WO2000012114A1 - Peptide scaffolds for transfer of molecules into eukaryotic cells - Google Patents

Abstract

Compositions and methods of using compositions with a nuclear targeting peptide containing a nonclassical nuclear localization signal to deliver selected molecules to the nucleus of eukaryotic cells are provided. The compositions are particularly useful in gene transfer methods.

Description

PEPTIDE SCAFFOLDS FOR TRANSFER OF MOLECULES INTO

EUKARYOTIC CELLS

Field of the Invention

The present invention relates to peptide based compounds that overcome the final rate limiting step of nuclear entry encountered during lipofection of plasmids to nondividing cells thereby increasing the efficiency of nonviral gene transfer methodologies (i.e. lipofection). The present invention also relates to methodologies for delivering plasmids into the nucleus of mammalian cells, including nondividing mammalian cells via these peptide based compounds. The compounds and methods of the present invention rely upon nuclear targeting peptides (NTPs) comprising nonclassical nuclear localization signals of eukaryotic cells such as mammalian cells to exploit endogenous import and export mechanisms of ribonucleic acid trafficking to transfer selected molecules including genes to the nucleus of the cells. An example of an NTP useful in the present invention is the M9 sequence of the heteronuclear ribonuclear protein type 1 (hnRNP Al) . One embodiment of the present invention involves the chemical conjugation of a small cationic peptide with known DNA complexing/compacting activity, termed the scaffold, to an NTP with an engineered carboxy terminal cysteine residue which facilitates conjugation. As demonstrated herein, compositions comprising an NTP containing a nonclassical nuclear localization signal or scaffold-NTP conjugates of the present invention are highly efficient in transfecting nondividing mammalian cells with selected molecules .

Background of the Invention

Gene transfer by nonviral methodologies (i.e. lipofection) has been found to be very inefficient in nondividing cell populations. Unfortunately, cells are not actively dividing in many in vivo tissues that are potential clinical targets for gene therapy. While receptor targeting, fusigenic peptides, or endosome disrupting agents help overcome some of the first barriers that limit liposome-based gene delivery, thus far virus free gene transfer using liposo es has had limited clinical utility because of the difficulty of transporting genetic material into the nucleus of a nondividing cell. Cationic liposomes are commonly used as a means of delivering DNA to dividing cells in culture. DNA/lipid aggregates form spontaneously after mixing aliquots of the liposome reagent with an aqueous solution of DNA (Feigner et al. Proc . Na tl Acad. Sci . U.S.A. 1987 84, 7413; Feigner et al . J. Biol . Chem . 1994 269, 2550; Feigner, P.L. and Ringold, G.M. Na ture 1989 337, 387). However, the efficiency of transfection is low with nondividing cells and certain target cells such as endothelium. For example, COS-7 cells lipofected with chloramphenicol acetyl transferase (CAT) gene are nearly 80% CAT positive compared to essentially no CAT expression in endothelial cells (Nathwani et al . Bri t . J. Haem . 1994 88, 121) .

Arterial gene transfer has been proposed for the treatment of atherosclerosis and restenosis following angioplasty. The two major strategies of arterial gene transfer are: (1) the cultivation, transfection and reintroduction of autologous endothelial cells to the recipient, (Dichek et al. Circula tion 80, 1347, 1989; Wilson et al. Science . 244, 1344, 1989) and (2) the direct catheter delivery of genetic material with transfection vehicle or virus to the cells of the artery (Nabel et al . Science 244, 1342, 1989; Le archand et al . Proc . Natl Acad. Sci . 89, 6482, 1992; Flugelman et al . Circula tion . 85, 1110, 1992; Takeshita et al. J. Clin . Invest . 93, 652, 1994; Nabel E.G., J. Vase . Sur. , 15, 931, 1992; Chapman et al . Circ . Res . 71 , 27, 1992; Schulick et al . Circ . Res . 11 , 475, 1995) . While receptor targeting of vectors to endothelium may provide benefit, the endothelium is ideally accessible by catheter. In vivo, endothelial cells have a very low mitotic rate. Hence, while adenoviral gene transfer (Flugelman et al . Circ . 1992 85:1110- 1117; Nabel et al . Science 1989 244:1342-1344; Subramanian, A. and Diamond, S.L. Tissue Engineering 1997 3:39-52; Lemarchand et al . Circ . Res . 1993 72:1132-1138; Steg at al . Circ . 90:1648-1653; Sung et al . Circ . Res . 1993 73:797-807; Schachtner et al . Circ . Res . 1995 76:701-709) is highly efficient in transfecting these cells, gene transfer via liposome mediated routes and retroviral vectors which are more efficient in mitotic cells (Flugelman et al . Circula tion . 85, 1110, 1992 1992) result in low levels of gene transfer.

Thus, various attempts have been made to improve efficiency of gene transfer of nondividing cells via liposome mediated routes.

Recent improvements of lipofection protocols for subconfluent epithelium tend to plateau at about 30% transfection efficiency: α₅β₁ integrin targeting peptide results in 25% transfection of corneal endothelium (Hart et al. Hum . Gene Ther . 1998 9:575-585); replication-deficient AdV-complexed plasmid results in 20% transfection of HUVEC (Edgell et al . Biotechniques 1998 25:264-268) and 25 to 35% transfection of BAEC (Go et al . Am . J. Physiol . 1998 274 :H1-

H7 ) ; histone complexation of plasmid results in 20% transfection of bovine aortic endothelial cells (BAEC)

(Subramanian, A. and Diamond, S.L. Tissue Engineering 1997 3:39-52); AdV fiber added to plasmid results in 30% transfection of BAEC (Hong et al. Chin . Med . J. 1995 108:332- 338) and plasmid condensed with recombinant histone HI containing SV40 T-antigen NLS resulted in 10 to 30% transfection of COS-7 or NIH3T3 (Fritz et al . Hum Gene Ther . 1996 7:1395-1404). It is not known whether the benefits of these protocols can be achieved when using confluent cells at the time of transfection. The 30% plateau is believed to represent the persistence and elevated level of intact cytoplasmic plasmid available to accomplish gene transfer in cells dividing at times one to two days after transfection. An unprotected plasmid with short half-life in the cytoplasm would transfect only the cells dividing in the first few hours after the lipofection. A large amount of labeled-plasmid transfected into the cells via these protocols was found to be present in the endosomes as indicated by punctate staining. However when the plasmids were transcribed cytoplasmically using a T7 RNA polymerase based system, 80% of cells were found to express β-galactosidase, indicating that some of the plasmid does get out of the endosome into the cytoplasm in the cells.

