Following publication of this article [1], several concerns were raised about the Western blots in Figs 1, 5, and 6. The authors confirmed that in preparing these figures they had spliced image fragments to remove empty lanes, rearrange the sample order, and in some cases to combine lanes from short and long exposures of a given blot. The Director of the Institute for Developmental Biology of Marseille discussed this matter with the corresponding author and examined the original data underlying the results in question. The Director concluded that the images were modified for the purpose of presentation, and that the scientific results presented in the article and underlying data are sound.

The authors provided the following explanations as to how the original figures were prepared; updated figures are included with this notice, and the original raw Western blot data underlying most of the affected figure panels are included in S1 File.

• Fig 1B: The original α-HA panel was comprised of input data (lanes 1–3) from a short exposure and IP Flag data (lanes 4–6) from a longer exposure. In addition, an empty lane was removed from the α-Flag Western blot panel. The new Fig 1B uses only the long exposure of the α-HA blot, and the empty lane (lane 4) has not been removed in either panel.
• Fig 1D: In the original blot, the experimental samples were separated by empty lanes. In generating the published figure, the empty lanes were removed and experimental samples were spliced next to one another. The new figure presents the original image with the empty spacer lanes intact.
• Fig 1F: To generate the published figure, the authors removed empty lanes from the IP Flag lanes of the α-HA blot. For the α-Flag blot, the authors spliced together lanes to remove empty lanes and to combine data from different exposures. The images were cropped so as not to show unspecific bands in IP lanes. In the new figure, the authors present data from a replicate experiment. Raw images used to generate both the original figure and the revised figure are provided with this notice.
• Fig 1G: In the published figure, a blank lane was removed between the input and IP Flag lanes of both panels. Lanes 4–6 in the α-HA blot appear to have been inverted vertically in preparing the figure. The new figure presents the intact blot images, with the empty lanes in each blot (lane 4) intact.
• Fig 5A: In the published figure, lanes 3 and 4 were flipped horizontally. The new panel presents the same data in a direct representation of the original blot image. Note that the loading order in the new figure differs from that in the original version of the figure.
• Fig 5B: Concerns were raised regarding an irregularity in lane 6 of the α-HA blot, and potential splicing between lanes 1 and 2 of the α-Flag blot. An empty lane was also excised between the input and IP Flag lanes of the α-Flag blot in the published figure. The original α-HA blot is no longer available, and so the authors provide data from a replicate experiment in the new figure.
• Fig 5C: In the published figure, empty lanes were removed from each panel between lanes 3 and 4. The new figure represents the original blot with the empty lane intact in each panel.
• Fig 5D: In the published figure, an empty lane was removed between lanes 3 and 4. The new figure represents the original blot with the empty lane intact.
• Fig 6A: The published α-HA panel is comprised of lanes from a short exposure (input lanes) and long exposure (IP Flag lanes) of the same experiment. The IP Flag lanes were inverted horizontally, and empty lanes between experimental samples were removed. The new figure presents the intact image from the long exposure, with empty lanes included. Note that the loading order for the new panel differs from the loading order of the α-Flag blot shown in Fig 6A of the original publication.
• Figure S4A: The raw data image provided for the input lanes of the α-Flag blot panel (S2 File) does not match the image in the published figure. Raw data supporting the α-HA blot in this figure panel are in S3 File.

Panels F and G in the original version of Fig 5 [1] were mislabeled and should have been labeled panels E and F. The corresponding figure legend in [1] likewise contained an error in referencing these panels. These errors have been corrected in the updated version of Fig 5 provided here; the corrected labels (E, F) align with the in-text citation to these figure panels in the Results section of [1].

Raw data supporting other results reported in the article are provided in S4–S20 Files.

The underlying raw data are no longer available for the α-Flag blot in Fig 5D, or for Figs 1H/I, 2D/E/F, 4, 5G, 6B, 6C, 7C, S1, S2Q/R, S3, S4B/E, or S5. For Figure S5 in [1], the authors provided an updated figure that reports replication data. The new version of this figure is included with this notice as S21 File, and the raw data from the replication experiments are in S22–S25 Files.

Given the unavailability of these data and the extent of concerns with the images in the published article [1], the PLOS ONE Editors issue this Expression of Concern.