Selective Modulation of Excitatory Transmission bym-OpioidReceptor Activation in Rat Supraoptic Neurons

QING-SONG LIU, SHENG HAN, YOU-SHENG JIA, AND GONG JUInstitute of Neurosciences, The Fourth Military Medical University, Xian 710032, People’s Republic of China

Liu, Qing-song, Sheng Han, You-sheng Jia, and Gong Ju.Selec-tive modulation of excitatory transmission bym-opioid receptor acti-vation in rat supraoptic neurons.J. Neurophysiol.82: 3000–3005,1999. Opioid peptides have profound inhibitory effects on the pro-duction of oxytocin and vasopressin, but their direct effects on mag-nocellular neuroendocrine neurons appear to be relatively weak. Wetested whether a presynaptic mechanism is involved in this inhibition.The effects ofm-opioid receptor agonistD-Ala2, N-CH3-Phe4, Gly5-ol-enkephalin (DAGO) on excitatory and inhibitory transmission werestudied in supraoptic nucleus (SON) neurons from rat hypothalamicslices using whole cell recording. DAGO reduced the amplitude ofevoked glutamatergic excitatory postsynaptic currents (EPSCs) in adose-dependent manner. In the presence of tetrodotoxin (TTX) toblock spike activity, DAGO also reduced the frequency of spontane-ous miniature EPSCs without altering their amplitude distribution,rising time, or decaying time constant. The above effects of DAGOwere reversed by wash out, or by addition of opioid receptor antag-onist naloxone or selectivem-antagonist Cys2-Tyr3-Orn5-Pen7-NH2

(CTOP). In contrast, DAGO had no significant effect on the evokedand spontaneous miniature GABAergic inhibitory postsynaptic cur-rents (IPSCs) in most SON neurons. A direct membrane hyperpolar-ization of SON neurons was not detected in the presence of DAGO.These results indicate thatm-opioid receptor activation selectivelyinhibits excitatory activity in SON neurons via a presynapticmechanism.

I N T R O D U C T I O N

The magnocellular neurons of supraoptic nucleus (SON) andparaventricular nucleus (PVN) synthesize oxytocin and vaso-pressin and release these hormones from their terminals in theneurohypophysis. A number of studies indicate that opioidsmay be involved in controlling release of oxytocin and vaso-pressin, and that both the terminals in neurohypophysis and thesoma in hypothalamus are important sites of opioid inhibition(Arnauld et al. 1983; Bicknell and Leng 1982; Bicknell et al.1985; Renaud and Bourque 1991; Wakerley et al. 1983; Zhaoet al. 1988). The SON and PVN receive opioid innervationfrom fibers originating from other brain regions (Finley et al.1981; Sawchenko et al. 1982) and contain high levels ofm- andk-opioid receptor–binding sites (Mansour et al. 1988; Sumneret al. 1990, 1992). Both in vivo and in vitro extracellularrecordings have shown that opioids inhibit firing of a portion ofSON and PVN neurons via activation ofm-, d-, or k-receptors(Arnauld et al. 1983; Inenaga et al. 1990; Leng and Russell1989; Wakerley et al. 1983). Intracellular recordings have alsorevealed that them-agonistD-Ala2, N-CH3-Phe4, Gly5-ol-en-

kephalin (DAGO) decreases or suppresses spontaneous firingin less than one-half of the putative neurosecretory magnocel-lular neurons in SON and PVN, but this inhibition is notaccompanied by an appreciable change of the resting mem-brane potential or input resistance. In contrast, DAGO de-creases or suppresses spontaneous firing in most of the low-threshold Ca21 spike neurons (putative nonneurosecretory) inPVN; this inhibition is accompanied by a marked hyperpolar-ization (Wuarin and Dudek 1990; Wuarin et al. 1988). Thusneighboring nonneurosecretory neurons are more effectivelyinhibited by this opioid than magnocellular neurosecretoryneurons per se. These findings suggest that modulation ofsynaptic inputs may be important in opioid-induced inhibitionof the neurosecretory neurons.

