Cationic lipids for intracellular delivery of biologically active molecules

Description Submit all comments and votes

BACKGROUND OF THE INVENTION

The present invention relates to cationic lipids which are used to enhance delivery of biologically active agents, particularly polynucleotides, proteins, peptides, and drug molecules, by facilitating transmembrane transport or by encouraging adhesion to biological surfaces. It relates particularly to cationic lipids comprising ammonium groups.

Some bioactive substances do not need to enter cells to exert their biological effect, because they operate either by acting on cell surfaces through cell surface receptors or to by interacting with extracellular components. However, many natural biological molecules and their analogues, including proteins and polynucleotides, or foreign substances, such as drugs, which are capable of influencing cell function at the subcellular or molecular level are preferably incorporated within the cell in order to produce their effect. For these agents the cell membrane presents a selective barrier which is impermeable to them.

Just as the plasma membrane of a cell is a selective barrier preventing random introduction of potentially toxic substances into the cell, the human body is surrounded by protective membranes which serve a similar defensive function to the whole organism. These membranes include skin, gastric mucosa, nasal mucosa and the like. While these membranes serve a protective function preventing entry of toxic substances, they can also prevent passage of potentially beneficial therapeutic substances into the body. The complex composition of the cell membrane comprises phospholipids, glycolipids, and cholesterol, as well as intrinsic and extrinsic proteins, and its functions are influenced by cytoplasmic components which include Ca.sup.++ and other metal ions, anions, ATP, microfilaments, microtubules, enzymes, and Ca.sup.++ -binding proteins. Interactions among structural and cytoplasmic cell components and their response to external signals make up transport processes responsible for the membrane selectivity exhibited within and among cell types.

Successful intracellular delivery of agents not naturally taken up by cells has been achieved by exploiting the natural process of intracellular membrane fusion, or by direct access of the cell's natural transport mechanisms which include endocytosis and pinocytosis (Duzgunes, N., Subcellular Biochemistry 11:195-286 (1985).

The membrane barrier can be overcome in the first instance by associating these substances in complexes with lipid formulations closely resembling the lipid composition of natural cell membranes. These lipids are able to fuse with the cell membranes on contact, and in the process, the associated substances are delivered intracellularly. Lipid complexes can not only facilitate intracellular transfers by fusing with cell membranes but also by overcoming charge repulsions between the cell membrane and the molecule to be inserted. The lipids of the formulations comprise an amphipathic lipid, such as the phospholipids of cell membranes, and form hollow lipid vesicles, or liposomes, in aqueous systems. This property can be used to entrap the substance to be delivered within the liposomes; in other applications, the drug molecule of interest can be incorporated into the lipid vesicle as an intrinsic membrane component, rather than entrapped into the hollow aqueous interior.

Intracellular delivery of beneficial or interesting proteins can be achieved by introducing expressible DNA and mRNA into the cells of a mammal, a useful technique termed transfection. Gene sequences introduced in this way can produce the corresponding protein coded for by the gene by using endogenous protein synthetic enzymes. The therapy of many diseases could be enhanced by the induced intracellular production of peptides which could remain inside the target cell, be secreted into the local environment of the target cell, or be secreted into the systemic circulation to produce their effect.

Various techniques for introducing the DNA or mRNA precursors of bioactive peptides into cells include the use of viral vectors, including recombinant vectors and retroviruses, which have the inherent ability to penetrate cell membranes. However, the use of such viral agents to integrate exogenous DNA into the chromosomal material of the cell carries a risk of damage to the genome and the possibility of inducing malignant transformation. Another aspect of this approach which restricts its use in vivo is that the integration of DNA into the genome accomplished by these methods implies a loss of control over the expression of the peptide it codes for, so that transitory therapy is difficult to achieve and potential unwanted side effects of the treatment could be difficult or impossible to reverse or halt.

Liposomes have been discussed as possible in vivo delivery vehicles and some encouraging results using this approach to the intracellular expression of DNA have been obtained (Mannino, R. J. Fould-Fogerite, S., Biotechniques 6, 682-690 (1988); Itani, T., Ariga, H., Yamaguchi, N., Tadakuma, T. & Yasuda, T. Gene 56 267-276 (1987); Nicolau, C. Legrand, A. & Grosse, G. E. Meth. Enz. 149 157-176 (1987); Straubinger, R. M. & Papahadjopoulos, D. Meth. Enz. 101 512-527 (1983); Wang, C. Y. & Huang, L. Proc Natl. Acad. Sci. USA 84 7851-7855 (1987)); however, the methodology has fundamental problems. Chief among the difficulties is the failure of liposomes to fuse with the target cell surface, but to be taken up phagocytically instead. Phagocytized liposomes are delivered to the lysosomal compartment, where polynucleotides are subjected to the action of digestive enzymes and degraded, leading to low efficiency of expression.

