Proteinase-activated Receptor 2 Stimulates Na,K-ATPase and Sodium Reabsorption in Native Kidney Epithelium*

Proteinase-activated receptors 2 (PAR2) are expressed in kidney, but their function is mostly unknown. Since PAR2 control ion transport in several epithelia, we searched for an effect on sodium transport in the cortical thick ascending limb of Henle's loop, a nephron segment that avidly reabsorbs NaCl, and for its signaling. Activation of PAR2, by either trypsin or a specific agonist peptide, increased the maximal activity of Na,K-ATPase, its apparent affinity for sodium, the sodium permeability of the paracellular pathway, and the lumen-positive transepithelial voltage, featuring increased NaCl reabsorption. PAR2 activation induced calcium signaling and phosphorylation of ERK1,2. PAR2-induced stimulation of Na,K-ATPase Vmax was fully prevented by inhibition of phospholipase C, of changes in intracellular concentration of calcium, of classical protein kinases C, and of ERK1,2 phosphorylation. PAR2-induced increase in paracellular sodium permeability was mediated by the same signaling cascade. In contrast, increase in the apparent affinity of Na,K-ATPase for sodium, although dependent on phospholipase C, was independent of calcium signaling, was insensitive to inhibitors of classical protein kinases C and of ERK1,2 phosphorylation, but was fully prevented by the nonspecific protein kinase inhibitor staurosporine, as was the increase in transepithelial voltage. In conclusion, PAR2 increases sodium reabsorption in rat thick ascending limb of Henle's loop along both the transcellular and the paracellular pathway. PAR2 effects are mediated in part by a phospholipase C/protein kinase C/ERK1,2 cascade, which increases Na,K-ATPase maximal activity and the paracellular sodium permeability, and by a different phospholipase C-dependent, staurosporine-sensitive cascade that controls the sodium affinity of Na,K-ATPase.

Proteinase-activated receptors (PARs) 4 constitute a unique class of the seven-transmembrane domain G protein-coupled receptors. Their ligand, tethered in their extracellular domain, is unmasked upon cleavage of their N-terminal domain by specific proteases (1). To date, four isoforms of PARs (PAR1-4) displaying different profiles of activation by serine proteases have been cloned. PAR1, PAR3, and PAR4 are activated by thrombin, whereas PAR2 is activated by trypsin. PARs can also be activated experimentally, without proteolysis, by synthetic peptides that mimic their ligand domain.
PAR1 and PAR2 are expressed in most tissues, where their stimulation is coupled to the activation of phospholipase C (PLC) ␤ and, secondarily, to that of various protein kinases C (PKCs), which often trigger ERK 1,2 phosphorylation (2)(3)(4). Besides PLC␤, PAR1 and PAR2 may also activate other signaling pathways in a tissue-specific manner (4).
Although PAR1 and PAR2 are expressed at high level in the kidney (5)(6)(7)(8), their effects on renal function remain mostly unknown. PAR1 has been involved in inflammatory cell-mediated renal injury in crescentic glomerulonephritis (9,10), and PAR2 has been involved in the development of interstitial fibrosis in IgA nephropathy (7). PAR1 and PAR2 have also been shown to induce vasoconstriction and vasodilatation of renal vessels, respectively, thereby decreasing and increasing glomerular filtration rate (11). However, whether they exert a physiological effect on epithelial cells function has not been established. PAR2 may be involved in the control of kidney ion transport because 1) it is expressed in renal tubule epithelial cells (5), 2) in many epithelia, including distal colon, pancreatic duct, and airways, activation of PAR2 induces chloride secretion through activation of calcium-dependent channels (12)(13)(14)(15)(16), and 3) similar observations have been made in M1 cells, a mouse cell line derived from renal collecting duct principal cells (5). It must be stressed, however, that the renal epithelium is a reabsorptive and not a secretory epithelium and that native renal tubule epithelia display neither chloride secretion nor apical calcium-sensitive chloride channels.
