INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

IT HAS BEEN NOTED THAT DURING free-flow micropuncture experiments in cortical distal tubules, "in vivo" K⁺ secretion at a given flow rate is considerably higher than during microperfusion with artificial Ringer solutions (23, 26). It is well known that distal K⁺ secretion is a function of flow rate, distal sodium load, pH. and transepithelial potential difference (PD), among other factors (14, 43). However, it has also been shown that during distal perfusion with native proximal fluid collected before the experiment, distal K⁺ secretion is markedly higher than when an artificial Ringer solution of comparable composition is perfused (27). This finding suggested the presence, in native proximal fluid, of endocrine or paracrine factors that might stimulate this transport process by acting at the luminal surface of tubule cells. A similar suggestion has been made for fluid and sodium reabsorption in proximal tubule (19).

In the present work, we have investigated the role of arginine vasopressin (AVP) in distal K⁺ secretion. AVP is a peptide hormone that has been found to affect this process and is present in luminal fluid in physiological conditions (6, 21). Vasopressin has been shown by several groups to stimulate K⁺ secretion in cortical distal tubule (9, 12) and in cortical collecting duct (8, 35, 41) and is found in final urine in significant concentrations (21). ADH action on K⁺ secretion has been found mostly in rat cortical collecting duct, especially when electrolyte transport is stimulated by mineralocorticoids such as DOCA (35, 40). In addition, luminal action of AVP on electrolyte transport was observed in rabbit cortical collecting duct (2, 21).

Most studies have detected AVP action when applied at the basolateral surface of distal tubules and collecting ducts. This action has been shown to be mediated mostly by V₂ receptors via the adenylate cyclase/cAMP signaling system (8, 17). However, in recent years V₁ receptors have been detected both in apical and basolateral membrane domains and have been shown to mediate AVP activity at the luminal surface of cortical collecting duct via phospholipase C/inositol 3,4,5-triphosphate (IP₃)/calcium signaling (17, 21, 32).

In the present work, we have applied AVP and its antagonists from the luminal surface of the cortical distal tubule (initial collecting duct) and measured the rate of K⁺ secretion by an in vivo microperfusion technique. The results have shown that AVP is an important modulator of distal K⁺ secretion also when applied from the luminal cell surface in this segment, acting via V₁ receptors.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Male Wistar rats weighing 180-320 g were anesthetized with Inactin, 100 mg/kg, and prepared for in vivo micropuncture as described previously (15). Stationary microperfusion experiments were performed as summarized in Fig. 1. A proximal tubule was punctured by means of a double-barreled micropipette, one barrel being used to inject FDC green-colored perfusion solution, and the other to inject Sudan black-colored castor oil used to block the injected fluid columns in the lumen. The control solution contained 100 mM NaCl, 20 mM Na HEPES, 0.5 mM KCl, 1 mM CaCl₂, and raffinose (added to minimize fluid reabsorption) to reach an osmolality of 300 mosmol. pH was adjusted to 6.5. A single micropipette containing the same Ringer solution plus the polypeptide agent was impaled into a neighboring proximal loop or into an early distal loop. A late distal segment of the same nephron, recognized by the colored perfusion and by having a transepithelial PD of >20 mV, lumen negative, was impaled by a double-barreled asymmetric microelectrode, the larger barrel containing at its tip the K⁺-sensitive ion-exchange resin (Fluka, Buchs, Switzerland) and the smaller (reference), 0.24 M NaCl and 0.76 M Na acetate, colored by FDC green. This K⁺-free solution was calculated to have mean similar cation and anion mobilities, because the mobility in solution of Cl is larger than that of Na⁺ and that of acetate, smaller. The microelectrode had a tip diameter of ~1 µm, and the reference barrel had a resistance of <5 M. Additional properties of the microelectrode were described previously (45). Standards had a composition of 3, 10, or 30 mM KCl, and 100 mM NaCl was added to each of these solutions to compensate for the Na⁺ sensitivity of the resin. Mean decade voltage difference for K⁺ was 42.8 ± 0.51 (n = 132) mV. The electrodes were calibrated before and after every impalement by superfusion of the kidney surface with standards at 37°C. A luminal oil block was split by perfusions performed by hand-held, air-filled syringes connected to the micropipette holders by polyethylene tubing, applying pressure to the solutions in the micropipettes. Perfusion rate was considered adequate when the color of the perfused segment was that of the perfusion solution, the segment was only moderately expanded, and the perfusion rate was sufficient to lower luminal K⁺ concentrations to values near those of the perfusion fluid, i.e., 0.5 mM. After the concentration reached this level, perfusion was stopped and an additional oil block was introduced into the tubule lumen, and the increase in luminal K⁺ activities, representing K⁺ secretion, followed until a stable level was reached.

