AU2242799A - Nucleic acid delivery system, method of synthesis and uses thereof - Google Patents

Description

I

P100iohi Regulation 3.2

AUSTRALIA

PATENTS ACT 1990 COMPLETE SPECIFICATION FOR A DIVISIONAL PATENT

ORIGINAL

I..

TO BE COMPLETED BY APPLICANT DANA-FARBER CANCER INSTITUTE s: Wayne, A. MARASCO; Si-Yi CHEN Name of Applicant: Names of actual Inventor Address for Service: CALLINAN LAWRIE, 711 High Street, Kew, 3101, Victoria, Australia Invention Title: NUCLEIC ACID DELIVERY SYSTEM, METHOD OF SYNTHESIS AND USES THEREOF The following statement is a full description of this invention, including the best method of performing it known to me:-

I;

F;.

t i -i Oi ii 4 WO 95/22618 PCTIUS95O214O I A NUCLEIC ACID DELIVERY

SYSTEM,

1fMTOD OF SYNTHESIS A"D USE-' THEREOF In recent years, a new form of therapy, gene therapy, has been proposed to treat a variety Of maladies including cystic fibrosis (CF) [Rosenfeld, et al., cell 68:143-155 (1992); Rosenfeld,

M.A.,

at al., science 252:431-434 (1991), Ferkcol, et al., j. CIlin. Invest. 92:2394-2400 (1993)J], tumors such as retinoblastoma, diseases caused by infection by a virus such as the human immunodeficiency virus (HIV), for example Hly-1 infection (Baltimore, Mature 335, 395-396 (1988)], etc. In this form of therapy, a gene is introduced into calls so that the cells will express that gene. The gene can positively potentiate the cells, supply a missing protein, stimulate the immune system, or it may act in a negative manner, for instance expressing a viral inhibitor, which can result in the inhibition of the virus such as HIV-l replication and thus infection. Several approaches including anti-senlse RNA, ribozymes and dominantnegative mutants have been shown to be able to inhibit EXV-i infection at the cellular level (Hasselcf f I3, et al., Nature 334:585-591 (1988); Von der Krol, A. et al., BioTecbfl1ques 6:958-976 (1988); Malim, M. at al., Cell 58:205-214 (1989); Troflo, et al., Cell 59:113-120 (1989) Sullenger, et al., Cell 63:601- 608 (1990); Green, at al., Cell 58:215-223 (1989); Buonocore, et al., Nature 345:625-628 (1990)2]. The intracellular delivery and expression of a human antibody, such as an anti-gpl2o single chain antibody, is able to inhibit viral replication, etc. For example, the anti-gVl 2 0 antibody inhibits Ely-1 envelope glyc-oprotein maturation and function [Marasco, et al., PCT Application No. PClT/US93f06735, filed July 1993].

WO 9522618 PCTIUS95/02140 However, despite these advances, a major impediment to the development of gene therapy protocols for treatment and prevention of malignancies, diseases, etc. using any of these strategies is the relatively inefficient means to effectively transduce the desired genes into the desired target cells. Although murine retroviral vectors have been widely used to transfer gene into cells, they indiscriminately infect many cell types and limitedly infect desired targeted cells. In addition, retroviral vectors contain potentially hazardous viral DNA along with therapeutic genes.

Therefore, these vectors may not be optimal-as an efficient transfer system for the human gene therapy of, for instance, AIDS [Miller, Nature 357:455- 460 (1992); Eglitis, et al. Science 230:1395-1398 (1985); Dizerzak, E. et al., Nature 331:35-41 (1988)]. To resolve the problem of specific delivery for HIV infected cells, defective HIV vectors which can specifically transfer a gene into HIV susceptible cells have been developed. (Poznansky, et al., J. Virol. A 65:532-536 (1991); Shimada, et al., J. Clin.

Invest. 88:1043-1047 (1991)]. However, this approach may not be practical with all viruses and malignancies.

SFurther, the theoretical potential of recombinant ii rescue of the defective vector, however low, may impede its use. The delivery and expression of a recombinant gene into cells has also been achieved using liposomes, lipofectin, and calcium phosphate-precipitated methods either in vitro or in vivo [Nicolau, et al., Proc.

Natl. Acad. Sci. USA 80:1068-1072 (1983); Brigham, K.

L, et al., Am. J. Med. Sci. 298:278-281 (1989); Nabel, et al. Science 249:1285-1288 (1990); Benvenisty, N. et al., Proc. Natl. Acad. Sci. USA 83:9551-9555 (1986); Chen, et al., J. Virol. 65:5902-5909 (1991)]. These methods have several advantages over WO 95t22618 PCTIUS95I2140 3 retroviral systems for gene th erapy. Plasmid DNA constructs containing suitable promoter elements are technically easier and less time consuming to prepare and test than retroviral vectors. Plasm-id DflAs are more suitable for large-scale preparation than are the infectious retroviruises." Plasmid DNAs can also permit the delivery of larger-sized segments of DNA than is possible with retrovirus -based system. one additional advantage is that plasmnid DNA may exclude deleterious side effects of retroviral vectors such as virus infection or cancer in a small percentage of patients.

However, the potential for delivery of genes In vivo using these methods is limited by a lack of cell ,specificity and efficiency.

In an attempt to overcome th2 problem of cellspecific gene transfer, Wu and Wu, Biol.- Chemn.

262:4429-4432 (1987) have proposed a chemically coupled receptor-mediated gene transfer system which uses receptor-mediated endocytosis to carry DINA or RUA molecule into target hepatocytes or primary hematopoietic cells. The strategy of this system is based an the fact that such cells possess unique asialglycoprotein receptors on their surface that bind and internalize asialoglycoproteins, its ligand. The proteins (ligands) are preferably coupled to poly-Llysine which can bind DNA or MN to form soluble complexes by a strong, electrostatic interaction. This system has been reported to transfer genes into the targeted hepatocytes or primary hematopoietic cells at the cellular level as vell as in animal studies'(Wu, et al., J7. Blo. Chbem. 263:14621-14624 (1988); Zenke, et Proc. Mati. Acad. Sal. 87:36551-3659 (1990); Wu, G. etal., J7. Biol. Chem. 266:14338- 14342 (1991); Curiel, D. et al., Prac. Nrati. Acad.

Sdl. USA 88:8850-8854 (1991); Wagner, et al., Proc.

Mat2. Acad. Sci. USA 87:3410-3414 (1990); Curiel, D.T., WO 95122618 PCTIS9SO2 140 et al., Human Gene Therapy 2:230-238 (1992)]. However, the overall efficiency of this method has been reported to be relatively low because endocytosis is relatively inefficient in that the DNA frequently does not get out of the endosomal compartment and is ultimately degraded in lysosomes. Thus, multiple administrations are necessary and antigenicity of this system can be a problem. Furthermore, the synthesis of the delivery system is relatively time consuming as one has to first couple the poly-L-lysine to the asialoglycoprotein and then subsequently couple the ligand-polylysine complex to the exogenous DNA. Furthermore, in its typical application, the exogenous DNA introduced into the cell is not presented in a manner which is stably incorporated into the chromosome. Thus, expression is transient. Accordingly, repeated administration is necessary for this reason also. However, as mentioned the polylysine moiety as an artificial moiety may trigger an antigenic reaction limiting the ability to repeatedly use this system.

Another form of therapy that has been proposed is delivering an already expressed protein to the target cell. In one common form of cancer therapy, one introduces cytopathic or cytotoxic agents to the malignant cells in order to kill them. However, care must be taken to minimize the harm to healthy tissues and cells. Thus, strategies have been developed to try to specifically target the unhealthy cells. The use of immunotoxins is one method of such therapy. An immunotoxin is a class of cytotoxic agents consisting of a toxin protein linked to a monoclonal antibody or a ligand, which binds specifically to a target on the cell surface [Vitetta, et al. Science 238:1098 (1987); Pastan, et al., Cell 47:641 (1986); Pastan, et al., Science 254:1173 (1992)]. Due to the predicted specificity for the cell and the potential

I

*f a r WO 95/22618 PCTIUS95/02140 for efficacy, this therapy has been predicted to play an important role in therapy against cancer and various diseases. However, in practice this has not proven to be the case. Rather, the toxins are highly antigenic proteins. Neutralizing antibodies against these toxins typically arise within two weeks after the first exposure, severely limiting their effectiveness after only one or two therapy sessions. Thus, strategies, such as use of immunosuppressive agents to suppress immune reaction have been proposed. However, this is not only difficult to achieve, but may not be beneficial to the ultimate outcome of the therapy, function such as fighting infection, other tumor cells and pathogens.

Accordingly, it would be desirable to have a nucleic acid delivery agent that can be assembled more simply than other nucleic acid delivery systems, such as the delivery system of Wu and Wu.

It would also be desirable if such a delivery system could be synthesized more readily than is possible with a chemical coupling process.

It would also be desirable that the nucleic acid delivery system could readily be adapted to be used to specifically target a variety of target cells.

It would also be desirable if the delivery system had lower antigenicity than many currently available deli'ery systems. For example, it would be desired if it could be used to deliver a cytotoxic agent, e.g. an immunoatoxin, to a cell without the antigenicity currently associated with such systems.

