Design principles of the ESCRT-III Vps24-Vps2 module
Curation statements for this article:-
Curated by eLife
Evaluation Summary:
Different ESCRT-III subunits share sequence similarity but have been characterized in distinct conformations and perform distinct roles in the polymerization process that is central to ESCRT membrane fission pathways. Here it is shown that mutations in one ESCRT-III subunit can compensate for loss of a different subunit by encoding the specialized roles of both subunits within one polypeptide. These findings will be of interest to investigators studying ESCRT pathway mechanisms and those more generally interested in the adaptability of protein sequences and functions.
(This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. Reviewer #3 agreed to share their name with the authors.)
This article has been Reviewed by the following groups
Listed in
- Evaluated articles (eLife)
- Cell Biology (eLife)
Abstract
ESCRT-III polymerization is required for all endosomal sorting complex required for transport (ESCRT)-dependent events in the cell. However, the relative contributions of the eight ESCRT-III subunits differ between each process. The minimal features of ESCRT-III proteins necessary for function and the role for the multiple ESCRT-III subunits remain unclear. To identify essential features of ESCRT-III subunits, we previously studied the polymerization mechanisms of two ESCRT-III subunits Snf7 and Vps24, identifying the association of the helix-4 region of Snf7 with the helix-1 region of Vps24 (Banjade et al., 2019a). Here, we find that mutations in the helix-1 region of another ESCRT-III subunit Vps2 can functionally replace Vps24 in Saccharomyces cerevisiae . Engineering and genetic selections revealed the required features of both subunits. Our data allow us to propose three minimal features required for ESCRT-III function – spiral formation, lateral association of the spirals through heteropolymerization, and binding to the AAA + ATPase Vps4 for dynamic remodeling.
Article activity feed
-
-
Author Response:
Reviewer #1 (Public Review):
This Research Advance builds on the findings of this group's 2019 eLife paper which showed that conserved acidic and basic helices associate to enable heteropolymer formation by Snf7 and Vps24. This work provides some general structure/sequence relationships among the homologous ESCRT-III proteins that will be of interest to those in the ESCRT field. While there are no new mechanistic principles obtained from this study, the data allow the authors to propose a model of the minimal or core units needed for ESCRT-III membrane remodeling.
The focus is largely on similarities and differences between the closely related Vps24 and Vps2, where they show that a few key point mutations or chimeric swaps (for Vps4 binding by the C-terminal region of Vps2) can exchange their functions. The last …
Author Response:
Reviewer #1 (Public Review):
This Research Advance builds on the findings of this group's 2019 eLife paper which showed that conserved acidic and basic helices associate to enable heteropolymer formation by Snf7 and Vps24. This work provides some general structure/sequence relationships among the homologous ESCRT-III proteins that will be of interest to those in the ESCRT field. While there are no new mechanistic principles obtained from this study, the data allow the authors to propose a model of the minimal or core units needed for ESCRT-III membrane remodeling.
The focus is largely on similarities and differences between the closely related Vps24 and Vps2, where they show that a few key point mutations or chimeric swaps (for Vps4 binding by the C-terminal region of Vps2) can exchange their functions. The last portion of the paper further tests similarities within the subgroups of ESCRT-III proteins to experimentally test functional groupings defined by sequence relationships.
We thank the reviewer for their generous comments. We’d like to emphasize that one of the main focus behind this study is to be able to generate minimal ESCRT-III system that can be functional. We study Vps24 and Vps2 to generate a model ESCRT-III module with their specific properties. We previously engineered Snf7 to replace Vps20 (and other ESCRT components, eLife 2016). In this paper, we also extend some of the analysis to other ESCRT-III components. We agree that this current manuscript combines previously described mechanisms to understand the minimal ESCRT-III system and provides us a direction to understand why in some cases (for example archaeal system), there may be only two ESCRT-III subunits. This work, following up on previous works from our lab and others, takes us one additional step toward that direction.
In addition, we’d also like to highlight from our work that in yeast, MVB biogenesis does have strong contributions from Did2 (CHMP1) and Vps60 (CHMP5), but not from Ist1 (IST1) and Chm7 (CHMP7) (Fig. 5). These have previously been under-emphasized in the literature.
Reviewer #2 (Public Review):
The manuscript by Emr and colleagues addresses the important question of how core ESCRT-III members Vps2 and Vps24 interact to form functional polymers using protein engineering and genetic selection approaches.
