Blood-derived non-extracellular vesicle proteins as potential biomarkers for the diagnosis of early ER+ breast cancer and detection of lymph node involvement

Extracellular vesicles (EV’s) are membrane surrounded structures released by different cell types and are emerging as potential therapeutic and diagnostic targets in cancer. In the present study, plasma samples derived from 7 patients with metastatic and non-metastatic ER+ (estrogen receptor positive) breast cancer (BC) were collected and their respective (EVs) isolated and the protein content analyzed by mass spectrometry and FunRich analysis. Two putative plasma biomarkers (absent in healthy controls samples) were identified which could be used to detect early ER+ breast cancer and for those with lymph node (LN) involvement However, given the current limitations of the EV isolation method used, it is possible that these biomarkers did not originate from EVs and may represent blood-derived extracellular proteins. Further work in a larger patient cohort is warranted to confirm these findings and examine the diagnostic potential of these biomarkers.

Extracellular vesicles (EVs) are membrane surrounded structures released by different cell types that are involved in cellular communication and are emerging as potential therapeutic and diagnostic targets in cancer 1 as in the case of early pancreatic cancer 2 .
EVs can be classified in several subtypes based on their size, shape, and supposed origin. Exosomes are defined as ∼30-100 nm vesicles which originate from multivesicular bodies (MVB) and contain late endosomal markers 3,4 , although biochemically indistinguishable vesicles can bud directly from the plasma membrane 3,5 . Microvesicles or shedding vesicles are generally larger (>200 nm), are more variable in shape and density, and likely originate from the plasma membrane 4,6,7 . EVs may contain proteins, lipids, and RNAs, however how these components are sorted into EVs remains unclear.
Tumor-derived EVs are also critical components for preparing the tumor microenvironment because they enable tumor cells to escape from the immunological surveillance 8 and help in the setting of a pre-metastatic niche for the engraftment of detached cancer cells 9 . Both exosomes and MVs have been extensively studied and attributed various important physiological roles in cancer 10,11 . For instance, EVs have been found to play an important role in every phase of cancer development from cancer initiation, invasion and metastasis 12 . For these reasons, EVs are potential therapeutic and diagnostic targets in cancer and EV-derived biomarkers maybe useful for predicting future metastatic development and identify metastasis sites 13 . ER+ (estrogen receptor positive) breast cancer (BC) represents 60-80% of all BC cases 14,15 . Here we describe our preliminary findings exploring the role of tumour derived EVs biomarkers that could ultimately be used as part of a test kit for the detection of early ER+ BC and lymph node involvement.

Samples
Plasma samples from 4 control patients (2 adult women and 2 men) which were confirmed as not having any form of BC, ER+ BC metastases, BC1 and BC2 explants EVs, SKBC and parental BC (Lyden lab, WCM, USA). Samples CF37, CF5, CF1, CF25, CF33, CF27 and CF110 and C7 (female control plasma sample) were collected at Champalimaud Clinical Centre, Portugal, as part of a study on the role of tumor-derived microvesicles and bone marrow progenitor cells as diagnostic and prognostic biomarkers in advanced BC and inflammatory BC Patients (RECI/ BIM-ONC/0201/2012, FCT, Portugal). ER+ BC patient samples were selected based on their stage of disease progression -confirmed by CT-scan and surgery. EVs derived from conditioned media of cells lines SKBr3, MCF7, MDA468, MDA231 and MCF10A were also used in this study (details about these samples can be found in Table 1).

Ethics approval and informed consent
This study was approved by an Ethics Review Board at Champalimaud Foundation, Portugal. All study patients provided their written, informed consent.
EV purification and analysis EV purification and analysis were performed at the Lyden lab (WCM) accordingly to Andre et al., 2016 16 . Briefly, plasma was pelleted at 500 × g for 10 min, then the supernatant was centrifuged at 20,000 × g for 20 min. Exosomes were then harvested by centrifugation at 100,000 × g for 70 min. The exosome pellet is resuspended in PBS and collected by ultracentrifugation at 100,000 × g for 70 min. The exosome pellet is resuspended in PBS and then stored at −80°C. The LM10 nanoparticle characterization system (NanoSight) equipped with a blue laser (405 nm) Proteomics and proteomic analysis Proteomic analysis was performed at the Rockefeller University, Proteomics Center as described in Hamidi et al., 2017 17 . Proteomic analysis was performed with the help of FunRich Program version 3. Only proteins with Mascot scores of approximately 90 or >90 were considered 18 .

