Act0 Physzol S c u d 1991, 141, 185-1'35

ADONIS 000167i291OOO2YN

Lithium absorption by the rabbit ga I - bladder C. P. H A N S E N , N.-H. H O L S T E I N - R A T H L O U , 0. S K YTT, P. P. LEYSSAC and 0. F R E D E R I K S E N Institute of Experimental Medicine, T h e Panum Institute, University of Copenhagen, Denmark H A N S ~cN. ,P., HOLSTEIN-RATHLOU, N.-H., SK0TT, o.,LEYSSAC, P. P. & FREDERIKSEN, 0. 1991. Lithium absorption by the rabbit gall-bladder. Actu Physzol Scund 141, 185-195. Received 5 January 1990, accepted 30 August 1990. ISSN 0001-6772. Institute of Experimental Medicine, University of Copenhagen, Denmark. 1,ithium (Lit) absorption across the low-resistance epithelium of the rabbit gall-bladder was studied in order to elucidate possible routes and mechanisms of Li+ transfer. Li' at a concentration of 0.4 mM in both mucosal and serosal media did not affect isosmotic mucosa-to-serosa fluid absorption. At this low concentration net mucosa-to-serosa Li' absorption was insignificant when the ambient Na+ concentration was 11 5 mM, although the gall-bladder had a significant 1,i' permeability (2.7 x cm s-l) and a significant mucosa-to-serosa Li' gradient developed as a result of fluid absorption. Net Li' absorption was induced at reduced mucosal Na' concentrations (by lowering the Na' concentration down to 50 mM with or without substitution with sucrose, or by adding sucrose to the mucosal medium). This L i b absorption occurred even in the absence of a mucosa-to-serosa Li+ gradient. Na' and Li' absorptions occurring at 50 mM N a ' were inhibited to the same degree by mucosal 1 mM amiloride. Substitution of 5-50 mM (440/,) Na' by Li' in the external medium dose-dependently depressed Na+ absorption by up to i 6 % , while substitution by 50 mM choline had no significant effect. Li' inhibition of Na' absorption was elicited from the mucosal side and was not accounted for by compensatory Li' absorption; water and Na+ absorption rates decreased nearly in parallel. The effects of 0.4 mM amiloride and of substitution with 20 mM Li' were only partly additive. It is concluded that Li+ absorption in the rabbit gall-bladder cannot be explained by passive (paracellular) transport, but must be the result of transcellular, active transport. Both at low and at high concentrations Li' may enter the cell via an Na+/H+exchanger in the apical cell membrane. At high concentrations Li+ may inhibit Na' absorption by interference with the exchange mechanism and/or via effects at the cytoplasmic level. The Li+ transfer mechanism across the basolateral cell membrane remains unknown.

Key words :amiloride, gall-bladder, lithium, rabbit, sodium absorption, sodium-proton exchange, water absorption.

T h e mechanism by which Li' is absorbed in leaky epithelia is largely unknown. I n mammalian renal proximal tubules Li' is absorbed in parallel with N a + salts and water (Hayslett & Kashgarian 1979, Leyssac et a!. 19901, which results in an end proximal tubule-to-plasma Li' concentration ratio of 1. Li' reabsorption beyond the Correspondence : Ole Frederiksen MD, University Institute of Experimental Medicine, The Panum Institute, building 10,5, Blegdamsvej 3, DK-2200 Copenhagen N, Denmark.

proximal straight segment is insignificant under normal sodium balance conditions and therefore L i + clearance becomes a measure of fluid outflow from the proximal tubule (Thomsen et af. 1981). Both passive paracellular permeation (Burg & Green 1976, Corman et af. 1981) and transcellular transport by an N a + / H + exchange mechanism (see review by Aronson & Igarashi 1986) have been suggested to be involved in Li+ reabsorption across the proximal tubular epithelium. I n another mammalian high-conductance epithelium, the rabbit gall-bladder, the passive

185

186

C. P.Hunsen et al.

(paracellular) permeability for Li+ is almost as high as for Na+ (Moreno & Diamond 1976), as is also the case for the proximal tubule (Corman et ul. 1981). Previous results on Li+ transfer have been obtained from everted rabbit gall-bladders, in which incubation takes place with solutions on the mucosal side only (Peters & Walser 1966). This preparation does not allow a distinction between cellular and paracellular transport of Li+, and results from this preparation can be interpreted in several ways. T h e aim of the present study was to characterize transepithelial Li+ transport in the rabbit gall-bladder. Transport of Li+ was studied at low I,i+ concentrations similar to those prevailing under lithium clearance experiments. Furthermore, the effect of Li+ on isosmotic Na+ and water absorption was studied when Na+ was substituted by Li+ in concentrations ranging from 5 to 50 mM. The results demonstrate that the rabbit gallbladder is unable to absorb Li+ in the presence of 0.4 mM I,i+ at high Na+ concentrations, while lowering the external Na+ concentration induces net Li+ absorption. Li+ concentrations exceeding 5 mM acutely and reversibly inhibit Na+ absorption, while substitution of 50mM Na+ by choline has insignificant effect on Na+ absorption. MATERIALS AND METHODS Female white rabbits weighing about 3 kg were killed by a blow on the neck. The gall-bladder was isolated as previously described (Frederiksen 1983). T h e effects on Na+ absorption of Na+ substitution by Li+ or choline were studied in gall-bladders mounted in Ussing chambers, while Li+ absorption was investigated in gall-bladders used as sac preparations. T h e composition of the control Ringer solution was (in mM): 115 Na+, 7 K+, 2 Caz+, 1.2 Mg2+,104.7 Cl-, 17.5 HCO;, 1.2 H,PO;, 5 glutamate and 11 glucose. At these Na+ and K+ concentrations the reabsorption rates were maximal and were maintained high for up to 8 h (Frederiksen & Leyssac 1969). Bumetanide was a kind gift from Leo Pharmaceutical Products, Copenhagen, Denmark ; and amiloride was kindly provided by Merck, Sharp & Dohme, Copenhagen, Denmark. Experiments zn Usszng chambers. Gall-bladders were cut open and mounted vertically between Ussing halfchambers (Frederiksen 1983). Each half-chamber contained 15.0 ml of Ringer solution. T h e exposed gall-bladder area was 0.9 cm2. Constant stirring, oxygenation, and a p H of 7.4 were obtained by gas

