In Vivo therapeutic potential of mesenchymal stem cell-derived extracellular vesicles with optical imaging reporter in tumor mice model

Mesenchymal stem cells (MSCs) can be used as a therapeutic armor for cancer. Extracellular vesicles (EVs) from MSCs have been evaluated for anticancer effects. In vivo targeting of EVs to the tumor is an essential requirement for successful therapy. Therefore, non-invasive methods of monitoring EVs in animal models are crucial for developing EV-based cancer therapies. The present study to develop bioluminescent EVs using Renilla luciferase (Rluc)-expressing MSCs. The EVs from MSC/Rluc cells (EV-MSC/Rluc) were visualized in a murine lung cancer model. The anticancer effects of EVs on Lewis lung carcinoma (LLC) and other cancer cells were assessed. EV-MSC/Rluc were visualized in vivo in the LLC-efffuc tumor model using optical imaging. The induction of apoptosis was confirmed with Annexin-V and propidium iodide staining. EV-MSC/Rluc and EV-MSCs showed a significant cytotoxic effect against LLC-effluc cells and 4T1; however, no significant effect on CT26, B16F10, TC1 cells. Moreover, EV-MSC/Rluc inhibited LLC tumor growth in vivo. EV-MSC/Rluc-mediated LLC tumor inhibitory mechanism revealed the decreased pERK and increased cleaved caspase 3 and cleaved PARP. We successfully developed luminescent EV-MSC/Rluc that have a therapeutic effect on LLC cells in both in vitro and in vivo. This bioluminescent EV system can be used to optimize EV-based therapy.


Results
Characterization of MSC/Rluc cells. The Rluc gene was stably transduced into mouse bone marrow-derived MSCs using Rluc-expressing lentiviral particles (Suppl. Figure 1A). The bioluminescent imaging (BLI) signal was higher in MSC/Rluc cells than in parental MSCs, and increased in a cell number-dependent manner (Suppl. Figure 1B; R 2 = 0.90). The presence of Rluc protein was confirmed by western blot analysis in MSC/Rluc cells, and no band was observed in parental MSCs (Suppl. Figure 1D). Furthermore, Rluc mRNA transcript expression in these cells was analyzed by RT-PCR. Rluc gene expression was detected in MSC/Rluc cells but not in parental MSCs (Suppl. Figure 1C). These results confirmed that Rluc was stably expressed in the MSC/Rluc cells. Next, we confirmed the expression of MSC phenotype markers on these cells using flow cytometry. The MSC/ Rluc cells expressed Sca-1, CD29, CD34, and were negative for CD117 (Suppl. Figure 1E). We also analyzed the differentiation potential of MSC and MSC/Rluc cells. MSC and MSC/Rluc cells stained positively for Alizarin Red S after differentiation treatments (Suppl. Figure 1F) for 14 days. The calcium deposits appear as bright orange-red stained areas in the light microscopy images. Both differentiated cultures stained positive (1F (ii) and 1F (iv)) whereas no stain was detected in undifferentiated control cells. These results confirmed that MSC/Rluc cells were capable of osteogenic differentiation, similar to parental MSCs. Furthermore, we compared the proliferation rates of MSC/Rluc cells and parental MSCs. The proliferation rate of MSC/Rluc cells was similar to the rate of parental MSCs after 48 h (Suppl. Figure 1G).

LLC-effluc activity in vitro.
LLC-effluc activity, visualized by an IVIS system, increased with increasing cell numbers compared to parental LLC cells (Suppl. Figure 2A). The quantitative data (Suppl. Figure 2B; R 2 = 0.98) shows a sequential increase in effluc activity. These results confirmed the presence of effluc protein activity in the stable LLC-effluc cells.

