INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

THE FUNCTIONAL ELEMENTS of resistance arteries include perivascular nerves, smooth muscle cells, and endothelial cells. Perivascular nerves and smooth muscle cells are electrically excitable, such that depolarization promotes calcium entry into the cytoplasm, leading to neurotransmitter release and vasoconstriction, respectively. Endothelial cells are generally thought to be electrically unexcitable because they typically do not exhibit voltage (i.e., depolarization)-activated currents (6, 15), though exceptions have been reported (2, 3). Nevertheless, endothelial cells exhibit robust hyperpolarization when activated by acetylcholine (ACh) (5, 17, 24), bradykinin (1), or shear stress (16), which often lead to hyperpolarization and relaxation of smooth muscle (10, 24). Whereas the stimuli that evoke hyperpolarization of endothelial cells can activate multiple signaling pathways, the effect of endothelial cell hyperpolarization, in and of itself, on vasoactive signaling has not been determined.

In the present study, we tested the hypothesis that endothelial cells can be hyperpolarized directly with a suitable electrical stimulus (i.e., independent of ligand-receptor interactions) and thereby evoke vasodilation. To investigate this relationship, we applied focal electrical stimuli across one end of isolated resistance arteries using glass microelectrodes. Stimulation parameters were adjusted to selectively activate perivascular nerves, smooth muscle cells, or endothelial cells. Once responses to focal field stimulation were evaluated, current was injected into a single endothelial cell to determine its effectiveness as a point source of electrical signaling for evoking vasomotor responses of the entire vessel segment. Feed arteries of the hamster retractor muscle were chosen for these experiments based upon their role in governing blood flow into the microcirculation (23, 26) and because of our ability to obtain isolated segments suitable for in vitro study (8, 9). We show that endothelial cells can be "activated" with electricity to evoke vasodilation and suggest that this approach may be useful in promoting tissue blood flow.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Animal care and use. Procedures were approved by the Animal Care and Use Committee of The John B. Pierce Laboratory and performed in accordance with the Guide for the Care and Use of Laboratory Animals of the National Research Council. Male Golden hamsters (40-50 days old; 80-100 g) were obtained from Charles River Breeding Laboratories (Kingston, NY). Hamsters were anesthetized with pentobarbital sodium (65 mg/kg ip); supplemental doses (20 mg/kg ip) were given as needed. After vessels were harvested (see Surgery), the hamster was given an overdose of pentobarbital intraperintoneally.

Solutions and drugs. Three physiological (pH 7.4) salt solutions were prepared before each experiment. Solution 1 (for superfusion of vessels during experiments) contained (in mmol/l) 148 NaCl, 4.7 KCl, 2.0 CaCl₂, 1.17 MgSO₄, 0.026 EDTA, 2.0 3-(N-morpholino)propanesulfonic acid (MOPS), 5.0 glucose, and 2.0 pyruvate. These reagents were obtained from Sigma (St. Louis, MO) or J. T. Baker (Phillipsburg, NJ). Solution 2 contained all of the above plus 1% albumin (no. 10856, Amersham; Cleveland, OH) and was used for perfusing the vessel lumen. Solution 3 contained 1% albumin but excluded CaCl₂ and was used during dissection and cannulation of the vessel.

Surgery. While we viewed through a stereo microscope (SV8, Zeiss; Thornwood, NY), a 2-cm incision was made over the left or right scapula, and the skin was retracted to expose the underlying retractor muscle of the cheek pouch (25). Superficial connective tissue was removed using microdissection, and the muscle was gently reflected to expose feed arteries extending from the thoracodorsal artery. Connective tissue around a feed artery and its adjacent collecting vein was trimmed away, and the vessel pair was then severed at proximal and distal ends and placed in a chilled (4°C) dissecting dish containing solution 3. The vein was pinned onto a Sylgard (Dow Corning; Midland, MI) surface to secure the vessel pair. The feed artery was dissected away with great care and transferred into a vessel chamber containing solution 3 (4°C).

