Dexmedetomidine Alleviates the Long-Term Neurodevelopmental Toxicity Induced by Sevoflurane in the Developing Brain

Over the past decades, whether anesthetics can affect the neurodevelopment of pediatric patients remains controversial. Although there are no studies confirming a clear correlation between anesthetics and neurodevelopmental impairments in children, a retrospective study revealed an increased probability of learning deficits in children after prolonged or repeated exposure to anesthetics [1‒3]. In addition, a series of studies in immature animals, including nonhuman primates, have revealed that exposure to anesthetics during development causes neurodegeneration and cognitive deficits [4, 5]. The potential toxicity of anesthetics to neurodevelopment has raised widespread concerns about its safety.

Sevoflurane (SEVO) is widely used in the anesthesia of infants and children due to its favorable pharmacodynamic and pharmacokinetic properties [6, 7]. Many researches have shown that SEVO caused neurotoxicity in the developing brain by promoting apoptosis, inducing neuronal autophagy, and diminishing synapse density [5, 8, 9]. Moreover, various researches indicated that the side effects of SEVO on neurodevelopment are sex related [10, 11]. As there is prevalence of SEVO in pediatric surgery, an increasing number of studies were executed to explore effective strategies to prevent the underlying neurotoxicity of SEVO [12, 13].

Dexmedetomidine (DEX) is an α₂-adrenergic receptor agonist with potent sedative, analgesic, and neuroprotective effects, which is highly sensitive to neuronal damages and attenuates hypoxic-ischemic brain injury in the developing brain, and has been commonly used in combination with anesthetics, such as SEVO, to mitigate its toxic effects [12, 14]. However, several researches proposed that DEX could not decrease but increase the neurotoxicity of SEVO [15, 16].Whether DEX could attenuate the neurological damages of developing brain induced by SEVO and its mechanisms still needs to be identified.

As a multifunctional protein, glycogen synthase kinase-3β (GSK-3β) and its upstream and downstream regulators play critical roles during neurodevelopment and are related to the neurotoxic effects of anesthetics [17]. Disrupted in schizophrenia 1 (DISC1) is associated with mental illness that binds to GSK-3β to regulate neurodevelopmental processes such as neuron proliferation, dendrite growth, and synaptogenesis [18], but the association between DISC1 and anesthetic-induced developmental toxicity has not been clearly established. In the present study, DEX in combination with SEVO in different ways to the developing brains of female and male rats was to explore whether the DISC1/GSK-3β signaling was involved in the DEX attenuating neurodevelopmental toxicity induced by SEVO, as well as the effects of different DEX administration modes and sex differences on SEVO-induced neurotoxicity.

All animals and the experimental procedures were approved by Laboratory Animal Use and Care Ethical Committee of Harbin Medical University (Approval No.: 2022-WZYSLLSC-13). The pregnant Sprague Dawley rats were housed in a facility with 12-h light-dark cycle at 25 ± 1°C and 50% ± 10% humidity and took food and water ad libitum. Postneonatal day 7 (PND7) rat pups (12–16 g, 40 females and 40 males in each group) were randomly divided into six groups by the principle of litter control: ① Control (exposed to air for 6 h), ② DEX (twice intraperitoneal injection of DEX at a dose of 10 μg/kg body weight [b.w.]), ③ SEVO (exposed to 2.5% sevoflurane with 30% oxygen for 6 h), ④ DEX + SEVO (after 30 min intraperitoneal injection of DEX, the pups were exposed to 2.5% sevoflurane for 6 h), ⑤ SEVO + DEX (DEX 10 μg/kg b.w. was injected intraperitoneally 30 min after the end of a 6-h exposure to 2.5% SEVO), ⑥ DEX + SEVO + DEX (DEX 10 μg/kg b.w. was injected before and after the 2.5% SEVO exposure). The SEVO exposure method was described as follows: rat pups were simultaneously placed in a homemade anesthesia chamber with an anesthesia pump attached to the upper vent of the chamber and a gas monitor attached to the lower vent. The chamber was maintained at 37°C. Experimental grouping and scheme are shown in Figure 1A, B.

After 1 h of recovery from anesthesia, the litters were euthanized with pentobarbital; brain tissues were then dissected immediately for molecular pathological analysis.

