- Email: [email protected]
Conditioned taste aversion but not adrenal activity develops to ICV adminstration of interleukin-1 in rats
Physiology & Behavior, Vol. 49, pp. 691-694. ©Pergamon Press plc, 1991. Printed in the U.S.A.
0031-9384/91 $3.00 + .00
Conditioned Taste Aversion But Not Adrenal Activity Develops to ICV Administration of Interleukin- 1 in Rats L. J. J A N Z , * R. B R O W N , t
L. ZUO,'~ J. F A L K , * A. H. G R E E N B E R G t
A N D D. G. D Y C K .1
*Department o f Psychology and ~Manitoba Institute o f Cell Biology University o f Manitoba, Winnipeg, Manitoba, Canada R e c e i v e d 22 January 1990
JANZ, L. J., R. BROWN, L. ZUO, J. FALK, A. H. GREENBERG AND D. G. DYCK. Conditioned taste aversion but not adrenal activity develops to ICV administration of interleukin-1 in rats. PHYSIOL BEHAV 49(4) 691-694, 1991.--In a previous investigation with mice, the paired presentation of either odor or taste cues with the peripheral lIP) administration of the immunoactive peptide interleukin-1 (IL-1) led to the conditioned enhancement of glucocorticoid production. The present study found that an initial central infusion of IL- 1 in the presence of saccharin cues produced a robust taste aversion but not a conditioned elevation of either ACTH or corticosterone production. These results indicate that the glucocorticoid response induced by centrally administered IL-1 in rats is independent of the behaviorally aversive properties of this cytokine which are conditionable. The differential effects of IP versus ICV administration of IL-1 on glucocorticoid conditioning requires a clearer specification of the respective signaling mechanisms and pathways activated by these two routes of administration. Conditioned taste aversion
Adrenal activity
Interleukin-1
RESEARCH supports the existence of bidirectional communication between the central nervous system (CNS) and the immune system via neuroendocrine and sympathetic pathways (3,12). The impact of CNS signaling on primarily immunosuppressive components of the immune system has also been demonstrated through various behavioral conditioning studies (1, 2, 9, 10). The procedure in these latter studies involves pairing a distinctive environmental stimulus (conditional stimulus; CS) with a potent immunomodulatory agent (unconditional stimulus; UCS). Conditioning is most often inferred on the basis of an altered immune response to a later presentation of the CS alone (1,2). Interlenkin-1 is an immunoactive 17 kD peptide which is produced peripherally and centrally by many cell types. Its numerous target organs include the CNS where it is capable of inducing fever, altered hippocampal and hypothalamic electrical activity, and slow wave sleep (3). Peripheral administration of exogenous IL-1 can also activate the pituitary-adrenal axis resulting in elevated levels of corticotropin releasing factor (CRF), adrenocorticotropic hormone (ACTH), and glucocorticoids (4). It has been hypothesized that IL-1 is an afferent signal to the CNS and that ACTH and corticosterone are efferent hormonal outputs which could down-regulate peripheral immune responses [reviewed in (3-5)]. Although it is controversial whether IL-1 crosses the bloodbrain barrier, ICV IL-1 rapidly decreases some peripheral cellular immune responses (17). Since most conditioned alterations of
Saccharin
Glucocorticoid
immune responses are immunosuppressive, it is possible that the above pathway may participate in these conditioning phenomena. In support of this hypothesis, glucocorticoid conditioning has recently been reported in mice with IP IL-1 using both taste and odor as conditioning stimuli (11). It has also been reported that IP IL-1 supports a conditioned taste aversion in rats (18). The following experiments assessed the effect of pairing saccharin with the ICV administration of recombinant human interleukin-1 Beta (rhlL-1) on the development of a conditioned taste aversion as well as ACTH and corticosterone production. METHOD Seventy-two experimentally naive, male, Sprague-Dawley (Charles River) rats (Experiment 1 =31; Experiment 2 = 4 1 ) between 90 and 100 days old, weighing 300-350 g, were used. Animals were housed individually in polypropylene cages with a 12-h light/dark cycle with food and water available ad lib. All animals were handled daily in a manner which mimicked treatment conditions prior to surgery and habituated to a two-bottle drinking regimen. Three days after cannula implantation fluid availability was restricted to 30 min dally between 08:30 h and 10:00 h. While under general anesthesia (sodium pentobarbital, 50
1Requests for reprints should be addressed to Dr. D. G. Dyck, Department of Psychology, University of Manitoba, Winnipeg, Manitoba, Canada R3T 2N2.
