EP1011735A1

EP1011735A1 - METHOD OF $i(IN VIVO) TRANSFORMATION UTILIZING LIPID VEHICLES

Info

Publication number: EP1011735A1
Application number: EP98944788A
Authority: EP
Inventors: Howard J. Uni. of Rochester Medical Cnt FEDEROFF; Mary Haak-Frendscho
Original assignee: Promega Corp
Current assignee: Promega Corp
Priority date: 1997-09-08
Filing date: 1998-09-08
Publication date: 2000-06-28
Also published as: WO1999012575A1; AU9224398A

Abstract

An in vivo method for the transformation of animals utilizing a transformation reagent containing an admixture of a DNA or RNA encoding a gene product of interest and a lipid composition containing a fusogenic lipid and a cationic lipid is disclosed.

Description

METHOD OF IN VIVO TRANSFORMATION UTILIZING LIPID VEHICLES

FIELD OF THE INVENTION

The invention is directed to clinical and therapeutic in vivo transformation of animal cells utilizing DNA or RNA molecules combined with lipid vehicles to express gene products encoded on the DNA or RNA molecules.

BIBLIOGRAPHY Complete bibliographic citations of the non-patent prior art cited below are contained in the Bibliography section. All of the patent and non-patent references cited below are incorporated herein by reference in their entireties.

DESCRIPTION OF THE PRIOR ART Prior art methods for in vivo transfection of mammalian cells rely predominately upon the use of viral vectors. The wild-type genomes of various Herpes Simplex virus (HSV) and adenovirus have been extensively manipulated to enable the virus to function as suitable vectors. While these genetically- engineered viral vectors are not pathogenic, they are antigenic. HSV and adenovirus vectors, as well as other viral vectors, readily induce a immuno- response in mammalian hosts which are inoculated with the virus. The antigenic properties of viral vectors represents a distinct shortcoming to their in vivo utility. Most mammals, including humans, generate antibodies in response to the viral vectors. Consequently, after the first inoculation with viral vector, the host's immune system generates anti- virus antibodies which then destroy the efficacy of any virus subsequently injected. See, for instance, E. Marshall (1995).

Additionally, the safety of such viral vectors is unknown. The possibility exists that the virus and any genetic construct housed therein might be propagated in the initial host and unintentionally transmitted to others. Therefore, the widespread use of viral vectors for in vivo transfection in non-quarantined animal populations must wait for a definitive answer with respect to the ultimate fate of the vector.

Various non-viral vectors for in vitro transfection have been described in the prior art. Most of these non-viral vectors are lipid-based compounds. As a group, they generally suffer from limited transfection ability and/or cytotoxicity. For example, a commercially-available lipid formulation for in vitro transfection consists of a 1 to 1 formulation of N-{l-(2,3-dioleyl)propyl}-N,N,N- trimethylammonium bromide and dioleylphosphatidylethanolamine (DOPE). This formulation suffers from the above-noted cytotoxicity and also inhibits protein kinase C activity, an undesirable side effect. Examples of lipid-based agents for in vitro transfection of animal cells are presented in U.S. Patent No. 5, 171,678; U.S. Patent No. 5,279,833; and U.S. Patent 5,527,928.

Of these, U.S. Patent No. 5,527,928, issued June 18, 1996, to Nantz et al. is particularly relevant. This patent describes a synthetic cationic lipid having the systematic name N,N,N',N'-tetramethyl-N,N'-bis-(2-hydroxyethyl)-2,3- di(oleyloxy)-l,4-butanediammonium iodide. The corresponding structure appears as follows:

The patent also describes the in vitro transfection of NIH 3T3 cells utilizing lipids of the above compound which contain plasmid DNA.

Compositions containing the above compound are commercially marketed by the Promega Corporation, Madison, Wisconsin, under the trademarks "Tfx"- 10, "Tfx"-20 and "Tfx"-50 reagents for the transfection of eukaryotic cells (product nos. E1811, E2381, E2391, and E2400.) For purposes of brevity, these compositions will be collectively referred to herein as "Tfx" reagents. All of the "Tfx" reagents available from Promega contain the same concentration of Nantz et al. 's above-noted compound and varying concentrations of DOPE, the concentration of DOPE being designated by the number following the "Tfx" trademark. For a further description of the "Tfx" reagents, see Promega Technical Bulletin No. 216.

As a general proposition, in vitro transfection is performed in rapidly dividing cells in order to maximize uptake of the exogenous DNA. Methods for the in vivo transfection of quiescent cells, however, such as central nervous system (CNS) cells, are of particular interest to gene therapy. In vivo gene transfer into central nervous system (CNS) cells represents a valuable tool for examining gene product function as well as for modeling gene therapy. See, for example, Andersen and Breakefield (1995), Bergold et al. (1993), Brooks et al. (1997), Davidson et al. (1993), During et l. (1994) Federoff (1995), Federoff et al. (1992), Geller and Breakefield (1988), Geller et al. (1993), Geschwind et al. (1996), Glorioso et al. (1994), Ho et al. (1995), Kaplitt et al.

