The actions of mu- and delta-opioid agonists (DAMGO and DPDPE, respectively) on GABAergic interneurons in stratum oriens of area CA1 of the hippocampus were examined by using whole-cell voltage-clamp recordings in brain slices. Both agonists consistently generated outward currents of similar magnitude (15–20 pA) in the majority of cells. However, under control conditions, current–voltage (I/V) relationships revealed that only a small number of these cells (3 of 77) demonstrated clear increases in membrane conductance, associated with the activation of the potassium current known as G irk. These interneurons also exhibited a slowly activating, inwardly rectifying current known as I h on hyperpolarizing step commands.I h was blocked by the extracellular application of cesium (3–9 mm) or ZD 7288 (10–100 μm) but was insensitive to barium (1–2 mm). In an effort to determine whether the holding current changes were attributable to the modulation of G irkand/or I h, we used known blockers of these ion channels (barium or cesium and ZD 7288, respectively). Extracellular application of cesium (3–9 mm) or ZD 7288 (25–100 μm) blocked I h and significantly reduced the opioid-induced outward currents by 58%. Under these conditions the opioid agonists activated a potassium current with characteristics similar toG irk. Similarly, during barium (1–2 mm) application the opioid-induced outward currents were reduced by 46%, and a clear reduction in I hand the whole-cell conductance was revealed. These data suggest that the opioids can modulate both I h andG irk in the same population of stratum oriens interneurons and that the modulation of these ion channels can contribute to the inhibition of interneuron activity in the hippocampus.
- opioid receptor
- nonselective cation current
- oriens/alveus interneurons
- queer current
Opioid receptors are members of a class of Gi/Go-coupled receptors that inhibit both adenylyl cyclase and voltage-dependent Ca2+ channels (VDCCs) in central and peripheral neurons (Childers, 1993; Moises et al., 1994). In addition to these actions, opioids activate a G-protein-coupled inwardly rectifying potassium conductance known as G irk, resulting in the hyperpolarization of neurons throughout the CNS (Williams et al., 1982; Madison and Nicoll, 1988; North, 1989; Wimpey and Chavkin, 1991). These opioid effects on ion channels generally inhibit neuronal activity. However, in some cellular circuits, when the opioid-sensitive neuron releases an inhibitory neurotransmitter such as GABA, the net result can be the excitation or disinhibitionof the postsynaptic neuron. Such is the case for the principal cells of the CA1 region of the hippocampus, where opioids produce profound increases in pyramidal neuron activity after the inhibition of GABA release from local circuit inhibitory interneurons (Zieglgansberger et al., 1979; Lee et al., 1980; Nicoll et al., 1980; Pang and Rose, 1989;Lupica and Dunwiddie, 1991). These interneurons constitute only 10–20% of the neuronal population in the hippocampus and are heterogeneous with regard to their membrane properties, neurotransmitter sensitivity, and neuropeptide content (for review, seeFreund and Buzsáki, 1996). However, these cells exert pandemic control over hippocampal activity, because they each make multiple synaptic contacts on hundreds of principal neurons.
Although it is likely that opioids play an important role in the modulation of processes like learning and memory, epilepsy, and drug reinforcement via actions on interneurons in the hippocampus (Siggins et al., 1986; Stevens et al., 1991; Xie and Lewis, 1991), there have been few direct studies of opioid effects on these cells. Most have relied, instead, on indirect measures such as GABA-mediated IPSPs/IPSCs from pyramidal cells (Lee et al., 1980; Nicoll et al., 1980; Cohen et al., 1992; Lupica et al., 1992; Lupica, 1995). To date, studies using direct recordings from CA1 interneurons have examined only the effects of opioids on K+ channels and VDCCs (Madison and Nicoll, 1988;Wimpey and Chavkin, 1991; Lambert and Wilson, 1996). These investigators found that G irk was activated by the μ-selective agonist DAMGO in acutely dissociated CA1/subicular interneurons (Wimpey and Chavkin, 1991) or by a nonselective opioid agonist (d-Ala2-Met5-enkephalinamide) in stratum pyramidale interneurons (Madison and Nicoll, 1988). However, another study found that DAMGO did not activateG irk, nor did it inhibit VDCCs in stratum radiatum interneurons (Lambert and Wilson, 1996).
