Automated Synthesis of [18F]Flumazenil Application in GABAA Receptor Neuroimaging Availability for Rat Model of Anxiety

Clinical studies have demonstrated that the γ-aminobutyric acid type A (GABAA) receptor complex plays a central role in the modulation of anxiety. Conditioned fear and anxiety-like behaviors have many similarities at the neuroanatomical and pharmacological levels. The radioactive GABA/BZR receptor antagonist, fluorine-18-labeled flumazenil, [18F]flumazenil, behaves as a potential PET imaging agent for the evaluation of cortical damage of the brain in stroke, alcoholism, and for Alzheimer disease investigation. The main goal of our study was to investigate a fully automated nucleophilic fluorination system, with solid extraction purification, developed to replace traditional preparation methods, and to detect underlying expressions of contextual fear and characterize the distribution of GABAA receptors in fear-conditioned rats by [18F]flumazenil. A carrier-free nucleophilic fluorination method using an automatic synthesizer with direct labeling of a nitro-flumazenil precursor was implemented. The semi-preparative high-performance liquid chromatography (HPLC) purification method (RCY = 15–20%) was applied to obtain high purity [18F]flumazenil. Nano-positron emission tomography (NanoPET)/computed tomography (CT) imaging and ex vivo autoradiography were used to analyze the fear conditioning of rats trained with 1–10 tone-foot-shock pairings. The anxiety rats had a significantly lower cerebral accumulation (in the amygdala, prefrontal cortex, cortex, and hippocampus) of fear conditioning. Our rat autoradiography results also supported the findings of PET imaging. Key findings were obtained by developing straightforward labeling and purification procedures that can be easily adapted to commercially available modules for the high radiochemical purity of [18F]flumazenil. The use of an automatic synthesizer with semi-preparative HPLC purification would be a suitable reference method for new drug studies of GABAA/BZR receptors in the future.

Generally, the goal of fear and anxiety research is to identify the neurobiological mechanisms of brain pathogenesis, and understand how to treat the potentially devastating effects of anxiety disorders in humans [8]. Numerous animal models have been used to investigate anxiety-related disorders in rodents, including paradigms based on exploration (open field, hole-board, elevated plus maze, light-dark, social interaction, mirror chamber) vide more insight into the properties of GABA A receptors in vivo, and their relevance in anxiety disorders.

[ 18 F]Flumazenil Radiosynthesis and Quality Controls
All synthesis and purification procedures were performed on a Tracerlab FX-FN automated synthesizer ( Figure 1A,B). [ 18 F]flumazenil was produced from the nitroflumazenil precursor via a standard nucleophilic substitution fluorination reaction, as shown in Figure 1A. The produced activity of the radioligand was 0.8-2.3 GBq, with a radiochemical yield (RCY) of 15-20% decay, corrected at the end of bombardment (EOB); the total synthesis time was 90 min. This procedure resulted in a higher quality [ 18 F]flumazenil, with a radiochemical purity > 95% after radio-HPLC and radio-thin layer chromatography (TLC). In an amber glass vial, the radiochemical purity of [ 18 F]flumazenil in 30% ethanol remained > 90% at 8 h after the end of synthesis (EOS) at room temperature ( Figure 1C,D). distribution of GABAA receptors in a rat model of anxiety. We expect that these res will promote straightforward labeling with, and purification of, [ 18 F]flumazenil, and vide more insight into the properties of GABAA receptors in vivo, and their relevanc anxiety disorders.

[ 18 F]Flumazenil Radiosynthesis and Quality Controls
All synthesis and purification procedures were performed on a Tracerlab FX-FN tomated synthesizer ( Figure 1A,B). [ 18 F]flumazenil was produced from the nitroflu zenil precursor via a standard nucleophilic substitution fluorination reaction, as show Figure 1A. The produced activity of the radioligand was 0.8-2.3 GBq, with a radioche cal yield (RCY) of 15-20% decay, corrected at the end of bombardment (EOB); the t synthesis time was 90 min. This procedure resulted in a higher quality [ 18 F]flumaz with a radiochemical purity > 95% after radio-HPLC and radio-thin layer chromatogra (TLC). In an amber glass vial, the radiochemical purity of [ 18

