Hypoxia reduces K_Ca channel activity by inducing Ca²⁺ spark uncoupling in cerebral artery smooth muscle cells

Abstract

Arterial smooth muscle cell large-conductance Ca²⁺-activated potassium (K_Ca) channels have been implicated in modulating hypoxic dilation of systemic arteries, although this is controversial. K_Ca channel activity in arterial smooth muscle cells is controlled by localized intracellular Ca²⁺ transients, termed Ca²⁺ sparks, but hypoxic regulation of Ca²⁺ sparks and K_Ca channel activation by Ca²⁺ sparks has not been investigated. We report here that in voltage-clamped (−40 mV) cerebral artery smooth muscle cells, a reduction in dissolved O₂ partial pressure from 150 to 15 mmHg reversibly decreased Ca²⁺ spark-induced transient K_Ca current frequency and amplitude to 61% and 76% of control, respectively. In contrast, hypoxia did not alter Ca²⁺ spark frequency, amplitude, global intracellular Ca²⁺ concentration, or sarcoplasmic reticulum Ca²⁺ load. Hypoxia reduced transient K_Ca current frequency by decreasing the percentage of Ca²⁺ sparks that activated a transient K_Ca current from 89% to 63%. Hypoxia reduced transient K_Ca current amplitude by attenuating the amplitude relationship between Ca²⁺ sparks that remained coupled and the evoked transient K_Ca currents. Consistent with these data, in inside-out patches at −40 mV hypoxia reduced K_Ca channel apparent Ca²⁺ sensitivity and increased the K_d for Ca²⁺ from ∼17 to 32 μM, but did not alter single-channel amplitude. In summary, data indicate that hypoxia reduces K_Ca channel apparent Ca²⁺ sensitivity via a mechanism that is independent of cytosolic signaling messengers, and this leads to uncoupling of K_Ca channels from Ca²⁺ sparks. Transient K_Ca current inhibition due to uncoupling would oppose hypoxic cerebrovascular dilation.

CHANGES IN OXYGEN (O₂) partial pressure (P_O2) modulate tissue perfusion and regional blood flow, but reactivity varies depending on the anatomic origin of the vasculature (9, 23, 25, 30). In general, hypoxia constricts small distal pulmonary arteries and arterioles but dilates small systemic arteries and arterioles, including those in the brain (2, 25, 31). Hypoxic pulmonary vasoconstriction acts to divert blood supply to O₂-rich regions of the lung. In contrast, in the systemic vasculature hypoxic vasodilation functions to match O₂ supply to metabolic demand. Although O₂ regulation of vascular reactivity serves diverse physiological roles, mechanisms that mediate these responses are poorly understood.

Multiple signaling messengers have been proposed to mediate hypoxia-induced systemic artery dilation, including endothelium-dependent nitric oxide, cyclooxygenase products, adenosine 3′,5′-cyclic monophosphate (cAMP), and guanosine 3′,5′-cyclic monophosphate (cGMP) (12, 13, 25). Activation of potassium channels, including ATP-sensitive and large-conductance calcium (Ca²⁺)-activated potassium (K_Ca) channels, is also proposed to contribute to hypoxic vasodilation (3, 4, 14, 16, 18, 35). However, there are conflicting reports of arterial smooth muscle cell K_Ca channel regulation by hypoxia, with studies reporting activation, inhibition, or no modulation (3, 16, 25, 35, 41, 45). Thus the role of K_Ca channels in hypoxic vasodilation is unclear.

In arterial smooth muscle cells, K_Ca channels are activated by localized intracellular Ca²⁺ transients, termed Ca²⁺ sparks, which occur because of the opening of ryanodine-sensitive Ca²⁺ release (RyR) channels in the sarcoplasmic reticulum (SR) membrane (21, 36). Ca²⁺ sparks generate the micromolar subsarcolemmal intracellular Ca²⁺ concentration ([Ca²⁺]_i) elevation necessary for K_Ca channel activation and therefore are critical modulators of K_Ca channel activity (39, 51). Because of their rapid and localized temporal and spatial properties, Ca²⁺ sparks do not contribute directly to [Ca²⁺]_i (21). Ca²⁺ spark-induced K_Ca channel activation causes membrane hyperpolarization, leading to a decrease in voltage-dependent Ca²⁺ influx, a reduction in global [Ca²⁺]_i, and vasodilation. Conversely, Ca²⁺ spark inhibition or a reduction in the coupling of Ca²⁺ sparks to K_Ca channels results in vasoconstriction (21, 36).

