Developmental Brain Research 98 Ž1997. 145–149
Short communication
Neonatal caffeine alters passive avoidance retention in rats in an ageand gender-related manner Susan Fisher, Ronnie Guillet
)
Department of Pediatrics (Neonatology), Box 651, The Children’s Hospital at Strong, UniÕersity of Rochester School of Medicine and Dentistry, 601 Elmwood AÕe., Rochester, NY 14642, USA Accepted 27 August 1996
Abstract Chronic administration of an adenosine receptor antagonist disturbs spatial learning and memory in adult mice and neonatal caffeine exposure results in long-term behavioral and biochemical sequelae in mice and rats. We thus postulated that early treatment with caffeine would have latent effects on learning and memory as measured in a passive avoidance paradigm. Rats were not handled or received caffeine Ž15–20 mgrkgrday. by gavage over postnatal days 2–6. At 28 or 70–90 days of age, rats were trained to avoid an electrified grid and tested for retention 24 h, 72 h, and 7 days later. At 28 days, caffeine-exposed rats required more trials to meet criterion than did control rats, regardless of gender. There was minimal effect on retention of either neonatal treatment or gender at this age. At 70–90 days, there was no effect of either gender or treatment on learning; however, there was a significant effect of gender Ž P - 0.05. on retention at 24 h that was more pronounced in neonatally caffeine-treated rats Ž P - 0.01.. At 72 h, the effect of caffeine on retention differed between male and female rats. Neonatal caffeine exposure significantly improved retention in females Ž P - 0.01. and significantly decreased retention in males Ž P - 0.05.. Thus, caffeine exposure limited to the first week of life resulted in alterations in passive avoidance retention that became apparent over pubertal development. These changes were a function of the gender of the animal. Keywords: Caffeine; Passive avoidance; Ontogeny; Gender; Learning; Memory
Caffeine, a methylxanthine, is used therapeutically to treat apnea in preterm infants. In addition, it is widely consumed by humans in soft drinks, tea and coffee, as well as in over-the-counter medications. At moderate doses Ž30–100 mmolrkg., it is a central nervous system stimulant w14x that diffuses throughout the body in a volume of distribution similar to that of body water and is able to cross the placenta to the fetus. It quickly penetrates the brain either by diffusion or through a saturable carriermediated transport system w9x and maintains a stable brain concentration for at least an hour in the rat w15x. Central levels of behaviorally effective doses of various methylxanthines, including caffeine, are consistent with their affinities for adenosine receptors w14x, suggesting that methylxanthines elicit their behavioral responses by blocking central adenosine receptors.
) Corresponding author. Fax: Ž1. Ž716. 461-3614; E-mail:
[email protected]
Previous research in this laboratory w2–6x has shown that neonatal exposure to therapeutic doses of caffeine produces long-term behavioral and neurochemical consequences. Caffeine administered to rat pups 2–6 days of age, a time period during which brain development is comparable to that of the 26–38 week gestation human, at dosages comparable to those used therapeutically for apnea of prematurity, delays the development of a stimulatory effect of acute administration of caffeine on locomotor activity w3x. In contrast, early exposure to caffeine accelerates the development of a locomotor response to acute administration of D-phenylisopropyl adenosine ŽD-PIA., an analog of adenosine w6x. Neonatal caffeine exposure also alters adult seizure thresholds for a variety of chemoconvulsants w4x. Furthermore, neonatal caffeine exposure accelerates attainment of adult adenosine A1 receptor density in rat cortex, cerebellum, thalamus, and hippocampus, but not in the brain stem or hypothalamus w2,5x. Caffeine, given acutely, affects cognitive function. Acute administration of caffeine enhances learning and
0165-3806r97r$17.00 Copyright q 1997 Elsevier Science B.V. All rights reserved. PII S 0 1 6 5 - 3 8 0 6 Ž 9 6 . 0 0 1 5 8 - 7
146
S. Fisher, R. Guilletr DeÕelopmental Brain Research 98 (1997) 145–149
memory abilities of mice in various tasks w1x and, depending on the degree of novelty of the task and on the involvement of a food reward, has been shown to enhance memory and learning in male rats w11x. The acute effects of caffeine on learning and memory are thus postulated to be related in part to its actions involving arousal and vigilance. Chronic treatment with adenosine receptor ligands appears to produce behavioral consequences opposite to those produced by acute administration. Von Lubitz et al. w17x studied effects on spatial memory acquisition produced by daily administration of either N6-cyclopentyladenosine ŽCPA, 5–30 mgrkg, i.p.. or 8-cyclopentyl-1,3-dipropylxanthine ŽCPX, 0.2–1.0 mgrkg, i.p.. for 9 days. Chronic administration of CPA, an adenosine receptor agonist, enhanced, while chronic treatment with CPX, an antagonist, impaired performance in