|Home | About | Journals | Submit | Contact Us | Français|
Growing evidence indicates that ketamine causes neurotoxicity in a variety of developing animal models, leading to a serious concern regarding the safety of pediatric anesthesia. However, if and how ketamine induces human neural cell toxicity is unknown. Recapitulation of neurogenesis from human embryonic stem cells (hESCs) in vitro allows investigation of the toxic effects of ketamine on neural stem cells (NSCs) and developing neurons which is impossible to perform in humans. In the present study we assessed the influence of ketamine on the hESC-derived NSCs and neurons.
hESCs were directly differentiated into neurons via NSCs. NSCs and two-week-old neurons were treated with varying doses of ketamine for different durations. NSC proliferation capacity was analyzed by Ki67 immunofluorescence staining and bromodeoxyurindine assay. Neuroapoptosis was analyzed by TUNEL staining and caspase 3 activity measurement. The mitochondria-related neuronal apoptosis pathway including mitochondrial membrane potential, cytochrome c distribution within cells, mitochondrial fission, and reactive oxygen species (ROS) production were also investigated.
Ketamine (100 μM) increased NSC proliferation after 6 h exposure. However, significant neuronal apoptosis was only observed after 24 h of ketamine treatment. In addition, ketamine decreased mitochondrial membrane potential and increased cytochrome c release from mitochondria into cytosol. Ketamine also enhanced mitochondrial fission as well as ROS production compared with no-treatment control. Importantly, Trolox, a ROS scavenger, significantly attenuated the increase of ketamine-induced ROS production and neuronal apoptosis.
These data for the first time demonstrate that (1) ketamine increases NSC proliferation and causes neuronal apoptosis; (2) mitochondria are involved in ketamine-induced neuronal toxicity which can be prevented by Trolox; and (3) the stem cell-associated neurogenesis system may provide a simple and promising in vitro model for rapidly screening anesthetic neurotoxicity and studying the underlying mechanisms as well as prevention strategies to avoid this toxic effect.
Growing evidence suggests that prolonged exposure of developing animals during brain growth spurt to general anesthetics induces widespread neuronal cell death followed by long-term memory and learning abnormalities.1–4 Ketamine is widely used in pediatric anesthesia to provide sedation/analgesia to children for painful procedures.5 In addition, ketamine is one of the most studied anesthetics for addressing neurotoxicity issues in both rodent and primate models. For instance, ketamine administered subcutaneously to 7-day-old mice for 5 h resulted in a significant increase in neuronal cell death.6 Intravenous administration of ketamine for 24 h caused an increase of cell death in the cortex of rhesus monkeys at 122 days of gestation and postnatal day 5.4In vitro experimental evidence from cultured neonatal animal neurons confirmed the in vivo findings.7–14 Apoptosis was involved in anesthetic-induced neuronal cell death.15–18 However, the mechanistic details by which anesthetics induce neurotoxicity have yet to be established.
So far, there is no direct evidence showing any such toxic effect in human neonates and infants at any dose of anesthetics. Every year millions of children are exposed to a variety of anesthetics. The findings from animal-related studies lead to a serious concern regarding the safety of pediatric anesthesia and raise a real question whether similar neuroapoptosis also occurs in the developing human brain. However, it is almost impossible to obtain neonatal human neurons to study anesthetic neurotoxicity. In addition, many developmental events, including neural stem cell (NSC) proliferation, neurogenesis, and synaptogenesis, occur during the brain growth spurt. Therefore, neuroapoptosis may not be the only variable to be considered in evaluating potential adverse effects of anesthetics on neuronal development. Thus, it is imperative to find a good model to study anesthetic-induced developmental neurotoxicity in human neuronal cells.
Human embryonic stem cells (hESCs) are pluripotent stem cells and can replicate indefinitely and differentiate into various cell types. Differentiation ability of hESCs into committed cell types is potentially valuable for studying cellular and molecular events involved in early human development under physiological and pathological conditions which is almost impossible to perform in humans.19–22 In the present study, we used hESCs to recapitulate neurogenesis in vitro following the principles of neural development and obtained human NSCs and neurons by culturing hESCs in chemically defined medium. Using this in vitro hESC-related neurogenesis model, we then studied the ketamine-induced toxicity in NSCs and neurons. We hypothesized that ketamine interfered with the proliferation of hESC-derived NSCs and induced apoptosis in the human neurons differentiated from NSCs via reactive oxygen species (ROS) and mitochondrial pathway.
