Local regulation of extracellular vesicle traffic by the synaptic endocytic machinery

This article has been Reviewed by the following groups

Read the full article

Listed in

Log in to save this article


Neuronal extracellular vesicles (EVs) can be locally released from presynaptic terminals, carrying cargoes that are important in intercellular signaling and disease. EVs are derived from endosomes, but it remains unclear how synaptic cargoes are directed to the EV pathway, rather than undergoing conventional retrograde endosomal transport and degradation. Here, we find that the clathrin-mediated endocytic machinery plays an unexpected role in maintaining a release-competent pool of synaptic EV cargoes. Endocytic mutants, including nervous wreck ( nwk ), Shibire /Dynamin, and AP-2 , exhibit local depletion of multiple cargoes in EV donor terminals. Accordingly, nwk mutants phenocopy synaptic plasticity defects associated with loss of the EV cargo Synaptotagmin-4, and suppress lethality upon overexpression of the EV cargo Amyloid Precursor Protein. These EV defects are genetically separable from canonical functions of endocytic proteins in synaptic vesicle recycling and synaptic growth. This endocytic pathway opposes the endosomal retromer complex to regulate EV cargo levels, and acts upstream of synaptic cargo removal by retrograde axonal transport. Our data suggest a novel molecular mechanism that protects EV cargoes from local depletion at synapses.

Article activity feed

  1. Author response


    We thank the researchers within the ASAPbio community for taking the time to provide valuable feedback on our manuscript and also Iratxe Puebla for both facilitating this review of our preprint and for consolidating the comments we received. Here we provide comments to some of the points raised by the reviewers.

    In regards to the reviewers’ comment that our “work focuses on the nwk mutant”, we note that Figures 3 and 4 show the unexpected EV cargo depletion phenotype for mutants of numerous components of the clathrin-mediated endocytic machinery. We chose the nwk mutant for our in-depth analysis because it best shows separability of functions in synaptic vesicle (mild defect) vs extracellular vesicle traffic (severe defect), and also produces null mutant viable adult flies for our APP functional studies. However, our work indicates that EV cargo regulation is a broader function for the endocytic machinery and raises the possibility that previously identified neuronal phenotypes for many endocytic mutants could be due to loss of EV cargoes from synapses. Related to this, in reference to the comment that the nwk mutant “affects EV release” we also wanted to highlight that while the EV phenotype we observed for nwk and other endocytic mutants shows both pre- and postsynaptic depletion of EV cargoes, our retromer(vps35);nwk double mutant result suggests that endocytic machinery such as Nwk is not directly regulating release of EV cargoes. Instead, we conclude that the reduction of postsynaptic EV cargoes is a secondary consequence of presynaptic depletion due to defective intracellular traffic. Your helpful feedback has alerted us that we could make these points more clear in the writing and organization of our manuscript.

    In response to the points that “...the question arises as to how specific this pathway is to EVs” we should clarify that our findings seem to be specific to cargoes for which sorting to extracellular vesicles is at least a major trajectory (ie, Syt4 and hAPP, of which 30-50% of the synaptic complement is in EVs). We agree that they both have an intracellular component (either en route to EVs or for intracellular signaling functions, which have been well-documented for APP). In response to the comment that “other cargoes that undergo clathrin-dependent endocytosis and are not packaged into EVs would need to be tested”, indeed both Syt1 or Tkv require CME machinery for their traffic (PMID12795692, PMID16459302, PMID18498733), but we find that they are not detectable in EVs and are not depleted at CME machinery mutant synapses. This indicates that local synaptic depletion is specific to cargoes for which least a significant portion of their total pool is normally packaged in EVs.

    The reviewers commented that APP and perhaps Syt4 also have intracellular itineraries and functions that may be affected by their depletion at synapses - we agree that our results have implications for both extracellular vesicle and intracellular functions of these cargoes. We fully agree that “the [Figure 2] results might not be specific to EV functions of Syt4 or hAPP” and that a more general statement (such as was suggested in the comments) here would make this possibility more explicit. Our results at least indicate that reduction of these cargoes in presynaptic terminals (but not axons, cell bodies, or dendrites) is sufficient to abrogate their functions. It will be critical in the future to identify trafficking mutants that specifically disrupt EV release without impacting levels in the donor cell, in order to directly query the physiological functions of EV sorting.

