Kv1.5 is a major component underlying the A-type potassium current in retinal arteriolar smooth muscle

Abstract

Little is known about the molecular characteristics of the voltage-activated K⁺ (K_v) channels that underlie the A-type K⁺ current in vascular smooth muscle cells of the systemic circulation. We investigated the molecular identity of the A-type K⁺ current in retinal arteriolar myocytes using patch-clamp techniques, RT-PCR, immunohistochemistry, and neutralizing antibody studies. The A-type K⁺ current was resistant to the actions of specific inhibitors for K_v3 and K_v4 channels but was blocked by the K_v1 antagonist correolide. No effects were observed with pharmacological agents against K_v1.1/2/3/6 and 7 channels, but the current was partially blocked by riluzole, a K_v1.4 and K_v1.5 inhibitor. The current was not altered by the removal of extracellular K⁺ but was abolished by flecainide, indicative of K_v1.5 rather than K_v1.4 channels. Transcripts encoding K_v1.5 and not K_v1.4 were identified in freshly isolated retinal arterioles. Immunofluorescence labeling confirmed a lack of K_v1.4 expression and revealed K_v1.5 to be localized to the plasma membrane of the arteriolar smooth muscle cells. Anti-K_v1.5 antibody applied intracellularly inhibited the A-type K⁺ current, whereas anti-K_v1.4 antibody had no effect. Co-expression of K_v1.5 with K_vβ1 or K_vβ3 accessory subunits is known to transform K_v1.5 currents from delayed rectifers into A-type currents. K_vβ1 mRNA expression was detected in retinal arterioles, but K_vβ3 was not observed. K_vβ1 immunofluorescence was detected on the plasma membrane of retinal arteriolar myocytes. The findings of this study suggest that K_v1.5, most likely co-assembled with K_vβ1 subunits, comprises a major component underlying the A-type K⁺ current in retinal arteriolar smooth muscle cells.

smooth muscle tone in resistance arteries and arterioles is an important determinant of peripheral vascular resistance and blood pressure (29). By controlling membrane potential and the level of free intracellular Ca²⁺ available to the contractile apparatus, voltage-activated K⁺ channels (K_v) are crucial in regulating arterial smooth muscle tone (38). Based on electrophysiological studies, two major K_v current components have been identified in vascular smooth muscle cells; namely, the delayed rectifier K⁺ current and the transient A-type K⁺ current. However, this broad classification fails to reveal the highly diverse nature of the K_v channels that underlie these currents, and in fact, a host of delayed rectifier and A-type K⁺ currents have been reported in arterial smooth muscle with distinct biophysical and pharmacological properties (9). This diversity stems in part from the large number of gene families that encode the pore-forming K_vα subunits (K_v1–11) but also from other processes such as alternative splicing, heterotetrameric assembly of members within the same K_vα subfamily, and interaction of the channels with accessory subunits (8). Despite this complexity, recent efforts have been directed toward identifying the molecular composition of the K_v channels that underlie functional K_v currents in arterial smooth muscle, and evidence has begun to emerge suggesting that some delayed rectifier currents may arise as a consequence of heterotetrameric assembly of distinct K_v1α subunits (2, 4–7, 15, 22, 26, 27, 31, 39). The molecular composition of A-type currents in myocytes of the systemic circulation is not clear at this time (9); however, K_v3.4 and K_v4 channels appear to be the most likely determinants of these currents in pulmonary arterial smooth muscle (20).

K⁺ channel α-subunits with A-type properties are found in several K_v channel subfamilies, including Shaker (K_v1.3, K_v1.4, and K_v1.7), Shaw (K_v3.3 and K_v3.4), and Shal (K_v4.1, K_v4.2, and K_v4.3), and transcripts for all of these subunits have been detected in vascular smooth muscle (9). In addition, co-expression of K_vβ1 or Kβ3 accessory subunits with certain K_v1α subunits confers rapid inactivation on otherwise noninactivating delayed rectifier channels (18, 25, 33). K_vβ subunit expression has been identified in several types of arterial smooth muscle (9). Rapid inactivation of A-type K_v channels is thought to be conferred by “ball” domains in the NH₂ termini of the relevant K_vα subunits and K_vβ1 and K_vβ3 subunits (N-type inactivation) (32). Upon depolarization, these domains occlude the channel pore by binding to a receptor near or at the cytoplasmic side of the pore (21).

We have recently described a rapidly inactivating A-type current in retinal arteriolar smooth muscle cells, which appears to play a physiological role in suppressing cell excitability and contractility (28). In the present study we have used this preparation as a model system to begin to resolve the molecular nature of the A-type current in myocytes of the peripheral systemic circulation. By combining a range of pharmacological, molecular, immunohistochemical, and neutralizing antibody approaches, we show that K_v1.5, most likely co-assembled with K_vβ1, comprises a principal component underlying the A-type K⁺ current in this tissue.

