RED BLOOD CELL-DERIVED VESICLE

Abstract

The invention relates to red blood cell-derived vesicles comprising encapsulated active agents, their use in therapy and methods of production thereof.

Claims

1. A red blood cell-derived vesicle comprising an encapsulated active agent.

2. The vesicle according to claim 1, comprising at least one targeting ligand or moiety attached to the surface thereof.

3. The vesicle according to claim 2, wherein the at least one targeting ligand or moiety is configured to selectively target the vesicle to a cell involved in blood clotting or a factor present in the characteristic microenvironment of a thrombus site.

4. The vesicle according claim 2 , wherein the at least one targeting ligand or moiety is configured to selectively target the vesicle to a platelet.

5. The vesicle according to claim 2, wherein the at least one targeting ligand is a peptide selected from a group of peptides consisting of: SEQ ID No: 1; SEQ ID No: 2; SEQ ID No: 3; SEQ ID No: 4; SEQ ID No: 5; SEQ ID No: 6; SEQ ID No: 7; SEQ ID No: 8; SEQ ID No: 9; SEQ ID No: 10; SEQ ID No: 11; SEQ ID No: 12; SEQ ID No: 13; SEQ ID No: 14; SEQ ID No: 15; and SEQ ID No: 16.

6. (canceled)

7. The vesicle according to claim 1, wherein the active agent is a thrombolytic agent.

8. The vesicle according to claim 1, wherein the active agent is selected from a group consisting of: a fibrinolytic agent; a von Willebrand factor-cleaving protease (VWFCP); and a DNase that is capable of degrading neutrophil extracellular traps (NETs).

9. The vesicle according to claim 7, wherein the thrombolytic agent is tPA.

10. The vesicle according to claim 1, wherein the active agent is an antiplatelet agent selected from a group consisting of: a GPIIb-IIIa α.sub.IIbβ.sub.3) inhibitor; an irreversible cyclooxygenase inhibitor; an adenosine diphosphate (ADP) receptor inhibitor; an adenosine reuptake inhibitor; a phosphodiesterase inhibitor; a protease-activated receptor-1 (PAR-1) antagonist; and a thromboxane inhibitor.

11. The vesicle according to claim 10, wherein the anti-platelet agent is acetylsalicylic acid (aspirin, ASA).

12. The vesicle according to 6 claim 1, comprising an active agent selected from a thrombolytic agent, a fibrinolytic agent, a von Willebrand factor-cleaving protease (VWFCP); and a DNase that is capable of degrading neutrophil extracellular traps (NETs) comprising an anti-platelet agent selected from a group consisting of: a GPIIb-IIIa (a.sub.IIbβ.sub.3) inhibitor; an irreversible cyclooxygenase inhibitor; an adenosine diphosphate (ADP) receptor inhibitor; an adenosine reuptake inhibitor; a phosphodiesterase inhibitor; a protease-activated receptor-1 (PAR-1) antagonist; a thromboxane inhibitor; and acetylsalicylic acid (aspirin, ASA).

13. A method of preparing a red blood cell-derived vesicle (RBCV) comprising an encapsulated active agent, the method comprising: (i) contacting a red blood cell with a hypotonic solution to produce a red blood cell ghost; and (ii) encapsulating an active agent using the red blood cell ghost, to thereby produce a red blood cell-derived vesicle comprising an encapsulated active agent.

14. The method according to claim 13, wherein the active agent is as defined in claim 7.

15. The method according to claim 13 , wherein step (ii) further comprises contacting the red blood cell ghost with at least one targeting ligand or targeting moiety, and wherein the at least one targeting ligand or targeting moiety is configured to selectively target the vesicle to a cell involved in blood clotting or a factor present in the characteristic microenvironment of a thrombus site.

16-25. (canceled)

26. A method of treating, preventing, ameliorating, or reducing a thrombotic disorder or a blood clot, the method comprising administering, or having administered, to a subject in need thereof, a therapeutic amount of the red blood cell vesicle according to claim 1.

27. The method of claim 26, wherein the thrombotic disorder is selected from the group consisting of: ischemic stroke; myocardial infarction and pulmonary embolism.

Description

[0215] 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 Figures, in which:-

[0216] FIG. 1 shows a schematic illustration of the mechanism of platelet aggregation (top left), an embodiment of a fibrinogen-mimicking red blood cell vesicle (RBCV) according to the invention (FM-RBCV, top right), and also targeted delivery of tissue plasminogen activator (tPA) to a blood clot using the tPA-loaded, fibrinogen-mimicking RBCV (tPA-RGD-RBCV or tPA-cRGD-PEG-RBCV) for targeted thrombolysis to remove the blood clot (bottom left and right).

[0217] FIG. 2 shows a schematic illustration of one embodiment of the preparation of tPA-loaded, fibrinogen-mimicking, RBC-derived vesicles (tPA-RGD-RBCVs or tPA-cRGD-PEG-RBCVs) according to the invention.

[0218] FIG. 3 shows a typical dynamic light scattering (DLS) plot of tPA-RGD-RBCVs in pH 7.4 PBS buffer. Inset: representative photograph of tPA-RGD-RBCVs solution.

[0219] FIG. 4 shows a representative transmission electron microscope (TEM) image of the tPA-RGD-RBCVs according to the invention. The scale bar was 200 nm.

[0220] FIG. 5 shows the half maximal inhibitory concentration (IC50) of linear RGD and cyclic RGD (cRGD) measured by the platelet aggregation assay.

[0221] FIG. 6 shows (A) Typical flow cytometry histogram profiles of resting platelets incubated with FITC-labelled tPA-RBCVs, FITC-labelled tPA-RGD-RBCVs and FITC-labelled tPA-cRGD-PEG-RBCVs, respectively. Resting platelets treated with pH 7.4 PBS buffer were used as the control. (B) Fluorescence intensities of resting platelets treated with FITC-labelled tPA-RBCVs, FITC-labelled tPA-RGD-RBCVs and FITC-labelled tPA-cRGD-PEG-RBCVs relative to the control, respectively.

[0222] FIG. 7 shows (A) Typical flow cytometry histogram profiles of activated platelets incubated with FITC-labelled tPA-RBCVs, FITC-labelled tPA-RGD-RBCVs and FITC-labelled tPA-cRGD-PEG-RBCVs, respectively. The platelets treated with pH 7.4 PBS buffer were used as the control. (B) Fluorescence intensities of activated platelets respectively treated with FITC-labelled tPA-RBCVs, FITC-labelled tPA-RGD-RBCVs and FITC-labelled tPA-cRGD-PEG-RBCVs relative to the control.

[0223] FIG. 8 shows (A) Typical confocal laser scanning microscopy images of resting or activated platelets incubated with FITC-labelled tPA-RBCVs, FITC-labelled tPA-RGD-RBCVs and FITC-labelled tPA-cRGD-PEG-RBCVs, respectively. (B) Mean fluorescence intensity (MFI) in the fluorescence images of resting or activated platelets after incubation with FITC-labelled tPA-RBCVs, FITC-labelled tPA-RGD-RBCVs and FITC-labelled tPA-cRGD-PEG-RBCVs, respectively.

[0224] FIG. 9 shows relative fluorescence intensity of activated platelets incubated with the FITC-labelled tPA-RGD-RBCVs containing different RGD peptide arm densities as measured by flow cytometry.

[0225] FIG. 10 shows tPA release profiles of tPA-RBCVs, tPA-RGD-RBCVs and tPA-cRGD-PEG-RBCVs in the presence of resting or activated platelets.