(1994), and Linnik et al. (1995). Unlike in vitro transfection in log-phase growth cells, CNS cells are quiescent. Stereotaxic neurosurgical approaches in non-human primates and rats have allowed for fine spatial control of the operative environment, thereby enabling reproducible gene transfer. Virtually all gene transfer to date has relied on the introduction of a reagent using a stereotaxic frame-mounted microsyringe that is manually driven. Microprocessor-controlled stereotaxic microsyringes are also available and offer much improved precision and reproducibility compared to manually-driven syringes. Microprocessor- controlled microsyringe devices are available commercially. (For example, the "UltraMicroPump" brand syringe controller, available from World Precision

Instruments.)

BRIEF DESCRIPTION OF THE DRAWINGS

Fig. 1 is a schematic of World Precision Instruments' "UltraMicrosyringe" Pump. The pump is designed to dispense down to picoliter volumes from a microliter syringe. The pump utilizes a stepper motor-driven lead screw having an ultrafine pitch. The lead screw drives the microsyringe plunger. The pump is controlled by a microprocessor controller which is programmable for volume, rate, and syringe size.

Figs. 2A, 2B, 2C, 2D, 2E, and 2F - X-gal histochemistry of HSVlac- infected tissue. Sections (40 μm in width) representative of the injection site (Figs. 2B and 2E), a site anterior of the injection (Figs. 2 A and 2D) and a site posterior of the injection (Figs. 2C and 2F) were processed in a developing solution containing X-gal to visualize expression of j8-galactosidase. Animals were either injected with a microprocessor-controlled pump (Figs. 2A, 2B, and 2C) or manually (Figs. 2D, 2E,and 2F). All sections were counterstained with thionin and acquired at a magnification of 2.5x. Expression of 3-galactosidase is evident along the needle track (Fig. 2E, arrow) in an animal injected manually.

Figs. 3A, 3B, 3C, 3D, 3E, and 3F - X-gal histochemistry of Adlac- infected tissue. Sections (40 μm in width) representative of the injection site

(Figs. 3B and 3E), a site anterior of the injection (Figs. 3 A and 3D) and a site posterior of the injection (Figs. 3C and 3F) were processed in a developing solution containing X-gal to visualize expression of jS-galactosidase. Animals were either injected with a microprocessor-controlled pump (Figs. 3A, 3B, and 3C) or manually (Figs. 3D, 3E, and 3F). All sections were counterstained with thionin and acquired at a magnification of 2.5x. Expression of /3-galactosidase is evident anterior to (Fig. 3D) and along the needle track (Fig. 3E, arrow) in an animal injected manually.

Figs. 4A, 4B, 4C, 4D, 4E, and 4F - X-gal histochemistry of Tfx-20- transduced tissue. Sections (40 μm in width) representative of the injection site (Figs. 4B and 4E), a site anterior of the injection (Figs. 4 A and 4D) and a site posterior of the injection (Figs. 4C and 4F) were processed in a developing solution containing X-gal to visualize expression of /3-galactosidase. Animals were either injected with a microprocessor-controlled pump (Figs. 4A, 4B, and 4C) or manually (Figs. 4D, 4E, and 4F). All sections were counterstained with thionin and acquired at a magnification of 2.5x. Expression of β-galactosidase is evident within white matter fiber tracts anterior to the needle track in an animal injected manually (Fig. 4D). Figs. 5A and 5B - Infection and 3-galactosidase expression at a site distant from injection by hand. Manual injections resulted in a spread of /3-galactosidase expression as seen by X-gal histochemistry. Positive cells were identified at several sites away from the injection site in tissues adjacent to ventricular spaces in animals infected with either HSNlac (Fig. 5 A) or Adlac (Fig. 5B).

Figs. 6A, 6B, 6C, 6D, 6E, and 6F - Expression of _/3-galactosidase for all gene transfer vehicles. 3-galactosidase expression was measured quantitatively for all of the gene transfer vehicles and both modes of delivery described in the Examples. Four days after injection, tissue punches were harvested from the injection site (including the needle track), and from the corresponding anatomical location on the contralateral side of the injection. Activity levels of jS-galactosidase were measured as a function of relative light units. For all vehicles tested {HSVlac (Fig. 6A) F=23.5, df= l,6, p = .0029; Adlac (Fig. 6B) F = 13.7, df= l,6, p = .01; Tfx-20 (Fig. 6C) F = 13.7, df= l,6, p= .01; Tfx-10 (Fig. 6D) F=7.8, df= l,6, p= .04; and naked DΝA (Fig. 6E) F=5.0, df= l,6, p= .05} a significant increase was observed in the levels of gene expression in animals that were injected by the microprocessor controlled method. A significant increase in the level of gene expression also was observed on the contralateral side in animals injected by hand in the HSVlac infected group (Fig. 6A) F = 100.1, df= l,6, p = .0001. Differences in gene expression were also observed between the viral vectors, DΝA/lipid mixtures and naked DΝA groups (Fig. 6E) F = 196.3, df =4,15, p = .0001. Statistical analyses were determined using a one-way analysis of variance (AΝOVA). Expression levels contralateral to naked DΝA injections are at background.

