ogy, Part A 148 (2007) 64–71www.elsevier.com/locate/cbpa

Comparative Biochemistry and Physiol

Beta-adrenergic activation of solute coupled water uptake by toad skinepithelium results in near-isosmotic transport☆

Robert Nielsen ⁎, Erik Hviid Larsen

Institute of Molecular Biology and Physiology, University of Copenhagen, August Krogh Building, Universitetsparken 13, DK-2100 Copenhagen Ø, Denmark

Received 6 August 2006; received in revised form 14 December 2006; accepted 25 December 2006Available online 16 January 2007

Abstract

Transepithelial potential (VT), conductance (GT), and water flow (JV) were measured simultaneously with good time resolution (min) inisolated toad (Bufo bufo) skin epithelium with Ringer on both sides. Inside application of 5 μM isoproterenol resulted in the fast increase in GT

from 1.2±0.3 to 2.4±0.4 mS·cm−2 and slower increases in equivalent short circuit current, ISCEqv=−GT×VT, from 12.7±3.2 to 33.1±6.8 μA cm−2,

and JV from 0.72±0.17 to 3.01±0.49 nL cm−2 s−1. Amiloride in the outside solution abolished ISCEqv (− 1.6±0.1 μA cm−2) while JV decreased to

0.50±0.15 nL cm−2·s−1, which is significantly different from zero. Isoproterenol decreased the osmotic concentration of the transported fluid,Cosm≈2× ISC

Eqv/JV, from 351±72 to 227±28 mOsm (Ringer's solution: 252.8 mOsm). JV depicted a saturating function of [Na+]out in agreementwith Na+ self-inhibition of ENaC. Ouabain on the inside decreased ISC

Eqv from 60±10 to 6.1±1.7 μA cm−2, and JV from 3.34±0.47 to 1.40±0.24 nL cm−2·s- 1. Short-circuited preparations exhibited a linear relationship between short-circuit current and JV with a [Na+] of the transportedfluid of 130±24 mM ([Na+]Ringer's solution=117.4 mM). Addition of bumetanide to the inside solution reduced JV. Water was transported uphill andJV reversed at an excess outside osmotic concentration, ΔCS,rev=28.9±3.9 mOsm, amiloride decreased ΔCS,rev to 7.5±1.5 mOsm. It is concludedthat water uptake is accomplished by osmotic coupling in the lateral intercellular space (lis), and hypothesized that a small fraction of the Na+ fluxpumped into lis is recirculated via basolateral NKCC transporters.© 2007 Elsevier Inc. All rights reserved.

Keywords: Solute coupled water transport; Isosmotic transport; Hyposmotic transport; Uphill water transport; High resistance epithelium; Lateral intercellular space;Isoproterenol; Cutaneous water uptake; Anuran Amphibia

1. Introduction

Water uptake coupled to transport of NaCl in absence of atransepithelial osmotic concentration difference is associatedwith ‘leaky’ epithelia producing a near-isosmotic absorbate.These epithelia, e.g., vertebrate kidney proximal tubule (Wind-hager et al., 1959) gallbladder (Diamond, 1962, 1964), andsmall intestine (Curran and Solomon, 1957; Curran, 1960) sharea number of features like small transepithelial electrical po-tential difference (VT) associated with low-resistance tightjunctions (Frömter and Diamond, 1972), abundant expressionof Na+/K+-pumps in the plasma membrane lining the lateral

☆ This paper was presented in the session “Water transport” at the Society ofExperimental Biology's Annual Meeting at the University of Kent, Canterbury,UK April 2nd–7th 2006.⁎ Corresponding author. Tel.: +45 3532 1708; fax: +45 3532 1567.E-mail address: rnielsen@aki.ku.dk (R. Nielsen).

1095-6433/$ - see front matter © 2007 Elsevier Inc. All rights reserved.doi:10.1016/j.cbpa.2006.12.041

intercellular space (DiBona and Mills, 1979), and paracellularfluxes of plasma membrane impermeable molecules (e.g., Hilland Hill, 1978; Shachar Hill and Hill, 1993; Whittembury et al.,1980, 1988). Their general function is to translocate isosmoticfluid from one extracellular body compartment to another. Incontrasts, water transport by ‘tight’ epithelia is driven by atransepithelial osmotic concentration gradient stemming from adiluted external (luminal) solution, which results in a hypotonicabsorbate. These epithelia, e.g., distal epithelia of vertebratenephron and amphibian skin and bladder are characterized bylarge VT (Ussing and Zerahn, 1951; Leaf et al., 1958; Granthamet al., 1970; Lewis and Diamond, 1976) and neurophypophysealhormone regulated osmotic permeability (Ewer, 1952; Hays andLeaf, 1962; Bentley, 1969; Al-Zahid et al., 1977; Johnsen andNielsen, 1984; Harris et al., 1986; Nielsen et al., 1995).Generally, this class of epithelia serves osmoregulatory func-tions with water and ion transports regulated according to theextracellular volume and osmotic balance of the organism.

