Genetic variation of rs12918566 affects GRIN2A expression and is associated with spontaneous movement response during sevoflurane anesthesia induction

Abstract N‐methyl‐D‐aspartate (NMDA) receptors mediate excitatory neurotransmission in the nervous system and are preferentially inhibited by general anesthetics such as sevoflurane. Spontaneous movement is a common complication during sevoflurane anesthesia induction and seriously affects operations. In this study, we investigated the relationship between NMDA polymorphisms and spontaneous movement during sevoflurane induction. This prospective clinical study enrolled 393 patients undergoing sevoflurane anesthesia as part of their surgical routine. In the GRIN1, GRIN2A, and GRIN2B genes, 13 polymorphisms that form a heteromeric complex as part of the NMDA receptor were selected using Haploview and genotyped using matrix‐assisted laser desorption ionization–time of flight mass spectrometry MassARRAY. Both RNAfold and Genotype‐Tissue Expression portals were used to identify gene expression profiles. Our data showed that 35.8% of subjects exhibited spontaneous movement. The GRIN2A rs12918566 polymorphism was associated with spontaneous movement during sevoflurane induction. A logistic regression analysis of additive, dominant, and recessive models indicated a significant association (odds ratio [OR] (95% confidence limit [CI]): 0.58 (0.42–0.80), p = .00086; OR (95% CI): 0.51 (0.31–0.84), p = .0075, and OR (95% CI): 0.47 (0.27–0.81), p = .0060, respectively). After false discovery rate (FDR) correction, the additive model was still significant with a P FDR =0.010. Bioinformatics demonstrated that the rs12918566 genomic variation affected GRIN2A expression in brain tissue. We also revealed that GRIN2A rs12918566 was significantly associated with spontaneous movement during sevoflurane induction. We believe the NMDA receptor plays an important role in regulating the anesthetic effects of sevoflurane.


Sevoflurane is a volatile anesthetic approved by the Food and Drug
Administration for the induction and maintenance of general anesthesia and is one of the most commonly used volatile anesthetics in pediatric, adult, and outpatient surgical patients (Brioni et al., 2017;De Hert & Moerman, 2015). Sevoflurane provides hypnosis, amnesia, analgesia, immobility, and autonomic blockade during surgical and procedural interventions. It is commonly accepted that amnesia, sedation/hypnosis, and immobility are the three vital anesthesia requisites (Constant & Sabourdin, 2015). However, like other volatile anesthetics, brief spontaneous movements often occur during sevoflurane anesthesia induction and recovery, with some patients experiencing epileptiform electroencephalogram changes (Jantti & Sonkajarvi, 2004;Vlajkovic & Sindjelic, 2007;Yli-Hankala et al., 1999). Studies have reported that spontaneous movement incidences in children receiving sevoflurane may be as high as 60%-91% (Akeson & Didriksson, 2004;Chawathe et al., 2005;Constant et al., 1999). An ideal anesthetic agent is characterized by anesthetic and analgesic actions, no adverse effects on the respiratory and circulatory system, and no irritation to the skin and mucous membranes. Severe spontaneous movements during anesthesia may lead to patients falling out of bed or surgical complications and seriously affect patient safety and surgical progress (Borgeat et al., 1991;Chan et al., 1996;Nagata et al., 2004;Reynolds & Koh, 1993).
Studies have shown the spinal cord is the primary site affected by volatile anesthetics, often causing immobility, whereas sedation (hypnosis) and amnesia involve supraspinal mechanisms (Jinks et al., 2008). However, the molecular mechanism underpinning spontaneous movement during sevoflurane induction is unclear.
Scientists have discovered that sevoflurane alters postsynaptic currents by acting on ion channel receptors on the striatal synapse, generating synaptic excitation and spontaneous movement (Oose et al., 2012).

