This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Guia, A. Articles by Leblanc, N. Search for Related Content PubMed PubMed Citation Articles by Guia, A. Articles by Leblanc, N. Pubmed/NCBI databases Compound via MeSH Substance via MeSH Hazardous Substances DB CALCIUM COMPOUNDS CALCIUM, ELEMENTAL Related Collections Animal models of human disease Cell biology/structural biology Ion channels/membrane transport

(Circulation Research. 1999;84:1032-1042.)
© 1999 American Heart Association, Inc.

Original Contributions

Local Ca²⁺ Entry Through L-Type Ca²⁺ Channels Activates Ca²⁺-Dependent K⁺ Channels in Rabbit Coronary Myocytes

Antonio Guia, Xiaodong Wan, Marc Courtemanche, Normand Leblanc

From the Research Center, Montreal Heart Institute, Montréal, Québec, Canada.

Correspondence to Normand Leblanc, Research Center, Montreal Heart Institute, 5000 Bélanger St E, Montréal, Québec, Canada H1T 1C8. E-mail leblancn{at}alize.ere.umontreal.ca

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
Abstract—Large-conductance Ca²⁺-dependent K⁺ channels (K_Ca), which are abundant on the sarcolemma of vascular myocytes, provide negative feedback via membrane hyperpolarization that limits Ca²⁺ entry through L-type Ca²⁺ channels (I_CaL). We hypothesize that local accumulation of subsarcolemmal Ca²⁺ during I_CaL openings amplifies this feedback. Our goal was to demonstrate that Ca²⁺ entry through voltage-gated I_CaL channels can stimulate adjacent K_Ca channels by a localized interaction in enzymatically isolated rabbit coronary arterial myocytes voltage clamped in whole-cell or in cell-attached patch clamp mode. During slow-voltage-ramp protocols, we identified an outward K_Ca current that is activated by a subsarcolemmal Ca²⁺ pool dissociated from bulk cytosolic Ca²⁺ pool (measured with indo 1) and is dependent on L-type Ca²⁺ channel activity. Transient activation of unitary K_Ca channels in cell-attached patches could be detected during long step depolarizations to +40 mV (holding potential, -40 mV; 219 pS in near-symmetrical K⁺). This local interaction between the channels required the presence of Ca²⁺ in the pipette solution, was enhanced by the I_CaL agonist Bay K 8644, and persisted after impairment of the sarcoplasmic reticulum by incubation with 10 µmol/L ryanodine and 30 µmol/L cyclopiazonic acid for at least 60 minutes. Furthermore, we provide the first direct evidence of simultaneous openings of single K_Ca (67 pS) and I_CaL (3.9 pS) channels in near-physiological conditions, near resting membrane potential. Our data imply a novel sensitive mechanism for regulating resting membrane potential and tone in vascular smooth muscle.

Key Words: local channel regulation • subsarcolemmal Ca²⁺ • vascular smooth muscle • feedback regulation • indo 1 fluorescence

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
Large-conductance Ca²⁺-dependent K⁺ channels (K_Ca) play an important role in the regulation of arterial tone and vascular resistance.¹ These channels serve as an important negative feedback mechanism that indirectly sets the level of Ca²⁺ entry through voltage-operated L-type Ca²⁺ channels (I_CaL) by shifting membrane potential in a direction that reduces their steady-state open probability. Because steady-state activity of I_CaL current (I_CaL) can be detected near the resting membrane potential (RMP; –50 mV to –30 mV) in arterial^{2 3 4} and airway⁵ smooth muscle cells, small changes in RMP can profoundly affect tension development in tonic vascular smooth muscle.⁶ Vasodilation may be produced either directly by interfering with I_CaL^{7 8} or indirectly by stimulating K⁺ channels, thereby hyperpolarizing RMP^{9 10 11} and limiting Ca²⁺ entry through I_CaL channels. The resulting reduction in cytosolic free Ca²⁺ concentration ([Ca²⁺]_i) leads to relaxation of the smooth muscle cells. Hence, RMP represents a key factor in the control of vascular tone, especially in the coronary vasculature^{12 13 14 15} and in arteries submitted to constant high-pressure loads.^{1 16}

K_Ca activity is regulated by Ca²⁺ entry and by Ca²⁺ release from sarcoplasmic reticulum (SR).¹⁷ Recent evidence has indicated that changes in K_Ca activity are better described by alterations in subsarcolemmal Ca²⁺ levels than by overall or bulk cytosolic Ca²⁺ content.¹⁸ In 1989, van Breemen and Saida¹⁹ proposed the existence of a superficial buffer barrier, suggesting that a fraction of the Ca²⁺ that crosses the sarcolemma during agonist- or depolarization-induced stimulation may not reach the contractile machinery or bulk cytoplasm but may instead be taken up by the Ca²⁺-ATPase of the peripheral or subsarcolemmal elements of the SR,²⁰ thereby creating a subsarcolemmal compartment with elevated [Ca²⁺]_i. This compartmentalization of Ca²⁺ was demonstrated in coronary smooth muscle myocytes¹⁸ by probing the subsarcolemmal [Ca²⁺], which was inferred from the activity of K_Ca, and the bulk cytosolic [Ca²⁺], which was measured with fura 2. A buffering role of superficial SR was also recently demonstrated in studies showing that intracellular Ca²⁺ accumulation during Ca²⁺ influx is enhanced after cell exposure to agents interfering with SR function.^{21 22 23} Although these studies indicate that SR may be involved in the Ca²⁺-dependent activation of K_Ca, they do not rule out the possibility that K_Ca may be locally activated by Ca²⁺ entry through I_CaL channels.

In the present study, we tested the hypothesis that sarcolemmal I_CaL channels can themselves stimulate K_Ca channels by increasing [Ca²⁺]_i in the environment of neighboring K_Ca channels. Measurements of whole-cell and single-channel K_Ca and I_CaL currents in conditions that modulated Ca²⁺ entry through I_CaL channels, or Ca²⁺ release from the SR, provide evidence for a close interaction between voltage-operated I_CaL channels and K_Ca channels. This localized effect may play a primary role in modulating the RMP, amount of Ca²⁺ entry, development of muscle tone, and therefore coronary blood flow. Some of the results of this work have been presented in abstract form.²⁴

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
Cell Isolation
Coronary smooth muscle cells were isolated from New Zealand White rabbits of either sex (1.5 to 2 kg; Charles River, St-Constant, Québec, Canada) by enzymatic digestion (method adapted from Remillard and Leblanc²⁵ ). Rabbits were killed by cervical dislocation, and the heart was removed via a transverse and bilateral longitudinal incision to the chest. The heart was rinsed by gentle retrograde aortal injection of 10 µmol/L Ca²⁺-containing dissecting solution (chemicals and solutions are defined later) and placed into the same solution for further dissection at room temperature (22°C). All solutions used during the isolation procedure were bubbled with 95% O₂, balance CO₂. Under a stereoscopic dissecting microscope the left anterior descending and the right circumflex coronary arteries were cut free of ventricular tissue and, as much as possible, of the tunica adventitia.

