Abstract/Video opens by clicking at the talk title.
Abstract/Video opens by clicking at the talk title.
|12:00 p.m.||PS 06-01||
Introductory Lecture (#665)
1 Max Planck Institute for Biological Cybernetics, Tuebingen, Saxony, Germany
|12:18 p.m.||PS 06-02||
Correlation of resting-state functional connectivity with serotonin transporter and D2 receptor binding potentials in the healthy rat brain – a simultaneous PET/fMRI study (#426)
Tudor M. Ionescu1, Mario Amend1, Rakibul Hafiz2, Hans Wehrl1, Bharat Biswal2, Bernd J. Pichler1
1 Eberhard Karls University Tuebingen, Department for Preclinical Imaging and Radiopharmacy, Werner Siemens Imaging Center, Tuebingen, Germany
Resting-state functional connectivity (FC) measured using BOLD-fMRI relies on hemodynamic effects to reveal interregional brain activity correlations. However, brain activity is initiated on a molecular level by neurotransmitters. Therefore, correlation of FC with neurotransmitter activity appears feasible. In this study we show for the first time the correlation of FC with serotonin transporter binding potentials (BPs) imaged by [11C]DASB and D2 receptors (D2R) BPs imaged by [11C]Raclopride using simultaneous PET/fMRI scans.
We performed 80-minute PET/MRI scans using [11C]DASB and [11C]Raclopride (n=30 each) in healthy male Lewis rats under isoflurane anesthesia (1.3%) using a 7T MRI with a PET insert. fMRI data were acquired using an EPI-BOLD sequence (TR=2.5s, TE=18ms). The tracers were applied via bolus-infusion and the PET data were reconstructed as static scans for the 40-80 minute interval after scan start to ensure steady state, the same period being used for FC computation. Relationships between regional BPs and several FC metrics were determined by computing their inter-individual correlations. A multiple iteration approach of repeatedly splitting each cohort into two random subgroups (n=15) and comparing their outputs was applied for reproducibility computations (only reproducible results reported).
Increased regional [11C]DASB BPs were heterogenously correlated with FC. In the frontal cortex (Fig 1A) BPs in the mPFC were positively associated with short-range FC of both mPFC and OFC to each other (r=0.44 with mPFC-OFC FC), while being anticorrelated with their long-range connectivities to default-mode network regions (r=-0.48 with OFC-RS FC). Moreover, BPs in serotonergically densely innervated posterior regions had differing effects on salience network (SN) FC. While [11C]DASB BPs in the MB correlated with SN FC (r=0.34 with Ins FC strength, Fig 1B), enhanced Th [11C]DASB BPs were associated with reduced SN FC strength (r=-0.5 with Ins FC strength, Fig 1C), in line with previous cell studies indicating inhibitory effects of serotonin on thalamocortical neurons . For [11C]Raclopride CPu BPs were anticorrelated with SN FC (Fig 2), the strongest effect being observed for FC between Ins and Cg (r=-0.57), two main anchors of this network. (abbreviation lists included in figures)
Our method enhances the understanding of the molecular basis of FC by systematically evaluating its correlation with neurotransmitters on multiple levels. The approach could be used to identify inter-individual variance in FC by distinguishing such effects from pathological alterations. Finally, future studies may determine if these relationships are affected by neurodegeneration and thereby serve as potential novel biomarkers.
 Monckton JE, McCormick DA. Neuromodulatory role of serotonin in the ferret thalamus. J Neurophysiol. 2002;87(4):2124-36. doi: 10.1152/jn.00650.2001. PubMed PMID: 11929930.
Correlation of FC with [11C]DASB BPs.
Brain maps indicate correlation values between regional BPs and FC metrics in respective networks. (A) Increased BPs in the mPFC correlated with short-range FC between mPFC and OFC, while anticorrelating with the FC of these two areas to more distant regions of the DMN, such as the RS. (B) Higher BPs in the MB were associated with increased SN FC strength, with the relationship to the Ins FC strength the most dominant. (C) In contrast, BP increases in Th were related to a decline in SN FC, the strongest association being also found to the Ins FC strength. (n = 30)
Correlation of FC with [11C]Raclopride BPs.
