| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
From the *Department of Anesthesia, Cincinnati Children's Hospital and University of Cincinnati, Cincinnati, Ohio; and Department of Pediatrics, Division of Neurology, Cincinnati Children's Research Foundation and University of Cincinnati, Cincinnati, Ohio.
Address correspondence and reprint requests to John J. McAuliffe, Department of Anesthesia, Cincinnati Children's Hospital, 3333 Burnet Ave., Cincinnati, OH 45229. Address e-mail to john.mcauliffe{at}cchmc.org.
 |
Abstract |
|---|
 Top Abstract IntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
METHODS: Nine-day-old C57x129T2 F1 mice received either 1.8% Iso, hypoxic (10% O2 in N2), or sham (room air) preconditioning. The following day, the mice were subjected to permanent right common carotid ligation or sham ligation followed by 65 min of hypoxia, or room air. At 70 days of age, learning was tested using a series of Morris water maze tests. Striatal function was assessed by response to apomorphine injection. Histological analysis was performed on adult brain (P120) sections of striatum and dorsal hippocampus.
RESULTS: Iso preconditioning 24 h before severe neonatal hypoxia–ischemia reduced preweaning mortality from 20% to 0% (P < 0.04) and improved striatal function in adult mice, as assessed by circling after apomorphine injection (P < 0.028), but no improvements in performance were noted in the spatial-reference memory water maze tests. Hypoxic preconditioning improved learning relative to the sham-preconditioned group on the hidden maze, but not the more difficult reduced maze test of spatial memory. It had no significant effect on preweaning mortality and apomorphine response. Histologic analysis showed the hippocampus of non-preconditioned and Iso-preconditioned animals to be equally injured.
CONCLUSION: Iso and hypoxia confer selective functional neuroprotection in a delayed preconditioning paradigm in neonatal mice.
|
Introduction |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Zhao and Zuo (2) demonstrated that neonatal rat pups are protected from hypoxic-ischemic (HI) injury in a delayed protection paradigm. Isoflurane was administered at 1 or 1.5% for 30 min 24 h before HI. The isoflurane-preconditioned animals had significantly greater hemisphere weight ratios than non-preconditioned animals seven days after the HI insult. However, the long-term outcome of anesthetic preconditioning has not been studied in a neonatal model of HI injury.
The current study was designed to compare the long-term outcomes of mice that were preconditioned with hypoxia, isoflurane, or room-air (sham preconditioning) 24 h before 65 min of HI on postnatal day 10 (P10). We hypothesized that both isoflurane and hypoxic preconditioning would result in improved adult dopaminergic function and improved adult learning after neonatal HI compared to sham-preconditioned mice. Additionally, we investigated the immediate and long-term consequences of isoflurane preconditioning in the absence of HI.
Learning was assessed using a Morris water maze sequence designed to test cued and spatially based learning. Spatial memory is generally accepted to be dependent on an intact hippocampus (6). Cued learning is partially dependent on the hippocampus and on other structures, such as the dorsal striatum (7,8) and superior colliculus (6). The integrity of the dopaminergic component of the motor system after unilateral HI can be tested with dopamine agonists such as apomorphine (9). Animals that have sustained unilateral loss of dopaminergic fibers in the striatum circle more than control animals toward the side of the injury (10).
|
METHODS |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
Preconditioning
Our initial studies showed that breeding pair, but not gender, was a significant determinant of performance on water maze testing. Therefore, the present study was blocked on mating pair only. Twenty pups per mating pair (five pairs) were randomized to one of five treatments. Thus, each mating pair contributes four mice to each group. The actual numbers randomized to each treatment group were:
Litters larger than 10 pups were culled to 10. On postnatal day 9 (P9) mice from a litter were randomized to a group based on the mating pair producing the litter. The P9 mice were preconditioned for a period of 2 h according to the group assignments. After recovery from the preconditioning protocol the pups were returned to the dams and kept in pathogen-free housing until the next day.
