FAM134B regulates ER-to-lysosome-associated degradation of misfolded proteins upon pharmacologic or genetic inactivation of ER-associated degradation

  1. Elisa Fasana
  2. Ilaria Fregno
  3. Carmela Galli
  4. Maurizio Molinari

Reviewed by

Read the full article

Listed in

Log in to save this article

Abstract

About 40% of the eukaryotic cell’s proteome is synthesized and assembled in the endoplasmic reticulum (ER). Native proteins are transported to their intra- or extra-cellular site of activity. Folding-defective polypeptides are dislocated across the ER membrane into the cytoplasm, poly-ubiquitylated and degraded by 26S proteasomes (ER-associated degradation, ERAD). Large misfolded proteins like mutant forms of collagen or aggregation-prone mutant forms of alpha1 antitrypsin cannot be dislocated across the ER membrane for ERAD. Rather, they are segregated in ER subdomains that vesiculate and deliver their cargo to endolysosomal compartments for clearance by ER-to-lysosome-associated degradation (ERLAD). Here, we show the lysosomal delivery of a canonical ERAD substrate upon pharmacologic and genetic inhibition of the ERAD pathways. This highlights the surrogate intervention of ERLAD to remove defective gene products upon dysfunctional ERAD.

Article activity feed

  1. Review Commons

    Note: This response was posted by the corresponding author to Review Commons. The content has not been altered except for formatting.

    Learn more at Review Commons

    Reply to the reviewers

    Manuscript number: #RC-2023-02281

    Corresponding author(s): Maurizio Molinari

    Point-by-point description of the revisions

    This section is mandatory. *Please insert a point-by-point reply describing the revisions that were already carried out and included in the transferred manuscript. *

    Reviewer #1 (Evidence, reproducibility and clarity (Required)):

    In this manuscript from Fasana et al., the authors present data that investigates potential compensatory degradation pathways for misfolded glycoproteins in the ER - postulating that the ER-to-lysosome associated degradation (ERLAD) pathway becomes employed in the absence of a path for substrates to reach the ER-associated degradation (ERAD) mechanism. Using the classic ERAD substrate alpha1-antitrypsin NHK variant (NHK), the authors first demonstrate that pharmacologically preventing access of NHK to ERAD either with KIF (early) or PS-341 (late) elevates the number of LAMP-1 positive endolysosomes also immunoreactive for NHK (via HA), similar to what is observed for the ATZ variant that forms polymers in the ER (Fig 2). The authors next use shRNAs that silence essential ERAD factors (EDEM1, OS-9) involved in glycan recognition to demonstrate comparable enrichment of NHK in endolysosomes through genetic disruption (Fig 3). Next, the authors employ FAM134B-deficient MEFs to demonstrate the requirement for this ER-phagy receptor when ERAD is unavailable (Fig 4). Reconstituting FAM134B-/- MEFs treated with KIF/PS-341 + Baf, with a full length FAM134B rescue plasmid restores endolysosomal accumulation of NHK while a FAM134B-∆LIR does not, providing supporting evidence for substrate rerouting to ERLAD. Finally, the authors use knockouts of Atg7 and Atg13 to demonstrate dependence on LC3 lipidation and independence from macro-ERphagy (Fig 6), that points towards a pathway that is like that used to remove ATZ polymers. From these data, the authors conclude that ERLAD is increasingly engaged for substrate degradation when ERAD is impaired.

    MAJOR COMMENTS

    1. All assays rely on quantification of the NHK-HA substrates by microscopy. Would it be possible for the authors to also include biochemical analysis of NHK - potentially including data assessing its changing abundance and glycosylation state?

    To consider this, and other comments, the new submission includes biochemical data (pulse-chase analyses) on NHK (new panels A-D in Fig. 2) and on BACE457delta, an additional ERAD substrate (new Fig. 6). Please also refer to Comment 3.

    In Figure 3D, the knockdown of OS-9.1/2 is modest compared to that of EDEM1 (Fig 3A). Moreover, there is only data from single shRNAs presented. Could the authors please at least include another shRNA to confirm and demonstrate whether the targeting to ERLAD is accordingly scaled to loss of access to ERAD (based on the degree of OS-9 or EDEM1 remaining)?

