Maternal Exposure to Sevourane Disrupts Oligodendrocyte Myelination of the Postnatal Hippocampus and Induces Cognitive and Motor Impairments in Offspring

Maternal exposure to sevourane can impose signicant neurocognitive risks on the developing brain of infants. Several studies have indicated that oligodendrocytes may be involved in sevourane-induced neurotoxicity, but the concrete effects of sevourane on the development and myelination of oligodendrocytes remain unclear. In this study, we assessed fetal myelination and neural behavior after maternal exposure to sevourane. Pregnant C57BL/6J mice (gestational day 15.5) were exposed to sevourane (2.5%) for 6 h. The cognitive function and motor coordination of offspring (8 weeks of age) were determined via the novel object recognition test, the Morris water maze test and the accelerating rotarod test. Proliferation and differentiation of cultured oligodendrocyte precursor cells (OPCs) were detected via immunocytochemistry. Expression and ultrastructure of myelin in the fetal hippocampus were analyzed using immunohistochemistry and transmission electron microscopy (TEM). Myelin-associated genes and proteins were tested via qRT-PCR, immunouorescence and western blotting. The functionality of myelin was evaluated by electrophysiology. The results showed that maternal exposure to sevourane induced cognitive and motor impairments in infants, accompanied by inhibitions of OPC proliferation and differentiation, and damages of myelin structure. Myelin-associated genes and proteins (including MBP, Olig1, PDNFRα, Sox10, etc.) were downregulated. The conduction velocity of axons also declined. These results suggested that maternal exposure to sevourane could induce detrimental effects on cognitive and motor functions in offspring, which might be associated with disrupted myelination of oligodendrocytes in the hippocampus.


Introduction
Advances in surgical technologies have led to a substantial increase in the number of pregnant women undergoing general anesthesia ). Subsequently, there has been increasing concerns regarding the safety of maternal anesthesia for offspring (Chai et  Sevo urane is one of the most commonly used inhalation anesthetic agents in obstetric and pediatric surgeries (Lee et al. 2017). Due to its lipophilicity, sevo urane can easily cross the placenta and bloodbrain barrier, and affect the development of the brain of infants (Fang et al. 2017). Several mechanisms have been proposed to be involved in sevo urane-induced neurotoxicity, including apoptosis, in ammation, disruption of synaptogenesis and so on (Neag et al. 2020). However, most of these mechanisms are focused on neurons, not glial cells (Zheng et al. 2013). Given the integrity of the central nervous system (CNS) and the crucial role of oligodendrocytes in the development and functional maintenance of neurons (Xin et al. 2020), the potential effects of sevo urane exposure on oligodendrocytes should not be neglected.
In the developing CNS, oligodendrocytes extend membrane processes by ensheathing neuronal axons with lipid-rich myelin membranes. Myelination enables axons to transmit information more rapidly, which facilitates the evolution of a complex yet compact CNS in vertebrate animals (Elbaz et al. 2019). In addition, myelin plays a critical role in proper neuronal function by providing trophic and metabolic support to axons and facilitating energy-e cient saltatory conduction (Ishii et al. 2019). Impairment of oligodendrocyte myelination has been shown to be associated with many neurological diseases, including autism, Alzheimer's disease and depression (Lu et al. 2016). However, whether exposure to sevo urane during the gestational period in uences the myelination and development of oligodendrocytes is still elusive.
In this study, we evaluated the effects of maternal exposure to sevo urane on myelination in the hippocampus and the long-term cognition and motor performance of postnatal mice. Our results demonstrated that sevo urane exposure during the prenatal period induced impairments in the structure of myelin, as well as de cits in long-term memory and ne motion. In addition, we observed inhibitions of proliferation and differentiation of oligodendrocyte precursor cells (OPCs), and reduction of the expression of RNAs and proteins involved in the regulation of oligodendrocyte development. The function of axonal conduction velocity was also attenuated. Altogether, these results suggested that early gestational exposure to sevo urane could induce detrimental effects on cognitive and motor functions of postnatal mice, which might be associated with disrupted oligodendrocyte myelination in the hippocampus.

