INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

ONE OF THE MOST SPECIFIC renal functions in mammals is the urine-concentrating mechanism, a mechanism found only in birds among all other animals (14). The high level of urine concentration that takes place in the inner medulla is a unique characteristic of the mammalian urine-concentrating mechanism. The inner medulla is a novel structure in which the ascending limb of Henle's loop is shaped like a thin limb, namely, the thin ascending limb (tAL). The tAL is present only in the mammalian inner medulla, where the highest osmotic environment in the kidney is formed. This segment is present, not in the short-looped, but in the long-looped nephron.

The tAL has long been considered important in the formation of concentrated urine by means of specific transport properties that dilute the urine without any movement of water across the epithelium. The tAL, the first diluting segment of the long-looped nephron in Henle's loop, dilutes urine concentrated in the descending limbs by extruding NaCl with simple passive extrusion (6). The mechanism of NaCl transport in the tAL has been debated since the 1960s. Most of the debate has focused on the presence or absence of active NaCl reabsorption in the tAL. Gottshalk and Mylle (2) provided the first direct evidence for very high permeability of Na⁺ in the hamster tAL. In 1965, Marsh and Solomon (12) identified the lumen-negative transepithelial voltage (V_t) in the tAL under free-flow conditions, suggesting that Na⁺ was actively reabsorbed in this segment. Later in vivo studies using electrodes with increased reliability found a lumen-positive V_t under free-flow conditions in the tAL (3, 11, 15), but this did not solve the issue of whether active Cl reabsorption was responsible for lumen-positive potentials.

In the 1980s, more extensive studies revealed that Na⁺ and Cl may be separately transported across the tAL epithelium (7, 9, 20). These reports suggested that luminal Na⁺ is passively diffused via tight junctions (16), whereas luminal Cl is transported across the cell membranes (5, 18, 20). Indeed, extensive studies in the 1990s further clarified this issue. In 1993, Uchida et al. (17) revealed the presence of the specific Cl channel protein CLC-K1 in this segment. Later studies clarified that CLC-K1 is expressed only in the tAL and that the major properties of this channel resemble those of the segment itself (18). Taken collectively, these studies make it plain that CLC-K1 is a kidney-specific chloride channel that mediates transepithelial chloride transport in the tAL. Our recent research collaboration demonstrated that the deletion of CLC-K1 by gene targeting results in a severe nephrogenic diabetes insipidus in mice (13). This observation strongly suggests that CLC-K1 in the tAL plays an important role in urine concentration and that the countercurrent system in the inner medulla is involved in the generation and maintenance of hypertonic medullary interstitium (13).

To further clarify the mode of NaCl reabsorption in the tAL in the present study, we conducted a series of experiments in CLC-K1 null mice. Our results strongly suggest that the indirect electrical coupling of Na⁺ and Cl is a basic mechanism of NaCl reabsorption in the tAL.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Isolation and microperfusion of renal tubules. Knockout mice deficient in CLC-K1 were generated by targeted gene disruption (13). Genotype analysis of tail DNA was performed by PCR. Homozygous wild-type (CLC-K1^+/+) and null (CLC-K1^/) mice resulting from breeding of heterozygotes were maintained on a regular diet and tap water until the day of the experiment. The ages of the mutant animals were matched with their wild-type controls. The mice were anesthetized by intraperitoneal injection of 50 mg/kg of pentobarbital sodium, and their left kidneys were removed. The renal tubule segment was microdissected with fine forceps under a stereoscopic microscope. A fragment of the tubule was transferred to a chamber on the stage of an inverted microscope and microperfused in vitro using Burg's method (1) with some modifications (6) at 37°C. Both sides of the tubule were initially microperfused with HEPES-buffered Ringer solution containing (in mM) 135 NaCl, 3.0 KCl, 2.0 KH₂PO₄, 1.5 CaCl₂, 1.0 MgCl₂, 10 HEPES, 100 urea, 1.0 sodium acetate, 5.5 glucose, and 5.0 L-alanine (solution 1, Table 1). The solution was titrated to pH 7.4 with NaOH at 37°C.