MATERIALS AND METHODS

Retinal arteriole preparation.

Male Sprague-Dawley rats (200–300 g) were anesthetized with CO₂ and killed by cervical dislocation. Retinas were rapidly removed, and arterioles, devoid of surrounding neuropile, were isolated as previously described (37). In brief, retinas were lightly triturated using a fire-polished Pasteur pipette (internal tip diameter 0.3 mm) in a low Ca²⁺ Hanks' solution. The resulting homogenates were centrifuged at 2,800 rpm (952 g) for 1 min, the supernatant was aspirated off, and the tissue was washed again with low Ca²⁺ medium. The suspension was then stored at 21°C until needed. Arteriolar segments remained useable for up to 10 h under these conditions.

Patch-clamp recording.

Voltage-clamp experiments were performed using the whole cell-perforated patch-clamp technique as previously described (28). One milliliter of homogenate was pipetted into a rectangular glass-bottomed recording bath on the stage of an inverted microscope (Nikon Eclipse, TE2000). Arterioles were anchored down with tungsten wire slips (50 μm diameter, 2 mm length) and continuously superfused with normal Hanks' solution at 37°C to remove extraneous retinal tissue from the bathing medium. Before the electrophysiological recording, vessels were digested for 20-min with an enzyme cocktail of collagenase 1A (0.1 mg/ml) and protease type XIV (0.01 mg/ml) to remove surface basal lamina. Enzyme and drug solutions were delivered via a seven-way micromanifold with an exchange time of 1 s as measured by switching over to dye solution. The flow from the manifold into the bath was through a single tube (350 μm in diameter, 6 mm in length, 0.2-μl volume) long enough to allow the temperature to equilibrate with the solution flowing through the bath. Gigaseals were formed directly on arteriolar smooth muscle cells still embedded within their native arterioles. Electrodes (1–2 MΩ in free bathing solution) were pulled from filamented borosilicate glass capillaries (1.5 mm od w 1.17 mm id, Clark Electromedical Instruments, UK), and the patch-clamp amplifier used was an Axopatch-1D (Axon Instruments, Foster City, CA). The pipette solution was K⁺ based with amphotericin B as the perforating agent (see Drugs and solutions). Contaminating currents through Ca²⁺-activated K⁺ and Cl⁻ channels were minimized in all experiments by including 100 nM Penitrem A and 1 mM 9-anthracene carboxylic acid in the external bathing medium. Recordings were delayed until full perforation of the membrane patch had been achieved, as judged from the development of repeatable currents in response to step depolarizations: this usually took 3–5 min. Recordings were low pass filtered at 0.5 kHz and sampled at 2 kHz by a National Instruments PC1200 interface using software provided by John Dempster (University of Strathclyde, UK). Leakage currents were subtracted off-line from the active currents with the use of the standard leak subtraction protocol contained within the Patch software suite. Liquid junction potentials (<2 mV) were compensated electronically. Series resistance (∼30 MΩ) was routinely compensated by >70%. For determination of whole cell current densities, cell membrane capacitance was determined from the time constant of a capacitance transient elicited by a 20-mV depolarization from −80 mV with a sampling frequency of 20 kHz.

To evaluate the effects of pharmacological agents, outward A-type K⁺ current was first elicited by voltage steps from −80 mV to +20 mV (500 ms, 10-s intervals) in drug-free medium and then in bath solution containing the drug of interest. We have previously shown that the A-type K⁺ current in retinal arteriolar smooth muscle is closely approximated by the peak minus the sustained components of the net outward current in Penitrem A and 9-anthracene carboxylic acid (28). With the use of this method to isolate the A-type K⁺ current, the peak current densities for each vessel was averaged from five voltage steps before and then after drug application, and changes were expressed as a percentage for a minimum of four vessels per treatment group. For experiments using antibodies, conventional whole cell recordings were performed. Pipettes were dipped in an antibody-free intracellular solution and then back filled with the pipette solution containing the antibody of interest [anti-K_v1.4_(589–655) or anti-K_v1.5_(513–602); Alomone Labs, Jerusalem, Israel]. Aliquoted antibodies were defrosted daily to avoid degradation. Because the antibody experiments did not permit pairwise comparisons to be made within the same vessel, to improve resolution of changes across treatment groups, voltage steps were applied from −80 mV to +80 mV, where average peak current densities in control vessels are twice those observed at +20 mV.

Characterization of patch-clamp preparation.

