Methods for the preparation of a pharmaceutical-vesicle formulation and associated products and uses

Abstract

The invention relates to methods for the preparation of a pharmaceutical-vesicle formulation comprising steps of: preparing and processing vesicle components and a pharmaceutical agent to entrap the pharmaceutical agent in the vesicle and form a pharmaceutical-vesicle formulation, wherein the pharmaceutical-vesicle formulation is reconstituted in a known quantity of the pharmaceutical agent dissolved in a pharmaceutically-acceptable carrier to provide a biphasic pharmaceutical-vesicle formulation. The invention also relates to the associated pharmaceutical-vesicle formulations, pharmaceutical kits and uses as a medicament, in particular for the prevention or treatment of infection by bacteria such as Burkholderia pseudomallei and Francisella tularensis, and viruses such as Venezuelan Equine Encephalitis Virus (VEEV).

Claims

1. A method for the preparation of a pharmaceutical-vesicle formulation, the method comprising the steps of: a) heating vesicle components comprising monopalmitoyl glycerol, cholesterol and dicetyl phosphate at a temperature in the range of 50° C. to 150° C.; b) dissolving a pharmaceutical agent in a pharmaceutically-acceptable carrier and heating the resultant pharmaceutical agent-carrier mixture at a range of 30-99° C.; adding the pharmaceutical agent-carrier mixture to the vesicle components to provide a formulation mixture; and, d) processing the formulation mixture to form a pharmaceutical-vesicle formulation, whereby the pharmaceutical agent and carrier is entrapped within a plurality of vesicles wherein the pharmaceutical agent is an orally-available antibiotic, wherein the pharmaceutical-vesicle formulation is reconstituted in a known quantity of the pharmaceutical agent dissolved in a pharmaceutically-acceptable carrier to provide a biphasic pharmaceutical-vesicle formulation, wherein the vesicles are modified with i) glucosamine and/or ii) transferrin, wherein, for i), palmitic acid is covalently linked to the glucosamine to provide palmitoylated glucosamine, and wherein, for ii), cholesterol is modified by replacing a portion of the cholesterol by, cholesterol-PEG-malemide, and the vesicle components further comprise thiolated transferrin, or any variant which is at least 70% identical to transferrin wild-type or base sequence, or any fragment which is any amino acid portion which retains the desired properties of transferrin wild-type or base sequence, such that a covalent bond forms between the thiolated transferrin, or variant or fragment, and the modified cholesterol.

2. The method according to claim 1, wherein the monopalmitoyl glycerol, cholesterol and dicetyl phosphate are provided in a 5:4:1 molar ratio respectively.

3. The method according to claim 1, wherein sodium deoxycholate is provided in the vesicle components and/or when dissolving a pharmaceutical agent in a pharmaceutically-acceptable carrier.

4. The method according to claim 1, wherein the pharmaceutical agent-carrier mixture and vesicle components are provided in a respective ratio of 3:1.

5. The method according to claim 1, wherein the pharmaceutical-vesicle formulation undergoes at least one freeze-drying and thawing cycle following processing.

6. The method according to claim 1, wherein the antibacterial agent is selected from a group consisting of fluoroquinolones and tetracyclines.

7. The method according to claim 1, wherein the antibacterial agent is selected from a group consisting of levofloxacin, ciprofloxacin and doxycycline.

8. The method according to claim 1, wherein the antibacterial agent is at a concentration of 30 mg/ml.

9. A pharmaceutical-vesicle formulation prepared by the method of claim 1.

10. The pharmaceutical-vesicle formulation according to claim 9, wherein the monopalmitoyl glycerol, cholesterol and dicetyl phosphate are present in a 5:4:1 molar ratio respectively.

11. The pharmaceutical-vesicle formulation according to claim 9, wherein the vesicle is of a size of 50-4000 nm.

12. The pharmaceutical-vesicle formulation according to claim 9, wherein the antibacterial agent is selected from a group consisting of fluoroquinolones and tetracyclines.

13. The pharmaceutical-vesicle formulation according to claim 12, wherein the antibacterial agent is selected from a group consisting of levofloxacin, ciprofloxacin and doxycycline.

14. The pharmaceutical-vesicle formulation according to claim 9, wherein the antibacterial agent is present at concentration of 30 mg/ml.

15. A pharmaceutical kit comprising the pharmaceutical-vesicle formulation of claim 9.

16. The method according to claim 1, wherein, for ii), a ratio of cholesterol to cholesterol-PEG-malemide is 4:1.

17. The pharmaceutical-vesicle formulation according to claim 9, wherein, for ii), a ratio of cholesterol to cholesterol-PEG-malemide is 4:1.

18. The pharmaceutical-vesicle formulation according to claim 9, wherein sodium deoxycholate is provided in the vesicle components and/or when dissolving a pharmaceutical agent in a pharmaceutically-acceptable carrier.

19. The pharmaceutical-vesicle formulation according to claim 9, wherein the pharmaceutical agent-carrier mixture and vesicle components are provided in a respective ratio of 3:1.

Description

BRIEF DESCRIPTION OF THE FIGURES

(1) FIG. 1 shows graphs demonstrating the stability of biphasic formulations when stored at a range of temperatures in liquid or freeze-dried format;

(2) FIG. 2 shows graphs of in vivo assessment of antimicrobial function of levofloxacin formulated in NISVs in the Galleria mellonella model;

(3) FIG. 3 shows graphs of repeat dose tolerability studies for biphasic NISV and bilosome formulations of antibiotic:body weight changes;

(4) FIG. 4 shows a graph demonstrating weight change in mice control groups for a repeat safety/toxicology study;

(5) FIG. 5 shows a graph demonstrating weight change in mice following daily dosing with levofloxacin formulations in a repeat safety/toxicology study;

(6) FIG. 6 shows a graph demonstrating weight change in mice following daily dosing with deoxycycline formulations in a repeat safety/toxicology study;

(7) FIG. 7 shows a graph demonstrating weigh change in mice following daily dosing with ciprofloxacin formulations in a repeat safety/toxicology study;

(8) FIG. 8 shows graphs of broad microbiome analysis following treatment with antibiotic formulations;

(9) FIG. 9 shows graphs of the modulation of dendritic cell activation in the presence of NISVs;

(10) FIG. 10 shows graphs of the effect of bilosome-levofloxacin compared to free levofloxacin in an aerosol model of B. pseudomallei;

(11) FIG. 11 shows graphs of the effect of bilosome-doxycycline compared to free doxycycline in an aerosol model of B. pseudomallei;

(12) FIG. 12 shows graphs of animal weight data following exposure to an aerosol challenge of B. pseudomallei and treatment with bilosome-levofloxacin or free levofloxacin;

(13) FIG. 13 shows a graph of the effect of bilosome-levofloxacin compared to free levofloxacin in an aerosol model of F. tularensis;

(14) FIG. 14 shows graphs of rabbit IgG antibody deposition in mouse tissues;

(15) FIG. 15 shows graphs of viral load in mice organs following treatment with different formulations of Hu1A3B-7 VEEV mAb; and

(16) FIG. 16 shows a graph of survival in mice in an aerosol model of VEEV following treatment with different Hu1A3B-7 VEEV mAb.

DETAILED DESCRIPTION

(17) The method of the invention provides a method for the preparation of a pharmaceutical-vesicle formulation, the method comprising the steps of: a) heating vesicle components comprising monopalmitoyl glycerol, cholesterol and dicetyl phosphate at a temperature in the range of 50° C. to 150° C.; b) dissolving a pharmaceutical agent in a pharmaceutically-acceptable carrier and heating the resultant pharmaceutical agent-carrier mixture at a range of 30-99° C.; c) adding the pharmaceutical agent-carrier mixture to the vesicle components to provide a formulation mixture; and d) processing the formulation mixture to form a pharmaceutical-vesicle formulation, whereby the pharmaceutical agent and carrier is entrapped within a plurality of vesicles; wherein the pharmaceutical-vesicle formulation is reconstituted in a known quantity of the pharmaceutical agent dissolved in a pharmaceutically-acceptable carrier to provide a biphasic pharmaceutical-vesicle formulation.

