A Rapid Capsaicin-Activated Current in Rat Trigeminal Ganglion Neurons (Pain/Taste/Nociceptive Fibers/Ph/Capsazepine) L

Home , Resiniferatoxin

Proc. Natl. Acad. Sci. USA Vol. 91, pp. 738-741, January 1994 Neurobiology A rapid capsaicin-activated current in rat trigeminal ganglion neurons (pain/taste/nociceptive fibers/pH/capsazepine) L. LIu* AND S. A. SIMONt Departments of *Neurobiology and tAnesthesiology, Duke University Medical Center, Durham, NC 27710 Communicated by Irving T. Diamond, October 15, 1993

ABSTRACT A subpopulation of pain fibers are activated Na2HPO4, 5.6 mM D-glucose, and 10 mM Hepes. They were by capsaicin, the ingredient in red peppers that produces a then incubated for 40 min at 37°C in HBSS containing 1 mg burning sensation when eaten or placed on skin. Previous of collagenase per ml (type XI-S), triturated with a flamed studies on dorsal root ganglion neurons indicated thatcapsaicin Pasteur pipette, and finally incubated at 37°C for 6 min with activates sensory nerves via a single slowly activating and 0.1 mg of DNase I per ml (type IV). Then they were inactivating inward current. In rat trigeminal neurons, we retriturated and washed/centrifuged three times in HBSS. identified a second capsaicin-activated inward current. This They were then resuspended in F-14 medium (GIBCO) current can be distinguished from the slow one in that it rapidly containing 10%o fetal calf serum in a Petri dish. The cultures activates and inactivates, requires Ca2+ for activation, and is were maintained in an incubator at 37°C equilibrated with 5% insensitive to the potent capsaicin agonist resiniferatoxin. The CO2. Most patch clamp experiments were done on cells rapid current, like the slower one, is inhibited by ruthenium cultured for 12-24 hr. At the beginning of each experiment, red and capsazepine. The two capsaicin-activated inward cur- the neurons were placed in a chamber containing Krebs- rents share many similarities with the two inward currents Henseleit (KH) buffer on an inverted microscope. The com- activated by lowering the pH to 6.0. These similarities include position of KH is 140 mM NaCl, 5 mM KCl, 2.0 mM CaCl2, kinetics, reversal potentials, responses to Ca2+, and inhibition 1.0mM MgCl2, 10 mM Hepes, and 10mM D-glucose (pH 7.4). by ruthenium red and capsazepine. These results suggest that For the Ca2+-free solutions, the KH buffer contained 1 mM acidic stimuli may be an endogenous activator of capsaicin- CaCl2 and 10 mM EGTA. Under these conditions, the free gated currents and therefore may rationalize why pain is Ca2+ and Mg2+ concentrations were calculated to be 7.4 nM produced when the plasma acidity is increased, as occurs and 0.056 mM, respectively. The chamber had a vol of200 p1 during ischemia and inflammation. and KH flowed at a rate of 6 ml/min. Capsaicin, resiniferatoxin (RTX), and capsazepine were dissolved in KH and a Capsaicin is eaten daily by over a third of the world's small concentration of dimethyl sulfoxide. They were added population (1) and produces its pungent taste sensation when to the main flow stream by injection to ensure that the effects it binds to on a subset of and warm were reversible. Experiments were performed at 21°C. receptors nociceptors Whole-cell configuration patch-clamp recordings were per- thermoreceptors (2-8). Upon prolonged application, it blocks formed with an Axoclamp 1D patch-clamp amplifier (Axon nerve fiber conduction and depletes peptides from nerve Instruments, Burlingame, CA). The output was digitized with terminals, which accounts for its clinical use as an analgesic a Digidata 1200 A/D converter (Axon Instruments). Series and anti-inflammatory agent (9, 10). Capsaicin is thought to resistance was compensated but leak currents were not. produce these effects by activating a single type of cation- Electrode resistances were 2-5 Mfl. The microelectrode selective ion channel present in a subset ofnociceptive fibers solution always contained 140 mM KCI, 1.0 mM CaCl2, 2.0 (3). mM MgCl2, 10 mM EGTA, 10 mM Hepes, and 