Vessel cannulation. A water-jacketed vessel chamber (1 ml volume; Technical Services, Pierce Laboratory; New Haven, CT) contained the feed artery. A pair of cannulation pipettes (heat polished outer diameter, 30-60 µm) were pulled (P-97, Sutter Instruments; Novato, CA) from borosilicate glass capillary tubes (GC150T-10, Warner Instruments; New Haven, CT) and positioned in the vessel chamber using micromanipulators (MT-XYZ, Newport; Irvine, CA). Each pipette was filled with solution 2 and connected to a hydrostatic column mounted on a vertical pulley to control pressure. While we observed through the stereo microscope and used fine forceps, one end of the vessel was pulled onto a pipette and secured with 11-0 monofilament nylon suture (7715, Ethicon; Somerville, NJ). Pressure was raised to 10 cmH₂O to expel residual blood, and then the free end was secured onto the other pipette in a similar manner.

Electrical stimulation. Focal electrical field stimulation (EFS) was performed using two stimulus micropipettes filled with 0.9% saline. Each micropipette was connected via Ag/AgCl wires to the respective poles of a stimulation isolation unit (SIU5, Grass; Quincy, MA) driven by a square-wave stimulator (S48, Grass). At the downstream (with respect to superfusate flow) end of the vessel, the cathode was positioned in the adventitia, and the anode was positioned directly across the vessel (tip separation: ~150 µm; Fig. 1A). Stimulus pulses (1 ms at 16 Hz for 5 s) were delivered at "low" (30-50 V), "medium" (50-80 V), and "high" (80-110 V) intensity to directly activate perivascular nerves (14, 22), smooth muscle cells (21), and endothelial cells (see RESULTS), respectively. Direct current EFS (DCEFS; 30 V, 5 s) enabled stimulation of smooth muscle cells and endothelial cells without activating perivascular nerves (see RESULTS). When the stimulus micropipettes were elevated ~100 µm (so that the vessel was no longer directly between them), EFS produced no vasomotor response (n = 4), confirming that the electrical field was highly localized between the microelectrodes.

View larger version (12K): [in this window] [in a new window]   Fig. 1.   Vasomotor responses of isolated feed artery to progressive electrical stimulation. Illustration depicts ~300-µm long region in vicinity of stimulus micropipettes positioned toward the distal end of the isolated vessel segment. A: under control conditions, focal 16-Hz electrical field stimulation (EFS) at low voltage (V) activated perivascular nerves, causing constriction that spread along the entire 3- to 4-mm vessel segment. B: constriction was limited to the site of stimulation during perivascular nerve blockade (with tetrodotoxin, -conotoxin, or phentolamine). Low V (30-50) produced a one-sided dimple beneath the cathode, medium V (50-80) evoked circumferential constriction, and high V () evoked circumferential constriction with dilation along the rest of the vessel. C: direct current EFS (DCEFS) caused dilation similar to that with 16 Hz, high V stimuli, although constriction at the local site was less pronounced; responses were unaffected by tetrodotoxin, -conotoxin, or phentolamine.

Video microscopy. Images were acquired in brightfield using a Leitz L25 objective (numerical aperature = 0.35); illumination was provided by a 100-W halogen lamp (Nikon LWD condenser; numerical aperature = 0.52). The optical image was projected onto a video camera (KP-D50, Hitachi; Japan) and displayed on a video monitor (PVM-1343MD, Sony; Japan) at a total magnification of ×1,000. Internal diameter was measured (resolution, ~1 µm) using video calipers (Microcirculation Research Institute; College Station, TX) calibrated with a stage micrometer. The output of the calipers was directed to a data acquisition system (PowerLab 400, CB Sciences; Dover, NH).

