Ž .Brain Research 851 1999 141–147www.elsevier.comrlocaterbres

Research report

Peripheral benzodiazepine receptors in cerebral cortex, but not in internalorgans, are increased following inescapable stress and subsequent

avoidancerescape shuttle-box testing

Julia Lehmann a, Ronit Weizman b, Christopher R. Pryce a, Svetlana Leschiner c, Isabelle Allmann a,Joram Feldon a, Moshe Gavish c,d,)

a Laboratory of BehaÕioural Biology, The Swiss Federal Institute of Technology, CH-8603 Schwerzenbach, Switzerlandb Tel AÕiÕ Community Mental Health Center, Tel AÕiÕ and Sackler Faculty of Medicine, Tel AÕiÕ UniÕersity, 69978 Tel AÕiÕ, Israel

c Department of Pharmacology, The Bruce Rappaport Faculty of Medicine, Technion-Israel Institute of Technology, POB 9649, 31096 Haifa, Israeld Rappaport Family Institute for Research in the Medical Sciences, Technion-Israel Institute of Technology, 31096 Haifa, Israel

Accepted 21 September 1999

Abstract

Ž .Stress-induced alterations in peripheral benzodiazepine receptor PBR density have been reported in humans and in rats. However, thePBR response is highly specific, and its function remains largely unexplained. The aim of the present study was to investigate the

Ž .relationship between behavior in the two-way active avoidance paradigm 2WAA and post-test PBR densities in adrenal, testis, kidney,Ž .and cerebral cortex. Adult male Wistar rats were tested in the 2WAA either in the naive state AA or 24 h following shock preexposure

Ž .PE , known to interfere with avoidancerescape response acquisition, and decapitated immediately after testing. Control subjects weredecapitated without experimental experience. The stressful characteristic of the experiment was validated by significantly increasedpost-test corticosterone levels in AA and PE subjects compared with controls, with a trend towards higher corticosterone levels in PErelative to AA rats. Similarly, PE compared with AA subjects tended to show retarded acquisition of the escaperavoidance response.

Ž .PBR densities in adrenal, kidney, and testis and central benzodiazepine receptors CBR in the cerebral cortex remained unaffected byavoidance testing. Cerebral cortex PBR density was significantly increased in PE subjects. These findings suggest that avoidance testing,although stressful to the animals, led to changes confined to cerebral cortex PBR, indicating that the hypothalamic–pituitary–adrenalŽ .HPA response occurs independently of the PBR response in peripheral organs, and also suggest that the opportunity for coping alters theimpact of the stressor on the subject and prevents the expression of PBR response in peripheral organs. q 1999 Elsevier Science B.V. Allrights reserved.

Keywords: Benzodiazepine receptor; Avoidance learning; Coping; Corticosterone; Unconditioned stimulus preexposure effect; Rat

1. Introduction

The benzodiazepine receptors are divided into centralŽ .benzodiazepine receptors CBR , associated with the

synaptosomal membrane in the central nervous systemŽ . Ž .CNS , and peripheral benzodiazepine receptors PBR ,widely distributed in peripheral tissues but also present in

w xthe CNS 29 . There is clear evidence that CBR mediatethe therapeutic actions of benzodiazepines and are fully

) Corresponding author. Department of Pharmacology, The Bruce Rap-paport Faculty of Medicine, Technion-Israel Institute of Technology, P.O.Box 9649, 31096 Haifa, Israel. Fax: q972-4-851-3145; e-mail:mgavish@tx.technion.ac.il

w xinvolved in the neurobiological control of anxiety 9 . Thefunction of the PBR is not yet fully understood; however,their localization on the outer mitochondrial membrane in

w xa number of organs 29 supports the hypothesis that PBRmediate an integrated metabolic, physiological, and en-

w xdocrine response to environmental stressors 32 .The effect of acute stress on PBR is rapid, short-lasting,

organ-specific, and stressor-dependent. Several studies havedescribed a bidirectional phenomenon, whereby PBR den-sity is upregulated in response to acute stress and downreg-

w xulated following repeated or chronic stress 9 . Studies inrodent models have shown a significant increase in renal

w xPBR density in response to five tail shocks 6 and anincrease in cerebral cortex and cardiac ventricular PBR

0006-8993r99r$ - see front matter q 1999 Elsevier Science B.V. All rights reserved.Ž .PII: S0006-8993 99 02160-5