Another conductance that is prominent in hippocampal interneurons found in stratum oriens is known as I h (Maccaferri and McBain, 1996). This slowly developing inward cation current is activated by hyperpolarization, is carried by Na+and K+ ions, and does not inactivate, even with prolonged hyperpolarization (Halliwell and Adams, 1982; Mayer and Westbrook, 1983; Maccaferri and McBain, 1996). Because of these characteristics, I h probably contributes to the resting membrane potential and the generation of rhythmic pacemaker-like depolarizations in central neurons and cardiac cells (DiFrancesco, 1981; Bal and McCormick, 1996; Maccaferri and McBain, 1996; Gasparini and DiFrancesco, 1997) (for review, see Pape, 1996). Therefore, modulation of I h by neurotransmitters would be expected to alter the oscillatory activity of individual neurons and, in turn, the network of cells with which they communicate (Freund and Buzsáki, 1996) (for review, seePape, 1996). In addition, because of its properties,I h might interact with other conductances by opposing the membrane hyperpolarization initiated by inhibitory neurotransmitters. Among the several neuromodulators that modulateI h are the opioids, which inhibit this current in peripheral neurons by inhibiting adenylyl cyclase (Ingram and Williams, 1994). However, at this time no studies of the opioid modulation of I h in central neurons have been described. In the present investigation we demonstrate that μ- and δ-opioid receptor agonists inhibit I h and activate G irk in the same hippocampal interneurons, and we postulate that these effects contribute to the sustained inhibition of interneuron activity by the opioids.
MATERIALS AND METHODS
Electrophysiology. Hippocampal slices were prepared and maintained as previously described (Miller et al., 1997). Briefly, 14- to 30-d-old male Sprague Dawley rats (Sasco, Omaha, NE) were killed by rapid decapitation. Their brains were removed and placed in ice-cold oxygenated artificial CSF (aCSF; see below). Brain slices containing the hippocampus were cut transverse to the anterior–posterior axis at 300 μm nominal thickness, using a vibrating tissue slicer (Technical Products International, St. Louis, MO). Then the slices were suspended on netting in a beaker containing aCSF that was aerated continuously with 95% O2/5% CO2 at room temperature. Control aCSF consisted of (in mm): 126 NaCl, 3.0 KCl, 1.5 MgCl2, 2.4 CaCl2, 1.2 NaH2PO4, 11.0 glucose, and 26 NaHCO3, saturated with 95% O2/5% CO2. Interneurons were visualized in stratum oriens of area CA1, using a fixed stage upright microscope equipped with differential interference contrast optics and infrared illumination, as previously described in detail (Miller et al., 1997). Whole-cell recordings were obtained at room temperature (20–23°C) with an Axopatch-200A amplifier (Axon Instruments, Foster City, CA) and electrodes pulled from thick-walled borosilicate capillary tubing (inner diameter, 0.75 mm; outer diameter, 1.5 mm; Sutter Instrument, Novato, CA). The electrodes had resistances of 4–7 MΩ when filled with (in mm): 125.0 K-gluconate, 10.0 KCl, 10.0 HEPES, 1.0 EGTA, 0.1 CaCl2, 2.0 Mg2+-ATP, and 0.2 Na+-GTP (adjusted to pH 7.2–7.4 with 1m KOH, and brought to 270–280 mOsm with de-ionized water). The reversal potential for K+ currents (E K) using these stated internal (135.0 mm) and external (3.0 mm) K+concentrations was predicted to be approximately −96 mV at 20°C, according to the Nernst equation. Unless otherwise stated, all interneurons were voltage-clamped at −66 mV after correction for a liquid junction potential. Series resistance was generally <15 MΩ and was monitored throughout the experiments by using the capacitative currents generated by small (−5 to −10 mV, 250 msec) voltage steps. Cells were rejected from analysis if the series resistance changed by ≥10–15%. Voltage-clamp protocols were delivered by a pulse generator (Master 8, A.M.P.I., Jerusalem, Israel), and signals were acquired by a PC-based data acquisition system (Strathclyde Electrophysiology Software, courtesy of John Dempster, Strathclyde University, UK).