To examine the long-term effects of SEVO and DEX on learning and cognitive defects, an hour post-recovery, the PND7 rat pups were returned to their dams for feeding until PND21 and PND35. Morris water maze (MWM) test was utilized to assess learning and memory ability of the rats, followed by euthanasia for pathological evaluations.

The protein expression in brain tissue was detected by Western blotting as previously described [19]. Briefly, the hippocampus of rats (n = 3/per group) was dissected on ice and homogenized using RIPA buffer (Beyotime), which contained protease and phosphorylase inhibitors (Sigma-Aldrich). The proteins were separated by 10% polyacrylamide gel electrophoresis, then transferred onto nitrocellulose membrane (PALL, Shanghai, China), and blocked with 5% nonfat milk, followed by incubation with primary antibodies: GSK-3β (1:1,000, Cell Signaling Technology Com [CST], Shanghai, China), pGSK-3β (1:1,000, CST), GAPDH (1:50,000, ABclonal), β-actin (1:1,000, CST), DISC1 (1:500, ABclonal), Bax (1:1,000, ABclonal), and Bcl-2 (1:1,000, Abways) at 4°C overnight. Membranes were incubated with horseradish peroxidase-conjugated anti-rabbit secondary antibody (1:1,000, CST) at room temperature for 2 h. The immunoreactive signals were detected using enhanced chemiluminescence reagents (Thermo Fisher Scientific Inc.) and quantified by ImageJ software (1.48v, NIH, USA).

The learning and memory ability of rats was evaluated by MWM test. A black circular pool (diameter: 180 cm, depth: 60 cm) was filled with tepid water at 25 ± 1°C and was equidistantly divided into four quadrants (I, II, III, and IV). An invisible platform (diameter: 10 cm, submerged: 1 cm) was placed underwater in the “I” quadrant. The MWM test included 4 days of training and probe trials on the fifth day. During the first 4 days of training, the rats (male and female, n = 10/per group, respectively) were placed in the water from four different quadrants, and the time it took to find the hidden platform was recorded. If the rat did not find the platform after 90 s, it would be guided to the platform and stay for 15 s, and the escape latency time was recorded as 90 s. On the fifth day, the platform was removed, allowing the rat swim in the pool for 90 s; the percentage of time spent in the quadrant was recorded where the platform is located. During the test period, each swimming trajectory of the rat was recorded using a ceiling-mounted video camera, and the data were analyzed by MWM software. After experiment, use a towel to dry these animals and put them back to the chamber.

Rats (n = 3/per group) were anesthetized with pentobarbital sodium (50 mg/kg, i.p.), perfused with phosphate-buffered saline (pH 7.40) followed by 4% paraformaldehyde, and the brain tissue was subsequently isolated for dehydration and embedding. The coronal sections of brain were cut at a thickness of 4 μm, and after the sections were deparaffinized and hydrated, endogenous peroxidase blocking agents were used to prevent nonspecific binding. Thereafter, sections were incubated with primary antibodies: anti-cleaved caspase-3 (1:100, CST) and anti-DISC1 (1:100, ABclonal) at 4°C overnight. On the next day, sections were treated with anti-rabbit secondary antibody and followed by a chromogenic reaction by 3,3′-diaminobenzidine; all sections were counterstained with hematoxylin finally. The sections were observed using a digital pathological slide scanner. For the quantification of cleaved caspase-3, the number of positive cells in the cortical and hippocampal areas of each section was counted. The whole hippocampus and five randomly selected visual fields in the cortical area at ×400 magnification of each section were quantified for DISC1-positive cells.

A TUNEL staining kit supplied by Millipore (Serologicals Corporation, Norcross, GA) was used to measure neuronal apoptosis as previous studies [17]. After sequentially treating the tissue with proteinase K and 3% hydrogen peroxide, enough terminal deoxynucleotidyl transferase reaction mix was added to each section to cover the entire tissue and incubate at 37°C for 1 h. The sections (n = 3/per group) were stained with diaminobenzidine and then counterstained with hematoxylin. Finally, the number of apoptotic cells in the cortical and hippocampal areas of each section was counted in a blinded manner.

