Antitumor effect of oncolytic virus and paclitaxel encapsulated in extracellular vesicles for lung cancer treatment

This is a self-archived version of an original article. This version may differ from the original in pagination and typographic details. Author(s): Title: Year: Version: Copyright: Rights: Rights url: Please cite the original version: CC BY 4.0 https://creativecommons.org/licenses/by/4.0/ Antitumor effect of oncolytic virus and paclitaxel encapsulated in extracellular vesicles for lung cancer treatment © 2018 The Authors. Published by Elsevier B.V. Published version Garofalo, M.; Saari, H.; Somersalo, P.; Crescenti, D.; Kuryk, L.; Aksela, L.; Capasso, C.; Madetoja, M.; Koskinen, Katariina; Oksanen, T.; Mäkitie, A.; Jalasvuori, Matti; Cerullo, V.; Ciana, P.; Yliperttula, M. Garofalo, M., Saari, H., Somersalo, P., Crescenti, D., Kuryk, L., Aksela, L., Capasso, C., Madetoja, M., Koskinen, K., Oksanen, T., Mäkitie, A., Jalasvuori, M., Cerullo, V., Ciana, P., & Yliperttula, M. (2018). Antitumor effect of oncolytic virus and paclitaxel encapsulated in extracellular vesicles for lung cancer treatment. Journal of Controlled Release, 283, 223-234. https://doi.org/10.1016/j.jconrel.2018.05.015 2018 Contents lists available at ScienceDirect Journal of Controlled Release journal homepage: www.elsevier.com/locate/jconrel Antitumor effect of oncolytic virus and paclitaxel encapsulated in extracellular vesicles for lung cancer treatment M. Garofalo a,b,⁎ , H. Saari a,1 , P. Somersalo a,b,1 , D. Crescenti b , L. Kuryk a,c,d , L. Aksela a , C. Capasso e , M. Madetoja f , K. Koskinen g , T. Oksanen a , A. Mäkitie h , M. Jalasvuori a,g , V. Cerullo e , P. Ciana b , M. Yliperttula a,⁎ a Division of Pharmaceutical Biosciences and Centre for Drug Research, University of Helsinki, Viikinkaari 5, Helsinki 00790, Finland b Department of Oncology and Hemato-Oncology, Center of Excellence on Neurodegenerative Diseases, University of Milan, Via Balzaretti 9, Milan 20133, Italy c National Institute of Public Health –National Institute of Hygiene, Department of Virology, 24 Chocimska str, 00-791 Warsaw, Poland d Targovax Oy, R&D, Clinical Science, R&D, Saukonpaadenranta 2, 00180 Helsinki, Finland e Laboratory of ImmunoViroTherapy, Drug Research Program, Faculty of Pharmacy, University of Helsinki, Viikinkaari 5, Helsinki 00790, Finland f Made Consulting, Tykistökatu 4 B, FI-20520 Turku, Finland g Biological and Environmental Science, Nanoscience Center, University of Jyväskylä, Survontie 9C, 40500, Finland h Department of Otorhinolaryngology –Head and Neck Surgery, Helsinki University Hospital and University of Helsinki, P.O.Box 263, FI_00029 HUS, Helsinki, Finland ARTICLE INFO Keywords: Extracellular vesicles Oncolytic viruses Cancer therapy Drug delivery Paclitaxel Xenograft animal model Lung cancer ABSTRACT Standard of care for cancer is commonly a combination of surgery with radiotherapy or chemoradiotherapy. However, in some advanced cancer patients this approach might still remaininefficient and may cause many side effects, including severe complications and even death. Oncolytic viruses exhibit different anti-cancer mechanisms compared with conventional therapies, allowing the possibility for improved effect in cancer therapy. Chemotherapeutics combined with oncolytic viruses exhibit stronger cytotoxic responses and oncolysis. Here, we have investigated the systemic delivery of the oncolytic adenovirus and paclitaxel encapsulated in extracellular vesicles (EV) formulation that, in vitro, significantly increased the transduction ratio and the infectious titer when compared with the virus and paclitaxel alone. We demonstrated that the obtained EV formulation reduced the in vivo tumor growth in animal xenograft model of human lung cancer. Indeed, we found that combined treatment of oncolytic adenovirus and paclitaxel encapsulated in EV has enhanced anticancer effects both in vitro and in vivo in lung cancer models. Transcriptomic comparison carried out on the explanted xenografts from the different treatment groups revealed that only 5.3% of the differentially expressed genes were overlapping indicating that a de novo genetic program is triggered by the presence of the encapsulated paclitaxel: this novel genetic program might be responsible of the observed enhanced antitumor effect. Our work provides a promising approach combining anticancer drugs and viral therapies by intravenous EV delivery as a strategy for the lung cancer treatment. 1. Introduction Despite major advances in conventional cancer treatments with surgery, radiotherapy, chemotherapy,and their combination, the outcome is still partially ineffective against numerous cancer types, like lung cancer [1]. Lung cancer is highly invasive and rapidly metastasizing, often diagnosed at an advanced stage with poor prognosis and without efficient treatment options [2]. Given the poor survival rate of patients, new therapeutic strategies with systemic drug delivery are warranted. Oncolytic virotherapy is emerging as a promising and potential approach to treat cancer, and the approval of the first oncolytic virus, Imlygic (T-Vec, talimogene laherparepvec), in the Western world by US Food and Drug Administration (FDA) and European Medicines Agency (EMA) provides new perspectives for improved treatment of cancer [3,4]. Indeed, its application can be particularly relevant for tumors without curative options, including metastatic lung cancers [1]. In oncolytic virus therapy, viruses are specifically engineered to preferentially infect, replicate in and kill cancer cells instead of normal cells where their normal functions are restricted [5–9]. Virus replication in tumor cells eventually leads to cell lysis, allowing https://doi.org/10.1016/j.jconrel.2018.05.015 Received 21 December 2017; Received in revised form 10 May 2018; Accepted 14 May 2018 ⁎ Corresponding authors at: Division of Pharmaceutical Biosciences and Centre for Drug Research, University of Helsinki, Viikinkaari 5, Helsinki 00790, Finland. 