Histology. Biocytin (0.25%; Sigma, St. Louis, MO) was added to the internal solution for post hoc evaluation of the anatomical location of the neurons. A majority of the interneuron recordings were >1 hr in duration. In these cases, at the end of the recording period the patch pipette was withdrawn slowly from the slice, which then was transferred to chilled (4°C) 4% paraformaldehyde in PBS. In recordings of shorter duration the pipette was left in place in the slice, and the preparation was left undisturbed in the recording chamber for an additional one-half hour to permit diffusion of biocytin into the cell. Slices were stored in the chilled fixative for 3–7 d. After fixation, they were rinsed in PBS (three times for 5 min) and then exposed to avidin–biotin horseradish peroxidase complex (ABC, Vector Laboratories, Burlingame, CA) for 4 d. On the fourth day, they were rinsed again in PBS and then incubated in 0.04% diaminobenzidine (DAB) for 20–30 min. The DAB reaction was stopped by rinsing in PBS. Then the slices were dehydrated in incrementing concentrations of ethanol (70, 95, 100, 100%). After dehydration, they were cleared in Hemo-D (three times for 5 min) and then mounted on slides. Cells were reconstructed with the aid of a camera lucida drawing tube attached to an Olympus microscope.
Chemicals. Drugs were obtained from the following sources: DPDPE (D-Pen2,D-Pen5-enkephalin) and DAMGO (Tyr-D-Ala2,N-CH3-Phe4,Gly-ol-enkephalin), National Institute on Drug Abuse Drug Supply System (Rockville, MD); tetrodotoxin (TTX), Alomone Laboratories (Jerusalem, Israel) or Sigma; CsCl and BaCl2 (Sigma); ZD 7288, Tocris Cookson (Ballwin, MO). All drugs and channel blocking agents were made at 100 times their final concentration in de-ionized water and added to the aCSF bathing the slice (flow rate = 2 ml/min) with calibrated syringe pumps (Razel Scientific Instruments, Stamford, CT).
All of the recordings were obtained from neurons that had somata located within stratum oriens of the CA1 region of the hippocampus. Previous studies have shown that a vast majority of neurons found outside of stratum pyramidale are positive for glutamic acid decarboxylase (GAD) and, therefore, GABAergic (Somogyi et al., 1983;Frotscher et al., 1984; Kunkel et al., 1986). The morphology of the stratum oriens cells was diverse, and the axonal arborizations were, in many cases, elaborate. The axons of several of these cells extended through all CA1 layers, and no consistent patterns of termination were noted. Although the axonal distribution pattern was diverse, most of the interneurons possessed multiple dendritic processes restricted primarily to stratum oriens. A few of these cells exhibited dendrites extending into the alveus (Fig. 1). The morphology of these cells with somata and dendrites confined to stratum oriens is consistent with previous descriptions (Lacaille et al., 1987;McBain et al., 1994; Zhang and McBain, 1995). Examples of two biocytin-filled interneurons reconstructed with the use of camera lucida are shown in Figure 1.
Opioids generate outward currents in stratum oriens interneurons
Bath application of the selective μ-opioid receptor agonist DAMGO (1 μm) or the selective δ-agonist DPDPE (1 μm) generated reversible outward currents in the majority of stratum oriens interneurons. Of the 66 interneurons in which DPDPE was tested, 40 (60.6%) responded with outward currents, whereas 24 of 55 (43.6%) neurons that were tested were sensitive to DAMGO. The magnitudes of the outward currents were similar with either agonist (Fig. 2 A). Only 12 of 44 neurons (27%) that were tested responded to both agonists.