Mitotically inactivated mouse embryonic fibroblasts (MEFs) by mitomycin C (Sigma) were used as feeder cells to support the growth and maintenance of hESCs (H1 cell line, WiCell Research Institute Inc.). Inactivated MEFs were plated in 0.1% gelatin-coated 60 mm culture Petri dishes containing Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (Gibco) in a humidified incubator under normoxic condition (20% O2 /5% CO2) at 37°C. On the following day, hESCs were plated on MEFs with hESC culture medium and incubated in a hypoxic incubator (4% O2/5% CO2). hESC culture medium consisted of DMEM/F12 supplemented with 20% KnockoutTM serum replacement (Gibco), 1% non-essential amino acids, 1% penicillin-streptomycin, 1 mM L-glutamine (Chemicon), 0.1 mM ß-mercaptoethanol (Sigma), and 4 ng/mL human recombinant basic fibroblast growth factor (bFGF; Invitrogen). The medium was changed daily. hESCs were passaged every 5–7 days using a mechanical microdissection method. hESCs with passage numbers between 70 and 80 were used in this study.
hESCs underwent four-step progression that includes embryonic body (EB) culture, rosette cell formation, NSC expansion, and neuronal differentiation as illustrated in Figure 1A and described as follows: (1) EB culture. hESCs in the culture were digested using dispase (1.5 unit/mL) (Invitrogen) for 30 min. Digested hESCs were then transferred to 60 mm ultra-low-attachment dishes (Corning) and cultured in hESC medium without bFGF under normoxic conditions. The medium was changed every day. EBs were visible one day after culturing. Four days later, EBs were switched to neural induction medium consisting of DMEM/F12 supplemented with 1% N2 (Invitrogen), 1% non-essential amino acids, 5 ng/mL bFGF, and 1mg/mL heparin (Sigma) for 4 days. (2) Rosette formation. EBs were transferred to matrigel-coated 60 mm culture dishes and cultured with neural induction medium at day 8. The medium was changed every other day. EBs attached on the dishes and formed neural tube-like rosettes with radial arrangements of columnar cells within 5 days. (3) NSC expansion. Two days after rosette formation, rosette cells were gently blown off with a 5 ml serological and then transferred to other dishes containing DMEM/F12 supplemented with 1% N2, 2% B27 (Invitrogen), 1% non-essential amino acids, 20 ng/mL bFGF, and 1mg/mL heparin. One day later, rosette cells rolled up to form round spheres called NSCs. Half of the medium was changed every other day. NSCs were passaged every 5–6 days by digestion with accutase (Innovative Cell Technology). (4) Neuronal differentiation. Half a million NSCs were cultured in 60 mm matrigel-coated dishes with neuronal differentiation medium consisting of Neurobasal, 2% B27 (Invitrogen), 0.1 μM cyclic adenosine monophosphate, 100 ng/mL ascorbic acid (Sigma), 10 ng/mL brain-derived neurotrophic factor, 10 ng/mL glial cell-derived neurotrophic factor, and 10 ng/mL insulin-like growth factor 1 (Pepro Tech Inc.). The medium was changed every other day. Two-week-old neurons differentiated from NSCs (passages 6 to 12) were used in this study.
NSCs were cultured in either 60 mm dishes (with or without 12 mm coverslips) (5 × 105 cells/dish) or 96-well tissue culture plates (1 × 104 cells/well) with neuronal differentiation medium. NSCs or 2-week old differentiated neurons were incubated with or without ketamine (range 0 to 100 μM) (Phoenix Pharmaceutical Inc.) and ROS scavenger Trolox for 3, 6, or 24 h. Cell proliferation, cell viability, apoptosis, mitochondrial membrane potential (ΔΨm), distribution of cytochrome c in cells, mitochondrial fission, or ROS measurement were performed immediately after ketamine exposure for indicated durations. Although it was reported that peak blood levels of ketamine could be as high as 103 μM,23 the levels required to maintain anesthesia are usually in the range of 10–20 μM.24 Experimental evidence provided from in vitro cell culture and in vivo animal studies demonstrated that ketamine could induce neurotoxicity when administered at high doses and/or for prolonged periods.25–29 Thus, in this study, we treated neurons with ketamine (100 μM) to study the effect of ketamine on the NSC proliferation and neuronal apoptosis for 3, 6, and 24 h. The underlying mechanisms of this side effect were investigated by exposing neurons to ketamine for 24 h.