Studies on the isolated spinal cord have suggested that actions
involving excitatory α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid and N-methyl-D-aspartate (NMDA) receptors but not inhibitory (glycine and γ-aminobutyric acid type A [GABA A ]) transmission may produce immobility . NMDA receptors are composed of a major class of excitatory ligand-gated ion channels in the nervous system. Their transient activation leads to long-term modification of synaptic connections and refinement of neuronal circuits, which are involved in learning and memory formation (Irifune et al., 2007). Ishizaki et al. reported a maximal 30% decrease in isoflurane minimum alveolar concentration (MAC) in rats given intrathecal NMDA antagonist bolus doses (Ishizaki et al., 1996). A study investigating antagonized NMDA receptors indicated that spinal cord NMDA receptors might mediate a portion of the immobilizing effects of isoflurane (Stabernack et al., 2003).
Molecular studies have revealed that NMDA receptors exist as heteromeric complexes composed of GRIN1, GRIN2A-D, and GRIN3A-B subunits (Paoletti et al., 2013). GRIN1 is the essential NMDA receptor channel-forming subunit that contains a glycine binding site. The GRIN2 subunit contains a glutamate binding site with affinities for agonists and antagonists. GRIN2A/Bs are abundant in the central nervous system (Nishikawa, 2011;Teng et al., 2010). In contrast, NMDA receptor sensitivity to anesthetics does not appear to depend on GRIN3 subunits (Petrenko et al., 2014). In a Xenopus laevis model expressing human GRIN1/GRIN2B receptor channels, equal concentrations of inhaled anesthetics inhibited NMDA receptors to varying degrees (the inhibitory effects of sevoflurane on GRIN1/GRIN2B expression were significantly higher than isoflurane: 14%-28%), and this result may be useful for defining the role of NMDA in producing general anesthetics (Solt et al., 2006). From this literature, GRIN1 and GRIN2A/B subunits are more likely to contribute to the effects of inhalation anesthesia.
Therefore, we hypothesized that NMDA polymorphisms might be associated with spontaneous movement responses induced by sevoflurane. Our aim was to determine associations between NMDA (GRIN1, GRIN2A, and GRIN2B), tag-SNPs (tag-single-nucleotide polymorphisms), and spontaneous movement responses after sevoflurane induction.  pregnancy, fever, water and electrolyte disorder, hyperthyroidism, severe anemia, serious history of cardiopulmonary disease, and abnormal liver and kidney function.

| Treatment
Patients received the following treatment before general anesthesia: fasted for 8 hr and received no liquids for 2 hr prior to anesthesia. An intravenous infusion channel was established before anesthesia, and a multi-function monitor assessed blood pressure, heart rate, and electrocardiogram output. Bispectral index (BIS) was used to monitor anesthesia depth. Patients inhaled 3% sevoflurane (5 L/min) for 4 min using a pressurized sealing mask. The sevoflurane was then replaced with air and administered at 5 L/min for 2 min and then propofol for routine anesthesia. Throughout anesthesia, the airway was opened and ventilation assisted based on end-of-breath carbon dioxide partial pressure (PaCO 2 ) (maintained at 35-45 mm Hg).
Heart rate (HR) and blood pressure were monitored. When the HR was <50 beats/min, patients were given atropine (0.3 mg-0.5 mg).
When the mean arterial pressure (MAP) was <30% of baseline, 5 mg ephedrine was given, and 0.25 mg nitroglycerin was given when the MAP was >30% of baseline.
Modified observer's assessment of alertness/sedation scoring was performed every 30 s; (5, readily responds to the name spoken in a normal tone; 4, lethargic response to the name spoken in a normal tone; 3, responds only after the name is loudly and/or repeatedly called; 2, responds only after mild prodding or shaking; 1, responds only after a painful trapezius squeeze; 0, no response after a painful trapezius squeeze). Patients were deemed unconscious when the score was <2. Spontaneous movement criteria were also applied: grade 0: no movement, quiet, cooperative; grade I: mild movement, limited to hands or feet, no rotation in arms or legs; grade II: moderate/mild movement, arm or leg flexion or extension; Grade III: strenuous/moderate exercise, plus internal or external rotation of shoulders or hips and knee, severely struggling and required an external force to stop (Borgeat et al., 1991). In our study, a 6-min evaluation period was observed after sevoflurane administration, and the occurrence of any movement in grade I, II, and III was defined as a spontaneous movement. The BIS value, end-tidal anesthetic concentration (ETAC), and MAC were also recorded when patient activity appeared.