Thus, we investigated the role of PAR2 in the regulation of solute transport in a native kidney epithelium and its signaling pathway. In preliminary experiments in rat kidney, we found a high level of PAR2 mRNA expression along the whole renal tubule. This study was focused on the rat cortical thick ascending limb of Henle's loop (cTAL), a nephron segment that avidly reabsorbs sodium via both the transcellular and the paracellular route. Transcellular NaCl reabsorption is primarily energized by basolateral Na,K-ATPase and generates a lumen-positive transepithelial potential difference (PD te ) that serves as driving force for paracellular sodium reabsorption (17). Thus, sodium reabsorption in cTAL was evaluated through monitoring of the PD te , the conductance of the paracellular shunt pathway and the activity of Na,K-ATPase. Signaling events were evaluated pharmacologically.

EXPERIMENTAL PROCEDURES
Animals-Experiments were performed on male Sprague-Dawley rats (Charles River, L'Arbresle, France) weighing either 60 -80 g for in vitro microperfusion experiments and intracellular calcium ([Ca 2ϩ ] i ) measurements or 160 -180 g for Na,K-ATPase assay and immunoblotting. In preliminary experiments, we found no difference in Na,K-ATPase response to PAR2 activation between the two groups of rats (data not shown). Animals were fed the standard laboratory diet ad libitum with free access to tap water. All experiments were performed in accordance with the French legislation for animal care and experimentation.
Microdissection of cTAL-cTALs were dissected either from fresh kidney slices (microperfusion and [Ca 2ϩ ] i measurements) as described previously (18) or from liberase-treated kidneys (Na,K-ATPase measurement and immunoblotting). Kidney treatment with liberase (Blendzyme 2, Roche Diagnostics, Meylan, France) was designed so as to make microdissection of the large number of cTALs required for the different assays compatible with the preservation of PAR activity. Briefly, the left kidney was perfused in situ with 6 ml of Hank's solution supplemented with 1 mM glutamine, 1 mM pyruvate, 0.5 mM MgCl 2 , 0.1% bovine serum albumin, 20 mM Hepes, and 0.015% liberase (w/v), pH 7.4. Thin pyramids were cut from the kidney and incubated in 0.006% liberase solution for 20 -25 min at 30°C and thoroughly rinsed in microdissection solution supplemented with inhibitors of protease (2 g/ml aprotinin and 5 g/ml leupeptin). Preliminary studies showed that this treatment reduced PAR2 peptidic agonist-induced increase in ([Ca 2ϩ ] i ) by 65%, but nevertheless, allowed us to detect 2-fold stimulations of Na,K-ATPase activity and of ERK phosphorylation in response to PAR2 activation.
Measurement of [Ca 2ϩ ] i -Intracellular calcium concentration was determined on single cTALs by the fura 2-AM fluorescence, as described previously (19). Briefly, after isolation, each cTAL loaded with acetoxymethyl ester of fura 2-AM (10 M, 1 h at room temperature) was transferred to a perfusion chamber and immobilized by sucking each end within the tip of a holding micropipette. The peritubular fluid maintained at 37°C was continuously exchanged at a rate of ϳ10 ml/min. After a 5-10-min equilibration, fura 2-AM fluorescence of ϳ15 cells was measured with a standard photometric setup (MSP 21, Zeiss). Tubule autofluorescence was subtracted from the fluorescence intensities of fura 2-AM at 340 and 380 nm. Values were calculated as described previously (19).
