15th European Molecular Imaging Meeting
supported by:

To search for a specific ID please enter the hash sign followed by the ID number (e.g. #123).

Disease Models - Translational Approaches in Neuroimaging I

Session chair: Jens Bankstahl (Hannover, Germany); Pedro Ramos-Cabrer (San Sebastian, Spain)
Shortcut: PW05
Date: Wednesday, 26 August, 2020, 5:30 p.m. - 7:00 p.m.
Session type: Poster


Abstract/Video opens by clicking at the talk title.


Histological vs. multiparametric MRI profiling of demyelination and remyelination in the cuprizone murine model of multiple sclerosis

Ander Egimendia1, 2, Daniel Padro1, Lorena Colás1, Iñaki Osorio-Querejeta2, David Otaegui2, 4, Pedro Ramos-Cabrer1, 3

1 CIC biomaGUNE, Magnetic Resonance Imaging Laboratory, Donostia-San Sebastian, Spain
2 ISS Biodonostia, Multiple Sclereosis Group, Donostia-San Sebastian, Spain
3 Ikerbasque, Basque Foudation for Science, Bilbao, Spain
4 Spanish network of Multiple Sclerosis, Barcelona, Spain


The cuprizone model is widely used to study de/remyelination. Cuprizone causes demyelination (peak after 5 weeks) partially resolved by spontaneous remyelination after withdrawal. [1] Many authors have described MRI parameters to quantify myelin but no consensus exist on a gold standard protocol, due to important differences in key aspects like the field, resolution, references, timing or brain regions analyzed. We present a high spatial and temporal resolution MRI protocol at 11.7T (vs. Luxol Fast Blue histology) to contribute to the discussion of a suitable imaging protocol for this model.


C57BL/6 male mice were fed with standard vs. 0.2% (w/w) cuprizone diet. Groups: Cuprizone treated animals from W0-W5 (standard diet from W5 to month 6) and sacrifice at W5 (n=6), W10 (n=6) and M6 (n=3). Controls: standard diet from W0 to month 6 and sacrifice at W5 (n=6), W10 (n=6) and M6 (n=3). Luxol fast blue and cresyl violet histology (25 µm slices) was performed after sacrifice, images were analyzed with image-J. High resolution (75 µm) MRI was performed weekly (W1-W10) and M6 at 11.7 T including: SWI, MTR, T1w, T2w and DTI (FA, MD, RD, AD). FSL was used for image analysis (brain segmentation, co-registration, ROI selection using AMBMC atlas [1] and group averaging). Diffusion images denoised using local PCA [2]. Mean signal on T1w, T2w and SWI was normalized (whole brain = 10000).


MRI parameters sensitive to demyelination (Week 5) were (ordered by magnitude of differences between cuprizone vs. controls): T2w > SWI > MTR > FA > T1w > RD. Remyelination (Week 10) was detected by T2w > SWI > FA > RD. Chronic effects (M6) were only detected by T2w > RD > T1w. Luxol fast blue histology showed (Fig1) that T2w(r2=0.84), T1w (r2=0.72) and RD (r2=0.76), were also the only ones showing good correlation with myelin content. At 11.7T the T1/T2 ratio (clinically used for myelin imaging) did show worse performance that T2w alone, probably due to very short T2 times at high fields. Longitudinal follow-up (Fig 2) revealed that medial corpus callosum (peak at W5, renormalized at W10), genu of CC (peak at W2 slowly renormalized from W4) and Caudate putamen (peak at W8, slight drop at W10) present different evolution profiles, indicating that myelin loss affects differently to these brain regions, with important implications for the study of de- and remyelination.


1) At very high field, most sensitive and specific detection of myelin is achieved combining T2w, RD and T1w, in the cuprizone model.
2) Up to 3 different temporal patterns of demyelination-remyelination were found in different regions of the brain (multi-ROI and high resolution is a must in this model).
3) Demyelination leaves long-lasting effects visible up to 6 months after the demyelinating insult.


This work has been performed under the Maria de Maeztu Units of Excellence Program from the Spanish State Research Agency (Grant No. MDM-2017-0720), and Grant SAF2017-87670-R (Spanish State Research Agency ). PRC is funded by Ikerbasque.

[1] Lindner, M. et al. Neuropathol Appl Neurobiol 2008;34(1):105-14
[2] Australian Mouse Brain Mapping Consortium: http://www.imaging.org.au/ AMBMC/)
[3] J. V. Manjón et al. PLoS One 2013; 8(9): e73021
Correlation between Myelin histology and T2w MRI signal
Normalized T2w signal intensity was the MR parameter showing the highest specificity for Myelin imaging, as compared with Luxol Fast Blue staining (correlation index r2=0.84). Both parameters suscesfully depicted demyelination and remyelination, specially at the level of the corpus callosum. 
Temporal evolution of MRI parameter in different regions of the brain
Longitudinal follow-up of imaging parameters at high spatial resolution showed up to 3 different patterns of evolution during de- and remyelinating processes in the brain of the cuprizone mouse. This result may explain differences observed in this model for different authors.
Keywords: Curprizone, MRI, Multiple Sclerosis, demyelination, myelin

Alteration in SV2A density in transgenic rat model of AD TgF344-AD and with age measured by [18F]UCB-H in vivo PET imaging

Daniela Bochicchio1, James Minshull1, Federico Roncaroli1, Rainer Hinz1, Herve Boutin1

1 University of Manchester, Manchester, United Kingdom


Synapses are the critical interface between neurons allowing essential transmission of information from a neuron to another. Alterations in synapses are associated with a variety of neurodegenerative disorders including Alzheimer’s disease. 18F-UCB-H is a tracer able to bind to synaptic vesicles 2A (SV2A), a synaptic vesicle glycoprotein involved in vesicle trafficking and exocytosis, and SV2A has been shown to be reduced in postmortem brain of dementia patients1. We here investigate synaptic density with 18F-UCB-H in TgF344-AD (TG) rats and WT littermates at 7 and 15 months of age.