| Sample collection, DNA extraction, singlenucleotide polymorphism selection, and genotyping
We collected 2 ml venous blood in ethylenediaminetetraacetic acid anticoagulant tubes. DNA was extracted using a HiPure Blood DNA Mini Kit (Magen, China) according to the manufacturer's protocols.
Assay Designer software (version 3.1) was used for the selected SNP primer design. Polymorphisms were detected using the Sequenom MassARRAY Genotype Platform (Sequenom, San Diego, California, USA) using an allele-specific matrix-assisted laser desorption ionization-time of flight mass spectrometry method.

| Statistical analyses
Statistical analyses were performed using SPSS 18.0 for Windows (IBM, Inc., Chicago, IL, USA) and PLINK 1.07. Counting data were presented as the mean ± standard deviation. Differences between the two groups were compared using the independent sample T test, while the one-way analysis of variance was performed to compare more than two groups. p <.05 indicated statistical significance.
Associations between genotypes and spontaneous movement were performed using a logistic regression model, and corrections for potential confounders were performed by adjusting for covariates, for example, age, ETAC.

| Single-nucleotide polymorphism selection and genotyping
We selected 13 tagSNPs in GRIN1 and GRIN2A/B for genotyping (

| GRIN2A rs12918566 genomic variation is associated with GRIN2A expression
Bioinformatics analysis indicated that the GRIN2A rs12918566 polymorphism significantly altered GRIN2A mRNA secondary structure, and the free energy of the mutant (G allele) increased (−8.60 kcal/ mol) when compared with the wild type (T allele, −10.10 kcal/mol) ( Figure 1). Furthermore, eQTL analysis in 170 tissues showed that the TG and GG genotypes of rs12918566 displayed lower GRIN2A expression in brain basal ganglia when compared with TT individuals (Figure 2a). This association also existed in other brain sections and the spinal cord (Figure 2b).

