Abstract

Positron emission tomography (PET) is an effective tool for noninvasive examination of the body and provides a range of functional information. PET imaging with [18F]fluoro-2-deoxy-d-glucose ([18F]FDG) has been used to image alterations in glucose metabolism in brain or cancer tissue in the field of clinical diagnosis but not in the field of toxicology. A single dose of N-methyl-d-aspartate (NMDA) receptor antagonist induces neuronal cell degeneration/death in the rat retrosplenial/posterior cingulate (RS/PC) cortex region. These antagonists also increase local cerebral glucose utilization. Here, we examined the potential of [18F]FDG-PET as an imaging biomarker of neurotoxicity induced by an NMDA receptor antagonist, MK-801. Using [18F]FDG-PET, we determined that increased glucose utilization involved the neurotoxicity induced by MK-801. The accumulation of [18F]FDG was increased in the rat RS/PC cortex region showing neuronal cell degeneration/death and detected before the onset of neuronal cell death. This effect increased at a dose level at which neuronal cell degeneration recovered 24h after MK-801 administration. Scopolamine prevented the neurotoxicity and [18F]FDG accumulation induced by MK-801. Furthermore, in cynomolgus monkeys that showed no neuronal cell degeneration/death when treated with MK-801, we noted no differences in [18F]FDG accumulation between test and control subjects in any region of the brain. These findings suggest that [18F]FDG-PET, which is available for clinical trials, may be useful in generating a predictive imaging biomarker for detecting neurotoxicity against NMDA receptor antagonists with the same pharmacological activity as MK-801.

Positron emission tomography (PET) imaging with [18F]fluoro-2-deoxy-d-glucose ([18F]FDG), which is widely used in clinical settings to detect pathophysiological changes based on regional glucose metabolism in the brain, holds promise for the diagnostic assessment of patients with Alzheimer disease (AD) and dementia (Foster et al., 2007; Newberg et al., 2012). Using the same mechanisms as 2-deoxy-d-[14C]glucose, [18F]FDG is known to be accumulated in neuronal cells according to neuronal glucose utilization, which is closely coupled to local neuronal activity, and cerebral blood flow is continually adjusting to meet the dynamic alterations in the local metabolic demand associated with normal physiological events (Sokoloff, 1981; Yarowsky and Ingvar, 1981). Therefore, cerebral glucose hypometabolism detected by [18F]FDG in AD and dementia reflects decreased neuronal activity with neuronal cell degeneration. Recent studies have shown that, compared with computed tomography and magnetic resonance imaging (MRI), [18F]FDG-PET is a superior prospective imaging biomarker for assessing possible structural changes in the initial evaluation of dementia, demonstrating an ability to differentiate persons with AD from healthy subjects with high sensitivity and specificity (Bohnen et al., 2012). In contrast, [18F]FDG-PET has not been used to assess toxicity.

N-methyl-d-aspartate (NMDA) receptor antagonists protect against neurotoxicity in various experimental models of focal ischemia (Church et al., 1988; Collins and Olney, 1982; Gill et al., 1991). However, these antagonists also induce neurotoxicity in specific areas of the rodent brain. In rats, a single administration of NMDA receptor antagonists induces neuronal degeneration in the retrosplenial/posterior cingulate (RS/PC) cortex region (Carliss et al., 2007; Fix, 1994; Olney et al., 1989) and is characterized by vacuolar changes (Fix et al., 1996; Olney et al., 1989). Interestingly, these changes were not detected 24h after administration of low doses, indicating that vacuolation was reversible (Fix et al., 1996; Olney et al., 1989). In contrast, high doses of NMDA receptor antagonists caused irreversible neuronal necrosis in the RS/PC cortex region (Fix et al., 1996; Horvath et al., 1997; Wozniak et al., 1998). The vacuolar changes in the RS/PC cortex region induced by NMDA receptor antagonists were prevented by several neurotransmitter receptor ligands, such as agonists of GABAA and α2-adrenergic receptors, and antagonists of cholinergic muscarinic and glutamatergic non-NMDA receptors (Farber et al., 2002; Olney et al., 1991; Sharp et al., 1995). However, there are few studies, to our knowledge, showing the preventive effects of these agents against the neuronal cell death induced by NMDA receptor antagonists.

In addition to the above-mentioned effects, NMDA receptor antagonists induce marked regional alterations in local cerebral glucose utilization, including hypermetabolism in the vacuolar regions of rats (Kurumaji et al., 1989; Nehls et al., 1990; Sharkey et al., 1994). This alteration in glucose utilization occurs at a comparatively early phase after the administration of NMDA receptor antagonists (Eintrei et al., 1999; Kurumaji et al., 1989). Given that the hypermetabolism induced by NMDA receptor antagonists possibly reflects overactivation of neuronal cells (McCulloch and Iversen, 1991), we hypothesized that [18F]FDG-PET can be used as an imaging biomarker of NMDA receptor antagonist-induced cortical neurotoxicity. MK-801, a noncompetitive blocker of the opened ion channel of the NMDA receptor, is one of the most neurotoxic NMDA receptor antagonists (Olney et al., 1989). If the alteration of glucose metabolism by MK-801 could be detected by [18F]FDG-PET imaging, this methodology would represent a novel imaging biomarker of cortical neurotoxicity.

Here, we conducted dose- and time-based histopathological assessments of MK-801-induced neurotoxicity and confirmed the preventive effects of scopolamine, a cholinergic muscarinic antagonist, on MK-801-induced neuronal toxicity. Furthermore, to clarify the relationship between neuronal toxicity and glucose utilization, using [18F]FDG-PET imaging, we assessed the effect on glucose utilization of MK-801 with or without scopolamine in rats and the involvement of glucose utilization and neuronal toxicity in cynomolgus monkeys as representatives of nonhuman primates.