Na,K-ATPase Assay-Na,K-ATPase activity was determined using pools of 4 -6 permeabilized cTALs by the previously described microassay (20). For determining the V max of the enzyme, total ATPase activity was determined in a solution containing 120 mM NaCl, 5 mM KCl, 10 mM MgCl 2 , 1 mM EDTA, 100 mM Tris-HCl, 10 mM Na 2 ATP, and 5 nCi/l [␥-32 P]ATP (2-10 Ci/mmol) at pH 7.4. For basal ATPase activity measurements, NaCl and KCl were omitted, Tris-HCl was 150 mM, and 2 mM ouabain was added. For determination of the sodium apparent affinity of Na,K-ATPase, cTALs were thoroughly rinsed in a sodium-free medium (100 mM choline chlo- ride, 1 mM CaCl 2 , 50 mM Tris-HCl, pH 7.4) after treatment with inhibitors and PAR2 agonists, and total ATPase activity was determined in a solution containing 10 mM MgCl 2 , 1 mM EDTA, 100 mM Tris-HCl, 10 mM Tris-ATP, 5 nCi/l [␥-32 P]ATP, and various concentrations of NaCl and KCl, the sum of their concentrations being maintained constant at 140 mM, at pH 7.4. Na,K-ATPase was taken as the difference between mean total and basal ATPase activities, each determined in triplicate. Values are means Ϯ S.E. from several animals.
Cell Surface Expression of Na,K-ATPase-The abundance of cell surface Na,K-ATPase ␣ subunit was quantified after cell surface biotinylation, streptavidin precipitation, and Western blotting, as described previously (21). Briefly, after treatment with or without agonist peptide SLIGRL-NH 2 (AP) (10 min at 37°C), pools of 40 cTALs were incubated in bovine serum albumin-free microdissection solution containing 1.5 mg/ml EZ-Link Sulfo-NHS-SS-biotin (Pierce) for 60 min on ice, and thereafter, the biotin was quenched with 0.1% bovine serum albumin solution. Tubules were lysed in HB solution (20 mM Tris-HCl, pH 7.4, 2 mM EDTA, 2 mM EGTA, 1% Triton X-100, 0.1% SDS, protease inhibitor mixture (Roche Diagnostics)), and proteins were precipitated by 20% streptavidin agarose resin (Pierce) overnight at 4°C in TLB solution (50 mM Tris-HCl, pH 7.4, 100 mM NaCl, 5 mM EDTA, protease inhibitor mixture (Roche Diagnostics)). The agarose resin was washed thrice in TLB solution and resuspended in Laemmli buffer. Proteins were denaturated for 15 min at 60°C and separated on 7.5% acrylamide SDS-PAGE. Blots were successively incubated with anti-Na,K-ATPase ␣ subunit antibody (1/10,000, gift from K. Geering) and anti-rabbit IgG horseradish peroxidase (1/10,000, Promega). In each experiment, duplicate samples were run in parallel. The densitometry of the ϳ100-kDa bands was quantitated by scanning and using the ImageJ software and normalized to cTAL length, and the mean value in AP-treated tubules was expressed as the percentage of that under basal conditions. Data are means from 4 experiments.
Immunoblotting-After treatment, pools of 50 -80 cTALs were solubilized at 90°C for 5 min after the addition of 1 volume of 2ϫ Laemmli buffer. SDS-PAGE was performed on 10% polyacrylamide gels, and proteins were electrotransferred to polyvinylidene difluoride membranes (GE Healthcare). After blocking in TBS-Nonidet P-40 (50 mM Tris base, 150 mM NaCl, 0.2% Nonidet P-40) containing 10% low fat milk, blots were successively incubated with anti-ERK antibody (p44/42 MAP kinase antibody, Cell Signaling Technology, 1/500 in TBS-Nonidet P-40 ϩ milk) and anti-rabbit IgG antibody coupled to horseradish peroxidase (Promega France, Charbonnières, France) and revealed with the Western Lightning chemiluminescence reagent Plus (PerkinElmer Life Sciences). After stripping (four successive incubations in 25 mM glycine pH 2, 0,2% SDS buffer), the polyvinylidene difluoride membranes were incubated with anti-phosphoERK antibody (phospho44/42 MAP kinase antibody, Cell Signaling Technology) and post-treated as above. Densitometry of the different bands was quantitated by scanning and using ImageJ software.
