Production of Extracellular Vesicles Loaded with Therapeutic Cargo

Abstract

Extracellular vesicles (EVs) are biological nanoparticles comprising exosomes, microvesicles, and other heterogeneous nanoscopic vesicle populations that are produced by most cell types. In addition to their putative roles as critical mediators of intercellular communication, EVs have begun to be harnessed as drug delivery vehicles, with early evidence indicating they may have significant advantages over synthetic nanoparticle delivery systems for particular applications. Targeted delivery of EV-encapsulated cargo has already been realized and may have broad applicability; however, methods for producing and purifying EVs and loading them with therapeutic molecules have yet to be standardized. In this chapter, we outline steps for EV isolation and characterization and compare current methods for active and passive loading of EVs with payloads of short interfering RNA (siRNA) or small molecules, with the results revealing that active loading via electroporation increases loading efficiency of siRNA but not of Rhodamine B, a model for a small molecule drug, in HEK293T-derived EVs. The methods described here may inform future design of targeted delivery of nucleic acids or small molecules via EVs.

1. Introduction

Extracellular vesicles (EVs) are natural nanoscopic particles produced by most cells that hold immense promise for utilization as drug carriers. EVs include exosomes (30–120 nm), which are released to the extracellular environment upon fusion of multivesicular endosomes with the plasma membrane, as well as microvesicles (50–1000 nm), which are produced by the outward budding of membrane vesicles from the cell surface [1, 2]. Exosomes and microvesicles have similar properties and are difficult to completely separate with current isolation methods, thus we refer them here as EVs following the recommendations laid out by the International Society for Extracellular Vesicles [3].

EVs play significant roles in intercellular communication via cell-cell transfer of proteins and especially nucleic acids such as microRNAs (miRNAs), long noncoding RNAs (lncRNAs), and mRNAs [4, 5]. The status of EVs as native nucleic acid carriers prompted investigation of their utility for short interfering RNA (siRNA) delivery, and a seminal study by Wood and colleagues demonstrated that exosomes could be targeted to the brain for delivery of therapeutic siRNA [6]. Additional work has shown that EVs can be utilized for targeted delivery of microRNA (miRNA) [7] and small molecule drugs [8] to combat cancer, opening a pathway for EVs to be applied as drug carriers for treatment of numerous diseases as well as in a variety of tissue repair and regeneration applications [9].

However, one limiting factor in further development of EV-based therapeutics is a lack of standardized methods for EV isolation and for loading EVs with therapeutic cargo. This chapter focuses on methods for loading of EVs with siRNA and small molecules; reviews of EV isolation methodologies are available elsewhere [10, 11]. In the study by Wood and colleagues, EVs were loaded with siRNA via electroporation [6, 12], a common molecular biology technique that has been used to deliver DNA, drugs or chemicals into prokaryotic or eukaryotic cells [13, 14]. However, other studies have reported an inability to efficiently load EVs with siRNA [15] or miRNA [7] using this approach, potentially due to electric field-induced aggregation of these short RNA molecules. Here, using HEK293T-derived EVs, we detail an approach for siRNA loading into EVs via electroporation that addresses siRNA aggregation and also define parameters for siRNA loading capacity in EVs.

In addition, we report on methods for loading EVs with small molecules. Small molecule drugs such as curcumin and doxorubicin have been successfully loaded into EVs by different methods [8, 16]. Passive loading, i.e., incubation of EVs with drug in solution with no additional stimulation, is sufficient for EVs to encapsulate curcumin [16]. Alternatively, electroporation has been employed to incorporate doxorubicin into EVs [8]. We have compared these passive and active loading methods using Rhodamine B as a model small molecule drug. Overall, this chapter provides an overview of various methods that could be used to incorporate therapeutic cargo into EVs for a wide variety of targeted delivery applications.

2. Materials

All aqueous solutions of reagents should be prepared using ultrapure water and filter-sterilized through a 0.22 μm filter into sterile container.

2.1. Cell Culture

1. EV producing cells: HEK293T cells (ThermoScientific HCL4517).

2. Cell culture media: DMEM high glucose with sodium pyruvate (110 mg/ml) + L-glutamine (6 mM) + penicillin (100 units/ml) + streptomycin (100 μg/ml) + 10% fetal bovine serum (FBS) as final concentrations.

2.2. EV Isolation and Characterization

1. Ultracentrifuge with rotor capable of 100,000 × g spins: Beckman Optima L-90K ultracentrifuge with T70i rotor.

2. Ultracentrifuge tubes: Optiseal tubes (Beckman 41121703).

3. Nanoparticle imaging apparatus with 50 nm sensitivity: Nanosight LM10 with Nanoparticle Tracking Analysis (NTA) software.

