🧬 biology

Simulating the Effects of Gamma Oscillation Treatment on Selective Neurotransmitters Using The Virtual Brain and Complex Pathway Simulator

This study utilized The Virtual Brain and COPASI to simulate gamma oscillation treatment in an Alzheimer's disease model, revealing a promising therapeutic effect characterized by increased acetylcholine and glutamate levels.

Original authors: Brenden Nago, Collin Qu

Published 2026-09-30

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Brenden Nago, Collin Qu

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ⚕️ This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer

The human brain is a vast, humming network of billions of cells that communicate by sending chemical signals back and forth. In a healthy mind, these signals are balanced, allowing us to think, remember, and move with ease. But in Alzheimer's disease, this delicate balance breaks down. The brain begins to lose specific chemicals that help cells talk to one another, while other chemicals become dangerously overactive, causing damage to the very cells they are meant to support. For decades, doctors have tried to fix this by giving patients drugs that either boost the missing chemicals or block the harmful ones, but these treatments often offer only temporary relief. Recently, scientists have turned their attention to a different kind of therapy: using rhythmic flashes of light or sound to nudge the brain into a specific electrical rhythm. This rhythm, known as a gamma oscillation, is a fast, steady pulse that the brain naturally uses when it is focused or remembering things. The hope is that by forcing the brain to vibrate at this speed, we might be able to reset its internal chemistry and slow the disease.

A team of researchers set out to see if this idea could work on a computer before trying it on people. They built a detailed digital model of a human brain, one that was programmed to show the signs of Alzheimer's, specifically the stage where memory loss is noticeable but not yet severe. To make this model as realistic as possible, they fed it real data from actual patients, including scans of their brain structure and how the different parts were connected. Once the virtual brain was ready, the researchers applied a treatment that mimicked the gamma rhythm. They sent a steady, 40-hertz electrical pulse to the parts of the brain responsible for hearing and seeing, keeping this stimulation going for a long period to see how the system reacted. The goal was to watch what happened to the two key chemicals: the one that is usually missing in Alzheimer's patients, and the one that is usually causing trouble.

The results of the simulation were immediate and striking. As soon as the rhythmic pulse began, the levels of both chemicals rose sharply. The chemical that is typically depleted in Alzheimer's patients increased, which is exactly what doctors hope to achieve with any new treatment. However, the simulation also showed that the harmful chemical, which is usually too high in the disease, also went up. This created a complex picture. The researchers found that the brain's electrical activity had become more excited overall, causing both types of signals to spike at the same time. The model showed that within a fraction of a second, the levels of these chemicals climbed from their starting points to a new, higher plateau and stayed there for the duration of the test.

This finding suggests that the treatment acts as a powerful short-term booster for the brain's entire network, rather than a precise switch that turns only the good chemicals on and the bad ones off. The rise in the helpful chemical is promising because it could improve how brain cells communicate. Yet, the simultaneous rise in the harmful chemical raises a question about safety, since too much of that substance can damage brain cells. The researchers noted that their study was a simulation, a mathematical experiment run on a computer, and it did not prove that this would happen in a living human body over a long period. They observed that the brain model reacted quickly to the stimulus, but they could not determine if these changes would last or if they would lead to permanent healing.

The study concludes that while this rhythmic stimulation can successfully trigger a rapid increase in the brain's chemical activity, the outcome is a double-edged sword. It shows that the brain can be coaxed into a more active state, but it also warns that this activity is broad and affects everything at once. The researchers suggest that future work needs to figure out how to keep the benefits of the treatment while avoiding the potential risks of over-stimulating the brain. For now, this computer model provides a clear map of what happens in the first few moments of the therapy, offering a starting point for scientists to design better treatments that might one day help patients with Alzheimer's disease.

Technical Summary: Simulating the Effects of Gamma Oscillation Treatment on Selective Neurotransmitters

Problem Statement
Alzheimer's Disease (AD) is characterized by neuronal injury and altered neurotransmitter production, specifically involving a deficit in acetylcholine and an elevation of glutamate. While current pharmacological treatments are limited, 40 Hz gamma oscillation therapy has emerged as a novel experimental intervention showing potential to reduce amyloid-beta plaques. However, the long-term effects of this treatment on specific neurotransmitter concentrations and the underlying network mechanisms remain unknown. This study addresses the gap in understanding how gamma oscillation therapy influences acetylcholine and glutamate levels within the context of AD pathology.

