The causal roles of glutamate and GABA genes on brain and behavior in autism
Poster No:
419
Submission Type:
Abstract Submission
Authors:
Viola Hollestein1, Tom Claassen2, Jan Buitelaar3, Nicolaas Puts4
Institutions:
1Donders Institute, Nijmegen, N/A, 2Donders Institute, Nijmegen, Gelderland, 3Donders Institute for Brain, Cognition, and Behavior, Radboud University Nijmegen, Nijmegen, NL, Nijmegen, Netherlands, 4Institute of Psychiatry, London, N/A
First Author:
Co-Author(s):
Jan Buitelaar
Donders Institute for Brain, Cognition, and Behavior, Radboud University Nijmegen, Nijmegen, NL
Nijmegen, Netherlands
Donders Institute for Brain, Cognition, and Behavior, Radboud University Nijmegen, Nijmegen, NL
Nijmegen, Netherlands
Introduction:
One of the most influential theories of the underlying mechanisms of autism suggests that an imbalance between excitation and inhibition (E/I) in the brain causes behavioral differences in autism. However, how these E/I differences arise, and how these relationships link to brain function and differ across behavioral characteristics is not well understood. Understanding these relationships is important for developing more targeted support options. Using large multimodal datasets we aimed to infer probable causal relationships between the genetic underpinnings of glutamate (excitation) and GABA (inhibition) communication pathways in the brain, functional activity during inhibitory control, and several behavioral measures of autism. We further examined whether these links were mediated by other variables.
Methods:
We used two samples. First, the discovery sample was the Longitudinal European Autism Project (LEAP) cohort, part of the AIMS-2-TRIALS clinical research programme (https://www.aims-2-trials.eu/) [1] consisting of 638 participants (autistic = 359, non-autistic = 279), aged 6-30 years old. Second, the replication sample was the TACTICS cohort (www.tactics-project.eu), including 164 participants (autistic = 60, non-autistic = 104), aged 8-13 years old.
We selected gene-sets of glutamate and GABA communication pathways in the brain using Ingenuity Pathway Analysis software (http://www.ingenuity.com/). With these, we calculated individual's polygenic score for autism based on the glutamate and GABA gene-sets, using PRSet in PRSice [2].
In both cohorts, functional MRI during inhibitory control was measured, and successful and failed inhibitory control contrasts created. In TACTICS, MR Spectroscopy measures of glutamate concentrations in striatum and ACC were also available. We used the placement of the MRS voxels in ACC and striatum to extract beta values of the successful and failed inhibitory control contrasts in these specific regions of interest using MarsBar [3].
We included behavioral measures capturing what is typically called 'core characteristics' of autism; repetitive behaviors, social behaviors and sensory processing, which were measured through questionnaires [4-6]. In the autistic participants the diagnostic interviews ADI-R [7] and ADOS-2 [8] were also included as measures of autistic traits. Additionally, to account for potential influence of the most common co-occurring conditions (ADHD, anxiety, depression), questionnaire measures of these were included.
To investigate direct and indirect causal relationships between all these observational measures we used Bayesian Constraint-based Causal Discovery (BCCD) algorithms [9]. This method combines the strengths of constraint-based methods giving clear causal relationships, and of score-based methods estimating confidence measures of the inferred causal relationships. The output is a graphically presented model of the causal structure, reporting on estimated reliability of inferred causal relationships.
We selected gene-sets of glutamate and GABA communication pathways in the brain using Ingenuity Pathway Analysis software (http://www.ingenuity.com/). With these, we calculated individual's polygenic score for autism based on the glutamate and GABA gene-sets, using PRSet in PRSice [2].
In both cohorts, functional MRI during inhibitory control was measured, and successful and failed inhibitory control contrasts created. In TACTICS, MR Spectroscopy measures of glutamate concentrations in striatum and ACC were also available. We used the placement of the MRS voxels in ACC and striatum to extract beta values of the successful and failed inhibitory control contrasts in these specific regions of interest using MarsBar [3].
