Inhibition of proteasome rescues a pathogenic variant of respiratory chain assembly factor COA7

Abstract Nuclear and mitochondrial genome mutations lead to various mitochondrial diseases, many of which affect the mitochondrial respiratory chain. The proteome of the intermembrane space (IMS) of mitochondria consists of several important assembly factors that participate in the biogenesis of mitochondrial respiratory chain complexes. The present study comprehensively analyzed a recently identified IMS protein cytochrome c oxidase assembly factor 7 (COA7), or RESpiratory chain Assembly 1 (RESA1) factor that is associated with a rare form of mitochondrial leukoencephalopathy and complex IV deficiency. We found that COA7 requires the mitochondrial IMS import and assembly (MIA) pathway for efficient accumulation in the IMS. We also found that pathogenic mutant versions of COA7 are imported slower than the wild‐type protein, and mislocalized proteins are degraded in the cytosol by the proteasome. Interestingly, proteasome inhibition rescued both the mitochondrial localization of COA7 and complex IV activity in patient‐derived fibroblasts. We propose proteasome inhibition as a novel therapeutic approach for a broad range of mitochondrial pathologies associated with the decreased levels of mitochondrial proteins.

The MIA machinery for many proteins drives protein import into the mitochondrial intermembrane space. Mohanraj and colleagues describe here the characterisation of a novel MIA40/CHCHD4 substrate, called RESA1/COA7. In figures 1-6 the interaction with MIA40 and the oxidation of RESA1 is extensively characterized demonstrating that RESA1 indeed behaves in many aspects like classical MIA substrates. In the last figures, an interesting RESA1 mutant that is hampered in mitochondrial import is introduced. Cytosolic accumulation of the RESA1 mutant prompts its proteasomal degradation. Inhibition of this degradation appeared to partially rescue both, mitochondrial localization of RESA1 and complex IV activity (for which RESA1 is an assembly factor). This is a well-written and data-rich manuscript. Especially the characterisation of RESA1 as MIA40 substrate is comprehensive. However, this part is also less exciting than the findings presented later in the manuscript. This is mainly because Petrungaro and colleagues already found RESA1 (as C1orf163) in their interactome of MIA40 as an interaction partner that coprecipitates under denaturing conditions indicating disulfide-linkage between MIA40 and RESA1 (Cell Metab, 2015), and Kozjak-Pavlovic and colleagues described its function and the consequences of its absence already in 2014 (JMB, 2014). Still, the mechanistic unravelling of the precise import and oxidation mechanism is of high value.
The last figures present the most exciting and novel insights into the biology of MIA-dependent import. I feel however that this part is underdeveloped, and that it would require some additional work to render this manuscript suitable for publication. It remains otherwise unclear whether only the levels of RESA1 in both cytosol and IMS increase upon MG132 treatment or whether the ratio is tilted towards the IMS. If at all, the ratio seems to shift towards the cytosol (8F). 3/ Figure 8D: Why does MG132 only have an effect in the absence of CHX? If the recognition for degradation were an early event, than the authors should perform radioactive pulse-chase experiments 4/ Figure 8D: MG132 has no effect on endogenous RESA1, yet it exerts an effect on overexpressed wild type RESA1 (almost as strong as with the RESA1 mutants). How high are overexpression levels? The authors should titrate RESA1-WT levels to endogenous amounts and then confirm that they do not observe effects of MG132 treatment. 5/ Figure 8: Does RESA1 become ubiquitinated upon MG132 treatment? 6/ Figure 9A: Why does the HSP70 signal disappear without proteasomal inhibitor treatment? 7/ Figure 9C,D: what is the statistical reasoning for using 'standard error' and not 'standard deviation'? On what was the normalization, i.e. how do you know that mitochondrial isolation worked equally well? Unfortunately, the actual complex IV activity assay was described only poorly. The description should be improved. 8/ Figure 9. Respiratory chain activity upon MG132 treatment: This part should be extended by BN-PAGE analysis (is there more assembled complex IV present upon MG132 treatment), oxygen consumption assays, viability assays on galactose etc.. Moreover, comparison with control fibroblasts is missing. 9/ There appear to be certain 'redundancies' in the first figures. Many panels show in orthogonal approaches the interaction between MIA40 and RESA1. These figures could be condensed to allow expansion of the latter figures describing the interesting 'proteasomal effect'.