( )J. Lehmann et al.rBrain Research 851 1999 141–147142

w xdensity in response to acute maximal electroshock 1 ;however, 80 repeated tail shocks significantly decreasedPBR density in kidney, heart, pituitary gland, and cerebral

w xcortex 6 . Likewise, PBR density increased in the cerebralcortex, olfactory bulb, kidney, platelets, and lymphocytes

w xafter acute single forced swimming stress 22,26 anddecreased in the kidney after daily forced swimming for 21

w xdays 2 . A stress-induced alteration in PBR density hasalso been reported in humans, using platelets as the target

w xtissue 4,16,33,34 .PBR activation during acute exposure to stress may

provide individuals with the neural and metabolic prepara-tion for coping with stress. In cases of chronic stress, PBRbinding capacity may be diminished in order to avoidlong-term hypercortisolemia, which may produce CNSdamage, metabolic changes, and immune function impair-ment. Since stress is accompanied by an increase in gluco-corticoid synthesis and release, it has been suggested thatthe downregulation of mitochondrial PBR associated withchronic stress may reflect a neuroendocrine defense mech-anism that reduces mitochondrial cholesterol transport and

w xoverproduction of glucocorticoids 32 . However, themechanisms underlying the stress-related modulations inPBR expression have not yet been established. In somesituations, to enable efficient psychological and physicalfitness, the changes in PBR are accompanied by a simulta-

w xneous increase in CBR activity 9 . However, this is notw xalways true. For example, Mennini et al. 19 reported that

handling-habituated rats exposed to acute noise stressshowed an 80% increase in PBR density in the cerebralcortex and a 30–40% decrease in CBR density in the sametissue.

The physiological impact of a stressor can be modulatedby the subject’s ability to predict andror control theaversive situation, i.e., by the possibility to display acoping response. A coping response generally includes theability to alter onset, pattern, intensity, or duration of the

w xaversive stimulus 21 , and it has been suggested thatuncontrollable aversive events are more stressful than pre-

w xdictable or controllable events 30 . It has been shown invarious studies that coping is likely to reduce the physio-

w xlogical stress of subjects 20,31 . Most of the studies onPBR response to acute stress have employed paradigms inwhich the animals were subjected to an uncontrollable

w xsituation, for example, to inescapable shocks 1,6,7 orw xforced swimming behavior 2,27 , and subjects could nei-

ther predict nor control the situation. In the present study,we selected a model, namely, the two-way active avoid-

Ž .ance paradigm 2WAA , that includes both predictabilityand controllability components, i.e., subjects can adopt astrategy of escape andror avoidance from the negativereinforcer, which can be interpreted as a coping response.However, subjects exposed to active avoidance learning do

w xshow increased plasma corticosterone levels 3 , whichconfirms that the active avoidance paradigm is a stressfulexperience, regardless of coping behavior.

The aim of the present study was to investigate therelationship between behavior in the active avoidanceparadigm and post-test PBR densities in various organs.The stressful characteristic of the paradigm was validatedby post-test measurements of endocrine activity in the

Ž .hypothalamic–pituitary–adrenal HPA axis. Learning per-formance of subjects and consequently also the degree ofstress experienced by the subjects was experimentallymodified by preexposing half of the experimental subjectsto repeated presentations of the unconditioned stimulusŽ .mild electroshock , which has been reported to induce a

w xsubsequent escape deficit 23,28,35 . On the basis of evi-dence that at least renal PBR respond to a large range ofstressors, including those that are predictable or escapablew x13 , we hypothesized that exposure to the active avoid-ance paradigm would include organ-specific changes inPBR density. We also hypothesized that subjects preex-

Ž .posed to the unconditioned stimulus electroshock woulddemonstrate an escape deficit that, in turn, would exertquantitatively different effects on PBR densities than ap-parent in non-preexposed subjects.

