Introduction

There is increasing public and scientific attention regarding potential neurotoxic effects associated with anesthesia in the developing brain during pregnancy and after birth in animal models and humans [1, 2]. In the United States, approximately 1–2% of pregnant women require non-obstetric surgery or fetal intervention procedures under anesthesia [3, 4]. Clinical studies indicate that children who were subjected to surgery and anesthesia at an early age are at increased risk for neurodevelopmental deficits and cognitive dysfunction [5,6,7]. The etiology of cognitive decline or impairment in children who were exposure to anesthesia in utero remains unclear [8]. It is well appreciated that the fetal period is a critically vulnerable time point for the developing brain. Limited existing literature has already uncovered relationships between some commonly used anesthetics (e.g., ketamine, nitrous oxide, GABA agonists) and their causal relationship with long-term brain and behavioral changes [9, 10].

Desflurane is one of the anesthetics that is currently used in surgery during pregnancy as an inhaled anesthetic drug. Current literature reports that inhaled anesthetics may have neurotoxic effects in the brains of rodent models [11, 12]. Previously, we reported that the anesthetic sevoflurane induced neurotoxic effects in fetal and offspring mice [13, 14], while isoflurane and desflurane may cause cognitive impairment [15, 16]. However, the effects of desflurane use during pregnancy on neuronal development and synaptic formation in offspring remain unknown.

In the present investigation, we hypothesized that desflurane exposure and surgery during the gestational period to children’s early-life may adversely affect neural and synaptic development and impair the cognition in fetal and offspring mice.

Materials and Methods

Mouse Anesthesia and Sciatic Nerve Hemi-transection Surgery

All animal experiments were conducted in accordance with institutional animal care guidelines of Beijing Chaoyang Hospital affiliated with Capital Medical University and approved ethically by the Administration Committee of Experimental Animals. Thirty C57BL/6J female mice and male mice (Beijing Vital River Laboratory Animal Technology Co., Ltd., Beijing, China), which were 3 months old, were paired for reproduction. Then pregnant mice were housed individually. Animals were kept in a temperature and humidity-controlled (22–23 °C, 50–60%) room under a 12-h light–dark cycle; standard diet food and water were available ad libitum. Pregnant mice were randomly assigned to the desflurane group, the desflurane plus sciatic nerve hemi-transection (SNT) group or the sham group at gestational day (G)14. The desflurane group received 10% desflurane in 100% oxygen for 3 h in a transparent anesthesia box. Mice randomly assigned to the desflurane plus SNT group were exposed to desflurane, and under aseptic conditions, the right sciatic nerve was exposed at thigh level and subjected to surgical hemi-transection. Anesthesia was maintained via a nose-cone throughout surgery, and desflurane exposure time is 3 h. The sham group received 100% oxygen at an identical flow rate for 3 h in the same chamber. The mice breathed unassisted, and concentrations of oxygen and anesthesia were continuously monitored (Drager Vamos, Germany). The mice’s heart rate and pulse oxygen saturation were also monitored continuously. The anesthesia chamber was placed in a small animal temperature controlled operating table and rectal temperature of animals was kept at 37 °C ± 0.5 °C. After anesthesia administration, the animals were placed in a box supplied with 100% oxygen until they regained the righting reflex. The aim of this study is to demonstrate whether surgery and 10% desflurane (approximately 1.2 minimum alveolar concentration) [17] with 3 h for pregnant mice which mimics clinical medical useage of desflurane could lead to fetal mouse neurotoxicity that would alter cognition in P31 offspring mouse.

Primary Neuronal Culture

The hippocampus of postnatal (P0-P1) C57BL/6 mice was quickly dissected on ice and trypsinized at 37 °C in trypsin-ethylenediaminetetraacetic acid (EDTA; 0.25%) (Life Technologies, Grand Island, NY, USA). Tissues were then washed in 1 × Dulbecco’s modified Eagle medium (DMEM) with 10% FBS to inactivate trypsin. Afterward the cell suspension was pelleted at 600 g for 5 min and resuspended in serum-free Neurobasal medium (Invitrogen). A neurobasal medium supplemented with B-27® supplement (Life Technologies, Grand Island, NY, USA) and penicillin/streptomycin (KeygenBiotech, Beijing, China) was used to culture cells. The cells were seeded on poly-d-lysine coated coverslips or petri dishes at 37 °C and 5%CO2 cell incubator (Thermo Fisher Scientific, USA). Every 3 days, half of the medium was replaced by the pre-warmed medium.