DETAILED DESCRIPTION OF THE INVENTION It has been unexpectedly discovered that certain lipid compositions, when intimately combined with DNA molecules, can direct the in vivo expression in animal cells of gene products encoded by the DNA molecules. A first embodiment of the invention, and an invention claimed herein, is therefore drawn to a method of in vivo transformation of animal cells, including in vivo transformation of human cells, which comprises intimately associating a DNA or RNA molecule encoding a gene product of interest with a lipid composition comprising a fusogenic lipid and a cationic lipid of Formula I or Formula II:

FORMULA I

FORMULA II wherein R is a fatty acid alkyl or alkenyl group (i.e. , lauryl, myristyl, palmityl, stearyl, arachidyl, behenyl, lignoceryl, palmitolyl, oleyl, linolyl, linolenyl, and arachidonyl) and X is an anion, to yield a lipid/DNA transformation reagent. The transformation reagent is then injected into an animal host, preferably a mammalian host, at a selected injection site, whereby cells surrounding the injection site are transformed to express the gene product of interest encoded by the DNA or RNA molecule.

The method can be practiced on any type of animal, without limitation. Mammalian hosts are preferred.

Particularly preferred for use in the present invention is a lipid composition wherein the fusogenic lipid is dioleylphosphatidylethanolamme (DOPE) and the

Formula I or Formula II cationic lipid is the compound wherein R is oleyl

(preferred for Formula I) or myristyl (preferred for Formula II) and X is iodide; i.e. , the preferred Formula I compound is N,N,N' ,N'-tetramethyl-N,N'-bis-(2- hydroxyethyl)-2,3-di(oleyloxy)-l ,4-butanediammonium iodide; the preferred Formula II compound is N,N-[bis(2-hydroxyethyl)-N-methyl-N-(2,3- di(tetradecanoyloxy)propyl] ammonium iodide. While DOPE is the preferred fusogenic lipid, other lipid compounds now known to, or later discovered to, facilitate fusion of lipid vesicles with animal cell membranes can be incorporated into the lipid composition and utilized in the present invention. It is preferred that the fusogenic lipid have a net neutral charge. Preferably, the lipid composition comprises an aqueous solution containing a concentration of from about 0.1 to about 10 mM of the Formula I or Formula II cationic lipid and from about 1 to about 100 mM of the fusogenic lipid. Most preferred is when the concentration of the cationic lipid is about 1.0 mM.

A second embodiment of the claimed invention is directed to the in vivo transformation of quiescent mammalian cells, particularly CNS cells. Here, the method proceeds in the same fashion as above, with the exception that the transformation reagent is injected into a central nervous system site such as the brain (via intracranial injection) or the spinal column of a mammalian host. Of great significance, it has been found that the present method readily induces the expression of exogenous gene products in non-replicating, quiescent mammalian cells. (See the Examples for illustrative experiments utilizing mice and exogenous DNA's encoding /3-galactosidase.)

As used throughout the specification, the word "transformation" denotes a desired goal wherein an in vivo animal cell is induced to express, transiently or otherwise, a functional gene product encoded by a DNA or RNA sequence introduced into the animal host along with the transformation reagent described above.

The cationic lipids of Formula I and Formula II and the fusogenic lipid DOPE can be synthesized by prior art methods. See, for example, Nantz et al. , U.S. Patent No. 5,527,928, incorporated herein by reference. And, as noted above, compositions containing DOPE and the preferred Formula I and Formula II cationic lipid are available commercially from the Promega Corporation. The DNA or RNA encoding the gene product of interest which is to be transformed into the mammalian host may be in the form of linear polynucleotide fragments or in the form of circulized molecules such as DNA or RNA plasmids. The DNA or RNA may be single-stranded, double-stranded, mismatched, hetero- duplexed, right- or left-handed, contain methylated or non-natural nucleotide bases, or may be otherwise manipulated in ways known to the art. The DNA or RNA may be derived from any source, either by isolation from natural sources or by synthetic means, without limitation.

By the term "intimately associating," it is meant that the DNA or RNA to be transformed into the animal host is brought into intimate contact with the lipid composition described above by thorough mixing, sonication, and the like. It is much preferred that the DNA or RNA and the lipid composition be intimately associated with each other by forming liposomes of the lipid composition which contain the DNA or RNA within their lamellar or vesicle structure. Liposome formation is extensively described in the relevant prior art and need not be discussed in great detail here. Briefly, liposomes, either multilamellar vesicles (MLV's), sonicated vesicles, or otherwise, can be formed from DOPE and the cationic lipids of Formula I or Formula II by mixing together separate chloroform solutions of the two lipid components and then evaporating the chloroform solvent at a constant temperature of about 37 °C. The resulting films are placed under high vacuum to remove any remaining traces of chloroform. The lipid mixture is then re-dissolved in distilled water and thoroughly mixed to yield a suspension of multilamellar vesicles (MLVs). The suspension may be sonicated to thoroughly disperse the suspended liposomes. Adding the desired DNA or RNA to be transformed to the liposome suspension, followed by thorough mixing, causes the polynucleotide to form complexes with the suspended liposomes. It is believed this is due, in large part, to the electronic charge polarity between the net cationic nature of the liposome (due to the cationic lipid of Formula I or Formula II) and the net anionic nature of the polynucleotide due to its abundance of phosphate groups. However, the underlying physical phenomena whereby the DNA or RNA molecules become associated with, incorporated into, or otherwise complexed with the liposomes of the lipid composition to yield the transformation reagent is not critical to a complete understanding of the present invention.