2. Materials and methods

2.1. Animals

Experiments were performed on 24 adult male WistarŽ .rats Institute of Toxicology, Zurich, Switzerland aged 3

Žmonths. The animals were housed in Macrolon cages type.IV, 59.0=38.5=20.0 cm under reversed-cycle lighting

Ž .lights on 1900–0700 h in a temperature- and humidity-controlled animal facility, with free access to food andwater in the home cage. The animals were equally dividedinto two treatment groups and one control group, as fol-

Ž .lows: the active avoidance subjects AA group werehabituated to the shuttle boxes 24 h before being tested in

Ž .the 2WAA, and the preexposed subjects PE group weresubjected to inescapable footshock treatment in the activeavoidance apparatus 24 h before testing in the 2WAA;

Ž .control subjects CON group were decapitated for bloodand tissue sampling immediately after removal from thehome cage, i.e., without prior experimental experience. Allanimals, including the CON subjects, were handled gentlyprior to the experiment in order to reduce handling stressduring transfer from the home cage to the test apparatus.All experiments were carried out in agreement with SwissFederal Legislation for Animal Experimentation.

2.2. Test apparatus

The apparatus consisted of four identical shuttle boxesŽmodel E10-16TC, Coulbourn Instruments, Allentown, PA,

.USA , each set in a ventilated sound- and light-attenuatingŽ .shell model E10-20 . The internal dimensions of each

experimental chamber, as measured from the raised grid

( )J. Lehmann et al.rBrain Research 851 1999 141–147 143

floor, were 35=17=21.5 cm. The box was divided by anŽ .aluminum hurdle 17 cm long=4 cm high that was only

1 mm thick to prevent subjects from balancing on top of itin order to avoid shock. Scrambled shocks were delivered

Žfrom a constant-current shock generator model E13-14,. Ž .Coulbourn Instruments and scanner model E13-13 set at

0.5 mA. The conditioned stimulus was an 85-dB toneŽ .produced by a 2.9-kHz tone module model E12-02 placed

behind the shuttle box on the floor of the shell. Thechambers were illuminated during the experimental session

Ž .with two diffuse light sources house lights mounted 19cm above the grid floor in the middle of the side walls.

2.3. Test procedure

The procedure for the 2WAA, with or without preexpo-sure, comprised two sessions 24 h apart. Behavioral testing

Žwas carried out between 0900 and 1600 h during the dark.phase of the animals .

2.3.1. Preexposure sessionEach rat was placed in the shuttle box. PE subjects

received 50 presentations of a 1-s 0.5-mA footshock, withŽ .a variable interstimulus interval of 50 s range, 10–90 s .

AA animals were confined to the chamber for an identicalperiod of time without receiving the shock stimuli.

2.3.2. Test sessionEach animal was placed in the shuttle box and under-

went 100 avoidance trials at a variable interval of 50 s.ŽEach avoidance trial began with a 12-s tone conditioned

.stimulus , the last 2 s of which were concurrent with a0.5-mA footshock. If the animal crossed the barrier to theopposite compartment during the tone, the stimulus was

Žterminated and no shock was delivered avoidance re-.sponse . A crossing response during shock terminated the

Ž .tone and the shock escape response . If the animal failedto cross during the entire tone-shock trial, the tone and theshock terminated after 12 s from the onset of the tone. Thenumber of avoidance and escape responses was recordedfor 100 trials.

2.4. Data collection during test session

2.4.1. AÕoidance responsesThe 100 avoidance trials were divided into 10 blocks of

10 trials each. The number of avoidance responses perblock was calculated for each of the 10 blocks and ex-pressed as percentage of avoidance responses.

2.4.2. AÕoidance plus escape responsesDue to the expectation of an escape deficit in the PE

animals, the total number of avoidance and escape re-Ž .sponses equal to completed trials per block was calcu-

lated and expressed as the total number of trials per blockin which a crossing response occurred.

2.5. Plasma corticosterone

Plasma corticosterone level was determined within 5–10min of the active avoidance test to provide an independentindicator of the stress induced by the paradigm. Outcomeswere compared with the basal levels of this hormonederived from values obtained simultaneously from the

Ž .CON untested rats taken directly from the home cage.Blood was obtained by decapitation, collected into 5-ml

Žheparinized plastic tubes Sarstedt AG, Sevelen, Switzer-.land on ice, and centrifuged; plasma samples were stored

at y258C until assayed. Plasma total corticosterone titersŽ .were determined in a single radioimmunoassay RIA with

w3 x Ža H corticosterone kit 07-120016, ICN Biomedical, Es-.chwege, Germany . The sensitivity of the RIA was 15

pgrml, and intra-assay precision was 4%.