Animals and Experimental Design
All procedures of the study were approved by the Animal Care and Use Committee of the Fourth Military Medical University (Xi'an, China) and followed institutional guidelines. C57BL/6J male and female mice (8-week-old) were provided by the Animal Centre of the Fourth Military Medical University. Pairs of female mice mated with one male, and pregnant mice were identi ed and placed into another cage. All mice were housed and allowed free access to a standard animal diet and tap water. Room temperature was maintained at 20~23 °C with a 12 h/12 h light/dark cycle.
A total of 108 offspring mice (6 per group in functional tests and 4 per group in morphological tests) were used in this study on the basis of scienti c literatures and our pre-test results. Brie y, morphological changes of oligodendrocytes were detected via immuno uorescence, western blot, PCR and electron microscope at PND14, PND30 and PND60 after maternal exposure to sevo urane, while functional changes were detected via electrophysiological and behavioral tests at PND30 and PND60 (Fig. 1a). All experiments were carried out in triplicate, and randomization and double-blinding were conducted to minimize subjective bias in the design.

Sevo urane Exposure
Pregnant C57BL/6J mice were randomly assigned to the control group or the sevo urane-treated group on gestational day 15.5 (G15.5). Mice in the sevo urane-treated group received 2.5% sevo urane in 97.5% oxygen for 6 h in an anesthetizing box, while mice in the control group received 100% oxygen for 6 h. The size of the anesthetizing box was 15 × 15 × 35 cm 3 . The gas ow rate was 2 L/min for induction and 1 L/min for maintenance. The concentrations of sevo urane and oxygen were continuously monitored with a gas analyzer (Drager, Germany). A warming blanket was used during anesthesia to prevent hypothermia.
Cultured OPCs were treated with sevo urane through a vaporizer (Abbott, USA). Cells were placed in an airtight incubation chamber (Billups-Rothenberg, USA) at 37 °C and subsequently perfused with air (21% O2, 5% CO2, 69% N2) containing 4.1% sevo urane for 6 h. The gas concentrations of O2, CO2 and sevo urane were continuously monitored by an anesthetic gas measurer module (Datex Ohmeda, Spain). The cells in the control group were perfused for the same time with fresh air. Once the exposure was nished, the cells were returned to the incubator.

Immunocytochemistry and BrdU Incorporation
For immunocytochemistry, cells were xed with 4% paraformaldehyde for 10 min. For BrdU incorporation analysis, cells were treated with 2 N HCl after xation for 10 min at 37 °C to denature DNA, followed by neutralization with borate buffer (0.1 M, pH 8.5) for 10 min at room temperature. After blocking with 3% BSA and 0.3% Triton X-100 in PBS for 30 min, cells were incubated with the following primary antibodies AB_2762826) secondary antibodies for 1 h at room temperature. Nuclei were stained with DAPI (4',6diamidino-2-phenylindole, 1:1000, Sigma-Aldrich, USA). A confocal system (Olympus Fluoview Ver4.2b, Japan) was used for image acquisition. Brie y, slides were scanned under a laser confocal microscope with wavelengths at 405 nm, 488 nm and 543 nm. The parameters were setup as follows: Object lens (20 magni cation); lter mode (Kalman and line 2); Sequential (Line); Pixel (1024by*1024by). All images were captured in a dark room at a temperature of 25℃. Image J software was used for image analysis.

Immunohistochemistry
Animals were deeply anesthetized with pentobarbital and transcardially perfused with PBS followed by 4% paraformaldehyde. Then, the brains were collected and post xed for 2 h. After transfer to a gradient of sucrose (20% and 30% in PBS), 30 μm thin serial coronal sections encompassing the entire hippocampus were collected using a freezing microtome (Leica, Germany). The sections were kept in citrate buffer at 86 °C for 15 min for antigen retrieval, followed by blocking with 3% BSA and 0.3% Triton X-100 for 1 h. Next, the sections were incubated with the following primary antibodies overnight at 4 °C: RRID: AB_2762826) secondary antibodies for 2 h at room temperature. Nuclei were stained with DAPI (1:1000, Sigma-Aldrich), and uorescence images were captured using a confocal system (Olympus Fluoview Ver4.2b, Japan) as mentioned above.