All cultures were maintained for at least 28 days before treatments, unless otherwise indicated. When neurons became matured, they were treated with 10% desflurane for 3 h as described by our previous studies [13], control cells had no desflurane, which also formed a basis for our determination of anesthesia concentrations that could potentially affect synapse formation via regulation of the postsynaptic density-95 (PSD-95) protein. The cells were harvested at the end of exposure to anesthesia and were then fixed and studied via immunofluorescence, real-time PCR, and Western blot analysis.

Cell Immunofluorescence

Cells on coverslips were fixed for 15 min at 37 °C with 4% PFA, rinsed with PBS and then permeabilized with 0.3% Triton X-100 in PBS buffer for 5 min at room temperature. Nonspecific binding sites for 60 min in 5% goat serum in PBS buffer were blocked, followed by incubation with primary antibodies diluted in blocking solution containing 0.025% triton X-100 overnight at 4 °C. The next day, secondary antibody Alexa Fluor 488 Goat-anti-Rabbit IgG (1:500, ab150077) was added to the cells on coverslips, for 60 min at RT in a dark room, followed by PBS rinses. As appropriate, DAPI (with Prolong Gold Antifade Reagent) staining of cell nucleus was performed for 5 min at RT. Primary antibodies used were monoclonal anti-NeuN (1:300, ab177487) and monoclonal anti-synaptophysin (1:100, ab32127).

Western Blot

At 24 h after desflurane anesthesia and surgery, fetal mice were collected after cesarean section and brain tissues were isolated. The whole brain was homogenized for Western blot experiments. In long-term experiments, postnatal day (P)31 offspring were sacrificed via decapitation. The hippocampal tissue proteins were analyzed by Western blot and RNA by real-time PCR. For Western blot, the harvested tissues were homogenized on an ice-cold RIPA lysis buffer (C1053, Applygen, Beijing, China) plus a protease inhibitors cocktail (P1265, Applygen, Beijing, China). Lysates were centrifuged at 4 °C, 12,000 x g for 15 min, and the supernatant was obtained. Protein concentrations were determined using a bicinchoninic acid (BCA) protein assay kit (Applygen, Beijing, China).

An equal mass of proteins (40 μg) for each sample was loaded in each well and electrophoresed on SDS-PAGE gels, then transferred to polyvinylidene fluoride membranes or nitrocellulose membranes. Before the membranes were sealed with 5%BSA in Tris-buffered saline with Tween-20 (TBST), the membranes were stained with ponceau. Then they were incubated with primary antibody at 4 °C for 12 h. Primary antibodies used were interleukin (IL)-6 antibody (ab7737, 1:20 dilution; 24 kDa, abcam, Cambridge, MA), PSD-95 antibody (ab18258, 1:1000 dilution; 85 kDa, abcam), caspase-3 antibody (ab13847,1:500 dilution; 17 kDa, abcam), β-actin (ab8227, 1:5000 dilution; 42 kDa, abcam). The second day, TBST was used to wash the membranes three times, then membranes were incubated for 1 h at room temperature with horseradish peroxidase (HRP) conjugated secondary antibodies (1:5000, goat anti-rabbit, SouthernBiotech, USA). The Bio-Rad (ChemiDoc™ XRS + , USA) image software (Quantity One) was used to semi-quantify protein bands. β-actin levels were used to normalize other protein levels and to control loading quantity of protein samples. We used the ratio of a protein level to the β-actin level to compare experimental groups with the sham group to describe the changes of protein levels.

