Functional Evidence for an Inward Rectifier Potassium Current in Rat Renal Afferent Arterioles

Materials and Methods

The in vitro hydronephrotic rat kidney¹⁶ was used to examine the role of Kir in the afferent arteriole. Use of animals complied with Canadian Council on Animal Care regulations. Unilateral hydronephrosis was induced in male Sprague-Dawley rats by ligating the left ureter under halothane anesthesia. Rats were obtained from the Animal Resource Center breeding facility at the University of Calgary. Kidneys were harvested after 6 weeks, when tubular atrophy had advanced to a stage allowing direct visualization of the renal microvasculature. The renal artery was cannulated, and the kidney was excised with continuous perfusion. Kidneys were perfused with modified DMEM (GIBCO) at 37°C, equilibrated with 5% CO₂, containing (in mmol/L) Ca²⁺ 1.6, bicarbonate 30, glucose 5, pyruvate 1, and HEPES 5. Endogenous prostanoids modulate vascular reactivity in both normal kidneys and in our model.¹⁶ Therefore, 10 μmol/L ibuprofen was included in the perfusate. Perfusion pressure was monitored at the level of the renal artery and was controlled by adjusting pressure within the perfusion reservoir. Kidneys were allowed 1 hour to recover before initiation of experimental protocols. Diameters were measured by online image processing.

Afferent arteriolar RMPs were measured as described previously.¹⁷ Hydronephrotic kidneys were excised as described above. The capsule was removed, and the perfused kidney was mounted on a Plexiglas post, allowing access to the surface. A custom microscope with a water-immersion ceramic objective, coupled to a projection lens and video system, provided high magnification for impalements and diameter measurements. A microelectrode (2 mmol/L KCl and beveled at 30° to 80 to 100 MΩ) was placed within a patch-type pipette (tip diameter 10 to 20 μm) using a custom holder with motorized positioner. The patch pipette was then positioned against the vessel wall and held with suction to stabilize the impalement site. A piezo device was used to impale the vessel. We have shown that RMP measurements obtained using this approach are similar to those obtained using conventional microelectrodes.¹⁷ An impalement was considered successful if each of the following criteria were met: sudden negative deflection in membrane potential, stable recording of membrane potential for at least 1 minute before treatment with Ba²⁺, and no change in microelectrode tip resistance on withdrawal. The effects of Ba²⁺ on RMP and diameter were recorded simultaneously using this approach.

Glibenclamide and ibuprofen were obtained from Research Biochemicals International. All other chemicals were obtained from Sigma. KCl solutions 10 and 15 mmol/L were prepared by isotonic substitution for NaCl. Ouabain was made fresh for each experiment (3 mmol/L in DMEM). Phentolamine and propranolol were added during the ouabain experiment to avoid effects mediated by neurotransmitter release and to the CEC experiments to avoid effects of adrenoceptor modulation.

Data are expressed as mean±SEM. The number of replicates (n) refers to the number of vessels. Generally, no more than 2 vessels were studied per kidney (where appropriate, N refers to number of kidneys). Differences between means were evaluated by Student’s paired t test unless otherwise noted. P<0.05 was considered significant.

Results

Figure 1⇓ illustrates the vasodilator response of an afferent arteriole to 15 mmol/L KCl. Vasoconstriction was induced by elevating renal arterial pressure (RAP) from 80 to 160 mm Hg, reducing diameters from 16.3±0.4 to 8.1±0.5 μm (n=25, P<0.0001). In this setting, 15 mmol/L KCl caused afferent arteriolar dilation to 13.6±0.5 μm (P<0.0001). Returning to 5 mmol/L KCl restored myogenic vasoconstriction to previous levels (8.5±0.5 μm, P=0.26).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

K⁺-induced dilation of afferent arterioles. A, Original tracing illustrating RAP-induced vasoconstriction and K⁺-induced dilation in the afferent arteriole. B, Summary of data obtained in 25 arterioles subjected to the protocol illustrated in panel A. *P<0.0001 vs elevated RAP alone.

