WO1997029118A1

WO1997029118A1 - Lipopolyamine compound, pharmaceutical composition and nucleic acid transfer vector containing same, and preparation method

Info

Publication number: WO1997029118A1
Application number: PCT/FR1997/000225
Authority: WO
Inventors: Rainer Bischoff; Yves Cordier
Original assignee: Transgene S.A.
Priority date: 1996-02-05
Filing date: 1997-02-05
Publication date: 1997-08-14
Also published as: AU1728197A; FR2744453A1; FR2744453B1

Abstract

A compound of formula (II), wherein R1 is hydrogen and R2 is hydrogen or, for only one of its n substituents, a deoxy residue in the 3 position of the cholesterol or a mono- or polyvalent residue of a cholesterol derivative having one or more reactive groupings, respectively, said groupings being capable of reacting with a secondary amine group of a polyamine molecule; n is an integer from 2 to 6; and p is an integer from 2 to 6 and identical or different in the n (a) fragments.

Description

LIPOPOLYAMINE COMPOUND, PHARMACEUTICAL COMPOSITION AND NUCLEIC ACID TRANSFER VECTOR INCLUDING, AND

PREPARATION PROCESS

The present invention relates to a lipopolyamine compound ci to pharmaceutical compositions comprising it. More particularly the present invention relates to the use of this composition for the transfer of a nucleic acid into a host cell. Finally, the present invention relates to a process for the preparation of a compound according to the invention, and to a process for the purification of an isomer of said compound.

The transfer of genes into a given cell is the very basis of gene therapy. However, one of the major problems is to introduce a sufficient quantity of therapeutic nucleic acid into the host cell to be treated. The most effective vectors in this regard are viral, in particular retroviral and adenoviral vectors. In fact, viruses have developed sophisticated mechanisms to cross cell membranes, escape degradation at the level of lysosomes and make their genome enter the nucleus. However, the viral approach has its limitations, in particular a cloning capacity in the viral genome restricted, a production of infectious viral particles capable of dissemination in the host organism and the environment, a risk of insertional mutagenesis in the case of vectors retroviral, and induction of immune and inflammatory responses in the host. These significant drawbacks in the context of human use justify the search for alternative nucleic acid transfer systems.

Many methods are currently available, including calcium phosphate co-precipitation, electroporation, microinjection and even lipofection. The latter uses cationic lipids which interact spontaneously with nucleic acid by interaction of charges to form positively charged complexes capable of fusing with anionic cell membranes and causing the nucleic acid which they transport to penetrate. The use of cationic lipids as a transfer vector for various molecules of therapeutic interest has already been described in the literature. It has been shown that cationic lipids with various chemical structures allow the transfection of cultured mammalian cells with very different efficiencies depending on the type of cell and the formulation of lipid-DNA complexes. However, the efficiencies are generally variable and low in animals, in particular as regards muscular administration, and the results obtained with the synthetic molecules currently available, are not satisfactory. Taken as a whole, these studies did not make it possible to identify factors influencing the efficiency of the gene transfer mechanism and optimization trials as a function of structural parameters such as particle size, charge, DNA etc ... did not provide clear answers.

EP 0 349 1 1 1 and WO95 / 18863 (Behr et al.) Relate to the lipid composition comprising a 5-carboxyspermylglycine dioctadecylamide compound (also designated DOGS) which results from the combination of spermine with a straight chain lipid. WO95 / 1 8863 describes successful in vivo transfection experiments with a +/- charge ratio of less than 2, only in mice born by intracerebral injections. No transfection by intramuscular or intravenous administration is obtained, in particular on mice other than newborns.

The article by Gao and Huang, 1991 in BBRC 179, 280-285 and the international application WO93 / 05162 describe a lipopolyamine, DC-CHOL resulting from the reaction of cholesteryl chloroformate and a tertiary amine (dimethylethylene diamine). DC-CHOL or 3 R ([N - (N \ N '- dimethylaminoethane) - carbamoyl] cholesterol has only a positive charge and therefore has a low transfecting power.

The aim of the present invention is to provide new, more efficient cationic lipids, in particular for application as a DNA transfer vector, in particular for intramuscular, intratracheal or even intrapulmonary administration. The present invention therefore relates to a lipopolyamine compound of formula;

C - (S) _m (I) in which:

C represents a deoxy residue in position 3 of cholesterol or a monovalent or polyvalent residue of a cholesterol derivative carrying one or more reactive groups capable of reacting with a primary or secondary amine group of a polyamine molecule,

S represents an amine residue of said amine group of a polyamine molecule, and

- m represents an integer from 1 to 3.

The term “cholesterol derivative” is intended to mean a compound having the formula the cholesterol backbone linked to one or more substituent (s), at least one of which comprises a reactive group capable of reacting with a primary or secondary amine group of a polyamine molecule.

By way of illustration, we cite more particularly as:

- compounds where m ≈ 1, the compounds in which C represents a monovalent deoxy residue in position 3 of cholesterol;

- compounds where m = 2, the compounds in which C represents a divalent didesoxy residue of 5-cholesten-3β, 4β-diol, or 5-cholesten- 3β, 25-diol, and,

- compounds where m = 3, the compounds in which C represents a trivalent tridesoxy- residue of cholesten - 3β, 5f _\ 6β-triol.

In a preferred embodiment S is a residue of a polyamine molecule of formula:

H2N - [- (CH ₂ ) p - NH] _n - H (IX) and the lipopolyamine compound according to the invention corresponds to the formula

HN [(CH ₂ ) _F Nh _n H (n: IR i R2

in which :

- Ri and R ₂ represent hydrogen or a monovalent residue C;

- the symbol R ₂ can be identical or different in the n fragments [(CH ₂ ) _P - N];

R2

- only one of R _! and n substituents R ₂ represents undit monovalent residue C;

- n is an integer from 2 to 6; and

- p is an integer from 2 to 6 identical or different in the n fragments [(CH ₂ ) _P - N]

R2

Preferably in formula (II), Rj represents hydrogen and R ₂ represents hydrogen or, for only one of the n substituents, a deoxy residue in position 3 of cholesterol or a monovalent or polyvalent residue of a cholesterol derivative carrying one or respectively more reactive groups capable of reacting with a secondary amine group of a polyamine molecule;

- n is an integer from 2 to 6; and

- p is an integer from 2 to 6 identical or different in the n

[(CH2) _P - N] ^r 1 Preferably the compounds n = 2 or 3 and p = 3 or 4.

More preferably, S represents a residue of natural spermine or spermidine compound corresponding to the compounds of formula (IX) in which n = 2 and p = 3 and 4 (spermidine) and n = 3 and p = 3, 4 and 3 respectively in the different links (spermine).

