About this Author
DBL%20Hendrix%20small.png College chemistry, 1983

Derek Lowe The 2002 Model

Dbl%20new%20portrait%20B%26W.png After 10 years of blogging. . .

Derek Lowe, an Arkansan by birth, got his BA from Hendrix College and his PhD in organic chemistry from Duke before spending time in Germany on a Humboldt Fellowship on his post-doc. He's worked for several major pharmaceutical companies since 1989 on drug discovery projects against schizophrenia, Alzheimer's, diabetes, osteoporosis and other diseases. To contact Derek email him directly: Twitter: Dereklowe

Chemistry and Drug Data: Drugbank
Chempedia Lab
Synthetic Pages
Organic Chemistry Portal
Not Voodoo

Chemistry and Pharma Blogs:
Org Prep Daily
The Haystack
A New Merck, Reviewed
Liberal Arts Chemistry
Electron Pusher
All Things Metathesis
C&E News Blogs
Chemiotics II
Chemical Space
Noel O'Blog
In Vivo Blog
Terra Sigilatta
BBSRC/Douglas Kell
Realizations in Biostatistics
ChemSpider Blog
Organic Chem - Education & Industry
Pharma Strategy Blog
No Name No Slogan
Practical Fragments
The Curious Wavefunction
Natural Product Man
Fragment Literature
Chemistry World Blog
Synthetic Nature
Chemistry Blog
Synthesizing Ideas
Eye on FDA
Chemical Forums
Symyx Blog
Sceptical Chymist
Lamentations on Chemistry
Computational Organic Chemistry
Mining Drugs
Henry Rzepa

Science Blogs and News:
Bad Science
The Loom
Uncertain Principles
Fierce Biotech
Blogs for Industry
Omics! Omics!
Young Female Scientist
Notional Slurry
Nobel Intent
SciTech Daily
Science Blog
Gene Expression (I)
Gene Expression (II)
Adventures in Ethics and Science
Transterrestrial Musings
Slashdot Science
Cosmic Variance
Biology News Net

Medical Blogs
DB's Medical Rants
Science-Based Medicine
Respectful Insolence
Diabetes Mine

Economics and Business
Marginal Revolution
The Volokh Conspiracy
Knowledge Problem

Politics / Current Events
Virginia Postrel
Belmont Club
Mickey Kaus

Belles Lettres
Uncouth Reflections
Arts and Letters Daily
In the Pipeline: Don't miss Derek Lowe's excellent commentary on drug discovery and the pharma industry in general at In the Pipeline

In the Pipeline

« A Look Under the VC Hood For 2013 | Main | Press Release Fraud - The Verdict Stands »

December 18, 2013

Lab Mice Are Being Kept Too Cold, Apparently

Email This Entry

Posted by Derek

Earlier, the unsuitability of mice in inflammation models was shown in a paper that should have been noted by anyone in the field. Just last month, a paper in Science detailed the problems with many animal studies (mouse and otherwise), particularly the smaller ones, which can suffer from bad statistics and poor protocols.

Now we have this, from PNAS. The authors, from the Roswell Park Institute and the EPA, say that standard rodent facility conditions are actually causing unintended chronic physiological stress:

We show here that fundamental aspects of antitumor immunity in mice are significantly influenced by ambient housing temperature. Standard housing temperature for laboratory mice in research facilities is mandated to be between 20–26 °C; however, these subthermoneutral temperatures cause mild chronic cold stress, ac- tivating thermogenesis to maintain normal body temperature. When stress is alleviated by housing at thermoneutral ambient temperature (30–31 °C), we observe a striking reduction in tumor formation, growth rate and metastasis. . .Overall, our data raise the hypothesis that suppression of antitumor immunity is an outcome of cold stress-induced thermo- genesis. Therefore, the common approach of studying immunity against tumors in mice housed only at standard room temperature may be limiting our understanding of the full potential of the antitumor immune response.

As mentioned in that last line, the problem seems to be with the adaptive immune system - this effect is driven by CD8+ T cells in almost every case, and sometimes by changes in CD4+ cells as well. Overall, housing mice at the recommended temperatures, which are on the cool side, seems to promote a general immunosuppression, which I think it's safe to say is not a factor that many people are taking into account. The animals have similar core body temperatures, but the extra burden of maintaining that in the cooler rooms is tipping some sort of balance - keeping all those immune systems running is apparently energetically costly, and they get downregulated.