65R. Nielsen, E.H. Larsen / Comparative Biochemistry and Physiology, Part A 148 (2007) 64–71

Amphibian skin constitutes a classical example of a tightepithelium with hormone controlled water and ion transport(reviews: Larsen, 1991; Jørgensen, 1997; Uchiyama and Konno,2006). It has been known for long that anuran skin has capacityto absorb water when bathed in solutions on the two sides ofsimilar composition (Reid, 1892). This fluid uptake is stimulatedby vasopressin and it depends on the active transport of Na+

(Nielsen, 1997). Cutaneous water uptake in vivo not drivenprimarily by an osmotic gradient was more recently studied indehydrated Bufo marinus (Hillyard and Larsen, 2001). It wasestimated that the absorbate was near-isosmotic and shown thatthe rate of water uptake is significantly decreased by the sodiumchannel blocker amiloride in the external bath. In the terrestrialEuropean toad, Bufo bufo, dehydration as well as injection of theβ-adrenergic agonist isoproterenol increases the cutaneouscirculation as indicated by an increased blood cell fluxmonitored by laser Doppler flow-cytometry (Viborg andRosenkilde, 2004). In dehydrated toads and in hydrated toadstreated with a β-receptor agonist an increased cutaneous bloodcell flux was paralleled by enhanced water uptake from tapwater.

The purpose of the present study was to investigate wheth-er β-adrenergic receptors of the skin regulate the epithelialwater uptake per se and to analyze in more details the mech-anism of Na+-dependent cutaneous water transport underthese conditions.

2. Material and methods

2.1. Preparation

European toads (B. bufo) fed with mealworms ad lib werekept in an outdoor or an indoor (winter) terrarium with freeaccess to a pool of tap water. The toads were killed by doublepithing and the skin removed by dissection. The epithelium wasisolated by 2 h serosal exposure of the skin at room temperatureto a Ringer's solution containing 2 mg/ml collagenase.

2.2. Monitoring of water flow and measurements of electro-physiological parameters

The preparation was mounted in a modified Ussing chamberwith an exposed area of 1.5 cm2 and supported against astainless steel net with a pressure of 2 cm H2O. Spontaneoustransepithelial potential (VT) or short circuit current (ISC), con-ductance (GT), and water flow (JV) were measured simulta-neously. The water flow was measured as describes previously(Johnsen and Nielsen, 1982). Briefly, one of the chamber halvesis closed except for an outlet consisting of a capillary tube, thismeans that when a drug is added to the closed chamber half themeasurement of the water flow is suspended for a period.During the experiment the solution is allowed to flow into thecapillary tube, whereby the light transmission of the tubechanges markedly. The transmission is detected by a light-emitting diode and a photosensitive transistor. The signal fromthe detector controls a motor-driven syringe which appropri-ately adjusts the water volume in order to keep the position of

the meniscus constant. The motor drives also a precisionpotentiometer which is used as potential divider. The position ofthe syringe is recorded by feeding the DC-signal from thepotential divider into a pen recorder and to the analogue input ofa computer controlled AD-converter. For passing currentsthrough the epithelium Ag/AgCl electrodes were mountedeither in the chamber or outside the chamber. When the currentelectrodes were mounted outside the chamber the current waspassed through salt bridges with a 3% agarose gel made in thesame solution as that present in the chamber. We used theAgarose type VIII with very low sulphate content in order toprevent water movement across the current bridges caused byelectro-osmosis (Nielsen, 1995).

Experiments were carried out under either open- or short-circuit conditions by a means of a computer controlled voltage-current clamp amplifier (VCC600, Physiologic Instruments SanDiego). Under open circuit conditions the conductance wasestimated by passing a current pulse (±10 μA) across the epi-thelium and by measuring the concomitant change in thetransepithelial potential. Under short-circuit conditions theconductance was estimated by stepping the clamp voltage(±10 mV) and recording the transepithelial current response.Under open circuit conditions the equivalent short-circuit cur-rent (I SC

Eqv) was calculated from the open circuit transepithelialpotential and the measured transepithelial conductance accord-ing to: I SC

Eqv =−GT×VT.

2.2.1. Sign conventionsVT is indicated relative to the electrical potential of the

internal solution (VT=ψoutside bath−ψinside bath), and inward

currents and inward volume flows are given a positive sign.

2.3. Solutions and chemicals

The Ringer's solution had the following composition (inmM): 110 NaCl, 2.4 NaHCO3, 2.0 KCl, 2.0 MgCl2, 1.0 CaCl2,5 Na-acetate, 5 glucose and pH 8.2 with a nominal osmoticconcentration of 252.8 mOsm. The osmolality of the solutionswas controlled by a VAPRO-5520 vapour pressure osmometer.Collagenase-Type 2 was fromWorthington Chemical (NJ, USA)and isoproterenol, bumetanide, ouabain, amiloride and Agarose-Type VIII were obtained from Sigma-Aldrich Co (St. Louis, MO,USA).

3. Results

3.1. Relationship between transport of NaCl and water flow

In the experiments presented below the epithelia weremounted in an Ussing chamber and bathed in isosmotic Ringer'ssolution, the epithelia were mounted either with the inside or theoutside of the epithelium side facing the open half of thechamber. Addition of the β-agonist isoproterenol (5 μM, inside)resulted in an almost immediate significant increase in thetransepithelial conductance, GT, followed by a slower significantincrease in the transepithelial water uptake and equivalent short-circuit current, I SC

Eqv (Fig. 1, Table 1). The conductance activation

Fig. 2. Effect of amiloride (0.1 mM) applied to the outside on equivalent short-circuit current (I SC

Eqv/F), and water uptake (JV) across epithelia incubated in thepresence of isoproterenol (5 μM, basolateral). The experiment was performedunder open circuit conditions, with the outside of the epithelium facing the openchamber half of the Ussing chamber.