[0226] FIG. 11 shows NBD fluorescence intensity after incubation of RGD-RBCVs or cRGD-PEG-RBCVs (containing 0.01 mmol NBD-PE and 0.01 mmol Rhod-PE) with resting platelets, activated platelets or eptifibatide-pretreated activated platelets, respectively, at various time intervals.

[0227] FIG. 12 shows (A) Representative photograph of fibrin clots treated with PBS buffer only, tPA-RBCVs, tPA-RGD-RBCVs, tPA-cRGD-PEG-RBCVs and free tPA, respectively, in the presence of activated platelets. (B) The area of the fibrin lysis ring after treatment with PBS buffer only, tPA-RBCVs, tPA-RGD-RBCVs, tPA-cRGD-PEG-RBCVs or free tPA in the presence of activated platelets. Statistical analysis was performed using the Student's t-test. The triple asterisk symbol (***) denotes p < 0.001.

[0228] FIG. 13 shows (A) Representative photograph of Halo blood clots after treatment with PBS buffer only, RBCVs, tPA-RBCVs, tPA-RGD-RBCVs, tPA-cRGD-PEG-RBCVs and free tPA, respectively. (B) Time-dependent clot lysis in the Halo model after treatment with PBS buffer only, RBCVs, tPA-RBCVs, tPA-RGD-RBCVs, tPA-cRGD-PEG-RBCVs and free tPA, respectively.

[0229] FIG. 14 shows (A) A schematic illustration of the mechanism of platelet aggregation (left) and an embodiment of a fibrinogen-mimicking RBCV according to the invention (right). (B) A schematic illustration of the acetylsalicylic acid (ASA, Aspirin)-loaded, fibrinogen-mimicking RBCV (ASA-RGD-RBCV or ASA-cRGD-PEG-RBCV) for inhibition of platelet aggregation.

[0230] FIG. 15 shows a schematic illustration of one embodiment of the preparation of ASA-loaded, fibrinogen-mimicking RBCVs (ASA-RGD-RBCVs or ASA-cRGD-PEG-RBCVs) according to the invention.

[0231] FIG. 16 shows (A) Typical flow cytometry histogram profiles of resting platelets incubated with NBD-labelled RBCVs, NBD-labelled RGD-RBCVs and NBD-labelled cRGD-PEG-RBCVs, respectively. The platelets treated with pH 7.4 PBS buffer were used as the control. (B) Fluorescence intensities of activated platelets treated with NBD-labelled RBCVs, NBD-labelled RGD-RBCVs, and NBD-labelled cRGD-PEG-RBCVs relative to the control, respectively.

[0232] FIG. 17 shows (A) Typical flow cytometry histogram profiles of activated platelets incubated with NBD-labelled RBCVs, NBD-labelled RGD-RBCVs and NBD-labelled cRGD-PEG-RBCVs, respectively. The platelets treated with pH 7.4 PBS buffer were used as the control. (B) Fluorescence intensities of activated platelets treated with NBD-labelled RBCVs, NBD-labelled RGD-RBCVs, and NBD-labelled cRGD-PEG-RBCVs relative to the control, respectively.

[0233] FIG. 18 shows (A) Typical confocal laser scanning microscopy images of resting or activated platelets incubated with NBD-labelled RBCVs, NBD-labelled RGD-RBCVs and NBD-labelled cRGD-PEG-RBCVs, respectively. (B) Mean fluorescence intensity (MFI) in the fluorescence images of resting or activated platelets after incubation with NBD-labelled RBCVs, NBD-labelled RGD-RBCVs and NBD-labelled cRGD-PEG-RBCVs, respectively.

[0234] FIG. 19 shows inhibition ratios of PBS, ASA-RBCVs, ASA-RGD-RBCVs, ASA-cRGD-PEG-RBCVs and ASA against (A) adenosine diphosphate (ADP)-induced, (B) arachidonic acid (AA)-induced or (C) thrombin-induced platelet aggregation.

[0235] FIG. 20 shows a schematic illustration of the mechanisms of inhibition of cyclooxygenase (COX) by ASA.

[0236] FIG. 21 shows the level of (A) thromboxane B2 (TXB2) and (B) prostaglandin F1α (PGF1α) in plasma after incubation with PBS, ASA-RBCVs, ASA-RGD-RBCVs, ASA-cRGD-PEG-RBCVs and ASA, respectively. Statistical analysis was performed using the Student's t-test. The double asterisk symbol (**) denotes p < 0.01.

[0237] FIG. 22 shows (A) a representative photograph of the wet thrombi after treatment with PBS, ASA-RBCVs, ASA-RGD-RBCVs, ASA-cRGD-PEG-RBCVs and ASA, respectively, and (B) the weight of thrombi formed in the presence of PBS, ASA-RBCVs, ASA-RGD-RBCVs, ASA-cRGD-PEG-RBCVs and ASA, respectively. Statistical analysis was performed using the Student's t-test. The single asterisk symbol (*) denotes p < 0.05, the double asterisk symbol (**) denotes p < 0.01, and NS represents no significant difference between two groups.

EXAMPLES

Materials and Methods

Materials

[0238] Tissue plasminogen activator (tPA, alteplase) was a product of Boehringer Ingelheim (Germany). 1, 2-distearoyl-sn-glycero-3-phosphoethanolamine-N-[4-(p-(cysarginylglycylaspartate-maleimidomethyl)cyclohexane-carboxamide] (DSPE-RGD), acetylsalicylic acid (aspirin, ASA), 1, 2-distearoyl-sn-glycero-3-phosphoethanolamine-N-[amino(polyethylene glycol)-2000] (DSPE-PEG-NH2), cyclo(Arg-Gly-Asp-d-Phe-Val) (cRGD), 4-dimethylaminopyridine (DMAP), N-hydroxy succinimide (NHS), 1-(3-dimethylaminopropyl)-3-ethyl carbodiimide hydrochloride (EDC), L-α-phosphatidylethanolamine-N-(7-nitro-2-1,3-benzoxadiazol-4-yl) (NBD-PE), L-α-phosphatidylethanolamine-N-(lissamine rhodamine B sulfonyl) (Rhod-PE), Dulbecco's phosphate-buffered saline (D-PBS), sodium chloride (NaCl), calcium chloride (CaCl.sub.2), fluorescein isothiocyanate (FITC), Triton X-100, thrombin, arachidonic acid (AA), adenosine diphosphate (ADP), plasminogen, fibrinogen (Fg), tris(hydroxymethyl) aminomethane, agar, tPA chromogenic activity assay kit S-2251, acid citrate dextrose (ACD), and ethylene diamine tetraacetic acid (EDTA) were purchased from Sigma-Aldrich (Dorset, UK). Centrifugal concentrators were purchased from Fisher Scientific (Loughborough, UK). Sheep blood was obtained from TCS Biosciences Ltd (Buckingham, UK). The extruder set, polycarbonate membranes and filter supports were purchased from Avanti Polar Lipids Inc. (Alabaster, USA). Thromboxane B2 (TXB2) ELISA kit and 6-keto-PGF1α ELISA kit were purchased from Abcam (Cambridge, UK). Chloroform (CHCl.sub.3), hydrochloric acid, sodium hydroxide and other chemicals were obtained from VWR (Lutterworth, UK).

