Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Animal models: Not close enough

Despite some outstanding drug-development successes, the mouse version of multiple sclerosis has been worryingly unreliable at screening human treatments.

If we are to believe the headlines, multiple sclerosis (MS) has been cured many times over. Dozens of interventions, some as simple as the spice turmeric and others as sophisticated as gene therapy, have shown promise in animal models of MS. In 2011, for example, two reports described compounds that could block the progression of the disease in mice1,2. But such stories have been appearing for decades, so why is there still no cure for human MS?

It's telling that the predominant animal model was not developed with MS in mind. Called experimental autoimmune encephalomyelitis (EAE), it grew from an attempt in 1933 to understand the 'neuroparalytic catastrophe' — weak limbs, high fevers, seizures and even death — that sometimes resulted from vaccination against rabies or smallpox. Researchers found that by repeatedly inoculating monkeys with brain tissue without any virus, they could induce a suite of neurological symptoms. These symptoms are reminiscent of MS but result from markedly different biological processes.

Three-quarters of a century and more than 5,000 publications later, EAE has been refined to more closely resemble human MS. Animals (usually mice, but sometimes rats, guinea pigs, rabbits or primates) are injected with a two-part cocktail. First, some form of myelin antigen — either purified myelin protein, whole or otherwise, or pulverized spinal cord — primes the immune system to recognize and launch an attack on the animals' own myelin. Second, a powerful adjuvant containing killed tuberculosis bacteria kicks the immune system into high gear. The result is an MS-like progressive paralysis that starts at the tail and works its way up.

More than half of the human MS drugs were first evaluated in animals with a different nervous system disorder – experimental autoimmune encephalomyelitis, or EAE – which produces similar neurological symptoms. Credit: Hooke Laboratories, Inc.,

In one sense, this model has been a spectacular success: more than one-half of the drugs that can alleviate human MS were first evaluated in animals with EAE. In the case of fingolimod, an oral medication that hit the market in 2010, less than a decade elapsed between success in rats and approval in humans.

But these victories do not tell the whole story because EAE has been troublingly unreliable for screening human MS treatments. For every drug that has translated well into humans, there have been many more that did not work. Finding that a drug can cure EAE gives almost no indication of whether it will help treat MS. In fact, the EAE success stories are so random as to be “almost coincidental”, says Richard Ransohoff, director of the Neuroinflammation Research Center at the Cleveland Clinic in Ohio.

In fact, some interventions that proved therapeutic in animals with EAE — such as interferon γ and tumour necrosis factor (TNF)-α blockade — actually make human MS worse. Even the success stories sometimes have a dark side. The drug natalizumab, for example, which can reduce relapses and slow the progression of disability in human MS, was temporarily withdrawn from the market when some patients developed a potentially fatal brain infection not seen in the animal model. “If you wanted to have a robust, predictable platform for drug screening, you'd consider the animal-model situation to be very disappointing,” says Ransohoff.

“We would like to have a model that would tell us it's worth pushing a drug, or that we can drop the drug without a second thought,” says immunologist Roland Liblau of the Rangueil University Hospital in Toulouse, France, president of the French Society of Immunology. “But it's not so easy in real life.”

A mightier mouse

One of the main differences between the animal model and the human disease is that whereas MS occurs spontaneously, EAE is induced. In humans, the cause of MS remains mysterious (see The X factor). But for EAE in mice, “I know why the disease is starting: it's because I did it,” says neurologist Michael Racke of Ohio State University in Columbus. And it is much easier to fix something if you know exactly how you broke it in the first place.

A truly spontaneous mouse model of MS would more accurately represent the situation in humans. Until recently, attempts to create a mouse that spontaneously develops EAE resulted in strains so artificial that they cannot tell us much about the human disease, says Racke. These transgenic mice were engineered with T cells that recognize and attack a myelin antigen. But the rest of the mouse's immune system must be hobbled for the disease to develop, leaving it severely immunocompromised. “You'd never see a human like that,” says Racke.

In 2006, a mouse model was reported that spontaneously develops a specialized variant of MS in which myelin destruction is confined to the optic nerve and spinal cord, and progresses steadily3,4. These transgenic mice have both T and B cells engineered to target a myelin antigen. Unlike earlier spontaneous models, their immune systems remain otherwise intact — an important advance. But this model still failed to capture important aspects of the most prevalent form of human MS, in which symptoms relapse and remit over time. To address this shortcoming, some of the researchers used a different strain of mouse as a starting point and in 2009 came up with the first spontaneous model of relapsing–remitting MS5.

But the more a mouse's immune system has been tweaked to induce an MS-like disease, the less it is possible to determine with any degree of confidence how the immune system produces that disease. Other autoimmune diseases, notably diabetes and lupus, benefit from a naturally occurring, spontaneous mouse model. For MS, researchers know of no such mouse.

Other researchers advocate a completely different approach. Instead of trying to make EAE more like MS, they are trying to make mice more like humans. Young mice that lack an immune system can be spurred to develop a human-like immune system when injected with human stem cells or blood from the umbilical cord. Such mice could be further manipulated to carry genes known to confer susceptibility to MS.