(18) Various studies have been carried out to demonstrate, in particular, that the method of the invention provides a delivery system, in particular an oral delivery system, which improves the delivery and efficacy of orally-availably pharmaceutical agents, including orally-available antibiotics. Furthermore, the method provides a delivery system with an unexpected and highly significant benefit in protection against pharmaceutical side effects, for example antibiotic-induced weight loss. In particular, the inventors have shown efficacy of the present invention against B. pseudomallei, F. tularensis and VEEV.

(19) Material and Methods

(20) Antibiotics

(21) The antibiotics levofloxacin ≥98.0% (HPLC), doxycycline hyclate ≥98.0% (HPLC) and ciprofloxacin (Sigma-Aldrich) were obtained in dry powder form and used in all studies. For making the bilosome formulations and dosing as unformulated drugs, the antibiotics were suspended in PBS at pre-determined concentrations.

(22) VEEV Monoclonal Antibody Synthesis

(23) Humanised anti-VEEV monoclonal antibody was produced as previously described (S. A. Goodchild, L. M. O'Brien, J. Steven, M. R. Muller, O. J. Lanning, C. H, Logue, R. V. D'Elia, R. J. Phillpotts, S. D. Perkins. A humanised murine monoclonal antibody with broad serogroup specificity protects mice from challenge with Venezuelan equine encephalitis virus. Antiviral Res., 90(2011), 1-8). Briefly, CHO DG44 cells transfected with vector expressing Hu1A3B-7 were cultured in IMDM media supplemented with 10% (v/v) FBS, 1% antimycotic, 25 mg Gentamycin, 1 mM sodium pyruvate, 2 mM glutamine, 1% (v/v) pen/strep, 1% (v/v) non-essential and 1% (v/v) essential amino acids (Gibco) and 10 mM methotrexate (Sigma). Humanised antibody was purified using protein A affinity chromatography using Prosep-A (Millipore) and dialysed into PBS.

(24) Bacteria

(25) The clinical isolate Burkholderia pseudomallei K96423 was used for in vitro and in vivo studies. Bacteria were grown in Luria broth at 37° C. on a rotary platform for aerosol challenges and enumerated on L-Agar plates.

(26) F. tularensis strain LVS was derived from a vaccine ampoule stored at −20° C. in DSTL's culture collection. Bacteria were cultured overnight at 37° C. on supplemented blood cysteine glucose agar (BCGA), harvested into PBS and diluted to obtain an OD.sub.600 of between 0.15-0.20 (1×10.sup.9 CFU/ml). Bacterial numbers for challenge were determined on agar following serial dilution (1:10) of samples.

(27) Viral Preparations

(28) Tissue culture: The L929 (murine fibroblast) and Vero (simian kidney) cell lines (European Collection of Animal Cell Cultures, UK) were propagated by standard methods using the recommended culture media. For experimental purposes, cells were maintained in Leibovitz L-15 media supplemented with 2% (v/v) foetal calf serum, 2 mM I-glutamine, 50 U/mI penicillin and 50 μg/ml streptomycin (L15MM). Media and supplements were all supplied by Sigma-Aldrich (UK).

(29) Stocks: To prepare virulent virus stocks of VEEV strain TrD, suckling mice were infected intracerebrally with ˜103 pfu virus. Infected brains were harvested at 24 hours, prepared as 10% tissue suspensions in L15MM and clarified by centrifugation at 10,000×g for 15 min. The titre was determined by plaque formation under a carboxymethyl cellulose overlay in Vero cells.

(30) Titration of virus: The titre of virus was determined by plaque formation under a 1.5% (w/v) carboxymethyl cellulose overlay in Vero cells. Briefly, cells were seeded into 24-well plates (1×10.sup.5 cells well) and incubated overnight. Media was removed and the cells overlaid with 100 μl of virus diluted in L15MM. After an incubation of 30 min at room-temperature, 1 ml of double-strength L15MM diluted in 3% (w/v) carboxymethyl cellulose was carefully added to each well. Cells were incubated for 72 hrs, fixed with 10% (v/v) formal saline overnight and then stained with 0.1% (w/v) crystal violet. The plaques were counted and the amount of virus was calculated.

(31) Preparation of Monophasic and Biphasic NISV and Bilosome Formulations

(32) Briefly, 1-monopalmitoyl glycerol, cholesterol and dicetyl phosphate in a 5:4:1 molar ratio were combined and heated to 130° C. For the preparation of NISVs, the relevant antibiotic at 30 mg/ml in PBS was added. For bilosomes, the relevant antibiotic at 30 mg/ml with sodium deoxycholate in 0.025 M carbonate buffer was added. Antibiotics were pre-warmed to 60° C. prior to addition to the vesicle constituents. Preparations were subsequently vortexed vigorously for 2 min. Antibiotics were either entrapped as vesicles were formed or after by 5 cycles of freeze-thawing. Non-entrapped antibiotic was removed through centrifugation and the pelleted vesicles re-suspended in the appropriate buffer for monophasic preparation, or the appropriate buffer containing an amount of the appropriate antibiotic at a concentration equivalent to that entrapped within the vesicle to give biphasic preparations (i.e. 50/50 entrapped:free preparation).

(33) NISVs were made by melting 24.8 mg monopalmitoyl glycerol, 8.2 mg dicetylphosphate, and 23.2 mg cholesterol at 130° C., adding 5 ml of warmed PBS and vortexing for 2 minutes. Sonicated vesicles were made by mixing the same vesicle components in 5 ml PBS and heated to 90° C. for 60 minutes and after which were sonicated using a probe sonicator for 4 minutes at 20% maximum capacity. To ensure sterility, all vesicles samples were irradiated using an X225 irradiator at 2.2 Gy per minute until a final dosage of 10 Gy.

(34) N-Palmitoyl Glucosamine (NPG) Synthesis

(35) 86.3 mg of glucosamine was dissolved in 93 μl of triethanolamine and 15 ml of dimethylsulphoxide. This mixture was then added to 283 mg of palmitic acid N-hydroxysuccinimide and dissolved in 4 ml of chloroform. The solution was stirred for 48 hours at room temperature in the dark. The palmitoylated-glucosamine was precipitated out by chilling the mix and adding ice-cold water. The mix was then applied to a vacuum filter to separate the precipitate. Separate water, chloroform and ethanol washes were applied to the palmitoylated-glucosamine before drying at 40° C. for 48 hours.

(36) Human Holo-Transferrin Thiolation

(37) Human holo-transferrin was modified with a thiol by mixing transferrin with a 10-fold molar excess of Traut's reagent at 25° C. for 1 hour. Excess Traut's reagent was removed using a Vivaspin 6 column with a 5 kDa molecular weight cut-off.

(38) Vesicle preparation using microfluidics Vesicles were prepared using the NanoAssemblr™ benchtop (Precision Nanosytems) with a 300 μm staggered herringbone micromixer chip held at a temperature of 60° C. The vesicle components; palmitin, cholesterol and dicetyl-phosphate were solubilised in menthol in 5:4:1 ratio with a total concentration of 5.62 mg/ml and heated to 60° C. Antibodies (either Rabbit IgG, mAb Hu1A3B-7 or Human IgG-FITC) were solubilised in PBS at 2 mg/ml. The aqueous and solvent streams were mixed at a ratio of 3:1, with a combined flow rate of 12 ml/min. Drug encapsulation and vesicle formulation occurs simultaneously within the microfluidic chip. The methanol and unentrapped drug were removed using a Vivaspin20 (Sartorious) spin column with a 300 kDa molecular weight cut off. Concentration of entrapped antibody was determined using standard Bradford's assay. gNISV were prepared in the same manner with the addition of 6.4 mg palmitoylated glucosamine to the vesicle components dissolved in methanol. tNISV were synthesised by including PEG-Maleimide in the NISV mix. Following removal of the methanol and unentrapped cargo the vesicles were incubated with the thiolated transferrin for 1 hour at 25° C. Excess transferrin was then removed using a Vivaspin 20 column with a 100 kDa molecular weight cut off.

(39) Freeze Drying for Vesicle Formulations

(40) Sucrose was added to NISV preparations to a concentration of 100 mM. Preparations were placed at −80° C. at 24 hours prior to being lyophilised in a Christ Epsilon 2-4LD Freeze Dryer and dried at −40° C. for 48 hours with a condenser temperature of −70° C.