A major advance in this area was the discovery that a positively charged synthetic cationic lipid, N-[1-(2,3-dioleyloxy)propyl]-N,N,N-trimethylammonium chloride (DOTMA), in the form of liposomes, or small vesicles, could interact spontaneously with DNA to form lipid-DNA complexes which are capable of fusing with the negatively charged lipids of the cell membranes of tissue culture cells, resulting in both uptake and expression of the DNA (Felgner, P. L. et al. Proc. Natl. Acad. Sci., USA 84:7413-7417 (1987) and U.S. Pat. No. 4,897,355 to Eppstein, D. et al.). Others have successfully used a DOTMA analogue, 1,2-bis(oleoyloxy)-3-(trimethylammonio)propane (DOTAP) in combination with a phospholipid to form DNA-complexing vesicles. The Lipofectin.TM. reagent (Bethesda Research Laboratories, Gaithersburg, Md.), an effective agent for the delivery of highly anionic polynucleotides into living tissue culture cells comprises positively charged DOTMA liposomes which interact spontaneously with negatively charged polynucleotides to form complexes. When enough positively charged liposomes are used, the net charge on the resulting complexes is also positive. Positively charged complexes prepared in this way spontaneously attach to negatively charged cell surfaces, fuse with the plasma membrane, and efficiently deliver functional polynucleotide into, for example, tissue culture cells.

Although the use of known cationic lipids overcomes many problems associated with conventional liposome technology for polynucleotide delivery in vitro, several problems related to both in vitro and in vivo applications remain. First, although the efficiency of cationic lipid mediated delivery is relatively high compared to other methods, the absolute level of gene product produced is typically only several hundred copies per cell on average. Thus it would be desirable to improve delivery and expression by a factor of 10 to 1000-fold to achieve useful methodologies. Secondly, known cationic lipids such as DOTMA are toxic to tissue culture cells; thus, any improvements that reduce in vitro toxicity would strengthen the methodology.

A significant body of information is emerging regarding the use of other cationic lipids for the delivery of macromolecules into cells. Loyter prepared vesicles containing a quaternary ammonium surfactant that are capable of transferring functional tobacco mosaic virus into plant protoplasts. (Ballas, N., Zakai, N., Sela, I. and Loyter, A. Biochim. Biophys Acta 939 8-18 (1988)). Huang used cetyltrimethylammonium bromide to obtain functional expression from the chloramphenicol acetyl transferase gene transfected into mouse fibroblasts (Pinnaduwage, P., Schmitt, L. and Huang, L. Biochim. Biophys Acta 985 33-37 (1989)). Behr has shown that a novel lipophilic derivative of spermine can transfect primary pituitary cells (Behr, J-P, Demeneix, B., Loeffler, J-P and Perez-Mutul, J. Proc. Natl. Acad. Sci. USA 86 6982-6986 (1989)). Finally, John Silvius has shown that a cationic lipid (DOTAP), originally synthesized by Eibl (Eibl, H. and Woolley, P. Biophys. Chem. 10 261-271 (1979)) forms liposomes which can- fuse with negatively charged liposomes and can deliver functional DNA and RNA into tissue culture fibroblasts (Stamatatos, L., Leventis, R., Zuckermann, M. J. & Silvius, J. R. Biochemistry 27 3917-3925 (1988)). Other laboratories have studies the physical properties of vesicles formed from synthetic cationic amphophiles (Rupert, L. A. M., Hoekstra, D. and Engberts, J. B. F. N. Am. Chem. Soc. 108: 2628-2631 (1985); Carmona-Ribeiro, A. M., Yoshida, L. S. and Chaimovich, H. J. Phys Chem 89 2928-2933 (1985); Rupert, L. A. M., Engberts, J. B. F. N. and Hoekstra, D. J. Amer. Chem. Soc. 108:3920-3925 (1986)).

It is not feasible to extend in vitro transfection technology to in vivo applications directly. In vivo, the diether lipids, such as DOTMA or Lipofectin the current commercial standard, would be expected to accumulate in the body due to the poorly metabolized ether bonds. And finally, it has been reported that the cationic lipid transfection methodology is inhibited by serum; for in vivo applications conditions must be identified which allow transfection to occur in a complex biological milieu such as 100% serum.

