Confirmatory Results: Regulation of TCP genes by miR319

Update on a Developmental Cell paper with duplicated image

We have found issues with our article ‘Sequence and Expression Differences Underlie Functional Specialization of Arabidopsis MicroRNAs miR159 and miR319’, published in Developmental Cell in July 2007.

1. Even though it should be clear from the text, the blot in Figure S10H is the same as the right hand blot in Figure 5H; this information could have been added to the Figure S10H figure legend.

2. The sentence ‘Even the weaker expression of miR319b (which is identical to miR319a) from a 35S:miR319b construct had a pronounced effect on TCP4:GFP RNA levels (Figure 4B)’ in the main text should read ‘Even the weaker expression of miR319b (which is identical to miR319a) from a 35S:miR319b construct had a pronounced effect on TCP4:GFP RNA levels (Figure 4A)’.

3. We noticed spliced lanes in an RNA blot in Figure 4A and Figure S7 that were not mentioned in the figure legend.

We have gone back to the original data, and properly re-assembled Figure S7 with a clear indication of the spliced lanes. A correctly assembled Figure S7 is shown here. (Because spliced gels were not in violation of journal policies at the time of publication, the journal does not judge that this issue requires a formal Correction.)

4. The most serious issue affects Figure 4B. During its assembly, the same image was mistakenly used for the upper and lower left panels of Figure 4B, with one of them rotated by 180°. The two images document experiments performed in 2004, with both panels intended to show similar TCP4:GFP fluorescence signal with or without miR159a overexpression. (There are substantial levels of endogenous miR159a in this tissue, and if miR159a affected TCP4:GFP, we would have expected to see an effect already without miR159a overexpression.)

Unfortunately, the original data of this experiment are no longer available. We have therefore repeated the complete experiment depicted in Figure 4, which includes microscopy images of N benthamiana (Figure 4B) as well as RNA and small RNA blots of the same samples (Figure 4A). TCP4:GFP fluorescence signal was indeed similar with or without miR159a overexpression. A corrected Figure 4 is shown here.

We have discussed this issue with the journal and offered a retraction, since the issue could not be remedied with data available at the time of publication, and since a simple formal Correction is therefore not possible. The journal leadership was of the opinion that retraction of the entire article would go too far, and that they prefer to publish an Editorial Note of Concern regarding these issues. We published a short manuscript describing the repeated experiment with a corrected Figure 4 on bioRxiv, in order to provide a permanent link to these data, and for colleagues to be able to more fully judge the data. We will also mark the paper in our own CVs as having a linked Note of Editorial Concern.