Characterization of EV-MSC/Rluc. To investigate the effects of MSC and MSC/Rluc derived EVs on
tumors, we first isolated EVs from culture supernatants of these cells. The successful purification of MSC/ Rluc-derived EVs was examined by western blotting. First, we compared the protein expression profiles of EVs derived from MSC and MSC/Rluc cells with the profiles of their corresponding donor cells. Equal amounts of proteins extracted from MSC and MSC/Rluc derived EVs were separated by 10% SDS-PAGE and stained with Coomassie Blue (Fig. 1A). As shown in Fig. 1B, MSC/Rluc cells and EVs were positive for expression of Alix, a multivesicular body marker. CD63 was detected in EV-MSC/Rluc, whereas GM-130, a Golgi marker, and calnexin, a marker of endoplasmic reticulum, were not detected. The purity of EVs was thus confirmed. Fractions of putative EVs were isolated from conditioned media of MSC and MSC/Rluc cells to investigate the paracrine effect of MSCs via EV release. High-resolution transmission electron micrographs showed that MSC and MSC/Rluc derived EVs exhibited rounded and lipid bilayer structures with a mean size of approximately 100 nm (Fig. 1C). Nanosight analysis was performed to further define the size of these EVs, and the size distribution profile displayed a mean size of around 100 nm for EV-MSC and EV-MSC/Rluc (Fig. 1D). These results demonstrated the successful purification of EV-MSC and EV-MSC/Rluc. Next, we confirmed Rluc activity in the EVs using different concentrations of EV-MSC/Rluc and EV-MSC by IVIS upon adding coelenterazine (CTZ). Rluc activity was detected in EV-MSC/Rluc, whereas no bioluminescence was detected in EV-MSC ( Fig. 2A). The Rluc activity of EV-MSC/Rluc increased with increasing concentration (10, 15, and 20 μ g EVs) compared to EV-MSC upon quantification ( Fig. 2B; R 2 = 0.99). In addition, we confirmed the presence of Rluc proteins in EV-MSC/Rluc by western blotting (Fig. 2C). Furthermore, we also performed a dot blot analysis for Rluc (Fig. 2D). The Rluc protein increased between adding 4.5 μ g and 9 μ g EV-MSC/Rluc.

Cellular uptake of EV-MSC/Rluc into LLC-effluc cells.
To study the internalization of EV-MSC/Rluc by LLC cells, EV-MSC/Rluc were labeled with the fluorescent dye DiD. DiD-labeled EV-MSC/Rluc were incubated with LLC cells for 4 h and cellular uptake of EV-MSC/Rluc was observed by confocal laser microscopy (Fig. 3). We found that DiD-labeled EVs were localized in the LLC cells, indicating that EV-MSC/Rluc were internalized into tumor cells.

Effect of EV-MSC and EV-MSC/Rluc on LLC luciferase activity and viability of other cancer cell types.
LLC-effluc cells were treated with EV-MSC and EV-MSC/Rluc for 24 h, and then LLC-effluc activity was measured by IVIS imaging (Fig. 4A-D). Various concentrations of EV treatment significantly decreased the viability of LLC-effluc cells. EVs treatment at 5, 10, and 20 μ g significantly decreased the LLC-effluc activity (p < 0.05, 0.001), compared to untreated LLC-effluc cells. The present results confirmed that EVs from parental MSC and Rluc labeled MSC have a similar effect. We also analyzed the therapeutic effect of EV-MSC and EV-MSC/Rluc on other cancer cell types such as melanoma, cervical, breast, and colon cancer, and we found that the breast cancer cell line 4T1 exhibited significantly decreased viability upon EV treatment compared to other types of cancer cells (Suppl. Figure 3).

Effects of EV-MSC/Rluc on LLC cell apoptosis.
Flow cytometric analysis of FITC-Annexin V staining was performed on LLC cells to assess the apoptotic effects of EV treatment (Fig. 4E). LLC cells were treated with EV-MSC/Rluc at 10 or 20 μ g/mL for 24 h. After 24 h, cells were incubated with FITC-Annexin V in a buffer containing propidium iodide (PI) and then were subjected to flow cytometric analysis. Untreated cells were primarily FITC-Annexin V-and PI-negative, indicating that they were viable and not undergoing apoptosis. After 24 h treatment with EVs, two populations of cells were observed. Viable cells that were not undergoing apoptosis (FITC-Annexin V and PI negative) constituted 84% and 76% of the total population in cells treated with 10 and 20 μ g/mL EV-MSC/Rluc respectively. In contrast, cells undergoing early apoptosis (FITC-Annexin V positive and PI negative) constituted 14.22% and 19.52% of the total cell population treated with 10 or 20 μ g/mL EV-MSC/ Rluc, respectively, compared to untreated control cells (p < 0.05, 0.001). Therefore, the present study confirmed that EV-MSC/Rluc treatment induces apoptosis in LLC cells.

