“Cancer has been cured a thousand times.”
So says Christopher Austin, the director of the National Center for Advancing Translational Sciences (NCATS) at the National Institutes of Health. Austin should know — as the director of NCATS, his focus is on exactly these kinds of groundbreaking laboratory studies.
His proclamation comes with a significant caveat that will pop the bubbles in your champagne. Austin is so interested in these studies because they all happened in mice, in a lab. When the hundreds of different drugs that made mouse tumors disappear were carried forward to human trials, they went in and came out without doing what they promised. Or worse, they turned out to be toxic.
The failure of drugs and procedures to translate from animal models to humans plagues the entirety of medical research. An astounding 90 percent of drug trials never make it from the first phase of development to FDA approval, and the humble lab mouse shoulders much of the blame. Long held as the standard model for animal research, scientific advances have called the reign of the mouse into question. But should scientists ditch their furry lab model altogether, or can they simply design a better mouse?
Mice were introduced into the lab back in the 1920s, when an ambitious young geneticist named Clarence Cook Little believed he had found the perfect model for studying cancer. Little strongly believed that cancer was an inheritable disease, and mice, short-lived and low maintenance, turned out to be the ideal subjects for his experiments. Little would go on to found The Jackson Laboratory and sold mouse strains to researchers all over the country, even going so far as to secure the mouse as the official animal model for research funded with government grants following the passage of the National Cancer Institute Act in 1937.
Other animals were, and still are, used to study disease. Cats and dogs are popular, as well as chimpanzees; of course, ethical considerations get knottier the further up the food chain you go. Add to that the problem of cost: A standard lab mouse can be had for about $20 — it costs far more to buy a healthy feline or canine and in sufficient numbers to conduct thousands of trials. Due to economies of scale, an entire industry has sprung up around lab mice. Today, the Jackson Laboratory is a nonprofit biomedical research institution that delivers crates of genetically-identical mice to laboratories around the world. The primacy of the mouse in scientific research has become well-established.
Small, Important Differences
We share more than 97 percent of our working DNA with mice, a consequence of a shared ancestor some 75 million years ago. This similarity has been both a blessing and a curse. While we share many fundamental biological processes, it is the subtle, hard-to-find divergences that can make the difference when researching highly specific disease mechanisms. That 3 percent of the genome that sets us apart from mice can have a big influence on how our bodies work. In some cases, the difference between a human and a mouse comes down to a single nucleotide. Austin likens it to trying to read a word with a single letter missing.
“It turns out, that on a genetic level, these differences, these spelling differences, even if they’re small, can really be important,” says Austin. “If you have a very long word and you mess up a letter, that changes the meaning of the word. And that’s true in DNA spelling too. Sickle cell anemia is caused by one difference out of three billion.”
What’s more, the sequences of DNA themselves, the “words”, can be interpreted differently. Genes can be expressed in more than one way, depending on how transcription factors are applied to them. Transcription factors determine how DNA copies itself to messenger RNA, effectively altering how the influence a particular gene has on your body. In this way, a 3 percent difference grows into a yawning chasm between species.
Mysteries Inside of Mysteries
Scientists now appreciate the delicate machinery of life that’s working in mice and humans. Where once we accepted broad, physical similarities as adequate evidence of our kinship with mice, the ability to peer into the world of bacteria, cells and molecules, has added much-needed nuance to our understanding of biology. For example, the micro-environment, or the sum of all of the cells in a system and how they interact, is exposing crucial differences between mice and men, especially in cancer research.
“Tumor cells, by themselves, will never survive,” says Ellen Puré, the chair of the Department of Animal Biology at the University of Pennsylvania School of Veterinary Medicine. “If you took a nice tumor cell and put it in a normal tissue, it won’t grow. It actually requires the cells around it to regulate whether or not the cancer cell will grow.”
The micro-environment plays a vital role in whether cancer will establish a foothold in the body. And, you guessed it, mice and humans don’t have the same micro-environment. After all, mice and humans live in different conditions, consume vastly different diets and behave differently. Currently, scientists don’t enough about the human micro-environment to explain why something that works in mice may not work in humans.
Take, for example, a class of lipid-lowering drugs called fibrates, prescribed to some patients with diabetes. In mice, they cause the proliferation of organelles called peroxisomes and subsequently a high rate of liver cancer. When we take the drug, the deadly chain reaction that happens in mice is completely absent, and researchers still don’t know why — they just know that the drug works. Cardiovascular research has suffered from a similar blind spot, and numerous other trials have been thrown out when unknown mechanisms cause them to fail.
The way scientists conduct most of their experiments — in sterile labs with genetically identical mice housed in decidedly unnatural conditions — could also be muddying the results they get.