MATERIALS AND METHODS

Animals

A total of 205 eleven-week-old female Sprague Dawley rats (Crl: CD, Charles River Laboratories Japan, Inc., Kanagawa, Japan) were used in this study, with 85 relegated for histological examination and 120 for [18F]FDG-PET and MRI. These rats were acclimated to the animal room 1 week before experimentation. The animal room was maintained at temperature 23±3 °C and relative humidity 55±15%, with light provided from 7:00 a.m. to 7:00 p.m. Rats were allowed ad libitum access to a pellet diet (CRF-1, Oriental Yeast, Co., Ltd., Tokyo, Japan) and microfiltered tap water.

Four female cynomolgus monkeys each, aged 10 or 11 years and weighing 4–5kg (Macaca fascicularis; Guangxi Grandforest Scientific Company Ltd., Guangxi, China), were used for histological and [18F]FDG-PET imaging examinations. They were housed in an environmentally controlled room (12-h light/dark cycle, 23±5°C temperature, and 55±15% relative humidity) with ad libitum access to water and supplied with 100g of a pellet diet for monkeys (PS, Oriental Yeast Co., Ltd., Tokyo, Japan) every day. In addition, a 6-year-old male cynomolgus monkey (M. fascicularis; Maccine Pte Ltd., Singapore) was used for an MRI study.

Animal Treatment

A single dose of MK-801 maleate (Wako Pure Chemical Industries, Ltd., Osaka, Japan) dissolved in physiological saline was sc administered at 0.4 or 1mg/kg for rats and 3mg/kg for cynomolgus monkeys. A control group for rats was similarly administered physiological saline alone. For the rat study, scopolamine hydrobromide (LKT Laboratories, Inc., St Paul, MN) was dissolved in physiological saline and administered at 3 or 30mg/kg ip 10min after 1mg/kg MK-801 administration (Olney et al., 1991). A control group was administered physiological saline in the same manner.

All animal experiments were performed in accordance with the Guidelines for the Care and Use of Laboratory Animals of Astellas Pharma Inc. The MRI study was approved by the Institutional Animal Care and Use Committee (IACUC) of Maccine Pte Ltd. (Singapore Science Park II, Singapore) before commencing the study. The Kashima Facilities and Tsukuba Research Center of Astellas Pharma Inc., and Maccine Pte Ltd., have been awarded Accreditation Status by the AAALAC International.

Histological Examination

To confirm the dose levels causing reversible and irreversible neuronal changes and to clarify the temporal morphological changes in the RS/PC cortex region, rats for histological examination (n = 5 animals per group) were examined at 4 and 24h after administration of MK-801 at 0.4mg/kg and 2, 4, and 24h after administration of MK-801 at 1mg/kg. To confirm the suppressive effect of scopolamine administration against MK-801-induced neuronal cell degeneration and cell death, rats (n = 5 animals per group) were examined histologically by the method described below at 4 and 24h after MK-801 and scopolamine hydrobromide coadministration. The animals were perfused through the heart for a short period with saline under deep anesthesia with ether and then with formaldehyde-neutral buffer solution (Sigma Aldrich Inc., Tokyo, Japan) as a fixative. The perfused brains were quickly removed and immersed in the same fixative overnight.

Four cynomolgus monkeys for histological examination were euthanized under deep anesthesia with isoflurane 24h after MK-801 administration. The brains were quickly removed and immersed overnight in formaldehyde-neutral buffer solution (Sigma Aldrich Inc.). The fixed brains of rats and cynomolgus monkeys were embedded in paraffin and then cut into 4-µm sections with a microtome. The specimens were stained with hematoxylin and eosin and examined histopathologically.

Morphological changes in neurons were categorized as “absent or little present (−)” or “distinctly present (+).”

PET imaging with [18F]FDG and MRI in rats.

Regional glucose utilization of the brain lesion was evaluated using an Inveon small animal PET scanner (Siemens, Knoxville, TN) with [18F]FDG. [18F]FDG was synthesized in-house using a reagent kit for [18F]FDG synthesis (Rotem Industries Ltd., Arava, Israel) and TRACERlab MX (GE Healthcare, Waukesha, WI). Before performing imaging studies, rats for imaging examination were fasted overnight. At 1, 3, and 23h after MK-801 administration, the rats were injected through the tail vein with 10–20 MBq of [18F]FDG. After an uptake period of 50min, they were anesthetized (2.0–2.5% isoflurane in 100% oxygen gas) and placed on a bed with the head centered in the field of view (FOV) in a spread-prone position. A 5-min static acquisition was started 60min after [18F]FDG injection in three-dimensional mode. All images were reconstructed using the three-dimensional Ordered Subset Expectation Maximization algorithm with dead-time decay random and scatter corrections applied. The data obtained from the brain in units of Bq/g were converted to a standardized uptake value (SUV) using the following equation:

 
formula
(1)

PET images were coregistered with proton MR images (T2-3D) obtained from another rat. MRI data were acquired on a compact MRI system (MRmini SA; DS Pharma Biomedical). Two-dimensional multislice T1W images were obtained using a spin-echo sequence with the following parameters: pulse repetition time (TR) = 400ms, echo time (TE) = 9ms, matrix size = 256×128, FOV = 30mm, slice thickness = 1.5mm, number of acquisitions = 8, number of slices = 5, slice gap = 0mm, and acquisition time for one set = 6.8min. Image reconstruction and analysis were performed using an NMR Imager (MR Technology) and MRVision (Ver. 1.5, MRVision).

[18F]FDG-PET imaging and MRI in cynomolgus monkeys.

We used four animals, which differed from those used for histology, for [18F]FDG-PET imaging. Regional glucose utilization in the brain lesions of monkeys was evaluated using a high-resolution animal PET scanner (SHR-17000; Hamamatsu Photonics K.K., Shizuoka, Japan), with a transaxial resolution of 2.6mm full-width at half-maximum in the center of the scan field and a center-to-center distance of 3.05mm. A blank scan (120min) was performed using a rotating 68Ge/68Ga line source before each study. Before scanning, and after mask induction (isoflurane gas), animals were rapidly tracheally intubated and mechanically ventilated with approximately 2% isoflurane in a N2O/air/O2 (2:2:1) gas mixture (Apollo; Dräger Medical Japan, Tokyo, Japan).

