Naloxone and shock-elicited freezing in the rat

Seediscussions,stats,andauthorprofilesforthispublicationat:https://www.researchgate.net/publication/22663686

Naloxoneandshock-elicitedfreezingintherat

ArticleinJournalofComparativeandPhysiologicalPsychology·September1979

DOI:10.1037/h0077609·Source:PubMed

CITATIONS

289

READS

104

2authors,including:

MichaelFanselow

UniversityofCalifornia,LosAngeles

281PUBLICATIONS23,952CITATIONS

SEEPROFILE

AllcontentfollowingthispagewasuploadedbyMichaelFanselowon23January2015.

Theuserhasrequestedenhancementofthedownloadedfile.

Journal of Comparative and Physiological Psychology1979, Vol. 93, No. 4, 736-744

Naloxone and Shock-Elicited Freezing in the Rat

Michael S. Fanselow and Robert C. BollesUniversity of Washington

The freezing behavior of the rat that occurs following painful electric shockwas found to increase when the animal was pretreated with the opiate antago-nist naloxone. Freezing was a positive linear function of drug dose and shockintensity (Experiment 2). Naloxone pretreatment enhanced freezing onlywhen the animal was given two or three shocks but did not affect freezingwhen the animal was given only one shock or not shocked at all (Experiments3, 4, and 5). Naloxone must be present during shock, not just during the ob-servation period, in order to increase freezing (Experiment 6). These resultssuggest that when an animal is shocked, it releases endogenous analgesics (en-dorphins) that make a subsequent shock less aversive. Naloxone, by blockingthe endorphin system, makes the shock more aversive than it would normallybe.

Peptides with properties similar to thenarcotic analgesics have recently been iso-lated (Hughes, 1975). These substances,designated endorphins, are blocked by thenarcotic antagonist naloxone (Belluzzi,Grant, Garsky, Sarantakis, Wise, & Stein,1976; Loh, Tseng, Wei, & Li, 1976; Malick &Goldstein, 1977; Walker, Berntson, Sand-man, Coy, Schally, & Kastin, 1977). Onephysiological role of these endogenous an-algesics may be to reduce the aversiveness ofpainful stimuli. Such a mechanism wouldallow an animal to engage in defensive be-haviors despite the pain of injuries it mighthave received. If this were the case, nalox-one would be expected to potentiate an ani-mal's reaction to painful stimuli. In supportof this hypothesis, naloxone has been re-ported to enhance reactivity to nocioceptivethermotic stimuli (Frederickson, Burgis, &Edwards, 1977; Grevert & Goldstein, 1977b;Jacob, Tremblay, & Colombel, 1974; Walkeret al., 1977) and visceral pain induced byacetic acid (Kokka & Fairhurst, 1977).

Naloxone's effect on the reaction to elec-

This study was supported by Grant BN.S 76-19912from the National Science Foundation. We would liketo thank T. M. McEvoy, Cynthia Carroll, and M. E.Bouton for their assistance in the research, and EndoLaboratories for generously providing naloxone.

Requests for reprints should be sent to Michael S.Fanselow, Department of Psychology, University ofWashington, Seattle, Washington 98195.

trie shock is less clear. Pain thresholds forelectric shock in humans are not changed bynaloxone pretreatment (El-Sobky, Dost-rovsky, & Wall, 1976). Nor does naloxoneaffect the threshold shock intensity for apole-climbing escape task in rats (Goldstein,Pryor, Otis, & Larsen, 1976). However,Fanselow (1979) found that naloxone at-tenuates rat's preference for signaled shock.This result suggests that a signal whichpredicts shock causes an anticipatory releaseof endorphin, which is capable of modifyingthe aversiveness of the impending shock.Naloxone may attenuate the preference byreducing the effectiveness of this anticipa-tory endorphin release.

The present study provides further evi-dence for this interpretation. Our methodis based on the principle that rats typicallyfreeze following exposure to painful electricshock (Bolles & Collier, 1976). This effectis graded, so the more intense the shock thehigher the probability that the rat will freeze(Blanchard & Blanchard, 1969). If naloxoneenhances the aversiveness of electric shock,rats treated with naloxone before exposureto a series of shocks should freeze more aftershock than saline-treated rats.

