CA2223923A1 - Cationic lipid:dna complexes for gene targeting - Google Patents

Abstract

The invention herein describes pharmaceutical compositions and methods for targeted delivery of functional genes into cells and tissues in vivo. The invention discloses DNA:lipid complexes, methods of making such complexes and methods of using such complexes for facilitating the targeted delivery and entry of recombinant expression constructs into cells and tissues in vivo, and particularly delivery of such recombinant expression constructs into cells and tissues in vivo, and particularly delivery of such recombinant expression constructs by intravenous, intraperitoneal or direct injection. The delivery vehicle for the targeting of the recombinant construct encoding the gene of interest, is composed of a mixture of a cationic imidazolinium lipid of formula (I) wherein R and R1 mean a C11-29 aliphatic hydrocarbyl group, together with neutral lipids (e.g. DOPE, cholesterol).

Description

CA 02223923 1997-12-0~
W O 96/40962 PCT~US96/09526 CATIONIC LIPID:DNA COMPT,F.~,.S FOR GENE TARGETING
BAC~GRO~ND OF T~lF, I1~1V~,l~TION
1. Field of the I~
A pe~ ial goal in the ph,.. ~ rological arts has been the development of mrtho~ and compositions to facilitate the specific delivery of theldpeuLic and other agents to the ~piO~lial~ cells and tissues that would benefit from such L~ ...r.~l, and the avoidance of the general physiological effects of the i~ lol,lia~ delivery of such agents to other cells or tissues of the body. Recently, the advent of Lecolllbil~lL DNA
technology and genetic ~ g has provided the pharmacological arts with a wide new spectrum of agents that are functional genes carried in lccollll)il1allL ~ ion constructs capable of m.ofli~ting ~.ies~ion of these genes in host cells. These develc~,lRnL~ have carried the prolllise of "molecular medicine," specifically gene therapy, wL~lel)y a defective gene could be replaced by an exogenous copy of itscognate, functional gene, thereby alleviating a variety of genetic ~ e~rs.
However, the greatest drawback to the achievement of effective gene therapy has been the limited ability in the art to introduce recombinant ~lc;,sion constructs encoding functional genes into cells and tissues in vivo. While it has been recognized in the art as being desirable to increase the efficiency and specificiy of ~-imini~tration of gene therapy agents to the cells of the relevant tissues, the goal of specific delivery has not been achieved in the prior art.
Liposomes have been used to attempt cell Ldl~Lillg. Rahman et al., 1982, Life Sci. ~1: 2061-71 found that liposomes which cont~in~-d galactolipid as part of the lipid appeared to have a higher affinity for ~dl~,nch~ al cells than liposomes which lacked galactolipid. To date, however, efficient or specific delivery has not been predictably achieved using drug-enr~ps~ od liposomes. There remains a need for the development of a cell- or tissue-targeting delivery system.
Thus there remains in the art a need for rn~othn~s and reagents for achieving cell and tissue-specific targeting of gene therapy agents, particularly recombinant ~I,ression constructs encoding functional genes, in vivo.

BRIEF SUMMARY OF THE INVENTION
The present invention is directed to improved methods for targeted delivery of functional genes to cells and tissues in vivo. This delivery system achieves such specific delivery by the formation of DNA:lipid complexes between nucleic acid cc,~ g a recolllbi l~ expression construct e~rotlin~ a functional gene or fragment thereof complexed with a mixture of a cationic lipid and a neutral lipid. Methods of use are also provided. This invention has the specific advantage of targeted delivery of functional genes into cells in vivo, achieving effective intracellular delivery of constructs encoding functional genes more efficiently and with more specificity than conventional delivery systems.
In a first embo-lim~nt the invention provides a ~hA~ r~,~lir~l composition, COlll~li..illg a form~ tinn of a soluble CU111~1CA of a recombinant c~ es~iOn construct and a l~ Lulc of a neutral lipid and a cationic lipid in a ph~ reutir~lly acceptable carrier suitable for ~ l . dLion to an animal by injection. In this ellll)odilllclll of the invention, the recolllbillallt exl,lcssion construct comprises a nucleic acid encoding a transcription product, the nucleic acid being o~e/dlively linked to gene expression regulatory el~mrnt~ and whereby the nucleic acid is capable of ll~nsc.i~Lion in vivo.
As used herein, the term "llallsc~i~Lion product" is int~n-1-od to encompass an RNA
product resulting from Llallsclil,Lion of a nucleic acid seq~lenre, and explicitly includes RNA sequences that are not Llallsclil)ed into protein (such as ~nticen~e RNAs orribozymes), as well as RNAs that are snhseqll~ntly tr~n~l~trd into polypeptides or pl.,~eills.
In this first embodiment, the cationic lipid is a nitrogen-cont~ining, 2û imidazolinium-derived cationic lipid having the formula:

>

~R

OCR
-CA 02223923 1997-12-0~

W O 96/40962 PCT/U'~6J'~3S26 wherein each of R and R~ i"~ Pntly is a straight-chain, aliphatic hydrocarbyl group of 11 to 29 carbon atoms inclusive. Preferred are those cations wherein each of R and Rl independently have from 13 to 23 carbon atoms inclusive. In particularly pl~rcllcd embodiments, the cationic lipid is 1-(2-(oleoyloxy)ethyl)-2-oleyl-3-(2-~ 5 lly~l~o~ycLllyl)imidazoliniUm chloride. In additional ,ulcÇcllcd embodiments, the neutral lipid is cholesterol, and the 1-(2-(oleoyloxy)ethyl)-2-oleyl-3-(2-hydroxyethyl)imidazolinium chloride and cholesterol are present in the complex at a molar ratio of 1:1. Further p.~r~ d emb~limPnt~ COIll~)liSc a l~col,,bil,dl,L c~ ,SsiOll construct encoding human CFTR and a mixture of a neutral lipid and a cationic lipid having a ratio of DNA to lipid of from about 1:6 to about 1:15 (~gDNA:nmoles lipid).
Particularly p-crc~cd are embodiments where the DNA COIll~)liSillg the recombinant expression construct is present in the complex at a concentration of about 0.5 to lmg/mL. In further ~lcr~llcd embodiments, the cationic lipid is 1-(2-(oleoyloxy)ethyl)-2-oleyl-3-(2-hydroxyethyl)imidazolinium chloride and the neutral lipid is dioleoylphosphatidyl ethanolamine, and the 1-(2-(oleoyloxy)ethyl)-2-oleyl-3-(2-hydroxyethyl)imidazolinium chloride and dioleoylphosphatidyl ethanolamine are present in the complex at a ratio of 1:1. Further plcrclled cmbo-limPnt~ comprise a recol~ ,d"L expression cul~l~u~;l and a mixture of a neutral lipid and a cationic lipid having a ratio of DNA to lipid of about 1:1 (~gDNA:nmoles lipid). Particularly L"cre.,cd are embollimPntc where the DNA CUIll~li.. illg the l- colllbill~-L expression construct is present in the formulation at a collcellLldtion of about 0.5 to 5mg/mL.
In a second embodiment, the invention provides methods for introducing a reco.llbi,~lL expression construct into a cell cullllulisi"g lung tissue in an animal, the method COlllpli~.illg the step of ~ h~g the pharm~re-ltir~l composition of the invention to the animal by intravenous injection. In ~lcr~,.lcd embo~imPnt~, thecationic lipid is 1-(2-(oleoyloxy)ethyl)-2-oleyl-3-(2-hydroxyethyl)imidazoliniumchloride. In additional preferred embodiments, the neutral lipid is cholesterol, and the 1-(2-(oleoyloxy)ethyl)-2-oleyl-3-(2-hydroxyethyl)imidazolinium chloride and cholesterol are present in the complex at a molar ratio of 1:1. Further L~cr~.~cd ~ 30 emborlimrntc comprise a recombinant expression construct and a mixture of a neutral lipid and a cationic lipid having a ratio of DNA to lipid of from about 1:6 to about CA 02223923 1997-12-0~
W O 96/40962 PCTAU53G~5~>6 1:15 (~gDNA:nmoles lipid). Particularly plc~-~d are embo~iim~ntc where the DNA
comprising the recombinant expression construct is present in the formulation at a concellLl~,Lion of about 0.5-lmg/mL.
In another aspect of the second embodiment of the invention is provided methods for introducing a recul~lbil~llL c~Lyl~ssion construct into a cell culll~ hlg spleen tissue in an animal, the method COll~liSillg the step of ~ illg the ph~rm~rentir~l composition of the invention to the animal by intravenous injection.
In plcr~ d embo~limPnt~, the cationic lipid is 1-(2-(oleoyloxy)ethyl)-2-oleyl-3-(2-hyd,~cycLllyl)imidazolinium chloride. In additional plcfe.lcd embo-limt~ntc, theneutral lipid is dioleoylphr,sph:~ti-lyl ethanolamine, and the cationic lipid and the neutral lipid are present in a molar ratio of 1:1. Further plcL~.led embc!-~im~nt~
co~ lise a recombinant expression construct and a mixture of a neutral lipid and a cationic lipid having a ratio of DNA to lipid of about 1:1 (~gDNA:nmoles lipid).Particularly plcl~lcd are embodiments where the DNA colll~lisillg the lcculllbh~allL
expression col~LlucL is present in the formulation at a conce~lLldLion of about 1-2.5mg/mL.
In further embo-lim~nt~ of this aspect of the invention, the DNA:lipid complex is L~lgclt:d to ~.iLoneal macrophages by ~ L.~lion by hlll~c~iLoneal injection.
In these embodiments, the cationic lipid is 1-(2-(oleoyloxy)ethyl)-2-oleyl-3-(2-hydroxyethyl)imidazolinium chloride, the neutral lipid is cholesterol, the cationic lipid and the neutral lipid are present in a molar ratio of about 1:1, the complex of a reculllbil~lL e,~rci,~ion construct and a mixture of a neutral lipid and a cationic lipid has a ratio of DNA to lipid of about 1:1 (,ugDNA:nmoles lipid), and the DNA
cunce~lL.~Lion in the DNA:lipid formulation is about 1-2.5mg/mL. In additional embo~lim~nt~ of this aspect of the invention, the DNA:lipid complex is targeted to spleen macrophages and ~ d by intraperitoneal injection. In these embodiments, the cationic lipid is 1-(2-(oleoyloxy)ethyl)-2-oleyl-3-(2-hydroxyethyl)imidazolinium chloride, the neutral lipid is cholesterol, the cationic lipid and the neutral lipid are present in a molar ratio of about 1:1, the complex of a recombinant c~lcssioll col~LlLIcL and a mixture of a neutral lipid and a cationic lipid has a ratio of DNA to lipid of about 1:1 (,ugDNA:nmoles lipid), the DNA

W O 96/40962 PCT/U'3~ 3-.~6 conccllLl~tion in the DNA:lipid form~ tiQn is about 1 to 2.5mg/mL.
In this aspect, the invention also provides methods for targeting gene transfer into pancreatic tissue by il,Lldp~.;Lol1eal injection. In ~.ef~,led emb~1imPntc, the cationic lipid is 1-(2-(oleoyloxy)ethyl)-2-oleyl-3-(2-hydroxyethyl)imidazoliniumchloride, the neutral lipid is dioleoyl~ho~.~h~Lidyl ethanolamine, the cationic lipid and the neutral lipid are present in a molar ratio of about 1:1, the complex of a l~,cul~ illdllL C~ sioll cu~ u~;L and a mixture of a neutral lipid and a cationic lipid has a ratio of DNA to lipid of about 1:1 (ugDNA:nmoles lipid), and the DNA
concentration in the DNA:lipid form~ tion is about 1.5 to about 2.5mg/mL.
The invention also provides a method of introducing a rec-~lllbJl~lL expression construct into a cell colll~li..illg a tissue in an animal, the method COlllpli.illg the step of ~I...;..icl~,i"g the ph~rm~renti~l composition of the invention to the animal by direct injl~ctinn. In l,lcrclled embo~lim~nts, the cationic lipid is 1-(2-(oleoyloxy)ethyl)-2-oleyl-3-(2-hydroxyethyl);~ 7olinium chloride and the neutral lipid is cholesterol.
Also pl~,rcll~ d are ll~i~LLulcS of the cationic lipid and the neutral lipid in a molar ratio of about 1:1. Pl~L,l~d complexes include a complex of a reconlbillallL expression construct and a mixture of a neutral lipid and a cationic lipid h-aving a ratio of DNA
to lipid of about 1:1 (,ugDNA:nmoles lipid). The preferred DNA cullce,lll~tion in the DNA:lipid formlll~tinn is about 1-2.5mg/mL in this embodiment of the invention.
Specific plcrc-lcd embo~1imlont.c of the present invention will become evident from the following more detailed des~ lion of certain IJ-cfe,-cd embo-lim-ontc and the claims.

