How a Whole-Brain Computer Model Links Acetylcholine Receptor Density to Neural Dynamics

For decades, large-scale brain models treated every cortical region as if it worked the same way. A team of researchers has now built a computer model that breaks with that simplification, linking the brain's molecular chemistry to its whole-brain patterns of activity. The study, published in the Proceedings of the National Academy of Sciences, provides new evidence that regional differences in receptor density help shape how activity and information move across the brain .

How the Model Was Built

The researchers built the model by integrating empirical PET-derived maps of muscarinic acetylcholine receptor density across 68 cortical regions into The Virtual Brain (TVB), an open-source whole-brain simulation platform . They layered these spatially heterogeneous receptor maps onto the brain's actual structural connections (the human connectome) and used biophysically grounded mean-field models to simulate neural dynamics across states ranging from wakefulness to sleep .

Instead of assuming uniform cortical properties, each of the 68 regions had its own receptor-density profile, allowing the model to capture how identical chemical signals produce distinct network effects depending on local molecular architecture . This approach builds on a growing body of research showing that neurotransmitter receptor distributions are not random but follow organized spatial gradients across the cortex .

Key Findings

Enhanced Synchronization and Information Flow

Incorporating realistic receptor heterogeneity significantly improved interareal functional connectivity and information routing compared to a model where all regions behaved identically . This biological variability supports more flexible and coordinated brain states . The finding aligns with earlier work showing that hierarchical heterogeneity of local synaptic strengths improves model fit to fMRI-measured resting-state functional connectivity .

Spontaneous Localized Slow Waves

Perhaps the most striking result: the model naturally reproduced a real-world brain phenomenon — sleep-like slow oscillations appearing in specific cortical regions while the rest of the cortex remained in an awake-like state . This mimics what occurs during attentional lapses, sleep deprivation, and around brain lesions . The finding is consistent with prior research showing that low levels of acetylcholine during sleep drive microcircuit activity into slow oscillations and network synchrony .

Region-Specific Neuromodulation

Acetylcholine does not act uniformly across the brain. Its effects depend on where its receptors are concentrated, meaning the same chemical signal can produce different network dynamics in different cortical areas . High ACh concentrations, such as those during wakefulness or attentional tasks, desynchronize network activity, while low ACh levels allow synchronized slow-wave patterns to emerge .

A Bridge Between Scales

The study provides a concrete computational framework connecting microscopic molecular chemistry (receptor distribution) to macroscopic, brain-wide functional activity . The authors suggest this approach could help explain state transitions in disorders of consciousness, focal brain injuries, and neurodegenerative diseases .

Why This Matters

Most large-scale brain models simplify things by treating every cortical region as if it worked the same way . This new model takes a different approach: it incorporates detailed maps of muscarinic acetylcholine receptor densities, creating a more realistic simulation of how the brain might transition between conscious and unconscious states .

The model was developed as part of The Virtual Brain Twin Project and the EBRAINS 2.0 research infrastructure . The work was led by Leonardo Dalla Porta at the Institut d'Investigacions Biomèdiques August Pi i Sunyer (IDIBAPS), in collaboration with researchers at CEITEC - Central European Institute of Technology .