Boosting the signal: Endothelial inward rectifier K⁺ channels

Abstract

Endothelial cells express a diverse array of ion channels including members of the strong inward rectifier family composed of K_IR2 subunits. These two-membrane-spanning-domain channels are modulated by their lipid environment, and exist in macromolecular signaling complexes with receptors, protein kinases and other ion channels. Inward rectifier K⁺ channels (K_IR) currents display a region of negative slope conductance at membrane potentials positive to the K⁺ equilibrium potential that allows outward current through the channels to be activated by membrane hyperpolarization, permitting K_IR to amplify hyperpolarization induced by other K⁺ channels and ion transporters. Increases in extracellular K⁺ concentration activate K_IR allowing them to sense extracellular K⁺ concentration and transduce this change into membrane hyperpolarization. These properties position K_IR to participate in the mechanism of action of hyperpolarizing vasodilators and contribute to cell-cell conduction of hyperpolarization along the wall of microvessels. Expression of K_IR in capillaries in electrically active tissues may allow K_IR to sense extracellular K⁺, contributing to functional hyperemia. Understanding the regulation of expression and function of microvascular endothelial K_IR will improve our understanding of the control of blood flow in the microcirculation in health and disease and may provide new targets for development of therapeutics in the future.

Introduction

Ion channels in the membranes of endothelial cells contribute to all aspects of the function of these cells [57]. Calcium influx and release through members of these membrane proteins determines intracellular Ca²⁺ concentration. This second messenger importantly regulates a number of endothelial cell processes related to the regulation of vascular function including endothelial cell production of autacoids such as NO, prostaglandins and epoxides of arachidonic acid (EETs) [31], as well as the activity of Ca²⁺-dependent ion channels [31]. Intracellular Ca²⁺ also regulates endothelial cell barrier function [39,64], gene expression [82,93], and proliferation [83,87]. Cell volume regulation also depends on the activity of plasma membrane ion channels [52]. Importantly, ion channels determine and regulate endothelial cell membrane potential. Modulation of this separation of charge, through changes in ion channel activity, functions as an important signal for endothelial cell-endothelial cell communication and endothelial cell-smooth muscle cell communication, because these cells are electrically coupled by gap junctions [23]. Membrane potential also affects the electrochemical gradient for movement of all ions across the plasma membrane, potentially modulating Ca²⁺ influx through endothelial cell Ca²⁺ channels [6,46,47,51,87], although this topic remains controversial [16,26,77,80,107]. Thus, the physiology and pathophysiology of endothelial cells depends heavily on the expression and function of ion channels.

As in all cells [50], K⁺ channels play a central role in setting and modulating membrane potential of endothelial cells. At the resting membrane potential of microvascular endothelial cells in intact pressurized vessels (~30–40 mV [32,102,103,116]), with physiological ion gradients (3–5 mM K⁺ in the extracellular fluid, ~140 mM K⁺ in the cytosol), the electrochemical gradient for diffusion of K⁺ across endothelial cell membranes is outward. Thus, K⁺ will flow out of cells when K⁺ channels open, producing membrane hyperpolarization. Closure of open K⁺ channels will have the opposite effect, membrane depolarization.

Endothelial cells may express four or more classes of K⁺ channels [57]. The expression and function of strong inward rectifier K⁺ (K_IR2.X) channels will be the focus of this review. Because ion channel expression and function change dramatically during cell culture [7,8,14,87,99], emphasis will be placed on data originating from intact microvessels and freshly isolated endothelial cells. Earlier literature may be accessed from prior reviews [87–89].