a water maze. This is the opposite of the effects seen following acute administration of these agents. Little research has been done to assess the effects of neonatal caffeine exposure on learning and memory. The goal of the present study was to investigate the permanent andror latent behavioral consequences of neonatal caffeine exposure. Since chronic administration of an adenosine receptor antagonist disturbs spatial learning and memory in adult mice, and since neonatal caffeine exposure has previously been shown to result in long-term behavioral and biochemical sequelae in mice and rats, we postulated that early treatment with caffeine would have latent effects on learning and memory as measured in a passive avoidance paradigm. Long Evans rats, 2–90 days of age, born to mature females obtained from Harlan Sprague-Dawley ŽAltamont, NY. and bred to in-house Long Evans males, were used in this study. Litters were culled to 10–12 pups at postnatal day 1–2 Žpostnatal day 0 s day of birth. and randomly assigned to an experimental protocol Žsee below.. Rats were weaned at 26–28 days and housed according to litter and gender. Subjects were moved to individual cages the day before passive avoidance training and testing began. All animals received food and water ad libitum and were maintained on a 12 h light cycle Žlights on at 0600 h. under conditions of consistent temperature Ž72 " 28F. and humidity. All investigations were approved by the University of Rochester Committee on Animal Resources. Pups 2–6 days of age were either left not handled ŽNH. or given caffeine citrate ŽCAFF, Sigma Chemicals. by gavage as follows: 20 mgrkg on day 2 and 15 mgrkg on days 3–6, administered in a volume of 0.05 mlr100 g body weight. This procedure has been shown to result in blood caffeine levels of 5–15 mgrl over a 24 h period w3x. A separate cohort of pups were given water administered in a volume of 0.05 mlr100 g body weight Žgavage controls.. All pups in the same litter received the same treatment. Rats from 3 or more litters were used for each experimental condition. Eight subjects comprised each ex-
perimental group at 27–28 and at 70–90 days, except as noted. Subjects were tested by an observer blind to neonatal treatment. Behavioral tests were performed between 1300 and 1600 h in a dimly illuminated Ž40 lux., acoustically isolated room. An exhaust fan provided masking noise. Horizontal activity was measured the day before passive avoidance training to ascertain that differences in performance were not due to differences in locomotor activity. A transparent Plexiglas testing chamber Ž10 = 18 inches. with infrared photosensors evenly spaced 2 inches apart, 1 inch above the floor, arranged in a grid pattern, was used to measure horizontal activity. The number of grid crossings was automatically totaled as the number of interruptions of a photocell beam per test session. Rats were trained on the passive avoidance paradigm 24 h after spontaneous activity measurement. The testing apparatus Ž9.5 = 20 inches. consisted of two compartments equal in size: one with a floor made of 0.12 inch stainless steel rods spaced 0.5 inches apart, and one with a translucent Plexiglas raised platform Ž1 inch above grid floor.. Footshock Ž0.5 mA, alternating current. was delivered to the grid floor through a Coulbourn Instruments Programmable Animal Test Cage Grid Floor Shocker Žmodel PM-720.. The walls of the apparatus were constructed of translucent Plexiglas. A black Plexiglas guillotine door separated the two compartments. Rats were housed in their individual cages in a soundproof room until training began. They were then allowed to explore a translucent Plexiglas holding cage for 60 s before being placed on the platform of the passive avoidance apparatus. After 30 s the door separating the two compartments was raised and step-through latency was recorded. A maximum of 180 s was allowed for the rat to step onto the gird with four feet. When the rat stepped onto the grid, the door closed and 3 footshocks of 1 s duration each were delivered 10 s apart. The rat was returned to the holding cage for 30–90 s while the apparatus was cleaned with a damp cloth. To assess learning, the rat was replaced on the platform and training repeated until the animal withheld crossing onto the grid. The number of trials required for the rat to withhold crossing Žtrials to criterion. was recorded. The criterion for withhold in the initial experiments was 60 s. In subsequent testing to assess learning during training, the criterion was increased to 5 min to correspond to the duration of the testing period for the retention sessions. To assess retention, rats were tested 24 h, 72 h, and 7 days post-training using the same procedure except that the step-through latency Žup to 5 min. was measured and no footshock was administered. The two-tailed Wilcoxon–Kruskal-Wallis ranked sums test was used to analyze activity, initial cross and trial to criterion. Testing at 24 h, 72 h, and 7 days was evaluated using the x 2 likelihood ratio. Statistical significance was attributed when P - 0.05.
