HBHL Symposium 2023 | May 4-5th, 2023

Healthy Brain Healthy Lives, McGill University

Published online: May 4-5th, 2023

Effects of anticholinergic burden on hippocampal subfields and surrounding white matter volume in first-episode psychosis

Agnès Belkacem1, Katie M. Lavigne1, 2, Delphine Raucher-Chéné1, Carolina Makowski3, Mallar Chakravarty1, Ridha Joober1, Ashok Malla1, Jai Shah1, Martin Lepage1

1Douglas Research Centre, McGill University, Montreal, QC, Canada
2Montreal Neurological Institute-Hospital, McGill University, Montreal, QC, Canada
3Department of Radiology, the University of California, San Diego, CA, United States

Corresponding Author: Martin Lepage, email: martin.lepage@mail.mcgill.ca


Using multiple medications is not uncommon in psychosis. However, considering their anticholinergic burden (cumulative effect of medications that block the cholinergic system) is. Evidence suggests that patients receiving high doses of antipsychotics had reduced hippocampal volume. However, limited studies have examined the possible association with a high anticholinergic burden. Our study, therefore, aimed to explore this association in patients with first-episode psychosis. We expected patients with the highest anticholinergic burden to show a more significant reduction in hippocampal volume than patients with low anticholinergic burden and healthy controls at months 3 (baseline) and 12. Patients (n=83) followed at the PEPP-Montreal clinic, and controls (n=55) completed a 3T MRI at both assessments. Results at baseline suggested that high anticholinergic burden patients showed reduced left fimbria volumes compared to low anticholinergic burden patients and controls, even after controlling for antipsychotic dosage. Findings over time also revealed that high anticholinergic burden patients showed reduced left fimbria volumes compared to low anticholinergic burden patients and controls, even after controlling for antipsychotic dosage. Overall, the associations found between high anticholinergic burden and hippocampal volume provide further evidence for the need to prescribe lower doses of medication and, therefore, a lower anticholinergic burden in early psychosis.

Structural correlates of polygenic risk for temporal lobe epilepsy

Alexander Ngo1, Lang Liu2, Sara Larivière3, Valeria Kebets3, Serena Fett3, Jessica Royer3,4, Clara F. Weber3,5, Eric Yu2, Budhachandra Khundrakpam6, Alan C. Evans6, Raúl Rodríguez-Cruces3, Luis Concha7, Zhiqiang Zhang8, Leon Qi Rong Ooi9, B.T. Thomas Yeo9, Neda Bernasconi10, Birgit Frauscher4, Andrea Bernasconi10, Ziv Gan-Or2, Boris C. Bernhardt3

1Integrated Program in Neuroscience, McGill University, Montreal, QC, Canada
2Deparment of Human Genetics, McGill University, Montreal, QC, Canada
3Multimodal Imaging and Connectome Analysis Laboratory, Montreal Neurological Institute and Hospital, McGill University, Montreal, QC, Canada
4Analytical Neurophysiology Laboratory, Montreal Neurological Institute and Hospital, McGill University, Montreal, QC, Canada
5Social Neuroscience Laboratory, Translational Psychiatry Unit, Lübeck University, Lübeck, Germany
6McGill Centre of Integrative Neuroscience Laboratory, Montreal Neurological Institute and Hospital, McGill University, Montreal, QC, Canada
7Instituto de Neurobiología, Universidad Nacional Autónoma de México, Campus Juriquilla, Queretaro, Mexico
8Department of Medical Imaging, Jinling Hospital, Nanjing University School of Medicine, Nanjing, China
9Department of Electrical and Computer Engineering, National University of Singapore, Singapore, Singapore
10Neuroimaging of Epilepsy Laboratory, Montreal Neurological Institute and Hospital, McGill University, Montreal, Canada

Corresponding Author: Boris Bernhardt, email: boris.bernhardt@mcgill.ca


Epilepsy is associated with brain-wide structural compromise and complex genetic architecture. Interactions between genetic factors and dynamic structural changes, however, have not yet been considered in the neurodevelopmental features of epilepsy. Here, we assessed the genetic risk burden for the temporal lobe epilepsy (TLE) on brain morphology in healthy children and adolescents. Using a neuroimaging-genetics approach based on thousands of typically developing children from the Pediatric Imaging, Neurocognition, Genetics (PING) and Adolescent Brain Cognitive Development (ABCD) datasets, we computed individualized measures of polygenic genetic risk for TLE (PRS-TLE). Brain-wide linear modelling assessed the relationship between PRS-TLE and cortical morphology. These imaging-genetic maps were spatially correlated to patterns of grey matter atrophy from different, independent multi-site datasets aggregating data from healthy controls and patients with an epilepsy diagnosis. Widespread thinning across the cortex was observed in children with high PRS-TLE, mainly targeting temporo-limbic cortices (pFWE < 0.05) and consistent across studies and sites (F1A). Moreover, stratification of PRS-HS effects according to an atlas of functional zones further identified peak values in paralimbic regions (F1B). However, spatial comparisons between PRS-mediated shifts and TLE-specific atrophy revealed no significant correlations (all p > 0.05; F2). Capitalizing on recent imaging-genetic initiatives, our analyses revealed widespread thinning associated with increased PRS-TLE that did not reflect atrophy patterns observed in adult case-control studies. PRS-TLE effects might not capture effects of overall disease burden, but rather represent genetically mediated structural risk factors that are clustered around temporo-limbic epicentres, overall providing novel insights into the aetiology of TLE.

Generating harmonized neuroimaging and clinical-behavioral data annotations using Neurobagel

Alyssa Dai1, Jonathan Armoza1,2, Arman Jahanpour1, Nikhil Bhagwat1, Brent McPherson1, David Keator3, Yaroslav O. Halchenko4, Satrajit Ghosh5, Jeffrey Grethe6, David Kennedy7, Sebastian Urchs1, Mallar Chakravarty8,9,10, Jean-Baptiste Poline1,11

1McConnell Brain Imaging Centre, Montreal Neurological Institute, Faculty of Medicine, McGill University, Montreal, QC, Canada
2Department of English, New York University, New York, NY, United States
3Department of Psychiatry and Human Behavior, University of California, Irvine, CA, United States
4Department of Psychological and Brain Sciences, Dartmouth College, Hanover, NH, United States
5McGovern Institute for Brain Research, Massachusetts Institute of Technology, Cambridge, MA, Estados Unidos
6Department of Neurosciences, University of California, San Diego, CA, United States
7Department of Psychiatry and Radiology, University of Massachusetts Chan Medical School, Worcester, MA, Estados Unidos
8Cerebral Imaging Centre, Douglas Mental Health University Institute, Montreal, QC, Canada
9Department of Biological and Biomedical Engineering, McGill University, Montreal, QC, Canada
10Department of Psychiatry, McGill University, Montreal, QC, Canada
11Department of Neurology and Neurosurgery, McGill University, Montreal, QC, Canada

Corresponding Author: Jean-Baptiste Poline, email: jean-baptiste.poline@mcgill.ca


Combining participants who meet specific criteria across datasets is essential for larger sample sizes in neuroimaging studies, but requires dataset metadata harmonization. However, adoption of standards for phenotypic (e.g., neurocognitive, clinical) metadata has been hampered by the difficulty of applying community-driven vocabularies and linked data principles. We introduce a browser-based tool for harmonizing phenotypic data that facilitates the use of controlled terminologies. This annotator is part of Neurobagel, an ecosystem aiming to provide participant-level cohort search across distributed datasets via a common, reusable data schema created using controlled vocabularies. The Neurobagel annotator (annotate.neurobagel.org) provides an intuitive interface for annotating tabular data through visually-guided steps: 1) mapping table columns to common data elements (e.g., age, diagnosis, cognitive assessment), 2) mapping unique column values to terms from existing, well-established controlled vocabularies, 3) storing annotations as BIDS-compliant data dictionaries to yield harmonized phenotypic and neuroimaging metadata for the dataset. These metadata are then ingested by Neurobagel’s knowledge graph to enable federated cohort search, and are reusable by other software to enable cross-dataset analyses. The data schema used for annotation was created in collaboration with the ReproNim and NIDM communities to maximize extensibility and reusability. We demonstrate proof-of-concept by harmonizing the Quebec Parkinson Network and Parkinson’s Progression Markers Initiative datasets (n>3000). The simplicity, modularity, and BIDS compatibility of this tool offer easy integration with existing data sharing platforms (OpenNeuro, NeuroHub, etc). Overall, the Neurobagel annotator will provide researchers with an efficient workflow for describing participant data using standardized terms, supporting FAIR compliance across datasets.

Exploratory Analyses of Obstructive Sleep Apnea and Sleep Structure in Parkinson’s Disease Motor Subtypes

Amanda Scanga1, Ann Robinson2, Marianne Gingras2, Andrea Benedetti3,4, Anne-Louise Lafontaine5, R John Kimoff2, Marta Kaminska1,2,4

1Division of Experimental Medicine, McGill University, Montreal, QC, Canada
2Respiratory Division, Sleep Laboratory, McGill University Health Centre, McGill University, Montreal, QC, Canada
3Department of Epidemiology, Biostatistics and Occupational Health, McGill University, Montreal, QC, Canada
4Respiratory Epidemiology and Clinical Research Unit, McGill University Health Centre, McGill University, Montreal, QC, Canada
5Montréal Neurological Institute-Hospital, McGill University Health Centre, McGill University, Montreal, QC, Canada

Corresponding Author: Marta Kaminska, email: marta.kaminska@mcgill.ca


Background: Parkinson’s disease (PD) can be divided into motor subtypes: postural instability/gait difficulty (PIGD), tremor dominant, and indeterminate. This study aimed to assess differences in obstructive sleep apnea (OSA) between the PIGD and non-PIGD subtypes. Methodology: PD participants with or without OSA (defined as apnea-hypopnea index (AHI) ≥ 15 events/hour on overnight polysomnography) were included. Patients were separated into two groups: PIGD and non-PIGD. Logistic regression adjusted for age, sex, body mass index, levodopa equivalent dose, proportion of sleep time in stage N3, proportion of sleep time in stage rapid-eye movement (REM), and proportion of sleep time in the supine position was used to determine if the prevalence of OSA differs across groups. Linear regression was used to explore differences between groups in AHI and other respiratory parameters (adjusted for the same confounders). 3 Subset analyses excluding patients on psychoactive medication, dopaminergic medication, or both, were performed. Results: We studied 146 participants: mean AHI 31.2±21/h. Fewer patients had OSA in the PIGD versus non-PIGD subtypes [adjusted OR 0.5, 95%CI (0.2, 1.1), p=0.096 in full sample; adjusted OR 0.3, 95%CI (0.06, 0.9), p=0.038) in subset 3 (n=60)]. The AHI was lower in the PIGD group (p=0.048 in full sample; p<0.05 in all subsets). Discussion: OSA was more frequent in the non-PIGD subtype when assessing subsets free of psychoactive medication and dopaminergic medication. The PIGD group also had a lower AHI, adjusting for potential confounders including differences in sleep architecture, in subsets excluding psychoactive medication and nighttime dopaminergic medication.

Investigating the Role of Intrinsic and Extrinsic Factors in Neuronal Heterogeneity: A Computational Study on Sensory Pyramidal Cells

Amin Akhshi1, Maurice Chacron1, Anmar Khadra1

1Department of Physiology, McGill University, Montreal, QC, Canada

Corresponding Author: Anmar Khadra, email: anmar.khadra@mcgill.ca


Sensory neurons display ubiquitous heterogeneity in their spiking activity across different sensory modalities in the brain, even when encoding the same stimulus property. While recent evidence suggests that such heterogeneity can benefit neural coding and behavior, how such heterogeneous spiking activities can arise due to different factors including morphological, molecular, and electrophysiological differences is not yet fully understood. Over the past decades, computational modeling has emerged as a powerful tool for understanding the relative contributions of these factors during heterogeneous firing activities in neuronal populations. However, whether this approach can successfully explain such relationships in vivo, where neurons are in a high-conductance state due to massive synaptic bombardment, remains unknown to date. In this study, we address this important question using a novel and realistic computational workflow to develop biophysically detailed modeling of pyramidal cells within the electrosensory system of weakly electric fish combined with extracellular in vivo recordings to investigate the role of different intrinsic and extrinsic sources that affect the firing activity of pyramidal cells. Specifically, by varying parameters through optimization, we showed that our conductance-based computational model successfully reproduced the highly heterogeneous spiking activities seen experimentally. This technique enabled us to identify model parameters involved in spiking heterogeneities of pyramidal cells in the ELL. Furthermore, we tested model predictions experimentally by performing two different pharmacological manipulations to show how the model explains the effect of synaptic bombardments as well as neuromodulatory serotonergic input on the spiking activity patterns of pyramidal cells.

Contrast detection in noise with pink and white temporal spectra

Annabel Wing-Yan Fan1, Alex S Baldwin1

1Department of Ophthalmology & Visual Sciences, McGill University, Montreal, QC, Canada

Corresponding Author: Alex S. Baldwin, email: alexander.baldwin@mcgill.ca


Human visual performance in a detection task can be analysed on the basis that the performance is limited by both an intrinsic “noise” in the visual system and the efficiency of the strategy used to process the noisy information. By measuring detection performance in the presence of different amounts of “external” noise (applied to the target), one can infer the equivalent “internal” noise and processing efficiency. Previous research has focused on defining the optimal spatial characteristics of external noise, but relatively little attention has been paid to the temporal characteristics. Our study explores visual detection performance in temporally constant, pink, or white noise when detecting a target grating. We varied the temporal spectra of noise (constant, pink, white) and presented that noise at four levels (no noise, 18dB, and +12dB and +15dB of a participant’s noise detection threshold) in a 2-alternative forced choice task. In that task, the participant was asked to detect a grating target in that temporal noise. From the data, we generated noise masking functions and estimated equivalent internal noise, and efficiency. Our analysis revealed no significant main effect of temporal spectra on efficiency or equivalent noise. We compared participants' performance with an ideal observer model and investigated the impact of varying presentation duration. In future studies, we will compare the results from young adults to those of older adults. This research has important implications for understanding how the visual system processes information in noisy environments and how this may change over the course of the lifespan.

Social defeat stress alters PFC molecular markers dissimilarly between sexes in adult mice

Ashraf Mahmud1,2, Radu Gabriel Avramescu2, Jose Maria Restrepo1,2, Cecilia Flores2,3

1Integrated Program in Neuroscience, McGill University, Montreal, QC, Canada
2Douglas Mental Health University Institute, Montreal, QC, Canada
3Department of Psychiatry and of Neurology and Neurosurgery, McGill University, Montreal, QC, Canada

Corresponding Author: Cecilia Flores, email: cecilia.flores@mcgill.ca


Elevated levels of the Netrin-1 guidance cue receptor gene DCC in the adult prefrontal cortex (PFC) is a consistent trait of major depressive disorder in humans and plays a causal role in susceptibility to chronic social defeat stress (CSDS) in adult male mice. Although depression is twice more likely in women than men, the molecular mechanisms underlying this disparity remain unknown. Here we investigated if DCC receptors or microRNA-218, a posttranscriptional repressor of DCC, are dysregulated in the PFC of adult female mice following CSDS. Adult (PND75±15) C57BL/6J female mice were exposed to CSDS for 5 minutes each day for 10 days. Control mice were housed with a different conspecific every day. Following CSDS, mice were assessed in the social interaction test (SIT) and 24h later, PFC tissue was collected to measure Dcc mRNA expression. Defeated females were segregated into resilient and susceptible groups based on the SIT. Compared to control mice, susceptible and resilient mice showed an increase in body weight and a deficit in the nestlet shredding test, a measure of self-care-like behaviour. There was no difference in PFC Dcc and Netrin-1 mRNA levels or Unc5c, another predominant Netrin-1 receptor gene. However, microRNA-218 expression was elevated in susceptible mice compared to the control mice. These data indicate that, contrary to males, CSDS in females does not directly alter the Netrin-1/DCC guidance cue system in the adult PFC, and that currently unidentified sexually dimorphic molecular processes involving microRNAs may mediate stress vulnerability. Keywords: Netrin-1/DCC guidance cue system, Depression, Chronic social defeat stress, microRNA-218.

Cortical quantitative MRI metrics are sensitive to amyloid and tau PET in preclinical Alzheimer’s Disease

Aurélie Bussy1,2, Alfie Wearn5, Raihaan Patel1,4, Giulia Baracchini5,6,7, Sylvia Villeneuve3,6, Judes Poirier3,6, John CS Breitner6, Gabriel A. Devenyi1,3, R. Nathan Spreng5,6,7, M. Mallar Chakravarty1,2,3,4

1Computational Brain Anatomy Laboratory, Cerebral Imaging Centre, Douglas Mental Health University Institute, Montreal, QC, Canada
2Integrated Program in Neuroscience, McGill University, Montreal, QC, Canada
3Department of Psychiatry, McGill University, Montreal, QC, Canada
4Department of Biomedical Engineering, McGill University, Montreal, QC, Canada
5Department of Neurology and Neurosurgery, McGill University, Montreal, QC, Canada
6Centre for the Studies on the Prevention of AD, Douglas Mental Health University Institute, Verdun, QC, Canada
7McConnell Brain Imaging Centre, Montreal Neurological Institute, McGill University, Montreal, QC, Canada

Corresponding Author: Mallar Chakravarty, email: mallar.chak@gmail.com


While the Canadian population is living longer, Alzheimer’s disease (AD) prevalence increases. Here, we combined structural MRI to quantify atrophy, quantitative MRI metrics to estimate different sources of microstructure (myelin, iron, water) and PET imaging to extract amyloid and tau content. The goal of this study is to investigate a) age sensitivity of morphometry and/or microstructure metrics, and b) amyloid and tau in individuals with familial risk-factor for AD to determine if MRI-derived metrics could be indicative of early pathological accumulation in preclinical AD. To investigate brain myelin (magnetization transfer saturation: MTsat and R1), iron (R2*) and free water content (proton density: PD), a whole-brain quantitative multi-parametric protocol was acquired. Amyloid was quantified using [18F]NAV4694 PET scans and tau with [18F]AV1451 scan. Non-negative matrix factorization (NMF) was used to parcelate the cortex in a part-based representation using vertex-wise metrics (CT, SA, R1, R2*, PD, MTsat). The relationships between NMF subject-weights and age, sex and the interaction between amyloid and tau groups was tested with linear models. Increased age was related to lower CT and R1 throughout the cortex, lower PD except in the occipital lobe and lower MTsat in the temporal and occipital lobes. In high tau PET individuals, higher cortical PD was found throughout the cortex except in the anterior cingulum and frontal lobe while higher MTsat was found in the temporal lobe. Beyond the expected age-related CT and myelin decrease, our results demonstrated that tau accumulation was related to higher PD (interpreted as a sign of macromolecular content loss).

Developmental axonal swellings depend on action potential propagation signaling

Bruna de Souza1,2, Alanna Watt2

1Integrated Program in Neuroscience, McGill University, Montreal, QC, Canada
2Biology Department, McGill University, Montreal, QC, Canada

Corresponding Author: Alanna Watt, email: alanna.watt@mcgill.ca


The development of axons and the establishment of contacts between neurons is an important step during brain development and the subsequent formation of functional brain circuits. Recently we demonstrated that axonal swellings appear transiently on cerebellar Purkinje cell axons during postnatal development where they propagate action potentials with higher fidelity and are associated with enhanced cerebellar function. An understanding of how axonal swellings form is thus important. We aimed to further investigate the role of firing in the formation of axonal swellings. We performed 2-photon time-lapse imaging in acute cerebellar slices from juvenile mice (P10 - P14) and added sub-saturating levels of tetrodotoxin (TTX; 1, 2.5, 5 and 10 nM) to parametrically block different concentrations of Na+ channels and as a result, impair action potential propagation in the axon to different degrees. Paradoxically, we found that low concentrations of TTX that block fewer somatic action potentials resulted in more axonal swellings that formed faster. Higher sub-saturating concentrations of TTX blocked more somatic action potentials but resulted in fewer and slower-forming axonal swellings. Our results support a model in which action potential propagation failure is required for axonal swelling formation, but a balance between successful action potentials and failures exists. These data suggest that mild impairment of action potential propagation may be the optimal trigger to activate signalling cascades to form swellings and boost action potential fidelity.

The influence of generativity on purpose in life is mediated by social support and moderated by prefrontal connectivity between the vmPFC and dlPFC in older adults at risk for Alzheimer’s Disease

Caitlin Walker4, Linda Li2, Adrian Noriega de la Colina1,3, Jennifer Tremblay-Mercier3, Sylvia Villeneuve1,3, R. Nathan Spreng1, Maiya R. Geddes1,3

1Montreal Neurological Institute, McGill University, Montreal, QC, Canada
2Department of Pharmacology and Therapeutics, McGill University, Montreal, QC, Canada
3Douglas Mental Health University Institute, McGill University, Montreal, QC, Canada
4Neurology and Neurosurgery, Integrated Program in Neuroscience, McGill University, Montreal, QC, Canada

Corresponding Author: Maiya R. Geddes, email: maiya.geddes@mcgill.ca


Objectives: Generativity is the desire and actions to improve the well-being of younger generations. Generativity is associated with purpose in life, with both being shown to independently improve cognition in older adults. Our aims were to identify the neural substrates supporting generativity and determine the mechanism underlying the relationship between generativity and purpose in life in older adults. Method: Forty-three older adults (Mage = 70.33, 79.1% female) at risk for Alzheimer’s disease underwent resting-state functional magnetic resonance imaging and completed questionnaires assessing generativity, social support, and purpose in life. Seed-to-voxel analyses examined if resting-state functional connectivity (rsFC) of the ventromedial prefrontal cortex (vmPFC) and ventral striatum, key nodes at the intersection of subjective valuation and self-transcendence, were associated with generativity. Moderated mediation models examined if social support or rsFC mediated or moderated the association between generativity and purpose in life, respectively. Results: Generative desire was associated with enhanced rsFC between the vmPFC and right dorsolateral prefrontal cortex (rdlPFC). Affectionate social support fully mediated the relationship between generative desire and purpose in life, and rsFC between the vmPFC and rdlPFC significantly moderated this association. Discussion: This study is the first to examine the rsFC underlying generativity and provides mechanistic insight into how purpose in life is enhanced through generative desire. Generative desire is supported by rsFC implicated in value-based social decision making and is associated with purpose in life through enhanced love and affection from others. This knowledge contributes to future developments of personalized interventions that promote resilience in at-risk aging.