The transformation reagent, in combination with a suitable pharmaceutically-acceptable diluent, is then injected into an animal host. The injection site may be intramuscular, parenchymal, intracranial, sub-dermal or sub- buccal, nasal, intravenous or intraarterial, or otherwise parenteral.

Injection is accomplished utilizing well known syringes suitable for this purpose. The injections may be done manually, although the use of microprocessor-controlled means for injection are much preferred. Such a device is depicted in Fig. 1. As shown here, the device includes a housing to support a microliter syringe and a connection to microprocessor control means (not shown). Within the housing is a stepper motor operationally linked to the lead screw, which is, in turn, operationally linked to reciprocating means for driving the plunger of the syringe. Operation of the stepper motor is controlled by the microprocessor control means. The device shown in Fig. 1 also includes means for mounting the device upon a stereotaxic from for precise alignment of the syringe with the injection site. The device shown in Fig. 1 is completely programmable for delivery volume, delivery rate, etc.

Numerous utilities of this non-viral, in vivo transformation method are manifest: the method can be used for gene therapy in animals, including humans, without encountering the unknown risks associated with the use of viral vectors. The method successfully transforms quiescent cells in vivo. To Applicants' knowledge, such a transformation has never been demonstrated in vivo using non- viral delivery means. Clearly, by delivering genetic elements to non-replicating cells for ultimate expression of the products encoded by the genetic elements, the invention has utility for the study of gene expression in such cells and also for clinical and therapeutic applications wherein a desired gene product is expressed in situ in a living animal subject.

Having described the invention, the following Examples provide illustrations of the in vivo transformation method in practice. Reference is made to the attached drawing figures. The Examples are included solely to provide a more complete and thorough understanding of the invention disclosed herein. The Examples do not limit the scope of the invention disclosed and claimed herein in any fashion.

EXAMPLES AND COMPARATIVE EXAMPLES The following materials and methods were utilized in the Examples:

Lipid Compounds -

"TfX"-10, "TfX"-20 Brand Transfection Compositions:

"Tfx" reagents were resuspended according to the manufacturer's instructions (Promega Corp. , Madison, Wisconsin). Two μg of HSVlac DNA

(/3-galactosidase gene under the transcriptional control of the HSV IE 4/5 promoter) was ethanol precipitated and resuspended in 1.5 μl of either "Tfx"- 10 or "Tfx "-20. The mixture was incubated at room temperature for 10-15 minutes and then immediately injected. The target charge ratio of DNA:lipid was 1: 1.

Herpes Amplicon Virus - Cell lines:

RR1 cells used for packaging the amplicon constructs are a BHK-derived cell line engineered to stably express the HSV IE3 gene, Paterson and Everett (1990). The NIH 3T3 line (ATCC 1658), as well as the RR1 line, were maintained in Dulbecco's modified Eagles Medium (DMEM) with 10% (v/v) fetal bovine serum (FBS), penicillin (100 μ/ml) and streptomycin (100 μg/ml). Bioactive geneticin (G418; 400 μg/ml, Gibco BRL, Gaithersberg, Maryland) was included in the RR1 medium to maintain selective pressure during routine passage.

Herpes Simplex Amplicon Virus Type 1 Deletion Mutant Production and

Titering:

Amplicon DNA (HSVlac) was packaged into HSV-1 particles by transfecting 5 μg of plasmid DNA into RR1 cells with lipofectamine as recommended by the manufacturer (Gibco BRL). After a twenty-four hour incubation, the monolayer was superinfected with the strain 17, IE3 deletion mutant D30EBA, Johnson et al. (1992); Johnson et al. (1994), at a multiplicity of infection (MOI) of 0.2. Once cytopathic, the monolayer was harvested, taken through a single freeze-thaw cycle and sonicated using a cup sonicator (Misonix Inc. , Farmingdale, New York). Viral supernatants were clarified by centrifugation prior to repassage on RR1 cells. This second viral passage was harvested as above and concentrated overnight by ultracentrifugation in a 25% sucrose/PBS gradient. Viral pellets were resuspended in PBS (Ca²⁺ and Mg²⁺- containing) and stored at -80°C for future use. Amplicon titers were determined by plating 3T3 cells in a 24-well plate at a density of 1 x 10⁵ cells/well followed by infection with dilutions of concentrated viral stocks. After twenty-four hours post-infection, the monolayers were washed twice in PBS, fixed with 4% paraformaldehyde and stained by X-gal histochemistry (5 mM potassium ferricyanide; 5 mM potassium ferrocyanide; 0.02% NP-40; 0.01 % sodium deoxycholic acid; 2 mM MgCl₂ and 1 mg/ml X-gal dissolved in PBS). Blue-forming unit titers (BFUs) were calculated by averaging duplicate wells. HSVlac stocks were titered for helper virus by standard plaque assay methods as described previously, Geschwind et al. (1994).