Each rat retina yielded approximately two to three first-order arterioles (35–50 μm diameter, 50–500 μm in length). The arterioles were uncontracted and easily distinguished from venules by their thick wall of circularly arranged smooth muscle cells (Fig. 1). As noted above, just before the patch-clamp recording, vessels were anchored down in the recording bath and digested for 20 min with an enzyme cocktail of collagenase and protease. In addition to removing the basal lamina to facilitate gigaseal formation, this treatment electrically uncouples the endothelial cells from the overlying arteriolar smooth muscle (28). In other types of arterioles, the smooth muscle cells are also known to be electrically coupled via gap junctions (10, 19, 41, 42). In our earlier work, we did not investigate whether the retinal arteriolar smooth muscle layer remains electrically coupled following enzyme treatment. An indirect method to evaluate this is to compare cell capacitance measurements with the estimated cell surface area for a single arteriolar myocyte. In the present study, cell capacitance was measured as 13.1 ± 0.48 pF (n = 31), indicating a total patch-clamped membrane surface area of ∼1,310 μm². The dimensions of individual retinal arteriolar myocytes from first-order arterioles were estimated by confocal scanning laser microscopy in vessels loaded for 10 min with the membrane-tracking dye di-4-ANEPPs (10 μM) (11); average dimensions for length (based on the vessel circumference), width, and height were 121.3, 5.7, and 2 μm, respectively (21 cells; n = 4 vessels). When we used these values and assumed a scalene ellipsoid structure, the approximate surface area for a single retinal arteriolar myocyte was calculated according to the Knud Thomsen formula:

(1)

where p = lg(3) = ln(3)/ln(2) and a, b, c are the semiaxes.

Fig. 1.

Photomicrographs of a freshly isolated rat retinal arteriole and venule. Scale Bars, 10 μm.

The approximated cell surface area using this method was calculated to be 1,220 μm², suggesting our electrophysiological recordings are most likely confined to individual myocytes. To further verify this, experiments were undertaken using the gap junction inhibitor 18β-glycyrrhetinic acid (18β-GA). In mesenteric arterioles, 40 μM 18β-GA causes a rapid block of electrical communication within the smooth muscle layer, as denoted by a switch from predominantly slow to fast capacitative transients (41). In the present study, no changes in the capacitative currents were observed in enzyme-digested arterioles exposed to 100 μM 18β-GA (capacitances were 12.53 ± 0.81 pF and 11.86 ± 0.97 pF, before and after 18β-GA, respectively; n = 9; P = 0.18). Taken together, the above results strongly suggest that following collagenase and protease treatment retinal arteriolar smooth muscle cells within intact vessel segments are electrically uncoupled from their neighboring cells.

PCR gene amplification.

Retinal homogenates were placed in a 2-ml recording chamber on the stage of an inverted microscope and between 5 and 13 vessels collected for each PCR experiment using single tungsten wire slips (50 μm in diameter, 5 mm length). Total RNA was extracted using RNeasy minikit (Qiagen, Crawley, UK) according to the manufacturer's protocol. Total RNA was also extracted from brain pia. Samples were split into two aliquots, and first-strand cDNA was prepared from one aliquot using Sensiscript Reverse Transcription kit (Qiagen). The other aliquot was used in an equivalent reaction lacking enzyme to control for potential genomic or extraneous DNA contamination [no reverse transcriptase (RT)]. The cDNA RT products were amplified with K_v1.4-, K_v1.5-, K_vβ1-, and K_vβ3-specific primers by RT-PCR using Qiagen HotStar Taq reagents. The primer pairs, relevant Genbank entries, and expected product sizes are listed in Table 1. All products were resolved on 2.5% agarose gels and visualized by ethidium bromide fluorescence.

Table 1.

Genbank entries, primer sequences, and expected product sizes

MRNA (Accession No.)	Primer Sets	Expected Size, BP
Kv1.4 (NM012971)	For. GGAAATTAACTTTTGAAAAGGCTGCT	216
	Rev. TTGGTGCGTTAGTAAACATTCACAG
Kv1.5 (NM012972)	For. TGAAGGAAGAACAAGGCAACCA	167
	Rev. GGCCTTGAGGTGACACTTTTCTAGA
Kvβ1 (NM017303)	For. GCACACTACCTCAGCTGGCTGT	176
	Rev. GGGCTTGTTGCGCAATATGT
Kvβ3 (NM031652)	For. CACAAAATTGGTGTTGGCTCAGT	134
	Rev. TGCACTTTTTCTTTGAGCCACTG

Immunohistochemistry.