Brain map indicates correlation values between [11C]Raclopride BPs in the CPu and salience network FC. Higher BPs in the CPu were negatively associated with SN FC. The strongest correlation was found to the FC between Ins and Cg, the two main hubs of the SN. (n = 30)
Keywords: PET/fMRI, Functional connectivity, Serotonin transporter, D2 receptor
|12:30 p.m.||PS 06-03||
Cholinergic modulation of the default mode-like network in rats (#325)
Monica van den Berg1, Lore M. Peeters1, Rukun Hinz1, Georgios A. Keliris1
1 University of Antwerp, Bio-Imaging Lab, Wilrijk, Belgium
The default mode network has gained a lot of interest during the past years since the discovery of its involvement in various neurological disorders . During attention-demanding tasks, the increased activity in the attention networks is associated with decreased activity in the DMN [2, 3]. Although this antagonistic relationship between could be repeatedly replicated, the underlying mechanisms remain to be fully understood. We investigated the influence of stimulation of basal forebrain cholinergic neurons on functional connectivity (FC) on the rat default-mode like network (DMLN).
Adult Chat-Cre Long Evans rats (n=28) were subjected to stereotactic surgery targeting the right basal forebrain with an adeno-associated (AAV) virus. Rats received either (n=16) AAV8-hSyn_DIO_hM3Dq(gq)-mCherry or a control virus (n=12), AAV8-hSyn-DIO-mCherry. Imaging procedures started at least six weeks after viral injections and were performed on a 9.4 T MRI system. Rats were anesthetized with a bolus of 0.05 mg/kg medetomidine, followed by an infusion of 0.1 mg/kg/h medetomidine and 0.4% isoflurane. RsfMRI were acquired using a T2* weighted EPI sequence. Two scan sessions were acquired for DREADDs-expressing animals, during which the effect of CNO or vehicle was evaluated, while one scan session was acquired for the animals in the control group to evaluate off-target effects of CNO.
Recently, our group demonstrated that visual stimulation decreases functional connectivity (FC) in the default mode-like network (DMLN) in sedated rats, and discovered that the DMLN is modulated by bottom-up stimulation . In the present study, we observed a significantly decreased FC in the right DMLN in the DREADDs-expressing rats starting at 5 min post-CNO injection, which reached a plateau after 10 min. We demonstrated that CNO-induced stimulation of the right BFB cholinergic system significantly decreased the FC between various DMLN ROIs (fig 1A). Additionally, the right intra- and inter-hemispheric FC of the DMLN was significantly decreased (fig 1B). In contrast, CNO injection in the sham group and saline injection in the DREADDs-expressing animals, did not elicit FC alterations. The similarity between these two findings suggest involvement of the basal forebrain cholinergic system in mediating decreased FC in DMLN upon presentation of visual stimuli in sedated rats.
Our study demonstrates decreased right intra- and interhemispheric FC in the DMLN upon activation of basal forebrain cholinergic neurons in sedated rats. We conjecture that our DREADDs-induced stimulation mediates decreased DMLN FC through similar pathways as task-related DMLN suppression. To conclude, our findings provide new critical insights into the interplay between attentional networks and DMLN in rodents.
This study was supported by the Fund of Scientific Research Flanders (FWO G048917N) and the Research Fund of the University of Antwerp (BOF DOCPRO FFB150340). Johan Van Audekerke provided excellent technical support. The computational resources and services used in this work were provided by the HPC core facility CalcUA of the University of Antwerp, the VSC (Flemish Supercomputer Center), funded by the Hercules Foundation and the Flemish Government – department EWI. We thank prof. dr. Mansvelder from the university of Amsterdam for kindly providing the Chat-Cre rats.
 Mohan, A., et al., 2016, ´The Significance of the Default Mode Network (DMN) in Neurological and Neuropsychiatric Disorders: A Review´ J Biol Med, 89, 49-57.
 Fox MD, et al. 2005, ´The human brain is intrinsically organized into dynamic, anticorrelated functional networks´ Proceedings of the National Academy of Sciences of the United States of America,. 102, 9673-9678
 Fransson P., 2005, ´Spontaneous low-frequency BOLD signal fluctuations: an fMRI investigation of the resting-state default mode of brain function hypothesis.´ Hum Brain Mapp, 26, 15-29.
