NetLogo User Community Models

(back to the NetLogo User Community Models)

Download If clicking does not initiate a download, try right clicking or control clicking and choosing "Save" or "Download". Try It in NetLogo Web

## WHAT IS IT?

This model simulates microglia, a type of immune cell in the central nervous system (CNS), and their interactions within a 2-dimensional slice of the hippocampus. It builds on a previous model that simulated microglia and their interaction with neurons and cytokines, a kind of signaling protein. This version of the model is focused on microglia metabolism and glucose consumption in the context of Alzheimer’s Disease (AD).

Microglia are responsible for phagocytosing (eating) cellular debris and fighting infections in the CNS. They can also modulate blood brain barrier (BBB) permeability, which regulates the flow of molecules between the bloodstream and the brain. To perform their immune tasks, microglia utilize glucose as their primary energy source. However, impaired microglia metabolism can cause harm to the CNS. This includes causing excess lactate in the brain, more acidic pH, greater levels of Aβ plaques, and a "leaky" BBB with lower barrier integrity. These are all symptoms and hallmarks of AD, which currently has no cure or definitive cause.

However, recent research has shown that exercise and metabolic boosters are potential treatments to AD. Exercise regimens can reduce the growth rate of Aβ plaques, while metabolic boosters such as GLP-1 receptor agonists can help rescue immune tolerant microglia and restore their immune functions. Therefore, this model helps simulate the effects of impaired microglia metabolism on the progression of AD disease, as well as how these potential therapies may rescue microglial function by acting on metabolic pathways.

## HOW IT WORKS

This simulation relies on microglia agents and patch variables representing Aβ plaques, the BBB, and lactate accumulation. The primary outcomes of this model are the effects of microglia metabolism on Aβ plaque accumulation, brain pH, and BBB permeability. One tick of the simulation represents a time step of one hour, and simulations last 8,760 ticks (one year).

### Microglia

This model includes two agent breeds, homeostatic and proinflammatory (M1) microglia. Homeostatic microglia are represented by an arrowhead shape, while M1 microglia are represented by a directional arrow shape. Both breeds accomplish their defined tasks through pseudo-random movement, and an agent can switch between these breeds depending on specific environmental cues.

Homeostatic microglia undergo oxidative phosphorylation (OXPHOS) to metabolize glucose, allowing them to generate ATP (energy). These microglia can also strengthen the BBB by increasing their INTEGRITY value. However, strengthening the BBB for too long can cause microglia to become proinflammatory, switching to the M1 breed. Homeostatic microglia can also become proinflammatory when adjacent to an Aβ plaque.

M1 microglia undergo glycolysis to metabolize glucose, which is a faster way to generate ATP to supply their immune response. This comes at the cost of greater glucose consumption (around 10 times more than OXPHOS) and lactate generation. Excessive lactate can cause a pH imbalance within the hippocampus, contributing to greater levels of Aβ plaques. M1 microglia can phagocytose (clear) Aβ plaques, upon which they switch back to the homeostatic breed.

### Blood Brain Barrier

As part of SETUP, the BBB is initialized with the INITIALIZE-VESSELS procedure. Temporary agents are created and randomized on one of the sides of the environment. These agents will then randomly move throughout the environment, setting the INTEGRITY of the patches they encounter by some positive amount. Patches with a nonzero INTEGRITY value will be a shade of red, with stronger colors representing greater integrity. The initializing agents die once reaching an edge of the environment, creating a trail of patches representing a blood vessel. The number of these agents can be adjusted through the INIT-VESSELS slider.

Homeostatic microglia that encounter a blood vessel patch will strengthen BBB integrity. However, fortifying for too long can cause homeostatic microglia to switch to the proinflammatory M1 breed. M1 microglia instead decrease the INTEGRITY value of the BBB, representing an increase in permeability.

### Glucose Levels

Every tick, microglia use some amount of glucose from the GLOBAL-GLUCOSE value. The amount of glucose used varies depending on the chosen metabolic pathway. Glycolysis, used by proinflammatory microglia, uses 10 times more glucose compared to OXPHOS, used by homeostatic microglia. If microglia cannot obtain enough glucose from the environment, they will stop in place and be unable to perform any immune responses.

