.A N 1 FORCIS RAIOIOLI

Best Available Copy INCIENTC REPORT

SR86-13,\'euri-scwnc & Sirhath,,nid R',iu,,,. Vol. I0. pp. 55--!,5. 19W. Ankhk Intcmtjional Inc. Pnnlcd in the U.S.A. 0149.141 M$1.00 .00

S DTICMechanisms of Radiation-Induced AELECTE

Ln ~APR2 118Conditioned Taste Aversion Learning Rto(

BERNARD M. RABIN*t AND WALTER A. HUNT " '

*Bhvior,,I S(i('ces I)cpartm,nnt. Arined Forcet RaIifioloy Research ItstituteBcthesda. %11) 20814-5145

U and el)partm'nt of Psv('hoo,.'. Univcr.sitv oJf.hIar\'and Baltimore ('ountv. Catiu sville. AD 21228

Received 25 March 1985

RABIN. B. NI. ANI) W. A HUNT. thWi ianim ,/! radiai,i -inti cd i onditiotm'd ta.te tiver.sio, h'irnint,. NEUROSCIBIOBEHAV REV I111 55-h,, 1986.-The literature on taste aiersion learning is reviecd and discussed. with particu-lar emphasis on those studies that have used exposure to ionizing radiation as an unconditioned stimulus to produce aconditioned taste aversion. The primary aim of the review is to attempt to define the mechanisms that lead to the initiationof the taste aversion response follosing exposure to ionizing radiation. Studies using drug treatments to produce a tasteaversion have been included to the etent that they are relevant to understanding the mechanisms by which exposure toionizing radiation can affect the behavior of the organism.

Ionizing radiation Conditioned taste aversion Lithium chloride, Area postrema , Behavior, .

THE toxicity of ionizing radiation generally is expressed as onizing radiation may have subtle actions on the brain thatthe loss of rapidly turning over cells. This results from an may have clinical significance.impairment of the synthesis of these cells, especially those in In addition to emesis. exposure to lower doses of ionizingbone marrow. However. since neurons in the mature central radiation induces a conditioned taste aversion (ETA) in cx-nervous system do not undergo cellular turnover, the brain perimental animals and humans. A CTA can result when agenerally has been believed to be insensitive to exposure to novel tasting solution is paired with a toxin, so that when theionizing radiation 1711, thereby allowing for the routine use solution is presented subsequently. further ingestion isof radiation in treating brain tumors. However, little atten- avoided. This avoidance behavior is typically acquired aftertion has been given to evidence appearing over the years that a single pairing of the solution with the toxin. Taste aver-raises doubts about this conclusion. For example, exposure sions can be produced. using saccharin or sucrose solutions.to ionizing radiation at doses of 15-500 rad significantly re- after injection of a wide variety of drugs., such as lithiumduces the electroshock seizure threshold, an effect lasting up chloride (LiCI). amphetamine, copper sulfate, and apomor-to 8 months [971. Also. in this dose range emesis is induced. phine [44.1131. as well as after exposure to ionizing radiationTransient increases in the spontaneous locomotor activity of 11241. Since radiation-induced taste aversions are obtainedC57BL mice and arousal in rats have been observed after typically after exposure to 37-150 rad doses., examining theexposure to 1000-1500 rad 160.831. And. on the molecular underlying mechanisms associated with development of thislevel, the functioning of sodium channels is impaired after type of CTA might provide information on other possibleradiation exposure to as little as 10 rad [ 1401. These observa- behavioral consequences of low-dose irradiation. Such ations suggest that exposure to ionizing radiation may have model has the additional benefit of having an extensive be-behavioral consequences not previously appreciated. havioral and physiological database relating to taste aver-

High lethal doses of ionizing radiation can severely dis- sions in general 'see [I. 47. 1171 for recent reviews of drug-rupt behavior. Depending upon the species, the quality of induced CTA learning).radiation, and the nature of the behavioral measurements. The goal of this paper is to review the variety of studies1000-10000 rad degrades performance on a number of tasks. that have used radiation to produce a CTA in order to defineUnder some conditions the effects are transient, generally the general mechanisms by which exposure to nonlethallasting up to one hr after irradiation (7. 14. 15. 36. 641. but levels of radiation might alter behavior. Our purpose is not to

C.) after sufficiently high doses of radiation a permanent in- review CTA learning in general. However, studies withcapacitation is induced, culminating in the death of the or- toxins in addition to radiation will be presented when they

1t ganism. Possible effects of lower doses (< 1000 rad) of radia- are relevant to the discussion.J tion have not been adequately explored. Therefore. given the A number of methods are used to produce a CTA. Since

LA available evidence, such as. that cited above, exposure to interpretation of information derived from some procedures

F 'Requests for reprints should be addressed to Walter A. Hunt.

86 4 21 144

. J Jr -.W U *' = 7 . 7 .: M: W. .,WLW;i , .-, -' .'P. %k T

.- - ' .: 'W ; .a --,.-" -. t .:-.

,

56 RABIN AND HUNT

can be misleading, we will begin with a brief discussion of injection by varying the number of LiCI preexposures tothe various methods employed for producing a CTA and which the animals are subjected [1301. This observationtheir advantages and disadvantages. suggests that CTA acquisition is not dependent upon the

LiCI-induced changes in pituitary/adrenal function. Consis-METHDOLGICL COSIDRATONStent with this finding is the report by Rabin et al. 1 1041 that

METHODOLtOGI'AL CONSIDERATIONS hypophysectomy has no effect on the acquisition of an LiCI-

, The general procedure for producing a conditioned taste induced CTA in rats tested using a two-bottle procedure.,' aversion involves placing the subjects on a water deprivation Similar findings have been reported with radiation-

schedule. Once the animal has adapted to the schedule, a induced taste aversion learning. Garcia and Kimeldorf 1401novel tasting solution is substituted for the regular water working with non-deprived rats in a single-bottle design re-bottle as the conditioned stimulus (CS). Ingestion of the ported that hypophysectomy had no effect on the acquisitionnovel solution is followed by treatment with the toxin, the of a radiation-induced CTA in rats. Similarly. Ranin et a/.unconditioned stimulus (UCS). The dependent measure, the 11041 found that hypophysectomy had no effect on radiation-

conditioned response (CR). is the amount of the solution induced CTA learning in rats tested with a two-bottle proce-ingested at a subsequent presentation of the CS. dure. In contrast. Cairnie and Leach [161, using a single-

• Within this general paradigm, a number of variations are bottle test with del rived animals, reported that injections ofused in quantifying the dependent measure, although the dexamethasone produced a significant attenuation of asuggestion has been made that these different procedures radiation-induced CTA in rats. However. using a two-bottlemay be measuring different aspects of CTA learning 11311. test Rabin et al. (unpublished manuscript) found no effect of

• One such variation involves the use of either a one-bottle or the drug. Since hormones of the pituitary/adrenal system, two-bottle test to study taste aversion learning. In a single- have been shown to play a role in mediating conflict-induced

bottle test, the animal is presented with only a single bottle arousal 155.1381. these results suggest that activation of thecontaining the CS during the period in which fluid is avail- pituitaryladrenal system or manipulation of corticosteroidable. In a two-bottle test. the animal can choose between the levels in a CTA paradigm is related more to conflictCS or a neutral fluid such as tap water. In a single-bottle test produced using a single-bottle experimental design than tothe subject is forced to choose between ingesting the solution the CTA learning itself.

, that has been paired with the UCS or "'going thirsty": A second potential confounding artifact in studies at-whereas with a two-bottle test. an alternative is available to tempting to evaluate the mechanisms involved in mediatingdrinking the CS. CTA learring may result from a failure to control for the

A second variation in the dependent measure involves effects of state-dependent learning 1941. State-dependentdetermining either the absolute or the relative intake. In a learning results from a failure to maintain identical condi-single-oottle test. the most typical dependent measure is the tions (states) on both the conditioning and test days. Conse-actual amount of the CS ingested by the subject, although quently. a CTA that is learned on the conditioning day is notpostconditioning intake may also be expressed as a percent- recalled on the subsequent test day because the test condi-age of CS intake on the conditioning day prior to treatment. tions are different than the conditions under which the re-When a two-bottle design is used. the data are typically pre- sponse was originally learned. In experiments designed to

" sented in terms of the relative intake of both the neutral and evaluate the effects of a variety of potentially disruptiveconditioned stimuli. While the data from a two-bottle test treatments on CTA acquisition the usual procedure is to in-

* may be presented in terms of the actual intakes of each solu- terpolate some treatment during the interval between thetion. the data can also be presented as a sucrose preference presentation of the CS and UCS on the conditioning day.score, calculated by dividing the CS intake by total fluid Under these conditions, the subjects are exposed to the CSintake. Because this preference measure is relatively insen- in the same treatment-free environment on both conditioningsitive to variations in total fluid intake, the preference score and test days. Some experimental manipulatio, -.. however.has been considered to be a more sensitive indicator of CTA may require that the disruptive treatment be administeredlearning 131.531. particularly under conditions in which the prior to. or coincident with. the presentation of the CS on thetaste aversion produced by the UCS is relatively weak. conditioning day. If the identical treatment is not subse-