We included behavioral measures capturing what is typically called 'core characteristics' of autism; repetitive behaviors, social behaviors and sensory processing, which were measured through questionnaires [4-6]. In the autistic participants the diagnostic interviews ADI-R [7] and ADOS-2 [8] were also included as measures of autistic traits. Additionally, to account for potential influence of the most common co-occurring conditions (ADHD, anxiety, depression), questionnaire measures of these were included.
To investigate direct and indirect causal relationships between all these observational measures we used Bayesian Constraint-based Causal Discovery (BCCD) algorithms [9]. This method combines the strengths of constraint-based methods giving clear causal relationships, and of score-based methods estimating confidence measures of the inferred causal relationships. The output is a graphically presented model of the causal structure, reporting on estimated reliability of inferred causal relationships.
Results:
In our discovery sample we found likely, direct, causal interactions with Bayesian statistical reliability of at least 80% probability between glutamate polygenic scores and ADI-R (see Figure 1), which was subsequently replicated in the TACTICS replication sample. The fMRI derived beta contrasts were causally linked to each other, but not to other behavioral or genetic measures.
Conclusions:
Our results suggest a direct link between glutamate communication pathways in the brain and autistic characteristics measured in the ADI-R. This was replicated in an independent cohort, further strengthening the confidence of these results. This gives strong evidence for differences in the balance between excitation and inhibition in the brain causing autism characteristics. This is one of the first in vivo human studies suggesting causal links between the genetics of E/I (glutamate and GABA) to differentially underlie autism phenotypes.
Disorders of the Nervous System:
Neurodevelopmental/ Early Life (eg. ADHD, autism) 1
Genetics:
Genetics Other 2
Higher Cognitive Functions:
Executive Function, Cognitive Control and Decision Making
Novel Imaging Acquisition Methods:
BOLD fMRI
MR Spectroscopy
Keywords:
Autism
Computational Neuroscience
FUNCTIONAL MRI
GABA
Glutamate
Magnetic Resonance Spectroscopy (MRS)
MR SPECTROSCOPY
Other - Bayesian Constraint based Causal Discovery (BCCD)
1|2Indicates the priority used for review
Provide references using author date format
[2] Shing Wan Choi, Paul F O'Reilly (2009). PRSice-2: Polygenic Risk Score software for biobank-scale data, GigaScience, 8(7).
[3]Brett, M., Anton, J.-L., Valabregue, R., Poline, J.-B., 2002. Region of interest analysis using an SPM toolbox. NeuroImage 2002 (16), 497.
[4] Bodfish JW, Symons FJ, Parker DE, Lewis MH. Varieties of repetitive behavior in autism: comparisons to mental retardation. J Autism Dev Disord. 2000. https:// doi.org/10.1007/978-1-4419-1698-3_894
[5] Constantino J, Gruber C. Social responsiveness scale 2nd ed. Los Angeles: Western Psychilogical Services; 2002.
[6] Tomchek SD, Dunn W. Sensory processing in children with and without Autism: a comparative study using the short sensory profile. Am J Occup Ther. 2007; 61:190–200.
[7] Rutter M, Le Couteur A, Lord C. Autism diagnostic interview-revised. West Psychol Serv. 2003. https://doi.org/10.1007/978-1-4419-1698-3_894
[8] Lord C, Risi S, Lambrecht L, Cook EH, Leventhal BL, DiLavore PC, et al. The Autism diagnostic observation schedule–generic: a standard measure of social and communication deficits associated with the spectrum of autism. J Autism Dev Disord. 2000;19:205–23.
[9] Sokolova, E., Groot, P., Claassen, T., & Heskes, T. (2014). Causal discovery from databases with discrete and continuous variables. In European Workshop on Probabilistic Graphical Models (442-457). Springer, Cham.