Fig. 1. Effect of the β-agonist isoproterenol applied on the inside followed byamiloride applied on the outside on transepithelial potential difference (VT),conductance (GT), active Na+ flux estimated from calculated equivalent short-circuit current (I SC

Eqv/F), and water uptake (JV). The experiment was performedunder open circuit conditions, with the inside of the epithelium facing the openchamber half of the Ussing chamber. Note the relatively fast isoproterenol effecton VT and GT reflecting rapid response of the chloride conductance ofmitochondria rich cells. The stimulation of I SC

Eqv/F and JV proceeds with slowerbut almost similar time courses.

66 R. Nielsen, E.H. Larsen / Comparative Biochemistry and Physiology, Part A 148 (2007) 64–71

by isoproterenol was abolished by a subsequent addition ofthe β-receptor antagonist propranolol (GT in mS cm−2: con-trol, 0.41±0.04; 5 μM isoproterenol, 0.86±0.13; 5 μMisoproterenol+10 μM propranolol, 0.33±0.04, n=8, p=0.0002for propranolol inhibition of GT). In Fig. 1 it is also seen thataddition of the Na+-channel blocker amiloride (0.1 mM outside,closed chamber half, Nielsen, 1982) resulted in a highlysignificant reduction of these parameters; since amiloride wasadded to the closed chamber half the measurement of JV wassuspended for a period. Addition of amiloride (0.1 mM outside)to an isolated epithelium which mounted in the chamber with theoutside (apical) side facing the open chamber half resulted in aprompt decrease in I SC

Eqv and in somewhat slower decrease in JV(Fig. 2). This indicate that JV depends on the active transport ofNa+ but also that the coupling between salt transport and wateruptake is not direct (electro-osmosis or friction) but rather is dueto local osmosis, which would explain the delayed amiloride

Table 1Effect of isoproterenol, and amiloride on bioelectrical parameters and water flowacross isolated toad skin epithelia

VT

(mV)GT

(mS cm−2)I SCEqv

(μA cm−2)JV(nL cm−2·s-1)

Control −11±2 1.2±0.3 12.7±3.2 0.72±0.17p (Control–Iso) NS 0.03 0.004 0.0008Isoproterenol −13±1 2.4±0.4 33.1±6.8 3.01±0.49p (Iso–Amil) b0.0001 NS 0.004 0.004Amiloride −1.5±0.7 1.5±0.2 1.5±0.8 0.50±0.15

The data are the mean±SE (n=7). Experiments were conducted under open-circuit conditions, the epithelia were mounted with the inside facing the openhalf of the Ussing chamber. Isoproterenol (5 μM) was added to the inside of theepithelium and amiloride (0.1 mM) was added to the outside, as indicated onFig. 1. p (Control–Iso) indicates the p-value for the paired t-test between controland the isoproterenol data, and p (Iso–Amil) indicates the p-value between theisoproterenol and amiloride data.

inhibition of water flow relative to the fast inhibition of the activesodium flux (Nielsen, 1997).

In order to get an estimate of the osmotic concentration of thetransported fluid as a first approximation the rate of NaCluptake under open circuit conditions was estimated from I SC

Eqv

(see Discussion) which would correspond to a flux of Na+ andCl− of, JNa+JCl≈2× I SC

Eqv/F, with F being the Faraday. Thecalculated osmotic concentration of the transported fluid in thepresence of isoproterenol was 227±28 mOsm (n=7), which isnot significantly different from the osmotic concentration of theexternal baths (252.8 mOsm).1 Thus, as a result of isoproterenolstimulation of ion and water fluxes the transported fluid tendson average to become iso-osmotic. More details about thisseries of experiments are listed in Table 2 showing the range ofosmotic concentrations obtained after isoproterenol stimulation.It is noteworthy that three out of the seven preparations produceda hypotonic fluid with the following osmotic concentrations 208,109, and 179 mOsm, respectively. Using the same calculationfor estimating the osmotic concentration after amiloride wearrive at 62±12 mOsm, that is, a strongly hyposmotic absorbate.

3.2. The transepithelial water flow is rate limited by the activeflux of sodium

The data presented above indicate that the uptake of water iscoupled to the active transport of Na+. For substantiating thisconclusion a series of experiments were performed withisoproterenol stimulated preparations where the effect of theNa+/K+ -pump inhibitor ouabain was examined. From the datapresented (Fig. 3) it is seen that the significant ouabain-inhibition of VT and I SC

Eqv brings about inhibition of JV as well.

1 These and all other calculations of osmotic concentrations disregard anosmotic coefficient of 0.9 for the Ringer's solution and solutions of similar ionconcentrations. As long as comparisons are made between nominal osmoticconcentrations of solutions of similar strength this has no influence onconclusions regarding the question of isosmotic transport.

