TSPO expression was dynamically upregulated during epileptogenesis, persisted in the chronic phase and correlated with microglia activation rather than reactive astrocytes. TSPO expression was correlating with spontaneous seizures and its high expression during the latent phase might possibly suggest being an important switching point in disease ontogenesis which could be further investigated by PET imaging.

Synaptic pathology is associated with several brain disorders, thus positron emission tomography (PET) imaging of synaptic vesicle glycoprotein 2A (SV2A) using the radioligand [11C]UCB-J may provide a tool to measure synaptic alterations. Given the pivotal role of mouse models in understanding neuropsychiatric and neurodegenerative disorders, this study aims to validate and characterize [11C]UCB-J in mice. We performed a blocking study to verify the specificity of the radiotracer to SV2A, examined kinetic models using an image-derived input function (IDIF) for quantification of the radiotracer, and investigated the in vivo metabolism. Regional TACs during baseline showed rapid uptake of [11C]UCB-J into the brain. Pretreatment with levetiracetam confirmed target engagement in a dose-dependent manner. VT (IDIF) values estimated with one- and two-tissue compartmental models (1TCM and 2TCM) were highly comparable (r=0.999, p < 0.0001), with 1TCM performing better than 2TCM for K1 (IDIF). A scan duration of 60 min was sufficient for reliable VT (IDIF) and K1 (IDIF) estimations. In vivo metabolism of [11C]UCB-J was relatively rapid, with a parent fraction of 22.5 ± 4.2% at 15 min p.i. In conclusion, our findings show that [11C]UCB-J selectively binds to SV2A with optimal kinetics in the mouse representing a promising tool to noninvasively quantify synaptic density in comparative or therapeutic studies in neuropsychiatric and neurodegenerative disorder models.

Brain inflammation is an important factor in the conversion of a healthy brain into an epileptic one, a phenomenon known as epileptogenesis, offering a new entry point for prognostic tools. The development of anti-epileptogenic therapies to treat before or at disease onset is hampered by our inability to predict the severity of the disease outcome. In a rat model of temporal lobe epilepsy we aimed to assess whether in vivo non-invasive imaging of brain inflammation at disease onset was predictive of spontaneous recurrent seizures (SRS) frequency and severity of depression-like and sensorimotor-related comorbidities. To this end, translocator protein, a biomarker of inflammation, was imaged by means of positron emission tomography (PET) 2 and 4weeks post-status epilepticus using [F]-PBR111. Translocator protein was highly upregulated 2weeks post-status epilepticus in limbic structures (up to 2.1-fold increase compared to controls in temporal lobe, P<0.001), whereas 4weeks post-status epilepticus, upregulation decreased (up to 1.6-fold increase compared to controls in temporal lobe, P<0.01) and was only apparent in a subset of these regions. Animals were monitored with video-electroencephalography during all stages of disease (acute, latent - first seizures appearing around 2weeks post-status epilepticus - and chronic phases), for a total of 12weeks, in order to determine SRS frequency for each subject (range 0.00-0.83SRS/day). We found that regional PET uptake at 2 and 4weeks post-status epilepticus correlated with the severity of depression-like and sensorimotor-related comorbidities during chronic epilepsy (P<0.05 for each test). Regional PET imaging did not correlate with SRS frequency, however, by applying a multivariate data-driven modeling approach based on translocator protein PET imaging at 2weeks post-status epilepticus, we accurately predicted the frequency of SRS (R=0.92; R=0.86; P<0.0001) at the onset of epilepsy. This study not only demonstrates non-invasive imaging of translocator protein as a prognostic biomarker to ascertain SRS frequency, but also shows its capability to reflect the severity of depression-like and sensorimotor-related comorbidities. Our results are an encouraging step towards the development of anti-epileptogenic treatments by providing early quantitative assessment of SRS frequency and severity of comorbidities with high clinical relevance.

