Stem cell researchers have papers retracted for image manipulation

A group of dentistry researchers in Japan, whose work on stem cells has been the subject of an institutional investigation, have now lost two papers in PLOS ONE for image problems. 

The authors, from Aichi Gakuin University in Nagoya, were led by Makio Mogi, a medicinal biochemist at the school. Mogi asked for at least one of the retractions. 

The first article, published in 2013, was titled “Matrix metalloproteinase-3 in odontoblastic cells derived from Ips cells: unique proliferation response as odontoblastic cells derived from ES cells.” It has been cited 18 times, according to Clarivate Analytics’ Web of Science. According to the notice: 

Following publication of this article [1], the authors requested its retraction due to concerns raised about several figures. Aichi Gakuin University investigated this matter and confirmed the following issues.

1. In several cases, the grayscale of a PCR gel image was inverted, and the consequent image was used to represent Western blot data.

— In Figure 1B, the β-tubulin Western blot panels for Odontoblasts derived from iPS cells and from B6G-2 cells were generated from the GAPDH PCR image shown in Figure 1A for the B6G-2 experiment.

— In Figure 2, the gray scale of each PCR image in panel A was reversed; the resultant images were adjusted for contrast and black level and were used to represent western blot images in panel B.

— The gray scale of the GAPDH panel for KN-3 cells in Figure 6A was reversed, and the resultant image was used to generate the β-tubulin panel for the E14Tg2a cell experiment in Figure 6B.

— In Figure 6, the gray scale of lanes 1–4 of the MMP-3/KN-3 PCR image in Figure 6A was reversed. The resultant image was adjusted for contrast and black level, vertically enlarged, and used as lanes 1–4 of the MMP-3/E14Tg2a blot in Figure 6B.

2. In Figure 6A, vertical lines suggestive of image splicing were observed in the 1st, 4th, 5th, and 7th panels (MMP-3/iPS, GAPDH/B6G-2, MMP-3/E14Tg2a, and MMP-3/KN-3, respectively).

3. In Figure 6A, the same image was used in the 2nd and 6th panels (GAPDH/iPS, GAPDH/E14Tg2a), with the 6th panel rotated 180 degrees relative to the 2nd panel.

4. In Figure 6A, the first four lanes in the 1st and 7th panels (MMP-3/iPS, MMP-3/KN-3) appear to present duplicate data, with the image adjusted for contrast and brightness and vertically enlarged in the 1st panel relative to the 7th.

The original data underlying these results are not available.

In light of these concerns, and in line with the University’s recommendation, the PLOS ONE Editors retract this article.

NO, MM, HY, RK, KN, AK, HN agreed with the retraction. TH did not respond.

The second notice, for a 2015 paper titled “Interleukin-1β-induced autophagy-related gene 5 regulates proliferation of embryonic stem cell-derived odontoblastic cells,” makes the same case: 

The corresponding author requested retraction of this article [1] due to concerns about the integrity of the data and the validity of the conclusions. The underlying data supporting the results in this article are not available.

The following specific concerns were noted about the reported results:

— The brightness and contrast on the western blot figures in the article do not allow a confirmation of the integrity of the results as presented.

— The β-tubulin blots appear similar in Fig 7 panels A & D, panels B & E, and panels C & F.

— BrdU and apoptosis microscopy images in the following panels of Figs 4, 5, and 6 appear to report the same data:

— BrdU and apoptosis images in Fig 4B, Fig 5B, and 6B for experiments i, ii, iii

— BrdU images in panel vi of Fig 4B and 6B, which represent different experiments

— Apoptosis images in panel vi of Fig 4B, Fig 5B, and Fig 6B, which represent different experiments

The authors were unable to provide any clarifications about these issues.

Aichi Gakuin University is investigating this work. The investigating committee has tentatively confirmed that there are concerns about the quality of the western blot images reported in this article and that the underlying data are no longer available.

In light of the above concerns, the PLOS ONE Editors retract this article.

NH, HY, RK, AK, TM, KN, and MM agree with the retraction. NO and TH did not respond or could not be reached.

The paper has been cited five times.

TH is Taiki Hiyama. Our emails to Hiyama and Ozeki bounced back as undeliverable. We also attempted to contact Mogi but have yet to receive a reply.

Update, 2130 UTC, 11/1/19: This appears to be a summary report of the university’s investigation.

Like Retraction Watch? You can make a tax-deductible contribution to support our work, follow us on Twitter, like us on Facebook, add us to your RSS reader, or subscribe to our daily digest. If you find a retraction that’s not in our database, you can let us know here. For comments or feedback, email us at [email protected].