INTRODUCTION Technical Field

The present invention relates to methods and compositions for systemic introduction of exogenous genetic material into mammalian, particularly human, cells in vivo .

Background

An ever-expanding array of genes for which abnormal expression is associated with life-threatening human diseases is being cloned and identified. The ability to express such cloned genes in humans will ultimately permit the prevention and/or cure of many important human diseases, diseases which now either are treated poorly or are untreatable by currently available therapies. As an example, in vivo expression of cholesterol-regulating genes, genes which selectively block the replication of HIV, or of tumor-suppressing genes in human patients should dramatically improve treatment of heart disease, HIV, and cancer, respectively. However, currently available gene delivery strategies have been unable to produce either a high level of or generalized transgene expression in vivo in _. a wide variety of tissues after systemic administration to a mammalian host. Thiε inability has precluded the development of effective gene therapy for most human diseases. Approaches to gene therapy include both different goals and different means of achieving those goals. The goals generally include gene replacement, gene correction and gene augmentation. In gene replacement, a mutant

gene sequence is specifically removed from the genome and replaced with a normal, functional gene. In gene correction, a mutant gene sequence is corrected without any additional changes in the target genome. In gene augmentation, the expression of mutant genes in defective cells is modified by introducing foreign normal genetic sequences.

The means to reach the above goals used by others, have included "ex vivo" transfection of a target cell followed by introduction of the transformed cells into a suitable organ in the host mammal. Ex vivo techniques include transfection of cells in vitro with either naked DNA or DNA liposome conjugates, followed by introduction into a host organ ("ex vivo" gene therapy) . The criteria for a suitable target organ or tissue include that the target organ or tissue is easily accessible, that it can be manipulated in vitro , that it is susceptible to genetic modification methods and ideally, it should contain either non-replicating cells or cycling stem cells to perpetuate a genetic correction. Further, it should be possible to reimplant the genetically modified cells into the organism in a functional and stable form. Exemplary of a target organ which meets these criteria is the mammalian bone marrow. A further criterion for ex vivo gene therapy, if for example a retroviral vector is used, is that the cells be pre- itotic; post-mitotic cells are refractory to infection with retroviral vectors. Although this has not been reported, in some instances it may be possible to transfect cells from other than the target organ or tissue using ex vivo gene therapy if the corrective gene product can be secreted and exert the desired effect on/in the target cell following circulation in blood or other body fluids.

There are several drawbacks to ex vivo therapy; for example, if only differentiated, replicating cells

are infected, the newly introduced gene function will be lost as those cells mature and die. Ex vivo approaches also can be used to transfect only a limited number of cells and cannot be used to transfect cells which are not removed first from the body. The above methods generally involve integration of new genetic material into the cell genome and thus constitute permanent changes.

Liposomes have been used effectively, particularly to introduce drugs, radiotherapeutic agents, enzymes, viruses, transcription factors and other cellular effectors into a variety of cultured cell lines and animals. The agent to be introduced is typically entrapped within the liposome, or lipid vesicle, or the agent may be bound to the outside of the vesicle. Successful clinical trials examining the effectiveness of liposome-mediated drug delivery have been completed. Several strategies have been devised to increase the effectiveness of liposome-mediated drug delivery by targeting liposomes to specific tissues and specific cell types. However, while the basic methodology for using liposome-mediated vectors is well developed, the technique has not been perfected for liposome-based transfection vectors for in vivo gene therapy.

In vivo expression of transgenes as reported by others has been restricted to injection of transgenes directly into a specific tissue, such as direct intratracheal, intramuscular or intraarterial injection of naked DNA or of DNA-cationic liposome complexes, or to ex vivo transfection of host cells, with subsequent reinfusion. The expression is low and generally has been limited to one tissue, typically the tissue that was .. injected (for example muscle) ; liver or lung where iv injection has been used; or lung where intratracheal injection has been used, and less than 1% of all cells within these tissues were transfected. In some cases, .

transfection of cells has been obtained in tissues afferent to the site of intravenous administration. Currently available gene delivery strategies consistently have failed to produce a high level and/or generalized transgene expression in vivo . It therefore would be of interest to develop compositions and delivery methods for in vivo gene therapy that provide for a high level of transcription of the transgene and/or expression in a variety of cell and tissue types for the in vivo treatment, prevention, or palliation of numerous human diseases. Also of interest, is the use of gene modulation as an alternate means of gene therapy. In gene modulation, expression of an already expressed gene is increased by introducing exogenous normal genetic sequences and decreased by introducing antisense genes or gene fragments, or by introducing vectors that can produce ribozymes that can cleave specific mRNAs. Gene modulation can also be achieved by the introduction of exogenous normal genetic sequences that code for proteins that modulate the extent of gene expression, or affect the processing, assembly or secretion of gene products.

Relevant Literature

A large number of publications relate to in vivo and ex vivo transfection of mammals. In some cases, only transcription of a transgene has been achieved, in others, the data appear to show only a low level of expression and/or expression in a limited number of tissues or cell types. The following are examples of the publications in this area. A variety of approaches for introducing functional new genetic material into cells, both in vitro and in vivo have been attempted (Friedmann, (1989) Science , 244:1275-1281) . These approaches include integration of the gene to be expressed into modified retroviruses

(Friedmann, (1989) supra; Rosenberg, (1991) Cancer Research, 5JL(18) , suppl. : 5074S-5079S) ; integration into non-retrovirus vectors (Rosenfeld, et al., (1992) Cell, 68.143-155; Rosenfeld, et al . , (1991) Science, 252:431- 434) ; or delivery of a transgene linked to a heterologous promoter-enhancer element via liposomes (Friedmann, (1989), supra; Brigham, et al . , (1989) Am. J. Me . Sci . , .298:278-281; Nabel, et al., (1990) Science, 249:1285- 1288; Hazinski, et al. , (1991) Am. J. Resp. Cell Molec. Biol., 4 _. :206-209; and Wang and Huang, (1987) Proc. Natl . Acad. Sci. (USA) , 84 _. :7851-7855) ; coupled to ligand- specific, cation-based transport systems (Wu and Wu, (1988) J. Biol. Che . , 263:14621-14624) or the use of naked DNA expression vectors (Nabel et al., (1990), supra); Wolff et al., (1990) Science, 247:1465-1468..

Direct injection of transgenes into tissue produces only localized expression (Rosenfeld, (1992) supra) ; Rosenfeld et al., (1991) supra; Brigham et al., (1989) supra; Nabel, (1990) supra; and Hazinski et al., (1991) supra). The Brigham et al. group (Am. J. Med. Sci., (1989)

298:278-281 and Clinical Research, (1991) 39 (abstract)) have reported in vivo transfection only of lungs of mice following either intravenous or intratracheal administration of a DNA liposome complex. An example of a review article of human gene therapy procedures is: Anderson, (1992) Science 256:808-813.

PCT/US90/01515 (Feigner et al.) is directed to methods for delivering a gene coding for a pharmaceutical or immunogenic polypeptide to the interior of a cell of a vertebrate in vivo. Expression of the transgenes is limited to the tissue of injection. PCT/US90/05993 (Brigham) is directed to a method for obtaining expression of a transgene in mammalian lung cells following either iv or intratracheal injection of an expression construct. PCT 89/02469 and PCT 90/06997 are

directed to ex vivo gene therapy, which is limited to expressing a transgene in cells that can be taken out of the body such as lymphocytes. PCT 89/12109 is likewise directed to ex vivo gene therapy. PCT 90/12878 is directed to an enhancer which provides a high level of expression both in transformed cell lines and in transgenic mice using ex vivo transfection. PCT/US92/08806 is directed to particle-mediated transformation of mammalian unattached cells. EP application 91301819.8 is directed to the use of recombinant techniques to produce cystic fibrosis transmembrane conductance regulator (CFTR) .

SUMMARY OF THE INVENTION Methods and compositions are provided for introduction of a transgene into a plurality of mammalian tissues in vivo . The method includes the step of incorporating a transfection cassette comprising a nucleotide sequence of interest into a largely non- integrating plasmid and introducing the plasmid into a mammalian host, other than by directly introducing it into a specific tissue. The transfection cassette comprises as operably joined components, a transcriptional and initiation regulatory region, a nucleic acid sequence of interest, and a transcriptional termination regulatory region, wherein said regulatory regions are functional in the cells of a mammalian host, ^• and wherein said nucleic acid sequence of interest is free of introns. Optionally, the transcriptional cassette is complexed with a cationic lipid carrier. The method finds use to modulate the phenotype of mammalian cells.

DESCRIPTION OF THE DRAWINGS Figure 1 shows photomicrographs of sections of mouse lung (A) from a mouse injected with a transfection cassette, or (B) a control mouse. The photomicrograph in (A) shows a section of mouse lung 48 hours following iv injection of pZN27:DDAB:Cholesterol expression vector- cationic lipid carrier complexes. The plasmid contained sequences for the expression of CAT (chloramphenicol acyl transferase) . The lipid carrier composition was l to l molar DDAB:Cholesterol. The carrier: plasmid ratio was 5 nmoles cationic lipid to 1 μg DNA. A dose of 100 μg DNA was injected per mouse. This field shows alveoli and alveolar lining cells, the majority (50-70%) of which stain red positively for the presence of CAT protein when probed with anti-CAT antibody and visualized using alkaline phosphatase (indicated by arrows) . The treated animals' lungs stain uniformly with diffuse involvement of alveolar and vascular endothelial cells. Airway epithelial staining is also seen indicating airway cells are also transfected. The CAT protein normally is not present in mammalian cells and therefore the presence of CAT protein in these cells indicates that they have been transfected in vivo . The photomicrograph in (B) shows a section of mouse lung from a control animal treated with iv-injected lipid carrier only, and probed with anti-CAT antibody. Cells do not show significant staining, although low-level background staining is detectable in some alveolar macrophages (indicated by arrows) , which possess endogenous alkaline phosphatase activity. Figure 2A shows a phase contrast micrograph of mouse T lymphocytes isolated 48 hours after intraperitoneal (ip) injection of DNA-cationic lipid carrier complexes. Lipid carriers were DDAB:DOPE, 1 to 1 molar. DNA-cationic lipid carrier complexes were 1 mg pCIS-CAT with 1 μmole cationic lipid. .1 mg DNA was

injected ip. Cells were incubated with anti CAT mouse monoclonal antibody, followed by Texas red-conjugated goat anti mouse IgG. Figure 2B shows a fluorescence micrograph of the same field showing that essentially all T lymphocytes present removed from the treated animal were stained red by im unofluorescence demonstrating expression of the transfected CAT gene following ip injection of DNA/cationic lipid carrier complexes. Figure 2C shows a phase contrast micrograph of T lymphocytes isolated from an uninjected control mouse. Figure 2D shows a fluorescence micrograph of the same field, showing that there is no CAT gene expression in these lymphocytes from control animals.

Figure 3A shows hematopoietic bone marrow-derived cells from a mouse treated with a DNA: lipid carrier complex DDAB:Cholesterol:pZN51 intravenously 48 hours earlier. Cells were stained for CAT protein as in Figure 1. The photomicrograph shows that approximately 70 percent of the bone marrow-derived cells, including primitive or blast cells are stained red (blast cell indicated by arrow) and are transfected in vivo . Lipid carriers were 1:1 DDAB:Cholesterol. DNA:cationic lipid ratio was 1 μg of DNA to 5 nmoles cationic lipid. 100 μg of DNA was injected iv into each mouse. Figure 3B shows isolated bone marrow cells from an untreated control mouse. The photomicrograph shows no red staining indicating no CAT gene expression in untreated animals.

Figure 4 shows transfection of human T lymphocytes (CD4 ⁺ ) by CAT expression plasmid-DNA complexes in culture. The lipid carrier composition was 1:1 molar ratio DDAB:DOPE small uni-la ellar vesicles (SUV) . 25 μg of pZN27 complexed with 50 nmoles cationic lipid was added to 10 million cells in culture. Cells were incubated with anti CAT mouse monoclonal antibody, followed by Texas red-conjugated goat anti mouse IgG. In

Figure 4A is shown a phase contrast micrograph of freshly isolated human CD4 ⁺ T lymphocytes 48 hours after transfection. Figure 4B shows a fluorescence micrograph of the same field which demonstrates ^' that essentially 100% of the T lymphocytes stain positively for CAT protein, i.e., they have been transfected.

Figure 5 shows construction of plasmid pZN20. Figure 6 shows a restriction map of the immediate early enhancer and promoter region of HCMV (Towne) in Figure 6A and HCMV(AD169) in Figure 6C. Figure 6B shows a sequence comparison of the two HCMV promoters. The position of the Ncol site is indicated by an asterisk.

Figure 7 shows electron micrographs which demonstrates that cationic lipid carrier: DNA complexes (DOTMA:DOPE:pRSV-CAT) are internalized by CV-1 (African Green Monkey Kidney) cells via classical receptor- mediated endocytosis following binding to cell surface receptors. For construction of pRSVCAT see Gorman et al . , (1982) Proc . Nat . Acad . Sci . USA 79_: 6777-6781. Lipid carriers were 1:1 DOTMA:DOPE. 20 μg DNA were complexed with 20 nmoles cationic lipid, and added to 2 x 10 ⁵ cells. The arrows in panel (a) show particles binding to clathrin-coated pits; panel (b) , the particles that have been ingested and are present in endosomes. Figure 8 shows photomicrographs of a histochemical analysis of lungs from B-16 melanoma-bearing animals which received an intravenous injection of CMV-CAT cationic lipid carrier:DNA complexes. Lipid carriers were DOTAP:Cholesterol, 1 to 1 molar ratio. Cationic lipid: DNA ratio was 6 nanomoles: 1 μg DNA. The plasmid used was pZN20 (see Figure 5) . 100 μg DNA was injected per mouse. The immunohistochemical analysis for CAT protein revealed intense red staining of many focal parenchymal tumors (Figure 8A indicated by arrows) and tumor emboli within blood vessels (Figure 8B indicated by

arrows) , indicating that large numbers of B16 melanoma tumor cells in the lung, as well as blood-borne metastases, are transfected after iv injection. Figure 8C shows that in tissue from B16 melanoma-bearing mice which did not receive an injection of DNA-lipid carrier complexes, no CAT protein is present in the surrounding normal lung or in any of the lung tumor cells.

Figure 9 shows the construction of plasmid pZN27, encoding CAT. Figure 10A and Figure 10B show the construction of plasmid pZN46, encoding human IL-2.

Figure 11 shows the construction of plasmid pZN32, encoding human CFTR.

Figure 12 shows photomicrographs of immunohistochemically stained (for CFTR protein) frozen sections of lung tissue from mice treated by intravenous administration of either pZN32:cationic lipid carrier complexes or lipid carrier alone (control mice) . Figures 12A, 12C, and 12E are lung sections from mice treated with DNA lipid carrier complexes at 50X, 100X, and 250X magnification, respectively. Figures 12B and 12D are lung sections from control mice at 50X and 100X magnification show that there is no detectable CFTR expression in lungs of control mice. Lipid carriers were 1 to 1 molar DDAB:Cholesterol (SUV) . Lipid carrier-DNA complexes were 5 nanomoles cationic lipid to 1 μg DNA. 100 μg DNA was injected per mouse. Animals were sacrificed 24 hrs after injection. The figures show that >70% of both airways and airway lining cells stain intensely red, indicating that they are expressing the CFTR gene.

Figure 13 shows the construction of plasmid pZN51, encoding CAT.