The amount of glucose entering the system is influenced by ADDED-GLUCOSE, a variable indicating the amount of glucose entering the system per tick. The total glucose in the environment is represented by GLOBAL-GLUCOSE, and its initial value upon setup is equal to ADDED-GLUCOSE.

### Lactate and pH

When microglia are proinflammatory, the use of glycolysis causes an excess of lactate to be secreted into the extracellular space. This is visualized through the light blue trails emitted from M1 microglia. Under typical circumstances, excess lactate can be used by the environment, such as by neurons. However, too much lactate can cause semi-permanent lactate deposits (white) to appear. These deposits cause the pH of the environment to decrease, which can increase the probability that Aβ plaques spread. The chance that a blue lactate patch becomes a lactate deposit is influenced through the LACTATE-PROBABILITY slider.

To visualize the spread of lactate, patches can emit a light blue trail that dissipates after a certain amount of time. For more information on the procedures related to lactate spread trails, see our work on the cytokine and inflammation procedures found in [our previous model.](https://ccl.northwestern.edu/netlogo/models/community/Microglia%20Model)

### Aβ Plaques

The number of initial patches with an Aβ plaque (orange) can be adjusted with the INIT-PLAQUE slider. At regular intervals, existing Aβ plaques will attempt to spread to other patches, representing the progression of AD. The chance that spreading succeeds relies on the PLAQUE-SPREAD-PROB variable. This variable is inversely related to the current pH. With exercise, PLAQUE-SPREAD-PROB is reduced by a constant amount. The greater the exercise intensity, the lower the spread probability.

M1 microglia are responsible for phagocytosing Aβ plaques that they encounter. The chance that a proinflammatory microglia succeeds in phagocytosing plaque is dependent on the EAT-PROBABILITY slider and PLAQUE-EAT-SCALAR variable, which is based on the selected exercise amount. If successful, the microglia becomes homeostatic and plaque will be removed from the patch.

### Exercise

The EXERCISE chooser represents the level of daily exercise, which influences the spread of Aβ plaques. When set to “none”, the probability of plaques spreading relies only on the current pH of the system, with lower pH causing a higher chance of spreading. With exercise, chances of plaque spreading decreases while their chances of removal by microglia increases. Additionally, exercise also periodically increases the INTEGRITY of BBB patches. Higher exercise intensity will result in a greater effect.

### Immune Tolerance

If proinflammatory microglia are adjacent to Aβ plaques for an extended period of time (one day), they become immune tolerant. When they switch back to homeostatic, these microglia cannot become proinflammatory for a temporary period of time. The time in which a microglia agent is tolerant increases as they continue to stay adjacent to plaques.

Metabolic booster effects are set with the METABOLIC-BOOSTER chooser. Choices represent common dose frequencies, such as “daily” or “weekly”. Given one tick representing one hour, the chosen boosting frequency determines the interval between doses. Whenever a metabolic booster is applied, all currently tolerant microglia will cease being tolerant.

## HOW TO USE IT

The Interface tab includes sliders and switches that modify the simulation. Below is a description of these variables:

* INIT-MICROGLIA: The number of microglia that are initialized in the model.

* INIT-PLAQUE: The number of initial patches that contain an Aβ plaque.

* INIT-VESSELS: The number of blood vessels created by the INITIALIZE-VESSELS procedure during SETUP.

* EAT-PROBABILITY: The probability that a proinflammatory microglia successfully phagocytoses an Aβ plaque.

* LACTATE-PROBABILITY: The probability that part of a lactate trail becomes a lactate deposit.

* FORTIFY-PROBABILITY: The probability that a homeostatic microglia successfully increases the INTEGRITY of the current vessel patch.

* ADDED-GLUCOSE: The amount of glucose that is added per tick. This also determines the starting GLOBAL-GLUCOSE value.

* METABOLIC-BOOSTER: The interval in which a metabolic booster is applied to the system, rescuing immune tolerant microglia.

* EXERCISE: The level of exercise to represent within the system, influencing the spread of Aβ plaques.

