METHOD FOR PREPARING LIPOSOMES

Abstract

The invention relates to liposomes, methods of producing liposomes, and methods of loading cell-derived liposomes with cargo molecules. The invention extends to such liposomes per se, and to the use of these liposomes as cellular delivery systems for the delivery of biologically and therapeutically active payload molecules, such as small molecules, RNAi molecules (e.g. siRNA), bioactive proteins, genome editing tools (e.g. Cas9) and drugs into cells for treating a range of disorders. The liposomes may also be used in a range of diagnostic and theranostic applications. The invention extends to pharmaceutical compositions comprising such liposomes, including populations of extracellular vesicles (EV), exosomes and to fusion proteins.

Claims

1. A method of preparing a liposome, the method comprising contacting at least one cell with: (i) a pore-forming protein, or a pore-forming domain or a variant or fragment thereof; and (ii) a shuttle protein, optionally attached to a bioactive payload molecule, wherein the pore-forming protein, or the pore-forming domain or the variant or fragment thereof creates a pore through a phospholipid bilayer of the at least one cell, and the shuttle protein interacts with the pore-forming protein, or the pore-forming domain or the variant or fragment thereof, and is internalised into the cell to thereby produce a liposome, optionally loaded with a bioactive payload molecule.

2. A method according to claim 1, wherein the method further comprises isolating the liposome from the cell.

3. A method according to any preceding claim, wherein the liposome comprises a vesicle.

4. A method according to any preceding claim, wherein the liposome comprises an extracellular vesicle (EV), an intracellular vesicle or an intraluminal vesicle (ILV).

5. A method according to any preceding claim, wherein the liposome comprises an exosome.

6. A method according to any preceding claim, wherein the liposome has an average diameter of between 10 nm and 500 nm, or between 20 nm and 400 nm, or between 30 nm and 300 nm, or between 40 nm and 200 nm, or between 50 nm and 150 nm, or between 60 nm and 120 nm.

7. A method according to any preceding claim, wherein the cell comprises a biological cell, optionally a mammalian or human cell.

8. A method according to any preceding claim, wherein the pore-forming protein, or the pore-forming domain or the variant or fragment thereof comprises, or is derived from, a non-toxic protein.

9. A method according to any preceding claim, wherein the pore-forming protein, or the pore-forming domain or the variant or fragment thereof is derived from B. anthracis or Ricin.

10. A method according to any preceding claim, wherein the pore-forming protein, or the pore-forming domain or the variant or fragment thereof is derived from B. anthracis virulence factor Protective Antigen (PA).

11. A method according to any preceding claim, wherein the pore-forming protein, or the pore-forming domain or the variant or fragment thereof is derived from B. anthracis 83 or 63.

12. A method according to any preceding claim, wherein the pore-forming protein, or the pore-forming domain or the variant or fragment thereof, comprises or consists of an amino acid sequence substantially as set out in any one of SEQ ID No: 1, 2, 5-10, or a variant or fragment thereof.

13. A method according to any preceding claim, wherein the shuttle protein is configured to facilitate transport, preferably of a bioactive payload molecule, through the pore through the pore of the preforming protein.

14. A method according to any preceding claim, wherein the shuttle protein comprises an attenuated toxin protein.

15. A method according to any preceding claim, wherein the shuttle protein is B. anthracis derived lethal factor (LF) or oedema factor (EF).

16. A method according to any preceding claim, wherein the shuttle protein comprises or consists of an amino acid sequence substantially as set out in SEQ ID No: 11, or a variant or fragment thereof.

17. A method according to any preceding claim, wherein the shuttle protein comprises a linker protein.

18. A method according to claim 17, wherein, in the attenuated toxin, at least one toxin domain, optionally one or more of toxic domains II-IV of the B. anthracis lethal factor protein toxin, is replaced by the linker protein.

19. A method according to claim 17, wherein the linker protein comprises a nucleic-acid-binding domain.

20. A method according to claim 19, wherein the nucleic-acid-binding domain is Saccharomyces cerevisiae GAL4.

21. A method according to any preceding claim, wherein the shuttle protein comprises or consists of an amino acid sequence substantially as set out in SEQ ID No: 12 or 13, or a variant or fragment thereof.

22. A method according to any preceding claim, wherein the bioactive payload molecule is a therapeutically active molecule which is active within the cell cytosol, within the nucleus, within an organelle or intracellular structure such as a vesicle or a vacuole, within a cell surface lipid membrane or an intracellular lipid membrane.

23. A method according to any preceding claim, wherein the molecular weight of the bioactive compound is between 1 Da and 10 MDa.

24. A method according to any preceding claim, wherein the bioactive molecule is: (i) a small molecule, a protein, a nucleotide, DNA or a DNA construct, plasmid, RNA or an RNA construct, mRNA, miRNA, a guide RNA, snRNA, siRNA, antisense oligonucleotide (ASO), or (ii) a large molecule, such as a protein or enzyme or a fragment thereof, a nuclease, or an antibody or antigen-binding fragment thereof.

25. A method according to any preceding claim, wherein the bioactive payload molecule comprises a genome editing tool, optionally a nuclease, preferably Cas9, Cpf1, a TALEN, or a zinc finger nuclease.

26. A method according to any preceding claim, wherein the bioactive molecule comprises or consists of an amino acid sequence substantially as set out in SEQ ID No: 14, or a variant or fragment thereof.

27. A liposome obtained, or obtainable by, the method of any one of claims 1-26.

28. A liposome comprising a phospholipid bilayer surrounding a lumen, a pore-forming protein, or a pore-forming domain or a variant or a fragment thereof, and a shuttle protein.

29. A liposome according to claim 28, wherein the liposome is as defined in any one of claims 1-26.

30. The liposome according to any one of claims 27-29, for use in therapy or diagnosis.

31. The liposome according to any one of claims 27-29, for use in treating, preventing, or ameliorating a disease.

32. The liposome, for use according to claim 31, wherein the disease to be treated is obesity, more preferably FMO5-regulated obesity.

33. The liposome, for use according to claim 31, wherein the disease to be treated is a prostaglandin-D2-regulated disease selected from a group consisting of: androgenetic alopecia (AGA); acne; rosacea; and prostate cancer.