Structure of K_IR channels

Inward rectifier K⁺ channels in the K_IR2.1 – 2.4 family represent the products of four genes (loci: KCNJ2, KCNJ12, KCNJ4, KCNJ14, respectively) in the 15 member family of K_IR channels [49]. These channels consist of a tetramer of pore-forming α-subunits [49,86]. Each α-subunit has two membrane spanning domains (M1 and M2, Figure 1) with intracellular carboxy and amino termini [49,86]. A P-loop links the membrane spanning domains (Figure 1). The P-loop and M2 form the ion-conducting pore [49,86]. Conserved amino acids in the P-loop (T142-I143-G144-Y145-G146-F147-R148 in K_IR2.1) comprise the channels’ K⁺ ion selectivity filter [49,86,108] (Figure 1). Voltage-dependent block of outward K⁺ currents by intracellular polyamines [34,37,73] and Mg²⁺ [78,112] causes the distinguishing K_IR channel current inward rectification (Figure 2) [75]. Magnesium ions and polyamines interact with basic residues in M2 (D172 in K_IR2.1, D173 in K_IR2.2) and in the carboxy terminus (E224, E299 and D255 in K_IR2.1; E225, E300 and D256 in K_IR2.2) to produce voltage-dependent block of outward current flow at membrane potentials positive to the K⁺ equilibrium potential [75,108] (Figures 1 and 2). Based on studies of the crystal structure of chicken K_IR2.2 [108], K⁺ ions also will interact with these same residues, explaining why inward rectification shifts to more positive membrane potentials as extracellular K⁺ concentration is elevated.

Figure 1.

Structure of K_IR2 channels. Shown is a schematic representation of one K_IR2 channel subunit positioned in the lipid bilayer of a cell membrane as shown. These channels have two membrane spanning helical domains, denoted M1 and M2 connected by a P-loop that contains a helical domain (P in the drawing). The channel’s pore is formed by the P-loop and M2, with the selectivity filter sequence highlighted in the drawing. Cytosolic amino (NH₂) and carboxy (COOH) termini are also shown. Approximate locations of sites where regulatory molecules interact with the channel protein are shown as indicated. See text for references and more information.

Figure 2.

K_IR currents in arteriolar endothelial cells. Shown are mean ± SE (n = 5) current-voltage (I–V) relationships for Ba²⁺ -sensitive currents recorded from enzymatically isolated endothelial cell tubes (see [55] for more information). At membrane potentials (E_ms) negative to the K⁺ equilibrium potential (E_K), inward currents are carried by the channels (inward rectification). At membrane potentials positive to E_K, outward currents are carried by the channels, however, the outward currents diminish as the membrane becomes more positive producing a region of negative slope conductance, where hyperpolarization increases outward current flow through the channels. At normal resting E_m (5 mM K⁺ outside and 140 mM K⁺ inside, ~−30 mV), little current flows through these channels due to block by intracellular polyamines and Mg²⁺. However, membrane hyperpolarization will recruit current through the channels and amplify the initial hyperpolarization. Also shown is the shift in the I–V relationship with an increase in extracellular K⁺ (due to the shift in E_K). Note that at the resting level of E_m (−30 mV) with 15 mM K⁺ outside, there is now outward current through K_IR channels that will tend to hyperpolarize the membrane toward the new E_K. Data from [55] and graph modified from the same publication, with permission.

Negative-slope-conductance and the K_IR channel current-voltage relationship

Although K_IR channels are named for the inward rectification of current, it is the small outward “hump” in the current-voltage relationship (i.e., the region of negative slope conductance [29,98], Figure 2) that is present between the K⁺ equilibrium potential and resting membrane potential (~−40 to −30 mV [32,102,103,116]) of endothelial cells that is important for the physiological function of K_IR channels. Membrane hyperpolarization from the resting potential will increase outward K_IR channel current, amplifying the original hyperpolarization [12,55,58,72,104,106] (Figure 2). Microvascular endothelial cells have high membrane resistance at resting membrane potential (2–100 GΩ; [63,106,114]). Thus, activation of only a few K_IR channels can effectively modulate endothelial cell membrane potential; in mouse mesenteric arteries, there are on the order of only 144 functional K_IR channels per endothelial cell that significantly contribute to regulation of vessel function [106]. In K_IR channels containing K_IR2.2 subunits, steady-state membrane hyperpolarization decreases the channel’s open-state probability in a time-dependent fashion, a feature unique to this K_IR2 family member [49].

