Write a comment

Otto et al. have developed a proximity labelling method called Organelle Contact-dependent Affinity Purification (ORCA) to systematically separate subpopulations of mitochondria based on their membrane contact site formation with other organelles. The method relies on specific biotinylation of a mitochodndrial subpopulation through a BirA expressed on the surface of the target organelle and an AviTag localised to the surface of mitochondria, rather than on APEX2-based labelling which has been used for similar purposes in the past (Cunningham et al, 2025). Using ORCA, the authors identify a novel Golgi-associated subpopulation of mitochondria in U2OS cells. This subpopulation has high levels of gene expression factors and translation activity, which seems to be stimulated by the thus far uncharacterised peripheral Golgi protein KIAA0930, renamed to GMO1 in this study.

Strength:

The manuscript is generally easy to follow, the introduction motivates the study in an excellent manner, and experimentally ORCA provides a useful tool for identifying organellar sub-populations. The authors thoroughly demonstrate that the method reliably enriches for mitochondrial proteins, and that co-immunoprecipitated non-mitochondrial proteins can mostly be attributed to the specific contact site under investigation. The discovery of translation-specialised mitochondria associated with the Golgi is a convincing proof of principle of ORCA and a very exciting foundation for studying the significance of this mitochondrial subpopulation further.

Major points:

1)

It is very intriguing that ORCA enriches not only for mitochondrial proteins but also for specific proteins of the BirA-targeted organelle. It would be very useful if the authors carried out some basic analysis of these non-mitochondrial proteins in order to better understand what kind of contact site proteins are enriched by ORCA. In the results it is stated that ORCA “captures cytosol-facing proteins associated with mitochondria at organelle contact sites” but there is no evidence provided for this in the data presented.

As the ER luminal protein Calreticulin and the peroxisomal transmembrane protein Pex14 can be detected in the immunoprecipitated samples in BirA-cytosol cells (Figure 1C), it is plausible that mitochondria-associated microsomes and even whole peroxisomes may be co-immunoprecipitated. An enrichment analysis comparing luminal proteins, transmembrane proteins, peripheral membrane proteins and/or known mitochondrial membrane contact site proteins over background for each organelle could be carried out to clarify this, similar to the sub-mitochondrial protein enrichment analysis in Figure 1E.

2)

The number of cells analysed in Figure 3 is very inconsistent. While 16 cells have been quantified in B, only 7 cells were quantified in D and 4 cells in F. The authors should quantify a similar number of cells for each stain for comparable statistical analysis.

3)

The authors find that GMO1 knockout leads to a reduction in OXPHOS complexes and results in lower respiratory efficiency as well as loss of cristae. These results are derived from global mitochondrial analysis. It would be interesting and in line with the study’s aims to investigate whether the defects are observed due to a specific impact on the Golgi-associated mitochondrial subpopulation or whether all mitochondria are affected. In addition, GMO1 is referred to as a regulator of mitochondrial translation, but from the experimental evidence presented, the role of GMO1 as a tether or as a regulator at Golgi-mitochondrial contact sites cannot be dissected. To address these questions, we suggest the following experiments:

· Analysis of Golgi-mitochondria contact sites in fluorescent or electron microscopy in GMO1 KO cells, and/or ORCA of mitochondria in BirA-Golgi GMO1 KO cells to assess whether contacts between mitochondria and Golgi are lost in the absence of GMO1

· Overexpression of an artificial or natural Golgi-mitochondria tether to test whether this leads to an increase in translation factors and translation of OXPHOS proteins in global mitochondria or in the Golgi-associated subpopulation using ORCA

· ORCA of mitochondria in BirA-peroxisome/BirA-ER/BirA-lysosome GMO1 KO cells to determine whether other mitochondrial subpopulations show a reduction in OXPHOS complexes or have an otherwise altered proteome. This would provide information on the dynamics of mitochondrial subpopulations and show whether the translation-specialised Golgi-associated mitochondria fulfil a specific function at the Golgi or whether they represent a Golgi-regulated translation factory from which OXPHOS proteins are distributed to other mitochondrial subpopulations

Minor points:

4)

An explanation of the rationale behind specifically choosing Golgi-BirA with peroxisome-BirA and ER-BirA with lysosome-BirA for pairwise comparison would be helpful.

5)

Some Figures could be more reader-friendly.

· Larger size of the images in Figures 1A, 1F, 5 and 5M would be appreciated.

· In Figure 5M and Figure S5J the colouring of the mitochondria prevents analysis of the membrane ultrastructure, perhaps marking only the outlines of mitochondria may improve visualisation

· A more intuitive representation of the GOEA data in Figures 1F, 2E and 2H would be helpful. The representation of enrichment through colour as well as size is confusing.

6)

In Figure 2D it should be clarified whether proteins annotated as dual mito/peroxisome localising are experimentally proven to localise to both organelles or whether putative dual localisation proteins are included in the annotation. Furthermore, the authors may want to discuss the dual localisation outlier MGST1 enriched in the Golgi-BirA sample.

7)

In Figure 3 A, C and E the choice of the different organelle markers could be explained and the insets in each image labelled.

8)

We found that it would improve reading flow if Figures S2 B, C, D, F and H were included in the main text rather than in supplements, as they are quite integral to the validation of the method and represent some of the main findings with regards to the ER and lysosome-associated mitochondrial population.

9)

In Figure 4D it is not very easy to judge the level of GMO1 reduction upon BrefA treatment without a quantification of the immunodetected signal.

10)

The authors should comment on the choice of rescue with GMO1 isoform 2 over other isoforms in Figure 5, especially as isoform 1 was used in Figure 4.

11)

Relating to GMO1 knockdown in the indirect flight muscles of Drosophila, a more extensive description of the phenotype and differences to U2OS cells would enhance the manuscript. In Figure S5J, the cristae still appear quite tightly packed in the mitochondria of Drosophila cells, which would differ from the U2OS cells. Also, the authors could describe whether they observed a physical phenotype in the animals given that OXPHOS may be compromised in their indirect flight muscles.

Reporting issues

The authors should consider making the mass spectrometry data openly available in a repository.

Competing interests

The authors declare that they have no competing interests.

Use of Artificial Intelligence (AI)

The authors declare that they did not use generative AI to come up with new ideas for their review.