In a particular embodiment, the compound according to the invention is characterized in that the monovalent residue C, is a monovalent deoxy residue in position 3 of cholesterol.

Amine S can then be coupled, for example, by nucleophilic substitution of a tosylate group in position 3 of a tosylate cholesterol compound with an amine function of S, thus creating a direct amine bond. This has the advantage of preserving the number of positive charges of spermidine (3 charges) and spermine (4 charges).

However, the coupling of the amine S can also be done by coupling of a reactive group of C with an amine function of the compound of formula (IX). This is the case where C represents a residue of a cholesterol derivative. More precisely, the residue C can represent a residue of a cholesterol derivative in position 3 of formula L-R in which:

L represents an aliphatic radical comprising a residue of a group capable of reacting with a primary or secondary amine group and

- R represents the remainder in position 3 of formula

Mention is made in particular of the compounds in which the remainder of the cholesterol derivative in position 3 has the formula:

This residue (V) is a residue of the chloroformate derivative in position 3 of cholesterol.

Mention may also be made of the compounds of formula (III) in which L represents a hemisuccinate type bond of formula

- C - (CH ₂ ) ₂ - CI l II

OO This residue is then a residue of an OH - C - derivative (CH ₂ h - C - R

U I I

O O which can be obtained by grafting an anhydrous succinite radical on cholesterol. The reaction is accompanied by an opening of the succinite ring which can then react with the polyamine. This embodiment introduces an arm between the latter and the cholesterol.

The present invention also provides monosubstituted isomers of purified cholesterol corresponding to the formula:

H - N - (CH ₂ ) ₃ - N - (CH ₂ ) 4 - ^• NH (VI)

_||

Ri R2 ¹ R ₂ 2

in which Ri and R ₂ ', with i = 1 and 2 have the meanings of Ri and R ₂ given above and of compounds corresponding to the formula:

H - N - (CH ₂ ) ₃ - N - (CH ₂ ) ₄ - N - (CH ₂ ) ₃ - N - H (VII) 1 1 II

Ri R2 ¹ R2 ² R2 ³

in which R i and R ₂ 'with i = 1, 2 and 3 have the meanings of R _t and R ₂ given above.

Mention is made more particularly of the compounds of formula (VI) for which: - R ₂ ^l represents undit remains monovalent of cholesterol or of a cholesterol derivative in position 3 and,

- Ri and R ₂ 2 represent H. as well as the compounds for which:

- Ri and R ₂ ¹ represent H and - R? ² represents a residue of cholesterol or of a cholesterol derivative in position 3 and, the compounds for which:

R i represents undit monovalent cholesterol residue of a cholesterol derivative in position 3, and - R ₂ ¹ and R ₂ ² represent H.

Likewise, a compound according to the invention is more particularly cited, characterized in that in the formula (VU) - R ₂ t and R ₂ ² represent H, and

- one of Ri and R ₂ ³ represents said monovalent residue of a cholesterol derivative in position 3 and the other represents H. as well as the compounds according to the invention, characterized in that in formula (VII) - Ri and R ₂ ³ represent H, and

- one of R ₂ ¹ and R ₂ ² represents said residue of cholesterol or of a cholesterol derivative in position 3 and the other represents H. These compounds in which the cholesterol or cholesterol derivative is substituted on a terminal amine compounds (VI) or (VII) are preferred.

The experimental data described below show that the transfecting power of these isomers is greater when the remainder of a cholesterol derivative in position 3 is grafted onto a primary terminal amine of the compound of formula (IX), in particular on the long arm [(CH ₂ ) .ι] spermidine.

The compounds according to the invention are advantageously used for their transfecting power for the transfer into a cell of a therapeutically active substance negatively charged.

The present invention therefore also relates to a pharmaceutical composition comprising a compound according to the invention and a therapeutically active substance negatively charged.

Mention is made in particular of a composition in which said active substance is an acid. Preferably, said active substance is a nucleic acid.

Said nucleic acid can be a sense or anis sense DNA or RNA. The sense and anti sense applications are well known to those skilled in the art. In short, by:

- "anti sense" a nucleic acid having a sequence complementary to a target sequence, for example an mRNA sequence whose expression is sought to block by hybridization on the target sequence; and - "sense" a nucleic acid having a sequence homologous or identical to a target sequence, for example a sequence which binds to a protein transcription factor and is involved in the expression of a given gene.

According to a preferred embodiment, the nucleic acid comprises a gene of interest and elements allowing the expression of said gene of interest. In this embodiment, said nucleic acid fragment is advantageously in the form of a plasmid.

In the context of the present invention, the nucleic acid can be homologous or heterologous to the target cell. It may be advantageous to use a nucleic acid encoding a cytokine (interleukin including IL-2, interferon, colony stimulating factor, etc.), a cellular or nuclear receptor, a ligand, a coagulation factor (Factor VII , Factor VIII, Factor IX ...), the CFTR protein (Cystic Fibrosis Transmembrane Conductance Regulator in English), insulin, dystrophin, a growth hormone, an enzyme (renin urease, thrombin ...), an inhibitor enzyme (inhibitor of a viral protease, α 1-antitrypsin, etc.), a polypeptide with an anti-tumor effect (product of tumor suppressor genes, polypeptide stimulating the immune system, etc.), a polypeptide capable of inhibiting or slow the development of a bacterial, viral or parasitic infection (antigenic polypeptide, trans-dominant variant, etc.), an antibody, a toxin, an immunotoxin and ! 0 finally a marker (luciferase, β-galactosidase, product conferring resistance to an antibiotic ...). Of course, this list is not exhaustive and other genes can also be used.

Advantageously, the nucleic acid is placed under the control of the elements necessary for its expression in the host cell. By "necessary elements" is meant all of the elements allowing the transcription of said DNA fragment into RNA (antisense RNA or mRNA) and the translation of mRNA into polypeptide. These elements include a regulatable or constitutive promoter, which can be hierologous or on the contrary homologous to the parental virus. Examples include the promoter of the human or murine PGK (Phospho Glycerate Kinase) gene, the early promoter of the SV40 virus (Simian Virus), the RSV LTR (Rous Sarcoma Virus), the TK promoter (Thymidine Kinase) of the HSV-1 virus (Herpes Simplex virus) and the adenoviral promoters El A and MLP. The necessary elements may, in addition, include additional elements (intronic sequence, secretion signal sequence, nuclear localization sequence, site of initiation of translation, poly A signal of termination of transcription, etc.).