However, plasmid DNA cannot readily enter the nucleus since they are typically excluded by the nuclear pore (Feigner et al. Proc . Na tl Acad. Sci . USA 1987 84:7413-7417; Feigner, P.L. and Ringold, G.M. Nature 1987 337:387-388; Jo et al . J. Biol . Chem . 1997 272:1395-1401; and Zabner et al . J. Biol . Chem . 1995 270:189997-19007). Increasing cytoplasmic concentrations of plasmid can directly enhance total expression in dividing cells by enhancing plasmid levels in the nucleus, post-mitotically . It is believed that the frequency of nuclear import events in nondividing cells increases with elevated plasmid levels through enhancement of the probability of plasmid encounter with the nuclear pore entrance. However, inefficient nuclear pore targeting due to cytoplasmic sequestration (scaffolding) and inefficient transit of plasmid across the pore remain important rate limits during nonviral gene transfer.

Attempts to conjugate classical nuclear localization signals (NLS) to a plasmid to alleviate this problem have had limited success (Subramanian, A. and Diamond, S.L. Tissue Engineering 1997 3:39-52; Fritz et al . Hum . Gene Ther . 1996 7:1395-1404). Classical nuclear localization signals involve peptide sequences of clustered residues that interact with two proteins, importin- and importin-β, also known as karyopherin and karyopherin β, respectively. The protein importin-β binds the nuclear pore. The importin complex also binds a GTPase RNA (Nakielny et al. Exp . Cell Res . 1996 229:261-266) . Examples of classical NLS include SV40 large T antigen (PPKKKRKV; SEQ ID NO: 6), adenovirus E1A (SCKRPRP; SEQ ID NO:7), human lamin A (SVTKKRKL; SEQ ID NO:8); polyoma large

T antigen (PPKKARED; SEQ ID NO : 9 ) , polyoma large T antigen

(VSRKRPRP; SEQ ID NO:10), human c-myc (PAAKRVKL; SEQ ID

NO:ll), rat glucocorticoid receptor (RKTKKKIK; SEQ ID NO:12), and human estrogen receptor (IRKDRRG; SEQ ID NO:13). The classical nuclear localization sequence can also contain a bipartite form, with two basic amino acids separated by an amino acid spacer from a second cluster of three or more basic amino acids. A prototypical bipartite NLS is found in nucleoplasmin (AVKRPAATKKAGQAKKKKLD; SEQ ID NO:14). It has been found that histones (Subramanian, A. and Diamond, S.L. Tissue Engineering 1997 3:39-52; Fritz et al . Hum . Gene Ther. 1996 7:1395-1404) or SV40 T antigen (Sebestyen et al . Na ture Biotech . 1998 16:80-85) when linked covalently or by charge interactions to plasmid have not been able to get plasmid across the nuclear pore of intact cells to effect fully efficient gene transfer with 100% transfection.

In the present invention methods and compositions are providing for delivering selected molecules to the nuclei of eukaryotic cells via nuclear targeting peptides containing nonclassical nuclear localization signals.

Summary of the Invention

An object of the present invention is to provide methods and compositions for delivering selected molecules to the nucleus of eukaryotic cells via targeting of nonclassical nuclear import pathways. In one embodiment, compositions comprising a nuclear targeting peptide containing a nonclassical nuclear localization signal which interacts with transportin are used to mediate nuclear pore targeting and import of molecules into the nucleus of the cells . By "selected molecule" it is meant to include, but is not limited to, organic molecules, polymers, proteins and nucleic acids useful for either therapeutic or research purposes. Another object of the present invention is to provide a compound comprising a cationic peptide scaffold and a nuclear targeting peptide conjugated to the cationic scaffold via a hydrolytic resistant linkage. This compound is referred to herein as a Scaffold-NTP conjugate. Another object of the present invention is to provide compositions suitable for lipofection of mammalian cells with a selected nucleic acid sequence which comprise a selected nucleic acid sequence, a peptide scaffold and an NTP. In a preferred embodiment, a complex is formed between a plasmid containing the selected nucleic acid sequence and a Scaffold- NTP conjugate.

Another object of the present invention to provide a method for expressing a selected nucleic acid sequence in mammalian cells which comprises contacting cells with a mixture of a selected nucleic acid sequence, a peptide scaffold and an NTP.

Another object of the present invention is to provide a method for expressing a selected nucleic acid sequence in mammalian cells which comprises forming a complex between a plasmid containing the selected nucleic acid sequence and a scaffold-NTP conjugate; and contacting cells with the complex.

Another object of the present invention is to provide cells transfected with a complex comprising a plasmid containing a selected nucleic acid sequence and a scaffold-NTP conjugate. Yet another object of the present invention is to provide a method of treating a patient suffering from a condition associated with an absence in the expression of a normal selected nucleic acid sequence comprising administering to the patient a composition comprising a complex formed between a plasmid containing the selected nucleic acid sequence and a Scaffold-NTP conjugate.

Brief Description of the Drawings

Figure 1 is a bargraph showing the enhancement of total expression of β-galactosidase activity following complexation of pCMVβgal with a nuclear targeting peptide containing the M9 sequence. In all samples, cells counts were kept at 10⁷ cells/ml .