A catheter was inserted into the left femoral vein, and the animal was positioned foot first, supine on the scanner bed, and secured by straps. Appropriate veterinary equipment was used to maintain body temperature. A transmission scan (30min) was performed using a rotating 68Ge/68Ga line source. At 3h after physiological saline or MK-801 administration, the monkey was iv injected with 244±19 MBq of [18F]FDG. After an uptake period of 60min, PET emission scans were started and dynamic PET emission data were collected for 20min in two-dimensional list-mode acquisition. Dynamic PET images were reconstructed using a dynamic row-action maximum likelihood algorithm and anatomically standardized using a Neurostat (Cross et al., 2000; Minoshima et al., 1994, 1995). A region of interest (ROI) was taken for the whole brain on the anatomically standardized images. The data obtained from the brain in units of Bq/g were converted to an SUV using Equation 1.

Averaged PET images were generated from anatomically standardized images from each group. MRI scanning was conducted using a single animal as anatomical reference on a 3T Siemens TIM Trio scanner (Siemens Healthcare; Erlangen, Germany) at Maccine Co. Ltd. (Singapore). The whole head was imaged using an i-PAT-compatible, eight-channel clinical knee coil (Siemens Healthcare; Erlangen, Germany). At the start of the scan, the animal was positioned head first, supine on the scanner bed, and secured with straps, with the head positioned in the center of the coil. Appropriate veterinary equipment was used to maintain body temperature and monitor vital signs during the scan. Following scout and localizer images, an isotropic, high-resolution structural image of the head was acquired with FOV sufficiently large to capture full nasal cavity, brain, ear canal, and upper spinal cord. The high-resolution image was a three-dimensional FLASH (fast low-angle shot) with T1 weighting, with 160 slices collected in transverse orientation (0.6×0.6×0.6mm voxels; 0mm slice gap; TR = 20ms; TE = 5.73ms; flip angle = 25°; FOV = 124mm; bandwidth = 140 Hz/Px). Total acquisition time was approximately 90min.

Statistical Analysis

Histological data for MK-801 and scopolamine coadministration were analyzed using the Holm’s method. SUV data were analyzed using one-way ANOVA followed by the Dunnett’s multiple comparison test or the post hoc Tukey’s test. Statistical analyses were performed using Pharmaco Analyst II (Hakuhousha, Tokyo, Japan) for histological data and GraphPad Prism 5 (GraphPad Software, Inc., San Diego, CA) for SUV data, and P values less than 0.05 were considered to be significant. SUV data are presented as the mean and SEM.

RESULTS

Neuromorphological Changes Induced by MK-801 in the RS/PC Cortex Region in Rats

The time course of MK-801-induced neurotoxicity was evaluated by histopathological examination performed at 0 (saline), 2, 4, and 24h after MK-801 administration at 1mg/kg. At 2h after MK-801 administration, only one animal showed neuronal vacuolar degeneration in the RS/PC cortex region. This vacuolar degeneration was observed in most animals at 4h after MK-801 administration (Figs. 1A and 1a and Table 1). However, no cells with vacuolar degeneration were detected at 24h after administration in any animal. Neuronal cell death was detected at 4h after administration and at 24h (Figs. 1B and 1b and Table 1). To confirm the dose levels causing reversible neuronal changes in the RS/PC cortex region, histopathological examination in rats was conducted after MK-801 administration at 0.4mg/kg. At 4h after administration, vacuolar degeneration in cells in the RS/PC cortex region was detected at 1mg/kg (Table 1). However, at 24h after MK-801 administration, neither neuronal cell death nor vacuolar degeneration was detected in any animal treated with 0.4mg/kg.

Table 1

Dose Dependency and Time Course of Neurotoxicity Induced by MK-801 in the RS/PC Cortex Region in Rat Brains Induced by MK-801

MK-801 0.4mg/kg 1mg/kg 
After administration 0 h 4 h 24 h 0 h 2 h 4 h 24 h 
Number of animals 
Grade 
 Neuronal vacuolation 
  − 
  + 
 Neuronal cell death 
  − 
  + 
MK-801 0.4mg/kg 1mg/kg 
After administration 0 h 4 h 24 h 0 h 2 h 4 h 24 h 
Number of animals 
Grade 
 Neuronal vacuolation 
  − 
  + 
 Neuronal cell death 
  − 
  + 

Note. −, absent or little present; +, distinctly present.

Fig. 1.

Neuronal changes revealed by hematoxylin and eosin staining of the RS/PC cortex region of rats treated with 1mg/kg MK-801. Neuronal vacuolation appeared in the RS/PC cortex region of rat 4h after MK-801 administration (A and a). No neuronal vacuolation was observed 24h after MK-801 administration; instead, neuronal cell death was detected in the same region (B and b). The arrowhead and arrow point to neuronal vacuolation and neuronal cell death, respectively. Scale bar: 50 µm at low magnification (A and B) and 20 µm at high magnification (a and b).

Fig. 1.

Neuronal changes revealed by hematoxylin and eosin staining of the RS/PC cortex region of rats treated with 1mg/kg MK-801. Neuronal vacuolation appeared in the RS/PC cortex region of rat 4h after MK-801 administration (A and a). No neuronal vacuolation was observed 24h after MK-801 administration; instead, neuronal cell death was detected in the same region (B and b). The arrowhead and arrow point to neuronal vacuolation and neuronal cell death, respectively. Scale bar: 50 µm at low magnification (A and B) and 20 µm at high magnification (a and b).