Preparation and Characterisation of tPA-Loaded, Fibrinogen-Mimicking, Red Blood Cell-Derived Vesicles (tPA-RGD-RBCVs and tPA-cRGD-PEG-RBCVs)

a) Preparation of RBC-Derived Vesicles (RBCVs)

[0239] Briefly, the blood was centrifuged at 3000 rpm for 5 min and then washed with phosphate buffered saline (PBS, pH=7.4) for three times to remove serum. RBCs were re-suspended in 0.2 mM EDTA solution to induce membrane rupture. Subsequently, the cell solution was adjusted to 1×PBS by using 10 × PBS and then centrifuged at 14800 rpm for 7 min at 4° C. to remove the supernatant. The EDTA treatment step was repeated for times until the supernatant was free of red colour. The resulting RBC ghosts were re-suspended in pH 7.4 PBS and then sonicated until a clear and transparent solution of RBCVs was obtained.

[0240] In an alternative method to generate red blood cell ghost from red blood cells, an ice bath hypotonic treatment protocol with a few modifications (Hu et al., Proc Natl Acad Sci USA, 2011, 108, 10980-5) was performed. RBCs were washed three times with ice-cold 1 x PBS and centrifuged for 5 min at 3000 rpm at 4° C. The resulting RBC pellet was treated with 0.25 x PBS and incubated in a shaking ice bath for 20 minutes at 50 rpm. The resulting solution was centrifuged at 12,100 rpm at 4° C. for 8 minutes and the supernatant was removed. The resulting pellet of red blood cell ghost was washed once with 1x PBS and resuspended in the same solution before being stored in the fridge.

b) Preparation of tPA-loaded, Fibrinogen-Mimicking RBCVs (tPA-RGD-RBCVs and tPA-cRGD-PEG-RBCVs)

[0241] RBC ghosts were re-suspended in pH 7.4 PBS buffer, which was then added with tPA at a final concentration of 5 mg mL.sup.-1. The DSPE-PEG-cRGD lipid was first synthesised by an amidation reaction between the free -NH.sub.2 in DSPE-PEG-NH.sub.2 and the free -COOH in cyclo(Arg-Gly-Asp-d-Phe-Val) (cRGD) peptide. DSPE-RGD lipid or DSPE-PEG-cRGD lipid (5.0 × 10.sup.-7 mol) was dissolved in chloroform in a 25-mL round-bottom flask. A lipid film was formed by removal of the organic solvent with rotary evaporation, and then hydrated with 5 mL of tPA-containing RBC ghost solution at 40° C. for 1 h. The mixture solution was sonicated at 4° C. for 30 min, followed by extrusion for times through the 200-nm polycarbonate membrane by using the extruder set. The unencapsulated tPA was removed by ultracentrifugation (Eppendorf, UK) at 4° C. Meanwhile, tPA-loaded RBCVs without RGD coating (tPA-RBCVs) were synthesised by using the same protocol for comparison.

c) Vesicle Characterisation

[0242] The hydrodynamic size and polydispersity index (PDI) of tPA-cRGD-PEG-RBCVs, tPA-RGD-RBCVs and tPA-RBCVs were measured by dynamic light scattering (DLS, Zetasizer Nano S, Malvern, UK). Their morphology was measured by transmission electron microscopy (TEM, JEOL JEM-2100F, Japan). Their zeta potential was measured by zeta potential analyser (ZetaPALS, Brookhaven, USA).

[0243] The nanoparticle tracking analysis (NTA) of 1 mL of RBCVs in pH 7.4 PBS buffer was performed by a NanoSight NS300 equipped with a 532 nm green laser (Malvern, UK).

[0244] The tPA content in the vesicles was determined by the chromogenic substrate S-2251 assay. The encapsulation efficiency (EE) was calculated according to the following equation:

[00001]$\text{EE}\left(\%\right)=\frac{m_l}{m_t}\times 100$

where m.sub.l is the weight of tPA loaded and m.sub.t is the total weight of tPA in the initial loading solution.

Binding Affinity with Activated Platelets

a) Isolation of Platelets

[0245] Platelets were isolated from whole blood with anti-coagulant ACD by differential centrifugation. Briefly, the blood was placed into a 2-mL tube and centrifuged at 200 × g for 15 min. Platelet-rich plasma (PRP) was obtained by removing RBCs, and platelets were then collected by centrifugation of PRP at 800 × g for 15 min. Platelets were resuspended in PBS buffer and counted for experimental use.

b) Preparation of FITC-Labelled tPA

[0246] To evaluate the potential of the fibrinogen-mimicking vesicles for targeted tPA delivery and selective thrombolysis, their specific binding to activated platelets was determined by flow cytometry and confocal laser scanning microscopy (CLSM). For the purpose, tPA was fluorescently labelled by fluorescein isothiocyanate (FITC). Briefly, 1 mL of FITC solution in DMSO (1.0 mg mL.sup.-1) was added dropwise into 2 mL of tPA solution in PBS buffer at pH 8.0 (1.0 mg mL.sup.-1) and stirred at 4° C. in the dark overnight. The mixture was dialysed against pH 7.4 PBS solution for 24 h (MWCO = 3500 Da) and the purified product was stored at 4° C. before use.

c) Binding Affinity of Free Peptides

[0247] Binding affinity of free liner RGD or cRGD peptides to activated platelets can be indicated by a half maximal inhibitory concentration (IC.sub.50) value, which is the peptide concentration required to inhibit fibrinogen mediated platelet aggregation in platelet-rich plasma (PRP) by 50%. PRP was incubated with various concentrations of liner RGD or cRGD peptides in the presence of platelet agonist, thrombin (4 .Math.M) for 30 min under stirring. The platelet aggregation percentage was determined with a GloMax-Multi Microplate Multimode Reader (Promega, USA) by measuring the absorbance at 595 nm. The platelet aggregation percentage was calculated by the following equation:

[00002]$\text{Aggregation}\left(\%\right)=\frac{A_0-A_t}{A_0}\times 100$

where A.sub.o is the initial absorbance at 595 nm, and A.sub.t is the absorbance at 595 nm after incubation.

[0248] The percentage platelet aggregation inhibition was calculated by the following equation:

[00003]$\text{Inhibition}\left(\%\right)=\frac{P{A}_{PBS}-P{A}_S}{P{A}_{PBS}}\times 100$

where PA.sub.PBS is the platelet aggregation in the PRP control in the presence of thrombin and PBS only (no peptides), and PA.sub.s is the platelet aggregation in the PRP sample in the presence of thrombin and peptides.

d) Determination of Binding Affinity with Activated Platelets by Flow Cytometry

[0249] 2 mL platelets (1.0 × 10.sup.8 mL.sup.-1) were seeded in 6-well plates and activated by incubation with 100 .Math.L thrombin (1 U/mL) for at least 20 min. Inactivated platelets or thrombin-activated platelets were incubated with various FITC-labelled tPA-loaded vesicle formulations (equivalent tPA concentration of 0.2 mg mL.sup.-1). Free vesicles, which were not attached to platelets, were removed by centrifugation. Data for 1.0 × 10.sup.4 gated events were collected and analysed using a BD Fortessa II flow cytometer.

e) Determination of Binding Affinity with Activated Platelets by CLSM

[0250] 2 mL platelets (1.0 × 10.sup.8 mL.sup.-1) were seeded in 6-well plates, with a collagen-coated glass coverslip on the bottom of each well. After 30 min, 100 .Math.L thrombin (1 U/mL) was added onto the platelet-adhered coverslips to ensure activation of platelets at least 20 min. Inactivated platelets or thrombin-activated platelets were incubated with various FITC-labelled tPA-loaded vesicle formulations (equivalent tPA concentration of 0.2 mg mL.sup.-1). The platelets were then fixed with 4.0% formaldehyde for 30 min, followed by rinsing with pH 7.4 PBS buffer for three times. Subsequently, the resulting slides were mounted and observed with a Leica SP.sub.5 MP confocal microscope.