Figure 2 provides results from experiments demonstrating lipofection of confluent BAEC with pCMVβgal in the presence and absence of peptides. Figure 2A shows the fluorescence of untransfected BAEC. These cells were used as a measure of endogenous galactosidase activity and FDG autofluorescence . Figure 2B shows the fluorescence of confluent BAEC transfected with 1 μg plasmid alone. Figure 2C shows the fluorescence of confluent BAEC transfected with 1 μg plasmid plus 60 μg of M9 peptide. Figure 2D shows the fluorescence of confluent BAEC transfected with 1 μg plasmid plus 30 μg of M9 plus 30 μg of ScT (unconjugated) . Figure 2E shows the fluorescence of confluent BAEC transfected with 1 μg plasmid + 60 μg M9-ScT conjugate. Percent transfection is defined in each panel by the percent of cells with > 100 f.u.

Detailed Description of the Invention

The present invention relates to a new approach for high efficiency nonviral gene transfer across the nuclear pore by exploiting endogenous import and export mechanisms of ribonucleic acid trafficking which utilize nuclear targeting peptides containing nonclassical nuclear localization signals. Peptide based compounds that overcome the final rate limiting step of nuclear entry encountered during lipofection of plasmids to nondividing cells have now been designed and demonstrated to be effective at transfecting mammalian cells, and in particular nondividing mammalian cells.

The compounds of the present invention comprise a nuclear targeting peptide, termed the NTP. The NTP comprises a peptide sequence containing a nonclassical, nuclear localization signal (NLS) of eukaryotic cells. Examples of nonclassical nuclear localization signals include, but are not limited to, M9 of hnRNP Al, KNS of hnRNP K, and HNS of HuR. In contrast to classical nuclear localization signals, these nonclassical NLS do not interact with proteins such as importin- and importin-β. Instead, the nonclassical NLS of hnRNP Al, termed the M9 sequence, binds a 90 kDa intracellular protein called transportin which mediates nuclear pore targeting and import. The nonclassical NLS of HuR, termed the HNS, shows some homology with the M9 sequence. The nonclassical NLS of HnRNP K, termed the KNS, utilizes an import pathway independent of either the M9 or the classical NLS import pathway and has been suggested to interact directly with the nuclear pore (Michael et al . The EMBO J. 1997 16:2587-3598) . In a preferred embodiment the NTP comprises the M9 sequence (NQSSNFGPMKGGNFGGRSSGPYGGGGQYFAKPRNQGGY; SEQ ID NO: 3) of the heteronuclear ribonuclear protein type 1 (hnRNP Al) protein or a similar NTP with the ability to interact with transportin to mediate nuclear pore targeting and import of selected molecules.

The M9 epitope of human hnRNP Al is a small, potentially nonimmunogenic sequence with no homology with known fusigenic epitopes . This nonclassical, NLS is not highly cationic and does not display strong DNA binding properties. Thus, in one embodiment of the present invention, the M9 sequence is engineered with a carboxy terminal cysteine residue (NQSSNFGPMKGGNFGGRSSGPYGGGGQYFAKPRNQGGY-GGGC; SEQ ID NO : 1 ) to facilitate conjugation to a scaffold peptide to improve the DNA binding properties of the NTP. It is preferred that the NTP be con ugated to a small cationic peptide with known DNA complexing/compacting activity, termed the scaffold. Peptide sequences ranging in length from 5 to 200 ammo acids and enriched m basic am o acid residues including lysme, arg ine and histid e provide useful scaffolds . Also useful as scaffolds are peptide sequences containing the consensus sequence RNA-b dmg domain (CS-RBD) which is found in hnRNP Al . Examples of peptides capable of acting as scaffolds in the compounds of the present invention include, but are not limited to, HIV rev (TRQARRNRRRRWRERQ; SEQ ID NO: 15), HIV tat (ALGISYGRKKRRQRRP; SEQ ID NO: 16), AN (MDAQTRRRERRAEKQAQW; SEQ ID N0:17), and φ21N (GTAKSRYKARRAELIAER; SEQ ID NO:18). A preferred scaffold sequence useful in the present invention comprises a scrambled SV40 T antigen (ST), also referred to herein as a mutant SV40 T antigen, which has five positively charged ammo acids (VKKGKCRPGKGYG; SEQ ID NO : 2 ) .

The peptide scaffold and the NTP are preferably conjugated via a hydrolytic-resistant chemical linkage. In one embodiment, chemical conjugation of the peptide scaffold to the NTP is achieved using a crosslmker such as succmimidyl 4- (N-maleimidomethyl) cyclohexane-I (SMCC). However, other homobifunctional or heterobifunctional crosslmkers are well known and used routinely by those of skill m the art. For example, for conjugation of an NTP containing a terminal cysteine and a scaffold peptide containing a cysteine, the following homobifunctional linkers can be used in a two step reaction: sulfhydryl to sulfhydryl homobifunctional crosslmker: bis-maleimidoethane; bis- maleimidohexane; 1, 4-bιs-maleιmιdyl-2 , 3-dιhydroxybutane; or 1, 4-bιs-maleιmιdobutane . For conjugation of an NTP containing a terminal cysteine and a scaffold peptide containing a terminal am e, the following heterobifunctional linkers can be used in a two step reaction: sulfhydryl to ammo heterobifunctional crosslinker: N-succinimidyl [ 4 - iodoacetyl] aminobenzoate; N-succinimidyl iodoacetate; succinimidyl 3- [bromoacetamido] propionate; succinimidyl 4-[p- maleimidophenyljbutyrate; succinimidyl-6- [ ( β - maleimidopropionamido) hexanoate] ; N-succinimidyl 3- [2- pyridyldithio] propionate; N- [γ-maleimidobutyryloxy] sulfo- succinimide ester; N- [ K-maleimidoundecanoyloxy ] - sulfosuccinimide ester; m-maleimidobenzoyl-N- hydroxysulfosuccinimide ester; or N-succinimidyl- [4- vinylsulfonyl] benzoate .

Scaffold-NTP conjugates can also be prepared via recombinant expression through bacterial or eukaryotic expression systems of a protein containing a cationic charge and an NTP such as the M9 sequence. Bacterial and eukaryotic expression systems are well known and used routinely by those skilled in the art.