Fig. 4. Volume flow (JV) as function of outside Na+-concentration. The experiments

were carried out under open circuit conditions in the presence of 5 μM of the β-agonist isoproterenol on the inside. The full line is the best fit to, JV=JV

max×[Na+]out/(KNa+[Na

+]out), showing that fluid uptake is governed by the Na+-permeability ofthe apical membrane, which obeys a similar [Na+]out-dependence. Note, however,JVN0 for [Na

+]out=0. Values are the mean±SE (n=6).

Table 2Estimated osmotic concentration of the transported fluid of individual epithelialpreparations in the presence of isoproterenol on the inside

Exp. # I SCEqv/F (pmol cm−2 s−1) JV (nL cm−2·s-1) Osmotic concentration

(mOsm)

1 416.6 3.18 262.32 578.3 4.44 260.73 556.5 4.49 248.24 373.9 3.60 207.65 135.8 2.49 109.26 183.4 1.14 321.87 155.5 1.74 178.9

Data from seven individual experiments. Same protocol as in Fig. 1. Theosmotic concentration was estimated as, 2× I SC

Eqv/JV, which should be comparedto the nominal osmotic concentration of the Ringer's solution of 252.8 mOsm.

67R. Nielsen, E.H. Larsen / Comparative Biochemistry and Physiology, Part A 148 (2007) 64–71

The initial relatively fast decrease in I SCEqv and the subsequent

slower current inhibition are paralleled by concomitant de-creases in JV of fast and slow time course, respectively. In sixisoproterenol stimulated preparations ouabain decreased I SC

Eqv

from 60±10 to 6.1±1.7 μA cm−2 (p=0.0014, n=6), while JVdecreased from 3.34±0.47 to 1.40±0.24 nL cm− 2 s− 1

(p=0.0004, n=6).The active transepithelial Na+ uptake is controlled at the

level of the apical Na+ permeability of principal cells which isgoverned by density and open/close activity of the epithelialsodium channel, ENaC (Palmer and Garty, 1997). The apicalNa+ permeability decreases with increasing external sodiumconcentration. This [Na+]out-dependent closure of sodiumchannels results in saturation kinetics of the macroscopic influxof Na+ across the epithelium (Fuchs et al., 1977). The rela-tionship between the uptake of water and the Na+ concentrationin the outside bathing solution is shown in Fig. 4. The ex-periments were carried out in the presence of isoproterenol(5 μM) with [Na+]out being varied from 0 to 112 mM. ExternalNaCl was replaced by an isosmotic concentration of sucrose inorder to maintain similar osmolarity of outside and inside

Fig. 3. Effect of the β-agonist isoproterenol followed by ouabain both applied onthe inside on transepithelial potential difference (VT), conductance (GT), activeNa+ flux estimated from calculated equivalent short-circuit current (I SC

Eqv/F), andwater uptake (JV). The experiment was performed under open circuit conditions,with the inside of the epithelium facing the open chamber half of the Ussingchamber. Note similar time courses of inhibition of I SC

Eqv/F and JV, respectively.

solutions. It is seen that the fluid uptake depicts a saturatingfunction of [Na+]out in a way similar to the above mentionedrelationship between [Na+]o and the transepithelial active Na+

flux. This finding provides further support for the hypothesisthat water uptake under transepithelial osmotic equilibriumconditions is rate limited by the active sodium uptake. Never-theless, for [Na]out=0 mM, JV=0.248±0.060 nL cm−2 s−1

which is significantly different from zero (pb0.01, n=6). Thisobservation will be dealt with in Discussion.

3.3. Coupling between active Na+ flux and transepithelialwater flow under short-circuit conditions

To study more directly the relationship between fluidtransport and the active Na+ flux in isoproterenol stimulatedpreparations VT was clamped to zero. In the short-circuitedpreparation the transepithelial current, ISC, is carried by Na

+. Asan example, Fig. 5 shows parallel and sustained stimulation of

Fig. 5. Effect of the β-agonist isoproterenol (5 μM) followed by bumetanide(0.2 mM) both added to the inside bath on conductance (GT) and active Na

+ fluxcalculated from the short-circuit current, JNa= ISC/F, and water uptake (JV). Theexperiment was performed under short-circuited conditions with the inside ofthe epithelium facing the open chamber half.

Fig. 6. Volume flow (JV) in the presence of the β-agonist isoproterenol (5 μMonthe inside) as a function of the active sodium flux calculated from recorded short-circuit current in continuously short-circuited preparation, JNa= ISC/F. The slopeof the regression line corresponds to a sodium concentration of 130±24 mM ofthe transported fluid. Data obtained from 12 preparations.

Fig. 7. Effect of an increase in the osmolarity of the outside solution (addition ofsucrose) on transepithelial water flow (JV, full line and right hand axis) in theabsence and presence of 100 μM amiloride. The active Na+ flux was estimatedfrom calculated equivalent short-circuit current and given as I SC

Eqv/F (blacksymbols and left hand axis). The preparation was incubated in the presence of5 μM isoproterenol on the inside under open circuit conditions, with the outsideof the epithelium facing the open chamber half. The sucrose concentration of theapical bathing solution is given by the numbers in the open bars.