For each sample, the densitometry ratio of phosphoERK/ ERK was calculated, and in each experiment, these ratios in different experimental conditions were expressed as the percentage of the control group. Data are means Ϯ S.E. from several experiments.
In Vitro Microperfusion-cTALs dissected without liberase treatment were microperfused in vitro as described previously (18). Briefly, the dissected tubules were transferred to a perfusion chamber mounted on the stage of an inverted microscope and perfused by a gravity-driven system at a rate of ϳ15 nl/min. The bath flow rate was 12 ml/min to ensure a rapid and complete change of bath solutions, and its temperature was main- tained at 37°C. The PD te was measured at the perfusion end of the tubule via microelectrodes connected through an Ag/AgCl half-cell to an electrometer.
In a first series of experiments, cTALs were perfused under symmetrical conditions. The bath and perfusate solutions contained (in mM): 118 NaCl, 23 NaHCO 3 , 1.2 MgSO 4 , 2 K 2 HPO 4 , 2 calcium lactate, 1 sodium citrate, 5.5 glucose, 5 alanine, 10 Hepes, pH 7.4 (bath continuously gassed with 95% O 2 , 5% CO 2 ). In a second experimental series, cTAL were perfused in the presence of a NaCl diffusion gradient from the bath to the perfusate. For this purpose, the bath solution was unchanged, whereas NaCl concentration in the perfusate solution was reduced to 50 mM, and mannitol was added to achieve the same osmolarity.
Statistics-Results are expressed as means Ϯ S.E. from several animals. Comparison between groups was performed either by non-paired t test or by variance analysis followed by protective least squares difference Fisher's test, as appropriate.

RESULTS
Expression of PAR2 in the cTAL-Because PAR2 are classically coupled to activation of PLC and calcium signaling, we attempted to demonstrate functional expression of PAR2 in the cTAL by searching for changes in [Ca 2ϩ ] i induced by activation of PAR2. Basolateral addition of trypsin to cTAL dose-dependently and transiently increased intracellular calcium concentration (Fig. 1, A and B), with threshold and maximal effects observed with 0.15 and 10 nM, respectively. This effect of trypsin was related to its proteolytic activity since it was abolished in the presence of soybean trypsin inhibitor (Fig. 1B). It was mimicked by the PAR2-specific AP and its furoylated derivative (F-AP) but not by the reverse inactive peptide LRGILS-NH 2 (RP) (Fig. 1B). Thrombin (1 M), an agonist of PAR1, PAR3, and PAR4, was devoid of effect (data not shown). In the absence of extracellular calcium, the peak calcium response to trypsin was not altered (data not shown), demonstrating that it was accounted for by recruitment of an intracellular calcium pool.
Following a first stimulation by trypsin, PAR2 was rapidly desensitized (data not shown).
The addition of AP to cTAL increased almost 2-fold the phosphorylation level of ERK 1,2 (Fig. 1C). Pretreatment with the specific PLC inhibitor Et-18-OCH 3 (150 M) decreased the basal phosphorylation of ERK and prevented its stimulation by AP (Fig. 1D). All together, these results demonstrate that rat cTAL expresses functional basolateral PAR2 coupled to PLC stimulation and subsequent activation of ERK 1,2 .
Effect of PAR2 on Transepithelial Voltage-As described previously, in vitro microperfused rat cTAL displayed a lumenpositive PD te that spontaneously decreased with time (Fig. 2, A  and B), likely due to progressive turn off of in vivo stimulatory pathways. When a steady rate of PD te decrease was achieved, at

TABLE 1 Effect of protein kinase inhibitors on trypsin-induced increase in PD te in microperfused cTAL
cTALs were pretreated with inhibitors, at the indicated concentration, for at least 35 min during the PD te equilibration period before stimulation with trypsin (10 nM). Changes in PD te were recorded for at least 15 min after trypsin addition. Inhibitors targets and sensitivity (IC 50 or K i ) are from the Calbiochem catalogue. Only targets possibly inhibited at inhibitor concentrations used in this study are listed. MEK, MAP kinase/ERK kinase; PTK, protein tyrosine kinases; PI3K, phosphatidylinositol 3-kinase.