W a 9 5 2 2 6 1 8P C T U S 9 5 0 2 1 4 0 6 It Would also be beneficial i ~isSse i o have the pote tia ofcausing disease on its own by mainn transformtion 1 o a cell sc scnoc With viral delivery systems.Suhacnocr We have now developed a highly efficient nlUcleic acid delivery system to a desired targept cell. This system can be used, for example, to deliver a gene coding for the essential portion of a toxin protein.

The nucleic acid, either DNA Or RUA, is Coupled to a fusion protein. The fusion protein consists of a target moiety and a nucleic acid binding moiety, for example a DNA binding moiety. For examIple, the target moiety Preferably can be an antibody, more preferably a single chain antibody, Ia Fab portion of the antibody or a (Fab')7. segment. -If the target animal is a_ human, the DNA binding moiL'ety should preferably be a human

DNA

binding moiety, such as Protamine.

Brief Descri tio of the Draw*n~ Figure is a Schematic representation of one embodient of the use of nucleic acid delivery system according to the present invention.

Figure 2 is a schematic representation of one emboiment of the expression vector for the fusion protein. It is a schematic representtion of a bi- cistronic mammalian expression vector, which will encode an antibody for the Eiv gpl2O protein fused -to a protamine protein.

Figure 3 is an autoradiogaph shoving radiola1beling, and immunoprecipitation of .4xpressed PablO 5 -protamine fusion proteins WO 95.22618 PCT/US95/02140 7 Figure 4 shows purification and SDS-PAGE analysis of the recombinant fusion proteins.

Figure 5 shows binding activity of the purified fusion proteins to HIV Figure 6 is an autoradiograph showing the DNA binding activity of the Fabl05-protamine fusion proteins under varying concentrations.

Figure 7 is an autoradiograph showing the DNA binding activity of the FablOS-protamine fusion proteins under varying concentrations.

Figure 8 are FACS analysis showing the binding ability of FablO5-protamine DNA complexes to protein as compared to that of Fabl05 complexes in both Iuninfected and HIV-infected cells.

Figure 9 is a schematic of the expression vectors of the PEA catalytic fragment schematically showing the PEA encoding gene and two vectors made containing partial domains of this gene.

Figure 10 is a graph showing selective cytotoxicity of one of the nucleic acid delivery systems of the present invention, toxin expressor, to HIV infected cells and shows cell viability.

Figure 11 is a graph showing selective cytotoxicity of FablOS-protamine-toxin expressor complexes to HIV-infected cells and shows a protein inhibition assay.

-7 ^"i WO 95/22618 PCT/US95/02140 8 Figure 12 shows selective cytotoxicity of the FablOS-protamine-toxin expressor complexes to HIVinfected cell as measured by ADP-ribosylation activity.

Detailed Description of the Invention We disclose herein a new nucleic acid delivery system. The system comprises a fusion protein which binds the desired nucleic acid sequence.

The fusion protein comprises a target moiety and a binding moiety. The target moiety is preferably a protein that will specifically bind to a site on the target cell. For example, i can be a ligand for a ligand specific receptor for instance a fibroblast growth factor receptor (FGF-R)) and the specific FGF for that receptor, e.g. basic FGF for a basic FGR-R.

Alternatively, the protein can be an antibody specific to the target cells. For example, -it can be an antibody to an HIV envelope protein, an antibody to an oncogenic determinant such as extracellular ligandbinding domain of an activated receptor, erbB, kit, fas, neu FrbB2), etc. It can also be an antibody to a receptor, for example, an antibody to the GM-CSF receptor. Preferably, the target moiety is an antibody. Still more preferably, the antibody is a single chain antibody comprising the binding sequence of the antibody, a (Fabr) 2 segment or the Fab fragment of the antibody. More preferably the antibody is a single chain antibody or a (Fab') 2 segment.

The particular target moiety chosen can be determined empirically based upon the present disclosure depending upon the target cell. For example,, with somatic cell therapy or In vivo with readily accessible cells or tissues such as an WO 95/22618 PCT/US95/02140 intravascular target, the important attributes of the target moiety are affinity and selectivity. In such instances the use of single chain antibodies as the target moiety is preferable. However, when the target cell is not readily accessible, such as when the cell is part of a large solid tumor mass with a poor blood supply and high interstitial pressure, the serum halflife is extremely important to consider. In such instances, the full antibody and (Fab') 2 segments are typically preferred. In a preferred embodiment, one could synthesize the fusion protein so that the binding moiety is attached to the carboxy terminus of an intact immunoglobulin such as IgG 1 In order to limit antigenic reaction, the targeting moiety is preferably selected to take into account the host animal whose cells will be targeted.

Thus, if the target animal is a mouse, one preferably uses murine antibodies, whereas if the target animal is a human, one preferably uses a human antibody or a humanized antibody.

SThe second part of the fusion protein consists of a nucleic acid binding moiety, either a DNA or RNA binding moiety. Preferably, one uses a moiety that can bind either DNA or RNA. This binding moiety can be any protein from the target animal that will bind either DNA or RNA. For example, it can be protamine, which is a small basic DNA binding protein, which serves to condense the animal's genomic DNA for packaging .into the restrictive volume of a sperm head [Warrant, R.W., et al., Nature 271:130-135 (1978); Krawetz, et S- al., Genomics 5:639-645 (1989)]. The positive charges of the protamine can strongly interact with negative charges of the phosphate backbone of nucleic acid, such as DNA resulting in a neutral and stable DNA protamine complex. The nucleic acid can be either DNA WO3 95122618 PCTIUS95102140 or R~NA depending on the purpose. For example, the nucleic acid to be transferred can be used to express an antibody intracellularly, dominant negative mutants, anti-sense RNA, ribozymes or a cytotoxic agent. For example, the cytotoxic agent can be a portion of a bacteria or plant toxin which is extremely potent such as ricin, the catalytic fragment of Pseudomonas exotoxini A (PEA), etc.

ime nucleic acid can be used for transient or stable transfection of the cell. For example, when the nucleic acid encodes a factor which is lethal to the cell such as a DNA segment encoding a toxin, transient expression is sufficient. In contrast, where it expresses a factor such as a suppressor gene (e.g.

retinoblastoma), or a-protein that is not being expressed at sufficient levels, e.g. adenosine deaminase (ADA) [Belmont, et al., Mo2. Cell.

Biol. 8:5116-5125 (1988) Palmer, at al., Proc.

Hati. Acad. Scl MS:0515 (1987), rdn diphosphate (UDP) -glucuronyl-transf erase (Ponder, K.P., et al., Proc. Hatl. Acad. Sci USA 88:1217-1221 (1991), or insulin, stable integration into the cells ch~romosome may be desired. In those instances where stable integration is desired the nucleic acid can be a DNA segment wherein the gene coding for the desired factor is inserted into a cassette that will facilitate integration into the cell. For instance, the integration cassette which s urr ounds the gene can be a and 3' LTR (long terminal repeat) of a retrcavirus i.e. MMV, an ITR (inverted terminal repeat unit, i. e.

adeno associated virus), etc. [See, Scherdin, et J7. Vizal. 64:907-912 (1990); Stief, et al., Nature 341:343-345 (1989); Phi-Van, et al., ffol. &Cell. B101. .10:2302-2307 (1990); Phi-Van, et al., The EMBO Journal 7:655-664 (1988) This cassette, can be prepared by standard techniques. For example, WO 95122618 PCT/US95/02140 11 mammalian expression vectors where a gene of interest can be inserted between LTRs or ITRs. One can construct a cassette containing flanking LTR or ITR regions at both ends, a promoter/enhancer, preferably with a polylinker for the gene of interest to be inserted in between, and when desired a selectable marker based upon the present disclosure using known techniques. This cassette with the desired nucleic acid, e.g. gene or genes, of interest is the nucleic acid segment.

The target moiety specifically brings the delivery system to the target cell.

One can also use localization sequences to intracellularly deliver the released RNA or DNA to a cellular site of interest.

I Thereafter, the targeted cell can internalize the delivery system, which is bound to the cell.

Typically, the delivery system binds to a specific receptor on the cell.

For example, membrane proteins on the cell I surface, including receptors and antigens can be internalized by receptor mediated endocytosis after i interaction with the ligand to the receptor or antibodies. [Dautry-Varsat, et al., Sci. Am.

250:52-58 (1984)]. This endocytic process is exploited by the present delivery system. Because this process can damage the DNA or RNA as it is being internalized, S:it is preferable to include a strong promoter for the nucleic acid that is to be expressed. Similarly, the use of a segment containing multiple repeats of the gene of interest may be desirable. One can also include sequences or moieties that disrupt endosomes and lysosomes. See, Cristiano, at al., ai m 1 WO 95/22618 PCT/US95/02140 12 Proc. Natl. Acad. Sci. USA 90:11548-11552 (1993); Wagner, et al., Proc. Natl. Acad. Sci. USA 89:6099- 6103 (1992); Cotten, et al., Proc. Natl. Acad. Sci.

USA 89:6094-6098 (1992).

