MRSA kills 20,000 Americans per year, but scientists have yet to figure out exactly how it does the deed. But researchers can't exactly infect a bunch of people with antibiotic-resistant bacteria to figure out how to cure it. So when a group of researchers led by Alice Prince at Columbia University decided to try and crack the case, they used a strange hybrid: mice with a human immune system.

From a biological standpoint, humans would be the best model organisms for studying human diseases; to truly understand diseases and what treatments might work, we need to study diseases in ways that mimic real biological conditions as best as possible.

Enter animal models, which allow us to examine diseases and conditions in ways that simply can’t be done with humans. For example, animal facilities are carefully regulated and controlled, which provides benefits to the animals as well as to the researchers. The animals receive quality care that is standardized, which means the animals are treated well and are all treated at the same baseline: same food, same water, same air, same schedules of care.

Having this much control over their environment minimizes any variability that could potentially skew results from experiments. Try getting a group of humans to follow the same exact schedule and diet! Often, one benefit of using animals is simply that you can use more of them, and repeat your experiments enough times to show that your result is reproducible, and therefore more likely to be true.

But animal models aren’t always perfect. Take Salmonella bacteria. A common cause of food-borne illness, a variety of Salmonella strains cause over a million cases of salmonellosis combined in the United States alone each year. In humans, the bacteria colonize the intestines and cause inflammation, diarrhea, nausea, and other symptoms of gastrointestinal distress that generally last for about a week.

The story is different in mice. The bacteria can't grow as well in mice intestines, so they escape and invade the liver and spleen, causing a more widespread, or systemic, infection. Since the bacteria cause such a different type of disease in mice compared to humans, results from experiments done with mice may not necessarily match what happens in us.

That doesn't mean that information gathered using animal models is useless, just that we need to be creative and smart about how we use animals. Prince and her team developed a type of mouse that would let them ask questions about how MRSA—methicillin resistant Staphylococcus aureus—makes people sick. They did this by giving mice a human immune system.

There is a rare, but natural, genetic mutation that causes mice to lose a large portion of the cells that make up their immune system. These mice are called severe combined immune deficiency (SCID) mice, and they can be bred to keep a population carrying this mutation going. Since they have very poor immune systems, these mice are kept in separate areas and given extra care to prevent them from becoming sick.

Then the mice are injected with human stem cells that will develop into the different cells that make up the immune system. Researchers can tailor their humanized mice to the condition they want to study by giving the mice different types of human cells. For example, the authors of the MRSA paper also gave the mice some tissue from the thymus, which is critical for making T cells. Twelve weeks after giving the mice human cells, 5 to 63 percent of their immune system cells were human.

The authors then used the humanized mice to see if they could address unanswered questions about how MRSA makes people sick.

MRSA is a health issue for many reasons, but perhaps the most pressing issue now is its risk of becoming resistant to additional antibiotics. Antibiotic resistance, the process by which bacteria become able to withstand antibiotics, makes the infections they cause become increasingly difficult to treat. The World Health Organization describes antibiotic resistance as one of the biggest threats to human health, food security, and development worldwide. The problem is so troubling that the Centers for Disease Control and Prevention recently created an entire program devoted to monitoring and researching antibacterial resistant bacteria, called the Antibiotic Resistance Solutions Initiative.

Understanding how exactly these bacteria make us sick is therefore of great importance. The authors first showed that more bacteria survived in lung fluid and tissue from humanized mice than in samples taken from regular mice, or even the SCID mice, showing that human cells were worse at fighting off MRSA than mouse cells. While others had already shown that the potent PVL (Panton-Valentine leucocidin) toxin from MRSA is better at attacking human cells than many types of non-human cells, the new results both confirmed those earlier findings and showed that the humanized mouse model was working as predicted.

As the authors mention throughout their paper, though, they aren’t the first to study PVL, nor the first to show that PVL targets and kills certain human cells. So why did they go through the effort to make this new mouse? The best way to make sure results from an experiment are true are not just to repeat the experiment, but to repeat it as new methods that more accurately mimic natural systems are developed. If we get the same answers by using better methods, then we can have more confidence that those results are true.

For example, one group of researchers studied PVL using rabbits. While the toxin killed the rabbit cells like it does human cells, a humanized mouse model provides some advantages over the rabbit. First, mice are easier to maintain in the lab. Also, humanized mice more closely mimic a human immune system than a rabbit does, meaning the results are more likely to reflect what is actually happening in a human.

Another group studied PVL on cells removed from humans and grown in the lab. This approach carries the benefit of working directly on human cells, but it lacks the benefit of working within a complex living organism, which is the natural system in which MRSA infection occurs. A humanized mouse, while not the exact natural system for MRSA infection, is much closer than cells in a petri dish would be. The authors suggest that the major next step would be to create mice that carry cells from different parts of the human immune system to allow researchers to learn specifically how different parts and different types of immune cells work in response to MRSA, as well as other infections.

Controversial ethics

It's impossible to make mice 100 percent human. But can the ethics of testing on humanized mice stretch too far?

I will never forget my first day of animal training. I was fresh out of college, starting my first full-time job. I was working in a cancer biology lab, and my main responsibility was to care for and keep track of the different types of mice we used. Being an animal lover, I suspected it would be tough learning how to use mice in a medical research lab.

I was right. After my first day of animal training, just learning the basics made me feel terrible: putting ID tags on ears, collecting blood samples from tails, learning how to firmly hold the mice so they can't reach around and bite your fingers (and boy, do they try!) and, of course, learning how to sacrifice a mouse once an experiment is done. My guilt was strong enough that I purchased a hamster from the pet store on my way home that night. Somehow, I thought having one little rodent to care for at home would balance out the work I did during the day. Coming home to Duncan did help, especially in those first few weeks.

While working with animals often isn’t easy or pretty, it helps to know that institutions with animal facilities have to follow strict regulations ensuring that animals are given adequate food and water and that cages are cleaned on a regular basis. And experiments involving mice have to meet a high level of approval that prevents undue harm toward the animal. There are also often dedicated staff members whose sole responsibility is to care for the animals, including making sure that no animal is becoming too sick or stressed from a particular experiment.

And, most importantly, we can use animals to study diseases and conditions in ways we simply cannot do with humans. Safety trials for new drugs must pass preliminary tests in animals, so we can thank animal models for essentially every FDA-approved drug available to us today. Even medications for our pets go through initial safety studies in lab animals. Furthermore, many groups of scientists have created mice carrying genetic mutations in BRCA1 and BRCA2, two genes linked with hereditary breast and ovarian cancers. These mice have allowed scientists to better study the association between these mutations and the cancers they can cause. Similarly, mice are used to study Alzheimer’s disease, the leading form of dementia.

While using animal models in medical research can make people uncomfortable, the truth is that they play an essential role in our ability to study and treat diseases. These animals deserve the best care, and our respect, for helping us in the fight. Next time you reach for an aspirin for your headache, or a decongestant for that flu, remember to be thankful for the lab animals who tried it first.