Engineering hybrid exosomes by membrane fusion with liposomes

Exosomes are a valuable biomaterial for the development of novel nanocarriers as functionally advanced drug delivery systems. To control and modify the performance of exosomal nanocarriers, we developed hybrid exosomes by fusing their membranes with liposomes using the freeze–thaw method. Exosomes embedded with a specific membrane protein isolated from genetically modified cells were fused with various liposomes, confirming that membrane engineering methods can be combined with genetic modification techniques. Cellular uptake studies performed using the hybrid exosomes revealed that the interactions between the developed exosomes and cells could be modified by changing the lipid composition or the properties of the exogenous lipids. These results suggest that the membrane-engineering approach reported here offers a new strategy for developing rationally designed exosomes as hybrid nanocarriers for use in advanced drug delivery systems.

synthetic liposomes and exosomes post secretion from parent cells. This method allows us to optimize the properties of the exosome surface in order to decrease its immunogenicity and increase its colloidal stability, and improve the half-life of exosomes in blood. However, the hydrophobic properties of lipid molecules hinder direct loading of lipids into exosomes. Genetic modification of the exosomes' lipid membrane is also difficult because lipid biosynthesis involves a large number of proteins, and the process by which lipids are sorted from the host cell to the exosome is not clearly understood. In this context, we proposed a new approach to prepare engineered hybrid exosomes by fusing the membranes of exosomes and liposomes using the freeze-thaw method (Fig. 1). The properties of the exosome surface can be easily modified using liposomes embedded with peptides or antibodies as targeting moieties or poly(ethylene glycol) (PEG) 20,21 . To provide a proof-of-concept for engineering exosome using this membrane fusion approach, we prepared exosomes in cells expressing the tyrosine kinase receptor HER2 22 , and these HER2-containing exosomes were fused with phospholipid liposomes. In addition to evaluating the membrane fusion of exosomes with liposomes, we performed flow cytometry and fluorescence microscopy to examine the performance of the obtained exosomes as nanocarriers.

Results
Interaction of Raw 264.7 cell-derived exosomes and liposomes. Exosomes were isolated from the supernatants of Raw 264.7 cell cultures by differential centrifugation and micro-filtration. Western blot analysis of Raw 264.7 cell-derived exosomes revealed high expression of exosome markers, especially heat-shock proteins (e.g. Hsc70) and tetraspanin (CD9) (Fig. 2a). Nanoparticle tracking analysis (NTA) revealed that the hydrodynamic diameter of the exosomes was 150 ± 56 nm (mean ± standard deviation) (Fig. 2b).
The efficiency of exosome and liposome fusion was quantitatively evaluated in terms of the lipid mixing ratio measured using a fluorescence resonance energy transfer (FRET) assay using a set of NBD and rhodamine-labelled lipid, which is widely used for the analysis of liposome-liposome, liposome-cell, and viruscell fusion [23][24][25][26] . For example, when NBD-DMPE is excited at a wavelength of 460 nm, the resulting solutions emit fluorescence at 530 and 588 nm, reflecting the emission from NBD-DMPE and rho-DMPE, respectively. After subjecting the exosome-liposome mixture to one or more freeze-thaw cycles, the fluorescence intensity at 530 nm increased and the intensity at 588 nm decreased as the number of cycles was increased (Fig. 3a). These results demonstrate that the fluorescent lipids in the liposome were diluted by lipid molecules or other membrane components (e.g., membrane proteins) present in the exosome. The lipid dilution ratio as fusion efficiency was evaluated using a calibration set of fluorescence spectra. Figure 3b shows the relationship between the efficiency and the number of freeze-thaw cycles.
When unlabelled DOPC liposomes were interacted with fluorescently labelled liposomes, the lipid dilution ratio was almost 2.1 ± 0.1 after 10 freeze-thaw cycles, whereas, the efficiency for the exosome-liposome system was 3.3 ± 0.2, which was higher than that of the liposome-liposome system. As for the mechanism to explain the difference of the fusion efficiencies, interaction of liposomes with the protein existed in exosomal membrane or glycosylation is considered to be a plausible factor. The efficiency of each liposomal system is presented in Table 1, which shows relatively high fusion efficiencies for all of the liposomal systems used in this study, irrespective of the type of lipid. Although the diameter changes of exosomes after fusion with liposomes were evaluated by nanoparticle tracking analysis, it was difficult to detect statistically the differences of the averaged sizes and the size distribution after the freeze thawing (Supplementary Table S1 for averaged size and Figure S1 for size distribution). The morphological change of exosome after the freeze thawing was also examined by negatively stained transmission electron microscopy (TEM). Though TEM study could not represent averaged morphology, there were not any apparent morphological difference between the exosomes before and after membrane fusion by the freeze thawing treatment (Supplementary Figure S2).

