Characterization of postischemic behavioral deficits in gerbils with and without hypothermic neuroprotection

Transcription

1 r98r$19.00 q 1998 Elsevier Science B.V. All rights reserved. Ž. Brain Research Research report Characterization of postischemic behavioral deficits in gerbils with and without hypothermic neuroprotection Frederick Colbourne a,), Roland N. Auer a, Garnette R. Sutherland b a Department of Pathology, Faculty of Medicine, UniÕersity of Calgary, 3330 Hospital DriÕe N.W., Calgary, Alberta, Canada, T2N 4N1 b Department of Clinical Neurosciences DiÕision of Neurosurgery, Faculty of Medicine, UniÕersity of Calgary, 3330 Hospital DriÕe N.W., Calgary, Alberta, Canada, T2N 4N1 Accepted 2 June 1998 Abstract Five minutes of global ischemia in gerbil results in delayed hippocampal CA1 neuronal degeneration, which is accompanied by working memory impairments and hyperactivity in novel environments. In this study, postischemic activity was characterized in familiar and in novel environments to determine whether hyperactivity was due to impaired spatial habituation or another form of motor hyperactivity. This study also determined whether 6-h delayed hypothermia, which reduces CA1 neuronal injury, would attenuate functional impairments. Gerbils were subjected to 5 min of normothermic ischemia or sham operation 2 days following implantation of brain temperature probes. One of two ischemic groups was cooled Ž ) 48 h. starting at 6-h postischemia. Locomotor activity in a familiar cage was measured for 6 days while activity in three novel environments was intermittently measured on days 4, 5 and 6. Open field behavior and working memory in a T-maze were also assessed. Untreated ischemia caused marked hyperactivity in the familiar cage on day 1, which reverted to near-normal by day 2. Nonetheless, these gerbils showed hyperactivity during novel environment sessions on days 4 6. This maze behavior, which predicted hippocampal CA1 injury, was not due to different habituation rates nor baseline hyperactivity. Conversely, open field sessions on day 8 revealed ischemic habituation rate deficits. Ischemia also impaired working memory in the T-maze. Delayed hypothermia, which reduced neuronal loss in the CA1 sector to 12% from 81%, reduced all functional impairments. Ischemic gerbils quickly developed spontaneous locomotion hyperactivity that returned to near-normal after 1 day. This motor hyperactivity did not explain the elevated activity found with delayed testing in novel environments. Regardless, only the open field test on day 8 revealed a habituation-like deficit. q 1998 Elsevier Science B.V. All rights reserved. Keywords: Cerebral ischemia; Postischemic hypothermia; Learning; CA1; Gerbil; Habituation; Working memory 1. Introduction The CA1 zone of the hippocampus is exquisitely sensitive to cerebral ischemia with brief periods of global ischemia in humans and rodents leading to a 2 3 day delayed loss of CA1 neurons w14,19,20 x. Moderate to extensive loss in the CA1 sector causes dense anterograde amnesia in humans w26x with comparable functional impairments Ž e.g., working memory. in rodents w1,5,22,23 x. Locomotor hyperactivity is a consistent and predictive feature of ischemic injury to the CA1 sector in gerbil w2,3,11,13,16,18,24 x. The degree and duration of this motor hyperactivity depends upon insult severity but can last at least 48 h w18 x. Since this hyperactivity occurs in familiar ) Corresponding author. Fax: q ; colbourn@acs.ucalgary.ca environments Ž i.e., home cage. w11,18 x, it is unlikely to be due to an exploration deficit. However, several studies suggest that ischemia-induced hyperactivity in unfamiliar environments represents some form of spatial habituation deficit as increased activity is dependent upon novelty w2,24 x. For example, Wang and Corbett w24x found that postischemic hyperactivity in an open field was blocked with pre-ischemic exposure to the test. These authors did not control for the confounding effects of locomotor hyperactivity that is unrelated to learning. Thus, in this study, we measured postischemic motor activity in a familiar environment Ž home cage. while intermittently measuring novelty-induced activity Ž new mazes. to differentiate the two forms of hyperactivity. In addition, delayed open field testing was used to further characterize the effects of ischemia on short- and long-term habituation. Previous studies w5,10x have demonstrated short- but not long-term