In Vitro Drug Release Upon Interaction with Activated Platelets

[0251] Drug release was evaluated in the presence of activated platelets to confirm the tPA release specifically upon interaction with activated platelets. 200 .Math.L of platelets (1.0 × 10.sup.8 mL.sup.-1) were placed into a collagen-coated 96-well microplate and activated by treatment with 20 .Math.L thrombin (1 .Math.M) for 20 min. Inactivated platelets or thrombin-activated platelets were incubated with various tPA-loaded vesicle formulations (equivalent tPA dose of 0.5 mg mL.sup.-1). The released tPA in each well was determined by measuring the absorbance at 405 nm with a GloMax-Multi Microplate Multimode Reader (Promega, USA), as described above in the tPA activity assay. The percentage of tPA release was calculated according to the following equation:

[00004]$\text{tPA}\text{release}\left(\%\right)=\frac{A-A_c}{A_p-A_c}\times 100$

where A is the absorbance of platelet samples after incubation with the tPA-loaded vesicles; A.sub.c is the absorbance of negative control after incubation with the pH 7.4 PBS buffer only; A.sub.p is the absorbance of positive control after incubation with the tPA-loaded vesicles lysed by Triton X-100.

Selective Fibrin Clot Lysis

[0252] Fibrin clot lysis by the tPA-loaded vesicles was measured by an agar plate assay. Briefly, 300 mg agar was dissolved in buffer mixture (15 mL of 0.05 M Tris-HCl buffer at pH 7.2 and 5 mL of 0.025 M CaCl.sub.2 solution). 50 mg Fibrinogen was dissolved in 10 mL of Tris-HCl buffer (0.05 M, pH 7.2). The agar solution was mixed with the fibrinogen solution, and then 10 .Math.L thrombin (4.0 .Math.M) was added under stirring for 1 min. The resulting mixture was spread carefully on a transparent plastic plate and homogeneous gels were obtained at 37° C. after 3h. Four sample wells were created in each plate and 5 .Math.L of plasminogen solution (1 mg mL.sup.-1) was then added into each sample well. PBS buffer, tPA-RBCVs, tPA-RGD-RBCVs, tPA-cRGD-PEG-RBCVs and free tPA (equivalent tPA dose of 1 mg mL.sup.-1) were added into respective sample wells and incubated at 37° C. overnight. The area of the lysed zone in each well was measured to evaluate the degree of fibrin clot lysis.

Selective Blood Clot Lysis in a Halo Blood Clot Model

[0253] Briefly, a clotting mixture (5 mL of buffer containing 66 mM Tris-HCl, 130 mM NaCl and 45 mM CaCl.sub.2; 5 .Math.L of 1 .Math.M thrombin, pH 7.4) was freshly prepared. In a 96-well microplate, 5 .Math.L of this clotting mixture was placed on one side of the well bottom, then 15 .Math.L of whole blood was added on the opposite side of the well bottom. Clotting was initiated by mixing the two drops with a pipette tip in a circular motion to form a homogenous halo shape of blood around the edge of the well bottom, leaving the centre area empty. The plate was covered and incubated at 37° C. for 30 min. After clot formation, 80 .Math.L tPA-RBCVs, tPA-RGD-RBCVs, tPA-cRGD-PEG-RBCVs and free tPA (equivalent tPA dose of 1 mg mL.sup.-1) were added simultaneously into respective wells containing halo clots. The dissolution of the halo clots was determined with a plate reader GloMax-Multi Microplate Multimode Reader (Promega, USA) by measuring absorbance at 495 nm caused by RBCs progressively covering the centre of the well after clot degradation at 37° C. The Negative controls for the assay was obtained by adding 80 .Math.L PBS only to halo thrombi (no tPA), and the positive control was obtained by mixing 15 .Math.L blood and 85 .Math.L of PBS in a well (no halo clots). The percentage of clot dissolution was calculated according to the following equation:

[00005]$\%Clotlysis=\frac{As-A_n}{A_p-A_n}\times 100$

where A.sub.s is the absorbance of well at 495 nm after incubation with samples; A.sub.n is the absorbance of negative control well at 495 nm; A.sub.p is the absorbance of positive control well at 495 nm.

Preparation and Characterisation of ASA-Loaded, Fibrinogen-Mimicking, Red Blood Cell-Derived Vesicles (ASA-RGD-RBCVs and ASA-cRGD-PEG-RBCVs)

a) Preparation of ASA-loaded, Fibrinogen-Mimicking RBCVs (ASA-RGD-RBCVs and ASA-cRGD-PEG-RBCVs)

[0254] RBC ghosts were re-suspended in pH 7.4 PBS buffer, which was then added with ASA at a final concentration of 0.5 mg mL.sup.-1. DSPE-RGD lipid or DSPE-PEG-cRGD lipid (5.0 × 10.sup.-7 mol) was dissolved in chloroform in a 25-mL round-bottom flask. A lipid film was formed by removal of the organic solvent with rotary evaporation, and then hydrated with 5 mL of ASA-containing RBC ghost solution at 40° C. for 1 h. The mixture solution was sonicated at 4° C. for 30 min, followed by extrusion for times through the 200-nm polycarbonate membrane by using the extruder set. The unencapsulated ASA was removed by dialysis against pH 7.4 PBS solution for 6 h (MWCO = 12 kDa). Meanwhile, ASA-loaded RBCVs without RGD coating (ASA-RBCVs) were synthesised by using the same protocol for comparison.

b) Vesicle Characterisation

[0255] The hydrodynamic size and polydispersity index (PDI) of ASA-cRGD-PEG-RBCVs, ASA-RGD-RBCVs and ASA-RBCVs were measured by DLS (Zetasizer Nano S, Malvern, UK). Their zeta potential was measured by zeta potential analyser (ZetaPALS, Brookhaven, USA).

[0256] The absorbance of ASA was measured by UV/Vis spectrophotometry at wavelength of 290 nm. The absorbance value was used to calculate the weight of loaded ASA in vesicles based on the calibration curve obtained for a range of ASA concentrations.

[0257] The encapsulation efficiency (EE) was calculated according to the following equation:

[00006]$\text{EE}\left(\%\right)=\frac{m_l}{m_t}\times 100$

where m.sub.l is the weight of ASA loaded and m.sub.t is the total weight of ASA in the initial loading solution.

Binding Affinity with Activated Platelets

a) Determination of Binding Affinity with Activated Platelets by Flow Cytometry

[0258] 2 mL platelets (1.0 × 10.sup.8 mL.sup.-1) were seeded in 6-well plates and activated by incubation with 100 .Math.L thrombin (1 U/mL) for at least 20 min. Inactivated platelets or thrombin-activated platelets were incubated with various NBD-labelled RBCV formulations (equivalent NBD-PE of 1.0 × 10.sup.-7 mol). Free vesicles, which were not attached to platelets, were removed by centrifugation. Data for 1.0 × 10.sup.4 gated events were collected and analysed using a BD Fortessa II flow cytometer.

c) Determination of Binding Affinity with Activated Platelets by CLSM

[0259] 2 mL platelets (1.0 × 10.sup.8 mL.sup.-1) were seeded in 6-well plates, with a collagen-coated glass coverslip on the bottom of each well. After 30 min, 100 .Math.L thrombin (1 U/mL) was added onto the platelet-adhered coverslips to ensure activation of platelets for at least 20 min. Inactivated platelets or thrombin-activated platelets were incubated with various NBD-labelled RBCV formulations (equivalent NBD-PE of 1.0 × 10.sup.-7 mol). The platelets were then fixed with 4.0% formaldehyde for 30 min, followed by rinsing with pH 7.4 PBS buffer for three times. Subsequently, the resulting slides were mounted and observed with a Leica SP5 MP confocal microscope.

