Labeling of Extracellular Vesicles for Monitoring Migration and Uptake in Cartilage Explants

Osteoarthritis (OA) is an articular degenerative disease that implies a progressive and irreversible inflammation and destruction of the extracellular matrix of the articular cartilage¹. Although various forms of arthritis have numerous treatments^2,3,4, these are restricted by their side effects and limited efficacy. Tissue engineering techniques using autologous chondrocyte implantation are routinely applied for the regenerative treatment of injured cartilage in early OA cartilage lesions⁴. Cell-based therapies are restricted mainly due to the limited number of phenotypically stable chondrocytes or chondroprogenitors capable of effectively repairing the cartilage³. Therefore, the development of new therapeutic strategies to prevent disease progression and regenerate large cartilage lesions is of paramount importance.

Extracellular vesicles (EVs) have been suggested as a treatment for OA by different authors^5,6. EVs are membranous bodies secreted by the majority of cell types, are involved in intercellular signaling, and have been shown to exert stem cells' therapeutic effects^7,8,9, due to which they have recently elicited interest in regenerative medicine¹⁰. EVs derived from mesenchymal stromal cells (MSCs) are the main therapeutic EVs investigated for OA, although other joint-related cells have been used as EV sources, e.g., chondroprogenitors or immune cells^11,12.

Platelet concentrates, such as platelet lysates (PLs), are used to enhance wound healing in different injuries, such as corneal ulcers^13,14,15 or in tendon tissue regeneration¹⁶, because of the hypothesis that the EV component of platelet concentrates may be responsible for these effects¹⁷. Some studies related to joint-related diseases use platelet-derived EVs (PL-EVs) as a treatment to ameliorate osteoarthritic conditions. PL-EVs improve chondrocyte proliferation and cell migration by activating the Wnt/β-catenin pathway¹⁸, promoting the expression of chondrogenic markers in osteoarthritic chondrocytes¹⁹, or showing higher levels of chondrogenic proteins and fewer tissular abnormalities in osteoarthritic rabbits treated with PL-EVs¹⁸.

EVs contain proteins, lipids, and nucleic acids that are liberated to the target cell, establishing cell-to-cell communication, which is the main feature related to their therapeutic applications²⁰. The effects of EVs depend on their reaching cells and subsequent cargo release. This effect can be confirmed indirectly by changes caused in cells, such as metabolic activity or gene expression modification. However, these methods do not allow the visualization of how EVs reach cells to exert their function. Thus, this paper presents a method to label these PL-derived EVs to be used as a treatment for inflammation-driven OA cartilage explants. Confocal microscopy was used to monitor EV uptake and progression to the chondrocytes present in the explants in a time-lapse of 5 h.

NOTE: Cartilage explants were obtained from the IdISBa Biobank (IB 1995/12 BIO) in compliance with institutional guidelines after ethical approval of the project by the CEI-IB (IB 3656118 PI).

1. Column preparation

1. Equilibrate columns following the manufacturer's instructions or as follows:
   1. Remove the column cap and equilibrate the column. Remove the storage buffer by elution.
   2. Wash the column 3 times with 2.5 mL of phosphate-buffered saline (PBS). During each wash, wait for the column to absorb the whole volume.
      ​NOTE: Do not let the column dry.
   3. Cover the column with the cap after the last wash and until sample preparation.

2. EV labeling

NOTE: This EV labeling protocol uses a PL-EV sample previously isolated by size exclusion chromatography (SEC) with previously described conditions^21,22. However, any EV sample from any source may be used with this protocol.

