Perinatal asphyxia, hyperthermia and hyperferremia as factors inducingbehavioural disturbances in adulthood: A rat model
Michał Caputa∗, Justyna Rogalska, Katarzyna Wentowska, Anna NowakowskaDepartment of Animal Physiology, Institute of General and Molecular Biology, N. Copernicus University, ul. Gagarina 9, 87-100 Toru´n, Poland
Received 25 April 2005; received in revised form 16 May 2005; accepted 19 May 2005Available online 20 July 2005
Abstract
Alertness was studied in adult male Wistar rats after neonatal critical anoxia applied under three different thermal conditions: (i) atphysiological neonatal body temperature of 33◦C, (ii) at body temperature elevated to 37◦C, and (iii) at body temperature elevated to 39◦C(both during anoxia and for 2 h postanoxia). To elucidate the effect of iron-dependent postanoxic oxidative damage to the brain, half of thegroup (iii) was injected with deferoxamine, a chelator of iron. Postanoxic behavioural disturbances were recorded in open-field, elevatedplus-maze, and sudden silence tests when the rats reached the age of 4 month. Moreover, spontaneous motor activity of the rats was recordedr ubjected ton and suddens rbances werep
adiotelemetrically in their home-cages. Both open-field stress-induced and spontaneous motor activity were reduced in rats seonatal anoxia under hyperthermic conditions. In contrast, these rats were hyperactive in the plus-maze test. Both the plus-mazeilence tests revealed that these rats show reduced alertness to external stimuli signalling potential dangers. The behavioural disturevented by the body temperature of 33◦C and by postanoxic administration of deferoxamine.These data support the conclusion that permanent postanoxic behavioural disturbances are due to iron-dependent oxidative d
rain, which can be prevented by the reduced neonatal body temperature.2005 Elsevier B.V. All rights reserved.
eywords:Rats; Neonatal asphyxia; Body temperature; Hyperferremia; Deferoxamine; Behavioural disturbances
. Introduction
In our previous study[30] we have shown that neona-al asphyxia in rats results in stress-induced hyperactivity,eveloping in juvenile (5–45 days old) animals. Physiologi-al neonatal body temperature of 33◦C prevented, while bodyemperature elevated to a febrile level of 39◦C enhancedhe behavioural disturbances. Because chelation of iron witheferoxamine (DF) protected the animals against the delayedehavioural abnormalities we have concluded that iron-ependent oxidative damage to the brain contributed to thesebnormalities[30]. Indeed, there is a growing body of evi-ence that oxidative stress plays an important role in neu-onal damage both in animal models of neonatal hypoxia and
reoxygenation[1,24,38]and in babies suffering from birasphyxia[28,43]. A complementary piece of informatiois finding that iron supplementation in the critical neonperiod leads to oxidative stress in adult rats[9].
According to some authors[4,8] postischemic hypothemia provides effective and long-lasting neuroprotectionothers[12,40] claim that hypoxic-ischemic neurotoxicitymarkedly delayed but not prevented by post-insult hypomia. In contrast, excessive body temperature enhances pchemic damage to the brain[13,29]. Therefore, the preseinvestigation was undertaken to examine the effects: (both the physiological body temperature (33◦C) and excessive body temperature (37 and 39◦C) in neonatal rats exposto a critical anoxia, and (ii) of postanoxic chelation of ironneonatal rats exposed to both the critical anoxia and hthermia on stress responses of the animals at the agemonths. Moreover, we decided to apply some additional(in addition to open-field test, which was used in our prev
M. Caputa et al. / Behavioural Brain Research 163 (2005) 246–256 247
investigation[30]) to detect anxiety and alertness in the adultrats subjected neonatally to the above-mentioned factors. Tohave a reliable reference to the stress responses we continuedalso radiotelemetric recording of spontaneous motor activ-ity [30] in the above-mentioned adult rats when they wereprotected from any stress in their home-cages.
2. Materials and methods
All experiments described in the present paper have beenapproved by the Local Committee on the Use and Care of Labo-ratory Animals (permission No. 6/2003).
2.1. Animals
A total of 128 newborn Wistar rats of 2 days of age, of both sexes,weighting 7–8 g were used. They were taken from their mothers(housed in acrylic cages lined with wood shavings, at a room temper-ature of 20–22◦C) and were exposed to a critical anoxia at variousbody temperatures. To minimize stress of bereavement the moth-ers were left with at least one suckling. The remaining sucklingswere divided equally among the experimental groups described inthe next section. Postanoxic behavioural disturbances in open-field,elevated plus-maze, and sudden silence tests as well as changes inthe spontaneous motor activity and body temperature rhythms underc n theyr k con-d afterw idedw exesa . Ther .
ferent levels: (i) 33◦C – which is the normal body temperature ofnewborn rats[31], (ii) 37 ◦C – which is the body temperature ofhealthy adults (it is also regarded as an appropriate body temperaturefor human neonates nursed in incubators), and (iii) 39◦C – which isbody temperature typical of febrile adults. Since we have previouslyshown that anoxia at 39◦C resulted in a transient hyperferremia[3], in the present investigation additional eight rats, exposed toanoxia at this particular body temperature, were injected with DF(deferoxamine mesylate, Sigma-Aldrich, Chemie, GmbH, Stein-heim, Germany), a chelator of iron, to match the non-treated anoxicanimals. The drug was injected subcutaneously (100 mg/kg s.c.)twice: immediately after anoxia and 24 h later.
The anoxia was continued for 25 min[2,22]or it was interruptedearlier (to prevent lethal asphyxia) when a critical accelerated andshallow gasping phase occurred[3,31].
After anoxia the animals were exposed to atmospheric air atunchanged temperature for 120 min.
Control rats, which were also placed in the plethysmographicchambers, were exposed to atmospheric air throughout the sameperiod (145–130 min) in the respective thermal conditions.