Figure 14 shows the construction of plasmids pZN60, pZN61, pZN62 and pZN63, all encoding CAT. Figures

14A and B show the construction of intermediate plasmids pZN52, pZN54, pZN56 and pZN58. Figure 14C shows the construction of the final plasmids, pZN60 through pZN63, from the intermediates. Figure 15 shows an autoradiograph of the thin layer chromatograph of the a CAT assay for six different plasmids injected intravenously in mice. Lanes 1-12 show the CAT activity in lung tissue; Lanes 13-24 show the CAT activity in liver tissue. Lanes 1, 2, 13, 14-pZN51; lanes 3, 4, 15, 16-pZN60; lanes 5, 6, 17, 18-pZN61; lanes 7, 8, 19, 20-pZN62; lanes 9, 10, 21, 22-pZN63; lanes 11, 12, 23, 24-pZN27. Lipid carriers were DDAB:Cholesterol (1:1) . Lipid carriers-DNA complexes were 5 nmoles cationic lipid to lμg DNA. lOOμg DNA was injected per mouse. Animals were sacrificed after 48 hrs. Each lane represents a single mouse. Chromatograph runs from bottom to top of Figure as shown. As shown by the brackets in Figure 15, unreacted ¹⁴ C chloramphenicol only migrates a short distance from the origin of the TLC plate. The ¹ C Acetylated chloramphenicol species migrate which result from CAT enzymatic activity further up the plate as shown by the brackets.

Figure 16 (A-E) shows photomicrographs of frozen lung section that have been immunohistochemically stained for CAT protein. Mice were injected iv by tail vein with pZN27 (16A) which shows positive red staining for CAT is endothelial, alveolar and airway cells within the lung. In contrast, liposome complexes of the CAT gene linked to the natural CFTR promoter (pBE3.8CAT) showed CAT expression primarily in airway epithelial cells (16B) . A lung section from an uninjected control animal did not show red staining, indicating no CAT expression in untransfected cells (16C) . Figure 16D shows a high magnification photomicrograph of alveoli from a pZN27 treated mouse, and shows a high level of positive red

staining for CAT gene product in both alveolar and epithelial cells. Figure 16E shows a high magnification photomicrograph of alveoli from a pBE3.8CAT injected mouse, and shows no significant CAT gene expression in either alveolar or endothelial cells. One hundred μg of plasmid was injected iv complexed with DDAB:Cholesterol as SUVs. Animals were sacrificed after 24 hrs.

Figure 16(F-K) shows autoradiographs of the thin layer chromatographs of CAT activity in heart (16F), lymph nodes (16G) spleen (16H) , kidney (161) , lung (16J) , and liver (16K) in tissue from uninjected mice (lanes 1- 3), mice injected IV with pBE3.8CAT (Chou, et al . , (1991) J. Biol . Chem . , 266:24471-24476. (lanes 4-6) , or mice injected with pCIS-CAT (lanes 7-9) . Lipid carriers were DDAB:Cholesterol, 1:1 SUV (1 μg DNA to 5 nmoles cationic lipid; 100 μg injected per mouse.

Figure 17 shows the construction of plasmid pZN84, encoding murine GM-CSF.

Figure 18 shows the construction of pZN102, encoding human CFTR.

Figure 19A shows an autoradiograph of a TLC plate showing CAT activity in animals after injection of pZN27 without any lipid carrier. One milligram was injected iv twice over a 4-hour period. Mice were sacrificed 24 hrs. later. The various tissues examined are indicated under each lane.

Figure 19B shows an autoradiograph of a TLC plate showing CAT activity in an animal injected with pZN27 complexed with a lipid carrier. Lipid carrier was DDAB:Cholesterol 1 to 1 molar, 5 μg of plasmid complexed to 1 n ole of lipid. 100 μg was injected iv, and animals were sacrificed 24 hrs later. The various tissues ^■ examined are indicated under each lane.

Figure 20 shows an autoradiograph of a TLC plate of injection into the right (R) ventricle of pCIS-CAT

complexed to DOTMA:DOPE (1:1) lipid carriers in the indicated ratios of plasmid:lipid. Animals were sacrificed 48 hrs. after the stereotactic injection of 2.5 μg. Each pair of lanes represents a single animal. The left (L) ventricle was not injected and therefore transfection of the left ventricles demonstrates that the entire brain is being transfected.

Figure 21 shows human lung cancer cell line transfections. An autoradiograph of a TLC plate shows CAT activity in extracts from three different human lung cancer cell lines transfected using various cationic liposomes. The lung cancer cell lines NCI-H69, NCI-H82, and NCI-H520 were plated at 2 x 10 ⁶ cells/plate and transfected in suspension. Individual plates of cells were transfected with 5 μg of RSV-CAT (RC) alone or 5 μg of RSV-CAT complexed to a total of 10 or 20 n ole of pure DOTMA (D) , pure L-PE (L) , or L-PE/CEBA (L/CA) liposomes. The cells were harvested 48 h after transfection and assayed for CAT activity. Figure 22 shows serum murine GM-CSF levels in a single goat (A) injected iv with l mg of GM-CSF plasmid (pZN84) complexed with DDAB:Cholesterol liposome (1 μg plasmid to 1 nmole lipid) . Serum samples were collected at 0, 12, 24, 48, 72, 168, and 840 hrs. after injection. Murine GM-CSF levels were measured by ELISA using a commercial kit (Endogen) . Figure 23 show the serum murine GM-CSF levels in two additional goats (B and C) injected with pZN84 as per goat A.

DESCRIPTION OF THE SPECIFIC EMBODIMENTS In accordance with the subject invention, nucleic .. acid constructs together with methods of preparation and use are provided which provide for in vivo change of and/or modulation of phenotype of cells in a plurality of tissues of a mammalian host, following introduction of _.

the constructs into the host at a dose sufficient to cause transfection of tissues and cells contacted by the nucleic acid. The components of the transfection vector generally will include as operably linked components in the direction of transcription, a transcriptional initiation region, a DNA sequence of interest and a transcriptional termination region, wherein the transcriptional regulatory regions are functional in the host mammal's cells that are targeted for transfection. Optionally, an intron may be included in the construct, preferably 5' to the coding sequence. Generally, the construct does not become integrated into the host cell genome and is introduced into the host as part of a non- integrating plasmid where it is maintained primarily in an extrachromosomal or episomal form. The constructs can be either naked nucleic acid or nucleic acid associated with a lipid carrier. By a sufficient dose is meant that which will result in a desired effect, for example prevention, palliation, and/or cure of an animal or human disease, or modulation of endogenous levels of an agent of interest ( "in vivo" gene therapy) . The modulation may be generalized, i.e. obtained in a multiplicity of cell types or tissues, or the modulation may be selective, for example inducible and/or in only selected cell or tissue types.

Transfection of multiple tissues and cells other than those solely at or afferent to the site of introduction of the nucleic acid constructs into the host is obtained and expression is at a high level and in a large number of cells and cell types. The tissues which can be transformed include the lungs, heart, liver, bone marrow, spleen, lymph nodes, kidneys, thy us, skeletal muscle, ovary, uterus, stomach, small intestine, colon, pancreas, and brain in normal animals, as well as metastatic tumors and intravascular tumor emboli in

tumor-bearing mammals. Particular cells which are transfected include macrophages, alveolar type I and type II cells, hepatocytes, airway epithelial cells, vascular endothelial cells, cardiac myocytes, myeloblasts, erythroblasts, B-lymphocytes and T-lymphocytes. The route of administration typically is into a circulating bodily fluid such as blood or cerebrospinal fluid, but other routes of administration also can be used. The constructs can be either naked nucleic acid or nucleic acid associated with a lipid carrier. Optionally, the lipid carrier molecule and/or construct may provide for targeting and/or expression in a particular cell type or types.

The nucleic acid constructs can be prepared from nucleic acid sequences which are synthetic, naturally derived or a combination thereof. The sequences can be obtained from a donor host homologous with an intended recipient host, or depending upon the nature of the nucleic acid sequence, it may be desirable to synthesize the sequence with host preferred codons. The host preferred codons may be determined from the codons of highest frequency in the proteins expressed in the largest amount in the particular host of interest. Where cloning is done in a bacterial host system, it may be beneficial to change codon choice to promote stability in the bacterial host system. Where the intended use is to treat disease resulting from an infectious organism, an appropriate source of sequence may be viral nucleic acid, either RNA or DNA. Genes are high molecular weight, polyanionic molecules, for which carrier-mediated delivery is usually but not always required for DNA transfection of cells either in culture or in vivo . Cationic lipid carriers such as liposomes can be used to deliver transcriptional regulatory proteins. Further, liposomes themselves

(unlike viral vectors) appear to be non-immunogenic in vivo . Liposome formulations, including those containing a cationic lipid, have been shown to be safe and well tolerated in human patients (Treat et al . , (1990) J . Natl . Cancer Instit . 81:1706-1710). Although a wide variety of transfection techniques can produce high level expression of transgenes in cultured cells, only a few such methods, including those using liposomes, are compatible with in vivo gene delivery. Previous attempts at using either cationic liposomes or naked DNA to obtain in vivo transfection have resulted in transfection of a single tissue and/or transfection only of the tissue of introduction of exogenous DNA. This pattern of transfection suggests that rather than the DNA and/or liposomes being taken up by normal cellular mechanisms, cells in the tissue of introduction are damaged by the introduction method and are then able to take up some of the exogenous DNA which has been introduced.

Applicants surprisingly discovered changes which provide for both transfection of a wide variety of tissues and cell types and a high-level of transgene expression after systemic administration into a mammalian host. These changes include the use of DNA:cationic lipid carrier complexes, wherein the ratio of DNA to cationic lipid of the lipid carrier can significantly affect obtention of in vivo expression; use of a higher dose of complexes than has been used previously; use of an appropriate promoter element, for example one which is both strong and constitutively active, such as the HCMV- IE1 element where it is desired to provide enhanced expression in a wide variety of cell types in vivo ; and placement of greater than 100 bp of an intrύn 5' to the coding region of the transgene or removal of the intron altogether to facilitate production of a desired gene product. Additionally, it has surprisingly been

discovered that a number of tissues, including the lung, liver, spleen, kidney, lymph nodes and heart can be transformed following direct administration of high doses of naked DNA into a circulating bodily fluid, such as blood or cerebral spinal fluid. Alternatively, selective expression can be obtained in specific cell and tissue types and at a desired time (i.e. expression is inducible) by alterations in the components, particularly promoters and/or enhancers of the constructs used, or by the use of targeting moieties on the liposomes.

Most gene therapy strategies have relied on transgene insertion into retroviral or DNA virus vectors. Potential disadvantages of retroviruses, as compared to the use of naked DNA or the use of cationic lipid carriers, include the limited ability of retroviruses to mediate in vivo (as opposed to ex vivo) transgene expression; the inability of retrovirus vectors to transfect non-dividing cells; possible recombination events in replication-defective retrovirus vectors, resulting in infectious retroviruses; possible activation of oncogenes or inhibition of tumor suppressor genes due to the random insertion of the transgene into host cell genomic DNA; size limitations: less than 15 kb of DNA typically can be packaged in a retrovirus vector; and potential immunogenicity, leading to a host immune response against the vector. In addition, all ex vivo approaches require that the cells to be transfected be removed from the body and that they be maintained in culture for a period of time. While in culture, the cells may undergo deleterious or potentially dangerous phenotypic and/or genotypic changes. Adenoviral and other DNA viral vectors share several of the above potential limitations. Thus, the subject invention which does not use adenoviral or other DNA viral vectors has several advantages over existing techniques.

Additionally, the subject invention offers ease of administration and results not achievable by other means.

The constructs for use in the invention include several forms, depending upon the intended use of the construct. Thus, the constructs include vectors, transcriptional cassettes, expression cassettes and plasmids. The transcriptional and translational initiation region (also sometimes referred to as a "promoter,"), preferably comprises a transcriptional initiation regulatory region and a translational initiation regulatory region of untranslated 5' sequences, "ribosome binding sites," responsible for binding mRNA to ribosomes and for translational initiation. It is preferred that all of the transcriptional and translational functional elements of the initiation control region are derived from or obtainable from the same gene. In some embodiments, the promoter will be modified by the addition of sequences, such as enhancers, or deletions of nonessential and/or undesired sequences. By "obtainable" is intended a promoter having a DNA sequence sufficiently similar to that of a native promoter to provide for the desired specificity of transcription of a DNA sequence of interest. It includes natural and synthetic sequences as well as sequences which may be a combination of synthetic and natural sequences.

For the transcriptional initiation region, or promoter element, any region may be used with the proviso that it provides the desired level of transcription of the DNA sequence of interest. The transcriptional initiation region may be native to or homologous to the host cell, and/or to the DNA sequence to be transcribed, or nonindigenous to the host cell or foreign or heterologous to the DNA sequence to be transcribed. By nonindigenous to the host cell, is intended that the

transcriptional initiation region is not normally found in the host into which the construct comprising the transcriptional initiation region is to be inserted. By foreign to the DNA sequence is intended a transcriptional initiation region that is not normally associated with the DNA sequence of interest. Efficient promoter elements for transcription initiation include the SV40 ^• (simian virus 40) early promoter, the RSV (Rous sarcoma virus) promoter, the Adenovirus major late promoter, and the human CMV (cytomegalovirus) immediate early 1 promoter.

Inducible promoters also find use with the subject invention where it is desired to control the timing of transcription or the tissue or cell type in which the nucleic acid of interest is transcribed. Examples of inducible promoters include those obtainable from a 3-interferon gene, a heat shock gene, a metallothionein gene or those obtainable from steroid hormone-responsive genes, including insect genes such as that encoding the ecdysone receptor. Such inducible promoters can be used to regulate transcription of the transgene by the use of external stimuli such as interferon or glucocorticoids. Since the arrangement of eukaryotic promoter elements is highly flexible, combinations of constitutive and inducible elements also can be used. Tandem arrays of two or more inducible promoter elements may increase the level of induction above baseline levels of transcription which can be achieved when compared to the level of induction above baseline which can be achieved with a single inducible element.

Generally, the transcriptional regulatory sequence comprises DNA up to about 1.5 kb 5' of the transcriptional start of a gene, but can be significantly smaller. As necessary, this regulatory sequence may be modified at the position corresponding to the first codon

of the desired protein by site-directed mutagenesis (Kunkel, (1985) Proc . Natl . Acad . Sci . (USA) , 82:488-492) or by introduction of a convenient linker oligonucleotide by ligation, if a suitable restriction site is found near the N-terminal codon. In the ideal embodiment, a coding sequence with a compatible restriction site may be ligated at the position corresponding to codon #1 of the gene. This substitution may be inserted in such a way that it completely replaces the native coding sequence so that the substituted sequence is flanked at its 3' end by the gene terminator and polyadenylation signal.

Transcriptional enhancer elements optionally may be included in the transcription or expression cassette. By "transcriptional enhancer elements" is intended DNA sequences which are primary regulators of transcriptional activity and which can act to increase transcription from a promoter element, and generally do not have to be in the 5' orientation with respect to the promoter in order to enhance transcriptional activity. The combination of promoter and enhancer element(s) used in a particular cassette can be selected by one skilled in the art to maximize specific effects. Different enhancer elements can be used to produce a desired level of transgene transcription and/or expression in a wide variety of tissue and cell types. For example, the human CMV immediate early promoter-enhancer element can be used to produce high level transgene transcription and expression in many different tissues in vivo .

