~f~r[),s~~~~~~ Vol. 37. No. 1, pp. 45-53, Printed in Great Britain

03~S22/~ $3.00 + 0.00 Pqamon Presspk c, 1990tBR0

1990

STIMULATION OF ENDO~ENOUS OPIOID RELEASE DISPLACES MU RECEPTOR BINDING IN RAT HIPPOCAMPUS J. J. WAGNER, R. M. CAUDLE, J. F. NEUMAIERand C. CHAVKIN Department of Pharmacology, SJ-30, Un~ve~ity of Washington School of Medicine, Seattle, Washington 98195, U.S.A. Abstract-Physiological release of endogenous opioids in the rat hippocampus was detected by an in vitro radioligand displacement assay using [‘H][p-Ala2,N-methyl-Phe4,glyols]enkepha1m ([)H]DAGO), a mu selective opioid agonist. In this assay, radioligand binding to opioid receptors in the in vitro hippocampal slice was reduced by competition with endogenous opioids released following tissue depolarization. Veratridine-indu~d opioid release caused displacement of [-‘HIDAGO that could be blocked by either tetrodotoxin addition or calcium removal from the incubation buffer. Maximal displa~ment of [-?H]DAGO also required the presence of peptidase inhibitors in the incubation buffer. None of the buffer composition changes directly affected 13H]DAG0 binding to rat brain membranes. Calcium-dependent displacement of (3H]DAG0 binding from mu receptor sites elicited by focal electrical stimulation depended on the intensity and frequency of stimulation and positioning of the electrode in the slice. Maximal displacement of [‘HIDAGO binding was observed following high intensity (I50-300 /.4A), high frequency (IO-50 Hz) stimulation of the perforant path, a major afferent fiber system to the hippocampus previously shown to contain pr~nkephalin-delved opioids. Low frequency stimulation (0.1-I Hz) was ineffective. Stimulation of the mossy fibers (containing both dynorphins and enkephalinsf also significantly reduced mu receptor binding, but to a lesser extent. Electrical stimulation of the hippocampal slice at sites not containing opioid peptides did not cause mu receptor displacement. These results demonstrate that under physiological conditions, the release of endogenous opioids from the major opioid containing pathways can be detected in a single hippocampal slice following high frequency stimulation

progress has been made in identifying the different endogenous opioid peptides, the types and specificities of opioid receptors, and the pharmacological effects of opioid receptor activation.‘.‘0.34 These results have strongly suggested that endogenous opioids act as inhibitory transmitters in the nervous system; however, the specific physiologic functions of opioids are still unclear. Understanding the actions of endogenous opioid peptides in the control of neuronal circuitry requires further characterization of the physiological conditions regulating opioid peptide release, the location and identity of the cellular targets, and the specific opioid receptor types mediating their subsequent effects. One likely site of endogenous opioid action in the nervous system is the hippocampus, a brain structure particularly suitable for in vitro analysis. In the hippocampus two major opioid containing pathways have been identified by immunocytochemical techniques: the lateral perforant path containing proenkephalin-derived opiojdsl8.19.31 and the mossy fibers containing prodynorphin and a lesser amount of proenkephalinderived opioids. 303’Electrophysiological studies have shown that exogenous opioids regulate neuronal excitabifity in the hippocampus by inhibiting a population of GABA containing interneurons.27.43 The Considerable

Abbreviu~ion: [)H]DAGO, [“H][o-Alar, N-methyl-Phe4-glyoP]-

enkephalin. 45

presence of specific opioid receptors in this region has been demonstrated by receptor autoradiography under conditions selective for either mu, delta or kappa binding sites and has revealed potential sites of opioid action within the hipp~mpus.‘4.32 Pha~acologicai actions of receptor-selective opioids at these sites in the hippocampus have also been identified.‘O However, the disparity between the distribution of opioid peptide-containing terminals and opioid receptors in the hippocampus makes the identification of the specific cellular targets of the endogenous opioids uncertain. Characterization of the neurotransmitter-like properties of the opioids in the hippocampus has shown that the enkephalin and dynorphin opioid peptides can be released by pharmacological depolarization of in uitro hippocampal tissue in a calciumdependent manner, as measured by high ~rfo~ance liquid chromatography resolution and specific radioimmunoassay.6,8 These results indicate that endogenous opioids are likely to act as neurotransmitters in the hippocampus, but previous attempts to measure synaptic conductances evoked by release of endogenous opioids in h~pp~ampus have been inconclusive,9.‘6.25,3gpresumably because the optimal stimulation and recording sites were not known. Further advance in the characterization of endogenous opioid peptide action requires a better understanding of how to stimulate opioid release and where

J. J. WAGNERet al.

46

in the tissue the endogenous opioids act. The goal of this study was to define the physiological stimulation

parameters required for release of endogenous opioid peptides. We have adapted a radioligand displacement assay originally used to study opioid action in tiiz~~~.~*to measure endogenous opioid release in the in vitro hippocampal slice under physiological conditions. The results in this paper extend our previous observations of opioid radioligand displacement caused by veratridine or potassium induced depolarization.3’ EXPERIMENTAL Slice binding