Mechanism of EV-MSC/Rluc on LLC apoptosis.
To determine the EV-MSC/Rluc-mediated cell death mechanism, we examined ERK1/2. Figure 4F depicts the western blot analysis for ERK, phosphorylated ERK, cleaved caspase 3, and cleaved PARP in cells treated with or without EVs. In the present study, EV-MSC/Rluc treatment increased the levels of cleaved caspase 3 (8.4 fold), and PARP (6.5 fold). ERK1/2 is involved in one of the mitogen-activated protein kinase (MAPK) pathways that promote cell cycle progression. In this study, EV-MSC/Rluc decreased pERK1/2 levels (0.03 fold); however, it did not change total ERK levels. The observed decrease in pERK1/2 explains the reduction in LLC, cell viability, and induction of apoptosis.

In vivo visualization of EV-MSC/Rluc and the anti-tumor effect of EVs on LLC tumors.
To determine the effect of EVs on tumor growth in vivo, immune-competent (C57BL6) mice were utilized. Intratumor injection of 50 μ g EV-MSC/Rluc was performed at days 21 and 26 after xenograft implantation. A schematic diagram of the in vivo study is depicted in Suppl. Figure 4. The Rluc activity of mice injected intratumorally with EV-MSC/Rluc is shown in Suppl. Figure 5. Rluc activity of EV-MSC/Rluc was successfully visualized in LLC tumors. The LLC-effluc activity of the tumors upon IVIS imaging is shown in Fig. 5. Five days after the second injection of EV-MSC/Rluc, the LLC-effluc activity was analyzed. The LLC-effuc activity of the EV-MSC/Rluc treated group was decreased compared to the activity of the vehicle-treated group (Fig. 5A,B) on day 26 (p < 0.05) and day 32 (p < 0.05). The tumor weight (Fig. 5C) in the EV-MSC/Rluc-treated group was also decreased significantly compared to the vehicle-treated group (p < 0.05). In addition, the ex vivo effluc activity (Fig. 5D,E) of EV-MSC/ Rluc treated tumors was also decreased significantly compared to effuc activity in control tumors (p < 0.001).