For immunofluorescence experiments, retinal arterioles were visualized while embedded within retinal flatmount preparations. The advantage of this technique over the use of isolated arterioles is that negative results within the vasculature can be cross-referenced with other cell types in the retina to gauge the degree of antibody reactivity. Freshly enucleated eyes were placed in low Ca²⁺ Hanks' solution, and the anterior segment lens complexes were removed. The posterior eye cup was fixed in 4% paraformaldehyde for 20 min and washed extensively in phosphate-buffered saline (PBS) throughout a 4-h period. Retinas were subsequently detached and soaked in PBS containing 0.5% Triton X-100 to permeabilize the tissue and 5% normal donkey serum (Chemicon International, Temecula, CA) to block nonspecific binding of the primary antibody. K_v1.4, K_v1.5, and K_vβ1 were detected in separate experiments using rabbit polyclonal antibodies targeted to rat, mouse, and human sequences, respectively [anti-K_v1.4_(589–655) and anti-K_v1.5_(513–602); Alomone Labs, Jerusalem, Israel; anti-K_vβ1_(311–360); abcam, Cambridge, UK]. Tissue was incubated in primary antibody at a dilution of 1:100 (anti-K_v1.4 and anti-K_v1.5) or 1:200 (anti-K_vβ1) in the permeabilization buffer overnight at 4°C and then extensively washed for 4 h at 21°C. A 1:200 dilution of a donkey anti-rabbit IgG labeled with Alexa-488 (Molecular Probes Europe BV, Leiden, The Netherlands) was used for secondary detection; the staining time and subsequent washing were the same as that used for the primary antibody. The specificity of the antibodies was investigated by parallel control experiments in the absence of primary antibody. To facilitate rapid identification of arterioles within the retinal neuropile, propidium iodide nuclear stain was used. Retinas were incubated in 5 nmol/l of propidium iodide (Invitrogen, Carlsbad, CA) in PBS for 30 min at 37°C. Retinas were flattened by placing four radial cuts from the retinal periphery to points within 1 mm from the optic disk and mounted in Vectashield (Vector Laboratories, Burlingame, CA). Images were acquired using an Olympus BX60 fluorescence microscope (Olympus, London, UK) fitted with a MicroRadiance confocal-scanning laser microscope (Bio-Rad).

Drugs and solutions.

Hanks' solution contained (in mM) 140 NaCl, 5 KCl, 5 d-glucose, 2 CaCl₂, 1.3 MgCl₂, 10 HEPES; pH 7.4 with NaOH. Low Ca²⁺ medium differed only in that it contained 0.1 mM CaCl₂. K⁺-free medium was the same as normal Hanks' solution without 5 mM KCl. For patch-clamp recordings the pipette solution contained (in mM) 138 KCl, 1 MgCl₂, 0.5 EGTA, 10 HEPES (pH adjusted to 7.2 using NaOH) to which 300 μg/ml amphotericin B was added in perforated patch mode.

Amphotericin B, collagenase 1A, penitrem A, protease type XIV, 9-anthracene carboxylic acid, 18β-glycyrrhetinic acid, flecainide, and riluzole were purchased from Sigma-Aldrich (Poole, UK). Correolide was a kind gift from Dr. Maria Garcia and Dr. Jianming Bao of Merck Research Laboratories, Rahway, NJ. Phrixotoxin-2, rHeteropodatoxin-2, rHongotoxin-1, rNoxiustoxin, and BDS-I were obtained from Alomone Labs.

Data analysis.

Data are reported as means ± SE; n refers to the number of vessels tested. Significant differences between control and experimental treatments were determined using the paired t-test. Antibody experiments were analyzed using one-way ANOVA. P values <0.05 were considered significant.

RESULTS

Pharmacology of the A-type K⁺ current in retinal arteriolar smooth muscle.

Over recent years there has been a substantial increase in the number of toxins available that inhibit K_v channels. Taking impetus from this, we tested a range of pharmacological blockers as a first step in resolving likely K_v channel components underlying the A-type K⁺ current in retinal arteriolar myocytes. Initially, we screened agents that selectively block A-type K_vα subunits within the main K_v channel subfamilies. Cells were held at −80 mV, and command voltage steps to +20 mV were applied. Phrixotoxin-2 and heteropodatoxin-2 are peptides from spider venoms that specifically inhibit K_v4 channels (12, 36). Neither of these toxins applied at concentrations higher than reported IC₅₀ values affected the A-type K⁺ current in retinal arteriolar myocytes nor did BDS-I, a K_v3.4 channel antagonist (13) (see Fig. 2 and Table 2). Correolide is a novel nortriterpine from the Costa Rican tree Spachea correa that selectively binds to and blocks members of the K_v1 channel subfamily (14, 17). In four retinal arterioles, 10 μM correolide inhibited the peak A-type K⁺ current by >70% (Fig. 2 and Table 2). In accord with its slow-onset kinetics (14), and consistent with previous reports in cerebral arteriolar myocytes (2), the blocking effects of correolide on the K_v current developed slowly taking 5–10 min to reach maximal levels.

Fig. 2.