Examples of other enhancer elements which confer a high level of transcription on linked genes in a number of different cell types from many species include enhancers from SV40 and RSV-LTR. The SV40 and RSV-LTR are essentially constitutive. They can be combined with other enhancers which have specific effects, or the specific enhancers may be ^' used alone. Thus*, where .

specific control of transcription is desired, efficient enhancer elements that are active only in a tissue-, developmental-, or cell-specific fashion include immunoglobulin, interleukin-2 (IL-2) and β-globin enhancers are of interest. Tissue-, developmental-, or cell-specific enhancers can be used to obtain transgene transcription and/or expression in particular cell types, such as B-lymphocytes and T-lymphocytes, as well as myeloid, or erythroid progenitor cells. Alternatively, a tissue-specific promoter such as that derived from the ■ human cystic fibrosis transmembrane conductance regulator (CFTR) gene can be fused to a very active, heterologous enhancer element, such as the SV40 enhancer, in order to confer both a high level of transcription and tissue- specific transgene transcription. In addition, the use of tissue-specific promoters, such as LCK, allow targeting of transgene transcription to T lymphocytes. Tissue specific transcription of the transgene can be important, particularly in cases where the results of transcription of the transgene in tissues other than the target tissue would be deleterious.

Tandem repeats of two or more enhancer elements or combinations of enhancer elements can significantly increase transgene expression when compared to the use of a single copy of an enhancer element; hence enhancer elements find use in the cassettes. The use of two different enhancer elements, which can be either from the same or from different sources flanking or within a single promoter can in some cases produce transgene expression in each tissue in which each individual enhancer acting alone would have an effect, thereby increasing the number of tissues in which transcription is obtained. In other cases, the presence of two different enhancer elements results in silencing of the enhancer effects. Evaluation of particular combinations

of enhancer elements for a particular desired effect or tissue of expression is within the level of skill in the art.

Although generally it is not necessary to include an intron in the transcription and/or expression cassette, an intron optionally can be included which comprises a 5' splice site (donor site) and a 3 ' splice site (acceptor site) separated by a sufficient intervening sequence to produce high level extended in vivo transcription and/or expression of a transgene. Generally, an intervening sequence of about lOObp produces the desired expression pattern and/or level, but the size of the sequence can be varied as needed to achieve a desired result. The optional intron placed 5' to the coding sequence results in high level extended in vivo expression of a transgene but generally is not necessary to obtain expression. Optimally, the 5' intron specifically lacks cryptic splice sites which result in aberrantly spliced mRNA sequences. If used, the intron splice donor and splice acceptor sites, arranged from 5' to 3' respectively, are placed between the transcription initiation site and the translational start codon. Alternatively, the intervening sequence can be placed 3' to the translational stop codon and the transcriptional terminator or inside the coding region. The intron can be a hybrid intron with an intervening sequence or an intron taken from a genomic coding sequence. An intron 3' to the coding region, particularly one of less than 100 bp, or any intron which contains cryptic splice sites can under certain condition substantially reduce the level of transgene expression produced in vivo . However, unexpectedly, a high level of in vivo expression of a transgene can be achieved using a vector that lacks an

intron. Such vectors therefore are of particular interest for in vivo transfection.

Downstream from and under control of the transcriptional and/or translational initiation regulatory regions is a multiple cloning site for insertion of a nucleic acid sequence of interest which will provide for one or more alterations of host phenotype, for example to dramatically improve the treatment of acquired immune deficiency syndrome (AIDS) , cystic fibrosis, cancer, heart disease, autoimmune diseases and a variety of life threatening infections. Conveniently, the multiple cloning site can be employed for a variety of nucleic acid sequences in an efficient manner. The nucleic acid sequence inserted in the cloning site can encode a ribozyme sequence or encode a polypeptide, for example, a protein having enzymatic activity, with the proviso that if the coding sequence encodes a polypeptide, it should lack cryptic splice sites which can block production of appropriate mRNA molecules and/or produce aberrantly spliced or abnormal mRNA molecules. The polypeptide can be one which is active intracellularly, a transmembrane protein, or it may be a secreted protein. It also can be a mutant protein, for example, which is normally secreted, but which has been altered to act intracellularly. The nucleic acid sequence can be DNA; it also can be a sequence complementary to a genomic sequence (an "antisense sequence") , where the genomic sequence may be one or more of an open reading frame, an intron, a non- coding leader sequence, or any other sequence where the complementary sequence will inhibit transcription, or messenger RNA processing, for example splicing, or translation.

For the treatment of AIDS, anti-TAT, REV, TAR or other critical anti-HIV sequences may be used,

particularly for expression of the appropriate coding sequences in T lymphocytes, macrophages and monocytes which can be achieved following iv administration of the appropriate coding sequences. Thus, DNA sequences which code for mRNA (an antisense or ribozyme sequence such as one to HIV-REV or a BCR-ABL sequence) or code for proteins such as transdominant negative mutants which specifically prevent the integration of HIV genes into the host cell genomic DNA, replication of HIV sequences, translation of HIV proteins, processing of HIV mRNA, or virus packaging in human cells can be used.

Expression of wild-type conductance regulator (CFTR) gene in the lungs of cystic fibrosis patients can be used in the treatment of cystic fibrosis (see Collins, (1992) Science 25^:774-783). CFTR cDNA can be obtained from Dr. Collins at University of Michigan or Dr. Tsui at Toronto Sick Children's Hospital.

Expression of wild-type p53 in tumors of cancer patients with absent or aberrant expression of this gene can be used as a means of treating these patients. p53 is obtainable from Dr. Vogelstein at John Hopkins Univ. Other methods of cancer treatment include transcription of antisense sequences to over-expressed, transforming oncogenes, such as myc or ras in tumors. Viral pneumonias remain a major cause of death and disability of the very young and in the aged population. Proteins such as granulocyte macrophage colony stimulating factor (GM-CSF) , which stimulates production of white blood cells from the bone marrow of immunocompromised patients can be useful in controlling the course of viral infections. In particular, GM-CSF delivered via gene transfer would serve with prophylactic modalities for those at risk for occurrence of viral pneumonia, as well as in therapeutic modalities for

treatment of the infection. GM-CSF also has significant anti-tumor activity.

Other examples of nucleic acid sequences of interest, include those encoding the LDL (low density lipoprotein) receptor, which specifically lowers serum cholesterol, and which can reduce the risk of heart attack in individuals with elevated serum cholesterol levels; sequences such as an antisense IL-1 receptor sequence to down-regulate the IL-1 receptor as a treatment of conditions involving inflammation, including sepsis, rheumatoid arthritis and asthma; downregulation of the gene coding for apo(a), circulating levels of which are correlated with the risk of coronary artery disease (CAD) ; and genes which block activity of activated T cell clones which attack myelin in multiple sclerosis or other targets in autoimmune diseases. A T-cell lymphocyte clone activated to recognize and attack myelin can be targeted by using an antisense sequence, ribozyme sequence or transgene coding for a transdominant negative mutant which specifically blocks surface expression on the T-cell of T-cell receptor components which mediate recognition and/or attack of myelin- sheathed cells. Other beneficial therapeutic nucleic acid sequences which can be expressed in appropriate cells in vivo using this invention include nucleic acid sequences encoding molecules which have superoxide dismutase activity or catalase activity to protect the lung from oxidant injury; endothelial prostaglandin synthase to produce prostacyclin and prostaglandin E2; and antiprotease alpha-1 antitrypsin, and erythropoietin. The termination region which is employed primarily will be one of convenience, since termination regions appear to be relatively interchangeable. The termination region may be native to the intended nucleic acid sequence of interest, or may be derived from another

source. Convenient termination regions are available and include the 3' end of a gene terminator and polyadenylation signal from the same gene from which the 5' regulatory region is obtained. Alternatively, a terminator and polyadenylation signal from a different gene or genes may be employed with similar results. Specific sequences which regulate post-transcriptional mRNA stability may optionally be included. For example, certain polyA sequences (Volloch and Housman, (1981) Cell 2.:509-514) and β-globin mRNA elements can increase mRNA stability, whereas certain AU-rich sequences in mRNA can decrease mRNA stability (Shyu et al . , (1989) Genes and Devel . 2:60-72). In addition, AU regions in 3' non- coding regions may be used to destabilize mRNA if a short half-life mRNA is desirable for a particular application. The construct may additionally include sequences for selection, such as a neomycin resistance gene or a dihydrofolate reductase gene and/or signal sequences to generate recombinant proteins that are targeted to different cellular compartments or secreted when the wild type sequence is not. Any of a variety of signal sequences may be used which are well-known to those skilled in the art. These signal sequences may allow generation of new vaccine strategies or produce soluble antagonists directed against specific cell surface receptors such as transformed oncogenes. The sequences for selection may be on a separate plasmid and cotransfected with the plasmid carrying the therapeutic nucleic acid. Where a carrier is used, the selection plasmid may be complexed to a different carrier or to the same carrier as the therapeutic plasmid.

In some cases, it may be desirable to use constructs that produce long term transgene effects in vivo , either by integration of the transgene into host cell genomic DNA at high levels or by persistence of the

transgene in the nucleus of cells in vivo in stable, episomal form. When desired, integration of the transgene into genomic DNA of host cells in vivo may be facilitated by administering the transgene in a linearized form (either the coding region alone, or the coding region together with 5' and 3' regulatory sequences, but without any plasmid sequences present) . It may be possible to further increase the incidence of transgene integration into genomic DNA by incorporating a purified retroviral enzyme, such as the HIV-1 integrase enzyme, into the lipid carrier-DNA complex. Appropriate flanking sequences are placed at the 5' and 3' ends of the transgene DNA. These flanking sequences have been shown to mediate integration of the HIV-l DNA into host cell genomic DNA in the presence of HIV-l integrase.

Alternatively, duration of transgene expression in vivo can be prolonged by the use of constructs that contain non-transforming sequences of a virus such as Epstein- Barr virus, and sequences such as oriP and EBNA-1, which appear to be sufficient to allow heterologous DNA to be replicated as a plasmid in mammalian cells (Burhans, et al . , (1986) Cell 62:955-965) .

The recombinant coding-sequence flanked at its 5' end by the promoter and regulatory sequences and at its 3' end by a terminator and regulatory sequences may be introduced into a suitable cloning plasmid (e.g., pUClδ, pSP72) for use in direct DNA uptake in host cells following introduction into the host. It is a theory of the invention that the naked nucleic acid when introduced into the host is protected from degradation by nucleases by associating with lipid carriers in the blood, such as chylomicrons. Thus, more effective transfection may be achieved when lipid in the blood is at an elevated level after ingestion of food.

The nucleic acid construct also may be complexed with a carrier such as a lipid carrier, particularly a cationic lipid carriers to form a complex. By a complex is intended an association between a nucleic acid construct such as a plasmid containing an expression cassette or a transcriptional cassette and a lipid mixture. The physical form of the complex may be a liposome with nucleic acid complexed to the outside or entrapped within the liposome, or the complex may be in the form of interleavened lipid and nucleic acid, or the complex may be a mixture of any or all of the above physical forms. For intravenous administration, generally the complex is prepared by sonicating the lipid mixture to be used and then mixing the sonicated mixture with the nucleic acid in an appropriate DNA:lipid ratio in a physiologically acceptable diluent immediately prior to use. The lipid carriers can be prepared from a variety of cationic lipids, including DOTAP, DOTMA, DDAB, L-PE, and the like. Lipid carrier mixtures containing a cationic lipid, such as N-[l-(2,3-dioleyloxy) propyl]- N,N,N-triethylammonium chloride (DOTMA) also known as "lipofectin", dimethyl dioctadecyl ammonium bromide (DDAB), l,2-dioleoyloxy-3-(trimethylammonio) propane (DOTAP) or L-lysinyl-phosphatidylethanolamine (L-PE) and a second lipid, such as dioleoylphosphatidylethanolamine (DOPE) or cholesterol (Choi) , are of particular interest. DOTMA synthesis is described in Feigner, et al . , (1987) Proc . Nat . Acad . Sciences, (USA) 81:7413-7417. DOTAP synthesis is described in Stamatatos, et al . , Biochemistry , (1988) .27:3917-3925. DOTMA:DOPE lipid carriers can be purchased from, for example, BRL. DOTAP:DOPE lipid carriers can be purchased from Boehringer Mannheim. Cholesterol and DDAB are commercially available from Sigma Corporation. DOPE is commercially available from Avanti Polar Lipids.

DDAB:DOPE can be purchased from Promega. Biodegradable cationic amphophiles also can be used (see, for example, copending PCT application attorney docket number MEBI- 002/00WO filed December 17, 1992). Lipid carriers such as cationic liposomes can mediate high level cellular expression of transgenes or mRNA by delivering the nucleic acid into a wide variety . of cells in culture. The use of specific cationic lipids can confer specific advantages for in vivo delivery of complexes. For example, iv injection of nucleic acid complexed to DOTAP-containing liposomes or ethyl- phosphatidylcholine (E-PC) lipid carriers can target transgene expression primarily to the lung. Furthermore, DOTAP, as well as L-PE and cholesterol ester 0-alanine (CEBA) are fully metabolized by cells, whereas DOTMA cannot be fully metabolized by cells. Therefore, DOTAP, E-PC, and L-PE, but not DOTMA, are suitable for repeated injection into mammalian hosts. Additionally, using a lipid carrier comprising a cationic lipid and a second lipid, particularly cholesterol or DOPE can maximize transgene expression in vivo . Also, mixing a steroid, such as cholesterol, instead of DOPE, with DOTAP, DOTMA, or DDAB, substantially increases transgene expression in vivo. Particular cells and tissues can be targeted, depending upon the route of administration and the site of administration. For example, a tissue which iε closest to the site of injection in the direction of blood flow can be transfected in the absence of any specific targeting. Additionally, if desired, the lipid carriers may be modified to direct the complexes to particular types of cells using site-directing molecules. Thus antibodies or ligands for particular receptors or other cell surface proteins may be employed, with a target cell associated with a particular surface protein.

For example, the AIDS virus is primarily directed to cells having the CD4 surface protein. By having anti-CD4 antibody bound to the surface of the lipid carrier, a nucleic acid lipid carrier complex can be directed primarily to T-helper cells.

A particular ligand or antibody can be conjugated to the lipid in accordance with conventional ways, either by conjugating the site-directing molecule to a lipid or by providing for a linking group on a lipid present in the bilayer for linking to a functionality of the site- directing compound. Such techniques are well known to those skilled in the art. Ligand-directed DNA-polycation complexes have been shown to transfect to hepatocytes in the liver after iv injection; the ability to transfect other cell types or tissue types by this approach has not been demonstrated.

Non-cationic lipid carriers, particularly pH εensitive liposomes, offer another potentially attractive approach to in vivo gene therapy. However, as compared to cationic liposomes, pH sensitive liposomes are less efficient in capturing DNA and delivering DNA intracellularly and may be inactivated in the presence of serum, thus limiting their iv use.

A number of factors can affect the amount of expression in transfected tissue and thus can be used to modify the level of expression to fit a particular purpose. Where a high level of expression is desired, all factors can be optimized, where less expression is desired, one or more parameters can be altered so that the desired level of expression is attained. For example, if high expression would exceed the therapeutic window, then less than optimum conditions can be used. The factors which can be modified are as follows:

Either the lipid composition of the complex or the mean diameter of the lipid. carriers (when in particle

form such as a liposome) injected can dramatically affect the level of transgene expression produced in vivo . Thus, the liposomal lipid compositions generally have a composition of 50% molar ratio of cationic lipid to non- cationic lipid, but may range from 5% to 100%. The diameter of the lipid carriers should generally be within the range of lOOnm to 10 μm. Cationic lipid carrier-DNA complexes wherein the lipid carriers range from 100 nm to several μm in diameter can produce significant levels of transgene expression after syεtemic introduction into a mammalian host.

The use of lipid carriers of greater than 500 nm (i.e. multilamellar vesicles (MLV) or large unilamellar vesicles (LUV) ) can in certain cases significantly increase the level of transgene expression achieved in a mammalian host when compared to small unilamellar vesicles (SUV) . MLV and LUV are prepared by vortexing rather than sonicating after addition of the aqueous material to the dry lipid film. Where it is desired to use particles, the reεulting lipid carrierε can be extruded under high pressure through sized polycarbonate membranes to achieve particular uniform size distributions for a particular use.

