Strap in ‘cos at ~2000 words this one is fairly long.
The notion that mice could handily emulate human physiology has always strained credulity. After all, a mouse is a short-lived, four-legged, nocturnal, burrowing rodent. It beggars the imagination that the evolutionary selection forces that shaped its physiology remotely resemble those that shaped that of humans. It then follows that mouse physiology likely offers little by way of concordance to that of humans. And yet since post-WWII, the role of mice, and to a lesser extent, rats, has only become more deeply embedded in biomedical research.
There are two parts to this predominance of mice and other rodents in experimental biomedical research; one, how this came about in the first place, and two, why it continues to remain the norm. As is often seen in history, the first resulted from the accidental confluence of a series of historical precedents in the first part of the 20th century while the second is dictated by prevailing cultural and economic imperatives in biomedical research, a process where the longer it continues, the more the ensuing historical precedence assumes perceived importance. A utilitarian mindset that presumes the expediency of such animals also helps sustain the process.
Amateur mouse fanciers turned breeders, the rediscovery of Mendel, the birth of mammalian genetics, the creation of the modern scientific infrastructure: the chain of early 20th century events that led to mouse predominance in experimental biomedical research
In the early decades of the 20th century, Abbie Lathrop, a mouse and rat fancier, was breeding fancy mice and rats on her Massachusetts farm and unusually for that time and especially given her non-scientific background, she kept excellent breeding records to boot.
Around the same time, after languishing in obscurity for >3 decades, the rediscovery of Mendel‘s now famous pea breeding experiments jump-started American genetics research, creating a confluence of right resources, right finding and right timing since Lathrop’s treasure trove of inbred and quasi-inbred mouse strains proved a godsend for mammalian genetics pioneers such as William Ernest Castle who were then just beginning their careers. In turn, Castle trained the next generation of world famous mammalian geneticists such as George Snell and Clarence Cook Little.
Little was an excellent entrepreneur who took some mouse colonies from Lathrop’s farm and ended up creating a rodent breeding facility at Bar Harbor, Maine, that over the subsequent decades morphed into the Jackson Laboratory, even today a pre-eminent global source of lab mice, rats and other experimental animals.
Snell’s research on those pioneering inbred mouse strains led to the discovery of the. Having inbred mice with a single MHC haplotype proved not just crucial for this discovery but was also handy to use to maintain segregated mouse colonies, a key tool necessary for reducing confounding variables in biological experimentation.
Meantime, the US emerged post-WWII as the wealthiest nation in the world and set about establishing the scientific research infrastructure and agenda that pretty much every other country ended up replicating in some shape or form.
- Vannevar Bush laid the groundwork for an unprecedented policy of government funding of scientific research and development in Science, the Endless Frontier, his 1945 report to the US president.
- Implementation of this policy over subsequent decades made the NIH and, to a lesser extent, the then-newly formed NSF, behemoths of US biomedical research funding. Such R&D funding at US universities transformed them into preeminent centers of scientific research.
- Legislative acts such as the GI bill further empowered US universities to train and graduate a scientifically trained workforce, a process that has only grown in the subsequent decades creating a workforce glut that academia can barely absorb.
Guinea pigs had come to dominate biomedical research from the late 19th century. However, guinea pig colonies were expensive to maintain; they bred in small numbers, availability of few inbred strains meant that uncontrollable variables were introduced into each experiment, their housing and feed were expensive. In contrast, mice are relatively hardier, and easier to breed, feed and house.
Mouse (and rat) animal husbandry thus proved far easier to standardize at an industrial scale and replicate in universities and other research institutions in the post-WWII years. The knowledge, expertise and influence of people like Little played no small role in the spread of mouse and rat as the models of choice in experimental biomedical research.
Back then the American economic model was much better at fostering competition so other lab animal suppliers such as Charles River, Harlan and Taconic soon emerged, creating the current industrial pipeline of lab animal supply, consisting largely of mice and rats, that currently dominates biomedical research.
In short, the fundamentals favored post-WWII predominance of mice and rats in biomedical research. Where the earlier guinea pig predominance left its mark on the English language, they have themselves long since ceded ground to mice and rats (below from).
In a way, the baton passing from guinea pig to mouse as the pre-eminent animal model also represented the torch passing from Europe to the US as the global biomedical research leader. Where even into the 1920s, widespread acceptance of research findings required the imprimatur of European researchers, typically French or German, post-WWII till date it’s the US all the way.
Cul-de-sac or the mouse that roared: Be it ever so counter-productive, prevailing historical, cultural and increasingly economic imperatives dictate mouse predominance in experimental biomedical research
Follow the money. Cliched, overused and yet it remains on the nose for explaining mouse predominance in biomedical research.
The industrialization of molecular biology since the 1980s has only served to further accelerate and deepen this predominance. For example, as the technological capability to knock-out and knock-in genes became easier starting in the late 1980s, the trend began among the powerful stakeholders within the scientific community such as the granting agencies, and scientific journals and their peer reviewers to initially seek, then to insist and lastly to demand that results of in vitro cell culture knock-out and knock-in studies be validated in vivo.
The by then well-established industrial pipeline of inbred mouse strains easily stepped in as the obvious choice for such in vivo tests. That started off a new wave of creating mouse strains with various genes knocked-out or knocked-in, a trend that continues unabated till date.