2 70 F. Colbourne et al.rbrain Research habituation impairments with measures of locomotion in the open field. Thus, in an effort to better understand exploratory behavior and the effects of ischemia, locomotion and rearing, which are both thought to reflect exploration andror arousal were measured. Hippocampal CA1 cell injury and behavioral deficits in gerbil are markedly attenuated with postischemic hypothermia. While brief postischemic hypothermia Ž 3 h. only transiently reduces CA1 sector cell loss w12 x, longer intervals Ž e.g., 24 h. provide extensive and lasting protection w4,5,8,9,11 x. A previous study found minimal histological and behavioral protection with a 24-h period of hypotherwx 5. Thus, this study also mia induced at 4-h postischemia examined whether a 2-day period of hypothermia, induced after a 6-h intervention delay, provides histopathologic and functional protection. Since not all neurons of the CA1 sector may be permanently salvaged even with prolonged cooling w4,5,9 x, a 30-day survival time was used. 2. Methods 2.1. Subjects The present animals were part of an experiment that also examined the effects of behavioral testing on ischemic neuronal injury. Those data, which include the specific details of the model and the induction of hypothermia, are presented elsewhere w27 x. Female Mongolian gerbils were obtained from High Oak Ranch Ž Baden, ONT. at f weeks of age Ž f 54 g. and divided into three groups: SHAM Ž Ns14., ISCH Ž Ns12., and ISCHqHYPO Ž Ns20.. During the experiment, animals had free access to food and water and were housed individually under diurnal light conditions Ž lights on at 0800 and off at This experiment conformed to the standards of the University of Calgary s Animal Care Committee and the guidelines of the Canadian Council on Animal Care Ischemia and temperature control Gerbils were implanted with a guide cannula as dew6,7 x. After 2 days, telemetry brain scribed previously temperature probes Žmodel XM-FH, Mini-Mitter, Sunriver, OR. were inserted. Subsequently, animals were housed separately in cages Ž23-cm wide = 14-cm high = 31-cm long., which rested upon receivers interfaced to a computer running DataQuest IV Ž DataSciences, St. Paul, MN.. Baseline striatal temperature and general activity Ž30 s sum. were recorded every 30 s and averaged every 5 min. Significant movement of the gerbil s head in the vertical Ž e.g., rearing. and horizontal directions Ž e.g., walking. over the receiver were counted. Ischemia or sham occlusion surgery was performed between 0800 and 1300 on the fourth day following cannula implantation according to previously described methods w5,7 x. Briefly, sham occlusion or bilateral artery occlusion Ž 5-min duration. was performed under halothane anesthesia while core and brain Ž 36.08C. temperatures were regulated at normothermia wx 7. Following ischemia, one group Ž ISCH. had their temperature regulated at a mild hyperthermic pattern as develw4,5x while the ISCH q HYPO group ops spontaneously was subjected to over 2 days of mild hypothermia Ž 32r348C each for 24 h., which was slowly Ž 18Cr30 min. induced starting at 6-h postischemia. Temperature regulation after surgery was achieved in the awake, freely-moving animal by a precise Ž "0.28C from target. servo-controlled system, which utilized infrared lamps, fans and fine water misters wx 7. All groups were monitored for f 6.5 days at which time they were briefly anesthetized and their brain probes removed. In addition to ad lib rat chow, gerbils were given f 7 g of mixed hamster food ŽStaple VME Diet, Hagen, Montreal, Quebec., which contained about 10 sunflower seeds, for the first 4 days following surgery. This seed mixture is preferred over their regular diet and has the advantage that it does not get soaked by the water misters. All gerbils first and primarily ate the sunflower seeds BehaÕioral testing All groups were exposed to three Ž A, B and C. initially novel and distinct mazes Ž Fig. 1. on days 4 6 following ischemiarsham occlusion surgery. Gerbils were placed in each maze for 15 min once a day Ž i.e., 3 mazesrday. over three consecutive days. The three maze sessionsrday, which were between 1300 and 1600, were separated by f1 h Ž start to start.. The mazes were administered in a quiet lab with overhead cues that included the temperature regulation equipment. The presentation orders were the six permutations of these mazes Ži.e., ABC, ACB, BAC, BCA, CAB, CBA.. A particular order was randomly chosen for each gerbil Ž constant over days. such that presentation order was counterbalanced. This was a prerequisite for a statistical analysis of habituation trends within each maze over days and for habituation trends within a day from the first to third maze session. The mazes, which were constructed of black plexiglass with a transparent cover, mea- Fig. 1. Diagrams of maze A, B and C. These mazes were constructed of black plexiglass with a transparent cover. Gerbils were placed in each maze for 15-min sessions while the maze rested upon the receiver.

