Inward rectification and vascular function: As it was in the beginning

A perspective on the classic papers of Edwards & Hirst (1988) and Edwards et al. (1988)

Potassium channels are known to play an important role in the regulation of vascular tone through their ability to hyperpolarize smooth muscle cells and produce vasodilatation, primarily by causing the closure of voltage-dependent calcium channels. Of the four different classes of potassium channels found in blood vessels (ATP-sensitive potassium channels, inwardly rectifying potassium channels (K_IR), voltage-activated potassium channels and calcium-activated potassium channels), K_IR channels have been implicated in a range of vasomotor functions, from vasodilatation of cerebral, coronary and skeletal muscle vascular beds in response to local release of potassium ions (K⁺; for example, see Quayle et al. 1997; Armstrong et al. 2007) to augmentation and facilitation of the propagation of vasodilatory responses over distance along microcirculatory arterioles and feed arteries (Rivers et al. 2001; Goto et al. 2004; Jantzi et al. 2006). Finally, down-regulation or inactivation of these channels is increasingly seen as instrumental in the increased excitability of vascular smooth muscle found in pathological conditions such as hypertension, hypercholesterolaemia and diabetes (Chrissobolis & Sobey, 2003; Jackson, 2006; Matsushita & Puro, 2006).

K_IR channels are activated by hyperpolarization and most readily pass inward currents, selective for K⁺, at potentials negative to the equilibrium potential for K⁺ (E_K). However, at membrane potentials between the activation voltage and E_K, K_IR channels sustain small outward currents. Thus, although the channel name refers to the greater ability to pass inward than outward current, the involvement of K_IR channels in vasomotor function centres on the smaller outward currents activated at membrane potentials between about −40 and −80 mV. K_IR channels are blocked by low concentrations of barium ions (Ba²⁺) and most importantly, their activation potential is shifted to more positive potentials by increases in extracellular potassium concentration ([K⁺]_o) and to more negative potentials by decreases in [K⁺]_o (Quayle et al. 1997).

Over 60 years ago, it was known that small increases in [K⁺]_o could initiate vasodilatation, although the effect depended on the vascular bed, the level of vascular tone and the concentration of KCl (Dawes, 1941). While the underlying mechanism was unknown at the time, Dawes suggested that this phenomenon could explain the increased blood flow seen in contracting muscle, particularly if sufficient K⁺ were released during muscular activity. Subsequent studies confirmed the link between small increases in perivascular [K⁺] and vascular tone in other vascular beds, including the cerebral circulation (Kuschinsky et al. 1972), underscoring the importance of activity-dependent alterations in [K⁺]_o in tuning blood supply to organ activity.

As potassium channels make a significant contribution to resting membrane potential, an increase in [K⁺]_o would be expected to shift E_K to more positive potentials and thus lead to depolarization and constriction. The fact that small increases in [K⁺]_o produced hyperpolarization and vasodilatation in certain arteries was seen as somewhat of a paradox. In 1988, although K_IR channels had been described in skeletal muscle fibres, starfish eggs and neurones of the olfactory cortex (Katz, 1949; Hagiwara & Takahashi, 1974; Constanti & Galvan, 1983), they had not been described in arterioles, nor were they considered to be activated at membrane potentials more positive to E_K. In earlier studies, Hirst & Neild (1978) had detected rectification in segments of guinea pig submucosal arterioles with resistances falling with membrane hyperpolarization, while Hirst & van Helden (1982) had observed that a reduction in [K⁺]_o resulted in depolarization and an increase in membrane resistance of similar arteriolar segments. Thus the aim of the first study by Edwards & Hirst in 1988 was to determine whether this hyperpolarization-induced rectification and the paradoxical effects of changing [K⁺]_o could be attributed to a K⁺-selective inward rectifier.

Experiments were carried out on short isopotential segments of arterioles in which it was possible to investigate the current–voltage relationships in either voltage or current clamp mode. In these short unpressurized segments the resting membrane potential was −70 mV. The results demonstrated membrane rectification, with the membrane resistance starting to fall at an activation voltage of around −50 mV and membrane conductance increasing with hyperpolarization. The rectifying current was blocked by Ba²⁺ and showed substantial changes in activation potential with only small changes in [K⁺]_o, i.e. from 1.25 to 10 mm. More importantly, Edwards & Hirst (1988) showed that the activity of this channel could impact on arteriolar excitability, the very aspect now considered in pathological studies to be crucial. Thus, the time course of action potentials initiated from hyperpolarized potentials, at which K_IR channels would be active, were briefer than those initiated from more depolarized potentials. Similar effects were observed when the membrane potential was held at −70 mV and [K⁺]_o altered; time courses were shorter in 10 mm K⁺ (more active K_IR) and longer in 2.5 mm K⁺ (less active K_IR). The effect of the K_IR current in attenuating the decay of excitatory junction potentials following nerve stimulation could also be seen with summed excitatory junction potentials (more depolarized, less active K_IR) having larger time constants than single excitatory junction potentials (more hyperpolarized, more active K_IR).