    __The reviewer is right. The phenotype (i.e., lysosomal delivery of NHK, Figs. 3B, 3C) is quite modest upon EDEM1 silencing. However, one has to consider that in contrast to OS9 lectins, EDEM1 is an enzyme, and residual protein may partially facilitate NHK de-mannosylation and access to the ERAD pathways and therefore reduce the ERLAD contribution for NHK clearance in these cells. Moreover, cells also express EDEM2 and 3 that may partially compensate the loss of EDEM1. __

    While degradation is implied, it is not specifically demonstrated at any point in the manuscript. Perhaps the authors might include some demonstration of NHK stabilization in one of the figures via a translational shutoff or pulse-chase assay.

    __In the new submission, we show biochemical analyses (pulse-chase) that reveal the decay of radiolabeled NHK (Fig. 2A, lanes 1-3) and BACE457delta (Fig. 6A, lanes 1-3), the inhibition by PS341 (lanes 4, 5) and by KIF (lanes 8, 9), and the intervention of lysosomal enzymes when ERAD is inhibited (lanes 6, 7 and 10, 11). Moreover, we confirm that the protein delivered to the endolysosome is eventually degraded by performing a Bafilomycin washout experiment (new Fig. 2J-2O). __

    10-30% of NHK-HA positive endolysosomes are detected even with Baf alone (e.g. Fig 2E)? Does this mean that Baf impairs ERAD to some extent since or is it evidence for continuous ERLAD involvement when ERAD is intact? If so, how much is its contribution?

    Pulse-chase analyses (new Fig. 2D) and published data show that BafA1 or chloroquine do not inhibit clearance of the ERAD substrates NHK and BACE457delta (e.g., Liu et al 1999, Molinari et al 2002, references in the manuscript). A basal level of endolysosomal delivery between the 20 and 30% as quantified with LysoQuant is observed in all experiments (Figs. 2I, 2O, 3C, 3F, 4C, 4K, 5H, 6G, 6O), which have been performed in 3 different cell lines (3T3, HEK293, MEF). We measure similar basal levels also when ER-phagy is monitored on quantification of lysosomal delivery of endogenous ER marker proteins (e.g., CNX), possibly to be ascribed to constitutive ER phagy that controls physiologic ER turnover.

    An accounting of how much ERLAD is contributing to NHK degradation with or without ERAD impairment is not really present.. Effectively, how much degradation capacity is ERLAD making up? These would be interesting data to include if possible as they would speak to the "division of labour" for ER substrate degradation its potentially dynamic nature.

    The biochemical analyses show the contribution of ERLAD on NHK (new Figs 2B, 2C, grey zones) and BACE457delta (new Figs. 6B,C, grey zones) clearance, when ERAD is dysfunctional.

    MINOR COMMENTS

    1. In Figure 4, an increase is observed for the rescue of FAM134B-/-MEFs with WT FAM134B that is 50% greater that of WT MEFs, suggesting that its availability might be rate limiting. Could the authors compare the relative levels of FAM134B for the WT and KO-rescue MEFs to address this possibility?

    __The referee is right in assuming that FAM134B, expressed at low levels in these cells, is limiting. We now show the levels of endogenous FAM134B and of recombinant FAM134B in WB (new Fig. 4A). __

    In Figures 1 and 6, the terms siOS9 and siEDEM1 are used but Figure 3 shows data from shRNAs and not siRNAs.

    We apologize for the mistake. We have corrected this in the new Figures 1 and 7.

    Samples from Figure 3 treated with Baf but this is not indicated in the figure or figure legend.

    We have corrected this, thank you.

    VCP/p97 inhibitors typically stabilize ERAD glycoprotein substrates better than proteasome inhibitors do. Is the same degree of endolysosomal targeting present ?

    __For the convenience of the reviewer (we did not put these data in the new manuscript). In our experiments, the p97 inhibitor DBeQ is less efficient in deviating NHK to the endolysosomal degradative compartments, if compared with KIF (see below). At higher doses, DBeQ also inhibits other AAA-ATPases (e.g., VPS4, which plays a role in certain types of autophagy). This, or other cross-reactivities of DBeQ could explain the moderate capacity to activate ERLAD pathways as a response of ERAD inhibition, if compared with the phenotypes observed when ERAD is inhibited with KIF or PS341. __

    Reviewer #1 (Significance (Required)):