In deciding what type of nucleic acid segment to use, the skilled artisan will take into account the protein being expressed in light of the present specification. For example, when one is introducing a toxin protein, because of its extreme cytotoxicity, the expression of only a few molecules are needed to kill a cell. In other cases such as with expressing ADA, larger amounts of protein expression are needed and the use of LTRs, ITRs, as part of the DNA cassette, and/or lysoscmal disrupting agents such as replication- 1^E defective adenoviruses may be used.

The particular protein chosen for the targeting moiety will depend upon the target cell. For example, if one is targeting an infected cell, such as an HIV infected cell, one can use a monoclonal antibody that will specifically target HIV infected cells. This would include an antibody against the envelope Sglycoprotein. One can use any of a number of known antibodies against HIV-1 gpl20 or HIV-2 gpl20, such as 21h [Thali, et al., J. Virol. 67:3978-3988 (1993)], F105, 176 and 48d. If one wants to deliver the nucleic acid sequence prophylactically such as a gene for intracellular expression of an antibody, a decoy sequence, etc., one can target highly susceptible cells by targeting receptors present on such cells such as the CD4 receptor for HIV susceptible cells. In such a situation, the protein can be a ligand that will preferentially bind to the receptor, for example, CD4, as well as using an antibody to the receptor, such as an antibody to the CD4 receptor.

WO 95122618 prU9I24 13 This strategy f or choosing the targeting moiety is very adaptable. For example, certain tumors are frequently associated with cells possessing a large amount of a particular cell surf ace receptor neu with breast cancers), or an abnormal form of a particular protein.

other receptors of interest include those for lymphokineS such as interleukins and interferons, for example, the interleukin- 2 (IL-2) receptor (IL-2R) The p55, IL-ZR a chain also ref erred to as the Tac, protein is associated with -3g or mitogen-activated

T-

calls but not resting T-cells. It is expressed in high levels on malignanlt cells of lymphoid cancers such as adult T-cell leukemia, cutaneous T-cell lymphoma and Hodgkins disease. The anti-TaC antibody will bind to this protein. Humanized version of such antibodies are known and described in Queen, at al., Proc. HatI.

Acad. Sci. USA:10029-1 0 0 3 9 (1989); Halcimi, et al., j. of. Immun. 151:l075-1085 (1993) (Xi41 which is a Nab against IL-2R 0 chain); Kreitmanf, et al., J. of Izumun. 149:2810-281s (1992); Hakinhi, et al., JT. Of IzunD 147:1352-1359 (1991).

Antibodies to these various proteins are known and available. These antibodies can readily be adapted f or use in this system by following the general procedures described herein, and substituting the gene coding for the desired binding site for the exemplified gene. For example, where the targeted cell is an HIV-infected cell, the targeting moiety can target the EIV envelope glycoprotein. Any number of antibodies to this protein -can be used. For instance, a recombinant atibody based on the FI05 antibody is made by known teachings techniques. (Posner, et al., J. Izumunol.

146:4325-4332 (1991); Thali, et al, J. Virol.

65:6188-6193 (1991); Ma.rasco, at al., Proc. Nati.

WO 95/22618 PCTIUS9502140 14 Acad. Sci. USA 90:7889-7893 (1993)] other antibodies that can be made include, 15e, 21h, 17b, 48d, etc.

A vector for expression of the antibody can be made as described herein. For example, a bicistronic mammalian expression vector which will express the Fd portion of the antibody (VH and CH) and the binding region of the light chain a kappa chain) of, for example, the F105 antibody can be constructed by using an Fd fragment without a stop codon and amplifying the segment by standard techniques, for example by polymerase chain reaction (PCR). The upstream primer preferably will correspond to the leader sequence of the immunoglobulin of the animal from which the cells of the delivery agent is to be used (for example, where the target cell is a human cell a human immunoglobulin of amino acids with an additional convenient cloning site such as a HindIII site. The downstream primer can correspond to amino acids by the carboxy terminus of the heavy chain constant region. For example, with an antibody based upon F105, amino acids 4 226-233 of human heavy chain CHI domain with a convenient cloning site inserted, such as the XbaI site. The PCR reaction is performed according to standard means. By this means the gene or gene segment encoding the targeting moiety of the fusion protein is prepared.

As described above, the second portion of the fusion protein is the binding moiety. Preferably,- one uses a single vector containing gene segments that will express both the targeting moiety and the binding i moiety. However, one can use a vector system to co- B ~transfect a cell with at least two vectors and select Sfor cells expressing the fusion protein. Preferably, one uses a single vector. One preferably attaches the sequence encoding the target moiety to a gene, or gene p" WO 95/22618 PCT/US95/02140 segment, encoding the binding moiety by standard means.

For example, a gene for human protamine [Balhorn, J. of Cell. Biol. 93:298-305 (1982)]. Other nucleic acid binding proteins include GCN4, Fos and Jun which bind DNA through a common structural motif consisting of several basic residues and an adjacent region of about residues containing a heptad repeat of leucines, the "leucine zipper" that mediates dimerization [Talanian, et al., Science 249:769-771 (1990)]; the TFIIS nucleic acid binding domain, which is seen in the Cterminal residues 231-280 [Qlan, et al., Nature 365:277-279 (1993)]; the ribonucleoprotein (RNP) family that is present in domains in human FMRI, the yeast protein HX, 14 domains of the chicken gene vigillin, merl, a yeast protein, bacterial polynucleotide phosphoylase, and the ribosomal protein S3 [Ashley, et al., Science 262:563-566]; and the binding motifs in heat.shock protein [Rabindran, et al., Science 259:230-234 (1993)]. The host animal of the target cells will be used to determine which protein or protein fragment with a binding motif is used. For example, with a human host and for expression of human protamine one can use the known plasmid pTZ 19R-HP1 [Krawetz, et al. Genomics 5:639-645 (1989)].

Preferably one would delete an intron in this gene so that the expression vector can also be used for expression of the fusion protein in prokaryotic systems as well as eukaryotic systems. PCR amplification would be performed by standard means. For example, using an upstream primer, which corresponds to sequences from the amino terminus, for example, corresponding to amino acids 1-6 of the protamine protein with a convenient restriction site, such as the Xbal cloning site and a downstream primer corresponding to the carboxy portion of the first exon, for example, amino acids 29-37 -with additional sequences complimentary to the 5' amino acids in the second terminus amino acids 38-40 in WO 95/22619 PCTfUS95/0214 0 16 the second exon) A second PCR reaction can then be performed using the upstream primer correspondifig to the amino terminus and the downstream primer corresponding to an overlapping portion to the carboxy terminus. For example, using a sequence corresponding to amino acids 31-40 with the sequence of amino acids 41 to the stop codon in the second exon and an additional convenient cloning site such as NotI. The first PCR amplif ied DNA segment can be used as a template-. By using convenient restriction sites, one can cut out the targeting moiety and the binding moiety by known methods such as purifying them using standard techniques, agarose gel. For example, in the example described above, the PCR amplified Fd of Fins without a stop codon can be cut with HindIII/fXbdl and purified by agarose gel. The PCR amplif ied protamine coding gene, without intron can be cut with XbaI/Nctl and purified from agarose gel. The FablC5 plasmid can be cut with HindIII/NotI and the DNA segment purified from an agarose gel. The HindIII/XbaI-cut Fd fragment and XbaZ/Noti-cut protamine fragment can then be cloned into the HindIIl/Notl sites of the plasmid containing FlOS by three-piece ligation. See Figure 2. The resulting expression vector thus contains a cartridge of an Fd-protamine fusion gene (in-frame) and kappachain gene under the control of an independent promoter, such as a CMV promoter. The particular promoter that will be used depends upon the desired cell system for expression of the fusion protein.

Promoters are known to t~he skilled artisan and can readily be selected based upon the present disclosure.

For example, preferred promoters include CK~V, Sea, RSV, 1mv LTR, ma4 and HIV-*1 5' LTR.

This construct can readily; be confirmed by standard means, such as DNA sequencing.

-74

K

WO 95122618 PCT/US95/02140 17 This expression vector can then be used to stably transform a cell line. The cell line can be any desired cell line including prokaryotic as well as eukaryotic cells. Preferred cell lines include mammalian cell including COs cells, kidney cell lines such as CHO, myeloma cell lines such as SP/0, and SP/2, EMMA2-11 TG10, and insect cell lines such as Drosophilla. Preferably, to reduce antigenicity one would use a mammalian cell line. More preferably, one would use a myeloma cell- Preferred cells include SP/0, SP/2, Sp2/0-Agl4, X63Ag8.653, FO, NSI/l-Ag4-l, NSO/1, FOX-NY, YB2/0 and 1R983F.

The transformation of the cell can be by any standard techniques. It is preferred that one stably transforms the cell, although in certain instances transient transformation by the DEAE-Dextran technique will be acceptable. Thus, one preferably uses a method Sfor stably transforming the cell, such as the calcium phosphate precipitation method followed by selection of transformed cell lines such as by G418 selection. The transformed cell line can be cultured and the fusion protein harvested by standard techniques. For.

instance, the Fd-protamine protein and kappa-chain of F105 are expressed and secreted into the culture of COS transformed FablOS protamine cells and detected by radiolabelling and immunoprecipitation with anti-human IgG antibody.

For example, a COSI cell can be transfected with an expression vector containing the cartridge of the targeting moiety-binding moiety using lipofectin.