Isolation of exosomes from HER2-expressing CMS7 cells.
To further examine the applicability of the membrane fusion technique used here, we investigated the fusion behaviour between exosomes bearing a specific membrane protein and the liposomes. Exosomes were isolated from the culture supernatants of wild-type CMS7 cells (CMS7-wt cells) and genetically modified CMS7 cells overexpressing the human HER2 receptor (CMS7-HE cells) 27,28 . The surface protein CD63, a representative exosome marker, was detected in exosomes derived from both cell types by fluorescence-activated cell sorting analysis using an exosome bead-coupling technique (Fig. 4a,b). The hydrodynamic diameters of exosomes derived from CMS7-wt and CMS7-HE, as measured by NTA, were similar (Fig. 4c,d).
The loading of HER2 protein on the exosome membrane was evaluated by western blot analysis using anti-Hsc70 as an exosome marker and anti-HER2. As expected, HER2 was only detected in CMS7-HE-derived exosomes as a predicted band (Fig. 4f), whereas Hsc70 was detected in exosomes from both cell types (Fig. 4e). Furthermore, HER2 activity was determined using a corresponding anti-phospho antibody (Fig. 4g), which confirmed that HER2 was activated by auto-phosphorylation of tyrosine residues. The presence of phosphorylated Scientific RepoRts | 6:21933 | DOI: 10.1038/srep21933 HER2 indicates that the exosome carried active HER2 protein. The protein concentration of HER2 on exosomes was also evaluated using a standard curve prepared using samples containing known concentrations of HER2. These studies revealed that the obtained exosomes contained about 2-3.3 ng of HER2 per 1 μ g of total protein. Interaction of genetically modified exosomes embedded with HER2 protein and liposomes. Table 2 shows the fusion efficiency for each lipid system. As expected, significant membrane fusion was observed for all of the lipid systems evaluated here. These efficiencies of CMS7-HE were relatively higher than that of Raw 264.7 exosomes system. We do not have clear answer for the difference of fusion efficiency at this moment. Membrane component of exosome (e.g. lipid composition, membrane protein content) were significantly different depending on its origin so that the difference would alter the interaction between liposome and exosome related to fusion efficiency.
Next, we performed western blotting to evaluate the properties of exosomes embedded with HER2 after the freeze-thaw cycles with liposomes. HER2 and phosphorylated HER2 were detected in the exosome-liposome mixtures ( Fig. 5a,b), as in untreated exosomes (Fig. 4f,g). These results indicate that genetic modification of exosomes and fusion-based membrane modification techniques can be combined, and that exosome-liposome hybrids carrying specific proteins can be obtained by freeze-thaw methods. Cellular uptake of membrane-engineered exosomes. Exosomes have valuable delivery functions, which can be exploited in the development of novel nanocarriers for advanced drug delivery. In general, exosomes deliver their molecular cargo to the cytosol via endocytosis or membrane fusion 4,5,6,29 . Therefore, we determined the cellular uptake efficiency of the membrane-engineered exosomes.
HeLa cells were incubated with 5-or 6-(N-succinimidyloxycarbonyl)-fluorescein 3′ ,6′ -diacetate (CFSE)-labelled exosomes or exosome-liposome hybrids for 4 h and cellular uptake was quantified by FACS analysis and observed by confocal laser scanning microscopy. Prior to the cellular uptake experiment, it was confirmed that the labeling efficiency did not significantly change in the various systems. In Fig. 6, the cellular uptake profile was similar between DOPC-exosome hybrids and DOPS-exosome hybrids, which indicates that neutral and anionic lipids did not affect the cellular uptake of these exosomes. However, incorporation of a cationic lipid (i.e., DOTAP) into the hybrid exosome decreased the cellular uptake efficiency. On the other hand, the cellular uptake of the hybrid exosomes with PEG-DOPS was significantly increased, by almost two-fold, compared with the unmodified exosomes. To rule out the possibility that the positive effect of PEG-DOPS is attributed to a fraction of the unfused PEG-DOPS liposomes and exosomes but not the hybrid exosomes cellular uptakes with or without freeze-thaw cycles were compared by using FACS under the same condition. The uptake of the hybrid exosomes with PEG-DOPS treated by freeze-thaw process was significantly increased more than that of the mixture of exosome and DOPS without freeze-thaw, showing the increased cellular uptake was attributed to the hybridization of the exosomes (Supplementary Figure S3). These results indicate that the interactions between exosomes and cells can be modified by introducing exogenous lipids to hybrid exosomes.

