Mice are not men


Is it the mouse’s fault if mice are not men? Glue Grant researchers conducted a rigorous, systematic evaluation that demonstrates what most investigators already know from their experiences – current mouse models poorly reflect human immune diseases. Yet, mouse efficacy models continue to be used to create and test candidate drugs that can cure diseases like sepsis in mice, but not in patients.

Study results from mouse disease models have dominated the scientific literature in recent decades. A mouse model of a disease is very helpful to help understand certain features of the human condition it resembles – features that might not be safe to evaluate directly in humans. In the drug testing process, animal models, commonly mouse models, are often used for efficacy testing to study the candidate drug’s capacity to produce the desired effect(s) by screening the drug before it is tested in humans.  Read More

Failed trials in the field of inflammation

However, few of the follow-on human studies based upon mouse disease models have shown success in creating new drugs to treat or cure inflammatory diseases in patients. For example, there have been nearly 150 clinical studies testing candidate drugs intended to block the inflammatory response in critically ill patients, and every one of them failed. Why? Most simply stated, mice are not men.

Clinicians and researchers know that human immunology is complex. Systemic inflammation, a condition present in many diseases and following traumatic injury, causes significant complications in patients and contributes to hundreds of thousands untimely deaths each year. The scientific community and public regulators (like the Federal Food and Drug Administration, as one example) have assumed, in the absence of systematic evidence, that results from animal research may necessarily reflect human disease. The good news is that Glue Grant researchers and others are beginning to identify some of the individual genes and gene groups that play important roles in how the human body responds to systemic inflammation. Some of the gene patterns help predict who would survive after life-threatening injury and which patients would go on to develop complications in the intensive care unit.

How well do mouse models mimic human inflammatory diseases?

&copy 2013 Seok et al. Originally published in the Proceedings of the National Academy of Sciences of the United States of America. 110:3507-12. doi: 10.1073/1222878110

The responses in corresponding mouse models correlate poorly with the human conditions, and also, one another

Ten years of accumulated data from the NIH large-scale, collaborative research program − involving many different disciplines including clinical medicine, animal science, genomics, immunology, molecular and cellular biology, bioengineering, computational biology, and biostatistics − allowed investigators for the first time, to systematically compare the white blood cell transcriptome (sequences of RNA that tell us when and where each gene is turned on or off in the cells) in three acute, severe inflammatory disease mouse models (trauma, burns, and endotoxemia, a condition that acts like sepsis) with the human diseases these models are designed to mimic. In a broad evaluation of all known genes, investigators performed time course comparisons and compared multiple characteristics of the response in humans and mice.

The results showed that, in humans, the genomic responses in patients correlated well with each other, i.e., different severe inflammatory stresses produce similar genomic responses, which likely indicates the same genes or gene groups are involved in the response. Most importantly, investigators were surprised by their results showing the poor correlation (close to random) between the genomic responses in the mouse models and those responses in human injuries, especially given the world-wide prevalence of the use of mice to model human inflammation. When multiple datasets of other acute inflammatory diseases were analyzed, the same lack of correlation between human and mouse responses was found.

Focusing on complex human diseases

A mouse is not a man

Reprioritizing the infrastructure, resources, tools, and methodologies to study human inflammatory diseases in humans

This systematic comparison of the genomic response between three human inflammatory diseases and their corresponding mouse models has the potential to change the way life science is performed.  Glue Grant investigators contend that systematic evaluations like this one on how well the genomic responses in mouse models reflect human diseases is critically important to explain why molecular results from current mouse models fail to routinely translate directly to human conditions, since virtually every FDA-approved drug as well as every drug candidate functions at the molecular level.

This systematic evaluation picked up by The New York Times has been controversial, widely quoted, and has sparked wide scientific debate about the reliance on mouse models to reproduce diseases for drug discoveries. “The study’s findings do not mean that mice are useless models for all human diseases,” says NYT author Gina Kolata. In the investigators’ view, it is not the mouse’s fault, but rather the lack of understanding about the essential physiology that should be represented in the mouse model. To correct this lack of understanding about human diseases, researchers should be placing more and more emphasis to be sure that we understand human disease actually by studying humans. With a better understanding of the human disease, researchers can then be sure to recreate a more comparable environment in the mouse to more closely mimic the human disease on its molecular basis.

Relevant publications

Seok J, Warren HS, Cuenca AG, Mindrinos MN, Baker HV, Inflammation and Host Response to Injury, Large Scale Collaborative Research Program. Genomic responses in mouse models poorly mimic human inflammatory diseases. Proc Natl Acad Sci U S A. 2013 Feb 26;110(9):3507-12. PubMed PMID: 23401516; PubMed Central PMCID: PMC3587220

Kolata G. Mice Fall Short as Test Subjects for Some of Humans’ Deadly Ills. The New York Times. February 11, 2013.  Accessed January 26, 2015

Warren HS, Tompkins RG, Mindrinos MN, Xiao W, Davis RW. Reply to Cauwels et al.: Of men, not mice, and inflammation. Proc Natl Acad Sci U S A. 2013 Aug 20;110(34):E3151. PubMed PMID: 24137663; PubMed Central PMCID: PMC3752239

Tompkins RG, Warren HS, Mindrinos MN, Xiao W, Davis RW. Reply to Osterburg et al.: To study human inflammatory diseases in humans. Proc Natl Acad Sci U S A. 2013 Sep 3;110(36):E3371. PubMed PMID: 24137798; PubMed Central PMCID: PMC3767510

Warren HS, Tompkins RG, Moldawer LL, Seok J, Xu W, Mindrinos MN, Maier RV, Xiao W, Davis RW. Mice are not men. Proc Natl Acad Sci U S A. 2014 Dec 24. pii: 201414857. Epub ahead of print.  PubMed PMID: 25540422


Ronald Tompkins, M.D., Sc.D. H. Shaw Warren, M.D. Wenzhong Xiao, Ph.D.
rtompkins@mgh.harvard.edu hswarren@mgh.harvard.edu wenzhong.xiao@mgh.harvard.edu
617-726-3447 617-724-5774 617-724-7261

See Less

andrwkirkMice are not men