We tested the hypothesis that a presynaptic mechanism isinvolved in m-opioid inhibition of magnocellular neurosecre-tory neurons. Intracellular recordings have shown that thek-agonist dynorphin suppresses evoked excitatory postsynapticpotentials (EPSPs) and inhibitory postsynaptic potentials(IPSPs) in a portion of the SON neurons, whereasd-agonistD-Ala, D-Leu enkephalin andm-agonist morphine have little orno effect (Inenaga et al. 1994). In that study, the effect ofmorphine was tested in only three neurons and in relatively lowconcentrations. Here, using whole cell recordings from hypo-thalamic slices, we examined the effects of DAGO, a highlyselectivem-opioid receptor agonist, on evoked and spontane-ous excitatory postsynaptic currents/inhibitory postsynapticcurrents (EPSCs/IPSCs) in rat SON neurons.

M E T H O D S

Preparation of brain slices and whole cell recordings were per-formed as described previously (Liu et al. 1997). Briefly, Sprague-Dawley rats (4–6 wk old) were deeply anesthetized with ketamineand chloral hydrate and decapitated, and the brains were quicklyremoved and placed in ice-cold artificial cerebrospinal fluid (ACSF)for a few minutes. The ACSF was well oxygenated with 95% O2-5%CO2 and consisted of (in mM) 125 NaCl, 5 KCl, 1.25 Na2HPO4, 2CaCl2, 1.3 MgSO4, 25 NaHCO3, and 10 glucose. The hypothalamuswas blocked out, and 350-mm slices were cut coronally using aVibratome (TPI, St. Louis, MO). After sectioning, the slices werehemisected and allowed to recover for at least 1 h in warm ACSF(30–32°C). The slice was transferred into a submerged chamberperfused with warm ACSF (30–32°C) at 1–2 ml/min. Tight-seal(2–10 GV) whole cell patch-clamp recordings (Blanton et al. 1989)from SON neurons were obtained using an Axopatch 200A amplifier(Axon Instruments). Membrane currents were amplified at 5–50 mV/pA, filtered at 1–2 kHz, and digitized at 5–10 kHz using pClampsoftware and DigiData 1200 interface (Axon Instruments). Borosili-cate glass capillaries (1.6 mm OD, 1.2 mm ID, with filament) were

The costs of publication of this article were defrayed in part by the paymentof page charges. The article must therefore be hereby marked “advertisement”in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

3000 0022-3077/99 $5.00 Copyright © 1999 The American Physiological Society

by 10.220.33.6 on October 28, 2016

http://jn.physiology.org/D

ownloaded from

pulled in two stages to a resistance of 3–5 MV. For most experimentsthe pipette solution was (in mM) 120 potassium gluconate, 5 KCl, 2MgCl2, 10 HEPES, 10 EGTA, and 2 Mg-ATP, pH 7.2 with KOH. Forrecording of spontaneous miniature IPSCs, 120 mM potassium glu-conate and 5 mM KCl were replaced by 125 mM KCl. Evoked EPSCsand IPSCs were elicited with a bipolar tungsten stimulation electrode(WPI, Sarasota, FL) placed outside the dorsolateral border of SON.Square pulses were delivered at a duration of 0.2–1 ms, intensity of0.1–1 mA, and frequency of;0.1 Hz. These parameters were ad-justed to produce evoked EPSCs and IPSCs ranging from 80 to 300pA. Series resistance was routinely compensated by 80–85% in allexperiments except recording of mEPSCs. Series resistance (,20MV) was carefully monitored during experiment, data were not in-cluded if a significant increase in series resistance ($15%) wasdetected. The junction potential was nullified before attempting toform giga-seals.

The evoked EPSCs and IPSCs were averaged, measured, andfitted using Clampfit. Data are given as means6 SE. The paired orunpairedt-test as appropriate was used to compare the differencesbetween these data. The statistical criterion for significance wasP , 0.05. Spontaneous miniature EPSCs (mEPSCs) and IPSCs(mIPSCs) were analyzed off-line with a synaptic current detectionsoftware provided by Dr. P. Vincent. The details for this event-detection procedure were described elsewhere (Vincent and Marty1993). Synaptic events were detected with an adjustable threshold,often set at 6 –10 pA and kept constant for the same group of data(control, treatment, and recovery). The analysis of mEPSCs andmIPSCs was performed with cumulative probability plots (Van derKloot 1991). The cumulative probability of amplitude and inter-event interval were compared using the Kolmogorov-Smirnov test(K-S test), which estimates the probability (P) that two distribu-tions are similar. With this test, two cumulative sets of data wereconsidered significantly different only whenP , 0.01.