Membrane lipids modulate K_IR channels

The lipid environment around K_IR channels strongly regulates their function. Phosphatidylinositol 4,5-bisphosphate (PIP₂) activates K_IR2 channels by interacting with basic amino acids in M2 (K177, K178, R179 for K_IR2.1) and in the cytoplasmic tails of the channels (H53, R67, K187, K188, R189, K219, R228, and R312 for K_IR2.1) [49] (Figure 1). This interaction opens the potential for K_IR channel modulation by phospholipase-mediated hydrolysis of PIP₂ and its synthesis via kinases [49]. It is possible that activation of G_q/11-coupled receptors and the subsequent activation of phospholipase Cβ-mediated PIP₂ hydrolysis could inhibit K_IR2.2 and 2.3, because they have a relatively low PIP₂ affinity [28]. In contrast, K_IR2.1 affinity for PIP₂ binding is sufficiently high that phospholipase-induced PIP₂ hydrolysis may not remove activator PIP₂ from these channels [28]. Angiotensin II, vasopressin and GTPγS in the patch pipette inhibit K_IR currents in porcine cerebral capillary endothelial cells, in vitro [54]. These data suggest that microvascular endothelial cell K_IR channels may be physiologically modulated by activation of G-protein coupled receptors. However, it is not known if these effects were due to stimulation of PIP₂ hydrolysis.

Membrane cholesterol also modulates K_IR channel function. Increases in membrane cholesterol inhibit, while decreases in membrane cholesterol stimulate currents through K_IR2 channels in macrovascular endothelial cells [33,36,95–97] and in heterologous expression systems [94]. The hinge region of M1 (L85, V93, S95 in K_IR2.1) and M2 (I166, V167, I175, M183 in K_IR2.1) and the region between M1 and the cytosolic domains (L69, A70, V77 in K_IR2.1) are the sites where cholesterol interacts with the channels [96] (Figure 1).

K_IR channels reside in signaling complexes

K_IR2 channels reside in cholesterol-rich lipid rafts [94]. Loss of cholesterol results in channel movement out of these microdomains [109]. The caveolar protein, Caveolin-1, also inhibits K_IR2 channel function [45]. Inward rectifier K⁺ channels exist in membrane signaling microdomains in cardiac myocytes [67,117], where they interact with receptors, other ion channels, protein kinases, etc. [9]. Heterologously expressed K_IR2.1 channels interact with AKAP79 that targets these channels to protein complexes, which include PKA, calcineurin and other signaling proteins [22]. The carboxy terminus of K_IR2.1 contains the PDZ domain recognition sequence, (E424-S425-E426-I427; Figure 1), that interacts with postsynaptic density protein (PSD) 95 [84]. This adaptor protein, which is present in endothelial cells [110], interacts with AKAPS, nitric oxide synthase and other signaling proteins [11,110]. Thus, K_IR2.1 channels are likely located in signaling complexes in endothelial cells, providing significant potential for modulation of channel function, although this has not been explored in microvascular endothelial cells.

Protein kinases modulate K_IR channels

Studies in other systems suggest that there is considerable potential for K_IR channel regulation by kinases. There is a protein kinase A (PKA) consensus sequence in the C-terminus of K_IR2.1 (S425 in the PDZ domain in K_IR2.1; Figure 1) [119]. However, the effects of phosphorylation of this residue on K_IR channel function is far from clear. Phosphorylation of S425 by PKA inhibits inward K_IR currents at potentials more negative than the K⁺ equilibrium potential [113,118], but increases outward currents at more positive potentials [113] when the channels are expressed in COS cells. Activation of PKA inhibits native K_IR channels in the heart [113]. However, in contrast, inward K_IR2.1 channels expressed in Xenopus oocytes are activated by PKA [35], whereas native K_IR2.1 channels in cultured pulmonary endothelial cells are unaffected by activation of PKA [60]. Thus, how K_IR channels are modulated by PKA depends on the particular cell type in which the channels are expressed, and, most likely, also on the composition and configuration of the signaling complexes in which they are located. Regulation of microvascular endothelial cell K_IR channels by PKA has not been studied.

