WO2002079470A1 - Disease-specific promoter chimeras for gene expression - Google Patents

Abstract

The present invention is directed to construction of novel gene delivery (expression) vectors and a method for producing disease-specific promoter chimeras sensitive to genotoxic stress by combining promoters, introns or other transcription regulatory sequences from disease-specific genes with viral gene promoter, intron or other transcription regulatory elements. The present invention also provides a method for producing chimeras of viral and cellular promoters, introns or other transcription regulatory sequences that have improved tissue-specific, targeting properties.

Description

PROMOTER CHIMERAS FOR GENE EXPRESSION HAVING VIRAL AND DISEASE-SPECIFIC PROMOTER ELEMENTS

FIELD OF THE INVENTION The present invention relates to the production of vectors having improved promoter chimeras composed of viral promoter gene sequences and disease-specific gene sequences that are sensitive to genotoxic stress. More particularly, an enhancer-like motif from the hMDRl gene promoter has been identified that is sensitive to genetic mutations that accumulate during tumor development. This same hMDRl promoter sequence is induced by treatment with cancer therapy such as radiation treatment and drug therapy. To improve the strength of gene expression in vivo or in vitro, the disease-specific hMDRl element has been combined in forward and reverse orientation with sequences from the human cytomegalovirus (HCMV) immediate early (IE) gene promoter, enhancer, and intron A (P/E/I). The hMDRl disease- specific element was introduced into multiple locations of the HCMV IE gene promoter and intron to improve gene expression and the targeting properties in vitro and in vivo to lung, heart and liver. Sequences not contributing to gene expression in the promoter chimera were deleted.

BACKGROUND OF THE INVENTION Gene therapy protocols and in vitro recombinant protein expression rely on incorporating a therapeutic, structural, or nonstructural gene into a vector composed of DNA sequences that that can be expressed in prokaryotic and/or eukaryotic cells. Such vectors must have regulatory sequences composed of promoters, enhancers, and introns responsible for directing the expression of the target gene (Manthorpe et al. (1993) Hum. Gene Ther. 4:419- 431). Most vectors currently used for production of therapeutic genes in vitro or in vivo have sequences from the promoter and/or intron A of the human cytomegalovirus immediate early gene (HCMV IE). The promoter from the HCMV IE gene is among the most powerful eukaryotic promoters, and both a short form of the promoter of approximately 480 base pairs (bp) (US Patent 5,385,839) and a long form of the promoter (US Patent 5,688,688) have been used for gene expression in vitro and in vivo. The long form of the HCMV promoter includes sequences from intron A from the HCMV IE gene. The addition of intron A improves eukaryotic gene expression in vitro and in vivo (Chapman (1991) Nucleic Acids Res 19:3979- 86; Hartikka et al. (1996) Hum. Gene Ther. 7:1205-17; and Liu et al. (1997) Nat. Biotechnol. 15:167-73). Other viral promoters can be used for therapeutic purposes such as the SV40 enhancer and promoter and the long terminal repeats from retroviruses. The HCMV IE promoter/enhancer/intron A (P/E/I) is preferred because the potency of target gene expression in vitro and in vivo is as much as 100- to 1, 000-fold greater than other viral promoters (Foecking et al. (1986) Gene 45:101-5; and Schmidt et al. (1990) Mol. Cell Biol. 10:4406-11). The subject optimized promoter chimeras having sequences unique to both the long and short form for the HCMN promoters now in use.