Typical records showing the effects of selective A-type K_vα subfamily inhibitors on the A-type K⁺ current in retinal arteriolar smooth muscle cells. K_v currents were elicited by voltage steps from −80 to +20 mV before and during application of the relevant antagonist. Arterioles were in the continuous presence of 100 nM Penitrem A and 1 mM 9-anthracene carboxylic acid (9-AC) to prevent contamination by Ca²⁺-activated K⁺ and Cl⁻ currents, respectively. No effects on the peak A-type K⁺ current were observed with the K_v4 channel blockers heteropodatoxin-2 (A) and phrixotoxin-2 (B) or the K_v3.4 antagonist BDS-I (C). In contrast, the A-type K⁺ current was blocked by 10 μM correolide (D), a selective inhibitor of K_v1 channels. Average data for these experiments are presented in Table 2.

Table 2.

Mean data showing the effects of specific Kv subfamily inhibitors on the A-type K⁺ current in retinal arteriolar myocytes

Inhibitor	Kvα Subunits Inhibited	Mean Block±SEM, %	P Value (No. of Vessels)
Heteropodatoxin 2	Kv 4.2	−4.77±7.28*	NS (4)
Phrixotoxin 2	Kv 4.2, 4.3	−0.54±1.97*	NS (4)
BDS 1	Kv 3.4	0.4±2.05	NS (4)
Correolide	Kv 1.x	76.1±5.23	0.00696 (4)

Peak current densities for each vessel were averaged from 5 voltage steps before and after drug application, and changes expressed as a percentage.

^*Negative % block indicates an increase in current. NS, not significant; i.e. P > 0.05.

The above data support the idea that Kv1 channels may underlie the A-type K⁺ current in retinal arteriolar myocytes. The K_v1 channel subfamily consists of multiple members (K_v1.1–1.7), and several K_v1α subunits are known to display A-type properties, including K_v1.3, K_v1.4, and K_v1.7 (3). Moreover, the accessory subunit K_vβ1.1 can induce rapid inactivation of delayed rectifier type K_v1 channels, including K_v1.1, K_v1.2, and K_v1.5, but not K_v1.6 (18). Since most K_v1 channels could potentially represent molecular candidates of the A-type K⁺ current in retinal arteriolar myocytes, we extended our pharmacological approach using known inhibitors that specifically target various members of this subfamily. Hongotoxin is a peptide from venom of the scorpion Centruroides limbatus, which is known to block homotetrameric K_v1.1, K_v1.2, and K_v1.3 channels (IC₅₀s in 31–170 pM range) and also K_v1.6 with low affinity (IC₅₀ = 6 nM) (23). In retinal arteriolar smooth muscle, 100 nM recombinant hongotoxin did not suppress the A-type K⁺ current (Fig. 3, Table 3). To further eliminate a possible contribution by Kv1.2–1.3 homotetramers, we tested the effects of noxiustoxin, another scorpion-derived toxin that binds to and blocks these channels at low concentrations (IC₅₀ values of 1 and 2 nM, respectively) (16). K_v1.7 is also potently blocked by noxiustoxin at nanomolar levels (IC₅₀ of 18 nM) (8). Recombinant noxiustoxin (20 nM) had no effect on the A-type K⁺ current in retinal arterioles (Fig. 3, Table 3). Riluzole is a neuroprotective drug that modulates K_v1.4 channels via the oxidation of a cysteine residue in the NH₂-terminal inactivation ball (IC₅₀ of 70 μM) (40). More recently, it has also been shown that this drug inhibits cloned K_v1.5 channels by preferentially binding to the inactivated and to the closed states of the channel (IC₅₀ 40 μM) (1). Riluzole (100 μM) inhibited the A-type K⁺ current by 32 ± 7% at +20 mV (Fig. 3, Table 3). To begin to isolate the possible contributions of K_v1.4 and K_v1.5 channels to the A-type K⁺ current, we investigated the effects of K⁺-free bathing medium and the anti-arrhythmic agent flecainide (Fig. 3, Table 3). Removal of extracellular K⁺ is known to suppress current through K_v1.4 channels but not K_v1.5 (30), and in retinal arteriolar myocytes, exposure to K⁺-free medium elicited no change in the peak A-type K⁺ current. Classically, flecainide has been used to discriminate Kv4-based A-type K⁺ currents (flecainide sensitive; IC₅₀ ∼ 10 μM) from those mediated by Kv1.4 channels (flecainide insensitive; IC₅₀ = 700 μM) (43). Similar to K_v4 channels, K_v1.5 is also known to be sensitive to flecainide at low micromolar concentrations (16), and application of 20 μM flecainide inhibited the A-type K⁺ current in retinal arterioles by >80%.

Fig. 3.