Adenovirus Deletion Mutant Production and Titering:

A replication-defective human adenovirus serotype 5 (Ad5)-derived adenoviral vector was used for in vivo gene transfer. The replication-defective adenoviral vector (AdCMVlacz) has been deleted of sequences in the El A and

E1B regions which prevents viral replication, Hurwitz and Chinnadural (1985). The AdCMVlacz, hereafter referred to as Adlac, contains the enhancer/promoter of the cytomegalovirus (CMV) and the E. coli-deήved lacz transcription unit with an SV40 polyadenylation signal. High titer adenoviral stocks were generated by amplification in HEK 293 cells and concentrated by CsCl gradient ultracentrifugation. All animals received 1 x 10⁵ p.f.u. (plaque-forming units, diluted in PBS) of Adlac in a volume of 1 μl.

Construction of stereotaxic-mounted, microprocessor-driven syringe pump: World Precision Instruments' Model UMP "UltraMicroPump" brand microsyringe driver, Fig, 1, in conjunction with Model UMC-1 microprocessor- based controller, is designed to deliver picoliter injection volumes using microliter syringes. The syringe is activated by a stepper motor driving a specially designed lead screw with ultrafine pitch. The microprocessor is completely programmable for delivery volume (number of steps per delivery) and rate (number of steps per time interval). These parameters may also be programmed for sample withdrawal. The injector unit was mounted on a precision small animal stereotaxic frame micromanipulator (ASI Instruments, Warren, Michigan) at a 90° angle using a mount for the injector designed by ASI.

EXAMPLE 1 - Intracerebral Injection of Lipid/DNA Reagent: and

COMPARATIVE EXAMPLES 1, 2 and 3 - Intracerebral Injection of Viral Gene Transfer Reagents and Naked DNA:

To demonstrate the utility and efficacy of the invention to transform animal cells in vivo, transformation in mice utilizing two "Tfx "-based transformation reagents and the general protocols described above was compared to transformation utilizing two well-known and extensively-studied viral vectors for mammalian transfection: herpes simplex virus (HSV) and Adenovirus (Ad). The two viral vector transformation methods therefore act as benchmarks against which to compare the efficiency of transformation according to the presently disclosed method. Additionally, the invention was compared to transformation using naked

DNA. The animal model described in the Examples and Comparative Examples was based upon intracranial injections in mice with DNA's encoding /3-galactosidase as the protein of interest.

Microprocessor-controlled and manual injections:

Mice were anesthetized with 3% halothane in 70% N₂0 and 30% O₂ in an induction chamber and maintained at 2% halothane during stereotactic intracerebral injections. After positioning in an ASI murine stereotactic apparatus, the skull was exposed via a midline incision, and burr holes were drilled over the designated coordinates (bregma, +0.5 mm; lateral -2.0 mm; and deep, -3.0 mm).

A 33 GA steel needle was gradually advanced to the desired depth, and 1 μl of "Tfx" reagent/DNA mixture (Example 1) or Formula II reagent/DNA mixture (Example 2), virus (Comparative Examples 1 and 2) or naked DNA (Comparative Example 3) was infused by hand or by a microprocessor-controlled pump over 10 min. The injections delivered by the pump were at a constant rate of 110 nl/min. The needle was removed slowly over an additional 10 minute period.

3-galactosidase histochemistry:

After four days post-injection, mice were anesthetized, a catheter was placed into the left ventricle, and intracardiac perfusion was initiated with 10 ml of heparinized saline (5,000 U/L saline) followed by 20 ml of chilled 4% paraformaldehyde in 0.075 M phosphate buffer (pH = 7.5). Brains were extracted, dehydrated for 48 hrs in 30% sucrose and 40 μm sections were prepared on a freezing sliding microtome (Micron, Wilmington, Delaware). Serial sections were washed in 1 x TBS for 20 min and then transferred to an iron d e v e l o p i n g s o l u t i o n c o n t a i n i n g X - G a l ( 5 - b r o m o - 4-chloro-3-indolyl-/3-D-galactoside) in conventional fashion. Sections were incubated for 40 minutes at 37 °C and washed for 1 hr with lx PBS. Sections were mounted, counterstained with thionin, rinsed, dehydrated in a series of alcohol washes and coverslips were applied and affixed with "CYTOSEAL" brand sealant (VWR, Springfield, New Jersey).

Measurement of β-galactosidase activity:

After four days post-injection, mice were euthanized by cervical dislocation, the brains were immediately harvested and chilled on ice. The brains were placed into a Jacobowitz brain block (Zivic Miller, Allison Park, Pennsylvania) and 1 mm brain slices at the level of the injection site on both the ipsilateral and contralateral sides were prepared for quantification of /3-galactosidase activity. Brain slices were placed in 100 μl of a lysis solution (100 mM potassium phosphate (pH = 7.8), 0.2% "TRITON" X-100, 1 mM DTT, 0.2 mM PMSF and 5 μg/ml leupeptin) and homogenized for 20 seconds followed by centrifugation for 10 min at 12,500 x g at 4°C. The supernatant was removed and heated to 48 °C for 60 minutes to reduce endogenous /3-galactosidase activity. A 10 μl aliquot of the supernatant was added to "GALACTO-LIGHT TM" reaction buffer (Tropix, Bedford, Massachusetts) and incubated for 10 minutes at room temperature. Light emission accelerator (100 μl) was added to each sample and then immediately read in a luminometer (LKB, Germantown, Maryland). All samples were read for 5 seconds.