Vasomotor responses. Vasomotor responses were recorded as changes in diameter; the magnitude of each response was calculated by subtracting the resting diameter from the peak response diameter. Responses were evaluated at the site of stimulation and at defined distances (determined with a calibrated eyepiece reticule) along the vessel upstream of the stimulus. The microscope rested on a motorized stage (MT-150MD, Warner), which enabled repositioning of the field of view without disturbing the vessel preparation or micropipettes secured to the fixed stage. Thus a given stimulus was repeated up to four times while sites were monitored at defined distances along the vessel. The order in which respective sites were studied was randomized between experiments; individual stimuli were separated by ~2 min to allow complete recovery. Preliminary experiments established that tachyphylaxis to these stimulus protocols was negligible for the duration of an experiment (2-4 h). All observations were made in the central 2 mm of vessel segments; this precaution minimized potential effects from cannulated ends.

Electrophysiology and cell labeling. Membrane potential was recorded with an electrometer (IE-210, Warner) using microelectrodes pulled (P-97, Sutter) from glass capillary tubing (GC100F-10, Warner). Electrode tips were filled with 1% propidium iodide (P-1304, Molecular Probes; Eugene, OR) in 2 mol/l KCl and backfilled with 2 mol/l KCl; tip resistance was 100-150 M. An Ag/AgCl pellet positioned in the chamber effluent served as the reference electrode. The output of the electrometer was connected to the data acquisition system and an audible baseline monitor (ABM-D, World Precision Instruments; Sarasota, FL). Data were acquired at 100 or 400 Hz.

Air bubble treatment. To disrupt the endothelium without damaging the smooth muscle cell layer (9), an air bubble (0.05 ml) was introduced into one of the cannulating pipettes, and pressure in the opposite cannula was reduced to 20 mmHg; the pressure gradient (55 mmHg) pushed the air bubble through the vessel lumen. After the bubble had passed, the lumen was reperfused with solution 2 reequilibrated for 30 min, and responses to ACh and PE microiontophoresis were then reevaluated (see RESULTS). DCEFS was then reapplied at the downstream end of the vessel while responses were reevaluated along the vessel.

Pharmacology. Tetrodotoxin (TTX, 10⁶ mol/l; Calbiochem; La Jolla, CA) or -conotoxin GVIA (CON, 1.5 × 10⁷ mol/l; Auspep; Louisville, KY) was used to inhibit perivascular nerves. Atropine (10⁶ mol/l; Sigma), phentolamine (Phe, 10⁶ mol/l; Research Biochemicals International; Natick, MA), nifedipine (10⁶ mol/l; Sigma), and N-nitro-L-arginine (L-NNA; 10⁴ mol/l; Sigma) were used to inhibit muscarinic receptors, -adrenoceptors, L-type calcium channels, and synthesis of nitric oxide, respectively. At the conclusion of each experiment, sodium nitroprusside (10⁵ mol/l; Sigma) was added to the vessel chamber to evaluate maximal diameter and intrinsic vasomotor tone.

Statistics. Summary data are presented as means ± SE. Statistical analyses were performed on a personal computer using SigmaStat (version 2.03, SPSS; San Rafael, CA) as detailed in RESULTS and the figures. Differences were considered statistically significant with P < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Vasomotor responses. Upon pressurization, isolated vessels (n = 55) first dilated to their maximal diameter (95 ± 2 µm) and then constricted to a stable resting diameter (63 ± 1 µm). At the site of microiontophoresis, vessels dilated (by 26 ± 2 µm; n = 26) to ACh and constricted (by 28 ± 3 µm; n = 24) to PE, confirming the integrity of endothelial cells and smooth muscle cells, respectively (9, 24). Electron microscopy confirmed that the smooth muscle layer was typically one or two cells thick (B. Doran and S. Segal, data not shown).

View larger version (16K): [in this window] [in a new window]   Fig. 2.   Vasodilation to focal DCEFS encompasses the entire feed artery. A: drawing (not to scale) of the vessel wall before (dotted line) and during (solid line) focal DCEFS with vasodilation (arrows). B: representative diameter responses at distances indicated along vessel during DCEFS. Horizontal bar indicates 5 s DCEFS. C: summary data showing the magnitude (peak  resting value) of diameter responses from 26 vessels (resting diameter = 70 ± 4 µm; maximal diameter = 105 ± 4 µm). Note that site of stimulation remained constricted relative to the rest of vessel (*P < 0.05; one-way ANOVA with Tukey's post hoc comparisons).