Measurement of transepithelial potentials. The V_t was measured by connecting the bath and perfusion pipette with a saturated KCl flowing boundary and an agar bridge containing NaCl-Ringer solution, respectively. A high-input impedance electrometer was used to monitor the V_t (Duo773; WP Instruments, New Haven, CT). To measure diffusion potential evoked by the transepithelial electrolyte gradient, the bath solution was exchanged, and the deflection of the V_t was observed. The KCl flowing boundary minimized the changes in the liquid junction potential caused by the solution exchange.

Measurement of relative permeability for Cl. Because the tAL is more permeable to Cl than to Na⁺, the lumen-positive diffusion voltages (V_d values) are generated when a NaCl concentration gradient is imposed across the tAL. The orientation of the voltage was compatible with the preferential conductance for electrolytes. Thus the NaCl V_d provides a good approximation for the relative permeability of Cl over Na⁺ (P_Cl/P_Na). The NaCl V_d was measured by reducing the NaCl concentration of the bathing fluid by 100 mM while the lumen was perfused with solution 1 (solution 2, Table 1).

Direct measurements of efflux coefficients for ²²Na and ³⁶Cl. To elucidate the permeabilities of Na⁺ and Cl in the tAL directly, the efflux coefficient (K_e) of ²²Na or ³⁶Cl was measured in the in vitro microperfused tAL in CLC-K1^+/+ and CLC-K1^/ mice. ²²Na (370 mBq/ml) or ³⁶Cl (185 mBq/ml) was added to the perfusate (final concentration). The lumen-to-bath efflux coefficient for isotope x (K_e x) was calculated as
where V_o is the collection velocity rate, L is the length of the tubule, and C^*_i and C^*_o are radioactivity (concentrations) of isotopes in the perfusate and collected fluid, respectively.

Chemicals. The compositions of the solutions used in this study are listed in Table 1. HEPES was obtained from Sigma (St. Louis, MO). All other chemicals were purchased from Wako Pure Chemical Industries (Osaka, Japan).

Statistical analyses. The data for the V_d represent the values obtained when the voltage became stable for at least 30 s. On rare occasions when the voltages were not stable, the maximal voltage deflections were chosen as representative values. Data are expressed as means ± SE. Comparisons within a group with one treatment used the paired Student's t-test. Comparisons between two groups employed the unpaired Student's t-test. A value of P < 0.05 was accepted as statistically significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Effects of NPPB on NaCl V_d. In the absence of a transmural NaCl concentration gradient, the V_t values were not different from zero, as Imai and Kokko (4) reported previously. To characterize the passive ionic permeabilities of Na⁺ and Cl, the transepithelial diffusional potential (V_d) was measured when the basolateral NaCl concentration was decreased by 100 mM. When the NaCl in the basolateral solution was reduced from 135 to 35 mM by isotonic replacement with urea, the V_d sharply turned toward the lumen-positive direction in CLC-K1^+/+ and CLC-K1^+/ mice, indicating the preferential permeation of Cl, whereas in CLC-K1^/ mice it continued moving in the lumen-negative direction. The steady-state V_d values were 15.5 ± 1.0 (n = 10), 12.8 ± 0.7 (n = 11), and 7.6 ± 1.4 mV (n = 16) in CLC-K1^+/+, CLC-K1^+/, and CLC-K1^/ mice, respectively (Fig. 1). These values in wild-type and heterozygous mice were very different from those in knockout mice (P < 0.0001, Student's unpaired t-test). The values were also slightly different between wild-type and heterozygous mice (P = 0.048).

View larger version (14K): [in this window] [in a new window]   Fig. 1.   Diffusion voltage (Vd) caused by transepithelial NaCl concentration gradient. Reduction of NaCl in the basolateral solution from 135 to 35 mM by isotonic replacement with urea caused marked positive deflections of the Vd values in CLC-K1+/+ and CLC-K1+/ mice and negative deflection in CLC-K1/ mice. The steady-state Vd values were 15.5 ± 1.0 (n = 10), 12.8 ± 0.7 (n = 11), and 7.6 ± 1.4 mV (n = 16) in CLC-K1+/+, CLC-K1+/, and CLC-K1/ mice, respectively. The Cl-to-Na+ permeability ratio (PCl/PNa) values are also depicted at the bottom of each bar.