2.3. Behavioural tests
The experiments were started when the rats were 4 months old.The behavioural tests were performed in the following order: (i)control radiotelemetric recording of daily changes in locomotoractivity and body temperature for 2–4 days; (ii) exposure to open-fi itya ele-v equentr andb
roomw tantt in ther med.T f thee al. Toa peri-m 09:00a
vatedp d thena gen,N
2of the
a -fieldap toryt pen-fi lourm airC thee
ed:1 erem otald uring
ontrol and poststress conditions were recorded in the rats wheeached the age of 4 months. The rats were raised at light:daritions of 12 h:12 h (lights on 07.00–19.00 h), both before andeaning. They were fed with commercial chow pellets and provith water ad libitum. The weanlings were sorted according to snd 63 males were selected for the behavioural experimentsats were housed individually until the end of the experiments
.2. Exposure to anoxia
The methods used to elicit anoxia, to manipulate neoody temperature, and pharmacological treatment of the ratsescribed in details in our previous work[30]. Briefly, rectal temperture of newborn rats was recorded by means of miniature (0.
n diameter) copper–constantan thermocouples (Physitemp Ients, Clifton, USA) using Keithley Instruments (Cleveland, Opparatus (182 sensitive digital voltmeter and 705 scanner)ected to a data acquisition system, which in turn was connusing GPIB card) to a computer (PC 386/25 MHz). Temperaata were collected at 30 s intervals. During the recordings ratslaced individually in 200 ml plethysmographic chambers, were submerged, under controlled thermal conditions, in a w
hermostat.To elicit anoxia, the plethysmographic chambers were flu
ith 100% nitrogen. Gasping movements were continuoecorded in asphyxiated unrestrained rat pups, using the baroethod.Five to seven animals were used for each experimental g
rior to experiments thermocouples, connected to the recordinem, were inserted 5 mm into the rectum of the rat pups. Theups, divided into three experimental (exposed to anoxia) andontrol groups, were placed in the plethysmographic chamberst different temperatures to stabilize body temperature at thre
eld followed by recording of daily changes in locomotor activnd body temperature for 2 days; (iii) 5 min exposures to theated plus-maze in three successive days combined with subsadiotelemetric recording of daily changes in locomotor activityody temperature; (iv) exposure to a sudden silence.
Each behavioural test was performed in sound-isolatedith lighting conditions and environmental cues held cons
hroughout testing. During the tests the experimenter stayedoom adjacent to the one in which the experiments were perforo remove the smell traces left by rats, the floor and walls oquipment were carefully cleaned before testing the next animvoid influences of circadian rhythms on performance of the exental animals each behavioural test was carried out betweennd 10:00.
The behaviour of rats during exposure to the open-field, elelus-maze as well as to the sudden silence was videotaped annalyzed with the EthoVision 2.3 software (Noldus, Wageninetherlands).
.3.1. Open-field testThe open-field test was used to evaluate responsiveness
nimals to a stressogenic and novel environment. The openpparatus was the square wooden box (1.14 m× 1.14 m× 0.44 m),ainted white. In addition to the ambient light in the labora
wo bulbs of 100 W were placed 1 m above the centre of the oeld. For the optimal tracking performance, the animals were coarked with commercial hair dye (Schwarzkopf PolyColor Holour Cream, Dusseldorf, Germany) 3 days before start ofxperiment.
The following variables of the rat’s behaviour were analyz. Total movement – the total period during which animals woving of the total time of the open-field exposure (%). 2. Tistance moved – the total distance travelled by the animal d
248 M. Caputa et al. / Behavioural Brain Research 163 (2005) 246–256
Fig. 1. A side view of the elevated plus-maze apparatus: c – the closed arms,o – the open arms, s – the central open square.
the entire exposure to open-field (m). 3. Mean linear velocity – thedistance moved per unit of time (mm/s). 4. Mean angular velocity– the mean angular change in direction of movement of the rat perunit time (◦/s).
Each test lasted 10 min. The experiments started with the place-ment of animals in the centre of the arena. After the test they werereturned to their home-cages.
2.3.2. Elevated plus-mazeFor further testing the anxiety-guided behaviours, rats were
exposed to the elevated plus-maze task. The procedure of the testhas been described in detail by Pellow et al.[26]. The elevatedplus-maze consisted of two opposed open arms (0.5 m× 0.1 m)and two opposed closed arms (0.5 m× 0.1 m), and an open square(0.1 m× 0.1 m) in the centre (Fig. 1). The enclosed arms were sur-rounded by 0.4 m high walls. The maze was elevated 0.5 m abovethe floor. The light intensity in the open arms was 200–350 lx. In theclosed arms it was less than 1 lx. Immediately before being testedin the plus-maze, each rat was placed in the open-field for 5 min(procedure described above). This procedure allows to increase theoverall activity of rats in the plus-maze and to increase the likelihoodthat they would explore open arms[26]. The animals were placedon the central square of the plus-maze, facing one of the open arms.The following variables were analyzed: the number of entries intothe open and the closed arms, and the time spent in the open armsand on the central square. An entry into any of the compartmentswas defined as putting all four paws on its floor. Each rat was testedduring three successive days (5 min each).
2noise
wp aldae con-s wereh xper-i roundn Thed entalc wass in.
2
werei ture
and Activity Transmitter Without Battery, Mini Mitter, Co, Inc.,Bend, OR, USA). Once implanted, the transponders allow collectingdata for the lifetime of the rat. The implantation procedure has beendescribed in our previous paper[30].
The telemetric recordings were performed when the rats were4 months old. They stayed in their home-cages and were protectedfrom any stress. Both temperature and activity data were recordedat 30 s intervals for 2–4 days. The body temperature and activitysignals were transmitted via the receiver board (ER-4000 EnergizerReceiver, MiniMitter, USA) placed under the home-cage of eachrat and analyzed by the data acquisition system (VitalView Series4000, software 3.1, Mini Mitter, USA).
2.5. Data analysis
Values are reported as the means± S.E.M. Moreover, motoractivity index was calculated for each rat exposed to the open-field,to the sudden silence, and to the plus-maze according to the formuladescribed in our previous paper[30]. To calculate indexes of the fourcomponents of motor activity (i.e., total duration of motor activity,distance, linear and angular velocities), the control rats maintainedat physiological neonatal body temperature of 33◦C were given val-ues 1.0, and individual data in each group were indexed accordingly,and then they were averaged for each rat.