Behavioral Study
Fear-conditioned rats received 1 to 10 tone-foot-shock trainings, each training sisting of a 20 s tone (CS) of 80 dB, and a foot-shock of 1.2 mA (US) administered fo ( Figure 2A). The behavioral study of freezing behavior between the closed arm of 1 a using different foot-shock (0.6 and 1.2 mA) and repeat times were listed in (Supplem  tary Table S1). The rats who received 10 CS-US presentations froze significantly more t control animals (t-test: p < 0.001; Figure 2B).
The effect of the number of CS-US pairings was also elucidated. The rats exhib

Behavioral Study
Fear-conditioned rats received 1 to 10 tone-foot-shock trainings, each training consisting of a 20 s tone (CS) of 80 dB, and a foot-shock of 1.2 mA (US) administered for 3 s (Figure 2A). The behavioral study of freezing behavior between the closed arm of 1 and 2 using different foot-shock (0.6 and 1.2 mA) and repeat times were listed in (Supplementary Table S1). The rats who received 10 CS-US presentations froze significantly more than control animals (t-test: p < 0.001; Figure 2B). animal model of anxiety. Inspection of individual behavioral data showed considerable variability within each group. Therefore, we selected the best responders for subsequent imaging and autoradiographic analyses. Fear-conditioned rats received 1-10 tone-foot-shock training, consisting of a 20 s tone (CS) of 80 dB, and a foot-shock of 1.2 mA (US) administered for 3 s. This study observed that increasing the number of CS-US pairings caused the animals to exhibit a higher freezing response. (B) The rats which received 10 CS-US presentations froze significantly more than control animals. The control group (Naive) showed without freezing than CS-US group presentated. Data points represent the mean values, and the error bars represent standard deviations. **, p < 0.001 (ttest).

Ex Vivo Autoradiography
We performed ex vivo autoradiography to assess GABAA receptor availability in fear-conditioned rats. In the autoradiograms of sections taken 30 min after injection of [ 18 F]flumazenil, the specific localization of GABAA receptors in each rat brain was examined ( Figure 3). The analysis of GABAA receptor distribution in rat brains indicated that the prefrontal cortex, cortex, hippocampus, and amygdala, which are known to have high densities of GABAA receptors, displayed the most intense accumulation of [ 18 F]flumazenil. Compared to control rats, there was a significantly lower cerebral accumulation of [ 18 F]flumazenil in rats expressing contextual fear conditioning ( Figure 3A).
Brain tissues from four regions, including the prefrontal cortex, cortex, hippocampus, and amygdala, were examined bilaterally. The specific binding ratio was obtained for all target regions using the pons as a reference region. For the region-dependent differences, normal rats displayed a markedly higher specific binding ratio than fear-conditioned rats in the prefrontal cortex, cortex, hippocampus, and amygdala (p < 0.05) ( Figure 3B). The effect of the number of CS-US pairings was also elucidated. The rats exhibited high levels of freezing with successive foot-shock presentations, indicating a comparable level of learning during training. Thus, we settled on a 10-trial procedure to establish an animal model of anxiety. Inspection of individual behavioral data showed considerable variability within each group. Therefore, we selected the best responders for subsequent imaging and autoradiographic analyses.

Ex Vivo Autoradiography
We performed ex vivo autoradiography to assess GABA A receptor availability in fear-conditioned rats. In the autoradiograms of sections taken 30 min after injection of [ 18 F]flumazenil, the specific localization of GABA A receptors in each rat brain was examined ( Figure 3). The analysis of GABA A receptor distribution in rat brains indicated that the prefrontal cortex, cortex, hippocampus, and amygdala, which are known to have high densities of GABA A receptors, displayed the most intense accumulation of [ 18 F]flumazenil. Compared to control rats, there was a significantly lower cerebral accumulation of [ 18 F]flumazenil in rats expressing contextual fear conditioning ( Figure 3A).
Brain tissues from four regions, including the prefrontal cortex, cortex, hippocampus, and amygdala, were examined bilaterally. The specific binding ratio was obtained for all target regions using the pons as a reference region. For the region-dependent differences, normal rats displayed a markedly higher specific binding ratio than fear-conditioned rats in the prefrontal cortex, cortex, hippocampus, and amygdala (p < 0.05) ( Figure 3B).  The normal rats displayed markedly higher specific binding ratios than the fear-conditioned rats in prefrontal cortex, cortex, hippocampus, and amygdala (p < 0.05, unpaired t test). Black = anxiety group, white = control group. (* p value < 0.05 compared to control group).