 Hinz R, et al., 2019, ´Bottom-up sensory processing can induce negative BOLD responses and reduce functional connectivity in nodes of the default mode-like network in rats.´, NeuroImage, 197, 167-176.
Decreased FC in DMLN upon DREADD-induced activation of basal forebrain cholinergic neurons
Fig. 1: (A) Pairwise z-transformed FC matrix after saline injection (lower half) and after CNO injection (upper half). The colour bar represents z-scores. The stars indicate significant differences (repeated measures Two-way ANOVA with Sidak’s correction for multiple comparisons). (B) Bar-graphs of the average intra-hemispheric FC (z-scores) in the DMLN of the right hemisphere and left hemisphere, as well as the average inter-hemispheric FC. The z-scores of both groups were statistically compared using a paired Student’s T-test. * p≤0.05, ** p≤0.01, *** p≤0.001, ****p≤0.0001.
Keywords: Default mode network, Acetylcholine, DREADDs, Resting state fMRI, Basal forebrain cholinergic system
|12:42 p.m.||PS 06-04||
Exploration of the antidyskinetic effects of NLX-112, a biased agonist of 5-HT1A receptors, using [18F]FDG-PET in awake hemiparkinsonian rats (#182)
Sarah Chaib1, 2, Caroline Bouillot3, Benjamin Vidal1, Adrian Newman-Tancredi4, Luc Zimmer1, 2, 3, Elise Levigoureux1, 2
1 Université de Lyon, Université Claude Bernard Lyon 1, Lyon Neuroscience Research Center, CNRS UMR5292, INSERM U1028, Lyon, France
The gold-standard treatment for Parkinson's disease is L-DOPA, which in the long term leads to side effects known as levodopa-induced dyskinesia (LID). This side effect is explained by the conversion of L-DOPA into dopamine and its release as a ‘false neurotransmitter’ by serotonergic neurons (1). Previous promising studies in animals showed antidyskinetic effects of a novel 5-HT1A receptor biased agonist, NLX-112 despite a poorly understood mechanism. The aim of this study was to visualize, with PET imaging, the neural circuits affected by NLX-112 in hemiparkinsonian rats with or without LID.
Regional brain activity changes were examined in 5 hemiparkinsonian (HPK) rats (unilaterally lesioned with 6-OHDA) versus 5 dyskinetic hemiparkinsonian rats (HPK-LID) which received a chronic daily administration of L-DOPA (6 mg/kg) and benserazide (12 mg/kg) for 3 weeks in order to induce stable reproducible abnormal involuntary movements (AIMs). Each animal underwent 4 different PET scans in a randomized order with [18F]FDG, i.e. with an intraperitoneal injection of NaCl 0.9% (control condition), L-DOPA (6 mg/kg), NLX-112 at antidyskinetic dose (0.16 mg/kg) or L-DOPA + NLX-112. Rats were awake during the [18F]FDG uptake period (30 minutes). PET data were analyzed with a voxel-based analysis and a (MRI-based) volume of interest (VOI) method.
The voxel-based analyses showed that the metabolic profiles of HPK and HPK-LID were almost identical in terms of [18F]FDG uptake ratios in the control conditions (fig. 1.A). The injection of L-DOPA modified these profiles with an hypermetabolism in cerebellum, brainstem, substantia nigra, raphe and mesencephalic locomotor region. A decrease of [18F]FDG uptake was detected in thalamus and cortex. These discrepancies were more pronounced in HPK-LID (fig.1.B). We then compared the effect of L-DOPA to the effect of NLX-112 with L-DOPA in HPK-LID (fig.2). NLX-112 partially erased the cerebral effects of L-DOPA in HPK-LID rats, while it completely blocked the induced AIMs. MRI-based VOI analysis showed a significant increase of 8% of [18F]FDG uptake ratios in the cortical areas after NLX-112 and L-DOPA administration in HPK-LID compared of L-DOPA. A significant decrease of glucose metabolism was observed in the raphe (-14%) and non-significant in the striatum (-2%).
