Autophagy–mediated plasma membrane removal promotes the formation of epithelial syncytia
This article has been Reviewed by the following groups
Listed in
- Evaluated articles (Review Commons)
Abstract
Article activity feed
-
-
Note: This rebuttal was posted by the corresponding author to Review Commons. Content has not been altered except for formatting.
Learn more at Review Commons
Reply to the reviewers
Manuscript number: #RC-2021-00992
Corresponding author(s): Parisa Kakanj and Maria Leptin
Reviewer #1 (Evidence, reproducibility and clarity (Required)):
In this study, the authors use the fruit fly as a model to understand the role and regulation of autophagy in epidermal integrity during development and wound healing. They discover that hyper activation of autophagy via overexpression of Atg1 leads to disruption of epithelial organization, junctional protein localization, and syncytium formation. In addition, these epidermal defects were found to be dependent on TORC1 as knockdown or inhibition of TORC1 antagonists resulted in similar epidermal defects …
Note: This rebuttal was posted by the corresponding author to Review Commons. Content has not been altered except for formatting.
Learn more at Review Commons
Reply to the reviewers
Manuscript number: #RC-2021-00992
Corresponding author(s): Parisa Kakanj and Maria Leptin
Reviewer #1 (Evidence, reproducibility and clarity (Required)):
In this study, the authors use the fruit fly as a model to understand the role and regulation of autophagy in epidermal integrity during development and wound healing. They discover that hyper activation of autophagy via overexpression of Atg1 leads to disruption of epithelial organization, junctional protein localization, and syncytium formation. In addition, these epidermal defects were found to be dependent on TORC1 as knockdown or inhibition of TORC1 antagonists resulted in similar epidermal defects which could be rescued by knockdown of Atg1 or Atg5. Wound healing in fruit fly epidermis is known to induce cell fusion and here the authors show that syncytium formation is dependent on autophagy. GFP-Atg8a autophagosomes were found to accumulate in cells adjacent to the wound site, but Atg1-induced syncytium formation was dispensable for wound repair. However, the authors found that hyper activation of autophagy prior to injury slowed wound closure. This may be due to defects in actomyosin organization or another developmental defect the authors observed in the epidermis. Overall, the key conclusions of this study are convincing, but the experiments would be strengthened by validation of all the RNAi strains used as well as demonstration that epidermal barrier remains intact as described.
**Major Comments**
This study uses a number of UAS-RNAi strains as well as dominant negative and overexpression transgenes. There is no validation that these genetic perturbations work as expected.
Almost all of the lines we use have been extensively used and validated by others as documented in the literature. We append a table (below, page 14) with these references. It would be close to impossible for us to show their tissue specific efficacy in the larval epidermis because it is extremely difficult to obtain clean dissections of epidermis without contamination from other tissues (muscles, nerves, etc.), and we believe we can rely on the known validation of most of the lines. It is true that some of the lines are less well characterised, and we comment on those below, and will eliminate our speculation on their effects in the manuscript.
In fact, the authors state on pg 5 that RNAi to Atg6, Atg7, and Atg12 may be less effective, but do not verify the knockdown efficiency to the gene of interest (i.e. Atg5 RNAi knock downs Atg5 transcript or protein).
Atg12 and Atg7 have been shown (PMID: 25882046) by quantitative RT-PCR to effectively reduce RNA levels in the midgut during larval to pupal transition. We will therefore have to eliminate our speculation that the weak effect in the epidermis may be due to ineffective knock-down. Rather, it seems that these components are accessory but not necessarily essential for the completion of autophagy, as also observed by others (PMID: __25882046, __PMID: 1805642, PMID: 23599123, PMID: 15296714, PMID: 23873149, PMID: 23406899)
This is particularly important as authors use a single UAS-rictor RNAi strain to conclude that autophagy is dependent on TORC1 and not TORC2. If rictor RNAi is also weak or ineffective than this conclusion would be erroneous.
The function of rictor has been validated by classic genetics: Animals homozygous for deletions of rictor show no defects throughout their normal life cycle (Hietakangas and Cohen, 2007). We have also shown that epidermis of homozygous rictor∆1 larvae (marked with Src-GFP, DsNuc-Red2) shows no abnormalities in cell shapes or cell membranes. We include an image here.
Figure A ____| Effect of rictor deletion on the epidermis. a,b, Fluorescence micrographs of larval epidermis expressing the indicated markers in a larva homozygous for a rictor deletion (*rictorEY08986 *, also named rictor∆1). a, Lower magnification showing the entire width of larval segments A3 or A4. n=16-18 larvae each genotype. Scale bars:__ a __50 μm; b, 20 µm.
A major conclusion of this study is that autophagy remodels the lateral cell membranes and not the basal or apical, so the membrane integrity remains intact. This is described and shown in Fig S3a, but it is hard to see that the apical membrane is intact. It would be helpful if authors could show a true membrane marker, such as UAS-CD8mGFP to see if there is a continuous membrane.
We will include new experiments with this marker.
Alternatively, is there a barrier assay that could help demonstrate that syncytium formation does not disrupt epithelial integrity?
This follows from the fluorescence recovery we performed (Supplementary Video 13), where we observe rapid diffusion between areas in the epidermis, but never any leakage of fluorescence in the y-axis into the body cavity. We will emphasize this more clearly in the text.
This could be performed in the fly gut, using the smurf assay (Rera M et al. 2011), since the authors also describe (pg 9) a similar role for autophagy in disruption of epithelial lateral membranes.
We had done a smurf assay, and observed no leakage from the gut, but didn’t document this at the time because of difficulties during the period of Covid restrictions of accessing a dissecting scope/camera set up in a lab outside our own. We will try to repeat this now in the hope that with current reduced restrictions we can record the result.
Is autophagy dependent syncytium formation cell autonomous?