+/

/

/

4

/

View larger version (13K): [in this window] [in a new window]   Fig. 2.   Effect of 5-nitro-2-(3-phenylpropylamino)-benzoate (NPPB) on the NaCl Vd values. The Vd values were measured in the presence of a transmural NaCl concentration gradient (100 mM higher in the lumen). When 0.1 mM NPPB was added to the bath solution, the Vd values were suppressed by 17.3 ± 1.8 (n = 9), 13.8 ± 0.5 (n = 11), and 0.2 ± 0.1 mV (n = 11) in CLC-K1+/+, CLC-K1+/, and CLC-K1/ mice, respectively. The changes in the Vd values of CLC-K1/ mice brought about by applying NPPB were not significantly different from 0. n.s., Not significant.

View this table: [in this window] [in a new window]   Table 2.   Effect of various Cl and Na+ transport inhibitors on NaCl diffusion voltage in tAL

Effects of ambient pH on V_d. It has been reported that Cl conductance in the tAL is regulated by H⁺. The Cl conductance is suppressed by acidic pH (8, 9). Briefly, we observed the effects of the pH of the bathing fluid on the V_d across the tAL generated when a transmural NaCl gradient was imposed. As shown in Fig. 3, the decrease in the pH of the bathing fluid from 7.4 to 6.0 sharply reduced the V_d values by 6.9 ± 2.0 (n = 10) and 4.2 ± 0.5 mV (n = 11) in CLC-K1^+/+ and CLC-K1^+/ mice, respectively. In CLC-K1^/ mice, however, the same maneuver changed the V_d values by only 0.1 ± 0.1 mV (n = 9), and this was not significant.

View larger version (13K): [in this window] [in a new window]   Fig. 3.   Effect of acidification of bathing fluid on the NaCl Vd values. The Vd values was measured in the presence of a transmural NaCl concentration gradient (100 mM higher in the lumen). The decrease in pH of the bathing fluid from 7.4 to 6.0 changed the Vd values by 6.9 ± 2.0 (n = 10) and 4.2 ± 0.5 mV (n = 11) in CLC-K1+/+ and CLC-K1+/ mice, respectively. The same maneuver, however, showed no effect in CLC-K1/ mice (0.1 ± 0.1 mV, n = 9).

Effects of protamine on V_d. We also examined the effect of the Na⁺ channel blocker protamine on the NaCl V_d across the tAL. When the composition of the perfusate and bathing fluid was identical, V_t was zero. Replacement of the bathing fluid with solution 2 caused a marked positive deflection of the V_d values in CLC-K1^+/+ and CLC-K1^+/ mice and a negative deflection in CLC-K1^/ mice. Under this condition, the addition of 300 µg/ml protamine into the bath caused a rapid lumen-positive deflection of the V_d in all three types of mice that reached a plateau in a few minutes. The V_d did not recover in any of the experiments when protamine was eliminated from the bath, but it recovered rapidly when 30 U/ml heparin was added under this condition. Figure 4 shows the results of experiments conducted to observe the effect of protamine on the NaCl V_d. Representative data showing the effects of protamine and heparin in the tAL of CLC-K1^+/+ and CLC-K1^+/ mice are depicted in Figs. 5 and 6, respectively.

View larger version (13K): [in this window] [in a new window]   Fig. 4.   Effect of protamine on the NaCl Vd values. The Vd values were measured in the presence of a transmural NaCl concentration gradient (100 mM higher in the lumen). The addition of 300 µg/ml protamine into the bath caused a rapid lumen-positive deflection of the Vd by 5.6 ± 0.7 (n = 9), 5.7 ± 1.1 (n = 9), and 12.6 ± 1.5 mV (n = 9) in CLC-K1+/+, CLC-K1+/, and CLC-K1/ mice, respectively.