RRTF~ D~ CRTPTION OF T~ T)R~wINGs Figure 1 is a graph showing the stability of DNA:lipid complexes of the invention assayed by intravenous ~.l...;..i.~l.~tion and lung CAT assays over a period of 11 weeks.
Figure 2 is a graph of a cullll)alison of chloride efflux in the presence and absence of stimuli in cells tr;~ncf~-ct~d with human CFTR-encoding plasmid vectors ~ 30 complexed with EDMPC:cholesterol.
Figure 3 is a sch~ le~ sclllation of the plasmid p4119.

CA 02223923 1997-12-0~
W O 96/40962 PCTAJS96/'~5.'.>6 Figure 4 is a histogram showing that mice ~ f led CAT-encoding plasmids complexed with DOTIM:cholesterol liposomes exhibited CAT gene expression in the lung.
Figure S lcl~le3c~ . autoradiograms of Southern blot hyblidi~dtion assays 5performed on mouse tissues obtained after hlLlavc~lous ~ - dLion of CAT-encoding DNA:lipid complexes of the invention.
Figure 6 is a histogram showing ~-g~ to~ e ~A~lession in mouse lung in 9 eAI,~,.i...~.lL~I mice ~.I...;..i~l~ red ,~-gal~rtosi-~e-encoding DNA:lipid complexes of the invention.
Figure 7 is a sr'n~m~ti~ re~l.,.. c.lLdlioll of the plasmid pMBl9.
Figure 8 is a graph showing long-term, pel~ cll~ cA~ ssion of CAT activity in mouse lung obtained after intravenous ~ l ~ c,Lion of CAT-encoding DNA: lipidcomplexes of the invention.
Figure 9 is a histogram showing CAT gene cA~lcssion in brain after CAT-15encoding DNA:lipid complexes of the invention were ~ ed intracranially.
Figure 10 is a l~i~.Lo~sl~ll colll~,alillg tissue specificity of CAT gene expression in DNA:lipid complexes ~rlmini.ctered intravenously (samples 1 and 2) or illLld~c.iloneally (samples 4 and 5).
Figure 11 is a histogram showing formulation-dependent variability in the 20extent of spleen expression of CAT after intravenous ~ dLion of DNA:lipid complexes of the invention.
Figure 12 is a histogram showing human HLA antigen cA~rcssion in bone ~lldllUW, spleen and Iymph nodes following hlLld~ ous ~ .aLion of various formulations of DNA:lipid complexes of the invention.
25Figure 13 is a IC~JICSC ' 1l;l~ i~n of tissue-specific targeting of CAT-encoding DNA
complexed with different liposome complexes and ~-1mini.ct-ored intravenously.
Figure 14 is a histogram illustrating CAT gene Lal~e~ g to peritoneal macrophages after intraperitoneal injection.
Figure 15 is a histogram showing macrophage-specific targeting by 30~ L. dLion of CAT-encoding DNA using DNA:lipid complexes of the invention.
Figure 16 is a histogram showing pancreas-specific tdl~cLiilg by ~ , dLion W O 96/40962 PCT/U'.,G~'~3~'6 of CAT-encoding DNA using DNA:lipid complexes of the invention.
Figure 17 is a histogram showing spleen-specific targeting by a~hllh,i..Llation of CAT-encoding DNA using DNA:lipid complexes of the invention.
Figure 18 is a lC~llC:~e' ~l i l inn of tissue-specific ~l~,Lillg of CAT-encoding DNA
S complexed with different liposome complexes and ~ ;ni!~lelcd i"L,dl)el;Lclleally.
Figure 19 is a nistogram showing CAT gene expression in human prostate tissue in which CAT-encoding DNA using DNA:lipid complexes of the invention weredirectly ~rl~";~ .cd ex corpora.
Figure 20 is a hi~.~o~lalll showing a comparison of spleen-specific and lung-specific targeting of DNA:lipid complexes of the invention using intravenous andhltlap.,~iLoneal routes of a-1mini~rration.
Figure 21 shows the results of RT-PCR of transfected lung tissuse sections showing ~lallsgelle seqllenres specifically targeted to vascular endothelial cells.

DF,TAIT,F,~ D~,.5CRTFrrION OF T~F,p~F,FF,RRF,n F~M RODrn~TS
The present invention provides compositions of matter and methods for facilit~ting the entry into cells of nucleic acids, particularly recombinant c~ics~.ion constructs encoding functional genes. For the purposes of this invention, the term "recombinant c~ ,..sion construct" is intended to encompass a replicable DNA
construct comprising a nucleic acid encoding a functional gene or fragment thereof, operably linked to suitable control seqlltonres capable of effecting the expression of the gene in a suitable host cell. ~xpressly intrnr~.-d to fall within the r~rfinition of a "gene"
are embo.limrntc comprising cDNA and genomic DNA sequ~-nrrc of functional genes,as well as chimeric hybrids thereof. Also int~n~l.od to fall within the scope of the lecolllbi,~ expression constructs of the invention are fragments or mutants of such genes which, when c~,cssed, may inhibit or suppress the function of an endogenous gene in a cell, including, inter alia, trans-d-)min~nt mutants, :~nti~rnse gene fragmPn and ribozymes.
In the reculllbinan~ expression constructs as provided by the present invention,~ 30 the need for such control sequences will vary depending upon the host and cell types selected and the llal~rull~ ion method chosen. Generally, control sequences include , CA 02223923 1997-12-0~

a transcriptional promoter, optional or ancillary l~ s~ Lion control seqllenrçs, such as ~ldnsclip~ion factor binding domains, enhancer sequences, and other eukaryotic "operator" sequences to control llalls~ ,Lion, a sequen-~e ~nro~ling suitable mRNA
libosolllal binding sites, and seq~l~nres which control the Lcllllil~tion of l~dnsc.i~Lion and translation. See, Sambrook et al., 1990, Moler~ r Clor~ A T ~ho~dto~y al (Cold Spring Harbor Press: New York).
Vectors useful for practicing the present invention include plasmids, viruses (inr!llfling phage), retroviruses, and i--l~;~ldlible DNA fr~gmPnt~ (i.e., fr~gm~on i..Leg.d~il)le into the host genome by homologous or non-homologous .~,combi..dlion).
Also useful are vectors which replicate ~ulollo~llously in host cells. Suitable vectors will contain replicon and control se~ onres which are derived from species compatible with the intPn-led ex~ ion host cell.
The recombinant ~ ;,sion c~ llu~ of the present invention are useful in gene therapy, and s~ecircally, for delivering exogenous, functional copies of a defective gene to a specific tissue target in vivo. See generally Thomas & Capecchi, 1987, Cell 51: 503-512; Bertling, 1987, Bioscience Reportsl: 107-112; ~Smithi~os et al., 1985, Nature 317: 230-234.
The invention provides complexes of recombinant DNA constructs encoding functional genes or fr~gm~nt~ thereof and also cOll~ ing a mixture of a cationic lipid and a neutral lipid. For the purposes of this invention, the term "cationic lipid" is int~n-led to encompass lipids which are positively charged at physiological pH, and more particularly, col~LiLuLi~ely positively charged lipids COlllpLi~illg, for example, a ~ludte.nary ammonium salt moiety.

WO 96/40962 PCT/U',G~ J~6 Specifir~lly, the invention provides lliLIogell~ont~ining, imidazolinium-derivedcationic lipids having the formula:
OH
~

[~R

OOCR~
wherein each of R and Rl independently is a straight-chain, aliphatic hydrocarbyl group of 11 to 29 carbon atoms inclusive. P'ler~-led are those cations wh~ill each of R and Rl inde~elldelllly have from 13 to 23 carbon atoms inclusive. The R and Ri groups are sdluldlt:d or are unsaturated having one or more ethylenically unsaturated linkages and are suitably the same or are dirrelc;lll from each other. Illustrative R
groups include lauroyl, lllyl ;~oyl, palmitoyl, stearoyl, linoleoyl, eicosanoyl,tricosanoyl and non~ros~n~ yl. In plc~ d embo~iimlont~ the cationic lipid is 1-(2-(oleoyloxy)ethyl)-2-oleyl-3-(2-hydroxyethyl)imidazolinium chloride (abbreviated as DOTIM herein).
The cationic lipids cclll~lisillg the liposome formulations of the invention canbe synth~si7~od by a l~allangt;ll~lll reaction. This reaction comprises ~ylllhesi~ of DOTIM from N,N-bis(2-hydroxyethyl)ethylene ~ min~o through an amino-~lotecl~d diacylated intPrmP~ tr to the desired product. The method in general involves synthesis of an imidazolinium ion by heating a precursor compound of formula:

W O 96/40962 PCTAJS9G~03~6 RCOO

R,coo~ NH

in an organic solvent at a Lc~ ueldLu,c above the boiling point of water, wherein each of R and R~ independently Icl~lescnl~ an organic group such that the plCCUI:~Ol compound is soluble in the solvent and the R and Rl are stable against reaction in the solvent at the tclllllclaLul~. Specifically, imidizolinium-col~ illg cationic lipids of the invention are plc~al~d according to the following reaction scheme:

W O 96/40962 PCT~US96/09526 HO ~ HN X' HO ~ N
~ >

HO ~ NH HO ~ N ~ X

RCOZ

H.HY
RCOC = N HY RCOC ~ N
<
R~COO ~ NH.HY R1COO ~ N

heat OH

N~
R OH~

OCR

CA 02223923 lss7-l2-oj wo 96/40962 PCT/US96/09526 In this reaction scheme, X is any amino pl~ lhlg group that reacts preferably with and p-oLe-;L~ a secollddly amino group in the ~lcs~nce of a hydroxyl group, preferably one that is removable by acid hydrolysis (e.g., with a strong acid such as HCL); X' is the precursor of the X proLe~Lillg group (e.g., X' is an anhydride or acid chloride S where X is an acyl group); RCOZ is an acid chloride (Z is Cl) or anhydride (Z is RCOO) in which R is defined as either R or R~; and HY is a strong acid (e.g., sulfuric acid or a derivative thereof or a hydrogen halide). A pl~l;..~,d amino pluLe~_Lillg group is t-butyloxycarbonyl (from di-t-b-llyl~Jylocal'L~olldL~). Preferred acylating groups are acid chlorides of fatty acids (such as the fatty acid sllhstihlPn~c of the imidazolinium herein described). A ~ler._.. ed acid for the l~,allang~ enL and deploLecLion steps of the pltydldLi~fe scheme disclosed above (and which can be combined in a single step) is HCL. Heat for the rearrangement reaction is preferably provided by reflux in a solvent having a boiling point in the range of 100~ to 200~C, more preferably in the range of 100~-150~C. The initial imidizolinium ion is formed as a hydroxide saltand/or a chloride salt (if HCL is used as the strong acid), but the anion can be replaced by an exchange reaction. This specific reaction scheme is shown below:

