Angiogenic Activity of Cytochalasin B-Induced Membrane Vesicles of Human Mesenchymal Stem Cells

The cytochalasin B-induced membrane vesicles (CIMVs) are suggested to be used as a vehicle for the delivery of therapeutics. However, the angiogenic activity and therapeutic potential of human mesenchymal stem/stromal cells (MSCs) derived CIMVs (CIMVs-MSCs) remains unknown. Objectives: The objectives of this study were to analyze the morphology, size distribution, molecular composition, and angiogenic properties of CIMVs-MSCs. Methods: The morphology of CIMVs-MSC was analyzed by scanning electron microscopy. The proteomic analysis, multiplex analysis, and immunostaining were used to characterize the molecular composition of the CIMVs-MSCs. The transfer of surface proteins from a donor to a recipient cell mediated by CIMVs-MSCs was demonstrated using immunostaining and confocal microscopy. The angiogenic potential of CIMVs-MSCs was evaluated using an in vivo approach of subcutaneous implantation of CIMVs-MSCs in mixture with Matrigel matrix. Results: Human CIMVs-MSCs retain parental MSCs content, such as growth factors, cytokines, and chemokines: EGF, FGF-2, Eotaxin, TGF-α, G-CSF, Flt-3L, GM-CSF, Fractalkine, IFNα2, IFN-γ, GRO, IL-10, MCP-3, IL-12p40, MDC, IL-12p70, IL-15, sCD40L, IL-17A, IL-1RA, IL-1a, IL-9, IL-1b, IL-2, IL-4, IL-5, IL-6, IL-7, IL-8, IP-10, MCP-1, MIP_1a, MIP-1b, TNF-α, TNF-β, VEGF. CIMVs-MSCs also have the expression of surface receptors similar to those in parental human MSCs (CD90+, CD29+, CD44+, CD73+). Additionally, CIMVs-MSCs could transfer membrane receptors to the surfaces of target cells in vitro. Finally, CIMVs-MSCs can induce angiogenesis in vivo after subcutaneous injection into adult rats. Conclusions: Human CIMVs-MSCs have similar content, immunophenotype, and angiogenic activity to those of the parental MSCs. Therefore, we believe that human CIMVs-MSCs could be used for cell free therapy of degenerative diseases.