    Deconvolution of the different pathways taken by misfolded proteins to escape the ER is of great interest not only to the ER community but also represents consequences to consider for those interested in therapeutics involving UPS inhibition. While concise, this manuscript does a good job of trying to demonstrate the principal of substrate rerouting and the prioritisation of degradation pathways. Overall, the manuscript is well written, the experiments presented are performed to a sufficient standard, the data are lean but of good quality, and the appropriate statistical analyses have mostly been included where necessary and are described. The Methods and Materials is brief but describes the experiments that have been performed. The manuscript is brief in its results and would obviously benefit from additional complementary assays that would strengthen and broaden the authors arguments for rerouting. But too their credit, the authors do not grossly overstate their findings and merely present the culmination of a set of experiments to answer a single question - what happens to a misfolded glycoprotein substrate when ERAD is impaired. This is a key question with broad implications.

    While their limited data clearly demonstrates an acquired dependence on ERLAD, one can't help but wonder how broadly these findings hold true, as only a single glycoprotein substrate example is used.

    We have now added a complete set of experiments (imaging + biochemical to monitor clearance of the model polypeptides by pulse-chase analyses) performed with a second ERAD substrate (BACE457delta, Fig. 6). These data fully recapitulate the results obtained with NHK.

    Moreover, it is not clear what percentage ERLAD contributes to overall NHK degradation (with or without ERAD) as the total NHK amount remaining is not assessed or measured.

    Pulse-chase analyses (new Fig. 2D) and published data (e.g., Liu et al 1999, Molinari et al 2002, references in the manuscript) show that BafA1 or chloroquine do not inhibit clearance of the ERAD substrates NHK and BACE457delta. The biochemical analyses now show the contribution of ERLAD on NHK (new Figs 2B, 2C, grey zones) and BACE457delta (new Figs. 6B,C, grey zones) clearance, when ERAD is dysfunctional.

    Nevertheless, the manuscript is an advancement of understanding of the fate of substrates unable to access ERAD and raises many future questions of interdependency between the ERAD and ERLAD pathways. The data just need a bit of shoring up.

    Expertise - ERAD, UPS, protein quality control

    Reviewer #2 (Evidence, reproducibility and clarity (Required)):

    The endoplasmic reticulum (ER) is a crucial site for protein synthesis and folding within the cell, and strict protein quality control is essential for maintaining ER homeostasis. In this context, ER-associated degradation (ERAD) and the unfolded protein response (UPR) play pivotal roles. Recent researches have highlighted the significance of ER-phagy in protein quality control. In this manuscript, the authors demonstrate the role of FAM134B in degrading misfolded proteins such as ATZ through the ER-phagy pathway when the ERAD pathway is obstructed. This work partially addresses a prominent issue in the field, unveiling the interconnections between different regulatory pathways in maintaining ER homeostasis.

    Major issues: 1: In a multitude of experiments, the authors employed Bafilomycin A1 (BafA1) to block the fusion between autophagosomes and lysosomes, attempting to demonstrate that the clearance of misfolded proteins mediated by FAM134B is independent of autolysosomes. However, in Figure 4, the lack of rescue of FAM134B knockout by overexpressing FAM134B△LIR suggests a dependence on the interaction between FAM134B and LC3. The conclusions drawn before and after appear contradictory.

    We apologize if our explanations were unclear. We have now modified the text and performed new experiments to clarify these issues.

    __The inhibitor of the V-ATPase BafA1 is used here to inhibit the activity of lysosomal hydrolases and to accumulate undegraded material in the LAMP1-endolysosomes (note that these endolysosomes also display RAB7 at their limiting membrane) (Fregno et al 2018, Forrester et al 2019, Fregno et al 2021, …). __

    __In Figs. 2A-2D, we now monitor the lack of NHK stabilization by cell exposure to BafA1 (Fig. 2D), which correlates with lack of accumulation of NHK in the LAMP1-positive compartment (e.g., Fig. 2F, 2J, and quantifications in 2I and 2O). The biochemical data also show that BafA1 stabilizes NHK in cells where ERAD has been inactivated with PS341 or KIF (Fig. 2A, lanes 6, 7, 10, 11 and grey zones in Figs. 2B and 2C), which correlates with accumulation of NHK in LAMP1-positive organelles (Figs. 2G, 2H, 2I, 2K, 2M, 2O). __

    __In Figs. 2J-2O, we have now added panels showing that NHK clearance from the LAMP1-positive endolysosome lumen is restored upon BafA1 washout. __