W O 96/40962 PCTtUS96tO9526 BOC
HO~ N HO ~N
> BOC20 >
HO~ NH H~~~N~ BO

RCOCI/NEt3 BOC
H HCI
RCOO~\ \ RCOO ~ \

' RCOO~ NH.HCI RCOO~N B
4 3 C~

heat OH

[~ R OH~3 N
., ~

OOCR

CA 02223923 1997-12-0~
W 096/40962 PCT~US96~'~5-.26 This synthetic scheme is not limited to the explicitly-disclosed i~mida_olinium compounds COlll~ illg the fo~.~n~ tic.n~ of the invention. This reaction scheme provides a general protocol for the production of imi~3~7~-1inium co---~,ou--ds of formula: e o ~3 ~N3 ~ '~X
N~

~

in which X~ ;s. .-ls the residue of an acyl group after the .~:a..~nge..~e..~ reaction as shown (from H to complex organic group) while ~. and X3 independently lc~ sellL
H or an organic group. X. would initially .cp.esenL R-CO-, but this group could be removed or be replaced by a different organic group using standard rht?mir~l reactions; since one of the two potential hydroxyl groups in the initial product is already protected, synthesis of compounds in which X~ and X3, represent dirr~clll groups can readily be accomplished. Ions in which both X, and X3 l~lcsenl H are pl~ft.led, as these can be used in the synthesis of numerous imi~7O1inium compounds. ~lthollgh there is no particular limit on the structure of the three "X"
groups in the general synthesis other than those imposed by solubility or reactivity under the heating conditions being used for the reaction (which will be readily apparent), plere-.~:d organic groups are hydrocarbyl groups con~ining 30 or fewer carbons and their oxygenated products (especially fatty acids and their reactionproducts as previously described, as well as other hydloca~'Lyl groups and oxygenated products cont~ining 15 or fewer carbon atoms, preferably 10 or fewer, more CA 02223923 1997-12-0~

W O 96/40962 PCT~U5~C/'~35~6 preferably hydrocarbyl groups co~ i..g no more than one phenyl ring with the r~m~in-l~r of the hydLocd-lJyl group being composed of alkyl groups, especially alkyl groups of 5 or fewer carbons). Organic groups formed oxygenated hydrocarbyl groups are ~.~LLdbly carboxylic acids, alcohols, esters, ethers, ketones and aldehydes cont~ining no more than one such functional group per organic group. Examples ofimi~l~701inium ions that can be ~ Jaled by the ~yllLllcsis as described above (with further mo(lific~ti~ n of the hydroxyl groups using simple organic reactions) include 1,3-dihydroxyethylim~ olinium, l-methoxyethyl-3-hydrox.yeLllyl;---i~701inium, 1-hydroxyethyl-2-phenyl-3-methylcarboxyethylimi~1~701inium, 1,3-dimethoxyethoxy-ethylimi-1~701inium, 1,3-hydroxyethyl-2-tridecylimil1~7~1inium, and 1-hydroxyethyl-2-cis, cis-8,11-heptadecyldienyl-3-oleoylo~y~Lhyl;---i-l~7c-linium.
Since the reaction is a simple self co~ .C~ion reaction with the elimin~tion of water, the solvent and/or other reaction conditions are not important to the overall reaction. Any solvent can be used that will dissolve the pl~;Ul:~Ol colllL)oulld and that has a boiling point above that of water (under the pressure conditions of the reaction, which are not limiting). If an acid catalyst is used to speed up the reaction, a protic solvent is ~.rer~ d in order to provide easier proton çxrh~nge. Ethylene glycol and other alcohols having a boiling point above 100 C are pl~r~..ed.
One of the explicitly-disclosed cationic lipids of the invention (termed DOTIM) is also col-,.. e.cially available (Avanti Polar Lipids, Al~h~m~).
Cationic lipids are particularly useful as carriers for anionic compounds, particularly polyanionic mac-~,-.-olecules such as nucleic acids. As cationic lipids are positively charged, a tight charge complex can be formed between a cationic lipid carrier and a polyanionic nucleic acid, resulting in a lipid carrier-nucleic acid complex which can be used directly for systemic delivery to a m~mm~l or m~mm~ n cell.
Neutral lipids, in contrast to the cationic lipids of the invention, are characterized as being electroch~mir~lly neutral, although this definition does not preclude protonation of such lipids to produce a positively-charged salt under certain conditions. Expressly included within this definition are, inter alia, steroids such as cholesterol and dioleoylphosphatidyl ethanolamine (DOPE).
Complexes of DNA and llli~Lulc:s of cationic and neutral lipids of the invention CA 02223923 1997-12-0~ .
W 096/40962 PCT/U~9~U33J6 are characterized by a number of pala~ . intrinsic to the formation of such complexes. These include the identity of the cationic lipid and the neutral lipid; the ratio of cationic lipid to neutral lipid; con~ n of DNA in the complex; the ratio of DNA to lipid; DNA purity; cationic liposome size; m~tho~ic of pl~alillg liposomes;
the methods of p~ alillg the DNA:lipid complexes; and other variâbles. Preferredcolllbhlations of cationic and neutral lipids include 1-(2-(oleoyloxy)ethyl)-2-oleyl-3-(2-hydroxyethyl)imidazolinium chloride and cholesterol and 1-(2-(oleoyloxy)ethyl)-2-oleyl-3-(2-ll.ydltl~Ly~;Lllyl)imitl~7nlinium chloride and diol~ll.lm~h~ yl ethanolamine.
A plerell.,d molar ratio of these lipids is 1:1. DNA collcellL aLion in the formnl~tions of the invention is from about 0.5mg/mL to about Smg/mL, more preferably from about 0.5mg/mL to about 2.5mg/mL. DNA:lipid ratios are preferably from about 1:1(,ug DNA/nmole lipid) for formulations to be injected hlL ai~ .iLoneally or by direct injection, to from about 1:6 to 1:15 (ug DNA/nmole lipid) for preparations to beinjected intravenously. DNA purity has a direct effect on liposome complex formation, but DNAs having a purity of about 15% to about 100% are a~.opliate for complex formation. DNAs having a purity of 90-100% by HPLC are preferably used in DNA:lipid complexes in a range of 1:12 to 1:15 ,ug DNA/nmole lipid.
The various lipid carrier-nucleic acid complexes, whel~ the lipid carrier is a liposome, are prepared using m~othndc well known in the art. Mixing conditions can be o~-Li lli~ed by visual ~o~5min~tinn of the resultant lipid-DNA mixture to establish that no ~.eci~3iL~tion or a~ ~leg~lion occurs. To make the lipid-DNA complexes more visible, the complexes can be stained with a dye which does not itself cause ag~ Lion, but which will stain either the DNA or the lipid. For example, Sudan black (which stains lipid) can be used as an aid to eY~ min~ the lipid-DNA mixture to dct~ llille if aggregation has occurred. Particle size also can be studied with m~thn~
known in the art, including elecl.onic Illicloscopy, laser light scattering, CoulterTM
collnting/sizing, and the like. Standard-size beads can be used to calibrate i.~LI un~
used for determining the size of any liposomes or complexes that form.
By "lipid carrier-nucleic acid complex" is meant a nucleic acid sequence as described above, generally bound to a lipid carrier preparation, as ~ cusser1 below.
The lipid carrier ~cpalation can also include other substances or cofactors.

CA 02223923 1997-12-0~

WO 96/~0962 PCT/U~,C~ 26 Furthermore, the lipid carrier-nucleic acid complex can include targeting agents to deliver the complex to particular cell or tissue types. Generally, the nucleic acid material is added to a ~u~en~ion of preformed liposomes which may be multi-lamellar vesicles (MLVs) or small llnil~m~llar vesicles (SUVs), usually SUVs formed by S sonication or by extravasation through ~L,~ ,lidL~ly-sized polycalhollate membranes.
The li~osollles themselves are ~le~al~d from a dried lipid film that is resuspended in an a~ o~,l ial~ mixing solution such as sterile water or an isotonic buffer solution such as lOmM Tris/NaCl or 5% dextrose in sterile water and sonicated to form the liposomes. Then the ~l~rulllled lipid carriers are generally mixed directly with the DNA.
Mixing and ~ alhlg of the lipid-DNA complex can be critically affected by the sequenre in which the lipid and DNA are combined. Generally, it is preferable (to ,n;--;",;,~ aggregation) to add the lipid to the DNA at ratios of DNA:lipid of up to 1:2 inclusive (microgram DNA:nanomoles cationic lipid). Where the ratio of DNA:lipidis 1:4 or higher, better results are generally obtained by adding the DNA to the lipid.
In either case, mixing should be rapidly achieved by shaking or vortexing for small volumes and by use of rapid mixing systems for large volumes. The lipid carrier and DNA form a very stable complex due to binding of the ne~d~ively charged DNA to the cationic lipid carriers. The DNA:lipid complexes of the invention find use with small nucleic acid fr~gm~ntc as well as with large regions of DNA (230kb).
Aggregation of the lipid carrier-nucleic acid complex is prevented by controlling the ratio of DNA to lipid carrier, ...i..;l..i~il.g the overall col1cellLldLion of DNA:lipid carrier complex in solution (usually less than 5 mg DNA/mL solution) and avoiding the use of chPI~ting agents such as EDTA and/or ~i~..ir.e~ .ullL~i of salt, either of which tends to prollloL~ macro-aggl~dLi~,ll. The ~l~f~ ,d excipient is water, dextrose/water or another solution having low or zero ionic strength. Further, the volume should be aC~ te~1 to the ",i,-i"""" n-oce~ry for injection into the hostm~mm~l, while at the same time taking care not to make the solution too conc~:llLlaL~d so that ag~l~,gaLes form.
~ 30 DNA:lipid complexes of the invention may be sized in accordance with conventional te~ hni~ es, depending upon the desired size. For intravenous or CA 02223923 1997-12-0~

hlLla~eliLull~,àl delivery, the complexes of the invention are preferably 150-300nrn in diameter.
The DNA:lipid complexes of the invention have utility in mt~ ting the errlcielJt delivery of the recolllbilL~nl expression constructs of the invention, encoding S functional genes of fragments thereof, into eukaryotic, preferably m~mm~ n, most preferably human cells. DNA:lipid complexes of the invention are useful for achieving gene transfer in vitro using established terhni~ e~c. More illl~OlLallLly, the DNA:lipid complexes provided by this invention, and the mPthcric of ~.h..i"i~lf .i"g the DNA:lipid complexes provided herein, are capable of specifir.lly delivering recombinant e,~l.ression cO~ ;L~ of the invention to particular tissues and cells C-J",~,isi"g those tissues in vivo, thereby providing tal~Li,lg of these genes to specific tissues. These ~ pellies of the ph~rm~relltir~l compositions and methods of the present invention provide for re~li7~tion of practical gene therapy, whereby, e.g., a particular ~efic ient gene is restored by the introduction of a functional copy of the normal cognate gene into the cells of the affected tissue, without the i,laL,p,opliate introduction of the construct into other cells and tissues of the body n~ ecirlcally.
Thus, the invention provides m~tho-lc and ph~nn~rentir~l colll~o~iLions having a number of advantages over the prior art. The liposomes and lipid complexes of the invention have been extensively studied in humans, and are non-immllnngenic, relatively non~toxic, and non-infectious. These complexes are stable, as illustrated by the exL1elhllellLal results shown in Figure 1. A particular DNA:lipid complex (DOTIM:Cholesterol (1:1) complexed with a CAT-encoding plasmid at a DNA:lipid ratio of 1:6 and a DNA conr~ontration of 0.625mg/mL~ was prepared and tested weekly over 11 weeks by injection into the tail vein of ICR mice. CAT activity was thend~t.,l",hled in mouse lung using protocols described in detail below. The Figureshows results dem~n~L~aLi~lg that this ple~dlaLion was stable over the course of the experiment, wh.,leby ~b~L;~ y identical levels of CAT gene expression were obtained at all time points tested.
The DNA:lipid complexes of the invention have additional advantages over the prior art. Recombinant expression constructs of any practicable size can be used, there being no limit~tion on large plasmid size due to the absence of p~rL-~ging the CA 02223923 1997-12-0~