There is also a tyrosine kinase consensus sequence in K_IR2.1 at Y242 [118]. However, as with PKA, how phosphorylation of this site affects K_IR2.1 channel activity remains unclear with evidence for both decreased [118] and increased [123] activity reported. In addition, the tyrosine kinase Src inhibits currents through channels containing K_IR2.2 independent from protein kinase C (PKC), but does not inhibit homomeric K_IR2.1 channels [125]. In contrast, PKC mediates α_1A-adrenergic receptor-induced inhibition of K_IR2.3 channels [125]. Also, Gβγ-subunits inhibit K_IR2.3-containing channels [17].

Channels containing K_IR2.2 or 2.3 are inhibited by PKC [28,126], with no effect on K_IR2.1 channels in mammalian expression systems [28,48], or in cultured endothelial cells [60]. However, activation of β₃-adrenoreceptors activates K_IR2.1 channels through a mechanism involving PKC in Xenopus oocytes [100]. In vascular smooth muscle cells that express K_IR2.1 and K_IR2.2, hypoosmotic-induced cell swelling inhibits K_IR channels through a mechanism involving PKC [120].

There is also evidence for modulation of the expression and function of K_IR channels by other protein kinases. Calmodulin-dependent protein kinase II increases protein expression and activates currents through K_IR2.1 channels in cultured macrovascular endothelial cells [92]. Surface expression of K_IR2.1 is decreased via AMP kinase-dependent phosphorylation of the ubiquitin ligase, Nedd4-2 [3].

Other modulators of K_IR channels

Intracellular acidosis inhibits K_IR channels through effects on residues in the N-terminal domain of the channel [49]. In taste buds, acid-induced closure of K_IR2.1 contributes to acid-induced depolarization that is involved with the transduction of sour taste [121]. Whether pH modulates the function of endothelial K_IR channels has not been studied.

Fluid shear stress activates K_IR2.1 channels in macrovascular endothelial cells [53,68,90] and in heterologous expression systems through a mechanism that may involve a tyrosine kinase [53]. Whether a similar mechanism functions in microvascular endothelial cells remains to be established, but would be predicted to produce endothelial cell hyperpolarization and vasodilation. Shear stress also modulates transcription of K_IR channels in cultured human coronary endothelial cells, with K_IR2.2, 2.3 and 2.4 being upregulated by increased shear stress [61]. It is not known if altered shear stress produces similar changes in native microvascular endothelial cells. The impact of sheer stress-induced changes in endothelial K_IR channel expression on microvascular endothelial cell function also has not been established. However, such altered K_IR channel isoform expression would be predicted to modify the modulation of K_IR channel function by protein kinases, for example, potentially depressing or augmenting K_IR channel-mediated changes in endothelial cell membrane potential.

Increases in nitric oxide result in S-nitrosylation of a cysteine residue in the amino terminus of K_IR2.1 (C76) and an increase in K_IR channel activity in the heart and in heterologously expressed channels [40]. This may account for the increase in K_IR channel activity induced by NO in vascular smooth muscle K_IR channels [101].

K_IR channel pharmacology

Extracellular Ba²⁺ potently blocks K_IR channels in a voltage-dependent fashion [43]: at physiological membrane potentials (−30 to −40 mV) the K_d = 19–30 μM for K_IR2.1 [2,69]; for K_IR2.2, the K_d = 9 μM [69], and for K_IR2.3, the K_d = 70 μM [69]. Barium ions also block K_ATP channels (IC₅₀ = 100 μM [10]). Barium block of K_IR2.1 channels involves interactions with two residues, one in the outer vestibule of the channel (E125) and one just before the selectivity filter (T141) (Figure 1) [2]. In addition to Ba²⁺, extracellular Cs⁺ ions block K_IR channels [49], as they do all K⁺ channels.