The use of particular nucleic acid to lipid carrier ratio also can increase the amount of transfection and/or the level of transcription and/or expression of a nucleic acid sequence of interest. The ratios used determine whether and to what level transgenes are expressed in vivo thus can be optimized, depending upon various factors including the nature of the construct, the size and lipid composition of the lipid carrier and whether it is MLV or SUV, the route of administration and the host mammal. As an example, using a reporter gene CAT (chloramphenicol acetyl transferase) , an approximately 1:1 (range 0.5:1 to 2:1) DNA to lipid .

carrier ratio (μg DNA to nmoles of the cationic lipid) produces the highest levels of gene expression in a mouse in all organs after ip administration, and an approximately 1:4 ratio, (range 2:1 to 1:7) produces the highest levels of gene expression in all organs after iv administration. In addition to achieving a high level of transgene expression in a wide variety of tisεues using optimal conditions, the majority of all cells present in the lung, spleen, lymph nodes and bone marrow are transfected in vivo , aε well as the majority of all endothelial cells present in the heart.

The DNA:lipid carrier ratio also can affect at what level transgeneε are expressed in mammalian hosts after syεtemic injection of the co plexeε. Several factors are important in order optimize the DNA:lipid carrier ratio for a particular expression level desired. Thus, specific DNA:lipid carrier ratios are required for each type of cationic lipid used as well as for each different lipid carrier size used. To optimize for maximum expression, for each lipid carrier composition used, DNA is be mixed together with the lipid carriers in multiple different ratios, ranging from 6:1 to 1:10 (μg DNA to nmole cationic lipid) , in order first to determine which ratios result in aggregation of the DNA:lipid carrier complexes. Ratios which result in aggregation cannot be used in vivo . The ratios which do not result in aggregation are tested in animal models to determine which of the DNA:lipid carrier ratios confers the deεired level of tranεgene expression in vivo . For example, the optimal DNA:lipid carrier ratios for SUV for DOTMA:DOPE, DDAB:DOPE, DOTAP:DOPE, DOTAP:Cholesterol, L-PE:CEBA, DDAB:Cholesterol and L-PE:DOPE are 1:4, 1:3, (very low activity at all ratios), 1:6, 1:1, 1:5, and 2:1, respectively. DNA:lipid carrier complexes are made in appropriate physiologic solutions. The DNA:lipid carrier

complexes are mixed in physiologic solutions (approximately 290 illiosmoles) which do not themselves induce aggregation of the DNA:lipid carrier complexes. The εolutionε include 5% dextrose in water or normal saline. Cell surface receptors for cationic lipid carriers can be used to both regulate and confer target cell specificity on transgene expression in mammalian hosts. Cationic lipid carrier:DNA complexes are internalized by cells by a classical receptor-mediated endocytosis (see Figure 7) using cell surface receptors which contain specific binding sites for, and are able to internalize, cationic molecules. Using agents such aε cytokineε, growth factors, other soluble proteins and certain drugs, it iε thuε poεεible to εelectively up or down regulate theεe cation-binding receptorε. The rate of up or down regulation of theεe receptors by the appropriate agent will allow selection of specific cells for enhanced or reduced levels of tranεfection in vivo . Cell εurface receptors for naked DNA also can be used both to regulate and to confer target cell specificity on transgenic expresεion in mammalian host.

The most frequent interaction between DOTMA lipid carriers, either uni- or multilamellar lipid carriers, complexed to plasmid DNA and the various cell types (for example, CV-1 monkey kidney cellε, U937 human myelomonocytic leukemia cells, K562, MEL (murine erythroblastic leukemia cells) , rat alveolar macrophages, and alveolar type II cellε) , iε that of lipid carrier adheεion and internalization. Thiε interaction is common to well-defined examples of receptor-mediated endocytosis. All cells which appear to have contacted cationic lipid carrier:DNA complexes ingest the complexes following binding of the complexes to the plasma membrane. All these cell types demonstrate the same

claεεical receptor-mediated endocytic pathway of internalization.

The mammalian hoεt may be any mammal, particularly a mammal having εymptomε of a genetically-baεed diεorder or an infectiouε diεease which is amenable to gene-based therapy. Thus, the subject application finds use in domestic animals, feed stock, such aε bovine, ovine, and porcine, aε well aε primateε, particularly humanε. The mammalian hoεt may be pregnant, and the intended recipient of the gene-baεed therapy may be either the gravid female or the fetuε or both.

In the method of the invention, tranεfection in vivo iε obtained by introducing a therapeutic transcription or expresεion vector into the mammalian host, either as naked DNA or complexed to lipid carriers, particularly cationic lipid carriers. The constructε may provide for integration into the host cell genome for stable maintenance of the tranεgene or for epiεomal expreεεion of the tranεgene. The introduction into the mammalian hoεt may be by any of εeveral routeε, including intravenouε or intraperitoneal injection, intratracheally, intrathecally, parenterally, intraarticularly, intranasally, intramuscularly, topical, tranεdermal, application to any mucous membrane surface, corneal instillation, etc. Of particular interest is the introduction of a therapeutic expresεion vector into a circulating bodily fluid or into a body orifice or cavity, such as lung, colon, vagina, and the like. Thuε, iv adminiεtration and intrathecal administration are of particular interest since the vector may be widely disseminated following such routes of administration, and aerosol administration finds use with introduction into a body orifice or cavity. Any physiologically acceptable medium may be employed for administering the DNA or lipid carriers, εuch as deionized water, saline, phosphate-

buffered saline, 5% dextrose in water, and the like, depending upon the route of administration Other components may be included in the formulation such aε bufferε, stabilizers, biocides, etc. Theεe components have found extenεive exemplification in the literature and need not be deεcribed in particular here. Any diluent or components of diluents that would cause aggregation of the complexes should be avoided, including high salt, chelating agents, and the like. The amount of naked DNA or complexes used will be an amount εufficient to provide for adequate diεεemination to a variety of tissueε after entry of the DNA or complexeε into the bloodεtream and to provide for a therapeutic level of expreεεion in tranεfected tissues. A therapeutic level of expresεion iε a εufficient amount of expression to, prevent, treat or palliate a diseaεe or infection of the host mammal. In addition, the dose of the nucleic acid vector used muεt be εufficient to produce a desired level of tranεgene expreεεion in the tiεεue or tiεses of interest, in vivo for example, > lmg of an expresεion plasmid alone injected into a mouse results in a high level of expression of the CAT gene in multiple tisεues. Other DNA sequences, such aε adenoviruε VA geneε can be included in the administration medium and be co-transfected with the gene of interest. The presence of genes coding for the adenoviruε VA gene product may εignificantly enhance the tranεlation of mRNA tranεcribed from the plaε id if thiε is desired.

The level and tissues of expresεion of the recombinant gene may be determined at the mRNA level and/or at the level of polypeptide or protein. Gene product may be quantitated by meaεuring its biological activity in tissues. For example, enzymatic activity can be measured by biological asεay or by identifying the gene product in tranεfected cellε by immunoεtaining

techniques such as probing with an antibody which specifically recognizes the gene product or a reporter gene product present in the expresεion caεεette. Alternatively, potential therapeutic effectε of the gene product can meaεured, for example where the DNA εequence of interest encodes GM-CSF, by determining the effects of gene expression on survival of lethally irradiated animals in which the GM-CSF transgene is expressed. Production of significant amounts of a transgene product will substantially prolong the survival of these mice. Where expression of the polypeptide/protein or even the mRNA itself confers a changed biochemical phenotype upon the hoεt, the presence of a new phenotype or absence of an old phenotype may be evaluated; for example, as a result of transfection of the host cells, there may be enhanced production of pre-existing desirable products formerly produced in insufficient quantities or there may be reduction or even suppression of an undesirable gene product using antisense, ribozyme or co-suppression technologies; in the case of suppression, a reduction of the gene product may be determined. Typically, the therapeutic cassette is not integrated into the host cell genome. If necessary, the treatment can be repeated on an ad hoc basis depending upon the resultε achieved. If the treatment iε repeated, the mammalian host can be monitored to ensure that there is no adverse immune or other response to the treatment.

As an example, in a clinical setting where it is desired to treat a particular disease state, both the biological efficacy of the treatment modality as well as the clinical efficacy need to be evaluated, if possible. For example, in the treatment of cystic fibrosis, there is ε generalized epithelial dyεfunction which manifests itself as abnormalities in the electrolyte and water

content of luminal liquid or "secretionε" of the airwayε, εweat glands, intestinal and reproductive tracts, and pancreas. The biological efficacy of gene therapy therefore can be evaluated by for example measuring the transepithelial electrical potential difference prior to treatment and following transaction. The evaluations can be done following transfection of nasal cells and also following transfection lung cells. Examples of techniques which can be used for measuring the bioelectric potential difference acroεε respiratory epithelial and cystic fibrosiε as described in Knowles, et al . , (1981) New England General of Medicine 305:1489- 1495; Knowles, et al . , (1983) General of Clinical Investigation 21:1410-1417; and Knowles, et al . , (1983) Science 221:1067-1070. The clinical efficacy, whether treatment of the underlying defect is effective in changing the course of disease, can be more difficult to measure. While the evaluation of the biological efficacy goes a long way as a surrogate end point for the clinical efficacy, and is not definitive. Thus, measuring a clinical endpoint such as the so-called "spirometry" factors which give an indication of lung function of a for example, a six-month period of time, may give an indication of a clinical efficacy of the treatment regimen. Typical measurements would include forced vital capacity (FVC) of the lung and forced expired volume in one second (FEV) . An example of the type of clinical study which can be performed to evaluate the efficacy of gene therapy for cystic fibrosiε iε one which iε being used for amiloride for the treatment of lung disease and cystic fibrosis: Knowles, et al . , (1990) New England General of Medicine 322:1189-1194. Similarly, one skilled in the art can evaluate the biological and clinical efficacy of a particular gene therapy protocol.

The subject compositionε can be provided for use in one or more procedures. Kits will uεually include the DNA either aε naked DNA or complexed to lipid carriers. Additionally, lipid carriers may be provided in a εeparate container for complexing with the provided DNA. The DNA either for direct adminiεtration or for complexing with lipid carrierε, or the lipid carrier/DNA complexeε may be present as concentrateε which may be further diluted prior to use or they may be provided at the concentration of use, where the vials may include one or more dosages. Conveniently, single doεageε may be provided in εterile vials so that the phyεician or veterinarian may employ the vialε directly, where the vials will have the deεired amount and concentration of agentε. Thus, a vial may contain the DNA or the

DNA/lipid carrier complexes in appropriate proportional amounts. When the vials contain the formulation for direct use, usually there will be no need for other reagents for use with the method. The invention finds uεe in in vivo prevention, treatment and/or palliation of a number of diseases. In vivo replacement of a gene can be accomplished by techniques such as homologous recombination or initial knockout of the aberrant gene and subsequent replacement with the desired transgene. An additional benefit of expressing nucleic acids in appropriate cells in vivo using this invention, is that the encoded proteins would be processed and subject to post translational modification in the correct manner.

The following examples are offered by way of illustration and not by limitation.

EXAMPLES Table of Contents Example 1 Preparation of Plasmids for in vivo Gene Therapy pRSVCAT p5'PRL3-CAT pSIS-CAT pZN20 (see Figure 5) pZN27 (see Figure 9) pZN46 (see Figures 10 A,B) pZN32 (see Figure 11) pZN51 (see Figure 13) pZN60, pZN61, pZN62, pZN63 (see Figures 14, A, B,

C) pCIS-CAT Example 2 Preparation of Lipid Carriers and DNA Complexing with Lipid Carriers Preparation of Lipid Carriers

Plasmid Preparation

Preparation of Lipid Carrier-Plasmid Complexes Example 3 Demonstration by Immunohistochemistry of CAT Gene Expresεion in the Lung after Intravenous

(iv) Injection of pZN27-DDAB:Cholesterol Lipid Carrier Complexes (see Figure 1) Example 4 Expression of pCIS-CAT following Intraperitoneal Administration Example 5 Demonstration of CAT Gene Expression in the spleen After Intravenous (iv) Injection of p5'PRL3-CAT:L-PE:CEBA Complexes

Example 6 Injection of DOTMA:DOPE + pSIS-CAT Plaεmid clearly did Not Produce Detectable CAT Gene Expreεεion in vivo

Example 7 Interaction of DNA:Lipid Carrier Complexeε with Cell Surface Receptorε (see Figure 7)

Example 8 Demonεtration that Mouεe T Lymphocyteε are Tranεfected in vivo (see Figureε 2 A,B)

Example 9 Demonεtration that Mouse Hematopoietic Bone Marrow-derived Cells are Transfected in vivo Example 10 Demonstration that Human CD4 ⁺ Lymphocytes, freshly isolated from Normal Donors, are Transfected (see Figure 4)

Example 11 Efficient Transfection of a variety of Human Lung Cancer Cell Lineε uεing Cationic Lipoεome-mediated Delivery of DNA (see Figure 21)

Example 12 Tranεfection of Lung Cancerε in Mice by Intravenouε Injection of Cationic Lipid Carrier:DNA Complexeε (see Figure 8 A, B) Example 13 Demonstration of High Level CAT Gene Expression in Multiple Tisεueε after Intravenouε (iv) Injection of pZN27 alone, or pZN27:DDAB:Choleεterol SUV Complexes

Example 14 Induction of High Levels of Human Interleukin-2 in the Spleen and Lymph Nodeε of Mice by Intravenouε Injection of Cationic Lipid Carrierε Complexed to a CMV- Interleukin-2 Gene

Example 15 Induction of High Level Expreεεion of the Human CFTR Gene in Mice Treated by iv Adminiεtration of pZN32:Cationic Lipid Carrier Complexes (see Figure 12E)

Example 16 Demonstration of CAT Gene Expression in Lung and Liver after Intravenous Injection of

Different CAT Gene-Containing Plasmids (see Figure 15)

Example 17 Generalized versus Tissue and Cell Type- specific CAT Gene Expresεion produced by iv Injection of CMV-CAT-Lipoεome or CFTR-CAT- Liposome Complexes, respectively (see Figure 16 A-K)

Example 18 Comparison of Transfection Focusing on iv Injection of Plasmid alone with iv Injection of Plaεmid Complexed to a Lipid Carrier (see Figure 19 A,B)

Example 19 IV Injection of GM-CSF Expression Plasmid- Cationic Liposome Complexes Produces Significant Antitumor Effects pZN84 (see Figure 17)

Example 20 Prolonged, High Level Murine GM-CSF Gene Expression in in vivo after Intravenous (iv) injection of pZN84 complexed to DDAB:Cholesterol (1:1) Liposomes into Goats (see Figures 22, 23)

Example 21 Influence of Liposome-GM-CSF Plasmid Complex on the Course of Experimental Viral Pneumonia in Mice

Example 22 High Level Expression of the CAT Gene in Mouεe Brain produced by Injection of DNA alone or DNA-Cationic Liposome Complexes directly into the Central Nervous System (see Figure 20) .

Example 23 Demonstration of CAT Gene Expression in the Lung after Intravenous (iv) Injection of pRSV-CAT:L-PE:CEBA Complexes

Example 24 Demonstration of CAT Gene Expression in Multiple Tissueε after Intravenous (iv) Injection of pZN20-CAT:DDAB:DOPE Complexes

INTENTIONALLY LEFT BLANK

Example 1 Preparation of Plasmids for in vivo Gene Therapy

Details regarding the plasmids that have been used for transfection of mammalian cells are as follows. pRSVCAT: construction of this plasmid is described in Gorman et al . , (1982) Proc. Nat . Acad . Sciences (USA) 79:6777-6781. In the pRSVCAT plasmid, the 3'-RSVLTR is juxtaposed as a promoter upstream from CAT encoding sequences. The distance between the LTR transcriptional start site and the CAT initiation codon (the first AUG downstream from the start site) is about 70 bp. p5'PRL3-CAT: construction of this plasmid is described in Sakai et al . , (1988) Genes and Development 2 : 1144-1154. pSIS-CAT: construction of this plasmid is described in Huang and Gorman, Nucleic Acids Research , (1990) 18:937-948.