5 mM K2-ATP In the process of investigating the responses of trigeminal (pH 7.2). All drugs were purchased from Sigma except for ganglion cells to chemical stimuli (11, 12), we found that capsazepine, which was purchased from Cookson Chemical capsaicin also activates a current with characteristics quite (Southampton, U.K.). Rat trigeminal cells had a mean mem- distinct from the one previously described. The two currents brane potential of -52 mV (12). activated by capsaicin exhibit many similarities with the two currents activated by acidic stimuli, suggesting that acid may be an endogenous activator of these capsaicin-activated RESULTS currents. In rat trigeminal neurons held at -60 mV and bathed in modified KH buffer, 0.25 ,uM capsaicin activated two inward MATERIALS AND METHODS currents (Fig. 1). This concentration was chosen because it is near the value of the dissociation constant for capsaicin in a Rat trigeminal cells were cultured and whole-cell currents variety of tissues (8). Approximately half the cells tested were recorded by conventional culturing (13) and patch- (57.3%; n = 131) were activated by 0.25 puM capsaicin, but clamp (14) methods. Cells were cultured as follows. Male and different cells displayed different ratios and magnitudes of female Sprague-Dawley rats (150-250 g) were anesthetized these two currents (see Figs. 1-5). The early current (RI-C) with sodium pentobarbital (50 mg per kg of body weight) and activated rapidly with a mean time to peak (ap) and mean their trigeminal ganglia were excised. Trigeminal ganglion inactivation time constant (X) of 1.7 and 3.0 s, respectively. cells were washed several times in cold (4°C) modified In contrast, the later current (SI-C), which has been de- Hanks' balanced salt solution (HBSS) containing 130 mM scribed (5, 15), activated and inactivated more slowly with its NaCl, 5 mM KCl, 0.3 mM KH2PO4, 4 mM NaHCO3, 0.3 mM Abbreviations: DRG, dorsal root ganglion; RTX, resiniferatoxin; The publication costs of this article were defrayed in part by page charge RR, ruthenium red; RI-C, rapidly inactivating capsaicin current; payment. This article must therefore be hereby marked "advertisement" SI-C, slowly inactivating capsaicin current; RI-H, rapidly inactivat- in accordance with 18 U.S.C. §1734 solely to indicate this fact. ing acid current; SI-H, slowly inactivating acid current. 738 Downloaded by guest on October 6, 2021 Neurobiology: Liu and Simon Proc. Natl. Acad. Sci. USA 91 (1994) 739 A B C +RR, 1 uM Capsaicin, 0.25 ,uM Capsaicin, 0.25 ,uM * I X .

I 1UUnnApA + Capsazepine, 10 ,M 100 s 100 pA 100 pA 200 pA 7 1 100 s III FIG. 1. Responses of three rat trigeminal ganglion cells to 0.25 AM capsaicin. Capsaicin elicits two inward currents; the extent 1000 pA depends on the particular cell. (A) A cell containing both a RI-C and 100 s a SI-C current of about the same magnitude. (B) A cell containing predominantly SI-C currents. (C) A cell containing predominantly FIG. 3. Inhibition of capsaicin-activated currents by 1 ,uM RR RI-C currents. Arrows indicate the onset of the application of (upper three traces) and 10 uM capsazepine (lower three traces). capsaicin. Vh = -60 mV. Capsaicin (0.25 ,uM; arrow) added to KH buffer (left traces) induced both RI-C and SI-C currents. RR and capsazepine (middle traces) mean rp and r being 24.5 and 40.0 s, respectively. Of the 75 reversibly inhibit the responses to capsaicin (right traces). An capsaicin (0.25 ,M)-activated fibers, 41 had RI-C and SI-C interesting feature seen only with capsazepine is that after extensive currents, where both currents were about the same magni- washout an additional inward current was present (lower right trace). tude (Fig. !A), 21 exhibited predominantly SI-C currents Vh = -60 mV. (Fig. 1B), and 13 exhibited predominantly RI-C currents (Fig. 1C). genic inflammation and pain. In addition, RTX competes with In dorsal root ganglion cells (DRGs), it was shown that the capsaicin for a binding site(s) on DRG and trigeminal ganglion SI-C current is permeable to