Quantitative Real-Time PCR

Trizol reagent (Invitrogen, Carlsbad, Calif, USA), according to the manufacturer’s instructions, was used to extract the total RNA. An ultraviolet spectrophotometer (NanoDrop, Wilmington, DE, USA) was used to detect RNA concentration at 260–280 nm. An optical density (OD) value was found to be 1.8-2.0 which was considered qualified. The primers for PCR were obtained from Sangon Biotech (Shanghai, China) (Table 1). The total reverse transcription reaction system was 20ul to synthesize cDNA according to the PrimeScript™ RT reagent kit with gDNA Erase (Perfect Real Time) (Takara, Japan). PCR amplification was carried out by the ABI PRISM 7500 thermocycler (Applied Biosystems, Foster City, CA, USA) using Talent qPCR PreMix (SYBR Green) (Tiangen, Beijing, China). We used the manufacturer’s recommended two-step method involving an initial denaturation at 95 °C for 3 min for 1 cycles, and a cycling stage at 95 °C for 5 s, 60 °C for 32 s for 40 cycles total. The results using this 2−∆∆Ct method were calculated and then normalized by compared with levels of GAPDH [18].

Table 1 Primer sequence
Full size table

Open Field Monitoring

We assessed changes in behavior of subjects from the anesthesia and surgery groups using open field monitoring [19]. The open field monitoring was conducted using a white opaque plastic chamber (40 cm × 40 cm × 40 cm, Beijing Macroambition S&T Development Co., Ltd., Beijing, China) with P31 offspring mice and the exploratory locomotor activity was assessed according to our previous publication [20]. Mouse was placed into the middle of the stage and left to move freely for 5 min while a video tracking system was used to record the activity of the mice. The total distance and velocity of stage was quantified during the test.

Fear Conditioning Test

Test mouse was put in a conditioning chamber (Beijing Macroambition S&T Development Co., Ltd., Beijing, China) and allowed to probe into the chamber for 3 min, then the mice were subjected to a battery of one 30 s tone (70 db, 1.4 kHz), followed by one 2 s footshock (0.75 mA), repeated after 120 s for five total foot shocks. Afterwards, the mice were left in the chamber for another 30 s before they were put back in the home cage. 75% ethyl alcohol was used to sanitize the chamber after each use. After 24 h, their fear of the conditioned context (a hippocampus-dependent task) was evaluated and quantified. For estimating contextual fear conditioning, the mice were placed in the previous chamber for five minutes without tone presentation or foot shock, and the duration of observed freezing behavior was tallied and scored. After 2 h, the auditory-cued fear test was performed. The mice were placed in an alternate chamber with an odor composed of 1% acetic acid solution instead of ethyl alcohol. After 3 min of free exploration and acclimatization, the tone was played for 180 s. After a short rest, the mice were returned to their home cages. Their freezing behavior was observed and scored throughout the experiment. Freezing behavior was defined as time when all movements were absent, except for respiration, which were tallied and summarized as a percentage of the total observed time. The FCT freezing times for both context and tone tests were used to assess fear memory function.

Statistics

Data were expressed as mean ± SEM. A one-way analysis of variance (ANOVA) followed by pairwise or Student’s t test was used to compare differences between control groups, and data were normally distributed. Values of p < 0.05 were considered statistically significant. SPSS software version 18.0 and GraphPad PRISM version 5 software were used to analyze the data and plot graphs.

Results

Sciatic Nerve Hemi-transection under Desflurane Anesthesia in Pregnant Mice Induced Cognitive Impairment in P31 Offspring Mice

Pregnant mice with the gestational day 14 (G14) mice were randomly divided into three groups: a sham group, a desflurane group, and a desflurane plus sciatic nerve hemi-transection (SNT) group. After birth, the offspring were assessed for fear conditioning at P31 (Fig. 1). Freezing times for both context and tone tests of the mice were used to determine impairments in fear memory. We found that surgery and desflurane anesthesia but not desflurane anesthesia alone significantly altered the freezing time of offspring mice in both context tests [F(2, 27) = 9.77, p < 0.05]and tone tests [F(2, 27) = 14.57p < 0.05] of FCT (Fig. 1). In the open field test 31 days after birth, there were no significant changes in the total distance [F(2, 27) = 0.10, p > 0.05] and velocity [F(2, 27) = 0.07, p > 0.05] traveled by mice after the anesthesia plus SNT groups compared to anesthesia only mice (Fig. 1). The results suggested that desflurane and surgery didn’t affect the mice’s ability to exercise.