Prevention of MK-801-Induced Neurotoxicity by Coadministration of Scopolamine in Rats

We next assessed the preventive effect of scopolamine on the neuronal changes induced by MK-801 in the RS/PC cortex region in rats at the dose of 1mg/kg, which induced both vacuolar degeneration and cell death. The vacuolar degeneration was detected at 4h after MK-801 administration alone, which was consistent with the previous finding, and this change was clearly suppressed by scopolamine treatment at 3 and 30mg/kg (Table 2). Neuronal cell death after MK-801 administration was similarly prevented by scopolamine at 30mg/kg but not at 3mg/kg (Table 2). No changes in neuronal cells were detected with scopolamine administration alone, even at 30mg/kg.

Table 2

Effects of Scopolamine Against Neurotoxicity Induced by MK-801 in the RS/PC Cortex Region in Rat Brains Induced by MK-801

MK-801 (mg/kg) 
Scopolamine (mg/kg) 30 30 30 30 
After administration 4 h 24 h 
Number of animals 
Grade 
 Neuronal vacuolation 
  − 5* 5* 
  + 
 Neuronal cell death 
  − 5* 
  + 
MK-801 (mg/kg) 
Scopolamine (mg/kg) 30 30 30 30 
After administration 4 h 24 h 
Number of animals 
Grade 
 Neuronal vacuolation 
  − 5* 5* 
  + 
 Neuronal cell death 
  − 5* 
  + 

Notes. −, absent or little present; +, distinctly present.

*Statistically significant difference (p < 0.05 vs. MK-801 1mg/kg and scopolamine 0mg/kg group).

Increased Accumulation of [18F]FDG in the RS/PC Cortex Region of Rat Brain by MK-801 Administration

First, we performed [18F]FDG-PET imaging to examine the time course of [18F]FDG disposition in the rat brain after MK-801 administration at 1mg/kg (Fig. 2). An increase in [18F]FDG accumulation on the rat RS/PC cortex region was observed 4h after MK-801 administration (Fig. 2A). We further examined the time course of the [18F]FDG accumulation. At 2h after administering MK-801, the accumulation of [18F]FDG was markedly elevated in the RS/PC cortex region in which neuronal cell vacuolation/death was histopathologically detected. This accumulation increased until 4h after MK-801 administration and then returned to the baseline level by 24h (Fig. 2B).

Fig. 2.

Effect of MK-801 on glucose metabolism in rat brain. (A) Representative MR and PET images for the slice corresponding to the RS/PC cortex region of rat brain acquired 2, 4, and 24h after sc administration of 1mg/kg MK-801 or saline control. Arrows, RS/PC cortex region. SUV is a normalized tracer concentration indicating glucose utilization for [18F]FDG, reflecting neuronal activity. The [18F]FDG uptake in RS/PC cortex region was elevated in MK-801-treated animals at 2 and 4h. (B) Quantitative analysis of [18F]FDG-PET images of the RS/PC cortex region at 0h (saline), 2h (MK), 4h (MK), and 24h (MK) after administration of MK-801 at 1mg/kg. Mean ± SEM (n = 6).

Fig. 2.

Effect of MK-801 on glucose metabolism in rat brain. (A) Representative MR and PET images for the slice corresponding to the RS/PC cortex region of rat brain acquired 2, 4, and 24h after sc administration of 1mg/kg MK-801 or saline control. Arrows, RS/PC cortex region. SUV is a normalized tracer concentration indicating glucose utilization for [18F]FDG, reflecting neuronal activity. The [18F]FDG uptake in RS/PC cortex region was elevated in MK-801-treated animals at 2 and 4h. (B) Quantitative analysis of [18F]FDG-PET images of the RS/PC cortex region at 0h (saline), 2h (MK), 4h (MK), and 24h (MK) after administration of MK-801 at 1mg/kg. Mean ± SEM (n = 6).

We next examined the effect of MK-801 on glucose utilization at 0.4mg/kg, which caused reversible changes (neuronal vacuolation) in the rat RS/PC cortex region (Fig. 3 and Table 1). At 0.4mg/kg, accumulation of [18F]FDG was significantly increased to the same level as that at 1mg/kg, which causes irreversible changes.

Fig. 3.

Dose-dependent effect of MK-801 on glucose metabolism in rat brain. Quantitative analysis of [18F]FDG-PET images of the RS/PC cortex region 4h after administration of saline (saline, n = 15), or MK-801 at 0.4mg/kg (MK 0.4, n = 5) or 1mg/kg (MK 1, n = 11). Mean ± SEM.

Fig. 3.

Dose-dependent effect of MK-801 on glucose metabolism in rat brain. Quantitative analysis of [18F]FDG-PET images of the RS/PC cortex region 4h after administration of saline (saline, n = 15), or MK-801 at 0.4mg/kg (MK 0.4, n = 5) or 1mg/kg (MK 1, n = 11). Mean ± SEM.

Suppression of MK-801-Induced Glucose Utilization by Scopolamine in Rats

To investigate the effect of scopolamine on the enhancement of [18F]FDG accumulation by MK-801 in rats, we performed [18F]FDG-PET imaging at 2 and 4h after coadministration of MK-801 and scopolamine (Fig. 4). At 2h after MK-801 administration, scopolamine at both 3 and 30mg/kg completely suppressed the increased [18F]FDG accumulation by MK-801 (Fig. 4A). The increased [18F]FDG accumulation at 4h after administration was suppressed by coadministration of scopolamine at 30mg/kg, but the effect of scopolamine coadministration at 3mg/kg was small (Fig. 4B).

Fig. 4.

