Proc. Natl. Acad. Sci. USA Vol. 85, pp. 3618-3622, May 1988 Neurobiology Single K+ channels activated by D2 dopamine receptors in acutely dissociated neurons from rat corpus striatum (patch-clamp/caudate/putamen/quinpirole/antipsychotic) JONATHAN E. FREEDMAN* AND FORREST F. WEIGHT Section of Electrophysiology, Laboratory of Physiologic and Pharmacologic Studies, National Institute on Alcohol Abuse and Alcoholism, 12501 Washington Avenue, Rockville, MD 20852 Communicated by Solomon H. Snyder, December 28, 1987 ABSTRACT Corpus striatum neurons acutely dissociated from the brains of young adult rats had membrane surfaces suitable for Gi-seal recording. Whole-cell current-clamp and voltage-clamp recordings indicated that the cells remained electrically excitable after dissociation. Cell-attached recordings frequently revealed single-channel openings in the presence of dopamine or of the D2 dopamine agonist quinpirole. Channel openings were rarely or never observed in the absence of drugs or in the presence of quinpirole plus the dopamine antagonist haloperidol. The D2 antagonist spiperone was more potent at blocking the appearance of the channel than was the D1 antagonist SCH-23390. The channel reversal potential varied with the extracellular K+ concentration as predicted by the Nernst equation. The channel currentvoltage relationship was linear, with a conductance of m85 ps in the presence of 140 mm KCL. These results are consistent with the opening of single K+ channels following D2 dopamine receptor activation. The corpus striatum (caudate and putamen) is a major postsynaptic target of dopaminergic projections in mammalian brain (1, 2). Because of the role of dopamine in the actions of antipsychotic drugs (1, 2) and because of the potential usefulness of studying these actions with single-molecule resolution, we have studied the effects of dopamine at the singlechannel level. Patch-clamp recording in adult mammalian brain has been hampered by the need for dissociated cells. One way to obtain dissociated cells is to work with cultured embryonic tissue (3). However, rat striatal dopamine receptor radioligand binding sites are largely absent in the neonate and do not reach adult levels until =28 days after birth (4). A method has been described for enzymatically splitting the hippocampus along the pyramidal cell layer (5) but may not be useful for brain regions such as the striatum, which are not organized into cellular layers. Kay and Wong (6) have described a method for acutely dissociating adult guinea pig brain, based upon trypsin treatment in a slightly acidic Pipes buffer. Kay et al. (7) and Huguenard and Alger (8) have used this method to record from guinea pig hippocampal pyramidal neurons. We have made minor modifications in their procedure to dissociate striatal neurons of young adult rats. We report here that we have used these dissociated cells to obtain an initial description at the single-channel level of the actions of dopamine in mammalian brain. METHODS Dissociation. This procedure is based upon the methods of Kay and Wong (6). Male Sprague-Dawley rats, 31-45 days The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. 1734 solely to indicate this fact. 3618 FIG. 1. Phase-contrast photomicrograph of a typical acutely dissociated striatal neuron. (Bar = 10 Itm.) post partum, were the tissue source. The corpus striatum was rapidly and carefully dissected in a beeswax-coated Petri dish under ice-cold Pipes saline [120 mm NaCI, 5 mm KCl, 1 mm CaCl2, 1 mm MgCl2, 25 mm D-glucose, 20 mm Pipes-NaOH (ph 7.0 at 25 C, preequilibrated with 100% 02)1 and hand-minced with a scalpel blade into blocks of =2 mm per edge. The tissue blocks were transferred by using a wide-bore pipette to a 25-ml jacketed double-side-armr Wheaton Celstir stirring chamber, containing 130,000 benzoylarginine ethyl ester units of trypsin (Sigma type XI) in 12 ml of Pipes saline. The tissue was stirred at =40 rpm under 100% 02 at 32 C, with the stirring rate adjusted so that the tissue blocks were raised above the bottom of the chamber but not so fast as to cause the solution to become cloudy. The incubation was continued for 1 hr, after which the solution was replaced with fresh saline and trypsin, and the incubation was continued for a second hour. Soybean trypsin inhibitor (Sigma type 1-S; 7.8 mg in 0.5 ml of Pipes saline) was then added. After 5 min, the solution in the chamber was removed, and the tissue was washed twice with 12.5 ml of Dulbecco's modified Eagle's medium (DMEM) [bicarbonate-free, with 25 mm Hepes, 25 mm glucose, 4 mm glutamine (adjusted with NaOH to ph 7.4 at 25 C, and with NaCl to 340 mosmol/kg)]. A third volume of DMEM was added, and incubation was continued at 32 C for up to 6 hr. As *To whom reprint requests should be addressed.