A Scaffold-NTP conjugate of the present invention, M9- ScT has been demonstrated to be highly efficient in transfecting confluent, nondividing large artery aortic mammalian endothelial cells.

In experiments lipofecting confluent bovine aortic endothelial cells with M9-ScT/rhodamine plasmid, the majority of fluorescence was endosomal indicating that the M9 sequence is not fusigenically active. When confluent bovine aortic endothelial cells (BAEC) were lipofected with rhodamine- plasmid for 16 hours, highly punctate staining was observed in the cytoplasm of every cell, demonstrating that endosomal compartmentalization of plasmid was sustained. Nuclear levels in these cells were extremely low in comparison. Highly confluent BAEC transfected for 20 hours with pTMβgal and T7 RNA polymerase to allow for cytoplasmic transcription of the plasmid resulted in nearly 80% of the cells expressing β- galactosidase . Without addition of the T7 RNA polymerase, the flow cytometry signal was due to the level of endogenous activity and FDG autofluorescence detected in untransfected cells. Endosome escape, though not efficient during lipofection, was widely detectable in the population of confluent BAEC. Xgal staining of confluent BAEC lipofected with pTMβgal/T7 RNA polymerase demonstrated that over half of the cells were intensely blue and the other half were moderately blue.

Highly confluent BAEC monolayers transfected with plasmid alone resulted in about 1% of the cells staining Xgal positive for β-galactosidase . In contrast, nearly all the cells displayed Xgal staining to varying extent when lipofected with pCMVβgal complexed with the M9-ScT conjugate. Confluent BAEC were then lipofected with pCMVβgal complexed to various combinations of peptides and analyzed by flow cytometry (see Figure 2A-E) . The majority of untransfected cells had a fluorescence of < 10 f.u. This low level of fluorescence was due to endogenous galactosidase activity and FDG autofluorescence (see Figure 2A) . When confluent BAEC were transfected with pCMVβgal alone (see Figure 2B) , the majority of cells displayed a signal above 10 f.u. demonstrating that lipofection provided for extremely low level, but widespread expression that is below the sensitivity of Xgal staining. Only 5.27% of cells lipofected with plasmid alone had fluorescence above 100 f.u. This level increased to 55% when the plasmid was precomplexed with M9 peptide alone (see Figure 2C) . Adding M9 and ScT to the plasmid without chemical conjugation resulted in a transfection efficiency of 76.4% (see Figure 2D) . This marked increase with M9 alone or M9 + ST (uncon ugated) was in the absence of the nonpeptide SMCC crosslmker. The highest transfection efficiency of 83% was obtained when confluent BAEC were lipofected with plasmid complexed with M9-ScT indicating that binding of M9 peptide to the cationic ScT peptide led to increased transfection efficiency (see Figure 2E) . In highly confluent monolayers of BAEC, the use of M9-ScT conjugate provided for a 15.7-fold increase the percent positive cells with >_. 10' f.u. measured by flow cytometry compared to lipofection with plasmid alone.

The total β-galactosidase expression in the cells was measured fluorometrically (see Figure 1) . Complexation of plasmid with M9 peptide alone provided a 18.3-fold increase in expression. This increase was not simply due to a condensation effect since ScT which condenses plasmid effectively by electrostatic interactions (as observed by an ethidium bromide fluorescence release assay) provided only a 4-fold increase in β-galactosidase expression. A similar level of increase in β-galactosidase expression was observed when the plasmid was complexed with SV40 T-antigen consensus NLS or with polylysine (13 amino acid peptide), indicating that classical NLS (or sequences with equivalent charge density and size) were markedly less efficient in targeting the plasmid to the nucleus as compared to M9 NLS. Transfection of confluent BAEC with plasmid precomplexed with M9-ScT conjugate resulted in a 63-fold enhancement in expression over transfection with plasmid alone, while M9+ST (unconjugated) complexed to plasmid resulted in a 47-fold increase.

To investigate nuclear localization activity of M9- peptide, confluent BAEC were treated with digitonin to permeabilize the plasmalemma and then incubated with fluorescent plasmid ± M9-ScT conjugate for 30 minutes. The cells incubated with plasmid alone displayed cytoplasmic and perinuclear staining but no nuclear staining. Nuclear import of the complex of plasmid + M9-ScT conjugate was not observed in cells pretreated with wheat germ agglutinin (WGA) to block nuclear pores. However, when cells were incubated with plasmid complexed with M9ScT, the amount of plasmid in the nucleus increase markedly.

Similarly, complexation of plasmid with a cationic peptide alone that aids in plasmid condensation and protection against DNases provided only a modest 4-fold enhancement in expression. Thus, by flow cytometry, transfection of confluent mammalian cells with plasmid alone resulted in only 5% of the cells testing positive, while 76.4% of the cells were positive when a mixture of M9, ScT and plasmid were added to the cells without chemical conjugation and 83% of the cells were positive when the plasmid was complexed with the M9-ScT compound of the present invention. This compound, M9-ScT, caused a 60-fold increase in total β-galactosidase expression over transfection with plasmid alone. These striking differences were seen in replicate cultures of confluent cells with the same cell density at the time of transfection or at the time of assay. The role of enhanced nuclear import by this compound of the present invention is further supported by results from a nuclear import assay. Permeabilized BAEC cells incubated with plasmid labeled with rhodamine showed very distinct cytoplasmic and perinuclear staining but no staining in the nucleus. When rhodamine-plasmid was complexed with the M9-ScT conjugate, accumulation was observed in the nucleus. This accumulation was inhibited by preincubation of the cells with an agent which blocks nuclear pores.