NanoPET/CT Imaging
The regional distribution of GABAA receptors in the brain, as imaged by [ 18 F]flumazenil PET, is shown in Figure 4. After injection of [ 18 F]flumazenil, each time frame was observed in 10 min in both fear-conditioned rats and normal rats. The normal rats displayed markedly higher specific binding ratios than the fear-conditioned rats in prefrontal cortex, cortex, hippocampus, and amygdala (p < 0.05, unpaired t test). Black = anxiety group, white = control group. (* p value < 0.05 compared to control group).

NanoPET/CT Imaging
The regional distribution of GABA A receptors in the brain, as imaged by [ 18 F]flumazenil PET, is shown in Figure 4. After injection of [ 18 F]flumazenil, each time frame was observed in 10 min in both fear-conditioned rats and normal rats.
Individual dynamic images were automatically co-registered to each rat brain region template. ROIs, including the prefrontal cortex, cortex, hippocampus, amygdala, and pons, were extracted from a set of previously constructed regions. Finally, time-activity curves were generated by projecting predefined ROIs onto the dynamic images ( Figure 5).
Based on comparisons with the reference region, the specific binding ratios in the control group were significantly higher than those in the fear-conditioned group. Timeactivity curves showed persistence of high specific binding ratios of [ 18 F]flumazenil in the four target regions, and conversely for the fear-conditioned rats ( Figure 5). This phenomenon was even more profound in the last frame (50-60 min post-injection), where the comparison of fear-conditioned animals with controls showed the largest differences in specific binding ratios for all target regions. Across the target regions, the average specific binding ratios of [ 18 F]flumazenil in control rats compared to the fear-conditioned group were 1.05 vs. 0.27 in the amygdala, 1.60 vs. 0.31 in the hippocampus, 2.83 vs. 0.78 in the prefrontal cortex, and 2.17 vs. 0.71 in the cortex, for control vs. fear-conditioned rats, respectively ( Figure 5).
The density patterns of GABA A , visualized from the PET experiments, were in agreement with the results of the ex vivo autoradiography study.  Individual dynamic images were automatically co-registered to each rat brain re template. ROIs, including the prefrontal cortex, cortex, hippocampus, amygdala, pons, were extracted from a set of previously constructed regions. Finally, time-act curves were generated by projecting predefined ROIs onto the dynamic images (Fi 5).
Based on comparisons with the reference region, the specific binding ratios in Figure 4. The brain regional distribution of GABA A receptors as imaged by [ 18 F]flumazenil NanoPET/CT in fear-conditioned (anxiety) and normal rats. The acquired NanoPET/CT dynamic series were reconstructed with NanoPET/CT images, consisting of six time frames (0-10, 10-20, 20-30, 30-40, 40-50, 50-60 min). After injection of [ 18 F]flumazenil, peak uptake was observed at the first time frame (0-10 min post-injection) in both fear-conditioned rats and normal rats. In all time frames, the normal rats displayed markedly higher uptake of [ 18 F]flumazenil than the fear-conditioned rats in the prefrontal cortex, cortex, amygdala, and hippocampus. The pons was delineated as ROIs and used as a reference region.
prefrontal cortex, and 2.17 vs 0.71 in the cortex, for control vs. fear-conditioned ra spectively ( Figure 5). The density patterns of GABAA, visualized from the PET experiments, were in ment with the results of the ex vivo autoradiography study. Separate displays of the time-activity curves from the amygdala, hippocampus, cort frontal cortex, respectively. The time-activity curves were generated by projecting the speci take ratios, calculated from the pre-defined ROIs, onto the dynamic images. Based on comp to the reference region, the specific binding ratios in the control group were significantly high those in the fear-conditioned group in the four target regions.