Experiment 1

Method

Subjects. The subjects in this experiment were 32

Copyright 1979 by the American Psychological Association, Inc. 0021-9940/79/9304-0736$00.75

736

NALOXONE AND SHOCK-ELICITED FREEZING 737

adult female rats of Long-Evans descent, weighing be-tween 214 and 284 g, raised at the University of Wash-ington, Psychology Department vivarium. The ratswere maintained in individual cages on ad lib food andwater. The experiment was conducted during thelights-on portion of a 12:12 hr day/night cycle.

Apparatus. The rectangular observation chamberhad a Plexiglas front viewing wall 26 cm long. The re-maining walls were made of stainless steel. Thechamber was 23 cm deep. The Plexiglas ceiling was 24cm above the floor. The floor was composed of 13stainless steel rods 1 cm in diameter and spaced 1.9 cmcenter to center. Illumination was provided by a 7.5-Wwhite light bulb suspended 5 cm above the ceiling. Theobservation chamber was placed inside a sound-atten-uating chest. A fan produced background noise andventilation.

Electric shock was produced by a Grason-Stadlershock generator/scrambler wired to each of the rods aswell as the stainless steel walls.

To minimize the influence of olfactory cues, 4.5 ml ofa germicidal detergent (Vestal, 1-stroke ves-phene) wereplaced in 270 ml of tap water and a 2-mm bath of thissolution was placed in the catch tray 4.5 cm below thegrid floor. The observation chamber was wiped cleanwith this solution before a subject was placed inside.The resistance between rods was checked with an ohmmeter to ensure that the detergent solution, feces, orurine did not short out any grid combinations.

Procedure. A rat was removed from its home cageand given an ip injection. For half of the animals theinjection was 3 mg/kg naloxone hydrochloride in anisotonic saline vehicle (3 mg/ml). The other half of theanimals received an equivalent volume of isotonic sa-line.

Immediately after the injection the rats were placedin the observation chamber. Two minutes afterplacement in the box, half of the saline and half of thenaloxone animals received three shocks, through thegrid floor. The shocks (1-mA intensity, .75-sec dura-tion) were spaced 20 sec apart. The remaining animalsreceived identical treatment but with the shock gener-ator off.

Immediately following the last shock (or the waitingperiod for the no-shock animals) the animals were ob-served for 8 min. Every fifth second an observer (blindto the type of injection given) recorded the subject'sbehavior according to the following five categories:freezing—absence of all observable skeletal movementexcept respiration, and minimal vibrissae movements;locomotion—use of the rear legs in forward motion;rearing—raising the front paws above the grid floor;grooming—all grooming, scratching, and licking of thebody; general activity—any behavior that could not beclassified as one of the above.

Results

For each rat a "freezing score" was calcu-lated as the percentage of all behavioral ob-servations that were categorized as freezing.These scores were subjected to a 2 X 2 fac-torial analysis of variance. The mean

freezing score was 67% for the shocked ani-mals and only 2% for the nonshocked ani-mals, F(l, 28) = 111.3, p < .005. The maineffect for naloxone versus saline was alsosignificant, F(l, 28) = 9.33, p < .005.

The significant interaction, F(l, 28) =7.66, p < .01, took the following form: Theshocked-naloxone animals froze more thanthe shocked-saline animals (means of 85%and 49%, respectively). This difference wassignificant by both a Scheffe test using theerror term of the analysis of variance (p <.01) and an individual t test, £(14) = 2.78, p< .01. The nonshocked-naloxone rats didnot freeze reliably more than the non-shocked-saline animals (3% and 1%, respec-tively). This difference was analyzed byboth a Fisher's test (F = .65) using the errorterm of the analysis of variance and an in-dividual t test (t = .86).

Naloxone not only had little effect uponfreezing in the nonshocked groups, it hadlittle or no effect on the other types of ac-tivity. T tests were calculated which com-pared the nonshocked saline and naloxonegroups on each of the other behavior cate-gories; none of these t values approachedsignificance.

Discussion

The large increase in postshock freezingshown by the naloxone animals could nothave been a general or nonspecific effect ofthe drug, such as interference with motorability, because nonshocked-control animalsshowed no behavioral effect of the naloxonetreatment.