Major points
Minor points: 1/ The labelling of figures/extent of experimental description in figure legends (what has been done?, MW marker, which kind of IP, how was MIA40 expressed in the different figures: transient, stable cell lines, stable inducible cell lines, etc. à heavy overexpression might result in mislocalization and influence interpretation of results etc) is underdeveloped. This has to be improved to enable understanding of the performed experiments. Likewise, n numbers and quantifications of experiments are missing. 2/ The nomenclature in the field is somewhat of a mess. I would ask the authors to mention that MIA40 is also referred to as CHCHD4, and that ALR is the human homolog of the yeast Erv1. 3/ Figure 1B: Please provide additional immunoblots against more classical substrates of MIA40 4/ Figure 1A: Provide the proteomics data in full, e.g. as an excel file in the SI or as a link to a database. 5/ Figure 2D shows a very uneven expression of MIA40 variants. Is this due to transient expression? 6/ Figure 3A: The scheme is somewhat hard to grasp. Could the authors provide a better version? 7/ Figure 4D: Isolated mitochondria are not a good model for the determination of cysteine redox states. How can the authors ensure preservation of the endogenous redox state? Moreover, to present the data, the exposure should be varied and quantifications of the ratio 'ox/red' should be provided. 8/ Figure 6A: Provide sequencing data for the MIA40 CRISPR clones. 9/ Figure 6B: Why is MIA40 almost completely gone if only one allele is affected?
Referee #2 (Comments on Novelty/Model System for Author): The authors use in vitro import assays in combination with assays in cultured cells. This allows them to tackle mechanistic and physiological questions.

Referee #2 (Remarks for Author):
This manuscript characterizes COA7 (RESA1) as an IMS protein and a non-canonical substrate of MIA40. Some of the authors had previously reported a matrix localization of the protein but here its IMS residence and the involvement of MIA40 for its import are undoubtedly demonstrated. The protein is biomedically relevant because mutations in the human gene have been associated with mitochondrial leukoencephalopathy associated with mitochondrial respiratory chain complex IV deficiency. The authors identified the COA7 mutations as responsible for import failure, which leads to retention of the newly synthesized protein in the cytoplasm. However, the protein does not accumulate in this compartment because is actively degraded by the proteasome. The authors elegantly show that overexpression of mutant COA7 or inhibition of the proteasome with either MG132 or other clinically approved proteasome inhibitors restores COA7 levels in mitochondria and also its activity in complex IV assembly. Therefore, the authors suggest that proteasome inhibition may be a new venue to combat mitochondrial diseases associated with poor mitochondrial protein import or excessive degradation by the proteasome.
The manuscript is technically and conceptually sound, and appropriate for publication in EMBO Molecular Medicine. I have only two requests to improve the manuscript: 1-The experiments presented do not completely exclude the possibility that mutant RESA1 is not degraded by mitochondrial proteases. The authors should silence some of these proteases (e.g. AAA proteases) and test whether the protein is still degraded with a similar efficiency.
2-The authors suggest that clinically approved proteasome inhibitors, such as bortezomib, may be applied as therapeutic agents to combat at least a subset of mitochondrial disorders. The authors should discuss the potential side effects on mitochondria and other organelles. Minor points: 1-Page 10: "TIMM8A (CX9C) and COX19 (CX3C)" should be "TIMM8A (CX3C) and COX19 (CX9C)" 2-I strongly suggest the authors to use COA7 to refer to the protein. The use of alternative names only serves to confuse the literature.