The tissues (n = 3/per group) were deparaffinized as previously described [19] and incubated with 3% hydrogen peroxide for 10 min, then blocked in 1% BSA. Sections were then incubated overnight at 4°C with primary antibodies: cleaved caspase-3 or DISC1 dilutions containing NeuN488 and then incubated with the appropriate secondary fluorescent antibodies for 1.5 h at room temperature. After washing with phosphate-buffered saline, sections were treated with 2-(4-amidinophenyl)-6-indolecarbamidine dihydrochloride-containing anti-fluorescence quenching mounting medium. The experimental results were processed and consistent with immunohistochemistry.

Total RNA was extracted from hippocampus by TRIzol reagent (HaiGene Biotech Co., Ltd) and cDNA was synthesized with PrimeScript RT Reagent Kit with gDNA Eraser (Takara, Tokyo, Japan). Subsequently, qPCR was performed using TB Green Premix Ex Taq II (Takara). The primer sequences were shown as below: DISC1 F 5′-CTG​GAG​GTC​ACT​TCC​TTG​ATT​T-3′, R 5′-CTG​TTC​CAG​GTC​TTC​CAA​TCT​C-3′; GAPDH F 5′-GTG​CCG​CCT​GGA​GAA​AC-3′, R 5′-AAG​GTG​GAA​GAG​TGG​GAG​T-3′. DISC1 mRNA expression was calculated using the 2^−ΔΔCt method and standardized by GAPDH.

Statistical analysis was carried out using SPSS 23 (IBM SPSS Inc.), and the graphs were created by GraphPad Prism 6.0 (GraphPad Software LLC). The escape latency of MWM was the average of four trials, and the difference between the groups was compared by repeated two-way ANOVA. Data obtained from TUNEL assay, immunochemistry, immunofluorescence, Western blot, reverse transcription-quantitative polymerase chain reaction (RT-qPCR), and the respiratory and mortality rates were analyzed by one-way ANOVA, followed by a post hoc Dunnett’s test or Newman-Keuls test for individual comparisons, and presented as mean values ± standard error.

After the onset of anesthesia, we recorded the mortality of rats in each group every 1.5 h. As shown in Figure 1C, SEVO decreased the survival rate of rats in a time-dependent manner. Since SEVO is an inhalational anesthetic, we recorded the respiratory rate at 1.5 h, 3 h, and 4.5 h during anesthesia. SEVO had an inhibitory effect on respiration (Fig. 1D).

Apoptosis in the hippocampus and cortex was evaluated by TUNEL assay. The number of TUNEL-positive cells in the hippocampus and cortex of PND7 rats in the SEVO group was increased compared with the controls, which could be effectively mitigated by treatment with DEX 10 μg/kg before or after SEVO exposure, but application of DEX 10 μg/kg before and after SEVO was not effective in reducing TUNEL-positive cells (Fig. 1E, F).

Immunohistochemistry was performed to evaluate the activation of caspase-3 in brain sections of rats in each group. In PND7 rats, compared with the control group, SEVO exposure promoted the activation of caspase-3, but the combination of DEX 10 μg/kg before or after SEVO inhibited the activation. However, the number of cleaved caspase-3-positive cells in the DEX + SEVO + DEX group was also higher than the controls (Fig. 2A, B). The immunofluorescence with cleaved caspase-3 (red) and neuron-specific nuclear protein (NeuN, green) confirmed that DEX 10 μg/kg attenuated the activation of caspase-3 by SEVO in neurons. Compared with the control, the application of DEX before and after SEVO exposure increased the number of caspase-3-positive neurons (Fig. 2C, D). Western blot analysis was performed to detect the expression of Bax and Bcl-2; the results showed an increased ratio of Bax/Bcl-2 in the SEVO group. However, a single combined application of DEX before or after SEVO exposure, rather than two, decreased the ratio of Bax/Bcl-2 induced by SEVO (Fig. 3B).