View larger version (18K): [in this window] [in a new window]   Fig. 1.   Schematic drawing of experimental system. Left: double-barreled perfusion pipette with perfusion solution (P) and castor oil (C). Right: double-barreled microelectrode with K+-sensitive ion-exchange resin (IE) and reference solution (Ref). An additional single-barreled micropipette (S) is used in proximal or early distal tubule for experimental solution. This figure is modified from Ref. 11.

Every tubule was perfused first with control solution, and then with the active agent, allowing for paired measurement of K⁺ secretion. By this technique several (~2-5 each) control, experimental, and recovery curves were obtained, the mean of control plus last recovery curves and experimental curves constituting the pair of values for this tubule. The value of n given for an experimental condition corresponds to the number of perfused tubules, approximately one to three being perfused in one rat. Statistical evaluation was performed by the paired t-test, comparing the means of control and experimental values of every tubule. As seen in Fig. 2, luminal K⁺ activity fell to 0.5 mM initially and then recovered progressively to a stationary level (K⁺_s). The voltage between the microelectrode barrels, representing the luminal K⁺ activity, was sampled every second by an analog-digital converter (Lynx, Sao Paulo, Brazil) in a microcomputer (model 333D, Dell). At the same time, the PD between the reference barrel and ground (the rat tail) was recorded, giving the evolution of transepithelial PD with time during the perfusion (see Fig. 2). The data were analyzed by a Visual Basic program by Excel software, fitting an exponential to the approach of K⁺ activities to their stationary level by plotting the differences between luminal K⁺ activity and K⁺_s against time, as given in Fig. 3. The half-time (t_1/2) of the approach of K⁺ activities to their stationary level was calculated from this exponential. Secretory K⁺ fluxes (J_K) were obtained by the following relationship
where t_1/2 is the half-time of K⁺ activity increase, K⁺_s is the stationary K⁺ activity, K⁺_o is the initial K⁺ activity (taken as 0.5 mM), and r is the tubule radius (45), a relationship we have used before for the measurement of bicarbonate reabsorption (11).

View larger version (16K): [in this window] [in a new window]   Fig. 2.   Record of experiment measuring K+ activity (A) and transepithelial potential difference (Vt; B) in renal distal tubule of rat. Note that during perfusion with 0.5 mM K+ solution (, P) there is a marked reduction of both PD and K+ activity, which both return to control levels with time after blocking of perfusion solution with oil in lumen (, B).

View larger version (18K): [in this window] [in a new window]   Fig. 3.   Difference in K+ concentration ([K+]; in mM) against time (s) during control and 1011 M arginine vasopressin (AVP) perfusion of cortical distal tubule, after perfusion with 0.5 mM [K+] and blocking with oil. Difference refers to [K+] in stationary condition minus [K+] at time t. Continuous curve is calculated variation of [K+] with time, assuming exponential approach of these differences to stationary [K+] level.