O₂ regulation of K_Ca channels in arterial smooth muscle cells has primarily been studied by measuring single-channel activity or whole cell currents (9, 16, 20, 38). Thus hypoxic regulation of Ca²⁺ sparks and K_Ca channels that are under Ca²⁺ spark control is unclear. Since Ca²⁺ sparks are a principal regulator of K_Ca channel activity in cerebral artery smooth muscle cells, the present study was undertaken to study hypoxic regulation of Ca²⁺ sparks, K_Ca channels, and the coupling relationship between Ca²⁺ sparks and K_Ca channels. Our data indicate that in cerebral artery smooth muscle cells hypoxia reduces the apparent micromolar Ca²⁺ sensitivity of K_Ca channels, which leads to Ca²⁺ spark to K_Ca channel uncoupling and a decrease in the frequency and amplitude of transient K_Ca currents. In contrast, in voltage-clamped cells, hypoxia does not alter Ca²⁺ spark frequency or amplitude or global [Ca²⁺]_i. These data suggest that, in response to hypoxia, uncoupling of K_Ca channels from Ca²⁺ sparks would oppose the cerebral artery dilation.

MATERIALS AND METHODS

Arterial smooth muscle cell isolation

Individual smooth muscle cells were enzymatically dissociated as previously described (7, 49, 50). Briefly, Sprague-Dawley rats (∼200−250 g) of either sex were anesthetized by an intraperitoneally injected overdose of pentobarbital sodium (150 mg/kg body wt). Animal protocols used were reviewed and approved by the Animal Care and Use Committee of the University of Tennessee. The brain was then removed and placed into ice-cold (4°C) HEPES-buffered physiological salt solution containing (in mM) 134 NaCl, 6 KCl, 2 CaCl₂, 1 MgCl₂, 10 HEPES, and 10 glucose (pH 7.4 with NaOH). Posterior cerebral, middle cerebral, and cerebellar arteries (100−200 μm in diameter) were removed and cleaned of connective tissue. Individual smooth muscle cells were dissociated from arteries with a HEPES-buffered isolation solution containing (in mM) 55 NaCl, 80 sodium glutamate, 5.6 KCl, 2 MgCl₂, 10 HEPES, and 10 glucose (pH 7.3 with NaOH), which was supplemented with papain (0.7 mg/ml) and collagenase (1.0 mg/ml), as described previously (19). Smooth muscle cells were maintained in ice-cold (4°C) HEPES-buffered isolation solution and used for experiments between 1 and 8 h after isolation.

Patch-clamp electrophysiology

Potassium currents were measured with the perforated-patch or the excised inside-out patch-clamp configuration (Axopatch 200B, Clampex 8.2). For perforated-patch experiments, HEPES-buffered physiological salt solution was used as the bath solution. The pipette solution contained (in mM) 110 KAsp, 30 KCl, 10 NaCl, 1 MgCl₂, 10 HEPES, and 0.05 EGTA (pH 7.2, KOH). For inside-out patch recordings, the bath solution contained (in mM) 130 KCl, 10 HEPES, 1 MgCl₂, 5 EGTA, and 1.6 HEDTA (pH 7.2 with KOH), with free Ca²⁺ concentrations of 1, 3, 10, 30, 100, or 300 μM. The pipette solution for inside-out patch experiments contained (in mM) 130 KCl, 10 HEPES, 1 MgCl₂, 5 EGTA, and 1.6 HEDTA, with 10 μM free Ca²⁺ (pH 7.4 with KOH). Free Ca²⁺ concentrations were measured with a Ca²⁺-sensitive (Corning no. 476041) and a reference (Corning no. 476370) electrode. Hypoxic solutions were obtained by purging bath solution with 100% N₂ in a gas-impermeant container for at least 1 h before use. Experimental chambers were continuously perfused with normoxic or hypoxic solution at a rate of 5−10 ml/min. Dissolved P_O2 was monitored in experimental chambers with an O₂-sensitive electrode (Extech Instruments). Changing the perfusion solution from normoxic to hypoxic reduced the dissolved P_O2 in the chamber from ∼150 to 15 mmHg. K⁺ currents were filtered at 1 kHz and digitized at 4 kHz. K_Ca current analysis was performed off-line with custom analysis software or Clampfit 9.2. A transient K_Ca current was defined as the simultaneous opening of three K_Ca channels, as previously defined (7, 27). Single-K_Ca channel amplitude was measured in normoxia and hypoxia with histograms.