After dissection, the arteries were subjected to 30 minutes of Ca²⁺-free dissecting solution to remove Ca²⁺ from interstitial spaces and then placed into 10 µmol/L Ca²⁺-containing dissecting solution for further dissection. The coronary arteries were cut into 1-mm lengths and transferred to a test tube containing 5 mL of 10 µmol/L Ca²⁺-containing dissecting solution with the following compounds added: 300 to 350 U/mL collagenase type 1A, 0.4 U/mL protease type XXVII, and 1000 BAEE U/mL trypsin inhibitor. Enzymatic digestion was allowed to proceed in the test tube at 35°C. After 10 minutes, a further 0.4 U/mL protease was added. The tissue was digested for an additional 5 minutes or until the tissue pieces started to stick to one another, indicating possible intertangling of cells or fragments that were partially liberated from the tissue pieces.

Digestion was quenched by rinsing the preparation at least 3 times with enzyme-free 10 µmol/L Ca²⁺-containing dissecting solution at room temperature. The partially digested preparations were triturated gently with a fire-polished glass Pasteur pipette until smooth muscle cells were mechanically liberated from the tissue pieces. This cell-isolation technique produces relaxed myocytes (average whole-cell capacitance was 24 pF²⁶ ) with a visibly healthy cellular morphology that do not contain cytosolic inclusions, as visible under a standard light microscope, and that contract and relax in response to agonists and changes in membrane potential. After isolation, the preparation containing isolated cells was refrigerated at 4°C in 10 µmol/L Ca²⁺-containing dissecting solution. Aliquots of the isolated cells were taken as needed and allowed to warm to room temperature for 10 to 20 minutes before experiments were performed. Cells were used within 8 hours of being isolated.

Whole-Cell Experiments
The myocytes were voltage clamped in the whole-cell configuration.²⁷ All experiments were performed at room temperature with the cells settled onto a coverslip bound with silicone glue to the bottom of a 1-mL acrylic recording chamber mounted on a moveable stage of an inverted microscope (Diaphot, Nikon). The recording chamber was designed to allow efficient exchange of solutions in an open, nonrecycling system in which the inflow (2 to 3 mL per minute) is driven by gravity and the outflow is suctioned out to a waste container. The cells were perifused with normal external solution for at least 10 minutes before formation of a seal was attempted. Sections of borosilicate glass tubes were pulled on a Narishige vertical 2-stage pipette puller (model PP-83; Narishige Scientific Instruments) and polished on a Narishige Microforge (model FP-83) to yield micropipettes having tip resistances of 3 to 5 M (1 to 2 µm diameter) when filled with normal intracellular pipette solution. The filled micropipette was inserted into an Ag/AgCl electrode assembly that was attached to a headstage preamplifier mounted on a motorized micromanipulator (model MS-314; Fine Science Tools, Inc) used to direct the micropipette tip onto the cell membrane. The pipette junction potential was zeroed with the tip in external solution. Liquid junction potential between pipette solution and external solution was compensated for in all experiments (–10 mV). Formation of a 2- to 10-G seal was accomplished by gently pressing the pipette down on the cell while applying gentle suction to the pipette. Tip capacitance and, after gaining whole-cell access, whole-cell resistance were compensated for during a repeatedly applied square voltage pulse using the voltage-clamp amplifier (model Axopatch 1D or 200A, Axon Instruments). Whole-cell capacitance and leak compensation were not performed. Experimental protocols were usually initiated after 2 minutes of gaining whole-cell access. Data were digitally acquired (Labmaster DMA acquisition board with a TL-1-125 interface or Digidata 1200 acquisition system, Axon Instruments) at a 1- to 5-kHz sampling rate with a 1-kHz 4-pole Bessel filter. Experimental protocols were executed with the use of pClamp software (version 5.5, Axon Instruments), and the data were processed on an IBM-compatible PC computer.

Single-Channel Measurements
Some cells were voltage clamped in the cell-attached patch configuration.²⁷ Techniques and equipment used for unitary current measurements were similar to those described for whole-cell experiments except for the following differences. Experiments were performed in a smaller recording chamber designed to hold 0.5 to 0.8 mL volume. Cells were perifused with Ca²⁺-free external solution for at least 10 minutes before formation of a seal was attempted. Micropipettes with tip resistances of 4 to 8 M (0.6 to 1 µm diameter) were filled with either Ca²⁺-containing or Ca²⁺-free depolarizing medium. When needed, racemic Bay K 8644 was added to the pipette solution. Because the junction potential was <2 mV, compensation for junction potentials was not performed. After a G seal was formed, the tip capacitance was compensated, except during measurements of single I_CaL channels, and the cell was perifused with a Ca²⁺-free depolarizing medium for 5 to 10 minutes before unitary currents were recorded. Signals were filtered (4-pole Bessel) at 500 Hz to 5 kHz. Acquisition of I_CaL channel data were done using 1-kHz filtering, and the data were then digitally filtered (5-point boxcar) offline at 500 Hz. The depolarizing medium was used to clamp the cell RMP near 0 mV so that the patch transmembrane potential was approximately equal to the inverse of the applied potential. Membrane potentials reported in this study are expressed as true transmembrane potentials (–V_applied).

Because there are numerous K_Ca channels on the sarcolemma, multiple channels were present in almost every membrane patch. As a consequence, open probability (P_o) is expressed as NP_o, calculated using the following formula (see Kajioka et al¹⁰ ).

N represents the number of single channels in the patch. A_c represents the area under the all-points amplitude histogram for the channel in the closed state. A₁ to A_n represent the area under the all-points amplitude histogram for the open state of 1 to n channels. The values of A₁ to A_n were calculated with pClamp software (v.6.0, Axon Instruments) using a simplex least-squares fit of the data to Gaussian distribution curves.

Indo 1 Fluorescence
In some experiments the concentration of free intracellular [Ca²⁺]_i was measured simultaneously with whole-cell membrane currents, as previously described.^{28 29} The salt form of indo 1 (indo 1-K₅, Molecular Probes, Inc) was used to monitor relative changes in [Ca²⁺]_i. The epifluorescence attachment of the Nikon inverted microscope (with a 100-W mercury arc lamp) was fitted with an electronic shutter that could be triggered immediately before the application of the voltage-ramp protocol. Indo 1 was excited by means of a high-pass dichroic mirror (allowing transmission of wavelengths longer than 380 nm) and an excitation filter (340 nm, 10-nm bandwidth). Epifluorescence emission light was focused through the side port of the microscope onto a red (allowing transmission of wavelengths longer than 600 nm) filter with a small hole drilled through the center. This allowed for fluorescence signals (at 500-nm and 400-nm wavelengths with 10-nm bandwidth) to be measured (with 2 Hamamatsu type R2560HA photomultiplier tubes energized by a 1000-V power supply) from an 8-µm area within the visual field into which the cells were aligned before voltage clamping. The excitation beam was focused onto an area slightly larger than the area inside this 8-µm-diameter area.

Indo 1 was included in the pipette (EGTA-free) solution and dialyzed into the cell. A G seal was formed near but not inside the 8-µm-diameter fluorescence measuring area. The background fluorescence signal was canceled for each cell after the seal was formed. Once whole-cell access was established, the cell was allowed to dialyze with the indo 1–containing pipette solution for 3 to 10 minutes. The loading of the indicator was assessed by monitoring the changes in fluorescence intensity of the dye as a function of time. The ratio of the fluorescence emission (400 nm/500 nm; dynamic range between 0.3 with cytosolic Ca²⁺-free EGTA and 2.5 with saturating [Ca²⁺]_i), which was used as an index of [Ca²⁺]_i, was filtered at 50 Hz and digitally recorded simultaneously with the voltage-clamp signal.

