Transfer of the Cystic Fibrosis Transmembrane Conductance Regulator to Human Cystic Fibrosis Cells Mediated by Extracellular Vesicles

Abstract

Cystic fibrosis (CF) is a genetic disease caused by mutations in the CF transmembrane conductance regulator (CFTR) gene, resulting in a deficiency in chloride channel activity. In this study, extracellular vesicles (EVs), microvesicles, and exosomes were used as vehicles to deliver exogenous CFTR glycoprotein and its encoding mRNA (mRNA^GFP-CFTR) to CF cells to correct the CFTR chloride channel function. We isolated microvesicles and exosomes from the culture medium of CFTR-positive Calu-3 cells, or from A549 cells transduced with an adenoviral vector overexpressing a GFP-tagged CFTR (GFP-CFTR). Both microvesicles and exosomes had the capacity to package and deliver the GFP-CFTR glycoprotein and mRNA^GFP-CFTR to target cells in a dose-dependent manner. Homologous versus heterologous EV-to-cell transfer was studied, and it appeared that the cellular uptake of EVs was significantly more efficient in homologous transfer. The incubation of CF15 cells, a nasal epithelial cell line homozygous for the ΔF508 CFTR mutation, with microvesicles or exosomes loaded with GFP-CFTR resulted in the correction of the CFTR function in CF cells in a dose-dependent manner. A time-course analysis of EV-transduced CF cells suggested that CFTR transferred as mature glycoprotein was responsible for the CFTR-associated channel activity detected at early times posttransduction, whereas GFP-CFTR translated from exogenous mRNA^GFP-CFTR was responsible for the CFTR function at later times. Collectively, this study showed the potential application of microvesicles and exosomes as vectors for CFTR transfer and functional correction of the genetic defect in human CF cells.

Introduction

Most eukaryotic cells naturally release into the extracellular environment a heterogeneous population of membrane-enveloped microparticles, referred to as extracellular vesicles (EVs).¹ These EVs are classified as microvesicles (MVs) and exosomes (EXos), based on their size, ultrastructural and biochemical properties, and their mode of release. MVs are relatively large (100–1000 nm), nonhomogeneous vesicles, generated by the outward budding and shedding from the cell plasma membrane.^2,3 They are sometimes referred to as ectosomes or shedding MVs.^4,5 EXos represent a more homogenous population of smaller membrane-coated particles (30–100 nm in diameter), generated by the inward budding of endosomal membranes within large multivesicular bodies (MVBs). The fusion of the MVBs with the cell plasma membrane results in the release of the EXos into the extracellular milieu.^2

–7

EVs participate in normal biological processes such as tissue repair,^8,9 immune surveillance,¹⁰ blood coagulation,¹¹ and stem cell maintenance,¹² by delivering effectors such as transcription factors, small and large noncoding regulatory RNAs, as well as mRNAs. They act as mediators of intercellular communications via the transfer of proteins, lipids, and nucleic acids.^3,6,10 They are also associated to several diseases such as tumorigenesis,^13,14 Alzheimer's and Parkinson's diseases,^15,16 the spread of the prion protein PrP^c,^17,18 the dissemination of HIV-1,¹⁹ or the reactivation of latent HIV-1.²⁰ The EVs have a natural ability to transfer membrane, cytoplasmic, and genetic material to neighboring and distant cells, and thus have been exploited as vehicles for the delivery of therapeutic molecules such as bioactive proteins, lipids, and nucleic acids.⁸ Several protocols of EV-based vaccination and antitumor therapy are already at the preclinical or clinical stage.²¹

Cystic fibrosis (CF) is an autosomal recessive genetic disease caused by mutations in a single gene, the CF transmembrane conductance regulator (CFTR) gene. The CFTR is a type III membrane glycoprotein, with large intracytoplasmic N- and C-terminal domains and 12 membrane-spanning domains constituting a transmembrane, cAMP-dependent chloride channel.^22,23 It is present at the apical surface of epithelial cells, and localized in the lipid raft microdomains of the plasma membrane.^24
–26 In a previous study, we showed that MVs and EXos were able to achieve the transfer of the human CFTR mature glycoprotein and CFTR-encoding mRNA molecules (mRNA^CFTR) in a Chinese hamster ovarian (CHO) cell model, resulting in the gain in CFTR chloride channel function by MV- and EXo-recipient cells.²⁷

In the present study, we investigated the feasibility of MV- and EXo-mediated delivery of the exogenous CFTR glycoprotein and CFTR-encoding mRNA molecules to CFTR-deficient cells derived from CF patients (CF cells) to restore the CFTR chloride channel function. We used a GFP-CFTR fusion construct, which has been shown to be biologically active and capable of correcting the transepithelial chloride transport,^28,29 to track the exogenous CFTR protein in EV-recipient cells. We found that both MVs and EXos were capable of packaging GFP-tagged CFTR and mRNA^GFP-CFTR, and transferred both types of molecules to target CF cells, resulting in the correction of the CFTR defect in CF cells. The CFTR chloride channel activity newly acquired by EV-transduced CF cells was measurable until day 3 after the EV-mediated transfer, and still detectable at day 5. Experimental results suggested that the maintenance of this function was performed by newly synthesized CFTR molecules translated from exogenous mRNA^GFP-CFTR. This is, to our knowledge, the first demonstration that EVs can be exploited as therapeutic vectors for a genetic disease.

Materials and Methods

Cells

A549 and HEK-293 cells were obtained from the American Type Culture Collection (ATCC; Manassas, VA). HEK-293 (ATCC CRL-1573) and A549 cells (ATCC CCL-185), adenocarcinomic human alveolar basal epithelial cells, were maintained as monolayers in Dulbecco's modified Eagle's medium (DMEM; Life Technologies) supplemented with 10% fetal bovine serum (FBS; Life Technologies), penicillin (200 U/ml), and streptomycin (200 μg/ml) at 37°C and 5% CO₂. Calu-3 cells, derived from human bronchial submucosal glands, were obtained from the laboratory of M.J. Welsh.^30,31 Calu-3 cell monolayers were grown at 37°C and 5% CO₂ in a combined medium made up of ¾ DMEM and ¼ Ham's F12 (Life Technologies), supplemented with FBS, penicillin, and streptomycin at the same doses as above. CF15 cells, human nasal epithelial cells homozygous for the DeltaF508 CFTR mutation,^30,32 were obtained from the laboratory of M. Chanson.^32,33 CF15 were grown as monolayers in collagen I-coated flasks (BioCoat; BD Biosciences) at 37°C and 5% CO₂ in CNT-17 medium (CellNTec), supplemented with penicillin and streptomycin.

Adenoviral vectors

Human adenovirus type 5 (HAdV5)-derived vectors HAdV5-GFP and HAdV5-GFP-CFTR have been described and characterized previously.^28,29 HAdV5-GFP-CFTR encoded the wild-type allele of the CFTR gene, fused to the 3′ end of the GFP gene. Vector stocks were produced and titrated on HEK-293 cell monolayers, as described previously.^28,29,34,35

Isolation of microvesicles and exosomes from extracellular vesicles

EVs were recovered from the culture medium of Calu-3, A549, and A549 transduced by HAdV5-GFP or HAdV5-GFP-CFTR, using a previously published protocol.²⁷ In brief, cells grown to 80–90% confluency in T150 flasks (5 × 10⁷ cells/flask) were rinsed 3 times with phosphate buffered saline (PBS), and fresh culture medium, preclarified by ultracentrifugation at 100,000 × g for 16 hr at 4°C, was added to the cells. The EVs in the culture medium were then recovered at 24, 48, and 72 hr, by two successive steps of ultracentrifugation. The culture medium was first centrifuged for 5 min at 13,000 × g and 4°C to remove the cell debris, before the first ultracentrifugation step at 30,000 × g and 4°C for 2 hr, to obtain the first pellet originally called MP₃₀ (now identified as MVs). The supernatant was then subjected to a second round of ultracentrifugation at 100,000 × g and 4°C for 2 hr, which gave the second pellet called MP₁₀₀ (now identified as EXos). The MV and EXo pellets were resuspended in 200 μl PBS with gentle mixing, and stored at 4°C. They were used within 10 days in our experiments of EV-mediated cell transfer assays.

Protein assay

Aliquots of each EV sample were kept for protein estimation using the Bradford Reagent (Sigma-Aldrich Corp.). Triplicates of MV and EXo preparations, usually diluted 1:10 to 1:20 in PBS, were lysed with 0.01% SDS at room temperature for 30 min, and then reacted with the Bradford reagent, according to manufacturer's recommendations. Values were extrapolated from a standard curve of serially diluted bovine serum albumin (BSA) samples.

Titration of microvesicles and exosomes by nanoparticle tracking analysis

The concentrations of MV and EXo preparations were subjected to NTA, using the NanoSight LM10 system (Malvern Instruments Ltd.), equipped with a 405 nm laser and a high-sensitivity digital camera (Scientific CMOS trigger camera). Videos were collected and analyzed using the NTA software (version 2.3), with the minimal expected particle size, minimum track length, and blur setting, all set to automatic. Camera sensitivity and detection threshold were set to maximum (16) to detect small particles. Ambient temperature was recorded manually, ranging from 22°C to 27°C. Each sample was diluted in sterile PBS (Life Technologies) to obtain concentration range between 10⁸ and 10⁹ EVs/ml, and the number of completed tracks per sample was between 500 and 1500. The screen gain was 6 for EXos and 12 for MVs, with a capture time of 60 sec.