Referee #3 (Remarks for Author):
This manuscript reports a careful analysis of the biogenesis of the mitochondrial respiratory chain assembly factor 1 (RESA1). RESA1 has been linked to mitochondrial leukoencephalopathy and complex IV deficiency. The authors show that RESA1, which contains 13 cysteine residues, is an unusual substrate of the mitochondrial intermembrane space assembly (MIA) system. In a remarkably complete characterization, they elucidated the molecular mechanisms of import of RESA1 into the mitochondrial intermembrane space, the interaction with Mia40 and disulfide bond formation. Importantly, the mitochondrial import of pathogenic mutant versions of RESA1 is slower than that of wild-type RESA1 and proteins accumulating in the cytosol are degraded by the proteasome. Using patient-derived fibroblasts, the authors discovered that inhibition of the proteasome rescued the localization of the mutant RESA1 to mitochondria and the activity of complex IV. This paper by leading experts of the field is of technically very high quality and written very well. It provides exciting novel findings on the role of the proteasome in the pathogenesis of mitochondrial diseases and opens the way for new therapeutic approaches by using clinically approved proteasome inhibitors.
I have only a few minor comments on this exciting paper.
1. The authors provide a complete characterization of the biogenesis of wild-type and mutant RESA1 with important medical implications. It would be helpful for the general readership to present a cartoon of the import pathway of RESA1 and the role of the proteasome, e.g. in the last figure.  Referee #1 (Remarks for Author): The MIA machinery for many proteins drives protein import into the mitochondrial intermembrane space. Mohanraj and colleagues describe here the characterization of a novel MIA40/CHCHD4 substrate, called RESA1/COA7. In figures 1-6 the interaction with MIA40 and the oxidation of RESA1 is extensively characterized demonstrating that RESA1 indeed behaves in many aspects like classical MIA substrates. In the last figures, an interesting RESA1 mutant that is hampered in mitochondrial import is introduced. Cytosolic accumulation of the RESA1 mutant prompts its proteasomal degradation. Inhibition of this degradation appeared to partially rescue both, mitochondrial localization of RESA1 and complex IV activity (for which RESA1 is an assembly factor).
This is a well-written and data-rich manuscript. Especially the characterisation of RESA1 as MIA40 substrate is comprehensive. However, this part is also less exciting than the findings presented later in the manuscript. This is mainly because Petrungaro and colleagues already found RESA1 (as C1orf163) in their interactome of MIA40 as an interaction partner that coprecipitates under denaturing conditions indicating disulfide-linkage between MIA40 and RESA1 (Cell Metab, 2015), and Kozjak-Pavlovic and colleagues described its function and the consequences of its absence already in 2014 (JMB, 2014). Still, the mechanistic unravelling of the precise import and oxidation mechanism is of high value.
The last figures present the most exciting and novel insights into the biology of MIA-dependent import. I feel however that this part is underdeveloped, and that it would require some additional work to render this manuscript suitable for publication. Major points Please provide a scheme of the RESA1 disease mutant. Figure 6A of the revised manuscript.

A: We have included a schematic presentation of mutated RESA1/COA7 in
The patient carries both mutations. Is there crosstalk between the two RESA1 mutants?

A: In patient fibroblasts we observed the decreased steady state levels of the exon2 Δ mutant compared to Y137C and wild-type. Similarly, in HEK293 cells when mutant versions were overexpressed separately, the exon2 Δ mutant was expressed to a lower level than the Y137C mutant. Therefore we think that the observed disproportion in protein levels is unlikely to result from a crosstalk between both mutant variants of RESA1/COA7. We have introduced this notion in Results section when we first describe levels of RESA1/COA7 in mt4229i cells (Fig EV3). Moreover, in response to the reviewer's question about possible dominant-negative effects of mutant proteins, we verified effects of mutant expression on cytochrome c oxidase in the presence of a wild-type protein and found no influence (please see the response below).
Are experiments depicted in figure 7 performed in the presence of endogenous RESA1?