PZN20: construction of this plasmid is illustrated in Figure 5. The plasmid was prepared as follows. pCATwt760 (Stinski and Roehr, (1985) J. Virol . 55:431- 441) was treated with Jϊindlll and the fragment containing the HCMV IE 1 enhancer and promoter element purified. The isolated fragment was then cloned into the ifindlll site of pSP72 (Promega) creating pZN9. Clones were screened in which the enhancer and promoter element is as shown in Figure 5. Following partial Hindlll digestion of pZN9, the blunt ends were filled in with DNA polymerase I Klenow fragment. The resulting clone pZN12 had lost the ifindlll site 5' to the enhancer and promoter element. pZN12 was then treated with Wcol and Hindlll and the large Nσol-Hindlll fragment purified and ligated to a purified small Wcol-Hindlll fragment from PBC12/CMV/IL-2 (Cullen, (1986) Cell 16:973-982. pBC12/CMV/IL-2 contains the HCMV promoter from the AD169

strain. The resulting clone waε pZN13. pZN13 waε partially digeεted with BamHI, filled in with DNA polymeraεe I Klenow fragment and the reεulting clones screened for the clone which had loεt the BamHI εite at the 5' end of the enhancer and promoter element. The reεulting clone was called pZN17.■ pZN17 was treated with Hindlll and BamHI and the resulting fJindlll-BamHI large fragment was purified and ligated to a purified small iindlll-BamHI fragment obtained from pSV2-CAT (Gorman, et al . , (1982), Molecular Cell Biology 2 :1044-1051) . The resulting clone was pZN20. The full restriction map of HCMV (Towne) is shown in Figure 6A. HCMV (AD169) is shown in Figure 6C. A comparison of the two promoterε iε shown in Figure 6B. Significantly more expreεεion iε obtained when a promoter from the AD169 εtrain iε uεed aε compared to one from the Towne εtrain. pZN20 contains a composite promoter which has the Towne sequence 5' of the Wcol site and the AD169 sequence 3' of the Wcol site. The Wcol site is indicated by the asterisk in Figure 6B. pZN20 has this composite HCMV promoter followed by the CAT gene, SV40 t-intron and SV40 polyA adenylation εite.

PZN27: Conεtruction of thiε plaεmid iε illuεtrated in Figure 9. pZN27 contains the composite HCMV promoter followed in order by the SV40 t-intron, the CAT coding sequence and the SV40 polyA adenylation site. pZN46: Construction of this plasmid iε εhown in Figure 10A and Figure 10B. pZN46 containε the compoεite HCMV promoter, followed by the human IL-2 gene, rat preproinεulin 2 intron and polyA addition site from the rat preproinsulin 2 gene. These last three components were derived from the pBC12/CMV/IL-2 plasmid of Cullen (Cell , (1986) 16:973-982. The rat preproinsulin 2 intron waε modified by deleting an internal 162 base pair Wdel fragment.

PZN32: Conεtruction of thiε plaεmid iε εhown in Figure 11. pZN32 containε the compoεite HCMV promoter followed in order by the modified rat preproinsulin 2 intron described for pZN46, human CFTR cDNA, and rat preproinsulin 2 gene polyA addition site as described for pZN46. CFTR cDNA was obtained from pBQ4.7 from F. Collins (Univ. of Michigan) . pZN51: Construction of this plaεmid iε shown in Figure 13. pZN51 contains the composite HCMV promoter followed by the CAT coding sequence and the SV40 polyA adenylation site.

PZN60, PZN61. pZN62, pZN63 : Construction of these plasmids is shown in Figure 14 ^' . pZN60 contains the HCMV composite promoter followed by the modified rat preproinsulin 2 intron, the CAT coding sequence, and the SV40 polyA addition εite. pZN61 is identical to pZN60 but contains an additional 166 base pairs 5' to the intron. This additional DNA is the 166 base pairs immediately 5' of the intron in the pBC12/CMV/IL-2 plasmid and contains rat preproinsulin 2 gene coding sequence. pZN62 is similar to pZN60 except that the intron is 3' of the CAT coding sequence rather than 5' as in pZN60. pZN63 is identical to pZN62 except for the additional 166 base pairs 5' to the intron. This iε the same additional sequence described for pZN61. pCIS-CAT: Thiε plasmid was made as described in Huang, M.T.F. and Gorman, CM., (1990) Nucl . Acids Res . JUB:937-947, with the exception that a CMV promoter and a hybrid intron sequence were used rather than the SV40 promoter in the plasmid pML.I.CAT, Gorman, et al . , (1990) DNA Protein Eng. Tech., 2 _. :3-10.

Example 2 Preparation of Lipid carriers and DNA Complexing with Lipid carriers

Lipid carriers containing a cationic lipid, such as N-[ (1-2-3-dioleyloxy) propyl]

-N,N,N-triethylammonium chloride (DOTMA) , dimethyl dioctadecyl ammonium bromide (DDAB) , or l,2-dioleoyloxy-3-(trimethylammonio)propane (DOTAP) or L- lysinyl-phosphatidylethanolamine (L-PE) and a second lipid, such as dioleoylphosphatidylethanolamine (DOPE) or cholesterol, were prepared as follows.

Preparation of lipid carriers:

Lipids, e.g. DDAB, L-PE, cholesterol-ester-jø- alanine (CEBA) , DOTAP, and choleεterol (Choi) were dissolved in chloroform. Proper amounts of each lipid (determined by the desired molar ratio of each lipid in the final lipid carrier formulation usually 1 to 1 moles cationic lipid to moles non-cationic lipid but ranging from 5 to 1 up to 1 to 5) were mixed together and evaporated to dryness on a rotary evaporator. The lipid film was then resuspended by vortexing after the addition of 5% dextrose in water or lipid carrier buffer (25 mM Tris-HCl pH 7.4, 100 μM ZnCl ₂ , NaCl to isotonicity. To make a final lipid concentration of 20mM of multi- lamellar vesicles (MLV) . For the preparation of small unilamellar vesicles (SUV) , the mixture was then sonicated in a bath sonicator for 15 min, and the lipid carrierε were εtored under argon at 4°C until use.

Plasmid Preparation:

The E. coli strain HB101 which carried the plasmid was grown in TB at 37°C. The method of plasmid purification is a modification of the protocol of "lysis by alkali" and "purification of plasmid DNA by precipitation with polyethylene glycol" described by Sambrook et al . , (1989) (Molecular Cloning, 2nd edition, Cold Spring Harbor Laboratory Presε) . The modification waε that the precipitation of DNA by PEG waε omitted. The final DNA preparation waε diεεolved in lOmM Triε-HCl pH 8.0.

Preparation of lipid carrier-plaεmid complexeε:

Plasmids were diluted separately in 5% dextrose in water solution to the desired concentration (uεually 1 μg/μl) . The lipid carrierε were also diluted in 5% dextrose in water to the same volume as the plasmid.

The amounts of lipid carrierε used were determined based on the ratio of nmoles of lipid to μg of plasmid added, e.g. for lipid carrier:plasmid=l:1, one nmole of cationic lipid is mixed with 1 μg of plasmid DNA.

Plasmid and lipid carriers were then mixed together to form DNA:lipid carrier complexes.

Dose injected. At least 50 μg, and routinely 100 μg of plasmid DNA complexed to cationic lipid carriers were injected intravenously by tail vein per mouse. For injection of plasmid alone, at least 500 μg and routinely

2 mg of plasmid DNA were injected intravenously by tail vein per mouse.

Example 3 Demonstration by Im unohistochemiεtry of CAT Gene Expression in the

Lung After Intravenous (iv) injection of pZN27-DDAB: Cholesterol Lipid carrier Complexes

Lipid carrier: DDAB:Cholesterol=l: 1, εtock 20 mM in lipid carrier buffer. Plaεmid: pZN27.

DNA:Lipid carrier Ratio: lipid carrier:plaεmid=5 nanomoleε cationic lipid: 1 μg DNA

DNA dose: 100 μg plasmid DNA in 200 μl 5% dextrose in water was injected iv by tail vein per mouse. Mice: ICR, female, 25 grams.

Immunohistochemical staining to detect CAT protein in lung sections of mice treated in vivo .

Procedure: Forty eight hours after injection of the pZN27-DDAB:Cholesterol complexes, the lungs are removed, perfused with 33% O.C.T. Embedding Medium (Miles, Inc.), embedded in O.C.T. and snap frozen. Frozen tisεueε were εectioned at 6 μm, collected onto glaεε εlideε, fixed for 10 minuteε in 4°C acetone and then placed in 0.2% Triton X-100 to permeabilize membraneε. Sectionε were then incubated for 12-48 hourε with the monoclonal anti-CAT antibody (available from Dr. Parker Antin, Univ. of Arizona) or iεotype negative control antibody. at the appropriate dilution. After waεhing, a biotinylated antibody directed againεt the primary antibody (Zymed, S. San Franciεco) was added for a minimum of 60 minutes, followed by application of the streptavidin-alkaline phosphataεe complex (Zymed) for 60 minuteε. The εubstrate-chromogen appropriate for the enzyme label waε then applied, per the manufacturer'ε inεtructionε.

Slides were coverslipped in water-soluble mounting media for examination.

Results: The resultε are shown in Figure 1 and demonstrate diffuse staining of the lung. The stain localizes to the alveolar walls, indicating that greater than 70% of pulmonary vaεcular endothelial cellε, aε well aε alveolar lining cellε, including type I and type II cells and alveolar macrophages are transfected by a single iv injection of DNA lipid carrier complexes. In addition, εignificant numbers of bronchiolar airway lining cells stain positively for CAT protein, and are therefore transfected in vivo by iv injection of lipid carrier:DNA complexes. Thus, the great majority of all cells in the lung transfected by an iv injection of pZN27-DDAB:Cholesterol complexes.

Example 4

Expression of pCIS-CAT

Following Intraperitoneal Administration

Effect of the Amount of pCIS-CAT-cationic lipid carrier complexes injected ip on the level of CAT gene expression in vivo .

Female ICR mice were injected ip with 1 ml of 5% dextrose in water containing 0.01, 0.1, or 1 mg of pCIS- CAT expresεion plasmid complexed with 0.01, 0.1 or 1 μ oles, respectively, of DDAB:DOPE lipid carriers. Mice were εacrificed 48 hours later, the organs removed, and tisεueε were homogenized in 0.25M Tris-HCl buffer pH 7.8, using a hand-held homogenizer. Cytoplaεmic extracts were made, normalized by protein content and the level of CAT activity then measured. The experimentε compriεe three animals per group and the results show the mean dpm ± SEM of acetylated chloramphenicol.

Methods: Lipid carrierε containing DDAB were prepared in 1:1 molar ratio with DOPE, aε followε: 10 μmoleε of DOPE dissolved in chloroform and 10 μmoles of the cationic lipid, dissolved in ethanol were evaporated to drynesε on a rotary evaporator. One ml εterile of water waε added, and the mixture was sonicated in a bath sonicator (Laboratory Supply, Hickεville, NY) for 20 min. Lipid carriers had mean diameters of approximately 100 ±25 nm. For CAT asεay, cell extractε were made, and their protein content determined by the Coomassie Blue assay (BioRad, Richmond) . One hundred μg of protein from the lung, spleen, liver, and heart extractε, and 50 μg of lymph node extract were reacted with ¹⁴ C labeled chloramphenicol and chromatographed aε previously described (Gorman, supra) . To calculate dpm, both the acetylated and unacetylated specieε were cut from TLC plateε and radioactivity counted in a scintillation counter.

Results: To assess potential dose-response relationships in vivo , animals were injected animals in groups of three with 0.01 mg, 0.1 mg, or 1 mg of pCIS-CAT plasmid complexed with 0.01 μ ole, 0.1 μmole, or 1 μmole respectively of DDAB:DOPE lipid carriers. Both the 0.1 mg and 1 mg DNA doseε produced highly significant levels of CAT protein (p < 0.005) in all the organs asεayed.

Maximal levelε of CAT gene expreεsion in each organ were produced by the 1 mg DNA dose: increaεing the DNA-lipid carrier doεe 10 fold reεulted in an approximately 2 fold increase in lymph node CAT levels and a 3 fold increaεe in the spleen.

Example 5 Demonεtration of CAT gene expression in the spleen after intravenous (iv) injection of p5'PRL3-CAT:L-PE:CEBA complexes.

Lipid carrier: L-PE:CEBA=1:1, stock 20 mM in lipid carrier buffer.

Plasmid: p5'PRL3-CAT. (Sakai, et al . , (1988) Genes and Development 2 : 1144-1154) DNA:Lipid carrier Ratio: lipid carrier:plasmid=l nmole cationic lipid:

1 μg plasmid DNA. DNA dose: 200 μg plasmid DNA in 200 μl 5% dextrose in water was injected by tail vein per mouse. Mice: Balb/c, female, 25 gramε. Tissue extraction procedure: Forty eight hours after tail vein injection, mice were sacrificed, whole spleens were homogenized in 1ml of 0.25M Tris-HCl pH 7.8, 5mM EDTA, 80 μg/ml PMSF and the resulting extract was centrifuged and then the supernatant was subjected to 3 cycles of freeze-thaw and then heated to 65°C for 20 min. CAT assay procedure: 100 μl of extract+10 μl of 20 mM acetyl CoA+4 μl of ¹⁴ C-chloramphenicol (25 μCi/ml, 55 mCi/mmole, Amersham) were incubated together at 37°C for 6 hr. At 3 hours, an additional 10 μl of acetyl CoA was added.

Results:

This experiment showed that a εignificant level of CAT activity waε preεent in the εpleen extract of the treated animal, but not in the extract of control εpleen, taken from an animal injected with lipid carrier alone.

Example 6 Injection of DOTMA:DOPE+pSIS-CAT Plasmid Clearly Did Not

Produce Detectable CAT Gene Expression in vivo)

Lipid carrier: DOTMA:D0PE=1: 1, in 5% dextrose in water Plasmid: pSIS-CAT (Huang, and Gorman, (1990) Nucleic

Acids Research 18:937-947) ♦ Ratio: Cationic lipid:plasmid=4 nmoles: 1 μg, dose: 100 μg DNA in 200 μl 5% dextrose in water. Mouse: ICR, female, 25 gramε. Injection: tail vein.

Tissue collection and procesεing: Mice were sacrificed at day 2 and day 6, and lung, spleen, liver, and heart were collected. The whole organs were homogenized in 0.5 ml, except livers which were homogenized in 2.0 ml, of 0.25M Tris-HCl pH 7.8, 5 mM EDTA, 2 μg/ml aprotinin, 1 μg/ml E- 64, and 0.5 μg/ml leupeptin (all proteaεe inhibitorε purchased from Boehringer Mannheim) . Extracts were subjected to three cycles of freeze-thaw, then heated to 65°C for 10 min.