Using a single-bottle design may also introduce a potential quently administered together with the presentation of theconflict that can result from forcing the animal to ingest the CS on the test day, then the different conditions on the twosolution whi.-il has been paired with the UCS. The only other days can interfere with the recall of the previously acquired

S ." alternative for the animal is no' to drink at all. Conflict is not CTA response. Under these conditions, identical treatmentsinvolved with a two-bottle design because the subject can must be administered on both conditioning and test days tochoose which solution to drink. The introduction of conflict maintain identical states on both days.into the CTA learning paradigm can complicate the interpre- That state-dependent learning can present a serious prob-tation of data relating to possible mechanisms underlying lem for studies designed to analyze the mechanisms of CTACTA acquisition. This is a significant problem is studies that learning has been shown by a number of studies using a

-. have attempted to assess a possible role of pituitary/adrenal variety of unconditioned stimuli and disruptive treatments.hormones in CTA learning. In a series of studies using Working with ionizing radiation. Levy e, al. 1751 reportedsingle-bottle designs. Smotherman ta i. 1128.1291 have re- that treatment of rats with an antihistamine (chlor-ported that acquisition of an LiCI-induced CTA is associated pheniramine maleate) on the conditioning day attenuated thewith the activation of the pituitary/adrenal system. Manipu- acquisition of a CTA to saccharin. However, they adminis-lation of corticosteroid levels by injection of ACTH or dex- tered the antihistamine prior to the initial exposure of the CSamethasone produces corresponding changes in the acqui- and failed to give a corresponding treatment on the test day.sition or extinction of the CTA response [56. 57, 1271. How- When the antihistamine is administered together with the CSever, more recent work has shown that it is possible to dis- on both conditioning and test days, treatment with the

-'C sociate the corticosteroid and behavioral responses to LiCI antihistamine has no effect on the acquisition of a radiation-'.

7m.

U

RADIATION-INDUCED TASTE AVERSIONS 57

induced CTA 11021. Alternatively, when the antihistamine is [21. A radiation-induced CTA can be acquired when thegiven after the presentation of the CS. so that the animals are radiation UCS is presented up to 6 hr preceding the presen-in a drug-free state on both conditioning and test days. tation of the CS. The strength of the aversion seems to be theneither intraperitoneal [16, 102. 1191 nor intraventricular greatest when the UCS precedes the CS y 90 min [191. LiCI.11011 injections of antihistamine have an effect on the acqu;- in contrast, is not capable of inducing a CTA under thesesition of a radiation-induced CTA. conditions. These observations indicate that exposing the

In addition to the use of a radiation UCS, state-dependent organism to ionizing radiation, unlike treatment with LiCI.effects have also been observed in studies of drug-induced causes a long-lasting change that produces the temporalCTA learning. Phillips and LePiane 1961 have reported that overlap necessary for the acquisition of a CR.the disruption of LiCI-induced CTA learning produced hy A related issue concerns the dose equivalence of differentstimulation of the basolateral amygdala is not observed when unconditioned stimuli when comparing their mechanisms ofthat stimulation is given on both the conditioning and test action. When a given treatment or manipulation shows anday. Similarly, in a pilot experiment. Rabin and Hunt (un- equivalent effectiveness in modulating a CTA produced bypublished manuscript) found that the reported disruption of different unconditioned stimuli, the empirical observation is.an amphetamine-induced CTA by pretreatment with alpha- by itself, sufficient evidence for a similarity of mechanism asmethyltyrosine 1501 is greatly attenuated if the treatment se- it relates to those specific experimental conditions. In con-quence is administered prior to the presentation of the CS on trast. the observation that an experimental manipulation isboth the conditioning and the test day. It would therefore effective in modulating a CTA produced by one UCS but notseem that the prob!ems posed by state-der'-dent learning by another UCS may mean either that the mechanisms ofare general. affecting a variety of unconditioned stimuli and action of these stimuli are different, or that there are differ-disruptive treatments. As a result, appropriate controls fo," ences in the initial effectiveness of those stimuli in producingstate-dependent effec',s must be employed where a disrup- a CTA. In part. the validity of the two hypotheses can betive treatment is administered prior to or during the presen- determined by comparing the experimental dose of thetation of the CS CTA-inducing agents with the threshold doses of those

agents. In general. a given manipulation is less likely to beTHE RAI)IATION-INDUCED CTA effective in modulating a CTA the greater the experimental

dose is above the threshold dose (e.g.. [35.1101). In addition.the two alternatives can be evaluated by comparing the ef-

Ionizing radiation is only one of many potential uncon- fects of the manipulations on a variety of response measures.ditioned stimtli which can be used to produce a CTA [44. Thus. for example. LiCI- and amphetamine-induced aver-113. 1241. Therefore. it would be important to review some sions are different not only in terms of the effects of areaof the parametric factors associated with the acquisition of postrema lesions on the acquisition of the response, but alsotaste aversions produced by exposure to ionizing radiation or in terms of the non-consummatory responses elicited by theby injection of chemical comptunds. A consideration of two stimuli 195.1141.some of the similarities and differences in taste aversionlearning produced by these various unconditioned stimuliwill facilitate a comparison of the mechanisms involved in Ruh. qIlhm'xxthe acquisition of a CTA. One of the major unknowns in the acquisition of a

As indicated above, the dose of radiation used to produce radiation-induced CTA concerns the nature of the UCS.CTA learning is well below the lethal dose 145.1241. What are the specific characteristics of the UCS that lead toThreshold doses for whole body exposures in rats have been the subsequent avoidance of a -food- with which the UCSvariously reported to range from 7.5 rad 1451 to between has been associated? With injections of toxic unconditioned2(--25 rad 11081. These variations probably reflect differences stimuli that produce obvious signs of illness in the organismin the radiation quality and in the conditions associated with such as LiCI or copper sulfate, for example, a direct relation-the behavioral testing 11241. Further increases in the dose ship is assumed between the gastrointestinal effects of theproduce a non-linear increase in the strength of the aversion UCS and the resultant CTA learning 122. 23. 1181. Consistent11081. such that whole body exposures of 5(0-IM rad produce with this hypothesis is the observation that lesions of thea nearly total avoidance of the CS (e.g.. 145. 124. 1261). area postrema. the chemoreceptive trigger zone for emesis

Similar results are obtained with LiCI. The threshold dose 191. disrupt the acquisition of a CTA produced by injectionsfor an LiCl-induced aversion is between 0.15 and 0.30 of methylscopolamine 14.1141 and by injections of LiCI

* mEq kg 191, 1081. Further increases in the dose of LiCI also 1103.1141. That the acquisition ofa CTA requires a relativelyproduce a non-linear increase in the strength of the aversion, specific UCS is shown by the observation that pairing awith the maximum aversion being reached with doses of 3 novel CS with a painful exteroceptive UCS. such as shock.imEq/kg 1911. does not readily result in CTA learning 141.731. This obser-

• Amphetamine. in contrast to both radiation and LiCI. vation which shows that pain per se is not an effective UCSdoes not show this dose/response relationship. The threshold for CTA learning suggests that the UCS may in some way

% dose for amphetamine-induced CTA is between 0.5 and 1.0 involve the gastrointestinal system. However. even assum-mg/kg. but further increases in dose up to 2.0 mg/kg do not ing. for the moment, that the toxic character of some stimuliproduce a corresponding increase in the strength of the constitutes the direct antecedent condition for CTA learning.Saversion 11081. a CTA can also be produced by pairing a novel CS with a

Despite the similar dose/response relationships between variety of unconditioned stimuli (such as amphetamine, can-, radiation and LiCI. there are significant differences in time- nabinols. sesame oil and morphine) that are not only non-

courses of the conditioning effects. This is primarily re- toxic, but will also be self-administered 13. 28. 38. 491. Thisflected in the capacity of the radiation UCS to produce a would suggest that the toxicity of a UCS is not a necessaryCTA when presented in a backwards conditioning paradigm condition for CTA learning to occur.