A: All the experiments presented in Figure 7 (Figure 6 in the revised manuscript) were performed in the presence of native RESA1/COA7. The expression of wild-type and mutant RESA1/COA7 was obtained by transient expression. We have clarified this point both in figure legends and in the corresponding text of Results section.
Are RESA1 mutants dominant-negative, i.e. are complex IV levels and activity changed upon mutant expression?

A: Our approach to address reviewer's comment was to overexpress separately the mutated forms of RESA1/COA7 in HEK293 cells carrying wild type alleles of the protein. We then verified steady state levels of subunits of cytochrome c oxidase and the activity of the complex (new data included in Appendix Fig S5A and B). We did not find any negative effects of mutants' overexpression upon either composition or activity of cytochrome c oxidase. We conclude that mutants do not impinge a dominant-negative effect in the background of wild-type protein. This notion is further supported by the fact that both parents of the patient, who carry single alleles of pathogenic variants of RESA1/COA7, do not present any symptoms of the disease (Martinez Lyons et al, 2016).
Are the depicted experiments performed with transient expression or with stable cell lines? If it were transient expression, how can you be sure that you always have similar expression levels (just as an example: compare e.g. Figures 7A and B, expression levels of the two mutants with respect to each other differ)? This point is especially important as later experiments in the presence of MG132 indicate changes in protein levels. Figure 8A and B of the revised manuscript).

A: In the experiments presented in the Figure 7 (Figure 6 in the revised manuscript) we used transient transfection in order to obtain overexpression of proteins in HEK293 cells. We have now clarified this point both in figure legends and in the Results section. We agree with the reviewer's remark that transient transfection induces more variable levels of expression than those observed e.g. in stable cell lines. In order to prevent misinterpretation we standardized transfection conditions i.e. cell densities, timing of experiments and amounts of DNA and transfectant used. We have included the detailed description of this procedure in the Appendix section. Our method of transfection reproducibly yielded lower levels of mutant protein expression than the wild-type RESA1/COA7. In the majority of experiments we also observed the lower levels of the exon2 Δ mutant as compared with the Y137C mutant. In few experiments we saw slight variability in the exon2 Δ mutant expression, yet expression of the Y137C mutant was very reproducible. The reviewer suggests that our interpretation of the apparent rescue of mutant proteins by proteasome inhibition (Figure 7 D and F of the revised manuscript) may be hindered by unequal transfection. In our opinion, there is a minimal chance that this may have influenced our interpretation in this particular experimental setting. However, our conclusions are supported by the experiments on patient fibroblasts wherein we demonstrated that proteasome inhibition rescued also mutant proteins of physiological abundance (
2/ Figure 8: Please provide quantifications. It remains otherwise unclear whether only the levels of RESA1 in both cytosol and IMS increase upon MG132 treatment or whether the ratio is tilted towards the IMS. If at all, the ratio seems to shift towards the cytosol (8F). Fig 8F, lanes 3, 7, 10, 14).", which could suggest that we meant "a fraction of mutant protein". Therefore in the revised manuscript we modified this sentence to avoid this misleading trait: "Indeed, in the presence of MG132 mediated proteasome inhibition, the levels of RESA1 mutant proteins increased in the cytosol, which was paralleled by an increased mitochondrial content of the mutant proteins ( Fig  7F, lanes 3, 7, 10, 14)".