Blocking Study
As shown previously, [ 18 F]flumazenil passes through the blood-brain barrier localized in the regions where GABAA receptors are concentrated. To further charac the nature of the GABAA receptor uptake, the effects of pre-treatment for 10 min wi µg diazepam (BZR agonist) on the distribution of [ 18 F]flumazenil were evaluated, ex vivo autoradiography and pre-treatment with 0.01-10 mg/kg diazepam for PET/CT in vivo imaging. In ex vivo autoradiography, treatment with diazepam disp a marked decrease of [ 18 F]flumazenil accumulated in all brain regions ( Figure 6). pected for a receptor ligand in NanoPET/CT imaging, increasing the dose of diaz resulted in reduced cerebral uptake of [ 18 F]flumazenil. A rapid and almost complete out of [ 18 F]flumazenil was visible at 10 mg/kg diazepam, afforded in all brain region specific uptake in the frontal cortex, cortex, and the hippocampus were significan minished by pre-treating rats with diazepam (0.01-10 mg/kg, i.v.). Pre-treatmen higher concentrations of diazepam resulted in fewer GABAA receptors being availa binding, and [ 18 F]flumazenil was eliminated from the brain at a rapid rate (Figure 7 (A-D) Separate displays of the time-activity curves from the amygdala, hippocampus, cortex and frontal cortex, respectively. The time-activity curves were generated by projecting the specific uptake ratios, calculated from the pre-defined ROIs, onto the dynamic images. Based on comparisons to the reference region, the specific binding ratios in the control group were significantly higher than those in the fear-conditioned group in the four target regions.

Blocking Study
As shown previously, [ 18 F]flumazenil passes through the blood-brain barrier and is localized in the regions where GABA A receptors are concentrated. To further characterize the nature of the GABA A receptor uptake, the effects of pre-treatment for 10 min with 200 µg diazepam (BZR agonist) on the distribution of [ 18 F]flumazenil were evaluated, using ex vivo autoradiography and pre-treatment with 0.01-10 mg/kg diazepam for NanoPET/CT in vivo imaging. In ex vivo autoradiography, treatment with diazepam displayed a marked decrease of [ 18 F]flumazenil accumulated in all brain regions ( Figure 6). As expected for a receptor ligand in NanoPET/CT imaging, increasing the dose of diazepam resulted in reduced cerebral uptake of [ 18 F]flumazenil. A rapid and almost complete washout of [ 18 F]flumazenil was visible at 10 mg/kg diazepam, afforded in all brain regions. The specific uptake in the frontal cortex, cortex, and the hippocampus were significantly diminished by pre-treating rats with diazepam (0.01-10 mg/kg, i.v.). Pre-treatment with higher concentrations of diazepam resulted in fewer GABA A receptors being available for binding, and [ 18 F]flumazenil was eliminated from the brain at a rapid rate (Figure 7). out of [ 18 F]flumazenil was visible at 10 mg/kg diazepam, afforded in all brai specific uptake in the frontal cortex, cortex, and the hippocampus were si minished by pre-treating rats with diazepam (0.01-10 mg/kg, i.v.). Pre-t higher concentrations of diazepam resulted in fewer GABAA receptors bein binding, and [ 18 F]flumazenil was eliminated from the brain at a rapid rate (

Discussion
Previously, [ 18 F]flumazenil synthesis was conducted via isotopic exchange of 19 F for 18 F, using flumazenil as the precursor during synthesis [40][41][42]. The development of nitroflumazenil as a precursor has resulted in high specific activity, which is essential in brain receptor imaging studies. However, a long synthesis time of approximately 75 min has been reported when using the same precursor. Moreover, after radiosynthesis, the needed purification of [ 18 F]flumazenil using semi-preparative HPLC can result in low specific activity, as it is time-consuming process (50 min), and the procedure has been difficult to automate, resulting 18 F radioactive decay [43,44]. Our study represents a straightforward carrier-free nucleophilic radio-fluorination method, using an automatic synthetic technique with solid extraction purification to replace semi-preparative HPLC, based on a fully automatic radiosynthesizer labeled with nitro-flumazenil precursor ( Figure 1A). The total synthesis time of [ 18 F]flumazenil was 90 min, and the obtained [ 18 F]flumazenil exhibited high radiochemical purity (radio-HPLC (> 90%) and radio-TLC (> 90%), in Figure  1C,D), although the radiochemical end-of-synthesis (EOS) yield was approximately 15-20% lower than the reference (30%) [43,44]. Overall, we report a very convenient method for obtaining adequate yields and radiochemical purity of [ 18 F]flumazenil.