_i

58 RABIN AND HUNT

Whether the toxicity of radiation is important for the de- treatment on CTA acquisition, there were also no effects onvelopment of a CTA is not clear. As indicated in the preced- the recall of a previously acquired radiation- or LiCI-induceding section, high doses of radiation (>500 rad) are clearly CTA. In this part of the experiment the procedures usedtoxic stimuli which can produce a variety of changes in were similar to those of Coil et a. [231, which would suggestneural [67. i12, 136, 1411 and behavioral [30. 82. 89. 901 that any possible effect of antiemetic treatment on CTAfunctioning. In addition, like both LiCI and copper sulfate, learning is marginal at best. Consistent with this interpreta-irradiation can lead to emesis 18.121 and. in humans, to tion is the observation that. in humans, nausea is not a nec-nausea 11201. In contrast, the dose of radiation (37-150 rad) essary symptom in subjects that acquire a CTA [51. There-typically used to produce a CTA. has no apparent effect on fore. the data do not provide compelling evidence that athe unrestrained behavior of the organism [1241. However. UCS-induced illness is a direct antecedent condition, to CTAlesions of the area postrema attenuate a radiation-induced learning.CTA just as they attenuate a CTA following an injection of The failure to find a consistent effect of antiemetic treat-LiCI 193.1031. This might suggest that a radiation-induced ment on CTA learning produced by a toxic UCS raises somechange, possibly related to emesis, is a factor in the acquisi- questions about the role of a UCS-produced illness as antion of a radiation-induced CTA. and that. therefore, there is antecedent condition to the acquisition of a radiation-some relationship between the toxic nature of radiation and induced CTA. Possibly the toxicity of the UCS is a side-the capacity of exposure to ionizing radiation to lead to CTA effect of the treatme-' that is not required for the acquisitionlearning, of the CTA. Some auditional evidence is concordant with

If an effect on the gastrointestinal system related to this possibility. Several sudies have provided evidenceemesis does play a role in the acquisition of a CTA when a which indicates that pairing a novel CS with the illnesstoxic UCS is used. then it should be possible to disrupt the produced by various poisons is not, by itself, sufficient toCTA by treating the organisms with an antiemetic to prevent produce a CT 166.921. Similarly, both radiation- and LiCI-the development of the malaise. Initial experiments did not induced taste aversions can be acquired by rats that are ex-support the hypothesis of a link between the antiemetic- posed to the UCS while under deep surgical anesthesia andsensitive gastrointestinal distress and CTA learning. levy i, maintained under the anesthesia for an additional 4-8 hr(1. 1751 reported that pretreating rats with the antiemetic 1100.1161. Under these conditions, it is difficult to under-trimethobenzamide did not disrupt the acquisition of a stand how the cxperience of a possible UCS-induced illnessradiation-induced CTA. and Gadusek and Kalat [371 re- could contribute to the acquisition of the CTA.ported that treatment with scopolamine did not attenuate therecall of a previously acquired LiCI-induced CTA. However. R4h, qfthe Ga.triietinal SY. ('1reasoning that these previous investigations used doses ofaniiemetics that were much higher than the clinically effec- Even though the data on radiation presented abovelive doses. Coil vi al. 1231 tested the effects of several doses suggest the possibility of a subtle effect on nervous systemof a variety of antiemetics on the recall of it LiCI-induced function following exposure to low doses of ionizing radia-CTA. They found that treatment with the clinically effective tion. these radiation-induced changes in nervous system ac-doses of scopolamine. trimethobenzamide. prochlorperazine tivity do not seem to be critical for the occurrence of CTAand cyclizine produced a significant attenuation of the previ- learning 1611. Garcia and Kimeldorf [401 reported that radia-ously acquired CTA. Doses of antiemetics either greater or tion exposure restricted to the abdomen of rats could serveless than !he clinically effective dose had no effect on CTA as the UCS for CTA learning. Although a higher dose ofrecall. These results would be consistent with the hypothesis abdomen-only radiation is needed to produce an aversionthat a UCS-induced illness plays a role in CTA learning when equivalent to that produced by whole-body exposure, ir-LiCI is used as the UCS. radiation restricted to the abdomen is much more effective in

More recent research however, has failed to confirm producing a CTA than is irradiation restricted to the head.these findings. Goudie ct al. 1511. working with taste aver- pelvis, or thorax alone. These basic results have been rep-sions produced by ait variety of diugs in:luding LiCI. reported licatLd in a number of investigations that. despite utiizing athat treatment with the clinically effective doses of variety of ptoced!tres. all show that body-only radiation ex-scopolamine or prochlorperazine had no effect on the recall posure will produce a stronger aversion and at a lower doseof it previously acquired CTA. In a more detailed study, than will head-only irradiations 1106.1261. Since exposure ofRabin and Hunt 1991 looked at the effect of antiemetic treat- pelvis or thorax would he expected to affect the spinal cordment on both the acquisition and recall of taste aversions and peripheral uerves to a similar extent as irradiation of theproduced by injection of LiCI or by exposure to ionizing abdomen, the greater elTectiveness of exposure of the abdo-radiation. Treatment with the previously reportedeffective men in producing a CTA would be consistent with the by-doses of trimethobenzamide. prochlorperazine or cyclizine pothesis that some effect of the radiation related to the gas-1231 did not attenuate The acquisition of either a radiation- or trointestinal system may be involved in mediating the acqui-liCI-induced CTA. Because radiation exposure can produce sition of the radiation-induced CTA.a CTA when administered up to 6 hr prior to ingestion of the There are two possible mechanisms by which exposure ofCS 121 and because the maximal effects of irradiation on the abdomen can lead to taste CTA learning. First. the radi-CTA learning are observed 90 min after exposure 1191. addi- altion may have a direct effect on the activity of the gastroin-tional groups of rats were included in this study that were testinal system. Second, it may act as it nonspecific toxingiven -one injection of prochlorperazine 15 min prior to it- causing the release of some humoral factor related to a gen-radiation or to injection of LiCI and t second antiemetic eralized irritation of the gastrointestinal system. In supporttreatment 3 hr later. This extended antiemetic treatment, of the first possibility. Hulse and Mizon 1581 have reportedwhich would be expected to overlap the period of radiation that exposing the abdomen of rats to ionizing radiation ateffects, also had no effect on the acquisition of a CTA. As doses of 20-IN) rad produces a delay in gastric emptyingwould be expected, given the lack of effect of antiemetic which is correlated with the strength of an a~ersion to a

RADIATION-INDUCED TASTE AVERSIONS 59

barium meal. Exposing the head and shoulders of rats to member of the parabiotic pair varies as a function of the doseradiation doses of up to 200 rad does not have an effect on to which the unshielded member is exposed [63].gastric emptying and produces only a weak aversion, at best. Additional support for the hypothesis that a radiation-to the barium meal. However, while treating rats with insulin released humoral factor which serves to mediate CTA learn-reverses the radiation-induced delay in gastric emptying [59]. ing following exposure to ionizing radiation comes from theit does not disrupt the acquisition of a radiation-induced observation that a CTA can be produced in rats by injectingCTA [161. them with serum from rats that have been previously ex-

A second approach to evaluating a possible role for direct pcsed to radiation [431. However, caution must be used ingastrointestinal effects in CTA learning has involved section- interpreting the results of this experiment because the donoring the vagus nerve. The vagus constitutes the most exten- rats were subjected to extreme doses of radiation (30000 rad)sive afferent source from the gut to the central nervous sys- and because the recipient rats were given injections of 12 mltem [88] and has been implicated in the regulation of a wide of serum in 3 injections separated by 15 min each on 4 sepa-range of visceral homeostatic functions [79. 80. 137]. If dis- rate experimental days. This procedure is so different fromruption of gastric function is the UCS for the acquisition of a the standard procedures for producing a CTA that the degreeradiation-induced CTA and this information is relayed to the to which we can generalize from these results to the morebrain by neural pathways from the stomach, then sectioning typical experiment is not certain.the vagus might be expected to produce some changes in the Another line of evidence that is consistent with the hy-acquisition of the CTA [22.801. However. Rabin, Hunt and pothesis ofhumoral mediation of the radiation-induced CTALee 11051 found that subdiaphragmatic vagotomy in rats did comei from studies of the effect of area postrema lesions onnot disrupt the acquisition of a CTA following exposure to CTA learning. The area postrema is the chemoreceptive trig-either 200 rad whole-body or body-only irradiation. The re- ger zone for emesis which functions to monitor the blood andsuits obtained with taste aversions produced by exposure to cerebrospinal fluid for toxins [9]. Lesions of the areaionizing radiation generally parallel those obtained with postrema disrupt the emetic response to a variety of blood-drugs. In general, vagotomy has no effect on the acquisition borne toxins [91 as well as to intraventricular histamine [6].of taste aversions induced by systemic toxins such as Similarly, lesions of the area postrema disrupt the emeticapomorphine. LiCI or ethanol 168, 69. 80. 1051. While these response to ionizing radiation in monkeys [121 and in dogsstudies do not eliminate the possibility that the relevant in- 1201, suggesting that a humoral factor mediates the emeticformation for CTA learning may be carried from the gut to response to irradiation. This conclusion is limited by thethe central nervous system by alternate pathways (e.g.. observation that area postrema lesions have been reported tosplanchnic), it does not seem likely that a radiation-induced have no effect on emesis produced by whole body irradiationchange in gastric function mediated by the vagus nerve, in the cat [8]. However, more recent work indicates that areawhich provides the most extensive afferent source, serves as postrema lesions are effective in disrupting emesis in catsthe proximate UCS for CTA learning following exposure to exposed to body-only radiation (Rabin. Hunt. Chedester andradiation. Lee. in preparation).

In contrast, the emetic response to gastric irritation As with blood-borne emetic stimuli, lesions of the areaproduced by intragastrically administered copper sulfate is postrema disrupt CTA acquisition to systemically-greatly attenuated by vagotomy 11371. Concordant with this administered drugs. Area postrema lesions have been shownfinding is the report that subdiaphragmatic vagotomy in rats to attenuate CTA learning following intraperitoneal injectionsdisrupts CTA acquisition following intragastric and intraperi- of methylscopolamine [41, LiCI [103.1141 and histaminetoneal copper sulfate, but not following intravenous adminis- 11031. as well as fllowing an intravenous injection of coppertration 1241. While this finding would suggest that vabally- sulfate [221. The area postrema thus seems to mediate CTAmediated changes in gastrointestinal function can play a role learning for a particular class of unconditioned stimuli thatin CTA learning. Rabin et al. [ 105.1091 have been unable to are apparently related in some manner to emesis because thereplicate this finding. Although more work is needed to re- area postrema lesions do not disrupt the CTA produced bysolve this discrepancy, the evidence suggesting a direct gas- injections of amphetamine [4,114]. A similar distinctior be-trointestinal involvement in CTA learning is not compelling. tween LiCI and amphetamine as representing different

classes of unconditioned stimuli for CTA learning has beenmade by Parker 1951 based upon the observation of different

Roh, of lluloral and Netural Mthlani,%nis nonconsummatory behavioral responses produced by treat-ment with these stimuli.

