Carbenoxolone does not Improve Sevourane-induced Neurotoxicity in the Developing Rat Hippocampus

Background: Connexin 43 is the most abundant connexin protein expressed in astrocytes. Our previous research found that sevourane-induced neurotoxicity was related to Cx43 via JNK/MAPK/Ap-1 signaling pathway. Cx43 functions through hemichannels (HCs) and gap junctions (GJs) and both of them could affect the homeostasis of central nervous system. Carbenoxolone (Cbx) is generally considered as a Cx43 hemichannel and gap junction inhibitor in traditional applications. Therefore, we hypothesized that preadministration of Cbx may attenuate sevourane-induced cognitive dysfunction. Methods: Seven-day-old SD rats (P7) were exposed to 3% sevourane for 4 hours with or without Cbx pretreatment at a dose of 50 mg/kg. Levels of Bcl-2, Bax, Cx43 and caspase-3 positive cells in P8 rat’s hippocampus were examined using Western blotting and immunohistochemistry. Morris water maze was performed from P28 to P33 to test the cognitive function. Results: Cx43 levels and caspase-3 positive cells in P7 rat hippocampus were increased 1 day after exposure to 3% sevourane for 4 h compared with control rats. Sevourane anesthesia decreased the expression of Bcl-2 and increased the expression of Bax in P8 rat’s hippocampus (P<0.05, n=5). Exposure to sevourane led to signicant cognitive impairment from P28 to P33.All these defects could not be alleviated by pretreatment with Cbx. Conclusions: These data suggested that Cbx could not improve sevourane-induced neurotoxicity by inhibiting Cx43 HCs in the developing brain. The Morris water maze (MWM) experiment is a classical experimental method to study animal behavior that is widely used in the eld of neurobiology and can well reect the spatial learning and memory ability of animals. The results showed that there was no signicant difference in the performance of the four groups in the rst three days. On trial day 5, when the rats were 32 days old, the time to nd the platform in the Sevo group was signicantly longer than that in the Ctrl group (P < 0.05, n = 8). The time of nding the platform in the Sevo + Cbx group was signicantly longer than that in the Ctrl group on trial days 4 and 5 (P < 0.05, P < 0.01, n = 8, respectively), indicating that inhalation of 3% sevourane for 4 hours had a signicant effect on the long-term spatial learning and memory ability of rats. However, i.c.v. injection of the Cx43 HC inhibitor Cbx before sevourane exposure did not improve the long-term learning and memory ability of rats. Compared with the Ctrl group, there was no signicant difference in nding the submerged platform in the Ctrl + Cbx group. Similarly, when the rats were 33 days old, the number of platform traversions in the Sevo group and Sevo+Cbx group were less than that in the Ctrl group, and the spatial memory ability signicantly decreased (P < 0.05, n=8). The number of rats crossing the platform in the Ctrl + Cbx group was similar to that in the Ctrl group (P > 0.05, n = 8). Additionally, there were no signicant differences in the average swimming speeds among the four groups (results not shown). These data suggested that sevourane could affect hippocampal function in adolescent P7 SD rats and that 50 mg/kg Cbx i.c.v. could not attenuate