We next assessed the effects of Ba²⁺ and CEC on K⁺-induced vasodilation. The control studies presented in Figure 2A⇓ illustrate that repetitive K⁺-induced responses were stable over time. As shown in Figure 2B⇓, Ba²⁺ attenuated this response in a concentration-dependent manner. Ba²⁺ also reduced basal diameters (considered in detail below), precluding a calculation of the ED₅₀. Nevertheless, Ba²⁺ reduced the K⁺-induced dilation over concentrations that selectively block Kir^{10 11 12} (10 to 100 μmol/L). Complete blockade was observed at 100 μmol/L (P=0.54, n=6). At 100 μmol/L, Ba²⁺ may block ATP-sensitive K⁺ channels (K_ATP).¹⁴ However, glibenclamide had no effect. Thus, 15 mmol/L KCl increased diameters from 8±1.4 to 15.1±2.1 μm in the presence of vehicle (ethanol) and from 8.3±2.6 to 15.1±2.5 μm in the presence of 10 μmol/L glibenclamide (n=3, P=0.9). The K⁺-induced dilation was also blocked by 100 μmol/L CEC, an agent recently shown to block Kir¹⁵ (Figure 2C⇓). Before application of CEC, 15 mmol/L K⁺ induced dilation in preconstricted arterioles (9.5±1.6 versus 15.5±0.9 μm with 15 mmol/L KCl, RAP=140 mm Hg, n=5). CEC caused additional constriction, reducing diameters to 6.4±1.1 μm, and completely blocked the K⁺-induced dilation. Notably, application of 15 mmol/L KCl during CEC treatment caused additional constriction (to 4.9±1.1 μm, n=5), consistent with the change in the K⁺ equilibrium potential.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

Barium and CEC blockade of K⁺-induced vasodilation. A, Time controls illustrating reproducible vasodilatory response of 4 vessels to 6 consecutive challenges with 15 mmol/L KCl. B, Effects of Ba²⁺ on K⁺-induced vasodilation. Ba²⁺ decreased basal diameter and blocked K⁺-induced vasodilation in a concentration-dependent manner (1, 3, 10, 30, and 100 μmol/L; ▵, ▿, ⋄, □, and ○, respectively; n=6). C, Blockade of K⁺-induced dilation by 100 μmol/L CEC. Propranolol and phentolamine (10 μmol/L) were present to prevent adrenoceptor-mediated effects.

The above findings suggest that the K⁺-induced dilation is mediated by Kir on the basis of the effects of Ba²⁺ and CEC. We next examined the effects of blockade of the Na⁺-K⁺ ATPase. Vessels were preconstricted by administering 3 mmol/L ouabain (RAP 80 mm Hg), a concentration sufficient to block the rat isoform.¹⁴ Kidneys were pretreated with phentolamine and propranolol (10 μmol/L) to prevent effects attributable to the release of neurotransmitters. As shown in Figure 3⇓, ouabain reduced diameters from 16.3±1.5 to 4.2±1.5 μm (n=4, P=0.022), but K⁺-induced vasodilation was preserved (increasing diameters from 4.2±1.5 to 13.7±0.9 μm, P=0.0246). In these studies, 10 mmol/L KCl was used to allow for the decrease in internal [K⁺] induced by ouabain. When kidneys were treated with both 100 μmol/L Ba²⁺ and ouabain, diameters decreased from 18±0.9 to 3.8±1 μm (n=4, P=0.0033), and KCl-induced dilation was blocked (3.9±1.2 μm, P=0.56). The ability of Ba²⁺ to block KCl-induced dilation was readily reversed when Ba²⁺ was removed from the perfusate (Figure 4A⇓).

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.

