Misconduct at Oxford prompts retraction of insulin paper

The C57BL/6J mouse displays glucose intolerance and reduced insulin secretion. QTL mapping identified Nicotinamide Nucleotide Transhydrogenase (Nnt), a nuclear-encoded mitochondrial protein thought to be involved in free radical detoxification, as a candidate gene. To investigate its functional role, we used siRNA to knock down Nnt in insulin-secreting MIN6 cells. This produced a dramatic reduction in insulin secretion and the rise in [Ca²⁺]_i evoked by glucose, but not tolbutamide. We identified two ENU-induced point mutations in Nnt (N68K, G745D). Nnt mutant mice were glucose intolerant and secreted less insulin during a glucose tolerance test. Isolated islets showed impaired insulin secretion in response to glucose, but not to tolbutamide, and glucose failed to enhance ATP levels. Glucose utilization and production of reactive oxygen species were increased in Nnt β cells. We hypothesize that Nnt mutations/deletion uncouple β cell mitochondrial metabolism leading to less ATP production, enhanced K_ATP channel activity, and consequently impaired insulin secretion.

The paper has been cited 95 times, according to Thomson Scientific’s Web of Knowledge. But as the retraction notice states:

This paper is being retracted because of concerns with Figures 2B and S3. In Figure S3, the SDS-PAGE gel image had molecular weight markers added from another autoradiograph and additional nonspecific bands obscured or removed. In the top panel of Figure 2, an actin Western blot was used instead of a GCK Western blot. This matter was investigated by the Medical Research Council (UK) under their Scientific Misconduct Policy and Procedures, and scientific misconduct pertaining to the issues described here against the first author Helen Freeman was found. Repetition of these experiments by other group members gave identical results to those shown in the paper. Nevertheless, although we feel that the scientific conclusions of the paper (that loss of NNT impairs insulin secretion and impairs glucose tolerance in C57BL/6J mice) still stand, given the conclusion of scientific misconduct against the first author, the authors think the most responsible course is to retract the paper. The authors sincerely apologize for the inconvenience caused by this retraction.

We note that Freeman’s name does not appear on the earlier retraction from Ashcroft’s team. Could it be a second lightning strike?

Meanwhile, there’s a bit more to this story. The group narrowly escaped (we’re not quite sure how) the retraction of a similar 2006 paper, this one in the journal Diabetes. Here’s the correction notice for “Deletion of Nicotinamide Nucleotide Transhydrogenase: A New Quantitive Trait Locus Accounting for Glucose Intolerance in C57BL/6J Mice:”

In the article listed above, it was discovered that the Western blot in Fig. 1C was inappropriately manipulated. It appears that a protein band labeled C3H/HeH, which was a control sample, was added to the figure in its preparation. A Medical Research Council investigation found that there had been scientific misconduct by the first author, Dr. Helen Freeman, and that none of the other authors were involved. The other authors sincerely apologize for this having occurred, and wish to correct the published record. The authors have replicated the results of the original figure using their remaining bacterial artificial chromosome (BAC) mouse line (see Fig. 1C replacement below) and have also previously replicated the rescue of the glucose tolerance phenotype in this line, thus showing that the original conclusions of the article are not compromised.

New Fig. 1C legend: BAC transgenic expression of wild-type NNT protein in male C57BL/6J mice.Top panel: Expression of NNT protein in C3H/HeH (lane 1), C57BL/6J (lane 2), C57BL/6NTac (lane 3), and BAC transgenic C57BL/6J (lanes 4 and 5) mouse liver. Bottom panel: Reprobing of the same blot with an anti–α-tubulin antibody as a loading control. Protein was separated on a 4–12% gradient polyacrylamide gel (Invitrogen) and electrophoresed in MOPS SDS running buffer (NuPage, Invitrogen). The NNT antibody was our own custom affinity purified rabbit polyclonal. The anti–α-tubulin monoclonal antibody was 12G10. The 12G10 anti–α-tubulin antibody developed by Joseph Frankel and E. Marlo Nelsen was obtained from the Developmental Studies Hybridoma Bank, was developed under the auspices of the National Institute of Child Health and Human Development, and maintained by The University of Iowa, Department of Biology, Iowa City, IA. The Western blotting was carried out by Dr. Michelle Goldsworthy.

That paper has been cited 66 times, according to Thomson Scientific’s Web of Knowledge.

We also note that Freeman and Ashcroft appear on a patent application for “Methods and compositions for the diagnosis of treatment of type 2 diabetes.”