In contrast to the above, there is good support for the It therefore seems that there are two classes of UCS forhypothesis that exposure to ionizing radiation causes the re- CTA learning: one mediated by the area postrema and thelease ofa humoral factor, which in turn serves to mediate the other which does not depend upon the integrity of the areaacquisition of a CTA. Using a standard CTA paradigm. Hunt postrerna. The CTA produced by exposure to ionizing radia-et i. 1621 exposed one member of a parabiotic pair of rats to tion. for the most part. falls within the class of UCS that is360 rad of ionizing radiation while the other member of the mediated by the area postrema. Lesions of the area postremapair was shielded. When tested for a saccharin preference, produce an equivalent attenuation of taste aversionsthe shielded member of the pair showed a significant aver- produced by exposure to gamma radiation and by injectionsion to the saccharin solution. Since the parabiotic pair share of LiCI and histamine [93.1011. Also consistent with this hy-a common blood supply, the inference is that some blood pothesis is the observation that lesions of the area postrematransferable humoral factor is produced in the irradiated after the initial pairing of a novel sucrose solution with eithermember that can act as a UCS for the shielded member. radiation or LiCI on the conditioning day has no effect on theThere appears to he a dose/response relationship between subsequent recall of the CTA [1071. This observation furtherthe dose of radiation and the release of the humoral factor confirms the function of the area postrema as a transfer pointbecause the strength of the aversion produced in the shielded by which information about potential toxins in the blood and

6) RABIN AND HUNT

cerebrospinal fluid is transmitted into the central nervous serotonin. A few studies have appeared examining choliner-system. gic mecmaanisms.

As indicated above, a comparison of aversions produced One difficulty inherent in this approach, however, is theby partial-body exposures indicates that irradiation of the possible confounding of the processes that lead to the initia-body-only is more effective in producing a CTA than is ir- tion of the behavior with those that lead to the association ofradiation of the head-only [40. 1261. This finding raises the the CS with the UCS or to the expression of the behavior.possibility that taste aversions produced by head- and Since the acquisition of a response can only be demonstratedbody-only exposures may involve different mechanisms. through its subsequent performance, the failure of an or-This possibility was evaluated in an experiment comparing ganism to show a CTA following neurochemical manipula-the effects of area postrema lesions on the acquisition of a tion of the brain may reflect a deficit in either process. SinceCTA produced by head- or body-only exposure [1061. Le- the present concern is with the mechanisms responsible forsions of the area postrema of rats exposed to body-only the initiation of the CTA response. care must be taken to.diation produced a complete disruption of CTA learning. sepa,~ate out these various processes that lead to the per-In contrast, area postrema lesions in rats exposed to head- formance of a learned response. Ideally, the experimentonly radiation, while producing a significant attenuation of should be designed to show that a given manipulation dis-the CTA. did not prevent the occurrence of a significant rupts CTA learning to a class of unconditioned stimulireduction in test day sucrose preference compared to condi- presumed to have a similar mechanism of action while hav-tioning day preference. These results indicate that the ac- ing no effect on the acquisition of a CTA induced by stimuliquisition of a CTA following partial-body exposures is presumed to have a different mechanism of action.mediated by the area postrema and also involves additional This approach has been utilized in the comparison of LiCImechanisms not dependent upon the integrity of the area and amphetamine as unconditioned stimuli for taste aversionpostrema. A direct effect of radiation on the brain mediating learning. Lesions of the area postrema disrupt the acquisi-the acquisition of a CTA would be consistent with studies tion of a CTA produced by injection of LiCI andindicating changes in electrocortical activity 124. 86, 87). in methylscopolanine. but have no effect on an aversionseizure thresholds 188.971. and in sodium channel function produced by amphetamine 14.1141. Conversely, manipula-11401 at radiation doses of less than 300 rad. tion of catecholaminergic systems, either by injection of

6-hydroxydopamine [ 115,1351 or by lesions of the dorsolat-Ride o1" ( 'entral Neuriptrn.inier.s eral tegmentum [ 1391 attenuate a CTA produce I by injection

of amphetamine, but have no effect on the acquisition of aSo far. there is little indication what brain mechanisms CTA following injection of LiCi. Because these experiments

may mediate radiation-induzed CTA learning in the absence show that the effects of manipulation of catecholaminergicof the area postrema. One possible approach to determining systems on CTA learning are restricted to a single class ofpotertial mechanisms would be to examine the effects of unconditioned stimuli, it would suggest that the effects ofirradiation on central neurotransmitters. For example. neu- manipulation of these systems are on the mechanisms re-rochemial studies have. in fact. indicated that exposing an sponsible for the initiation of the response and not on theorganism to ionizing radiation can affect biogenic amines in processes responsible for the association or performance ofthe brain 1651. the response.

A number of neurobehavioral studies have been under- One transmitter whose activity correlates with the devel-taken in an attempt to determine the mechanisms of CTA opment of an LiCI-induced CTA is serotonin. Various ma-learning produced by a variety of unconditioned stimuli. nipulations that alter serotonergic activity can modify theMost of these studies use lesions of various areas of the brain magnitudeofthe CTA. Lesions of the median raphe nucleus.and drugs that modify the actions of neurotransmitters. The but not the dorsal raphe nucleus prior to conditioningrelationship of these studies to the mechanisms of radiation- enhance the LiCI-induced CTA 1761. On the other hand, pre-induced taste aversions is not clear, since they generally treatment of either raphe-lesioned or unlesioned animals arehave involved unconditioned stimuli other than radiation. pretreated tefore conditioning with 5-hydroxytryptophan,However. based on the discussions in previous sections of the precursor of serotonin. or with inhibitors of serotoninthis review, it may be possible to infer potential mechanisms uptake results in the acquisition of an attenuated CTAof radiation-induced taste aversions through the actions of 177.781. Since medial ra'phe lesions deplete serotonin in lim-unconditioned stimuli that seem to act in a manner similar to hic structures such as the hippocampus and septum, the re-radiation. One way to group these stimuli is by the ability of suits of these studies are interpreted as indicating that thelesions of the area postrema to block specific aversions, role of serotonin may be to modulate the perceived intensitySince both radiation- and LiCl-induced aversions are pre- of the toxic stimulus.vented by area postrema lesions, actions of lithium relative In general, there is little evidence for a role of catechola-to the development of a CTA might have some bearing on mines in radiation- or LiCI-induced CTA learning. Blockadehow radiation induces a CTA. of catecholamine syntlhesis or dopaminergic receptors does

The manipulation of the actions of neurotransmitters has not alter the subsequent development of a CTA 199.1221.been a major means by which possible mechanisms underly- Also, as indicated above, depletion of forebrain norepineph-ing CTA learning have been examined. These manipulations rine with the neurotoxin 6-hydroxydopamine or electrolytichave been accomplished either by administering drugs that lesions of the dorsolateral tegmentum are ineffective 181,either facilitate or suppress the activity ofa given transmitter 1391. On the other hand, infusions of beta-adrenergicor by applying selective lesions to areas of the brain that agonists and antagonists into the ventricular syste!m of thesend afferents containing that transmitter or contain the syn- brain have hecn reported to modify the uevelopment ofpatic endings. Most of the research, though limited, has fo- LiCI-induced taste aversions. The agonist enhances the CTAcused on the biog,:nic amines that include the catechola- and the antagomist reduces it 1721. Because similar infusionsmines. dopamine and norepinephrine. and the indoleamine. into the lateral hypothalamus can alter the asersiveness of

RADIATION-INDUCED TASTE AVERSIONS 6 1

cer'ain tastes, these treatments may be related to the mag- sistent with the observation that lesions of the area postremanitude of an animal's response to taste. Also, because lesions disrupt CTA learning following injection ofof the amygdala produce a nonselective disruption of CTA methylscopolamine. which cannot cross the blood/brainlearning produced by X-irradiation [32] as well as to both barrier, but have no effect on CTA learning when the UCS isLiCI and amphetamine injection [541, and because the deple- a drug such as amphetamine, which can cross the barrierlion of norepinephrine produced by injection of [4,1141. However. this hypothesis does not account for the6-hydroxydopamine into the basolateral amygdala also dis- disruption of an LiCl-induced CTA by area postrema lesionsrunts CTA learning [111. it may be that catecholamines are because LiCI does cross the blood/brain barrier [251. There-more related to the association of the CTA response to a CS fore. other factors, in addition to the ability of a substance tothan to the nature of the UCS. cross the blood/brain barrier, might determine whether or