The transfection experiments show that advantageously the weight ratio of the lipid compound according to the invention to said DNA fragment is from 0.01 to 100. The optimal ratio is from 0.1 to 10.

A subject of the present invention is also the use of a pharmaceutical composition according to the invention for preparing a vector for transferring said nucleic acid into cells in vitro, ex vivo or in vivo.

In general, a composition according to the invention can be used in various types of host cells. It is preferably a mammalian cell and in particular a human cell. Said cell may be a primary or tumor cell of an origin hematopoietic (totipotent stem cell, leukocyte, lymphocyte, monocyte or macrophage ...) hepatic, epithclial, fibroblast ... and especially a muscle cell (myoblast, cardiomyocyte ...), a tracheal or pulmonary cell. Furthermore, a composition according to the invention can comprise a targeting element for a particular cell, for example a ligand to a cellular receptor. Such targeting elements are known.

A composition according to the invention can be administered by intragastric, subcutaneous, intravenous, intraperitoneal, intratumoral or nasal route. Particular preference will be given to the intracardiac, intramuscular, intrapulmonary or intratracheal routes of administration. The administration can take place in single dose or repeated one or more times after a certain interval of interval. The appropriate route and dosage will vary depending on various parameters, for example, the individual or disease to be treated or the nucleic acid to be transferred.

A composition according to the invention is intended for the preparation of a medicament intended for the treatment of the human or animal body and, preferably, by gene therapy. According to a first possibility, the medicament can be administered directly in vivo (for example into a muscle, into the lungs by aerosol, etc.). It is also possible to adopt the ex vivo approach which consists in taking cells from the patient (stem cells from the bone marrow, lymphocites of peripheral blood, muscle cells, etc.), transferring them in vitro according to the techniques of the art and to re-administer them to the patient.

The present invention therefore also relates to a vector for transferring nucleic acids into cells of microorganisms, plants, humans or animals, in particular mammals, characterized in that it comprises a pharmaceutical composition according to the invention comprising a nucleic acid and an adjuvant capable of improving the transfecting power of said composition. In one embodiment, said adjuvant is a neutral co-lipid.

By way of examples, mention may be made of phospholipids such as phosphatidyl ethanolamine, phosphatidyl choline, phosphatidyl serine, phosphatidyl glycerol as well as pharmaceutically acceptable detergents.

It is also possible to combine a third substance with the lipid / colipid mixture to further improve the transfection efficiency or the stability of the complexes.

The molar ratio of the lipopolyamine compound according to the invention to the adjuvant neutral co-lipid is preferably from 0.1 to 5, more preferably from 0.25 to 2.

The present invention also relates to a process for the preparation of a lipopolyamine compound according to the invention, characterized in that the coupling of a chloroformate compound of formula:

O

I I

Cl - C \

O - R (VII I)

in which R represents a deoxy residue in position 3 of cholesterol, with an amine group of the compound of formula:

H ₂ N-KCH ₂ ) p - NHtnH (IX)

in which p and n have the meanings given above.

To promote the production of mono-substituted cholesterol derivatives, the conditions of the process had to be adapted. The reaction being carried out in the presence of an excess of amine (IX) from 10 to 20 times, at a temperature of - 20 ° C, and in an inert gas atmosphere. In addition, the mono-substituted compound of formula (II) is advantageously obtained by extraction in a water / dichloromethane mixture followed by two stages of crystallization from methanol of the dichloromethane fraction. Indeed, we had to select a crystallization solvent (methanol) suitable for the hydrophobicity of the molecules according to the invention, capable of first crystallizing the most hydrophobic molecules (polysubstituted derivatives), then monosubstituted (compounds of formula (II)).

To obtain an isomer of a lipopolyamine compound according to the invention, a compound of formula (X) is prepared which corresponds to a compound of formula (IX) of which at least one amine function is not protected, but whose other functions amines are protected. Then the coupling of compound (X) is carried out with a chloroformate compound of formula (VIII). A compound is thus obtained in which the free amino functions are protected and said amino functions are deprotected to obtain a compound of formula (II).

Preferably, to prepare the compound of formula (VI) or (VII) in which p = 3 and 4 and n = 2 or 3, coupling is carried out of one mole of 1, 4 diaminobutane and 1 or 2 moles respectively. acrylonitrile, then the amine groups are protected, then the nitrile group (s) is reduced to amine group (s). A compound of formula (VI) or respectively (VII) is thus obtained in which the residue C is substituted on a terminal primary amine of the compound of formula (IX).

Finally, the present invention also relates to a process for the purification of an isomer of a compound according to the invention, from a mixture of isomers of said compounds, characterized in that a high performance liquid chromatography is carried out in reverse phase with a supelcosil LC-ABZ column. The method for preparing isomers in substantially pure form by reverse phase HPLC has proved to be particularly difficult to implement because of their hydrophobic but nevertheless polar characteristic. In this regard, the choice of the column proved to be decisive: it was necessary to use a column based on silica strongly derived in Cl 8 groups and deactivated at the surface, such as the supelcosil® LC-ABZ provided for the purification of compounds. basic. The gradient conditions applied are also important for optimizing the purification of isomers whose hydrophobicity is very close.

The results of transfection of cell lines in culture show a capacity for transfection of cells of tissue origin and of different species. In a first series of experiments carried out in parallel with DC Chol (3 β [N, N '- dimethylaminomethane) carbamoyie] cholesterol), which differs from the compounds of the invention by the presence of N, N' - dimethylethylenediamine ( 1 positive charge only) instead of spermine or spermidine (2 or 3 positive charges), there is a significant increase in gene transfer in muscle cells which are usually refractory to transfection. Consequently, the cationic lipids of the invention are very particularly suitable for the treatment of muscular diseases, in particular myopathies of Duchenne and Bccker. In addition, they are also effective in transfecting lung cells and are suitable for the treatment of lung diseases, including cystic fibrosis, lung cancer and asthma.

Another advantage of the compounds according to the invention, such as spermine-cholesterol and spermidine-cholesterol, is that they result from the association of molecules naturally present in the human body. Cholesterol constitutes cell membranes and spermine and spermidine are molecules ubiquitously present in cells. This characteristic, which is not shared by the cationic lipids of the prior art, is an additional advantage in the context of use. human and limits the risks of immunogenicity and toxicity in the patient. In this regard, the compounds of the invention are in fact capable of being metabolized by enzymes into natural products therefore not very toxic. Consequently, they present security guarantees which are essential for human use.

Other advantages and characteristics of the present invention will become apparent in the light of the detailed description which follows.

The following experimental data are illustrated by reference to the following Figures:

Figure 1 illustrates the analytical HPLC profile obtained with the spermidine-cholesterol compound after synthesis and 6 consecutive water / dichloromethane extractions.

