Why “Move Over, Mice”?
We push the frontiers of creating and deploying what NIH calls “NAMs” (Novel Alternative Methods): user-friendly human “physiomimetic” models, blending systems biology models of patients with microphysiological systems that act as living patient avatars. How can you possibly capture the essential pathophysiology of a complex systemic disease like endometriosis, chronic Lyme or TB in vitro, even on a sophisticated microfluidic platform? It’s iterative! Hard part: Conceptualize the biology and frame good questions. Also hard part: build very robust technologies to (i) capture the relevant tissue biology and (ii) read out the consequences of perturbations and interventions. And yet another hard part: translate in vitro results back to in vivo to demonstrate meaningful impact on understanding or treating a disease (see graphic).
The pace of access to high-resolution multi-omics data sets on well-phenotyped patients continues to accelerate, giving the impression that “answers” are lurking for complex problems if we “just apply AI”. Hypotheses can (at times) be generated – but they must be validated functionally. An increasing number of diseases are being linked to genetic variants of common genes, often immunological in ways difficult or impossible to replicate adequately in animal models. Enter well-designed NAMs – some are ready for prime time, but most need further development. Pain? Even pain is in the crosshairs of human models – not yet the whole tissue-spine-CNS (immune) circuit yet, but the parts are shaping up. Some surprisingly sophisticated information can arise from simple in vitro models – if the upfront biological conceptualization of the pathophysiology is clever. Further, MIT is at the forefront of “species translation” (will provide a link), devising mathematical models to glean ways that the copious existing animal multi-omics data – as well as in vitro avatar data – can be translated to human patients.
Our goal with MOM is to build a community of partners to journey with us to the promised land: accessible, validated, human physiomimetics. At MIT, this means our Division of Comparative Medicine, which now oversees animal use, will also house a core facility of living patient avatars and train the entire community on experimental design, implementation, and interpretation. This core will partner with MOM labs all over MIT to keep infusing the core with the newest technologies and approaches, whether developed at MIT or elsewhere.
Move Over Mice: A gradual transition, not a coup.
We will be stuck with mice – and other inadequate models – forever if we do not invest in development of human NAMs. Consider: “Horsepower” entered the thermodynamics lexicon because once upon a time, before internal combustion engines (and later electric engines) transformed the way motion-reliant processes derived power, actual horses were the engines of work, and rate of energy generation calibrated to that of a horse communicated the utility of new methods in a common vernacular for the times. Horses still have utility in the modern world, but our food is no longer produced by horses turning a millstone, except in quaint settings.
Animal models, from worms to flies to mice to monkeys, have provided stunning insights into molecular-systems physiology and pathophysiology of humans for centuries. The era of animal model utility in biomedical research is not over. But it is waning, as animal models may succeed in capturing general phenomena seen in humans – e.g., studies in c elegans revealed the process of apoptosis – but often fail to capture crucial particulars required for therapeutically relevant human translation. In 2020, NIH commissioned an Advisory Committee to the Director Working Group to analyze use of animal models in NIH-funded research, citing as motivation “About 40% of our R01 grants use murine models, creating a substantial opportunity cost, if some models poorly represent the corresponding human disease”. One example: the Juvenile Diabetes Research Foundation (JDRF) created a ground-breaking human cadaver pancreas collection program in 2007, without help from the NIH, explaining “critical questions regarding [Type 1 diabetes] disease pathogenesis are specific to humans and cannot be easily investigated in experimental animals”. Prof Griffith summarized this and additional problems with animal models of human Type 1 diabetes and other disease in a presentation to the NIH WG in 2020.
Mice will move over, but never leave entirely.
A co-leader of our MOM program is a veterinarian – the MIT DCM Director Kelly Pate, who studies platelets in infectious diseases. Even if we were to devise perfect human NAMs, human lives are inextricably intertwined with those of animals: they are our pets, we raise them for food, and they are used commercially in contexts from goats for weed control to bees pollinating our crops. Animals in the wild carry pathogens that can infect humans, or animals on which humans depend. Human health thus demands that we understand and perfect animal health.
