WO2003045415A2 - Self-assembling p53 peptides as gene delivery vehicles - Google Patents

Abstract

Novel delivery vehicles comprising peptide based domains that can self-assemble into multivalent assemblies are described. The multivalent assemblies are preferably tetrameric assemblies such as hp53tet. The peptide domain is preferably fused to a signal molecule such as a polycationic molecule. The vehicles are useful in delivering a wide range of agents to a cell including peptides, vaccines, cytotoxic molecules, plasmids, genes, drugs and diagnostic probes or agents.

Description

TITLE: SELF-ASSEMBLING PEPTIDE DELIVERY VEHICLES FIELD OF THE INVENTION

The invention relates to novel delivery vehicles comprising peptide based domains that can self-assemble into multivalent assemblies. The vehicles are capable of directing the cellular uptake and cellular localization of agents such as macromolecules. BACKGROUND OF THE INVENTION

Existing multitasking molecules that were designed by nature to be useful molecular templates can be re-engineered as therapeutic agents. Protein templates, such as antibodies and recombinant single chain polypeptides (hormones and receptor ligands) (1 , 2), have been used as homing agents to deliver protein toxins or cytotoxic agents to tissues and cells (3, 4). A major challenge associated with the use of structurally complex protein platforms is the fact that they represent a difficult starting point for the optimization of potentially useful protein conjugates. Practical constraints such as their tumour penetration (5), their intracellular routing and their cellular processing (6) have not been addressed in the design of most available constructs. A possible approach to achieve the vectorial delivery (7) of peptide-based constructs to tissues and cells would be to properly assemble and display minimal functional peptide sequences on a molecular scaffold resulting in templates able to carry out a series of defined routing tasks. We have previously designed branched peptides called loligomers, that penetrate eukaryotic cells and self-localize to the nucleus (8, 9). The utility of such constructs have been demonstrated in delivering plasmids (10) and the photodynamic therapy agent, chlorin e6 (11). Synthetic techniques used to assemble loligomers limit the length and type of functional domains that can be integrated into such structures. In addition, the purification and characterization of such branched peptides represent important challenges that will require alternate assembly strategies. It would be desirable to develop flexible scaffolds based on self- assembling sequences rather than covalent linkage which would broaden the number and size of peptide domains available for building functionally diverse delivery vehicles.

SUMMARY OF THE INVENTION

The present inventor has developed a self-assembling intracellular delivery vehicle based on the scaffold of the human tetramerization domain of p53 (/7p53^tet) which contains residues 325 to 355 of human p53. The delivery vehicles exhibit enhanced cellular uptake and are able to efficiently deliver macromolecules to a particular location in a cell.

Accordingly, the present invention provides a vehicle for the delivery of an agent to cell comprising a peptide domain that is capable of forming multivalent assemblies. In one embodiment, the peptide domain comprises the human tetramerization domain of p53, Λp53^tet.

In further embodiments of the invention, the peptide domain comprises an amino acid sequence selected from the group consisting of th e tetramerization domain containing residues 325 to 355 of human p53 (SEQ ID NO:1 , Table 1); the tetramerization domain of p63 (SEQ ID NO:2, Table 1); the tetramerization domain of p73 (SEQ ID NO:3, Table 1); the tetramerization domain of the p53 protein from Xenopus laevis (SEQ ID NO:4, Table 1); the tetramerization domain of the p53 protein from rainbow trout (SEQ ID NO:5, Table 1); p53tet mutant E343K (SEQ ID NO: 7, Table 1); p53tet mutant E346K (SEQ ID NO: 8, Table 1); and p53tet mutant E343K/E346K (SEQ ID NO: 9, Table 1). The invention also includes modifications to or peptide mimetics, analogs, homologs or derivatives of the SEQ ID NOs:1-9 as long as such modifications or peptidemimetics do not affect their ability to act as a scaffold in the delivery vehicles of the invention.

The present invention further includes delivery vehicles comprising two or more peptide domains that are capable of associating into heteromultimers, for example, heterodimeric tetramers. In specific embodiments of the invention, the heterodimeric tetramer comprises SEQ ID NO:9 and SEQ ID NO: 10 as shown in Table 1 , or analogs, homologs, derivatives or mimetics thereof. The peptide domain is preferably fused (at the C- or N-terminus) to one or more signal molecules that can deliver the agent to and into specific compartments of cells. In a preferred embodiment, the cellular import signal molecule is a polycationic molecule, preferably a polycationic amino acid sequence such as deca-lysine or deca-arginine. I n a further embodiment, the delivery vehicle additionally comprises one or more nucleus-directing signal sequences that can guide the uptake of the delivery vehicle to the nucleus of the cell. In specific embodiments of the invention, the delivery vehicle comprises an amino acid sequence selected from the group consisting of of SEQ ID N0:12, SEQ ID N0:13, SEQ ID NO:16, SEQ ID NO:17, SEQ ID NO:20, SEQ ID NO:21 and SEQ ID NO:24, as shown in Figure 1 or a fragment, peptide mimetic, analog, homolog or derivative thereof.

The present invention further includes a method of delivering an agent to a cell comprising: (a) forming a complex between a delivery vehicle according to the invention and the agent; and (b) contacting the cell with the complex of (a) under conditions that allow the complex to be delivered to the cell. In embodiments of the invention, the agent is selected from the group consisting of peptides, vaccines, cytotoxic molecules, nucleic acids, plasmids, genes, drugs and diagnostic probes or agents.

The invention further relates to uses of the delivery vehicles of the present invention to deliver agents to cells.

Other features and advantages of the present invention will become apparent from the following detailed description. It should be understood, however, that the detailed description and the specific examples while indicating preferred embodiments of the invention are given by way of illustration only, since various changes and modifications within the spirit and scope of the invention will become apparent to those skilled in the art from this detailed description.

BRIEF DESCRIPTION OF THE DRAWINGS The invention will now be described in relation to the drawings in which:

Figure 1A is a schematic representation of the linear arrangement of targeting domains present in typical p53^tet constructs prepared for this study. The IS (import sequence) is comprised of either 10 lysines, 10 arginines or the arginine-rich HIV Tat import sequence. The NLS sequence chosen is derived from the SV40 large T antigen.

Figure 1 B is a table outlining the sequence and abbreviation of synthetic peptides cited in the study. The superscript after each peptide name indicates their oligomerization state under physiological conditions (X^tet=tetramer, X^mono=monomer).

Figure 2 shows chromatograms depicting the elution profile and homogeneity of purified p53^tΘt peptides on a C reverse phase analytical column. The absorbance was recorded at 220nm. Figure 3 is a graph showing the internalization of fluorescent p53^tet - based constructs by CHO and Vero cells as measured by flow cytometry. Panels A-D. Dependence of cellular import on peptide concentration. CHO and Vero cells were incubated for 4 hours with increasing concentrations of p53 constructs. Panels A and B. Cellular uptake curves for tetrameric p53 constructs: NLS-10K- p53^tet (•), 10K-p53^tet (■), NLS-10R-p53^tet (A), 10R- p53^tet (Y), NLS-TAT-p53^tet (X ), TAT-p53^tet (♦). Panels C and D. Cellular uptake curves for monomeric p53 constructs: NLS-10K-p53^mono (O), 10K- _p53mon^o _(D)ι NLS-10R-p53^mono (Δ), 10R-p53^mono (V), NLS-TAT-p53^mono (X ),

TAT-p53 ^oπo (0). Panels E-H. Time dependence of cellular import. Fluorescein-labeled p53^tet constructs (0.5 μM; symbols are identical to those cited in panels A to D) were incubated with CHO or Vero cells and the level of cellular uptake measured by flow cytometry at different time intervals following their exposure to cells. Each data point represents the averaged relative mean fluorescence signal from experiments performed in triplicate. Figure 4 shows confocal images showing the distribution of fluorescein- labeled peptides within CHO cells. Images (1μm thickness) of cells were recorded after a 4-hour incubation period with peptides. The co-localization of labeled constructs (fluorescein, green color) with the nuclear stain (7-AAD, red color) could be observed for peptide 10K-p53^tet (Panel A), NLS-10K-p53^tet (Panel B),10R-p53^tet (Panel E), and NLS-10R-p53^tet (Panel F). Fluorescein- labeled monomeric peptides as depicted by NLS-10K-p53^moπo (Panel C) faintly accumulated on the perimeter of cells while the TAT-p53^tet construct (Panel D) was not readily imported into CHO cells. All stained cells were viable and images depict localization events representative of all cells in the field of view.

Figure 5 is a gel shift assay of p53^tet peptides with DNA. Increasing amounts of NLS-10K-p53^tet, NLS-10R-p53^tet, TAT-p53^tet and NLS-10K- p53^mono peptides were added to a tube containing 1μg of pEGFPLuc plasmid. The resulting complexes were resolved on an agarose gel. Shifting of plasmid DNA bands indicate the noncovalent association of p53 constructs with DNA. Figure 6A is a histogram showing luciferase expression levels for CHO cells transfected with DNA complexed to tetrameric p53 constructs. The peptide: DNA weight ratio for each construct was optimized for transfection efficiency and typically ranged in values between 15:1 and 25:1. Panel B is a histogram of luciferase expression levels for CHO cells transfected with DNA complexed to monomeric p53 constructs. Baseline readings for CHO cells transfected with DNA alone ranged in values from 0-10³ RLU/mg protein. Panel C shows images of β-galactosidase expression for CHO cells transfected with the three highest luciferase-producing p53^tet constructs presented in panel A. Figure 7 is a schematic of the structure of the tetramerization domain of human p53. a) Three-dimensional structure of the tetramerization domain of human p53 as determined by X-ray crystallography (44), PDB ID file 1 c26, using Swiss PDB Viewer (62) (http://www.expasy.ch/spdv). One ionic network involving residues Glu343, Glu346, and Lys351 is depicted as a wire frame model, b) Close up view of the ionic network. Relevant distances between atoms are indicated, c) Sequence of the 72-amino acid long p53tet protein constructs used in this study. Vector-associated sequences (histidine tag and thrombin cleavage site) are shown in italics, while the p53tet minimum structural domain (residues 325-355 of human p53) is underlined. Residues Glu343, Glu346, and Lys351 are shown in bold face. The site of cleavage by thrombin is indicated by an arrow. Alignments (performed with ClustalW (63)) with sequences of the tetramerization domains of human p73 (41) and p63 (40) as well as in the tetramerization domains of p53 proteins from Xenopus laevis (42) and rainbow trout (43) are presented to highlight compensatory mutations involving Glu343 and Lys351 which avoid unfavorable charge effects. Figure 8 are graphs showing representative sedimentation equilibrium ultracentrifugation data for p53tet variants, measured at 20 °C at the indicated speeds, a) p53tet-K351 E (33 000 r.p.m.), b) p53tet-E343K/E346K + p53tet- K351E (32 000 r.p.m.). Absorbance values are at 230 nm. The fit for a) is to a monomer-dimer equilibrium; the fit for b) is a free fit to a single oligomeric species. Residuals are multiplied by 10³ for clarity.

Figure 9 are chromatograms showing the molecular size of p53tet variants. Variants of human p53tet were analyzed by size-exclusion chromatography. Injected samples (0.8 mg in 400 μl) were dissolved in 25 mM sodium phosphate, pH 7.0, 100 mM NaCl), and absorbance was recorded at 280 nm. Dotted lines indicate the elution times of the wild-type tetramer p53tet-WT (labeled as ) and of the known dimer p53tet- M340Q/L344R (labeled as 'd'). The elution profiles of the E343K + K351 E and E346K + K351 E mixtures (not shown) are also similar to WT.

