Extracellular vesicles derived microRNAs as non-invasive markers of liver fibrosis in chronically infected HCV patients: a pilot study

Abstract

Extracellular vesicles (EVs) are an increasingly promising tool for liquid biopsy in liver diseases. Hepatitis C Virus (HCV) infection, alone or together with Human Immunodeficiency Virus (HIV) infection significantly impacts on the microRNA (miRNA) EVs content resembling chronic hepatitis C (CHC) progression. The objective of the study was to delve into the intricate EVs-miRNA profiles in CHC patients with different liver fibrosis stages, aiming to pinpoint non-invasive markers capable of distinguishing significant fibrosis. Plasma EV-miRNAs from 50 CHC patients (HCV+ and HCV+/HIV+) stratified in no significant (F < 2) and significant (F ≥ 2) fibrosis, were massively sequenced. General linear models (GLM) were used to identify significantly differential expressed (SDE) miRNAs according to liver fibrosis stages (F ≥ 2 and F < 2). Dysregulated biological pathways were subsequently analyzed in silico for the following groups: i) all patients; ii) HCV+; and iii) HCV+/HIV+. Multiple-ordered logistic regression analysis was performed to develop a score to identify F ≥ 2 cases. The diagnostic potential of both the SDE miRNAs and the developed score was assessed using ROC curve analysis. With respect to all CHC patients, two SDE miRNAs (hsa-miR-122-5p and hsa-miR-92a-3p) were identified which regulate genes related to cytoskeleton organization. Regarding their diagnostic performance to discriminate F ≥ 2, both miRNAs individually demonstrated acceptable diagnostic values. However, their combined use in a new score enhanced their diagnostic performance (AUROC = 0.833). In the HCV+ subgroup, 8 SDE miRNAs (hsa-miR-122-5p, hsa-miR-320c, hsa-miR-3615, hsa-miR-320a-3p, hsa-miR-374b-5p, hsa-let-7a-3p, hsa-miR-199a-5p, hsa-miR-142-5p), which regulate macrophage activity and cell growth/death regulation, were recognized. Among them, hsa-miR-3615 displayed the highest diagnostic performance to discriminate F ≥ 2 (AUROC = 0.936). With respect to HCV+/HIV+, 18 SDE miRNAs (hsa-miR-4508, hsa-miR-122-5p, hsa-miR-451a, hsa-miR-1290, hsa-miR-1246, hsa-miR-107, hsa-miR-15b-5p, hsa-miR-194-5p, hsa-miR-22-5p, hsa-miR-20b-5p, hsa-miR-142-5p, hsa-miR-328-3p, hsa-miR-335-3p, hsa-miR-125a-5p, hsa-miR-423-3p, hsa-let-7d-3p, hsa-miR-128-3p, hsa-miR-10a-5p) were recognized that regulate RNA silencing processes. In this case, hsa-miR-423-3p and hsa-miR-128-3p showed outstanding diagnostic performances (AUROC > 0.900). Distinct EVs-miRNA profiles were identified in patients with varying liver fibrosis stages, both in the overall CHC cohort and within HCV+ and HCV+/HIV+ subgroups. These specific miRNA signatures would allow the elucidation of potential mechanisms involved in clinical evolution and identification of specific biomarkers of unfavorable progression, plausible to be used in a diagnostic panel. Furthermore, the developed score demonstrates the ability to discriminate within the CHC group those individuals with significant fibrosis regardless of their HIV infection status.

Keywords: Biomarkers; Extracellular vesicles; Fibrosis; Hepatitis C infections; MicroRNAs.

Graphical abstract

Fig. 1

Analysis of EV-microRNA cargo in each group of patients. a) Schematic representation of RNA-seq analysis and results. b) Exploratory analysis. Multivariate analysis was performed by supervised partial least squares discriminant analysis (PLS-DA) from normalized log-transformed and scaled miRNA count per million (cpm) data.

Fig. 2

Analysis of EV-microRNA cargo in the CHC group and in HCV+ and HCV+/HIV+ subgroups. a) Volcano plot showing SDE miRNAs in F ≥ 2 patients. FDR, false discovery rate; log₂ FC, log₂ fold change. Red dots show miRNAs with a FDR ≤0.2 and a log₂FC ≥ 0.585 (equivalent to FC ≥ 1.5); blue dots show miRNAs with a FDR ≤0.2 and a log₂ FC < −0.585; grey dots show miRNAs without statistical significant different expression between fibrosis severity groups. b) Chord diagram of the top 25 target genes. Each targeted gene is represented by one different color, together with their corresponding interaction, while SDE miRNAs are shown on the left.

Fig. 3

Receiver operating characteristic curves analysis for discriminating F ≥ 2 cases. a) ROC curves of SDE miRNAs and the proposed new score in CHC patients. b) ROC curve of hsa-miR-3615 for HCV+ patients. c) ROC curves of hsa-miR-423-3p and hsa-miR-128-3p for HCV+/HIV+ patients. AUROC: area under the receiver operator characteristics curves; Se: Sensitivity; Sp: specificity; PPV: positive predictive value; NPV: negative predicted value.