Accordingly compositions of the present invention comprising a nuclear targeting peptide containing a nonclassical NLS are useful in gene transfection methods. In one embodiment, the NTP, preferably the M9 sequence, is simply mixed with a plasmid containing the selected nucleic acid sequence. In another embodiment, a peptide scaffold is also added to the mixture. Cells are then contacted with the mixture so that the plasmid and selected nucleic acid sequence are transfected into the cells and the selected nucleic acid sequence is expressed. More preferably, cells are contacted with a composition comprising a complex of a plasmid containing a selected nucleic acid sequence and a scaffold-NTP conjugate of the present invention under conditions in which the plasmid is transfected into the cells and the selected nucleic acid sequence is expressed. Compositions of the present invention are useful in transfecting any mammalian cells with a selected nucleic acid sequence. However, the compositions of the present invention are particularly useful in transfecting nondividing mammalian cells such as endothelial cells. The compositions of the present invention are useful in gene transfer methods used to treat patients suffering from conditions associated with an absence in the expression of a normal selected nucleic acid sequence. This condition can thus result from no expression of the nucleic acid sequence or expression of a mutated nucleic acid sequence. Compositions of the present invention are believed to particularly useful in arterial gene transfer methods used in the treatment of atherosclerosis and restenosis following angioplasty . Targeting of nonclassical nuclear import pathways via the compositions and methods of the present invention can also be used to deliver other selected molecules to the nucleus of eukaryotic cells. In addition to selected nucleic acid sequences, by "selected molecule" it is also meant to include, but is not limited to, organic molecules, polymers, and proteins. In a preferred embodiment, compositions comprising a nuclear targeting peptide containing a nonclassical nuclear localization signal which interacts with transportin such as the M9 sequence are used to mediate nuclear pore targeting and import of molecules into the nucleus of the cells. Eukaryotic cells can be contacted with a mixture of these compositions and a selected molecule to mediate nuclear pore targeting and import of the selected molecules into the nucleus of the cells . The following nonlimiting examples are provided to further illustrate the present invention. EXAMPLES

Example 1 : Cell Culture

Bovine aortic endothelial cells (BAEC) (passage < 10) are grown to confluence, passed to a 6-well culture dish at a split ratio of 1:3, and then grown to confluence as described by Subramanian, A. and Diamond, S.L. Tissue Engineering 3, 39-52, 1997. Growth medium is Dulbecco's modified Eagle's medium (DMEM) containing 10% heat inactivated newborn calf serum, 0.30 mg/ml of glutamine, 150 U/ml penicillin and 0.15 mg/ml streptomycin (GIBCO, Grand Island, NY). Hela cells (Hela S3 ATCC No. CCL 2.2) are grown in DMEM media (Gibco BRL) supplemented with 10 % fetal calf serum as described by Ranjan, V. and Diamond, S.L. Biochem . Biophys . Res . Comm . 196, 79, 1993.

Example 2: Transfection Procedure

Cationic liposomes (Lipofectamine Reagent, Gibco-BRL) , containing a 3:1 by weight mixture of polycationic lipid 2,3- dioleoyloxy-N- [2 (sperminecarboxamido) ethyl] -N, N-dimethyl-1 propanaminium trifluoroacetate (DOSPA) and the neutral lipid dioleoyl phosphatidylethanolamme (DOPE) was used in all transfections in accordance with procedures described by Subramanian, A. and Diamond, S.L. Tissue Engineering 3, 39-52, 1997. Pure liposome (10 μl) diluted to 100 μl are added to the DNA/scaffold complex, incubated for 45 minutes. Then 0.8 ml of OPTIMEM media is added to this mixture which is then overlaid on the cells. The cells are incubated for 2 hours at 37°C. After two hours, the mixture is aspirated, 4% serum in OPTIMEM media is added, and the cells are allowed to grow for 48 hours. The cells are then fixed and assayed for β-gal activity. During the transfections, the cells are continuously monitored for cell viability and adhesion. Example 3: βgal assays - % transfection, total expression and flow cytometry

To assay for βgal to measure % transfection, cells are stained with XGAL as described by Subramanian, A. and Diamond, S.L. Tissue Engineering 3, 39-52, 1997. The percent of cells stained per field of view is counted using a Zeiss 125M inverted microscope. The count is averaged from at least 4 fields of view for each transfected monolayer. To measure β- gal expression quantitatively, cell lysates are prepared and tested for β-gal activity against the fluorogemc substrate, fluorescein di-β-D-galactopyranoside (FDG, F- 1179 Molecular Probes) which is not fluorescent until cleaved to fluorescein monogalactoside and then to fluorescein (excitation/emission : 494/520 nM) . For flow cytometry, the trypsmized cell suspension is supplemented with serum, pelleted (100 g) and resuspended m staining medium to approximately 10 cells/ml. A 100 μL cell suspension is maintained at 37°C and 100 μl of 2 mM FDG maintained at 37°C is then added to the cell suspension, mixed rapidly and incubated at 37°C for 1 mmute. The FDG loading is stopped at the end of 1 mmute by adding 1.8 mL of ice-cold staining medium containing 1.5 μM Propidium Iodide and 1 mM PETG. The cells are kept on ice prior to flow cytometer analysis using a Becton Dickinson FACScan system. Unstained BAEC cells allow compensation for autofluorescence . The histogram showing fluorescence/cell to the number of cells is used to calculate the % transfection.

Example 4 : Domains for nuclear targeting

The M9 sequence of hnRNP Al is a 38 ammo acid sequence that provides for import/export of a fusion β-gal-M9 marker. The deletion of 5 ammo acids from the carboxy or ammo terminus abolishes its targeting activity (Michael et al . Cell

83,415-422, 1995). The M9 sequence is:

NQSSNFGPMKGGNFGGRSSGPYGGGGQYFAKPRNQGGY (SEQ ID NO : 3 ) For purposes of chemical conjugation, the M9 sequence has been synthesized with a carboxy terminus addition of Gly-Gly-Gly- Cys to give an accessible thiol group provided by the cysteine : Nuclear Targeting Peptide (NPS) =

NQSSNFGPMKGGNFGGRSSGPYGGGGQYFAKPRNQGGY-GGGC (SEQ ID NO:l)

The K nuclear shuttling domain (KNS) of the hnRNP K protein is a 39 amino acid sequence: YDRRGRPGDRYDGMVGFSADETWDSAIDTWSPSEWQMAY (SEQ ID NO : 4 )

