Improved Automated Radiosynthesis of [11C]PBR28

Microglial activation is commonly identified by elevated levels of the 18 kDa translocator protein (TSPO) in response to several inflammatory processes. [11C]PBR28 is one of the most promising PET tracers to image TSPO in both human and non-human primates. In this study, we optimized the radiolabeling procedure of [11C]PBR28 for higher radiochemical yield, radiochemical purity, and specific activity, which can be easily translated to any automated module for clinical trials. Time-activity curves (TACs) derived from the dynamic PET imaging of male rhesus monkey brains demonstrated that [11C]PBR28 had suitable kinetics with radiotracer accumulation observed in the caudate, putamen, cerebellum, and frontal cortex region.


Introduction
Expression of the peripheral benzodiazepine receptors (PBR), recently described as the 18 kDa translocator protein TSPO, is considered to be a hallmark for microglial activation [1,2]. TSPOs are ubiquitous in active cerebral phagocytic cells and located in the outer mitochondrial membranes of peripheral organelles, including the brain, heart, kidney, liver, and lungs [3][4][5]. Several putative biological functions such as cell proliferation, cholesterol transportation, immune alterations, and apoptosis are associated with the transmembrane channels of the mitochondrial membrane, which are considered as a depository for TSPO [1,[6][7][8][9][10]. Significant levels of TSPO are observed during neuroinflammation, but absent in the resting microglial CNS parenchyma [2,[11][12][13]. Moreover, neurodegenerative disorders including Alzheimer's disease (AD), Wernicke's encephalopathy, epilepsy, Huntington's disease, and cerebral ischemia have demonstrated increased TSPO in the cerebellum, olfactory lobes during disease development and progression stages, and have stimulated researchers to improve targeted neuroinflammation therapies [12][13][14]. However, accurate monitoring of the inflammatory processes related to each disease state will be required to determine if these therapies will be effective. Imaging microglial activation by targeting TSPO may provide an efficient index of the disease progression, and enhance the therapeutic planning for diseases affected by neuroinflammatory processes [15].
The TSPO targeting radiotracer, [ 11 C]PK11195 has been used for the PET imaging of cerebral inflammation and was utilized to investigate microglial activation in a variety of animal models of neuroinflammation [16][17][18][19][20][21]. Additionally, [ 11 C]PK11195 was evaluated in humans for neurological disease conditions such as multiple sclerosis, epilepsy, Parkinson's disease, and AD for more than two decades [22][23][24][25]. However, despite its widespread use in neuroinflammation imaging, it demonstrates a low brain extraction, resulting in a low signal-to-noise ratio [18]. Additional studies have reported the accumulation of [ 11 C]PK11195 in the regions of the brain that are not traditionally associated with disease processes resulting in low sensitivity, poor quantification, and the inability to image milder forms of neuroinflammation [26][27][28][29][30]. Moreover, several unidentified radiometabolites were observed after the administration of [ 11 C]PK11195 [31,32] [42,43,45]. Biodistribution studies and PET using [ 11 C]PBR28 enabled researchers to accurately quantify cerebral artery inflammation caused by occlusion, and its pharmacokinetics in humans was concordant with the data obtained in non-human primate models [35,46].
Recently, a lot of interest has been vested on investigating the imaging properties of [ 11 C]PBR28 as a radiopharmaceutical to image TSPO in other organs and animal models [40][41][42]. Several new automation methods have been published recently for the synthesis [47][48][49]; however, the reported methods use a stronger methylation agent and base for the reaction, i.e. [ 11 C]MeOTf as the methylating agent and sodium hydride as the base [47][48][49]. The radiochemistry method using [ 11 C]MeOTf and sodium hydride (NaH, 60% in mineral oil) followed by sonication in acetonitrile does not seem to be a robust technique to get consistent radiochemical yields. Our study describes a simple route for the synthesis and radiochemical synthesis of [ 11 C]PBR28 which resulted in a higher radiochemical yield over the previously reported methods [47][48][49] with high specific activity that can be easily adapted to any automated module, for example the most commonly used ones like GE FXC-pro, GE FXMeI-FXM, and TRASIS AIO modules. We used a simple reaction vial method to load the precursor in DMSO and then added a minimum of 1.9 µL of 5 N NaOH as the base. [ 11 C]MeI was bubbled into the reaction mixture vial and was heated only to 60°C for 5 min. The radiochemical yield was ~45-55% and the specific activity ranged from 8000-9500 mCi/μmol (n=15, decay corrected to EOS).