Introduction
Beginning with William T. G. Morton's rst successful public demonstration of general anesthesia with ether on October 16, 1846, volatile anesthetics have been widely used in clinical practice. Sevo urane is the most popular anesthetic in pediatric and geriatric anesthesia due to its unique advantages. Unfortunately, increasing evidence suggests that sevo urane exposure increases the risk of developing a learning disability, especially in the developing brain [1,2].
Previously, the main function of glial cells was to support the central nervous system and isolate neurons.
However, in recent years, it has been found that glial cells, also play roles in synaptic transmission, buffering and releasing neurotransmitters, nutrition and the removal of harmful substances, immune regulation and other functions. As the most abundant glial cells, the role of astrocytes in the central nervous system has attracted increasing attention.
Connexins are transmembrane proteins that are part of a multigenic family present in almost every cell. A total of 20 and 21 connexins have been identi ed in mouse and human genomes respectively, and each connexin is named according to its respective molecular weight. Connexins are synthesized in the endoplasmic reticulum and then tra c to the plasma membrane through the Golgi apparatus. Connexin 43 (Cx43) is one of the most abundant connexin proteins in astrocytes [3]. Six connexins can form a connexon, also known as hemichannel (HC), and two connexons or hemi-channels dock to form a gap junction (GJ) [4][5][6]. Many studies have shown that Cx43 is related to the development of the central nervous system, neuronal excitability, synaptic transmission, synaptic plasticity, learning, memory, and in ammation of the central nervous system. Our previous study found that sevo urane-induced neurotoxicity in the developing brain was associated with the upregulation of Cx43, and the JNK/c-Jun/AP-1 signaling pathway was involved [7]. A single Cx43 protein does not play a role, but after oligomerization into HCs or GJs, Cx43 proteins could participate in many important roles.
GJs allow adjacent cells to exchange small molecules directly (<1.5 kDa), including ions, ATP, glutamate, Additionally, GJs can buffer toxic substances when a small amount of damage occurs. When sustained damage occurs, the toxic substances will spread distance, causing more cell damage.
Sevo urane-induced neurotoxicity in the developing brain could increase the expression of the Cx43 protein, but the exact reason is still unclear. We have reason to believe that the opening of HCs can cause direct intracellular and extracellular communication and damage the CNS and that GJs can mitigate the damage from intracellular toxic substances between adjacent cells by diffusion, thus spreading harmful substances to distant places. Therefore, we hypothesize that sevo urane-induced CNS damage is associated with the roles of HCs and GJs. We observed whether sevo urane-induced neurotoxicity in the developing rat hippocampus could be reduced by intracerebroventricular injection of carbenoxolone (Cbx), a general HC and GJ inhibitor, to improve learning and memory ability in adolescence.

Materials And Methods
Our study was carried out in compliance with the ARRIVE guidelines 2.0 guidelines.

Experimental animals
All animal experimental procedures and protocols were approved by The Laboratory Animal Care Committee of China Medical University, Shenyang, China (NO. 2017PS019K) and were performed in accordance with the Guidelines for the Care and Use of Laboratory Animals from the National Institutes of Health, United States. Twenty pregnant Sprague -Dawley (SD) rats that were in their second trimester, weighing 250-300 g were purchased from Changsheng Shengwu (China). The pregnant rats were housed under a 14:10 constant light -dark cycle with free access to water and food for one week at room temperature (24 ± 1 °C). The day of birth was noted as postnatal day 0 (P0). Postnatal day 7 (P7) male or female rat pups weighing 14-18 g, were used in this study.
Sevo urane exposure and drug administration Seventy-two P7 pups from different litters were randomly allocated to four groups (18 pups were in each group): control (Ctrl), control+Cbx (Ctrl+Cbx), sevo urane (Sevo) and sevo urane+Cbx (Sevo+Cbx). Two hours before sevo urane exposure, Ctrl+Cbx and Sevo+Cbx groups were injected Cx43 inhibitor Cbx intracerebroventricularly(i.c.v.) 50 mg/Kg. The Ctrl and Sevo groups were injected with the same amount of 0.9% saline intracerebroventricularly. All seventy-two P7 rat pups were separated from their mothers and placed in a glass chamber (40×24×20 cm) resting in a water bath to maintain a constant environmental temperature of 38°C. In the chamber, the rats were exposed to either 3% sevo urane in 30% oxygen carrier gas (balanced with nitrogen) or a carrier gas without sevo urane for 4 hours. All method descriptions are similar to our previously published study [7]. After sevo urane exposure, all P7 pups remained in the chamber ventilated with the mixed gas (30% O 2 /70% N 2 ) at a rate of 2 L/min until they were fully awakened. Therefore, the rats were transferred to their original cages.

Western blots
Five pups from each group were randomly selected and killed by decapitation 1 d after sevo urane exposure and then the hippocampus was harvested and the protein was extracted (n = 5/group). All procedures are similar to our previously published study [7,12,13]. Fifty micrograms of each protein sample was separated by 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE); using a semidry blotting apparatus (Bio-Rad Laboratories, Munich, Germany), the proteins were transferred to a polyvinylidene uoride (PVDF) membrane (Millipore, Bedford, MA, United States) and then incubated overnight at 4 °C with the appropriate primary antibodies: anti-Cx43 (1:1000; SAB4501175), anti-Bcl-2 (1:1000; 4695), anti-Bax (1:1000; 9252), and anti-GAPDH (1:1,000; A9169). Next, the membranes were washed in TBST and incubated with secondary antibodies (A9169, Sigma, United States) for 2 h at room temperature. The positive reactive bands were detected by Amersham enhanced chemiluminescence (ECL) reagents. The blots were scanned using an Amersham Image 600 scanner (GE Healthcare Life Sciences), and the protein band density was quanti ed using ImageJ software. Protein expression levels were evaluated by the GAPDH ratio.