CAT assay; 100 μl of extracts for each aεεay with 0.3 μCi of ¹⁴ C-chloramphenicol and 10 μl of 20 mM acetyl CoA at 37°C for either 5 hrε or 24.5 hrs, and the materials were then extracted using ethyl acetate and analyzed on TLC plates. Results:

There were no acetylated chloramphenicol specieε preεented as determined by comparing the extracts from treated animals with that from control animalε. Thuε, under εimilar experimental conditionε that produce high level expreεsion of pZN27, the use of the pSIS-CAT expreεεion vector doeε not reεult in any detectable expression of the linked-CAT gene in any of the tisεueε aεεayed in vivo . The lack of expression of pSIS-CAT in

vivo may be due either to a different promoter-enhancer element (SV40) or to a different intron sequence when compared to the pZN27 vector, which yields high level in vivo expression.

Example 7

Interaction Of DNA: Lipid carrier Complexes With Cell Surface Receptorε

Cells and cell culture: CV-1 (African green monkey kidney) , U937 (human myelocytic leukemia) , murine erythroleukemia (MEL) cellε, and K562 cellε (human erythroleukemia cellε were obtained from the American Type Culture Collection (Rockville, MD) . CV-1 and MEL cellε were maintained in Dulbecco minimum essential medium DME-H-21 with 5% fetal bovine serum (FBS) at 37°C and 7% C0 ₂ - Rat alveolar type II cells and rat alveolar macrophages were isolated and purified aε previouεly described. (Debs et al . , (1987) Amer. Rev. Respiratory Disease 135:731-737 : Dobbs, (1986) Amer . Rev. Respiratory Disease 134:141-145) Type II cells were maintained in DME-H-16 with 5% FBS at 37°C and 7% C0 ₂ . Twenty nmole of DOTMA:DOPE lipid carrierε complexed to 20 μg of pRSV-CAT plasmid DNA were added to 2 million cells growing in 60 mm Falcon plastic disheε (either SUV or MLV) , and fixed for EM at time pointε from 15 minuteε to 2 hourε thereafter.

Fixation and Procesεing for Electron Microscopy

DOTMA lipid carrierε and cells in tissue culture or freshly iεolated from blood or pulmonary alveoli were fixed in 1.5% glutaraldehyde in 0.1 molar εodiu cacodylate buffer containing 1% sucrose, pH 7.4, at room temperature for 1 hr. Following tannic acid and uranyl acetate enhancement, tisεue waε dehydrated in a graded .εerieε of alcohols and* embedded in epoxy 81*2 resin

(Ernest F. Fullam, Inc., Latham, NY) sectioned on an MT 2 microtome using diamond knives and examined with a Jeol 100CX transmiεεion electron microεcope operating at 80 kV. Electron photomicrographs of the internalization of lipid complexes in CV-1 monkey kidney cells are shown in Figure 7. The arrows in panel (a) show particles binding to clathrin-coated pits; in panel (b) , the arrows show where the particles have been ingested and are present in endosomes. The most frequent interaction between DOTMA lipid carriers, either uni- or multilamellar lipid carriers, complexed to plasmid DNA and the variouε cell typeε (CV-1 monkey kidney cells, U937 human myelomonocytic leukemia cells, K562, MEL erythroblastic leukemia cells, rat alveolar macrophages, and alveolar type II cells) , is that of lipid carrier adheεion and internalization in a typical coated vesicle pathway (Fig. 7 a-f) . This interaction is common to well defined examples of receptor-mediated endocytosiε. All cellε which appear to have contacted cationic lipid carrier:DNA complexes ingest the complexes after binding to the plasma membrane. All these cell types (derived from rodent, monkey and human cells, demonstrated the same classical receptor-mediated endocytic pathway of internalization. DNA-cationic liposome complexes were generally taken up by human cells as well as or better than nonhuman and particularly rodent cells of similar lineage.

Example 8 Demonstration that Mouse T Lymphocytes are Transfected in vivo .

Female ICR mice were injected ip with 1 ml of 5% dextrose in water containing 1 mg of a pZN27 plasmid complexed to 1 μmole of DDAB:DOPE (1:1 molar, SUV) lipid

carriers. Mice were sacrificed 48 hrs. later, the spleen and lymph nodeε were removed, and rendered into single cell suspensions by homogenizing in serum containing medium. The cells were then incubated with FITC- conjugated anti-Thy 1.2 antibody (provided by Dr. J. Beck, San Franciεco Veteranε Adminiεtration Medical Center) and sorted by FACS. The Thy 1.2 ⁺ T lymphocyte fraction was cytospun onto microscope slideε and fixed and probed for the preεence of intracellular CAT protein after permeabilizing the cells using 0.25% Triton X-100. The cells were incubated with an anti-CAT monoclonal antibody (gift from Dr. P. Antin, Univ. of Arizona) for 1 hr. at 20°C and then stained with a Texas Red- conjugated goat-anti-mouse IgG for 1 hr. at 20°C. Figure 2A shows a field of T lymphocytes by phase contrast microεcopy and 2B shows the same field viewed by fluorescence microscopy. These results demonstrate that more than 70% of Thy 1.2 ⁺ T lymphocytes are transfected in vivo (as shown by the red fluorescence) , by a single ip injection of pZN27:DDAB:DOPE complexeε. Thy l.2 ⁺ lymphocyteε from untreated mice do not show immunofluorescent staining as shown in 2D; a phaεe contraεt micrograph of the same field iε εhown in 2C.

Example 9 Demonεtration that Mbuεe Hematopoietic Bone Marrow- Derived

Cells Are Transfected in vivo .

Female ICR mice were injected ip with 1 ml of 5% dextrose in water containing 1 mg of a pZN27 plaεmid complexed to 1 μmole of DDAB:DOPE SUV lipid carrierε. Mice were εacrificed 48 hrs. later, bone marrow-derived hematopoietic cellε were then obtained by perfuεing the femur cavity ith RPMI-1640 medium and then, homogenizing

clumpε to obtain a εingle cell suspenεion. Bone marrow cells were then centrifuged onto glass slideε and fluorescently stained as described in Example 8. This experiment demonstrated that approximately 20% of mouse bone marrow hematopoietic cells (including cellε that on the baεiε of morphology are primitive myeloblaεtic and erythroblastic precursor cells) were tranεfected in vivo by a single ip dose of pZN27:DDAB:DOPE complexeε.

Example 10 Demonεtration that Human CD4 ⁺ T Lymphocytes, Freshly Isolated from Normal Donors, are Transfected in vitro .

Buffy coat preparations were freshly isolated from normal human donors by gradient centrifugation. The cells were then panned using an anti-CD3 (Becton-

Dickinson, Mountain View, CA) monoclonal antibody to isolate the CD3 ⁺ T lymphocyte fraction. These cells were then transfected uεing the following protocol: 10 million cellε were plated on 100 mm diεheε and then 25 μg of pZN27 complexed to 50 nmoles of DDAB:DOPE (1:1) SUV lipid carriers were added for 48 hours. Control cells were not tranεfected. The cellε were then incubated with a FITC- conjugated monoclonal anti-CD4 antibody (Becton- Dickinεon) and sorted by FACS. The resulting CD4 ⁺ T lymphocyteε were cytoεpun onto microεcope εlideε and fixed and probed for the presence of intracellular CAT protein after permeabilizing the cells using 0.25% Triton x-100. The cells were incubated with an anti-CAT monoclonal antibody for 1 hr at 20°C and then stained with a Texaε Red conjugated goat anti-mouse IgG for 1 hr at 20°C. Resultε:

Reεultε are shown in Figure 4 and demonstrate that at least 70% of freshly isolated human CD4 ⁺ T lymphocytes are transfected after exposure to pZN27:DDAB:DOPE complexes in culture as shown by the red fluorescence of cells in 4B. Control untransfected cells showed no fluoreεcence. Theεe reεults suggest that this approach may dramatically improve the therapy of diseases including AIDS and cancer.

As the above resultε show, high level transgene expreεεion haε been achieved in heart, kidney, lymph nodes, bone marrow cells, liver, lung and spleen after εyεtemic (iv or ip) tranεgene adminiεtration. Transfection of the heart, kidney, lymph nodes or bone marrow cells individually after syεtemic (iv or ip) tranεgene adminiεtration into adultε haε not previously been accomplished. Transfection of T lymphocytes, lung airway or alveolar cell types, cardiac endothelial lining cells and cardiac muscle cells, and bone marrow hematopoietic precursor cellε in vivo by εystemic administration of DNA has not been shown previously. Specifically, greater than 50% of T lymphocytes, lung airway epithelial, alveolar and vascular endothelial cell types, cardiac endothelial lining cells and bone marrow hematopoietic precursor cells (including about 70% of blast cells) are transfected in vivo , following one iv or ip injection of CAT expreεεion plaεmid-cationic lipid carrier complexeε. Transfection of a high percentage of all the cells present in any εingle tissue has not been reported previously.

Example 11 Efficient tranεfection of a variety of human lung cancer cell lineε uεing cationic lipoεome-mediated delivery of DNA. Method:

Cell Culture: NCI-H69, NCI-H82, and NCI-H520 cellε were used. H69 and H520 cells were grown in RPMI-1640 with 10% fetal bovine serum (FBS) and H82 cells were grown in Dulbecco's minimum eεsential medium (DME)-H21 with 10% FBS.

Liposome preparation: Liposomes were prepared as followε: a total of 4 μmoles of lipid dissolved in chloroform, (or in ethanol (DOTMA) ) were evaporated to dryness on a rotary evaporator. One ml of 50 mM Tris, 0.5 mM EDTA, 50 mM NaCl, 100 μM ZnCl ₂ buffer per 20 mmoleε of lipid was added, and the mixture was sonicated in a bath sonicator (Laboratory Supply Co., Hickεville, NY) for 20 min. ^N The resulting liposomeε have an approximate mean diameter of 100 ± 25 nm. The following lipoεome preparationε were uεed: pure DOTMA,

DOTMA:Cholesterol in a 2 to 1 molar ratio, pure L-PE or L-PE:CEBA in a 6 to 4 molar ratio.

Cellular transfection: For transfection of cells, 2 x 10 ⁶ cells in 4 ml of serum-free medium were plated in 100 mm plaεtic petri dishes (Falcon, Oxnard, CA) . The plasmid DNA-liposome complex waε prepared by firεt adding 1) DNA and then 2) , lipoεomeε and mixing gently. The complex waε then εuspended in 1 ml of εerum-free medium and added to the cellε. Four hours later, the cells were washed twice, resuspended in 10 ml of serum-containing medium, and subεequently harveεted, 44 hours later. Just prior to harvesting, the cells were waεhed 2 ti eε, and the plateε were then εcraped with a rubber policeman. The cellε were centrifugated at 1,000 x g for 5 min, and 0.135 ml of 0.25 M Tris buffer pH 7.5, 5 mm EDTA was

added to each pellet. The cells were freeze-thawed 3x, heated at 65°C for 10 min, and spun at 12,500 x g for 10 min. The supernatant was assayed for protein and 20 μg of supernatant protein per sample was uεed to measure CAT activity, as described in Example 4.

Results: The results demonstrate the ability of cationic liposomeε to mediate high level tranεfection of two different human small cell lung cancer lines (H69 and H82) and a εquamouε cell lung cancer line (H520) . All three lines were very efficiently transfected by RSV-CAT when complexed to 3 different cationic liposome formulations (Figure 21) . These human cell lines were transfected either as, or more efficiently than rodent tumor cell lines tranεfected under comparable conditionε.

Example 12

Transfection of lung cancers in mice by intravenous injection of cationic lipid carrier:DNA complexes

Mouse: C57/black 6, female, 25 grams. Cancer: B16, mouse melanoma line which iε highly metaεtatic to lung. The cell line was grown in RPMI 1640 medium supplemented with 5% fetal calf εerum. Lipid carrier: DOTAP:Choleεterol=l.1, lOmM in 5% dextroεe in water. Plasmid: pZN20.

Ratio: Cationic lipid:DNA=6 nmoles:1 μg, 100 μg in 200 μl of 5% dextroεe in water were injected by tail vein into each mouεe. Inoculation of cancer cell line into mice and adminiεtration of CAT expreεεion plaεmid-cationic lipid carrier complexes:

B16 cells were trypsinized off the plateε and 50,000 cellε were inoculated into each mouse by

intravenouε injection into the tail vein. Two weekε after injection, cationic lipid carrier-DNA complexeε were injected via tail vein. Lungε were collected 48 hours postinjection, infused with 33% O.C.T., frozen in a dry ice-ethanol bath, cryosectioned, and processed for immunohiεtochemical analyεiε to detect intracellular CAT protein. Immunohistoche ical analysis: Procedure:

Organε were removed, appropriately trimmed, embedded in OCT and εnap frozen. Frozen tiεεues were sectioned at 6 μm, collected onto glasε εlideε, fixed for 10 minuteε in 4°C acetone and then placed in 0.2% Triton X-100 to permeabilize membraneε. Sectionε were then incubated for 12-48 hourε with the monoclonal anti-CAT antibody or iεotype negative control antibody at the appropriate dilution. After waεhing, a biotinylated antibody directed againεt the primary antibody (Zymed, S. San Francisco) was added for a minimum of 60 minutes, followed by application of the streptavidin-peroxidase complex (Zymed) for 60 minuteε and then application of the εubstrate-chromogen appropriate for the enzyme label used. Slides can then be counter-εtained in dilute hematoxylin or left unεtained and coverεlipped in water- soluble mounting media for examination. Resultε:

Immunohistochemical analysiε are shown in Figure 8 and demonstrate that B-16 melanoma lung tumors (8A _f indicated by arrows) aε well aε intravaεcular tumor emboli (8B, indicated by arrowε) are efficiently tranεfected after iv injection of DNA-lipid carrier complexes. Both lung tumors and intravaεcular tumor emboli εhow intenεe staining, indicating efficient, a generalized transfection in vivo . Tumor bearing mice, which did not receive an injection of DNA-lipid carrier complexeε, show no CAT activity in the lung or in any

lung tumor cells (8C) . The ability to transfect tumors present within mammalian hosts by systemic administration of a cloned gene has not previously been demonstrated.

Example 13 Demonstration of High Level CAT Gene Expression in Multiple

Tisεueε After Intravenous (iv) Injection of PZN27 Alone, or pZN27:DDAB:Choleεterol SUV Complexeε.

Lipid carrier: DDAB:Choleεterol=l: 1, εtock lOmM in 5% dextroεe. After addition of 5% dextroεe to the dried lipid film, the SUV were prepared by εonication in a bath sonicator for 20 minutes. Plasmid: pZN27.

DNA:Lipid carrier Ratio: Cationic lipid:plasmid DNA = 5 nanomoles:l μg DNA. DNA dose:

PZN27 alone: Individual mice received 500 μg, 1 mg, 2 mg, or 500 μg, followed 4 hours later by a second 500 μg dose, respectively of pZN27 in 200 μl 5% dextroεe in water by tail vein injection.

PZN27 complexed to lipid carriers: 100 μg plasmid DNA complexed to 500 nanomoles to DDAB:Choleεterol SUV lipid carrierε in 200 μl 5% dextroεe in water was injected by tail vein per mouse. Mice: ICR, female, 25 grams.

Tisεue extraction procedure: Each organ waε homogenized in 0.3 ml of 0.25 M Tris-HCl pH 7.8, 5 mM EDTA, and the resulting extract was centrifuged and the supernatant waε then εubjected to 3 cycles of freeze-thaw and then heated to 65°C for 20 min.

CAT aεεay procedure: The protein concentration of each tiεεue extract waε quantitated using a Coomassie Blue- based protein asεay (BioRad, Richmond) , and same amount of total protein from each tisεue extract waε added in the CAT aεεay, together with 10 μl of 20 mM acetyl CoA+12 μl of ¹⁴ C-chloramphenicol (25 μCi/ml, 55 mCi/mmole, Amerεham) ) , at 37°C for 13 hrε. Results:

Significant levels of CAT gene expression were εeen in each of the 6 different tiεεues assayed (lung, heart, liver, kidney, spleen, and lymph nodes) after injection of either pZN27 alone, or pZN27 complexed to DDAB:Cholesterol lipid carriers. Expresεion of a tranεgene in multiple tiεεueε in vivo after εyεtemic injection of a naked expreεεion plasmid previously has not been demonstrated.

