WO2019086542A1

WO2019086542A1 - Metal organic framework-based compositions and uses thereof

Info

Publication number: WO2019086542A1
Application number: PCT/EP2018/079881
Authority: WO
Inventors: David FAIREN-JIMENEZ; Gabriele KAMINSKI SCHIERLE; Michelle TEPLENSKY; Colin HOCKINGS; Salame HADDAD
Original assignee: Cambridge Enterprise Limited
Priority date: 2017-10-31
Filing date: 2018-10-31
Publication date: 2019-05-09
Also published as: GB201717975D0

Abstract

The present invention relates to compositions based on metal organic framework (MOF) materials for delivering RNAi molecules, processes for the manufacture of such materials and to the uses of such materials. The invention has particular, but not exclusive, applicability to therapeutic uses. In particular, use in the treatment of cancer is encompassed.

Description

_

METAL ORGANIC FRAMEWORK-BASED COMPOSITIONS AND USES THEREOF

BACKGROUND TO THE INVENTION Field of the invention

Related art Metal-organic frameworks (MOFs) are porous crystalline materials prepared by the self- assembly of metal ions and organic ligands. MOFs can have large pore volumes and apparent surface areas as high as 8,000 m²/g. MOFs combine a structural and chemical diversity that make them attractive for many potential applications, including gas storage, gas separation and purification, sensing, catalysis and drug delivery. The most striking advantage of MOFs over more traditional porous materials is the possibility to tune the host / guest interaction by choosing the appropriate building blocks, i.e. the metal ions and organic ligands, from which the MOF is formed.

Metal-organic frameworks are prepared by self-assembly processes from metal ions or metal clusters, which act as coordination centres for interconnection with organic ligands. The resulting products have a well-defined crystalline structure.

Metal-organic frameworks are known to encapsulate certain active agents by efficiently adsorbing them into their pore structures. For example, Horcajada et al. have loaded different anticancer and antiviral agents into metal-organic frameworks, and He et al. have reported the use of metal-organic frameworks for the co-delivery of the anticancer cisplatin and small interference RNA, to enhance therapeutic effect. According to He et al., the siRNA is apparently coordinated to metal sites on the exterior surface. The interior porosity of the metal-organic framework was apparently not used for loading in this situation.

Many eukaryotic cells use RNAi to regulate gene expression. While mRNA molecules transfer genetic information from nuclear DNA to the ribosome for translation into peptide and protein molecules, RNAi molecules interfere with that process by inducing degradation of the targeted mRNA. The neutralisation of the targeted mRNA molecules specifically reduces expression of the gene that was transcribed to produce the targeted mRNA molecule and may be referred to as 'gene silencing' and/or 'gene knockdown'. Delivery of exogenous RNAi molecules to cells has been used to reduce the expression of a given target gene in the therapeutic context of RNAi mediated gene therapy, including tumour gene therapy (CN104888234A). Two types of RNAi molecules that have been used therapeutically are short interfering RNA (siRNA) and microRNA

(miRNA).

Typically, siRNA molecules are around 22-23 base pairs in length. siRNA molecules are double stranded molecules formed of two complementary RNA molecules hybridized to each other as a duplex, which have been cleaved from endogenously expressed double stranded RNAs (dsRNAs) by the enzyme Dicer. The diameter of siRNA molecules is usually taken to be around 2.0 nm, while the length is around 7 nm. miRNAs exit the nucleus as a single molecule, termed "pre-miRNA", having around 70 bases and containing a hairpin structure and a duplex region formed by hybridization of one end of the pre-miRNA molecule with the other end. The pre-miRNA is processed in the cytosol to form the mature miRNA, which consists of about 22 nucleotides.

Both siRNA and miRNA exert their gene silencing action by forming an RNA-induced silencing complex (RISC) involving proteins such as Argonaute, which cleaves the target mRNA. While RNAi therapy has the potential to treat many diseases including cancer, there are current limitations with the technology owing to its lack of stability, ease of degradation by biologically native enzymes, and large size compared to other therapeutic compounds. To overcome these limitations, different groups and organizations have modified its chemical structure, or loaded it within polymers or nanoparticles. There are several drawbacks to each of these solutions, mainly that chemical modifications tend to lower the efficacy of the siRNA upon cytosol delivery, and that loadings within these other structures tend to be difficult and low because of its size. Synthetic RNAi molecules have the potential to inhibit various disease-associated genes, and may be designed to bypass the early stages of endogenous RNAi molecule production thereby allowing for the creation of a platform technology with any genetic sequence. Despite its great therapeutic potential, there are current limitations with RNAi technology owing to a lack of stability of the RNA molecules, their ease of degradation by biologically native enzymes and large size compared to many synthetic compounds.

SUMMARY OF THE INVENTION

The present invention has been devised in order to address at least one of the above problems. Preferably, the present invention reduces, ameliorates, avoids or overcomes at least one of the above problems. In particular, the present invention is based on the surprising observation that a MOF can be used in conjunction with an endosomal release factor in order to deliver RNAi molecules effectively to cells.

Accordingly, in a first preferred aspect, the present invention provides a composition comprising (i) a metal organic framework (MOF) particle; (ii) an RNAi molecule; and (iii) an endosome release factor. The MOF material is mesoporous, such that the diameter of pores opening at the surface of the MOF particles and extending into the MOF particles is at least 2 nm and not more than 10 nm and wherein the length of the pores formed within the MOF crystal structure is optionally greater than the diameter of the pores.

In a second preferred aspect, the present invention provides a method of making the composition of the invention, the method comprising providing a MOF and contacting it with (i) an RNAi molecule and (ii) an endosome release factor.

In a third preferred aspect, the present invention provides the composition of the invention for use in methods of treating a patient in need thereof; and provides methods for treating a patient in need thereof by administering the composition to said patient. In a fourth preferred aspect, the present invention provides the use of a MOF particle in the manufacture of a medicament, wherein said manufacture comprises the step of forming a composition by contacting the MOF with (i) an RNAi molecule and (ii) an endosome release factor. In line with the abovementioned aspects of the invention, the MOF material is mesoporous, such that the diameter of pores opening at the surface of the MOF particles and extending into the MOF particles is at least 2 nm and not more than 10 nm and wherein the length of the pores formed within the MOF crystal structure is optionally greater than the diameter of the pores.

In a fifth preferred aspect, the present invention provides a method of delivering an RNAi molecule to a cell, the method comprising treating the cell with the composition of the invention. Preferably, the method of this aspect is performed in vitro.