Electrophysiological traces showing the effects of K_v1 channel modulators on the A-type K⁺ current in retinal arteriolar smooth muscle cells. Whole cell currents were evoked by voltage steps from −80 to +20 mV. Hongotoxin (A) and noxiustoxin (B), which collectively inhibit K_v1.1/2/3/6 and 7 homotetrameric channels, did not affect the A-type K⁺ current, whereas some reduction was observed in the presence of 100 μM riluzole (C), a known blocker of K_v1.4 and K_v1.5 channels. No effects were seen following removal of extracellular K⁺ (0K_o; D), but the current was almost completely abolished by 20 μM flecainide (E). Mean data for these protocols are summarized in Table 3.

Table 3.

Pooled data showing the effects of selective Kv1α subunit inhibitors on the A-type K⁺ current in retinal arterioles

Inhibitor	Kv 1α Subunits Blocked	Mean Block±SEM, %	P Value (No. of vessels)
Hongotoxin	Kv 1.1, 1.2, 1.3, 1.6	0.89±0.99	NS (4)
Noxiustoxin	Kv 1.1, 1.2, 1.3, 1.7	−0.03±0.89*	NS (4)
Riluzole	Kv 1.4, 1.5	31.97±6.95	0.001045 (6)
0Ko	Kv 1.4 (not Kv1.5)	−2.04±6.05*	NS (5)
Flecainide	Kv 1.5 (not Kv1.4)	76.87±5.04	0.004859 (6)

0K_o, K⁺-free.

^*Negative % block indicates an increase in current.

Expression and neutralizing antibody studies further confirm a role for K_v1.5.

Our pharmacological data suggest that the A-type K⁺ current in retinal arteriolar myocytes is mediated, at least in part, by K_v1.5-containing channels. To provide further evidence that K_v1.5 underlies the A-type K⁺ current, RT-PCR was initially used to screen for mRNAs encoding K_v1.4 and K_v1.5 in freshly isolated retinal arterioles. RNA samples were collected from 5 to 13 arterioles, and rat brain pial membrane was used as a positive control (6). K_v1.5 mRNA was consistently detected in retinal arterioles but K_v1.4 was not evident (Fig. 4A). Lack of K_v1.4 expression could not be attributed to primer design, since K_v1.4 transcript expression was verified in the brain using the same primer pairs. No products were detected in any of the minus-reverse transcriptase control experiments (no amplification controls). Immunofluorescence staining with commercially available polyclonal antibodies was also used to test for cell-specific expression of K_v1.4 and K_v1.5 in retinal arterioles embedded within retinal flatmount preparations. To ensure that the images collected originated from the arterioles, the retinas were counterstained with propidium iodide nuclear stain. First-order arterioles radiating from the optic disk were readily identified by their distinctive monolayer of smooth muscle cell nuclei (see, for example, Fig. 4B,i), which was absent on interdigitating venules. Endothelial cell nuclei could also be seen just below the focal plane of the arteriolar smooth muscle cells and orientated in a longitudinal direction (Fig. 4B,i). Consistent with the RT-PCR results, K_v1.4 could not be detected in retinal arterioles (Fig. 4B, i and ii). This was not due to a lack of antibody reactivity because intense staining was observed in the outer nuclear layer of the retina (Fig. 4B, iii and iv), which was not apparent in flatmounts exposed to only secondary antibodies (2°Control; Fig. 4B,v). In contrast to K_v1.4, strong immunoreactivity was detected for K_v1.5 in retinal arterioles localized to the plasma membrane of the arteriolar smooth muscle cells (Fig. 4B, vi and vii). Anti-K_v1.5 staining was not visible in endothelial cells (data not shown), and 2°Controls were negative (Fig. 4B, viii and ix).

Fig. 4.

Screening by RT-PCR and immunofluorescence for K_v1.4 and K_v1.5 expression in rat retinal arterioles. A: 167-bp product of K_v1.5 was detected by RT-PCR in freshly isolated retinal arterioles, but K_v1.4 transcript expression was not observed. Rat brain pial membrane was used as a positive control and minus lanes represent no RT controls. Detection of K_v1.5 and not K_v1.4 was verified in cDNA reverse transcribed from three different RNA isolations. B: cellular localization of K_v1.4 and K_v1.5 in arterioles embedded within retinal flatmount preparations. All red images are propidium iodide (PI) nuclear stain. Arteriole was stained with PI (i) and anti-K_v1.4 (ii); no immunoreactivity for anti-K_v1.4 was detected. VSM refers to vascular smooth muscle. Outer Nuclear Layer (ONL) of retina was stained with PI (iii) and anti-K_v1.4 (iv). Control staining in the absence of primary antibody in the ONL (v) is shown. PI (vi) and anti-K_v1.5 (vii) staining in an arteriole with surrounding retinal ganglion cells are shown. Control staining for a PI-labeled retinal arteriole (viii) without primary antibody (ix) are shown. Scale bars, 10 μm