Chemicals and Solutions
The 3 dissecting solutions had similar composition (in mmol/L, NaCl 120, NaHCO₃ 25, KCl 4.2, KH₂PO₄ 1.2, MgCl₂ 1.2, glucose 11, taurine 25, and adenosine 0.01) except for the Ca²⁺ (CaCl₂) concentration, which was as stated. In the Ca²⁺-free dissecting solution, CaCl₂ was omitted in exchange for 0.1 mmol/L EGTA.

In whole-cell voltage-clamp studies designed to study outward K currents in Figures 1, 2, and 4, normal external solution used for superfusion of the myocytes contained the following (in mmol/L): NaCl 130, NaHCO₃ 10, KCl 4.2, KH₂PO₄ 1.2, MgCl₂ 0.5, CaCl₂ 1.8, glucose 5.5, and HEPES 10, pH 7.35 (NaOH) at 22°C. Ca²⁺-free solution was of similar composition, but CaCl₂ was substituted with MgCl₂. Normal intracellular pipette solution was composed of the following (in mmol/L): potassium gluconate 110, KCl 30, NaCl 10, MgCl₂ 0.5, ATP-K₂ 5, HEPES 5, and EGTA 0.1 or 5.0 (where stated), pH 7.20 (KOH).

View larger version (31K): [in this window] [in a new window]   Figure 1. Dependence of whole-cell membrane current elicited by a slow ramp protocol on Ca2+ entry. Currents were elicited by a voltage-ramp protocol from –80 to +60 mV over 5 seconds (protocol shown in panel A, inset, upper left) using normal (intracellular) pipette solution. A, Left, Currents were elicited in normal external solution (1.8 mmol/L Ca2+ [1.8Ca]) and after 10 minutes in nominally Ca2+-free solution (0Ca). Right, Data recorded in Ca2+-free medium (shown in left panel) were subtracted from data recorded under normal conditions to reveal the I-V relationship of the Ca2+-free sensitive current (0Ca). Mean data are reported in panel C. B, Left, Currents were recorded from a different cell before (Ctrl) and after the addition of 1 µmol/L nifedipine (Nif). Right, Trace obtained in the presence of nifedipine was subtracted from that under control conditions to demonstrate the I-V of the nifedipine-sensitive current. Mean data are reported in panel C. C, Averaged data for nifedipinesensitive current (1 µmol/L, n=6, Nif) and Ca2+-free sensitive current (n=4, 0Ca) are shown with their 1-sided SEs (dots). Currents are obtained by subtracting control traces from test traces. Inset, Same average currents are better resolved in the voltage range of –60 to –10 mV. D, Average IbTx-sensitive current in the absence (100 nmol/L, n=5, IbTx) and presence of nifedipine (1 µmol/L, n=4, IbTx [+1 µmol/L Nif]) are shown with their 1-sided SEs (dots).

View larger version (27K): [in this window] [in a new window]   Figure 2. Effect of increased Ca2+ entry on whole-cell membrane current elicited by voltage-ramp protocol. Voltage-ramp protocol and solutions are identical to those described in Figure 1. A, Left, Currents measured from a cell exposed to normal physiological (1.8 mmol/L; 1.8Ca) and elevated (5.4 mmol/L; 5.4Ca) Ca2+-containing external medium for 6 minutes. Right, Data under normal conditions were subtracted from high Ca2+ to demonstrate the I-V of the high-[Ca2+]o–sensitive current (5.4Ca). Mean data are reported in panel C. B, Left, Currents measured in a different cell under control conditions (Ctrl) and in the presence of 1 µmol/L Bay K 8644 (BayK). Right, Net I-V of the current enhanced by Bay K 8644 was obtained by digital subtraction of the control current from the current in the presence of Bay K 8644 (BayK). Mean data are reported in panel C. C, Average traces with their 1-sided SEs (dotted lines) are displayed for the Bay K–sensitive current (BayK, n=5) and for the high-[Ca2+]o–sensitive current (5.4Ca, n=7). Left axis shows scale of the Bay K–sensitive current, and right axis shows scale for high-[Ca2+]o–sensitive current.

View larger version (27K): [in this window] [in a new window]   Figure 4. Dissociation between Ca2+ entry and bulk cytosolic [Ca2+]i. Voltage-ramp protocol is identical to that described in Figure 1. Cells were separated into 2 populations, 1 lacking (A) and 1 demonstrating (B) voltage-sensitive Ca2+ transient (indicated by indo 1 ratio) during ramp protocols identical to those shown above. Plots reflect mean±SE (solid and dotted lines, respectively) of indo 1 ratio versus ramp voltage obtained in 6 and 5 myocytes, respectively. Horizontal dashed lines in panel B demonstrate the presence of a [Ca2+]i plateau above +10 mV. Traces are aligned to the same voltage scale.

For studies designed to measure whole-cell I_CaL in Figure 3, the external solution was modified by replacing external KCl and KH₂PO₄ with 5.4 mmol/L tetraethylammonium chloride (TEA-Cl). Pipette solution contained the following (in mmol/L): cesium aspartate 105, TEA-Cl 20, CsCl 20, MgCl₂ 0.5, ATP-2Na 5, HEPES 5, and EGTA 5, pH 7.2 (CsOH).

View larger version (35K): [in this window] [in a new window]   Figure 3. Macroscopic inward Ca2+ current elicited by voltage step or ramp protocol. Currents were measured using K+-free bathing and pipette solutions (K+ was replaced with Cs+). A, Sample family of inward current traces in response to 250-ms depolarizing steps (HP, –60 mV) to various voltages (protocol shown below traces). B, Average peak inward current from 10 experiments identical to those in panel A is plotted as a function of step voltage. The resulting bell-shaped I-V curve shows a peak at 0 mV and an apparent reversal potential at +35 mV. Inset, Under conditions identical to those in panel A, voltage steps were applied in the absence and in the presence of 100 µmol/L Cd2+. Graph shows the voltage dependence of peak Cd2+-sensitive inward Ca2+ current (Cd) obtained by digital subtraction of the currents acquired in the presence of Cd2+ from those acquired in control conditions (n=4). C, Voltage-ramp protocol identical to that described in Figure 1 was applied in the same 10 cells as in panel B. The average membrane current is shown (SEs are indicated by dotted lines), with a peak inward current at –12 mV and reversal at +14 mV. D, Under conditions identical to those in panel C, voltage ramps were applied in the absence (Ctrl) and in the presence of 100 µmol/L Cd2+ (Cd). Currents displayed are normalized to unit membrane surface area and averaged from 4 cells (same cells as in inset of panel B). Inset, Currents acquired in the presence of Cd2+ were subtracted from those obtained under control conditions to reveal the Cd2+-sensitive current (Cd) with a peak inward current at –14 mV and a reversal potential of +53 mV.

Single-channel K_Ca measurements in Figures 5 and 7 were performed under pseudosymmetrical K⁺ gradients. Ca²⁺-free external solution was the same as that used for whole-cell studies. Depolarizing external solution consisted of the following (in mmol/L): KCl 140, MgCl₂ 2, HEPES 10, EGTA 0.1, and glucose 5.5, pH 7.4 (KOH) at 22°C. Ca²⁺-free pipette solution was of the same composition as the depolarizing external solution. Ca²⁺-containing pipette solution was also of similar composition; however, EGTA was omitted and 2 mmol/L CaCl₂ was added.