Example 5 : Chemical conjugation of nuclear targeting epitopes to cationic scaffolds

Several different cationic scaffolds that are rich in amine suitable for conjugation reactions and which mediate electrostatic complexation or condensation with plasmid are used. These scaffolds can include: SV40 T antigen NLS (SVT = CGYGPKKKRKVGG (SEQ ID NO : 5 ) ) , a mutated version of the SV40 T antigen NLS (muT = VKKGKCRPGKGYG (SEQ ID NO : 2 ) ) , poly-L- lysine (MW 1, 4, and 30-70 kDa), histone HI, and hydrophilic amine-terminated dendrimers (87,340 MW) of small size (8.4 nm) available through Polysciences, Inc. (Warrington, PA). These scaffolds are tested for DNA condensation using a fluorescence quench assay of ethidium bromide labeled plasmid. Synthetic peptides with a C-terminus Gly-Gly-Gly-Cys-COOH linked to M9 or KNS are grafted at 1:1 to 3:1 to the activated amine-rich scaffolds. The crosslinker succinimidyl 4-(N- maleimidomethyl) cyclohexane-1 (SMCC, Pierce) is added at a final concentration of 10 mM (at 10 % DMSO) to 100 μg amine rich scaffold (pH 7.2, 25°C for 2 hours) to react the SMCC NHS-ester to the primary amine of the scaffold. Excess SMCC and DMSO are removed by sephadex G-15 chromatography . The activated scaffold is conjugated with equimolar or 2 to 3-fold excess of the KNS or M9 peptide at 4°C (16 hours) by maleimide reaction with the SH moiety of the C-terminal cysteine. Individual reaction species are isolated by electroelution or FLPC. Precipitation with this method has not been observed.

Example 6 : NTP containing M9 sequence and Scaf old containing a mutated sequence of SV40 T antigen NLS in confluent BAEC

A nuclear targeting peptide (NTP) of M9-GGC (approximately 4.5 kDa) was chemically synthesized and then crosslinked using SMCC to activate the primary amine on a scrambled sequence containing the amino acids of the SV40 T- antigen NLS (Scaffold: 13 amino acids; MW approximately 1.5 kDa) . The Scaffold-NTP conjugate was about 6 kDa. Addition of conjugated Scaffold-NTP provided marked benefit to transfection of growth-arrested confluent bovine aortic endothelium. Without addition of the crosslinking agent SMCC, the scaffold and the NTP remain as individual peptides and addition of these two peptides provided little enhancement of gene transfection of BAEC. The overall expression of the marker protein β-galactosidase was enhanced over 30-fold with the addition of Scaffold-NTP to plasmid prior to mixture with the lipofection vehicle of lipofectamine . The scaffold alone provided little benefit to gene transfer.

Example 7 : Nuclear Import Assay

The nuclear import assay was conducted in accordance with procedures described by Adam and Gerace {Meth . Enzy. 1992 219:97-111). BAEC cells were washed in ice cold import buffer

[20 mM Hepes (pH 7.3), 110 mM KC₂H₃0₂, 5 mM NaC₂H₃0₂, 2 mM

MgC₂H₃0₂, 0.5 mM ethylene glycol-bis ( β-aminoethyl ether)

N,N,N' ,N' , -tetraacetic acid (EGTA) , 2 mM DL-dithiothreitol

(DTT) and 1 μg/ml each of aprotinin, leupeptin and pepstatin) . The cells were permeabilized by incubation of cells with import buffer containing 40 μg/ml digitonin for 5 minutes and then washed with import buffer. A total of 10 μl of rhodamine labeled plasmid (0.1 μg/ml) was mixed with 40 μl complete import buffer [50%(v/v) rabbit reticulosate (Promega Corp.), 20 mM Hepes (pH 7.3), 11 mM KC₂H₃0₂, 5 mM NaC₂H₃0₂, 2 mM MgC₂H₃0₂, 0.5 mM EGTA, 2 mM DTT, 1 μg/ml each of aprotinin, leupeptin and pepstatin, 5 mM ATP, 5 mM creatinine phosphate, 20 U/ml creatinine phosphokinase) . Cells were incubated at 30°C for 30 minutes with a solution containing 10 μl of rhodamine-plasmid (0.1 μg/μl) + 5 μl of M9 ScT conjugate (12 μg/μl), and 35 μl of complete import buffer. In import inhibition experiments, cells were preincubated with wheat germ agglutinin (WGA, 40 μg/ml) in import buffer for 20 minutes prior to adding rhodamine-plasmid + peptide. The coverslips were rinsed with import buffer, mounted on a slide, and then observed by phase contrast and epifluorescence microscopy using an inverted Zeiss Axiophot microscope equipped with a 63X planapo (NA 1.4) lens.

Example 8 : Delivery of NTP or NTP-scaffold with plasmid to cells in culture

An NTP containing the M9 sequence or other nonclassical NLS (from 0.001 μg to 100 μg NTP per μg DNA plasmid) can be complexed to DNA plasmid via: (1) nonspecific interactions;

(2) electrostatic complexation via cationic residues on the

NTP-scaffold; (3) direct chemical crosslinking of the NTP to the plasmid at ratios from 1 NTP per plasmid to > 100 NTPs per plasmid; (4) co-aggregation of NTP monomers or aggregated multimers of NTP with cationic peptides or cationic polymers that condense plasmid; or (5) co-aggregation of NTP with cationic lipids containing both hydrophobic and charged moieties that facilitate the interaction of the NTP with the DNA. These complexes of the NTP and plasmid can then be delivered to cells at doses ranging from 0.01 to 100 μg of plasmid per 10⁶ cells via: (1) direct incubation; (2) scrape loading; (3) calcium phosphate precipitation; (4) electroporation; (5) radiofrequency poration; (6) micro j ection; (7) ultrasound permeabilization of membrane;

(8) detergent loading to transiently permeabilized membranes;