Since the amount of postshock freezing isa direct function of shock intensity(Blanchard & Blanchard, 1969), we suggestthat naloxone may have produced its effecton the shocked animals by increasing theaversiveness of the shock. Naloxone couldhave produced its effect on the aversivenessof shock by antagonizing the rat's endoge-nous analgesic system. The naloxone-treated animals may have evidenced morefreezing than the controls simply becausethey experienced a more aversive shock thanthe controls. Such an interpretation isconsistent with the finding that endorphinsproduce an analgesia that is antagonized by

738 MICHAEL S. FANSELOW AND ROBERT C. BOLLES

naloxone (e.g., Belluzzi et al., 1976; Loh et al.,1976; Malick & Goldstein, 1977).

Although Goldstein et al. (1976) found noeffect of naloxone on the shock-escapethreshold of rats, they used shock intensitiesthat were below .2 mA, which may be suffi-cient to motivate escape behavior withoutbeing very painful. Naloxone might wellaffect a pain threshold without altering thethreshold of detection or of aversiveness(Campbell & Masterson, 1969).

Experiment 2

In the first experiment, freezing was as-sumed to vary with the aversiveness of shock.This assumption is based on Blanchard andBlanchard's (1969) finding that the amountof postshock "crouching" increases withshock intensity. But as Bolles and Collier(1976) pointed out, freezing, as it has beendefined here, is not the same thing ascrouching, as defined by Blanchard andBlanchard, although crouching subsumesmost freezing postures. Experiment 2demonstrates that what we call freezing isindeed a function of shock intensity andtherefore can be used to estimate the aver-siveness of shock. The experiment was alsodesigned to show naloxone's enhancementof postshock freezing as a function of thedose of the drug.

Method

Subjects. The subjects were 120 rats from the samepopulation and were housed under the same1'conditionsas those of Experiment 1. They were between 88 and115 days old and weighed between 184 and 300 g.

Apparatus. The equipment used was the same asin the first experiment.

Procedure. Six animals were randomly assigned toeach of the 20 cells of a 5 X 4 factorial design. The twofactors were shock intensity (2.0 mA, 1.0 mA, .5 mA, .25mA, or no shock) versus naloxone dose (8 mg/kg, 3ipng/kg, .5 mg/kg, or saline).

An animal was removed from its home cage and giventhe appropriate injection. It was then placed in theobservation chamber and after 3 min given two shocks,at the appropriate intensity. The shocks, like those ofExperiment 1, were .75 sec in duration and spaced 20sec apart. Following the second shock the animals wereobserved for 8 min with the behavior-sampling proce-dure described in the previous experiment. After a ratwas removed from the box, the boluses in the catch traywere counted, and then the catch tray was cleaned.

Z 40-

M

LJCCU_ 20-

SHOCK (ma)

Figure 1. Percentage of freezing in an 8-min periodfollowing shocks of a given intensity (collapsed overdrug treatment).

Results and Discussion

Shock intensity had a significant effect onfreezing, F(4, 100) = 35.6, p < .005. Thehighly significant linear trend, F(l, 100) =133.5, p < .005, indicates that rats freezemore with more intense shocks (see Figure1). This finding validates our assumptionthat freezing reflects aversiveness. Theseresults are consistent with those of otherauthors who found that crouching increases(Blanchard & Blanchard, 1969) and activitydecreases (Anisman & Waller, 1973) withincreasing shock intensity.

Naloxone potentiated freezing in responseto shock, F(3, 100) = 4.2, p < .005, whichreplicates the effect found in the first ex-periment with two as opposed to threeshocks. The potentiation effect increasedlinearly, F(l, 100) = 11.9, p < .005, (Gaito,1965) with dose of naloxone (see Figure 2).

As can be seen in Figure 3, the greatest

60-

40-

UJLU

20-

0.5 3 8

DOSE (mg/kg)Figure 2. Percentage of freezing in an 8-min periodfollowing shocks, as a function of the dose of naloxone(collapsed over shock intensity).

NALOXONE AND SHOCK-ELICITED FREEZING 739

naloxone potentiation was at the mid-rangeshock intensities (.5 mA and 1.0 mA). At 2.0mA, freezing was at a ceiling, which reducedthe ability to detect a drug effect uponfreezing. The low shock intensity (.25 mA)seemed to be close to a threshold since itproduced considerable freezing in some an-imals and no freezing in others (range:0%-76% of the behavior was freezing at .25mA). Therefore the drug effect was not re-liable at .25 mA.