View larger version (11K): [in this window] [in a new window]   Fig. 5.   A representative tracing of the Vd in response to protamine and heparin in the thin ascending limb (tAL) of CLC-K1+/+ mice. The Vd was measured in the presence of a transmural NaCl concentration gradient (100 mM higher in the lumen) in CLC-K1+/+ mice. The addition of protamine immediately increased the Vd by ~7 mV. Subsequent removal of protamine from the bath did not alter the Vd. The addition of 30 U/ml heparin then caused rapid recovery of the Vd.

View larger version (11K): [in this window] [in a new window]   Fig. 6.   Representative tracing of the Vd in response to protamine and heparin in the tAL of CLC-K1/ mice. Although the same protocol described in Fig. 5 was performed in the tAL of CLC-K1/ mice, the difference of the basal Vd, protamine, and heparin had similar effects on the Vd in CLC-K1/ mice.

Effect of protamine in the presence of NPPB. Isozaki et al. (5) reported that NPPB suppresses Cl conductance in this segment. If NPPB selectively inhibits Cl transport, Na⁺ conductance can be expected to become dominant in the presence of NPPB. We examined the effect of protamine under this condition. At first, we observed the V_d by the same methods used in the first protocol. When we added 0.1 mM NPPB to the bathing fluid, CLC-K1^+/+ and CLC-K1^+/ mice showed a rapid change and negative deflection of the diffusion potential, whereas CLC-K1^/ mice showed no such change. Under this condition, 300 µg/ml protamine significantly increased this V_d to the same extent in all three types of mice. This was also reversed by 30 U/ml heparin. Table 2 summarizes the results of various Cl and Na⁺ transport inhibitors on the NaCl V_d. The calculated Cl/Na⁺ permeability ratios are also given. The NPPB and ambient pH clearly decreased the positively oriented the NaCl V_d in CLC-K1^+/+ and CLC-K1^+/ mice while imparting no such effect in CLC-K1^/ mice. The addition of protamine to the bath increased the NaCl V_d in all three types of mice, and the presence of NPPB did not affect this protamine-induced increase.

Measurements of efflux coefficients for ²²Na and ³⁶Cl. To elucidate the characteristics of NaCl transport in the tAL of CLC-K1^+/+ and CLC-K1^/ mice directly, permeabilities of Cl and Na⁺ were measured using ³⁶Cl and ²²Na. As depicted in Fig. 7, the efflux coefficient for ³⁶Cl (K_e Cl) was strongly impaired in CLC-K1^/ mice, whereas the efflux coefficient for ²²Na (K_e Na) in CLC-K1^/ mice was not different from that in wild-type mice. These results strongly indicate that targeted disruption of the CLC-K1 gene in mice leads to selective impairment of Cl transport in the tAL of the renal medulla.

View larger version (16K): [in this window] [in a new window]   Fig. 7.   Efflux coefficients of 36Cl and 22Na in the tALs of CLC-K1+/+ and CLC-K1/ mice. Efflux coefficients of 36Cl (Ke Cl) and 22Na (Ke Na) were measured in the tALs of CLC-K1+/+ and CLC-K1/ mice. Although Ke Cl was significantly decreased in CLC-K1/ mice (188.3 ± 25.6 in 4 tubules vs. 17.2 ± 7.0 × 105 cm/s in 6 tubules), Ke Na was not significantly different between CLC-K1+/+ and CLC-K1/ mice (56.2 ± 15.7 in 4 tubules vs. 31.3 ± 2.3 × 105 cm/s in 6 tubules). The results also imply that the preparation of the tAL of CLC-K1+/+ and CLC-K1/ mice in the present studies was reliable for obtaining good viability.

Effect of NPPB on K_e Cl and K_e Na. To characterize the K_e Cl and K_e Na of the tALs in CLC-K1^+/+ and CLC-K1^/ mice, effects of 10⁴ M NPPB in the basolateral solution were examined. As clearly shown in Fig. 8, NPPB only inhibited the K_e Cl in the CLC-K1^+/+ mice. This result indicates that NPPB-sensitive Cl permeability in the tAL is predominantly mediated by CLC-K1, whereas Na⁺ permeability is entirely unaffected by NPPB.