(41) Biophysical Characterisation of Bilosomes

(42) Large batches of bilosome preparations were formulated and aliquoted for lyophilisation and storage. Aliquots were recovered as required and resuspended in reverse osmosis water, followed by vigorous agitation. Vesicle size and zeta-potentials were determined using a Malvern Zeta-sizer (Zetasizer 30000HS, Malvern Instruments Ltd., UK). Vesicles were pelleted by centrifugation and HPLC used to determine entrapped antibiotic content.

(43) HPLC Methodologies

(44) HPLC analysis was carried out on an Agilent1290 Infinity Series HPLC, using a C18 column (150 mm×4.6 mm, 5p) maintained at 50° C. The mobile phase for antibiotics (doxycycline, levofloxacin and ciprofloxacin) consisted of 0.02M Na.sub.2HPO.sub.4, pH2 with H.sub.3PO.sub.4 and either acetonitrile or methanol.

(45) Stability Studies

(46) Large batches of each NISV and bilosome preparations were formulated and aliquots placed at −20° C., 4° C., room temperature and 37° C. either as liquid formulation or freeze-dried preparations. Aliquots were recovered at the various time-points and the freeze-dried preparations re-suspended in reverse osmosis water followed vigorous agitation. Vesicle size and zeta-potentials were determined using a Malvern Zeta-sizer (Zetasizer 30000HS, Malvern Instruments Ltd., UK). Vesicles were pelleted by centrifugation and HPLC used to determine entrapped drug content.

(47) PK/PD Studies

(48) Freeze-dried preparations of NISV and bilosome formulated antibiotic (50-100 mg/kg) were resuspended in reverse osmosis water and administered to animals by the intraperitoneal route or by gavage. Other groups of mice received the identical concentration of antibiotic as a solution. In initial experiments blood was obtained from superficial venesection at time points up to 24 hours. In subsequent experiments mice were sacrificed by terminal anaesthesia, blood collected from the heart and lung, liver and spleen harvested. Each tissue was weighed and 4× volume of saline added before homogenization. Samples were de-proteinated and passed through a 0.2 μm PFTE filter.

(49) Minimum Inhibitory Concentrations (MICs)

(50) MICs for antibiotic formulations were determined for B. pseudomallei strain K96243 and F. tularensis SchuS4 using the broth micro dilution method in accordance with the Clinical Laboratory Standards Institute (CLSI) guidelines. Assays were performed in 96 well micro-titre plates in Iso-sensitest broth (BSAC) or Mueller Hinton broth (CLSI), or broth suitable for pathogen requirements, with antibiotic concentrations in the range of 64 mg/L to 0.03 mg/L, and bacteria at a final concentration of approximately 5×10.sup.5 CFU/mL. Following incubation at 37° C. for 24 h the optical density (OD) of the plates was read in an automated plate reader at a wavelength of 590 nm. MICs were determined as the concentration that inhibited >80% of bacterial growth.

(51) Minimum Bactericidal Concentrations (MBCs)

(52) MBCs for antibiotic formulations were determined by plating 100 μL aliquots of the MIC dilutions showing no visible growth onto L-agar plates or BCGA plates in triplicate and incubating at 37° C. for 48 hours. The MBC was recorded as the lowest concentration of antibiotic that killed 99.9% of the bacteria in the original inoculum.

(53) Galleria mellonella Infection

(54) G. mellonella caterpillars were stored in the dark. Caterpillars 0.2-0.3 g in weight were employed in all assays. A 10 ml Hamilton syringe was used to inject 10 ml aliquots of the inoculum (3×10.sup.6 CFU/larvae of F. tularensis LVS) into the hemocoel of each caterpillar via the last left proleg. After infection, caterpillars were incubated in plastic containers. A 10 ml Hamilton syringe was also used to inject antimicrobial agents (levofloxacin 15 mg/kg) and controls (PBS) 24 hours post infection. Caterpillars were considered dead when the displayed no movement in response to touch.

(55) Microbiome and Toxicology Studies

(56) BALB/c mice were allocated into groups of twelve mice, and treated via the oral route (0.1 ml/day at 10 mg/ml concentration) with PBS, empty bilosomes, empty NISVs, antibiotic, NISV-antibiotic or bilosome-antibiotic on days 1-7. Faecal samples were collected three days prior to treatment, 24 hrs after the final dose (day 8), 14 days later (day 22) and 28 days after the cessation of treatment (day 36). Animal weights and condition were recorded daily. On day 8, half the mice in each group were sacrificed and serum, small intestine and large intestine were collected. Serum samples were stored frozen at −80° C., the large and small intestines were cleaned of contents, rolled into pinwheels and fixed in 10% formalin. The remaining mice were weighed and observed for an additional 28 days before they were sacrificed and serum and tissues harvested as described for the mice sacrificed on day 8.

(57) DNA Extraction and bTEFAP®

(58) Genomic DNA was isolated from faecal samples using the PowerSoil® DNA Isolation Kit (Qiagen) following the manufacturer's instructions. As an alternative to the recommended 250 mg of soil, approximately 200 mg of faecal sample was added to the PowerBeads tube to undergo cell lysis. Purified DNA was eluted from the spin filter using 50 uL of solution C6 and stored at −20° C. until PCR amplification.

(59) The 16S universal Eubacterial primers 515F GTGCCAGCMGCCGCGGTAA and 806R GGACTACHVGGGTWTCTAAT were utilized to evaluate the microbial ecology of each sample on the HiSeq 2500 with methods via the bTEFAP® DNA analysis service. Each sample underwent a single-step 30 cycle PCR using HotStarTaq Plus Master Mix Kit (Qiagen, Valencia, Calif.) were used under the following conditions: 94° C. for 3 minutes, followed by 28 cycles of 94° C. for 30 seconds; 53° C. for 40 seconds and 72° C. for 1 minute; after which a final elongation step at 72° C. for 5 minutes was performed. Following PCR, all amplicon products from different samples were mixed in equal concentrations and purified using Agencourt Ampure beads (Agencourt Bioscience Corporation, MA, USA). Samples were sequenced utilizing the IIlumina HiSeq chemistry following manufacturer's protocols.

(60) The Q25 sequence data derived from the sequencing was processed using a proprietary analysis pipeline (www.mrdnalab.com, MR DNA, Shallowater, Tex.). Sequences were depleted of barcodes and primers then short sequences <200 bp were removed, sequences with ambiguous base calls removed, and sequences with homopolymer runs exceeding 6 bp removed. Sequences were then de-noised and chimeras removed. Operational taxonomic units (OTUs) were defined after removal of singleton sequences, clustering at 3% divergence (97% similarity). OTUs were then taxonomically classified using BLASTn against a curated GreenGenes/RDP/NCBI derived database and compiled into each taxonomic level into both “counts” and “percentage” files. Counts files contain the actual number of sequences while the percent files contain the relative (proportion) percentage of sequences within each sample that map to the designated taxonomic classification.

(61) Statistical analysis was performed using a variety of computer packages including XLstat, NCSS 2007, “R” and NCSS 2010. Alpha and beta diversity analysis was conducted as described previously using Qiime (www.qiime.org).

(62) Dendritic Cell Activation

(63) Bone marrow cells were obtained from the femurs of 8-wk-old male BALB/c mice and differentiated into dendritic cells using GM-CSF-enriched complete DMEM (DMEM 10% foetal calf serum, 100 U/mI penicillin, 100 μg/ml streptomycin and 2 mM L-glutamine). This routinely results in 80% CD11c+ cells. BMDCs were then seeded at 1×10.sup.6 cells/well on 24-well tissue culture plates in complete RPMI (RPMI 1640 10% foetal calf serum, 100U/ml penicillin, 100 μg/ml streptomycin and 2 mM L-glutamine) and stimulated with LPS (3 μg/ml) or left unstimulated as controls. Cells were simultaneously treated with NISV produced by the melt method, or the sonication method, or left untreated. Cells were harvested 24 hours later and stained with a 1/100 dilution of anti-CD40-APC, anti-CD80-PerCP and anti-CD86-PE-Cy7 before analysis by flow cytometry. Compensation and PMT voltages were set up using single stained controls. Fluorescent minus one (FMO) control was used in order to identify the correct gating strategy and was used in order to produce a negative control.