View larger version (32K): [in this window] [in a new window]   Figure 5. Transient stimulation of single KCa channels recorded in cell-attached patches during voltage-clamp steps. Cells in Ca2+-free depolarizing medium (pseudosymmetrical K+) were voltage clamped to +40 mV (transmembrane potential) for 5 seconds (HP, -40 mV; voltage protocol is illustrated to lower right of traces). A, Sample traces recorded from 1 cell with 2 mmol/L Ca2+-containing (+Ca) medium in the pipette (left) and from another cell with Ca2+-free (–Ca) medium in the pipette (right). Similar results were obtained in 3 of 7 cells for Ca2+-containing pipette solution and in 8 cells for Ca2+-free pipette solutions. B, Mean±SE open probability (see Results) of KCa channels measured during the +40 mV voltage step (n=7 for Ca2+-containing pipettes, ; n=8 for Ca2+-free pipettes, ). C, Protocols and conditions are identical to those described in panel A except that Bay K 8644 (1 µmol/L) was added to the pipette solution. Shown are sample traces recorded from 1 cell with 2 mmol/L Ca2+-containing pipette solution (left) and from another cell with Ca2+-free solution in the pipette (right). Similar results were obtained in 9 of 11 cells for Ca2+-containing pipette solution and in 12 cells for Ca2+-free pipette solution. D, Mean±SE open probability (see Results) of KCa in the presence of 1 µmol/L Bay K 8644 measured during the voltage step to +40 mV (n=11 for Ca2+-containing pipettes, ; n=12 for Ca2+-free pipettes, ).

View larger version (24K): [in this window] [in a new window]   Figure 7. Transient activation of KCa is independent of SR function. In this cell, the SR was compromised by bathing the cell for at least 60 minutes in Ca2+-free depolarizing medium containing 10 µmol/L Ryan and 30 µmol/L CPA. Bay K 8644 (1 µmol/L) and Ca2+ (2 mmol/L) were present in the pipette. Voltage-clamp protocol is identical to that described in Figure 5. Shown are four sample traces obtained at different incubation times in the presence of Ryan and CPA. As shown, the transient activation was maintained during repeated application of step protocols. A similar degree of transient activation was observed in 3 cells, and a smaller degree of activation was observed in a further 4 cells (from a total of 11 cells).

Single–I_CaL channel currents in Figure 6 were measured in asymmetrical K⁺ gradients. Depolarizing external solution was the same as that described above for single-channel measurements. The pipette contained Ca²⁺-free external solution with the addition of 5 mmol/L CaCl₂ and 5 mmol/L 4-aminopyridine (4AP). Pipette tips were coated with Sylgard elastomer (Dow Corning) to reduce noise in the recordings.

View larger version (39K): [in this window] [in a new window]   Figure 6. Dependence of KCa single-channel openings on ICaL openings. Cells were bathed in depolarizing solution containing 5 mmol/L 4AP and 10 nmol/L IbTx. Cell-attached patches were voltage clamped with a pipette solution containing Ca2+-free extracellular solution with added 5 mmol/L 4AP, 10 nmol/L IbTx, 1 µmol/L Bay K 8644, and 5 mmol/L Ca2+. Aa, Single-channel KCa currents (upward deflections) and ICaL currents (downward deflections) in response to 200-ms voltage steps (HP, -60 mV) to –30 mV. Protocol is shown at the top left. Note that ICaL channel openings and closures may be seen during KCa openings. Similar results were observed in 5 cells (NPo=0.66±0.37 for KCa). b, Traces obtained under conditions similar to those in panel Aa but with 10 µmol/L nifedipine added to the pipette. Similar results were obtained in 5 cells (NPo=0 for KCa). B, Single-channel I-V relationship for ICaL channels (left) and KCa channels (right). Data are averaged (mean current±SE) and fit by linear regression (95% confidence bands shown) to yield single-channel conductances of 3.9±0.5 pS (r2=0.96, n=5) and 67±5 pS (r2=0.97, n=8) for ICaL and KCa channels, respectively.

TEA-Cl and 4AP were added as powder directly to the solution to the final desired concentration. Iberiotoxin (IbTx) was diluted to the desired concentration from a 100 µmol/L aqueous stock solution. Bay K 8644 and nifedipine were diluted to the desired concentration from 10 mmol/L stock solutions of DMSO. Calcium chloride was diluted from a 1 mol/L aqueous stock. Stock solutions were kept frozen at -20°C.

Glucose, MgCl₂ · 6H₂O, CaCl₂ · 2H₂O, KCl, KH₂PO₄, CdCl₂, and NaCl were purchased from Malinckrodt Chemical Co. NaHCO₃ was purchased from J.T. Baker Inc. Bay K 8644 racemate was purchased from Calbiochem. IbTx was purchased from Research Biochemicals International. All other chemicals were purchased from Sigma.

Statistical Analysis
Where possible, the results are expressed as mean±SE. In simple comparisons between 2 groups, statistical significance was determined using the Student paired 2-tailed t test, for comparisons between treatments using the same group of cells, or the Student unpaired 2-tailed t test, for comparisons using different groups of cells. In single-channel studies the Student t tests were complemented with the Bartlett homogeneity of variance test. Multiple comparisons were performed by an ANOVA test, followed by a Duncan or a Tukey post hoc test to determine significance between individual groups. Significance threshold was set at 95% confidence.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
Ca²⁺ Dependence of K⁺ Currents Elicited by Slow Voltage Ramps
Cytosolic [Ca²⁺]_i and L-type Ca²⁺ current (I_CaL) are important in regulating the activity of K_Ca¹⁶ ; hence, we first investigated the effects of reducing Ca²⁺ entry by removal of Ca²⁺ from the perfusate or by inhibition of I_CaL. Five-second ramp protocols were used to record membrane current. With 1.8 mmol/L [Ca²⁺]_o, the current-voltage (I-V) relationship of the ramp current is biphasic, exhibiting an initial "hump" followed by a second outwardly rectifying noisy component (Figure 1A, left). A large fraction of the hump is sensitive to millimolar concentrations of 4AP (data not shown) and results from the activity of voltage-dependent K⁺ channels (K_V). Charybdotoxin- or TEA-sensitive Ca²⁺-activated K⁺ current (K_Ca) is the dominant conductance responsible for the second phase.^{26 28 30} Removing Ca²⁺ in the perfusate reduced the outward current and the amount of current noise at positive potentials (above –15 mV) and shifted the current in the outward direction at potentials below –15 mV. Figure 1A, right, illustrates the voltage dependence of the current sensitive to [Ca²⁺]_o removal. Above –15 mV, removing Ca²⁺ ions revealed an outwardly rectifying current consistent with the properties of K_Ca.²⁶ At these potentials the current displays a hump that is consistent with the activation of K_Ca by Ca²⁺ entry through I_CaL. Below –10 mV, external Ca²⁺ removal suppressed a steady inward current that exhibited bell-shaped voltage dependence between approximately –40 mV and –15 mV (Figure 1A, right).

