WO2015002956A1 - Exosome delivery system - Google Patents

Abstract

Disclosed are exosomes loaded with a therapeutic polynucleotide, as well as compositions, systems, and methods for making same. The disclosed exosomes can contain an exosome targeted fusion protein and a chimeric polynucleotide. Also disclosed is a composition comprising an exosome containing the disclosed exosome targeted fusion protein. In some embodiments, the exosome is loaded with a disclosed chimeric polynucleotide. Also disclosed is an exosome producing cell engineered to produce the disclosed targeted exosomes. Also disclosed is a method for making the disclosed targeted exosome loaded with a therapeutic polynucleotide that involves culturing the disclosed exosome producing cells under conditions suitable to produce exosomes. The method can further involve purifying exosomes from the cell that comprise the targeted fusion protein.

Description

EXOSOME DELIVERY SYSTEM

CROSS-REFERENCE TO RELATED APPLICATIONS

This application claims benefit of U.S. Provisional Application No.

61/841,577, filed July 1, 2013, which is hereby incorporated herein by reference in its entirety.

STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH OR

DEVELOPMENT

This invention was made with Government Support under Grant No.

1UH2TR000914 awarded by the National Institutes of Health. The Government has certain rights in the invention.

BACKGROUND

Therapeutic approaches using oligonucleotides have been studied in detail. These approaches include small interfering RNA (siRNA) as well as antisense to miRNAs that are overexpressed or miRNA mimics of miRNAs that are reduced in disease. It is widely accepted that delivery of therapeutic oligonucleotides is a major bottleneck in the clinical development of these agents. Oligonucleotides are inherently unstable in circulation. They are difficult to penetrate cell membranes in the absence of lipid transfection agents due to their size and charge. Once

internalized into cells they typically entrapped within the lysosome. While lipid nanoparticles are the current standard method for oligonucleotide delivery, they possess certain limitations. Composed of synthetic ingredients, lipid nanoparticles will decompose in vivo to produce cytotoxic or immunogenic activities. For example, lipid nanoparticles were shown to produce a variety of toxicities including proinflammatory response and activation of toll-like receptor 4 (Kedmi R, et al.

Biomaterials. 2010 31 :6867-75). Therefore, improved delivery systems are needed.

SUMMARY

Disclosed are exosomes loaded with a therapeutic polynucleotide, as well as compositions, systems, and methods for making same. The disclosed exosomes do not require transfection for polynucleotide loading. Rather, an exosome producing cell can be engineered to produce the exosome and load it with a heterologous therapeutic polynucleotide. This is accomplished in part by producing an exosome targeted fusion protein that binds a chimeric polynucleotide containing the therapeutic polynucleotide when expressed by the cell.

The disclosed exosome can therefore contain an exosome targeted fusion protein and a chimeric polynucleotide. In some embodiments, the exosome targeted fusion protein contains a nucleic acid binding moiety and an exosomal

transmembrane moiety. In these and other embodiments, the chimeric polynucleotide can contain a therapeutic nucleic acid sequence and a substrate for the nucleic acid binding moiety in the targeted fusion protein.

Therefore, disclosed is a system for making the exosome that contains a first and second nucleic acid sequence, wherein the first nucleic acid sequence encodes the exosome targeted fusion protein and wherein the second nucleic acid sequence encodes the chimeric polynucleotide.

Non-limiting examples of therapeutic nucleic acid sequences include siRNA, dsRNA, dsDNA, shRNA, mRNA, microRNA, antimicroRNA, antagomir, antimir, antisense, aptamer, and dsRNA/DNA hybrids.

The nucleic acid binding moiety of the exosome targeted fusion protein and the substrate for the nucleic acid binding moiety in the chimeric polynucleotide can be any amino acid/nucleic acid pair where the amino acid sequence is capable of specifically binding the nucleic acid sequence. For example, in some embodiments, the nucleic acid binding moiety comprises HIV Tat protein, or fragment or variant thereof capable of binding an HIV trans-activation response (TAR) element, and the substrate for the nucleic acid binding moiety comprises an HIV TAR element, or a fragment or variant thereof capable of binding an HIV Tat protein. Another example is binding of the stem loop structure of the prion protein (PrP) mRNA to the nuclear lectin protein CBP35. A further example is the interaction between a Fab fragment protein and the loop region of the class I ligase ribozyme P5 RNA hairpin.

Non- limiting examples of exosomal transmembrane moieties include Lamp-1,

Lamp-2, CD13, CD86, Flotillin, Syntaxin-3, CD2, CD36, CD40, CD40L, CD41a, CD44, CD45, ICAM-1, Integrin alpha4, LiCAM, LFA-1, Mac-1 alpha and beta, Vti- 1A and B, CD3 epsilon and zeta, CD9, CD18, CD37, CD53, CD63, CD81, CD82, CXCR4, FcR, GluR2/3, HLA-DM (MHC II), immunoglobulins, MHC-I or MHC-II components, TCR beta, and tetraspanins.

The exosome targeted fusion protein can further comprise a tissue targeting moiety to target the exosome to a specific organ, tissue, or cell type. For example, the tissue targeting moiety can comprise a PC94 peptide, a rabies virus glycoprotein (RVG), an RGD peptide, luteinizing hormone -releasing hormone (LHRH) peptide, or galectin-3 -binding peptide. RGD peptide targets integrins (overexpressed in a wide variety of cancers). LHRH peptide is overexpressed in breast, ovarian, prostate and hepatic carcinoma. Galectin-3 -binding peptide (G3-C12) targets galectin-3, which is overexpressed in prostatic carcinoma. In some embodiments, the tissue targeting moiety comprises a single chain antibody (scFv) that binds a tissue specific antigen. For example the scFv AF-20 shown to target human hepatocellular carcinoma cells. In some embodiments, the tissue targeting moiety is a cell-specific internalization peptide that infiltrates tumor tissue.

The exosome targeted fusion protein can further comprise a protein tag. For example, the protein tag can be a FLAG-tag or a streptavidin binding peptide.

Therefore, the exosome targeted fusion protein can have the following formula:

P - TT - TM - NB - P,

wherein "P" consists of an optional protein tag,

wherein "TT" consists of an optional tissue targeting moiety ,

wherein "TM" consists of a exosomal transmembrane moiety,

wherein "NB" consists of a nucleic acid binding moiety, and

wherein "-" consists of a peptide linker or a peptide bond.

Each of the first nucleic acid sequence and the second nucleic acid sequence can be operably inserted in an expression vector. In some embodiments, the first nucleic acid sequence and second nucleic acid sequence are operably inserted in a common expression vector so they are expressed together.

Also disclosed is a composition comprising an exosome containing the disclosed exosome targeted fusion protein. In some embodiments, the exosome is loaded with a disclosed chimeric polynucleotide. Also disclosed is an exosome producing cell engineered to contain the first nucleic acid sequence and the second nucleic acid sequence of the disclosed system. Non- limiting examples of exosome producing cells include dendritic cells (e.g., immature dendritic cell), Human Embryonic Kidney 293 (HEK) cells, 293T cells, Chinese hamster ovary (CHO) cells, and human ESC-derived mesenchymal stem cells.

Also disclosed is a method for making the disclosed exosome loaded with a therapeutic polynucleotide that involves culturing the disclosed exosome producing cell engineered to contain the first nucleic acid sequence and the second nucleic acid sequence of the disclosed system under conditions suitable to produce exosomes. The method can further involves purifying exosomes from the cell that comprise the targeted fusion protein.

The details of one or more embodiments of the invention are set forth in the accompanying drawings and the description below. Other features, objects, and advantages of the invention will be apparent from the description and drawings, and from the claims.

DESCRIPTION OF DRAWINGS

Figures lAto 1C illustrates a scheme for targeting protein engineering. Figure 1 A shows three targeting proteins. The upper protein contains a N-terminus FLAG tag, an RVG targeting peptide sequence, and a membrane spanning Lamp-2 protein. The middle targeting protein contains an N-terminus FLAG -tag, a 12 amino acid PC94 targeting moiety, a membrane spanning Lamp-2 protein, and a C-terminus Tat peptide. The lower protein is a nontargeting control protein. Figure IB is an illustration of the core segments of a modified pri-miR-199a-2. On the left is a pri- miRNA sequence showing the basal segments & lower stem that are required for accurate Drosha processing. On the right, are bulge & loop motifs of the HIV TAR RNA. In the middle are passenger & guide strand, respectively for pre-mR-199a-2. Figure 1C shows sequences of three pri-mR-I99a-2 constructs. The wild type pri-mR- 199a-2 sequence (top, SEQ ID NO: l) is shown with the mature miR-199a-3p sequence underlined. Eight nucleotides contained within the loop and stem portions were modified in the pri-miR -199a-2 sequence (middle, SEQ ID NO:2) to resemble that portion of the TAR RNA that binds to the Tat protein (bottom, SEQ ID NO:3).

Figure 2 is a schematic overview of a targeted microvessicle delivery system. In step (a), The modified miR-199a-2 gene engineered into an intron of the targeting protein gene is spliced from the pre-mRNA in the nucleus of the HEK293T cells. In step (b), targeting protein is translated from the spliced mRNA, while in step (c) the modified pri-miR-199a-2 is processed to mature miR-199a-3p using the natural miRNA biogenesis. In step (d), both targeting peptide and pre-miR-199a-2 are directed into extracellular vesicles; the targeting peptide through alignment or the Lamp2 transmembrane domain and the pre-miRNA-199a-2 through binding of the modified precursor or the Tat peptide on the luminal C-terminus of the targeting protein. In step (e), extracellular vesicles are shed into the extracellular space. In step (f), extracellular vesicles following purification are used to treat recipient cancer cell lines expressing the target of the PC94 peptide present on the cytoplasmic membrane. In step (g), once the extracellular vesicles fuse to the recipient cell, the pre-miR-199a- 2 is released into the cytoplasm were it is processed to mature miR-199a-3p and suppresses protein levels following binding to the 3' UTR of target mRNA.

Figure 3 shows results from EMSA gel shift of modified pre-miR-199a-2. The ³²P-labeled transcript of the modified pre-miR-199a-2 was incubated in the absence (lane 1) or presence of 10- or 50- fold excess of Tat peptide, lanes 2 and 3

respectively.

Figure 4 shows results of rDicer processing of peptide bound modified pre- miR-199a-2. A 10-fold excess of Tat peptide was bound to the 62 nt modified pre- miR-199a-2 transcript (lane 3). Transcripts were exposed to rDicer for 18 hrs (lanes 2- 3) or mock treatment (lane 1). The 23 nt mature miR-199a (arrow) was generated by rDicer in the absence (lane 2) or presence (lane 3) of the Tat peptide.

Figure 5 is a schematic showing engineering of pri-miR-199a-2 gene into chimeric intron. Unique Hindlll and Sail restriction sites located in exons 3 and 4, respectively of the Lamp2B cDNA (upper). The 370 bp chimeric intron containing the pri-miR-199a-2 gene, 5' and 3' flanking sequences of exons 3 and 4, respectively and the 5' and 3' splice sites arecloned into the Hindlll and Sail sites of the targeting vector. Figure 6A is a Cryo-TEM image of HEK293T microvesicles (arrows) isolated using ultracentrifugation. Bar, 100 nm. Figure 6B is a graph showing particle size distribution using NanoSight analysis. The mean particle size was 122 +/- 2.8 nm and the concentration was 7.56 +/- 0.25 E8 particles/ml.

Figure 7 shows translation of engineered proteins in 293T cells and microvesicles. HEK293T cells were engineered to produce a Flag-PC94-Lamp2a-Tat- His fusion protein. Protein was extracted from the HEK293T cells (Cells) or microvesicles (MVs) and the presence of Flag, His, endogenous LAMP1, GAPDH and histone were determined by western blotting. 30 μg (Cells) and 15 μg (MVs) were loaded per lane of the gel.

Figure 8 is a bar graph showing pre-miR-199a mimic is preferentially loaded into targeting microvesicles. HEK293T were transfected with vectors containing the fusion gene Flag-PC94-Lamp2a-Tat-His (FPLTH), empty vector, or FPLTH vector containing the wildtype or loop modified pre-miR-199a. Total RNA was extracted from the cells and qRT-PCR was performed using a miR-199a modified loop specific TaqMan probe. The values in parenthesis are the CT values form the qRT-PCR.

