Neurophysiological consequences of synapse loss in progressive supranuclear palsy

Abstract

Synaptic loss occurs early in many neurodegenerative diseases and contributes to cognitive impairment even in the absence of gross atrophy. Currently, for human disease there are few formal models to explain how cortical networks underlying cognition are affected by synaptic loss. We advocate that biophysical models of neurophysiology offer both a bridge from clinical to preclinical models of pathology, and quantitative assays for experimental medicine. Such biophysical models can also disclose hidden neuronal dynamics generating neurophysiological observations like electro- and magneto-encephalography (MEG). Here, we augment a biophysically informed mesoscale model of human cortical function by inclusion of synaptic density estimates as captured by [11C]UCB-J positron emission tomography, and provide insights into how regional synapse loss affects neurophysiology. We use the primary tauopathy of progressive supranuclear palsy (Richardson's syndrome) as an exemplar condition, with high clinicopathological correlations. Progressive supranuclear palsy causes a marked change in cortical neurophysiology in the presence of mild atrophy and is associated with a decline in cognitive functions associated with the frontal lobe. Using (parametric empirical) Bayesian inversion of a conductance-based canonical microcircuit model of MEG data, we show that the inclusion of regional synaptic density - as a subject-specific prior on laminar specific neuronal populations - markedly increases model evidence. Specifically, model comparison suggests that a reduction in synaptic density in inferior frontal cortex affects superficial and granular layer glutamatergic excitation. This predicted individual differences in behaviour, demonstrating the link between synaptic loss, neurophysiology, and cognitive deficits. The method we demonstrate is not restricted to progressive supranuclear palsy or the effects of synaptic loss: such pathology-enriched dynamic causal models can be used to assess the mechanisms of other neurological disorders, with diverse non-invasive measures of pathology, and is suitable to test the effects of experimental pharmacology.

Reliability of dynamic causal modelling of resting‐state magnetoencephalography

“…The laminar interpretation of the conductance‐based model (e.g., superficial, deep geometry) is supported by (i) specification of priors for intrinsic connections (Table 1 ), (ii) the equation of the observer for MEG data, and (iii) endogenous inputs to the model which targets spiny stellate cells in layer four. This model parametrisation supports inferences about the laminar basis of degenerative brain neurological disorders (Adams et al, 2022 ; Adams, Hughes, et al, 2021 ; Adams, Pinotsis, et al, 2021 ; Shaw et al, 2021 ).…”

“…The use of PEB can be seen as an extension of BMR to cohort studies. PEB is well suited to address whether model evidence from cohort data can be improved by replacing non‐informative (or weakly informative) prior parameters in DCM by empirical priors; that is, empirical constraints (Adams et al, 2022 ; Adams, Pinotsis, et al, 2021 ; Friston et al, 2022 ; Jafarian et al, 2021 ; Jafarian, Hughes, et al, 2023 ). In PEB, a reduced model with higher model evidence is sought at the group level by constraining parameters using group information to enhance cohort model evidence.…”

“…Dynamic causal modelling (DCM) has been used widely in translational neuroscience to elucidate the underlying causes of physiological observations including electro/magnetoencephalography (MEG) (Adams et al, 2022 ; Adams, Hughes, et al, 2021 ; Gilbert et al, 2016 ; Jafarian, Hughes, et al, 2023 ; Shaw et al, 2017 ; Shaw et al, 2021 ). However, the use of imaging and analytical methods to reveal the effects of disease progression and treatment intervention rests on reliability.…”

“…However, with complex models—with high posterior covariance among parameters—the reliability of the expected value of particular parameters can be very low (Frässle & Stephan, 2022 ; Rowe, 2010 ; Rowe et al, 2010 ; Schuyler et al, 2010 ). The reliability of DCM parameters has also been examined using split sampling over evoked potentials (Adams et al, 2022 ) and resting‐state MEG (Jafarian, Hughes, et al, 2023 ). However, the frequentist approach to reliability is ill‐suited to complex nonlinear dynamic systems with posterior covariance among model parameters.…”

Reliability of dynamic causal modelling of resting state magnetoencephalography

“…However, with complex models -with high posterior covariance among parameters -the reliability of the expected value of particular parameters can be very low (Schuyler et al, 2010, Frässle and Stephan, 2022, Rowe, 2010. The reliability of DCM parameters has also been examined using split-sampling over evoked potentials (Adams et al, 2022) and resting state MEG (Jafarian et al, 2023). However, the frequentist approach to reliability is ill-suited to complex nonlinear dynamic systems with posterior covariance among model parameters.…”