Voltage independence of vasomotion in isolated irideal arterioles of the rat

Abstract

The cellular mechanisms underlying vasomotion of irideal arterioles from juvenile rats have been studied using electrophysiological methods, ratiometric calcium measurements and video microscopy. Vasomotion was not affected by removal of the endothelium. Spontaneous contractions were preceded by spontaneous depolarizations. Both were abolished by the intracellular calcium chelator, BAPTA AM (20 μm), but not by ryanodine (10 μm), suggesting a dependence on the cyclical release of calcium from intracellular stores, other than those operated by ryanodine receptors. Oscillations were little changed when the membrane potential of short segments of arteriole was either depolarized or hyperpolarized. When the segments were voltage clamped, oscillating inward currents were recorded, indicating that the changes in membrane potential were voltage independent. Vasomotion was preceded by intracellular calcium oscillations and both were abolished by inhibitors of phospholipase C (U73122, 10 μm), phospholipase A₂ (AACOCF₃, 30 μm) and protein kinase C (chelerythrine chloride, 5 μm, and myristoylated protein kinase C peptide, 10 μm). Inhibition of vasomotion by the dual lipoxygenase and cyclo-oxygenase inhibitor, NDGA (10 μm), the lipoxygenase inhibitor, ETI (1 μm) but not by the cyclo-oxygenase inhibitors, aspirin (10 μm) and indomethacin (10 μm), or the cytochrome P450 inhibitor 17-ODYA (10 μm), suggested an involvement of the lipoxygenase pathway. The observations suggest that vasomotion of iris arterioles is voltage independent and results from the cyclical release of calcium from IP₃-sensitive stores which are activated by cross talk between the phospholipase C and phospholipase A₂ pathways in vascular smooth muscle.

Rhythmical contractions or vasomotion have been observed in vivo in vessels of the microcirculation, as well as in larger vessels such as the basilar artery (Auer & Gallfhoffer, 1981; Hundley et al. 1988; Fujii et al. 1990b; Morita-Tsuzuki et al. 1992; Bertuglia et al. 1994). This contractile activity is considered to play an important role in the regulation of vascular resistance (Gratton et al. 1998) and in the maintenance of nutritive perfusion under low blood pressure and ischaemic conditions, even improving flow in neighbouring tissues which do not show such activity (Hudetz et al. 1998; Rucker et al. 2000). Vasomotion is independent of neural activity (Hayashida et al. 1986; Bouskela & Grampp, 1992; Lee & Earm, 1994; Hill et al. 1999), although neurotransmitters can induce similar vasomotor activity in quiescent vessels (Gustafsson, 1993; Omote & Mizusawa, 1993). Recent studies indicate that vasomotion is upregulated in hypertension (Shimamura et al. 1999).

The mechanisms underlying vasomotion are heterogeneous. While contraction of vascular smooth muscle is dependent on increases in intracellular calcium concentration, this can result from voltage dependent and independent events (Hirst & Edwards, 1989). The former depends on a change in membrane potential which subsequently leads to the opening of voltage dependent calcium channels, while in the latter case, calcium is released from intracellular stores, often following the activation of intracellular second messenger pathways (Somlyo & Somlyo, 1994). Rhythmical contractions have been described to result from at least three different mechanisms; one based entirely on voltage-dependent events, such as in the rabbit femoral artery, rat thoracic aorta and basilar artery (Fujii et al. 1990a; Omote & Mizusawa, 1995), one involving entirely voltage-independent intracellular mechanisms, such as in the arterioles of the rat iris (Hill et al. 1999) and the other relying on both voltage dependent and independent events, such as in small arteries of the rat mesentery, and in non-vascular smooth muscle tissues, such as the guinea-pig lymphatics and the guinea-pig urethra (Gustafsson & Nilsson, 1993; Van Helden, 1993; Hashitani & Edwards, 1999). In a number of vessels, additional factors including the influence of the endothelium, nitric oxide, eicosanoids and blood pressure, have been implicated (Fujii et al. 1990b; Chemtob et al. 1992; Gustafsson, 1993; Bertuglia et al. 1994; Omote & Mizusawa, 1995; von der Weid et al. 1996; Hudetz et al. 1998; Hill et al. 1999).

The spontaneous, rhythmical depolarizations and contractions recorded in developing iris arterioles in vitro (Hill et al. 1999) were suggested to be independent of voltage since they were not blocked by dihydropyridines, antagonists of voltage-dependent L-type calcium channels (Hill et al. 1999). Recently, however, a new class of voltage-dependent calcium channels, which are not blocked by dihydropyridines, has been shown to be present in arterioles (Morita et al. 1999). Since the spontaneous contractions in iris arterioles were preceded by large spontaneous depolarizations, we have readdressed the issue of their voltage independence using current and voltage clamp studies. We have also used pharmacological tools to investigate the role of the intracellular signal transduction pathways involving phospholipase C and phospholipase A₂ and have made direct measurements of calcium changes to determine the site of action of these compounds. The data suggest that the vasomotion is independent of voltage changes but instead results from the cyclical activation of the phospholipase C and phospholipase A₂ pathways and the cyclical release of intracellular calcium from the IP₃ store. Some of the results have previously been presented in abstract form (Haddock & Hill, 1999).