A: This comment tackles an important aspect of our discourse that obviously was inadequately explained in the manuscript. Inhibition of the proteasome increases an overall abundance of RESA1/COA7 mutants in the cell. In fact a substantial portion of the mutant protein, especially during transient expression, is present in the cytosol. Our data demonstrates however that in parallel the mitochondrial content of mutant proteins is also increased. Below we present quantification of localization of exon 2 deletion mutant with and without treatment with MG132. It demonstrates that indeed the protein abundance in the cytosol increases more than in mitochondria. However, it is the mitochondrial content of mutant protein that is beneficial to mitochondria function irrespective to changes in the ratio between cytosolic and mitochondrial fractions. In the submitted version of the manuscript we used a phrase "… more interestingly a larger portion of the mutant proteins was localized in mitochondria (
[Unpublished data removed upon the authors' request.] 3/ Figure 8D: Why does MG132 only have an effect in the absence of CHX? If the recognition for degradation were an early event, than the authors should perform radioactive pulse-chase experiments

A: In order to address this comment of the reviewer, we performed a pulse-chase experiment, in which we inhibited proteasome either during the radiolabeling of newly synthesized proteins (A) or directly after the labelling (B) or 1h after the labelling (C)(please see a scheme below). The chase was finalized by the affinity purification of overexpressed RESA1/COA7-Y137C HIS . When MG132 was applied 1h after the end of the labelling (C) we could not observe any increase in RESA1/COA7 in the eluate (line 10). This is consistent with the interpretation that RESA1/COA7 is protected from proteasomal degradation shortly after the synthesis when it is efficiently imported into mitochondria. Unfortunately, at earlier times of MG132 treatment (A and B) we observed a significant decrease of protein labelling (load fraction), which reflected a temporal decrease of translation caused by MG132. This side effect of proteasome inhibition has been previously described (Jiang & Wek, 2005; Wu et al, 2009). Therefore we could not conclude about the early effects of proteasomal inhibition on the stability of RESA1/COA7-Y137C HIS . Taking this result into consideration we have milden our conclusions in the Results section and stated that: "
In contrast under active translation mutant COA7 were degraded by proteasome, while wild-type protein was only marginally affected suggesting that proteins with slower rate of import to mitochondria were sensitive to proteasome-mediated degradation (Fig 7D, lanes 4 and  5)." We also modified the Discussion section accordingly. Figure 8D: MG132 has no effect on endogenous RESA1, yet it exerts an effect on overexpressed wild type RESA1 (almost as strong as with the RESA1 mutants). How high are overexpression levels? The authors should titrate RESA1-WT levels to endogenous amounts and then confirm that they do not observe effects of MG132 treatment Fig S4C). We then checked the effect of MG132 upon the overexpressed protein and found that MG132 did not stabilize overexpressed RESA1/COA7 when the load of plasmid DNA was decreased to 1µg per 60cm 2 (Appendix Fig S4C). This is consistent with our original assumption that proteasomal degradation of the wild-type RESA1/COA7 results from its overproduction and inefficiency to be imported into mitochondria. However this result does not influence the core conclusions of the manuscript as the effect of proteasome inhibition on mutant RESA1/COA7 was confirmed in patient fibroblasts. Figure 8: Does RESA1 become ubiquitinated upon MG132 treatment? A: In order to verify whether RESA1/COA7 can undergo ubiquitination we overexpressed ubiquitin tagged with His tag together with COA7 FLAG and then performed affinity purification of ubiquitin in the presence of MG132 (Fig EV3A).

We purified various species of RESA1/COA7 corresponding to the protein modified with ubiquitin chains of different lengths. At the same time we could not co-purify the native RESA1/COA7, which is consistent with our former interpretation that RESA1/COA7 is subject to proteasome degradation only when the efficiency of import to mitochondria is impaired by mutation in the protein itself or by an increased protein supply in the cytosol due to the overexpression.
6/ Figure 9A: Why does the HSP70 signal disappear without proteasomal inhibitor treatment?