Lack of effect of ouabain on K⁺-induced vasodilation. Tracings illustrate K⁺-induced vasodilation in afferent arteriole preconstricted with 3 mmol/L ouabain (A) and blockade of this response by 100 μmol/L Ba²⁺ (B). Mean data are depicted in panel C (n=4 each). All studies carried out in the presence of 10 μmol/L phentolamine and propranolol.

• Download figure
• Open in new tab
• Download powerpoint

Figure 4.

Tracings illustrating reversibility of Ba²⁺-induced inhibition of KCl-induced vasodilation (A) and Ba²⁺-induced vasoconstriction (B). A, □ indicates exposure to 15 mmol/L KCl. Data representative of 3 replicates each.

Kir is thought to contribute to RMP and resting tone in other vessel types.^{10 11 12} We thus determined the effects of Ba²⁺ blockade of Kir on resting diameter and RMP in the afferent arteriole. As shown in Figure 4B⇑, Ba²⁺ elicited a vasoconstrictor response of the afferent arteriole that was reversed on Ba²⁺ removal. Ba²⁺ blockade of Kir elicited dose-dependent vasoconstriction over concentrations of 1 to 100 μmol/L. At 80 mm Hg, 30 μmol/L Ba²⁺ reduced afferent arteriolar diameter from 15.6±0.5 to 7.6±1.6 μm (51±9% decrease, n=6, P=0.0041). Half-maximal constriction (EC₅₀) was seen at 19±3 μmol/L Ba²⁺ (Figure 5⇓). As described below, Ba²⁺ altered the threshold for myogenic vasoconstriction, shifting the response to lower pressures. We therefore characterized the effects of Ba²⁺ at low perfusion pressure (40 mm Hg). Under these conditions, 30 μmol/L Ba²⁺ reduced RMP from −47±2 to −34±2 mV (P<0.0001) and reduced diameter from 14.5±1 to 10.9±0.8 μm (n=10, N=8, P=0.002). These data are summarized in Figure 6⇓.

• Download figure
• Open in new tab
• Download powerpoint

Figure 5.

Ba²⁺-induced constriction in afferent arterioles (mean±SEM, n=6). *P<0.05 vs control. EC₅₀=19±3 μmol/L; RAP=80 mm Hg.

• Download figure
• Open in new tab
• Download powerpoint

Figure 6.

A, Original tracings of membrane potential and diameter of an afferent arteriole during the administration of 30 μmol/L Ba²⁺. B, Ba²⁺ (30 μmol/L)-induced depolarization (left) and vasoconstriction (right) in afferent arterioles perfused at a renal arterial pressure of 40 mm Hg. ○, individual measurements; •, mean±SEM (n=10). C, Summary of individual measurements.