In neurons throughout the CNS, opioid-generated outward currents reverse near the K+ equilibrium potential (E K) and usually are associated with anincrease in membrane conductance because of the activation of a G-protein-coupled K+ channel known asG irk (Williams et al., 1982; Madison and Nicoll, 1988; North, 1989; Wimpey and Chavkin, 1991). However, in the present study, when the effects of DAMGO and DPDPE were examined on current–voltage (I/V) relationships (using 250 msec voltage steps from −66 to −136 mV) (Fig. 2 B), we found that there was either very little change in the whole-cell conductance or a small net decrease, despite the clear change in holding current (control = 7.1 ± 0.6 nS, pooled DPDPE and DAMGO = 6.8 ± 0.7 nS; n = 17; measured at steady state from the linear portion of the I/V curves). This is illustrated in Figure 2 C, in which both DPDPE and DAMGO produced changes in the I/V curves during the peak outward currents that were inconsistent with an increase in whole-cell conductance expected to result from the activation of a K+ channel like G irk. To evaluate whether these neurons were capable of generating outward currents associated with the activation ofG irk, we compared the actions of the GABAB receptor agonist baclofen, which is known to activate this conductance (Gahwiler and Brown, 1985; Inoue et al., 1985;Newberry and Nicoll, 1985; Christie and North, 1988), with the opioid agonists in the same interneurons. At a holding potential of −66 mV, baclofen (60 μm) generated outward currents (25.0 ± 4.7 pA, n = 7) that were associated with an increase in membrane conductance of 1.7 ± 0.5 nS (Fig.3). In addition, theE rev for the baclofen-induced current was −91.1 ± 4.1 mV, which is close to the calculatedE K of −96 mV. These data indicated that, whereas both the opioid and GABAB receptor agonists produced outward currents, only baclofen activated an apparent K+ channel associated with an increase in membrane conductance and characteristics similar to G irk(compare Fig. 3 B,C). A small number of the stratum oriens interneurons (8 of 94, 8.5%) exhibited current responses to voltage steps hyperpolarized to approximately −76 mV that lacked the prominent inward sag that is indicative of the hyperpolarization-activated cation current known as I h (visible in Fig.2 B). Of those eight neurons lackingI h, three demonstrated clear outward currents on opioid application. In each of these cells theI/V curves revealed that the opioid-sensitive currents reversed near E K and were associated with an increase in membrane conductance, consistent with the activation ofG irk (Fig. 4).
Characteristics of Ih in stratum oriens interneurons
The hyperpolarization-activated cation current,I h, is known to activate slowly and displays virtually no time-dependent inactivation during a voltage step (Mayer and Westbrook, 1983; Maccaferri and McBain, 1996; Watts et al., 1996). In addition, in most preparations I h is not active at membrane potentials positive to approximately −55 to −75 mV, and it typically does not saturate at potentials as negative as −155 mV (Maccaferri and McBain, 1996; Watts et al., 1996). The slow activation kinetics of I h suggest that this current should be larger at the end of a 2 sec voltage step, at which point I h is fully active (i.e., steady state,I ss), compared with the current observed near the beginning of a hyperpolarizing voltage step (i.e., the instantaneous current, I ins), where the contribution of I h is smaller. Similarly, the voltage-dependent nature of I h suggests that the effects of modulators of this current should be seen at membrane voltages within its range of activation (i.e., negative to −55 mV). This can be seen in Figure 5 in which the effects of extracellularly applied ZD 7288 (25 μm) and cesium (3 mm), both blockers ofI h, have been examined on the instantaneous and steady-state currents. Further isolation of these effects on I h was achieved by subtractingI ins from I ss at each voltage step, in the presence and absence of the modulator, using the following equation: Equation 1where I ssm andI insm are the steady-state and instantaneous currents measured in the presence of the modulator, andI ssc andI insc are these currents in the control condition. I ss was inhibited to a greater extent than I ins by both cesium and ZD 7288 (shown for ZD 7288 in Fig. 5 B), and the cesium- and ZD 7288-sensitive currents were more prominent at step potentials more negative than approximately −75 mV (Fig. 5 C).
Voltage steps greater than −136 mV caused a rapid degradation of the recordings and, therefore, were not used routinely. Overall, cesium (3–9 mm) decreased the whole-cell conductance by ∼60% (control = 7.1 ± 1.1 nS; cesium = 4.3 ± 0.85 nS;n = 6), indicating that the contribution ofI h to the total whole-cell conductance was substantial. Consistent with previous reports, the block ofI h by ZD 7288 began after 4–7 min of superfusion, and a maximal blockade was achieved after 20–50 min (Harris and Constanti, 1995; Maccaferri and McBain, 1996; Gasparini and DiFrancesco, 1997). A complete block of I h by cesium was achieved only at concentrations >3 mm but was seen much more rapidly (3–5 min) than with ZD 7288 (data not shown). In addition to the alteration in whole-cell conductance, cesium also generated outward currents ranging from 3.5 to 19 pA in amplitude (mean = 7.5 ± 2.0 pA; n = 8), suggesting that I h may be activated partially in some interneurons at a clamp potential of −66 mV. These effects of extracellular cesium are consistent with previous reports demonstrating the blockade of I h in other preparations (Mayer and Westbrook, 1983; McCormick and Pape, 1990; Bal and McCormick, 1996). The reversal potential of I h(E h) was measured by using the method described by Mayer and Westbrook (1983). Thus, instantaneousI/V curves were generated by six to eight voltage steps (2 sec) from holding potentials of −56, in whichI h is minimally active, and −96 mV, in whichI h is activated substantially. Then the intersection of the extrapolated I/V curves was used as the estimate of E h. The estimate ofE h obtained by using this method was −33.0 ± 2.0 mV (n = 7), which is similar to the value previously reported in these interneurons (Maccaferri and McBain, 1996).