(64) Animal Infection Studies

(65) Six to eight week old female BALB/c mice (Charles River, UK) were transferred to a high containment Class III rigid isolator, where they were given unlimited access to food and water and allowed to acclimatise for at least 5 days. Mice were allocated to treatment groups (15 per group) and housed in cages of 5. Mice were challenged with 50-100 CFU (10MLD) of B. pseudomallei K96243 or F. tularensis SchuS4 via the aerosol route in a nose-only exposure system using a computerised delivery platform (Biaera Technologies). A sub-optimal therapy study design was used, in which antibiotic administration was started at 6h post-infection and administered once daily for only 7 days, to test therapeutic efficacy. The antibiotics levofloxacin and doxycycline were delivered daily by the oral route at 50 mg/kg and treatment was continued for 7 days. The treatment groups comprised bilosome-delivered antibiotics or unformulated antibiotics. A subgroup of 5 mice per treatment group was culled at day 3 p.i. to determine bacterial loads in lung, spleen and liver. Mice were challenged with VEEV Trinidad Donkey Strain by aerosol using a Biaeara aerosol device. Mice were checked twice daily and scored for clinical symptoms and mice were weighed daily. Mice reaching a humane end-point, based on a pre-determined set of objective clinical signs, were promptly culled. Survival times were recorded for some mice and others were culled for analysis of tissues at different time points. All procedures and housing complied with the UK Animal (Scientific Procedures) Act (1986).

(66) Titration of Virus in Tissues

(67) The amount of VEEV strain TrD present in mouse tissues (brain, lung, liver and spleen) was determined by titration in Vero cells. Tissues were removed and homogenized in 2 ml L15MM by passing them through a 70 μm nylon cell strainer (BD Falcon). Cell suspension was then diluted serially (1:10) in L15MM. Diluted homogenate (100 μl) was added to wells of a 24-well plate containing confluent monolayers of Vero cells. The cells were incubated for 72 hrs, after which time the monolayers were fixed by the addition of 10% (v/v) formal saline and stained with 0.1% (w/v) crystal violet.

(68) Cytokine Analysis

(69) For cytokine analysis, 200 ml aliquots of cell suspension were centrifuged for 5 min at 2000 rpm. Supernatants were removed for cytokine analysis and stored at −80 C. The levels of cytokine were measured via 23-plexmurine Luminex array (Bio-Rad), used in accordance with the manufacturer's instructions. In addition, a magnetic plate washer (Bio-Rad) was used for wash steps and samples were ultimately fixed using 4% paraformaldehyde in PBS for at least 24 h at 4° C.

(70) Immunohistochemistry

(71) Formalin-fixed Paraffin-embedded (FFPE) tissue samples were used to standardise a protocol to localise the rabbit IgG in the tissue. Tissue sections were deparaffinised and rehydrated using xylene and a graded series of ethanol and water, finishing with TBS. Several methods were used to retrieve the antigen in the 4-micron tissue sections, including enzymatic digestion with PK (DAKO), and High Temperature

(72) Antigen Retrieval (HTAR) with pH6.0 and pH 9.0 buffers (DAKO). The best results were obtained with PK pre-treatment. The enzymatic digestion was stopped (3 washes in TBS) and followed by the incubation with a polyclonal goat anti-rabbit IgG Alexa568 labelled antibody (30 min at RT & 1:100 dilution). After 3 TBS washes, slides were mounted using Vectashield antifade hardset mounting media (Vector laboratories) and studied under fluorescent light in a Nikon Ni-SE microscope.

(73) Similarly, tissue sections were used for immunohistochemical detection of rabbit IgG with light microscopy, using the same antigen retrieval method (PK). The dewaxing/rehydration protocol included a step of endogenous peroxidase quench using a solution of hydrogen peroxide in methanol. A polyclonal biotinylated goat anti-rabbit IgG antibody (Pierce)(30 min at RT diluted 1:200) followed by the avidin-biotin peroxidase complex reaction (Pierce). The reaction was developed using NOVARED (Vector). FFPE sections stained with immunohistochemistry for digital image analysis. The mean intensity was calculated using the image analysis software (Nikon Br-NIS). The maximum intensity would be 100.

(74) Statistical Analysis

(75) A variety of statistical analyses were performed using the program SPSS V21.0 (IBM) or Graphpad PRISM V6.0. Graphs have been constructed using Graphpad PRISM V6.0. Survival data were compared using log rank tests. Continuous data were analysed by parametric analysis (ANOVA, T tests, GLM) when conditions were met (QQ plots to assess Gaussian distribution and Levene's/Bartlett tests for unequal variation) or non-parametric tests (Kruskal-Wallis, Mann-Whitney, Moods) where these criteria were not met. In some cases it was possible for parametric criteria to be attained in the use of transformations such as logarithmic transformation. Contingency tables were used for binary data. Multiple testing corrections for familywise error were performed on individual comparisons with analyses. These included Bonferroni's and Dunn's corrections.

(76) Results

(77) Assessment of Encapsulation Efficiency and Stability

(78) Encapsulation efficiency was determined by removing non-entrapped antibiotic by centrifugation and re-suspending the pellet in an appropriate buffer for immediate analysis via HPLC. The value generated represents the amount of antibiotic that was entrapped into the vesicles. Both monophasic and biphasic formulations were tested for stability, as determined by the retention of the antibiotic in each formulation type over time using HPLC for quantification. The zeta potential and size of the NISV and bilosome samples was also measured.

(79) These analyses determined that the biphasic formulations were far superior to the monophasic formulations and so after the first two time points, only the biphasic formulations were continued in the stability assays. The biphasic formulations of both NISVs and bilosomes were tested both as liquid and as freeze-dried preparations and at a range of temperatures (37° C., room temperature, 4° C. and −20° C.). While optimum stability has been demonstrated by biphasic bilosomes stored freeze-dried, all of the biphasic formulations were relatively stable in terms of antibiotic retention, for the time elapsed in the study (3 months). Representative data for biphasic formulations of levofloxacin and doxycycline are shown are shown in FIG. 1 (percentage of initial entrapped antibiotic over time. Biphasic antibiotic formulations were made as previously described and either left as a liquid solution or subsequently freeze dried (FD) Both formulations were left as a variety of different storage temperatures (37° C., room temperature, 4° C. or −20° C.) and assayed at multiple time points post manufacture (1 week, 1 month, 3 months) to determine the level of antibiotic that has remained internalised in the vesicle. FD formulations were rehydrated at each time point for the assay. Data are presented as a percentage of the initial entrapped antibiotic).

(80) To further exemplify the successful adaption of the bilosome formulation technology to entrap the levofloxacin and doxycycline, the total antibiotic delivered by each formulation in mg/ml, together with percentage of this which was entrapped, is recorded in Table 1. For each formulation, more than 50% of the total antibiotic was entrapped. The mean size of bilosomes with antibiotic cargo was in the range 2700-3400 nm with zeta potentials in the range of −30 to −23, where negative values for zeta potential indicate formulation stability.

(81) TABLE-US-00001 TABLE 1 Total loading, percentage entrapment and zeta potential of bilosome formulations. The values presented in the table are generated from making bilosomes via the melt method with entrapped concentrations determined via HPLC following removal of free antibiotic. Size and zeta potential measurements were determined by a Malvern Zeta-sizer and data are presented as raw values for 3 independent samples. Means and SEM are highlighted. Total Antibiotic Per- antibiotic entrapped centage in form- in form- en- Antibiotics ulation ulations trapped Zeta formulations (mg/ml) (mg/ml) (%) Size potential Bilosome- 19.4 11.4 58.9 2846.0 ± −29.667 ± levofloxacin  124.42   0.31 Bilosome- 17.2  9.2 53.5 3329.33 ± −23.33 ± deoxycycline  85.62  0.29

(82) The stability (determined via Zeta potential) of the bilosome formulations was tested as freeze-dried preparations at a range of temperatures (room temperature, +4° C. and −20° C.) and at a range of time post-manufacture (1 week, 1 month and 3 months) (Table 2). The bilosome zeta potential remained relatively consistent across all time points and all temperatures. There was slight variability in the bilosome doxycycline formulation if kept at room temperature, however all formulations used in these studies were stored at −20° C. and used within a month of manufacture.

