There's been a report on the toxicity of various pesticides in the literature suggesting that they're far more toxic to human cells than had been thought. My eyebrows went up a bit when I heard this, because these sorts of assays had been done many times before. Then I realized that this was another paper from the Séralini group, and unfortunately, that alone is enough to account for the variance.
Update: commentors to this post have noted that the cell culture conditions used in the paper are rather unusual. Specifically, they're serum-free during the testing period, which puts the cells under stress to begin with. There's also the general problem, which others have brought up, about what it means to dispense these things directly onto cell cultures in diluted DMSO, since that's rather far from how they're going to be presented in the real world. Cell assays get run like that in the drug industry, to be sure, but you've got to be very careful drawing toxicological or other whole-animal conclusions from them. And we already have whole-animal studies on these formulations, don't we? I mean, juiced broccoli straight from the organic farmer's market might well have similar effects under these conditions.
Here's a story from Science with more background. Seralini is the guy who made headlines a couple of years ago with another report that genetically modified corn caused tumors in rodents, but that one was so poorly run and poorly controlled that its conclusions (which have not been seen in any other study) cannot be taken seriously. That's Séralini's problem right there: from all appearances, he's a passionate advocate for his positions, and he appears to be ready to go with whatever results line up with his beliefs. This is human nature, for sure, but science is about trying to work past those parts of human nature. The key is to keep the curious, inquisitive side, and correct for the confirmation bias I-know-I'm-right side. At this point, even if Séralini were to discover something real (and really worth taking seriously), it would have a hard time gaining acceptance, because his previous papers have been so unreliably over-the-top.
I'm not the only person who thinks that. An editor of the journal this latest Seralini paper appeared in has actually resigned because it got published:
When Ralf Reski read the latest paper from controversial French biologist Gilles-Eric Séralini, he quickly decided he wanted nothing to do with it. Séralini’s report in BioMed Research International describes how pesticides kill cultured human cells, with the hair-raising conclusion that pesticides may be vastly more toxic than assumed by regulatory authorities. Some scientists are criticizing the findings as neither surprising nor significant—but they have touched off a firestorm, with environmental groups calling for changes in how pesticides are regulated. That was too much for Reski. Within hours of reading the paper last week, the plant scientist at the University of Freiburg in Germany resigned as an editor of the journal and asked for his name to be removed from its website. "I do not want to be connected to a journal that provides [Séralini] a forum for such kind of agitation," he wrote in his resignation e-mail to the publisher, Hindawi Publishing Corporation.
Should pesticide toxicity be a subject of investigation? Absolutely. Should people be alert to assays that have not been run that should be investigated? Definitely. Are there things that we don't know about pesticide exposure that we should? I would certainly think so. But Séralini's history makes him (scientifically) one of the least effective people to be working on these questions. As a headline-grabber, though, he's pretty efficient. Which I suspect is the real point. If you're sure you're right, any weapon you can pick up is a good one.