Our clonal analysis in wound healing addresses this point (Figure 2; text page 5 and 6). Clones of GFP-expressing cells neighbouring a wound share their cytoplasmic contents with their neighbours during wound closure. However, a clonal cell that is Atg5-deficient in a wild-type background does not share its content with the neighbouring cells. This shows that for a cell to participate in syncytium formation, that every cell itself has to be competent to perform autophagy. We will expand the explanation of this point in the text.
The A58-Gal is not cell-type specific as authors describe (pg 9) similar effects in trachea, salivary glands, and intestine and it is unclear if effects are due to disruption of autophagy in epidermal cells or general disruption in fly's physiology. The authors should determine, using a more restrictive Gal driver, whether syncytium formation is due to activation of autophagy in the epidermal cells or another cell type (trachea, salivary glands, or intestine).
We apologize if our phrasing of ‘ectodermal’ led to the impression that A58-Gal4 is cell-type specific. A58 also drives expression in the tracheal system, as all other available epidermal drivers do. A58 expression in the salivary gland is presumably due to the origin of the Gla4 construct, which like many other Gal4 drivers (e.g. NP1-Gal4) includes salivary gland specific enhancers (PMID: 8223268 and PMID: 12324947). A58 is not active in the gut, and for the experiments in the gut we used the NP1 driver. We will rephrase the text in the paper to avoid confusion. There is no driver that is absolutely restricted to the epidermis.
Alternatively, if no other Gal4 is available for the larval epidermis then authors could at least show using enterocytes driver (NP1-Gal4) that overexpression of Atg1 is sufficient to induce syncytium formation and its effect on gut barrier integrity.
We did do this experiment but didn’t include the images because of the large number of figures we already had. We now show them here. As mentioned above, barrier integrity is not compromised. We can also provide images of the phenotype in tracheal cells.
__Figure B ____| Effect____ of uncontrolled autophagy on enterocytes and salivary glands. __Larval gut or salivary glands expressing the indicated markers and overexpression (Tsc1,2 or Atg1S) or RNAi (raptori) constructs using the NP1-Gal4 driver. Images are from live imaging of gut or salivary gland of 6 to 11 larvae for each genotype. Scale bars, 20 µm.
In Fig 8, authors nicely show that Atg1 RNAi can rescue Tor RNAi and raptor RNAi, but, what about the reverse. Is overexpression of Tor sufficient to inhibit the overexpression Atg1 and reduce autophagy-induced syncytium formation?
Overexpression of Tor would affect both TORC1 and TORC2. We have done this experiment using* UAS-Torwt* construct but found that it leads to excessive autophagy rather than suppression, consistent with similar results reported by others (PMID: 12324961 and PMID: 15186745). This approach can therefore not be used to do the proposed experiment. Instead, one could use downregulation of the Tor inhibitor TSC1, which acts on TORC1, and we have shown to reduce autophagosome formation in wound healing (Fig. 1d). Another option is to overexpress the TORC1-specific activator Rheb (PMID: 12893813, PMID: 17208179 and PMID: 31422886). We will set up the experiments with these constructs in the hope that they will yield interpretable results.
**Minor comments:**
- Check spelling of abbreviations, Sqh is often misspelled Shq in figures
We will correct them. Thanks for alerting us.
The order of images in Figure 3 should match the description in the text (pg. 6).
We would prefer to retain the current order because it is then consistent with all the other figures. Re-writing the text to reflect this order would make it less clear.
AtgW is described in text, but not shown in Fig 3a-c. Also, upstream activators of TORC1 are described first, but shown last in this Figure making it difficult to follow.
We will now only mention Atg1W later in the text where we also show it in a figure.
Fig7a should show junctional effect of Atg1W alone and in combination with Atg5i which is used in 7b.
We had left this out to save space, but we will now include these data.
It is unclear why authors switched to this weak overexpression for this photobleaching assay when Atg1S was predominantly used in the rest of the study.
The reason we used Atg1W in this particular experiment is that we had it on a chromosome where it was recombined with GFP which made it genetically much easier to use for FLIP experiments. However, perhaps these constructs merit some discussion. Atg1W and Atg1S were originally called “weak” and “strong” based on studies in other tissues and other stages (PMID: 33253201). However, we found that in the epidermis their effects are practically indistinguishable, as judged by TEM (Fig.3d,e) (Fig 5e,f) (Suppl. Fig. 5a,b and Suppl. Fig. 6b,c), and all markers we used in confocal analyses (which we will include them). Thus, to avoid confusion, we will change the nomenclature we use on our paper to the neutral Atg1GS and Atg16B.
Reviewer #1 (Significance (Required)):
This study elucidates the role and regulation of TORC1 and autophagy in epithelial membrane remodeling. This is important work that is significant to both developmental and wound healing research. Many cell types become multinucleate during differentiation, aging, and wound healing and here the authors find a novel role for authophagy in remodeling lateral cellular junctions to facilitate syncytium formation.
Reviewer #2 (Evidence, reproducibility and clarity (Required)):
In their present manuscript Kakanj and colleagues show that during epithelial wound healing autophagy pathway controls plasma membrane integrity and homeostasis. Furthermore, elevated autophagic activity is sufficient to induce syncytium formation, which is essential for wound closure and healing. Authors used the epidermis of fruit fly larvae as model to study wound healing and video microscopy to examine this process. The methodology is well established, since authors already published a related study in 2016 using similar tools.
The findings presented here are interesting and promising, the quality of most experiments are satisfactory, the confocal images/videos are excellent and I truly appreciate that authors used electron microscopy to support some of their claims. Findings are well presented and the text is well written and easy to read.
Overall, my opinion is very positive about this manuscript.
I believe most of the findings are very well supported, but I have some suggestions, which may can help strengthen the authors' points.