All procedures were approved by the Home Office and carried out in accordance with the Animals (Scientific Procedures) Act 1986. Under isoflurane anaesthesia WT and TG rats at 7 (WT, n=8; TG, n=7) and 15 (WT, n=10; TG, n=10) months of age were injected i.v. with 18F-UCB-H at the start of the PET scan (baseline scans). To measure the non-specific binding, some of these WT (n=4) and TG (n=4) rats were injected with cold UCB-J (1mg/kg) 10min before 18F-UCB-H injection (block scans). Data were then expressed as normalised uptake values (NUVND) of the ratio between the 9-22min averaged SUV of the baseline SUV over the blocked (non-displaceable) SUV for each ROI. PET images were quantified in BrainVisa and Anatomist software using 28 brain ROIs based on a MRI rat brain atlas2.


Pre-saturation with UCB-J blocked up to 70% of the specific binding and removed regional differences (Fig.1). Brain distribution in WT (Fisher 344) rats was similar to those previously described in Sprague-Dawley rats. Correlation between NUVND calculated with individual values from the blocking study or group averages returned excellent correlation (Fig.1C), we therefore used the population-based SUVUCB-J to calculate the NUVND for all rats.
TG rats had significantly lower NUVND than WT at 7m of age (Fig.2) in frontal cortex (-17%, p=0.002), motor cortex (-16%, p=0.006), hippocampus (-12%, p=0.020), thalamus (-13%, p=0.020), cerebellum (-17%, p=0.001), caudate putamen (-15%, =0.009) and globus pallidus (-13%, p=0.017) but not in temporal cortex (p=0.11). There was an effect of age in both WT and TG so that the 15m old were significantly lower than 7m in both genotypes and TG were no longer significantly different from WT at 15m (Fig.2).


The 18F-UCB-H brain distribution showed here in Fisher344 rats is in good agreement with previous reports3. We here demonstrate an early alteration in 18F-UCB-H uptake in TG rats vs WT at 7m, while normal ageing affects both genotype later on. These results are in line with alterations in functional connectivity4 and BBB5 in TG observed at 10 and 12m respectively. Confirmation of these results by immunohistochemistry is ongoing.

[1] Robinson J. L. et al. (2014). Brain.137:2578-2587.
[2] Schwarz A. J. et al. (2006). Neuroimage.32:538-550.
[3] Serrano M. E. et al. (2019). Molecules.24.
[4] Anckaerts C. et al. (2019). Neurobiol Dis.124:93-107.
[5] Dickie B. R. et al. (2019). Neuroimage.184:349-358.
Bochicchio et al. Fig 1

Fig.2: UCB-H normalised uptake values (NUVND) analysis in different brain region at 7 and 15 months of age. Data expressed as mean±SD. Data analysed with 2 way ANOVA and Sidak post-hoc test. * and ** indicates significant differences between TG and WT p<0.05 and p<0.01 respectively.

Bochicchio et al fig2

Fig.2: UCB-H normalised uptake values (NUVND) analysis in different brain region at 7 and 15 months of age. Data expressed as mean±SD. Data analysed with 2 way ANOVA and Sidak post-hoc test. * and ** indicates significant differences between TG and WT p<0.05 and p<0.01 respectively.

Keywords: synaptic density, density, UCB-H, PET, Alzheimer disease model

Longitudinal measure of protein synthesis rate in the transgenic rat model of Alzheimer’s disease TgF344-AD by [11C]leucine PET imaging

Daniela Bochicchio1, Rainer Hinz1, Christine Parker2, Herve Boutin1

1 University of Manchester, Faculty of Biology, Medicine and Health, Manchester, United Kingdom
2 GlaxoSmithKline (GSK), Stevenage, United Kingdom


In brain, protein synthesis (PS) is essential for long term memory1 formation. Memory alterations are one of the main hallmarks of Alzheimer’s disease2,3. Changes in PS pathways have been shown in AD animal models and patients4 ex vivo, however such changes have not been assessed in vivo in AD. [11C]leucine PET is the method of choice to measure cerebral protein synthesis rate (PSR). [11C]leucine has a high uptake across the blood brain barrier, a simple catabolism and almost exclusively incorporate in protein. We here aimed at investigating PSR in the TgF344-AD rats.


All procedures were approved by the Home Office and carried out in accordance with the Animals (Scientific Procedures) Act 1986. Under isoflurane anaesthesia, Wistar, TgF344-AD (TG) and WT rats were injected i.v. with [11C]leucine at the start of a 60min PET scan. TG and WT were scanned at 6, 12 and 18 months of age. Arterial blood activity was monitored online (Twilite®) and discrete blood samples were collected in some rats. Whole-blood, plasma samples and free [11C]leucine in plasma were measured by γ-counting. To measure the PSR inhibition, Wistar rats were injected with anisomycin (60mg/kg) 10min prior PET acquisition. PET images were quantified in BrainVisa/Anatomist software using 28 brain ROIs based on a MRI rat brain atlas and modelled with a two tissues compartimental1,5 model.