Gel electrophoresis and Western blot analysis

Polyacrylamide gel electrophoresis of SDS-denatured protein samples (SDS-PAGE), and immunoblot analysis have been described in previous studies.^36

–39 Briefly, samples were denatured in SDS/beta-mercaptoethanol-containing loading buffer at 100°C for 2 min, and proteins electrophoresed in SDS-denaturing 10% polyacrylamide gel and electrically transferred to nitrocellulose membrane (Hybond-C-extra; GE Healthcare Life Sciences). Blots were blocked in 5% skimmed milk in Tris-buffered saline (TBS) containing 0.05% Tween-20 (TBS-T), rinsed in TBS-T, and then successively incubated with primary antibody and peroxidase-labeled secondary antibodies. Blots were reacted with SuperSignal West Pico chemiluminescence substrate (Thermo Fisher Scientific Inc.), and exposed to autoradiographic film (Hyperfilm MP; GE Heathcare Bio-Sciences). The apparent molecular weights of proteins were estimated by comparison with prestained protein markers (Precison Plus Protein Standards, Dual Color; Bio-Rad Laboratories, Inc.).

Antibodies and reagents

Mouse monoclonal antibody against the human CFTR C-terminus (mouse IgG2A; Clone 24-1; MAB25031) was from R&D Systems. Goat polyclonal anti-CD63 antibody was purchased from Santa Cruz Biotechnology (Catalog No. SC-15363). Mouse monoclonal anti-GFP antibody (mixture of clones 7.1 and 13.1) was from Roche Applied Science. Peroxidase-coupled monoclonal anti-β-actin antibody was purchased from Sigma-Aldrich (Catalog No. A3854). Peroxidase-labeled rabbit antigoat IgG antibody (Catalog No. A5420) and sheep antimouse IgG antibody (Catalog No. A5906) were also purchased from Sigma-Aldrich.

Metabolic inhibitors of endocytic pathways included two inhibitors of the clathrin-dependent endocytosis, chlorpromazine and the lysosomotropic agent chloroquine; NH4Cl, an inhibitor of the acidification of the endosomes; Bafilomycin A1, a proton-pump inhibitor that blocks the vacuolar ATPase and the endosome/lysosome fusion; the isoflavone drug genistein and the polyene macrolide antibiotic filipin III, two inhibitors of the clathrin-independent, caveolin-mediated endocytosis; cytochalasin B (CytB), nocodazole (NDZ), and latrunculin A (LatA), which all disorganize the cell cytoskeleton, and have individual preferential negative impacts on endocytosis (CytB), intracytoplasmic trafficking (NDZ), and macropinocytosis (LatA); and the proteasome inhibitor MG132. All drugs were purchased from Sigma-Aldrich and prepared according to the manufacturer's instructions. They were used under sterile conditions, and at concentrations that did not decrease the cell viability to levels lower than 75%, as determined by the MTT assay (not shown). For testing the inhibitor effects, cell monolayers, taken at 80% confluence, were pretreated with each drug for 1 hr at 37°C before MVs^GFP or EXos^GFP addition. After a 6 hr incubation at 37°C with EVs (2 × 10⁴ EVs/cell) in the presence of the drug, the cells were collected, fixed, and analyzed by flow cytometry.

EV-mediated transfer of GFP and GFP-CFTR into target cells

In a typical experiment, samples of GFP-loaded EVs (MVs^GFP and EXos^GFP) and GFP-CFTR-loaded EVs (MVs^GFP-CFTR and EXos^GFP-CFTR) in PBS were mixed with an equal volume of prewarmed serum-free medium, and 50–100 μl aliquots added to A549 or CF15 cell monolayers (2 × 10⁵ cells/well). The total input (50–100 μl) contained 5–50 μg protein and 2 × 10⁹ to 10 × 10⁹ EVs, with transducing doses ranging from 5 × 10³ to 5 × 10⁴ EVs/cell. The cells were further incubated at 37°C, and the GFP signal was monitored in live cells using a Zeiss Axiovert-135 inverted microscope equipped with an AxioCam digital camera. Cells were harvested at different time periods, and GFP-positive or GFP-CFTR-positive cells were quantitated using flow cytometry. CFTR-associated chloride channel activity was measured using an iodide-selective microelectrode, as described below.

DNA and RNA isolation, and PCR analyses

DNA from cells and EVs was extracted using the QIAamp DNA Mini Kit (Qiagen). Total RNA was extracted from cells using the Nucleospin RNA II kit (Macherey Nagel), and from EVs using the QIAamp viral RNA Mini kit (Qiagen). Aliquots (100 ng) of RNA were reverse transcribed using the Iscript DNA synthesis kit (BioRad). Real-time PCR (RT-PCR) was performed using the KAPA Sybr Fast qPCR kit (KAPA Biosystems) and the StepOnePlus apparatus (Thermofisher). Quantitative RT-PCR (qRT-PCR) analysis of RNA was performed using the protocol described in ref.⁴⁰ and the ribosomal protein RPL27 RNA transcript as internal standard, amplified with the following primers: sense 5′-ATCGCCAAGAGATCAAAGATAA-3′; antisense 5′- TCTGAAGACATCCTTATTGACG-3′. For the CFTR gene, a PCR-amplified fragment of 415 nucleotides (nt) in length, overlapping exons 3 and 5 of the CFTR gene, was obtained with the following primers⁴¹: sense 5′-AGAATGGGATAGAGAGCTGGCTTC-3′ (exon 3); antisense 5′-TTCATCAAATTTGTTCAGGTTGTTG 3′ (exon 5). In the case of the GFP-CFTR fusion, a fragment of 196 nt in length, overlapping the GFP and CFTR junction sequence,^28,29 was obtained using a forward primer designed from the 3′ end of the GFP gene (nucleotide position 621; 5′-AACGAGAAGCGCGATCACATG), and an antisense primer designed from the 5′ end of the CFTR gene (nucleotide position 78 in the CFTR gene; 5′-GCGCTGTCTGTATCCTTTCCTCAA). For the GFP gene, a fragment of 99 nt within the GFP gene was obtained using the following primers: sense 5′- ATGGTGAGCAAGGGCGAGGA-3′; antisense 5′-CTCGCCGGACACGCTGAACT-3′. For the HAdV5 hexon gene, a fragment of 465 nt was obtained using the following primers: forward 5′- GCTGTATTTGCCCGAC-3′; reverse 5′-CATGGCCTCAAGCGTG-3′, corresponding to the genome nt 20,276–20,291 and 20,726–20,741, respectively.

Iodide efflux assay

The iodide efflux assay, which reflected the activity of the CFTR channel, was measured using an iodide-selective microelectrode.⁴² The experimental setup included the MicroIodide Ion Electrode (LIS-146ICM; Lazar Research Laboratories), connected to an ORP-meter (ORP Model 62; Jenco Instruments Inc.), and the protocol described by Aleksandrov et al.⁴³ Replicates of cell monolayers were loaded with sodium iodide by incubation for 30 min at 37°C in prewarmed iodide-loading buffer (ILB:136 mM NaI, 3 mM KNO₃, 2 mM Ca[NO₃]₂, 11 mM glucose, 20 mM HEPES, pH 7.4 with NaOH). Cells were then washed 10 times with efflux buffer (EB; 136 mM NaNO, 3 mM KNO₃, 2 mM Ca[NO₃]₂, 11 mM glucose, 20 mM HEPES, pH 7.4 with NaOH) to completely remove extracellular iodide. The CFTR stimulation mixture (10 μM forskolin, 100 μM dibutyryl-cAMP, 1 mM 3-isobutyl-1-methylxanthine [IBMX] as enhancer of intracellular cAMP) was added at time 0, to activate efflux through CFTR channels. The values of the voltage generated by iodide in solution were extrapolated from the standard curve generated by measuring the voltages of standard NaI solutions. Iodide efflux values, expressed in nmol/min, represented the mean ± SEM of the amount of iodide released from the cells during each 1 min interval (n = 3). At the end of each assay, efflux buffer containing 0.1% NP-40 was added to release the whole bulk of intracellular iodide.

Flow cytometry

Cells were seeded in complete media in 24- or 48-well plates to obtain a confluency of 70–80% the following day. The cells were then incubated with MVs or EXos at 37°C and further incubated for different lengths of time. At the end of the incubation periods, the cells were rinsed with PBS, and detached with PBS containing 0.05% trypsin and 1 mM Na₂EDTA. After centrifugation, the cells were fixed in PBS containing 1% paraformaldehyde, and analyzed in an LSRII flow cytometer (Becton Dickinson Biosciences).

Cell imaging

Fluorescence microscopy

Cell monolayers were fixed with 1% paraformaldehyde in PBS, and permeabilized with 0.1% (v/v) Triton X-100 in PBS. Cells were then blocked in 1% BSA in PBS for 1 hr at room temperature and incubated with primary monoclonal antibodies against Rab5 (clone C8B1; Cell Signalling), Rab4a (D-20; Santa Cruz Biotech), Rab11 (clone H-87; Santa Cruz Biotech), and Lamp1 (or CD107a; BD Biosciences), following dilutions recommended by the manufacturer. Samples were then rinsed in PBS and incubated with Alexa Fluor 568-conjugated goat antimouse IgG antibody, diluted 1:1000 (Molecular Probes, Invitrogen) for 1 hr at room temperature, rinsed, and treated with DAPI before being mounted on slides. Observations were performed using an Axioimager Z1 epifluorescence microscope (Carl Zeiss France) and a confocal microscope (LSM710; Carl Zeiss). Images were analyzed using the Zen software (Carl Zeiss).

