Abstract
The authors contacted the journal after becoming aware of an error in the panel displayed in Fig 3F. Based on the exchanges with the authors, a reassessment by the original reviewers and internal analysis at the journal, the editors agreed to correct the following figures and associated claims and to publish the related source data with this notice:
Figure 3F is withdrawn.
Figure EV3A has been added.
Figure 6H and J is withdrawn and replaced.
Author statement
We became aware that the wrong image was included in Figure panel 3F due to a misnaming of the file for the data at the point of acquisition. The original data have been lost. Upon redoing the experiment displayed in original Fig 3F (phosphorylation assay using CCND1 and CDK4 with isotope‐labeled ATP), we noted that both wt and S309A mutants of ATOH1 are phosphorylated, indicating that S309 is not the sole phosphorylation site for CCND1 and CDK1. This experiment is now included as a new panel in Fig EV3A. Fig EV3A and B in the original manuscript have been renamed as Fig EV3B and C.
Updated figure legend to Fig EV3A (new text in bold)
In vitro phosphorylation assay with either wild‐type ATOH1 (WT) or S309A mutant ATOH1 (S309A‐ATOH1). WT‐ or S309A‐ATOH1 was incubated with CCND1 and CDK4 (4D) in the presence of [γ‐32P]‐ATP. Clear bands were detected for WT‐ATOH1 and S309A‐ATOH1 when incubated with CCDN1/CDK4 (lanes 2,4).
Figure EV3. Original.
Figure EV3. Corrected. Source data are available online for this figure.
We performed in vitro kinase assays with anti‐S309‐phosphorylated‐ATOH1 antibody, which indicated that S309 is phosphorylated by CCND1 and CDK4, and therefore, this conclusion in the original paper is valid. Figure panel 3F is herewith retracted and republished, including the new data generated after publication of the original paper.
Updated figure legend to Fig 3F (new text in bold)
In vitro phosphorylation assay with either wild‐type ATOH1 (WT) or S309A mutant ATOH1 (S309A‐ATOH1). WT‐ or S309A‐ATOH1 was incubated with CCND1 and CDK4 (4D), followed by immunoblotting with anti‐Atoh1 (total Atoh1), p‐S309‐Ab (p‐Atoh1 S309). Total ATOH1 was detected in all lanes. In the presence of CCND1 and CDK4, strong signals were detected for WT‐ATOH1 but not S309A‐ATOH1 (lanes 2,4). Quantification of relative p‐Atoh1 S309 to total Atoh1 (lower panel). Tukey–Kramer test was performed. Bars and error bars represent mean ± SEM from n = 4 for each condition. ***P < 0.001, N.S., not significant.
Figure 3. Corrected.break;Source data are available online for this figure.
The manuscript text is updated to reflect the new experimental data as follows (new text in bold)
Although S309 and S328 are both highly conserved proline‐directed serines in the C‐terminal region (Fig 3E), we excluded S328 as a candidate target for CCND1/CDK4 phosphorylation‐mediated protein stabilization, because it was reported that phosphorylation of S328 leads to protein degradation of ATOH1 (Forget et al, 2014; Cheng et al, 2016). Therefore, we focused on S309 ATOH1.
Previously, we reported that the CCND1/CDK4 complex phosphorylated purified wild‐type ATOH1 (WT‐ATOH1) but not S309A mutant ATOH1 (S309A‐ATOH1) in an in vitro kinase assay (Miyashita et al, 2021). However, our current experiments revealed that not only WT‐ATOH1 but also S309A‐ATOH1 gave rise to phosphorylation signals when incubated with CCND1/CDK4 (Fig EV3A). To directly investigate the phosphorylation of the S309 of ATOH1, we generated a specific antibody against S309‐phosphorylated ATOH1 (p‐S309‐Ab; Fig EV3B and C). We performed immunoblotting with this specific antibody to lysates in which purified WT‐ATOH1 or S309A‐ATOH1 was incubated with or without CCND1/CDK4(Fig 3F). While WT‐ATOH1 exhibited an obvious signal with CCND1/CDK4, S309A‐ATOH1 did not, suggesting that CCND1/CDK4 directly phosphorylates the S309 residue of ATOH1.
Fig 6H–J
We note that the bands for total Atoh1 in Fig 6H were miscropped at around 50 kDa, while the correct molecular weight of ATOH1 is about 40 kDa. We located the source data. Figure 6H is herewith withdrawn and replaced with the correct blot. The authors have performed requantification of the signal in Fig 6J, which is also herewith replaced.
Figure 6H. Original.
Figure 6H. Corrected.break;Source data are available online for this figure.
Figure 6J. Original.
Figure 6J. Corrected.
Materials and Methods
The Materials and Methods section “in vitro kinase assay” has been updated to include the details for the in vitro phosphorylation assays with anti‐S309‐phosphorylated‐ATOH1 antibody (new text in bold).
The reacted solution was then applied to the sample buffer and subjected to SDS polyacrylamide gel electrophoresis. Phosphorylation signals were detected using a Typhoon system (GE Healthcare).
In vitro phosphorylation assay
Purified WT‐ATOH1 (3–5 μg) or S309A‐ATOH1 was incubated with reaction buffer containing 40 μM ATP, 0.02% BSA, 50 mM Tris (pH 7.5), and 10 mM MgCl 2 with or without 5 μg of CCND1/CDK4 at 37°C for 1 h. The reacted solution was then applied to the sample buffer and subjected to SDS polyacrylamide gel electrophoresis, followed by immunoblotting with p‐S309Ab or total Atoh1 antibody.
Molecular weight markers
Figure panels 3A, C, F and G, 4A–E, 6H, EV1E and F, EV2D, EV3A–C, and EV4A and B are updated to include molecular weight markers. The figures along with the corresponding source data are published with this corrigendum.
Figure 4. Corrected.break;Source data are available online for this figure.
Figure EV1. Corrected.break;Source data are available online for this figure.
Figure EV2. Corrected.break;Source data are available online for this figure.
Figure EV4. Corrected.break;Source data are available online for this figure.
Manuscript text
Upon reassessment of these issues by the original reviewers, they indicated that references are missing for the two statements in the manuscript text. These are now added (new text in bold).
“Although S309 and S328 are both highly conserved proline‐directed serines in the C‐terminal region (Fig 3E), we excluded S328 as a candidate target for CCND1/CDK4 phosphorylation‐mediated protein stabilization, because it was reported that phosphorylation of S328 leads to protein degradation of ATOH1 (Forget et al, 2014; Cheng et al, 2016).”
“Phosphorylation of serine 328 (S328) of ATOH1 has been reported to promote protein degradation mediated by the E3‐ubiquitin ligase HUWE1 (Forget et al, 2014; Cheng et al, 2016).”
The authors state that these changes do not affect the interpretation of the data or the conclusions of the study and apologize for this oversight and for any inconvenience it may have caused.
All authors agree with this corrigendum.
Supporting information
Source Data for Expanded View
Source Data for Figure 3
Source Data for Figure 4
Source Data for Figure 6
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Source Data for Expanded View
Source Data for Figure 3
Source Data for Figure 4
Source Data for Figure 6