The “elixir of life” is rapamycin, an immunosuppressant used to prevent transplanted organ rejection. The maximum human daily dose for rapamycin is 5 mg. To receive the same relative amount as the mice were fed in the study, people would have to ingest 200 mg daily. This means that anyone who takes “longevity” doses of rapamycin had better live in a controlled, sterile environment. It also means that they’ll be prone to developing tumors. This is not a problem for mice, in whom cancer can be easily cured. Not so, by a long shot, for humans.

Mice are extremely useful for dissecting molecular pathways and fairly useful as models for a long list of diseases. Their advantages are many and obvious: they’re mammals, so they share more with humans than the other organisms used in genetics — yeast, flies, worms, fish. Their generation times are short, their litters large. We can totally control their environment, though this affects their health and behavior. We can alter, delete or insert genes in mice and regulate when and where they express. And mouse research is far less emotionally and ethically fraught than experiments on primates.

However, as I never tire of saying, the devil’s in the details. Mice are sufficiently different that drug studies, in particular, are better as proofs of principle than one-on-one mappings. Unlike us, mice are fast-metabolizing and short-lived. Drugs, nutrients and hormones are radically divergent in every mammal in terms of bioavailability, activity, metabolism and toxicity. Many drugs that showed exceptional promise in mice fared no better than placebos or essentially acted as poisons in human clinical trials. The social and behavioral wiring of the two species are totally disparate. For example, mouse sexual arousal and mothering behavior depend on smell pheromones. Mouse models of complex diseases are prone to over- or mis-interpretation. Does a compound stop a mouse from jerking uncontrollably because it’s a pain killer — or is it acting as a muscle relaxant? It’s even harder to correlate such complicated responses as learning, anxiety or depression.


Cartoon credit: Sidney Harris

At the molecular level, many genes that are essential in mice are non-essential in humans and vice versa. Comparisons between the human genome and its recently completed mouse equivalent show that they diverge far more than previously recognized — close to 20%. Not only are there significant numbers of human-specific genes; equally importantly, genes whose structure is identical in the two species are nevertheless differently regulated. And last but decidedly not least, the two species have distinct regulatory RNAs (miRNAs for the truly curious).

All this may sound esoteric, but the conclusion is simple: mouse models, especially those based on highly inbred strains, recapitulate only limited aspects of many human functions and diseases. This is particularly true of metabolism, neoplasty, lifespan and that troublesome jewel in the crown — the brain. Elixirs and panaceas for obesity, cancer, longevity and dementia remain stubbornly species-specific. So next time a brushfire sweeps the Internet with news of yet another miracle cure, the first and last question of non-biologists should be: what were (or when will we know) the results of the human tests?