Introduction
Biological activity and differentiation potential of the mesenchymal stem/stromal cells (MSCs) made them an attractive tool for regenerative medicine [1]. However, the risk of undesirable effects, including malignant transformation [2] and differentiation in unwanted directions [3,4], limits their therapeutic application.
Recent discoveries have suggested a key role of the intercellular communication in the MSCs effects on the neighboring cells. Important mediators of intercellular communication are extracellular vesicles (EVs)-membrane-enclosed nano-and microstructures released by mammalian cells [5]. It was shown that MSCs-derived EVs contain biologically active molecules, similar to those in the donor cells. In fact, several studies showed that EVs retain the therapeutic effect of the donor MSCs [6]. For example, the regenerative potential of MSCs-derived EVs was demonstrated using the rat model of myocardial infarction, where EVs restored the blood flow and reduced the size of the tissue damage [7]. In the mouse model of hindlimb ischemia EVs induced formation of new blood vessels and increased blood reperfusion [8]. MSC-EVs mediate cartilage repair by enhancing proliferation of chondrocytes, attenuating apoptosis, and modulating immune reactivity [9]. MSC-EVs have also been reported to reduce the hepatic injury by improving the hepatocytes viability, suppressing oxidative injury and modulating the inflammatory response in vivo [10]. MSCs-derived EVs were also successfully applied to treat transient global ischemia in mice, where the neuroprotective effects of MSC-EVs were demonstrated [11]. Based on these findings, the cell-free therapy concept was developed, where the MSCs-derived EVs are used instead of cells [12][13][14].
EVs represent a heterogeneous population of vesicles including microparticles (or microvesicles), exosomes, and apoptotic bodies [15]. Microvesicles are microstructures budding from the cell surface that range from 0.1 to 1 µm in diameter [15]. Exosomes are nanostructures of endogenous origin 40-100 nm in size [15]. Apoptotic bodies are large microstructures more than 1 µm in diameter, which are released from dying cells and mostly contain fragmented DNA [16].
EVs can be used as a therapeutic vehicle as they have multiple advantages: (1) Microvesicles retain the surface proteins of parental cells, (2) hydrophobic molecules (bioactive lipids and membrane proteins) could be transferred by EVs, and (3) the EVs cytoplasmic membrane protects their content from degradation [12]. However, the yield of the naturally produced EVs by MSCs is low, limiting their clinical application. One of the strategies to enhance the EVs' yield is using bioreactor cultures of bone marrow MSCs [17]. Furthermore, the osmotic shock was applied to induce the release of vesicles from CHO cells [18]. In another approach, the cells suspension extrusion through the polycarbonate filter with 1, 2, or 3 µm-pore size was applied to isolate the plasma membrane vesicles [19]. A more moderate large scale EVs production was achieved by using cytochalasin B treatment of target cells [20].
The potential of cytochalasin B-induced membrane vesicles (CIMVs) as a vector for drug delivery has been demonstrated [21]. The main question concerning the production of cytochalasin B-induced membrane vesicles was how different they were from apoptotic bodies since there was a concern that treatment of cells with a drug may induce apoptosis. Peng et al. demonstrated that cytochalasin B does not cause cell death, membrane vesicles possess surface charge similar to their parental cells, and concluded that cytochalasin B-induced membrane vesicles are not apoptotic bodies [21]. Moreover, we have previously found that neither cytochalasin B treatment of donor cells, nor treatment of recipient cells with CIMVs induce cell death [22]. CIMVs have a size similar to that of natural EVs [22]. CIMVs also retain the biological activity of parental cells and are able to stimulate capillary tube formation in vitro and vasculogenesis in vivo [22]. These properties of CIMVs led to our interest in the evaluation of the biological activity of MSCs-derived CIMVs (CIMVs-MSCs) as this might provide a potential tool for cell-free therapy.
The objectives of this study were to determine the morphology, molecular content, and angiogenic activity of human CIMVs-MSCs, as well as describe the membrane receptor transferring and biological activity of human CIMVs-MSCs.

MSCs Isolation and Characterization
Human samples were collected and methods were carried out in accordance with an experimental protocol approved by the Biomedicine Ethic Expert Committee of Kazan Federal University and Republican clinical hospital (No. 218, 11.15.2012) based on article 20 of the Federal Legislation on "Health Protection of Citizens of the Russian Federation" № 323-FL, 21 November 2011. Signed informed consent was obtained from all donors. To obtain cell suspension, the adipose tissue was cut into small pieces of about 1-3 mm 3 in size, then 10 mL of minced fat were transferred into a 50 mL tube. Next, an equal volume of 0.2% collagenase II solution (Dia-M, Russia) was added and the mixture was incubated in a shaker-incubator at 37 • C, 120 rpm for one hour. After centrifugation (400× g for 5 min), the upper fat layer was discarded, the supernatant was removed, and the remaining cell pellet was washed once in PBS (PanEco, Moscow, Russia). Then cells were re-suspended in DMEM (PanEco, Moscow, Russia) supplemented with 10% fetal bovine serum (Gibco, UK) and 2 mM L-glutamine (PanEco, Moscow, Russia). To remove the remaining tissue parts, the suspension was filtered through a cell strainer (40 µm, 93040, SPL, Korea) into a fresh tube. The cell suspension was transferred into a culture flask (ratio for solid adipose tissue was 175 cm 2 surface area/10-15 mL of adipose tissue). The culture medium was changed after 1 day of culture and the cells were maintained in a humidified environment at 37 • C, 5% CO 2 with culture medium replaced every three days.
The immune phenotype of isolated cells was analyzed by staining with monoclonal antibodies