This study looked at several sorts of tumorigenesis, but only for solid tumors, so the effects on leukemia, etc., are still unknown. You'd have to think, though, that several other disease areas could be affected by this situation as well - for example, how much of the uselessness of mice in inflammation models is caused by these changes? I'm simultaneously glad to see these things being uncovered, while being worried about how long it's taken to uncover them: what else are we missing?

Comments (43) + TrackBacks (0) | Category: Animal Testing | Cancer


1. sepisp on December 18, 2013 8:59 AM writes...

Wouldn't this royally screw up all carcinogenity data up to date? In the State of California?

Permalink to Comment

2. Arty Facts on December 18, 2013 9:14 AM writes...

This is one reason I really like the Jax outbred strain idea (no financial ties). There are just too many stressors, and it seems to me that outbreeding is a way to improve robustness through hybrid vigor.

Other interesting stress factors I know about include travel and handling stress (obviously); pheromone exposure from bedding and adjacent cages; differential responses to anaesthetics (especially halogenated inhalants); noise stress; circadian disruption; exercise levels; and chow quality/supplier. In my experience these are quite impactful in immune function studies. Intra-strain variability in Tregs and DC subsets is already far worse than most people appreciate, now layer on some stress...

Now I'm thinking it's worth considering other factors. A few ideas: degree of environmental enrichment (toys); hiding places for when they get startled (rats like to hide near something large when startled); co-habitation with strangers v. littermates; weaning age and age of separation from their mother; chronic, low level noise and noise out of human hearing range (think HVAC again); lighting type (wavelength, oscillation frequency); handler scent and pheromones (do mice get white coat syndrome?); water quality and taste; cage size; dish washing soap choice. One could go insane with the possibilities, but I'd bet sound, scent, and perceived danger from predators seem likely to be significant.

Permalink to Comment

3. BigB@UNC on December 18, 2013 9:27 AM writes...

This is one of the most basic experiments that should have been run, oh, I don't know, in 1890... Every chemist knows one must control, or at least consider, temperature... why wouldn't toxicologists and biologists do the same for lab rats???

And Sepisp's comment is certainly valid - but the State of California has bigger concerns - this will probably never see the light of day in California courts.

Permalink to Comment

4. Chemjobber on December 18, 2013 9:27 AM writes...

"And next up, after this break, a study that shows that you should turn up your thermostat -- your immune system depends on it!"

Permalink to Comment

5. exGlaxoid on December 18, 2013 9:51 AM writes...

Now my wife will want to turn up the heat even more in winter. But I guess I can use the money I save on healthcare bills to pay the utility company...

Permalink to Comment

6. Jim on December 18, 2013 10:10 AM writes...

The interpretation and next steps that follow this are crucial. So temperature has an effect, but what is the best model system to use? One that is the most stringent, one that is most like humans, both? Will any of those ultimately matter towards the predictive validity of the models? I know that sounds obvious, but watch people react to this paper saying that all the old data are now worthless. Maybe, but that's not what's shown here.

Permalink to Comment

7. Wile E. Coyote, Genius on December 18, 2013 10:25 AM writes...

Before we throw out all data ever generated from animal toxicology: So the paper is about mice. What is the thermoneutral temperature range for dogs, for monkeys, for rats? Are they the same? Does this mean that rats housed in room 1 need a different temperature range than the mice housed in room 2 next door? Is that acceptable for the dogs housed in the next hall and the monkeys in the hallway past that? How about the rabbits used for repro tox also housed in this theoretical animal facility? Seems to raise some rather important facility management and logistical questions.

@3, these toxicology studies are controlled for temperature, in the sense that all the animals on one study (control group and dosed groups) are in the same animal room and therefore all have the same exposure to ambient temperature. The optimum temperature to house animals is not a question of proper experimental controls, but instead are questions of artifacts of the test system. Tumor development in the controls and experimental groups housed in the same environment should be valid. Or do we need to run a bunch of 2 year carci studies with matched control and dosed groups at different temperatures to determine if this is an effect and validate the appropriate temperature? Could this vary by strain and tumor type, such that some tumors develop more readily at one temperature than another in one strain than another? What does that mean for interpretation of the carcinogenicity risk identification/risk assessment? Those are relatively simple and inexpensive things to determine (sarcasm alert).