Currents through K_IR2 channels are inhibited by ML133 (IC₅₀ = 1.9 μM for K_IR 2.1; 2.3 μM for K_IR2.2 and 4 μM for K_IR2.3) [115]. This compound blocks K_IR channel currents in endothelial cells from rat middle cerebral arteries [63], and in rat mesenteric artery endothelial cells [106]. Residues in M2 (D172 and I176 in K_IR2.1) are involved in the mechanism of action of ML133 [115]. This compound also blocks K_ATP channels composed of KIR6.2 subunits (IC₅₀ = 7.7 μM) [115], channels that are likely expressed in some microvascular endothelial cells [38]. Inward rectifier K⁺ channels are also non-specifically blocked by a number of drugs including: antihistamines such as diphenhydramine and mepyramine; the antimalarial, choloroquin; and the class Ia antiarrhythmic, quinidine [49].

Potassium ions are an important physiological agonist of K_IR channels [72], increasing K_IR channel conductance proportional to the square root of the extracellular K⁺ concentration [44,66,72,74,76,98]. Extracellular K⁺ interacts with a residue near the K⁺ selectivity filter (R148 in K_IR2.1) [49]. Furthermore, increases in extracellular K⁺ concentration, by moving the K⁺ equilibrium potential to more positive potentials, will move the negative slope conductance region to more depolarized potentials (Figure 2). This results from interactions of K⁺ ions with the negatively charged residues that bind Mg²⁺ and polyamines and which are responsible for inward rectification [108]. The K⁺-induced shift in the current-voltage relationship will recruit outward current through the channels, also promoting membrane hyperpolarization (Figure 2). Thus, as with vascular smooth muscle K_IR channels, endothelial cell K_IR channels can serve as sensitive sensors of extracellular K⁺ concentration, transducing small changes (e.g., 3–5 mM to 8–15 mM K⁺) into membrane hyperpolarization [72].

The antiarrhythmic drug, flecainide activates cardiac K_IR2.1 channels by interacting with a cysteine residue in the carboxy terminus of the channel (C311 in K_IR2.1; Figure 1), a residue that is not found in K_IR2.2 or 2.3 [13]. This activation results from decreased affinity of the channels for polyamines, like spermine, through an allosteric mechanism [13]. Other drugs, such as the antiarrhythmic, propafenone and the β-adrenergic receptor antagonist, timolol may activate K_IR2.1 channels through a similar mechanism [41].

Expression of K_IR channels in microvascular endothelial cells

Early studies of cultured cells suggested that microvascular endothelial cells might not express K_IR channels [1,87]. However, more recent investigation of freshly isolated microvascular endothelial cells have clearly shown Ba²⁺-sensitive K_IR channel currents [18,21,54,55,63,70,72,114] (Figure 2). Barium also blocks endothelial cell K_IR currents in rat superior mesenteric arteries, in situ [15].

Capillary endothelial cells reportedly express mRNA for K_IR2.1, 2.2 and possibly 2.3 [69,81]. Rat cremaster muscle arteriolar endothelial cells express protein for K_IR2.1, but not K_IR2.2 (Figure 3). Similar results were observed in mouse arteriolar endothelial cells [57]. Inward rectifier K⁺ currents are blocked by the selective K_IR2 inhibitor, ML 133 [115], in rat middle cerebral artery [63] and mouse 3^rd-order mesenteric artery [106] endothelial cells. Knockout of endothelial cell expression of K_IR2.1 suppresses Ba²⁺-sensitive K_IR currents in mouse 3^rd-order mesenteric artery [106] and brain capillary [70] endothelial cells. Taken together, these data indicate that K_IR2.1 is an essential subunit of the native endothelial K_IR channels expressed in rats and mice. The presence and functions of other K_IR2 channel subunits in microvascular endothelial cells remains to be established.

Figure 3.