Figure 2 illustrates the profile obtained with the spermidine-cholesterol preparation in semi-preparative HPLC and visualizes the three mono-cholesterol isomers.

Figure 3 illustrates the profile obtained with the spermine-cholesterol preparation in semi-preparative HPLC and visualizes the two mono-cholesterol isomers.

FIG. 4 schematically represents the stages of synthesis of the N'-spermidine cholesterol isomer via protective groups.

Figure 5 shows a thin layer chromatography of spermine-cholesterol and spermidine-cholesterol isomers produced synthetically or purified from the lipid preparation (development by ninhydrin). FIG. 6 schematically represents the stages of synthesis of the N ^ spermine isomer cholesterol via protective groups.

Figure 7 illustrates the transfection efficiency of the spermidine-cholesterol / DOPE mixture (molar ratio of 1: 0.74). The Y axis represents the absorbance read at 405 nm, the X axis the amount of lipid / colipid (μg) per well and the Z axis the amount of DNA (μg) per well.

Figure 8 illustrates the transfection efficiency of the C2C12 muscle line with the cationic lipids (A) Transfectam, (B) DC-Chol, (C) spermidine-cholesterol and (D) spermine cholesterol.

Figure 9 illustrates the transfection efficiency of the 3 isomers of spermidine-cholesterol (pic 1 to 3 of Figure 2). The X axis represents the absorbance read at 405 nm and the Y axis the amount of DNA (μg) per well.

1. Synthesis and purification of the spermidine-cholesterol compound

The synthesis is carried out under conditions favoring the production of monosubstituted derivatives. To do this, we place ourselves in excess of amines (preferably 10 to 20 times of molar excess), at low temperature (-20 "C) to slow the reaction rate and in an inert gas atmosphere (for example argon ) to respect the anhydrous conditions. It is indicated that the basic products spermidine, cholesteryl chloroformate and N-hydroxysuccinimide (NHS) are commercially available (for example Sigma; reference S-2626, C-6633 and H-7377 respectively).

We introduce into a first balloon under argon:

- 1.6 g of cholesteryl chloroformate (i.e. 3.4 mmol) taken up in 15 ml of anhydrous dichloromethane (SDS, Peypin, France, reference 2910E16), and - 0.5 g of NHS (4.1 mmol) dissolved in 5 ml of dioxane (SDS, Peypin, France, reference 03610) rendered anhydrous by passage over a molecular sieve which traps the water molecules.

The mixture A thus formed is left under stirring for 15 h at 4 ° C.

In parallel, in a second flask, 10 g of spermidine (69 mmol) are introduced. After drying under vacuum for 13 to 15 h, the following are added in order to constitute mixture B:

5 ml of anhydrous dichloromethane, and

5 ml of diisopropylethylamine (40 mmol) (DIPEA, Applied Biosystem, St Quentin, France; reference 400136).

Mixture B is also placed under an argon atmosphere.

The mixture A is added slowly to the mixture B via a capillary of 100 μm in diameter, at a rate of approximately 1 ml per min, at -20 ° C. and with stirring. It is indicated that the mixture A is connected by the same type of capillary to a connected balloon containing 20 ml of anhydrous dichloromethane placed at a given height which defines the flow rate of passage from A to B.

The preparation thus obtained is first lyophilized at -80 ° C to evaporate the solvent and the DIPEA. The lyophilisate is taken up in approximately 50 ml of dichloromethane, then is extracted with the same volume of water. After separation of the phases, the dichloromethane fraction is recovered which is subjected to several additional extractions (in general 5) under the aforementioned conditions. The free amines which are extremely hydrophilic pass into the aqueous phase. The dichloromethane fraction is collected and then dried. Finally, the quality of the product can be improved by two successive stages of crystallization from methanol. The first carried out in a volume of 120 ml, makes it possible to crystallize the most hydrophobic products such as cholesterol and polysubstituted amino derivatives. The second crystallization is carried out on the supernatant of the first previously concentrated to 10 ml by evaporation at 65 ° C. The monosubstituted amines are recovered in the recrystallisate at -20 ° C. As an indication, 1 to 1.5 g of spermidine-cholesterol are obtained under these conditions (1.8 to 2.7 mmol).

It is of course possible to carry out, following crystallizations, additional extractions in the water / dichloromethane mixture (volume to volume) to further improve the purity of the product.

2. Synthesis and purification of the spermine-cholesterol compound

The process used is similar to that described above except for the following modifications:

Mixture A is formed by adding 2.09 g of cholesteryl chloroformate (4.65 mmol) placed in 15 ml of anhydrous dichloromethane and 0.6 g of NHS (5.58 mmol) taken up in 5 ml of dioxane.

Mixture B results from the reaction of:

- 17.7 g of spermine (87.5 mmol; Sigma, reference S-3256) previously dried under vacuum, and

- 5 ml of anhydrous dichloromethane, and

- 5 ml of DIPEA (40 mmol).

With regard to the yields, this process makes it possible to generate 0.5 to 1 g of product (0.8 to 1.6 mmol). 3. Ca ra ct er is ati on of spe pa ra ti ons linked to spermidine-cholesterol and spermine-cholesterol.

3.1 By thin layer chromatography

Analysis by thin layer chromatography is carried out after each key step of the purification protocol to rapidly determine the contamination of the synthetic product by the basic constituents. Conventional techniques are applied (E. Stahl, Dunnschicht-Chromatographie, ein Laboratoriumshandbuch, 2 ed., Spinger-Verlag Berlin, Germany, 1967).

Briefly, an aliquot is deposited on a silica plate on glass, which is placed in the migration solvent N-propanol / ammonia 32% (in water) / water (6/3/1 by volume). The various constituents are visualized by sulfuric acid which reveals all the molecules containing cholesterol or by ninhydrin 0, 1% in ethanol which reveals only the amino products.

The separation is a function of hydrophobicity, the most hydrophobic products migrating most rapidly. Free spermine and spermidine are detected in the aqueous phases collected during water / dichloromethane extractions, while the last dichloromethane phase mainly contains the amino substituted by cholesterol. The frontal retention (Rf) characteristic of each product is 0.71 and 0.68 for spermidine mono-cholesterol and spermine monocholesterol.