Effect of scopolamine coadministration on increased glucose metabolism by MK-801 in rat brain. (A) Quantitative analysis of [18F]FDG-PET images of the RS/PC cortex region 2h after administration in rats treated with saline (saline, n = 10), 1mg/kg of MK-801 (MK-801, n = 6), 30mg/kg of scopolamine (Sco30, n = 6), 1mg/kg of MK-801 plus 3mg/kg of scopolamine (MK + Sco3, n = 8), or 1mg/kg of MK-801 plus 30mg/kg of scopolamine (MK + Sco30, n = 7). (B) Quantitative analysis of [18F]FDG-PET images of the RS/PC cortex region 4h after administration in rats treated with saline (saline, n = 10), 1mg/kg of MK-801 (MK-801, n = 6), 30mg/kg of scopolamine (Sco30, n = 7), 1mg/kg of MK-801 plus 3mg/kg of scopolamine (MK + Sco3, n = 6), or 1mg/kg of MK-801 plus 30mg/kg of scopolamine (MK + Sco30, n = 8). Mean ± SEM.

Fig. 4.

Effect of scopolamine coadministration on increased glucose metabolism by MK-801 in rat brain. (A) Quantitative analysis of [18F]FDG-PET images of the RS/PC cortex region 2h after administration in rats treated with saline (saline, n = 10), 1mg/kg of MK-801 (MK-801, n = 6), 30mg/kg of scopolamine (Sco30, n = 6), 1mg/kg of MK-801 plus 3mg/kg of scopolamine (MK + Sco3, n = 8), or 1mg/kg of MK-801 plus 30mg/kg of scopolamine (MK + Sco30, n = 7). (B) Quantitative analysis of [18F]FDG-PET images of the RS/PC cortex region 4h after administration in rats treated with saline (saline, n = 10), 1mg/kg of MK-801 (MK-801, n = 6), 30mg/kg of scopolamine (Sco30, n = 7), 1mg/kg of MK-801 plus 3mg/kg of scopolamine (MK + Sco3, n = 6), or 1mg/kg of MK-801 plus 30mg/kg of scopolamine (MK + Sco30, n = 8). Mean ± SEM.

Neuromorphological Changes and Alteration of [18F]FDG Accumulation by MK-801 in Cynomolgus Monkeys

To evaluate the neuromorphological changes in cynomolgus monkeys after MK-801 administration, histopathological examination was performed 24h after MK-801 administration at 3mg/kg. No neuronal cell changes, including death, were detected in any animal (data not shown). We also performed quantitative brain [18F]FDG-PET imaging 4h after MK-801 administration (3mg/kg) (Supplementary fig. 1). [18F]FDG disposition in the brain of each monkey at 4h after MK-801 dosing was comparable with that after dosing with physiological saline (Supplementary fig. 1A), and ROI analysis revealed no differences between physiological saline and MK-801 dosing (Supplementary fig. 1B).

DISCUSSION

In this study, we showed that [18F]FDG was strongly accumulated in the rat RS/PC cortex region in which neuronal cell vacuolation/cell death was induced by treatment with MK-801. The [18F]FDG accumulation was also detected at 0.4mg/kg MK-801 that caused reversible neuronal vacuolation. We also found that the cholinergic muscarinic antagonist scopolamine suppressed not only neuronal vacuolation and cell death but also the increased accumulation in this region of [18F]FDG induced by MK-801. In cynomolgus monkeys, in contrast, no neuronal cell degeneration or cell death was observed in the PC cortex, and alteration of [18F]FDG accumulation was not detected in any region of the brain. These results indicate that the pronounced increase in glucose utilization in the RS/PC cortex region was closely related to neuronal toxicity and that the increase in glucose utilization occurred at a dose at which neuronal toxicity was reversible. Thus, the clinically available [18F]FDG-PET appears to represent a useful imaging biomarker of neurotoxicity induced by NMDA receptor antagonists with the same pharmacological activity as MK-801.

The transient occurrence of neuronal cell vacuolation and neuronal cell death in the rat RS/PC cortex region following NMDA antagonist administration has been well documented (Fix et al., 1993; Olney et al., 1989). Although vacuolar changes induced by MK-801 at low doses have been reported to be reversible, high doses could cause irreversible neuronal necrosis (Allen and Iversen, 1990). The number of vacuolized neurons in the rat RS cortex peaked 4h after MK-801 administration at 0.5mg/kg. The number of vacuolized neurons gradually decreased after 4h and then the vacuolized neurons were undetectable at 16 and 24h (Hashimoto et al., 2000).

Further, at 10mg/kg as a higher dosage, cytoplasmic vacuoles of pyramidal neurons were observed in the RS/PC cortex region 4h after MK-801, after which vacuoles were not seen but neuronal necrosis was detected in the same region (Fix et al., 1996). Consistent with these reports, we showed that MK-801 at 0.4mg/kg as the low dose induced transient neuronal vacuolation in the rat RS/PC cortex region, a change that was subsequently resolved. At 1mg/kg as the high dose, neuronal vacuolation was detected in the same region at 4h after MK-801 administration. Although this vacuolation was not detected at 24h, neuronal cell death in this region was seen. These results indicate that 0.4mg/kg of MK-801 was a dose level capable of inducing reversible neuronal change, whereas 1mg/kg of MK-801 was the level that induced irreversible neuronal changes. The mechanistic difference between the reversible and irreversible changes is unclear at present. The difference in the neurotoxicity of MK-801 by dose level may contribute to the duration of pharmacological action.

The mechanism of neuronal degeneration induced by NMDA receptor antagonists has been widely studied, and it involves the cholinergic, glutamatergic, GABAergic, dopaminergic, and noradrenergic systems (Farber et al., 2002,, 2003; Griffiths et al., 2000; Kim et al., 1999; Olney et al., 1991). Administration of muscarinic cholinergic antagonists, such as scopolamine, atropine, and procyclidine, prevents the neuronal vacuolation induced by NMDA receptor antagonists, whereas pilocarpine, a cholinergic agonist, exacerbated this neurotoxicity (Corso et al., 1997; Olney et al., 1991). A microdialysis study showed that the acetylcholine concentration in the RS/PC cortex region after MK-801 administration was three times that at baseline and that the release of acetylcholine after MK-801 administration was blocked by α2-adrenergic or GABAA agonists (Kim et al., 1999).