METHODS

In vitro tissue preparation

All experiments have been performed in accordance with the guidelines of the Animal Experimentation Ethics Committee of the Australian National University. Wistar rats of either sex, aged 14–17 days postnatal, were anaesthetized with ether and killed by cervical dislocation. Both eyes were removed and the iris was dissected from each eye and cut in half. Each section of iris, containing on average two intact, unpressurized arterioles, was pinned with tungsten wire along the ciliary and sphincter muscle edges, onto a Sylgard silicone resin (Dow Corning Corporation, Midland, MI, USA) coated coverslip which formed the base of a 1 ml recording chamber. The recording chamber was perfused at a constant flow (3 ml min⁻¹) with Krebs solution, composition (mm): 120 NaCl; 5 KCl; 25 NaHCO₃; 1 NaH₂PO₄; 2.5 CaCl₂; 2 MgCl₂; 22 glucose; gassed with 95 % O₂ and 5 % CO₂ and maintained at 34–37 °C. Since blood vessel function has been suggested to be altered in the presence of high glucose concentrations (Mene et al. 1993), control experiments were performed in the presence of a lower glucose concentration (11 mm). No significant differences in the frequency or amplitude of spontaneous contractions were found in Krebs solution containing 22 mm glucose compared to those in 11 mm glucose (frequency, 3.7 ± 0.1 and 3.6 ± 0.1 min⁻¹; amplitude, 23.2 ± 3.1 % resting vessel diameter and 24.2 ± 2.9 % resting vessel diameter respectively; n = 4; P > 0.05).

Measurements of contractions and membrane voltage

Arterioles, located midway between the ciliary and sphincter edges of the iris (see Fig. 1 in Gould & Hill, 1994), were visualized using video microscopy. The vessel diameter (20–35 μm in the stretched preparation), was measured continuously using DIAMTRAK computer tracking program (Neild, 1989). Recordings were digitized and stored for later analysis. All preparations were equilibrated for 20 min or until spontaneous activity was constant in frequency and amplitude. In all experiments, observations were obtained from a minimum of four preparations, each from a different rat. The effects of stretching the arterioles during the pinning procedure were evaluated by comparing the characteristics of the contractions of two populations of arterioles of significantly different diameter, all taken from 14-day-old rats. The degree of tone exhibited by spontaneously contracting arterioles was determined by the addition of the vasodilator, SNAP (10 μm).

Since previous studies have shown that spontaneous contractions showed no significant change in frequency or amplitude during the first hour of incubation (Hill et al. 1999), experiments were completed within this time frame. Preparations were incubated in drug solutions for a maximum period of 20 min before being returned to control Krebs solution.

In some experiments, arteriolar smooth muscle cells were impaled with sharp microelectrodes, filled with 0.5 m KCl, and having resistances of 120–220 MΩ (Flaming Brown micropipette puller, Sutter Instrument Co.). Membrane potential records were low-pass filtered (cut-off frequency 1 kHz) and amplified with an Axoclamp 2B (Axon Instruments). Simultaneous changes in membrane potential and changes in vessel diameter in the region where the cell was impaled were acquired with sample rates of greater than 100 Hz and stored on computer disk for analysis.

Membrane potentials were also recorded from short segments of arteriole, which were cut to have lengths of about 300 μm (mean 328 ± 19 μm, n = 10), using electrodes which had been dipped in silicone fluid to reduce the electrode tip capacitance. When membrane potentials were measured in single electrode current clamp mode (SECC), the electrode capacitance was compensated, by monitoring the head stage voltage. Typically, switching frequencies of 200 to 500 Hz were employed, the head-stage voltage was continuously monitored when current was passed to change the membrane potential, so ensuring that the settling of tip capacitance did not distort the voltage recording. When the voltage clamp technique was used, using a single electrode voltage clamp, the same precautions were taken.

Measurement of intracellular calcium

Measurements of changes in global intracellular concentration of calcium ions ([Ca²⁺]_i) were made using the ratiometric fluorescent dye, Fura-2 acetoxymethyl ester (Fura-2 AM). Preliminary experiments showed that the Fura-2 AM did not penetrate the cells of the arteriolar wall when the vessel remained within the iris. Moreover vessels did not contract if dissected entirely free from the wall of the iris. Consequently, the iris was pinned to the bottom of the recording chamber and a small area of dilator muscle and stroma surrounding the arteriole was gently removed to expose the arteriole to the Fura solution. The iris preparation was then incubated at room temperature (22 °C) in modified Krebs solution containing reduced calcium (0.5 mm CaCl₂), the detergent pluronic F-127 (0.01 %) and Fura-2 AM (5 μm) for 40 min. After loading, preparations were superfused with warmed (34 °C) Krebs solution, containing 2.5 mm CaCl₂, for 20 min and only arterioles that exhibited stable spontaneous contractions were studied further.