Results and discussion
Clinical data on the EVs isolated from BC patient's plasma samples and cell lines can be found in Table 1. The method used for EV isolation also precipitates lipoproteins and immunocomplexes (IC) which are known possible contaminants 19 . However, samples submitted for mass spectrometry analysis showed none of the recognised contaminants of high speed centrifugation. In the two patients with early BC (Table 2a), we detected HCG1745306 isoform CRA-a, a protein from the family of alpha type haemoglobins and for the patient with lymph node involvement, we detected histone H1.2 (Table 2 a-b). HCG1745306 isoform CRA-a was only present in the two patients with early BC with Mascot scores of 3208.8 and 3966.5,

Amendments from Version 2
In this new version we better addressed the referees concerns by explaining that although we have identified two potential biomarkers possibly derived from EVs, our study does suffer from a number of recognised limitations. Firstly, ultracentrifugation is insufficient to purify EVs from other contaminants. Secondly, given the size of the particles we isolated possibly they correspond to low-density lipoproteins which have the same size as EVs. Moreover, it is unlikely that EVs would contain a histone, which are normally confined to DNA in the nucleus. However, the presence of DNA in EVs was claimed by electron microscopy (EM) though the EM image is not of sufficient magnification to allow for an accurate morphologic analysis and may simply represent cellular debris or apoptotic bodies or even unspecific staining. Additionally, it is also unlikely HCG1745306 isoform CRA-a, would be present in EVs and it may simply be a precipitant similar to the α-globin seen in β-thalassemia. Therefore our current data does not support the idea that these biomarkers derived from EVs and could in fact be blood-derived extracellular proteins and for these reasons we changed the title to "Blood-derived non-extracellular vesicle proteins as potential biomarkers for the diagnosis of early ER+ breast cancer and detection of lymph node involvement". In order to support our arguments we provide additional NTA analysis files as well as Figure 1 showing a summary of our NTA results. Although we have identified two potential biomarkers possibly derived from EVs, our study does suffer from a number of recognised limitations. Firstly, ultracentrifugation is insufficient to purify EVs from other contaminants 29 . For example, co-isolation of high-density lipoprotein and other particles with EVs isolated from blood by density gradient centrifugation has been reported 29,30 suggesting that the biomarkers we identified might not be associated with EVs but with a constituent of another particle type such as a lipoprotein. Secondly, as mentioned above, exosomes are defined as ∼30-100 nm vesicles that originate from MVB. In contrast, microvesicles or shedding vesicles are generally larger (>200 nm), more variable in shape and density and arise from the plasma membrane. The size of the particles we isolated ranged from 76.7-213.4 and 73.8-192.3 nm, for samples CF5 and CF37, respectively and for all the samples between 12.3-298.4nm ( Figure 1 and original NTA files) possibly correspond to lowdensity lipoproteins which have the same size as EVs 31 . Moreover, it is unlikely that EVs would contain a histone (which are normally confined to DNA in the nucleus). However, Thakur et al., claim to have identified genomic DNA in EVs by electron microscopy (EM) though the EM image is not of sufficient magnification to allow for an accurate morphologic analysis and may simply represent cellular debris or apoptotic bodies or even unspecific staining 32,33 . Additionally, it is also unlikely HCG1745306 isoform CRA-a, would be present in EVs and it may simply be a precipitant similar to the α-globin seen in β-thalassemia 34 . Therefore our current data does not support the idea that these biomarkers derived from EVs and could in fact be blood-derived extracellular proteins.

REVISED
Nevertheless, a strength of our study is that samples were drawn from those with confirmed non-metastatic and metastatic disease at different sites and so are likely to be representative patients.

Conclusion
In conclusion, our observations suggest the possibility that HCG1745306 isoform CRA-a, and histone H1.2, irrespective of their origin, could represent potential biomarkers for the detection of early ER+ BC. Further work in a larger cohort of patients is clearly needed to confirm these initial findings.

Competing interests
No competing interests were disclosed. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. While the results are clearly presented, the interpretation that these are EV-based biomarkers is not supported. The EV isolation method used is insufficient to purify EVs from other contaminants. Furthermore, it is not intuitive that EVs would be likely to contain a histone (normally confined to the nucleus) or hemoglobin. The authors' discussion of this major concern is insufficient ("However, samples submitted for mass spectrometry analysis showed none of the recognised contaminants and it was therefore concluded that the main EV type present in these samples is the MV.").

Grant information
The EV literature is quite confusing due, in part, to the over-interpretation of observations from samples prepared using inadequate isolation methods. Therefore, in it's current version, I do not recommend indexing of this report. Ideally, the authors should repeat the work with approaches that better enrich for EVs over non-EV protein contaminants (e.g. density-gradient based ultracentrifugation or immuno-affinity purification) and protease protection assays in the presence/absence of detergent should be used evaluate if the proteins are indeed associated with EVs and if they are on the surface or interior of vesicles. If further experiments are not possible, I recommend revising the paper to remove the unsupported conclusion that the proteins identified are EV-associated. (For example the title could be: Blood-derived extracellular proteins as potential biomarkers for the diagnosis of early ER+ breast cancer and detection of lymph node involvement).
For reference, I direct the authors to an updated publication prepared by members of the EV community on the minimal information for studies of EVs in the Journal of Extracellular Vesicles .

I cannot comment. A qualified statistician is required.
Are all the source data underlying the results available to ensure full reproducibility? Yes Are the conclusions drawn adequately supported by the results? Yes No competing interests were disclosed. Competing Interests: I have read this submission. I believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard.

The benefits of publishing with F1000Research:
Your article is published within days, with no editorial bias You can publish traditional articles, null/negative results, case reports, data notes and more The peer review process is transparent and collaborative Your article is indexed in PubMed after passing peer review Dedicated customer support at every stage For pre-submission enquiries, contact research@f1000.com