691
692
JANZ ET AL.
TABLE 1 IL-1 CONDITIONINGPROTOCOLIN TASTEAVERSIONMODEL
Group Unconditioned IL- 1 Conditioned IL- 1 Conditioned (No Cues) Negative Control IL- 1
N
Day 1 (condition)
Day 2
TABLE 2 PLASMAACTH AND CORTICOSTERONELEVELSINDUCEDBY ICV ADMINISTRATIONOF IL-1
Day 3 (test)
16 15 15
H20*/IL- 1~ sac/IL- 1 sacflL- 1
sac+ H20 H20
sac sac H20
13 13
sac/sal:~ H20/nil §
H20 H20
sac IL- 1¶
*Water in both bottles. +Five ng IL-1 in 5 I~1 apyrogenic saline delivered ICV. tSaccharin in preferred bottle at 0.1% w/v. :~Five p~l apyrogenic saline ICV. §No injection. ¶IL-1 following water exposure.
IP), the cannulae were chronically implanted and anchored over the left cerebral ventricle [A-P - 0 . 8 mm from bregma, L + 1.3 mm, D-V - 3 . 0 mm, (14)]. The injector stilette which projected 1.5 mm below the guide cannula and penetrated the ventricle, delivered 5 ~1 in approximately 90 s. The rats were unrestrained in their cages during injections. During the deprivation phase each of the bottles were weighed to determine the amount consumed and thus the preferred bottle. This continued for two days and was followed by the conditioning phase. On the conditioning day, groups receiving saccharin had the preferred bottle filled with a 0.1% w/v saccharin solution and the nonpreferred bottle filled with plain tap water for the entire 30-min drinking period. Animals not receiving saccharin had tap water in both bottles. Animals receiving injections were picked up immediately following bottle removal and were injected with either saline or 5 ng rhlL-1. Recombinant human interleukin-1 was obtained from Drs. D. Urdal and C. Henney, Immunex Corporation. On the intervening day between conditioning and the test day, the unconditioned group received saccharin in the preferred bottle and tap water in the other. On test day the preferred bottle was again filled with saccharin for those receiving the CS while the other bottle contained tap water. All other animals received tap water in both bottles. Following injection, the injector stilette was left in place for 1 rain to allow the cerebrospinal fluid to diffuse the drug. For animals not receiving an injection the bottles were simply removed after 30 min. The experimental design is shown in Table 1. The conditioning group that was reexposed to saccharin (IL-1 CONDITIONED) was compared to a group that was conditioned but not reexposed [IL-1 CONDITIONED (NO CUES)], and a group that received tap water followed by IL-1 on the training day (UNCONDITIONED). The experimental design also included a NEGATIVE CONTROL which received only the operation, apyrogenic saline injection, and saccharin, and a US only group which received the operation, limited water availability, and the drug on the test (IL-1). The rats were decapitated and trunk blood collected in heparinized tubes 2 h after the test trial. The blood was centrifuged at 3000 rpm for 3 min and the supernatant removed for the ACTH and corticosterone assays. The corticosterone assay followed standard procedure (11) and was measured as nmoles/1. ACTH was assessed using a radioimmunoassay kit obtained from Jolden Diagnostics, Scarborough, Ontario with all volumes in the kit halved and was measured as pg/ml. This procedure was validated using the standards in the kit. Dilution curves were run and found to be appropriate for assessing rat plasma ACTH levels.
