Recruitment of clathrin to intracellular membranes is sufficient for vesicle formation
Curation statements for this article:-
Curated by eLife
Evaluation Summary:
This manuscript reports a striking finding, which should be of interest to cell biologists and biophysicists. The authors use an innovative approach to recruit clathrin to mitochondrial membranes and observe the budding and fission of clathrin-coated vesicles. The study leads to a much clearer view of how the clathrin lattice functions in endocytosis.
(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. The reviewers remained anonymous to the authors.)
This article has been Reviewed by the following groups
Listed in
- Evaluated articles (eLife)
- Cell Biology (eLife)
Abstract
The formation of a clathrin-coated vesicle (CCV) is a major membrane remodeling process that is crucial for membrane traffic in cells. Besides clathrin, these vesicles contain at least 100 different proteins although it is unclear how many are essential for the formation of the vesicle. Here, we show that intracellular clathrin-coated formation can be induced in living cells using minimal machinery and that it can be achieved on various membranes, including the mitochondrial outer membrane. Chemical heterodimerization was used to inducibly attach a clathrin-binding fragment ‘hook’ to an ‘anchor’ protein targeted to a specific membrane. Endogenous clathrin assembled to form coated pits on the mitochondria, termed MitoPits, within seconds of induction. MitoPits are double-membraned invaginations that form preferentially on high curvature regions of the mitochondrion. Upon induction, all stages of CCV formation – initiation, invagination, and even fission – were faithfully reconstituted. We found no evidence for the functional involvement of accessory proteins in this process. In addition, fission of MitoPit-derived vesicles was independent of known scission factors including dynamins and dynamin-related protein 1 (Drp1), suggesting that the clathrin cage generates sufficient force to bud intracellular vesicles. Our results suggest that, following its recruitment, clathrin is sufficient for intracellular CCV formation.
Article activity feed
-
-
Evaluation Summary:
This manuscript reports a striking finding, which should be of interest to cell biologists and biophysicists. The authors use an innovative approach to recruit clathrin to mitochondrial membranes and observe the budding and fission of clathrin-coated vesicles. The study leads to a much clearer view of how the clathrin lattice functions in endocytosis.
(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. The reviewers remained anonymous to the authors.)
-
Reviewer #1 (Public Review):
Clathrin-mediated endocytosis has been extensively studied by many labs, resulting in a highly complex picture of a network of interacting proteins that need to be recruited to the right place at the right time. Some of these proteins are thought to act as pioneers, initiating the recruitment of the rest of the coat; others as curvature generators or sensors; while the final pinching off of the vesicle requires the GTPase dynamin. Much less is known about CCV formation from intracellular membranes, specifically the trans-Golgi network and endosomes, but it is generally assumed that the situation is very similar.
In the present study, the authors turn this assumption on its head by showing that in fact, it is possible to form a CCV from essentially any intracellular membrane by simply targeting a protein with …
Reviewer #1 (Public Review):
Clathrin-mediated endocytosis has been extensively studied by many labs, resulting in a highly complex picture of a network of interacting proteins that need to be recruited to the right place at the right time. Some of these proteins are thought to act as pioneers, initiating the recruitment of the rest of the coat; others as curvature generators or sensors; while the final pinching off of the vesicle requires the GTPase dynamin. Much less is known about CCV formation from intracellular membranes, specifically the trans-Golgi network and endosomes, but it is generally assumed that the situation is very similar.
In the present study, the authors turn this assumption on its head by showing that in fact, it is possible to form a CCV from essentially any intracellular membrane by simply targeting a protein with a clathrin-binding domain to that membrane; the clathrin in the cell will do the rest. They succeeded in reconstituting CCV formation from mitochondria, the ER, the Golgi apparatus, and lysosomes. However, they mainly focus on mitochondria, as this result was particularly surprising, given that mitochondria aren't even part of the endomembrane system. The authors employed multiple complementary approaches. For instance, they used four methods to confirm that the "MitoPits" that form from mitochondria are coated with clathrin: immunofluorescence, electron microscopy, clathrin knockdown, and several different constructs with clathrin-binding sites, together with control constructs with point mutations in the clathrin-binding site. Live cell imaging was used to show that not only buds but actual vesicles were being formed from mitochondria, even in the absence of a "pinchase" like dynamin. Some of the data are presented in the form of "SuperPlots", showing each actual point from biological replicates rather than bar graphs with error bars. Thus, the authors' conclusion, that the other proteins implicated in CCV formation are modulators rather than mediators, is well justified.
-
Reviewer #2 (Public Review):
In this work, Kuey et al develop a synthetic system to test if clathrin, on its own, can bend and cut membranes inside living cells. The authors expand a previously-developed "hotwire" system to recruit a clathrin-binding hook (Beta2 hinge/appendage domain) to the cytoplasmic face of the mitochondrial outer membrane through an FKBP/rapamycin/FRB dimerization system. After the addition of the ligand, clathrin was observed to form puncta on mitochondria and create double-membrane clathrin-coated vesicles. This process was independent of many classic endocytic proteins including dynamin. The authors propose that clathrin itself is the only protein needed to generate the forces required to drive vesicle formation and scission inside cells. This work supports the idea that the dozens of accessory proteins …
Reviewer #2 (Public Review):
In this work, Kuey et al develop a synthetic system to test if clathrin, on its own, can bend and cut membranes inside living cells. The authors expand a previously-developed "hotwire" system to recruit a clathrin-binding hook (Beta2 hinge/appendage domain) to the cytoplasmic face of the mitochondrial outer membrane through an FKBP/rapamycin/FRB dimerization system. After the addition of the ligand, clathrin was observed to form puncta on mitochondria and create double-membrane clathrin-coated vesicles. This process was independent of many classic endocytic proteins including dynamin. The authors propose that clathrin itself is the only protein needed to generate the forces required to drive vesicle formation and scission inside cells. This work supports the idea that the dozens of accessory proteins important for CME at the plasma membrane play roles specific to the forces, lipids, or cargo encountered at the plasma membrane.