A: In the figure 9A (figure 8A of the revised manuscript) we present changes in protein levels following treatments with various inhibitors of proteasome. As it is mentioned in the manuscript HSP70 is known to accumulate in response to proteotoxic stress evoked by inhibition of proteasome, which explains a drastic difference of HSP70 levels between DMSO and inhibitor treated samples (Kim et al, 1999; Awasthi & Wagner, 2005). We included HSP70 Western blot in the previously submitted version of the manuscript as an additional proof that concentration of inhibitors used were effective.
7/ Figure 9C,D: what is the statistical reasoning for using 'standard error' and not 'standard deviation'? On what was the normalization, i.e. how do you know that mitochondrial isolation worked equally well? Unfortunately, the actual complex IV activity assay was described only poorly. The description should be improved. Figure 9A,

B and C are normalized to control samples obtained from DMSO treated fibroblasts. The normalization to the DMSO treated control was necessary as over time we observed some variability in absolute activity of complex IV. This was most probably due to various batches of digitonin used over time. Commercially available digitonin is a natural plant extract, which contains variable amount of impurities and therefore the activity/quality of different batches may vary.
8/ Figure 9. Respiratory chain activity upon MG132 treatment: This part should be extended by BN-PAGE analysis (is there more assembled complex IV present upon MG132 treatment), oxygen consumption assays, viability assays on galactose etc.. Moreover, comparison with control fibroblasts is missing.

In addition we performed the proliferation assay on patient and control fibroblasts grown in galactose medium (Appendix Fig S5C and D). Substitution of glucose with galactose forces mammalian cells to depend on oxidative phosphorylation as a main source of ATP. Both cell lines grew slower in galactose medium as compared to glucose medium. We then verified how bortezomib influenced cell proliferation in galactose medium. Concentrations of bortezomib used to increase the assembly of respiratory supercomplexes and activity of complex IV caused a significant decrease in growth rate (data not shown). This is in agreement with a well known role of ubiquitin-proteasome system in the cell cycle (Bassermann et al, 2014). In order to avoid toxic effects of bortezomib we decreased the concentration to 2,5 nM and 1 nM, which was close to the lowest concentration of the drug that was still resulting in increased ubiquitination of proteins. In these conditions 2,5 nM bortezomib treatment for 24 h slightly increased cell growth in patient fibroblasts, yet the effect was statistically insignificant.
9/ There appear to be certain 'redundancies' in the first figures. Many panels show in orthogonal approaches the interaction between MIA40 and RESA1. These figures could be condensed to allow expansion of the latter figures describing the interesting 'proteasomal effect'.

A: We have revised figure legends to identify all the weaknesses of description of experimental design. We introduced information requested by the reviewer.
2/ The nomenclature in the field is somewhat of a mess. I would ask the authors to mention that MIA40 is also referred to as CHCHD4, and that ALR is the human homolog of the yeast Erv1.

A: Indeed there is a certain level of inconsistency in the nomenclature referring to the MIA pathway. We use the term MIA40 instead of CHCHD4 because it relates to function of the protein rather than to its structural features and also it corresponds to the name of yeast ortholog where the protein and the pathway was first described. We agree with the reviewer that we should include the latter term and this has been introduced to the revised manuscript when MIA40 is mentioned for the first time. Analogically we introduced a term GFER, which is a systematic name for ALR, and information about homology of ALR to yeast Erv1.
3/ Figure 1B: Please provide additional immunoblots against more classical substrates of MIA40. Figure EV1A.

Interaction between MIA40 and its substrates is very transient and thus strong antibodies are required to demonstrate it via affinity purification and Western blotting. Unfortunately most of antibodies against MIA40 substrates in our possession are quite weak and thus we could not demonstrate interaction of MIA40 with other precursor proteins.
4/ Figure 1A: Provide the proteomics data in full, e.g. as an excel file in the SI or as a link to a database. A: We have included proteomics data in excel file format in the Source Data 1. Figure 2D shows a very uneven expression of MIA40 variants. Is this due to transient expression?