W O 96/40962 PCT~US96/09526 DNA into the genome of a vector olgal~isms like a lcLl~vi~-lS or an adenovirus. Gene transfer can be achieved in non-dividing cells, unlike prior art systems which relied on viral vectors whose life cycle required the infected cells to be dividing. In addition, the specific formulation of the DNA:lipid complexes of the invention can be altered to affect ~l~e~ g and duration of the gene-e~ ession effect. The DNA:lipid complexes of the invention are also amenable to many delivery routes, and are less likely to encou~Le. the types of safety issues related to viral-based delivery systems.
The DNA:lipid complexes of the invention may be ~ ed to an animal to effect delivery of functional genes into specific tissues by any a~ o~. ia~e thc.~l~t;uLic routine, inrl~ ing intravenous, hlLl~eliloneal, sllbc~lt~nPou~, or"",~ r injection, as well as direct illje~,liol1 into target tissue(s). Typically, the DNA:lipid complexes of the invention are injected in solution where the conc~lll,dLion of the DNA to be delivered dictates the amount of the complex to be ~mini~t~oredThis amount will vary with the tissue to be targeted and the effectiveness of the targeted DNA, the required concel~LldLion for the desired effect, the number of s~lmini~trations, and the like.
The mPthods and ph~ relltir~l compositions of the invention are particularly useful and ap~Lopliate for introducing functional human genes, particularly human CFTR, to lung tissue. These methods and pharm~reutir~l compositions thus have utility in the L~ Llll~llL of human ~ ces, including cystic fibrosis and chronic clliLis.
The following Examples illustrate certain aspects of the above-described methods and advantageous results. The following examples are shown by way of illustration and not by way of limit~tion.

rl~r,~ . of DOTIl~ Cholesterol (1:1) Small Unil~n~ r Vesirl~
To a lL round bottom flask was added 500 ,umoles cholesterol dissolved in an excess of chloroform and then 500 ,umoles DOTIM also dissolved in an excess of chloroform. The amount of DOTIM was determined by high pleS~ule liquid chromatography (HPLC) or be UV s~ecL.~scopy at 237nm.

CA 02223923 1997-12-0~

After brief, gently mixing, the flask was ~tt~rh~d to a rotary evaporating apparatus and chloroform withdrawn under slow speed and water vacuum conditions until almost all of the solvent was evaporated. Evaporation was completed at m~imllm rotation speed using a vacuum pump to completely dry the lipid mixture to a thin film on the wall of the round bottom flask.
As an i"t~ ~ ",~.li,.l~ step to the formation of the title co~ ,osi~ion, mllltil~rnP~ r vesicles (MLVs) were prepared from this film by the addition of 16mL endotoxin-free water to the flask, which was then warmed to 37~C in a water bath with gentle hand-swirling. The MLVs thus formed were removed from the flask using a 9" Pasteur pipette and ~Idl~r~ d to a 20mm screw cap tube at room 1~ .. p.,~l.. c. The flask was cleared of any rem~ining MLVs by washing with an additional ml endotoxin-free water, which was added to the 16mL previously Lld~r~,...,d from the flask. Thesesolutions were mixed, and ~ uott~rl equally into 20 16mL screw cap tubes using aPasteur pipette.
MLVs were converted into the SUVs of the title composition by sonication.
Each of the 16mL screw cap tubes cont~inin~ MLVs were placed individually into asoni~ting water bath ~ d at 36~C for 5 min, and the k.ll~c.~ture of the bath chPc~ bclween the introduction of each tube. Sonicated droplets within each tubewere collected by brief vortex mixing, and the individual solutions of SWs were then combined into a single 20mm screw cap tube using a 9" Pasteur pipette, and then filtered using a 0.2 micron disposable filter (Nalgene). Finally, an amount of an endotoxin-free solution of 25% dextrose in water, equal to one-quarter of the final volume of SUVs, was added to the tube of SWs. This resulted in a ~ ellsion of SWs con~l;~i..g 20mM DOTIM and 20mM cholesterol (40mM total lipid) in a 5%
dextrose solution, which was kept at 4~C until use.

T qrge Sr~ ? Pl~cmi(l DNA P~
Plasmid DNA was p.~,~,a.ed in large-scale (i.e., milligram) qu~ntiti~-c using a mnr1ifin~tiQn of the alkaline Iysis procedure (Sambrook et al., 1990, ibid.). Briefly~
bacteria COlll~ illg a single colony were grown for 12-18 hours or overnight in l5mL

, W O 96/40962 PCT~US9C~'0~5~6 TB broth (47g/L TB (Sigma Chemical Co., St. Louis, MO)/ 8 % glycerol) supplemtont~cl with 100~Lg/mL carbenirillin at 37~C with shaking (250 rpm). 2-2.5mL
of this culture was then added to 400mL TB (supp1em~lf~ with 100,ug/mL
cd.l,ell.cillin) in each of six 2L flasks (for a total of 2.4L culture) and grown at 37~C
- 5 witn shaking overnight (16-18h).
After ~ hL growth, bacteria were collected by centrifugation for 10 min.
at 4~C in a Beckm~n J2-MI centrifuge equipped with a JA-10 rotor. The bacterial pellet in each centrifuge bottle was gently 1- s..~ ed in 20mL of an ice-cold solution of 50mM dextrose in 25mM HCL buffer (pH8)/10mM EDTA. To the resuspended bacterial cell pellets were added 40mL of a freshly-made solution of 0.2N NaOH/1 %
sodium dodecyl sulfate at room lellllJeldlul~, resulting in cell Iysis upon gentle agitation of this mixture on ice for about 5 min. After the added Iysis solution has been thoroughly mixed into the bacterial suspension and the cells Iysed, the mixture was allowed to stand at room Lrnl~,.dLule for S min. To this mixture of Iysed bacteria was added 20mL of an ice-cold solution of 3M potassium acetate, which was mixed into the Iysed bacterial solution gently by hand and then stored on ice for 10 min. A
flocculant white precipitate formed, comprising bacterial chromosomal DNA, RNA
and SDS/protein/m~lll'uldlle çomplexes, which were cleared from the solution by centrifugation at 8000rpm for 15 min at 4~C in the JA-10 rotor as above.
After centrifugation, the su~elllaL~IlL was ll~l~.fel.~d with filtering through Miracloth to 250mL centrifuge bottles, and 50mL isu~lol~anol added at room e,ature, mixed and inrnh~tt-(l for 10 min. The plasmid DNA precipitate was recovered by centrifugation at 5000rpm for 10min at room t~ cldth.e in a JA 14 rotor (Rer'~n~n). The alcohol-cont~ining sul-~-llaL~llL was ~ec~nt~ and residual~u~t.. aL~.. L removed by vacuum aspiration.
The plasmid DNA pellets were l~.u~ ded in 6mL of a solution of 6mM Tris-HCL (pH8) and Ll~ led to 50mL centrifuge tubes upon dissolution. To each tube was added and equal volume of cold (-20~C) 5M LiCI, the solutions mixed by hand and then centrifuged at 8000rpm for 10min at room lem~ lule in a JA-20 rotor ~ 30 (Re~'~n~n). The supernatant solution from each tube was Llal~r~ d to a fresh tube and the plasmid DNA then re-plcci~iLaL~d by the addition of an equal volume of CA 02223923 1997-12-0~
W O 96/40962 PCT/U~3G/'~53 6 isoplul)allol, mixed and collected by centrifugation at 5000rpm for 10min at room te-~pcl~Lule in a JA-20 rotor. The alcohol-cont~ining supernatant solution was then ~eç~nted, residual alcohol removed by aspiration, and the plasmid DNA pellets allowed to air dry for 5min.
S Co.. l;.. i.~t;.. ~ bacterial RNA was removed from the plasmid DNA bydissolving the pellets in lmL 10mM Tris-HCL (pH8), adding about 0.5-0.75,ug of pancreatic RNase per mL, followed by inrllb~ting the mixture at 37~C for lh.
Disa~pealdllce of RNA was dcL~ h~ed by ethi~ lm bromide-stained agarose gel analysis (see Sambrook et al., ibid.). Plasmid DNA was purified by phenol-chloroform extraction. Briefly, to each aliquot of plasmid DNA solution was added an equal volume of Tris-saturated phenol:chloroform (1: 1), the immi~cible solutions mixed by vortexing, and centrifuged in a laboratory tabletop microfuge for 5min at room Lc~ JcldLulc. The aqueous (upper) layer was removed, Lldl~rt;llcd to a fresh microfuge tube, and extraction with phenol:chloroform repeated at least twice. These extractions were followed by two extractions of the aqueous layer with Tris-saturated chloroforrn. Plasmid DNA was corlccllLldLcd by precipitation, with the addition of SM
sodium acetate to a final cunccllLlaLion of 0.3M and the addition of two volumes of cold (-20~C) absolute ethanol. DNA was allowed to pl~cipiL~te in this solution at -20~C for lh or overnight.
After pleci~iL~tion, plasmid DNA was collected by centrifugation at about 6000rpm in a clinical microcentrifuge. The alcohol-conf~ining ~ul~.l~LallL was aspirated by vacuum, and the pellet washed twice with 70% ethanol/water (4~C). The washed pellets were air dried for at least 30min. Plasmid DNA pellets were dissolved in a total of 6mL of a solution of 10mM Tris-HCL (pH8), and concentration deL~.lllilled by spectrophotometric analysis of a 1-to-200 dilution of the recovered plasmid at A260 Pl~v;~ of I~NA:I,lP~n C~ c DOTIM:cholesterol:plasmid DNA liposomes were prepared as follows. A
DOTIM:cholesterol mixture (1:1, 20~moles/,uL each lipid) was ~lc~lcd as described , CA 02223923 1997-12-0~
W O 96/40962 PCT~US96/09526 in Example 1 above. Complexes with plasmid DNA were prepared in DNA:lipid ratios of 1:1 and 1:6. DNA and DOTIM:cholesterol were each first brought from storage conditions (-20~C for DNA, 4~C for liposome form~ tions) to room temperature before use over the course of about l.Sh. DNA concentration in the complex ~ ,aldLiorls were optimally 100-550,ug/200,uL complex solution (for ratios of 1:1 DNA:lipid) and 100-150,ug/,uL complex (for ratios of 1:6 DNA:lipid). DNA
c..nr~ ns were typically ~ ",;"tod just prior to DNA:lipid complex formation, by ultraviolet spectrophotometry as described in Example 2. DOTIM:cholesterol Lules were typically used at a total lipid concentration of 40,umole/mL, co"e~l,ollding to 20,umole/mL DOTIM and 20,umole/mL cholesterol.
DNA:lipid complexes were ~lcpdlcd from these reagents as follows. Each component was prepared in individual microfuge tubes to a total volume per tube of 100,uL. An a~,vpliate amount of DNA (equivalent to a final DNA concentration of 500,ug DNA/mL complex) was added to one tube, and brought to volume with water or a solution of 5 % dextrose in water. The al)p~u~ amount of the DOTIM:Cholesterol mixture (lOOnmoles lipid/100,ug DNA at a 1:1 ratio; 600nmoles lipid/100~g DNA at a 1 :6 ratio) was added to a second tube, and water or a solution of 5% dextrose in water was added to bring this solution to a total volume of 100,uL.
The contents of the lipid-cont~ining tube were mixed by vortexing for about 2sec, while the contents of the DNA-cont~ining tube were mixed gently using a lmL
pipettor. The contents of the lipid mixture-cont~ining tube were then added to the DNA-cont~ining tube using a lmL pipettor. It was found that it was ~c~enti~l that this addition was pelrc"",ed slowly, in a constant stream, to the top of the DNA solution in tube A. As the lipid solution mixed with the DNA, formation of the DNA:lipid complex was ~ tt~et~d by the solution becolllillg slightly cloudy and op~lPscenf It was also d~t.,~",ined that, at this stage, the mixture could not be vigorously mixed (for example, by vortexing) without seriously comprolllisi~g the integrity and usefulness of the complexes so formed; however, it was advantageous to gently mix the entire contents of the tube 3-4 times after completion of addition of the lipid mixture to the DNA mixture.
After the complexes were formed, the final conct;~ tion of DNA was WO 96/40962 PCT~U5~5.'0~ 6 determined by ultraviolet ~ecllol)hotometry as described above, and the size of the DNA:lipid complexes detellllilled by light scalle.illg measured at 400nm.