lifts containing 969;) 0, and 404, CO,. The temperature was kept constant at 37 "C by means of water jackets. All experiments were substitution experiments in which 5, 10, 20 or 50 mM Na' was replaced by equimolar concentrations of J,i+ or choline on both sides of the gall-bladder, unless otherwise stated. Unidirectional Na' fluxes were measured as '%a+ fluxes in individual gall-bladders (carrier-free "NaCI was obtained from Amersham, UK) (Frederiksen 1983). Net Na' absorption (J),: was calculated as 3s.NC, = 37"" -3s.bM, where 32 is the mucosa-to-serosa Na+ flux and 33 is the serosa-to-mucosa Na+ flux. T h e transepithelial electrical potential difference (p.d.) and resistance (R,)were measured as described previously (Frederiksen 1983), and R, was corrected for fluid resistance in the absence of the tissue. Experiments in sac preparations. Net water absorption was measured gravimetrically in gall-bladder sac preparations (Frederiksen & Leyssac 1969). The gall-bladder was filled with Ringer solution which was saturated with 96O/L 0, and 404, CO,, and suspended in a beaker containing 150 ml Ringer solution maintained at 37 "C. The outside bathing solution was continuously stirred and oxygenated by a stream of bubbles of 96yb 0, and 40,; CO,. The following six types of experiments (sac protocols a-f) were performed : (a) The solutions bathing the mucosal and serosal sides were identical and contained 115 mM Na' in the control period and 65 mM Na+ plus 50 mM Li' in the experimental period (n = 6). (b) I n the control period mucosal and serosal bathing media contained 115 mM Na' and 0.4 mM Li'. In the experimental period 100 mM sucrose was added to the mucosal medium ( n = 8). (c) I n the control period mucosal and serosal bathing media contained 115 mM Na+. In two subsequent experimental periods the Na+ concentration in both media was reduced to 75 and 50 mM, respectively, by removing NaCI. At the start of the control and experimental periods the media contained 0.4 mM Li+ ( n = 6). (d) Protocol similar to protocol (c), except that in the experimental periods sucrose was added to keep the osmolar concentration constant ( n = 7). (e) In the control and experimental periods both media contained 50 mM Na' and 0.4 mM Li', and in the experimental period 1.0 mM amiloride was added to the mucosal medium ( n = 4). (0 I n the control and experimental periods both sides of the gall-bladder were exposed to 115 mM Naand the serosal medium contained 0.8 mM Li' initially. Amiloride (1.O mM) was added to the mucosal side in the experimental period (n = 4). Weighing intervals of 10 min were used. The luminal content was renewed between the weighing periods with the exceptions stated below. Fluid absorption rate was allowed to stabilize during control

~,~

Rabbit gall-bladder lithium absorption periods, whereupon absorption was measured in a 30rnin weighing period without renewal of the luminal content. Then the sac was emptied, rinsed and refilled with the experimental solution. In the experimental periods involving lowering of the Na+ concentration (protocols c and d) or addition of amiloride (protocols e and f), fluid absorption was allowed to stabilize during two 10-min weighing periods followed by a 30rnin period without renewal of the luminal fluid. In the series involving mucosal addition of sucrose (protocol b), fluid absorption in the experimental period was determined from weighing periods of 5 min; the total period with sucrose averaged 92 min (range 61-185 rnin), during which period the luminal content was not renewed. Total luminal fluid content at the start and at the end of the 30-min weighing periods of control and experimental periods (in the sucrose series the entire experimental period) were determined at the end of the experiment by subtracting the weight of the emptied sac preparation. Absolute and fractional changes in fluid and ion absorption rates induced in the experimental periods were determined as changes in proportion to the stabilized transport rates in control periods. This procedure is justified by previous observations that rabbit gall-bladders show a spontaneous, time-dependent decrease in ion and water absorption of maximally 5"/" per hour when incubated for up to 8 h under control conditions, as in the present study (Frederiksen & Leyssac 1969, Leyssac et al. 1974, Frederiksen 1978, 1983). Measurement of net Li+ and N u + transports. Li+ and Na+ concentrations were measured in the bathing Ringer solutions and in the luminal solutions aspirated at the end of the 30-min control and experimental weighing periods (in the sucrose series in protocol (b) the entire experimental period). Li+ was measured by atomic absorption spectrophotometry and Na+ by flame emission spectrophotometry (Perkin-Elmer 2380 Atomic Absorption Spectrophotometer). At the end of the experiment the gall-bladder area was calculated from the height and the maximal diameter of the gallbladder, assuming the surface to be the combined surface of a cone and a hemisphere. Transport rates per min cmz were calculated. Calculations and statistics. T h e apparent I,i+ permeability (Pr,i)was determined from the group of sac experiments designated sac protocol (0. PLi was calculated from the equation: l/Vv-p,.iA)

+ PLiC,

x [1n(G--LlA) C,(t) l/Jv x [In (V(0)-JV

A)];1n(')

+

t ) ] ;= 0,

(1)

which is obtained by solving the differential equation :

-a t ) C,(t))ldt

d(( V(0)

= -p,.i

A(C,(t) - CJ.