Cys2-Tyr3-Orn5-Pen7-NH2 (CTOP) and (6)-2-amino-5-phospho-nopentanoic acid (AP-5) were from RBI (Natick, MA). All otherreagents were from Sigma.

R E S U L T S

Effect of DAGO on evoked EPSCs

SON neurons were voltage clamped at270 mV, andEPSCs were evoked following at least 20 min pretreatmentwith the GABAA receptor antagonist bicuculline (20mM).Focal stimulation of the dorsolateral border of the SONelicited EPSCs in a total of 38 SON neurons. In 28 of the 38neurons, only “pure” monosynaptic EPSCs were observed.In the remaining 10 neurons, the monosynaptic EPSCs werefollowed by brief, high-frequency bursts of polysynapticactivity. Both the monosynaptic EPSCs (Fig. 1A, n5 6) andthe burst of polysynaptic activity (n 5 4) were completelyand reversibly blocked by 10mM 6,7-dinitroquinoxaline-2,3-dione (DNQX), indicating that they were mediated bynon–N-methyl-D-aspartate (NMDA) glutamate receptors.The monosynaptic EPSCs had a fixed latency of,4 ms, andfast rising and slow decaying phases. The average rise timeand decay time constants were 2.16 0.1 ms and 4.86 0.6ms, respectively (n 5 28). Bath perfusion ofm-agonistDAGO (0.1–1 mM) for 5 min consistently reduced theamplitude of the evoked EPSC in each of the 22 cells tested(Fig. 1). This effect peaked at 2– 4 min after perfusion ofDAGO and was reversed by 10 –20 min wash out (Fig. 1,Band D). In the continuous presence of 0.5mM DAGO,

subsequent addition of the opioid antagonist naloxone (5mM, Fig. 1D) or selectivem-antagonist CTOP (1mM, Fig.1, C and D) reversed the effect of DAGO. DAGO had nosignificant effect on the kinetics of the evoked EPSCs. In thepresence of DAGO, the rising time of the evoked EPSCswas 976 3% of control, and the decay time constant was103 6 6% of control (P . 0.05, n 5 22).

We also tested the effects of DAGO on the burst of polysyn-aptic activity. The burst generally lasted;80 ms after terminationof the stimulation, and in shape resembled a collection of spon-taneous EPSCs at high-frequency in the same neuron (Fig. 2).Their amplitudes ranged from 6 to 10 pA (threshold for detection)to ;100 pA. Bath perfusion of DAGO (0.5mM) decreased thefrequency of the polysynaptic events from 1606 9 Hz to 366 3Hz (266 4% of control;P , 0.001;n 5 6), and this effect wasreversed on wash out (1236 11 Hz; 856 11% of control). Thefrequency was calculated as the number of events from 0 to 80 msafter the stimulation. In these neurons, DAGO also reduced theamplitude of monosynaptic EPSCs, but it was not always possibleto measure the amplitude accurately because the burst of polysyn-aptic events was superimposed on the monosynaptic EPSCs.Therefore these data were not included for amplitude analysis.DAGO had no effect on the holding currents at the holdingpotential of270 mV.

FIG. 1. Effects of m-agonist D-Ala2, N-CH3-Phe4, Gly5-ol-enkephalin(DAGO) on evoked monosynaptic excitatory postsynaptic currents (EPSCs)recorded from supraoptic nucleus (SON) neurons.A: non–N-methyl-D-aspar-tate (NMDA) glutamate receptor antagonist 6,7-dinitroquinoxaline-2,3-dione(DNQX; 10 mM) reversibly blocked evoked monosynaptic EPSCs. Tracemarked “wash” was taken after 20 min wash out of the DNQX. In this andsubsequent traces, stimulation artifacts are blanked out.B andC: DAGO (0.5mM) reduced the amplitude of evoked EPSCs. This inhibition was reversed by10 min wash out of DAGO (B) or by subsequent addition ofm-antagonistCys2-Tyr3-Orn5-Pen7-NH2 (CTOP; 1mM, C). Each trace was obtained by dataaveraged from 10 consecutive stimuli.D: mean amplitude was expressed as apercent of that obtained in control. Each bar represents the means6 SE of5–22 cells; the numbers inside the bars are numbers of cells tested. *P , 0.05,**P , 0.001.