Opioids inhibit Ih in stratum oriens interneurons
When the effects of DAMGO (1 μm) and DPDPE (1 μm) were examined on currents evoked by 2 sec hyperpolarizing voltage steps, it was found that in some neurons, similar to the effects of cesium and ZD 7288,I ss was inhibited to a larger extent than I ins at voltage steps between approximately −76 and −136 mV (Fig.6 B,C). Thus, the opioid-sensitive I h currents were similar to those that were sensitive to ZD 7288 and cesium (compare Figs.5 C, 6 D).
This is also evident in Figure 7, in which the current responses that were inhibited by ZD 7288 (25 μm), cesium (3 mm), and DPDPE (1 μm) are shown (V hold = −66 mV). Here it can be seen more clearly that the current inhibited by ZD 7288 and cesium is similar in its kinetic activation and voltage dependence to that inhibited by DPDPE. Exponential time constants (see below) fit to these currents at the largest hyperpolarizing voltage steps also indicate that the time course of these inhibited currents was similar (Fig. 7, legend). This suggests that all of these agents inhibited a current possessing the same temporal characteristics asI h. In contrast, the currents that were modulated by baclofen (60 μm) are clearly different from those modulated by the inhibitors of I h (Fig.7 D).
Because a previous report has suggested that the inhibition ofI h by baclofen may be secondary toG irk activation (Watts et al., 1996) and because it was unclear whether the modulation of G irk orI h was responsible for the opioid-induced outward currents in the present study, barium was used to block the opioid activation of G irk and to permit an assessment of opioid effects on I h in relative isolation. These experiments, and those using cesium (below), were conducted in the presence of the Na+ channel blocker tetrodotoxin (TTX) to enhance the stability of the recordings and to determine whether the effects of DPDPE and DAMGO were direct. Outward currents were generated reliably by the opioids in the presence of TTX, whether administered with barium or cesium (below). Alone, bath application of barium (1–2 mm) caused a decrease in membrane conductance of 1.8 ± 0.5 nS (n = 8) (Fig. 8 B). TheE rev of the current that was inhibited by barium (measured at steady state, using 2 sec voltage steps) was −91 ± 3.2 mV, suggesting that, in agreement with previous reports, barium blocked voltage-dependent K+ channels (e.g.,I IR) (Hille, 1992), andI h was, in large part, unaffected (Halliwell and Adams, 1982; Mayer and Westbrook, 1983; Maccaferri and McBain, 1996). To evaluate the effects of barium on the opioid response, we established sensitivity to DPDPE or DAMGO; then we washed the opioid from the preparation and applied it again in the presence of barium (Fig. 8 A). Wherever possible, the effect of the opioid also was evaluated after barium was washed from the preparation (Fig. 8 A1). Barium was found to have two major effects on the opioid response. First, in every cell, barium significantly reduced the amplitude of the outward currents generated by DPDPE and DAMGO (Fig. 9 A). However, these currents were reduced only to 45.8 ± 8.8% of control (p < 0.05; paired Student’st test; n = 5) (see Figs.8 A1, 11). Second, when the effects of the opioids were evaluated by the use of I/V curves in the presence of barium, it was found that they caused a decrease in the remaining whole-cell conductance in every cell that was tested, with a mean ± SEM reduction of 24.4 ± 5.7% (n = 8) (Figs.8 D, 9 B). The reduction inI h caused by the opioids in barium was manifest as a 55.2 ± 7.2% decrease at −86 mV and a 29.3 ± 9.8% decrease at −136 mV (n = 8). In addition, the shape of the subtracted opioid-sensitive current was similar to that obtained for I h, as defined in the ZD 7288 and cesium experiments (see Fig. 5 C).