    Importantly, the involvement of the lipidation machinery, of the ER-phagy receptor FAM134B and of the LC3-binding function of FAM134B (the LIR), does not necessarily imply the involvement of autophagosomes in the process under investigation, as the comment by the referee seems to suggest. For example, both the clearance from the ER of ATZ polymers and of mutant forms of procollagen rely on the LC3 lipidation machinery and on the LC3-binding function of FAM134B, but ERLAD of ATZ polymers does not rely on autophagosomes intervention (new Fig. 1B, arrow 1 and Fregno et al 2018), whereas ERLAD of procollagen relies on intervention of autophagosomes (new Fig. 1B, arrow 2 and Forrester et al 2019).

    2: Some Western blot data are insufficient to substantiate the author's conclusions. For instance, in Figure 5D, the ATG7 KO line is inadequately supported

    The WB show____s the absence of ATG7 in the ATG7-KO cells (a well-established cell line generated in the lab of Masaaki Komatsu (____Komatsu M, et al. J Cell Biol 169: 425-434____) and used in many____ laboratories, including our lab in Fumagalli et al 2016, Fregno et al 2018, Fregno et al 2021, Loi et al 2019, Kucinska et al 2023). We agree with the reviewer that the anti-Atg7 shows cross-reactions. We have now added a WB showing the lack of LC3 lipidation in the Atg7-KO cells exposed to nutrient deprivation (new Fig. 5D).

    3: The author employed Lamp1 antibody for lysosomal staining in cells and observed a significant abundance of lysosomes in some experiments, as depicted in Figure 2C, 2D, 4I, etc. Is the phenomenon of lysosomes extensively filling the entire cell a common occurrence? Is it indicative of a normal physiological state?

    There may be variations depending on the cell type used for the experiments. In the new version of the manuscript, we now present imaging data for 3 cell lines (NIH 3T3 with stable expression of NHK and ATZ (Figs. 2E-2H), MEF (Figs. 2J-2N, 4, 5, 6) and HEK293 with transient expression of ERAD clients (Figs. 3).

    Minor issues： 1: Some immunofluorescence experimental data are unclear. Please request the authors to replace these with more distinct images, as seen in Figure 3B and 3E.

    We hope that the quality of the new images will be considered sufficient for publication.

    2: Some expressions appear to be questionable. For instance, the necessity of utilizing endolysosomes requires clarification.

    For the use of endolysosomes (lysosome would be incorrect in our opinion to indicate these LAMP1/RAB7-positive degradative organelles), we now refer to the papers by Bright et al ____Endolysosomes Are the Principal Intracellular Sites of Acid Hydrolase Activity____ Curr Biol 2016, and the original definition by Huotari and Helenius ____Endosome maturation EMBO J 2011 (Introduction, page 2).

    3: Some writing lacks precision, such as referring to FAM134B as FAM134.

    __Corrected, thank you____ __ Reviewer #2 (Significance (Required)):

    o General assessment: o Advance: provide an meaningful evidence that how two degradative pathways are coordinated in maintaining ER homeostasis. o Audience: cell biologist o Reviewer's expertise: autophagy, vesicle trafficking, organelle biolgy Reviewer #3 (Evidence, reproducibility and clarity (Required)):

    In their study, Fasana and colleagues investigate protein quality control in the ER. Specifically, they test whether folding-incompetent proteins that are normally cleared by ER-associated degradation (ERAD) can also be targeted for degradation by direct vesicular transport from the ER to lysosomes in case ERAD is blocked. They show that blocking ERAD pharamacologically or genetically indeed leads to re-rerouting of an ERAD model substrate (the NHK variant of alpha-antitrypsin) to lysosomes and that this pathway requires the reticulon-like protein FAM134B, the ability of FAM134B to interact with the ubiquitin-like protein LC3 and the machinery for LC3 lipidation.

    The paper is, for the most part, easy to follow. There are, however, a few minor issues and I think the authors could do more to connect their work with similar studies in the literature. Accordingly, I have some general and specific suggestions to make the manuscript more accessible for the reader.