The concentration of unlabelled leucine in plasma was statistically different between arterial and venous samples and between strains (Wistar vs Fischer-344) but no significant difference was found between WT and TG. Anisomycin administration significantly reduced the net uptake rate constant (Kcplx) (-86%, p=0.003; Fig.1) proving the sensitivity of the method. For the longitudinal study, since permanent arterial cannulation for individual arterial blood sampling was not doable, we validated and used an averaged population-based whole-blood and plasma input function (IF) to calculate Kcplx together with an averaged population-based arterial cold leucine concentration to calculate PSR. Kcplx, but not PSR, was decreased in TG (genotype effect by 2 way ANOVA) in hippocampus (p=0.044; Fig.2), whole Brain (p=0.048) and caudate-putamen (p=0.042). In the globus pallidus Kcplx and PSR were significantly reduced (-13%, p=0.042 and p=0.048 respectively; t-Test; Fig.2) in TG vs WT at 18 months.


[11C]leucine PET is sensitive enough to measure brain PSR in rat, able to detect PSR inhibition upon administration of a PS inhibitor. However, modelling is challenging. Population-based IF returned only modest changes in Kcplx & PSR in the TgF344-AD vs WT. Acquiring individual blood values in a cross-sectional study may have showed greater differences but would have compromised the longitudinal aspect of the study and tripled the number of rats.


DB and this study are funded by a GSK‐University of Manchester PhD studentship. MV was funded by the EPSRC project EP/M005909/1. The authors wish to thank all the personnel of the WMIC, especially Ms. Lidan Christie, Ms. Carol Brough, Mr Michael Green and Mr Hamza Al‐qasmi for facilitating the study.

[1] Bishu S. et al. (2008). J Cereb Blood Flow Metab.28:1502-1513.
[2] Ma T. et al. (2013). Nat Neurosci.16:1299-1305.
[3] Garcia-Esparcia P. et al. (2017). Am J Neurodegener Dis.6:15-25
[4] Hoozemans J. J. et al. (2005). Acta Neuropathol.110:165-172.
[5] Sundaram S. K. et al. (2006). J Nucl Med.47:1787-1795
Fig1. Bochicchio et al.

Fig.1: Kcplx (net uptake rate constant) in Wistar rats without (baseline) and with injection of anisomycin (i.v., 10 mins before [11C]leucine @60mg/kg) in different brain regions. Data expressed as mean±SD. Data analysed unpaired t-tests. ** indicates significant differences (p<0.01).

Fig2. Bochicchio et al.

Fig.2: (A) Kcplx (net uptake rate constant) and protein synthesis rate (PSR) in the hippocampus of WT and TgF344-AD rats at 6, 12 and 18 months of age. There was an overall effect of genotype for Kcplx (p=0.044, 2-way ANOVA) but no effect for PSR. (B) Kcplx and PSR in globus pallidus in WT and TG at 18m of age; there was a significant difference between WT and TG Kcplx and PSR (Kcplx , p=0.042 and PSR, p=0.048; data were analysed with unpaired t-tests). Data are expressed as mean±SD.

Keywords: leucine, protein synthesis rate, alzheimer model

Dynamic functional connectivity and graph theory metrics in a rat model of temporal lobe epilepsy

Emma Christiaen1, Marie-Gabrielle Goossens2, Benedicte Descamps1, Robrecht Raedt2, Christian Vanhove1

1 Ghent University, MEDISIP, Department of Electronics and Information Systems, Ghent, Belgium
2 Ghent University, 4Brain Team, Department of Head and Skin, Ghent, Belgium


Epilepsy is a neurological disorder characterized by recurrent epileptic seizures. The involvement of abnormal functional brain networks in the development of epilepsy and its comorbidities has been demonstrated by electrophysiological and neuroimaging studies in epilepsy patients1. In this longitudinal resting state functional MRI (rsfMRI) study, changes in dynamic functional connectivity (dFC) and network topology during epileptogenesis were investigated using the intraperitoneal kainic acid (IPKA) rat model of temporal lobe epilepsy (TLE).


24 adult male Sprague-Dawley rats were used. 17 were i.p. injected with kainic acid according to the protocol of Hellier et al. (1998)2, resulting in status epilepticus (SE), and 7 with saline (controls). Before and at 5 time points post-SE, rsfMRI scans were acquired on a 7T system. Correlation between fMRI timeseries was calculated within a sliding window of 50s with a 2s step length. The resulting correlation matrices were classified into 6 states using k-means clustering. For all time-varying correlation matrices, graph theoretical metrics were calculated and classified into 6 states of network topology. Percentage dwell time in and number of transitions between each state were calculated and their correlation with seizure frequency, based on hippocampal EEG recordings, was assessed.


The 6 states of FC were sorted from highest to lowest mean value. Percentage dwell time in State 1, 2 and 3 was significantly lower in the IPKA group compared to controls, while dwell time in State 5 and 6 was significantly higher. A significant effect of time post-SE could be found in the IPKA group, where a significant decrease in dwell time in State 1, 2, 3 and 4 and increase in State 5 and 6 could be observed during epileptogenesis (Fig. 1A). The number of transitions was significantly lower in the IPKA group compared to controls and decreased significantly during epileptogenesis in the IPKA group (Fig. 1B). Seizure frequency was positively correlated with dwell time in State 2 one week post-SE and in State 4 16 weeks post-SE, and negatively with dwell time in State 5 one week post-SE and State 6 16 weeks post-SE (Fig. 2). The number of transitions 16 weeks post-SE was positively correlated with seizure frequency. Similar results were obtained for the states of network topology.