PROCEDURES

experiments

Freshly dissected, transverse hippocampal slices (500 pm) cut parallel to the visible alvear fibers, were divided into treatment groups. Slices were incubated, two slices per well, at 34°C in a 24-well tissue culture plate (Falcon) containing continuously oxygenated Krebs bicarbonate buffer containing (in mM): NaCl 124, KC1 4.9, KH,PO, 1.2, MgCl, 2.4, CaClz 2.5, glucose 10, NaHCO, 25.6, saturated with 95% O,-5% CO,. Calcium-free Krebs bicarbonate buffer contained 5 mM MgCl, and no added CaCl, After 15 min preincubation, peptidase inhibitors dissolved in buffer were added yielding the following concentrations: bestatin (20pM, Sigma), thiorphan (0.3/1M, Sigma), captopril (IOpM, Squibb) and kelatorphan (20pM, kindly provided by Dr B. Roques). Following an additional 15-min equilibration period, depolarization began with the addition of either 3pM veratridine (Sigma) or focal electrical stimulation. Electrical stimulation was delivered through a narrow (IOOpm) concentric, bipolar electrode (SNE-100, Rhodes Medical Supply) connected to a Grass stimulus isolation unit (PSIU6) controlled bv a Grass SIIA stimulator. Stimulating electrodes attached to a micromanipulator (Brinkmann MM33) were manually positioned with the aid of a dissection microscope (Leitz-Wild) at 16x. Ten minutes following the initiation of depolarization, [‘H][o-Ala*, N-methyl-Phe’-glyol’)enkephalin ([)H]DAGO; 33-41 Ci/mmol, Amersham), a selective mu receptor agonist,*‘.*’ was added to each well to yield a final concentration of 10 nM. Non-specific binding was determined in parallel wells in the presence of IOpM naloxone (Sigma). Each incubation well contained a final volume of 500~1. After incubation with [)H]DAGO for 15 min, each slice was homogenized by removal from the incubation well through a 26-gauge needle into a l-ml syringe and then ejected into GF/C filters (Whatman) presoaked in distilled water saturated with isoamyl alcohol. Filters were washed by vacuum three times with 4 ml of ice-cold buffer. The harvest and wash time was less than 20s for each slice. Bound [‘HIDAGO was quantified by liquid scintillation counting [Ecolite ( + ), ICN]; each slice binding assay was performed with two to six replicate slices per test group. For both veratridine and electrical stimulation, the total depolarization period was 25 min; shorter periods were not tested. Preliminary results showed that 15 min incubation with [)H]DAGO was the minimum required for adequate specific binding in the slices and that optimal displacement required that d&olarization begin before radioligand addition.‘” Data were exoressed as specific f)HlDAGO binding per slice as simultaneous determmation of the protein content of each slice was not performed. However, in separate determinations usina a modified Bradford method,s the protein content of typical slices was found to vary minimally (0.74 k 0.05 mg/slice, n = IO slices.33For ail comparisons described, statistical analyses were carried out using ANOVA followed by a Neuman-Keuls test or paired r-test where appropriate; P values less than 0.05 were considered significant.

Auroradiographj, For autoradiography, freshly dissected rat hippocampi were quickly frozen in liquid freon and cryostat sectioned (2011m) in the transverse plane. Sections were thawmounted onto gelatin-coated glass slides, refrozen, and stored at -20°C overnight to allow section adherence to slides. Slides were then hydrated in 50 mM Tris (pH 7.4) for 60 min and then incubated in 50 mM Tris containing 3 nM (‘HIDAGO for 60min. The slides were washed twice for IO min in ice-cold Tris, rinsed briefly in ice-cold distilled water, and dried under a stream of cold air. The sections and tritium standards (American Radiochemicals O-489 pCi/g) were exposed to tritium-sensitive film (LKB Ultrofilm) for 10 weeks. Non-specific binding was determined in the presence of IO PM naloxone. Under these conditions non-specific binding was evenly distributed throughout the sections and was not significantly different from film background. Images from the sections were digitized using MCID software developed by Dr P. Ramm.-” The digitized images from specific hippocampal regions were outlined and the image density was measured. Background was subtracted and the average density for each lamina was calibrated to the tritium standards.

RESULTS

The rat hippocampus is an elongated structure approximately 1 cm long. Transverse slices can be prepared from the full extent of the structure, and multiple slices from each hippocampus are typically used for physiological and neurochemical assays. To determine how the mu opioid receptors were distributed along the longitudinal axis of the rat hippocampus, we prepared sequential sections from the septal (dorsal) to temporal (ventral) poles of this structure. Photographs of representative images show that the distribution of [3H]DAG0 binding to hippocampal slices was non-uniform (Fig. 1). As previously reported,‘4.32 mu receptor binding was concentrated in the stratum pyramidale of CAl-CA3, in the stratum granulare, and in the stratum moleculare of the dentate gyrus and CA3 of each tissue section. However, mu binding density formed an increasing gradient from the septal to the temporal pole. The difference in binding densities was particularly striking in the stratum lacunosum-moleculare of the CA3. An additional region containing high specific mu binding abruptly appeared in the border between the subiculum and the entorhinal cortex approximately 7 mm from the septal pole (see Fig. 5A for a description of the slice anatomy). The identity of this hippocampal area is not clear because the plane of section does not permit precise comparison with the available brain atlases. However, this region of high density of specific binding (approximately 60Ofmol/mg) that was evident between subiculum and entorhinal cortex at 6.5-IOmm from the septal pole, coincides with the anatomical appearance of the presubiculum at this level in the hippocampus.” The regional differences in binding were quantified using computer-assisted densitometry (Fig. 2). Specific binding of [3H]DAG0 gradually increased along the longitudinal axis of the hippocampus from the septal to temporal pole. In contrast, the density