5/
A: Data presented in figure 2D (figure 1D in the revised manuscript)

refer to Flp-In T-REx 293 cells. These cells carry one copy of a cassette in the genome that allows for introducing a gene of interest, selection of stable clones and on-demand induction of expression with tetracycline. We have generated several clones, which express wild-type or mutant MIA40 tagged with a FLAG tag. Reproducibly C55S and SPS mutants show lower expression then wild-type and C53S mutant. We do not fully understand the source of this phenomenon. We have added this information in the Materials and Methods section.
6/ Figure 3A: The scheme is somewhat hard to grasp. Could the authors provide a better version? A: We have simplified the scheme to make it more approachable (Figure 2A of the revised manuscript). Figure 4D: Isolated mitochondria are not a good model for the determination of cysteine redox states. How can the authors ensure preservation of the endogenous redox state? Moreover, to present the data, the exposure should be varied and quantifications of the ratio 'ox/red' should be provided. Figure 3D of the revised manuscript) and we moved the data obtained from the isolated mitochondria to Appendix Fig  S2D. 8/ Figure 6A: Provide sequencing data for the MIA40 CRISPR clones. A: We included sequencing data in the Source Data 2 and 3. Figure 6B: Why is MIA40 almost completely gone if only one allele is affected? Figure 5D. This manuscript characterizes COA7 (RESA1) as an IMS protein and a non-canonical substrate of MIA40. Some of the authors had previously reported a matrix localization of the protein but here its IMS residence and the involvement of MIA40 for its import are undoubtedly demonstrated. The protein is biomedically relevant because mutations in the human gene have been associated with mitochondrial leukoencephalopathy associated with mitochondrial respiratory chain complex IV deficiency. The authors identified the COA7 mutations as responsible for import failure, which leads to retention of the newly synthesized protein in the cytoplasm. However, the protein does not accumulate in this compartment because is actively degraded by the proteasome. The authors elegantly show that overexpression of mutant COA7 or inhibition of the proteasome with either MG132 or other clinically approved proteasome inhibitors restores COA7 levels in mitochondria and also its activity in complex IV assembly. Therefore, the authors suggest that proteasome inhibition may be a new venue to combat mitochondrial diseases associated with poor mitochondrial protein import or excessive degradation by the proteasome.

A: The reviewer is correct to point out a decrease of wild-type MIA40 in the CRISPR clone (figure 5D of the revised manuscript). The mutant allele of MIA40 lacks the CPC motif, which is responsible for substrate binding and oxidation. In mammalian cells MIA40 is imported to mitochondria via the MIA pathway. We suspect that possessing one allele of protein, which cannot actively support MIA pathway, renders entire pathway less effective. In this sense the mutant allele would exert a dominant negative effect. We have introduced this information in the Results section when we refer to
The manuscript is technically and conceptually sound, and appropriate for publication in EMBO Molecular Medicine.
I have only two requests to improve the manuscript: 1-The experiments presented do not completely exclude the possibility that mutant RESA1 is not degraded by mitochondrial proteases. The authors should silence some of these proteases (e.g. AAA proteases) and test whether the protein is still degraded with a similar efficiency. Fig EV3B). Silencing of YME1L was only partial so perhaps with more efficient approach we could observe more significant effects on RESA1/COA7 mutant levels. These data suggest that YME1L could be a protease, which degrades RESA1/COA7 in the IMS.

A: In original manuscript we omitted the role of mitochondrial proteases in degradation of RESA1/COA7. In order to address this issue we performed silencing of YME1L in patient fibroblast. We observed a tendency towards increased levels of both mutated forms of RESA1/COA7 whenever YME1L was silenced (
2-The authors suggest that clinically approved proteasome inhibitors, such as bortezomib, may be applied as therapeutic agents to combat at least a subset of mitochondrial disorders. The authors should discuss the potential side effects on mitochondria and other organelles.

