Background
Cells that are present in close proximity to each other, including immune cells recruited to epithelial tissues, are in constant communication with each other, both in disease or homeostatic condition [
1‐
4]. This communication can be induced by direct cell-to-cell contact, secreted free proteins, or shuttling of bioactive molecules via extracellular vesicles (EVs). EV-based cellular cross talk is an evolutionarily conserved form of communication, and EVs can induce signaling through their cargoes of protein, DNA, and RNA [
5‐
9]. With their varied cargo, EVs have the potential to activate multiple signaling events in recipient cells. However, how these different signaling events integrate together to drive a cellular phenotype is not well studied.
EVs have been reported to induce reprogramming of cells via epithelial-mesenchymal transition (EMT) [
10‐
14]. During EMT, cells undergo programmed changes in cellular mRNA and protein content [
15‐
17]. This process is highly coordinated and reversible and is associated with tissue damage and with numerous pathological conditions [
18‐
21]. Extracellular cues recruit immune cells into the micro-niche of the lungs during chronic inflammatory diseases, including asthma, allergic reactions, and chronic obstructive pulmonary disorder [
22‐
27]. During this process, lung epithelial cells are constantly exposed to inflammatory insults from recruited and resident immune cells, including mast cells [
22,
28]. This damage induces structural re-arrangements of the epithelial lining along with tissue remodeling, increased metalloproteinase activity, the creation of fibrotic lesions, and altered cytokine levels [
29‐
31]. Mast cells are known to interact physically with many cell types and thus to influence the inflammatory phenotype [
29‐
31]. Some of the described inflammatory epithelial phenotypes are indeed features of the EMT process [
32]. In this study we determined the potential of mast cell-derived EVs to regulate EMT, and we identified the potential signaling events in recipient epithelial cells.
In a previous study we showed that the mast cell-derived EV-associated membrane protein c-Kit is transferred to epithelial cells, resulting in downstream phosphorylation of AKT and GSK3-β and thus enhancing proliferation [
33]. Our earlier studies also identified transforming growth factor (TGFβ-1) on the surface of mast cell-derived EVs that might play a role in EMT [
34]. Here, we hypothesized that mast cell-derived EVs have the capacity to induce an EMT-like phenotype and that they activate multiple signaling events in lung epithelial cells. To this end, we used HMC1 cell derived EVs, as these cells are constitutively active and does not need growth-factor because of constitutive activity of the receptor tyrosine kinase Kit [
35]. Our transcript, protein, and phosphoprotein analysis of epithelial cells showed activation of EMT pathways. This study highlights the potential signaling cascades activated by mast cell-derived EVs in regulating the EMT phenotype in epithelial cells.
Methods
Cell culture
The alveolar epithelial cell line A549 was obtained from ATCC, USA, and the cells were cultured in DMEM/F12 (Sigma Aldrich, St. Louis, MO, USA) supplemented with 10% fetal bovine serum (FBS; Sigma Aldrich). The culture media was changed to EV-depleted FBS-containing medium 24 h prior to the experiments. For all experiments the cells were seeded at a density of 15,000 cells/cm2.
Human mast cells, HMC-1 (J. Butterfield, Mayo Clinic, Rochester, MN, USA), were cultured in Iscove’s modified Dulbecco’s medium (HyClone Laboratories, Logan, UT, USA) supplemented with 10% EV-depleted FBS, 1.2 mM α-thioglycerol (Sigma Aldrich), and 2 mM L-glutamine (HyClone Laboratories). These cultures were supplemented with 100 units/ml penicillin and 100 μg/ml streptomycin (HyClone Laboratories) and cultured at 37 °C in a 5% CO
2 humidified conditions. The FBS used for HMC-1 cultures was ultracentrifuged for 18 h at 120,000×
g (Type 45 Ti rotor, Beckman Coulter) to remove the serum EVs, as reported earlier [
36].