S. Fisher, R. Guilletr DeÕelopmental Brain Research 98 (1997) 145–149
147
Fig. 1. Percentage of adult rats Ž70–90 days of age. that met criterion for retention as a function of neonatal treatment and gender. Control animals ŽNH. were not handled over postnatal days 2–6 Ž n s 8 each, males and females.. Caffeine-exposed rats ŽCAFFEINE. received 15–20 mgrkg caffeine by gavage over postnatal days 2–6 Ž n s 8 each, males and females.. Data are presented as % of each group that did not cross onto the grid within 5 min, passive avoidance retention, at ŽA. 24 h, ŽB. 72 h, and ŽC. 7 days after the training session. ) Significant difference, P - 0.01; ) ) significant difference, P - 0.001 for the gender andror neonatal treatment group at the testing interval.
In the 70–90 day old rats there were no differences between NH and gavage controls Ž n s 5 males and 4 females. in any of the measures tested Ždata not shown.. Locomotor activity was a function of gender, but not treatment. Female rats had significantly higher activity levels than male subjects Ž121 " 7 vs. 101 " 6 photocell interruptionsrmin, mean " S.E.M., P - 0.05.. Two rats, both male NH, failed to cross onto the grid within 180 s of being placed in the chamber Žprior to shock. and were eliminated. There was no effect of neonatal treatment or gender on pre-shock step-through latency or on trials to criterion. Most animals required only one trial to meet the criterion. Retention tested 24 h, 72 h and 7 days after training is illustrated in Fig. 1. The percentage of rats that met criterion Ži.e. did not cross onto the grid within 5 min, passive avoidance retention. is shown as a function of neonatal treatment and gender. No effect of treatment was seen at 24 h. A significant effect of gender Ž P - 0.05. was seen at 24 h, with 81% of female rats, but only 50% of males, displaying retention of the training. This effect was more pronounced in neonatally caffeine-treated rats Ž P 0.01.. All female CAFF showed passive avoidance retention at 24 h, while only 38% of male CAFF did. At 72 h, neonatal treatment with CAFF significantly affected retention rates in male and female rats, relative to NH. However, the effect of CAFF on passive avoidance
differed between male and female rats. In females, the neonatal caffeine treatment significantly improved retention scores Ž P - 0.01.. 88% of female CAFF showed passive avoidance, while only 25% of female NH met criterion. In males, the neonatal caffeine treatment significantly decreased retention Ž P - 0.05.. Only 13% of male CAFF met criterion at 72 h, compared with 67% of male NH. No significant effects of neonatal treatment or gender were seen at 7 days; however, female CAFF continued to have the highest retention rate. Passive avoidance retention was 50% and 13% in female CAFF and NH, respectively. In males, 13% of CAFF and 17% of NH met criterion. In 27–28 day old rats there were no differences between NH and gavage controls Ž n s 3 males and 3 females. in any of the measures tested Ždata not shown.. Locomotor activity and initial step-through latency were not affected by either gender or neonatal treatment when animals were tested at 28 days of age. Five animals Ž1 male CAFF, 1 female CAFF, 3 female NH. were eliminated from the experiment for failing to cross onto the grid within 180 s during the training session. There was an effect of neonatal treatment on trials to criterion. CAFF rats took more trials to meet criterion than did NH rats Ž1.9 " 0.3 vs. 1.4 " 0.2, mean " S.E.M., P 0.05.. Gender had no significant effect on trials to criterion.