- Authors used GFP-Atg8a reporter to follow autophagy during wound healing. While I also believe that, the appearing GFP-Atg8a dots represent autophagic vesicles after wounding but GFP-Atg8a has some certain limitations. First: Atg8a (or LC3 in mammals) is removed from the outer surface of autophagosomes by Atg4 and the Atg8a trapped inside the autophagosomes will be degraded in the autolysosomal lumen. Thus Atg8a not always localizes to autolysosomes, actually Atg8a immunostaining mostly labels autophagosomes (and phagophores) but not autolysosomes in insect cells. Accordingly, GFP-Atg8a reporter is also subject of autolysosomal degradation and furthermore most of the GFP signal is quenched in the acidic lumen of autolysosomes, since at lower pH GFP loses fluorescence. Nevertheless, if lysosomal degradation proceeds normally, GFP-Atg8 will be degraded completely. Thus, some of the autolysosomes cannot be detected using this reporter, for this mCherry-Atg8a reporters can be used, since mCherry is more resistant than GFP and thus accumulate inside lysosomes, and retains its fluorescence in acidic environments.
This is a good suggestion and we had done these experiments. However, the red fluorophores have a serious problem in that they all tend to form small aggregates or puncta – not in all tissues and at all stages, but this is a very wide-spread phenomenon, and is even observed in in vitro experiments (own observations). This makes quantification of vesicles or other small structures such as autophagosomes complete impossible. Nevertheless, here are a few figures from our analyses. While some of the spots clearly appear to be autophagosomes, as judged by their positions, they cannot be objectively distinguished from the other spots.
Figure C ____| Autophagy during epidermal wound healing. Time-lapse series of single-cell wound healing in larva expressing mCherry-Atg8a (black) to mark autophagosomes and autolysosomes (A58>mCherry-Atg8a). a, z-projections of a time-lapse series. b, Higher magnification of the areas marked by magenta boxes in (a). n=11 larvae. Each frame is a merge of 57 planes spaced 0.28 μm apart. Scale bars:__ a __20 μm; b, 10 µm.
However, I still believe that for video microscopy GFP-Atg8a was a perfect choice, I just suggest to confirm the appearance of autophagosomes after wounding by other means: for instance, immunostaining of the epidermis after wounding (120 min) against Atg8a should confirm the presence of autophagosomes. There are a few specific available antibodies working in flies which are listed in the reviews of Nagy (PMID: 25481477) or more recently in Lorincz (PMID: 28704946)
This is technically a huge challenge. We would have to induce a single cell wound, then filet and fix the epidermis, during which it rolls up and often destroys the area of interest. If it doesn’t, then the prep can be flattened out, but it still can be very difficult to find the wound in the prep.
- One of the major claims of the authors is that elevated autophagy leads to the breakdown or removal of lateral plasma membranes to promote syncytium formation. It is clearly seen on the confocal or EM images that lateral membranes disappear after wounding. However, it is also suggested that the lateral plasma membrane material is incorporated into autophagosomes or plasma membrane is a potential membrane source of autophagosome formation. I believe this is the least supported claim of the manuscript since no direct evidence for this is presented. This is based on BodyPy staining only, that BodyPy positive vesicles accumulate inside the cells. If this is indeed the case plasma membrane components should be detected in autophagic vesicles. Thus, I recommend co-staining membrane components with autophagic markers.
This is indeed the clear next step, and we did a number of experiments along those lines, but they were once again compromised by the problem with the mCherry aggregates. This made the interpretation in the unwounded epidermis with artificially upregulated autophagy impossible. However, experiments with naturally upregulated autophagy in wound healing yielded results that are consistent with plasma membrane components being associated with autophagosomes (with the caveat that not every red dot we see represents an autophagosome). We have just repeated some of these using the septate junction marker FasIII and have obtained some beautiful movies that show FasIII labelled membrane (green) being surrounded by mCherry spots, and as the membrane begins to dissociate, the mCherry spots turn from red to yellow. We have included stills from results of these analyses here and will include them in a new figure in the revised manuscript.
__Figure D ____| Colocalization of Atg8a and the septate junction component FasIII during epidermal wound healing. a, Time-lapse series of single-cell wound healing in a larva expressing mCherry-Atg8a (red) (A58>mCherry-Atg8a) and endogenously tagged FasIII (GFP gene trap; green), a transmembrane component of septate junctions. b, Higher magnification of the time-lapse marked by magenta boxes in (a). n=11 larvae. a,b, Each frame is a merge of 68 planes spaced 0.28 μm apart. Scale bars: a,b __20 μm.
However if authors observe no colocalization of plasma membrane components with autophagy markers I still believe this study worth to be published. I would like to recommend the review of Ungermann and Reggiori (PMID: 29966469) in which the trafficking of Atg9 is discussed,
Yes, indeed. And there is in fact now a further paper that goes in a similar direction (PMID: 34257406). We had left this out because we did not have direct data on Atg9, but will be happy to include it in the discussion in which we cite the paper that shows that Drosophila Atg9 is localized on the lateral plasma membrane in nurse cells, and loss of it leads to syncytium formation.
since the source of autophagosomal Atg9 is in part the plasma membrane in mammalian cells. Therefore, these findings may strengthen the authors' claims.
**Minor points:**
Figure 2A: I believe authors wanted to use the word 'maintaining' not mating in their scheme.
Indeed. Thanks for alerting us.
Discussion: Authors suggest that: another function of autophagy in the cells surrounding the wound may be to clear up debris as in planarian and other cell types autophagy is activated in healthy cells, which simultaneously phagocytose cell debris. Honestly, I do not believe that this is the case here. Some of the Atg proteins are indeed required for phagocytosis during LC3-assiciated phagocytosis (LAP) (see: PMID: 30787029), but LAP is independent form Atg1
Good point, we will include this in the discussion.
and if LAP happened in the cells, surrounding the wound then GFP-Atg8a positive phagosomes would appear in those cells. However, it is clearly not the case here.