Septal 6 mm

I mm

7 mm 2 mm

8 mm 3 mm

9 mm

10 mm 5 mm

Temporal Fig. 1. Autoradiographic characterization of the longitudinal distribution of [3H]DAG0 binding from septal to temporal poles. Hippocampi from male Sprague-Dawley rats (250-300 g, n = 3) were removed and cryostat-sectioned (20 pm) in the transverse plane, then sections were mounted on gelatin-coated slides. Slides were incubated in 50 mM Tris containing 3 nM [‘H]DAGO (4WiOCi/mmol) for 60min at room temperature. The slides were then rinsed, dried, and then exposed to tritium-sensitive film for 10 weeks. Non-specific binding was determined in the presence of 1OpM naloxone. Under these conditions, non-specific binding was evenly distributed throughout the sections and was not significantly different from film background. Images from the sections were digitized using MCID software, I7 background was subtracted and the resulting computer image photographed. Photographs are representative images of sections at approximately I-mm intervals from the septal to the temporal pole of a single, representative hippocnmpus (images from two other hippocampi gave simihtr binding gradients). 41

3x

J.

J. WAGNER

of non-specific [‘HIDAGO binding did not significantly change throughout the septotemporal axis (data not shown). The lacunosum-moleculare of CA3

showed the greatest change in binding site density along the longitudinal axis from septal to temporal pole. The differences in [‘HIDAGO binding along the longitudinal axis shown by receptor autoradiography were confirmed in freshly dissected slices by comparing the binding in slices taken from the septal, mid septotemporal and temporal thirds of the hippocampus. Live slices (500 pm) were incubated in oxygenated physiological Krebs bicarbonate buffer at 34-C for I h with 3 nM [-‘HIDAGO as described above. Non-specific binding was determined in the presence of IO PM naloxone. At the end of the incubation period, slices were homogenized, filtered, and counted. The specific [“HJDAGO binding found was: septal, 3.3 F 0.8; mid septotemporal, 6.6 + 0.9; temporal, 6.0 i: 0.8 (fmol/slice + S.E.M.; II = 6 independent experiments). The specific binding did not differ significantly in the mid septotemporal and temporal thirds, but both had significantly greater binding than the septal third (P < 0.05). The regional differences in binding resulted from differences in both slice size and receptor distribution. Higher concentrations of endogenous opioids have also been noted in the temporal half of the hippocampus.‘* These differences may have important implications for the functioning of opioids in the different regions of the hippocampus, but to limit the experimental variability, slices of the mid septotemporal hippocampus were used in the in vitro binding assays described below. Veratridine (3 PM)-induced depolarization of in vitro hippocampal slices resulted in displacement of specific [3H]DAG0 binding (Table I), just as it caused the displacement of [‘Hjdiprenorphine in a

tv td.

Table I. Veratridine-induced from

slices

was

displacement of [‘HIDAGO

peptidase-sensitive tetrodotoxin

and

blocked

~HIDAGO specific binding (fmol/slice)

Test condition Unstimulated control Veratridine (3 PM) Veratridine (3 p M) minus peptidase inhibitors Veratridine (3 p M) with tetrodotoxin (1 PM)

7.2 rf: 0.4 5.0 + 0.8* 7.0 + 0.9

%

n

100 70 98

8 7 4

96

4

6.9 + 0.6

Assays were performed as described in Experimental Procedures; veratridine (3 p M) was added 10 min before the 15-min [lH]DAGO (10 nM) incubation was begun. Values are the means +S.E.M. of (n) independent experiments. Non-specific [‘HIDAGO binding as determined in the presence of naloxone (IOpM) was subtracted from total binding in each test condition to give specific binding. Statistical comparisons: *P < 0.05 (Neuman-Keuls) between the indicated condition and each of the other three test conditions.

previous study.33 The effect of veratridine on [‘HIDAGO binding was blocked by the addition of tetrodotoxin (1 PM). The antagonism by tetrodotoxin indicated that the veratridine effect on [‘HIDAGO binding was specifically mediated by activation of sodium channels rather than a direct effect on opioid receptors. Supporting this interpretation, neither 3 PM veratridine nor 1 PM tetrodotoxin had significant effects on 1 nM [3H]DAG0 binding to rat brain membranes (data not shown). As also shown in Table 1, exclusion of peptidase inhibitors from the incubation buffer significantly reduced depolarization-induced displacement of [3H]DAG0 binding. Specific binding of [3H]DAG0 to rat brain membranes was not affected by the presence of peptidase inhibitors (data not shown).