Therefore, while the known lipofection technique of transfection described is more efficient and satisfactory than previously known procedures, and permits transient as well as stable transfection and peptide expression, it is not understood what factors regulate the efficiency of the transfection process and how it may be optimized. It would be desirable to determine these factors in order to develop an intracellular delivery system having the advantages of the above-described systems but without their inherent limitations.

Accordingly, it is an object of the invention to provide cationic lipids which carry out both stable and transient transfections of polynucleotides such as DNA and mRNA into cells more effectively.

It is also an object of the invention to provide cationic lipids which deliver other molecules of therapeutic interest, including proteins, peptides and small organic molecules, into cells more effectively.

Further, it is an object of the invention to provide cationic lipids that are not only more effective in accomplishing intracellular delivery but are also metabolizable so as to have reduced in vivo and in vitro toxicity.

It is another object of the invention to provide transfection formulation, comprising novel cationic lipids, that are optimally effective in both in vivo and in vitro transfection.

BRIEF DESCRIPTION OF THE FIGURES

FIG. 1 presents data showing the effect the presence of serum during lipid complex formation on subsequent RNA transfection.

FIG. 2 demonstrates the effect of serum on the effectiveness of RNA transfection.

FIG. 3 demonstrates the effect of cationic lipid concentration on the effectiveness of RNA transfection using DOTAP and DOTMA as cationic lipids.

FIG. 4 demonstrates the effect of neutral lipids on the comparative effectiveness of a series cationic lipids in promoting RNA transfection.

FIG. 5 demonstrates the comparative effectiveness of DPTMA, DOTMA and corresponding derivatives of the Rosenthal Inhibitor in RNA transfection.

FIGS. 6a-6d demonstrate the effect of increasing relative concentrations of lysophosphatidylcholine in lipid formulations on DNA transfection efficiency as demonstrated by expression of gene product in cell culture.

FIGS. 7a-7c demonstrate the comparative DNA transfection activity of various cationic lipid analogs.

FIGS. 8a-8d demonstrate the effect of neutral phospholipids in the transfection lipid formulation on the efficiency of DNA transfection.

FIGS. 9a-9c demonstrate the effect of cholesterol in the transfection lipid formulation on the efficiency of DNA transfection.

SUMMARY OF THE INVENTION

The present invention provides compositions of novel cationic lipids, suitable for use in the intracellular delivery of bioactive agents, comprising polynucleotides, proteins, small organic molecules and drugs, in both in vivo and in vitro applications, and into the cells of plants and animals.

These compositions have the general structure ##STR2## wherein

Y.sup.1 and Y.sup.2 are the same or different and are --O--CH.sub.2 --, --O-- C(O)--, or --O--;

R.sup.1 and R.sup.2 are the same or different and are H, or C.sub.1 to C.sub.23 alkyl or alkenyl; and

R.sup.3, R.sup.4, R.sup.5, R.sup.6, and R.sup.7 are as defined below. Preferred embodiments are compositions wherein R.sup.3 and R.sup.4 are individually C.sub.1 to C.sub.23 alkyl groups, R.sup.5 is --(CH.sub.2).sub.m --, R.sup.6 is absent, R.sup.7 is H, and R.sup.1 and R.sup.2 individually have from 0 to 6 sites of unsaturation, and have the structure

CH.sub.3 --(CH.sub.2).sub.a --(CH.dbd.CH--CH.sub.2).sub.b --(CH.sub.2).sub.c --

wherein the sum of a and c is from 1 to 23; and b is 0 to 6.

Particularly preferred embodiments are compositions wherein the long chain alkyl groups are fatty acids, that is, wherein Y.sup.1 and Y.sup.2 are alike and are --O--C(O)--. These compounds are easily metabolized by cells and therefore lack the toxicity of presently known transfection agents. A specific example of this class of compounds is DL-1,2-dioleoyl-3-dimethylaminopropyl-B-hydroxyethylammonium and its salts.

Other particularly preferred embodiments are those compounds wherein Y.sup.1 and Y.sup.2 are alike and are --O--CH2--. These compounds, having ether-linked alkyl groups, have been found to be superior in transfective properties to presently known cationic lipids. A specific example of a compound of this class is 1,2-O-dioleyl-3-dimethylaminopropyl-.beta.-hydroxyethylammonium and its salts. Useful cationic lipids for intracellular delivery also comprise compounds wherein Y.sup.1 and Y.sup.2 are different and are either --O--CH2-- or --O--C(O)--. These compounds, having alkyl groups attached by both ether and ester linkages, will have combined properties of low toxicity and improved transfective properties. A particularly preferred composition of this class is 1-O-oleyl-2-oleoyl-3-dimethylaminopropyl-.beta.-hydroxyethylammonium and its salts.