View larger version (16K): [in this window] [in a new window]   Fig. 8.   Effects of NPPB on efflux coefficients of 36Cl and 22Na in the tALs of wild-type and knockout (KO) mice. The effects of 0.1 mM NPPB in bath on Ke Cl and Ke Na were examined in wild-type and KO mice. Ke Cl was decreased by 79.7 ± 19.3 (n = 4) and 2.7 ± 3.6 × 105 cm/s (n = 6) in CLC-K1+/+ and CLC-K1/ mice, respectively. Ke Na was increased by 6.7 ± 8.1 (n = 4) and 8.2 ± 3.9 × 105 cm/s (n = 6) in CLC-K1+/+ and CLC-K1/ mice, respectively.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Urine concentration is achieved by water reabsorption in the collecting duct according to the osmolar gradient across the tubule. In this process, the tAL plays an important role in the inner medulla by diluting urine and by maintaining the hypertonicity of the interstitium through reabsorption of NaCl. Whether NaCl is reabsorbed actively in the tAL has long been an important question. Kondo et al. (7, 9) and Yoshitomi et al. (20) provided indirect and direct evidence for the presence of stilbene-sensitive Cl channels in both luminal and basolateral membranes of the tAL. Furthermore, molecular biological studies demonstrated the presence of the CLC-K1, a kidney-specific chloride channel that mediates transepithelial Cl transport, in the tAL (17, 18). In the present attempt to identify the mechanism of Na⁺ transport across the tAL, the fragile properties of the tAL in CLC-K1 null mice unfortunately thwarted our attempts to obtain the direct data on the lumen-to-bath ²²Na and ³⁶Cl flux. However, in conjunction with previous characterizations of the tAL, the present data on the electrical properties of the tAL in CLC-K1 null mice probably already suffice to elucidate the nature of Na⁺ and Cl transport. The results reported here support the view that the paracellular shunt pathway of this segment has, in fact, selective Na⁺ permeability.

Characterization of Cl transport in the tAL. As shown in Fig. 1, CLC-K1^+/+ and CLC-K1^+/ mice showed lumen-positive deflection on reduction of basolateral NaCl concentration, whereas CLC-K1^/ mice generated an opposite deflection in the same condition. This phenomenon supports the view that the tAL is more permeable to Cl than to Na⁺ in normal mice. To further characterize the mechanism of this Cl conductance across the tAL, we examined effects of various Cl transport inhibitors. NPPB is a specific transport inhibitor of the Cl channel CLC-K1 in the tAL (5), and the ambient pH regulates the Cl conductance (8, 9). As shown in Figs. 2 and 3, the reduction of the pH or application of NPPB suppressed the V_d strongly and to approximately the same extent in CLC-K1^+/+ and CLC-K1^+/ mice, whereas the same maneuvers for CLC-K1^/ mice scarcely changed the V_d at all. As shown in Table 2, acidification of the bathing solution or application of NPPB induced a suppression of the NaCl V_d that corresponded to the decreases in P_Cl/P_Na in CLC-K1^+/+ and CLC-K1^+/ mice but showed no effect in CLC-K1^/ mice. Theoretically, the decrease in the P_Cl/P_Na ratio reflects the following changes: 1) decrease in P_Cl, 2) increase in P_Na, 3) a combination of 1 and 2, or 4) decrease in both parameters with a preferential decrease in P_Cl. In the previous studies, it was proved that the reduction of the pH or application of NPPB selectively decreased Cl permeability without affecting Na⁺ permeability (8, 9). Therefore, we conclude that acidic pH or NPPB decreases Cl transport, not in CLC-K1^/, but in CLC-K1^+/+ and CLC-K1^+/ mice. As the sensitivities to acidic pH or NPPB are characteristics of a major Cl transport pathway in the tAL, our results indicate that Cl transport in the tAL is indeed abolished in CLC-K1^/ mice.

Properties of Na⁺ transport in tAL. It has been reported that protamine increases the resistance of the paracellular shunt pathway in various epithelia. Because Koyama et al. (10) reported that protamine selectively inhibits Na⁺ conductance of the paracellular pathway, we decided to examine the effects of protamine on the V_d values during in vitro microperfusion.

/

+/

/

/

+/