Inhibitor
Concentration ϳ0.07 mV/min (residual PD te ϭ 8.3 Ϯ 0.6 mV Ϯ S.E., n ϭ 30), the addition of trypsin (3-10 nM) to the perfusion bath increased PD te by ϳ1.5 mV within 5 min (Fig. 2, A and B). Thereafter, PD te decreased at the same rate as spontaneously observed in time control cTALs. Consequently, trypsin-induced hyperpolarization remained almost constant for at least 15 min. In contrast, at a higher trypsin concentration (30 nM), the decrease was more rapid, and the PD te returned to time control level within 15 min (Fig. 2B). Like low trypsin concentrations, F-AP induced a stable hyperpolarization. The hyperpolarizing effect of trypsin was abolished by soybean trypsin inhibitor (Fig. 2, A and B). The hyperpolarizing effect of trypsin was not altered by BAPTA, which buffers changes in intracellular calcium concentration. It was also insensitive to specific inhibitors of all classical (cPKCs), novel, and atypical PKCs known to be expressed in TAL (22)(23)(24) and was only mildly inhibited by the ERK 1,2 kinase inhibitor U0126 (Fig. 2C and Table 1). Inhibitors of cAMP-and cGMP-dependent protein kinases (PKA and PKG), calmodulin kinase (CaMK), myosin light chain kinase (MLC-K), protein tyrosine kinases, and phosphatidylinositol-3 kinase (i.e. other signaling pathways triggered by PAR2 in different cell types (4)) were also devoid of effect (Table 1). However, total prevention of trypsin-induced hyperpolarization of cTAL could be achieved with a low concentration of staurosporine (15 nM), a broad spectrum inhibitor of serine threonine kinases (Fig. 2D). Thus, trypsin-induced hyperpolarization of cTAL is not mediated by the classical PLC/ Ca 2ϩ /cPKC/ERK-dependent signaling pathway but by a calciumindependent pathway involving an unidentified staurosporine-sensitive kinase.
The addition of vasopressin (AVP), at a concentration that induces a maximal increase in PD te via activation of its V2 receptors and activation of PKA in rat TAL, increased the PD te by ϳ2.5 mV. Subsequent addition of trypsin, during AVP-induced stimulation, further increased PD te , but only by 0.5 mV, i.e. to a lesser extent than trypsin alone (Fig. 3, A and B). Conversely, the addition of vasopressin during the plateau response to trypsin increased PD te by only 1.5 mV, which is less than the initial response to AVP (Fig. 3, A and B). This demonstrates a partial additivity of PD te responses to stimulation of AVP V2 receptor and PAR2. Because these two receptors are coupled to a different signaling pathway, the partial additivity of their effects on PD te suggests that these effects result in part from the stimulation of at least one common target. Effect of PAR2 Stimulation on Na,K-ATPase-Treatment of cTALs with AP increased the V max of Na,K-ATPase by Ͼ80% within 5 min, and this stimulation was sustained for at least 20 min (Fig. 4A). Stimulation of Na,K-ATPase V max was also observed in response to trypsin and F-AP, whereas RP had no effect (Fig. 4B), indicating that it was secondary to activation of PAR2. Furosemide, a specific and potent inhibitor of apical sodium entry in cTAL, did not alter the stimulatory effect of AP on Na,K-ATPase V max (Fig. 4C), demonstrating that stimulation of Na,K-ATPase is a primary effect of PAR2 activation and is not secondary to increased sodium entry, an event known to trigger targeting of cytosolic Na,K-ATPase subunits at the plasma membrane (25). AP-induced increase in Na,K-ATPase V max was not associated with any increase in the expression of Na,K-ATPase ␣ subunit at the plasma membrane (Fig. 4D). In addition to its stimulatory action on Na,K-ATPase V max , AP also enhanced about 2-fold the apparent affinity of Na,K-ATPase for sodium, from ϳ40 to ϳ20 mM ( Fig. 5A and Table 2).