Preconditioning
Hypoxia was maintained with 10% oxygen in nitrogen administered in a temperature-controlled chamber designed to maintain the ambient temperature between 33.8°C and 34.5°C. Anesthesia was maintained with 1.7–1.8% isoflurane in oxygen administered in an incubator set to maintain an environmental temperature of 34.5°C. Gas concentration was monitored with a Datex-Ohmedia gas analyzer. Animals maintained in air were placed in chambers designed to maintain environmental temperature in a range of 33.8–34.5°C. After the 2 h preconditioning period, all mice were recovered in room air with an ambient temperature of 34.0–34.5°C until they demonstrated a righting reflex in <5 s. After recovery, the preconditioned animals were returned to the dam.
HI
On day P10, 21.5 h after the end of the preconditioning period, mice were anesthetized with 3% isoflurane in oxygen, the right common carotid artery was double-ligated with 5-0 silk ligatures then divided between the ligatures (RCL) followed by wound closure. The exposure to isoflurane was <10 min for the carotid ligation procedure (RCL) and <6 min for the sham RCL. Sham operated mice had the carotid artery exposed but not ligated (Sham RCL). After recovery from anesthesia in room air, the pups were returned to the dam for 2 h. All pups nursed during this time. Twenty-four hours after the end of the preconditioning period mice in AS and SS groups were exposed to room air (sham) and the HH, AH and SH groups were exposed to 65 min of hypoxia (10% O2/90% N2) in the same temperature-controlled chambers described above. After a 10-min recovery period in room air at an ambient temperature of 34.0–34.5°C, the surviving pups were returned to the dam. Pups were weaned at P26–28 and transferred to conventional housing 5–7 days later.
Blood Chemistry Data
A separate cohort, of P9 mice were used to acquire blood gas, electrolyte, and blood glucose data for each of the preconditioning treatments. None of the mice in the long-term outcome study was used for this part of the study. Blood samples (150–200 µL) were taken either by carotid laceration (anesthetized animals, n = 10) or cardiac puncture (hypoxic animals, n = 10) either at 1 h or 2 h into the preconditioning period. The sample was processed immediately using an I-STAT (Abbott Laboratories) blood gas analyzer with a CG-8 cartridge. The sham-preconditioned animals (n = 5) were briefly anesthetized to obtain blood samples.
Behavioral Testing
Testing was conducted by a single individual who was blinded to the treatment received by the mice. Mice were acclimated to handling 10 min per day for 1 wk before the start of behavioral testing. Testing began on the first Monday after the mice reached 71 days of age (P71). The test sequence was as follows: Week 1, the cued water maze; Week 2, the hidden water maze with probe trial; Week 3, the reduced water maze with probe trial; Week 4, rest; Week 5, the 2-week-delayed probe trial followed by apomorphine challenge.
Morris Water Maze.
Testing was performed in a tank of 122 cm diameter with the water temperature maintained at 21°C as previously described (11). The water was tinted with white tempera paint to obscure the platform. A 10 x 10 cm2 platform was used as the goal for cued and hidden platform testing. A 5 x 5 cm2 platform was used for the reduced platform maze trials. The platform location and entry point were varied according to a preset randomization scheme for the cued phase. The platform location was marked by a visible cue on the platform and curtains were closed around the maze to reduce distal cues. Each mouse was allowed four trials per day for 6 days with a maximum of 60 s per trial and intertrial interval of 15 s. If the mouse failed to find the goal within the allotted time, it was placed on the platform.
The platform was fixed in the southwest quadrant for the hidden platform acquisition phase and in the northeast quadrant for the reduced platform acquisition phase. Animals entered the tank at four different locations each day for 5 days according to a predetermined randomization schedule using locations that were not in the same quadrant as the platform (i.e., northwest, north, east, and southeast, but not west or south for the hidden maze). A single probe trial from a unique start position (northeast for hidden and southwest for reduced) was performed on Day 6. Three distal visual cues were placed on the walls surrounding the tank in addition to the many inherent room cues available to the animals, and the curtains used during cued trials were opened so that animals had access to distal cues. The added cues were placed on each of the three walls nearest the tank. A delayed probe trial was performed in the same tank using the same distant visual cues as used for the hidden and reduced platform testing.