The first, second, third, fourth and/or fifth aspect of the invention may have any one or, to the extent that they are compatible, any combination of the following optional features.

In some embodiments, the MOF particles have a longest linear dimension that falls within the range 100 nm to 600 nm, or the range 100nm to 300nm. In other embodiments, the MOF particles have a longest linear dimension that falls within the range 200 nm to 600 nm, or the range 300 nm to 500 nm, assessed as a number average. Preferably, the MOF particles have a longest linear dimension of at most 400 nm, assessed as a number average. More preferably, the MOF particles have a longest linear dimension of at most 300 nm or at most 200 nm, assessed as a number average. Particle sizes may be determined from the SEM images of a sample of a metal-organic framework. Exemplary SEM methods are described herein. Preferably, the length of the pores formed within the MOF crystal structure is at least 10 nm.

The RNAi molecule may be present within the pores and/or associated with the outer surface of the MOF particle.

In some embodiments, the diameter of pores opening at the surface of the MOF particles and extending into the MOF particles is not more than 7 nm. In other embodiments, the diameter of pores opening at the surface of the MOF particles and extending into the MOF particles is not more than 5 nm. The diameter of the pores opening at the surface and extending into the MOF particle may fall within the range 2 nm to 9 nm, 2 nm to 8 nm,

2 nm to 7 nm, 2 nm to 6 nm, 2 nm to 4 nm, or 2 nm to 3 nm. In some embodiments, the diameter of these pores falls within the range 3 nm to 9 nm, 3 nm to 8 nm, 3 nm to 7 nm,

3 nm to 6 nm, 3 nm to 5 nm, or 3 nm to 4 nm. In some embodiments, the diameter of these pores falls within the range 4 nm to 9 nm, 4 nm to 8 nm, 4 nm to 7 nm, 4 nm to 6 nm, or 4 nm to 5 nm. In some embodiments, the diameter of these pores falls within the range 5 nm to 9 nm, 5 nm to 8 nm, 5 nm to 7 nm or 5 nm to 6 nm. In some embodiments, the diameter of these pores is about 2 nm or about 3 nm.

The abovementioned dimensions can be measured using any suitable technique known in the art, for instance X ray diffraction, electron microscopy, laser diffraction, etc. One suitable method for determining the longest linear dimension of a MOF is to use scanning electron microscopy to determine the length of several individual MOF particles and then determine the average. Without being bound by theory, the inventors believe that the MOF particles used in the composition of the invention are effective for delivering RNAi molecules because the RNAi molecules are able to become located within the pores. In some instances, the RNAi molecules are partially located within the pores. In some instances, the RNAi molecules are entirely located within the pores. The pores are too small for RNA- degrading enzymes to enter so delivery of siRNA using a MOF can protect the siRNA from degradation.

In some embodiments of the invention, the MOF unit cell comprises at least one metal ion selected from the group consisting of Zr ions, Hf ions, Bi ions, Fe ion and Mg ions. In one embodiment, the metal-organic framework is a zirconium-containing metal-organic framework, for example a metal-organic framework having zirconium oxo-clusters. The Additionally or alternatively, the metal-organic framework may be a bismuth-containing metal-organic framework. In some embodiments, the MOF has the structure Zr6^3-OH)8(OH)₈(linker)2. In some embodiments, the MOF has the structure Zr6^3-OH)8(OH)₈(TBAPy)2, wherein TBAPy is 1 ,3,6,8-tetrakis(p-benzoic acid)pyrene. In some embodiments, the MOF is NU-1000. In some embodiments, the MOF is NU-1003, which has a very similar structure to NU-1000 (see Peng et at), except that the pyrene- based linker has longer strands. In some embodiments, the MOF is PCN-128. PCN-128 is also similar to NU-1000 and NU-1003 except that the linker is of intermediate size. In some embodiments, the MOF is PCN-777, which has the formula

Zr₆(0)4(OH) i o(H₂0)6(TATB)2, wherein TATB is 4,4',4"-s-triazine-2,4,6-triyl-tribenzoate. The pores of PCN-777 are not straight, instead 'snaking' into the MOF particle. MOF particles of a given chemical structure may be provided in different particle size. For instance, NU-1000 may be provided with a longest linear dimension of about 150 nm (which may be denoted nNU-1000).

The precise composition of the MOF is not critical. The crystalline, mesoporous structure enables the MOF to form an effective component of the composition of the invention.

The composition of the invention includes an RNAi molecule. The RNAi molecule may be an siRNA or a microRNA. The present invention allows 'mature' RNAi molecules to be delivered into cells such that the RNAi molecules can interact with the RISC complex directly after being released into the cytosol, without the need for processing of the RNAi molecule first. Preferred RNAi molecules are capable of downregulating the expression of a gene implicated in tumour cell survival, for instance a multidrug resistance gene.

The composition of the invention includes an endosome release factor. One or more endosome release factors may be included in the composition. The endosome release factor may be a peptide, for instance a cell penetrating peptide (CPP), a membrane opening peptide (MOP), and/or an amphipathic peptide. One such peptide is the 'KALA' peptide, which has the amino acid sequence YEAKLAKALAKALAKHLAKALAKALKACEA and which is commercially available. Other endosome release peptides include HA2E5- TAT, HA2-penetratin, HA-K₄, HA2E4, GALA, INF-7, PMAP (considered to be membrane opening peptides) and HIV TAT protein, penetratin, SV40 (cell penetrating peptides). Besides these, many other cell penetrating and membrane opening peptides are known (Miletti, 2012), and these are able to facilitate cellular intake/uptake of various materials.

Alternatively, a non-peptide endosomal release factor(s) may be used in the

compositions of the invention. For instance, the endosome release factor may be a chemical such as ammonium chloride (NH₄CI). Non-peptide endosomal release factors may exert their action by modulating endosomal acidity. Exemplary chemicals that can be used as endosome release factors include the so-called proton-sponge® (chemical name 1 ,8-Bis(dimethylamino) naphthalene, N,N,N',N'-Tetramethyl-1 ,8- naphthalenediamine), which has the molecular formula CioH₆[N(CH₃)2]2. More broadly, as described herein, other chemical agents (such as cationic polymers or lipids with excess uncharged protonable amine groups) are able to act as endosome release factors via the proton sponge mechanism, including Poly L-histidine, PEI, PPAA and ammonium chloride (NH4CI).