GSK-3β is a serine/threonine kinase and widely expressed throughout the central nervous system, regulates cell survival through various signaling pathways [20]. DISC1 is a multifunctional scaffold protein, which regulates neurogenesis and neural development by inhibiting GSK-3β both in the embryonic brains and adult hippocampus [18]. To examine whether the DISC1/GSK-3β axis is involved in SEVO-induced neurotoxicity, the expression of DISC1 and GSK-3β was determined in the brain tissues. RT-qPCR analysis revealed that DEX significantly ameliorated SEVO-induced decrease in DISC1 mRNA level. Western blot results demonstrated that SEVO inhibited the expression of DISC1 and phosphorylated GSK-3β, and the application of DEX 10 μg/kg before or after SEVO attenuated the inhibitory effect obviously (Fig. 3A, B). In addition, immunohistochemical and immunofluorescence results manifested that DEX 10 μg/kg weakened the inhibitory effect of SEVO on DISC1 expression in neuronal cells and neurons in the hippocampus and cortex (Fig. 3C, D, 4a, b).

During the MWM test of PND21 rats, SEVO-treated female and male rats were inferior to the control, DEX, DEX + SEVO, and SEVO + DEX group with significantly longer time to locate the platform. Compared with controls, the time spent in the targeted quadrant in the SEVO group of both sexes was significantly shortened, but the DEX + SEVO, SEVO + DEX group spent more time in the targeted quadrant than the SEVO group (Fig. 5a–c). In addition, by comparison with the control group, the levels of pGSK-3β and DISC1 in the hippocampus of the SEVO group rats were decreased, but the levels of pGSK-3β and DISC1 in the DEX + SEVO and SEVO + DEX groups were significantly higher than the SEVO group (Fig. 5d–g).

To investigate the sustained damaging effect of SEVO on memory function, the PND7-treated pups were kept until PND35 for the MWM test. In the training trials, female and male rats exposed to SEVO exhibited a long escape latency and a diminished target quadrant swimming time. Interestingly, a single application of DEX before or after SEVO exposure ameliorated the MWM performance (Fig. 6a–c). Furthermore, DEX before or after SEVO exposure attenuated SEVO-induced reduction in DISC1 and pGSK-3β expression (Fig. 6d–g). To prove our hypothesis, the binding of SEVO and DISC1 was analyzed by molecular docking, and DISC1 was found to be the target of SEVO (Fig. 7a–c).

Here, we proposed that exposure of juvenile rats to SEVO induces neurotoxicity, resulting in persistent learning and memory dysfunction. DEX 10 μg/kg treatment before or after 2.5% SEVO exposure in developing rats effectively attenuated SEVO-induced long-term neurotoxicity and improved the cognitive function in rats by regulating the DISC1/GSK-3β signaling. Moreover, sex differences in the neurotoxicity of SEVO and the protective effect of DEX were not observed.

As one of the most commonly used anesthetics in clinical trial, SEVO has been found to cause long-term neurotoxicity and cognitive dysfunction in developing animals [21, 22]. Previous studies have shown that exposure to 2.5% SEVO for 2 h induces neurotoxicity in adult mice brain tissue, causing nerve cell apoptosis in the offspring of pregnant mice [23, 24]. In our work, we observed that exposure to 2.5% SEVO for 6 h during critical developmental periods caused a significant increase in the number of apoptotic neurocytes and neurons in the hippocampus and cortex of PND7 rats, which are involved in spatial memory formation [25]. Since neurons are chiefly nonrenewable, this may be the main cause of SEVO-induced sustained damage to memory function. Our previous studies have shown that neuronal apoptosis is associated with GSK-3β activation [17].

GSK-3β is the homologous mammalian isoform of GSK-3 and is involved in a variety of processes such as division, proliferation, and differentiation [26]. Activated GSK-3β increases neuronal apoptosis and is associated with neurological diseases, but phosphorylation at Ser⁹ negatively regulates its pro-apoptotic activity [27]. DISC1 was originally identified as a candidate gene for schizophrenia [28]. It is currently believed that DISC1 is a multifunctional protein involved in multiple signaling pathways and regulates diverse processes of neurodevelopment such as neuronal proliferation, migration, dendritic growth, and synaptogenesis. Furthermore, DISC1 regulates the transport of neurotransmitters, vesicles, and mitochondria, thus affecting the function of neurons [29]. DISC1 regulates the behavioral outputs via modulation of GSK-3β [18]. Here our findings demonstrated that DEX attenuated the neurotoxicity of SEVO and improved cognitive function by upregulating the expression of DISC1.