2.3. Cargo Preparation

1. siRNA.

   1. Purified siRNA: sequence: GGUGCCAGUUC UCCAAGAUUdTdT (Dharmacon GE Life Sciences CTM-120916). Resuspend into ultrapure RNase-free water to make a final concentration of 200 pmol/μl (200 μM).
2. Small molecules.

   1. Purified hydrophobic molecules less than 1000 Da: Rhodamine B (Sigma 83689–1G). Prepare 10 mM stock solution in DI H₂O and store at room temperature with protection from light.

2.4. Electroporation

1. Electroporation buffer: 1.15 mM potassium phosphate pH 7.2, 25 mM potassium chloride, 21% Optiprep, as described in Alvarez-Erviti et al. [6].

2. Electroporator and cuvettes: GenePulser Xcell electroporator (Biorad) with Gene Pulser/Micropulser Cuvettes (Biorad 165–2089).

3. 300 kDa MWCO filter: Pall Nanosep centrifugal device with Omega membrane, MWCO 300 kDa (OD300C33).

2.5. Loaded Cargo Detection

1. Kit that can detect nucleic acid concentration at picogram sensitivity: Quant-it PicoGreen Assay kit (Life Technologies P7589).

2. Labeling buffer: 0.5% DMSO in 1× TE (10 mM Tris, pH 7.5 and 1 mM EDTA, pH 8.0.

3. Black-walled clear bottom non-treated polystyrene 96-well plates.

4. 0.5 ml centrifuge tubes, 0.2 ml thin walled, nuclease-free PCR tubes.

5. Microplate reader with fluorescence capability (or any other modality needed for small molecule quantification): Molecular Devices SpectraMax M5.

3. Methods

3.1. Measurement of EV Concentration and Size

1. Grow cells in EV-depleted media (see Note 1) in T150 flasks.

2. Isolate EVs using differential centrifugation method as described previously [17]. In brief, collect media from cultured cells and centrifuge at 300 × g for 10 min. Transfer supernatant into a new tube and centrifuge again at 2000 × g for 20 min and then 10,000 × g for 30 min to remove larger vesicles and debris. Finally, transfer the supernatant into ultracentrifuge tubes and centrifuge at 100,000 × g for 2 h to pellet EVs consisting primarily of exosome and microvesicle fractions.

3. Discard the supernatant and resuspend EVs into ice-cold 1× PBS (see Note 2).

4. Determine size and concentration of EVs (Fig. 1) (see Note 3).

5. For electroporation, resuspend EVs into electroporation buffer (EB) and, if not immediately used, store EVs either at −20 or −80 °C (see Note 4). For passive loading, resuspend EVs into 1× PBS and, if not immediately used, store EVs either at −20 or −80 °C.

Fig. 1.

EV size distribution. Size distribution of HEK293T־derived EVs as measured by NanoTraoking Analysis reveals a peak at 76 nm in diameter, with the majority of isolated EVs being less than 200 nm in diameter

3.2. Passive and Active Loading

1. Prepare samples for loading using the appropriate proportions for each drug molecule type listed below. Examples of appropriate controls are indicated in Figs. 2 and 3 (see Notes 5 and 6).

   1. siRNA: 10 μl of 1 μg/μl EVs (~3.2 × 10⁸ vesicles for HEK293T) + 5 μl of200 pmol/μl siRNA + 25 μl 2× electroporation buffer (or PBS for passive loading) + 10 μl DI H₂O (nuclease free).
   2. Rhodamine B: 10 μl of 1 μg/μl EVs (~3.2 × 10⁸ vesicles for HEK293T) + 10 μl of 10 mM Rhodamine B (see Note 7) + 25 μl 2× electroporation buffer (or PBS for passive loading) + 5 μl DI H₂O (nuclease free).
2. Perform loading step.

   1. Active loading: Transfer samples (samples 1, 2, and 3 in Fig. 2) into ice chilled cuvettes and electroporate each sample at 400 V and 125 μF with two pulses.
   2. Passive loading: Incubate samples at room temperature for 15 min (samples 4, 5, and 6 in Fig. 2).

3. Transfer each electroporated sample from its cuvette into a 0.5 ml tube.

4. Recover the residual sample from each cuvette by adding 100 μl of 1× TE, pipetting up and down and transferring to the appropriate 0.5 ml tube.

5. For siRNA only, add EDTA to a final concentration of 1 mM to alleviate siRNA aggregation and incubate for 15 min at room temperature.

6. Transfer the sample to a 300 kDa MWCO filter tube. This filter retains EVs but small molecules and buffers that have not been incorporated into EVs will pass through the filter during washing steps.