Na+, K+, and Ca2+ and has a neurons (20). In trigeminal ganglion neurons, which showed reversal potential near 0 mV (5, 16). In trigeminal ganglion both types of capsaicin-activated currents, 2.5 nM RTX (10 neurons, both the RI-C and the SI-C currents also have cells) and 100 nM RTX (5 cells) did not activate the RI-C reversal potentials that ranged between +10 and 0 mV and current. Rather, RTX activated only an inward current, between -10 and 0 mV, respectively (n = 8) (Fig. 2). This having the same time to peak as the SI-C current, but it suggests that they are also permeable to Na+, K+, and remained activated for much longer (Fig. 4A), which could possibly Ca2+ (see below). reflect the very hydrophobic character of RTX. The RTX- Capsaicin-induced fluxes and currents have been shown to induced current was not reversible, even after prolonged be inhibited noncompetitively by ruthenium red (RR) and washing (up to 30 min) with KH buffer. To further demon- competitively by the capsaicin analogue capsazepine (15, 17, strate that the two capsaicin-induced currents are distinct, a 18). Although RR also inhibits a variety of calcium channel trigeminal ganglion cell that elicited both the RI-C and the currents (19), capsazepine appears to be a specific inhibitor SI-C currents was then washed to remove capsaicin and ofcapsaicin-activated currents (15). In rat trigeminal ganglion exposed to 2.5 nM RTX. After the RTX-induced current neurons, both 1 ,uM RR and 10 ,uM capsazepine reversibly became relatively constant, 0.25 uM capsaicin was added to inhibited both the capsaicin (0.25 ,uM)-activated currents the bathing solution and the RI-C current was again activated (Fig. 3). RR inhibited 75.3% + 27% and >95% (n = 4) ofRI-C (Fig. 4B). and SI-C currents, respectively, whereas capsazepine inhib- The different physiological responses seen between cap- ited both the capsaicin-activated currents by >95% (n = 7). saicin and RTX (20) could be the consequence of the RI-C In contrast to capsazepine, which is a capsaicin analogue current and the longer activation time produced by RTX. The and antagonist, RTX is an ultrapotent capsaicin analogue and longer time of activation produced by RTX (Fig. 4A) will agonist (Kd = 0.27 nM) that mimics most (20), but not all (21), permit more current (including Ca2+) to enter the nerve fiber of the physiological responses of capsaicin, including neuro- and this feature could explain why a single stimulus of Vh=-60 mV -10 mV Capsaicin, 0.25 ,tM RTX, 2.5 nM A

-40 mV 0 mV i Capsaicin, 0.25 AM RTX, 2.5 nM Capsaicin, 0.25 jtM B t -20 mV 10 mV

1000 pA 5 nA 100 s 100 s FIG. 4. (A) Response of a trigeminal ganglion cell to 0.25 jLM FIG. 2. Response of a single trigeminal ganglion cell to the capsaicin. After washing the fiber, 2.5 nM RTX was added. (B) Same application of 0.25 M capsaicin (arrows) at different holding poten- as in A except that 2.5 min after the addition of RTX 0.25 ,uM tials ranging from -60 to +10 mV. capsaicin was added and induced a RI-C current. Vh = -60 mV. Downloaded by guest on October 6, 2021 740 Neurobiology: Liu and Simon Proc. Natl. Acad. Sci. USA 91 (1994) capsaicin only partially desensitizes and RTX completely selectivity changes to monovalent cations but still requires desensitizes these nociceptive fibers (20). Ca2+ for activation (28-30). The SI-H current is mainly The association of acid with pain has long been recognized produced by the influx of monovalent cations (25). The RI-H (22) and thus it is ofinterest to ascertain whether the currents and the SI-H currents have reversal potentials [range, -12 to elicited by capsaicin are associated with those elicited by +8 mV (25)] and therefore are similar to those found for the acidic stimuli. Analyses ofresponses from primary afferent C capsaicin-activated currents (Fig. 2). fibers (23, 24) and cultured DRG cells (25) that were stimu- Since the RI-H current is a modified Ca2+ channel, we lated by acid and/or capsaicin suggested that protons may be investigated whether standard Ca2+ blockers