The increase in freezing was necessarily atthe expense of other behaviors. Analyses ofthe other behavioral categories showedsimilar patterns of results. General activity,F(4,100) = 20.31, p < .005, rearing, F(4,100)= 38.3, p < .005, locomotion, F(4,100) = 10,p < .005, and grooming, F(4,100) = 9.68, p< .005, all decreased with shock intensity.Naloxone's effect was evident as increasedsuppression of grooming, F(3,100) = 6.45, p< .005, locomotion, F(3,100) = 5, p < .005,and general activity, F(3, 100) = 2.77, p <.05. Rearing showed no drug effect, F(3,100) = 1.75,p >.10).

Naloxone's effect on freezing was presentonly in shocked animals. Figure 3 showsthat when the animals were not shocked, thedrug did not affect freezing. General ac-tivity, locomotion, and rearing were also not

affected by the drug when the animals werenot shocked (all Fs < 1).

It was surprising that grooming in non-shocked animals was suppressed by all dosesof the drug, F(3, 20) = 8.22, p < .005. Thatthese grooming behaviors were replaced bygeneral activity is indicated by a trend forincreased activity in the nonshocked-nal-oxone animals, F(3, 20) = 2.85, p < .10.This effect was not found in the first exper-iment. As grooming represented a smallpart of the animal's behavior in this situation(23% of the nonshocked-saline animal's be-havior) and was immediately suppressed toa very low level even by the low-intensityshock, it cannot account for naloxone's en-hancement of freezing. We are unable tohypothesize a mechanism for this effect ofnaloxone on grooming.

Another surprising effect of naloxone onnonshocked animals was that the highestdose significantly decreased defecation, F(3,20) = 3.71, p < .05. It is interesting thatnaloxone, a pure morphine antagonist, pro-duces an effect similar to morphine on def-ecation. This effect was not seen inshocked animals. Defecation increased withshock intensity, F(4, 100) = 11.05, p < .05,but no drug-related change in defecation wasseen in the shocked animals.

o•z.NLdUJcr

100

80

60-

40-

20-

0 J

T '

1 j t1 1 E

RM c1 K'u \

f-> ^C. ,_! | f |;H £j

'I

1111313 •LM.

0.5 3 8 0 .5 3 8

NO SHOCK .35 MA

I

llsjKV;':fi

T1^

T

T1

1

11;i|

i•Vi

T

I1

£•'

y-•:

s;

•"•'•i

1*>.

?f';'•?

;V;'o•

I

0 . 5 3 8 0.5 3 8 0 . 5 3 8

.5 MA 1 MA 2 MA

Figure 3. Freezing as a joint function of shock intensity and dose of naloxone.

740 MICHAEL S. FANSELOW AND ROBERT C. BOLLES

Experiment 3

We have shown that naloxone increasesthe amount of freezing in animals given twoor three shocks but not in unshocked controlanimals. One possibility suggested by thesefindings is that endorphins are maintainedat some basal level and that the effect ofnaloxone is merely to lower this level. Ac-cording to this interpretation, the endorphinsystem would serve the function of deter-mining pain thresholds (El-Sobky et al.,1976). An alternative possibility is that,rather than being maintained at some basallevel, endorphins are released by an initialshock and that they then tend to reduce theaversiveness of later shocks. If naloxoneblocks the effect of the initial endorphin re-lease, it would make later shocks moreaversive than they would otherwise be.

Experiment 3 was designed to test thesealternatives. If a painful event is necessaryto cause the release of endorphins, thennaloxone should be effective in potentiatingfreezing only when the animal is given atleast two shocks. If endorphins are exertinga constant influence, then naloxone shouldbe effective even if the animal is shockedonly once.

Method

Subjects and apparatus. Sixty-four rats, from thesame stock as in Experiment 1, were used. These ratsweighed between 210 and 280 g. As in the other ex-periments, the animals were run during the daylightportion of the 12:12 hr day/night cycle. The equipmentused was that of the first experiment.

Procedure. The procedure was an exact replicationof the first experiment except that the rats received 0,1, 2, or 3 shocks of 1-mA intensity and .75-sec duration.For the animals given multiple shocks, the intershockinterval was 20 sec. The behavior sampling procedurebegan after the last shock.

