Male Sprague–Dawley rats (n = 61, Harlan Laboratories), weighing 364 ± 38 g, were treated in accordance with the European Ethics Committee (decree 86/609/CEE). The study protocol was approved by the Animal Experimental Ethical Committee of the University of Antwerp, Antwerp, Belgium. The animals were kept in individually ventilated cages under controlled conditions (12 h normal light–dark cycles, 20–23°C, and 50–55% relative humidity) with water and rodent food pellets ad libitum. The study design (Figure 1) was identical for both radiotracers ([¹¹C]RAC: n = 29 − [¹¹C]ABP688: n = 32). For each radiotracer the design is subdivided in three separate experimental groups: a reference μPET group without implants; a dummy implant μPET group; and a beta-microprobe 0.75 mm OD implant group. Further subdivisions were made based on different experimental conditions such as injected dose (TD vs. HD) and implant diameter. All animals received either a high dose (target dose of 65 MBq; cf. Table 1) or a low mass tracer dose being lower than 0.5 nmol/kg (Schiffer et al., 2005) for [¹¹C]RAC and lower than 3.0 nmol/kg (Wyckhuys et al., 2013) for [¹¹C]ABP688. Every radiotracer administration was combined with either a reference μPET-CT scan without implantations (Figure 1A), a dummy probe implant combined with a μPET-CT scan (Figure 1B), or a beta-microprobe implant and acquisition (Figure 1C). The dummy μPET experimental group was further subdivided in separate implant groups depending on the OD and thus the inflicted tissue trauma (Figure 3). These dummy groups were mandatory since simultaneous μPET and beta-microprobe acquisitions are practically unfeasible as the actual scintillation probes are non-disposable and cannot be permanently fixed with dental cement (cf. Section Experimental Procedure). Three out of 64 subjects (n = 61) were left out for further analysis due to surgical complications (bleeding) or when the maximal allowed cold mass was exceeded for TD conditions.

For the reference μPET-CT scans (Figure 1A), the animals were anesthetized (2% isoflurane mixed with medical oxygen). After the insertion of a catheter in the tail vein for tracer administration, the animals were positioned on the heated bed (37.5°C) of the μPET-CT scanner.

For beta- and dummy probe implantations (Figures 1B,C), the animal was anesthetized (2% isoflurane mixed with medical oxygen) ~90 min before radiotracer administration and placed on a heated blanket (37.5°C). Before surgery, a catheter was placed in the tail vein for tracer administration. Subsequently, the head of the animal was fixated in a stereotaxic frame (Kopf Instruments, California, USA). After making an anteroposterior incision toward the neck region, the skull was exposed by retraction of the skin and muscle layers. Two holes were drilled in the skull for probe positioning, one at the level of the right striatum and another at the level of the cerebellum. The anteroposterior (AP) and mediolateral (ML) stereotaxic coordinates were determined relative to the bregma. The surface of the dura functioned as a reference point for the dorsoventral (DV) coordinate. Implant coordinates were obtained from a rat brain atlas (Paxinos and Watson, 2013): AP +0.2, ML 3.0, DV −6.0 for the striatum and AP −11.8, ML 0.0, and DV −3.5 for the cerebellum. Finally, the dummy probes (0.75 mm or 1.00 mm OD) or beta-microprobes (0.75 mm OD) were positioned. Beta-microprobe animals remained on the heated blanket and were coupled to the light-tight Beta Scintillator System Twin (Swisstrace, Zurich, Switzerland; Weber et al., 2003; Wyss et al., 2007), awaiting immediate tracer administration. Dummy probe animals were transferred to the μPET-CT scanner (Siemens Preclinical Solution, Knoxville, TN, USA) after fixation of the dummy probes on the skull with dental cement (prorep propal, Belgica Dental, Belgium).

Concerning the synthesis of both radiotracers, [¹¹C]RAC was synthesized by O-methylation of (S)-O-desmethylraclopride precursor (ABX, Raderberg, Germany) with ¹¹C-CH₃O₃SCH₃. ¹¹C-CO₂ was produced in an Eclips HP cyclotron (Siemens, Knoxville, TN, USA) by irradiation of a N₂ target for 50 min with an 11 MeV proton beam and a beam intensity of 60 μA. The ¹¹C-CO₂ was converted to ¹¹C-CH₃O₃SCH₃ for reaction with (S)-O-desmethylraclopride via ¹¹C-CH₃I. The radiochemical purity was determined by reverse phase high-performance liquid chromatography (HPLC) (Waters Symmetry C18, 3.5 μm, 4.6 × 50 mm) and was ≥99.9%. The purified [¹¹C]RAC was dissolved in a mixture of NaH₂PO₄ 0.005M pH 4.5, 0.9% NaCl (B. Braun, Germany) and ethanol 96% v/v (BP, Eur.Ph., Merck-Chemicals, Darmstadt, Germany). The solution was sterile filtered for in vivo use. [¹¹C]ABP688-E is prepared by reaction of 0.5 mg desmethyl-ABP688 (E/Z) with ¹¹C-CH₃SO₃CF₃ in 400 μL acetone in the presence of 10 μL 1N NaOH for 4 min at room temperature. For purification, the crude reaction mixture was injected into an analytical HPLC column (WatersXBridge C18, 5 μm, 4.6 × 150 mm) with an eluent of 0.05 M sodium acetate (pH 5.5) and ethanol (55/45, v/v) at a flow rate of 1 mL/min to isolate the [¹¹C]ABP688-E. The purified [¹¹C]ABP688-E is filtrated through a sterile Millipore Millex-GV filter (0.22 μm) and diluted with 0.9% NaCl through the same filter to reduce the ethanol concentration to <10%. The average specific activity, injected dose, and injected mass dose for all experimental groups are shown in Table 1. In order to keep the injected mass dose below the imposed limit of 0.5 nmol/kg for [¹¹C]RAC and 3.0 nmol/kg for [¹¹C]ABP688, injected doses in the TD groups accurately compensate for the varying specific activity of each radiotracer production.

