The authors of a 2006 Journal of Immunology study have retracted it after it dawned on them that they used the wrong mice.
The study, “Endogenous IL-1R1 Signaling Is Critical for Cognate CD41 T Cell Help for Induction of In Vivo Type 1 and Type 2 Antipolysaccharide and Antiprotein Ig Isotype Responses to Intact Streptococcus pneumoniae, but Not to a Soluble Pneumococcal Conjugate Vaccine,” has been cited 11 times, according to Thomson Scientific’s Web of Knowledge.
According to the retraction notice:
In Figs. 1–5, we stated that we had used IL-1R1 knockout mice (B6129 background, The Jackson Laboratory) in our experiments. Recently, using IL-1R1 knockout mice (C57BL/6 background, The Jackson Laboratory), we were unable to reproduce a critical experiment from that published paper. Upon investigation we discovered that the mice that had been ordered for the experiments described in our published article were, in fact, mistakenly mice that were doubly deficient in IL-1R1 and TNF a p55 (type 1) receptor. As a result, our data cannot be interpreted in the context of a sole deficiency in IL-1R1 signaling. Although this was a purely unintentional mistake, we need to retract this paper.
Cliff Snapper, the paper’s senior author, gave us some more details by email:
This was not Jackson’s fault. The individual conducting most of the studies had unfortunately misunderstood the designation on the descriptive sheet for these mice, and thought she was getting IL-1R1-/- mice (B6129 background) only . Each time she ordered more mice, she handed in a xerox copy of the order, so kept making the same mistake. More recently we decided to continue exploring the role of IL-1 in this model system, and this time she ordered bona-fide IL-1R1-/- mice only, on the C57BL/6 background. We now could not reproduce the main finding. She went back into her records to try and understand this major discrepancy, and only then realized her mistake. I did not mean to ascribe any fault on Jackson’s part.
It’s actually easy to see how someone would make that mistake: The two strains are listed close to one another on the same page, with all sorts of symbols, superscripts, numbers, and letters.
In another email, Snapper wrote:
Needless to say, the retraction was a painful experience, but absolutely necessary.
We know Snapper feels badly about this, and thought carefully through it, because the only reason he mentioned that this wasn’t Jackson Labs’ fault was because we had originally read the retraction notice too quickly and made assumptions. Good thing we checked it out.
Good for him. He gets high marks from Retraction Watch for transparency.
Hat tip: Tom Oates
This might not be a popular thing to say in scientific circles, but I hope someone’s been making contributions to humane societies in memory of all the animals who died pointlessly in all these retracted studies we’ve been reading about.
I hate to point this out, but these kinds of errors are unfortunately not that uncommon. When I was a grad student years ago, one of the postdocs in our lab was in the process of preparing a paper for submission involving the use of hundreds of “knockout mice” over several years. I was asked to read and provide comments to a late draft. Upon reading the manuscript and looking up the Jackson mouse information for the strain used and cited, I realized she had been using a floxed strain for the entire set of experiments (without ever mating said strain with any Cre strain). Somehow several PIs and postdocs missed this crucial detail, and somehow the postdoc who performed the experiments was completely unfamiliar with the Cre-Lox system. Needless to say there were lots of unhappy people that day. Luckily, it was caught prior to submission.
Where I work, we go through a LOT of mice. And nobody likes it.
Not the PI who has the mega-expense ($25 to $15,000 a mouse, plus boarding fees) nor the scientist who have to handle the animals. I’ve walked in people who were sobbing and wondered who died. It was the mice.
But to improve human health and save lives, there’s often no other way apart from animal experiments. It sucks. Big time. But it’s true.
Now, cosmetic testing on the other hand . . .
Several hundred times more animals are killed (after living horrible lives in most cases) for food than are killed for science every year, in the UK, and it’s the same in the US I’m sure.
Lab animals killed per year in the UK: about 3.5 million.
Broiler chickens killed, for the UK market, every day: 2.3 million (sic).
(that doesn’t include fruit flies, but, really, I don’t think anyone weeps over them.
As far as I’m concerned, if you’re angry about animal testing, you should be several hundred times more angry about meat, and that’s leaving aside the fact that meat is an unhealthy luxury while animal testing saves lives.
I’m angry and grieved about both. And I wish people would stick to the point when discussing animal testing: even if it is expedient, is it ethically correct?
This isn’t an issue of whether it’s expedient or not. There is, simply put, no other way to conduct research on a variety of areas necessary to improve human health.
In general, the closer the model system is to human, the more applicable the results will be to human health. In particular, the adaptive immune system is unique to mammals, and requires a mammal model, such as mice, to be studied.
Mammal models are also essential in drug testing, to make sure that new drugs are reasonably safe before proceeding to human testing.
I appreciate the integrity of the Snapper Lab. Mistakes do happen and it’s better to correct the record when they are discovered.
Very dignified attitude of Snapper. Mistakes indeed do happen, and it is almost honourable that upon discovering the inconsistency of the results, they started looking into the data to see what might have gone wrong, rather than shrugging their shoulders. Lets hope other people will follow this example.
Oldish thread, I know, but I feel the need to point out a mistake above:
*Vertebrates* have adaptive immune systems, whether they are mammalian or not. In fact, B-cells, an important part of the adaptive immune system, are named as such after the discovery of where they develop in chickens, clearly not a mammal, the Bursa of Fabricius.