(83) TABLE-US-00002 TABLE 2 Stability of bilosomes with time and storage conditions, as measured by zeta potential. Bilosome formulation Storage Levofloxacin Doxycycline time RT 4° C. −20° C. RT 4° C. −20° C. Start −28.1 −19.8 −30.2 −22.6 −32.7 −27.8 Mean −30.22 N/A N/A −23.40 N/A N/A SEM 0.42 1 week −24.40 −28.90 −17.90 −24.00 −28.30 −28.10 −26.10 −31.90 −18.00 −30.00 −27.90 −25.90 −25.90 −31.50 −19.00 −27.30 −26.50 −28.50 Mean −25.47 −30.77 −18.30 −27.10 −27.57 −27.50 SEM 0.17 0.30 0.11 0.55 0.17 0.26 1 month −24.20 −26.50 −29.80 −16.50 −26.70 −22.20 −26.30 −30.00 −30.20 −18.40 −28.20 −22.80 −28.00 −31.00 −32.70 −18.70 −30.00 −26.50 Mean −26.17 −29.17 −30.90 −17.87 −28.30 −23.83 SEM 0.35 0.43 0.29 0.22 0.30 0.43 3 month −28.90 −29.90 −26.00 −35.30 −23.40 −33.60 −29.40 −30.90 −28.50 −36.40 −28.10 −36.40 −29.20 −32.60 −27.90 −37.00 −29.00 −35.00 Mean −29.17 −31.13 −27.47 −36.23 −26.83 −35.00 SEM 0.05 0.25 0.24 0.16 0.55 0.26 Bilosome antibiotic formulations were made as previously described and subsequently freeze dried (FD). Formulations were left at a variety of different storage temperature (37° C., Room Temperature, 4° C. or −20° C.) and assayed at multiple time points post manufacture (1 week, 1 month and 3 months) to determine the zeta potential of the vesicles. FD formulations were rehydrated at each time point for the assay. Measurements were determined by a Malvern Zeta-sizer and data are presented as raw values (zeta potential) for 3 independent samples per time point and per storage conditions. Means and SEM are highlighted in bold.

(84) The ability to store the freeze-dried vesicles containing antibiotic at room temperature, or even at 37° C., with no reduction in entrapped antibiotic, is a significant benefit of this formulation technology, obviating the need for cold storage.

(85) In Vitro Assessment of Formulations for Minimum Inhibitory Concentrations (MIC) and Minimum Bactericidal Concentrations (MBC)

(86) To rule out the possibility that the formulations of antibiotics in NISVs or bilosomes would adversely affect antimicrobial function, the MIC and MBCs of the formulations were tested in vitro against F. tularensis and B. pseudomallei. In each of these assays, NISV- and bilosome-formulated antibiotics exerted as much antimicrobial activity as unformulated antibiotics (Table 3), indicating that the process of formulation had not adversely affected the antibiotic cargo.

(87) As an adjunct to the in vitro assessment of the formulated antibiotics, the inventors have also carried out some preliminary in vivo screening in a simple model. Using the wax moth larva (Galleria mellonella) model, challenged with F. tularensis LVS, it was demonstrated that the formulated antibiotics increased both survival and time to disease of F. tularensis LVS-infected G. mellonella i.e. exerted significantly enhanced antimicrobial activity compared with unformulated antibiotics (FIGS. 2A and 2B; data represented on the graph are Kaplan Meier plots generated from 4 independent experiments with 10 G. mellonella per group (total N=40 per group)). Furthermore, the inventors have shown that fluorescently-labelled NISVs are rapidly taken up within 30 minutes into macrophages in vitro, where they continue to accumulate for a further 90 minutes. Using Image Stream technology, the inventors have shown that NISVs are readily visualised in individual cells.

(88) TABLE-US-00003 TABLE 3 In vitro assessment of antimicrobial properties of formulated antibiotics by determination of MC and MBC. Values represented on the table are the median value generated from 3 independent experiments with 3 technical repeats in each experiment. MIC is the lowest concentration that inhibits growth i.e. value is recorded when growth is less than 10% of positive control measured by OD. MBC is the lowest concentration that prevents 99.9% of positive control growth i.e. No bacterial colonies present from 10 μl drops on agar plates. Bacteria B. pseudomallei K96243 F. tularensis SCHU-S4 Antibiotic MIC MBC MIC MBC Formulations (μg/ml) (μg/ml) (μg/ml) (μg/ml) Free ciprofloxacin 2 16 0.06 4 NISV biphasic 2 16 0.125 4 ciprofloxacin Free levofloxacin 2 32 0.06 4 NISV biphasic levofloxacin 2 8 0.03 4 Bilosome biphasic 4 8 0.06 4 levofloxacin Free doxycycline 1 16 0.5 >64 NISV biphasic doxycycline 1 16 1.5 >64 Bilosome biphasic 1 16 0.5 >64 doxycycline

(89) In Vivo Screening of Lead NISV and Lead Bilosome Formulations for Tissue Distribution Pharmacokinetic/Pharmacodynamics (PK/PD) Following Parenteral or Oral Dosing

(90) NISVs and bilosomes incorporating levofloxacin were selected as the lead formulations for PK/PD analysis in vivo. In a time-course study, female Balb/c mice (n=5 per group) were dosed with NISVs or bilosomes by the intraperitoneal or oral dosing routes respectively, with blood and tissues collected at multiple intervals during a 24 hr period. Blood and tissue samples were processed for analysis of antibiotic concentration by HPLC. The metrics of Cmax (maximum drug concentration during dosing interval), Tmax (time at which the Cm is observed) and area under the curve (AUC i.e. total area under the drug concentration-time course) for levofloxacin which had been formulated in biphasic NISVs or bilosomes and freeze-dried, or for free levofloxacin are shown in Table 4 for blood (panel 1), liver (panel 2) or spleen (panel 3). Levofloxacin dosed orally in bilosomes demonstrated the greatest performance, increasing the AUC value for liver deposition two-fold.

(91) TABLE-US-00004 TABLE 4 PK/PD analysis for biphasic formulations of levofloxacin in NISVs and bilosomes. Summary of key PK/PD metrics following dosing via the intraperitoneal or oral route with free, NISV and bilosome levofloxacin formulations in the blood, liver and spleen. Groups of 5 Balb/c mice were dosed and culled at 0.25, 0.5, 1, 2, 4, 8, 16 and 24 hours post treatment. Naïve mice were used as controls. Cmax, Tmax and AUC data were calculated using standard formulae from raw data. Panel 1 50 mg/kg dose for all group Blood plasma Cmax Tmax AUC Formulation Route μg/ml Hours μg/ml/h Levofloxacin Intraperitoneal 50.2155  0.3  122.1861 FD biphasic NISV Intraperitoneal 50.5442  0.45 120.8761 levofloxacin levofloxacin Oral 46.07238  0.6  119.6124 FD biphasic bilosome Oral 44.21877  0.3  116.8437 levofloxacin Panel 2 100 mg/kg dose for all groups Liver Cmax Tmax AUC Formulation Route μg/ml Hours μg/ml/h Levofloxacin Intraperitoneal  2.099737 0.25 5.73601 FD biphasic NISV Intraperitoneal  1.125444 0.6  5.339507 levofloxacin Levofloxacin Oral  1.947519 0.25 6.588408 FD biphasic bilosome Oral  1.698263 0.35 11.59386 levofloxacin Panel 3 100 mg/kg dose for all groups Spleen Cmax Tmax AUC Formulation Route μg/ml Hours μg/ml/h Levofloxacin Intraperitoneal  1.289808 0.25 1.092219 FD biphasic NISV Intraperitoneal  0.655502 0.3  1.182148 levofloxacin Levofloxacin Oral  0.569373 0.35 1.552381 FD biphasic bilosome Oral  0.699086 0.4  1.562795 levofloxacin

(92) Safety/Toxicology Studies on Bilosome Formulations

(93) A study was performed to evaluate the tolerability of ciprofloxacin, levofloxacin and doxycycline formulations given daily for 14 days in Balb/c mice. One hundred and twenty (120) female Balb/c mice were allocated to twelve (12) groups of ten (10) mice and were treated as shown in methods on days 1-14. Animal weights and condition were recorded daily. No mice died during this study. However, weight changes were noticed between treatment routes and antibiotic formulations (results shown in FIG. 3).