Thus, eight animals were randomly assigned to eachcell of a 2 X 4 factorial design, with the factors being thepresence or absence of naloxone (3 mg/kg) versus thenumber of shocks received.

Results

The data were scored in terms of the per-centage of freezing over the 8-min observa-tion period. These scores were subjected toa factorial analysis of variance. Freezingwas a positive function of the number of

3 - N

1 2 3 4 5 6 7 8

MINUTES

Figure 4. Percentage of freezing in groups of animalsgiven 0, 1, 2, or 3 shocks and injected with either nal-oxone (N) or saline (S).

shocks, F(3, 56) = 61.63, p < .005. Naloxoneanimals froze significantly more than salineanimals, F(l, 56) = 13.25, p < .005.

The interaction of the two main effectswas significant, F(3, 56) = 3.67, p < .025, andtherefore was analyzed by the following posthoc comparisons. The three-shock naloxoneand two-shock naloxone groups froze morethan their respective saline controls, ac-cording to the "conservative" Scheffemethod (p set at .01). The no-shock nal-oxone and one-shock naloxone groups didnot differ from their respective saline con-trols by either the "liberal" Fisher's test (F< 1) or individual t test (t < I). The data foreach minute of the observation period ispresented in Figure 4.

Grooming in the no-shock group was againfound to be suppressed by naloxone, t (14) =2.69, p < .02. The no-shock saline animalsgroomed 13% of the time while no-shocknaloxone animals groomed only 5% of thetime. No reliable differences were found inany of the other behavioral categories of theanimals receiving no shock. With one shock,where freezing was similar for naloxone andsaline animals, no effect on grooming wasfound.

Discussion.

Naloxone enhanced postshock freezing ifa rat received two or three shocks, but not

NALOXONE AND SHOCK-ELICITED FREEZING 741

one shock. This result supports the hy-pothesis that the first shock causes releaseof an endogenous analgesic, which in turnmodifies the aversiveness of the second andthird shocks. If some basal level or ongoingrelease of an endogenous analgesic were afactor in determining pain thresholds (e.g.,El-Sobky et al., 1976), then there shouldhave been an effect of naloxone in the one-shock groups. But this was not the case.

The present data are not consistent withthe hypothesis that triggering the endorphinsystem requires prolonged exposure tonoxious stimulation (Grevert & Goldstein,1977b; Jacob et al., 1974; Kokka & Fairhurst,1977). Rather, the results of the two-shockgroup suggest that a brief electric shockcauses release of an endogenous opiate whoseanalgesic activity is effective within 20 sec.This result is consistent with Fanselow's(1979) finding that rats preferred a placewhere a 20-sec signal predicted shock, asopposed to a place where shock was unsig-naled, but only if they were not pretreatedwith naloxone. The signal is hypothesizedto cause a conditioned release of an endoge-nous analgesic, which reduces the aversive-ness of the shock that follows.

That naloxone did not affect the behavior(with the exception of grooming) of animalsthat were not shocked, or those receivingonly one shock, indicates that the drug wasnot causing a nonspecific depression of ac-tivity.

The finding that freezing increased withnumber of shocks as well as with shock in-tensity further supports the contention thatfreezing reflects the aversiveness of the testsituation.

Experiment 4

We have suggested that the first shock, ina series of shocks, triggers the rat's endoge-nous analgesic system, which renders sub-sequent shocks less aversive. It should benoted that this interpretation is based pri-marily on the lack of effect of naloxone onanimals receiving only one shock, a null re-sult. Experiment 4 was designed to providemore conclusive evidence that naloxone doesnot affect freezing in animals that have re-ceived only a single shock. This was done by

30-

.25 .5

SHOCK (ma)

Figure 5. Freezing as a function of shock intensity inanimals given one shock after injection of either nal-oxone (solid line) or saline (broken line).

replicating the null result over a range ofshock intensities.

Method

Subjectx and apparatus. Fifty rats from the samestock as in Experiment 1 were run in the same apparatusas in Experiment 1. Unlike the first experiment, therats were handled for 1 min a day for 4 days before theexperiment.

Procedure. The procedure was an exact replicationof the one-shock groups of Experiment 3. However, fivedifferent shock intensities (.25, .5, 1.0, 1.3, or 2.0 mA)were used. The five shock intensities were combinedfactorially with an injection of either naloxone (3 mg/kg)or saline. The same behavior-sampling procedure wasalso used.