CA3 Lacunosum

moleculare

CA3 CAI Dentate

0

I 2

I 4

Dstnnce

I

bl

I

6 6 from septal pole (mm)

gyrus

I

IO

Fig. 2. Longitudinal distribution of PH]DAGO binding density in rat hippocampus. Autoradiographic images from the transverse hippocampal sections prepared as described for Fig. 1, were digitized and binding density was determined in comparison with the calibration standards as described. Results of specific [‘HIDAGO binding in CA1 and CA3 stratum pyramidale, dentate gyrus, and CA3 lacunosummoleculare are shown. Points represent binding densities, in a representative hippocampus, at approximately 500-pm intervals from the septal to the temporal pole.

49

Opioid release displaces mu binding in hippocampus The requirement for the peptidase inhibitors in the slice binding assays indicated that displacement of [)H]DAGG binding was caused by the release of endogenous opioid peptides which could be degraded by peptidases present in the tissue. To obtain better spatial and temporal control of the stimulation parameters, we tested the effects of focal electrical stimulation on the release of endogenous opioids in the hippocampus. Two major opioidcontaining fiber tracts are known to be present in the hippocampus, the lateral perforant path and the In the hippocampal slice, stimumossy fibers. ‘8~19.30.31 lation of the subiculum near the apex of the dentate gyrus should activate the proenkephalin-containing perforant path fibers that pass through this region.2’ Initial tests to determine the optimal electrode position for perforant path stimulation showed that the maximum displacement of [3H]DAGG binding was obtained by high frequency stimulation at the site shown (PP) in Fig. 5A, although electrode placement at this site is likely to activate other fibers in addition to the opioid-containing lateral perforant path. The current intensities used were within a reasonable physiological limit, as similar current amplitudes were required to evoke synaptic responses measured

T **

**

l.Mrwm~ted

T

10 VTN

VTN-Cc

Ffstim

FP-cu

Fig. 3. Stimulated displacement of [‘HJDAGO from slices was calcium-dependent. Assays were performed as described in Experimental Procedures, Veratridine (3 PM) was added 15 min before the IS-min [‘HIDAGO (10 nM) incubation, 300-PA stimulus trains of 0.3 ms duration at 50 Hz for 1 s every 10 s were begun 10 min before the 15min [‘HIDAGO (10nM) incubation was started (for stimulus location see Fig. 5A). Non-specific [3H]DAG0 binding as determined in the presence of naloxone (10pM) was subtracted from total binding in each test condition to give specific binding values (fmol/slice). The 3 p M veratridine (VTN) value was 63% of the unstimulated control, veratridine stimulation in the absence of added calcium (VTN-Ca) was 1IO%, perforant path location shown in Fig. 5A (PP stim) was 66%, and stimulation in the absence-of added calcium (PP-Ca) was 108%. Bar values are the means -+ S.E.M. from at least nine independent experiments, each test condition having two

replicate slices within the experiment. Statistical comparisons: **P< 0.01(Neuman-Keuls) between the indicated condition and the unstimulated control, and between the indicated condition and its corresponding minus calcium test condition. Neither of the stimuli given in the absence of calcium caused a significant difference in the amount of radiohgand binding compared with the unstimulated control.

‘I-

4I 0

50

IO0

150

200

250

xx)

Stirrutusintensity (pAI Fig. 4. Displacement of [‘HIDAGO resulting from perforant path stimulation was dependent on stimulus intensity. Assays were performed as described in Experimental Procedures; stimulus trains of 0.3 ms duration at SOHz for 1 s every 10 s were begun 10 min before the 15-min (3H]DAG0 (10 nM) incubation was started (for perforant path stimulus location see Fig. 5A). Non-specific [‘HIDAGO binding as determined in the presence of naloxone (10PM) was subtracted from total binding in each test condition to give specific binding values (fmol/slice). Compared with the unstimulated control, 30 PA reduced binding by an average of 15%, lZOyA, 22% and 3OOpA, 32%. Point values are the means f S.E.M. from at least nine independent experiments, each time intensity having two replicate slices within the experiment. Statistical comparisons: *P c 0.05, **P-z0.01 (Neuman-Keuls) between the indicated intensity and the unstimulated control.

physiologically using this stimulation equipment.“” High frequency stimulus trains (50 Hz for 1 s every 10 s) of the perforant pathway reduced specific [3H]DAGO binding by an average of 34% (P < 0.01) compared with non-stimulated slices (Fig. 3). The degree of displacement was similar to the effects of 3 PM veratridine tested at the same time: specific [‘HIDAGO binding was reduced by an average of 37% by veratridine stimulation (P < 0.01). Replacing CaCl* in the incubation buffer with MgC12 blocked the displacement caused by either veratridine or electrical stimulation on [)H]DAGO binding (Fig. 3). The calcium dependence, combined with the observed displacement of mu ligand binding upon depolarization of a tissue known to contain endogenous opioids, is consistent with a mechanism involving the release of endogenous opioid peptides. The displacement of [‘HIDAGO specific binding by focal electrical stimulation was also dependent on the stimulus current intensity (Fig. 4). Stimulus trains to the perforant path as above (0.3-ms pulse duration at 50 Hz for 1 s every 10 s) were delivered at 0, 30, 150, or 300 PA intensities. Specific binding was significantly decreased by 22% at an intensity of 150pA (P < 0.05) and maximal displacement of 32% was measured at 3OOpA (P < 0.01, Fig. 4). Control slices in which the stimulating electrode was positioned but delivered no current, showed no significant [3H]DAG0 displacement. The greater [‘HIDAGO displacement at higher stimulus intensities indicates that more opioid was released, either