3001PRESYNAPTIC INHIBITION OF SON NEURONS BYm-OPIOID

by 10.220.33.6 on October 28, 2016

http://jn.physiology.org/D

ownloaded from

Effect of DAGO on spontaneous mEPSCs

To investigate whether the effect of DAGO on evokedEPSCs is mediated by a presynaptic mechanism, its action onmEPSCs was analyzed. After pretreatment of slices with 20mM bicuculline and 1mM TTX for at least 20 min, mEPSCswere recorded in SON neurons voltaged clamped at270 mV.In agreement with a previous report (Wuarin and Dudek 1993),mEPSCs are mediated by non-NMDA glutamate receptors,because they were blocked by DNQX (10mM; n 5 5). Thefrequency of mEPSCs ranged from 0.5 to 15 Hz, and theamplitude ranged from 6 to 10 pA (the threshold for detection)to ;100 pA. Seven cells with relative high-frequency ofmEPSCs were tested with DAGO. DAGO (0.5mM) decreasedmEPSC frequency from 6.26 1.8 Hz to 3.16 1.2 Hz (4765% of control; pairedt-test,P , 0.001;n 5 7). Subsequentaddition of the selectivem-opioid receptor antagonist CTOP (1mM) in the continuous presence of DAGO reversed the effectof DAGO on mIPSCs (5.76. 2.1 Hz, Fig. 3A). Cumulativefrequency plot analysis showed that DAGO did not change thecumulative amplitude distributions (K-S test,P . 0.01; Fig.3C) but shifted the cumulative distribution of intervals betweensuccessive events to the right in each of seven cells tested (K-S

test, P , 0.001; Fig. 3B). DAGO also had no effect on thekinetics of mEPSCs (Fig. 3D).

Effect of DAGO on evoked IPSCs and mIPSCs

Next we tested whether IPSCs could also be modulated bym-receptor activation. After 20 min perfusion of slices withglutamate receptor antagonists DNQX (10mM) and AP-5 (100mM), focal stimulation evoked fixed-latency (,4 ms) mono-synaptic IPSCs, which were completely and reversibly blockedby 20mM bicuculline (Fig. 4A, n5 7). Generally IPSCs couldbe evoked by lower stimulating intensities than those requiredfor EPSCs, and they were not followed by a burst of polysyn-aptic activity. The evoked IPSCs reversed at about270 mVand appeared outward at more depolarized holding potentials.In the present experiment, the neurons were voltage clamped at240 mV to record evoked IPSCs. The average rise time anddecay time constants of the evoked IPSCs were 2.26 0.3 msand 20.06 4.2 ms, respectively (n5 15). Bath perfusion ofm-agonist DAGO (0.5–1mM) for 5 min had no significanteffect on the evoked IPSCs (Fig. 4B). In the presence ofDAGO, the amplitude of evoked IPSCs was 98.26 4.2% ofcontrol, the rise time was 106.16 5.6% of control, and the

FIG. 2. Effects of m-agonist DAGO onevoked polysynaptic activity.A: DAGO (0.5mM) decreased the frequency of polysynap-tic activity. This effect was reversed by 10min wash out. Arrows indicate the time atwhich stimuli were given.B: traces are thesame as the lowest traces inA but withdifferent vertical scale. Both frequency ofpolysynaptic events and amplitude of mono-synaptic EPSCs were decreased by DAGO.C: summary of the effect of DAGO. Each barrepresents the means6 SE of 6 cells. **P ,0.001. The total number of synaptic eventsfrom 0 to 80 ms after termination of thestimuli was used to calculate the frequency.

FIG. 3. m-Agonist DAGO affected thefrequency, but not the amplitude or kineticsof mEPSCs.A: bath application of DAGO(0.5 mM) decreased the frequency ofmEPSCs from 13.7 to 7.3 Hz. This effectwas reversed by subsequent addition of se-lective m-antagonist CTOP (1mM). B:DAGO shifted the cumulative frequencydistribution to the right, i.e., toward longerinterevent intervals (K-S test,P , 0.001).C: DAGO had no significant effect on thecumulative amplitude distributions.D: su-perimposed averages of 200 consecutivemEPSCs obtained during control andDAGO treatment. Dashed line represents amonoexponential fit to the control average.Note the similar amplitudes of both aver-ages (28.5 vs. 28.7 pA), and the similardecay constants (t 5 1.05 ms).