DETAILED DESCRIPTION

Disclosed are exosomes loaded with a therapeutic polynucleotide, as well as compositions, systems, and methods for making same. The disclosed exosome can contain an exosome targeted fusion protein and a chimeric polynucleotide. Also disclosed is a composition comprising an exosome containing the disclosed exosome targeted fusion protein. In some embodiments, the exosome is loaded with the disclosed chimeric polynucleotide. Also disclosed is an exosome producing cell engineered to produce the disclosed exosomes. Also disclosed is a method for making the disclosed exosome loaded with a therapeutic polynucleotide that involves culturing the disclosed exosome producing cells under conditions suitable to produce exosomes. The method can further involve purifying exosomes from the cell that comprise the targeted fusion protein.

Fusion proteins, also known as chimeric proteins, are proteins created through the joining of two or more genes which originally coded for separate proteins.

Translation of this fusion gene results in a single polypeptide with function properties derived from each of the original proteins. Recombinant fusion proteins can be created artificially by recombinant DNA technology for use in biological research or therapeutics. Chimeric mutant proteins occur naturally when a large-scale mutation, typically a chromosomal translocation, creates a novel coding sequence containing parts of the coding sequences from two different genes.

The functionality of fusion proteins is made possible by the fact that many protein functional domains are modular. In other words, the linear portion of a polypeptide which corresponds to a given domain, such as a tyrosine kinase domain, may be removed from the rest of the protein without destroying its intrinsic enzymatic capability. Thus, any of the herein disclosed functional domains can be used to design a fusion protein.

A recombinant fusion protein is a protein created through genetic engineering of a fusion gene. This typically involves removing the stop codon from a cDNA sequence coding for the first protein, then appending the cDNA sequence of the second protein in frame through ligation or overlap extension PCR. That DNA sequence will then be expressed by a cell as a single protein. The protein can be engineered to include the full sequence of both original proteins, or only a portion of either.

If the two entities are proteins, often linker (or "spacer") peptides are also added which make it more likely that the proteins fold independently and behave as expected. Especially in the case where the linkers enable protein purification, linkers in protein or peptide fusions are sometimes engineered with cleavage sites for proteases or chemical agents which enable the liberation of the two separate proteins. This technique is often used for identification and purification of proteins, by fusing a GST protein, FLAG peptide, or a hexa-his peptide (aka: a 6xhis-tag) which can be isolated using nickel or cobalt resins (affinity chromatography). Chimeric proteins can also be manufactured with toxins or anti-bodies attached to them in order to study disease development.

Alternatively, internal ribosome entry sites (IRES) elements can be used to create multigene, or polycistronic, messages. IRES elements are able to bypass the ribosome scanning model of 5' methylated Cap dependent translation and begin translation at internal sites (Pelletier and Sonenberg, 1988). IRES elements from two members of the picornavirus family (polio and encephalomyocarditis) have been described (Pelletier and Sonenberg, 1988), as well an IRES from a mammalian message (Macejak and Sarnow, 1991). IRES elements can be linked to heterologous open reading frames. Multiple open reading frames can be transcribed together, each separated by an IRES, creating polycistronic messages. By virtue of the IRES element, each open reading frame is accessible to ribosomes for efficient translation. Multiple genes can be efficiently expressed using a single promoter/enhancer to transcribe a single message (U.S. Pat. Nos. 5,925, 565 and 5,935,819;

PCT/US99/05781). IRES sequences are known in the art and include those from encephalomycarditis virus (EMCV) (Ghattas, I. R. et al, Mol. Cell. Biol, 11 :5848- 5849 (1991); BiP protein (Macejak and Sarnow, Nature, 353:91 (1991)); the

Antennapedia gene of drosophilia (exons d and e) [Oh et al, Genes & Development, 6: 1643-1653 (1992)); those in polio virus [Pelletier and Sonenberg, Nature,

334:320325 (1988); see also Mountford and Smith, TIG, 11 : 179-184 (1985)).

In some embodiments, the exosome targeted fusion protein contains a nucleic acid binding moiety and an exosomal transmembrane moiety.

The nucleic acid binding moiety of the exosome targeted fusion protein and the substrate for the nucleic acid binding moiety in the chimeric polynucleotide can be any amino acid/nucleic acid pair where the amino acid sequence is capable of specifically binding the nucleic acid sequence. For example, in some embodiments, the nucleic acid binding moiety comprises HIV Tat protein, or fragment or variant thereof capable of binding an HIV trans-activation response (TAR) element, and the substrate for the nucleic acid binding moiety comprises an HIV TAR element, or a fragment or variant thereof capable of binding an HIV Tat protein. For example, the HIV TAR element can have the nucleic acid sequence 5 '-GGCAGAUCUG

AGC CUGGG AG CUCUCUGCC-3' (SEQ ID NO:3) or a variant thereof capable of binding HIV Tat, and the HIV Tat protein can have the amino acid sequence

RPRGTRGKGR RIRR (SEQ ID NO:5), or a variant thereof capable of binding HIV TAR.

Another example is binding of the stem loop structure of the prion protein

(PrP) mRNA to the nuclear lectin protein CBP35. A further example is the interaction between a Fab fragment protein and the loop region of the class I ligase ribozyme P5 RNA hairpin.

The exosome targeted fusion protein can be expressed on the surface of the exosome by expressing it as a fusion protein with an exosomal transmembrane protein. A number of proteins are known to be associated with exosomes; that is they are incorporated into the exosome as it is formed. Examples include but are not limited to Lamp-1, Lamp-2, CD13, CD86, Flotillin, Syntaxin-3, CD2, CD36, CD40, CD40L, CD41a, CD44, CD45, ICAM-1, Integrin alpha4, LiCAM, LFA-1, Mac-1 alpha and beta, Vti-IA and B, CD3 epsilon and zeta, CD9, CD18, CD37, CD53, CD63, CD81, CD82, CXCR4, FcR, GluR2/3, HLA-DM (MHC II), immunoglobulins, MHC-I or MHC-II components, TCR beta and tetraspanins. In some embodiments, the transmembrane protein is selected from Lamp-1, Lamp-2, CD 13, CD86, Flotillin, Syntaxin-3. For example, the transmembrane protein can be Lamp-2. An example sequence for Lamp-2 is set forth below:

MVCFRLFPVPGSGLVLVCLVLGAVRSYALELNLTDSENATCLYAKWQMNFT

VRYETTNKTYKTVTISDHGTVTYNGSICGDDQNGPKIAVQFGPGFSWIANFTK

AASTYSIDSVSFSYNTGDNTTFPDAEDKGILTVDELLAIRIPLNDLFRCNSLSTL

EK DVVQHYWDVLVQAFVQNGTVSTNEFLCDVDKTSTVAPTIHTTVPSPTTT

PTPKEKPEAGTYSVNHGNDTCLLATMGLQLNITQDKVASVININPNTTHSTGS

CRSHTALLRLNSSTIKYLDFVFAVK ENRFYLKEVNISMYLVNGSVFSIAN N

LSYWDAPLGSSYMCNKEQTVSVSGAFQINTFDLRVQPFNVTQGKYSTAQ

(SEQ ID NO: 7). Therefore, the transmembrane protein can have the amino acid sequence SEQ ID NO:7, or a variant thereof capable of targeting the protein to exosomes.

The disclosed exosomes can be targeted to a desired cell type or tissue. This targeting can be achieved by expressing on the surface of the exosome a targeting moiety which binds to a cell surface moiety expressed on the surface of the cell to be targeted. Typically the targeting moiety is a peptide within the disclosed exosome targeted fusion protein. However, it can also be independently expressed as a fusion protein with an exosomal transmembrane moiety.

Examples of suitable targeting moieties are short peptides, scFv and complete proteins, so long as the targeting moiety can be expressed on the surface of the exosome and does not interfere with insertion of the membrane protein into the exosome. Typically the targeting peptide is heterologous to the transmembrane exosomal protein. Peptide targeting moieties may typically be less than 100 amino acids in length, for example less than 50 amino acids in length, less than 30 amino acids in length, to a minimum length of 10, 5 or 3 amino acids.

Targeting moieties can be selected to target particular tissue types such as muscle, brain, liver, pancreas and lung for example, or to target a diseased tissue such as a tumor. For example, the tissue targeting moiety can comprise a PC94 peptide, a rabies virus glycoprotein (RVG), an RGD peptide, luteinizing hormone -releasing hormone (LHRH) peptide, or galectin-3 -binding peptide. RGD peptide targets integrins (overexpressed in a wide variety of cancers). LHRH peptide is

overexpressed in breast, ovarian, prostate and hepatic carcinoma. Galectin-3 -binding peptide (G3-C12) targets galectin-3, which is overexpressed in prostatic carcinoma.

In some embodiments, the PC94 peptide has the amino acid sequence

SFSIIHTPIL PL (SEQ ID NO:4), or a variant thereof capable of binding to HCC cells.

In some embodiments, the tissue targeting moiety comprises a single chain antibody (scFv) that binds a tissue specific antigen. For example the scFv AF-20 shown to target human hepatocellular carcinoma cells. In some embodiments, the tissue targeting moiety is a cell-specific internalization peptide that infiltrates tumor tissue.

Additional examples of targeting moieties include muscle specific peptide, discovered by phage display, to target skeletal muscle, a 29 amino acid fragment of Rabies virus glycoprotein that binds to the acetylcholine receptor or a fragment of neural growth factor that targets its receptor to target neurons and secretin peptide that binds to the secretin receptor can be used to target biliary and pancreatic epithelia. As an alternative, immunoglobulins and their derivatives, including scFv antibody fragments can also be expressed as a fusion protein to target specific antigens, such as VEGFR for cancer gene therapy. As an alternative, natural ligands for receptors can be expressed as fusion proteins to confer specificity, such as NGF which binds NGFR and confers neuron-specific targeting. The exosome targeted fusion protein can further comprise a protein tag, such as an affinity tag, epitope tag, chromatography tag, fluorescence tag.

Suitable affinity tags include, but are not limited to, chitin binding protein (CBP), maltose binding protein (MBP), and glutathione-S-transferase (GST), and poly(His) tag.

Epitope tags are short peptide sequences which are chosen because high- affinity antibodies can be reliably produced in many different species. These are usually derived from viral genes, which explain their high immunoreactivity. Epitope tags include V5-tag, Myc-tag, and HA-tag. These tags are particularly useful for western blotting, immunofluorescence and immunoprecipitation experiments, although they also find use in antibody purification.

Chromatography tags are used to alter chromatographic properties of the protein to afford different resolution across a particular separation technique. Often, these consist of polyanionic amino acids, such as FLAG-tag. For example, the FLAG- tag can have the amino acid sequence DYKDDDDK (SEQ ID NO:6), or a variant thereof containing an epitope for an anti-FLAG antibody.

Fluorescence tags are used to give visual readout on a protein. GFP and its variants are the most commonly used fluorescence tags.

In some embodiments, the protein tag is placed at the amino terminal end of the fusion protein for use in verifying that the genes are in frame and correctly expressed.

In some embodiments, the chimeric polynucleotide contains a therapeutic nucleic acid sequence and a substrate for the nucleic acid binding moiety in the targeted fusion protein, as discussed above.

Therapeutic nucleic acids

Therapeutic approaches using nucleic acids, e.g., oligonucleotides, have been studied in detail. These approaches include small interfering RNA (siRNA) as well as antisense to miRNAs that are overexpressed or miRNA mimics of miRNAs that are reduced in disease. It is widely accepted that delivery of therapeutic oligonucleotides is a major bottleneck in the clinical development of these agents. Oligonucleotides are inherently unstable in circulation. They are difficult to penetrate cell membranes in the absence of lipid transfection agents due to their size and charge. While lipid nanoparticles are the current standard method for oligonucleotide delivery, they possess certain limitations. Composed of synthetic ingredients, lipid nanoparticles will decompose in vivo to produce cytotoxic or immunogenic activities. For example, lipid nanoparticles were shown to produce a variety of toxicities including proinflammatory response and activation of toll-like receptor 4 (Kedmi R, et al.

Biomaterials. 2010 31 :6867-75). The disclosed targeted microvessicles provide a superior method for delivering therapeutic nucleic acids.