Figure 10 are graphs showing the thermal stability of p53tet variants. Temperature melting curves for individual and combinations of p53tet constructs were determined by plotting ellipticity values derived from CD measurements at 222 nm as the fraction of unfolded protein (f_u) versus temperature, assuming a two-state folding model. Experiments were conducted with a protein (monomer) concentration of 10 μM in 25 mM sodium phosphate, pH 7.0, 100 mM NaCl. a) Mutants of p53tet, compared to p53tet- WT: (D) p53tet-WT, (■) p53tet-E343K, (O) p53tet-E346K, (•) p53tet- E343K E346K, (A) p53tet-K351 E. b) Mixtures of p53tet mutants: (□) p53tet- WT, (■) p53tet-E343K + p53tet-K351 E, (O) p53tet-E346K + p53tet-K351 E, (•) p53tet-E343K/E346K + p53tet-K351E.

Figure 11 shows an SDS-PAGE analysis of combinations of His-tagged and non-His-tagged p53tet complexes captured using metal affinity resin (Talon). For each pair of p53tet constructs tested, the first lane represents the mixture of both components prior to the addition of the Talon resin. The second lane is an aliquot of the supernatant after incubation with Talon resin, and the third lane depicts the eluate recovered in the presence of imidazole. a) Lanes 1-3: His6-p53tet-E343K/E346K and His6-p53tet-K351 E, lanes 4-6: non-His-tagged p53tet-E343K/E346K and non-His-tagged p53tet-K351 E, lanes 7-9: His₆-p53tet-E343K/E346K and non-His-tagged p53tet-K351 E, lanes 10-12: non-His-tagged p53tet-E343K/E346K and His₆-p53tet-K351 E. b) Lanes 1 :3: His₆-p53tet-E343K and non-His-tagged p53tet-K351 E, lanes 4-6: His₆-p53tet-E346K and non-His-tagged p53tet-K351 E, lanes 7-9: His₆-p53tet- WT and non-His-tagged p53tet-K351 E, lanes 10-12: His₆-p53tet-WT and non- His-tagged p53tet-E343K/E346K. DETAILED DESCRIPTION OF THE INVENTION (i) PEPTIDE BASED DELIVERY VEHICLES

The inventor has developed self-assembling intracellular delivery vehicles based on the scaffold of the human tetramerization domain of p53 (/7p53^tet, residues, 325-355). This self-associating peptide template displays 8 termini within its tetramer structure that allows for the simultaneous presentation of various cell targeting signals or functional domains. The resulting constructs exhibit enhanced cellular uptake and are able to efficiently deliver macromolecules to a particular location within a cell. The addition of a deca-lysine or deca-arginine sequence (import signal) to the N-terminus of the A7p53^tet peptide promotes cellular import into eukaryotic cells. This event was not observed for monomeric p53^tet peptides, created by a single point mutation within the 7p53^tet domain (L344P). Peptides containing a nuclear localization sequence (NLS) along with the polycationic sequence were found to shuttle efficiently to the nucleus of cells and were capable of delivering reporter plasmids into CHO cells for the purpose of transfection. The development of self-assembling peptide-based vehicles provides a strategy for creating non-viral multitasking delivery vehicles with potential uses in vaccine designs or gene therapy.

Accordingly, the present invention provides a vehicle for the delivery of an agent to cell comprising a peptide domain that is capable of forming multivalent assemblies.

The peptide domain can be any peptide sequence that can spontaneously associate into stable multivalent assemblies such as tetrameric assemblies. Multimeric peptides are used as they have a high binding avidity and enhanced functional activity as compared to monomeric peptides. Examples of peptide domains that can be used in the present invention include, but are not limited to, domains from p53 such as the tetramerization domain, coiled-coil sequences, helix bundles, and the pentameric arrangement of bacterial protein toxins such as cholera toxin, the labile enterotoxin and Shiga and related toxins. The multimeric assembly may comprise any number of peptide monomers which self associate into a multivalent species. For example the assembly may comprise two (dimer), three (trimer), four (tetramer), five (pentamer) etc. peptide monomers. Preferred are tetrameric assemblies. Further, the peptide monomers need not all be the same as the invention extends to heteromultimeric species.

Accordingly, the present invention includes vehicles for the delivery of agents to a cell comprising two or more peptide domains that are capable of associating into heteromultimers. The term "heteromultimer" refers to a multivalent peptide assembly wherein one or more of the peptide domains is different from the others. In particular, the heteromultimer is a heterodimeric tetramer, wherein a dimer of one peptide domain, associates with a dimer of a second peptide domain.

In one embodiment, the peptide domain comprises a tetramerization domain from a protein such as p53, p63 or p73. Preferably the peptide domain comprises the tetramerization domain containing residues 325 to 355 of human p53 (SEQ ID NO:1 , Table 1). The tetramerization domains of p63 (SEQ ID NO:2) (40) and p73 (SEQ ID NO:3) (41) are also shown in Table 1 , along with the tetramerization domains of p53 proteins from Xenopus laevis (SEQ ID NO:4) (42) and rainbow trout (SEQ ID NO:5) (43).

The invention also includes modifications to or peptide mimetics, analogs, homologs or derivatives of the SEQ ID NOs:1-10 as long as such modifications or peptidemimetics do not affect their ability to act as a scaffold in the delivery vehicles of the invention (i.e. does not affect its ability to form multimers). The structure of multivalent peptide assemblies, for example, the p53 tetramer, have been well studied using X-ray crystallographic and NMR techniques. Accordingly a person skilled in the art would be able to devise peptide mimetics, analogs, homologs or deriviatives that would not affect (and may even enhance) the formation of a stable assembly. For example, as shown in Figure 7a, each monomer within the context of the human p53 tetramer (p53tet) domain adopts an identical structure, namely a short N- terminal β-strand (residues 326-333) followed by a turn and a C-terminal - helical domain (residues 335-354). Two monomers associate in an antiparallel fashion through contacts between β-sheet strands as well as hydrophobic interactions involving α-helical residues to form a "primary dimer" (44,45). One significant salt-bridge in the p53tet region occurs between Arg337 of one subunit and Asp352 of its adjacent subunit (44) (side-chain 0- N distance of 2.72 A), stabilizing the structure of the primary dimer (46,47). Two primary dimers then self-associate through an interface derived from residues located in their α-helical domains to form a "dimer of dimers," referred to as a p53 tetramer. Mutations of amino acids at this interface have highlighted the importance of hydrophobic residues leading to the formation of the tetramer as well as stable p53 dimers (48-53). In particular, Davison et al. (22) have noted that the substitution of proline for lysine at position 344 (SEQ ID NO:6) abolishes its activity to form tetramers.

A further examination of the crystal structure (44) of the human p53 tetramerization domain reveals the presence of one arginine (Arg342), one lysine (Lys351), and four glutamates (Glu339, Glu343, Glu346 and Glu349) within the boundaries of the dimer-dimer interface (residues 338 to 351). Of these residues, the only pairs of complementary charged side chains proximal enough to form an intermonomer salt-bridge involve Lys351 with Glu343 and/or Glu346. The side chain oxygen of Glu343 on one monomer was found to be located 2.58 A from the nitrogen side chain of Lys351 on another monomer. NMR structures of p53tet (45, 54) in solution have revealed that Glu346 is apparently closer than Glu343 to Lys351 , although these ionic residues are further apart in these structures than in the crystal structure. Finally, the alignments of p53 tetramerization domain sequences from Xenopus laevis (42) and rainbow trout (43) as well as from human p73 and p63 sequences (40,41) (Table 1) indicate that the naturally-occurring Glu343Lys mutation in their tetramerization domain is always coupled with a corresponding loss of the positively charged lysine residue at position 351. Taken together, these findings suggest that the presence of salt-bridges involving Lys351 and Glu343 and/or Glu346 would contribute four or more ionic interactions within the context of the p53 tetramer interface favoring the stabilization and self-association of primary dimers. The present inventor has prepared, using recombinant DNA technology, of the p53tet domain harboring charge-reversal mutations at positions 343, 346 and 351 (Table 1. SEQ ID Nos:7-10). The p53tet mutants, E343K (SEQ ID NO:7), E346K (SEQ ID NO:8) and E343K/E346K (SEQ ID NO:9) existed as tetramers, although their stability was lower than the wild type p53tet. A heterodimeric tetramer, with enhanced thermal stability, relative to either of the two components in isolation, was formed between the mutants p53tet-E343K/E346K (SEQ ID NO:9) and p53tet-K351 E (SEQ ID NO: 10). Accordingly, the present invention provides a heterodimeric tetramer comprising SEQ ID NO:9 and SEQ ID NO: 10 as shown in Table 1 , or analogs, homologs, derivatives or mimetics thereof.

As stated above, the present invention encompasses any derivatives or peptidomimetics of SEQ ID NOs: 1-10, in particular SEQ ID NO:1 , in which the peptide bond or backbone of the peptide has been replaced with a molecular skeleton so that the functional residues of the peptide are preserved, and conformationally constrained, in approximately the correct positions for interaction of the important sites on the original peptide. This substitution may include but is not restricted to any atoms of the peptide, such as the O and NH atoms of the peptide backbone, or other atoms which are rarely involved in close interactions with the important sites. "Peptidomimetics" may encompass any such substitutions resulting in the preservation of the residue interactions which are paramount in the proper tetramer formation of the peptide. An example of a peptidomimetic is the substitution of a non-peptidal architectural spacer (i.e., a benzodiazepine- based β-turn mimetic) so that the functional side-chain residues are positioned so that their C-α-atoms can occupy equivalent positions to those occupied in the native peptide. This example is not meant to be limiting. Any such peptidomimetics or analogues are encompassed herein.

The term "analog" as used herein refers to a peptide having the amino acid sequence of a reference sequence except for one or more amino acid substitutions, insertions, and/or deletions. Amino acid substitutions may be of a conserved or non-conserved nature. Amino acid residues that are "conservative variants" or "conservative substitutions" for corresponding residues in a reference sequence are those that are physically or functionally similar to the corresponding reference residues, e.g. that have similar size, shape, electric charge, hydrophobicity, hydrophilicity, polarity, reactive chemical properties including the ability to form covalent or hydrogen bonds, and other properties. Particularly preferred conservative variants are those fulfilling the criteria defined for an "accepted point mutation" (Dayhoff et a , in Atlas of Protein Sequence and Structure, Suppl. 3, Natl. Biomed. Res. Foundation, Washington, D.C., Chapter 22, pp. 352-54). Conservative variants of amino acids typically include substitutions within the following groups:

I. glycine, alanine, valine, isoleucine, leucine

II. aspartic acid, glutamic acid, asparagine, glutamine

III. serine, threonine IV. lysine arginine

V. phenylalanine, tyrosine. The term "homolog" as used herein refers to a peptide having an amino acid sequence having at least 70%, preferably 80-90% identity with a reference amino acid sequence.

The mulitmer-forming peptide domain is preferably fused to one or more signal molecules that can deliver the agent to and into specific compartments of cells. The one or more signal molecules may be fused to the either the N- or C-terminus of the peptide. Examples of signal molecules that can be used in the present invention include, but are not limited to, peptide-based import signals including cationic and/or amphipatic peptide import sequences such as polylysine, polyarginine or polyomithine sequences or derived from the HIV Tat protein, or the third helix of the Antennapedia homeodomain peptide, or from transportin, or the Herpes simplex virus VP22 protein, or from the Influenze HA-2 sequence ligands able to direct the delivery of constructs to specific cells; endosomolytic domains such as imidazole containing polymers exemplified by a poly histidine sequence or acting as proton sponges; CTL or other peptide epitopes; peptide routing signals such as NLS, mitochondrial and/or ER retention signals; protease- sensitive cleavage sites associated with the release of drugs, toxins, proteins or epitopes; molecules inducing, enhancing or skewing immune responses such as CPG motifs, lipid chains or mannose groups introduced in the scaffold or the arms of the branched polymers for example; enzymatic, metal chelating or fluorescent domains and domains able to modulate the function of specific signalling pathways; domains facilitating the purification and recovery of constructs; domains able to bind to nucleic acids with the intent of using such peptide vehicles as non-viral antisense or gene delivery vectors.