3002 Q.-S. LIU, S. HAN, Y.-S. JIA, AND G. JU

by 10.220.33.6 on October 28, 2016

http://jn.physiology.org/D

ownloaded from

decay time constant was 104.36 5.7% of control respectively(n 5 8). These values were not statistically different fromcontrol (P. 0.05).

We also tested the effect of DAGO on spontaneous mIP-SCs, recorded in the presence of glutamate receptor antag-onists and also 1mM TTX to block spike activity. At aholding potential of270 mV, mIPSCs appeared inwardwhen KCl-based pipette solution was used for whole cellrecording (Fig. 4C). They were also reversibly blocked bybicuculline (10 mM) and thus mediated by activation ofGABAA receptors. Interestingly, TTX, applied at a concen-tration that completely blocked the action potential or so-dium current, had no significant effect on the spontaneousIPSCs recorded from the SON neurons, concurring withrecent data (Brussaard et al. 1996; Kabashima et al. 1997)and suggesting that the spontaneous IPSCs in SON areproduced by random release of single vesicles from axonterminals, which are independent of somatodendritic actionpotentials (Brussaard et al. 1996). Overall, after TTX treat-ment, the frequency was 95.36 5.6% of that control (n 57, pairedt-test, P . 0.05); the amplitude distribution wasalso not altered as determined by cumulative plot and K-Stest (P . 0.01). In the continuous presence of TTX, bathperfusion of DAGO (0.5mM) for 5 min had no significanteffect on the mIPSCs in five of seven cells tested (Fig. 4C).The mean frequency of mIPSCs before and after DAGOperfusion was 5.261.6 Hz and 4.96 1.5 Hz, respectively(n 5 5, pairedt-test,P . 0.05); the cumulative amplitudeand interevent interval distribution were not significantly

changed (K-S test,P . 0.01). In the remaining two cells,DAGO perfusion significantly decreased the mIPSC fre-quency by 35.3 and 43.6%, respectively, which was alsoaccompanied by small but significant left shift (towardsmaller values) of cumulative amplitude distribution. How-ever, in contrast to the effect of DAGO on mEPSCs whichwas reversible, the inhibition of mIPSCs by DAGO showedpartial recovery on wash out of DAGO in only one of thetwo cells. This effect could be due to either rundown ofGABA responsiveness or a specific action of DAGO on aminority of presynaptic terminals. The functional signifi-cance is not clear. In each of these neurons, DAGO had nosignificant effect on the holding current at the holdingpotential of240 and270 mV.

D I S C U S S I O N

Our data indicate that them-agonist DAGO selectively in-hibits excitatory transmission in SON in a receptor-mediated,presynaptic manner. DAGO is a highly selectivem-agonist,and its effect is reversible by addition of selectivem-antagonistCTOP and nonselective opioid receptor antagonist naloxone.Therefore activation ofm-opioid receptors accounts for theeffects of DAGO. To determine whether a presynaptic mech-anism is involved in this inhibition, we have analyzed theeffect of DAGO on spontaneous mEPSCs, recorded in thepresence of TTX to block spike activity. DAGO reduces thefrequency of mEPSCs without altering their amplitude distri-bution or kinetics. This pattern of a decrease in the frequencyof miniature synaptic events without a significant change inother properties is typical of a presynaptic effect (Bekkers andStevens 1990; Del Castillo and Katz 1954; Van der Kloot1991).

As reported previously (Dudek and Gribkoff 1987), focalstimulation produced brief, high-frequency bursts of polysyn-aptic activity in SON neurons following the monosynapticEPSP. This polysynaptic activity may result from repetitivefiring of presynaptic neurons within or near the stimulationsites, which lie outside of the SON. Although these polysyn-aptic events are present in the absence of bicuculline (Dudekand Gribkoff 1987), the presence of bicuculline in the presentstudy removes GABA-mediated tonic inhibition and promotesfiring of presynaptic neurons. DAGO-induced reduction of thefrequency of these polysynaptic events can be explained by itsability to reduce firing from presynaptic neurons. This is inagreement with the idea that DAGO inhibits firing of neigh-boring nonneurosecretory neurons and thus reduces their exci-tatory input to SON neurons.