50

J. J.

WAGNEK

by activating more fibers or by releasing more peptide from each terminal. [jH]DAGO displacement was also dependent on the frequency of the pulses in the high frequency stimulus train. Stimulus trains (0.3-ms pulse duration at an intensity of 300pA for 1 s every 10s) begun 10 min before the 15min (“HIDAGO (10 nM) incubation were initiated from a bipolar electrode placed at the perforant path location shown (Fig. 5A). Specific binding to unstimulated slices was 6.2 -/c 0.4 fmol per slice (n = 41) under these conditions. A train pulse frequency of 1 Hz (equivalent to 0.1 Hz continuous stimulation) resulted in no significant [3H]DAG0 displacement: specific binding was 108 + 15% (n = 9) of unstimulated hippocampal slices. However, significant displacement of [‘HIDAGO was observed at train pulse frequencies of 10 HZ [72 + 11% of control (n = 17), P < 0.051 and 50 Hz [65 + 11% of control (n = 9), P < 0.011. These results show that high frequency stimulation was required to detect endogenous opioid release using this assay. We compared [3H]DAG0 displacement caused by perforant path stimulation with that induced by stimulation of the mossy fibers, lacunosummoleculare, or entorhinal cortex regions (Fig. 5). The lacunosum-moleculare at the CA3-CAl border contains a high density of mu opioid receptors (Fig. 1) and enkephalin-containing terminals of the lateral perforant path.‘8.‘9 As an additional control for stimulation artifacts, a site lacking opioids and opioid receptors was chosen in layers 5-6 of the medial entorhinal cortex.Z As expected from the distribution of mu receptors and proenkephalin-derived peptides in the slice, the perforant path was the most effective location of focal electrical stimulation, causing a decrease in radioligand specific binding of 34% (P < 0.01). Mossy fiber stimulation resulted in a smaller but significant reduction in [‘HIDAGO binding (21% decrease, P < 0.05). Stimulation of the lacunosum-moleculare at the CAl-CA3 border, a region having some diffuse, opioid-containing terminals, gave a small reduction in binding that was not statistically significant (6%, P > 0.05). Stimulation at the entorhinal cortex site did not significantly change [)H]DAGO binding. The results indicate that focal electrical stimulation effectively releases endogenous opioid peptides following high frequency stimulation at discrete locations in the tissue. The optima1 stimulation sites correspond to the known distribution of endogenous opioid-containing pathways in the hippocampal formation.‘8.‘9.-‘0.~’ DISCUSSION

The principal goal of this study was to define a stimulation paradigm which results in the release of endogenous opioids present in the rat hippocampus. In characterizing our experimental system, we determined that the distribution of mu opioid binding

e/

al

250 PP

MF

Lt.4

EC

Fig. 5. Displacement of [)H)DAGO from slices varied depending on stimulus electrode location. Assays were performed as described in Experimental Procedures; 300 PA stimulus trains of 0.3 ms duration at 50 Hz for 1 s every 10 s were begun 10min before the 15 min [3H]DAG0 (IOnM) incubation was begun. (A) Locations of the stimulus electrode are indicated by EC, medial entorhinal cortex; LM, lacunosum-moleculare of CAICA3; MF, mossy fibers; PP, perforant path fibers; other anatomical features referred to are also labeled. The presubiculum is present in temporal slices 6.5-10mm from the septal pole. (B) Results at each location are expressed as a percentage of the unstimulated control specific binding (defined as 100%). Perforant path fiber stimulation reduced binding by 34% compared with the unstimulated control; mossy fiber, reduced binding by 21%; lacunosum-moleculare reduced binding by 6%; and the medial entorhinal cortex increased binding by 4%. Bar values represent means + S.E.M. from at least nine independent experiments, each test condition having two replicate slices within the experiment. Statistical comparisons: *P < 0.05; l*P c 0.01 (paired r-test) between the indicated location and its unstimulated control.

sites in the hippocampus from one pole to the other pole was nonuniform. In general, the density of [3H]DAG0 binding sites increased along the longitudinal axis from the septal towards the temporal end. Along with this change in binding site density, a change in the pattern of binding site distribution was also observed. This was illustrated by the abrupt appearance of very dense binding in a region between the subiculum and entorhinal cortex approximately 6-7 mm from the septal pole. The mu binding in this area correlated well with appearance of the presubiculum region at this level of the hippocampus. These results suggest that there may be a relatively high density of mu receptor binding in the presubiculum region of the hippocampus that may be a site of endogenous opioid action. The results also showed that high frequency, focal electrical stimulation of the perforant path or the mossy fiber tract caused a reduction in (‘HIDAGO specific binding to mu type opioid receptors. A reduction in specific binding could have conceivably