Fig. 2. Percentage of juvenile rats Ž27–28 days of age. that met criterion for retention as a function of neonatal treatment and gender. Control animals ŽNH. were not handled over postnatal days 2–6 Ž n s 8 each, males and females.. Caffeine-exposed rats ŽCAFFEINE. received 15–20 mgrkg caffeine by gavage over postnatal days 2–6 Ž n s 8 each, males and females.. Data are presented as % of each group that did not cross onto the grid within 5 min, passive avoidance retention, at ŽA. 24 h, ŽB. 72 h, and ŽC. 7 days after the training session. A significant effect of neonatal treatment on retention was only apparent at 24 h post-training Ž ) P - 0.05..
148
S. Fisher, R. Guilletr DeÕelopmental Brain Research 98 (1997) 145–149
A significant effect of neonatal treatment on retention was only apparent at 24 h post-training Ž P - 0.05.. At 24 h ŽFig. 2., 88% male NH withheld crossing, while only 38% of male CAFF retained passive avoidance. Female CAFF and NH were no different Ž63% each group.. There was no difference in retention as a function of neonatal treatment at either 72 h or 7 days. No effect of gender was seen on any test day. We have demonstrated that early neonatal exposure to caffeine has effects on performance in a passive avoidance paradigm, a task generally thought to test learning and memory, when tested in the juvenile period Ž28 days. or in young adulthood Ž70–90 days.. These effects were gender-specific in adult, but not in juvenile, rats. Specifically, female rats exposed to caffeine at an early age showed enhanced retention of a passive avoidance task when tested as adults, while male rats displayed decreased retention when so exposed. These differences could not have been due to alterations in activity since no significant differences between treatment groups were noted in spontaneous locomotor activity or pre-shock step-through latencies. In fact, females showed better retention of passive avoidance than did males even though their baseline activity was greater. In those animals tested at 28 days of age, caffeine-exposed male and female rats required more trials than nonhandled rats to meet the criterion for learning; however, once learned, there was no difference in retention of this task as a function of either early exposure or gender. Caffeine may exert its effects on passive avoidance retention in adult rats through its interaction with adenosine receptors. Chronic administration of the adenosine receptor antagonist CPX impairs the performance of male rats in a water maze w17x. In addition, neonatal administration of caffeine increases adenosine A1 receptor sensitivity in the cerebral cortex, cerebellum, and hippocampus of adult rats, apparently through an increase in maximal binding density w6x. The neonatal period is a critical time in the development of the adenosine A1 receptor system. Adult levels of A1 receptors are not attained in the rat cerebellum until at least 42 days of age; in the hippocampus, adult levels are evident at 21 days of age; and in the cerebral cortex, brain stem, and hypothalamus, at 14–21 days of age w8x. Neonatal caffeine may accelerate the attainment of adult densities of A1 receptors in the cerebellum, hippocampus, and cerebral cortex w5x. Thus, as previously suggested by Guillet and Kellogg w5x, the proper development of behavioral functions such as memory may depend on the appropriate ontogeny of the adenosine receptor system. Previous research in this laboratory has also shown that neonatal caffeine exposure raises seizure thresholds in adult rats for a variety of chemoconvulsants w4x. Sanders and Murray w12x have reported a temporal correlation between upregulation of adenosine A1 receptors and decreased bicuculline seizure susceptibility in adult rats
chronically treated with the methylxanthine theophylline. These results taken together suggest a possible anticonvulsant effect of methylxanthines produced through changes in the adenosine receptor system. Of note, the increase in seizure threshold in adult rats treated with caffeine as neonates was sexually dimorphic, with the greatest effect on threshold occurring in the female rats. Sexual dimorphism in the effects of caffeine exposure on adult behaviors has been previously reported. In a similar experiment, Sinton et al. w13x found no significant effects of gestational caffeine Ž60, 80, or 100 mgrkgrday in drinking water. on the adult offspring of one strain of mice, but in the more active strain, caffeine increased open field activity and spontaneous alternation in females only. A