In some embodiments, the therapeutic nucleic acid is a heterologous polynucleotide not typically associated with the exosomes. Thus the therapeutic nucleic acid is in some embodiments not normally associated with exosomes.

The therapeutic nucleic acid may be single or double stranded. Non-limiting examples of therapeutic nucleic acid sequences include siRNA, dsRNA, dsDNA, shRNA, mRNA, microRNA, antimicroRNA, antagomir, antimir, antisense, aptamer, and dsRNA/DNA hybrids.

The therapeutic nucleic acid is chosen on the basis of the desired effect on the cell into which it is intended to be delivered and the mechanism by which that effect is to be carried out. For example, the therapeutic nucleic acid may be useful in gene therapy, for example in order to express a desired gene in a cell or group of cells. Such nucleic acid is typically in the form of plasmid DNA or viral vector encoding the desired gene and operatively linked to appropriate regulatory sequences such as promoters, enhancers and the like such that the plasmid DNA is expressed once it has been delivered to the cells to be treated. Examples of diseases susceptible to gene therapy include haemophilia B (Factor IX), cystic fibrosis (CTFR) and spinal muscular atrophy (SMN-1).

Therapeutic nucleic acid can also be used for example in immunization to express one or more antigens against which it is desired to produce an immune response. Thus, the therapeutic nucleic acid can encode one or more antigens against which is desired to produce an immune response, including but not limited to tumor antigens, antigens from pathogens such as viral, bacterial or fungal pathogens.

The therapeutic nucleic acid can also be used in gene silencing. Such gene silencing may be useful in therapy to switch off aberrant gene expression or in animal model studies to create single or more genetic knock outs. The therapeutic nucleic acid molecules can act as effectors, inhibitors, modulators, and stimulators of a specific activity possessed by a target molecule, or the therapeutic nucleic acid molecules can possess a de novo activity independent of any other molecules.

Therapeutic nucleic acid molecules can interact with any macromolecule, such as DNA, RNA, polypeptides, or carbohydrate chains. Often therapeutic nucleic acids are designed to interact with other nucleic acids based on sequence homology between the target molecule and the therapeutic nucleic acid molecule. In other situations, the specific recognition between the therapeutic nucleic acid molecule and the target molecule is not based on sequence homology between the therapeutic nucleic acid molecule and the target molecule, but rather is based on the formation of tertiary structure that allows specific recognition to take place.

Antisense molecules are designed to interact with a target nucleic acid molecule through either canonical or non-canonical base pairing. The interaction of the antisense molecule and the target molecule is designed to promote the destruction of the target molecule through, for example, RNAseH mediated RNA-DNA hybrid degradation. Alternatively the antisense molecule is designed to interrupt a processing function that normally would take place on the target molecule, such as transcription or replication. Antisense molecules can be designed based on the sequence of the target molecule. Numerous methods for optimization of antisense efficiency by finding the most accessible regions of the target molecule exist.

Exemplary methods would be in vitro selection experiments and DNA modification studies using DMS and DEPC. It is preferred that antisense molecules bind the target molecule with a dissociation constant (Kj) less than or equal to 10^~6, 10^~8, 10^~10, or 10^{" 12}. A representative sample of methods and techniques which aid in the design and use of antisense molecules can be found in U.S. Patent Nos. 5,135,917, 5,294,533, 5,627,158, 5,641,754, 5,691,317, 5,780,607, 5,786,138, 5,849,903, 5,856,103, 5,919,772, 5,955,590, 5,990,088, 5,994,320, 5,998,602, 6,005,095, 6,007,995, 6,013,522, 6,017,898, 6,018,042, 6,025,198, 6,033,910, 6,040,296, 6,046,004, 6,046,319, and 6,057,437.

Aptamers are molecules that interact with a target molecule, preferably in a specific way. Typically aptamers are small nucleic acids ranging from 15-50 bases in length that fold into defined secondary and tertiary structures, such as stem-loops or G-quartets. Aptamers can bind small molecules, such as ATP (U.S. Patent No.

5,631,146) and theophiline (U.S. Patent No. 5,580,737), as well as large molecules, such as reverse transcriptase (U.S. Patent No. 5,786,462) and thrombin (United States patent 5,543,293). Aptamers can bind very tightly with Ka's from the target molecule of less than 10-12 M. It is preferred that the aptamers bind the target molecule with a ¾ less than 10^"6, 10^"8, 10^"10, or 10^"12. Aptamers can bind the target molecule with a very high degree of specificity. For example, aptamers have been isolated that have greater than a 10,000 fold difference in binding affinities between the target molecule and another molecule that differ at only a single position on the molecule (U.S. Patent No. 5,543,293). It is preferred that the aptamer have a Ka with the target molecule at least 10, 100, 1000, 10,000, or 100,000 fold lower than the ¾ with a background binding molecule. It is preferred when doing the comparison for a polypeptide for example, that the background molecule be a different polypeptide. Representative examples of how to make and use aptamers to bind a variety of different target molecules can be found in U.S. Patent Nos. 5,476,766, 5,503,978, 5,631,146, 5,731,424 , 5,780,228, 5,792,613, 5,795,721, 5,846,713, 5,858,660 , 5,861,254, 5,864,026, 5,869,641, 5,958,691, 6,001,988, 6,011,020, 6,013,443, 6,020,130, 6,028,186, 6,030,776, and 6,051,698.

Gene expression can also be effectively silenced in a highly specific manner through RNA interference (RNAi). This silencing was originally observed with the addition of double stranded RNA (dsRNA) (FireA, et al. (1998) Nature, 391 :806-11; Napoli, C, et al. (1990) Plant Cell 2:279-89; Hannon, G.J. (2002) Nature, 418:244- 51). Once dsRNA enters a cell, it is cleaved by an RNase III -like enzyme, Dicer, into double stranded small interfering RNAs (siRNA) 21-23 nucleotides in length that contains 2 nucleotide overhangs on the 3' ends (Elbashir, S.M., et al. (2001) Genes Dev., 15: 188-200; Bernstein, E., et al. (2001) Nature, 409:363-6; Hammond, S.M., et al. (2000) Nature, 404:293-6). In an ATP dependent step, the siRNAs become integrated into a multi-subunit protein complex, commonly known as the RNAi induced silencing complex (RISC), which guides the siRNAs to the target RNA sequence (Nykanen, A., et al. (2001) Cell, 107:309-21). At some point the siRNA duplex unwinds, and it appears that the antisense strand remains bound to RISC and directs degradation of the complementary m NA sequence by a combination of endo and exonucleases (Martinez, J., et al. (2002) Cell, 110:563-74). However, the effect of iR A or siR A or their use is not limited to any type of mechanism.

Short Interfering RNA (siRNA) is a double-stranded RNA that can induce sequence-specific post-transcriptional gene silencing, thereby decreasing or even inhibiting gene expression. In one example, an siRNA triggers the specific

degradation of homologous RNA molecules, such as mRNAs, within the region of sequence identity between both the siRNA and the target RNA. For example, WO 02/44321 discloses siRNAs capable of sequence-specific degradation of target mRNAs when base-paired with 3' overhanging ends, herein incorporated by reference for the method of making these siRNAs. Sequence specific gene silencing can be achieved in mammalian cells using synthetic, short double-stranded RNAs that mimic the siRNAs produced by the enzyme dicer (Elbashir, S.M., et al. (2001) Nature, 411 :494 498) (Ui-Tei, K., et al. (2000) FEBS Lett 479:79-82). siRNA can be chemically or in vitro-synthesized or can be the result of short double-stranded hairpin-like RNAs (shRNAs) that are processed into siRNAs inside the cell. Synthetic siRNAs are generally designed using algorithms and a conventional DNA/RNA synthesizer. Suppliers include Ambion (Austin, Texas), ChemGenes (Ashland, Massachusetts), Dharmacon (Lafayette, Colorado), Glen Research (Sterling,

Virginia), MWB Biotech (Esbersberg, Germany), Proligo (Boulder, Colorado), and Qiagen (Vento, The Netherlands). siRNA can also be synthesized in vitro using kits such as Ambion' s SILENCER® siRNA Construction Kit.

The production of siRNA from a vector is more commonly done through the transcription of a short hairpin RNAs (shRNAs). Kits for the production of vectors comprising shRNA are available, such as, for example, Imgenex's

GENESUPPRESSOR™ Construction Kits and Invitrogen's BLOCK-IT™ inducible RNAi plasmid and lentivirus vectors. Disclosed herein are any shRNA designed as described above based on the sequences for the herein disclosed inflammatory mediators.

microRNAs (miRNAs) are small, regulatory noncoding RNAs. miRNA genes are often located within introns of coding or noncoding genes and have also been identified in exons and intergenic regions (Kim VN, et al. Trends Genet. 2006 22:165- 73). Endogenous miRNAs are transcribed by RNA polymerase II into a long primary transcript or pri-miRNA. The pri-miRNA is processed to a ~ 75 nt pre -miRNA by the ribonucleoprotein complex Drosha/DGCR8. Both the pri- and pre-miRNA contain the characteristic hairpin structure. Following cytoplasmic transport by exportin 5, the pre-miRNA is loaded into the Dicer complex which removes the loop of the hairpin. The duplex miRNA, is loaded into the miRISC complex and the strand with the poorer 5' end stability is removed (Schwarz DS, et al. Cell. 2003 115: 199-208). The complex then scans messenger RNA to locate the miRNA' s target. Binding of the mature miRNA (via complete hybridization of the 7 nt 5 ' seed sequence) typically occurs in the 3' UTR of mRNA and results in translational repression. Altered miRNA expression has been observed in all cancers studied to date. miRNA may be oncogenic or tumor suppressive depending upon the miRNA, its' expression level and the type of cancer. Much has been learned in the past 10 years regarding the role of miRNA in HCC, reviewed in (Braconi C, et al. Seminars in oncology. 2011 38:752- 63). As is true of most cancers, certain miRNAs have increased expression in the tumors of patients with HCC including miR-221 (Budhu A, et al. Hepatology. 2008 47:897-907; Gramantieri L, et al. Cancer Res. 2007 67:6092-9; Jiang J, et al. Clin Cancer Res. 2008 14:419-27; Pineau P, et al. Proc Natl Acad Sci U S A. 2009; Wang Y, et al. J Biol Chem. 2008 283: 13205-15), miR-21 (Budhu A, et al. Hepatology. 2008 47:897-907; Jiang J, et al. Clin Cancer Res. 2008 14:419-27; Meng F, et al.

Gastroenterology. 2007 133:647-58; Pineau P, et al. Proc Natl Acad Sci U S A. 2009), and miR-181b (Ji J, et al. Hepatology. 2009 50:472-80; Wang B, et al. Oncogene. 2010 29(12): 1787-97). Primary HCC tumors had reduced expression of other miRNAs such as miR-199a-3p (miR-199a*) (Jiang J, et al. Clin Cancer Res. 2008 14:419-27; Murakami Y, et al. Oncogene. 2006 25:2537-45; Wang Y, et al. J Biol Chem. 2008 283: 13205-15), miR-122 (Bai , et al. J Biol Chem. 2009 284:32015-27; Coulouarn C, et al. Oncogene. 2009 28:3526-36; Fornari F, et al. Cancer Res. 2009 69:5761-7; Kutay H, et al. J Cell Biochem. 2006 99:671-8) and miR-26a (Chen L, et al. Molecular therapy: the journal of the American Society of Gene Therapy. 2011 19: 1521-8).

Antagomirs are a specific class of miRNA antagonists that are used to silence endogenous microRNA. For example, custom designed Dharmacon meridian™ microRNA Hairpin Inhibitors are commercially available from Thermo Scientific. These inhibitors include chemical modifications and secondary structure motifs. Specifically, incorporation of highly structured, double-stranded flanking regions around the reverse complement core significantly increases inhibitor function and allows for multi-miRNA inhibition at subnanomolar concentrations. Other such improvements in antagomir design are contemplated for use in the disclosed methods.