3 F. Colbourne et al.rbrain Research sured 22-cm wide=32-cm long=15-cm high Žsame size as home cage.. Each maze had a unique internal design thought to be distinguishable by the gerbils. Movement in the mazes, which was thought to reflect exploration, was quantified by the activity counts collected by DataQuest IV. Data were averaged over three 5-min intervals during Fig. 2. Brain temperature Ž A. during Ž 0 h. and after ischemiarsham occlusions presented as 5-min averages. Telemetry-measured activity data Ž B and C. are relative Ž vs. baseline mean of f4.5 counts. measures that were determined from detecting transmitter movements over the receiver. The increases in temperature in the SHAM, ISCH and ISCHqHYPO groups at 102, 126 and 150 h were due to placement in the small mazes. Brain temperature of SHAM and ISCH groups increased by f18c after placement in these mazes. Daily activity count averages are given in C. The hypoactivity of ISCHqHYPO gerbils during and for a short period following cooling is similar to that previously found in normal or sham-operated gerbils subjected to hypothermia Žunpublished data from wx 7 and Colbourne et al., in preparation.. This hypoactivity recovered to normal prior to the small maze sessions. Conversely, untreated ischemia caused hyperactivity which largely abated prior to day 4. The slightly elevated activity in the ISCH group on days 4, 5 and 6 Ž C. is, at least in part, due to their significantly greater activity during the small maze session Ž Fig. 3..

4 72 F. Colbourne et al.rbrain Research each maze session. Data were analyzed by various ANOVAs with specific contrasts. The activity data from one SHAM gerbil were not recorded due to experimenter error. On the morning of day 8 following ischemiarsham operation, all gerbils were placed in an open field Ž1 m=1 m box; 60-cm high. for three 10-min sessions Ž1-h inter-test interval. while their activity was videotaped. The open field was located in a quiet room with constant environmental conditions. The videotapes where analyzed by two people blinded to the animals treatments. As a measure of exploration, the number of squares entered by the gerbil s head during each min of the 10-min sessions was counted by dividing the box into 16 equal squares Žon video monitor.. Additionally, the number of bipedal rears Žon the two hind legs. were counted as this is also believed to measure exploration. The effects of ischemia on rearing behavior had not been previously examined. The data were statistically analyzed with ANOVA and specific contrasts. Four animals were tested but the data were lost due to video camera failure. On the morning of days 9 and 10, all groups were familiarized to the T-maze Ž60-cm stem length, 30-cm arm lengths, 10-cm wide, 12-cm high. for three 5-min sessionsrday. During these sessions gerbils had free access to both arms where they found and ate a shelled sunflower seed located in the food cup at the end of each arm. During the next 6 days, gerbils were given 10 pairs of trialsrday. Each pair consisted of a forced Ž FT. and choice trial Ž CT., which were separated by 15 s. On the FT, gerbils were allowed entry into only one arm Ždoor blocked opposite arm. where they ate their reward Ž 1r2 sunflower seed. and were then returned to the start area Ž door closed.. The forced arm was randomly chosen with the constraint that there were 5 left and 5 right arm FT entriesrday. On the CT, the start area door was opened and animals could enter either choice arm, but were only rewarded Ž 1r2 seed. if they choose the opposite arm to the FT Ž win-shift rule.. Animals were sequentially habituated and trained in groups of 4. The T-maze was located in a quiet room under invariant environmental conditions. It is important to note that gerbils were not food deprived as they always had ad lib rat chow in their home cages. A similar T-maze procedure, which measures spatial working memory, was previously used in the gerbil and found to reflect hipw1,5 x. The data were analyzed with pocampal integrity ANOVAs, specific contrasts and simple effects comparisons for the % correct, FT latency and CT latency data Histology Gerbils were sacrificed 1 month following ischemia and perfused with 4.0% formaldehyde. Brains were then processed, embedded in paraffin, coronally sectioned at 6 mm and stained with haematoxylin and eosin. Viable Žnot eosinophilic. neurons in the CA1 area were counted in medial, middle and lateral sectors Ž each 0.2-mm long. at y1.7, and y2.2 mm to bregma w15x as described previwx 5. Counts were averaged and expressed as a per- ously centage of SHAM. 3. Results Brain temperature Ž Fig. 2a. was regulated during Žaver- age of C. and after ischemia as desired Žsee Section 2.. One SHAM gerbil was excluded for technical reasons while one ISCH q HYPO gerbil died unexpectedly Fig. 3. Telemetry-measured activity data during maze A over days 4, 5 and 6. Data Ž mean" S.D.. are given for each 5-min interval of the 15-min sessions. Note that activity in the home cage on days 4 to 6 was quite similar among all groups Ž - 3 count differences. and, therefore, the significantly elevated activity of the ISCH group was not solely due to motor hyperactivity. While ISCH gerbils had higher activity levels Ž by f15 counts. than SHAM and ISCHqHYPO groups, all groups had similar habituation trends. Identical results were found in maze B and C. In addition, the decline in scores within a day from the first to third maze session was similar between groups indicating similar habituation trends. Only the absolute activity levels differed between groups.