Motivational mechanism for social versus monetary reward in older adults at risk of Alzheimer's Disease

Caitlin Walker1, Andréanne Powers2, Nicolas Lavoie1, Maiya R. Geddes3,4

1McGill University, Montreal, QC, Canada
2Laval University, Quebec, QC, Canada
3Montreal Neurological Institute-Hospital, McGill University, Montreal, QC, Canada
4Harvard University, Cambridge, MA, United States

Corresponding Author: Maiya R. Geddes, email: maiya.geddes@mcgill.ca


Previous research demonstrates that enhanced motivational processes are associated with a reduced risk of Alzheimer’s disease (AD). In older adults, social stimuli are intrinsically rewarding and activate regions of the brain’s reward network more than extrinsic rewards such as money. However, whether older adults decide to exert greater effort for social than monetary rewards, and whether this varies by award value, is underexplored. To investigate this question, we administered a novel, computerized reward-effort decision making paradigm to 37 cognitively normal older adults who are at risk for AD. On each trial, a horizontal line presented at a certain height indicates the number of key presses required to move an arrow from the bottom of the screen to the height of the line. Participants decide whether they would exert this physical effort for social (donation to charity) or monetary rewards that vary between $0.50 and $3.00. The results revealed statistically significant three-way interactions for both the proportion of ‘yes’ responses and decision response times (RTs). The willingness to exert high effort was significantly higher for the social than monetary condition at the highest level of reward. RTs for high-value rewards were significantly higher for the social condition at all levels of effort, except for medium effort levels where RTs for monetary rewards were higher. Given that apathy is a common neuropsychiatric symptom in early AD, a greater understanding of what intrinsically motivates at-risk older adults to exert effort can help promote health behaviours that reduce the risk of cognitive impairment.

Brain chaoticity and avalanche criticality are markers of anesthetic-induced unconsciousness

Charlotte Maschke1, Jordan O’Byrne2, Karim Jerbi2, George A. Mashour3, Stefanie Blain-Moraes1

1McGill University, Montreal, QC, Canada
2University of Montreal, Montreal, QC, Canada
3University of Michigan, Ann Arbor, MI, United States

Corresponding Author: Stefanie Blain-Moraes, email: stefanie.blain-moraes@mcgill.ca


Neurons appear to have the capacity to self-organize to produce highly complex and flexible patterns – a phenomenon known as criticality. Critical dynamics occur when a system is poised between two dynamical regimes, such as stability and chaos. Electrophysiological dynamics poised at criticality have been proposed to underpin healthy brain function as well as the emergence of consciousness. The aim of this project is to investigate the sensitivity of avalanche criticality and the edge of chaos, to anesthetic-induced loss of consciousness in healthy adults. We analyzed an existing dataset of nine healthy adults who underwent a protocol of propofol and isoflurane anesthesia. Electroencephalogram was recorded using a 128-channel sensor net before, during and after the anesthetic protocol. Avalanche criticality was estimated using α, the power law exponent of the empirical avalanche duration distribution. Chaoticity was estimated using the modified 0-1 chaos test. Statistical significance was tested using RMANOVA. Anesthesia had a significant effect on chaoticity [F(4, 32) =20.44, p <0.001, η2=0.62], with stronger chaoticity during anesthetic-induced unconsciousness. Similarly, Anesthesia had a significant effect on avalanche criticality [F(4, 32) =13.19, p <0.001, η2=0.52], with a steeper power law during anesthetic-induced unconsciousness. Post-hoc tests revealed that both measures significantly differentiated the pre-anesthetic state from light and deep sedation (p<0.01). Our results support the theory that consciousness requires the brain to be poised near criticality and suggest that different types of criticality – namely the edge of chaos and avalanche criticality - can be used as a marker of anesthetic-induced loss of consciousness.

Characterization of altered Purkinje cell physiology and motor performance in a mouse model of chronic stress

Connor O'Donnell1, Dr. Aparna Suvrathan2

1Integrated Program in Neuroscience, McGill University, Montreal, QC, Canada
2Research Institute of the McGill University Health Centre, Departments of Neurology and Neurosurgery and Pediatrics, McGill University, Montreal, QC, Canada

Corresponding Author: Aparna Suvrathan, email: aparna.suvrathan@mcgill.ca


The ability to adapt in response to stressful environments is crucial for survival. However, when exposure to stress becomes chronic, maladaptive changes to neurophysiology and behaviour can arise. The cerebellum, a brain region known for its critical role in motor coordination, has been found to undergo structural and functional changes in response to stress. The precise alterations to cellular physiology in the cerebellum following chronic stress, and the associated consequences for cerebellum-mediated behaviours, are not well understood. In the present study we employed a mouse model of chronic stress to address this gap in understanding. To identify chronic stress-induced alterations in Purkinje cells, the sole output neurons of the cerebellar cortex, we used whole-cell recordings to study their electrophysiological properties. Strikingly, we observed that the firing rate of Purkinje cells is significantly increased in mice exposed to chronic stress. As these neurons are directly responsible for mediating cerebellar learning, we investigated wether this increase in firing rate would be correlated with a behavioural effect. The accelerating rotarod, a cerebellum-dependent motor learning task, was used to assess impairments in cerebellar function induced by chronic stress, while a standard open field test was used to assess baseline locomotor activity levels. Our preliminary findings suggest that chronic stress has an impact on motor performance in these mice. These results provide a critical first step towards characterizing the full extent of cerebellar alterations induced by chronic stress and the behavioural consequences associated with them.

Investigating the temporal dynamics of dichoptic masking

Daniel Gurman1, Alexandre Reynaud2

1Department of Ophthalmology and Visual Sciences, McGill University, Montreal, QC, Canada
2McGill University, Montreal, QC, Canada

Corresponding Author: Alexandre Reynaud, email: alexandre.reynaud@mail.mcgill.ca


In the standard model of binocular combination, inputs from the two eyes not only sum together but also suppress each other. Interocular suppression is often characterized through dichoptic masking, a phenomenon in which the detection of a target presented to one eye is reduced by a noise mask presented to the other eye. However, the temporal dynamics of interocular suppression are not well understood. In particular, the relationships between simultaneous masking, backward masking, and forward masking are not known for dichoptic stimuli. Our objective was to better understand these relationships. We employed a dichoptic suppression paradigm using a two-alternative-force-choice task. Stimuli were displayed on a passive 3D screen. Participants indicated the orientation of a target grating presented to one eye while a pink noise mask was presented to the other eye at the same spatial location but at a different time. In a second viewing condition, the mask and target were presented monocularly. The contrast of the target was adjusted using a 2-up 1-down staircase. Thirteen interstimulus intervals between the mask and the target were used to individually investigate the three masking types. Our results revealed the presence of a masking effect for all three masking types. Surprisingly, the strongest masking effect was found in the backward masking condition. These findings provide novel insight into the temporal dynamics of dichoptic masking, particularly in revealing the strong effect of backward masking, and may have implications in the general understanding of amblyopic suppression and in the development of suppression-related treatments for amblyopia.

Role of radial astrocytes on the encoding of visual signals and plasticity in the developing Xenopus laevis optic tectum

David Foubert1, Edward Ruthazer2

1Integrated Program in Neuroscience, McGill University, Montreal, QC, Canada
2Department of Neurology and Neurosurgery, McGill University, Montreal, QC, Canada

Corresponding Author: Edward Ruthazer, email: edward.ruthazer@mcgill.ca


During late stages of brain development, efficient circuit wiring relies on instructive cues from sensory experience and neural activity to drive synaptic plasticity and achieve mature functional connectivity. Initially thought to only implicate neurons, more recently astrocytes have been shown to be directly involved in the neural plasticity mechanisms that mediate structural and functional circuit refinement. Given that astrocytes invade brain regions, providing guidance and structural support, do they play a role in tuning their functional properties as well? Using the visual system of Xenopus laevis we conducted live calcium imaging while presenting visual stimuli to study the role of radial glia in neural plasticity. To selectively control the activity state of radial astrocytes of the tectum they were preferentially transfected with chemogenetic calcium channels. We observed only a small number of cells responding to the visual stimuli, which is consistent with previous findings of sparse information encoding. Glial activation increased the strength and duration of network wide short- and long-term depression after plasticity inducing stimuli. On the other hand, the activity of the neuropil, where retinal ganglion axons synapse onto tectal neurons, saw a significant increase in responsiveness to the stimuli after induction and glial activation, in contrast to the control group which had no changes in responsiveness after training. Our results indicate that glia are acting between the synapses and the cell body to modulate signal transduction and plasticity.

Ultra-high field precision neuroimaging of the individual human brain

D.G. Cabalo1,2, Y. Wang1,2, S. Tavakol1,2, J. Royer1,2, J. DeKraker1,2, J. Smallwood3, R.R. Cruces1,2, B. Bernhardt1,2

1Multimodal Imaging and Connectome Analysis Laboratory, McConnell Brain Imaging Centre, Montreal Neurological Institute and Hospital, McGill University, Montreal, QC, Canada
2Montreal Neurological Institute and Hospital, McGill University, Montreal, QC, Canada
3Queens University, Kingston, ON, Canada

Corresponding Author: Boris Bernhardt, email: boris.bernhardt@mcgill.ca


Multimodal neuroimaging has significantly advanced our knowledge of the human brain by allowing non-invasive visualization and examination of brain structure and function. Precision neuroimaging builds upon this foundation by mapping neural activity and connectivity patterns at higher spatial and temporal resolutions, resulting in a more precise and individualized understanding of brain function[1,2]. Ultra-high field (UHF) neuroimaging operates at magnetic field strengths of 7Tesla or higher, is a rapidly evolving subfield within precision neuroimaging that offers greater spatial and temporal resolution than traditional MRI[3,4]. This study aims to provide a multimodal neuroimaging dataset utilizing UHF 7T MRI, acquired across multiple sessions, as a valuable resource for the neuroscience community. The UHF imaging protocol was implemented at the Montreal Neurological Institute. To date, we have collected high-precision data from 5 healthy participants (target 10), who underwent quantitative T1 relaxometry (qT1), diffusion-weighted MRI, magnetization transfer ON/OFF, T2* weighted imaging, and resting-state fMRI across three sessions. All data collected have been preprocessed using micapipe[5], an open-access surface-based preprocessing software. Alongside anonymized raw data, we plan to release cortex-wide connectomes derived from resting-state and diffusion tractography data. Researchers will also have access to averaged qT1, diffusion, and resting-state scans from the three sessions. This dataset has the potential to enhance our understanding of the brain and aid in the development of more effective treatments for neurological disorders. 1. Gordon (2017). Neuron, 95(4), 2. Poldrack (2017). Neuron, 95(4), 727–729, 3. Dumoulin (2018). NeuroImage, 168, 345–357, 4. Nakada (2007). Brain Development, 29(6), 325–335. 5. Cruces, (2022). NeuroImage, 263, 119612.

Low-cost inertial measurement units for autonomous at-home monitoring of upper extremity activities of daily living in stroke patients

Redman, E. X.1, Archambault, P. S.1

1Integrated Program in Neuroscience, McGill University, Montreal, QC, Canada

Corresponding Author: Philippe Archambault, email: philippe.archambault@mcgill.ca


Upper extremity (UE) impairments are common in stroke patients, affecting up to 80% acutely and 40% chronically. Performance of activities of daily living (ADLs) is a reliable predictor of stroke recovery, and at-home exercises are often prescribed to improve ADL performance. However, adherence to these exercises is often low, leading to reduced recovery outcomes. Inertial measurement units (IMUs), sensors capable of measuring linear and angular displacement, offer a convenient means of monitoring stroke patients long-term and may support adherence to prescribed exercises while providing valuable recovery insights. This research study focuses on developing a standardized protocol for detecting and classifying complex UE ADLs in stroke patients using low-cost IMUs. The aim is to achieve 85% sensitivity in ADL classification while maintaining comparable specificity to distinguish from non-UE functional movements. With a sample of 30 participants with chronic stroke, the study involves motor assessments, fitting participants with seven IMUs, and having them complete 200 trials of various movements. Data analysis includes preprocessing of raw IMU data, sensor fusion, and evaluation of model architectures such as DCNN and CNN-LSTM. Upon completion, the open-sourced Python scripts for calibration, data collection, preprocessing, and analysis will be provided. This project aims to validate the use of low-cost IMUs for autonomous at-home monitoring of ADLs in stroke patients, ultimately improving recovery outcomes through increased adherence to prescribed exercises and better-informed clinical decision making.

The Autonomous Sensory Meridian Response Questionnaire (ASMR-Q): Validation in both naive and familiar individuals

Elise Desbarats1, Mathieu Roy1, Isabelle Plante2

1McGill University, Montreal, QC, Canada
2Université de Québec à Montréal, Montreal, QC, Canada

Corresponding Author: Mathieu Roy , email: mathieu.roy3@mcgill.ca


Autonomous Sensory Meridian Response (ASMR) is a pleasant tingling paresthesia that spreads from the central nervous system to the periphery in response to various auditory and visual triggers. Individuals report that ASMR helps them to cope with depression, anxiety and chronic pain, yet to investigate these claims we require a validated way to measure the sensation. We currently lack measures to quantify ASMR responding in those who are unfamiliar with the concept or who lack experience with the genre of online media designed to trigger ASMR, limiting study recruitment to those who engage with ASMR on the internet. Two studies were conducted to validate a questionnaire that is accessible to those who are unfamiliar with ASMR by assessing self-report of mundane offline experiences. In Study 1, the ASMRQ was administered to ASMR-naive and ASMR-familiar individuals (n = 307), and exploratory analysis identified factors of the ASMR experience. In Study 2, these factors were replicated in a new sample (n = 255) and we identified subscales common to both online and mundane settings. Confirmatory factor analysis was then performed on a final version of the ASMRQ retaining only mundane items. The ASMRQ shows good psychometric properties for use as a measure of ASMR sensation, with 5 subscales demonstrating high internal consistency. The ASMRQ also showed concurrent validity with another measure of ASMR, the ASMR-15 (r = .41). Our results are significant in demonstrating that individuals do not need to have been exposed to online ASMR content to be assessed for ASMR responding.

Disorganized communication and social dysfunction in schizophrenia: Emerging concepts and methods

Emmanuel Olarewaju1, Lena Palaniyappan2,3,4, Guillaume Dumas5,3,6

1Integrated Program in Neuroscience, McGill University, Montreal, QC, Canada
2Douglas Mental Health University Institute, McGill University, Montreal, QC, Canada
3Department of Psychiatry, McGill University, Montreal, QC, Canada
4Robarts Research Institute, Department of Medical Biophysics, Western University, London, ON, Canada
5Centre Hospitalier Universitaire Sainte Justine Research Center, Montreal, QC, Canada
6Division of Social and Transcultural Psychiatry, McGill University, Montreal, QC, Canada

Corresponding Author: Lena Palaniyappan, email: lena.palaniyappan@mcgill.ca


Social dysfunction is a core feature of schizophrenia. This dysfunction arises from disorganized interpersonal interaction that typifies the social profile of affected individuals. Traditionally, disorganized speech and behaviour have been studied as the product of hidden cognitive processes (‘private mind’). A dysfunction in these processes was attributed to the brain afflicted by the illness (‘brain-bound mechanisms’). But this approach has contributed to challenges in measuring and quantifying disorganization. To date, the single-brain focus has not provided satisfactory clarity or led to effective treatments for persistent social dysfunction in schizophrenia. In this research, we embrace the emerging field of second-person neuroscience to address disorganization in schizophrenia. We argue that the focus of interest for disorganization is the interpersonal space where shared mental processes (‘social mind’) occur based on the bio-behavioural synchrony between two (or more) interacting people. We deploy several bio-behavioural measures that can capture the component parts of this process. In particular, we emphasize the real-time imaging technology of hyperscanning that enables multi-person analysis of naturalistic social interaction. We illustrate how these measures can be used in empirical studies by posing disorganization as a problem of interpersonal processing. We anticipate that these emerging concepts and methods will offer unique insights into the mechanisms of social dysfunction in schizophrenia, paving the way for potential advancements in diagnostic and treatment strategies with broader mental health applicability.

Complexity of inflammation, social conditions, and experienced symptoms in schizophrenia

Enzo Cipriani1,2, Philippe Kerr1,2, Cécile Le Page1,2, Robert-Paul Juster1,2

1Centre de Recherche de l'Institut Universitaire en Santé Mentale de Montréal, Montreal, QC, Canada
2Department of Psychiatry & Addiction, Université de Montréal, Montreal, QC, Canada

Corresponding Author: Robert-Paul Juster, email: robert-paul.juster@umontreal.ca


Schizophrenia (SCZ) is a severe mental disorder having a considerable impact on functional capacities, social integration, and quality of life of the concerned. It is characterized by three families types of symptoms: being positive (e.g., hallucinations), negative (e.g., apathy), and cognitive (e.g., memory loss). Several studies have identified significant differences in cytokines (CTK; indicators of inflammation and immune activity) concentrations between diagnosed individuals with SCZ and healthy controls. Significant correlations between pro-inflammatory CTK, negative, cognitive, and potentially positive symptoms were also identified. In the current study, we analyzed the concentrations of 3 pro-inflammatory CTK (Interleukin [IL] 6, Tumor Necrosis Factor[TNF]-α and C-Reactive Protein [CRP]) among psychiatric emergency patients diagnosed with SCZ (n=80) and healthy controls (n=145) from the Signature Bbiobank. We identified significantly higher concentrations of IL-6 (U=4449, p=0.001) and TNF-α (t=2.116, p=0.036) amongst our SCZ patients compared to our controls group. We also searched for the presence of examined correlations between CTK levels, psychotic symptoms (and positive, cognitive), symptoms of depression and anxiety anxious, and depressive symptoms in our SCZ group. We identified a significant correlation between CRP levels, depressive symptoms (r=0.254, p=0.044), and psychotic symptoms (r=0.345, p=0.006). These preliminary results concord with the literature and present a complex portrait of the relation associations between immune processes (through CTK concentrations), SCZ diagnostics, and experienced symptoms. Upcoming results will include a greater sample size (311 SCZ and 145 controls) and psychosocial measures to offer a greater statistical power and put results in perspective with the social conditions of the patients.

Sex and gender correlates and stereotypes towards sex differences in cognitive functioning

Fanny Saulnier1, Mina Guerin1, Louis Cartier1, Elise Isoard1, Audrey Labelle1, Nevena Chuntova1, Robert-Paul Juster1

1Department of Psychiatry and Addictology, Université de Montréal, Montréal, QC, Canada

Corresponding Author: Robert-Paul Juster, email: robert-paul.juster@umontreal.ca


According to the scientific literature, gender stereotypes influence individuals cognitive performance. Some studies have shown that men experienced decreased performance in the presence of unfavorable gender stereotypes (e.g., women are better at verbal tasks). The same was true and for women on mathematical tasks. In the absence of such stereotypes, no significant differences were observed. However, few studies measure this influence with neurocognitive tests with gender-diverse individuals. Thus, our study aims to fill this gap in the literature. Our protocol contains a gender stereotype induction component and uses a battery of neurocognitive tests. At the same time, the interrelated effects of sex hormones, sex assigned at birth, and gender-related psychosocial factors on cognition are assessed. The sample will consist of 264 adult participants distributed in subgroups according to sex assigned at birth, gender identity and sexual orientation. Eight neurocognitive tests will be administered. Before each test, a gender stereotype is stated by the experimenter. Saliva samples are collected before, during and after the experiment. Two questionnaires are administered after the neurocognitive tests. One is designed to measure psychosocial variables and the other is designed to assess participants' stereotypes of different skills. Preliminary results, with 87 sexually diverse participants, showed that estradiol was positively associated with verbal abilities and that male sex assigned at birth predicted spatial abilities. Also, research was conducted to obtain the opinions of gender diverse individuals on relevant research topics. Finally, in a previous phase, we collected data without stereotype induction from 222 participants.

Multivariate prediction of language function post-stroke: A model-based structural connectivity approach

Franziska Hildesheim1,2, Anna Zumbansen3, Anja Ophey4, Thomas Funck5, Amy Kuceyeski6, Keith Jamison6, Alexander Thiel1,7

1Lady Davis Institute for Medical Research, Jewish General Hospital, McGill University, Montréal, QC, Canada
2Elise Department of Neurology & Neurosurgery, McGill University, Montréal, QC, Canada
3Music and Health Research Institute, University of Ottawa, ON, Canada
4Department of Medical Psychology, Neuropsychology and Gender Studies, Center for Neuropsychological Diagnostics and Intervention, Medical Faculty, University Hospital Cologne, Cologne, Germany
5Forschungszentrum Juelich, Juelich, Germany
6Weill Cornell Medical College, Medical Faculty, Ithaca, United States
7Department of Neurology & Neurosurgery, McGill University, Montréal, QC, Canada

Corresponding Author: Alexander Thiel, email: alexander.thiel@mcgill.ca


Background and aims: Prognostication of post-stroke language function may be important for clinicians to develop individualized treatment plans for aphasic stroke patients. The use of connectome disruption information for the prognosis in clinical practice is limited by the need to acquire diffusion tensor imaging (DTI) data, which is usually not part of clinical imaging protocols. We present a model-based approach to quantify connectome disruption with minimal user interaction, exclusively relying on routinely collected MR-images. Methods: Language function at baseline and follow-up was assessed in 76 aphasic stroke patients, using the Boston Naming Test (BNT), Token Test (TT), and Semantic Verbal Fluency test (sVF). The Network Modification tool was used to calculate change in connectivity (ChaCo) scores as estimates of structural connectivity disruption for 13 left-hemispheric grey matter regions and right-hemispheric homologues by superimposing individual infarct masks onto a DTI tractogram reference set. Machine-learning multivariate prediction models were developed. Results: Prediction models explained 46.4%-60.4% of variance at baseline and 60.1%-81.8% at follow-up. Baseline language and cognitive function were the most important variables to predict language function at follow-up, reducing the prediction error by 35.7%-38.1% and 19.0%-42.4%, respectively. ChaCo scores explained up to 12.5% additional variance, with left superior temporal gyrus (9.9%) and right angular gyrus (12.5%) as most important network nodes. Conclusion: Connectome disruption estimates improve prediction accuracy of post-stroke language function in addition to baseline language and cognitive function. Model-based connectivity analysis using routine morphological MRI might be a useful method for clinical prognostication of the individual patient’s recovery trajectory.

Understanding the association between neural reward responsiveness and suicide attempt history among university students

Grace O. Allison1, Anna Weinberg1

1McGill University, Montreal, QC, Canada

Corresponding Author: Anna Weinberg, email: anna.weinberg@mcgill.ca


Suicide is a leading cause of death among youth and emerging adults. While many individuals endorse suicidal ideation, not all who contemplate suicide will engage in suicidal behaviors. In fact, most individuals with suicidal ideation will not attempt suicide, and research has yet to adequately characterize factors that increase suicide attempt risk. Prior work has identified differences in neural reward responsivity in at-risk and suicidal individuals. However, findings show both hyporeactivity and hyperreactivity to reward among these individuals, and it remains unclear whether alterations in neural reward response relate to suicide attempt risk specifically, independent of risk for ideation and other suicidal behaviors. The present study examined whether a suicide attempt history was associated with neural reward response among university students with and without suicidal thoughts and behavior histories. Participants (N = 153) completed a monetary reward guessing task to elicit the Reward Positivity (RewP), a neural response to reward receipt. A lifetime suicide attempt history was associated with a blunted neural reward response, even after adjusting for lifetime suicidal ideation, lifetime suicidal behaviors (e.g., preparatory behaviors), and neural response to loss. These findings suggest that neural reward response is useful in differentiating individuals with suicide attempt histories from those with ideation or other suicidal behaviors. However, future research should determine whether this pattern of neural response contributes to suicide attempt risk or occurs because of attempt histories. These findings are an initial step towards clarifying risk factors specific to suicide attempts and may aid in future intervention efforts.