This study revealed for the first time the brain activation patterns of HPK-LID rats thanks to PET functional imaging. These results complete the few rare data available on imaging of LID. We showed that NLX-112 seems to prevent the LID occurrence with a double action, i.e., by activating pre-synaptic 5-HT1A autoreceptors in the raphe and post-synaptic 5-HT1A heteroreceptors in the cortex, confirming a model that has so far been hypothetical.
We thank Marco Valdebenito for his zootechnical assistance.
 Carta M, Carlsson T, Kirik D, Bjorklund A. Dopamine released from 5-HT terminals is the cause of L-DOPA-induced dyskinesia in parkinsonian rats. Brain. 2007; 130: 1819-33.
 Iderberg H, McCreary AC, Varney MA, Kleven MS, Koek W, Bardin L, et al. NLX-112, a novel 5-HT1A receptor agonist for the treatment of L-DOPA-induced dyskinesia: Behavioral and neurochemical profile in rat. Exp Neurol. 2015; 271: 335-50.
Figure 2 : [18F]FDG metabolic patterns of HPK-LID rats (n=5)
(A) L-DOPA bolus administration (to reactivate dyskinesia) and with (B) co-administration of L-DOPA bolus and NLX-112 at 0.16 mg/kg. Voxel-to-voxel statistical comparisons of [18F]FDG uptake ratio between L-DOPA injections or co-administration L-DOPA + NLX-112 and NaCl injection.
Keywords: Parkinson's disease, L-DOPA induced dyskinesia, PET imaging, biased agonist, 5-HT1A receptors
|12:54 p.m.||PS 06-05||
fMRI mapping of anesthesia-induced burst suppression across multiple mammalian species (#302)
Nikoloz Sirmpilatze1, 2, Jürgen Baudewig1, Judith Mylius1, Daniel Golkowski3, Andreas Ranft3, Rüdiger Ilg3, Jaakko Paasonen4, Olli Gröhn4, Susann Boretius1, 2
1 German Primate Center, Functional Imaging Laboratory, Göttingen, Germany
Burst Suppression (BS) is an EEG activity pattern that appears in deeply anesthetized and comatose states1. It consists of quasi-periodic bursts of high-voltage slow waves, separated by periods of depressed activity (suppressions). Bursts are coupled with rises in blood flow, and can manifest as widespread signal increases in fMRI recordings under anesthesia2. This effect has been observed in both humans3 and animals2,4, and can be leveraged to study BS. In this work we use fMRI to identify and map anesthesia-induced BS across multiple mammalian species.
We first explored how BS manifests itself in fMRI data by using a published dataset of 20 healthy human volunteers3,5. This dataset consists of simultaneous EEG-fMRI scans across three depths of sevoflurane anesthesia, including scans during BS. We used Principal Component Analysis (PCA) to decompose the fMRI data into its predominant spatiotemporal patterns, and identified components that correspond to the EEG pattern of BS. We then searched for similar patterns in task-free fMRI scans of isoflurane-anesthetized animals, which included 13 long-tailed macaques, 21 common marmosets, and 17 Wistar rats (imaged at two different sites). We successfully found a subset of scans which exhibited patterns of BS for each species, and used these to create species-specific brain-wide maps of BS.
Based on the human EEG-fMRI data, we found that scans acquired during BS were uniquely characterized by the presence of a widespread fMRI signal fluctuation. The first principal component of each scan explained a large portion of the variance, and positively correlated with most cortical voxels (Fig. 1). In addition, the time-courses of all such components closely followed the EEG pattern of BS. Components exhibiting these same characteristics were also identified in a subset of the animal fMRI scans, and were thus regarded as evidence of BS. By using the identified components as regressors for fMRI analysis, we created species-specific maps of BS (Fig. 2). The resulting maps showed striking similarities among the primates: BS was widely distributed in the cortex but was absent from early sensory areas, most notably the primary visual (V1) and somatosensory (S1) cortices. In rats, however, BS was present throughout the entire neocortex, including all sensory areas.
We hope that our description of the fMRI signatures of BS across multiple species will facilitate translational efforts within anesthesia and coma research. Our mapping of BS revealed that early sensory areas in primates are decoupled from the rest of the cortex, and that this is at odds with the pan-cortical synchrony which is seen in rats. This interesting discrepancy calls for further, more targeted invasive investigations.