(2)

A. is the volume flow, which is assumed to be constant. C,(t) is the Li+ concentration in the mucosal fluid at time t , and C, is the Li+ concentration in the beaker

187

(serosal fluid), which is assumed to be constant. A is the membrane area. V(0)is the gall-bladder volume at t = 0. The term (V(0)-A. t ) is then the volume of the gall-bladder at time t. Equation (2) expresses the passive I,i+ influx as solely driven by a chemical gradient. Since the major part (around 95%) of passive ion permeation in the gall-bladder epithelium occurs via the paracellular tight junction route (see, for example, Moreno & Diamond 1976), it seems justified to disregard the influence of cell membrane potentials on serosa-to-mucosa Li+ fluxes. The influence of the small transepithelial potential difference (2-3 mV, mucosa positive) is insignificant (max. 5%) and is therefore disregarded. The absence of a solvent drag component in rabbit gall-bladder passive (paracellular) ion permeation is supported by the following arguments : (1) The tight junction pathway is not the main route for water transport across the rabbit gallbladder epithelium (Wright et al. 1972, van 0 s & Slegers 1973, Frederiksen et al. 1979). (2) Serosa-to-mucosa Na+ flux is not altered when water (and net Na+) absorption is inhibited by mucosal amiloride, and amiloride does not affect the paracellular conductance (Frederiksen 1983). (3) The apparent Li+ permeability calculated from eqn (1) is not significantly different in the absence and presence of mucosal 1 mM amiloride (see Results). Finally, mucosa-to-serosa transport of Li' which has entered the small mucosal compartment during the flux periods can be neglected since Li+ absorption in the presence of 115 mM Na+ is not significantly different from zero (see Results). Results are presented as means & SEM. Paired and unpaired t-tests were applied as appropriate, and P < 0.05 was considered statistically significant.

RESULTS Efects of Lz+ on Na+Jtuxes

Figure 1 shows the effects of bilateral substitution of 50 mM N a + by Li' on unidirectional and net Na+ fluxes. T h e lower part shows that JE:, rapidly and reversibly was inhibited by 76% by this N a + substitution. T h e upper part demonstrates that the decrease inJE%, primarily resulted while 3 : ; decreased in from a decrease in proportion to the change in bathing medium Na' concentration. Similar substitution experiments were carried out with 5, 10 and 20 mM Li'. Figure 2 shows that the fractional decrease in J::t was larger than the fractional change in Na+ concentration ; e.g. with 10 mM Lit and 105 mM N a + (9% substitution) J::, was reduced by 29%. T h e

Jz,

C . P. Hunsen et al.

188

OLI'

700r

0 Ll' 115mM Na'

50mM LI+ 65mM Na'

115rnM Na+

700r

rMS

t-----l

1

,001

OL'

'

'

'

'

'

=E

200

m; .ltx-x-x--x-

l o0o t ,

1

I

I

I

80

LO

'

'

I

I

control

IOrnMLi' uniiot

IOrnML,' bilat

Fig. 3. Side specificity of the effect of substitution of 10 mM Na+ by Li+ absorption in rabbit gall-bladders evaluated from measurements of mucosa-to-serosa Na' fluxes. Mucosal (M) and serosal (S) substitutions were each performed in five experiments. Meansf SEM (bars).

rNet

\

'

I

120 rninutes

I

160

1

200

Fig. 1. Effects of bilateral substitution of 50 mM Na+ by Li+ on Na' fluxes in rabbit gall-bladders measured in Ussing chambers. The upper graph shows unidirectional Na' fluxes in the mucosa-to-serosa (MS, closed circles, n = 5) and serosa-to-mucosa (SM, open circles, n = 4) directions as means SEM (bars). The lower graph shows net Na+ absorption calculated as the difference between unidirectional fluxes.

calculation 0f3;;, for each Li' concentration was based on the means obtained from 4 7 determinations 0f3:; and 2-4 determinations 0 f 3 g T h e steady-state Na+ fluxes after 45 min were used.

Eflects of' choline on

I n order to compare the Li+ effects with the effects of a supposedly inert and non-transportable monovalent cation, 50 mM Na' was substituted by 50 mM choline bilaterally for 45 min. Both 38; and 3:; were reduced reversibly by substitution with choline. Thus, 3g decreased from 483 2 52.3 to 360 2 26.2 and reversed to 473 21.6 nmol min-' cm-' ( n = 5 ) , and 324 decreased from 189 19.4 to 102 2 7.9 and reversed to 1 9 0 k 18.7 nmol min-' cm-' ( n = 3). However, calculated 3EEt was only insignificantly reduced (12%, P > 0.6), while by 76"/0 substitution by 50 mM Li' reducedJ;;, (see Fig. 2 ) .

201

L

0

10

20

30

[ill

ImM)

LO

50

Fig. 2. Dose-response curve for the effect of bilateral substitution of Na+ by Li' on net Na+ absorption in rabbit gall-bladders. The sum of the Na' and Li' concentrations was kept constant at 115 mM. Net Na+ absorptions were determined as demonstrated in Fig. 1, and calculation of the fractional change in net Na' absorption at each Li' concentration was based on 4-7 mucosa-to-serosa and 2 4 serosa-to-mucosa Na+ flux experiments. Means SEM (bars).