Discussion
In this study, we developed engineered hybrid exosomes by fusing the membranes of exosomes with liposomes using the freeze-thaw method. The freeze-thaw method 30-32 is a relatively simple physicochemical process that was selected to prepare the hybrid exosomes by membrane fusion. The freeze-thaw method has been used to convert multi-lamellar liposomes to large uni-lamellar liposomes or incorporate water-soluble molecules into the internal water phase of liposomes, which suggests that this method disrupts plasma membranes by temporary formation of ice crystals [33][34][35] . Accordingly removing the water molecules from the hydrophilic surface of the lipid bilayer membrane by freezing appears to cause significant structural and functional changes, including  hydration-dependent phase changes, lateral phase separation of membrane components, and membrane fusion; the latter may occur as a result of membrane reconstitution 36,37 . These findings prompted us to use the freezethaw method to induce fusion or mixing of exosome and liposome membranes. In addition to being simple, this physical technique avoids contaminating the membranes with unwanted chemicals such as calcium or PEG, for example, which are used in other chemical fusion methods. Exosomes embedded with a specific membrane protein (in this case HER2) were also fused with various liposomes at relatively high efficiencies, confirming that membrane engineering can be combined with genetic modification of exosomes.
Cellular uptake studies of the hybrid exosomes confirmed that the delivery function of the exosomes could be modified by changing the lipid composition or the properties of the exosome by membrane fusion. Although it is unclear why the cellular uptake of cationic DOTAP-exosome hybrids was reduced, it is possible that cationic lipids might influence the proteins involved in the interactions between the exosomes and cells for cellular uptake. On the other hand, the cellular uptake of the PEG-modified exosomes was significantly increased. Generally, the presence of PEG on the liposome surface increases their circulation time in blood while reducing their uptake by mononuclear phagocytes (namely, stealth liposomes) 38 . These findings indicate that PEG reduces interactions between the liposomes and cellular membranes for effective cellular uptake of the PEG-modified exosomes. One of the possible explanations for the unexpectedly high uptake efficiency is that PEG existed on exosomal surface reduced electrostatic repulsions of anionic exosome surface with anionic cell surface to increase the exosome-cell interaction for increased cellular uptake. Another possibility is PEG-induced membrane fusion 39 which is caused by small perturbations in lipid packing within the bilayer leaflets contacted by the liposomes.
Recently, it was reported that exosomes transport not only water-soluble cargoes like proteins and RNA, but also lipophilic molecules 13 . Therefore, exosomes can carry bioactive lipids such as lipid signals or mediators between proximal or distal cells, and their fate is at least partly regulated by these lipids. Furthermore, exogenous membrane proteins can be incorporated into exosomes using this engineering method. Recently, we reported a new one-step method for preparing proteoliposome by cell-free membrane protein synthesis in the presence of liposomes. For example, liposomes containing connexin were developed using this method, and were capable of delivering a hydrophilic oligo-peptide to cells via connexin-mediated communication through gap junctions between the proteoliposome and cells 40 . Fusion of the proteoliposome and exosome membranes facilitated the incorporation of the membrane protein into the exosomes. The exosome-liposome fusion method should also be useful for loading therapeutic agents into exosomes, a field that is now being researched by our group.
In conclusion, the present results suggest that the membrane-engineering method used here represents a new strategy to yield hybrid exosomes as novel biological nanotransporters (bio-nanotransporter). These engineered hybrid exosomes can be used to transport exogenous hydrophobic lipids incorporated in vitro by the membrane fusion method to recipient cells as well as hydrophilic cargoes within the exosomes as advanced drug delivery systems.  bovine serum (FBS) and 1% penicillin-streptomycin (PS). HeLa cells were cultured in MEM medium with 1 % Non-essential amino acids solution and 10 % FBS and 1% PS. All cell lines were cultured in a 5% CO 2 incubator at 37 °C.

Materials
Exosome isolation. Exosomes were isolated by differential centrifugation and micro-filtration as previously described 10,41,42 . Briefly, the cells (Raw 264.7, CMS7-wt, and CMS7-HE) were incubated for 15 h in exosome-free FBS, which had been centrifuged at 100,000 × g before use. The culture supernatants obtained from each cell culture were collected, centrifuged at 300 × g for 10 min to remove cells, at 1,200 × g for 20 min, and finally at 12,000 × g for 30 min (all at 4 °C), followed by filtration through a 0.22-μ m filter. Exosomes were pelleted by ultracentrifugation at 120,000 × g for 70 min at 4 °C. The pellets were rinsed with phosphate buffered saline (PBS), ultracentrifuged again, and dispersed in PBS. The protein concentrations of the exosome suspensions were determined using BCA assays (Thermo Scientific, Waltham, MA, USA) and were adjusted to 300 μ g/mL before use.