Discussion
Previously, [ 18 F]flumazenil synthesis was conducted via isotopic exchange of 19 F for 18 F, using flumazenil as the precursor during synthesis [40][41][42]. The development of nitroflumazenil as a precursor has resulted in high specific activity, which is essential in brain receptor imaging studies. However, a long synthesis time of approximately 75 min has been reported when using the same precursor. Moreover, after radiosynthesis, the needed purification of [ 18 F]flumazenil using semi-preparative HPLC can result in low specific activity, as it is time-consuming process (50 min), and the procedure has been difficult to automate, resulting 18 F radioactive decay [43,44]. Our study represents a straightforward carrier-free nucleophilic radio-fluorination method, using an automatic synthetic technique with solid extraction purification to replace semi-preparative HPLC, based on a fully automatic radiosynthesizer labeled with nitro-flumazenil precursor ( Figure 1A). The total synthesis time of [ 18 F]flumazenil was 90 min, and the obtained [ 18 F]flumazenil exhibited high radiochemical purity (radio-HPLC (>90%) and radio-TLC (>90%), in Figure 1C,D), although the radiochemical end-of-synthesis (EOS) yield was approximately 15-20% lower Pharmaceuticals 2023, 16, 417 9 of 15 than the reference (30%) [43,44]. Overall, we report a very convenient method for obtaining adequate yields and radiochemical purity of [ 18 F]flumazenil.
The purification of crude [ 18 F]flumazenil was previously carried out using acetonitrile and water as the mobile phase [31,[45][46][47]. However, acetonitrile is flammable, volatile, toxic, and must be removed during the purification process. In contrast to acetonitrile, ethanol is less toxic, less expensive, and has lower disposal costs [29]. Therefore, we used ethanol instead of acetonitrile, and the purified product was directly diluted for injection. Using this radiosynthesis process, there are many promising aromatic ring compounds that could be substituted more conveniently [48][49][50]. However, as analyzed by radio-TLC, [ 18 F]flumazenil demonstrated good stability, maintaining a purity of up to 90%, after 8 h in ethyl acetate:ethanol (80/20) at room temperature ( Figure 1D).
In the present study, we used [ 18 F]flumazenil, which has similar in vivo binding characteristics to [ 11 C]flumazenil, to address the half-life disadvantage and to reduce errors during the long PET scanning time (90 min) [24]. Fear conditioning has become widely regarded as a valid experimental model for clinical anxiety disorders [13]. It is important to note that most neurobiological studies employ simple temporal pairings of CS and US to produce fear conditioning [51,52]. Previous studies have shown that fear conditioning can be rapidly instilled in animals, even after a single conditioning trial, and is usually maintained for long periods [14,53]. This study observed that by increasing the number of CS-US pairings, the animals exhibited higher freezing responses, indicating a comparable level of learning during the training. Additionally, the effect of shock intensity on learning was elucidated. The rats receiving 1.2 mA shocks exhibited increased freezing responses compared with those receiving 0.6 mA shocks, suggesting a significant impact of the shock intensity on associative memory formation [27]. In this study, we investigated GABA A receptor availability using [ 18 F]flumazenil PET in a rat model of anxiety. We found a significant reduction in GABA A receptor availability in the amygdala, prefrontal cortex, cortex, and hippocampus in fear-conditioned animals. Our rat autoradiography results also supported the PET imaging findings.
Clinical studies have demonstrated that the GABA A receptor complex plays a central role in the modulation of anxiety. Behavioral and physiological lines of evidence suggest that GABAergic transmission in the amygdala undergoes plastic changes which contribute to the control of fear and anxiety [45,46]. Neuroimaging studies in humans have indicated that some brain regions may be revealed as a conditioned-fear circuit, which includes the prefrontal cortex, anterior cingulate, and additional cortical and subcortical structures [28,40,45]. The currently available therapeutic agents that act on this receptor are effective anxiolytics; infusions of GABA or GABA A receptor agonists into the amygdala decreased measures of fear and anxiety in several animal species, while infusions of GABA antagonists tended to have anxiogenic effects [46,47]. This suggests that the neuroinhibitory processes of GABA A receptors may serve as target sites for drug action [2,39]. This is the first study to utilize [ 18 F]flumazenil PET to visualize GABA A receptor distribution in vivo in an animal model of anxiety. By utilizing [ 18 F]flumazenil NanoPET imaging, we observed a global reduction in GABA A receptor availability in rats expressing contextual fear compared to controls. Furthermore, the specific binding ratios of [ 18 F]flumazenil were significantly reduced in the brain regions of the prefrontal cortex, cortex, amygdala, and hippocampus in fear-conditioned rats, which clearly depicts the effects of fear conditioning on GABA A receptors. Our findings have demonstrated that the specific binding ratios of the GABA A receptor are profoundly influenced by brain regions, such as the amygdala, prefrontal cortex, cortex, and hippocampus, suggesting that these brain regions may be involved in processing and storing conditioned fear memory.
To further characterize the nature of GABA A receptor uptake, the effects of pretreatment with 200 µg diazepam (BZR agonist) were studied in ex vivo autoradiography, and pre-treatment with 0.01-10 mg/kg diazepam for NanoPET/CT in vivo imaging. Presaturation studies have shown a high displaceable component for diazepam administration in the frontal cortex, cortex, amygdala, and hippocampus contrasted to [ 18 F]flumazenil, indicating that the structural modification of [ 18 F]flumazenil results will show in high affinity for GABA A /cBZRs (central benzodiazepine receptors) [23,30,31].
Associative learning in fear-conditioned animals is thought to involve different forms of activity-dependent synaptic plasticity [49]. Although previous data and the results from the present NanoPET study implicate these brain regions in contextual conditioning, the exact contribution is still unclear. This reduction in GABAergic neurotransmission may be caused by any or all of the following: loss of GABAergic inter-neurons, loss of GABA A receptors, or changes to GABA A receptor subunits, leading to alterations in receptor properties [27]. The most interesting result of this study is the identification and elucidation of the involvement of GABA A receptors in animals in the expression of contextual fear responses. The results of this study emphasize the importance of careful study design and control of experimental conditions when imaging GABA A receptors. Further studies incorporating cued fear and fear extinction as behavioral measures may elucidate the exact role of GABA A receptors. We suggest that these results will facilitate the development of new treatment options for patients with anxiety disorders in the long run.