Reviewer #2 (Significance (Required)):
I highly recommend this manuscript to be uploaded to a relevant journal and I believe the findings presented here will be interesting for biologists specialized in regeneration and readers from the autophagy fields alike.
Reviewer #3 (Evidence, reproducibility and clarity (Required)):
**Summary:**
The larval epidermis of Drosophila is a prime model for studying wound healing by combining live imaging with cellular, genetic and molecular analysis of the processes involved. Autophagy is known to be activated and necessary for efficient wound healing in animal models through secretion of cytokines and clearance of bacteria. This manuscript implicates autophagy in cellular syncytium formation during wound healing. Live imaging demonstrates autophagy activation in cells surrounding the wound. Inhibition of autophagy by RNAi against atg1 or atg5, required for autophagy initiation and autophagosome formation had no effect on the rate of constriction and closing of the wound site. However, elegant live imaging demonstrates that autophagy is required cell autonomously for cell fusion, a normal process during wound healing in flies. Autophagy can also be instructive for cell fusion. Strong induction of autophagy by TORC1 inhibition, TSC1/2 overexpression or Atg1 overexpression induce cell fusion that is genetically dependent on atg5, a gene acting downstream of atg1 in autophagosome formation. As Chloroquine treatment, a chemical inhibiting autophagosome fusion to the lysosome and lysosomal breakdown showed no effect, the authors suggest that later steps of autophagy are not involved. Live imaging with a selection of cellular fluorescently tagged markers of apical, lateral and basolateral membrane domains, combined with electron microscopy show clearly that lateral membrane are disrupted and removed within the epithelium. During this process, membranous large vesicles "drift" away from the plasma membrane. If these vesicles relate to autophagy is not addressed. In addition to the effect on cell fusion, strong autophagy induction also leads to autophagy within the nucleus, chromatin condensation and distortion of the nuclear membrane. The manuscript is well written and easy to follow. Figure panels and data are clearly presented. All experiments are well described throughout and skillfully executed with appropriate controls and statistical analysis. It remains unknown what induces autophagy in response to wounding. It also remains unclear whether autophagy deconstructs or engulfs parts of the plasma membrane, or if parts of the autophagy machinery has additional roles in plasma membrane fusion.
**Major comments:**
Are the key conclusions convincing? -Conclusions are generally balanced and convincing.
-I have seldom seen a paper so well written, presented and balanced by first pass. Hence my experimental suggestions are few.
Should the authors qualify some of their claims as preliminary or speculative, or remove them altogether? -Claims are well founded.
Would additional experiments be essential to support the claims of the paper? Request additional experiments only where necessary to evaluate the paper as it is, and do not ask authors to open new lines of experimentation.
-The inhibition of autophagy is performed using knockdown of two genes acting in autophagy initiation (atg1, a part of the ULK1 kinase complex) and atg5, required for autophagosome formation. Later acting genes in the autophagy process such as autophagosome closure, fusion with the lysosome or degradation were not analyzed. In the abstract, the authors state "Proper functioning of TORC1 is needed to prevent autophagy from destroying the larval epidermis which depends on membrane isolation and phagophore expansion, but not fusion of autophagosomes to lysosomes". As far as I can see, the last statement on fusion derives from experiments with Chloroquine. Although frequently used for qualitative experiments, CQ is not suited for conclusive experiments. Without genetic experiments targeting genes for autophagosome-lysosome fusion such as snap29,stx17,vamp7 this statement is in my mind not well supported.
We agree this would strengthen our findings, and we had indeed ordered these strains from the Bloomington stock collection. However, they were dead on arrival and both our labs in Heidelberg and Cologne currently have major problems with shipments from Bloomington and German customs. Other colleagues whom we asked did not have them available either. We will continue to search for appropriate constructs, but even if we find them and they arrive alive, and are processed by customs within a reasonable time, it will take many weeks to establish and then expand them and subsequently do the multi-generation crosses to obtain the stocks with all the relevant drivers and markers to set up the experiment. Three months is the absolute lower limit provided everything works according to plan, and first time round 6 months is a more realistic assumption. We hope that the referees and the editors agree that while this is a desirable experiment, it is not essential for the publication of the other results we present.
- Are the suggested experiments realistic for the authors? It would help if you could add an estimated cost and time investment for substantial experiments. -Given the expertise of the authors, these experiments should be easy to perform within 3 months.
• Are the data and the methods presented in such a way that they can be reproduced?
- The manuscript is well written and an excellent example of how how methods and experiments should be presented. Methods, tools and experiments are all well described.
• Are the experiments adequately replicated and statistical analysis adequate? -Replicates and statistics are adequate and custom for the type of analysis performed.
**Minor comments:**
Specific experimental issues that are easily addressable. Figure 3 h. The live imaging documents the striking disappearance of lateral cell membranes using SRC-GFP. In 3h, large vesicle formation and movement towards the cell interior is shown. How frequent is this?
This can only be seen clearly in experiments with time-controlled (Gal80ts) induction of authophagy where we can observe the process unfolding. We see these structures very frequently, but great variability in morphology and the structures are not always captured clearly in the plane of imaging. We here provide further examples.
Figure E ____| Autophagy in unwounded epidermis. a-c, Three additional examples showing apparent extrusions from lateral membranes after induction of autophagy (same experiment asn Figure 3h). Time-lapse series of epidermal cells expressing Src-GFP and Atg1S. Transgene expression is induced at the end of the second larval instar, live imaging started 6 h later (t=0) and continued for an additional 6 hours. a-c, Src-GFP containing material appears to be taken out of and eventually detached from lateral cell membranes (arrows).