The activities resulting from combining different promoter elements can be affected by several factors. For example, there may be transcriptional interference that reduces the potency of gene expression, or there may be augmented expression, or the tissue-selectivity of the expressed gene may be altered either positively or negatively. Promoters designed to have targeting properties use enhancers from tissue-specific genes (Kuriyama et al. (1991) Cell. Sruct. Funct.l6:503-10; Chen et al. (1995) J. Clin. Invest. 96:2775-82; Harris et al. (1994) Gene Ther. 1:170-75; Osaki et al. (1994) Cancer Res. 54:5258-61; Ido et al. (1995) Cancer Res. 55:3105-9; Richards et al. (1995) Human Gene Therapy 6:881-3; and Siders et al. (1998) Cancer Gene Therapy 5:281-91). These enhancers have not proved successful, because of their weak transcription activity. Recently, the hexokinase II gene promoter and a cAMP- responsive promoter have been used to target gene delivery (Suzuki et al. (1997) Gene Therapy 4:1195-1201; Kabati et al. (1999) Human Gene Therapy 10:155-164). There are a few examples of promoters assembled from multiple genes (promoter chimeras) that are more potent than the HCMV promoter, but they are not chimeras of viral and disease-specific elements. By combining four tissue-specific enhancers from muscle, a promoter chimera was produced that is approximately 8-fold stronger in murine muscle in vivo than the HCMV IE promoter/enhancer (Li et al. (1999) Nat. Biotechnol. 17:241-5). A few tissue-specific sequences (300 to 3,000 bp) have been combined with the short form of the HCMV promoter/enhancer to produce chimeras that have improved properties of target gene expression in human cell lines in vitro. There is one example of combining a tissue-specific enhancer from muscle with the long form of the HCMV promoter/enhancer/intron A that caused a reduction in gene expression in muscle in vivo (Hartikka et al. (1996)). i another example, individually combining 19 different muscle specific enhancers with the HCMV promoter/enhancer identified only a single combination that improved gene expression in vivo (Barnhart et al. (1998) Hum. Gene Ther. 9:2545-53). The subject inventive promoter chimeras differ from those previously described by at least three different factors: altering the location of the inserted sequence, changing the HCMV IE sequence, and use of a disease-specific DNA element of less than 50-bp. Muscle-specific expression was also improved by combination of the α-actin and β-actin promoters with a CMV IE promoter/enhancer (Sawicki et al. (1998) Exp. Cell Res. 244:367-69; Hagstrom et al. (2000) Blood 95:2536-42). These examples of promoter combinations are with tissue-specific enhancers, not disease-specific enhancers.

The HCMV IE promoters and other viral promoters currently used for recombinant protein expression and gene therapeutics have little selectivity for expression in vivo (Schmidt et al. (1990) Mol. Cell Biol. 10:4406-11; and Baskar et al. (1996) J. Virol. 70:3207-14). Combining promoter elements from multiple genes is a broad strategy being applied to development of improved vectors for targeting gene therapeutics. There are tissue-specific antigens upregulated in certain forms of cancer. Addition of large regions of promoter DNA sequences from such tumor antigens (up to 1.2 kilobases) to the HCMV promoter has recently been attempted with improvement in gene expression. Combining approximately 3,000 bp from the promoter of one of several forms of carcinoembryonic antigen upregulated in colon carcinoma, increased gene expression in vitro approximately 5-fold. Combination of prostate- specific antigen sequences with HCMV sequences produced vectors having increased expression in vitro, but no improvement in tissue-specificity (Lantham, 2000). Combining HCMV promoter sequences and 19 muscle-specific promoter elements identified a single combination active in vivo. Each of these chimeras introduced the cellular promoter element 5' to the HCMV promoter. One of the objectives of the present invention is to develop a rational method for altering the location of the disease-specific promoter sequence relative to the HCMV P/E/I and deleting portions of the HCMV P/E/I that either do not contribute to gene expression in a target tissue or negatively affect gene expression in said tissue. Use of this method can improve detection of cooperative interactions between DNA sequences in the promoters from different genes. Promoter chimeras of minimal size can be produced having both superior properties of gene expression in said tissue in vitro and superior targeting properties in vivo. Cancer is caused by multiple genetic mutations in key regulatory genes. Some of these genes encode for growth suppressors, and their mutation leads to genetic alterations in cells (Vogelstein et al. (1992) Cell 70:523-26; Levine et al. (1993) Ann. Rev. Biochem. 62:623-51; and Cox et al. (1995) Bioassays 17:501). One such tumor suppressor, the p53 protein, encodes a transcription factor that regulates the expression of several downstream target genes by binding to specific DNA sequences in the promoters of these genes (Funk et al. (1992)