    General suggestions

    1. To avoid confusion, it would be helpful to more clearly distinguish between vesicular transport to endolysosomes and autophagy. Previous work by the authors has defined a trafficking pathway from the ER to endolysosomes that appears to rely on conventional vesicle-mediated transport (Fregno et al, EMBO J 2018). This pathway delivers material from the ER lumen to the lumen of endolysosomes, which are both topologically equivalent to the extracellular space. Hence, this pathway is distinct from autophagy, which is the transport of cytoplasmic components to endolysosomes and thus the transport of material from intracellular to extracellular space. This distinction is particularly important as both vesicular ER-to-lysosome transport and autophagy of the ER involve LC3 and FAM134B, which is typically referred to as an ER-phagy receptor. To make this less confusing, it may be helpful to explain that FAM134B appears to be a multifunctional molecule that can function as a receptor for macroautophagy but also in the vesicular transport pathway studied here. In addition, it would be helpful to point out that LC3 appears to also have roles unrelated to autophagosome formation.

    The reviewer is referring to the original definition of ERLAD to describe the mechanisms of clearance of ATZ polymers (Fregno et al 2018). The definition of ERLAD has now been expanded and is given, for example, in Klionsky DJ, et al (2021) Guidelines for the use and interpretation of assays for monitoring autophagy (4th edition). Autophagy 17: 1-382 and is explained in detail in our recent review Rudinskiy M, Molinari M (2023) ER-to-lysosome-associated degradation in a nutshell: mammalian, yeast, and plant ER-phagy as induced by misfolded proteins. Febs Letters: 1928-1945.

    __Notably, the acronym ERAD for ER-associated degradation has originally been used to describe ____the proteasomal clearance from the ER of misfolded pro-alpha factor in a reconstituted yeast system in McCracken AA, Brodsky JL (1996) Assembly of ER-associated protein degradation in vitro: dependence on cytosol, calnexin, and ATP. The Journal of cell biology 132: 291-298. Only later on, the acronym has been used as an umbrella term that now covers all the pathways that control proteasomal clearance of misfolded proteins from the ER. A short historical excursus is presented in the new introduction to better explain these issues. __

    It is well established that LC3 and the LC3 lipidation machinery have functions that go beyond macroautophagy (which involves double membrane autophagosomes). Micro-autophagy (or micro-ER-phagy to remain on the topic of our paper) is an example of autophagic pathway relying on ER-phagy receptor that engage LC3, on the LC3 lipidation machinery, without involving autophagosomes. This is schematically represented in the new Fig. 1B.

    Several recent papers that appear relevant to the present study are not mentioned. In particular, Sun et al., Dev Cell 2023 (PMID: 37922908) appears worthy of discussion, as does Gonzalez et al., Nature 2023 (PMID: 37225996).

    Thank you. Both papers are not directly linked to our study addressing the intervention of ERLAD pathways when ERAD activity is impaired. In particular the work of Gonzales et al describes post-translational modification of ER-phagy receptors for their activation. The Sun et al paper is not really related to the topic covered in our manuscript, but we cite it as an alternative pathway that removes ATZ from the ER (page 8).

    Specific suggestions

    1. Abstract: The abstract begins with "About 40% of the eukaryotic cell's proteome is synthesized ... in the ER." Similar statements can be found in many papers and purportedly reflect common knowledge. However, it is unclear where the figure of 'about 40%' comes from. It would be proper to provide a reference and demonstrate that giving such a fairly precise estimate is supported by experimental data. Alternatively, the statement could be modified to avoid being precise than is justified.

    No reference is allowed in the abstract. We therefore modified the sentence as suggested by the reviewer.

    1. p2: "The ER is site of gene expression in nucleated cells and ... native proteins to be delivered at their site of activity ...". There is something missing at the beginning of this sentence. Also, it should be 'delivered to their site of activity', not 'delivered at'.

    Thank you

    1. p2: "... by mechanistically distinct ER-phagy pathways collectively defined as ER-to-lysosome-associated degradation ERLAD." This statement suggests that all pathways subsumed under the term ERLAD are ER-phagy pathways, which I believe is misleading (see comment above on the distinction between autophagy and vesicular transport pathway).

    See point 1.

    1. p2: "KIF selectively ...". Please spell out KIF and explain what kind of compound it is.

    __Thank you, we changed to “____The alkaloid kifunensine (KIF) is a cell permeable selective inhibitor of the members of the glycosyl hydrolase 47 family of ____a____1,2-mannosidases____”____ __

    1. p3: "Notably, ERAD inhibition delays, rather than blocking degradation of ERAD clients ...". Please correct, for example: Notably, ERAD inhibition delays rather than blocks degradation of ERAD clients ...