(94) It was shown that ciprofloxacin administered via the I.P or oral route significantly induced weight loss as the experiment progressed (P<0.001). However, encapsulation of ciprofloxacin into NISVs and bilosomes significantly negated this weight losing effect by approximately 25% and 40% respectively (P=0.002). The effect of levofloxacin on overall weight loss via both the I.P and oral routes was minimal with NISV and bilosome encapsulation having no effect (P=0.499 and P=0.616 respectively). A strong indication for doxycycline-based weight loss was observed as the experiment progressed when administered via both routes (P<0.001). Encapsulation in bilosomes did not significantly alter the weight loss but encapsulation of deoxycycline into NISVs had a minor positive effect (P=0.091).

(95) On day 15, half the mice in each group were sacrificed and necropsies performed. The remaining mice were weighed and observed for an additional 7 days before necropsies were performed. No abnormal findings were noted on day 15 or day 22 in groups treated with any ciprofloxacin or levofloxacin antibiotic formulations. Some minor pathological observations were made in the doxycycline-treated groups. On day 15 no abnormal findings were noted in mice receiving doxycycline orally, but mice receiving doxycycline via IP injection had rounded edges on the liver and peritoneal adhesions, involving the peritoneal wall and the intestines (large and small). These findings were consistent in all animals in both IP groups (NISV and PBS).

(96) On day 22, again no abnormal findings were noted in mice receiving doxycycline bilosomes orally. In the group receiving doxycycline in PBS orally, two mice had slightly rounded edges to the lobes of the liver. In the groups treated with doxycycline via the intraperitoneal routes, minor peritoneal adhesions were seen in some of the NISV encapsulated groups, but all mice treated with the PBS-doxycycline formulation showed peritoneal adhesions involving spleen, pancreas, intestines and kidney. Overall, these findings suggest that NISV and bilosome formulations of antibiotics significantly improve tolerability, with tissue protecting effects, for repeated dosing of antibiotics.

(97) A repeat-dose safety/toxicology study was carried out to evaluate the tolerability of the bilosome antibiotic formulations. Groups of 10 mice were dosed daily with 0.1 ml (at a concentration of 1 mg per mouse) of each antibiotic. Bilosome formulations were administered by the oral route. Suitable controls (free antibiotic in PBS) were included for all groups. Mice were weighed daily and data recorded as percentage change from starting weight.

(98) The effects of treatment with different antibiotics, formulated in either saline or bilosomes, were evaluated on weight gain, microbiome composition, small intestine histology and serum serotonin levels. Levofloxacin, doxycycline or ciprofloxacin were given orally for seven days, and half the mice were sacrificed on day 8 to evaluate the immediate impact of treatment. Faecal pellets were collected from mice sacrificed at day 8. The remaining mice were monitored for 28 days after the cessation of treatment, with faecal pellets collected at day 22 and terminal samples collected on day 36.

(99) No mortality was noted, small intestine histology was normal in all animals, and there were no statistically significant differences between groups in serum serotonin levels (data not shown). However statistically significant differences in weight gain were noted between treatment groups (FIG. 4 (PBS and empty bilosome control groups); FIG. 5 (levofloxacin formulations); FIG. 6 (deoxycycline formulations); FIG. 7 (ciprofloxacin formulations); percentage daily weight change for each animal and the means for each treatment group were calculated. Error bars represent the standard error for the mean (SEM)).

(100) In the group treated with PBS, mice had a mean weight loss of 0.6% of their starting weight by Day 8, and a mean weight gain of 3.8% on Day 36. The mice treated with empty bilosomes lost an average of 0.3% of their starting weight by Day 8, and had a mean weight gain of 7.5% on Day 36. The mice treated with levofloxacin in PBS lost an average of 3.3% of their starting weight by Day 8, and had a mean weight gain of 3.2% on Day 36. In the group treated with levofloxacin bilosomes, mice had a mean weight loss of 0.9% of their starting weight by Day 8, and a mean weight gain of 6.6% on Day 36. In the group treated with doxycycline in PBS, mice had a mean weight loss of 5.2% of their starting weight by Day 8, and a mean weight gain of 1.7% on Day 36. The mice treated with empty-doxycycline bilosomes lost an average of 0.9% of their starting weight by Day 8, and had a mean weight gain of 1.5% on pay 36. The mice treated with ciprofloxacin in PBS lost an average of 4.4% of their starting weight by Day 8, and had a mean weight gain of 2.4% on Day 36. In the group treated with ciprofloxacin bilosomes, mice had a mean weight gain of 0.1% of their starting weight by Day 8, and a mean weight gain of 9.1% on Day 36.

(101) In particular the bilosome levofloxacin group loss significantly less weight than mice treated with unformulated levofloxacin at days 9, 10, 11, 15, 23, 27 and 36 post treatment (p<0.05). Interestingly therapy ended at day 8 and that is when 3 consecutive days of significant weight difference between groups was observed. No significant difference was seen between bilosome doxycycline and free doxycycline treated groups.

(102) To evaluate the significance of the differences seen in weight gain the Area Under the Curve (AUC) for the percentage weight change for each animal was calculated and the groups compared using a one-way ANOVA test. In this analysis, the data were evaluated to the end of the dosing period on Day 8 and for the animals that were sacrificed on Day 36. In the analysis of the data to Day 8, the ANOVA test indicated that significant differences were seen between groups (p=0.0001), and in the data to Day 36 the ANOVA test also indicated significant differences (p=0.0020). Group-to-group comparisons were performed on the data to evaluate the impact of single variables. Importantly this analysis indicated that there were statistically significant differences between the group treated with ciprofloxacin in PBS and ciprofloxacin bilosomes (p=0.0152). Similar positive trends were seen when comparing levofloxacin in PBS vs levofloxacin bilosomes and PBS vs empty bilosomes.

(103) Microbiome Analysis

(104) To investigate the ability of bilosomes to ameliorate antibiotic-induced weight loss, the microbiome of mice was monitored for 36 days following treatment with free antibiotic or bilosome-encapsulated antibiotic. Mice were administered antibiotic and the microbiome characterised by next generation sequencing. Genus data was categorised into specific phyla to which they belong and percentage abundance for each treatment group was calculated. Data was graphed on area charts/stacked bar charts.

(105) The results demonstrate that bilosomes had no adverse effects on the microbiome (FIG. 8). In general, the effect of antibiotic treatment was to reduce the Gram-negative bacteriodetes and increase the Gram-positive firmicutes at day 8, with a gradual restoration to pre-treatment ratios by day 36. However, these changes were accompanied by the appearance of small amounts of Verucomicrobia species in bilosome treated groups, which are thought to have beneficial effects in the microbiome.

(106) A total of fifty-nine (59) genera were defined as high frequency (present in at least 75% of specimens). There was some variation in the presence of these 59 genera in the baseline samples, but at least 57 genera were present in all groups. Slight reductions in representation were seen on days 8 and 22, with some recovery by day 36. The content of the doxycycline-bilosome group was substantially lower than all the other groups at all time points. Upon evaluation of the data, it became clear that this was due to a near-complete absence of clostridia sequences in this group, which were present at high levels in all other groups. Most samples in the doxycycline-bilosome group, including all samples on days 8 and 22, were devoid of clostridia, while the few samples that did have clostridia, had 100-1000 fold less clostridia than samples from other groups.

(107) A total of eighty-one (81) genera were defined as low frequency (present in less than 25% of specimens). No more than 15 of these 81 genera were present in any one group at any time point, with most genera being absent from most specimens. In total, these genera accounted for very little of the total microbial complement, however the data suggest that bilosome levofloxacin treated groups had a more diverse microflora for these low frequency genera.

(108) Thus, overall, microbiome analysis demonstrated that the bilosome platform is well tolerated in vivo and may encourage the restoration of bacteriodetes and growth of beneficial bacteria (Verucomicrobia) in the gut following antibiotic disturbance.