7. Centrifuge samples at 5000 × g at 4 °C for 5 min to remove buffer and unincorporated cargo.

8. Discard flow through, add 500 μl of 1× TE into each tube, and centrifuge again at 5000 × g at 4 °C for 5 min.

9. Repeat step 8 two times for a total of three washes.

10. Add 50 μl of 1× TE into the same filter tube, pipette up and down, and transfer the sample into a fresh 0.5 ml tube.

11. Add additional 45 μl of 1× TE to recover residual amount of sample from filter tube and transfer total volume of sample, i.e., 95 μl into a 0.2 ml thin walled PCR tube (see Note 8).

Fig. 2.

Schematic for controlled experiments to assess active and/or passive loading of EVs. Sample preparation is described in Notes 5 and 6. Model drugs (drug) used included siRNA and Rhodamine B. Asterisk indicates this step is relevant for siRNA loading only

Fig. 3.

siRNA incorporation into EVs. (a) siRNA is detectable from EVs that were electroporated in the presence of siRNA following extensive washing and filtration to remove unassociated molecules. Electroporation of siRNA alone does cause positive signal, potentially as the result of aggregation. EVs incubated passively with siRNA do not retain detectable amounts. (b) siRNA associated with EVs is influenced by the initial loading amount until saturation between 2500 and 5000 pmol loaded. These data were normalized to the background signal generated by electroporated siRNA at each loading amount. Note: 10 μg of EVs corresponds to ~3.2 × 10⁸ vesicles

3.2.1. siRNA Quantification

1. To lyse EVs and release incorporated siRNA, add 5 μl of 0.4% SDS into 95 μl of EVs mixture, mix well, and incubate in a thermocycler for 15 min at 85 °C (see Note 9).

2. Prepare working solution (0.5% DMSO in 1× TE buffer). 10 μl of Quant-iT PicoGreen dye is added to 450 μl of working solution to make dye reagent (see Note 10). Protect dye from light by covering with foil or placing in the dark, as the dye is susceptible to photodegradation. Freshly prepare these solutions immediately prior to the labeling reaction.

3. Add equal volume of dye reagent to each sample prepared from step 1 of this section. Dye reagent (100 μl) is added to 100 μl of each sample to make final volume of 200 μl.

4. Prepare a control that contains siRNA only (sample 7, Fig. 2). Mix 10 pmol of siRNA (10 μl, 1 pmol/μl) with 85 μl of 1× TE and 5 μl of 0.2% SDS. Finally, combine this solution with 100 μl of dye reagent to make final 200 μl final volume (see Note 11).

5. Transfer samples into 96-well plates (see Note 12) and incubate at room temperature for 10 min away from light (cover with aluminum foil). Measure the sample fluorescence using a fluorescence microplate reader (excitation ~480 nm, emission ~520 nm).

6. Measure the fluorescence value from 1× TE with 0.01% SDS as background and subtract this value from all samples (1–7).

7. After subtraction from background, almost no signal should be detected from samples 3, 5, and 6. No signal from sample 5 means that washing conditions are good enough to remove siRNA from the filter tube. Example data is shown in Fig. 3a (see Note 13).

3.2.2. Small Molecule Quantification

1. Lyse EVs by following step 1 of Subheading 3.2.1 as described previously.

2. Prepare a control that contains Rhodamine B only (sample 7, Fig. 2). Mix 10 μl of Rhodamine B (1 nmol, 100 pmol/μl) with 85 μl of 1× TE and 5 μl of 0.2% SDS.

3. Transfer samples into 96-well plates and measure fluorescence using excitation and emission wavelengths of 540 and 625 nm respectively.

4. Measure the fluorescence value from 1× TE with 0.01% SDS at these wavelengths as background and subtract this value from all samples (1–7).

5. After subtraction from background, almost no signal should be detected from samples 3, 5, and 6. No signal from sample 5 means that washing conditions are good enough to remove Rhodamine B from the filter tube. Example data is shown in Fig. 4a (see Note 14).

Fig. 4.

Small molecule incorporation into EVs. (a) Rhodamine B is detectable from EVs that were electroporated or incubated in the presence of Rhodamine B after extensive washing and filtration to remove unassociated molecules. Unlike siRNA, there appears to be no increase in Rhodamine B incorporation into EVs associated with electroporation. (b) Rhodamine B associated with EVs is influenced by the initial loading amount until saturation between 500 and 1000 nmol loaded. These data were normalized to the background signal generated by non- electroporated Rhodamine B at each loading point. Note: 10 μg of EVs corresponds to ~3.2 × 10⁸ vesicles