In a preferred embodiment, the signal molecule is an import signal that can guide the cellular uptake of the delivery vehicle through the plasma membrane of eukaryotic cells. In a specific embodiment, the import signal is a polycationic molecule, more preferably a polycationic amino acid sequence. Examples of polycationic amino acid sequences that can be used include polylysine, polyarginine, and the TAT import sequence GRKKRRQRRRAP. In a specific embodiment, the polycationic amino acid sequence is deca-lysine or deca-arginine.

In a further embodiment, the delivery vehicle additionally comprises one or more nucleus-directing signal sequences that can guide the uptake of the delivery vehicle to the nucleus of the cell. The one or more nucleus- directing signals may be fused to the either the N- or C-terminus of the peptide and either before or after the one or more signal molecules (if present). The nucleus-directing signal is preferably a nuclear localization sequence (NLS). In embodiments of the invention, the NLS is fused to the signal molecule, such as a polycationic sequence. An example of an NLS is the NLS of the SV40 large T antigen which can target non-nuclear proteins to the cell nucleus.

In further embodiments of the invention, the delivery vehicle comprises an amino acid sequence selected from the group consisting of SEQ ID NO: 12, SEQ ID NO:13, SEQ ID NO:16, SEQ ID NO:17, SEQ ID NO:20, SEQ ID NO:21 and SEQ ID NO:24, as shown in Figure 1. The present invention also includes delivery vehicles comprising any fragment, peptide mimetic, analog, homolog or derivative of the above peptides as long as such modifications or mimetics do not affect the ability of the peptides to act as delivery vehicles. In still further embodiments of the invention, the delivery vehicle comprises an amino acid sequence selected from the group consisting of SEQ ID NO: 13, SEQ ID NO: 16 and SEQ ID NO: 17, as shown in Figure 1 , and any fragment, peptide mimetic, analog, homolog or derivative thereof.

The peptides described herein (including truncations, analogs, etc.) may be prepared using well-known synthetic chemical techniques, for example using solid phase synthesis methodologies as described in Example 1 hereinbelow, or using recombinant DNA methods, for example as described in Example 2, hereinbelow. Exemplary methods for the recombinant expression of peptides and proteins are described in Sambrook et al (Molecular Cloning: A Laboratory Manual, 2nd Ed., Cold Spring Harbor Laboratory Press, 1989). The term "delivery" refers to transportation of a molecule to a desired cell or any ceil. Delivery can be to the cell surface, cell membrane, cell endosome, within the cell membrane, nucleus or within the nucleus, or any other desired area of the cell. Delivery includes transporting any macromolecule including, but not limited to, nucleic acids, proteins, lipids, carbohydrates and various other molecules.

The term "macromolecule", refers to any natural and/or synthetic molecule capable of being in a biological environment and includes but is not limited to, proteins, oligonucleotides, dextrans, lipids or carbohydrates that can be delivered using the delivery vehicles systems described herein. The term "nucleic acid" as used herein refers to DNA or RNA. This would include naked DNA, a nucleic acid cassette, naked RNA, or nucleic acid contained in vectors or viruses. These are only examples and are not meant to be limiting. (ii) METHODS OF DELIVERING AGENTS TO CELLS The delivery vehicles of the invention are advantageous as the tetrameric peptide domain acts as a scaffold to which additional domains can be integrated. Specifically, the tetrameric domain has 8 termini to which the domains can be fused. The delivery vehicles of the invention have been demonstrated to be non-toxic and provide a novel strategy for creating non- viral delivery vehicles for the delivery of various agents including peptides, vaccines, cytotoxic molecules, plasmids, genes, drugs, and diagnostic probes or agents. Further, the delivery vehicles themselves may act by themselves as therapeutic or diagnostic agents without the need to carry any "cargos".

Accordingly, the present invention further provides a method of delivering an agent to a cell comprising (a) forming a complex between a delivery vehicle according to the present invention and the agent; and (b) contacting the cell with the complex of (a) under conditions that allow the complex to be delivered to the cell.

The phrase "complex between the delivery vehicle and the agent" means that the delivery vehicle can associate with the agent. The phrase

"associate with" includes, but is not limited to, bind, attach, connect or link molecules through non-covalent (preferred) or covalent means. By virtue of their cationic nature, the p53tet peptides non-covalently associate or form complexes with the negatively charged phosphate backbone of DNA or RNA. This provides a straightforward means to bring the delivery vehicles into association with agents comprising nucleic acids for delivery to cells. Other association methods may be employed. Accordingly the delivery vehicles of the present invention may be used for many applications requiring the delivery of agents to cells, including, for example, gene therapy and induction of a productive immune response against a pathogen.

The term "under conditions that allow the complex to be delivered to the cell" will vary depending on the identity of the agent and the cell type. In general, the conditions involve incubating the cells in the presence of the delivery vehicle-agent complex under conditions that allow the cell to grow and multiply. When the agent is a nucleic acid, the delivery vehicle-agent complex may be "delivered" to the cell using well known cell transformation techniques.

The abilities of the delivery vehicles of the present invention to deliver nucleic acids to specific cells and to the nucleus also allows transgenic animal models to be used for exploring model therapeutic avenues as well as livestock agricultural purposes. Furthermore, the above peptide- macromolecule complexes may be used for administration of agents for the treatment of various diseases. In addition, the above peptide-macromolecule complexes can be used to transform cells to produce particular proteins, polypeptides, and/or RNA. Likewise, the above peptide-macromolecule complexes can be used in vitro with tissue culture cells. In vitro uses allow the role of various nucleic acids to be studied by targeting specific expression into specifically targeted tissue culture cells.

The present invention also encompasses a transgenic animal whose cells contain a nucleic acid delivered via the peptide-macromolecule complex. These cells include germ or somatic cells. Transgenic animal models can be used, for example, for dissection of molecular carcinogenesis and disease, assessing potential chemical and physical carcinogens and tumor promoters, exploring model therapeutic avenues and livestock agricultural purposes. The methods of use also include a method of treating humans, which is another aspect of the present invention. The method of treatment includes the steps of administering the delivery vehicles as described above so as to deliver a desired nucleic acid to a cell or tissue for the purposes of expression of the nucleic acid by the cell or tissue. Cell or tissue types of interest can include, but are not limited to, liver, muscle, lung, endothelium, joints, skin, bone and blood. The present invention further includes a use of the delivery vehicles of the present invention to deliver a desired nucleic acid to a cell or tissue for the purposes of expression of the nucleic acid by the cell or tissue. One embodiment of the present invention includes cells transformed with nucleic acid associated with the delivery vehicles of the present invention described above. Once the cells are transformed, the cells will express the protein, polypeptide or RNA encoded for by the nucleic acid. Cells included, but are not limited to, liver, muscle and skin. This description is not intended to be limiting in any manner.

The nucleic acid which contains the genetic material of interest is positionally and sequentially oriented within the host or vectors such that the nucleic acid can be transcribed into RNA and, when necessary, be translated into proteins or polypeptides in the transformed cells. A variety of proteins and polypeptides can be expressed by the sequence in the nucleic acid cassette in the transformed cells. These products may function as intracellular or extracellular structural elements, ligands, hormones, neurotransmitters, growth regulating factors, apolipoproteins, enzymes, serum proteins, receptors, carriers for small molecular weight compounds, drugs, immunomodulators, oncogenes, tumor suppressors, toxins, tumor antigens, antigens, antisense inhibitors, triple strand forming inhibitors, ribozymes, or as a ligand recognizing specific structural determinants on cellular structures for the purpose of modifying their activity.

Transformation can be done either by in vivo or ex vivo techniques. One skilled in the art will be familiar with such techniques for transformation. Transformation by ex vivo techniques includes co-transfecting the cells with DNA containing a selectable marker. This selectable marker is used to select those cells which have become transformed. Selectable markers are well known to those who are skilled in the art.

Administration as used herein refers to the route of introduction of the delivery vehicle into the body. Administration includes intravenous, intramuscular, systemic, subcutaneous, subdermal, topical, or oral methods of delivery. Administration can be directly to a target tissue or through systemic delivery.

In particular, the present invention can be used for administering nucleic acid for expression of specific nucleic acid sequence in cells. Routes of administration include intramuscular, aerosol, olfactory, oral, topical, systemic, ocular, intraperitoneal and/or intratracheal. A preferred method of administering nucleic acid carriers is by intravenous delivery. Another preferred method of administration is by direct injection into the cells.

In addition, another means to administer the nucleic acid carriers of the present invention is by using a dry powder form for inhalation. One compound which can be used is polyvinylpyrrolidone ("PVP"), an amorphous powder. PVP is a polyamide that forms complexes with a wide variety of substances and is chemically and physiologically inert. Specific examples of suitable PVP's are Plasdone-C™ 15, MW 10,000 and Plasdone-C™ 30, MW 50,000. Furthermore, administration may also be through an aerosol composition or liquid form into a nebulizer mist and thereby inhaled.

The special delivery route of any selected vector construct will depend on the particular use for the nucleic acid associated with the nucleic acid carrier. In general, a specific delivery program for each nucleic acid carrier used will focus on uptake with regard to the particular targeted tissue, followed by demonstration of efficacy. Uptake studies will include uptake assays to evaluate cellular uptake of the nucleic acid and expression of the specific nucleic acid of choice. Such assays will also determine the localization of the target nucleic acid after uptake, and establishing the requirements for maintenance of steady-state concentrations of expressed protein. Efficacy and cytotoxicity is then tested. Toxicity will not only include cell viability but also cell function. The chosen method of delivery should result in cytoplasmic accumulation and optimal dosing. The dosage will depend upon the disease and the route of administration but should be between 0.1-1000 mg/kg of body weight/day. This level is readily determinable by standard methods. It could be more or less depending on the optimal dosing. The duration of treatment will extend through the course of the disease symptoms, possibly continuously. The number of doses will depend upon disease delivery vehicle and efficacy data from clinical trials.

Establishment of therapeutic levels of DNA within the cell is dependent upon the rate of uptake and degradation. Decreasing the degree of degradation will prolong the intracellular half-life of the DNA.

The following non-limiting examples are illustrative of the present invention: (iii) EXAMPLES Example 1

In this Example, the use of the tetramerization domain of human p53 ( ?p53^tet) as a self-associating peptide template to design novel multitasking delivery agents is described. MATERIALS AND METHODS Cell lines

CHO and Vero cell lines were obtained from ATCC (Rockville, MD) and were maintained in α-MEM media supplemented with 10% Fetal Calf Serum and antibiotics. Cells were grown at 37°C in the presence of 5% C0₂. Synthesis and fluorescein labeling of p53^tet peptides Peptides were synthesized on an Applied Biosystems 431A synthesizer using 9-fluorenylmethoxycarbonyl (Fmoc) chemistry and RINK- amide resin support (Applied Biosystems, Mississauga ON). After the final deprotection following synthesis completion, a portion of each peptide was labeled with fluorescein (6-carboxyfIuorescein (6-FAM), Molecular Probes, Eugene, OR) using standard active ester (HBTU) chemistry. Peptides were cleaved from their support by resuspending resins in TFA: phenol: water: thioanisole: ethanediol (82.5%: 5%:5%:5%:2.5%). The crude lyophilized peptides were dissolved in water containing 0.1% (v/v) TFA and purified on a C₄ semi-preparative reverse phase column (Waters, Milford MA) equilibrated in the same mobile phase. The peptides were eluted with an acetonitrile gradient going from 5 to 80% AcN in water in 30 minutes with a flow rate of 7ml/min. The mass of all peptides was confirmed by mass spectrometry. Purified peptides were further analyzed on a C analytical reverse phase column (Waters, Milford MA). The peptides were dissolved in water containing 0.1%TFA (v/v) and eluted at a flow rate of 1 ml/min using a 5-80% AcN gradient over a 20-minute period. Peptides peaks were detected at 220nm (Figure 2).