As discussed above, the nucleic acid binding moiety of the exosome targeted fusion protein and the substrate for the nucleic acid binding moiety in the chimeric polynucleotide can be any amino acid/nucleic acid pair where the amino acid sequence is capable of specifically binding the nucleic acid sequence. For example, in some embodiments, the nucleic acid binding moiety comprises HIV Tat protein, or fragment or variant thereof capable of binding an HIV trans-activation response (TAR) element, and the substrate for the nucleic acid binding moiety comprises an HIV TAR element, or a fragment or variant thereof capable of binding an HIV Tat protein.

In some embodiments, the substrate is a nucleic acid aptamer capable of binding the nucleic acid binding moiety of the exosome targeted fusion protein. The term "aptamer" refers to oligonucleic acid or peptide molecules that bind to a specific target molecule. These molecules are generally selected from a random sequence pool. The selected aptamers are capable of adapting unique tertiary structures and recognizing target molecules with high affinity and specificity. A "nucleic acid aptamer" is a DNA or RNA oligonucleic acid that binds to a target molecule via its conformation, and thereby inhibits or suppresses functions of such molecule. For example, a nucleic acid aptamer may be constituted by RNA.

Each of the first nucleic acid sequence and the second nucleic acid sequence can be operably inserted in an expression vector. In some embodiments, the first nucleic acid sequence and second nucleic acid sequence are operably inserted in a common expression vector so they are expressed together. For example, in some embodiments, the second nucleic acid encoding the chimeric polynucleotide is inserted in frame into an intron of the first nucleic acid encoding the exosome targeted fusion protein. Methods to construct expression vectors containing genetic sequences and appropriate transcriptional and translational control elements are well known in the art. These methods include in vitro recombinant DNA techniques, synthetic techniques, and in vivo genetic recombination. Such techniques are described in Sambrook et al., Molecular Cloning, A Laboratory Manual (Cold Spring Harbor Press, Plainview, N.Y., 1989), and Ausubel et al, Current Protocols in Molecular Biology (John Wiley & Sons, New York, N.Y., 1989).

Vectors include but are not limited to plasmids, viral nucleic acids, viruses, phage nucleic acids, phages, cosmids, and artificial chromosomes. Expression vectors generally contain regulatory sequences necessary elements for the translation and/or transcription of the inserted coding sequence. For example, the coding sequence is preferably operably linked to a promoter and/or enhancer to help control the expression of the desired gene product. Promoters used in biotechnology are of different types according to the intended type of control of gene expression. They can be generally divided into constitutive promoters, tissue-specific or development- stage-specific promoters, inducible promoters, and synthetic promoters.

Depending on the vector system and host utilized, any number of suitable transcription and translation elements may be used. In mammalian cell systems, promoters from mammalian genes or from mammalian viruses are preferable.

Vectors based on SV40 or EBV may be used with an appropriate selectable marker to generate a cell line that contains multiple copies of the sequence encoding a polypeptide.

Also disclosed is an exosome producing cell engineered to contain the first nucleic acid sequence and the second nucleic acid sequence of the disclosed system.

Exosomes are produced by many different types of cells including immune cells such as B lymphocytes, T lymphocytes, dendritic cells (DCs) and most cells. Exosomes are also produced, for example, by glioma cells, platelets, reticulocytes, neurons, intestinal epithelial cells and tumor cells. Exosomes for use in the disclosed compositions and methods can be derived from any suitable cell, including the cells identified above. Exosomes have also been isolated from physiological fluids, such as plasma, urine, amniotic fluid and malignant effusions. Non-limiting examples of suitable exosome producing cells for mass production include dendritic cells (e.g., immature dendritic cell), Human Embryonic Kidney 293 (HEK) cells, 293T cells, Chinese hamster ovary (CHO) cells, and human ESC-derived mesenchymal stem cells.

In some embodiments, exosomes are derived from DCs, such as immature DCs. Exosomes produced from immature DCs do not express MHC-II, MHC-I or CD86. As such, such these exosomes do not stimulate na^'ive T cells to a significant extent and are unable to induce a response in a mixed lymphocyte reaction. Thus exosomes produced from immature dendritic cells can be used for use in delivery of genetic material.

Exosomes can also be obtained from any autologous patient-derived, heterologous haplotype-matched or heterologous stem cells so to reduce or avoid the generation of an immune response in a patient to whom the exosomes are delivered. Any exosome-producing cell can be used for this purpose.

Exosomes produced from cells can be collected from the culture medium by any suitable method. Typically a preparation of exosomes can be prepared from cell culture or tissue supernatant by centrifugation, filtration or combinations of these methods. For example, exosomes can be prepared by differential centrifugation, that is low speed (<20000 g) centrifugation to pellet larger particles followed by high speed (> 100000 g) centrifugation to pellet exosomes, size filtration with appropriate filters (for example, 0.22 μιη filter), gradient ultracentrifugation (for example, with sucrose gradient) or a combination of these methods.

The disclosed exosomes may be administered to a subject by any suitable means. Administration to a human or animal subject may be selected from parenteral, intramuscular, intracerebral, intravascular, subcutaneous, or transdermal

administration. Typically the method of delivery is by injection. Preferably the injection is intramuscular or intravascular (e.g. intravenous). A physician will be able to determine the required route of administration for each particular patient.

The exosomes are preferably delivered as a composition. The composition may be formulated for parenteral, intramuscular, intracerebral, intravascular

(including intravenous), subcutaneous, or transdermal administration. Compositions for parenteral administration may include sterile aqueous solutions which may also contain buffers, diluents and other suitable additives. The exosomes may be formulated in a pharmaceutical composition, which may include pharmaceutically acceptable carriers, thickeners, diluents, buffers, preservatives, and other

pharmaceutically acceptable carriers or excipients and the like in addition to the exosomes.

Parenteral administration is generally characterized by injection, such as subcutaneously, intramuscularly, or intravenously. Preparations for parenteral administration include sterile solutions ready for injection, sterile dry soluble products, such as lyophilized powders, ready to be combined with a solvent just prior to use, including hypodermic tablets, sterile suspensions ready for injection, sterile dry insoluble products ready to be combined with a vehicle just prior to use and sterile emulsions. The solutions may be either aqueous or nonaqueous.

If administered intravenously, suitable carriers include physiological saline or phosphate buffered saline (PBS), and solutions containing thickening and solubilizing agents, such as glucose, polyethylene glycol, and polypropylene glycol and mixtures thereof. Pharmaceutically acceptable carriers used in parenteral preparations include aqueous vehicles, nonaqueous vehicles, antimicrobial agents, isotonic agents, buffers, antioxidants, local anesthetics, suspending and dispersing agents, emulsifying agents, sequestering or chelating agents and other pharmaceutically acceptable substances. Examples of aqueous vehicles include sodium chloride injection, ringers injection, isotonic dextrose injection, sterile water injection, dextrose and lactated ringers injection. Nonaqueous parenteral vehicles include fixed oils of vegetable origin, cottonseed oil, corn oil, sesame oil and peanut oil. Antimicrobial agents in

bacteriostatic or fungistatic concentrations must be added to parenteral preparations packaged in multiple-dose containers which include phenols or cresols, mercurials, benzyl alcohol, chlorobutanol, methyl and propyl p-hydroxybenzoic acid esters, thimerosal, benzalkonium chloride and benzethonium chloride. Isotonic agents include sodium chloride and dextrose. Buffers include phosphate and citrate.

Antioxidants include sodium bisulfate. Local anesthetics include procaine

hydrochloride. Suspending and dispersing agents include sodium

carboxymethylcelluose, hydroxypropyl methylcellulose and polyvinylpyrrolidone.

Emulsifying agents include Polysorbate 80 (TWEEN® 80). A sequestering or chelating agent of metal ions include EDTA. Pharmaceutical carriers also include ethyl alcohol, polyethylene glycol and propylene glycol for water miscible vehicles; and sodium hydroxide, hydrochloric acid, citric acid or lactic acid for pH adjustment. The concentration of the pharmaceutically active compound is adjusted so that an injection provides an effective amount to produce the desired pharmacological effect. The exact dose depends on the age, weight and condition of the patient or animal as is known in the art.

The unit-dose parenteral preparations can be packaged in an ampoule, a vial or a syringe with a needle. All preparations for parenteral administration should be sterile, as is known and practiced in the art.

A therapeutically effective amount of composition is administered. The dose may be determined according to various parameters, especially according to the severity of the condition, age, and weight of the patient to be treated; the route of administration; and the required regimen. A physician will be able to determine the required route of administration and dosage for any particular patient. Optimum dosages may vary depending on the relative potency of individual constructs, and can generally be estimated based on EC₅₀s found to be effective in vitro and in vivo animal models. In general, dosage is from 0.01 mg/kg to 100 mg per kg of body weight. A typical daily dose is from about 0.1 to 50 mg per kg, preferably from about 0.1 mg/kg to 10 mg/kg of body weight, according to the potency of the specific construct, the age, weight and condition of the subject to be treated, the severity of the disease and the frequency and route of administration. Different dosages of the construct may be administered depending on whether administration is by

intramuscular injection or systemic (intravenous or subcutaneous) injection.

Preferably, the dose of a single intramuscular injection is in the range of about 5 to 20 μg. Preferably, the dose of single or multiple systemic injections is in the range of 10 to 100 mg/kg of body weight.

Due to construct clearance (and breakdown of any targeted molecule), the patient may have to be treated repeatedly, for example once or more daily, weekly, monthly or yearly. Persons of ordinary skill in the art can easily estimate repetition rates for dosing based on measured residence times and concentrations of the construct in bodily fluids or tissues. Following successful treatment, it may be desirable to have the patient undergo maintenance therapy, wherein the construct is administered in maintenance doses, ranging from 0.01 mg/kg to 100 mg per kg of body weight, once or more daily, to once every 20 years.

The term "primary precursor miR A" or "pri-miRNA" refers to the form of miRNA that is transcribed from the gene.

The term "precursor miRNA" or "pre-miRNA" refers to the approximately

75nt miRNA hairpin that results from Drosha processing of the pri-miRNA.

The term "mature miRNA" or "miRNA" refers to the approximately 21nt active miRNA that is processed from the pre-miRNAby Dicer and miRISC.

The term "Drosha" refers to an enzyme capable of processing pri-miRNA into pre-miRNA.

The term "Dicer" refers to an enzyme capable of processing pre-miRNA to mature miRNA by removing the loop precursor.

The term "miRNA silencing complex" or "miRISC" refers to a complex capable of degrading the passenger strand of an miRNA duplex.

The term "guide strand" refers to the active miRNA strand that is not degraded by miRISC.

The term "passenger strand" refers to the inactive strand of duplex miRNA that is degraded by miRISC.

The term "subject" refers to any individual who is the target of administration or treatment. The subject can be a vertebrate, for example, a mammal. Thus, the subject can be a human or veterinary patient. The term "patient" refers to a subject under the treatment of a clinician, e.g., physician.

The term "therapeutically effective" refers to the amount of the composition used is of sufficient quantity to ameliorate one or more causes or symptoms of a disease or disorder. Such amelioration only requires a reduction or alteration, not necessarily elimination.

The term "pharmaceutically acceptable" refers to those compounds, materials, compositions, and/or dosage forms which are, within the scope of sound medical judgment, suitable for use in contact with the tissues of human beings and animals without excessive toxicity, irritation, allergic response, or other problems or complications commensurate with a reasonable benefit/risk ratio. The term "carrier" means a compound, composition, substance, or structure that, when in combination with a compound or composition, aids or facilitates preparation, storage, administration, delivery, effectiveness, selectivity, or any other feature of the compound or composition for its intended use or purpose. For example, a carrier can be selected to minimize any degradation of the active ingredient and to minimize any adverse side effects in the subject.

The term "treatment" refers to the medical management of a patient with the intent to cure, ameliorate, stabilize, or prevent a disease, pathological condition, or disorder. This term includes active treatment, that is, treatment directed specifically toward the improvement of a disease, pathological condition, or disorder, and also includes causal treatment, that is, treatment directed toward removal of the cause of the associated disease, pathological condition, or disorder. In addition, this term includes palliative treatment, that is, treatment designed for the relief of symptoms rather than the curing of the disease, pathological condition, or disorder; preventative treatment, that is, treatment directed to minimizing or partially or completely inhibiting the development of the associated disease, pathological condition, or disorder; and supportive treatment, that is, treatment employed to supplement another specific therapy directed toward the improvement of the associated disease, pathological condition, or disorder.