68 R. Nielsen, E.H. Larsen / Comparative Biochemistry and Physiology, Part A 148 (2007) 64–71

both ISC and JV after addition of the agonist at a concentrationof 5 μM to the inside bath. In Fig. 6 the transepithelial waterflow in 12 different isoproterenol stimulated preparations isplotted as a function of the active inward Na+ flux calculated foreach preparation as, JNa= ISC/F. The slope of the regression lineis 0.0077±0.0014 nL pmol−1 Na+, which corresponds to aconcentration of 130±24 mM Na+, or to 427±78 water mole-cules being translocated per sodium ion actively transported.These numbers are significantly different from those obtainedwith AVT-stimulated skin of the frog Rana esculenta, 347±64 mM Na+ and 160±15 H2O/Na

+, respectively (Nielsen,1997). The above average concentration of Na+ of the fluidabsorbed by the toad preparation, 130±24 mM, is not sig-nificantly different from the Na+ concentration of the externalbaths, [Na+]Ringer=117.4 mM.

3.4. Effect of bumetanide

Addition of the co-transport inhibitor bumetanide to theshort-circuited isoproterenol stimulated preparation resulted in ahighly significant decrease in the transepithelial conductance(p=0.006, n=8) (Fig. 6 and Table 3). The observed significantreduction in GT indicates that bumetanide inhibits the transe-pithelial conductive permeability for Cl−, which is localized tothe mitochondria-rich cells (Willumsen et al., 2002; Nagel et al.,2002). Such an increase in the shunt resistance will have major

Table 3Effect of bumetanide on bioelectrical parameters and water flow in isolated toadskin epithelium

JV (nL cm−2 s−1) ISC (μA cm−2) GT (mS cm−2)

Control 1.42±0.40 27.3±7.6 1.35±0.29Isoproterenol 5.13±0.55 59.7±6.3 3.20±0.55Bumetanide 3.77±0.43 44.3±5.8 1.84±0.27Amiloride 0.47±0.28 −0.3±2.0 1.84±0.27

The data are the mean±SE (n=8). The experiments were conducted under short-circuited conditions according to the protocol indicated in Fig. 4. Isoproterenol(5 μM) and bumetanide (0.2 mM) were added to the inside, and amiloride(0.1 mM) was added to the outside of the epithelium.

effect on the intraepithelial loop-current, that is, the NaCl uptakeunder open circuit conditions will be significantly decreased (seeDiscussion). Therefore, in order to avoid interference fromchanges in the Cl− permeability these experiments were per-formed under short-circuit conditions. Additional bumetanide(0.1 mM) to the inside solution of isoproterenol stimulated short-circuited preparations resulted in a highly significant decrease inJV (p=0.006, n=8) and in a significant decrease in ISC (p=0.03,n=8) (Table 3). The transepithelial uptake of water (JV) inthe presence of both bumetanide and amiloride (0.47±0.28 nLcm−2 s−1) is not significantly different from 0 (p=0.13, n=8).

3.5. Local osmosis

The data presented above indicate that the transport of NaCl,and Na+, across isolated toad skin epithelium results in thebuilding up of an osmotic concentration difference between theoutside bath and an intraepithelial compartment in whichosmotic coupling of salt and water transport takes place. This

Fig. 8. Toad skin epithelium transports water uphill. The fluid uptake (JV)is depicted as a function of the outside concentration of sucrose([sucrose]out). The straight line is the linear fit to the experimentaldata points from which the osmotic concentration at which the water flowreverses direction, ΔCS,rev=34.0 mM, is obtained.

Table 4Hydrosmotic parameters of isolated toad skin epithelium

n=5 JV(nL cm−2 s−1)

ΔCS,rev

(mOsm)Slope(nL·cm−2 s−1 mOsm−1)

Isoproterenol 3.42±0.15 28.9±3.9 0.116±0.012Amiloride 0.95±0.12 7.5±1.5 0.125±0.023

The protocol of the experiments is shown in Fig. 8, where the volume flow, JV, isdepicted as a function of the concentration of a non-permeable osmolyte addedto the solution bathing the outside of the epithelium, [sucrose]out. ΔCS,rev is theintercept between the regression line and the abscissa of Fig. 8. Mean±SE forfive preparations.

69R. Nielsen, E.H. Larsen / Comparative Biochemistry and Physiology, Part A 148 (2007) 64–71

would be in agreement with the idea developed in studies ofwater absorption by small intestine (Curran, 1960; Curran andMacintosh, 1962). To investigate how large an osmotic gradientthe epithelium is able to produce, experiments were performedwhere the osmolarity of the apical bathing solution was in-creased until the transepithelial water movement was abolishedor reversed. The experiments were carried out on isolatedepithelia under open circuit conditions in the presence ofisoproterenol, and the osmolarity of the outside bath wasincreased by the addition of an appropriate volume of a Ringer'ssolution containing 1000 mOsm sucrose. Fig. 7 shows that bystepwise increasing the external sucrose concentration the fluiduptake was decreased in a similar stepwise fashion withoutaffecting I SC

Eqv. The excess osmotic concentration at which thevolume flow reverses, denoted ΔCS,rev, was calculated as theintercept between the regression line and the abscissa (Fig. 8,Table 4). The addition of amiloride reduced significantly theosmotic concentration at which the volume flow reversed(p=0.002, n=5). This result confirms that the amiloridesensitive active Na+ transport builds up the osmotic concentra-tion difference between the external solution and the intrae-pithelial coupling compartment. With the Na+/K+-pumpslocated in the plasma membranes lining the lateral intercellularspace of amphibian skin epithelium (Mills et al., 1977) weconclude that the lateral intercellular space constitutes thecoupling compartment.