Vectors include the vector pCMV-Fabl05-protamine. The transfected cells can then he incubated in DMEM, supplemented with 10% fetal calf serum (FCS) for two days and replaced with a selection medium such as DMEM with 10% FCS and 500 pg G418. This is readily .L I-F WO 9522618 PCTIUS95/02140 18 available, for example, from BRL. The G418 resistant colonies will appear after about two weeks and can be readily selected. Colonies can then be cloned with limiting dilution and examined by radiolabelling and immunoprecipitation, ELISA and immunofluorescent staining for expression of the recombinant fusion protein. The proteins can be secreted and purified in these cells by standard means. For example, the transformed COS cells can be grown in a flask with DMEM medium supplemented with 10% fetal calf serum and 500 ig/ml of neomycin. After reaching confluence, the cultures can be replaced with fresh DMEM without FCS every three days for two weeks. The collected culture mediums can be clarified by, for example, centrifugation at, for instance, at 500 rpm for minutes at 4°C and then concentrated using, for example, a membrane filter with a molecular weight cutoff of 10,000 dalton such as an Amico concentrator.

The concentrated medium can then be loaded into an affinity column coupled with anti-human IgG kappa-chain monoclonal antibody, such as sold by Kirkegaard Perry, Inc. The affinity column can be washed with PBS and loaded with the concentrated culture medium. The medium will then pass through the column, followed by PBS washings until no protein is detected in the eluate. The column is then washed with pre-elution buffer, for example, 10 mM phosphate at pH 8.0 and eluted from the column with 100 mN glycine at pH 2.4.

The protein peak fractions are detected by standard means such as by for example Bradford protein assay (Biolab) and pooled together and dialyzed against 0.2 M NaCl.

In a preferred embodiment you would adapt the cell for growth in a serum-free medium. This can be done by the skilled artisan. For example, one can readily adapt COS and CHO cells. In doing this, relatively r'' WO 95122618 PCTIUS95102140 19 pure Fab-fusion proteins are secreted into the medium.

Thus, the purification procedure is simpler.

The purified fusion protein is now ready to be combined with the desired nucleic acid sequence such as one for a positive potentiator (such as a gene for a cytokine, a gene for a missing or defective protein, etc.) or a sequence for a negative potentiator (such as a toxin, an anti-sense RNA, a suicide gene such as HSV thymidiac kinase, a ribozyme, a dominant-negative mutant, etc.). For example, when the nucleic acid encodes a toxin, one preferably takes care to alter the toxin gene to minimize its potential to affect nontargeted cells. This can be done by standard techniques such as deleting those sequences encoding recognition domains. Toxins are well known and include diphtheria toxin and truncated versions thereof, pseudomonas exotoxin, and truncated versions thereof, Ricin/abrin, Blocked ricin/abrin, Ricin ToxinA-chain, ribosome inactivating protein, etc. All these proteins have different domains. For example, the gene encoding PEA has several domains: Domain I is responsible for cell recognition; Domain II for translocation of the toxin cross-membrane and Domain III for adenosine diphosphate (ADP)-ribosylation of elongation factor 2, which is the step actually responsible for cell death.

[Gary, et al., Proc. Natl. Acad. Sci. USA 81:2645-2649 (1984); Allured, et al., Proc. Natl.

Acad. Sci. USA 83:13220-1324 (1986); Siegall, et al., J. Biol. Chem. 264:14256-14261 (1989)].

Accordingly, by alterations in Domain I or Domain II, S- that render those domains incapable of expression, for example, by a frameshift mutation, insertion of termination sequences, or deletions one can minimize the ability of the toxin to affect neighboring cells.

Thereafter, the skilled artisan can use standard techniques to insure that the other domains, or .t

I-

.I WO 95/22618 PCT/US9502140 portions of domains where expression is desired, are used.

For example, as indicated above, with PEA only Domain III is absolutely required. However, we have found that including partial sequences from other domains makes the toxin more effective. For example, we prepared two PEA mammalian expression vectors. This is one in which Domain III (mature PEA amino acid residues 405 to 613) only, referred to as pCMV-PEA III is expressed and one which encodes Domain III and partial Domain IB, a sequence of amino acids 385 to 613 (pCMV-PEAIbIII) is expressed. These sequences should be operably linked to a promoter which will permit expression in the target cell. For example, mammalian promoters such as CMV, SRa, RSV, SV40, MMLV LTR, HIV-1 are preferred. More preferably, CMV, HIV-1 RSV, and SV40. The toxin proteins encoded by these gene fragments lack a recognition domain. They are non-toxic to surrounding cells and are only toxic I when expressed inside a cell. These expression vectors can readily be tested to determine how well they express a product intracellularly by a simple in vitro assay. For example, the expression of those DNA sequences encoding PEA toxin fragments can be tested by transforming a cell with the delivery system and observing the cytotoxicity of the cell. We have found that the pCMV-PEIbIII vector shows a higher level of ADP-ribosylation than the vector expression only Domain III and thus, we prefer using it.

Figure 2 is a schematic representation showing the use of a nucleic acid delivery system according to the present system, wherein the nucleic acid sequence is a toxin expressor DNA.

x^ IP--P- 4 I ~C WO 95122618 PCrTUS95102140 ~i 21 In some instances, even with immunotoxins, resistant mutants can develop. In such instances, one can readily insert a different toxin gene or different types of nucleic acid segments into the nucleic acid cassette which is attached to the fusion protein.

Thus, the present system permits the production and use Sof a wide range of DNA and RNA segments.

SIn some preferred embodiments one would administer a cocktail of nucleic acid delivery systems where the targeting moiety may be changed to broaden the number of targeted cells or alternatively the nucleic acid segment that is delivered is changed to widen the spectrum of products delivered to the target cell.

When the protein is a toxin, transient expression in the cell is all that is needed. However, when it is desired to stably transform a cell, the gene is placed into a cassette containing LTRs or ITRs at either side to foster stable integration. Alternatively the cassette can be an episomal vector such as one that contains an Epstein Barr virus for example, pEBV His A, B, and C, pREP4, pREP7, pREP10, which are sold commercially by Invitrogen Corporation.

The recombinant fusion proteins are combined with the nucleic acid segment by standard techniques. For example, the fusion protein can be mixed with given amounts of the desired nucleic acid sequence, either DNA or RNA, by known means such as mixing in solution.

For example, in 0.2 H NaC solution. The DNA or RNA is Sreadily bound by the protamine. Thereafter, the Scarrier can be administered to the desired cell either for somatic cell therapy or used in vivo.

The delivery system can be delivered by any of a number of means. For example, it can be administered C~bSC~BI ~1 IPI WO 95122618 PCTIUS95/02140 22 by parenteral injection (intramuscular intraperitoneal intravenous or subcutaneous oral or other routes of administration well known in the art. Parenteral administration is preferred.

,The amount used will typically be in the range of about 0.1 mg to about 10 mg/kg of body weight. The delivery system will preferably be formulated in a unit dosage form based upon the nucleic acid or nucleic acids being delivered.

For example, solid dose forms that can be used for oral administration include capsules, tablets, pills, powders and granules. In such solid dose forms, the active ingredient, targeting moiety, is mixed with at least one inert carrier such as sucrose, lactose or starch. Such dose forms can also comprise additional substances other than inert diluents, e.g., lubricating agents, such as magnesium stearate.

Furthermore, the dose forms in the case of capsules, 4

I

tablets and pills may also comprise buffering agents.

The tablets, capsules and pills can also contain timerelease coatings.

For parenteral administration, one typically includes sterile aqueous or non-aqueous solutions, suspensions or emulsions in association with a pharmaceutically acceptable parenteral vehicle.

Examples of non-aqueous solvents or vehicles are propylene glycol, polyethylene glycol, vegetable oils such as olive oil and corn oil, gelatin and injectable organic esters, such as ethyl oleate. These dose forms may also contain adjuvants such as preserving, wetting, emulsifying and dispersing agents. They may be sterilized by, for example, filtration through a bacterial-retaining filter, by incorporating -7 r WO 95122618 2PCTUS95iO2140 23 sterilizing agents into the composition, by irradiating the compositions, etc., so long as care is taken not to inactivate the antibody. They can also be manufactured in a medium of sterile water or some other sterile injectable medium before use. Further examples of these vehicles include saline, Ringer's solution, dextrose solution and 5% human serum albumin.

Liposomes may also be used as carriers. Additives, such as substances that enhance isotonicity and chemical stability, buffers and preservatives, may also be used.

The preferred range of active ingredient in such vehicles is in concentrations of about 1 mg/ml to about xg/ml. More preferably, about 3 mg/ml to about mg/ml.

Although receptor mediated gene delivery in certain embodiments may be relatively inefficient, by utilizing the present gene delivery system, one does not have to worry about antigenic reactions occurring from the use of higher dosages or repeated injections.

This is because the targeting moiety and the DNA binding moiety can be designed so that they are either from the animal that one is injecting, such as a human, or made to be like that animal, i.e. using a humanized S murine antibody or binding protein for a human.

DNA

itself is weakly or non-immunogenic. Thus, the entire agent is either non or weakly immunogenic. Since the delivery system can be efficiently produced and adapted to have high binding activity, it can be used S repeatedly.