Nu' fruxes

Side spectficity

3::

ofthe

Li+ effect on 3%

was measured in a control period, a period with 10 mM Li+ on only one side of the gallbladder, and finally with 10 mM Li+ on both sides of the gall-bladder. As seen from Fig. 3, substitution with Li+ solely on the serosal (S) side did not change 3% significantly, whereas mucosal (M) presence of Li+ significantly into the same extent as did bilateral hibited presence of Li'. Unilateral substitution of 10 mM Na+ by Li' can be calculated to cause a change

3g

R a b b i t gall-bladder lithium absorption

189

Table 1. Absorption of Na', Li+ and water in rabbit gall-bladder sac preparations

3""+

Incubation

3,

(nl min-' cm-')

(nmol min

115 Na+/O Li' 65 Na+/SO Li' b3NFt

1498+2ll 581 k32 - 6 1 3;)

168 5 34 .54 & 7 - 68 yh

conditions (mM)

NPt

cm-')

3kt

(nmol min-' cm ')

-

22+6 ~

Means fSEM ( n = 6 ) . in the passive (paracellular) component 0 f 3 & of maximally 10 nmol min-' cm-' (PKain gallbladders mounted in Ussing chambers is cm s-'; calculated from 3:; in Fig. 1.65 x 1). Thus, a correction for this changc in the passive component of 3;; under the different experimental conditions does not affect the conclusion that T i + exerts its inhibitory effect on Na+ absorption from the mucosal side of the gall-bladder.

EfJcts of bumetanide and amiloride on Li+ inhibition of Naf absorption It was tested whether the inhibitory effect of Lit on Na+ absorption could be prevented by addition of transport inhibitors. Bumetanide ( M) on the mucosal side did not change Na' fluxes, but subsequent substitution of 20 mM Na+ by Li' in the continuous presence of bumetanide reduced 3E:t by 340/& This value is not significantly different from the inhibitory effect (37%) of the same substitution in the absence of bumetanide. Amiloride (0.4 mM) on the mucosal side reduced 3E:t by 53 yo.Subsequent substitution with Li+ (20 mM) led to a 26°/b additional reduction of 3E:,.This 26% reduction is not significantly different from the 37% inhibition found in the control group (0.05 < P < 0.10). The combined effect of amiloride and Li' was a 65 0; inhibition, which is significantly greater than that of Li' alone (P < 0.01) or amiloride alone ( P < 0.01). Thus, the effects of amiloride and Li+ are at least partly additive.

Effects of Na' substitutions on p.d. and R, Potential differencc was 2-3 mV (mucosa positive) in control periods. Substitution of 50 mM Naf by Li+ did not change p.d. significantly, whereas substitution of 50 mM Na+ by choline led to a marked reversible mucosal hyper-

polarization to a mean p.d. of 8.7 mV. The control R, was 45-55 R cm2. Replacement of 50 mM Na+ by Li+ increased R, slightly, while substitution with 50 mM choline reversibly increased the mean R, by about 20 R cm'. These changes in the electrical parameters are in accordance with the notion that the passive paracellular permeabilities to Na' and Li+ are almost equal, while choline has a very low permeability. Potential difference and R, were unaffected by bumetanide and amiloride.

Effect o f substitution of Na' by Li' on water and N a + absorption (sac protocol a ) Table 1 shows that substitution of 50 mM Na+ by Li+ reduced 3i:t by 68%, in reasonable agreement with the 76% found in Ussing chamber experiments. T h e decrease in 3::, was . net water accompanied by a decrease of 61 yo In absorption Uv), and by a net Li' absorption Uget) of 22 nmol min-' cm-'. This Li+ absorption can only account for 22/( 168- 54) = 19O/, of the decrease in Na' absorption (Table 1). T h e controlJE:, (168f34 nmol min-' cm-') was lower than the value found in Ussing chamber experiments (407 nmol min-' cm-').

Effects of mucosal hypertonicity on the rate of Li+ absorption (sac protocol b ) In the presence of 115 mM Na' (and 0.4 mM Li') in the Ringer solutions bathing the mucosal and the serosal side of the gall-bladder the average value of net fluid transport rate c7,) in the control period was 1309 144 nl min-' crn-'. Addition of 100 mM sucrose to the mucosal solution caused a drop in Jv to a final level of 498 f50 nl min-' cm-'. Table 2 demonstrates that in the control period JFet was not significantly different from zero. Mucosal addition of sucrose evoked a statistically significant (P< 0.01) increase inJ??,. Addition of sucrose caused

190

C. P.Hunsen e t al.

Table 2. Effects of mucosal addition of 100 mM sucrosc on net absorption of Li' and Na' and the development of mucosa-to-serosa ion concentration gradients in rabbit gall-bladder sac preparations

.& (nmol min

3& (pmol min-'

~

End

._ ..

Start

End ~-

-~

153.0k 18.1 125.6f13.8

50.9k45.2 109.6k25.1

"a+I,,/"a+l, ...

_ _ _ _ ~

Start

cm-2

cm-2 ~

Control period Sucrose period

[Li+l,,,/[LiAls I

1 0.91 k0.02

1.32k0.02 0.98k0.02

1 1.01 k0.01

0.Y 5 0.0 1 0.51+0.03

Values are means f SEM (n = 8). T h e initial ambient Na' and I,i' concentrations in both periods were 115 and 0.4 mM respectively; and the initial mucosal sucrose concentration in the sucrose period was 100 mM. The length ofthe transport measurement period was 30 min in the control period and averaged 92 min in the sucrose period.

N

contrast, the Li+ concentration ratio increased significantly during the control period (from 1 to 1.32 f0.02, corresponding to a mucosal Lit concentration of 0.52 mM). I n the presence of sucrose the ratio for Na+ decreased as expected, whereas the ratio for Lit remained almost unchanged and close to unity (0.98k0.02; n.s.).

2Or

Effect ofNa+ concentration on the rate of Li+ absorption (sac protocols c and d )

1

115

100

75 "a']

50

(mM1

Fig. 4. Effects of lowering the Na+ concentration in

the mucosal and serosal bathing media on net absorption rates of water, Na+ and Li' in rabbit gallbladder sac preparations. The Na+ concentration was reduced by lowering the concentration of NaCI. The initial Li' concentration in the bathing media was 0.4 mM. For further explanations see Materials and Methods (sac protocol c). Values are meansfSEM (bars); n = 6. a slight but non-significant decrease in

3:f,.