AcknowledgmentWe wish to thank Kristin Kötz and Kerstin Fuhrmann for technical assistance, and Iris Steinmann for advice regarding EEG analysis. This project was partly funded through the DFG Research Center for Nanoscale Molecular Physiology of the Brain (CNMPB).
 Kroeger, D., Florea, B. & Amzica, F. Human brain activity patterns beyond the isoelectric line of extreme deep coma. PLoS One 8, e75257 (2013).
 Liu, X., Zhu, X.-H., Zhang, Y. & Chen, W. Neural origin of spontaneous hemodynamic fluctuations in rats under burst-suppression anesthesia condition. Cereb. Cortex 21, 374–384 (2011).
 Golkowski, D. et al. Coherence of BOLD signal and electrical activity in the human brain during deep sevoflurane anesthesia. Brain Behav. 7, e00679 (2017).
 Zhang, Z. et al. Isoflurane-Induced Burst Suppression Increases Intrinsic Functional Connectivity of the Monkey Brain. Front. Neurosci. 13, 296 (2019).
 Ranft, A. et al. Neural Correlates of Sevoflurane-induced Unconsciousness Identified by Simultaneous Functional Magnetic Resonance Imaging and Electroencephalography. Anesthesiology 125, 861–872 (2016).
Fig 1. The fMRI signatures of burst suppression (BS)
An fMRI scan of a human subject during BS (top left), exhibiting a widespread BOLD signal fluctuation. The first principal component (PC1) accounts for 33.5% of the explained variance and correlates strongly with most cortical voxels. Components with such properties are only found in scans acquired during BS (as confirmed by EEG), which occurs at higher anesthetic doses (bottom left). These components represent the direct hemodynamic correlates of the EEG pattern of BS (top right), and can be used as regressors for fMRI analysis, to generate brain-wide maps of BS (bottom right).
Figure 2. Burst Suppression (BS) maps across four mammalian species
Species-specific maps of BS are shown for humans, long-tailed macaques, common marmosets, and rats. The maps were generated by pooling together fMRI scans that exhibited signatures of BS. In primates, BS appears to exclude primary visual (V1) and somatosensory (S1) cortices. In rats, however, BS covers the entire neocortex.
Keywords: anesthesia, fMRI, burst suppression, rat, primate
|1:06 p.m.||PS 06-06||
Coupled functional ultrasound and electrophysiological recording in awake behaving monkeys (#514)
Julien Claron1, Thomas Deffieux1, Mickael Tanter1, Pierre Pouget2
1 ESPCI Paris, PSL Research University, Physic for Medecine Paris, INSERM U1273, CNRS, Paris, France
Whole brain neuroimaging remains hampered by the contradictory requirements of high temporal and spatial resolutions. Whole brain imaging techniques such as fMRI are based on the use of neurovascular coupling. However, its combination with electrophysiological recording for a direct assessment of local brain electric activity remains complex due to electromagnetic compatibility issues(1). Here, by coupling functional ultrasound (fUS) and electrophysiological recording we were able to directly compare the hemodynamic blood flow at high spatiotemporal resolutions and the local neuronal activity.
We used an ultrasound neuroimaging scanner (Iconeus, France). For ultrasensitive Doppler imaging, the scanning sequence consisted of 11 tilted planar ultrasonic waves (-10° to 10°, with a step of 2°), with a pulse repetition frequency of 500 Hz (Pulse repetition Frequency 5500 Hz) and a linear ultrasonic probe (128 elements, 0.1 mm pitch, 15 MHz). Electrophysiology was performed using regular tungsten microelectrode with electrical signal recorded on a MAP system (Plexon). Electrophysiological signals were filtered using the basic feature 0.5 to 135 for LFPs and 40kHz for spike activity. Here we aim to investigate the relation between the Power Doppler signal, proportional to the local Cerebral Blood Volume (CBV) and neuronal firing using simultaneous fUS and electrophysiology recording.
First, the ultrasonic imaging of local tissue motion and strain during electrode insertion enabled us to precisely locate the electrode tip with a typical 100 µm precision. After positioning of the electrode in the supplementary eye field (SEF) region at 2mm depth in the prefrontal cortex, we made electrophysiological and fUS coupled recording.