Results of Examples and Comparative Examples and Discussion Thereof:

EXAMPLE 1 - In Vivo Liposome Gene Transfer:

This Example clearly demonstrates that cationic liposomes containing a Formula I cationic lipid and a fusogenic lipid can be used to deliver DNA constructs which are subsequently expressed, without the need for a viral vector. HSVlac DNA:lipid mixtures of "Tfx"-10 and "Tfx"-20 reagents (Promega) were prepared as described above and injected in a volume of 1.0 μl into the striatum of mice using either manual or microprocessor delivery. Reference is made to Figs. 4 A through 4E, which clearly show the expression of /3-galactosidase in the injected mice. Similar to the observations made with HSVlac and Adlac, below, microprocessor delivery resulted in more localized expression (compare Figs. 4B and 4E) and higher overall levels of /3-galactosidase expression than manual delivery (see Figs 6C & 6D). In addition, gene expression was greater with Tfx-20 than Tfx- 10 at the equivalent charge ratio of DNA: lipid. See Fig. 6F for a comparison of the relative expression of all this Example and the Comparative Examples.

EXAMPLE 2 - In Vivo Liposome Gene Transfer: Example 2 was performed in identical fashion as in Example 1 with the exception that a compound of Formula II wherein R is myristyl was used as the cationic lipid. The results were comparable to those for Example 1 , discussed below.

Comparative Example 1 - HSVlac Transfection:

The plasmid-based HSV amplicon vector, a highly versatile gene transfer platform, is packaged within a HSV virion particle. The HSVlac construct, which places the _/3-galactosidase gene under the transcription control of the HSV IE 4/5 promoter, was packaged, purified and concentrated. One μl of HSVlac, containing 1 x 10⁵ infectious particles was unilaterally delivered to the striatum (coordinates; 0.5 mm anterior of bregma; 2.0 mm lateral of midline; 3.0 mm down from the surface of the skull) of mice using either the standard slow manual delivery method or the microprocessor delivery method. All animals survived the gene transfer procedure and four days later were euthanized. Some brains were processed for X-gal histochemical detection of /3-galactosidase, whereas others were assayed quantitatively for _/3-galactosidase activity using the chemiluminescent substrate "GALACTO-LIGHT" (Tropix).

Figures 2A through 2F are representative tissue sections taken from microprocessor (Figs 2A, 2B, and 2C) and manually (Figs. 2D, 2E, 2F) HSVlac-injected mice. Inspection of the photomicrographs reveals that gene expression is more focal in mice injected using the microprocessor (Figure 2B) than in those done manually (Figure 2E). Substantial expression is noted along the injection path (arrow) in the manually-injected sections; no such injection path staining is observed in the sections from microprocessor injected mice (Fig. 2B).

In addition, quantitation of /3-galactosidase expression (see Fig. 6 A) revealed that microprocessor-injected animals had greater levels of gene product than manually-injected animals. Moreover, inter-animal variability was greater in manually-injected mice, as was the propensity of virus to direct gene expression on the contralateral side (again, see Fig. 6A). This is likely due to some entry of virus into the CSF in the manually injected mice. Transduced cells are observed in ependyma and parenchymal tissues adjacent to the ventricle in manually injected mice. (See Fig. 5A).

Comparative Example 2 - Adenoviral Transfection:

First generation Ad vectors, constructed by recombination of a heterologous transcription unit in place of the EIA gene grow to high titer and have been used for many CNS gene transfer studies. See, for example, Akli et al. (1993), Andersen and Breakefield (1995), Bajocchi et al. (1993), Brody and Crystal (1994). The Adlac vector utilizes the hCMV promoter to drive /3-galactosidase expression. One μl of Adlac, containing 1 x 10⁵ plaque forming units was unilaterally delivered to the striatum of mice using manual or microprocessor methods as described above. After four days, animals were euthanized and brain tissues were analyzed histochemically and enzymatically for

/3-galactosidase expression (Figs. 3 A through 3F and Fig. 6B). As observed in the HSVlac-injected mice of Comparative Example 1, microprocessor delivery as compared to the manual method resulted in more localized expression (compare Figs. 3B and 3D), less needle track reflux (see Fig. 3E); and overall greater gene expression (Fig. 6B). Animals that were manually-injected with Adlac also exhibited transduced cells in ependyma and parenchymal tissues adjacent to the ventricle (see Fig. 5B), similar to animals manually-injected with HSVlac.

Comparative Example 3 - Naked DNA Injection: Plasmid DNA, which has previously been shown to be an ineffective means for achieving gene expression in the CNS, was evaluated using the microprocessor delivery method. Ten μg of HSVlac in a volume of 1.0 μl was injected into the striatum of mice. Four days later these animals were euthanized and /3-galactosidase expression was analyzed. No histochemically detectable cells could be identified after a maximum of four hours in the X-gal reagent. However, minimal expression could be detected using chemiluminescent substrate. Although above the background, the extent of gene expression for naked DNA is quite low, particularly when compared with the other vehicles for gene delivery described above (see Fig. 6E).