Additional novel cationic lipids provided by the invention are adducts of the general structure comprising additional cationic groups attached at the hydroxyl of the .beta.-hydroxyethanolamine moiety. In preferred embodiments of this class of compounds, the additional cationic groups are provided by lysyl groups attached to the hydroxyl group through a diaminocarboxylic acid linker. A glycyl spacer may connect the linker to the hydroxyl group. Particularly preferred compositions of this class are 3,5-(N,N-dilysyl)-diaminobenzoyl-3-(DL-1,2-dioleoyl-dimethylaminopropyl-.b eta.-hydroxyethylamine) and 3,5-(N,N-dilysyl)diaminobenzoylglycyl-3-(DL-1,2-dioleoyl-dimethylaminoprop yl-.beta.-hydroxyethylamine).

Alternatively, the additional cationic groups of the adduct can be provided by attaching cationic amine-containing groups such as, for example, spermine, spermidine, histones, or other molecules known to bind DNA. Preferred embodiments of this class of compositions are L-spermine-5-carboxyl-3-(DL-1,2-dioleoyl-dimethylaminopropyl-.beta.-hydrox yethylamine). These cationic groups can in turn provide further hydrophobic regions to the cationic lipid composition through alkyl quaternizing groups on the attached lysine, spermine, or other amine-containing groups.

Also included within the scope of the invention are analogues of known cationic lipids having ester linkages substituted for ether linkages between alkyl substituents and the glycerol moiety of the structure to provide less toxic, more easily metabolized compositions suitable for use in vivo. These analogues have the general structure ##STR3## or an optical isomer thereof, wherein

Y.sup.1 and Y.sup.2 are different and are either --O--CH2--, --O--C(O)-- or --O--;

R.sup.1 and R.sup.2 are individually C.sub.1 to C.sub.23 alkyl or alkenyl, or H; and

R.sup.3, R.sup.4, R.sup.5 and X are as defined below.

According to yet another aspect of the invention there are provided lipid formulations for transfection comprising a cationic lipid and an effective transfection-promoting amount of a lysophosphatide, having the structure ##STR4## wherein Y is selected from the group consisting of --O--CH2-- and --O--C(O)--;

R is C.sub.10 to C.sub.23 alkyl or alkenyl; and

Z is a headgroup.

Preferred formulations for transfection of polynucleotides and peptides into cells comprise novel cationic compounds of the invention having the structure set forth herein, together with an effective transfection-promoting amount of a lysophosphatide. The lysophosphatide may have a neutral or a negative headgroup. Lysophosphatidylcholine and lysophosphatidylethanolamine are preferred, and 1-oleoyl lysophosphatidylcholine is particularly preferred. Lysophosphatide lipids are advantageously present in the formulation in a molar ratio of 0.5 lyso lipid to cationic lipid.

Lyso forms of cationic lipids, selected from the novel cationic lipids of the invention, DOTMA, or DOTAP can also be used to increase the effectiveness of the transfection. These lyso forms are advantageously present in effective amounts up to about one-third of the total cationic lipid in the formulations.

According to another aspect of the invention, there is provided a liposomal formulation, comprising a cationic lipid of the invention, wherein the cationic lipid is in the form of vesicles in an aqueous media. The lipids of the liposomal formulation can further comprise a neutral lipid species selected from the group consisting of phosphatidylcholine, phosphatidylethanolamine, sphingomyelin, or cholesterol. A preferred molar ratio of cationic to neutral lipid species in these formulations is from about 9/1 to 1/9; a molar ratio of about 5/5 is particularly preferred. The liposomal formulation can further comprise a lyso lipid selected from the group consisting of lysophosphatidylcholine, lysophosphatidylethanolamine, or a lyso form of a cationic lipid species.

According to yet another aspect of the invention, there are provided pharmaceutical products comprising the cationic lipids of the invention having any of the structures disclosed herein together with a pharmacologically effective amount of a therapeutic agent. Cationic lipids present in these compositions facilitate the intracellular delivery of the active therapeutic agent. Products are provided for topical, enteral and parenteral uses. In one pharmaceutical product the therapeutic agent is a steroid; in another, the therapeutic agent is a non-steroidal anti-inflammatory agent.