Vasopressin, the V₁-receptor antagonist (anti-V₁; -mercapto---cyclopenta-methylenepropionyl¹, O-Me-Tyr²-Arg⁸ vasopressin; MCMV), and the V₂-receptor antagonist (anti-V₂; adamantaneacetyl¹, O-Et-D-Tyr², Val⁴ aminobutyryl⁶, Arg^8,9 vasopressin; AAV) (28) were obtained from Sigma Chemical (St. Louis, MO). Benzamil (benzylamiloride hydrochloride) was obtained from Research Biochemicals (Natick, MA). Other chemical products were of analytic grade.

Statistical comparisons were made by the paired t-test, or, when nonpaired groups were compared, by ANOVA followed by the Bonferroni contrast test. The probability of 0.05 (5%) was taken as the limit of significance.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The basic experiment of this series involves injecting a low (0.5 mM)-K⁺ solution into the tubule lumen and following the changes of this concentration back to its stationary level after the fluid column is blocked in the lumen (Fig. 2). This procedure allows for the construction of graphs such as that in Fig. 3, which gives the evolution of the difference between the stationary K⁺ concentration and the concentration at time t, which decays exponentially toward zero. The data are obtained by digitization of the measurements performed by means of the potassium microelectrode, which are then transferred to an Excel chart and processed by a "macro" within this software. Figure 3 compares data obtained during perfusion with control Ringer solution, that is, a perfusion in the absence of any active agent, with data obtained during perfusion with solution to which 10¹¹ M AVP was added. Figure 4 shows a sequence of perfusions in one distal tubule to which 10¹¹ M AVP was applied. It is clear that during luminal perfusion with AVP a consistent rise in J_K is obtained. Such a sequence gives rise to one pair of data points in all our tables or figures, control and experimental. The control value includes the mean of the points before application of AVP plus those corresponding to recovery of control levels after AVP, and the experimental value includes the points during AVP perfusion.

View larger version (10K): [in this window] [in a new window]   Fig. 4.   K+ secretion rates (JK) measured during sequence of perfusions in a distal tubule in control conditions and during perfusion with 1011 M AVP.

To study the effect of conditions that are known to modify K⁺ transport in distal tubule, luminal perfusions with benzamil, an amiloride analog with almost only Na⁺-channel blocking activity and no effect on Na⁺/H⁺ exchange (24), and with Ba²⁺, a K⁺ channel blocker, were performed. Table 1 gives mean K⁺_s values measured during perfusion, when luminal K⁺ levels returned to their stationary level, t_1/2, half-times of the approach of luminal K⁺ concentrations to their stationary level, and transepithelial PD and J_K in control conditions and after perfusion with 10⁴ M benzamil; n is the number of perfused tubules. K⁺ concentrations are markedly reduced in this condition, and t_1/2 is significantly higher than control values, indicating that the influx of K⁺ is impaired by this agent. In addition, it is noted that the mean J_K falls markedly; as expected, transepithelial PD falls to near zero in this condition. Table 2 shows experiments in which K⁺ channels are blocked by luminal perfusion with 3 mM Ba²⁺ with and without addition of AVP 10¹¹ M. It is clear that perfusion with Ba⁺² reduces K⁺ secretion markedly, with reduction of stationary K⁺ concentration ([K⁺]_s)and increase in t_1/2. Transepithelial PD is moderately increased, as expected. However, when a comparison is made with PD measured immediately after perfusion was started , a value of 38.3 ± 3.5 (n = 17) mV was obtained in a control group, against 56.8 ± 4.0 (n = 21) mV in tubules perfused with 3 mmol/l Ba⁺², a significant difference compatible with the expected decrease in luminal K⁺ conductance.