Cells cultured on matrigel-coated coverslips were fixed with 1% paraformaldehyde for 30 min. Cells were then washed three times with phosphate buffered saline (PBS) alone or PBS containing 0.5% Triton X-100 (Sigma) and blocked with 10% donkey serum for 30 min at room temperature. Next, cells were incubated with primary antibodies in a moist chamber for 1 h at 37°C. The primary antibodies were mouse anti-stage-specific embryonic antigen 4 (SSEA-4), microtubule-associated protein 2 (MAP2), β-tubulin III (abcam), cytochrome c (BD Pharmingen), Homer 1 (Synaptic Systems), and rabbit anti-Oct-4, nestin, Ki67, synapsin-1, and PAX6. After three wash cycles, cells were incubated with Alexa Fluor 485 (or 594) donkey anti-mouse (or rabbit) immunoglobulin (IgG) (Invitrogen) for 1 h at room temperature. Cell nuclei were stained with TO-PRO®-3 (Invitrogen). The coverslips were then mounted onto the slides and observed under a laser-scanning confocal microscope (Nikon Eclipse TE2000-U). Neuronal differentiation efficiency was determined by calculating the percentage of MAP2-positive cells over the total cells. Results were obtained from at least 100 cells in each of three independently differentiated neuron samples.
The influence of ketamine on NSC proliferation was analyzed by two methods: Ki67 staining and bromodeoxyurindine (BrdU) assay. (1) Ki67 is a nuclear non-histone protein and preferentially expressed during late G1, S, G2, and M phase of the cell cycle; while resting, noncycling cells (G0 phase) lack Ki67 expression. NSCs (4×104 per 12 mm coverslip) were cultured in NSC medium with or without 100 μM ketamine for 3 and 6 h. Ki67 expression in NSCs was visualized using immunofluorescence staining. Ki67-positive stem cells were counted from 6 random fields per coverslip in each of three independently differentiated NSC samples. (2) BrdU incorporation as an indicator of cell proliferation was studied using a colorimetric BrdU kit (Roche Diagnostics) according to the manufacturer's instructions. BrdU is a thymidine analog and can be incorporated into newly synthesized DNA strands of actively proliferating cells. The incorporation of BrdU into cellular DNA is then detected using anti-BrdU antibody, allowing assessment of cell proliferation rate. NSCs (1×104 per well) were grown in 96-well plates containing NSC medium with or without 100 μM ketamine with all media being supplemented with 10 μM BrdU for either 3 or 6 h. Cells were then fixed with FixDenat solution for 30 min at room temperature followed by incubation with anti-BrdU antibody conjugated with peroxidase for 30 min. Cells were washed 3 times for 5 min each and 100 μl substrate solution was then added to the wells. Fifteen minutes later, the absorbance at 450 nm (representing proliferating potential) was measured using a microplate reader.
Neurons cultured on matrigel-coated coverslips from each experimental condition were fixed with 2% glutaraldehyde in 0.1M sodium cacodylate buffer at 4°C, washed in the same buffer, and post-fixed with 1% osmium tetroxide for 60 min on ice. Cells were then washed briefly in distilled water and dehydrated through graded methanol (50%, 20 min; 70%, 20 min; 95%, 20 min; 100%, 3 × 20 min) and acetonitrile (2×10 min). The cells were infiltrated with epoxy resin (EMbed-812; Electron Microscopy Sciences, Hatfield, PA), and polymerized overnight at 70°C. Ultra-thin sections (~60 nm) were cut, stained with uranyl acetate and lead citrate according to standard procedures and viewed with a Hitachi H600 EM.
Caspase 3 activity was measured using Caspase 3 Colorimetric Assay kit (Genescript) following the manufacturer's protocol. Neurons cultured in 60 mm dishes were washed twice with PBS and lysed with 50 μL lysis buffer followed by adding 50 μL reaction buffer as well as 5 μL substrate. After 4 h incubation at 37°C, caspase 3 activity was measured using a microtiter plate reader at 400 nm. The values of optical density at 400 nm were normalized to total protein content of the samples as determined by DC Protein Assay Reagents Package kit (Bio-Rad).
DNA fission associated with apoptosis was analyzed using terminal deoxynucleotidyl transferase-mediated deoxyuridine triphosphate in situ nick end labeling (TUNEL) detection kit (Roche) following the manufacturer's instruction. The cells cultured on coverslips were rinsed with PBS, fixed with ice-cold 1% paraformaldehyde. Terminal deoxynucleotidyl transferase, a template-independent polymerase, was used to incorporate nucleotides at the sites of DNA breaks. Nuclei were stained with TO-PRO®-3. Fluorescent images were taken from three different fields on each coverslip under the confocol microscope. Apoptotic index was calculated as a percentage according to the following formula: TUNEL-positive nuclei number / the number of total cell nuclei.