| D ISCUSS I ON
Sevoflurane has been clinically used for more than 20 years; however, its precise induction mechanism and general anesthesia maintenance characteristics remain unclear. Glutamate is the most concentrated excitatory neurotransmitter in the brain and spinal cord and mediates postsynaptic membrane effects mainly via F I G U R E 1 The secondary structure of GRIN2A mRNA contains different rs12918566 alleles. (a) GRIN2A structure containing the rs12918566 T allele. The minimum free energy (MFE) of the structure is −10.10 kcal/mol. (b) GRIN2A structure containing the rs12918566 G allele. The MFE of the structure is −8.60 kcal/mol. # Secondary structures and MFEs were predicted using the 50 base pair sequences upstream and downstream of the rs12918566 mutation site F I G U R E 2 Single tissue expression quantitative trait locus analysis for GRIN2A rs12918566. (a) GRIN2A mRNA expression in basal ganglia between different rs12918566 genotypes. (b) GRIN2A rs12918566 polymorphism was significantly associated with GRIN2A expression in basal ganglia, spinal cord, and substantia nigra ionotropic and metabotropic receptors (Reiner & Levitz, 2018). As a class of ionotropic receptors, NMDA receptors have a wide range of functions in the basic synaptic transmission, neuroplasticity, and neurodegenerative diseases of the central nervous system; the domain of the GRIN2 subunit contains a glutamate binding site (Daw et al., 1993;Fang et al., 2014;Lee et al., 2010). This study was the first to investigate the effects of NMDA receptor genetic polymorphisms on spontaneous movement response during sevoflurane anesthesia. We revealed that GRIN2A rs12918566 was significantly associated with sevoflurane-induced spontaneous movement during anesthesia induction.
Transient NMDA receptor activation induces the long-term modification of synaptic connections and the refinement of neuronal circuits and may be the basis of learning and memory (Jeon et al., 2008). In a rat model, intrathecal administration of the NMDA receptor antagonist, MK801, reduced the MAC value of isoflurane in a dose-dependent manner, but this effect was reversed by NMDA receptor agonist administration (Ishizaki et al., 1996(Ishizaki et al., , 1997Stabernack et al., 2003). The MAC value for an inhaled anesthetic is similar to the ED50 (median effective dose) for an intravenous anesthetic; that is, the lowest alveolar inhalation anesthetic concentration at 1 atmosphere when 50% of patients no longer experience body movement during surgical procedures (Hodgson & Liu, 2001). Thus, NMDA appears to have important roles in regulating anesthetic capacity (Rudolph & Antkowiak, 2004).
The GRIN2 (A-D) subunits are the main determinants of receptor functional heterogeneity and exhibit strikingly different spatiotemporal expression (Monyer et al., 1994). The GRIN2A gene encodes the NMDA receptor GRIN2A subunit, which is located on chromosome 16p13.2 and consists of 12 exons and 13 introns (Liu & Zhao, 2013).
A patch-clamp study reported that the inhibitory effects of xenon on NMDA receptor-mediated currents were weakened when transmembrane segment four of GRIN2A or transmembrane segment three of GRIN1 were mutated (Dickinson et al., 2007). GRIN2A is present throughout the dorsal horn except in lamina II at the spinal cord level, while GRIN2B does not appear to exist in lamina II and is restricted to certain areas in the superficial dorsal horn (Boyce et al., 1999;Watanabe et al., 1992). Physiologically, this suggests GRINA2A may play a potential role in immobilization effects because volatile anesthetic-induced immobility results from actions on the spinal cord (Antognini et al., 2002). GRINA2A subunit knockout mice were resistant to the hypnotic effects (loss of righting reflex) produced by ketamine (Sato et al., 2004). Mutation of the transmembrane segment of the GRIN2A subunit (A825W) significantly reduced recombinant NMDA receptor sensitivity to halothane, isoflurane, cyclopropane, and xenon (Ogata et al., 2006). Also, the release of GABA stimulated by NMDA in striatum slices was significantly decreased in GRIN2A knockout mice (Miyamoto et al., 2001).
NMDA receptors are possible targets of inhaled anesthetic components and are involved in the braking of general anesthesia and loss of consciousness (Petrenko et al., 2014).
Our results suggest NMDA is involved in regulating spontaneous movement during sevoflurane induction. Individuals with GRIN2A rs12918566 T alleles were not sensitive to sevoflurane and were more prone to experiencing spontaneous movement during sevoflurane administration.
From in silico predictions using RNAfold, we observed a significant difference in the free energy of the secondary structure, and GTEx database analysis showed that the rs12918566 locus was highly associated with GRIN2A expression in the brain. We speculate the rs12918566 polymorphism regulates GRIN2A gene expression by altering GRIN2A mRNA secondary structure and functioning in the patient's responses to sevoflurane.
In summary, we showed the GRIN2A rs12918566 locus was significantly associated with spontaneous movement during sevoflurane induction in Chinese Han women. Therefore, we speculate the GRIN2A subunit may have important roles in volatile anesthesia.

ACK N OWLED G M ENTS
We thank all patients participating in this study. We are also grateful to all supported funds from the Hunan Province Natural Science

Foundation of China (2019JJ40459) and the National Natural
Science Foundation of China (81971028).

CO N FLI C T O F I NTE R E S T
The authors disclose no conflict.

AUTH O R CO NTR I B UTI O N S
CMH, CF, and NYW analyzed the data and wrote this manuscript.
NYW and MHC aided in processing the data. YHX and YJZ aided in the collection of the materials. NYW helped to revise this manuscript. CF, MHC, and WOY conceived and designed the experiments.

PE E R R E V I E W
The peer review history for this article is available at https://publo ns.com/publo n/10.1002/brb3.2165.