In Vitro Antiplatelet Aggregation Assay

[0260] The final platelet count was adjusted to 2 × 10.sup.8 platelets/mL. 150 .Math.L of PRP was seeded in 96-well plates and was incubated with 50 .Math.L of various formulations: ASA-RBCVs, ASA-RGD-RBCVs, ASA-cRGD-PEG-RBCVs or ASA (equivalent ASA concentration of 0.1 mM) in the presence of 10 .Math.L platelet agonist thrombin in PBS (4 .Math.M), arachidonic acid (AA) in PBS (400 .Math.M) or adenosine diphosphate (ADP) in PBS (15 .Math.M). After incubation under shaking at 37° C. for 40 min, the absorbance at 595 nm was measured by a GloMax-Multi Microplate Multimode Reader (Promega, USA). The aggregation percentage was calculated by the equation 2. From the platelet aggregation percentage and comparison to aggregation of PBS without samples, the inhibition percentage was calculated by the equation 3.

TXB.SUB.2 Assay

[0261] The procedure of the thromboxane B.sub.2 (TXB.sub.2) ELISA Kit (Abcam, UK) was followed to perform the assay. By using the standard wells of the ELISA plates, a standard curve was obtained to identify the TXB.sub.2 level in the plasma treated with various formulations: PBS, ASA-RBCVs, ASA-RGD-RBCVs, ASA-cRGD-PEG-RBCVs or ASA (equivalent ASA concentration of 0.1 mM).

[0262] 2 mL of sheep blood was centrifuged at 200 × g for 15 min to collect the plasma. The 1 mL plasma was added in the tube and 500 .Math.L of a specific formulations was then added. The tube was incubated at 37° C. for 10 min, and then 250 .Math.L of arachidonic acid (AA) in PBS (0.2 mg/mL) was added and incubated for another 30 min. The samples were treated according to the procedure of the TXB.sub.2 ELISA Kit, and the absorbance at 405 nm was measured by a GloMax-Multi Microplate Multimode Reader (Promega, USA). The level of TXB.sub.2 was calculated by using the standard curve.

6-Keto-PGF.SUB.1α Assay

[0263] The procedure of the 6-keto-PGF.sub.1α ELISA Kit (Abcam, UK) was followed to perform the assay. By using the standard wells of the ELISA plates, a standard curve was obtained to identify the 6-keto-PGF.sub.1α level in the plasma treated with various formulations: PBS, ASA-RBCVs, ASA-RGD-RBCVs, ASA-cRGD-PEG-RBCVs or ASA (equivalent ASA concentration of 0.1 mM).

[0264] 2 mL of sheep blood was centrifuged at 200 × g for 15 min to collect the plasma. The 1 mL plasma was added in the tube and 500 .Math.L of a specific formulation was then added. The tube was incubated at 37° C. for 10 min, and then 250 .Math.L of arachidonic acid (AA) in PBS (0.2 mg/mL) was added and incubated for another 30 min. The samples were treated according to the procedure of the 6-keto-PGF.sub.1α ELISA Kit, and the absorbance at 405 nm was measured by a GloMax-Multi Microplate Multimode Reader (Promega, USA). The level of 6-keto-PGF.sub.1α was calculated by using the standard curve.

In Vitro Antithrombotic Assay

[0265] 200 .Math.L of sheep blood was added into a weighted tube. Then, 50 .Math.L specific formulation (PBS, ASA-RBCVs, ASA-RGD-RBCVs, ASA-cRGD-PEG-RBCVs or ASA at an equivalent ASA concentration of 0.1 mM) was added in the tube, and then 10 .Math.L of thrombin in PBS (4 .Math.M) was added in each well and incubated at 37° C. under shaking for 30 min. Finally, the un-clotted blood was removed to obtain the weight of the wet clotted thrombus.

Example 1 - Red Blood Cell-Derived Vesicles (RBCVs) for Targeted Thrombolysis

[0266] The inventors set out to create a novel delivery means for a thrombolytic agent, such as tPA. Inspired from the fibrinogen binding with activated platelets at a thrombus site, as shown in FIG. 1 (top left), the inventors developed novel fibrinogen-mimicking systems based on surface coating of natural red blood cell-derived vesicles (RBCVs) with peptides containing RGD sequences. The fibrinogen-mimicking RBCVs (FM-RBCVs) can be encapsulated with thrombolytic agents, including tPA, the resulting RBCVs having high selectivity and affinity towards activated platelets for targeted delivery of tPA to a thrombus site and triggered release locally for effective thrombolysis without minimised off-target bleeding side effects.

Results & Discussion

[0267] The inventors made novel red blood cell derived vesicles (RBCV) 1, as shown in FIG. 1 (top right), where the RBCV (1) comprises an outer lipid bilayer membrane (6), derived from a red blood cell (7) defining an intravesicular space (8), the RBCV (1) being decorated with a plurality of spaced apart platelet-targeting moieties or ligands (2), such as RGD and cRGD. The RBCVs are able to mimic the action of fibrinogen (9) by the presence of the RGD ligands (2), which functionalise the surface of the RBCV (1). Fibrinogen (9) is capable of binding to GPIIb-IIIa (α.sub.IIbβ.sub.3) integrin (10) on activated platelets (12) through the arginine-glycine-aspartic acid (RGD) motifs located on the two Aα chains (14). This leads to platelet aggregation (13) through the “bridging effect”, which leads to thrombus (5) formation and the occlusion of a blood vessel (4), as shown in FIG. 1 (top left and bottom left). Thus, RBCVs (1) comprising an RGD ligand (2) are capable of binding to activated platelets (12) to enable targeted delivery of the active agent (3) contained within the intravesicular space, for example tPA (3), to activated platelets (12) in the thrombus (5) present in a blood vessel (4), as shown in FIG. 1 (bottom left). After delivery of the active agent (3) to the thrombus (5), the agent may act to lyse the clot (15), removing the thrombus (5), thus enabling the flow of blood cells (7) along the blood vessel (4), as shown in FIG. 1 bottom left.

[0268] The inventors also devised an elegant method to produce RBCVs for targeted thrombus delivery, as shown in FIG. 2. Red blood cells (7) were treated with a hypotonic solution (16) comprising EDTA (17) to substantially remove haemoglobin from the red blood cells (7). After steps of centrifugation, removing the supernatant and repeating EDTA treatment until the solution is free of red colour, the resulting red blood cell ghosts (19) were suspended in a PBS solution (18). tPA (3) was added to the solution and this was added to a dehydrated DSPE-RGD or DSPE-PEG-cRGD lipid (20). The solution was sonicated (21) and then extruded (22) through a 200 nm polycarbonate membrane to produce a filtrate, and unencapsulated tPA was removed by ultracentrifugation of the filtrate producing the tPA (3)-loaded, fibrinogen-mimicking RBCVs (1) (tPA-RGD-RBCVs or tPA-cRGD-PEG-RBCVs).