Quantitative real-time PCR
Total RNA from hippocampal tissue was extracted using TRIzol reagent (15596026, TSF). The quality and quantity of RNA were detected using a NanoDrop spectrophotometer (ND-NDL-2YRW-CCC, TSF). Equal amounts of RNA were reverse transcribed to cDNA by using a SuperScript rst-strand cDNA synthesis kit (18080051, TSF) with Oligo-dT. qRT-PCR was performed with SYBR Green Master Mix (a46109, TSF) using the ABI Prism 7900 Sequence Detector System (PE Applied Biosystems, USA). Expression of GAPDH was served as control to normalize values. Relative RNA expression was calculated using the 2 −ΔΔCt method.

Western blotting
Tissues were lysed in lysis buffer (89901, TSF) containing inhibitors of protease and phosphatase (78442, TSF). Protein concentrations were estimated using a bicinchoninic acid (BCA) protein assay kit (23227, TSF). An equivalent amount of protein (30 μg) from each sample was resolved on SDSpolyacrylamide gels and then transferred to PVDF membranes (88520, TSF). Next, the membranes were blocked in a 5% skimmed milk solution for 1 h at room temperature, followed by incubation overnight at 4°C with the following rabbit primary antibodies: β-tubulin (1:1000, Abcam Cat# ab179513) ( TFS Cat# PA5-78079, RRID: AB_2736230). After 3 washes with TBS containing 0.1% Tween-20, the membranes were incubated at room temperature for 2 h with a horseradish peroxidase-conjugated goat anti-rabbit secondary antibody (1:5000, Abcam Cat# ab150077, RRID: AB_2630356). The bands of protein on the membranes were tested using a chemiluminescent substrate (1812401, Millipore, USA). The optical density of the protein bands was measured by ImageJ software (NIH, USA).

Transmission electron microscopy
Mice were sacri ced with an overdose of pentobarbital and perfused transcardially with xative solution (2% paraformaldehyde, 2.5% glutaraldehyde in 0.1 M phosphate buffer, pH 7.4). Brains were quickly removed from the skull, and 3 mm-thick slabs containing the whole hippocampus were cut. The hippocampal tissues were immersed in the same xative solution overnight at 4 °C. After rinsing in phosphate buffer, the tissues were post xed in 1% osmium tetroxide (419494, Sigma-Aldrich) for 1 h and dehydrated in a series of graded acetone solutions. The specimens were embedded and cut with an ultramicrotome (DuPont-Sorvall, USA), stained with uranyl acetate and lead salts, and then observed under a transmission electron microscope (JEOL, Japan). Myelin in the CA1 region was assessed at 10000 magni cation. The structure of myelin in the CA1 region was analyzed based on at least 20 images per animal that contained more than 200 axons. The g-ratio (inner axonal diameter to total outer diameter including myelin) was measured on the same myelin axonal structures as previously described (Chomiak et al. 2009).

Compound action potential (CAP) recording
The optic nerves of mice were dissected and incubated with oxygenated recording solution (in mM) at room temperature as follows: 125 NaCl, 2.5 KCl, 1 MgCl 2 , 2 CaCl 2 , 10 D-glucose, 1.25 NaH 2 PO 4 , and 25 NaHCO 3 . After transfer to the recording chamber and visualization under a light microscope (Nikon, Japan), the ends of the optic nerves were suctioned into separate re-polished borosilicate glass suction electrodes using 2 ml precision syringes (Gilmont, USA). A grid with 1 mm spacing was used to measure the nerve length. Constant current stimulation (50 s) was delivered to one end of the nerve using a stimulus isolator (ISO-Flex, Israel). Electrical signals were acquired using a differential AC ampli er (model 1700, AM Systems) and digitized using a digitizer (Axon Digidata 1440A, Molecular Devices). The recording chamber was constantly perfused with oxygenated recording solution delivered by gravity and removed via a peristaltic pump. Recordings of CAPs were taken by stimulating the nerve at one end and measuring the response at the other end. Conduction velocity was calculated by dividing the length of the nerve by the latency between the start of the stimulus artifact and the peak of the CAP. The normalized half-width was determined by measuring the latency between the half-amplitude preceding and halfamplitude following the peak of the CAP and dividing it by the length of the nerve.