The effect of the opioid agonists on the time constant ofI h activation also was determined in a subset of opioid-sensitive interneurons. The activation ofI h was studied by evoking current responses to voltage steps (2 sec) between −86 and −136 mV from a holding potential of −66 mV. Then the inward sag associated withI h activation was fit by using exponential functions of the form: Equation 2where I t is the amplitude of the current at time t, I ss is the steady-state current measured at the end of a 2 sec voltage step,I h is I ins subtracted from I ss, and τ is the time constant of activation. The rate of I h activation demonstrated strong voltage dependence, with control time constants of 0.551 ± 0.134 sec at the −96 mV voltage step and 0.196 ± 0.036 sec at the −116 mV voltage step (n = 8). The opioid agonists significantly (p < 0.001; paired t test) slowed the activation ofI h, such that the time constants were 1.36 ± 0.122 sec at −96 mV and 0.45 ± 0.059 sec at −116 mV. Collectively, these results suggest that, whenG irk was blocked by the addition of barium, the opioids consistently decreased the whole-cell conductance and slowed the rate of I h activation. In addition, the persistence of the opioid effect on I h in the presence of TTX suggests that it resulted from the direct stimulation of opioid receptors located on the interneurons.
Opioids activate K+ channels in stratum oriens interneurons
In an effort to determine whether the activation ofG irk by the opioids was obscured by the concomitant inhibition of I h, we examined the effects of DPDPE and DAMGO on stratum oriens interneurons during the blockade of I h by extracellular cesium or ZD 7288. A protocol similar to the one described for the barium experiments was used except that, in each case, the concentration of cesium or ZD 7288 was increased until a complete block ofI h was achieved (cesium, 3–9 mm; ZD 7288, 25–100 μm). At a voltage-clamp potential of −66 mV, cesium inhibited the opioid-induced outward current in every cell that was examined (n = 6) (Fig.10 A). On average, cesium inhibited this response by 58.1 ± 10.8% (p < 0.01; paired Student’s t test) (Fig. 11), suggesting that the inhibition of I h was partly responsible for the opioid-induced change in holding current. In addition, as reported above, the application of cesium alone generated outward currents that were similar in magnitude to those caused by the opioids during barium application. This observation suggests that the inhibition of inward current associated with I h by both cesium and the opioids was partially responsible for the holding current changes. When the interneurons were treated with a combination of barium (2 mm) and cesium (3–9 mm), the opioid-induced outward currents were blocked nearly completely (14.6 ± 7.5% of control, n = 6) (Fig. 11), and there was no effect of the opioids on the I/V relationship.
As reported above, under control conditions the opioid agonists DPDPE and DAMGO caused small decreases in the whole-cell membrane conductance measured at a holding potential of −66 mV (see Figs.2 B, 10 B). However, during the cesium block of I h this decrease in conductance was converted to an increase (from 4.1 ± 1.0 nS to 5.3 ± 1.3 nS; n = 6) (see Fig. 10 B). Furthermore, this increase in conductance was associated with a current that reversed near the predicted E rev for K+ ions (measured E rev = −101 ± 6.1 mV; predicted E K = −96 mV), suggesting that in the absence of I h the opioid activation of K+ channels was evident. Similar results were obtained in three interneurons when the selective blocker of I h, ZD 7288, was used. In these cells the opioid-induced current also reversed near the predictedE K (measured E rev = −99.7 ± 9.0 mV).
A previous study has suggested that extracellular cesium can reduce the inward component of the baclofen-activated K+current observed at step potentials more negative than approximately −110 mV without affecting baclofen-induced outward currents (Sodickson and Bean, 1996). Therefore, in the present study the magnitude of the opioid-induced increase in G irk in the presence of cesium probably was underestimated. Moreover, the observation that cesium did not alter the outward currents produced by baclofen implies that the opioid-induced outward currents caused by K+ channel activation similarly were unaffected by cesium and that the cesium attenuation of the opioid-induced outward currents likely was attributable to the blockade ofI h. These findings, considered together, suggest that the opioids activate a K+ conductance and inhibit I h in the same stratum oriens interneurons and that both of these conductances contribute to the holding current changes seen at a holding potential of −66 mV.
The results of this study demonstrate that opioid agonists selective for either μ- or δ-receptors can generate outward currents in hippocampal stratum oriens interneurons. This would result in membrane hyperpolarization and the inhibition of these cells, which is consistent with the effects of opioids on neuronal populations throughout the CNS (Williams et al., 1982; Madison and Nicoll, 1988;North, 1989; Wimpey and Chavkin, 1991; Johnson and North, 1992). However, our results diverge from these previous studies in that this effect is not associated with an increase in membrane conductance and the apparent activation of K+ channels under control conditions. Instead, our results indicate that the outward currents produced by the opioid agonists in the absence of channel blockers are associated with a small decrease in conductance. One explanation for these disparate findings is that, in stratum oriens interneurons, either the opioids modulate a different current than that described in other systems or they simultaneously act at more than one ion channel. The large contribution of I h to the total membrane conductance (Maccaferri and McBain, 1996) and the previously described actions of the opioids on G irk made these currents prominent candidates for further investigation. The present results suggest that concomitant modulation of these ion channels can explain the ambiguous effects of the opioids on the whole-cell conductance and that the outward currents produced by these agonists are the result of this modulation.