Internalization of p53^tet peptides into eukaryotic cells

CHO and Vero cells (1x10⁶ cells/ml) were incubated with either increasing concentrations of fluorescein-labeled peptides for 4 hours (concentration dependence) or exposed to 0.5μM peptide for various time intervals (time dependence). The relative fluorescence was then measured by flow cytometry (FACScan, Becton Dickinson). Viable cells were gated using the cell impermeant DNA intercalating dye, 7-AAD (0.1 mg/ml, Molecular Probes). Data shown represents the relative mean fluorescence intensity based upon three or more separate experiments. Confocal Microscopy

CHO cells (1x10^δ cells/ml, 0.5ml/sample) were maintained in suspension and incubated in the dark with 1μM fluorescein-labeled peptide for 4 hours at 37°C. Cells were then recovered by centrifugation, washed once with PBS with 1 % BSA, and then resuspended in 100μl PBS containing 5μM Syto17 (Molecular Probes). After a 10-minute incubation at room temperature, cells were centrifuged and the pellet washed twice with 1% BSA in PBS. The cells were resuspended in a final volume of 100μl of PBS, dispensed on glass slides and treated with Prolong antifade (Molecular Probes) to preserve the fluorescence signal. Covered slides were then left overnight to dry. Cells were visualized using a Zeiss LSM 510 Confocal Microscope (Ar-Kr lasers) and images (1μm thickness) were recorded. Gel Retardation Assay

One μg of plasmid DNA (pEGFPLuc, Clontech) was mixed with increasing amounts of peptide in PBS for 10 minutes at room temperature. Samples were loaded onto a 0.8% agarose gel (with EtBr) and the complexes were electropherically resolved at 100V. DNA bands were visualized with a UV light and photographed. Transfection Experiments

The noncovalent interaction between plasmid and peptides was achieved by mixing 4μg of pEGFPLuciferase plasmid (Clontech) with an optimized amount of peptide (60μg-100μg) in a final volume of 50μl with PBS. The complexation step was allowed to proceed at room temperature for 20 minutes. Aliquots of CHO cells (1x10^δ cells/ml) were then added to the reaction mixture (450μl) and samples placed in the incubator at 37°C. After 4 hours, the cells were transferred into 6-well plates (Nunc) containing 4.5ml of medium per well. Transfected cells were grown for 48hrs at 37°C prior to assaying luciferase activity (Promega kit, Promega, Madison, Wl). Briefly, cells were lysed with 100μl of a cell lysis buffer (25mM Tris-phosphate pH 7.8, 2mM DTT, 2mM 1 ,2-diaminocyclohexane-N,N,N',N'-tetraacetic acid, 10% glycerol, 1 % Triton X-100), vortexed and then freeze-thawed. Samples were resuspended once more and cell debris removed by centrifugation (14000rpm for 2 minutes). The luciferase activity present in cell lysates (20μl) was determined using luciferin (Promega Assay reagent, Promega, Madison, Wl) and light units were read on a luminometer. Relative light units (RLU) were standardized to RLU/mg protein. The protein concentration for each cell lysate was determined using a Bradford Protein Assay (BioRad). Cells transfected with pGeneGrip (Gene Therapy Systems, San Diego, CA) were stained for β-galactosidase using a commercial kit (β-gal staining kit, Roche). RESULTS AND DISCUSSION

The alteration of a cellular gene or the induction of a productive immune response against a pathogen represent only two of many applications in need of vehicles able to selectively deposit and route macromolecules into cells. The growing list of known peptide routing signals provides a broad spectrum of functional elements or building blocks that can be used to design non-viral delivery vehicles (7). Assembly designs and scaffolds are also required to effectively display such peptide-based elements. The focus of this study is to provide the proof-of concept that efficient peptide-based multitasking vehicles can be created from incorporating or fusing minimal functional domains to the scaffold of a self-assembling peptide domain. Selection of a self-assembling protein domain

Structural templates such as synthetic branched peptides that harbor multiple copies of functional domains typically result in high binding avidity and enhanced functional activity in relation to their monomeric counterparts (7, 8, 12-14). Methods involving solid-phase synthesis have limitations associated with the integration of a large number of peptide or protein domains into branched peptide scaffolds. Linear peptides containing self- associating domains represent a simpler strategy of introducing multivalent signals into peptide scaffolds that is amenable to both synthetic and recombinant approaches. For instance, short peptide sequences coding for coiled coil dimers have been used as peptide elements to allow the pairing of distinct functional domains (15-18). Heterodimeric coiled-coils afford a second dimension in terms of complexity in relation to their homodimer counterparts. To create a platform that is more reminiscent of branched peptides (4 to 8 arms) or dendrimers, one needs to start with a self- assembling domain able to form at least tetramers in solution. The 30-amino acid long tetramerization domain of human p53 (/7p53^tet) represents a well- characterized example of a peptide segment that spontaneously assembles into tetramers. This domain is best described as a symmetric dimer of dimers (19). Its primary amino acid sequence stretching from residues 325-355, folds into a secondary structure that consists of a β-strand (326-334), followed by a turn (335-336) and then an α-helix (337-354). The β-strands of two monomers then interact to form an antiparallel dimer. The formation of the tetramer is stabilized by hydrophobic interactions at the interface of the dimer of dimers (20, 21). However, the substitution of proline for leucine at position 344 abolishes tetramer formation (22). Biophysical studies have confirmed that the minimal region of 325-355 spontaneously forms stable tetramers (23- 25). The short length of this sequence makes it suitable for both synthetic and recombinant methods of assembly and provides a unique platform for building multivalent structures. A number of peptide constructs containing the Λp53^tΘt domain (325-355) were thus assembled by solid-phase peptide synthesis to demonstrate the concept of designing multitasking peptide-based intracellular delivery vehicles (Figure 1). Design of cell routing vehicles incorporating the hp53^tet domain

Several cationic import sequences (IS) have been shown to traverse cell membranes (7). In particular, poly-lysine, poly-arginine and the HIV TAT peptide have been frequently used as peptide import signals (IS) in delivering molecular cargos into cells (26-31). A direct comparison of these three ISs was undertaken to determine the relative efficiency of each sequence in causing the penetration of the 7p53^tel-containing constructs. Import sequences were introduced at the N-terminus of the Λp53^tet domain (Figure 1). These ISs were defined as either a string of 10 lysine residues (10K), 10 arginine residues (10R) or the TAT import sequence GRKKRRQRRRAP (31). The tetramerization of these p53^tet constructs would thus create assemblies harboring up to 40 cationic residues. In addition, some constructs incorporated the nuclear localization signal (NLS) derived from the SV40 large T antigen (32, 33). The HIV TAT sequence is thought to code for both cell penetration and nuclear uptake (31) and thus was introduced onto the p53^tet scaffold without the insertion of an NLS sequence. Several control peptides were also synthesized and are listed in Figure 1 B. For example, the NLS- p53^tet peptide is tetrameric but lacks an import sequence and hence should not be imported into cells. In addition, the monomer versions of all p53^tet peptides were generated by the single point mutation of Leucine 344 to Proline as previously described (22) in order to demonstrate the functional importance of a tetrameric structure in promoting the vectorial delivery of peptide constructs into cells. The mass and purity (Figure 2) of each construct were confirmed prior to studying the effect of tetramerization and peptide routing signals on the ability of each peptide vehicle to perform their projected routing tasks. Peptides with a functional /?p53^tet domain spontaneously assembled into tetramers in a 25mM phosphate buffer (pH 7.0) with 200mM NaCl as determined by ultracentrifugation, gel permeation and cross-linking studies with glutaraldehyde (Table 1). In comparison, all peptides harboring a Leu344Pro substitution were unable to form tetramers in solution. Internalization and localization ofp53^tet constructs into cells

Flow cytometry was used to monitor the ability of eukaryotic cells to internalize fluorescein-labeled constructs. Tetrameric peptides with either a deca-lysine (10K-p53^tet) or deca-arginine (10R-p53^tet) IS were readily imported by Chinese Hamster Ovary (CHO) and Vero cells, at concentrations as low as 0.1-0.5μM (Figure 3A-D). The addition of an NLS sequence onto the N-terminal end of the p53^tet peptides (NLS-10K-p53^tet and NLS-10R-p53^tet) slightly decreased the total amount of labeled construct imported by cells. Fluorescence signals from monomeric peptides were not detected in cells until the peptide concentration reached a minimum of 2.5 μM. Surprisingly, the uptake of the TAT-p53^tet peptide by CHO cells was minimal in relation to the lysine or arginine containing peptides and was only marginally comparable to other IS signals in the case of Vero cells (Figure 3B,D). The NLS-p53^tet control (lacking an import sequence) was not readily internalized and thus illustrated the need to integrate IS domains into constructs.

The time dependence of 7p53^tet peptides import into cells was recorded by flow cytometry. The time study (Figures 3E to 3H), revealed that the NLS- containing peptides were taken up more rapidly than the IS-p53^tet constructs. However, the relative fluorescence signal observed for IS-p53^tet peptides typically exceeded that of the NLS-containing peptides after 2 hours. The peptide signal within cells reached a plateau by 4 hours and hence this time point was used in later experiments. In contrast, fluorescein-labeled monomeric peptides could not be detected within cells even after 6 hours suggesting that the tetrameric state of the p53^tet peptides was required for import. Thus, the spacial clustering of four IS domains appears necessary for cell internalization. The cellular location of the fluorescein-labeled p53 constructs was confirmed by confocal microscopy (Figure 4). NLS-10K-p53^tet and 10R-p53^tet constructs were routed to the nucleus after 4 hours, as defined by the co- localization of the green signal (fluorescein-labeled peptide) and red nuclear stain (Figures 4B and E). The nuclear localization of the 10R-p53^tet peptide suggests that the 10R sequence may act as a NLS (Figure 4E). This evidence correlated well with its ability to transfect cells with a reporter plasmid (Figure 6). This apparent nuclear localization has been previously reported for arginine-rich and poly-arginine sequences (34). The 10K-p53^tet peptide was found to localize mostly within vesicles or endosomes throughout the cell as expected since poly-lysine does not function as an NLS (35). The nuclear stain did appear to co-localize within these vesicles (Figure 4A). The treatment of cells with NLS-10K-p53^mono (Figure 4C) resulted in its accumulation at the perimeter of CHO cells, a weak staining pattern representative of all monomeric peptides tested. Typically, the fluorescence signal observed for monomeric peptides was minimal in comparison with the signal recorded for oligomeric constructs and was absent within the nucleus of cells. In agreement with the flow cytometry data (Figure 3), the labeled TAT- p53^tet peptide (Figure 4D) was not readily visible and was poorly internalized in comparison to poly-Lys and poly-Arg containing p53^tet peptides.