1 Shared co-authorship. E-mail addresses: mariangela.garofalo@helsinki.fi(M. Garofalo), marjo.yliperttula@helsinki.fi(M. Yliperttula). Journal of Controlled Release 283 (2018) 223–234 Available online 01 June 2018 0168-3659/ © 2018 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/BY/4.0/). T the new virus progeny to spread to surrounding cells and even to distant metastases through circulation [10]. However, as a single therapeutic agent oncolytic adenoviruses have not been observed to efficiently destroy large tumor mass in patients [11,12]. Thus there is a need to enhance their antitumor efficacy by combining viral therapy with other anticancer agents [13–15]. Cisplatin has improved oncolysis of Herpes simplex virus type 1 (HSV-1) in non small cell lung cancer (NSCLC) [16]. Additionally, combination of cisplatin with adenovirus facilitated the replication of the virus and significantly reduced the tumor progression [17]. Enhanced effects have been also reported in a malignant pleural mesothelioma (MPM) with NV1066 (HSV-1 based virus) [18]. However, the use of oncolytic viruses as a potential approach to treat cancer has also disadvantages [19]; the immune response will presumably limit ongoing viral replication and spread in cancer cells. Administered viruses will be detected by immune system and inactivated by neutralizing antibodies, decreasing its replication and efficacy. Additionally, given the intratumoral administration of oncolytic viruses [20–22], they are eligible only in injectable lesions and thus limiting the approach to treat many solid tumors. Systemic delivery of virus together with anticancer drugs would circumvent some of these limitations. Extracellular vesicles (EVs) are naturally occurring cargo delivery agents with the potential to be used as drug delivery vehicles [23,24] since they can transfer biological molecules even over long distances within the body [25]. The lipid membrane of EVs can protect the cargo from degradation by body fluids and further improved uptake by the target cells [26,27]. In recent studies, it has been shown to that oncolytic viruses can be delivered into the nucleus of tumorigenic cells by tumor microparticles while simultaneously avoiding the production of neutralizing antibodies and mediating the virus entry into cancer cells independently from the virus-specific receptor [28]. In this study, we set to investigate whether it is possible to encapsulate the oncolytic adenovirus and chemotherapeutic agent into EVs in an attempt to utilize them as carriers for targeted drug delivery. The obtained formulations were tested in vitro in lung cancer cell line and subsequently in vivo in lung cancer xenograft animal model using both intra tumor (it) and intra venous (iv) injections. Abraxane (paclitaxel, albumin-bound nanoparticle formulation) and EVs without encapsulated virus and/or drug were used as control samples. We found that the systemic delivery of both oncolytic virus and paclitaxel encapsulated in EVs resulted in improved drug efficacy and reduced offtarget toxicity. 2. Materials and methods 2.1. Cell culture A549 human lung cancer cell line was purchased from the American Type Culture Collection (ATCC, USA). The cells were cultured at 37 °C and 5% CO 2 in Dulbecco's modified eagle medium (DMEM, Lonza, Switzerland) supplemented with 10% fetal bovine serum (FBS, Gibco Laboratories, USA), 1% of 100 u/mL penicillin/streptomycin (Gibco Laboratories) and 1% L-glutamine (Gibco Laboratories).PNT2 (European Collection of Authenticated Cell Cultures, ECACC, UK) human prostate cell line was purchased from Sigma-Aldrich. The cells were cultured at 37 °C and 5% CO 2 in RPMI 1640 (Gibco Laboratories) supplemented with 10% FBS (Gibco Laboratories), 1% of 100 u/mL penicillin/streptomycin (Gibco Laboratories) and 2% L-glutamine (Gibco Laboratories). The prostate cancer cell line PC-3 (ATCC), was cultured at 37 °C and 5% CO 2 in Ham's F-12 K (Kaighn's) basal medium (Gibco Laboratories) supplemented with 10% FBS (Gibco Laboratories) and 1% of 100 u/mL penicillin/streptomycin (Gibco Laboratories). 2.2. Oncolytic virus Ad5D24-CpG, was generated according to standard protocols [29] by recombining a CpG-rich shuttle plasmid (pTHSN-CpG1) with a plasmid containing the 24 adenovirus backbone. Viral stocks were expanded in human lung cancer cell line A549 and purified on cesium chloride gradients. The viral particle concentration was determined by OD 260 -reading and standard TCID 50 (tissue culture infectious dose 50) assay was performed to determine infectious particle titer. Virus was characterized by PCR and restriction enzyme analysis Ad5D24-RFP, expressing a red fluorescent protein (RFP) was kindly provided by Dr. Masataka Suzuki from Baylor College of Medicine (Houston, TX [30];). 2.3. Paclitaxel (PTX) solutions A 50 mM stock solution of Paclitaxel (PTX; Selleck Chemicals) was prepared by dissolving PTX into di-methyl sulfoxide (DMSO) (SigmaAldrich). This was used as the stock solution in A549 cell experiments as well as in EV-encapsulation of PTX.Abraxane was provided for us by the Hospital Pharmacy of the Helsinki University Hospital based on the prescription of MD A. Mäkitie, School of Medicine, University of Helsinki, Finland. A stock suspension of Abraxane (Albumin-PTX conjugate, Celgene, USA) was prepared by suspending the powder, corresponding to 100 mg of PTX, in phosphate buffer saline (PBS, Lonza) to a final volume of 8.5 mL, resulting in 11.76 mg/mL of PTX. This stock was then further diluted in PBS to produce the solutions used in the in vivo animal experiments exclusively. 