5 F. Colbourne et al.rbrain Research at 19 days postischemia. Since this gerbil had completed behavioral testing and had an intact CA1 zone, its behavioral data were included. Average baseline activity Žmovement of telemetry probe. in the familiar home cage ranged from 4.1 to 4.8 countsr30-s period. Following ischemia, all groups showed profound Ž f 7-fold increase. and significant Žt-tests on 6-h means, p elevations in motor activity over SHAM gerbils. This hyperactivity, which started at f 1-h postischemia and peaked at 6 h Ž Fig. 2b., occurred in the familiar home cage. There were no significant differences between ischemic groups Ž t- 1. up to this time. Activity of ISCH gerbils steadily declined to near-normal levels by the second postischemic day Ž i.e., within 2.5 activity units.. Nonetheless, the mean Ž 24-h average. activity scores ŽFig. 2c. of the ISCH group were statistically higher than SHAM gerbils for 6 days Ž p F Hypothermia caused a faster decline in activity, which started at the onset of cooling. The ISCH q HYPO group was slightly hypoactive on days 2 and 3 Ž pf vs. SHAM. due to the effects Fig. 4. Open field locomotion scores Ž squares crossedrmin, A. for the three 10-min sessions, which were given on day 8 following ischemiarsham operation. Each session was separated by 1 h. The number of bipedal rears is given in B Ž) denotes p-0.05 for SHAM vs. ISCH with a for significant ISCHqHYPO vs. ISCH comparisons.. A short-term habituation impairment Ž squares crossed. was found in the ISCH gerbils on session 1. The significant Group=Minute interaction on the third session is apparently due to complete habituation in SHAMs while the ISCH group were still learning.

6 74 F. Colbourne et al.rbrain Research of hypothermia. Subsequently, the hypothermic group had similar activity as the SHAM group. The ISCH gerbils displayed significantly Ž1 between y2 within ANOVA. higher activity Žmean of f 15 counts higher. than SHAM gerbils during all mazes Žp F ; Fig. 3.. Note that on these days all groups had activity scores in the home cage, which differed by less than 3 counts. The ISCH q HYPO gerbils were similar to the SHAM group in the three mazes Ž F-1. and were always significantly lower than the ISCH group Ž p F Overall, mean scores in each maze correlated highly with other maze scores Ž r s , p In each maze, there were significant Ž p Day Ž 4, 5 and 6. and Time effects Ž3=5-min intervals during each session. because of declining movement within each session and over days. However, a significant Group=Day interaction was noted in maze A Ž p s and B Ž ps , but not in maze C Ž F-1.. In maze A and B, this was due to the greater decline in scores over days in the SHAM and ISCH group, which were similar Ž pg , than in the ISCHqHYPO gerbils Ž pf Importantly, the Group= Time interactions were not significant in maze A Ž ps , B Ž F-1. or C Žps because all groups had similar trends within each session Ž i.e., no short-term habituation rate differences.. The Day = Time = Group interactions were not significant Ž p G in any maze. The Presentation order effect Ž 1st, 2nd and 3rd maze exposurerday. was significant Ž p but the Group = Presentation order interaction was not Ž F- 1. as all groups showed a similar decline in activity from the first to third maze session on each day Ž i.e., no long-term habituation rate differences.. The Presentations order = maze Order Ž ABC, etc.. interaction was not significant Ž ps When the data were analyzed as a 2 between Ž Group, maze Presentation order. y2 within Ž Day, Maze. ANOVA, no significant Order Ž F-1., Group=Order Ž F-1. or Day=Group=Order Ž p s interactions were found, thus affirming the validity of the aforementioned statistical comparisons. There were significant group differences in the open field on the number of squares crossed Ž Fig. 4a. and the rearing counts Ž Fig. 4b.. The ISCH group crossed more squares than SHAM animals on all sessions, but this was not significant at all time points. In fact, there were significantly