Nanoparticle tracking analysis. The size distribution of the exosomes and liposomes were analysed using a Nano Sight LM10 instrument Marvern instruments Ltd, Worcestershire, UK) equipped with NTA software version 2.3. The particle suspensions were diluted with PBS to a concentration of 1-8 × 10 8 particles/mL for analysis.
Liposome preparation. Liposomes were prepared using an established protocol 43  Membrane fusion. The exosome and liposome membranes were fused using an established freeze-thaw method [30][31][32] . Exosomes were mixed with 100 μ M liposomes labelled with 1 mol% NBD-DMPE and rho-DMPE (1:1 by volume). The mixtures were frozen in liquid nitrogen and thawed at room temperature for 15 min. The freeze-thaw cycle was repeated for several times. The fusion efficiency was evaluated using established FRET assays [30][31][32] . Briefly, fusion was monitored using a fluorescence couple of NBD-DMPE and rho-DMPE within the lipid bilayer membrane. The fluorescence of the mixture was measured using a FP8000 fluorescence spectrometer (JASCO, Tokyo, Japan). Excitation of NBD-DMPE at 460 nm induces fluorescence emission at 530 and 588 nm, corresponding to the emissions from NBD-DMPE and rho-DMPE, respectively. Dilution of rho-DMPE owing to membrane fusion with NBD-DMPE increases the fluorescence intensity at 530 nm and decreases the intensity at 588 nm. The fusion efficiency was evaluated using calibration samples with known fluorescence 23,26 , which allowed us to assess membrane fusion following the freeze-thaw cycles based on the lipid composition of the liposomes.  Table 2. Fusion efficiencies between of the CMS7-HE cell-derived exosomes and liposome after freezethaw cycles.  NBD-DMPE and rho-DMPE, and were plotted against the lipid dilution ratio. The lipid dilution ratio of the labelled liposome was then calculated from the standard curve.
Flow cytometry. Exosome bead coupling and flow cytometry. Exosomes (6 μ g, protein) were incubated with 2 μ L of aldehyde/sulfate latex beads (4 μ m diameter) for 15 min at room temperature. The volume was made up to 1 mL using MEMS buffer and the mixture was incubated overnight on a rotary shaker at 4 °C. To block remaining binding sites, the exosome-coated beads were incubated for 30 min with 92 mM glycine at room temperature. After three washes in stain buffer, the presence of CD63 protein on the membrane surface of exosome bound to the beads was examined by single-color immunofluorescence labelling with PE-conjugated anti-CD63 monoclonal antibody. As isotype controls, exosome-coated beads were stained with PE-conjugated rat IgG2a after incubation for 1 h at room temperature. All data were collected on a LSR Fortessa cell analyser (BD Bioscience), and analysed with FlowJo software (Treestar, Inc., San Carlos, CA, USA).
Cellular uptake. Purified Raw 264.7-derived exosomes (300 μ g/mL, protein) were fused with 100 μ M of liposomes (1:1 by volume) using the freeze-thaw method described above and labelled with 62.8 μ M CFSE for 30 min at 37 °C, and the volume was made up to 100 μ L with PBS buffer. The labelled exosome suspension was transferred to a PD Spin Trap G-25 to remove free CFSE. The resulting Raw 264.7 exosome-liposome hybrids or Raw 264.7 exosomes (4.5 μ g protein in exosome) were incubated with 1 × 10 5 HeLa cells for 4 h at 37 °C. The cells were then washed twice with PBS, detached using trypsin, and resuspended in stain buffer. The fluorescence intensity was measured on the LSR Fortessa cell analyser.
Confocal laser scanning microscopy. The cellular uptake of exosomes into HeLa cells was visualized by confocal laser scanning microscope equipped with plan-APOCHROMAT 40× oil immersion objective (LSM780, Carl Zeiss, German) with excitation by Argon laser at 488 nm. CFSE-labelled Raw 264.7 exosome-liposome hybrids or CFSE-labelled Raw 264.7 exosomes (3 μ g protein in exosomes) were incubated with 1 × 10 5 HeLa cells for 4 h at 37 °C under 5% CO 2 . The cells were washed twice with PBS, the medium was replaced with fresh medium, and the cells were cultured for 4 h at 37 °C under 5% CO 2 .