These findings suggest an important role of Kir in the setting of RMP and basal tone in the afferent arteriole. However, this interpretation is based exclusively on the premise that 30 μmol/L Ba²⁺ selectively blocks Kir. To independently test this postulate, we examined the effects of reducing external K⁺ ([K⁺]_o). The inward rectification that is characteristic of Kir is attributed to polyamine or Mg²⁺ block. Elevated [K⁺]_o reduces rectification by displacing these blocking agents and enhancing outward current, thereby causing vasodilation.^{10 11 12 13} Conversely, reducing [K⁺]_o will enhance the effects of the blocking agents and reduce outward current carried by Kir. Thus, if Kir is the predominant K⁺ channel influencing RMP, then reducing [K⁺]_o should cause vasoconstriction. On the other hand, reducing [K⁺]_o shifts the K⁺ equilibrium potential to more negative values and will enhance the hyperpolarizing current carried by other K⁺ channel types. Kir is the only K⁺ channel in vascular smooth muscle in which reduced [K⁺]_o would enhance intrinsic block. Hence, if K⁺ channels other than Kir are dominant, reducing [K⁺]_o would cause vasodilation, not vasoconstriction. As shown in Figure 7⇓, reducing [K⁺]_o from 5 to 1.5 mmol/L induced a constriction at 40 mm Hg, reducing diameters from 15±0.6 to 12.8±0.9 μm (n=6, N=5, P=0.0088). To rule out the possibility that lowering [K⁺]_o induced vasoconstriction by reducing activity of the electrogenic Na⁺-K⁺ ATPase, we conducted additional experiments in the presence of 3 mmol/L ouabain. The ouabain-induced vasoconstriction was minimized by reducing RAP to 30 mm Hg. In the presence of ouabain, lowering [K⁺]_o from 5 to 1.5 mmol/L caused additional constriction, reducing diameters from 14±8.2 to 8.2±1.2 μm (P=0.0014, n=5). We next examined the effects of lowering [K⁺]_o at elevated RAP, conditions in which channels other than Kir, such as voltage- or Ca²⁺-activated K⁺ channels (Kv and K_Ca), might be active. We have shown that K_Ca is activated at elevated pressure, modulating myogenic vasoconstriction.¹⁸ Elevating RAP from 80 to 120 mm Hg reduced afferent diameters from 15.8±0.8 to 9.9±0.8 μm (n=6, P=0.0008). In this setting, reducing [K⁺]_o to 1.5 mmol/L increased diameters from 9.9±0.8 to 14.1±0.5 μm (n=6, P=0.0006). In concert with the Ba²⁺ data, these observations provide compelling evidence that Kir is the predominant K⁺ channel responsible for setting resting afferent arteriolar tone at low levels of perfusion pressure.

• Download figure
• Open in new tab
• Download powerpoint

Figure 7.

Effects of reducing [K⁺]_o on afferent arteriolar diameter. Left panel depicts original tracings illustrating the vasoconstriction evoked by 1.5 mmol/L KCl at 40 mm Hg (top) and the vasodilation evoked in the same vessel at 120 mm Hg (bottom). Mean data (±SEM) depicted on right.

Accordingly, alterations in Kir might contribute to myogenic activation. Therefore, we examined the effects of Ba²⁺ on the myogenic response. As shown in Figure 8A⇓, graded afferent vasoconstriction is elicited when RAP is increased from 20 to 180 mm Hg. Figure 8A⇓ additionally illustrates that in a series of time controls, 6 repeated challenges to increased renal arterial pressure (administered over a 7-hour period) elicited identical myogenic responses (n=4). As shown in Figure 8B⇓, 1⇑ to 100 μmol/L Ba²⁺ elicited vasoconstriction at low RAP but did not prevent additional constriction when RAP was increased. This is additionally illustrated in Figure 8C⇓, which depicts the response as the percent change in diameter. Ba²⁺ caused a leftward shift in the activation curve, perhaps reflecting the impact of blockade of Kir on basal tone and RMP. However, despite the Ba²⁺-induced preconstriction, the arterioles responded in an incremental manner to elevations in RAP, suggesting that Kir modulation is not a requisite component of myogenic signaling in this vessel.

• Download figure
• Open in new tab
• Download powerpoint

Figure 8.

Effect of Ba²⁺ on the myogenic response. A, Time controls illustrating reproducible responses of 4 arterioles to 6 consecutive challenges with stepped elevation in RAP. B, Increasing concentrations of Ba²⁺ (1, 3, 10, 30, and 100 μmol/L; □, ▵, ▿, ⋄, and ○, respectively) elicited vasoconstriction at low RAP and augmented the maximal response to RAP (controls, •; n=6). C, Data from panel B expressed as percent change relative to the diameter at 40 mm Hg. Increasing Ba²⁺ concentrations elicited a leftward shift in the pressure-activation curve.

Discussion

The present study provides the first evidence for a functional role of Kir in the renal afferent arteriole. We demonstrated that 15 mmol/L KCl elicits a vasodilation in this vessel and that this response was not prevented by ouabain but was sensitive to CEC and Ba²⁺ at concentrations reported to selectively block Kir. Moreover, manipulations that would reduce Kir currents under basal conditions (30 μmol/L Ba²⁺, low [K⁺]_o) elicit both depolarization and vasoconstriction, suggesting that Kir is a significant determinant of RMP and resting tone in the afferent arteriole.