The aim of epilepsy models is to investigate disease ontogenesis and therapeutic interventions in a consistent and prospective manner. The kainic acid-induced status epilepticus (KASE) rat model is a widely used, well-validated model for temporal lobe epilepsy (TLE). As we noted significant variability within the model between labs potentially related to the rat strain used, we aimed to describe two variants of this model with diverging seizure phenotype and neuropathology. In addition, we evaluated two different protocols to induce status epilepticus (SE). Wistar Han (Charles River, France) and Sprague-Dawley (Harlan, The Netherlands) rats were subjected to KASE using the Hellier kainic acid (KA) and a modified injection scheme. Duration of SE and latent phase were characterized by video-electroencephalography (vEEG) in a subgroup of animals, while animals were sacrificed 1 week (subacute phase) and 12 weeks (chronic phase) post-SE. In the 12 weeks post-SE groups, seizures were monitored with vEEG. Neuronal loss (neuronal nuclei), microglial activation (OX-42 and translocator protein), and neurodegeneration (Fluorojade C) were assessed. First, the Hellier protocol caused very high mortality in WH/CR rats compared to SD/H animals. The modified protocol resulted in a similar SE severity for WH/CR and SD/H rats, but effectively improved survival rates. The latent phase was significantly shorter (p < 0.0001) in SD/H (median 8.3 days) animals compared to WH/CR (median 15.4 days). During the chronic phase, SD/H rats had more seizures/day compared to WH/CR animals (p < 0.01). However, neuronal degeneration and cell loss were overall more extensive in WH/CR than in SD/H rats; microglia activation was similar between the two strains 1 week post-SE, but higher in WH/CR rats 12 weeks post-SE. These neuropathological differences may be more related to the distinct neurotoxic effects of KA in the two rat strains than being the outcome of seizure burden itself. The divergences in disease progression and seizure outcome, in addition to the histopathological dissimilarities, further substantiate the existence of strain differences for the KASE rat model of TLE.

Purpose: This study aimed at investigating binding specificity, suitability of reference regionbased kinetic modelling, and pharmacokinetics of the metabotropic glutamate receptor 1 (mGluR1) radioligand [ 11 C]ITDM in mice. Procedures: We performed in vivo blocking as well as displacement of [ 11 C]ITDM during positron emission tomography (PET) imaging using the specific mGluR1 antagonist YM-202074. Additionally, we assessed in vitro blocking of [ 3 H]ITDM at two different doses of YM-202074. As an alternative to reference region models, we validated the use of a noninvasive image-derived input function (IDIF) compared to an arterial input function measured with an invasive arteriovenous (AV) shunt using a population-based curve for radiometabolite correction and characterized the pharmacokinetic modelling of [ 11 C]ITDM in the mouse brain. Finally, we also assessed semi-quantitative approaches. Results: In vivo blocking with YM-202074 resulted in a decreased [ 11 C]ITDM binding, ranging from − 35.8 ± 8.0 % in pons to − 65.8 ± 3.0 % in thalamus. Displacement was also markedly observed in all tested regions. In addition, in vitro [ 3 H]ITDM binding could be blocked in a dosedependent manner. The volume of distribution (V T) based on the noninvasive IDIF (V T (IDIF)) showed excellent agreement with the V T values based on the metabolite-corrected plasma input function regardless of the metabolite correction (r 2 9 0.943, p G 0.0001). Two-tissue compartmental model (2TCM) was found to be the preferred model and showed optimal agreement with Logan plot (r 2 9 0.960, p G 0.0001). A minimum scan duration of 80 min was required for proper parameter estimation. SUV was not reliable (r 2 = 0.379, p = 0.0011), unlike the SUV ratio to the SUV of the input function, which showed to be a valid approach. Conclusions: No suitable reference region could be identified for [ 11 C]ITDM as strongly supported by in vivo and in vitro evidence of specific binding in all brain regions. However, by