In other pharmaceutical products of the invention, the therapeutic agent is an antiviral nucleoside analogue or preferably a lipid derivative of an antiviral nucleoside analogue, which is a phosphatidyl derivative, or a diphosphate diglyceride derivative. The antiviral nucleoside can be a dideoxynucleoside, a didehydronucleoside, a halogenated or azido- derivative of a nucleoside, or an acyclic nucleoside. In preferred embodiments, the lipid derivatives of antiviral nucleosides are (3'-azido-3'-deoxy)thymidine-5'-diphospho-3-diacylglycerol (AZT diphosphate diglyceride) and dideoxythymidine diphosphate diglyceride. In particularly preferred embodiments, the lipid derivative of an antiviral nucleoside is an acyclovir or gancyclovir diphosphate diglyceride or diphosphate diglyceride derivatives of 1-(2-deoxy-2'-fluoro-1-.beta.-D-arabinofuranosyl)-5-iodocytosine (FIAC) or 1(2'-deoxy-2'-fluoro-1-.beta.-D-arabinofuranosyl)-5-iodouracil (FIAU).

In other pharmaceutical products of the invention the therapeutic agent is a polynucleotide. In one of these embodiments, the therapeutic polynucleotide is a ribozyme, or an antisense RNA or DNA. In preferred embodiments, the formulation comprises an antisense DNA or RNA or a ribozyme directed against HIV. In a particularly preferred embodiment, the therapeutic polynucleotide is an antisense DNA or RNA or a ribozyme directed against the rev transactivator of HIV. An example of such an agent is the 28-mer phosphorothioate antisense polynucleotide. Alternatively, the therapeutic polynucleotide can be one coding for an immunogen, a natural hormone, or a synthetic analogue of a natural hormone; or it can be a polynucleotide sequence coding for a gene product that is deficient or absent in a disease state, and administration of said product to a human in need of therapy relating to said gene product has a therapeutic effect.

The pharmaceutical products disclosed may also comprise a therapeutic protein or polypeptide corresponding to those coded for by the therapeutic polynucleotides described above.

In a preferred embodiment, the invention provides pharmaceutical preparations for topical use comprising a novel cationic lipid of the invention, having any of the structures disclosed herein together with a pharmacologically effective amount of a therapeutic agent in a pharmaceutically acceptable vehicle. Preferred therapeutic agents are steroids, non-steroidal anti-inflammatory agents, antiviral nucleosides or phospholipid derivatives of these antiviral nucleosides, a therapeutic polynucleotide which is a ribozyme or an antisense RNA or DNA sequence, a polynucleotide coding for a therapeutic protein or polypeptide, or a therapeutic protein or polypeptide itself. The therapeutic protein or polypeptide may be, for example, one that is absent or deficient in a genetic disease, an immunogen, a natural hormone, or a synthetic analogue of a natural hormone.

Included among the particularly preferred embodiments according to this aspect of the invention are topical formulations for the treatment cf herpes simplex, comprising a cationic lipid of the invention together with a pharmacologically effective concentration of acyclovir, gancyclovir,1-(2-deoxy-2'-fluoro-1-.beta.-D-arabinofuranosyl)-5-iodocytosi ne (FIAC) or 1(2'-deoxy-2'-fluoro-1-.beta.-D-arabinofuranosyl)5-iodouracil (FIAU) in an pharmaceutically acceptable vehicle. In preferred embodiments, the preparation comprises phosphoglyceride derivatives of acyclovir, gancyclovir, FIAC or FIAU.

According to another aspect of the invention, there is provided a method for introducing a biologically active agent into a cell, either plant or animal, comprising the steps of preparing lipid vesicles comprising a cationic lipid of the invention, and using these lipid vesicles to facilitate the transfection or transport of bioactive agents into the cells. The intracellular transport may be accomplished by incorporating or encapsulating the bioactive agent in the lipid vesicle and contacting the cell with the lipid vesicles, as in conventional liposome methodology; or alternatively, by contacting the cells simultaneously with empty lipid vesicles, comprising the cationic lipids together with the bioactive agent, according to conventional transfection methodology. In the process of either strategy, the bioactive agent is taken up by the cell. In preferred embodiments of the method, the bioactive agent is a protein, polynucleotide, antiviral nucleoside or a drug. In particularly preferred embodiments, the bioactive agent is an antisense RNA or DNA sequence or a ribozyme. According to one embodiment of the method, the contacting step occurs in vitro. The method may be applied in the treatment of disease in a vertebrate, comprising the step of administering a pharmaceutical preparation comprising any one of the cationic lipids having the structure set forth above together with a pharmacologically effective amount of a therapeutic agent specific for the treatment of the disease to the vertebrate and permitting the therapeutic agent to be incorporated into a cell, whereby the disease is effectively treated. The bioactive agent is delivered to the cells of the animal in vivo or in vitro. The in vitro delivery of a bioactive agent may be carried out on cells that have been removed from an animal. The cells are returned to the animal body whereby the animal is treated.