34. The liposome, for use according to claim 31, wherein the liposome is used for treating, preventing, or ameliorating: Zika fever (or Zika virus disease), Ebola virus disease, Acquired immunodeficiency syndrome (human immunodeficiency virus), Stat3-responsive cancer, P53-deficient cancer, virally-mediated cervical cancer (i.e. human papilloma virus), familial hypercholesterolemia, Duchene muscular dystrophy, spinal muscular atrophy, Crohn's disease, or an inflammatory disease especially one of the bowel implicated in the overexpression of intracellular adhesion molecule-1 (ICAM-1).

35. The liposome, for use according to any one of claims 30-34, wherein the cell used to create the liposome is obtained from a subject being treated.

36. The liposome, for use according to claim 35, wherein healthy cells are obtained from or from a stem cell line, or from the target tissue of the subject and expanded in culture prior to being used for the production of the liposome that is loaded with a therapeutic compound appropriate to treating the clinical condition in question.

37. The liposome, for use according to claim 35, wherein the cells obtained from the subject do not comprise healthy cells, optionally from a patient biopsy.

38. A kit comprising the liposome according to any one of claims 27-29, and instructions for use.

39. The liposome according to any one of claims 27-29, for use in a genome editing technique.

40. A genome editing method comprising loading a liposome according to any one of claims 27-29 with (i) a guide RNA; and/or (ii) a nuclease or genetic construct encoding a nuclease, and using the loaded liposomes in a gene editing therapy.

41. A method according to claim 40, wherein the nuclease comprises Cas9 or Cpf1 or a TALEN or a zinc finger nuclease.

42. A shuttle protein comprising an attenuated toxin protein attached to Protein Kinase R (PKR).

43. A shuttle protein according to claim 42, wherein the protein comprises or consists of an amino acid sequence substantially as set out in SEQ ID No: 11, or a fragment or variant thereof.

Description

[0122] For a better understanding of the invention, and to show how embodiments of the same may be carried into effect, reference will now be made, by way of example, to the accompanying Figure, in which:

[0123] FIG. 1 is an image from an optical microscope (LSM880 via an Airyscan detector (Carl Zeiss Ltd) showing the localisation of Texas red-labelled cargo (i.e. Texas Red labelled LFn-PKR) to intraluminal structures within HeLa cells stained with α-CD63 (3 h chase). The panels show Texas Red labelled LFn-PKR (centre); α-CD63 (right); and the merged image (left);

[0124] FIG. 2 is an image of liposomes/exosomes isolated form HeLa cells stained with (Cy5) cell mask;

[0125] FIG. 3 is an image from an optical microscope showing exosome preparations from HeLa cells exposed to PA83 and Texas Red-labelled LFn-SaCAS9 using two different magnifications;

[0126] FIG. 4 is an image showing the results of exosome preparations from Hela cells exposed to PA83 and Texas Red-labelled LFn-PKR (Panel A) or PA83 and Texas Red-labelled BSA (Panel B, control) after 3 h;

[0127] FIG. 5 is the result of TCA precipitating liposomes/exosomes prepared with PA83 and either: Texas Red-labelled LFn-SaCAS9, Texas Red-labelled LFn-PKR, or Texas Red-labelled BSA;

[0128] FIG. 6 is the biological activity of β-galactosidase in cells treated with control liposomes/exosomes, and cells treated with liposomes/exosomes loaded with siRNA targeting β-galactosidase translation;

[0129] FIG. 7 is a schematic diagram showing the cellular production of liposomes (i.e. exosomes) of the invention;

[0130] FIG. 8 shows the activity of exosomes loaded with LFn-diphtheria toxin A chain (DTA) after incubation for 60 min at 37° C. with excess trypsin (n=3±SEM) prior to incubation with HeLa cells. Cell viability was approximately 55% of an untreated control in both instances;

[0131] FIG. 9 shows the photon correlation spectroscopic analysis of HeLA exosome and extracellular vesicle fractions loaded with wild type PA83, LFnPKR and anti-TdTom antisense oligonucleotides (ASOs);

[0132] FIG. 10 shows the photon correlation spectroscopic analysis of HeLA exosome and extracellular vesicle fractions loaded with forced octamer PA83 mutants, LFnPKR and anti-TdTom antisense oligonucleotides (ASOs);

[0133] FIG. 11 shows normalized β-galactosidase expression of HEK293 cells 48 hours after exosome treatment; and

[0134] FIG. 12 shows β-galactosidase activity after transfection of HEK293 cells with HeLa exosomes loaded with α-GFP siRNA and isolated using the ExoEasy Kit. Feeder cells treated with PA:LF nPKR::50 nM si RNA overtime.

EXAMPLES

[0135] The inventors have developed a novel method of producing liposomes (e.g. exosomes) and a novel cellular delivery system comprising these liposomes for stealth delivery of biologically and therapeutically active payload molecules, such as small molecules, antisense oligonucleotides (ASOs), RNA molecules (e.g. siRNA), bioactive proteins, genome editing tools (e.g. cas9) and drugs into cells for treating a range of disorders.

[0136] Referring to FIG. 7, there is shown a schematic drawing summarising the four stages (explained below) by which the liposomes of the invention are produced: [0137] 1. The pore-forming protein (e.g. PA83) oligomerises in the membrane of the cell, thereby creating a pore, and the shuttle protein (e.g. LFn or PKR), which is optionally attached to a bioactive payload molecule (e.g. siRNA or Cas9 or an ASO etc.), interacts with the pore-forming protein and is internalised into the cell by endocytosis thereby forming an endocytic vesicle. [0138] 2. Endosomal sorting complexes required for transport (ESCRT) machinery transports the endocytic vesicle, which is loaded with the pore-forming protein and the shuttle protein, which is optionally attached to a bioactive payload molecule, to multivesicular bodies (MVBs) where the endocytic vesicles form intraluminal vesicles (ILVs). [0139] 3. Back-fusion events between ILVs, within a MVB, and the limiting membrane of the MVB during an apoptosis-linked gene 2-interacting protein X (ALIX) dependent process subvert the endomembrane system (which would usually result in fusion to lysosomes and destruction of the ILVs) to access the cytosol. [0140] 4. The MVB, comprising ILVs loaded with the pore-forming protein and the shuttle protein, which is optionally attached to a bioactive payload molecule, are then released from liposomes/exosomes containing the pore-forming protein and the shuttle protein, which is optionally attached to a bioactive payload molecule.