Immunohistochemistry staining (IHC)
Neuronal apoptosis was detected by immunohistochemistry (IHC) staining of caspase-3 positive cells in the hippocampus. Five rats in each group were euthanized by transcardial perfusion with saline, followed immediately by 4% paraformaldehyde 1 day after sevo urane exposure. The hippocampus of the rats was cut from the whole brain at a distance of 2 mm from the forehead, after which it was dehydrated in graded ethanol and was embedded in para n. Brain coronal slices of the para n-embedded tissue (2.5 um) were sectioned using a microtome. These tissue sections were then baked, depara nized, rehydrated, and quenched of endogenous peroxides. It was the same way that we've studied before [7].
Each slice was incubated with activated caspase-3 antibody (1:200 dilution, catalog no. 9662; Cell Signaling Technology) overnight at 4°C. The next day, the sections were exposed to peroxidaseconjugated goat antirabbit secondary antibody (Santa Cruz Biotechnology). The hippocampal CA1 region, was colorized with diaminobenzidine solution for 8 min and counterstained with hematoxylin. Finally, the sections were photographed by Nikon C1 microscope with 3 randomly chosen elds imaged per slide by an investigator who was blinded to the experimental intervention (n = 5/group). Caspase-3 positive cell accumulated optical density was calculated in each hippocampal slide vision.

Morris water maze (MWM)
Morris water maze experiment was conducted to evaluate spatial learning and memory ability of the rats. Thirty-two rats from four groups including Ctrl, Ctrl + Cbx, Sevo and Sevo + Cbx, were subjected to the MWM after 28 d (n = 8/group), as previously described [7]. Escape latency trials were conducted once per day for ve consecutive days. The time spent nding the hidden platform and the swimming distance before reaching the platform were recorded. If the rat did not nd the platform within 90 s, escape latency was recorded as 90 s. Then the rats were guided to stand on the concealed platform for 10 s. The spatial probe test was conducted on day 6 to test the ability of the rats to maintain their memory of the platform space after learning to nd the platform. The platform was removed after the positioning navigation experiment was completed. After the platform was taken away from the maze, the rats were put into the quadrant opposite the platform and permitted to swim freely for 90 s. The time in the target quadrant was noted. The entire behavioral test was recorded and analyzed using a Noldus Ethovision XT video analysis system (Netherland). After the MWM test, all 32 rats were sacri ced without biochemical analysis.

Statistical analysis
The statistical analysis was conducted using Graph Pad Prism 7.0 software (version 7.0; Graphpad Software, Inc.). Statistical signi cance was determined by One-way ANOVA followed by Tukey's test.
Interaction between time and group factors in a two-way ANOVA with repeated measurements was used to analyze the difference of learning curves (based on escape latency) between rats in the MWM [7]. At least ve individual trials were performed for each experiment and data represented as mean ± SEM.
Differences were considered statistically signi cant for P < 0.05.