Confocal Ca²⁺ imaging

Cells were incubated in HEPES-buffered isolation solution and fluo-4 AM (10 μM) for 25 min at room temperature, followed by a 30-min wash. Imaging was performed with HEPES-buffered physiological salt solution in the experimental chamber. Fluo-4 fluorescence was imaged with a Noran Oz laser scanning confocal microscope with a ×60 water-immersion objective (numerical aperture = 1.2) by illuminating with 488-nm light and collecting emitted light >500 nm. Images (256 × 240 pixels, 56.3 × 52.8 μm) were recorded every 8.3 ms (i.e., at 120 images/s). Simultaneous current and fluorescence measurements were synchronized with a light-emitting diode placed above the recording chamber that was triggered during acquisition. Each cell was imaged for at least 10 s under each condition. Ca²⁺ sparks and global Ca²⁺ concentration were analyzed off-line with custom software written with IDL 5.3 that was a kind gift from Dr. M. T. Nelson (University of Vermont, Burlington, VT). Ca²⁺ sparks were detected by dividing off an area 1.54 μm (7 pixels) × 1.54 μm (7 pixels) (i.e., 2.37 μm²) in each image (F) by a baseline (F₀) that was determined by averaging 10 images without Ca²⁺ spark activity. The entire area of each image was analyzed to detect Ca²⁺ sparks. A Ca²⁺ spark was defined as a local increase in F/F₀ > 1.2. Global Ca²⁺ fluorescence was calculated from the same images used for Ca²⁺ spark analysis and was the mean pixel value of 100 different images acquired during a 10-s period (7, 8, 49, 50).

Fura-2 imaging

Isolated smooth muscle cells were incubated in HEPES-buffered isolation solution containing fura-2 AM (5 μM) and 0.05% Pluronic F-127 for 20 min, followed by a 15-min wash. Experiments were performed with HEPES-buffered physiological salt solution in the chamber. Fura-2 was alternately excited at 340 or 380 nm with a PC-driven hyperswitch (Ionoptix, Milton, MA). Background-corrected ratios were collected every 1 s at 510 nm with a Dage MTI integrating CCD camera (Ionoptix). SR Ca²⁺ load ([Ca²⁺]_SR) was estimated by measuring the amplitude of caffeine (10 mM)-induced [Ca²⁺]_i transients (7, 8, 49, 50).

Reagents

Unless otherwise specified, all reagents were purchased from Sigma-Aldrich (St. Louis, MO). Fluo-4 AM, fura-2 AM, and Pluronic F-127 were purchased from Molecular Probes (Eugene, OR) and papain from Worthington Biochemical (Lakewood, NJ).

Statistical analysis

Values are expressed as means ± SE. Student's t-test and Student-Newman-Keuls test were used for comparing paired or unpaired data and multiple data sets, respectively. Simultaneous Ca²⁺ spark and transient K_Ca current amplitude data were fit with a linear regression function and the slope ± SE of each fit was compared with a Student's t-test. The relationship between K_Ca channel open probability (P_o) and free Ca²⁺ concentration was fit with a Hill equation, y = V_max * x_n/ (x_n + k_n), where V_max is the maximal P_o of K_Ca; n is the Hill coefficient (n_H), and k is the dissociation constant (K_d). V_max, n_H, and K_d were compared between normoxia and hypoxia with a Student's t-test. P < 0.05 was considered significant.

RESULTS

Hypoxia inhibits transient K_Ca currents in cerebral artery smooth muscle cells

Transient K_Ca currents were measured in isolated smooth muscle cells using the perforated-patch clamp configuration. At −40 mV, a reduction in dissolved P_O2 from ∼150 (normoxia) to 15 (hypoxia) mmHg reversibly reduced the frequency and amplitude of transient K_Ca currents (Fig. 1). Specifically, hypoxia decreased mean transient K_Ca current frequency from 0.79 ± 0.16 to 0.48 ± 0.11 Hz, or to ∼61% of normoxia (n = 10 cells, P < 0.05). Hypoxia reduced mean transient K_Ca current amplitude from 33.0 ± 6.3 to 25.1 ± 5.7 pA, or to ∼76% of normoxia (n = 10 cells, P < 0.05).