2.4. Production of extracellular vesicles (EV) and PTX loaded EVs formulations In order to produce EVs 2.6 × 10 6 A549 cells were plated into T-175 flask in medium supplemented with 5% FBS. The FBS growth media was ultra-centrifuged overnight (110,000 ×gat 4 °C for 18 h, Optima LE-80 K ultracentrifuge, rotor type 50.2, Beckman Coulter) to remove EVs present in serum. Cells were cultured at 37 °C and 5% CO 2 until cytopathic effect was seen, where upon the media was collected. EVs were isolated from the conditioned medium using differential centrifugation. First the conditioned medium was centrifuged at 500 ×gin 4 °C for 10 min to pellet cells (Allegra X-15R Centrifuge, Beckman Coulter). Then, the supernatant was collected and ultra-centrifuged for 2 h at 100000 ×gin 4 °C, using Optima L-80 XP ultracentrifuge (Beckman Coulter) with rotor SW32Ti (Beckman Coulter). The supernatant was aspirated and EVcontaining pellets containing resuspended in PBS (Lonza) 100 μL and stored at −80 °C. PTX-loaded EVs were prepared as previously described by us [23] by incubating 1 × 10 8 –5×10 9 EVs in 1 mL of 5 μM PTX-DPBS solution for in vitro samples and 10 μM PTX-DPBS solution for in vivo samples, for 1 h at 22 °C. Next, the samples were centrifuged at 170000 ×gfor 2 h to pellet the EVs. The supernatant containing unbound PTX was removed, and the EV-pellet was washed by suspending it in DPBS and pelleting it again at 170000 ×g. 2.5. Production of EV-Virus and EV-Virus-PTX formulations In order to produce EV-encapsulated virus (EV-Virus), 2.6 × 10 6 of A549 cells were infected with with 10 viral particles/cell of Ad5D24CpG and were cultured at 37 °C and 5% CO 2 48 h later when most of the cells were detached from the culture flask, the culture media were collected for EV-Virus isolation using differential centrifugation. First the conditioned medium was centrifuged at 500 ×g and 4 °C for 10 min, to separate the cells (Allegra X-15R Centrifuge, Beckman Coulter). Then, the supernatant containing EV-Virus was collected and ultra-centrifuged for 2 h at 100000 ×gand 4 °C, using Optima L-80 XP ultra-centrifuge (Beckman Coulter) with rotor SW32Ti (Beckman Coulter). The supernatant was aspirated and pellets containing EV-Virus re-suspended in PBS 100 μL and stored at −80 °C.EVVirus samples were incubated in 100 mM NaOH at room temperature for 20 min in order to inactivate any free not EV encapsulated virus present. Free virus used as controls was always inactivated for each M. Garofalo et al. Journal of Controlled Release 283 (2018) 223–234 224 experiment performed as previously reported [31]. Samples were subsequently neutralized by the addition of HCl 0.1 M. To generate EV-Virus-PTX, the EV-Virus formulation was incubated in a 10 μM PTX solution, prepared by diluting 10 mM PTX in DMSO with PBS with the ratio of 1:1000. Incubation was carried out at RT with mixing for 1 h. Samples were then centrifuged at 150000 ×gfor 2 h at RT, in order to pellet EV-Virus-PTX. The washing procedure was repeated using PBS as diluent. The final EV-Virus-PTX pellet was resuspended in 100 μL of PBS and stored at −80 °C. 2.6. Quantification of PTX present in EV-Virus-PTX 50 μL of 1.5 × 10 10 EV-Virus-PTX/mL as well as supernatant from the second washing step associated with removal of free PTX, to enable validation of the washing procedure, was processed for ultra performance liquid chromatography (UPLC, Acquity UPLC System) using a Cortecs UPLC C18+ column, 2.1 × 50 mm, particle size 2.7 μm (Waters, USA). Additionally, a 10 μM PTX solution was prepared, and used as a control for sample processing. Sodium dodecyl sulfate (SDS) was added to samples to a final concentration of 5% (w/v) in order to lyse EV-Virus PTX. Vortexing was followed by incubation at RT for 1 h. Acetonitrile was then added to a final concentration of 75% (v/v) and samples vortexed. Precipitated proteins were pelleted by centrifugation at 10000 ×gfor 5 min at RT. The supernatant was analyzed by UPLC using gradient flow of acetonitrile from 30% to 80% in phosphate buffer, pH = 2, at 30 °C within 3 min. Detection and quantification of PTX, with a retention time of 1.7 min, was performed spectrophotometrically at the wavelength of 229 nm by using a reference standard curve. 2.7. Size distribution analysis by nanoparticle tracking analysis (NTA) Size distribution and concentration of EV, EV-Virus and EV-VirusPTX formulations were analyzed by NTA using Nanosight model LM14 (Nanosight) equipped with blue (404 nm, 70 mV) laser and sCMOS camera. The samples containing virus were incubated at +95 °C for 10 min in order to inactivate the viruses. NTA was performed for each sample by recording three 90 s videos, subsequently analyzed using NTA software 3.0 (Nanosight). The detection threshold was set to level 5 and camera level to 15. 2.8. Zeta potential analysis by electrophoretic light scattering The zeta potential was measured using ZetaSizer Nano (Malvern, UK). All the samples were diluted in a volume of 800 L of MilliQ H 2 O and injected with a 1 mL syringe in the capillary flow (DTS1070 folded capillary cell) for the measurement. An equilibration time of 120 s was set on the software to allow the samples to stabilize at 25 °C inside the measurement chamber. Three parallel measurements were performed on each sample. 