As demonstrated in locus coeruleus and hippocampal neurons,DAGO could affect the excitability postsynaptically by hyperpo-larizing the neuronal membrane via opening of K1 channels(North and Williams 1985; Svoboda and Lupica 1998). However,we have not detected any outward currents from SON neurons byperfusion of DAGO, suggesting the lack of direct membranehyperpolarization. If DAGO had hyperpolarized the SON neu-rons, it should have produced an outward current at holdingpotentials used in the present experiment (i.e.,240 and270 mV).This result is in agreement with previous intracellular work show-ing that DAGO has no significant effect on the membrane poten-tial of SON neurons (Wuarin and Dudek 1990). Together, these

FIG. 4. m-Agonist DAGO did not affect the evoked monosynaptic inhibi-tory postsynaptic currents (IPSCs) and miniature IPSCs (mIPSCs).A: GABAA

receptor antagonist bicuculline (20mM) reversibly blocked evoked IPSCs.Trace marked “wash” was taken after 15 min wash out of the bicuculline.B:bath application of DAGO (1mM) for 5 min had no significant effect onevoked IPSCs. InA andB, each trace was obtained by data averaged from 5–10consecutive stimuli.C: mIPSCs were recorded from the same cell before (left)and 5 min after 0.5mM DAGO perfusion (right).

3003PRESYNAPTIC INHIBITION OF SON NEURONS BYm-OPIOID

by 10.220.33.6 on October 28, 2016

http://jn.physiology.org/D

ownloaded from

results suggest thatm-receptor activation inhibits SON neuronsprimarily via a presynaptic mechanism.

DAGO selectively inhibits EPSCs without affectingIPSCs in most of the SON neurons tested. Selective modu-lation of synaptic transmission bym-agonists has been well-documented in hippocampal neurons wherem-agonists re-duce GABA release from interneurons, and therefore excitepyramidal neurons through a mechanism of disinhibition(Cohen et al. 1992; Copogna et al. 1993; Lupica 1995). Inaddition, m-agonists do not affect EPSPs (Copogna et al.1993) or increase the amplitude of evoked EPSPs via thereduction of GABAergic inhibition (Lupica et al. 1992).Although the common feature ofm-opioid’s acting on syn-apse transmission is a reduction of transmitter release, theyprobably act selectively on either the excitatory or inhibi-tory input, but not both. This specificity suggests that thefunctional expression of opioid receptors is often limited toeither the excitatory or inhibitory terminals depending onthe neuronal cell type. Obviously this specificity may haveimportant physiological implications.

The firing frequency and pattern of magnocellular neurose-cretory neurons can change dramatically in response to certainphysiological and pathological stimuli. For example, during themilk ejection reflex and parturition, oxytocin neurons generateintermittent bursts of action potentials. The action potentialsare synchronized among all oxytocin neurons, resulting in aperiodic release of hormone (Poulain and Wakerley 1982).These temporal patterns are well correlated with the function ofthe hormone. Thus magnocellular neurosecretory neurons arestrongly dependent on synaptic inputs to generate their output;modulation of the synaptic inputs may have great impact onfiring patterns and hormone secretion. Thereforem-receptoractivation, by selectively reducing excitatory synapse inputs,will inhibit firing of SON neurons and subsequent hormonerelease, particularly in response to certain physiological andpathological stimuli. This indirect effect could be more rele-vant for inhibition of SON neurons than its rather weak actionon the membrane potential.

We thank Drs. Darwin K. Berg and Jian-Tian Qiao for comments on themanuscript and Dr. P. Vincent for providing a program to analyze mEPSCs.

This work was supported by a grant from the National Natural ScienceFoundation of China.

Present address of Y.-S. Jia: Dept. of Neurobiology and Behavior, Univer-sity of California, Irvine, CA 92697.

Present address and address for reprint requests: Q.-S. Liu, Dept. of Biology0357, University of California, San Diego, 9500 Gilman Dr., La Jolla, CA92093.