Is this believed to be the mechanism of lateral membrane removal?
We would of course like to believe that, but we have no proof, and would therefore only be able to speculate.
If so, is it dependent on the autophagy machinery. Are these vesicle positive for autophagy markers?
Some autophagy markers have indeed been reported to be associated with the plasma membrane (e.g. Atg9, Atg16), but a conclusive study, while highly desirable, in our view goes beyond the scope of this study.
Resolving this issue may lift the conclusions of the paper. Using 3xCherry-Atg8 together with SRC-GFP, this should be possible.
We are intrigued by this suggestion and will be setting up the necessary crosses to do the experiments. However, it will take a long time to generate the necessary stocks (see genetics described below), and we will then again encounter the problem with the mCherry aggregates (see response to referees # 2). We are curious about the outcome, but we do not think it will be reasonable to promise as part of this revision that we will be able to provide conclusive results in the foreseeable future. Along with the many other things to do, this may just have to become part of a future paper, especially if there turn out to be other problems to be solved along the way. Like, for example, having to make an infrared (like mIFP or mKate, with which we have had much better experience in other contexts) Atg8 construct.
Using CQ, the authors should be able to detect plasma membrane and junctional components in autophagosomes or autolysosomes (by confocal and electron microscopy) as degradation is inhibited. This should help to distinguish whether lateral membranes are engulfed and digested or if cells simply fuse, by using a part of the autophagy machiney.
We have many interesting EM images on which we have had extensive discussions with the Paolo Ronchi and Yannick Schwab at the EMBL (whom we embarrassingly forgot to acknowledge in our manuscript, which will now be corrected), and one of the authors of this paper (BM) is an expert in EM imaging of the larval epidermis. It was agreed that some structures could indeed be interpreted as autophagosomes with content resembling junctional material. However, in the absence of absolute proof, we did not include them in the paper. We now show them here.
Figure F ____| ____Autophagosomes with junctional material ____in unwounded epidermis. Transmission electron micrographs of sections through the epidermis of a larva with elevated autophagy (A58>Atg1S) at two different magnifications. Arrows mark the autophagosomal membrane with content resembling junctional material.
The authors, state that strong autophagy activation also leads to syncytium formation of tracheal cells, salivary glands and gut EC cells. Representative images in a supplementary figure would be useful for future reference.
See response to other comments above (response to referees # 1). We have added some images in this document (Figure B) and will be happy to add additional ones in the revised manuscript.
Are prior studies referenced appropriately? -Yes. Key literature and findings are cited and discussed.
- Are the text and figures clear and accurate? -Yes
• Do you have suggestions that would help the authors improve the presentation of their data and conclusions?
-See suggested experiments above.
Reviewer #3 (Significance (Required)):
Describe the nature and significance of the advance (e.g. conceptual, technical, clinical) for the field. -The findings clearly documents a role of autophagy in syncytium formation in the physiological process of wounding. This has parallels to muscle syncytium formation, but has to my knowledge not been demonstrated in any other cell type to be performed by autophagy. Moreover, the authors show that strong autophagy induction can lead to fusion of epithelial cells. This may have relevance for processes and diseases where polyploidy are observed.
Place the work in the context of the existing literature (provide references, where appropriate).
State what audience might be interested in and influenced by the reported findings. -The data are very strong and the demonstration that autophagy controls syncytium formation outside of muscle development is surprising and significant. It is of interest to the field of cell biology and development in general and the autophagy field in particular. It will also be of interest for the medical field that deals with multinuclear phenotypes, such as cancer.
Define your field of expertise with a few keywords to help the authors contextualize your point of view. Indicate if there are any parts of the paper that you do not have sufficient expertise to evaluate. -Development, cell signaling, autophagy, vesicle trafficking.
Table 2 | Fly stocks used in experiments
Transgenes
Stock ID
__Source __
__Publications using this construct __
Reference
*UAS-GFP-Kuk *
(UAS-GFP-KukEY07696(w+))
Jörg Großhans
PMID: 16421189
29
UAS-Atg1i
(UAS-Atg1RNAi)
V # 16133
(GD7149)
PMID: __19363474 __
PMID: __31995752 __
PMID: __32032548 __
PMID: __32915229 __
UAS-Atg5i
(UAS-Atg5RNAi)
V # 104461
(KK108904)
PMID: __31995752 __
PMID: 32032548
UAS-Atg6i
(UAS-Atg6RNAi)
V # 110197
(KK102460)
PMID: 28581519
PMID: 23599123
PMID: __27542914 __
PMID: 25644700
Dissertation of Philipp Trachte, Abb. 23. https://refubium.fu-berlin.de/handle/fub188/27709
Dissertation of Sirena Soriano Rodríguez. https://roderic.uv.es/bitstream/handle/10550/50749/Tesis%20SSoriano.pdf?sequence=1
UAS-Atg7i
(UAS-Atg7RNAi)
V # 45558
(GD11671)
PMID: 25882046
PMID: __31995752 __
PMID: __32032548 __
PMID: 23599123
UAS-Atg12i
(UAS-Atg12RNAi)
V # 29791
(GD15230)
PMID: 25882046
PMID: __17568747 __
PMID: __31995752 __
*UAS-TSC1,2 *
(UAS-TSC1, AUS-TSC2)
Iswar K. Hariharan
PMID: 15296714
PMID: 11348592
64
UAS-TSC1i
(UAS-TSC1RNAi)
V # 22252
(GD11836)
PMID: 23144631
PMID: 29144896
PMID: 29456138
*UAS-Tori *
(UAS-TorRNAi)
BL # 33951
Nobert Perrimon
PMID: 25882046
PMID: 26395483
65
*UAS-TORDN *
(UAS-TORTED)
BL # 7013
Thomas P. Neufeld
PMID: 15296714
PMID: 29144896
66
*UAS-raptori *
(UAS-raptorRNAi)
BL # 34814
Nobert Perrimon
PMID: 25882046
PMID: 31048465
65
*UAS-raptori-2 *
(UAS-raptorRNAi)
BL # 41912
Nobert Perrimon
PMID: 32097403
65
UAS-rictori
(UAS-rictorRNAi)
BL # 36699
Nobert Perrimon
PMID: 25882046
65
UAS-Atg1S
(UAS-Atg16B)
Thomas P. Neufeld
PMID: 33253201
67
UAS-Atg1W, UAS-GFP
(UAS-Atg1GS10797)
Thomas P. Neufeld
PMID: 33253201
67
UAS-S6Ki
(UAS-S6KRNAi)
BL # 41895
Nobert Perrimon
PMID: 25284370
65
*UAS-SqaKA *
(UAS-SqaT279A/CyO)
Guang-Chao Chen
PMID: 21169990
30
UAS-RhoAi
(UAS-RhoARNAi)
V # 12734
(GD4726)
PMID: 23853710
PMID: 33789114
UAS-Roki
(UAS-RokRNAi)
V # 104675
(KK107802)
PMID: 24995985
PMID: 33789114
UAS-RhebAV4
BL # 9690
Fuyuhiko Tamanoi
PMID: 31909714
PMID: 28829944
69
-
Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.