To calculate indexes of the open arms penetration in the plus-maze test we used two variables: frequency of entries and timespent in the open arms. All results were tested for normalityw ouri witht ra-t nda tem-p dm lts ofr DFi testsw lus-m e lastf s ind wedb test.T s.D tSoft,C
3
3
oxiac en-fi
d (two-w rrp eent ith
.3.3. Response to a sudden silenceThe behavioural responses to a sudden drop in background
ere recorded in a square wooden box (0.85 m× 0.60 m× 0.60 m),ainted black. The procedure was previously described by Buwt al.[2]. Each test consisted of two trials performed during twoecutive days. On the first day of the experiment the animalsabituated to the test conditions and placed for 5 min in the e
mental box. A sound generator produced a constant backgoise (65 dB, 2–8 kHz) during the whole time of the exposure.ay after the habituation rats were exposed again to the experimonditions for 5 min. After the first 2 min the background noisewitched off, leaving the animal in total silence for the final 3 m
.4. Telemetric recordings
At the age of 31 days five rats from each experimental groupmplanted intraperitoneally with transmitters (E-Mitter Tempera
ith Kolmogorov–Smirnov test. All variables of rat’s behavin the open-field and sudden silence tests were analyzedwo-way ANOVA in a 3× 2 factorial design, where tempeures (33, 37, and 39◦C) and experimental conditions (control anoxia) were the factors. A three-way analysis of variance (erature× experimental conditions× trial, 3× 2× 3) with repeateeasures on the third factor was applied to analyze the resu
at’s behaviour in plus-maze test. The effect of treatment withn hyperthermic groups in the open-field and sudden silenceas analyzed by one-way ANOVA and the effect of DF in the paze test – by two-way ANOVA with repeated measures on th
actor. Two-way ANOVA was also applied to analyze changeiurnal activity and body temperature. Each analysis was folloy multiple comparison using Tukey–Kramer HSD post hoche significance level was set atp< 0.05 for all statistical testata analyses were performed with the use of Statistica (Staracow, Poland).
. Results
.1. Open-field
Both neonatal body temperature and neonatal anlearly influenced motor activity of the adult rats in the opeld test (Fig. 2).
Percentage of the active time during the test (Fig. 2A)ecreased with increasing neonatal body temperatureay ANOVA, F2,32= 16.69; p< 0.001) and was furthe
educed in rats exposed to neonatal anoxia (F1,32= 4.81;< 0.05). There was no significant interaction betw
hose factors (F2,32= 1.62; n.s.). Further analysis w
M. Caputa et al. / Behavioural Brain Research 163 (2005) 246–256 249
Fig. 2. Effects of neonatal anoxia, neonatal body temperature and chelation of iron with deferoxamine (DF) upon total duration of motor activity (A),totaldistance travelled (B), mean linear velocity of movements (C) and mean angular velocity of movements (D) in adult (4 months old) rats exposed for 10 mintoopen-field stress. Values are presented as mean + S.E.M.;*p< 0.05,** p< 0.01,*** p< 0.001 for two-way ANOVA or#p< 0.05,##p< 0.01 for one-way ANOVA.Numbers inside columns indicate numbers of animals in individual groups.
Tukey–Kramer HSD test revealed that animals exposed toneonatal anoxia at 39◦C spent less time on motor activitythan rats keptpostpartumat body temperature of 33◦C bothin control and anoxic groups (p< 0.001). One-way ANOVA,performed to test the effect of DF at neonatal body tem-perature of 39◦C (Fig. 2A, right side), depicted significantdifferences between the three hyperthermic groups of rats(F2,14= 6.75; p< 0.01). The hypoactivity in rats exposedto anoxia at 39◦C was prevented by DF administration(p< 0.01).
Two-way ANOVA applied to analyze changes in dis-tance travelled during the test (Fig. 2B) showed a significanteffect of neonatal body temperature (F2,36= 10.44;p< 0.001)and neonatal anoxia (F1,36= 22.70;p< 0.001), and revealeda significant interaction between the factors (F2,36= 4.17;p< 0.05). Rats exposed to neonatal anoxia at the highest bodytemperatures of 37 and 39◦C travelled about six times shorterdistances than control rats allowed to maintain neonatal bodytemperature at 33◦C (p< 0.001). There was a significantdifference between hyperthermic (39◦C; right side of thepanel) groups of rats exposed to anoxia (one-way ANOVA;F2,16= 5.54; p< 0.05). DF-treated anoxic rats were moreactive than their untreated counterparts at the same neonatal
body temperature (p< 0.05; Tukey–Kramer HSD test) and thedistance travelled by animals from this group was the sameas that in the control animals allowed to maintain neonatalbody temperature at 33◦C. Two-way ANOVA yielded theonly significant effect of neonatal anoxia on linear velocityof movements (Fig. 2C; F1,35= 12.89;p< 0.001). However,there were no significant effects of neonatal body temperature(F2,35= 1.65; n.s.), and no significant interactions betweenboth factors (F2,35= 1.37; n.s.). Post hoc analysis revealedthat rats exposed to neonatal anoxia at body temperature of39◦C moved slower than control rats maintained at physi-ological body temperature of 33◦C (p< 0.05). There was asignificant effect of experimental conditions on linear veloc-ity (one-way ANOVA;F2,14= 5.75;p< 0.05) (Fig. 2C, rightside) in three groups of rats forced to maintain neonatal bodytemperature of 39◦C. The Tukey–Kramer test demonstratedthat the administration of DF prevented the decrease in linearvelocity in rats exposed to neonatal anoxia at the same bodytemperature (p< 0.05).
Neither neonatal body temperature nor neonatal anoxiaexerted significant influences on mean angular velocity ofmovements (Fig. 2D) in rats exposed to the open-field test.One-way ANOVA performed to test the effect of DF in rats
250 M. Caputa et al. / Behavioural Brain Research 163 (2005) 246–256
Fig. 3. Effects of neonatal anoxia, neonatal body temperature and chelationof iron with deferoxamine (DF) on motor activity index of rats exposed toopen-field stress at the age of 4 months. To prepare the indexes we usedthe data analyzed inFig. 2 (A–D). Values are presented as mean + S.E.M.;** p< 0.01 for two-way ANOVA or##p< 0.01 for one-way ANOVA. Thehorizontal dotted line shows the reference value 1.0. Numbers inside columnsindicate numbers of animals in individual groups.
exposed to neonatal anoxia at body temperature of 39◦Con mean angular velocity depicted no significant differencesbetween the three hyperthermic groups (F2,17= 0.08; n.s.)(Fig. 2D, right side).