[3 ]2.6. H PK 11195 binding assay

Immediately following decapitation and blood sam-pling, each subject’s adrenals, testes, kidneys, and brainwere removed and stored at y708C for PBR bindingassay. Briefly, the organs were homogenized separately for

Ž .15 s with a Brinkmann Polytron setting 10 . The ho-mogenates were centrifuged at 49,000=g for 15 min, andthe pellets were resuspended in 50–100 volumes of ice-cold

w3 xbuffer. H PK 11195 binding assays were conducted asw xdescribed previously 12 . The reaction mixture consisted

Ž .of 400 ml homogenate 100–200 mg protein and 25 mlw3 x Ž .H PK 11195 0.2–6 nM final concentration in the ab-

Ž . Ž .sence total binding or presence nonspecific binding ofŽ .75 ml unlabeled PK 11195 final concentration, 10 mM .

After incubation for 60 min at 48C, samples were filteredunder vacuum over Whatman GFrC filters and washedthree times with 3 ml of potassium phosphate buffer.Filters were placed in vials containing 4 ml of scintillation

Žcocktail Opti-Fluor, Packard, Groningen, The Nether-.lands and counted in a scintillation counter following 12 h

Ž .equilibration. The maximal number of binding sites BmaxŽ .and the equilibration constant K were calculated ford

each animal individually using Scatchard analysis of satu-w3 xration curves of H PK 11195 binding.

[3 ]2.7. H Ro 15-1788 binding assay

w3 xH Ro 15-1788 binding in cerebral cortex was assayedin 50 mM potassium phosphate buffer, pH 7.4, at 48C in afinal volume of 500 ml. The reaction mixture contained

w3 x400 ml membrane homogenate and 50 ml H Ro 15-1788Ž . Ž0.1–3 nM final concentration in the absence total bind-

. Ž .ing or presence nonspecific binding of 50 ml clon-Ž .azepam final concentration, 10 mM . The rest of the

w3 xprocedure was as described for the H PK 11195 bindingassay.

( )J. Lehmann et al.rBrain Research 851 1999 141–147144

2.8. Statistical analysis

Data of the avoidance responses were analyzed by aŽ .2=10 two-way analysis of variance ANOVA with a

Ž .main factor of preexposure 0, 50 and a repeated-measure-Ž .ment factor of block 1–10 . Due to the expectation of an

escape deficit in the PE animals, the mean number ofcompleted trials was calculated across all 100 trials andanalyzed using the nonparametric Mann–Whitney U-testcomparison. Plasma corticosterone data were analyzed us-ing a one-way ANOVA with the main factor of treatmentŽ .CONrAArPE . To test PBR binding, a one-way ANOVA

Ž .with the main factor of treatment CONrAArPE wascarried out for each of the organs separately.

3. Results

3.1. ActiÕe aÕoidance

3.1.1. AÕoidance responsesThe ANOVA revealed a highly significant main effect

Ž Ž . .of blocks F 9,126 s20.3, P-0.001 , indicating anoverall increase in avoidance responses as a function ofprogressive training. Due to the relatively high between-subject variation, the effect of preexposure did not reach

Ž .significance P)0.4 . However, as can be seen in Fig. 1,the PE subjects tended to demonstrate fewer avoidanceresponses.

3.1.2. AÕoidance plus escape responsesThe mean number of completed trials, i.e., trials in

which the subject responded with either escape from oravoidance of the shock, differed significantly between AA

Ž .and PE subjects Mann–Whitney U-test, P-0.05 . AsŽ .can be seen in Fig. 1 inset histogram , the AA rats

Fig. 1. Means"S.E.M. of avoidance responses in 10 blocks of 10 trialsŽ . Žeach for shock-preexposed open symbols, PE and non-preexposed filled

.symbols, AA adult Wistar rats. Inset histogram displays the averageŽ .number "S.E.M. escape plus avoidance responses for shock-preex-

Ž . Ž . Uposed PE and non-preexposed AA adult Wistar rats. P -0.05Ž .Mann–Whitney U-test .

Fig. 2. Means"S.E.M. of plasma corticosterone concentration in controlŽ . Ž .CON subjects and following avoidance testing with PE or withoutŽ . UAA previous preexposure to the unconditioned stimulus. P -0.05Ž .Fisher’s protected LSD .

responded to the stimuli, on average, significantly moreoften than their PE counterparts.

3.2. Corticosterone

ANOVA revealed a significant effect of treatmentŽ Ž . .F 2,21 s8.95, P-0.002 , indicating differences amongthe three groups in peripheral cortisol titers at sacrifice.