Permalink to Comment

8. MDACC Alum on December 18, 2013 10:31 AM writes...

This is a consequence of cost cutting. The air going into these rooms is cooled to pull out the moisture and then heated back up. When I was at MDACC we saw that when the administrators wanted to save some money that month they wouldn't allow the air to be brought back up to full temp.
I found it annoying, since we were doing cancer and a big issue is weight loss (in both human patients and mice) and they get cold extremely fast.

Permalink to Comment

9. luysii on December 18, 2013 10:45 AM writes...

Look at the paper carefully. The paper is from Atlanta. The mice down there are kept at temperatures of 20 - 26 C (68 - 78 F). They define ambient temperature as 30 - 31 C (86 - 88F), MUCH higher that most of us live. The old work is still valid IMHO, and theirs is somewhat suspect.

They are saying that 68 - 78 F causes thermogenesis from brown fat in the mice -- really? I don't any of us would define these temperatures as cold stress (which they do)

Permalink to Comment

10. sep332 on December 18, 2013 11:11 AM writes...

BigB@UNC: They have been controlling for temperature. The rats have been kept at 20-26C consistently.

Permalink to Comment

11. Anonymous on December 18, 2013 11:55 AM writes...


Permalink to Comment

12. johnnyboy on December 18, 2013 12:37 PM writes...

The fact that the recommended housing temperatures for mice are too cold for them has been known for a while, as are all the other potential stressors helpfully listed by #2. The reason these haven't been changed are in large part due to human factors, ie. people don't want to work at uncomfortable temperatures, or work at night (rodents are nocturnal), etc... It's unfortunate, but that's the reason you have control groups.

Permalink to Comment

13. Lu on December 18, 2013 4:00 PM writes...

Forget mice, my cat needs 30 C as well. Last couple of weeks he is not leaving my lap at all.

Permalink to Comment

14. NJBiologist on December 18, 2013 4:00 PM writes...

@9 luysii: You may be thinking of the old definition of thermoneutral (temperatures at which heat production equals heat loss, used to generate the Guide range). The 30degC number seems to be based on the range of temperatures at which VO2 is flat.

Permalink to Comment

15. Steve on December 18, 2013 4:16 PM writes...

What else are we missing? The fact that animals aren't people.

Permalink to Comment

16. bank on December 18, 2013 4:27 PM writes...

The image linked to below shows the natural range of Mus musculus. It extends from the arctic to the equator. So I personally find it hard to believe that one particular temperature in the animal facility is more representative than another for the "natural" temperature that a "wild-type" mouse lives at (and by extension the "normal" activity of whatever aspect of physiology is being investigated, in particular if there are only a few degrees difference.

Permalink to Comment

17. Jim on December 18, 2013 4:35 PM writes...

@bank: remember, we're not trying to mimic the natural state of mice, we're trying to model a human disease in mice and assess therapies for that disease. So, if temperature affects immune response, we want to pick the best temperature for making a rodent immune system respond as closely as a human's immune system would. Frankly, given the robust immune system in rodents, depressing it might not be a bad idea. But that's the problem...we don't know if it's a "good" effect or a "bad" effect.

Permalink to Comment

18. Terry on December 18, 2013 5:36 PM writes...

@Jim: Bingo! for the win. We want the animal model to be in the best possible condition to predict human biology. Cold-stressed mice might significantly diverge from non-stressed humans. In addition to the immune system, metabolic activity is significantly altered in these mice (see thermogenesis and uncoupling protein 1). The fact that the temperature is selected in part for the convenience of the animal husbandry personnel should provide a clue to question just how we set up models and protocols. Think of growth curves that sample at 1, 2, 4, 8, and 24 hours. What happens 16-24 hours? Investigator goes home and sleeps, leaving unobserved any events during that period. I've seen many animal models selected because they make the investigator's life more predictable, and not necessarily for predicting translational results between species.

Permalink to Comment

19. anony-mous(e) on December 18, 2013 7:09 PM writes...

