Study highlights strengths, limitations of mouse models used to study MS

The two main mouse models used to study myelin damage in multiple sclerosis (MS) each replicate important but distinct biological features of the disease — but both fall short of capturing its full biological complexity, a new study shows.

The findings suggest that each of these two models may have unique strengths and limitations, and each may be useful for studying different aspects of MS.

“Our analysis of these two models of myelin loss and regeneration provides a road map based on robust scientific evidence that we hope will advance the study of MS and related diseases,” Katrina Adams, PhD, co-author of the study at the University of Notre Dame in Indiana, said in a university news story.

The study, “A comparative transcriptomic analysis of mouse demyelination models and multiple sclerosis lesions,” was published in Nature Communications. The National Multiple Sclerosis Society funded the work.

Recommended Reading

April 28, 2026 News by Marisa Wexler, MS

Brain inflammation may help repair myelin, animal study shows

2 MS mouse models often used interchangeably by scientists

In MS, inflammation in the brain and spinal cord drives damage to the myelin sheath, a fatty substance that wraps around nerve fibers and helps them send electrical signals more efficiently. Myelin damage, also known as demyelination, disrupts normal nerve signaling, ultimately giving rise to symptoms of MS.

To study how myelin is damaged and repaired, researchers commonly rely on one of two mouse models. In one model, a toxic chemical called cuprizone (CPZ) is added to the mice’s diet, leading to widespread myelin damage that develops over several weeks. In the other model, a chemical called lysophosphatidylcholine (LPC) is injected directly into the brain, and myelin damage at the injection site develops rapidly over a few days.

The CPZ and LPC models are often used interchangeably by scientists. But it’s not known whether they are actually interchangeable and accurately capture patterns of myelin damage seen in people with MS.

To address this, a team of scientists at Notre Dame analyzed gene activity in individual brain cells from both models and compared them with large single-cell data sets from MS patient brain tissue.

“By matching each model to features seen in diseased tissue from real patients, we can be sure that we’re targeting things that are actually causing disease in human patients,” Adams said. “There are so many potential paths to follow, so we want to make sure that the path chosen has direct relevance to MS patients.”

Recommended Reading

May 4, 2026 News by Marisa Wexler, MS

Brain sugar levels act as signal for myelin growth, study finds

Mouse models represent simplified versions of more complex system

A central finding of the study is that oligodendrocytes do not simply die or recover during myelin loss. Instead, they enter distinct “damage-associated” cellular programs that reflect different types of stress and injury.

Some of these states were shared across both models and human MS tissue and were characterized by activation of immune programs in oligodendrocytes. However, the models diverged substantially in other aspects.

In the CPZ model, oligodendrocytes entered a state associated with more severe cellular stress, which closely resembled oligodendrocyte dysfunction observed in MS lesions. By contrast, the LPC model primarily induced immune-driven oligodendrocyte responses, with less evidence of deep cellular stress.

“We found that CPZ demyelination induces a distinct [disease-associated oligodendrocyte pattern of activity] characterized by a pronounced [genetic activity] shift not observed following LPC injection,” the researchers wrote. “This CPZ-specific state exhibited substantially greater gene overlap with [disease-associated oligodendrocytes] identified in [people with MS] compared to LPC-induced changes.”

Although oligodendrocyte activity during myelin damage was different, oligodendrocyte activity during myelin repair was similar in the CPZ and LPC models, with both models sharing many of the characteristics seen in human samples.

The strategic use of these two preclinical models is essential for translating insights into therapies that might restore lost myelin. We need to better understand the very process of demyelination in order to treat one of the root causes of this debilitating disorder.

Still, human disease showed far greater diversity in oligodendrocyte states, indicating that while mouse models capture key aspects of disease biology, they represent simplified versions of a much more complex system.

Notably, although the LPC model didn’t reflect the dynamics of oligodendrocytes as well as the CPZ model, it showed changes in inflammatory cell activity that more closely resembled human MS. For example, the LPC model showed changes in the activity of immune cells called perivascular macrophages that better reflected human disease, while immune responses were more restricted in the CPZ model.

Overall, these data suggest that both the CPZ and LPC models reproduce key aspects of MS biology, but each captures different elements of the disease. CPZ may be better suited for studying oligodendrocyte stress, whereas LPC may be better suited for studying acute inflammatory responses and immune cell dynamics.

“If you’re studying the myelin-producing cells and what’s happening to them in MS — are they stressed, dying or trying to repair? — CPZ is better, since the loss of myelin is more gradual,” Adams said. “For studying the immune cells that respond to the myelin loss, LPC may be better, since the immune response is more aggressive than in CPZ. The strategic use of these two preclinical models is essential for translating insights into therapies that might restore lost myelin. We need to better understand the very process of demyelination in order to treat one of the root causes of this debilitating disorder.”