Expression of K_IR2.1 in arteriolar endothelial cells – Shown are fluorescent micrographs of enzymatically isolated rat cremaster arteriolar smooth muscle and endothelial cells that were fixed in 4% paraformaldehyde, permeabilized with Triton X-100, and labeled with DAPI (all cell nuclei – Left panels), FITC-labeled primary antibody for α-smooth muscle actin to positively identify smooth muscle cells (1:1000 – Sigma, Middle Panels), and primary antibodies for (A) K_IR2.1 (1:200 - Alomone) and (B) K_IR2.2 (1:400 – Alomone) (Right Panels). The secondary antibodies for the K_IR channels were Texas-red-labeled donkey anti-rabbit (Jackson Immunoresearch). Bottom row of panels in A and B shows results in the absence of primary antibodies. Panel A shows that endothelial cells display K_IR2.1 immunoreactivity, consistent with expression of K_IR2.1, while staining of smooth muscle cells in the same preparation was weak. Panel B shows that K_IR2.2 immunoreactivity is not present in endothelial cells, but is robustly expressed in smooth muscle cells as a positive control. Data are representative of 3 experiments. Similar results were obtained for mouse arteriolar endothelial cells [55].

Functions of endothelial K_IR channels

Vascular smooth muscle K_IR channels amplify hyperpolarization induced by the activation of other K⁺ channels, and transduce small increases in extracellular K⁺ into cell hyperpolarization [59,62,72,85,104,122]. However, the function of endothelial K_IR channels is not as clear, based solely on studies utilizing rat and mouse mesenteric arteries where functional K_IR channels are confined to the endothelium [18,25,27,42,104,106,107]. Barium attenuates endothelium-dependent hyperpolarization and conducted vasodilation in rat mesenteric arteries in normotensive rats [42]. In addition, Ba²⁺ inhibits K⁺-induced hyperpolarization and vasodilation in these arteries [25,42]. Studies in mouse 3^rd-order mesenteric arteries demonstrate that endothelial cell K_IR channels amplify the effects of drugs that act via endothelial cell hyperpolarization [106]. Endothelial K_IR channels can also mediate K⁺-induced vasodilation in these vessels [106]. Importantly, Sonkusare et al. [106] showed that endothelial cell-selective knockout of K_IR2.1 abolished these effects, demonstrating the crucial role for K_IR2.1 in murine resistance artery endothelial cells as end-stage boosters of membrane hyperpolarization. These studies are consistent with a significant role for endothelial cell K_IR channels in the vasoreactivity of small mesenteric arteries.

In contrast, other studies in rat mesenteric arteries do not support a major role for endothelial cell K_IR channels. Takano, et al. [107] found that Ba²⁺ had no effect on conducted vasodilation induced by agents that hyperpolarized either endothelial cells or smooth muscle cells. Barium also had no effect on conducted dilation or K⁺-induced vasodilation in mesenteric arteries, in contrast to cerebral and coronary arteries that show robust expression of smooth muscle K_IR channels [104]. Thus, there is evidence both for and against a significant role for endothelial cell K_IR channels.

The level of agonist-induced activation of the smooth muscle could be one possible explanation that might reconcile the opposing views of the role-played by endothelial K_IR channels. Increased agonist-induced activation of smooth muscle has been shown to inhibit Ba²⁺-sensitive K⁺-induced dilation of rat mesenteric arteries [25]. However, Goto et al. [42] (who observed a role for endothelial K_IR channels), Takano et al. [107] and Smith et al. [104] (who both found no role for endothelial K_IR channels) used similar concentrations of phenylephrine (~1μM) to pre-constrict their rat mesenteric arteries, and yet observed disparate results. Thus, there does not appear to be a simple reason that can reconcile these opposing views.

Nonetheless, the disparate findings outlined in the preceding paragraphs do suggest that endothelial cell K_IR channel function may be modulated, allowing fine-tuning of vascular function to maintain homeostasis, as well as dysregulation of these channels in disease states. Consistent with this latter proposition, endothelium-dependent hyperpolarization and conducted vasodilation are depressed in mesenteric arteries from spontaneously hypertensive rats, and Ba²⁺ is without effect in these arteries suggesting that hypertension decreases endothelial cell K_IR channel function [42]. The mechanism responsible for the hypertension-induced K_IR channel dysfunction has not been determined.