Learn more at Review Commons
Referee #3
Evidence, reproducibility and clarity
Summary:
The larval epidermis of Drosophila is a prime model for studying wound healing by combining live imaging with cellular, genetic and molecular analysis of the processes involved. Autophagy is known to be activated and necessary for efficient wound healing in animal models through secretion of cytokines and clearance of bacteria. This manuscript implicates autophagy in cellular syncytium formation during wound healing. Live imaging demonstrates autophagy activation in cells surrounding the wound. Inhibition of autophagy by RNAi against atg1 or atg5, required for autophagy initiation and autophagosome formation had no effect …
Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.
Learn more at Review Commons
Referee #3
Evidence, reproducibility and clarity
Summary:
The larval epidermis of Drosophila is a prime model for studying wound healing by combining live imaging with cellular, genetic and molecular analysis of the processes involved. Autophagy is known to be activated and necessary for efficient wound healing in animal models through secretion of cytokines and clearance of bacteria. This manuscript implicates autophagy in cellular syncytium formation during wound healing. Live imaging demonstrates autophagy activation in cells surrounding the wound. Inhibition of autophagy by RNAi against atg1 or atg5, required for autophagy initiation and autophagosome formation had no effect on the rate of constriction and closing of the wound site. However, elegant live imaging demonstrates that autophagy is required cell autonomously for cell fusion, a normal process during wound healing in flies. Autophagy can also be instructive for cell fusion. Strong induction of autophagy by TORC1 inhibition, TSC1/2 overexpression or Atg1 overexpression induce cell fusion that is genetically dependent on atg5, a gene acting downstream of atg1 in autophagosome formation. As Chloroquine treatment, a chemical inhibiting autophagosome fusion to the lysosome and lysosomal breakdown showed no effect, the authors suggest that later steps of autophagy are not involved. Live imaging with a selection of cellular fluorescently tagged markers of apical, lateral and basolateral membrane domains, combined with electron microscopy show clearly that lateral membrane are disrupted and removed within the epithelium. During this process, membranous large vesicles "drift" away from the plasma membrane. If these vesicles relate to autophagy is not addressed. In addition to the effect on cell fusion, strong autophagy induction also leads to autophagy within the nucleus, chromatin condensation and distortion of the nuclear membrane. The manuscript is well written and easy to follow. Figure panels and data are clearly presented. All experiments are well described throughout and skillfully executed with appropriate controls and statistical analysis. It remains unknown what induces autophagy in response to wounding. It also remains unclear whether autophagy deconstructs or engulfs parts of the plasma membrane, or if parts of the autophagy machinery has additional roles in plasma membrane fusion.
Major comments:
• Are the key conclusions convincing? -Conclusions are generally balanced and convincing. -I have seldom seen a paper so well written, presented and balanced by first pass. Hence my experimental suggestions are few.
• Should the authors qualify some of their claims as preliminary or speculative, or remove them altogether? -Claims are well founded,
• Would additional experiments be essential to support the claims of the paper? Request additional experiments only where necessary to evaluate the paper as it is, and do not ask authors to open new lines of experimentation.
-The inhibition of autophagy is performed using knockdown of two genes acting in autophagy initiation (atg1, a part of the ULK1 kinase complex) and atg5, required for autophagosome formation. Later acting genes in the autophagy process such as autophagosome closure, fusion with the lysosome or degradation were not analyzed. In the abstract, the authors state "Proper functioning of TORC1 is needed to prevent autophagy from destroying the larval epidermis which depends on membrane isolation and phagophore expansion, but not fusion of autophagosomes to lysosomes". As far as I can see, the last statement on fusion derives from experiments with Chloroquine. Although frequently used for qualitative experiments, CQ is not suited for conclusive experiments. Without genetic experiments targeting genes for autophagosome-lysosome fusion such as snap29,stx17,vamp7 this statement is in my mind not well supported.
• Are the suggested experiments realistic for the authors? It would help if you could add an estimated cost and time investment for substantial experiments. -Given the expertise of the authors, these experiments should be easy to perform within 3 months.
• Are the data and the methods presented in such a way that they can be reproduced?
- The manuscript is well written and an excellent example of how how methods and experiments should be presented. Methods, tools and experiments are all well described.
• Are the experiments adequately replicated and statistical analysis adequate? -Replicates and statistics are adequate and custom for the type of analysis performed.
