A Plasma Exosomes-based Novel Screening Tool For Colorectal Cancer Revealed By Shotgun And Data-independent Acquisition Mass Spectrometry

Part 1: Proteomics and phosphoproteomics based on TMT-labeled LC-MS/MSBackground:Early screening for colorectal cancer(CRC)is essential to improve its prognosis.Liquid biopsies are increasingly being considered for diagnosing cancer due to low invasiveness and high reproducibility.In addition,plasma exosomes expressing tumor-specific proteins are potential biomarkers for various cancers.Methods:The discovery cohort(cohort 1)consisted of 10 healthy individuals(group N),10 CRC patients without liver metastasis(group Ca)and 10 CRC patients with livermetastases(group M).The plasma samples from all individuals from a group are pooled together for TMT-based LC-MS/MS.Exosomes,the most acknowledged small extracellular vesicles(EVs)were isolated from plasma of cohort 1 by overnight ultracentrifugation coupled with sucrose density gradient centrifugation.The morphology of exosomes was visualized by TEM,and further characterized by high surface expression of known markers.The diameters of the exosomes were measured by Nano sight technology.Tumor-specific plasma exosomes proteins were prioritized using Tandem Mass Tag(TMT)-based shotgun proteomics and phosphoproteomics,combining with the biological characteristics of the differential proteins.Results:The EVs isolated from plasma samples in cohort 1 were identified as plasma exosomes.A total of 545 Swiss Prot proteins were identified,and expression levels of 462 proteins were obtained.In addition,we also identified 252 unique phospho-peptides and 347 phosphorylation sites from 195 proteins,of which 315 sites from 177 proteins were quantified.Gene Ontology(GO)analysis showed that many EV proteins are involved in cell-cell communication,metabolic processes,stimuli responses,and biogenesis.Cluster analysis showed variations in protein expression across the three groups,with significant differences in fibrinogen α chain(FGA),fibronectin(FN)1,S100A9 and haptoglobin(HP)expression levels,as well as the phosphorylation sites.Conclusions:The proteome and phosphoproteome between healthy individuals and CRC patients are significantly different.The expression levels of FGA,FN1,S100A9 and HP are associate with CRC occurrence.Part 2: Validation of differential proteins and the candidate biomarkerPurpose:Validate the differential proteins in plasma exosomes chosen above using independent cohorts.Pick the most differential and potential plasma exosomal protein as a candidate biomarker and explore the reason of its change.Methods:Exosomes were isolated from plasma of cohort 2 and 3 by overnight ultracentrifugation coupled with sucrose density gradient centrifugation.We validated the expression levels of FN1,HP,S100A9 and FGA in cohort 2 consisting of 11 healthy donors(group N),11 CRC patients without liver metastases(group Ca),and 6 CRC patients with liver metastases(group M).FGA was picked as a candidate biomarker and validated in an enlarged cohort 3 consisting of 19 healthy donors(group N),19 CRC patients without liver metastases(group Ca),20 adenoma patients(group A)and 3 CRC patients with liver metastases(group M).We evaluated the presence of the FGA-overexpressing plasma exosomes in the culture supernatants of two CRC cell lines(SW620 and Lo Vo)and a mouse embryonic fibroblast cell line(NIH 3T3),as well as in the plasma of a tumor-bearing mouse model.Samples were collected from the mice before and one week after subcutaneously injecting them with SW620(n=5)or Lo Vo(n=4)cells.Results: FN1,HP,S100A9 and FGA all showed significant differences in cohort 2,of whichFGA was the most dominant distinguishing biomarker.In enlarged cohort 3,All CRC-exosomes samples(regardless of metastasis)showed FGA overexpression,as did 65% of the adenoma samples.However,FGA-overexpressed plasma exosomes could only detect part of adenoma but failed to tell the liver metastases of CRC.Compared to NIH 3T3 derived exosomes,SW620 derived and Lo Vo derived exosomes have higher expression of FGA.Plasma exosomal FGA also increased significantly in the respective samples after in vitro culture and subcutaneous tumor growth SW620(n=5),Lo Vo(n=4).Conclusion: Validated by molecular experiments,FN1,HP,S100A9 and FGA all showed significant differences between healthy individuals and CRC patients,of which FGA was the most dominant distinguishing biomarker.The presence of the FGA-overexpressing exosomes in the culture supernatants of two CRC cell lines(SW620 and Lo Vo)and a mouse embryonic fibroblast cell line(NIH 3T3)as well as a tumor-bearing mouse model indicated that FGA highly expressed plasma exosomes are likely released from the parent CRC cells,and may contribute to primary tumorigenesis.Thus,FGA+ plasma exosomes was picked as a candidate biomarker for CRC in the following study.Part 3: DIA-MS Optimization and validation for CRC screeningPurpose: To establish a DIA-MS-based diagnostic method for liquid biopsies,which can be potentially adapted as a rapid and non-invasive screening tool to detect CRC in early stages.Meanwhile,validate the tool by PRM and confirm the need to isolate plasma exosomes.Methods: The validation cohort(cohort 4)consisted of 12 healthy individuals(group N),13 adenoma patients(group A)and 12 CRC patients without liver metastasis(group Ca).An initial DIA-MS run on 20 plasma samples(n= 6,7,7)was used to test whether the differential FGA level appear without the isolation of plasma exosomes.Exosomes were isolated from plasma of cohort 4 by overnight ultracentrifugation coupled with sucrose density gradient centrifugation.S6 sample was analyzed by DIA-MS using four different LC gradients of 20 min,45 min,60 min and 90 min.Both the 20 min and 90 min conditions were tested on samples S1 to S10(three each from healthy individuals and adenoma patients,and four from CRC patients without liver metastases).After we have systematically performed the experiment by comparing results using 90 min,60 min,45 min,and 20 min LC gradient,20 min LC gradient was used to determine the utility of DIA-MS detected plasma exosomal FGA in the clinical screening of colon adenoma and CRC in cohort 4.Finally,the result of DIA-MS was validated by PRM-MS.Results: We found considerable overlap between the peptides and proteins detected by thefour conditions,with no significant difference in the numbers of peptides or protein groups(Figure 4B;p>0.05).It was discovered that by reducing the LC gradient from 90 min to 20 min,we could still detect 87% of the protein groups(353/406).Furthermore,the small number of proteins not detected using the 20 min LC gradient did not affect the clinical usage of the data.Fast DIA-MS showed significant differences between the N and A groups,as well as between the N and Ca groups,but not between the two patient groups.It was confirmed that highly expressed FGA in plasma exosomes could distinguish colon adenoma with an area of curve(AUC)in the receiver operating characteristic(ROC)curve of 0.949 and CRC patients(AUC of ROC is 1.000)from healthy individuals.The performance outperformed conventional tumor biomarkers.The DIA-MS quantification of highly expressed FGA in plasma exosomes among three groups agreed with that from PRM-MS.Conclusion: Optimized DIA-MS detection of highly expressed FGA in plasma exosomes is a potential non-invasive screening tool to identify early stage CRC,which could distinguish colon adenoma and CRC from healthy individuals with better performance than conventional tumor biomarkers.Meanwhile,the isolation of plasma exosomes is a necessary step before the analysis.