(9) cold-shock loading or heat-shock loading of cells; (10) lipofection with cationic/neutral lipids combinations such as DMRIE (N- [1, (2, 3) -dimyπstyloxy ) propyl ] -N, N-dimethyl-N- (2- hydroxye thy1 ) ammonium bromide (DMRIE) /dioleoyl phosphatidylethanolamme (DOPE), DOSPER ( 1 , 3-dιoleolyoxy-2-

( 6-carboxyspermyl ) -propyl amme/DOPE, lipofectamine formulation (DOSPA/DOPE) , lipofectm formulation (DOTMA/DOPE) , or DOTAP/cholesterol; (11) lipofection with anionic lipids;

(12) lipofection with neutral lipids such as DOPE or cholesterol; (13) delivery with cationic polymers such as polyethyleneimme (PEI), polyhistidme, polylysme, transferrm conjugated polylysme; receptor-targeted polylysme containing RGD sequence or antibodies against cellular receptors; (14) anionic polymers such as heparm;

(15) charged proteins such as mixtures of histone isoforms or purified histone fractions; (16) endosome escape agents such as chloroqume or peptides, hydrophobic sequences of amphipathic peptides, viral subfractions that contain fusogemc activity to promote endosome escape; or (17) combinations of the above approaches.

Example 9: Intravenous Delivery of NTP or NTP-scaffold with plasmid to animals An NTP containing the M9 sequence or other nonclassical

NLS (from 0.1 μg to 100 μg NTP per μg DNA plasmid) can be complexed to DNA plasmid via: (1) nonspecific interactions;

(2) electrostatic complexation via cationic residues on the

NTP-scaffold; (3) direct chemical crossl kmg of the M9 containing NTP to the plasmid at ratios from 1 NTP per plasmid to > 100 NTPs per plasmid; (4) co-aggregation of NTP monomers or aggregated multimers of NTP with cationic peptides or cationic polymers that condense plasmid; or (5) co-aggregation of NTP with cationic lipids containing both hydrophobic and charged moieties that facilitate the interaction of the NTP with the DNA. The formed complex can then be injected intravenously for delivery to the pulmonary endothelium, hepatocytes, hepatoreticuloendothelium, microvascular capillaries, arterioles, and venules . Such formulations for injection can range in dosage from 0.1 to 100 mg of plasmid. The complexes of the NTP and plasmid can be injected in combination with: saline buffer, (2) ultrasound permeabilization of tissue structures; (3) detergent loading to transiently permeabilized membranes; (4) cold-shock loading or heat-shock loading of cells; (5) lipofection with cationic/neutral lipids combinations such as DMRIE (N- [1, (2, 3) -dimyristyloxy)propyl] -N,N-dimethyl-N- ( 2 - hydroxye thy1 ) ammonium bromide (DMRIE) /dioleoyl phosphatidylethanolamme (DOPE), DOSPER (1 , 3-dioleolyoxy-2- ( 6-carboxyspermyl ) -propyl amine/DOPE, lipofectamine formulation (DOSPA/DOPE) , lipofectin formulation (DOTMA/DOPE) , DOTAP/cholesterol, 3β [N- (N' ,N' - (dimethylaminoethane ) carbamoyl cholesterol (DC-cholesterol) /DOPE; Lysyl- phosphatidylethanoloa ine (PE), DOTMA/cholesterol, or polylysine/DOPE; (6) lipofection with anionic lipids; (7) lipofection with neutral lipids such as DOPE or cholesterol;

(8) delivery with cationic polymers such as polyethyleneimine

(PEI), polyhistidine, polylysme, transferrin conjugated polylysme; receptor-targeted polylysme containing RGD sequence; (9) anionic polymers such as heparin; (10) or charged proteins such as mixtures of histone isoforms or purified histone fractions; (11) endosome escape agents such as chloroquine or peptides, hydrophobic sequences of amphipathic peptides, or viral subfractions that contain fusogenic activity to promote endosome escape, or (12) combinations of the above approaches. Example 10 : Intramuscular Delivery of NTP or NTP-scaffold with plasmid to animals

An NTP containing the M9 sequence or other nonclassical NLS (from 0.001 μg to 100 μg NTP per μg DNA plasmid) can be complex to DNA plasmid via: (1) nonspecific interactions; (2) electrostatic complexation via cationic residues on the NTP- scaffold; (3) direct chemical crosslinking of the M9 containing NTP to the plasmid at ratios from 1 NTP per plasmid to > 100 NTPs per plasmid; (4) co-aggregation of NTP monomers or aggregated multimers of NTP with cationic peptides or cationic polymers that condense plasmid; or (5) co-aggregation of NTP with cationic lipids containing both hydrophobic and charged moieties that facilitate the interaction of the NTP with the DNA such as charged anesthetics including bupivacaine-HCl . These complexes can then be injected intramuscularly for delivery to the skeletal muscle cells, and other muscle resident cells such as lymphocytes, macrophages, antigen presenting cells. Such formulations for injection can range in dosage from 0.1 to 100 mg of plasmid. These complexes of the NTP and plasmid can be injected in combination with:

(1) saline buffer, (2) ultrasound permeabilization of tissue structures; (3) detergent loading to transiently permeabilized membranes; (4) cold-shock loading or heat-shock loading of cells; (5) lipofection with cationic/neutral lipids combinations such as DMRIE (N- [1, (2 , 3) -dimyristyloxy) propyl] -

N,N-dimethyl-N- (2-hydroxyethyl) ammon ium bromide

(DMRIE) /dioleoyl phosphatidylethanolamme (DOPE), DOSPER

( 1, 3-dioleolyoxy-2- ( 6-carboxyspermyl ) -propyl amine/DOPE, lipofectamine formulation (DOSPA/DOPE) , lipofectin formulation (DOTMA/DOPE) , DOTAP/cholesterol, 3β[N-(N' ,N'-

(dimethylaminoethane) carbamoyl cholesterol (DC-cholesterol) /DOPE; Lysyl-phosphatidylethanoloamine (PE) ,

DOTMA/cholesterol; or polylysine/DOPE; (6) lipofection with anionic lipids; (7) lipofection with neutral lipids such as DOPE or cholesterol; (8) delivery with cationic polymers such as polyethyleneimine (PEI), polyhistidine, polylysme, transferrin conjugated polylysme or receptor-targeted polylysme containing RGD sequence; (9) anionic polymers such as heparin; (10) charged proteins such as mixtures of histone isoforms or purified histone fractions; (11) endosome escape peptides, hydrophobic sequences of amphipathic peptides, or viral subtractions that contain fusogenic activity to promote endosome escape; or (12) combinations of the above approaches.