Preparations loaded with Fura-2 AM were illuminated with alternating wavelengths of 340 and 380 nm, at a frequency of 30 Hz (Polychrome II illumination system; T.I.L.L Photonics GMBH, Germany). Care was taken to limit the exposure to the fluorescent light. Emitted light was passed through a 510 nm bandpass filter to a photometry system (T.I.L.L Photonics GMBH, Germany). Data were recorded and analysed using pClamp 8 (Axon Instruments, Inc., USA). Changes in [Ca²⁺]_i are expressed as the ratio of the fluorescence emission recorded at 340 nm and 380 nm excitation which is proportional to [Ca²⁺]_i. Simultaneous measurements of vessel diameter were made by illuminating the preparation with infrared light (775 nm) and collecting continuous images with a Hamamatsu High Performance Vidicon camera.

Impairment of endothelium function

To investigate the role of the endothelium in the generation of spontaneous contractions, rats were deeply anaesthetized with intraperitoneal injection of 44 mg kg⁻¹ ketamine and 8 mg kg⁻¹ xylazine. The descending aorta was exposed and clamped with artery forceps so that blood flow to the abdominal cavity and hind limbs was occluded. Animals were perfused (1.25 ml min⁻¹) via the innominate artery, which supplies the right upper extremity, and the right atrium was cut to allow blood to escape. The perfusate, composed of Krebs solution, also contained 1 % sodium nitrite, to cause dilation of blood vessels and 1 % bovine serum albumin and 0.1 % heparin, to prevent blood clot formation. When the exudate from the atrium was clear, the endothelium was disrupted by switching the perfusion fluid to distilled water for 4 min. Following this, the animal was reperfused with Krebs solution for a final 4 min. The right eye was then removed, the iris dissected free and set up as described above. Following the 20 min incubation period, arteriolar activity was recorded for 5 min. The tissue was then incubated in the α₂-adrenoceptor agonist UK-14304 (100 nm) for 4 min to increase vessel tone (39.2 ± 5.3 % of resting vessel diameter; n = 9), before acetylcholine (ACh; 1 μm, 10 μm) was added to the bath to assess the effectiveness and selectivity of the endothelial cell damage. Control preparations containing vessels with intact endothelium were also prepared in the same manner (n = 14), except for perfusion with water. The effect of Ach (1 μm, 10 μm) was also assessed in these preparations in the presence of UK-14304 (100 nm) to increase vessel tone (40.6 ± 2.9 % of resting vessel diameter; n = 14). Results confirmed that the perfusion with water had successfully impaired the endothelium since vasodilatation following addition of ACh (1 μm, 10 μm) was not significantly different (15.1 ± 5.5 %, 25.3 ± 5.5 %, n = 9) from that occurring with time in intact preparations in the absence of ACh (15.4 ± 6.2 %, 27.3 ± 8.5 %; n = 5, P > 0.05). Application of ACh (1 μm, 10 μm) to vessels with intact endothelium, however, produced a significant reduction (88.5 ± 3.1 %, 99.3 ± 2.5 %, n = 9; P < 0.05) in the amplitude of the UK-14304 constriction.

Analysis of results

The frequency of spontaneous contractions in both control and drug solutions was determined over a 5 min time period. The amplitude of spontaneous contractions was expressed as a percentage of resting vessel diameter and was the average of 15 consecutive contractions. Measurements were made at 20 min, immediately following the equilibration period, and at 45 min, the time point at which drugs would have been present in the tissue bath for 20 min. The effects of drugs were determined by expressing spontaneous contractions in drug solution as a percentage of the contractions in control Krebs solution. Results are given as mean ± s.e.m of n preparations where each preparation was from a different rat. Statistical analysis was determined using 95 % confidence limits (P < 0.05) and one-way ANOVA followed by pairwise t tests using Bonferroni correction for multiple groups. Paired or unpaired t tests were used when comparing two data groups. Data analysis and production graphs were performed using the scientific statistical package, GraphPad Prism.