significant effect of caffeine on passive avoidance retention was demonstrated when mice of the more active strain were tested 15 min after training, but not at later testing times. Females showed the highest retention at the highest dose of caffeine, while males showed the highest retention at the lowest dose. In our initial studies with 70–90 day old rats, the rats were allowed 180 s to cross onto the grid with all four feet. Passive avoidance learning following footshock, established if the rat failed to step through within 60 s, may not have been indicative of passive avoidance. Step-through latencies of male rats have been shown to increase as a function of pre-shock trial w16x, suggesting that habituation may influence rats’ exploratory behavior. It is possible that latencies to cross while assessing ‘learning’ of passive avoidance increased due to habituation to the procedure, rather than because of any passive avoidance that may have taken place. However, of those rats that initially stepped through within 180 s, only 2 Ž7%. took longer than 60 s and only 5 Ž17%. took longer than 45 s. In the study reported by Van Haaren and Van de Poll w16x, step-through latencies did not increase dramatically between the first and the third pre-shock trials. Therefore, it seems that even including the rats with pre-shock latencies greater than 45 s, most rats would demonstrate passive avoidance by not crossing within 60 s on the learning trial. Thus, 60 s appears to have been sufficient time to assess learning. The possible complication with this 60 s learning assessment period was alleviated in the experiments with 28 day old rats. The duration of the assessment period was increased to 5 min, corresponding to the duration of the test period. The failure of neonatal caffeine exposure to affect later spontaneous locomotor activity in juvenile rats is in agreement with previous findings. In rats neonatally exposed to caffeine using the same model, no significant effect of caffeine on baseline locomotor activity was demonstrated at 15, 18 or 28 days of age w3x. Since no effect was found in the present experiments in which animals were tested at 26–28 days and at 70–90 days of age, it is likely that early caffeine exposure does not have a long-term influence on spontaneous locomotor activity. There is a paucity of literature on the passive avoidance
S. Fisher, R. Guilletr DeÕelopmental Brain Research 98 (1997) 145–149
performance of young rats neonatally administered caffeine. In the present study caffeine-treated animals took more trials than nonhandled rats to meet the criterion for ‘learning’ passive avoidance. Apparently, neonatal treatment with caffeine diminishes a rat’s ability to enact passive avoidance behaviors. This could be a result of a deficit in memory formation or an impaired ability to inhibit exploratory behaviors. Since neither activity nor initial step-through latencies varied among treatment groups, the latter possibility does not seem to explain the difference between the number of trials to criterion in the two experimental groups. Although we have postulated that early caffeine exerts its effects on behavior via the adenosine receptor system, it may alternatively alter retention of a passive avoidance task by influencing the central serotonergic system. Acute administration of caffeine increases in vitro concentrations of serotonin in the brain stem, cerebral cortex, and cerebellum by reducing serotonin availability at post-synaptic receptor sites w10x. Further, in untreated rats, sexual dimorphism in passive avoidance behavior has been shown to be dependent on the integrity of the central serotonergic system w7x. In summary, neonatal exposure to caffeine alters retention of a passive avoidance task in adult rats, but not in juvenile rats. Conversely, it impairs learning of this task in young, but not adult rats. These behavioral responses may result from an interaction of caffeine with the developing adenosine receptor system. Additionally, since passive avoidance retention is not sexually dimorphic in prepubertal rats, but is in adult rats, it appears that the sex differences in passive avoidance behavior in response to neonatal caffeine could be affected by hormonal influences on the adenosine receptor system. Extrapolation of these results to humans suggests that caffeine administered therapeutically for apnea of prematurity could have long-term behavioral consequences. Thus, these findings may have implications for the clinical use of caffeine in human neonates.
