Fatty Acid Unsaturation Degree of Plasma Exosomes in Colorectal Cancer Patients: A Promising Biomarker

Abstract

Even though colorectal cancer (CRC) is one of the most preventable cancers, it is currently one of the deadliest. Worryingly, incidence in people <50 years has increased unexpectedly, and for unknown causes, despite the successful implementation of screening programs in the population aged >50 years. Thus, there is a need to improve early diagnosis detection strategies by identifying more precise biomarkers. In this scenario, the analysis of exosomes is given considerable attention. Previously, we demonstrated the exosome lipidome was able to classify CRC cell lines according to their malignancy. Herein, we investigated the use of the lipidome of plasma extracellular vesicles as a potential source of non-invasive biomarkers for CRC. A plasma exosome-enriched fraction was analyzed from patients undergoing colonoscopic procedure. Patients were divided into a healthy group and four pathological groups (patients with hyperplastic polyps; adenomatous polyps; invasive neoplasia (CRC patients); or hereditary non-polyposis CRC. The results showed a shift from 34:1- to 38:4-containing species in the pathological groups. We demonstrate that the ratio Σ34:1-containing species/Σ38:4-containing species has the potential to discriminate between healthy and pathological patients. Altogether, the results reinforce the utility of plasma exosome lipid fingerprint to provide new non-invasive biomarkers in a clinical context.

Keywords: colorectal cancer; exosomes; lipidome; monounsaturated fatty acids; polyunsaturated fatty acids.

Figure 1

Analysis of the main membrane lipid classes of patients’ plasma derived EVs. The bars represent healthy (n = 13), with hyperplastic polyps (HP, n = 5), with adenomatous polyps (AD, n = 16), hereditary non-polyposis colorectal cancer or Lynch syndrome (Her, n = 9), and carcinoma (Neo, n = 19). Values are expressed as % of total membrane lipids (mean ± SEM). Statistical differences were assessed by one-way ANOVA followed by a Bonferroni post-test. * p < 0.05; ** p < 0.01. For simplicity, only statistical differences compared to healthy patients are represented. Detailed results showing all comparisons are included in Supplementary Table S2. Cer: ceramides, PC: phosphatidylcholine, PE: phosphatidylethanolamine, PS: phosphatidylserine, PI: phosphatidylinositol, LPC: lysophosphatidylcholine, SM: sphingomyelin.

Figure 2

Lipid species composition of EVs isolated from patient plasma. (a) Phosphophatidylcholine (PC) species; (b) Phosphatidylethanolamine (PE) species; (c) Phosphatidylinositol (PI) species; (d) LysoPC (LPC) species; (e) Sphingomyelin (SM) species; (f) Ceramide (Cer) species. For simplicity only the species accounting for more than 5% of total species and the major AA-containing species were included. The bars represent healthy patients (n = 13), patients with hyperplastic polyps (HP, n = 5), patients with adenomatous polyps (AD, n = 16), patients of hereditary non-polyposis colorectal cancer (Her, n = 9), and carcinoma patients (Neo, n = 19). Values are expressed as % of total phospholipid or sphingolipid class (mean ± SEM). Statistical differences were assessed by one-way ANOVA followed by a Bonferroni post-test. * p < 0.05; ** p < 0.01; *** p < 0.001. For simplicity, only statistical differences compared to healthy patients are represented. Detailed results showing all comparisons are included in Supplementary Table S3. Cer: ceramides, PC: phosphatidylcholine, PE: phosphatidylethanolamine, PS: phosphatidylserine, PI: phosphatidylinositol, LPC: lysophosphatidylcholine, SM: sphingomyelin.

Figure 3

Assessment of the discriminatory capability of the ratio between the summation of 34:1 species and the summation of 38:4 species between healthy and pathological groups. The ratio was calculated by dividing the total content of 34:1 and 38:4 in PC, PE, and PI. Values are expressed mean ± SEM, n = 13 for healthy group, n = 5 for HP, n = 16 for AD, n = 9 for Her, and n = 19 for Neo. Statistical differences were assessed by one-way ANOVA followed by a Bonferroni post-test. ** p < 0.01; *** p < 0.001.