Neurobiology: Freedman and Weight needed, 3 or 4 tissue blocks were transferred with =0.25 ml of DMEM to a 1.5-ml conical tube and tapped by finger until the solution became cloudy and the tissue blocks were reduced in size. The supernatant was poured into a polyornithine-coated Petri dish, and the cells were allowed to settle for 10 min before beginning electrophysiological recording, which was performed within 1 hr. Electrophysiology. The cells were continuously superfused at room temperature in a solution containing 150 mm NaCl, 5 mm KCI, 2.5 mm CaCI2, 1 mm MgCI2, 10 mm D-glucose, 10 mm Hepes'NaOH (ph 7.4, adjusted with sucrose to 340 mosmol/kg). Borosilicate glass patch pipettes were filled with the solutions indicated in the figure legends and had resistances of 3-5 Mfl. Patch-clamp recording was performed in the whole-cell and cell-attached configurations (9) by using a List EPC-7 amplifier. Amplifier output went through an 8-pole Bessel filter (4-kHz lowpass for whole-cell recording or 2-kHz lowpass for cell-attached recording) to a 12-bit A/D converter (5 khz for whole-cell voltage-clamp, 8 khz for current-clamp, and 10 khz for single-channel re- A Proc. Natl. Acad. Sci. USA 85 (1988) 3619 cordings) for storage and analysis on a Digital PDP-11/73 computer. Recordings were made from phase-bright cells free of apparent membrane blisters and retaining visible neuritic processes. Whole-cell recordings were deemed suitable for use if membrane potential was not less negative than -40 mv and if the action potential amplitude exceeded 80 mv. Cell-attached recordings were used if seal resistance exceeded 5 Gfi and if the cell remained phase-bright throughout the recording. Drugs and Reagents. Quinpirole was a gift of Eli Lilly. Spiperone and (R)-( + )-SCH-23390 were from Research Biochemicals (Natick, MA). All other drugs and reagents were from Sigma. RESULTS A typical dissociated striatal cell is shown in Fig. 1. Cells had truncated neuritic processes and were typically 15-20 gm in diameter. Cell surfaces proved suitable for Gfl-seal formation. Although the yield of cells was variable, it was usually B 4 40 ms 3 na _ pa r 0 L -25 25 ms mv - +30 C - 8 0-6 0-4 0-2 0 2 0 4 0 mv D na 4- / 3-2 - -5 I rt--'- s T- - 8 0-6 0-4 0-2 0. ] 20 40 mv na -9 FIG. 2. Whole-cell recordings from dissociated striatal neurons. Patch pipettes contained 140 mm KCI, 1 mm CaC12, 2 mm MgC12, 11 mm EGTA, and 10 mm Hepes-KOH (ph 7.4, 320 mosmol/kg). (A) Action potential recorded in current-clamp mode. Voltage (upper trace) and current (lower trace) are shown. The cell had a resting membrane potential of - 50 mv. A 25-pA hyperpolarizing pulse was applied for 80 msec and elicited an anode-break spike. (B) Voltage-clamp record of another cell, held at - 90 mv and step-depolarized for 75 msec to values up to + 30 mv in 5 mv increments, eliciting fast inward and delayed outward currents. Current (upper traces) and voltage (lower traces) are shown. (C) Leak-subtracted current-voltage curve for the total inward current of the cell in B. Current amplitude was measured 1 msec after the onset of the depolarizing step. Leakage current was estimated by extrapolating a linear regression fit of data at -60 mv and more negative. (D) Leak-subtracted current-voltage curve of the total outward current of the cell in B. Current amplitude was measured 70 msec after the onset of the depolarizing step.