Comparable levels of cellular uptake into CHO cells had been observed by flow cytometry for both deca-lysine and deca-arginine-containing peptides (Figure 3). These constructs however differed in their cellular localization. The 10R-p53^tet construct lacking an NLS sequence was able to self-localize into the nucleus of cells (Figure 4E) and was capable of acting as a transfection vehicle (Figure 6). The addition of an NLS sequence to the arginine-tagged p53^tet construct (NLS-10R-p53^tet) also targeted the peptide to the nucleus (I hour time point - data not shown). Surprisingly, this construct re-distributed itself outside the nucleus within 4 hours (Figure 4F). Thus, an inherent difference may exist in the routing of poly-arginine and poly-lysine within a cell. The multivalent presentation of peptide-based routing domains on a Λp53^tet self-assembling scaffold greatly enhances their cellular uptake by CHO and Vero cells (Figures 3 and 4). Monomeric p53 constructs containing a single import sequence (monomeric peptides) were typically unable to effectively cross the cell membrane. These results parallel similar conclusions observed for branched peptides such as loligomers (8). Confocal images also illustrated the penetration and localization of tetrameric peptides within a cell (Figure 4). Monomeric peptides stained only the perimeter of cells with only a minimal amount of fluorescence signal being observed within the cytosol of viable cells. The TAT-p53^tet construct was poorly internalized by cells when compared to the deca-lysine or deca-arginine-tagged peptides. The TAT peptide from HIV had been originally shown by other groups to translocate across membranes as well as localize to the nucleus of Hela cells (30, 31). This particular sequence is very rich in arginine residues and may simply behave as a poly-arginine sequence (24). In the present study, the monomeric as well as tetrameric TAT-p53 constructs were either weakly or minimally imported into CHO cells with marginally better uptake being observed in Vero cells. These results suggest that the ability of the HIV TAT sequence to function as an import signal is only minimally affected by its presentation on a p53 tetramerization scaffold and the nature of the cell line being targeted. Non-covalent association ofp53^tet constructs with plasmid DNA

The impact of oligomerization and sequence composition (peptide- based routing signals) on the cell trafficking properties of p53^tet constructs was also monitored indirectly using a functional assay. Briefly, each peptide construct was evaluated in terms of its ability to condense and deliver plasmid DNA (pEGFPLuciferase, pGeneGrip) encoding reporter genes into cells (Figures 5 and 6). By virtue of their cationic character, p53^tet peptides were expected to non-covalently associate with the negatively charged phosphate backbone of plasmid DNA (pEGFPLuciferase). Increasing amounts of each construct were thus added to 1μg of plasmid DNA and then resolved on an agarose gel to determine the ratio of peptide to DNA needed to fully condense the plasmid DNA (Figure 5; shifting of DNA band to the top of the gel). For all lysine-containing peptides, a weight ratio of 2.5:1 peptide to DNA was necessary to complex the DNA (Figure 5, NLS-10K-p53^tet shown). For polyarginine peptides, slightly less peptide was needed (NLS-10R-p53^tet shown). Monomeric peptides as well as NLS-p53^tet and TAT-p53^tetwere also capable of shifting the DNA band at the same weight ratio as observed for the tetrameric constructs. hp53^tet-based vehicles are able to deliver plasmids into cells

Peptide-based vehicles based on the p53^tet scaffold are being designed to facilitate the delivery of small and large molecular entities into cells. Plasmid DNA represents a large molecular cargo that provides a simple approach to compare the transfection efficiencies of p53^tet constructs. As expected, imported complexes of either NLS-10K-p53^tel or NLS-10R-p53^tet associated with the plasmid pEGFPLuciferase, resulted in luciferase readings greater than 10⁷ RLU per mg of protein and produced blue-colored CHO cells when complexed with the pGeneGrip plasmid encoding for the expression of β-galactosidase (Figure 6, panels A and B). Interestingly, peptide 10R-p53tet was comparable to the NLS-containing peptides as a delivery vehicle in these transfection experiments (Figure 6). In contrast, the use of 10K-p53^tet as a transfection agent yielded low levels of expression of both luciferase (RLU levels comparable to negative controls) and β-galactosidase (no blue colonies observed) in CHO cells. CHO cells or CHO cells transfected with DNA alone produced luciferase reading ranging from 0-10³ RLU. The dramatic differences in the transfection potential (Figure 6) and localization profile (Figure 4) of 10R-p53^tet and 10K-p53^tet peptides suggest that a fundamental difference exist between poly-lysine and poly-arginine as import sequences.

NLS-Tat-p53^tet and Tat-p53^tet tetrameric constructs as well as all monomeric forms of the p53^tet peptides proved to be poor transfection vehicles in accordance with their lack of internalization by cells (Figures 3 and 4). An analysis of luciferase signal ratios for each pair of p53 tetramer constructs (Table 2) highlights the fact that the deca-arginyl- and deca-lysyl- ISs were typically 20 to 300 times more effective than the Tat sequence in yielding constructs able to deliver plasmids to cells, a finding that correlates with our FACS data (Figure 3). The importance of oligomerization on the transfection potential of each p53 construct was also confirmed by calculating the ratio of luciferase signal values between tetramer and monomer forms of each construct (ratio values in parentheses) with the following order of impact due to tetramerization: NLS-10R-p53 (1400) » NLS-10K-p53 (53) >10R-p53 (8) ~ 10K-P53 (3) > Tat-p53 (0.6) ~ NLS-Tat-p53 (0.3). Surprisingly, the multivalent display of the Tat import sequence on a p53^tet scaffold only marginally enhanced its cell transduction function (Figure 3) with little or no impact on its potential as a transfection agent. In summary, only the tetrameric forms of the p53^tet peptides tested (NLS-10K-p53^tet, 10R-p53^tet and NLS-10R-p53^tet) were able to effectively route these peptide vehicles to the nucleus of cells and to efficiently transfect cells. Considerations for future designs One can envision the introduction of additional tasking domains into vehicles based on the self-assembling peptide scaffold of human p53. Endosomal release represents one important feature that should be encoded onto future /7p53^tet-based vehicles. Poly-L-histidine or imidazole-containing polymers have been found to aid in delivering DNA to the cytosol by destabilizing lipid bilayers in acidic conditions (pKa~6.0) and facilitate endosomal leakage (36, 37). Thus, the insertion of such a domain into the /7p53^tet structure may enhance plasmid DNA escape from endosomes and raise the transfection efficiency of such vehicles. Limitations associated with peptide synthesis restrict the length and type of domains that can be introduced into peptides. These challenges can now be overcome by using recombinant methods. Several of the p53^tet constructs presented in this study have now been expressed in bacteria (Kawamura and Gariepy, unpublished results). Thus, an assortment of other functional domains can be added to the peptide such that other aspects of delivery may be addressed. Another important issue in the design of delivery vehicles is the cytotoxicity of the vehicle itself. The cytotoxicity of large cationic polymers such as poly-lysine has limited their use as drugs (26, 28, 29). Loligomers utilize short multiple penta-lysine domains that act as import signals and enable cell penetration. They display a modest level of cytotoxicity (8). Other DNA delivery agents such as polyethylenimine (PEI) function as 'proton sponges', able to induce the osmolytic rupture of endosomal vesicles causing the leakage of their content into the cytosol (38, 39). This ability to escape endosomes however, leads to cell toxicity (36). The cytotoxicity of p53^tet constructs towards CHO and Vero cells was found to be minimal (showing no toxicity at our maximum dose of 25μM; WST-1 assay, data not shown) and these peptides may therefore provide a safer scaffold for the development of non-viral delivery vectors.

In summary, the design of a self-assembling peptide that is capable of cellular uptake and nuclear localization was reported. The assembly scaffold was based on the 30-amino acid long sequence from the tetramerization domain of human p53 (residues 325-355). Short cationic import sequences were then added to the N-terminus of the p53^tet domain. The NLS sequence from the SV40 large T antigen was also attached to complete the design of peptides capable of cellular import and nuclear routing. This unique strategy for designing multivalent peptides can be achieved through both synthetic (Figure 2) and recombinant methods. The Λp53^tet domain (325-355) represents a self-assembling scaffold able to effectively present multiple copies of peptide-based import and intracellular routing signals that greatly enhances cellular import and nuclear localization. The resulting constructs serve as non-viral delivery vehicles and provide a strategy to design novel protein-like multitasking molecules. Example 2:

In the example, variants of the p53tet domain harbouring charge- reversal mutations at positions 343, 346 and 351 are engineered and their stability an oligomerization states are analyzed in a series of biophysical experiments to resolve the role of salt bridges at the dimer-dimer interface. MATERIALS AND METHODS Mutagenesis The plasmids pET-15b-p53(310-360) and pET-15b-p53(310-360)- M340Q/L344R (53) were gifts from the laboratory of Dr. Cheryl Arrowsmith, Ontario Cancer Institute. The plasmids contain a synthetic gene coding for residues 310-360 of human p53 inserted into the Ndel and BamHI restriction sites of the bacterial expression vector pET-15b (Novagen, Madison, Wl, USA). The p53tet(310-360) sequence is preceded by a vector-encoded Hisδ metal-ion affinity purification tag and a thrombin cleavage site (Figure 7). The E343K, E346K, and K351 E single mutants and the E343K/E346K double mutant of p53(310-360) were assembled by PCR mutagenesis using a two step, three primer method (55), using ProofStart DNA polymerase (Qiagen, Mississauga, ON, Canada). PCR products were purified from reaction mixtures or agarose gels by Qiaquick PCR purification or Qiaquick gel extraction kits (Qiagen). The final PCR products were cloned into a pET-15b vector. Mutations in the gene were confirmed by DNA sequencing. Plasmid constructs were transformed into competent BL21 (DE3) pLysS (Novagen) cells according to standard methods (56). Protein expression and purification. Wild-type and mutant His6-p53(310-360) (i.e. His6-p53tet) proteins were expressed and purified by the same methods. Briefly, stocks of BL21 (DE3) pLysS cells carrying the appropriate plasmid were plated on LB-agar plates supplemented with 100 μg/ml carbenicillin and 34 μg/ml chloramphenicol. A single colony was subsequently used to inoculate 40 ml TB broth supplemented with the same antibiotics. The cultures were grown overnight with shaking at 37 °C. A 15 ml aliquot of each culture was then used to inoculate 1.5 L of preheated (37 °C) TB broth containing carbenicillin and chloramphenicol. The resulting cultures were then grown at 37 °C with shaking until ODβoonm reached 0.6~0.9 (2-3 h), at which point 0.5 mM isopropyl-β-D-thiogalactopyranoside was added to the media to induce protein expression. Cells were harvested by centrifugation after a 3 h induction period.