In aspects of this invention involving delivering the composition of the invention to a cell, the cell is preferably a mammalian cell. For instance, the cell may be a tumour cell. In some embodiments, the invention provides a method for delivering an RNAi molecule, such as an siRNA into a human tumour cell. The method may be performed in vitro. One or more of the component parts of the composition may be conjugated to a probe, e.g. a fluorescent dye. Alternatively, a probe, e.g. a fluorescent dye, may be included as an additional component of the composition. The use of a probe (e.g. a dye) enables intracellular imaging and/or tracking of the composition.

In aspects of this invention relating to methods of treating patients using the composition of the invention, the composition may be administered to the patient via any suitable route of administration. For instance, the composition may be administered orally, intravenously or directly at the tumour site e.g. by intratumoural injection. For certain cancers, e.g. skin cancer, topical administration may be used. For certain cancers, e.g. lung cancer, inhalation may be used as the route of administration.

The composition of the invention may be used in the treatment of cancer. The cancer may be breast cancer, mesothelioma, pancreatic cancer, cervical cancer, skin cancer or lung cancer. Preferably, the RNAi molecule that is delivered as part of the composition of the invention comprises a sequence that corresponds with a sequence of a gene that is expressed by the tumour. The skilled person is able to determine the degree of correspondence needed. For instance it has been reported that only 6 nucleotides of correspondence can be sufficient for an RNAi molecule to exert a gene silencing effect. Preferably, the RNAi molecule that is delivered as part of the composition of the invention comprises a sequence that corresponds with: at least 10 nucleotides of the target sequence, at least 15 nucleotides of the target sequence, at least 16 nucleotides of the target sequence, at least 17 nucleotides of the target sequence, at least 18 nucleotides of the target sequence, at least 19 nucleotides of the target sequence, or at least 20 nucleotides of the target sequence.

The composition of the invention may be formulated as a pharmaceutical composition by way of addition of one or more pharmaceutically acceptable excipient(s) and/or additives as appropriate for the chosen route of administration. The excipient can be selected from the group of stabilizers, wetting agents, binders, disintegrants, diluents and solubilisers.

Further optional features of the invention are set out below.

BRIEF DESCRIPTION OF THE DRAWINGS

Embodiments of the invention will now be described by way of example with reference to the accompanying drawings in which:

Fig. 1 shows the varying degree of gene knockdown using different siRNA molecules. HEK 293 cells incorporating an inducible mCherry gene are able to express RFP when induced with doxycycline (Fig. 1 A, upper panel; top-left quadrant) and treatment with siRNA having a scrambled sequence does not visibly reduce RFP expression (Fig. 1 A, upper panel; lower-left quadrant). No RFP expression is visible without doxycycline (Fig. 1 A, upper panel; right-hand quadrants). Different siRNA molecules, each with sequences that correspond with different regions of the RFP gene, are able to knockdown RFP expression to differing degrees (Fig. 1 A, lower panel). Knockdown of RFP expression using different siRNA molecules is also shown by western blot (Fig. 1 B).

Fig. 2 shows the effects of siRNA loading on MOF characterization by powder X-ray diffraction (PXRD). Loss of peaks, decreasing intensity, and slight broadening of peaks is observed. Decrease in contrast can be attributed to decreasing relative peak intensity, which suggests siRNA within the framework.

Fig. 3 is an image of a gel electrophoresis analysis of siRNA molecules following treatment with RNA degrading RNase enzyme. The lane at the left-hand side of the figure contains the untreated control. When reacted with enzyme, the siRNA in its unprotected state is degraded, and no band is visible at the relevant 21 nt length. The right-hand lanes contains the MOF/siRNA composition with one that has undergone enzyme treatment, and one that has not. In both instances, the visible band at 21 nt in length (indicated by the arrow) shows that the siRNA is still present post treatment. Thus, the siRNA in the MOF/siRNA composition is protected from enzymatic degradation. Fig. 4 shows mCherry (RFP) expression in HEK 293 cells. First column (left hand side; "HEK 293-mC (-Tet)") shows that no mCherry expression is detected in cells that have not been induced with tetracycline. Expression in induced cells that were not treated with other agents is shown in the second column ("Induced HEK 293-mC (+Tet)"). All values were normalized to this level. Subsequent columns (left to right) show induced cells treated with: naked siRNA; siRNA with lipofectamine; MOF alone; siRNA with MOF; siRNA with MOF and Proton-Sponge®; Proton-Sponge® alone; siRNA with MOF and KALA peptide; siRNA with MOF and both Proton-Sponge® and KALA peptide; and siRNA with MOF and NhUCI.

Fig. 5 shows intracellular fluorescence at 647 nm following treatment of HEK 293 cells treated with siRNA tagged with Alexafluor-647, delivered as: naked tagged siRNA;

tagged siRNA with lipofectamine; MOF alone; and tagged siRNA with MOF.

Fig. 6 shows fluorescence-lifetime imaging microscopy (FLIM) analysis of three different NU-1000 conditions - MOF only, siRNA loaded NU-1000, and enzyme-reacted siRNA loaded NU-1000.

Fig. 7 shows fluorescence of 'naked' siRNA tagged with Alexafluor-647 (normalised to 100), compared with; PCN-777 Zr at a size of 300 nm; PCN-777 Zr at a size of 300 nm treated with Shortcut RNase III enzyme; tagged siRNA with PCN-777 Zr at a size of 300 nm; tagged siRNA with PCN-777 Zr at a size of 300 nm treated with Shortcut RNase III enzyme; PCN-777 Hf at a size of 130 nm; PCN-777 Hf at a size of 130 nm treated with Shortcut RNase III enzyme; tagged siRNA with PCN-777 Hf at a size of 130 nm; tagged siRNA with PCN-777 Hf at a size of 130 nm treated with Shortcut RNase III enzyme. Fig. 8 shows FLIM analysis of different dextran-conjugated Oregon Green 488 (dextran- OG488) conditions - dextran-OG488 only (Control), KALA-added dextran-OG488 (KALA), NH4CI-added dextran-OG488 (NH4CI), and Proton Sponge-added dextran- OG488 (PS). Scatterplots of the lifetimes are shown with means and standard error. Non-equivalent variances between some of the conditions meant that statistical analysis was carried out using unpaired t-tests. (^*** P<0.001 , ^**** P<0.0001 )

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS, AND FURTHER OPTIONAL FEATURES OF THE INVENTION The large porosity of MOFs has attracted interest and research into the use of MOF particles for delivering therapeutics molecules into cells. The present invention provides compositions that use the MOF in conjunction with an endosomal release factor to allow delivery of RNAi molecules into cells. Metal-Organic Framework

Metal-organic frameworks (MOFs) are a class of porous self-assembling materials composed of inorganic metal clusters connected by organic linkers, which have recently emerged in porous-materials science for multiple unrelated applications including catalysis, gas storage and separation, ion exchange, use in sensors and use in medicine. A crystalline metal-organic framework is a network solid in which inorganic centres (or nodes) are connected by organic ligands. The inorganic centres may be metal ions or metal clusters. Within the crystalline metal-organic framework the inorganic centres are provided in a highly regular and extended arrangement. The crystalline framework is said to possess long range order.