Recently, neutralizing antibody approaches have been successfully used to demonstrate the contribution of specific K_vα subunits to observed macroscopic K_v currents in native vascular smooth muscle cells (44). Conventional whole cell recordings were used to dialyze anti-K_v1.4 and anti-K_v1.5 antibodies into the cytosol of individual retinal arteriolar myocytes. The anti-K_v1.5 antibody employed has previously been demonstrated to specifically block K_v1.5 currents when applied intracellularly to mesenteric artery smooth muscle cells (26). Since gene and protein expression for K_v1.4 could not be detected in retinal arterioles, anti-K_v1.4 antibody was chosen as a negative control. Without antibody in the pipette, the A-type current was stable for periods >20 min. In the presence of anti-K_v1.5 antibody in the pipette solution, the A-type current was inhibited within 2–5 min after establishment of the whole cell configuration by 64 ± 11% (Fig. 5). Application of anti-K_v1.4 antibody in the pipette solution did not have any effect on the A-type current even after 20 min of recording (Fig. 5). Anti-K_v1.5 antibody also inhibited the sustained component of the net outward current (Fig. 5A; current densities were 44.1, 52.2, and 22.3 pA/pF for controls, anti-K_v1.4 antibody, and anti-K_v1.5 antibody, respectively; P < 0.05 for anti-K_v1.5 vs. other groups), suggesting that some of the delayed rectifier channels in this tissue may also contain K_v1.5 subunits.

Fig. 5.

Effects of anti-K_v1.4 and anti-K_v1.5 antibodies on the A-type K⁺ current in retinal arteriolar smooth muscle. A: whole cell K⁺ currents elicited by voltage steps from −80 to +80 mV in the absence and presence of either anti-K_v1.4 or anti-K_v1.5 antibody in the pipette solution. Records were obtained 5 min after establishment of the whole cell configuration. B: summary of effects of the antibodies on peak current density for the A-type current. Six vessels were tested per treatment group. **P < 0.01.

Contribution of K_vβ1 to the A-type K⁺ current.

The above results strongly suggest that K_v1.5 constitutes a principal molecular component underlying the A-type current in retinal arteriolar smooth muscle. In heterologous expression systems, K_v1.5 classically gives rise to delayed rectifier, not A-type currents. However, co-expression of K_v1.5 with K_vβ1 or K_vβ3 accessory subunits can transform the current into a rapidly inactivating A-type current (18, 25). To examine whether this mechanism might be of relevance in retinal arteriolar myocytes, we initially sought evidence for mRNA transcript expression of K_vβ1 and K_vβ3 subunits in freshly isolated retinal arterioles. K_vβ1 product was consistently detected; however, no definite expression of K_vβ3 mRNA was observed (Fig. 6A). K_vβ3 was detected in rat brain pial membrane (Fig. 6A). K_vβ1 protein expression in retinal arterioles was confirmed by immunohistochemistry of retinal flatmount preparations (Fig. 6B). At low magnifications, K_vβ1-associated fluorescence was most intense in retinal astrocytes, particularly in the end feet surrounding blood vessels (Fig. 6B,i). At higher magnifications, K_vβ1 was also observed to be specifically localized to the plasma membrane of retinal arteriolar myocytes (Fig. 6B, ii and iii), but was absent in the endothelial cell layer (Fig. 6B,iv).

Fig. 6.

A: K_vβ transcript expression in freshly isolated rat retinal arterioles. Expression of K_vβ1 transcript, but not K_vβ3, was evident. Brain was used as a positive control and minus lanes represent no RT controls. B, i: low magnification confocal image of a retinal flatmount preparation stained with PI (red) and anti-Kvβ1 (green) focused at the level of the ganglion cell layer. Strong Kvβ1-associated immunofluorescence was seen in the retinal astrocytes. Scale bar, 50 μm. Higher magnification image of retinal arteriole stained with PI (ii) and anti-K_vβ1 (iii). iv: luminal confocal image of a retinal arteriole showing that Kvβ1-associated immunofluorescence was limited to the arteriolar smooth muscle cells. Scale bars for B, ii–iv, 10 μm.

DISCUSSION

This study is the first to examine the molecular composition of the K_v channels underlying the A-type K⁺ current in vascular smooth muscle cells of the peripheral systemic circulation. Similar studies have previously been undertaken to identify the molecular nature of the K_v channels underlying delayed rectifier currents in arterial myocytes. The general consensus from these studies is that slowly inactivating tetraethylammonium (TEA)-insensitive delayed rectifier currents arise through heterotetrameric complexes of K_v1.2 with K_v1.4 or K_v1.5 channels associated with a K_vβ1 subunit, whereas TEA-sensitive delayed rectifier currents most likely contain at minimum K_v2.1 α-subunits (9). From the present study it appears that the molecular identity of the A-type K⁺ current in retinal arteriolar smooth muscle closely resembles the TEA-insensitive delayed rectifier current observed in other types of arterial smooth muscle in that K_v1.5 appears to be a major component and is probably co-assembled with K_vβ1 subunits.