The time required for the mouse to reach the platform from the entry point (platform latency) was recorded by hand for the cued maze. All other testing was performed with automated tracking and data analysis (Polytrack, San Diego Instruments, San Diego, CA). The software was used to extract the selected data parameters, such as platform latency, swim speed and path length.
Platform latency is determined by both motor and cognitive performance (12). Swim speed was used to control for motor function. A mouse may be able to process the visual clues needed to locate the platform with training but swim poorly. The latency (time required to swim to the platform) for this animal will appear abnormal but with training the latency decreases. However, due to the motor deficit (poor swimming) latency will be greater than the controls, but the mouse may have a normal learning curve. A learning curve can be obtained by computing a metric called fractional change in latency (FCL) day (n), defined as [average latency Day 1 – average latency day (n)]/average latency Day 1. This metric can be used as a measure of cognitive function as swim speed tended to be constant across days. FCL was calculated for Days 2–6 for the cued maze and Days 2–5 for both the hidden and reduced maze data. FCL was analyzed by both repeated measures general linear models and ANOVA.
Apomorphine Challenge.
Assessment of the response to apomorphine injection was the last test performed before necropsy because of the potential for apomorphine's effects on subsequent behavioral testing. Mice were observed in a 16 in. x 16 in. clear acrylic chamber for clockwise rotation for a period of 10 min before apomorphine injection (9,10). An intraperitoneal injection of 1.2 mg/kg apomorphine was given to each animal (12,13). The number of clockwise circles made by each animal during the first 15 min after apomorphine injection was recorded. The test was considered positive if an animal circled more times than 90% of the non-HI mice. The data analysis for the apomorphine challenge was performed using the 
2 distribution.
Brain Weights and Histology
After the assessment of circling behavior following apomorphine injection, all mice were killed under anesthesia and perfused with 4% paraformaldehyde. The brains from selected mice from each group were divided along the midline, blotted, and the hemispheres weighed separately. The hemispheres were postfixed in paraformaldehyde, cryoprotected in sucrose, and frozen together in cryomatrix. The brains of the remaining mice were weighed, postfixed intact in paraformaldehyde, cryoprotected in 30% sucrose, and frozen in cryomatrix. Coronal sections were cut to correspond to the structures shown in Figures 27–29 and 49–51 of the Paxinos and Franklin atlas of the mouse brain (14). These sections were stained with hematoxylin and eosin or fast cresyl violet (Nissl).
An additional eight P9 mice were preconditioned for 2 h with 1.8% isoflurane in oxygen, allowed to recover as above, returned to the dam to feed for a period of 3 h (14,15) and then killed under anesthesia and perfused with 4% paraformaldehyde. Sections through dorsal and ventral hippocampus were cut and stained for activated caspase-3, the nuclear envelope antigen (NeuN) and nucleic acid (TOTO-3). Triple-labeled immunofluorescent staining was performed using anticleaved caspase-3 antibody (1:100, cat no. 9661S, Cell Signaling Technologies, Beverly, MA), NeuN (1:1000, cat no MAB377, Chemicon, Temecula, CA), followed by application of Alexa Fluor488 conjugated goat anti-rabbit secondary (1:200, A11034, Molecular Probes, Eugene, OR) and Alexa Fluor 568 goat anti-mouse secondary (1:200, A11031, Molecular Probes) followed by TOTO-3 (1:10,000, T3604, Molecular Probes). Confocal images were acquired using a Zeiss LSM-510 using excitation frequencies of Ar 488nm, He–Ne 543 nm and He–Ne 633 nm. Sections through the hippocampus of mice from a separate study (killed 3 h after 6 h of 1.8% isoflurane in 30% oxygen) were used as a positive control for caspase-3 activity. Cells with activated caspase-3 will appear green in the green channel and in the absence of colocalization will also appear green in the merged channel. If the cell is a mature neuron and the caspase-3 has translocated to the nucleus, the cell will have a white appearance in the merged image due to colocalization of the green caspase-3 signal, the red NeuN signal, and the blue TOTO-3 signal. Nonneuronal cells and NeuN negative neurons with activated caspase-3 in. the nucleus will appear blue-green in the merged image due to colocalization of caspase-3 and TOTO-3.