Results
Cbx did not affect the expression of Cx43 protein in hippocampus of P7 rats after exposure to 3% sevo urane for 4 hours.
We observed the expression of the Cx43 protein in the hippocampus of P7 rats exposed to 3% sevo urane 1 day after Cbx (50 mg/kg) was injected i.c.v. Western blot experiments showed that Ctrl + Cbx had no effect on Cx43 protein expression in the hippocampus of rats after exposure to sevo urane for 1 day (n = 5, P > 0.05). Cx43 protein expression was signi cantly increased in the hippocampus of P7 rats 1 day after sevo urane compared with that in the Ctrl group (n=5, P < 0.05), but there was no signi cant difference between the Sevo and Sevo + Cbx group rats regarding the expression of Cx43 (n=5, P > 0.05). These results suggest that 50 mg/kg Cbx i.c.v. had no signi cant inhibitory effect on Cx43 protein expression in the hippocampus of P7 rats 1 day after inhalation of sevo urane (Figure1).
Cbx did not affect the expression of Bcl-2 and Bax in hippocampus of P7 rats exposed to 3% sevo urane for 4 hours.
Apoptosis is the main pathway of programmed cell death and can be triggered by multiple death stimuli.
The triggers of apoptotic pathways include the mitochondrial-mediated (intrinsic) pathway and the cell surface death receptor (extrinsic) pathway. The intrinsic apoptotic pathway is mediated by Bcl-2 family members and the mitochondrial permeability transition pore (PT pore). Bcl-2 is a family of proteins with proapoptotic (e.g., Bax) or antiapoptotic (e.g., Bcl-2) properties. Western blot experiments showed that the expression of the Bax protein in the hippocampus of rats increased signi cantly, while Bcl-2 decreased 1 day after inhalation of sevo urane compared with the Ctrl group (n=5, P < 0.05). There were no differences in Bax and Bcl-2 expression in the hippocampi of the Cbx group, and the Sevo + Cbx group did not exhibit changes in Bax and Bcl-2 expression compared with the Sevo group (n = 5, P > 0.05). It was concluded that 50 mg/kg Cbx injected i.c.v. did not decrease the expression of apoptosis-related proteins in the hippocampus of P7 rats after exposure to 3% sevo urane for 4 hours (Figure 2).
Cbx did not decrease the number of cleaved caspase-3 positive cells in the hippocampus of rats exposed to 3% sevo urane for 4 hours.
Cleaved caspase-3-positive cells are the ultimate indicator of apoptosis. In our study, the cumulative optical density of cleaved caspase-3-positive cells in the hippocampus of P7 rats was observed by IHC after inhalation of 3% sevo urane for 4 hours. The results showed that i.c.v. injection of Cbx alone did not affect the number of positive caspase-3 cells in the hippocampus (n = 5, P > 0.05). However, the cumulative optical density of cleaved caspase-3-positive cells in the hippocampus of the Sevo group increased signi cantly 1 day after exposure to sevo urane (n=5, P<0.05). Compared with the Sevo group, the cumulative optical density of cleaved caspase-3-positive cells in the hippocampus of the Sevo+Cbx group did not decrease signi cantly 1 day after exposure to sevo urane, suggesting that Cbx did not decrease the apoptosis of neurons in the hippocampus of P7 rats after exposure to sevo urane ( Figure   3). Cbx did not improve learning and memory impairment in rats after inhalation of 3% sevo urane for 4 hours.
The Morris water maze (MWM) experiment is a classical experimental method to study animal behavior that is widely used in the eld of neurobiology and can well re ect the spatial learning and memory ability of animals. The results showed that there was no signi cant difference in the performance of the four groups in the rst three days. On trial day 5, when the rats were 32 days old, the time to nd the platform in the Sevo group was signi cantly longer than that in the Ctrl group (P < 0.05, n = 8). The time of nding the platform in the Sevo + Cbx group was signi cantly longer than that in the Ctrl group on trial days 4 and 5 (P < 0.05, P < 0.01, n = 8, respectively), indicating that inhalation of 3% sevo urane for 4 hours had a signi cant effect on the long-term spatial learning and memory ability of rats. However, i.c.v. injection of the Cx43 HC inhibitor Cbx before sevo urane exposure did not improve the long-term learning and memory ability of rats. Compared with the Ctrl group, there was no signi cant difference in nding the submerged platform in the Ctrl + Cbx group. Similarly, when the rats were 33 days old, the number of platform traversions in the Sevo group and Sevo+Cbx group were less than that in the Ctrl group, and the spatial memory ability signi cantly decreased (P < 0.05, n=8). The number of rats crossing the platform in the Ctrl + Cbx group was similar to that in the Ctrl group (P > 0.05, n = 8). Additionally, there were no signi cant differences in the average swimming speeds among the four groups (results not shown). These data suggested that sevo urane could affect hippocampal function in adolescent P7 SD rats and that 50 mg/kg Cbx i.c.v. could not attenuate sevo urane anesthesia-induced cognitive impairment in young SD rats (Figure 4).