The term "biocompatible" generally refers to a material and any metabolites or degradation products thereof that are generally non-toxic to the recipient and do not cause any significant adverse effects to the subject.

The term "biodegradable" generally refers to a material that will degrade or erode under physiologic conditions to smaller units or chemical species that are capable of being metabolized, eliminated, or excreted by the subject. The degradation time is a function of polymer composition and morphology. Suitable degradation times are from days to months.

The term "antibody" refers to natural or synthetic antibodies that selectively bind a target antigen. The term includes polyclonal and monoclonal antibodies. In addition to intact immunoglobulin molecules, also included in the term "antibodies" are fragments or polymers of those immunoglobulin molecules, and human or humanized versions of immunoglobulin molecules that selectively bind the target antigen.

The terms "peptide," "protein," and "polypeptide" are used interchangeably to refer to a natural or synthetic molecule comprising two or more amino acids linked by the carboxyl group of one amino acid to the alpha amino group of another.

The term "protein domain" refers to a portion of a protein, portions of a protein, or an entire protein showing structural integrity; this determination may be based on amino acid composition of a portion of a protein, portions of a protein, or the entire protein.

The term "nucleic acid" refers to a natural or synthetic molecule comprising a single nucleotide or two or more nucleotides linked by a phosphate group at the 3 ' position of one nucleotide to the 5 ' end of another nucleotide. The nucleic acid is not limited by length, and thus the nucleic acid can include deoxyribonucleic acid (DNA) or ribonucleic acid (R A).

The term "specifically binds", as used herein, when referring to a polypeptide

(including antibodies) or receptor, refers to a binding reaction which is determinative of the presence of the protein or polypeptide or receptor in a heterogeneous population of proteins and other biologies. Thus, under designated conditions (e.g. immunoassay conditions in the case of an antibody), a specified ligand or antibody "specifically binds" to its particular "target" (e.g. an antibody specifically binds to an endothelial antigen) when it does not bind in a significant amount to other proteins present in the sample or to other proteins to which the ligand or antibody may come in contact in an organism. Generally, a first molecule that "specifically binds" a second molecule has an affinity constant (Ka) greater than about 10⁵ Μ ¹ (e.g., 10⁶ M^-1, 10⁷ ΜΓ¹, lO⁸ IVT¹, 10⁹ IVT¹, 10¹⁰ IVT¹, 10¹¹ IVT¹, and 10¹² IVT¹ or more) with that second molecule.

A "chimeric molecule" is a single molecule created by joining two or more molecules that exist separately in their native state. The single, chimeric molecule has the desired functionality of all of its constituent molecules. Frequently, one of the constituent molecules of a chimeric molecule is a "targeting molecule" or "targeting moiety." The targeting molecule is a molecule such as a ligand or an antibody that specifically binds to its corresponding target, for example a receptor on a cell surface. A "fusion protein" refers to a polypeptide formed by the joining of two or more polypeptides through a peptide bond formed between the amino terminus of one polypeptide and the carboxyl terminus of another polypeptide. The fusion protein can be formed by the chemical coupling of the constituent polypeptides or it can be expressed as a single polypeptide from nucleic acid sequence encoding the single contiguous fusion protein. A single chain fusion protein is a fusion protein having a single contiguous polypeptide backbone. Fusion proteins can be prepared using conventional techniques in molecular biology to join the two genes in frame into a single nucleic acid, and then expressing the nucleic acid in an appropriate host cell under conditions in which the fusion protein is produced.

The term "specifically deliver" as used herein refers to the preferential association of a molecule with a cell or tissue bearing a particular target molecule or marker and not to cells or tissues lacking that target molecule. It is, of course, recognized that a certain degree of non-specific interaction may occur between a molecule and a non- target cell or tissue. Nevertheless, specific delivery, may be distinguished as mediated through specific recognition of the target molecule.

Typically specific delivery results in a much stronger association between the delivered molecule and cells bearing the target molecule than between the delivered molecule and cells lacking the target molecule.

A "spacer" as used herein refers to a peptide that joins the proteins comprising a fusion protein. Generally a spacer has no specific biological activity other than to join the proteins or to preserve some minimum distance or other spatial relationship between them. However, the constituent amino acids of a spacer may be selected to influence some property of the molecule such as the folding, net charge, or hydrophobicity of the molecule.

The term "vector" or "construct" refers to a nucleic acid sequence capable of transporting into a cell another nucleic acid to which the vector sequence has been linked. The term "expression vector" includes any vector, (e.g., a plasmid, cosmid or phage chromosome) containing a gene construct in a form suitable for expression by a cell (e.g., linked to a transcriptional control element).

The term "operably linked to" refers to the functional relationship of a nucleic acid with another nucleic acid sequence. Promoters, enhancers, transcriptional and translational stop sites, and other signal sequences are examples of nucleic acid sequences operably linked to other sequences. For example, operable linkage of DNA to a transcriptional control element refers to the physical and functional relationship between the DNA and promoter such that the transcription of such DNA is initiated from the promoter by an RNA polymerase that specifically recognizes, binds to and transcribes the DNA.

"Polypeptide" as used herein refers to any peptide, oligopeptide, polypeptide, gene product, expression product, or protein. A polypeptide is comprised of consecutive amino acids. The term "polypeptide" encompasses naturally occurring or synthetic molecules.

As used herein, the term "amino acid sequence" refers to a list of

abbreviations, letters, characters or words representing amino acid residues. The amino acid abbreviations used herein are conventional one letter codes for the amino acids and are expressed as follows: A, alanine; B, asparagine or aspartic acid; C, cysteine; D aspartic acid; E, glutamate, glutamic acid; F, phenylalanine; G, glycine; H histidine; I isoleucine; K, lysine; L, leucine; M, methionine; N, asparagine; P, proline; Q, glutamine; R, arginine; S, serine; T, threonine; V, valine; W, tryptophan; Y, tyrosine; Z, glutamine or glutamic acid.

The term "variant" refers to an amino acid or peptide sequence having conservative amino acid substitutions, non-conservative amino acid subsitutions (i.e. a degenerate variant), substitutions within the wobble position of each codon (i.e. DNA and RNA) encoding an amino acid, amino acids added to the C-terminus of a peptide, or a peptide having 60%, 70%, 80%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%o, or 99%) percent identity to a reference sequence.

The term "percent (%) sequence identity" or "homology" is defined as the percentage of nucleotides or amino acids in a candidate sequence that are identical with the nucleotides or amino acids in a reference nucleic acid sequence, after aligning the sequences and introducing gaps, if necessary, to achieve the maximum percent sequence identity. Alignment for purposes of determining percent sequence identity can be achieved in various ways that are within the skill in the art, for instance, using publicly available computer software such as BLAST, BLAST-2,

ALIGN, ALIGN-2 or Megalign (DNASTAR) software. Appropriate parameters for measuring alignment, including any algorithms needed to achieve maximal alignment over the full-length of the sequences being compared can be determined by known methods.

A number of embodiments of the invention have been described.

Nevertheless, it will be understood that various modifications may be made without departing from the spirit and scope of the invention. Accordingly, other embodiments are within the scope of the following claims.

EXAMPLES

Example 1:

Disclosed is a microvessicle-based system to synthesize and deliver therapeutic, nucleic acid cargo. A nucleic acid delivery system using MVs derived from dendritic cells has been described (Alvarez-Erviti L, et al. Nat Biotechnol. 2011 29:341-5). Mouse dendritic cells were transfected with a vector expressing a protein engineered to contain an N-terminal FLAG purification tag adjacent to a targeting peptide sequence (e.g. CNS-specific rabies viral glycoprotein). At the C-terminus was Lamp-2b, a protein found abundantly in MV membranes (Fig. 1 A). Lamp-2b directed the targeting protein to the membrane of MVs produced when these dendritic cells are cultured in the presence of GM-CSF. The engineered MVs are pulled down from the pool of dendritic cell MVs using anti-FLAG-tagged beads (Alvarez-Erviti L, et al. Nat Biotechnol. 2011 29:341-5). The purified MVs were electroporated with siRNA oligonucleotides. Intravenous injection of the targeting MVs localized to mouse brain delivering fully function siRNA in the absence of lipid transfection reagents (Alvarez-Erviti L, et al. Nat Biotechnol. 2011 29:341-5). Of note, greater amounts of the MVs were delivered to the brain compared to the liver, kidney and spleen, three highly perfused organs that typically accumulate large amounts of oligonucleotides and lipid nanoparticles. Engineered MVs have also been used to target a let-7 miRNA mimic to MVs expressing an EGFR targeting peptide (Ohno SI, et al. Mol Ther. 2013 21(1):185-91).

The disclosed technology overcomes two issues that were apparent in these studies. Both studies transfected therapeutic nucleic acids into the purified MVs.

Secondly, they used synthetic oligonucleotides, which are costly for the doses needed to treat animals and humans. To overcome these shortcomings, a targeted MV delivery system loaded with nucleic acid that is synthesized by the cells that produce the MVs is disclosed. An overview of the disclosed technology is shown in Fig. 2. The targeting protein gene can be modified to incorporate an intron containing a modified pri-miRNA gene (Fig. 1 A).

The modified pri-miRNA can be spliced from the targeting protein mRNA and undergo natural miRNA biogenesis (e.g. processed by Drosha to pre-miRNA, exported to the cytoplasm and processed by cytoplasmic Dicer to the mature miRNA). Processing can occur as producer cells express mature miRNA that was processed from intronic pri-miRNA genes. Human MVs contain pre -miRNA. An innovative feature of this delivery system is that the purified MVs will contain the therapeutic pre-miRNA in abundance. To achieve this specificity, the pre-miRNA gene can be engineered to contain modified nucleotides in the loop region (Fig. 1B,C). These changes can allow the loop portion of the pre-miRNA to resemble the HIV-1 transactivation response (TAR) RNA. The interaction between TAR RNA and the HIV-1 transactivator protein Tat has been widely studied (Aboul-ela F, et al. J Mol Biol. 1995 253:313-32; Bigalke JM, et al. Methods. 2011 53:78-84; Ippolito JA, et al. Proc Natl Acad Sci U S A. 1998 95:9819-24). Cytoplasmic, modified pre-miR-199a- 2 can be directed to the engineered MVs through biding of its' loop region to a portion of the Tat protein engineered to the luminal C-terminus of the targeting protein (Fig. ID). The miRNA loaded targeting MVs can be purified by anti-FLAG tagged magnetic beads. This design overcomes two major hurdles of cellular produced delivery systems (J) purification of the engineered MVs from the cellular MVs and (ii) loading of the pre-miRNA of interest to the engineered MVs.

This project addresses an important question in contemporary biomedical research, namely the optimal delivery of therapeutic nucleic acids. The technology is highly innovative as both the delivery device and the therapeutic nucleic acid are synthesized by cells. miRNA-loaded MVs are a natural component of the

bloodstream and are comprised of non-synthetic and non-viral components.

