For some years, clonidine and dexmedetomidine, α2-adrenoceptor agonists, have been in general use in the fields of anaesthesia, intensive care and pain management. Clonidine has been used for epidural or intrathecal anaesthesia in combination with opioids or local anaesthetics  and has a potent antinociceptive effect when administered epidurally [2,3] or intrathecally  as the sole drug. Reportedly, α2-adrenoceptor agonists produce an antinociceptive effect through several antinociceptive mechanisms at the spinal level . The main mechanism is the inhibition of cyclic adenosine 3′,5′-monophosphate (cAMP) formation via G-protein. As a result, the calcium conductance of primary nerve fibres is decreased through an inhibition of voltage-gated calcium channels, whereupon the release of substance P from such fibres is reduced. In addition, potassium channels in dorsal horn neurons might open, followed by cell hyperpolarization and a decrease in impulse transmission . Previous studies have revealed that in peripheral neurons, tetrodotoxin-resistant (TTX-R) Na+ channels are key elements in the transmission of nociceptive sensory information [6,7]. We therefore designed the present study in order to elucidate the potential direct effects of α2-adrenoceptor agonists on TTX-R Na+ channels in rat dorsal root ganglion (DRG) neurons (using the whole-cell patch-clamp technique).
Preparation of cells
DRG neurons were isolated as previously described . Briefly, adult Sprague-Dawley rats (200-250 g, n = 40) were anaesthetized using intraperitoneal pentobarbital, then killed by decapitation and DRGs were rapidly removed along the cervical, thoracic and lumbar sections of the spinal cord. The DRGs were incubated at 37°C for 23-30 min in Tyrode solution (for composition, see below) containing 2 mg mL−1 collagenase (Type 1; Sigma, St Louis, MO, USA) and 5 mg mL−1 dispase II (Boehringer Mannheim, Indianapolis, IN, USA). After washing three times with fresh, enzyme-free Tyrode solution, single neuronal cells were obtained by gentle agitation in Tyrode solution through a small-bore Pasteur pipette. After filtering the cell suspension, the collected cells were resuspended in Tyrode solution, placed on glass coverslips and incubated in a humidified atmosphere containing 5% CO2 at 37°C for 2-8 h before being used for patch-clamp experiments.
Recording of membrane currents
A coverslip with cells was placed in a small organ bath (0.8 mL) on the stage of an inverted microscope (TMD; Nikon, Tokyo, Japan). Two to six cells were studied from each animal. Recordings of whole-cell membrane currents were made at an experimental temperature of 23 ± 2°C, n = 65) using standard patch-clamp techniques . Patch pipettes were made from glass capillaries using a four-step puller (P-97; Sutter Instrument Company, CA, USA), their tips being fire-polished using a microforge (MF-830; Narishige, Tokyo, Japan) to give a final resistance of 1.0-2.0 MΩ. Membrane currents were amplified using a current amplifier (Axon Instruments, CA, USA), and signals were digitized by means of a 12-bit analogue-to-digital converter (Digidata 1200B; Axon Instruments). The current signals were filtered at a cut-off frequency of 5 kHz, digitized at a sampling rate of 20 kHz using PClamp v.8.0 software (Axon Instruments) and stored on a personal computer. Series resistance was compensated as far as possible (by 90-100%). A P+1/4P protocol  was used for leak subtraction. Data analysis and preparation of figures were performed using Origine v.6 software (Microcal Software Inc., MA, USA). Potassium channel currents were suppressed by the inclusion of Cs+ in the pipette solution and tetraethylammonium (TEA) in the external solution. Calcium channel currents were suppressed by the inclusion of F− in the pipette solution and Mg2+ in the external solution.
DRG neurons were held under a voltage clamp at −120 mV after the whole-cell patch-clamp configuration had been achieved, then Na+ currents were evoked by depolarizing pulses (50 ms in duration) to −10 mV. We observed TTX-R Na+ currents under external solution containing 0.2 μmol TTX. Small DRG neurons play an important role in nociceptive transmission  and we have reported that rat DRG neurons of smaller size (<30 μm) preferentially express TTX-R Na+ currents . We therefore recorded from DRG neurons with a diameter of ≤30 μm.
