Isolation and characterization of vesicular and non-vesicular microRNAs circulating in sera of partially hepatectomized rats

Abstract

Circulating microRNAs are protected from degradation by their association with either vesicles or components of the RNAi machinery. Although increasing evidence indicates that cell-free microRNAs are transported in body fluids by different types of vesicles, current research mainly focuses on the characterization of exosome-associated microRNAs. However, as isolation and characterization of exosomes is challenging, it is yet unclear whether exosomes or other vesicular elements circulating in serum are the most reliable source for discovering disease-associated biomarkers. In this study, circulating microRNAs associated to the vesicular and non-vesicular fraction of sera isolated from partially hepatectomized rats were measured. Here we show that independently from their origin, levels of miR-122, miR-192, miR-194 and Let-7a are up-regulated two days after partial hepatectomy. The inflammation-associated miR-150 and miR-155 are up-regulated in the vesicular-fraction only, while the regeneration-associated miR-21 and miR-33 are up-regulated in the vesicular- and down-regulated in the non-vesicular fraction. Our study shows for the first time the modulation of non-vesicular microRNAs in animals recovering from partial hepatectomy, suggesting that, in the search for novel disease-associated biomarkers, the profiling of either vesicular or non-vesicular microRNAs may be more relevant than the analysis of microRNAs isolated from unfractionated serum.

Figure 1. TEI precipitation recovers extracellular vesicles more efficiently as compared to ultracentrifugation.

(a) Schematic representation of experimental procedure is shown. Vesicles were isolated from independent pools of rat sera by either addition of TEI or by ultracentrifugation for 2 (standard) or 18 (extended) hours. Following NTA was used to measure residual vesicles in the supernatant, compared to the number of vesicles present in the whole sera (Not Centrifuged). Measurements from two independent pools are shown. (b) Residual EVs distribution and (c) Number of residual EVs as measured by NTA. N.D. Non-detectable.

Figure 2. Vesicles isolated from rat sera via TEI-precipitation are up taken by rat primary hepatocytes.

To assess whether the vesicles isolated via TEI-induced precipitation were active, freshly isolated rat primary were incubated over night with 1.9 × 10¹² vesicles, which were either stained with Syto RNAselect (left panel) or left unstained (right panel). Vesicle uptake was recorded by live imaging.

Figure 3. Filtration does not affect distribution of vesicles isolated by TEI-induced precipitation.

Pool of rat sera were either filtered through 0.2 μm (FL200) or 0.45 μm (FL450) syringe filters or left unfiltered (NFLS) before isolation by addition of appropriate quantity of TEI. Following, vesicle distribution was measure by using NTA. (a) Schematic representation of experimental procedure is shown. (b) Vesicles size distribution and (c) vesicles peak as measured by NTA for three independent pools. Data are represented as average ± standard deviation. *P ≤ 0.05.

Figure 4. Expression of VEP-associate miRNAs isolated from rats sera by using Ultracentrifugation and TEI.

RNAs isolated from whole sera (TOTAL), from vesicles pelleted by either Total Exosome Isolation reagent-based (TEI) or Standard ultracentrifugation (ULTRA) were reverse transcribed following the miQPCR protocol, and the expression of a panel of miRNAs was analyzed by individual qPCR assays. qPCR data were median normalized as described in ref. . Data are represented as average ± standard deviation calculated from four independent isolations (n = 4). *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.0001; n.s., non-significant.

Figure 5. Effect of filtration on expression profiling of VEP- and VDS-associated miRNAs.

Vesicular and non-vesicular RNAs were extracted from sera isolated from wild type rats. Expression of miRNAs present in the VDS and VEP fractions isolated from (a) non filtered (NFLS) serum, (b) serum filtered through 0.2 μm (FL200) or (c) serum filtered through 0.45 μm (FL450) were measured by using miQPCR. Data are represented as average ± standard deviation from 4 independent isolations (n = 4).

Figure 6. Expression of vesicles-associated miRNAs isolated from the serum of rats after partial hepatectomy.

Vesicles were isolated via TEI precipitation from sera of rats recovering from PHx and RNAs were isolated from VEP fractions of sera. qPCR data were median normalized as described in ref. . Data are represented as average ± standard deviation. Statistical analysis was performed by One-Way Anova (day 0; n = 3) versus individual groups (day 2, 4 and 6; n = 3 within each group) for each miRNA. *P ≤ 0.05; **P ≤ 0.01.

Figure 7. Expression of miRNAs circulating in VDS fractions of sera isolated from rats recovering from partial hepatectomy.

Vesicles were precipitated from sera isolated from rats recovering PHx and RNAs were isolated from vesicles-depleted sera. qPCR data were median normalized as described in ref. . Data are represented as average ± standard deviation. Statistical analysis was performed by One-Way Anova (n = 3 within each group). *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.0001; ^§, indicate that the P value is included between: 0.05 < P ≤ 0.06.

Figure 8. NTA analysis of EVs circulating in sera of rats recovering from partial hepatectomy.

Sera isolated from rats recovering from PHx were subjected to centrifugation (20000 rcf/20 min). EVs present in the clarified supernatant were measured by using NTA. Data are represented as average ± standard deviation. Statistical analysis was performed by unpaired T-test (day 0; n = 3) versus individual regeneration groups (day 2, 4 and 6 after PHx; n = 3 within each group). *P ≤ 0.05.

Figure 9. Modulation of miRNA expression and secretion in activated rat hepatic stellate cells.

Following activation, rat HSCs were cultured for 24 hours in serum free medium before stimulation with either FCS or the GSK-3β inhibitor TWS-119 or cytokines (IL-6 and TGF-β1) for 72 hours. (a) Effect of treatments on HSCs morphology was assessed via contrast phase microscopy. Expression profiling of selected miRNAs was analyzed by miQPCR in either (b) cellular or (c) vesicles-associated RNAs. The values of unstimulated controls HSCs (CRT) were set arbitrarily to 100. Statistical analysis was performed by unpaired T-test of control group (CRT; n = 12) versus individual treatment groups (n = 9 to 12) for each miRNA. Data are represented as average ± standard deviation. *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.0005; and ****P ≤ 0.0001. N.D. Non-detectable.