Results and Discussion

The results are presented in Figure 5.The effect of shock intensity on freezing wassignificant, F(4,40) = 3.86, p < .01, and lin-ear, F(l, 40) = 11.44, p < .005 (Gaito, 1965).Neither the drug's effect nor its interactionwith shock intensity was reliable (Fs < 1),which replicates the lack of effect of nalox-one on freezing over a range of shock inten-sities.

Experiment 5

In the previous experiment, an effectiveceiling for the contribution of shock intensityto freezing was reached at 1 mA. However,even at the three higher intensities theprobability of freezing with only a singleshock was rather low, only 23%. It should benoted that the effect of a drug on a particulardependent variable is often a function of the

742 MICHAEL S. FANSBLOW AND ROBERT C. BOLLES

expected rate of that dependent variable inthe absence of the drug (e.g., Dews & Wen-ger, 1977). The low rate at which freezingoccurs with a single shock may be a sufficientexplanation of the lack of effect of naloxonein the one-shock groups.

Experiment 5 was designed to test thispossibility. Parameters were selected thatwould produce similar rates of freezing inanimals receiving either one or two shocks.If naloxone enhanced freezing in rats re-ceiving two mild shocks but not in rats re-ceiving one intense shock, the rate-depen-dency (Dews & Wenger, 1977) explanationcould be ruled out.

Method

Subjects and apparatus. Thirty-two rats from thesame stock and weighing between 190 and 250 g wererun in the same apparatus as in Experiment 1. The ratswere handled daily for 1 min a day for 4 days before theexperiment.

Procedure. The procedure was similar to that of theother experiments. Half of the animals received nal-oxone (3 mg/kg) and the other half saline before beingplaced in the observation chamber. Two minutes afterplacement in the chamber, half of each drug group re-ceived a single 1.6-mA shock (.75 sec). The remaininganimals received two shocks of .4 mA (.75 sec) spaced20 sec apart. The behavior-sampling procedure de-scribed earlier began after the last shock and continuedfor 8 min.

Results and Discussion

The freezing scores were analyzed by atwo-factor analysis of variance; only the in-teraction of shock type and drug was signif-icant, F(l, 28) = 4.61, p < .05. The animalsthat received two mild shocks and naloxonefroze 33% of the time which is reliably morethan the other groups, F(l, 28) = 10.61, p <.005. The two-shock saline, one-shock sa-line, and one-shock naloxone animals did notdiffer; their mean freezing scores were 9%,11%, and 9%, respectively.

It should be noted that these rats froze lessthan rats that received a similar treatmentin the other experiments. This differencemay be due to a difference in the sample ofrats that served as subjects. The importantpoint is that despite the low overall rate offreezing and the similarity in the rate offreezing of the saline-treated one-shock and

two-shocks groups, naloxone potentiatedfreezing only in the two-shock group. Thislack of effect of naloxone on freezing in ani-mals receiving only one shock cannot be ex-plained solely by the low rate of freezingfound in the rats receiving only a singleshock.

Taken as a whole, Experiments 3, 4, and5 support the hypothesis that the initialshock in a series of shocks triggers the en-dorphin system, which reduces the aver-siveness of subsequent shocks. This inter-pretation would predict that following a se-ries of shocks, animals would not only beanalgesic but also have a higher level ofcentral nervous system endorphins, and thatis precisely what Madden, Akil, Patrick, &Barchas (1977) found. Although the datacannot rule out the possibility that somebasal level of endorphins determines painthresholds, such a threshold does not seemto contribute to the effect reported here.

Experiment 6

We have suggested that endorphin has itseffect by reducing the aversiveness of shockand that naloxone, by blocking this system,makes shock more painful. Grevert andGoldstein (1977a) found that naloxone af-fected the mood of human subjects who hadbeen exposed to a painful tourniquet but didnot affect the rating of the pain itself. Theseauthors suggested that endorphins may haveantianxiety properties. These antianxietyproperties may have reduced fear and hencefreezing during the observation period.Therefore, naloxone might produce its effectby antagonizing endorphins' antianxietyproperties. A single shock may not havebeen sufficient to trigger the endorphinsystem, or it might not have producedenough fear to be modified by endorphins.This is a possible account of why there wasno effect of naloxone in the groups receivingonly one shock.