3620 Neurobiology: Freedman and Weight possible to obtain 4 or 5 good cell-attached recordings per day, with somewhat fewer recordings in the whole-cell configuration. Whole-cell current-clamp and voltage-clamp recordings indicated that the cells remained electrically excitable after dissociation (Fig. 2). For 22 cells tested in the current-clamp mode, the mean resting membrane potential was -55 mv (range, -75 to -40 mv), and the mean input resistance was 700 Mfl (range, 200 Mfl to 1.6 GQ). Action potentials (Fig. 2A) had a mean amplitude of 99 mv (range, 84-116 mv), a rate of rise of 209 V/sec (range, 72-420 V/sec), and a duration (at -40 mv) of 1.5 msec (range, 0.5-4.8 msec). Voltage-clamp recordings (n = 14 cells) revealed fast inward and delayed outward currents (Fig. 2 B-D). For the total inward current, the mean peak amplitude was 6.0 na (range, 1.0-8.1 na), with a threshold at -49 mv (range, -65 to -40 mv). For the total outward current, the mean amplitude A No drug 5pA B Quinpirole (10 gm) 10 ms C Quinpirole+Haloperidol (10 gm) 0 pa FIG. 3. Cell-attached recordings of channel currents associated with dopamine receptor activation. Patch pipettes contained 140 mm KCl, 2.5 mm CaCl2, 1 mm MgCl2, 10 mm Hepes-KOH (ph 7.4), plus drugs as indicated. (A) Control patch, with no drug in the electrode. (B) Second cell with an electrode containing 10 /AM quinpirole. (C) Third cell with 10,uM quinpirole plus 10 ILM haloperidol. Zero-current level is indicated by dashed lines at the right of each trace. Upward deflection corresponds to inward current. Pipette potential was 0 mv, so patch membrane potential was the resting potential of the cell. at +30 mv was 4.3 na (range, 1.8-8.8 na); the threshold was -35 mv (range, -60 to -15 mv). Outward current showed partial inactivation within 75 msec in 6 of 14 cells (data not shown). Whole-cell currents and potentials showed a marked run-down after =5 min; the above data were all obtained within 5 min of beginning the recording. When cell-attached recordings were performed with 140 mm KCI inside the patch electrode, no channel openings were seen at the resting membrane potential (Fig. 3A). (The KCI in the pipette is expected to depolarize the cell by not more than 1 mv, based upon the relative areas of the pipette tip and the cell surface.) In contrast, when the pipette also contained 10 AM of either dopamine or the D2 dopamine agonist quinpirole (10, 11), channel activity was frequently observed (Fig. 3B). Channels typically spent a high percentage of the time in the open state, often in prolonged bursts, and there was frequently more than one channel per patch. When the pipette contained 10,M haloperidol, a dopamine antagonist and antipsychotic drug (1,2) (Fig. 3C), in addition to quinpirole, cells displaying channels were observed less frequently (15% versus 71%, Table 1). Since it was not possible to change the solution inside the electrode, we tested a number of cells with various pipette solutions (Table 1). Because the presence or absence of channel activity correlated with the presence or absence of dopamine receptor activation, there is a high probability that this channel was activated by dopamine receptors. Since spiperone, which selectively antagonizes D2 receptors as well as some nondopaminergic receptors (1, 2, 11), was more potent at preventing the observation of channels than was the selective D1 antagonist SCH-23390 (11, 12) (Table 1), these receptors appear to be of the D2 subtype. In preliminary experiments, we did not observe channel openings when 10,uM quinpirole was applied in the bath rather than in the pipette (zero cells responding of four tested). Channel openings were observed over a wide range of patch membrane potentials (Fig. 4A, n = 23 cells), approaching zero amplitude as the patch was depolarized by 40-60 mv (similar to resting membrane potentials seen in whole-cell recordings, suggesting a patch potential near 0 mv, with K+ roughly isotonic). The current-voltage relationship was linear, with a channel conductance of =85 ps in the presence of 140 mm KCI (Fig. 4B). With further depolarization the recording became noisy; discrete outward channel openings were not seen. When the pipette K+ concentration was reduced from 140 mm to 40 mm, the extrapolated reversal potential of the channel shifted to =15-25 mv above resting potential (Fig. 4B, n = 5 cells). Table 1. Proc. Natl. Acad. Sci. USA 85 (1988) Pharmacology of channel openings No. of cells responding/ Pipette solution no. of cells tested No drug 0/14 (0%o) Quinpirole (10,uM) 10/14 (71%) Quinpirole (10,M)/haloperidol (10,M) 2/13 (15%) Quinpirole (10 AM)/spiperone (0.5 jm) 0/6 (0%7) Quinpirole (10,uM)/SCH-23390 (1,uM) 3/6 (50%o) Quinpirole (10,uM)/sulpiride (25,uM) 4/10 (40%6) Dopamine (10 jm) 4/8 (50%o) Experiments were performed as described in Fig. 3, with the indicated drugs added to the pipette solution. Cells were observed for at least 10 min. Cells were rated as responding if a channel opening was observed; cells rated as not responding showed no channel openings per 10 min. Numbers in parentheses are percentage of cells responding. In fact, all responding cells showed >5 channel openings per min, starting as soon as the recording was begun. Dopamine solutions also contained 100 jim ascorbic acid to retard oxidation. Ascorbic acid was without effect in control experiments.