Drugs and solutions

The following drugs were used: (1,2-bis(o-aminophenoxy) ethane-N,N,N’,N'-tetaacetic acid tetra (acetoxymethyl) ester (BAPTA AM), 1-(6-((17β-3-methoxyestra-1, 3, 5(10)-trien-17-yl) amino) hexyl)-1H-pyrrole-2, 5-dione (U73122), 1-(6-((17β-3-methoxyestra-1, 3, 5 (10)-trien-17-yl) amino) hexyl)-2,5-pyrrolidine-dione (U73343), myristoylated protein kinase C (18–28) (MPK-C), (BIOMOL, USA); chelerythrine chloride, indomethacin, nordihydroguaretic acid (NDGA), acetylsalicyclic acid (aspirin), acetylcholine (ACh), S-nitroso-N-acetylpenicillamine (SNAP), 4,4′-diisocthiocyanatostilbene-2,2′-disulphonic acid (DIDS), (Sigma Chemical Co., USA); 17-octadecynoic acid (17-ODYA), arachidonyl trifluoromethyl ketone (AACOCF₃), 5,8,11-eicosatrienoic acid (ETI), (Fluka Chemie, Switzerland), 2-aminoethyldiphenyl borate (2-APB), (Calbiochem, Germany), Fura-2 AM (Molecular Probes, USA) and UK14304–18, (Pfizer Central Research, England).

Stock solutions of U73122, U73343, chelerythrine chloride, BAPTA AM, DIDS, indomethacin and aspirin were dissolved in DMSO; NDGA, 17-ODYA, SNAP, 2-APB and ETI were dissolved in ethanol and UK-14304–18 in 2 % acetic acid. All other drugs were made up as ×1000 stock solutions in distilled water and diluted into Krebs solution, except AACOCF₃ which was purchased as a solution in ethanol and diluted directly into Krebs solution. Control experiments showed no significant effects of DMSO (0.001 %, 0.0001 %) or ethanol (0.001 %) on spontaneous activity or resting vessel diameter. Light-sensitive drugs were protected through the use of long wavelength (> 610 nm) light.

RESULTS

General observations

The percentage of preparations exhibiting spontaneously active arterioles was similar to that previously described by Hill et al. (1999). The average frequency of spontaneous contractions was 3.2 ± 0.1 min⁻¹ (n = 64). The average amplitude of spontaneous contractions for all preparations in control Krebs solution was 14.7 ± 0.6 % resting vessel diameter (n = 64), where the mean resting vessel diameter was 26.3 ± 0.7 μm (n = 64). In experiments where SNAP (10 μm) was added to maximally relax vessels, the tone of spontaneously active preparations was determined to be 7.5 % of resting vessel diameter (n = 4). Stretching of the vessels during pinning of the iris had no effect on spontaneous contractions. In arterioles with significantly different resting vessel diameters (38.8 ± 1.3 μm and 21.4 ± 0.9 μm; P < 0.05), there was no significant difference in the frequency (3.2 ± 0.3, n = 6 and 3.3 ± 0.2 min⁻¹; n = 8, respectively) or amplitude (14.3 ± 2.2 and 13.9 ± 1.7 % resting vessel diameter, respectively) of spontaneous contractions.

Role of voltage-dependent mechanisms

When intracellular recordings were made from the vessels, each contraction was seen to be preceded by a wave of membrane depolarization (Fig. 1A and B). Depolarizations had peak amplitudes of 31 ± 1.8 mV (n = 24) and the resting membrane potential (RMP) of arterioles showing vasomotion was −60 ± 1.1 mV (n = 24).

Figure 1. Effect of BAPTA AM on spontaneous depolarizations and contractions of iris arterioles.

In control Krebs solution (A and B), spontaneous depolarizations precede spontaneous contractions. After the addition of BAPTA AM (20 μm), both depolarizations and contractions are rapidly abolished (C and D). Voltage scale bar applies to the top two traces and diameter scale bar applies to the bottom two traces while time scale bars apply to all records.

Ionic basis of the spontaneous depolarizations

The intracellular calcium chelator, BAPTA AM (20 μm) caused a rapid abolition of the spontaneous depolarizations and contractions (P < 0.05, Fig. 1C and D). Although the vessels relaxed, no significant change in RMP was detected (−63.3 ± 1.4 mV, n = 6). In the presence of BAPTA, an increase in the extracellular concentration of K⁺ (50 mm) caused a large depolarization and contraction, as previously described (Hill & Gould, 1997).