The methods according to other embodiments of the invention include the topical application of a preparation to the skin; the injection of a preparation into a body cavity or into the tissues of said vertebrate; or the oral administration of said preparation. The biologically active agent can be a polynucleotide, such as, for example, DNA or mRNA coding for a polypeptide, and said polypeptide is expressed after said DNA or said mRNA is taken up into said cell. In yet other embodiments, the biologically active agent is a drug.

The cationic lipids of the invention provide more effective intracellular delivery than the use of presently available agents for the purpose. Further these lipids include species that are less toxic to cells when used in in vivo and in vitro procedures.

These and other advantages and features of the present invention will become more fully apparent from the following description and appended claims.

DETAILED DESCRIPTION OF THE INVENTION

The cationic lipids (CLs) of the invention, comprising compositions having an ammonium group together with hydrophobic alkyl groups, as well as adducts of these cationic lipids, are advantageously used in formulations to prepare lipid vesicles or liposomes to be used in transfection procedures, or to similarly facilitate the intracellular delivery of proteins, polypeptides, small organic molecules, and drugs of therapeutic interest. The adducts further comprise additional cationic and hydrophobic groups that enhance the effectiveness of the lipids in interacting with cell membranes.

We have discovered that certain derivatives and adducts of a compound having the structure ##STR5## wherein R is a long chain fatty acid, are highly effective compounds for use in lipid formulations for transfection and other intracellular delivery procedures. A single species of a compound of this type, comprising C.sub.18 (stearoyl) fatty acids was described by Rosenthal, A. F. and R. P. Geyer, J. Biol. Chem. 235(8):2202-2206 (1960). The Rosenthal compound, which is an inhibitor of phospholipase A (Rosenthal Inhibitor, RI), is itself ineffective as a promoter of transfection or intracellular delivery. Modifications to the RI molecule that we have discovered to be most effective in conferring transfective properties are substitution of preferred long chain aliphatic groups, selection of preferred acyl (ester) or alkyl (ether) links between the glycerol moiety of RI and the aliphatic groups, and the addition of groups to the hydroxyl moiety which promote interaction with cell membranes. These compounds have proved to be superior in transfective performance to any presently known, including the cationic lipids described in European Application No. 0 187 702 (1986).

Nomenclature

To simplify description, compounds are referred to herein by acronyms, as follows: RI: The Rosenthal Inhibitor

______________________________________ RI: The Rosenthal Inhibitor DORI: Dioleoyl derivatives of RI having two C.sub.18 un- saturated (18:1) aliphatic groups, comprising: DORI diester: DL-1,2-dioleoyl-3-dimethyl- aminopropyl-.beta.- hydroxyethylammonium DORIE diether: DL-1,2-O-dioleyl-3-di- methylaminopropyl-.beta.- hydroxyethylammonium DORI ester/ether: DL-1-O-oleyl-2-oleoyl-3- dimethylaminopropyl-.beta.- hydroxyethylammonium OR DL-1-oleoyl-2-O-oleyl-3- dimethylaminopropyl-.beta.- hydroxyethylammonium DPRI: Derivatives of RI having C.sub.16 (16:0) aliphatic groups, comprising: DPRI diester: DL 1,2-dipalmitoyl-3-di- methylaminopropyl-.beta.- hydroxyethylammonium DPRI diether: DL 1,2-O-dipalmityl-3-di- methylaminopropyl-.beta.- hydroxyethylammonium DOTMA: N-[1-(2,3-dioleyloxy)propyl]-N,N,N-trimethyl- ammonium DOTAP: DL-1,2-dioleoyl-3-propyl-N,N,N-trimethyl- ammonium DPTMA: DL-(2,3-dipalmityl)-3-propyl-N,N,N- trimethylammonium DLYS-DABA-DORI diesters, diethers, or ester/ethers: Lysine- containing adducts of DORI, having lysine groups attached at the hydroxyl group of the .beta.-hydroxyethyl moiety through a diaminobenzoic acid linker, which is optionally joined to DORI through a glycyl spacer. DLYS-DABA-DPRI diesters, diethers, or ester/ethers: analogues of above DORI compounds, but comprising DPRI. SPC-DORI diesters, diethers, or ester/ethers: Spermine- containing adducts of DORI, having spermine attached at the hydroxyl group of the .beta.-hydroxyethyl moiety. SPC-DPRI diesters, diethers, or ester/ethers: analogues of DORI compounds above, but comprising DPRI. SPC-DABA-DORI diesters, diethers, or ester/ethers: Spermine- containing adducts of DORI, having spermine groups attached at the hydroxyl group of the .beta.-hydroxyethyl moiety through a diaminobenzoic acid linker, which is optionally joined to DORI through a glycyl spacer. ______________________________________