View this table: [in this window] [in a new window]   Table 1.   Stationary K+ concentration, half-time of K+ equilibration, transepithelial potential difference, and K+ secretion rate in cortical distal tubules of rats during perfusion with 104 M benzamil

Table 3 shows a summary of [K⁺]_s and t_1/2 data during AVP action on distal K⁺ secretion, giving the mean control data compared with the mean experimental values; n represents the number of perfused tubules. In most experiments [K⁺]_s levels are not significantly altered, whereas t_1/2 are significantly shortened when AVP is perfused and increased in the presence of the anti-V₁ receptor agent. Figure 5 shows J_K (in nmol · cm² · s¹) at different luminal AVP concentrations. Figure 6 gives the effect of the anti-V₁ receptor peptide MCMV, alone and in combination with AVP, on J_K, and Fig. 7, the effect of the anti-V₂ receptor peptide AAV. It is noted that 10¹¹ and 10⁹ M AVP stimulate K⁺ secretion significantly, which is an important finding because physiological plasma levels of AVP are in the range of 10¹² to 10¹¹ M, and final urine AVP levels are up to 1,000 times higher than plasma levels (29). AVP at 10⁶ M shows some stimulation, but without reaching significance. The addition of anti-V₁ 10⁶ M to 10¹¹ M AVP reverts the action of AVP and in addition reduces K⁺ secretion significantly below control, but anti-V₁ alone has the same effect. On the other hand, the V₂-receptor antagonist, anti-V₂ (see METHODS) has no significant effect when given alone and does not abolish the action of AVP when given together with this peptide. These data indicate that the effect of AVP observed during luminal perfusion of the peptide is mediated by V₁-type receptors and that, even in the absence of this agent, the perfusion with anti-V₁ reduces distal K⁺ secretion, suggesting background activity of the hormone in distal tubule. In addition, data in Table 2 indicate that AVP+Ba²⁺ in the lumen does not increase K⁺ secretion significantly, suggesting that AVP might act by affecting luminal K⁺ channels.

View this table: [in this window] [in a new window]   Table 3.   [K+]s values and equilibration t1/2 in cortical distal tubule perfused with AVP and anti-V1 and V2 receptor agents

View larger version (10K): [in this window] [in a new window]   Fig. 5.   Distal JK: effect of luminal perfusion with AVP at different concentrations [1011 (A), 109 (B), and 106 M (C)].

View larger version (11K): [in this window] [in a new window]   Fig. 6.   Distal JK: effect of luminal perfusion with AVP (A) and with V1-receptor antagonist (AV1) -mercapto---cyclopenta-methylenepropionyl1, O-Me-Tyr2-Arg8 vasopressin (MCMV; B and C). See METHODS.

View larger version (12K): [in this window] [in a new window]   Fig. 7.   Distal JK: effect of luminal perfusion with AVP (A) and with V2-receptor antagonist (AV2) adamantaneacetyl1, O-Et-D-Tyr2, Val4 aminobutyryl6, Arg8,9 vasopressin (AAV; B and C).

Table 4 gives mean transepithelial PD in control and AVP perfused distal tubules. Although there was a tendency of PD to decrease during the experimental period in some of the groups, differences between control and experimental groups were not statistically significant, which is compatible with the absence of differences in [K⁺]_s in most groups.

The activity of the V₁-receptor antagonist (AV₁) alone could be due to an unspecific action of this substance, that is, to an action not related to the blocking of V₁ receptors for AVP. To study this possibility, we used homozygous Brattleboro rats, which do not produce AVP but do have receptors for the hormone (34). Results obtained in these rats are given in Table 5 and in Fig. 8. It is clear that in these rats K⁺ secretion is significantly stimulated by 10¹¹ M AVP and that AV₁ abolishes the stimulatory action of AVP but, when perfused alone, does not inhibit potassium secretion, as had been found in control rats; this supports the view that this effect might be due to the presence of a basal level of AVP in the control group, which is absent in the Brattleboro strain. In these rats, distal transepithelial PD was not significantly different in the groups that were investigated, being on the order of 40-50 mV.