Example 14 Induction of High Levels of Human Interleukin-2 in the Spleen and Lymph nodes of Mice by Intravenous Injection of Cationic

Lipid carriers Complexed to a CMV-Interleukin-2 Gene

Mouse: C57/black 6, female, 25 grams. Cancer: B16, mouse melanoma line which is highly metastatic to lung. The cell line waε grown in RPMI-1640, 5% fetal calf serum. Lipid carrier: DDAB:Cholesterol=l.1, 10 mM in 5% dextrose in water.

Plasmid: pZN46 (the HCMV promoter enhancer fused to the human interleukin-2 coding sequence) .

Ratio: Cationic lipid:DNA=5 nmoleε:l μg

DNA in 200 μl of 5% dextroεe in water adminiεtered per injection. Inoculation of the tumor cell line into mice and administration of a human interleukin-2 expression plasmid-cationic lipid carrier complexes:

B16 cells were trypsinized off the plates, and 50,000 cells were inoculated into each mouse by intravenous injection into the tail vein. Starting 2 dayε after the tumor cell injection, cationic lipid carrier-DNA complexeε were injected via tail vein 2 timeε per week for a total of 2 weekε. The animals were sacrificed 2 weekε post tumor cell injection, the spleen and lymph nodes were removed, rendered into single cell suspensions using a tisεue grinder, and then cultured for 24 hours in RPMI-1640, 10% fetal calf serum in 100 mm plastic dishes in a 37°C incubator. After 24 hours the supernatant was collected and the concentration of human interleukin-2 in the supernatant was determined using a human IL-2 ELISA. Results:

One hundred picograms of human IL-2/ml per ml was present in the spleen cell supernatant and 91 pg/ l IL-2 was present in the lymph node cell supernatant from the mouse which was injected with the pZN46-DDAB:Cholesterol lipid carrier complexeε. No human IL-2 was detected in either spleen cell or lymph node cell supernatants derived from mice which received an identical injection of B-16 melanoma cells, but which did not receive the pZN46-DDAB:Cholesterol lipid carrier complexes. Thus, subεtituting the human IL-2 gene coding region for the CAT gene coding region in a HCMV expression plasmid reεulted in high level expreεεion of the IL-2 gene in vivo and the production of large amountε of a εecreted human IL-2 protein in a mouse.

Example 15 Induction of High Level Expression of the Human CFTR gene in Mice Treated by iv Administration of pZN32:Cationic Lipid carrier Complexeε

Mice: ICR female, 25 grams

Lipid carrier: DDAB:Choleεterol=l:1 SUV, 10 mM in 5% dextroεe in water

Plasmid: pZN32 (the HCMV promoter enhancer fused to the human CFTR coding sequence)

Ratio: Cationic lipid:DNA=5 nmoles:lμg plasmid DNA. Dose: A total of 100 μg of plasmid DNA in 200 μl of 5% dextrose in water administered per iv tail vein injection. Procedure:

Forty-eight hours after injection the mice were εacrificed, the lungs were removed and trimmed, embedded in OCT, and snap frozen. Frozen tisεues were sectioned at 6 μm, collected onto glaεε εlideε, and the fixed for 10 minutes in 4% acetone. Immmunolocalization of CFTR was then performed using the affinity purified rabbit polyclonal anti-CFTR antibody, -1468 (Cohn, et al . , (1991) Biochem . Biophys . Res . Comm . 181:36-43) . The procedure uεed was identical to the one deεcribed in Marino et al . , (1991) J . Clin . Invest . 28:712 with the following variation. After washing, a biotinylated antibody directed against the rabbit polyclonal antibody (Zymed) was added for 60 minutes, followed by application of the streptavidin phosphataεe complex (Zymed) for 60 minuteε and then application of the εubstrate-chromogen. Slideε were then coverslipped in water-soluble mounting media for examination. Results:

Photomicrographε of frozen εectionε (viewed at different magnificationε) of mouse lung 48 hours following iv injection to pZN32-DDAB:Cholesterol (1:1) liposome complexes and lung from untreated control are shown in Figures 12A-E. Aε demonstrated by the intense staining with the polyclonal anti-CFTR antibody, α-1468, the overwhelming majority of the airways were tranεfected with the human CFTR gene. In 12A, 12C and 12E. by visual inspection, esεentially all the cells in transfected airways stain positively, demonstrating that the overwhelming majority of airway cells are transfected with the human CFTR gene in vivo with a single iv dose of pZN32 complexed to DDAB-choleεterol (1:1) liposome- treated and control animals could not be distinguiεhed histologically. Significant expresεion of the human CFTR gene iε present in at least 50% of all the airways and at least 50% of all of the airway lining cells (by visual inspection) in mouse lungs for at least 60 days following a single iv dose of pZN32 complexed to DDAB-Cholesterol (1:1) liposomes. Frozen sectionε of mouse lungs from control animals (12B and 12D) do not show any detectable staining for CFTR, confirming that all the CFTR expression present in 12A, 12C and 12E iε due to transfection of lung cells with the human CFTR gene. Aε shown by the above resultε, a εingle iv doεe of an expression construct, containing a gene of interest, complexed to cationic liposomes transfectε the majority of the cellε lining the conducting airwayε of the lung, the gene product is present in the lung for at leaεt 60 dayε, the expreεεion appearε to be airway cell-εpecific, and there iε no histological evidence of damage following expoεure. Thiε is important because liposomeε are well tolerated and non-immunogenic. Furthermore, the appearance, behavior, and life εpan of mice treated with intravenouεly injected pZN32:DDAB-Cholesterol (1:1)

complexed appear normal, and are indistinguishable from untreated, normal control animals. This lack of toxicity, as demonstrated by analyεiε of appearance, behavior, life εpan and detailed hiεtologic analyεis of a wide variety of tissueε from theεe animals demonstrates the lack of toxicity produced by in vivo delivery of pZN32-DDAB:Cholesterol complexes. Additionally, the effects of repeated iv administration of the DNA/liposome complexes is effective and iε non-toxic. The cationic liposome-mediated DNA delivery by iv injection provides high level, lung-specific tranεgene expreεεion in vivo .

Example 16 Demonstration of CAT Gene Expression in Lung and Liver After Intravenous Injection of Different CAT gene- Containing Plasmids

Lipid carrier: DDAB:Cholesterol= 1:1, stock 5mM in 5% dextroεe in water. Plasmids: Plasmids are indicated below. DNA-Lipjd carrier Ratio: Cationic lipid: plasmid DNA = 1 nanomole: 1 μg

Dose: 100 μl DNA in 200 μl volume injected intravenously by tail vein injection. Mice: ICR, female, 25 g Procedure: The animals were sacrificed 24 hourε after injection. The tiεεue extraction procedure and CAT aεεay were aε described in Example 4 except that the CAT asεay waε incubated for 3 hr at 37°C and 2.0 mM paraoxon (Lai, et al . , (1988) Carcinogenesis 9_: 1295-1302) waε added to the liver εamples. The resultε are εhown in Figure 15. Lanes 1-12 are lung samples, lanes 13-24 are liver samples. Lanes 1, 2, 13, 14 are pZN51; lanes 3, 4, 15, 16 are pZN60; laneε 5, 6, 17, 18 are pZN61; laneε 7, 8,

19, 20 are pZN62; laneε 9, 10, 21, 22 are pZN63; and laneε 11, 12, 23, 24 are pZN27.

Resultε: pZN51, which does not contain an intron, is expresεed aε well aε, or better than, plaεmids containing an intron either 3' or 5' to the coding sequence.

Example 17 Generalized versus tisεue and cell type-specific CAT gene expresεion produced by iv injection of CMV-CAT-liposome or CFTR-CAT-liposome complexes, respectively.

Mouse: ICR female, 25 grams.

Liposome: DDAB:Cholesterol=l:1 SUV, lOmM in 5% dextrose in water. Plasmid: 1) pZN27 or 2) pBE3.8CAT (see Chou et al . , (1991) J. Biol Chem 166:24471-24476, for construction) . Experimental conditions: Mice in groups of 3 received either (a) , no treatment, or (b) , a single iv tail vein injection of DDAB:Cholesterol liposomes complexed to 100 μg of) a 3.8 kb sequence of the 5' upstream region of the human CFTR gene fused to the CAT gene (pBE3.8CAT) or (c) pZN27. Mice were sacrificed 24 hours later and CAT activity assayed in lung, liver, εpleen, lymph nodeε, kidney and heart, aε deεcribed in Example 4. Immunohistochemical analysis of lung section from each of the groups waε performed as described in Example 3. Results:

Immunohistochemical staining of frozen lung sectionε from these mice showed that iv injection of CMV- CAT-liposome complexes produced high levels of red staining, indicating CAT gene expression in endothelial, alveolar and airway cells within the lung (16A) . In contrast, CFTR-CAT-liposome complexes produced CAT gene

expreεεion primarily localized in airway epithelial cellε (16B) . This approximates the pattern of endogenous CFTR gene expresεion in rat lung, aε determined by in situ hybridization εtudies (Trezise and Buchwald, (1991) Nature 353:434-437. A lung section from an uninjected mouse does not show red staining, indicating that CAT gene expression is present only in tranεfected cellε (Fig. 16C) . Figure 16D iε a high magnification photomicrograph of alveoli from a CMV-CAT treated mouεe and εhowε a high level of CAT gene expreεεion in both alveolar cellε and lung endothelial cellε. A high magnification photomicrograph of alveoli from CFTR-CAT treated mice (Figure 16E) shows no significant CAT gene expresεion in either alveolar or endothelial cellε, demonεtrating that the CFTR promoter targets transgene expresεion to airway epithelial cellε. Thiε is the first demonstration that transgeneε can be expreεsed within mouse lung in either a generalized or cell type-specific fashion depending on the regulatory element used, after iv injection.

CAT assay demonstrated that CMV-CAT produced significant CAT gene expresεion in the lung, liver, heart, spleen, lymph nodeε and kidney, whereaε CFTR-CAT produced lung-εpecific gene expression. Photographs of autoradiographic analysiε of each tiεεue are εhown in Figure 16 (F-K) . Thus, the CMV promoter induces expresεion of a linked gene in a wide range of tissues, whereas the 5' flanking region of the human CFTR gene directs tissue-εpecific tranεgene expreεεion after iv, liposome-based administration.

Example 18 Comparison of transfection focusing on iv injection of plasmid

alone with iv injection of plasmid complexed to a lipid carrier.

Demonstration of wideεpread, high level CAT gene expression in in vivo after intravenous (iv) injection of pZN27 alone.

Plaεmid: pZN27.

DNA:Liposome Ratio: Plasmid DNA alone, without liposomes, was injected.

DNA doεe: 1 mg plaεmid DNA in 200 μl 5% dextroεe in water was injected times 2 over a 4 hour period by tail vein per mouse. Mice were εacrificed 24 hourε later and 17 different tiεεues were asεayed for CAT gene activity.

Mice: ICR, female, 25 gramε. Tissue extraction procedure: Each organ was homogenized in 0.3 ml of 0.25 M Tris-HCl pH 7.8, 5 mM EDTA, and the resulting extract was centrifuged and then subjected to 3 cycles of freeze-thaw and then treated to 65°C for 20 min. CAT assay procedure: The protein concentration of each tissue extract was quantitated using a ninhydrin- based protein assay (BioRad, Richmond) , and the same amount of total protein from each tissue extract waε added in the CAT assay, together with 10 μl of 20 mM acetyl CoA+12 μl of ¹ C-chloramphenicol (25 μCi/ml, 55 mCi/mmole, Amersham) , at 37°C for 13 hrs.

Resultε: An autoradiograph of thiε experiment iε εhown in 19A. Compared to control levels (lane 1) , iv injection of pZN27 alone produced highly significant levels of CAT gene expression in the following tisεueε: lung, thy uε, eεophaguε, heart, liver, spleen, stomach, small intestine, caecum, ovary, vagina, skeletal muεcle, pancreas and lymph nodes. Thuε, iv injection of the

pZN27 expresεion plasmid alone can efficiently transfect a very large and diverse number of tisεueε in the body.

Demonεtration of .widespread, high level CAT gene expreεsion in vivo after intravenous (iv) injection of pZN27 complexed to DDAB:Cholesterol (1:1) liposomes. Plasmid: pZN27.

Liposome: DDAB:Cholesterol=l: 1, stock 10 mM in 5% dextrose.

DNA:Liposome Ratio: liposome:plasmid 5 nmoles:1 μg

DNA dose: 100 μg plasmid DNA in 200 μl 5% dextrose in water was injected by tail vein per mouse. Mice were sacrificed 24 hourε later and 17 different tiεεueε were aεsayed for CAT gene activity. Mice: ICR, female, 25 grams.

Tissue extraction procedure: Each organ was homogenized in 0.3 ml of 0.25 M Tris-HCl pH 7.8, 5 mM EDTA, and the resulting extract was centrifuged and then subjected to 3 cycles of freeze-thaw and then treated to 65°C for 20 min.

CAT assay procedure: The protein concentration of each tissue extract was quantitated using a ninhydrin- baεed protein aεεay (BioRad, Richmond) , and εame amount of total protein from each tiεεue extract was added in the CAT asεay, together with 10 μl of 20 mM acetyl CoA+12 μl of ¹⁴ C-chloramphenicol (25 μCi/ml, 55 mCi/mmole, Amerεham) , at 37°C for 13 hrε.

Reεultε: An autoradiograph of thiε experiment is presented in Figure 19B. Compared to control levelε (lane 1), iv injection of pZN27:DDAB:Choleεterol complexes produced high levels of CAT gene expression in the following tissues: lung, thymus, eεophaguε, heart, liver, εpleen, εtomach, εmall inteεtine, large inteεtine, caecum, uterus, ovary, vagina, skeletal muscle, pancreas

and lymph nodes. Thus, iv injection of pZN27:DDAB:Cholesterol complexes can efficiently transfect a very large and diverse number of tiεsues in the body.

Example 19

IV Injection Of GM-CSF Expression Plasmid-Cationic Liposome

Complexes Produces Significant Antitumor Effectε.

Mouse: C57/black 6, female, 25 grams. Cancer: B16, mouse melanoma line which is highly metastatic to lung. The cell line was grown in RPMI-1640, 5% fetal calf serum. Liposome: DDAB:Cholesterol=l:1, 10 mM in 5% dextroεe in water. Twenty five thousand B-16 cells were injected iv by tail vein. Plasmid: pZN84 (the HCMV promoter enhancer fused to the murine GM-CSF coding sequence as shown in Figure 17) . Ratio: Cationic lipid:DNA=5 nmoles:1 μg.

A total of 100 μg of plasmid DNA in 200 μl of 5% dextrose in water administered per injection. Experimental outline: Mice in groups of eight received a single iv tail vein injection of 2.5 x 10 ⁴ B-16 melanoma cells. Group 1 received no treatment, group 2 received bi-weekly injections of 100 μg of pZN84 complexed to DDAB:Choleεterol lipoεomes, beginning 4 days prior to tumor cell injection and continuing for 2 weeks following tumor injection. All mice were sacrificed 3 weeks following tumor cell injection and surface lung tumor nodules which were black and macroscopic were counted using a diεsecting microscope.