Opioid release displaces mu binding in hippocampus been caused by a depolarization-induced change in the receptor site (e.g. receptor internalization or desensitization). However, the reduction in specific binding was likely to have resulted from competition between the radioligand and released endogenous opioid peptides since: (1) the reduction in radioligand binding was dependent on the presence of calcium in the buffer. Calcium is required for the release of synaptic vesicles from nerve terminals;“” (2) the reduction in radioligand binding was potentiated by the presence of peptidase inhibitors previously shown to protect endogenous opioids;3’3 (3) the reduction in radioligand binding was not caused by a direct effect on receptor affinity as shown by the lack of effect of these manipulations on radioligand binding to rat brain membranes; (4) the reduction in radioligand binding caused by focal electrical stimulation depended on the precise positioning of the electrode near the two opioid-containing fiber tracts. In sum, these results indicate that [3H]DAGO binding was reduced by competition with an endogenous opioid able to bind to the mu receptor and are not consistent with alternative mechanisms such as depolarization-induced receptor internalization or receptor desensitization. Unlike the results from previous in uiuo studies,“.‘* the reduction of radioligand binding observed in viwo could not have been caused by a change in [3H]DAG0 penetration through the tissue. Previous studies have shown that each of the proenkephalin-derived and prodynorphin-derived opioids are present in the rat hippocampus and can be detected in extracts using selective radioimmunoassay.M Similarly, potassium-induced depolarization of rat hippocampal slices was shown to release endogenous opioids that could be detected by selective radioimmunoassays of the slice superfusion buffers.6 Although each of the endogenous opioids present may have been released by the stimulation paradigms used, the displacement assays described in this study do not directly identify the endogenous opioid responsible for the competition. The displacement of the mu-selective radioligand, [3H]DAG0, indicates that an endogenous opioid able to bind to that site was released. The observation that mossy fiber stimulation was less effective than perforant path stimulation at displacing [‘HIDAGO binding to mu opioid receptor is consistent with the relatively higher concentration of proenkephalin-derived opioids in the perforant path. The proenkephalin-derived opioids have been shown to have higher potency at the mu receptors than the prodynorphin-derived opioids.” Pharmacological studies show that the proenkephalin-derived opioid metorphamide is the most selective endogenous mu opioid$* however, depending on the concentration of peptide achieved at the site, each of the endogenous opioids could be effective mu receptor ligands.’ The peptide responsible for the radioligand displacement was shown to be protected

51

by the peptidase inhibitors added. This peptidase inhibitor mixture was originally designed to protect the enkephalins,‘.” and its effectiveness at protecting the prodynorphin-derived opioids under these conditions has not been established. Preliminary results indicate that [‘251jdynorphin-A(l-17) was not protected by these peptidase inhibitors when incubated with hippocampal slices under these conditions (unpublished observations). By protecting the “enkephalin-like” forms, it is possible that the peptidase inhibitors were trapping partially degraded opioid fragments in more mu-selective forms. Thus, while these results suggest than an enkephalin form of opioid peptide (instead of a dynorphin form) may have been responsible for the radioligand displacement, the identity of the specific opioid peptide responsible was not determined in these assays. Stimulation of the perforant path fibers in the hippocampal slice produced a greater change in radioligand binding than either mossy fiber or lacunosum-moleculare stimulation, although each of these regions contain endogenous opioid peptides. Both the mossy fibers and the perforant path terminals in the lacunosum-moleculare of CA3 have been shown to stain less densely than the perforant path fibers in the subiculum for proenkephalinderived opioid immunoreactivity.‘8.‘9.30.3’ Presumably, the greater displacement following perforant path stimulation site was caused by the release of more endogenous ligand for the mu receptor. Similarly, the lack of a significant reduction in mu receptor binding following stimulation of the lacunosum-moleculare of CA3 suggests that although this region contains a high density of mu receptors, only a small amount of endogenous opioid was released. However, a direct correlation between the extent of radioligand displacement and the amount of peptide released has not been established in this system. The radioligand displacement assay is a highly sensitive method able to detect endogenous peptide release using a single hippocampal slice (0.75 mg protein). Comparable assays of opioid release from hippocampal slices using selective radioimmunoassays required approximately 50 times more tissue in each chamber.6 However, the displacement assay did require a minimum of 15 min exposure to the radioligand to detect specific binding and required that the stimulation start before radioligand addition to detect significant displacement;” presumably, a shorter duration of stimulation will be sufficient to detect opioid release in physiological assays. The minimal stimulus frequency tested which resulted in significant radioligand displacement was relatively high (IO Hz trains). While only a limited range of stimuli were tested in this study, this result is consistent with previous reports that low frequency stimulation of the opioid pathways (0.1 Hz) did not release enough opioid peptide to detect opioid. receptor activation 925 but that higher frequencies (IO-400 Hz) were able to produce naloxone-sensitive

J. J. WAGNERr/ ni

51

effects.“.“~‘s.” The frequencies of stimulation required are higher than those required to release the more classical transmitters, glutamate and GABA, measured electrophysiologically in the hippocampus.” Similarly, high frequency stimulation paradigms were required to release neuropeptide Y26 and teleost luteinizing hormone-releasing hormone.24 Milner et ~1.‘~ have shown that nerve terminals in the medulla, containing both catecholamine and (Leulenkephalin immunoreactivities, package the transmitters separately. The enkephalin-containing large dense core vesicles tended to be located more distal to the synaptic junctions. Thus, the greater stimulus required to release opioids in the hippocampus may reflect a difference in the packaging and location within the nerve terminal of the opioid-containing synaptic vesicles. Peptide-containing vesicles may be released by a slightly different process, perhaps requiring sustained depolarization of the nerve terminal to allow sufficient calcium entry.