Cell pellets (~7.5 g wet weight) were subjected to three freeze/thaw cycles, and were resuspended in ~3.7 vol. buffer A (50 mM Tris-HCI, pH 8.0, 500 mM NaCl, 0.1 % Triton X-100) with 20 mM imidazole, 1 mM phenylmethylsulfonyl fluoride, 10 mM MgCI₂, and 2.5 units/ml benzonase nuclease (Novagen). This suspension was placed on ice, sonicated three successive times for 45 s and the resulting sonicate was centrifuged at 15 000 g for 30 min. The supernatant was loaded onto a 2.5 ml column of Talon metal affinity resin (Clontech, Palo Alto, CA, USA) equilibrated with buffer A with 25 mM imidazole and 1 mM PMSF, and the resin was washed with 25-50 ml of the same buffer. Pure protein was eluted with 20 ml buffer A containing 200 mM imidazole. The eluate was dialyzed extensively against 20 mM NH₄HC0₃ and the protein was lyophilized and stored at -20 °C until use. Purity was determined by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) with Coomassie staining (28), and matrix- assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI- TOF-MS) (MMRC Mass Spectrometry Lab, University of Toronto). Circular dichroism spectroscopy

Circular dichroism (CD) spectra were recordedon an Aviv 62A DS circular dichroism spectrometer using a 0.5 cm path-length rectangular cuvette with a 2 ml sample volume. Protein samples (10 μM) were prepared in sample buffer (25 mM sodium phosphate, pH 7.0, 100 mM NaCl). Wavelength scans were recorded with a 1 nm spectral bandwidth, 1 nm between points, and an averaging time of 8 s. Ellipticity measurements at 222 nm were collected a function of temperature for each p53tet construct (or mixtures) using a 1 nm bandwith and a 50s averaging time. Measurements were recorded from 20 °C to 98 °C at 3 °C intervals, with a 1 min temperature pre-equilibration. Ellipticity values were plotted as the fraction of unfoldedprotein (f_u) versus temperature assuming a two-state folding model. Ultracentrifugation

Sedimentation equilibrium ultracentrifugation experiments were performed on a Beckman Optima XL-I analytical ultracentrifuge using an AN50-Ti rotor with six-channel charcoal-Epcon cells. Protein concentrations were 0.125, 0.25, and 0.5 mg/ml prepared in sample buffer (25 mM sodium phosphate, pH 7.0, 100 mM NaCl), as measured by UV spectroscopy using the molar extinction coefficient for a free tyrosine residue (ε₂76nm = 1450). Samples were centrifuged at 20 °C at three different speeds for 24 hours before equilibrium absorbance measurements were taken at 230 nm. Association constants and molecular masses were estimated using Beckman XL-I data analysis software in which absorbance versus radial position data were fitted to the sedimentation equilibrium equation using non-linear least- squares techniques (57).

Size-exclusion chromatography

Analytical size-exclusion chromatography (SEC) experiments were performed on a Superdex-75HR (Pharmacia) column (10 mm X 30 cm) operating at a flow rate of 1 ml/min. Samples (0.8 mg in 400 μl) were injected onto the column and absorbance was monitored at 280 nm. The column was calibrated with gel filtration standards from Bio-Rad (Mississauga, ON, Canada). Thrombin cleavage of mutant p53tet proteins

His₆-p53tet-E343K/E346K and His₆-p53tet-K351 E were cleaved with thrombin using a thrombin cleavage capture kit (Novagen). Two milligrams of each protein were dissolved in 5 ml of a 1X thrombin cleavage buffer (20 mM Tris-HCI pH 8.4, 150 mM NaCl, 25 mM CaCI₂). Biotinylated thrombin (0.5 units, 0.25 units/mg) was then added to the reaction mixture and the reaction was left to proceed at room temperature for 16 h. The biotinylated thrombin was subsequently removed with streptavidin-agarose and the cleaved Hisetag was eliminated with Talon metal affinity resin. The filtrate containing pure cleaved p53tet-E343K/E346K or p53tet-K351E, without His₆tag, was dialyzed against 20 mM NH HC0₃, lyophilized, and stored at -20 °C until use. Samples were analyzed for cleavage and purity by SDS-PAGE. Complete cleavage was achieved and cleavage at other sites in the protein did not occur as no other low molecular weight bands were detected. Metal affinity experiments

Lyophilized His₆-p53tet-WT, His₆-p53tet-E343K, His₆-p53tet-E346K, His₆- p53tet-E343K E346K, His₆-p53tet-K351 E, p53tet-E343K/E346K, and p53tet- K351 E were dissolved in sample buffer (25 mM sodium phosphate, pH 7.0, 100 mM NaCl) to final concentrations of 0.2 mM (1.6 mg/ml). Combinations of p53tet constructs (at molar equivalence of each construct) were mixed and incubated at room temperature for 1 h in microcentrifuge tubes. Fifty μL of Talon metal affinity resin was added to 300 μL of each mixture, and the resulting samples were gently mixed at room temperature for 10 min. The resin was pelleted by centrifugation (10 000 X g for 30 s), and subsequently washed 5X with 200 μL of sample buffer. Bound proteins were eluted in the presence of 100 μL sample buffer containing 0.5 M imidazole. Samples corresponding to the original mixture of proteins prior to treatment with Talon resin, as well as aliquots of the supernatant after incubation with Talon resin, and of the eluate from the imidazole wash were analyzed by SDS-PAGE. RESULTS

Design and structure of human p53tet constructs

A wild type human p53tet(310-360) construct with a N-terminal His₆-tag and a thrombin cleavage site as well as four corresponding p53tet variants harboring either the mutation E343K, E346K, E343K/E346K, and K351 E (Figure 7c) were expressed in bacteria. The five 72-amino acid long constructs were purified to homogeneity by metal-affinity chromatography (Cobalt-based, Talon resin; Clontech) and their mass confirmed by MALDI- TOF MS and SDS-PAGE. The circular dichroism spectra of the p53tet- E343K/E346K and p53tet-K351 E analogs were similar to that of our p53tet-wt construct (data not shown), suggesting a comparable secondary structure. Oligomeric state ofp53tet mutants

Wild-type and mutant p53tet constructs were subjected to analytical ultracentrifugation (20 °C) to assess their oligomeric state. Representative results are shown in Figure 8. Data for all variants were fitted to a single species, the apparent masses are listed in Table 4 and indicate that p53tet- WT (apparent mass 31.8 kDa) is a tetramer in solution. The analog p53tet- E343K/E346K is also predominantly a tetramer. However, its apparent oligomeric state (3.5) is lower than that of p53tet-WT (3.9), suggesting that the p53tet-E343K/E346K tetramer is less stable than its p53tet-WT counterpart. The data for p53tet-WT and p53tet-E343K/E346K can best be analyzed by assuming a monomer-tetramer equilibrium. The free energy change upon tetramerization, ΔG°, for p53tet-WT was calculated to be -23.4 kcal/mol, which corresponds to a K_d (total protein concentration at which half the protein exists as a tetramer) of 2.0 μM. This agrees well with published values for similar p53tet peptides (58,59). Using the same data treatment, the ΔG° for tetramer formation of p53tet-E343K/E346K was found to be -20.0 kcal/mol, which corresponds to a Kd of 13.3 μM. Thus, these calculations confirm the decreased stability of p53tet-E343K/E346K relative to p53tet-WT. The apparent oligomeric state of p53tet-K351 E was calculated to be 2.0, indicating that this mutant is a dimer in solution. Data were also fitted to a monomer- dimer equilibrium; the Kd was calculated to be very low (<10 -15 M). The oligomeric states of the p53tet mutants were further confirmed by size- exclusion chromatography (SEC). Representative elution profiles are depicted in Figure 9. Under these conditions, the tetrameric p53tet-WT construct eluted at 10.3 min as a tetramer with an apparent molecular weight of 44.3 kDa. As has been noted previously (48,50), the observed value is greater than the expected molecular weight for tetrameric p53tet-WT, owing to the relatively unique nature of its tetrameric structure. The M340Q/L344R mutant of p53tet, which is known to form dimers (53), eluted at 11.5 min, establishing the retention time for a dimeric form of such constructs. The p53tet mutants E343K, E346K, and E343K/E346K (Figure 9) elute at times similar to p53tet-WT indicating that these mutants exist as tetramers. The analog p53tet-K351E eluted at 11.3 min, pointing out that this construct, as in the case of M340Q/L344R, is a dimer. This finding supports our analytical ultracentrifugation results (see above). p53tet constructs with charge inversions at positions 351 and 343/346 are less stable

As shown above, sedimentation equilibrium ultracentrifugation data indicate that the p53tet-E343K/E346K tetramer is less stable than p53tet-WT, with a difference in ΔG° of tetramerization is 3.4 kcal/mol. In addition, circular dichroism (CD) was used to measure the thermal stability of wild type and mutant forms of p53tet. All the constructs displayed a sigmoidal unfolding curve, which is indicative of co-operative unfolding. Under the conditions of this study (25 mM sodium phosphate, 100 mM NaCl, 10 μM of each p53 constructs based on monomer concentration), p53tet-WT has a thermal unfolding temperature (T_m) of 68 °C. The effect of temperature on the fraction of folded structure as calculated from changes in ellipticity at 222 nm (Figure 10, and Table 5) suggests that the inversion of charges at residues 343, 346 and 351 resulted in less stable tetramers. The E343K and E346K mutations both posted lower Tm values in relation to the wild type p53tet construct. The E346K mutation has the greater effect on Tm (60 °C) than the identical mutation at position 343 (E343K, 67 °C). When both E343K and E346K mutations are included, the destabilizing effect is greater (Tm of 57 °C). The K351 E mutation to p53tet alone displays the largest destabilizing effect (T_m of 53 °C), demonstrating an important role for Lys351 in stabilizing the tetramer. p53tet-K351E and p53tet-E343K/E346K specifically form a heterotetramer

Since p53tet-K351 E forms a dimer in solution, it can be used to study the formation of heterotetramers with other p53tet species. Size exclusion chromatography results show that when p53tet-K351 E is mixed at equal proportions with any one of the p53tet mutants E343K (data not shown), E346K (data not shown), or E343K/E346K (Figure 9), the result is a single peak corresponding to a tetrameric species. This finding indicates that p53tet- K351 E associates with these mutants to form a 2:2 heterotetramer. This association is specific. For instance, p53tet-K351 E does not associate with p53tet-WT as demonstrated by the elution profile of an equimolar mixture of these two proteins (Figure 9). Two peaks are observed for this mixture, with elution times very similar to the individual p53tet components (10.5 min [for the wild type tetramer] and 11.2 min [for the K351E dimer]). The substantial amount of dimeric species in the size exclusion elution profile of the mixture of p53tet-K351 E and p53tet-WT indicates that an association of p53tet-WT with p53tet-K351E does not occur.

Sedimentation equilibrium ultracentrifugation data (Figure 8, Table 4) also show that an equimolar mixture of p53tet-E343K/E346K and p53tet- K351 E produces a new species that occurs as a tetramer (single state free-fit revealed an apparent oligomeric state of 3.9, identical to that of p53tet-WT). Importantly, both SEC and ultracentrifugation showed the absence of any dimeric species, suggesting that no uncomplexed p53tet-K351 E remains in solution after mixing. A fit of the ultracentrifugation data for the p53tet- E343K/E346K + p53tet-K351 E mixture to a monomer-tetramer equilibrium revealed an apparent ΔG° of tetramerization of -23.5 kcal/mol, which corresponds to a K_d of 1.8 μM. Due to the complex nature of the mixture (monomers, homodimers, homotetramers, and heterotetramers could all potentially be present), the significance of these values is open to question. However, the close correspondence between the ΔG° and K_d values obtained with those observed for p53tet-WT (ΔG° = -23.4 kcal/mol, K_d = 2.0 μM) suggests that the stability of the heterotetramer is similar to that of p53tet-WT.