Electron microscopy

For negative staining of MVs and EXos, samples were adsorbed on gold grids, and negatively stained with 2% uranyl acetate before EM observation.⁴⁴ For ultrathin sections, cell samples incubated with MVs or EXos were fixed with 2% glutaraldehyde in 0.1 M sodium cacodylate buffer pH 7.4, pelleted, and postfixed with osmium tetroxide (1% in 0.1 M cacodylate buffer, pH 7.4). The specimens were dehydrated and embedded in Epon resin, sectioned, and processed as previously described.^27,38 Ultrathin sections were examined under a JEM-1400 Jeol electron microscope equipped with an Orius-Gatan digital camera (Gatan France, 78113 Grandchamp).

Statistics

Results were expressed as mean ± SEM of n observations. Comparison between two samples was performed using the Student's t test (Excel); multiple samples were compared using the one-way ANOVA test. Differences were considered statistically significant when p < 0.05.

Results

Isolation, characterization, and nomenclature of EVs recovered from human cells

Human epithelial cell lines Calu-3 (human bronchial submucosal glandular cells) and A549 (human adenocarcinoma alveolar basal epithelial cells) were tested as EV-producing cells. Calu-3 cells are considered as a pertinent in vitro model of respiratory functions and chloride channel activity.⁴⁵ On the other hand, A549 cells are permissive to human adenovirus serotype 5 (HAdV5), and to the adenoviral vectors HAdV5-GFP or HAdV5-GFP-CFTR, which express the green fluorescent protein (GFP) or GFP-CFTR, a GFP-fused version of the human CFTR glycoprotein.^28,29 EVs that are naturally released in the culture supernatant of A549 and Calu-3 cells were separated into two populations, using a two-step ultracentrifugation procedure. The first fraction, MP₃₀, was recovered in the pellet after centrifugation at 30,000 × g, and the second fraction, MP₁₀₀, was pelleted after a second round of centrifugation at 100,000 × g. The two EV populations could be distinguished by their morphology under the EM, their size, and protein surface markers.

As exemplified by EV samples from A549 cells, MP₃₀ consisted of large particles characterized by their size heterogeneity, ranging from 200 to 1200 nm in diameter (Fig. 1a, left panel). The MP₁₀₀ population, however, showed more homogenous regular particles, rather spherical and relatively small in size, ranging from 15 to 60 nm (Fig. 1a, right panel). Size measurements were performed using dynamic light scattering (DLS), and representative DLS patterns obtained with EVs from Calu-3 cells are shown in Fig. 1b. A similar range of values as those estimated by EM was found for both types of EVs: 100–1000 nm for the diameter of MP₃₀ (320 ± 280 nm; mean ± SEM; n = 2,272 k counts/s), and 20–200 nm (65 ± 24 nm; n = 1,770 k counts/s) for MP₁₀₀ (Fig. 1b, compare the left and right panels).

Figure 1.

Morphological, biophysical, and biochemical characterization of extracellular vesicles (EVs) isolated from human cells. (a) Electron microscopy (EM). MP₃₀ pellet (microvesicles [MVs], left panel) and MP₁₀₀ pellet (exosomes [EXos], right panel) were isolated from A549 cell culture medium by differential ultracentrifugation at 30,000 × g and 100,000 × g, respectively, and negatively stained for observation under the EM. (b) Diffusion light scattering (DLS). Populations of MP₃₀ (MVs, left panel) and MP₁₀₀ (EXos, right panel) isolated from Calu-3 cell culture medium were analyzed by DLS. (c) Protein content. MP₃₀ and MP₁₀₀ pellets isolated from Calu-3 cells were analyzed by SDS-PAGE and Western blotting, using anti-CD63 antibody (left panel), or anti-CFTR antibody (right panel); β-actin (42 kDa) was used as internal control of protein loads. Lane mm, protein markers, with their molecular masses indicated in kDa. CFTR, CF transmembrane conductance regulator. Color images available online at www.liebertpub.com/hum

MP₃₀ and MP₁₀₀ issued from Calu-3 cells were also analyzed with respect to their CD63 content, a specific marker of the exosomal compartment.⁴⁶ Only the MP₁₀₀ fraction, and not MP₃₀, was found to contain CD63 (Fig. 1c, left panel), indicating that MP₁₀₀ fraction was essentially composed of exosomes, whereas the MP₃₀ fraction contained a majority of MVs. Thus, for the rest of this study, the terminology of microvesicles (abbreviated MVs) and exosomes (abbreviated EXos) was used instead of MP₃₀ and MP₁₀₀, respectively. The term “extracellular vesicles” (EVs) included both MVs and EXo populations.

Production and recovery of extracellular EVs

Culture supernatants of Calu-3 or A549 cells were collected three times over a period of 72 hr, successively at 24, 48, and 72 hr. After pooling the three samples, the recovery of MVs and EXos was assayed by nanoparticle tracking analysis (NTA). From 2.5 × 10⁸ EV-producing cells, we obtained 4.92 ± 0.96 × 10¹⁰ MVs and 10.8 ± 2.1 × 10¹⁰ EXos with A549, and 4.46 ± 0.59 × 10¹⁰ MVs (mean ± SEM; n = 6) and 8.20 ± 1.35 × 10¹⁰ EXos with Calu-3 cells. The yields did not significantly differ for the two types of EV-donor cells, which corresponded to about 200 MVs per cell and 300–400 EXos per cell, over a 72 hr period culture. After resuspension in 200 μl PBS, the EV titer was ∼2.5 × 10⁸/μl for MVs, and 5 × 10⁸/μl for EXos. The protein concentrations ranged between 0.25 and 0.5 μg/μl, that is, a ratio of 2 × 10⁸ MVs or EXos per μg protein. This ratio was consistent with the values reported in the literature for exosomes preparations obtained by ultracentrifugation and titrated by NTA (6 × 10⁸ EXos/μg; see ref.⁴⁷).

EV-mediated delivery of chloride channel activity to human CF cells

In an earlier study, we generated a CHO cell line that stably expressed the GFP-CFTR protein and released MVs and EXos containing the GFP-CFTR glycoprotein and GFP-CFTR-encoding mRNA (mRNA^GFP-CFTR). We showed that both types of molecules were transferable to target CHO cells.²⁷ Because Calu-3 cells are known to naturally express CFTR at relatively high levels,⁴⁵ we assumed that MVs and EXos isolated from Calu-3 would contain high amounts of CFTR and mRNA^CFTR. This was confirmed by SDS-PAGE and Western blot analysis, which showed that CFTR glycoprotein was detectable in both MVs and EXos, but in higher amounts in MVs compared with EXos (Fig. 1c, right panel). RT-PCR performed on MV and EXo samples normalized to the same EV titer showed that both types of EVs contained mRNA^CFTR, but MVs had an apparent higher mRNA^CFTR content, compared with EXos (Fig. 2a).

Figure 2.

Characterization of Calu-3 as EV-donor cells. (a) RNA^cftr content of EVs. MP₃₀ (MVs) and MP₁₀₀ (EXos) isolated from Calu-3 cell culture medium were analyzed by RT-PCR, using CFTR-specific primers. Control MP₁₀₀ and control MP₃₀ corresponded to the fractions recovered by ultracentrifugation at 100,000 × g and 30,000 × g, respectively, from medium that had not been in contact with any cells. Control MP₁₀₀ and MP₃₀ were subjected to the same RT-PCR as MP₁₀₀ and MP₃₀ isolated from Calu-3 cell culture media. The 415-nucleotide (nt) fragment amplified from the CFTR mRNA is indicated by arrows. (b) CFTR chloride channel activity in Calu-3 (EV-donor cells) and CF15 cells (EV-recipient cells). Aliquots of CF15 cells (1.5 × 10⁶) were incubated with MVs or EXos isolated from the 72 hr culture medium of Calu-3 (1.5 × 10⁷ cells), at 200 EVs per cell, for 6 hr at 37°C. Control Calu-3 (EV-donor cells), untreated CF15, and MV- or EXo-treated CF15 cells were first loaded with iodide, and cAMP-containing stimulation cocktail was then added at time 0 (arrow) to activate the anion efflux through CFTR channels. The amount of iodide released from the cells was measured at 1 min intervals, using an iodide-selective electrode (mean ± SEM; n = 3). Shown are representative curves of the iodide efflux. (c) CFTR activity ratios. The CFTR Cl⁻ channel activity in Calu-3 and CF15 cells was deduced from experimental curves, as exemplified in (b). The values of CFTR Cl⁻ channel activity determined in untreated CF15, and in MV- or EXo-transduced CF15 cells were expressed as the percentage of the mean value obtained with Calu-3 cells, attributed the 100% value. *p ≤ 0.05; **p ≤ 0.005. Note the log scale of the y axis.

We next determined whether MVs and/or EXos isolated from the Calu-3 cell culture medium (referred to as [Calu-3]MVs and [Calu-3]EXos) were capable of conferring the CFTR-associated chloride channel function to CF15 cells, a human nasal epithelial cells homozygous for the ΔF508 CF mutation.^30,32,33 CF15 cells were incubated for 6 hr at 37°C with [Calu-3]MVs or [Calu-3]EXos at the transducing dose of 200 EVs/cell. Fresh culture medium was then added, and cells were further incubated for 72 hr at 37°C. At the end of this period, the medium was removed, and the cells were tested for the cAMP-dependent CFTR chloride channel activity, measured by the iodide efflux assay using an iodide-selective microelectrode.^42,43

Curves representative of the cyclic cAMP-stimulated iodide efflux response in EV-transduced CF15 cells were compared with the curves of untreated CF15 cells (negative control), and of Calu-3 cells (positive control). A peak of CFTR channel activity was detected in EV-treated CF15 cells, suggesting the restoration of the CFTR function in these target cells (Fig. 2b). The CFTR chloride channel activity measured in EXo-treated CF15 cells was slightly lower, compared with that of MV-treated CF15 cells (Fig. 2b, c). Interestingly, the peak of channel activity was obtained at 4 min in EV-treated CF15 cells, a slight delay compared with the peak of Calu-3 that occurred at 2–3 min (Fig. 2b, c). Even though the levels of CFTR channel activity observed in EV-treated CF15 cells were 10- to 20-fold lower than that of the Calu-3 cells, this positive result incited us to pursue our studies, in an attempt to improve the efficiency of the CFTR transfer in human cells. The factors that could influence the transfer of the CFTR function included the nature of the EV-donor and EV-recipient cells, the EV cargo and the EV doses, and the experimental conditions of EV-cell interaction. These different parameters were investigated in the following experiments.