Discussion:

Example 1 shows that the method described herein can be used to efficiently transform animal cells, even non-replicating animal cells, in vivo. While the extent of expressed gene product is less than that seen with the conventional herpes simplex and adenovirus vectors, the use of a DNA/lipid transformation reagent obviates many risks which are inherent in the use of viral vectors. There was only three-fold less expression seen in mice injected with the "Tfx "-20 reagent compared to the most efficient vehicle used (the HSVlac virus vector). Both the "Tfx" -10 reagent and the "Tfx "-20 reagent contain the same preferred cationic lipid of Formula I. But, "Tfx "-20 has an increased content of the fusogenic lipid DOPE as compared to "Tfx"-10. This difference likely underlies the greater gene transfer efficiency noted with "Tfx "-20 and suggests that studies directed toward optimizing the content of DOPE will improve the gene transfer efficiency of the present invention.

Example 1 further shows that transformation efficiency can be manipulated by altering the ratio of the fusogenic lipid and the cationic lipid of Formula I.

This demonstrates the breadth of the invention in that the content of the lipid compositions for delivery of the DNA or RNA construct can be optimized to suit a particular purpose, application, injection site, host animal, etc. without adversely effecting its gene delivery properties. The efficiency of gene transfer using the present invention is particularly notable. While the Examples contained herein showed that lipid transformation was three-fold less than viral transfection, it must be remembered that transfection utilizing HSV and adenovirus vectors has been extensively utilized and optimized to increase transformation efficiency. In contrast, in vivo transfection utilizing lipid vectors has not been subjected to such extensive optimization. By altering the ratio of fusogenic lipid to cationic lipid, as noted above, as well as by optimizing the DNA/lipid charge, it is believed that the efficiency of the method described herein can be optimized to match or even to exceed the transformation efficiency of the HSV and adenovirus vectors. Another advantage of the present invention is its distinct promise for effective long-term transformation and its use in chronic gene therapy. As briefly noted above, the antigenicity of viral vectors makes them extremely ill-suited for chronic gene therapy. While a first in vivo treatment using a viral vector may be highly effective in introducing exogenous polynucleotides, the efficacy of subsequent treatments drops precipitously as the host organism mounts an immuno-response to the antigens presented by the vector. Subsequent injections yield no response because the host organism now has a high titer of anti-viral antibodies. As soon at the vector injection is made, the virus particles are rapidly inactivated by the host's immune response. Therefore, viral vectors cannot be used to treat ailments which require chronic gene therapy.

In contrast, the extremely low antigenicity of the lipid transformation compositions as compared to viral vectors eliminates or greatly ameliorates this problem. Because the lipid compositions are much less antigenic than viral vectors, the animal host is far less likely to develop an immune response to the DNA/lipid injection. Therefore, the transformation compositions as described herein can be injected repeatedly without an immediate loss in transformation efficiency due to a powerful immune response by the host animal. This represents a vast improvement over prior art in vivo transfection methods.

Another advantage of the present invention is its ability to transform quiescent cells. The Examples presented above were conducted on mice CNS cells, cells which do not replicate. By affording transformation of quiescent cells, the invention allows gene therapy to be utilized against a host of afflictions of the central nervous system.

BIBLIOGRAPHY

Akli, S., Caiilaud, C , Vigne, E. , Stratford, Perricaudet Ld. , Poenary, L. , Perricaudet, M. , Kahn, A. , Peschanski, M.R. , (1993) Transfer of a foreign gene into the brain using adenovirus vectors, Nat Genet, 3: 224-228.

Andersen, Land Breakefield, X., (1995) in Somatic Gene Therapy J. Wolfe, Ed. (CRC Press, Inc.) pp. 135-160.

Bajocchi, G. , Feldman, S. , Crystal, R. and Mastrangeli, A. (1993) Direct in vivo gene transfer to ependymal cells in the central nervous system using recombinant adenovirus vectors, Nat Genet, 3: 229-234.

Bergold, P., Casaccia-Bonnefil, P. , Federoff, H. and Stelzer, A. , (1993) Transduction of C A3 neurons by a herpes virus vector containing a GluRό subunit of the kainate receptor induces epileptiform discharge, Soc Neuro Abstr,

Washington, DC.

Brody, Sand Crystal, R. (1994) Adenovirus-mediated in vivo gene transfer, Ann NY Acad Sci, 716: 90-101.

Brooks, A. , Muhkerjee, B. , Panahian, N., Cory-Slechta, D. and Federoff, H. (1997) Nerve growth factor somatic mosaicism produced by herpes virus-directed expression of ere recombinase, Nature Biotechnology, 15: 57-62. Davidson, B. , Allen, E. , Kozarsky, K. , Wilson, K. and Roessler, B.

(1993) A model system for in vivo gene transfer into the central nervous system using an adenoviral vector, Nat Genet, 3: 219-223.

During, M. , Naegele, J., O'Malley, K. and Geller, A. (1994) Long-term behavioral recovery in parkinsonian rats by an HSV vector expressing tyrosine hydroxylase, Science, 266: 1399-1403.