Fig. 1.

Hypoxia inhibits transient Ca²⁺-activated potassium (K_Ca) currents in voltage-clamped cerebral artery smooth muscle cells. A: original trace illustrating reversible transient K_Ca current inhibition by a reduction in dissolved O₂ from ∼150 (normoxia) to ∼15 (hypoxia) mmHg. Expanded segments are shown below the full trace to illustrate individual transient K_Ca currents in normoxia and hypoxia. B and C: hypoxia reduces mean transient K_Ca current frequency (B) and amplitude (C) (n = 10 for each). *P < 0.05 compared with normoxia.

Hypoxia attenuates the coupling ratio and amplitude relationship between Ca²⁺ sparks and transient K_Ca currents

To investigate mechanisms mediating transient K_Ca current inhibition by hypoxia, simultaneous measurements of Ca²⁺ sparks and transient K_Ca currents were obtained by performing confocal Ca²⁺ imaging in combination with patch-clamp electrophysiology. At −40 mV, in normoxia, 89.1 ± 3.5% of Ca²⁺ sparks activated a transient K_Ca current (n = 155 sparks, 16 cells). In the same cells, hypoxia reduced Ca²⁺ spark coupling to 62.6 ± 9.1% (n = 123 sparks). Hypoxia also reduced the slope of the amplitude relationship between coupled Ca²⁺ sparks and evoked transient K_Ca currents from 66 ± 4 to 45 ± 5 pA/F/F₀ (P < 0.05, Fig. 2B). In contrast, hypoxia did not alter mean Ca²⁺ spark frequency (normoxia 0.74 ± 0.13 Hz, hypoxia 0.76 ± 0.17 Hz; n = 16 cells, P > 0.05) or amplitude (F/F₀, 1.70 ± 0.05 vs. 1.72 ± 0.05; n = 155 and 123 sparks, respectively, P > 0.05) (Fig. 2, C and D). In hypoxia, global fluo-4 fluorescence was 1.03 ± 0.04 that in normoxia, indicating that global Ca²⁺ did not change (n = 16 cells, P > 0.05).

Fig. 2.

Hypoxic regulation of Ca²⁺ sparks and the effective coupling of Ca²⁺ sparks to K_Ca channels. A: original simultaneous recordings of transient K_Ca currents (top) and Ca²⁺ sparks that occurred at 2 locations (bottom) in the same cell voltage-clamped at −40 mV in normoxia (left) and hypoxia (right). * and #, coupled and uncoupled Ca²⁺ sparks, respectively. B: amplitude correlation of coupled Ca²⁺ sparks and evoked transient K_Ca currents in normoxia (black squares; n = 134) and hypoxia (gray circles; n = 86) with linear regression and 95% confidence bands fit to each data set (slope ± SE: normoxia 66 ± 4, hypoxia 45 ± 5). Hypoxia decreased the effective coupling of Ca²⁺ sparks to K_Ca channels (P = 0.0037). C and D: hypoxia did not change mean Ca²⁺ spark frequency (C) or amplitude (D) (n = 155 and 123 sparks in normoxia and hypoxia, respectively; n = 16 cells).

Consistent with the observation that hypoxia did not alter Ca²⁺ spark frequency, hypoxia also did not change [Ca²⁺]_SR, as determined by measuring caffeine (10 mM)-induced [Ca²⁺]_i transients in isolated smooth muscle cells (Fig. 3). Collectively, these data suggest that in cerebral artery smooth muscle cells hypoxia reduces the effective coupling of Ca²⁺ sparks to K_Ca channels but does not alter Ca²⁺ spark properties or global [Ca²⁺]_i.

Fig. 3.

Hypoxia does not alter cerebral artery smooth muscle cell sarcoplasmic reticulum (SR) Ca²⁺ load. A: caffeine (10 mM)-induced intracellular Ca²⁺ concentration ([Ca²⁺]_i) transients were similar in normoxia and hypoxia in an isolated cerebral artery smooth muscle cell. 340/380, 340- to 380-nm ratio. B: mean change in fura-2 ratio in normoxia and hypoxia (n = 6 cells).