Ž .Post-hoc analysis Fisher’s protected LSD showed thatboth the AA and PE subjects had higher plasma cortisoltiters than CON subjects. As can be seen in Fig. 2, PEsubjects also tended to have higher post-test cortisol values

Ž .than AA animals P-0.2 . There was no time-of-dayeffect on the cortisol titers.

3.3. PBR density

Ž .The mean "S.E.M. PBR values for each of theexperimental groups are shown in Table 1. None of the

Table 1B and K values in rat adrenal gland, kidney, testis, and cerebralmax d

Ž . Ž .cortex following avoidance testing with PE or without AA preexpo-Ž .sure to the unconditioned stimulus and in controls CON

Results are means"S.E.M.

CON AA PE

( )B fmolrmg proteinm a x

PBR adrenal 60213"2885 72 013"6965 67152"3004PBR testis 6286"302 6122"124 6105"461PBR kidney 7269"499 7716"543 8179"476

aPBR cortex 372"69 385"69 680"53CBR cortex 1629"79 1548"1205 1529"228

( )K nMd

PBR adrenal 4.26"0.28 4.19"0.43 3.86"0.34PBR testis 1.53"0.16 1.63"0.11 1.69"0.32PBR kidney 2.80"0.37 3.11"0.23 2.31"0.22

aPBR cortex 1.10"0.14 0.96"0.10 1.75"0.21CBR cortex 3.03"0.23 2.46"0.25 2.84"0.49

a Ž .P -0.01 vs. CON and AA ANOVA .

( )J. Lehmann et al.rBrain Research 851 1999 141–147 145

one-way ANOVAs for PBR in adrenal, kidney, or testisrevealed a significant effect of treatment. PBR in cerebralcortex demonstrated a significant increase in density in the

w Ž .PE animals compared with the other two groups F 2,21xs7.36, P-0.004 , which did not differ from each other.

K values in all groups were in the nanomolar range.d

3.4. CBR in cerebral cortex

One-way ANOVA did not reveal a significant effect oftreatment in cerebral cortex CBR.

4. Discussion

The present study demonstrated that PBR densities inkidney, adrenal, testis, and cerebral cortex, as well as CBRdensity in the cerebral cortex, were not influenced byexposure to avoidable electroshock, even though this expe-rience is inherently stressful, as shown by the elevation in

Ž .HPA activity i.e., in corticosterone levels .w xBrain PBR are localized mainly in glial cells 29 and

are involved in the biosynthesis of g-aminobutyric acid-ac-w xtive neurosteroids 24 ; thus, it is conceivable that the

PE-induced upregulation of cerebral cortex PBR is gearedto stimulate the release of neuroactive steroids in this brain

w xregion in order to attenuate the stress-related anxiety 5,25 .It should be pointed out, though, that it was not possible toestablish whether this increase in brain PBR in the PEgroup as compared with the AA group was due to the

Žshock preexposure procedure taking place 1 day before.the actual avoidance training or to the fact that the PE

group was exposed to more shocks during avoidance train-Ž .ing as expressed by an escape deficit . The fact that, other

than an increase in cerebral cortex PBR, preexposure didnot influence PBR or CBR densities in any other organmight be attributable to the less intense preexposure proce-dure used here. Previous studies in our laboratory havedemonstrated that animals preexposed to a stronger stimu-lus have a much more robust escape deficit, compared withnonexposed rats, than the animals tested here. This has

w xalso been reported by several other authors 15,18 . Never-theless, as the kidney is generally considered the organ

w xmost sensitive to stress across a range of paradigms 9 ,our finding is surprising. Some authors have claimed thatinescapable shock exposure can result in learned helpless-

w xness and thereby changes in renal PBR 9 . In the presentstudy, although the animals preexposed to the uncondi-tioned stimulus made fewer avoidancerescape responsesŽ .learned helplessness and exhibited an upregulation ofcerebral cortex PBR, renal PBR remained unchanged.

w xBasile et al. 1 reported similar findings following maxi-mal electroshock exposure. It should be noted, however,that compared with other reports in the literature, we used

Ž .a relatively mild footshock intensity 0.5 mA . It is also of

w xnote that Drugan et al. 11 showed an alteration in CBRw3 xdensity, as assessed by H Ro 15-1788 binding in various

Ž .brain areas hippocampus, cerebral cortex, and striatumfollowing shuttle-box escape testing, using higher intensi-ties of shock stress than the present study, as well asmeasuring the binding in vivo, rather than the in vitroanalysis in the current study.