CIMVs Production
CIMVs were prepared as described previously [22]. Briefly, MSCs of passage 4 were used in the study. After reaching a confluence of 80-90%, the MSCs were detached using trypsin/EDTA solution (2 mL/T75 flask). After 5 min incubation at 37 • C, 5% CO 2 , trypsin was inactivated by adding the culture medium. MSCs were washed twice with PBS and maintained in DMEM supplemented with 10 µg/mL of cytochalasin B (Sigma-Aldrich, St. Louis, MO, USA) for 30 min (37 • C, 5% CO 2 ). Cell suspension was vortexed vigorously for 30 sec and pelleted (100× g for 10 min). The supernatant was collected and subject to two subsequent centrifugation steps (100× g for 20 min and 2000× g for 25 min). The pellet from the last step, containing CIMVs-MSC, was washed once in PBS (2000× g for 25 min).

Scanning Electron Microscopy (SEM)
CIMVs were fixed (10% formalin for 15 min) and dehydrated using graded alcohol series and dried at 37 • C. Prior to imaging, samples were coated with gold/palladium in a Quorum T150ES sputter coater (Quorum Technologies Ltd., Lewes, United Kingdom). Slides were analyzed using Merlin field emission scanning electron microscope (CarlZeiss, Oberkochen, Germany). For the size analysis, three independent batches of CIMVs-MSCs (MSCs were obtained from three donors) were produced and used to generate at least six electron microscope images for each batch. Data collected was used to determine the CIMVs size.
Liquid chromatography mass spectrometry analysis (LC-MS/MS) of peptide extracts was done using the 3000 Nano LC nanochromatographic system (Thermo Scientific, Waltham, MA, USA) and Maxis Impact mass spectrometer with an electrospray ionization source Captive Spray (Bruker, Billerica, MA, USA). Peptides were separated by reverse phase chromatography using an Acclaim PepMap 100 NanoViper column (C18, 2 µm, 100 Å, 75 µm × 15 cm) (Thermo Scientific, Waltham, MA, USA). The tryptic peptides were eluted in a linear gradient with a mixture of solution A (4.5% acetonetrile in diH2O with 0.5% formic acid) and increasing percentage (from 5 to 35%) of solution B (94.5% acetonetrile with 0.5% formic acid). Elution was done for 60 min at 40 • C and a 300 nL/min flow rate, 1600 V of the Captive Spray source, capillary temperature 150 • C, and 3.0 L/min dry gas flow. The positive polarity and total spectrum measurements, as well as data dependent acquisition (DDA) were set on Maxis Impact mass spectrometer (Bruker, USA). Peptides mass spectrum was compared to the theoretical peptide masses of all human proteins using the SWISS-PROT and NCBI databases.

Animals
Adult rats (Rattus norvegicus) (Pushchino, Russia) were used. All experiments were carried out in compliance with the procedure protocols approved by the Kazan Federal University (KFU) local ethics committee (protocol #5, date: 27 May 2014) according to the rules adopted by KFU and Russian Federation Laws. All experiments were repeated three times. Rats were euthanized using CO 2 in compliance with the procedure protocols approved by the KFU local ethics committee (protocol #5, date: 27 May 2014). Immunocompetent animals were used in the experiment due to the fact that MSCs have low immunogenicity and immunosuppressive effect [25], as well as the short duration of the experiment (specific immunity on the introduced cells was not developed) [26].
A multiplex approach was used to characterize the molecular content of CIMVs-MSCs. Currently, there is little known about the CIMVs molecular content, while it could be suggested that it should retain, in part, the MSCs cytoplasm components, including intracellular cytokines. Due to these reasons, we sought to analyze the cytokine content of CIMVs as well as the parental MSCs.  Table 1). Levels of IL-3 and IL-13 were below the detection range in MSCs and CIMVs-MSCs. Interestingly, levels of TGF-α, CCL7, sCD40L, IL-1b, and TNF-β were higher in MSCs as compared to CIMVs-MSCs.