Just imagine the impact temperature has on all your Metabolic targets kids !!!!!!!! We've known about this for years, and yet continue to do NOTHING about it where I work.

Permalink to Comment

20. newnickname on December 18, 2013 11:03 PM writes...

We had a small animal facility but no one was allowed near the animals except the authorized handlers; not even to 'show off' the mice to visitors (i.e., potential investors). There was a small (8"x8"?) window but peeping wasn't allowed. Management was conducting their own experiment on us lab rats by keeping the labs on the chilly side, exclusive of all the other stressors.

Permalink to Comment

21. sepisp on December 19, 2013 3:13 AM writes...

>Tumor development in the controls and experimental groups housed in the same environment should be valid.

Yes, but there are two important consequences:

1) Extrapolating to humans just got a bit harder.
2) Regarding the previous, there are no guarantees that the response is linear. For instance, it could follow catastrophe theory instead. That is, there is a trigger level of stress (either magnitude, time distribution or some combination), and we're moving it around with the extra cold stress, with chaotic consequences. (I don't say random, that implies nondeterminism. Deterministic systems that are impossible to model analytically are called chaotic.)

Tumors would be expected to appear earlier, more often and would have more catastrophic consequences. Even if we assume each of these is some simple function, we've added 4-6 parameters (degrees of freedom) to the model already.

Unfortunately due to the lack of Mengeles it's extremely unlikely that a control experiment in humans could be run.

Another unfortunately, there most probably won't be any funding for a control study, since it doesn't create "discoveries". In this way process chemists are more fortunate.

Permalink to Comment

22. Dee on December 19, 2013 3:31 AM writes...

There's another angle to look at this, i.e. to what extent lab conditions are representative of experimental animals' (or rather their relatives' in nature) habitat. It's probably safe to say that they experience "subthermoneutral temperatures" most of the time which is one of the reasons why a lot of time is either spent on searching for food or cuddling in the nest with the relatives. Also, other confounders like e.g. free access to food may well be more important than the mild temperature stress (cf discussions suggesting that most laboratory animals have features of metabolic syndrom).

Permalink to Comment

23. Lars on December 19, 2013 7:02 AM writes...

@Dee: Have you heard of the Rat Park studies?

Permalink to Comment

24. cliffintokyo on December 19, 2013 7:20 AM writes...

From a med chemist's perspective: for mice used in anti-cancer drug animal studies, perhaps the optimum temperature should be that at which the control animals showed greatest ability to resist growth of the cancer, in order to give our test compounds the toughest hurdle to cross; fail early and fast and all that. Perhaps a biologist/pharmacologist could comment on feasibility?

Permalink to Comment

25. newnickname on December 19, 2013 8:57 AM writes...

In his "Cargo Cult Science" piece, Richard Feynman praises the extremely careful rat running experiments of (Paul Thomas) Young and, by implication, denigrates those that ignored Young's work when doing their own rat running experiments. He says, "But not paying attention to experiments like [Young's] is a characteristic example of cargo cult science."

Permalink to Comment

26. Red Fiona on December 19, 2013 11:23 AM writes...

Not an animal biologist (E. coli is as complicated as I get) but are there any studies on what the internal temp is in mouse/rat nests?

Permalink to Comment

27. NJBiologist on December 19, 2013 12:14 PM writes...

@26 Red Fiona--Healthy normal rats run right around 37degC.

Permalink to Comment

28. Red Fiona on December 19, 2013 1:03 PM writes...

@27 Sorry I meant inside their nests, or hollows or burrows or what have you.

Permalink to Comment

29. Anonymous on December 19, 2013 2:57 PM writes...

Gas mark 4.

Permalink to Comment

30. matt on December 21, 2013 5:29 PM writes...

@bank #16: who said these were vigorous wild mice? These are mice carefully and completely inbred to represent rigid standardization. Where, in the biological world, do you find such total consistency? See #2 Arty facts on the benefits of hybrid vigor.

Will this make any difference on animal handling? I'd hope so, in the pharmaceutical world. Probably not a bit, in the academic world, because the incentives aren't aligned as well. No one has backed off mouse longevity studies, it seems, despite well-known confounders of overeating controls.