2.9. HIM microscopy For Helium Ion Microscopy, A549 cells were cultured to 70% confluence in DMEM (Gibco Laboratories, USA) supplemented with heat inactivated 10% FBS (Gibco Laboratories) and 100 u/mL penicillinstreptomycin (Gibco Laboratories) on poly-L-lysin (Sigma Aldrich, Germany) glass coverslips. Cells in culture were infected with EV-Virus (5 × 10 9 /mL on 75 cm 2 80% confluent cell culture) by replacing the cell culture medium with a solution containing EV-Virus in DMEM of 2% FBS and 100 u/mL penicillin-streptomycin. After 2 h incubation, EV-Virus solution was again replaced with DMEM of 10% FBS and 100 μ/mL penicillin-streptomycin. Cells were fixed at different time points by replacing the EV-Virus solution with 2% glutaraldehyde (GA, Merck, USA) in 0.1 M Sodium Cacodylate (NaCac) buffer (pH 7.4). After GA fixation, the cells were washed twice with 0.1 M NaCac and further fixed with 1% OsO 4 in 0.1 M NaCac buffer (pH 7.4). After 1 h in OsO 4 , cells were washed twice with 0.1 M NaCac buffer and chemically dried in an increasing EtOH concentration series of 50%, 70%, 96%, and twice with 100%. After 100% EtOH, the cells were submerged in 98% hexamethyldisiloxane (Sigma Aldrich) and left to dry for overnight. After fixation and chemical drying, coverslips were mounted on stands and imaged with Orion NanoFab Helium Ion Microscope (Zeiss, Germany) using 30 kV acceleration voltage with beam current 0.2–1 pA. 2.10. Transduction assay Cells were seeded at a density of 1 × 10 4 cells/well in 96-well plates and maintained under appropriate condition. On the following day cells were treated in triplicates with an oncolytic adenovirus Ad5D24Rfp encoding for the red fluorescent protein (10vp/cell) and control EVs (10 particles/cell), EV-Virus (10 particles/cell), EV-PTX (10 particles/cell, 5μM of PTX), EV-Virus-PTX (10 particles/cell, 5 μM of PTX) or Virus and PTX separately (Virus+PTX) (10 vp/cell, 5 μM of PTX). EVs, EVVirus and EV-Virus-PTX had the same doses of particles as virus alone (10 vp were calculated as 10 EV-particles per cell in the well). The cells were then imaged with EVOS FL fluorescence microscope at 8, 24 and 48 h after the treatment in order to count the portion of cells expressing RFP. 2.11. Immunocytochemistry staining (ICC) The determination of the infectivity was based on the visual quantification of infected cells as previously described [32,33]. Cells were seeded at a density of 2 × 10 5 cells/well in 24 well plates, and maintained under appropriate condition in DMEM, completed with 5% FBS, 1% L-glutammine and 1% of penicillin/streptomycin. All the chemicals were purchased from Gibco Laboratories. On the following day cells were treated with Virus (10vp/cell), Virus + PTX (10vp/cell and 5 μM PTX solution), EV-Virus formulations (10 particles/cell) and EV-VirusPTX formulations (10 particles/cell, 5 μM of PTX). Plates were centrifuged for 90 min with 1000 ×gin 37°C and incubated for 48 h before staining at 37 °C and 5% CO 2 .48 h following the incubation cells were fixed by adding 250 μL of ice-cold methanol per well and incubated 15 min. Then cells were washed three times with PBS 1%-BSA (Bovine serum albumin, 9048-46-8 Sigma-Aldrich) solution and incubated in the dark for one hour with 1st antibody, mouse monoclonal anti-hexon 1:2000 (Novus Biological, NB600-413). After the incubation time, cells were washed three times with PBS 1%-BSA and incubated in the dark for other 1 h with 2nd antibody: Biotin-SPconjugated goat anti-mouse 1:500 (Jackson Immuno Research, 115065-062). After the incubation time, cells were washed three times with PBS1%-BSA and incubated in the dark for 30 min with extravidin-peroxidase (Sigma Aldrich, E2886). Finally cells were washed three times as indicated earlier and treated with Dab peroxidase substrate solution (Sigma Aldrich, A7284-50ML). To quench the reaction, cells were treated once with PBS. The detection of the infectious titer was performed using microscope EVOS, and each well was photographed with 5 pcs at 5 non-overlapping sites. The following formula was used to determine the infectious titer: L mL Infectious titer:x A(well) A(field) 11 v Where x = number of infected (stained cells). A(24 well) = 190 mm 2 . A(field) = surface area of the field. L = dilution. v = volume of virus dilution applied per well. M. Garofalo et al. Journal of Controlled Release 283 (2018) 223–234 225 2.12. MTS cell viability assay A549, PNT2 and PC-3 cells were seeded at a density of 1 × 10 4 cells/ well in 96-well plates and maintained under appropriate condition. On the following day cells were treated in triplicates with Virus (10vp/ cell), Virus + PTX (10vp/cell + 5 μM PTX solution), control EVs (10 particles/cell), EV-PTX (10 particles/cell, 5 μM of PTX), EV-Virus formulations (10 particles/cell), EV-Virus-PTX formulations (10 particles/ cell, 5 μM of PTX). Cell viability was determined by MTS assay according to the manufacturer's protocol (Cell Titer 96 AQueous One Solution Cell Proliferation Assay; Promega, Nacka, Sweden). The absorbance was measured with a 96-wells plate spectrophotometer Varioskan Flash Multimode Reader (Thermo Scientific) at 490 nm. The experiments were independently performed three times with triplicates of each condition in each experiment. 2.13. Analysis of apoptotic and necrotic cells A549 cells were plated into 6 well plates, 2 × 10 5 cells/well. Cells were treated with an oncolytic adenovirus Ad5D24CpG 10 vp/cell, Virus + PTX (10vp/cell + 5 μM PTX solution), control EVs (10 particles/cell), EV-PTX (10 particles/cell, 5 μM of PTX), EV-Virus formulations (10 particles/cell), EV-Virus-PTX formulations (10 particles/cell, 5μM of PTX). The amount of apoptotic and necrotic cells was measured after 24 h post-treatment with a TACS Annexin V-FITC kit (Trevigen Inc., Gaitherburg, MD, US) and BD LSRII flow cytometer according to the manufacturer's instruction. 