Lipid-nucleic acid particles prepared via a hydrophobic lipid-nucleic acid complex intermediate and use for gene transfer

Description Submit all comments and votes

FIELD OF THE INVENTION

This invention relates to lipid-nucleic acid particles which are useful for the introduction of nucleic acids into cells, and methods of making and using them. The invention provides a circulation-stable, characterizable delivery vehicle for the introduction of plasmids or antisense compounds into cells. These vehicles are safe, stable, and practical for clinical use.

BACKGROUND OF THE INVENTION

Gene transfer into genetically impaired host cells in order to correct the genetic defects has vast potential for successfully treating a variety of thus far hitherto untreatable medical conditions. There are currently six major non-viral methods by which genes are introduced into host cells: (i) direct microinjection, (ii) calcium phosphate precipitation, (ii) DEAE-dextran complexes, (iv) electroporation, (v) cationic lipid complexes and (vi) reconstituted viruses and virosomes (see Chang, et al., Focus 10:88 (1988)).

Most reported examples of gene transfer have been performed in vitro. In vivo gene transfer is complicated by serum interactions, immune clearance, enzymatic degradation of the genes, toxicity and biodistribution. In in vivo administration, selection is not possible, and a reasonably high frequency of transformation is necessary to achieve sufficient expression to compensate for a defective endogenous gene.

The in vivo gene transfer methods under study in the clinic consist almost entirely of viral vectors. Although viral vectors have the inherent ability to transport nucleic acids across cell membranes and some can integrate exogenous DNA into the chromosomes, they can carry only limited amounts of DNA. In addition, their use poses significant risks. One such risk is that the viral vector may revert to a pathogenic genotype either through mutation or genetic exchange with a wild type virus.

In view of these limitations and risks, alternative non-viral-based gene transfer methods have been developed. These methods use often plasmid vectors, which are small circular sequences of DNA, as vectors for DNA delivery. However, most plasmids do not possess the attributes required for intracellular delivery and therefore sophisticated delivery systems are required.

Cationic lipid complexes are presently the most effective generally used means of introducing non-viral nucleic acids into cells. A number of different formulations incorporating cationic lipids are commercially available. These include:(i) LIPOFECTION.RTM. (which uses 1,2-dioleyloxy-3-(N,N,N-trimethylamino)propane chloride, or DOTMA, see Eppstein, et al., U.S. Pat. No. 4,897,355); LIPOFECTAMINE.RTM. (which uses DOSPA, see Hawley-Nelson, et al., Focus 15(3):73 (1993)); and LIPOFECTACE.RTM. (which uses N,N-distearyl-N,N-dimethyl-ammonium bromide, or DDAB, see Rose, U.S. Pat. No. 5,279,833). Others have reported alternative cationic lipids that work in essentially the same manner but with different efficiencies, for example 1,2-dioleoyloxy-3-(N,N,N-trimethylamino) propane chloride, or DOTAP (see Stomatatos, et al., Biochemistry 27: 3917-3925 (1988)); glycerol based lipids (see Leventis, et al., Biochem. Biophys. Acta 1023:124 (1990); lipopolyamines (see, Behr, et al., U.S. Pat. No. 5,171,678) and cholesterol based lipids (see Epand, et al., WO 93/05162, and U.S. Pat. No. 5,283,185). It has been reported that DOTMA and related compounds are significantly more active in gene transfer assays than their saturated analogues (see, Feigner, et al., WO91/16024). However, both DOTMA and DOSPA based formulations, despite their efficiency in effecting gene transfer, are prohibitively expensive. DDAB on the other hand is inexpensive and readily available from chemical suppliers but is less effective than DOTMA in most cell lines. Another disadvantage of the current lipid systems is that they are not appropriate for intravenous injection.