We next determined the specific role of I_CaL in regulating K_Ca during similar ramp protocols. As shown in panel B, left, outward current elicited by a voltage ramp was differently affected on a 2-minute exposure to 1 µmol/L nifedipine when compared with Ca²⁺-free medium (Figure 1B). Nifedipine reduced the outward ramp current at potentials between –15 and +60 mV, with a maximum effect near +15 mV. The outward current decreased at potentials positive to –20 mV, an effect that resembled that produced by external Ca²⁺ removal (panel A). Figure 1B, right, displays the digitally subtracted (nifedipine-sensitive) current plotted as a function of ramp voltage. Net inward current was evident between approximately –50 and –25 mV. However, the nifedipine-sensitive outward current displayed a bell-shaped voltage dependence with a maximum near +20 mV. Prolonged exposure (>10 minutes) to nifedipine likely depleted the SR and subsarcolemmal space as it resulted in changes that were similar to those produced by Ca²⁺ removal (data not shown). Pooled data for the 2 conditions show similar results (Figure 1C). The concentration of nifedipine used was found to have no effect on K_V measured during depolarizing steps from –20 to +20 mV (holding potential [HP], -60 mV) in cells dialyzed with 5 mmol/L EGTA and superfused with 100 µmol/L Cd²⁺ and 100 nmol/L IbTx (n=4), a potent and selective inhibitor of maxi K_Ca channels.^{16 31}

Figure 1D isolates the IbTx-sensitive current (100 nmol/L) in the absence and presence of nifedipine (Nif, 1 µmol/L) as indicated. The hump was abolished when the cells were preincubated with nifedipine, leaving a monophasic outwardly rectifying conductance consistent with that of K_Ca at resting [Ca²⁺]_i level. Except for the lack of net inward current at potentials less than –15 mV, the outward current profile of the IbTx-sensitive current in the absence of nifedipine is similar to that obtained with the Ca²⁺-free–sensitive current (panel A, right) or the nifedipine-sensitive ramp current after a long incubation (data not shown).

As shown on the left side of Figure 2A, elevation of [Ca²⁺]_o from 1.8 to 5.4 mmol/L increased an outward current that displayed outward rectification (panel A, right). The enhanced outward current was accompanied by a negative shift of the 0-current potential (by –18 mV in this cell), indicating possible cell hyperpolarization.²⁶ This hyperpolarization was confirmed by measurements in current clamp mode, in which the RMP of 5 cells hyperpolarized by 24±8 mV from a control RMP of –31±4 mV to –55±2 mV (P<0.001) after raising extracellular Ca²⁺ to 5.4 mmol/L (data not shown).

We also examined the effect of 1 µmol/L Bay K 8644 on ramp current. Bay K increased outward ramp current and shifted the net 0-current potential by –30 mV (Figure 2B, left). A Bay K–induced hyperpolarization was confirmed in current clamp experiments (–28±4 mV in control, –51±0.6 mV in Bay K, n=3). Analysis of the voltage dependence of the Bay K–sensitive current (Figure 2B, right) reveals that the elicited current reversed near potassium equilibrium potential (E_K) (–75 mV) displayed outward rectification up to –10 mV and gradually declined toward 0 as the membrane was further depolarized. Mean data for high-Ca²⁺– and Bay K–sensitive currents are displayed in Figure 2C. Bay K 8644 failed to stimulate outward K current when the cells were preincubated with 1 mmol/L TEA (n=4) or in cells dialyzed with 5 mmol/L EGTA (data not shown, n=5).

L-Type Ca²⁺ Current
To test the hypothesis that the modulation of outward K⁺ currents during the voltage ramp results from slow Ca²⁺ entry through I_CaL channels, we measured I_CaL under conditions that minimized the activity of K⁺ channels. All cells studied displayed time-dependent inward Ca²⁺ currents similar to the traces shown in Figure 3A. Panel B shows the mean I-V relationship for peak inward current (n=10). The voltage dependence of this current reveals activation starting near –40 mV, a peak inward current near 0 mV, and apparent reversal potential at +35 mV. The voltage dependence of Cd²⁺-sensitive currents shows net inward Ca²⁺ current spanning the entire voltage range studied as previously described.³

Figure 3C shows the mean I-V relationship of current elicited by a voltage-ramp protocol identical to that used for Figures 1 and 2. Net inward current was apparent at –40 mV, peaked at –15 mV, and reversed at +15 mV. The more negative peak of the I-V relationship of membrane current elicited by ramp versus pulse protocols is consistent with the steady-state versus nonstationary voltage dependence of open probability of single I_CaL channels.³² For currents elicited by step (panel B) and voltage-ramp (panel C) protocols, the relatively negative reversal potential of the current is related to the existence of an unknown superimposed outwardly rectifying component, which can be better appreciated in panel D, in which I_CaL is abolished by the addition of 100 µmol/L Cd²⁺ in the bath. The inset shows that the reversal potential of the Cd²⁺-sensitive current (+53 mV) is similar to that reported for I_CaL in the same preparation.³

Local Regulation of Whole-Cell K_Ca by I_CaL
K_Ca current and soluble fluorescent cytosolic Ca²⁺ indicators have been used to assess changes in subsarcolemmal versus bulk intracellular [Ca²⁺]_i, respectively.^{18 33 34} This approach was used to test whether an increase in bulk cytosolic Ca²⁺ is required to activate K_Ca during a ramp protocol.

Figures 4A and 4B report the mean±SE indo 1 fluorescence ratio in response to voltage ramps identical to those used for Figures 1 and 2. Of 11 myocytes, 6 failed to exhibit a Ca²⁺ transient during the ramp despite showing a nifedipine-sensitive current similar to that shown in Figure 1B (panel A). This illustrates a complete dissociation between the bulk cytosolic (measured with indo 1) and subsarcolemmal (measured by the activity of K_Ca) Ca²⁺ during the slow activation of I_CaL channels. The remaining 5 myocytes displayed a Ca²⁺ transient during the ramp (panel B). This rise in intracellular Ca²⁺ was apparent at voltages positive to –60 mV and reached a plateau near +10 mV (indicated by the upper horizontal dashed line), consistent with the voltage dependence of activation of I_CaL in this preparation (Figure 3D, inset). The fact that [Ca²⁺]_i failed to return to its initial resting level whereas nifedipine-sensitive current declined at more positive voltages (>+20 mV; Figure 1B) further agrees with the concept of a dissociation between the 2 compartments for calcium.^{18 23 35}

The lack of a Ca²⁺ transient in Figure 4A does not indicate experimental error or failure of the indicator to represent intracellular Ca²⁺ for the following reasons: (1) the amount of dye loading, as indicated by the fluorescence emission at 500 nm, was relatively constant from cell to cell (data not shown); (2) the resting ratio, as seen at the start of the voltage ramp where membrane potential is close to RMP, was similar in all cells studied and compared well with that measured by our group in rabbit portal vein cells²⁹ ; and (3) all cells displayed a prominent L-type Ca²⁺ current (see Figure 3) and nifedipine-inhibitable activation of K_Ca current (see Figure 1B).