To what extent are the locked doors and limited access to animal testing facilities due to PETA and the fear of adverse publicity, rather than scientific rigor? I suspect lawyer FUD >> animal studies' correctness, but maybe that's only outside pharma.

Permalink to Comment

31. Menachem Begin on December 22, 2013 8:00 AM writes...

The Economist recently publish 'Trouble at the Lab' (mandatory reading) where it was explained that the rate of untrue conclusions in peer-reviewed journals has skyrocketed to something around 40%. Life sciences are the worst. Surveys to reproduce findings come up with poor reproducibility.

We all know that Science is self-correcting through the Scientific Method. Few of us realize the immediate implication that: Therefore, at any point in time, a significant fraction of accepted conclusions must be incorrect.

Frankly, the entire Scientific field has become a bit of a schmozzle lately. Especially the Life Sciences. It's embarrassing.

Permalink to Comment

32. Simon on January 6, 2014 7:23 PM writes...

Interesting finding as some other studies such as this science paper ( have found lowered core temp to be associated with longer life span (up to 20% longer!)

Permalink to Comment

33. psychic know on January 28, 2015 7:32 PM writes...

It's amazing to pay a visit this web site and reading the views of all colleagues concerning this post, while I am also eager of getting

Permalink to Comment

34. chronic kidney disease stage 1 on February 5, 2015 5:21 PM writes...

Common, generalized symptoms of multiple sclerosis are: pronounced fatigue, body weakness,
sensations of tingling, burning, pain, itching and numbness of
the muscles, loss of dexterity and uncoordinated body movement.

Dianna Malkowski is a Board Certified Physician Assistant and
Mayo Clinic trained nutritionist specializing in diabetes, cancer,
wound healing, therapeutic diets and nutrition support.
In fact, there ae some forms of yoga tat have been identified to gar towrds the limitations and abilities of patints with this disorder.

Permalink to Comment

35. eBay Selling Tips on February 5, 2015 5:28 PM writes...

Hurrah! At last I got a webpage from where I can genuinely obtain useful facts regarding
my study and knowledge.

Permalink to Comment

36. Michell on February 9, 2015 6:43 PM writes...

Thanks on your marvelous posting! I genuinely enjoyed reading it,
you may be a great author.I will always bookmark your blog
and may come back later in life. I want to encourage one to
continue your great posts, have a nice weekend!

Permalink to Comment

37. kosmetyki sklep internetowy on February 13, 2015 2:42 PM writes...

Hello to every one, because I am in fact eager of reading this web site's post to be updated on a regular basis.
It contains nice information.

Permalink to Comment

38. kosmetyki sklep internetowy on February 13, 2015 2:42 PM writes...

Hello to every one, because I am in fact eager of reading this web site's post to be updated on a regular basis.
It contains nice information.

Permalink to Comment

39. sklep snowboard on February 14, 2015 1:56 AM writes...

Very soon this website will be famous amid all blogging and site-building viewers, due to it's
good articles or reviews

Permalink to Comment

40. sklep snowboard on February 14, 2015 1:57 AM writes...

Very soon this website will be famous amid all blogging and site-building viewers, due to it's
good articles or reviews

Permalink to Comment

41. Soundscape on February 14, 2015 3:12 AM writes...

I was pretty pleased to find this page. I wanted to thank you for your time just for this wonderful read!!
I definitely appreciated every part of it and i also have you bookmarked to
see new things on your website.

Permalink to Comment

42. cydia repo sources on May 15, 2015 10:56 PM writes...

The JailBreakMe exploit was a vulnterability in Apple's PDF renderer.
It is easy to do when you know the steps to follow and such.
A few of these apps will go to the jailbroken equivalent app store.

Permalink to Comment

43. bunion braces on August 27, 2015 7:20 PM writes...

When you look at other professions which display a high pattern of tailor's bunion formation the picture becomes clearer.

Permalink to Comment


Remember Me?


Email this entry to:

Your email address:

Message (optional):

The Last Post
The GSK Layoffs Continue, By Proxy
The Move is Nigh
Another Alzheimer's IPO
Cutbacks at C&E News
Sanofi Pays to Get Back Into Oncology
An Irresponsible Statement About Curing Cancer
Oliver Sacks on Turning Back to Chemistry