The calcium-dependent chloride channel inhibitor DIDS (100 μm) abolished both the spontaneous depolarizations and contractions (Fig. 2). This was accompanied by a small but significant hyperpolarization (RMP in DIDS −68.2 ± 2.5 mV, RMP in control −62.6 ± 3.2 mV, n = 5, P < 0.05, paired t test, Fig. 2C) and relaxation (Fig. 2D). A parallel series of experiments was carried out on Fura loaded tissues in which changes in [Ca²⁺]_i and arteriolar diameter were monitored. Under this protocol, approximately 50 % of the preparations failed to initiate spontaneous contractile activity, presumably because Fura itself buffered [Ca²⁺]_i to low levels. However in the preparations where spontaneous contractions were recorded, oscillations in [Ca²⁺]_i were also observed (Fig. 3Aa and b). Oscillations in [Ca²⁺]_i always preceded spontaneous contractions. The mean delay between the peak of the [Ca²⁺]_i oscillation and the peak of the contraction was 1.4 ± 0.07 s (n = 20) With time in vitro, the basal level of [Ca²⁺]_i showed a trend to an increase and the resting vessel diameter a trend to a decrease (Fig. 3Ab and a). [Ca²⁺]_i oscillations and the contractions were unaffected by nifedipine (1 μm; n = 4; P > 0.05, Fig. 3Bb and a). On the other hand DIDS (100 μm) abolished the [Ca²⁺]_i oscillations (Fig. 3Bb) and associated contractions (P < 0.05, Fig. 3Ba).

Figure 2. Effect of Cl_Ca channel inhibition on spontaneous depolarizations and contractions.

Addition of the calcium-dependent chloride channel inhibitor, DIDS (100 μm) abolished depolarizations (A and C) and contractions (B and D), producing relaxation (D) and a small hyperpolarization (C) of the iris arteriole (P < 0.05). Voltage scale bar applies to the top two traces and diameter scale bar applies to the bottom two traces, while the time scale bar applies to all records.

Figure 3. Inhibition of L-type voltage-dependent calcium channels and Cl_Ca channels.

A, time control in which changes in wall [Ca²⁺]_i (b) and vessel diameter (a) are recorded at 20 min and at 40 min. B, nifedipine had no effect on spontaneous activity, while the addition of DIDS (100 μm, n = 4,) abolished both spontaneous [Ca²⁺]_i oscillations (b) and contractions (a) and caused the arteriole to relax. Control resting vessel diameters were 24.5 μm and 26.2 μm and basal [Ca²⁺]_i 340/380 ratio was 0.48 and 0.44 in A and B, respectively.

Voltage changes in isopotential segments

Isolated segments of arteriole, cut to have lengths of about 300 μm, also generated rhythmical depolarizations like those recorded from intact arteriolar trees (Fig. 4). These segments had resting membrane potentials of −59 ± 2.0 mV (n = 12). The preparations had input resistances in the range 27 to 110 MΩ (55.6 ± 8.0 MΩ; n = 10). If the preparations had similar electrical properties to those determined previously in iris arterioles (Hirst et al. 1997), their electrical lengths would be less than half of a length constant. Since the electrode was placed in the centre of the preparation, the maximum attenuation along the preparations would be less than 5 % and the preparations will approach isopotentiality except for very rapidly rising transients (Hirst & Neild, 1980).

Figure 4. Lack of effect of membrane potential of a myogenically active short segment of irideal arteriole on spontaneous oscillations.

The resting membrane potential of the preparations was −55 mV. When the membrane potential was either depolarized or hyperpolarized over a range of potentials from −40 to −99 mV, rhythmical depolarizations persisted without changing their frequency of occurrence. The time and voltage scale bars apply to all traces.

In eight preparations, taken from five different animals, the effect of changing the membrane potential on the frequency of rhythmical depolarizations was examined. In each preparation, depolarizing or hyperpolarizing the preparation from its resting potential failed to abolish the rhythmical depolarizations and produced no consistent change in their frequency of occurrence (Fig. 4). At normal resting potential the frequency of slow waves was 4.9 ± 0.2 waves min⁻¹ (n = 8); after hyperpolarizing the membrane potential by about 20 mV (21.3 ± 2.4 mV), the frequency of slow waves was 5.0 ± 0.3 slow waves min⁻¹. These observations suggest that the spontaneous depolarizations are unlikely to result from the sequential activation of voltage-dependent ion channels.

Voltage clamp of short segments

When preparations were voltage clamped at their peak negative membrane potentials, rhythmical inward currents were detected (Fig. 5). Before applying the voltage clamp, the rate of occurrence of spontaneous depolarizations was 5.5 ± 0.6 waves min⁻¹ (n = 7), their mean amplitude was 20.5 ± 4.1 mV. After applying the voltage clamp, currents occurred at 5.5 ± 0.6 waves min⁻¹ and they had peak amplitudes of 0.35 ± 0.03 nA. When the current records were compared with the membrane potential recordings it was apparent that not only did they occur at the same frequency but also that they had very similar time courses. Visual observation of the preparation indicated that rhythmical movements of the preparations continued whilst the voltage clamp was applied. These observations indicate that during each entire depolarization, there is a flow of current which does not depend upon the activation of voltage dependent ion channels. The failure to inhibit vasomotion suggests that cyclic changes in [Ca²⁺]_i occur even when membrane potential changes are prevented.

Figure 5. Rhythmical discharge of inward current recorded from a segment of irideal arteriole.