Discussion
In the current study, we demonstrated the presence of the Rluc reporter protein in EVs derived from Rluc-expressing MSCs, as well as the visualization of EV-MSC/Rluc in the LLC-xenograft tumor model. We also studied the anticancer effect of EV-MSC/Rluc, and to our knowledge, this is the first study to examine MSC-derived EVs with an Rluc reporter. EVs, or exosomes, are nano-sized vesicles of endocytic origin and are  30,31 . MSC-derived EVs carry a large cargo load and protect their contents from chemicals or degradative enzymes 32 . Microparticles derived from MSCs could represent a novel therapeutic tool. These vesicles have significant effects on target cells, and this treatment approach has been exploited in many diseases, such as cardiovascular diseases, renal diseases, and cancer 6,18,32,33 .
First, we successfully transduced the Rluc reporter gene into MSCs and isolated EVs from the culture supernatant of these MSC/Rluc cells. We observed that MSC-derived EVs showed unique protein expression patterns as reported by studies with other cells 34,35 . The EV purity was confirmed by western blot analysis for CD63, which is a positive marker of EVs 36 . Alix is a cargo protein of the multivesicular body (MVB) that was observed in both cells and EVs. Further, the absence of GM130 and calnexin in EVs showed that they were free from contamination with cell organelles. These results confirmed that the EVs isolated from MSC/Rluc cells were pure. The size and morphology of EVs were examined with TEM, and the mean size of EV-MSC and EV-MSC/Rluc, as confirmed by NanoSight analysis, was about 100nm, which is similar to the findings of previous studies [37][38][39] .
Further, we confirmed Rluc activity and the presence of Rluc protein in the EV-MSC/Rluc. These results suggest selective distribution of the reporter protein in EVs and support the observations of previous studies with other reporter proteins 40,41 . Therefore, these results showed that this novel bioluminescent reporter protein can be used to label EVs.
MSCs can be isolated from different sources, mainly the bone marrow and adipose tissue, and can differentiate into various cell types of the mesodermal lineage. The characteristics of MSCs make them good candidates for therapeutic intervention in regenerative medicine or immunologic disorders 42 . Furthermore, MSCs have the potential for oncologic application as they are known to inhibit tumor progression when administered into established tumors [43][44][45] . However, MSCs also can have oncogenic potential, as was found for several tumor types, including osteosarcomas, lipomas, and gastric adenocarcinomas [46][47][48] , strongly suggesting their involvement in tumor development. Therefore, any clinical application of MSCs needs to be evaluated cautiously. In addition, the timing of MSC injection is crucial for therapeutic intervention 22 . Some studies have demonstrated that the co-injection of MSCs with tumor cells promotes tumor growth 49,50 . Therefore, administering EVs may have some advantages over administering MSCs. There is growing interest in using EVs as biological delivery vehicles. EVs can be taken up by acceptor cells, and thereby a number of cellular processes can be altered 51   There is evidence that EVs do not elicit acute immune rejection; they are non-viable, and thus they do not have tumorigenic potential. MSC-derived EVs can be mass-produced in culture and can be incorporated with multiple therapeutic miRNAs, thus enabling personalized treatment 51,53,54 . Moreover, MSC-derived EVs have no risk of aneuploidy, teratoma formation, or immune rejection after allogeneic administration in vivo 20,55,56 . Hence, use of MSC-derived EVs for disease treatment is a promising therapeutic strategy. In the present study, exposure to EV-MSC/Rluc resulted in inhibition of LLC tumor growth in vitro and in vivo. EV-MSC and EV-MSC/Rluc significantly inhibited LLC-effluc activity with increasing concentrations. In addition, we analyzed the effects of EV-MSC and EV-MSC/Rluc on the viability of other cancer cells and found that 4T1 cells were significantly inhibited by EV treatment; however, no significant changes were observed in other cancer cells such as B16F10, TC-1, and CT26. Although MSC-derived EVs are not effective on all types of cancers, they have definitive therapeutic effects on certain types of cancers. Further, induction of LLC cell apoptosis was observed upon EV-MSC/ Rluc treatment at doses of 10 and 20 μ g/mL. These data indicate that MSC/Rluc derived EVs have the potential to inhibit LLC.
In order to demonstrate the mechanism of cell death induced by EV treatment, we analyzed ERK, pERK, cleaved caspase 3, and cleaved PARP. The most important signaling molecule of the caspase cascade is caspase 3, which is a downstream effector caspase 57,58 . Activation of caspase 3 then leads to the cleavage of cellular substrates like PARP (poly (ADP)-ribose polymerase) 59 . In the present study, EV-MSC/Rluc treatment inhibited mitogenic signaling by downregulating pERK1/2 and activated apoptosis signaling by increasing cleaved caspase 3 and cleaved PARP. The EV-MSC/Rluc-mediated induction of apoptosis was also confirmed by flow cytometry with annexin V staining.
The advent of molecular imaging technologies, coupled with the development of cell-based therapies, has brought about a revolution in in vivo visualization. Among optical imaging tools, bioluminescence imaging with reporter genes is an indirect or direct labeling technique and is a promising method for visualizing biological targets in small animal models. Bioluminescence is generated by the conversion of chemical energy into visible light by the action of luciferase enzymes on their substrates [60][61][62][63] . In the current study, EV-MSC/Rluc isolated from MSC/Rluc were injected intratumorally and were well visualized with BLI upon addition of CTZ in the LLC xenograft tumor model. We confirmed that the bioluminescent EVs retained Rluc activity in the in vivo tumor model.
For EV treatment of tumors in vivo, LLC cells labeled with effluc were used and the therapeutic effect of EVs on the tumor xenograft was monitored non-invasively by BL imaging. EV-MSC/Rluc (50 μ g) were administered twice with a 5 day interval by direct intratumor injection in the rapidly growing LLC tumors, and PBS was In the current study, we visualized both the EVs and xenografts in animals in a non-invasive manner using two different luciferase reporter genes combined with the appropriate substrates and a two-color reporter system. This is an ideal system to visualize the therapeutic agent and the target together in the same animal and can be used to optimize EV therapy and other therapeutic options as well.
Additional studies are still needed to optimize the delivery of EVs to multiple tumors via systemic administration for better clinical translation. Technological improvements for culture and purification of EVs are needed to increase the efficiency of EV isolation and EV safety, in order to determine the feasibility of MSC-derived EV therapy. Optimization of EV therapy can improve the therapeutic effect either when used alone or in combination with other therapeutic modalities.
In the current study, an Rluc reporter system in MSC-derived EVs was successfully utilized and the presence of Rluc protein was confirmed in the EVs. EV-MSC/Rluc were non-invasively visualized in an LLC tumor xenograft model. In addition, the MSC-derived EVs were found to inhibit proliferation of LLC cells in vitro and to inhibit progression of LLC tumor xenografts in vivo. Thus, our findings suggest that MSC-derived EVs can be effective therapeutic agents against lung cancer.