Federoff, H. , (I 995) in Viral Vectors. (Academic Press, Inc., New York, pp. 109-117.

Federoff, H., Geschwind, M., Geller, A. and Kessler, J. (1992) Expression of nerve growth factor in vivo, from a defective HSV-1 vector prevents effects of axotomy on sympathetic ganglia, Proc Natl Acad Sci USA, 89: 1636-1640. Geller, A. and Breakefield, X. (1988) A defective HSV-1 vector expresses escherichia coli beta-galactosidase in cultured peripheral neurons, Science, 241: 1667-1669.

Geller, A. , During, M. , Haycock, J. , Freese, A. and Neve, R. (1993) Long-term increases in neurotransmitter release from neuronal cells expressing a constitutively active adenylate cyclase from a herpes simplex virus type 1 vector. , Proc Natl Acad Sci U S A, 90: 7603-7607. Geschwind, M. , Hartnick, C.J. , Liu, W. , Amat, J. , Van De Water, T.R. , Federoff, H.J., (1996) Defective HSV-1 vector expressing BDNF in auditory ganglia elicits neurite outgrowth: model for treatment of neuron loss following cochlear degeneration, Hum Gene Ther, 7: 173-182.

Geschwind, M. , Lu, B. and Federoff, H. (1994) in Providing pharmacological access to the brain; A volume of Methods in Neurosciences P. Conn, Ed. (Academic Press, Inc. , Orlando, Florida, vol. 21 , pp. 462-482. Glorioso, J., Goins, W. , Meaney, C , Fink, D. and DeLuca, N. (1994) Gene transfer to brain using herpes simplex virus vectors, Ann Neurol, Suppl 35: S28-S34.

Ho, D. , Fink, S.L. , Lawrence, M.S., Meier, T.J. , Saydam, T.C. , Dash, R. , Sapolsky, R.M. , (1995) Herpes simplex virus vector system: analysis of its in vivo and in vitro cytopathic effects, J Neurosci Methods, Accepted for publication:

Hurwitz, D. and Chinnadurai, G. (1985) Evidence that a second tumor antigen coded by adenovirus early gene region Ela is required for efficient cell transformation, Proc. natn. Acad. Sci. , 82: 163-167.

Jin, B.K. , Belloni, M. , Conti, B., Federoff, H.J. , Starr, R., Son, J.H., Baker, H. , Joh, T.H., (1996) Long-term in vivo gene expression driven by a tyrosine hydroxylase promoter in a defective herpes simplex virus amplicon vector. , In press, 7: 2015-2024.

Johnson, P. , Miyanohara, A., Levine, F. , Cahill, T. and Friedmann, T. (1992) Cytotoxicity of a replication-defective mutant of herpes simplex virus type 1, J Virol, 66: 2952-2965.

Johnson, P. , Wang, M. and Friedmann, T. (1994) Improved cell survival by the reduction of immediate-early gene expression in replication-defective mutants of herpes simplex virus tupe 1 but not by mutation of the virion host shutoff function, J Virol, 68: 6347-6362.

Kaplitt, M., Leone, P. , Samulski, R.J., Xiao, X. , Pfaff, D.W. , O'malley, K.L. , During, M.J. , (1994) Long-term gene expression and phenotypic correction using adeno-associated virus vectors in the mammalian brain, Nat Genet, 8: 148-154.

Linnik, M., Zahos, P. , Geschwind, M. and Federoff, H. (1995) Expression of bcl-2 from a defective herpes simplex virus- 1 vector limits neuronal death in focal cerebral ischemia, Stroke, 26: 1670-1675.

Marshall, E. (1995), Science, 269: 1052-53. Paterson, T.and Everett, R. (1990) A prominent serene-rich region in Vmwl75, the major transcriptional regulator protein of herpes simplex virus type 1 , is not essential for virus growth in tissue culture, J Gen Virol, 71: 1775-1783.

Claims

CLAIMSWhat is claimed is:

1. A composition for use in in vivo transfection of central nervous system cells in mammals comprising, in combination: a non-ionic, fusogenic lipid; and a cationic lipid selected from the group consisting of

and

wherein R is a fatty acid alkyl or alkenyl group and X is an anion.

2. The composition according to Claim 1, wherein R is oleyl or myristyl.

3. The composition according to Claim 1 or Claim 2, wherein the non-ionic, fusogenic lipid is dioleylphosphatidylethanolamme.

4. The composition according to any one of Claims 1, 2, or 3, wherein X is iodide.

5. The composition according to any one of the preceding claims wherein the cationic lipid is N,N,N',N'-tetramethyl-N,N'-bis-(2-hydroxyethyl)-2,3- di(oleyloxy)-l,4-butanediammonium iodide or N,N-[bis(2-hydroxyethyl)- N-methyl-N-(2,3-di(tetradecanoyloxy)propyl] ammonium iodide.

6. The composition according to any one of the preceding claims further comprising, in combination, DNA or RNA to be transfected into the mammal.

7. Use of a composition as recited in any one of the preceding claims in the manufacture of a medicament for the in vivo transfection of mammalian central nervous system cells.