    Thank you

    Figures 2 - 5: The number of quantified cells is given but it is not clear if experiments were done once or in biological replicates. Please indicate this in the figure legends.

    __N is now given for all panels in the corresponding figure legends.____ __

    1. p4: "To verify if ERAD inactivation ..." sounds odd. Less ambiguous would be 'To test whether' or 'To ask if'.

    Thank you

    1. p7, beginning of discussion: Please correct "delivered at" to 'delivered to'.

    Thank you

    Reviewer #3 (Significance (Required)):

    This is a concise and convincing manuscript with a clear message. The idea that proteins that cannot be processed by ERAD can be eliminated by other means, for instance by autophagy, is not new. Similarly, the FAM134B- and LC3-dependent pathway for ER-to-lysosome transport has been described by the authors before (Fregno et al, EMBO J 2018). Furthermore, the study exclusively relies on microscopy and does not attempt to tackle new mechanistic questions. Still, this study presents a definite functional advance in our understanding of the interplay of various ER quality control pathways.

    The findings presented here will be of interest mainly to molecular cell biologists working on protein quality control and organelle homeostasis. However, given the disease-relevance of misfolded proteins, and alpha-antitrypsin in particular, the impact of this study may eventually go beyond basic research and may also interest translational researchers.

  2. Review Commons

    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

    In their study, Fasana and colleagues investigate protein quality control in the ER. Specifically, they test whether folding-incompetent proteins that are normally cleared by ER-associated degradation (ERAD) can also be targeted for degradation by direct vesicular transport from the ER to lysosomes in case ERAD is blocked. They show that blocking ERAD pharamacologically or genetically indeed leads to re-rerouting of an ERAD model substrate (the NHK variant of alpha-antitrypsin) to lysosomes and that this pathway requires the reticulon-like protein FAM134B, the ability of FAM134B to interact with the ubiquitin-like protein LC3 and the machinery for LC3 lipidation.

    The paper is, for the most part, easy to follow. There are, however, a few minor issues and I think the authors could do more to connect their work with similar studies in the literature. Accordingly, I have some general and specific suggestions to make the manuscript more accessible for the reader.

    General suggestions

    1. To avoid confusion, it would be helpful to more clearly distinguish between vesicular transport to endolysosomes and autophagy. Previous work by the authors has defined a trafficking pathway from the ER to endolysosomes that appears to rely on conventional vesicle-mediated transport (Fregno et al, EMBO J 2018). This pathway delivers material from the ER lumen to the lumen of endolysosomes, which are both topologically equivalent to the extracellular space. Hence, this pathway is distinct from autophagy, which is the transport of cytoplasmic components to endolysosomes and thus the transport of material from intracellular to extracellular space. This distinction is particularly important as both vesicular ER-to-lysosome transport and autophagy of the ER involve LC3 and FAM134B, which is typically referred to as an ER-phagy receptor. To make this less confusing, it may be helpful to explain that FAM134B appears to be a multifunctional molecule that can function as a receptor for macroautophagy but also in the vesicular transport pathway studied here. In addition, it would be helpful to point out that LC3 appears to also have roles unrelated to autophagosome formation.
    2. Several recent papers that appear relevant to the present study are not mentioned. In particular, Sun et al., Dev Cell 2023 (PMID: 37922908) appears worthy of discussion, as does Gonzalez et al., Nature 2023 (PMID: 37225996).

    Specific suggestions

    1. Abstract: The abstract begins with "About 40% of the eukaryotic cell's proteome is synthesized ... in the ER." Similar statements can be found in many papers and purportedly reflect common knowledge. However, it is unclear where the figure of 'about 40%' comes from. It would be proper to provide a reference and demonstrate that giving such a fairly precise estimate is supported by experimental data. Alternatively, the statement could be modified to avoid being precise than is justified.
    2. p2: "The ER is site of gene expression in nucleated cells and ... native proteins to be delivered at their site of activity ...". There is something missing at the beginning of this sentence. Also, it should be 'delivered to their site of activity', not 'delivered at'.
    3. p2: "... by mechanistically distinct ER-phagy pathways collectively defined as ER-to-lysosome-associated degradation ERLAD." This statement suggests that all pathways subsumed under the term ERLAD are ER-phagy pathways, which I believe is misleading (see comment above on the distinction between autophagy and vesicular transport pathway).
    4. p2: "KIF selectively ...". Please spell out KIF and explain what kind of compound it is.
    5. p3: "Notably, ERAD inhibition delays, rather than blocking degradation of ERAD clients ...". Please correct, for example: Notably, ERAD inhibition delays rather than blocks degradation of ERAD clients ...
    6. Figures 2 - 5: The number of quantified cells is given but it is not clear if experiments were done once or in biological replicates. Please indicate this in the figure legends.
    7. p4: "To verify if ERAD inactivation ..." sounds odd. Less ambiguous would be 'To test whether' or 'To ask if'.
    8. p7, beginning of discussion: Please correct "delivered at" to 'delivered to'.