[ 18 F]Flumazenil Radiosynthesis
All of the chemicals were purchased from commercial vendors and were of pharmaceutical grade.

Animals
Animal housing and experiments were approved by the Ethical Animal Use Committee of the Institute of Nuclear Energy Research Atomic Energy Council (INER), and performed in compliance with Taiwan's laws for the care and use of laboratory animals (LAC-2019-0253). All experiments were carried out in male Sprague Dawley rats (3-4 weeks), purchased from a licensed breeder (BioLASCO Taiwan Co., Taipei, Taiwan). The rats were maintained at a room temperature of 21 ± 2 • C with 50 ± 20% humidity under an automatic 12 h light and 12 h dark cycle, with food and water provided ad libitum. All experiments were performed during the activity phase of the animals, between 9:30 and 17:30. The rats were allocated to two groups: one group used for fear conditioning studies, and the other the control group.
The experimental setup of fear-conditioned animals has been explained in detail in previous studies [54][55][56]. The rats were placed in a fear conditioning chamber with a metal grid for shock application. After a 2 min adaptation period, a tone cue was presented at 80 dB for 20 s. An electric foot-shock (1.2 mA) was administered via the metal grid during the last 3 s of the tone presentation, and co-terminated with the tone. After the shock presentation, a second identical trial preceded an inter-trial interval of 1 min. Rats were trained with 1-10 CS-US pairings to investigate the effect of the number of CS-US pairings.
The freezing behavior of the animals was recorded using a video camera. Freezing behavior was defined as immobility, except for respiration movements. Freezing responses in different conditions were measured over the entire 3 min interval and expressed as a percentage of the observation interval. Freeze (%) = (Freeze time/Total time) × 100%. Control animals underwent the same procedure but did not receive a CS-US presentation. The rats received 10 tone-foot-shock pairings and subsequently NanoPET/CT imaging and ex vivo autoradiography of their brains was performed.