extremely sensitive measure of release; radioligand displacement is readily quantifiable; by using appropriate radioligands the displacement assay is highly selective and less sensitive to interference caused by the release of other neurotransmitters than electrophysiological assays of release; the intensity, frequency, and location of stimulation can be varied and the effects on release efficiently evaluated. With this assay, we have determined a set of conditions resulting in endogenous opioid peptide release that can be used to investigate opioid effects in the hippocampus. This info~atjon will guide subsequent physiological studies required to define the function of endogenous opioids in this brain region. Further application of the assay to autoradiographic studies will yield more definitive information concerning the sites of release, the specific peptides involved, and the anatomically relevant receptors within the rat hippocampus. Acknon4edgemenrs-This work was supported by U.S. Pub-

CONCLUSION

lic Health Service Research grants NS-23483, DA-04123

The electrically stimulated displacement assay has several advantages: radioligand displa~ment is an

and GM-07750. We thank Dr B. Roques for providing kelatorphan, and Dr D. Baskin for access to the MCID analysis system.

REFERENCES I.

2.

3. 4. 5. 6.

7. 8. 9. 10. Il.

12.

13. 14. IS. 16. 17. 18. 19. 20.

Akil H., Watson S., Young E., Lewis M., Khachaturian H. and Walker J. (1984) Endogenous opioids: biology and function. A. Rev. Neurosci. 7, 223-255. Bayer S. (1985) Hippocampal region. In The Rut Nervous System (ed. Paxinos G.), Vol. I, pp. 335-352. Academic Press, Sydney. Bouboutou R., Waksman G., Devin J., Fournie-Z&ski M. and Roques B. (1984) Bidentate peptides: highly potent new inhibitors of enkephalin degrading enzymes. Lr@ Sci. 35, 1023-1030. Bramham C., Errington M. and Bliss T. (1988) Naloxone blocks the induction of long-term potentiation in the lateral but not in the medial perforant pathway in the anesthetized rat. Brain Res. 449, 352-356. Brogdon W. and Dickinson C. (1983) A microassay system for measuring esterase activity and protein concentration in small samples and in high-pressure liquid chromatography eluate fractions. Anulyi. Biochem. 131, 499-503. Chavkin C., Bakhit C., Weber E, and Bloom F. (1983) Relative contents and concomitant release of prodyno~hin~n~~o~hin-delved peptides in rat hipp~mpus. Proc. nam. Acud. Sei. U.S.A. 80, 7669-7673. Chavkin C., Henriksen S., S&ins G. and Bloom F. (1985) Selective inactivation of opioid receptors in rat hippocampus demonstrates that dynorphin-A and -B may act on mu-receptors in the CAi region. Brain Res. 331, 366-370. Chavkin C., Shoemaker W., McGinty J., Bayon A. and Bloom F. (1985) Characterization of the prodynorphin and proenkephalin neuropeptide systems in rat hippocampus. J. hkwosci. 5, 808-816. Chavkin C. and Bloom F. (1986) Opiate antagonists do not alter neuronal responses to stimulation of opioid-containing pathways in rat hippocampus. Nekopeptide; 7, 19-22. Chavkin C., Neurnaier J. and Swearcngen E. (1988) Opioid receptor mechanisms in the rat hippocampus. In Opioids in the ~i~~curnu~s feds McGintv J. and Friedman D.), Vol. 82, pp. 94-117. NIDA Research Monog~ph. Caudle RI and -Chavkin C. (1990) Mu opioid recepror a&vat& reduces inhibitory post synaptic potentials in hippocampal CA3 pyramidal cells of rat and guinea pig. J. Pharmac. exp. Ther. 252, 1361-1369. Caudle R., Swearengen E. and Chavkin C. (1989) Endogenously released opioids inhibit inhibitory post synaptic potentials in guinea pig CA3 pyramidal cells and rat dentate granule cells. In Progress in Clinical and Biological Research (ed. Quirion R., Jhamandas K. and Gianoulakis C.), Vol. 328, pp. 5-9. Alan R. Liss, New York. Corbett A., Paterson S., McKnight A., Magnan J. and Kosterlitz H. (1982) Dynorphin (l-8) and dynorphin (I-9) are ligands for the kappa-subtype of opiate receptor. Nurure 299, 79-81. Crain B., Chang K.-J. and McNamara J. (1986) Quantitative autoradiographic analysis of mu and delta opioid binding sites in the rat hippocampal formation. J. camp. Neuroi. 246, 170-180. Derrick B. and Martinez J. (1989) A unique opioid peptide dependent form of long-term potentiation is found in the rat hippocampus. In Progress in Opioid Research (eds Cros J., Meunier J. and Hamon M.). Adv. Biosci. 75, 213-216. Dingledine R. (1981) Possible mechanisms of enkephalin action on hippocampal CA1 pyramidal neurons. J. Neurosci. 1, 1022-1035. Dingledine R. 11984) Hippocampus; synaptic pharmacology. In Bruin S/ices (ed. Dingledine R.), pp. 87-l 12. Plenum Press, New York. Gall C., Brecha N.. Karten H. J. and Chang K.-J. (1981) Localization ofenkephaljn-like immunoreactivity to identified axonal and neuronal populations of the rat hippocampus. J. camp. Neural. 198, 335-350. Gall C., Brecha N., Chang K.-J. and Karten H. J. (1984) Ontogeny of enkephalin-like immunoreactivity in the rat hippocampus. Neuroscience 11, 359-380. Handa B.. Lane A.. Lord A., Morgan B., Rance M. and Smith C. (1981) Analogues of beta-LPH,,,, possessing selective agonist activity at mu-opiate receptors. Eur. J. Pharmac. 70, 531-541.