Hypoxia decreases the apparent Ca²⁺ sensitivity of cerebral artery smooth muscle cell K_Ca channels

To determine mechanisms mediating hypoxia-induced Ca²⁺ spark uncoupling and attenuation of the amplitude relationship between Ca²⁺ sparks and K_Ca channels, K_Ca channel activity was measured in cerebral artery smooth muscle cells. To block transient K_Ca currents, Ca²⁺ sparks were inhibited with thapsigargin (100 nM), which depletes [Ca²⁺]_SR.K_Ca channel activity was measured with the perforated-patch clamp configuration with myocytes voltage-clamped at 0 mV to improve the signal-to-noise ratio. In the same smooth muscle cells, hypoxia reduced mean K_Ca channel activity (NP_o) from 0.18 ± 0.03 to 0.08 ± 0.03, or to ∼44% of that in normoxia (n = 7 cells, P < 0.05; Fig. 4). In contrast, hypoxia did not alter K_Ca channel amplitude (normoxia 5.2 ± 0.3 pA, hypoxia 5.2 ± 0.4 pA; n = 7 cells, P > 0.05).

Fig. 4.

Hypoxia inhibits K_Ca channels in rat cerebral artery smooth muscle cells. A: original traces obtained from the same smooth muscle cell illustrating reversible K_Ca channel inhibition by hypoxia at 0 mV. Transient K_Ca currents were blocked by pretreatment with thapsigargin (100 nM), to deplete SR Ca²⁺. B: mean data illustrating hypoxic inhibition of K_Ca channel activity (NP_o, n = 7 cells). *P < 0.05 compared with normoxia.

We then tested the hypothesis that hypoxia inhibits transient K_Ca currents by decreasing K_Ca channel apparent Ca²⁺ sensitivity. We also sought to determine whether K_Ca channel inhibition by hypoxia requires cytosolic signaling pathways. To test these hypotheses, cerebral artery smooth muscle cell K_Ca channel open probability (P_o) was measured over a range of intracellular free Ca²⁺ concentrations from 1 to 300 μM in excised inside-out membrane patches voltage-clamped at −40 mV. Hypoxia increased the mean K_d for Ca²⁺ of K_Ca channels from 17.3 ± 1.3 to 31.9 ± 0.81 μM without altering n_H or the maximum P_o (Fig. 5A and B). Relative K_Ca channel inhibition by hypoxia was also Ca²⁺ dependent. With 3, 10, and 30 μM free Ca²⁺, which were the Ca²⁺ concentrations where inhibition was greatest, hypoxia decreased K_Ca channel P_o to 29%, 38%, and 65% of normoxia, respectively (Fig. 5A).

Fig. 5.

Hypoxia decreases the apparent Ca²⁺ sensitivity of K_Ca channels in excised membrane patches from rat cerebral artery smooth muscle cells. A: original recordings obtained from the same membrane patch voltage-clamped at −40 mV, illustrating K_Ca channel inhibition by hypoxia at 3, 10, and 30 μM Ca²⁺. c, Closed level, with number of open levels also shown. B: average single-K_Ca channel open probability (P_o) vs. free [Ca²⁺] in normoxia and hypoxia. Data were fit with a Hill equation. In normoxia, the mean apparent dissociation constant (K_d) for Ca²⁺ at −40 mV was 17.3 ± 1.29 μM, with a Hill coefficient of 1.52 ± 0.14 and a maximum P_o of 0.80 ± 0.02 (n = 3−6 cells). In the same patches, hypoxia increased the mean K_d for Ca²⁺ of K_Ca channels to 31.9 ± 0.81 μM but did not alter the Hill coefficient (1.68 ± 0.06) or the maximum P_o (0.75 ± 0.01).

DISCUSSION

The major novel findings of our study investigating O₂ regulation of cerebral artery smooth muscle cell K_Ca channels that are under Ca²⁺ spark control indicate that hypoxia: 1) reversibly reduces transient K_Ca current frequency and amplitude, 2) does not alter Ca²⁺ spark frequency or amplitude or change global [Ca²⁺]_i in voltage-clamped cells, 3) attenuates both the coupling ratio and effective coupling of Ca²⁺ sparks to K_Ca channels, and 4) reduces the apparent Ca²⁺ sensitivity of K_Ca channels through a mechanism that is independent of cytosolic signaling pathways. These data indicate that a reduction in P_O2 causes a decrease in K_Ca channel apparent Ca²⁺ sensitivity, leading to K_Ca channel uncoupling from Ca²⁺ sparks and a decrease in transient K_Ca current frequency and amplitude. These data suggest that hypoxia-induced K_Ca channel uncoupling would oppose the resulting vasodilation.