CCK-8 cell proliferation assay for naïve MSC and MSC/Rluc cells. An in vitro proliferation assay was
performed on MSCs and MSC/Rluc cells using a CCK-8 assay kit (Dojindo Laboratories, Kumamoto, Japan). Briefly, MSCs and MSC/Rluc cells (5,000 cells/well) were plated in 96-well plates and were incubated for 48 h at 37 °C, 5% CO 2 . After 48 h, 10 μ L of CCK-8 solution was added to each well followed by incubation for 1 h at 37 °C. Absorbance at 450 nm was then measured using a microplate reader.

Transduction of enhanced firefly luciferase (effluc) gene into LLC cells. LLC cells were cultured in
DMEM-High (Hyclone) supplemented with 10% fetal calf serum and 1% antibiotic-antimycotic solution (Gibco), at 37 °C and 5% CO 2 . LLC cells were transduced with effluc-expressing retroviral particles isolated from Phoenix A-effluc cells. The transduced LLC-effluc cells were sorted with magnetic beads bound to a CD90.1 MicroBeads (thy1) antibody (MACS, Miltenyi Biotec, South Korea) for selecting effluc positive cells, and their effluc activity was quantified by IVIS imaging using D-luciferin as a substrate.

Isolation of EVs. MSC and MSC/Rluc-derived EVs were isolated by an ultracentrifugation method. Briefly,
MSCs and MSC/Rluc cells were cultured in the presence of FBS depleted of EVs (by centrifugation at 100,000 × g for 18 h at 4 °C). EVs were then purified from the cell-free supernatant according to a previously described protocol with slight modifications 65 . After 2-3 days of culture, the cell supernatants were collected and centrifuged at 300 × g for 10 min to remove live cells, then at 1500 × g for 15 min to remove cell debris, and finally at 2500 × g for 20 min to remove apoptotic bodies. This supernatant was filtered through a 0.45 μ m syringe filter and centrifuged at 100,000 × g for 60 min at 4 °C to obtain EVs. To remove protein contamination, the EV pellet was washed by resuspension in phosphate buffered saline (PBS) followed by another centrifugation at 100,000 × g for 60 min. All centrifugations were performed at 4 °C. The residual pellet was resuspended in 50-100 μ L of PBS, stored at − 20 °C, and functional assays were performed within a week. Ultracentrifugation was performed (SW28 rotor; Ultra-Clear tube) using the Optima L-100 XP ultracentrifuge (Beckman Coulter, USA). The protein content of EVs was quantified by the bicinchoninic acid method (BCA kit, Pierce, Appleton, WI, USA).