The properties of the K⁺-induced dilation in the afferent arteriole are similar to those attributed to Kir in other vessel types. Application of 6 to 16 mmol/L KCl to preconstricted isolated posterior cerebral arteries,^{6 14} middle cerebral arteries,¹⁹ and coronary arteries⁶ results in dilation, and this response is blocked by micromolar levels of Ba²⁺. In these preparations, as in the present study, neither ouabain nor glibenclamide prevented this response. Electrophysiological characterization of Kir in vascular smooth muscle was first reported by Hirst and Neild,⁴ who described currents exhibiting inward rectification in submucosal arterioles of guinea pig ileum. Similar inwardly rectifying currents have been described in carotid artery⁵ and middle cerebral arteriole.^{20 21} Studies using patch-clamp techniques have identified K⁺-induced increases in outward current and inhibition of this current by micromolar Ba²⁺ in myocytes isolated from posterior cerebral artery^{2 6} and coronary artery.^{3 6 22} Recently, the targeted disruption of the Kir2.1 gene in mice was shown to be associated with a loss of both the inwardly rectifying K⁺ current and K⁺-induced dilation in cerebral arteries, suggesting its essential role in the response.²³ Our finding that CEC inhibited K⁺-induced vasodilation additionally supports a role for Kir. CEC is an irreversible agonist at α₂ adrenoceptors and an antagonist at α₁ adrenoceptors²⁴ but was recently reported to selectively inhibit inward rectifier K⁺ currents in skeletal muscle fibers and whole-cell K⁺ currents in cultured cell lines transfected with Kir2.1.¹⁵ These effects did not involve adrenoceptors, because they were not mimicked or altered by adrenoceptor ligands and could be demonstrated in cells that do not express adrenoceptors.¹⁵ Similarly, the inhibition of K⁺-induced dilation by CEC in the afferent arteriole that we observed occurred in the presence of phentolamine and propranolol, indicating that adrenoceptors are not involved.

Two previous studies have addressed the possible role of Kir in the renal circulation. Kurtz and Penner²⁵ described inward rectifying K⁺ currents in mouse juxtaglomerular cells, renin-secreting cells in the terminal segment of the afferent arteriole. They recorded whole-cell currents that were sensitive to changes in external K⁺ but did not assess Ba²⁺ sensitivity. More recently, Prior et al⁷ evaluated the Ba²⁺ and ouabain sensitivity of K⁺-induced dilation of rabbit renal arcuate arteries. In contrast to our findings with the afferent arteriole, these authors found that K⁺-induced dilation of the arcuate artery was attenuated by ouabain but not blocked by Ba²⁺, suggesting that Kir is not involved. As mentioned previously, Kir is thought to be more prevalent in resistance vessels than in conduit arteries, such as the arcuate artery. Our findings, together with those of Prior et al,⁷ indicate that this pattern holds in the kidney.

In our studies, 30 μmol/L Ba²⁺ produced marked constriction of the afferent arteriole and membrane depolarization (13 mV) at low RAP, suggesting a role for Kir in maintaining RMP and resting tone. A more modest Ba²⁺-induced constriction (8%, from 208±7 to 191±19 μm with 120 μmol/L Ba²⁺) was described in isolated perfused rat middle cerebral arteries by Johnson et al.¹⁹ In our studies, Ba²⁺ inhibited K⁺-induced vasodilation at 10 to 100 μmol/L and elicited vasoconstriction in the same concentration range (ED₅₀=19±3 μmol/L). The ED₅₀ for Ba²⁺-induced vasoconstriction in the middle cerebral artery is ≈60 μmol/L (see Figure 6⇑ in Johnson et al). Similarly, the IC₅₀ for Ba²⁺ block of K⁺-induced dilation in middle cerebral arteries was 20 to 40 μmol/L (see Figure 3⇑ in Johnson et al¹⁹ ). These values are close to the range for Ba²⁺ blockade of Kir channels. At −45 mV, the K_d for Ba²⁺-induced inhibition of Kir in posterior cerebral artery myocytes was estimated at 8 μmol/L.^{2 6} Similar K_d values, 3 μmol/L⁶ Ba²⁺ and 4.7 μmol/L²² Ba²⁺ at −40 mV, were reported for coronary artery myocytes.