would inhibit an endogenous activator of the capsaicin-activated (SI-C) the capsaicin-induced currents. Antagonists of voltage-gated current. DRGs respond to acidic stimuli by activating two Ca2+ channels such as nifedipine (10 ,uM; L-type), diltiazem inward currents (25): one that rapidly activates and inacti- (1 mM; T-type), and verapamil (100 ,uM; N-type) (31) did not vates (RI-H) and one that slowly activates and inactivates significantly inhibit either the RI-C (data not shown) or the (SI-H). RI-H current (28). These antagonists did not inhibit the SI-C To further explore the association between acid and cap- current in DRGs (18, 32) or rat trigeminal ganglion cells (data saicin, we directly compared the responses to acid and not shown). capsaicin on trigeminal ganglion cells. Virtually all (24/25) of To determine whether there are still other similarities the cells tested that responded to 0.25 ,uM capsaicin also between the currents elicited by capsaicin and acid, we responded to acidic stimuli (pH 6) (Fig. 5A). Under these investigated the effects of removing Ca2+ from the extracel- specific conditions, the magnitude and kinetics of the RI-C lular solution. In physiological studies, it was found that and RI-H and SI-C and SI-H currents in the same cell have reducing extracellular Ca2+ prevents peptide release induced a similar appearance. In addition, cells have been identified by acid and/or capsaicin but does not prevent the excitation in which both the RI-H and SI-H currents are completely of nociceptive fibers by these stimuli (26, 33). Decreasing the inhibited by 1 uM RR and 10 ,uM capsazepine (data not extracellular Ca2+ by adding 10 mM EGTA, which greatly shown). In this regard, physiological experiments on muscle decreases the free concentrations of both Ca2+ and Mg2+, or revealed that RR inhibits peptide release induced by either by substituting 2 mM CaCl2 with isotonic glucose (which low pH or capsaicin (26) and, in addition, capsazepine has reduces Ca2+ but not Mg2+), eliminated (>95%; n = 9) the been shown to inhibit capsaicin- and acid-induced bronchoc- RI-C current (Fig. 5B) and >90% (n = 6) ofthe RI-H current onstriction and nasal irritation (27). The RI-H current has (Fig. SC). The similarities of the RI-C and RI-H currents in been characterized as a Ca2+-selective channel whose ion regard to their kinetic responses, selectivity for Ca2+ over Mg2+, reversal potentials, and inhibition by RR and cap- Capsaicin, 0.25 AM pH 6.0 sazepine suggest that the same current pathway may be activated by acid and capsaicin. A + Preliminary experiments indicate that the RI-C current is permeable to Ca2 . Upon replacing the Na+ in the extracellular KH solution by the large impermeant cation N-methyl- D-glucamine (but with 2.0 mM CaCl2 still present), the RI-C 1 nA current was inhibited 72% (n = 7) as compared to >95% in the absence of extracellular Ca2+. 100 s It is well established that the SI-C currents in DRGs are permeable to Ca2+ (5, 18) and the same appears true for the Capsaicin, 0.25 AM SI-C current in trigeminal ganglia. That is, only 16.2% ± 9.9% (n = 6) of the SI-C current was inhibited after reducing Ca2+ (with EGTA). In contrast, two diametrically opposite responses were obtained for the SI-H current after reducing extracellular Ca2 . In 4/6 cells, the SI-H current was inhibited >90% (Fig. SC), whereas in 2/6 cells it was not inhibited at all. The relationship between Ca2+ and its ability to 1 nA activate the SI-H current clearly requires additional investi- 100 s gation. pH 6.0 In summary, a capsaicin-activated, inward current was identified in rat trigeminal ganglion neurons. These data suggest that this rapid current is also activated by acidic stimuli. The slower capsaicin and RTX-activated currents involve action potential generation, since this property can be produced even in the absence ofextracellular Ca2+. Although the physiological significance for the rapid capsaicin- 1 nA activated current remains to be determined, one possibility is 100 s that