Materials and Methods

General Chemicals, Fluorescent Probes and Reagents

[0141] General laboratory reagents were from Sigma Aldrich (Dorset, UK) unless otherwise stated. Texas Red-N-hydroxysuccinimide ester (TxR-SE) was from Invitrogen (Paisley, UK). Dulbecco's Minimal Essential Medium, Eagles-Minimal Essential Medium, Non-essential amino acids, penicillin, streptomycin and glutamine solutions were all from Gibco (ThermoFisher Scientific, Paisley UK), Blasticidin solution was from Invitrogen (Paisley UK) and Ionomycin was from Sigma Aldrich (Dorset UK). Mouse Monoclonal anti-CD63 was from AbCam and monoclonal anti-Lamp2 was form DHSB (University of Iowa, Iowa, USA). Alexaflour 488-labeled goat anti-mouse antibody was from Invitrogen (Paisley UK). Goat anti-Texas Red monoclonal antibody and HRP-conjugated donkey anti-goat antibodies were from Vector Labs. The exoEasy Maxi kits (20) (Cat No: 76064) were from QIAgen and Total exosome Isolation reagent (from cell culture media) (Cat No: 4478359) (PEG solution) was from Invitrogen (Paisley UK). Stealth RNAi™ siRNA GFP Reporter Control (Cat No: 12935145) was from Invitrogen (Paisley UK) and was supplied as a 20 μM solution.

Exosome-Free Media

[0142] Bovine liposomes/exosomes in FCS were sedimented at 180 000×g for 18 h at 4° C. The supernatant was collected and added to serum free but otherwise complete media (MEM) and filter sterilized (0.2 μm filter, Sartorus) under negative pressure.

Cell Culture

[0143] The culture and passage of HeLa (ATCC: CCL2) and HEK293 (AMSBIO: SCO08) cells were performed as described by the supplier. Cells for microscopy were seeded onto sterile coverslips at a density of 1×10.sup.5 cells/well. Fixation, antibody hybridization and detection were performed as previously described [2]. CD63 immunostained cells were fixed in cold methanol.

Protein Production, Isolation and Enrichment

[0144] The DNA sequence coding for the protein PA83 (based upon GenBank accession numbers AAF86457 and AAT98414) has been previously described [2]. LFn-PKR was synthesized by BioBasic Inc., (Ontario, Canada) using the GenBank accession number AAY15237 (for LFn) and NM_002759 (for PKR). The open reading frame coding for LFn-PKR was sub-cloned into the bacterial expression cassette pET151/D (Invitrogen, Paisley, UK) as described previously in [2] and in PCT/GB2014/051918. Plasmids encoding LFn-Staphylococcus aureus (Sa) Cas9 and GST-PA63 were synthesized by Invitrogen using the pET151 bacterial expression system as the parent plasmid. The GST sequence used was from pGEX3× and the SaCAS9 sequence was codon optimised (i.e. SEQ ID No:14) from Genbank accession number CCK74173.1. The addition of a V5 epitope tag and a 6× histidine affinity tag allowed immunodetection and affinity purification of the protein from bacterial lysate.

[0145] LFn-PKR and PA83 were enriched from cultures of E. coli with a yield of approximately 2 mg/L (LFn-SaCas9 was approximately 0.5 mg/L). Using chemically competent E. coli BL21*DE3pLys (Invitrogen, Paisley, UK) transformed with 10 ng of plasmid and cultured overnight in 2×YT containing 200 μg/mL ampicillin (Sigma, Dorset, UK) and then grown in 1000 mL of 2×YT at 37° C. and 200 rpm for 3 h. Subsequently, isopropylthio-β-galactoside (IPTG) (Sigma, Dorset, UK) was added to a final concentration of 1 mM and incubated for a further 3 h. Bacterial pellets prepared by centrifugation (6 000×g for 6 min at 4° C.) were lysed using a French Press (Thermo Scientific, Paisley, UK) set to 15 000 psi. Lysates were cleared (20 000×g for 20 min at 4° C.) and the supernatant passed over a 6× histidine affinity chromatography column (Talon® resin; Clontech, Saint-Germain-en-Laye, France). The 6×His containing proteins were eluted using 150 mM imidazole (Sigma, Dorset, UK) in PBS, in fractions of 1 mL. Protein fractions were analyzed for purity and concentration, pooled, dialyzed to exhaustion against PBS and finally filter sterilized (0.2 μm filter, Sartorus). The final protein preparation was evaluated by SDS-PAGE and subjected to Coomassie staining (to determine purity) and Western blot analysis using the antibodies described.

Synthesis and Characterization of Probes

[0146] LFn-PKR-TxR and LFn-SaCas9-TxR were prepared using methods previously described [14]. Briefly, TxR-SE (5 mg) was dissolved in DMSO (5 mL). To 2.5 mL of PBS, 100 μL of the TxR solution was added to approximately 5 mg of recombinant protein and left in the dark at 25° C. for 1 h. The product was purified using PD-10 columns (GE Healthcare, Chalfont St Giles, UK) and PBS as eluent to collect 0.5 mL fractions. The most optically dense fractions were then selected and pooled to give either the LFn-PKR-TxR or the LFn-SaCas9-TxR conjugates. Fluorescent conjugates were then filter sterilized (0.2 m filter, Sartorus) and frozen at −80° C.

Cell Culture for Exosome Loading

[0147] Cells were seeded in 175 cm.sup.2 TC treated dishes and left to grow under standard incubation conditions i.e. 37° C. in 5% (v/v) CO.sub.2. At 90% confluence, the cell monolayer was washed three times in PBS prior to exosome loading.