The simultaneous recording of local electrophysiological signals and global functional ultrasound -imaging data allows us to record distinct biological phenomena at the same time, but also to monitor the insertion and precise positioning of the electrophysiological probe in the brain, in real time. It paves the way to a competitive bimodal acquisition setup for behavior studies in non-human primates.
 NK Logothetis et al, 2001,'Neurophysiological Investigation of the basis of fMRI signal', Nature,
 A. Dizieux et al, 2019, 'Functional ultrasound imaging of the brain reveals propagation of task-related brain activity in behaving primates', Nature Communication
Keywords: Functional Ultrasound Imaging, Electrophysiology, Neurovascular coupling
|1:18 p.m.||PS 06-07||
Motor task induced modulations of human resting state networks (#425)
Silke Kreitz1, 2, Angelika Mennecke1, Laura C. Konerth2, Armin Nagel3, Frederic Laun3, Michael Uder3, Arnd Dörfler1, Andreas Hess2
1 University Hospital Erlangen, Department of Neuroradiology, Erlangen, Germany
Though resting state (RS) connectivity is intensively studied in the context of pathology, its biological relevance is not fully understood. It is hypothesized, that RS dynamics modulated by prior tasks promote learning1. Such modulations have been detected in human studies after performance of a motor learning task2. However, due to the tiny nature of RS modulations, those studies require spatial and temporal highly sensitive acquisition and robust analysis methods. Here, we used 7T RS fMRI and graph theoretical analysis to detect human RS modulations after a simple motor performance task.
18 healthy subjects were scanned using a 7T Siemens Magnetom. Each session consisted of an anatomy followed by 3 EPI scans (TR=2000 ms, TE=21 ms) with isotropic voxel resolution (1.5 mm). First, a RS scan (10 min) was performed, the second scan included a right hand finger tapping motor task followed by a post task RS scan. 9 subjects performed no motor task in the second session. After standard preprocessing (inter-slice time and motion correction, smoothing) RS data were bandpass filtered between 0.009 Hz and 0.08 Hz. After white matter and ventricle time course regression they were analyzed using a graph theoretical multi seed region approach3. Statistically significant modulations were determined using NBS4 and permutation corrected paired t-tests of seed region correlation maps.
The comparison of post task RS graphs between motor task and sham group revealed the left primary motor cortex (M1) and middle frontal gyrus (MFG) with the most enhanced connections (Fig. 1). Following up, we performed a voxel wise correlation of seeds within M1 and MFG and carved out regions with significantly enhanced correlations in post task RS compared to pre task RS (Fig. 2a). In accordance with previous studies5,6, these regions include structures of the frontoparietal and cerebellar RS components and the somatosensory and motor cortex. Additionally, we found altered connections between M1 and brainstem and MFG and temporal cortex. Next, we examined structures that are significantly stronger anticorrelated to M1 and MFG (Fig. 2b). These structures comprise the auditory and visual cortex for both seed regions. M1 is specifically anticorrelated to Amygdala and striatum, MFG to the tegmentum, hippocampus and occipital cortex.
7T fMRI revealed specific post motor task RS modulations not only in frontoparietal cortex and cerebellum as previously described, but also in brainstem, amygdala, hippocampus and striatum. These RS modulations include enhanced correlation and anticorrelation indicating that complex circuitries of excitations and inhibitions might be involved. Higher field strength and sophisticated analysis may allow deeper insights in RS dynamics and functions.
 Tambini, A., et al. (2010). "Enhanced brain correlations during rest are related to memory for recent experiences." Neuron 65(2): 280-290.
 Muraskin, J., et al. (2016). "Brain dynamics of post-task resting state are influenced by expertise: Insights from baseball players." Hum Brain Mapp 37(12): 4454-4471.
 Kreitz, S., et al. (2018). "A New Analysis of Resting State Connectivity and Graph Theory Reveals Distinctive Short-Term Modulations due to Whisker Stimulation in Rats." Front Neurosci 12: 334.
 Zalesky, A., et al. (2010). "Network-based statistic: identifying differences in brain networks." Neuroimage 53(4): 1197-1207.
 Albert, N.B., et al. (2009). "The resting human brain and motor learning." Curr Biol 19(12): 1023-1027.
Keywords: Resting State, fMRI, motor task, short term modulation