Previous studies have demonstrated that hypoxia inhibits K_Ca channels in rabbit and lamb pulmonary artery smooth muscle cells, rat cerebral artery smooth muscle cells, and rat carotid body chemoreceptor cells (1, 9, 20, 30, 31, 38, 42, 47). Similarly, recombinant human and rat K_Ca channels expressed in immortalized cell lines were inhibited by hypoxia (24, 26, 32, 48). In contrast, hypoxia activated cat cerebral artery smooth muscle cell and piglet pial artery K_Ca channels, leading to vasodilation (3, 4, 16). Our data indicate that in rat cerebral artery smooth muscle cells, hypoxia inhibits transient K_Ca currents in intact cells and K_Ca channels in both intact cells and excised membrane patches. Conceivably, K_Ca channel responses to O₂ may be species specific, but further studies will be required to investigate this hypothesis.

Hypoxia may have blocked transient K_Ca currents by inhibiting Ca²⁺ sparks. Data obtained with confocal imaging indicated that hypoxia did not alter Ca²⁺ spark frequency or amplitude. An elevation in global [Ca²⁺]_i activates Ca²⁺ sparks, whereas a reduction in global [Ca²⁺]_i inhibits Ca²⁺ sparks (21). Hypoxic regulation of global [Ca²⁺]_i was measured to determine whether mechanisms that regulate Ca²⁺ sparks were altered by a reduction in P_O2. However, hypoxia did not change global [Ca²⁺]_i. Since a [Ca²⁺]_SR reduction inhibits Ca²⁺ sparks, and blocking SR Ca²⁺ release elevates [Ca²⁺]_SR, we also measured hypoxic regulation of [Ca²⁺]_SR (6, 28). Hypoxia did not alter [Ca²⁺]_SR, providing further support for our finding that hypoxia did not regulate Ca²⁺ sparks. Rather, hypoxia decreased transient K_Ca current frequency by reducing the percentage of Ca²⁺ sparks that activated a transient K_Ca current. Hypoxia also reduced the amplitude relationship between Ca²⁺ sparks and K_Ca channels that remained coupled, resulting in a decrease in transient K_Ca current amplitude. Therefore, hypoxia reduced K_Ca channel activity in cerebral artery smooth muscle cells by reducing both the coupling percentage and the effective coupling of Ca²⁺ sparks to K_Ca channels. In smooth muscle cells, Ca²⁺ sparks activate transient K_Ca currents by elevating subsarcolemmal [Ca²⁺]_i to within the micromolar concentration range (39, 51). Here, hypoxia reduced K_Ca channel coupling in intact smooth muscle cells where K_Ca channels were exposed to subsarcolemmal micromolar Ca²⁺ concentrations generated by Ca²⁺ sparks. Hypoxia also reduced K_Ca channel apparent Ca²⁺ sensitivity in excised membrane patches that were exposed to micromolar Ca²⁺ concentrations. These data demonstrate that hypoxia reduces K_Ca channel apparent Ca²⁺ sensitivity via a mechanism that is independent of cytosolic signaling pathways, and this a primary mechanism leading to Ca²⁺ spark uncoupling.