States with a lower mean FC occurred more often in IPKA animals compared to controls. FC became less variable during epileptogenesis, which might be related to cognitive problems. Seizure frequency was positively correlated with dwell time in states with high FC and number of transitions between states, indicating that dwelling in states of higher FC and more switching between states seem to increase the probability that seizures are generated.


This research is funded by a Doctoral Grant Strategic Basic Research of Research Foundation – Flanders.

[1] Chiang, S. & Haneef, Z. 2014, ‘Graph theory findings in the pathophysiology of temporal lobe epilepsy’, Clin. Neurophysiol., 125, 1295–1305
[2] Hellier, J. L., Patrylo, P. R., Buckmaster, P. S. & Dudek, F. E. 1998, ‘Recurrent spontaneous motor seizures after repeated low-dose systemic treatment with kainate: assessment of a rat model of temporal lobe epilepsy’, Epilepsy Res., 31, 73–84
Fig. 1
A) % dwell time in states of FC as a function of time. Data are visualized as a stacked bar graph with mean values. Each segment of a bar represents % dwell time in a state averaged over the animals in the group at one time point. The 6 segments always add up to 100%. B) Number of transitions between states of FC as a function of time. Data are visualized as a stacked bar graph with mean values. Each segment of a bar represents number of transitions between 2 states averaged over the animals in the group at one time point.
Fig. 2

Correlation between average daily seizure frequency (number per 24 hours) on the one hand and percentage dwell time in and total number of transitions per scan between states of functional connectivity 1 week and 16 weeks post-SE on the other hand.

Keywords: Resting state functional MRI, temporal lobe epilepsy, intraperitoneal kainic acid rat model, dynamic functional connectivity, sliding window analysis

Enhanced transhemispheric connectivity after cortical stroke in mice

Niklas Pallast1, Frederique Wieters1, Mathias Hoehn2, Gereon R. Fink1, 2, Markus Aswendt1

1 University of Cologne, Department of Neurology, Cologne, Germany
2 Research Center Juelich, Institute of Neuroscience and Medicine (INM-3), Juelich, Germany


Stroke is a devastating disease disrupting connections between neurons by direct tissue loss and indirect disconnection of remote areas leading to sever long-term disability. The stroke lesion and the white matter tracts can be visualized and quantified using Diffusion MRI and Diffusion Tensor Imaging (DTI), respectively. Previous studies could show that structural connectivity such as the integrity of the corticospinal tract, predict motor impairment in chronic stroke patients (1). Pre-clinical studies in mice are necessary in order to understand the underlying processes of reorganization.


We applied high-resolution and longitudinal T2-weighted MRI and DTI in adult mice which received cortical stroke lesions by photothrombosis or sham protocol and repetitive behavioral testing (Rotating Beam, Cylinder and Grid Walk) up to 4 weeks post-surgery. The T2 stroke lesion and DTI data was processed using an in-house developed pipeline (2) and registered with a modified version of the Allen Mouse Brain Atlas (3). DSI studio was used to extract the whole-brain structural network and subsequently derive the connectivity for individual pairs of brain regions. For the DTI network approach, graph theory was used to derive the degree and edge weights for nodes related to the regions of the sensorimotor network. Immunostaining for GFAP was used to quantify the lesion on cryosections (4).


All stroke mice showed significant sensorimotor deficit during the first week after stroke and spontaneous functional improvement, which depending on the behavior test, reached pre-stroke levels. In accordance with previous studies (5), we detected increased glial activity in the ipsilesional thalamus, which in our study was related to the lesion size. In addition, structural connectivity between the sensory-motor cortex related part of the thalamus (DORsm) and the sensorimotor cortex was significantly reduced with the strongest effects in the DORsm to primary somatosensory (SSp) upper and lower limb regions, which corresponds to the stronger behavioral deficit in larger strokes. In line with the spontaneous improvement, we detected increased structural connectivity from the contralesional DORsm and primary motor cortex (MOp) to the lesioned cortex regions. The increase in these connections was stronger in the larger strokes.


Here, we provide the first longitudinal and quantitative mapping of fiber tract changes related to cortical stroke combined with correlations to repetitive behavioral testing. This study highlights the importance of the contralesional hemisphere for stroke recovery and presents a first step for the development of connectivity measures as a novel biomarker.


This work was financially supported by the Friebe Foundation (T0498/28960/16). The authors acknowledge Marieke Nill, Olivia Käsgen, Joshua Strelow, Veronika Fritz and Alina Jakobs for technical assistance with MRI, animal handling and histology.