3.2 By high performance liquid chromatography in reverse phase

The quality of the lipid preparation can also be monitored by HPLC (for High Performance Liquid Chromatography in English) under very specific analytical conditions. A water / isopropanol gradient is applied in the presence of 0.1% trifluoroacetic acid (TFA) under the following experimental conditions: column: supelcosil LC-ABZ 15 cm x 4.6 mm (Supelco; 5-9140) analysis time: 35 minutes flow rate: 0.5 ml / minute injection: 10 to 100 μg in 100 μl of solvent at initial conditions ( 80% A 20% B) /: gradient A = water + 0.1% trifluoroacetic acid gradient B = isopropanol (HPLC 205 quality) + 0.1% trifluoroacetic acid

- initial conditions: 80% A 20% B

- 6 minutes: 60% A 40% B

- 20 minutes: 0% A 100% B

- 30 minutes: 0% A 100% B - 32 minutes: 80% A 20% B

detection: absorbance at 205 nm.

In general, the different constituents are separated according to their hydrophobicity: the free spermine and spermidine are not retained while the hydrophobic molecules are. The monosubstituted derivatives are eluted first before the polysubstituted derivatives.

Figure 1 shows the typical HPLC profile obtained on the dichloromethane phase generated during the synthesis of the spermidine-cholesterol compound. The results show that the preparation tested largely comprises the spermidine mono-cholesterol sought (retention time 15 to 18 min). The polysubstituted derivatives are also present (22 to 28 min) but will be eliminated during the subsequent crystallization stages. An indirect analysis technique of the non-retained fraction makes it possible to estimate the amount of free spermidine contaminating the preparation. To do this, the first harvested fractions are deposited on a thin layer chromatography plate, revealed by ninhydrin and compared to a concentration range established from commercial spermidine. If the contamination is greater than 1%, one or more additional dichloromethane extraction steps are introduced in the purification protocol.

The same is done to analyze the spermine-cholesterol preparations.

4. Purification and characterization of the monocholesterol isomers from lipid preparations.

4.1 Purification of the isomers by semi-preparative HPLC The various monosubstituted isomers have very close hydrophobicity and exhibit a similar retention time under the conditions applied previously. New conditions must therefore be defined which favor their separation in order to be able to characterize them and study their respective transfectional efficiency. For this purpose, several hundred μg of lipid preparations were deposited on a 25 cm × 10 mm supelcosil LC-ABZ HPLC column (Supelco; reference 5-9170) and eluted in a water / acetonitrile gradient in the presence of 0.1% TFA . The experimental procedures for performing the chromatograms of Figures 2 and 3 which illustrate the isolation of the isomers of spermine and spermidine mono-cholesterol respectively are as follows:

Figure 2 (Spermidine - monocholesterol)

column: supelcosil LC-ABZ 25 cm x 10 mm (Supelco; 5-9170); analysis time: 12 minutes; flow rate: 5 ml / minute; injection: ~ 100 μg in 100 μl of solvent at initial conditions (70% A 30% B): gradient A = water + 0.1% trifluoroacetic acid gradient B = acetonitrile (HPLC quality 205) + 0.1% trifluoroacetic acid 9?

- initial conditions: 70% A30% B - 1.5 minutes: 65% A35% B -2 minutes: 33% A67% B - 11 minutes: 28% A72% B

- 11.5 minutes: 70% A30% B

detection: absorbance at 205 nm

Figure 3 (Spermine - monocholesterol):

column: supelcosil LC-ABZ 25 cm x 10 mm (Supelco: 5-9170); analysis time: 9 minutes; flow rate: 5 ml / minute; injection: ~ 500 μg in 100 μl of solvent at initial conditions (65% A35% B): gradient A = water + 0.1% trifluoroacetic acid gradient B = acetonitrile (quality HPLC 205) + 0.1% trifluoroacetic acid

- initial conditions: 65% A35% B

- 0.2 minute: 65% A 35% B

- 1 minute: 39% A 61% B -8.5 minutes: 34% A66% B

- 8.8 minutes: 65% A35% B

. detection: absorbance at 205 nm

The chromatogram (Figure 2) shows 3 peaks of spermidine-cholesterol: a first well isolated peak (peak 1 migrating at 7.2 min) and two less close peaks (peaks 2 and 3 migrating at 9.6 min and 10.6 respectively min). Optionally, reinjection of the fractions covering the latter makes it possible to reduce the contamination of one by the other. HPLC analysis of the spermine-cholesterol preparation (Figure 3) produces two well separated peaks exiting at 4.9 min and 7 min.

4.2 Characterization of the mass of isomers by ionization-electrospray mass spectrometry (ESI-MS)

The HPLC fractions covering each of the peaks were pooled and analyzed by mass spectrometry (Edmonds and Smith, Hectrospray ionization mass spectrometry, in Methods in Enzymology, Vol 193, pp 412-432; ed. McCloskey, Académie Press, San Diego, USA , 1990).

The 3 peaks of spermidine-cholesterol contain products with a molecular mass of approximately 558 Da which is consistent with that expected for a monosubstituted product (calculated mass: 557.84 Da). These data confirm the production during chemical synthesis of the three isomers by attachment of the cholesterol group to one of the three amines of spermidine.

Similarly, the HPLC fractions collected from the lipid preparation of spermine-cholesterol contain products with a molecular mass of approximately 615 Da corresponding to that expected for spermine substituted by a cholesterol molecule (calculated mass 614.96 Da) . Due to the symmetry of spermine, only two isomers can be produced depending on the fixation of cholesterol on a primary or secondary amine.

4.3 Characterization of the isomers by 13 C NMR

The semi-preparative HPLC fractions are extracted into a 20 mM sodium dichloromethane mixture in order to deprotonate the amines before being subjected to 13C NMR (Gunther, NMR Spectroscopy,

2ed., Wiley, Chichester, UK, 1995; Kimberley and Goldstein, 1981, Anal. Chem.

53, 789-793; Dagnall et al., 1984, J. Chem. Soc. Perkin Trans. II, 435-440). 7/29118 PCÏ7FR97 / 00225

24

This analysis makes it possible to determine which type of amine, primary or secondary, the cholesterol group is attached to. It is carried out in parallel with spermine, spermidine and cholesterol standards and the spectra obtained with the molecules to be tested and the standards are compared. The displacement of the signals compared to the standards indicates a change in the chemical environment depending on the cholesterol binding site.

With regard to the spermine-cholesterol compound, peak 1 displayed on the HPLC profile reported in Figure 3, contains the isomer in which cholesterol is linked to the secondary amine and peak 2 to the primary amine. As regards spermidine-cholesterol, the substituted isomer on the secondary amine is detected in the fractions covering the most rapidly emerging peak (peak 1 in FIG. 2) and the isomers modified on the primary amines correspond to the peaks 2 and 3. The latter were differentiated by comparison with a standard constituted by the spermidine-cholesterol isomer obtained in pure form by chemical synthesis

(see 5. 1). Peak 2 corresponds to the synthetic isomer in which cholesterol is fixed on the primary amine of the short arm (CH ₂ ) ₃ of spermidine and peak 3 to that modified on the long arm (CH ₂ ) ₄ .