Cationic lipids according to one aspect of the invention have the general formula ##STR6## wherein

Y.sup.1 and Y.sup.2 are the same or different and are --O--CH.sub.2 -- --, --O--C(O)--, or --O--;

R.sup.1 and R.sup.2 are the same or different and are H, or C.sub.1 to C.sub.23 alkyl or alkenyl;

R.sup.3 and R.sup.4 are the same or different and are C.sub.1 to C.sub.24 alkyl, or H;

R.sup.5 is C.sub.1 to C.sub.24 alkyl straight chain or branched chain;

R.sup.6 is --C(O)--(CH.sub.2).sub.m --NH--, a diaminocarboxylic acid which is alkyl, aryl, or aralkyl, or --C(O)--(CH.sub.2).sub.m --NH-- linked to said diaminocarboxylic acid, or is absent;

R.sup.7 is H, spermine, spermidine, a histone, or a protein with DNA-binding specificity, or wherein the amines of the R.sup.7 moiety are quaternized with R.sup.3, R.sup.4, or R.sup.5 groups; or

R.sup.7 is an L- or D-alpha amino acids having a positively charged group on the side chain, such amino acids comprising arginine, histidine, lysine or ornithine or analogues thereof, or the same amino acids wherein the amine of the R.sup.7 moiety is quaternized with R.sup.3, R.sup.4 or R.sup.5 groups; or

R.sup.7 is a polypeptide selected from the group comprising L- or D-alpha amino acids, wherein at least one of the amino acids residues comprises arginine, histidine, lysine, ornithine, or analogues thereof;

n is 1 to 8;

m is 1 to 18; and

X is a non-toxic anion.

We have determined structure-transfection activity relationships within classes cf cationic lipids having a quaternary ammonium group and have found these relationships to be useful in predicting efficient transfection. We accordingly provide synthetic cationic lipids of this class suitable for use in transfection formulations. CLs having long chain aliphatic (R.sup.1 and R.sup.2) groups comprising ether linkages are preferred to those having ester linkages; CLs having unsaturated R.sup.1 and R.sup.2 groups are preferred to CLs having corresponding saturated groups; and CLs such as analogues of RI, having polar hydroxyethyl group substituents on the quaternary ammonium group are more effective than those substituted with alkyl groups, for example, the methyl substituent of DOTMA.

Therefore, in particularly preferred embodiments, the cationic lipids of the invention are derivatives of RI having a structure comprising at least one alkyl ether group. A specific memeber of this class of cationic lipids is a DORI diether (DORIE) having long chain alkyl groups with one site of unsaturation, and having the structure: ##STR7##

For applications demanding metabolizable, less toxic compounds, CLs having long chain R.sup.1 and R.sup.2 aliphatic groups attached by acyl bonds are preferred. Therefore, in other preferred embodiments, the cationic lipids of the invention comprise derivatives of RI having the structural characteristics of Formula I, but comprising at least one acyl group, as, for example, a DORI diester having the structure: ##STR8##

In yet other preferred embodiments, cationic lipids of the invention are substituted at the hydroxyl group of an ethanolamine moiety with various species which act to enhance binding to cell membranes. In preferred embodiments the amine group of ethanolamine is quaternized.