Solutions and drugs
The Tyrode solution was of the following composition (mmol): NaCl 140.0, KCl 4.0, MgCl2 2.0, glucose 10.0 and N-2-hydroxyethylpiperazine-N′-2-ethanesulfonic acid (HEPES) 10.0, and it was adjusted to pH 7.4 with NaOH. As reported elsewhere , the pipette solution was of the following composition (mmol): CsF 135.0, NaCl 10.0 and HEPES 5.0 (adjusted to pH 7.0 with CsOH). The external solution was: NaCl 25.0, tetramethylammonium chloride 75.0, TEA chloride 20.0, CsCl 5.0, CaCl2 1.8, MgCl2 1.0, glucose 25.0 and HEPES 5.0 (adjusted to pH 7.4 with TEA-OH).
The drugs used were clonidine hydrochloride, yohimbine hydrochloride, lidocaine hydrochloride and tetrodotoxin (Sigma, St Louis, MO, USA). Dexmedetomidine hydrochloride was provided by Abbott Laboratories (Abbott Park, IL, USA).
Application of drugs
Extracellular application of drugs was achieved by replacing the bath solution in the recording chamber (0.8 mL) with drug-containing solution 7-10 times within 20 s.
Analyses were performed as previously described . Inactivation curves were drawn according to the Boltzmann equation:
where max INa is the maximal value for INa, Vh is the membrane potential achieved using a 150 ms prepulse (conditioning) potential, Vh0.5 is the potential at which INa is half of max INa and kh is the slope factor.
The dose-response curves for the blocking actions of drugs on TTX-R Na+ currents were fitted to the Hill equation:
where IC50 is the half-maximum concentration for the inhibitory action of a given drug, [Drug] is the drug concentration and h is the Hill coefficient.
The dissociation constants for the binding of drugs to the TTX-R Na+ channels in the inactivated state were calculated using the equation:
where [Drug] is the concentration of a given drug, Kr the dissociation constant for drug binding to the TTX-R Na+ channels in the closed-available state, V0.5 the drug-induced shift in amplitude in the voltage-dependent inactivation curve and kh the slope factor for the inactivation curve.
All values are expressed as means ± SD. Statistical significance was assessed using a Student's paired or unpaired t-test, differences being considered significant at P < 0.05.
Concentration-response relationships for the blocking actions of clonidine and dexmedetomidine on TTX-R Na+ currents were obtained at two different holding potentials (HPs), −120 and −70 mV, with stepping to −10 mV in each case. At HP = −120 mV, the IC50 values obtained for clonidine and dexmedetomidine were significantly higher than the corresponding values obtained at HP = −70 mV (Fig. 1a, Table 1).
The effects of these α2-adrenoceptor agonists on the voltage dependence of the steady-state inactivation were investigated using a conventional double-pulse protocol (HP = −120 mV, duration of the prepulse =150 ms) (Fig. 2a, b). We selected concentrations of 300 and 1000 μmol for clonidine, and 100 and 300 μmol for dexmedetomidine since these inhibited the TTX-R Na+ currents by approximately 30% and 70%, respectively (Fig. 1a). As shown in Figure 2a, b and Table 1, both clonidine and dexmedetomidine shifted Vh0.5 (the potential at which the peak amplitude of the Na+ current is half-maximal) in the hyperpolarizing direction. Drug effects on the voltage dependence of the steady-state inactivation were reversible during wash out. Using Eq. (3), we calculated the Ki values for clonidine and dexmedetomidine by entering, respectively, values of 1000 and 300 for [Drug], 611.8 and 152.3 for Kr [the IC50 values for clonidine and dexmedetomidine for the TTX-R Na+ current at an HP of −120 mV, which can be used if we assume that the Na+ channels are all in the closed-available state (Fig. 2)], 4 and 8.5 for V0.5, and 5.4 and 5.91 for kh. The Ki values for clonidine and dexmedetomidine are thereby calculated to be 220.9 and 26.3 μmol, respectively.