## THINGS TO NOTICE

Some things to notice when the simulation starts are:

* Aβ plaques can spread exponentially over time.

* When Aβ plaques comprise a large portion of the environment, microglia stay immune tolerant for longer periods of time.

* The number of lactate deposits accumulates the longer the model runs.

* Blood vessels fade to black as their INTEGRITY level decreases.

## THINGS TO TRY

Try varying the level of EXERCISE in the model without any metabolic boosters, then observe the spread of Aβ plaques. Next, try varying the levels of METABOLIC-BOOSTER without any exercise. Which treatments seem more effective in clearing plaques? Which treatments affect BBB permeability more? Does combining the effects of exercise and metabolic boosters provide an even better result?

Try changing INIT-MICROGLIA and ADDED-GLUCOSE, then watch the levels of GLOBAL-GLUCOSE on the monitor. What happens when microglia use glucose faster than it is replenished? Does having too much glucose in the system affect the microglias’ behavior? Additionally, try to find values for INIT-MICROGLIA and ADDED-GLUCOSE that strike a balance between glucose consumption and replenishment.

To see all of the model’s functions working together, try the following starting values:

* INIT-MICROGLIA: 10

* INIT-PLAQUE: 3

* INIT-VESSELS: 5

* EAT-PROBABILITY: 0.10

* LACTATE-PROBABILITY: 0.20

* FORTIFY-PROBABILITY: 0.10

* ADDED-GLUCOSE: 2400

* METABOLIC-BOOSTER: none

* EXERCISE: moderate

## EXTENDING THE MODEL

The CNS is a multi-faceted system with a myriad of different cells, pathways, and interactions. Our model is a simplification of the interaction between microglia, their metabolism, and AD. Ways to extend this model include:

* Astrocytes, another cell in the CNS, play a key role in modulating BBB permeability and protecting the brain from injuries and infections. These cells can respond to chemical signals sent by microglia and neurons, as well as send their own chemical signals. Modeling the interactions between astrocytes and microglia may help with understanding the role of these cells in AD contexts.

* Angiogenesis is the process by which new blood vessels are formed from existing vessels. Excess lactate and exercise can stimulate angiogenesis, which can improve brain function. However, defective blood vessels may form around Aβ plaques, disrupting the blood brain barrier. Developing this model to include angiogenesis can help better model how the BBB is affected by AD progression.

* Neurons, astrocytes, and microglia all use glucose as a source of energy. This model currently only takes into account microglia metabolism using glucose. However, some evidence points to microglia and neurons being able to use lactate as an emergency energy source. Future models can take glucose consumption by other cell types into consideration, alongside alternative energy sources.

* The timing in which metabolic boosters are started for people susceptible to/have AD may affect whether the treatment is successful. The model currently only supports the interval between booster timing rather than when this treatment was started in terms of current AD progression. Future models can take into account different AD stages and how the effects of metabolic boosters vary depending on which stage patients began receiving them.

## RELATED MODELS

Our previous microglia model is posted here: [Microglia Model](https://ccl.northwestern.edu/netlogo/models/community/Microglia%20Model)

## CREDITS AND REFERENCES

This project was supported by NSF grant 2245839 from the Mathematical Biology program. We are deeply grateful for this support.

We would also like to thank Amanda Case and Emmanuel Mezzulo for their work on our previous model, on which this one builds from.

For more information about microglia, see our publication in Spora: A Journal of Biomathematics: [https://ir.library.illinoisstate.edu/spora/vol11/iss1/3/](https://ir.library.illinoisstate.edu/spora/vol11/iss1/3/).

## HOW TO CITE

If you mention this model or the NetLogo software in a publication, we ask that you include the citations below.

For the model itself:

Penland, A.\*, Ty, C.\*, Gerving, J., Mendoza-Ceja, M., Pratt, M., & Larripa, K. (2025). Microglia Metabolism Model. Cal Poly Humboldt, Arcata, CA.

\* These authors contributed equally to this work.

Please cite the NetLogo software as:

Wilensky, U. (1999). NetLogo. http://ccl.northwestern.edu/netlogo/. Center for Connected Learning and Computer-Based Modeling, Northwestern University, Evanston, IL.