    To “provide some more information on how the [postsynaptic ɑ-HRP puncta intensity] quantification was done”, we selected an intensity threshold sufficient to distinguish postsynaptic puncta from background muscle fluorescence. We did not directly select puncta manually. Puncta with brightness above this intensity threshold were measured within a 3 μm region around the neuronal HRP. Puncta brightness was not normalized to neuronal HRP brightness, but instead was normalized to the neuronal HRP volume. This analysis was not blinded as many endocytic mutants exhibit synaptic overgrowth phenotypes that are easily visible, thus complicating the blinding process. Using a complementary automated analysis for presynaptic Syt4-GFP, we found very similar results to our manual thresholding analysis. We were however unable to successfully automate the postsynaptic signal measurements due to signal-to-noise-ratio heterogeneity, especially for HRP. Here we’d also like to clarify that in regards to “postsynaptic objects smaller than 0.015 μm were excluded”, we meant postsynaptic objects smaller than 0.015 μm3.

    In response to the comment that “...saying this trafficking opposes retromer complex sorting appears to extend beyond the results” we would like to clarify that while direct opposition of endocytic machinery to retromer on endosomes is one possible interpretation, it is not the one we favor in the discussion. We agree that endocytosis and retromer are more likely to oppose each other more indirectly by regulating overall flux through the recycling pathway. We intended to convey opposition as a genetic rather than a mechanistic argument, and we think this conclusion is supported by our data. However based on this feedback we see that we could make this more clear in our manuscript.

    We thank the reviewers for pointing out that “In Figure 4B clc depletion does not yield a significant difference in pre-synaptic Syt4 levels. However, Figure 4D the levels of Syt4 are significantly lower in clc both pre- and postsynaptically”. One possibility is that this just reflects variance in the assay, and that the subtle Syt4 phenotype in the clc mutant reached our arbitrary threshold of significance in one experiment but did not in another. There is however also a potentially interesting biological explanation. The 4B clc experiment was conducted at 25℃ while the 4D experiment was conducted at 20℃, since we found that at the lower temperature we were able to recover more clc; nwk double mutant third instar larvae. Endocytosis is well-known as a temperature-dependent process, and perhaps there is some residual endocytosis at this lower temperature in the clc mutant, making it more similar to slowed endocytosis in the endocytic accessory protein mutants (see PMID16269341), compared to a more complete block in the chc or clc 25º condition. This would suggest that slow endocytosis drives cargo into the degradative pathway, fast endocytosis into the rapidly recycling and EV pathway, while no endocytosis traps cargo in unproductive membrane cisternae. Proving this would likely require more quantitative endocytosis assays than are currently available.

    We are also appreciative of reviewers comments that will help to make our manuscript more clear, such as suggestions to present the plots consistently, to mention that individual points represent individual NMJs, and to report that C155 is a neuron-specific driver, among other helpful points.

  2. This review reflects comments and contributions by Joachim Goedhart, Sónia Gomes Pereira, Karen Lange, Arthur Molines, Gregory Redpath, Zara Weinberg.

    The authors studied the release of extracellular vesicles (EV) from nerve terminals. Drosophila is used as a model and the work focuses on the nwk mutant, which affects EV release. The extracellular intensity of GFP-tagged proteins was quantified as a proxy for secretion. Overall, the study is well performed, with the correct analysis and controls, and the images are great.

    The interpretation of the data could be further clarified or toned down, especially in Figure 2, 4 and 5 and the statements around endocytic machinery opposing retromer complex sorting. The authors have performed a good job of characterising the role of Nwk in endocytosis of cargoes that can be incorporated in EVs. However, these proteins also have intracellular roles (at least for APP, not sure for Drosophila Syt4, although the human version has clear intracellular roles) meaning that the question arises as to how specific this pathway is to EVs as such. The study most definitely uncovered a protein important in SV endocytosis and recycling, which has an impact on EVs.