Bioassays 17:501). The sequence-specific DNA element recognized by p53 is approximately 9 bp in length. The p53 tumor suppressor is composed of three unique protein domains, an amino terminal transcription activation domain, a central DNA binding domain, and a carboxyl regulatory domain responsible for tetramerization of the protein (Hupp et al. (1994) Curr. Biol. 4:865-75). Approximately 50% of human cancers have a mutation in the DNA binding domain of p53 that causes the protein to lose its ability to bind to specific DNA sequences and activate expression of downstream genes. These genes are down-regulated in cells having mutant forms of p53. A small cluster of genes is activated when p53 mutations accumulate in cells, including the hMDRl gene, the vascular endothelial growth factor (VEGF) gene, lipoxegenase gene, c- myc oncogene, and others (Dittmer et al. (1993) Nature Genetics 4:42-6; and Frazier et al. (1998) Mol. Cell Biol. 18:3735-43). The hMDRl gene encodes P-glycoprotein, an ATP- dependent, multidrug efflux pump, that, when overexpressed, causes chemical multidrug resistance (Roninson (1992) Biochem. Pharmacol. 43:95-102). Transcription regulatory sequences from the hMDRl gene are sensitive to activation by genotoxic stresses such as oncogene and tumor suppressor mutations. Transcription regulatory DNA sequences sensitive to genotoxic stress are disease-specific elements, and these are biochemically distinct from tissue-specific elements or enhancers. Certain viral gene promoters also have disease-specific promoter sequences and are sensitive to mutations of tumor suppressor genes such as p53 (Debs et al. (1992) J. Virol. 66:6164-70; and Subler et al. (1994) J. Virol. 68: 103-110) but the precise sequences responsible have not been identified. Synthetic templates composed of sequence-specific DNA elements such as those that bind the transcription factors Spl, TBP, NF-kβ, ATF, and CREB are also sensitive to transcriptional activation by mutant p53 in an experimental model. The mechanisms responsible for gene activation by mutant ρ53 are not understood. Mutant p53 has lost its capacity to bind specifically to DNA, so that indirect mechanisms dependent on cellular factors present in the cells must operate. Genes activated in cells having the mutant forms of the p53 tumor suppressor represent disease-specific genes that are sensitive to genetic alterations in cells. The hMDRl gene is an example exploited herein for production of improved promoter chimeras having disease-specific DNA elements. Changes in the expression of such genes can be mediated by the promoter for that gene or by altering the stability of the protein or the messenger RNA for that gene product. The present invention relates to identification of disease-specific gene promoter DNA sequences sensitive to genotoxic stress such as mutations and incorporating such DNA sequences into viral promoter and intron sequences to improve tissue-specific and tumor-specific targeting of vectors for therapeutic gene delivery. Another embodiment of the invention is that the orientation of the disease-specific elements can be altered to improve their general or differential expression properties. The hMDRl gene promoter is sensitive to conditions of genotoxic stress. It is activated by mutant p53, by chemotherapeutic drug treatments such as vinblastine and etoposide, by radiation treatment, by activated oncogenes such as raf and ras, and by chemical tumor promoters such as phorbol myristate acetate (Chin et al. (1992) Science 255:459-62; Cornwell et al. (1993) J. Biol. Chem. 268:19505-11; Dittmer et al. (1993) Nature Genetics 4:42-46; Miltenberger et al. (1995) Cell Growth Differ. 6:549-56; Nguyen et al. (1994) Oncol. Res. 6:71-77; and McCoy et al. (1995) Mol. Cell. Biol. 15:6100-08). There are three elements in the proximal hMDRl promoter responsible for sensitivity to genotoxic stress, a closely linked Y-box and two overlapping GC-boxes spanning the promoter -88 to -36 relative to the transcription start site (Cornwell et al. (1993); Sundseth et al. (1997); Goldsmith et al. (1993) J. Biol. Chem. (1993) 268:5856-60; and Jin et al. (1998) Mol. Cell Biol. 18:4377-84). Induction of the hMDRl gene by chemotherapeutic drugs depends on the Y-box (Asakuno et al. (1994) Biochem. Biophys. Res. Commun. 199:1428-35; and Hu et al. (2000) J. Biol. Chem. 275:2979- 85). Induction of the hMDRl gene by chemical tumor promoters or the oncogene raf ^'depends on the overlapping GC-boxes (-66 to-36). Induction of the hMDRl gene by radiation depends on both of these elements (Hu et al. (2000)). This region of the hMDRl promoter has been described as a constitutive element responsible for basal expression of the hMDRl gene (Jin et al. (1998)). The hMDRl gene is not constitutively expressed in all tissues, but it is a gene sensitive to genotoxic stress such as those responsible for tumor development. The Y- and overlapping GC-boxes of the hMDRl promoter represent a disease-specific promoter element. It functions autonomously in tumor cell lines that overexpress the hMDRl gene (Sundseth et al. (1997)). A mutant form of a larger region of the hMDRl promoter was shown to be inducible by drug treatment (Walther et al. (1997) Gene Ther. 4:544-52). As is discussed herein, it has been found that the 52 bp from the proximal hMDRl promoter is equipotent to the larger form (Sundseth et al. (1997)) described by Walther et al. (1997). It is also approximately as potent as the SV40 promoter and enhancer (Sundseth et al. (1997)), making it much weaker than the HCMV IE gene promoter/enhancer/intron A. The present invention combines properties of the 52-bp element from the hMDRl gene promoter and the HCMV promoter/enhancer/intron A. The DNA sequences of both elements are further modified in this invention when combined to produce novel promoter chimeras.