Fig. 1
figure 1

Desflurane anesthesia and sciatic nerve hemi-transection surgery to pregnant mice impaired fear memory in offspring mice. a The experimental timeline for desflurane anesthesia and sciatic nerve hemi-transection. b, c The freezing time to context and tone were decreased in the desflurane plus SNT group compared with the sham group (n = 10). d The total freezing time from the training phase. e, f Locomotor measurement of travel distances and velocity (n = 10). *,#p < 0.05, *compared with the sham group, #compared to the desflurane alone group

Full size image

Desflurane Anesthesia in Primary Hippocampus Neurons Reduced PSD-95 and Synaptophysin Levels

The hippocampal neurons in culture for 28 days were treated with desflurane (10% concentration). Immunofluorescence analysis showed that the desflurane exposure of 3 h significantly reduced level of the presynaptic protein synaptophysin (p < 0.05) while number of NeuN+ neuronal cells was unchanged (p = 0.77) [21, 22]. Quantifications by Western blot and real-time PCR showed that 10% desflurane for 3 h significantly increased both mRNA and protein levels of proinflammatory cytokine IL-6 and activated caspase-3, but decreased level of PSD-95 in the primary hippocampal cultures compared to the sham group (Fig. 2). These results suggested that desflurane induced detrimental effects in hippocampal neurons of the mice.

Fig. 2
figure 2

Desflurane exposure in primary hippocampal neurons. Desflurane exposure reduced synaptophysin in neurons. a The experimental timeline for desflurane treatment. b–e Desflurane anesthesia led to a reduction in synaptophysin. *p < 0.05, *compared with the sham group

Full size image

Desflurane Anesthesia in Pregnant Mice Caused Neurotoxic Effects in Fetal Mice

Real-time PCR and Western blot analysis were used to measure alterations in brain tissues of the G14 fetal mice, following desflurane exposure with surgery or without surgery. Compared to the sham group, treatment with desflurane alone increased IL-6 and caspase-3 levels and reduced PSD-95 levels in the fetal mice. Surgery further aggravated desflurane-induced augments in IL-6, caspase-3 and the reduction in PSD-95 levels [mRNA level: F(2, 6) = 74.17, F(2, 6) = 319.0, F(2, 6) = 31.52, protein level: F(2, 15) = 5.71, F (2, 15) = 4.41, F(2, 15) = 11.47, respectively, p < 0.05] (Fig. 3).

Fig. 3
figure 3

Molecular mechanisms of desflurane exposure in primary hippocampal neurons and mice. Desflurane exposure increased IL-6 and caspase-3 levels, and decreased PSD-95 protein levels. a–h Desflurane treatment increased IL-6 and caspase-3 levels, decreased PSD-95 levels in primary neurons. i–p Desflurane anesthesia together with the surgery further increased IL-6 and caspase-3 levels, and decreased PSD-95 levels in fetal mice (n = 6). *,#p < 0.05, *compared with the sham group, #compared to the desflurane alone group

Full size image

Sciatic Nerve Hemi-transection and Desflurane Anesthesia in Pregnant Mice Lowered PSD-95 Levels of Offspring Mice

Considering that sciatic nerve hemi-transection surgical insult alongside desflurane anesthesia in pregnant mice resulted in impaired memory function in their offspring mice, we assessed the effects of the surgery and desflurane anesthesia on the levels of PSD-95 in the hippocampus. The hippocampal tissues of P31 offspring mice from different groups were examined. Western blot and real-time PCR for PSD-95 indicated that, compared with the control condition, the desflurane plus SNT group exhibited significantly lower levels of PSD-95 in the brain tissues of offspring mice [mRNA level: F(2, 6) = 30.08, protein level: F(2, 15) = 11.47, respectivitely, p < 0.05] (Fig. 4).

Fig. 4
figure 4

Long-term molecular mechanisms after desflurane exposure. Anesthesia and sciatic nerve hemi-transection surgery in pregnant mice decreased PSD-95 levels of the postnatal day (P)31 offspring mice. a Immunofluorescence showed the neuronal marker NeuN (red) and cell nucleus DAPI (blue). Dashed line on the hippocampal tissue section of P31 offspring mice indicated the dissected region for qPCR (b) and Western blot (c, d) analyses, b–d quantification results showed that sciatic nerve hemi-transection and desflurane anesthesia decreased PSD-95 levels (n = 5). *,#p < 0.05, *compared with the sham group, #compared to the desflurane alone group (Color figure online)