Behavioral analysis
Behavioral experiments were performed at PND60. To allow habituation and reduce stress, mice were moved to the experimental room 48 h before the start of the experiment (the brightness of the experimental room was 70 lux). Each experimental apparatus was cleaned with 75% ethanol after being exposed to a mouse to remove odor cues.
Novel object recognition (NOR). Two objects, which were different in shape and color but similar in size, were placed in an activity chamber. Each mouse was allowed to explore the chamber and objects for 10 min of training. Then, the mouse was moved to its home cage, and the chamber and objects were cleaned with ethanol to remove odor cues. One hour later, the mouse was allowed to explore the chamber and objects again for 5 min, with a novel object replacing one of the objects used in the training session. Discrimination scores were calculated by subtracting the number of nose pokes of the familiar object from the number of nose pokes of the novel object and dividing the difference by the total number of nose pokes of both objects.
Morris water maze (MWM). The water maze (diameter: 150 cm; height: 60 cm) was located in an isolated room and surrounded by a black curtain with four quadrants. Water (25 °C) was lled to the level of 1 cm over the platform. A video recording device connected to a computer with Any-Maze tracking software (Stoelting, USA) was used to track the movement of mice during swimming. For training, each mouse was given 60 s to locate the platform using the spatial cues in the room, after which the animal remained on the platform for 15 s (mice that could not nd the platform within 60 s were guided to the platform for learning). Mice were trained by performing four trials daily for 4 days. On the fth day, mice were allowed to swim freely in the maze, and the total time they spent in the platform quadrant and the crossing time were recorded.
Accelerating rotarod (AR). Motor function was tested using an accelerating rotarod (4-40 rpm, in 5 min; model 7650, Ugo Basile Biological Research Apparatus, Italy). Mice performed two trials per day with a 45~60-min intertrial interval for 5 consecutive days (at the same hour every day). For each day, the average time spent on the rotarod, or the time the mouse successfully made 3 consecutive wrapping/passive rotations (latency in seconds), was calculated. The maximum duration of a trial was 5 min.

Statistical analysis
GraphPad Prism 7.00 (GraphPad Software, USA) was used for statistical analysis. Analyses were performed in a manner where the person conducting the analyses was blinded to treatment assignments in all experiments. All data are expressed as the mean ± standard deviation (SD). Comparisons between two groups were performed using an unpaired t-test, Comparisons between multiple groups were performed using a one-way ANOVA followed by Tukey-Kramer's post hoc test. Comparisons between multiple groups at different time point were performed using a two-way ANOVA followed by post hoc Bonferroni's test. P <0.05 was considered to be statistically signi cant.