Results from experiments studying the effect of drugs that not the area postrema plays a necessary role in the acquisi-block acetylcholine receptors have depended upon the do,:es tion of a CTA. It may be that the central effects of LiCI.used. With low doses of atropine (0.6 mg/kg. SC) or unlike those of amphetamine. are not relevant for the acqui-scopolamine (0.05-1.0 mg/kg. SC). the development of a sition of a CTA.radiation- or LiCI-induced CTA is not altered 137,1251. After While the data indicate that activation of the area %

higher doses of atropine (15 mg/kg. SC or 104) mg/kg. IP), postrema is not a necessary condition for CTA learning, ,-radiation- and LiCl-induced taste aversions are attenuated there are some very limited data which suggests that it might[29.521. However. interpretation of these results is further be a sufficient condition. In the cat. but not in the rat, thecomplicated by the observation that !reatment with atropine cardiovascular effects of angiotensin 11 are mediated by thesulfate (25 mg/kg. IP) itself produces a CTA [981. area postrema [1211. In agreement with these findings is the

Although its relevance to radiation-induced CTA learning observation that injection of angiotensin I1 produces changeshas yet to be directly established. endogenous opioids have in the activity of single units in the area postrema of the catbeen implicated as mediating some of the behavioral changes 1101, but has no such effects in the rat [131. Concordant withobsersed following exp-)sure to ionizing radiation [83,841. the electrophysiological data on the area postrema is theMickley 1831 has reported that treatment with nalxone pre- observation by Rabin et al. [ I I I that injection ofangiotensinvents the occurrence of a radiation-induced stereotypic 11 will produce a CTA in cats. but not in rats. To more fullyhy peractivity in C57BL/6J mice that is similar to the behav- evaluate this possibility that activation of the area postremaioral response of these mice to treatment with morphine. may constitute a sufficient condition for CTA learning, how-Also. morphiine-tolerant rats show a smaller radiation- ever. would require additional electrophysiological studiesinduced performance decrement than do non-tolerant rats monitoring the response of area postrema neurons to ioxins[841. Similar results have been reported y Tesky and that do not cause a CTA.Kavaliers 11321 who observed a radiation-induced analgesia One other question concerns the specific nature of thein CF-I mice wh~ch could be reversed by treatment with UCS. The data reviewed in previous sections do not providenaloxone. Because treatment with morphine produces a clear evidence as to whether the potential toxicity of a cer-CTA that can b,: attenua.ed by treatment with naloxone tain class of unconditioned stimuli, such as ionizing radiation174.1341. it may be that endogenous opioids play a role in the and LiCI. is the proximal antccedent condition for CTAacquisition of a radiation-induced CTA. However. Rabin and learning. Because a CTA can be produced by nontoxicHuit (unpublished manuscript) found that pretreating rats stimuli that organisms will sef-administen. a UCS-producedwith a single injection of naloxone did not attenuate the ac- illness cannot be considered a necessary condition for CTAquisition of t radiation-induced CTA. learning. Even where a toxic UCS is utilized, the data do not

From this discussion, there is no clear role for the major provide unequivecal support for the common assumptiontransmitters in initiatingt a CTA induced by radiation- or that a sti;nulus-induced illness is a sufficient condition fordrug-released toxins a(ting through the area postrema. The CTA learning. This is true not only for a UCS such as LiCI.effects observed after transmitter manipulation appear to be which produces overt signs of distress. but even more so forrelated more to the expression of the CTA. exposure to ionizing radiation, which produces no overt

changes in unrestrained behavior. It may be that radiation isCONCLU.'SIONS able to produce a CTA because it activates the neural cir-

cuits associated with illness even in the absence of the expe-Overall. the research reviewed in this report would be rience of the illness by the awake organism [1001. Alterna-

consistent with the hypothesis that there are at least two tively. it may be. as suggested by Gamzu 1381, that the basisJlasses of unconditioned stimuli that can lead to the acquisi- for taste aversion learning lies in thL novelty of thelion of a CTA. The first class consists of those unconditioned treatment-induced state: that any treatment will produce astimuli, such as amphetamine, that do not require the media- CrA as long as it produces a novel, discriminable statetion of the area postrvima for CTA learning. !n the second within the organism. In this view, ionizing radiation is able toclass, there are those stis-.uli. such as LiCI. which require the lead to CTA learning because irradiation produces a discri-mediation of the area postrema for the acquisition ofa CTA. minable change in nervous system activity. The potentiallyFor the most part. radiation, as a UCS for CTA learning, toxic character of the radiation UCS. or any other UCS suchseems to belong to the second class of stimuli, as LiCI. is not directly relevant to the CTA learning as long

The area postrema is one of i group of circumventricular as that UCS is capable of producing a discriminably differentorgans that is characterized by a relatively weak blood/brain state. As such, the UCS-produced illness may not be thebarrier 1281. As such. it may be assumed that the role of the direct antecedent cause of the CTA. but may rather simplyarea postrema in CTA learning is to transfer to the central be an unavoidable side-effect of the treatment that producesnervous system information about the presence of toxins in such a novel state.blood and cerebrospinal fluid when those toxins cannot cross The final question concerns the nature of the interactionthe blood/brain barrier 11071. This hypothesis would be con- between the UCS and the area postrema. When a CTA is

62 RABIN AND HUNT

acquired following treatment with a systemic toxin. such as tides causes emesis. In rats. treatment with the gastrointesti-LiCI, which involves the mediation of the area postrema. an nal hormone cholecystokinin may produce a CTA ([27.331.implicit assumption is that the toxin exerts direct effects on but see 1481 for contrary data) that is mediated by the areaneuronal activity of the area postrema leading to CTA learn- postrema [1331. A CTA has also been reported to result froming. This assumnption. however, might not be correct. Smith repeated systemic injection of arginine vasopressin [341.[123] reported that injection of LiCI directly into the fourth Whether any of these peptides can, in fact, serve as theventricle, unlike systemically administered LiCI, did not factor which is the proximal UCS for CTA learning followingproduce a CTA. The implication of this finding is that CTA exposure to ionizing radiation or following systemic treat-learning following systemic (intraperitoneal or intragastric) ment with LiC remains to be established.treatment with LiCI does not result from a direct action of It is, however, clear that the CTA. as a CR to a CS asso-lithium on area postrema neurons, but rather that systemic ciated with toxic consequences to the organism, has definitetreatment with LiCI causes the release of an endogenous implications for the survival of the organism. Given the widemediator to which the area postrema is sensitive. Thus. the range of unconditioned stimuli which can elicit a CTA. it isLiCI-induced CTA. like the radiation-induced CTA, may de- unlikely that the area postrema would have evolved specificpend upon a treatment-released humoral factor which serves receptors for each potential toxin. Rather, it is more likelyas the proximate UCS for CTA learning. It is interesting to that a class of toxic stimuli produce a series of similar effectsspeculate that, because manipulations which affect LiCI- resulting in the release of an endogenous factor. This factorinduced CTA learning also affect the radiation-induced CTA. may alter the activity of neurons within the area postremathe same endogenous humoral factor mediates CTA learning with the consequent development of a CTA. Radiation isfollowing treatment with either UCS. or. indeed, any UCS capable of functioning as a UCS for CTA learning because itthat requires the mediation of the area postrema. too is capable of causing the release of this humoral factor.

Electrophysiological studies have shown that the area Ionizing radiatio, is, therefore, just one member of a class ofpostrema is sensitive to a variety of endogenous peptides. environmental toxins that induce a CTA through the releaseWorking with dogs. Carpenter and his coworkers [17,181 of some endogenous factor affecting the activity of the areahave reported that iontophoretic application of a variety of postrema. The validity of this hypothesis and the definitionpeptides causes changes in the activity of area postrema of a potential endogenous factor remain to be e,..blished byneurons and that systemic treatment with these same pep- further research.

REFERENCES

I. Ashe, J. H. and M. Nachman. Neural mechanisms in taste 14. Bruner. A.. V. Bogo and R. K. Jones. Delayed match-to-aversion !earning. In: Progre.is in P.vc'h/,hihhgy and Phy'.%i- sample early performance decrement in monkeys after "'Co,doei'al P."ychsly. vol 9. edited by J. M. Sprague and A. N. irradiation. Radiat Rex 63: 83-96. 1975.Epstein. New York: Academic Press. 1980. pp. 23.3-262. 15. Burghardt. W. B. and W. A. Hunt. Characterization of

2. Barker. L. M. and J. C. Smith. - comparison of taste aver- radiation-induced performance decrement using a two-leversions induced by radiation and lithium chloride in CS-US and shock-avoidance task. Radht RvE' 103: 149-157, 1985.US-CS paradigms. J ('Op Phvsi~d, P vchl 87: 644-654. 1974. 16. Cairnie, A. B. and K. E. Leach. Dexamethasone: A potent

3. Berger. B. D. Conditioning of food aversions by injections of blocker for radiation-induced taste aversion in rats. Pharmmadpsychioactive drugs. J (omp Phvsiod Pvchld .;': 21-26. 1972. Biochem Behav 17: 305-311. 1982.