Loading Liposomes/Exosomes with LFn-PKR-TxR or LFn-SaCas9-TxR

[0148] Cells were incubated with PA83 (50 μg/mL) and LFn-PKR-TxR (50 μg/mL) or LFn-SaCas9-TxR (50 μg/mL) in 3 mL of serum free DMEM at 37° C. for 1 hour. After 1 h, 5 mL of exosome free DMEM with 10% (v/v) FCS was added to a final volume of 8 mL and plates were left to incubate at 37° C. for 3 h.

Loading Liposomes/Exosomes with LFn-PKR::siRNA

[0149] LFn-PKR (50 μg/mL) was left to incubate with GFP siRNA (50 nM) in serum free DMEM for 5 minutes previous to adding PA83 (50 μg/mL). The mixture was then added to the cell monolayer and cells were incubated at 37° C. for 1 hour. After 1 h, 5 mL of exosome free DMEM with 10% (v/v) FCS was added to a final volume of 8 mL and plates were left to incubate at 37° C. for 3 h.

Liposome/Exosome Isolation

[0150] Ionomycin (50 μM) was added to the media and left to incubate for 30 minutes under standard conditions. The media was then collected and cell debris sedimented after centrifugation at 1,500×g for 2 min at 4° C. The resulting supernatant was filter sterilized (0.8 μm, Sartorous) prior to freezing or exosome isolation. Exosome isolation was performed using one of three methodologies:

[0151] 1. Differential Centrifugation

[0152] This was modified from [15]. Briefly, frozen filtered conditioned media was thawed on ice, and subject to 10 000×g for 30 min to sediment EVs. The supernatant was then subject to 110 000×g at 4° C. for 70 min and the pellet collected in 1 mL of PBS. The re-suspended pellet was then subject to a second round of sedimentation at 100 000×g for 70 min 4° C. The resulting pellet was then suspended in 1000 μL of exosome-free media, filtered through a 0.2 m filter (Sartorus) and stored frozen at −20° C. until required.

[0153] 2. Isolation by Polyethylene Glycol Precipitation

[0154] Briefly, the volume of the cleared and filtered cell culture media was estimated and to this, 0.5 volumes of the Isolation reagent was added. This preparation was left overnight at 4° C. The mix was then centrifuged at 10,000×g for 1 h at 4° C. The resulting pellet was then suspended in a final volume of 1000 μL of exosome-free media and stored frozen at −20° C. until required.

[0155] 3. Isolation by Membrane Adsorption

[0156] This was performed using the QIAgen exoEasy Kit according to manufacturer's specifications. Briefly, 8 mL of XPB buffer was mixed with the cell culture reagent and after the isolation, an additional step was added to remove the elution (XE) buffer from the exosome preparation. This was achieved by centrifuging the eluate at 110 000×g for 70 min at 4° C. The resulting pellet was then suspended in 1000 μL of exosome-free media and stored frozen at −20° C. until required.

Protein Quantification

[0157] The bicinchoninic acid assay (BCA) assay was performed according to the Bicinchoninic Acid Kit (BCA-1) (Sigma Aldrich, Dorset UK) specification to determine protein concentrations of final exosome samples previous to storage. Additionally, a ulite (BioDrop Inc.) apparatus was used to determine DNA and protein concentrations at OD.sub.260 and OD.sub.280, respectively, following the manufacturer's recommended protocols.

Microscopy

[0158] Microscopic visualisation of liposomes/exosomes was performed by mixing an equal volume of the exosome preparation with an equal volume of cell mask reagent which incorporated a Cy5 fluorophore (Cat. No. C10046; Invitrogen, Paisley, UK). This allowed the imaging of liposomes/exosomes under fluorescence using a LSM880 laser scanning confocal microscope fitted with an Airyscan unit (Carl Zeiss Ltd, Germany). The super-resolution capabilities of the Airyscan unit made the resolution of liposomes/exosomes possible. For liposomes/exosomes loaded with Texas-red labelled proteins, the liposomes/exosomes were visualised directly using the super resolution capabilities of the LSM880's Airyscan unit (Carl Zeiss Ltd, Germany). In both instances a Plan-Apochromat 63×/1.40 numerical aperture Oil DIC f/ELYRA objective was used. Immunostaining was performed on either paraformaldehyde fixed or cold (20° C.) methanol fixed cells grown on coverslips as previously described.

Assaying siRNA Activity

[0159] To assess the delivery of siRNA, control siRNA specific for GFP was purchased. This was directed against a stably expressed transgene expressed in the HEK293 (SCOO8) cells and used as a standard to report on gene activity. HEK293 cells overexpressed GFP fused in frame to beta-galactosidase, an enzyme responsible for the hydrolysis of x-gal from a colourless precursor to an insoluble blue compound, was detected spectrophotometrically at 620 nm. Consequently, it was possible to monitor GFP siRNA activity by measuring beta-galactosidase mediated X-gal conversion. Finally, beta-galactosidase activity was expressed as a percentage of the untreated control, after normalising it (OD.sub.620) to protein concentration.

Cell Culture for Gene Modulation Assay

[0160] Activity was assessed using 6-well plates. Cells were seeded at 5×10.sup.5 cells/well (HEK293, AMSBIO) and left to incubate at 37° C. in 5% (v/v) CO.sub.2 for 24 h prior to treatment.

Dosing with Liposomes/Exosomes

[0161] Cells were treated with 200 μL of exosome preparation, diluted in 2 mL of complete media.

[0162] This preparation was filter sterilised (0.2 μm filter, Sartorus) prior to being incubated with the cells for the desired time (24 h, 48 h, 72 h).

Assaying Gene Modulation

[0163] Media was discarded, and cell monolayer was carefully washed 3 times in chilled PBS, prior to adding 500 μL of RIPA buffer (R0278-50ML, Sigma Aldrich) to each well. Following a 15 min incubation period on ice, the cell lysate from each well was aspirated 10 times and decanted into a labelled Eppendorf. After centrifugation at 21,000×g for 10 min at 4° C., the supernatant was transferred into a new Eppendorf and the pellet discarded. Then, 10 μL of lysate were added to 100 μL of 2% BCA reagent in a 96-well plate and incubated at 37° C. The remaining 400 μL of supernatant were mixed with 12 μl of X-gal (50 mM in DMSO) (R0404, ThermoFisher) and transferred into a 96-well plate at 100 μL/well. X-Gal conversion was assayed over time (readings taken every 15 min over 5 hours) at 620 nm using the spectrophotometer set to 37° C.