Ischemia [91] and stress [71] also depress K_IR channel function in cerebral vascular smooth muscle cells. The stress-induced suppression of K_IR channel function results from glucocorticoid-mediated decreases in K_IR2.1 channel transcript levels and reduced expression of functional K_IR channels in the smooth muscle cells [71]. Diabetes alters retinal pericyte K_IR channel function through increased polyamine synthesis and increased inward rectification [79]. Mutations in KCNJ2 result in a multi-system disorder (Andersen-Tawil syndrome) that includes cardiac arrhythmias, periodic paralysis and dysmorphogenesis related to decreased K_IR channel function [111]. These data support the hypothesis that K_IR channels are targets for modulation during disease states, and that mutations that lead to altered K_IR channel function produce significant pathologies. The impact of diseases and mutations on microvascular endothelial cell K_IR channel function remains to be established.

In arterioles, endothelial cell K_IR channels also may be positioned to sense changes in extracellular K⁺ concentration in the restricted space between the endothelium and overlying smooth muscle cells, as has been hypothesized for smooth muscle K_IR channels [30] (Figure 4). As noted above, in mesenteric arteries, endothelial K_IR channels have been shown to mediate dilation induced by elevated extracellular K⁺ [25,42]. Thus, local increases in K⁺ produced by the activity of other endothelial cell or smooth muscle cell K⁺ channels also may activate endothelial cell K_IR channels (Figure 4). This provides another mechanism (in addition to activation by hyperpolarization), to amplify the activity of other K⁺ channels, in either the endothelium or smooth muscle layers, and could contribute to the regulation of endothelial cell membrane potential and other microvascular functions.

Figure 4.

Endothelial K_IR channels: Amplifiers and sensors of extracellular K⁺. Hyperpolarization of endothelial cells, due to Ca²⁺-dependent activation of endothelial cell IK_Ca and SK_Ca, or by activation of other endothelial cell K⁺ channels such as ATP-dependent K⁺ channels (K_ATP) can activate endothelial cell K_IR channels, amplifying the initial hyperpolarization in a positive feed-back manner. Conduction of this hyperpolarization to adjacent endothelial cells that are electrically coupled by gap junctions, can also recruit K_IR channels amplifying the hyperpolarization and promoting conduction of this electrical signal. Membrane hyperpolarization may promote Ca²⁺ entry into the endothelial cells through store-operated channels, although this is controversial. The resulting endothelial cell hyperpolarization can then be conducted, through myoendothelial gap junctions, to overlying smooth muscle cells, deactivating smooth muscle voltage-gated Ca²⁺ channels, reducing intracellular Ca²⁺ and promoting vasodilatation. Endothelial cell K⁺ channels also can be recruited by elevation of extracellular K⁺ concentration in their microenvironment by adjacent endothelial cell or smooth muscle cell K⁺ channels. This mechanism may allow endothelium-derived vasodilators, such as NO, prostacyclin (PGI₂) or epoxides of arachidonic acid (EETs), which act, in part, by activating smooth muscle K⁺ channels, to utilize endothelial cell K_IR channels in their mechanism of action. See text for more information. Figure redrawn and adapted from [56], with permission.

Capillary endothelial cell K_IR channels

Cells in excitable tissues like brain, heart and skeletal muscle rely on the opening of K⁺ channels and efflux of K⁺ ions to repolarize the cells during each action potential. During periods of increased activity, this results in accumulation of K⁺ in the interstitium that can routinely be on the order of 8–10 mM, more than sufficient to activate K_IR channels [72]. Expression of K_IR channels in capillary endothelial cells opens the exciting possibility that capillaries may be ideally positioned to sense these increases in extracellular K⁺, transducing this signal into hyperpolarization that could be transmitted, via gap junctions, to upstream arterioles to elicit vasodilation and increases in capillary blood flow to match the increased activity of the tissue (functional hyperemia) [72].