Results
Maternal exposure to sevo urane causes cognitive and motor impairments in offspring Pregnant mice (G15.5) received 2.5% sevo urane for 6 h. The NOT test and the MWM test were conducted in the offspring at PND60 to assess their cognition, and the AR test was conducted to assess their motor coordination (Fig. 1a). In the NOR test, compared with control mice, sevo urane-treated mice spent more time exploring the familiar object than the novel object (Fig. 1b-c). In the MWM test, the escape latency to reach the platform of sevo urane-treated mice was signi cantly increased compared with that of the control mice during both training (Fig. d-e) and probe phases (Fig. f). We also observed that time spent in the target quadrant was substantially shorter in the sevo urane-treated group than that in the control group (Fig. 1g), while total distance (Fig. 1h) and average speed (Fig. 1i) were not changed between two groups. In the AR test, mice in the sevo urane-treated group showed a signi cant decrease in the latency to fall off the rotarod (Fig. 1j-k). Taken together, these results suggested that maternal exposure to sevo urane induced impairments of cognitive and motor functions in developing offspring.
Suppressed OPC proliferation and differentiation after sevo urane exposure Primary cultured OPCs were treated with 4.1% sevo urane for 6 h. BrdU incorporation and immunocytochemistry were performed to assess proliferation and differentiation of OPCs, respectively. For proliferation, the number of BrdU/Olig2 double-labeled cells was signi cantly decreased after sevo urane exposure compared with that of the control group ( Fig. 2a-b). For differentiation, the number of MBP/Olig2 double-labeled cells was also decreased (Fig. 2c-d). These results suggested that sevo urane exposure caused inhibitions of proliferation and differentiation in cultured OPCs.
Damages of myelin structure in the postnatal hippocampus after maternal exposure to sevo urane To explore the effect of prenatal sevo urane exposure on architecture of myelin in the hippocampus, we stained MBP at PND14, PND30 and PND60 after maternal exposure to sevo urane. The results showed that myelinated bers traversed throughout the stratum lacunosum-moleculare (SLM) and hilus in the hippocampus either in small groups or in individual strands (Fig. 3a). At PND14, the intensity of MBP immunoreactivity was signi cantly decreased in the CA2/CA3 regions after sevo urane treatment. The intensity of MBP immunoreactivity showed decreasing trends in the CA1 and DG regions, but there were no signi cant differences (Fig. 3a-b). At PND30 and PND60, the intensity of MBP immunoreactivity in the whole hippocampus was signi cantly reduced, including CA1, CA2/CA3 and DG regions (Fig. 3a, c-d). These results demonstrated that prenatal exposure to sevo urane caused damage to myelinated bers during development of the hippocampus.
Next, transmission electron microscopy was utilized to evaluate ultrastructure of myelin after maternal exposure to sevo urane. The results showed that at PND30 and PND60, the numbers of myelinated axons with compact layers in the sevo urane group were signi cantly reduced compared with those in the control group (Fig. 3e). In addition, g-ratio was used as an index to assess structure and function of axonal myelin, as myelin sheaths thinner or thicker than the theoretical optimal g-ratio of 0.77 cause a decline in conduction velocity in the CNS (Chomiak et al. 2009;Hunt et al. 2017). We found that prenatal exposure to sevo urane caused a signi cant increase in the g-ratio in the hippocampus compared with that of the control group at PND30 and PND60 (Fig. 3f, g). These results suggested that prenatal exposure to sevo urane caused loosening of the myelin sheath in the hippocampus of the mouse brain.
Maternal exposure to sevo urane reduces the expression of myelin-associated genes and proteins To systematically evaluate the effects of prenatal exposure to sevo urane on myelination of oligodendrocytes, we detected myelin-associated genes and proteins via immuno uorescence, western blotting and qRT-PCR. We found that maternal exposure to sevo urane signi cantly reduced the expression of the PDGFRα + oligodendrocyte precursor at PND30 (Fig. 4a-b). Colocalization of CC1 and Olig2 was conducted to assess the maturation of oligodendrocytes, and the number of CC1 + /Olig2 + cells in the CA2/CA3 region at PND30 was reduced after sevo urane treatment, but there were no signi cant changes in the CA1 and DG regions (Fig. 4c-d). We further con rmed probable changes in myelin morphology by calculating the distribution and contents of MBP and NF200 in the hilus region of DG.
Immuno uorescence showed that the intensity of MBP + /NF200 + cells was signi cantly reduced in the sevo urane-treated group at PND30 and PND60 compared with that in the control group (Fig. 4e-f), revealing an indirect effect of prenatal exposure to sevo urane on myelination of postnatal hippocampus.
The mRNA and protein levels of myelination-regulating genes/transcription factors in the hippocampus were also detected, including Olig1, Olig2, Sox10, MBP, CNPase, myelin associated glycoprotein (MAG), myelin oligodendrocyte glycoprotein (MOG), chondroitin sulfate proteoglycan (NG2), myelin gene regulatory factor (MYRF), PDGFRα, and proteio lipid protein (PLP). qRT-PCR showed that almost all of these transcription factors were downregulated after sevo urane treatment at PND30 and PND60 (Fig.  5a). Western blotting showed that the protein levels of PDGFRα, Sox10, Olig1 and MBP were decreased at PND30 and PND60, while the levels of NKX2.2 and Olig2 were not signi cantly changed (Fig. 5b-g). In conclusion, the above results suggested that maternal exposure to sevo urane inhibited the expression of genes and proteins involved in the regulation of myelination in offspring.

Maternal exposure to sevo urane attenuates the conductivity of axons in optic nerves
To investigate whether maternal exposure to sevo urane caused functional changes in axonal myelin, we recorded the CAPs, the sum of the ring action potentials from isolated optic nerves (Fig. 6a). We observed that sevo urane exposure resulted in a 37% reduction in the conduction velocity compared with that of the control group (Fig. 6b). Furthermore, the half-width of the CAP response in sevo urane-exposed mouse nerves was 1.4 times as wide as that in control mice (Fig. 6c). These ndings suggested that maternal exposure to sevo urane decreased the conduction velocity of optic nerve bers in offspring.