Therapeutic exercise interventions in rat models of arthritis

Hannah Derue1,2,3, Alfredo Ribeiro-da-Silva1,2,3,4

1Integrated Program in Neuroscience, McGill University, Montreal, QC, Canada
2Department of Pharmacology and Therapeutics, McGill University, Montreal, QC, Canada
3Alan Edwards Centre for Research on Pain, McGill University, Montreal, QC, Canada
4Department of Anatomy and Cell Biology, McGill University, Montreal, QC, Canada

Corresponding Author: Alfredo Ribeiro-da-Silva, email: alfredo.ribeirodasilva@mcgill.ca


Arthritis is the leading cause of musculoskeletal pain and disability worldwide. Nearly 50% of individuals over the age of 65 have arthritis, which contributes to limited function, articular pain, physical inactivity, and diminished quality of life. Therapeutic exercise is often recommended in clinical settings for patients experiencing arthritic pain, however, there is little practical guidance regarding the use of therapeutic exercise to alleviate arthritic musculoskeletal pain. Rodent models of arthritis allow researchers to control for experimental variables, which cannot be done with human participants, providing an opportunity to test therapeutic approaches in preclinical models. This literature review provides a summary of published findings in therapeutic exercise interventions in rat models of arthritis as well as gaps in the existing literature. We reveal that preclinical research in this field has yet to adequately investigate the impact of experimental variables in therapeutic exercise including its modality, intensity, duration, and frequency on joint pathophysiology and pain outcomes.

Parvalbumin controls inhibitory tone in the spinal cord

Haoyi Qiu1, Lois Miraucourt1, Albena Davidova1, Reza Sharif-Naeini1

1Department of Physiology, McGill University, Montreal, QC, Canada

Corresponding Author: Reza Sharif-Naeini, email: reza.sharif@mcgill.ca


The nervous system processes sensory information by relying on the precise coordination of neural networks and their synaptic firing patterns. In the spinal cord, disturbances to the firing pattern of the tonic firing parvalbumin-expressing inhibitory interneuron (PV neurons) disrupt the ability of the dorsal horn (DH) to integrate touch information and may result in pathological phenotypes. The parvalbumin protein (PVp) is a calcium (Ca2+)-binding protein that buffers the accumulation of Ca2+ following a train of action potential to allow for tonic firing. Here, we find that peripheral nerve injury causes a decrease in PVp expression in PV neurons and transitions them from tonic to adaptive firing. We show that reducing the expression of PVp causes otherwise healthy mice to develop mechanical allodynia. This critical role for PVp extends outside of the DH and applies to other CNS PV neurons. Indeed, in both the DH and hippocampus, decreasing PVp causes the PV neurons to lose their high frequency activity. We show that this transition is mediated by Ca2+-activated potassium (SK) channels activation in PV neurons of the DH and hippocampus. Further, their tonic firing can be partially restored after nerve injury by selectively inhibiting the SK2 channels in DH PV neurons. We also reveal that a decrease in the transcriptional coactivator, PGC-1α, causes decreased PVp expression and mechanical allodynia. By preventing the decrease in PVp expression before nerve injury, we were able to protect mice from developing mechanical allodynia. Our data indicate an essential role for PVp-mediated calcium buffering in PV neuron firing activity and in mechanical allodynia after nerve injury.

Context reduces the negative impact of masks on inferring emotions

India Audet1, Sarah McCrackin1, Jelena Ristic1

1McGill University, Montreal, QC, Canada

Corresponding Author: Jelena Ristic, email: jelena.ristic@mcgill.ca


Since the onset of the COVID-19 pandemic, face masks have become a daily part of our lives and have been crucial in reducing the spread of the virus. Yet research has shown that one of their unfortunate side effects is that they impair emotion recognition. However, while emotion recognition involves recognizing emotions based on facial cues, emotional inferences are also based on emotional context. Thus, it is possible that providing emotional context could compensate for the lack of facial cues in masked situations to facilitate emotional inferences. To investigate this question, we presented participants with happy and sad faces with or without masks, and with or without accompanying emotional context. We asked participants to rate the emotional valence and intensity of the protagonists. Supporting the idea that masks impair emotional inferences, masked individuals were inferred to feel more neutral and less intense emotion than unmasked individuals. Furthermore, providing emotional context amplified ratings of valence in each emotion’s direction and heightened the perceived intensity of these emotions. Most importantly and confirming our hypotheses, providing emotional context lessened the degree of impact of masks on emotional inferences, suggesting that congruent emotional context can compensate for the lack of visual face cues. These results imply that providing more verbal information or context when interacting with masks may benefit our ability to infer the emotional states of others, which is a skill that contributes to important aspects of our social lives and well-being.

The Amyloid-beta Interacting Peptide (AIP) attenuates Aβ43-induced toxicity in transgenic Drosophila melanogaster

Isabel Sarty1, Adeola Shobo1, Mark Hancock2, Gerhard Multhaup1

1Integrated Program in Neuroscience, McGill University, Montreal, QC, Canada
2The Surface Plasmon Resonance - Mass Spectrometry, McGill University, Montreal, QC, Canada

Corresponding Author: Gerhard Multhaup, email: gerhard.multhaup@mcgill.ca


As the leading cause of dementia, Alzheimer’s disease (AD) prevalence will continue to rise with the aging global population. Although two disease modifying treatments have recently been FDA-approved for AD, cost and safety concerns have called their risk-benefit profiles into question. Thus, novel, cost-effective therapeutic strategies are urgently needed to prevent AD and its progression. Our lab previously demonstrated (in vitro and in vivo) that a small, eight D-amino acid peptide (D-AIP) attenuates the toxicity of 42-amino acid amyloid beta (Aβ42) oligomers, which are suspected to cause synaptic disfunction and neurodegeneration in AD. In the present longitudinal study using transgenic Drosophila melanogaster, we demonstrated that D-AIP can also target and attenuate toxic Aβ43 oligomers, which are present in AD and a commonly co-existing disease, cerebral amyloid angiopathy (CAA). Importantly, Aβ43 has been proposed to act as a template for Aβ aggregation and misfolding. Live confocal imaging of transgenic Drosophila treated with D-AIP showed a significant decrease in Aβ43-induced toxicity at five, 14, and 28 days old, compared to non-treated groups in both sexes. Western blot analyses of treated flies revealed a significant decrease in levels of insoluble Aβ43 protofibril/fibrils compared to non-treated flies. Consumption and biostability of D-AIP in transgenic flies was confirmed using matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) mass spectrometry, while Kaplan-Meier curves demonstrated that D-AIP had no toxic effect on survival. This study shows the utility of D-AIP to target multiple toxic Aβ species, representing a critical advance towards earlier interventions for AD, CAA, or mixed AD/CAA patients.

Modulation of interneurons in Christianson Syndrome mouse model

Mustian J.1, Mitchell M.1, Masson LC.1, Gao AYL.1, Sharif-Naeini, R.1, McKinney, RA1

1Department of Pharmacology, McGill University, Montreal, QC, Canada

Corresponding Author: Anne McKinney, email: anne.mckinney@mcgill.ca


Christianson Syndrome (CS) is an X-linked neurodevelopmental/neurodegenerative disorder. Patients have intellectual disability, autism and epilepsy. It arises from mutations in the SLC9A6 gene which normally encodes the endosomal pH regulator (Na+, K+)/H+ exchanger isoform 6 (NHE6). NHE6 maintains the accurate pH of endosomes to allow proper cargo trafficking. All mutations found in CS result in a loss-of-function for NHE6 and an overacidification of endosomes leading to the disruption of cargo essential for the development of the neuronal circuitry. Disturbances in excitation-inhibition (E/I) balance are linked to epilepsy and autism. Such disturbances could be due to an elevation of excitation or a decrease in inhibition. Using a multidisciplinary approach in murine model of CS, we found an age-dependent reduction in inhibitory cells indicated by reduced GAD67 in CS hippocampal tissue at P180. Immunohistochemistry revealed that PV+ interneurons were significantly decreased in the CA1 at P60 and CA1/CA3 at P180 in CS hippocampi. To determine if PV reduction was due to death of PV interneurons or downregulation of PV expression, we generated a PV mouse as the distinction poses opposing effects on the circuitry. We observed a reduction of PV+ interneurons at P60 in KO mice compared to WT but no changes at P180. This reduction at P60 may suggest a delay in PV+ interneurons development and a downregulation of PV expression at P180. Together these results reveal mechanisms for neurotransmission imbalance in CS and the possible role PV interneurons in the disruption.

Longitudinal inference of hippocampal volume and morphometric hippocampal-cortical connectivity in psychosis subtypes

Jana Fabienne Totzek1, M. Mallar Chakravarty1,2,3, Ridha Joober1,2, Ashok Malla1,2, Jai L. Shah1,2, Alexandra L. Young4, Martin Lepage1,2, Katie M. Lavigne1,2,5

1Department of Psychiatry, McGill University, Montreal, QC, Canada
2Douglas Mental Health University Institute, Montreal, QC, Canada
3Department of Biological and Biomedical Engineering, McGill University, Montreal, QC, Canada
4Department of Neuroimaging, Institute of Psychiatry, Psychology and Neuroscience, King’s College London, London, United Kingdom
5Montreal Neurological Institute and Hospital, McGill University, Montreal, QC, Canada

Corresponding Author: Martin Lepage , email: martin.lepage@mcgill.ca


Psychosis is characterized by reduced hippocampal volume and morphometric hippocampal-cortical connectivity. To date, it remains open how these two concepts relate to each other temporally and across the psychosis spectrum. To explore this question, we applied a recently developed machine-learning algorithm, Subtype and Stage Inference (SuStaIn), which identifies patient clusters with common disease progression patterns and thus uniquely integrates clustering and disease progression modeling. From two cross-sectional datasets of first- and multi-episode psychosis, we sampled neuroimaging data of 175 patients and 117 non-clinical controls. Then we applied SuStaIn to 18 hippocampal volumes (subfields + white matter regions), 18 measures of hippocampal-cortical connectivity, and 8 measures of volume and connectivity combined, across three analyses. We found that deteriorations in hippocampal volume can but do not necessarily precede deteriorations in connectivity. More precisely, we identified three possible patterns of volume development (no change, subfields deterioration, white matter deterioration) and four possible patterns of connectivity deterioration (no change, input region deterioration, left hemisphere deterioration, broad deterioration). The analysis of volume and connectivity combined revealed that if white matter volume deteriorates first, deterioration in one of the three patterns of connectivity necessarily follows. However, a deterioration of hippocampal subfield volumes precedes only deteriorations of left and broad connectivity patterns while also yielding the possibility of no further deteriorating connectivity. These results identify distinguishable disease progression patterns, which may characterize how the deterioration of neural biomarkers (i.e., hippocampal volume, morphometric hippocampal-cortical connectivity) influences the heterogeneous clinical manifestations of psychosis.

Investigating the cannabis withdrawal trajectory in individuals with cannabis and tobacco co-use

Zac J S Yeap1,2, Isabela Lara Uquillas1,2, Tony P George3, Romina Mizrahi2,4, Rachel Rabin1,2,4

1Integrated Program in Neuroscience, McGill University, Montreal, QC, Canada
2Douglas Mental Health University Institute, Montreal, QC, Canada
3Centre for Addiction and Mental Health and Department of Psychiatry, University of Toronto, Toronto, ON, Canada
4Department of Psychiatry, McGill University, Montreal, QC, Canada

Corresponding Author: Rachel Rabin, email: rachel.rabin@mcgill.ca


Background: Cannabis withdrawal follows a distinct trajectory and strongly predicts of relapse. Symptoms begin 24 hours after cessation, peak at 7-days, and dissipate 28-days later. Cannabis withdrawal has been linked to reduced endocannabinoid availability. Thus, normalizing endocannabinoid levels may reduce withdrawal. A sizeable proportion of individuals who use cannabis co-use tobacco (CT+). We and others report that CT+s experience greater cannabis withdrawal severity compared to individuals who only use cannabis (CT-). Because these studies were cross-sectional, it is unknown whether cannabis withdrawal remains elevated in severity for a protracted period of time. Aim 1: To compare cannabis withdrawal severity between CT+ and CT- following 28-days of cannabis abstinence Aim 2: To test associations between endocannabinoids and cannabis withdrawal severity and tobacco use. Methods: We recruited n=10 CT+ (>5 cigarettes/day) and n=9 CT- (≤5 cigarettes/day) to undergo 28-days of cannabis abstinence. Cannabis withdrawal was assessed weekly (Marijuana Withdrawal Checklist, MWC). In an independent sample of men with daily cannabis use (N=8), correlations between cannabis withdrawal, serum-derived anandamide and 2-AG levels, and tobacco consumption were assessed. Results: Cannabis withdrawal symptoms in CT+ remained elevated in severity compared to CT- (group x time interaction effect, F(4,68)=2.77 p=0.03). Correlations between cannabis withdrawal and endocannabinoid levels revealed negative associations between cannabis craving and anandamide (r=-0.78; p=0.02) and 2-AG (r=-0.84; p=0.01). Tobacco consumption also negatively correlated with anandamide levels (r=-0.71; p<0.05). Conclusions: Co-use leads to more severe and protracted withdrawal symptoms which may reflect aberrant endocannabinoid activity. Findings may help with developing interventions for co-use.

Contrast detection in spectrally pink and white spatial noise in human vision

Jeong Ung Song1, Alex S. Baldwin1

1Department of Ophthalmology & Visual Sciences, McGill University, Montreal, QC, Canada

Corresponding Author: Alex S Baldwin, email: alexander.baldwin@mcgill.ca


Our ability to detect low-contrast targets is limited by an internal noise in the visual system. This noise can be measured indirectly using the equivalent noise method. This measures the amount of external noise that must be applied to a stimulus to be equivalent to the internal noise in the visual system. This gives a measure of both the effective “noisiness” of input to visual processing, and the efficiency with which that input is processed. Pink noise, with a power that declines as a function of spatial frequency, has a spectrum resembling that occurring within natural scenes. Most noise masking studies use white noise however, which has a flat spectrum. This study explores how these two types of noise impact detection. We measured contrast detection thresholds for different grating targets (sin, square, and missing-fundamental square) in the two noise types. Human performance can be compared against models that use different strategies. The simplest model uses the expected target as a linear template that is matched to the noisy stimulus. Compared to this model, humans show similar efficiency for detecting all three targets in pink noise. In white noise however there is a marked inefficiency for detection of the missing fundamental target. The performance of this simple model is compared against that of models featuring spatial-frequency-tuned channels (that are thought to exist in human vision). The methodological implications of the results will inform the development of parallel work, where we are studying noise in individuals suffering from visual snow syndrome.

Neuromelanin-sensitive MRI as a measure of dopamine turnover in first episode psychosis

Jessica Ahrens1,2, Sabrina Ford3, Betsy Schaefer3, David Reese3, Ali Khan3,4, Clifford Cassidy5, Lena Palaniyappan1,2

1McGill University, Montreal, QC, Canada
2Douglas Mental Health Research Institute, Montreal, QC, Canada
3Robarts Research Institute, London, ON, Canada
4University of Western Ontario, London, ON, Canada
5The Royal's Institute of Mental Health Research, Ottawa, ON, Canada

Corresponding Author: Lena Palaniyappan, email: lena.palaniyappan@mcgill.ca


Background: The dopamine hypothesis of schizophrenia is based upon the discovery of antipsychotics and states that hyperactive dopamine causes schizophrenia symptoms. Advances in magnetic resonance imaging (MRI) have resulted in neuromelanin-sensitive MRI (NM-MRI), a proposed tool to noninvasively measure dopamine turnover. Recent research shows that the NM-MRI signal correlates to psychosis severity in those with schizophrenia; however, it is unknown if these results extend to those with first episode psychosis, or how neuromelanin changes over time in schizophrenia. Objectives and Hypotheses: This study aims to: (1) investigate if prior research results on excess neuromelanin in established schizophrenia extends to a first episode sample; and (2) determine the change in neuromelanin after a year in those with schizophrenia versus controls. The following hypotheses are made: (1) first episode patients will have higher neuromelanin than controls; and (2) neuromelanin will increase at a faster rate in those with schizophrenia. Methods: Participants will be recruited from the Early Psychosis Program in London, Ontario. Participants will undergo a task-free functional MRI scan with a neuromelanin sequence. Results: Not available yet. So far, the quantification of neuromelanin has been extremely successful based on quality measures and Contrast-to-Noise estimates. Significance: To date, no studies have studied neuromelanin in those with first episode psychosis or assessed how neuromelanin changes over time in those with schizophrenia. Current global efforts focused on neuromelanin have highlighted the value of using this as a marker in upcoming clinical trials, which can significantly enhance therapeutic options for patients.

Suicidality and persistent negative symptoms following a first episode of psychosis

Joseph Ghanem1, Massimiliano Orri1,2, Laura Moro2,3, Katie M. Lavigne1,2, Delphine Raucher-Chéné1,2, Ashok Malla1,2, Ridha Joober1,2, Martin Lepage1,2

1McGill University, Montreal, QC, Canada
2Douglas Mental Health University Institute, Montreal, QC, Canada
3Universite de Montreal, Montreal, QC, Canada

Corresponding Author: Martin Lepage, email: martin.lepage@mcgill.ca


Suicide is a leading cause of death in first-episode psychosis (FEP), with an elevated risk in the first year following illness onset. Previous research suggests that negative symptoms are protective against suicidality, supported by evidence from deficit schizophrenia, a disease subtype with elevated negative symptoms and lower suicidality, possibly due to reduced insight and diminished interest in social engagement. The Persistent Negative Symptoms (PNS) subtype, characterized by idiopathic negative symptoms, is similar to deficit schizophrenia and may represent a group with lower suicidality. By contrast, secondary PNS (sPNS) that result from depressive, positive, or extrapyramidal symptoms may characterize a heterogeneous subtype with potent suicidality due to a greater symptom burden. In a large cohort of FEP patients from an early intervention program, we investigated suicidality in PNS, sPNS, and non-PNS. Patients were categorized into their groups at month 12, and suicidality was evaluated using the Brief Psychiatric Rating Scale. Covarying for age and sex, we found that sPNS had higher suicidality than PNS and non-PNS, PNS had lower suicidality than sPNS, and PNS and non-PNS did not differ. The difference between sPNS and PNS remained after controlling for depression, but sPNS and non-PNS no longer differed. Consistent with the literature, we observed that PNS had low suicidality but did not significantly differ from non-PNS. However, we identified sPNS as a group with elevated suicidality beyond depression, suggesting that sPNS would benefit from targeted intervention and that PNS categorization identifies a subgroup for whom negative symptoms are not protective against suicidality.

Temporal lobe epilepsy gray and white matter alterations across the lifespan: an ENIGMA study

Judy Chen1, Alexander Ngo1, Sara Lariviere1, Raul Rodriguez-Cruces1, Reinder Vos de Wael1, Paul M. Thompson2, Sophia I. Thomopoulos2, Sanjay M. Sisodiya3, Carrie McDonald4, Lorenzo Caciagli5, Andrea Bernasconi1, Neda Bernasconi1, Boris Bernhardt1

1Department of Neuroscience, McGill University, Montreal, QC, Canada
2University of Southern California, Los Angeles, CA, United States
3University College London, London, United Kingdom
4University of California San Diego, San Diego, CA, United States
5University of Pennsylvania, Philadelphia, PA, United States

Corresponding Author: Boris Berrnhardt, email: boris.bernhardt@mcgill.ca


Magnetic resonance imaging (MRI) analysis can measure brain atrophy in temporal lobe epilepsy (TLE). Research has mainly focused on linear effects of aging. We capitalized on a novel age-window approach to probe the association between ageing and gray and white matter alterations in temporal lobe epilepsy. Participants. As part of ENIGMA-Epilepsy, we analyzed T1w- and diffusion- MRI data in 885 healthy individuals (378 male; mean±SD age: 36.0±11.7 years) and 769 TLE (314 males; mean±SD: 38.2±11.1 years; 430 left-sided focus). Sliding age-window analysis. Using a window range of ± 2 years from the age of interest, yielding a total of five unique ages, the mean values for CT and SV regions, as well as FA and MD across all white matter tracts were calculated for each age. These mean values were then multiplied by normally distributed weights to yield a weighted average for each brain region. This was repeated for every age of interest sliding across from 19-71 years of age in TLE patients. The sliding age-window analysis revealed a distinct, non-linear pattern with an accelerated decline with aging in patients across cortical and subcortical gray matter region. White matter tracts also demonstrated similar pattern decreases across the lifespan. The nonlinear progression in common epilepsies through adulthood highlights the need for further distinction between younger and older patients. Follow up analyses with larger sample in older patients as well as longitudinal investigations can help delineate age-related effects in epilepsy.