Previous studies concerned with the molecular composition of K_v channels in arterial smooth muscle have begun by systematically screening K_v channel gene expression. Although this approach represents a logical starting point for identifying potential candidate α-subunits, data interpretation is often confounded by the multitude of K_v channel isoforms expressed when compared with the number of K_v current components observed (9). For this reason we initially adopted a pharmacological strategy to narrow down likely K_v channel components underlying the A-type K⁺ current. This type of approach has only recently become feasible with the emergence of an increasing number of drugs and toxins that selectively target specific members of the K_v channel superfamily. A caveat to this kind of approach, however, is that some of the drugs and toxins are only active against homotetrameric K_v channels (35), and hence, a lack of effect does not necessarily exclude the possibility that the subunit(s) of interest forms part of heterotetrameric K_v complex. Thus, although we can confidently exclude K_v1.1/2/3/4/6 and 7 homotetrameric channels as contributing to the A-type K⁺ current in retinal arteriolar smooth muscle, these subunits could still play a crucial role in the observed currents by forming heterotetrameric complexes with K_v1.5. In fact, we have previously shown that the A-type K⁺ current in retinal arterioles is partially suppressed by low levels of TEA (10 mM) (28), yet it is established that K_v1.5 homotetrameric channels are resistant to the actions TEA (16). This strongly suggests that the channels underlying the A-type K⁺ current in retinal arteriolar smooth muscle cells are heterotetrameric complexes of K_v1.5 with TEA-sensitive K_v1α subunits such as K_v1.1 and K_v1.6 (8, 16).

A distinctive feature of A-type K⁺ currents bestowed by K_v1.5 channels assembled with K_vβ1 subunits is the voltage for half-inactivation, which at −31.5 mV (18) is around 20–30 mV more positive than A-type K⁺ currents mediated by the majority of α-subunits in other K_v subfamilies (8). We have previously reported a half-inactivation voltage for the A-type K⁺ current in the retinal arteriolar smooth muscle of −28.3 mV (28), and this adds further weight to the notion that K_v1.5-containing channels co-assembled with K_vβ1 subunits form the basis of this current. There is, nonetheless, a discrepancy in the relation to the time constant for recovery from inactivation. A-type K⁺ currents mediated by the K_v1 family of channels recover relatively slowly from inactivation (3), and for K_v1.5/K_vβ1 channels, the time constant is 2.4 s (18). The time constant for recovery from inactivation of the A-type K⁺ current in retinal arteriolar myocytes is 118 ms (28), which is more reminiscent of K_v4-derived currents. Few studies have specifically addressed those factors involved in modulating recovery from inactivation for K_v1 channels, although it is notable that Ca²⁺/calmodulin-dependent kinase (CAMKII) phosphorylation of an NH₂ terminal residue of K_v1.4 leads to an accelerated recovery from N-type inactivated states (34). Retinal arteriolar smooth muscle cells exhibit a very high level of spontaneous subcellular Ca²⁺ signaling activity (11), and this would be consistent with the idea that CAMKII-dependent regulation of K_v1.5/K_vβ1 may be more relevant in this tissue than in heterologous expression systems where the biophysical properties these channels have been previously determined. Although it remains unclear if the phosphorylation status of Kvβ subunits modifies recovery from N-type inactivation of K_v1 channels, protein kinase A phosphorylation of serine-24 in the NH₂ terminus of K_vβ1.3 is known to alter the kinetics of inactivation of K_v1.5 (24).

By using a multi-faceted approach, we have identified K_v1.5 as being a principal component underlying the A-type K⁺ current in retinal arteriolar smooth muscle cells. It seems likely that K_v1.5 forms a heterotetrameric complex with other TEA-sensitive K_v1α subunits and that the transient nature of the current arises through association of the channels with K_vβ1 subunits. Intriguingly, the molecular components underlying the A-type K⁺ current in retinal arterioles appear similar to those that mediate delayed rectifier K⁺ currents in other types of vascular smooth muscle. Although the physiological significance of these observations warrants further investigation, it is evident that we should not necessarily consider A-type and delayed rectifier K⁺ currents in vascular smooth muscle as separate K_v current components. Instead, these currents may be derived from K_v channels with similar α-subunit compositions but reflect opposing ends of a spectrum of kinetically distinct currents that vary according to their degree of regulation by K_vβ subunits.

GRANTS

We thank The Juvenile Diabetes Research Foundation (US), Fight for Sight (UK) and The Wellcome Trust for financial support.