View larger version (12K): [in this window] [in a new window]   Fig. 8.   Distal JK in Brattleboro rats: effect of luminal perfusion with AVP (A) and with V1-receptor antagonist MCMV (B and C).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The use of stationary microperfusion to study distal K⁺ secretion is a methodological innovation. This method has been used before for the analysis of H⁺ secretion and bicarbonate reabsorption in nephron segments (11, 15) and for the determination of proximal tubule K⁺ transport (45). It allows for paired determinations of ion transport, that is, for comparison of control and experimental perfusions in the same tubule, while luminal K⁺ activities in the perfused segment is measured. In addition, the studied segment is punctured only with a thin microelectrode (tip diameter ~1 µm), because perfusion can be made at a distance from the measurement site. The technique was validated by studying the effect of some factors known to affect K⁺ secretion. Thus luminal perfusion in the presence of 3 mM barium reduced J_K to 22% of control (see Table 2). This finding indicates that blocking of luminal K⁺ channels probably abolishes cellular K⁺ secretion, which is responsible for 78% of distal K⁺ secretion, whereas the remainder probably represents K⁺ transport via the paracellular path along a favorable electrical potential gradient. Similar observations were made in the rabbit cortical collecting duct (30). This finding also suggests that K⁺ secretion by KCl cotransport is only a minor component of this process and that AVP stimulation does not involve this pathway (see Table 2) (10). In addition, reduction of transepithelial PD by a specific blocker of apical Na⁺ channels, benzamil (See Table 1), which depolarizes PD, also causes a marked reduction of distal K⁺ secretion (24), as expected from the electrical PD of this transport (13).

Stimulation of electrolyte, including K⁺, transport by vasopressin has been observed in several renal epithelia. Increase in K⁺ secretion was detected during parenteral vasopressin administration in distal tubule perfusions (12), and luminal application of the hormone to the cortical collecting duct activated luminal chloride conductance (2, 5, 31). We are showing here that in late distal tubule, corresponding to the initial collecting duct, this hormone is active in K⁺ transport when applied to the luminal cell surface. In addition, the action of the hormone is abolished by apical anti-V₁ (10⁶ M) but not by anti-V₂ (10⁶ M) receptor blockers.

V₁ receptors have been localized by immunostaining in distal tubule connecting segment and in cortical collecting duct at both apical and basolateral membrane (16). V_1a receptor mRNA has been detected by PCR in rat initial cortical collecting duct (the micropuncturer's late distal tubule) (39). In addition, AVP has been shown to act also via oxytocin receptors (38); phospholipase C-coupled V₁-type non-V_1a and non-V_1b receptors, leading to an increase in cell calcium, have also been suggested to mediate AVP action in inner medullary collecting duct cells (25). The V₁ antagonist used in the present work acts unspecifically on the different V₁ subtypes that have been described, not allowing for distinction between them. However, the V₁ receptor described by most investigators for the cortical collecting duct is the V_1a subtype (1, 25, 39).

Previous data from our laboratory have shown that luminal AVP acts on H⁺ secretion in both early and late distal tubule, stimulating this process at 10⁹ M (3). Evidence was obtained that AVP stimulates the Na⁺/H⁺ exchanger of the apical membrane of this segment. In addition, it was found that a V₁ antagonist (the same used in the present experiments) abolished the AVP effect, whereas a V₂ antagonist had no effect. In a comparison of these data to the present findings, it may be suggested that the regulation of H⁺ secretion via the Na⁺/H⁺ exchanger shares a common mechanism with that of K⁺ secretion in this segment.