The mGlu2 positive allosteric modulator AZD8529 alleviates L-DOPA-induced dyskinesia and psychosis-like behaviours in the Parkinsonian marmoset

Judy Shaqfah5, Imane Frouni1,2, Cynthia Kwan1, Dominique Bédard1, Stephen G. Nuara4, Jim Gourdon4, Adjia Hamadjida1, Philippe Huot1,2,5,6

1Montreal Neurological Institute-Hospital, Montreal, QC, Canada
2Département de Pharmacologie et Physiologie, Faculté de Médecine, Université de Montréal, Montreal, QC, Canada
3Department of Neurology and Neurosurgery, McGill University, Montreal, QC, Canada
4Comparative Medicine and Animal Resource Centre, McGill University, Montreal, QC, Canada
5Department of Neurology and Neurosurgery, McGill University, Montreal, QC, Canada
6Division of Neurology, Department of Neurosciences, McGill University, Montreal, QC, Canada

Corresponding Author: Philippe Huot, email: philippe.huot@mcgill.ca


AZD8529 is a selective metabotropic glutamate 2 (mGlu2) receptor positive allosteric modulator (PAM) that has undergone clinical trials for schizophrenia and smoking cessation. Previously, we demonstrated that the selective mGlu2 PAMs LY-487,379 and CBiPES alleviated L-3,4-dihydroxyphenylalanine (L-DOPA)-induced dyskinesia and psychosis-like behaviours (PLBs) in the 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP)-lesioned marmoset model of Parkinson's disease (PD). However, neither drug has clinically relevant pharmacological properties, contrary to AZD8529, which could be repurposed if efficacious. To assess the effect of AZD8529 on L-DOPA-induced dyskinesia and PLBs in the MPTP-lesioned marmoset, parkinsonism was induced by MPTP injections followed by oral administration of L-DOPA/benserazide (L-DOPA/vehicle) to elicit dyskinesia and PLBs. On experimental days, marmosets were injected with L-DOPA (15/3.75 mg/kg s.c.) along with vehicle or AZD8529 (0.1, 0.3, 1, and 10 mg/kg s.c.). After treatment, marmosets were recorded for 6h and this footage was analysed for dyskinesia, PLBs, and parkinsonism. The results showed a reduction in global dyskinesia severity (up to 70%, P<0.001), and on-time with disabling dyskinesia (up to 97%, P<0.001) when compared to L-DOPA/vehicle. Similarly, there was a reduction in global PLB severity (up to 64%, P<0.001), and on-time with disabling PLBs (up to 94%, P<0.001) when compared to L-DOPA/vehicle. Additionally, AZD8529 increased the duration of the anti-parkinsonian action of L-DOPA at doses of 0.3 mg/kg and above (up to 29%, P<0.05). Our results demonstrate the potential of AZD8529 and mGlu2 positive allosteric modulation for alleviating L-DOPA-induced dyskinesia and PLBs and amplifying L-DOPA anti-parkinsonian effects.

Neurodevelopmental risks of advanced paternal age through placental DNA methylation

Julia Barnwell1,2, Sophia Rahimi1,3, Sherri Lee Jones4,5, Josée Martel1, Donovan Chan1, Tuong-Vi Nguyen1,5,6,7, Tina Montreuil1,5,8,9,10, and Jacquetta Trasler1,2,8,11

1Research Institute of the McGill University Health Centre, Montreal, QC, Canada
2Department of Human Genetics, McGill University, Montreal, QC, Canada
3Department of Medicine, University of Toronto, Toronto, ON, Canada
4Douglas Research Centre, Montreal, QC, Canada
5Department of Psychiatry, McGill University, Montreal, QC, Canada
6Lady Davis Institute for Medical Research, Jewish General Hospital, Montreal, QC, Canada
7Reproductive Psychiatry Program, McGill University Health Centre, Departments of Psychiatry and Obstetrics and Gynecology, Montreal, QC, Canada
8Department of Pediatrics, McGill University, Montreal, QC, Canada
9Department of Educational and Counselling Psychology, McGill University, Montreal, QC, Canada
10Centre Hospitalier Universitaire Sainte Justine Research Center, Montreal, QC, Canada
11Department of Pharmacology and Therapeutics, McGill University, Montreal, QC, Canada

Corresponding Author: Jacquetta Trasler, email: jacquetta.trasler@mcgill.ca


The influence of preconception paternal factors such as older age on child susceptibility to neurodevelopmental disorders, including autism and schizophrenia, is poorly understood. Advanced paternal age is associated with epigenetic modifications like DNA methylation (DNAm) in sperm that can be transmitted to the feto-placental unit at conception. The placenta is a known sensitive indicator of DNAm changes that have been linked to neurodevelopmental disorders. Paternal-age-induced DNAm alterations could thus have long-term impacts on brain and behavioral development that may manifest during adrenarche, the first stage of puberty and a critical period for cortico-limbic development (occurring ~age 6-8). As children face increased social and cognitive demands in elementary school, this age is optimal to assess the effect of perinatal protective and risk factors on child neurodevelopment. Data from the 3D (Design, Develop, Discover) prospective birth cohort, including parental age and placental DNAm (Illumina 850K array, n=64) are used, along with follow-up structural MRI and behavioral outcome measures from children aged 6-8. Preliminary analyses have identified advanced paternal age to be mainly associated with decreased DNAm in the placenta, affecting genes linked to neurodevelopment (FOXG1, MYT1L, CNTN4). We will next investigate potential associations with structural differences in the amygdala, a brain region implicated in emotional regulation, and with socio-emotional behaviors in the children, measured with the Strengths and Difficulties Questionnaire. Mediation analyses will probe for causal mechanisms whereby paternal age affects child amygdala volume or socio-emotional outcomes through placental DNAm. These data are important for understanding how paternal preconception factors biologically influence child neurodevelopment.

Lateral head impact traumatic brain injury mouse model causes hyponatremia and activates supraoptic nucleus

Julie O’Reilly1, Nicholas J. Simpson1, Zahra S. Thirouin1, Paolo A. Bastone1, Charles W. Bourque1

1McGill University

Corresponding Author: Charles Bourque, email: charles.bourque@mcgill.ca


Background: Hyponatremia is commonly reported in the days following a traumatic brain injury (TBI). Because the brain is highly vulnerable to osmotic swelling, hyponatremia following TBI may result in additional cognitive deficits. Hyponatremia is most frequently due to inappropriate secretion of vasopressin, which is secreted by the hypothalamic supraoptic nucleus (SON) and promotes water reabsorption by the kidney. The long-lasting hyponatremia observed after TBI suggests prolonged activation of SON neurons, which may undergo plastic changes after the impact. However, the involvement of the SON in TBI-related hyponatremia has not been investigated yet. Methods: 2-4 months old C57Bl/6 male mice were lightly anesthetized with isoflurane and subjected to a lateral head injury using a Gothenburg Impactor (Collision Analysis Inc). This instrument was used to deliver a reproducible, calibrated blow to the side of the head protected by a helmet via a 50 g projectile launched at 9 m/s. Mice treated the same way but without the impact served as controls (shams). The time courses of serum natremia and c-Fos protein expression in the SON after TBI/sham were performed. Results: TBI mice showed significantly lower serum natremia compared to sham at 3h, 6h and 12h after TBI/sham, but not at 24h or 48h. 3h after TBI/sham, our analysis revealed a significantly higher c-Fos nucleus density in the SON of TBI mice compared to sham. This difference was not present 6h after TBI/sham. Conclusion: The SON may be involved in the development of hyponatremia after TBI.

White matter abnormalities suggestive of multiple sclerosis in Wolfram syndrome: report of two unrelated cases

Justin Simo1,2, Catherine Ashton1,2,3, Guy Rouleau1,2,4, Yann Nadjar5, Roberta La Piana1,2,6

1Montreal Neurological Institute-Hospital, McGill University, Montreal, QC, Canada
2Department of Neurology and Neurosurgery, McGill University, Montreal, QC, Canada
3Department of Neurology, Royal Perth Hospital, Perth, WA, Australia
4Department of Human Genetics, McGill University, Montreal, QC, Canada
5Department of Neurology, Pitié-Salpêtrière University Hospital, Paris, France
6Department of Diagnostic Radiology, McGill University, Montreal, QC, Canada

Corresponding Author: Roberta La Piana, email: roberta.lapiana@mcgill.ca


Background: Wolfram syndrome (WFS) is a genetic spectrum disorder clinically characterized by optic atrophy (OA), diabetes mellitus, sensorineural deafness, and diabetes insipidus. WFS typically manifests in the first two decades and progresses into adulthood, often with important neurological and/or psychiatric complications. It is caused by mutations in WFS1 (mono- or biallelic) or CISD2 (biallelic) genes. Neuroradiological features include cerebellar and/or brainstem atrophy with visual pathway and white matter involvement. We report two subjects with WFS in which multifocal, progressive, and contrast-enhancing white matter abnormalities (WMA) led to the consideration of multiple sclerosis (MS). Methods: We retrospectively analyzed the clinical, genetic, and radiological data from two unrelated subjects with genetically confirmed WFS and multifocal WMA. Results: Subject I: 43-year-old woman, heterozygous for a known WFS1 variant, had a history of congenital deafness and OA. Brain MRI documented progressive (over a seven-year period) multifocal WMA including pericallosal and periventricular lesions. Subject II: 28-year-old woman, compound heterozygous for two WFS1 variants, was known for OA and diabetes mellitus. Brain MRI revealed multifocal periventricular, callosal, subcortical, and juxtacortical WMA, with some enhancing after gadolinium injection. Spinal MRI uncovered a C3-level demyelinating spinal cord lesion. Conclusions: We expand the WFS spectrum of white matter involvement to include progressive, seemingly inflammatory lesions. Although we cannot exclude a dual diagnosis, the roles of WFS1 and CISD2 in myelination suggest a selective white matter vulnerability in WFS. Our report champions the use of reverse phenotyping (through international collaboration) to discern MS-like radiological patterns in rare white matter disorders.

The effect of tau hyperphosphorylation on organelle transport

Karthikeyan Swaminathan1, Adam Hendricks1,2

1Integrated Program in Neuroscience, McGill University, Montreal, QC, Canada
2Department of Bioengineering, McGill University, Montreal, QC, Canada

Corresponding Author: Adam Hendricks, email: adam.hendricks@mcgill.ca


Tau is a microtubule-associated protein that bundles and stabilizes microtubules in neuronal axons and regulates intracellular transport by kinesins and dyneins. We seek to understand how aberrant phosphorylation of tau alters intracellular transport and its contribution to the pathogenesis of Alzheimer’s Disease and other tauopathies. Recently, we discovered that tau differentially regulates motility of specific cargoes based on the phosphorylation status of tau and the type of motors that carry them. Using TIRF microscopy, we examined how tau hyperphosphorylation effects the transport of lysosomes and early endosomes. We exogenously expressed phosphomimetic tau (E14; mimicking tau hyperphosphorylation by GSK-3β kinase), phosphoresistive tau (AP14) and wild-type tau (WT) in COS-7 cells and tau-deficient iPSC-derived neurons. At expression levels similar to endogenous neuronal expression, E14 tau in COS-7 localized primarily in the cytoplasm in contrast to AP14 tau which enriched strongly on microtubules. All tau constructs reduced the processivity of early endosomes and lysosomes. AP14 tau strongly inhibited the processivity of lysosomes and early endosomes compared to WT and E14 tau, due to its tight association with microtubules. Lysosomes and early endosomes motility were inhibited weakly by E14 tau compared to WT tau due to the weak affinity of E14 tau to microtubules. Our preliminary results suggest that tau phosphorylation may initially be a neuroprotective response to enhance axonal transport.

CONP Experiments: An open-source neuroimaging experiment-sharing platform

Katie M. Lavigne1,2,4, Joshua Unrau1, Caroline Dakoure1, Patrick Bermudez2, Samir Das2, Emmet O’Brien2, Cecile Madjar2, Tristan Kuehn3, Martin Lepage1,5, Alan C. Evans2,4

1Douglas Research Centre, Montreal, QC, Canada
2McGill Centre for Integrative Neuroscience, Montreal, QC, Canada
3Western University, London, ON, Canada
4Montreal Neurological Institute-Hospital, Department of Psychiatry, McGill University, Montreal, QC, Canada
5Department of Psychiatry, McGill University, Montreal, QC, Canada

Corresponding Author: Martin Lepage, email: martin.lepage@mcgill.ca


Harmonizing experimental measures or tasks is a vital, but often overlooked, element of replicable science. Particularly in neuroscience, minor differences in stimulus presentation can affect the brain mechanisms and activity recruited by the experiment. The few existing experiment-sharing platforms tend to be restricted to specific software packages and are not optimized for neuroscience research. CONP Experiments is an open-source experiment-sharing platform for multimodal, software-independent, neuroimaging tasks contained within the Canadian Open Neuroscience Platform (CONP) Portal. To search for experiments, researchers can use keywords, filters, and other search capabilities to find experiments based on standardized metadata including neuroimaging modality (e.g., fMRI, EEG, MEG), operating software (e.g., E-Prime, NBS Presentation, PsychoPy, Matlab), and assessed behaviour (e.g., cognitive, sensory, motor). The platform’s universality is possible because it does not involve running experiments and because modality-specific tasks differ in terms of task parameters contained within the code itself. Contributing new experiments to the platform is simple through a standardized web form and automated retrieval via tags from open repositories (e.g., OSF, Zenodo). CONP Experiments fulfills an acute need in neuroscience for an open, searchable, platform- and domain-agnostic repository of neuroimaging experiment materials and will facilitate large-scale functional neuroimaging initiatives, which have to date been hindered by use of independent and inconsistent tasks. It is the first platform dedicated to the open sharing of a broad range of functional neuroimaging experiments with modality-, software-, and function-specific search.

Effect of Tai Chi on stress, anxiety, and self-perceived happiness: A longitudinal intervention study

Laís Almeida1,2, Gláucia Maria Bovi Ambrosano1, Guilherme Bovi Ambrosano1, Anderson Taíra3, Rosana de Fátima Possobon1, Marcelo de Castro Meneghim1, Harmehr Sekhon2, Soham Rej2, Karine Laura Cortellazzi1

1Department of Health Sciences and Children's Dentistry, Piracicaba Dental School, State University of Campinas, SP, Brazil
2Department of Psychiatry, McGill University, Montreal, QC, Canada
3Department of Genetics, Luiz de Queiroz College of Agriculture, University of São Paulo, SP, Brazil

Corresponding Author: Karine Laura Cortellazzi and Sojam Rej, email: karinecortellazzi@gmail.com; soham.rej@mcgill.ca


This intervention follow-up study evaluated anxiety and stress levels and self-perceived happiness of individuals linked to the health field who did not practice Tai Chi and compared these variables before and after practicing this art. One hundred twenty-three participants were subdivided into two groups: G1) Older health professionals (n=41) and G2) Younger pre-university students interested in the health field (n=82). Participants received a 60-minute Tai Chi class. The Perceived Stress Scale (PSS-14) was applied before the practice, and the State-Trait Anxiety Inventory (STAI) and Andrews Face Scale were used before and after the class. Generalized linear models and the Bowker symmetry test were applied for data analysis. 46.3% and 30.5% of participants in G1 and G2 had mild stress levels, and 43.9% and 46.3% had moderate stress levels, respectively. After practicing Tai Chi, both groups presented significantly lower mean anxiety scores. In G1, 31.7% of participants went from moderate to mild anxiety levels after practice, 19.5% from severe to moderate, and 7.3% from severe to mild. In G2, 28.2% of participants went from moderate to mild anxiety levels after practice, 18.3% from severe to moderate, and 6.1% from severe to mild. Furthermore, 100.0% of individuals in G1 and 96.3% in G2 were happy after the class. In this sample, Tai Chi was effective in reducing anxiety levels and improving self-perceived happiness.

The influence of exercise habits on acute stress responses in Parkinson’s patients and healthy older adults

Lara Telli1, Lyla Hawari1, Mario Bogdanov1, Madeleine Sharp1, Nasri Balit1, Nil Mehrpouya1

1Integrated Program in Neuroscience, McGill University, Montreal, QC, Canada

Corresponding Author: Madeleine Sharp, email: madeleine.sharp@mcgill.ca


Parkinson’s patients have reported using exercise to reduce stress but whether exercise influences the physiologic stress response in Parkinson’s disease is unknown. The goal of this study was to determine whether higher levels of self-reported physical and mental exercise are associated with lower susceptibility to acute stress. Fifty patients and fifty older adults underwent an in-lab acute stress manipulation. Physiological measures of the acute stress responses included pre- and post-stress salivary cortisol and systolic blood pressure measurements. Self-reported affect was also measured pre- and post-stress. Participants provided information about the duration and frequency of physical exercise and about the frequency of mental exercise, which were used to derive a composite for overall exercise levels. Planned analyses will examine the association between exercise level and physiologic and affective susceptibility to stress. Our results could suggest possible methods of alleviating the effects of stress in Parkinson’s patients and older adults.

Prior head injury influences cognitive outcomes and white matter microstructure in older adults at risk of Alzheimer's Disease

Larissa Chiu1, Meishan Ai2, Jennifer Tremblay-Mercier3, Sylvia Villeneuve3

1Department of Neurology and Neurosurgery, McGill University, Montreal, QC, Canada
2Northeastern University, Boston, MA, United States
3Douglas Mental Health Institute Research Centre, Montreal, QC, Canada

Corresponding Author: Maiya Geddes, email: maiya.geddes@mcgill.ca


Background: The long-term combined effects of mild traumatic brain injury (mTBI) and APOE ε4-allele carrier status (APOE+/-), a genetic risk factor of Alzheimer’s disease (AD), on cognition and white matter structural integrity in older adults at-risk of AD is not fully understood. Method: To investigate mTBI effects on late-life neurobehavioural outcomes, a subsample of the PREVENT-AD cohort was analyzed. Participants were cognitively normal adults over age 55 with a parental or multiple-sibling history of AD. The mTBI group included individuals who self-reported at least one prior mTBI with loss of consciousness and/or memory gap (n=48, 33.3% APOE4+). Rey Auditory Verbal Learning Test measured immediate and delayed memory, Color-Word Interference Test measured cognitive flexibility and inhibitory control, and Trail Making Test measured processing speed and attention. The control group included demographically-matched (age, sex, years of education) participants from the same cohort without self-reported mTBI (n=62, 29.0% APOE4+). Data were analyzed using 2-way ANOVAs with weighted means on years of education, corrected for multiple comparisons. To evaluate underlying structural differences in white matter tracts between mTBI and control groups, diffusion tensor imaging (DTI) metrics, including fractional anisotropy, axial and radial diffusivity, using the JHU-atlas parcellation. Results: mTBI/APOE4+ individuals performed significantly worse on cognitive outcomes of delayed recall, cognitive flexibility, and inhibitory control compared to controls; however, they performed significantly better on tests of processing speed and visual attention compared to mTBI/APOE4- individuals. DTI results showed that the mTBI group, compared to the control group, had significantly decreased axial diffusivity in the left internal capsule, right superior longitudinal fasciculus, and superior frontooccipital fasciculus (SFOF) (p<0.05 uncorrected). Within the mTBI group, significant positive correlation between lower processing speed and decreased axial diffusivity of SFOF, accounting for age and sex. Conclusion: Our findings suggest that prior mTBI impacts cognitive aging in older adults at-risk of AD differentially depend on APOE4 status, where APOE4+ individuals with prior mTBI had worse working memory, cognitive flexibility, and inhibitory control but better processing speed outcomes. Lower cognitive outcomes in mTBI individuals may be supported by decreased diffusivity in specific white matter tracts. These results can inform dementia prevention in these individuals.

Mlp60a as a potential cofactor for Spire in the development of dendritic arbors

Lauren Cinq-Mars2, Claire Marquilly1,2, Dierdre Hatton3, Donald J. van Meyel1,3

1Department of Neurology and Neurosurgery, McGill University, Montreal, QC, Canada
2Integrated Program in Neuroscience, McGill University, Montreal, QC, Canada
3Research Institute of the McGill University Health Centre, Montreal, QC, Canada

Corresponding Author: Donald van Meyel, email: don.vanmeyel@mcgill.ca


Dendrites develop highly complex arbors which are critical for the maintenance and function of neural circuits. Dendrite branching is controlled through the dynamic regulation of actin filaments (F-actin). In Drosophila class 4 dendritic arborization (cIVda) neurons, our lab previously demonstrated that the F-actin nucleation protein Spire is important for the formation of new dendrite branches. However, Spire overexpression alone is not sufficient for this process, and so we are investigating possible interactors with Spire for dendrite branching. A screen from our lab has identified muscle LIM protein 60a (Mlp60a,) with the potential to antagonize Spire, since increased branching upon Mlp60a knockdown is suppressed by the partial loss of Spire. Therefore, we are testing further genetic interactions between Spire and Mlp60a, analyzing the number and distribution of cIVda neuron dendrites. We are also investigating the subcellular compartments where Spire can be found in dendrites, aiming to pinpoint the cellular mechanism(s) by which Spire acts to initiate the formation of dendrites. An improved understanding of the mechanism and function of Spire in developing dendrites could provide improved comprehension of neurodegenerative diseases where the disruption of dendritic arbors is observed.

Glutamatergic neurometabolites in Risperidone-responsive and nonresponsive Schizophrenia

Lejia Fan1, Liangbing Liang2, Liu Yuan3, Lijun Ouyang3, Xiaogang Chen3

1Department of Psychiatry, McGill University, Montreal, QC, Canada
2Robarts Research Institute, Western University, London, ON, Canada
3Xiangya Second Hospital of Central South University, Changsha, Hunan, China

Corresponding Author: Lena Palaniyappan , email: lena.palaniyappan@mcgill.ca


Background: Schizophrenia is a debilitating psychiatric disorder. Approximately one-third of patients do not respond to antipsychotic medications. Evidence suggests that treatment-resistant schizophrenia, characterized by lack of response to dopamine-blocking antipsychotics, may not be due to subcortical dopaminergic abnormalities but relate to cortical glutamate. However, the glutamatergic basis of antipsychotic response in first episode schizophrenia remains poorly understood. We compare glutamatergic neurometabolites in risperidone responsive and non-responsive schizophrenia. Methods: We recruited 41 drug-naïve patients with first episode schizophrenia (FEP) and 41 healthy controls (HC). All patients received 12 weeks of risperidone treatment at a dosage of 4–6 mg/day, and psychotic symptoms were assessed using the Positive and Negative Syndrome Scale (PANSS) at baseline and follow-up. Patients were divided into responders (N=27) and non-responders (N=14) based on whether the percent PANSS total score reduction was above or below 50%. All subjects underwent proton magnetic resonance spectroscopy of glutamatergic metabolites at 3-Tesla (Glx and Glu) in the anterior cingulate cortex (ACC) at a dorsal (Salience Network) and a ventral (DMN) site. We assessed the effect of group and conducted logistic regression to assess if baseline glutamate level predicted later treatment response. Results: Non-responders had higher Glu in ACC (F(2,79)=3.20, p=0.041; mean(SD) in non-responders=19.15(5.80); responders=15.53(3.71); HC=16.68(4.15)) and Glx in DMN (F(2,79)=3.34, p=0.046; mean(SD) in non-responders=17.98(3.03); responder=15.61(3.27); HC=16.54(2.36)) compared to responders. Responders who had lower Glx in DMN had larger changes in PANSS negative (r=-0.416, p=0.031). Discussion: Higher glutamatergic level is associated with an unfavorable prognosis in FEP. This could be attributed to excitotoxicity or disinhibitory circuit level dysfunction.