4. Discussion

4.1. Triple effects on epithelial permeabilities of β-adrenergicreceptor activation

The present study has shown that stimulation of β-adrenergicreceptors leads to the activation of the transepithelial waterpermeability as well as the apical sodium permeability ofprincipal cells (Figs. 1 and 2). Previous studies have indicatedthat β-adrenergic receptors also couple to cAMP activatedCFTR-like chloride channels in the apical plasma membrane ofmitochondria-rich cells (Willumsen et al., 1992, 2002). Bysimultaneous recording with good time resolution of the waterflow and the ion permeability it was observed that the activationof the chloride permeability is faster than the activations of theactive sodium flux and fluid uptake (Fig. 1). The water channelprotein activated by the β-adrenergic receptor agonist is mostlikely the AQP2/3 water channels which were cloned from the

skin epithelium of the tree frog Hyla japonica and shown to betranslocated to the apical plasma membrane of the principalcells upon arginine vasotocin (AVT) stimulation (Hasegawaet al., 2003). With primers generated from B. marinus AQP1-4 cDNA and subsequent PCR-amplification Willumsen et al.(2007—this issue) identified expression of these aquaporins inthe skin of B. bufo. Similar to the β-adrenergic receptors, the V2receptor-controlled water pathway is involved in Na+-coupledfluid uptake by the skin of R. esculenta (Nielsen, 1997).

In the comparative physiological context, an interestingfinding of the present study is the above triple effect of β-adrenergic receptor stimulation which together with the pre-viously described isoproterenol activation of the skin circulation(Viborg and Rosenkilde, 2004) points to a physiologicallysignificant role of this receptor type for the regulation of wateruptake by terrestrial and semi-terrestrial anuran Amphibia.Increased skin circulation in vivo was not seen with AVT injec-tion which, however, stimulated the cutaneous water uptake byB. bufo (Viborg and Rosenkilde, 2004). Thus, the relationshipbetween these two receptors and their specific functions inmaintaining anuran water balance remain as open questions.

4.2. Solute coupled fluid uptake by toad skin epithelium

The osmotic concentration of the transported fluid wasestimated by calculating the equivalent short-circuit current andassuming that the sum of Na+ and Cl− fluxes under open circuitconditions is about twice this bioelectric quantity, i.e., JNa+JCl≈2× ISC

Eqv/F. This method provides a conservative estimate ofthe osmotic concentration of the absorbed fluid with the error ofthe estimate depending on the transepithelial shunt resistance.For example, with Na+ being the only actively transported ionthe net uptake of NaCl at spontaneous VT would be zero if theshunt resistance tends to infinity. This, of course, is not soneither for the short circuit current nor for the equivalent shortcircuit current. It means, however, that our calculation of theosmotic concentration of the transported fluid is overestimatedunless the transepithelial shunt resistance is zero correspondingto infinite chloride conductance of mitochondria-rich cells.With a finite shunt resistance, therefore, the true osmoticconcentrations of the transported fluid are less than the numbersgiven above, especially under control conditions.

Thus, we can conclude that another interesting finding of thepresent study is that the hormone stimulated toad skin epithe-lium has a capacity to produce isosmotic and hyposmoticabsorbates under conditions of transepithelial osmotic equilib-rium (Table 1). Our experiments further demonstrated that therate of fluid uptake depends on the active sodium flux throughNa+/K+-pumps (Figs. 3 and 8), and that fluid is being absorbedeven against an adverse osmotic concentration gradient (Fig. 8).Taken together, our study has shown that this ‘high resistance’epithelium shares a physiologically significant function with‘low resistance’ epithelia specialized for constitutive isosmoticfluid transport by osmotic coupling of water uptake to the activeuptake of Na+ (Windhager et al., 1959; Curran, 1960; Diamond,1964; Parsons and Wingate, 1958; Green et al., 1991;Weinstein, 1992).

70 R. Nielsen, E.H. Larsen / Comparative Biochemistry and Physiology, Part A 148 (2007) 64–71

4.3. Water uptake in absence of transepithelial salt uptake

It is a puzzling finding and a challenge to the above osmoticcoupling theory that a stationary (small) water uptake wasmaintained with sodium channels blocked by amiloride (Table 1,Fig. 2), or with [Na+]out=0 mM (Fig. 4). We will discuss thisreproducible observation together with isosmotic and hyposmo-tic transports and the effect of bumetanide on the transepithelialfluid uptake. Principal cells of amphibian skin epitheliumdisplay a NKCC transporter in the basolateral membrane, whichtogether with basolateral Cl channels is of significance for cellvolume regulation and cellular accumulation of Cl- (Ussing,1982, 1985; Dörge et al., 1985). If a fraction of the lateral Na+/K+-pump flux is associated with returning of Na+ from the insidesolution via the NKCC transporter we would predict that serosalapplication of bumetanide reduces the osmotic concentration ofthe lateral space and thereby the uptake of water from the outsidesolution, which was observed in the present study (Fig. 5 andassociated text). With this set of transporters the lateral Na+/K+-pumps would build up a hypertonic lateral intercellular space byan electroneutral recirculation of ions across the basolateralmembrane even in the absence of apical Na+ inflow. Thus, thishypothesis would also explain fluid uptake in absence oftransepithelial ion uptake (Figs. 1, 2 and 4, Table 1). Further-more, similar recirculations of ions can give rise to isosmotic aswell as hyposmotic fluid uptake (Table 2) by returning ions to thelateral intercellular space and thereby compensating for the factthat the fluid emerging from the lateral intercellular spacethrough the interspace basement membrane is hyperosmotic. Inthe accompanying paper (Larsen et al., 2007-this issue) this Na+

recirculation theory is discussed in more details and analyzedquantitatively.