Where a crystalline metal-organic framework holds a component, this component may be provided within the pores of the framework. Additionally, the component may also be provided on the surface of the framework, meaning the external surface of the crystalline metal-organic framework. In some embodiments, the majority, such as substantially all, of the component is provided within the pores of the framework. In some embodiments, the component that is held within the pores of the framework is the RNAi component. In some embodiments, the component that is held within the pores of the framework may be the endosome release factor.

The MOF that is used in the present invention is mesoporous. This means the MOF has pores of at least 2 nm in diameter. The size and geometry of the pores of the MOF used in the present invention is described elsewhere herein. The long range order of the MOF is observable in X-ray diffraction, specifically the Bragg reflections that are observable in the crystalline framework. Thus, the X-ray diffraction pattern for a crystalline metal-organic framework contains discernible Bragg peaks.

Analysis of a crystalline metal-organic framework by XRD can show whether a

component is present within the pores of the framework.

A metal-organic framework for use in the present invention may contain a metal that is selected from the group consisting of zirconium, hafnium, zinc, cobalt, nickel, palladium, platinum, copper, indium, iron, bismuth, magnesium and potassium, and mixtures thereof. A metal-organic framework for use in the present invention may contain a metal that is selected from the group consisting of zirconium, hafnium, zinc, iron, magnesium, cobalt, nickel, palladium, platinum, copper and indium, and mixtures thereof. Additionally or alternatively the metal-organic framework may contain bismuth as a metal. Typically a metal-organic framework contains only a single type of metal. A metal may be present in the framework as a metal cluster, such as a metal oxo-cluster.

Examples of zinc-containing metal-organic frameworks includes ZIF-1 , ZIF-3, ZIF-4, MOF-5 and MOF-177. An example of a cobalt-containing metal-organic framework is Co-ZIF-4. Examples of nickel-, palladium-, platinum-, and copper-containing metal- organic frameworks include nickel(ll) bisimidazolate, palladium(ll) bisimidazolate, platinum(ll) bisimidazolate, and copper(ll) bisimidazolate respectively. An example of a magnesium-containing metal-organic framework is CPO-27. An example of a potassium- containing metal-organic framework is CDMOF-1 .

An example of a bismuth-containing metal-organic framework is CAU-7. In the crystal structure of CAU-7, Bi³⁺ ions are nine-fold coordinated by oxygen atoms provided by BTB^3- ions (1 ,3,5-benzenetrisbenzoic acid ions), thereby forming threefold capped trigonal prisms. Face sharing of the Bi0₉ polyhedral leads to chains along the c-axis.

In one embodiment, the metal in the metal-organic framework has an oral lethal dose of at least 0.5 g/kg, at least 1 .0 g/kg, at least 2.0 g/kg, or at least 10 g/kg in a rat model, for example as measured in a salt form.

In one embodiment, the metal-organic framework is a zirconium-based organic framework. Zirconium-based systems are particularly attractive as zirconium low toxicity. For example, the oral lethal dose for zirconyl acetate, determined as LD₅o, is around 4.1 g/kg in a rat model. Furthermore, the human body contains around 300 mg of zirconium, and the amount of zirconium ingested daily is typically around 3.5 mg/day. Indeed there is a daily requirement for Zr at 0.05 mg per day. In one embodiment, a crystalline metal-organic framework has an IC₅o of at least 0.1 , at least 0.5, at least 1 .0, or at least 1 .5 mg/mL as measured against a test cell line, such as HeLa cells. Methods for measuring IC₅o values are described in the present case, and are well known in the art. Metal-organic frameworks are known to degrade under physiological conditions

("bioerosion"), thereby allowing release of the other components of the composition. Preferably, the MOF is amenable to decomposition in the presence of phosphate ions. As such, metal-organic frameworks are especially useful for releasing components in vivo, for example in a method of treatment, and in particular, release of the components in the endosome of a cell. The cell is preferably a tumour cell.

In one embodiment, the composition allows continuous release of the RNAi component into the cytoplasm over a period of at least 1 hour, at least 2 hours, at least 4 hours, at least 8 hours, at least 12 hours, at least 1 day, at least 2 days or at least 3 days.

Further details regarding the synthesis and characterisation of relevant MOF particles are known in the art, e.g. as disclosed in the description of WO2016/207397A1 , which is hereby incorporated by reference. RNA interference molecules In the last 20 years, RNA interference (RNAi), has appeared in various academic research and biotechnology start-ups and gained large amounts of interest and investment, si RNA is a double stranded RNA fragment typically twenty-one to twenty- three nucleotides in length, cleaved from endogenously expressed long double stranded RNAs (dsRNAs) by the enzyme Dicer. However, synthetic siRNAs have the potential to be inhibitors of various disease-associated genes, bypassing the first step of

endogenous cleavage from dsRNAs, and allowing for the creation of a platform technology with any genetic sequence. Gene knockdown utilizes siRNA and is the process by which siRNA effectively lowers a gene's expression. This is a huge asset to the medical field because it i) has the potential for high efficiency of knockdown - 100-1000 fold more efficient than antisense

oligonucelotides ii) is incredibly specific and thus limits off-targeting effects and iii) has a lack of systemic toxicity and immunoreactivity, as it is a biological molecule natively found in humans. Various medical diseases in addition to cancer, including neurological disorders and viral infections, are known to benefit from siRNA gene therapy. siRNA is therefore a versatile platform technology as the mechanism for its delivery and subsequent gene knockdown is universal for any selected sequence. RNA interference through short interfering RNA (siRNA) is a two-step process. First, dsRNA is cleaved within the cell to yield short interfering RNAs (siRNAs) of about 21 -23 nucleotides in length with 5' terminal phosphate and 3' short overhangs (of approximately 2 nucleotides). The siRNAs target the corresponding mRNA sequence specifically for destruction (Zamore P.D., 2001 ). RNAi may be also be efficiently induced using chemically synthesized siRNA duplexes of the aformentioned structure with 3'-overhang ends. Synthetic siRNA duplexes have been shown to specifically suppress expression of endogenous and heterologous genes in a wide range of mammalian cell lines (Elbashir, 2001 ). Thus siRNA duplexes containing between 20 and 25 bps, more preferably between 21 and 23 bps, of a target gene may form one part of the invention.