    Significance

    This is a concise and convincing manuscript with a clear message. The idea that proteins that cannot be processed by ERAD can be eliminated by other means, for instance by autophagy, is not new. Similarly, the FAM134B- and LC3-dependent pathway for ER-to-lysosome transport has been described by the authors before (Fregno et al, EMBO J 2018). Furthermore, the study exclusively relies on microscopy and does not attempt to tackle new mechanistic questions. Still, this study presents a definite functional advance in our understanding of the interplay of various ER quality control pathways.

    The findings presented here will be of interest mainly to molecular cell biologists working on protein quality control and organelle homeostasis. However, given the disease-relevance of misfolded proteins, and alpha-antitrypsin in particular, the impact of this study may eventually go beyond basic research and may also interest translational researchers.

  3. Review Commons

    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

    The endoplasmic reticulum (ER) is a crucial site for protein synthesis and folding within the cell, and strict protein quality control is essential for maintaining ER homeostasis. In this context, ER-associated degradation (ERAD) and the unfolded protein response (UPR) play pivotal roles. Recent researches have highlighted the significance of ER-phagy in protein quality control. In this manuscript, the authors demonstrate the role of FAM134B in degrading misfolded proteins such as ATZ through the ER-phagy pathway when the ERAD pathway is obstructed. This work partially addresses a prominent issue in the field, unveiling the interconnections between different regulatory pathways in maintaining ER homeostasis.

    Major issues:

    1. In a multitude of experiments, the authors employed Bafilomycin A1 (BafA1) to block the fusion between autophagosomes and lysosomes, attempting to demonstrate that the clearance of misfolded proteins mediated by FAM134B is independent of autolysosomes. However, in Figure 4, the lack of rescue of FAM134B knockout by overexpressing FAM134B△LIR suggests a dependence on the interaction between FAM134B and LC3. The conclusions drawn before and after appear contradictory.
    2. Some Western blot data are insufficient to substantiate the author's conclusions. For instance, in Figure 5D, the ATG7 KO line is inadequately supported
    3. The author employed Lamp1 antibody for lysosomal staining in cells and observed a significant abundance of lysosomes in some experiments, as depicted in Figure 2C, 2D, 4I, etc. Is the phenomenon of lysosomes extensively filling the entire cell a common occurrence? Is it indicative of a normal physiological state?

    Minor issues:

    1. Some immunofluorescence experimental data are unclear. Please request the authors to replace these with more distinct images, as seen in Figure 3B and 3E.
    2. Some expressions appear to be questionable. For instance, the necessity of utilizing endolysosomes requires clarification.
    3. Some writing lacks precision, such as referring to FAM134B as FAM134.

    Significance

    • General assessment:
    • Advance: provide an meaningful evidence that how two degradative pathways are coordinated in maintaining ER homeostasis.
    • Audience: cell biologist
    • Reviewer's expertise: autophagy, vesicle trafficking, organelle biolgy
  4. Review Commons