Tetramethylrhodamine ethyl ester (TMRE) is a lipophilic positively charged dye and can penetrate living cells and enter the negatively charged mitochondria where it accumulates in an inner-membrane potential-dependent manner. When the ΔΨm collapses in apoptotic cells, TMRE no longer accumulates inside the mitochondria and becomes more evenly distributed throughout the cytosol. When dispersed in this manner, overall cellular fluorescence levels decrease dramatically and this event can easily be visualized by fluorescence microscopy. The wavelengths of excitation (λex) and emission (λem) for TMRE are 548 nm and 573 nm, respectively. Neurons cultured on glass coverslips were loaded with 50 nM TMRE (Invitrogen) for 20 min at room temperature. TMRE fluorescence intensity representing ΔΨm was recorded under the confocol microscopy with a 60×1.4 numerical aperture oil immersion objective (Nikon). Imaging conditions such as gain levels (115) and confocal aperture size (90 μm) were held constant. Images were taken from six random fields per coverslip. Data were analyzed by ImageJ software 1.41 (Wayne Rasband).
Ten days after NSCs were cultured in neuronal differentiation medium, differentiated neurons were transduced with the virus CellLight™ mitochondria-GFP (green fluorescence protein) (Invitrogen) for 24 h to label the mitochondria according to the protocol provided by the company. This fluorescent protein-based reagent contains the leader sequence of E1-alpha pyruvate dehydrogenase fused to emerald GFP. Transduced neurons expressed GFP within mitochondria. Four days later, the labeled neurons were then used for the analysis of the effect of ketamine on the cytochrome c translocation and mitochondrial fission. Transduction efficiency was calculated as the ratio between GFP-positive cells and total cells. GFP expression in the mitochondria was confirmed by the colocalization of GFP (green) and TMRE fluorescence (red) signals.
The distribution of cytochrome c in the neurons expressing GFP within mitochondria was analyzed with antibody against cytochrome c (abcam) using immufluorescence staining.
The labeled neurons (expressing GFP within mitochondria) treated with or without 100 μM ketamine for 24 h were imaged with the confocal microscope. Using Image J 1.41o software (Wayne Rasband; National Institutes of Health, USA), the fluorescent images were then converted to binary (black and white) images. Mitochondrial particles were analyzed for length, width, area, and perimeter. We then identified two factors of every single mitochondria of the investigated cells (43 cells per group). These two factors were used to score mitochondrial morphology to evaluate the status of mitochondrial fission and fusion as previously described:30,31 aspect ratio (AR) and form factor (FF). AR, the ratio between the major and the minor axes of the ellipse, represents the length of the mitochondria. FF (perimeter2/4π*area) indicates the degree of mitochondrial branching.32 Both parameters having a minimal value of 1 represent a perfect circle. Mitochondria in healthy neurons have high values of AR and FF representing the elongated tubular mitochondria and the increase of mitochondrial complexity (length and branching), respectively. Three investigators were responsible for performing the experiments, acquiring the images, and analyzing the data independently. Two investigators were responsible for performing the experiments and acquiring the images; and the third (blinded) individual analyzed the data.