Ex Vivo Autoradiography
The fear-conditioned and control rats (three rats per group) received an injection of 37 MBq/200 µL [ 18 F]flumazenil, via the lateral tail vein, 2 h after the fear-conditioning experiments. The animals were sacrificed by the administration of carbon dioxide 30 min after the injection. The brains were rapidly removed, placed in optimal cutting temperature compound (OCT) embedding medium, and frozen with powdered dry ice. The brains were cut into 20 µm thick sagittal sections on a cryostat (LEICA CM3050S, Wetzlar, Germany), thaw-mounted on gelatin-coated microscope slides, and air dried at 25 • C. The slides containing the brain sections were exposed to a film (BAS-SR2040, Fuji Photo Film) in an autoradiographic cassette for 72 h. The optical densities of the autoradiograms were determined with an image analysis system using the FLA-5000 and MultiGauge ver 2.0 software (Fuji Photo Film, Akasaka, Minato, Tokyo, Japan).
The specific binding ratio in all target regions (prefrontal cortex, cortex, hippocampus, and amygdala) could be obtained using the pons as a reference region. Specific binding ratios were calculated using the following equation: (target region-reference region)/reference region.

NanoPET/CT Image Acquisition and Analysis
In an independent experiment, rats were transported to an imaging facility 2 h after receiving 10 tone-foot-shock pairings. The fear-conditioned and control rats (five rats per group) were anesthetized with 3% isoflurane gas and injected intravenously with [ 18 F]flumazenil (37 MBq/200 µL) via the lateral tail vein. The apparent behavioral status and vital signs of the rats were monitored during the scan to adjust for anesthesia. The temperature of the animals was maintained at 37 • C. For the PET study, the rats were positioned in a NanoPET/CT PLUS system (Bioscan Europe, Ltd., Paris, France), using the dynamic mode for data acquisition from 0 to 60 min (a series of six frames, 6 × 10 min).
An acquired PET dynamic series was reconstructed with PET images consisting of six time frames (six frames × 10 min). The frames were then reconstructed using 3D Monte Carlo reconstruction with a matrix size of 128 × 128 pixels. Reconstructed PET data were further processed using PMOD ver 3.3 software (PMOD Technologies Ltd., Fällanden, Switzerland) and co-registered to a CT template. The PMOD v3.3 rat brain regions template was used to define the regions of interest (ROI). On the CT template, the amygdala, hippocampus, cortex, prefrontal cortex, and pons were delineated as ROIs. The PET data were matched to the respective CT data sets. The specific binding ratio in each ROI was computed for each frame, resulting in the generation of time-activity curves.

Blocking Study
In vivo competitive binding of 200 µg diazepam (benzodiazepine receptor agonist, BZR agonist) in the brain was investigated in normal rats using [ 18 F]flumazenil labeling. After 10 min of pretreatment with diazepam (BZR agonist) through intraperitoneal injection (0.01, 0.1, 1, and 10 mg/kg), the rats received an injection of 37 MBq/200 µL [ 18 F]flumazenil. The pre-treated rats were imaged using NanoPET/CT following the same procedures described above for the control rats. Occupancy (%) = (ROI of control group brain region − ROI of pretreated group brain region/ROI of control group brain region) × 100%.

Statistical Analysis
Data points represent mean values and error bars represent standard deviations. Statistical analyses were performed using the unpaired two-tailed Student's t-test, unless otherwise noted; p values less than 0.05 were considered statistically significant, and significance levels in graphs are marked as follows: * p < 0.05, ** p < 0.01, and *** p < 0.001.

Conclusions
The straightforward radio-labeling and purification procedure described in this study can be easily adapted by commercially available modules for the high radiochemical purity of [ 18 F]flumazenil. The results demonstrate that this fully automated nucleophilic [ 18 F]fluorination system can replace traditional methods, and the product of [ 18 F]flumazenil can be utilized in PET imaging. To quantify GABA A availability in in vivo NanoPET/CT and ex vivo autoradiography in an animal model of anxiety.
Supplementary Materials: The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/ph16030417/s1, Table S1: The behavioral study of freezing behavior between the closed arm of 1 and 2 using different foot-shock (0.6 and 1.2 mA) and repeat times.  Institutional Review Board Statement: The animal study protocol was approved by the Taipei medical university institutional animal care and use committee (protocol code LAC-2019-0253) for studies involving animals.
Informed Consent Statement: Not applicable.

Data Availability Statement:
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.