Table 2 also shows the mucosa-to-serosa (m-tos) concentration ratio of both 1.i' and Na+ at the start and at the end of both the control and the sucrose period. At the start of the control period the Na+ concentration ratio was equal to 1 but fell to just below 1 at the end of the period. I n

Figure 4 demonstrates that, in spite of a more than 50 reduction in the Na+ concentration (from 115 via 75 to 50 mM by removal of NaCl), Jv remained almost constant, while 3::t decreased. 3:?,,on the other hand, increased from 42 f 6 1 to 261 f31 pmol min-' cm-' ( P < 0.05) when the Na' concentration was lowered from 115 to 50 mM in the presence of 0.4 mM Li+ on both sides. Table 3 (upper part) shows that the m-to-s Li+ concentration ratio increased from 1 at the beginning of the periods to 1.25f0.03, 1.24 f0.04 and 1.17 0.03 (corresponding to mucosal Li' concentrations close to 0.5 mM) at the end of the 30-min transport periods at Na+ concentration of 115,75 and 50 mM respectively, while the similar m-to-s concentration ratios for Na+ decreased slightly to 0.95 f0.02,0.92 f0.02 and 0.83 0.02. When similar reductions in the ambient Na+ concentration were performed at equiosmolar conditions (from 115 via 75 to 50 mM Na+ by removal of NaCl and addition of sucrose; protocol d) 3& again decreased, from 158 f21 to 45 f9 nmol min-' cm-2. Under these conditions, however, 3valso decreased from 1385 f50 to 330 k 4 0 nl min-' cm-2. Jyet again increased significantly in the absorptive direction from - 84 f 108 pmol min-' cm-2 (small but non-

Rabbit gall-bladder lzthzurn absorptzon

191

Table 3. Development of mucosa-to-serosa ion concentration gradients for Li+ and Na+ in rabbit gall-

bladder sac preparations during 30-min transport periods at barious ambient Na+ concentrations Extracellular Na+ concentration --

Sac protocol (c) : no osmolar substitution with sucrose (n = 6) Sac protocol (d): osmolar substitution with sucrose (n = 7)