The suppressive basis of ocular dominance changes induced by short-term monocular deprivation in normal and amblyopic individuals

Ling Gong1,2, Alexandre Reynaud2, Robert F. Hess2, Jiawei Zhou1,2

1State Key Laboratory of Ophthalmology, Optometry and Visual Science, Eye Hospital, Wenzhou Medical University, Ouhai District, Wenzhou, China
2McGill Vision Research, Department of Ophthalmology and Visual Sciences, McGill University, Montreal, QC, Canada

Corresponding Author: Alexandre Reynaud, email: alexandre.reynaud@mcgill.ca


The standard of care for amblyopia, a neurodevelopmental disorder with monocular and binocular visual function deficits, is to deprive the fellow good eye for 3-6 hours a day, this has been in operation for the past 200 years. However, this approach is not beneficial for binocular vision even if it does improve the visual acuity of the amblyopic eye in some children. To understand the neural basis of the neuroplastic effect of monocular deprivation, we investigated the change of suppressive interocular interactions and monocular contrast threshold. Nine adults with anisometropic or mixed amblyopia and 10 control adults participated in our study. Suppressive interaction and monocular contrast sensitivity, which were measured by dichoptic masking under masking and monocular conditions respectively, were measured before and after 2 hours monocular deprivation. We demonstrate that the ocular dominance changes that induced by short-term monocular deprivation, namely the strengthening of the deprived eyes contribution, are associated with asymmetric changes in suppressive interaction where the suppression from the nondeprived eye is reduced. On the other hand, the changes in contrast sensitivity are symmetric between the eyes with the previous deprived eye’s sensitivity being increased and the nondeprived eye’s sensitivity being reduced. Similar results occur in normals and amblyopes. This provides a better understanding of how inverse patching (patching of the amblyopic eye) could, by reducing the suppressive drive from the normally sighted (nondeprived) eye and increasing the monocular sensitivity of the amblyopic eye, form the basis of a new treatment for the binocular deficit in amblyopia.

Sex and gender correlates in cognitive functioning: A psychoneuroendocrine approach

Louis Cartier1,2, Fanny Saulnier1,2, Mina Guérin1, Nevena Chuntova1, Elise Isoard1, Audrey Labelle1, Robert-Paul Juster1,2

1University of Montreal, Montreal, QC, Canada
2Centre de Recherche de l'Institut Universitaire en Santé Mentale de Montréal, Montreal, QC, Canada

Corresponding Author: Robert-Paul Juster, email: robert-paul.juster@umontreal.ca


Several studies have highlighted gender differences in cognitive abilities. These differences are the basis of the concept of sexually polymorphic cognition (SPC) and result from the interaction between biological (birth-assigned sex (BAS), sex hormones) and psychosocial (gender identity, gender roles, sexual orientation) factors. However, the literature remains quite mixed regarding the magnitude of the effects of these variables. This project used a battery of cognitive tests designed to assess the influence of sex hormones on cognitive performance. At the same time, we aimed to assess the interrelated and respective effects that BAS, sex hormones, and gender-related psychosocial factors have on cognition. The sample consisted of 222 adults. They completed eight cognitive tasks that assessed a variety of cognitive domains during a 150-minute session. Subgroups have been formed based on BAS, gender identity and sexual orientation. Saliva samples have been collected before, during, and after the test, measuring testosterone, estradiol, progesterone, cortisol, and DHEA. Psychosocial variables have been measured by self-report questionnaires. The test battery presented reflects gender differences that are partially consistent with the literature. Interestingly, biological factors seem to explain differences in male-typed cognitive tasks, while psychosocial factors seem to explain differences in female-typed cognitive tasks. The results of this study will provide a solid foundation for understanding the importance of other types of factors that may alter SPC, such as gender stereotypes.

The effect of acute stress on cognition in Parkinson’s disease patients and older adults

Lyla Hawari1, Mario Bogdanov1, Lara Ekin Telli1, Nasri Balit1, Nil Mehrpouya1, Madeleine Sharp1

1Integrated Program in Neuroscience, McGill University, Montreal, QC, Canada

Corresponding Author: Madeleine Sharp, email: madeleine.sharp@mcgill.ca


It is a widely observed phenomenon that patients with Parkinson’s disease are very sensitive to stress. Recent evidence from a large cohort study also suggests that patients experience higher stress than older adults without Parkinson’s disease. However, it remains unknown whether the subjectively enhanced susceptibility to stress in Parkinson’s disease is associated with enhanced activation of the hypothalamic-pituitary axis, and whether it has direct effects on cognitive function. To address this gap, we compared stress responses in 50 Parkinson’s disease patients and 51 older healthy controls (HC). Participants underwent two sessions, an acute stress induction, using the socially evaluated cold pressor task, and a control session. Stress responses were measured using salivary cortisol, heart rate, blood pressure and self-reported affect at multiple timepoints. To evaluate the effects of stress on cognitive performance, participants completed several tasks measuring working memory and executive function. Preliminary analyses show that both groups exhibited a significant increase in self-reported stress, cortisol and blood pressure following the stress manipulation. Furthermore, consistent with previous stress research in older adults, our results show positive stress effect on working memory performance seen in older adults but not in Parkinson’s patients.

The highs and lows of cannabis use: Pilot data on the effects of 28-days of cannabis abstinence on affective symptoms

Lyne Baaj1,2, Rachel Rabin1,2

1Integrated Program in Neuroscience, McGill University, Montreal, QC, Canada
2Douglas Research Centre, Montreal, QC, Canada

Corresponding Author: Rachel Rabin, email: rachel.rabin@mcgill.ca


Background and Objectives: Despite consistent findings that recreational cannabis use does not improve symptoms of depression and anxiety, many continue to use cannabis to cope with low mood and stress. This is concerning because cannabis use is, paradoxically, associated with poorer mood and anxiety, although causal evidence is equivocal. To better understand the relationship between cannabis use and affective symptoms, we employed a cannabis abstinence paradigm to study the effects of 28-days of cannabis abstinence on depressive and anxiety symptoms in individuals with cannabis use disorder (CB+). Methods: CB+ participants (N=7) were randomized to an “abstinence” group (n=4) or to a “cannabis use as normal” control group (n=3). Depressive symptoms were assessed using the Hamilton-Depression Rating Scale and anxiety was assessed using the State Trait Anxiety Inventory. Affective symptoms were measured before and following 28-days of cannabis abstinence or use as normal. Cannabis abstinence was determined using a self-report questionnaire. Data was interpreted graphically. Results: Three participants (75%) successfully maintained cannabis abstinence for the 28-day period. Cannabis abstinence did not improve affective symptoms, relative to control. Discussion: Our pilot findings suggest that affective symptoms may not change with 28-days of cannabis abstinence. Significance: Chronic cannabis use may contribute to the onset and maintenance of affective symptoms, which may persist with one month of abstinence. Data collection is ongoing and as we increase our sample size, we hope to have more insight into the effects of cannabis abstinence on affective symptoms.

Predicting brain health from resting-state neuroimaging using machine learning

Olivier Valentin1,2,3,4, Antonin Perrin5, Alexandre Lehmann1,2,3,4

1Laboratory for Brain, Music and Sound Research - Centre for Research on Brain, Language or Music, Montréal, QC, Canada
2Centre for interdisciplinary research in music media and technology, Montreal, QC, Canada
3McGill University, Faculty of Medicine, Department of Otolaryngology–Head and Neck Surgery, Montreal, QC, Canada
4Research Institute of the McGill University Health Centre, Montreal, QC, Canada
5École Polytechnique Universitaire de Lyon, Lyon, France

Corresponding Author: Alexandre Lehmann, email: alexandre.lehmann@mcgill.ca


Alzheimer's disease is a prevalent neurodegenerative disorder affecting up to 6.5 million of people in the United States. The disease is associated with faster rates of brain atrophy compared to healthy individuals, suggesting that Alzheimer's may be a form of accelerated brain aging. While current treatments for Alzheimer's only manage symptoms or slow disease progression, early diagnosis is critical for both patients and families to better understand the disease, plan for the future, and initiate therapies that may slow progression and improve management. Unfortunately, the current diagnostic methods for Alzheimer's rely on the documentation of significant cognitive decline, by which time severe brain damage has already occurred. Therefore, there is a critical need to develop a diagnostic tool that can detect early manifestations and accurately diagnose Alzheimer's in its earliest stages. Here, we present a machine learning approach able to predict participant’s age based on electroencephalography (EEG) data, with potential applications for developing a predictive tool to detect early premature aging of the brain. The best accuracy result was obtained using a Gradient Boosting classifier (76.6%, p<0.01) with a 0.5 learning rate. The correlation between the predicted and estimated age confirms the ability of our models to predict age. These results confirm that (1) aging affects human brain EEG signals, and (2) age can be predicted using resting state EEG data. Although this project used a neuroimaging database of healthy individuals, the outcome of this research project might enable early detection in individuals who have not yet been diagnosed.

Towards building an in vitro 3D neuromuscular junction model of ALS

Maria Jose Castellanos Montiel1, Anna Krystina Franco-Flores1, Mathilde Chaineau1, Lale Gursu1, Ghazal Haghi1, Carol X-Q Chen1, Michael Nicouleau1, Wolfgang E. Reintsch1, Thomas M. Durcan1, Eric Deneault2

1The Neuro’s Early Drug Discovery Unit, McGill University, Montreal, QC, Canada
2Centre for Oncology, Radiopharmaceuticals and Research; Biologic and Radiopharmaceutical Drugs Directorate, Health Products and Food Branch, Health Canada, Ottawa, ON, Canada

Corresponding Author: Thomas M. Durcan, email: thomas.durcan@mcgill.ca


Motor neurons (MNs) are the neuronal cells that control muscle contraction through specialized connections called neuromuscular junctions (NMJs). In amyotrophic lateral sclerosis (ALS), these connections are disrupted due to selective degeneration of MNs in the cortex, brainstem, and spinal cord. Induced pluripotent stem cells (iPSCs) have emerged as a novel tool to model neurodegenerative diseases as they retain the genetic background of the patients and can be differentiated into any cell type, including many cells of the central nervous system (CNS). Furthermore, iPSCs can be used to generate 3D models in which cell-to-matrix and cell-to-cell interactions are recapitulated in a more physiological matter than with 2D models. In our lab, we differentiate iPSCs into MN progenitor cells (MNPCs) that are seeded into low attachment plates to promote their aggregation and maturation as MN spheroids. Using this method, we generated MN spheroids from a CRISPR/Cas9-edited iPSC line carrying an ALS-associated mutation in the superoxide dismutase 1 (SOD1) gene and its isogenic control. Additionally, we are able to culture human primary myoblasts and differentiate them into skeletal muscle micro-tissues, which can be co-cultured with MN spheroids to generate human 3D in vitro models of the NMJ. With this model in place and the tools developed for its analysis, we aim to study the morphology and functionality of the NMJ, an early and primarily affected structure in the development of ALS.

Modulatory feedback

Mashbayar Tugsbayar1,2, Eilif Muller3,2, Blake Richards1,2

1McGill University, Montreal, QC, Canada
2Mila quebec, Montreal, QC, Canada
3Université de Montréal, Montreal, QC, Canada

Corresponding Author: Blake Richards, email: blake.richards@mila.quebec


Many recent advances in artificial neural networks have been inspired by computations in the human brain. Even so, processes in modern neural networks often do not resemble the biological process it derives inspiration from. Differences from the human visual system, for instance, result in differing representations at higher levels and differing behaviors, such as inability to deal with occluded, blurred and otherwise challenging stimuli that humans can identify based on context. We hypothesize these differences are due to the omission of top-down feedback connections which are prevalent in the brain. Using diffusion imaging and histological data, we aim to build a model of the visual system that has human-like connectivity and uses top-down feedback. Preliminary results reveal that modulatory top-down feedback connections carrying contextual information or an auditory clue help different networks identify ambiguous images. By incorporating human brain-based inductive biases and modulatory top-down feedback, we hope to build a model of the visual system that more closely resembles the representations and capabilities of the human brain. The model aims to be effective both for tasks that require human-like decision making and studies investigating the role of various connections in the visual system.

A consensus algorithm for genetic demultiplexing of scRNA-seq from pools of multiple subjects for cohort-scale analysis

Michael Fiorini1, Saeid Amiri2, Rhalena Thomas3, Sali Farhan1,3

1Department of Human Genetics, McGill University, Montreal, QC, Canada
2Montreal Neurological Institute, Montreal, QC, Canada
3Department of Neurology and Neurosurgery, McGill University, Montreal, QC, Canada

Corresponding Author: Sali Farhan, email: sali.farhan@mail.mcgill.ca


Single-cell RNA sequencing (scRNA-seq) of in vitro cellular disease models can inform the transcriptional landscape of unique cell types in neurodegenerative diseases. However, scRNA-seq of induced pluripotent stem cells (iPSC) remains costly and has low throughput, restricting the feasibility of cohort-scale studies. Multiplexing iPSC lines bypasses the costly parallel growth of individual iPSC lines by pooling batches of lines in a single dish. Demultiplexing algorithms can then resolve the sample identity of each cell based on genetic differences. However, the accuracy of available demultiplexing software remains unknown as benchmarking has been based on small sample sizes. To facilitate cohort-scale iPSC-based scRNA-seq, we are developing a consensus demultiplexing software that assigns sample identities to individual cells based on the predictions of six available demultiplexing algorithms. We are testing our consensus algorithm on synthetically pooled iPSC-derived dopaminergic neuron scRNA-seq transcriptomes from controls and Parkinson’s disease (PD) patients. We will produce and test in silico pools ranging from 4 to 80 iPSC lines to determine if our algorithm remains accurate for cohort-scale cell batches. Preliminary results demonstrate that our algorithm outperforms four other demultiplexing algorithms for pools comprised of three different iPSC lines. Eventually, we will use our consensus demultiplexing algorithm to test the hypothesis that multiplexed iPSC lines and scRNA-seq can be used to screen for transcriptional responses to chemical compounds in dopaminergic neurons and microglia in PD patients. Importantly, we plan to share our demultiplexing algorithm to facilitate cohort-scale iPSC-based scRNA-seq across neurodegenerative diseases, contributing to important open science efforts.

Gaze behaviors in dyadic Interactions: Role of face masks on prevalence of recurrent looks

Mikayla Strandberg1, Florence Mayrand1, Jelena Ristic1

1Department of Psychology, McGill University, Montreal, QC, Canada

Corresponding Author: Jelena Ristic , email: jelena.ristic@mcgill.ca


Gaze conveys important social messages, such as intentions and emotions. In natural dyadic or two-person interactions, somewhat surprisingly people tend to look little at their partner’s face, and in general avoid eye contact. Recent research shows that face masks exert a sizable detriment on reading of facial emotions. In this work, we are investigating how face occlusion by masks may impact attentional looking. In our study, dyads engaged in a natural interaction while wearing a face mask. Their eye movements were measured using dual mobile eye trackers. The recordings were analyzed for the amount of time participants look towards the mouth, eyes, and body regions of their interactive partner, and were classified as various forms of mutual gaze behaviors. The amount of time spent in mutual gaze behaviors as a function of regions of looking will be analyzed. The results showed that participants spent more time overall looking at the eyes of their partners than they did looking at any other region of interest. In addition, mutual engagement seemed to occur primarily through direct eye-to-eye gaze and eye-to-mouth gaze. Taken together, these results provide additional insight into gaze in real-life interaction and the role of mutual engagement in dyadic pairs, as well as the impact of facial occlusion on such behaviors. 

Distinct injury response of human oligodendrocytes to metabolic stress: Relevance for progressive multiple sclerosis

Milton Guilherme Forestieri Fernandes1, Abdulshakour Mohammadnia1, Laura Eleonora Schmitz-Gielsdorf1, Caroline Hodgins1, Florian Pernin1, Qiao-Ling Cui1, Moein Yaqubi1, Manon Blain1, Jeffery Hall1, Roy Dudley1, Myriam Srour1, Stephanie E. J. Zandee1, Wendy Klement1, Alexandre Prat1, Jo Anne Stratton1, Moses Rodriguez1, Tanja Kuhlmann1, Timothy E. Kennedy1, Jack P. Antel1

1McGill University, Montreal, QC, Canada

Corresponding Author: Jack Antel, email: jack.antel@mcgill.ca


Oligodendrocyte (OL) loss is a central feature of progressive MS. A potential cause for this loss is metabolic stress. Here we aim to define the injury response of primary human OLs (hOLs) to metabolic stress (reduced glucose/nutrients) in vitro and relate these to OLs in cases of MS. In MS lesions and in vitro under stress conditions, we observe cell shrinkage and a reduction in OL nuclear area. Under metabolic stress, we detect reduction in ATP per cell that precede changes in survival. Autophagy was initially activated, although ATP levels were not altered by autophagy modulators. Prolonged stress resulted in autophagy failure. Although prolonged stress resulted in increased ROS and cleavage of spectrin, cell death was not prevented by inhibitors of ferroptosis or MPT-driven necrosis, the regulated cell death (RCD) pathways likely to be activated by metabolic stress. hOLs had decreased expression of VDAC1, VDAC2, and genes regulating iron accumulation and cyclophilin. Consistent with in vitro results, we detected an increase in autophagosomes in OLs in MS lesions compared to controls. Gene expression analysis indicated lack of up-regulation of RCD pathways in OLs in active MS lesions. We conclude that this distinct response of hOLs, including resistance to RCD, reflects the combined impact of autophagy failure, increased ROS, and calcium influx, resulting in metabolic collapse and degeneration of cellular structural integrity. Defining the bases of OL injury and death in MS provides guidance for development of protective strategies.

Copy number variants in CACNA1H implicated in essential tremor

Miranda Medeiros1, Calwing Liao1,2, Dan Spiegelman2,3, Allison Dilliott2, Jay Ross1,2, Sali Farhan1,2, Patrick A. Dion2,3, Guy A. Rouleau1,2,3

1Department of Human Genetics, McGill University, Montreal, QC, Canada
2Montreal Neurological Institute, McGill University, Montreal, QC, Canada
3Department of Neurology and Neurosurgery, McGill University, Montreal, QC, Canada

Corresponding Author: Guy Rouleau, email: guy.rouleau@mcgill.ca


Background/Objectives: Essential tremor (ET) is the most common movement disorder globally, affecting approximately 5% of the population by the age of 65. ET leads to significant impairment and diminished quality of life due to tremors occurring in the hands during voluntary motion. Despite ET having an overall heritability of between 45-90%, only a limited number of genetic risk factors have been identified. Here, we leveraged a large ET cohort of 2,133 cases and 10,336 controls to identify copy-number variants (CNVs) associated with ET. Methods: Samples were limited to those of European ancestry inferred through PCA. CNVs were called from SNP array data using the program PennCNV and annotated for functional consequence and population frequency using gnomAD. Only rare CNVs intersecting protein coding regions of the genome were tested. The enrichment of CNVs per gene between cases and controls was investigated using Fisher’s exact test. Only the candidate genes that passed Bonferroni correction were considered. Results: Following QC, 3,516 CNVs across 1,638 genes were tested across 1,365 cases and 9,585 controls. Gene duplications in the Calcium Voltage-Gated Channel Subunit Alpha1H gene (CACNA1H) were significantly more common in ET cases (18 CNVs, p = 7.1×10⁻⁹) and were absent in any controls examined. Approximately 1.3% of studied cases carried this rare structural variant. Conclusion: The observation of duplications of the CACNA1H gene implicates this gene as a risk factor for ET and adds to the potential involvement of CACNA genes in ET pathology. Grant References: Foundation grants - Canadian Institutes of Health Research

Age-related differences in the relationship between the basal forebrain volume, functional connectivity, and cognition

Miriam Taza1, Taylor W. Schmitz2, Roni Setton3, Laetitia Mwilambwe-Tshilobo4,5, Gary R. Turner6, R. Nathan Spreng1

1McGill University, Montreal, QC, Canada
2Western University, London, ON, Canada
3Harvard University, Cambridge, MA, United States
4University of Pennsylvania, Philadelphia, PA, United States
5Princeton University, Princeton, NJ, United States
6York University, Toronto, ON, Canada

Corresponding Author: Nathan Spreng, email: nathan.spreng@mcgill.ca


The basal forebrain (BF) comprises large cholinergic projection neurons that innervate the entire cortical mantle. Acetylcholine is involved in low-frequency sleep-wake cycles of alertness, in addition to a rapid modulation of cortical processes involved in attention. The cholinergic BF is impacted early in the progression of Alzheimer’s disease. However, age-related differences in the relationships between BF structure, functional connectivity, and attention are poorly understood. Neuropsychological assessment of attention, anatomical and resting-state multi-echo functional MRI were analyzed in a sample of 145 younger (mean age=22y, SD=3y) and 75 older (mean age=68y, SD=6y), cognitively intact, healthy adults. In older adults, BF volume was smaller than younger adults (p<0.01) and was related to a measure of executive attention (self-ordered search), where larger BF volume was associated with faster reaction time (r =-.24, p<.01) and higher task accuracy (r =.23, p<.05), after accounting for sex, age, education, and intracranial volume. No associations were observed for semantic and episodic memory, vocabulary or processing speed (p’s>.20). Age-group differences were also observed in a multivariate partial least squares analysis comparing BF connectivity with known cortico-cortical resting state networks (p<.001). In young adults, the magnitude of BF connectivity was higher in regions of the limbic, salience, and somatomotor networks. In contrast, older adults showed greater BF connectivity to regions of the default network. These results provide novel evidence that functional connectivity between the BF and neocortex changes with advancing age, and that these functional changes are related to BF structural integrity and executive attention.

Do commonly-used dynamic functional connectivity methods produce comparable results?

Mohammad Torabi1, Jean-Baptiste Poline1, Georgios Mitsis1

1McGill University, Montreal, QC, Canada

Corresponding Author: Jean-Baptiste Poline and Georgios Mitsis , email: jean-baptiste.poline@mcgill.ca; georgios.mitsis@mcgill.ca


Functional connectivity (FC) has become a popular measure for understanding brain function and as a potential biomarker. Dynamic variations in FC (dFC) are believed to play a critical role in brain functional organization and may provide a link between neural dynamics and cognition in health and disease. Although various methods have been developed and used for assessing dFC, the impact of methodological choices on results remains poorly understood and understudied. This study aimed to assess the analytical flexibility of dFC estimation by comparing dFC patterns obtained using the same dataset but different methods. Seven widely-used dFC assessment methods were implemented in Python and used to estimate the dFC matrix for 395 preprocessed fMRI subjects data from the Human Connectome Project (HCP). The analytical flexibility of dFC results across methods was assessed by measuring pairwise similarity using Spearman correlation. The similarity between each pair of methods was assessed in terms of total similarity, temporal similarity, spatial similarity, and graph-based similarity. DFC methods were also compared in terms of the correspondence between their inter-time point relations and subject-wise clustering. The obtained results suggest considerable variability across methods with correlation values ranging between weak to strong. Hierarchical clustering based on these correlation matrices suggests that some methods yielded similar results and can be grouped together. The identified patterns of similarity can be mostly attributed to the similarity in the assumptions of methods. These findings highlight the importance of careful methodological considerations when assessing dFC, particularly in clinical and behavioral applications.