Major findings are:
Vps2 overexpression can functionally replace Vps24 in MVB sorting.
Helix 1 N21K, T28A, E31K mutations, Vps2, were identified to be sufficient for suppression, concluding that Vps2 and its' over expression can replace the function of Vps24 and Vps2.
Vps24 over expression does not rescue delta Vps2. The authors propose that this is due to the lack of the MIM and helix5 binding sites for Vps4 present in Vps2.
Vps24 E114K mutation was identified to rescue deltaVps2 upon over expression and even better as a Vps24/Vps2 chimera suggesting that auto-activated Vps24 that can recruit Vps4 can functionally replace Vps2.
Analyzing the effect of single ESCRT-III deletions on Mup1 sorting confirmed Snf7, Vps20, Vps2 and Vps24 as essential for sorting.
In summary, the manuscript provides new insight into the assembly of ESCRT-III. It confirms some redundancy of VPS2 and Vps24 and shows how Vps2 can substitute Vps24 but not vice versa.
We thank the reviewer for this summary of our work. One point we’d like to emphasize is that while we agree that Snf7, Vps20, Vps2 and Vps24 form a minimal core subunit to form MVBs, there are important functions of other ESCRT-III molecules Did2 and Vps60 (Figure 5 and supplement) for MVB biogenesis.
Comments:
The three minimal principles for ESCRT-III assembly stated in the abstract are not novel. Spiral formation of ESCRT-III has been described before for yeast Vps2-Vps24 as well as its mammalian homologues. The requirement for VPS4 recruitment is also well documented and finally, the manuscript does not provide proof for lateral association of the spirals via hetero-polymerization.
We agree with the first two comments about spiral formation and Vps4 recruitment. We’d like to emphasize that the lateral association through heteropolymerization mechanism extends from our previous work (eLife 2019) and supported by this work through mutational analysis of Vps2’s helix-1 motif. In our previous work, we provided evidence of the association of Snf7’s helix-4 region with Vps24’s helix-1 region, and also lateral association of Snf7 and Vps24/Vps2 with in vitro assays. In the previous work, we didn’t characterize Vps2-Snf7 interaction, which we do further in this work. We find that charge-inversion mutations in Vps2 increases its affinity to Snf7, and this effect is sufficient to replace Vps24. We believe that these analyses strengthen our model and also enhance our knowledge of ESCRT-III polymerization. Therefore, this manuscript a strong extension/advance on our previous eLife paper, and both papers should be analyzed together.
The authors show that 8-fold over expression is necessary to rescue Mup1 sorting to an extent of 40%. The authors hypothesize that over expression of Vps2 can rescue Vps24 deletion because Vps2 may have a lower affinity for Snf7 than Vps24. This is in agreement with data on mammalian homologues which showed that indeed CHMP3 binds with 10x higher affinity to CHMP4B than CHMP2A (Effantin et al, 2012). This could have been included in the discussion, since the function of yeast and mammalian core ESCRT-III proteins is most likely not different.
We apologize for this oversight and have included appropriate reference to this paper in the next version.
The authors designed several chimeric Vps24/Vps2 constructs and show that some of the Vps24 chimera including Vps2 helix 5 and the MIM are fully functional in Mup1 sorting in delta Vps24 cells, but lack the ability to functionally replace Vps2 in Vps2 delta cells. It is unclear whether the chimeras are in the closed conformation in the cytosol. It would be interesting to know whether they are activated more easily and possibly prematurely.
With our current assays we cannot distinguish the open vs. closed conformations in solution vs. membrane for Vps24. We do not think that these chimeras are activated prematurely because they do remain functional (as highlighted by the reviewer) in vps24∆ strain.
We’d like to thank the reviewer for pointing us to these mutants, which have encouraged us to further study these and related chimeras. To understand the role of swapping the Vps2 helix5 and MIM region further, we have added a couple of more experiments that would allow us to further understand the role of these motifs.
We replaced the helix-5 and MIM regions of Vps2 onto Snf7 to ask whether this construct remains functional, and whether they can replace function of Vps24-Vps2 (by directly recruiting Vps4).