different patterns Ži.e., rates of habituation; Group= Minute interaction. between SHAM and ISCH gerbils within sessions 1 Ž p and 3 Ž ps However, the decline in movement over sessions was similar between these groups ŽGroup= Session interaction; F- 1.. Hypothermia treatment reduced this impairment in the open field but this was only occasionally significant. On the first session, all groups had similar rearing count trends. On the second and third sessions, the ISCH group often had higher rearing activity than SHAM animals, but this was only significant during 2 min of the second session. Thus, as all groups scores declined over days Fig. 5. T-maze performance Ž top; % correct"s.d.. on days 11 to 16 following ischemiarsham operation. Significant impairments Žlower; % correct. were noted in the ISCH group that were attenuated in the ISCHqHYPO group Ž) denotes p-0.05 for comparison with ISCH group.. These differences were not due to FT Ž middle. or CT latencies Ž bottom., which were similar among groups. there appeared to be a greater shift to rearing in the untreated ischemic animals. There were significant group differences Ž p in the acquisition of the T-maze win-shift rule Ž Fig. 5. that were not explained by the FT Ž F-1. and CT Ž ps

7 F. Colbourne et al.rbrain Research latencies Ž group means differed by -1 s.. The FT and CT latencies were the highest on the first training day after which response times improved and averaged less than 3 s. The ISCH group was significantly impaired Žlower % correct. vs. the SHAM Ž p s and ISCH q HYPO groups Ž p while the latter groups were not significantly different overall Ž p s These overall comparisons generally mirrored individual day contrasts, which were done with One-way ANOVA s and t-tests because of heterogeneity of variance ŽISCHq HYPO group had less variability.. There was a significant Day effect Ž p but not a significant Day = Group interaction Ž p s as all groups had comparable trends over days. The ISCH group performed worse than SHAM animals and never reached the same level of performance. The ISCH q HYPO group sometimes performed better than SHAM, but this was only significant Ž p s on day 13. Extensive Žcell counts averaged 19% of SHAM" 16S.D.. CA1 sector damage Žsee Colbourne et al., submitted. occurred in the ISCH group Ž p Hypothermia significantly Ž p and almost completely Ž88% of SHAM" 14S.D.. attenuated neuronal injury in the CA1 zone at a 30-day survival. Many, but not all of the behavioral measures, correlated with neuronal counts in the CA1 sector when all groups were included in the linear regression analysis ŽTable 1 and Fig. 6.. However, except for some of the activity measures Že.g., mean spontaneous activity during the first day. in the ISCH group, most of the behavioral tests failed to significantly correlate with cell counts in the CA1 zone when only the ISCH or ISCHq HYPO animals were considered. Thus, while significant Group effects were noted on all tests, CA1 sector cell count variability within injured Ž ISCH. or protected Ž ISCHq HYPO. groups was not consistently related to behavioral performance. 4. Conclusions Untreated gerbils experienced a significant degree of motor hyperactivity during the first day following ischemia and this was highly predictive of CA1 sector injury. Since this occurred in the familiar home cage environment, this hyperactivity is not easily explained as an anterograde spatial learning impairment even though this hyperactivity predicted CA1 area cell loss. While this early motor hyperactivity might be related to motor system damage, this hyperactivity does not correlate with striatal or cortical damage w16 x, which is minimal in the gerbil 5 min bilateral carotid artery model. The use of the three novel small mazes was a relatively good predictor of CA1 neuronal injury. Accordingly, this simple testing strategy may prove useful for gauging neuroprotection in other ischemia studies. However, the ischemia-induced hyperactivity, which occurred in these mazes cannot be