Inhibition of PLC with Et-18-OCH 3 (150 M) prevented the increase in Na,K-ATPase V max and apparent affinity for sodium ( Fig. 5B and Table 2). This inhibitory action of Et-18-OCH 3 was not related to a toxic, nonspecific effect since it did not alter cAMP-induced stimulation of Na,K-ATPase (data not shown). AP-induced stimulation of Na,K-ATPase V max was also abolished by blunting changes in intracellular calcium with BAPTA   Table 2). In contrast, AP-induced increase in sodium affinity was not prevented either by BAPTA or by GÖ 6976 or U0126 (Fig. 5, C-E, and Table 2). It was abolished by the wide spectrum kinase inhibitor staurosporine ( Fig. 5F and Table 2).
All together, these findings demonstrate that PAR2-induced activation of PLC has dual effects on Na,K-ATPase in rat cTAL. Firstly, it increases its V max activity via a signaling cascade that includes increased intracellular calcium concentration, stimulation of PKC␣/␤, and phosphorylation of ERK 1,2 . Positive regulation of Na,K-ATPase by ERK 1,2 has already been reported in the renal collecting duct (26). Stimulation of Na,K-ATPase V max activity by PKC␣ has also been observed in TAL in response to insulin-connecting peptide (24). Secondly, PAR2 activation unexpectedly increases Na,K-ATPase apparent affinity for sodium via a different, although PLC-dependent, pathway, which is independent of increased intracellular calcium, of PKC␣/␤, and of ERK 1,2 but sensitive to staurosporine, as is the trypsininduced increase in PD te . Based on these data, we conclude that the increase in sodium apparent affinity of Na,K-ATPase is a key component of the PAR2-induced increase in cTAL transepithelial PD te that involves a calcium-independent staurosporine-sensitive kinase, the nature of which remains to be determined.
Effect of PAR2 on Paracellular Ion Permeability-To measure the paracellular permeability to sodium, the transcellular transport of NaCl was blocked by the addition of luminal furosemide (100 M), and a diffusion gradient for NaCl was established between the bath and the lumen through partial substitution of perfusate NaCl with mannitol (Fig. 6A). Under these conditions, cTALs displayed a steady lumenpositive PD te (18.7 Ϯ 0.9 mV Ϯ S.E., n ϭ 13), generated by the diffusion of bath sodium to the cTAL lumen through the paracellular pathway. This lumen-positive potential con-

TABLE 2
Effect of PAR2 stimulation on Na,K-ATPase kinetic properties in rat cTAL firmed the preferential conductance of the paracellular pathway for sodium over chloride (17). After the addition of trypsin to the bath, the PD te increased by Ͼ1.5 mV within 10 min (Fig. 6, B and C), indicating increased permeability of the paracellular pathway preferentially for sodium.
This effect of trypsin on the paracellular conductance to sodium was fully blocked in the presence of inhibitors of the classical PKCs (GÖ 6976) and of ERK kinase (U0126) (Fig. 6, B and C). Because inhibition of PLC fully abolished PAR2-induced phosphorylation of ERK 1,2 , the effect of PAR2 on paracellular sodium permeability is likely mediated by the same PLC/ PKC/ERK cascade as that mediating the increase in Na,K-ATPase V max . In the light of these data, the lack of effect of cPKC inhibition on PAR2induced increase in PD te (Fig. 2C) suggests that the increase in PD te brought about by the stimulation of Na,K-ATPase V max is counterbalanced by its decrease related to stimulation of paracellular permeability (Fig. 7A).

