Adolescent maturation of cortical microcircuits based on individualized biophysical network modeling
Poster No:
1236
Submission Type:
Abstract Submission
Authors:
Amin Saberi1,2,3, Kevin Wischnewski1, Kyesam Jung1, Leon Lotter1, Lina Schaare2, Juergen Dukart1, Oleksandr Popovych1, Simon Eickhoff1, Sofie Valk2
Institutions:
1INM-7, Research Centre Jülich, Jülich, Germany, 2Max Planck Institute for Human Cognitive and Brain Sciences, Leipzig, Germany, 3Institute of Systems Neuroscience, Heinrich Heine University Düsseldorf, Düsseldorf, Germany
First Author:
Amin Saberi
INM-7, Research Centre Jülich|Max Planck Institute for Human Cognitive and Brain Sciences|Institute of Systems Neuroscience, Heinrich Heine University Düsseldorf
Jülich, Germany|Leipzig, Germany|Düsseldorf, Germany
INM-7, Research Centre Jülich|Max Planck Institute for Human Cognitive and Brain Sciences|Institute of Systems Neuroscience, Heinrich Heine University Düsseldorf
Jülich, Germany|Leipzig, Germany|Düsseldorf, Germany
Co-Author(s):
Introduction:
Adolescence is a critical period of development with substantial macro- and microscale changes in the brain, including maturation of cortical microcircuits and the excitation-inhibition (E-I) balance [1, 2]. However, the in vivo and non-invasive assessment of E-I is a challenge limiting the extent of current available evidence in humans [3, 4]. Biophysical network modeling (BNM) is a promising approach that can bridge the functional imaging data at the macroscale to the hidden features of cortical microcircuits at the microscale [5, 6]. Here, we used individualized BNM on magnetic resonance imaging data to study typical E-I maturation and its association to psychopathology throughout adolescence.
Methods:
We studied adolescents from the cross-sectional Philadelphia Neurodevelopmental Cohort (PNC; N = 764, 421 female, age: 15.3±2.6 [10-19]) and the longitudinal IMAGEN dataset (N = 148, 74 female) with follow-ups from age 14 to 19. The imaging data was processed to calculate individual structural connectomes (SC), empirical functional connectomes (FC), and empirical functional connectivity dynamics (FCD) matrices.
Next, we performed BNM simulation-optimization at the level of each individual (Fig. 1). We simulated the activity of 100 cortical regions of the Schaefer atlas as network nodes governed by the reduced Wong-Wang model with feedback inhibition control [7, 8] and the Balloon-Windkessel model for calculation of simulated BOLD signals [9]. The model was controlled by 15 free parameters, including global coupling (G), in addition to bias and coefficient terms which determine the regional excitatory-to-excitatory (wEE) and excitatory-to-inhibitory (wEI) connection weights as a combination of six biological maps (T1w/T2w, cortical thickness, FC principal gradient, gene expression principal axis, NMDA-R and GABAA PET maps). The goodness-of-fit of the simulations to the empirical functional data of each subject was defined as the correlation of FC matrices subtracted by the absolute difference in their means and Kolmogorov-Smirnov distance of FCD matrices (Fig. 1A). Model optimization was performed using covariance matrix adaptation evolution strategy (Fig. 1B). We subsequently used the optimal simulation of each subject to calculate the mean simulated excitatory firing rates (<rE>) as regional markers of E-I balance (Fig. 1C). Following, we studied the effect of age on <rE> through adolescence. We assessed spatial correlation of age effects derived from different datasets or conditions through using spin-permutated surrogates. Last, we explored potential alterations of E-I maturation associated with psychopathology by evaluating age-by-group interaction effects on <rE> in typical and atypical developing subgroups, as defined by Global Assessment Scale ratings (PNC) or presence of DSM diagnoses (IMAGEN).