Minor comments:
• Specific experimental issues that are easily addressable. Figure 3 h. The live imaging documents the striking disappearance of lateral cell membranes using SRC-GFP. In 3h, large vesicle formation and movement towards the cell interior is shown. How frequent is this? Is this believed to be the mechanism of lateral membrane removal? If so, is it dependent on the autophagy machinery. Are these vesicle positive for autophagy markers? Resolving this issue may lift the conclusions of the paper. Using 3xCherry-Atg8 together with SRC-GFP, this should be possible.
Using CQ, the authors should be able to detect plasma membrane and junctional components in autophagosomes or autolysosomes (by confocal and electron microscopy) as degradation is inhibited. This should help to distinguish whether lateral membranes are engulfed and digested or if cells simply fuse, by using a part of the autophagy machiney.
The authors, state that strong autophagy activation also leads to syncytium formation of tracheal cells, salivary glands and gut EC cells. Representative images in a supplementary figure would be useful for future reference.
• Are prior studies referenced appropriately?
-Yes. Key literature and findings are cited and discussed.
• Are the text and figures clear and accurate?
-Yes
• Do you have suggestions that would help the authors improve the presentation of their data and conclusions?
-See suggested experiments above.
Significance
• Describe the nature and significance of the advance (e.g. conceptual, technical, clinical) for the field.
-The findings clearly documents a role of autophagy in syncytium formation in the physiological process of wounding. This has parallels to muscle syncytium formation, but has to my knowledge not been demonstrated in any other cell type to be performed by autophagy. Moreover, the authors show that strong autophagy induction can lead to fusion of epithelial cells. This may have relevance for processes and diseases where polyploidy are observed.
• Place the work in the context of the existing literature (provide references, where appropriate).
• State what audience might be interested in and influenced by the reported findings. -The data are very strong and the demonstration that autophagy controls syncytium formation outside of muscle development is surprising and significant. It is of interest to the field of cell biology and development in general and the autophagy field in particular. It will also be of interest for the medical field that deals with multinuclear phenotypes, such as cancer.
• Define your field of expertise with a few keywords to help the authors contextualize your point of view. Indicate if there are any parts of the paper that you do not have sufficient expertise to evaluate.
-Development, cell signaling, autophagy, vesicle trafficking.
-
Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.
Learn more at Review Commons
Referee #2
Evidence, reproducibility and clarity
In their present manuscript Kakanj and colleagues show that during epithelial wound healing autophagy pathway controls plasma membrane integrity and homeostasis. Furthermore, elevated autophagic activity is sufficient to induce syncytium formation, which is essential for wound closure and healing. Authors used the epidermis of fruit fly larvae as model to study wound healing and video microscopy to examine this process. The methodology is well established, since authors already published a related study in 2016 using similar tools.
The findings presented here are interesting and promising, the quality of most experiments are …
Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.
Learn more at Review Commons
Referee #2
Evidence, reproducibility and clarity
In their present manuscript Kakanj and colleagues show that during epithelial wound healing autophagy pathway controls plasma membrane integrity and homeostasis. Furthermore, elevated autophagic activity is sufficient to induce syncytium formation, which is essential for wound closure and healing. Authors used the epidermis of fruit fly larvae as model to study wound healing and video microscopy to examine this process. The methodology is well established, since authors already published a related study in 2016 using similar tools.
The findings presented here are interesting and promising, the quality of most experiments are satisfactory, the confocal images/videos are excellent and I truly appreciate that authors used electron microscopy to support some of their claims. Findings are well presented and the text is well written and easy to read.
Overall, my opinion is very positive about this manuscript.
I believe most of the findings are very well supported, but I have some suggestions, which may can help strengthen the authors' points.
Authors used GFP-Atg8a reporter to follow autophagy during wound healing. While I also believe that, the appearing GFP-Atg8a dots represent autophagic vesicles after wounding but GFP-Atg8a has some certain limitations. First: Atg8a (or LC3 in mammals) is removed from the outer surface of autophagosomes by Atg4 and the Atg8a trapped inside the autophagosomes will be degraded in the autolysosomal lumen. Thus Atg8a not always localizes to autolysosomes, actually Atg8a immunostaining mostly labels autophagosomes (and phagophores) but not autolysosomes in insect cells. Accordingly, GFP-Atg8a reporter is also subject of autolysosomal degradation and furthermore most of the GFP signal is quenched in the acidic lumen of autolysosomes, since at lower pH GFP loses fluorescence. Nevertheless, if lysosomal degradation proceeds normally, GFP-Atg8 will be degraded completely. Thus, some of the autolysosomes cannot be detected using this reporter, for this mCherry-Atg8a reporters can be used, since mCherry is more resistant than GFP and thus accumulate inside lysosomes, and retains its fluorescence in acidic environments. However, I still believe that for video microscopy GFP-Atg8a was a perfect choice, I just suggest to confirm the appearance of autophagosomes after wounding by other means: for instance, immunostaining of the epidermis after wounding (120 min) against Atg8a should confirm the presence of autophagosomes. There are a few specific available antibodies working in flies which are listed in the reviews of Nagy (PMID: 25481477) or more recently in Lorincz (PMID: 28704946)
One of the major claims of the authors is that elevated autophagy leads to the breakdown or removal of lateral plasma membranes to promote syncytium formation. It is clearly seen on the confocal or EM images that lateral membranes disappear after wounding. However, it is also suggested that the lateral plasma membrane material is incorporated into autophagosomes or plasma membrane is a potential membrane source of autophagosome formation. I believe this is the least supported claim of the manuscript since no direct evidence for this is presented. This is based on BodyPy staining only, that BodyPy positive vesicles accumulate inside the cells. If this is indeed the case plasma membrane components should be detected in autophagic vesicles. Thus, I recommend co-staining membrane components with autophagic markers. However if authors observe no colocalization of plasma membrane components with autophagy markers I still believe this study worth to be published. I would like to recommend the review of Ungermann and Reggiori (PMID: 29966469) in which the trafficking of Atg9 is discussed, since the source of autophagosomal Atg9 is in part the plasma membrane in mammalian cells. Therefore, these findings may strengthen the authors' claims.