DISCUSSION
Sodium reabsorption in TAL occurs via both the transcellular and the paracellular routes (Fig. 7B). The sodium gradient generated by basolateral Na,K-ATPase in TAL cells is mainly dissipated by the apical, furosemide-sensitive, electroneutral Na/K/2Cl cotransport system (NKCC2) that couples the downhill entry of sodium to the uphill transport of potassium and chloride. Potassium ions accumulated above Nernst equilibrium within the cell by NKCC2 are recycled across the apical membrane via inwardly rectifying, voltage-insensitive potassium channels (ROMK). Chloride ions leave the cells across the basolateral membrane mainly via chloride channels (ClC-K). Conductive diffusion of chloride and potassium depolarizes the basolateral membrane and hyperpolarizes the apical one, respectively. These two diffusion potentials in series FIGURE 6. Activation of PAR2 increases the paracellular sodium permeability. A, cTALs were microperfused in vitro in the presence of 100 M furosemide (to fully inhibit transcellular NaCl transport) and of a NaCl concentration gradient between the bath and the luminal fluid (NaCl bath, 120 mM; NaCl perfusate, 50 mM). Under these conditions, the PD te is generated by the conductive diffusion of Na ϩ and/or Cl Ϫ ions from the bath to the lumen through the paracellular pathway. B, the positivity of the mean PD te is accounted for by the much higher permeability of the paracellular pathway to sodium than to chloride. Trace a shows that conversely to the PD te generated by transcellular transport (Fig. 2), this sodium diffusion-generated PD te is stable as a func-   Ca 2ϩ ] i ), which activates cPKc and induces the phosphorylation of ERK. Activation of ERK is responsible both for increased permeability of the tight junction to sodium and for enhanced V max of Na,K-ATPase. Because inhibition of this pathway has no significant effect on the transepithelial voltage (PD te ), it is suggested that the opposite effects on PD te of increased V max of Na,K-ATPase and increased permeability to sodium of the paracellular pathway (thick arrows) are quantitatively similar. Activation of PLC also stimulates an unidentified staurosporine-sensitive protein kinase, which mediates an increase in the apparent affinity of Na,K-ATPase for sodium. Because inhibition of this pathway also abolishes PAR2-induced increase in PD te , the increased sodium affinity of Na,K-ATPase is likely a major actor of the change in PD te and to increased sodium reabsorption. B, basolateral Na,K-ATPase generates a sodium gradient allowing apical entry of sodium, potassium, and chloride via the furosemidesensitive cotransporter NKCC2. Diffusive exit of potassium by apical ROMK and of chloride by basolateral ClC-K generates a transepithelial voltage (PD te ) that drives paracellular reabsorption of sodium. Because the tight junctions are almost impermeable to chloride, there is no back diffusion of chloride toward the lumen. combine to generate a lumen-positive transepithelial voltage characterizing the TAL. This transepithelial voltage is the driving force for paracellular ion transport; because the permeability of the paracellular route is higher for cations than for anions, the shunt current is mainly carried by net sodium reabsorption.
Present results go to prove that activation of PAR2 stimulates both transcellular and paracellular sodium reabsorption in the rat cTAL. Trypsin and PAR2-specific peptide agonist not only increase the transepithelial voltage (Fig. 2), reflecting increased transcellular reabsorption of NaCl, but also the paracellular permeability to sodium (Fig. 6). In the presence of a physiological lumen-positive transepithelial voltage, this enhanced sodium conductance increases the amount of sodium reabsorbed through the shunt pathway. This effect is further amplified by the concomitant increase in transepithelial voltage brought about by the stimulation of transcellular transport.
Although PAR2-induced changes in paracellular permeability are mediated by the PLC/cPKC/ERK pathway in cTAL (Fig.  6), as in intestine and corneal epithelia (27,28), they display different features. 1) Paracellular permeability increases within minutes following activation of PAR2 in cTAL ( Fig. 6) but only after hours in other epithelia (27)(28)(29). 2) PAR2 stimulation induces intracellular redistribution of the tight junction protein ZO-1 in intestine, whereas we did not observe such an event in a TAL cell line that expresses PAR2 (data not shown). 3) Along with this last finding, changes in cTAL are more subtle as they mostly concern the selectivity of the paracellular pathway to sodium, whereas they result in major alterations of the barrier function of tight junctions in other epithelia, allowing, for example, the transepithelial crossing of macromolecules and possibly of bacteria (27,29). These findings suggest that the molecular targets of PAR2 pathway in TAL tight junctions are the proteins specifically responsible for their ion permselectivity, such as claudins, rather than the ubiquitous structural proteins.