Theses 'humanized' mice might be better at predicting how people will respond to drug candidates — perhaps warning of the deadly side effects that often go undetected in traditional EAE models. But a humanized mouse can only go so far: its immune cells might be human, but its central nervous system — the focus of MS — remains strictly murine.

Not-so-great expectations

Despite all these refinements, EAE is not, and never will be, MS. The two diseases have different triggers and proceed through different mechanisms. MS is governed by an unknown number of susceptibility genes and triggered by poorly understood environmental influences, a level of complexity impossible to reconstruct in an inbred mouse living under controlled laboratory conditions. Given these differences, mouse models might be better understood not as testing grounds for human drugs, but rather as tools for picking apart the mechanisms that drive autoimmune attacks on the nervous system.

In 2009, for instance, Liblau and colleagues at the French National Institute of Health and Medical Research (INSERM) used a clever twist on EAE to show that killer T cells — white blood cells with the ability to kill cells that have been infected or otherwise compromised — can be programmed to target myelin-producing brain cells and destroy them6. Several other laboratories have since reproduced the finding. These cells were long suspected to contribute to the progression of MS because they show up in MS brain tissue sections, but their role in damaging brain tissue had previously been uncertain.

In December 2011, researchers at Harvard Medical School used mice with EAE to uncover clues about the mysterious lymph-like structures that often appear in the nervous systems of MS patients. They found that only one subtype of T cell, called T-helper 17 (TH17), could induce these structures to form in EAE mice, suggesting that these cells play a crucial role in MS pathogenesis.

Meanwhile, rats with EAE have allowed researchers at the Karolinska Institute in Stockholm to probe the genetics of MS susceptibility. Rats used to be the animal of choice for EAE but were largely supplanted by mice, which are more readily available, although some laboratories continue to use the larger rodents. By crossing susceptible rats with resistant rats, the researchers zoomed in first on a chromosomal region and then on a particular gene associated with increased propensity to develop MS. The gene has since been shown to confer MS susceptibility in humans as well.

So although EAE might not be a reliable platform for drug development, it has already led to improved treatments for human MS through more indirect means. “I have the belief that proper use of the animal model, which is to ask scientific questions and not to screen drugs, is very helpful for drug development and for identifying treatment opportunities,” says Ransohoff.

Natalizumab is a prime example of this approach in action. The drug, which prevents immune cells from crossing the blood–brain barrier, was developed for mice in the early 1990s. After proving effective against EAE, it was fast-tracked through human clinical trials. Although it was extraordinarily successful in treating MS, it had the unfortunate side effect of leaving some patients vulnerable to a rare but deadly brain infection called progressive multifocal leukoencephalopathy. When researchers did spinal taps on patients who took the drug, they found that it was also blocking immune cells from entering the spinal fluid — thought to be a critical first step in the development of MS.

Another group took this concept back into the mouse model to probe the molecular underpinnings of the process. They discovered a receptor that is required for cells to cross into spinal fluid. That receptor is now a potential drug target, says Ransohoff. A drug that keeps immune cells out of the spinal fluid but still allows them into the brain might knock out MS without compromising the body's ability to fight a brain infection.

Liblau describes this back-and-forth approach as a game of table tennis — and it is a game he would like to keep playing. “I really believe,” he says, “that this ping-pong game is the way to go to better understand the disease process.”

Sidebar: Go fish

Existing treatments for MS attempt to halt the emergence or progression of the disease. By contrast, there have been few attempts to repair the myelin and oligodendrocyte damage that has already taken place. To learn how we might make such repairs, researchers are now looking to another animal model entirely: the zebrafish larva.

Zebrafish larvae are tiny, develop quickly and cost about 1,000-fold less to work with than mice. What is more, they are transparent; so, by tagging relevant proteins with fluorescent molecules, researchers can watch myelin formation and oligodendrocyte differentiation take place right before their eyes.

In the past few years, several groups have harnessed the power of this system to screen large libraries of compounds quickly for potential myelin-repairing drugs. These little see-through fish larvae might open the door to an entirely new approach to treating MS.


  1. Solt, L. A. et al. Nature 472, 491–494 (2011).

    ADS  CAS  Article  Google Scholar 

  2. Cruz-Orengo, L. et al. J. Exp. Med. 208, 327–339 (2011).

    CAS  Article  Google Scholar 

  3. Krishnamoorthy, G. et al. J. Clin. Invest. 116, 2385–2392 (2006).

    CAS  Article  Google Scholar 

  4. Bettelli, E. et al. J. Clin. Invest. 116, 23931–2402 (2006).

    Article  Google Scholar 

  5. Pöllinger, B. et al. J. Exp. Med. 206, 1303–1316 (2009).

    Article  Google Scholar 

  6. Saxena, A. et al. J. Immunol. 181, 1617–1621 (2008).

    CAS  Article  Google Scholar 

Download references

Author information

Authors and Affiliations


Rights and permissions

Reprints and Permissions

About this article

Cite this article

Rice, J. Animal models: Not close enough. Nature 484, S9 (2012).

Download citation

  • Published:

  • Issue Date:

  • DOI:

Further reading


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing