Optimized Protocol for Isolation of Small Extracellular Vesicles from Human and Murine Lymphoid Tissues

Abstract

Small extracellular vesicles (sEVs) are nanoparticles responsible for cell-to-cell communication released by healthy and cancer cells. Different roles have been described for sEVs in physiological and pathological contexts, including acceleration of tissue regeneration, modulation of tumor microenvironment, or premetastatic niche formation, and they are discussed as promising biomarkers for diagnosis and prognosis in body fluids. Although efforts have been made to standardize techniques for isolation and characterization of sEVs, current protocols often result in co-isolation of soluble protein or lipid complexes and of other extracellular vesicles. The risk of contaminated preparations is particularly high when isolating sEVs from tissues. As a consequence, the interpretation of data aiming at understanding the functional role of sEVs remains challenging and inconsistent. Here, we report an optimized protocol for isolation of sEVs from human and murine lymphoid tissues. sEVs from freshly resected human lymph nodes and murine spleens were isolated comparing two different approaches-(1) ultracentrifugation on a sucrose density cushion and (2) combined ultracentrifugation with size-exclusion chromatography. The purity of sEV preparations was analyzed using state-of-the-art techniques, including immunoblots, nanoparticle tracking analysis, and electron microscopy. Our results clearly demonstrate the superiority of size-exclusion chromatography, which resulted in a higher yield and purity of sEVs, and we show that their functionality alters significantly between the two isolation protocols.

Keywords: exosomes; extracellular vesicles; isolation; lymph node; purification; size-exclusion chromatography; small extracellular vesicles; solid tissue; spleen; sucrose density cushion; ultracentrifugation.

Figure 1

Experimental outline of the comparison of size-exclusion column-based (SEC) versus density-based small extracellular vesicle (sEV) isolations. Supernatant of dissociated lymphatic tissues was separated by differential ultracentrifugation and the resulting 100 K pellet was resuspended and split into two equal parts for direct method comparison. Equal volumes were loaded on either SEC columns or on a sucrose density cushion. Resulting sEV fractions were compared for yield, purity and functionality.

Figure 2

Comparative isolation and characterization of sEVs from human lymph nodes (LN). Size-exclusion column (SEC) fractions (F0 = void volume, 1 mL; F1, F2, F3, F4, F5, F6, F7 = serial fractions, 250 µL) and cushion fraction (pellet resuspended in 250 µL) were analyzed by nanosight tracking analysis (NTA), bicinchoninic acid (BCA) protein quantification, immunoblotting, and transmission electron microscopy (TEM). (A) Left: particle concentrations in IZON fractions F0–F7 and “cushion” fraction for three different human LN samples measured by NTA. Right: Absolute number of detected particles as sum of fraction 1 and fraction 2. For each sample, the particle concentration was normalized to the final volume of elution. (B) Left: BCA protein quantification for IZON fraction F0–F7 and the “cushion” fraction. Right: Absolute amount of protein in fraction 1 and fraction 2 (the protein concentration was normalized to the final volume of elution). (C) Mean particle size for IZON fraction F0–F7 and the “cushion” fraction analyzed by NTA. (D) Ratios of particles per protein amount are plotted for IZON and “cushion” fraction. (E) Representative particle distribution profile for IZON fraction 2 sample (left) and “cushion” sample analyzed by NTA (Sample LN221). (F) Immunoblotting analysis of FLOTILLIN-1, CD81, CD9, TSG101, CALNEXIN, CYTOCHROME C, and GM130 for indicated IZON fractions, the “cushion” fraction and parental cell lysates for one LN sample. (G) TEM images of IZON fraction 2 and the “cushion” fraction for the three indicated samples. (H) Immunogold electron microscopy for HLA-DR of one human LN sample (sample LN221). Scale bar: 200 µm.

Figure 3

Isolation and characterization of murine spleen sEVs. (A) Left: SEV concentration in the different IZON Fractions and “cushion” preparations for three samples analyzed by NTA. Right: Absolute number of particles in indicated preparations. For each sample, the particle concentration in the two peak fractions or in the cushion product was normalized to the final volume of elution. (B) Left: protein quantification in the indicated preparations assessed by BCA assay. Right: absolute amount of protein in indicated preparations (the protein concentration was normalized to the final volume of elution). (C) Mean particle size of all fractions and the “cushion” preparations analyzed by NTA. (D) Ratios of particles per protein amount are plotted for IZON and “cushion” fraction. (E) One representative particle distribution profile for an IZON fraction 2 (left) and a “cushion” preparation analyzed by NTA (Spleen 42704). (F) Immunoblotting analysis of FLOTILLIN-1, ALIX, TSG101, CALNEXIN, and ATP5A for the different IZON fractions, the “cushion” preparation and parental cells for one spleen sample (spleen 224). (G) Transmission electron microscopy (TEM) images of IZON peak fraction and “cushion” preparation for the three indicated samples. (H) TEM image of ferritin-like structures found in “cushion” preparations. (I) Pictures of the sEV pellet, resuspended sEVs prior to application on the sucrose density cushion, and final pellet in PBS before resuspension.

Figure 4

Response of murine monocytes upon tumor-derived sEVs (TEX) treatment. Bone marrow-derived monocytes were treated with 5 µg of the indicated sEV preparations for 8 h and analyzed by flow cytometry gating on CD11b+F4/80+⁺CX3CR1+Ly6C+ cells (n = 3 mice per sEV preparation). (A) Top: percentage of PD-L1 positive cells among CD11b+F4/80+CX3CR1+Ly6C+ monocytes. Bottom: representative histogram including isotype antibody staining as negative control (IgG). (B) Percentage of MHC-II/HLA-DR positive cells among CD11b+F4/80+CX3CR1+Ly6C+ monocytes. Bottom: representative histogram including fluorescence-minus-one (FMO) staining as negative control. (C) Top: ICAM-1/CD54 expression presented as normalized mean fluorescence intensity (nMFI). Bottom: representative histogram. p-values were determined by one-way ANOVA with Tukey’s multiple comparisons test. * p < 0.05; ** p < 0.0021; *** p < 0.0002; **** p < 0.0001.