Single-Channel Measurements
The hypothesis of local regulation of K_Ca by I_CaL was further tested in cell-attached patch experiments carried out with myocytes exposed to K⁺-depolarizing medium. Figure 5A (left) shows sample recordings from the activity of at least 2 K⁺ channels recorded in the cell-attached configuration in the presence of 140 mmol/L K⁺ and 2 mmol/L Ca²⁺ in the pipette solution. The estimated conductance of these channels in this patch was 224 pS, a value typical for single maxi K_Ca channels.^{16 28 36} In 4 patches under pseudosymmetrical K⁺, the mean conductance and reversal potential were 219±11 pS and +1±4 mV, respectively. From top to bottom are shown 6 consecutive traces that were elicited by the voltage-clamp protocol shown at the bottom right side. A step to +40 mV was chosen to (1) allow sufficient Ca²⁺ entry through I_CaL to occur (maximum open probability³ ) and (2) facilitate the recording of K_Ca by providing a reasonable driving force for K⁺ (+40 mV). With physiological [Ca²⁺] in the pipette solution, K_Ca channels tended to open early during the pulse. The transient stimulation of K_Ca was observed in 43% of the patches and disappeared when Ca²⁺ was omitted from the pipette solution (Figure 5A, right). Overall NP_o of K_Ca was significantly higher in the experiments carried out with Ca²⁺ in the tip (including the patches that did not exhibit transient behavior) as opposed to Ca²⁺-free experiments (Figure 5B). Inclusion of Bay K 8644 (Figure 5C, left) further increased the magnitude of the transient activation of K_Ca, as well as the open probability of the channels (Figure 5D). With Bay K 8644, the probability of detecting transient behavior of K_Ca with Ca²⁺ in the pipette solution nearly doubled (9 out of 11; 82%). As in the absence of Bay K 8644, transient K_Ca activity was never detected in patches exposed to 1 µmol/L Bay K 8644 with no added Ca²⁺ (Figure 5C, right). The mean NP_o reported for patches exposed to Ca²⁺-free solution in the presence of Bay K (Figure 5D) was not significantly different from that obtained in the absence of the agonist (Figure 5B).

Because of the very small driving force for Ca²⁺ at +40 mV, it was not possible to detect simultaneous openings of K_Ca and I_CaL channels. Rubart et al³² have successfully recorded single-channel I_CaL with physiological [Ca²⁺]_o in the bathing medium at voltages negative to –10 mV. We have therefore attempted to record I_CaL and K_Ca channels simultaneously. Asymmetrical K⁺ gradient should produce a K_Ca current at negative voltages flowing in a direction opposite that of I_CaL through I_CaL channels, thus making it easy to discern the 2 types of unitary currents at the same voltage. In the presence of Bay K, extracellular Ca²⁺ was increased from 2 to 5 mmol/L to increase the amplitude of unitary I_CaL to a range at which it may be visualized next to K_Ca channel openings. Voltage-dependent K⁺ channels were blocked by including 5 mmol/L 4AP in the pipette solution. To exclude the interference by inward current flowing through Ca²⁺-dependent chloride channels, the pipette solution contained 10 µmol/L niflumic acid, a blocker of these channels in vascular myocytes,^{37 38} and a low concentration of IbTx (10 nmol/L) was included to counter the stimulatory effect of niflumic acid on K_Ca channels.³⁹

Figure 6Aa shows sample recordings of single-channel openings of I_CaL (downward openings) and K_Ca (upward openings) channels in response to 200-ms depolarizing steps (HP, -60 mV) to –30 mV. We were able to measure simultaneous openings of single I_CaL channels with K_Ca channels in 4 other patches similar to that displayed in panel Aa; however, we have detected many patches that displayed K_Ca openings at voltages negative to –20 mV but in which I_CaL channel openings were impossible to discern from background noise level. However, because of the numerous K_Ca channels present in each patch, and overlapped openings, it was impossible to establish a link between the openings of I_CaL channels and K_Ca channels. The P_o of K_Ca channels at this voltage is known to be close to 0 at resting subsarcolemmal [Ca²⁺]_i.⁴⁰ However, in elevated subsarcolemmal [Ca²⁺]_i the activation curve of K_Ca may be shifted to allow openings at –30 mV. In separate experiments carried out in the absence of niflumic acid and IbTx, we have measured an NP_o for K_Ca channels of 0.275±0.100 (n=5). Single I_CaL and K_Ca channels could not be observed when a high concentration (10 µmol/L) of nifedipine was added to the pipette solution (panel Ab, n=5). We estimated the mean open time of I_CaL channels in 3 cells in which K_Ca activity was low. The mean open time of I_CaL channels was in the range of 3.5 to 7 ms (5±2 ms). Because of the presence of multiple I_CaL channels within a patch, the mean open time may be overestimated. We observed an average number of maximum simultaneous openings of K_Ca in our patches to be 7±2 channels using voltage-ramp protocols, which would estimate the P_o to be 0.04.

Figure 6B shows the mean I-V relationship and corresponding slope conductances for unitary I_CaL (left, n=5 cells) and K_Ca (right, n=8 cells) currents in response to 200-ms voltage steps (HP, -60 mV) to various membrane potentials. The linear correlation is shown with its upper and lower 95% confidence bands giving slope conductances of 3.9±0.5 pS and 67±5 pS for I_CaL and K_Ca, respectively. These conductances are consistent with those reported by Rubart et al³² for I_CaL channels and by Morales et al³⁶ for K_Ca channels in similar conditions.

Is Transient K_Ca Activity Related to SR Function?
To test whether transsarcolemmal Ca²⁺ entry alone is sufficient to activate K_Ca, we compromised the function of the SR by incubating the myocytes for at least 30 minutes with ryanodine (Ryan; 10 µmol/L) and cyclopiazonic acid (CPA; 30 µmol/L), which are known to deplete the SR stores by impairing Ca²⁺ release and uptake mechanisms in smooth muscle cells in the same solutions used in Figure 5.^{23 41 42} With the SR compromised, we were still able to observe transient stimulation of K_Ca that was well maintained in time (Figure 7). Of 11 patches, 7 (64%) exhibited transient behavior, a proportion that was intermediate to that observed in the absence of CPA and Ryan (82%) and control conditions (43%). We also measured the time constant of relaxation of ensemble average currents to determine whether Ca²⁺ uptake into the SR influences the kinetics of the activity of the decline of K_Ca during the pulse. The time constant was 2.65±1.26 (n=5) and 1.10±0.42 seconds (n=6) for Bay K alone and Bay K+Ryan+CPA, respectively. Despite a tendency for a faster relaxation in the latter condition, the 2 values were not significantly different (P=0.237). These results indicate that the SR is not an essential prerequisite for the transient phenomenon.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

 
The main goal of our study was to assess whether Ca²⁺ entry through voltage-gated I_CaL channels can stimulate adjacent K_Ca channels by a localized interaction in rabbit coronary myocytes. During slow-voltage-ramp protocols, we identified an outward K_Ca current that is activated by a subsarcolemmal Ca²⁺ pool that is dissociated from bulk cytosolic Ca²⁺ pool (measured with indo 1) and is distinctly dependent on L-type Ca²⁺ channel activity. Transient activation of unitary K_Ca channels in cell-attached patches could be detected during long step depolarizations to +40 mV (HP, -40 mV), required the presence of Ca²⁺ in the pipette solution, and was enhanced by the I_CaL agonist Bay K 8644. This local interaction between K_Ca and I_CaL channels persisted after impairment of the SR. Furthermore, we provide the first direct evidence of simultaneous openings of single K_Ca and I_CaL channels in near-physiological conditions. Our data imply a novel sensitive mechanism for regulating RMP and tone in vascular smooth muscle.