4. Berger. B. D.. C. D. Wise and L. Stein, Area postrema damape 17. Carpenter. D. 0.. D. B. Brigg., and N. Strominger. Responsesand bait shyness. J (Comp Pl'.%i,, Psvth,,i 82: 475-479, 1973. of canine areapostrema to neurotransmitters and peptides. 6,1

5. Bernstein. 1. L. and M. M. Webster. Learned taste aversions in MIo .\;-,ur,htio 3: 113 -126, 1983.humans. PhysioI Behv 25: 36.3-366. 1980. 18. Carpenter. D. 0.. D. B. Briggs and N. Strominger. Peptide-

6. Bhargava. K. P. and K. S. Dixit. Role of the chemoreceptive induced emesis in dogs. 11chE Brain Re.s I1: 277-281, 1984.trigger zone in histamine-induced emesis. Br J Pharmaco M: 19. Carroll. M. E. and J. C. Smith. Time course of radiation-508-513. 1968. induced taste aversion conditioning. PhYsxid BettoE 13: 809-

7. Bogo, V. Effects of bremsstrahlung and electron radiation on 812. 1974.rat motor performance. REidia Res 100. 31.3-320. 1984. 20. Chinn. H. I. and S. C. Wang. Locus of emetic action following

8. Borison. H. L. 5ite of emetic action of X-radiation in the cat../ irradiation. PrEE Soc i'.p Bid Med 85: 472-474. 1954..aup .\,urd 107: 439-453. 1v57. 21. Coil. J. D. and R. Norgren. Taste aversions conditioned with

9. Borison. H. L. Area postrema: Chemoreceptive trigger zone intravenous copper sulfate: Attenuation by ablation of the areafor vomiting-is that all'? Ifi, St i 14: 1807-1817. 1974. postrema. Brain Res 212: 425-433. 1981.

10. Borison. H. L.. M. J. Hawken. J. I. Hubbard and N. E. Sirett. 22. Coil. J. D., R. C. Rogers. J. Garcia and D. Novin. ConditionedUnit activity from cat postrema influenced by drugs. fraut Rd'. taste aversions: Vagal and circulatory mediation of the toxic92: 15.1- 156, 1975. unconditioned stimulus. Beice, Bi,d 24: 509-519, 1978.

II. Borsini. F. and E. T. Rolls. Role of norad.enaline and seroto- 23. Coil. J. D.. W. G. Hankins. D. J. Jenden and J. Garcia. Thenin in the basolateral region of the amygdala in food prefer- attenuation of a specific cue-to-consequence association byenczs and learned taste aversions in the rat. Phy i,,l Bc/ia. 31: antiemetic agents. P'%, ,/,harnicEE(Ed,, I hErlin) 56: 21-25.37-43. 1984. 1978.

12. Brizzee, K. R Etfects of localized brain stem lesions and sup. 24. Daugherty. J. H. and C. D. Barnes. Effect of X-irradiation onradiaphragmatic vagotomy on X-irradiation emesis in the mon- thie interaction of flash-evoked cortical potentials. Radial Re.Skey. Ain J Payssitd 187: 567-570. 196. 49: 359-366. 1972.

13. Brmiks. M. J.. J. I. Hubbard and N. F. Sirett Extracellular 25. Davenpot. V. D. Distribution of parenterally administeredrecording in rat area postrema in vitro and the effects of lithium chkide in plasma, brain and muscle of rats. Am .1cholinergic drugs. serotonin and angiotensin II. Brain R,.v 261: Phv.iED 163: 633-641. 1950.8.-90. 1983.

RADIATION-INDUCED TASTE AVERSIONS 63 K126. Dempsey. E. W. Neural and vascular ultrastructure of the area 50. Goudie, A. J., E. W. Thornton and J. Wheatley. Attenuation

postrema in the rat. J Crnip Neurot' 150: 177-200, 1973. by alpha-methyltyrosine of amphetamine-induced conditioned27. Deutsch. J. A. and W. T. Hardy. Cholecystokinin produces taste aversion. Psvchopharnacoloia 45: 119-123, 1975.

bait shyness in rats. Nature 266: 196. 1977. 51. Goudie. A. J.. i. P. Stolerman. C. Demellweek and G. _ '_

28. Deutsch. J. A.. F. Molina and A. Puerto. Conditioned taste DMello. Does conditioned nausea mediate drug-induced tasteaversion caused by palatable nontoxic nutrirnts. Behav Bi,'l 16: aversion. Ps%-chophariacolo v iBerlin) 78: 277-281. 1982. i7,7

161-174. 1976. 52. Gould. M. N. and N. B. Yatvin. Atropine-caused central nerv-29. Deutsch, R. Effects ot atropine on conditioned taste aversion. ous system interference with radiation-induced learned and un- ,

PhormaoI Bioihni Bthov 8: 685-694. 1978. learned behaviours. Int J Radiat Biu/ 24: 463-468, 1973.

30. Doyle. T. F.. C. R. Curran and J. E. Turns. The prevention of 53. Grote. F. W. and R. T. Brown. Conditioned taste aversions:radiation-induced, early. transient incapacitation of monkeys Two-stimulus tests are more sensitive than one-stimulus tests.by an antihistamine. Pro( So( L.rp Biol M d 145: 1018-1024, Bchat Res Meth ln tr 3: 311-312. 1971.

1974. 54. Grupp. L. A., M. A. Linseman and H. Cappell. Effects of31. Dragoin. W.. G. E. McCleary and P. McCleary. A comparison amygdala lesions on taste aversions produced by amphetamine

of two methods of measuring conditioned taste aversions, and LiCI. Pliarmacol Bio)hern Behav 4: 541-544, 1976.Bchav Rs Meth Instr 3:309-310. 1971. 55. Hennessy. J. W. and S. Levine. Stress arousal and the pituitary

32. Elkins, R. L. Attenuation of X-ray-induced taste aversions by adrenal system. A psychoendocrine hypothesis. In: Progress inolfactory bulb or amygdaloid lesions. Phisiol Bhav 24: 515- P .chohiopitv and Phystobgical Psychology, vol 8. edited by

521. 1980. J. M. Sprague and A. N. Epstein. New York: Academic Press.33. Eitenberg. A. and G. F. Koob. Different effects of choiecys- 1979. pp. 133-178.

tokinin and satiety on lateral hypothalmic self-stimulation. 56. Hennessy, J. W.. W. P. Smotherman and S. Levine. Con-Pihvi, Bhav 32: 127-130. 1984. ditioned taste aversion and the pituitary-adrenal system. Beha•-

34. Ettenberg. A.. D. van der Kooy. M. Le Moal. G. F. Koob and Bi,,l IS: 41.3-424, 1976.F. E. Bloom. Can aversive pronerties of (peripherally-injected) 57. Hennessy, J. W.. W. P. Smotherman and S. Levine. Investiga-vasopressin account for its putative role in memory'? Bhav tions into the nature ofdexamethasone and ACTH effects uponBrain R(-. 7: 331-350. 1983. learned taste aversion. Phi'siol Behav 24: 645-649, 1980.

35. Fenner. D. P. and 1. L. Bernstein. Interference in food aversion 58. Hulse. E. V. and L. G. Mizon. Radiation-conditioned aversionconditioning by reducing drug dose or conditioning trials. iq rats after part. body X-irradiation and its relationship to gas-B~tha v Nural Bi, /40: 114-1 18, 1984. tric emptying. Int J Radiat Biol 12: 515-522, 1%7.

36. Franz. C. G. Effects of mixed neutron-gamma total-body ir- 59. Hulse. E. V. and G. Patrick. A model for treating post-radiation on physical activity performance of rhesus monkeys. irradiation nausea and vomiting in man: The action of insulin inRadiat Rc. 101: 434-441. 1985. abolishing radiation-induced delay in gastric emptying in the

37. Gadusek. F. J. and J. W. Kalat. Effects of scopolamine on rat. Br J Radiol 50; 645-651, 1977.retention of taste-aversion learning in rats. Phy,io, P.svchol 3: 60. Hunt. E. L. and D. J. Kimeldorf. Behavioral arousal and neural130-132. 1975. activation as radiosensitive reactions. Radiat Res 21: 91-110.