ICV
N
ACTH pg/ml ~ SE
Corticosterone nmoles/1 -+ SE
saline IL-I 5 ng IL-1 100 ng
5 6 5
124 ___ 69 561 _+ 191 >1900
470 _+ 115 1124 _+ 42 1051 _+ 53
RESULTS On the basis of preliminary dose response, time course and reported studies (13), a 5 ng dose of IL-1 and a two-hour interval between the end of CS reexposure and sacrifice was selected. ACTH and corticosterone are significantly elevated to central but not peripheral administrations of this dose. Further, while corticosterone levels peak at 2 h, ACTH levels have already begun to decline but remain significantly above baseline. Thus a measurable CR should be detectable 2 h with both of these hormones. Table 2 shows the dose response curves for ICV injections of saline, 5, and 100 ng of IL-1. Each of the dependent measures were analysed using two-way factorial analysis of variance (ANOVA) with Groups and Replications as the factors. These analyses were followed where appropriate, by post hoc comparisons using the Newman-Keuls' procedure with the significance level set at the 0.05 level. The analysis of saccharin consumption was conducted only among the three groups provided with saccharin on the test. This analysis revealed a significant Group main effect, F(2,37)= 4.7, p<0.05. Neither the replications factor nor the interaction were significant. Post hoc comparisons revealed that the IL-1 CONDITIONED group compared to both the IL-1 CONDITIONED (NO CUES) group and the UNCONDITIONED group rank significantly less saccharin on the test (p<0.05, in both cases) (see Fig. 1A). Although not shown here, a preference score was calculated relating saccharin consumption to total fluid consumption for both the conditioning and test days. Total fluid consumption did not differ either on the conditioning or the test day. More importantly, while the preference for saccharin did not differ among groups on the conditioning day ( F < I . 0 ) , clear differences were observed on the test day, F(2,40)=5.25, p<0.01. That is, the IL-1 CONDITIONED group had a significantly lower preference for saccharin on the test (14%) than the UNCONDITIONED (44%) and NEGATIVE CONTROL groups (41%; both ps<0.01). The analysis of corticosterone levels indicated only a significant Group main effect, F(4,61)=21.7, p<0.01, which was accounted for by the unconditioned increase in corticosterone levels produced by the administration of IL-1. These results are shown in Fig. lB. Similarly, the analysis of the ACTH data indicated that there was a significant unconditioned, but not conditioned response, F(4,61) = 2.98, p<0.05. Thus, although the conditioning protocol was effective in producing an aversion to saccharin, it was without effect on ACTH and corticosterone production. The ACTH results are shown in Fig. 1C. DISCUSSION A significant taste aversion developed as a result of a single pairing of saccharin with an ICV injection of IL-1. To our knowledge this is the first demonstration of behavioral conditioning uti-
IL-1 CONDITIONING
693
Z ~9
A.
900-
B.
C.
18001600-
8 700 -
1400
E
-
¢.n
~8
+~ 1 2 0 0 -
8 .is
~
500 -
3
g400~300-
2
~200-
1
100-
~4
-
0
-
- -
1000 -
Q.
400-
0 i i | | i
i
i
i
200O-
"
W
°-
:'i
~
o
":~ ":~" ~
.<.> ~'--
---~-~ o ':~ " ~ " ~
-'
FIG. 1. Effects of taste aversion conditioning protocol on (A) saccharin consumption, (B) serum corticosterone levels, and (C) ACTH levels. Saccharin consumption was evaluated in groups reexposed to saccharin on the test day. Serum corticosterone and ACTH levels were taken 2 hours following the test (see Table 1).
lizing a centrally administered immunopharmacological UCS. The absence of glucocorticoid conditioning with ICV IL-1 is in striking contrast to recent similar studies with mice where cues paired with IP IL-1 resulted in significant conditioned elevations of corticosterone (11). Clearly a similar protocol with ICV administration of the peptide does not produce parallel effects. Several reasons for this contrast can be posited. The first is that rats and mice do not respond in a parallel manner. A second possible reason for the difference is that a one-bottle procedure was used with the mice and a two-bottle procedure here. It has been reported that the pituitary-adrenal response to the taste paired with sickness is a function of the conflict imposed by a one-bottle procedure (15). Therefore, a combination of interspecies and procedural factors could have contributed to the differences found in this and the previous work. Notwithstanding the aforementioned caveats, the difference in glucocorticoid conditioning effects as a function of the route of IL-1 administration, may be due to the fact that different pathways are involved in mediating these different conditioned responses. Since it seems unlikely that IL-1 crosses the blood-brain barrier (3, 5, 8), the conditioning effects observed with IP IL-1 may reflect the action of an intermediate agent. Although it has
been hypothesized that IL-1 exerts its central pyrogenic actions by entering the brain through areas in which the blood-brain barrier is absent, e.g., the organum vasculosum of the laminae terminals of the hypothalamus (the OVLT) (16); it has also been suggested that intermediate mechanisms such as prostaglandins may be responsible for mediating some of the central effects of peripheral administration of IL- 1 (7). Consistent with this view is the observation that indomethacin, a blocker of prostaglandin production, effectively eliminates the IV or ICV IL-1 induced glucocorticoid activation (13). It is possible, therefore, that the steroidogenesis resulting from peripheral administration of IL-1, which is conditionable, is mediated by different factors. It has recently been questioned whether IL-1 directly stimulates the CNS (18) or whether the toxic and aversive properties of this cytokine indirectly contribute to the observed glucocorticold activation. The dissociation observed here between behavioral and glucocorticoid responses suggests that the latter effects are not epiphenomenal to the former. Future work is needed to determine the basis of the differential conditioned response to peripheral and central administration procedures as well as a more precise characterization of the signaling mechanisms and pathways activated by the two routes of IL-1 administration.