Whole-Cell Experiments
It has long been recognized that macroscopic I_CaL can trigger a prominent transient K_Ca current in many smooth muscle cell types.^{11 43 44 45} Our data further extend this concept by demonstrating the existence of a local mode of regulation of vascular smooth muscle membrane potential and Ca²⁺ influx. We used ramp voltage-clamp protocols to investigate the relationship between K_Ca and I_CaL. In control conditions, the resulting biphasic quasi–steady-state I-V of net membrane current is composed of the following 3 major conductances: (1) a small inward I_CaL, (2) a 4AP-sensitive voltage-dependent K⁺ current that is partly responsible for the initial hump in the I-V observed at potentials in the range of -40 to +10 mV, and (3) a TEA-, IbTx-, and charybdotoxin-sensitive noisy Ca²⁺-dependent K⁺ current that exhibits prominent outward rectification at positive voltages with a significant contribution to the hump at potentials negative to +20 mV (this study and others²⁶ ). The nifedipine–sensitive and Ca²⁺-free–sensitive I-Vs revealed an initial net inward I_CaL between –60 and –40 mV, which in general reversed near –15 mV, a result similar to that previously described by our group.²⁶

The outward portion of the nifedipine-sensitive bell-shaped I-V was almost the mirror image of that expected for I_CaL with the ascending and descending portions mainly reflecting the gating properties of I_CaL and declining driving force for Ca²⁺, respectively. This markedly contrasted with that produced by cell exposure to Ca²⁺-free medium, which exhibited an initial hump and a prominent outwardly rectifying current. Whereas a short exposure to nifedipine (<10 minutes) revealed a bell-shaped difference I-V, a longer incubation led to an I-V similar to the one induced by Ca²⁺-free medium or IbTx and characterized by an initial hump followed by a late outwardly rectifying phase.²⁶ This hump disappeared from the IbTx-sensitive current when the cells were preincubated with nifedipine, which confirms the notion that the I-V of the nifedipine-sensitive outward ramp current is carried by K_Ca. Our data suggest that a short incubation likely reduced a component of outward current that is tightly linked to the function of I_CaL channels; as the incubation with nifedipine is prolonged, an overall depletion of subsarcolemmal [Ca²⁺]_i probably ensued. Enhancement of I_CaL with Bay K 8644 or elevation of extracellular [Ca²⁺] had effects similar to those of nifedipine or Ca²⁺ removal, albeit opposite in direction.

These contrasting effects of specific (dihydropyridines) and nonspecific ([Ca²⁺]_o) methods of altering transmembrane Ca²⁺ fluxes on outward K⁺ current suggested the existence of a compartmentalized restricted pool of Ca²⁺ that is accessible to both L-type Ca²⁺ and K_Ca channels. This concept of a functional subsarcolemmal pool of Ca²⁺ was advanced further by comparing simultaneous recordings of membrane current and bulk free [Ca²⁺]_i with combined whole-cell patch-clamp and microfluorometric techniques. While the I-V relationship of the nifedipine-sensitive outward current was bell shaped, 55% of the cells failed to reveal any change in bulk free [Ca²⁺]_i as estimated with indo 1; in the remaining cells displaying a small Ca²⁺ transient, the indo 1 ratio increased in a sigmoid fashion from –40 to +10 mV but failed to return to its preramp control level, as was observed for the nifedipine-sensitive K⁺ current.

Local Activation of Single K_Ca Channels by Nearby I_CaL Channels
Cell-attached patch clamp experiments were performed in intact cells to determine whether single K_Ca channels could be activated by adjacent I_CaL channels through a mechanism involving a local interaction. For the first time in smooth muscle, we were able to demonstrate transient activation of multiple levels of unitary K_Ca channels during step depolarizations. With physiological [Ca²⁺] in the pipette solution, 40% of the patches exhibited transient behavior in response to 5-second steps to +40 mV (HP, -40 mV). Such a transient activity was never observed when Ca²⁺ was omitted from the pipette solution. Moreover, the fact that Bay K 8644 enhanced both the probability of detecting transient activity and the open probability of K_Ca is consistent with the notion that K_Ca channels were stimulated by Ca²⁺ influx across adjacent I_CaL channels. The observation that in Ca²⁺-free pipette solution NP_o of K_Ca was not affected by Bay K 8644 indicates that it had a specific interaction with I_CaL channels rather than a nonspecific effect on K_Ca. These K⁺ channels are known to activate in a time-dependent manner at constant [Ca²⁺]_i but do not exhibit intrinsic voltage-dependent inactivation.^{16 26 36 40} During a depolarizing step, Ca²⁺ would cross the sarcolemma via I_CaL channels and result in a transient subsarcolemmal accumulation of Ca²⁺ which would then activate nearby K_Ca channels. The kinetics of their activation would depend on several factors, as follows: (1) gating kinetics of I_CaL³ (and in particular the slowed inactivation due to reduced Ca²⁺-dependent inactivation kinetics at higher membrane potentials should prolong the inactivation half-life of I_CaL, because a reduction in Ca²⁺-dependent inactivation [in the range of hundreds of milliseconds] shifts the dominant mode of inactivation toward voltage-dependent inactivation [in the range of seconds])^{46 47 48} ; (2) [Ca²⁺] kinetics within the subsarcolemmal space, which may depend on several factors (diffusion or binding of Ca²⁺ or both, active membrane transport systems, etc); (3) a possible Ca²⁺-dependent activation of enzymes (eg, via calmodulin-dependent protein kinase) that may phosphorylate K_Ca channels or associated regulatory proteins^{49 50} ; and (4) the rates of binding and unbinding of Ca²⁺, which determine the rates of activation and deactivation of K_Ca channels at fixed voltage.⁵¹

Convincing evidence is provided in experiments in which we demonstrated the simultaneous openings of K_Ca and I_CaL channels under near-physiological [Ca²⁺]_o and [K⁺]_o, as similarly reported in hippocampal neurons.⁵² Single-channel conductance for I_CaL channels in our preparation (3.9 pS) was comparable with that reported in cerebral arterial smooth muscle cells (4.6 pS).³² The identity of this channel is further supported by the fact that the Bay K induction of the channels was competitively antagonized by a high concentration of nifedipine. Similar to whole-cell experiments, Bay K was also able to enhance single K_Ca current at negative voltages, an effect that was completely inhibited by nifedipine. We conclude from these results that Ca²⁺ entry through I_CaL channels may produce sufficient accumulation of cytosolic Ca²⁺ to activate neighboring K_Ca channels. We have also found that Ca²⁺ entry alone is sufficient for the activation of K_Ca channels without an overall accumulation of Ca²⁺ into the bulk of the cytosol as determined by indo 1.

Is the SR Playing a Role?
Many studies have postulated the existence of a steep Ca²⁺ gradient between the sarcolemma and center core of smooth muscle cells, which has been attributed, at least in part, to the function of the superficial portions of the SR¹⁹ that were shown to make close contacts with specific areas of the sarcolemma (gaps of <50 nm have been detected in some cases).²⁰ A discrepancy exists between bulk [Ca²⁺]_i as estimated with fura 2 or indo 1 and subsarcolemmal [Ca²⁺]_i as probed from the activity of K_Ca under various conditions that affected the function of SR and I_CaL in bovine¹⁸ and guinea pig coronary myocytes³⁵ and guinea pig urinary bladder myocytes.²³ Activation and decay of K_Ca current induced by rapid caffeine exposure is faster than the resulting bulk Ca²⁺ transient.³⁵ Spontaneous transient outward K⁺ currents, which are believed to reflect the transient activation of Ca²⁺-dependent K⁺ channels by spontaneous unloading of Ca²⁺ from superficial SR elements,^{11 53} always preceded bulk Ca²⁺ transients in cells exhibiting global oscillations in [Ca²⁺]_i³⁵ and were kinetically correlated with the time course of near-membrane Ca²⁺ sparks as viewed by fluorescent confocal microscopy.³⁴ Our single-channel experiments showed that impairment of SR Ca²⁺ release and reuptake by preincubation of the myocytes with Ryan and CPA produced little if any effect on the probability of observing transient K_Ca activity. This suggests that under our conditions, the transient stimulation of K_Ca appeared to be independent of SR function. Despite a tendency for slower kinetics, the relaxation phase of the averaged transient K_Ca current was not significantly affected by CPA and Ryan, which possibly suggests that subsarcolemmal Ca²⁺ removal was little affected by SR function.