In tumor cells that overexpress hMDRl, two nuclear factors interact with the Y- and GC-boxes, namely NF-Y and Spl. NF-Y binds to the Y-box element (Sundseth et al. (1997); Jin et al. (1998)), and Spl binds only to the penultimate GC-box (Comwell et al. (1993); Sundseth et al. (1997)). Induction of the hMDRl gene by the 7-α/oncogene requires both overlapping GC-boxes (Miltenberger et al. (1995)), and Egrl binds to this motif. In some cell types, the proteins that bind to the hMDRl disease-specific DNA sequences, NF-Y and Spl, exist as heterodimers (Roder et al. (1999) Gene 234:61-69), and Y and GC elements can cooperatively regulate gene expression (Yamada et al. (2000) J. Biol. Chem. 275:18129-137. There are fewer than twenty genes known to have linked Y and GC elements (Sundseth et al. (1997)). Cooperative interactions between DNA promoter sequences do not require that the DNA sequence elements be closely linked, and cooperative interactions can take place over very long stretches of DNA. The disease-specific DNA promoter element from the hMDRl gene spanning -88 to -36 relative to the transcription start site differs from DNA sequences from tumor antigens overexpressed in cells and used previously to modify the 5' region of the HCMV promoter/enhancer. It is sensitive to genotoxic stress such as DNA mutation of tumor suppressors or oncogenes, and it is responsible for hMDRl gene activation by drug and radiation treatment. The disease-specific element is inducible and significantly smaller than promoters from tumor antigens used to modify the HCMV promoter previously (300 to 3,000 bp).

Producing unnatural promoter chimeras is a superior method of inducing targeting properties into vectors for gene delivery. Such promoter chimeras will rely on activation or repression by cellular factors that exist in the target tissue, cell, or tumor. Bulky proteins used for targeting such as antibodies, even when humanized, are immunogenic (Kurane et al. (1998) Jpn. J. Cancer Res. 89: 1212-19). Altering viral tropism will also likely lead to immune inactivation of the vector. Thus, producing unnatural promoter chimeras of viral and disease- specific DNA sequences yields regulatory elements that depend on host factors. Introduction of such promoter chimeras into vectors will reduce potentially toxic immune responses to gene therapeutics. One objective of the present invention is to improve both the potency and targeting properties of vectors used to deliver a therapeutic gene in vivo or to express therapeutic proteins in vitro.

Yet another objective of the invention is to produce promoter chimeras that are induced by chemotherapy or radiation treatment. Another objective of the present invention is to produce promoter chimeras between viral and disease-specific promoter sequences that are sensitive to genetic mutations that occur during cancer development.

One objective of this invention is to combine portions of the HCMV IE gene P/E/I with an element from the human multidrug resistance gene (hMDRl) promoter (-88 to -36) that is sensitive to mutations in oncogenes and tumor suppressors, and is induced by genotoxic stresses such as drug or radiation treatment.

Another objective of the present invention is to produce chimeras from viral DNA sequences and disease-specific DNA sequences such as that from the hMDRl gene promoter to produce a promoter chimera that is more potent than either DNA element alone.

Another objective of the present invention is to produce promoter chimeras between viral gene promoter sequences such as those from the HCMV IE gene promoter/enhancer/intron A and disease-specific promoter sequences such as that from the hMDRl gene promoter sequence (-88 to -36) that have improved properties of gene expression in vitro and in vivo.

Another objective of the present invention is to produce promoter chimeras between the HCMV promoter/enhancer/intron A and the hMDRl promoter element (-88 to -36) that display improved targeting properties in vivo subsequent to systemic injection with cationic lipids in lung, heart, and liver. Another objective of the present invention is to provide a rational method for production of promoter chimeras having disease-specific properties that can be exploited for therapeutic gene delivery in disease.