Example 5 Preparation of Isomer Synthetically

5.1 Synthesis of the substituted spermidine-cholesterol isomer at the level of the primary amine of the short arm.

The main steps are shown in Figure 4:

1) synthesis of an intermediate product with two possible intermediates according to the site of addition of the acrylonitrile part;

2) protection of the amines by a protective group Boc (Boc ≈ tert-butyloxycarbonyl);

3) catalytic reduction of the terminal nitrile;

4) reaction with cholesteryl chloroformate, and 5) deprotection of amines by TFA. The protocol applied to obtain 1, 4-Diboc-spermidine is published by Goodnow et al. (1990, Tetrahedron 46, 3267-3286) with the difference that the step of purifying the intermediates is omitted. 50 mg (177 μmoles) are taken up in 3 ml of anhydrous dichloromethane and 40 μl (417 μmoles) of diisopropylethylamine then cooled to 0 ° C. The reaction is started by slowly adding 90 mg (200 μmol) of cholesterylchloroformate placed in 3 ml of anhydrous dichloromethane. It is checked after one hour that the reaction is complete by thin layer chromatography. The deprotection of the amines is carried out by adding 300 μl TFA (final concentration of 4.7%). The reaction mixture is left at room temperature overnight.

The synthesis product is purified by chromatography on silica gel (5 g of gel for a column 2 cm in diameter and 3.5 cm in height) using a dichloromethane / ethanol mixture (77/23) to load the column and dichlorornethane / methanol / triethylamine (15/5/1) for the elution. The isomer is characterized by reverse phase HPLC and by thin layer chromatography. As illustrated in FIG. 5, the synthetic product has an Rf of 0.72 characteristic of spermidine-cholesterol (migration identical to the isomer of peak 2).

5.2 Synthesis of the isomer of spermine-cholesterol substituted at the level of the primary amine.

The synthesis protocol is shown schematically in the

Figure 6. The intermediate NN 'dipropionitrile 1, 4 diamino butane is formed by reaction of 2 moles of acrylonitrile and 1 mole of 1, 4 diaminobutane. In practice, 75.2 mmol of acrylonitrile (Fluka; reference 01710) are added with stirring (i) at 0 ° C. to (ii) 37.6 mmol of 1,4 diaminobutane (Fluka; reference 32790) taken up in 2 ml of methanol and 3.8 ml of dichloromethane at 0 ° C. After returning to ambient temperature, 99 mmol of Boc-On (2-Boc-oxyimino) -2-phenylacetonitrile) (Flu ka; reference 1 5475) suspended in a mixture are added to the mixture. dichloromethane / methanol (1/3). The Boc group is fixed on the free amines in this case the secondary amines. The "protected" product (dipropionitrile di-Boc 1,4 diamino butane) is subjected to different cycles of evaporation and extractions to remove the solvents and contaminating products. Then, the nitrile reduction step is carried out in the presence of 140 mmol of UAIH4 (Sigma; LO260) taken up in ether by controlling the temperature (0 ^β C). When the reduction is considered to be complete (by thin layer chromatography), the reaction is stopped by adding sodium hydroxide. The mixture is filtered, the ethereal fraction is recovered and the reaction by-products are eliminated by several extractions with a 20% aqueous NaCl solution.

The product is purified by chromatography on silica gel (5 g of silica for 1 g of product in 10 ml of equilibration solvent). The silica containing the adsorbed product is applied to a column 2 cm in diameter and 23 cm in height loaded with 25 g of gel. The equilibration solvent consists of the mixture of dichloromethane / ethanol (2/1) and the eluting buffer of dichloromethane / methanol / triethylamine (5/15/1). The fractions containing spermine-diboc on which cholesterylchloroformate is reacted are determined by thin layer chromatography (1.48 moles of spermine diboc for 492 μmoles of cholesterylchloroformate). The reaction mixture is lyophilized, suspended in dichloromethane, then deprotected in the presence of 5% TFA and finally purified on silica gel chromatography under the same conditions as above except that the column used is more resolving. The N i spermine-cholesterol isomer is characterized by analytical HPLC and thin layer chromatography. The frontal retention of the synthetic product is characteristic of the spermine-cholesterol (Figure 5) constituted here of the isomer of peak 2. Example 6 Transfer of Nucleic Acid In Vitro

6.1 Plasmids

The reporter plasmid pCH UON is used to evaluate the transfection efficiency of the compounds of the invention. It includes the promoter / early enhancer block of the SV40 virus (Simien Virus) directing the expression of the LacZ gene coding for the β-galactosidase of Escherichia coli provided in N-terminal with a nuclear localization signal originating from SV40. But any other plasmid can also be used. Mention may be made, for example, of a plasmid expressing (i) a gene for resistance to an antibiotic, (ii) a luciferase gene or (iii) a gene which is easily detectable. Such plasmids are known in the art. All the experiments are carried out in parallel with DC-Chol as a transfection control and, when indicated, transfectam and Lipofectin are also included. DC-Chol is obtained by chemical synthesis on a scale of 20 g according to the process described in Gao et al. (1991, BBRC 179, 280-285). Transfectam and Lipofectin are commercially available (Sepracor, MA, USA; 26395 1 and Gibco BRL, MD, USA; 18292.01 1 respectively). In order to increase the transfection efficiency, the cationic lipids can be combined with a colipid. The colipid used is DOPE (Sigma, P5078) which is widely used in the state of the art.

6.2 Determination of the optimal lipid to colipid ratio (DOPE)

The transfections are carried out in 96-well plates (Falcon) each seeded with 2 × 10 ⁴ to ⁴ × 10 ⁴ A549 cells (ATCC CCL 185) derived from a human lung carcinoma. The culture is left overnight at 37 ° C. in a conventional medium supplemented with fetal calf serum (FCS) at a final concentration of 10%. First, the cationic lipid is mixed with a variable amount of DOPE to evaluate the effect of this on transfection efficiencies. 500 μg of spermine-cholesterol or spermidine-cholesterol are dissolved in 50 to 100 μl of dichloromethane, placed in a sterile glass tube and placed in the presence of a variable amount of DOPE taken up in 50 to 100 μl of chloroform (see table below) -after).

spermine-cholesterol spermidine-cholesterol

Lipid DOPE Ratio Lipid DOPE Ratio

(M) (μg) (μg) (M) (μg) (μg)

1: 0.5 290 500 1: 0.5 334 500

1: 0.86 500 500 1: 0.74 500 500

1: 1,580,500 1: 1,668,500

1: 2 1160 500 1: 2 1336 500

1: 4 2320 500 1: 4 2672 500

The mixture is dried under a stream of nitrogen. 1 ml of 0.9% NaCl is added and the whole is left overnight at 4 ° C. before being sonicated for 3 to 5 min until a milky to translucent solution is obtained. The lipid mixture is stored at 4 ° C until use. Dilutions in series (2 in 2) of the plasmid solution and the lipid mixture are prepared. A wide range of lipid / DNA complexes is formed by transferring a given volume of DNA dilutions into an identical volume of lipid dilutions.