Isolation of EVs
Conditioned medium from HMC-1 cells was obtained after 3–4 days of culture, and cells were removed by centrifugation at 300×g for 10 min. The cell-free supernatant was further centrifuged at 16,500×g for 20 min to remove microvesicles and apoptotic bodies. Finally, this supernatant was centrifuged at 120,000×g for 3 h (Type 45 Ti rotor, Beckman Coulter), and the pelleted EVs were washed once with PBS. The final EV pellet was suspended in PBS and stored at − 80 °C for further experiments. The protein concentration of the EVs was measured using the BCA protein assay kit (Pierce, Thermo Fisher Scientific, Waltham, MA, USA).
EV labeling and cellular uptake
EVs obtained from HMC-1 cells were labeled with the PKH67 Green Fluorescent Cell Linker Kit (Sigma Aldrich) as per the manufacturer’s protocol. The labeled EVs were loaded onto the bottom of an iodixanol density gradient (0, 20, 30, and 50% iodixanol) and centrifuged at 28,000 rpm for 2 h in a swinging bucket rotor (SW40Ti, Beckman Coulter). The EVs floating over the interphase (20–30%) were collected and washed in PBS followed by centrifugation at 120,000×g for 3 h (Type 45 Ti rotor, Beckman Coulter). A549 cells were grown on coverslips at 15,000 cells/cm2 for 24 h. The labeled EVs were incubated with the A549 cells grown on the coverslip for 2 h or for 16 h. The cell membranes and nuclei were stained with the Image-IT LIVE kit (Invitrogen, Thermo Fisher Scientific) using Alexa Fluor-594 wheat germ agglutinin and Hoechst 33342, respectively, according to the manufacturer’s protocol. The cells were fixed in a paraformaldehyde (3.5%) solution for 10 min and washed before the cover slip containing the cells was mounted on a slide and imaged under a structural illumination microscope (Zeiss Elyra 3D SIM, Germany).
Gelatin zymography
A549 cells were exposed to mast cell-derived EVs, and conditioned medium was collected at 24 h and at 48 h. The conditioned media was separated on gelatin-contacting zymogram gels (BioRad Laboratories, Hercules, CA, USA) with 5× non-reducing loading buffer (Sigma Aldrich). Renaturation of matrix metalloproteinases in the gel was performed at room temperature in 2.5% Triton X-100 (Sigma Aldrich) for 1 h followed by overnight incubation at 37 °C in development solution (50 mM Tris (pH 7.4), 5 mM CaCl2, 200 mM NaCl). Gels were then stained with Coomassie brilliant blue and destained (30% methanol and 10% acetic acid) until the white bands that reflect gelatinase activity appeared. Finally, 2% acetic acid was added to stop the destaining process. The degree of gelatinase activity was measured by quantifying the band intensity using ImageJ software.
Reversed cell migration assay
The Boyden chamber migration assay (Neuroprobe, Gaithersburg, MD, USA) was used to determine the migratory potential of A549 cells upon EV stimulation as described earlier [
34].
Immunofluorescence microscopy
A549 cells were incubated with EVs for 24 h before being fixed with 3.7% paraformaldehyde at room temperature for 10 min, permeabilized for 5 min with 0.2% Triton X-100, washed, and blocked for 1 h in 3% BSA. Finally, incubation with primary antibody for N-cadherin (NCAD) was performed for 1 h at room temperature, and the sample was washed again before it was stained with AF488-labeled secondary antibody. After three washes with PBS, the cells were further stained with DAPI (Sigma Aldrich) and cover slips were mounted using ProLong Gold anti-fade mounting reagent (Invitrogen, Carlsbad, CA, USA) and observed under a fluorescence light microscope (Axio Observer, Zeiss, Oberkochen, Germany).
Secretion of TGFβ-1 measured by ELISA
The amount of TGFβ-1 secreted by A549 cells was measured by TGFβ-1 ELISA Ready-SET-Go kit (eBioscience Affymetrix) according to the manufacturer’s instructions.