Cellular ROS were assayed using both 5-(and-6)-chloromethyl-2′, 7'-dichlorodihydrofluorescein diacetate, acetyl ester (CM-H2DCFDA) and MitoSOX red (Invitrogen). CM-H2DCFDA, a ROS-sensitive membrane permeable fluorescent probe, has been used extensively to measure cellular ROS (e.g, H2O2), but it does not measure superoxide directly. Upon penetration into the cells, the acetate groups of CM-H2DCFDA are cleaved by cytoplasmic esterases to form membrane impermeable non-fluorescent CM-H2DCF trapped within the cells. CM-H2DCF is then rapidly oxidized by ROS into a fluorescent dye dichlorodihydrofluorescein (DCF). The increase of DCF fluorescence serves as a measure of ROS generation. MitoSOX red is a selective indicator of mitochondrial superoxide which is catalyzed to form many ROS members like H2O2 and OH˙− radicals by superoxide dismutase enzymes and also by autodismutation reactions. MitoSOX™ red is a poorly fluorescent compound carrying a charge that results in the selective accumulation of this probe within the mitochondria. After reaction with the superoxide anion, oxidized MitoSOX™ red produces DNA sensitive fluorochromes that generate a red fluorescence. (A) CM-H2DCFDA assay. Neurons were loaded with 10 μM CM-H2DCFDA for 30 min in the presence/absence of ketamine with or without Trolox followed by 20-min washout. The neuron-contained coverslips were placed in a polycarbonate recording chamber (Warner Instruments) on the stage of confocal microscope. DCF fluorescence (λex/λem=495/520 nm) was visualized using the confocol microscope within 8 min. Imaging conditions for ROS data acquisition such as gain levels (120) and confocal aperture size (90 μm) were held constant between different groups. DCF intensity was then quantified with ImageJ software 1.41. Results were obtained from at least 30 neurons in each of three experiments per group. (B) MitoSOX™ red assay.33,34 Neurons were treated with or without 100 μM ketamine for 24 h. Twenty-four h later, cells were labeled with 5 μM MitoSOX red in the neuronal differentiation medium for 10 min at 37°C. After cells were washed twice at room temperature, MitoSOX red fluorescence in the cells was recorded under the confocol microscope within 8 min. The excitation and emission wavelength are λex/λem = 510/580 nm for MitoSOX red. Imaging conditions for MitoSOX red data acquisition such as gain levels (100) and confocal aperture size (90 μm) were held constant. Results were obtained from at least 30 neurons in each of three experiments per group.
Previous studies showed that ketamine significantly decreased neuronal death based on the data that were obtained from n=3.11,15 The study had more than n=3 when multiple groups/conditions were considered. For instance, when 2 to 6 groups were considered, there were 6 to 18 samples examined. In addition, we examined the normality of residuals from the presented data by drawing a QQ-plot of residuals using SPSS software version 15. No substantial deviations from the expected straight line pattern were found in the QQ-plots. This simultaneously evaluated the assumptions of normality and variance homogeneity. Additionally, Levene's test was used to formally assess the homogeneity of variance. With reliance on these prior studies and the residual analysis, we collected data from three independent neuronal differentiations. Reported values were expressed as the means ± standard errors of the means showing normal distribution. The statistically significant differences of raw data among groups were tested by using one-way analysis of variance using SPSS software. Tukey's post hoc test was used for pairwise comparisons, and only adjusted p-values are reported. A level of P<0.05 was considered to be statistically significant.
hESCs were induced to differentiate into neurons in vitro via a four-step progression as shown in Figure 1A. Cells at different differentiation stages demonstrated different morphology and expressed different cell-specific markers. hESCs grew as colonies while cultured on MEFs and stained positive for pluripotent stem cell markers Oct-4 and SSEA-4 (Fig. 1B-a to c). hESCs were then directly differentiated into neural rosettes which were radial arrangements of columnar NSCs. Rosette cells expressed neuroepithelial cell marker PAX6 (Fig. 1B-d to e). In the rosette culture, cells were heterogeneous. Rosettes were surrounded by other non-organized cells. The radially organized rosette cells were then mechanically isolated under the microscope and cultured on matrigel-coated dishes. These enriched NSCs showed triangle-like morphology and had strong proliferation potential. NSCs were confirmed by the expression of both NSC marker nestin and proliferating marker Ki67 (Fig. 1B-f to h).
Obvious neuronal differentiation could be observed in the culture 3 days after NSCs were cultured in neuronal differentiation medium. Differentiated neurons exhibited round cell shape with small projections. Cell projections extended and the extensive neuron networks were further formed in the culture over time. Two-week-old neurons expressed neuron-specific markers β-tubulin III and MAP2 (Figs. 1C-a to c). In addition, they expressed synaptic marker synapsin-1 (Fig. 1C-d) which is exclusively localized in the regions occupied by synaptic vesicles.35 Differentiated neurons also expressed postsynaptic protein Homer 1 (a family of postsynaptic scaffolding proteins) (Fig. 1C-e).36 There were approximately 92% of NSCs differentiating into neurons within 2 weeks of protocol initiation.
NSCs were treated with 100 μM ketamine for 3 and 6 h followed by Ki67 immunofluorescence staining and BrdU assays. There were more Ki67-positive cells in the 6 h-ketamine-treated culture than in the control and 3 h-ketamine culture (Figs. 2A and B) (P=0.016 vs. 6 h no-treatment control, n=3). The influence of ketamine on the proliferation of NSCs was further analyzed by BrdU assay. Ketamine (100 μM) increased NSC proliferation after 6 h of exposure (P=0.026 vs. 6 h no-treatment control, n=3) (Fig. 2C). There were no significant differences of Ki67-positive cells and BrdU incorporation between 3 and 6 h control groups. Caspase 3 activity assay showed that ketamine treatment for up to 24 h did not induce NSC apoptosis (Fig. 2D).