This hypothesis has recently been tested in mouse cerebral microcirculation. Using a novel ex vivo system composed of pressurized cerebral parenchymal arterioles with attached capillaries, Dabertrand and colleagues [21] have shown that microapplication of 10 mM K⁺ onto the capillaries results in vasodilation of the upstream arterioles that can be blocked by Ba²⁺ and which is absent in vessels isolated from endothelial cell K_IR2.1^−/− mice. Similarly, Longden et al, [70] demonstrated that application of 10 mM K⁺ to brain capillaries, in vivo, results in upstream arteriolar dilation and increased red blood cell flux through the capillaries. Whether similar mechanisms operate in heart and skeletal muscle remains to be established. Coronary capillary endothelial cells display robust K_IR channel currents and express mRNA for K_IR channels [69,114] suggesting a potential role in the heart. In skeletal muscle, K_IR channels have been implicated in both the rapid-onset of vasodilation and steady-state increases in blood flow associated with muscle contraction [4,19,20]. However, the location of the K_IR channels (endothelium, smooth muscle, skeletal muscle fibers, etc.) that mediate skeletal muscle functional hyperemia is not known. While, conduction of signals from capillaries to arterioles has long been proposed [5,24,65,105,124], and conduction of electrical activity from capillaries to arterioles has been observed [5], Dabertrand and colleagues’ study is the first to define the ion channel that mediates such a response and to exclude other cell types for initiation of signal transmission from capillaries to arterioles.

Conclusions

While K_IR channel expression in endothelial cells has been known for some time, the physiological function of these ion channels is just beginning to be unraveled. Current evidence suggests that these K⁺ channels serve to boost hyperpolarization induced by opening of other K⁺ channels or transporters (such as the Na⁺/K⁺ ATPase), and to transduce changes in extracellular K⁺ concentration into endothelial cell membrane hyperpolarization. The hyperpolarization booster function of K_IR channels will be particularly effective for small hyperpolarizations due to opening of only few K⁺ channels, for example. Large hyperpolarizations, that drive the membrane potential close to the K⁺ equilibrium potential on their own, will be little affected by the recruitment of current through K_IR channels, because of the shape of the K_IR channel current-voltage relationship (Figure 2).

These functions position endothelial cell K_IR channels (along with their smooth muscle counterparts) to be involved in the mechanism of action of all hyperpolarizing vasodilators, with modulation of K_IR channel function providing an additional mechanism to fine tune the reactivity of arterioles in the microcirculation. Endothelial K_IR channels also may contribute to conduction of hyperpolarization along the wall of microvessels, participating in the coordination of local blood flow regulation. The expression of K_IR channels in capillary endothelial cells positions these ion channels to play a major role in sensing extracellular K⁺ and contributing to functional hyperemia in electrically active tissues such as the brain, heart and skeletal muscle.

How the expression and function of native microvascular endothelial cell K_IR channels is modulated during cell signaling processes in health and disease remains to be established. However, based on studies in other systems, it seems likely that microvascular endothelial cell K_IR channels also may be modulated in diseases like hypertension, diabetes, and stress. It also is likely that endothelial K_IR channels reside in macromolecular signaling complexes. However, knowledge of the protein partners with which these channels interact is simply lacking. At a more fundamental level, our knowledge and understanding of the expression and function of the full complement of ion channels in microvascular endothelial cells, smooth muscle cells and pericytes remains in its infancy. Molecular approaches need to be applied to define the complete ion channel transcriptome and proteome in native, not cultured microvascular cells, in multiple vascular beds. For example, we know much about the function of endothelial K_IR channels from the study of mesenteric arteries. However, how these findings apply to other vascular beds, is simply not known. Functional assays of ion channel function using patch-clamp approaches need to be applied so that the actual currents, in the cells of interest, which are responsible for a physiological response, are identified and characterized to verify what is often inferred from the application of pharmacology to complex systems. This will mean developing new, or adapting old techniques to isolate microvascular cells from the particular vessels of interest, including capillaries. The use of cell specific and hopefully, conditional knockouts of K_IR channels should help delineate the function of K_IR channels in specific microvascular cells. However, given that these ion channels are likely part of much larger signaling complexes, such knockouts have the potential to disrupt more than just the function of a single ion channel. Thus, multiple approaches, including careful pharmacology, must be applied, rather than relying on any single strategy. Understanding the control of expression and function of microvascular endothelial cell K_IR channels will improve our understanding of local blood flow control in health and disease, and may, in the future, provide new targets for the development of therapeutics directed at these important ion channels.