Acknowledgements

Technical assistance in the laboratory by Mrs. PernilleRoshof is greatly appreciated. The project is supported by framegrant 272-05-0417 from the Danish Natural Science ResearchCouncil.

References

Al-Zahid, G., Schafer, J.A., Troutman, S.L., Andreoli, T.E., 1977. The effect ofantidiuretic hormone on water and solute permeation, and activationenergies for these processes, in mammalian cortical collecting tubules:evidence for parallel ADH sensitive pathways for water and solute diffusionin luminal plasma membranes. J. Membr. Biol. 3, 103–129.

Bentley, P.J., 1969. Neurohypophyseal hormones in amphibia: a comparison oftheir actions and storage. Gen. Comp. Endocrinol. 13, 39–44.

Curran, P.F., 1960. Na, Cl, and water transport by rat ileum in vitro. J. Gen.Physiol. 43, 1137–1148.

Curran, P.F., Macintosh, J.R., 1962. A model system for biological watertransport. Nature 193, 347–348.

Curran, P.F., Solomon, A.K., 1957. Ion and water fluxes in the ileum of rats.J. Gen. Physiol. 41, 143–168.

Diamond, J.M., 1962. The absorptive function of the gall-bladder. J. Physiol.161, 442–473.

Diamond, J.M., 1964. The mechanism of isotonic water transport. J. Gen.Physiol. 48, 15–42.

DiBona, D.R., Mills, J.W., 1979. Distribution of Na-pump sites in transportingepithelia. Fed. Proc. 38, 134–143.

Dörge, A., Rick, R., Beck, F., Thurau, K., 1985. Cl transport across the basolateralmembrane in frog skin epithelium. Pflügers Arch. 495 (Suppl. 1), S8–S11.

Ewer, R.F., 1952. The effect of pituitrin in fluid distribution in Bufo regularisReus. J. Exp. Biol. 29, 173.

Frömter, E., Diamond, J.M., 1972. Route of passive ion permeation in epithelia.Nature 235, 9–13.

Fuchs, W., Larsen, E.H., Lindemann, B., 1977. Current-voltage curve of sodiumchannels and concentration dependence of sodium permeability in frog skin.J. Physiol. 267, 137–166.

Grantham, J.J., Burg, M.B., Orloff, J., 1970. The nature of transtubular Na and Ktransport in isolated rabbit renal collecting tubule. J. Clin. Invest. 49,1815–1826.

Green, R., Giebisch, G., Unwin, R., Weinstein, A.M., 1991. Coupled watertransport by rat proximal tubule. Am. J. Physiol. 261, F1046–F1054.

Harris, H.W., Wade, J.B., Handler, J.S., 1986. Transepithelial water flowregulates apical membrane retrieval in antidiuretic hormone-stimulated toadurinary bladder. J. Clin. Invest. 78, 703–712.

Hasegawa, T., Tanii, H., Suzuki, M., Tanaka, S., 2003. Regulation of waterabsorption in the frog skins by two vasotocin-dependent water-channelaquaporins, AQP-h2 and AQP-h3. Endocrinology 144, 4087–4096.

Hays, R.M., Leaf, A., 1962. Studies of the movement of water through theisolated toad bladder and its modification by vasopressin. J. Gen. Physiol.45, 905.

Hill, A.E., Hill, B.S., 1978. Sucrose fluxes and junctional water flow acrossNecturus gall bladder epithelium. Proc. R. Soc. Lond., B 200, 163–174.

Hillyard, S.D, Larsen, E.H., 2001. Lymph osmolality and rehydration fromNaClsolutions by toads, Bufo marinus. J. Comp. Physiol. B 171, 283–292.

Johnsen, A.H., Nielsen, R., 1982. Enhanced sensitivity to stimulation of sodiumtransport and cyclic AMP by antidiuretic hormone after Ca2+ depletion ofisolated frog skin epithelium. J. Membr. Biol. 69, 137–143.

Johnsen, A.H., Nielsen, R., 1984. Correlation between cAMP in isolated frogskin epithelium and stimulation of sodium transport and osmotic water flowby antidiuretic hormone and phosphodiesterase inhibitors. Gen. Comp.Endocrinol. 64, 144–153.

Jørgensen, C.B., 1997. 200 years of amphibian water economy: from RobertTownson to the present. Biol. Rev. 72, 153–237.

Larsen, E.H., 1991. Chloride transport by high-resistance heterocellularepithelia. Physiol. Rev. 71, 235–283.