Fig. 3 shows index of the stress-induced motor activ-ity, computed from the data presented inFig. 2A–D. Two-way ANOVA proves that stress-induced motor activity ofrats decreased with increasing neonatal body temperature(F2,16= 5.36;p< 0.01) and was reduced in animals exposed toneonatal anoxia (F1,16= 5.31;p< 0.05). However, there wasno significant interaction between both factors (F2,16= 0.43;n.s.). The activity of rats exposed to neonatal anoxia at bodytemperature of 39◦C was reduced (about two-fold) comparedto that of control rats having physiological neonatal body tem-perature of 33◦C (p< 0.01). Moreover, the rats exposed toneonatal anoxia at 37 and 39◦C tended to be less active thananoxic rats allowed to maintain neonatal body temperatureat 33◦C, although the differences were not significant. Theone-way ANOVA revealed a significant effect of experimen-tal conditions on motor activity index in three groups of ratsexposed to neonatal body temperature of 39◦C (F2,70= 6.10;p< 0.01). The chelation of iron with DF significantly pre-vented the anoxia-induced hypoactivity (p< 0.01). It mustbe stressed that the motor activity index of DF-treated ratsremained unchanged at a control level of 1.0.
3
en-t ses:n eda quareo
As far as the number of entries into open arms isconcerned (Fig. 4A) the analysis shows a significantmain effect of all three experimental factors: temperature(F2,49= 11.10; p> 0.001), anoxia (F1,49= 11.88; p> 0.01)and trial (F2,98= 7.28;p< 0.01). Interaction of temperatureand anoxia was also significant (F2,49= 3.60;p> 0.05). Fur-ther analysis with Tukey–Kramer HSD test revealed thatanimals exposed to neonatal anoxia at 37C and 39◦C enteredthe open arms more frequently than rats keptpostpartumatbody temperature of 33◦C, both in control and in anoxicgroups. The differences were highly significant. The ratsentered the open arms more frequently during the first sessionthan during the subsequent exposures. During the first ses-sion anoxic animals at neonatal body temperature elevatedto 37 and to 39◦C entered the open arms with the highestfrequency.
Two-way ANOVA (experimental conditions× trial, 2× 3,repeated measures on the last factor) was performed to test theeffect of DF in rats exposed to neonatal anoxia at body tem-perature of 39◦C. There were significant differences betweenthe three hyperthermic groups (F2,25= 6.82;p< 0.01).Posthocanalysis demonstrated that the high frequency of enteringthe open arms in rats exposed to anoxia at 39◦C was pre-vented by administration of DF (p< 0.01). The main effect oftrial (F2,50= 2.09; n.s.) and the interaction term (F4,50= 1.35;n.s.) were not significant.
rmsc rms(3 weda ture( ofa l( ena weren withT dt 9s keptpp d noe Two-w ntalc nb
thep rial( wo-w exot VA;F ty( erew for-mt oth
.2. Plus-maze test
The following variables of the behaviour in all experimal groups were subjected to three-way ANOVA analyumber of entries (±S.E.M.) into the open arms and closrms, time spent in the open arms and on the central sf the plus-maze.
Time the animals of all groups spent in the open ahanged parallel with the number of entries into the aFig. 4B). Three-way ANOVA (temperature× anoxia× trial,× 2× 3, with repeated measures on the last factor) shosignificant main effect of neonatal body tempera
F2,57= 5.94; p< 0.01) and a significant main effectnoxia (F1,57= 5.66; p< 0.05). A significant effect of triaF2,11= 5.53; p< 0.01) indicates that time spent in oprms decreased during subsequent sessions. Thereo interactions between the factors. Further analysisukey–Kramer HSD test (temperature× anoxia) reveale
hat animals exposed to neonatal anoxia at 37 and 3◦Cpent more time in the open arms than control ratsostpartumat body temperature of 33◦C (p< 0.05). Theostanoxic administration of DF to hyperthermic rats haffect on the percentage of time spent in the open arms.ay ANOVA shows that the main effects of experimeonditions (F2,26= 2.64), trial (F2,52= 1.01), and interactioetween the factors (F4,52= 0.78) were not significant.
The index of entries into the open arms oflus-maze, pooled from the first through the third tFig. 5A) increased with neonatal body temperature (tay ANOVA; F2,348= 13.37;p< 0.001) and so did the indf motor activity (F2,326= 6.87; p< 0.01) (Fig. 5B). Both
he index of entries into the open arms (two-way ANO1,348= 10.90; p< 0.01) and the index of motor activiF1,326= 7.07;p< 0.01) were influenced by anoxia, but there no interactions between the factors both in theer (F2,348= 1.81) and the latter of these (F2,326= 0.99). The
herapeutic effect of DF was significant in terms of b
M. Caputa et al. / Behavioural Brain Research 163 (2005) 246–256 251
Fig. 4. Effects of neonatal anoxia, neonatal body temperature and chelation of iron with deferoxamine (DF) upon number of entries (left panel) and time spentin the open arms (right panel) in adult rats subjected to the elevated plus-maze test in three successive sessions. Values are presented as mean + S.E.M. Numbersinside upper columns indicate numbers of animals in individual groups.
reduced penetration of the open arms (one-way ANOVA;F2,168= 6.62;p< 0.01) and reduced motor activity (one-wayANOVA; F2,164= 5.06;p< 0.01).
