Local anesthetic actions on the K+ channels of dorsal root ganglion (DRG) and dorsal horn neurons may modulate sensory blockade during neuraxial anesthesia. In dorsal horn neurons, local anesthetics are known to inhibit transient but not sustained K+ currents. The authors characterized the effects of local anesthetics on K+ currents of isolated DRG neurons.
The effects of lidocaine, bupivacaine, and tetracaine on K+ currents in isolated rat DRG neurons were measured with use of a whole cell patch clamp method. The currents measured were fast-inactivating transient current (I(Af)), slow-inactivating transient current (I(As)), and noninactivating sustained current (I(Kn)).
One group of cells (type 1) expressed I(Af) and I(Kn). The other group (type 2) expressed I(As) and I(Kn). The diameter of type 2 cells was smaller than that of type 1 cells. Lidocaine and bupivacaine inhibited all three K+ currents. Tetracaine inhibited I(As) and I(Kn) but not I(Af) For bupivacaine, the concentration for half-maximal inhibition (IC50) of I(Kn) in type 2 cells was lower than that for I(Kn) in type 1 cells (57 vs. 121 microM). Similar results were obtained for tetracaine (0.6 vs. 1.9 mM) and for lidocaine (2.2 vs. 5.1 mM).
Local anesthetics inhibited both transient and sustained K+ currents in DRG neurons. Because K+ current inhibition is known to potentiate local anesthetic-induced impulse inhibition, the lower IC50 for I(Kn) of small type 2 cells may reflect preferential inhibition of impulses in nociceptive neurons. The overall modulatory actions of local anesthetics probably are determined by their differential effects on presynaptic (DRG) and postsynaptic (dorsal horn neurons) K+ currents.
DURING spinal and epidural anesthesia, the axon terminals of dorsal root ganglion (DRG) neurons and the dorsal horn neurons with which they synapse are both exposed to relatively high concentrations of local anesthetics because of their proximity to the injection site. Blockade of Na+channels by local anesthetics has been studied extensively and likely represents the main mechanism by which these agents block impulse propagation in axons. 1–2Local anesthetic effects on other ion channels in sensory neurons may modulate the ability of these agents to produce anesthesia and analgesia. For example, alteration of K+channel function by local anesthetics may influence the transmission of sensory impulses by altering the resting membrane potential, the timing of repolarization after an action potential, neuronal excitability, and the firing pattern of a neuron.
Olschewski et al. 3reported that local anesthetics inhibit transient but not sustained K+currents of rat dorsal horn neurons. Using frog sensory neurons, Guo et al. 4reported that bupivacaine inhibits IK(the delayed rectifier K+current). In outside-out patches obtained from amphibian sciatic nerve, which should include axons of DRG, Bräu et al. 5reported that local anesthetics inhibit a sustained K+current. These authors did not report the effects of local anesthetics on transient K+currents. The effects of local anesthetics on mammalian sensory neurons is not known. Therefore, we studied the effects of local anesthetics on transient and sustained K+currents in isolated rat DRG neurons.
Methods and Materials
Weanling Sprague-Dawley rats of either sex were deeply anesthetized with an intraperitoneal injection of pentobarbital (200 mg/kg), after which the spinal column was removed. The procedure was approved by the Animal Care and Use Committee of the University of Wisconsin, Madison, Wisconsin. DRGs were isolated in Dulbecco’s Modified Eagle Medium and treated with trypsin (2.5 mg/ml) and collagenase (2 mg/ml) for 45 min at 35°C. After the enzyme treatment, DRGs were dissociated by trituration with fire-polished Pasteur pipettes. DRG neurons were collected by low-speed centrifugation; resuspended in Dulbecco’s Modified Eagle Medium containing fetal bovine serum (10%), penicillin (50 U/ml), and streptomycin (50 μg/ml); and allowed to settle on cover glass coated with polylysine. Cells were maintained at 35°C in a humidified incubator equilibrated with a gas mixture containing 5% CO2. The cells were used within two days after isolation.