To examine the effect of yohimbine, an α2 antagonist, on the inhibitory effects of clonidine and dexmedetomidine on the TTX-R Na+ current, we added 1 μmol yohimbine to 300 μmol clonidine or 100 μmol dexmedetomidine (concentrations that inhibited the TTX-R Na+ peak current by approximately 30%) (Fig. 3a, b). We chose the concentration of 1 μmol for yohimbine at HP −120 mV because yohimbine itself inhibited the TTX-R Na+ current when used at concentrations above 1 μmol (three experiments from each cell: data not shown). We also determined the IC50 value for yohimbine at HP −70 mV (Fig. 4). Yohimbine 1 μmol did not alter the inhibition of the TTX-R Na+ peak current induced by either clonidine or dexmedetomidine.
To examine possible interactions of clonidine and dexmedetomidine with lidocaine, we administered each agent at its IC50 concentration: lidocaine at 72.6 μmol (Table 1) either with clonidine at 257.2 μmol (Fig. 4a) or with dexmedetomidine at 58.0 μmol (Fig. 4b), in each case at an HP of −70 mV. The percentage inhibitions of the peak current were 68.1 ± 4.8% for lidocaine with clonidine (six experiments from each cell) and 68.4 ± 4.7% for lidocaine with dexmedetomidine (six experiments from each cell).
Both clonidine and dexmedetomidine caused a use-dependent inhibition of the TTX-R Na+ current. Currents were evoked repeatedly by stepping (for 10 ms) to −10 mV from −70 mV at one of the three frequencies (0.2, 5 or 20 Hz). The peak amplitude of the current evoked by the first-step pulse was normalized as 100% in the absence or presence of 100 μmol clonidine or 30 μmol dexmedetomidine (concentrations that induced approximately 20−30% inhibition of the peak current). Approximately 1 min was allowed to elapse between one pulse train and the next at each frequency. The percentages of first pulse inhibition were 22.7 ± 8.6% for clonidine and 31.3 ± 17.4% for dexmedetomidine (n = 18 experiments from each cell). At frequencies of 5 and 20 Hz, there were significant differences (*P < 0.05) in peak amplitude when the current evoked by the 15th pulse was compared between the absence and presence of drugs (clonidine or dexmedetomidine) (Table 2).
The present results demonstrate that both clonidine and dexmedetomidine block the TTX-R Na+ current in rat DRG neurons in a concentration-dependent manner, and that the IC50 for dexmedetomidine (58 ± 10 μmol) is 4.43 times lower than that for clonidine (257 ± 31 μmol) at HP = −70 mV. Yohimbine did not prevent these blocking effects of clonidine and dexmedetomidine, but at a higher dose yohimbine itself induced a significant reduction in this current. Clonidine and dexmedetomidine shifted the voltage-dependent inactivation curve for the TTX-R Na+ channels in the hyperpolarizing direction, and also produced a use-dependent block of these channels. Combinations of clonidine with lidocaine and dexmedetomidine with lidocaine produced an additive blockade-type interaction on the TTX-R Na+ current.
Both clonidine and dexmedetomidine produced a use-dependent blockade. In our previous study , 30 μmol lidocaine inhibited the relative 15th peak current amplitude by 62.8% ± 12.0 at 5 Hz and by 37.4 ± 9.1% at 20 Hz (using the same protocol as that used here). This inhibition by lidocaine was much more powerful than those induced by clonidine and dexmedetomidine in the present study (Table 2). Thus, both clonidine and dexmedetomidine would seem to produce a much weaker open-channel blockade than lidocaine.