The upper pair of traces (A) show the rhythmical changes in membrane potential before and the error signal detected after the application of a voltage clamp using a single electrode voltage clamp circuit. It can be seen that when the membrane potential change was controlled to less than 2.5 % of the control responses, a rhythmical inward current was detected (B). This current had the same frequency of occurrence as the rhythmical depolarization. The time bar applies to all traces.

Role of intracellular calcium

Ryanodine (10 μm), applied for up to 30 min, had no effect on spontaneous contractions (n = 5). In the presence of both ryanodine and caffeine (1 mm), which was used to deplete ryanodine-sensitive calcium stores, spontaneous contractions were rapidly abolished (Fig. 6B). Spontaneous contractions reappeared after the caffeine was washed out, even though ryanodine (10 μm; Fig. 6C) continued to be present in the physiological saline.

Figure 6. Effect of ryanodine on spontaneous contractions.

In the presence of caffeine (1 mm) and ryanodine (10 μm), spontaneous contractions were rapidly abolished (A and B). When caffeine was washed out, spontaneous contractions reappeared in spite of the continued presence of ryanodine (C). Scale bars apply to all records.

The phospholipase C inhibitor, U73122 (10 μm; n = 6), the protein kinase C inhibitor, chelerythrine chloride (5 μm; n = 4) and the myristoylated protein kinase C peptide (MPK-C; 10 μm; n = 7) reduced both the frequency (P < 0.05; Fig. 7A) and amplitude (P < 0.05; Fig. 7B) of spontaneous contractions. The inactive isomer of U733122, U73343 (10 μm; n = 5) had no effect on spontaneous contractions (P > 0.05; Fig. 7A and B). Non-specific effects of U73122 were seen at 100 μm when the inactive U73343 also inhibited the vasomotion. The IP₃ receptor antagonist, 2-APB (60 μm; n = 4) caused a rapid but transient abolition of spontaneous contractions, after which spontaneous contractions were significantly increased in frequency (3.6 ± 0.2 and 4.3 ± 0.3 min⁻¹, P < 0.05, paired t tests) and decreased in amplitude (13.2 ± 0.5 and 8.2 ± 0.1 % resting vessel diameter, P < 0.05). Both [Ca²⁺]_i oscillations and contractions were abolished following incubation in U73122 (Fig. 7Cb and a) and MPK-C (Fig. 7Db and a).

Figure 7. Effect of the phospholipase C on Ca²⁺ oscillations and spontaneous contractions.

The effect of the phospholipase C pathway on the frequency (A) and amplitude (B) of spontaneous contractions was analysed 20 min after the addition of the phospholipase C inhibitor, U73122 (10 μm), the inactive isomer of U73122, U73343 (10 μm), the protein kinase C peptide, myristoylated protein kinase C (MPK-C, 10 μm) and the protein kinase C inhibitor, chelerythrine chloride (CC, 5 μm). Values represent means and s.e.m. * Significantly different from control. C, U73122 (10 μm) and (D) MPK-C (10 μm) abolished both spontaneous [Ca²⁺]_i oscillations (b) and contractions (a). Traces are representative of four experiments using tissues from different animals. Control resting vessel diameters were 20.4 μm and 18.7 μm and basal [Ca²⁺]_i 340/380 ratios were 0.44 and 0.68 in C and D, respectively.

The phospholipase A₂ inhibitor, arachidonyl trifluoromethyl ketone (AACOCF₃; 30 μm; n = 6), the dual lipoxygenase and cyclo-oxygenase inhibitor, nordihydroguairetic acid (NDGA; 10 μm; n = 10) and the lipoxygenase inhibitor, 5,8,11-eicosatrienoic acid (ETI; 1 μm; n = 7) all produced significant decreases in the frequency (P < 0.05; Fig. 8A) and amplitude (P < 0.05; Fig. 8B) of spontaneous contractions. The cyclo-oxygenase inhibitors, indomethacin (10 μm; n = 8) and aspirin (10 μm; n = 4) and the cytochrome P450 inhibitor, 17-ODYA (10 μm; n = 4) had no effect on the frequency of spontaneous contractions (Fig. 8A), however, the amplitude of spontaneous contractions was reduced in the presence of 17-ODYA (P < 0.05; Fig. 8B). Parallel experiments using Fura loaded arterioles showed that inhibition of the phospholipase A₂ pathway with AACOCF₃ (30 μm; n = 4) abolished [Ca²⁺]_i oscillations and associated contractions (Fig. 8Ca and b).

Figure 8. Effect of the phospholipase A₂ pathway on spontaneous contractions.