The study of AVP antagonists gives important information about the signaling pathways that may be involved in the regulation of distal tubule K⁺ secretion. It is well known that V₂ receptors are present mostly at the basolateral membrane, where they mediate the hydrosmotic effect of AVP at picomolar concentrations. This mechanism is known to involve the adenylatecyclase-cAMP-protein kinase A pathway (4, 8). On the other hand, V₁ receptors mediate AVP action mostly via a G_q-11 protein-phospholipase C-IP₃-protein kinase C-Ca²⁺ pathway. Evidence for this path was described for rabbit proximal tubule (46), rat thick ascending limb (6, 36), and rabbit cortical collecting duct principal cells (7, 21). The impairment of AVP action by an anti-V₁-receptor antagonist and the absence of a similar effect by a V₂-receptor antagonist suggests that the signaling pathway involving protein kinase C and/or cell Ca²⁺ might be decisive for luminal action of AVP on K⁺ secretion.

Data obtained by the patch-clamp technique have brought important information about the mechanism by which AVP may affect K⁺ transport. It was shown that the addition of AVP to the bath medium of split-open cortical collecting ducts induced the activity of previously silent apical channels in the cell-attached configuration; that is, AVP induced a larger number of low-conductance K⁺ channels to function, increasing apical membrane conductance of principal cells (8). These authors also obtained evidence that this mechanism was mediated by the cAMP-protein kinase A pathway. On the other hand, in chick kidney cells in primary culture Ca²⁺-activated K⁺ channels have been found on the apical membrane, and it was shown that these channels increased their open probability when AVP was added to the medium in the cell-attached condition. These are 107-pS channels, and it is possible that they might mediate a Ca²⁺-dependent effect of apical membrane K⁺ conductance (18). It has also been proposed that this channel might be responsible for regulatory volume decrease, as would occur after cell swelling due to inhibition of the Na⁺/K⁺ pump, (37) but possibly also for stimulation of sodium entry into principal cells of the collecting duct by vasopressin (2, 33). More recently, an additional type of Ca²⁺-dependent, voltage-independent small K⁺ channel has been described; it has been found in a large number of tissues, and its properties are compatible with those described in the present paper (22, 44).

Besides AVP action on luminal channels, an indirect mechanism for activation of luminal K⁺ secretion has been proposed: luminal AVP, increasing cytosolic Ca²⁺ levels, could inhibit basolateral intermediate conductance (~150 pS) channels, impairing basolateral K⁺ recirculation and thereby increasing cell K⁺, and thus increasing the luminal gradient driving this ion's secretion (20). These considerations suggest that AVP action on distal K⁺ secretion might be mediated by increases in cytosolic Ca²⁺. Of course, more detailed studies will be necessary to define the K⁺ channel and signaling pathway responsible for the described action of AVP.

An interesting finding was the observed inhibition of K⁺ secretion by perfusion with the anti-V₁-receptor antagonist alone. This could be due to an unspecific inhibitory action of this peptide but might also indicate that in control conditions there may exist a basal level of AVP binding to luminal V₁ receptors, causing some tonic activation of K⁺ secretion. Inhibitory actions of this nature have not been reported before. However, perfusion of rat inner medullary collecting ducts with anti-V_1a receptor agents showed a moderate reduction of cell Ca²⁺ levels, which might impair a cellular Ca²⁺-dependent mechanism (25). To eliminate the possibility of an unspecific inhibitor effect of the anti-V₁ agent used in our experiments, we performed similar experiments in homozygous Brattleboro rats, which are known to be devoid of endogenous AVP production (42). In these rats, perfusion with exogenous AVP stimulated distal K⁺ secretion, showing the presence of receptors for this peptide. Although their distal control J_K values were not much lower than those found in Wistar rats, the anti-V₁ agent had no significant effect, showing that this peptide had no AVP-independent action. An additional question involves the duration of AVP binding to V₁ receptors in distal tubule apical membranes. Figure 4 indicates that AVP washout (perfusion with control solution after AVP-containing solution) in our experiments is not immediate but takes several minutes to approach preperfusion levels, suggesting that the maintenance of residual levels of AVP binding in control conditions might be expected. These findings support the view that in control Wistar rats a portion of the normally observed rate of distal K⁺ secretion may be dependent on endogenous AVP levels.