Received 5 April 1999; accepted in final form 13 August 1999.

REFERENCES

ARNAULD, E., CIRINO, M., LAYTON, B. S., AND RENAUD, L. P. Contrastingactions of amino acids, acetylcholine, noradrenline and leucine enkephalinon the excitability of supraoptic vasopressin-secreting neurons.Neuroendo-crinology 36: 187–196, 1983.

BEKKERS, J. M. AND STEVENS, C. F. Presynaptic mechanism for long-termpotentiation in the hippocampus.Nature346: 724–729, 1990.

BICKNELL, R. J.AND LENG, G. Endogenous opiates regulate oxytocin but notvasopressin secretion from the neurohypophysis.Nature 298: 161–162,1982.

BICKNELL, R. J., CHAPMAN, C., AND LENG, G. Effects of opioid agonists andantagonists on oxytocin and vasopressin release in vitro.Neuroendocrinol-ogy 41: 142–148, 1985.

BLANTON, M. G., LO TURCO, J. J.,AND KRIEGSTEIN, A. R. Whole-cell recordingfrom neurons in slices of reptilian and mammalian cerebral cortex.J. Neu-rosci. Methods30: 203–210, 1989.

BRUSSAARD, A. B., KITS, K. S., AND DE VLIEGER, T. A. Postsynaptic mecha-nism of depression of GABAergic synapses by oxytocin in the supraopticnucleus of immature rat.J. Physiol. (Lond.)497: 495–507, 1996.

CAPOGNA, M., GAHWILER, B. H., AND THOMPSON, S. M. Mechanism ofm-opi-oid receptor–mediated presynaptic inhibition in the rat hippocampus in vitro.J. Physiol. (Lond.)470: 539–558, 1993.

COHEN, G. A., DOZE, V. A., AND MADISON, D. V. Opioid inhibition of GABArelease from presynaptic terminals of rat hippocampal interneurons.Neuron9: 325–335, 1992.

DEL CASTILLO, J.AND KATZ, B. Quantal components of the end-plate potential.J. Physiol. (Lond.)124: 560–573, 1954.

DUDEK, F. E.AND GRIBKOFF, V. K. Synaptic activation of slow depolarizationin rat supraoptic nucleus neurones in vitro.J. Physiol. (Lond.)387: 273–296,1987.

FINLEY, J.C.W., MANDERDRUT, J. L., AND PETRUSZ, P. The immunocytochem-ical localization of enkephalin in the central nervous system of the rat.J. Comp. Neurol.198: 541–565, 1981.

INENAGA, K., IMURA, H., YANAIHARAT , N., AND YAMASHITA , H. Kappa-selective opioid receptor agonists leumorphin and dynorphin inhibitsupraoptic neurons in rat hypothalamic slices.J. Neuroendocrinol.2:389 –395, 1990.

INENAGA, K., NAGATOMO, K., YANAIHARA , N., AND YAMASHITA , H. Kappa-selective agonists decrease postsynaptic potentials and calcium componentsof action potentials in the supraoptic nucleus of rat hypothalamus in vitro.Neuroscience58: 331–340, 1994.

KABASHIMA , N., SHIBUYA , I., IBRAHIM, N., UETA, Y., AND YAMASHITA , H.Inhibition of spontaneous EPSCs and IPSCs by presynaptic GABAB recep-tors on rat supraoptic magnocellular neurons.J. Physiol. (Lond.)504:113–126, 1997.

LENG, G. AND RUSSELL, J. A. Opioids, oxytocin and parturition. In:BrainOpioid System in Reproduction, edited by R. G. Dyer and R. J. Bicknell.Oxford, UK: Oxford, 1989, p. 231–256.

LIU, Q. S., JIA, Y. S., XIE, Z. P., AND JU, G. Whole-cell recordings of thesupraoptic nucleus neurons from rat hypothalamic slices in vitro.ActaPhysiol. Sinica49: 467–470, 1997.

LUPICA, C. R.d andm enkephalins inhibit spontaneous GABA-mediated IPSCsvia a cyclic AMP-independent mechanism in the rat hippocampus.J. Neu-rosci. 15: 737–749, 1995.