The fact that we did not detect an increase in renal PBRin the AA subjects does not rule out the possibility that atransient increase in their density did occur, followed by arapid return to prestress density. The active avoidance testlasts about 1.5 h on average, depending on individual

w xperformance. Given the finding of Holmes et al. 13 that areversal in the PBR density response occurs within 2 h, wemight assume that the PBR system responds immediatelyafter onset of the experiment and returns to baseline assoon as the subject starts to learn. Yet if this were true, theresponse of the PBR system would be much faster than theresponse of the HPA axis. On the other hand, Drugan et al.w x6 reported that changes in renal PBR occur in response toinescapable shocks that persist for 2 h or more. In addition,considering our results in terms of time course, we wouldexpect a greater interindividual variation in PBR density inthe experimental than the control animals, with PBR den-sity being correlated with time elapsed since the final

Ž .shock depending on learning performance ; however, thiswas not the case here.

The major difference between the studies that describe aPBR response and the present study was the involvementhere of learning and coping, i.e., the subjects could adopt astrategy of escape from or even avoidance of the shock,based on classical conditioning. Nevertheless, the avoid-ance itself proved to be a stressor, as shown by theelevation in cortisol levels. In an attempt to dissociate thepsychological and physiological impacts of a stressor on

w xthe PBR and CBR, Holmes et al. 13 manipulated thecontrollability or predictability of the stressor. They foundthat the response of the kidney did not depend on theseparameters. In the lung, stressor controllability appeared toprotect against PBR reduction, whereas in the olfactorybulb, the PBR were responsive to a purely psychological

w xstressor, specifically, the conditioned fear response 13 .w xHowever, in contrast to the study of Holmes et al. 13 ,

where the animals could predict or escape the shock, in ourexperiment the animals could also avoid it. This differencemay be decisive. The psychological component of thestressor may be of much lesser weight under conditions inwhich the shock is avoidable than when the noxious eventis only predictable; indeed, the latter may even enhanceindividual fear. There is evidence that the controllability ofa stressor may lead not only to changes in behavior, butalso to changes in the physiological impact of the stressorw x w x17 . This is supported by the study of Drugan et al. 8 ,showing that the noxious component of the stressor maybe the crucial factor for PBR response. This componentwas clearly present under the conditions of Holmes et al.

( )J. Lehmann et al.rBrain Research 851 1999 141–147146

w x13 , but much reduced in the active avoidance paradigmbecause the subject had the option to avoid the shock.

w xWhen Holmes et al. 13 tested the conditioned fear re-sponse, which lacks a noxious component during testing,they did not find changes in renal, adrenal, or cerebralcortex PBR densities, which is in line with our presentresults.

A separation of the steroid hormone and the PBRresponse to stress has been proposed by Holmes and

w x w xDrugan 14 . Drugan et al. 7 showed that the stress-in-duced renal PBR response is enhanced followingadrenalectomy, suggesting that these two systems functionindependently, i.e., stress can stimulate PBR expression inthe kidney without a parallel elevation in adrenal glucocor-ticoids, indicating a possible role of the renin–angiotensin

w xsystem in renal PBR response to stress 13 . Interestingly,the trend towards higher post-test blood corticosteronelevels in rats preexposed to the unconditioned stimulusŽ . Ž .shocks in an unavoidable situation PE suggests thatsubsequent avoidance testing was more stressful for thosesubjects compared with rats tested only in the avoidance

Ž .paradigm without preexposure AA , which may have beenŽdue to the deficit in performance more trials in which no

.escape or avoidance response occurred displayed by PEsubjects.

In summary, we have shown that a stressful experience,such as is the result of active avoidance learning, does notlead to changes in PBR density in peripheral organs andthat the steroid response to stress is independent of PBR

w xreactivity. This supports the claim of Drugan et al. 10that ‘the apparent chaotic regulation of PBR in a variety oftissues that are differentially activated by stress may, uponfurther reflection, represent a fine-tuned correlate of thedifferent states of stress reactivity’; the glucocorticoidresponse, by contrast, is a more general reaction to stress.The present results suggest that there is a dissociationbetween the psychological and physiological PBR re-sponse to stress and that the noxious component may bethe critical factor for activating the PBR system.