Patch pipettes were pulled from borosilicate glass tubing. Whole cell K+currents were measured from DRG neurons at room temperature (21–25°C) using an Axopatch 200B patch clamp amplifier (Axon Instruments, Foster City, CA) and acquired using pCLAMP 6 software (Axon Instruments). Currents were filtered at 5 kHz and acquired at 10 kHz. A P/−4 protocol was used for leak subtraction. Cell capacitance and series resistance were read from the dials of the patch clamp amplifier after cancellation of the capacitative current transient obtained during a small depolarizing test pulse. Cell diameter was estimated before patch clamping using an eyepiece micrometer at 400× magnification. Gigaseal formation and whole cell configuration were achieved in a medium containing the following (in mm): NaCl, 130; KCl, 5; MgCl2, 1; CaCl2, 2; HEPES, 10 (pH 7.4 with NaOH). Control external solution for the measurements of K+currents contained the following (in mm): choline Cl, 130; KCl, 5; MgCl2, 1; CoCl2, 2; HEPES, 10 (pH 7.4 with choline base). The recording chamber (0.7 ml) was perfused with 5 ml of the external solution containing various concentrations of local anesthetics before the current measurements. The pipette solution contained the following (in mm): KCl, 120; MgCl2, 2.5; EGTA, 10; HEPES, 10; MgATP, 2; LiGTP, 0.3 (pH 7.3 with NaOH).
Dulbecco’s Modified Eagle Medium was obtained from Life Technologies (Grand Island, NY). All other reagents were obtained from Sigma Chemical Company (St. Louis, MO). Stock solutions of bupivacaine (10 mm) were prepared in water. Because of the high concentrations used in the experiments, stock solutions of lidocaine (20 mm) and tetracaine (10 mm) were prepared in the control external solution, and pH was adjusted to 7.4 by the addition of choline base.
The resting potential was maintained at −80 mV. Based on the study of Gold et al. , 6we used test pulses to +30 mV with a duration of 400 ms preceded by 1 s conditioning prepulses to −100 mV or −30 mV to separate the different types of K+currents. A test pulse preceded by a prepulse to −100 mV was used to elicit all types of voltage-activated K+currents (total current). A prepulse to −30 mV eliminated the currents that undergo inactivation, leaving only a noninactivating sustained current (IKn). Subtraction of IKnfrom the total current revealed either a fast-inactivating transient current (IAf) or a slow-inactivating transient current (IAs). The magnitudes of the transient currents (IAfand IAs) were measured at the peak amplitude, whereas the sustained currents (IKn) was measured at its plateau amplitudes at the latter part of the test pulse.
The inhibitory effect of local anesthetics on the amplitude of each type of K+current was analyzed using a Hill equation of the form:
where IC50is the concentration of local anesthetic for half-maximal inhibition, and n is the Hill coefficient. Under control conditions, the rates of inactivation of IAfand IAswere determined by fitting the decaying phase of the current to a single exponential of the form:
where A is the magnitude, C is the offset constant, and τ is the time constant. With bupivacaine present, the decaying phase of IAswas fit to a biexponential equation of the form:
where A1represents the magnitude of fast phase with the time constant τ1, A2represents the magnitude of slow phase with the time constant τ2, and C is the offset constant. The relative proportion (%) of the fast phase was 100 × A1/(A1+ A2). The fast phase of current decay in the presence of bupivacaine most likely reflects the bupivacaine block of the open K+channel. 7The effect of bupivacaine on the magnitude of the fast-inactivating component of IAswas analyzed using the Hill equation of the form:
where Kd is the dissociation constant. We also have obtained Kd for bupivacaine binding from the plot of a reciprocal of time constant of the fast-inactivating inactivating component (τB) of the current against bupivacaine concentration using the relation:
Data are expressed as mean ± SD. K+current decaying phases were fit to either single or double exponential equations as described using the Chebyshev algorithm (pCLAMP6, Axon Instruments). Dose–response relations were fit using the Quasi-Newton algorithm (MacCurveFit; Kevin Raner Software, Mt. Waverley, Victoria, Australia). Data were excluded from analysis if the amplitude of the K+current after washout of local anesthetic was less than approximately 50% or greater than approximately 150% of control. An unpaired t test was used to evaluate the difference between two mean values. The concentration dependence of the effects of local anesthetics was evaluated by repeated measures analysis of variance followed by the Dunnett t test. Differences were considered significant when P < 0.05.
Voltage-activated K+Currents Expressed in Two Subpopulations of DRG cells
From the patterns of K+current expression, two types of cells were recognized (fig. 1). One cell type (type 1) generally expressed two different K+currents. These were a fast-transient current (IAf; inactivation time constant of 12 ± 8 ms, n = 24) and a sustained current that does not undergo steady-state inactivation (IKn). The current elicited after a prepulse to −30 mV is IKn(fig. 1A). The difference between the total current (measured after a prepulse to −100 mV) and IKnwas IAf(fig. 1A). The other cell type (type 2) expressed a slow-transient current (IAs; inactivation time constant of 279 ± 86 ms, n = 20) and a noninactivating sustained current (IKn). As in type 1 cells, the current elicited after a prepulse to −30 mV was IKn(fig. 1B). However, in type 2 cells, the difference between the total current and IKnwas IAs(fig. 1B). Type 1 cells had significantly larger diameters (36 ± 4 vs. 28 ± 3 μm;P < 0.05) and whole cell capacitance (59 ± 18 vs. 26 ± 7 pF;P < 0.05) compared with type 2 cells.