Discussion
Every year, millions of pregnant women suffer from nonobstetric surgeries with general anesthetic exposure (Kuczkowski 2006;Okeagu et al. 2020). Although the safety of anesthetic agents has been guaranteed, the adverse effects of anesthetics on the development of the fetus are of great concern (Xu et al.; Yu et al.). It has been widely accepted that prenatal exposure to sevo urane can induce impairment of cognitive function in offspring (Kang et al. 2017). However, studies of the effects of anesthesiainduced neurotoxicity have mainly focused on neurons, not glial cells (Vutskits et al. 2016). Given the integrity and essential role of oligodendrocytes in the development and functional maintenance of mouse brain (Philips et al. 2017;Xin et al. 2020), the potential effects of sevo urane exposure on oligodendrocytes should not be neglected. In this study, by evaluating myelin-associated morphological and functional indices, we provided evidences that gestational exposure to sevo urane could result in disruptions in oligodendrocyte development and myelination in the hippocampus after birth. In addition, we observed impairments of cognitive and motor functions in adult offspring. Our data suggested that myelination of oligodendrocytes might be involved in neurodevelopmental toxicity induced by sevo urane.
Previous studies have reported that exposing pregnant mice to sevo urane leads to increased risks on cognition in offspring (Zuo et al. 2020a). Myelin plays important structural and functional roles in the development of the brain, and degradation of myelin is a key feature of neurological disorders involving cognitive dysfunction (Park et al. 2016). In sevo urane-induced neurotoxicity, the involvement of myelin alterations has also been proposed ). Here, we focused on the effects of maternal exposure to sevo urane on myelin produced by oligodendrocytes in the fetus. We found that exposure to 2.5% sevo urane for 6 h induced damages in the structure of myelin and downregulations of myelinassociated genes and proteins, which was supported by other investigations (Zuo et al. 2020b). In functional study, sevo urane exposure attenuated the conductivity of axons. We guess these disruptions of myelination contribute to the impairment of cognitive function induced by sevo urane exposure, which is worthy of further exploration.
In the CNS, myelination is a highly specialized and tightly regulated process that involves the proliferation, migration, and differentiation of OPCs into myelin-generating cells during development and throughout adulthood (Tsai et al. 2019). Myelination in the hippocampal axons has been reported to profoundly affect learning and memory via regulation of information processing in neural circuits (Fields 2015), and demyelination in the hippocampus is involved in several neurodegenerative and neurological disorders, including multiple sclerosis, Alzheimer's disease, and schizophrenia (Duncan et al. 2016). Herein, we investigated the effects of sevo urane exposure during the mid-gestational period on hippocampal myelination. We found that sevo urane exposure signi cantly decreased the number of BrdU + /Olig2 + and MBP + /Olig2 + double-labeled cells in vitro, revealing inhibitions of OPC proliferation and differentiation. MBP is one of the most abundant structural proteins in myelin and can re ect myelinization and compaction in neuronal cells (Zuo et al. 2020b). We examined the formation and development of myelin sheaths in the hippocampus by studying the immunoreactivity of MBP in vivo. Signi cant reductions of the numbers of myelin bers in the CA1, CA2/CA3 and DG regions at PND14, PND30 and PND60 were observed, and the g-ratio of the myelin sheath was markedly increased. Our ndings indicated that prenatal exposure to sevo urane induced de cits in myelin development.
We further con rm the effects of sevo urane exposure on genes and proteins involved in the regulation of the myelination process via immuno uorescence, western blot analysis and qRT-PCR. We found that the levels In conclusion, our results suggested that gestational exposure to sevo urane had detrimental effects on oligodendrocyte development and axonal myelination in the hippocampus, which might be associated with sevo urane-induced cognitive and motor abnormalities. Further studies should be conducted to clarify the exact mechanisms of these effects and offer a promising strategy for the treatment of sevo urane-induced neurotoxicity.  Sevo urane inhibits the proliferation and differentiation of hippocampal-derived OPCs in vitro (a)