Acknowledgements The authors thank Dr. Carol Kellogg for helpful discussion and Ms. Karen Ferguson for excellent technical assistance. This research was supported in part by The Strong Children’s Research Center.
149
References w1x Castellano, C., Effects of caffeine on discrimination learning, consolidation, and learned behavior in mice, Psychopharmacology, 48 Ž1976. 255–260. w2x Etzel, B. and Guillet, R., Effects of neonatal exposure to caffeine on adenosine A1 receptor ontogeny using autoradiography, DeÕ. Brain Res., 82 Ž1994. 223–230. w3x Guillet, R., Neonatal caffeine exposure alters adenosine receptor control of locomotor activity in the developing rat, DeÕ. Pharmacol. Ther., 15 Ž1990. 94–100. w4x Guillet, R. and Dunham, L., Neonatal caffeine exposure and seizure susceptibility in adult rats, Epilepsia, 36 Ž1995. 743–749. w5x Guillet, R. and Kellogg, C.K., Neonatal exposure to therapeutic caffeine alters the ontogeny of brain adenosine A1 receptors in rats, Neuropharmacology, 30 Ž1991. 489–496. w6x Guillet, R. and Kellogg, C.K., Neonatal caffeine exposure alters developmental sensitivity to adenosine receptor ligands, Pharmacol. Biochem. BehaÕ., 40 Ž1991. 811–817. w7x Heinsbroek, R.P.W., Feenstra, M.G.P., Boon, P., Van Haaren, F. and Van de Poll, N.E., Sex differences in passive avoidance depend on the integrity of the central serotonergic system, Pharmacol. Biochem. BehaÕ., 31 Ž1988. 499–503. w8x Marangos, P.J., Boulenger, J.-P. and Patel, J., Effects of chronic caffeine on brain adenosine receptors: regional and ontogenetic studies, Life Sci., 34 Ž1984. 899–907. w9x McCall, A.L., Millington, W.R. and Wurtman, R.J., Blood-brain barrier transport of caffeine: dose-related restriction of adenine transport, Life Sci., 31 Ž1982. 2709–2715. w10x Nehlig, A., Daval, J.L. and Debry, G., Caffeine and the central nervous system: mechanisms of action, biochemical, metabolic and psychostimulant effects, Brain Res. ReÕ., 17 Ž1992. 139–170. w11x Oliviera, E.M., Rubin, M.A., Belloi, C.R.B., Belloi, M.H.B. and Rocha, J.T.B., Effect of caffeine administration on latent learning ability of male rats in a simple maze task, Brazil. J. Med. Biol. Res., 23 Ž1990. 975–980. w12x Sanders, R.C. and Murray, T.F., Temporal relationship between A1 adenosine receptor upregulation and alterations in bicuculline seizure susceptibility in rats, Neurosci. Lett., 101 Ž1989. 325–330. w13x Sinton, C.M., Valatx, J.L. and Jouvet, M., Gestational caffeine modifies offspring behavior in mice, Psychopharmacology, 75 Ž1981. 69–74. w14x Snyder, S.H., Katims, J.J., Annau, Z., Bruns, R.F. and Daly, J.W., Adenosine receptors and behavioral actions of methylxanthines, Proc. Natl. Acad. Sci. USA, 78 Ž1981. 3260–3264. w15x Thithapandha, A., Maling, H.M. and Gillette, J.R., Effects of caffeine and theophylline on activity of rats in relation to brain xanthine concentrations, Proc. Soc. Exp. Biol. Med., 139 Ž1972. 582–586. w16x Van Haaren, F. and Van de Poll, N.E., The number of pre-shock trials affects sex differences in passive avoidance behavior, Physiol. BehaÕ., 33 Ž1984. 269–272. w17x Von Lubitz, D., Paul, I.A., Bartus, R.T. and Jacobson, K.A., Effects of chronic administration of adenosine A1 receptor agonist and antagonist on spatial learning and memory, Eur. J. Pharmacol., 249 Ž1993. 271–280.