Genetic modulators of α-Synuclein aggregate propagation in Parkinson’s disease: A genome-wide CRISPR approach

Nathan Karpilovsky1, Thomas Goiran1, Jace Jones-Tabah1, Julien Sirois1, Graham MacLeod2, Wen Luo1, Esther Del Cid Pellitero1, Benoit Vanderperre3, Irina Shlaifer1, Stephane Angers2, Thomas Durcan1, Edward Fon1

1McGill University, Montreal, QC, Canada
2University of Toronto, Toronto, ON, Canada
3Université du Québec à Montréal, Montréal, QC, Canada

Corresponding Author: Edward Fon, email: ted.fon@mcgill.ca


One of the major hallmarks of Parkinson’s Disease (PD) is the presence of Lewy bodies, which are comprised of protein aggregates including α-synuclein (αSyn). Aggregates of αSyn derived from Lewy bodies have been shown to spread PD pathology in the brain, contributing to disease progression. Lab-generated preformed fibrils (PFF) of αSyn have also been shown to propagate transcellularly between neurons, which is thought to augment their toxic effect in mouse models. While many pathways have been suggested to play a role in the internalization and propagation of αSyn aggregates, our understanding of the various mechanisms remains limited. Our lab previously conducted a CRISPR knockout (KO) screen to identify genes required for PFF uptake; whereas, in this work, we utilize an opposite and complementary approach to activate genes with CRISPR activation (CRISPRa). We identified putative genetic modulators of αSyn fibril uptake by performing a FACS-based genome-wide CRISPRa screen on the accumulation of fluorescently labelled PFF in immortalized cells. This approach yielded distinct genetic targets which are suggested to play a role in inflammation, endo-lysosomal trafficking, and protein glycosylation. Select genes will be further validated using iPSC-derived neurons as a physiologically relevant model to provide further insight into the mechanisms of αSyn accumulation. This research will shed light on the function of key genes and proteins that are important in the neuronal uptake of αSyn fibrils. Identifying novel genetic targets will then allow for potential development of therapeutics, combating PD symptoms and/or halting its progression.

Developing a novel dichoptic reading application to treat amblyopia

Nicole Dranitsaris1, Ken Chong2, Robert F. Hess1, Alexandre Reynaud1

1Department of Ophthalmology and Visual Sciences, McGill University, Montreal, QC, Canada
2Research Institute of the McGill University Health Centre, Montreal, QC, Canada

Corresponding Author: Alexandre Reynaud, email: alexandre.reynaud@mcgill.ca


The neurodevelopmental disorder amblyopia, or lazy eye, occurs when the brain can only process visual information from one eye. Current amblyopia treatment research is focused on binocular tasks instead of the typical patching treatment which is not effective across age groups and has low compliance rates. The goal of this study was to assess reading as a binocular treatment for amblyopia. Here, a dichoptic e-book application was developed as a platform to study the feasibility of this task for amblyopes. An application prototype was developed and uploaded onto tablets for participant assessments. Using anaglyph red/green/black presentation to allow for independent adjustments of monocular and binocular contrast participants read e-books. Controls and amblyopes were questioned on application use comfortability and their reading speed was measured. We discovered that amblyopes read significantly slower than controls in all the dichoptic presentations and amblyopic participants read slower in all dichoptic presentations than in control presentations. This indicates that their visual systems were forced to integrate information from both eyes. Furthermore, when the contrast of text seen by the fellow eye was reduced, reading speed increased in accordance with current research on binocular training approaches. In conclusion, this research showed that amblyopes can read binocularly using the e-book application platform suggesting that it may be an effective treatment for amblyopia. Future steps in this research are aimed towards training amblyopes on reading in this application to improve their binocular vision.

Spatial clustering of white matter hyperintensities based on their microstructural properties

Olivier Parent1,2, Aurélie Bussy1,2, Gabriel A. Devenyi1,3, Grace Pigeau1,2, Manuela Costantino1,2, Christine L. Tardif4,5,6, Mahsa Dadar3, Mallar Chakravarty1,2,3,4

1Cerebral Imaging Center, Douglas Mental Health University Institute, Montreal, QC, Canada
2Integrated Program in Neuroscience, McGill University, Montreal, QC, Canada
3Department of Psychiatry, McGill University, Montreal, QC, Canada
4Department of Biomedical Engineering, McGill University, Montreal, QC, Canada
5McConnell Brain Imaging Center, Montreal Neurological Institute, Montreal, QC, Canada
6Department of Neurology and Neurosurgery, McGill University, Montreal, QC, Canada

Corresponding Author: Mallar Chakravarty, email: mallar.chak@gmail.com


White matter hyperintensities (WMHs) are age-related radiological abnormalities widely recognized as indicators of small vessel disease. However, the tissue alterations underlying WMHs are highly heterogeneous, and part of this variability can be explained by the different spatial locations of WMHs. Here, we describe voxel-wise microstructural abnormalities of WMHs using diffusion magnetic resonance imaging (MRI). We used a subset of 2627 UK Biobank participants. We performed automated WMH segmentation, preprocessing of diffusion MRI data, and non-linear registration to MNI space. We extracted Diffusion Tensor Imaging (DTI) metrics, including fractional anisotropy (FA) and mean diffusivity (MD), and Neurite Orientation Dispersion and Density Imaging (NODDI) metrics, including the intra-cellular volume fraction (ICVF), the isotropic volume fraction (ISOVF), and the orientation dispersion index (OD). Voxel-wise average microstructural deviations from the healthy white matter were calculated with Z-values. Microstructurally-similar clusters of WMH voxels were uncovered using spectral clustering. The first cluster was touching the ventricles and had low microstructural abnormality overall. The second cluster was mostly in periventricular-posterior regions and had medium microstructural abnormality only on MD and ICVF metrics. The third cluster was mostly in superficial white matter (SWM) regions and had medium microstructural abnormality on all metrics except ISOVF (Z ≈ +/- 2). The fourth cluster was mostly in periventricular-anterior regions, had very high MD and ICVF abnormality (Z > +/- 4), as well as medium abnormality in FA and ISOVF (Z > +/- 1). Using a parcellation-free approach in a large aging cohort, we have identified spatial WMH components that share similar diffusion MRI-derived microstructural properties.

Genome-wide environment interaction studies (GWEIS) identify susceptibility genetic variants related to executive functioning deficits in individuals exposed to early adversity

Patrícia Maidana Miguel1, Janou Liu2, Guillaume Elgbeili3, Irina Pokhvisneva3, Michael J. Meaney1,3,4,5, Patrícia Pelufo Silveira1,3

1Department of Psychiatry, Faculty of Medicine, McGill University, Montreal, QC, Canada
2Undergraduate Studies in Cognitive Science, Faculty of Science, McGill University, Montreal, QC, Canada
3Ludmer Centre for Neuroinformatics and Mental Health, Douglas Mental Health University Institute, McGill University, Montreal, QC, Canada
4Singapore Institute for Clinical Sciences, Agency for Science, Technology and Research, Singapore, Singapore
5Brain-Body Initiative, Agency for Science, Technology and Research, Singapore, Singapore

Corresponding Author: Patricia Pelufo Silveira, email: patricia.silveira@mcgill.ca


Exposure to early life adversity (ELA) is a risk factor for altered executive functioning (EF), which is highly involved in both physical and mental health. However, there are large individual differences in the effects of ELA, and genetic differences may be involved. Aiming at identifying individuals that are more vulnerable to the damaging effects of ELA, we performed two independent genome-wide environment interaction studies (GWEIS) to quantify the interaction of single nucleotide polymorphisms (SNP) with trauma exposure (felt hated/unloved as a child) to predict type 2 diabetes (GWEIS-T2D) or mood disorders (GWEIS-Mood) in adults from the UK Biobank (n=97,583). Using a p<5x10-5 threshold, GWEIS-Mood identified 448 SNPs that mapped to 15 genes, and the GWEIS-T2D discovered 120 SNPs that mapped to 18 genes. Genes from the GWEIS-Mood are mainly upregulated in brain, whereas genes from the GWEIS-T2D are upregulated especially in peripheral tissues (aorta, thyroid, and gastroesophageal junction). We then calculated GWEIS-derived polygenic risk scores (PRS-T2D or PRS-Mood) by selecting the GWEIS SNPs that were specifically significant in individuals exposed to adversity. In the ALSPAC cohort of young individuals (n=3,190), both PRS-Mood and PRS-T2D interacted with ELA to predict working memory, divided attention, and inhibitory control problems. Only individuals with the high PRSs showed EFs deficits in the presence of adversity, confirming that susceptibility genetic variants found in adults predicted altered EF in young individuals. The GWEIS seems to be a promising approach to capture specific SNPs that interact with the environment to predict impaired EFs and later chronic diseases.

mGluR5 dysfunction underlies mGluR-LTD deficit in the absence of synaptopodin

Pei You Wu1, Yanis Inglebert1, R. Anne McKinney1

1Department of Pharmacology & Therapeutics, McGill University, Montreal, QC, Canada

Corresponding Author: Anne McKinney, email: anne.mckinney@mcgill.ca


Synaptopodin (SP) is an actin-associated protein found only in a subset of excitatory synapses, mainly in the larger and more stable dendritic spines of the telencephalic neurons. It is necessary for the formation of spine apparatus, an organelle located at the base of dendritic spines that is involved in local protein synthesis and calcium regulation in individual spines. It has been shown that SP knock-out mice (SPKO) exhibits normal synaptic transmission and dendritic spine density. However, synaptic plasticity such as NMDAR-LTP was found impaired in SPKO. It is currently unknown whether SP is involved in other types of plasticity such as metabotropic glutamate receptor-dependent long-term depression (mGluR-LTD). mGluR-LTD is critically involved in learning and memory formation in brain. Enhanced mGluR-LTD is associated with Fragile X syndrome, where an elevated level of SP is also observed. To understand the role of SP in mGluR-LTD, we used electrophysiology technique and found that hippocampal mGluR-LTD is impaired in SPKO. Furthermore, we found that mGluR-LTD becomes mGluR5- and protein synthesis-independent in SPKO. Lastly, we show evidence that, although mGluR-LTD is induced primarily through postsynaptic events, the lack of SP switched mGluR-LTD to be presynaptically driven.

A novel adaptive method for measuring point of subjective equality

Penghan Wang1, Alexandre Reynaud1

1McGill University

Corresponding Author: Alexandre Reynaud, email: alexandre.reynaud@mcgill.ca


Points of Subjective Equality (PSEs) are usually measured with the constant stimuli (CS) method. However, it’s found time-consuming and inefficient. Thus, we developed a more efficient adaptive (AD) method to estimate the PSE and its variability, in which levels are chosen on a pre-defined scale. Instead of testing each level with the same number of repetitions, each stimulus would be chosen depending on the previous response of the participant. Specifically, if the participant responded “up”, a random level in the lower range would be picked for the next trial. And if responded “down”, a random level in the upper range would be picked. A bell-shaped distribution of the tested levels around the estimated PSE was expected. The AD method aims to maximize the number of trials near the PSE while keeping the distribution relatively large for the calculation of the slope. In this study, we compared this method with the CS method on a task based on the Pulfrich phenomenon with a fixed number of trials, while the PSEs could be varied using different ND filters. As a result, we observed a significant correlation between the PSEs obtained by the two methods. And in most cases, our adaptative method yielded comparable performance for the estimates of the PSE and the slope of the psychometric function. Therefore, we can conclude that the adaptative method is an efficient way of measuring PSEs with our Pulfrich paradigm. It could be useful in clinical settings where testing time is limited.

The role of IRF1 and IRF8 in cerebral malaria

David Langlais1,2, Rebecca Bellworthy1

1Department of Human Genetics, McGill Genome Centre, Quebec, Canada
2Department of Microbiology and Immunology, McGill University, Montréal, QC, Canada

Corresponding Author: David Langlais, email: david.langlais@mcgill.ca


Cerebral malaria (CM) is the most severe complication of malaria in humans that is responsible for >400,000 deaths each year for which there is currently no effective treatment. Studies in mice with knockouts of IRF1 and IRF8 genes show immunodeficient phenotypes with particular susceptibility to intracellular pathogens, and extensively disrupted immune cell maturation and function. We have observed that these mice are also highly resistant to the experimental cerebral malaria (ECM) model, which is typically rapidly fatal due to excessive neuroinflammation. By determining the function of these genes in immune regulation, we hope to further our understanding of CM disease pathogenesis and ultimately discover potential therapeutic targets. This presentation includes characterization of the immunophenotype of these mice, with results from ECM infections in our cohort of single and double knockout mice alongside genetic analyses through techniques such as chromatin immunoprecipitation (ChIP) of the immune cells extract from the mice.

A characterization of cerebellar perineuronal nets in humans, mice, and macaques

Refilwe Mpai1,2, Christa Hercher1,2, Claudia Belliveau1,2, Maria-Antonietta Davoli2,3, Naguib Mechawar2,4

1Integrated Program in Neuroscience, McGill University, Montreal, QC, Canada
2McGill Group for Suicide Studies, McGill University, Montreal, QC, Canada
3Douglas-Bell Canada Brain Bank, Montreal, QC, Canada
4Department of Psychiatry, McGill University, Montreal, QC, Canada

Corresponding Author: Naguib Mechawar, email: naguib.mechawar@mcgill.ca


Introduction: Perineuronal nets (PNNs) have been shown to restrict neuroplasticity and stabilize synapses. PNNs have been well-characterized in sensory cortices, though little is known about them in the cerebellum (CB), especially in humans. This study aims to characterize CB PNNs through a cross-species comparison of mice, macaques and humans. Methods: Post-mortem human CB from neurologically and psychiatrically healthy individuals were provided by the Douglas-Bell Canada Brain Bank. CB from wild-type mice and cynomolgus macaques were obtained through collaborations. Using immunofluorescence, we labelled PNNs using Wisteria Floribunda Lectin (WFL) and anti-aggrecan (ACAN) antibodies and labelled parvalbumin(PV)-expressing neurons in the animal samples with an anti-PV antibody. Fluorescent in situ hybridization (FISH) was employed to label vesicular glutamate transporter 1 (SLC17A7), glutamate decarboxylase 1 (GAD1), and PV to determine the phenotype of cells surrounded by PNNs in the human CB. Preliminary results: Across the species studied, we observed WFL+ and ACAN+ PNNs in the CB nuclei. In mice and macaques, we also found WFL+ and ACAN+ PNNs in the CB cortex with differences in expression between markers. In the mouse CB, no PNNs were surrounding PV+ neurons. In macaques, PNNs in the CB nuclei mostly surrounded PV+ neurons, while those in the CB cortex were mostly PV-. FISH experiments revealed that human CB PNNs mostly surround PV+/SLC17A7+ neurons. Conclusion: This work highlights species differences in the nature and distribution of CB PNNs, and paves the way for future studies on PNN-related CB neuroplasticity in the healthy and disordered brain.

Exploring surround suppression mechanisms in amblyopia: A psychophysical study

Rinku Sarkar1, Frederick A.A. Kingdom1, Alexandre Reynaud1

1McGill Vision Research, Department of Ophthalmology and Visual Sciences, McGill University, Montreal, QC, Canada.

Corresponding Author: Alexandre Reynaud, email: alexandre.reynaud@mcgill.ca


Amblyopia, a developmental vision disorder, is characterized by persistent spatial deficits in the amblyopic eye due to strong interocular suppression from the fellow eye. This study aimed to compare the surround suppression mechanisms in amblyopia to discern whether the suppression arises from the fellow eye or the amblyopic eye itself. Using a dichoptic center-surround masking paradigm, subjects performed a psychophysical task of detecting a central target in a two-interval forced-choice procedure. Stimuli were horizontally oriented 0.5 cycles per degree (cpd) gratings, with the central test stimulus measuring 2 degrees and the surround mask measuring 6.5 degrees in diameter. The experimental setup consisted of a pair of interleaved test conditions, where the target was presented either to the amblyopic eye or to the fellow eye, along with three distinct surround mask conditions i.e., absence of a mask, a fellow eye mask, and an amblyopic eye mask. The results indicated that contrast thresholds in the fellow eye were more elevated when the mask was also in the fellow eye, indicating monocular suppression. In contrast, amblyopic eye thresholds were raised in both fellow and amblyopic eye masking conditions, but to a greater extent in the fellow eye masking condition, indicating dichoptic suppression. Overall, the findings suggest that amblyopes exhibit greater dichoptic suppression in the amblyopic eye from fellow eye masks compared to suppression in the fellow eye from amblyopic eye masks. Furthermore, amblyopes are vulnerable to suppression, regardless of whether the suppression arises from the amblyopic eye or from the fellow eye.

Investigating if the small Rho GTPase Rac1 is misregulated in Christianson Syndrome

Roy Shi1, Andy Yuan Lee Gao1, Pei You Wu1, Jamie Mustian1, John Orlowski1, R. Anne McKinney1

1McGill University, Montreal, QC, Canada

Corresponding Author: Anne McKinney, email: anne.mckinney@mcgill.ca


Christianson syndrome (CS) is a monogenic X-linked neurodevelopmental/neurodegenerative disorder caused by the loss-of-function mutations of the SLC9A6 gene, which encodes for the endosomal (Na+, K+)/H+ exchanger 6 (NHE6). NHE6 regulates endosomal pH to carefully control the accurate trafficking of cargo necessary for neuronal function. Loss of NHE6 leads to over-acidification of the endosomal lumen, resulting in cargo mistrafficking and preventing structural and functional long-term potentiation (LTP). Growing evidence suggests that the disruption of Rac1, a Rho family GTPase, known to play a central role in cytoskeletal remodelling at glutamatergic synapses during the learning signalling pathway, contributes to Autism Spectrum Disorder and prevents structural and functional LTP. As Rac1 is located in recycling endosomes and we found that NHE6 binds to Rac1, we wondered if the loss of NHE6 modulates Rac1, thus preventing functional LTP and spine remodelling post-LTP. Using IHC and WB, we found no difference in Rac1 total protein between CS and WT. In CS before and after ChemLTP-induced learning, there are decreased dendritic and spine Rac1 puncta. We found more Rac1 puncta present in lysosomes in the hippocampus of CS vs WT, suggesting that Rac1 may be undergoing erroneous proteolysis. Using leupeptin, a protease inhibitor, we found it was able to rescue the total dendritic Rac1 puncta density before and after ChemLTP in CA1 hippocampal pyramidal cells. Spine Rac1 puncta density was unaffected in cells with or without ChemLTP treatment. The present study gives insight into potential treatments for learning deficits in CS by modulating Rac1 trafficking.

Introduction of a fast multi-shot three-dimensional diffusion MRI sequence

Sajjad Feizollah1, Christine Tardif2,3,4

1Integrated Program in Neuroscience, McGill University, Montreal, QC, Canada
2Department of Neurology and Neurosurgery, McGill University, Montreal, QC, Canada
3McConnell Brain Imaging Center, Montreal Neurological Institute, McGill University, Montreal, QC, Canada
4Department of Biomedical Engineering, McGill University, Montreal, QC, Canada

Corresponding Author: Christine Tardif, email: christine.tardif@mcgill.ca


Several acquisition techniques have been developed to improve the signal-to-noise ratio (SNR) and resolution of 2D diffusion MRI of the brain. Most high-resolution techniques use a multi-shot approach, which requires a long scan time. We propose a fast 3D multi-shot implementation that includes an additional inversion pulse prior to the excitation to improve the SNR at short repetition times. Bloch simulations were used to optimize acquisition parameters and demonstrate the gain in SNR efficiency. Initial phantom and human scans at 3T show an approximately two-fold increase in SNR compared to conventional 2D multi-slice diffusion imaging within the same scan time.

Sociodemographic and clinical correlates of hallucinations at admission to an early intervention program for first episode psychosis

Samantha Aversa1, Joseph Ghanem2,1, Gili Grunfeld2,1, Ann-Catherine Lemonde2,1, Srividya Narayanan Iyer2,1, Ashok Malla2,1, Martin Lepage2,1, Ridha Joober2,1, Jai Shah2,1

1Douglas Mental Health University Institute, Montreal, QC, Canada
2Department of Psychiatry, McGill University, Montreal, QC, Canada

Corresponding Author: Jai Shah, email: jai.shah@mcgill.ca


Hallucinations are a core feature of first-episode psychosis (FEP), and their acute phase severity is associated with a host of poor outcomes. Previous research indicates that factors such as depression, anxiety, negative symptoms, and longer durations of untreated psychosis may predict various positive symptoms in FEP. Despite this work, the precise sociodemographic and clinical factors associated with hallucinations in a clinical population at first presentation to an early intervention service have not been extensively researched. We therefore investigated sociodemographic and clinical factors in 636 minimally-medicated patients followed longitudinally from 2003-2018 at PEPP-Montréal, an early intervention service for FEP. Hallucinations were measured using the Scale for the Assessment of Positive Symptoms (SAPS), along with negative, depressive, and anxiety symptoms. Sociodemographic variables such as age, sex and education level were collected through interviews and medical files. A majority (n = 381, 59.9%) of the sample presented with clinically significant hallucinations (SAPS global hallucinations score ≥ 3) at program entry. These patients had an earlier age of onset, fewer years of education, higher severity of delusions, depression and negative symptoms than those without clinical-level hallucinations. However, they did not significantly differ in DUP or anxiety. These results indicate that clinically significant hallucinations are less prevalent than delusions, but that a higher severity of hallucinations at presentation to an early intervention program is associated with a greater overall symptom burden.

Using brains as sensors of the magnetic fields produced by other brains: Reporting a robust and simpler method

Samuel Calmels1, Enora Jeuland1, J. Bruno Debruille1

1McGill University, Montreal, QC, Canada

Corresponding Author: J. Bruno Debruille, email: bruno.debruille@mcgill.ca


The effects of transcranial magnetic stimulations (TMS) show that the human brain is impacted by some magnetic fields (EMFs). Moreover, after a delay, it produces potentials that reveal a subsequent processing of this impact. The human brain also seems to be sensitive to very weak magnetic fields of extremely low frequencies (vwEMFelf). Namely, to the vwEMelf produced by the brain of other persons when they process visual stimuli. In effect, two previous works report that the event-related brain potentials (ERPs) that are evoked by presenting a picture to a participant can be modulated by simultaneously and privately presenting a picture to a partner. A reprocessing of the data of these works showed that these modulations exist only when the block of trials where these two pictures are the same occurs after the one where they differ. A new experiment was thus run to replicate these results and to control for block order effects. ERPs evoked by presenting, at each trial, the photograph of a face, were recorded. Simultaneously and, again, privately, the same or a different face photograph was presented to the partner in an adjacent room. The ERPs of these partner-participants were found to strongly depend on the sameness of the two photographs (p<0.001), unbeknownst to them. These joint processing effects (JPEs) confirm that a simple and robust method can be used to study the sensitivity of the human brain to the vwEMFelf produced by other brains. This should help future works answering the numerous questions raised this sensitivity.