[1] R. Lindenberg, et al., Structural integrity of corticospinal motor fibers predicts motor impairment in chronic stroke. Neurology74, 280–7 (2010).
[2] N. Pallast, et al., Processing Pipeline for Atlas-Based Imaging Data Analysis of Structural and Functional Mouse Brain MRI (AIDAmri). Front Neuroinform 13, 42 (2019). 
[3] E. S. Lein, et al., Genome-wide atlas of gene expression in the adult mouse brain. Nature 445, 168–176 (2006). 
[4] N. Pallast, F. Wieters, G. R. Fink, M. Aswendt, Atlas-based imaging data analysis tool for quantitative mouse brain histology (AIDAhisto). J Neurosci Meth 326, 108394 (2019). 
[5] Schroeter, S. Jander, O. Witte, G. Stoll, Heterogeneity of the microglial response in photochemically induced focal ischemia of the rat cerebral cortex. Neuroscience 89, 1367–1377 (1999). 
Keywords: stroke, neuroimaging, DTI, recovery

Magnetic resonance imaging reveals the temporal dynamic of ischemic brain lesion and tissue loss after transient MCAO in Tlr2-deficient mice

Srecko Gajovic1, Marina Dobrivojević Radmilović1, Siniša Škokić1, Anton Glasnović1, Dunja Gorup1, Paula Josic1, Helena Justić1, Anja Barić1

1 University of Zagreb School of Medicine, Croatian Institute for Brain Research, Zagreb, Croatia


The effects of modified neuroinflammation on the outcomes of mouse brain ischemic lesion, being a model for ischemic stroke in humans, are still controversial. The two faces of inflammation, detrimental and beneficial, are reflected differently on the post-ischemic damage and recovery and vary in relation to the time after ischemia, having even opposite consequences in chronic compared to the acute phase (1,2). Tlr2-deficiency is an example where the differences in ischemic lesion size were shown between the acute and chronic phase when compared to the wild type (WT) animals (3).


In order to clarify temporal dynamic of ischemic brain lesion, in this study we used the advantage of magnetic resonance imaging (MRI, Bruker 7T Biospec 70/20 USR) to follow up the animals up to 28 days after transient (60 min) middle cerebral artery occlusion (tMCAO). Tlr2-deficient animals were compared to WT controls. MRI analysis was based on T2 scans. At the end of monitoring (28 days after MCAO) the animals were sacrificed, brain isolated, sectioned and stained by Nissl. The animals were functionally assessed by neurological scoring.


For MRI analysis we established a standardized manual segmentation pipeline referred as “Zagreb approach”. Hyperintense areas were manually delineated for each time point from the T2 maps during the first 7 days following MCAO. The volume of the lost tissue at day 14 and 28 after MCAO was calculated by subtracting the ipsilateral hemisphere volume at the time point of interest from the ipsilateral hemisphere volume at baseline.

The analysis confirmed the dynamic differences between Tlr2-deficient and WT mice. The survival of Tlr2-deficient animals was better than of WT mice. Contrary to this, Tlr2-deficient animals performed functionally worse having higher neurological score in the acute phase. However, the neurological scores were comparable to WT animals at the end of the experiment. MRI showed higher ischemic lesion and bigger tissue loss in Tlr2-deficient animals and the differences were more pronounced in the chronic phase, although at the borderline of statistical significance.


The modified inflammation in Tlr2-deficient mice resulted in worse functional outcomes in the acute phase and trends toward bigger tissue loss in the chronic phase after ischemia. However, Tlr2-deficient mice survived better the acute phase, and had comparable functional outcome at the end of the experiment. The evaluation of modified inflammation after ischemic lesion demonstrates a confusing combination of beneficial and detrimental effects.


This study was supported by EU European Regional Development Fund, Operational Programme Competitiveness and Cohesion, grant agreement No.KK., CoRE – Neuro, and by the Croatian Science Foundation under the project IP-06-2016-1892 (RepairStroke). The work of doctoral students Anja Barić and Paula Josić has been fully supported by the “Young researchers' career development project – training of doctoral students” of the Croatian Science Foundation funded by the European Union from the European Social Fund. Multimodal imaging was done at Laboratory for Regenerative Neuroscience - GlowLab, University of Zagreb School of Medicine.

[1] Gorup, D, Škokić, S, Kriz, J, Gajović, S 2019, ‘Tlr2-deficiency is associated with enhanced elements of neuronal repair and caspase 3 activation following brain ischemia’, Sci Reports, 9, 2821
[2] Gorup, D, Bohaček, I, Miličević, T, Pochet, R, Mitrečić, D, Križ, J, Gajović S 2015, ‘Increased expression and colocalization of GAP43 and CASP3 after brain ischemic lesion in mouse’, Neurosci Lett, 597, 176
[3] Bohacek, I, Cordeau, P, Lalancette-Hébert, M, Gorup, D, Weng, YC, Gajović, S, Kriz, J 2012, ‘Toll-like receptor 2 deficiency leads to delayed exacerbation of ischemic injury’, J Neuroinflammation, 9, 191
Keywords: middle cerebral artery occlusion, ischemic stroke, inflammation, brain, toll-like receptors

Altered patterns of neural activity and functional connectivity revealed by dynamic rsfMRI in the Q175 mouse model of Huntington's disease

Tamara Vasilkovska1, Bram Callewaert1, Somaie Salajeghe1, Dorian Pustina2, Longbin Liu2, Mette Skinbjerg2, Celia Dominguez2, Ignacio Munoz-Sanjuan2, Annemie Van der Linden1, Marleen Verhoye1

1 University of Antwerp, Bio-Imaging Lab, Antwerp, Belgium
2 CHDI Foundation, Princeton, NJ, United States of America


Resting state fMRI (rsfMRI) is a method that uses Low Frequency (LF) fluctuations of Blood Oxygenated Level Dependent (BOLD) signal within the brain. The correlation of these LF BOLD changes between neuro-anatomically defined regions represents static FC. However, FC does not provide temporal information of the BOLD changes. A more sophisticated approach, the Quasi Periodic Patterns (QPPs), represents spatiotemporal patterns of recurring brain dynamics correlated to neural activity1. In this study, we investigated both FC and QPPs in the Q175 Heterozygous (HET) mouse model2.