Characterization of EV-donor cells

Fluorescent EVs loaded with GFP or GFP-tagged CFTR were used to help us track and quantitate the uptake of EVs by target cells.²⁷ We assumed that EVs isolated from A549 cells transduced by HAdV5 vectors overexpressing either the GFP or GFP-CFTR protein would incorporate relatively higher amounts of GFP or GFP-CFTR glycoproteins and their corresponding mRNAs, compared with EVs from Calu-3 cells. SDS-PAGE and Western blot analysis, using anti-CFTR and anti-GFP antibodies, were performed on A549 cell lysates and compared with Calu-3 cell lysates. As expected, the CFTR glycoprotein content of HAdV5-GFP-CFTR-transduced A549 cells could be estimated to be 10–30 times higher than the endogenous CFTR content of Calu-3 cells, after normalization to equivalent protein loads (Fig. 3a).

Figure 3.

Characterization of EV-donor cells. Calu-3 and A549 cells, as indicated at the top of the panels, were probed for their CFTR or GFP-CFTR glycoprotein contents (a), and corresponding mRNAs (b). (a) SDS-PAGE and Western blot analysis. Whole cell lysates were analyzed using anti-CFTR (left panel), or anti-GFP antibodies (right panel). The CFTR band is indicated by a dot, and free GFP by an arrowhead. (b) RT-PCR analysis. RT-PCR was performed using different pairs of primers, specific to the CFTR gene (left panel), or specific to the GFP gene and GFP-CFTR junction (right panel). The amplified fragments were 415 nt in length for CFTR, 99 nt for GFP, and 196 nt for the GFP-CFTR junction, respectively.

RT-PCR amplified a 415 nt fragment from the endogenous mRNA^CFTR present in Calu-3 cells, and the same fragment was amplified from the mRNA^GFP-CFTR transcript coding for the GFP-CFTR fusion protein in HAdV5-GFP-CFTR-transduced A549 cells. No endogenous mRNA^CFTR was detectable in control, nontransduced A549 cells, or in CF15 cells (Fig. 3b, left panel). A 99 nt fragment, amplified from the GFP gene, and a 196 nt fragment, specific of the GFP-CFTR junction, were found only in the RT-PCR samples of HAdV5-GFP-CFTR-transduced A549 cells, and not in the other samples (Fig. 3b, right panel). We therefore used HAdV5-GFP-CFTR-transduced A549 cells as EV-donor cells, and both A549 and CF15 as target cells for homologous and heterologous EV cell transfer, respectively.

Characterization of EVs^GFP and EVs^GFP-CFTR from A549 cells

MVs and EXos prepared from control, nontransduced A549 cells and from HAdV5-GFP- or HAdV5-GFP-CFTR-transduced A549 cells were analyzed for their GFP fluorescent signal by flow cytometry (Fig. 4a). Over 80% MVs and EXos from HAdV5-GFP-transduced A549 cells carried the GFP fluorescent signal (Fig. 4a). In contrast, the populations of EVs recovered from HAdV5-GFP-CFTR-transduced A549 cells contained only 10–12% fluorescent MVs, and 2–3% fluorescent EXos, namely, 4–5-fold less for the latter (Fig. 4a). These results were consistent with the fact that GFP is an abundant cytoplasmic and nuclear protein, whereas GFP-CFTR is a tightly regulated channel protein, localized in specialized microdomains of the cell plasma membrane.^24
–26

Figure 4.

Characterization of EVs isolated from A549 cells. (a) Flow cytometry. MVs and EXos were isolated from the culture medium of untreated A549 cells used as negative controls, A549 cells transduced with HAdV5-GFP, and A549 cells transduced with HAdV5-GFP-CFTR at 72 hr posttransduction (pt). The isolated MVs and EXos were analyzed for the intensity of the GFP signal by flow cytometry. (b) GFP-CFTR glycoprotein content. MVs and EXos isolated from HAdV5-GFP-CFTR-transduced A549 cells at 72 hr pt were probed for GFP-CFTR, using SDS-PAGE and Western blot analysis with antibodies (Ab) against CFTR (left panel), or GFP (right panel). Lane WCL, whole cell lysates were used as positive controls. Lane mm, molecular markers. (c) RNA^GFP-CFTR content (RT-PCR). MVs and EXos from untreated A549 or HAdV5-GFP-CFTR-transduced A549 cells were collected at different times pt, and subjected to RT-PCR analysis, using CFTR-specific primers. (d) Quantitative RT-PCR of RNA^GFP-CFTR. The qRT-PCR analysis of MVs and EXos was performed using the pair of primers amplifying the GFP-CFTR junction (196 nt fragment), with RPL27 RNA as the internal standard.

Samples of MVs^GFP-CFTR and EXos^GFP-CFTR were then analyzed by SDS-PAGE and Western blotting, using anti-GFP or anti-CFTR antibodies to detect the GFP-CFTR fusion protein (Fig. 4b). The GFP-CFTR protein signal in MVs^GFP-CFTR was estimated to be 10–20-fold higher, on a per-EV basis, compared with that of EXos^GFP-CFTR (Fig. 4b). Of note, a single band of the CFTR protein migrating with an apparent molecular mass of 210–220 kDa was visible in the EXos^GFP-CFTR fraction, which was compatible with that of the GFP-fused mature CFTR glycoprotein (25 + 190 kDa). In the MVs^GFP-CFTR samples, however, several bands were detected, including a major doublet band at 200–220 kDa, and extra bands of immature CFTR and/or cleavage products (Fig. 4b).

MVs^GFP-CFTR and EXos^GFP-CFTR isolated from the A549 cell culture medium at different times after transduction with HAdV5-GFP-CFTR were then probed for their mRNA^GFP-CFTR content. RT-PCR analysis, performed using primers overlapping the GFP-CFTR junction, or primers within the CFTR gene sequence, suggested that the maximum incorporation of mRNA^GFP-CFTR into both MVs and EXos was attained during the first 48 hr after cell transduction with HAdV5-GFP-CFTR (Fig. 4c). This was confirmed by quantitative RT-PCR, which also showed that the mRNA^GFP-CFTR content was twofold higher in MVs^GFP-CFTR, compared with EXos^GFP-CFTR (Fig. 4d).

Because MVs^GFP-CFTR and EXos^GFP-CFTR were isolated from A549 cells transduced by the HAdV5-GFP-CFTR vector, we determined whether these EVs could also contain adenoviral vector DNA or fragments thereof. Quantitative PCR using primers specific of the GFP-CFTR gene junction was performed on MVs^GFP-CFTR and EXos^GFP-CFTR samples harvested at 24 and 48 hr after HAdV5-GFP-CFTR transduction. Vector DNA was found in both types of EVs, and their DNA content increased with time posttransduction. There was no significant difference in vector DNA content between MVs^GFP-CFTR and EXos^GFP-CFTR (Supplementary Fig. S1a; Supplementary Data are available online at www.liebertpub.com/hum). To further characterize the vector DNA incorporated in EVs, MVs^GFP-CFTR and EXos^GFP-CFTR were tested by plaque assays on HEK-293 cells, a trans-complementing cell line for rescuing E1A+E1B-deleted HAdV5 vectors such as HAdV5-GFP-CFTR. No viral plaques were obtained at any of the EV doses tested, implying that no viable adenoviral vector could be rescued from the EVs (not shown).

Cellular uptake of MVs^GFP-CFTR and EXos^GFP-CFTR by target cells

EV dose dependence of the GFP-CFTR transfer

MVs^GFP-CFTR and EXos^GFP-CFTR from HAdV5-GFP-CFTR-transduced A549 cells were added to A549 or CF15 cell monolayers (5 × 10⁵ cells per well) at different transducing doses, ranging from 5 × 10³ to 4 × 10⁴ EVs/cell, and incubated for 6 hr at 37°C. The cells were then harvested, and cell-associated fluorescent signal was analyzed by flow cytometry. There was a progressive increase in GFP-positive cells from 5% to 40% with increasing doses of MVs^GFP-CFTR. A dose-dependent progressive augmentation of GFP-positive cells was also observed with EXos^GFP-CFTR, reaching a maximum of 10% GFP-positive cells at the highest dose tested (Fig. 5a).

Figure 5.

Dose dependence, cytotoxicity, and temperature dependence of EV-mediated transduction of target cells. (a) EV dose dependence. A549 cells were incubated with increasing doses of MVs^GFP-CFTR or EXos^GFP-CFTR, and the cellular GFP signal was analyzed by flow cytometry (mean ± SEM; n = 4). (b) Cytotoxicity. A549 or CF15 cells incubated with MVs^GFP-CFTR or EXos^GFP-CFTR at different doses were analyzed for cell survival using the MTT assay. Data shown are the percentages of viable cells compared with untreated samples. (c) Temperature dependence. Target cells, A549 (homologous EV-to-cell transfer; top panels) or CF15 (heterologous EV-to-cell transfer; bottom panels), were incubated with samples of MVs^GFP-CFTR or EXos^GFP-CFTR (as indicated under the x axis) at the indicated temperatures. Cells were collected after 6 hr incubation, and analyzed by flow cytometry to quantitate the percentage of GFP-positive cells (left panels), and the transduction efficiency (right panels). The transduction efficiency (TE) was expressed as arbitrary units (AU), using the following formula: TE = MFI × % GPF-positive cells. MFI, mean fluorescence intensity.