In a back-to-back paper, Hirst and colleagues investigated the role of K_IR channels in cerebral arterioles (Edwards et al. 1988). Again current–voltage relationships in short arterial segments and the effect of alterations in [K⁺]_o, from a control value of 5 mm, were tested. Both arteriolar segments located close to the middle cerebral as well as those located at some distance from this larger vessel were chosen, as previous studies had demonstrated loss of sympathetic fibres with increasing distance from the middle cerebral artery (Hill et al. 1986) and neuronal activity was considered to be capable of producing appropriate changes in [K⁺]_o. Interestingly, the results showed that distal arterioles were more excitable than proximal arterioles, had more depolarized membrane potentials, often with superimposed oscillations, and little evidence of inward rectification. Moreover, regional differences in responses to increases in [K⁺]_o were found; distal segments hyperpolarized and now demonstrated inward rectification, while proximal segments depolarized and showed inward rectification at more positive potentials. Since Ba²⁺ abolished inward rectification of proximal segments, and of distal segments exposed to elevated [K⁺]_o, it was clear that K_IR channels were expressed in both areas but that these channels differed in their sensitivity to [K⁺]_o, i.e. higher concentrations were required to activate K_IR channels in distal than in proximal segments.

In both studies, inhibition of the K_IR channels with Ba²⁺ resulted in a large depolarization of submucous arterioles and proximal cerebral arterioles leading the authors to conclude that K_IR channels constituted the predominant resting potassium conductance. However, the resting membrane potential detected in these early studies was around −70 mV, representing recordings from unpressurized vessels. This value is considerably more hyperpolarized than that measured subsequently in vivo in systemic vessels at around −30 to −40 mV (Welsh & Segal, 1998; Emerson & Segal, 2000; Siegl et al. 2005). Although K_IR channels would be expected to be largely closed at these more physiological potentials, low concentrations of Ba²⁺ which selectively inhibit K_IR channels (Quayle et al. 1997) have been reported to reduce the diameter of skeletal muscle arterioles (Loeb et al. 2000) and large cerebral arteries in vivo and of other vessels in vitro (Chrissobolis & Sobey, 2003), suggesting perhaps a minor contribution at rest. Such a contribution seems especially surprising in the cerebral vessels since the [K⁺]_o of cerebrospinal fluid is around 3 mm and at this concentration the activation potential of the K_IR channels would be expected to be shifted to more negative potentials.

Irrespective of whether K_IR channels contribute to resting vascular tone, these two classical papers provided proof that these channels could influence vascular reactivity, especially in situations where [K⁺]_o was increased. They were the first to describe the existence of a potassium-sensitive inward rectifier in both cerebral and systemic arterioles, demonstrate the activation of this current at potentials more positive than E_K, show how the activation potential could be moved to more depolarized potentials by small increases in [K⁺]_o and illustrate how this current could modify arteriolar excitability. However, the original studies failed to take into account the fact that the preparations contained both smooth muscle and endothelial cells and so the precise cellular location of the channels was not determined. Although subsequent studies confirmed the existence of K_IR currents in smooth muscle cells isolated from cerebral arteries (Quayle et al. 1993), others have shown that K_IR channels are also expressed in isolated endothelial cells (Nilius et al. 1997) and in intact rat mesenteric arteries, they may even be confined to this layer (Doughty et al. 2001; Crane et al. 2003; Goto et al. 2004).

Such differential cellular expression of K_IR channels may be a characteristic of different vascular beds, with vessel size perhaps contributing to further variation. Given that heterocellular coupling has been reported to play a critical role in vasodilatation evoked by endothelial-derived hyperpolarization and that the incidence of myoendothelial gap junctions varies along and between vascular beds (Sandow & Hill, 2000; Sandow et al. 2002, 2003), the specific cellular location of K_IR channels may differentially affect smooth muscle contractility. Indeed, a recent study has shown that K_IR channels expressed in vascular smooth muscle cells of cerebral or coronary arteries act as electrical amplifiers of hyperpolarizing responses activated in either the endothelium or the smooth muscle (Smith et al. 2007). However, this was not the case in mesenteric arteries in which K_IR channels were not highly expressed in the smooth muscle cells (Smith et al. 2007). If these vessels are analogous to those evaluated in previous studies of the mesenteric circulation, then one would expect the endothelium to express K_IR channels. In this case, the physiological role of these endothelial channels warrants further investigation.

The pioneering studies of Hirst and colleagues on arteriolar K_IR channels raised other questions which remain topics for future research. For example, what is the significance of the regional variation in the sensitivity of the K_IR channels to [K⁺]_o? Interestingly, pericytes surrounding microvessels in the rat retina have also been shown to display a topographical heterogeneity in K_IR currents (Matsushita & Puro, 2006), while studies in the spiral modiolar artery of the ear have described a heterogeneously coupled population of smooth muscle cells with a bimodal distribution of membrane potentials resulting from the continuous activation or not of K_IR channels (Jiang et al. 2001). In the latter study, the two cell populations clearly differed in their sensitivity to [K⁺]_o in a manner analogous to that of the proximal and distal arteriolar cells. In the cerebral circulation, the distal arteriolar segments which demonstrated the less sensitive K_IR channels lacked a sympathetic innervation (Hill et al. 1986) leading the authors to suggest a possible trophic influence of the nerves on muscle properties (Edwards et al. 1988). It is not known whether such correlations exist in other vascular beds. Nevertheless, the end result is that the proximal cerebral vessels will hyperpolarize and relax to smaller increases in [K⁺]_o than will the distal vessels, a characteristic conducive to producing an effective increase in blood flow.

Future studies will need to address in more detail the differential regional and cellular expression of K_IR channels throughout the vascular system and the contribution of these channels to arteriolar excitability in vivo. The factors which might regulate the expression and biophysical characteristics of K_IR channels are also of great interest, particularly if these may vary with cell type or pathophysiological condition. In this respect it should be borne in mind that even a small change in the magnitude of fluctuations in perivascular [K⁺]_o could a have profound effect on the ability of K_IR channels to influence vascular excitability.