Another methodology known in the art for interference of expression of target genes is the use of "microRNA" (miRNA). This technology employs artificial miRNAs, which sequences can be generated using well-defined rules in the light of the disclosure herein. Thus, for example RNA molecules encoding a stem loop structure including a sequence portion of one of a target gene of around 20-25 nucleotides, optionally including one or more mismatches may form one part of the invention. Means for designing miRNA molecules are known in the art, see http: . wmd.weiaelworld.org bin mirnatools.pl for example. Thus, the present invention further provides the use of any of the RNA molecules described herein, for example: double stranded RNA (i.e. synthetic siRNA or mature miRNA) with appropriate specificity as described herein; a nucleic acid precursor of siRNA or miRNA as described herein; for down-regulation of gene expression, particularly down-regulation of expression of a target gene. The RNAi molecule may be present at a molecular ratio of least 1 :1 with the MOF component of the invention. More preferably, the RNAi molecule is present at a ratio of at least 3:1 or 10:1 with the MOF component of the invention (denoting a 3-fold or 10-fold greater abundance of the RNAi molecule, respectively). Yet more preferably, the RNAi molecule is present at a ratio of at least 30:1 or 100:1 with the MOF component of the invention.

In some embodiments, the amount of RNAi associated with the MOF in the composition of the invention is described by the amount of RNAi that is treated per mass of MOF. For instance, the composition of the invention may be prepared by treating (i.e. contacting) each milligram of MOF with between 10 picomoles to 500 picomoles of RNAi, or between 30 picomoles to 300 picomoles RNAi. Preferably, the composition of the invention may be prepared by treating each milligram of MOF with approximately 150 picomoles of RNAi. Endosome Release Factors

Several endosomal escape pathways are known in the art (Varkouhi et al, 201 1 ). For instance, viral and bacterial proteins have been identified that are involved in this process. Chemical and biochemical agents that facilitate endosomal escape may be termed "endosomal release factors" or "endosomal escape agents". To the inventors' knowledge, the successful incorporation of an endosomal release factor into a composition

comprising a MOF and an RNAi molecule has not before been disclosed.

One class of endosome release factors are pH buffering agents, which can mediate endosomal rupture via the so-called proton sponge effect. For instance, it is thought that agents with a high capacity for protonation can induce an influx of hydrogen ions into the endosome, causing osmotic swelling and ultimately endosomal lysis. One class of such agents are tertiary amines having a hydrophobic chain. Some such agents (1 ,8-

Bis(dimethylamino)naphthalene, N,N,N',N'-Tetramethyl-1 ,8-naphthalenediamine) are sold commercially under the trade name (registered trade mark) "Proton-sponge".

Polyethyleneimine (PEI) may also be used or adapted as a proton sponge.

Protein- and peptide-based agents form another class of endosome release factors.

Some such (poly)peptide based agents act to destabilise the endosomal membrane via membrane fusion. Many viruses have integral membrane peptides which undergo conformational changes in response to changing acidity, to induce fusion with the lipid bilayer. Such fusogenic proteins can be adapted for use as an endosome release factor. For instance, such fusogenic proteins can form the basis of a cell penetrating peptide that may form part of the composition of this invention. Other viral and bacterial proteins are able to form a pore in the endosomal membrane, allowing biomolecules to pass from the endosomal lumen, through the pore and into the cytosol.

Some known agents that may be adapted for use as the endosome release factor component of the claimed composition are set forth in Table 1 of Varkouhi et al, 201 1 , and listed here (with proposed mechanism of endosome escape given in parentheses): Hemagglutinin (Fusion); HA2/poly L-lysine (Fusion); dilNF-7 (Fusion); penton base (Pore); gp41 (Pore/fusion); gp41 /polyethylenimine (Pore/fusion/proton sponge); TAT (Unclear); Papillomavirus L2 (Fusion); envelope protein E of West Nile virus (Fusion); listeriolysin O (Pore); Pneumococcal pneumolysin (Pore); Streptococcal streptolysin O (Pore); Diphtheria toxin (Fusion); Pseudomonas aeruginosa exotoxin A (Pore); Shiga toxin (Pore); cholera toxin (Pore); Ricin (Unclear); Saporin (Unclear); Gelonin (Unclear); human calcitonin derived peptide, hCT (Unclear); fibroblast growth factors receptor

FGFR3 (Unclear); Melittin (Pore); R-Ahx-R 4 / AhxB (Unclear) glycoprotein H gpH from herpes simplex (Fusion); KALA (Fusion); GALA (Fusion); Synthetic surfactants (Fusion); Penetratin pAntp (Unclear); R6-Penetratin with arginine-residues (Unclear); EB1

(Unclear); bovine prion protein (Pore); Poly L-histidine (Proton sponge); Sweet Arrow Peptide SAP proline-rich (Fusion); polyethylenimine PEI (Proton sponge);

Polyamidoamines PAAs (Proton sponge); polypropylacrylic acid PPAA (Proton sponge); ammonium chloride (Proton sponge); chloroquine (Proton sponge); methylamine (Proton sponge). EXAMPLES

The inventors have provided compositions, which are capable of effectively delivering RNA interference molecules to cells and which can achieve a clear knockdown of the target gene of interest. These compositions are illustrated and better understood by reference to the non-limiting examples presented here:

Design of cell system and corresponding si RNA

The inventors produced a HEK 293 cell line that expresses red fluorescent protein (RFP). The cell line was produced using the commercially available T rex Flp-ln system to incorporate an inducible mCherry expression cassette into the cells.

The inventors also produced siRNA molecules with sequences that correspond with different regions of the mCherry sequence. From the genetic code of the mCherry used, five of the 32 (21 -nucleotide length) siRNA sequences were identified as 'promising' because of their relatively low GC content (<50%) and no stretches of greater than four T or A basepairs. Taking these five newly-designed mCherry-targeting siRNA molecules, a qualitative evaluation of their capacity to silence RFP expression was performed using fluorescence microscopy and western blot (Figure 1 A and 1 B respectively). Ponceau stain was used on the western blot gel to show equal loading (data not shown). Figure 1 A demonstrates that without inducing the cell line, no mCherry signal is present. The scrambled siRNA sample also shows the need for genetically specific siRNA to cause a signal knockdown. siRNA samples 2, 3, 17, 28, and 29 had varying levels of effectiveness at decreasing fluorescence. This effect is more clearly shown in Figure 1 B, where the intensity of the bands of anti-RFP antibody for samples 3 and 28 is

substantially reduced. The most effective (double-stranded) sequence having a sense strand sequence 5'-AAGGAGTTCATGCGCTTCAAG-3' was used for future

experimentation.