This is an excellent and well-done study. The questions are important and the experiments creative. It leads to a much clearer view of how the clathrin lattice functions in endocytosis. It is a controlled dissection of the endocytic machinery. It is particularly useful because many of the proteins involved in endocytosis have overlapping and complex roles.
-
Reviewer #3 (Public Review):
Kuey et al describe an interesting new biochemical system that allows the chemically induced recruitment of clathrin to various organelles of eukaryotic cells. Their major focus is on recruitment to outer mitochondrial membranes. Upon recruiting clathrin to these surfaces, bright puncta of clathrin are formed, which subsequently form buds and vesicles that separate from the parent membrane, as confirmed by fluorescence and electron microscopy. The authors then ask what is responsible for the fission of these vesicles. They knockdown dynamin, and the mitochondrial dynamin, drp1, with little effect on the efficiency of fission. From these results, they conclude that fission is spontaneous, likely owing to the clathrin coat itself. The authors also examine the recruitment of clathrin adaptor and accessory …
Reviewer #3 (Public Review):
Kuey et al describe an interesting new biochemical system that allows the chemically induced recruitment of clathrin to various organelles of eukaryotic cells. Their major focus is on recruitment to outer mitochondrial membranes. Upon recruiting clathrin to these surfaces, bright puncta of clathrin are formed, which subsequently form buds and vesicles that separate from the parent membrane, as confirmed by fluorescence and electron microscopy. The authors then ask what is responsible for the fission of these vesicles. They knockdown dynamin, and the mitochondrial dynamin, drp1, with little effect on the efficiency of fission. From these results, they conclude that fission is spontaneous, likely owing to the clathrin coat itself. The authors also examine the recruitment of clathrin adaptor and accessory proteins to these mitochondrial buds, finding that only epsin and fcho are recruited in measurable quantities. Further, when these proteins are knocked down, the number of fission events is unchanged, suggesting that they did not play a major role in membrane vesiculation in this specific setting.
Overall the experimental work is innovative, clear, and well controlled from my perspective as a biophysicist. That being said, cell biologists will have a clearer understanding of the necessary controls and should be consulted and deferred to on this point.
Unfortunately, the authors spoil the considerable appeal of their paper with a discussion section that is highly speculative and at times illogical. Overall it promotes an over-extension of the authors' findings to questions that are not addressed by their results. The most significant issue is the authors' assertion that their data prove that the many adaptors of the clathrin pathway, many of which have been directly shown to bend membranes to very high curvatures, are only "modulators" rather than true drivers of membrane bending in cells. The authors should consult the recent preprint of Cail and Drubin, which shows the exact opposite: https://www.biorxiv.org/content/10.1101/2021.07.15.452420v3. Note: I am not an author of this paper. Here the authors are also working in mammalian cells and have pre-defined the curvature by culturing cells on top of ridged substrates. They make a nearly complete knockdown of the clathrin heavy chain with the result that clathrin is no longer detected at the plasma membrane. Nevertheless, they observe that puncta form and vesiculate from the membrane surface, which is enriched in the many adaptor proteins of the clathrin pathway. These results show that when curvature is pre-imposed, as is the case (and to a very similar degree) for budding from mitochondria, clathrin is NOT required. Taking the two papers together, my interpretation is that BOTH clathrin AND adaptors make substantial contributions to membrane curvature. When curvature is pre-imposed, EITHER clathrin or adaptors can drive vesiculation. However, on the relatively flat plasma membrane, BOTH clathrin and adaptors are required, as has been shown in many studies.
A second major issue is the authors' attempt to put forth a biophysical model of bending. Here they simply compare clathrin's estimated contribution to curvature energy with the energetic cost of membrane bending. Notably, this model overlooks the sizable energy barrier to initiating the initial curvature. The authors should look at the work of Derenyi and Prost, among others to understand this better. But putting that significant issue aside, the authors argue that they can use a very low membrane bending rigidity of 15 kbT to account for the cost of bending the mitochondrial membrane. 15kbT would be a relevant value for membrane consisting of pure, unsaturated lipids like DOPC or at most POPC. This is certainly a vast underestimate, given that the mitochondrial membrane, like all biological membranes, is filled with proteins that substantially raise its rigidity. 50kbT or greater would be a more relevant value, which would reverse the authors' conclusions.
A third major issue is the authors' dismissal of molecular crowding as a possible explanation for their results. They acknowledge the molecular bulk of the engineered adaptor that they use for the recruitment of clathrin. However, they conclude that because the adaptor alone does not cause membrane budding, crowding cannot contribute. In making this statement, the authors reveal a fundamental misunderstanding of the crowding hypothesis. In order for crowding to occur, the adaptor proteins need to be clustered together within some sort of boundary. Clathrin can potentially serve in this role. If clathrin is not present, the adaptors diffuse apart and the steric pressure returns to zero. So, the authors have not presented data to evaluate the contribution of crowding to the curvature of the structures that they observe. Instead, they should measure the stoichiometric ratio of linkers to triskelia and determine what fraction of the membrane below the coat is covered by adaptors. When this fraction exceeds 20-30%, steric pressure will rise dramatically and crowding will make a significant contribution.
-