The area of the striatum (or hippocampus) in each hemisphere was measured using a digital software package (Micron, Westover Scientific) to integrate the area of a circumscribed structure. Striatal area was measured for both the treated and contralateral sides in sections corresponding to Figures 27–29 of the of the Paxinos and Franklin atlas of the mouse brain (13,14). The area of the hippocampus was measured for both the treated and contralateral sides in sections corresponding to Figures 49 –51 of the same atlas. A cystic infarction of a structure was given an area of zero. The area of the treated hemisphere divided by the area of the contralateral side is defined as the area ratio for a structure (striatum or hippocampus). Area ratios were calculated for sections from eight different animals for each of group (HH, AH, SH, and SS). The ratios for the groups were compared using ANOVA.
Statistical Analysis
Data sets were analyzed to compare the outcome of the animals subjected to HI, HH and AH to SH, as one block. The control group for these comparisons was the SH group. The non-HI mice, AH and SS were considered separately; the SS groups served as the control for these comparisons. The mice not subjected to HI, groups AS and SS were compared to determine if there were any long-term detrimental effects of isoflurane preconditioning in the absence of HI.
All data were analyzed using SPSS 11 for MAC. Dichotomous data were analyzed using the 2 distribution. Continuous variables were analyzed using ANOVA. Behavioral data were analyzed using the general linear models two-factor interaction repeated measures model including day and trial as within-subject variables. Variance leveling transformations were used as dictated by residual analysis (15). Post hoc comparisons for ANOVA were made using the Dunnett's t (2-sided) procedure using either group SH (comparing HI groups) or group SS (comparing non-HI groups) as the control. The critical value for 
was set to 0.05
Water maze data were also analyzed using distribution-free techniques. The upper limit of "normal" latency for any given day's testing was defined as the upper 95th percentile latency value for the sham control group. Animals with longer latencies were defined as having "abnormal" performance. The fraction abnormal among groups was compared using 2 statistics using a corrected value for the standard error of the mean proportion (16).
|
RESULTS |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
Caspase-3 Activation
Activation of caspase-3 and DNA fragmentation has been reported in the hippocampus of adult rats 3 h after normalization of blood glucose following profound hypoglycemia. Therefore we performed triple immunofluorescent staining sections of brains removed from a cohort of P9 mice 3 h after recovery from the 2 h of isoflurane preconditioning. The results are shown in Figures 2A–G. Panel G shows the location of the sections in Panels A–F relative to other structures (Ref. 19, with permission) (18). No caspase activity was noted in the sections from the isoflurane-preconditioned mice in the retrosplenial granular cortex (Fig. 2A), hippocampus (Figs. 2B–D), or dorsal-lateral thalamus (Fig. 2E). An isolated activated caspase-3 positive cell is shown in the superficial visual cortex (Fig. 2F). This cell is not a NeuN+ cell, as shown by the lack of NeuN signal adjacent to the arrow identifying the cell position in each panel. Only two to four isolated activated caspase-3 cells were identified in any section through the dorsal hippocampus for all eight of the isoflurane-preconditioned mice killed 3 h after preconditioning. The location of the activated caspase-3 positive cells varied from section to section and ranged from the molecular layer of the dentate gyrus to the sensory cortex. No activated caspase-3 positive cells were noted to be NeuN positive (no yellow cells in merged image). The activated caspase-3 signal was restricted to the cytoplasm in all cases, as there was no co-localization of caspase-3 and TOTO-3 signal (blue-green cells). Significant caspase-3 activation is noted in the mice exposed to 6 h of isoflurane (positive controls) in the CA1 and visual cortex (insets Figs. 2B and 1F) indicating that the antibody is functional.