Cytotoxicity of MVs^GFP-CFTR and EXos^GFP-CFTR

The potential cytotoxicity of MVs^GFP-CFTR and EXos^GFP-CFTR was tested at different doses on A549 and CF15 cells, using MTT assays. The viability of EXos^GFP-CFTR-transduced A549 cells decreased to 60% at the highest dose of 4 × 10⁴ EVs/cell, but the other cell samples showed more that 80% viable cells within the range of 2 × 10⁴ to 4 × 10⁴ EVs/cell. Interestingly, the CF15 cells showed the highest degree of tolerance to both MVs^GFP-CFTR and EXos^GFP-CFTR, with no significant cytotoxity detected at the highest dose of 4 × 10⁴ EVs/cell (Fig. 5b).

Temperature dependence

We next explored the effect of the incubation temperature on the cellular attachment and uptake of EVs in both homologous and heterologous systems. Aliquots of MVs^GFP and EXos^GFP from HAdV5-GFP-transduced A549 cells were added to A549 or CF15 cell monolayers (2 × 10⁴ EVs per cell), and cells were incubated for 2 hr at 37°C, 22°C, and 4°C, respectively. At the end of this period, cells were collected, fixed, and analyzed by flow cytometry. The uptake of MVs^GFP and EXos^GFP was temperature dependent for both types of cell targets, with a progressive decrease of the EV-mediated transduction with the lowering of the incubation temperature (Fig. 5c). In comparison to MVs^GFP-transduced A549 cells at 37°C, a reduction of about 25% was observed at 22°C, and 50% at 4°C. In MVs^GFP-transduced CF15 cells, the reductions were 50% and 90% at 22°C and 4°C, respectively. Because of the relatively low values at 37°C, the variations with the temperature lowering were less pronounced for the cellular uptake of EXos^GFP by both targets (Fig. 5c). In addition, the proportion of transduced cells and the transduction efficiency were significantly higher in the homologous cellular uptake compared with the heterologous uptake at all temperatures tested (Fig. 5c; compare upper and lower pairs of panels). These results suggested that the cellular uptake of A549-issued MVs^GFP and EXos^GFP by A549 and CF15 cells was a temperature-dependent, dynamic process, which required an optimal fluidity of the membrane lipid bilayers. However, the absence of total inhibition of cellular uptake of EVs at 4°C suggested that a temperature-independent mechanism might also be involved, such as the recognition between receptor and/or adaptor molecule(s) between the EV membrane and cell plasma membrane.

Homologous versus heterologous cellular uptake

To evaluate the efficiency of homologous and heterologous EV-to-cell transfer, GFP-tagged EVs (MVs^GFP and EXos^GFP) were prepared from HAdV5-GFP-transduced A549 cells. The advantage of using GFP-loaded EVs was the high proportion of fluorescent EVs for GFP (over 80%), compared with GFP-CFTR (around 10%; refer to Fig. 4a). For homologous cellular uptake of EVs, EVs isolated from A549 cells were incubated with A549 cells, whereas for heterologous cellular uptake, EVs isolated from A549 cells were incubated with CF15 cells. Incubation was carried out at 37°C for 1, 2, 3, 4, 6, 12, and 24 hr, at an input of 2 × 10⁴ EVs/cell. At each time point, the target cells were fixed for fluorescence microscopy and assayed by flow cytometry for the proportion of GFP-positive cells and the index of transduction efficiency (TE). The TE, expressed in arbitrary units (AU), was calculated using the following formula: TE = GFP-positive cells (%) × MFI.³⁴

Homologous cellular uptake (Fig. 6a)

About 60% of A549 cells were GFP-positive after 3 hr incubation with MVs^GFP, and 75% after 6 hr. The proportion of GFP-positive cells reached a plateau of 80% after 12 hr. The values obtained for EXos^GFP-transduced cells were lower, with 5% and 10% at 3 and 6 hr, respectively, and a plateau at 25% after 12 hr. The transduction efficiency was significantly higher with MVs^GFP, compared with EXos^GFP, with values ranging from 4- to 8-fold higher at all time points (Fig. 6a, upper and lower leftmost panels). Fluorescence microscopy observation showed that GFP dots were more abundant in MVs^GFP-transduced cells than in EXos^GFP-transduced cells at all time points tested (Fig. 6a, upper and lower rightmost panels).

Figure 6.

Homologous versus heterologous EV-mediated cell transduction. (a) Homologous EV-to-cell transfer. A549 cells were incubated with MVs^GFP or EXos^GFP isolated from HAdV5-GFP-transduced A549 cells, collected at different times, and analyzed by flow cytometry (left panels), and epifluorescence microscopy (right panels). (b) Heterologous EV-to-cell transfer. CF15 cells were incubated with MVs^GFP or EXos^GFP isolated from HAdV5-transduced A549 cells, collected at different times as in (a) and analyzed by flow cytometry (left panels), and epifluorescence microscopy (right panels). Data presented in the left panels are mean ± SEM (n = 3). The images of epifluorescence microscopy shown in the right panels were taken after 24 hr incubation.

Heterologous cellular uptake (Fig. 6b)

The uptake of heterologous EVs was slower, and its efficiency was significantly lower (about 2-fold), compared with homologous EV uptake: only 40% MVs^GFP-transduced CF15 cells were found to be GFP-positive at 12 hr, and a maximum of 50% at 24 hr. This difference was also observed with EXos^GFP-transduced CF15 cells, which showed no fluorescent signal over the background before the 4 hr time point, and a maximum value at 5% GFP-positive cells at 12 and 24 hr. As observed with A549 cells, the difference in transduction efficiency between MVs^GFP and EXos^GFP ranged between 6- and 8-fold within the late time period (Fig. 6b, upper and lower leftmost panels), which was confirmed by fluorescence microscopy (Fig. 6b, upper and lower rightmost panels). These results suggested that the cellular uptake of EVs was more efficient in homologous EV-to-cell transfer system, compared with the heterologous type, with MVs being more efficient than EXos in both cases.

Mechanisms of cellular uptake of EVs

Cell entry inhibitors

Metabolic inhibitors^48,49 were used to further explore the cell entry pathway of MVs^GFP and EXos^GFP in homologous and heterologous transfer systems. Only cytochalasin B, a cytoskeletal drug that acts as an endocytosis blocker,^48,49 and the macropinocytosis inhibitor amiloride⁵⁰ provoked a significant diminution (about twofold) of the GFP signal in EV-treated cells. This suggested that clathrin-dependent endocytosis and macropinocytosis were involved to some extent in the cellular uptake of EVs. The effects observed with the other drugs were moderate, and varied with the cell types and the nature of the EVs, in the negative or positive sense (data not shown). This implied that MVs and EXos used different mechanisms and/or alternative internalization pathways, of which the relative importance would depend on the nature of EVs and the type of target cells. In the presence of a given metabolic inhibitor responsible for the blockade of one pathway, the EVs might be redirected to alternative routes for cell entry, which, paradoxically, would result in an apparent higher transduction efficiency. This confirmed the diversity of mechanisms of cellular attachment and uptake of EVs previously observed.^21,51

Imaging of EV–cell interaction by electron microscopy

Monolayers of A549 cells were incubated with A549-issued MVs^GFP or EXos^GFP (2 × 10⁴ EVs/cell) at 37°C for 2 hr, and the cells were fixed and processed for EM. For both MVs and EXos, EM observation showed close contacts between EVs and the cell plasma membrane (Supplementary Fig. S2a, b), and between intravesicular EVs and the vesicular membrane (Supplementary Fig. S2d, f). Electron-dense material connecting MVs and EXos with plasma and vesicular membranes was also visible. These images were reminiscent of a receptor–ligand recognition, usually followed by cell internalization via membrane fusion, or alternatively via endocytosis (Supplementary Fig. S2d, f). Images evoking macropinocytosis were also observed with both MVs and EXos (Supplementary Fig. S2c, e).

Confocal fluorescence microscopy of EV–cell interaction

The intracellular trafficking of EVs after their cellular uptake was investigated using GFP-loaded EVs, and antibodies against markers of the early, intermediate, and late vesicular compartments of the endocytic pathway. These included antibodies against Rab5, Rab4a, Rab11, and Lamp1. A549 cells were incubated with MVs^GFP and EXos^GFP at an input of 2 × 10⁴ EVs/cell, harvested at 2, 6, and 24 hr posttransduction, permeabilized, and reacted with primary antibodies followed by secondary, red-fluorescent antibodies. The cell samples were then processed for confocal fluorescence microscopy. No colocalization with Rab5, Rab4a, or Rab11 was detectable with MVs^GFP or EXos^GFP at any of the time points considered (not shown). At 24 hr posttransduction, however, some degree of colocalization of EXos^GFP and Lamp1 was observed (Supplementary Fig. S3a). Colocalization with Lamp1 was not observed with MVs^GFP, even at late times posttransduction (Supplementary Fig. S3b). These observations indicated that the MVs and EXos did not accumulate in the same vesicular compartments at late times after cell entry, and suggested that MVs and EXos differed in their intracellular trafficking.