Another objective of the present invention is to provide a method for optimizing gene expression in target tissues by producing vectors having unnatural promoter chimeras composed of viral DNA sequences and disease-specific DNA sequences. DNA sequences that do not contribute to gene expression in said tissue are eliminated.

Another objective of the present invention is to provide unique viral promoters by removing DNA sequences from the HCMV IE gene promoter/enhancer/intron A that do not contribute to gene transcription in vitro. Another objective of the present invention is to alter the orientation of the positive viral and disease-specific DNA sequences relative to the transcription start site to optimize expression in a target tissue.

Another objective of the present invention is to produce promoter chimeras between viral and disease-specific DNA sequences that are regulated in ways that are unlike the native viral or disease-specific promoters. By producing such promoter chimeras, viral DNA sequences can replace portions of the disease-specific DNA sequences and improve the properties of gene expression in vitro and in vivo.

Yet another objective of the present invention is to identify the DNA sequences from the hMDRl promoter element (-88 to -36) responsible for augmenting the activity of the HCMV IE gene promoter/enhancer/intron A by mutation of sequence-specific DNA elements responsible for transcription factor binding.

Another objective of the present invention is to identify enhancer-like properties of the disease-specific sequence identified from the hMDRl promoter (-88 to -36) by showing that it operates in both forward and reverse orientation.

Another objective of the present invention is to produce a promoter chimera having multiple copies of the hMDRl gene promoter DNA sequence (-88 to -36) and a single copy of the HCMV IE gene promoter/enhancer/intron A to alter the targeting properties of the promoter chimera. Yet another objective of the present invention is to mutate the overlapping GC-boxes from the disease-specific hMDRl gene DNA sequence (-88 to -36) to show that they are unnecessary for the activity of a promoter chimera between the HCMV IE gene promoter/enhancer/intron A and the hMDRl DNA sequence.

Another objective of the present invention is to show that the Y-box within the hMDRl gene promoter DNA sequence (-88 to -36) is solely responsible for augmenting the activity of the HCMV IE gene promoter/enhancer/intron A in the context of the promoter chimera.

Yet another objective of the present invention is to define a method for combining promoter elements from a viral promoter and disease-specific DNA promoter, enhancer, and intron sequences to produce unique combinations of promoter elements that have improved tissue- specific properties of gene expression when introduced into a vector and delivered in vivo with a cationic lipid formulation.

SUMMARY OF THE INVENTION

The present invention provides novel promoter chimeras that can be used to construct vectors for the expression of therapeutically-relevant gene products in mammalian cells. The promoter chimeras comprise sequences from (or derived from) viral promoters, introns or other transcription regulatory sequences, and sequences from (or derived from) disease-specific promoters, introns or other transcription regulatory sequences. By disease-specific promoter, intron or other transcription regulatory sequences is meant those promoter, intron or other transcription regulatory sequences that are sensitive to genetic aberrations that occur in disease, such as mutation, gene rearrangement, chemical modification, or gene deletion, and also those promoter, intron or other transcription regulatory sequences that are activated by conditions of genotoxic stress such as radiation or drug treatment. In preferred embodiments, the promoter chimeras comprise disease-specific promoter, intron or other transcription regulatory sequences that are approximately 100 bp or less in size. The viral promoter, intron or transcription regulatory sequences and the disease specific promoter, intron or transcription regulatory sequences can be present as single or multiple copies (including multimers); in forward or reverse orientation; and can be mutated relative to the wild-type viral promoter, intron or other transcription regulatory sequence and wild-type disease-specific promoter, intron or other transcription regulatory sequence. Furthermore, the disease specific promoter, intron or other transcription regulatory sequences in the promoter chimera can be located at any site relative to the viral promoter, intron or other transcription regulatory sequences; e.g., 5' or 3' of or within the viral promoter, intron or regulatory sequence. As is illustrated herein, transcription regulatory sequences, whether from the promoter, intron or elsewhere, can be defined by mutation or deletion of those sequences in chimera promoter constructs operably linked to a reporter gene.