|
Behavioral Testing
The isoflurane-preconditioned/sham group (AS) performed as well as the sham-preconditioned/sham group (SS) on all water mazes, including the reduced maze (Fig. 3B) and delayed probe trial (Fig. 3C), as well as on apomorphine challenge. (Fig. 4)
|
|
Post-HI Injury outcome
A total of 96 mice were randomized to one of the five groups. Eighty-five survived behavioral testing. The numbers of mice in each treatment group at the time of behavioral testing were unequal due to differences in survival. Type IV sums-of-squares were used to minimize the effects of the imbalance among groups. Behavioral testing data were obtained from 83 mice of the 85 mice that survived to participate in the behavioral testing protocol.
Blood gas data were not obtained from study animals at the completion of HI as these data have been published (17). All animals were allowed to grow to adulthood and complete behavioral testing before they were killed for brain weight measurement and histological analysis.
Mortality
Deaths were ascribed to the effects of HI if the deaths occurred during the period of hypoxia or during the preweaning period as a result of injury sustained during the 65 min of HI. The raw mortality rates for the mice that had HI on P10 are group HH 3 of 19, group AH 0 of 19 and group SH 4 of 20. The AH group had reduced mortality compared to group SH (2 = 4.23, dof = 1, P < 0.04); the mortality for group HH was not significantly different from SH. Three of the four deaths in the SH group and two of the three deaths in the HH group occurred during the 65 min of hypoxia and the other death in each of the HH and SH groups occurred in the preweaning period.
One animal from each of the HH, AH, SH, and SS groups died between P35 and P63. There were no deaths during the behavioral testing phase.
Brain Weights
Body weights at P120 were not different among the groups (data not shown). The P120 (adult) brain weight data are summarized in Table 2. The non-HI mice (AS and SS) had significantly greater total brain weights than the HI mice (groups HH, AH, and SH) (P < 0.001). Total brain weights were not different among the groups that had HI on P10 (HH, AH, and SH). Right to left hemisphere weight ratios were not statistically different among all groups (P = 0.24), in large measure due to the high variance within the SH group. The brain/body weight ratio was significantly greater for the SS group than the HH, AH, and SH groups (P < 0.05).
|
Morris Water Maze
The raw platform latency data for the five groups on the hidden water maze and the reduced water maze are shown graphically in Figures 3A and B. Latency reflects the combined effects of motor and cognitive functions. Analysis of latencies revealed differences among the HI groups during the cued maze testing protocol. The average cued water maze mean-latencies (±sem) of the four trials per day are plotted in Figure 3D for the four treatment groups. Using repeated measures analysis, the control (SS) group performed better than all other groups (P < 0.001). The hypoxia and isoflurane-preconditioned treatment groups (HH and AH) performed significantly worse than the controls but the AH group outperformed the SH mice using repeated measures General Linear Modeling (GLM) analysis (P = 0.045). No differences in latencies over time among groups HH, AH, and SH were found on repeated measures GLM analysis of the hidden and reduced maze data.
The statistical summary of the proportion-abnormal analysis for the cued maze test is shown in Table 3. Using proportion-abnormal analysis, hypoxia preconditioning differs from control on Days 1 and 3. The AH group differs from control on Days 1 and 4 and the SH group is different from control on Days 1–5. Comparisons among the HI groups revealed that one group differed significantly (2 > 5.99, dof = 2, P < 0.05) from the other two groups in percentage of mice with abnormal latencies on Days 1 and 3 of testing. The AH group had a significantly lower percentage of mice with abnormal latencies than the SH group on Days 1 and 3 of testing. There was no case in which the SH group had a lower percentage of mice with abnormal latencies than the AH group. The percentage abnormal analysis is consistent with the repeated measures analysis; the anesthesia-preconditioned animals had lower latencies across the six days than the sham-preconditioned animals on the cued maze.