Effects of Local Anesthetics on IAf
Figure 2shows the effects of local anesthetics on IAf. Bupivacaine and lidocaine but not tetracaine reduced the amplitude of IAf. Higher concentrations of tetracaine were not tested because of the nonspecific cytotoxicity 8of this local anesthetic. Bupivacaine also increased the rate of inactivation of IAf(figs. 2A and 3A). In contrast to bupivacaine, lidocaine and tetracaine did not decrease the inactivation time constant of IAf. In some cells, lidocaine and tetracaine slowed the inactivation rate of IAf(figs. 2B and C). This effect showed statistical significance at high concentrations of lidocaine but not at any concentration of tetracaine tested (figs. 3B and C).
Effects of Local Anesthetics on IAs
Figure 4shows the effects of local anesthetics on IAs. All three local anesthetics that were tested decreased the amplitude of IAs. In the presence of bupivacaine, the time course of decay consisted of an initial rapid phase followed by a slow phase (fig. 4A). Lidocaine and tetracaine did not induce the pronounced rapid phase of inactivation observed with bupivacaine (figs. 4B and C). With bupivacaine present, the decay could be described by the sum of two exponentials, with the time constant of the fast phase an order of magnitude smaller than the time constant for the slow component (fig. 5A). As the bupivacaine concentration was increased, the magnitude of the fast component of decay increased, with the half-maximal effect at the bupivacaine concentration of 41 μm (fig. 5B). From the plot of the reciprocal of τBagainst bupivacaine concentration (fig. 5C), we obtained k = 0.89 × 106m−1· s−1, l = 58 s−1, and Kd = 65 μm.
Concentration-dependent Effects of Local Anesthetics on Amplitudes of IAfand IAs
Figure 6shows the concentration–response curves for the effects of the three local anesthetics on the current amplitudes of IAfand IAs. The values for the IC50and Hill coefficients are summarized in table 1. Tetracaine, in the range of concentration tested, did not significantly alter the amplitude of IAf(fig. 6A). Bupivacaine had a small effect on the amplitude of IAf, and the IC50value was estimated by assuming the data represent a portion of the dose–response relation that can be described by a Hill equation. The ratio of IC50values indicates that bupivacaine was 30 times more potent than lidocaine in its effects on IAf(table 1). All three local anesthetics also reduced the magnitude of IAs(fig. 6B), with bupivacaine being 20 times more potent than lidocaine and tetracaine being 2 times more potent than lidocaine (table 1).
Effects of Local Anesthetics on IKn
The magnitude of IKnin type 1 cells under control conditions (10.2 ± 4.3 nA, n = 30) was significantly greater than that of IKnfrom type 2 cells (5.1 ± 1.4 nA, n = 19;P < 0.05). Figure 7shows that lidocaine more potently inhibited the IKncurrent of type 1 cells than that of type 2 cells. This preferential inhibition was found with all three local anesthetics (fig. 8). Table 2shows the values for IC50and Hill coefficients that describe the inhibition of IKnin type 1 and type 2 cells for all three local anesthetics. For IKnof type 1 cells, bupivacaine was 42 times more potent than lidocaine, and tetracaine was 3 times more potent than lidocaine. For IKnof type 2 cells, bupivacaine was 39 times more potent than lidocaine, and tetracaine was 4 times more potent than lidocaine.
Of the three voltage-activated K+currents we studied, IAfwas least sensitive to the inhibitory action of local anesthetic. Tetracaine, up to 1.5 mm, did not inhibit IAf. With bupivacaine and lidocaine, the values of IC50for IAfwere about 5 times higher than IC50for the most sensitive currents, IAsand IKnof type 2 cells (tables 1 and 2). In this regard, the effects of local anesthetics we observed resemble those of tetraethyl ammonium ion. 9The effects we observed with DRG cells contrast with the selective inhibition of a transient current reported in cells of the spinal dorsal horn. 3Although concentrations of lidocaine higher than 1 mm were required to inhibit IKncurrent of type 1 cells (fig. 8), it should be noted that 1% lidocaine (molecular weight of 271 for lidocaine hydrochloride) corresponds to 37 mm. One possibility that can account for the Hill coefficient of approximately 2 observed for the effects of lidocaine and tetracaine on the transient currents (table 1) is that binding of more than one molecule of these local anesthetic is involved in reduction of the current amplitude.