The instant invention provides promoter chimeras that can be induced by chemotherapy and/or radiation treatment, as well as promoter chimeras that are sensitive to genetic mutations that occur during cancer development. Also provided herein are promoter chimeras between viral and disease-specific DNA sequences that are regulated in ways that are unlike the native viral or disease-specific promoters. By producing such promoter chimeras, viral DNA sequences can replace portions of the disease-specific DNA sequences and improve the properties of gene expression in vitro and in vivo. In preferred embodiments, the promoter chimeras of the instant invention comprise sequence from (or derived from) -88 to -36, -137 to +158, or -5,000 to +250 of the human MDR1 gene, wherein said sequence from (or derived from) the hMDRl promoter is activated or repressed by mutation of the p53 tumor suppressor, mutation of the raf oncogene, radiation treatment, and drug treatments, and further comprising sequences from (or derived from) the HCMV IE gene P/E/I. The liMDRl sequence in the promoter chimera can be located at any site relative to the P/E/I, e.g., 5' of the P/E/I, 3' of the P/E/I, or within the P/E/I. Furthermore, the hMDRl promoter sequence can be present as a single copy, or in multiple copies (including multimers); in forward or reverse orientation; and mutated relative to the wild-type promoter sequence. ("Multimerize" or variations thereon as used herein means duplication of sequences.) Similarly, the P/E/I components of the HCMV IE gene can be present in multiple copies, in forward or reverse orientation, and mutated relative to the wild-type HCMV IE gene sequence. By mutated, it is meant that a sequence contains base substitutions relative to its wild-type counterpart. The present invention demonstrates that the disease-specific sequence identified from the hMDRl promoter (-88 to -36) has enhancer-like properties because it operates in both forward and reverse orientation. The present invention also identifies the DNA sequences from the hMDRl promoter element (-88 to -36) that are responsible for augmenting the activity of the HCMV IE gene promoter/enhancer/intron A by mutation of sequence-specific DNA elements responsible for transcription factor binding.

The present invention demonstrates that the overlapping GC-boxes from the disease- specific hMDRl gene DNA sequence (-88 to -36) are unnecessary for the activity of a promoter chimera between the HCMV IE gene promoter/enhancer/intron A and the hMDRl DNA sequence. Furthermore, it is demonstrated herein that the Y-box within the hMDRl gene promoter DNA sequence (-88 to -36) is solely responsible for augmenting the activity of the HCMV IE gene promoter/enhancer/intron A in the context of the promoter chimera.

The present invention provides promoter chimeras for the expression of gene(s) encoding structural or non-structural products with therapeutic relevance to mammalian disease. The vectors of the instant invention comprise the aforementioned promoter chimeras operably linked to genes with therapeutic relevance, and further comprise sequences for replication in eukaryotic and prokaryotic host cells.

The promoter chimeras and expression vectors of the instant invention have a number of desirable properties that make them superior to prior art promoter chimeras. Specifically, the HCMV IE-hMDRl promoter chimeras are more potent than either promoter standing alone. Similarly, relative to either promoter standing alone, the HCMV IE-hMDRl promoter chimeras display improved targeting properties in vivo in lungs, heart and liver following systemic injection with cationic lipids. Futhermore, the HCMV IE promoter in a number of the promoter chimeras provided herein is only approximately one half of the size of the original HCMV IE promoter. The elimination of unnecessary sequence from the HCMV IE promoter greatly contributes to the utility of the promoter chimeras in gene therapy applications.

The present invention also provides a method for combining promoter elements from viral and disease-specific DNA promoter, enhancer, and intron sequences to produce unique combinations of promoter elements that have improved tissue-specific properties of gene expression when introduced into a vector, the method comprising the steps: a) identifying a human gene undergoing disease-specific transcriptional activation; b) identifying promoter and/or enhancer and/or intron sequences or other transcription regulatory sequence from said human gene that at least partially responsible for said disease-specific transcriptional activation; c) preparing a plurality of promoter chimeras comprising: (i) one or more copies in forward or reverse orientation of sequences from a viral promoter and/or enhancer and/or intron or other transcription regulatory sequence; (ii) one or more copies in forward or reverse orientations of said human disease- specific sequences; wherein said sequences in (ii) are located at a variety of positions in said promoter chimera relative to said sequences in (i); d) preparing vectors for the expression of said therapeutically relevant gene product, each vector comprising a promoter chimera of c) operably linked to said therapeutically relevant gene product, with or without sequences necessary for replication; e) optionally mutating said human disease-specific sequences and/or said viral sequences in said vectors; f) selecting vectors from d) and e) that have a desirable expression level and/or expression site of said therapeutically relevant gene product; and g) optionally minimizing the size of the vectors selected as having desirable expression levels and/or sites in step f), wherein said minimized vector substantially retains a desirable expression level and/or site, and wherein minimization is achieved by deleting portions of said sequence from said human sequences, and/or said viral sequences.