Quantitative real time PCR
RNA from epithelial A549 cells was isolated using the column-based miRCURY™ RNA isolation kit for cell and plant (Exiqon, Vedbaek, Denmark) and treated with TURBO DNase (Ambion, Life Technologies) to remove contaminating DNA. The concentration and purity of the RNA was quantified using a NanoDrop system (Thermo Scientific). The iScript cDNA Synthesis Kit (BioRad) was used to generate cDNA from 200 ng RNA. SsoAdvanced Universal SYBR Green Supermix was used to perform quantitative real time PCR in a BioRad CFX96 system for data collection and analysis. Briefly, cDNA was denatured at 95 °C for 30 s followed by 40 cycles of 95 °C for 15 s and 60 °C for 30 s. KiCqStart® primers (Sigma) were obtained for the following genes: TGFB1, TWIST1, SMAD2, MMP2, WNT5A, FOXC2, BMP7, and VIM. The endogenously expressed EF1 gene was used for transcript normalization. Relative fold change in gene expression was calculated by the 2T−ΔΔC method.
Western blotting
A549 cells were seeded at 0.2 × 106 cells/well in 6-well plates and incubated for 24 h. Cells were then treated with 30 μg/ml EVs or 10 ng/ml TGFβ-1 at different time points. The A549 cell pellet was washed in PBS, and the cells were lysed in 1× RIPA buffer (Cell Signaling Technology ST, Danvers, MA, USA) with 1× Halt protease and phosphatase inhibitor cocktail (Halt, Thermo Fisher Scientific). Protein lysates were subjected to SDS-PAGE and transferred onto nitrocellulose membranes. Nonspecific binding sites were blocked with Tris-buffered saline with 0.05% Tween-20 and 5% non-fat milk or 5% bovine serum albumin (BSA) for 1 h at room temperature. Primary antibodies in their respective blocking buffer were incubated at 4 °C overnight. Horseradish peroxidase-conjugated secondary antibodies (1:10,000 dilution, NA931V, NA9340, NA9310V, GE Healthcare) were diluted in blocking buffer and incubated with the membrane for 1 h at room temperature. Chemiluminescence signal for proteins was detected with SuperSignal West Femto Maximum Sensitivity Substrate (Thermo Fisher Scientific) according to the manufacturer’s protocol. The following antibodies were used: NCAD (3B9; 1:1000 dilution, #33–3900 Invitrogen, Thermo Fisher Scientific), E-cadherin (ECAD) (1:1000 dilution, #610182, BD Biosciences), Slug-Snail (1:1000 dilution, #ab180714, Abcam), and β-actin (1:3000 dilution, #sc47778, Santa Cruz Biotechnology).
Phospho-proteomics microarray
Sample preparation
Sample processing and analysis was performed as per the guidelines provided by Sciomics (Heidelberg, Germany). Epithelial A549 cells were seeded at a density to achieve ~ 70% confluency in 24 h. Approximately 2 million cells were either treated with EVs (30 μg/ml) or left untreated. Each set was a biological duplicate (i.e. two untreated and two EV treated). After 60 min, the cells were rinsed with PBS and pelleted. Pelleted samples were snap-chilled in dry ice and stored at − 80 °C until use and sent to Sciomics for further analysis. Extraction of proteins was performed with proprietary scioExtract buffer (Sciomics), and the protein concentration was measured using the BCA assay. Samples were labeled with scioDye for 1 h adjusting the protein concentration, and the reaction was then stopped by the addition of hydroxylamine. Excess dye was removed 30 min later and the buffer was exchanged to PBS. All labeled protein samples were used immediately.
Antibody microarrays
The samples were analyzed using a scioDiscover antibody microarray (Sciomics) targeting 1033 different proteins with 1516 antibodies. Each antibody was represented on the array in four replicates. The arrays were blocked with scioBlock (Sciomics) on a Hybstation 4800 (Tecan, Austria). The antibodies were first added to the microarray and then the samples and scioPhosphomix were incubated. scioPhosphomix provides information on protein-specific phosphorylation levels of serine, threonine, and tyrosine residues. After incubation for 3 h, the slides were thoroughly washed with 1× PBS (with Tween-20 and Triton X-100), rinsed with 0.1× PBS, rinsed with water, and dried under nitrogen.