2.14. In vivo xenograft animal experiments FIRST DAY OF TREATMENT Day 0 SECOND DAY OF TREATMENT Day 2 THIRD DAY OF TREATMENT Day 4 FOURTH DAY OF TREATMENT Day 15 EVs (1x109particles/tumor) XXX Virus (1x108vp/tumor) XXPBS Abraxane (10 mg of PTX/kg) PBS PBS X Virus+Abraxane (1x108vp/tumor +10 mg of PTX/kg) XXX EV-Virus (1x108particles/tumor + 1x108vp/tumor) XXPBS EV-Virus-PTX (1x108particles/tumor including 1x108 vp/tumor, and 10 mg of PTX/kg) XXX All the animal experiments performed under the ethical permission (ESAVI/10482/04.10.07/2015) of the National Laboratory Animal Board of Finland (Care and Use Committee) by Made Consulting Ltd. Oy (Turku, Finland) in GLP level animal facility. Mice were obtained from Janvier Labs (Barrier 4E-1, France) at 4 weeks of age. The acclimatization period was 13 days prior to A549 cancer cell injections, the cell line was the same used for the in vitro cell experiments. The A549 cell line purity was tested usingby IDEXX BioResearch - IMPACT III panelbefore inoculation and the cells viability of 99% were detected 50 min prior to first inoculation and 90% 20 min after the last inoculation using NucleoCounter NC-200. Health status of the mice was monitored daily and as soon as signs of pain or distress were evident they were euthanized. For the efficacy experiment, human xenografts were established by injecting 1.5 × 10 6 A549 cells s.c. into the flanks of 6-week old female BALB/c nude mice. The treatment groups were as follows: Virus (n = 6); Virus+Abraxane (n = 6); Abraxane (n = 6); EV (n = 6) and EV-Virus (n = 9); EV-Virus-PTX (n = 9). Treatment groups were administered i.v (100 μL) and i.t (50 μL) to mice with tumors (one tumor per mouse about 5 mm in diameter). The dosing days were 0, 2, and 4 for Virus (according to previous protocols [29]) and EV-Virus; 0 and 15 for Abraxane as previously reported [34]; 0, 2, 4 and 15 for EV and EV-Virus-PTX (table above and the Supplementary Table 1). The equation: 0.52 × length × (width) 2 , was used to calculate the tumor volumes to study the efficacy of the used EV-formulations. However, the average diameter of 15 mm was used as a limit to euthanize mice.Tumors, livers and spleens from each mouse were collected for histopathological examinations. 2.15. Quantitative PCR qPCR for adenovirus E4 copy number was carried out according to the protocol previously described [35] (primer FW:50-GGA GTG CGC CGA GAC AAC-30, primer RV: 50-ACT ACG TCC GGC GTT CCA T-30, probe E4: 50-(6FAM)-TGG CAT GAC ACT ACG ACC AAC ACG ATC T- (TAMRA)230). Total DNA was extracted from BALB/c nude murine samples (tumors, livers, blood) using the QIAamp DNA Blood Mini Kit (Qiagen, Hilden, Germany) according to manufacturer's protocol. Subsequently isolated DNA was analyzed for adenoviral E4 copy number normalized to murine beta-actin (liver, blood) and human beta-actin (tumor), respectively ((primer FW: 50-CGA GCG GTT CCG ATG C-30, primer RV: 50-TGG ATG CCA CAG GAT TCC AT-30, probe murine betaactin: 50-(6FAM)-AGG CTC TTT TCC AGC CTT CCT TCT TGG-(TAMRA) 230; (primer FW: 50-CAG CAG ATG TGG ATC AGC AAG-30, primer RV: 50CTA GAA GCA TTT GCG GTG GAC-30, probe human beta-actin: 50- (6FAM)- AGG AGT ATG ACG CCG GCC CCT C-(TAMRA)230). Samples were analyzed using LighCycler qPCR machine (LighCycler 480, Roche, Basel, Switzerland). 2.16. Histopathological studies Left lateral lobe of liver, half of spleen and one lobe of lungs were M. Garofalo et al. Journal of Controlled Release 283 (2018) 223–234 226 taken into 4% buffered formalin at necropsy. Finnish Centre for Laboratory Animal Pathology, Helsinki, Finland performed the histopathological evaluations of the lung, liver, spleen and tumor in vivo samples of nude (athymic) mice, implanted with human tumor xenograft and treated with virus alone, control EVs and EV-Virus-PTX by histotechnology. The samples were embedded into paraffin, cut at 4 μm (spleen 3 μm), and stained with haematoxylin and eosin. The histopathological evaluation was performed as a blind using 40×magnification using the microscope Zeiss Axio Imager.A2, Carl Zeiss Microscopy GmbH, Jena, Germany. 2.17. Total RNA-sequencing RNA from tumor tissues, from mice treated with control EVs, EVVirus and EV-Virus-PTX, were extracted using RNeasy Plus Micro Kit (Qiagen) according to the manufacturer's instructions. Indexed libraries were prepared from 10 ng/ea. purified RNA with SMARTer Stranded Total RNA-Seq Kit - Pico Input Mammalian (Clontech Laboratories, Inc.) according to the manufacturer's instructions. Libraries were quantified using the TapeStation 4200 (Agilent Technologies) and pooled such that each index-tagged sample was present in equimolar amounts, with final concentration of the pooled samples of 2 nM. The pooled samples were subject to cluster generation and sequencing using an Illumina NextSeq 500 System (Illumina) in a 2 × 150 single read format at a final concentration of 1.8 pmol. 2.18. RNA-Seq Analysis The raw sequence files generated (fastq files) underwent quality control analysis using FastQC (http://www.bioinformatics.babraham. ac.uk/projects/fastqc/).To analyze RNA-Seq data we used the strategy called “Direct mapping”as previously described [36]. Reads were first mapped on human genome (assembly hg38) using STAR [37]. The quantification of transcripts expressed for each replicate of the sequenced samples was performed using HTSeq-count [38]. R was used to create a matrix of all transcripts expressed in all samples with the corresponding read-counts and the Bioconductor package limma [39] was used to normalize the data and then to perform the differential expression analysis: an Empirical Bayes moderation t-test was performed.Data were also normalized in FPKM (Fragments Per Kilobase Of Exon Per Million Fragments Mapped) using Cuffnorm [40]. Genes upregulated (log2FC ≥1) and down-regulated (log2FC ≤−1) with a Pvalue < 0.05 were selected as differentially expressed. From these lists of DEGs, Genesis software [41,42]was used to generate heat maps and to investigate the gene ontology (GO) terms in the two gene sets of differentially expressed genes. 2.19. Statistical analysis Statistical analysis was performed by using one-way ANOVA followed by Bonferroni post-hoc test. Survival curves and their statistical analysis were performed using Kaplan–Meier test. The in vitro therapeutic synergy was calculated using fractional tumor cell viability (FTV) method [43,44]. Adjusted P-values in Supplementary Table 3 were calculated performing Fisher's exact test and the correction for multiple hypothesis testing using the Benjamini-Hochberg method [45]. All statistical analysis, calculations and tests were performed using GraphPad Prism 5 (GraphPad Software, San Diego, CA). 