ascribed to a habituation impairment as SHAM and ISCH groups showed a similar decline in activity within each session, within each day, and over days. Thus, it would appear that ISCH gerbils were simply hyperactive Ž by f15 counts., if it were not for the fact that activity in the home cage was similar Ž -3 counts. between groups on the test days. The cause of ischemic hyperactivity in these mazes remains to be determined. The open field data revealed a short-term habituation rate impairment in ISCH gerbils as SHAM gerbils habituated faster Ž i.e., greater decline in squares crossed. on the first open field session. Since both groups started at the same activity level, the ISCH gerbils were not simply hyperactive. This confirms previous work, which found a within session habituation rate impairment on the first open field session w5,10 x. Rearing did not significantly differ on the first session. The decline in locomotor activity over sessions, or long-term Ž3-h span from first to third Table 1 Correlations Ž Pearson r and p values given. between the various behavioral measures and average CA1 sector cell counts were done with all animals included and by group VS CA1 All groups SHAM ISCH ISCHqHYPO 0 6-h spontaneous activity rsy0.235, ps0.073 rsy0.572, ps0.052 rsy0.432, ps0.160 rs0.428, ps h spontaneous activity rsy0.739, p rsy0.657, ps0.020 rsy0.781, ps0.003 rsy0.458, ps0.056 Maze A mean activity rsy0.711, p rsy0.462, ps0.130 rsy0.759, ps0.007 rsy0.171, ps0.498 Maze B mean activity rsy0.668, p rsy0.057, ps0.860 rsy0.611, ps0.046 rsy0.135, ps0.593 Maze C mean activity rsy0.573, p rsy0.349, ps0.266 rsy0.527, ps0.096 rsy0.188, ps0.454 Mean activity of mazes A, B and C rsy0.710, p rsy0.399, ps0.199 rsy0.699, ps0.024 rsy0.200, ps0.426 Open field Ž a1. squares crossed rsy0.455, ps0.004 rs0.182, ps0.551 rsy0.407, ps0.189 rsy0.100, ps0.734 Open field Ž a2. squares crossed rsy0.527, ps0.001 rs0.002, ps0.996 rsy0.490, ps0.106 rsy0.148, ps0.614 Open field Ž a3. squares crossed rsy0.478, ps0.002 rsy0.210, ps0.492 rsy0.518, ps0.084 rsy0.337, ps0.238 Open field mean squares crossed rsy0.528, ps0.001 rs0.002, ps0.995 rsy0.509, ps0.091 rsy0.215, ps0.460 Open field Ž a1. rearing rsy0.049, ps0.765 rs0.151, ps0.622 rs0.123, ps0.704 rs0.130, ps0.659 Open field Ž a2. rearing rsy0.293, ps0.070 rs0.102, ps0.742 rsy0.271, ps0.395 rs0.095, ps0.747 Open field Ž a3. rearing rsy0.342, ps0.033 rs0.261, ps0.390 rsy0.413, ps0.182 rs0.057, ps0.847 Open field mean rearing rsy0.283, ps0.080 rs0.188, ps0.539 rsy0.276, ps0.386 rs0.120, ps0.684 T-maze mean % correct rs0.624, p rs0.187, ps0.541 rs0.519, ps0.084 rsy0.027, ps0.916

8 76 F. Colbourne et al.rbrain Research Fig. 6. Scatter plots comparing the mean CA1 sector cell counts with: Ž A. spontaneous activity on the first day following ischemiarsham operation; Ž B. average activity in the small mazes; Ž C. average movement activity Ž squares crossed. in the open field sessions; and Ž D. average T-maze performance. There was a significant linear relationship Ž Table 1. between CA1 sector injury and each of the behavioral measures when all groups were considered. However, none of the behavioral tests significantly predicted CA1 sector neuronal counts within the ISCHqHYPO group, and only some of the activity scores predicted CA1 sector outcome within the ISCH group. session. habituation, was similar between normal and ischemic animals as previously found with longer inter-trial intervals w4,5,10,17 x. This apparent lack of habituation rate differences may be partially due to a shift from locomotion to rearing since ISCH gerbils seemed to rely more on rearing to explore the open field on subsequent sessions. Accordingly, this may have masked an impairment in spatial exploration as measured by locomotion. In contrast to the first open field session, Group effects in the small mazes cannot be ascribed to habituation deficits. This may relate to features of the testing procedure Ž e.g., size of maze, delay postischemia. or be due to