Data acquisition and analysis
Slide scanning was performed with a Powerscanner (Tecan, Austria) with identical instrument laser power and adjusted photomultiplier tube settings. Spot segmentation was performed with GenePix Pro 6.0 (Molecular Devices, Union City, CA, USA). Acquired raw data were analyzed using the linear models for microarray data (LIMMA) package of R-Bioconductor after uploading the median signal intensities. For normalization, a Cyclic Loess normalization was applied. For analysis of the samples, a one-factorial linear model was applied with LIMMA resulting in a two-sided t-test or F-test based on moderated statistics. All presented p-values were adjusted for multiple testing by controlling for the false discovery rate according to Benjamini and Hochberg. Proteins were defined as differentially expressed with an IlogFCI > 0.5 and an adjusted p-value < 0.05. Differences in protein abundance or phosphorylation level between different samples or sample groups are presented as log-fold changes (logFC) calculated for base 2. In a study comparing samples versus controls, a logFC = 1 means that the sample group had on average a 21 = 2-fold higher signal than the control group. logFC = − 1 stands for 2− 1 = 1/2 of the signal in the sample compared to the control group.
Discussion
Epithelial cells in the lungs act as barrier and maintain homeostasis as they respond to extracellular factors from neighboring immune cells, including mast cells, and to foreign antigens [
44‐
46]. In the current study, we observed that mast cell-derived EVs were able to induce migration of the airway epithelial cell line A549 in-vitro, and this was accompanied by multiple indicators of EMT. We also demonstrated the differential regulation of EMT-associated transcripts (
TWIST1,
TGFB1,
MMP9, and
MMP2) and protein markers (MMP, Slug-Snail, NCAD, and ECAD) in response to the mast-cell EVs. We also found that EVs mediated the early signaling response by inducing the phosphorylation of multiple proteins in epithelial A549 cells that have been suggested to be involved in regulating EMT. For example, TGFβ-1 present on EV surfaces acts as an early signal to induce the phosphorylation of SMAD2 in A549 cells. Taken together, the results of this study show that mast cell-derived EVs induce an EMT response in epithelial cells, and this phenotypic cell response appears to be due to the early phosphorylation of numerous proteins known to be involved in EMT.
Various immune cells, including mast cells, T-cells, B-cells, and dendritic cells, are known to produce EVs under various inflammatory conditions [
7,
47‐
49]. Mast cells have been referred to as the “rheostat” of the local immune system, and they can release factors for host defense as well as for processes such as remodeling, wound-healing, angiogenesis, and cancer progression [
50‐
52]. The number of mast cells is high in the vicinity of epithelial cells in the peripheral airways, and this proximity provides an opportunity for cell-to-cell cross talk between mast cells and epithelial cells [
44]. In our in vitro studies, we found that mast cell-derived EVs were taken up by epithelial cells at different time points (Fig.
1a). Furthermore, the uptake of these EVs by epithelial cells induced a morphological change with elongated protrusions (Fig.
1b). The A549 cells showed detachment from neighboring cells after incubation with mast cell-derived EVs, and this was associated with upregulation of EMT-promoting markers (Fig.
2 and Fig.
3 a). An earlier study by Kasai et al. showed similar morphological changes in A549 cells and molecular changes in line with EMT induced by free TGFβ-1 [
23]. However, in the current study, the induction of EMT by EVs was observed as early as 24 h after treatment, compared to 48 h in the previous study in which free TGFβ-1 was given. In our previous study, we showed that TGFβ-1 is present on the surface of EVs, resulting in potent activation of TGF signaling in mesenchymal stem cells [
34]. Similarly, in the present study we observed the activation of SMAD2 in both alveolar A549 and bronchial BEAS-2B epithelial cells (Fig.
4 and Supplementary Figure
1).
EV membranes harbor multiple proteins, including growth factors that have bioactive functions. Interactions between TGFβ-1 and the EV surface are partially due to the glycoprotein-like heparin and heparan sulfate glycosaminoglycan present on the EV surface [
53]. These glycoproteins are known to interact with multiple growth factors like TGFβ-1, FGF and VEGF [
54‐
56]. It is likely that EV-associated TGFβ-1 activates A549 cells to initiate early EMT-inducing pathways; however, other molecular pathways might also be involved in activating EMT. Mast cell-derived EVs contain other secreted proteins like tryptase α/β-1, galectin-1, proteoglycan-2, and macrophage migration inhibitory factor that could potentially have additive effects [
34]. Further, DNA present on the mast cell EV surfaces might also be involved in the biological activity of the released EVs [
57]. A recent study by Kobayashi et al. showed that extracellular CpG DNA in combination with TGFβ-1 enhanced the EMT cascade in A549 cells at lower concentrations compared to TGFβ-1 alone. This suggests that EVs and multiple associated cargos, including TGFβ-1 and surface-exposed DNA, might be driving the epithelial phenotype because they can interact with cells simultaneously. However, the role of EV surface-associated DNA in the regulation of EMT as suggested in this study remains to be confirmed in further studies.