Larsen, E.H., Møbjerg, N., Nielsen, R., 2007. Application of the Na+

recirculation theory to ion coupled water transport in low- and highresistance osmoregulatory epithelia. Comp. Biochem. Physiol. A 148,101–116 (this issue). doi:10.1016/j.cbpa.2006.12.039.

Leaf, A., Anderson, J., Page, L.B., 1958. Active sodium transport bu isolatedtoad bladder. J. Gen. Physiol. 41, 657–668.

Lewis, S.A., Diamond, J.M., 1976. Na+ transport by rabbit urinary bladder, atight epithelium. J. Membr. Biol. 28, 1–40.

Mills, J.W., Ernst, S.A., DiBona, D.R., 1977. Localization of Na+-pump sites infrog skin. J. Cell Biol. 75, 968–973.

Nagel, W., Somieski, P., Katz, U., 2002. The route of chloride movement acrossamphibian skin: localization and regulatory mechanisms. Biochim. Biophys.Acta 1566, 44–54.

Nielsen, R., 1982. Effect of amiloride, ouabain and Ba++ on nonsteady-state Na–K pump flux and short-circuit current in isolated frog skin epithelia.J. Membr. Biol. 65, 227–234.

Nielsen, R., 1995. Correlation between chloride flux via the mitochondria-richcells and transepithelial water movement in isolated frog sin (Ranaesculenta). Acta Physiol. Scand. 155, 351–361.

Nielsen, R., 1997. Correlation between transepithelial Na+ transport andtransepithelial water movement across isolated frog skin (Rana esculenta).J. Membr. Biol. 159, 61–69.

Nielsen, S., Chou, C.-L., Marples, D., Christensen, E.I., Kishore, B.M.,Knepper, M., 1995. Vasopressin increases water permeability of kidneycollecting duct by inducing translocation of aquaporin-CD water channels toplasma membrane. Proc. Natl. Acad. Sci. U. S. A. 92, 1013–1017.

Palmer, L.G., Garty, H., 1997. Epithelial sodium channels: function, structureand regulation. Physiol. Rev. 77, 359–396.

71R. Nielsen, E.H. Larsen / Comparative Biochemistry and Physiology, Part A 148 (2007) 64–71

Parsons, D.S., Wingate, D.L., 1958. Fluid movements across wall of rat smallintestine. Biochim. Biophys. Acta 30, 666–667.

Reid, E.W., 1892. Report on experiments upon ‘absorption without osmosis’.Br. Med. J. 1, 323–326.

Shachar Hill, B., Hill, A.V., 1993. Convective fluid flow through the paracellularsystem of Necturus gall-bladder epithelum as revealed by dextran probes.J. Physiol. 468, 463–486.

Uchiyama, M., Konno, N., 2006. Hormonal regulation of ion and watertransport in anuran amphibians. Gen. Comp. Endocrinol. 147, 54–61.

Ussing, H.H., 1982. Volume regulation of frog skin epithelium. Acta Physiol.Scand. 114, 363–369.

Ussing, H.H., 1985. Volume regulation and basolateral co-transport of sodium,potassium, and chloride ions in frog skin epithelium. Pflügers Arch. 405(Suppl 1), S2–S7.

Ussing, H.H., Zerahn, K., 1951. Active transport of sodium as the source ofelectric current in the short-circuited isolated frog skin. Acta Physiol. Scand.23, 110–127.

Viborg, A.L., Rosenkilde, P., 2004. Water potential receptors in the skin regulateblood perfusion in ventral pelvis patch of toads. Physiol. Biochem. Zool. 77,39–49.

Weinstein, A.M., 1992. Sodium and chloride transport, In: Seldin, D.W.,Giebisch, G. (Eds.), The Kidney, second ed. Physiology and Pathophysi-ology, vol. 2. Raven Press, New York, pp. 1925–1973.

Whittembury, G., De Martínez, C.D., Linares, H., Paz-Aliaga, A., 1980. Solventdrag of large solutes indicates paracellular water flow in leaky epithelia.Proc. R. Soc. Lond., B 211, 63–81.

Whittembury, G., Malnic, G., Mello-Aires, M., Amorena, C., 1988. Solvent dragof sucrose during absorption indicates paracellular water flow in the ratkidney proximal tubule. Pflügers Arch. 412, 541–547.

Willumsen, N.J., Vestergaard, L., Larsen, E.H., 1992. Cyclic AMP- andβ-agonist-activated chloride conductance of a toad skin epithelium. J. Physiol. 449,641–653.

Willumsen, N.J., Amstrup, J., Møbjerg, N., Jesperen, Å., Kristensen, P., Larsen,E.H., 2002. Mitochondria-rich cells as experimental model in studies ofepithelial chloride channels. Biochim. Biophys. Acta 1566, 28–43.

Willumsen, N.J., Viborg, A., Hillyard, S., 2007. Vascular and molecular aspectsof water uptake mechanisms in the toad skin. Comp. Biochem. Physiol. A148, 55–63 (this issue). doi:10.1016/j.cbpa.2006.12.032.

Windhager, E.E., Whittembury, G., Oken, D.E., Schatzmann, H.J., Solomon,A.K., 1959. Single proximal tubules of the Necturus kidney: III. Dependenceof H2O movement on NaCl concentration. Am. J. Physiol. 197, 313–318.