1. Concentrate the EV sample and the control (PBS) using a concentrating tube.
   1. Place the samples in a 15 mL or 500 µL concentrating tube, depending on the starting volume of the EV sample starting. Centrifuge the tubes according to the manufacturer's instructions until an almost-dry sample is obtained.
      NOTE: The control sample is necessary to check for any dye background. Although this method does not require any specific initial volume, it should be considered that around 10% of particles will be lost during purification steps. 
2. Resuspend the concentrated samples with diluent C. Resuspend EV samples with 200 µL and the control group with 100 µL and transfer them to new 1.5 mL centrifuge tubes.
3. Separate the EV sample into two aliquots of 100 µL. Mark one with dye and use it as treatment (PKH-PL-EV); leave the other unmarked but process it (NTA-PL-EV) like the EV sample and use it to quantify the EV concentration by NTA.
4. Prepare 2x dye solution, resulting in 8 µM PKH26 solution in diluent C.
5. Mix 1 µL of 1 mM PKH26 linker per 125 µL of diluent C in the sample. Prepare a volume required to add to the samples in a 1:1 ratio.
6. Add 2x dye solution to PKH-PL-EV and control samples in a 1:1 ratio to achieve 1x dye concentration and 4 µM PKH26 concentration. Add the same volume of PBS to the NTA-PL-EV sample. Incubate for 5 min at room temperature.
7. Add 5% bovine serum albumin-PBS solution to the samples in a 1:1 ratio and ensure that the final volume is ~400 µL.
   NOTE: This step allows the removal of nonspecific dye interactions or unbound dye.
8. Proceed to separate the labeled EVs from the unbound dye and nonspecific interactions of the dye with the column.

3. Labeled-EV isolation

1. Remove the cap from the column, add 400 µL of the sample (PKH-PL-EV, NTA-PL-EV, or control), and discard all eluted liquid.
2. Wait for the sample to enter the column completely before proceeding to the next step. Add 600 µL of PBS and discard all eluted liquid.
3. Wait for the PBS to enter the column completely before proceeding to the next step. Add 600 µL of PBS and collect a fraction of 600 µL in a 1.5. mL centrifuge tube (EVs or control).
   NOTE: These steps are needed to remove the excess dye from the samples. Another separation by column is needed to obtain purer EVs. Thus, the following steps should be performed in a new equilibrated column (step 4.1.) or the same column after an initial washing step (step 4.2).
4. Prepare the column for a new EV separation step to obtain purer EVs. If it is a new column, repeat steps 2.1. and 2.2. If it is the same column, wash the column with 2.5 mL of 20% isopropanol and then repeat steps 2.1 and 2.2.
5. Add 600 µL of previously eluted EVs obtained in step 2.5 to the column and discard the eluted volume. Wait for the liquid to enter the column completely before proceeding to the next step.
6. Add 400 µL of PBS and discard all eluted volume. Wait for the liquid to enter the column completely before proceeding to the next step.
7. Add 600 µL of PBS and collect a fraction of 600 µL in a 1.5 mL centrifuge tube. Use the EVs and control samples for further analyses or store them overnight at 4 ᵒC.
8. Store the used columns for future use.
   1. Wash the column with 25 mL of 20% isopropanol and discard the eluted volume. Wash the column 3 times with 2.5 mL of PBS.
   2. Add the storage buffer removed in step 1.1.1 and wait for the buffer to enter the column. Cover the column with the cap and store at 4 ᵒC until subsequent use.

4. EV quantification

1. Prepare 1:1,000 dilutions of the NTA-PL-EV sample and the initial PL-EV sample as described by the following two steps.
   1. Prepare 1 mL of 1:10 diluted NTA-PL-EV and 1 mL of 1:10 diluted initial PL-EV with PBS filtered through a 0.2 µm filter.
   2. Prepare 1 mL of a 1:100 dilution of the previous diluted samples with PBS filtered through a 0.2 µm filter.
2. Inject the 1:1,000 diluted NTA-PL-EV sample or the initial PL-EV sample using a sterile syringe into the NTA pump. Follow the manufacturer's instructions and recommendations for particle concentration and size distribution determination.
   ​NOTE: As EV concentration depends on the sample starter volume, it may be necessary to read intermediate dilutions and make adjustments to obtain a correct NTA determination.