Western Blotting and TCA Precipitation

[0164] Western blotting and immunodetection was performed using the mini-tetracell apparatus (BioRad) following the manufacturer's instructions. For protein separation, a 10% (w/v) acrylamide gel was used and run for 60 min at 200 V. Transfer onto nitrocellulose membrane was performed at 400 mM for 60 min. Blocking was performed for 45 min using 5% (w/v) non-fat dried milk solution in PBS containing 0.1% (v/v) tween 20 reagent. Antibody hybridisations were performed in 3 mL at 37° C. for 60 min under shaking conditions using the antibody dilutions suggested by the manufacturer. The detection of HRP-labelled secondary antibodies was performed using enhanced ECL reagent (Pierce, ThermoFisher Scientific) following the manufacturer's instructions. Gels and blots were calibrated by running broad range pre-stained protein markers (Invitrogen). TCA precipitation of exosome proteins was performed by adding 0.6 volumes of TCA to the exosome preparation. This was then left to incubate at 4° C. for 30 min. The preparation was then sedimented at 21 000×g for 10 min at 4° C. and the pellet washed twice in acetone, also at 4° C. The resulting pellet was dissolved in Laemmli buffer, subject to western immunoblotting and probed with either an anti-LAMP2 specific primary antibody (DHSB, University of Iowa, Iowa, USA) under non-reducing conditions; or a Texas Red specific primary antibody (Vector labs) using dilutions suggested by the manufacturers.

Example 1—Liposomes are Taken Up by Cells

[0165] As shown in FIG. 1, Texas red-labelled cargo (i.e. Texas Red labelled LFn-PKR) localises to intraluminal structures within HeLa cells. As these intraluminal structures are positive for the exosome immunomarker CD63 (also known as LAMP3), it is likely that this intraluminal signal is within multivesicular endosomes, i.e. late endosomes. This shows that Texas Red labelled LFn-PKR is, when added to cells with PA83, able to preferentially label intraluminal membrane within MVE/MVBs 3 h after being added to cells.

Example 2—Liposomes are a Similar Size to Physiological Exosomes

[0166] FIG. 2 shows the inventor's findings which demonstrate that liposomes/exosomes isolated from Hela conditioned media, stained with Cy5-Cell Mask, and visualised using an Airyscan detector were approximately the right size for exosomes (60-120 nm). It should be noted that the limits of this system's resolution are 120 nM in the x-y plain. Liposomes/exosomes isolated from the exoEasy kit were also subject to immunoblotting analysis using LAMP2 as a probe and, as would be predicted, a band is visible within the exosome preparation at approximately the right molecular weight. This means that the liposomes/exosomes isolated using the exoEasy kit were not only approximately the correct size but also contained well characterised exosome immunomarkers as would be predicted.

Example 3—Liposomes can be Loaded with Cas9

[0167] Liposomes/exosomes have been effectively loaded with Staphylococcus aureus Cas9, i.e. SaCAS9. In FIG. 3, the red signal from the LFn-SaCAS9 can be clearly seen in exosome preparations from Hela cells exposed to PA83 and Texas Red-labelled LFn-SaCAS9. Control liposomes/exosomes R with: no PA83, no cargo or a non-translocation cargo (BSA-Texas Red) did not produce any red signal even when incubated with cell mask to check the plane of focus (FIG. 3B).

Example 4—Liposomes can be Loaded with Small Molecules

[0168] Liposomes/exosomes have been further shown to be effectively loaded with small molecules. In FIG. 4, the small molecule, Texas Red, is shown to be taken up by exosome preparations from Hela cells exposed to PA83 and Texas Red-labelled LFn-PKR or PA83 after 3 h. Texas Red-labelled BSA was used as a negative control. Here Texas Red-labelled LFn-PKR can be readily detected within cell mask positive populations of liposomes/exosomes whereas Texas Red-labelled BSA cannot.

Example 5—Loaded Protein is Present in Isolated Liposomes

[0169] FIG. 5 shows the result of TCA precipitating liposomes/exosomes prepared with PA83 and either: Texas Red-labelled LFn-SaCAS9, Texas Red-labelled LFn-PKR, or Texas Red-labelled BSA. From the PA83 Texas Red labelled LFn-SaCAS9 and PA83 Texas Red labelled LFn-PKR preparations, Texas Red is clearly visible in the pellet. Texas Red is not readily detectable for the “no treatment” or PA83 and BSA-Texas Red “treated” controls. Similarly, after immunoblotting and detection using a Texas Red specific primary antibody, Texas Red was detected labelling proteins of the predicted molecular weight from the same TCA precipitate as before.

Example 6—Liposomes Loaded with siRNA are Effective at Down-Regulating Proteins

[0170] The biological activity of liposomes/exosomes isolated form cell culture media using the exoEasy kit and differential centrifugation is shown in FIG. 6. Here a reduction of beta-galactosidase activity per unit cell protein was recorded demonstrating 1) the inventors' ability to load siRNA into liposomes/exosomes and 2) the biological activity of the siRNA loaded liposomes/exosomes i.e. their ability to deliver siRNA into the cytosol of a second population of cells.

Example 7—Treatment of Zika Virus Infection

[0171] As a theoretical example, the liposomes produced by the method of the first aspect are used to treat a patient who has contracted the Zika virus. Firstly, a biopsy is carried out on the patient, and a number of the patient's cells are isolated. Liposomes are then produced from the cells using the method of the first aspect, and loaded with anti-Zika virus siRNA. The liposomes containing the anti-Zika virus siRNA are then administered to the patient and are up-taken by the patents cells via endocytosis. The anti-Zika virus siRNA is therefore present in the patient's cells and the ability of the Zika virus to multiply in the patient's body is inhibited.

