blog

When you come to write up your lab results and publish them (whether in a journal or your dissertation), one of the things you have to do is write up how you did everything–the methods section. In fact, it’s usually a good idea to write up your methods well before this point, but that’s another discussion.

As a matter of course, in the methods section you should note the suppliers of all the various reagents you use. One reason for this is so that, months or years down the line, when you come to do more experiments using one of those particular reagents–usually after the grad student next door has snuck in and used the last of it without telling you–know where you bought it from. And to let other scientists know so they can repeat your experiments.

And of course, science being what it is, it also allows others to look at your methods and say things like “You’re using catalase from Sigma? Are you completely mad?”

The thing is, the cheapest reagents aren’t necessarily the most pure. Anybody who has worked in a lab can tell you this. Most of the time it doesn’t matter, at least, not if the phenomenon you’re interested in is reasonably robust. But there have been, for example, reports of plastics leaching into culture media, and no doubt we all have war stories about which suppliers to avoid for particualr experiments (and also check out the reason behind the XMRV ‘partial’ retraction). Some of these are anecdotal or even superstitious; others are worringly reproducible.

Last week, Pernille Hansen and Ole Skøtt at the University of Southern Denmark evaluated a paper in the American Journal of Physiology–Renal Physiology, Soluble epoxide hydrolase contamination of specific catalase preparations inhibits epoxyeicosatrienoic acid vasodilation of rat renal arterioles(^{10.1152/jprenal.00201.2011}). Briefly, a study designed to examine the effects of hydrogen peroxide on arteries in the kidney discovered that some preparations of catalase–the enzyme that deactivates peroxide–are contaminated with soluble epoxide hydrolase, an enzyme that hydrolyses the very signalling molecule that peroxide was thought to affect.

The authors claim that nearly two-thirds of articles in this journal may have used the contaminated catalases from Sigma (presumably that’s two-thirds of articles that report a use of catalase), which “may have led to flawed results.”

Oops.