Next, we performed BNM simulation-optimization at the level of each individual (Fig. 1). We simulated the activity of 100 cortical regions of the Schaefer atlas as network nodes governed by the reduced Wong-Wang model with feedback inhibition control [7, 8] and the Balloon-Windkessel model for calculation of simulated BOLD signals [9]. The model was controlled by 15 free parameters, including global coupling (G), in addition to bias and coefficient terms which determine the regional excitatory-to-excitatory (wEE) and excitatory-to-inhibitory (wEI) connection weights as a combination of six biological maps (T1w/T2w, cortical thickness, FC principal gradient, gene expression principal axis, NMDA-R and GABAA PET maps). The goodness-of-fit of the simulations to the empirical functional data of each subject was defined as the correlation of FC matrices subtracted by the absolute difference in their means and Kolmogorov-Smirnov distance of FCD matrices (Fig. 1A). Model optimization was performed using covariance matrix adaptation evolution strategy (Fig. 1B). We subsequently used the optimal simulation of each subject to calculate the mean simulated excitatory firing rates (<rE>) as regional markers of E-I balance (Fig. 1C). Following, we studied the effect of age on <rE> through adolescence. We assessed spatial correlation of age effects derived from different datasets or conditions through using spin-permutated surrogates. Last, we explored potential alterations of E-I maturation associated with psychopathology by evaluating age-by-group interaction effects on <rE> in typical and atypical developing subgroups, as defined by Global Assessment Scale ratings (PNC) or presence of DSM diagnoses (IMAGEN).
Results:
In the PNC dataset, <rE> significantly decreased with age in transmodal areas but increased in unimodal regions (Fig. 2A). This pattern was replicable across random subsamples (Fig. 2A) and was robust to age-related variations of SC, as similar age effects were observed with group-averaged SC (r = 0.82, p < 0.01). Longitudinal changes of <rE> from age 14 to 19 in IMAGEN showed a pattern largely similar to PNC (r = 0.67, p < 0.01; Fig. 2B). The age-by-group interaction effects, testing for associations between E-I maturation and psychopathology, were not significant in both datasets in any node after FDR correction (Fig. 2).
Conclusions:
We find E-I ratio during adolescence to decrease in association cortices and increase in unimodal regions. This corroborates previous animal and human studies showing a decreased E-I ratio in association areas [3, 4, 10]. Individualized modeling holds promise for further interrogation of the interrelationship between microcircuit maturation and cognitive as well as clinical variation.
Lifespan Development:
Early life, Adolescence, Aging 1
Modeling and Analysis Methods:
fMRI Connectivity and Network Modeling 2
Keywords:
Computational Neuroscience
Development
FUNCTIONAL MRI
WHITE MATTER IMAGING - DTI, HARDI, DSI, ETC
1|2Indicates the priority used for review
Provide references using author date format
2. Caballero A, Orozco A, Tseng KY. Developmental regulation of excitatory-inhibitory synaptic balance in the prefrontal cortex during adolescence. Semin Cell Dev Biol. 2021;118:60–63.
3. Larsen B, Cui Z, Adebimpe A, Pines A, Alexander-Bloch A, Bertolero M, et al. A developmental reduction of the excitation:inhibition ratio in association cortex during adolescence. Science Advances. 2022. February 2022. https://doi.org/10.1126/sciadv.abj8750.
4. Ghisleni C, Bollmann S, Poil S-S, Brandeis D, Martin E, Michels L, et al. Subcortical Glutamate Mediates the Reduction of Short-Range Functional Connectivity with Age in a Developmental Cohort. J Neurosci. 2015;35:8433–8441.
5. Popovych OV, Manos T, Hoffstaedter F, Eickhoff SB. What Can Computational Models Contribute to Neuroimaging Data Analytics? Frontiers in Systems Neuroscience. 2019;12:68.
6. Deco G, Kringelbach ML. Great Expectations: Using Whole-Brain Computational Connectomics for Understanding Neuropsychiatric Disorders. Neuron. 2014;84:892–905.
7. Deco G, Ponce-Alvarez A, Hagmann P, Romani GL, Mantini D, Corbetta M. How local excitation-inhibition ratio impacts the whole brain dynamics. J Neurosci. 2014;34:7886–7898.
8. Demirtaş M, Burt JB, Helmer M, Ji JL, Adkinson BD, Glasser MF, et al. Hierarchical Heterogeneity across Human Cortex Shapes Large-Scale Neural Dynamics. Neuron. 2019;101:1181-1194.e13.
9. Friston KJ, Harrison L, Penny W. Dynamic causal modelling. Neuroimage. 2003;19:1273–1302.
10. Hoftman GD, Datta D, Lewis DA. Layer 3 excitatory and inhibitory circuitry in the prefrontal cortex: Developmental trajectories and alterations in schizophrenia. Biological Psychiatry. 2017;81:862.