Example 8—Treatment of Obesity

[0172] As another theoretical example, the liposomes produced by the method of the first aspect are used to treat a patient who is suffering from FMO5-regulated obesity. In this example, liposomes are then derived from cultured mesenchymal stem cells using the method of the first aspect, and loaded with anti-FMO5 siRNA. The liposomes containing the anti-FMO5 siRNA are then administered to the patient and are up-taken by the patient's cells via endocytosis. The anti-FMO5 siRNA is therefore present in the patient's cells and the FMO5 enzyme is down-regulated and the patient no longer presents obesity-related symptoms.

Discussion

[0173] The materials which the inventors have loaded into liposomes/exosomes using PA83::LFn fusions include: low molecular weight covalent conjugates (i.e. Texas Red), LFn-PKR::siRNA, LFn-PKR conjugated-Texas Red, LFn-Gal4::eGFP-Rab5 and LFn-SaCAS9 conjugated-Texas Red. Consequently, the data presented here supports the novel idea that LFn fusion proteins can deliver selective cargo (rather than just LF or EF) into a biologically derived, stealth delivery system termed an exosome. Here, for the first time, a methodology that can achieve exosome loading without cargo overexpression or exosome disruption is disclosed and the evidence to support this conclusion discussed.

[0174] FIG. 1 demonstrates that Texas Red-labelled LFn-PKR can, in the presence of PA83, associate with intraluminal vesicles within a CD63 positive, membrane-delimited structure. These data support the hypothesis that LFn-PKR uses a similar cytosolic translocation pathway to wild type LF [3]. FIG. 2 validates the isolation of liposomes/exosomes using the QIAgen exoEasy kit through both microscopy (measuring vesical size) and immunoblotting detection of the exosome marker LAMP2 within a population of isolated liposomes/exosomes. This data indicates that liposomes/exosomes have indeed been isolated. FIG. 3 provides evidence that the isolated liposomes/exosomes contain cargo protein labelled with the fluorophore Texas Red. In this instance the cargo protein is LFn-SaCAS9, which has previously been reported to be unlikely to act as a PA translocase substrate. Here LFn-SaCas9 has been documented within populations of isolated liposomes/exosomes. FIG. 4 (panel a) duplicates this methodology showing reproducibility only this time using different cargo protein: Texas Red-labelled LFn-PKR. This further demonstrates the ability of this system to move selected small molecules, covalently conjugated to PA pore substrates like LFn-fusion proteins, into populations of liposomes/exosomes. FIG. 4 (panel B) serves as a negative control showing that: 1) BSA labelled-Texas Red doesn't act as a PA translocase substrate, 2) that the signal documented is specific to Texas Red and not attributable to either autofluorescence of bleed from the Cy5 cell mask channel.

[0175] FIG. 5 demonstrates that the Texas Red signal from isolated liposomes/exosomes was able to be precipitated using trichloroacetic acid (TCA) (i.e. attached to a protein) and that after Western analysis, it was of the predicted molecular weight. This indicates that the protein was intact and remained associated with the Texas Red fluorophore after PA pore translocation. FIG. 6 demonstrates that PA and LFn-PKR can be used to load liposomes/exosomes with siRNA and that these liposomes/exosomes can be isolated by either the exoEasy kit or by differential centrifugation. It also shows that the liposomes/exosomes were active, recipient cell fusion competent, and capable of delivering pharmacologically active siRNA. This was proof of concept, i.e. the described exosome loading methodology can be used to load and deliver cargo into liposomes/exosomes and that the liposomes/exosomes can be isolated and used to transfer biologically active cargo from one population of cells to another. This would support the idea that this methodology could be used to load drugs into an exosome derived from a patient's cells grown ex vivo in order to facilitate the third order targeting and stealth delivery of personalised, precision medicine such as siRNA, gene editing proteins and gRNA, shRNA, miRNA, genes and therapeutic proteins.

[0176] In an attempt to optimise the loading of material into liposomes/exosomes, both PA63 (produced as PA63-TEV recognition site-GST), PA83, PA83 D.sup.512K, PA83 G to N and PA83 N to S [16] were also investigated for their capacity to load liposomes/exosomes. Finally, a PA83 hybrid molecule incorporating the trans-membrane domain of haemolysin [17] replacing the PA63 trans-membrane domain was encoded into a bacterial expression cassette (pET151), also constructed to investigate the roll of the PA83 trans-membrane assembly in regard to the rate limits associated with the phenomena of Brownian ratchetting [6] during cargo pore transit.

Example 9

[0177] The inventors loaded exosomes with the conditionally lethal cargo LFn-Diphtheria toxin A chain (DTA) using PA83 as described above, with the exception of the plasmid used, i.e. the plasmid was from Addgene (pET-15b LFn-DTA, Addgene number 11075) (https://www.addgene.org/11075/)

Loading Exosomes

[0178] The exosomes were loaded by incubating LFnDTA—SEQ ID No: 15 (protein) and SEQ ID No: 16 (DNA)—with HeLa cells at a concentration of 10 μg/ml and 50 μg/ml of SEQ ID No. 2 (PA83) for 4 hours at 37° C. in humidified atmosphere containing 5% (v/v) CO.sub.2. After this time the cells were washed with PBS and incubated with 5 μM Ionomycin (Sigma chemical company catalogue number I9657-1MG) in serum free media. Exosomes were then isolated from the now conditioned media using differential centrifugation as previously described.

Trypsin Digestion

[0179] To half of the exosome preparation, 5 μl of cell culture trypsin/EDTA (TE) buffer (ThermoFisher Scientific catalogue number 25200056) was added to the exosomes and the volume adjusted to 100 μl with PBS. To the other half of the preparation, PBS was added to 100 μl. The exosome preparations were then incubated for 60 min under conditions that had already been demonstrated to be sufficient to digest 5 μg of LFnDTA, vastly in excess of the amount of LFnDTA contained within the exosome preparation. The exosomes were then added to a culture of HeLa cells with a trypsin control (found to be non-toxic) and cell viability assayed after 24 h. Results were expressed as DTA activity (%) normalized to the untreated control (LFnDTA containing exosomes killed about 45% of the cells).