    Minor comments:

    • The paper would benefit from an expansion to the legends, to indicate the conditions tested. In Figures 3A&B, 4D it would be helpful to outline the abbreviations used for the graphs.
    • It is unclear that the bars in Figure 1/3/4/5 add useful information; suggest showing the data + error bars, this would be more consistent with the plots shown in Figure 2.
    • The authors may want to include this recent report in the introduction and/or discussion: Ganguly et al. (2021) Neuron.Clathrin packets move in slow axonal transport and deliver functional payloads to synapses


    'EV cargo Amyloid Precursor Protein' - It could be mentioned (perhaps in the intro) that APP also has well established intracellular roles in endocytosis, as well as being an EV cargo. If the same is true for Drosophila Syt-4 (as it is for human Syt-4), a similar reference could be made.


    ‘During our studies of the endocytic F-BAR and SH3 domain-containing protein Nervous Wreck (Nwk), we found that nwk null mutants exhibited a strong reduction in the intensity of postsynaptic α-HRP puncta (within 3 μm of the boundary of motor neurons innervating muscles 6 and 7) (Fig. 1A) - Was this a hypothesis-driven exploration? It may be worth expanding on the groundwork for this initial key observation.

    Figure 1

    • ‘Quantification of postsynaptic α-HRP puncta intensity’ - Is it possible to provide some more information on how the quantification was done. Were the puncta manually selected? Was selection blinded to condition? Is puncta brightness normalized to neuronal HRP brightness?
    • ‘Data is represented as mean +/− s.e.m.; n represents NMJs’ - n values are not stated, suggest saying "individual dots represent".
    • ‘See Tables S1 and S3 for detailed genotypes and statistical analyses’ - The authors did a great job in providing clear details of what genotypes were used in each experiment as well as exactly which analysis and parameters were considered. This is great to increase transparency and reproducibility!

    ‘These phenotypes were rescued by neuronal re-expression of Nwk using the binary GAL4/UAS system (Fig 1B)’ - While the number of Syt4 puncta appears to be fully rescued, this is less clear for the pre- or postsynaptic intensities (only a partial rescued is observed). The text reads as both phenotypes are rescued, suggest clarifying that this rescue is not similar for both phenotypes, and if possible, to speculate on the reasons for this.

    ‘cell autonomously regulates’ - This fragrant may benefit from some clarification, the results indicate Nwk alters Syt4 endocytosis, perhaps it could be stated in that manner.

    'The calcium sensor Synaptotagmin-1 (Syt1)’ - Are these cargoes internalised by comparable endocytic pathways to Syt4, i.e. clathrin-dependent endocytosis? If so, this could be stated. If not, cargoes undergoing endocytosis by a comparable pathway could be used.

    ‘reduction of Syt4, hAPP, Evi, and Nrg in nwk mutants is specific to the EV trafficking itinerary’ - Suggest replacing "is" with "may be." If an expanded range of cargoes was investigated this could change.

    ‘Together these results suggest that Nwk specifically and locally regulates the levels of EV cargo proteins at synaptic terminals’ - This fragment could be reworded to reflect that some of these cargoes (at least APP) have intracellular roles, e.g. "Together these results suggest that Nwk specifically and locally regulates the levels of cargo proteins that can be incorporated into EVs at synaptic terminals." I realise the APP expressed here is human, but in human cells it has intracellular endocytic roles, so it could also play a role when expressed in Drosophila.

    Figure 2

    • Figure 2A ‘before (top trace) and after (bottom trace)’ - This information could be added to the figure.
    • Figure 2B - Consider removing some of the labels on the x-axis, to avoid the 45 degrees rotation (to improve readability).
    • ‘Figure 2C ‘C57-GAL4 and Vglut-GAL4 are muscle and neuron-specific drivers, respectively’ - For readers unfamiliar with Drosophila drivers, it would be useful if the figure was labelled with "muscle" and "neuron" and then the legend listed the specific driver.
    • Figure 2F ‘GAL4C155-driven’ - Would it be possible to report where this driver is expressed?