After aspirating the culture medium, the cells are placed in the presence of 100 μl of the above complexes and incubated in a humid atmosphere at 37 ° C., 5 to 10% CO ₂ . The medium is added in two stages: 50 μl of medium comprising 30% FCS 4 to 6 h after transfection and then 100 μl to 10% FCS 24 h after. The last two columns are generally reserved for witnesses and the standard range. The β-galactosidase activity is revealed, 48 h post-transfection, after lysis of the cells by addition of 50 μl of lysis buffer (0.1% Triton X-100, 250 mM Tris-HCl pH8). 50 μl of PBS buffer comprising 0.5% BSA (Bovine Serum Albumin) is added. The standard range consists of serial dilutions of a β-galactosidase standard (Boehringer Mannheim; reference 567 779) prepared in the same PBS-BSA buffer. As an indication, the range extends from 50 to 0.39 ng / ml. Finally, 150 μl of ONPG substrate (OrthoNitroPhenol Galactopyrannoside Sigma: reference NI 127) are added at a concentration of 2 mg / ml in PBS, 0.5% BSA. The absorbance is measured at 405 nm in an Elisa reader after 30 min, 4 h or 24 h of incubation.

The best results are obtained when the cells are transfected in roughly equimolar conditions (lipid / colipid 1: 1 to 1: 074). Thus, in the case of the spermidine-cholesterol / DOPE mixture at a ratio 1: 0.74 (FIG. 7), the β-galactosidase activity is high (absorbance reaching 3 to 3.5) this for a wide range of concentrations of lipids and DNA. The transfecting power of the mixture in a 1: 2 ratio is effective, especially in the range with high lipid concentration. As for the extreme ratios (1: 0.5 or 1: 4), the transfection activity, although lower, is still acceptable.

As regards the spermine-choiesterol / DOPE mixture, a ratio of between 1: 0.86 and 1: 2 is preferred, for which the transfection efficiency is highest for a range of dilutions of the lipid mixture and of the Fairly large DNA.

6.3 Transfection of established muscle cell lines.

This series of experiments is carried out in parallel with the two compositions of the invention, DC-Chol and Transfectam on the line C2C 12 (ATCC CRL-1772) derived from mouse myoblasts. The cells are distributed in 6-well culture plates due to ⁵ × 10 ⁵ cells per well, cultivated overnight under the usual conditions and transfected the next day with different dilutions of lipid / colipid / DNA complexes (cationic lipid / DOPE ratio roughly equimolar except in the case of the transfectant which is used as it is). After 48 h of culture, the cells are fixed and the enzymatic activity is revealed by X-Gal. The nuclei of the transfected cells (which produce a nuclear β-galactosidase), appear in blue and the proportion of stained cells is estimated. Figure 8 illustrates the results obtained for each lipid composition. The number of blue cells is significantly higher with the spermine-cholesterol or spermidine-cholesterol complexes (transfection rate roughly equivalent in both cases) than with those consisting of DC-Chol. In comparison, the transfection rate of Transfectam-based complexes is very low.

6.4 Transfection of primary myoblasts.

The myoblasts are taken from the diaphragms of adult C57 Black 10 mice or from the thighs of m d.v adult mice (4 weeks). The muscle samples are dissociated in a mixture of dispase and collagenase and the cells are cultured in 24-well plates and in DMEM or F14 medium supplemented with growth factors (insulin, FGF, EGF, glutamine) and serum. calf at a concentration of 10%.

The cells are transfected in a serum-free medium with 1.875 μg of plasmid pCHUON complexed with various lipids (spermine-cholesterol, spermidine-cholesterol, DC-Chol and Transfectam) in the presence of DOPE (1: 1 lipid / DOPE ratio and DNA ratio) / lipid mixture 1: 4). By way of comparison, a naked DNA witness is also constituted. 4 h after transfection, the culture is continued in a medium supplemented with serum. Myoblasts are able to proliferate and then merge into multinucleate structures constituting muscle fibers or myotubes. The cells are fixed after 48 h, treated with X-Gal and the estimated number of colored cells. The best transfection rates are obtained with spermine-cholesterol, spermidine-cholesterol and Transfectam complexes. The blue cells are much weaker in the case of DC-Chol. Naked DNA transfects even less well (less than 1% of positive β-galactosidase cells).

Another series of experiments is carried out on primary cardiomyocites of cardiac muscles of newborn C57 Black 10 mice, with the same lipid mixtures and also lipofectin / DOPE. The transfection is carried out after 4 days of culture in a conventional medium and the transfection efficiency determined 2 days later. Under these conditions, the Transfectam / DOPE mixture gives a percentage of blue cells slightly higher than those based on spermine-cholesterol and spermidine-cholesterol, but much higher than the DC-Chol and Lipofectin / DOPE mixtures. As before, naked DNA transfects very poorly.

6.5 Transfectional efficacy of isomers purified by HPLC

a) on lung cells;

The protocol applied is similar to that described in paragraph 6.2 with the difference that 500 μg of isomers purified by semi-preparative HPLC are used in place of the lipid preparation. In the case of spermidine-cholesterol, the substituted isomer at the level of the primary amine of the long arm of spermidine (corresponding to peak 3 in FIG. 2) has the best transfection activity in A549 cells compared to the other two isomers (Figure 9) and to the mixture, (not shown).

These experiments are reproduced using a plasmid expressing the luciferase gene placed under the control of the early CMV promoter. The spermidine-cholesterol isomers have very similar transfection activities comparable to those obtained with the mixture. The isomer corresponding to peak 2 is slightly weaker than isomers 1 (substitution on the secondary amine) and 3 (substitution on the primary amine of the long arm).

Experiments with spermine cholesterol isomers show a clear superiority of the substituted isomer over the primary amine (pic 2), particularly at low DNA concentrations (for example 0.525 μg of DNA for 4 to 0.5 μg in lipids). The isomer corresponding to peak 1 has a transfection activity of the same order of magnitude as that of the mixture.

b) on muscle satellite cells

The protocol applied is similar to that described in paragraph 6.4. apart from the fact that 500 μg of isomers purified by semi-preparative HPLC and the reporter plasmid luciferase are used.