The reference μPET and the dummy implant groups received a dynamic μPET-CT scan (Figures 1A,B) performed on two Siemens Inveon μPET-CT scanners (Siemens Preclinical Solution, Knoxville, TN, USA) with the following specifications: 60-min dynamic μPET acquisitions with 2 × 10 s, 3 × 20 s, 3 × 30 s, 3 × 60 s, 3 × 150 s, 9 × 300 s frames, followed by a 7-min μCT. Scans were started immediately after positioning the animal on the scanner and administration of a bolus of radiotracer via the tail vein. All animals were randomly distributed on both scanners.

For beta-microprobe recordings (Figure 1C), the light-tight Beta Scintillator System Twin was switched on at least 30 min before recording to ensure a sufficient warm-up and stabilization of the photomultiplier tubes (Ginovart et al., 2004). The system was equipped with 0.75 mm OD scintillator probes. After implantation, recordings at a sampling rate of 1 Hz were started 1 min before tracer administration for count stabilization followed by a 60-min acquisition using PMOD v3.0 (PMOD technologies, Zurich, Switzerland). Finally, the animals were removed from the stereotaxic frame and the scintillator probes were cleansed with water. All animals were injected with an overdose of pentobarbital (>30 mg/kg) at the end of each acquisition (beta-microprobe and dummy μPET). Upon immediate removal, their brain was submerged in cooled 2-methylbutane and stored at −20°C for a subsequent histological verification of the probe position in coronal cryosections (20 μm). Cryosections confirmed correct dummy probe localization for all subjects.

Both μPET scanners were calibrated using a uniform cylinder with known activity (±37 MBq; room temperature) to obtain a calibration factor for conversion of images to kBq/cc. Calibration (sampling rate: 1 Hz; duration: 60 s) of beta-microprobes was achieved by immersion in an aqueous solution (10 mL) with a known activity concentration of [¹¹C]ABP688 (±20 MBq; room temperature). Probes were repetitively calibrated (similar experimental conditions) in order to determine an averaged calibration factor for each probe.

For quantitative image analysis, μPET images were reconstructed using two-dimensional ordered subset expectation maximization with four iterations and 16 subsets after Fourier rebinning. The images were reconstructed on a 128 × 128 × 159 grid with a voxel size of 0.776 × 0.776 × 0.796 mm. Normalization, dead time correction, random subtraction, CT-based attenuation correction, and single-scatter simulation scatter corrections were applied. For both tracers, reconstructed images were processed in PMOD v3.3. A static image corresponding to the time-averaged frames of each dynamic acquisition was spatially transformed through brain normalization to a rat brain template for [¹¹C]RAC and [¹¹C]ABP688 (in-house developed; Verhaeghe et al., 2014). Both PET templates already corresponded to a standardized MR template space (Schiffer rat MR; available in PMOD v3.3) with corresponding VOI definition. The obtained matrix from the brain normalization step was applied to transform all dynamic scans to the [¹¹C]RAC and [¹¹C]ABP688 template space, respectively. The time-activity curves of the striatum and cerebellar VOI were extracted from the resulting images via the superimposition of a VOI template. The extracted time-activity curves served as input for the SRTM (Lammertsma and Hume, 1996). This kinetic modeling method allows quantification of receptor kinetics in a region of interest via the expression of tracer uptake in this tissue in terms of its uptake in a reference region (devoid of the receptor-of-interest) with a similar level of non-specific binding. Based on this method (which is implemented in PMOD software), the BP_ND and relative delivery R1 of the tracer to the striatal VOI were calculated, with the cerebellum as a reference region. In addition, pixel wise kinetic modeling, using SRTM2 (Wu and Carson, 2002), was applied to generate parametric BP_ND and R1 maps.

For each beta-microprobe experiment, the recorded time-activity curves of the striatum and cerebellum were decay corrected for ¹¹C with respect to the timing of radiotracer injection (PMOD v3.5). The individual probe sensitivity, represented by a mean calibration factor for each probe, was applied to convert measured counts per second (cps) into radioactive concentration (kBq/cc) (PMOD v3.5). The resulting time-activity curves were resampled to μPET frame duration (2 × 10 s, 3 × 20 s, 3 × 30 s, 3 × 60 s, 3 × 150 s, 9 × 300 s) to reduce the increasing statistical counting noise, arising from the rapid decay of ¹¹C. The corrected time-activity curve data points for both probes (striatum and cerebellum) were fitted using the SRTM to obtain the BP_ND and R1 parameters for individual measurements.