Adolescent social stress alters the growth of dopamine axons to the prefrontal cortex in a sex-dependent manner

Samuel Richer1,2, Andrea H. Pantoja Urban1,2, Aoran Song3, Sehar Gul3, Giovanni Hernandez2, Cecilia Flores2,4,5

1Integrated Program in Neuroscience, McGill University, Montreal, QC, Canada
2Douglas Mental Health University Institute, Montreal, QC, Canada
3McGill University, Montreal, QC, Canada
4Department of Psychiatry, McGill University, Montreal, QC, Canada
5Department of Neurology and Neurosurgery, McGill University, Montreal, QC, Canada

Corresponding Author: Cecilia Flores, email: cecilia.flores@mcgill.ca


Bullying in adolescence can hamper prefrontal cortex (PFC) development, impair its role in controlling impulse behavior and negatively affect mental health trajectories. In adolescent mice, dopamine (DA) axons expressing the guidance cue receptor DCC undergo targeting events in the nucleus accumbens (NAcc), forming enduring connections. However, mesocortical DA axons cross the NAcc to continue to grow to the PFC during adolescence, remaining vulnerable to insults. Using an accelerated social defeat stress (AcSD) model for adolescent male mice, we showed that social stress downregulates DCC receptors in DA neurons inducing aberrant PFC DA connectivity and inhibitory control deficits in adulthood (Vassilev et al., 2021; 2022). Here, we tested if AcSD triggers mistargeting of NAcc DA axons and their ectopic growth to the PFC. To track adolescent PFC DA axon growth, we used an intersectional viral tracing approach in postnatal day 21 DATCre male C57BL/6J mice whereby eYFP is only expressed in DA neurons that have reached the NAcc by early adolescence. At postnatal day 25, mice were exposed to AcSD or to control conditions. In adulthood, PFC DA axon growth was assessed by quantifying eYFP+ terminals with stereology. AcSD in adolescence show a greater number of eYFP+ axons in the male adult PFC compared to controls whereas females alternatively show a decrease. This is the first demonstration that exposure to physical/psychosocial harm in adolescence can deviate DA axons from their intended target, inducing their input into off-target regions, and likely altering adult cognitive processing.

Regulation of the intracellular accumulation of Tau by Numb and its role in tauopathies

Sarah C. Hales1,2, Marine Lacomme1, Michel Cayouette1,2,3,4

1Institut de recherches cliniques de Montréal, Montreal, QC, Canada
2Integrated Program in Neuroscience, McGill University, Montreal, QC, Canada
3Department of Medicine, Université de Montréal, Montreal, QC, Canada
4Department of Anatomy and Cell Biology, McGill University, Montreal, QC, Canada

Corresponding Author: Michel Cayouette, email: michel.cayouette@ircm.qc.ca


Tauopathies are a large family of neurodegenerative diseases which includes Alzheimer’s disease (AD) and are characterized by dysregulation of the microtubule binding protein Tau, which becomes hyperphosphorylated, resulting in elevated intracellular levels of Tau and formation of aggregates called neurofibrillary tangles. These tangles are known to be toxic to neurons, and therefore understanding how Tau is regulated in health and disease is an important step in better understanding the pathogenesis of tauopathies. Recently, our lab has shown that the long isoform of a well-known endocytic adaptor protein, Numb72, functions as a negative regulator of intracellular Tau levels (Lacomme et al., Sci. Adv., 2022). We found that overexpressing Numb72 in retinal neurons of mouse models of AD and tauopathy significantly reduces Tau accumulation and increases neuronal survival, but whether Numb72 can also function as a neuroprotective agent in neurons outside of the retina remains unknown. To address this question, we generated a new transgenic mouse line that expresses Numb72 in a Cre-dependent manner, allowing us to elevate the levels of Numb72 in various regions of the brain. We hypothesize that overexpression of Numb72 will generally reduce intraneuronal accumulation of Tau and prevent degeneration of hippocampal neurons in tauopathy mouse models, thereby reducing the progression of learning and memory deficits. Overall, this project will help contribute to the development of Numb72 as a novel therapeutic target for tauopathies.

Synaptic dysfunction and altered transcriptome in human iPSC-derived motor neurons with ALS-associated mutations in TARDBP/TDP-43

Sarah Lépine1,2, Angela Nauleau-Javaudin1,3, Gilles Maussion1, Eric Deneault4, Carol X.-Q. Chen1, Narges Abdian1, Anna Krystina Franco-Flores1, Georgina Jiménez Ambriz5, Dan Spiegelman5, Ghazal Haghi1, María José Castellanos-Montiel1, Mathilde Chaineau1, Thomas M. Durcan1

1Early Drug Discovery Unit, The Neuro-Montreal Neurological Institute and Hospital, Department of Neurology and Neurosurgery, McGill University, Montreal, Quebec, Canada
2Faculty of Medicine and Health Sciences, McGill University Montreal, Quebec, Canada
3Faculty of Medicine, Université de Montréal, Montreal, Quebec, Canada
4Centre for Oncology, Radiopharmaceuticals and Research, Biologic and Radiopharmaceutical Drugs Directorate, Health Products and Food Branch, Health Canada, Ottawa, ON, Canada
5The Neuro Bioinformatics Core Facility, The Neuro-Montreal Neurological Institute and Hospital, McGill University, Montreal, Quebec, Canada

Corresponding Author: Thomas M. Durcan, email: thomas.durcan@mcgill.ca


Amyotrophic lateral sclerosis (ALS) is the most common adult-onset motor neuron (MN) disease, characterized by progressive MN loss in the brain and the spinal cord leading to weakness and paralysis that is typically fatal within 2 to 4 years after diagnosis. At the neuropathological level, the cytoplasmic mislocalization and aggregation of the RNA-binding protein TDP-43 is a hallmark of this disease. Furthermore, mutations in the TARDBP gene (coding for TDP-43) have been associated with ALS. Yet, the pathogenic properties of these mutations remain poorly understood. Using CRISPR/Cas9, we generated two homozygous knock-in iPSC lines with mutations in TARDBP encoding TDP-43A382T and TDP-43G348C, two common ALS TDP-43 variants. MNs differentiated from knock-in iPSCs did not display neurodegeneration nor significant changes in TDP-43 subcellular localization or aggregation compared with isogenic control MNs. However, combining multi-electrode array recordings and immunocytochemistry, we found progressive alterations in spontaneous neuronal activity and synapse abnormalities in both mutant MN cultures. Additionally, RNA sequencing experiments revealed significantly overlapping transcriptome alterations in TDP-43A382T and TDP-43G348C MNs, including commonly dysregulated mRNAs, lncRNAs, and microRNAs. Integrated microRNA/mRNA analysis identified several altered microRNAs negatively correlated to the expression levels of their corresponding mRNA targets, suggesting post-transcriptional mechanisms of disease. Interestingly, these microRNAs were predicted to regulate cellular processes related to synaptic function, cell-cell adhesion, and neuronal development. Collectively, our results suggest that MN dysfunction precedes the occurrence of TDP-43 pathology and MN loss in ALS and point to synaptic defects and RNA dyshomeostasis as early events in the pathogenesis of this disease.

Rai1 regulates energy homeostasis through Bdnf-producing neurons and TrkB- signaling

Sehrish Javed1,2, Ya-ting Chang1,2, Yoobin Cho1,2, Yu-Ju Lee1,2, Hao-Cheng Yang1,2, Minza Haque1,2, Edward Lin1,2, Wei-Hsiang Huang1,2

1Department of Neurology and Neurosurgery, Centre for Research in Neuroscience, McGill University, Montreal, QC, Canada
2Brain Repair and Integrative Neuroscience Program, The Research Institute of the McGill University Health Centre, Montreal, QC, Canada

Corresponding Author: Wei-Hsiang Huang , email: wei-hsiang.huang@mcgill.ca


Haploinsufficiency of retinoic acid induced 1 (RAI1), the causative gene for Smith-Magennis Syndrome (SMS), results in severe hyperphagia, metabolic defect, and obesity. Rai1 regulates the transcription of brain-derived neurotropic factor (Bdnf), an important regulator of energy homeostasis in the brain. Bdnf is downregulated in the hypothalamus of SMS mice. This observation led us to further analyze the pathological contribution of Bdnf-producing neurons and the effect of therapeutically targeting Bdnf signalling in SMS. We first performed reverse phase protein analyses (RPPA) and identified dysfunctional biological pathways. Importantly, we discovered neurotrophin signalling pathway to be less active in SMS. We next deciphered the function of Rai1 in a discrete set of Bdnf-producing cells which regulates energy homeostasis through multiple hypothalamic cells including neurons residing in the paraventricular nucleus of hypothalamus (PVH). We generated Rai1 conditional knock out (cKO) model (Rai1-deletion in Bdnf producing neurons) and demonstrated that loss of Rai1 from Bdnf-producing cells contributes to obesity in SMS by selective altering fat deposition in several organs. 3-weeks-old cKO mice also showed reduced neuronal excitability in the PVH. Next, we explored the therapeutic potential of targeting Bdnf downstream signaling by using a partial agonist of TrkB, LM22A-4. Our results demonstrate that LM22A-4 treatment partially rescues the body weight in SMS mice. Moreover, reduction in body weight is followed by improved blood leptin and lipoprotein levels in the treatment group. Our work shows the pathological contribution of Bdnf pathways in SMS and demonstrate that targeting Bdnf signaling has a potential to ameliorate obesity associated with SMS.

Reduced salience network integrity and social cognitive deficits in symptomatic frontotemporal dementia patients

Shanny Foo1, Colleen Hughes1, Alfie Wearn1, Simon Ducharme2, R. Nathan Spreng1, Dave Cash3, Genetic Frontotemporal Dementia Initiative3

1Montreal Neurological Institute, McGill University, Montreal, QC, Canada
2Douglas Mental Health University Institute, McGill University, Montreal, QC, Canada
3Dementia Research Centre, University College London, London, UK

Corresponding Author: R. Nathan Spreng, email: nathan.spreng@mcgill.ca


Introduction: A robust feature characterising frontotemporal dementia (FTD) from other dementias are early deficits in social cognition and atrophy in the insula, a core salience network node. Although resting-state functional connectivity (RSFC) differences between FTD and controls has been examined, little is known about RSFC in presymptomatic FTD. We investigated RSFC differences between non-mutation carriers (NMC), presymptomatic mutation carriers (PMC), and symptomatic mutation carriers (SMC). Methods: Adults (N=981; mean age=49y±14; 55%F, 45%M) of a first-degree relative with a known pathogenic mutation in MAPT, GRN, or C9orf72 were recruited in a large-scale genetic FTD initiative. The sample comprised NMC (n=374), PMC (n=392) and SMC (n=215). A subsample completed social cognition measures and ANCOVA evaluated between-group differences. A common MRI protocol was used across sites and RSFC data were analyzed using partial least squares. Results: Across multiple domains, SMC showed poorer social cognitive ability compared to NMC and PMC, who did not differ from each other. A distributed pattern of RSFC dissociated NMC and PMC from SMC. In contrast to NMC and PMC, SMC exhibited: (1) weaker RSFC within the salience, default, and frontoparietal control networks and, (2) stronger between-network RSFC of the somatomotor network with the dorsal attention, salience and frontoparietal control networks. Conclusions: We provide evidence that symptomatic FTD is associated with social cognition deficits and wide-scale RSFC differences. Speculatively, weaker RSFC between salience and neurocognitive networks central to executive control and decision-making may explain the broad social cognitive impairment is observed FTD. The presymptomatic phase may be best captured using longitudinal changes.

Identifying the shared and distinct cognitive mechanisms underlying mood and behavioural symptoms in Parkinson's disease using a reward learning task

Sophie Sun1, Madeleine E. Sharp1

1McGill University, Montreal, QC, Canada

Corresponding Author: Madeleine Sharp, email: madeleine.sharp@mcgill.ca


Apathy, anhedonia, depression, and impulsivity are prevalent symptoms of Parkinson’s disease (PD) linked to impairments in reward learning and dopamine dysfunction. However, a striking feature of these symptoms is how variably they manifest, despite a link to the dopamine system, which is uniformly affected across patients. One possibility is that dopamine modulates multiple cognitive mechanisms involved in reward learning (beyond reward sensitivity), and that impairments in these processes may be differentially predictive of the different purportedly dopamine-related mood and behavioural symptoms of PD. To address this, we tested 65 PD patients and 37 controls on a standard probabilistic reward learning task. We used a Bayesian hierarchical reinforcement learning-drift diffusion model to extract performance parameters reflecting different processes. Participants also completed questionnaires measuring apathy, anhedonia, depression, and impulsivity. PD patients did not improve in performance over time whereas controls did, but model-derived learning rates did not differ between groups. Separate regressions predicting each mood symptom severity in PD patients revealed that apathy, depression, and impulsivity, but not anhedonia were each negatively associated with positive learning rate. Additionally, impulsivity was negatively associated with drift rate and positively associated with decision boundary. These results suggest that mood symptoms in PD relate to different cognitive mechanisms of reward learning. Future directions involve relating performance parameters and symptoms to neuroimaging measures of neurodegeneration in the dopaminergic system.

Sex differences in timing and degree of alpha-synuclein-induced brain atrophy patterns in a mouse model of synucleinopathy

Stephanie Tullo1, Daniel Gallino2, Janice Park1, Megan Park1, Kristie Mar1, Esther del Cid-Pellitero1, Edward A. Fon1, M. Mallar Chakravarty1,2

1McGill University, Montreal, QC, Canada
2Douglas Mental Health University Institute, McGill University, Montreal, QC, Canada

Corresponding Author: Mallar Chakravarty, email: mallar.chak@gmail.com


Parkinson's disease (PD) is a progressive neurodegenerative disorder with a higher prevalence and a more severe phenotype in men compared to women. Despite this, differences in sex-specifc brain pathology is understudied, and along with these sex differences not being fully understood, the mechanisms underlying PD pathology as a whole has yet to be elucidated although recent evidence suggests aggregated alpha-synuclein (aSyn) propagation may play a key role in synucleinopathy disease progression. Hemizygous mutated aSyn transgenic mice were inoculated with aggregated aSyn (preformed fibrils [PFF]) (or saline) in the right striatum to seed and induce aSyn propagation. Spatial covariance patterns of brain pathology derived from non-negative matrix factorization of voxel-wise volume measures from T1-weighted MRIs (100 μm3 voxels) was performed to examine sex-specific PFF-induced brain pathology patterns at an early (90 days post-injection [dpi]) and a late stage (120 dpi) time point (n>8 mice/group/sex/time point). We observed initial pre-symptomatic neurodegeneration in females with faster post-symptomatic decline and advanced pathology in males. At 90 dpi, sex-specific spatial similarity of PFF-induced pathology points to female-driven pathology, with a higher spatial overlap (62%). Conversely, when examining disease progression (survival rates), male PFF-injected mice succumbed to their symptoms at significantly higher rates than their female counterparts (p=0.0015), supporting the higher male-specific pathology patterns observed at 120 dpi. These findings suggest sex differences may play a significant role in the clinical features and prognosis of synucleinopathy, which could have implications for neuroprotection research and sex-specific analyses in pre-clinical drug trials.

I can’t see you! What about my N400?

Sujata Sinha1,2, Sarah Del Goleto3, Milena Kostova3, J. Bruno Debruille1,2,4

1Department of Neurosciences, Faculty of Medicine, McGill University, Montréal, QC, Canada
2Research Center of the Douglas Mental Health University Institute, Montréal, QC, Canada
3Université Paris, Saint-Denis, France
4Department of Psychiatry, Faculty of Medicine, McGill University, Montréal, QC, Canada

Corresponding Author: J. Bruno Debruille, email: bruno.debruille@mcgill.ca


Schizophrenia patients make more errors and require more time to learn cognitive tasks than healthy individuals when tested by a stranger-experimenter. This lack of cognitive plasticity tends to disappear when they are with someone they trust, such as a caretaker. We aim to see if these differences are due to social context impacts on semantic processes indexed by the N400 event-related brain potential (ERP). Indeed, five recent studies found that the presence of an unknown confederate next to participants could boost their N400 amplitudes. These social-N400 increases occurred when they were asked or not asked to take the confederate's perspective and they knew that the other did not have all stimuli-information as them. These increases appeared even when participants and the confederate accessed the same stimuli. All this suggests that perspective-taking may occur automatically. However, in these set-ups, participants could see the confederate. We examined whether this sight is crucial or whether these social-N400 increases arise even when they are aware that the confederate is out of their sight, seated slightly behind them. For the task, the healthy participants and the confederate decided if the last word of short stories was coherent, equivocal, or incoherent. The N400s elicited by those words were not different between participants with a confederate (n=50) and those who were alone (n=51). Equivocal words elicited similar N400s but larger late posterior positivities (LPPs) than the coherent ones, as in previous studies. Thus, perspective-taking is not automatic, and having the other within participants’ visual periphery is necessary.

THC in adolescence dysregulates the signaling pathway that orchestrates the development of the dopaminergic system

Tanya Capolicchio2,3, Giovanni Hernandez3, Katerina Estrada3, Emilie Dube3, Delaram Shirzad3, Cecilia Flores1,3

1Department of Psychiatry and Department of Neurology and Neurosurgery, McGill University, Montréal, QC, Canada
2Integrated Program in Neuroscience, McGill University, Montréal, QC, Canada
3Douglas Mental Health University Institute, Montreal, QC, Canada

Corresponding Author: Cecilia Flores, email: cecilia.flores@mcgill.ca


Early adolescence is a critical life stage wherein young individuals are honing their ability to make informed choices. Elucidating the mechanisms by which drugs of abuse can enduringly impact their formative choices and behaviors is imperative. Adolescence is a period of dynamic mesocorticolimbic dopamine development, when axons of ventral tegmental area (VTA) dopamine neurons grow from the striatum towards the prefrontal cortex (PFC), ultimately fine-tuning adult PFC function and inhibitory control. This protracted growth is regulated by the Netrin-1 receptor gene, DCC. Here we assessed whether adolescent exposure to the main constituent of cannabis, tetrahydrocannabinol (THC), alters Dcc mRNA and its microRNA repressor, miR-218, in dopamine neurons, and modifies aspects of inhibitory control in adulthood. Male and female mice (PND 22) received intraperitoneal injections of THC (0, 2.5, 5, mg/kg) once every other day, for 10 days. One week later, VTA miR-218 and Dcc levels were quantified. THC alters genes controlling mesocorticolimbic dopamine development in a sex-specific manner. In males, changes in Dcc mRNA are accompanied by altered PFC dopamine connectivity, wherein dopamine axons occupy a smaller span within the PFC and have less varicosities, sites of neurotransmitter release and synapse formation. The effects of THC on adult inhibitory control are dissociable, males show an improvement in action inhibition, and an impairment in waiting impulsivity. However, there is no change in inhibitory control in females treated with THC in early adolescence. Our results suggest that THC exposure in adolescence alters adult inhibitory control in a sex- and impulsivity domain-specific manner.

The effect of obstructive sleep apnea on cognitive functioning among Parkinson’s disease individuals in the Canadian Longitudinal Study on Aging

Gomes Teresa1,2, Benedetti Andrea2,3, Lafontaine Anne-Louise4, Gosselin Nadia5, Postuma Ron6, Kimoff John2,7, Kaminska Marta2,7

1Department of Integrated Program in Neuroscience, McGill University, Montreal, QC, Canada
2Translational Research in Respiratory Diseases, Research Institute of the McGill University Health Centre, Montreal, QC, Canada
3Department of Medicine and Department of Epidemiology, Biostatistics & Occupational Health, McGill University Health Centre, Montreal, QC, Canada
4Montreal Neurological Hospital, McGill University Centre, Montreal, QC, Canada
5Research Center, Centre intégré universitaire de santé et de services sociaux Nord-de-l'Ile-de-Montreal, Montreal, QC, Canada
6Department of Neurology and Neurosurgery, McGill University, Montreal General Hospital, Montreal, QC, Canada
7Respiratory Division and Sleep Laboratory, McGill University Health Centre, Montreal, QC, Canada

Corresponding Author: Marta Kaminska, email: marta.kaminska@mcgill.ca


Introduction: Obstructive sleep apnea (OSA) is associated with cognitive decline in the general older population and with lower cognition in PD patients in clinical cohorts. We aimed to evaluate associations between high risk for OSA and cognition in individuals with Parkinson’s disease (PD) from a population cohort. Methodology: Participants with PD were identified in Canadian Longitudinal Study of Aging (CLSA) comprehensive cohort at baseline or at 3-year follow-up using a validated algorithm. High risk of OSA was determined at the corresponding timepoint using the STOP (Loud Snoring, Tiredness/Sleepiness, Observed apneas and high blood Pressure)>2. Cognitive measures included: the Rey Auditory Verbal Learning Test (RAVLT); Animal Fluency Test (AFT), the Mental Alternation Test (MAT); Controlled Oral Word Association Test (COWAT), Stroop Test (STP), Prospective Memory Test (PMT and TMT), and Choice reaction times (CRT) task. Linear regression was performed to assess relationships between STOP dichotomized and cognitive measures, adjusted for potential cofounders. Results: We identified 150 individuals with PD, 89 individuals at baseline and 61 additional patients at the 3-yr follow-up. Overall, 55 had a high risk of OSA (mean age 70.9 (8.8) years, 72.7% male) and 95 did not (mean age 69.5 (9.2) years, 63.2% male). There was significant association between STOP>2 and cognitive measures MAT (p=0.02) and COWAT (p=0.04) in adjusted analyses. Discussion: High risk of OSA were associated with poorer results on executive function tests. Longitudinal assessment within the CLSA may provide additional insights into the potential impact of OSA on the evolution of cognitive function in PD.

Deciphering Mucopolysaccharidosis III spectrum of diseases through the analysis of hiPSCs-derived neurons

Travis Moore1,2, Poulomee Bose1, Jill Wood3, Thomas Durcan4, Alexey Pshezhetsky1,2

1Centre Hospitalier Universitaire Sainte Justine Research Center, Montreal, QC, Canada
2Department of Anatomy and Cell Biology, McGill University, Montreal, QC, Canada
3Phoenix Nest Inc, Brooklyn, NY, United States
4Montreal Neurological Institute and Hospital, McGill University, Montreal, QC, Canada

Corresponding Author: Alex Pshezhetsky, email: alexei.pchejetski@umontreal.ca


Mucopolysaccharidosis (MPS) III (A-D) is a spectrum of rare inherited neurological lysosomal storage diseases (LSDs) resulting from the deficiency of the enzymes that catabolize heparan sulphate (HS). All MPS III subtypes have a disease course that begins in early childhood and results in premature death due to ineffective treatments. Our lab has extensively studied the MPS IIIC mouse model (HgsnatP304L) to uncover neuron pathophysiological changes. Results have shown several neuronal-specific biomarkers of the disease progression, such as reduction for both the pre-synaptic/post-synaptic markers of excitatory and inhibitory neurons (VGLUT1/PSD-95) and (VGAT/Gephyrin), respectively, along with a decreased brain-derived neurotrophic factor (BDNF). These biomarkers have yet to be analyzed in patient-derived neuronal models of MPS III. Our lab has identified a potential therapeutic approach for the disease, a small molecule drug called AVP6, which positively affects synaptic markers and BNDF. To test the above deficits, we generated human-induced pluripotent stem cells (hiPSCs) from MPS IIIA & C patient fibroblasts and subsequently derived them into cortical-specific neurons. Differentiation from healthy controls and patient hiPSCs into neuronal progenitor cells (NPCs) with final differentiation into cortical neurons showed a robust expression of fidelity and differentiation markers (Nestin/Pax6/Tuj1/Synapsin1/TBR1). NPCs from patient cells showed disease-specific presentations with primary enzymatic deficits and an increased abundance of LAMP2+ area. MPS III-derived neurons had significantly reduced excitatory and inhibitory markers and BDNF puncta. AVP6 treatments displayed an amelioration of most deficits. Together, our results demonstrate a recapitulation of human synaptic deficits and the therapeutic potential of AVP6.