Interestingly, percentage of time spent on the centralsquare of the plus-maze was not significantly differentwhen analyzed for temperature (F2,56= 1.86; n.s.) and trial(F2,11= 0.98; n.s.). However, anoxia had a significant effecton time spent on the central square (F1,56= 4.34;p< 0.05)and the interaction between temperature and experimentalconditions (F2,56= 3.34;p< 0.05) was also significant. Theeffect of DF on time spent on the central square was sig-nificant (two-way ANOVA,F2,27= 5.52;p< 0.01). The maineffect of trial (F2,54= 2.96; n.s.) and interaction between thefactors (F4,54= 1.12; n.s.) were not significant. DF-treatedrats spent less time on the central square than their untreatedanoxic counterparts at the same neonatal body temperature(p< 0.05). Moreover, the percentage of time spent by DF-treated rats on the central square was the same as that in thecontrol animals allowed to maintain neonatal body tempera-ture at 33◦C.
The number of entries into the closed arms wasan additional behavioural variable evidencing the distur-bances in motor activity. It was significantly influenced byneonatal temperature (F2,50= 11.13;p< 0.001) and anoxia(F1,50= 17.26;p< 0.001) with significant interaction betweenthose factors (F2,50= 6.26;p< 0.01). Further analysis withTukey–Kramer HSD test revealed that animals exposed toneonatal anoxia at 39◦C entered the closed arms morefrequently than rats allowed to maintain neonatal bodytemperature at 33◦C both in control and anoxic groups(p< 0.001). The number of entries into the closed armswas also significantly greater in rats exposed to neonatalanoxia at body temperature of 39◦C comparing to anoxicanimals maintained at body temperature of 37◦C (p< 0.01).The frequency of the closed arms penetration by animalsdecreased during subsequent trials (F2,10= 4.19; p< 0.05).Two-way ANOVA performed to test the effect of DF atneonatal body temperature of 39◦C showed a significantmain effect of anoxia (F2,25= 20.84; p< 0.001) and trial(F2,50= 3.50;p< 0.5) on frequency of entries into the closed
252 M. Caputa et al. / Behavioural Brain Research 163 (2005) 246–256
Fig. 5. Effects of neonatal anoxia, neonatal body temperature and chelation of iron with deferoxamine (DF) on index of open arms penetration (A) (to preparethe indexes we used the data analyzed inFig. 4, which were pooled from trials 1 to trial 3) and on the index of motor activity (B) in adult rats subjected to theelevated plus-maze test. Values are presented as mean + S.E.M;*p< 0.05,** p< 0.01,*** p< 0.001 for two-way ANOVA or#p< 0.05,##p< 0.01 for one-wayANOVA. The horizontal dotted lines show the reference values 1.0. Numbers inside columns indicate numbers of animals in individual groups.
arms. However, the interaction of factors was not significant(F4,50= 0.91; n.s.).Post hocTukey–Kramer HSD test demon-strated that anoxia in hyperthermic conditions causes a signif-icant hyperactivity in rats exposed to plus-maze (p< 0.001)(Fig. 5B), which was prevented by administration of DF(p< 0.001).
3.3. Sudden silence test
The results of the test are shown inFig. 6. There weresignificant effects of neonatal body temperature on both thelatency of start of movement (Fig. 6A) (two-way ANOVA;F2,36= 19.58;p< 0.001) and general motor activity after ces-
sation of the background noise (Fig. 6B) (F2,179= 16.88;p< 0.001). However, neonatal anoxia had no significant effecton the latency (F1,36= 1.26; n.s.) and on the motor activity(F1,79= 1.25; n.s.). As far as the motor activity is concerned,there were significant interactions between the both neonatalfactors (F2,179= 4.96;p< 0.01).
Immediately after cessation of background noise rats keptpostpartumat body temperature of 33◦C both in control andanoxic conditions responded with immobility to the suddenchange in their environment (Fig. 6A). Further analysis withTukey–Kramer HSD test revealed that in animals exposedto neonatal anoxia at 37 and 39◦C the response was lesspronounced than in rats keptpostpartumat body temperature
F helatio ent (A) ant ction in .;* animal
ig. 6. Effects of neonatal anoxia, neonatal body temperature and che index of motor activity (B) in adult rats exposed to a sudden redu* p< 0.01,*** p< 0.001. Numbers inside columns indicate numbers of
n of iron with deferoxamine (DF) upon latency of the start of movemdbackground noise for 3 min. Values are presented as mean + S.E.M*p< 0.05,
s in individual groups.
M. Caputa et al. / Behavioural Brain Research 163 (2005) 246–256 253
Fig. 7. Effects of neonatal anoxia, neonatal body temperature, and deferox-amine (DF) on locomotor activity (in counts per 30 s) in adult rats, recordedin their home-cages prior to and following the elevated plus-maze test. Num-bers inside columns indicate numbers of animals in individual groups. Valuesare means + S.E.M.
of 33◦C both in control (p< 0.01) and anoxic (p< 0.05 andp< 0.01) groups.
The animals exposed to neonatal anoxia at 37 and 39◦Ctended to be less active after the cessation of backgroundnoise but the differences did not reach the threshold ofstatistical significance (Fig. 6B). On the other hand, sud-den silence markedly reduced motor activity of rats allowedto maintain their neonatal body temperature at physiolog-ical level of 33◦C, under both normoxic (p< 0.001) andanoxic (p< 0.001) conditions. Control hyperthermic groupsalso showed reduced motor activity (p< 0.001 at neonatalbody temperature of 37◦C andp< 0.01 at 39◦C). One-wayANOVA showed no effect of administration of DF on thelatency of start of movement (Fig. 6A) (F2,17= 1.21; n.s.) butgeneral motor activity of DF-treated rats was significantlyreduced (Fig. 6B) (p< 0.05) after cessation of backgroundnoise.
3.4. Telemetric data
Spontaneous motor activity of rats in their home-cages,recorded radiotelemetrically prior to the behavioural tests,was influenced by neonatal body temperature (F2,26= 17.72;p< 0.001) and by neonatal anoxia (F1,26= 5.14;p< 0.05) andthere was a significant interaction between the both factors( int la otherg uents ss ofp dings oxicg tureo h a
normal body temperature also tended to reduce the activityafter the test, and the difference was close to the threshold ofstatistical significance (p< 0.097).
Spontaneous changes in body temperature were notaffected by the neonatal experimental conditions.
