Gle1 is required for tRNA to stimulate Dbp5 ATPase activity in vitro and to promote Dbp5 mediated tRNA export in vivo

Reviewer #1 (Public Review):

Summary:
This study focuses on the defining cellular pathways critical for tRNA export from the nucleus. While a number of these pathways have been identified, the observation that the primary transport receptors identified thus far (Los1 and Msn5) are not essential and that cells are viable even when both the genes are deleted supports the idea that there are as yet unidentified mediators of tRNA export from the nucleus. This study implicates the helicase Dbp5 in one of these parallel pathways arguing that Dbp5 works in a pathway that is independent of Los1 and/or Msn5. The authors present genetic data to support this conclusion. At least one result suggests that the idea of these pathways in parallel may be too simplistic as deletion of the LOS1 gene, which is not essential decreases the interaction of tRNA export substrate with Dbp5 (Figure 2A). If the two pathways were working in parallel, one might have expected removing one pathway to lead to an increase in the use of the other pathway and hence the interaction with a receptor in that pathway. The authors provide solid evidence that Dbp5 interacts with tRNA directly and that the addition of the factor Gle1 together with the previously identified co-factor InsP6 can trigger helicase activity and release of tRNA. The combination of in vivo studies and biochemistry provides evidence to consider how Dbp5 contributes to the export of tRNA and more broadly adds to the conversation about how coding and non-coding RNA export from the nucleus might be coordinated to control cell physiology.

Strengths and weaknesses:
A major strength of this manuscript is the multi-pronged approach to explore a potential role for the helicase Dbp5 in one of the multiple export pathways for tRNA from the nucleus.

The obvious missing experiment here with respect to genetics is the test of whether deletion of the MSN5 gene in the cells, which combines deletion of LOS1 and the dbp5_R423A allele, shown in Figure 1D would be lethal. This key experiment would lend substance to the argument that Dbp5 functions in a tRNA export pathway that is parallel to the Los1 and Msn5 pathways.

While some of the binding assays show rather modest band shifts (Figure 4B for example), the data in Figure 4A showing that there is no binding detected unless a non-hydrolyzable ATP analogue is employed, argues for specificity in nucleic acid binding. The question that does arise is whether the binding is specific for tRNA.

With the exception of the binding studies, which also employ a mixture of yeast tRNAs, this study relies primarily on a single tRNA species to come to the conclusions drawn. Many other studies have used multiple tRNAs to explore whether pathways characterized are generalizable to other tRNAs.

The authors provide evidence of a model where the helicase Dbp5 plays a role in tRNA export from the nucleus. Further evidence is required to determine whether Dbp5 could function in the same pathway as the previously defined tRNA export receptors, Los1 and Msn5. There are genetic tests that could be performed to explore this question. Some of the biochemistry presented would show when Los1 is absent that the interaction of Dbp5 with tRNA decreases, which could support a model where Dbp5 plays a role in coordination with Los1.

This work allows insight into key questions which still remain about the multiple pathways that are required for tRNA trafficking as well as how transport pathways for coding and non-coding RNAs might be coordinated. These questions are important as many of these pathways may be regulated in response to cellular conditions or during development and defining the fundamental pathways will be critical to understanding these dynamic processes.

Reviewer #2 (Public Review):

This submission is about the role of Dbp5/Gle1 in tRNA export. The manuscript provides data showing that Dbp5/Gle1 are involved in tRNA export from the nucleus which is an essential process critical to translation. The authors provide data that largely supports conclusions, however, there are some pieces of data that are misinterpreted. (Figure 1A and B look the same; in Fig 1E, the DAPI staining is abnormal; in Fig 4 the bands can't be seen.)

Additionally, the methods used are fairly standard so the article does not contain any new technical achievements.

Author Response

We would like to thank reviewers and editors for their thoughtful and constructive review of our manuscript. Below we have provided responses to specific points in the reviewers’ comments and eLIFE assessment, highlighting areas of the manuscript that will be edited for clarity and where efforts will be made to provide data to address reviewer concerns upon a future resubmission.

eLife assesment:

The authors report that Dbp5 functions in parallel with Los1 in tRNA export, in a manner dependent on Gle1 and requiring the ATPase cycle of Dbp5, but independent of Mex67, Dbp5's partner in mRNA export. The evidence for this conclusion is still incomplete, as is the biochemical evidence that Dbp5 interacts directly with tRNA in vitro with Gle1 and co-factor InsP6 triggering Dbp5 ATPase activity in the Dbp5-tRNA complex. The evidence that Dbp5 interacts with tRNA in cells independently of Los1, Msn5 and Mex67 is, however, solid.