Quality Control Analysis of [ 11 C]PBR28
[ 11 C]PBR28 purity was assessed using an analytical reversed-phase Phenomenex Prodigy C18 analytical HPLC column (250 × 4.6 mm, 5 µm) and UV detection set at 254 nm. The mobile phase (1.5 mL/min) consisted of 70% acetonitrile and 30% 0.1 M aqueous ammonium formate pH 6.0-6.5 buffer solution. [ 11 C]PBR28 showed retention at 6.0 min, and authentication of the product was performed with co-injection of the non-radioactive standard PBR28, which demonstrated similar retention times.

Image Processing and Time-Activity Curves
PET PBR28 images were acquired on a General Electric 16-slice PET/CT Discovery ST Scanner which has 24 detector rings that provide 47 contiguous image planes over a maximum 70 cm transaxial field of view with CT attenuation correction. Axial spatial resolution of this scanner is 3.27 mm at the center of the gantry. Approximately 30 min prior to the scan, the monkey was anesthetized with ketamine (10 mg/kg, i.m.) and transported to the PET Center. Anesthesia was maintained during the scan by inhaled isoflurane (1.5%). The monkey was placed in the scanner and a catheter was inserted into an external vein for tracer injection and fluid replacement throughout the study. Body temperature was maintained at 40°C and vital signs (heart rate, blood pressure, respiration rate, and temperature) were monitored throughout the scanning procedure. An initial low dose CT-based attenuation correction scan was acquired. Next, [ 11 C]PBR28 was injected and a 120-min dynamic acquisition scan was acquired. Thirty-three frames were acquired over 120 min (6 x 30 s, 3 x 60 s, 2 x 120 s, 22 x 300 s) in 3D mode (i.e., septa retracted). Image reconstruction of the 3D data was done using the 3D-reprojection method with full quantitative corrections. Emission data was corrected for attenuation and reconstructed into 128 × 128 matrices using a Hanning filter with a 4-mm cut-off transaxially and a ramp filter with an 8.5-mm cut-off axially. Data analysis was conducted using PMOD Biomedical Image Quantification Software (version 3.5; PMOD Technologies, Zurich, Switzerland). Brain uptake was defined by its standardized uptake value (SUV) calculated by dividing the tracer concentration in each pixel by the injected dose per body mass. ROIs for the basal ganglia and cerebellum were drawn and time-activity curves were generated.

Chemistry
Using a slight modification of the previously reported methods [48,49], the [ 11 C]PBR28 p-phenol precursor, 6, and the corresponding nonradioactive standard, PBR28, were synthesized in higher chemical yields and a shorter reaction time (Scheme 2). Details of the synthesis have been provided in the "Supporting Information" folder. Briefly, 4-chloro,3-nitro pyridine 1 underwent a base-assisted substitution reaction with phenol resulting in 3-nitro-4-phenoxypyridine, 2, which upon reduction gave the corresponding amino compound 3. The amine, 3, was then subjected to a condensation reaction with salicylaldehyde, followed by an in situ NaBH 4 reduction of the Schiff base to give the secondary amine, 4. The amine, 4, was then acetylated with acetyl chloride to give the diacetylated intermediate 5, which upon selective LiOH-assisted O-deacetylation resulted in the corresponding [ 11 C]PBR28 phenol precursor, 6. The amine intermediate, 3, was condensed with 2-methoxy-benzaldehyde to form the Schiff's base in situ, which was reduced by NaBH 4 to yield the secondary amine, compound 7. Compound 7 was Nacetylated to give the non-radioactive standard, PBR28.

Radiochemistry
The radiochemical synthesis of [ 11 C]PBR28 was investigated using different reaction conditions (n=3) ( In our hands, the use of 5 N NaOH led to the synthesis of [ 11 C]PBR28 in a 45-55% radiochemical yield with >99% radiochemical purity, greater than has been previously reported in the literature [47,48] and a specific activity of approximately 8000-9500 mCi/μmol (n=15, decay-corrected to EOS). Further, we optimized the radiolabeling procedure of [ 11 C]PBR28 in GE-FXC, GE-FXMeI/FXM, and TRASIS AIO modules at the Wake Forest PET Center for clinical trials, using the least amount of 1.9 µL 5 N NaOH for the methylation step, to bring down the UV mass to <4.0 µg/mL in the final dose. Due to the improvement in the radiochemical yield, the synthesis of [ 11 C]PBR28 can be easily translated to any automated radiochemistry modules around the world.