Preferred species for this purpose are compounds such as spermines and spermidines, or other compounds having multiple amino groups, or histones, or similar proteins rich in basic amino acids such as arginine and histidine. Cationic substances such as the histones, spermines, and spermidines are known to bind and modulate negatively charged cell membrane surfaces. For example, lipid-derivatized spermine-like structures are reported to efficiently modulate gene transfer into mammalian endocrine cells (Behr, J.-P. et al. Proc. Natl. Acad. Sci. USA 86:6982-6986 (1989). We have designed a series of molecules which combine advantageous properties of both cationic lipids and cationic structures derived from amino acids and spermines. These molecules are prepared by coupling spermine, through a carboxylic acid group, to the hydroxyl moiety of the ethanolamine group of a lipid such as DORI, DORIE or DPRI.

One such series of compounds, is represented by L-spermine-5-carboxyl-3-(DL-1,2-dipalmitoyldimethylaminopropyl-.beta.-hydr oxylamine, designated SPC-DPRI-diester, which has the structure ##STR9##

In an example of another lipid of this type, the basic amino acid lysine is linked to the same hydroxyl moiety of the lipid through a linker molecule. The linker molecule can be any diaminocarboxylic acid, either alkyl, aryl or aralkyl, having two amino sites by which lysine is anchored as a pendant in a branched molecule that can bind to multiple binding sites simultaneously. In preferred embodiments, the linker molecule is joined to the hydroxyl group of the hydroxy lipid through a spacer arm which can be any alkyl amino acid. Glycine is a preferred spacer arm. A representative cationic lipid of this type comprises lysine linked to the hydroxyl moiety of DPRI through diaminobenzoic acid and a glycine spacer, to form 3,5-(N,N-di-lysyl)-diaminobenzoyl-glycyl-3-(DL-1,2-dipalmitoyldimethylamin opropyl-.beta.-hydroxyethylamine). This lipid, designated DLYS-DABA-GLY-DPRI-diester, has the structure ##STR10## Particularly preferred compounds of this class are DLYS-DABA-GLY-DORI-diester, having the structure ##STR11## and DLYS-DABA-GLY-DORI-diether, having the structure ##STR12## Other molecules of this type can comprise linkers or spacer arms to which are joined other basic amino acids, such as histidine and arginine or analogues or derivatives or these basic amino acids comprising related molecules, which are structurally modified, for example by having substituent groups, such as 1-methyl histidine or 3-methyl histidine. Polymers of these amino acids or their analogues can be attached to the linker in the same manner. Amine-containing groups added to the cationic lipids of the invention through spacers and linkers at the .beta.-hydroxyethylammonium moiety can in turn provide further hydrophobic regions to the lipid structure by quaternization of the amine with the alkyl, alkenyl, aryl and aralkyl groups of R.sup.3, R.sup.4, and R.sup.5. Thus, the assembled lipid adducts, comprising additional cationic groups, and in some cases, additional hydrophobic groups as well, incorporate additional sites capable of interaction with the cell membrane, thereby increasing the intracellular delivery potency of the cationic lipid.

For some applications it is important that cationic lipids used are metabolizable and therefore non-toxic, both for in vitro applications and especially when used in vivo, and yet retain the substantial transfective properties associated with lipid species having an ether-linked alkyl group. Accordingly, we have synthesized cationic lipids according to another aspect of the invention having the formula ##STR13## or an optical isomer thereof, wherein

Y.sup.1 and Y.sup.2 are different and are either --O--CH2--, --O--C(O)--, or OH;

R.sup.1 and R.sup.2 are individually absent or are C.sub.1 to C.sub.23, alkyl or alkenyl;

R.sup.3, R.sup.4 and R.sup.5 are the same or different and are H, C.sub.1 to C.sub.14 alkyl, C.sub.7 to C.sub.11 aryl or aralkyl, or at least two of R.sup.3, R.sup.4, and R.sup.5 are taken together to form quinuclidino, piperidino, pyrrolidino, or morpholino;

n is 1 to 22; and

X is a non-toxic anion.

According to one aspect of the invention, the CLs are combined with other lipids in formulations for the preparation of lipid vesicles or liposomes for use in intracellular delivery systems. The formulations preferably are prepared from a mixture of positively charged lipids, negatively charged lipids, neutral lipids and cholesterol or a similar sterol. The positively charged lipid can be one of the cationic lipids of the invention alone, a mixture of these, or one of the cationic lipids of the invention in combination with the cationic lipids DOTMA, DOTAP, or analogues thereof. Neutral and negatively charged lipids can be any of the natural or synthetic phospholipids or mono--, di-, or triacylglycerols. The natural phospholipids are typically those from animal and plant sources, such as phosphatidylcholine, phosphatidylethanolamine, sphingomyelin, phosphatidylserine, or phosphatidylinositol. Sy