Therefore, this delivery system alleviates shortcomings of viral or lipid nanoparticle- based nucleic acid delivery systems. Their small size allows them to cross major biological membranes and the lipid bilayer membrane protects the therapeutic nucleic acids from degradation.

miR-199a-3p was selected as the model nucleic acid MVs for a number of important reasons. Many studies have reported reduced miR-199a-3p in HCC (Jiang J, et al. Clin Cancer Res. 2008 14:419-27; Murakami Y, et al. Oncogene. 2006 25:2537-45; Wang Y, et al. J Biol Chem. 2008 283: 13205-15). Next generation sequencing of human HCC determined that miR-199a-3p was one of the most important differentially expressed miRNAs in HCC (Hou J, et al. Cancer Cell. 2011 19:232-43). miR-199a-3p acts as a key mediator to regulate the expression of multiple proliferation-related genes in HCC (Huang Y, et al. Nucleic Acids Res. 2012 40(20): 10478-93). miR-199a-3p has been shown to inhibit replication of both HCV (Murakami Y, et al. J Hepatol. 2009 50:453-60) and HBV (Zhang GL, et al. Antiviral research. 2010 88: 169-75). Only the aggressive, CD44+ positive HCC cell lines are sensitive to the anti-pro liferative and anti-invasive properties of miR-199a-3p oligonucleotide mimic (Henry JC, et al. Biochem Biophys Res Commun. 2010

403: 120-5). This is a result of miR-199a-3p directly targeting the CD44 3' UTR and suppressing both CD44 mRNA and protein (Henry JC, et al. Biochem Biophys Res Commun. 2010 403: 120-5). As CD44 is an important marker for collection of HCC stem cells from patients (Zhu Z, et al. International Journal of Cancer. 2010 126:2067- 78), conceivably miR-199a-3p may be used to target the HCC cancer stem cell population.

miR-199a-3p is the mature form that is processed from the pri-miR-199a-2. To allow binding to the Tat peptide and MV loading, the wild type loop of the pre- miR-199a-2 can be swapped with 15 nts from the loop of the TAR RNA (Fig. 1C). While this modification does not change the sequence of the active, mature miRNA, the sequence and shape of the pre-miRNA can be slightly modified. To predict binding between the modified pre-miR-199a-2 and Tat peptide in silico, the interaction was computationally modeled through Rosetta and manual manipulation of NMR structures. A molecular dynamics simulation of the final model was conducted to determine the stability of the structure and to analyze the key

interactions between the Tat peptide and the RNA. The key interactions between Tat peptide and the RNA loop involve Arg3 and Arg6 sandwiching U22 of the UCU bulge. Additionally, Arg3 forms hydrogen bonds with G27, while Arg6 forms hydrogen bonds with G25. All of these interactions are stable throughout the entire molecular dynamics simulation and the peptide remains bound in the major groove of the R A hairpin. Importantly, the stem portion of the R A is not bound to peptide allowing the option to potentially insert any pre-miRNA sequence into the targeting vector without effecting peptide binding. The interaction between the Tat peptide and the modified pre-miR-199a-2 was validated by EMS A gel shift assays. T7 in vitro transcribed and ³²P labeled pre-miR-199a-3p (synthesized from the PCR-generated DNA template) was incubated with increasing concentrations of the HPLC purified 14 amino acid Tat peptide (LifeTein, South Plainfield, NJ). To confer specificity, reactions were incubated with an excess of tRNA. The reactions were resolved on non-denaturing polyacrylamide gels. Tat peptide resulted in a shift in the mobility of the wild type and modified pre-miR-199a-2, conferring binding (Fig.3).

The proposed technology is designed such that the pre-miRNA loaded MVs will be delivered into the cytoplasm of the recipient cells. Once inside the cells, the peptide bound pre-miRNA can be liberated from the MV by Dicer processing.

Active, mature miRNA can be produced from the protein bound pre-miRNA following Dicer processing. Since the peptide can be bound predominately to the bulge of the pre-miRNA (Fig. 2), Dicer cleavage can release the miRNA from the targeting protein and allow it to be processed to mature miRNA via miRISC. The current model of Dicer processing postulates that Dicer anchors on the base of the hairpin's stem portion and the cleavage site is selected by measuring from the 5' end of the dsRNA (Park JE, et al. Nature. 2011 475:201-5). Measurement from the base of the RNA is achieved by a 5 ' pocket motif present in human Dicer that recognizes the 5' terminal phosphate group (Park JE, et al. Nature. 2011 475:201-5). For these reasons, modifications made to the pre-miRNA should not affect Dicer processing, as the stem portions are essentially unchanged (Fig. 1C). To test this, rDicer digests of peptide bound and free T7 transcribed modified pre-miRNA199a-2 containing the bulge and loop region of TAR RNA was performed (Fig. 2C). The ~ 22nt mature miRNA was processed from the precursor following a 12 h incubation with rDicer even in the presence of a 50-fold excess of Tat peptide (Fig. 4). This demonstrates that peptide bound to the loop of the modified pre-miRNA does not affect its processing since Dicer anchors from the base of the stem.

The optimal sequences for the modified pre-miR-199a-2 can be developed that allows the pri-miRNA to undergo Drosha processing and generate the mature miRNA following Dicer processing and passenger strand degradation by miRISC. Both in silico and biochemical approaches can be used to optimize the modified pre-miR- 199a-2 sequence. Targeting MVs containing the modified pre-miR-199a-2 can be synthesized from HEK293T cells. MV yield can be optimized and scaled-up to produce large quantities of purified MV for the disclosed in vivo studies. Following successful characterization of the MVs, they can be directed to hepatocellular carcinoma cell lines that express the target of the PC94 peptide. The PC94 targeting or nontargeting control MVs can be evaluated for their activity and targeting ability in vitro.

The next important question to be addressed is whether modifications to the pre-miR-199a-2 alter the Dhrosha processing. Correct processing by Dhrosha is required to give the specific 5 ' and 3 ' ends of the pre-miRNA so that Dicer processing liberates the desired mature miRNA. To perform these experiments, ³²P-labeled, wild type or modified pri-miR-199a-2 can be synthesized by T7 in vitro transcription from PCR-generated DNA templates. Transcripts can contain the loop, mature and guide strand and ~ 20 nts outside of the Dhrosha cleavage site (Fig. IB). Reactions can be processed in vitro using immunoprecipitated FLAG-tagged purified rDhrosha or Dhrosha-DGCR8. Products can be resolved on denaturing PAGE and the size of the products can be determined using 10 nt RNA markers. Products can be sequenced following gel purification, ligation to 5 ' and 3' linker oligos, reverse transcription, amplification and cloning (Han J, et al. Cell. 2006 125:887-901). The sequence that contains validated pre-miRNA 5 ' and 3 ' Dhrosha cleavage steps can be used to proceed. Once it is confirmed that the modifications to the pre-miR-199a-2 will be correctly processed by Dhrosha, the next step is to demonstrate that the pre-miR- 199a-2 can be processed by Dicer and miRISC. This includes Dicer processing at the correct 5' and 3' sites and degradation of the passenger strand (i.e. miR-199a-5p).

RNA oligos phosphorylated at the 5 ' position can be synthesized and HPLC purified by a commercial supplier (IDT or Dharmacon). The ~ 60 nt synthetic, wild type pre- miR-199a-2 or modified pre-miR-199a-2 can be 3' end labeled with [a- P] pCp and calf alkaline phosphatase (Park JE, et al. Nature. 2011 475:201-5). In this manner liberation of the mature miR-199a-3p by Dicer can be visualized when resolved on denaturing PAGE. Addition of an extra nucleotide to the 3' end (in this case

Cytidine) does not interfere with the correct Dicer processing as it was shown that the counting rule from the 5' end of the precursor predominates (Park JE, et al. Nature. 2011 475:201-5). Transcripts can be reacted with rDicer and the presence of the correct mature miRNA can be determined by PAGE and small RNA sequencing. Degradation of the passenger strand can be determined by measuring both the passenger (5p) and guide strands (3p) by qPCR.

The next step can be to determine if the correct processing of the modified pre-miR-199a-2 is achieved ex vivo. The wild type or modified pri-miR-199a-2 genes, including ~ 100 bp upstream and downstream of the hairpin, can be cloned into a vector that expresses the pri-miRNA from a pol II promoter (i.e. BLOCK-iT™ Pol II miR RNAi Expression Vector, Invitrogen). The wild type or modified pri-miR-

199a-2 containing vectors can be transiently transfected into 293T or HeLa cell lines (both lack mature miR-199a-3p and -5p) and the relative expression of the miR- 199a- 3p and -5p can be verified by TaqMan qRT-PCR (Life Technologies) and small RNA sequencing. The ratio of miR-199a-3p to miR-199a-5p in a variety of human tissues, including liver is ~ 80-fold (Lee EJ, et al. RNA. 2007 14:35-42).

A vector similar can be created using standard PCR cloning techniques. The Lamp2b pEGFP-Cl targeting vector was constructed by inserting the Lamp2b gene downstream of the CMV reporter, replacing the GFP in the process (Alvarez-Erviti L, et al. Nat Biotechnol. 2011 29:341-5). The targeting sequence (e.g. RVG) was cloned in between the Xhol and BspEI sites at the N terminus of Lamp2b. The RVG sequence can be removed and cloned in the PC94 targeting peptide into the same Xhol and BspEI restriction sites. This peptide (amino acid sequence SFSIIHTPILPL, SEQ ID NO:4) discovered by phage display specifically binds to HCC tissues and cell lines (Lo A, et al. Mol Cancer Ther. 2008 7:579-89). PC94 has been used as a targeting peptide in a variety of targeted MVs delivery devices for HCC (Ashley CE, et al.

Nature materials. 2011 10:389-97; Lo A, et al. Mol Cancer Ther. 2008 7:579-89;

Toita R, et al. Bioconjugate chemistry. 2012 23: 1494-501). The Tat peptide sequence (RPRGTRGKGRRIRR, SEQ ID NO:5) (Fig. 1 A lower) can be cloned into the C- terminus using standard techniques. DNA sequencing can be used to determine that the vector is in the correct reading frame. The vector can be transfected into

HEK293T cells and the expression of the protein can be detected using anti-Flag antibody and Western blotting.

The therapeutic miR-199a-3p can be synthesized by the producer HEK293T cells from an intron containing the pri-miR-199a-2 gene that has been engineered into the Lamp2b gene. Following transcription, this intron can be spliced from the primary transcript and the resulting mRNA can be translated to produce the targeting protein. Meanwhile, the intron can be processed to liberate the mature miR-199a-3p (Fig. 5). The Lamp2b gene contained within the pEGFP-Cl vector was cloned from cDNA and therefore lacks introns. pri-miR-199a-2 can be cloned into a chimeric Lamp2b intron by the strategy shown in Fig. 5. Exons 3 and 4 of Lamp2b contain unique restriction enzyme sites for Hindlll and Sail, respectively. A 370 bp segment containing the 3' and 5' ends of the Lamp2b exons, 3 and 4, respectively, the intron/exon junctions including the correct splice site and ~ 260 bp of DNM2 intron 15 containing the pri-miR-199a-2 gene including ~ 100 bp upstream and downstream of the hairpin miRNA sites, can be cloned into the plasmid (following digestion with these restriction enzymes). This 370 bp of DNA can be synthesized using a PCR- based method for long DNA segment synthesis (Xiong AS, et al. Nat Protoc. 2006 1 :791-7) or via a commercial vendor. This segment can contain the same Hindlll and Sail restriction enzyme sites at the 5' and 3' ends, respectively that can allow it to be ligated to the pEGFP-Cl vector. Next, it can be confirmed that the vector produces the correct protein and that the mature miR-199a-3p is processed from the intron. The vector can be transfected into HEK293T cells. Correct protein expression can be performed by western blotting using the anti-Flag antibody or an antibody to the Tat peptide. Correct generation of the mature miR-199a-3p can be determined by Northern blotting, qPCR and small RNA sequencing.

Once it has been validated that the correct protein and mature miRNA are expressed from the vector, it can be introduced into producer cells to synthesize the engineered MVs. Primary mouse dendritic cells can be used to produce sufficient

MVs for IV injection into mice. The advantage of dendritic cell MVs include (z) they are immunologically inert and (ii) immature dendritic cells produce large amounts of MVs that lack T-cell activators such as MHC-II and CD86 (Quah BJ, et al. Blood cells, molecules & diseases. 2005 35:94-110). However, dendritic cells are harvested from the bone marrow of C57BL/6 mice and then differentiated by a 4 to 7 day treatment with murine GM-CSF. This process is tedious, time consuming and not amendable to scale up that will give the numbers of MVs for the large preclinical evaluation in Phase II. As an alternative to primary dendritic cells, HEK293T cells can be used to produce the miR A- loaded targeting MVs. Advantages of HEK293T cells include ease of culture and scale up, high MV yield and the ability to stably express the targeting vector. HEK293T cells, a human embryonic kidney cell line, were only one order of magnitude less efficient at producing MVs than the mesenchymal stem cell (Yeo RW, et al. Adv Drug Deliv Rev. 2013 65(3):336-41). MVs derived from HEK293T cells were recently reported to target breast cancer expressing EGF receptor in vivo (Ohno SI, et al. Mol Ther. 2013 21(1): 185-91). To demonstrate the ability to isolate MVs from HEK293T cells, culture media was centrifuged at 120,000 x g for 90 mins. Electron microscopy was used to validate the appearance and approximate size of the MVs (Fig. 6A). Size was confirmed by NanoSight analysis (NanoSight USA, Costa Mesa, CA). This instrument sizes nanoparticles by tracking their motion in real-time by a CCD camera. The size distribution (90%) ranged from 30-150 nm and particle concentration was ~ 10⁶ particles per cm² of culture plastic.