Reεults :

Control animalε had 64.5 ± 19.7 (S.E.M.) nodules per lung. In contrast, the GM-CSF-liposome-treated animals had 11.9 ± 3.6 tumor nodules per lung (p < 0.01 5 versus the control group, as assessed by Student's t test) . Thus, iv injection of a cytokine gene produced a highly significant antitumor effect. This is the firεt demonεtration that iv injection of a gene can produce anti-tumor activity in vivo .

10 _. Example 20

Prolonged, high level murine GM-CSF gene expression in vivo after intravenous (iv) injection of pZN84 complexed to DDAB:Cholesterol (1:1) lipoεomes into goats.

15 Animals: Goats: female, 110 pounds Plasmid: pZN84

Liposome: DDAB:Cholesterol=l: 1, εtock lOmM in 5% dextroεe DNA:Liposome Ratio: Liposome:plasmid=5:1

20 DNA dose: 1.0 mg plasmid DNA in 5 ml 5% dextrose in water was injected by jugular vein per goat. Blood Sampling: Blood was drawn just prior to injection of pZN84 liposome complexes, and at 12, 24, 48, 72, 168 and 840 hours post-injection.

25 Murine GM-CSF ELISA assay procedure: Murine GM-CSF waε meaεured in theεe serum samples uεing a commercial murine GM-CSF ELISA kit provided by Endogen. There is no cross reactivity between goat and murine GM-CSF by ELISA. Results: As εhown in Figureε 22 and 23 (Goats A, B and

30 C) , as compared to control levels, iv injection of pZN84:DDAB:Cholesterol complexes produced high εuεtained circulating levelε of mouεe GM-CSF protein for at leaεt 168 hourε following injection. However, by 840 hrε (35 days) GM-CSF levels were subεtantially reduced (Figure

27, Goat A). No mouse GM-CSF protein was detected in the circulation of a goat which received 1 mg of the CAT gene (pZN51) complexed to DDAB:Choleεterol lipoεomeε by the identical iv injection protocol. The goat receiving the CAT gene did expreεε CAT antigen on the majority of its circulating white blood cells at 24 hours after injection, aε determined by immunoεtaining for CAT antigen. No CAT antigen waε detected in the white blood cellε of the goats which were injected with GM-CSF gene- liposome complexes. Thus, iv injection of pZN84:DDAB:Cholesterol complexes can produce high level expreεεion of the murine GM-CSF gene for prolonged periodε in goatε. IV injection- of pZN51:DDAB:Choleεterol complexeε can produce high levels of the CAT gene product in the majority of all circulating white blood cells in goats.

Example 21 Influence of liposome-GM-CSF plasmid complex on the courεe of experimental viral pneumonia in mice.

Endogenouε cytokines are believed to be included in the host response to a viral infection. It is therefore of interest to determine whether increased expression of a εingle cytokine, GM-CSF would influence the time course of viral pneumonia. The effect of delivering the murine GM-CSF plasmid-liposome complex via two different routes on the host response to a Sendai virus challenge is tested. The plasmid-lipoεome complex iε delivered iv to one group of mica, and intranasally (in) to a second group. Dose, time of delivery, virus challenge dose and endpoints are the same for both groups. Dose of plasmid-liposome complex is 100

μg/mouεe/dose. Doεing for both protocolε occurε on the following scale:

days 3, 7, 10 *PL - plasmid-lipoεome Each experimental group conεists of the following: murine GM-CSF plasmid-liposome + virus: 3 mice

CAT plasmid-lipoεome + viruε: 3 mice Untreated + viruε: 3 mice Thiε results in 9 mice per timepoint (27 total for each experiment, 54 total mice) . All animal procedures are performed via NIH and AAAI.AC guidelines including use of appropriate analgesia.

Virus challenge is administered intranasally with Sendai virus strain 771076, a strain known to be pathogenic in mice. Endpoints include the following (these have been shown to uεeful endpoints for Sendai pneumonia) :

1. lung weight vs. body weight and brain weight 2. lung histopathology

3. quantitative viruε recovery

4. Sendai εpecific antibody response

5. lung GM-CSF

6. lung TNF

Example 22 High level expresεion of the CAT gene in mouse brain produced by injection of DNA alone or DNA-cationic liposome complexeε directly into the central nervous system. Mice: ICR, female, 25 grams Plaεmid: pCIS-CAT Lipoεome: DOTMA:DOPE (1:1) For each mouse injected, 2.5 μg DNA was diluted in 5% dextrose, then mixed with liposomes diluted to the same volume in 5% dextrose. Five icroliterε were injected εtereotactically into the right ventricle of each mouse. Ratio: Plasmid:lipid = 1:0 (1 mg DNA: 0 μmoles DOTMA)

Plasmid: lipid = 1:1 (1 mg DNA: 1 μmoles DOTMA)

Plasmid: lipid = 1:3 (1 mg DNA: 3 μmoleε DOTMA) Plaεmid: lipid = 1:4 (1 g DNA: 4 μmoles

DOTMA)

Plasmid: lipid = 1:6 (1 mg DNA: 6 μmoles DOTMA)

Mice were sacrificed 48 hr. poεt injection. Brainε were removed and separated into left and right hemisphereε. Each hemisphere was homogenized in 250 μl 0.25 M Tris pH 7.8, 5 mM EDTA followed by three cycleε of freeze-thaw then 10 min. at 65°C.

The amountε of extractε uεed for CAT assay were normalized to protein levels. 0.3 ml of ¹⁴ C- chloramphenicol was used for each ssay. Aεsays were carried out at 37°C overnight. The reaction productε were εeparated on TLC plate and expoεed to film, as described in Example 4.

Reεultε: These resultε demonstrate that high level expression of a heterologous gene can be produced throughout the brain by injection of DNA-liposome complexes (at appropriate ratios) directly into the central nervouε εyεtem. They alεo demonstrate that injection of an expresεion plasmid alone can produce significant transgene expreεεion in the brain. The results are εhown in Figure 20.

Example 23 Demonεtration of CAT gene expression in the lung after intravenous (iv) injection of pRSV-CAT:L-PE:CEBA complexes.

Lipid carrier: L-PE:CEBA=1: 1, stock 20 mM in lipid carrier buffer. Plasmid: pRSV-CAT.

DNA:Lipid carrier Ratio: Lipid carrier:plasmid=l nanomole cationic lipid: 1 μg plasmid DNA.

DNA dose: 100 μg plasmid DNA in 200 μl 5% dextrose in water was injected by tail vein per mouse. Mice: Balb/c, female, 25 grams.

Tissue extraction procedure: Forty eight hours after tail vein injection, the animals were sacrificed, whole lung waε homogenized in 1ml of 0.25M Triε-HCl pH 7.8, 5mM EDTA, 80 μg/ml PMSF and the reεulting extract waε centrifuged and then the εupernatant waε subjected to 3 cycles of freeze-thaw and then heated to 65°C for 20 min..

CAT assay procedure: 100 μl of extract+10 μl of 20 mM acetyl CoA+4 μl of ¹⁴ C-chloramphenicol (25 μCi/ml, 55 mCi/mmole, Amersham) were incubated together at 37°C for 6 hr. At 3 hours, an additional 10 μl of acetyl CoA was added.

Results:

Thiε experiment εhowed that a significant level of CAT activity (indicative of expreεεion of the tranεgene) waε preεent in the lung of the animal injected with lipid carrier:DNA complexes, but not present in the lungs from control animals.

Example 24 Demonstration of CAT gene expression in multiple tissues after intravenous (iv) injection of PZN20-CAT:DDAB:DOPE complexes♦

Lipid carrier: DDAB:D0PE=1: 1, stock lOmM in 5% dextrose.

Plaεmid: pZN20.

DNA:Lipid carrier Ratio: Lipid carrier:plasmid=(A) 3 nmole cationic lipid: 1 μg plasmid DNA (SUV) . (B) 6 nmole cationic lipid: 1 μg plasmid DNA (MLV) .

DNA dose: 100 μg plaεmid DNA in 200 μl 5% dextroεe in water was injected by tail vein per mouse. Three mice each received this dose of MLV:pZN20 and 3 mice each this dose of SUV:pZN20.

Tisεue extraction procedure: Each organ was homogenized in 0.3 ml of 0.25 M Tris-HCl pH 7.8, 5 mM EDTA, and the resulting extract was centrifuged and then the supernatant was subjected to 3 cycles of freeze-thaw and then heated to 65°C for 20 min..

CAT assay procedure: The protein concentration of each tissue extract was quantitated using a Coomasεie Blue-baεed protein aεεay (BioRad, Richmond, CA) , and the same amount of total protein from each tisεue extract was added in the CAT asεay, together with 10 μl of 20 mM acetyl CoA+12μl of ¹⁴ C-chloramphenicol (25 μCi/ml, 55 mCi/mmole, Amerεham) , at 37°C for 13 hrs.

Results :

This experiment demonstrated that iv injection of pZN20:DDAB:DOPE complexeε gave significant levels of CAT gene expreεεion in each of 6 different tiεεueε including lung, heart, liver, εpleen, kidney and lymph nodeε.

Furthermore, MLV lipid carrierε mediated equal or higher levels of in vivo transgene expresεion than did SUV lipid carriers composed of the same lipids.

Example 25 Production of Vaccines

The subject invention may find general utility in the production of vaccine, in vivo , for a variety of diεeases. It has been shown that chimeric, genetically- engineered proteins of antigen fragments linked to granulocyte-macrophage colony-stimulating factor (GM- CSF) , augment antigen presentation and increase immune response to the antigen when injected in vivo (Tao and Levy, (1993) Nature 362:755-758. Thus a nucleic acid sequence coding for a particular antigen may be fused to the nucleic acid coding for GM-CSF and expressed in appropriate cells in vivo using this invention. This approach could dramatically improve the control of proliferation of malignant cells in the treatment of cancer. Other uses would include in vivo vaccine production against a variety of viral and other infectious diseases, particularly those diseaεeε in which an immune response haε been unεucceεεful or weak by conventional vaccine εtrategieε. Of particular intereεt would be thoεe diεeases, εuch aε HIV, which are characterized by proteins with hypervariable domains. Nucleic acid sequences encoding a variety of possible hypervariable domain sequences could be fused to GM-CSF and expreεεed εimultaneouεly in vivo to elicit an immune reεponse against many different εtrainε of the yiruε.

Example 26 Treatment of Tumor Proliferation Nucleic acid εequenceε εuch as those coding for protein 53 may be useful in treating tumor proliferation. Sequences coding for the tumor suppressor protein p53 may be expresεed intracellularly in vivo uεing appropriate expression vector-lipid complexes. Such complexes can be conveniently targeted to tumor cells by appropriate antibodies or ligands conjugated to the lipid complexes. Initial information regarding transfection of cells is developed uεing cultured cellε, such aε the tumor cell line K562. Once p53 gene expreεεion is verified in K562 cells and the optimal vector have been identified, the following in vivo anti-tumor experiment is conducted. SCID mice receive approximately 25 million K562 cells per mouse by intraperitoneal injection. Approximately three (3) weeks after injection, tumor-bearing mice in groups of ten (10) receive the following: no treatment/control group; p53 vector liposome complexes by intravenous injection; p53 vector liposome complexes by intraperitoneal injection; reported gene vector liposome complexes. The anti-tumor effects of the expresεion of p53 iε then be aεεeεεed based upon survival of the mice.

Example 27 Erythropoiesiε, the production of red blood cells, can cause continuously throughout the life span to offset cell destruction. Erythropoiesis iε a very preciεely controlled phyεiological mechanism and having sufficient numbers of red blood cells to be available in the blood for proper tisεue oxygenation, but not εo many that the cells would impede circulation. The formation of red blood cells occurε in the bone marrow and is under the control of the hormone erythropoietin. Because of erythropoietin is esεential in the proceεε of red blood

cellε formation, the hormone haε potential use for application in both the diagnosiε of the treatment of blood diεorderε characterized by low or defective red blood cell production. Gene therapy uεing the coding sequence for erythropoietin does find use in a variety of disease states , disorderε and εtateε of hematologic irregularity including anemia , in particularly correction of anemia of a type aεεociated with chronic renal failure and the like. A nucleic acid εequence coding for a polypeptide having EPO activity is inserted into an appropriate transcription or expression cassette and introduced into a host mammal aε naked DNA or complexed with an appropriate liquid carrier . Monitoring of the production of active EPO polypeptide can be performed aε described for example in USPN 4 , 703 , 008.

As shown by the above results , a plurality of tissues can be transformed following systemic administration of transgenes , either complexed to a lipid carrier or as naked nucleic acid. Expression of exogenous DNA following intravenous injection of a cationic lipid carrier/exogenous DNA complex into a mammalian host has been shown in multiple tissues, including T lymphocytes, metastatic tumors and intravascular tumor emboli. Expression of exogenous DNA in multiple different tissues, including those of the reticuloendothelial system has been obtained following intravenous injection of an expression plasmid as naked DNA. The ability to transfect T lymphocytes in vivo will have a dramatic impact on the treatment of AIDS, cancer, multiple sclerosis, and arthritis. In vivo transfection of cardiac endothelial cells will have a dramatic impact on the treatment of heart disease and heart attacks. In vivo transfection of lung cells will have a dramatic impact on the treatment of cystic fibrosis, asthma, emphysema and lung cancer. In vivo transfection of bone marrow cells will have a dramatic impact on the treatment of cancer, blood diseases and infections.

The in vivo gene therapy delivery technology as described above is non-toxic in animals and transgene expression has been shown to last for at least 60 days after a single admimstration. The transgene does not appear to integrate into host cell DNA at detectable levels in vivo as measured by Southern analysis, suggesting that this technique for gene therapy will not cause problems for the host mammal by altering the expression of normal cellular genes activating cancer-causing oncogenes, or turning off cancer- preventing tumor suppressor genes. Additionally, transgene expression after systemic admimstration of a DNA expression vector alone has been shown; transgene expression was produced in the lung for at least three weeks after a single administration of a DNA expression vector without a carrier system. Systematic injection of heterologous genes into adult animals can produce very high level transgene expression in a wide range of tissues, and transfect the majority (>70%) of all cells present in many of these tissues as demonstrated above. In contrast, previous studies attempting direct transfer of heterologous genes into adult animals have reported transfection limited to one or a few tissues, low levels of transgene expression in these tissues and (whenever histochemical analysis was included) transfection limited to less than 1% of the cells present in transfected tissues. In addition to transfecting the majority of all cells present in the lung, using the methods and constructs described above, high level transgene expression has been obtained in a wide variety of other tissues and cell types. These include:

Transfection of the majority of all cells present in the spleen and lymph nodes, including transfection of greater than 70% of all the T lymphocytes. The ability to efficiently transfect T lymphocytes in vivo permitting for the first time specific molecular approaches to both anti-HIV therapy and to selective modulation of the immune response.

Efficient transfection of visceral tumors and of intravascular tumor emboli after iv injection of DNA into tumor-bearing hosts. Previously, gene transfer studies involving cancer have been restricted solely to ex vivo approaches. Our work now 5 permits direct transfection with transgenes which provide tumor suppressor, anti-oncogene and/or anti-metastases activity within tumors in tumor-bearing hosts. Transfection of the majority of cardiac vascular endothelial cells, as well as bone marrow-derived hematopoietic cells, 0 ^" including the great majority of blast cells present in the bone marrow, by systemic delivery of heterologous genes. These results create dramatic and new means for controlling ischemic heart disease and hematopoiesis at the molecular level. 5 - Demonstration of the ability to produce high level in vivo expression of a variety of biologically important transgenes, as exemplified by human CFTR, IL-2 and GM-CSF. All publications and patent applications are herein incorporated by reference to the same extent as if each individual publication or patent o application was specifically and individually indicated to be incorporated by reference.

The invention now being fully described, it will be apparent to one of ordinary skill in the art that many changes and modifications can be made thereto without departing from the spirit or scope of the appended claims.