rsfMRI data was acquired in 24 3-month and 35 12-month old Q175 HET and wild type (WT) male mice (two groups at each age, N HET/N WT (age): 12/12 (3M) and 19/16 (12M)) using a 9.4T Biospec MRI scanner. Mice were anesthetized with a mixture of medetomidine (0.075mg/kg s.c. bolus; 0.15 mg/kg/h s.c. infusion) and 0.5% isoflurane. rsfMRI was acquired using a T2*-weighted single shot EPI sequence (TR/TE 500/15ms, matrix dimensions [90x70], 12 horizontal slices, pixel dimensions of (300x300x400) µm3, 1200 repetitions). Static FC and all QPP-related analysis were performed using MATLAB 2018b. A two-sample t-test was performed to evaluate potential FC differences between the groups. The spatiotemporal dynamics of BOLD patterns were extracted with a specific pattern finding algorithm3.


Fig. 1 shows the FC differences between the two groups (HET and WT) at the age of 3 (Fig.1a) and 12 months (Fig.1b). At 12 months, the FC profiles indicate significantly lower connectivity between: CPu(L)-CPu(R), CPu(R)-Piri(R), Ctx_Cl(R)-M1(R) and Ctx_Cl(R)-S1(R) in HET compared to WT mice, while at 3 months there was no significant FC difference between the two groups. Fig.2a shows a representative spatiotemporal QPP segment for both HET and WT at 3 months, indicating activity patterns which differ between the groups. Fig.2b shows representative QPP segments at a given time point for 12 months for both HET and WT. We observed qualitatively different patterns as well as lateralization in the HET compared to WT group. Our results suggest significant decrease in FC between HD-related regions at 12 months, as well as differential QPP starting at 3 months and with a more distinctive QPP pattern alteration between groups at 12 months.


To the best of our knowledge, this is the first study in which FC and QPP in an HD rodent model are assessed. Both measures suggest connectivity/activity changes in HET in both the striatum and cortical regions and additionally, in areas linked to the olfactory processing system4. Further histological assessment is needed to link these changes in brain connectivity and spatiotemporal patterns with mutant huntingtin (mHTT) deposition.

[1] Thompson, G.J., Pan, W.-J., Magnuson, M.J. et al., 2014, 'Quasi-periodic patterns (QPP): Large-scale dynamics in resting state fMRI that correlate with local infraslow electrical activity' Neuroimage; 84, 1018-1031
[2] Menalled L., Kudwa AE, Miller S, et al., 2012, 'Comprehensive behavioral and molecular characterization of a new knock-in mouse model of Huntington's disease: zQ175', PLoS One, 7(12): e49838.
[3] Majeed, W., Magnuson, M., Hasenkamp, W., et al., 2010, 'Spatiotemporal dynamics of low frequency BOLD fluctuations in rats and humans', Neuroimage, 54(2): 1140-1150
[4] Ferris, C.F., Kulkarni, P., Steven, T., et al., 2014, 'Studies on the Q175 Knock-in Model of Huntington’s Disease Using Functional Imaging in Awake Mice: Evidence of Olfactory Dysfunction', Frontiers in Neurology, 5: 94

FC changes between WT and HET in the Q175 mouse model at 3 and 12 months of age

Fig.1: ROI-based approach reveals differences in FC between distinct regions in HET compared to WT at 3 (a) and 12 months (b). Top right: color represents mean FC difference of WT-HET; Bottom left: red squares - brain region pairs presenting a significant genotype difference of FC (two-sample t-test, p-value <0.05, FDR corrected, also marked with (*) in top panel). Abbreviations: CPu – Caudate Putamen; Ctx_Cg–Cingulate cortex; Ctx_Cl – Claustrum; M1 – Motor cortex 1; OB – Olfactory Bulb; Piri – Piriform Cortex; RF – Reticular Formation; S1 - Somatosensory cortex 1

Representative WT and HET-QPP segments in Q175 mice at 3 and 12 months of age

Fig.2: Representative QPP segments at different brain levels for WT and HET at 3 (a) and 12 months (b). Mean BOLD color maps show BOLD activity over all QPP occurrences. ­ (red) / ¯ (blue) higher/lower BOLD activity relative to the global mean BOLD signal. The major ROIs are marked in the adjacent mouse brain atlas overlays - identical regions (black labels), several regions only present in WT (green) or HET QPP (pink); Corresponding two-tailed one sample T-test maps of the QPP patterns (FDR corrected, p-value <0.05) for WT and HET mice. Accompanying scale of the T-maps for both groups.

Keywords: rsfMRI, quasi-periodic patterns, Huntington's disease, Q175 mouse model, functional networks

In vivo assessment of glymphatic system efficacy and dynamics of brain-fluids-circulation during the late-onset of Alzheimer’s disease in Tet-Off-APP mice and TgF344-AD rat model

Aneta J. Keliris1, Ines R. H. Ben-Nejma1, Monica van den Berg1, Verdi Vanreusel1, Jasmijn Daans2, Peter Ponsaerts2, Marleen Verhoye1, Georgios A. Keliris1, Annemie Van der Linden1

1 University of Antwerp, Bio-imaging Lab, Antwerp, Belgium
2 University of Antwerp, Laboratory of Experimental Hematology, Antwerp, Belgium


Glymphatic system (GS) is a recently identified brain-wide network of perivascular channels promoting fluid exchange between CSF and brain interstitial space that facilitates removal of solutes, including aberrant Amyloid-β (Aβ) and Tau proteins present in Alzheimer’s disease (AD) [1]. While failure of glymphatic drainage is postulated to play a key role in the pathology of AD, the exact mechanisms behind are still largely unknown. Here, we investigate the efficacy of glymphatic flow in two different rodent models of AD and upon ageing by means of dynamic contrast enhanced MRI (DCE-MRI).