~v~ n of Ti~ ~pl.oe for CAT A~ nll Protrin nel~
Tissues were p~c~al~,d for assay as follows. E~ c;llLtl animals were .,i,r(l quickly and llulllanely. Mice were typically placed in a kill box flooded with CO2 for 2-3 min. Tissues were halve~.L~d by dissection and weighed, and then placed in lmL cold homogenation buffer (250mM Tris/5mM EDTA) supplemented with PMSF (35,ug/mL) and L~ul,.,~lill/Aplutillill (5~g/mL). Tissues were then homogenized for 20-30 sec using a tissue di~lu~L()l (such as a Polytron) until a ulliru homogenate was obtained. This homogenate was then tldl~rell~d to a centrifuge tube and quickly frozen on dry ice (or at -70~C) and then thawed at 37~C. Insoluble debris was cleared from t'ne homogenate by centrifugation at 10,000g for 5-10 min. 50,uL
of the rçslllting ~upelllaldllL solution was aliquotted into a microfuge tube and stored at -70~C until used for protein ~3e~r~ ions.
The rrm~inrlrr of the .u~elllaLdllL was heat inactivated at 65~C for 15 min and re-centrifuged at 10,000g for 10min, and stored at -70~C until use for CAT assay d~L~ ldlions~
To perform CAT assays, samples were analyzed in parallel with a series of standard CAT activity samples. From the ~Ldlldalds was developed a standard curve of CAT activity versus CAT protein, which was used to deLe.lllille the level of CAT
protein expression in tissue samples based on the observed CAT activity in tissue homogenates. To prepare the ~L~l1dald curve, serial dilutions of CAT enzyme were~ al~d ranging from 0.1 to 0.000025U. These standards were l,lepdl~d in a reaction mixture col~ hlg of 50,uL BSA buffer (having formula: 250mM Tris/SmM
EDTA/2mg/mL BSA, Fraction V (U.S. Biorh~ mir~l, Cleveland, OH), 5,uL standard CAT enzyme (and a~l)lù~lidl~ dilutions; obtained from Sigma Chemical Co, St. Louis, MO), 50,uL '4C-labeled chlorarnphenicol (New F.ngl~ntl Nuclear; diluted 1: 10 in BSA
buffer prior to use) and 25 ,uL n-butyryl-CoA (Sigma). Tissue samples were prepared i~1entir~11y, with the exception that 30,uL of tissue homogenate was substituted for the W O 96/40962 PCTAJ~'033 6 5~L of standard CAT enzyme activity. Samples were ~ cl at 37~C for 2h. After this inruhatinn, 300,uL of mixed xylenes (Aldrich Chemical Co.) were added to each tube, vortexed for 30sec, and centrifuged for 3min at lO,OOOrpm in an IEC centrifuge equipped with a 24-slot rotor. The mixed xylenes (upper) phase of each sample tube was Lldl~Ç~ d to a fresh microfuge tube and 750,uL homogenation buffer added. The samples were then vortexed and centrifuged as described above.
200,uL of the upper phase from each tube were Ll~r~.lcd to liquid scintill~tinn vials and 0.5rnL scintill~tion cocktail (Ready-Safe, Recl~m~n) added. The amount of CAT-specific radioactivity in each sample was ~fle~---;--fd by liquid scintill~ti--n counting assay.

EXAMPLE S
X-G~I St~ini~ of Tj!i~ f' ~ ples Tissue samples were stained with X-gal (5-bromo4-chloro-3-indolyl-a-D-gala~;Lc,~yldnoside) using the following protocol. Tissues are fixed by h~ elsion for 0.5-lh on ice in freshly-made fixative solution (2% neutral buffered formalin/0.02%
gluteraldehyde/0.02% Nonidet-P40). After fixation, tissues were rinsed twice at 4~C
in a solution of 2mM MgCl2/0.1 % desoxycholate/0.2% NP-40 in lOmM phosphate-buffered saline (PBS; pH 7.3). Tissues were then stained using rinse solution supplemPntf d with lmg/mL X-Gal (U.S. Biochf mir~l), SmM fe.licy~llide and SmM
ferrocyanide. Tissues were stained for 12-48h at 37~C or room temperature. Afterst~ining, tissues are rinsed in PBS. Tissues were then frozen and sectioned or fixed in 70% ethanol, embedded in paraffin and sectioned.
Protein r1f t~ ns were pelr~,lllled using a dye binding assay (BSA Protein Assay Reagent, Pierce Chemical Co.). The Pierce reagent was pl~alcd by mixing 50 parts of Reagent A with 1 part Reagent B as provided by the supplier. lOO,uL of this pl~paled reagent were aliquotted into each well of a 96 well llflCL'O~ plate. 100 ,uL of a solution conr~ining 20 ,ug BSA were added to the first well of the first row (i.e., well A1) and lOO,uL of a 1:2 to 1:8 dilution of each tissue extract were added to the other wells in the row. Serial dilutions at ratios of 1:2 were made in each of the adjoining rows consecutively using the wells in the preceding row.

W O 96/40962 PCTAU~,G~ 3>6 Typically, 96-well plates having 12 wells/row resulted in 6 serial dilutions (1 to 1/64);
the last row was a blank loaded with PBS as a control. The plates were incubated at 37~C for 30 minutes, and the extent of dye binding ~ d s~e~;LIul~hotomPtrically as abso,l,allce at 562nm. Prûtein concellllalions in sample wells were d~lelllli-led in S comparison with a standard curve generated using the OD readings from the serial dilutions of the BSA standard.

Mirro~ A~y for ~ t~ e~ F,~,. ~ n in Ti.c~
Tissue was homogenized in an appru~liate volume of homogenization buffer (250mM TristSmM EDTA) (e.g., 300,uL were used to homogenize a mouse lung).
The homogenate was then incubated on ice for 30min and centrifuged in a rnicrocelll,il;lg~ for 10min at 13,000rpm to clear the homogenate of insoluble debris.
Supell,~Lal.L~ from these homogenates were collected and assayed as follows.
Microplates were ~ ed for these analyses as follows. For each plate to be covered, 50,uL of anti-~-g~ tc)sidase monoclonal antibody was diluted in 5mL of SOmM sodium bica-l,ondt~ buffer (pH 9.4). 50,uL of the diluted antibody solution was added to each well of a lluclolill~ plate (e.g., Immulon 3, Dynatech), the plate sealed and in~llh~tPd overnight at 4~C. After overnight in~llb~tion, 200,uL BLOTTO solution (5% v/v nonfat dry milk and 0.2% Tween-20 in PBS) were added to each well and ill. ub~rd for lh at room ~ ,t;lalllle. The BLOTTO solution was then removed andthe plates washed three times with a solution of PBS/0.2% Tween-20, with the exception that the first row was not washed with this solution. 100,uL of a standard solution of 10mU/mL 13-g~ tosi~l~ce were added to the first well of the first row (i.e., well A1) and 100,uL of each tissue extract were added to the other wells in the row.
Serial ~ lti~mc at ratios of 1:2 were made in each of the adjoining rows consecutively using the wells in the p,t;cedillg row. Typically, 96-well plates having 12 wells/row resulted in 6 serial dilutions (1 to 1/64); the last row is a blank loaded with PBS as a control. The plates were il l- ~Ib~-~f d at roûm lell~ Lulc; for lh, and then washed three times with a solution of PBS/0.2% Tween-20 as above.
To develop the assay, 100,uL of CPRG assay buffer (2.5mg/mL chlorphenol W O 96/40962 PCT~ 'U3i26 red-~-D-gala.;w~yldlloside monosodium salt (CPRG)/1.8mg/mL MgCI2/7.1,hL/mL 2-mercaptoethanol in PBS) were loaded into each washed well and the plates then inr-~b~ted at 37~C for 2h. The extent of ~ rtosidase t;~ es~ion was then determined spectrophutol-le~lically as absorbance at 562nm.
Whole tissues and tissue sections were assayed using a mo-lifir~tion of this protocol. Frozen tissue or tissue sections were fixed by hll~ illg the frozen tissues in fixative solution (2% neutral buffered formalin/0.02% glutaraldehyde/0.02% NP-40) without thawing. Tissues were i...-ubatr~l in fixative solution for 2h at room Lul~ with gentle ~if~ti~n After inrllb~tio~, the tissues were rinsed twice with PBS, then inruh~t~d at 37~C overnight in X-Gal staining solution (SmM potassium r~,.li~;yal~ide/SmM potassium ferrocyanide/0.01% sodium desoxycholate/0.02% NP-40/lmg/mL X-Gal in PBS, supplemented with MgCk to 20,uM imm.~ trly before use). After staining, tissues were washed twice with PBS, and then embedded in paraffin or quick frozen for sectioning and hictorhrmir~l analysis.

Detecti- n o~ F..-~ n~ CFTR E~ in Transfected Cells U.~ir~ a Chloride ~ A.~
A chloride ion efflux assay was used to detect functional expression of CFTR
in transfected cells.
About 24h prior to introducing CFTR into cells, cells were split into a 6-well tissue culture dish, each well receiving lmL of lOmL of the cells on the dish plus 3mL
media. Cells were returned to the incubator and allowed to grow overnight at 37~C/5% CO~, or until they were about 70-80% col~lue~lL. For assay, media were removed from the wells and each well was washed with 2mL serum-free media. lmL
of serum-free media was then added per well, and the cells i.~. ..h;.~ d at 37~C for 1-2h.
200,u1 of a DNA-lipid complex c~nl~ illg a recombinant expression construct encoding CFTR were then added to each well and inrllh~f~-d at 37~C for 6-8h. After this inrnb~tion, media were removed from each well, the wells were washed twice with 2mL serum-free media and inr--hatrd in 4mL serum-cont~ining media at 37~C
for 48h.
The chloride ion efflux assay was p~lrulll,ed as follows. Media were aspirated from each of the wells co..l;.;..i..g cells treated with DNA-lipid complexes, and washed twice with efflux solution (135mM NaCI/2.4mM K~HPO4/0.6mM KH,PO,/1.2mM

CA 02223923 1997-12-0~ .
WO 96/40962 PCTAUS9~t~53~6 CaCI./1.2mM MgC~/10mM glucose/lOmM HEPES (pH 7.4)). Cells were then inrub~tPd with lmL efflux solution cont~ining Na36CI at a final col,ce,lLldLion of 2.5,uCi/mL 36CI- for 2h at 37~C. After inrllhi-tion, the36CI~-co..~i.;..i.,g efflux solution was aspirated from the cells and the cells then washed each of 4 times with lmL efflux solution. The cells were then ;I~ rd with lrnL efflux solution for 3min at room Lt:lllp~aLule, and the efflux solution then removed from the cells and Lldll,rt~ d into a scinfill~tion vial cont~ining 5mL scintill~tion cocktail. A fresh aliquot of efflux solution was added to each well and inr~lb~t~d for an additional 3 min. After each inr~lb.~tion, efflux solution was lldl~,L,ll~d to a scintillation vial cont~inin~ 5mL
scintill~tion cocktail, and a fresh lmL aliquot of efflux media was added to the cells and inrllbat~l for 3min. These steps of the assay were repeated ten times for a total of 30min. In certain of the wells,3ficl- ion efflux was stim--l~tt-d by in.~ul-~li..g these cells in the p~cs~nce of 40,uM Forskolin (Sigma), 500,uM cpt-cAMP (Sigma), and 100,~4M IBMX (Sigma) in efflux solution, efflux being stim--li-~t~cl at repetitions 3 through 7.
The extent of 36CI- ion efflux over this period was determined by scintillation col",~ , and the basal rate of 36CI- ion efflux colll~al-,d with the rate of efflux in cells stim--l~t~ by Forskolin/cpt-cAMP/IBMX. Extent of efflux was norrn~li7~-d relative to the amount of 36CI- ion l~ g inside the cells after the 30min inrllb~tion. This quantity was dt:Lc.lllhled by Iysing the cells by inrllh~ting them with lmL of scintillation fluid for 15min. The Iysate from each well was then Llal~,r~ ,d into a scintill~ti- n vial, the well washed with lmL of efflux solution which was added to the cell Iysate, and the 36CI- ion-associated radioactivity counted.
The results of one such assay are shown in Figure 2. Two plasmids encoding CFTR and differing in the details of the construct (see Table I) were tested with (closed circles and boxes) and without (open circles and boxes) stim~ ion. As isshown in the Figure, stimulation results in the rapid in~ ctinn of chloride ion efflux over the basal rate of efflux, which efflux persists even after the stim--l--c is removed (time points 24-30). These results demonstrate the utility of this assay to detect functional expression of CFTR in heterologous cells, and thus forms an in vitro ~.landaLd for d~tc.~llhlillg the vigor of dirr~le.ll recombinant expression constructs in e~L,ulessing human CFTR.

