Pathogen-free Sprague–Dawley (SD) postnatal day 7 (P7) rat pups were obtained from the Charles River Laboratories (Wilmington, MA). All experimental protocols were approved by the Boston Children’s Hospital institutional animal care and use committee (Institute of Laboratory Animal Research Commission on Life Sciences NRC, 1996). Ketamine (Ketalar, Bedford Labs, Bedford, OH) and lithium chloride (Sigma-Aldrich, St. Louis, MO) were obtained from commercial sources.

The treatments were conducted in a temperature-controlled acrylic container maintained at 36.7°C. As in previous reports, similar conditions resulted in core body temperatures between 36.5 and 37.5°C (Head et al., 2009). Since sex does not affect neuronal death in anesthetic-induced models of developmental neurotoxicity, we utilized both female and male rat pups for the experiments (Lee et al., 2014). The temporal effect of ketamine was determined in P7 rat pups randomly divided into 5 groups (6 rats per group) receiving 4 intraperitoneal (ip) injections of either ketamine (20 mg · kg⁻¹) or vehicle (saline 10 mL · kg⁻¹) at 90 min intervals over 6 h. Group 1 received 5 consecutive injections of saline (control), group 2: received 2 consecutive doses of ketamine followed by 3 injections of saline, group 3: received 3 consecutive doses of ketamine followed by 2 injections of saline, group 4: received 4 doses of ketamine followed by 1 injection of saline and group 5: received 5 consecutive doses of ketamine. This dosing regimen is similar to that used in previous investigations utilizing a similar experimental paradigm that yields a ketamine plasma concentration of 5.80 ± 3.10 μg⁻¹· ml⁻¹and brain concentration 2.65 ± 1.60 μg⁻¹· g⁻¹ (Zou et al., 2009).

To determine the effect of dosage on DISC1 and pGSK-3β (ser 9) levels, a second cohort of rat pups received ketamine (0, 5, 10, or 20 mg·kg⁻¹ at 90 min intervals over 6 h, 6 rats per group).

To examine the effect of lithium a non-specific blocker of GSK-3β on DISC1 expression, rat pups were randomly assigned into 4 groups (8 rats per group). The rats received 5 ip injections (10 mL · kg⁻¹) of saline, lithium (120 mg · kg⁻¹), lithium (120 mg · kg⁻¹) and ketamine (20 mg · kg⁻¹) or ketamine (20 mg· kg⁻¹) alone, at 90 min intervals over 6 h.

The rat pups were kept from their dam and visually monitored for respiratory effort and activity. After the treatment period, the rat pups were euthanized with pentobarbital (100 mg · kg⁻¹). The brains from each group were rapidly frozen in liquid nitrogen and processed for protein analysis or immersed in 4% paraformaldehyde and embedded in paraffin for histological processing.

Protein was extracted from flash-frozen brain tissue with radioimmunoprecipitation assay buffer (Sigma) containing complete (protease and phosphorylase inhibitor cocktail) and 1 mM phenylmethanesulfonyl fluoride. Protein concentrations were measured by BCA protein assay (Bio-Rad, Hercules, CA). The equal amounts of protein were boiled in sodium dodecyl sulfate loading buffer (Bio-Rad), resolved on 8–12% polyacrylamide denaturing (SDS PAGE) gels and transferred to nitrocellulose (Bio-Rad). Antibodies used for Western blotting included rabbit antibodies to cleaved-caspase-3, total and pGSK-3β (ser 9), DISC1 and β-actin (1:2,000, Cell Signaling, Beverly, MA). 12% SDS PAGE gels were utilized to detect DISC1, due to its relatively high molecular weight of DISC1 of 75 kDa. The blots were washed, and the species-matched peroxidase-conjugated secondary antibody was added. Labeled bands from each blot were detected by enhanced chemiluminescence for visualization and quantitation (Thermo Scientific, Waltham, MA). The densities of the specific protein bands were quantified with an image analysis software (Image J 1.42 NIH, United States).