DEX, an α₂-adrenergic receptor agonist, has been found to reduce brain neurotoxicity induced by anesthetics [30]. However, whether it could prevent the neurological damages of anesthetics is still controversial, which may be closely related to the dose and the mode of administration [31]. Based on our previous study, DEX 10 μg/kg was chosen as the experimental dose. In this work, our findings suggested that DEX 10 μg/kg attenuates SEVO-induced apoptosis and cognitive dysfunction by regulating the DISC1/GSK-3β signaling in the developing brain. Our previous studies have shown that DEX at low doses were neuroprotective, but high doses were neurotoxic [19]. Another research suggested that cumulative high doses of DEX (5 μg/kg and above) promote SEVO (1.1%)-induced neuronal apoptosis at subanesthetic doses in the brain of neonatal rats [16]. Moreover, DEX (1 μg/kg) decreased the apoptosis of nerve cells in multiple brain regions caused by 2.5% SEVO exposure for 6 h, while DEX (5–25 μg/kg) combined with 2.5% SEVO increased mortality [32]. Therefore, application of DEX both before and after SEVO was not neuroprotective due to excessive cumulative dose.

Some studies have suggested that sex has a momentous impact on cognitive functions after SEVO treatment [33], while a clinical trial demonstrated that there is no difference between men and women in memory performance during SEVO anesthesia [34]. We experimented with adolescent rats of both sexes. In the current study, the learning and memory functions of the experimental groups did not have difference by sex. Unfortunately, there is not a clear explanation for how these differences occurred, which may be because the rats at PND35 are all pubertal.

There are several limitations in the current study. First, SEVO inhibited respiration in rats, and arterial blood gas analysis was supposed to be performed during anesthesia. Zhou et al. [35] have assessed that 2.3% SEVO did not alter the oxygen tension and carbon dioxide tension, and Goyagi [36] indicated that 3% SEVO did not significantly increase the partial pressure of carbon dioxide. In order to prevent hypercarbia, the pups were exposed to SEVO with 30% oxygen, so blood gas analysis was not tested in this study. Second, only the MWM test was used to assess the learning and memory ability of PND21 and PND35 rats, and other behavioral tests are needed to further assess cognitive function. Third, cellular experiments are needed to validate the findings of animals and to further investigate the underlying mechanisms such as the cell models of SEVO. In addition, the neurotoxic effects of SEVO on the immature brain, such as neuroapoptosis and cognitive deficits, have been found only in nonhumans and are not yet clearly established a connection with children. Nevertheless, our study contributes to the design of new strategies to prevent potential long-term neurotoxicity of SEVO in child patients.

In summary, our findings demonstrated that exposure to 2.5% SEVO during the critical period of neurodevelopment increased neuroapoptosis and impaired cognitive function in weaning and adolescent rats. A single dose of DEX (10 μg/kg) before or after SEVO exposure attenuated the neurotoxicity and improved cognitive function by reducing neuronal apoptosis and modulating the GSK-3β/DISC1 signaling pathway. Our findings may provide new insights into designing new therapies for preventing the neurotoxicity of SEVO and avoiding the abuse of anesthetics.

All animals and the experimental procedures were approved by Laboratory Animal Use and Care Ethical Committee of Harbin Medical University (Approval No.: 2022-WZYSLLSC-13).

The authors declared that there are no conflicts of interest.

This work was supported by the Special Grant Project of the Fourth Affiliated Hospital of Harbin Medical University, China (No.: HYDSYTB202301).

Jia-Ren Liu, Shu-Jun Zhang, and Si-Hua Qi designed and supervised the study. Ting-Ting Yang, Ran Wei, Fei-Fei Jin, and Wei Yu performed the experiments and analyzed the data. Zhang Fang and Peng Yu conducted repeated experiments and maintained the rats. Ting-Ting Yang wrote the manuscript. Jia-Ren Liu and Shu-Jun Zhang revised the manuscript. All the authors read and approved the final manuscript.

Ting-Ting Yang, Ran Wei, Fei-Fei Jin, and Wei Yu: co-first authors.

All data generated or analyzed during this study are included in this article. Further inquiries can be directed to the corresponding author.