Targeting vector loaded mouse MVs can be prepared by transfecting the targeting vector developed above into the HEK293T cells. The PC94 or nontargeting control vector (Fig. 1) can be transfected into HEK293T cells using TrabsIT LT1 transfection reagent (Minis Bio) or a suitable transfection reagent and selected on

Neomycin. MVs can be collected from the stable HEK293T cells and the miR-199a- 3p or nontargeting control MVs can be further purified using Flag-tagged beads (Alvarez-Erviti L, et al. Nat Biotechnol. 2011 29:341-5). The percentage of Flag- tagged MVs in the total can be estimated using the NanoSight instrument; Flag-tag can be labeled with a fluorescently labeled Anti-flag-tagged antibody and the fluorescence can be read on the NanoSight. The miR-199a-3p concentration per MV can be calculated to allow calculation of the dose to cultured cells and mice. The miR-199a-3p content in the targeting and nontargeting control MVs can be determined by TaqMan qRT-PCR (Schmittgen TD, et al. Nucleic Acids Res. 2004 32:E43). The results can be expressed as copy number of pre-miR-199a-2 per MV; MV concentration can be determined by NanoSight.

Several variables can be modified to optimize the MV yield prior to further evaluation. These include altering the ratio of plasmid to producer cells and evaluating different MV purification methods (ultracentrifugation versus the

ExoQuick-TC™ system (System Biosciences, Mountain View, CA). Different culture times (i.e. days post transfection) and perhaps media optimization (i.e.

differing the amount of fetal bovine serum) can be attempted as well. Once these variables have been optimized, production can be scaled up into 1,800 cm² roller bottles. Based upon a yield of 10⁶ MVs / cm² of cultured HEK293T cells, it is predicted that 16>< 10⁹ MVs can be produced using 4, 1,800 cm² roller bottles.

Assuming a 15% yield from the Flagtag purification (Ohno SI, et al. Mol Ther. 2013 21(1): 185-91), this would give 2.4 x lO⁹ MVs 4, 1,800 cm² roller bottles. The number of copies of miR-199a-3p loaded per MV can then be determined by qPCR.

Next, whether the targeting vector MVs are functional in vitro and whether they are directed to HCC cells that overexpress PC94 targeting peptide can be evaluated. SK-Hep- 1 cells can be used for initial evaluation since they express the target of the PC94 targeting peptide (Lo A, et al. Mol Cancer Ther. 2008 7:579-89), are CD44+ and are sensitive to the antiproliferative effects of the miR-199a-3p mimic. Normal human primary nasomucosal cells (NNM) (Lee TY, et al. Cancer Res. 2004 64:8002-8) that lack PC94 binding (Lo A, et al. Mol Cancer Ther. 2008 7:579-89) can be used as a negative control. The IC₅₀ can be determined by exposing SK-Hep- 1 cells to the pre-miR-199a-2 targeting MVs for 1 to 4 days at a 10-log concentration range and cell proliferation can be determined by WST assay. SK-Hep- 1 and NNM cells can then be exposed to IC₅₀ concentrations of the PC94 targeting peptide or nontargeting control exoxomes. Positive control includes lipofectamine transfection of 100 nM miR-199a-3p duplex oligo mimic (Ambion). Negative controls can include empty MVs and the lipid transfected scrambled control oligo mimic (Ambion). To assess uptake, the amount of mature miR-199a-3p can be measured in the cells using qRT-PCR. The amount of CD44 mRNA and protein can be determined in the cells by qPCR and Western blotting, respectively. To evaluate activity, cells can be co-transfected with the psiCHECK-2 Vector (Promega) containing the CD44 3' UTR downstream of the luciferase (Henry JC, et al. Biochem Biophys Res Commun. 2010 403: 120-5); greater reduction in luciferase equates to more active miR-199a-3p delivered to the cells. Reduced proliferation of the cells can be determined by a WST-1 cell assay as described (Henry JC, et al. Biochem Biophys Res Commun. 2010 403: 120-5). miR-199a-3p mimic reduces in vitro invasion in the CD44+ HCC cell lines (Henry JC, et al. Biochem Biophys Res Commun. 2010 403: 120-5). The in vitro invasion of SK-Hep-1 cells exposed to the various treatments mentioned above can be determined using a matrigel coated Boyden chambers (Henry JC, et al. Biochem Biophys Res Commun. 2010 403:120-5).

To study efficacy in vivo, an orthotopic liver mouse model of human HCC can be used (Park JK, et al. Cancer Res. 2011 71 :7608-16). SK-Hep-1 cells can be stably transfected with a GFP-luciferase expressing construct to generate SK-Hepl-luc cells. Again, SK-Hep-1 cells were chosen as they are CD44+, PC94+ and are sensitive to the anti-pro liferative effects of miR-199a-3p mimic in vitro. Orthotopic tumors can be established by the direct intrahepatic injection of SK-Hepl-luc cells (1,000,000 cells suspended in matrigel) into the left hepatic lobe of nude mice. Starting 10-14 days after orthotopic implantation, and every week thereafter if necessary, tumor burden can be determined by bioluminescence imaging using the IVIS200 imaging system (Xenogen Corp., Alameda, CA), 10 minutes after LP. administration of 150 mg/kg body weight D-luciferin (Gold Biotechnology, St. Louis, MO). Once bioluminescence exceeds 1 x 10^"6 photons/sec, mice can be randomized to receive various doses of the PC94 targeting MVs.

MVs can be purified from HEK293T cells, and only batches that fall within 3 standard deviations of the mean for average particle size, number of vesicles harvested, and copy number of miR-199a-3p can be selected. Whether batches are free from bacteria can be verified prior to their use in animal experiments, and MVs can be used within some time limit after harvesting (e.g., the time limit will be based on MV stability data). MVs can be frozen and thawed to determine optimal storage conditions. For in vivo experiments, freshly isolated (or frozen) MVs can be quantified and resuspended in normal sterile saline to achieve a 10X dosing solution based on miR-199a-3p copy number. Solutions can be kept on ice then diluted 1 : 10 in 37°C sterile saline immediately before dosing. Sterile conditions can be maintained by using syringes only once for all dosing experiments.

The maximally tolerated single dose (acute MTD) of MVs that does not result in overt toxicity within a 24 hr time period can be determined. Signs of overt toxicity include paralysis, lethargy, tremors, labored breathing, lack of food or water consumption, abnormal urine or feces (e.g. blood in urine or diarrhea), or other observable signs that the animal is in distress. If toxicity is not observed, the acute MTD can be determined with practical considerations for maximum achievable MV concentration in the dosing solution or maximum amount of MVs that can be produced for in vivo studies. The acute MTD study can be started by dosing non- diseased nude mice intravenously (IV) with the highest dose achievable. If no toxicity is observed, it can be confirmed whether this dose is tolerable in orthotopic diseased mice then proceed to pharmacokinetic and pharmacodynamic (PK/PD) studies. If toxicity is observed in one or more of the five animals within a 24 hr time period after dosing with the highest achievable dose, the dose can be reduced to 1/3 and evaluated in othe ranimals. This process can be repeated until an IV acute MTD is identified and confirmed in diseased mice (i.e. the highest dose level where no animals exhibit toxicity). This study can then be repeated with intraperitoneal (IP) dosing to determine an IP acute MTD.

This study can generate PK/PD data allowing 1) selection of appropriate sampling times for the subsequent PK/PD studies described below and 2) an early determination of systemic miR-199a-3p availability with IP dosing. IP dosing can allow for a higher frequency of dosing in efficacy studies. The primary PD parameter to be modeled is the level of tumor CD44 mRNA as determined by qPCR. 199a-3p oligo reduces CD44 mRNA in SK-Hep-1 and is thus more quantitative than measuring CD44 protein by Western blotting. Single IV and IP doses of MVs at their respective MTDs can be given to 62 mice with orthotopic tumors (e.g., 30 mice each route and 2 mice as controls without MVs). Tumor bearing orthotopic mice can be randomized once bioluminescence exceeds 1 x 10^"6 photons/sec. Two animals can be sacrificed pre-dose then at 15 different time points between 2 mins and 96 hrs after dosing for collection of blood (plasma) and liver tissue. Other tissues, including kidney, spleen, lung, brain, heart, abdominal fat and skeletal muscle can also be collected for potential later analysis. All tissues and plasma samples collected can be flash frozen in liquid nitrogen and stored at -80°C until analysis. Plasma

concentrations of miR-199a-3p can be quantified in each plasma and liver sample using qPCR. Endogenous miR-199a-3p is present in circulation (Hunter MP, et al. PLoS ONE. 2008 3:e3694) and normal liver (Lee EJ, et al. RNA. 2007 14:35-42), however it is not present in SK-Hep-1 cells. Since the sequence of the 21 nt mature miR-199a-3p is identical between mouse and human, qPCR does distinguish the therapeutic miR-199a-3p from the endogenous miR-199a-3p in the treated mice. To allow measurement of the miR-199a-3p delivered from exosomes, the sequence of the miR-199a-3p can be slightly modified. The 3' most nt of the mature miR-199a-3p can be changed from an A to a C. This change does affect the 5' seed region (to be verified by luciferase reporter assays and western blotting/qPCR of transfected oligos in cells) that is critical to activity, however it can affect the binding of the TaqMan probe to the cDNA (Chen C, et al. Nucleic Acids Res. 2005 33:el79). This slight difference can allow discrimination of the therapeutic miR-199a-3p from endogenous mouse miR-199a-3p. The assay can be validated using synthetic RNA oligos of the wild type and modified sequence. Total RNA can be extracted from the livers and the amount of endogenous miR-199a-3p and CD44 mRNA can be determined by qPCR Plasma and liver miR-199a-3p concentration versus time and tumor CD44 gene expression versus time can be used to identify time points and to crudely model PK/PD.

10 to 15 time points can be selected based on observed data and a larger PK/PD study designed with approximately n=4 diseased orthotopic mice per time point (replication can depend on pilot study results) to more thoroughly characterize miR-199a-3p disposition at the IV and IP acute MTD doses. It can be important to capture the full profiles for both PK and PD, which can include approximately 5 half- lives for PK and recovery near 10% of baseline for PD. This study can provide livers at multiple time points, measure miR-199a-3p quantity in target tumor tissue, compare miR-199a-3p in tumor versus normal liver tissue, determine if exposure

(area under the concentration-time curves, AUC) compares to that required for activity in in vitro experiments, and determine how CD44 is modulated as a primary PD endpoint. miR-199a-3p concentrations can be measured in plasma and liver, and several additional tissues can be collected from each mouse and stored for potential later analysis (kidney, spleen, lung, heart, brain, abdominal fat and skeletal muscle). CD44 gene and protein expression, as well as endogenous miRNA and target protein expression, can be evaluated in liver. PK and PD models can be developed, and parameter values can be determined as discussed above. The linked PK/PD model can be used to simulate PK/PD concentration vs. time and effect vs. time curves with a variety of dose levels and schedules of administration and to choose dose levels for tumor efficacy studies as described below. Livers from pre-dose mice and mice sacrificed at the latest time points in each of the dose/route groups can be examined for signs of overt toxicity (defined above) as well as apoptosis. Signs of liver failure can also be evaluated by examining serum levels of bilirubin, ALT and AST. Off target effects can be examined using the following strategy. Total RNA can be collected from the tumors and subjected to whole genome cDNA arrays analysis. Data can be analyzed using Sylamer plots (van Dongen S, et al. Nat Methods. 2008 5: 1023-5) to ensure only miR-199a-3p target genes (Target Scan) are affected by the miR-199a-3p treatment. The cellular and subcellular localization of miR-199a-3p in the FFPE liver can be studied using in situ hybridization.