    ‘Together these results show that the depletion of EV cargoes Syt4 and hAPP at nwk mutant synaptic terminals correlates with a loss of function for these EV cargos in the nervous system’ - The results might not be specific to EV functions of Syt-4 or hAPP, an alternative wording for the statement might be "Together these results show that the depletion of Syt4 and hAPP at nwk mutant synaptic terminals correlates with a loss of function for these cargos in the nervous system."

    Figure 4 ‘all measurements were further normalized to the mean of their respective controls, which is indicated with a dashed line’ - In Figure 4A, the control data on the graph are unclear. It appears that the results were normalized to Post AP-2α, would it be possible to clarify why this would be the control? A similar question regarding Figure 4B where it is not clear where the control data appear on the graph.

    ‘We found that simultaneous loss of clc and nwk phenocopied loss of clc alone with regards to levels and localization of Syt4-GFP pre- and postsynaptically (Fig. 4D)’ - In Figure 4B clc depletion does not yield a significant difference in pre-synaptic Syt4 levels. However, Figure 4D the levels of Syt4 are significantly lower in clc both pre- and postsynaptically. This could be commented on further, with an explanation offered (if possible).

    ‘This suggests that the Syt4-GFP accumulations in clc mutant synapses are not accessible to the mechanism by which EV cargoes are depleted at nwk mutant synapses’ - This could be clarified, does it mean that Nwk cannot retrieve cargo from CLC accumulations, i.e. Nwk acts downstream of clathrin?

    'Together these results highlight a novel clathrin-dependent mechanism that regulates the traffic and levels of EV cargoes at synapses' - The data suggests Nwk acts in clathrin-dependent endocytosis to regulate uptake of Syt-4 and some other EV cargoes, and when this is defective these cargoes don't get packaged into EVs. This appears to be expected, when endocytosis is defective, downstream sorting will be defective. To support the statement ‘levels of EV cargoes at synapses’, other cargoes that undergo clathrin-dependent endocytosis and are not packaged into EVs would need to be tested.

    ‘suggesting that Nwk is involved in loading EV precursors and maintaining presynaptic cargo levels’ - Recommend adding "endocytic" in relation to ‘in loading EV precursors’, based on the statement above "in endocytic loading of cargoes into the EV-permissive precursor compartment that populates MVBs destined for EV release".

    Figure 5A ‘Cartoon depicting potential EV phenotype outcomes for nwk; Vps35 experiment, testing if Nwk regulates MVB-PM release or EV cargo sorting’ - The schematic is good, consider making the label a touch clearer, e.g. "release defect", as release is not happening.


    ‘Syt4-dependent synaptic plasticity’ - As mentioned in Fig 2, the loss of synaptic plasticity is interesting in the Nwk mutant, but not directly shown to be dependent on Syt-4 here. Recommend revising to "Nwk-dependent" synaptic plasticity.

    ‘endocytic machinery opposes the retromer complex in sorting cargoes’ - Endocytosis is clearly perturbed in the Nwk mutants. This endocytosis is indeed clearly important in the synaptic region, before cargoes attach to dynein mediated retrograde transport. However, saying this trafficking opposes retromer complex sorting appears to extend beyond the results. Synaptic endocytic machinery will ensure a rapid endocytic/exocytic cycle occurs, with the occasional endosome/synaptic vesicle targeted for retrograde trafficking and degradation. This endocytic machinery is part of an overarching endocytic network that ensures cellular function. A possible explanation here is that with the Nwk mutant, the cell has aberrant EV cargo trafficking and must deal with this somehow - likely degradation. Recommend incorporating further interpretations in addition to the one stated.

    ‘endocytic machinery protects' - Endocytic machinery feeds cargo into the synaptic vesicle cycle, when this trafficking is perturbed, an alternate pathway is utilised by the cell. Alternate explanations for the results could be provided in addition to what is already suggested.

    ‘Finally, our observation that AP2’ - µ missing here.

    Materials and methods

    ‘​​Postsynaptic objects smaller than 0.015 μm were excluded’ - Could this be clarified? A size of 15 nm cannot be measured with the kind of conventional microscopy (spinning disk) used here.