|
After neonatal HI injury, adult water maze latencies are affected by both cognitive and motor components (12). FCL values were calculated for the acquisition phase of the three water maze protocols in order to isolate the cognitive component of water maze performance. FCL (on the last day of the acquisition phase) results for the three mazes are tabulated in Table 4. The FCL Day 6 values for the cued maze were not different among groups HH, AH, and SH. ANOVA analysis of the fractional change in latency values for Day 5 of the hidden maze show that the HH group had a significantly greater FCL Day 5 than group SH (P = 0.038), while the FCL Day 5 for groups SH and AH were not significantly different (Table 4). No differences in FCL were noted among the animals in groups HH, AH, and SH on the reduced maze.
|
Repeated measures GLM analysis was performed using the calculated values for FCL for Days 2–6 for the cued maze and for Days 2–5 for the hidden and reduced mazes. No differences were found among the HI groups or between groups SS and AS for the cued maze. Analysis of the hidden maze repeated measures GLM showed that group HH was marginally different from group SH (P = 0.052) on post hoc analysis of the HI groups. No other comparisons approached significance for either the hidden or reduced maze FCL repeated measures.
The data from delayed probe trials were not normally distributed and therefore were analyzed using nonparametric methods. The proportion of mice in each group with one or more platform crossings was used as the dependent variable in the analysis of the delayed probe trial. The HI groups had a significantly lower proportion of mice with one or more crossings on the delayed probe trial compared to the non-HI groups (2 = 10.36, dof = 4, P < 0.035). Groups HH, AH and SH were not significantly different from each other (Fig. 3C).
Apomorphine Challenge
At least one mouse in each group, except for AS, had a positive test, or circling, in response to apomorphine using the criteria in the Methods section. The incidence of circling behavior is summarized in Figure 4A. The incidence of circling was significantly different among all groups (2 = 25.58, dof = 4, P < 0.001). The non-HI mice (SS) circled less than the HI groups. Among the mice that had HI on P10, the incidence of circling was significantly reduced only in the isoflurane-preconditioned, (AH) (50% circling) mice compared to the sham-preconditioned HI (SH) mice (87%, 2 = 4.95, dof = 1, P = 0.026). There was no significant difference in the incidence of circling after apomorphine between the AS and SS groups.
The cued maze data were reanalyzed after splitting the HH and AH mice into two groups based on response to apomorphine. The repeated measures ANOVA comparing the cued maze performance of the preconditioned HI mice that circled in response to apomorphine with those that did not circle was highly significant (F1,44 = 8.048, P = 0.007). The difference in proportion abnormal between circling and noncircling mice is significant (2 = 4.56, dof = 1, P = 0.035) for the cued maze. The results are shown in Figure 4B. This difference in cued maze performance between circling and noncircling mice was not due to a difference in FCL between the two groups (P = 0.831). The difference in cued performance is then due to either delayed initiation of motor movement in the circling group relative to the noncircling group, or swim speed. Swim speed was not directly measured for the cued maze, but the average swim speed of the circling and noncircling HI mice was not significantly different on the hidden maze acquisition phase.
Histology
Nissl and hematoxylin–eosin stained sections of adult brains are shown in Figures 5A–F. No overt histological abnormalities were noted in the striatum or hippocampus any of the animals in groups AS or SS (Fig. 5A and D). Most animals exposed to HI on P10 demonstrated injury to the right caudate-putamen and hippocampus but some of the isoflurane-preconditioned mice exhibited little or no injury (Fig. 5B). This injury in the caudate-putamen ranged from minimal tissue loss and no dilation of the lateral ventricle (Fig. 5B) to more than 50% loss with significant dilation of the lateral ventricle (Fig. 5C). The typical pattern of hippocampal injury included significant loss of CA1-CA3 pyramidal cell layers with sparing of the dentate gyrus granule cell layer (Figs. 5E and F). In some animals the hippocampus was replaced with a porencephalic cyst (not shown).
|
The area ratios for both the hippocampus and striatum were not found to be different among the HI groups (Table 5), although the numbers are small. The striatal area ratio was highly correlated with circling in response to apomorphine. A logistic regression using area ratio as the independent value gave a significant model (Hosmer–Lemeshow 2 = 7.246, P = 0.510; Nagelkerke R2 = 0.646)). Area ratios <0.6 were associated with a high probability of circling, therefore the number of brains with striatal area ratios <0.6 were counted for each group. The AH group had significantly fewer mice with striatal area ratios <0.6 compared to group SH (2 = 4.27, P = 0.039).