4. Discussion
Neonatal asphyxia in mammals causes delayedbehavioural disturbances such as abnormal responsesto stress and impaired learning[2,11,22,35]. The distur-bances persist over the entire life span[22]. The results of thepresent investigation confirm the existence of the abnormalresponses to stress in adult rats previously subjected toneonatal asphyxia. However, this investigation shows thatthe disturbances are more complex and variable than it hasbeen reported previously. The abnormal stress response inrodents exposed to neonatal anoxia has been regarded as abiphasic process: a transient hyperactivity in the infantileage[11,22,35]followed by a hypoactivity in the adulthood[21,22,42]. At first glance our previous[30] as well asthe present experiments confirm the existence of such apattern. We have been able to show that the open-fieldstress-induced hyperactivity culminates from 5th through30th day of the rat’s age, then it vanishes at the age of 45d cea ths.T tures tr s arei trastt oura rmia,w (seeFO rdedr hichw orns,a otorr andh ratsa ctive.S iousn
F1,26= 5.26; p< 0.05) (Fig. 7). Rats allowed to maintaheir neonatal body temperature at 33◦C, both in contrond anoxic group, were more active than those in anyroup. Open-field stress did not influence the subseqpontaneous activity (not shown in the figure for clearneresentation), while the activity was reduced in the recoression following the elevated plus-maze test, in the anroup of rats having physiological neonatal body temperaf 33◦C (p< 0.001). Control rats allowed to maintain suc
ays[30], andFigs. 2 and 3of the present paper evidenclear-cut hypoactivity in rats at the age of 4 mon
he hypoactivity seems to be a manifestation of premaenescence[22]. Both the previous[30] and the presenesults show that the postanoxic behavioural disturbancenduced and amplified by neonatal hyperthermia. In cono their hypoactivity during the open-field stress testdult rats, neonatally exposed to anoxia and hypertheere hyperactive during the elevated plus-maze testig. 5B) and during the sudden silence test (Fig. 6B).n the other hand, spontaneous motor activity, reco
adiotelemetrically in home-cages of the adult rats, were exposed to anoxia and hyperthermia as newbgain was significantly reduced. Altogether, locomesponses of adult rats neonatally exposed to anoxiayperthermia are not fully consistent: in general there hypoactive but some stressors make them hyperauch disturbances might reflect abnormalities in vareurotransmitter systems of the brain[22].
The present investigation shows that adult rats eaxposed to neonatal anoxia at body temperature of 3◦Cre protected against the disturbances in the behavesponses to stress. The controversies emerging from trature dealing with the long-term neuroprotective effeceduced body temperature might result from differental models and different procedures used by experimeome doubts as to permanent neuroprotection concern
erm consequences of brief post-hypoxic-ischemic reduf body temperature, which delays but does not preventral injury in rats subjected to the insult in their infan
254 M. Caputa et al. / Behavioural Brain Research 163 (2005) 246–256
[40] or as adults[12]. In contrast, prolonged postischemicbody cooling in adult rats[8] as well as adult[4] and old[7] gerbils provides persistent neuroprotection. Accordingto De Bow and Colbourne[10] some hippocampal neuronsof the gerbil, previously salvaged with body cooling, arevulnerable to secondary, transient ischemic attacks. Such sec-ondary insults are likely to appear spontaneously during thelife of animals previously subjected to experimental hypoxia-ischemia.
In rats used in the present investigation a critical asphyxiadeveloped after 14.95± 0.40 min at body temperature of37◦C and after 9.25± 0.30 at body temperature of 39◦C,so that the exposure to anoxia had to be interrupted in eachrat, while all newborns at body temperature of 33◦C toler-ated well the exposure to anoxia scheduled for 25 min[3].However, the present investigation proves that such a short-ened exposure to anoxia under hyperthermic conditions didnot prevent the behavioural disturbances in adult rats.
We manipulated the body temperature of newborn ratsnot only during the exposure to critical anoxia but also dur-ing 2 h period of reoxygenation. It must be stressed, how-ever, that the lowest body temperature of 33◦C, chosen inthe present investigation, is actually normal body temper-ature of newborn rats[31]. Accordingly, the experimentalanimals must have continued to maintain their body tem-perature at this particular level for many days after thea ouslyr anentp long-t weree , andt fer-o thedT iousfi ionsi sat n oura rther-m rox-a undi tingh n inn tos lyp etal[ rain[
pre-v aa tion[ dy( de oft ease[ r-
ization of cerebral neurons[15,23], massive calcium entryinto the neurons[5,16,22], cerebrocortical release of NO[39], and brain oedema[27,33,38]. The reduced neona-tal body temperature provides an excellent neuroprotectionnot only in rodents but also in babies suffering from birthasphyxia, when applied 4–76 h postpartum[6]. Interestingly,neuroprotective effect of the spontaneously reduced bodytemperature in neonatal rats is combined with their uniquegasping ability, making them extremely tolerant to asphyxia[31].
The present investigation provides an evidence that con-trol exposures of neonatal rats to hyperthermia affects theirbehaviour in adulthood. This has also been shown in juve-nile rats in our previous study[31]. A possible explanationis that neonatal hyperthermia triggers a damage to the brainnot only after the experimental anoxia but also following thepreceding natural parturitional asphyxia.
Some of the behavioural disturbances, recorded in thispaper in adult rats after neonatal exposure to anoxia, werequite surprising. The open arms penetration in the plus-maze test was significantly enhanced in adult rats neonatallyexposed to anoxia and hyperthermia (Fig. 5A). In experi-ments performed by Buwalda et al.[2], although the adultrats subjected to neonatal anoxia tended to spend more timeon the open arms of the maze, the difference did not reach thethreshold of statistical significance. Because the inventors oft danceo -t ucesa e oft roupo hant rsiont rol orai nces . Thisc lencet b-j ia thel lencew vinguT ratsr f thelc hyxiaa imuli.P stur-b ns fer-o t.
ourala ievedb c DFa
sphyxia. The present paper shows that the spontaneeduced neonatal body temperature provides a permrotection against behavioural disturbances. Such a
erm protection was also the case in our rats, whichxposed to combined neonatal anoxia and hyperthermiahen injected with deferoxamine, a chelator of iron. Dexamine is not only effective in chelation therapy, butrug has also been proven to be therapeutically safe[19].he therapeutic effect of DF is consistent with our prevnding that neonatal anoxia under hyperthermic condit
s followed by a bout of hyperferremia[3]. Because iron icatalyst for oxidative stress[9,18,24,37,43]we conclude
hat the delayed behavioural disturbances, recorded idult rats after neonatal exposure to anoxia and hypeia, are due to free radical damage to the brain. Defemine is extremely effective in reducing non-protein-bo
ron in plasma as well as in the brain, and in prevenypoxia-ischemia-induced reperfusion injury of the braiewborn lambs[34]. Newborns have an impaired abilityequester iron[24,32,37,41]. Therefore, they are extremerone to both excessive intraneuronal storage of the m
17,24,25]and to the delayed free radical damage to the b9,18,24].