The properties of I h, as it has been described in the present study, are similar to those identified in this and many different neuronal and cardiac preparations (DiFrancesco, 1981; Halliwell and Adams, 1982; Mayer and Westbrook, 1983; McCormick and Pape, 1990; BoSmith et al., 1993; Ingram and Williams, 1994; Harris and Constanti, 1995; Maccaferri and McBain, 1996; Watts et al., 1996;Gasparini and DiFrancesco, 1997) (for review, see Pape, 1996). Among these characteristics are a reversal potential indicative of a mixed Na+/K+ current, unique voltage and time dependency, and sensitivity to the blocking agents ZD 7288 and cesium. The evidence in support of the modulationI h by the opioid agonists in the present study is derived from several observations, including (1) the shape of theI/V relationship for the opioid-inhibited currents was similar to that obtained for I h, as defined by cesium and ZD 7288; (2) the opioids reduced the magnitude ofI h and the rate at which it activates; (3) the opioids had a larger effect on the steady-state I/Vrelationship, in which I h is more fully active, versus the instantaneous relationship; (4) extracellular cesium and ZD 7288 blocked I h and eliminated the opioid-induced decrease in membrane conductance. We also provide evidence that K+ channels were modulated in the same interneurons by the opioid agonists after I h was blocked by cesium or ZD 7288. Thus, the opioids clearly increased the whole-cell conductance under these conditions, and this was found to be the result of the activation of a current possessing aE rev near E K. The activation of G irk and the inhibition ofI h by the opioids are similar to the effects of adenosine on mesopontine cholinergic neurons (Rainnie et al., 1994), suggesting that this dual modulation may be common to several inhibitory neuromodulators. Collectively, our data suggest that bothI h and G irk are modulated by opioids in the same interneurons and that the simultaneous activation of G irk and the inhibition ofI h by the opioid agonists results in outward currents and little or no change in the whole-cell conductance.
In some cells the opioid modulation of I h was apparent in the absence of the blockade of G irkby barium (e.g., Fig. 6), suggesting either thatG irk was not coupled to opioid receptors in these neurons or that the coupling of the opioid receptor toI h was greater. However, in the majority of interneurons the modulation of I h was not obvious until the competing K+ channel response was blocked by barium. Similarly, the sole activation of a K+ conductance was seen in a small number of cells that had little apparent I h (e.g., Fig. 4), but rarely was seen in the absence of cesium blockade ofI h. This was despite the fact that K+ currents, presumably mediated byG irk, could be activated in these same neurons by the application of baclofen. This implies that, whereas the opioids may modulate both I h andG irk, GABAB receptor activation appears to modulate only G irk in these cells. This contrasts with previous reports in whichI h was inhibited by baclofen directly or as the indirect result of G irk activation in dopaminergic neurons (Jiang et al., 1993; Watts et al., 1996). In the present study the opioid inhibition of I h after the blockade of G irk by barium suggests that this action was direct, and not the consequence ofG irk activation. In addition, the observation that both GABAB and opioid receptors could activateG irk, whereas only the opioids inhibitedI h, suggests that these effects may occur via coupling to different G-proteins. Alternatively, both GABAB and opioid receptors may activate K+ channels via a shared set of G-proteins, but opioid receptors may use a different pool of G-proteins to modulateI h.