In addition to the relative insensitivity of IAfto the action of local anesthetics, the results indicated a preferential inhibition of K+currents expressed in type 2 cells. This is well-demonstrated for IKn, which was expressed in both type 1 and type 2 cells (table 2). The IC50values for local anesthetic inhibition of IKnin type 2 cells were approximately half of those for inhibition of IKnin type 1 cells. In fact, both of the K+currents expressed by type 2 cells (IAsand IKn) were generally more sensitive to local anesthetics than the two currents expressed by type 1 cells (IAfand IKn). Because type 2 cells are smaller and express IAs, similar to a subpopulation of DRG cells reported by Gold et al. 6that also respond to capsaicin, these cells may represent C-fiber-type nociceptive neurons. Therefore, the preferential inhibition of K+currents in type 2 cells may reflect selective potentiation of local anesthetic-induced impulse inhibition in nociceptive nerves. This is because the K+channel blocker, tetraethyl ammonium ion, is known to increase the lidocaine inhibition of compound action potentials of sciatic nerve. 10
The observed effects of three local anesthetics on noninactivating sustained K+current (IKn) were similar to those reported for patches of amphibian sciatic nerve. 5Therefore, bupivacaine was much more potent than lidocaine, and tetracaine, despite similar lipid solubility, 11was less effective than bupivacaine. In dorsal horn, on the other hand, Olschewski et al. 3found that sustained K+currents were more resistant to local anesthetics than were transient K+currents. Whether selective inhibition of presynaptic sustained K+currents and postsynaptic transient K+currents by local anesthetics influences their ability to alter sensory nerve transmission has not been tested.
We have studied the effects of local anesthetics on three different types of K+currents in DRG neurons using a simple inactivation protocol based on the method used by Gold et al. 6However, more than three types of K+currents are thought to exist in DRG neurons. Gold et al. , 6using biophysical parameters as well as specific K+channel blockers, identified a total of six different types of voltage-activated K+currents. 6Using immunocytochemical methods, Ishikawa et al. 12showed as many as seven voltage-activated K+channels in DRG (Kv1.1, Kv1.2, Kv1.3, Kv1.4, Kv1.5, Kv1.6, Kv2.1). The three currents we studied likely correspond to three of the currents described by Gold et al. 6Although we observed two of the other three currents described by these authors (IAhtand IKi), we did not report the effects of local anesthetics on these currents because of the small number of observations (IAht) or because of our inability to isolate the current (IKi) adequately. Therefore, it is possible that the K+currents we recorded were contaminated by currents through other K+channels that we could not distinguish using our protocol. Similarly, the current we identified as IKnin type 1 and type 2 cells actually may be two different currents.
The effects of bupivacaine on the time course of inactivation we observed with IAfresembles those reported for Ito13and for rKv1.4 and rKv4.3. 14It is possible that bupivacaine block contributes to the faster inactivation in the presence of this local anesthetic. As a possible mechanism for the pronounced decrease in the rate of inactivation of IAfobserved in the presence of lidocaine, we speculate that this local anesthetic may interfere with the endogenous fast-inactivation mechanism. The effect of bupivacaine on the time course of the decrease in current amplitude of IAswas similar to the effect of this local anesthetic on hKv1.5. 7The values of Kd (41 and 65 μm) obtained by two different methods for racemic bupivacaine in this study, however, were higher than the values reported for S (−)-bupivacaine (24.9 μm) and for R (+)-bupivacaine (5.0 μm) with hKv1.5 expressed in mouse Ltk -cell line. 7The initial fast phase of inactivation (blocking) is also reported for bupivacaine action on rat brain Kv1.1, which does not inactivate under control conditions, and is considered to reflect slow open channel block. 15The lack of the fast decrease of current observed with lidocaine was attributed to fast open channel block, so that the time course of block could not be measured. 15It is possible that the minimal effect of lidocaine and tetracaine on the time course of inactivation of IAsin the current study also may reflect fast open channel block.
In summary, the greater sensitivity of IKnfrom small type 2 cells suggests that local anesthetics may preferentially inhibit impulses in primary nociceptive neurons. Furthermore, all the local anesthetics used in this study inhibited the sustained K+current (IKn) of DRG neurons, an effect different from that reported in dorsal horn neurons by Olschewski et al. 3It is likely that the overall modulatory effects of local anesthetics at the level of the DRG–dorsal horn synapse are influenced by their differential effects on presynaptic (DRG) and postsynaptic (dorsal horn) K+currents.
The authors thank Craig V. Levenick, B.S., Department of Anesthesiology, University of Wisconsin, Madison, Wisconsin, for his excellent technical assistance.