The effect of ketamine on the ultrastructure in the neurons was analyzed using electron microscope. As shown in Figure 3, the neurons without ketamine treatment had very elongated mitochondria. Other organelles such as Golgi apparatus appeared regular. In the 24 h of 100 μM ketamine-treated culture, clear signs of the toxic effect on the cellular ultrastructure were observed and described as follows: Large oval-shaped autophagosomes having some membrane-like material within them were found in almost every cell occupying the majority of the cytosol volume. Other notable observations were the fragmented mitochondria and the disappearance of Golgi.
In order to investigate if ketamine induces the occurrences of apoptosis in neurons, we measured caspase 3 activity and examined DNA damage by TUNEL staining. When compared with no-treatment, 24 h of 100 μM ketamine exposure significantly increased caspase 3 activity (P=0.016, n=3) (Fig. 4A). Consistent with caspase 3 activation, more TUNEL-positive cells were observed in ketamine-treated cultures (Fig. 4B). Most cells stained positive for TUNEL were located in condensed nuclei, a hallmark of apoptosis. There were 19 ± 2% apoptotic cells after treatments with ketamine. In control, some spontaneous programmed cell death resulted in 12 ± 0.4% TUNEL-positive cells (Fig. 4C).
Given that ketamine may cause mitochondrial damage, we then measured ΔΨm and distribution of cyochrome c in the cells. As shown in Figure 5A, treatment of differentiated neurons with 100 μM ketamine for 24 h significantly decreased ΔΨm (P=0.030, n=3). To investigate the distribution of cytochrome c within the cells, the neurons were transduced with the virus CellLight™ mitochondria-GFP (green). The expression of GFP within mitochondria was confirmed by the colocalization of GFP and mitochondrial probe TMRE signals (red) (Fig. 5B). Transduction efficiency (GFP-positive cells) was 40% (Fig. 5C). The distribution of cytochrome c was examined using immunofluorescence staining. Figure 5D showed that cytochrome c (red fluorescence) was located within mitochondria in the control culture. However, in the ketamine-treated cells, cytochrome c was released from the mitochondria into cytosol.
We next examined whether ketamine altered the mitochondrial morphology. Labeled neurons expressed GFP within mitochondria, indicating mitochondrial morphology. In the control culture, mitochondria shape was mostly tubular (Figs. 6A-a and c). After treatment with 100 μM ketamine for 24 h, much shorter and smaller mitochondria were prevalent in the cells (Figs. 6A-b and d). Mitochondrial shapes were further analyzed by measuring the AR and FF. Most mitochondria in the 24 h of 100 μM ketamine-treated cells had lower values of FF (P=0.016, n=3) and AR (P=0.00020, n=3) compared to the control, indicating mitochondrial fission (Figs. 6B and C).
ROS production was determined by DCF fluorescence in the neurons with or without ketamine treatment. There was an increase of ROS production in the neurons compared with control (P=0.000059 vs. control, n=3). Trolox, a ROS scavenger, significantly attenuated ketamine-induced ROS production (P=0.000079 vs. ketamine group, n=3) (Figs. 7A and B). Next, we examined if the increased ROS was produced within mitochondria. The labeled neurons were then treated with or without 100 μM ketamine. Twenty four hours later, the cells were loaded with MitoSOX™ red. As shown in Figures 7C and D, ketamine increased superoxide production in the mitochondria (P=0.014 vs. control, n=3). Trolox significantly decreased ketamin-mediated superoxide production in mitochondria (P=0.042 vs. ketamine + Trolox group, n=3).
To determine if ROS production mediates the induction of neuroapoptosis, differentiated neurons were treated with 100 μM ketamine with or without Trolox for 24 h. Trolox (250 μM) significantly reversed ketamine-induced neuronal apoptosis (TUNEL-positive cells: 12 ± 0.4, 19 ± 2, and 12 ± 1% in control, ketamine, and ketamine plus Trolox groups, respectively, n=3) (Figs. 7E and F).
In the current study, we investigated for the first time the ketamine-induced toxicity in human NSCs and neurons using an in vitro hESC-related neurogenesis model. We provided novel findings on several points: (1) ketamine increased human NSC proliferation; (2) ketamine caused neuronal apoptosis via a mitochondrial pathway; (3) ketamine-induced intracellular ROS and mitochondrial superoxide productions that were accompanied by the ultrastructural abnormalities such as increased mitochondrial fission; and (4) ketamine-induced neuroapoptosis could be attenuated by the ROS scavenger Trolox.