|
|
DISCUSSION |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
We have not attempted to define a mechanism for the neuroprotection. It would be inappropriate to assume the mechanism to be induction of inducible nitric oxide synthase as shown by Zhao and Zuo (2), because agents which are neuroprotective in rats are not always neuroprotective in mice (19). It is possible that the cellular responses to the isoflurane-induced hypoglycemia are linked to the observed neuroprotection.
Study Design Issues
The duration of the HI insult was chosen so that an unprotected animal would be severely compromised or die as a result of the insult. Previous studies in this laboratory have found that mortality at 60 min of HI was 12–14%, while the mortality at 75 min of HI was 40%. (12) These data were used to select the 65-min timeframe. The data from the sham-preconditioned HI mice (group SH) were consistent with previous experience with the 129 x C57 F1 hybrids. The mortality for this group was 20%. Among the survivors 87% had evidence of striatal injury using apomorphine challenge as a marker. This group also performed the worst of all groups during water maze testing.
The selection of the preconditioning time is based on studies which indicate that delayed preconditioning effects can be observed with exposure to isoflurane from 30 min (2) to 3 h (20–22) at concentrations of 1.5% to 1 MAC. We have determined that the MAC for isoflurane in P10 129 x C57 F1 mice is 2.26 ± 0.23. (18) Similarly, 3 h of 5.7% desflurane was reported to produce changes in proteins in the brains of adult rats, the effect was maximal at 24 h (23). Hypoxia also induces multiple changes in gene expression after a period of 3 h of 8% oxygen in neonatal rats (24). Rats are more tolerant of hypoxia than mice, thus we used 10% oxygen and limited the exposure to 2 h. The final choice of preconditioning time (2 h) and concentration (1.8% isoflurane, 10% oxygen) is within the range of values reported to produce preconditioning effects in different model systems.
The control group in this study had room air preconditioning and sham carotid ligation. The sham ligation procedure required a brief exposure to isoflurane (<6 min); thus, there was no isoflurane-naïve group. The exposure time was very short and the animals were fully recovered before returning to the dams for 2 h. We have not found any differences in adult water maze performance between animals exposed to brief periods of isoflurane as neonates (P9–P10) versus those with no history of isoflurane exposure as neonates.
Preconditioning Effects
Although prolonged exposures to isoflurane have been reported to cause neuronal injury in P7 rodents (20,21,25,26), we found no evidence that the two-hour preconditioning with isoflurane produced any neuronal injury in P9 mice. Two hours of preconditioning resulted in hypoglycemia but there was no evidence of activation of caspase-3 within 3 h of resumption of oral feeds. It is possible that the timeframe of activation of caspase-3 in. the neonatal mice was different from that in adult mice (27) after normalization of glucose after profound hypoglycemia, but this explanation is not supported by the data. The mice exposed to 6 h of isoflurane showed clear caspase-3 activation 3 h after resumption of oral feeds. Alternatively, the cause of hypoglycemia is the determinant of whether or not apoptosis follows restoration of normoglycemia. Insulin-induced hypoglycemia, but not fasting hypoglycemia, has been found to adversely affect the outcome of normothermic neonatal HI in the rat (22,28). Although not analogous, the Yager et al. study (22) suggests that the cause of hypoglycemia may be an important determinant of its effect on the system under study.
The long-term outcome data support the position that 2 h of isoflurane preconditioning had no deleterious effects on the mice. The isoflurane-preconditioned sham hypoxia (AS) mice performed as well as sham-preconditioned-sham hypoxia (SS) mice in all water maze procedures. The reduced maze is especially useful in identifying animals with subtle cognitive deficits, as the platform is one-fourth the area of the hidden maze platform and requires precise spatial localization. Given that there was no difference in performance between the AS and SS groups on the reduced maze it is highly unlikely that 2 h of isoflurane preconditioning is associated with any long-term neurological deficits.