The phospholipase A₂ inhibitor, AACOCF₃ (30 μm), the dual lipoxygenase and cyclo-oxygenase inhibitor, NDGA (10 μm), the lipoxygenase inhibitor, ETI (1 μm), the cyclo-oxygenase inhibitors, indomethacin (Indo, 10 μm) and aspirin (10 μm) and the cytochrome P450 inhibitor, 17-ODYA (10 μm) were tested for their effects on the frequency (A) and amplitude (B) of spontaneous contractions. Values represent means and s.e.m. * indicates a significant difference from control. C, AACOCF₃ (30 μm) abolished spontaneous [Ca²⁺]_i oscillations (b) and contractions (a). Traces are representative of four experiments using tissues from different animals. The control resting vessel diameter was 19.9 μm and basal [Ca²⁺]_i 340/380 ratio was 0.54.

Role of the vascular endothelium in spontaneous contractions

In order to determine whether spontaneous contractions in irideal arterioles could result from the spontaneous release of endothelium-derived substances, spontaneous activity was studied in the presence and absence of the vascular endothelium. Disruption of the vascular endothelium had no significant effect (n = 9) on the frequency (3.1 ± 0.1 min⁻¹) or amplitude (9.9 ± 1.1 % of resting vessel diameter, which was 28.4 ± 2.2 μm) of spontaneous contractions when compared to those vessels with intact endothelium (3.0 ± 0.2 min⁻¹, 11.2 ± 1.4 % of resting vessel diameter, which was 26.4 ± 1.5 μm, n = 14; P > 0.05).

DISCUSSION

Like the rhythmical activity recorded in hamster cheek pouch arterioles in vivo and pig coronary artery strips in vitro (Segal & Beny, 1992; von der Weid & Beny, 1993), the spontaneous, rhythmical contractions developed by isolated irideal arterioles were always preceded by rhythmical membrane potential fluctuations. Where such activity is dependent on the activation of voltage-dependent ion channels in the cell membrane, for example in cardiac muscle, it can be prevented by hyperpolarization or by applying a voltage clamp (Brown et al. 1984). In contrast, when short isopotential segments of iris arterioles were examined, neither the frequency of spontaneous depolarizations nor the associated contractions were altered by membrane depolarization or hyperpolarization. Furthermore, under voltage clamp conditions, spontaneous inward currents were recorded at frequencies identical to those of the original voltage oscillations, indicating that the underlying currents were independent of voltage changes. This voltage independence of the contractions in these vessels suggests that they would be little altered by pressurization which results in depolarization of the resting membrane potential.

As the preparations consisted of a monolayer of smooth muscle cells, apparently unconnected to the endothelium (Hirst et al. 1997), complicating factors related to the control of the membrane potentials of complex syncytia do not apply. Furthermore, since the segments of arteriole were cut to have physical lengths much shorter than their electrical length constants (Hirst et al. 1997), the changes in membrane potential imposed at the centre of the arteriolar segment would be little, if at all, attenuated at the ends of the segments. Thus the experiments show that the rhythmical activity of iris arterioles results from the cyclical activation of sets of ion channels in the virtual absence of a potential change. Since the intracellular calcium chelator, BAPTA AM, abolished both spontaneous rhythmical depolarizations and contractions both would appear to depend on the cyclical release of Ca²⁺ from an internal store.

The oscillating depolarizations were abolished by the chloride channel antagonist, DIDS, suggesting at first that Cl⁻ channels may be activated by the cyclical increases in [Ca²⁺]_i. However, as DIDS also abolished rhythmical changes in [Ca²⁺]_i and associated contractions, it may that DIDS has additional effects on Ca²⁺ handling within arteriolar smooth muscle cells, either a non-selective action on intracellular calcium stores or, alternatively, blocking chloride channels in the sarcoplasmic reticulum which may function to maintain electrical neutrality across the store (Kawano & Hiraoka, 1993).

While ryanodine had no effect on the vasomotion of iris arterioles, inhibition of the phospholipase C pathway abolished both spontaneous [Ca²⁺]_i oscillations and contractions, suggesting that IP₃ stores and not ryanodine stores are important (Gould & Hill, 1996; Hill et al. 1999). The phospholipase C pathway has been shown to be important in the generation of rhythmical activity of other smooth muscles, such as the guinea pig gastric pylorus (Van Helden et al. 2000) and in the generation of calcium oscillations in canine pulmonary smooth muscle cells (Hamada et al. 1997). Surprisingly, however, the cell-permeable compound, 2-APB, which has been demonstrated to be involved in inhibition of IP₃ receptor-mediated calcium release and capacitive calcium re-entry (Maruyama et al. 1997; Ma et al. 2000; Potocnik & Hill, 2001), had little effect on the rhythmical contractions. This may be due to the involvement of a subtype of IP₃ receptor with reduced sensitivity to 2-APB, as has recently been reported (Kukkonen et al. 2001). Alternatively, 2-APB may have differing effects on the various subtypes of transient receptor potential (TRP) channels which may be involved in calcium influx (Putney & McKay, 1999).