Transfer of Cell Surface Receptors to the Recipient Cell Membrane by CIMVs-MSCs
Microvesicles can transfer soluble factors as well as surface receptors by the fusion of cytoplasmic membranes [29,30]. Therefore, we sought to determine whether CIMVs-MSCs could transfer the surface receptors to the recipient HEK293FT cells. HEK293FT cells were pre-stained with DiO (Invitrogen, USA) and cultured for 24 h with DiD labeled CIMVs-MSCs (10 µg/mL) (Invitrogen, USA). Expression of CD90 was selected to demonstrate receptor transfer, as it is specific for CIMVs-MSCs and absent on HEK293FT cells. Expression of CD90 was analyzed using the laser scanning confocal microscope Zeiss LSM 780 (Carl Zeiss, Oberkochen, Germany) and flow cytometry BD FACS Aria III (BD Bioscience, San Jose, CA, USA). Donor MSCs demonstrated homogenous staining with DiD membrane dye (red fluorescence) and anti-CD90 antibody (blue fluorescence) ( Figure 5A-C). We found that CIMVs-MSCs and HEK293FT membranes became fused and CD90 surface receptor was transferred to HEK293FT ( Figure 5D-G). Regions of DiD and partial staining with anti-CD90 antibody staining were found in the cytoplasmic membrane of HEK293FT treated with CIMVs-MSCs ( Figure 5D-G). Recipient cells (HEK293FT) treated with CIMVs acquired CD90 positive phenotype due to the internalization of CIMVs-MSCs membrane into the cytoplasmic membrane of the recipient cells. We determined that 99.14% of HEK293FT recipient cells acquired CD90+ immunophenotype ( Figure 5H,I).

CIMVs-MSCs Stimulated Angiogenesis In Vivo
Since the CIMVs-MSCs contain multiple growth factors (EGF, FGF-2, and VEGF), we postulated that CIMVs-MSCs could have angiogenic activity. We used an in vivo approach to demonstrate angiogenetic capacity of MSCs and CIMVs-MSCs. MSCs (1 × 10 6 cells) and CIMVs-MSCs (50 µg) were stained with vital membrane dye DiO (Invitrogen, USA), mixed with Matrigel matrix (400 µL), and injected into rats subcutaneously. Eight days later, MSCs and CIMVs-MSCs containing Matrigel matrix plugs were collected from the subcutaneous space of the rats and fixed in 10% formalin (group 1) or frozen in liquid nitrogen (group 2). Formalin-fixed Matrigel matrix plugs were stained with the hematoxylin-eosin kit (BioVitrum, Saint-Petersburg, Russia) ( Figure 6A-C). Frozen Matrigel matrix plugs were cut using HM560 Cryo-Star microtome (ThermoScientific, Waltham, MA, USA) and stained with DAPI (D1306, Invitrogen, USA) ( Figure 6D-F). MSCs and CIMVs-MSCs were detected in the Matrigel matrix implants eight days after subcutaneous injection ( Figure 6D-F). Moreover, newly developed blood capillaries were observed in Matrigel matrix containing MSCs and CIMVs-MSCs ( Figure 6A-C). We found that the number of the newly developed blood vessels in control Matrigel matrix (without MSCs or CIMVs) was 0.67 ± 0.15 cap/mm 2 (Figure 6A,D,G). In Matrigel matrix containing MSCs, the number of newly developed blood vessels was 11.3-fold higher (7.55 ± 0.46 cap/mm 2 , p < 0.01) than that in control ( Figure 6B,E,G). Similar to MSCs, the number of the new capillaries in Matrigel matrix containing CIMVs-MSCs was increased (5.7-fold higher (3.84 ± 0.16 cap/mm 2 , p < 0.01)) than that in control ( Figure 6C,F,G). We suggest that the human CIMVs-MSCs retain the angiogenic activity of the parental MSCs. Counting of the total number of vessels was carried out using the AxioVision 4.8 program (CarlZeiss, Oberkochen, Germany). (G) Quantitation of the capillary density in Matrigel matrix plugs. The data represents mean ± SD. For statistical analysis, ten hematoxylin and eosin stained slides per animal were analyzed.