Among the spermidine-cholesterol isomers, that substituted on the secondary amine (corresponding to peak 1) is 3 to 4 times more effective than the mixture. The best activity is obtained for 2 μg of DNA and 4 to 0.25 μg in total lipid. The spermine-cholesterol isomer substituted on the secondary amine is more effective than the mixture (in particular in a range 1 μg of DNA for 4 to 0.25 μg in total lipid).

Taken together, these data show that each isomer can be adapted to the transfection of a particular cell type. In particular, the isomers substituted on the primary amines of spermine cholesterol and spermidine cholesterol efficiently transfect the lung cells while the isomers substituted on the secondary amines are particularly effective with regard to muscle cells.

Claims

1. Compound of formula:

HN - [(CH2) p - NJ- _n II (II)

I I

Ri R ₂ in which:

- R _! represents hydrogen, and

- R ₂ represents hydrogen or for only one of the n substituents, a deoxy residue in position 3 of cholesterol or a monovalent or polyvalent residue of a cholesterol derivative carrying one or more reactive groups capable of reacting with a secondary amine group a polyamine molecule;

- n is an integer from 2 to 6; and

- p is an integer from 2 to 6 identical or different in the n fragments

[(CH ₂ ) _P - N]

R ₂

2. Compound according to claim 1, characterized in that only one of the n substituents R ₂ does not represent hydrogen.

3. Compound according to claim 2 characterized in that n = 2 or 3 and p = 3 or 4.

4. Compound according to one of claims 1 to 3, characterized in that R ₂ is a monovalent deoxy residue in position 3 of cholesterol.

5. Compound according to one of claims 1 to 4, characterized in that said residue of cholesterol derivative in position 3 has the formula:

6. Compound according to one of claims 1 to 5, characterized in that it corresponds to the formula:

- N - (CH ₂ ) ₃ - N - (CH ₂ ) ₄ - - NH (VI)

Ri R ₂ ι R ₂ ²

in which :

- Ri and R ₂ ² represent hydrogen and;

- R ₂ ¹ represents undit monovalent residue of cholesterol or of a cholesterol derivative in position 3.

7. Compound according to one of claims 1 to 5, characterized in that it corresponds to the formula:

H - - H (VII)

in which :

one of R ₂ ι and R ₂ ² represents undit monovalent residue of cholesterol or of a cholesterol derivative in position 3 and the other represents hydrogen, and - Ri and R ₂ ³ represent hydrogen.

8. Pharmaceutical composition comprising a compound according to one of claims 1 to 7 and a therapeutically active substance negatively charged.

9. Composition according to claim 8, characterized in that said active substance is a nucleic acid.

10. Composition according to claim 9, characterized in that said nucleic acid fragment is a DNA or a sense or antisense RNA.

1 1. Composition according to claim 10, characterized in that said nucleic acid comprises a gene of interest and elements of expression of said gene of interest.

12. Composition according to one of claims 9 to 11, characterized in that said nucleic acid is in the form of a plasmid.

13. Composition according to one of claims 9 to 12, characterized in that the weight ratio of the compound of claims 1 to 7 to said nucleic acid is from 0.01 to 100.

14. Composition according to claim 13, characterized in that said weight ratio is from 0.1 to 10.

15. Use of a composition according to one of claims 8 to 14 for preparing a transfer vector of said nucleic acid in cells in vitro, ex vivo or in vivo.

16. A vector for transfer of nucleic acids into cells of microorganisms, plants, humans or animals, in particular mammals, characterized in that it comprises a composition according to one of claims 9 to 15 and an adjuvant capable of 'Improving the transfecting power of said composition.

17. Vector according to claim 16, characterized in that said adjuvant is a neutral co-lipid.

18. Vector according to claim 17, characterized in that said neutral co¬ lipid is a phospholipid.

19. Vector according to claim 18, characterized in that said phospholipid is dioleyl phosphatidyl ethanolamine (DOPE).

20. Vector according to one of claims 17 to 19, characterized in that the molar ratio of the compound according to one of claims 1 to 7 to the adjuvant neutral co-lipid is from 0.1 to 5, preferably 0.25 to 2.

21. Use of a composition according to one of claims 8 to 14 for the preparation of a said nucleic acid transfer vector in mammalian muscle cells.

22. Use of a composition according to one of claims 8 to 14 for the preparation of a said nucleic acid transfer vector in mammalian lung cells.

23. Method for purifying an isomer of a compound according to one of claims 1 to 7 from a mixture of isomers of said compounds, characterized in that a high performance liquid chromatography is carried out in reverse phase with a LC-ABZ supelcosil column.

24. Use of a composition comprising a gene of interest and the expression elements of said gene of interest and a compound corresponding to the formula:

H - N - (CH ₂ ) ₃ - NH - (CH ₂ ) 4 - NH l I

Ri R ₂ 2

in which :

- Ri represents H and R ₂ ² represents undit monovalent residue of cholesterol or of a cholesterol derivative in position 3, or

- R ₂ ² represents H and R i represents undit monovalent residue of cholesterol or of a cholesterol derivative in position 3, or

of a composition comprising a gene of interest and the expression elements of said gene of interest and a compound corresponding to the formula:

H - N - (CH ₂ ) ₃ - NH - (CH ₂ ) ₄ - NH - (CH ₂ ) ₃ - N - I l ι i Ri R ₃

in which :

- Ri represents H and R ₃ represents undit monovalent residue of cholesterol or of a cholesterol derivative in position 3, or

- R ₃ represents H and Ri represents undit monovalent residue of cholesterol or of a cholesterol derivative in position 3;

for the preparation of a vector for transferring said gene of interest into muscle or lung cells of mammals.

25. Use according to claim 24 for the preparation of a vector for transferring said gene of interest into mammalian lung cells.

26. Use according to claim 24 or 25, characterized in that said gene of interest and its expression elements are included in a plasmid.

27. Use according to one of claims 24 to 26, characterized in that the weight ratio of said compound to said gene of interest or plasmid is from 0.01 to 100 and preferably 0.1 to 10.

28. Use according to one of claims 24 to 27, characterized in that said composition further comprises an adjuvant capable of improving the transfecting power of said composition.

29. Use according to claim 28, characterized in that said adjuvant is dioleyl phosphatidyl ethanolamine (DOPE).

30. Use according to claim 29, characterized in that the molar ratio of said compound to said adjuvant is from 0.1 to 5, preferably 0.25 to 2.