Lipid-based vectors used in gene transfer have generally been formulated in one of two ways. In one method, the nucleic acid is introduced into preformed liposomes made of mixture of cationic lipids and neutral lipids. The complexes thus formed have undefined and complicated structures and the lipofection efficiency is severely reduced by the presence of serum. A second method involves the formation of DNA complexes with mono- or poly-cationic lipids without the presence of a neutral lipid. These complexes are often prepared in the presence of ethanol and are not stable in water. Additionally, these complexes are adversely affected by serum (see, Behr, Acc. Chem. Res. 26:274-78 (1993)).

An examination of the relationship between the chemical structure of the carrier vehicle and its efficiency of gene transfer has indicated that the characteristics which provide for effective gene transfer would make a carrier unstable in circulation (see, Ballas, et al., Biochim. Biophys. Acta 939:8-18 (1988)). Additionally, degradation either outside or inside the target cell remains a problem (see, Duzghines, Subcellular Biochemistry 11:195-286 (1985)). Others who have attempted to encapsulate DNA in lipid-based formulations have not overcome these problems (see, Szoka et al., Ann. Rev. Biophys. Bioeng. 9:467 (1980); Deamer, U.S. Pat. No. 4,515,736, and Legendre, Pharm. Res. 9:1235-1242 (1992)).

Ideally, a delivery vehicle for a nucleic acid or plasmid will have the following characteristics: a) ease of preparation, b) capable of carrying a large amount of DNA per particle to enable gene transfer of all sizes of genes and reduce the volume of injection, c) homogenous, d) reproducible, e) is serum stable with minimal serum interactions and shields DNA from extracellular degradation, and f) is capable of transfecting target cells in such a way that the DNA is not digested intracellularly.

The present invention provides such compositions and methods for their preparation and use.

SUMMARY OF THE INVENTION

The present invention comprises novel, lipid-nucleic acid particles. The invention also comprises methods of making and using these particles.

In some embodiments, the particles are made by formation of hydrophobic intermediate complexes in either detergent-based or organic solvent-based systems, followed by removal of the detergent or organic solvent. Preferred embodiments are charge-neutralized.

In one embodiment, a plasmid is combined with cationic lipids in a detergent solution to provide a coated plasmid-lipid complex. The complex is then contacted with non-cationic lipids to provide a solution of detergent, a plasmid-lipid complex and non-cationic lipids, and the detergent is then removed to provide a solution of serum-stable plasmid-lipid particles, in which the plasmid is encapsulated in a lipid bilayer. The particles thus formed have a size of about 50-150 nm.

In another embodiment, serum-stable plasmid-lipid particles are formed by preparing a mixture of cationic lipids and non-cationic lipids in an organic solvent; contacting an aqueous solution of plasmid with the mixture of cationic and non-cationic lipids to provide a clear single phase; and removing the organic solvent to provide a suspension of plasmid-lipid particles, in which the plasmid is encapsulated in a lipid bilayer, and the particles are stable in serum and have a size of about 50-150 nm.

Another method of forming lipid-nucleic acid particles involves:

(a) contacting nucleic acids with a solution of non-cationic lipids and a detergent to form a nucleic acid-lipid mixture;

(b) contacting cationic lipids with the nucleic acid-lipid mixture to neutralize the negative charge of said nucleic acids and form a charge-neutralized mixture of nucleic acids and lipids: and

(c) removing the detergent from the charge-neutralized mixture to provide the lipid-nucleic acid particles in which the nucleic acids are protected from degradation.

Another method of forming lipid-nucleic acid particles involves:

(a) contacting an amount of cationic lipids with nucleic acids in a solution; the solution comprising of from about 15-35% water and about 65-85% organic solvent and the amount of cationic lipids being sufficient to produce a +/- charge ratio of from about 0.85 to about 2.0, to provide a hydrophobic, charge-neutralized lipid-nucleic acid complex;

(b) contacting the hydrophobic, charge-neutralized lipid-nucleic acid complex in solution with non-cationic lipids, to provide a lipid-nucleic acid mixture; and

(c) removing the organic solvents from the lipid-nucleic acid mixture to provide lipid-nucleic acid particles in which the nucleic acids are protected from degradation.