(109) Immunomodulatory Properties

(110) Bone marrow cells were obtained from the femurs of 8-wk-old male BALB/c mice and differentiated into bone marrow-derived dendritic cells (BMDCs) using GM-CSF-enriched media. This routinely resulted in 80% CD11c.sup.+ cells. BMDCs were then seeded at 5×10.sup.5 cells/well on 96-well microtiter tissue culture plates and stimulated with LPS (3 ug/ml) or left unstimulated as controls. Cells were simultaneously treated with NISV produced by the melt method, or the sonication method (sn NISV), or left untreated. Cells were harvested 24 hrs later and stained with antibodies to CD40, CD80 or CD86 before analysis by flow cytometry. CD40, 80 or 86 levels were not affected in non-LPS stimulated cells treated with either NISV formulation (FIG. 9; *p<0.05, **p<0.01). However, CD40 and CD80 levels were downregulated in LPS-stimulated cells treated with either NISV formulation (FIG. 9). In contrast, CD86 levels were augmented in LPS-stimulated cells treated with either NISV formulation, which is typical of the maturation of myeloid DCs (FIG. 9).

(111) In Vivo Assessment of Bilosome Formulations to Treat B. pseudomallei and F. tularensis Infection

(112) The efficacy of bilosome formulations of levofloxacin and doxycycline was tested in an aerosol model of melioidosis, using a sub-optimal therapy study design, in which antibiotic administration was started at 6 hours post-infection and administered once daily via the oral route for only 7 days, to test therapeutic efficacy. The treatment groups comprised bilosome-delivered antibiotics or unformulated antibiotics, with control groups receiving PBS or empty bilosomes. This study was repeated twice with exactly the same design, conditions and identical treatment groups, the data were stratified and combined for analysis, to give an overall significant survival advantage for bilosome-encapsulated versus free levofloxacin (p=0.014) (FIG. 10) and for bilosome-encapsulated versus free doxycycline (p<0.001) (FIG. 11). Data represented on the graphs are Kaplan Meier plots, with 10 mice per group. Both figures show the results of Experiment 1 (top left panel), Experiment 2 (top right panel) and combined stratified data (bottom panel). PBS and empty bilosome treatment groups were included as controls.

(113) In addition to the survival advantage conferred by delivering levofloxacin in bilosomes, over free levofloxacin, bilosome-levofloxacin-treated groups lost significantly less body weight compared to levofloxacin (FIG. 12). Data represented are plots for individual mice (Experiment 1 (top left panel); Experiment 2 (top right panel); and combined data uses box and whisker plots with inter quartile ranges (bottom panel)). PBS and empty bilosome treatment groups were included as controls. In the levofloxacin treatment groups, the protection against antibiotic-induced weight loss in mice receiving the bilosome formulations was very marked, such that the combined data from the bilosome-levofloxacin treated groups gave a highly significant difference in weight loss (p<0.001, FIG. 12).

(114) Similarly to B. pseudomallei, Balb/c mice were treated with bilosome-delivered levofloxacin p.i. with F. tularensis. Antibiotic therapy (20 mg/kg) was commenced at 72 hours after exposure to 100 cfu of aerosolised F. tularensis SchuS4 (approximately 20 MLD) and delivered daily for only 3 days to groups of 15 mice. Animals receiving bilosome-delivered levofloxacin by the oral route had a significantly (p<0.05) extended time to death, compared with those receiving free levofloxacin (FIG. 13). This result provides further evidence of the advantages of bilosome delivery of antibiotics.

(115) Bacteriological, immunological and blood chemistry analyses show supporting changes between bilosome and free antibiotic-treated groups. At day 3 p.i. with B. pseudomallei, cytokine analysis of lung, spleen and liver tissue samples showed a decrease in pro-inflammatory cytokines in bilosome-levofloxacin treated groups and this correlated with a reduction in ALT and GGT, enzymes associated with liver damage.

(116) Other evidence of beneficial effects of bilosome delivery of antibiotic has been derived from the F. tularensis study in which mice administered bilosome-entrapped levofloxacin had significantly reduced levels of some key inflammatory cytokines, compared to those mice receiving unformulated levofloxacin (p<0.05: IL-6 lung, IL-13 spleen, Eotaxin Liver, GM-CSF lung and liver, MIP1b liver). Analysis of cytokine levels in Burkholderia-infected mice has shown similar trends.

(117) Overall, the data indicate that bilosome delivery of antibiotic not only provides a significant survival advantage, but also reduces side effects associated with repeated antibiotic dosing and modulates the immune response to reduce cytokine-induced morbidity that is normally evident in these infections.

(118) Entrapment of a Range of Antivirals in Modified NISVs

(119) In order to develop an effective therapy for the encephalitic viruses (e.g. VEEV), NISVs have been modified to facilitate their passage across the blood brain barrier (BBB). This approach has been achieved by coating with palmitoylated glucosamine, or covalently linking the cholesterol moiety in the NISV formulation to transferrin, to create gNISV and tNISV respectively (‘brainsomes’). Palmitoylated glucosamine has been incorporated into gNISV with a view to binding Glut-1 receptors (prevalent in the endothelial tissues of the BBB) in vivo. Similarly, tNISV have been prepared by thiolation of human holo-transferrin to allow it to bind a maleimide modified cholesterol component substituted for a portion (29%) of the normal cholesterol included in NISVs.

(120) To optimize the entrapment of antibodies into NISVs and brainsomes and specifically. as model for the VEEV mAb, we have used non-specific rabbit IgG. This has been successfully entrapped into NISV, gNISV and tNISV at high efficiency (achieving 25-32% entrapment) (Table 5). Using these conditions, VEEV Mab has been incorporated into gNISV with a similarly high efficiency and formulated into tNISV (Table 5). IFNα2β has also been successfully entrapped into NISV with excellent efficiency (43%).

(121) TABLE-US-00005 TABLE 5 Physiochemical characterization of NISV variants loaded with antiviral cargo (or surrogate). Size Zeta Percentage Formulation (nm) potential PDI entrapment gNISV IgG 600 17 0.6 30 tNISV IgG 316 10 0.5 25 NISV IgG 300 20 0.7 32 gNISV VEEV 200 13 0.4 25.9 NISV IFNa2b 100 12 0.6 43

(122) Characterisation of Automated (Scaled-Up) NISV Preparation

(123) Nanovesicle production (brainsomes and bilosomes) has been successfully scaled-up in the NanoAssemblr®, with the aim of achieving a routinely automated and consistent manufacturing process to prepare nanovesicles with selected coatings and of optimized size ranges. NISV, brainsomes and bilosomes have all been successfully made using the NanoAssemblr® (Table 6). Modification of NISVs to provide brainsomes had no significant effect on properties of zeta potential or vesicle size, compared with non-coated NISVs.

(124) TABLE-US-00006 TABLE 6 Characterisation of automated (scaled-up) NISV preparations Size (z-average Zeta potential (- Concentration Vesicles nm) mV) (mg/ml) NISV deoxycyline 175 22.7 20 NISV levofloxacin 124 20 11 Bilosome 136 20 13 levofloxacin

(125) In Vitro and In Vivo Characterisation of NISV Deposition Including BBB-CNS Delivery Assessment

(126) Initially, gNISV were stained with Hoechst dye as a model cargo, since the dye is unable to cross the BBB under normal circumstances. Stained gNISV were injected intravenously in the Balb/c mouse and brain tissue removed at 90 minutes. As expected, Hoescht dye alone did not cross the BBB and was not detected in brain. gNISV with encapsulated Hoechst dye crossed the BBB to stain the nuclei of cells, whereas little nuclear staining was detected with unmodified NISV encapsulating Hoechst dye. These data indicate that the gNISV are able to cross the BBB and release their cargo.

(127) Having demonstrated that gNISV can deliver this model cargo, gNISVs have subsequently been used to encapsulate a polyclonal rabbit antibody (MW 150 kDa), with the intention ultimately of encapsulating the humanised anti-VEEV monoclonal IgG, Hu1A3B-7.

(128) Initial studies whereby the rabbit IgG encapsulated in gNISV has been delivered, on a single occasion by intravenous injection to the mouse, have revealed the presence of the antibody in mouse brain at 90 minutes (FIG. 14; tissue slices were fluorescently stained for rabbit IgG and total fluorescence in tissue slice measured with image analysis software. Data represented are means of 3 individual mice per time point and per group. Free IgG antibody was used as a control). Whilst antibody levels in brain had declined by 4 hours, significantly enhanced levels of antibody delivered by gNISV, were present in liver and lung at this timepoint.

(129) The encapsulation of Hu1A3B-7 into gNISV and tNISV has been Walled. Immunohistochemical analysis of mouse brain after intravenous delivery of rabbit IgG in gNISV has demonstrated the presence of the rabbit antibody not only associated with brain vasculature, but also widespread in the brain tissue from these mice, in contrast to brain tissue from mice injected with unformulated rabbit antibody.