5. EVs used as a treatment for inflammation-driven OA

1. Wash the cartilage twice with PBS and excise it using a 3 mm diameter biopsy punch under sterile conditions.
   NOTE: Perform the procedure from steps 5.1 to 5.6 in a cell culture hood.
2. Place the explants in 96-well culture plates with DMEM-F12 medium supplemented with 1% penicillin-streptomycin at 37 ᵒC, 5% CO₂, and 80% humidity.
   1. To establish an inflammation-driven OA model, supplement the cell culture medium with 10 ng/mL oncostatin M and 2 ng/mL tumor necrosis factor-alpha (TNFα).
3. Treat the explants with 1 × 10⁹ particles/well of labeled EVs (PKH-PL-EV) or control in cell culture medium supplemented with oncostatin M and TNFα.
   NOTE: Measure the volume of the sample containing 1 × 10⁹ particles/well and use the same volume for the control.
4. Remove the medium from the 96-well cell culture plates containing cartilage explants. Add 200 µL of the cell culture medium described in step 5.3 to each well.
   NOTE: If the 96-well plates have been in contact with fetal bovine serum (FBS), wash each well three times with 200 µL of PBS to remove any EVs from the FBS.
5. Stop the in vitro assay at different times: 0, 1, 2, 3, 4, and 5 h.
6. Wash the cell culture wells containing the cartilage explants twice with 200 µL of PBS.
7. Add 100 µL of 4% paraformaldehyde (PFA) to the tissue to fix it for 3 h at 4 ᵒC.
   ​NOTE: Steps involving PFA should be performed in a fume hood following the Safety Data Sheet recommendations.
8. Remove the PFA, add 100 µL of PBS, store the fixed tissue at 4 °C, and process the samples within 48 h.

6. Microscopy preparation and visualization

NOTE: This histological procedure consists of dehydration, paraffin embedding, and rehydration steps. These steps may reduce overall dye fluorescence (a limitation mentioned in the datasheet for PKH26). Therefore, other procedures, such as frozen sectioning, may be more suitable for EV visualization by confocal microscopy.

1. Embed the fixed tissues in paraffin blocks. Cut the tissue into 6 µm-thick sections.
2. Deparaffinize the tissue sections.
   NOTE: All steps using xylene should be performed in a fume hood.
   1. Immerse the tissue sections in xylene for 30 min, 100% ethanol for 2 min, 96% ethanol for 2 min, 75% ethanol for 1 min, and finally in distilled water for 30 s.
3. Permeabilize the tissue sections.
   1. Prepare a 0.1% Triton-0.1% sodium citrate solution to permeabilize the tissue. Add a 20 µL drop to each tissue section and incubate for 10 min at room temperature. Wash each section twice with 20 µL of PBS.
4. On a microscopic slide, add a drop of mounting medium containing 4′,6-diamidino-2-phenylindole (DAPI) with an aqueous mounting medium for preserving fluorescence. Cover the slide containing 3 tissue sections from step 6.3.1.
5. Incubate the slides overnight at room temperature, protected from light.
6. Store at 4 °C, protected from light until confocal microscopy.

A schematic overview for EV labeling and uptake monitoring is displayed in Figure 1. The particle concentration and EV size detected by NTA in Table 1 show that the EV concentration decreases during the process due to the purification step performed twice after labeling with the column. However, the amount obtained is in the optimal range of the number of particles to use for treatment. This particle concentration is used to calculate the volume of PKH-PL-EV and control that are used to treat osteoarthritic cartilage explants.

Once the cartilage explants are treated with EVs or the control group, they are fixed for different periods: 0, 1, 2, 3, 4, and 5 h. Each group is then paraffinized, sliced, and prepared for confocal microscopy with a mounting medium containing DAPI. Representative images for each group at each time point are presented in Figure 2, showing how EVs enter the tissue until they reach the chondrocytes and enter them over time.

As seen in Figure 2, labeled EVs are already localized around chondrocytes (shown in blue with DAPI staining) after 1 h of incubation (shown in red with PKH26 dye). Some background due to remnant dye can be observed for the control group, which does not have EVs but is processed following the same protocol as the EV sample. These results confirm the success of the protocol to label EVs, which can be used to monitor their migration through tissue in in vitro assays, as shown here, and in in vivo experiments.

Figure 1: Schematic overview for EV labeling and uptake monitoring protocol. Abbreviations: PBS = phosphate-buffered saline; EV = extracellular vesicle; RT = room temperature; BSA = bovine serum albumin; NTA = nanoparticle tracking analysis; OA = osteoarthritis; TNFα = tumor necrosis factor-alpha; PL = platelet lysate; PFA = paraformaldehyde; DAPI = 4′,6-diamidino-2-phenylindole. Please click here to view a larger version of this figure.

Figure 2: Representative images of EV uptake at different times. Confocal representative images taken after 0, 1, 2, 3, 4, and 5 h of osteoarthritic cartilage explants incubated with PKH-labeled EVs or with a control group. Images were taken at 400x. Scale bars = 50 µm. Abbreviations: OA = osteoarthritis; PL = platelet lysate; EV = extracellular vesicle. Please click here to view a larger version of this figure.

