Characterisation of Colorectal Cancer Cell Lines through Proteomic Profiling of Their Extracellular Vesicles

Round 1

Reviewer 1 Report

In this manuscript the authors aimed to understand whether the proteomic profile of small EVs could be used to backtrack the pathological state of the cell of origin. This proof-of concept was pursued with the final goal of understanding the feasibility of profiling EVs in liquid biopsy to monitor the phenotype of colorectal tumours in a clinical setting. For that the authors isolated small EVs from 3 distinct colorectal cancer cell lines and compared their proteomic profiles. The manuscript is very well written, organized and the data is technically sound. Nevertheless, the reviewer argues regarding the novelty of the presented data as it is (see Q2) and add additional points which believes that would be important to address in this manuscript:

 

Q1: No rational was provided for using the indicated 3 cell lines (HCT116, HT29 and SW620) in the presented study nor any relationship on the proteomic findings related to the cells’ phenotype (clinical staging that cell line represents, ability to invade/metastasis, mutational status of key genes involved in CRC progression, etc..). This would be of added novelty comparatively to similar works and should be consistently added throughout the manuscript.    

 

Q2: One important piece of literature on this topic, which was missing in the authors reference list, has been published by David Leyden’s group two years ago (Reference: Hoshino A et al. Extracellular Vesicle and Particle Biomarkers Define Multiple Human Cancers. Cell. 2020 Aug 20;182(4):1044-1061.e18. doi: 10.1016/j.cell.2020.07.009. Epub 2020 Aug 13. PMID: 32795414; PMCID: PMC7522766.). This work includes the proteomic comparison of more than 497 normal and cancer-associated human and murine-derived samples, including cell lines, tissues, plasma, and other bodily fluids. Importantly, these include proteomic analysis from CRC HCT116, SW620 cell lines which could be of interest for the authors to compare and discuss in the light of their work.

 

Q3: Since the aim of this study was to examine to which extent the proteomic content of small EVs reflect their parental cell the reviewer was surprised by the absence of phospho-proteomic analysis of small EVs from the distinct cell lines to check whether it matches the native cell or not (addressing on the implication to serve as biomarker in a clinical setting).

 

Q4: Considering the intended use of liquid biopsy-based small EVs for clinical monitoring of CRC progression and response to therapy the reviewer believes that would be important to either demonstrate the feasibility of detecting these specific CRC-derived small EVs in the whole plasmatic EVs by proteomic (or phospho-proteomic) approach or even discuss which would be the difficulties of this technical approach and strategies to overcome it. Do the authors believe that such alterations in the proteome of EVs from CRC could be observed/detected in whole plasma samples?

 

Minor Q5: In figure 4 the authors should explain why the TGFB1 is elevated in EVs from HCT116 while the TGFBR2 is completely absent. Is there any relation with the phenotype or stage of the cell line in question? Could the authors elaborate an explanation on this? How this would impact CRC progression?       

Minor Q6: Page 14 line 473: The authors should not address to the absence of a given protein in proteomics data as their absence from the biological sample due to recognized low dynamic range of proteomics technique for detecting low expressing proteins. This would be valid for several parts of the discussion/results section.  

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

Heck et al present in the manuscript “Characterisation of colorectal cancer cell lines through proteomic profiling of their extracellular vesicles” their proteomic study, in which they characterize whole cell lysates and extracellular vehicles (EV) of three colorectal cancer cell lines by a label free shotgun experiment. Although the paper has some interesting aspects, the study lacks validation and a statistical analysis, and should not be considered for publication at this point.

 

 

·         The paper solely relies on cell lines, but the authors claim its data is useful for biomarker discovery. For these purposes a comparison with healthy tissue needs to be performed.

·         The authors should discuss why they chose the selected cell lines, and what the significance of the differences in their proteomes is.

·         The mass spectrometry method, which the authors chose to use, is prone to missing values which is not addressed in the discussion. Furthermore, missing values seem to be considered as an expression of zero, which is not always the case. Validation of the results by targeted mass spectrometry or western blots should be incorporated.

·         The Western blot in Figure 1B and supplementary Figure 1 shows that for the whole cell lysate the bands for ALIX and TSG101 and GM130 originate from different membranes. This fact should be made clear in the main figure, by showing the original ALIX band separately.

·         In Figure 2 the authors perform a GO term enrichment analysis and conclude: “Investigation of Gene Ontology (GO) annotations of the 531 proteins observed in sEVs from all three cell lines (Table S3) showed that the clusters with the highest enrichment score encompassed GO terms associated with cell adhesion, cell translation, GTP/GDP signal transduction activity as well as the Wnt, TNF- β, MAPK and NIK/NF-κB signalling pathways”. The 21 proteins that were assigned to the NIK/NF-κB signalling are all constituents of the proteasome. Although the proteasome is an important part of this pathway, no protein was detected that is involved in the direct signaling cascade, and the results of the GO Term analysis should be interpreted more cautiously. Similarly, the proteins that are annotated for the WNT pathway are also proteasomal proteins, but they do not incorporate the core pathway. All described enriched pathways should be reevaluated more carefully and the GO Terms should be used as a starting point of the analysis and no its result.

·         The authors exclusively concentrate on proteins that are absent or present in EVs or cells, but do not account for proteins, that are differentially expressed. Volcano plots should be incorporated. In addition, a statistical analysis needs to be performed to fully understand the differences between the proteomes.

·         Subfigure 4c shows the Expression of TGF1B and TGFBR2. From table S2 it is clear that this quantitation is based on 1 peptide each, which covers 2% and 3% of the protein sequence. TGFB1 was detected in exactly in one out of three replicates in HT29, but the authors decided to report a value and not to include an error bar or any indication of missing values. The underlying data for this subfigure is extremely weak, and a require a validation by other methods such as Western Blot.

·         Similarly, it is not clear from the whole Figure 4 how significant the presented data is. No statistical evaluation of the protein expression was performed. Some proteins have high expression in one replicate only.

·         The phosphoproteomics analysis does not contain any validation of targets by western blots or other methods. The statistical analysis is missing. The phosphopeptide quantification seems to vary significantly between replicates. Due to the label free method that was used, these changes could be completely arbitrary.

·         It is not clear if the phoshopeptide intensities were normalized to the unphosphorylated protein abundance.

 

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Authors performed the asked alterations to the text. In some cases the answers are enough to suffice the reviewers concerns (although the answers were not complete, critical in nature or particularly thorough e.g. Q2). Taken together, the manuscript fulfils the minimum requisites to be published in Proteomes journal.      

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

The authors improved the manuscript by describing the limitations of the study and clarifying several points. Most of my questions and concerns were resolved.

However, the authors did not address one of my main critiques of the study:

A validation of key results by an independent method should be performed. The authors describe in the material and method parts Western Blots that use phospo-ERK antibodies, which indicates that such experiments were performed and images of those membranes should be included in the publication.

 

Author Response

Please see the attachment

Author Response File: Author Response.pdf