W O 96/40962 PCT~US96/09526 TABLE I

Vectors with the CFTR cDNA
enhancer promoter intron polyA antibiotic MB19: HCMV HCMV ppi ppi amp MB31: HCMV HCMV ppi SV40 amp MB65: HCMV HCMV ppi ~Imyc ppi amp MB66: HCMV HCMV ppi Cmvc SV40 amp MB76: HCMV HCMV ppi 3xSV40 amp MB77: --- CC10 ppi 3xSV40 amp MB78: HCMV CC10 ppi 3xSV40 amp MB81: ---- CF~R ppi 3xSV40 amp MB87: HCMV CE~R ppi 3xSV40 amp MB90: HCMV HCMV --- 3XSV40 amp MB93: HCMV HCMV pgl3 SV40 amp MB97: HCMV HCMV pgl3 SV40 amp/tet MB113: HCMV HCMV pgl3 SV40 tet Reversc T~ C~liv~z~c~-polym~r~cf~ Ch~in R~~ti-)n ~n~lysic Human CFTR gene ~ ,ssion was assayed using a reverse L~nsc~ hse polymerase chain reaction assay (RT-PCR) on ll~nsre~;led tissue culture cells and whole tissues. These assays were pelrol,lled using vector specific primers and CFTR
specific primers. The vector specific primers used were:
5' AGA TCG CCT GGA GAC GCC AT 3' forward primer (3651-3671bp in pMB19; Figure 7 and SEQ ID No.:1) and 5' GCT CCT AAT GCC AAA GGA AT 3' reverse primer (1246-1266 bp in pMBl9, u~ ll from hCFTR ATG site;
SEQ ID No.:2).
The CFTR specific primers were used:
S' CCT GTC TCC TGG ACA GAA A 3' forward primer (3337-3355bp in pMBl9; SEQ ID
No.: 3) and 5' GTC TTT CGG TGA ATG TTC TGA C 3' reverse primer (3651-3671 bp in pMB19; SEQ ID
No.: 4) Tissues were frozen on dry ice for RT-PCT and stored at -70~C. Tissue samples were homogenized and used directly in this evaluation.
Briefly, RT-PCR was ~.,.rulllled by ~le~ g first-strand cDNA from cellular RNA isolated from frozen tissues using ~L~ndald t~chniqlles (see Sambrook et al., ibid.), inrlu-lin~ specifically the use of random hPx~mPr for plilllillg and MMLV-derived reverse L~ s~ ~se. cDNA was used in PCR reactions pelrul~lled as follows.
The entire 25,uL of the first-strand cDNA reaction was mi~ced with the components of the PCR reaction (under standard conditions; see Innis et al., 1990, PCR Protocols:
A Guide to Mtothn~1c ~n-l A~ tior~, Ac~demi~ Press, New York), inr!ll~1ing 25,uMapiece of each of the specific pairs of PCR primers. PCR reactions were overlayed with light mineral oil to prevent con-1Pn~tion and then subjected to the following PCR

cycling protocol:

1 cycle 10min 94~C

S 30 cycles lmin 94~C
2min 55~C
3min 72~C

1 cycle 10min 72~C
2min 27~C.
After completion of the reaction, the a~alaLuS was programmed to take and hold the reaction lniA.~ul~es at 4~C until analysis.
PCR products were analyzed by electrophoresis in agarose or acrylamide gels.
In these assays, the vector-specific primers were expected to yield a band ,c~l~s.,.,~ re of plasmid DNA (485bp) and a hCFTR RNA-specific band (142bp).
The CFTR-specific primers were expected to yield a DNA fragment band of 334bp.

F-ln-~ti--nal P-ol;very of CAT Gene Cn~ u~ to CPI1~ In vivo Functional delivery of a variety of CAT lC~Ull~l gene constructs was achieved using different embo-lim~n~ of the DNA:lipid complexes of the invention.

A. DOTIM:Ch~ rrol Formn~ n I
DOTIM:cholesterol liposomes were prepared as described above in 1:1 ratio and used to prepare DNA:lipid complexes. DOTIM:cholesterol (1:1) liposomes were used to make DNA complexes using the chloramphenicol acetyl transferase (CAT) expression vector p4119 (Figure 3). DNA:lipid complexes were ~lcpa~d having a DNA:lipid ratio of 1 :6, and using 125,ug of DNA per 200,uL complex. Lipûsome size was deLe,lllilled by optical density (OD) at 400 nm. A total of 200,uL of the complex were injected into the tail veins of 3 ICR mice. At 24 hrs post-injection, tissues were harvested and plc~aled for CAT assays as described in Example 4 above. Tissues harvested inrln~le~l lung, liver, kidney, spleen, ovary, brain, smooth muscle, heart and ear.

CA 02223923 1997-12-0~
W O 96/40962 PCT~US~''05~6 Results of these CAT assays are shown in Tables II and III below. Table II
shows CAT activity as total '4C-labeled chloramphenicol counts converted to acetyl and diaceyl forms by CAT expression vector-encoded enzyme activity in lung for each of the three e~ h.lel.L~l animals tested.
TABLE II
animal number CAT (cpm) 20. 1-1 800,000 20. 1-2 1 ,400,000 1020. 1-3 400,000 Table III shows CAT assay data from a variety of tissues from one of the experimental animals (animal 20.1-2). These results demonstrate tnat intravenous inoculation of mice in the tail vein with DOTIM:cholesterol:DNA complexes in this formulation results in l~-e~lel.Lial targeting of the DNA:lipid complexes to the lungs, with CAT
activity in lung tissue le,~les~ g over 80% of the CAT activity ~letPctPd in all mouse tissues tested.
TABLE III
20tissue CAT (cpm) lu 1,400.000 li 20,000 sp 63,000 ki 15,000 ov 3,000 br 7,000 sm 58,000 he 1 15.000 ear 1 500 Key: lung (lu), liver (li), kidney (ki), spleen (sp), ovary (ov), brain (br), smooth muscle (sm), heart (he).

_ -The results of these e~ are also shown y,lA~hically in Figure 4, which ~U-l .la.i,~s the results obtained with over 700 e~.,.illle.lkll and control mice As can be seen in the Figure, treated mice reproducibly showed greater than 1000-fold higher CAT activity in lung of mice treated with the DNA:lipid complexes of the invention comprising CAT-encoding l~cOlllbillallL expression con~Llucl~ (a total of 555 mice), compared with control (untreated) mice (a total of 163 mice) The delivery and uptake into cells of various mouse tissues of the CAT plasmid DNA a~ ,d as DNA:lipid COlllp' .~S of the invention by injection into the tail vein of mice was analyzed by Southern blot analysis using routine procedures (see Sambrook et al. ibid.). DNA from mouse tissues was ç~tr~ctt~(~, purified and ~ ested with BamHI restriction endo ,~rl~ce The resulting DNA restriction frAgTnPnt~ were separated by agarose gel electrophoresis and Llal~r~lred to a ~ e by capillary action Such lllelllbl~nes were dried, ~-ellybridized and then hybridized with a r~lioactively-labeled, CAT DNA-specific probe (about 108-109 dpm/,ug) at an a~proplialt: stringency (2X-6X SSC at 62~C) overnight, washed to high ~ ge.. ey (0.1-0 5X SSC at 65~C) and exposed to autoradiographic film at -70~C using hl~ siry-ing screens Results of these e~.,l..llc;llls are shown in Figure 5 The lower panel is ntirAI to the upper panel, but has been allowed to expose the X-ray film for a longer period of time These results del.lol1sl,dL~ that CAT DNA is introduced specifically into lung, with signifirAnt amounLs of DNA uptake in spleen Much lower amounts of CAT DNA were observed in certain other tissues (liver, kidney) but many tissues showed rc~entiAlly no CAT-specific hybridization, even at the longer exposure time Lung tissue from untreated animals was analyzed to ~1( t. , ~i ~ the specific cell types L1A ~ir~cl~ Histological sections were analyzed for vector-specific mRNA by reverse ~ s~ ~se-poly---t;-_se chain reaction (RT-PCR), ~,.,~r(Jlnled as described in Example 9 The results shown in Figure 21 ;ll~ r~l that expression of the transgene was pred(Jlllhlalllly found in vascular endothelial cells A second series of t~ e,i---ents were pe,rc",lled using this lipid forTmllAtinn In these e~ , the DNA construct used was the ,B-galactosidase expression vector MB10 (see Table I) that encodes a form of ~-galactosidase that is translocated CA 02223923 1997-12-0~
W O 96/40962 PCT/U~,G~5;~6 into the nucleus in in vitro studies. Complexes were formed as described above, and mice were injected witn 200,uL complexes in the tail vein. The resulting ,~-g~l~rtosiri~e levels present in lungs are shown in Figure 6, which re~lcsellLi7 the results of e~ c.-L. with 9 ~AIJclill~nLdl and 1 control (a.l~ LI .cd liposome omy) mouse.
In a third series of e~ .illlcnts, ~ ,resi.ion of the human CFTR gene was shown following IV delivery of DNA/DOTIM:cholesterol complexes. A recol~ ldlll e~ress.ion plasmid encoding the human CFTR gene (M1319; see Table I and Figure 7) was used to make DNA:lipid complexes as described above (DNA/lipid ratio of 1 :6, 125,ug DNA/200,uL complex). These complexes were tested by transfection/chloride ion efflux assay in human 293 cells in vitro, as described in Example 8, and 200,uL was injected into each of ICR 3 mice. Cells and lungs werehdl vt; .l~d at 24 hrs. RNA was made using conventional methods as embodied in kits from either Stratagene (for cell culture results) or 5'-3' Prime (for lung tissues).
Samples were analyzed by RT-PCR as described above in Example 9. In this analysis, amplification of plasmid Sf qn~nrç~ yielded a 484bp PCR product, while amplifir~tion of cDNA collei.~ullding to spliced CFTR mRNA for CFTR yielded a 142bp PCR
product. Similar results were obtained from lungs following IV ~-1mini~tration of the CFTR/lipid complexes.
The time course of ~ .sion of exogenously added CAT-encoding plasmid in mouse lung was rlr~ od A number of mice were injected intravenously in the tail vein with DNA/lipid complexes comprising p4119 CAT DNA at a 1 :6 ratio witnDOTIM:cholesterol at a concellL~dLion of 125,ug/200,u1. Mice were sacrificed in duplicate over a period of 55 days, and lung tissue analyzed by CAT assay as described above. These results are shown in Figure 8, which in~ t~li that high-level, pc~ .Lel1L expression of the reporter gene construct had been achieved Complexes of this DOTIM:DNA formulation were also ~ .ed by direct intracranial delivery. Complexes were made using CAT ex~lession plasmid p4119 and complexed with DOTIM:cholesterol (1: 1) at a ratio of 1: 1 DNA:lipid at a DNA
collce~ aLion of 500,ug/200,uL. 200,uL of these complexes were directly implanted intracranially, and the extent of CAT activity is brain tissue analyzed 24h later. The W O 96/40962 PCT~US9GJ~53~6 results of these t~ .h~ L~ are shown Figure 9.
The results of these different assays in~ t~d that this DOTIM:cholesterol form~ tion was capable of delivering a variety of recombinant ex~l~,ssion constructs to the lung after intravenous ~ lion, as well as by direct injection into a tissue of interest (brain).