LUPICA, C. R., PROCTOR, W. R., AND DUNWIDDIE, T. V. Dissociation ofm andd opioid receptor–mediated reductions in evoked and spontaneous synapticinhibition in the rat hippocampus in vitro.Brain Res.593: 226–238, 1992.

MANSOUR, A., KHACHATURIAN, H., LEWIS, M. E., AKIL , H., AND WATSON, S. J.Anatomy of CNS opioid receptors.Trends Neurosci.11: 308–314, 1988.

NORTH, R. A. AND WILLIAMS , J. T. On the potassium conductance increased byopioids in rat locus coeruleus neurones.J. Physiol. (Lond.)364: 265–280,1985.

POULAIN, D. A. AND WAKERLEY, J. B. Electrophysiology of hypothalamicmagnocellular neurons secreting oxytocin and vasopressin.Neuroscience7:773–808, 1982.

RENAUD, L. P.AND BOURQUE, C. W. Neurophysiology and neuropharmacologyof hypothalamic magnocellular neurons secreting vasopressin and oxytocin.Prog. Neurobiol.36: 131–169, 1991.

SAWCHENKO, P. E., SWANSON, L. W., AND JOSEPH, S. A. The distribution andcells of origin of ACTH (1–39)-stained varicosities in the paraventricularand supraoptic nuclei.Brain Res.232: 365–374, 1982.

SUMNER, B.E.H., COOMBES, J. E., PUMFORD, K. M. AND RUSSELL, J. A. Opioidsubtypes in the supraoptic nucleus and posterior pituitary gland of mor-phine-tolerant rats.Neurosci.37: 635–645, 1990.

SUMNER, B.E.H., DOUGLAS, A. J., AND RUSSELL, J. A. Pregnancy alters thedensity of opioid binding sites in the supraoptic nucleus and posteriorpituitary gland of rats.Neurosci. Lett.137: 216–220, 1992.

SVOBODA, K. R. AND LUPICA, C. R. Opioid inhibition of hippocampal inter-neurons via modulation of potassium and hyperpolarization-activated cation(Ih) currents.J. Neurosci.18: 7084–7098, 1998.

VAN DER KLOOT, W. The regulation of quantal size.Prog. Neurobiol.36:93–130, 1991.

VINCENT, P. AND MARTY, A. Neighboring cerebellar Purkinje cells communi-cate via retrograde inhibition of common presynaptic interneurons.Neuron11: 885–893, 1993.

3004 Q.-S. LIU, S. HAN, Y.-S. JIA, AND G. JU

by 10.220.33.6 on October 28, 2016

http://jn.physiology.org/D

ownloaded from

WAKERLEY, J. B., NOBEL, R., AND CLARKE, G. Effects of morphine andD-ala,D-leu enkephalin on the electrical activity of supraoptic neurosecretory cellsin vitro. Neuroscience10: 73–81, 1983.

WUARIN, J. P., DUBOIS-DAUPHIN, M., RAGGENBASS, M., AND DREIFUSS, J. J.Effect of opioid peptides on the paraventricular nucleus of the guinea pighypothalamus is mediated bym-type receptors.Brain Res.445: 289–296,1988.

WUARIN, J. P.AND DUDEK, F. E. Direct effects of an opioid peptide selectivefor m-receptors, intracellular recordings in the paraventricular and supraop-tic nuclei of the guinea-pig.Neuroscience36: 291–298, 1990.

WUARIN, J. P. AND DUDEK, F. E. Excitatory amino acid antagonists inhibitsynaptic responses in the guinea pig hypothalamic paraventricular nucleus.J. Neurophysiol.65: 946–951, 1991.

WUARIN, J. PAND DUDEK, F. E. Patch-clamp analysis of spontaneous synapticcurrents in supraoptic neuroendocrine cells of the rat hypothalamus.J. Neu-rosci. 13: 2323–2331, 1993.

ZHAO, B. G., CHAPMAN, C., BROWN, D., AND BICKNELL, R. J. Opioid-norad-renergic interactions in the neurohypophysis. II. Does noradrenline mediatethe actions of endogenous opioids on oxytocin and vasopressin release?Neuroendocrinology48: 25–31, 1988.

3005PRESYNAPTIC INHIBITION OF SON NEURONS BYm-OPIOID

by 10.220.33.6 on October 28, 2016

http://jn.physiology.org/D

ownloaded from