3. Results 3.1. Oncolytic adenoviruses can be encapsulated inside EVs with PTX For investigating the possibility to create a new type of systemic drug delivery strategy for lung cancer, we encapsulated chemotherapy drug and oncolytic virus into EVs. The size distributions of EV-Virus and EV-Virus-PTX formulations were determined by using NTA (Fig 1AB). Size distribution of both control EVs and EV-Virus formulations were detected to be in the range of 50–1000 nm. The size distribution of EVVirus overlaps with the size of the Virus (93.8 ± 4.3 nm) with most of the EVs being smalleror the same size as the virus (Fig. 1A). In addition even though heating was used to inactivate any free viruses in samples, the inactivated free virus particles are still present and this may affect the size distribution of EV-Virus, bringing it closer to the size distribution of free virus (Fig. 1A). The effect of heating on the size distribution of EVs was also assessed with A549 control EVs: while the particle count of heated EVs was within the standard error of nonboiled EVs, the size distribution shifted to slightly smaller (approximately 20 nm) after boiling, which can affect the result of EV-Virus as well (unpublished data not shown). In any case, the size distributions of EV-Virus and control EVs were very similar, with EV-Virus being slightly more oriented towards its peak at 75 nm. With EV-Virus-PTX and EV-PTX (EVs loaded with PTX), no significant size differences were seen due to the addition of PTX (Fig. 1B). Furthermore, control EVs and EV-Virus both had a similar strongly negative zeta-potential of approximately −40 mV, while the free virus had a zeta-potential of −20 mV (Fig. 1C), suggesting that EV-Virus preparation consisted mostly of EVs, otherwise it zeta-potential should have shifted towards a Fig. 1. Oncolytic adenovirus encapsulated into the Extracellular vesicle allows complex formation. (AB) Size distribution of virus alone, EV-virus, EV-paclitaxel, EV-Virus-paclitaxel and Virus were determined by using Nano tracking analysis (NTA). (C) The surface charge of the virus alone, EV-virus, EVpaclitaxel, EV-Virus-paclitaxel and Virus was measured using ZetaSizer Nano Malvern. (D) Helium Ion Microscopy pictures imaged with Orion NanoFab Helium Ion Microscope (Zeiss, Germany) using 30 kV acceleration voltage with beam current 0.2–1 pA. Images are of non-infected (left) and EV-V infected (right) cells. The deformed phenotype with multiple protrusions extending from the cell surface was often observed in infected cells at least after 24 h post-infection. M. Garofalo et al. Journal of Controlled Release 283 (2018) 223–234 227 less negative value. The zeta-potential was also unaltered in EV-VirusPTX, and EV-PTX formulations, as is to be expected since PTX is a chargeless molecule (Fig. 1C). Helium Ion Microscopy (HIM) is an imaging technique comparable to Scanning Electron Microscopy, with the distinction that samples do not require conductive coating. HIM was used to image uninfected cells as well as EV and EV-Virus infected cells in an attempt to observe potential changes due to the EV-Virus interactions on the cell surface. EV exposed cells did not appear to differ from the un-infected cells (results not shown). Yet, occasionally the surface of EV-Virus infected cells was covered with vesicle-like protrusions (Fig. 1D). This phenomenon may derive from virus-induced changes inside the cell, resulting in phenotypic alterations on the surface. The amount of PTX encapsulated into the EVs was determined by UPLC as previously described [23] (Supplementary Table 1). The washing protocol used in the production of EV-Virus-PTX formulations were successful, since the PTX concentration of the second washing step supernatant was below 0.05 μM, and thus insignificant when compared to the PTX concentration of EV-Virus-PTX (4.7 μM). The UPLC assessed concentration of the 10 μM PTX control sample, shows a 38% loss of PTX (Supplementary Table 1). However, the concentration of the 10 mM PTX DMSO stock solution was confirmed by UPLC analysis. PTX precipitation is most likely not the reason for the PTX loss seen in the control sample, since acetonitrile was added to a final concentration of 75% to samples prior to UPLC analysis. However the analysis of PTX in EV-Virus-PTX was just meant as a qualitative proof that PTX is indeed encapsulated in the vesicles. 3.2. In vitro and in vivo enhanced antitumor effect of virus and PTX in EVVirus and EV-Virus-PTX formulations The responsiveness of solid tumors to chemotherapeutic agents depends to great extent of the optimization of the drug delivery. As of such, we here set to evaluate in a factorial experiment the in vivo efficacy of several combinations of EV-Virus-PTX formulations to alter tumor growth in a lung xenograft animal model. Abraxane is a clinically approved nanoformulation [46] chosen according to the clinical settings previously described [34] and it was introduced, since previous study showed that patients with non small cell lung cancer may benefit from the treatment [47,48]. Nude mice bearing A549 cells originating tumor in the right flank were treated by intravenous (iv) injections on day 0, 2, 4 and 15 with: i) EVs alone (1 × 10 9 particles/tumor), Virus alone (1 × 10 8 vp/tumor), Abraxane (10 mg of PTX/kg); Virus + Abraxane (1 × 10 8 vp/tumor + 10 mg of PTX/kg); EV-Virus (1 × 10 8 particles/tumor + 1 × 10 8 vp/tumor); EV-Virus-PTX formulation (1 × 10 8 particles/tumor including 1 × 10 8 vp/tumor, and 10 mg of PTX/kg) (Supplementary Table 2). EVs alone were not able to control the tumor growth, and were thus used as negative control in our experiments. The intratumoral (it) treatment did not show significant differences between EV-Virus, EV-Virus-PTX and Abraxane treatments (Supplementary Fig. 1). Interestingly, the iv injection of the EV-Virus-PTX formulation significantly reduced (P< 0.001) tumor growth in comparison to naked virus and Virus + Abraxane (Fig. 2A). The highest survival rate was observed in EV-Virus and