Kinetics of EV-mediated delivery of GFP-CFTR to CF target cells

MVs^GFP-CFTR and EXos^GFP-CFTR isolated from HAdV5-GFP-CFTR-transduced A549 cells were incubated with CF15 target cells at 2 × 10⁴ EVs/cell for 6 hr at 37°C, and cells collected at time points 6, 12, 24, and 48 hr after transduction. At 6 hr posttransduction, the percentage of GFP-CFTR-positive cells was 10–15% for MVs^GFP-CFTR-transduced cells and 5–6% for EXos^GFP-CFTR-transduced cells. At 12 hr, the percentage of GFP-CFTR-positive cells decreased for both types of EVs, with only 3% fluorescent cells with MVs^GFP-CFTR, and less than 2% for EXos^GFP-CFTR (Fig. 7a, left panel). Interestingly, the proportion of GFP-CFTR-positive cells progressively increased at later times posttransduction with both types of EVs, with 8–9% GFP-CFTR-positive cells at 48 hr for MVs^GFP-CFTR-transduced cells, and 5–6% for EXos^GFP-CFTR-transduced cells (Fig. 7a, left panel). The TE values followed a similar U-shaped pattern (Fig. 7a, right panel), and the transduction efficiency was consistently 5–10-fold higher with MVs^GFP-CFTR than with EXos^GFP-CFTR.

Figure 7.

Kinetics and efficiency of EV-mediated transfer of GFP-CFTR into target cells. (a) Flow cytometry. CF15 cells were incubated with MVs^GFP-CFTR or EXos^GFP-CFTR isolated from HAdV5-GFP-CFTR-transduced A549 cells for different time periods at 37°C. Cell samples were collected at times indicated in the x axis, and analyzed for the percentage of GFP-CFTR-positive cells (left panel), and transduction efficiency (right panel). (b) Quantitative RT-PCR analysis. The RNA^GFP-CFTR content was determined by qRT-PCR, using a pair of primers overlapping the GFP-CFTR junction (amplified fragment: 196 nt). The RPL27 RNA was used as the internal standard. Data shown are GFP-CFTR to RPL27 RNA ratios (left panel). Normalized values of GFP-CFTR/RPL27 RNA ratios, with the 100% value attributed to the maximal ratio obtained (right panel). The double-head arrow represents the estimated half-life of RNA^GFP-CFTR molecules in MV- and EXo-transduced cells, respectively (mean ± SEM; n = 3).

The transfer of mRNA^GFP-CFTR molecules to target cells was evaluated by RT-PCR and qRT-PCR analyses, using primers overlapping the GFP and CFTR sequences, and RPL27 RNA as the internal control. Reaction was performed on samples of MVs^GFP-CFTR- and EXos^GFP-CFTR-transduced CF15 cells, harvested at different time points after EV-transduction. Both types of samples contained mRNA^GFP-CFTR as early as 6 hr posttransduction (Fig. 7b, left panel). Interestingly, in MVs^GFP-CFTR-transduced cells, the number of mRNA^GFP-CFTR copies increased between 6 and 24 hr posttransduction and remained stable until 48 hr, before decreasing progressively. In EXos^GFP-CFTR-transduced cells, however, no plateau was observed between 24 and 48 hr, and the mRNA^GFP-CFTR content rapidly decreased after 24 hr, to almost undetectable levels at 72 hr. The difference in mRNA^GFP-CFTR levels between MVs^GFP-CFTR- and EXos^GFP-CFTR-transduced CF15 cells was estimated to be 5–7-fold between 6 and 24 hr, and 25–30-fold between 48 and 72 hr (Fig. 7b, left panel).

The half-life of intracellular mRNA^GFP-CFTR transferred, deduced from the kinetic curves, was t _½ = 48 hr for MVs^GFP-CFTR, and t _½ = 16 hr for EXos^GFP-CFTR (Fig. 7b, right panel), that is, three times shorter in EXos^GFP-CFTR-transduced cells compared with MVs^GFP-CFTR-transduced cells. These results suggested that the cellular uptake of heterologous EVs by the target cells CF15 was a progressive process that prolonged until 24 hr after the initial EV-cell contact. These results also suggested that the mRNA^GFP-CFTR molecules conveyed by MVs^GFP-CFTR into CF15 cells did not follow the same metabolic pathway as those conveyed by EXos^GFP-CFTR, and/or did not accumulate in the same cellular compartments at late times after transfer. These two possibilities, which are not mutually exclusive, were consistent with the observation by confocal fluorescence microscopy, suggesting different cell trafficking pathway for MVs and EXos (refer to Supplementary Fig. S3).

Vector DNA, carried over by EVs, was also detected in EV-transduced CF15 cells. Quantitative PCR analysis using primers specific of the GFP-CFTR junction showed that there was 50–80-fold less vector DNA in EXos^GFP-CFTR-transduced cells compared with MVs^GFP-CFTR-transduced cells (Supplementary Fig. S1b). There was no increase in the vector DNA content between 24 and 48 hr posttransduction. There was no detectable adenovirus hexon gene transcript in MVs^GFP-CFTR- or EXos^GFP-CFTR-transduced cells (not shown). These results suggested that the template of the amplified DNA fragments in both EVs and EV-transduced cells consisted of incomplete or fragmented vector DNA, naked or partially uncoated.

Restoration of chloride (Cl⁻) channel function in CF cells transduced by CFTR-loaded EVs

The CF15 cells, which carry the deltaF508 mutation in the CFTR gene, are defective in the CFTR Cl⁻ channel function. The next important issue to address was whether transduction of CF15 cells by CFTR-loaded EVs was able to restore or correct the CFTR Cl⁻ channel function in these cells. We therefore tested for the gain in Cl⁻ channel function in CF15 cells incubated with different doses of MVs^GFP-CFTR or EXos^GFP-CFTR, by measuring at 24 hr postincubation the iodide efflux using an iodide-selective microelectrode.^26,42,43 As positive controls of vector-transduced CFTR activity,^28,29,35 CF15 cells were transduced by HAdV5-GFP-CFTR at several vector doses. At 50 vector particles (vp) per cell, a strong and rapid response of iodide efflux was detectable as early as 1–2 min after activation (Fig. 8a). There was no detectable Cl⁻ channel activity at low vector doses (5 and 10 vp/cell), implying a threshold effect as previously observed in human airway primary CF cells transduced by the same HAdV5-GFP-CFTR vector.^28,29

Figure 8.

Measurement of the CFTR channel activity in EV-transduced CF cells. CF15 cells were incubated with MVs^GFP-CFTR or EXos^GFP-CFTR at increasing transducing doses (1 × 10⁴, 2 × 10⁴, or 4 × 10⁴ EVs/cell). As positive control, CF15 cells were transduced with HAdV5-GFP-CFTR at increasing vector doses (5, 10, or 50 PFU/cell). At 24 hr posttransduction, cells were loaded with iodide, and cAMP-containing stimulation cocktail was added at time 0 (arrow), to activate the anion efflux through CFTR channels. The amount of iodide released from the cells was measured at 1 min intervals, using an iodide-selective electrode (mean ± SEM; n = 3). (a) CF15 cells transduced with HAdV5-GFP-CFTR; (b) CF15 transduced with MVs^GFP-CFTR; (c) CF15 transduced with EXos^GFP-CFTR. Color images available online at www.liebertpub.com/hum

A dose–response study was performed with increasing transducing doses of MVs^GFP-CFTR or EXos^GFP-CFTR at 1 × 10⁴, 2 × 10⁴, and 4 × 10⁴ EVs/cell. For both MVs^GFP-CFTR or EXos^GFP-CFTR, none or low Cl⁻ channel activity (10–20 nmol I⁻/min) was detected at the lowest dose of 1 × 10⁴ EVs/cell. A modest Cl⁻ channel activity was detected with MVs^GFP-CFTR and EXos^GFP-CFTR at the transducing dose of 2 × 10⁴ EVs/cell (10–15 nmol I⁻/min), and more importantly, a relatively high CFTR activity (40–50 nmol I⁻/min) was obtained with MVs^GFP-CFTR and EXos^GFP-CFTR used at 4 × 10⁴ EVs/cell (Fig. 8b, c). Compared with HAdV5-GFP-CFTR-transduced cells, the peak of CFTR activity in EV-transduced cells was about 2- to 3-fold lower (40–50 vs. 160–200 nmol I⁻/min), and slightly delayed (4 min after activation; Fig. 8b, c). Nevertheless, the Cl⁻ channel activity in EV-transduced CF15 cells transduced at 2 × 10⁴ EVs/cell was within the same range of values as those measured in Calu-3 cells, which naturally express endogenous CFTR (refer to Fig. 2b,c). These promising results demonstrated that transduction of CF cells with MVs^GFP-CFTR or EXos^GFP-CFTR was able to restore the CFTR Cl⁻ channel function as early as 24 hr posttransduction. The CFTR Cl⁻ channel activity in these EV-transduced CF15 cells was found to persist at detectable levels up to 5 days (not shown).

Discussion

In an earlier study, we demonstrated that EVs isolated from CHO cells modified to express transmembrane glycoproteins such as the Coxsackie adenovirus receptor (CAR), CD46, and the more complex dodecaspanin CFTR had the capacity to transfer the CAR-, CD46-, and CFTR-encoding mRNA molecules to naive CHO cells. This EV-to-cell transfer resulted in the gain of specific functions associated with CAR, CD46, and CFTR in the target CHO cells.²⁷ In the present study, we investigated the capacity of EVs in mediating the delivery of CFTR to human CF cells. Preliminary experiments using EVs from Calu-3 cells incubated with CF15 cells, which carry the CFTR ΔF508 mutation, showed a correction of the CFTR chloride channel defect in CF15 cells. However, the level of this newly acquired CFTR activity was 10–20-fold lower than that of Calu-3 cells. We therefore explored different conditions and parameters to improve the delivery of CFTR to target cells.