REFERENCES 1. Ader, R.; Cohen, N. Behaviorally conditioned immunosuppression. Psychosom. Med. 37:333-340; 1975. 2. Ader, R.; Cohen, N. CNS-immune system interactions: Conditioning phenomena. Behav. Brain Sci. 8:379--394; 1985. 3. Bateman, A.; Singh, A.; Kral, T.; Solomon, S. The immune-hypothalamic-pituitary-adrenal axis. Endocr. Rev. 10:92-112; 1989. 4. Berkenbosch, F.; van Oers, J.; del Rey, A.; Tilders, F.; Besedovsky, H. Corticotropin-releasing factor producing neurons in the rat activated by interleukin-1. Science 238:524-526; 1987. 5. Coceani, F.; Lees, J.; Dinarello, D. A. Occurrence of interleukin-1 in cerebrospinal fluid of the conscious cat. Brain Res. 446:245-250; 1988. 6. Dinarello, C. A. Interleukin-1. Rev. Infect. Dis. 6:51-95; 1984. 7. Dinarello, C. A.; Bernheim, H. A. Ability of human leukocyte pyrogen to stimulate prostaglandin synthesis in vitro. J. Neurochem. 37:702-708; 1981. 8. Dinarello, C. A.; Weiner, P.; Wolff, S. M. Radiolabelling and disposition in rabbits of purified human leukocyte pyrogen. Clin. Res.
26:522A; 1978. 9. Dyck, D. G.; Driedger, S. M.; Nemeth, R.; Osachuk, T. A. G.; Greenberg, A. H. Conditioned tolerance to drug-induced (poly I:C) natural killer cell activation: Effects of drug-dosage and context-specificity parameters. Brain Behav. Immun. 1:251-266; 1987. 10. Dyck, D. G.; Greenberg, A. H.; Osachuk, T. A. G. Tolerance to drug-induced (poly I:C) natural killer cell activation: Congruence with a Pavlovian conditioning model. J. Exp. Psychol.: Anim. Behay. Proc. 12:25-31; 1986. 11. Dyck, D. G.; Janz, L.; Osachuk, T. A. G.; Falk, J.; Labinsky, J.; Greenberg, A. H. The Pavlovian conditioning of IL-1 induced glucocorticoid secretion. Brain Behav. Immun. 4:93-104. 12. Felton, D. L.; Felton, S. Y.; Carlson, S. L.; Olschowka, J.; Livvnat, S. Noradrenergic and peptidergic innervation of lymphoid tissue. J. Immunol. Suppl. 135:755-765; 1985. 13. Katsuura, G.; Gottschall, P. E.; Dahl, R. R.; Arimura, A. Adrenocorticotropin release induced by intracerebroventricular injection of recombinant human intedeukin-1 in rats: Possible involvement of
694
prostaglandin. Endocrinology 122:1773-1779; 1988. 14. Paxinos, G.; Watson, C. The rat brain in stereotaxic coordinates. New York: Academic Press; 1982. 15. Smotherman, W. P.; Hennessy, J. W.; Levine, S. Plasma corticosterone levels during recovery from LiC1 produced taste aversions. Behav. Biol. 16:401-412; 1976. 16. Stitt, J. T. Evidence for the involvement of the organum vasculosum laminae terminalis in the febrile response of rabbits and rats. J. Physiol. (Lond.) 368:501-511; 1985.
JANZ ET AL.
17. Sunder, K. S.; Becker, K. J.; Cierpial, M. A.; Carpenter, M. D.; Rankin, L. A.; Fleenr, S. L.; Ritchie, J. C.; Simson, P. E.; Weiss, J. M. Intracerebroventricular infusion of interleukin-1 rapidly decreases peripheral cellular immunity. Proc. Natl. Acad. Sci. USA 86:6398-6402; 1989. 18. Tazi, A.; Dantzer, R.; Crestani, F.; Le Moal, M. Interleukin-1 induces taste aversion in rats: a possible explanation for its pituitaryadrenal stimulating activity. Brain Res. 473:369-371; 1988.