Acknowledgements

This research was supported by a grant from the SwissFederal Institute of Technology, Zurich. We thank theanimal care team for their assistance and Ruth Singer forediting the manuscript.

References

w x1 A.S. Basile, B.A. Weissman, P. Skolnick, Maximal electroshockw3 xincreases the density of H Ro 5-4864 binding to muse cerebralŽ .cortex, Brain Res. Bull. 19 1987 1–7.

w x2 R. Burgin, R. Weizman, M. Gavish, Repeated swim stress and

peripheral-type benzodiazepine receptors, Neuropsychobiology 33Ž .1996 28–31.

w x3 G.D. Coover, S. Levine, H. Ursin, Plasma corticosterone levelsŽ .during active avoidance learning, Acta Physiol. Scand. 82 1971

16A–17A.w x4 D.E. Dar, A. Weizman, L. Karp, A. Grinshpoon, M. Bidder, M.

Kotler, S. Tyano, A. Bleich, M. Gavish, Platelet peripheral benzodi-Ž .azepine receptors in repeated stress, Life Sci. 48 1991 341–346.

w x5 S.I. Deutsch, J. Mastropaolo, A. Hitri, GABA-active steroids, en-dogenous modulators of GABA-gated chloride ion conductance,

Ž .Clin. Neuropharmacol. 15 1992 352–364.w x6 R.C. Drugan, A.S. Basile, J.N. Crawley, S.M. Paul, P. Skolnick,

w3 xInescapable shock reduces H Ro 5-4864 binding to ‘peripheral-type’ benzodiazepine receptors in the rat, Pharmacol. Biochem.

Ž .Behav. 24 1986 1673–1677.w x7 R.C. Drugan, A.S. Basile, J.N. Crawley, S.M. Paul, P. Skolnick,

w3 xCharacterization of stress-induced alterations in H Ro5-4864 bind-ing to peripheral benzodiazepine receptors in rat heart and kidney,

Ž .Pharmacol. Biochem. Behav. 30 1988 1015–1020.w x8 R.C. Drugan, A.S. Basile, J.H. Ha, R.J. Ferland, The protective

effects of stress control may be mediated by increased brain levelsŽ .of benzodiazepine receptor agonists, Brain Res. 661 1994 127–136.

w x9 R.C. Drugan, P.V. Holmes, Central and peripheral benzodiazepinereceptors: involvement in an organism’s response to physical and

Ž .psychological stress, Neurosci. Biobehav. Res. 15 1991 277–298.w x10 R.C. Drugan, P.V. Holmes, D.M. Scher, S. Luczak, H. Oh, R.J.

Ferland, Environmentally induced changes in peripheral benzodi-azepine receptors are stressor and tissue specific, Pharmacol.

Ž .Biochem. Behav. 50 1995 551–562.w x11 R.C. Drugan, A.L. Morrow, R. Weizman, A. Weizman, S.I. Deutsch,

J.N. Crawley, S.M. Paul, Stress-induced behavioral depression in therat is associated with a decrease in GABA receptor-mediated chlo-ride ion flux and brain benzodiazepine receptor occupancy, Brain

Ž .Res. 15 1989 45–51.w x12 F. Fares, M. Gavish, Characterization of peripheral benzodiazepine

binding sites in human term placenta, Biochem. Pharmacol. 35Ž .1986 227–230.

w x13 P.V. Holmes, A.P. Stringer, R.C. Drugan, Impact of psychologicaldynamics on the peripheral benzodiazepine receptor, Pharmacol.

Ž .Biochem. Behav. 42 1992 437–444.w x14 P.V. Holmes, R.C. Drugan, Stress-induced regulation of the renal

peripheral benzodiazepine receptor: possible role of the renin–Ž .angiotensin system, Psychoneuroendocrinology 19 1994 43–54.

w x15 R.L. Jackson, T.R. Minor, Effects of signaling inescapable shock onsubsequent escape learning: implications for theories of coping and‘learned helplessness’, J. Exp. Psychol. Anim. Behav. Process 14Ž .1988 390–400.

w x16 L. Karp, A. Weizman, S. Tyano, M. Gavish, Examination stress,platelet peripheral benzodiazepine binding sites, and plasma hor-

Ž .mone levels, Life Sci. 44 1989 1077–1082.w x17 S.F. Maier, S.M. Ryan, C.M. Barksdale, N.H. Kalin, Stressor con-

trollability and the pituitary–adrenal system, Behav. Neurosci. 100Ž .1986 669–674.