Monkey PET Image Analysis
Dynamic small animal PET imaging was performed on male rhesus monkeys (n=3), which received an intravenous injection of [ 11 C]PBR28. [ 11 C]PBR28 demonstrated high radioactive uptakes in monkey brain regions, with greater activity localizing in the grey matter. Figure 2 represents the dynamic PET image of [ 11 C]PBR28 (100 min) in a monkey brain.  (Figure 2). After 10 min postinjection of [ 11 C]PBR28, the radioactive uptake obtained in the cerebellum was 0.1 (%ID/cc) and the basal ganglion was 0.091 (%ID/cc). Additionally, the [ 11 C]PBR28 radioactive uptake pattern in these regions of the brain was consistent with cerebral TSPO distributions [35,40,44,45]. Regional cerebral distribution of [ 11 C]PBR28 was rather homogenous by the first 12 min after injection and no blatant accumulation was observed. As depicted by the TACs for the [ 11 C]PBR28 tracer, initial uptake in the brain was satisfactory and by approximately 2 h p.i., the radioactivity was washed out from all regions of the brain. The lipophilicity of a tracer affects its binding and distribution and especially with brain-related tracer development, it is an important determinant of brain penetration (through the blood-brain barrier) [49,50]. Clinical PET imaging studies, especially with C-11 PET probes, depend on tissue retention and clearance times [51].
From the TACs of [ 11 C]PBR28, the pharmacokinetics are acceptable for probing the TSPO receptor density in the brain.

Conclusion
In summary, we report the modified radiolabeling procedure for [ 11 C]PBR28 with a high radiochemical yield, radiochemical purity, and specific activity to be directly translated and easily adapted to any automated modules for human injections and clinical trials. We further validated the radioactive uptakes of [ 11 C]PBR28 in brains of male rhesus macaques using PET imaging studies, and [ 11 C]PBR28 demonstrated favorable pharmacokinetics to image TSPO, which thus possesses a high potential to be a valuable in vivo PET tracer for imaging several neuroinflammatory processes, both in research and clinical settings. This study strongly reinforces the utility of the TSPO radiotracers to be further evaluated as neuroinflammation imaging agents.

Acknowledgement
The authors thank Wake Forest Cyclotron Facilities for [ 11 C]CO 2 production and the Non-Human Primate Animal Imaging Core for conducting the monkey PET imaging studies.

Chemistry
All reagents were purchased from Sigma-Aldrich and were used without additional purification. All reactions were carried out using anhydrous solvents unless otherwise stated. 1 H-NMR was measured by the Varian ® 300 MHz NMR spectrometer and all chemical shifts are reported as ppm (δ). Melting points were measured using the Electrothermal Mel-Temp ® 3.0 melting point apparatus. All reactions were monitored by analytical thin-layer chromatography (TLC), and all UV-active spots were detected using the Mineralight® Lamp UVGL-25 UV lamp.

PBR28 Synthesis
Based on the reported literature chemical methods [48,49], the syntheses of radiolabeling precursor 6 and its non-radioactive standard PBR28 were carried out as depicted in Scheme 1 with slight modifications for better chemical yields and shorter reaction times.

N-(2-Methoxybenzyl)-4-phenoxypyridin-3-amine (7)
2-Methoxybenzaldehyde (0.80 mL, 5.90 mmol) was added to a solution of 3 (1.0 g, 5.37 mmol) in anhydrous toluene (20 mL) and refluxed using a Dean-Stark apparatus for 24 h. The reaction mixture was evaporated under reduced pressure and the residue was then dissolved in methanol (50 mL). NaBH 4 (0.75 g, 19.8 mmol) was added slowly to the above residue in methanol at 0°C, followed by stirring at RT for 1 h. The reaction mixture was further quenched with 5% aqueous acetic acid (17 mL) and then extracted with ethylacetate (3 × 50 mL). The organic layers were combined, washed with saturated NaHCO 3 solution (2 × 150 mL), brine (2 × 150 mL), and then concentrated to yield product 7 as a yellow thick oil (1.0 g, 3.26 mmol, 61%) and was used in the next step without any additional purification. 1