Increased transcellular reabsorption of NaCl in TAL can be accounted for by stimulation of Na,K-ATPase, ClC-K, ROMK, or NKCC2. In the present study, we documented marked increases in both the V max (Fig. 4) and the apparent affinity for sodium of Na,K-ATPase ( Fig. 5 and Table 2), two events subsequent to PLC activation (Fig. 7). PAR2-induced stimulation of Na,K-ATPase V max is independent of increased apical sodium entry (Fig. 4C), suggesting a primary regulation of the pump by trypsin-like proteases. This does not exclude that besides Na,K-ATPase, ROMK, ClC-K, and/or NKCC2 may also be target(s) of PAR2. As a matter of fact, PAR2 activation of chloride and potassium channels has already been reported in secretory epithelia (12)(13)(14)(15)(16).
It has been shown previously that membrane translocation of PKC␣ by insulin-connecting peptide increases the V max of Na,K-ATPase in medullary TAL, without altering the number of units located at the plasma membrane (24). This suggests that C peptide increases the maximal turnover rate of pre-existing Na,K-ATPase units. Since PAR2 activation does not increase the abundance of Na,K-ATPase at the cell surface (Fig.  4D), a similar mechanism might account for the increased Na,K-ATPase V max induced here by PAR2 via PLC, cPKC, and ERK (Fig. 5). However, in another study on rat medullary TAL, activation of PKC␣ by superoxides has been shown to stimulate NKCC2 without associated change in Na,K-ATPase activity (23,30). This indicates that other cellular events should occur together with activation of PKC␣ to account for the agonist specificity of the TAL response.
Increased affinity of Na,K-ATPase for sodium has been described before in the proximal tubule (31,32) but not in the TAL. Interestingly, PAR2-induced increase in Na,K-ATPase affinity for sodium is secondary to activation of PLC, but it is independent of changes in [Ca 2ϩ ] i and insensitive to inhibitors of classical PKCs, depending on a yet unidentified staurosporine-sensitive kinase, (Fig. 5 and Table 1), as transepithelial voltage (Fig. 2).
Functionally, the dual stimulation of Na,K-ATPase allows increased transcellular reabsorption of sodium even in the presence of reduced intracellular sodium concentration, i.e. in the absence of primary stimulation of apical sodium entry via NKCC2. This mechanism is different from that induced by the AVP-induced cAMP/PKA pathway, which primarily relies on activation of apical sodium entry, secondarily to phosphorylation and targeting of NKCC2 to the apical plasma membrane (33). Therefore, cAMP-generating agonists and PAR2-activating proteases, by stimulating distinct signaling pathways in the same cell, may be expected to cooperate in enhancing NaCl transport in TAL. In support of this hypothesis, trypsin and AVP were found to induce partially additive effects on PD te in in vitro microperfused cTAL (Fig. 3). In vivo, sodium transport in TAL is under strong stimulatory regulation by multiple hormones and receptors that maintain the cyclic AMP/PKA pathway under maximal activation (34,35). Therefore, PAR2-activating proteases may currently be considered as the sole agonist capable of increasing TAL transport capacity, under in vivo conditions. The proteases that may activate kidney PAR2 have not been searched for in this study. Trypsin is unlikely a physiologically relevant stimulus, but several circulating serine proteases displaying trypsin-like activity are putative candidates. Alternately, renal kallikrein is a good candidate since it is secreted by neighboring connecting tubules (36) and is able to activate PAR2 (37).