BRIEF DESCRIPTION OF THE FIGURES Figure 1A is a diagram of three promoter chimeras having different orientations of the hMDRl element. The 52-bp hMDRl promoter element (denoted by triangle) was cloned 5' to HCMV IE P/E (Cl), in between HCMV IE P/E and intron A (C2), or 3' to intron A (C3). The gray box denotes the luciferase reporter. Figure IB illustrates reporter activity (mean +/- SEM) of chimeric constructs (Cl, C2, C3) and the parental vector having only HCMV IE promoter and intron sequences in transiently transfected HCT 116, KB, and Hep G2 cells. Unmodified HCMV IE promoter activity in each cell line was set to one, and the promoter activities of Cl, C2, and C3 are displayed as fold activity of the HCMV IE promoter alone. Absolute luciferase values (in relative light units) were 4.5 x 10(5) for HCT 116 cells, 6.8 x 10(3) for KB cells, and 1.5 x 10(4) for Hep G2 cells. * p<0.05 vs. VR1255, | P<0.05 vs. Cl and C2.

Figure 2 summarizes a mutation analysis of the hMDRl disease-specific element in the promoter chimera C3. Constructs are shown on the left and described in detail in the text. Promoter activities in transiently transfected HCT 116, KB, and Hep G2 cells are shown on the right. Data are presented as mean ± SEM. C.31 activity in each cell line was set to one, and the promoter activity of deletion constructs are displayed as fold C3 activity. * p<0.05 vs. C3. a = p<0.05 vs. C3 when transfections analyzed individually, but p>0.05 when pooled data analyzed. This is related to a high variability in absolute luciferase values obtained from individual KB cell transfections.

Figure 3 summarizes deletion analysis of HCMV IE promoter in the promoter chimera. Constructs are shown on the left, and relative promoter activities in transiently transfected HCT 116, KB, and Hep G2 cells are shown on the right. Data are presented as mean ± SEM. C3 activity in each cell line was set to one and the promoter activity of deletion constructs are displayed as fold C3 activity. * p<0.05 vs. the next larger construct.

Figure 4 shows how 5' and 3' intron A sequences contribute to C3 promoter activity. Constructs are shown on the left and relative promoter activities in transiently transfected HCT 116, KB, and Hep G2 cells are shown on the right. Data are presented as mean ± SEM. C3 activity in each cell line was set to one and the promoter activity of deletion constructs are displayed as fold C3 activity. * p<0.05 vs. C3.

Figure 5 shows construction of a minimized chimeric promoter. Constructs are shown on the left, and relative promoter activity in transiently transfected HCT 116, KB, and Hep G2 cells is shown on the right. Data are presented as mean ± SEM. C3 activity in each cell line was set to one, and the promoter activity of deletion constructs are displayed as fold C3 activity. * p<0.05 vs. C3. f p<0.05 vs. DEL-13. a =p<0.05 vs. C3 when transfections analyzed individually, but p>0.05 when pooled data analyzed. This is related to a high variability in absolute luciferase values obtained from individual KB cell transfections.

Figure 6 illustrates the luciferase expression in lung, liver, colon and heart at 24 hours after injection of four plasmids, i.e., the parent plasmid having only the HCMV IE PEI, and the Cl, C2 and C3 liMDRl/HCMN chimeras, into the tail veins of mice using a formulation with the cationic lipid DOTAP.

Figure 7 is a diagram of the parent vector used in the Examples. DETAILED DESCRIPTION OF THE INVENTION

A malignant phenotype requires the acquisition of multiple mutations in growth regulatory pathways, signal transduction pathways, or DNA repair systems. The inactivation of such regulatory systems provides numerous potential targets for cancer gene therapy. Replacement of mutated genes represents one therapeutic option for cancer gene therapy (Lesoon-Wood et al. (1995) Hum Gene Ther. 6:395-405; Roth et al. (1996) Nature Med. 2:985-91; Kennedy et al. (1997) Adv. Drug Deliv. Rev. 26:119-133). Cancer gene therapy vectors need to target malignant cells and subsequently drive high-level, long lasting therapeutic gene expression (Dachs et al. (1997) Oncol. Res. 9:313-25; and Gomez-Navarro et al. (1999) Eur. J. Cancer 35:2039-57). Such characteristics are related to both the gene delivery system (viral vs. non-viral) and the promoter used to drive gene expression (Somia et al. (2000) Nature Rev. Genetics 1:91-99).