The lipid-nucleic acid particles of the present invention are useful for the therapeutic delivery of nucleic acids. In one embodiment, the particles are constructed via a hydrophobic lipid-nucleic acid intermediate (or complex). Upon removal of a solubilizing component (i.e., detergent or an organic solvent) the nucleic acid becomes protected from degradation. The particles thus formed are suitable for use in intravenous nucleic acid transfer as they are stable in circulation, of a size required for pharmacodynamic behavior resulting in access to extravascular sites and target cell populations.

In particular, it is an object of this invention to provide in vitro and in vivo methods for treatment of diseases which involve the overproduction or underproduction of particular proteins. In these methods, a nucleic acid encoding a desired protein or blocking the production of an undesired protein, is formulated into a lipid-nucleic acid particle, and the particles are administered to patients requiring such treatment. Alternatively, cells are removed from a patient, transfected with the lipid-nucleic acid particles described herein, and reinjected into the patient.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 illustrates a nucleic acid-lipid particle-mediated gene transfer using "sandwich-type" complexes of DNA.

FIG. 2 illustrates an aggregation and precipitation which commonly occurs during the entrapment of large nucleic acids in lipid complexes.

FIG. 3 provides a schematic representation of the preparation of plasmid-lipid particles according to certain embodiments of the present invention.

FIG. 4 illustrates the recovery of .sup.3 H-DNA from encapsulated particles following the reverse-phase preparation of the particles and extrusion through a 400 nm filter and a 200 nm filter. Lipid composition is POPC:DODAC:PEG-Cer-C.sub.20. PEG-CerC.sub.20 was held constant at 10 mole % and POPC and DODAC were changed relative to each other. 20 mg lipid; 50 .mu.g plasmid DNA (7.5 kbp).

FIG. 5 illustrates the recovery of 3H-DNA from particles prepared using a reverse-phase procedure. The particles were extruded through a 200 nm filter and eluted on a DEAE Sepharose CL-6B anion exchange column. The percent recovery reported is based on the amount recovered after filtration. Lipid composition is as in FIG. 4.

FIG. 6 illustrates the recovery of .sup.14 C-lipid from encapsulated particles following the reverse-phase preparation of the particles and extrusion through a 400 nm filter and a 200 nm filter. Lipid composition is as in FIG. 4.

FIG. 7 illustrates the recovery of .sup.14 C-lipid from particles prepared using a reverse-phase procedure. The particles were extruded through a 200 nm filter and eluted on a DEAE Sepharose CL-6B anion exchange column. The percent recovery reported is based on the amount recovered after filtration. Lipid composition is as in FIG. 4.

FIG. 8 illustrates the effect of DODAC concentration on the encapsulation of plasmid DNA. Encapsulation efficiency was measured by anion exchange chromatography. Vesicles were composed of DOPE, DODAC and 10 mole % PEG-Cer-C.sub.20 (symbol) or EPC, DODAC and 10 mole % PEG-Cer-C.sub.20 (symbol). Total lipid and DNA concentrations were 10 mmole/ml and 50 .mu.g/ml, respectively.

FIGS. 9A and 9B illustrate the effect of serum nucleases on free pCMVCAT DNA as assessed by column chromatography before (A) and after (B) incubation in 80% mouse serum. Free .sup.3 H-DNA (PCMVCAT) was eluted on a Sepharose CL-4B column in HBS, pH 7.4.

FIG. 10 illustrates the effect of serum nucleases on encapsulated pCMVCAT DNA (prepared by reverse-phase) as assessed by column chromatography. Sepharose CL-4B column profile of encapsulated pCMV plasmid incubated in 80% mouse serum for 30 min. (A) External DNA was removed by ion exchange chromatography prior to incubation in serum. (B) External DNA was not removed prior to incubation in serum. Lipid composition was POPC:DODAC:PEG-Cer-C.sub.20. Total lipid and plasmid concentrations were 20 .mu.mole/ml and 50 .mu.g/ml prior to anion exchange chromatography.