Example 11: Intraluminal Delivery of NTP or NTP-scaffold with plasmid to animals

An NTP containing the M9 sequence or other nonclassical NLS (from 0.001 μg to 100 mg NTP per μg DNA plasmid) can be complex to DNA plasmid via: (1) nonspecific interactions; (2) electrostatic complexation via cationic residues on the NTP- scaffold; (3) direct chemical crosslinking of the M9 containing NTP to the plasmid at ratios from 1 NTP per plasmid to > 100 NTPs per plasmid; (4) co-aggregation of NTP monomers or aggregated multimers of NTP with cationic peptides or cationic polymers that condense plasmid; (5) co-aggregation of NTP with cationic lipids containing both hydrophobic and charged moieties that facilitate the interaction of the NTP with the DNA; or (6) monomers for in si tu polymerization to the vascular wall including light activated or free radical initiated polymerization. These complexes can then be delivered intraluminally with a single port catheter, single balloon catheter, double balloon catheter, porous balloon catheter, or ultrasound catheter for delivery to the endothelial cells or underlying smooth muscle cells, and other vessel wall resident cells such as foam cells, neutrophils, lymphocytes, macrophages, and antigen presenting cells. Such formulations for catheter delivery to the vessel can range in dosage from 0.1 to 100 mg of plasmid. These complexes of the NTP and plasmid can be perfused through the catheter to the local site of delivery in combination with: (1) saline buffer, (2) ultrasound permeabilization of vessel structures; (3) detergent loading to transiently permeabilized membranes; (4) cold-shock loading or heat-shock loading of cells; (5) lipofection with cationic/neutral lipids combinations such as DMRIE (N- [1, (2, 3) -dimyristyloxy) propyl] -N, N-dimethyl-N- (2- hydroxye thyl ) ammonium bromide ( DMRIE ) /dioleoyl phosphatidylethanolamme (DOPE) , DOSPER ( 1 , 3-dioleolyoxy-2- ( 6-carboxyspermyl ) -propyl amine/DOPE, lipof ectamine formulation (DOSPA/DOPE) , lipofectin formulation ( DOTMA/DOPE) , DOTAP/cholesterol, 3β [N- (N' , N' - (dimethylaminoethane) carbamoyl cholesterol (DC-cholesterol) /DOPE; Lysyl- phosphatidylethanoloamine (PE) , DOTMA/cholesterol, or polylysine/DOPE; (6) lipofection with anionic lipids; (7) lipofection with neutral lipids such as DOPE or cholesterol; (8) delivery with cationic polymers such as polyethyleneimine

(PEI) , polyhistidine, polylysine, transferrin conjugated polylysine or receptor-targeted polylysine containing RGD sequences or antibodies; (9) anionic polymers such as heparin;

(10) charged proteins such as mixtures of histone isoforms or purified histone fractions; (11) endosome escape peptides, hydrophobic sequences of peptides, or viral subtractions that contain fusogenic activity to promote endosome escape, or (12) combinations of the above approaches .

Claims

What is claimed is :

1. A composition for enhancing delivery of a molecule to the nucleus of a eukaryotic cells comprising a nuclear targeting peptide containing a nonclassical, nuclear localization signal.

2. The composition of claim 1 wherein the nuclear targeting peptide interacts with transportin to mediate nuclear pore targeting and import of molecules into the nucleus of the cells.

3. The composition of claim 2 wherein the nuclear targeting peptide comprises SEQ ID NO: 3.

4. A method of delivering selected molecules to nuclei of eukaryotic cells comprising contacting the eukaryotic cells with the selected molecules and a nuclear targeting peptide containing a nonclassical, nuclear localization signal.

5. The method of claim 4 wherein the nuclear targeting peptide interacts with transportin to mediate nuclear pore targeting and import of the selected molecules into the nucleus of the cells.

6. The method of claim 5 wherein the nuclear targeting peptide comprises SEQ ID NO: 3.

7. A compound comprising:

(a) a cationic peptide scaffold; and

(b) a nuclear targeting peptide containing a nonclassical, nuclear localization sequence, said cationic peptide scaffold being conjugated to said nuclear targeting peptide via a hydrolytic-resistant chemical linkage .

8. The compound of claim 7 wherein the nuclear targeting peptide comprises SEQ ID NO : 1.

9. A composition comprising a peptide scaffold, a nuclear targeting peptide containing a nonclassical nuclear localization sequence and a plasmid containing a selected nucleic acid sequence.

10. The composition of claim 9 wherein the peptide scaffold is conjugated to the nuclear targeting peptide and a complex is formed between the plasmid and the conjugate.

11. A method for expressing a selected nucleic acid sequence in eukaryotic cells comprising contacting cells with a mixture of a selected nucleic acid sequence, a peptide scaffold and a nuclear targeting peptide containing a nonclassical nuclear localization signal.

12. A method for expressing a selected nucleic acid sequence in eukaryotic cells comprising forming a complex between a plasmid containing the selected nucleic acid sequence and a scaffold-nuclear targeting peptide conjugate; and contacting cells with the complex.

13. Eukaryotic cells transfected with a complex comprising a plasmid containing a selected nucleic acid sequence and a scaffold-nuclear targeting peptide conjugate.

14. A method for treating a patient suffering from a condition associated with an absence in the expression of a normal selected nucleic acid sequence comprising administering to the patient the composition of claim 9.