We intend to edit the text to make clear our conclusions and accommodate clarifications on a few details of this assessment.

(1) We would clarify that our data supports a model in which Dbp5 recruitment to tRNA is independent of Mex67 as an adapter in cells; however, this does not mean that Mex67 and Dbp5 do not still co-function in tRNA export. For example, it is possible Dbp5 and Mex67 could still co-function in the same pathway, but instead of Dbp5 working down stream of Mex67, Dbp5 may in fact work upstream as an adapter for Mex67. Edits to the text will be made to ensure this distinction is clear and highlight the possibility for future investigation to elucidate this relationship.

(2) We would like to highlight that based on structural and biochemical data detailing synergistic activation of Dbp5 ATPase cycle by Gle1/InsP6 and single stranded RNA, it is difficult to imagine a scenario where the apparent synergistic activation of Dbp5 ATPase cycle by tRNA and Gle1/InsP6 (Figure 5) is achieved independent of direct RNA binding. For this reason, we still support the claim that the observed synergistic activation, in combination with other in-vivo and in-vitro data provided in the manuscript, support a model where Dbp5 directly binds tRNA. However, we intend to edit the text to highlight this nuance and potential alternative conclusions based on reviewer feedback.

Reviewer #1 (Public Review):

“At least one result suggests that the idea of these pathways in parallel may be too simplistic as deletion of the LOS1 gene, which is not essential decreases the interaction of tRNA export substrate with Dbp5 (Figure 2A). If the two pathways were working in parallel, one might have expected removing one pathway to lead to an increase in the use of the other pathway and hence the interaction with a receptor in that pathway…. The obvious missing experiment here with respect to genetics is the test of whether deletion of the MSN5 gene in the cells, which combines deletion of LOS1 and the dbp5_R423A allele, shown in Figure 1D would be lethal…. The authors provide evidence of a model where the helicase Dbp5 plays a role in tRNA export from the nucleus. Further evidence is required to determine whether Dbp5 could function in the same pathway as the previously defined tRNA export receptors, Los1 and Msn5. There are genetic tests that could be performed to explore this question. Some of the biochemistry presented would show when Los1 is absent that the interaction of Dbp5 with tRNA decreases, which could support a model where Dbp5 plays a role in coordination with Los1”

We agree that this is an important point that should be made clear and discussed in the text. We also agree that further experiments would be needed to be to confirm Dbp5 functions broadly in tRNA export in parallel to both Msn5 and Los1. We will aim to address these points in resubmission and discuss possible alternative conclusions of the presented results.

Reviewer #1 (Public Review):

“While some of the binding assays show rather modest band shifts (Figure 4B for example), the data in Figure 4A showing that there is no binding detected unless a non-hydrolyzable ATP analogue is employed, argues for specificity in nucleic acid binding. The question that does arise is whether the binding is specific for tRNA.”

The specificity of the in-vitro interactions of Dbp5 are an important point of discussion. We will work to expand the topic of specificity of the in-vitro experiments during resubmission.

Reviewer #1 (Public Review):

“With the exception of the binding studies, which also employ a mixture of yeast tRNAs, this study relies primarily on a single tRNA species to come to the conclusions drawn. Many other studies have used multiple tRNAs to explore whether pathways characterized are generalizable to other tRNAs.“

It was previously shown that Dbp5 functions to support the export of multiple tRNA species (https://doi.org/10.7554/eLife.48410). As such, we agree that additional tRNAs should be tested to explore whether phenotypes reported here are also generalizable to other tRNAs. We will add data targeting additional tRNAs during resubmission.

Reviewer #2 (Public Review):

“there are some pieces of data that are misinterpreted. (Figure 1A and B look the same; in Fig 1E, the DAPI staining is abnormal; in Fig 4 the bands can't be seen.)”

Figure 1A and B represent separate experiments, showing that deletion of Los1 does not alter Dbp5 localization and conversely loss of Dbp5 does not alter Los1 localization. As such localization patterns under loss-of-function conditions look the same as wild-type localization for each protein respectively as noted. We believe that we have come to the same conclusion as the reviewer on Figure 1A and B (and this data is not misinterpreted), but also understand this panel will need to be adjusted for clarity and readability. We will make efforts to edit this figure and accompanying text make the data and conclusions clearer, including addressing the EMSAs in figure 4 and associated text for clarity.