DCE-MRI was performed to assess the dynamics of brain-fluids-circulation and GS efficacy in two transgenic (TG) rodent models of AD at advance stage of disease. To this end, we used: a) a mouse model with inducible amyloid expression, Tet-Off APP, and b) a translational rat model (TgF344-AD) that exhibits both Amyloid and Tau pathology. DCE-MRI was acquired on 9.4 T continuously by repeated acquisition of 3D T1-weighted MR images (3D FLASH, TA=4min), following the labelling of cerebrospinal fluid (CSF) with Dotarem (Gd-DOTA), in TG mice (N=6, 14m old), TgF344-AD rats (N=6, 22m old) and control littermates (CTL, N=6), until signal intensity returned back to baseline level (~2.5h/mouse; ~4h/rat, anesthesia: isoflurane). Ex vivo analysis was performed to assess microscale aberrations.


We have observed altered dynamics of brain-fluids-circulation in both rodent models of AD. Specifically, the TG mice showed a decrease in parenchymal penetration of Gd-DOTA compared to CTL, as shown on the Figure 1, indicative of the decrease in GS efficacy. Similar global impairment of glymphatic flow was observed in TG rats. The differences between the TG and the CTL animals were moderate, which could be attributed to the old age of the investigated animals since the efficacy of GS was shown to be reduced also upon ageing [2]. Further, our ex vivo results showed extensive astrogliosis along with wide-spread amyloid plaques burden present in the cortex and hippocampus of TG mice, and hippocampus, cortex, cerebellum, striatum and olfactory bulb of TG rats. These microscale aberrations might be responsible for the reduction in glymphatic flow observed in TG animals, and thus aggravation of AD pathology.


We have observed moderate changes in the efficacy of glymphatic system in both investigated rodent models of AD at an advanced stage of disease, and these changes coincided with the presence of pathological hallmarks of AD. Future research will focus on capturing the early alteration in the dynamics of brain-fluids-circulation and unraveling their role in initiation and progression of AD.

AcknowledgmentThis research was supported by the Fund for Scientific Research Flanders (FWO) (grant agreement G067515N and G048917N). We would like to thank Johan Van Auderkerke for the excellent technical support. The computational resources and services used in this work were provided by the HPC core facility CalcUA of the Universiteit Antwerpen, the VSC (Flemish Supercomputer Center), funded by the Hercules Foundation and the Flemish Government – department EWI
[1] Iliff JJ, Wang M, Liao Y, Plogg BA, Peng W, Gundersen GA, Benveniste H, Vates GE, Deane R, Goldman SA, Nagelhus EA, Nedergaard M, ‘A parvascular pathway facilitates CSF flow through the brain parenchyma and the clearance of interstitial solutes, including amyloid b’, Sci Transl Med, 2012 4(147), 147ra111
[2] Kress BT, Iliff JJ, Xia M, Wang M, Wei HS, Zeppenfeld D, Xie L, Kang H, Xu Q, Liew JA, Plog BA, Ding F, Deane R, Nedergaard M, ‘Impairment of paravascular clearance pathways in the aging brain’, Ann Neurol, 2014 76(6), 845.
Figure 1. Spatial distribution of Gd-DOTA in the Tet-Off-APP and control mice (CTL) over time.

 After the injection of Gd-DOTA in the cisterna magna, the spatial distribution of the MRI tracer was followed-up in the CTL (top row) and TG mice (bottom row) by DCE-MRI up to 150 min post injection, presented here at two different bregma level: midline (A) and more laterally at 1.92 mm (B). The colour scale represents the average percentage intensity change from the last scan (reference image), with dark red/white indicating low/high percentage intensity change.


Keywords: glymphatic system, DCE-MRI, Alzheimer's disease

Characterization of the rotenone mouse model of Parkinon's disease at early and late disease's stage using radioligands for the α4Β2 nicotinic acetylcholine receptor ((-)-[18F]Flubatine) and the adenosine A2A receptor ([18F]FESCH and [18F]FLUDA)

Magali Toussaint1, Mathias Kranz1, Daniel Gündel1, Thu Hang Lai1, 2, Susann Schröder2, Sladjana Dukic-Stefanovic1, Winnie Deuther-Conrad1, Rodrigo Teodoro1, Qi Shang3, 4, Marianne Patt5, Heinz Reichmann6, Richard Funk7, Osama Sabri5, Francisco Pan-Montojo3, Peter Brust1

1 Helmholtz-Zentrum Dresden-Rossendorf, Institute of Radiopharmaceutical Cancer Research, Department of Neuroradiopharmaceuticals, Research site Leipzig, Leipzig, Germany
2 ROTOP Pharmaka GmbH, Department of Research and Development, Dresden, Germany
3 Ludwig-Maximilians-Universität Munich, University Hospital Großhadern, Neurological Clinic & Policlinic, Department of Neurology, Munich, Germany
4 Technische Universität Dresden, University Hospital Carl Gustav Carus, Clinic of Neurology, Dresden, Germany
5 University Hospital Leipzig, Department of Nuclear Medicine, Leipzig, Germany
6 Technische Universität Dresden, University Hospital Carl Gustav Carus, Department of Neurology, Dresden, Germany
7 Technische Universität Dresden, University Hospital Carl Gustav Carus, Institute of Anatomy, Dresden, Germany