Postinjury Effects of Preconditioning
The first observation made was a reduction in the immediate mortality after HI among the isoflurane-preconditioned mice. The isoflurane-preconditioned mice (AH) had no preweaning mortality. This effect is potentially the result of isoflurane-induced cardiac protection.
Isoflurane preconditioning did not improve cognitive function on tests of spatial learning and reference memory after neonatal HI compared to sham-preconditioned mice. In contrast, the hypoxia-preconditioned HI mice (Group HH) demonstrated significantly better cognitive performance (FCL Day 5, P = 0.038) than the sham-preconditioned HI mice (Group SH) and not significantly different (P = 1.000) from the control mice (Group SS) on the hidden maze acquisition trials.
Hypoxic preconditioning has been shown to preserve spatial navigation in adult rats subjected to HI as neonates (P7) (29). The hypoxia-preconditioned animals in the Gustavsson study exhibited swim paths during a hidden maze acquisition protocol not significantly different in length from controls and significantly shorter than non-preconditioned rats that had HI on P7. Swim paths remove the effect of swim speed and measure cognitive performance as does FCL. Thus our long-term cognitive function results in mice after hypoxia preconditioning and neonatal HI are congruent with the results of similar interventions in rats. The results also suggest that, in our model, P10 HI degrades motor function more than cognitive function after hypoxic preconditioning.
The severity of the cognitive deficits among the mice in the sham-preconditioned HI group (SH) may seem out of proportion to the histological injury, as the lesion is unilateral. Previous studies have shown that, unlike adult mice, unilateral hippocampal lesions in the neonate produce profound and lasting deficiencies in performance on spatial reference memory tasks that approximate bilateral lesions in adult mice (30–32).
The isoflurane-preconditioned mice (AH) performed better than sham-preconditioned mice (SH) during cued maze testing (P = 0.045) and had a reduced incidence of circling in response to apomorphine (P = 0.04). After neonatal HI it appears that cued maze performance is largely motor component-dependent (12). Some or most elements of the motor component of cued maze performance are represented in the striatum. This argument is supported by the observation that the cued maze FCLs (Days 5 and 6) are nearly identical for the circling and noncircling HI mice despite a statistically significant difference in latencies by repeated measures GLM. Performance on cued learning tests have been shown to be degraded by isolated electrolytic or chemically induced lesions of the caudate (33,34). Thus, we propose that the difference in cued maze performance between isoflurane- preconditioned and sham-preconditioned HI mice is likely due to differences in striatal dependent motor performance. Taken together, the evidence supports the premise that delayed isoflurane preconditioning confers protection to striatal structures in a paradigm of moderate to severe neonatal HI.
We have tested for long-term differences in motor, cued and spatial navigation functional testing and found that isoflurane preconditioning 24 h before severe HI injury on P10 improves the functional capacity of the striatum (caudate-putamen) as assessed by response to apomorphine and performance on the cued water maze. However, cognitive performance on tests of hippocampal-dependent spatial reference memory was not improved by isoflurane preconditioning but was improved by hypoxia preconditioning in this paradigm.
|
|
Footnotes |
|---|
Supported by Department of Anesthesia, Cincinnati Children's Hospital Medical Center.
|
REFERENCES |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  A. W. Loepke, F. X. McGowan Jr, and S. G. Soriano CON: The Toxic Effects of Anesthetics in the Developing Brain: The Clinical Perspective Anesth. Analg., June 1, 2008; 106(6): 1664 - 1669. [Full Text] [PDF]
|
|
|
|
|
|
V. Degos, G. Loron, J. Mantz, and P. Gressens Neuroprotective Strategies for the Neonatal Brain Anesth. Analg., June 1, 2008; 106(6): 1670 - 1680. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. W. Loepke and S. G. Soriano An Assessment of the Effects of General Anesthetics on Developing Brain Structure and Neurocognitive Function Anesth. Analg., June 1, 2008; 106(6): 1681 - 1707. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