Serum neurofilament light as a diagnostic biomarker to differentiate the behavioural variant of frontotemporal dementia from primary psychiatric disorders

Victoria Light1,2, Sherri Lee Jones2, Elham Rahme1, Katherine Rousseau3, Sterre de Boer4, Charlotte Teunissen4, Yolande Pijnenburg4, Simon Ducharme1,2

1McGill University, Montreal, QC, Canada
2Douglas Research Center, Montreal, QC, Canada
3Institut universitaire en santé mentale de Montréal, Université de Montréal, Montreal, QC, Canada
4University of Amsterdam, Amsterdam, Netherlands

Corresponding Author: Simon Ducharme, email: simon.ducharme@mcgill.ca


Introduction: The behavioural variant of frontotemporal dementia (bvFTD) is challenging to diagnose as it combines behavioral and personality features, overlapping with characteristics of primary psychiatric disorders (PPD). Serum neurofilament light (sNfL) has emerged as a candidate discriminative biomarker, but large-scale studies of sNfL in PPD are lacking. The goal of this study is to use the world’s largest biobank of psychiatric emergency patients, to establish a clinical cutoff score to differentiate PPD from bvFTD. Methodology: PPD patients (n=861) and healthy control (n=69) data were acquired from Biobanque Signature, and bvFTDs (n=82) from both the Amsterdam Medical Center (n=70) and Canadian Consortium on Neurodegeneration in Aging (n=12), all above age 40. sNfL levels were measured using Simoa technology on an HD-X instrument. General linear models were used to test for differences in sNfL between PPD and controls, accounting for covariates. ROC curve analyses were used to determine global and age-specific clinical sNfL cut-offs to distinguish bvFTD from PPD, using the Youden Index. Results: Log-transformed sNfL was slightly higher in PPD versus controls when controlling for age, BMI and diabetes status. ROC curves analyses established a 16.99 pg/mL global cut-off to differentiate bvFTD from PPD (82.9% sensitivity, 74.9% specificity, AUC=.844). A cutoff of 15.952 pg/mL (80% sensitivity, 79.1% specificity, AUC=.864) was determined for ages 40-60. Conclusion: The global and age-specific discriminatory accuracies from ages 40-60 were good, but accuracy decreases after age 60. Clinical applications of this work may improve the accurate diagnosis of bvFTD from PPD up to age 60.

Effect of sex and Apolipoprotein E status on levels of astrocyte biomarkers in Alzheimer’s Disease

Wan Lu Jia1,2,3,4,5,6, Cécile Tissot1,7,8, Nesrine Rahmouni8,9, Thomas K Karikari10, Andrea Lessa Benedet10, Nicholas J. Ashton10, Henrik Zetterberg10, Kaj Blennow11, Pedro Rosa-Neto12

1McGill University, Montreal, QC, Canada
2McGill University Centre for Studies in Aging, Montreal, QC, Canada
3Translational Neuroimaging Laboratory, McGill University Centre for Studies in Aging, Montreal, QC, Canada
4Department of Family Medicine, McGill University, Montreal, QC, Canada
5Division of Geriatric Medicine, McGill University, Montreal, QC, Canada
6Department of Neurology and Neurosurgery, McGill University, Montreal, QC, Canada
7Montreal Neurological Institute, Montreal, QC, Canada
8The McGill University Research Centre for Studies in Aging, Montreal, QC, Canada
9Translational Neuroimaging Laboratory, Montréal, QC, Canada
10University of Gothenburg, Mölndal, Sweden
11Institute of Neuroscience and Physiology, The Sahlgrenska Academy at the University of Gothenburg, Mölndal, Sweden
12Translational Neuroimaging Laboratory, Montreal, QC, Canada

Corresponding Author: Pedro Rosa-Neto, email: pedro.rosa@mcgill.ca


Background: Astrocyte biomarkers of Alzheimer’s disease (AD), such as glial fibrillary acidic protein (GFAP) and chitinase-3 like protein 1 (YKL-40), are both positively associated with multiple neuroinflammation-related proteins. While GFAP levels reflect a response to amyloid-β pathology, YKL-40 mainly demonstrates a response to tau pathology. Cerebrospinal fluid (CSF) GFAP levels showed a significant increase in patients with Alzheimer’s disease compared to cognitively unimpaired individuals, the same as YKL-40. However, there is limited knowledge of the differences in these astrocyte biomarkers by gender and apolipoprotein E (ApoE) status. Methods: Data analysis was effectuated on study participants as part of the Translational Biomarkers in Aging and Dementia (TRIAD) cohort, McGill University, Canada. Participants were from the community or outpatients at the McGill University Research Centre for Studies in Aging. Results: We assessed 33 cognitively unimpaired individuals and 26 cognitively impaired individuals, in relation to Braak stages, 17 individuals have mild cognitive impairment (MCI), and 9 individuals were with AD dementia. Linear regressions were effectuated between GFAP, YKL-40, and ApoE status, sex, and an association between ApoE status and sex. YKL-40 presented a statistically significant association with ApoE status and sex (adjusted r2 = 0.1274, p = 0.03021). Wilcoxon rank tests between sex and ApoE status groups for YKL-40 showed no statistical significance. Conclusion: While GFAP was not found to have a statistically significant association, YKL-40 was found to be associated positively with patient sex and ApoE status. This indicates that sex and APOE impose differential effects across astrocyte biomarkers.

Letter identification by the amblyopic visual system: the role of spatial scrambling

Zhu, Xingqi R.1, Hess, Robert F.1, Baldwin, Alex S.1

1Department of Ophthalmology and Visual Sciences, McGill University, Montreal, QC, Canada

Corresponding Author: Alex S. Baldwin , email: alexander.baldwin@mcgill.ca


Unilateral amblyopia is a disorder of the visual system that arises as a consequence of impoverished input from one eye in childhood. It results in poor vision in the weaker eye. One hallmark of the disorder is a spatial scrambling of visual receptive fields. Our research aims to investigate where that scrambling may occur in the hierarchical organization of receptive fields in the visual system. We adopted a letter identification task and asked participants with amblyopia and control participants with healthy vision to perform the task. We additionally trained convolutional neural networks to do the task on the two types of scrambled letters. These simulated scrambling at either the input or output of the simple cell receptive fields. We extracted relative efficiency for each type of scrambling by comparing human performance against the performance of the neural network. For both participant groups, efficiency for input-scrambled letters was significantly higher than that for output-scrambled letters. Curiously, the amblyopic group had higher efficiency than the control group in the output scrambling condition. This suggests that amblyopic vision is specifically resilient to the effects of that scrambling. We also characterized the mistakes participants made by constructing confusion matrices based on the letter responses made on each trial. Through a k-means clustering analysis, we found that mistakes are less consistent in amblyopic participants. These findings suggest output processing of oriented receptive field differs between normal and amblyopic vision, and that individuals with amblyopia make a distinct pattern of mistakes compared to controls.

Effects of disaster-related prenatal maternal stress on default mode network structural connectivity in young adults: Project Ice Storm

Xinyuan Li1, Xiangyu Long1, Suzanne King1, Catherine Lebel1

1Integrated Program in Neuroscience, McGill University, Montréal, QC, Canada

Corresponding Author: Suzanne King, email: suzanne.king@mcgill.ca


The default mode network (DMN) plays a crucial role in cognition, decision-making and emotion regulation. Studies show that prenatal depression alters the functional organization of the DMN in children. However, it is unknown whether these alterations persist beyond childhood. Examining a natural disaster as the source of prenatal maternal stress, we aimed to determine whether the timing of stress exposure moderates DMN structural connectivity in young adults. Ice storms struck Quebec in January 1998. We recruited women who were within 3 months of conception or who were pregnant during the ice storm, and measured their objective hardship. At 19 years, their young adult offspring underwent diffusion MRI. Graph theory analysis was used to examine DMN organization as it related to prenatal objective hardship controlling for parental socioeconomic status and child sex. Moderation analyses were implemented with bootstrapping of 20,000 resamplings to examine the moderating effect of exposure timing. Significant interactions of timing of exposure and objective hardship were found on global efficiency and nodal degree centrality of the DMN. When stress exposure occurred during preconception and the first month of pregnancy, higher hardship was significantly related to lower global efficiency and lower nodal degree centrality. The result suggests the long-lasting effects of prenatal maternal stress, and highlights preconception and the first month of pregnancy as vulnerable windows for prenatal stress weakening DMN structural connectivity.

Calcium buffering tunes intrinsic excitability of spinal dorsal horn parvalbumin-expressing interneurons: A computational model

Xinyue Ma1,2, Loïs Miraucourt2,3, Haoyi Qiu2,3, Reza Sharif-Naeini1,2,3, Anmar Khadra1,3,4

1Integrated Program in Neuroscience, McGill University, Montreal, QC, Canada
2Alan Edwards Center for Research on Pain, McGill University, Montreal, QC, Canada
3Department of Physiology, McGill University, Montreal, QC, Canada
4Department of Quantitative Life Sciences, McGill University, Montreal, QC, Canada

Corresponding Author: Anmar Khadra and Reza Sharif-Naeini, email: anmar.khadra@mcgill.ca; reza.sharif@mcgill.ca


Parvalbumin-expressing interneurons (PVINs) play a crucial role within the dorsal horn of the spinal cord by preventing touch inputs from activating pain circuits. After nerve injury, their output is decreased via mechanisms that are not fully understood. In this study, we show that PVINs from nerve-injured mice change their firing pattern from tonic to adaptive. To examine the ionic mechanisms responsible for this decreased output, we employed a reparametrized Hodgkin-Huxley (HH) type model of PVINs, which predicted (1) the firing pattern transition is due to an increased contribution of small conductance calcium-activated potassium (SK) channels, enabled by (2) impairment in intracellular calcium buffering systems. Analyzing the dynamics of the HH-type model further demonstrated that a generalized Hopf bifurcation differentiates the two types of state transitions observed in the transient firing of PVINs. Importantly, this predicted mechanism holds true when we embed the PVINs model within the neuronal circuit model of the spinal dorsal horn. To experimentally validate this hypothesized mechanism, we used pharmacological modulators of SK channels and demonstrated that (1) tonic firing PVINs from naïve mice become adaptive when exposed to an SK channel activator, and (2) adapting PVINs from nerve-injured mice return to tonic firing upon SK channel blockade. Our work provides important insights into the cellular mechanism underlying the decreased output of PVINs in the spinal dorsal horn after nerve injury and highlights potential pharmacological targets for new and effective treatment approaches to neuropathic pain.

Effects of 70Hz transcranial alternating current stimulation (tACS) on movement-related brain oscillations and motor task performance

Xuanteng Yan1, Georgios D. Mitsis1, Marie-Hélène Boudrias2

1Department of Bioengineering, McGill University, Montreal, QC, Canada
2School of Physical & Occupational Therapy, McGill University, Montreal, QC, Canada

Corresponding Author: Georgios Mitsis, email: georgios.mitsis@mcgill.ca


Introduction: TACS is a non-invasive technique where a form of continuous sinusoidal current is applied over the scalp. Evidence from local field potential data in animals and computational modeling suggests that the applied sinusoidally varying tACS current can induce temporal alignment of neural firing patterns and is therefore able to entrain oscillatory activity in the stimulated brain area. However, it is still unclear what is going on inside the brain when tACS is applied. In the present study, we investigated the effects of 70Hz tACS on brain oscillations as well as motor performance. Methods: A healthy, young, right-handed female subject participated in this study. The motor task involved 50 times handgrips with her dominant hand. During the task, the participant was required to squeeze the gripper and reach the force level of 15% of her MVC. This motor task was performed during and 15min post-tACS. TACS was applied at 70Hz over the left motor cortex area. Electroencephalography (EEG) and Electromyography (EMG) signals were recorded along the process. Results: Results show that 70Hz tACS significantly reduced motor task error and decreased the reaction time. Additionally, smaller movement-related beta desynchronization (MRBD) was induced when the participant was receiving tACS. Conclusion: In the present study, we investigated the effects of 70Hz tACS on brain oscillations and motor task performance. We found that 70Hz tACS improved motor task performance by reducing motor error and reaction time. As previous studies showed that smaller MRBD was associated with better motor function, our findings provide evidence for the potential use of tACS as a therapeutic tool for motor rehabilitation in the future.

Regional cytoarchitecture tracks cortical network homogeneity and heterogeneity

Yezhou Wang1, Casey Paquola2, Raul Rodriguez Cruces1, Nicole Eichert3, Jessica Royer1, Katrin Amunts2,4, Alan C Evans1, Boris Bernhardt1

1McConnell Brain Imaging Centre, Montreal Neurological Institute and Hospital, McGill University, Montreal, QC, Canada
2Institute of Neuroscience and Medicine, Forschungszentrum Jülich, Jülich, Germany
3Wellcome Centre for Integrative Neuroimaging, Centre for Functional Magnetic Resonance Imaging of the Brain, John Radcliffe Hospital, University of Oxford, Oxford, UK
4C. and O. Vogt Institute for Brain Research, Medical Faculty, University Hospital Düsseldorf, Heinrich Heine University of Düsseldorf, Germany

Corresponding Author: Boris Bernhardt, email: boris.bernhardt@mcgill.ca


Defining the functional units of the human cortex is a key goal of neuroscience Recent studies have derived low-dimensional, continuous representations of cortical organization, also referred to as gradients, using cortex-wide decompositions of functional, microstructural, and structural connectivity features. The current work profiled functional, microstructural, and structural gradient features within a recently proposed probabilistic atlas of cortical cytoarchitecture, and assessed the uniqueness and redundancy of gradient fingerprints across cytoarchitecturally-defined cortical areas. We studied T1-weighted Magnetic Resonance Imaging (MRI), resting-state functional MRI, myelin-sensitive quantitative T1, and diffusion MRI of 50 unrelated healthy adults. We estimated five vertex-wise gradients for three imaging modalities using nonlinear dimensionality reduction techniques. We then used Julich-Brain to assign vertex-wise gradients to cytoarchitecturally-defined regions, defined as gradient profiles. To explore homogeneity of cortical areas, we estimated the cosine similarity of parcel-wise gradient profiles and found higher homogeneity in paralimbic regions and lower homogeneity in idiotypic i.e., sensory and motor cortices. Hierarchical clustering revealed four clusters, indicating high homogeneity within each cluster and high heterogeneity between the four clusters. We calculated the cosine distance between gradient profiles and performed spin-tests to explore the heterogeneity of cortical areas. Visual and primary sensorimotor cortices showed highest values (p<0.05, FDR corrected), while paralimbic network showed minimal heterogeneity. Our findings point to a sensory-paralimbic differentiation of cortex-wide gradient fingerprints, with sensory/motor regions being more heterogenous compared to less distinctive paralimbic cortices. Our work may provide new insights into the neuroanatomical basis of specialized and integrative cortical functions.

Precise mapping of Allen Human Brain Atlas sample coordinates to MNI space

Yohan Yee1, Gabriel A Devenyi1, Yashar Zeighami1, M Mallar Chakravarty1

1Cerebral Imaging Centre, Douglas Mental Health University Institute, McGill University, Montreal, QC, Canada

Corresponding Author: Mallar Chakravarty, email: mallar.chak@gmail.com


The Allen Human Brain Atlas (AHBA) is a widely-used resource in neuroimaging. Consisting of genome-wide mRNA expression sampled at 3702 points (“samples”) across six donor brains, this resource has been extensively used to determine genes and modules/pathways thought to underlie various structural and functional phenotypes observed through neuroimaging. Importantly, these inferences rely on spatial correlations between neuroimaging data and AHBA gene expression and therefore require precise brain coordinates for the acquired samples in a standard space. To this end, the AHBA provides sample locations in MNI space; updated coordinates based on a re-registration of donor brains to an MNI space template are also provided in the alleninf package and re-used in other popular software toolboxes. Here, we show that there is substantial error in these two sets of previously-reported MNI coordinates, and based on these coordinates, many of the 3702 samples are mislocated. We provide new coordinates for these samples, which are derived from multispectral registrations of donor brains to the MNI ICBM152 symmetric 2009c template. We show that our new coordinates provide dramatically better intensity correlations between source (donor) and target (template) brains as compared to previous reported coordinates, implying a better registration and mapping of sample coordinates to MNI space. We further assign samples to atlas-based structures based on their MNI coordinates and (using expert annotations as ground truth) demonstrate that fewer samples are mislabeled when using our new coordinates. These results suggest that inferences based on previously-reported coordinates may be compromised.

Robustness of differential neurophysiological effects between Parkinson’s disease and healthy control groups

Yueyue Sapphire Hou1, Jason da Silva Castanheira1, Alex I. Wiesman1, Sylvain Baillet1

1McConnell Brain Imaging Centre, Montreal Neurological Institute, McGill University, Montreal, Quebec, Canada

Corresponding Author: Sylvain Baillet, email: sylvain.baillet@mcgill.ca


Differential effects in non-invasive magnetoencephalographic (MEG) recordings between clinical and healthy cohorts hold promise as biomarkers of neurological conditions. However, the robustness of these between-group effects against the variable duration of patient recordings needs to be established to ensure they are clinically valid. We examined the minimum recording duration required to obtain robust group differences in the frequency spectrum of resting-state brain activity between patients of Parkinson’s disease (PD) and age-matched healthy controls (HC). We define the time-to-stability indicator (T2S) as the minimal length of data required beyond which less than 1% change is observed in between-group estimates. By simulating realistic MEG time series to instantiate varying sample and effect sizes, we estimated the T2S of four disease-relevant group-wise spectral contrasts. We found that the robust detection of shifts in alpha central frequency required at least 480s consecutive data length, 240s for differences in amplitude at alpha and beta frequency bands, and 300s for detectable differences in aperiodic exponents. Then, by using empirical MEG recordings from PD (N=79) and HC (N=54), we verified that between-group differences in frequency-band amplitude required about 240s to enable robust detection, and amplitude shifts did not stabilize within 390s of the available recordings. Also, the inverse relationship between whole-brain results of T2S and effect size was revealed. In sum, we propose a novel framework for testing the robustness of clinical neurophysiological effects by leveraging both simulated and empirical MEG data. We anticipate this approach will help define guidelines for other clinical research designs using similar data modalities such as EEG.

Indirect regulation of AMPA-mediated EPSP in the cerebellum interneuron in Fragile X Syndrome

Zhe Zhao1, Arjun A. Bhaskaran2, Nils A. Koch1, Erik A. Larson2, Anmar Khadra3, Derek Bowie2

1Integrated Program in Neuroscience, McGill University, Montreal, QC, Canada
2Department of Pharmacology and Therapeutics, McGill University, Montreal, QC, Canada
3Department of Physiology, McGill University, Montreal, QC, Canada

Corresponding Author: Derek Bowie, email: derek.bowie@mcgill.ca


Fragile X syndrome (FXS) is the most common single-gene cause of inherited intellectual disability and autism. Affected individuals and preclinical mouse models show brain hyperexcitability, although the underlying mechanism is poorly understood. Here we show that loss of Fragile X Messenger Ribonucleoprotein (FMRP) alters dendritic signalling and excitability in cerebellar molecular layer interneurons (MLIs - stellate cells). We observed a large but briefer excitatory postsynaptic potential (EPSP) in Fmr1-/- stellate cells (SCs) following excitatory parallel fiber (PF) stimulation. We demonstrate that the change in EPSP halfwidth is caused by an increased GABAergic transmission. The enlarged peak amplitude in Fmr1-/- SCs is due to decreased surface expression of TEA-sensitive delayed rectifier potassium channel (Kdr) currents and a delay in their activation properties. Computational H-H modelling revealed that the dendritic Kdr is critical for EPSP amplitude and hyperexcitability. Pharmacological block by TEA converted the profile of EPSP in WT SCs to that of Fmr1-/- mice further validating the importance of Kdr channel. Interestingly, introduction of an N-terminal fragment of FMRP into Fmr1-/- mice restore the EPSP amplitude and associated hyperexcitability to that of WT mice. Contrary to conventional understanding, our study demonstrates that neuronal signalling deficits in FXS can also include translational independent mechanisms that can be corrected acutely by using a fragment of FMRP.

Can combined treatment with hypothermia and sildenafil improve brain recovery in an animal model of term neonatal encephalopathy?

Zoe Ward1, Ruofan Song1, Jibin Zheng2, Pia Wintermark2,3

1Integrated Program in Neuroscience, Faculty of Medicine, McGill University
2Research Institute of the McGill University Health Centre, McGill University
3Department of Pediatrics, Division of Newborn Medicine, Montreal Children's Hospital, McGill University

Corresponding Author: Pia Wintermark, email: pia.wintermark@mcgill.ca


Objective: To evaluate the potential synergistic effect(s) on brain injury recovery when combining hypothermia treatment (HT) with sildenafil in an animal model of term neonatal encephalopathy. Methods: Rat pups were randomized to four experimental categories: sham-operated pups, HI-untreated pups, HI pups treated with HT, and HI pups treated with HT and sildenafil. HT was administered immediately after HI for 5h at 32ºC. Sildenafil (or vehicle) (50mg/kg) was administered twice daily by oral gavage starting 24h after HI and continued for 7 days. Hemisphere and hippocampal surfaces were measured on H&E-stained sections at P30. Neurons, astrocytes, microglia, and oligodendrocytes were quantified near the infarct boundary zone at P30 using immunohistochemistry. Results: Compared to no treatment or HT alone, combining HT with sildenafil further improved the size of the left hemisphere (p < 0.0001) and hippocampus (p < 0.05), as well as improved the number of cortical neurons (p < 0.01). The combined treatment also further reduced the number of reactive astrocytes and microglia in the cortex (respectively, p < 0.05, and p < .001). Conclusion: Sildenafil in addition to HT may further improve brain injury recovery compared to no treatment or HT alone.

Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.