Discussion
The current study demonstrated that anesthesia with 3% sevo urane for 4 hours caused signi cant upregulation of Cx43 and Bax, decreased the level of the Bcl-2 protein and increased the expression of cleaved caspase-3 in the hippocampus after 1 day. Cbx [14], injected i.c.v. 2 hours before sevo urane exposure could not decrease the upregulation of Cx43, Bax and cleaved caspase-3 or increase Bcl-2 expression in the hippocampus. Cbx also could not improve rat cognitive impairment after 30 days, and the role of Cx43 HCs and GJs in sevo urane-induced development of neurotoxicity needs further investigation.
Cx43 protein HCs exhibit a low open probability under normal conditions to maintain normal synaptic function. HCs could be deleterious by releasing excitotoxins (e.g., ATP and glutamate), disturbing the concentration of intracellular Ca 2+ or altering cytoplasmic ionic and osmotic balance. When the CNS is stimulated by various in ammatory mediators (cytokines, NO, ROS, etc.), it rst enhances the HCs activity of astrocytes and microglia. For example, microglia stimulated by detrimental agents will elevate the release of TNF-α and IL-1β via microglial Cx32 hemichannels, leading to an increase in astrocyte Cx43 protein HCs activity. Then, astrocyte Cx43 HCs opening mediated by microglia could increase Ca 2+ in ux and release glutamate, thus disrupting normal synaptic function.
Cx43 HCs opening under pathological conditions could release many glial transmitters, such as ATP and glutamate. When the extracellular Ca 2+ concentration decreases or the synaptic glutamate concentration increases, Cx43 HCs open and ATP is released into the synaptic space. ATP activates the P2Y1 purine receptor of inhibitory intermediate neurons [15], thereby enhancing synaptic inhibition. At the same time, the Cx43 HCs-dependent release of ATP can also enhance the synaptic transmission of glutamatergic neurons, but this mechanism is still unclear. It has been speculated that ATP activates the AMPA receptor of the postsynaptic membrane through the P2X7 receptor, which increases excitotoxicity. D-serine is also released in a HC-dependent form intracellularly to extracellularly to enhance synaptic transmission and synaptic plasticity. D-serine is a coagonist of the NMDA receptor in the postsynaptic membrane of neurons. Studies have shown that D-serine could act on glutamatergic receptors in the postsynaptic membrane and play a synaptic role [16]. Data have also shown that Cx43 HCs of hippocampal astrocytes also control long-term potentiation (LTP) by releasing D-serine. These results indicate that Cx43 HCs play an important role in gliotransmitter release in the synaptic space [17].
Similarly, Cx43 GJs play an important role in the CNS [18][19][20]. GJs in astrocytes contribute to the formation of functional syncytial bodies, promote the removal of ions and neurotransmitters during neuronal activity, and allow the transmission of Ca 2+ waves between astrocytes. Neurotransmitters released by neurons can bind to receptors in astrocyte membranes and produce metabolites such as inositol 1,4,5 triphosphate (IP3), ATP, glutamate and D-serine, which diffuse through GJs and affect nerve function. This intercellular communication based on Cx43 GJs enables astrocytes to perceive and integrate local and global synaptic activity and respond to glial transmitters affecting presynaptic and postsynaptic transmission. Therefore, Cx43 GJs and gliotransmitters are important participants in the regulation of the nerve activity, behavior and homeostasis of brain function.
An increase or decrease in Cx43 expression does not directly indicate whether its role is protective or damaging in the CNS, because Cx43 plays a role in the formation of HCs and GJs [21]. It is generally believed that under normal conditions, only a small number of Cx43 HCs are open in order to play some basic roles. HCs will open in large quantities enabling intracellular and extracellular communication directly when the CNS suffers various stimuli, such as ischemia, hypoxia, trauma and growth factors.
Harmful toxic substances should be released from the cells to the extracellular environment, which produces adverse effects [22]. GJs have been shown to be both neuroprotective and neurodestructive in ischemic situations. When damage is limited to a small extent, GJs can buffer the toxic substances between adjacent cells and allow the cells to survive. When the damage increases, GJs can spread toxic substances a longer distance, causing more cell death. Many diseases or pathophysiological conditions seem to be associated with Cx43 HCs and GJs which have become new pharmacological targets in recent years [23].
To date, there have been many reagents and drugs that inhibit Cx43 HCs and GJs, including Cbx, quinine, me oquine, connexin mimetic peptides GAP26, and GAP27 and so on [24,25]. Although the structures of Cx43 HCs and GJs have been studied for a long time, there is still no speci c inhibitors within Cx43 HC and GJ. Cbx, a semisynthetic derivative of glycyrrhetinic acid (GA), is widely used as a Cx43 HCs and GJs blocker in various pathological processes of the brain. It has been con rmed that Cbx displays different protective effects in a model of Parkinson's disease [26], ischemic brain injury [27], posttraumatic epilepsy [28], allergic airway in ammation [29] and fatty liver disease in obsess mice [30]. Also, Cbx did not induce signi cant side effects even at the highest dose tested (50 mg/Kg) [31]. Unfortunately, little is known about the clear mechanism of Cbx on connexin HCs and GJs [32]. Meanwhile, Cbx was previously shown to leave Cx43 expression unaffected [33]. Similar ndings were reported in our study.
Why Cbx failed to attenuate sevo urane-induced neurotoxicity in developing hippocampus in our study. First, Cbx, a more water-soluble derivate of GA, has different cellular targets and also acts as an inhibitor of 11β-hydroxysteroid dehydrogenase [34]. Second, in addition to blocking Cx43 HCs and GJs, Cbx can block many other channels and receptors including voltage-gated Ca2+ currents, homomeric Panx1 and heteromeric Panx1/Panx2 channels, P2X7 receptors, Cx26 and Cx38 HCs [35][36][37]. Third, Cbx was widely used as a potent blocker of the major astroglia-to-astroglia and astroglia-to-neuron HCs and GJs in previous studies [38]. Future study is need to determine which type of glial cell (microglia or astrocyte is most affected by Cbx. Last but not least, because there are no speci c inhibitors for HCs and GJs, it is necessary to develop Cx43 HCs and GJs speci c inhibitors to clarify the role of Cx43 HCs and GJs in sevo urane-induced CNS toxicity in the future.