Loading of siRNA into MOF and characterization

The effects of loading the siRNA into the MOF on the characterization of the material was investigated. Figure 2 shows the powder X-ray diffraction (PXRD) patterns of NU-1000 and loading conditions. The NU-1000 (150nm) was loaded into an oven at 100C for approximately 3 days to ensure that all potential solvent inside the porosity was removed (Figure 2, nNU-1000 solvent evaporation). Some loss of minor peaks in the system is observed, but the majority of crystallinity is maintained. With a different sample, the NU- 1000 was soaked in RNase-free water at the same MOF concentration as the siRNA- loaded sample (20 mg/mL). A high degree of agreement with the PXRD pattern of the solvent evaporated NU-1000. However, siRNA loading of the MOF leads to a decrease in intensity of the major peaks at ca. 5^° and 7.5^° 2Θ, elimination of the ca. 10.5^° peak. The amount of siRNA loaded was approximately 150 pmol per mg of NU-1000, as quantified by Qubit.

The decrease in the intensity of the peaks suggests that the siRNA is inside the porosity. This is because the intensity is related to the electronic contrast (density) between two phases, in this case the framework skeleton and the porosity. If the porosity is empty, the contrast is high and the peaks would in general have higher intensity. If the porosity is filled with guest molecules, the contrast would be lower.

Protection of RNAi molecules from degradation

Regardless of the location of the siRNA, the inventors have determined that the siRNA is protected from enzymatic degradation (which is one of the major drawbacks of this type of biological therapy). As shown in Figure 3, a band is visible on the gel at the 21 nt position for the naked siRNA - unprotected by any system - only when the naked siRNA sample has not been in contact with the RNase enzyme. When treated with the enzyme, the naked siRNA is degraded and no band is visible at the 21 nt position. NU-1000 run through the gel without siRNA does not show a false positive band at the 21 nt siRNA position. Most of the MOF gets trapped higher up on the gel, demonstrated by the bright and smeared bands above approximately 80nt marked by the ladder of RNA standards.

A band is present at the 21 nt position of the lane loaded with the siRNA@NU-1000 composition. Contacting the siRNA@NU-1000 composition with the enzyme, which was shown to cleave naked unprotected siRNA in a previous lane, the siRNA band at 21 nt is still visible. This demonstrates the potential of NU-1000 to protect the siRNA from enzymatic degradation.

Preparation of the composition of the invention

To produce the compositions of the invention, the MOF was suspended in water and contacted first with the RNAi solution for 3 hours at 37C. Then, without removing the RNAi solution, the endosomal release factor was added to the mixture. This mixture was allowed to stand for an additional hour at 37C.

In vitro effect of siRNA on mCherry cell line

RFP expression in the transformed HEK 293 cell line was induced with tetracycline, and incubated with siRNA@nNU-1000, and controls, to quantify signal knockdown. Figure 4 shows the ineffectiveness of the naked siRNA when added to the cells, as the

normalized mCherry fluorescence does not deviate far from the normalized 100% level of no-siRNA added HEK 293 cells. The positive control, siRNA@lipofectamine, shows a significant decrease in signal down to a mean value of 60%. It was also verified that the MOF itself was not having any effect on the cellular expression of mCherry, as the mean value stays near 100%.

When the siRNA@nNU-1000 complex is applied in vitro, a wide range of results was observed. Even as run in quintuplicates, sometimes the mCherry fluorescence would be unchanged and other times it would be nearly as effective as the positive control siRNA@lipofectamine.

When the Proton-Sponge® was added (to form "siRNA@nNU-1000 Proton Sponge", which is a composition of the invention), the average expression decreased to ca. 78% of the normal HEK activated mCherry expressing cells. Compared with the minimal effect on expression of the Proton-Sponge® or NU-1000 alone (ca. 95% each), there is statistical significance (^** and ^***, respectively) for the difference of the complexed MOF with these values. (^* denotes a P value of ≤ 0.05; ^** denotes a P value of ≤ 0.01 ; and ^*** denotes a P value of ≤ 0.001 .) The amphipathic peptide KALA was added to form another composition of the invention ("siRNA@nNU-1000 KALA"), which decreased the mCherry signal to ca. 73%.

Interestingly, when the MOF was co-loaded with both of the abovementioned endosome release factors (KALA and PS) the gene silencing efficacy was reduced to ca. 82%. This value is not statistically different to those of siRNA@nNU-1000 Proton Sponge or siRNA@nNU-1000 KALA, indicating that there is not a cumulative effect of these two endosome release factors. Another endosome release factor, NH₄CI, was used to form a third composition of the invention, "siRNA@nNU-1000 NH₄CI". This reduced mCherry expression to ca. 75%. (These results cite the average expression percentage; they additionally can be visualized as a box and whiskers plot, demonstrating that in some cases - even when removing outliers - knockdown effects on par with that of the positive control are observed.)

Investigation into intracellular delivery

To investigate the efficacy of siRNA delivery, a tagged siRNA was used (conjugated with a fluorophore at 647nm). Figure 5 indicates that the siRNA@nNU-1000 composition does deliver siRNA into the cell, as the siRNA@nNU-1000 signal is significantly higher to that of the negative controls. This shows that siRNA is entering the cell, but sometimes, as shown by the upper part of the siRNA@nNU-1000 plot of Figure 4, it is not having a gene-silencing effect. Since siRNA must be in the cytoplasm to be effective in its signal knockdown pathway, this indicated to the inventors that mCherry gene knockdown success could be attributed to the siRNA@nNU-1000 complex - especially the siRNA - being caught in endosomes and possibly degraded before it could enter the cytoplasm. This finding led the inventors towards the inventive concept of including various other factors - either proton sponges or membrane opening peptides in the compositions of the invention, together with the siRNA and MOF.