    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 manuscript from Fasana et al., the authors present data that investigates potential compensatory degradation pathways for misfolded glycoproteins in the ER - postulating that the ER-to-lysosome associated degradation (ERLAD) pathway becomes employed in the absence of a path for substrates to reach the ER-associated degradation (ERAD) mechanism. Using the classic ERAD substrate alpha1-antitrypsin NHK variant (NHK), the authors first demonstrate that pharmacologically preventing access of NHK to ERAD either with KIF (early) or PS-341 (late) elevates the number of LAMP-1 positive endolysosomes also immunoreactive for NHK (via HA), similar to what is observed for the ATZ variant that forms polymers in the ER (Fig 2). The authors next use shRNAs that silence essential ERAD factors (EDEM1, OS-9) involved in glycan recognition to demonstrate comparable enrichment of NHK in endolysosomes through genetic disruption (Fig 3). Next, the authors employ FAM134B-deficient MEFs to demonstrate the requirement for this ER-phagy receptor when ERAD is unavailable (Fig 4). Reconstituting FAM134B-/- MEFs treated with KIF/PS-341 + Baf, with a full length FAM134B rescue plasmid restores endolysosomal accumulation of NHK while a FAM134B-∆LIR does not, providing supporting evidence for substrate rerouting to ERLAD. Finally, the authors use knockouts of Atg7 and Atg13 to demonstrate dependence on LC3 lipidation and independence from macro-ERphagy (Fig 6), that points towards a pathway that is like that used to remove ATZ polymers. From these data, the authors conclude that ERLAD is increasingly engaged for substrate degradation when ERAD is impaired.

    Major comments:

    1. All assays rely on quantification of the NHK-HA substrates by microscopy. Would it be possible for the authors to also include biochemical analysis of NHK - potentially including data assessing its changing abundance and glycosylation state?
    2. In Figure 3D, the knockdown of OS-9.1/2 is modest compared to that of EDEM1 (Fig 3A). Moreover, there is only data from single shRNAs presented. Could the authors please at least include another shRNA to confirm and demonstrate whether the targeting to ERLAD is accordingly scaled to loss of access to ERAD (based on the degree of OS-9 or EDEM1 remaining)?
    3. While degradation is implied, it is not specifically demonstrated at any point in the manuscript. Perhaps the authors might include some demonstration of NHK stabilization in one of the figures via a translational shutoff or pulse-chase assay.
    4. 10-30% of NHK-HA positive endolysosomes are detected even with Baf alone (e.g. Fig 2E)? Does this mean that Baf impairs ERAD to some extent since or is it evidence for continuous ERLAD involvement when ERAD is intact? If so, how much is its contribution?
    5. An accounting of how much ERLAD is contributing to NHK degradation with or without ERAD impairment is not really present.. Effectively, how much degradation capacity is ERLAD making up? These would be interesting data to include if possible as they would speak to the "division of labour" for ER substrate degradation its potentially dynamic nature.

    Minor comments:

    1. In Figure 4, an increase is observed for the rescue of FAM134B-/-MEFs with WT FAM134B that is 50% greater that of WT MEFs, suggesting that its availability might be rate limiting. Could the authors compare the relative levels of FAM134B for the WT and KO-rescue MEFs to address this possibility?
    2. In Figures 1 and 6, the terms siOS9 and siEDEM1 are used but Figure 3 shows data from shRNAs and not siRNAs.
    3. Samples from Figure 3 treated with Baf but this is not indicated in the figure or figure legend.
    4. VCP/p97 inhibitors typically stabilize ERAD glycoprotein substrates better than proteasome inhibitors do. Is the same degree of endolysosomal targeting present

    Significance

    Deconvolution of the different pathways taken by misfolded proteins to escape the ER is of great interest not only to the ER community but also represents consequences to consider for those interested in therapeutics involving UPS inhibition. While concise, this manuscript does a good job of trying to demonstrate the principal of substrate rerouting and the prioritisation of degradation pathways. Overall, the manuscript is well written, the experiments presented are performed to a sufficient standard, the data are lean but of good quality, and the appropriate statistical analyses have mostly been included where necessary and are described. The Methods and Materials is brief but describes the experiments that have been performed. The manuscript is brief in its results and would obviously benefit from additional complementary assays that would strengthen and broaden the authors arguments for rerouting. But too their credit, the authors do not grossly overstate their findings and merely present the culmination of a set of experiments to answer a single question - what happens to a misfolded glycoprotein substrate when ERAD is impaired. This is a key question with broad implications.

    While their limited data clearly demonstrates an acquired dependence on ERLAD, one can't help but wonder how broadly these findings hold true, as only a single glycoprotein substrate example is used. Moreover, it is not clear what percentage ERLAD contributes to overall NHK degradation (with or without ERAD) as the total NHK amount remaining is not assessed or measured. Nevertheless, the manuscript is an advancement of understanding of the fate of substrates unable to access ERAD and raises many future questions of interdependency between the ERAD and ERLAD pathways. The data just need a bit of shoring up.

    Expertise - ERAD, UPS, protein quality control

  5. Version published to 10.1101/2023.11.28.569025v1 on bioRxiv