If endorphin exerts its effect on fear ratherthan on pain, then naloxone need only bepresent during the observation period to beeffective. However, if naloxone influencesthe perception of the nocioceptive propertiesof shock, then it must be present at the timeshock is experienced in order to have its ef-

NALOXONE AND SHOCK-ELICITED FREEZING 743

feet. The sixth experiment tests these al-ternatives.

Method

Subjects. Twenty-seven rats, from the same stockas in the first experiment, were used. These animalswere approximately 77 days old and weighed between200 and 260 g. They were housed in conditions iden-tical to those of the first experiment and were also runduring the light portion of the day/night cycle. Eachrat was handled for 2 min a day for the 3 days precedingthe experimental day.

Apparatus. The apparatus was the same as in Ex-periment 1.

Procedure. The procedure was an exact replicationof the three-shock group of Experiment 1 with the fol-lowing exceptions: Subjects were randomly assignednine to a group. The animals in two of the groups re-ceived saline injections before being placed in the ob-servation chamber. Immediately after the third shockthe animals were removed from the box and given asecond injection. One of the groups (S-N) received a3 mg/kg injection of naloxone, and the other (S-S) re-ceived a second saline injection.

The remaining group (N-S) of animals received analoxone injection before placement in the observationchamber. After the third shock they were removed andgiven a saline injection. Immediately after receivingthe second injection, all the animals were returned tothe observation chamber for an 8-min observation pe-riod. A blind behavior-sampling procedure, like thatof Experiment 1, was used.

Results and Discussion

The animals injected with naloxone beforeshock (N-S) froze 84% of the time in the ob-servation period. That level was signifi-cantly higher than those of the S-S and S-Ngroups, F(2, 24) = 6.33, p < .025, which froze59% and 58% of the time, respectively. Thedata for each minute of the observation pe-riod is presented in Figure 6.

The lack of difference between the S-Sand the S-N groups indicates that naloxonemust be present during the shock and notjust the observation period in order to beeffective in increasing the amount of post-shock freezing. Naloxone was evidently notproducing its effect by acting upon the fearpresent during the observation period.

In the previous experiments, naloxone wasfound to have no effect on freezing whenanimals were given only one or no shock.This was taken as evidence that freezing wasnot being produced by any direct action ofthe drug or that naloxone was nonspecifically

NlLJLJ

MINUTES

Figure 6. Freezing in animals given naloxone onlybefore receiving shock (N-S) or only before being ob-served (S-N), or given no naloxone (S-S).

suppressing activity. However, there waslittle or no freezing in these groups of ani-mals, and this effect of naloxone might nothave been detectable due to a floor effect.In the present study, naloxone had no effectwhen given after shock on animals that werefreezing at a high level. This suggests thatnaloxone does not directly increase freezingor nonspecifically suppress activity in thissituation.

General Discussion

The freezing behavior of rats was shownto be representative of the aversiveness ofthe situation, and naloxone pretreatmentwas found to increase the amount of post-shock freezing. This effect cannot be due tosuppression of motor activity or a direct ac-tion of the drug as naloxone did not affectfreezing in rats that were given only oneshock, or those not shocked at all, or thosegiven naloxone following exposure toshock.

That naloxone is effective only if presentduring the time shock is received suggeststhat the drug affects the rat's perception ofthe shock—the rat behaves as if the shockwere more intense or more painful. Thatnaloxone produces this effect by blocking theendogenous analgesic activity of endorphinsis most consistent with the known pharma-cological actions of the drug (Lewis, Bently,& Cowan, 1971; Pert & Snyder, 1973).

744 MICHAEL S. FANSELOW AND ROBERT C. BOLLES

The lack of a naloxone effect when ratswere given only one shock suggests that theendorphin system is not continually active.Rather, it seems that a first shock is neces-sary to trigger the system, resulting in analoxone effect upon the perception of sub-sequent shocks.

In conclusion, it appears that when ex-posed to a painful event (Bodnar, Kelley, &Glusman, 1978; Madden et al., 1977; thepresent study) or a signal that predicts apainful event (Fanselow, 1979), rats releasean endogenous analgesic capable of reducingthe painfulness of subsequent aversivestimuli.