38. Gamzu. E. The mul'ifaceted nature of taste-aveiion-inducing 1964. 1777agents: Is there a single common factor'? In: Learniu., .thch- 61. Hunt. E. L. and D. J. Kimeldorf. The humoral factor inantii, i, Food S ' timi. edited by L. M. Barker. M. R. Best radiation-induced motivation. Radiat Res 30: 404-419. 1967.and M. Domjam. Waco. TX: Baylor Univ. Press. 1977. pp. 62. Hunt. E. L.. H. W. Carroll and D. J. Kimeldorf. Humoral477-509. mediation of radiation-induced motivation in parabiont rats. r-

39. Garcia. J. and D. J. Kimeldorf. Temporal relationship within S iciec ISO: 1747-1748. 1%5.the conditioning of a saccharin aversion through radiation ex- 63. Hunt. E. L.. H. W. Carroll and D. J. Kimeldorf. Effects o'posure. J (woi Plivti ,,l I'v hod 50: 180-183. 1957. dose and of partial-body exposute on conditioning through a

40. Garcia. J. and D. J. Kimeldorf. Some factors which influence radiaticn.induced humoral factor. Piyviol Behav 3: 809-813.radiation-conditioned behavior of rats. Rhad Re. 12: 719--/27. 19619.1960. 64. Hunt. W. A. Comparative effects of exposure to high energy

41. Garcia. J. and R. A. Koelling. Relation of cue to consequence electrons and gamma radiation on active avoidance behavior.in avoidance learning. Pvt han S(i 4: 123- 124. 1966. In . Radit Btiol 44: 257-260, 1983. 6,

42. Garcia. J. and R. A. Koelling. A comparison of aversions in- 65 Hunt. W. A.. T. K. Dalton and J. H. Darden. Transient altera-duced l', X-ra, s. toxins, and drugs in the rat. Rtjdt Rci Suppl tions in neurotransmitter activity in the caudate nucleus of rat7: 439-450. 1967. hrain after a high dose of ionizing radiation. Rodiat Res 80:

43. Garcia. J.. F. R. Ervin and R. A. Koelling. Toxicity of serum 556-562, 1979.from irradiated donors. Vatur, 213: 682-683. 1967. 66. lonescu. E. and 0. Buresova. Failure to elicit conditioned taste

44. Garcia. J.. W. G. Hankins and K. W. Rusiniak. Behavioral aversion by severe poisoning. Pharma-'ol Bioc'wm Behav 6:regulation of the milieu interne in man and rat. .St ivac v85: 251-254. 1977.824-831. 1974. 67. Kemper. T. L.. R. O'Neill and W. F. Caveness. Effects of

45. Garcia. J.. D. J. Kimeldorf and F. L. Hunt. The use of ionizing single dose supervoltage whole brain radiation in macacaradiation as a motivating stimulus. Pin ho Rev 6 : 38.--395, mulatta. J .,uropathol i-.ip Nearol 36: 916-940, 1977.1961. 68. Kiefer. S. W.. K. W. Rusiniak and J. Garcia. Vagotomy

46. Garcia. J.. D. J. Kimeldorf and R. A. Koelling. Conditioned facilitates extinction of conditioned taste aversions in rats. Javersion to saccharin resulting from exposure to gamma radia- ('onip Phl'aiol Pivchol 95: 114-122. 198!.lion. Si ,m 122: 157-158. 1955. 69. Kiefer. S. W.. R. J. Cabral, K. W. Rusiniak and J. Garcia.

47. Gaston. K. E. Brain mechanisms of conditioned taste aversion Ethanol-induced flavor aversions in rats with stbdiaphragmaticlearning: A review of the literature. PhIv-uI Ivi wlI 6: 340-345, vagotomies. Bchav ,N'e.uratl Hiol 29: 246-254, 1980.197g. 70. Kimeldorf, D. J. and E. L. Hunt. Iiowint, Radiation: Neural

48. Gibbs. J.. R. C. Young and G. P. Smith. Cholecystokinin de- IFunt ,on and Bethavior. New York: Academic Press. 1965.creases food intake in ras, I (',,p I'nhv.,I 1'%%ho/ 84: 488-- 71. rimeldorf. D. J.. J. Garcia and D. 0. Ruhadeau. Radiation-495. 1973. induced conditioned avoidance behavior in rats, mice. and

49. Goudie. A. J. Aversive stimulus properties of drugs. cats. Radiat Rc.i 12: 710-718. 1966.%etrophartaioloev IS: 971-979. 1979. 72. Kral. P. A. and V. V. St. Omer. Beta adrenergic receptor in-

volvement in the mediation of learned taste aersions. Ps vcho-phartnacirifita 26: 79--3. 1972.

64 RABIN AND HUNT

73. Lasiter. P. S. and J. J. Braun. Shock facilitation of taste aver- 96. Phillips. A. G. and F. G. LePiane. Disruption of conditionedsion learning. Behav Neural Hiol 32: 277-281. 1981. taste aversion in the rat by stimulation of amygdala: A condi-

74. LeBlanc. A. and H. Cappell. Antagonism of morphine-induced tioning effect, not amnesia. J (ampp Phv s/al Psvchwl 94: 664-aversive conditioning by najoxone. Pharmual Biachein Behav 674. 1980.3: 185- 188. 1975. 97. Pollack. M. and P. S. Timiras. X-ray dose and electroconvul-

75. Levy. C. J_. M. E. Carroll. J. C. Smith and K. G. Hofer. sive responses in adult rats. Radiat Re-s 21: u11-119. 1964.Antihistamines block radiation-induced taste aversions. .Si- 98. Preston. K. L. and C. R. Schuster. Conditioned gustatorycm v 186: 10~44-1046, 1974. avoidance by three chelinergic agents. Plwrnuaudl Bit,ahin

76. Lorden. J. F. and D. L. Niargules Enhancement of con- &'hai 15: 827-8-28. 1981.ditioned taste aversions by lesions of the midbrain raphe nuclei 99. Rabin. B. M. and W. A. Hunt. Effects of antiemetics on thethat deplete seweonin. P/a.,iu'/ PA%'( /IN) 5: 2731-279. 0977. acquisition and recall of radiation- and lithium chloride-Induced

77. Lorden. J. F. and W. B. Nunn. -115tcts of central And periph- conditioned taste aversions. Pharmnacol Bihce Bewav 18: Vreral pretreatment with fluoxetine in gustatory conditioning. 629-635. 1983.Pharmnat l Bioc/n-a 8,4,ui 17: 431-443. 1982. 100. Rabin. B. M. and J. S. Rabin. Acquisition of radiation- and %-

78. Lorden. J. F. and G. A. Oltmans. Alteration of the charac-ithm clodenued odtoed ase vrins nteristics of learned taste aversion by manipulation of serotonin anesthetized rats. Aigim Learn Beha r 12: 439-441. 1984.levels in the brain. Pharmaucal Ejaclin em has, 8: 13-18. 1978. 101. Rabin. B. M.. W. A. Hunt and J. Lee. Studies on the role of

79. Louis-Syivestre. J. Validation of tests of completeness of va- central histamine in the acquisition of as radiation-induced con-putomy in rats. J Amon~ Verv Svst 9: 301-314. 1983. ditioned taste aversion. Radit Res 90: 609-620, 1982.

80. Martin. J. R.. F. Y. Cheng and D. Novin. Acquisition of 102. Rabin. B. M.. W. A. Hunt and J. Lee. State-dependent in-learned taste aversion following bilateral subdiaphragmatic va- teractions in the antihistamine-induced disruption of agotomy in rats. Phv.%iol Hchai, 21: 13-17. 1978. radiat ion- induced taste aversion. Radiat R-.% 90: 621-627. !982.

81. Mason. S. T. and H. C. Fibiger. Norasdrenaline and extinction 103. Rabin. B. M.. W. A. Hunt and J. Lee. Attenuation of radiation-of conditioned taste aversions in the rat. BelaiiNeural B/al 2;: and drug-induced conditioned taste aversions following area206-216. 1979. postrema lesions in the rat. Rad/at Res 93: 388-394. 1983.

82. Mickley. G. A. Behavioral and physiological changes produced 104. Rabin. B. M.. W. A. Hunt ind J. Lee. Acquisition of lithiumby a scipralethal dose of ionizing radiation: Evidence for chloride- and radiation-induced taste Aversions in hypophysec-hormone-influenced sex differences in the rat. Rodcai Resa 81: tomnized rats. P/uannal Bjaa/iin Hiau 18: 463-465. 1983.48-75. 1980. l05. Rabin. B. NI.. W. A. Hunt and J. Lee. Taste aversion learning

83. Mickley, G. A.. K. E. Stevens. G. A. White and G. L. Gibbs. in vagotomized rats. Paper presented at the 1983 Meeting of theEndogenous opiates mediate radiogenic behavioral change. American Psychological Association. Anaheim. CA.Sti/iii e 220: 1185i-1187. 1983. 106. Rabin. B. M.. W. A. Hunt and J. Lee. Effects of dose and of

84. Mickley. G. A.. K. E. Stevens. J. M. Burrows. G. A. White partial body ionizing radiation on taste aversion learning in ratsand G. L. Gibbs. Morphine tolerance offers protection from with lesions of the area postrema. l'/h.~/i Be/wi' 32: 119-122.radiogenic performance deficits. Radiiu Re-s 93: 38 1-387. 1983. 1994.