In these set of data, we present evidence that when incorporated into Snf7, the helices 5 and MIM motifs of Vps2 make this chimeric Snf7 dysfunctional (Fig. 3 – Supp. 3). These data are consistent with the reviewers’ interpretation that premature recruitment of Vps4 to ESCRT-III filaments is presumably dysfunctional. However, inclusion of these motifs to Vps24 most likely does not prematurely disassemble ESCRT-III filaments, hence they remain functional. Also, mere substitution of the H5 and MIM motif to Snf7 (and therefore the Vps4 binding) is not sufficient for ESCRT-III function in cells.
The larger point behind this set of analyses is that there are additional functions of Vps24-Vps2 beyond just recruitment of the AAA+ ATPase Vps4. Since we extensively analyzed the lateral association of Vps24-Vps2 to Snf7 in our previous manuscript (Banjade et al., eLife 2019), we ascribe these additional functions to lateral polymerization of Vps24-Vps2 on the Snf7 filament.
The authors show that Vps24 E114K can form some kind of polymers in the presence of Vps2 in vitro while no polymerization is observed for wt Vps24 at 1 µM. It would be interesting to know whether wt Vps24 polymerizes at higher concentrations in this assay.
We don’t observe polymers with 15 µM of Vps24 and 15 µM of Vps2, as the proteins start forming amorphous assemblies. We do refer to other manuscripts in the past who have observed similar linear polymers of Vps24 at higher concentrations (>300 µM) and longer incubation. So we believe that the ESCRT-III proteins Vps24 and Vps2 are able to form copolymers with a similar structure that is enhanced when these “activating” mutations are included.
While the conclusion that E114K shifts the equilibrium to the open state is plausible, there is no evidence provided that this mimics Vps2 as stated. If so, Vps24 E114k should form the same polymers as shown in figure 4 supp 1 in the absence of Vps2 and spiral formation with snf7 should not require Vps2.
We agree with this interpretation from the in vitro assays, and have appropriately changed the language in the manuscript. We now describe the effect of the E114K protein to “enhance” associated with existing Vps2. We hypothesize that this enhanced association to Vps2 occurs due to an “activation” process whereby Vps24 adopts a higher population of an open (or a semi-open) conformation, and have changed the language to reflect this interpretation. As an aside, we do note that Snf7 and Vps24 do form helices at higher concentrations without Vps2, as we showed in Banjade et al., eLife 2019.
The speculation in the results section that Vps24 may not extend its helices 2 and 3 in an activated form due to potential helix breaking Asn residues in the linker region is not backed up by data, and it would have been appropriate to indicate this in the manuscript.
We have now moved this analysis to the discussion and emphasized that this is a hypothesis. We also added the following sentence when describing the data regarding the mutations in the potential helix-breaking Asn residues: “We note that these data are indicative of mutations that control the conformations of the proteins. However, further biophysical analyses will be required for definitive evidence of this conformational flexibility.”
The proposal that Vps2-Vps24 heteropolymers are formed by interactions along helices 2 and 3 is not supported by data presented in the manuscript. The authors would need to use recombinant proteins to test their mutants in biophysical interaction studies.
We have now moved this interpretation to the discussion. Further dissection with biochemical and biophysical assays of Vps24-Vps2 would be a future direction in this area.
Reviewer #3 (Public Review):
This study sought to identify essential features of ESCRT-III subunits, with a focus on the yeast proteins Vps2 and Vps24, in order to reveal the required features of both subunits. The combined genetic and biochemical studies solidified the model that essential functions of ESCRT-III polymers - spiral formation, lateral association, and binding of Vps4 - are mostly distributed between different subunits (with some redundancy) and can be engineered into a single polypeptide. This study also sheds light on the long-standing and initially surprising finding that ESCRT-dependent budding of HIV does not require CHMP3 (Vps24), presumably because the distribution of distinct functions between different ESCRT-III subunits is not absolute.
Inspired by earlier studies, the ability of overexpression of one ESCRT-III subunit to compensate for deletion of another subunit was explored using sorting assays. The demonstration of partial rescue inspired a mutagenesis approach that identified three residues that cluster on one face of a helix that enhanced rescue, and therefore confer functionality that in wt is primarily provided in the deleted subunits, which in this case is binding to Snf7. Extension of this analysis by protein engineering further demonstrated that the essential role of recruiting the Vps4 ATPase is normally performed by Vps2 but can be transferred to Vps24 by substitution of residues near the ESCRT-III subunit C-terminus. Similarly, it is shown that sequences that alter the propensity for bending of a helix at a point where open and closed ESCRT-III subunits differ in conformation contributed to the ability of Vps24 to substitute for deletion of Vps2, presumably by conferring the ability to adopt the open, activated conformation as well as the closed conformation.