Our present results demonstrate that neuronal vacuolation after MK-801 administration was prevented by scopolamine, a muscarinic cholinergic antagonist, from a low dose of scopolamine (Table 2). In contrast to abundant studies on the prevention of neuronal vacuolation, there are few studies showing the preventive effects against neuronal cell death. Here, we show that neuronal cell death after the administration of MK-801 was prevented by a high but not low dose of scopolamine (Table 2). The pharmacological action of scopolamine peaks soon after administration and shows a clear, dose-dependent response that is prolonged with increasing dose (Brand, 1969; Ebert et al., 1998). In this study, the difference in the effect of scopolamine with increasing dose level might be attributable to the duration of its pharmacological action.

NMDA receptor antagonists have been found to produce marked alterations in local cerebral glucose. For example, Nehls et al. (1990) showed that MK-801 administered at 0.5mg/kg iv increased glucose utilization in the PC cortex (Nehls et al., 1990). Kurumaji and McCulloch (1989) reported that MK-801 increased glucose utilization in the rat limbic system. Another NMDA receptor antagonist, ketamine, was reported to increase glucose metabolism in the rat RS/PC cortex region (Eintrei et al., 1999). In this study, we describe, for the first time to our knowledge, the temporal involvement of glucose utilization and neuronal degeneration/cell death after NMDA receptor antagonist administration using [18F]FDG-PET. [18F]FDG accumulation was increased in the rat RS/PC cortex region at 2h after MK-801 administration, at which time neuronal degeneration had hardly appeared (Table 1 and Fig. 2). Further, [18F]FDG accumulation was detected after administration of MK-801 at 0.4mg/kg, which caused reversible morphological changes in neurons (Table 1 and Fig. 3).

An increase in glucose utilization induced by MK-801 may result from the overactivation of polysynaptic circuits (Kurumaji and McCulloch, 1989; Patel and McCulloch, 1995). Farber et al. (2002) reported that hyperactivation of a muscarinic receptor plays an important role in mediating the neurotoxicity induced by MK-801 (Farber et al., 2002). In addition, based on the results of direct intra-RS/PC cortex region injection of a non-NMDA glutamate receptor antagonist and MK-801, NMDA receptor hypofunction causes excitatory amino acid release in the RS/PC cortex region and hyperactivation of a non-NMDA glutamate receptor on neurons in the RS/PC cortex region. The hyperactivation in the RS/PC cortex region may represent an additional critical component of the mechanism that mediates this neurotoxic reaction (Farber et al., 2002; Noguchi et al., 1998).

Excitotoxic increase in glutamatergic neurotransmission leads to a massive increase in the level of intracellular Ca2+, which is rapidly incorporated by mitochondria and induces mitochondrial swelling (Ferreiro et al., 2012; Galindo et al., 2003). Observations using electron microscopy 4h after MK-801 administration showed that mitochondria and endoplasmic reticulum contributed to the formation of vacuoles in neuronal cell, and most mitochondria were irregularly swollen (Fix et al., 1993). Here, we clarified the relationship between [18F]FDG accumulation and the neuronal toxicity, namely that scopolamine prevented not only the neuronal cell degeneration/death but also the accumulation of glucose (Table 2 and Fig. 4).

In cynomolgus monkeys, no degeneration or death of neuronal cells was detected at 3mg/kg. The dose of MK-801 used in our study was sufficiently high, with 3mg/kg in cynomolgus monkeys being roughly equivalent to 9mg/kg in rats based on body surface area (Freireich et al., 1966). At this dosage level, MK-801 induced no detectable accumulation of [18F]FDG in any region of the brain in cynomolgus monkeys (Supplementary fig. 1). Taking these and our present results together, we consider that the increase in glucose utilization induced by NMDA receptor antagonists indicates the relationship between neuronal toxicity and synaptic overactivation, and the synaptic overactivation can be detected using [18F]FDG-PET (Supplementary fig. 2).

Recently, Zhang et al. (2009) reported that [18F]FDG uptake in the brain of rats did not significantly differ between rats treated with ketamine, an NMDA receptor antagonist, and rats treated with saline. They administered ketamine at 20mg/kg six times on postnatal day (PND) 7 and then acquired brain images on PND 35 (Zhang et al., 2009). Here, we conducted PET imaging from 2h after MK-801 administration in rats aged 12 weeks. NMDA antagonist–induced neurotoxicity is reported to differ depending on the age of animals (Ellison, 1995). Thus, this discrepancy may have resulted from differences in experimental conditions, including the timing of image acquisition and the age of the test animals.

In conclusion, we demonstrated that MK-801 caused either reversible or irreversible neurotoxicity, depending on dosage level. [18F]FDG-PET revealed that glucose utilization was increased not only at the dose level at which MK-801 induced an irreversible neuronal change but also at the dose level at which MK-801 induced reversible neuronal change in the RS/PC cortex region in the brain of rats. In addition, regarding the relationship between the neuronal toxicity and glucose utilization, studies of the effects of scopolamine or on cynomolgus monkeys indicate the validity of this relationship. The present results suggest that [18F]FDG-PET, which is available in clinical trials, may be a useful tool as a predictive imaging biomarker against the neuronal toxicity induced by NMDA receptor antagonists with the same pharmacological activity as MK-801.

SUPPLEMENTARY DATA

Supplementary data are available online at http://toxsci.oxfordjournals.org/.

FUNDING

Astellas Pharma Inc.

ACKNOWLEDGMENTS

The authors are grateful to H. Kurihara, K. Nakano, and M. Murata for their technical assistance and T. Kimura and J. Mizusawa for help with PET imaging studies. We also thank E. P. Manigbas for assisting with the MRI study of cynomolgus monkeys.