References

Anisman, H., & Waller, T. G. Footshock producedexcitation and inhibition of activity in rats. AnimalLearning and Behavior, 1973,1, 93-95.

Belluzzi, J. D., Grant, N., Garsky, V., Sarantakis, D.,Wise, C. D., & Stein, L. Analgesia induced in vivoby central administration of enkephalin in rat.Nature, 1976,260, 625-626.

Blanchard, R. J., & Blanchard, D. C. Crouching as anindex of fear. Journal of Comparative and Physi-ological Psychology, 1969, 67, 370-375.

Bodnar, R. J., Kelley, D. D., & Glusman, M. Stress-induced analgesia: Time course of pain reflex al-terations following cold water swims. Bulletin of thePsychonomic Society, 1978,11, 333-336..

Bolles, R. C., & Collier, A. C. The effect of predictivecues on freezing in rats. Animal Learning and Be-havior. 1976, 4,6-8.

Campbell, B. A., & Masterson, F. A. The psychophysicsof punishment. In B. A. Campbell & R. M. Church(Eds.), Punishment and aversive behavior. NewYork: Appleton-Century-Crofts, 1969.

Dews, P. B., & Wenger, G. R. Rate-dependency of thebehavioral effects of amphetamine. In T. Thompson& P. B. Dews (Eds.), Advances in behavioral phar-macology (Vol. 1). New York: Academic Press,1977.

El-Sobky, A., Dostrovsky, J. O., & Wall, P. D. Lack ofeffect of naloxone on pain perception in humans.Nature, 1976, 263, 783-784.

Fanselow, M. S. Naloxone attenuates rats' preferencefor signaled shock. Physiological Psychology, 1979,

7, 70-74.Frederickson, R. C. A., Burgis, V., & Edwards, J. D.

Hyperalgesia induced by naloxone follows diurnalrhythm in responsivity to painful stimuli. Science,1977, 198, 756-758.

Gaito, J. Unequal intervals and unequal n in trendanalyses. Psychological Bulletin, 1965, 63, 125-127.

Goldstein, A., Pryor, G. T., Otis, L. S., & Larsen, F. Onthe role of endogenous opioid peptides: Failure ofnaloxone to influence shock escape threshold in therat. Life Sciences, 1976,18, 599-604.

Grevert, P., & Goldstein, A. Effects of naloxone onexperimentally induced ischemic pain and on moodin human subjects. Proceedings of the NationalAcademy of Sciences of the United States ofAmerica, 1977, 74, 1291-1294. (a)

Grevert, P., & Goldstein, A. Some effects of naloxoneon behavior in the mouse. Psychopharmacology,1977,53,111-113. (b)

Hughes, J. Isolation of an endogenous compound fromthe brain with pharmacological properties similar tomorphine. Brain Research, 1975, 88, 295-308.

Jacob, J. J., Tremblay, E. C., & Colombel, M. C. Fa-cilitation de reactions nociceptives par la naloxonechez la souris et chez le rat. Psychopharmacologica,1974, 37, 217-223.

Kokka, N., & Fairhurst, A. S. Naloxone enhancementof acetic acid-induced writhing in rats. Life Sci-ences, 1977, 21, 975-980.

Lewis, J. W., Bentley, K. W., & Cowan, A. Narcoticanalgesics and antagonists. Annual Review ofPharmacology, 1971,11, 241-270.

Loh, H. H., Tseng, L. F., Wei, E., & Li, C. H. ^-En-dorphin is a potent analgesic agent. Proceedings ofthe National Academy of Sciences of the UnitedStates of America, 1976, 73, 2895-2898.

Madden, J., Akil, H., Patrick, R. L., & Barchas, J. D.Stress-induced parallel changes in central opioidlevels and pain responsiveness in the rat. Nature,1977,265, 358-360.

Malick, J. B., & Goldstein, J. M. Analgesic activity ofenkephalins following intracerebral administrationin the rat. Life Sciences, 1977,20, 827-832.

Pert, C. B., & Snyder, S. H. Opiate receptor: Dem-onstration in nervous tissue. Science, 1973, 179,1011-1014.

Walker, J. M., Berntson, G. G., Sandman, C. A., Coy, D.H., Schally, A. V., & Kastin, A. J. An analog of en-kephalin having prolonged opiate-like effects in vivo.Science, 1977, 796, 85-87.

Received October 16,1978 •