EV-Virus-PTX treatments (90% at 60 days) (Fig. 2B). The best survival rate (Fig. 2B) was observed in the combinatory group: EV-Virus-PTX over other studied formulations, suggesting that the best anti-tumor efficacy response was positively correlated with the survival. The local replication of the virus was quantified by the adenovirus E4 copy number in tumor, liver and serum by qPCR analysis. Adenoviral particles were not detected in serum and liver in any of the tested groups (Fig. 2C), suggesting that EV-Virus administered intraveniously infects and replicates only in tumor cells. In order to verify the in vitro cell death by the EV-Virus and EVVirus-PTX due to the apoptotic events, the flow cytometry measurements were carried out by measuring the amount of Annexin-V (early apoptotic stage) and propidium iodide (late apoptotic stage) for positive cells at 24 h post treatment. By that we were able to confirm that EV-Virus and Virus+PTX treatments induced both early and late in vitro apoptotic effect in A549 cells (Fig. 2 DE) (Supplementary Fig. 2).The in vitro therapeutic synergy between EVs and Virus was calculated using fractional tumor cell viability (FTV) method and demonstrated synergistic antitumor effect in the EV-Virus-PTX treatment group (Fig. 2 FG). Histopathological analysis of the liver, spleen and tumor samples from mice demonstrated no substantial changes (Fig. 3), with the exception of EV-Virus treatment, which showed moderate-sized inflammatory focus (mostly neutrophils) in otherwise normal liver lobule (Fig. 3 A4). Spleen samples displayed general histological pattern typical for nude mice such as periarteriolar lymphatic sheet areas (PALS) and in some cases mild lymphocyte hyperplasia (Fig. 3 B1–B4) [49]. Tumor samples exhibited typical features of a lung carcinoma (malignant epithelial tumor) and they were very uniform in their growth pattern and cellular features (Fig. 3 C4). 3.3. Oncolytic adenoviruses encapsulated in the EVs show increased transduction efficacy and enhanced infectious titer The transduction assay was conducted by using the red fluorescent protein [30] expressing virus Ad5D24RFP encapsulated in EVs. The transduction efficacy of the virus Ad5D24RFP alone was compared with EV-Virus and EV-Virus-PTX.Transduction was assessed at 8, 24 and 48 h post-infection. Interestingly, already at 8 h from infection, the amount of red fluorescent cells in EV-Virus and EV-Virus-PTX treated cultures was higher compared to cells treated with Virus Ad5D24RFP after 48 h (Fig. 4AB). The infectivity of the different formulations (Virus, Virus + PTX, EV-Virus and EV-Virus-PTX) were further investigated by Immunocytochemistry Assay (ICC). The infectious titer was found to be significantly higher for EV-Virus and EV-Virus-PTX formulations when compared to cells treated with virus Ad5D24RFP alone or with Virus + PTX (Fig. 4 CD). 3.4. Oncolytic adenoviruses encapsulated in the EVs show enhanced cytotoxicity To ensure that the cancer derived EVs did not affect healthy cells at least during the period of 48–96 h, in which we saw a clear anticancer effect in cancer cells, we performed the experiments also with the PNT2 cells, which is a non-cancerous cell line of prostate epithelium. It was shown by MTS cell viability assay (Fig. 5A). Then cytotoxicity of the EV-formulations was studied by MTS cell viability assays on the A549 cell line. EV-Virus and EV-Virus-PTX formulations reduced cell viability significantly more when compared to cells treated with the virus alone (Fig. 5B) (p< 0,001). Indeed, the cell killing activity of EV-Virus and EV-Virus-PTX formulations was investigated by the MTS cell viability assays on A549 and PC-3 cell lines treated with samples obtained from PC-3 and A549 cells, respectively. The EV-formulations obtained from PC-3 cells (EVs, EV-PTX, EV-Virus, EV-Virus-PTX) have been used to treat A549 cells, while EV-formulations obtained from A549 cells (EVs, EV-PTX, EV-Virus, EV-Virus-PTX) have been used to treat PC-3 cells in order to carry out the cross-experiments. In both experiments, it was observed that the cell killing effect of the EV formulations is not cancer cell line dependent (Fig 5CD) (p < 0,001). 3.5. Different molecular mechanisms underlying the anti-neoplastic effects produced by EV-Virus and EV-Virus-PTX treatments For the evaluation of the molecular mechanisms underlying the enhanced antitumor effect observed between viral and paclitaxel treatments delivered with EVs, we carried out an RNA-SEQ M. Garofalo et al. Journal of Controlled Release 283 (2018) 223–234 228 transcriptomic analysis on the RNAs extracted from tumor xenografts grown in EV, EV-Virus and EV-Virus-PTX treated mice. The analysis was carried out for 3 samples/group with the only exception of the EV-Virus group for which only 2 samples were of sufficient quality to be analyzed due to the small ratio of human versus mouse mRNA extracted from the xenograft. The analysis identified 615 and 317 transcripts for EV-VirusPTX and EV-Virus treatments, respectively, as differentially expressed in the two conditions, when compared to the EV control treatment. Most of the differentially expressed genes were up-regulated in the EVVirus, while the majority were down-regulated in the EV-Virus-PTX groups (Fig. 6A). Among the differentially expressed genes only 47 were differentially modulated by both treatments (Fig. 6B) as also detailed in the heat-maps (Fig. 6C). The two distinct genetic programs triggered by EV-Virus and EV-Virus-PTX indicated that a differential cellular response was produced by the two treatments. Indeed, although the types of biological processes involved in the response appeared remarkably similar for both treatments (Supplementary Table 3), the genes dysregulated were different. A general consideration on the net effect produced by the differential expression in terms of upand downregulation of each pathway is somehow hampered