Temperature melting curves derived from circular dichroism spectra

(Figure 10b) have precisely demonstrated that the increased stability associated with this heterotypic interaction is much more dramatic with the E343K/E346K double mutant than either p53tet-E343K or p53tet-E346K alone. The temperature curves presented in Figure 10 show that when p53tet-E343K or p53tet-E346K is combined with p53tet-K351 E, the melting temperatures of the resulting mixtures (Table 5) are lowered relative to the Glu-to-Lys single mutant alone (T_m of 66 °C for p53tet-E343K + p53tet-K351 E versus 67 °C for p53tet-E343K; T_m of 58 °C for p53tet-E346K + p53tet-K351 E versus 60 °C for p53tet-E346K). This finding indicates a lack of a stabilizing interaction between these p53tet components. In contrast, Figure 10 and Table 5 show that when p53tet-E343K/E346K and p53tet-K351 E are combined, the result is a new species with a T_m of 66 °C. This temperature is dramatically higher than that of either of the two mutants alone (57 °C for p53tet-E343K/E346K and 53 °C for p53tet-K351 E) and nearly equal the value observed for p53tet-WT (68 °C). These CD unfolding results support the sedimentation equilibrium ultracentrifugation data, which also suggested that the stability of the heterotetramer is similar to that of the p53tet-WT tetramer. The selective capture of His-tagged p53tet complexes on a metal-affinity resin demonstrates the existence of heterotetramers

The specific heterotypic association of p53tet-E343K/E346K and p53tet-K351 E was further assessed by monitoring the binding of protein complexes composed of Hisβ-tagged and non-Hisβ-tagged p53tet constructs to cobalt-bound affinity resin (Talon resin; Clontech). The resulting band patterns observed by SDS-PAGE (Figure 11) provide a unique signature describing the nature of the complexes formed. Figure 11a shows the results of combining together p53tet-E343K/E346K and p53tet-K351E. Lanes 1-3 clearly show that the combination of Hisβ-tagged p53tet-E343K/E346K and Hisβ-tagged p53tet-K351E both specifically bind the Talon resin (lane 2, no band) but are eluted with a high concentration of imidazole (lane 3). As expected, neither of the p53tet constructs lacking the Hisβ tag (lanes 4-6) were able to bind to the metal affinity resin (protein bands are found in the wash fraction, lane 5). p53tet-E343K/E346K migrates slightly faster than p53tet-K351E by SDS-PAGE, such that His6-p53tet-E343K/E346K runs at the same position as cleaved p53tet-K351 E (lane 7). When this mixture is bound to the affinity resin, both species bind to the resin (lane 8, no band) and both species elute with imidazole (lane 9). This finding demonstrates the specific association of these two proteins to form a complex with affinity for metal ions. The reciprocal combination yielded the same result, i.e. the association of cleaved p53tet-E343K/E346K and His6-p53tet-K351 E resulted in a complex retained by the metal-bound resin (lanes 10-12). In this case, separate bands for the two species can be clearly seen due to their differences in their electrophoretic mobilities.

The temperature melting curves for various p53tet combinations (Figure 10) indicated that the formation of a heterotetramer with p53tet-K351 E is most specific and strongest for the E343K/E346K double mutant. These findings are also supported by results from metal-affinity capture experiments. As Figure 11b shows, His₆-p53tet-E343K and His₆-p53tet-E346K both interact with p53tet-K351E, resulting in the retention of p53tet-K351E on the Talon matrix (lanes 1-3 and 4-6, respectively). However, both experiments yielded a significant amount of cleaved p53tet-K351 E in the wash fractions (lanes 2 and 5). This is almost as much as when a mixture of His6-p53tet-WT and cleaved p53tet-K351E was evaluated (lanes 7-9). More importantly, lanes 10-12 show that His6-p53tet-WT does not bind to cleaved p53tet-E343K E346K either, as a substantial amount of this protein is also found in the unbound fraction. Thus p53tet-E343K/E346K and p53tet-K351E form heterotetramers with each other but not with p53tet-WT. DISCUSSION

The tetramerization domain of human p53 is an important part of this key tumor suppressor protein. An analysis of the dimer-dimer interface of the human p53 tetramerization domain suggests that ion pair interactions between Glu343, Glu346 and Lys351 may contribute significantly to the stability of the tetramer. This hypothesis was further supported by the fact that sequences of the tetramerization domain from p53 in other organisms, as well as from human p63 and p73 sequences (Figure 7c) display the naturally occurring Glu343Lys mutation in their tetramerization domain. This mutation is always coupled with a corresponding loss of the positively charged lysine residue at position 351. This hypothesis was tested by designing and analyzing variants of the tetramerization domain of human p53, namely p53tet-E343K, p53tet-E346K, p53tet-E343K/E346K and p53tet-K351 E, harboring charge-reversal mutations at ionic residues.

In the first part of this study, the oligomeric state of these p53tet mutants was evaluated (Figure 9 and Table 4) in order to determine if these mutations, as is the case with many mutations to hydrophobic residues in the dimer-dimer interface (48,50,53), change the oligomerization specificity of p53tet from a tetramer to a dimer. Indeed, it was revealed that p53tet-K351E is a dimer in solution, demonstrating that a single mutation, to a charged residue, is sufficient to produce a dimeric p53. This finding confirms the hypothesis that the introduction of a charge reversal mutation (Lys to Glu) at position 351 of the p53tet domain introduces non-constructive charge repulsions at the dimer-dimer interface.

In the next part of this study, the stability of the resulting p53tet mutants was evaluated. It has been established that the oligomeric state and folding pattern of the p53 tetramerization domain are tightly linked features of this protein scaffold, with the monomeric form being essentially unfolded (48,51 ,60). Thus, thermal unfolding patterns as measured by CD represent an indicator of the tendency of the p53tet domain to oligomerize. As expected for a tetrameric protein, the T_m of p53tet is dependent on protein concentration (48,60). A p53 monomer concentration of 10 μM was thus selected in order for p53tet-WT to be fully unfolded at 98 °C, allowing the T_m values between p53tet-WT and its variant to be compared. The thermal unfolding temperature of p53tet is also dependent on the length of the protein or peptide used (48). However, the observed T_m value of 68 °C for the 72- amino acid long p53tet-WT construct used in this study was comparable to published values for related p53 constructs analyzed under these conditions (47,48,60). The thermal unfolding results shown in Figure 10 reveal that p53tet-E346K is less stable than p53tet-E343K, suggesting that Glu346 is involved in a more pronounced stabilizing interaction at the dimer-dimer interface. Glu346 rather than Glu343 may thus more strongly interact with Lys351 , in contrast with predictions arising from the crystal structure (44). CD thermal unfolding studies showed that p53tet-K351 E, in addition to being a dimer, is also very unstable, suggesting that the charge at position 351 is an important determinant of the stability of the p53 tetramer.

Subsequent experiments were undertaken to determine the potential of these p53tet mutants to form heterotetramers. Size-exclusion chromatography (Figure 9) and analytical ultracentrifugation data (Table 4) both indicated that either the E343K or the E346K mutation is sufficient to produce a species that specifically forms heterotetramers with p53tet-K351E. However, CD data (Figure 10 and Table 5) suggest that when both E343K and E346K mutations are included, the resulting heterotetramer with p53tet- K351E is much more stable relative to the two individual components. The specificity of the heterotetramer between p53tet-E343K E346K and p53tet- K351 E was also confirmed by metal affinity experiments (Figure 11), which strikingly depicts the necessity of both Glu-to-Lys mutations in determining the specificity of the heterotetramer. It was also found that these two mutants specifically associate with each other, and not with wild-type human p53tet. This interesting finding would suggest that human p53 mutants containing such mutations would indeed not have a "dominant negative" effect on cellular transformation since their intracellular expression would not directly compete or exchange with existing cellular pools of wild-type human p53 (61 ,62). This study demonstrates for the first time the important contribution of ionic interactions involving Glu343, Glu346 and Lys351 in the stability of the dimer- dimer interface of human p53tet domain.

While the present invention has been described with reference to what are presently considered to be the preferred examples, it is to be understood that the invention is not limited to the disclosed examples. To the contrary, the invention is intended to cover various modifications and equivalent arrangements included within the spirit and scope of the appended claims.

All publications, patents and patent applications are herein incorporated by reference in their entirety to the same extent as if each individual publication, patent or patent application was specifically and individually indicated to be incorporated by reference in its entirety.

TABLE 1 Tetramerization Domains

*H= Human

Table 2. Oligomerization state of synthetic p53 constructs.

*ND: not determined. tet=tetramer, mono=monomer

TABLE 3 Transfection efficiency ratios between pairs of p53 tetramer constructs in CHO cells

Tetramer/Tetramer Ratio (Luciferase signal RLU/mg protein)

Constructs NLS-10R- 10R-p53^KI NLS-10K-p53^tat 10K-p53"" NLS-TAT-p53^Kt TAT-p53 P53**

NLS-10R-p53^,et 1

10R-p53 ^let 4 1

NLS-10K-p53'^e 2 2 1

10K-p53^,et 86 24 41 1

NLS-TAT-p53^,e' 130 37 64 2 1

TAT-p53^,et 320 87 150 1 2

TABLE 4

Apparent mass and oligomeric state of p53tet constructs derived form sedimentation equilibrium ultracentrifugation studies.

Molecular weights are in kDa. ^b "Apparent MW/monomer MW' is referred to as the apparent oligomeric state of the protein; the molecular weight of the p53tet constructs used is 8.15 kDa

TABLE 5 thermal unfolding temperature (T_m) of p53tet constructs derived from circular dichroism studies

FULL CITATIONS FOR REFERENCES REFERRED TO IN THE SPECIFICATION

1. Bird, R. E., Hardman, K. D., Jacobson, J. W., Johnson, S., Kaufman, B. M., Lee, S. M., Lee, T., Pope, S. H., Riordan, G. S. & Whitlow, M. (1988) Science 242, 423-6. 2. Ward, E. S., Gussow, D., Griffiths, A. D., Jones, P. T. & Winter, G. (1989) Nature 341 , 544-6. 3. Chaudhary, V. K., Queen, C, Junghans, R. P., Waldmann, T. A., FitzGerald, D. J. & Pastan, I. (1989) Nature 339, 394-7.

4. Murphy, J. R., Bishai, W., Borowski, M., Miyanohara, A., Boyd, J. & Nagle, S. (1986) Proc Natl Acad Sci U S A 83, 8258-62.

5. Shockley, T. R., Lin, K., Nagy, J. A., Tompkins, R. G., Yarmush, M. L. & Dvorak, H. F. (1992) Cancer Res 52, 367-76.

6. Yazdi, P. T., Wenning, L. A. & Murphy, R. M. (1995) Cancer Res 55, 3763- 71.

7. Gariepy, J. & Kawamura, K. (2001) Trends Biotechnol 19, 21-8.

8. Sheldon, K., Liu, D., Ferguson, J. & Gariepy, J. (1995) Proc Natl Acad Sci U S A 92, 2056-60.

9. Singh, D., Kiarash, R., Kawamura, K., LaCasse, E. C. & Gariepy, J. (1998) Biochemistry 37, 5798-809.

10. Singh, D., Bisland, S. K., Kawamura, K. & Gariepy, J. (1999) Bioconjug

Chem 10, 745-54. 11. Bisland, S. K., Singh, D. & Gariepy, J. (1999) Bioconjug Chem 10, 982-92. 12.Tuchscherer, G., Grell, D., Mathieu, M. & Mutter, M. (1999) J Pept Res 54,

185-94. 13.Tuchscherer, G. & Mutter, M. (1995) J Pept Sci 1 , 3-10. 14. Chen, Q. R., Zhang, L, Stass, S. A. & Mixson, A. J. (2001) Nucleic Acids Res 29, 1334-40.

15. Domingues, H., Cregut, D., Sebald, W., Oschkinat, H. & Serrano, L. (1999)

Nat Struct Biol 6, 652-6. 16. Ballinger, M. D., Shyamala, V., Forrest, L. D., Deuter-Reinhard, M., Doyle,

L. V., Wang, J. X., Panganiban-Lustan, L., Stratton, J. R., Apell, G.,

Winter, J. A., Doyle, M. V., Rosenberg, S. & Kavanaugh, W. M. (1999) Nat

Biotechnol M, 1199-204. 17. Muller, K. M., Arndt, K. M. & Alber, T. (2000) Methods Enzymol 328, 261-

82. Iδ.Tang, A., Wang, C, Stewart, R. J. & Kopecek, J. (2001) J Control Release

72, 57-70. 19. Lee, W., Harvey, T. S., Yin, Y., Yau, P., Litchfield, D. & Arrowsmith, C. H. (1994) Nat Struct Biol 1 , 877-90.