Thus as molecular biology took off, mouse animal husbandry expanded, further cementing its predominance, creating hundreds of new genetically engineered mouse strains, strains whose study now consumes the entire careers of thousands upon thousands of biomedical researchers studying various physiological systems.
Biochemists to endrocrinologists to all manner of other specialists who’d previously used molecular biology tools to study their favorite molecules and processes at the cellular level using in vitro cell lines now swarmed into the mouse husbandry business, creating unique mouse strains, maintaining their colonies and studying them while specialists in other fields such as developmental biology and immunology had already glommed on to the mouse model decades earlier.
While developmental biologists still hedge their bets and continue using fruit flies, nematodes and even zebrafish as model experimental organisms, from about the 1970s many immunologists went all in wedding themselves to their mouse models so much so that grotesquely, knowledge of mouse immunology is today miles ahead of knowledge of human immunology.
To add insult to injury much of that vaunted knowledge rests on shockingly shaky grounds. Consider just one example, perhaps the most important of all, which stems from the most popular mouse strain used in all of biomedical research in general and in immunology in particular. This turns out to be a strain called C57Bl/6 (). Prone to obesity, it entirely lacks a whole MHC molecule, the one called I-E, due to a translocation event at some point in its breeding history. Thus, its immune responses could in no way, shape or form be considered representative of those of even a prototypical inbred mouse strain, let alone be considered an adequate surrogate for any human immune response.
Of course the most pernicious and stickiest ingress of the mouse and other rodent models proved to be in the drug development and drug approval processes, a choice sped up in the wake of the thalidomide scandal in the early 1960s, with the 1962 passage and implementation of the, which codified and granted the FDA the authority to approve new drugs for marketing and sale. Rather than their scientific suitability, a cultural dependence on preclinical rodent models has since driven regulator preference for such data and of course, the longer this goes on, the longer the weight of historical precedence accrues perceived importance.
We humans are preternaturally talented at outsourcing. Among ourselves, we effortlessly contrive convoluted and seemingly inconspicuous processes that unerringly steer the most undesirable jobs in the economy towards those deemed most undesirable, the poorest and most otherized, among them refugees and immigrants.
I’m sure we would likewise seek to outsource our diseases just as easily if we could but we can’t. And yet in our uniquely muddled way we’ve chosen to outsource as much of the risk in drug development to poor benighted animals who have no say in the matter and decidedly cannot provide informed consent, or at least not in a manner or language we could comprehend; muddled because no animal could conceivably be a scientifically adequate surrogate for human physiology, and yet we unabashedly choose to wallow in such a fanciful notion.
Why do we do this? Same answer as for so many other human decisions, because we can, being a species just smart enough to cause real and lasting havoc.
Thus, mouse strains serving as surrogates for human physiology, regardless that such extrapolations be ever so tenuous, has only become more deeply embedded in biomedical research, with newcomers simply unquestioningly learning the ropes from earlier generations and taking our almost single-minded reliance on mouse models to ever newer heights.
Every now and then, a certain amount of caterwauling about the lack of suitability of mice as experimental models bubbles up in the scientific literature. Issues include the inherent limitations of inbred mouse strains in being able to recapitulate basic wild mouse physiology, let alone their ability to emulate those of their human counterparts, how mice are housed, what they are fed, how lack of housing “enrichment” prevents a full manifestation of their “mouse” traits such as foraging.
In addition, everything from cage design, ventilation systems, light cycle, temperature and humidity settings to diet to how they are handled are just a few of the factors that profoundly influence the readouts of every single lab mouse experiment. But then prevailing economic, read scientific career, imperatives being what they are, things settle back to the uneasy status quo of mouse predominance.
This system is now creaking, sagging and almost hopelessly broken. Mouse husbandry costs are considerable for any biomedical research lab, even more so in this era of ever-shrinking research funding, which only intensifies competition.
Rigorous statistical procedures now considered the norm in clinical research such as randomization and blinding continue to remain as alien in experimental animal research as proverbial aliens on planet earth. Even the most basic of standard statistical analyses routine in clinical research such as meta-analyses and systematic reviews remain impossible for experimental animal research data simply because experimental procedures and data collection choices remain unstandardized and thus vary vastly between labs and often even between different experiments within the same lab.
Ever increasing pressure to publish plus tight budgets mean that not enough mice per group and not enough repeats of any given experiment have become more the norm than outlier. The possibility that things are stretched almost to breaking point shows up in the form of shoddy, irreproducible data (below from).
Astonishingly the economics of animal research remain hardly explored even though they could potentially explain every relevant issue about modern biomedical research. Even in that, EU numbers are somewhat easier to find compared to those in the US since the US Animal Welfare Act excludes mice, rats and fish, though it does have other regulations to ensure animal welfare. Thus, of the ~11.5 million experimental animals used in the EU in 2011, ~61% were mice and ~65% of total animal usage was for basic R&D (below from).
3. Hartung, Thomas. “Food for thought look back in anger–What clinical studies tell us about preclinical work.” Altex 30.3 (2013): 275.
4. Meigs, Lucy, et al. “Animal testing and its alternatives–the most important omics is economics.” ALTEX-Alternatives to animal experimentation 35.3 (2018): 275-305.