Translation repression via modulation of the cytoplasmic poly(A)-binding protein in the inflammatory response

[…] The study shows potential for progressing the field of cytoplasmic regulation of gene expression. The mouse knock-in approach has pioneering character and should inspire other researchers in the field. Yet, there are some unresolved mechanistic issues that would need to be addressed before publication in eLife. Also, there appears to be a substantial lack of scholarship in terms of citing earlier work.

The principal evidence for the proposed mechanism is a Zfp36 mutant (mF) that contains large deletions of the N- and C-termini. This mutant is not specific for Pabpc1, and based on published reports will also be deficient for many other interactions. Deletion of the N-terminus will abolish binding of GYF2/4EHP, which was reported to mediate translational silencing of TTP targets. Deletion of the C-terminus will abolish binding to the Ccr4-Not complex, which may also mediate translational silencing given its role in the translational repression of miRNA-bound mRNAs. Hence, the Zfp36-mF mutant is by no means specific, and does not prove that binding of Pabpc1 is critical for mediating translational suppression. Given the known role of Pabpc1 as a general enhancer of translation, it is difficult to conceive how it would act as a translational suppressor when bound to Zpf36. A key experiment the authors could carry out is an in vitro translation assay, where recombinant Zfp36 would specifically suppress translation of an ARE-containing mRNA only upon addition of recombinant Pabpc1, or fail to do so upon depletion of endogenous Pabpc1. Another experiment that would provide some additional support for the proposed mechanism (though not as conclusive as the in vitro translation assay) is if tethering of the Zfp36-mF deletion mutant would be sufficient to cause translational suppression. This mutant most likely does not interact with GYF2/4EHP or the Ccr4-Not complex, although it may still bind other proteins in addition to Pabpc1.

The fundamental question the reviewers asked here is whether or not Zfp36-mediated translational repression is dependent on Pabpc1.

Pabpc1 is the cytoplasmic poly(A) tail binding protein that associates with most of mRNAs. Importantly, perturbation on Pabpc1 may have pleiotropic effects to the cell. In the manuscript, we addressed the question via an system through modulating the poly(A) tail of a reporter mRNA. Specifically, we first generated an mRNA without the poly(A) tail by replacing the SV40 polyadenylation signal (SV40pA) on the FLuc reporter with a sequence from the 3’ end of MALAT1 (Figure 4C). Previous studies by Wilusz et al., 2012 showed that fusing the 3’ end sequence of MALAT1 can result in a poly(A) minus transcript stably present in the cytoplasm in mammalian cells. Indeed, when we fractionated the mRNAs via an oligod(T) column, we found that the Fluc mRNA with the 3’end MALAT1 sequence is predominantly in the poly(A) minus fraction, which is similar to the histone2ab mRNA, a known poly(A) minus transcript, but different from the Gapdh mRNA, a poly(A)+ transcript (Figure 4C). This result indicated that the Fluc-5BoxB-MALAT1 transcript we generated is a truly poly(A) minus mRNA. Next, we performed the tethering assay on this poly(A) minus transcript. We found that when tethered, Zfp36 fails to inhibit the translation of this poly(A) minus transcript (Figure 4F). This result is different from tethering Zfp36 to a poly(A) plus transcript (Figure 4A and B). Collectively, these observations indicate that Zfp36-mediated translational repression is dependent on the presence of the poly(A) tail.

We realize that although seems obvious, the poly(A) minus Fluc-5BoxB-MALAT1 transcript is devoid of Pabpc1 was still an assumption. To directly determine that Pabpc1 is not associated with the poly(A) minus Fluc-5BoxB-MALAT1 transcript we generated, we performed additional experiments. Specifically, we introduced a HA-Pabpc1 expressing plasmid with either the FLuc-5BoxB-SV40pA reporter or the Fluc-5BoxB-MALAT1 reporter into 293T cells. Then we performed IP on the HA-Pabpc1 from the transfected 293T cells (Figure 4D) followed by qRT-PCR to determine whether the poly(A) minus Fluc-5BoxB-MALAT1 transcript is associated with Pabcp1 (Figure 4E). We found that the Pabpc1 binds the poly(A)+ transcripts, such as the Gaphd1 mRNA and the FLuc-5BoxB-SV40pA transcript, but not the poly(A)- transcripts, such as the Histone2AB mRNA and the Fluc-5BoxB-MALAT1 transcript (Figure 4E). This experiment indicated that the FLuc-5BoxB-MALAT1 transcript is indeed devoid of Pabpc1. Since Zfp36 repressed the translation of the FLuc-5BoxB-SV40pA transcript (Figure 4B) but not that of the FLuc-5BoxB-MALAT1 transcript (Figure 4F), we conclude that Zfp36-mediated translation is dependent on Pabpc1.

In the manuscript, we identified the Zfp36-mF mutant because we wanted to use a minimal Zfp36 fragment that can be stably expressed and interacts with Pabpc1. Thereby, when over-expressed, this Zfp36 fragment can attenuate the endogenous Zfp36:Pabpc1 interaction. The Zfp36-mF fragment suits well for this purpose. As the reviewers described, the Zfp36-mF lacks both the N- and C-termini that interact with previous characterized translational repressors (GYF2/4EHP and Ccr4-Not complex, respectively). Thus, when over-expressed, the Zfp36-mF attenuates the endogenous Zfp36:Pabpc1 interaction (Figure 6A-C), but not the potential interactions of the endogenous Zfp36:GYF2/4EHP or Zfp36:Ccr4-Not complex. Therefore, the specific translational up-regulation of Zfp36 target mRNAs caused by Zfp36-mF over-expression is due to the disruption of the Zfp36:Pabpc1 interaction but not the potential Zfp36:GYF2/4EHP or Zfp36:Ccr4-Not complex interactions. These results support our conclusion that the Zfp36:Pabpc1 interaction is functionally important for Zfp36-mediated translational repression.

Pabpc1 promotes mRNA translation in multiple steps. Our data indicate that the translational repression mediated by Zfp36 is dependent on Pabpc1. We speculate that Zfp36 represses its target mRNA translation by compromising Pabpc1’s ability to maintain mRNA in a translation-competent state.

The role of Zfp36 as a translational silencer (in addition to its role as an inducer of mRNA degradation) has been reported previously, through the work of the Gaestel, Brooks and Lykke-Andersen labs. The authors fail to mention any of these studies, which is, at best, careless. The authors need to quote these publications adequately and discuss their findings in the light of these previous studies.

We apologize for this oversight. In the revised manuscript, in addition to citing review papers, we also included the papers describing the original findings from these labs.