I don't have concerns about design or technical aspects of the experimental approach.
We appreciate the reviewer’s comments and the summary of our work.
-
Reviewer #3 (Public Review):
This study sought to identify essential features of ESCRT-III subunits, with a focus on the yeast proteins Vps2 and Vps24, in order to reveal the required features of both subunits. The combined genetic and biochemical studies solidified the model that essential functions of ESCRT-III polymers - spiral formation, lateral association, and binding of Vps4 - are mostly distributed between different subunits (with some redundancy) and can be engineered into a single polypeptide. This study also sheds light on the long-standing and initially surprising finding that ESCRT-dependent budding of HIV does not require CHMP3 (Vps24), presumably because the distribution of distinct functions between different ESCRT-III subunits is not absolute.
Inspired by earlier studies, the ability of overexpression of one ESCRT-III …
Reviewer #3 (Public Review):
This study sought to identify essential features of ESCRT-III subunits, with a focus on the yeast proteins Vps2 and Vps24, in order to reveal the required features of both subunits. The combined genetic and biochemical studies solidified the model that essential functions of ESCRT-III polymers - spiral formation, lateral association, and binding of Vps4 - are mostly distributed between different subunits (with some redundancy) and can be engineered into a single polypeptide. This study also sheds light on the long-standing and initially surprising finding that ESCRT-dependent budding of HIV does not require CHMP3 (Vps24), presumably because the distribution of distinct functions between different ESCRT-III subunits is not absolute.
Inspired by earlier studies, the ability of overexpression of one ESCRT-III subunit to compensate for deletion of another subunit was explored using sorting assays. The demonstration of partial rescue inspired a mutagenesis approach that identified three residues that cluster on one face of a helix that enhanced rescue, and therefore confer functionality that in wt is primarily provided in the deleted subunits, which in this case is binding to Snf7. Extension of this analysis by protein engineering further demonstrated that the essential role of recruiting the Vps4 ATPase is normally performed by Vps2 but can be transferred to Vps24 by substitution of residues near the ESCRT-III subunit C-terminus. Similarly, it is shown that sequences that alter the propensity for bending of a helix at a point where open and closed ESCRT-III subunits differ in conformation contributed to the ability of Vps24 to substitute for deletion of Vps2, presumably by conferring the ability to adopt the open, activated conformation as well as the closed conformation.
I don't have concerns about design or technical aspects of the experimental approach.
-
Reviewer #2 (Public Review):
The manuscript by Emr and colleagues addresses the important question of how core ESCRT-III members Vps2 and Vps24 interact to form functional polymers using protein engineering and genetic selection approaches.
Major findings are:
Vps2 overexpression can functionally replace Vps24 in MVB sorting.
Helix 1 N21K, T28A, E31K mutations, Vps2, were identified to be sufficient for suppression, concluding that Vps2 and its' over expression can replace the function of Vps24 and Vps2.
Vps24 over expression does not rescue delta Vps2. The authors propose that this is due to the lack of the MIM and helix5 binding sites for Vps4 present in Vps2.
Vps24 E114K mutation was identified to rescue deltaVps2 upon over expression and even better as a Vps24/Vps2 chimera suggesting that auto-activated Vps24 that can recruit Vps4 …
Reviewer #2 (Public Review):
The manuscript by Emr and colleagues addresses the important question of how core ESCRT-III members Vps2 and Vps24 interact to form functional polymers using protein engineering and genetic selection approaches.
Major findings are:
Vps2 overexpression can functionally replace Vps24 in MVB sorting.
Helix 1 N21K, T28A, E31K mutations, Vps2, were identified to be sufficient for suppression, concluding that Vps2 and its' over expression can replace the function of Vps24 and Vps2.
Vps24 over expression does not rescue delta Vps2. The authors propose that this is due to the lack of the MIM and helix5 binding sites for Vps4 present in Vps2.
Vps24 E114K mutation was identified to rescue deltaVps2 upon over expression and even better as a Vps24/Vps2 chimera suggesting that auto-activated Vps24 that can recruit Vps4 can functionally replace Vps2.
Analyzing the effect of single ESCRT-III deletions on Mup1 sorting confirmed Snf7, Vps20, Vps2 and Vps24 as essential for sorting.