Limitations and Significance
Caution should be exercised in the interpretation of our single-channel data, as they do not allow us to speculate about the possible role of local Ca²⁺-induced Ca²⁺ release or Ca²⁺ uptake in modulating the activity of K_Ca under physiological conditions. Indeed, the myocytes were exposed to Ca²⁺-free, high-K⁺–depolarizing solution for long periods (generally more than an hour), which would inevitably lead to depletion of SR Ca²⁺ even before any experimental procedures were performed.

A similar type of local interaction has been proposed to exist in Helix neurons,⁵⁴ frog neuromuscular junctions,⁵⁵ and the presynaptic active zone of frog hair cells,⁵⁶ in which local interaction has been suggested to play an important role in modulating cell firing and synaptic transmission. In a recent study, single L-type Ca²⁺ channels were shown to only couple to small Ca²⁺-activated K⁺ or SK channels, whereas N-type Ca²⁺ channels seem to preferentially interact with maxi-K_Ca or BK channels.⁵² In those studies, Ca²⁺-dependent K⁺ and Ca²⁺ channels were postulated to be physically arranged or clustered into electrophysiological functional units that can regulate Ca²⁺ homeostasis. Although consistent with the notion that K_Ca and I_CaL channels may be colocalized on the membrane, without demonstration of a covalent link between the channels or direct visualization of the 2 proteins on the membrane, our data cannot be used to indicate whether the 2 channels are randomly distributed in the membrane or physically clustered into functional units, as proposed in other studies.^{52 54 55 56}

It is known that membrane depolarization can increase I_CaL in the whole cell. By extrapolation, if each very small I_CaL (relative to K_Ca channels) can have such influence on neighboring K_Ca channels, and if there are 4 times more K_Ca channels than I_CaL channels (estimates from our own data and those of others^{3 40} ), it is likely that the much larger K_Ca currents can summate to have a greater effect on membrane potential and thereby strongly influence surrounding channels. Feedback control of I_CaL channels by K_Ca would occur within a few milliseconds by deflecting membrane potential toward E_K, resulting in the closure of the I_CaL channels.

In conclusion, our study provides the first direct evidence of a functional local interaction between I_CaL channels and adjacent Ca²⁺-dependent K⁺ channels in coronary arterial myocytes. The localized effect between these channels likely reflects a more global feedback mechanism that potently controls Ca²⁺ entry and ultimately controls Ca²⁺ homeostasis in the cell.

	Acknowledgments

 
This project was funded by grants from the Medical Research Council of Canada, Québec Heart and Stroke Foundation, and Fond de Recherche de l'Institut de Cardiologie de Montréal (to N.L.) and by funds from the Natural Sciences and Engineering Research Council (to M.C.). A.G. is the recipient of a Medical Research Council-Novartis Fellowship. M.C. is a Fonds de la Recherche en Santé du Québec (F.R.S.Q.) junior scholar. N.L. is a F.R.S.Q. senior scholar. The authors gratefully acknowledge the help of Marie Andrée Lupien in preparing isolated cells and for her technical assistance.

Received April 20, 1998; accepted February 25, 1999.

	References

Top Abstract Introduction Materials and Methods Results Discussion References

 

1. Brayden JE, Nelson MT. Regulation of arterial tone by activation of calcium-dependent potassium channels. Science. 1992;256:532–535.[Abstract/Free Full Text]
2. Ganitkevich VY, Isenberg G. Contribution of two types of calcium channels to membrane conductance of single myocytes from guinea-pig coronary artery. J Physiol (Lond). 1990;426:19–42.[Abstract/Free Full Text]
3. Matsuda JJ, Volk KA, Shibata EF. Calcium currents in isolated rabbit coronary arterial smooth muscle myocytes. J Physiol (Lond). 1990;427:657–680.[Abstract/Free Full Text]
4. Langton PD, Standen NB. Calcium currents elicited by voltage steps and steady voltages in myocytes isolated from the rat basilar artery. J Physiol (Lond). 1993;469:535–548.[Abstract/Free Full Text]
5. Fleischmann BK, Murray RK, Kotlikoff MI. Voltage window for sustained elevation of cytosolic calcium in smooth muscle cells. Proc Natl Acad Sci U S A. 1994;91:11914–11918.[Abstract/Free Full Text]
6. Nelson MT, Patlak JB, Worley JF, Standen NB. Calcium channels, potassium channels, and voltage dependence of arterial smooth muscle tone. Am J Physiol. 1990;259:3–18.
7. Fleckenstein A. History of calcium antagonists. Circ Res. 1983;52:13–16.
8. Morgan KG. Role of calcium ion in maintenance of vascular smooth muscle tone. Am J Cardiol. 1987;59:24A–28A.[Medline] [Order article via Infotrieve]
9. Standen NB, Quayle JM, Davies NW, Brayden JE, Huang Y, Nelson MT. Hyperpolarizing vasodilators activate ATP-sensitive K⁺ channels in arterial smooth muscle. Science. 1989;245:177–180.[Abstract/Free Full Text]
10. Kajioka S, Kitamura K, Kuriyama H. Guanosine diphosphate activates an adenosine 5'-triphosphate-sensitive K⁺ channel in the rabbit portal vein. J Physiol (Lond). 1991;444:397–418.[Abstract/Free Full Text]
11. Hume JR, Leblanc N. Macroscopic K⁺ currents in single smooth muscle cells of the rabbit portal vein. J Physiol (Lond). 1989;413:49–73.[Abstract/Free Full Text]
12. Mekata F. The role of hyperpolarization in the relaxation of smooth muscle of monkey coronary artery. J Physiol (Lond). 1986;371:257–265.[Abstract/Free Full Text]
13. Hamilton TC, Weston AH. Cromakalim, nicorandil and pinacidil: novel drugs which open potassium channels in smooth muscle. Gen Pharmacol. 1989;20:1–9.[Medline] [Order article via Infotrieve]
14. Giudicelli JF, Larochelle CD, Berdeaux A. Effects of cromakalim and pinacidil on large epicardial and small coronary arteries in conscious dogs. J Pharmacol Exp Ther. 1990;255:836–842.[Abstract/Free Full Text]
15. Yanagisawa T, Teshigawara T, Taira N. Cytoplasmic calcium and the relaxation of canine coronary arterial smooth muscle produced by cromakalim, pinacidil and nicorandil. Br J Pharmacol. 1990;101:157–165.[Medline] [Order article via Infotrieve]
16. Nelson MT, Quayle JM. Physiological roles and properties of potassium channels in arterial smooth muscle. Am J Physiol. 1995;268:C799–C822.[Abstract/Free Full Text]
17. Singer JJ, Walsh JV Jr. Large conductance Ca²⁺-activated K⁺ channels in freshly dissociated smooth muscle cells. Membr Biochem. 1986;6:83–110.[Medline] [Order article via Infotrieve]
18. Stehnobittel L, Sturek M. Spontaneous sarcoplasmic reticulum calcium release and extrusion from bovine, not porcine, coronary artery smooth muscle. J Physiol (Lond). 1992;451:49–78.