The greatest vulnerability of developing brain to anesthetics occurs at the time of rapid synaptogenesis or so called brain growth spurt period. In rodents, this critical period lasts from approximately postnatal days 0 to 10. For rhesus monkeys, this period takes place after approximately 115 days of gestation and up to postnatal day 60. In humans, it starts from the third trimester and continues 2 to 3 years after birth.37–39 However, it is unclear if anesthetic-induced neuronal death is the direct cause of long-term neurocognitive dysfunction. Many developmental events (e.g., neurogenesis, synaptogenesis, and neuron structure remodeling) occur within this period. Multiple sequential steps are involved in neurogenesis including NSC proliferation and the differentiation of NSCs into neurons. Thus, we have considered that neuronal cell death may not be the only consequence caused by general anesthetics. Anesthetics may perturbate neural development by influencing NSC proliferation, neurogenesis, synaptogenesis, and neuronal survival.40–42 In the present study, we applied hESCs to establish an in vitro controlled neurogenesis system (Fig. 1). There are two main advantages of this experimental model used for studying anesthetic-induced developmental neurotoxicity: (1) mimicking neural developmental principles to obtain unlimited human NSCs and neurons allows the dissection of the toxic effect of anesthetics on the NSCs and neurons and the underlying mechanisms, and (2) rapidly screening the effect of various anesthetics and conditions (e.g., dose and time) on neurotoxicity.
Our current results showed that ketamine exerted potentially toxic effects on both NSCs and neurons (Figs. 2 and and3).3). Shorter-term exposure (6 h) caused NSC proliferation. However, prolonged ketamine exposure led to a significant increase in apoptosis only in neurons but not in NSCs (Figs. 2D and and4A),4A), suggesting that NSCs are more resistant to cell death than neurons. The ketamine-induced NSC proliferation was also observed in rats.43 Ketamine acts through blockade of N-methyl-D-aspartate receptors (NMDAR).11 D(−)-2-amino-5-phosphonopentanoic acid, an NMDAR antagonist like ketamine was also reported to increase NSC proliferation in both ventricular zone in neocortical slices and in primary cultures.44 However, a recent in vitro study showed that ketamine decreased rat cortical NSC proliferation.45 The varying effects of ketamine on NSC proliferation may be due to the different models used. Abnormal proliferation of NSCs could influence neurodevelopment. For instance, the total weight of the brain was found to be increased in autistic boys.46 Ethanol, an NMDA antagonist, has long been recognized to be neurotoxic to the developing brain. Exposure to ethanol during brain development might promote neurodevelopmental defects.47,48 In one recent study, Nash et al. used a similar in vitro hESC-based neurogenesis system to study ethanol-induced early developmental toxicity. They found that ethanol induced a complex mix of phenotypic changes, including an inappropriate increase in stem cell proliferation and loss of trophic astrocytes.49 Thus, ketamine-induced alterations in NSC proliferation might contribute to an abnormal brain development.
Neuronal apoptosis is commonly recognized as a harmful effect of anesthetics.12,25,50,51 However, how anesthetics cause neuronal apoptosis is not well understood. The central components of the programmed cell death are a group of proteolytic enzymes called caspases which can be activated by various types of stimulation. Loss of ΔΨm and release of cytochrome c from the mitochondria are key events in initiating mitochondria-involved apoptosis.52 The released cytochrome c activates caspase 9, which consequently induces caspase 3 activation, resulting in the cleavage of several cellular proteins, finally leading to the typical alterations related to cell apoptosis such as DNA fragmentation in cell nuclei.17,25 In this study, there was a significant increase in the caspase 3 activity as well as TUNEL-positive cells with condensed and fragmented nuclei after 24 h ketamine exposure (Fig. 4). This result was largely in agreement with several previous studies showing that only prolonged exposure could induce neuron death. For instance, treatment with 100 μM of ketamine for 48 h resulted in the loss of 45% of neurons from fetal rats after 18–19 days gestation.12 In addition, ketamine-induced neuronal apoptosis was accompanied by the significant decrease in ΔΨm and an increase in cytochrome c release from mitochondria into cytosol (Fig. 5), suggesting that ketamine might induce neuron apoptosis via mitochondrial pathway.