9 F. Colbourne et al.rbrain Research quantification differences. Notably, activity counts by the telemetry system include both bipedal rearing and locomotion. The reliance solely upon this movement score, as we have done in the small mazes, may give a false assessment of habituation ability and thereby, mask real exploratory impairments in ischemic animals. Nonetheless, activity in the small mazes, or open field, are quick and easy ways to predict histological neuroprotection Ž area CA1 function., which is needed in the evaluation of potential neuroprotectants. The T-maze revealed significant ischemic Ž ISCH. working memory impairments Ž lower % correct., which were abolished by 6-h delayed hypothermia Ž ISCHq HYPO. to a similar level previously found with 1 h delayed cooling wx 5. Group differences were not due to motor abnormalities or differences in motivation as all groups had similar CT and FT latencies. Furthermore, the gerbils were not food deprived and similarly ate all of the sunflower seeds available on the habituation trials. The observed differences were also not due to impaired habituation as gerbils were given an extensive habituation time Žsix 5-min sessions over 2 days.. The similar FT and CT latencies support this as one would expect higher latencies in gerbils with less habituation. This study also shows that 6-h delayed hypothermia can significantly and persistently Ž 30 day. reduce both ischemic injury to the CA1 area and associated functional impairments. This is a substantial improvement over previwx 5 where we found only limited CA1 neuronal ous work protection and minimal functional protection in the open field with a 4-h delayed hypothermic intervention Ž24-h duration. following ischemia. The superior efficacy in this study was likely due to the use of a longer duration of hypothermia Ž ) 48 h.. Activity in the small mazes and the open field Žsquares crossed. correlated significantly with CA1 sector cell counts as did the average performance in the T-maze. In fact, the quickest behavioral tests, the small mazes, were among the best predictors of CA1 zone cell counts. As expected w5,21,25 x, scores overlapped between normal and CA1 sector damaged gerbils on all behavioral tests and this is partly due to intrinsic inter-animal variability in performance. Likewise, most behavioral indices did not precisely correlate with CA1 zone neuronal counts when only ISCH or ISCHq HYPO gerbils were analyzed. Precise intra-group correlations cannot be expected due to the smaller sample sizes and the more limited range of CA1 zone cell counts. In addition, unassessed neuronal injury in other structures Ž e.g., hilus. and compensatory mechanisms Ž e.g., cingulate cortex. obfuscate the relationship between CA1 sector injury and performance. In summary, this study clearly showed a dissociation between novelty induced hyperactivity and spontaneous hyperactivity in ischemic gerbils. However, the results did not implicate a habituation-like impairment in the small mazes. Nonetheless, activity in such novel environments was a useful and simple way to gauge CA1 area integrity. Deficits in the open field and T-maze were more consistent with the role of the hippocampus in spatial learning and working memory. Finally, this study showed significant functional and histological neuroprotection with 6-h delayed postischemic hypothermia, which make this a clear candidate for clinical investigation. Acknowledgements The authors acknowledge the technical support of Bonnie Colbourne, Drs. Fangwei Yang and Dubravka Rakic. Research support was provided by the Heart and Stroke Foundation of Alberta. Dr. Colbourne gratefully acknowledges postdoctoral fellowship support from the Heart and Stroke Foundation of Canada. References wx 1 A.M. Babcock, H. Graham-Goodwin, Importance of preoperative training and maze difficulty in task performance following hippocampal damage in the gerbil, Brain Res. Bull. 42 Ž wx 2 A.M. Babcock, D.A. Baker, R. Lovec, Locomotor activity in the ischemic gerbil, Brain Res. 625 Ž wx 3 M.J. Chandler, J. DeLeo, J.M. Carney, An unanesthetized-gerbil model of cerebral ischemia-induced behavioral changes, J. Pharmacol. Toxicol. Methods 14 Ž wx 4 F. Colbourne, D. Corbett, Delayed and prolonged post-ischemic hypothermia is neuroprotective in the gerbil, Brain Res. 654 Ž wx 5 F. Colbourne, D. Corbett, Delayed postischemic hypothermia: a six month survival study using behavioral and histologic assessments of neuroprotection, J. Neurosci. 