Conclusions
Overall, for the rst time, we demonstrated that 50 mg/kg Cbx i.c.v. 2 hours before 3% sevo urane exposure could not decrease Cx43 upregulation, or the number of cleaved caspase-3 positive cells in the hippocampus of developing rats and also could not improve learning and memory impairment in adolescence. Both HC opening and GJ communication could affect normal CNS function, but which of them plays a pivotal role in sevo urane-induced neurotoxicity in the developing rat hippocampus still needs further study.

Author Contributions
Congjie Bi has contributed to conception, design, acquisition, analysis and drafted the manuscript.
Congjie Bi, Dan Xie, Donghai Yu have contributed to date collection and analysis. Yang Wang has contributed editing and interpretation. Figure 1 The effect of Cbx i.c.v. 2 hours before sevo urane exposure on the expression of the Cx43 protein in the hippocampus of P7 rats. (A) Western blot results showed that Cbx i.c.v. 50 mg/Kg alone had no effect on the expression of Cx43 protein in the hippocampus of P7 rats 1 day after sevo urane exposure compared with the Ctrl group. Sevo urane could increase the expression of the Cx43 in the hippocampus. Cbx i.c.v. before sevo urane did not reduce the up-regulation of Cx43 protein 1 day after sevo urane exposure. (B)

Funding
Quanti cation of the Western blotting results showed the differences of Cx43 between the Ctrl, Ctrl+Cbx, Sevo and Sevo+Cbx group 1 day after sevo urane exposure. N=5 in each group. *P<0.05 compared with the Ctrl group.

Figure 2
The effect of Cbx i.c.v. 2 hours before sevo urane exposure on the expression of Bcl-2 and Bax protein in the hippocampus of P7 rats. (A) Western blot results showed that Cbx i.c.v. 50 mg/Kg alone had no effect on the expression of the Bcl-2 and Bax protein in P7 rat hippocampus 1 day after sevo urane exposure compared with the Ctrl group. Sevo urane could decrease the expression of Bcl-2 and increase the expression of Bax after sevo urane in hippocampus. Cbx i.c.v. before sevo urane did not reduce the down-regulation of Bcl-2 protein or up-regulation of Bax protein 1 day after sevo urane exposure. (B) Quanti cation of the Western blotting results showed the differences of Bcl-2 and Bax between the Ctrl, Ctrl+Cbx, Sevo and Sevo+Cbx group 1 day after sevo urane exposure. N=5 in each group. *P<0.05 and **P<0.01 compared with the Ctrl group.  Cbx treatment did not mitigate the sevo urane-induced cognitive impairment after anesthesia with 3% sevo urane for 4 h in young rats. Cognitive behavior was tested with the MWM from day 28 to day 33. (A) There were no obvious differences among the four groups from day 1 to day 3. The rats in the Sevo group needed signi cantly more time to nd the submerged platform on trial day 5 than did the rats in the Ctrl group in the escape latency portion of the MWM. The rats in the Sevo+Cbx group also needed more time on trial day 4 and 5 in MWM compared with Ctrl group. (B) The rats in the Sevo and Sevo+Cbx groups crossed the platform fewer times than did the rats in the Ctrl group. N=8 in each group. *P<0.05 compared with the Ctrl group.