Mitochondria are highly dynamic and the alterations in mitochondrial shape (fusion or fission) can affect a variety of biological processes such as apoptosis and mitosis.53 Inhibition of mitochondrial fission was shown to prevent mitochondrial cytochrome c release and apoptotic cell death.54 Our data showed the fragmented discrete punctiform mitochondria in the ketamine-treated culture while mitochondria shape was elongated and interconnected to the network in the control cells (Fig. 6). In addition, in response to ketamine, there was a significant increase in ROS production. Superoxide within mitochondria was also high in the ketamine-treated cells, indicating the mitochondrial origin of ROS. Trolox, a ROS scavenger, prevented ketamine-induced ROS production and cell apoptosis (Fig. 7). ROS are involved in the regulation of many physiological processes. However, overproduction of ROS under various cellular stresses can cause cell death. Mitochondria are a main source of ROS, and the primary ROS is superoxide, which is converted to H2O2 either by spontaneous dismutation or by the enzyme superoxide dismutase. H2O2 can be further transformed to OH.55 Recent studies suggested that accumulation of ROS was associated with anesthetic-induced mitochondrial damage.17,56 Application of antioxidant 7-nitroindazole (nitric oxide synthase inhibitor) provided protection against ketamine-induced neuronal cell death.57 In addition, ROS was implicated in the mitochondrial fission process. Inhibition of mitochondrial fission prevented high glucose-induced ROS production.58 Thus, mitochondrial fission may mediate the ketamine-induced neuronal apoptosis by increasing ROS production. The experimental approach used here, however, does not allow for the resolution of the complete sequence of events regarding how ketamine induces neurotoxicity. Future experiments will dissect the cooperation of mitochondria fission and ROS production in the ketamine-induced neuronal apoptosis.
In conclusion, we used the in vitro hESC-related neurogenesis model to study ketamine-induced human neurotoxicity, demonstrating for the first time that ketamine exerts its influence on human NSC proliferation and neuronal viability via mitochondria pathway. This stem cell-related neurogenesis system might provide a simple and promising in vitro model for rapid screening of anesthetic-induced neurotoxicity and studying the underlying mechanisms as well as prevention strategies to avoid these toxic effects. In addition, given the toxic effect on the neurons seen in the culture treated with 100 μM ketamine for 24 hours, short-term low dose ketamine might be a safer strategy for brief procedure or other sedation in human children.
The authors thank Aniko Szabo, Ph.D. (Associate Professor, Institute for Health and Society, Division of Biostatistics, Medical College of Wisconsin) for help with the statistical analysis.
Funding: This work was supported in part by P01GM066730, R01HL034708 from the NIH, Bethesda, MD, and by FP00003109 from Advancing a Healthier Wisconsin Research and Education Initiative Fund (to Dr. Bosnjak).
The authors declare no conflicts of interest.
DISCLOSURES: Name: Xiaowen Bai, MD, PhD
Contribution: This author helped conceptualize the study, design the study, acquire data and images, analyze and interpret the data, and write the manuscript.
Attestation: Xiaowen Bai attests to having approved the final manuscript, approved the final manuscript. Xiaowen Bai attests to the integrity of the original data and the analysis reported in this manuscript. Xiaowen Bai is the archival author
Name: Yasheng Yan, BS
Contribution: This author helped perform the experiments, acquire the data, and analyze and interpret the data.
Attestation: Yasheng Yan attests to having approved the final manuscript. Yasheng Yan attests to the integrity of the original data and the analysis reported in this manuscript
Name: Scott Canfield, BS
Contribution: This author helped perform the experiments, and analyze and interpret the data.
Attestation: Scott Canfield attests to having approved the final manuscript
Name: Maria Y. Muravyeva, MD, PhD
Contribution: This author helped analyze the data.
Attestation: Maria Y. Muravyeva attests to having approved the final manuscript
Name: Chika Kikuchi, MD
Contribution: This author helped perform the experiments, and analyze and interpret the data.
Attestation: Chika Kikuchi attests to having approved the final manuscript
Name: Ivan Zaja, MD
Contribution: This author helped perform the experiments, acquire the data, and analyze and interpret the data.
Attestation: Ivan Zaja attests to having approved the final manuscript
Name: John A. Corbett, PhD
Contribution: This author helped analyze and interpret the data.
Attestation: John A. Corbett attests to having approved the final manuscript
Name: Zeljko J. Bosnjak, PhD
Contribution: This author helped conceptualize the study, design the study, and analyze and interpret the data, and write the manuscript.
Attestation: Zeljko J. Bosnjak attests to having approved the final manuscript
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.