The study can estimate residual variance and use that as a gauge to test whether the change in miR-199a-3p over time differs significantly from zero. From this study, the sample size for the full MTD PK PD study can be estimated to obtain sufficiently accurate estimates of PK/PD parameters. 4 mice per time point can have sufficient accuracy for all standard PK parameters so that their standard errors are less than 20% of their corresponding estimates. If the accuracy criterion is not met, a second stage of the experiment can be run based on standard errors from the first. A variety of compartmental PK models can be evaluated in order to identify the simplest model that describes the plasma and liver miR-199a-3p concentration-time data. Direct and indirect response PD models can be considered for describing CD44 modulation. Goodness-of-fit for each model can be assessed using model

convergence, visual inspection of residual plots, Akaike information criterion and

Schwarz criterion. Standard errors of estimate for each parameter can be used to gauge accuracy of each parameter value estimate. PK and PD variance estimates can guide replication requirements (numbers of animals required) in the definitive studies. The resulting PK and PD models can be linked to be able to simulate CD44 modulation with a variety of alternative IV and IP dose regimens.

Chronic toxicity can be evaluated in healthy mice prior to conducting the following proof of concept efficacy study in diseased mice. The chronic dosing scheme can use an MTD approach. In other words, mice can be dosed at a maximum tolerable dose level and frequency to ensure the best chance for efficacy without toxicity. The PK/PD model can be used to simulate the approximate effect vs. time profile (i.e. level of CD44 expression vs. time) when MVs are dosed at their acute IP or IV MTD at varying frequencies (i.e. twice daily, once daily, three times weekly, once weekly, etc.). If no toxicity is observed at the MVs' highest feasible dose, then a high frequency of dosing (e.g. twice daily) can be evaluated. On the other hand, if toxicity is observed at doses higher than the determine acute MTD, CD44 expression vs. time can be simulated with lower frequencies at the IV or IP acute MTD and with higher frequencies at dose levels lower than the acute IV or IP MTD. Therefore, the dose regimens ultimately evaluated in healthy mice may be as high as the acute IV or IP MTD and as frequent as twice daily. Based on this approximation for dose level and frequency, an approximate IV, IP or mixed (IV and IP) dosing regimen can be developed and evaluated in a chronic dosing study to ensure this regimen is non-toxic in healthy mice. High systemic availability of liposome encapsulated drugs can be achieved with IP administration. IP administration of MVs can result in high bioavailability. However, if miR-199a-3p plasma and liver concentrations and area under the concentration-time curves (AUCs) display poor systemic bioavailability with IP dosing relative to IV dosing, or if C_max of MiR-199a-3p at the IP acute MTD is anticipated to be below the IC₅₀ in liver, even after repeat dosing and potential accumulation are taken into consideration, then IP dosing can be abandoned as a viable option. Healthy mice can be dosed with miR-199a-3p or scramble miR loaded into targeting exosomes at the MTD dose for 8 weeks with the frequency determined above. Use of the scrambled miR can help to separate if any observed toxicity is due to the vesicles or due to the miR-199a-3p cargo. Eight weeks duration is used as this is the time that untreated control mice begin to die off. Mice in groups of n=5 can be dosed according to the schedule and evaluated for toxicity. If toxicity is observed, the level and/or frequency can be decreased and evaluated in another n=5 mice. When a tolerable chronic schedule is determined, this dose regimen can be defined as the chronic MTD and proceed to the efficacy study.

The chronic MTD regimen can be evaluated for efficacy in a 8-week dosing study. Mice with orthotopic tumors can be enrolled in this study once

bioluminescence exceeds 1 x 10^~6 photons/sec. Mice can be randomized into four groups (n=13 each group) for this study. The experimental group can be treated with miR-199a-3p containing MVs, and the three control groups can include, saline, targeting pre-miR control, nontargeting pre-miR-199a-3p). The experimental group can be dosed with the chronic MTD regimen, and the control groups can be dosed at their respective MTDs. The primary criteria for demonstration of in vivo efficacy and proof of concept can be a significant survival advantage compared to the non- targeting control. Assuming that all the scramble control mice die at day 70 (Parks, 2011), n=13 mice per group can give 84% power to detect 30% death reduction in the pri-miR-199a-2 mouse group (a=0.025). In a confirmatory analysis, the groups of targeting and non-targeting pre-miR-199a-3p can be compared, log-rank test can be used to compare the survival functions between groups.

Unless defined otherwise, all technical and scientific terms used herein have the same meanings as commonly understood by one of skill in the art to which the disclosed invention belongs. Publications cited herein and the materials for which they are cited are specifically incorporated by reference.

Those skilled in the art will recognize, or be able to ascertain using no more than routine experimentation, many equivalents to the specific embodiments of the invention described herein. Such equivalents are intended to be encompassed by the following claims.

Claims

WHAT IS CLAIMED IS:

1. A system, comprising:

(a) a first nucleic acid sequence encoding an exosome targeted fusion protein, wherein the exosome targeted fusion protein comprises a nucleic acid binding moiety and an exosomal transmembrane moiety; and

(b) a second nucleic acid sequence encoding a chimeric polynucleotide, wherein the chimeric polynucleotide comprises a therapeutic nucleic acid sequence conjugated to a substrate for the nucleic acid binding moiety.

2. The system of claim 1, wherein the therapeutic nucleic acid sequence is selected from the group consisting of an siRNA, dsRNA, dsDNA, shRNA, mRNA, microRNA, antimicroRNA, antagomir, antimir, antisense, aptamer, and dsRNA/DNA hybrids.

3. The system of claim 1 or 2, wherein the nucleic acid binding moiety comprises HIV Tat protein, or fragment or variant thereof capable of binding an HIV trans- activation response (TAR) element, and wherein the substrate for the nucleic acid binding moiety comprises an HIV TAR element, or a fragment or variant thereof capable of binding an HIV Tat protein.

4. The system of any one of claims 1 to 3, wherein the exosomal transmembrane moiety is selected from the group consisting of Lamp- 1, Lamp-2, CD 13, CD86, Flotillin, Syntaxin-3, CD2, CD36, CD40, CD40L, CD41a, CD44, CD45, ICAM-1, Integrin alpha4, LiCAM, LFA-1, Mac-1 alpha and beta, Vti-IA and B, CD3 epsilon and zeta, CD9, CD18, CD37, CD53, CD63, CD81, CD82, CXCR4, FcR, GluR2/3, HLA-DM (MHC II), immunoglobulins, MHC-I or MHC-II components, TCR beta, and tetraspanins.

5. The system of any one of claims 1 to 4, wherein the exosome targeted fusion protein further comprises a tissue targeting moiety.

6. The system of claim 5, wherein the tissue targeting moiety comprises a PC94 peptide, a rabies virus glycoprotein (RVG), an RGD peptide, LHRH peptide, or galectin-3 -binding peptide.

7. The system of claim 5, wherein the tissue targeting moiety comprises a single chain antibody (scFv) that binds a tissue specific antigen.

8. The system of any one of claims 1 to 7, wherein the exosome targeted fusion protein further comprises a protein tag.

9. The system of claim 8, wherein the protein tag comprises a FLAG-tag.

10. The system of claim 8, wherein the protein tag comprises a streptavidin binding peptide.

11. The system of any one of claims 1 to 9, wherein the exosome targeted fusion protein comprises the following formula:

P - TT - TM - NB - P,

wherein "P" consists of an optional protein tag,

wherein "TT" consists of an optional tissue targeting moiety ,

wherein "TM" consists of a exosomal transmembrane moiety,

wherein "NB" consists of a nucleic acid binding moiety, and

wherein "-" consists of a peptide linker or a peptide bond.

12. The system of any one of claims 1 to 11 , wherein the first nucleic acid sequence is operably inserted in an expression vector.

13. The system of any one of claims 1 to 12, wherein the second nucleic acid sequence is operably inserted in an expression vector.

14. The system of any one of claims 1 to 13, wherein the first nucleic acid sequence and second nucleic acid sequence are operably inserted in a common expression vector.

15. An exosome targeted fusion protein, comprising a nucleic acid binding moiety and an exosomal transmembrane moiety.

16. The exosome targeted fusion protein of claim 15, wherein the nucleic acid binding moiety comprises HIV Tat protein, or fragment or variant thereof capable of binding an HIV trans-activation response (TAR) element, and wherein the substrate for the nucleic acid binding moiety comprises an HIV TAR element, or a fragment or variant thereof capable of binding an HIV Tat protein.

17. The exosome targeted fusion protein of claim 15 or 16, wherein the exosomal transmembrane moiety is selected from the group consisting of Lamp- 1, Lamp-2, CD13, CD86, Flotillin, Syntaxin-3, CD2, CD36, CD40, CD40L, CD41a, CD44, CD45, ICAM-1, Integrin alpha4, LiCAM, LFA-1, Mac-1 alpha and beta, Vti-IA and B, CD3 epsilon and zeta, CD9, CD18, CD37, CD53, CD63, CD81, CD82, CXCR4, FcR, GluR2/3, HLA-DM (MHC II), immunoglobulins, MHC-I or MHC-II components, TCR beta, and tetraspanins.

18. The exosome targeted fusion protein of any one of claims 15 to 17, wherein the exosome targeted fusion protein further comprises a tissue targeting moeity.

19. The exosome targeted fusion protein of claim 18, wherein the tissue targeting moiety comprises a PC94 peptide or a rabies virus glycoprotein (RVG).

20. The exosome targeted fusion protein of claim 18, wherein the tissue targeting moiety comprises a single chain antibody (scFv) that binds a tissue specific antigen.

21. The exosome targeted fusion protein of any one of claims 15 to 20, wherein the exosome targeted fusion protein further comprises a protein tag.

22. The exosome targeted fusion protein of claim 21 , wherein the protein tag comprises a FLAG-tag.

23. The exosome targeted fusion protein of claim 21 , wherein the protein tag comprises a streptavidin binding peptide.

24. The exosome targeted fusion protein of any one of claims 15 to 22, wherein the fusion protein comprises the following formula:

P - TT - TM - NB - P,

wherein "P" consists of an optional protein tag,

wherein "TT" consists of an optional tissue targeting moiety ,

wherein "TM" consists of a exosomal transmembrane moiety,

wherein "NB" consists of a nucleic acid binding moiety, and

wherein "-" consists of a peptide linker or a peptide bond.

25. A composition comprising an exosome, wherein the exosome comprises the exosome targeted fusion protein of any one of claims 15 to 24.

26. The composition of claim 25, wherein the exosome is loaded with a chimeric polynucleotide comprising a therapeutic nucleic acid sequence conjugated to a substrate for the nucleic acid binding moiety.

27. The composition of claim 26, wherein therapeutic nucleic acid sequence is selected from the group consisting of an siRNA, dsRNA, dsDNA, shRNA, mRNA, microRNA, antimicroRNA, antagomir, antimir, antisense, aptamer, and dsRNA/DNA hybrids.

28. The composition of claim 26 or 27, wherein the nucleic acid binding moiety comprises HIV Tat protein, or fragment or variant thereof capable of binding an HIV trans-activation response (TAR) element, and wherein the substrate for the nucleic acid binding moiety comprises an HIV TAR element, or a fragment or variant thereof capable of binding an HIV Tat protein.

29. The composition of any one of claims 25 to 28, wherein the exosome is derived from a dendritic cell.

30. A recombinant cell, comprising an exosome producing cell engineered to contain the first nucleic acid sequence and the second nucleic acid sequence of the system of any one of claims 1 to 14.

31. The recombinant cell of claim 30, wherein the exosome producing cell is a dendritic cell.

32. The recombinant cell of claim 31 , wherein the dendritic cell is an immature dendritic cell.

33. The recombinant cell of claim 30, wherein the exosome producing cell is selected from the group consisting of Human Embryonic Kidney 293 (HEK) cells, 293T cells, Chinese hamster ovary (CHO) cells, and human ESC-derived

mesenchymal stem cells.

34. A chimeric polynucleotide, comprising a therapeutic nucleic acid sequence conjugated to a heterologous HIV trans-activation response (TAR) element.

35. The chimeric polynucleotide of claim 34, wherein therapeutic nucleic acid sequence is selected from the group consisting of an siRNA, dsRNA, dsDNA, shRNA, mRNA, microRNA, antimicroRNA, antagomir, antimir, antisense, aptamer, and dsRNA/DNA hybrids.

36. A method for making an exosome loaded with a therapeutic polynucleotide, comprising:

(a) culturing the recombinant cell of any one of claims 30 to 33 under conditions suitable to produce exosomes; and

(b) purifying exosomes from the cell that comprise the targeted fusion protein.