B. DOTIM:Ch~ fProl Fu~ n II
DOTIM:cholesterol liposomes were ~ a~,d as desc.ibed above in 1:1 ratio and used to prepare DNA:lipid complexes. DOTIM:cholesterol (1:1) liposomes were used to make DNA complexes using the chloramph~ni~ol acetyl Lldl~r~,ldse (CAT) expression vector p4119. DNA:lipid complexes were prepared having a DNA:lipid ratio of 1: 1, and using 200-550,ug of DNA per 200~L complex. Liposomes were injected into the tail vein of ICR mice, as described above.
CAT gene expression in lung tissue from mice injected with DOTIM:cholesterol:DNA complexes ~Ic,~dled at a DNA/lipid ratio of 1:1 was determined. Plasmid p4119 DNA was complexed with DOTIM:cholesterol formulation of the invention, the complexes having a DNA/lipid ratio of 1: 1. Tail vein injections were pe,rull,Rd and tissues harvested at 24 hrs as described.
The results of these assays are shown in Table IV below.
TABLE IV
Amount of DNA/complex OD4"" lung expression 200,ug/200,u1 0.24 39,000 300,ug/200~1 0.32 9.000 400,ug/200,L~1 0.53 500,000 500,ug/200,u1 0.66 700,000 550,ug/200~41 0.82 1.000,000 negative control 0.03 0 * light scdLL~ g as an estimate of complex size ** in cpm of acetylated and diacetylated '4C-labeled chloramphenicol CA 02223923 1997-12-0~
WO 96/40962 PCTAUS~5~ 6 DNA/DOTIM complexes were made using plasmid p4119 and DNA/DOTIM:cholesterol liposomes at ratios of 1:6 and 1:8, were held at 40~C for 11 days prior to testing and then tested again at 18 days. The results of CAT
,.ession assays using these formulations are shown in Table V.
TABLE V
DNA/DOTIM ratio time stored CAT/lung~
1 :6 11 days 515,000 1:8 11 days 1,050,000 1:6 18 days 11,000.000 1:8 18 days 3,450,000 ** in cpm of acetylated and diacetylated l4C-labeled chloramphenicol DOTIM:cholesterol complexes with DNA were also ~ "i..i~ .cd by i.. Llapc.iLolleal injection. DNA:liposome complexes ~rlminictrred intravenously and hiLl~l)e.iLolleally were con,~alcd, using CAT expression plasmid p4119 DNA
complexed with DOTIM:cholesterol (1:1) formulations of the invention. In these assays, the complexes had a DNA/lipid ratio of 1:1 and a DNA conce~lLl~tion of 300-500,ug/200,uL. A total of lmL of these complexes was injected hlL~pc;~iLolleally in two mice (mice 4 and 5), 200,uL were ~ tP~ ed intravenously (mice 1 and 2), and 1 mouse (mouse 3) was ~ ed a formulation c~ lisi.lg only liposomes.
Tissues were harvested at 48h post-injection. CAT assays were performed as des~;.il)ed above in Example 3, and the results of these assays are shown in Figure 10 for heart (he), spleen (sp), pancreas (pa) and lung (lu).
The effect on the effiri~ ry of DNA delivery to tissues in vivo of intravenously~-lmini~tering dirr~,re.lL formulations coll.pli~i-lg the same mixture of cationic and neutral lipids was determined by c~ln~illillg the extent of transferred CAT activity observed using the dirr~l~llL formulations. CAT plasmid DNA/DOTIM:DOPE (1: 1) complexes were p.~,~aLed in the following formulations:
A. DNA:Lipid ratio of 1:6 DNA concentration of 0.625mg/mL
B. DNA:Lipid ratio of 1:1 DNA Conce..L-~Lion of 2.5mg/mL

W O 96/40962 PCT/U~j'J~ 6 Each formulation was ~lcpdlcd as described in Examples 1 and 3 above, and were d by intravenous injection into the tail vein of cohorts of 3 ICR mice per tested fo~mnl~tinn. Liposomes that were not co..~ ed with DNA were injected intoa sepalalc cohort of 3 mice as a control.
S Animals were sacrificed 1-2 days after injection and analyzed by CAT assay of spleen, heart and lung tissue. The results of these e~ lcllL~i are shown in Figure 11. This Figure ~em~ dlcs that Formlll~tion B provides a c~ cl.lly higher level of CAT activity in spleen, heart and lung than Formlll~tirln A, ~Ifhml~h it appears that the relative efficiency of plasmid delivery is about the same for both formulations.
C. C~ of HLA Gene Delivery using Dilr~ l DNA:Lipid Complexes Three dirrt re-l~ lipid formlll~ti~ ns were used to deliver a human HLA-encodingColl~.LluCL to bone marrow, spleen and lymph node. The three f~rrmll~tions used were:
A. DOTIM:cholesterol (1:1) DNA:lipid ratio 1:6 DNA c(>nr~ n 0.625mg/mL
B. DOTIM:cholesterol (1:1) DNA:lipid ratio 1:1 DNA concellL dLion 2mg/mL
C. DOTIM:DOPE (1:1) DNA:lipid ratio 1:1 DNA collc~l.L.dLion 2mg/mL
(DOPE is dioleoylphosphatidylethanolamine). Each formulation was pl~a..,d as described in E~L~..ples 1 and 3 above, and were ~ L~.~d by intravenous injection into the tail vein of cohorts of 3 ICR mice per tested formulation. Liposomes that were not complexed with DNA were injected into a sepdldLc cohort of 3 mice as a control.
Animals were sdc~irlced 1-2 days after injection and analyzed by histochrmir~l st~ining for human HLA expression. Tissues were analyzed for pe~c~ dge of cells in the tissue positive for human HLA ckl!lc~.sion in the hi~torhrmir~1 staining assay.
Results of these e~ lh~-e-lL~ are shown in Figure 12, wh.,l~ill Form~ tion A is MB102, Form~ ti-)n B is MB107 and Fo~.~Tm-l~tion C is MB163. For each formulation tested, some cells in each of the tissues were found to stain positive for human HLA expression. Lymph node staining varies most among different a~mini~tered form~ tions, with DOTIM:cholesterol at the higher (2mg/mL) DNA
~ 30 col~r~-l,dlion providing t'ne most human HLA positive cells, and the DOTIM:DOPE
f~rml-l~tir~n providing the least human HLA positive cells. The results in spleen were _ CA 02223923 1997-12-0~
WO 96/40962 PCT~,GJ'~5~6 less variable, with the DOTIM:cholesterol formulation at the lower (0.625mg/mL) DNA co.lcellL.aLion providing the most human HLA positive cells. Bone marrow cells showed high levels of human HLA positive cells with all formulations tested.
In view of these results, a series of ~ were ~ rulllled to delllol~Lldt~
S formulation~ pPn(~Pnt targeting of DNA:lipid complexes to spleen and lung. Two form~ tions were used:
a. DOTIM:cholesterol (1:1) DNA:lipid ratio 1:6 DNA col-r~ ion0.625mg/mL
b. DOTIM:DOPE (1:1) DNA:lipid ratio 1:1 DNA concentration l.Smg/mL
Each forrn~ tion was ~ ,d as de~clibed in Examples 1 and 3 above, using a CAT-encoding construct, and were ~ d by intravenous injection into the tail vein of cohorts of 3 ICR mice per tested form~ tion. Liposomes that were not complexed with DNA were injected into a scl,al~te cohort of 3 mice as a control.
Animals were sacrificed 1-2 days after injection and analyzed by CAT assay of lung and spleen tissues as described above. The results of these c~ nelll~ are shown in Figure 13, wherein MB102 is Forml~l~tion A and MB153 is Fonm-l~tion B.
CAT activity is expressed as the pcrcc.l~ge of total l4C-chloramphenicol counts converted to acetylated and diacylated forms associated with each tissue. As can be seen from the Figure, the DOTIM:cholest~-rol formulation ~I...i..;~l~led intravenously resulted in 96% (of over 1 million counts) being localized to lung tissue; 2% of the counts resulting from this formnl~tion were found in the spleen, and the rest were found in other tissues. In contrast, the DOTIM:DOPE forml-l~tion ~lmini~tered intravenously resulted in 91% (of 160,000 counts) being localized to spleen tissue, with about 3 % of the counts being found in the lung and the rest being found in other tissues. These results (1emnn~trate that this DOTIM:cholesterol formlll~tic)n specifir~lly targets the DNA:lipid complex to the lung, while the DOTIM:DOPE
formnl~ti~)n ~.ecirlcally targets DNA:lipid complexes to the spleen. In addition, these results show that CAT activity is about 10-fold more robust when delivered in DOTIM:cholesterol complexes to the lung than the CAT activity resulting from DOTIM:DOPE complex-mr~ tecl delivery to spleen.

_ W O 96/40962 PCT~U~51'0~3~6 D. I..l,ay~.;lo..cal Delivery Fc~ ,..c Liposome form~ tinns were developed for targeted gene delivery by hlLl~,iL~eal ~ .dLion. DOTIM:cholesterol form~ tions (1:1) were tested using a CAT-encoding construct at a DNA:lipid ratio of 1:1 and a total DNA
conce.lLldLion in the complex of 2.5mg/mL. An amount (lmL) of these DNA
complexes were injected into the ~clilul~al cavity of each of 4 mice; an equal volume of the liposome fnrm~ tinn not complexed with CAT-encoding DNA was injected into4 mice in a s~Ja~ cohort as a control. Ptlil(~-1cal ..~c.~,~hages were isolated 24-48h after injection and tested for CAT activity as described above.
The results of these ~ , are shown in Figure 14. F'eliL~ eal ~laclupllages from control (untreated) mice showed e~nti~lly no CAT activity in this assay. Macrophages from mice ~ d the DNA:lipid complexes in this formulation i~L.~peliLoneally showed high levels of CAT activity, dem~ L dLi..g specific in vivo delivery of a functional CAT gene using this form~ tion.
Spleens from these animals were also tested and CAT activity compared to p~liLo~cal mac-u~hages. These results are shown in Figure 15, where it can be seen that CAT activity in ...a~ hages was much higher than in spleen, demol~,LldLi"g specificity in L~ Lillg to these cells.
Pancreatic tissues were targeted for gene delivery using the DNA:lipid formlll~tions of the invention as follows. A form~ tion C(~ illg a CAT~ ',hlg plasmid and DOTIM:DOPE (1:1), at a DNA:lipid ratio of 1:1, and a total DNA
col-ce~L aLion in the complex of 1.5mg/mL was injected hlLldp~liLol1eally into a cohort of 3 mice. Two mice were injected with the liposome formulation not complexed with DNA as a control. P~ as and lung tissues were analyzed 24-48h post-~ o~ ldlion for CAT activity as described above.
The results of these ex.~e~ L~ are shown in Figure 16. These results delllol~7L dLe~ that this form~ ti~n ~ecirlcally targets delivery of CAT-encoding DNA
Cc~ lu~;L~ to the pancreas when ~ r-c:d h~ olleally.
A CAT-encoding ~co~bi~.anL construct was targeted to spleen using yet another DNA:lipid form~ tion. Plasmid DNA that was complexed with DOTIM:cholesterol (1:1), at a DNA:lipid ratio of 1:1, and a total DNA con~ ion in the complex of 2.5mg/mL was injected i..L.~eliloneally into a cohort of 3 mice.