w x18 S.F. Maier, M.E.P. Seligman, Learned helplessness: theory andŽ .evidence, J. Exp. Psychol. Gen. 105 1976 3–46.

w x19 T. Mennini, M. Gobbi, C. Caharuchinda, Noise-induced oppositechanges in central and peripheral benzodiazepine receptors in rat

Ž .brain cortex, Neurosci. Res. Commun. 5 1989 27–34.w x20 S. Mineka, R.W. Hendersen, Controllability and predictability in

Ž .acquired motivation, Annu. Rev. Psychol. 36 1985 495–529.w x21 R. Murison, E. Isaksen, H. Ursin, ‘Coping’ and gastric ulceration in

rats after prolonged active avoidance performance, Physiol. Behav.Ž .27 1981 345–348.

w x22 M.L. Novas, J.H. Medina, D. Calvo, E. De Robertis, Increase ofperipheral type benzodiazepine binding sites in kidney and olfactory

Ž .bulb in acutely stressed rats, Eur. J. Pharmacol. 135 1987 243–246.w x23 J.B. Overmier, M.E.P. Seligman, Effects of inescapable shock upon

( )J. Lehmann et al.rBrain Research 851 1999 141–147 147

subsequent escape and avoidance responding, J. Comp. Physiol.Ž .Psychol. 63 1967 28–33.

w x24 V. Papadopoulos, P. Guarneri, K.E. Krueger, A. Guidotti, E. Costa,Pregnenolone biosynthesis in C6-2B glioma cell mitochondria: regu-lation by a mitochondrial diazepam binding inhibitor receptor, Proc.

Ž .Natl. Acad. Sci. U. S. A. 89 1992 5113–5117.w x25 R.H. Purdy, A.L. Morrow, P.H. Moore Jr., S.M. Paul, Stress-in-

duced elevations of g-aminobutyric acid type A receptor-activeŽ .steroids in the rat brain, Proc. Natl. Acad. Sci. U. S. A. 88 1991

4553–4557.w x26 L., Rago, A. Adojaan, P. Pokk, The effect of stress on omega3¨

benzodiazepine receptor in rat blood platelets and lymphocytes: theeffect of non-benzodiazepine tranquilizers, in:: L.H. AllikmentsŽ .Ed. , Molecular Pharmacology of Receptors, Vol. 3, Tartu Univ.Press, Tartu, Estonia, 1989, pp. 4–16.

w x27 L. Rago, R.-A. Kivet, J. Harro, M. Pold, Central- and peripheral-type¨benzodiazepine receptors: similar regulation by stress and GABA

Ž .receptor agonists, Pharmacol. Biochem. Behav. 32 1989 879–883.w x28 M.E.P. Seligman, G. Beagley, Learned helplessness in the rat, J.

Ž .Comp. Physiol. Psychol. 88 1975 534–541.

w x29 A. Verma, S.H. Snyder, Peripheral type benzodiazepine receptors,Ž .Annu. Rev. Pharmacol. Toxicol. 29 1989 307–322.

w x30 J.M. Weiss, Effects of coping responses on stress, J. Comp. Physiol.Ž .Psychol. 65 1968 251–260.

w x31 J.M. Weiss, Effects of coping behavior in different warning signalconditions on stress pathology in rats, J. Comp. Physiol. Psychol. 77Ž .1971 1–13.

w x32 R. Weizman, M. Gavish, Molecular cellular and behavioral aspectsof peripheral-type benzodiazepine receptors, Clin. Neuropharmacol.

Ž .16 1993 401–417.w x33 R. Weizman, L. Karp, D.E. Dar, B. Butin, H. Hermesh, H. Munitz,

M. Gavish, Electroconvulsive therapy down-regulates platelet pe-ripheral-type benzodiazepine receptors in medication-resistant major

Ž .depressed patients, Biol. Psychiatry 40 1996 221–224.w x34 R. Weizman, N. Laor, L. Karp, E. Dagan, A. Reiss, D.E. Dar, L.

Wolmer, M. Gavish, Alteration of platelet benzodiazepine receptorsŽ .by stress of war, Am. J. Psychiatry 151 1994 766–767.

w x35 D.A. Zhukov, Strain-dependent escape deficit in two rat models ofŽ .learned helplessness, Physiol. Behav. 53 1993 905–909.