0031-9384/91 $3.00 + .00
Conditioned Taste Aversion But Not Adrenal Activity Develops to ICV Administration of Interleukin- 1 in Rats L. J. J A N Z , * R. B R O W N , t
L. ZUO,'~ J. F A L K , * A. H. G R E E N B E R G t
A N D D. G. D Y C K .1
*Department o f Psychology and ~Manitoba Institute o f Cell Biology University o f Manitoba, Winnipeg, Manitoba, Canada R e c e i v e d 22 January 1990
JANZ, L. J., R. BROWN, L. ZUO, J. FALK, A. H. GREENBERG AND D. G. DYCK. Conditioned taste aversion but not adrenal activity develops to ICV administration of interleukin-1 in rats. PHYSIOL BEHAV 49(4) 691-694, 1991.--In a previous investigation with mice, the paired presentation of either odor or taste cues with the peripheral lIP) administration of the immunoactive peptide interleukin-1 (IL-1) led to the conditioned enhancement of glucocorticoid production. The present study found that an initial central infusion of IL- 1 in the presence of saccharin cues produced a robust taste aversion but not a conditioned elevation of either ACTH or corticosterone production. These results indicate that the glucocorticoid response induced by centrally administered IL-1 in rats is independent of the behaviorally aversive properties of this cytokine which are conditionable. The differential effects of IP versus ICV administration of IL-1 on glucocorticoid conditioning requires a clearer specification of the respective signaling mechanisms and pathways activated by these two routes of administration. Conditioned taste aversion
Adrenal activity
Interleukin-1
RESEARCH supports the existence of bidirectional communication between the central nervous system (CNS) and the immune system via neuroendocrine and sympathetic pathways (3,12). The impact of CNS signaling on primarily immunosuppressive components of the immune system has also been demonstrated through various behavioral conditioning studies (1, 2, 9, 10). The procedure in these latter studies involves pairing a distinctive environmental stimulus (conditional stimulus; CS) with a potent immunomodulatory agent (unconditional stimulus; UCS). Conditioning is most often inferred on the basis of an altered immune response to a later presentation of the CS alone (1,2). Interlenkin-1 is an immunoactive 17 kD peptide which is produced peripherally and centrally by many cell types. Its numerous target organs include the CNS where it is capable of inducing fever, altered hippocampal and hypothalamic electrical activity, and slow wave sleep (3). Peripheral administration of exogenous IL-1 can also activate the pituitary-adrenal axis resulting in elevated levels of corticotropin releasing factor (CRF), adrenocorticotropic hormone (ACTH), and glucocorticoids (4). It has been hypothesized that IL-1 is an afferent signal to the CNS and that ACTH and corticosterone are efferent hormonal outputs which could down-regulate peripheral immune responses [reviewed in (3-5)]. Although it is controversial whether IL-1 crosses the bloodbrain barrier, ICV IL-1 rapidly decreases some peripheral cellular immune responses (17). Since most conditioned alterations of
Saccharin
Glucocorticoid
immune responses are immunosuppressive, it is possible that the above pathway may participate in these conditioning phenomena. In support of this hypothesis, glucocorticoid conditioning has recently been reported in mice with IP IL-1 using both taste and odor as conditioning stimuli (11). It has also been reported that IP IL-1 supports a conditioned taste aversion in rats (18). The following experiments assessed the effect of pairing saccharin with the ICV administration of recombinant human interleukin-1 Beta (rhlL-1) on the development of a conditioned taste aversion as well as ACTH and corticosterone production. METHOD Seventy-two experimentally naive, male, Sprague-Dawley (Charles River) rats (Experiment 1 =31; Experiment 2 = 4 1 ) between 90 and 100 days old, weighing 300-350 g, were used. Animals were housed individually in polypropylene cages with a 12-h light/dark cycle with food and water available ad lib. All animals were handled daily in a manner which mimicked treatment conditions prior to surgery and habituated to a two-bottle drinking regimen. Three days after cannula implantation fluid availability was restricted to 30 min dally between 08:30 h and 10:00 h. While under general anesthesia (sodium pentobarbital, 50
1Requests for reprints should be addressed to Dr. D. G. Dyck, Department of Psychology, University of Manitoba, Winnipeg, Manitoba, Canada R3T 2N2.