Additionally, as discussed above there are methods j that can be utilized in the present system to improve the efficiency of the delivery system. For example, one can include targeting sequences such as nuclear

MOON

WO 9522618 WO 9522618 PC/US95/0 2 14 0 24 targeting sequences associated with the nucleic acid segment to more efficiently deliver the nucleic acid to its desired target. Targeting sequences are known in the art and include for example, the nuclear localization signal on HIV gag p17 between positions and 33 (SEQ ID NO:1) (KKKYYKLK). Thus, one can include a targeting sequence, preferably inside a liposome, as part of the fusion moiety to more effectively target the DNA.

The present invention is further illustrated by the following examples. These examples are provided to aid in the understanding of the invention.and are not to be construed as a limitation thereof.

EXAMPLES

A bi-cistronic mammalian cell expression vector which contains a chimeric gene encoding the Fd of F105 fused to the human protamine protein in one expression cassette and the F105 kappa chain encoding gene in another expression cassette i.

(Figure 2) was constructed as follows.

Construction of Mammalian Expression Vector For Protamine Fusion Protein The pCMV-Fabl05 plasmid was constructed as described below. This plasmid contains bi-cistronic expression cassettes for the Fd gene and kappa chain gene derived from F105 hybridoma. To construct a fusion protein expression vector, the Fd fragment of F105 without a stop codon was amplified by PCR using the pCMV-Fabl05 as a template. The upstream primer (SEQ ID NO:2)

-TTTGAATTCAAGCTTACCATGGACATCTGTGGTTC-

corresponding to the leader sequence of human immunoglobulin of amino acids 1 to 6 with an additional HindIII cloning site (Kabat, et al., 1987), and the downstream primer (SEQ ID NO:3) W'O 95n2618 PCIUS9S/02140 3 corresponding to the amino acids of 226 to -233 of human heavy chain constant region with an additional XbaI cloning site were used for pcR amplification. The PCR reaction was performed as described previously [Marasco, et al., J.

of C2in. Invest. 90:1467-1478 (1992) The human protamifle gene was amplified from the plasmid pTZl9R-HPl (Krawatz, et al., Genomics 5:639-645 (1989) To delete an intron in the protamifle gene in this clone (for expression in prokaryatiC system as well), the first PCR amplification was performed using the upstream primer P1, (SEQ ID NO:4) 1 GTICGATICAGAGCAGAAAG which corresponds to the sequence of amino acids 1 to 6 of protamifle protein with an additional XbaI cloning site, and the downstream primer (SEQ ID TAGTCTACACkCCTGCCCT CCGTGTCTGGCAGC3r) which corresponds to amino acids 29 to 37 with additional sequence complementary to amino acids 38 to 40 in the second exon. The second PCR reaction was performed using the upstream primer P1 and the downstream primer, (SEQ ID NO:6) -TTAATTGCGGCcQCTAGTGTT~cAC- ATCTCGGTTGTACC GGG TGAjCA CTCTTCCCGTGTCTG 3 corresponding to amino acid sequence 31 to 40 with the sequence of amino acids 41 to stop codon in the second exon and an additional Notl cloning site. The f irst PCR-amplif ied DNA was used as a template.

To construct a bi-cistronic fusion protein expression vector, the PCR-amplified Pd of F105 without a stop codon was cut with HindIIIfXbal and purified from an agaroSe gel. The PCR-amplified entire protaine coding gene without intron was out with XbaI/NotI and purified from an agarose gel. The WO 95122618 PCTIUS95/02140 26 plasmid was cut with HindIIl/NotI and the Dt;A fragment about 7.0 kd was purified from an agarose gel. The HindIIIfXbaI-cut Fd fragment and XbaI/NotI-cut pratamifle fragment were then cloned into the HindIII/NotI sites of pCKV-FablO5 by threepiece ligation. The resultant expression vector, designated as pcHV-FablO5-Protamile, contains Pdprotamirie fusion gene (in-frame) and kappa chain gene under the control of independent CMV promoter. This construct was confirmed by DNA sequencing.

The F105 Fd and human protamine DNA fragments which were cloned into the pCXV-FablO5 vector are shown in Figure 2. The resulting bi-cistronic expression vector (pCKV-FablO5-Protamiffe) contaiis an expression cassette for the FdlO5-Protamfine fusion protein and another cassette for the kappa chain of F105.

A transformed mammalian cell line COS-Fab1OS- Protamine. was generated after DNA transfection and G418 selection....

'Construction of Transformed Cell Lines To generate transformed cell lines, cos-I cells were grown on 6-well plates and transfec-ted with j pxv-pabloS5Protmfle, using lipofectin as described previously [Chen, et al., J7. VizoJ. 65:5902- 5909 (1991)3. The transf acted cells were incubated in DM supplemented with 10% FCS for two days and replaced with selection medium (DNM with 10% FcS and B00 gf/ml G418 (EEL). The G418 resistant colonies appeared after two to three weeks of selection. The colonies were subaloned with limited dilutioun and examined by radiolabelling and immunoprecipitation, E=ISA, and immunflfuoresceflt staining for expression of recombinant proteins as described [Marasco,

WA.,

WO 95/22618 PCT/US95102140 27 et al., Proc. Natl. Acad. Sci. USA 90:7889 (1993)].

The Fd-protamine protein and kappa chain of F105 are expressed and secreted into the culture medium of cells as detected by radiolabelling and immunoprecipitation with antihuman IgG antibody See Figure 3.

Purification of Fusion Proteins The transformed COS cells were grown in flasks with DMEM medium supplemented with 10% fetal calf serum (FCS) and 500 gg/ml of neomycin. After reaching confluence, the cell cultures were replaced with fresh DMEM without FCS every three days for two weeks. The collected culture medium was clarified by centrifugation at 5000 rpm for 20 minutes at 40 C, and then concentrated using an Amico concentrator with membrane filter molecular weight cut-off 10,000 i dalton. The concentrated medium was then loaded onto an affinity column coupled with anti-human IgG kappa chain monoclonal antibodies (Kirkegaard Perry Inc.).

Preparation of the affinity column was made by mixing 2 mg of the purified monoclonal antibody with 1 ml of wet beads of protein A-sepharose CL-4B (Pharmacia Inc. Uppsala, Sweden) as described.

Briefly, protein-A-sepharose 4B beads were washed with PBS and then mixed with purified antibodies in PBS at 4° C overnight. The mixture was washed with volumes of 0.2 M sodium borate (pH 9.0) and added with dimethypimelimidate to a final concentration of mM. The mixture was stirred for 30 minutes at room temperature on a rocker, washed once with 0.2 M ethanolamine (pH 8.0) and then incubated for 2 hours Sat room temperature in 0.2 M ethanolamine on a rocker. After final washing, the beads coupled with l WO 95/22618 PCTIUS95102140 28 antibodies were resuspended in PBS with 0.01% merthiolate.

The affinity column was washed with PBS, and loaded with the concentrated culture medium. The medium passed through the column followed with PBS washing until no protein was detected from the elute.

The column was washed with pre-elution buffer (10 mM phosphate, pH 8.0) and eluted from the column with pH 2.4, 100 mN glycine. The protein peak fractions were detected by Bradford protein assay (Bio-Lab) and pooled together and dialyzed against 0.2 M NaCl. The DNA-binding portion of the fusion protein was examined by incubation of the DNA-cellulose with the culture medium of radiolabeled cells.

The transformed cell line was generated with G418 (Gibco-BRL) selection after transfection with pCMV-Fabl05-Protamine DNA [Warrant, et al., Nature 271:130-135 (1978)]. The COS- 4 +f Fabl05 cell line was established as described previously [Warrant, Nature 271, supra]. To examine expressed proteins, the transformed cells were radiolabeled for 4 hours and precipitated with anti-human IgG (Southern Biotech) and protein A- Sepharose 4B beads or with DNA-cellulose (Pharmacia) and analyzed by SDS-PAGE as described previously [Chen, et al., J. Virol. 65:5902-5909 (1991)].

To purify secreted Fabl05-Protamine in the serum-free medium, the culture medium of COS-Fabl05-Protamine cells is clarified, concentrated, and loaded onto an affinity column of Protein-A-Sepharose 4B beads coupled with anti-human IgG monoclonal antibody, which was prepared according to described methods [Winter, at al., Nature 349:293-299 (1992)]. The bound-proteins on the colimn were eluted by 100 mM glycine (pH and then concentrated and dialyzed WO09SJ2618 PCr/US95IO2140 29 against 0.20 H NaCi solution. For ELISA: microtiter plates were coated with recombinant gp120 (American Biotechnology Inc.) and incubated with known concentration of FablO5 or FablO5-Protamifle proteins followed by incubation with anti-human IgG conjugated with aflkalinie phosphatase (Sigma) (Warrant, at al., Nature 271, supral.

Figure 3 shows radiolabeling and immunoprecipitation of the expressed fusion proteins.