Neurobiology: Freedman and Weight A B Proc. Natl. Acad. Sci. USA 85 (1988) 3621-20 mv + RMP n 5 pa K 10 ms pa 9 0 mv + RMP -- OpA * [K]=140 * [K]=40 +20 mv + RMP _X.~~~~~~~~~~~~~~~~-- +40 mv + RMP I1 r r L -40-20 20 40 60 mv (+RMP) +60 mv + RMP FIG. 4. Voltage and K+ dependence of channel currents. Recordings were performed as described in Fig. 3 with 10,uM quinpirole inside the pipette. (A) Current records at various potentials. Dashed lines denote zero-current level (excluding instantaneous leakage current); upward deflections correspond to inward current. Patch membrane potentials are expressed as the pipette potential multiplied by -1 plus the resting membrane potential (RMP), which is not known in this recording configuration. (B) Current-voltage relationship for single-channel currents. (Squares) Cell in A with 140 mm KCI in the patch pipette. (Triangles) Another cell, for which the pipette KCI concentration was decreased to 40 mm with substitution of 100 mm N-methyl-D-glucamine chloride as an impermeant cation. The lines were fit by linear regression. DISCUSSION We have described single-channel openings associated with D2 dopamine receptor activation, by using acutely dissociated rat corpus striatum neurons. This acute preparation made possible the use of cells from rats sufficiently mature to be expected to have adult levels of dopamine receptors (4). Because trypsin was used in the dissociation procedure, one must be aware of the possibility of an alteration of cell properties by proteolysis. We found, however, that the cells appeared to retain chemoresponsivity to dopamine and had whole-cell electrophysiologic parameters consistent with those observed in striatal cells by using intracellular recording with microelectrodes (13). The one exception was the high input resistance, which can be accounted for by the loss of dendritic membrane and by GQi-seal recording conditions, as noted by others using dissociated cells (6, 8). It is not clear whether the dissociation selected for some subpopulation of striatal neurons. We employed the cell-attached configuration when we examined the effects of dopaminergic compounds, so as not to risk losing any second messenger or other cytoplasmic component that might be necessary for a response. Drug solutions were placed inside the recording electrode, rather than in the bath outside the patch. We could thus rule out the possibility that channel openings we observed occurred secondarily to a change in membrane potential, and we were less likely to overlook channels coupled directly to the receptor without a diffusible second messenger. (Our preliminary experiments with bath-applied quinpirole should be interpreted with caution for these reasons and because of the small number of cells.) However, a disadvantage of our approach is that we could not change the pipette solution during an experiment, to observe a single cell before and after drug application. We, therefore, tested a number of cells with various pipette solutions; our results indicate that there was a high probability that channel openings were associated with receptor activation. The receptors associated with these channels showed pharmacologic properties of the D2 dopamine receptor subtype, as indicated by the efficacy of quinpirole and the greater potency of spiperone than SCH-23390. We found, however, that sulpiride, a highly selective D2 antagonist (11), was relatively ineffective as an antagonist in our experiments (Table 1). Although this fails to provide further support for a D2 subtype, it could be due to the fact that we used a high