Investigating the spatial position of the RNAi molecule

The question of whether the siRNA was located inside the MOF's porosity or outside on the external surface - or in both locations - can be answered through a microscopy technique called fluorescence-lifetime imaging microscopy (FLIM). This technique evaluates differences in exponential decay rate of fluorescence depending on the environment of the fluorescent molecule. In particular, the system uses time-correlated single photon counting (TCSPC) to compensate for variations in source intensity and the output of a fluorescence decay curve was recorded with relatively fast time resolutions. Due to the effect of Forster resonance energy transfer (FRET), the proximity of interaction between the loaded fluorophore tagged siRNA payload and MOF internally fluorescent linker can be measured as an alteration in emission lifetime. To ascertain whether the RNAi molecules are located in the pore or are adsorbed externally, change in the fluorescence decay lifetime of the MOF is measured following excitation at 488 nm. The fluorescence lifetime is around 4859 ps for the MOF alone.

NU-1000 was loaded with siRNA tagged with Alexa-Fluor 647. FLIM was used to evaluate: (i) unloaded NU-1000; (ii) NU-1000 loaded with tagged siRNA; and (iii) NU- 1000 loaded with tagged siRNA that was subsequently reacted with an RNase enzyme that can cleave siRNA of 21 nt in length. The results are presented in Figure 6.

The inventors believe that if the siRNA were externally located, even partially, the RNase enzyme would degrade it such that only any internally located siRNA would remain. The fluorescence lifetime (which is around 4859 ps for the MOF alone) decreases to 4180 ps for the siRNA-loaded nNU-1000, indicating that some of the fluorescence from the MOF is being passed to the siRNA fluorophore loaded in or around the structure. The lifetime for the enzyme-reacted siRNA loaded nNU-1000 is ca. 4393 ps. There is a statistically significant difference for both of the siRNA loaded MOFs compared with the unloaded nNU-1000. This demonstrates that the tagged siRNA structure is contributing a rate of decay for the nNU-1000's fluorescence lifetime. However, there is no statistically significant difference detected between the enzyme-reacted and non- enzyme-reacted conditions. This suggests that a negligible amount of tagged siRNA is located external on the MOF's surface and degraded by the enzyme, proposing that the majority of the siRNA is loaded within the internal porosity of the structure. This also indicates that this negligible amount of tagged siRNA removal did not have a significant effect on the MOFs fluorescence lifetime.

Investigation of the RNAi molecule with other MOF

The siRNA was also tested in the context of the PCN-777 Zr and PCN-777 Hf MOF, and protection from enzymatic degradation was demonstrated. The results are shown in Figure 7.

PCN-777 Zr and PCN-777 Hf were loaded with siRNA tagged with Alexa-Fluor 647 for ~2.5-3h at 37C. This preparation was then treated with the Shortcut RNase III enzyme for 20 min at 37C before being inactivated with EDTA. The Alexa-Fluor 647 signal strength was measured by loading the samples into a plate reader for a fluorescence intensity measurement with excitation at 635 nm and emission at 680 nm. The results are normalised to the level of fluorescence of the naked siRNA and are shown in Figure 7, which shows no significant reduction of the Alexa-Fluor 647 signal following RNase treatment of compositions comprising tagged siRNA and PCN-777 Zr at a size of 300 nm or PCN-777 Hf at a size of 130 nm. The Shortcut RNase III enzyme is known to degrade all siRNA (as shown in the gel of Figure 3 for instance). The Shortcut RNase III enzyme itself produces a fluorescent signal under these detection parameters, which was compensated for by subtracting the difference in signal between the naked tagged siRNA (without MOF) and the naked tagged siRNA with enzyme. While the invention has been described in conjunction with the exemplary embodiments described above, many equivalent modifications and variations will be apparent to those skilled in the art when given this disclosure. Accordingly, the exemplary embodiments of the invention set forth above are considered to be illustrative and not limiting. Various changes to the described embodiments may be made without departing from the spirit and scope of the invention.

Investigating the mechanism of the endosomal release factors

To better understand how the endosomal release cofactors were allowing for a lack of endosomal retention, an additional assay using FLIM was performed. Previously, advanced techniques using frequency domain FLIM have demonstrated the ability to assess pH changes in cellular compartments by measuring the change in modulation of excitation by a pH sensitive fluorophore. Increasing pH corresponds to decreasing modulation. However, modulation and fluorophore lifetime have an inverse relationship

ω, and m correspond to lifetime, repetition rate, and modulation, respectively.

Therefore, increasing fluorescence lifetime of Oregon Green 488-conjugated dextrans corresponds to increasing pH. Using TCSPC FLIM, the fluorophore-conjugated dextran can be monitored once it enters the cell through the endocytosis pathway. It is commonly established in literature that dextran enters cells routinely through fluid-phase

endocytosis. There is also evidence to suggest that dextran of the particular size used in this disclosure also enters through clathrin-mediated endocytosis, which is comparable to the pathway of entry that the nNU-1000 MOF uses. Therefore, the presence of the endosomal release cofactor - if it was causing a modification to the normal vesicular pH as a way to perturb endosomal retention - would cause a change in the lifetime of the dextran-conjugated fluorophore in the same vesicle. To test this hypothesis, cells were incubated with dextran-conjugated Oregon Green 488 (dextran-OG488); in some cases a cofactor was also added. Figure 8 shows the lifetimes of these experimental conditions - dextran-OG488 only (control), KALA-added dextran-OG488 (KALA), NH4CI-added dextran-OG488 (NH4CI), and Proton Sponge-added dextran-OG488 (PS) - as tested with FLIM. This was done to quantify the lysosome pH indirectly through dextran-OG488 lifetimes. The graph shows a scatterplot as well as the mean and standard error of all the replicates. The fluorescence lifetime for the dextran-OG488 alone is 2794 ± 50 ps. This value is non-significantly changed to 2803 ± 70 ps for the KALA-added dextran-OG488. There is a significant increase of the lifetime to 2860 ± 40 ps with the NH4CI-added dextran-OG488.

Additionally, while more variable, there is also a significant increase in the lifetime for Proton Sponge-added dextran-OG488 to 2921 ± 140.1 ps. The significant increases in the lifetimes indicate a trend in the intra-compartmental pH increasing. These results suggest that these two cofactors, PS and NH4CI, are acting in a mechanism that is increasing the vesicular pH from its normal value. Unsurprisingly, as these cofactors are both synthetic molecules with the ability to entrap protons, these results support the idea that the increase in pH is allowing the siRNA loaded MOF to escape degradation and be released into the cytosol. KALA, on the contrary, is a cell-penetrating peptide, which - as discussed above - fuses with the membrane and is not explicitly acting as a proton absorber. The lack of lifetime change for the KALA data-points of Figure 8 correlated to a lack of pH change, corresponding well to this notion. Overall, the results in Figure 8 confirm the suggested mechanisms proposed in the literature for endosomal release.

