Patient-specific multi-modal modeling uncovers neurotransmitter receptor involvement in motor and non-motor axes of Parkinson's disease

Multi-systemic neurodegeneration in Parkinson's disease (PD) is increasingly acknowledged, involving several neurotransmitter systems beyond the classical dopaminergic circuit and resulting in heterogeneous motor and non-motor symptoms. Nevertheless, the mechanistic basis of neuropathological and symptomatic heterogeneity remains unclear. Here, we use patient-specific generative brain modeling to identify neurotransmitter receptor-mediated mechanisms involved in PD progression. Combining receptor maps with longitudinal neuroimaging (PPMI data), we detect a diverse set of receptors influencing gray matter atrophy, microstructural degeneration, and dendrite loss in PD. Importantly, identified receptor mechanisms correlate with symptomatic variability along two distinct axes, representing motor/psychomotor symptoms with large GABAergic contributions, and cholinergically-driven visuospatial dysfunction. Furthermore, we map cortical and subcortical regions where receptors exert significant influence on neurodegeneration. Our work constitutes the first personalized causal model linking the progression of multi-factorial brain reorganization in PD across spatial scales, including molecular systems, accumulation of neuropathology in macroscopic brain regions, and clinical phenotypes.