(130) Detailed Assessment of Lead NISV Formulation p.i. With Aerosolised VEEV

(131) The VEEV mAb Hu1A3B-7 (S. A. Goodchild, L. M. O'Brien, J. Steven, M. R. Muller, O. J. Lanning, C. H, Logue, R. V. D'Elia, R. J. Phillpotts, S. D. Perkins. A humanised murine monoclonal antibody with broad serogroup specificity protects mice from challenge with Venezuelan equine encephalitis virus. Antiviral Res., 90(2011), 1-8; further disclosed in WO2011036435A1) was the chosen antiviral used during in vivo infection studies. A lethal aerosol model of VEEV infection was used with BALB/c mice exposed to ˜300PFU of VEEV Trinidad Donkey strain. This challenge causes clinical signs of infection at approximately 2 days p.i. with time to death between 5-6 days.

(132) It was demonstrated that the process of encapsulation and freeze drying had no detrimental effect on the activity of the mAb, with plaque assays giving similar results for free mAb and NISV and gNISV formulations of the mAb. Empty vesicles controls were shown not to have any anti-viral activity and acted the same as media-only controls (data not shown).

(133) A small scale preliminary study was conducted to test the utility of giving the therapy via the IV route (Table 7). 5 BALB/c mice per group (PBS control, Empty vesicle control, VEEV Hu1A3B-7 mAb, NISV Hu1A3B-7 mAb and gNISV Hu1A3B-7 mAb) were exposed to VEEV via the aerosol route with intravenous treatment given once p.i. at 24 hrs. All mice were culled at day 5 p.i. prior to control mice succumbing to disease. Brain and lung tissues were removed and processed for viral titres. The NISV platform delivered intravenously significantly decreased viral load in the lung and brain (P=0.001 compared to PBS control and P=0.007 compared to free mAb; data not shown).

(134) TABLE-US-00007 TABLE 7 Animal study plan for NISV and gNISV mAb intravenous treatment of aerosol VEEV infection. Therapy Length initiated of No. Treatment Route of at PC treatment of Groups regimen therapy (hours) (days) mice 1a PBS Intravenous 24 1 5 2a VEEV mAb Intravenous 24 1 5 Hu1A3B-7 3a NISV mAb Intravenous 24 1 5 Hu1A3B-7 4a gNISV mAb Intravenous 24 1 5 Hu1A3B-7 5a Empty gNISV Intravenous 24 1 5

(135) To better fully characterise the differences between NISV platform formulations and free compound, an intraperitoneal treatment model was used (Table 8). This route has previously been shown to be more consistent and suitable for assessing the efficacy of the Hu1A3B-7 mAb.

(136) TABLE-US-00008 TABLE 8 Animal study plan for NISV and gNISV mAb intraperitoneal treatment of aerosol VEEV infection. Therapy Length initiated of No. Route of at PC treatment of Groups Treatment regimen therapy (hours) (days) mice 6a + b PBS Intraperitoneal 24 1 10 7a + VEEV mAb Intraperitoneal 24 1 15 b + c Hu1A3B-7 (100 μg in 100 μl) 8a + NISV mAb Intraperitoneal 24 1 15 b + c Hu1A3B-7 (100 μg in 100 μl) 9a + gNISV mAb Intraperitoneal 24 1 15 b + c Hu1A3B-7 (100 μg in 100 μl) 10a + b Empty gNISV Intraperitoneal 24 1 10 (100 μg)

(137) Identical treatment groups (PBS control, Empty vesicle control, VEEV Hu1A3B-7 mAb, NISV Hu1A3B-7 mAb and gNISV Hu1A3B-7 mAb) were assessed using the aerosol VEEV model, with therapy again given once at 24 hours via the intraperitoneal route. In these studies a total of 15 mice per treatment group were used, with 5 mice being culled for virology analysis at day 5 p.i. and the remaining 10 mice monitored for survival.

(138) Two independent studies of identical design were carried out and showed similar results for all treatment groups. In experiment 1, both the NISV and gNISV entrapped Hu1A3B-7 mAb reduced viral load in the lung compared to PBS treated groups (p<0.001 and P=0.003 respectively). NISV-entrapped mAb was also significantly different from free Hu1A3B-7 mAb in terms of viral titre in the lung (p<0.001). NISV entrapped mAb also performed better in the spleen compared to PBS and free Hu1A3B-7 VEEV mAb (p<0.001 and <0.05 respectively). Importantly the gNISV-encapsulated mAb, specifically designed to target the brain performed significantly better than all other formulations and controls in the brain. This lead to increased survival in the NISV and gNISV treated groups with NISV being significantly better than free Hu1A3B-7 VEEV mAb (p=0.008).

(139) In experiment 2, the gNISV performed better than the PBS control in the lung (p=0.0031) and significantly better in the brain compared to all formulations. The NISV and gNISV formulation performed significantly better than PBS controls in the spleen tissue (p=0.0143 and P<0.001 respectively). Interestingly the free Hu1A3B-7 VEEV mAb also showed a beneficial effect in the spleen indicating that the therapy was efficacious despite the low VEEV-specific active component. In this experiment, gNISV mAb significantly improved survival compared to free Hu1A3B-7 VEEV mAb (p=0.0036) and NISV mAb groups showed a similar trend.

(140) When the data from experiments 1 and 2 are combined, the positive effects of free Hu1A3B-7 VEEV mAb are clear, but more importantly the advantages provided by the encapsulated formulations (NISVS and gNISVs) are highlighted. Both the NISV and gNISV mAb formulation significantly reduced viral load in the lung compared to PBS and free Hu1A3B-7 VEEV mAb controls. Furthermore, the gNISV significantly reduced viral load in the brain compared to both controls. Interestingly the spleen data showed that the free Hu1A3B-7 VEEV mAb provided therapeutic benefit by significantly reducing viral load compared to the PBS controls, although the viral load in the spleens of NISV and gNISV mAb groups did not differ significantly from the free Hu1A3B-7 group (FIG. 15; BALB/c mice challenged with aerosolised VEEV. A total of 10 mice per treatment group were culled at day 5 p.i. with viral titre determined in the lung, rain, spleen and liver. P values generated by ANOVA and multiple comparisons).

(141) The limited effect of the free mAb in the lung and brain was most likely due to the level of VEEV-specific active component in the preparation. This however further highlights the advantage of NISV encapsulation, by demonstrating the ability to increase the efficacy of sub-optimal levels of therapeutic compound. This is consistent with the antibiotic data where sub-optimal levels of antibiotic were used to demonstrate the advantage of NISV and bilosome entrapment.

(142) Reference to a Sequence Listing Submitted as a Text File Via Efs-Web

(143) The official copy of the sequence listing is submitted electronically via EFS-Web as an ASCII formatted sequence listing with a file named D1722_sequence listing.TXT, created on Mar. 3, 2021, and having a size of 4.096 kilobytes. The sequence listing contained in this ASCII formatted document is part of the specification and is herein incorporated by reference in its entirety.

(144) Combining the survival data from experiments 1 and 2 shows that both the NISV and gNISV Hu1A3B-7 VEEV mAb formulations performed significantly better that PBS and free Hu1A3B-7 VEEV mAb (p=0.0294 and p=0.0030 respectively; FIG. 16).

(145) Overall, this data indicates that NISV and gNISV delivery of Hu1A3B-7 VEEV mAb is advantageous compared to free mAb. Both platforms reduce viral load in the lung, and increase survival and reduce morbidity. Significantly, the gNISV platform reduces viral load in the brain. This would suggest that the vesicles specifically designed to deliver cargo such as pharmaceutical agents to the brain are capable of targeting this tissue and delivering therapeutically-relevant concentrations of cargo.

(146) It will be understood that the present invention has been described above purely by way of example, and modification of detail can be made within the scope of the invention. Each feature disclosed in the description, and (where appropriate) the claims, may be provided independently or in any appropriate combination. Moreover, the invention has been described with specific reference to methods, pharmaceutical composition and associated kits, and more specifically with reference to pharmaceutical-vesicle formulations and associated pharmaceutical kits for use against B. pseudomallei, F. tularensis and VEEV. Additional applications of the invention will occur to the skilled person.