Electrophysiological responses. DCEFS produced hyperpolarization of endothelial cells and smooth muscle cells that conducted along the length of the vessel segment (Fig. 3). In both cell types, the hyperpolarization was similar in magnitude and lasted for the entire 5-s stimulus. The magnitude of vasodilation increased significantly with the magnitude of hyperpolarization (Fig. 4). To confirm that electrophysiological responses to DCEFS were true cellular events rather than stimulus artifacts, DCEFS was performed 1) during intracellular recording with the stimulus micropipettes moved above the vessel by 100 µm (n = 4), or 2) with the recording electrode placed in the adventitia rather than in a cell (n = 4). In both circumstances, only capacitance spikes were recorded coincident with the beginning and end of the 5-s stimuli.

View larger version (95K): [in this window] [in a new window]   Fig. 3.   Electrophysiological responses during DCEFS. Vessels (n = 19; resting diameter = 56 ± 2 µm; maximal diameter = 85 ± 4 µm) were stimulated for 5 s with focal DCEFS (horizontal bars) while diameter (D) and membrane potential (Em) were measured 1,000 µm upstream from the stimulus site. A: representative tracings of diameter and Em from endothelial cells. Summary data of response magnitudes are presented below each tracing (n = 12; Drest = resting diameter, D = change in diameter, Em,rest = resting membrane potential, Em = change in membrane potential). Fluorescence image (right) depicts vessel staining after electrode (containing propidium iodide) was lodged in an endothelial cell for several minutes. White lines represent vessel edges. Endothelial cell nuclei are oriented along vessel axis, and dye spreads readily from the impaled cell to nearby endothelial cells. B: representative tracings and summary data of diameter and Em from smooth muscle cells (n = 13) after impalement. Note shape and magnitude of Em responses are similar in smooth muscle and endothelial cells. Smooth muscle cell nucleus is oriented perpendicular to vessel axis and dye is limited to the cell recorded from (right). Scale bar = 20 µm.

View larger version (13K): [in this window] [in a new window]   Fig. 4.   Intensity dependence of vasomotor and electrophysiological responses to DCEFS. A: representative tracings of diameter and Em recorded from an endothelial cell 1,000 µm proximal to stimulus microelectrodes during stimuli of increasing intensity (indicated below respective 5-s scale bars). B: summary data (n = 12; resting diameter = 55 ± 3 µm; maximal diameter = 80 ± 4 µm) for peak diameter and Em responses to DCEFS. Electrophysiological responses were similar between endothelial cells (n = 10) and smooth muscle cells (n = 2) and therefore pooled (*main effect of stimulus intensity on Diameter and Em; P < 0.05, one-way ANOVA).

Current injection. Injection of 0.8 to 1.6 nA into an endothelial cell caused hyperpolarization and reproduced vasodilator responses to DCEFS (Fig. 5), confirming that endothelial cell hyperpolarization in and of itself produces vasodilation. Injection of negative current into smooth muscle cells caused similar vasodilation, whereas injection of positive current (0.8 to 1.6 nA) into either cell type resulted in vasoconstriction (8; present data not shown).

View larger version (13K): [in this window] [in a new window]   Fig. 5.   Diameter responses to injection of negative current into a single endothelial cell (EC). Drawing at top illustrates microelectrode impaled into an EC. Vessel wall is shown before (dotted line) and during (solid line) current injection. Representative traces of diameter (µm), membrane potential (mV), and injected current (nA) are shown in the top, middle, and bottom traces, respectively. The magnitude of hyperpolarization during current injection could not be measured because of incomplete bridge balance on the electrometer. Summary data from 4 vessels (resting diameter = 58 ± 2 µm, maximal diameter = 80 ± 4 µm) of the magnitude of local (distance = 0 µm) and conducted (at 1,000 and 2,000 µm) vasomotor responses are shown next to each diameter tracing.