REFERENCES

Allen
H. L.
Iversen
L. L
.
1990
.
Phencyclidine, dizocilpine, and cerebrocortical neurons
.
Science
 
247
,
221
.
Bohnen
N. I.
Djang
D. S.
Herholz
K.
Anzai
Y.
Minoshima
S
.
2012
.
Effectiveness and safety of 18F-FDG PET in the evaluation of dementia: A review of the recent literature
.
J. Nucl. Med
 .
53
,
59
71
.
Brand
J. J
.
1969
.
The time course of action of (-)-hyoscine after intramuscular injection
.
Br. J. Pharmacol
 .
35
,
202
208
.
Carliss
R. D.
Radovsky
A.
Chengelis
C. P.
O’Neill
T. P.
Shuey
D. L
.
2007
.
Oral administration of dextromethorphan does not produce neuronal vacuolation in the rat brain
.
Neurotoxicology
 .
28
,
813
818
.
Church
J.
Zeman
S.
Lodge
D
.
1988
.
The neuroprotective action of ketamine and MK-801 after transient cerebral ischemia in rats
.
Anesthesiology
 .
69
,
702
709
.
Collins
R. C.
Olney
J. W
.
1982
.
Focal cortical seizures cause distant thalamic lesions
.
Science
 .
218
,
177
179
.
Corso
T. D.
Sesma
M. A.
Tenkova
T. I.
Der
T. C.
Wozniak
D. F.
Farber
N. B.
Olney
J. W
.
1997
.
Multifocal brain damage induced by phencyclidine is augmented by pilocarpine
.
Brain Res
 .
752
,
1
14
.
Cross
D. J.
Minoshima
S.
Nishimura
S.
Noda
A.
Tsukada
H.
Kuhl
D. E
.
2000
.
Three-dimensional stereotactic surface projection analysis of macaque brain PET: Development and initial applications
.
J. Nucl. Med
 .
41
,
1879
1887
.
Ebert
U.
Siepmann
M.
Oertel
R.
Wesnes
K. A.
Kirch
W
.
1998
.
Pharmacokinetics and pharmacodynamics of scopolamine after subcutaneous administration
.
J. Clin. Pharmacol
 .
38
,
720
726
.
Eintrei
C.
Sokoloff
L.
Smith
C. B
.
1999
.
Effects of diazepam and ketamine administered individually or in combination on regional rates of glucose utilization in rat brain
.
Br. J. Anaesth
 .
82
,
596
602
.
Ellison
G
.
1995
.
The N-methyl-D-aspartate antagonists phencyclidine, ketamine and dizocilpine as both behavioral and anatomical models of the dementias
.
Brain Res. Brain Res. Rev
 .
20
,
250
267
.
Farber
N. B.
Jiang
X.
Dikranian
K.
Nemmers
B
.
2003
.
Muscimol prevents NMDA antagonist neurotoxicity by activating GABAA receptors in several brain regions
.
Brain Res
 .
993
,
90
100
.
Farber
N. B.
Kim
S. H.
Dikranian
K.
Jiang
X. P.
Heinkel
C
.
2002
.
Receptor mechanisms and circuitry underlying NMDA antagonist neurotoxicity
.
Mol. Psychiatry
 .
7
,
32
43
.
Ferreiro
E.
Baldeiras
I.
Ferreira
I. L.
Costa
R. O.
Rego
A. C.
Pereira
C. F.
Oliveira
C. R
.
2012
.
Mitochondrial- and endoplasmic reticulum-associated oxidative stress in Alzheimer’s disease: From pathogenesis to biomarkers
.
Int. J. Cell Biol
 .
2012
,
735206
.
Fix
A. S
.
1994
.
Pathological effects of MK-801 in the rat posterior cingulate/retrosplenial cortex
.
Psychopharmacol. Bull
 .
30
,
577
583
.
Fix
A. S.
Horn
J. W.
Wightman
K. A.
Johnson
C. A.
Long
G. G.
Storts
R. W.
Farber
N.
Wozniak
D. F.
Olney
J. W
.
1993
.
Neuronal vacuolization and necrosis induced by the noncompetitive N-methyl-D-aspartate (NMDA) antagonist MK(+)801 (dizocilpine maleate): A light and electron microscopic evaluation of the rat retrosplenial cortex
.
Exp. Neurol
 .
123
,
204
215
.
Fix
A. S.
Ross
J. F.
Stitzel
S. R.
Switzer
R. C
.
1996
.
Integrated evaluation of central nervous system lesions: Stains for neurons, astrocytes, and microglia reveal the spatial and temporal features of MK-801-induced neuronal necrosis in the rat cerebral cortex
.
Toxicol. Pathol
 .
24
,
291
304
.
Foster
N. L.
Heidebrink
J. L.
Clark
C. M.
Jagust
W. J.
Arnold
S. E.
Barbas
N. R.
DeCarli
C. S.
Turner
R. S.
Koeppe
R. A.
Higdon
R.
et al
.
2007
.
FDG-PET improves accuracy in distinguishing frontotemporal dementia and Alzheimer’s disease
.
Brain
 .
130
(
Pt 10
)
2616
2635
.
Freireich
E. J.
Gehan
E. A.
Rall
D. P.
Schmidt
L. H.
Skipper
H. E
.
1966
.
Quantitative comparison of toxicity of anticancer agents in mouse, rat, hamster, dog, monkey, and man
.
Cancer Chemother. Rep
 .
50
,
219
244
.
Galindo
M. F.
Jordán
J.
González-García
C.
Ceña
V
.
2003
.
Reactive oxygen species induce swelling and cytochrome c release but not transmembrane depolarization in isolated rat brain mitochondria
.
Br. J. Pharmacol
 .
139
,
797
804
.
Gill
R.
Brazell
C.
Woodruff
G. N.
Kemp
J. A
.
1991
.
The neuroprotective action of dizocilpine (MK-801) in the rat middle cerebral artery occlusion model of focal ischaemia