-

-

~~~

115 mM

75 mM

50 mM

[Li+jn,/[Li+], [Na+],,/[Na+],

1.2.5 k0.03 0.95 k 0.02

1.24k0.04 0.92 k 0.02

0.83 k 0.02

[I,i+],/[Li+], [Na+],,/[Na'],

1.30k 0.04

0.98k0.02

1.08 k 0.03 0.71 k0.03

0.98 f0.03 0.72k0.07

1.17+0.03

Values are means +SEM. In each experiment Na+, Li+ and water absorptions were measured in a control period with an ambient Na' concentration of 115 mM. In two subsequent experimental periods the Na' concentration in the bathing media was reduced to 75 and 50 mM by removing NaCl without (upper part) or with (lower part) simultaneous osmolar substitution with sucrose. At the beginning of all 30-min transport periods the bath media contained 0.4 mM Li+, and the mucosa-to-serosa concentration gradients for Li+ and for Na' were equal to 1.

significant net secretion) at 115 mM Na' to 151 +47 pmol min-' cm-2 at 50 mM Na+ ( P < 0.05). T h e lower part of Table 3 shows that the m-to-s concentration ratio for Na+ developing during the 30-min transport 'periods at equiosmolar substitution of NaCl by sucrose were significantly smaller than without substitution with sucrose ( P < 0.05). Simultaneously the gradients for Li+ vanished. T h u s , at 50 mM Na+ (and 0.4 mM Li+) the m-to-s concentration ratio for Li+ after the 30-min transport period was 0.98 f0.03. T h e corresponding ratio for Na+, however, was 0.72k0.07, matching a final luminal Na+ concentration of 36mM. T h e development of such a low Na+ concentration may explain the decrease in both J,,.and during large Na+ substitutions with sucrose.

tions of zero on the mucosal side and 0.8 mM on the serosal side, influx rates of Li+ were determined, and PI,iwas calculated using eqn (1). PI,i was found to be (2.7k0.27) x cm s-l. I n the presence of 1 mM amiloride on the mucosal side P,,iwas (2.5 k0.13) x cm s-l. T h e two values are not significantly different. DISCUSSION

I t has been suggested that in the proximal tubule (Corman et al. 1981), in the small intestine (Diamond et al. 1983), and in leaky epithelia in general (Ehrlich & Diamond 1980), net Li+ absorption takes place through the paracellular tight junction pathway by means of passive forces. Support for this suggestion is the lack of any stimulatory effect of lithium on renal Na', Amzloride effect on the rate of Li+ absorptzon K+-ATPase (Gutman et al. 1973). I n the gall(sac protocol e ) bladder a mucosa-to-serosa Li+ concentration With 50 mM Na+ and 0.4 mM Li+ on both sides gradient is created as a consequence of fluid of the gall-bladder, 1 mM amiloride on the absorption (see Tables 2 and 31, and it might be mucosal side reduced J & from 92.0f3.5 to argued that this may account for part of the Li' 24.1 3.4 nmol min-' cm-2 (74%) andJ;,, from absorption by diffusion. I n the gall-bladder sac 251 88 to 86 f29 pmol min-' cm-' (660/,,). preparation the passive Li' permeability was T h e fractional changes were not significantly 2.7 x cm s-l and the passive permeability to different. Na+ was 2.3 x lo-' cm s-l (Frederiksen 1983). Bilateral substitution of up to 50 mM Na+ by Li+ did not alter the epithelial conductance, indiApparent Li+ permeabzlity (PL,)in the rabbit cating that the passive permeability to Li+ is gall-bladder (sac protocol 8 nearly identical to that of Na+ under the present In experiments with an ambient Na' con- conditions. This is in accordance with observacentration of 115 mM and initial Li' concentra- tions by Moreno & Diamond (1976). With a Li+

192

C. P . Hansen et al.

permeability of 2.7 x I O F cm s-', and assuming a linear increase in the mucosa-to-serosa Li+ concentration ratio during a transport period, the passive absorption of Li' will amount to maximally 70 pmol min-' cm-' during control conditions (0.4 mM Li' and 115 mM Na' bilaterally). Under this condition no significant net absorption of Li+ was observed (control periods in sac experiments using protocols c and d). Thus, passive transport of I,i+ is not a likely explanation for the absorption of Li+ induced by lowering the Na+ concentration. At reduced (mucosal) Na+ concentration, obtained by removal of NaCl (hyposmotic Na+ reduction), by substitution of NaCl with sucrose (isosmotic Na' reduction), or developed after addition of sucrose to the mucosal medium (mucosal hyperosmotic Naf reduction), a significant net Li' absorption is induced. This occurs even in the absence of a concentration gradient for Li+ across the epithelium. T h e results suggest that the Li+ absorption is active and induced by low mucosal Na+ concentration. Furthermore, the stimulation by low Na+ occurs independently of the osmolality of the exiracellular medium. T h e possibility that net 1 2 absorption induced by addition of 100 mM sucrose to the mucosal medium was the result of an increase in junctional permcability is unlikely, since this concentration of sucrose causes a decrease, and not an increase, in rabbit gallbladder epithelial conductance (Smulders et al. 1972, Wright et ul. 1972, Frederiksen, unpublished observations). An electroneutral Na+/H+ exchange mechanism is present in the plasma membrane of virtually all vertebrate cells and plays a role in cellular p H homeostasis (see reviews edited by Aronson & Boron [1986], and Grinstein & Rothstein [1986]). I n some mammalian leaky epithelia the Na+/H+ exchanger is present only in the apical cell membrane (Murer et ul. 1976, Kinsella & Aronson 1980). I n apical membrane vesicles from the proximal tubule and the small intestine it has been demonstrated (1) that Li+ competes with Na+ at the external transport site of the exchanger, (2) that Li+ and Na+ enter the vesicles in 1 : l exchange for H+, and ( 3 ) that amiloride is a competitive inhibitor (Giinther & Wright 1983, Aronson & Igarashi 1986). T h e present observations (1) that Li+ absorption in the rabbit gall-bladder at low Li+ concentration is induced by lowering the extracellular (mucosal)

