Proof of concept of using a membrane-sensing peptide for sEVs affinity-based isolation

Abstract

Introduction: One main limitation in biomarker studies using EVs is the lack of a suitable isolation method rendering high yield and purity samples in a quick and easily standardized procedure. Here we report an affinity isolation method with a membrane-sensing peptide (MSP) derived from bradykinin. Methods: We designed a protocol based on agarose beads carrying cation chelates to specifically bind to the 6His-tagged membrane-sensing peptide. This approach presents several advantages: 1) cation-carrying agaroses are widely used and standardized for His-tagged protein isolation, 2) the affinity protocol can be performed in small volumes, feasible and manageable for clinical routine and 3) elution with imidazole or EDTA allows a gentle and easy recovery without EV damage, facilitating subsequent characterization and functional analyses. Results: The optimized final procedure incubates 0.5 mg of peptide for 10 min with 10 µL of Long-arm Cobalt agarose before an overnight incubation with concentrated cell conditioned medium. EV downstream analyses can be directly performed on the agarose beads adding lysis or nucleic-acid extraction buffers, or gently eluted with imidazole or EDTA, rendering a fully competent EV preparation. Discussion: This new isolation methodology is based on the recognition of general membrane characteristics independent of surface markers. It is thus unbiased and can be used in any species EV sample, even in samples from animal or plant species against which no suitable antibodies exist. Being an affinity method, the sample handling protocol is very simple, less time-consuming, does not require specialized equipment and can be easily introduced in a clinical automated routine. We demonstrated the high purity and yield of the method in comparison with other commercially available kits. This method can also be scale up or down, with the possibility of analyzing very low amounts of sample, and it is compatible with any downstream analyses thanks to the gentle elution procedure.

Keywords: affinity; agarose beads; extracellular vesicles; isolation; peptide.

Graphical abstract

FIGURE 1

Selection of chelate resin. (A) Scheme of peptide (BPt) and EVs binding to cation agaroses. (B) Dot blot analysis with anti poly-Histidine antibody to compare BPt binding to copper, cobalt, and nickel agaroses. 0.1 mg of BPt was incubated with 50 µL of copper, cobalt, or nickel agarose for 1 h at RT.

FIGURE 2

Optimization of concentration and incubation times. Peptide concentration: different amounts of peptide as indicated were incubated with 10 µL of cobalt agarose for 1 h at RT. Then, 100 µL of concentrated conditioned media were added and incubated ON at 4°C. Anti-CD81 (5A6) (A) and anti-ApoB (B) antibodies were used for EVs and LPPs detection, respectively. Peptide incubation time: dot blot analyses to compare EVs (CD81) (C) and LPPs (ApoB) (D) binding to agaroses incubated with 0.5 mg of peptide on 10 µL of cobalt agarose for the indicated times before the addition of 100 µL of concentrated conditioned media. (E) EV binding time: Dot blot analysis of EV binding changing the times of incubation at 4°C with concentrated conditioned medium. As in previous experiments, 0.5 mg of peptide were incubated with 10 µL of cobalt agarose for 10 min before the addition of concentrated conditioned medium.

FIGURE 3

Optimization of chelate exposure. (A) Scheme of the two different cobalt agaroses tested which differ in the length of the linker between the agarose surface and the cobalt cation. Dot blot analysis of peptide (poly Histidine) (B) and EVs binding (CD81) (C) obtained with the two agaroses tested: Simple Cobalt Agaroses (Co Ag) and Long Arm Cobalt Agarose (LA-Co Ag). LA-Co Ag present half of cations than Co Ag. 250 or 500 µg were incubated with 10 µL of both agaroses. 100 μL of concentrated conditioned media was employed for EV binding. Results are represented as the % of a given marker signal in each dot. The means ± SEM of 3 independent experiments are shown. Significant differences in the T-student statistical test are indicated with * (p ≤ 0.05).

FIGURE 4

(A) Scheme and sequences of the two versions of BP tested and its binding to LA-Co agarose. BPt is a tandem version composed of 2 BP repetitions and BPb is a branched version composed of 4 BP repetitions. Dot blot analysis of peptide (Poly Histidine) (B) and EVs (CD81) (C) binding when tandem (BPt) or branched (BPb) peptide is employed. Again, 0.5 mg of peptides were incubated with 10 µL of LA-Co agarose before the addition of 100 µL of concentrated conditioned media.

FIGURE 5

Optimization of pH (A) Dot blot analysis of peptide and EV binding at different pH. 0.5 mg of BPt were incubated with 10 µL of LA-Co agarose in a final volume of 100 µL using PBS of different pH. After washing with different pH buffer, 100 µL of concentrated conditioned media diluted by half in PBS at the corresponding pH were added. (B) Quantifications of dot blot analysis. Means ± SEM of 3 independent experiments are shown. Significant differences in the T-student statistical test are indicated with * (p ≤ 0.05).

FIGURE 6

EV characterization. For EV isolation using BPT-based affinity chromatography, 40 µL of LA-Co agarose, 1 mg of BPt and 200 µL of concentrated conditioned media were used. After washing bounded EVs were eluted from agarose by incubation with 100 µL of filtrated PBS containing 0.5 M imidazole. For SEC isolation, 400 µL of concentrated conditioned media were employed for EV isolation. (A) Size profile of isolated EVs by BPt affinity or SEC isolation analyzed by NTA. (B) Representative TEM images of negatively stained EV samples and analysis of mean EV diameter from TEM Exosome Analyser on TEM images. Scale bars = 1 µm or 500 nm (on close up images). At least 9 images of 3 independent experiments were analyzed for each condition and the data was shown as mean ± SEM. Western blot analysis of (C) EV markers (Flotillin, TSG101, Syntenin, and ARF6), (D) non-EV markers (Calnexin or VDAC) in EV isolated samples.

FIGURE 7

EV yield in comparison to commercial kits. Customized SiMoA (Single Molecule Assay) pan-tetraspanin EV detection was used to estimate EV recovery after BPt protocol and other commercial tools. (A) SiMoA results of pan-tetraspanin immune-phenotyping of EVs sample dilutions ranging from 10⁴ to 10¹⁰ particle/mL, expressed as Average Enzyme per Bead (AEB) (B) EV recovery yield after normalization to starting sample volume for BPt protocol, IZON SEC, ExoQuick precipitation, immune-capture on magnetic beads preceded by ExoQuick precipitation, immune-capture on magnetic beads only, ExoSpin kit and phosphatidylserine affinity capture on magnetic beads by FujiFilm. Bar errors indicate SD from 3 technical replicates.

FIGURE 8

Analysis of EV uptake. SKMEL-147 cells were incubated with 25,000 EVs/cell, previously labelled with maleimide-Alexa633. EV uptake by SKMEL-147 cells was assessed by flow cytometry. (A) Uptake analysis of EVs isolated by BPt capture and imidazole elution (orange) or SEC (blue) after 2 h of incubation with SKMEL-147 cells. Fluorescence of the negative control (cells without EVs) is shown in the red histogram. (B) Uptake analysis of EVs after incubation with SKMEL-147 cells during 30 min (red), 1 h (blue) or 2 h (orange). EVs were previously isolated by SEC (left plot) or BPt capture and EDTA elution (right plot). Bars indicate negative signal threshold from SKMEL-147 cells without EVs.