Reduced neonatal physiological body temperatureents postanoxic metabolic acidosis and hyperferremi[3]nd thus might prevent secondary free radical forma
5,13,14,20,22,44]. Moreover, similar decrease of bobrain) temperature in adult mammals interrupts a cascahe following neurotoxic events: cerebral glutamate rel5,13,14,22]and its neurotoxicity[36], a general depola
he elevated plus-maze test have considered that avoif the open arm penetration reflects anxiety[26], the neona
al exposure to anoxia and hyperthermia seemingly rednxiety in adulthood. However, an increased avoidanc
he open arms after 3 days of repeated testing in this gf rats (Fig. 4) suggests that they did not perform better t
he control animals. On the contrary, they acquired aveo heights and open spaces much more slowly than contnoxic rats at neonatal body temperature of 33◦C. Accord-
ngly, the nature of this particular behavioural disturbaeems to be a reduced attention paid to potential dangersonclusion is supported by the results of the sudden siest (Fig. 6). In both groups of adult rats previously suected to combined neonatal asphyxia and hyperthermatency to start locomotion in response to the sudden sias markedly shorter than that in both groups of rats handisturbed neonatal body temperature of 33◦C (Fig. 6A).he subsequent motor activity of the former groups ofemained undisturbed by the sudden silence while that oatter groups was markedly reduced (Fig. 6B). Therefore, weonclude that adult rats being neonatally exposed to aspnd hyperthermia show reduced alertness to external stostanoxic DF injection prevented the behavioural diances in both the plus-maze (Fig. 5A and B) and the suddeilence (Fig. 6B) tests. These results prove again that dexamine also exerts a permanent neuroprotective effec
In conclusion, a permanent protection against behavibnormalities induced by neonatal asphyxia can be achy both reduced body temperature and by postasphyxidministration.
M. Caputa et al. / Behavioural Brain Research 163 (2005) 246–256 255
Acknowledgement
This work was supported by grant of 4.P05A.059.16 fromthe Polish State Committee for Scientific Research.
References
[1] Barth A, Bauer R, Gedrange T, Walter B, Klinger W, Zwiener U.Influence of hypoxia and hypoxia/hypercapnia upon brain and bloodperoxidative and glutathione status in normal weight and growth-restricted newborn piglets. Exp Toxic Pathol 1998;50:402–10.
[2] Buwalda B, Nyakas C, Vosselman HJ, Luiten PGM. Effects of earlypostnatal anoxia on adult learning and emotion in rats. Behav BrainRes 1995;67:85–90.
[3] Caputa M, Rogalska J, Nowakowska A. Effect of temperature onpostanoxic, potentially neurotoxic changes of plasma pH and freeiron level in newborn rats. Brain Res Bull 2001;55:281–6.
[4] Colbourne F, Corbett D. Delayed postischemic hypothermia: a sixmonth survival study using behavioral and histological assessmentsof neuroprotection. J Neurosci 1995;15:7250–60.
[5] Colbourne F, Sutherland G, Corbett D. Postischemic hypothermia.A critical appraisal with implications for clinical treatment. MolNeurobiol 1997;14:171–201.
[7] Corbett D, Nurse S, Colbourne F. Hypothermic neuroprotection. Aglobal ischemia study using 18- to 20-month-old gerbils. Stroke
withtran-0–6.FF,istar
[ killther-
[ .acial
rats.
[ D.hron-od
[ . In:newger;
[ rging
hem
[ xictabol
[ atarmia
[ er-pro-mia
[18] Kondo Y, Asanuma M, Nishibayashi S, Iwata E, Ogawa N. Late-onset lipid peroxidation and neuronal cell death following transientforebrain ischemia in rat brain. Brain Res 1997;772:37–44.
[19] Kontoghiorghes GJ, Pattichi K, Hadjigavriel M, Kolnagou A. Trans-fusional iron overload and chelation therapy with deferoxamine anddeferiprone (L1). Transfus Sci 2000;23:211–23.
[20] Lei B, Adachi N, Arai T. The effect of hypothermia on H2O2 pro-duction during ischemia and reperfusion: a microdialysis study inthe gerbil hippocampus. Neurosci Lett 1997;222:91–4.
[21] Loidl CF, Gavilanes AWD, Van Dijk EHJ, Vreuls W, BloklandA, Vles JSH, et al. Effects of hypothermia and gender on sur-vival and behavior after perinatal asphyxia in rats. Physiol Behav2000;68:263–9.
[23] Onitsuka M, Mihara S, Inokuchi H, Shigemori M, Higashi H. Mildhypothermia protects rat hippocampal CA1 neurons from irreversiblemembrane dysfunction induced by experimental ischemia. NeurosciRes 1998;30:1–6.
[24] Palmer Ch. Iron and oxidative stress in neonatal hypoxic-ischemicbrain injury. Directions for therapeutic intervention. In: Connor JR,editor. Metals and oxidative damage in neurological disorders. NewYork: Plenum Press; 1997. p. 205–36.
[25] Palmer Ch, Menzies SL, Roberts RL, Pavlick G, Connor JR. Changesin iron histochemistry after hypoxic-ischemic brain injury in theneonatal rat. J Neurosci Res 1999;56:60–71.
[26] Pellow S, Chopin P, File SE, Briley M. Validation of open: closedarm entries in an elevated plus-maze as a measure of anxiety in therat. J Neurosci Meth 1985;14:149–67.