Endothelial cell damage. Perfusion of a vessel with an air bubble caused visible rounding of endothelial cells and, in some regions, complete removal of the endothelial cell layer. Cell debris was then present in the vessel lumen and cannulating micropipettes. Strikingly, after air treatment, DCEFS no longer caused vasodilation along the vessel, and local vasoconstriction was enhanced relative to control (P < 0.05; Fig. 6, A and B). Air treatment abolished the dilator response to ACh (Fig. 6C) but did not significantly change resting diameter, the constrictor response to PE (Fig. 6D), or the dilator response to sodium nitroprusside (25 ± 5 vs. 16 ± 7 µm; n = 5).

View larger version (15K): [in this window] [in a new window]   Fig. 6.   Effect of EC damage on vasomotor responses to DCEFS. Feed arteries were perfused with an air bubble to disrupt endothelial cells. A: drawing of response to DCEFS before (control) and after treatment (air). B: summary data (n = 6; resting diameter 77 ± 14 µm, maximal diameter 105 ± 15 µm). EC damage abolished vasodilation to acetylcholine (ACh; C) but not vasoconstriction to phenylephrine (PE; D) and did not alter resting diameter (77 ± 14 µm). DCEFS delivered at distance = 0 µm in B (*P < 0.05, two-way repeated measures ANOVA with Tukey's post hoc comparisons).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The present experiments demonstrate selective electrical activation of perivascular nerves, smooth muscle cells, or endothelial cells of intact pressurized resistance arteries. Each stimulus was characterized by a particular vasomotor response that was selectively inhibited with established interventions. Whereas the activation of perivascular nerves and smooth muscle cells with electrical stimuli is well documented, we present novel evidence that endothelial cells can also be electrically "activated." In practice, direct hyperpolarization of even a single endothelial cell had the powerful effect of relaxing smooth muscle cells along the entire vessel.

Electrical stimulation of perivascular nerves. As expected (14, 21, 22), focal activation of perivascular nerves with repetitive (16 Hz) stimuli produced vasoconstriction along the entire vessel. In contrast, DCEFS failed to activate perivascular nerves, confirming that intermittent stimuli are required to maintain repetitive action potentials and sustain neurotransmitter release. Because the propagation of vasoconstriction was antagonized by -adrenergic blockade, our data indicate that norepinephrine was the primary vasoactive neurotransmitter released in these experiments. However, as stimulation parameters are changed, other neurotransmitters may be released (13). To eliminate this possibility with higher voltage or DC stimulation, vessels were treated with CON (27). In the presence of CON, responses were not different from those during exposure to TTX or Phe. Because neither atropine nor L-NNA altered responses to DCEFS, our findings argue further that neither ACh nor nitric oxide was released from perivascular nerves (20) or from endothelial cells (18).

Electrical activation of smooth muscle cells. In the presence of TTX, Phe, or CON, EFS directly activated smooth muscle cells in the vicinity of the microelectrode, whether using 16 Hz or DC stimuli. This vasomotor response was mediated by the opening of voltage-operated calcium channels as indicated by its sensitivity to nifedipine. Presumably, smooth muscle cell depolarization near the cathode gave rise to this response; however, local depolarization could not be measured because of the electrical artifact when stimulating and recording microelectrodes were <100 µm apart. Local vasoconstriction was graded (Fig. 1B), with low intensity EFS causing partial constriction (one-sided "dimpling") and medium intensity EFS causing pronounced circumferential constriction. In turn, this behavior suggests that low intensity EFS activated only those voltage-gated calcium channels closest to the tip of the cathode, whereas higher intensities activated smooth muscle cells circumferentially within the electrical field.

Electrical activation of endothelial cells. High intensity EFS or DCEFS resulted in hyperpolarization and vasodilation that conducted along the entire feed artery; responses were similar to those evoked by ACh (9). Whereas little is known of ionic currents of endothelial cells in vivo, voltage-operated (i.e., depolarization activated) ion channels have typically not been found in culture (6, 15). Nevertheless, endothelial cells abound with inward rectifying potassium channels (15). From the present data, we hypothesize that electrical stimulation activates these channels, giving rise to the conduction of hyperpolarization and vasodilation.