Platelet Targeting Moiety

[0269] Activation of the GPIIb-IIIa (α.sub.IIbβ.sub.3) integrin is the common pathway involved in platelet aggregation. Normally, α.sub.IIbβ.sub.3 integrins are inactive on the circulating resting platelet surface. However, in the event of thrombus formation, platelets will be in an active state α.sub.IIbβ.sub.3 integrins are abundantly expressed and activated with a conformational change on the platelet membrane. This conformational change of α.sub.IIbβ.sub.3 integrins allows specific binding of activated platelets to fibrinogens through the arginine-glycine-aspartic acid (RGD) motifs located in each of its two Aα chains, which leads to platelet aggregation through the “bridging effect”. Therefore, activated platelets are an ideal target for the selective delivery of thrombolytic agents to thrombi, because the aggregation of activated platelets and abundant active α.sub.IIbβ.sub.3 integrins on the surface of activated platelets are the significant hallmark events in thrombosis.

[0270] One embodiment of RBCVs made is referred to tPA-RBCVs - this includes a red blood cell-derived vesicle (1) encapsulating the active agent, tPA (3).

[0271] The second embodiment is known as tPA-RGD-RBCVs - this includes a red blood cell-derived vesicle (1) decorated with a plurality of RGD ligands and encapsulating the active agent, tPA (3).

[0272] The third embodiment is known as tPA-cRGD-PEG-RBCVs - this includes a red blood cell-derived vesicle (1) decorated with a plurality of cRGD ligands and encapsulating the active agent, tPA (3).

TABLE-US-00001 Mean hydrodynamic sizes measured by dynamic light scattering (DLS), encapsulation efficiencies and zeta potentials of tPA-RBCVs, tPA-RGD-RBCVs and tPA-cRGD-PEG-RBCVs Vesicles DLS size (nm) DLS size after 4 weeks (nm) Encapsulation efficiency (%) Zeta potential (mV) tPA-RBCVs 265.7±5.1 395.1±6.7 24.3±3.2 -30.1±3.9 tPA-RGD-RBCVs 253.7±2.4 282.5±4.3 27.9±2.8 -27.9±4.5 tPA-cRGD-PEG-RBCVs 260.3±3.8 265.1±3.4 29.2±2.5 -35.8±3.7

[0273] As shown in Table 1 and FIG. 3, the DLS sizes of tPA-RGD-RBCVs (253.7 ± 2.4 nm) and tPA-cRGD-PEG-RBCVs (260.3±3.8 nm) were similar to that of tPA-RBCVs (265.7±5.1 nm). This suggests that RGD or cRGD conjugation onto the vesicle surface did not cause a significant change in particle size. There was no significant size change for tPA-RGD-RBCVs or tPA-cRGD-PEG-RBCVs after 4 weeks of storage at 4° C., but a considerable size increase was detected for tPA-RBCVs. The TEM image confirmed that tPA-RGD-RBCVs were spherical in shape (see FIG. 4) and that the TEM particle size was consistent with the DLS result. The encapsulation efficiencies of tPA-RGD-RBCVs (27.9 ± 2.8%) and tPA-cRGD-PEG-RBCVs (29.2±2.5%) were similar to that for tPA-RBCVs (24.3 ± 3.2%). Zeta potential values of the tPA-RBCVs, tPA-RGD-RBCVs and tPA-cRGD-PEG-RBCVs were in the range of -35.8±3.7 to -27.9 ± 4.5 mV.

[0274] FIG. 5 shows the concentration-dependent effect of free peptides (RGD and cRGD) on the inhibition of platelet aggregation. Free cRGD caused 50% inhibition of platelet aggregation at a much lower peptide concentration (0.46±2.5 mM) compared to free linear RGD (381.7±6.3 mM). This confirms that both the α.sub.IIbβ.sub.3-specific peptides (linear RGD and cRGD) can kinetically outcompete with natural ligand fibrinogen in binding to active α.sub.IIbβ.sub.3 and hence prevent fibrinogen-mediated platelet aggregation. The higher affinity peptide can outcompete fibrinogen at low peptide concentrations while the lower affinity peptide requires much higher concentrations to gain the kinetic advantage. Hence, affinity was directly correlated with the IC.sub.50 value and the value for cRGD was found to be considerably lower than that for linear RGD.

[0275] As shown in FIG. 6A, resting platelets treated with FITC-labelled tPA-RBCVs, tPA-RGD-RBCVs and tPA-cRGD-PEG-RBCVs displayed a similar level of fluorescence intensity, which showed only a small increase compared to that of the control resting platelets treated with pH 7.4 PBS buffer alone (FIG. 6B). This suggests that the three vesicle formulations had a low level of attachment to resting platelets.

[0276] FIG. 7A shows that, even when platelets were activated by thrombin, only a slight increase in fluorescence intensity was observed for activated platelets treated with FITC-labelled tPA-RBCVs. However, FITC-labelled tPA-RGD-RBCVs and tPA-cRGD-PEG-RBCVs bound more avidly to activated platelets. FIG. 7B shows that the fluorescence intensity in activated platelets after treated with FITC-labelled tPA-RGD-RBCVs was a little lower than the treatment with FITC-labelled tPA-cRGD-PEG-RBCVs, but had an over 4-fold increase as compared to the treatment with FITC-labelled tPA-RBCVs. These results suggest that RGD peptides, especially cRGD peptides, can efficiently facilitate the specific binding of the tPA-loaded fibrinogen-mimicking RBC-derived vesicles to activated platelets.

[0277] As shown in FIG. 8A, the resting platelets treated with FITC-labelled tPA-RBCVs, tPA-RGD-RBCVs and tPA-cRGD-PEG-RBCVs only showed very weak green fluorescence. However, tPA-RGD-RBCVs and tPA-cRGD-PEG-RBCVs displayed significantly enhanced staining of activated platelets as compared to FITC-labelled tPA-RBCVs without RGD peptide coating. Analysis of the MFI in those images (FIG. 8B) suggested that tPA-RGD-RBCVs had a little lower binding affinity to activated platelets than tPA-cRGD-PEG-RBCVs, but showed an approximately 4-fold enhancement in binding affinity to activated platelets as compared to tPA-RBCVs. This is in good agreement with the flow cytometry results shown in FIG. 7B, further consolidating the ability of tPA-RGD-RBCVs and tPA-cRGD-PEG-RBCVs to efficiently facilitate targeted drug delivery to activated platelets at a thrombus site.

[0278] The amount of RGD arms can be controlled to maximise the binding affinity. As shown in FIG. 9, there was a sharp increase in the binding affinity of tPA-RGD-RBCVs to activated platelets when the RGD arm density increased from 1.0 × 10.sup.4 to 4.0 × 10.sup.4 arms per vesicle, beyond which the influence of RGD arm coating density was insignificant. This suggests that 4.0 × 10.sup.4 arms per vesicle was an important threshold value for RGD to achieve maximal targeting to activated platelets.

[0279] As shown in FIG. 10, upon binding of tPA-RGD-RBCVs or tPA-cRGD-PEG-RBCVs to activated platelets, around 70% of the entrapped tPA was released within 2 h, and the release of tPA continued to increase to about 85% in 6 h, which was considerably higher than the tPA release from tPA-RBCVs (about 30% within 6 h). By comparison, when incubated with resting platelets, tPA-RGD-RBCVs released only around 30% of the entrapped tPA in 6 h. These results indicate that drug release from the fibrinogen-mimicking vesicles was preferably induced by the interaction of RGD peptides on the surface of tPA-RGD-RBCVs with the α.sub.IIbβ.sub.3 on the surface of activated platelets.