Like the effects of phospholipase C inhibition, the specific cytosolic phospholipase A₂ inhibitor AACOCF₃ (LaBelle & Polyak, 1998) abolished the spontaneous contractions, as did antagonists of the lipoxygenase pathway. On the other hand, spontaneous contractions were resistant to cyclo-oxygenase inhibitors, while inhibition of the cytochrome P450 pathway, by 17-ODYA, caused a small reduction in the amplitude, but not the frequency, of contractions. The effect of 17-ODYA is consistent with a minor contractile effect of hydroxyeicosatetraenoic acids (HETEs) in smooth muscle cells rather than a relaxant effect mediated by epoxyeicosatrienoic acid production from endothelium (Campbell & Harder, 1999). The lipoxygenase pathway is reported to produce heterogeneous responses in vascular smooth muscle, with metabolites demonstrating both potent contractile (Scriabine et al. 1990) and relaxant effects (Barlow et al. 2000; Faraci et al. 2001). Such divergent responses may arise because lipoxygenase metabolites can activate receptors on either smooth muscle cells to produce contraction (Back et al. 2000) or on endothelial cells to produce relaxation (Walch et al. 1999). The failure of endothelial disruption to eliminate arteriolar contractions suggests that the lipoxygenase metabolites in the present study are active in the smooth muscle cells and not in the endothelial cells. The success of the procedure for damaging the endothelium was demonstrated by the absence of ACh-induced vasodilatation.

Since phospholipase A₂ inhibition produced an abolition of the intracellular [Ca²⁺]_i oscillations, as did inhibition of the phospholipase C pathway, it appears that this pathway is also integral to the generation of the calcium oscillations which underlie the depolarizations and contractions. Lipoxygenases convert arachidonic acid into hydroperoxyeicosatetraenoic acids (HPETEs) that can be metabolized into a variety of potent signalling molecules including leukotrienes, HETEs and trihydroxyeicosatrienoic acids (THETEs), all of which have vasoactive properties that are mediated at the cell surface on receptors or channels (Brash, 1999). In a variety of cell types (Mong et al. 1988; Bouchelouche et al. 1990; Sjolander et al. 1990; Oliva et al. 1994; Bouchelouche et al. 2001), leukotrienes have been shown to bind to G protein coupled membrane receptors that stimulate the phospholipase C/IP₃ cascade and the mobilization of [Ca²⁺]_i. It is possible to speculate that a similar mechanism, involving the production of leukotrienes, is occurring in the iris arteriole. Since recent evidence suggests that heterogeneity exists within the leukotriene receptor population (Labat et al. 1992; Coleman et al. 1995; Back et al. 2000) characterization of the active lipoxygenase metabolite must await future studies.

In addition to IP₃, activation of phospholipase C generates diacylglycerol which leads to the translocation and activation of protein kinase C (see Orallo, 1996 for review). In the present study, inhibition of protein kinase C abolished the spontaneous contractions. Simultaneous abolition of the underlying [Ca²⁺]_i oscillations suggests that the effects of protein kinase C are at the level of the calcium stores, although the possibility of additional modulation at the level of phosphorylation of myosin light chain cannot be excluded. Since inhibitors of phospholipase C or phospholipase A₂ also independently abolished [Ca²⁺]_i oscillations and vasomotion, it appears that cross talk must occur between the two pathways. The possibility exists that this may be brought about through protein kinase C, since up-regulation of phospholipase A₂ activity can occur through protein kinase C activation (Garcia et al. 1992).

Based on the data presented here, we propose a mechanism by which intracellular calcium is responsible for the spontaneous, rhythmical contractions generated by the irideal arterioles (Fig. 9). Constitutive activity of the phospholipase C pathway leads to a basal level of calcium, arteriolar tone and upregulation of the phospholipase A₂ pathway through activation of protein kinase C. Metabolism of arachidonic acid through the lipoxygenase pathway produces vasoactive substances which can augment the release of calcium from the IP₃ store. The augmented calcium release leads to a modulation of subsequent calcium release in a cyclical manner as has been previously described (Iino & Tsukioka, 1994). The calcium oscillations produce oscillating contractions and, coincidentally, activate calcium-dependent channels in the cell membrane to produce oscillatory voltage changes.

Figure 9. Mechanism underlying spontaneous contractions of iris arterioles.

Constitutive activity of the phospholipase C (PLC) pathway leads to the production of IP₃ and a basal release of calcium and arteriolar tone. Simultaneous activation of protein kinase C (PKC) in turn activates phospholipase A₂ (PLA₂) and breakdown of arachidonic acid via the lipoxygenase (LOX) pathway. The resultant metabolites further stimulate the PLC pathway resulting in cyclical oscillations of intracellular calcium and contractions. Intracellular calcium oscillations also trigger voltage oscillations of the cell membrane through an unknown ion channel.