Western blot analysis and dot blot for Rluc.
For western blot analysis, protein samples (30 to 80 μ g) were separated using 10% SDS-PAGE and then were transferred to PVDF membranes (Millipore, USA). The blots were incubated with primary antibodies against Rluc (GeneTex, USA), Alix (Abcam), GM-130 (Abcam), calnexin (Abcam), CD63 (Abcam), pERK, ERK (Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA), cleaved caspase 3 (Cell Signaling Technology, USA), and cleaved PARP (Cell Signaling Technology, USA) followed by incubation with the appropriate HRP-tagged secondary antibody (Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA). The protein-antibody complexes were visualized using an enhanced chemiluminescence kit (Pierce, USA). The dot blot method for detecting proteins was performed as described previously 66 . Here, Rluc protein was detected from EV-MSC/Rluc in PBS spotted onto nitrocellulose membranes at protein amounts of 4.5 and 9 μ g, by a 2-h incubation with Rluc antibody followed by incubation with streptavidin-HRP conjugate and detection by chemiluminescence. ImageJ 1.38 software was used to quantify the band intensity (Windows version of NIH Image; http://rsb.info.nih.gov/nih-image/).

Transmission electron microscopy (TEM).
The MSC and MSC/Rluc EV pellets were obtained by ultracentrifugation and fixed at 4 °C overnight. The fixative contained 2.5% glutaraldehyde in 0.01 M phosphate buffer at pH 7.4 (filtered through 0.22 μ m filters) and cells were washed with PBS. EVs were post-fixed in 1% OsO 4 (Taab Laboratories Equipment Ltd.) for 30 min. EV pellets were washed with distilled water and dehydrated with graded ethanol solutions. EV pellets were negatively stained with 1% uranyl-acetate in 50% ethanol for 30 min and embedded in Taab 812 (Taab), followed by overnight polymerization at 60 °C and ultrasectioning for TEM. Ultrathin sections were examined and images were captured with a HT 7700 transmission electron microscope (Hitachi, Tokyo, Japan), operated at 100 kV. Nanoparticle Tracking Analysis. EV-MSC and EV-MSC/Rluc resuspended in PBS were further diluted (1 μ l/mL) in distilled water for measuring size distribution using the NanoSight LM10 instrument. This method is Scientific RepoRts | 6:30418 | DOI: 10.1038/srep30418 based on conventional optical microscopy and uses a laser light source to illuminate nano-scale particles. The EV solutions (1 mL) were introduced into the viewing unit using a disposable syringe. Enhanced with a perfect black background, individual particles appeared as point-scatters moving under Brownian motion. The Nanoparticle Tracking Analysis (NTA) software suite allowed automatic tracking and size determination of the nanoparticles on an individual basis. Results are displayed as a frequency size distribution graph.

In vivo visualization of EV-MSC/Rluc and anti-tumor effects of EVs on LLC tumors. LLC-effluc
cells (1 × 10 6 ) were subcutaneously implanted into male C57BL/6 mice (Charles River, Jackson Laboratories). Tumors were allowed to grow for 2-3 weeks after implantation, and tumor establishment was assessed by measuring effluc activity using IVIS bioluminescence imaging. Animals with tumors exhibiting near similar photon flux values were used for EV treatment. The treatment protocol used for this study has been described previously, with slight modifications 15 . Mice were randomized into two treatment groups: (1) the control group was injected with vehicle (PBS) alone (n = 3); (2) the EV group received intratumoral injections of EVs (n = 3). The intratumor injection of EV-MSC/Rluc was initiated on day 21 after xenograft implantation. The first treatment was 50 μ g of EV-MSC/Rluc in 25 μ L of PBS. IVIS bioluminescence imaging was performed to assess the in vivo Rluc activity of the injected EV-MSC/Rluc. The second dose on day 26 after xenograft implantation was identical to the dosage given in the first treatment. LLC-effluc activity was monitored on days 26 and 32. Five days after the second EV treatment dose, the mice were sacrificed and tumors were harvested to examine ex vivo effluc activity and tumor weight. Ethics Statement. All described procedures were reviewed and approved by Kyungpook National University (KNU-2012-43) Animal Care and Use Committee, and performed in accordance with the Guiding Principles for the Care and Use of Laboratory Animals.