691
692
JANZ ET AL.
TABLE 1 IL-1 CONDITIONINGPROTOCOLIN TASTEAVERSIONMODEL
Group Unconditioned IL- 1 Conditioned IL- 1 Conditioned (No Cues) Negative Control IL- 1
N
Day 1 (condition)
Day 2
TABLE 2 PLASMAACTH AND CORTICOSTERONELEVELSINDUCEDBY ICV ADMINISTRATIONOF IL-1
Day 3 (test)
16 15 15
H20*/IL- 1~ sac/IL- 1 sacflL- 1
sac+ H20 H20
sac sac H20
13 13
sac/sal:~ H20/nil §
H20 H20
sac IL- 1¶
*Water in both bottles. +Five ng IL-1 in 5 I~1 apyrogenic saline delivered ICV. tSaccharin in preferred bottle at 0.1% w/v. :~Five p~l apyrogenic saline ICV. §No injection. ¶IL-1 following water exposure.
IP), the cannulae were chronically implanted and anchored over the left cerebral ventricle [A-P - 0 . 8 mm from bregma, L + 1.3 mm, D-V - 3 . 0 mm, (14)]. The injector stilette which projected 1.5 mm below the guide cannula and penetrated the ventricle, delivered 5 ~1 in approximately 90 s. The rats were unrestrained in their cages during injections. During the deprivation phase each of the bottles were weighed to determine the amount consumed and thus the preferred bottle. This continued for two days and was followed by the conditioning phase. On the conditioning day, groups receiving saccharin had the preferred bottle filled with a 0.1% w/v saccharin solution and the nonpreferred bottle filled with plain tap water for the entire 30-min drinking period. Animals not receiving saccharin had tap water in both bottles. Animals receiving injections were picked up immediately following bottle removal and were injected with either saline or 5 ng rhlL-1. Recombinant human interleukin-1 was obtained from Drs. D. Urdal and C. Henney, Immunex Corporation. On the intervening day between conditioning and the test day, the unconditioned group received saccharin in the preferred bottle and tap water in the other. On test day the preferred bottle was again filled with saccharin for those receiving the CS while the other bottle contained tap water. All other animals received tap water in both bottles. Following injection, the injector stilette was left in place for 1 rain to allow the cerebrospinal fluid to diffuse the drug. For animals not receiving an injection the bottles were simply removed after 30 min. The experimental design is shown in Table 1. The conditioning group that was reexposed to saccharin (IL-1 CONDITIONED) was compared to a group that was conditioned but not reexposed [IL-1 CONDITIONED (NO CUES)], and a group that received tap water followed by IL-1 on the training day (UNCONDITIONED). The experimental design also included a NEGATIVE CONTROL which received only the operation, apyrogenic saline injection, and saccharin, and a US only group which received the operation, limited water availability, and the drug on the test (IL-1). The rats were decapitated and trunk blood collected in heparinized tubes 2 h after the test trial. The blood was centrifuged at 3000 rpm for 3 min and the supernatant removed for the ACTH and corticosterone assays. The corticosterone assay followed standard procedure (11) and was measured as nmoles/1. ACTH was assessed using a radioimmunoassay kit obtained from Jolden Diagnostics, Scarborough, Ontario with all volumes in the kit halved and was measured as pg/ml. This procedure was validated using the standards in the kit. Dilution curves were run and found to be appropriate for assessing rat plasma ACTH levels.