To evaluate whether the morphological changes induced in epithelial cells via EVs were really due to EMT, we analyzed the transcripts of a set of genes reported to be involved in the induction of EMT. We observed significant increases in the transcript levels of TWIST1, TGFB1, MMP9, and BMP7, all of which are known to be involved in EMT. In parallel, we observed an increase in proteins involved in EMT such as TGFβ-1, Slug-Snail, MMP-2, and MMP-9. Interestingly, the increase in these mRNA levels was clear at the 24 h time point, but diminished at 48 h, possibly because the EMT response induced by the EVs is only transient. However, we observed a continuous increase in the release of TGFβ-1 in A549 supernatant up until 48 h post-EV exposure, suggesting that some biological functions might be longer lasting than others. This suggests that TGFβ-1 protein production might be important for the initiation but not for the maintenance of EMT (with respect to TGFB1 transcripts).
The EMT response is driven by upregulation of NCAD and the downregulation of ECAD. After the stimulation of epithelial cells with EVs, we observed a reduction in total protein for the epithelial marker ECAD, whereas levels of NCAD remained unchanged. It is possible that the cellular distribution of NCAD might have been altered, and thus we performed immunofluorescence microcopy to determine the cellular location of this molecule in epithelial cells after co-incubation with EVs. Indeed, we observed higher NCAD expression at the cellular junctions, indicating that EVs altered the cellular distribution of this molecule in the recipient cells. Previous studies have suggested that the presence of surface NCAD is associated with a higher migratory phenotype compared to higher levels of cytoplasmic NCAD [
58,
59], and this is supported by our present findings. Thus, based on this observation, we suggest that NCAD present on the cell surface can be sufficient to induce an EMT-associated migratory response. This migratory capacity is coupled to enhanced secretion of bioactive matrix metalloproteinases because the reduction of matrix component might indirectly assist cellular movement [
33].
With our current findings on EMT and previous observations that mast cell-derived EVs can induce A549 cell proliferation, we hypothesize that multiple cargos present on mast cell-derived EVs can activate multiple signaling cascades in recipient cells [
33]. Indeed, mast cell-derived EVs induced the phosphorylation of at least 47 different cellular proteins in epithelial A549 cells. Importantly, a majority of these are reported to be involved in EMT, for example, upregulation of TGM2, MMP-9, and Annexine-A1 and downregulation of VACM1, CADH1, and Chrombox3 proteins.
Beyond phosphorylation of EMT-associated proteins, we also observed phosphorylation of proteins involved in associated pathways, such as the PI3K-AKT, HIF-1, NFĸB, and Jak-STAT signaling pathways. In our previous study, we described the transfer of constitutively active c-KIT membrane receptor from mast cell-derived EVs to epithelial cells, and this was associated with the activation of PI3K-AKT signaling. Indeed, PI3K-AKT signaling can be activated by multiple upstream factors, including multiple growth factor such as TGFβ-1, in addition to the c-KIT transfer [
33]. PI3K-AKT signaling is central to many of the signaling pathways that induce EMT-like processes, so these results were not surprising. In this study we wanted to highlight the possibility of multiple signaling pathways activated by mast cell-derived EVs instead of focusing on a single pathway. A question that has remained unanswered is how the multiple and converging cellular responses that EVs can induce lead to a specific phenotype change such as EMT in recipient cells. Interestingly, we observed changes in the phosphorylation of proteins that regulate cell-to-cell contact, e.g. focal adhesion, cell adhesion, and tight junction proteins, and such changes might lead to rapid alterations of cell surface dynamics.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.