Results

[0180] Referring to FIG. 8, exosomes initially loaded with the cargo (i.e. LFn-diphtheria toxin A chain (DTA)—SEQ ID No: 15 and 16) using PA83 (i.e. SEQ ID No:2) were subsequently characterised to show that they were protected from the activity of external enzymes (i.e. trypsin). These data show that the exosomes can be successfully loaded with LFnDTA, which could be used for treating cancer, such as cervical carcinoma.

Example 10

[0181] In this Example, the inventors used dynamic light scattering AKA photon correlation spectroscopy to characterize the size of exosomes and extracellular vesicles (EVs) isolated using differential centrifugation.

Loading Exosomes

[0182] Exosomes from HeLa cells were loaded with prehybridized phosphorothioate-phosphodiester hybrid antisense oligonucleotides (ASOs) specific for tandem dimeric tomato at a concentration of 200 pMol/ml (total ASO)—SEQ ID No: 17 and 18. The following table explains the phosphorothioate codes (taken from Thermofisher website).

TABLE-US-00019 A sulphur is substituted for one of the oxygens See in the phosphodiester bonds between the nucleotides. Phosphorothioates below This linkage is to the 3′ side of the designated base. A-Phosphorothioate F C-Phosphorothioate O G-Phosphorothioate E T-Phosphorothioate Z

[0183] These ASOs were loaded into exosomes using 50 μg/ml LFnPKR and 50 μg/ml SEQ ID No: 2, (PA83) or 50 μg/ml PA forced octamer mutants. This mixture was incubated with the HeLa cells in serum free media for 4 hours at 37° C. in a humidified atmosphere containing 5% (v/v) CO.sub.2. The media removed and the cells washed with PBS prior to being incubated with 5 μM ionomycin in serum free media for 30 min.

Exosome Isolation

[0184] The conditioned media was first cleared by centrifugation at 1.5 k×g for 2 min and filtered through a 0.8 micron filter. The flow through was subject to centrifugation at 10 000×g for 30 min at 4° C. Finally exosomes were isolated by sedimentation at 110 000×g for 70 min at 4° C., resuspended in PBS and then re-sedimented at 110,000×g for 70 min at 4° C. Exosomes were stored at 4° C. until then were used. When used for cell culture the exosomes were diluted in the desired amount of serum free media and again filtered through a 0.8μ filter prior to incubation with cells.

Results

[0185] Referring to FIG. 9, exosomes were shown to have the predicted size when loaded with 200 pMol/ml anti-tandem dimeric tomato (TdTom) antisense oligonucleotides (ASOs) using 50 μg/ml PA83 (SEQ ID No:2) and 50 μg/ml LFnPKR (SEQ ID No:13).

[0186] Referring to FIG. 10, exosomes were also shown to have the predicted size when loaded with 200 pMol/ml anti-TdTom ASOs using forced octamer PA83 mutants, i.e. 25 μg/ml PA83D.sup.512K (SEQ ID No:7) and 25 μg/ml PA83 K245G; R252N (SEQ ID No:8) with 50 μg/ml LFnPKR (SEQ ID No:13). These data demonstrate that the membrane fractions that are generated are of the predicted size for exosomes and EVs.

Example 11

[0187] The inventors tested if exosomes loaded with active ASOs retain their pharmacological activity.

Loading Exosomes

[0188] Exosomes were loaded by incubating HeLa cells with 200 pMol/ml anti-TdTom ASOs (SEQ ID No: 17 and 18) and either 50 μg/ml heptameric (PA83—SEQ ID No:2), or forced octamer PA83 mutants (i.e. 25 μg/ml—SEQ ID No: 7, (PA83 D.sup.512K) and 25 μg/ml SEQ ID No: 8, (PA83 K.sup.245G; R.sup.252N)) and 50 μg/ml SEQ ID No: 13 (LFnPKR). After Isolation, these ASO loaded exosomes demonstrated antisense activity against an mRNA target encoding GFP fused beta-galactosidase and bicistronically expressing tandem dimeric tomato overexpressed in HEK293 cells (Catalogue number SC008 from ASMBIO) HEK cells (˜1×10.sup.6 per well) were treated with: 57 μg [total protein measured at OD.sub.280] of the heptamer::ASO exosome prep or 42 μg [total protein measured at OD.sub.280] of the octamer::ASO.

Results

[0189] Referring to FIG. 11, the exosomes loaded with ASOs are shown to have pharmacological activity. These data demonstrates that the methodology can be used to load ASOs into exosomes, and that the exosomes: (i) are fusion competent; (ii) contain ASOs; (iii) the ASOs are active; and (iv) the exosomes can be used to deliver ASOs. These data also show that variants of PA can be used to load exosomes.

Example 12

[0190] Exosomes were loaded using 50 nM stealth reporter anti-GFP siRNA (Invitrogen catalogue number 12935-145), 50 μg/ml PA83 (SEQ ID No: 2) and 50 μg/ml LFnPKR (SEQ ID No: 13) resuspended in serum free media. Exosomes were isolated as before i.e. after a 4 hour incubation with the protein::siRNA mixture and a 30 min incubation with 5 μM Ionomycin in serum free media. All incubations were carried out at 37° C. and the cells were washed with PBS between incubations as before. Exosomes were isolated using the exoEasy kit from Qiagen (catalogue number 76064). Exosomes were washed in PBS and sedimented as before (i.e. 110 000×g for 70 min at 4° C.) prior to being resuspended in 500 μl of PBS. Exosomes were diluted using complete media and added to HEK293 cells stably expressing GFP fused beta-galactosidase and tandem dimeric tomato (Catalogue number SC008 from ASMBIO). Beta-galactosidase activity in cell lysate was then assayed by measuring X-gal conversion at OD.sub.620 over time and normalising conversion to total cell lysate protein concentration.

[0191] Referring to FIG. 12, it can be seen that anti-GFP siRNA loaded into HeLa derived exosomes using PA83 and LFnPKR have pharmacological activity. These data demonstrate that, as with the ASOs of Example 11, siRNA retains its activity in the exosomes.

Summary

[0192] Advantages of the aspects and embodiments of the liposomes described herein reside in seeking to address one or more problems inherent in the prior art by sequestering ILVs loaded with Atx associated cargo as liposomes (e.g. exosomes) prior to ILV back-fusion. In addition, the use of membrane spanning sequences that perform the same function as Atx PA63 but contain recombinant trans-membrane sequence is also addressed.