ICV
N
ACTH pg/ml ~ SE
Corticosterone nmoles/1 -+ SE
saline IL-I 5 ng IL-1 100 ng
5 6 5
124 ___ 69 561 _+ 191 >1900
470 _+ 115 1124 _+ 42 1051 _+ 53
RESULTS On the basis of preliminary dose response, time course and reported studies (13), a 5 ng dose of IL-1 and a two-hour interval between the end of CS reexposure and sacrifice was selected. ACTH and corticosterone are significantly elevated to central but not peripheral administrations of this dose. Further, while corticosterone levels peak at 2 h, ACTH levels have already begun to decline but remain significantly above baseline. Thus a measurable CR should be detectable 2 h with both of these hormones. Table 2 shows the dose response curves for ICV injections of saline, 5, and 100 ng of IL-1. Each of the dependent measures were analysed using two-way factorial analysis of variance (ANOVA) with Groups and Replications as the factors. These analyses were followed where appropriate, by post hoc comparisons using the Newman-Keuls' procedure with the significance level set at the 0.05 level. The analysis of saccharin consumption was conducted only among the three groups provided with saccharin on the test. This analysis revealed a significant Group main effect, F(2,37)= 4.7, p<0.05. Neither the replications factor nor the interaction were significant. Post hoc comparisons revealed that the IL-1 CONDITIONED group compared to both the IL-1 CONDITIONED (NO CUES) group and the UNCONDITIONED group rank significantly less saccharin on the test (p<0.05, in both cases) (see Fig. 1A). Although not shown here, a preference score was calculated relating saccharin consumption to total fluid consumption for both the conditioning and test days. Total fluid consumption did not differ either on the conditioning or the test day. More importantly, while the preference for saccharin did not differ among groups on the conditioning day ( F < I . 0 ) , clear differences were observed on the test day, F(2,40)=5.25, p<0.01. That is, the IL-1 CONDITIONED group had a significantly lower preference for saccharin on the test (14%) than the UNCONDITIONED (44%) and NEGATIVE CONTROL groups (41%; both ps<0.01). The analysis of corticosterone levels indicated only a significant Group main effect, F(4,61)=21.7, p<0.01, which was accounted for by the unconditioned increase in corticosterone levels produced by the administration of IL-1. These results are shown in Fig. lB. Similarly, the analysis of the ACTH data indicated that there was a significant unconditioned, but not conditioned response, F(4,61) = 2.98, p<0.05. Thus, although the conditioning protocol was effective in producing an aversion to saccharin, it was without effect on ACTH and corticosterone production. The ACTH results are shown in Fig. 1C. DISCUSSION A significant taste aversion developed as a result of a single pairing of saccharin with an ICV injection of IL-1. To our knowledge this is the first demonstration of behavioral conditioning uti-
IL-1 CONDITIONING
693
Z ~9
A.
900-
B.
C.
18001600-
8 700 -
1400
E
-
¢.n
~8
+~ 1 2 0 0 -
8 .is
~
500 -
3
g400~300-
2
~200-
1
100-
~4
-
0
-
- -
1000 -
Q.
400-
0 i i | | i
i
i
i
200O-
"
W
°-
:'i
~
o
":~ ":~" ~
.<.> ~'--
---~-~ o ':~ " ~ " ~
-'
FIG. 1. Effects of taste aversion conditioning protocol on (A) saccharin consumption, (B) serum corticosterone levels, and (C) ACTH levels. Saccharin consumption was evaluated in groups reexposed to saccharin on the test day. Serum corticosterone and ACTH levels were taken 2 hours following the test (see Table 1).
lizing a centrally administered immunopharmacological UCS. The absence of glucocorticoid conditioning with ICV IL-1 is in striking contrast to recent similar studies with mice where cues paired with IP IL-1 resulted in significant conditioned elevations of corticosterone (11). Clearly a similar protocol with ICV administration of the peptide does not produce parallel effects. Several reasons for this contrast can be posited. The first is that rats and mice do not respond in a parallel manner. A second possible reason for the difference is that a one-bottle procedure was used with the mice and a two-bottle procedure here. It has been reported that the pituitary-adrenal response to the taste paired with sickness is a function of the conflict imposed by a one-bottle procedure (15). Therefore, a combination of interspecies and procedural factors could have contributed to the differences found in this and the previous work. Notwithstanding the aforementioned caveats, the difference in glucocorticoid conditioning effects as a function of the route of IL-1 administration, may be due to the fact that different pathways are involved in mediating these different conditioned responses. Since it seems unlikely that IL-1 crosses the blood-brain barrier (3, 5, 8), the conditioning effects observed with IP IL-1 may reflect the action of an intermediate agent. Although it has
been hypothesized that IL-1 exerts its central pyrogenic actions by entering the brain through areas in which the blood-brain barrier is absent, e.g., the organum vasculosum of the laminae terminals of the hypothalamus (the OVLT) (16); it has also been suggested that intermediate mechanisms such as prostaglandins may be responsible for mediating some of the central effects of peripheral administration of IL- 1 (7). Consistent with this view is the observation that indomethacin, a blocker of prostaglandin production, effectively eliminates the IV or ICV IL-1 induced glucocorticoid activation (13). It is possible, therefore, that the steroidogenesis resulting from peripheral administration of IL-1, which is conditionable, is mediated by different factors. It has recently been questioned whether IL-1 directly stimulates the CNS (18) or whether the toxic and aversive properties of this cytokine indirectly contribute to the observed glucocorticold activation. The dissociation observed here between behavioral and glucocorticoid responses suggests that the latter effects are not epiphenomenal to the former. Future work is needed to determine the basis of the differential conditioned response to peripheral and central administration procedures as well as a more precise characterization of the signaling mechanisms and pathways activated by the two routes of IL-1 administration.