All references referred to above are hereby incorporated by reference. References

Elbashir SM. et al. Nature, 41 1 , 494-498, (2001 )

He et al, J Am Chem Soc; Nanoscale Metal-Organic Frameworks for the Co-Delivery of Cisplatin and Pooled siRNAs to Enhance Therapeutic Efficacy in Drug-Resistant Ovarian Cancer Cells, 136(14): 5181 -5184, 2014.

P. Horcajada, R. Gref, T. Baati, P. K. Allan, G. Maurin, P. Couvreur, G. Ferey, R. E. Morris, and C. Serre Chem. Rev. 2012, 1 12, 1232.

Milletti F, Drug Discov Today; Cell-penetrating peptides: classes, origin, and current landscape. 17: 850-860, 2012.

Peng et al., ASC Nano; Nanosizing a Metal-Organic Framework Enzyme Carrier for Accelerating Nerve Agent Hydrolysis. 10 ( 10) , pp 9174-9182

Varkouhi et al, Journal of Controlled Release 151 (201 1 ) 220-228

Zamore P.D. Nature Structural Biology, 8, 9, 746-750, (2001 )

Zhang et al. J. Am Chem. Soc, Piezofluorochromic Metal-Organic Framework: A

Microscissor Lift. 2015, 137 (32), pp 10064-10067

Claims

1 . A composition comprising:

(i) a metal organic framework (MOF) particle;

(ii) an RNAi molecule; and

(iii) an endosome release factor,

wherein the MOF material is mesoporous, such that the diameter of pores opening at the surface of the MOF particles and extending into the MOF particles is at least 2 nm and not more than 10 nm and wherein the length of the pores formed within the MOF crystal structure is optionally greater than the diameter of the pores.

2. The composition according to claim 1 , wherein the length of the pores formed within the MOF crystal structure is at least 10 nm.

3. The composition according to claim 1 or claim 2, wherein the MOF unit cell comprises at least one metal ion selected from the group consisting of Zr ions, Hf ions, Bi ions, Fe ion and Mg ions.

4. The composition according to claim 3, wherein the MOF has the structure

Zr6^3-OH)8(OH)₈(linker)2, wherein the linker is an organic linker such as TBAPy (1 ,3,6,8- tetrakis(p-benzoic acid)pyrene).

5. The composition according to any one of the preceding claims, wherein the MOF particles have a longest linear dimension of at most 400 nm, assessed as a number average.

6. The composition according to any one of the preceding claims, wherein the diameter of pores opening at the surface of the MOF particles and extending into the MOF particles is not more than 7 nm.

7. The composition according to any one of the preceding claims, wherein the RNAi is a siRNA.

8. The composition according to any one of claims 1 -6, wherein the RNAi is a miRNA.

9. The composition according to any one of the preceding claims, wherein the RNAi is capable of downregulating the expression of a gene implicated in tumour cell survival.

10. The composition according to claim 9, wherein the gene is a multidrug resistance gene.

1 1 . The composition according to any one of the preceding claims, wherein the endosome release factor is a peptide.

12. The composition according to claim 1 1 , wherein the peptide is a cell penetrating peptide (CPP).

13. The composition according to claim 1 1 or claim 12, wherein the peptide is a membrane opening peptide (MOP).

14. The composition according to any one of claims 1 1 -13, wherein the peptide is an amphipathic peptide.

15. The composition according claim 14, wherein the amphipathic peptide is KALA.

16. The composition according to any one of claims 1 -10, wherein the endosome release factor is NH₄CI.

17. The composition according to any one of claims 1 -10, wherein the endosome release factor is a proton sponge.

18. An in vitro method of delivering an RNAi molecule to a cell, the method comprising treating the cell with the composition according to any one of claims 1 -17.

19. The in vitro method according to claim 18, wherein the cell is a mammalian cell.

20. The in vitro method according to claim 19, wherein the cell is a tumour cell.

21 . The composition according to any one of claims 1 -17 for use in the treatment of the human or animal body by therapy.

22. A method of treating a patient in need thereof, the method comprising administering the patient with the composition according to any one of claims 1 -17.

23. The composition according to any one of claims 1 -17 for use in a method of treating a cancer in a patient, the method comprising administering the composition to the patient.

24. A method of treating a cancer patient, the method comprising administering the cancer patient with the composition according to any one of claims 1 -17.

25. Use of a metal organic framework (MOF) particle in the manufacture of an anticancer medicament, wherein said manufacture comprises the step of forming a composition by treating the MOF with (i) an RNAi molecule and (ii) an endosome release factor,

wherein the MOF material is mesoporous, such that the diameter of pores opening at the surface of the MOF particles and extending into the MOF particles is at least 2 nm and not more than 10 nm and wherein the length of the pores formed within the MOF crystal structure is greater than the diameter of the pores.

26. The composition for the use according to claim 23, the method according to claim 24 or the use according to claim 25, wherein the composition is administered at the tumour site.

27. The composition for the use according to claim 23, the method according to claim 24 or the use according to claim 25, wherein the composition is administered orally.

28. The composition for the use according to claim 23, the method according to claim 24 or the use according to claim 25, wherein the composition is administered by intravenous injection.

29. The method, the composition for the use, or the use according to claim 26, wherein the composition is administered by intratumoural injection.

30. The method, the composition for the use, or the use according to claim 26, wherein the composition is administered topically.

31 . The method, the composition for the use, or the use according to claim 26, wherein the composition is administered by inhalation.

32. The method, the composition for the use, or the use according to any one of claims 23-31 , wherein the cancer is selected from the group consisting of breast cancer, mesothelioma and pancreatic cancer.

33. The method, the composition for the use, or the use according to claim 29 or claim 30, wherein the tumour is a skin tumour.

34. The method, the composition for the use, or the use according to claim 31 , wherein the tumour is a lung cancer.

35. The method, the composition for the use, or the use according to any one of claims 23-34, wherein the RNAi molecule comprises a sequence that corresponds with a sequence of about 15 nucleotides of a gene that is overexpressed by the tumour.

36. A method of making the composition according to any one of claims 1 -17, the method comprising providing a MOF, and treating the MOF with (i) an RNAi molecule and (ii) an endosome release factor,

37. The method according to claim 36, further comprising a step of formulating the composition as a pharmaceutical composition.