Enkephalins are thought to be the endogenous opioid receptor ligands in the CA1 region of the hippocampus, where they are synthesized by some interneurons (Gall et al., 1981). Furthermore, axon terminals containing leu-enkephalin have been found apposed to GABA-positive cell bodies, dendrites, and axon terminals in this brain area (Commons and Milner, 1996). Localization of opioid receptors, using antibodies raised against the cloned μ-, δ-, and κ-opioid receptors, has shown that only μ- and δ-receptors are expressed in the CA1 region of the hippocampus in appreciable concentrations (Mansour et al., 1995,1996; Commons and Milner, 1997). Moreover, both of these receptors are distributed widely in this brain area, with the δ-opioid receptor found to be associated with GAD-positive interneurons at particularly high levels in stratum oriens (Commons and Milner, 1997). The high selectivity of DAMGO and DPDPE for these receptor subtypes at the concentrations used in this study is well established (Mosberg et al., 1983; Cotton et al., 1985; Goldstein and Naidu, 1989; Lupica, 1995). This information and the observation that only 27% of the interneurons responding to either DPDPE or DAMGO also responded to the alternate agonist suggest that these peptides were selective for their respective receptor subtypes in the present study. In addition, the finding that outward currents were generated in the majority of interneurons by either DPDPE or DAMGO suggests that most stratum oriens interneurons express either μ- or δ-opioid receptors and that some of these cells express both receptor subtypes. Together, these data provide anatomical and pharmacological substrates for the previously described physiological actions of the opioids in the hippocampus, including the activation ofG irk, the activation of an outward rectifier K+ current, and the presynaptic inhibition of GABA release (Madison and Nicoll, 1988; Wimpey and Chavkin, 1991;Cohen et al., 1992; Lupica, 1995). All of these opioid actions ultimately diminish the release of GABA from these interneurons, thereby increasing the excitability of the pyramidal cells on which they impinge. The inhibition of I h by opioid receptors may provide an additional site to reduce the excitability of the interneurons and alter the activity of the hippocampal network.
Functional significance of Ih andGirk modulation by opioids
The most reliable effect of the opioid agonists was to induce outward currents in the stratum oriens interneurons. This response could be produced via the activation of a current with aE rev negative to the holding potential (e.g.,G irk) or by the inhibition of a current with an E rev positive to the holding potential (e.g., I h). The observations that barium and cesium each reduced the opioid-induced outward currents to a similar degree and that their combined application nearly completely blocked these responses suggest that both I h andG irk contributed to the holding current changes. Moreover, the observation that cesium alone could generate outward currents similar in magnitude to those caused by the opioids in the presence of barium suggests that I h is activated partially at the clamp potential of −66 mV and that the opioids may generate outward currents via the inhibition of these ion channels. Current-clamp experiments in several laboratories have indicated that the blockade of I h by ZD 7288 or cesium can result in the hyperpolarization of neurons held near the resting membrane potential (Harris and Constanti, 1995; Maccaferri and McBain, 1996; Gasparini and DiFrancesco, 1977). However, although resting membrane potentials were not recorded routinely in the present study, others have reported these values to range between approximately −52 mV and −66 mV in stratum oriens interneurons (Lacaille et al., 1987;McBain et al., 1994; Bergles et al., 1996). Furthermore, Maccaferri and McBain (1996) demonstrated in these same interneurons thatI h made a substantial contribution to the membrane conductance when the cells were held at −60 mV, which is likely close to rest. These data suggest the possibility that the inward current contributed by I h may help to set the resting membrane potentials of these cells at more depolarized levels.
In addition to its possible contribution to the membrane potential, the inhibition of I h by the opioids also would tend to make these GABAergic interneurons less responsive to excitatory inputs, particularly at hyperpolarized potentials. Also, the unusual property of the activation of this inward current on hyperpolarization implies that one of its functions might be to oppose the influences of inhibitory (i.e., hyperpolarizing) neuromodulators. Thus, the reduction in neuronal excitability caused by the activation of K+ channels and the subsequent membrane hyperpolarization initiated by the opioids and other inhibitory modulators might be reversed substantially byI h. Because of this possibility we hypothesize that the concurrent modulation of I h andG irk in the same cells may help to maintain decreased excitability by preventing I h from returning the membrane potential closer to action potential threshold. These combined actions would ensure that the inhibition of interneuron activity and GABA release initiated by K+ channel activation would not be diminished by the repolarizing influences ofI h.
This work was supported by National Institutes of Health Grant DA 07725, United States Public Health Service. We thank Drs. Thomas Dunwiddie, Kevin Staley, and Jeffrey Weiner for their remarks on this manuscript and Dr. Cathy Adams for assistance with the biocytin assays.
Correspondence should be addressed to Dr. Carl R. Lupica, Department of Pharmacology, Box C236, University of Colorado Health Sciences Center, 4200 East Ninth Avenue, Denver, CO 80262.