.
Br. J. Pharmacol
 .
103
,
2030
2036
.
Griffiths
M. R.
Cooper
A. J.
Barber
D. J.
Mitchell
I. J
.
2000
.
Pharmacological mechanisms mediating phencyclidine-induced apoptosis of striatopallidal neurons: The roles of glutamate, dopamine, acetylcholine and corticosteroids
.
Brain Res
 .
855
,
1
10
.
Hashimoto
K.
Fujimura
M.
Yamagami
K
.
2000
.
Dizocilpine-induced neuropathological changes in rat retrosplenial cortex are reversed by subsequent clozapine treatment
.
Life Sci
 .
66
,
1071
1078
.
Horvath
Z. C.
Czopf
J.
Buzsáki
G
.
1997
.
MK-801-induced neuronal damage in rats
.
Brain Res
 .
753
,
181
195
.
Kim
S. H.
Price
M. T.
Olney
J. W.
Farber
N. B
.
1999
.
Excessive cerebrocortical release of acetylcholine induced by NMDA antagonists is reduced by GABAergic and alpha2-adrenergic agonists
.
Mol. Psychiatry
 .
4
,
344
352
.
Kurumaji
A.
McCulloch
J
.
1989
.
Effects of MK-801 upon local cerebral glucose utilisation in conscious rats and in rats anaesthetised with halothane
.
J. Cereb. Blood Flow Metab
 .
9
,
786
794
.
Kurumaji
A.
Nehls
D. G.
Park
C. K.
McCulloch
J
.
1989
.
Effects of NMDA antagonists, MK-801 and CPP, upon local cerebral glucose use
.
Brain Res
 .
496
,
268
284
.
McCulloch
J.
Iversen
L. L
.
1991
.
Autoradiographic assessment of the effects of N-methyl-D-aspartate (NMDA) receptor antagonists in vivo
.
Neurochem. Res
 .
16
,
951
963
.
Minoshima
S.
Foster
N. L.
Kuhl
D. E
.
1994
.
Posterior cingulate cortex in Alzheimer’s disease
.
Lancet
 .
344
,
895
.
Minoshima
S.
Frey
K. A.
Koeppe
R. A.
Foster
N. L.
Kuhl
D. E
.
1995
.
A diagnostic approach in Alzheimer’s disease using three-dimensional stereotactic surface projections of fluorine-18-FDG PET
.
J. Nucl. Med
 .
36
,
1238
1248
.
Nehls
D. G.
Park
C. K.
MacCormack
A. G.
McCulloch
J
.
1990
.
The effects of N-methyl-D-aspartate receptor blockade with MK-801 upon the relationship between cerebral blood flow and glucose utilisation
.
Brain Res
 .
511
,
271
279
.
Newberg
A. B.
Arnold
S. E.
Wintering
N.
Rovner
B. W.
Alavi
A
.
2012
.
Initial clinical comparison of 18F-florbetapir and 18F-FDG PET in patients with Alzheimer disease and controls
.
J. Nucl. Med
 .
53
,
902
907
.
Noguchi
K.
Johnson
R.
Ellision
G
.
1998
.
The effects of MK-801 on aspartate and glutamate levels in the anterior cingulate and retrosplential cortices: An in vivo microdialysis study
Soc Neurosci. Abs
 .
24
,
233
.
Olney
J. W.
Labruyere
J.
Price
M. T
.
1989
.
Pathological changes induced in cerebrocortical neurons by phencyclidine and related drugs
.
Science
 .
244
,
1360
1362
.
Olney
J. W.
Labruyere
J.
Wang
G.
Wozniak
D. F.
Price
M. T.
Sesma
M. A
.
1991
.
NMDA antagonist neurotoxicity: Mechanism and prevention
.
Science
 .
254
,
1515
1518
.
Patel
T. R.
McCulloch
J
.
1995
.
AMPA receptor antagonism attenuates MK-801-induced hypermetabolism in the posterior cingulate cortex
.
Brain Res
 .
686
,
254
258
.
Sharkey
J.
Ritchie
I. M.
Butcher
S. P.
Kelly
J. S
.
1994
.
Differential effects of competitive (CGS19755) and non-competitive (MK 801) NMDA receptor antagonists upon local cerebral blood flow and local cerebral glucose utilisation in the rat
.
Brain Res
 .
651
,
27
36
.
Sharp
J. W.
Petersen
D. L.
Langford
M. T
.
1995
.
DNQX inhibits phencyclidine (PCP) and ketamine induction of the hsp70 heat shock gene in the rat cingulate and retrosplenial cortex
.
Brain Res
 .
687
,
114
124
.
Sokoloff
L
.
1981
.
Relationships among local functional activity, energy metabolism, and blood flow in the central nervous system
.
Fed. Proc
 .
40
,
2311
2316
.
Wozniak
D. F.
Dikranian
K.
Ishimaru
M. J.
Nardi
A.
Corso
T. D.
Tenkova
T.
Olney
J. W.
Fix
A. S
.
1998
.
Disseminated corticolimbic neuronal degeneration induced in rat brain by MK-801: Potential relevance to Alzheimer’s disease
.
Neurobiol. Dis
 .
5
,
305
322
.
Yarowsky
P. J.
Ingvar
D. H
.
1981
.
Symposium summary. Neuronal activity and energy metabolism
.
Fed. Proc
 .
40
,
2353
2362
.
Zhang
X.
Paule
M. G.
Newport
G. D.
Zou
X.
Sadovova
N.
Berridge
M. S.
Apana
S. M.
Hanig
J. P.
Slikker
W.
Jr
Wang
C
.
2009
.
A minimally invasive, translational biomarker of ketamine-induced neuronal death in rats: MicroPET Imaging using 18F-annexin V
.
Toxicol. Sci
 .
111
,
355
361
.