| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
We have identified several inaccuracies and concerns in our paper "Thyroid Hormone Regulates Hepatic Triglyceride Mobilization and Apolipoprotein B Messenger Ribonucleic Acid Editing in a Murine Model of Congenital Hypothyroidism" (Debnath Mukhopadhyay, Michelina Plateroti, Shrikant Anant, Fatiha Nassir, Jacques Samarut, and Nicholas O. Davidson, Endocrinology 144:711–719, 2003). These concerns are summarized below and primarily reflect the fact that the "n" represented in some of the figure legends is misstated.
Specific concerns
1. The legend to Fig. 1D states that data are derived from six animals per group. In fact, the data are derived from 5, 4, 5, 4 animals per group.
2. The legend to Fig. 2A states that the bar graph summary of the data is from a mean of six animals per group. In fact, the data are derived from the experiment shown in the top panel but from an "n" of 3, 2, 3, 2 per group. The statistical analysis and data are correct, but the "n" is misstated.
3. The Western blot in Fig. 2B, showing the appearance of apoB in serum, is a representative of two Western blots and not three independent experiments, with a total of two to three animals per group analyzed.
4. The data in Fig. 3A are stated as being derived from an n of 6 per group. In fact, the data were derived from an n of 3, 2, 3, 2 animals per group. The data in Fig. 3B were derived from a single experiment.
5. The data in Fig. 4, A–C, were representative of three replicate experiments but the data in panel D appear to have been derived from a single experiment. In addition, not all the primary data (Western blots for HuR) could be found in the original notebooks.
6. The data in Fig. 5 were performed on an "n" of 3, 2, 3, 2 samples per group and not an "n" of 4 per group as stated in the figure legend.
7. The data in Fig. 6 appear to have been performed on a single experiment and not three independent experiments as stated in the legend.
Response to concerns
In view of the serious and systemic nature of these concerns, five of the original authors and an additional scientist (V.B.) currently in the laboratory, but who was not involved in the original studies, undertook to repeat all the studies reported in this manuscript using entirely new sets of animals. These studies were initiated as soon as we discovered the discrepancies and, in view of the substantial requirements for neonatal Pax8–/– mice, have been completed over the last 15 months. These studies revealed the following findings, summarized below:
Figure 1. All the findings noted in the original manuscript have been replicated, including the data in panel D, which have been expanded to an "n" of 9 animals per group.
Figure 2A. All the findings in the original manuscript have been replicated, with an "n" of 4 animals per group.
Figure 2B. The changes in serum apoB isoform distribution appear variable, with an inconsistent change in the appearance of the apoB48 band. These experiments were undertaken on samples from three different groups of two to three mice per group.
Figure 3A. The findings in the original manuscript have been replicated with a decrease in ACF mRNA abundance in the Pax8–/– mice and a TH-dependent increase. These assays were undertaken using three groups of eight animals per group.
Figure 3B. The changes in ACF protein abundance are variable and we conclude that there is no consistent increase in total liver ACF protein. These experiments were undertaken on samples from three different groups of two to four animals per group.
Figure 4, A–C. The findings in the original manuscript have been replicated with three separate groups of animals, containing two animals per group.
Figure 4D. The findings have been replicated in two separate analyses of one animal per treatment group.
Figure 5. The findings in the original manuscript have been replicated with three independent experiments containing four animals per group.
Figure 6. The findings in the original manuscript have been replicated with two independent experiments.
All the data have been summarized and can be viewed for comparison alongside the data in the original publication (see The Endocrine Society’s Journals Online web site at http://endo.endojournals.org).
Conclusions from the replicated studies
The evidence from the replicated studies suggests the following conclusions.
Neonatal Pax8–/– mice demonstrate a striking increase in hepatic triglyceride content associated with a decrease in hepatic apoB mRNA editing. Thyroid hormone administration results in hepatic triglyceride mobilization and an increase in hepatic apoB mRNA editing. The induction of hepatic apoB mRNA editing by thyroid hormone is associated with an increase in the nuclear abundance of ACF protein, as confirmed by both immunohistochemical findings and Western blot analysis of fractionated nuclei. In vitro RNA editing assays demonstrate that supplementation with ACF restores the C to U RNA editing activity of liver S100 extracts from Pax8–/– mice.
Summary
In conclusion, our efforts to determine the validity of the observations and data sets reported in this paper suggest that the core observations originally reported are valid. We do not believe the inaccuracies in the reporting of the "n"s represent intentional misconduct. Nevertheless, we believe these important discrepancies need to be brought to the attention of the readership, along with the assurance that the laboratory stands by the original findings.
We regret any inconvenience that this may have caused.
Yours Sincerely,
Valerie Blanc, Ph.D.
Michela Plateroti, Ph.D.
Shrikant Anant, Ph.D.
Fatiha Nassir, Ph.D.
Jacques Samarut, Ph.D.
Nicholas O. Davidson, M.D.
| ||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