[0280] A fluorescence resonance energy transfer (FRET) assay was carried out to examine if the fibrinogen-mimicking RBCVs released drug payload via lipid membrane destabilisation involving membrane fusion between vesicles and activated platelets. As shown in FIG. 11, the NBD fluorescence was enhanced rapidly and considerably upon incubation of the RGD-RBCVs or cRGD-PEG-RBCVs nanovesicles with activated platelets. This was attributed to the decreased surface density of the donor NBD and the increased distance between the donor NBD and the acceptor Rhod resulting from the fusion between the RBCV membrane and the activated platelet membrane. By contrast, incubation RGD-RBCVs with resting platelets caused minimal membrane fusion and a consequently negligible increase in the NBD fluorescence. Inhibition of α.sub.IIbβ.sub.3 integrins on activated platelets with eptifibatide caused a negligible increase in the NBD fluorescence, confirming that this membrane fusion was induced by specific interaction between RGD peptide arms and α.sub.IIbβ.sub.3 integrins.

[0281] The fibrinolytic activity of tPA-RGD-RBCVs and tPA-cRGD-PEG-RBCVs was evaluated in the presence of activated platelets to confirm the selective fibrin lysis. As shown in FIG. 12A, no significant change of the fibrin clot was observed after incubation with PBS buffer, suggesting that PBS buffer did not break up the fibrin clot. Only a very small lysis ring was observed after incubation with tPA-RBCVs, which might be due to the marginal tPA release from tPA-RBCVs. Interestingly, upon treatment of the fibrin clot with tPA-RGD-RBCVs or tPA-cRGD-PEG-RBCVs, a clear clot lysis zone around the sample well was observed, showing a similar size of the lysis zone caused by free tPA. The areas of the fibrin clot lysis rings were calculated in FIG. 12B. It was shown that, in the presence of activated platelets, tPA-RGD-RBCVs and tPA-cRGD-PEG-RBCVs caused significant fibrin lysis and the areas of lysis ring were 1.13 ± 0.12 cm.sup.2 and 1.15 ± 0.18 cm.sup.2, respectively, which were similar to that of free tPA (1.18 ± 0.17 cm.sup.2) but significantly higher than that of tPA-RBCVs (0.14 ± 0.07 cm.sup.2) and PBS buffer (0.02 ± 0.01 cm.sup.2) . These results suggest that the tPA-loaded fibrinogen-mimicking vesicles can cause efficient fibrin lysis in the presence of activated platelets.

[0282] The representative photographs of Halo blood clots after treatment with PBS buffer, blank RBCVs, tPA-RBCVs, tPA-RGD-RBCVs, tPA-cRGD-PEG-RBCVs and free tPA, respectively are displayed in FIG. 13A. It was clearly found that the clots were dissolved and consequently RBCs were released after incubation with tPA-RGD-RBCVs or tPA-cRGD-PEG-RBCVs, similar to the clots treated with free tPA. By contrast, the remnant clots were visible to the naked eye after treatment with tPA-RBCVs, blank RBCVs or PBS buffer. As shown in FIG. 13B, the clot-lytic activity of tPA-RGD-RBCVs or tPA-cRGD-PEG-RBCVs was a bit lower than free tPA at the initial timepoints, which could be ascribed to the time required for tPA release from the fibrinogen-mimicking vesicles. However, after 45 min of treatment, both tPA-RGD-RBCVs and tPA-cRGD-PEG-RBCVs, like free tPA, caused the complete blood clot lysis. By comparison, the degree of blood clot dissolution by tPA-RBCVs (without RGD coating) was found to be about 50%, but the degree of blood clot lysis for blank RBCVs and PBS buffer were both less than 5% after incubation for 45 min. All these results demonstrated that these tPA-loaded fibrinogen-mimicking vesicles had considerably higher thrombolytic activity compared to the non RGD-coated vesicles.

Example 2 - Red Blood Cell-Derived Vesicles (RBCVs) for Antithrombotic Study

[0283] The inventors set out to create a novel delivery means for an anti-platelet agent, such as acetylsalicylic acid (ASA, aspirin), by using the fibrinogen-mimicking RBCVs. The fibrinogen-mimicking RBCVs can be encapsulated with anti-platelet agents, including ASA. The resulting ASA-loaded, fibrinogen-mimicking RBCVs have high selectivity and affinity towards activated platelets for targeted delivery and triggered release of ASA for effective antithrombotic effect.

Results & Discussion

[0284] The inventors used the same method, as described above, to produce fibrinogen-mimicking RBCVs and encapsulate ASA (FIGS. 14 and 15) for antithrombotic study as mentioned above.

[0285] One embodiment of RBCVs is known as ASA-RBCVs - this includes a red blood cell-derived vesicle encapsulating the active agent, ASA.

[0286] The second embodiment is known as ASA-RGD-RBCVs - this includes a red blood cell-derived vesicle decorated with a plurality of RGD ligands and encapsulating the active agent, ASA.

[0287] The third embodiment is known as ASA-cRGD-PEG-RBCVs - this includes a red blood cell-derived vesicle decorated with a plurality of cRGD ligands and encapsulating the active agent, ASA.

TABLE-US-00002 Mean hydrodynamic sizes measured by dynamic light scattering (DLS), encapsulation efficiencies and zeta potentials of ASA-RBCVs, ASA-RGD-RBCVs and ASA-cRGD-PEG-RBCVs Vesicles DLS size (nm) DLS size after 4 weeks (nm) Encapsulation efficiency (%) Zeta potential (mV) ASA-RBCVs 255.8±4.9 416.1±7.1 16.9±5.1 -20.1±5.4 ASA-RGD-RBCVs 256.6±2.7 281.8±5.9 17.7±4.8 -22.7±3.2 ASA-cRGD-PEG-RBCVs 258.1±4.2 264.9±3.1 21.4±3.7 -29.3±4.4

[0288] As shown in Table 2, the DLS size of ASA-RGD-RBCVs (256.6±2.7 nm) and ASA-cRGD-PEG-RBCVs (258.1±4.2 nm) were similar to that of ASA-RBCVs (255.8±4.9 nm). This suggests that RGD or cRGD conjugation onto the vesicle surface did not cause a significant change in particle size. There was no significant size change for ASA-RGD-RBCVs and ASA-cRGD-PEG-RBCVs after 4 weeks of storage at 4° C., but a considerable size increase was detected for ASA-RBCVs. The encapsulation efficiency of the ASA-RBCVs, ASA-RGD-RBCVs and ASA-cRGD-PEG-RBCVs were in the range of 16.9±5.1% to 21.4±3.7%. Zeta potential values of the ASA-RBCVs, ASA-RGD-RBCVs and ASA-cRGD-PEG-RBCVs were in the range of -29.3±4.4 to -20.1±5.4 mV.

[0289] As shown in FIG. 16A, resting platelets treated with NBD-labelled RBCVs, RGD-RBCVs and cRGD-PEG-RBCVs displayed a similar level of fluorescence intensity, which showed only a small increase compared to that of the control resting platelets treated with pH 7.4 PBS buffer alone (FIG. 16B). This suggests that the three vesicle formulations had a low level of attachment to resting platelets.

[0290] FIG. 17A shows that, even when platelets were activated by thrombin, only a slight increase in fluorescence intensity was observed for activated platelets treated with NBD-labelled RBCVs. However, NBD-labelled RGD-RBCVs and cRGD-PEG-RBCVs bound more avidly to activated platelets. FIG. 17B shows that the fluorescence intensity in activated platelets after treated with NBD-labelled RGD-RBCVs was a little lower than the treatment with NBD-labelled cRGD-PEG-RBCVs, but had an over 4-fold increase as compared to the treatment with NBD-labelled RBCVs. These results suggest that RGD peptides, especially cRGD peptides, can efficiently facilitate the specific binding of the fibrinogen-mimicking RBC-derived vesicles to activated platelets.