In summary, the manuscript provides new insight into the assembly of ESCRT-III. It confirms some redundancy of VPS2 and Vps24 and shows how Vps2 can substitute Vps24 but not vice versa.
Comments:
The three minimal principles for ESCRT-III assembly stated in the abstract are not novel. Spiral formation of ESCRT-III has been described before for yeast Vps2-Vps24 as well as its mammalian homologues. The requirement for VPS4 recruitment is also well documented and finally, the manuscript does not provide proof for lateral association of the spirals via hetero-polymerization.
The authors show that 8-fold over expression is necessary to rescue Mup1 sorting to an extent of 40%. The authors hypothesize that over expression of Vps2 can rescue Vps24 deletion because Vps2 may have a lower affinity for Snf7 than Vps24. This is in agreement with data on mammalian homologues which showed that indeed CHMP3 binds with 10x higher affinity to CHMP4B than CHMP2A (Effantin et al, 2012). This could have been included in the discussion, since the function of yeast and mammalian core ESCRT-III proteins is most likely not different.
The authors designed several chimeric Vps24/Vps2 constructs and show that some of the Vps24 chimera including Vps2 helix 5 and the MIM are fully functional in Mup1 sorting in delta Vps24 cells, but lack the ability to functionally replace Vps2 in Vps2 delta cells. It is unclear whether the chimeras are in the closed conformation in the cytosol. It would be interesting to know whether they are activated more easily and possibly prematurely.
The authors show that Vps24 E114K can form some kind of polymers in the presence of Vps2 in vitro while no polymerization is observed for wt Vps24 at 1 µM. It would be interesting to know whether wt Vps24 polymerizes at higher concentrations in this assay.
While the conclusion that E114K shifts the equilibrium to the open state is plausible, there is no evidence provided that this mimics Vps2 as stated. If so, Vps24 E114k should form the same polymers as shown in figure 4 supp 1 in the absence of Vps2 and spiral formation with snf7 should not require Vps2.
The speculation in the results section that Vps24 may not extend its helices 2 and 3 in an activated form due to potential helix breaking Asn residues in the linker region is not backed up by data, and it would have been appropriate to indicate this in the manuscript.
The proposal that Vps2-Vps24 heteropolymers are formed by interactions along helices 2 and 3 is not supported by data presented in the manuscript. The authors would need to use recombinant proteins to test their mutants in biophysical interaction studies.
-
Reviewer #1 (Public Review):
This Research Advance builds on the findings of this group's 2019 eLife paper which showed that conserved acidic and basic helices associate to enable heteropolymer formation by Snf7 and Vps24. This work provides some general structure/sequence relationships among the homologous ESCRT-III proteins that will be of interest to those in the ESCRT field. While there are no new mechanistic principles obtained from this study, the data allow the authors to propose a model of the minimal or core units needed for ESCRT-III membrane remodeling.
The focus is largely on similarities and differences between the closely related Vps24 and Vps2, where they show that a few key point mutations or chimeric swaps (for Vps4 binding by the C-terminal region of Vps2) can exchange their functions. The last portion of the paper …
Reviewer #1 (Public Review):
This Research Advance builds on the findings of this group's 2019 eLife paper which showed that conserved acidic and basic helices associate to enable heteropolymer formation by Snf7 and Vps24. This work provides some general structure/sequence relationships among the homologous ESCRT-III proteins that will be of interest to those in the ESCRT field. While there are no new mechanistic principles obtained from this study, the data allow the authors to propose a model of the minimal or core units needed for ESCRT-III membrane remodeling.
The focus is largely on similarities and differences between the closely related Vps24 and Vps2, where they show that a few key point mutations or chimeric swaps (for Vps4 binding by the C-terminal region of Vps2) can exchange their functions. The last portion of the paper further tests similarities within the subgroups of ESCRT-III proteins to experimentally test functional groupings defined by sequence relationships.
-
Evaluation Summary:
Different ESCRT-III subunits share sequence similarity but have been characterized in distinct conformations and perform distinct roles in the polymerization process that is central to ESCRT membrane fission pathways. Here it is shown that mutations in one ESCRT-III subunit can compensate for loss of a different subunit by encoding the specialized roles of both subunits within one polypeptide. These findings will be of interest to investigators studying ESCRT pathway mechanisms and those more generally interested in the adaptability of protein sequences and functions.
(This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. Reviewer #3 agreed to share their name with the authors.)
-
-