Minor points:
Figure 2A: I believe authors wanted to use the word 'maintaining' not mating in their scheme. Discussion: Authors suggest that: another function of autophagy in the cells surrounding the wound may be to clear up debris as in planarian and other cell types autophagy is activated in healthy cells, which simultaneously phagocytose cell debris. Honestly, I do not believe that this is the case here. Some of the Atg proteins are indeed required for phagocytosis during LC3-assiciated phagocytosis (LAP) (see: PMID: 30787029), but LAP is independent form Atg1 and if LAP happened in the cells, surrounding the wound then GFP-Atg8a positive phagosomes would appear in those cells. However, it is clearly not the case here.
Significance
I highly recommend this manuscript to be uploaded to a relevant journal and I believe the findings presented here will be interesting for biologists specialized in regeneration and readers from the autophagy fields alike.
-
Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.
Learn more at Review Commons
Referee #1
Evidence, reproducibility and clarity
In this study, the authors use the fruit fly as a model to understand the role and regulation of autophagy in epidermal integrity during development and wound healing. They discover that hyper activation of autophagy via overexpression of Atg1 leads to disruption of epithelial organization, junctional protein localization, and syncytium formation. In addition, these epidermal defects were found to be dependent on TORC1 as knockdown or inhibition of TORC1 antagonists resulted in similar epidermal defects which could be rescued by knockdown of Atg1 or Atg5. Wound healing in fruit fly epidermis is known to induce cell fusion and …
Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.
Learn more at Review Commons
Referee #1
Evidence, reproducibility and clarity
In this study, the authors use the fruit fly as a model to understand the role and regulation of autophagy in epidermal integrity during development and wound healing. They discover that hyper activation of autophagy via overexpression of Atg1 leads to disruption of epithelial organization, junctional protein localization, and syncytium formation. In addition, these epidermal defects were found to be dependent on TORC1 as knockdown or inhibition of TORC1 antagonists resulted in similar epidermal defects which could be rescued by knockdown of Atg1 or Atg5. Wound healing in fruit fly epidermis is known to induce cell fusion and here the authors show that syncytium formation is dependent on autophagy. GFP-Atg8a autophagosomes were found to accumulate in cells adjacent to the wound site, but Atg1-induced syncytium formation was dispensable for wound repair. However, the authors found that hyper activation of autophagy prior to injury slowed wound closure. This may be due to defects in actomyosin organization or another developmental defect the authors observed in the epidermis. Overall, the key conclusions of this study are convincing, but the experiments would be strengthened by validation of all the RNAi strains used as well as demonstration that epidermal barrier remains intact as described.
Major Comments
- This study uses a number of UAS-RNAi strains as well as dominant negative and overexpression transgenes. There is no validation that these genetic perturbations work as expected. In fact, the authors state on pg 5 that RNAi to Atg6, Atg7, and Atg12 may be less effective, but do not verify the knockdown efficiency to the gene of interest (i.e. Atg5 RNAi knock downs Atg5 transcript or protein). This is particularly important as authors use a single UAS-rictor RNAi strain to conclude that autophagy is dependent on TORC1 and not TORC2. If rictor RNAi is also weak or ineffective than this conclusion would be erroneous.
- A major conclusion of this study is that autophagy remodels the lateral cell membranes and not the basal or apical, so the membrane integrity remains intact. This is described and shown in Fig S3a, but it is hard to see that the apical membrane is intact. It would be helpful if authors could show a true membrane marker, such as UAS-CD8mGFP to see if there is a continuous membrane. Alternatively, is there a barrier assay that could help demonstrate that syncytium formation does not disrupt epithelial integrity? This could be performed in the fly gut, using the smurf assay (Rera M et al. 2011), since the authors also describe (pg 9) a similar role for autophagy in disruption of epithelial lateral membranes.
- Is autophagy dependent syncytium formation cell autonomous? The A58-Gal is not cell-type specific as authors describe (pg 9) similar effects in trachea, salivary glands, and intestine and it is unclear if effects are due to disruption of autophagy in epidermal cells or general disruption in fly's physiology. The authors should determine, using a more restrictive Gal driver, whether syncytium formation is due to activation of autophagy in the epidermal cells or another cell type (trachea, salivary glands, or intestine). Alternatively, if no other Gal4 is available for the larval epidermis then authors could at least show using enterocytes driver (NP1-Gal4) that overexpression of Atg1 is sufficient to induce syncytium formation and its effect on gut barrier integrity.
- In Fig 8, authors nicely show that Atg1 RNAi can rescue Tor RNAi and raptor RNAi, but, what about the reverse. Is overexpression of Tor sufficient to inhibit the overexpression Atg1 and reduce autophagy-induced syncytium formation?
Minor comments:
- Check spelling of abbreviations, Sqh is often misspelled Shq in figures
- The order of images in Figure 3 should match the description in the text (pg. 6).
AtgW is described in text, but not shown in Fig 3a-c. Also, upstream activators of TORC1 are described first, but shown last in this Figure making it difficult to follow. - Fig7a should show junctional effect of Atg1W alone and in combination with Atg5i which is used in 7b. It is unclear why authors switched to this weak overexpression for this photobleaching assay when Atg1S was predominantly used in the rest of the study.
Significance
This study elucidates the role and regulation of TORC1 and autophagy in epithelial membrane remodeling. This is important work that is significant to both developmental and wound healing research. Many cell types become multinucleate during differentiation, aging, and wound healing and here the authors find a novel role for authophagy in remodeling lateral cellular junctions to facilitate syncytium formation.
-