Brain sections from the ketamine/lithium/saline-treated cohorts were deparaffinized in xylene and rehydrated through graded alcohol. Endogenous peroxidases were inactivated by immersing the sections in hydrogen peroxide for 10 min, and then were incubated for 10 min with 10% normal goat serum to block non-specific binding. These sections were incubated with anti-DISC1 (Invitrogen) and neuron-specific mouse anti-NeuN (neuronal nuclei antigen) (1:100; Abcam, Cambridge, MA) antibodies overnight, followed by incubation for 90 min with the following secondary antibodies: donkey anti-rabbit cy3-conjugate (1:100, Jackson ImmunoResearch) and Streptavidin-Alexa fluor 488 conjugate (1:100, Life Technologies) antibodies. Both primary and secondary antibodies were diluted in phosphate buffer solution (PBS) containing 0.3% Triton x-100, 0.04% bovine serum albumin, and 0.1% sodium azide. Brain slices were rinsed in PBS before mounting on slides. After mounted sections with 90% glycerol, the slides were imaged with fluorescent microscope (Olympus IX81, Olympus, Japan), and the cell counts under magnification of 400× were performed as described previously (Liu et al., 2012). Three sections from each group were examined and the average number of positive cells was recorded in blinded fashion to the treatment group (6 brain per group).

DISC1 and GSK-3β amino acid sequences were obtained from National Center for Biotechnology Information (NCBI) database (NP_001158021 and XP_054202359). To predict DISC1 and GSK-3β interaction, we used Alpha Fold 2 software with auto mode without template (Jumper et al., 2021; Mirdita et al., 2022). We used 5 recycles and recycles early stop tolerance 1.0. The predicted structure was analyzed with pyMol.

Protein lysates extracted from the brains of control and ketamine (20 mg· kg⁻¹ at 90 min intervals over 6 h cohorts, 6 rats per group) were preincubated with protein A/G agarose beads (Thermo Scientific, Rockford, IL) were washed once with RIPA buffer. Solubilized rat brain tissues (1.0 mg of total protein) were pre-incubated with 20 μL of protein A/G agarose for 2 h at 4°C to reduce non-specific binding. Protein samples were incubated with anti-GSK-3β antibody (1:100, Cell signaling technology, Danvers, MA) overnight and 60 μL of protein A/G agarose was added for 2 h at 4°C. The beads were centrifuged and washed four times in the immunoprecipitation (IP) buffer. Samples were boiled in 2 × SDS sample buffer for 5 min at 95–100°C. Samples were loaded 10% SDS-PAGE gel for western blotting analysis. Approximately 50 μg of protein were used as a positive control. The negative control was prepared using the same protocol with omission of the anti-GSK-3β antibody. Proteins were then transferred to a 0.2 μm nitrocellulose membrane for electrophoresis. Non-specific binding sites were blocked by incubating the blots in 5% non-fat dry milk for 1 h at room temperature and incubated with an appropriate primary, anti-DISC1 (Invitrogen, Carlsbad, CA) and anti-GSK-3β antibodies, at 4°C overnight with a gentle rotation. The blots were incubated with anti-rabbit horseradish peroxidase-linked secondary antibodies and enhanced chemiluminescence substrate. The intensity of each protein band was quantified by densitometry using Image J software.

Given the role of DISC1 on neurite outgrowth, we examine the effect of ketamine on dendritic and axonal morphology in primary neuronal cells from cortices of embryonic day 18 (E18) SD rat fetuses. In brief, embryos were delivered from the dams and euthanized under CO₂. The cortices of brain were quickly dissected from embryos in ice-cold Hank’s balance salt solution (HBSS) under the dissection microscope (Leica, Buffalo Grove, IL). Cortical neurons (E18) were isolated from cortices after incubation with 0.25% trypsin–EDTA (ethylenediamine-tetraacetic acid) in HBSS for 30 min at 37°C. Cortical neurons were cultured with neurobasal medium (Invitrogen) containing 2% B27 supplement, 500 μM glutamine and 10 μg/mL gentamycin (Invitrogen) in slide chambers coated with poly-d-lysine. After a 3-day incubation period, the E18 cells treated with control solution, lithium 500 μM, ketamine 100 μM, or ketamine 100 μM combined with lithium 500 μM for 6 h. E18 cells were fixed with 1% paraformaldehyde solution and incubated with anti-NeuN and anti-Neurofilament antibodies (EMD Millipore, Billerica, MA) overnight at 4°C to recognize cellular nuclear and axon of neurons, respectively. The slides were incubated with Streptavidin-Alexa fluor 488 conjugate (1:100, Life Technologies) and donkey anti-rabbit cy3-conjugate (1:100, Jackson ImmunoResearch) antibodies for 90 min. The slides were washed with PBS and mounted with soluble 90% glycerol. Five adjacent slides were examined for each of the treatment groups. The cell images were taken under a fluorescent microscope. The length of axons and dendrites were determined by tracing the lengths of the dendrites and axons of each cell by Image J software in blinded fashion to the experimental group.