K⁺-induced vasodilation is an important physiological mechanism in the cerebral and coronary circulations but has not been studied in the renal circulation. Periods of high neural activity, such as seizure, and ischemic, hypoxic, or hypoglycemic conditions increase K⁺ levels in cerebral spinal fluid to >10 mmol/L.²⁶ The resultant K⁺-induced dilation increases blood supply to areas of high metabolic demand.¹⁴ Similarly, in the coronary circulation, interstitial K⁺ levels may rise during periods of ischemia to levels sufficient to activate K⁺-induced dilation.²⁷ In the canine kidney, elevation of plasma [K⁺] to ≈10 mmol/L has been shown to increase renal blood flow (RBF) and glomerular filtration rate (GFR).²⁸ Similar effects of acute hyperkalemia on RBF and GFR have been reported in conscious sheep²⁹ and in rats³⁰ and rabbits.³¹ Although these investigations did not consider the possible role of Kir, our finding that elevated K⁺ promotes afferent arteriolar vasodilation through Kir is consistent with the observed increase in RBF and GFR.

By modulating glomerular inflow resistance, the afferent arteriole regulates glomerular capillary pressure and GFR when renal perfusion pressure is altered. Pressure-induced afferent vasoconstriction mediates this effect. Our observation that Kir is a major determinant of RMP in this vessel prompted an investigation into the possible involvement of Kir in myogenic signaling. Our findings do not support this postulate. Thus, although Ba²⁺ elicited membrane depolarization and vasoconstriction at low pressure, it did not prevent elevations in RAP from causing additional vasoconstriction. These data agree with findings from Kir2.1 knockout mice.²³ In this model, cerebral arteries showed normal myogenic reactivity but no K⁺-induced dilation. The observed changes in myogenic responses, including a leftward shift in the myogenic threshold, likely reflect the increase in basal tone and more positive RMP in the presence of Ba²⁺.

Defining the determinants of RMP under physiological conditions is of critical importance in regard to potential activation mechanisms. Agonists have been shown to inhibit specific K⁺ channels in isolated myocytes.^{10 11 12} However, such mechanisms could contribute to physiological responses only if the affected K⁺ channel contributes to RMP in the intact setting. The model we used allowed direct assessment of RMP and contractions in the intact afferent arteriole in situ under physiological conditions in regard to perfusion and transmural pressure. Our findings clearly indicate that Kir is a primary determinant of RMP under these conditions. Our model requires the induction of hydronephrosis to visualize the microvasculature. This preparation is well characterized and exhibits responses that are not only comparable to those observed using other renal microvascular models^{8 32} but also similar to those of the normal, intact kidney.^{16 33} Thus, there is no indication that hydronephrosis alters vascular K⁺ channel expression. Nevertheless, this possibility cannot be excluded. As described above, K⁺-induced vasodilation is observed in the normal kidney,^{28 29 30 31} but the role of Kir in this response has not been investigated. Moreover, barium-sensitive K⁺ currents have not been studied in myocytes or endothelial cells of the afferent arteriole. Information on the expression and regulation of Kir in this vessel would be of major interest.

In conclusion, this study provides the first functional evidence for Kir in renal afferent arterioles. Our findings suggest that Kir is a major determinant of in situ RMP and resting tone in this vessel. Accordingly, alterations in Kir activity or expression would likely impact on renal microvascular reactivity and could contribute to physiological or pathophysiological alterations in renal hemodynamics.