As a veterinarian for research animals, I have an odd job, and part of the oddness is the amount of time I spend thinking about how to kill animals. Christie Wilcox posted on how carbon dioxide near-asphyxiation causes panic in mouse and human, so I had to pay close attention. I’m interested in all things carbon-dioxide, because that it is both one of the commonest ways of killing laboratory rats and mice, and one of the most controversial.

Killing animals is such an integral part of veterinary medicine (and animal shelter management, and laboratory animal use, and food production) that the American Veterinary Medical Association (AVMA)  has been publishing and updating guidelines on how to do that for almost 40 years. In 2001, they wrote that carbon dioxide asphyxiation is “appropriate” for a wide range of species. Those guidelines are now under revision [disclosure: I’m on one of the working groups for the revisions] and carbon dioxide will be the big controversy.

The Humane Society of the United States has pushed heavily to get carbon dioxide asphyxiation banned from laboratories. The science – and yes, there is a bit of research on best ways to end animals’ lives –is mixed, and much rides on how people measure pain and distress and fear in the animals. A group of UK scientists and animal welfare specialists has published their expert report. Their conclusion is that carbon dioxide exposure is not technically painful at less than about 50%, but it is “aversive,” which is to say, the animals really dislike it.

The study that Wilcox posted on doesn’t tell us whether mice and people panic when exposed to 10% carbon dioxide. That’s already established, as well as the finding that people with panic disorders will panic when CO2 is as low as 5% (usually, what you breathe is less than 0.04% carbon dioxide; a really stuffy room may get up to twice that). People are not generally exposed to levels as high as 5%, except in some seriously dangerous situations, where fear is certainly an appropriate response. 

The new study gives some info how the CO2 causes the fear – the bottom line being that the ASIC1a protein in the amygdala (one of the brain’s master centers of emotion) that is required for fear responses is also a pH-sensor. When you breathe in high levels of CO2, you acidify your blood and your amygdala, and the ASIC1a protein senses that and triggers fear.

Some of what they found in mice they has been studied in people: expose volunteers to carbon dioxide and MRI imaging will show activation of the amygdala. Other parts of what they found might now be further investigated with people: alkalyze the body (they used sodium bicarbonate in the mice) and perhaps you could neutralize the fear that carbon dioxide can cause. This might be especially useful for people whose panic sets in at more normal CO2 exposures; it’s certainly worth exploring how their ASIC1a proteins might differ from less anxious folks.

The mice studies were clever, looking at fear when exposed to carbon dioxide. We don’t usually use the word “panic” in talking about mice, but the authors found that mice would freeze in place when they sniffed the gas, typically scored as a “fear behavior.” Or they would hug the walls of an Open Field test-chamber. If you genetically modify the mice by knocking out the ASIC1a protein in the amygdala, you knock out the fear. If you “alkalyze” the mice, including your normal mice, with sodium bicarbonate a few minutes before the CO2 exposure, you also knock out the fear. Conclusion: the mouse ASIC1a protein senses rising CO2 levels by sensing the acid-base shift in the bloodstream, and triggers the amygdala to say “Fear.”

Animal care and use committees are tasked with asking why animals are being used in experiments instead of alternatives like cell cultures or computer simulations. But human volunteers are great animal-substitutes for some studies, and so I have to ask, how much of this could be done in human volunteers, sparing the mice? Quite a bit, and in fact, quite a bit has. 10% CO2 is unpleasant, but safe to breathe, and you can couple breathing it with self-reports of fear, and MRI scans of the amygdala. I imagine there will soon be studies (maybe in progress) asking the human subjects to take some bicarbonate to neutralize the carbon dioxide. Harder to investigate in humans would be the role of that particular ASIC1a gene and protein.

Fear in the face of carbon dioxide exposure is probably widespread. Mice, humans. Pigs appear to hate it too: if you expose them to CO2 in a room, you can’t entice them back in, even if you’ve replaced the CO2 with an apple; it’s that aversive to them.**  There’s not a whole lot a self-respecting pig won’t do for an apple.

The animal welfare issue: every year, tens of millions of laboratory rodents are killed by carbon dioxide exposure. X% is fatal, within about 15 seconds. >50% is scored as painful, at least by people. 35% causes unconsciousness, again, after several seconds exposure. If you exposed the animals to 100% carbon dioxide AND the effect was instantaneous, they’d never have the blood levels that cause fear or pain. But it’s not instantaneous. Time for yourself what 15 seconds feels like if you’re doing something painful. That’s about the time the Newcastle group thinks animals could be conscious during CO2 asphyxiation. The science to me is fairly clear – carbon dioxide causes several seconds of distress (and if done wrong, pain too) that other methods of killing can avoid. You can sedate an animal, then give a barbiturate anesthesia overdose, and it’s about the most peaceful death I’ve ever witnessed.

The ethical question goes beyond standard vet expertise, and that is: when, if ever, those several seconds of distress are justifiable. That’s one of the questions the AVMA’s expert panel will be debating, so I’ll have more to post on this soon. 

** Raj and Gregory, 1995 A.M. Raj and N.G. Gregory, Welfare implications of gas stunning of pigs. Determination of aversion to the initial inhalation of carbon dioxide or argon, Animal Welfare 4 (1995), pp. 273–280.