Na+ concentration, and (2) that Li+absorption is inhibited by amiloride, are compatible with the possibility that active Li+ absorption involves 1.i' uptake via an Na+/H+ exchanger. Low concentrations of I,i+ (0.4 mM) similar to those used in vivo under Li' clearance determinations do not affect isosmotic fluid absorption in rabbit gall-bladder. In contrast, substitution of 5-50 mM Na' by Li' inhibits isosmotic Na' and water absorption. Substitution of 50 mM Na' (44(%,)with Li+ caused a 76'1/;, decrease in net Na+ absorption (Fig. l), whereas a similar substitution by choline did not affect Na' absorption significantly. T h e major effect of substitution by Li+ was, therefore, not caused by the change in Na+ concentration. Furthermore, the decrease in Na+ absorption could not be explained quantitatively by a compensatory increase in I,i+ absorption. At substitution of 50 mM Na+ by Li+, Li+ absorption could only account for 19(;() of the decrease in Na' absorption (see Table 1). In isolated, microperfused rabbit proximal tubules it has be& demonstrated that isosmotic fluid absorption decreases by 50";) when 75 mM (50%) of the luminal and peritubular Na+ is substituted by T i ' (Corman et al. 1981). The authors concluded that Li+ was not actively absorbed; that 1.i' served as an inert substitute for Na+; and that Na+ and fluid absorption depended on extracellular Na+ concentration. These conclusions are not compatible with the observation that substitution of up to 500/;, Na+ in luminal and peritubular fluid with choline does not affect proximal tubular Na+ and water absorption (Gyory & Lingard 1976). T h e results by Corman rt al. may rather indicate that high concentrations of Li' exert an inhibitory effect on isosmotic Na' and water absorption also in the proximal tubule. Interference of Li+ with isosmotic Na+ and water absorption in the rabbit gall-bladder may be the result of competition between Li+ and Na+ somewhere in the transport process. Thus, Li+ might interfere with Na+ uptake across the apical cell membrane. Again, the Na+/H+ exchanger appears to be a likely candidate. The observation that the decrease in Na+ absorption was higher than should be expected from the fractional substitution of Na+ by Li+ (see Fig. 2 ) is compatible with Li' interference with Na+ absorption at the site of the exchanger, if the affinity for Li+ is higher than for Na+. This has

Rabbit gall-bladder lithium absorption not been measured in the rabbit gall-bladder, but it has been shown to be the case in renal apical membrane vesicles (Aronson & Igarashi 1986). Amiloride (0.4 mM) inhibited the Na+ absorption. Whether amiloride was present or not, 20 mM 1.i' had a similar fractional inhibitory effect on the Na+ absorption. T h e absolute inhibition caused by Li' in the presence of amiloride was, of course, much smaller than in its absence. Furthermore, the inhibitory effect of Li' was exerted only from the apical side of the epithelium (see Fig. 3). These observations are also compatible with the hypothesis that Idit crosses the apical cell membrane by an Na'/H+ exchange mechanism. I n the Necturus gall-bladder mucosal Li+ has been demonstrated to induce extracellular acidification (Weinman & Reuss 1982). An Na+/Cl- (or Na+/K+/2CI-) symport mechanism has been suggested to exist in the apical membrane in the rabbit gall-bladder (Frizzell et al. 1975, Cremaschi et al. 1983). Such a mechanism does not seem to be involved in the effects of T i + on Na+ absorption, since I,i+ caused the same decrease in net Na+ absorption in bumetanide-treated and in non-treated gallbladders. Bumetanide had no effect on transcellular Na+ transport either. This is in agreement with earlier results (Frederiksen 1985), and also with later results from Cremaschi et al. (1987). Thus, the results support the possibility that the effect of Li+ on Na+ and water absorption may involve interference and uptake of Li' via an apical, amiloride-sensitive N a + / H + exchange mechanism rather than by Na+/CI- co-transport. However, the effects of Li+ and amiloride could still be via separate mechanisms, although the final result is an inhibition of transepithelial Na' salt and water absorption. T h e observed inhibition of Na+ absorption by Li' may also be related to cytoplasmic effects of Li'. Intracellularly Li' may induce accumulation of inositol 1,4,5-triphosphate and release of Ca2+ from the endoplasmic reticulum (see review by Berridge & Irvine 1984). I t is not known whether Li' increases cytoplasmic free Ca+ concentration (Ca,) in gall-bladder cells, but it has been shown that the Ca2+ ionophore A23187 and the Ca'+ channel activator BAY K 8644 cause Ca2+dependent inhibitions of Na+ and water absorption (Frederiksen & Leyssac 1984, Hansen & Frederiksen 1990), and in isolated rabbit gall-

193

bladder epithelial cells BAY K 8644 elicits an increase in Ca,, which is dependent on extracellular Ca2+ (Bouchelouche et al. 1989). Thus, it is a possibility that high concentrations of Li' inhibit Na' and water absorption as a consequence of elevated Cai. Other effects such as interference with cellular p H regulation may also occur. Although mammalian proximal tubules and gall-bladder epithelia have many transport characteristics in common, there seems to be differences in their abilities to absorb I.i+ at low Li+ concentrations, where a significant net Li' absorption in the rabbit gall-bladder requires lowering of the extracellular Na+ concentration to nearly half the physiological concentration. The present study was supported by grants from the Id.P. 11. Poulsen and wife S. C. Poulsen Foundation (scholarship to C. P. H.), the Danish Medical Research Council and the NOVO Foundation (to O.F.). The skilful technical assistance of Mrs Lisbeth Wybrandt, Miss Pia Thuro Hansen and Mr Ian Godfrey is gratefully acknowledged. Parts of the present investigation were presented at the Scandinavian Physiological Society meeting in Copenhagen, November 1986 (Hansen & Frederiksen 1987) and at the XXXI International Congress of Physiological Sciences in Helsinki, July 1989 (Hansen el al. 1989). The present address for Dr C. P. Hansen is Physiologisches Institut, Albert-Ludwigs-Universitat, D-7800 Freiburg, FRG; and for Dr N.-H. HolsteinRathlou, Department of Physiology and Biophysics, University of Southern California, Los Angeles, CA 90033. USA. REFERENCES ARONSON, P.S. & BORON,W.F. (eds.) 1986. Current Topics tn Membrane and Transport, vol. 26 Nu+-H' Exchange, Intracellular p l l , and Cell Function, pp1-31 5. Academic Press, Orlando, Florida. ARONSON,P.S. & IGARASHI, P. 1986. Molecular properties and physiological roles of the renal Na+-H' exchanger. In: P.S. Aronson & W.F. Boron (eds.) Current Topics in Membranes and Transport, vol. 26 Na+-H' Exchange, Intracellular p H , and Cell Function, pp. 57-75. Academic Press, Orlando, Florida. M.J. & IRVINE,R.F. 1984. InositolBERRIDGE, phosphate, a novel second messenger in cellular signal transduction. Nature 312, 3 15-32 1. BOUCHELOUCHE, P.N., HAINAU, B. & FREDERIKSEN, 0. 1989. Effect of RAY K 8644 on cytosolic free calcium in isolated rabbit gall-bladder epithelial cells. Cell Calcium 10, 3746.