FIGS. 11A and 11B illustrate the effect of serum nucleases on encapsulated pCMVCAT DNA (prepared by detergent dialysis) as assessed by column chromatography. Sepharose CL-4B column profile of encapsulated pCMV plasmid incubated in 80% mouse serum for 30 min. (A) External DNA was removed by ion exchange chromatography prior to incubation in serum. (B) External DNA was not removed prior to incubation in serum. The lipid composition was DOPE:DODAC:PEG-Cer-C.sub.20 (84:6:10). Total lipid and plasmid concentrations were 10 .mu.mole/ml and 400 .mu.g/ml prior to anion exchange chromatography.

FIGS. 12A and 12B illustrate the resistance of plasmid complexed to preformed liposomes composed of DOPE:DODAC(50:50) (A) and plasmid encapsulated within DOPE:DODAC:PEG-Cer-C.sub.14 particles (B) to digestion by DNAse I. Plasmid DNA was extracted and subjected to PCR (polymerized chain reaction) to amplify for visualization on a gel. Free plasmid was used as a control. Lane 1:1 kb DNA marker; Lane 2: PCR negative control (no DNA); Lane 3: free plasmid alone; Lane 4: free plasmid in 0.05% detergent (Triton X-100); Lane 5: free plasmid incubated with DNAse I in the absence of detergent; Lane 6: free plasmid incubated with DNAse I in the presence of detergent: Lane 7: complexed (A) or encapsulated (B) plasmid alone; Lane 8: complexed (A) or encapsulated (B) plasmid in 0.05% detergent; Lane 9: complexed (A) or encapsulated (B) plasmid incubated in DNAse I in the absence of detergent; Lane 10: complexed (A) or encapsulated (B) plasmid incubated in DNAse I in the presence of detergent.

FIG. 13 illustrates the effect of plasmid DNA concentration on encapsulation efficiency (detergent dialysis). Vesicles were composed of DOPE: DODAC:PEG-Cer (84:6: 10) at a lipid concentration of 10 .mu.mole/ml.

FIG. 14 illustrates the effect of NaCl concentration on the optimal DODAC concentration for plasmid entrapment. Lipid composition was DOPE:DODAC:PEG-Cer-C.sub.14 (or PEG-Cer-C.sub.20). PEG-Cer was held constant at 10 mole %. Total lipid concentration was 10 .mu.mole/ml. Plasmid concentration was 50 .mu.g/ml.

FIG. 15 illustrates the size distribution of plasmid:lipid particles prepared by the detergent dialysis procedure (Volume weighted analysis). Lipid composition was DOPE:DODAC:PEG-Cer-C.sub.20) (84:6:10).

FIG. 16 illustrates the size distribution of plasmid:lipid particles prepared by the detergent dialysis procedure (Number weighted analysis). Lipid composition was DOPE:DODAC:PEG-Cer-C.sub.20 (84:6:10).

FIGS. 17A and 17B provide electron micrographs of liposomes composed of DOPE:DODAC:PEO-Cer-C.sub.20 without encapsulated plasmid (A) and the plasmid:lipid particles (B). The small arrows denote empty liposomes approximately 100 nm in diameter. These are compared to electron-dense particles surrounded by a membrane bilayer (large arrows). Scale bar=100 nm.

FIG. 18 shows the clearance of .sup.3 H-DNA and .sup.14 C-lipid from particles (prepared by reverse-phase methods) after injection into ICR mice. The figure includes free .sup.3 H-DNA after injection as a comparison. Lipid composition is POPC:DODAC:PEG-Cer-C.sub.20.