Important factors in CFTR delivery included the CFTR cargo in EVs, the homology between EV-donor cells and EV-recipient cells, and the EV transducing doses. To augment the CFTR cargo in EVs, and to visualize in situ the transfer of the CFTR glycoprotein, we used EVs isolated from A549 cells transduced by HAdV5 vectors expressing GFP-tagged CFTR at high levels.^28,29 Both types of EVs, MVs and EXos, were capable of incorporating and transferring GFP-CFTR to target cells in both homologous (EVs from A549 to A549 cells) and heterologous systems (EVs from A549 to CF15 cells). Not surprisingly, we found that the cellular uptake of EVs was more efficient in EV-to-cell homologous transfer. This parameter should be taken into consideration for future clinical applications of EV-based human biotherapy, and notably for the design of autologous EV-donor cells.

Upon uptake of MVs^GFP-CFTR and EXos^GFP-CFTR, the CFTR-deficient CF15 cells acquired the CFTR-associated Cl⁻ channel activity as rapidly as 24 hr posttransduction. The newly acquired CFTR activity was EV-dose dependent in the range of 1 × 10⁴ to 4 × 10⁴ MVs or EXos per cell. However, no significant Cl⁻ channel activity was detectable at doses inferior to 10⁴ EVs/cell. Of note, a similar threshold effect had been observed in human airway primary CF cells transduced by adenoviral vectors.²⁹ Interestingly, the transduction efficiency of CF cells over time by MVs^GFP-CFTR and EXos^GFP-CFTR showed a curve with two maxima separated by a minimum of GFP-CFTR-positive cells at 12 hr posttransduction (Fig. 8a). This U-shaped pattern suggested that the fluorescent signal detected at early and late times posttransduction resulted from the contribution of two types of molecules, the GFP-CFTR glycoprotein and its encoding mRNA: (1) mature GFP-CFTR glycoproteins transferred by EVs to target cells would be responsible for the early GFP signal; (2) de novo synthesized GFP-CFTR, translated from the exogenous mRNA^GFP-CFTR by the target cell translational machinery, would be responsible for the retarded GFP signal. These results implied that EV-packaged mRNA molecules in general, and mRNA^GFP-CFTR in particular, played a major role in the gain of a function by target cells, as exemplified by the CFTR Cl⁻ channel activity gained by CF cells. This corroborated our previous observations of the functional role of exogenous mRNA in conveying a desired biological activity in target cells.²⁷

However, EVs isolated from HAdV5-GFP-CFTR-transduced A549 cells were found to contain remnants of adenoviral vector DNA, which were transferred to target cells. Although no infectious viral vector could be rescued when MVs^GFP-CFTR and EXos^GFP-CFTR were incubated with trans-complementing HEK-293 cells, the possibility existed that the GFP-CFTR gene carried over by EVs was transcribed into mRNA^GFP-CFTR, and this neosynthesized mRNA translated into the GFP-CFTR protein at late times after transduction. Such a mechanism was supported by a recent study on the intracellular fate of EV-transferred biomolecules, and on the active role played by a plasmid DNA compared with RNA molecules, considering their relative instability and short lifetime within the recipient cells.⁵²

In the present study, we first demonstrated that functional correction of the CFTR Cl⁻ channel defect in CF cells could be achieved by EVs carrying mRNA^CFTR isolated from Calu-3 cells. In the case of CF15 cells transduced by EVs isolated from viral vector-transduced cells, such as HAdV5-GFP-CFTR/A549 cells, the mRNA^GFP-CFTR responsible for the newly acquired CFTR Cl⁻ channel activity might result from the translation of two separate pools of molecules, mRNA^GFP-CFTR presynthesized in EV-donor cells (A549) and transferred by EVs to target CF cells, and mRNA^GFP-CFTR neosynthesized in target CF15 cells. Whatever the respective contribution of one or the other mechanism, the neosynthesis of GFP-CFTR glycoprotein molecules would explain the persistence of the CFTR activity in CF cells for several days after the EV-mediated transfer.

Flow cytometry analysis of CF15 cells transduced with MVs^GFP-CFTR and EXos^GFP-CFTR at the same dose, and RT-PCR determination of the mRNA^GFP-CFTR cellular content, suggested a higher transfer efficiency by MVs, compared with EXos. Likewise, mRNA^GFP-CFTR had a longer half-life in MVs^GFP-CFTR-transduced cells, compared with EXos^GFP-CFTR-transduced cells. However, the CFTR Cl⁻ channel activity newly acquired by EV-transduced CF15 cells and its duration over time were found to be equivalent in EXos^GFP-CFTR- and MVs^GFP-CFTR-transduced cells. These apparently contradictory results implied that only a fraction of the mRNA^GFP-CFTR transferred by MVs^GFP-CFTR was metabolically active, and that the majority remained sequestered in a dead-end cell compartment. This hypothesis was supported by fluorescence confocal microscopy, which suggested that MVs^GFP-CFTR and EXos^GFP-CFTR followed different trafficking and/or metabolic pathways after their cellular uptake by CF cells.

Collectively, our study confirmed the potential application of EVs for the delivery of CFTR to CF cells and the correction of the chloride channel deficiency. The present limitation to the use of CFTR-loaded EVs was the threshold observed in the acquisition of the CFTR-associated channel function, which required relatively high transducing doses of EVs, compared with conventional viral vectors. This could be solved by in vitro transcription of mRNA^CFTR molecules from a CFTR-encoding plasmid, and their encapsulation at high concentration into EVs, as already developed for therapeutic RNAs and liposomes.⁵³ Such a technique could find its application for EVs isolated from individual patients, in the context of the personalized medicine of CF pathology using autologous EVs.

Footnotes

Acknowledgments

The work in Lyon was financed by the Cystic Fibrosis French Association (VLM Contract RF20130500796 and RF20140501171). C.V. was the recipient of doctoral fellowships from the Région Rhône-Alpes (ADR Cluster-ARC 2011–2014), CMIRA Explora'doc-2012, and VLM (Vaincre la Mucoviscidose). S.S.H. is an INSERM senior scientist (Chargée de Recherche) and the recipient of a Contrat d'Interface Hospices Civils de Lyon-INSERM (2009–2013). The work in Geneva was supported by Grant #310030_134907/1 from the Swiss Science Foundation. We are grateful to Sébastien Dussurgey and Thibault Andrieu (AniRA-Cytométrie, SFR BioSciences), Bariza Blanquier (AniRA-Analyse génétique SFR BioSciences), Catherine Ott and Laurence Générénaz (Laboratoire Commun de Recherche BioMérieux-HCL), and Nathalie Calin (ENS-Lyon) for their valuable advice and technical help. We also thank Sylvie Farget for her constant secretarial aid.

Author Disclosure

No competing financial interests exist.

References

Yáñez-Mó

, Siljander

, Andreu

, et al. Biological properties of extracellular vesicles and their physiological functions. J Extracell Vesicles, 2015; 4:27066.

Raposo

, Stoorvogel

. Extracellular vesicles: Exosomes, microvesicles, and friends. J Cell Biol, 2013; 200:373–383.

Thery

, Ostrowski

, Segura

. Membrane vesicles as conveyors of immune responses. Nat Rev Immunol, 2009; 9:581–593.

Cocucci

, Racchetti

, Meldolesi

. Shedding microvesicles: Artefacts no more. Trends Cell Biol, 2009; 19:43–51.

Hulsmans

, Holvoet

. MicroRNA-containing microvesicles regulating inflammation in association with atherosclerotic disease. Cardiovasc Res, 2013; 100:7–18.

El-Andaloussi

, Lee

, Lakhal-Littleton

, et al. Exosome-mediated delivery of siRNA in vitro and in vivo . Nat Protoc, 2012; 7:2112–2126.

Mathivanan

, Ji

, Simpson

. Exosomes: Extracellular organelles important in intercellular communication. J Proteomics, 2010; 73:1907–1920.

Bruno

, Camussi

. Role of mesenchymal stem cell-derived microvesicles in tissue repair. Pediatr Nephrol, 2013; 28:2249–2254.

Gatti

, Bruno

, Deregibus

, et al. Microvesicles derived from human adult mesenchymal stem cells protect against ischaemia-reperfusion-induced acute and chronic kidney injury. Nephrol Dial Transplant, 2011; 26:1474–1483.

10.

Bobrie

, Colombo

, Raposo

, et al. Exosome secretion: Molecular mechanisms and roles in immune responses. Traffic, 2011; 12:1659–1668.

11.

Del Conde

, Shrimpton

, Thiagarajan

, et al. Tissue-factor-bearing microvesicles arise from lipid rafts and fuse with activated platelets to initiate coagulation. Blood, 2005; 106:1604–1611.

12.

Ratajczak

, Miekus

, Kucia

, et al. Embryonic stem cell-derived microvesicles reprogram hematopoietic progenitors: Evidence for horizontal transfer of mRNA and protein delivery. Leukemia, 2006; 20:847–856.

13.

Camussi

, Deregibus

, Bruno

, et al. Exosome/microvesicle-mediated epigenetic reprogramming of cells. Am J Cancer Res, 2011; 1:98–110.

14.

Rak

, Guha

. Extracellular vesicles—vehicles that spread cancer genes. BioEssays, 2012; 34:489–497.

15.

Bellingham

, Guo

, Coleman

, et al. Exosomes: Vehicles for the transfer of toxic proteins associated with neurodegenerative diseases?. Front Physiol, 2012; 3:124.

16.

Emmanouilidou

, Melachroinou

, Roumeliotis

, et al. Cell-produced alpha-synuclein is secreted in a calcium-dependent manner by exosomes and impacts neuronal survival. J Neurosci, 2010; 30:6838–6851.

17.

Vella

, Hill

. Generation of cell lines propagating infectious prions and the isolation and characterization of cell-derived exosomes. Methods Mol Biol, 2008; 459:69–82.

18.