[0193] Liposomes/exosomes can shield intraluminal content from enzymatic destruction and the immune response and are often referred to as a naturally occurring, paracrine transport system that protects antigenic or enzymatically labile material in transit. Given that wild type LF has been documented within both ILVs and liposomes (e.g. exosomes), and it is known that ILVs can be secreted from cells as liposomes (e.g. exosomes) after the release of ER stored calcium [5], it has been reasoned by the inventors that recombinant LF could also be trapped or loaded into liposomes (e.g. exosomes). Further, the use of an ionophore (e.g. ionomycin) results in the release of ER calcium on demand, triggering the exocytosis of ILVs, as liposomes (e.g. exosomes). Consequently, ionomycin was used to temporally capture ILVs containing cargo, from cells that were previously treated with the Atx derived delivery system. This then allowed for the isolation of cargo loaded exosomes secreted into the cell culture media. Although Atx components protective antigen (PA) 83 or PA63, lethal factor (LF) and oedema factor (EF) have been reported to localise to liposomes/exosomes [4], the loading of associated molecules (i.e. LFn-GAL4, LFn-PKR, LFn-PKR-Texas Red, siRNA, CAS9 or Cas9-Texas Red) within liposomes/exosomes using pore-forming recombinant proteins has not been previously reported.

REFERENCES

[0194] [1] Tyagi P., Subramony J. A. (2018) Nanotherapeutics in oral and parenteral drug delivery: Key learnings and future outlooks as we think small. J. Control Release. 272:159-168. [0195] [2] S. C. Richardson, S. C. Winistorfer, V. Poupon, J. P. Luzio, R. C. Piper. (2004) Mammalian late vacuole protein sorting orthologues participate in early endosomal fusion and interact with the cytoskeleton, Mol Biol Cell. 15, 1197-1210. [0196] [3] Abrami, L., Lindsay, M., Parton, R. G., Leppla, S. H., & van der Goot, F. G. (2004). Membrane insertion of anthrax protective antigen and cytoplasmic delivery of lethal factor occur at different stages of the endocytic pathway. The Journal of Cell Biology, 166(5), 645-651. [0197] [4] Abrami, L., Brandi, L., Moayeri, M., Brown, M. J., Krantz, B. A., Leppla, S. H., & van der Goot, F. G. (2013). Hijacking Multivesicular Bodies Enables Long-Term and Exosome-Mediated Long-Distance Action of Anthrax Toxin, 5(4), 986-996. [0198] [5] Kuznetsov G., Brostrom M. A., and Brostrom C. (1992) Demonstration of a Calcium Requirement for Secretory Protein Processing and Export. The Journal of Biological Chemistry. 267(6); 3932-3939. [0199] [6] Blaustein, R. O., Koehler, T. M., Collier, R. J., and Finkelstein, A. Proc. Natl. Acad. Sci. USA 86, 2209-2213 (1989). [0200] [7] B. A. Krantz, et al., Acid-induced unfolding of the amino-terminal domains of the lethal and edema factors of anthrax toxin, J. Mol. Biol. 344 (3) (2004) 739-756. [0201] [8] Auger A., Park M., NitschkeF., MinassianL. M., BeilhartzG. L., Minassian B. A., and Melnyk R. A. (2105) Efficient Delivery of Structurally Diverse Protein Cargo into Mammalian Cells by a Bacterial Toxin. Mol. Pharmaceutics, 12 (8), pp 2962-2971 [0202] [9] Dyer P. D. (2013) Development of a Protein-Based Antisense Delivery Platform Modelled on Anthrax Toxin. PhD Thesis, University of Greenwich. [0203] [10] Gaur, R., Gupta, P., Goyal, A., Wels, W. & Singh, Y. Delivery of nucleic acid into mammalian cells by anthrax toxin. Biochem. Biophys. Res. Commun. 297, 1121-1127 (2002). [0204] [11] Baillie, L W., Huwar, T. B., Moore, S., Mellado-Sanchez, G., Rodriguez, L., Neeson, B. N., et al. (2010). An anthrax subunit vaccine candidate based on protective regions of Bacillus anthracis protective antigen and lethal factor. Vaccine, 28(41), 6740-6748. [0205] [12] Khandia R., Bhatia S., Chanu K. V., Sood R. and Dhama K. (2014). Anthrax Toxin Receptors, Functions and their Possible Use in Therapeutics: A Review. Asian Journal of Animal and Veterinary Advances, 9: 599-609. [0206] [13] Guo S. & Huang L. (2011) Nanoparticles Escaping RES and Endosome: Challenges for siRNA Delivery for Cancer Therapy. Journal of Nanomaterials. Article ID 742895. DoI: 10.1155/2011/742895 [0207] [14] S. C. Richardson, S. C. Winistorfer, V. Poupon, J. P. Luzio, R. C. Piper. Mammalian late vacuole protein sorting orthologues participate in early endosomal fusion and interact with the cytoskeleton, Mol Biol Cell. (2004) 15, 1197-1210. [0208] [15] Willms E., Johansson H. J., Mager I., Lee Y., et al., (2015) Cells Release Subpopulations of Exosomes with Distinct Molecular and Biological Properties. Scientific Reports 6: 22519. [0209] [16] Phillips D. D., Fattah R. J., Crown D., et al. (2013) Engineering Anthrax Toxin Variants That Exclusively Form Octamers and Their Application to Targeting Tumors. J. Biol. Chem., 288: 9058-9065. [0210] [17] Karginov, V. A., Nestorovich E. M., Schmidtmann, F. Robinson T. M., Yohannes, A., Fahmi N. E., Bezrukov S. M., and Hecht S. M. (2007) Inhibition of S. aureus α-Hemolysin and B. anthracis Lethal Toxin by β-Cyclodextrin Derivatives. Bioorg Med Chem.; 15(16): 5424-5431. [0211] [18] Feld G. K., Brown,M. J. & Krantz B. A. (2012) Ratcheting up protein translocation with anthrax toxin. PROTEIN SCIENCE. 21:606-624.