Molecular, cellular and network mapping of brain structural deviations in patients with Post-COVID19 syndrome

Abstract Post-COVID-19 syndrome encompasses persistent cognitive, neurological, and psychiatric symptoms following SARS-CoV-2 infection, profoundly affecting global quality of life. Clarifying the neurobiological basis of these symptoms is vital for effective therapeutic interventions. This study utilized normative modelling of brain structure (“CentileBrain”) to quantify individualized deviations in cortical thickness, surface area, and subcortical volumes among 20 patients experiencing persistent fatigue following mild COVID-19, compared to 20 matched healthy controls. Group-level analyses on deviation scores revealed subtle yet distinct regional alterations in cortical thickness, specifically decreased thickness within orbitofrontal cortices and increased thickness in occipital/sensory cortices. Although at the individual regional level, the proportion of patients exhibiting infranormal or supranormal thickness values was relatively low (<35%) and comparable to controls, deviations frequently clustered within structurally connected circuits, affecting up to 50% more of patients. Spatial analysis of regional cortical thickness alterations correlated significantly with the constitutive expression patterns of TMPRSS2, an essential protein facilitating SARS-CoV-2 cellular entry. Canonical correlation analyses further identified specific cell-type distributions and neuroreceptor densities predictive of regional thickness changes, highlighting neurons and molecular targets associated with serotoninergic, cannabinoid, cholinergic, and glutamatergic signalling pathways. Network-diffusion modelling constrained by a canonical structural connectome significantly outperformed null models based on permuted connectomes and Euclidean distance metrics, identifying posterior-parietal regions as probable initiation points (“seeds”) for network-wide structural changes. Seed likelihood correlated positively with TMPRSS2 expression levels, suggesting that these posterior-parietal regions may be particularly susceptible to SARS-CoV-2 infection. This highlights a plausible mechanism where structural alterations could propagate through connected neural networks, although direct evidence of such propagation requires further investigation. These findings provide novel insights into potential mechanisms underlying neural circuit disruptions in post-COVID-19 fatigue and suggest avenues for therapeutic neuromodulation. Highlights Normative modelling revealed heterogenous individualised cortical structural deviations in patients with PCS. Regional deviations in cortical thickness and surface area clustered within anatomically connected circuits, affecting up to 50% more of PCS participants. Structural changes were significantly associated with regional TMPRSS2 expression, implicating SARS-CoV-2 entry pathways in PCS-related brain alterations. Network diffusion modelling identified posterior-temporoparietal regions as likely epicentres of pathology propagation through the structural connectome. Cortical deviations were linked to neuronal and microglial cell-type densities and receptor systems including serotonergic, glutamatergic, cannabinoid and cholinergic pathways, suggesting targets for neuroimmune-informed interventions.