K+,Na+-free solution to resolve large singlechannel currents, whereas sulpiride, unlike most dopamine antagonists, requires a high Na+ medium for antagonist activity (14). Changing the pipette K + concentration from 140 to 40 mm is predicted by the Nernst equation to shift the reversal potential of a K+ channel by 32 mv, assuming that resting membrane potential and internal K+ concentration were consistent from cell to cell. (These assumptions are potentially large sources of error.) The observed shift of -30-35 mv supports permeability to K + but does not rule out permeability to other ions as well. Under physiological conditions, increased K + permeability is expected to give rise to an outward current resulting in hyperpolarization. D2 dopamine autoreceptors in the substantia nigra are known to mediate an inhibitory response (15, 16)
3622 Neurobiology: Freedman and Weight associated with an increased K+ conductance (16, 17). Although pre- and postsynaptic dopamine receptors may differ (18), it is possible that both these responses are mediated by the channel described here. However, both inhibitory (18, 19) and excitatory (20-22) responses have been attributed to D2 receptors in the striatum and other postsynaptic regions. Thus, our present results in no way rule out the existence of additional channels also controlled by D2 receptors. Our data suggest that D2 receptor activation opens these single K+ channels. Further experiments will be needed to determine whether these channels are directly gated by the receptor or are coupled to the receptor through calcium, guanine nucleotide-binding regulatory proteins, or a second messenger system. Indeed, single-channel recording offers a useful approach for studying the mechanisms of action of dopamine in mammalian brain. 1. Seeman, P. (1980) Pharmacol. Rev. 32, 229-313. 2. Creese, I., Sibley, D. R., Hamblin, M. W. & Leff, S. E. (1983) Annu. Rev. Neurosci. 6, 43-71. 3. Heyer, E. J. (1986) Brain Res. 382, 404-408. 4. Pardo, J. V., Creese, I., Burt, D. R. & Snyder, S. H. (1977) Brain Res. 125, 376-382. 5. Gray, R. & Johnston, D. (1987) Nature (London) 327, 620-622. 6. Kay, A. R. & Wong, R. K. S. (1986) J. Neurosci. Methods 16, 227-238. 7. Kay, A. R., Miles, R. & Wong, R. K. S. (1986) J. Neurosci. 6, 2915-2920. Proc. Natl. Acad. Sci. USA 85 (1988) 8. Huguenard, J. R. & Alger, B. E. (1986) J. Neurophysiol. 56, 1-18. 9. Hamill, 0. P., Marty, A., Neher, E., Sakmann, B. & Sigworth, F. J. (1981) Pflugers Arch. 391, 85-100. 10. Hahn, R. A. & MacDonald, B. R. (1984) J. Pharmacol. Exp. Ther. 229, 132-138. 11. Stoof, J. C. & Kebabian, J. W. (1984) Life Sci. 35, 2281-22%. 12. Iorio, L. C., Barnett, A., Leitz, F. H., Houser, V. P. & Korduba, C. A. (1983) J. Pharmacol. Exp. Ther. 226, 462-468. 13. Bishop, G. A., Chang, H. T. & Kitai, S. T. (1982) Neuroscience 7, 179-191. 14. Stefanini, E., Marchisio, A. M., DeVoto, P., Vernaleone, F., Collu, R. & Spano, P. F. (1980) Brain Res. 198, 229-233. 15. Bunney, B. S., Walters, J. R., Roth, R. H. & Aghajanian, G. K. (1973) J. Pharmacol. Exp. Ther. 185, 560-571. 16. Lacey, M. G., Mercuri, N. B. & North, R. A. (1987) J. Physiol. (London) 392, 397-416. 17. Chiodo, L. A. & Kapatos, G. (1987) Soc. Neurosci. Abstr. 13, 1180. 18. Skirboll, L. R., Grace, A. A. & Bunney, B. S. (1979) Science 206, 80-82. 19. White, F. J. & Wang, R. Y. (1986) J. Neurosci. 6, 274-280. 20. Bradshaw, C. M., Sheridan, R. D. & Szabadi, E. (1985) Br. J. Pharmacol. 86, 483-490. 21. Uchimura, N., Higashi, H. & Nishi, S. (1986) Brain Res. 375, 368-372. 22. Ohno, Y., Sasa, M. & Takaori, S. (1987) Life Sci. 40, 1937-1945.