Full size image

Discussion

In the present study, pregnant mice exposed to sciatic nerve hemi-transection and desflurane anesthesia exhibited detrimental effects on fetal and offspring mice. These data suggested that 10% desflurane for 3 h in pregnant mice induced neurotoxicity, via an increase in levels of proinflammatory cytokine IL-6 and caspase-3, with a concurrent reduction in presynaptic and postsynaptic markers. Pregnant mice submitted to desflurane anesthesia and an invasive surgical insult displayed an even more severe phenotype. These neurotoxic effects may be associated with cognitive impairments previously reported in the offspring mice [18, 19, 21, 22].

The gestational day 14 in pregnant mice corresponds to a second trimester of human pregnancy [23], which is a critical period for brain development. In clinical practice, traumatic surgical intervention during pregnancy occurs at a rate of about 6–7% [4]. In this study, the right sciatic nerve hemi-transection in the mid-thigh has been performed, which avoids the Y shaped uterus to potentially cause abdominal disturbances. It has been previously reported that inhaled anesthetics sevoflurane and isoflurane in pregnant mice induces behavioral changes in offspring mice [13, 24]. The Food and Drug Administration (FDA) in the US has warned that children of less than 3 years old and pregnant women who were submitted with anesthesia for more than 3 h or used anesthetics repeatedly may have negative impacts on brain development [1, 25]. Our data provided an evidence in the animal model that desflurane, a common inhaled anesthetic thought to be less toxic than isoflurane [26], may still be able to show adverse neurotoxic effects during pregnancy in developing brain of offspring babies.

We showed that desflurane increased IL-6 and caspase-3 levels and reduced PSD-95 and synaptophysin, suggesting that desflurane alone could induce detrimental effects on neurons and synapses by a caspase-3 dependent apoptotic pathway and neuroinflammation. We combined a surgery with this anesthetic exposure, which showed additional neurotoxicity. One of the reason is that inflammation triggered by surgery contributed to the worsen effect [27]. Many studies have shown that NF-κB pathway is important in the development of neuroimmune activation, and inhaled anesthetic could activate the NF-κB pathway. This would promote production of pro-inflammatory cytokines IL-6, TNF-α, IL-1β. IL-6 was reported to mediate cognitive dysfunction in animals and humans [28,29,30,31], and is considered to be a potential regulator of prenatal materal immune activation (MIA) to affect neural development of offspring. Animal experiments have shown that the lack of IL-6 signaling in trophoblasts blocks MIA-induced inflammatory responses in the fetal brain. The MIA induction could elevate the level of IL-6 in fetal brains. Injections of recombinant IL-6 into pregnant mice could lead to behavioral abnormalities in offspring [32]. Extensive arguing research suggests that caspase-3 plays a central role in cell death [33]. PSD-95, a postsynaptic marker, can orchestrate synaptic development and may also play a role in synapse stabilization and plasticity [34]. It has been demonstrated that lower levels of PSD-95 may be related to cognitive impairment [35, 36]. Therefore, our present study implies that desflurane together with surgery could increase the levels of IL-6 and caspase-3, leading to hippocampal neuroinflammation and neurotoxity, decreasing the levels of PSD-95 and synaptophysin, and interfering with synaptic plasticity, leading to decreased cognitive functions.

The effect of general anesthesia relies on various types of ion channels (e.g., sodium, potassium, calcium, receptors of NMDA, GABAA) [32, 37,38,39,40,41], which can affect function and regulation of neuronal cells. Disruptions in mouse brain functions led by anesthesia may involve the NF-κB signaling pathway [42], mitochondrial dysfunction [43, 44], Tau protein [42] or ubiquitin–proteasome pathways [43]. Anesthesia and surgery in pregnant mice also imposes stress on fetal mice via disruptions of the blood brain barrier and via toxic glial activation [45]. The long term effects of anesthesia of different anesthetics exposure during pregnancy on offspring merit further investigation.

In conclusion, clinically relevant desflurane exposure together with surgery procedures in pregnant mice could lead to acute and chronic neurotoxicity in fetal mice, including increases in inflammation, apoptosis and disruptions of synaptic formation. Desflurane displayed acute toxic effects on the synapses of hippocampal neurons.