194

C. P.Hansen et al.

BURG,M.B. & GREEN,N. 1976. Role of monovalent ions in the reabsorption of fluid by isolated perfused proximal renal tubules of the rabbit. Kidney Int 10, 221-228. CORMAN, B., ROINEL,N. & DEROUFFIGNAC, C. 1981. Dependence of water movement on sodium transport in kidney proximal tubule : A microperfusion study substituting lithium for sodium. 3 Membr Biol 62, 105-1 11. CREMASCHI, D., MEYER, G., BERMANO, S. & MARCATI, M. 1983. Different sodium chloride cotransport systems in the apical membrane of rabbit gallbladder epithelial cells. 3 Membr Bin1 73, 227-235. CREMASCHI, D., MEYER,G., ROSSETTI, C., BOTTA,G. & PALESTINI, P. 1987. The nature of the neutral Na+-Cl--coupled entry at the apical membrane of rabbit gallbladder epithelium. I. Na+/H+, C1 /HCO, double exchange and Na+-CI- symport. 3 Membr Biol95, 209-218. DIAMOND,J.M., EHRLICH,B.E., MORAWSKI, S.E., SANTA ANA,C.A. & FORDTRAN, J.S. 1983. Lithium absorption in tight and leaky segments of intestine. 3 Membr Biol 72, 153-159. EHRLICH,B.E. & DIAMOND,J.M. 1980. Lithium, membranes, and manic-depressive illness. 3 Membr Biol 52, 187-200. FREDERIKSEN, 0. 1978. Functional distinction between two transport mechanisms in rabbit gall-bladder epithelium by use of ouabain, ethacrynic acid and metabolic inhibitors. 3 Ph,ysiol (Lond)280, 373-387. FREDERIKSEN, 0. 1983. Effect of amiloride on sodium and water reabsorption in the rabbit gall-bladder. 3 Physiol (Lond) 335, 75-88. FREDERIKSEN, 0. 1985. Lack of involvement of cotransport and antiport mechanisms in rabbit gallbladder transepithelial isosmotic NaCl transport (abstract). Proceedings of the Vth European Colloquium on Renal Physiology, Frankfurt a.M., p. 145. FREDERIKSEN, 0. & LEYSSAC, P.P. 1969. Transcellular transport of isosmotic volumes by the rabbit gallbladder in vitro. 3 Physiol (Lond) 201, 201-224. 0. & LEYSSAC, P.P. 1984. Possible role FREDERIKSEN, of calcium in the control of gall-bladder fluid absorption. In: R.M. Case, J.M. Lindgard & J.A. Young (eds.) Secretion 1 Mechanisms and Control, pp. 207-21 1. Manchester University Press, ManChester. FREDERIKSEN, MBLLGKRD, K. & ROSTGAARD, J. 1979. Lack of correlation between transepithelial transport capacity and paracellular pathway ultrastructure in Alcian blue-treated rabbit gallbladders. 3 Cell Biol 83, 383-393. FRIZZELL, R.A., DUGAS,M.C. & SCHULTZ, S.G. 1975. Sodium chloride transport by rabbit gallbladder. Direct evidence for a coupled NaCl influx process. 3 Gen Physiol65, 769-795.

o.,

GRINSTEIN, s. & ROTHSTEIN, A. 1986. Mechanisms of regulation ofthe Na+/H+ exchanger. 3 Membr Rial 90, 1-12. GUNTHER, R.D. &WRIGHT, E.M. 1983. Na+, Li', and CI transport by brush border membranes from rabbit jejunum. 3 Membr Biol 74, 85-94. GUTMAN, Y., HOCHMAN, S. & WALD,H. 1973. The differential effect of Li' on microsomal ATPase in cortex, medulla and papilla of the rat kidney. Biochim Biophys Acta 298, 284290. GY~RY A.Z. , & LINGARD, J.M. 1976. Kinetics of active sodium transport in rat proximal tubules and its variation by cardiac glycosides at zero net volume and ion fluxes. Evidence for a multisite sodium transport system. 3 Physiol (Lond) 257, 257-274. HANSEN,C.P. & FREDERIKSEN, 0. 1987. Effects of lithium on sodium and water absorption in the rabbit gall-bladder (abstract). Acta Ph,ysiol Scand 129, 14A. IIANSEN,C. P. & FREDERIKSEN, 0. 1990. Calcium dependence of BAY K 8644 effects on the rabbit gall-bladder. PJugers Arch 406, 44-48, HANSEN,C.P., HOLSTEIN-RATHLOU, N.-H., SKBTT, O., LEYSSAC, P.P. & FREDERIKSEN, 0. 1989. Lithium absorption in the rabbit gall-bladder (abstract). Proc Int Union Pysiol Sci 17, 54. HAYSLETT, J.P. & KASHGARIAN, M. 1979. A micropuncture study of the renal handling of lithium. PJGgers Arch 380, 159-163. KINSELLA, J.L. & ARONSON,P.S. 1980. Properties of the Na+-H+ exchanger in renal microvillus membrane vesicles. Am 3 Ph.ysiol 238, F461LF469. 0. 1974. LEYSSAC, P.P., BUKHAVE, K. & FREDERIKSEN, Inhibitory effect of prostaglandins on isosmotic fluid transport by rabbit gall-bladder in vitro, and its modification by blockade of endogenous PGEbiosynthesis with indomethacin. Actu Physiol Scand 92, 496-507. LEYSSAC, P.P., HOLSTEIN-RATHLOU, N.-H., SKBTT, P. & ALFREY,A.C. 1990. A micropuncture study of proximal tubular transport of lithium during osmotic diuresis. A m J Physiol258, F109&F1095. MORENO,J.H. & DIAMOND, J.M. 1976. Cation permeation mechanism and cation selectivity in 'tight junctions' of gallbladder epithelium. In: G. Eisenman (ed.) Membranes - A series of Adcances, vol. 3, pp. 383497. Marcel Dekker, New York. MURER, H., HOPFER, U. & KINNE,R. 1976. Sodium/ proton antiport in brush border membrane vesicles isolated from rat small intestine and kidney. Biochem 3 154, 597-604. VAN Os, C.H. & SLEGERS, J.F.G. 1973. Path of osmotic water flow through rabbit gall bladder epithelium. Biochim Biophys Acta 291, 197-207. PETERS, C.J. & WALSER, M. 1966. Transport ofcations by rabbit gall-bladder : evidence suggesting a common cation pump. A m 3 Ph.ysiol210, 677-683.

Rabbit gall-bladder lithium absorption

195

SMULDERS, A.P., TORMEY, J.McD. & WRIGHT, E.M. WEINMAN, S. A. & R ~ u s sL. , 1982. Na+-H' exchange at the apical membrane of Necturus gallbladder. 1972. The effect of osmoticall) induced water flows Extracellular p H studies. JGen Ph,yszol80,299-32 1. on the permeability and ultrastructure of the rabbit WRIGHT,E M . , SMULDERS, A.P. & TORMEY, J.McD. gallbladder. 3 Membr Bad 7, 164197. THOMSEN, K., HOLSTEIN-KATHLOU, N.-13. & LLYSSAC, 1972. T h e role of the lateral intercellular spaces and solute polarization effects in the passive flow of P.P. 1981. Comparison of three measures of water across the rabbit gallbladder. 3Membr Bzol7, proximal tubular reabsorption : lithium clearance, 198-2 19. occlusion time, and micropuncture. Am 3 Ph,ysiol 241, F348-F355.