Methodology
The research employed a multi-stage computational pipeline integrating neuroimaging data, large-scale neural modeling, and kinetic simulation:

Data Acquisition and Preprocessing:
- Resting-state fMRI, T1-weighted MRI, and diffusion tensor imaging (DTI) data were sourced from the Alzheimer's Disease Neuroimaging Initiative (ADNI).
- The cohort was filtered to include only subjects with Late Mild Cognitive Impairment (LMCI).
- Images were converted to NIfTI and MRtrix3 formats. T1-weighted data underwent segmentation and parcellation using Freesurfer's recon-all pipeline, while diffusion data were processed via MRtrix3 (denoising, motion correction, and probabilistic tractography using the iFOD2 algorithm).
Network Simulation (The Virtual Brain - TVB):
- Structural connectome matrices and node coordinates were imported into TVB.
- Model: The Jansen–Rit neural mass model was utilized to simulate brain network dynamics.
- Stimulus: A 40 Hz sinusoidal regional stimulus was applied to the auditory and visual cortices for 100,000 ms.
- Parameters: Key settings included an excitatory synaptic gain (A) of 3.25 mV, an inhibitory synaptic gain (B) of 22.0 mV, and an axonal conduction speed of 3.0 mm/ms.
- Execution: Simulations ran for 300,000 ms using a Heun integration scheme (step size 0.522 ms). Time series data for variables $y_0$ (pyramidal output) and $y_1$ (interneuron output) were exported.
Neurotransmitter Kinetic Modeling (COPASI):
- Neural activity time series were downsampled and imported into COPASI via the BASICO Python tool.
- Mapping: $y_0$ was mapped as a proxy for glutamate release, and $y_1$ for acetylcholine release.
- Kinetics: A cytosolic compartment was defined with zero initial concentrations. Release kinetics were modeled with a rate constant ( $k$ ) of 0.5, and first-order decay reactions were set at $k=0.1$ .
- Analysis: Deterministic time-course simulations were run over 500 steps, with concentration profiles evaluated using paired statistical tests ( $\alpha = 0.05$ ).

Key Results

Neurotransmitter Dynamics: The simulation demonstrated that both acetylcholine and glutamate concentrations increased in response to the 40 Hz gamma oscillation treatment.
Temporal Profile: Both neurotransmitters exhibited a sigmoidal rise, reaching normalized levels near 1.0 within the first 100–150 ms and remaining stable throughout the 1,000 ms observation window.
Model Validation: A comparison of postsynaptic potentials (PSP) revealed a distinct gap between healthy and AD-affected simulations, confirming that the pathological connectivity was correctly established. Parameter sweeps confirmed the robustness of these temporal profiles across varying synaptic time constants and gains.

Significance and Claims
The authors claim that this study successfully simulated gamma oscillation treatment on a virtual AD subject, resulting in rapid, quantifiable spikes in neurotransmitter proxies. The findings highlight a "complex therapeutic duality":

Positive Outcome: The increase in acetylcholine is favorable, as it addresses the cholinergic deficits typical of AD.
Potential Concern: The concurrent rise in glutamate is unexpected and potentially problematic, given that glutamate is typically excitotoxic and elevated in AD.

The paper posits that these results support the potential of gamma oscillation modulation as a therapeutic strategy, though it explicitly notes that the observed increases may be tied to a general increase in network excitation rather than a targeted biochemical boost unique to AD pathology. The authors conclude that while 40 Hz stimulation acts as an effective short-term network excitor, the true clinical effectiveness and long-term biochemical impacts remain unclear. The study suggests that this computational pipeline can be adapted to evaluate gamma oscillation therapies for other neurodegenerative conditions, such as Parkinson's Disease.

Limitations
The authors acknowledge significant constraints, including the reliance on kinetic abstraction to estimate localized biochemical values from neural mass proxies, the use of a single subject dataset, and the brief temporal window of the simulation. They emphasize that the concurrent rise in both neurotransmitters requires validation through expanded simulation timeframes and in vivo animal models to empirically measure long-term synaptic effects.

Technical Summary: Simulating the Effects of Gamma Oscillation Treatment on Selective Neurotransmitters

More like this