FIGS. 19A and 19B show the clearance .sup.3 H-DNA and .sup.14 C-lipid from particles (prepared by detergent dialysis methods) after injection into ICR mice. Lipid compositions were (A) DOPE:DODAC-PEG-Cer-C.sub.20 (84:6:10) and (B) DOPE:DODAC-PEG-Cer-C.sub.14 (84:6:10).

FIG. 20 shows the results of in vivo gene transfer which occurs in the lungs of mice. Lipid composition is DOPE-DODAC-PEG-Cer-C.sub.20 or DOPE:DODAC:PEG-Cer-C.sub.14 (84:6: 10).

FIG. 21 shows the results of in vivo gene transfer which occurs in the liver mice. Lipid composition is DOPE-DODAC-PEG-Cer-C.sub.20 or DOPE:DODAC:PEG-Cer-C.sub.14 (84:6: 10).

FIG. 22 shows the results of in vivo gene transfer which occurs in the spleen of mice. Lipid composition is DOPE-DODAC-PEG-Cer-C.sub.20 or DOPE:DODAC:PEG-Cer-C.sub.14 (84:6:10).

FIG. 23 shows the effect of increasing amounts of LIPOFECTIN.RTM. (DOTMA/DOPE; 50:50 mol ratio) on the recovery of .beta.gal plasmid DNA in the aqueous phase following Bligh and Dyer extraction of the lipid-nucleic acid complexes.

FIGS. 24A and 24B show the effect of increasing amounts of cationic lipid on the recovery of plasmid DNA in the aqueous (A) and organic (B) phase following Bligh and Dyer extraction of the lipid-nucleic acid complexes.

FIGS. 25A, 25B, 25C and 25D show the recovery of plasmid DNA from aqueous (A and C) and organic (B and D) fractions following Bligh and Dyer extraction and expressed as a function of charge ratio (+/-).

FIGS. 26A and 26B illustrate the DNA condensation by poly-L-lysine and DODAC assayed by TO-PRO-1 dye intercalation. Condensation state was assessed in a Bligh and Dyer monophase (A) and in 100 mM OGP (B).

FIG. 27 illustrates the effects of increasing amounts of OGP on the recovery of plasmid DNA from the aqueous and organic phases following Bligh and Dyer extraction of lipid-nucleic acid complexes (plasmid/DODAC).

FIG. 28 shows the effects of increasing amounts of NaCl on the recovery of plasmid DNA from the aqueous phase following Bligh and Dyer extraction of lipid-nucleic acid complexes.

FIGS. 29A and 29B show the effect of poly-L-lysine and DODAC on the electrophoretic mobility of plasmid DNA.

FIG. 30 illustrates a protocol for preparing lipid-nucleic acid particles using detergent dialysis.

FIGS. 31A and B are bar graphs which illustrates the QELS results of a typical lipid-nucleic acid complex mixture prepared from .beta.-gal plasmid/DODAC/ESM.

FIG. 32 is a bar graph which illustrates the fluorescence spectroscopic evaluation of DNA condensation in the lipid-nucleic acid complexes using TO-PRO-1 dye intercalation. The results show that .beta.-gal plasmid in DODAC/ESM is condensed and protected against dye intercalation by the lipid, and that OGP can uncondense the particle.

FIG. 33 shows the results of electrophoresis of DNA extracted from lipid-nucleic acid complexes following digestion with DNase I. DNA within the complex is protected from DNase I degradation whereas uncomplexed DNA is not protected.

FIG. 34 provides the results of CHO cell lipofection using .beta.-gal plasmid/DODAC/ESM as assayed by .beta.-gal enzyme activity.

FIGS. 35A and B show changes in sample turbidity measured by 90.degree. light scattering at 600 nm during the preparation of nucleic acid-lipid particles in the presence of 100 mM (A) or 20 mM (B) n-octyl .beta.-D-glucopyranoside (OGP).

FIG. 36 shows solubilization of preformed DODAC (.circle-solid.) and SM (.box-solid.) vesicles in OGP as measured by 90.degree. light scattering. The concentrations of lipids used