Mechanisms by which acute changes in P_O2 regulate K_Ca channel activity are unclear. Heme oxygenase-2 (HO-2) is physically coupled to K_Ca channels and is proposed to act as an oxygen sensor in carotid body glomus cells (23, 48). However, genetic ablation of HO-2 did not change O₂ sensitivity of carotid body glomus cells or chromaffin cells (37). Oxygen sensitivity of K_Ca channels may also depend on the presence of a cysteine-rich, stress-regulated exon (STREX) present within the channel COOH terminus (32). However, K_Ca channels expressed in rat cerebral artery smooth muscle cells were recently cloned by our group and do not contain a STREX motif (20). Thus O₂ sensing in cerebral artery smooth muscle cell K_Ca channels occurs through a STREX-independent mechanism. NADPH oxidase and AMP-activated protein kinase (AMPK) may also act as O₂ sensors in pulmonary artery smooth muscle and carotid body type I cells (10, 22, 23). In the present study, the membrane-delimited effect of hypoxia on K_Ca channels that occurs in excised patches is unlikely to involve soluble signaling messengers or enzymes that require cofactors other than the ions present in the bath solution, arguing against a role for NADPH. In addition, phosphorylation is unlikely to be necessary since ATP was not present, suggesting that AMPK is not involved. Chronic hypoxia also reduces cerebral artery K_Ca channel β₁-subunit expression (35), an effect that would reduce K_Ca channel apparent Ca²⁺ sensitivity (5). However, in our experiments, hypoxic inhibition of transient K_Ca currents and K_Ca channels was immediate and unlikely to occur through a reduction in protein expression. Mitochondria are another potential O₂ sensor in vascular smooth muscle cells (33, 46). Hypoxia depolarizes mitochondria in renal artery smooth muscle cells but hyperpolarizes mitochondria in pulmonary artery smooth muscle cells (33). A small mitochondrial depolarization, such as that induced by diazoxide, an ATP-sensitive potassium (K_ATP) channel opener, or a nanomolar concentration of CCCP, a protonophore, activates Ca²⁺ sparks and transient K_Ca currents in cerebral artery smooth muscle cells (49). In contrast, a large mitochondrial depolarization induced by micromolar CCCP or rotenone, an electron transport chain complex I blocker, inhibits Ca²⁺ sparks and transient K_Ca currents (8, 49). Here, the primary effect of hypoxia was mediated by an effect on K_Ca channels, since hypoxia did not change Ca²⁺ spark frequency or amplitude. These data suggest that if hypoxia alters mitochondrial potential in cerebral artery smooth muscle cells, the net effects on Ca²⁺ sparks and K_Ca channels are very small and secondary to a direct membrane-delimited effect of the P_O2 reduction on K_Ca channels. While the K_Ca channel O₂ sensor is unclear, data indicate that the interaction between O₂ and K_Ca channels can occur in the absence of cytosolic signaling pathways and suggest that the K_Ca channel itself or a closely associated regulatory molecule is the O₂ sensor.

In human and porcine coronary and rabbit cerebral artery smooth muscle cells, hypoxia inhibited L-type Ca²⁺ channels and reduced [Ca²⁺]_i (43). In contrast, in pulmonary artery smooth muscle cells hypoxia stimulated SR Ca²⁺ release and increased cytosolic [Ca²⁺]_i (11, 17, 31, 34, 44). In fetal sheep pulmonary artery smooth muscle cells, a P_O2 elevation stimulated transient K_Ca currents (40), whereas in rabbit pulmonary artery smooth muscle cells hypoxia irreversibly blocked transient outward currents (45). While opposing responses to O₂ have been reported in pulmonary artery smooth muscle cells, in the present study hypoxia did not alter either cytosolic [Ca²⁺]_i in voltage-clamped cerebral artery smooth muscle cells or [Ca²⁺]_SR in isolated myocytes.

Hypoxia leads to cerebral artery dilation, a response that functions to match blood flow to metabolic requirements (15, 25). Although the signaling mechanisms mediating hypoxic vasodilation may depend on the precise reduction in P_O2, hypoxia induces systemic artery hyperpolarization, which would reduce smooth muscle cell voltage-dependent Ca²⁺ channel activity, leading to a reduction in global [Ca²⁺]_i and vasodilation (13, 29, 30, 43). Data here indicate that a hypoxic reduction in K_Ca channel activity due to Ca²⁺ spark uncoupling would attenuate, rather than contribute to, the hypoxic hyperpolarization and vasodilation. In contrast to effects in systemic arteries, hypoxia depolarizes and constricts pulmonary arteries (31). In pulmonary artery smooth muscle cells, hypoxia-induced K_Ca channel inhibition would contribute to the membrane depolarization, which would activate voltage-dependent Ca²⁺ channels, leading to vasoconstriction (31, 34). It remains to be determined whether hypoxia also modulates Ca²⁺ sparks and the coupling relationship between Ca²⁺ sparks and K_Ca channels in pulmonary artery smooth muscle cells, and whether such changes are mediated by the mechanism we describe here.

In summary, data indicate that hypoxia reduces K_Ca channel apparent Ca²⁺ sensitivity through a membrane-delimited mechanism, leading to a decrease in the effective coupling of Ca²⁺ sparks to K_Ca channels and a reduction in transient K_Ca current frequency and amplitude. These data indicate that K_Ca channel inhibition caused by uncoupling from Ca²⁺ sparks would oppose the hypoxic vasodilation.