The transformed cell line (CSFbO-rotamine) was generated as discussed above. The cells on-6-well plates were continuously radiolabeled with 31S_ aysteine for 4 hours and the culture medium of the cells was precipitated with either anti-human IgG antibody (Southern Biotech) followed by Sepharoseprotein-A or with DNA-cellulose (Pharmacia). The samples were analyzed by SDS-PAGE under reducing conditions. Lane 1, COS-FablO5- Protamine precipitated with anti-human IgG; lane 2, Cos-vector precipitated with anti-human IgG and DNA-cellulose; lane 3, CoS-Pab15-Protamine precipitated with DNAcellulose; lane 4 and 5, COS-FablG5 precipitated with DNA-cellulose or with anti-human IgG The DNA-cellulose coprecipitated the Pdprotamiiie fusion proteins and Kappa chain, but not the FablO5 fragment, suggesting that the DNA-binding portion of the Fd-pretlmile. fusion protein maintains its DNA binding ability and the fusion proteins are associated together. The binding activity against HIV gpl2O, approximately 0.1 pig/El/24 hours, was detected in the culture medium of protamine cells by enzyme-linked immunosorbant assay (ELISA), while no binding activity was observed in the medium of vector-transformed cells.

-I*

WO 95122618 PCT/US95/02 14 0 The secreted recombinant fusion proteins were purified from serum-free culture medium by using an affinity-column coupled with anti-human IgG kappa chain monoclonal antibody (Figure The fusion proteins bound to the column were eluted by 100 mM glycine (pH concentrated and analyzed by SDS- PAGE under nonreducing or reducing conditions following. As shown in Figure 2, under the reducing condition, two protein bands, corresponding to Fdprotamine fusion protein and kappa chain appeared on the gel. While under non-reducing conditions, the majority of the proteins shifted to a higher molecular weight band, which likely represents..

assembled Fab fragments. The specific binding activity of the purified Fabl05-Protamine to a although slightly lower than that of Fabl05, was -detected by ELISA.

Figure 4 shows the purification and SDS-PAGE -i analysis of the recombinant fusion -protein.

The Fabl05-Protamine fusion proteins in the culture medium were purified by an affinity-column coupled with anti-human IgG kappa chain monoclonal antibodies (Kirkegaard and Perry Lab) as described.: The proteins bound to the column were eluted by 100 mM glycine (pH 2.4) and then concentrated and dialzyed against 0.20 M NaCl solution. The purified proteins were analyzed by SDS-PAGE under reducing or nonreducing conditions following Coomassie blue Sstaining.

Lane a, 100 ng (left), Lane a' 10 ng (right) of purified FablOS-Protamine under the reducing conditions; Lane b 100 ng of purified FablOS- Protamine under non-reducing conditions. ~i WO 95122618 PCTIfUS9SJO2I4O Bindimg activity to Gpl2O of the purified fusion protein is shown in Figure ZLSA plates coated with recombinant HIV-1 gP120 (American Biotechnology, Inc.) were incubated with FabIOS or FablOS-Protamifle proteins followed by antihuman IgG conjugated with alkaline phosphatase (Sigma) The binding activity to gpl2O was detected SI at OD0 after incubation with substrate (Bio-Lab) The data shown are the mean values from duplicate determination. Lane a, 10 rig/mi. of FabIO5 or Fab105protamiie; Lane b, I ng/ml; Lane c, 0.1 rig/mi and Lane 3, 0.01 ng/ml. The first column in each lane is Fab-105, while the second column is Fab-105- Protamine.

These results indicate that the fusion proteins, which are assembled and secreted into the culture medium, have specific binding activity to HIV-l gp12O.

The DNA binding activity of FablO 5-Protamine was examined by a gel-shift assay [Wagner, et al., PNAS USA:89:6099- 6 1

O

3 (1992)]3.

DNA Binding Assay Gel-shift assay was used to analyze the DNA binding activity of the recombinant fusion proteins.

The increased amounts of purified fusion proteins in 0.2 N -Radl solution were mixed with a given amount of DNA either radio labeled or unlabeled in 0. 2 N NaCi solution. DNA radiolabelling with MP-dATP (Amrao'ham) was performed using a nick translation kit (Promega).

The protein-DNA mixtures was allowed to stand at room T. temperature f or 30 minutes and filtered through 0.*45 ux pore-size membrane to eliminate DNA-protein precipitates, and then loaded onto 1.0% agarose gel for electrophoresis at IX TAE buffer. To analyze WO 9512618 PCTIUS95O2140 32 cytotoxicity of the DNA-toxin expressor, the fusion protein-DNA mixtures were dialyzed against the normal saline solution at 4* C overnight before adding to cell cultures.

The DNA binding activity of protein is shown in Figure 6.

DNA-binding ability of FablO5-Protamine was examined by gel mobility-shift assay [Wui, et al., J. Biol. Chem. 262:4429-4432 (1987)]. The HindIII/XbaI-cut DNA fragments of Protamine was radiolabeled with 32P-dATP using a nicktranslation kit (Pharmacia).- 20 ng of labeled DNA f or each sample was incubated with increased amount of Fablo5-Protamine proteins in 0.20 R NaCl solution.

DNA were incubated with FablOS proteins as control.

The whole plasmid DNA pCKV-FablS-Protamin-e jug each sample) were also incubated with increased amount of rCKV-FablO5-Protamine proteins in 0.20 N NaCI (See, Figure The samples were analyzed by electrophoresis on 0.8% agarose gels. For autoradiography, the gel was dried, and exposed on Xray film. Figure 6: lane 1, DNA (5 ng) only; lanes 2 to 4, DNA (5 rig) with 0. 5 ng Fab-Protamine with 1.0 ng Fab-Protamine with 10 ng Fab- Protamina lane 5, DNA (5 ng)/lO ng FablO5 as control. Figure 7, lane 1, DNA (0.2 jig) only; lane 2, DNA 2 Ag) /2 -0 ;g FablO5 control; lanes 3 to 6, DNA 2 jig) with 0. 1 gg Fab-Protamine with 0. 2 pzg Fab-Protamine with 0 .4 pLg Fab-Protamine with 0.6 pg Fab-Protamine lane 7, FabI0S- Protaiine only (0.6 jg); and lane 8, DNA (0.2 ;ig) with 0.6 g of FAblO5-Protaminelphenol extract before loading onto the gel.

A

V

WO095122618 PCTIUS95102140 33 As shown in Figures 6 and 7, when increasing amounts of the fusion proteins-were mixed with the radiolabeled DNA fragments or whole plasmid DNA, the decreasing amounts of DNA fragments or whole plasmid DNA migrated into agarose gels and the DNA entered the agarose gels migrated slower, while the DNA incubated with FablO5 proteins shoved nio significant change of its mobility in the agarose gels. The binding activity of the fusion proteins to gp12a on the cell surface after coupling with DNA was further examined by fluorescent activated cell sorting

(FACS).

Binding ability of complexes to GP120 on the cell surface is shown in Figure 8.

The Ely-infected or mock-infected Jurkat cells were incubated with Feb10 5 or FablOS-Protamine protein-DNA complexes followed by anti-human IgG Fab (pastan, et al., Science 254:1173 (1992)] conjugated with Fito. The fluorescent staining on the cell surface was then analyzed by FACS.

The DNA mobility-shift assay was performed as described EWU, et al., J. B101. Chem. 262:4429 (1987); Wagner, et al., Pzoc. NatI. Acad. Scl.

USA 89:6099 (1992). Increased amounts of purified fusion proteins were mixed with given amounts of DNA in the 0.2 K NaCi solution. The mixtures standed at room temperature for 30 minutes and then filtered through 0.45 pZ membrane (Millipore) before loaded onto 0.8% agaros~e gels for electrophoresis. To detect the binding ability of fusion protein-DNA complexes to gpl2O on the cell surface, 1 gtg of purified FablOS-Protamine was mixed with 0.5 S gc of pCIKV-Fab105 plasmid DNA in 100 IL1 of 0.2 KN adl for 2 140 :34 minutes, and the mitrs wr iue o 12 0-9k~ H RaCi Solution and incubae lted ith 1:20 a inlfected or urninfected Jurkat aells followed by antihuman IgG-Fita conjugates. Fablo5 fragments Were Used as a control. The fluorescent staining was then analyzed by

FACS.

-As shown in Figure 8, the EIVl1-infected cells reacted~ with either FabIO 5 or the NA complexes showed pgsitive staining~r while finfected cells incubated with the Complexes showed negative staining. The infected cells directly incubated With conjugated antibody also showed negative staining (not shown). Thus, the Protamine fusionl proteins Maintain the binding act vit to gpi o a ter coupling with DNA molecules- The encoding gene of PEA was Selected to contuc mammalian toxin expression vectors due to the accumlaed knowledge Of the encoding gene seqluefcefncti relation [Gary, G.L, et al., Proc.

Natl., Acad. 5cd UrSA, 8-1 suPra; Allured, et al., Prc Nfatl. Acad. Scd. USA 8.3, szzpa; Siegal,,

C.B.,

et al-, J. Biol. Chem. 264, supra], PEA has several functional. domains: Domain I is responsible for cell recognition: domain II, for translOcation of the toxin cross membrane; and domain 111t for adenosine diPhosphate (ADP)--ribosylation of elongatioli factor 2F the step actually responsible for Cell det Callured, Proc. Natl. Acad. sci USA 8,s a -Siegall et T. B1ol. Chem- 264, supra] Tw PEA mamla expression vectors were designed and constuce in Which~ the domain 13:1 (Mature

PEA

amino acid residues of 405 to 613) only (pCMVpaj 1 1 or domain III and partial domain mb sequence (amino acids of 385 to 613) Only

(PQKV-PM~.I

1 jj was placed under the control Of CXV and T7 promoter.