α2-adrenoceptor agonists, such as clonidine and dexmedetomidine, produce antinociception when injected epidurally or intrathecally [1,13-15]. The mechanisms underlying their antinociceptive effects have been explained by α2-adrenoceptor activation . Since both dexmedetomidine and clonidine induced TTX-R Na+-current suppression in the present study (although with different potencies), their actions might seem likely to result from activation of α2-adrenoceptors. However, since yohimbine did not block the Na+-current suppressions induced by these two drugs, and since yohimbine itself completely blocked this Na+ current when we used a dose that also blocks the antinociceptive effects of epidural dexmedetomidine and clonidine , it is possible that the suppression of Na+ currents is due to a direct blocking effect on TTX-R Na+ currents that is independent of the adrenoceptor-mediated, G-protein-coupled mechanisms. Synergism with lidocaine in the clinical setting might be explained by different mechanisms of action (α-adrenoceptor vs. sodium channel).
Both clonidine and dexmedetomidine were found to inhibit the TTX-R Na+ current even in the presence of yohimbine, and a combination of either of these drugs with lidocaine produced an additive inhibition of the TTX-R Na+ current. Previous rat tail-flick studies have indicated that intrathecal clonidine and lidocaine act synergistically to reduce the nociceptive response [16,17]. It is possible that a difference in the sites of action between α2-adrenergic agonists and lidocaine might contribute to this synergistic interaction because a synergistic interaction can occur when drugs affect different critical points along a common antinociceptive pathway . It was demonstrated some years ago in an animal study that after an intrathecal injection of 300 μg clonidine (which produced a near-maximal antinociceptive effect after intrathecal administration in an experiment employing mechanical stimulation), its cerebrospinal fluid (CSF) concentration was roughly 100-10 000 ng mL−1 over the 4 h for which an analgesic effect was observed . Those concentrations (equivalent to 0.5-50 μmol) are much lower than the IC50 value we obtained for clonidine (257.2 μmol) at an HP of −70 mV. In a human study of epidural clonidine, the half-maximal effective concentration of clonidine in the CSF was found to be 80 ± 6 ng mL−1 (approximately 0.75 μmol) . On the other hand, intrathecal injection of 100 μg dexmedetomidine (same antinociceptive potency as 300 μg intrathecal clonidine) produced almost the same concentration range as that seen after 300 μg intrathecal clonidine (around 0.5-50 μmol) , the upper end of which is close to the IC50 value we obtained here for dexmedetomidine (58.0 μmol at HP = −70 mV). When 300 μg clonidine or 100 μg dexmedetomidine is given epidurally, the CSF concentrations are about the same or slightly less than those measured after intrathecal administration, but much greater than those measured after intravenous administration (<1 ng mL−1) [13,19]. Thus, we suggest that a therapeutic dose of dexmedetomidine may produce antinociception in part via an inhibition of TTX-R Na+ channels in DRG like local anaesthetic effects after its epidural or intrathecal administration.
Dexmedetomidine shifted the inactivation curve more strongly in the hyperpolarizing direction than clonidine (Fig. 2). This difference may be attributed to these α2-agonists having different affinities for the TTX-R Na+ channels in a given channel state. Judging from their Ki values, affinities of clonidine and dexmedetomidine for the inactivated state would be 2.8 and 5.8 times higher than those for the close-available state of the TTX-R Na+ channels. Thus, dexmedetomidine has about two times higher potency than clonidine for the TTX-R Na+ channels in the inactivation state. Rat DRG cells obtained from a chronic constriction-injury neuropathic-pain model reportedly show a shift the voltage dependence of activation and inactivation of the TTX-R Na+ currents to a more negative value . Since the proportion of TTX-R Na+ channels in the inactivation state might increase in such a neuropathic-pain model, dexmedetomidine may be more effective than clonidine at blocking TTX-R Na+ channels under neuropathic-pain conditions.
In conclusion, both clonidine and dexmedetomidine block TTX-R Na+ channels in rat DRG neurons in a dose-dependent and use-dependent manner. The mechanism underlying such blocking effects seems likely to be due to a direct action on TTX-R Na+ channels, and not α2-adrenoceptor activation. Such direct inhibitions of TTX-R Na+ channels may contribute to the antinociceptive effects of clonidine and dexmedetomidine when used as additives to regional anaesthesia.