W O 96/40962 PCT~US~','OS~>6 A Sc;p~alt: cohort of mice were injected with the liposome form~ ti~-n not complexed with DNA as a control. Spleen tissues from mice in each cohort were analyzed 2448h post~ . dLion for CAT activity as described above.
The results of these ~A~ are shown in Figure 17. These results demonstrate that this formlll~tion ~l,e~;ri~ ~lly targets delivery of CAT-encoding DNA
construets to spleen in vivo when ~ i";.~ ed inLld~eliLul1edlly.
The tissue specificity of hlLla~eliLul~~dl delivery was delllol~LldLt;d by COlll~dliSOll of two dirr~"e,.L formulations ~ "i..i.~ d illL~d~C.i~ eally. The following formulations were tested:
a. DOTIM:cholesterol (1:1) DNA:lipid ratio 1:1 DNA conce~Lldlion 2mg/mL
b. DOTIM:DOPE (1:1) DNA:lipid ratio 1:1 DNA concel,L,dLion 2mg/mL
Each formlll~tion was ~lcpalc:d as dc~,libed in E~ ,s 1 and 3 above, using a CAT-encoding coll~LlllcL, and were ~.I...;.);~I~l~,d by hlLldp~.iLoneal injection of cohorts of 3 ICR mice per tested fonmll~tion. Liposomes that were not complexed with DNA
were injected into a sepàlaLe cohort of 3 mice as a eontrol.
Animals were saerifieed 1-2 days after injection and analyzed by CAT assay of ~~al~cleas and spleen tissues as described above. The results of these e~p~ ll.,.
are shown in Figure 18, where MB153 is Formlll~tinn A, and MB152 is Fo~mll~tion B. CAT activity is expressed as the pel~el,L~ge of total l4C-chlOlalll~hc,licOl counts converted to acetylated and diacylated forms associated with each tissue. As can be seen from the Figure, the DOTIM:DOPE forrnlll~tinn a~ d i"l~dLIeliLoneally resulted in 96% (of 18 million counts) being loealized to pal~l.,aLic tissue; 3% of the counts resulting from this forrnlll:ltion were found in the spleen, and the rest were found in other tissues. In contrast, the DOTIM:cholesterol formlll~tinn ~.I"~i"i~t~ .~d i"l,ap~.iL)lleally resulted in 58% (of 28 million counts) being localized to spleen tissue, with about 42% of the counts being found in the ~al~cl.,as; esserlti~lly no CAT
aetivity was observed in other tissues. These results demonstrate that this DOTIM:DOPE formlll~tit n ~I,e~irlcally targets the DNA:lipid complex to the pancreas when ~ .ed inLIapcliLolleally, while the DOTIM:cholesterol formulation specifically targets DNA:lipid complexes to the pallc.~,as and spleen.

W O 96/40962 PCTrUS96/09526 E. Direct Delivery Fu~
Liposome formulations were developed for targeted gene delivery by direct injection into tissues. DOTIM:cholesterol form~ ti-ns (1:1) were tested using a CAT-encoding co~ u~;l at a DNA:lipid ratio of 1:1 and a total DNA conccl~L.dlion~ 5 in the complex of 2.5mg/mL. This formulation was directly injected in 1.5mL into a human pr~L~le ex corpora, and then assayed by CAT assay as dei,clil,ed above.
The results of this e~e~ L are shown in Figure 19. This Figure illustrates the resilts of four dirrer~"l p~ Ldte tissues tested, .lPt~oct~d as the amount of CAT
activity found in each of the Lldl~re~ d prostate tissues.
These results d~lllol~L aL~ that gene delivery can be ,.~I;AI~d by direct injection of DNA:lipid complexes on the invention into human tissues.

F. r~ ivJ-I of I~ d~.lu~ and Ildrd~ eal ~ iûn Rûutes The effect of ~.I...i..i!~l.dLion route on targeted delivery of CAT-encoding plasmid DNA using a single DNA:lipid complex form~ tiQn was d~Lelll~illed.
DOTIM:cholesterol formulations (1:1) were tested using a CAT-encoding construct at a DNA:lipid ratio of 1:1 and a total DNA conrel~l.A~ n in the complex of 2.5mg/mL. Cohorts of 3 mice were either injected intravenously in the tail vein, or i~lLldl)c~iLolleally. Spleen and lung tissues were analyzed 2448h post-~A,I-;,-i!il,dLion for CAT activity as described above. The results of these expcli",t;llL~ are shown in Figure 20. It can be seen from the Figure that the highest CAT activity levels were achieved in lung tissue following illlldVenOUS a~ alion of the form~ tion.
However, CAT activity after hlLld~.ilonedl ~ aLion was relatively higher in spleen than in lung. These results demonstrate that tissue-specific Ld,geLillg of DNA
delivery can be achieved with the same efficacious formulation of DNA:lipid c~ es, and that the targeted site can be i- .n~ ed by the route of ~-l. . .i"i~l . dtion.
It should be understood that the foregoing disclosure emphasizes certain specific embo-liml~nt~ of the invention and that all mo~1ifi-~tions or alternatives equivalent thereto are within the spirit and scope of the invention as set forth in the ~ 30 appended claims.

Claims (13)

WE CLAIM:

1. A pharmaceutical composition comprising a formulation of a soluble complex of a recombinant expression construct and a mixture of a neutral lipid and a cationic lipid in a pharmaceutically acceptable carrier suitable for administration to an animal by intravenous, intraperitoneal or direct injection into a tissue in the animal, wherein:
(a) the recombinant expression construct comprises a nucleic acid encoding a transcription product and wherein said nucleic acid is operatively linked to gene expression regulatory elements whereby the nucleic acid is transcribed in vivo; and (b) the cationic lipid is a compound having formula I:

where R and R1 independently straight-chain, aliphatic hydrocarbyl groups of from about 11 to 29 carbon atoms, wherein the cationic lipid and the neutral lipid are present in the complex at a molar ratio of about 1:1, the DNA and the cationic lipid are present in the complex in a ratio of from about 1:1 to about 1:8 µg DNA/nmole cationic lipid, and the nucleic acid comprising the recombinant expression construct is present in the complex at a concentration of about 0.5 mg/mL to about 5mg/mL.

2. The pharmaceutical composition of Claim 1 wherein the cationic lipid is 1-(2-(oleoyloxy)ethyl)-2-oleyl-3-(2-hydroxyethyl)imidazolinium chloride.

3. The pharmaceutical composition of Claim 1 wherein the neutral lipid is cholesterol or dileoylphosphatidylethanolamine.

4. The pharmaceutical composition of Claim 1 wherein the cationic lipid is 1-(2-(oleoyloxy)ethyl)-2-oleyl-3(2-hydroxyethyl)imidazolinium chloride and the neutral lipid is cholesterol or dileoylphosphatidylethanolamine.

5. The pharmaceutical composition of Claim 1 wherein the complex of the recombinant expression construct and a cationic lipid comprises a ratio of DNA to cationic lipid of about 1:6 to about 1:8 µg DNA/nmole cationic lipid.

6. The pharmaceutical composition of Claim 1 wherein the complex of the recombinant expression construct and a mixture of a neutral lipid and a cationic lipid comprises a ratio of DNA to cationic lipid of about 1:1 µg DNA/nmole cationic lipid.

7. Use of a pharmaceutical composition of Claim 1 to prepare a medicament for introducing a recombinant expression construct into a cell comprising lung tissue in an animal wherein the pharmaceutical composition is administered to the animal by intravenous injection, wherein the cationic lipid is 1-(2-(oleoyloxy)ethyl)-2-oleyl-3-(2-hydroxyethyl)imidazolinium chloride, the neutral lipid is cholesterol, the cationic lipid and the neutral lipid are present in a molar rario of about 1:1, the complex of the recombinant expression construct and a mixture of a neutral lipid ard a cationic lipid comprises a ratio of DNA to cationic lipid cf about 1:6 µg DNA/nmole cationic lipid, and the DNA concentration in the DNA:lipid complexes is from about 0.5mg/mL to about 1mg/mL.

8. Use of a phamaceutical composition of Claim 1 to prepare a medicament for introducing a recombinant expression construct into a cell within a spleen in an animal, wherein the pharmaceutical composition is administered to the animal by intravenous injection, wherein the cationic lipid is 1-(2-(oleoyloxy)ethyl)-2-oleyl-3-(2-hydroxyethyl)imidazolinium chloride, the neutral lipid is dioleoylphosphatidyl-ethanolamine, the cationic lipid and the neutral lipid are present in a molar ratio of about 1:1, the complex of the recombinant expression construct and a mixture of a neutral lipid and a cationic lipid comprises a ratio of DNA to cationic lipid of about 1:1 µg DNA/
nmole cationic lipid, and the DNA concentration in the DNA:lipid complexes is from about 1mg/mL to about 2.5mg/mL.

9. Use of a pharmaceutical composition of Claim 1 to prepare a medicament for introducing a recombinant expression construct into a cell that is a is a peritoneal macrophage in an animal, wherein the pharmaceutical composition is administered to the animal by intraperitoneal injection, wherein the cationic lipid is 1-(2-(oleoyloxy)ethyl)-2-oleyl-3-(2-hydoxyethyl)imidazolinium chloride, the neutral lipid is cholesterol, the cationic lipid and the neutral lipid are present in a molar ratio of about 1:1, the complex of the recombinant expression construct and a mixture of a neutral lipid and a cationic lipid comprises a ratio of DNA to cationic lipid of about 1:1 µg DNA/nmole cationic lipid, and the DNA concentration in the DNA:lipid complexes is from about 1mg/mL to about 2.5mg/mL.

10. Use of a pharmaceutical composition of Claim 1 to prepare a medicament for introducing a recombinant expression construct into a cell within the spleen of an animal, wherein the pharmaceutical composition is administered to the animal by intraperitoneal injection, wherein the cationic lipid is 1-(2-(oleoyloxy)ethyl)-2-oleyl-3-(2-hydroxyethyl) imidazolinium chloride, the neutral lipid is cholesterol, the cationic lipid and the neutral lipid are present in a molar ratio of about 1:1, the complex of the recombinant expression construct and a mixture of a neutral lipid and a cationic lipid comprises a ratio of DNA to cationic lipid of about 1:1 µg DNA/nmole cationic lipid.
and the DNA concentration in the DNA:lipid formulation is from about 1mg/mL to about 2.5mg/mL.

11. Use of a pharmaceutical composition of Claim 1 to prepare a medicament for introducing a recombinant expression construct into a cell within the pancreas of an animal, wherein the pharmaceutical composition is administered to the animal by intraperitoneal injection, wherein the cationic lipid is 1-(2-(oleoyloxy)ethyl)-2-oleyl-3-(2-hydroxmethyl) imidazolinium chloride, the neutral lipid is cholesterol, the cationic lipid and the neutral lipid are present in a molar ratio of about 1:1, the complex of the recombinant expression construct and a mixture of a neutral lipid and a cationic lipid, comprises a ratio of DNA to cationic lipid of about 1:1 DNA/nmole cationic lipid, and the DNA concentration in the DNA:lipid formulation is from about 1.5mg/mL toabout 2.5mg/mL.

12. Use of a pharmaceutical composition of Claim 1 to prepare a medicament for introducing a recombinant expression construct into a cell in an animal wherein the pharmaceutical composition is administered to the animal by direct injection into the tissue in the animal, wherein the cationic lipid is 1-(2-(oleoyloxy)ethyl)-2-oleyl-3-(2-hydroxyethyl)imidazolinium chloride, the neutral lipid is cholesterol, the cationic lipid and the neutral lipid are present in a molar ratio of about 1:1, the complex of the recombinant expression construct and a mixture of a neutral lipid and a cationic lipid, comprises a ratio of DNA to cationic lipid of about 1:1 µg DNA/nmole cationic lipid, and the DNA concentration in the DNA:lipid complexes is from about 1mg/ml to about 2.5mg/mL.

13. Use of a pharmaceutical composition of Claim 1 to prepare a medicament for introducing recombinant expression construct into a cell comprising brain tissue in an animal wherein the pharmaceutical composition is administered to the animal by direct intracranial injection.