On 2017 Jul 07, Wei Wang commented:

Several issues have been raised about this paper on PubPeer. Below are some of the comments:

Comment #1: "Jonathan Warner and colleagues failed to reproduce results reported in this paper. In their paper (PMID: 24860015), Warner et al. state:

"These results are at significant odds with the report that there can be 100-fold differences in the levels of the various RP mRNAs in several tissues of the developing mouse embryo (Kondrashov et al. 2011)."

Comment #3: "The authors wrote:

"Strikingly, we observed tremendous heterogeneity in RP expression among different RPs and within different tissues over a 250-fold range, expressed in log2 space (Figure 7D and Table S3)."

As Jon Warner and colleagues noted, this strikingly tremendous heterogeneity is at odds with any other dataset that we have seen. We downloaded the data from Table S3 and could not reproduce the 250-fold range either. Can the authors explain how they analyzed the data to find the 250-fold range ?

Hopefully, the authors can provide the raw data used for Figure 7D and Table S3.

Maybe some tissues express the mRNAs for all ribosomal proteins at higher levels because they need more ribosomes. Were the transcript data normalized for different global levels of ribosomal protein in different tissues ? "

Comment #4: "In addition to the mentioned issues with this paper, I would like to note some further points. I don't want to discredit the authors, they are certainly working on a very exciting topic, however, this paper makes conclusions and over interpretations, which are by no means proven by the data that they present. The authors claim to "uncover an important role for RPL38 in transcript-specific translational control". However, there are several important concerns with that claim:

(1) The authors don't provide any evidence that the ribosomes in the tissues/cells they are using are lacking RPL38, hence any assumptions about RPL38 specifically regulating the translation of these certain Hox mRNAs are purely speculative.

(2) As "evidence" that the translational defect of these mRNAs is specific to RPL38 deficiency, the authors present figure 4Q, where they are using additional mouse mutants. However, no data are provided to present how much the levels of the respectively affected ribosomal protein in the mutants are actually affected in the used tissues. Furthermore, out of the 8 Hox mRNAs, whose translation is impaired with RPL38 deficiency, the authors test the protein levels only for one (HoxA5) of these 8 transcripts in the additional mouse mutants, this seems like this was picked on purpose. And even for that 1 Hox protein, the data seem questionable. The RPS19 mutant seems to have also reduced HoxA5 protein levels when I look at the western, suggesting that the translational defect is not specific to RPL38 deficiency. The westerns for the other mutants seem to be slightly overloaded in the mutant samples compared to controls, hard to make a conclusion there.

(3) The polysome profiles in figure 3b are more than questionable. The polysomes are basically just a flat line, any quantitative data from RT-qPCR from isolated RNA of these polysomes don’t seem really reliable in terms of translation. Also, it appears hard to believe that the authors were able to dissect selectively the somites without including non-affected tissue material. "

On 2017 Jul 07, Wei Wang commented:

Several issues have been raised about this paper on PubPeer. Below are some of the comments:

Comment #1: "Jonathan Warner and colleagues failed to reproduce results reported in this paper. In their paper (PMID: 24860015), Warner et al. state:

"These results are at significant odds with the report that there can be 100-fold differences in the levels of the various RP mRNAs in several tissues of the developing mouse embryo (Kondrashov et al. 2011)."

Comment #3: "The authors wrote:

"Strikingly, we observed tremendous heterogeneity in RP expression among different RPs and within different tissues over a 250-fold range, expressed in log2 space (Figure 7D and Table S3)."

As Jon Warner and colleagues noted, this strikingly tremendous heterogeneity is at odds with any other dataset that we have seen. We downloaded the data from Table S3 and could not reproduce the 250-fold range either. Can the authors explain how they analyzed the data to find the 250-fold range ?

Hopefully, the authors can provide the raw data used for Figure 7D and Table S3.

Maybe some tissues express the mRNAs for all ribosomal proteins at higher levels because they need more ribosomes. Were the transcript data normalized for different global levels of ribosomal protein in different tissues ? "

Comment #4: "In addition to the mentioned issues with this paper, I would like to note some further points. I don't want to discredit the authors, they are certainly working on a very exciting topic, however, this paper makes conclusions and over interpretations, which are by no means proven by the data that they present. The authors claim to "uncover an important role for RPL38 in transcript-specific translational control". However, there are several important concerns with that claim:

(1) The authors don't provide any evidence that the ribosomes in the tissues/cells they are using are lacking RPL38, hence any assumptions about RPL38 specifically regulating the translation of these certain Hox mRNAs are purely speculative.

(2) As "evidence" that the translational defect of these mRNAs is specific to RPL38 deficiency, the authors present figure 4Q, where they are using additional mouse mutants. However, no data are provided to present how much the levels of the respectively affected ribosomal protein in the mutants are actually affected in the used tissues. Furthermore, out of the 8 Hox mRNAs, whose translation is impaired with RPL38 deficiency, the authors test the protein levels only for one (HoxA5) of these 8 transcripts in the additional mouse mutants, this seems like this was picked on purpose. And even for that 1 Hox protein, the data seem questionable. The RPS19 mutant seems to have also reduced HoxA5 protein levels when I look at the western, suggesting that the translational defect is not specific to RPL38 deficiency. The westerns for the other mutants seem to be slightly overloaded in the mutant samples compared to controls, hard to make a conclusion there.

(3) The polysome profiles in figure 3b are more than questionable. The polysomes are basically just a flat line, any quantitative data from RT-qPCR from isolated RNA of these polysomes don’t seem really reliable in terms of translation. Also, it appears hard to believe that the authors were able to dissect selectively the somites without including non-affected tissue material. "