The authors thank Drs S Komori and T Unno Laboratory of Pharmacology, Department of Veterinary Science, Faculty of Agriculture, Gifu University Graduate School of Medicine, Gifu, Japan for stimulating discussion throughout this work.
This work was supported by Grant-in-Aid for Scientific Research Nos. 14207059 and 18591697 from the Ministry of Education, Science and Culture, Tokyo, Japan.
1. Eisenach JC, De Kock M, Klimscha W. Alpha(2)-adrenergic agonists for regional anesthesia. A clinical review of clonidine (1984-1995). Anesthesiology
2. Mendez R, Eisenach JC, Kashtan K. Epidural clonidine analgesia after cesarean section. Anesthesiology
3. Asano T, Dohi S, Ohta S, Shimonaka H, Iida H. Antinociception by epidural and systemic alpha(2)-adrenoceptor agonists and their binding affinity in rat spinal cord and brain. Anesth Analg
4. Filos KS, Goudas LC, Patroni O, Polyzou V. Intrathecal clonidine as a sole analgesic for pain relief after cesarean section. Anesthesiology
5. Smith H, Elliott J. Alpha2 receptors and agonists in pain management. Curr Opin Anaesthesiol
6. Akopian AN, Souslova V, England S et al
. The tetrodotoxin-resistant sodium channel SNS has a specialized function in pain pathways. Nat Neurosci
7. Sangameswaran L, Delgado SG, Fish LM et al
. Structure and function of a novel voltage-gated, tetrodotoxin-resistant sodium channel specific to sensory neurons. J Biol Chem
8. Oda A, Ohashi H, Komori S, Iida H, Dohi S. Characteristics of ropivacaine block of Na+
channels in rat dorsal root ganglion neurons. Anesth Analg
9. Hamill OP, Marty A, Neher E, Sakmann B, Sigworth FJ. Improved patch-clamp techniques for high-resolution current recording from cells and cell-free membrane patches. Pflugers Arch
10. Bezanilla F, Armstrong CM. Inactivation of the sodium channel. I. Sodium current experiments. J Gen Physiol
11. Song JH, Huang CS, Nagata K, Yeh JZ, Narahashi T. Differential action of riluzole on tetrodotoxin-sensitive and tetrodotoxin-resistant sodium channels
. J Pharmacol Exp Ther
12. Osawa Y, Oda A, Iida H, Tanahashi S, Dohi S. The effects of class Ic antiarrhythmics on tetrodotoxin-resistant Na+
currents in rat sensory neurons. Anesth Analg
: 464-471 table of contents.
13. Eisenach JC, Shafer SL, Bucklin BA, Jackson C, Kallio A. Pharmacokinetics and pharmacodynamics of intraspinal dexmedetomidine in sheep. Anesthesiology
14. Castro MI, Eisenach JC. Pharmacokinetics and dynamics of intravenous, intrathecal, and epidural clonidine in sheep. Anesthesiology
15. Aveline C, Metaouna ES, Masmoudi A, Boelle PY, Bonnet F. The effect of clonidine on the minimum local analgesic concentration of epidural ropivacaine during labor. Anesth Analg
16. Kawamata T, Omote K, Kawamata M, Iwasaki H, Namiki A. Antinociceptive interaction of intrathecal alpha2-adrenergic agonists, tizanidine and clonidine, with lidocaine in rats. Anesthesiology
17. Hao S, Takahata O, Iwasaki H. Antinociceptive interaction between spinal clonidine and lidocaine in the rat formalin test: an isobolographic analysis. Anesth Analg
18. Berenbaum MC. What is synergy? Pharmacol Rev
19. Eisenach J, Detweiler D, Hood D. Hemodynamic and analgesic actions of epidurally administered clonidine. Anesthesiology
20. Kral MG, Xiong Z, Study RE. Alteration of Na+
currents in dorsal root ganglion neurons from rats with a painful neuropathy. Pain