Systemic administration of rotenone is able to reproduce the main pathological and behavioral hallmarks of Parkinson’s Disease (PD) in mice. Therefore, those mice are potentially useful for the development of therapies targeting the nicotinic acetylcholine receptor (α4β2nAChR) or the adenosine A2A receptor (A2AR).
Thus, we evaluated the ability of the rotenone model to resemble the decreased availability of α4β2nAChR and the increased availability of A2AR found in the brain of PD patients [1,2,3]. PET/MR imaging was performed to quantify these changes at early and later stages of the disease.


Two groups of 12-14-months-old male C57BL/6JRj mice (27-36 g) treated for 2 months (n=6) or 4 months (n=7) with rotenone, 5 days/week, 5 mg/kg p.o., and their corresponding control groups (n=7 and n=5, respectively) were investigated.
(-)-[18F]Flubatine (6.4±1.9 MBq; Am: 1185±713GBq/µmol) for α4β2nAChR investigation and [18F]FESCH (5.0±1.8 MBq; Am: 116±19 GBq/µmol, EOS) [4] or [18F]FLUDA (5.7±1.2 MBq; Am: 96±10 GBq/µmol, EOS) for A2AR investigation were injected intravenously followed by 60 min dynamic PET scans[4].
The cerebellum was used as a reference tissue. The time-activity curves (TACs) of the SUV ratio (SUVR) of thalamus or striatum over cerebellum were used as measure for specific uptake.


Specific uptake of (-)-[18F]Flubatine was observed in the thalamus of control and rotenone mice (SUVR60min p.i. ~3.5, all groups included). However, for none of the two treatment groups changes in α4β2nAChR availability compared to the control group were detected (figure 1).
Specific uptake of [18F]FESCH and [18F]FLUDA was observed in the striatum of control and rotenone mice (SUVR10-20 min p.i.  ~4.8, all groups included) (figure 2). PET scans revealed no significant differences in A2AR availability between control group and 2 months rotenone treatment group. However, the SUVR of the 4 months rotenone treatment group were higher compared to the control group (SUVR20-40 min p.i. 3.4 vs. 2.9, respectively), although statistically not significant due to the rather small and highly variable data set.
Altogether, the trend of these results indicates no accordance with clinical findings although a slightly increased availability of A2AR during the course of the disease can be mentioned.


Taking into account the high variability of the dataset, the investigation by PET/MR of the rotenone mouse model after 2 mo. and 4 mo. treatment shows no concordance with the clinical findings regarding α4β2nAChR and A2AR availabilities. We assume that the rotenone mouse model might not be suitable to assess PD-related changes in the availability of the two targets and thus perhaps not suitable for the investigation of the related targeting-drug.


The European Regional Development Fund and Sächsische Aufbaubank are acknowledged for financial support (Project No. 100226753).

[1] Vuorimaa A, Rissanen E, Airas L 2017, ‘In Vivo PET Imaging of Adenosine 2A Receptors in Neuroinflammatory and Neurodegenerative Disease’, Contrast Media Mol Imaging, 6975841.
[2] Meyer M, Strecker K, Kendziorra K, Becker G, Hesse S, Woelpl D, Hensel A, Patt M, Sorger D, Wegner F, Lobsien D, Barthel H, Brust P, Gertz H, Sabri O, Schwarz J 2009, ‘Reduced α4β2*–Nicotinic Acetylcholine Receptor Binding and Its Relationship to Mild Cognitive and Depressive Symptoms in Parkinson Disease’, Arch Gen Psych, 66 : 866-877.
[3] Cannon JR, Tapias V, Na HM, Honick AS, Drolet RE, Greenamyre JT 2009, ‘A highly reproducible rotenone model of Parkinson's disease’, Neurobiol Dis, 34 : 279-90.
[4] Khanapur S, van Waarde A, Dierckx RA, Elsinga PH, Koole MJ 2017, ‘Preclinical Evaluation and Quantification of 18F-Fluoroethyl and 18F-Fluoropropyl Analogs of SCH442416 as Radioligands for PET Imaging of the Adenosine A2A Receptor in Rat Brain’, J Nucl Med, 58 : 466-472
Figure 1

Representative horizontal PET images (time frame: 0-60 min) after intravenous injection of A) (-)-[18F]Flubatine showing a high radiotracer uptake in thalamus (green) and very low uptake in the cerebellum (yellow) for the 2 mo. rotenone treatment experiment, B) (-)-[18F]Flubatine showing a high radiotracer uptake in thalamus (green) and very low uptake in the cerebellum (yellow) for the 4 mo. rotenone treatment experiment.

Figure 2

Representative horizontal PET images (time frame: 20-40 min) after intravenous injection of A) [18F]FLUDA showing a high radiotracer uptake in striatum (red) and very low uptake in the cerebellum (yellow) for the 2 mo. rotenone treatment experiment, B) [18F]FESCH showing a high radiotracer uptake in striatum (red) and very low uptake in the cerebellum (yellow) for the 4 mo. rotenone treatment experiment.

Keywords: Adenosine 2A receptor, α4β2 nicotinic acetylcholine receptor, rotenone mouse model of Parkinson's disease, positon emission tomography