85. Miller. D. S. Effects of low level radiation on audiogenic con- 107. Rabin. B. M..- W. A. Hunt and J. Lee. Recall of a previously Ivulsive seizures in mice. In: Re-%poins,- of/gla ,erivuaa Sy ti-u, acquired conditioned taste aversion in rats following lt;..ions ofto Ia,,i: Riadjiation. edited by T. J. Haley and R. S. Snider. the area postrema. Phivshil Be/,av 32: 503-506. 1984.New York: Academic Press. 1962. pp. 513-531. 10(M. Rabin. B. M.. W. A. Hunt and J. i.ce. Interactions between N

86. Minamisawa. T. and T. Tsuchiya. Long-term changes in the unconditioned stimruli that produce ta:.te aversion learning. Sacaveraged evoked potentials of the rabbit after irradiation with Nettrami Absir It: 531. 1985.moderate X-ray doses. fH,-,trip-nc-phahui.r (1wn Ncaaro.siol-. 109. Rabin. B. MI.. W. A. Hunt an'I J. Lee. Intragastric copper42: 416-424. 1977. -sulfate produces a more reliable conditioned taste aversion in

87. Minamisawa. 1.. K. Yamamoto and T. Tsuchiya. Long-term vagotomized rats than in intact rats. He-hav Vv-ural B/ol. ineffects oif moderate X-ray doses on the averaged evoked po- press. 1'5.tentials in the rabbit. 04,-itri-ui-vitcehlitwr (u \i-sur,,hvil 110t. Rabin. B. M., W. A. Hunt and J. Lee. Effect of area postrema60: 59-70. il'8K0. lesions oin taste averions produced by treatment with WR-

8. Minkler. J. l,,gripui t/,um Io Neugrmia te-,ie. Saint-Louis. MO: C. 2721 in the rat. .Nurobea,-ii Ia 106 liragaol 8: in press. 1986.V. Mosby Co.. 1972. 111. Rabin. B. M.. W. A. Hunt. A. C. Bakarich. A. L. Chedester

89. Mossman. K. L.. A. J. Martini and R. 1. Heinkin. Radiation- and J. Lee. Angiotensin 11 produces a taste aversion in cats butinduced changes in sodium preference and fluid intake in the not in rats. Not Ne-ornm A/'Air 10: 1011. 1984.rat. ha J Rauwt jil36: 191-196. 1979. 112. Remler. NI. P. and W. H. Marcussen. 'Time course of early

90. Mossman. K. L..- J. D. Chencharick. A. C. Scheer. W. P. delayed blood-bran barrier changes in individual cats after

Radiation-induced changes in gustatory function. comparison 113. Riley. A. L. and L. L. Barid. Conditioned taste aversions: Aof effects of neutron and photon irradiation. /ti Rod/ait Ongi hiibliography. Anion l.a-arpi Bi/io 4: IS- 13S. 1976.Bi/ol Ph%% 5: 52 1-528. 1979. 114. kitter. S.. J. L. McGlone .And K. W. Kelly. Absence of

91. Nachman. NI. and J. H. Ashe. Learned taste aversions in rats lithium-induced taste aversion after area postrema lesion.as a function of dosage, concentration, and route of adminis- Biraina R.- 201: 501-"0. 19K). "A.

tration of LiCI. Ph% siol Bi-lhas, 10- 7.1-78. 1973. 115. Roberts. 1). C. and H. C. Fibiger. Attenuation of92. Nachman. NI. and P. L. Hartley. Role of illness in producing amphetamine-induced conditioned taste aversion following in-

learned taste aversion in rats: A comparisain of several rodent. traventricular 6-hydroxy-dopamine. Vviormui Lett 1: 343-147.icides. J (amuit, Pinva/aal /'%%-Ihitd 89: 1010--10118. 1975.19.

93. ()ssenkopp. K.-P. Taste aversion conditioned with gamma 116. Roll. I). IL. and J. C. Smith. Conditioned taste aversion inradiation: Attenuation by area postrema lesions in rats. Bc/war anesthetited rats. In: lt/liii.-i al lau/tai~'ILaihliBrojugs Rv-s 7: 295-3035. 1983. edited by NI. L. P. Seligman and J. L.. Hager. New York:

94. Overton. D. A. Experimental method-. for the study of state- Appleton-Century-Crafts. 1972. pp. 98-11)2.dependent learning. Fed l'roc 33: 1800-181.1. 197A. 117. Rondcau. 1). B..' F. RI. Jolicocur. A. I). Mcrkel and MI. J.

95. Parker. L. A. Nonconsammatory and consummatory bchav- Wayneur. ~rtugs and taste aversion. .Ariiami iaa/ie/iai Rev 5:ioral CRs elicited by lithium- and am phetamine- paired flavor%. 279-294. 19K I.Larn, Moir 13: 281-303. 19112.

RADIATION-INDUCED TASTE AVERSIONS 65

'e.'18. Rusiniak. K. W.. C. C. Palmerino. A. G. Rice. D. L. Forthman 130. Smotherman. W. P.. A. Margolis and S. Levine. Flavor preex-

and J. Garcia. Flavor-illness aversions. Poteriti-ition of odor by posures in a conditioned taste aversion iituation: A dissocia-taste %kith toxin but not shock in rats.. C(omip Ph% sj,/ P/i ho tion oftbehavioral an.. endocrine effects in rats. .1 Comip Phvsiol 496: 527-539, 1 ,82 PA%.chi, 94: 25-35 1980.

19. Sessions. G. R. Histamine and radiation-induced taste aversion 13 1. Spector. A. C., J. C. Smith and It. R. Hollander. A comparisonrconditioning . S, nm 190: 402-4013. 1975. of measures used to quantity rad iation- induced taste aversion.

120. Sheline. G. EC.. W. M. Wara and V. Smith. Fherapeutic irradi- Phv~hd Bi'hav 27: 887-901I. 1981.ation and brain injury. Inii J RadIiat Omoi I Ii I'in 6: 1 !15- i02. reskey. G. C. and MI. Kavaliers. Ionizing radiation induces122-8. 1990. cpioid-mediated analgesia in male mice. l.ic Sti 35- 1547-1552.

121. Simpson. J. B. The circum\sentricular organs and the central 1984.actions of angiotensin.zurird rin,hL!\ 32: 248-256. 19M1. 133. Van der Koo . D. Area postrema: site A here cholecy stokinin

122. Sklar. L. S. and Z.. Amit. Manipulations of, atecholamine svs- acts to decrease food intake. Brains Rvi 295: 345-347. 1984. F_ems block the condigioied taste aversion induced bv self- 134. Van der Koov. D. and A. Phillips. Temporal analysis of

administered drugs. A, itrjormwi4c 16: 64-h55, 1977. nalo,,one attenuation o)f morph ine- induced taste aversion.123. Smith. D. F. C.entral , d peripheral effects of lithium on con- I'har',ooIllji 8him Bhiiro 6: 637-641, 1977.

ditioned taste asersiins ;n rats./% (vit 11phltripi44 ivi4 611. 15- 135. Wagner. G. C., R. WA. Foltin. L. S. Seiden and C. R. Schuster.317. 1980. Dopamine depletion by 6-hydroxydopamine prevents con- i124. Smith. J. . Radiation: Its detection and its effects on taste ditioned taste aversion induced by meth~lamphetamine but not .,.

preferences. In: P'r. Crii in Phr mi'otjj (41 I' iwh, v ol 4. lithium chloride. Pko-ronimol Bio(Iwhiy, Beioi 14: 85-8. 1981.editeJ bs E. Stellar and J. M. Sprague. N--% York: Academic 36. Wakisaki. S.. R. R. O'Neil. T. L. Kemper. D. Mi. Verrelli andPress. 197 1. pp. 1- 117. W, F. Caseaiess. Delaxed brain damage in adult monkeys from

25. Smith. !. C. and D). D. Morris. Effects of atripine sulfate on the radiation in the therapeutic range. Ro4/itt Res 80: 277-291.c:onditioned aversioin ito Saccharin fluid Aith X-ra)ss as !he un- 1970.conditioned stimulus. Radwiti Ri 28: 186-1A). 1963. 117. Wang. S. C. and H. L. Borison. Copper sulfate emesis: A study

126. Smith. J. C.. G. R. Hollander and A. C. Spector. Taste aiver- of 4fferent pathvka~s from the gastrointestinal trazt. Ant J,ions conditioned \Aith partial body radiaion exposure:s. Pin wol 164: 52G-526. 195 1.P, ii /H Ititi 2 7: W)3- 9 1. 198 1. 138. Weinberg. J., MI. R. Gunnar. L. P. Brett. C. A. Gonzalez and

127. Smotherman. W. P. and S. Levine. ACTH and ACTH, ,, S. Levine. Sex differences in bio~ichavioral responses to con-modification of neopnobia and taste asersion responses in the fulc in a taste aversion paradigm. Pi Niii Ri/io 29: 20 1-2 10.rait. J1 ( omi Ph o I' % Pu hi 92: 22-13. 1979. 1982.

129. Smotherman. W. P.. J. W. Hennessy and S. I esune. Plasma 139. Wellman. P. J.. P. Mclntbsh and E. Guidi. Effects of dorsolat-ciirticuosierone le-vels during recovery from LtCI produced iaste cral tegmeatal lesions on amphetamine, and lithium-inducedis ersuons. Behaviii l 16: 40)1-412. 1976. lasie aversions. P/,rv.6ii Bcl.- 26: 141-144, 1991.

129. Smotherman. W. P.. J. W. Hennessy and S. L~evine. Plosma 140. Wixon. H. N. and W. A. Hunt. Ionizing radiation decreasescortcosterone level-, ais at, index (if the strength of illness in- veratridine-stimulated uptake of sodium in rat brain synapto-duced taste aversions. IhIn w Hi , hait 17: 9l~9(.1976, sones. Si tent i 210: 107 -1074, 1983.p

141. /ook. B. C_. E. W. Bradley. G. W. Casarett and C. C. Rogers.Pathologic findings in canine brains irradiated with fractionatedfast neutrons or rhotons. Radial Rc.a 84: 562-568. 1980.

Dist.