“If you take the same mouse and you feed it two different chows, the gene expression patterns of that mouse change dramatically on the basis of the food it eats,” says Austin. “If you put them in two different environments, in two different time zones, two locations, the results are different.”
A recent study by Herbert Virgin, a professor of pathology and immunology at the Washington University School of Medicine, showed that when mice were purposefully exposed to viruses that humans encounter regularly, they became much better models. The control mice kept in sterile conditions were more comparable to infants. The lab mice bred to be “perfect” controls end up living a pampered lifestyle that ultimately alters their physiology in ways that no human will experience. In other words, a “dirty” mouse isn’t such a bad thing.
“If mice aren’t perfect models, maybe part of the reason is that we’ve cleaned them up so much,” says Virgin.
Let’s Talk It Out
Another issue: not all scientists are experts in murine biology.
“Most scientists who are using mice as models for humans disease, we ourselves are not experts in the biology or husbandry of the mouse as an organism itself. We don’t have that background in the biology of the organisms that we’re dealing with,” says Kent Lloyd, the director of the Center for Mouse Biology at the University of California, Davis.
Without complete knowledge of everything that goes into making a mouse a mouse — a herculean task, to be sure — understanding exactly how a compound will work when introduced to its system becomes a game of educated guesswork.
“I think in the past, some investigators worked in a vacuum. I work with physiologists, but they didn’t understand genetics. Or they weren’t thinking necessarily about the fact that a mouse is not the same as a human, because they were [looking at] a different piece of the puzzle,” says Elizabeth Bryda, the director of the Rat Resource and Research Center (RRRC) in Missouri. “And sometimes it gets lost in there that maybe the species you’re working with isn’t the best species for the question, but you don’t know enough about the physiology or the genetics to recognize that.”
Like mice, rats are a common choice for researchers conducting studies. While some differences between the two do exist, such as minor physiological differences and, more notably, variations in their psychology that affect how they interact, the same issues that plaguing mouse research often appear in rat research, as well.
The RRRC was first funded in 2001 and serves both as a storage center for rat strains of interest and as a kind of rat customization shop. Want a rat with pulmonary hypertension and a deficiency of serotonin? They’ve got the model for you. Or maybe one with cataracts? No problem. Using CRISPR, the researchers tweak the rat genome to create so-called “transgenic” animals with human-like disease traits that researchers can work with. It’s another way researchers are working with the models they have to make them better.
“There weren’t any rats … out there that have those same genetic differences that people have, so you’re already starting with a major difference and now you’re trying to study a disease where you don’t even have identical genetics,” she says. For example, Bryda’s current focus is creating rats that display human-like symptoms of inflammatory bowel disease for researchers to study.
Fundamentals Still Unclear
CRISPR-ing mice into ideal research models could offer a powerful solution to the myriad problems plaguing medical research by narrowing the genetic divide between humans and their animal stand-ins. But another issue looms: Just as we don’t fully understand our own bodies — if we did, we wouldn’t have to do any research at all — we still don’t know exactly what makes a mouse tick.
To this end, researchers with the International Mouse Phenotyping Consortium (IMPC) are doubling down on the mouse as a research tool and picking through its genome with the finest of combs. There are around 20,000 genes in mice that bear strong similarities to humans. We know the function of about half of these, according to Lloyd, who works with the IMPC. He is collaborating with researchers across the world to elucidate the functions of the other half. The process is painstaking and involves the systematic “knockout” of every one of those genes in a stem cell culture. Once a gene is removed, researchers grow specialized mice from the stem cells to see what happens to them.
In this way, they hope to build up a phenotypic library of every single gene that unites us and mice. The process is a bit more complicated — some genes encode for more than one thing, and other genetic actors like transcription factors can also alter how a gene behaves. Still, by 2020, the group hopes to fully understand the 20,000-gene bridge between us and our lab experiments.
“If we don’t have the fundamental knowledge of what these genes do it’s very hard to interpret the pathology that could be associated with these mutations in people,” says Lloyd. “So that’s why this project in the mouse is so important to advance the initiative and our understanding of human beings.”
Probably Here to Stay
Mice may not always be the perfect model for understanding our own bodies, but they do hold real value to researchers. In fact, they are uniquely suited for the kinds of high-throughput screening experiments that are common in the lab these days. Their usefulness will only increase as scientists devise better ways to raise and modify them to compensate for their shortcomings. The crux of the issue may be that researchers need to be more critical as they evaluate the type of model that is best for their experiment, instead of resorting immediately to what has worked before.
In the end, the story of scientists’ relationship with lab mice will likely be one of incremental advances to improve a flawed system. Fittingly, this is exactly how most scientific research progresses — by critically examining what we know to expand our knowledge of what we don’t.