it might be involved in peptide release, since neither it nor peptide release is activated in the absence of Ca2+. The finding that there may be subtypes of capsaicin receptors FIG. 5. (A) Response of a trigeminal ganglion cell to 0.25 uM suggests that non-narcotic analgesic and anti-inflammatory capsaicin (left trace) and (after washing) to KH buffer adjusted to pH drugs can be designed to inhibit pain or the release of specific 6.0 with HCI. (B) Effect of removing extracellular Ca2+ on response neuromodulators. to 0.25 MM capsaicin (arrows). Left trace, control (KH buffer with Ca2+); middle trace, KH buffer without Ca2+; right trace, control. We thank Drs. R. Anholt, D. Lo, S. Patterson, and D. Riddell for (C) Effect of removing extracellular Ca2+ on response to pH 6.0 comments and Dr. C. Freudenrich for calculating the free Ca2+ (arrows). Left trace, control (KH buffer adjusted to pH 6.0); middle concentration. This work was supported by Grant DC-01065 from the trace, KH buffer at pH 6.0 where CaCl2 was removed as described National Institutes of Health and by a grant from the Smokeless above; right trace, control. Vh = -60 mV. Tobacco Research Council. Downloaded by guest on October 6, 2021 Neurobiology: Liu and Simon Proc. Natl. Acad. Sci. USA 91 (1994) 741 1. Rozin, P. (1990) in Chemical Senses: Irritation, eds. Green, 17. Dray, A., Forbes, C. A. & Burgess, G. M. (1990) Neurosci. B. G., Mason, J. R. & Kare, M. R. (Dekker, New York), Vol. Lett. 110, 52-59. 2, pp. 231-269. 18. Wood, J. N., Winter, J., James, I. F., Rang, H. P., Yeats, J. & 2. Baccaglini, P. I. & Hogan, P. G. (1983) Proc. Natl. Acad. Sci. Bevan, S. (1988) J. Neurosci. 8, 3208-3220. USA 80, 594-598. 19. Amann, R. & Maggi, C. A. (1991) Life Sci. 49, 849-856. 3. Bevan, S. & SzolcsAnyi, J. (1990) Trends Pharmacol. Sci. 11, 20. Szallasi, A. & Blumberg, P. M. (1990) Brain Res. 524, 106-111. 330-333. 21. Szolcsanyi, J., Bartho, L. & Petho, G. (1991) Acta Physiol. 4. Maggi, C. A. & Pierau, F.-K. (1991) Neurosci. Lett. 122, Hung. 77, 293-304. 199-201. 22. Lindhal, 0. (1962) Acta Rheum. Scand. 8, 161-169. 5. Marsh, S. J., Stanefeld, C. E., Brown, D. A., Davey, R. & P. W., Anton, F. & Handwerker, H. 0. (1992) McCarthy, D. (1987) Neuroscience 23, 275-289. 23. Steen, K., Reeh, 6. Szolcsinyi, J. (1977) J. Physiol. (Paris) 73, 251-259. J. Neurosci. 12, 86-95. 7. Holzer, P. (1988) Neuroscience 24, 739-768. 24. Geppetti, P., Tramontana, M., Patacchini, R., Del Bianco, E., 8. Szallasi, A. & Blumberg, P. M. (1990) Life Sci. 47, 1399-1408. Santicoli, P. & Maggi, C. A. (1990) Neurosci. Lett. 114, 9. Rang, H. P. (1991) in Proceedings ofthe VIth World Congress 101-106. on Pain, eds. Bond, M. R., Charlton, J. L. & Woolf, C. J. 25. Bevan, S. & Yeats, J. (1991) J. Physiol. (London) 433, 145-151. (Elsevier, Amsterdam), pp. 119-127. 26. Santicoli, P., Bianco, E. D., Geppetti, P. & Maggi, C. A. (1992) 10. Craft, R. M. & Porreca, F. (1992) Life Sci. 51, 1767-1775. Neurosci. Lett. 143, 19-22. 11. Wang, Y., Erickson, R. E. & Simon, S. A. (1993) J. Gen. 27. Lou, Y.-P. & Lundberg, J. M. (1992) Biochem. Biophys. Res. Physiol. 101, 1-24. Commun. 189, 537-544. 12. Liu, L., Ma, H., Pugh, W. & Simon, S. A. (1993) Brain Res. 28. Konnerth, A., Lux, H. D. & Morad, M. (1987) J. Physiol. 617, 37-42. (London) 386, 603-633. 13. Lindsay, L. M. (1988) J. Neurosci. 7, 2394-2405. 29. Pidoplichko, V. I. (1992) Gen. Physiol. Biophys. 11, 39-48. 14. Hamill, 0. P., Marty, A., Sakmann, B. & Sigworth, F. J. (1981) 30. Krishtal, 0. A. & Pidoplichko, V. I. (1980) Neuroscience 5, Pflugers Arch. 391, 85-100. 2325-2327. 15. Bevan, S., Hothi, S., Hughes, G., James, I. F., Rang, H. P., 31. Miller, R. J. (1987) Science 235, 46-52. Shah, K., Walpole, C. S. J. & Yeats, J. C. (1992) Br. J. 32. Dray, A., Bettaney, J. & Forster, P. (1990) Br. J. Pharmacol. Pharmacol. 107, 544-552. 101, 727-733. 16. Bevan, S. & Forbes, A. (1988) J. Physiol. (London) 398, 28P 33. Amann, R., Donnerer, J. & Lembeck, F. (1989) Neuroscience (abstr.). 32, 255-259. Downloaded by guest on October 6, 2021