Table 1: Characterization by nanoparticle tracking analysis. Abbreviations: OA = osteoarthritis; PL = platelet lysate; EV = extracellular vesicle; NTA = nanoparticle tracking analysis.

EV imaging helps to understand EV properties, such as their release and uptake mechanisms. Their imaging allows the monitoring of their biodistribution and the characterization of their pharmacokinetic properties as drug vehicles. However, EV imaging and tracking may be difficult due to their small sizes, although many imaging devices and labeling techniques have been developed to help researchers monitor EVs under in vitro and in vivo conditions^23,24,25.

Two possibilities exist when tracking EVs by optical microscopy: bioluminescence and fluorescence imaging. Both are used to detect EVs within the visible light spectrum (390-700 nm). Bioluminescence is a type of chemiluminescence produced after a luciferase catalyzes the oxidation of its substrate. Although this signal requires an ultrasensitive charge-coupled device (CCD) camera for detection, it has a high signal-to-noise ratio as the signal does not require an external light source²⁶.

Fluorescence imaging uses proteins or organic dyes that emit signals after excitation from an external light source. Compared to bioluminescence, fluorescence is easier to detect by a CCD camera. Moreover, in bioluminescence, substrate toxicity and half-life of the substrate's bioluminescence should be considered for real-time EV tracking^27,28.

In contrast, fluorescent protein- and organic dye-based labeling have been used with excellent resolution in optical microscopy. Although the fluorescence intensity depends on EV protein expression levels, the efficiency of EV labeling at membrane domains, and the excitation light source, fluorescence dyes provide stable and strong signals for EV imaging²⁵. Most organic fluorescent dyes were initially used for cell membrane imaging. These dyes generally combine fluorophores that label the lipid bilayer or proteins of interest on EVs via different functional groups²⁹.

One organic fluorescent dye family is the lipophilic PKH dye family consisting of fluorophores with a lipophilic carbocyanine that anchors into the lipid bilayer for fluorescence imaging^30,31. PKH dyes have been used for in vitro and in vivo studies as their in vivo half-life ranges from 5 to >100 days. Thus, the persistence of the dye in vivo may lead to misleading results in studies of shorter duration than the half-life of the dye. However, PKH dyes are useful as tracers to show EV migration³².

PKH26 is a member of this PKH lipophilic fluorophore family, found in the red spectrum with a peak of excitation at 551 nm and emission at 567 nm. This makes it compatible with other detection channels, such as rhodamine, phycoerythrin, or DAPI³³, allowing the detection of, in this case, the migration of EVs marked with PKH26 toward chondrocytes marked with DAPI. It is important to note that although PL was used as an EV source here, this protocol can be used with EVs from other sources and for other purposes, for example, to track labeled the in vivo distribution of EVs.

This protocol has some limitations; for instance, there are some concerns that PKH26 increases EV size, which may affect their biodistribution and cellular uptake. However, in such cases where EV size was increased by PKH26, the labeling procedure was different from that described in this protocol³⁴. These researchers did not include the washing and purification steps, thus leading to higher levels of free dye, which could cause the larger EV size. Moreover, the present protocol overcomes this problem by performing a parallel EV purification with and without PKH26. This allows the characterization of an unlabeled EV sample, which was processed identically as the labeled one. Thus, a misleading quantification due to confounding nonspecifically labeled particles (lipoprotein or protein sample contaminants) or by the presence of non-EV particles within the labeling mixture can be avoided, as demonstrated previously^35,36.

In this paper, two cycles of purification were performed by size exclusion chromatography using columns. The first one may be substituted by sucrose gradient centrifugation. However, all EV populations were collected in the same eluted aliquot with this column and not subjected to high g forces encountered in high-speed centrifugation. However, avoiding centrifugation may lead to a slower process, especially if column washes are needed between separation cycles. Another limitation of this protocol is the need for sample concentration before starting the labeling process due to the limited volume of the column. This handicap may be overcome by using columns with higher loading capacity.

Other EV-labeling protocols describe the use of different dyes; however, their use of ultracentrifugation steps may damage EV integrity³⁷. The protocol described here has allowed the monitoring of EV migration and uptake in cartilage explants easily with a confocal microscope without any particular function. Furthermore, this may be extrapolated to other tissues and lipidic samples or conditions, such as an in vivo assay.