Data are expressed as the averaged BP_ND and R1 ± standard error of the mean (s.e.m.) for each experimental group. The statistical significance of these SRTM output values (striatum) for both μPET and beta-microprobe experiments was evaluated using a non-parametric Mann–Whitney test in GraphPad Prism 6 (GraphPad Software, San Diego, USA). Differences were considered statistically significant at p < 0.05.

When comparing two dose conditions (TD and HD) using reference (without implant) μPET scans, the HD, and TD cerebellar time-activity curves both show the characteristic temporal profile of a reference region with a fast decline toward a low level (±0.1%ID/cc) for both radiotracers (Figure 2). Evidently, the corresponding striatal time-activity curves differ in temporal profile: the HD striatal time-activity curves visibly decline to a lower level for both tracers compared to the TD time-activity curves. This “dose effect” is more pronounced for the [¹¹C]RAC striatal time-activity curves (Figure 2A) reflected by the corresponding BP_ND values (Figures 2A,B box-plots): a HD (±65 MBq) administration of [¹¹C]ABP688 caused a 8.94% (±0.33%) decrease in the striatal BP_ND compared with TD (<3.0 nmol/kg) μPET experiments (Figure 2B). A similar HD BP_ND decrease is more pronounced for [¹¹C]RAC (−30.48 ± 2.04%), which reached significance (p = 0.0173; Figure 2A).

High-resolution μCT images (Figures 3A,B) confirm similar scintillating tip dimensions (1.00 mm length–0.50 mm diameter) in both beta-microprobe sets, which results in a similar sensitivity with however a difference in outer diameter (0.75 and 1.00 mm) by additional coating. Cryosections show tissue damage with both beta-microprobe sets (Figures 3C,D) disproportionally increasing with outer diameter. Additionally, our in-house dummy probes for μPET imaging were designed to have a similar shape and size (Figures 3E,F) and proved suitable to mimic the inflicted damage from beta-microprobe implantations, for both the 0.75 mm (Figure 3C) and 1.00 mm (Figure 3D) OD beta-microprobe set.

Figure 4A ([¹¹C]RAC) and Figure 5A ([¹¹C]ABP688) illustrate the corresponding time-activity curves for the cerebellum and the striatum of the reference μPET as well as the TD μPET experiments for both the 0.75 and 1.00 mm OD dummies compared to the intact hemisphere. The striatal time-activity curves of dummy implanted rats differed in temporal profile compared to reference μPET striatal time-activity curves and compared to the intact contralateral side, especially at the level of the peak and significantly more pronounced for 1.00 mm dummy probes.

In both [¹¹C]RAC dummy groups, when comparing the implant (white arrows) and intact striatal VOIs, the BP_ND maps (Figure 4B) and R1 maps (Figure 4C) show significant (p < 0.05) decreases in BP_ND (0.75 mm: −13.01 ± 0.94%; 1.00 mm: −13.89 ± 1.20%) and R1 values (0.75 mm: −29.67 ± 4.94%; 1.00 mm: −39.07 ± 3.17%) at the implant side (Figures 4D,E). Also, the averaged R1 values demonstrated significant (p < 0.05) reductions at the implant side for both dummy groups relative to the averaged reference μPET R1 value (Figure 4E).

For [¹¹C]ABP688, a comparison between both sides (Figures 5B,C) revealed significantly (p < 0.05) decreased BP_ND (−19.09 ± 2.45%) and R1 values (−38.12 ± 6.58%) at the implant side of the 1.00 mm dummy group (Figures 5D,E). Also, the R1 values of the striatal VOIs with implant (Figure 5E) significantly differed (0.75 mm: p < 0.05 and 1.00 mm: p < 0.01) from the averaged reference TD μPET R1 value (0.82 ± 0.03). For [¹¹C]ABP688, the BP_ND seems to be less influenced by the 0.75 mm implants.

Based on the results of the previous section, the 0.75 mm OD beta-microprobe set was selected for beta-microprobe recordings. The averaged and injected dose corrected time-activity curves for HD and TD beta-microprobe acquisitions are illustrated in Figure 6 for both [¹¹C]RAC and [¹¹C]ABP688. The corresponding averaged BP_ND and R1 values are shown in Table 2.

For TD recordings, the tail of the cerebellar reference curve distends (±3000 s) upon decay correction. A subsequent application of the SRTM to TD data resulted in inaccurate outcome parameters, producing a lower BP_ND value (Table 2) compared with HD recordings and conflicting with the reference μPET experiment (Figure 2).

HD recordings, where TD conditions are exceeded, yield lower average BP_ND values (Table 2) compared to reference TD μPET (Figure 2).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. The reviewer YS and handling Editor declared their shared affiliation, and the handling Editor states that the process nevertheless met the standards of a fair and objective review.