Discussion
CIMVs could be produced in large quantities while retaining the size of natural EVs [20,22]. However, our understanding of the biologic activity and therapeutic potential of MSCs-derived CIMVs remains limited.
Here, for the first time, we have shown that the size of the majority of human CIMVs-MSCs ranges between 100 and 1200 nm (89.36%), which is similar to that of EVs. Furthermore, the proteome content of human MSCs and CIMVs-MSCs appears to be similar. Analysis of the CIMVs-MSCs content revealed an increase of proteins linked to cytoskeleton, peroxisomes, cell membrane, and cytoplasm. In contrast, mitochondria and cytoplasm/nucleus proteins were decreased, while nucleus and secreted proteins were significantly depleted as compared to MSCs. We believe that the cytoskeleton proteins and membrane proteins enrichment of CIMVs-MSCs is due to the mechanism of their outward release from the cell surface. Similar data was demonstrated by Kim and colleagues [31]. We suggest that the enrichment of peroxisomes/cytoplasm proteins and depletion of mitochondria/cytoplasm/nucleus proteins could be due to the deep intracellular localization of these organelles.
We have demonstrated that CIMVs-MSCs transfer membrane receptors to the target cells. It is known that MSCs surface cell adhesion molecules and signaling receptors play an important role in MSCs biology and maintain the stem like phenotype [35]. The surface receptor transfer by EVs could be the mechanism of mimicry and reprogramming of target cells. Similar data was published by Ratajczak and colleagues, where stem cell-derived microvesicles reprogram target cells by delivering their content including mRNA [36].
MSCs and CIMVs-MSC contained growth factors, cytokines, and chemokines, suggesting similar biological activity. To support this assumption, we have demonstrated that human MSCs and CIMVs-MSCs share the angiogenic activity. We have found that human MSCs and CIMVs-MSCs stimulate the sprouting of new blood vessels in vivo. Our data corroborate results published by Gangadaran et al. where MSCs-derived EVs increased cellular migration, proliferation, endothelial tube formation in vitro, and enhanced angiogenesis in ischemic limb in vivo [8]. In addition, Lopatina et al. reported that MSCs-derived EVs could induce the formation of vessel-like structures in vitro and in vivo after the subcutaneous injection in mixture with Matrigel Matrix and human microvascular endothelial cells [37]. We believe that the angiogenic capacity of CIMVs-MSCs depends on growth factors present in their content. Human CIMVs-MSCs demonstrated an angiogenic effect in vivo, although it was lower than that of MSCs parental cells. Due to the risks of MSCs therapy connected with undesirable differentiation [3,4] and transformation [4], the therapeutic use of cell-free therapeutic instrument based on CIMVs is only mechanistically feasible [12]. On the other hand, CIMVs-MSCs could be used to stimulate angiogenesis as they have molecular content and angiogenic activity similar to the parent MSCs. Therefore, CIMVs-MSCs could be used as a method for cell-free regenerative medicine.

Conclusions
We analyzed the molecular content, receptors expression, and angiogenic potential of human MSCs and CIMVs-MSCs. Human CIMVs-MSCs have similar content, immunophenotype, and angiogenic activity to that of the parental MSCs. CIMVs-MSCs could transfer membrane receptors to the surface of target cells. Therefore, we believe that human CIMVs-MSCs could be developed for cell-free therapy of degenerative diseases.