15 Ž wx 6 F. Colbourne, S.M. Nurse, D. Corbett, Temperature changes associated with forebrain ischemia in the gerbil, Brain Res. 602 Ž wx 7 F. Colbourne, G.R. Sutherland, R.N. Auer, An automated system for regulating brain temperature in awake and freely moving rats, J. Neurosci. Methods 67 Ž wx 8 F. Colbourne, G. Sutherland, D. Corbett, Postischemic hypothermia: a critical appraisal with implications for clinical treatment, Mol. Neurobiol. 14 Ž wx 9 F. Colbourne, G.R. Sutherland, R.N. Auer, New features of delayed neuronal death in gerbils treated with prolonged postischemic hypothermia, Soc. Neurosci. Abstr. 23 Ž w10x D. Corbett, P. Crooks, Ischemic preconditioning: a long term survival study using behavioural and histological endpoints, Brain Res. 760 Ž w11x D. Corbett, S. Nurse, F. Colbourne, Hypothermic neuroprotection: a global ischemia study using month old gerbils, Stroke 28 Ž w12x W.D. Dietrich, R. Busto, O. Alonso, M.Y.-T. Globus, M.D. Ginsberg, Intraischemic but not postischemic brain hypothermia protects chronically following global forebrain ischemia in rats, J. Cereb. Blood Flow Metab. 13 Ž w13x S.C. Gerhardt, C.A. Boast, Motor activity changes following cerebral ischemia in gerbils are correlated with the degree of neuronal degeneration in hippocampus, Behav. Neurosci. 102 Ž w14x T. Kirino, Delayed neuronal death in the gerbil hippocampus following ischemia, Brain Res. 239 Ž

10 78 F. Colbourne et al.rbrain Research w15x W.J. Loskoto, P. Lomax, M.A. Verity, A Stereotaxic Atlas of the Mongolian Gerbil Brain Ž Meriones unguiculatus., Ann Arbor Sci. Publ., Ann Arbor, MI, w16x B.E. Mileson, R.D. Schwartz, The use of locomotor activity as a behavioral screen for neuronal damage following transient forebrain ischemia in gerbils, Neurosci. Lett. 128 Ž w17x S. Nurse, D. Corbett, Direct measurement of brain temperature during and after intraischemic hypothermia: correlation with behavioral, physiological, and histological endpoints, J. Neurosci. 14 Ž 12. Ž w18x S. Nurse, D. Corbett, Acute and chronic locomotor patterns as predictors of ischemic outcome, Soc. Neurosci. Abstr. 21 Ž w19x C.K. Petito, E. Feldmann, W.A. Pulsinelli, F. Plum, Delayed hippocampal damage in humans following cardiorespiratory arrest, Neurology 37 Ž w20x W.A. Pulsinelli, J.B. Brierley, F. Plum, Temporal profile of neuronal damage in a model of transient forebrain ischemia, Ann. Neurol. 11 Ž w21x M.R. Rod, I.Q. Whishaw, R.N. Auer, The relationship of structural ischemic brain damage to neurobehavioural deficit: the effect of postischemic MK-801, Can. J. Psychol. 44 Ž w22x B.T. Volpe, B. Waczek, H.P. Davis, Modified T-maze training demonstrates dissociated memory loss in rats with ischemic hippocampal injury, Behav. Brain Res. 27 Ž w23x B.T. Volpe, H.P. Davis, A. Towle, W.P. Dunlap, Loss of hippocampal CA1 pyramidal neurons correlates with memory impairment in rats with ischemic or neurotoxin lesions, Behav. Neurosci. 106 Ž w24x D. Wang, D. Corbett, Cerebral ischemia, locomotor activity and spatial mapping, Brain Res. 533 Ž w25x I.Q. Whishaw, M.R. Rod, R.N. Auer, Behavioral deficits revealed by multiple tests in rats with ischemic damage limited to half of the CA1 sector of the hippocampus, Brain Res. Bull. 34 Ž w26x S. Zola-Morgan, L.R. Squire, D.G. Amaral, Human amnesia and the medial temporal region: enduring memory impairment following a bilateral lesion limited to field CA1 of the hippocampus, J. Neurosci. 6 Ž w27x F. Colbourne, R.N. Auer, G. Sutherland, Behavioral testing does not exacerbate ischemic CA1 damage in gerbils, Stroke, in press.