WO 9512618 PCTIUS95IO214O Construction of Toxcin Exressiori Vectors A plasmid pJH8 containing the PEA encoding gene was obtained from the American Type Culture Collection, 12301 Parklawn Drive, Rockville, IMD 20852, having ATICC Depost No. 67208. See, Figure 9 for a schematic showing the PEA encoding domains.

The DNA sequences encoding the PEA catalytic fragment were obtained by PCR amplification using pJH8 DNA as a template. To construct the toxin expressor designated as pCHV-PEIII, an upstream primer (P1, (SEQ ID WC: 7) 5 'TTTAAGCTTATGGGCGACGTCAGCTTCAGCACC-3') containing an additional HindIII site and an initial codon followed by sequences compliment-y to the amino acids 405 to 411 of mature PEA and a downstream primer (SEQ ID NO:8) f-TmTCTAATACTTCACGTCCTCCGG-3') containing the sequence -ccm-p-imentary to amino acids 609 to stop codon of PEP. followed by an additional XbaI site were used to amplify domain III of PEA. To construct the toxin expressor pCPIV-PET-bIII, an upstream primer (SEQ ID HO:9) (P3-5' -TTTAAGCTTATGGCCGACGTGGTGAGCCTG,-3'), corresponding to amino acids 365 to 372 of PEA, and the downstream primer P-2 were used to amplify the partial domain Ib and domain III. The amplified DNA fragments were purified with Geneclean kits (Bio 101 Inc.), digested with HindIIIJXbaI and cloned into the pRc/ CKV expression vector (Invitrogen) under control of CHV promoter. The resulting constructs were confirmed by DNA sequencing.

These constructs ensure that any expressed toxin fragments without the recognition domain are nontoxic j to surrounding cells unless they are expressed inside a cell. To detect the toxin fragments expressed from the vectors, the plasmids pCK-PEAIII or pCMVpEAMbflI (See Figure 9) were first transformed into BL23. (DE3) expression bacterial hosts (Novagen), WO 9522618PCIUS95/0 2140 36 hihinducibly express T7 DNA polymerase f or transcription of the giene under the control. Of T7 promoter. ADPribosylation activity was detected from the tranSfOrmedi bacteria after induction. (not Shown) When the toxi~n expressors were transfeCted into mammalian cells (COS-1 and Heta) Using lipofectin [Chen, et al., LT. V.Lr1. 65:5902..

5909 (1991) 1, toxin~ fragjments were Produced and cYtotoxicity to the transfected cells was observed (not shown). The PC7V-EIbIll Vector which soe high r l vel cti ityof ADPribosylation than pCMV- PEIII was used for further e xperiments To investigate whether Fablo5-.poine ca function as a gene carrier to tase h oi expressor into target cells, the Purified protam-ine fusion proteinsweeicbtdwhpC PEAIbIII Plasmid DNA at 2:3 ratio (determined by titration) in 0.2 K NaCi so1utjo ofrmfso Pro ein D co ple es (See Figures 1-0-12). H IV-1-.

2.nfcte Jukatly'mphocytes which were shown over 95?6 Positive by imunofluOrescent staining with the antibody against gPI20 were used as target cells.

The target cells were incubated wit the Fabios- PrOtamIine..tOxjn expressor complexes, toxin~ expressors only or FabI05-Protamine proteins alone. Normal lymphocyt~es were also incubated with these molecules as Control. After 48 hours of iciaincl viability (Figure 10), protein Synthesis (Figure 11), and ADIP-ribasylation activity (Figure 12) in the.

culture cells were examined The selective cytotoxicity of toxin Expressor Complees to HIV-Infected Cells are shown in Figures 10-12.

-J1 WO 95122618 PCTIUS95J02 140 37 jurkat cells were inf ected with HIV-l virus and *at day 10 postinfection the surface gpl20 expression of the cells were examined by immnunof luorescent staining. The cells with over 95% gpl2O positive X 106) were incubated with the DNA-f usion protein complexes, or fusion protein only, or DNA only at 370 C for 48 hours.

Figure 10 shows cell viability in the culture examined by Trypan Blue Staining.

Percentage of viable calls were calculated from duplicate determination.

Figure 11 shows a protein inhibition assay, the cells 5 X 10") w.-ere replaced with leucine-f ree medium (See, Allured, et al., Proc. Nati. Acad.

Scj. USA 63:1320 (1986); Pastan, et al., Cell 47:643 (1986); Siegall, et al., J. Biol. Chem.

264:14256 (1989) and added with 4 ALci I'--'eucine for 4 hours. The cells were centrifuged at 3,000 rpm for minutes and lysed for scintillation counting.

Figure 12 shows detection of ADP-ribosylatiol activity in the culture cells [Collier, et al., J. Blo.l. Chem. 246:1496 (1971)J]. The cells (l X 106) were pelleted by centrifugation and lysed in 4 14 urea solution. The supernatants of the lysates were then suibj ected to tue AfP-ribasylatiofl assay and PEA proteins (Gibco-BEL) were used for positive control.

The mean scintillation counts of the samples are V 7 shown as calculated from duplicate determination.

In Figures 10-12 lane a: normal Jurkat cells incubated with Fab-Protamhife-toxifl expressor complexes (10 Ig FablO5-Protamiflel5 jLg expressor

DNA,

10:5 Ag); lanes b to a: HIV-infected Jurkat cells WO 95/226 18 PCTIUS95102140 38 incubated with Fab-Protaanine only with toxin expressor DNA only with FablOS-Protamine-ToXin expressor complexes (10:5 ig) with FablOS- Protamine-toxin expressor complexes (5:2.5 Ag) lane f:AfP-ribisylation activity of 1.0 ng of ureadenatured PEA control.

As shown in Figure 10-12, after incubation with FablOS-Protamine-toXin expressor complexes for 48 hours, the HIV-3.-ixxfected cells showed a significant decrease of cell viability and protein synthesis ability while the cells-incubated with toxin expresser or Fablos-protamine alone only showed slightly decreased cell viability and protein synthesis ability. In addition, uninfected lymphocytes showed no significant decreases of cell:.

viability and protein synthesis ability after incubation with the FablOS-Protamine-toxin expressor complexes. The observed selective cytotoxicity to HIV-infected cells should be a result of ADPribosylation activity from expressed toxin fragments since ADP-ribsylation activity, roughly equal to ng of PEA proteins, was detected from lysates of the I ElV-inf ected cells (I X 106) incubated with the complexes. Thus, the Fab105-Protamnine-toxin I expresser complexes selectively intoxicate HIV-l f infected cells in tissue culture.

A major obstacle of imnitaxins as efficacious agents in the treatment of human cancer and otherdiseases is the host antibody response to xenogeneic antibodies and toxin molecules. (yers, et al., Immunol. 65:329 (1988) Durrant, et al., C21n.

Exp. Immuno2. 75:258 (1989); Pai, et al., J.

Cln. Oncol. 9:2095-2103 (1991)]. The utilization of humanized murine or human antibodies may solve the problem of targeting moiety (Rybak, et al., WO 95J22618 PCT/US95/02140 39 Proc. Katl. Acad. Sci. USA 89:3165-3169 (1992)], but for highly immunogenic toxin moiety, the problem still remains. In this study we demonstrate that the Fabl05-Protamine fusion proteins can serve as a gene carrier to deliver toxin-expressor plasmid DNAs into HIV-l-infected cells by receptor-mediated endocytosis, resulting in selective killing of the target cells. The extreme potency of the toxin molecules compensates for the low efficiency of the receptor-mediated gene delivery [Wu, et al., J.

Biol. Chem. 262:4429-4432 (1987); Wagner, et al., PNAS USA 89:6099-6103 (1992)] to efficiently achieve the therapeutic goal. Since that antibody molecules or ligands (targeting moiety) and DNA-binding moiety of bifunctional fusion proteins can be of human origin, and the toxin expressor DNAs are very weakly or non immunogenic, the whole protein-toxin expressor complexes will be weakly immunogenic. Therefore, these complexes should be able to be repeatedly administered into patients without development of significant antibody response. Furthermore, the bifunctional recombinant fusion proteins as a gene carrier also have the advantage over chemically linked ones (Wu, et al., J. Biol. Chem.

262:4429-4432 (1987); Wagner, et al., PNAS USA 89:6099-6103 (1992)], such as efficient production, I and potentially better binding activity. In summary, C this gene therapy form of immunotoxins, termed herein "stealth immunotoxins" has significant advantages o over currently described immunotoxins for treatment Sof cancers, and other diseases. Moreover, the anti- Fabl05-Protamine-toxin expressor complexes have Sselective toxicity to HIV-1-infected cells which also represent a novel therapeutic agent for AIDS treatment.

BC-i-xt r- r Al; references described herein are incorporated herein by reference.