by the complexity of signals like “metabolic process”or “cellular component organization”. However, were the analysis was possible like for example for the mitogenic pathway, it was clear that both treatments triggered an anticancer effect while regulating different set of genes (Supplementary Fig. 3). In the mitogenic pathway, for example among the genes modulated by EV-Virus-PTX treatment, USP37, SNX33 and POLE were down regulated, whereas overexpression of these genes was shown to induce proliferation by promoting G1/S phase transition [50]; S-phase progression and mitosis respectively [51]. A net anti-proliferative effect was suggested also considering the genes differentially expressed by the EV-Virus treatment, which involves down regulation of BRSK2 expression, known to increase the percentage of cells in G2/M when overexpressed [52], and the upregulation of E2F4, a factor negatively influencing the G1 progression through cell cycle [53]. In conclusion, Fig. 2. In vitro and in vivo enhanced antitumor effects of oncolytic virus and paclitaxel encapsulated in Extracellular vesicles for lung cancer treatment.(A) A549 cell line was implanted subcutaneously into the right flank of BALB/c nude mice. All treatments were administered intravenously (i.v.). Tumor growth was followed over time. (B) Kaplan-Meier test was used to calculate the survival profile. (C) Adenoviral copies towards E4 gene were measured by qPCR from euthanized mice's organs (tumor, liver and serum) at the end of the treatment. (D-E) Early and late apoptotic or necrotic cell death were measured in A549 cells after 24 h post-treatment. The amount of early and late apoptotic or necrotic cells were analyzed by flow cytometry 24 h after the treatments. FITC-labeled Annexin-V was used to indicate the early apoptotic cell and PI for the necrotic or late apoptotic cells, ***P< 0.001. (F-G) The assessment of the in vitro therapeutic synergy was calculated with FTV method. Observed FTV (mean value of experimental cell viability)/(mean value of cell viability control). Expected FTV (mean FTV of Experimental condition)/(mean FTV of experimental control). A ratio > 1 indicates a synergistieffect, and a ratio < 1 indicates a less than additive effect). M. Garofalo et al. Journal of Controlled Release 283 (2018) 223–234 229 Fig. 3. Histopathological examination on liver, spleen and tumor. (A-C) Liver samples from mice treated with Virus (A1), EVs (A2), EV-Virus-PTX (A3) or EV-Virus (A4) exhibited no significant histopathological findings.A liver sample from an EVVirus -treated mouse (A4) shows a background lesion; a moderate-sized inflammatory focus (mostly neutrophils; arrow) in otherwise normal liver lobule. (B1) Spleen samples from mice treated with virus alone exhibit mild lymphocytehyperplasia (lymphatic follicles with lymphoblast-like cells, mitotic figs and tingible body macrophages), while spleen samples (B2) from mice treated with extracellular vesicles alone show no significant findings. Typical for nude mice, the PALS area (lightly-staining zone in white pulpa surrounding blood vessels) is sparse and marginal zone inconspicuous. In addition, no secondary follicles are present. (B3) A representative spleen sample from a mice treated with EV-VirusPTX displays mild hyperplasia of the marginal zone without lymphatic hyperplasia and a sample (B4) from mouse treated with EV-Virus mild hyperplasia of the marginal zone with mild lymphocyte hyperplasia. (C1) Tumor samples from mice treated with virus alone show necrotic remnants of cells in a small necrotic area and apoptotic cell remnants in degenerative area. (C2) A large cavity is filled with proteinacous fluid and lined by cubic to flattened cell. No necrosis. Tumor capsule (C3) A tumor sample treated with EV-Virus-PTX exhibits a large cavity with intraluminal blood and proteinacous fluid.Tumor tissue grows in densely packed nests or packets, and peripheral cords. No necrosis is present. (C4) Tumor sample treated with EV-Virus displays large number of neutrophils (arrowheads; examples) among foamy neoplastic cells. Fig. 4. Effect of oncolytic adenovirus encapsulated into extracellular vesicles on cell transduction and infectivity. (A) The transduction efficacy was evaluated by infection with an oncolyitc adenovirus encoding for the red fluorescent protein (RFP) encapsulated into the extracellular vesicles loaded or not with PTX.RFP was measured using Varioskan plate reader after 8, 24 and 48 h post infection, *** P< 0.001. (B) Most representative fluorescent microscope photograph (400 μm) of the infected cells. (C) The infectivity of virus alone, Virus + PTX, EVVirus and EV-Virus-PTX were assessed by ICC assay. (D) Most representative microscope photographs (400 μm) of the infected wells are presented. Figs represent difference in hexon protein expression (virus assembling), without distinguishing infectivity, replication, or gene expression manner, *** P < 0.001. (For interpretation of the references to colour in this fig legend, the reader is referred to the web version of this article.) Fig. 5. Effect of oncolytic adenoviruses encapsulated into extracellular vesicles on cell viability. (A-B) Cell viability was performed by MTS assay on PNT2 and A549 cell lines. The absorbance was measured with a 96-wells plate spectrophotometer Varioskan Flash Multimode Reader at 490 nm. (C-D) Cell viability was performed by MTS assay on A549 and PC-3, respectively treated with samples from PC-3 cells and A549 cell lines. EV-formulations from PC-3 cell line tested in A549 cell line: control EVs, EV-PTX, EV-Virus, EV-Virus-PTX. Other formulations tested in A549 cell line: Virus alone, PTX, Virus + PTX. EVformulations from A549 cell line tested in PC-3 cell line: control EVs, EV-PTX, EV-Virus, EV-Virus-PTX. Other forumulations tested in PC-3 cell line: Virus alone, PTX, Virus + PTX. M. Garofalo et al. Journal of Controlled Release 283 (2018) 223–234 230