In line with our and other laboratories’ practices, a minimum of six rat pups for each experimental group was utilized to detect a 40% difference in the mean values with 80% power at a significance level of 0.05 (Dell et al., 2002). Changes in cleaved-caspase-3 level were presented as percentage of control value. GSK-3β ratios of the phosphorylated and total forms were calculated. Cleaved-caspase-3 positive cells were presented as absolute values per microscopic field. Data were expressed as mean and standard error of the mean (S.E.M.). Normal distribution of the data was assessed by the Shapiro–Wilk test. The differences of the cleaved-caspase-3, pGSK-3β (ser 9), GSK-3β and DISC1 levels and E18 cell viability were analyzed with one-way ANOVA, followed by a post hoc Newman–Keuls multiple comparison test. Since the data from the immunofluorescence assay for cleaved-caspase-3 and NeuN had a skewed distribution, a Kruskal-Wallis’s test was applied to analyze differences, followed by Dunn test for multiple comparisons. Data analyses were generated, and plots were constructed using Prism 8 (GraphPad Software, La Jolla, CA).

In order to determine the impact of ketamine neuroapoptosis and DISC1 expression in the brain, the expression of cleaved-caspase-3, pGSK-3β (ser9) and DISC1 was examined by Western blot and immunohistochemistry. Ketamine significantly increased cleaved-caspase-3 and decreased DISC1 expression in p7 brain tissues after 3 h exposure in a time-dependent manner (p < 0.05, Figures 1A,B). Immunofluorescence microscopy of brain sections stained with antibodies to cleaved-caspase-3 (red) and NeuN, a neuron-specific nuclear protein (green) antibodies confirmed that ketamine significantly increased apoptosis and decreased DISC1 expression in neurons (Figures 2A,B). Increasing the interval dose of ketamine also significantly decreased DISC1 and pGSK-3β expression (p < 0.05 or p < 0.01) (Figures 3A,B). The dose dependent reduction of pGSK-3β expression is consistent with our previous report and indicate an increase in GSK-3β activity (Mao et al., 2009).

To examine the effect of ketamine on the DISC1/GSK-3β binding, lithium, a non-specific inhibitor of GSK-3β, was used to treat P7 rat with or without ketamine. Ketamine at dose of 20 mg · kg⁻¹ significantly decreased the expression of DISC1 protein in p7 brain tissues. The addition of lithium significantly attenuated ketamine-induced decreases DISC1 expression (p < 0.01) (Figure 4). Interestingly, lithium increased DISC1 protein expression in the absence of ketamine, which suggest that lithium may displace DISC1 from binding sites on GSK-3β.

Based on the several reports, we hypothesized that DISC1 and GSK-3β would directly bind to each other (Mao et al., 2009; Lipina et al., 2011). Using AlphaFold2, we predicted structural relationship between DISC1 and GSK-3β. Our docking suggested DISC1 structurally interacted with GSK-3β (Figure 5).

DISC1 has been shown to regulate GSK-3 β activity by directly binding to multiple domains on GSK-3β. To determine the effect of ketamine on DISC1-GSK-3β binding, protein lysates extracted from the brains of control and ketamine treated rats were subjected to co-immunoprecipitation assay. Ketamine induced a 40 and 60% reduction in absolute DISC1 expression standard western blots (Figures 1B, 3A). The experiment yielded a 90% reduction in bound DISC1 (Figure 6), which suggest a reduction in bound DISC1, which is lower than predicted by the standard DISC1 immunoblots. Therefore, the structural prediction of DISC1-GSK-3β interaction using Alpha Fold 2 software corroborates with our co-immunoprecipitation results.

DISC1 is an important regulator of dendrite formation and function, and axonal extension (Miyoshi et al., 2003). To examine dendrite formation and axonal extension in this experimental paradigm, E18 cells were treated ketamine with or without lithium for 6 h. Ketamine decreased dendritic and axonal length of E18 cells (p < 0.05) (Figure 7). Lithium, a non-specific inhibitor of GSK-3β, significantly mitigated the effects of ketamine on morphological changes in E18 cells. In addition, lithium itself did not change the morphology of E18 cells. Taken together, these results suggest that ketamine-induced reduction of dendritic and axonal growth are modulated by GSK-3β-DISC1 binding.