A: We introduced into the discussion a comment on potential side effects of bortezomib treatment on mitochondria.
Minor points: 1-Page 10: "TIMM8A (CX9C) and COX19 (CX3C)" should be "TIMM8A (CX3C) and COX19 (CX9C)" A: These mistakes were corrected in the revised manuscript.
2-I strongly suggest the authors to use COA7 to refer to the protein. The use of alternative names only serves to confuse the literature. A: In the revised manuscript we refer to the protein as COA7.
Referee #3 (Remarks for Author): This manuscript reports a careful analysis of the biogenesis of the mitochondrial respiratory chain assembly factor 1 (RESA1). RESA1 has been linked to mitochondrial leukoencephalopathy and complex IV deficiency. The authors show that RESA1, which contains 13 cysteine residues, is an unusual substrate of the mitochondrial intermembrane space assembly (MIA) system. In a remarkably complete characterization, they elucidated the molecular mechanisms of import of RESA1 into the mitochondrial intermembrane space, the interaction with Mia40 and disulfide bond formation. Importantly, the mitochondrial import of pathogenic mutant versions of RESA1 is slower than that of wild-type RESA1 and proteins accumulating in the cytosol are degraded by the proteasome. Using patient-derived fibroblasts, the authors discovered that inhibition of the proteasome rescued the localization of the mutant RESA1 to mitochondria and the activity of complex IV.
This paper by leading experts of the field is of technically very high quality and written very well. It provides exciting novel findings on the role of the proteasome in the pathogenesis of mitochondrial diseases and opens the way for new therapeutic approaches by using clinically approved proteasome inhibitors.
I have only a few minor comments on this exciting paper.
1. The authors provide a complete characterization of the biogenesis of wild-type and mutant RESA1 with important medical implications. It would be helpful for the general readership to present a cartoon of the import pathway of RESA1 and the role of the proteasome, e.g. in the last figure. Fig. 10. Thank you for the submission of your revised manuscript to EMBO Molecular Medicine. We have now received the enclosed reports from the referees that were asked to re-assess it. As you will see the reviewers are now globally supportive and I am pleased to inform you that we will be able to accept your manuscript pending minor editorial amendments. Any descriptions too long for the figure legend should be included in the methods section and/or with the source data.

Suggestions for corrections (indicated in CAPITAL letters
In the pink boxes below, please ensure that the answers to the following questions are reported in the manuscript itself. Every question should be answered. If the question is not relevant to your research, please write NA (non applicable). We encourage you to include a specific subsection in the methods section for statistics, reagents, animal models and human subjects.
definitions of statistical methods and measures: a description of the sample collection allowing the reader to understand whether the samples represent technical or biological replicates (including how many animals, litters, cultures, etc.).
Please fill out these boxes # (Do not worry if you cannot see all your text once you press return) a specification of the experimental system investigated (eg cell line, species name).

B-Statistics and general methods
the assay(s) and method(s) used to carry out the reported observations and measurements an explicit mention of the biological and chemical entity(ies) that are being measured. an explicit mention of the biological and chemical entity(ies) that are altered/varied/perturbed in a controlled manner.

Data
the data were obtained and processed according to the field's best practice and are presented to reflect the results of the experiments in an accurate and unbiased manner. figure panels include only data points, measurements or observations that can be compared to each other in a scientifically meaningful way. graphs include clearly labeled error bars for independent experiments and sample sizes. Unless justified, error bars should not be shown for technical replicates. if n< 5, the individual data points from each experiment should be plotted and any statistical test employed should be justified the exact sample size (n) for each experimental group/condition, given as a number, not a range; Each figure caption should contain the following information, for each panel where they are relevant:

Captions
The data shown in figures should satisfy the following conditions: Source Data should be included to report the data underlying graphs. Please follow the guidelines set out in the author ship guidelines on Data Presentation.

YOU MUST COMPLETE ALL CELLS WITH A PINK BACKGROUND #
We did not use any statistical test to predeterminate the sample size. The sample size was chosen on the basis of our experience and good laboratory practice. Yes, we present data with the Standard Error of the Mean (SEM) or the Standard Deviation (SD).