[27] Plateel M, Dehouck MP, Torpier G, Cecchelli R, Teissier E. Hypoxiaincreases the susceptibility to oxidant stress and the permeability
hem
[ ionRes
[ ramid-
399–
[ ess-body21–9.
[ re andxia in
[ diatr
[ theExp
[ i M,on-tis-
Lett
[ ats
[ em-uma
[ ease
[ DA.tionabbit
1997;28:2238–43.[8] Corbett D, Hamilton M, Colbourne F. Persistent neuroprotection
[9] Dal-Pizzol F, Klamt F, Frota Jr, MLC, Andrades ME, Caregnatoet al. Neonatal iron exposure induces oxidative stress in adult Wrat. Dev Brain Res 2001;130:109–14.
10] De Bow SB, Colbourne F. Delayed transient ischemic attackssome CA1 neurons previously salvaged with postischemic hypomia: neuroprotection undone. Brain Res 2003;959:50–7.
11] Dell’Anna ME, Calzolari S, Molinari M, Iuvome L, Calimici RNeonatal anoxia induces transitory hyperactivity, permanent spmemory deficits and CA1 cell density reduction in developingBehav Brain Res 1991;45:125–34.
12] Dietrich WD, Busto R, Alonso O, Globus MY-T, Ginsberg MIntraischemic but not postischemic brain hypothermia protects cically following global forebrain ischemia in rats. J Cereb BloFlow Metab 1993;13:541–9.
13] Ginsberg MD. Temperature influences on ischemic brain injuryHsu CY, editor. Ischemic Stroke: from basic mechanisms todrug development. Monogr Clin Neurosci, vol. 16. Basel: Kar1998. p. 65–88.
14] Globus MY-T, Alonso O, Dietrich WD, Busto R, GinsbeMD. Glutamate release and free radical production followbrain injury: effects of posttraumatic hypothermia. J Neuroc1995;65:1704–11.
15] Hiramatsu K, Kassell NF, Lee KS. Thermal sensitivity of hyporesponses in neocortical brain slices. J Cereb Blood Flow Me1993;13:395–401.
16] Kataoka K, Mitani A, Yanase H, Zhang L, Higashihara M, OgT, et al. Ischemic neuronal damage: how does mild hypothemodulate it? Mol Chem Neuropathol 1996;28:191–5.
17] Kondo Y, Ogawa N, Asanuma M, Ota Z, Mori A. Regional diffences in late-onset iron deposition, ferritin, transferrin, astrocyteliferation, and microglial activation after transient forebrain ischein rat brain. J Cereb Blood Flow Metabol 1995;15:216–26.
of the blood-brain barrier endothelial cell monolayer. J Neuroc1995;65:2138–45.
28] Ray G, Husain SA, Dutta AK, Batra S. Lipid peroxidatand antioxidant status in asphyxiated neonates. Med Sci1998;26:201–2.
29] Reglodi D, Somogyvari-Vigh A, Maderdrut JL, Vigh S, ArimuA. Postischemic spontaneous hyperthermia and its effects indle cerebral artery occlusion in the rat. Exp Neurol 2000;163:407.
30] Rogalska J, Caputa M, Wentowska K, Nowakowska A. Strinduced behaviour in juvenile rats: effects of neonatal asphyxia,temperature and chelation of iron. Behav Brain Res 2004;154:3
31] Rogalska J, Caputa M. Spontaneously reduced body temperatugasping ability as a mechanism of extreme tolerance to asphyneonatal rats. J Therm Biol; 2005, in press.
32] Saugstad OD. Oxygen toxicity in the neonatal period. Acta PeaScand 1990;79:881–92.
33] Schwab M, Bauer R, Zwiener U. Mild hypothermia preventsoccurrence of cytotoxic brain edema in rats. Acta Neurobiol1998;58:29–35.
34] Shadid M, Buonocore G, Groenendaal F, Moison R, FerralBerger HM, et al. Effect of deferoxamine and allopurinol on nprotein-bound iron concentrations in plasma and cortical brainsue of newborn lambs following hypoxia-ischemia. Neurosci1998;248:5–8.
35] Speiser Z, Korczyn AD, Teplitzky I, Gitter S. Hyperactivity in rfollowing postnatal anoxia. Behav Brain Res 1983;7:379–82.
36] Suehiro E, Fujisawa H, Ito H, Ishikawa T, Maekawa T. Brain tperature modifies glutamate neurotoxicity in vivo. J Neurotra1999;16:285–97.
37] Sullivan JL. Iron, plasma antioxidants, and the oxygen radical disof prematurity. AJDC 1988;142:1341–4.
38] Tan S, Zhou F, Nielsen VG, Wang Z, Gladson CL, ParksIncreased injury following intermittent fetal hypoxia-reoxygenais associated with increased free radical production in fetal rbrain. J Neuropathol Exp Neurol 1999;58:972–81.
256 M. Caputa et al. / Behavioural Brain Research 163 (2005) 246–256
[39] Thoresen M, Satas S, Puka-Sundvall M, Whitelaw A, Hallstrom A,Løberg E-M, et al. Post-hypoxic hypothermia reduces cerebrocorticalrelease of NO and excitotoxins. Neuroreport 1997;8:3359–62.
[41] Vannucci RC, Perlman JM. Interventions for perinatal hypoxic-ischemic encephalopathy. Pediatrics 1997;100:1004–14.
[42] Venerosi A, Valanzano A, Cirulli F, Alleva E, Calamandrei G.Acute global anoxia during C-section birth affects dopamine medi-
ated behavioural responses and reactivity to stress. Behav Brain Res2004;154:155–64.
[43] Yu T, Kui LQ, Ming QZ. Effect of asphyxia on non-protein-boundiron and lipid peroxidation in newborn infants. Dev Med Child Neu-rol 2003;45:24–7.
[44] Zhao W, Richardson JS, Mombourquette MJ, Weil JA, Ijaz S, ShuaibA. Neuroprotective effects of hypothermia and U-78517F in cere-bral ischemia are due to reducing oxygen-based free radicals: anelectron paramagnetic resonance study with gerbils. J Neurosci Res1996;45:282–8.