20. McCoy, M., Stavridi, E. S., Waterman, J. L, Wieczorek, A. M., Opella, S.

J. & Halazonetis, T. D. (1997) Embo J 16, 6230-6. 21. Mateu, M. G. & Fersht, A. R. (1998) Embo J 17, 2748-58. 22. Davison, T. S., Yin, P., Nie, E., Kay, C. & Arrowsmith, C. H. (1998) Oncogene 17, 651-6.

23. Bullock, A. N., Henckel, J., DeDecker, B. S., Johnson, C. M., Nikolova, P.

V., Proctor, M. R., Lane, D. P. & Fersht, A. R. (1997) Proc Natl Acad Sci U

24. Mateu, M. G. & Fersht, A. R. (1999) Proc Natl Acad Sci U S A 96, 3595-9. 25. Mateu, M. G., Sanchez Del Pino, M. M. & Fersht, A. R. (1999) Nat Struct

Biol 6, 191-8. 26. Leonetti, J. P., Degols, G. & Lebleu, B. (1990) Bioconjug Chem 1 , 149-53.

27. Mi, Z., Mai, J., Lu, X. & Robbins, P. D. (2000J Mol Therl, 339-47.

28. Ryser, H. J. & Shen, W. C. (1978) Proc Natl Acad Sci U S A 75, 3867-70. 29. Shen, W. C. & Ryser, H. J. (1978) Proc Natl Acad Sci U S A 75, 1872-6.

30.Vives, E., Brodin, P. & Lebleu, B. (1997) J Biol Chem 272, 16010-7. 31. Efthymiadis, A., Briggs, L. J. & Jans, D. A. (1998) J Biol Chem 273, 1623-

8. 32. Kalderon, D., Richardson, W. D., Markham, A. F. & Smith, A. E. (1984) Nature 311 , 33-8.

33. Kalderon, D., Roberts, B. L, Richardson, W. D. & Smith, A. E. (1984) Cell

39, 499-509. 34. Futaki, S., Suzuki, T., Ohashi, W., Yagami, T., Tanaka, S., Ueda, K. & Sugiura, Y. (2001) J Biol Chem 276, 5836-40.

35. Chan, C. K. & Jans, D. A. (1999) Hum Gene Ther W, 1695-702.

36. Pack, D. W., Putnam, D. & Langer, R. (2000) Biotechnol Bioeng 67, 217- 23.

37. Midoux, P. & Monsigny, M. (1999) Bioconjug Chem 10, 406-11. 38. Boussif, O., Lezoualc'h, F., Zanta, M. A., Mergny, M. D., Scherman, D., Demeneix, B. & Behr, J. P. (1995) Proc Natl Acad Sci U S A 92, 7297-301. 39. Godbey, W. T., Wu, K. K. & Mikos, A. G. (1999) J Control Release 60, 149-60.

40. Yang, A., Kaghad, M., Wang, Y., Gillett, E., Fleming, M. D., Dotsch, V.,Andrews, N. C, Caput, D., and McKeon, F. (1998) Mol. Cell 2, 305-316.

41. Kaghad, M., Bonnet, H., Yang, A., Creancier, L., Biscan, J. C, Valent, A., Minty, A., Chalon, P., Lelias, J. M., Dumont, X., Ferrara, P., McKeon, F., and Caput, D.(1997) Cell 90, 809-819.

42. Soussi, T., Caron de Fromental, C, Mechali, M., May, P., & Kress, M.

(1987) Oncogene 1. 71-78. 43. Caron de Fromental, C, Pakdel, F., Chapus, A., Baney, C, May, P., & Soussi, T. (1992) Gene 112, 241-245. 44. Jeffrey, P. D., Gorina, S., and Pavletich, N. P. (1995) Science 267, 1498- 1502.

45. Lee, W., Harvey, T. S., Yin, Y., Yau, P., Litchfield, D., and Arrowsmith, C. H. (1994) Nat. Struct. Biol. 1 , 877-890.

46. Davison, T. S., Yin, P., Nie, E., Kay, C, and Arrowsmith, C. H. (1998) Oncogene 17, 651-656.

47. DiGiammarino, E. L, Lee, A. S., Cadwell, C, Zhang, W., Bothner, B., Ribeiro, R. C, Zambetti, G., and Kriwacki, R. W. (2002) Nat. Struct. Biol. 9, 12-16.

48. Mateu, M. G., and Fersht, A. R. (1998) EMBO J. 17, 2748-2758. 49. Mateu, M. G., and Fersht, A. R. (1999) Proc. Natl. Acad. Sci. U. S. A. 96, 3595-3599.

50. McCoy, M., Stavridi, E. S., Waterman, J. L, Wieczorek, A. M., Opella, S. J., and Halazonetis, T. D. (1997) EMBO J. 16, 6230-6236. 51. Mateu, M. G., Sanchez Del Pino, M. M., and Fersht, A. R. (1999) Nat.

Struct. Biol. 6, 191-198. 52. Noolandi, J., Davison, T. S., Volkel, A. R., Nie, X., Kay, C, and Arrowsmith, C. H. (2000) Proc. Natl. Acad. Sci. U. S. A. 97, 9955-9960.

53. Davison, T. S., Nie, X., Ma, W., Lin, Y., Kay, C, Benchimol, S., and

Arrowsmith, C. H. (2001) J. Mol. Biol. 307, 605-617. 54. Clore, G. M., Ernst, J., Clubb, R., Omichinski, J. G., Kennedy, W. M.,

Sakaguchi, K., Appella, E., and Gronenbom, A. M. (1995) Nat. Struct. Biol. 2, 321-333.

55. Landt, O., Grunert, H. P., and Hahn, U. (1990) Gene 96, 125-128.

56. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) Molecular cloning: A laboratory manual, 2nd Ed., Cold Spring Harbor Laboratory, Cold Spring

Harbor, NY. 57. Johnson, M. L, Correia, J. J., Yphantis, D. A., and Halvorson, H. R. (1981)

Biophys. J. 36, 575-588. 58. Sakaguchi, K., Sakamoto, H., Lewis, M. S., Anderson, C. W., Erickson, J.

W., Appella, E., and Xie, D. (1997) Biochemistry 36, 10117-10124. 59 Sakamoto, H., Lewis, M. S., Kodama, H., Appella, E., and Sakaguchi, K. (1994) Proc. Natl. Acad. Sci. U. S. A. 91 , 8974-8978.

60. Johnson, C. R., Morin, P. E., Arrowsmith, C. H., and Freire, E. (1995)

Biochemistry 34, 5309-5316. 61. Chene, P., and Bechter, E. (1999) J. Mol. Biol. 286, 1269-1274. 62. Shaulian, E., Zauberman, A., Ginsberg, D., and Oren, M. (1992) Mol. Cell. Biol. 12, 5581-5592.

63. Guex, N., and Peitsch, M. C. (1997) Electrophoresis 18, 2714-2723. 64. Thompson, J. D., Higgins, D. G., and Gibson, T. J. (1994) Nucleic Acids

Res. 22, 4673-4680.

Claims

WE CLAIM:

1. A vehicle for the delivery of an agent to a cell comprising a peptide domain that is capable of associating into multivalent assemblies.

2. The vehicle according to claim 1 , wherein the peptide domain can associate into tetrameric assemblies.

3. The vehicle according to claim 1 or 2, wherein the peptide domain comprises an amino acid sequence selected from the group consisting of the tetramerization domain containing residues 325 to 355 of human p53 (SEQ ID NO:1 , Table 1); the tetramerization domain of p63 (SEQ ID NO:2, Table 1); the tetramerization domain of p73 (SEQ ID NO:3, Table 1); the tetramerization domain of the p53 protein from Xenopus laevis (SEQ ID NO:4, Table 1); the tetramerization domain of the p53 protein from rainbow trout (SEQ ID NO:5, Table 1); p53tet mutant E343K (SEQ ID NO: 7, Table 1); p53tet mutant E346K (SEQ ID NO: 8, Table 1); and p53tet mutant E343K/E346K (SEQ ID NO: 9, Table 1), or a peptide mimetic, analog, homolog or derivative thereof.

4. The vehicle according to claim 3, wherein the peptide domain comprises an amino acid sequence selected from the group consisting of the tetramerization domain containing residues 325 to 355 of human p53 (SEQ ID NO:1 , Table 1); the tetramerization domain of p63 (SEQ ID NO:2); the tetramerization domain of p73 (SEQ ID NO:3, Table 1); the tetramerization domain of the p53 protein from Xenopus laevis (SEQ ID NO:4, Table 1); and the tetramerization domain of the p53 protein from rainbow trout (SEQ ID NO:5, Table 1.

5. The vehicle according to claim 4, wherein the peptide domain comprises the tetramerization domain from human p53 which contains residues 325-355 of human p53 or an analog, homolog, derivative or mimetic thereof.

6. The vehicle according to claim 1 , comprising two or more peptide domains that are capable of associating into heteromultimers.

7. The vehicle according to claim 6, wherein the heteromultimer is a heterodimeric tetramer.

8. The vehicle according to claim 7, wherein the heterodimeric tetramer comprises SEQ ID NO:9 and SEQ ID NO:10 as shown in Table 1 , or analogs, homologs, derivatives or mimetics thereof.

9. The vehicle according to any one of claims 1 to 8 wherein the peptide domain is fused to one or more signal molecules.

10. The vehicle according to claim 9, wherein the signal molecule is a polycationic molecule.

11. The vehilce according to claim 10, wherein the polycationic molecule is a polycationic amino acid sequence.

12. The vehicle according to claim 11 wherein the polycationic amino acid sequence is a polylysine or polyarginine sequence.

13. The vehicle according to any one of claims 1 to 12 further comprising one or more nucleus-directing signal sequences.

14. The vehicle according to claim 13 wherein the nucleus-directing signal sequence is a nuclear localization peptide.

15. The vehicle according to claim 14, wherein the nuclear localization peptide is the nuclear localization peptide pf the SV40 large T antigen.

16. The vehicle according to claim 1 , wherein comprising an amino acid sequence selected from the group consisting of SEQ ID NO: 12, SEQ ID NO:13, SEQ ID NO:16, SEQ ID NO:17, SEQ ID NO:20, SEQ ID NO:21 and SEQ ID NO:24, as shown in Figure 1 or a fragment, peptide mimetic, analog, homolog or derivative thereof.

17. The vehicle according to claim 16, comprising an amino acid sequence selected from the group consisting of SEQ ID NO:13, SEQ ID NO:16 and SEQ ID NO:17, as shown in Figure 1.

18. A method of delivering an agent to a cell comprising:

(a) forming a complex between a delivery vehicle according to any of claims 1 to 17 and the agent; and

(b) contacting the cell with the complex of (a) under conditions that allow the complex to be delivered to the cell.

19. The method according to claim 18, wherein the agent is selected from the group consisting of peptides, vaccines, cytotoxic molecules, nucleic acids, plasmids, genes, drugs and diagnostic probes or agents

20. The method according to claim 19, wherein the agent is nucleic acid.

21. The method according to claim 20, wherein the nucleic acid is positionally and sequentially oriented within a vector such that the nucleic acid can be transcribed into RNA and translated into proteins or polypeptides in the cell.

22. A cell transformed using the method according to claim 21.

23. A use of a vehicle according to any one of claims 1 to 17 to deliver an agent to a cell.

24. A use according to claim 23 wherein the agent is selected from the group consisting of peptides, vaccines, cytotoxic molecules, nucleic acids, plasmids, genes, drugs and diagnostic probes or agents.