Where the terms "comprise", "comprises", "comprised" or "comprising" are used in this specification, they are to be interpreted as specifying the presence of the stated features, integers, steps or components referred to, but not to preclude the presence or addition of one or more other feature, integer, step, component or group thereof.

This is a Divisional Application of Australian Patent Application No. 19251195 and the disclosures thereof are incorporated herein by way of reference.

j

I

Tr WO 95122618 PCTIUS95O2140 41 SEQUENCE LISTING (31) GENERAL INFORMATION: APPLICANT: MARASCO, WAYNE A CHMN, SI-YI (ii) TITLE OF INVENTION: NUCLEIC ACID DELIVERY SYSTEM, METHOD OF SYNTHES IS AND USES THEREOF (Iii) NURSER OF SEQUENCES- 9 (iv) CORESPoNDEmCE

ADDRESS:

ADDRESSEE: DAVID G. CONLIN; DIKE, BRONSTEIN, ROBERTS

CUSHMAN

STREET: 130 WATER STREET CC) CITY: BOSTON STATE: MASSACHUSETTS COUNTRY: US ZIP: 02109 COMPUTER READABLE

FORM:

MEDIUM TypE:. Floppy disk COMPUTER: IB14 PC compatible OPERATING SYSTEM: PC-DOS/MS-00S SOFTWA.RE: PatentIn Release 11.0, Version~ 11.25 (vi) CURRENT APPLICATION

DATA:

APPLICATION NUMBER: us 08/199,070 FILING DATE: 02-FEB-1994 Cc) CLASSIFICATION: (viii.) ATTORNEY/AGENT

INFORMATION:

NAME: CONLIN, DAVID G REGISTRATION NUMBER: 27026 REFERENCE/DOCKET NUMBER: 43471 (iX) TELECOMMUNICATION

INFORMATION:

TELEPHONE: (617) 523-3400 TELEFAX: (617) 523-6440 j(C) TELEX: 200291 STRE UR INFORM4ATION FOR SEQ ID NO:l: SEQUENCE

CHARACTERISTICS:

LENGTH: 8 amino acids TYPE: amino acid STRANDEDNESS: unlknown TOPOLOGY: unknlown (XL) SEQUENCE DESCRIPTION: SEQ ID NO:1; Lys Ly. '.ys Tyr Tyr Lys Lau Lys INFORHATION FOR SEQ ID NO:2: f SEQUENCE

CHARACTERISTICS:

-7 tZ WO 95122618 PCTUS9SI2 140 42 LENGTH: 36 base pairs TYPE: nucleic acid STRANDEDNESS: unxknowrn TOPOLOGY: unknown (xi) SEQUENCE DESCRIPTION: SEQ ID NOz2: TTTGAAT-ACA AGCTTACCAT GGAACP.TCTG TGGTTC 3 INFORMATION FOR SEC ID NO:3: SEQUENCE CHARACTERISTICS: LENGTH: 33 base pairs TYPE: nucleic acid STRANDEDN2ESS: unknown TOPOLOGY: unknown (xi) SEQUENCE DESCRIPTION: SEQ 10 NO:3: GGTACOGAAT TCTCTAGAAC. AAGATTTGGG CTC 3 INFORMATION FOR SEQ ID NO:4: SEQUENCE CHARACTERISTICS: LENGTH: 36 base pairs TYPE: nucleic acidS.- STRANDEDNESS: unknown TOPOLOGY: unknown (xL) SEQUENCE DESCRIPTION: SEQ ID NO:4: GGTACGAAT TCTCTAGAAT GGCCAGGTAC AGATGC 3 INFORMATION FOR SEQ ID SEQUENCE CHARACTERISTICS: LENGTH: 48 base pairs TYPE: nucleic acid (C TAIDES nnw TOPODLOGY: unknown

-J,

(xi) SEQUENCE DESCRIPTION: SEQ ID 1(0:5: TTTAGGATCC TTAACAACAC CTCATGTC TCCTCCGTGT CTGGCAGC 48 INFORMATION FOR SEQ ID NO:6: SEQUENCE CHARACTERISTICS: LENGTH: 78 base pairs TYPE: nucleic acid STRAND EDNESS: unknown TOPOLOGY: unknown

T'

WO 95122618 PC1'1US95102140 43 (xi) SEQUEMCE DESCRIPTION: SEQ ID NO:6: '-T;ATTGrG ccrcTTAGTG TC-- CTACAT CTCGGTCTGT PACCTGGCT GACACCTCAT GCTTCCTC CGTG--TT 78 INFOR1MATION FOR SEQ ID O7 (ii SEQUENCE

CHA.RACTERISTICS:

LENGTH: 33 base pairs TYPE: nucleic acid STRANDEDNESS: unknownl TOPOLOGY: unknown (XI) SEQUENCE DESCRIPTION: SEQ ID NO:7: TTT,%ArCTTA TGaGcG.ACGT CAGCTTCAGC ACCO3 INFoR.-(ATrON FOR SEQ ID NO:8: SEQUENCE

CHARACTERISTICS:

LENGTH: 27 base pairs TYPE: nucleic acid STR.ANDEDUESS: unknown TOPOLOGY: unknown (xi) SEQUENCE DESCRIPTION: SEQ ID 110:8: TTT-CTAGAT TAC-7ICAGGT CCTCVOG 27 INFO RKATION FOR SEQ ID NO;: SEQUENCE

CHARACTERISTICS:

LENGTH: 30 base pairs TYPE: nucleic acid STP.ANDEDNESS2- unknown TOPOLOGY: unknown (xil SEQUENCE DESCRIPTION: SEQ ID NO-.92 TTAKGCTrTA TGG=cGT GGTGAGCCTG

IV

Claims (8)

1. A nucleic acid delivery system comprising: a fusion protein containing a targeting moiety, which will specifically bind to a site on a target cell, and a binding moiety which will bind to a nucleic acid segment, and the nucleic acid segment containing a nucleic acid sequence of interest.

2. The nucleic acid delivery system of claim 1, wherein the targeting moiety is an antibody.

3. The nucleic acid delivery system of claim 2, wherein the antibody is an antibody to a viral envelope protein, a cellular receptor, or an extracellular domain of an activated receptor.

4. The nucleic acid delivery system of claim 2, wherein the antibody is a single chain antibody, a Fab portion of an antibody or a segment. The nucleic acid delivery system of claim 1, wherein the binding moiety is a protein or the nucleic acid binding domain of a protein, and the binding moiety is fused to the carboxy portion of the targeting moiety. S_6. The nucleic acid delivery system of claim wherein the binding moiety is the nucleic acid binding domain of a protein selected from the group of nucleic acid binding domains present in proteins selected from the group consisting of GCN4, Fos, Jun, TFIIS, FMRI, yeast protein EX, Vigillin, Merl, TFS Fa, yeast P WO 9SM2618 PCTIUS9iOZI4O *bacteri al polynucleotide phosphaylase, ribosomal protein S3, and heat shock protein.

7. The nucleic acid delivery system of claim wherein the binding moiety is the protein protamine. a. The nucleic acid delivery system of claim 1, wherein the nucleic acid secuence of interest encodes an antibody, a dominant negative mutant, an antisense RNA, ribozymes, Or a cytotaxc agent.

9. The nucleic acid delivery system of claim 1, _herein the nucleic. acid segment comprises flanking 3' long termzinal repeat (LT-R) regions or inverted terminal repeat regions, a promoter operblylined to a desired gene in the nucleic acid sequence of interest. IL nucleic acid delivery system comprising a fusion protein wherein one portion of the fusion protein comprises an antibody, which will selectively bind to a desired site on a cell, and the Other portion of the fusion protein comprises a protamine protein capable of binding to a nucleic acid segment; and the nucleic acid segment. i1. The nucleic acid delivery system of claim wherein the nucleic acid segment is a DNA sequence corresponding to a cytotaoxin gene or a fragment I thereof which will encode a cytotoxic protein. 1j2. Thea nucleic acid delivery system of claim. 21, wherein the nucleic acid segment encodes at least Domain II1 of PseudamcnflS eXOtorlfl A~.

13.* A method of tzmnsformiflg a target cell which comprises adding an effective amount of the nucleic WO 9S/22618 PCTITJS9JO2140 46 aci;d delivery sy-stem Of claim 1 to a medium conraining t-he target call-, and waiting until the K nucleic acid sequence of the nuciecic acid delivery system tansforms the call.

14. A, method of preparing a nucleic acid delivery system which comprises ts'ansforming11 a cell with a vector containing a DNA segment which encodes the fusion protein Of Claim 1 operably linked to a promoter. incubating the cell, and collecting the expressed fusion Protein. A method of use of a nucleic acid delivery stem which comrises administering an eff ectiv I amount of the nucleic acid deliver sys-tem of claim 1 to serum containing a target cell, and waiting until- the nucleic acid delivery1 system contacts the target cell. 1.Amethod of use of a nucleic acid delivery system which compri.ses administering an effective amount of the nucleic acid delivery system of claim 3.(3 to serum containing a target cell, and waiting until the nucleic acid delivery system contacts the target cell. DATED this 25th day of February 1999 DANA-FARBER CANCER INSMIUTE By their patent attorneys CALLINAN LAWRIE