53

Opioid release displaces mu binding in hippocampus

21. Hjorth-Simonsen A. and Jeune B. (1972) Origin and termination of the hippocampal perforant path in the rat studied by silver impregnation. J. camp. Neural. 144, 215-232. 22. Hong J. and Schmid R. (1981) Intrahippocampal distribution of Met’-enkephalin. Brain Res. 205, 415-418. 23. James I. and Goldstein A. (1984) Site-directed alkylation of multiple opioid receptors. Molec. Pharmac. 25, 337-342. 24. Jan L. and Jan Y. (1982) Peptidergic transmission in sympathetic ganglia of the frog. J. Physiol. 327, 219-246. 25. Linseman M. and Corrigall W. (1981) Are endogenous opiates involved in potentialion of field potentials in hippocampus of the rat. Neurosci. Lerf. 27, 319-324. 26. Lundberg J.. Rudehill A., Sollevi A.. Theodorsson-Norheim E. and Hamberger B. (1986) Frequency- and reserpinedependent chemical coding of sympathetic transmission: differential release of noradrenaline and neuropeptide Y from pig spleen. Neurosci. LPI{. 63, 96-100. 27. Madison D. and Nicoll R. (1988) Enkephalin hyperpolarizes interneurones in the rat hippocampus. J. Physiol. 398, 123-130. 28. Martin M. (1983) Naloxone and long-term potentiation of hippocampal CA3 field potentials in cirro. Neuropeptides 4, 45-50.

Milner T., Pickel V. and Reis D. (1989) Ultrastructural basis for interactions between central opioids and catecholamines. I. Rostra] ventrolateral medulla. J. Neurosci. 9, 2114-2130. 30. McGinty J., Henriksen S., Goldstein A., Terenius L. and Bloom F. (1983) Dynorphin is contained within hippocampal mossy fibers: immunochemical alterations after kainic acid administration and colchicine induced neurotoxicity.

29.

Proc. natn.

Acad.

Sci. U.S. A. 80, 589-593.

McGinty J.. Van der Kooy D. and Bloom F. (I 984) The distribution and morphology of opioid peptide immunoreactive neurons in the cerebral cortex of rats. Neuroscience 4, 1104-l 117. 32. McLean S.. Rothman R.. Jacobson A.. Rice K. and Herkenham M. (1987) Distribution of opiate receptor subtypes and enkephalin and dynorphin immunoreactivity in the hippocampus of squirrel, guinea pig, rat and hamster. J. camp. Neurol. 255, 491-5 IO. 33. Neumaier J. and Chavkin C. (1989) Release of endogenous opioid peptides displaces [‘Hldiprenorphine binding in rat hippocampal slices. Bruin Res. 493, 292-302. 34. North A. (I 986) Opioid receptor types and membrane ion channels. Trends Neurosci. 9, I 14-l 17. 35. Panksepp J. and Bishop P. (1981) An autoradiographic map of [3H]diprenorphine binding in rat brain: effects of social interaction. Brain Res. Bull. 7, 405-410. 36. Paxinos G. and Watson C. ( 1986) The Rat Bruin in Srereofaxic Coordinates, 2nd edn. Academic Press, Orlando, FL. 37. Ramm P., Kulick J.. Stryker M. and Frost B. (1985) Video and scanning microdensitometer-based imaging systems in autoradiographic densttometry. J. Neurosci. Merh. 11, 89-100. 38. Seeger T., Sforzo G.. Pert C. and Pert A. (1984) In aivo autoradiography: visualization of stress-induced changes in opiate receptor occupancy in the rat brain. Bruin Res. 503, 303-31 I. 39. Siggins G. and Zieglgansberger W. (1981) Morphine and opioid peptides reduce inhibitory synaptic potentials in hippocampal pyramidal cells in vitro without alteration of membrane potential. Proc. natn. Acud. Sci. U.S.A. 78, 3 I.

5235-5239. 40.

Smith S. and Augustine G. (1988) Calcium ions, active zones and synaptic transmitter release.

Trends Neurosci.

11,

458464. 41. 42.

Swearengen E. and Chavkin C. (1989) Comparison of opioid and GABA receptor control of excitability and membrane conductance in hippocampal CAI pyramidal cells in rat. Neuropharmacology 28, 689-697. Weber E.. Esch F., Bohlen P., Paterson S., Corbett A., McKnight A., Kosterlitz H., Barchas J. and Evans C. (1983) Metorphamide: isolation, structure and biologic activity of an amidated opioid octapeptide from bovine brain: Proc. num.

43.

Acud.

Sci. U.S.A.

76, 7362-7366.

Zieglgansberger W.. French E.. Siggins G. and Bloom F. (1979) Opioid peptides may excite hippocampal pyramidal neurons by inhibiting adjacent inhibitory interneurons. Science 205, 415417. (Accepred

22 December

1989)