Vella

, Sharples

, Lawson

, et al. Packaging of prions into exosomes is associated with a novel pathway of PrP processing. J Pathol, 2007; 211:582–590.

19.

Mack

, Kleinschmidt

, Brühl

, et al. Transfert of the chemokin receptor CCR5 between cells by membrane derived microparticles: A mechanism for cellular human immunodeficiency virus infection. Nat Med, 2000; 6:769–775.

20.

Arenaccio

, Anticoli

, Manfredi

, et al. Latent HIV-1 is activated by exosomes from cells infected with either replication-competent or defective HIV-1. Retrovirology, 2015; 12:e87.

21.

György

, Hung

, Breakefield

, et al. Therapeutic applications of extracellular vesicles: Clinical promise and open questions. Annu Rev Pharmacol Toxicol, 2015; 55:439–464.

22.

Bradbury

. Intracellular CFTR: Localization and function. Physiol Rev, 1999; 79:S175–S191.

23.

Sheppard

, Welsh

. Structure and function of the CFTR chloride channel. Physiol Rev, 1999; 79:S23–S45.

24.

Chow

, Plumb

, Wen

, et al. Targeting transgene expression to airway epithelia and submucosal glands, prominent sites of human CFTR expression. Mol Ther, 2000; 2:359–367.

25.

Guo

, Su

, McNutt

, et al. Expression and distribution of cystic fibrosis transmembrane conductance regulator in neurons of the human brain. J Histochem Cytochem, 2009; 57:1113–1120.

26.

Mohamed

, Ferguson

, Seibert

, et al. Functional expression and apical localization of the cystic fibrosis transmembrane conductance regulator in MDCK I cells. Biochem J, 1997; 322:259–265.

27.

Gonzalez

, Vituret

, Di Pietro

, et al. Microparticle-mediated transfer of the viral receptors CAR and CD46, and the CFTR channel in a CHO cell model confers new functions to target cells. PLoS One, 2012; 7:e52326.

28.

Granio

, Norez

, Ashbourne Excoffon

, et al. Cellular localization and activity of Ad-delivered GFP-CFTR in airway epithelial and tracheal cells. Am J Respir Cell Mol Biol, 2007; 37:631–639.

29.

Granio

, Ashbourne Excoffon

KJD

, Henning

, et al. Adenovirus 5-fiber 35 chimeric vector mediates efficient apical correction of the cystic fibrosis transmembrane conductance regulator defect in cystic fibrosis primary airway epithelia. Hum Gene Ther, 2010; 21:1–19.

30.

Gruenert

, Willems

, Cassiman

, et al. Established cell lines used in cystic fibrosis research. J Cyst Fibros, 2004; 3:191–196.

31.

Shen

, Finkbeiner

, Wine

, et al. Calu-3: A human airway epithelial cell line that shows cAMP-dependent Cl- secretion. Am J Physiol, 1994; 266:L493–L501.

32.

Jefferson

, Valentich

, Marini

, et al. Expression of normal and cystic fibrosis phenotypes by continuous airway epithelial cell lines. Am J Physiol, 1990; 259:L496–L505.

33.

Pizurki

, Morris

, Chanson

, et al. Cystic fibrosis transmembrane conductance regulator does not affect neutrophil migration across cystic fibrosis airway epithelial monolayers. Am J Pathol, 2000; 156:1407–1416.

34.

Corjon

, Gonzalez

, Henning

, et al. Cell entry and trafficking of human adenovirus bound to blood factor X is determined by the fiber serotype and not hexon:heparan sulfate interaction. PLoS One, 2011; 6:e18205.

35.

Granio

, Porcherot

, Corjon

, et al. Improved adenovirus type 5 vector-mediated transduction of resistant cells by piggybacking on Coxsackie B-adenovirus receptor-pseudotyped baculovirus. J Virol, 2009; 83:6048–6066.

36.

DaFonseca

, Blommaert

, Coric

, et al. The 3-O-(3′,3′-dimethylsuccinyl) derivative of betulinic acid (DSB) inhibits the assembly of virus-like particles in HIV-1 Gag precursor-expressing cells. Antiviral Ther, 2007; 12:1185–1203.

37.

Dafonseca

, Coric

, Gay

, et al. The inhibition of assembly of HIV-1 virus-like particles by 3-O-(3′,3′-dimethylsuccinyl) betulinic acid (DSB) is counteracted by Vif and requires its Zinc-binding domain. Virol J, 2008; 5:162.

38.

Franqueville

, Henning

, Magnusson

, et al. Protein crystals in adenovirus type 5-infected cells: Requirements for intranuclear crystallogenesis, structural and functional analysis. PLoS One, 2008; 3:e2894.

39.

Hong

, Magnusson

, Henning

, et al. Adenovirus stripping: A versatile method to generate adenovirus vectors with new cell target specificity. Mol Ther, 2003; 7:692–699.

40.

Ftaich

, Ciancia

, Viarouge

, et al. Turnover rate of NS3 proteins modulates bluetongue virus replication kinetics in a host-specific manner. J Virol, 2015; 89:10467–10481.

41.

Chanson

, Scerri

, Suter

. Defective regulation of gap junctional coupling in cystic fibrosis pancreatic duct cells. J Clin Invest, 1999; 103:1677–1684.

42.

Long

, Walsh

. Iodide efflux measurements with an iodide-selective electrode: A non-radioactive procedure for monitoring cellular chloride transport. Meth Cell Sci, 1997; 19:207–212.

43.

Aleksandrov

, Kota

, Aleksandrov

, et al. Regulatory insertion removal restores maturation, stability and function of DeltaF508 CFTR. J Mol Biol, 2010; 401:194–210.

44.

Sumarheni

, Hong

, Josserand

, et al. Human full-length coagulation factor X and a GLA domain-derived 40-mer polypeptide bind to different regions of the adenovirus serotype 5 hexon capsomer. Hum Gene Ther, 2014; 25:339–349.

45.

Zhu

, Chidekel

, Shaffer

. Cultured human airway apithelial cells (Calu-3): A model of human respiratory function, structure, and inflammatory responses. Crit Care Res Pract 2010; 2010:ID 394578.

46.

Kesimer

, Scull

, Brighton

, et al. Characterization of exosome-like vesicles released from human tracheobronchial ciliated epithelium: A possible role in innate defense. FASEB J, 2009; 23:1858–1868.

47.

Webber

, Clayton

. How pure are your vesicles?. J Extracell Vesic, 2013; 2:19861.

48.

Conner

, Schmid

. Regulated portals of entry into the cell. Nature, 2003; 422:37–44.

49.

Dutta

, Donaldson

. Search for inhibitors of endocytosis: Intended specificity and unintended consequences. Cell Logist, 2012; 2:203–208.

50.

West

, Bretscher

, Watts

. Distinct endocytotic pathways in epidermal growth factor-stimulated human carcinoma A431 cells. J Cell Biol, 1989; 109:2731–2739.

51.

Mulcahy

, Pink

, Carter

. Routes and mechanisms of extracellular vesicle uptake. J Extracell Vesicl, 2014; 3:24641.

52.

Kanada

, Bachmann

, Hardy

, et al. Differential fates of biomolecules delivered to target cells via extracellular vesicles. Proc Natl Acad Sci U S A, 2015; 112:E1433–E1442.

53.

Wang

, Su

H-H

, Yang

, et al. Systemic delivery of modified mRNA encoding Herpes simplex virus 1 thymidine kinase for targeted cancer gene therapy. Mol Ther, 2013; 21:358–367.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.20 MB

0.06 MB

0.12 MB

0.00 MB

Transfer of the Cystic Fibrosis Transmembrane Conductance Regulator to Human Cystic Fibrosis Cells Mediated by Extracellular Vesicles

Abstract

Introduction

Materials and Methods

Cells

Adenoviral vectors

Isolation of microvesicles and exosomes from extracellular vesicles

Protein assay

Titration of microvesicles and exosomes by nanoparticle tracking analysis

Gel electrophoresis and Western blot analysis

Antibodies and reagents

EV-mediated transfer of GFP and GFP-CFTR into target cells

DNA and RNA isolation, and PCR analyses

Iodide efflux assay

Flow cytometry

Cell imaging

Fluorescence microscopy

Electron microscopy

Statistics

Results

Isolation, characterization, and nomenclature of EVs recovered from human cells

Production and recovery of extracellular EVs

EV-mediated delivery of chloride channel activity to human CF cells

Characterization of EV-donor cells

Characterization of EVsGFP and EVsGFP-CFTR from A549 cells

Cellular uptake of MVsGFP-CFTR and EXosGFP-CFTR by target cells

EV dose dependence of the GFP-CFTR transfer

Cytotoxicity of MVsGFP-CFTR and EXosGFP-CFTR

Temperature dependence

Homologous versus heterologous cellular uptake

Homologous cellular uptake (Fig. 6a)

Heterologous cellular uptake (Fig. 6b)

Mechanisms of cellular uptake of EVs

Cell entry inhibitors

Imaging of EV–cell interaction by electron microscopy

Confocal fluorescence microscopy of EV–cell interaction

Kinetics of EV-mediated delivery of GFP-CFTR to CF target cells

Restoration of chloride (Cl−) channel function in CF cells transduced by CFTR-loaded EVs

Discussion

Footnotes

Acknowledgments

Author Disclosure

References

Supplementary Material

Characterization of EVs^GFP and EVs^GFP-CFTR from A549 cells

Cellular uptake of MVs^GFP-CFTR and EXos^GFP-CFTR by target cells

Cytotoxicity of MVs^GFP-CFTR and EXos^GFP-CFTR

Restoration of chloride (Cl⁻) channel function in CF cells transduced by CFTR-loaded EVs