[0291] As shown in FIG. 18A, the resting platelets treated with NBD-labelled RBCVs, RGD-RBCVs and cRGD-PEG-RBCVs only showed very weak green fluorescence. However, RGD-RBCVs and cRGD-PEG-RBCVs displayed significantly enhanced staining of activated platelets as compared to RBCVs without RGD peptide coating. Analysis of the MFI in those images (FIG. 18B) suggested that RGD-RBCVs had a slightly lower binding affinity to activated platelets than that of cRGD-PEG-RBCVs, but showed an approximately 4-fold enhancement in binding affinity to activated platelets as compared to RBCVs. This is in good agreement with the flow cytometry results shown in FIG. 17B, further consolidating the ability of RGD-RBCVs and cRGD-PEG-RBCVs to efficiently facilitate binding to activated platelets.

[0292] The in vitro activities of ASA-RGD-RBCVs and ASA-cRGD-PEG-RBCVs inhibiting adenosine diphosphate (ADP)-, arachidonic acid (AA)- or thrombin-induced platelet aggregation were determined. As shown in FIG. 19A, the percentages of inhibition of ADP-induced platelet aggregation by ASA-RBCVs, ASA-RGD-RBCVs, ASA-cRGD-PEG-RBCVs and ASA (equivalent ASA concentration of 0.1 mM) were 16.31±2.48%, 21.76±5.19%, 29.41±3.35% and 27.56±2.71%, respectively. This suggests that for ADP-induced platelet aggregation ASA-RGD-RBCVs and ASA-cRGD-PEG-RBCVs displayed a similar level of inhibition to that of free ASA, but higher than that of ASA-RBCVs without RGD. As shown in FIG. 19B, the percentages of inhibition of AA-induced platelet aggregation by ASA-RBCVs, ASA-RGD-RBCVs, ASA-cRGD-PEG-RBCVs and ASA (equivalent ASA concentration of 0.1 mM) were 11.36±4.57%, 22.41±3.42%, 23.89±2.43% and 23.81±4.26%, respectively. This suggests that for AA-induced platelet aggregation ASA-RGD-RBCVs and ASA-cRGD-PEG-RBCVs displayed a similar level of inhibition to that of free ASA, but higher than that of ASA-RBCVs without RGD. As shown in FIG. 19C, the percentages of inhibition of thrombin-induced platelet aggregation by ASA-RBCVs, ASA-RGD-RBCVs, ASA-cRGD-PEG-RBCVs and ASA (equivalent ASA concentration of 0.1 mM) were 12.24±3.41%, 13.71±3.31%, 21.84±3.22% and 20.47±1.04%, respectively. This indicates that for thrombin-induced platelet aggregation ASA-cRGD-PEG-RBCVs displayed a similar level of inhibition to that of free ASA, but higher than that of ASA-RBCVs and ASA-RGD-RBCVs.

[0293] ASA's ability to suppress the production of prostaglandins and thromboxanes is attributed to its irreversible inactivation of the cyclooxygenase (COX) enzyme (FIG. 20), which results in the prevention of clotting. To evidence the inhibition of COX, the levels of thromboxane B2 (TXB2) and 6-keto-PGF1α were measured according to the manufacturer's guidance using the TXB2 ELISA kit and 6-keto-PGF1α ELISA kit (Abcam), respectively. As shown in FIGS. 21A and 21B, ASA-RGD-RBCVs and ASA-cRGD-PEG-RBCVs significantly decreased the levels of both TXB2 and 6-keto-PGF1α as compared to ASA-RBCVs without RGD, confirming the efficient inhibition of the COX-mediated AA metabolism by the ASA-loaded fibrinogen-mimicking RBC-derived vesicles.

[0294] The in vitro antithrombotic activities of ASA-RBCVs, ASA-RGD-RBCVs, ASA-cRGD-PEG-RBCVs and ASA were evaluated by comparing the thrombus weight. As shown in FIGS. 22A and 22B, ASA-RBCVs, ASA-RGD-RBCVs, ASA-cRGD-PEG-RBCVs and ASA exhibited varied antithrombotic activities, with the thrombus weight of 94.60±4.69, 75.61±9.01, 53.87±8.52 and 49.27±10.73 mg, respectively. This indicates that ASA-loaded fibrinogen-mimicking RBC-derived vesicles showed the significantly higher antithrombotic efficiency than that of ASA-RBCVs without RGD. The high inhibition of clotting could result from the selective delivery and the consequent triggered release of ASA due to the presence of RGD for ASA-RGD-RBCVs and the presence of cRGD for ASA-cRGD-PEG-RBCVs. The stronger binding affinity of cRGD suggested a higher antithrombotic ability of ASA-cRGD-PEG-RBCVs than ASA-RGD-RBCVs.

Conclusions

[0295] The inventors have, for the first time, described the use of RGD-coated RBCVs (linear and cyclic) as activated-platelet-sensitive nanocarriers for thrombus-targeted delivery of thrombolytic drugs, including tPA, for safer and more effective thrombolytic therapy, or as anti-platelet agents that decrease platelet aggregation and inhibit thrombus formation.

[0296] The naturally derived, camouflaged nanovesicles are biocompatible, biodegradable and non-immunogenic, and can achieve a very long circulation in the bloodstream. The RGD on the RBCV surface can enable a very high selectivity and affinity binding with the α.sub.IIbβ.sub.3 integrin overexpressed on activated platelets at a blood clot site, thus leading to the efficient, triggered release of thrombolytic agents locally and the consequent targeted thrombolysis.

[0297] In summary, therefore, the RBCVs enable selective delivery of thrombolytics to a blood clot site and controlled release locally. This targeted delivery enhances clot disrupting efficacy, limit drug dose and attenuate off-target bleeding side effects. In addition, the RBCVs enable selective delivery of anti-platelet agents to activated platelets and controlled release locally, which leads the inhibition of platelet aggregation, enhanced antithrombotic efficacy, and reduced drug dose. The RBCV enables more patients with thrombotic diseases to receive safer and more effective treatment. In addition, the RBCVs provide a support surface for anchorage of controllable amounts of RGDs such as linear RGD or cRGD. The resulting multi-arm nanovesicles have superior selectivity and strong binding with activated platelets.

[0298] The use of the RBCVs provide for a triggered release mechanism. Before reaching a thrombus, RBCVs protect thrombolytic agents in the blood circulation, leading to considerably improved stability and prolonged half-life and temporarily suppress thrombolytic activity, leading to reduced haemorrhagic side effects. Upon selective binding to activated platelets, the RBCVs fuse with the activated platelet membrane, leading to rapid and efficient release of thrombolytics. This is favourable for treatment of acute events including but not limited to ischemic stroke which requires immediate drug action. It may also be favourable for lysis of both fibrin-rich (responsive to tPA) and platelet-rich blood clots (resistant to tPA), thus with broader potential clinical applications.

[0299] The novel RBCVs advantageously: [0300] (i) encapsulate and protect drugs in the bloodstream with considerably improved stability and elongated half-life, [0301] (ii) temporarily suppress thrombolytic activity of thrombolytics in the bloodstream and thus reduce the risk of systemic bleeding, [0302] (iii) target thrombolytic drugs to blood clots and thus improve its efficacy without increasing off-targets, [0303] (iv) selectively bind to activated platelets, which can cause efficient and rapid controlled drug release locally as a result of membrane fusion; and [0304] (v) enhance penetration of drugs into clots and thus lead to efficient recanalisation; [0305] (vi) enable targeted delivery of antiplatelet drugs to activated platelets, which can cause efficient and rapid controlled drug release locally, thus decreasing platelet aggregation and inhibiting thrombus formation.