REFERENCES 1. Ader, R.; Cohen, N. Behaviorally conditioned immunosuppression. Psychosom. Med. 37:333-340; 1975. 2. Ader, R.; Cohen, N. CNS-immune system interactions: Conditioning phenomena. Behav. Brain Sci. 8:379--394; 1985. 3. Bateman, A.; Singh, A.; Kral, T.; Solomon, S. The immune-hypothalamic-pituitary-adrenal axis. Endocr. Rev. 10:92-112; 1989. 4. Berkenbosch, F.; van Oers, J.; del Rey, A.; Tilders, F.; Besedovsky, H. Corticotropin-releasing factor producing neurons in the rat activated by interleukin-1. Science 238:524-526; 1987. 5. Coceani, F.; Lees, J.; Dinarello, D. A. Occurrence of interleukin-1 in cerebrospinal fluid of the conscious cat. Brain Res. 446:245-250; 1988. 6. Dinarello, C. A. Interleukin-1. Rev. Infect. Dis. 6:51-95; 1984. 7. Dinarello, C. A.; Bernheim, H. A. Ability of human leukocyte pyrogen to stimulate prostaglandin synthesis in vitro. J. Neurochem. 37:702-708; 1981. 8. Dinarello, C. A.; Weiner, P.; Wolff, S. M. Radiolabelling and disposition in rabbits of purified human leukocyte pyrogen. Clin. Res.
26:522A; 1978. 9. Dyck, D. G.; Driedger, S. M.; Nemeth, R.; Osachuk, T. A. G.; Greenberg, A. H. Conditioned tolerance to drug-induced (poly I:C) natural killer cell activation: Effects of drug-dosage and context-specificity parameters. Brain Behav. Immun. 1:251-266; 1987. 10. Dyck, D. G.; Greenberg, A. H.; Osachuk, T. A. G. Tolerance to drug-induced (poly I:C) natural killer cell activation: Congruence with a Pavlovian conditioning model. J. Exp. Psychol.: Anim. Behay. Proc. 12:25-31; 1986. 11. Dyck, D. G.; Janz, L.; Osachuk, T. A. G.; Falk, J.; Labinsky, J.; Greenberg, A. H. The Pavlovian conditioning of IL-1 induced glucocorticoid secretion. Brain Behav. Immun. 4:93-104. 12. Felton, D. L.; Felton, S. Y.; Carlson, S. L.; Olschowka, J.; Livvnat, S. Noradrenergic and peptidergic innervation of lymphoid tissue. J. Immunol. Suppl. 135:755-765; 1985. 13. Katsuura, G.; Gottschall, P. E.; Dahl, R. R.; Arimura, A. Adrenocorticotropin release induced by intracerebroventricular injection of recombinant human intedeukin-1 in rats: Possible involvement of
694
prostaglandin. Endocrinology 122:1773-1779; 1988. 14. Paxinos, G.; Watson, C. The rat brain in stereotaxic coordinates. New York: Academic Press; 1982. 15. Smotherman, W. P.; Hennessy, J. W.; Levine, S. Plasma corticosterone levels during recovery from LiC1 produced taste aversions. Behav. Biol. 16:401-412; 1976. 16. Stitt, J. T. Evidence for the involvement of the organum vasculosum laminae terminalis in the febrile response of rabbits and rats. J. Physiol. (Lond.) 368:501-511; 1985.
JANZ ET AL.
17. Sunder, K. S.; Becker, K. J.; Cierpial, M. A.; Carpenter, M. D.; Rankin, L. A.; Fleenr, S. L.; Ritchie, J. C.; Simson, P. E.; Weiss, J. M. Intracerebroventricular infusion of interleukin-1 rapidly decreases peripheral cellular immunity. Proc. Natl. Acad. Sci. USA 86:6398-6402; 1989. 18. Tazi, A.; Dantzer, R.; Crestani, F.; Le Moal, M. Interleukin-1 induces taste aversion in rats: a possible explanation for its pituitaryadrenal stimulating activity. Brain Res. 473:369-371; 1988.