There is currently a need to improve promoters used in vectors designed for gene therapy, vaccines, or recombinant protein or nucleic acid expression that can drive high-level, tissue- or tumor-specific expression (Gomez-Navarro et al. (1999)). Reducing promoters in size without adversely affecting their activity would represent an important design improvement that could facilitate combination with tumor-specific or regulatory promoter elements and increase the size of the therapeutic gene that could be introduced into a given vector such as adeno-associated virus vectors. Thus, there are at least two aspects of tumor targeting: delivery to the target tissue and selective expression in the tumor tissue. Delivery to the target tumor tissue can be accomplished in vivo by intravenous injection of lipid-formulated, plasmid vectors (Logan et al. (1995) Gene Ther. 2:38-49; Brigham et al. (1989) Am. J. Med. Sci. 296:278-81; and Liu et al. (1995) J. Biol. Chem. 27:24864-70). Incorporation of HCMV IE promoter/enhancer into plasmid vectors improves systemic gene delivery to lung when complexed to cationic lipid

(Liu et al. (1995); and Liu et al. (1997) Nat. Biotechnol. 15:167-73). According to the subject invention, selective expression of the vector in target tumor tissue can be accomplished by a promoter from a disease-specific gene (e.g., hMDRl gene) that is activated during genotoxic stress such as chemotherapy, radiation therapy, mutation (e.g., p53 mutation), and the like. The subject invention comprises promoter chimera obtained by combining a strong viral promoter (e.g., HCMV IE gene promoter) and a disease specific promoter (e.g., hMDRl promoter). The subject invention also comprises a systematic method of defining the essential sequences of the viral promoter and the disease-specific promoter necessary to enhance tissue selectivity without reducing expression. Defining the essential sequences of the viral and disease-specific promoters involves eliminating or mutating those sequences that do not contribute to expression in the target tissue or do not contribute to disease-specific expression. The object of the subject invention is to reduce the size of the promoter chimera, while simultaneously enhancing expression and maintaining expression of the downstream therapeutic gene. Another object of this invention is to introduce a difference in expression in normal and disease tissue. Yet another object is to improve the targeting properties or differential expression of vectors to specific tissues for therapeutic purposes. Another object of this invention is to improve expression of recombinant materials in cultured cells.

The hMDRl gene proximal promoter (-88 to -36) has been observed to function autonomously in tumor cells that overexpress the hMDRl gene (Sundseth et al. (1997) Mol. Pharmacol. 51:963-71). Interestingly, the hMDRl element likely contains no tissue-specific enhancers. Instead, it appears to define a region of the hMDRl promoter that is sensitive to genotoxic stress (Hu et al. (2000) J. Biol. Chem. 275:2979-85; Asakuno et al. (1994) Biochem. Biophys. Res. Commun. 199:1428-38; Cornwell et al. (1993) J. Biol. Chem. 268:19505-11; McCoy et al. (1995)Mol. Cell. Biol. 15:6100-08; Miltenberger et al. (1995) Cell Growth Differ. 6:549-56; andUchiumi et al. (1993) Cell Growth Differ. 4:147-57).

As described in the Examples, the hMDRl element (-88 to -36) was combined with HCMV immediate early (IE) promoter/enhancer/intron A (PEI) to make a variety of promoter chimeras that were then tested for in vitro and in vivo activity. As is exemplified herein, five primary findings emerge from the in vitro studies. First, a small human promoter element from the hMDRl gene can augment expression of the HCMV IE P/E-intron A in vitro in multiple orientations. Second, the degree of augmentation of HCMV IE promoter activity by hMDRl sequences depends on the cell type and the position and spacing of the hMDRl element with respect to HCMV IE P/E sequences. This invention produces promoter chimeras sensitive to gene activation or repression in disease, and the hMDRl gene element has been used as an example for production of improved promoter chimeras between viral and cellular promoters. Other disease-specific elements can also be exploited to create promoter chimeras with viral sequences. Third, HCMV IE P/E-intron A sequences can be deleted without reducing promoter chimera activity in vitro. Fourth, intron A is required to maintain the activity of the promoter chimera. Fifth, HCMV IE promoter and intron sequences can be deleted without affecting promoter activity in a form that does not contain the HMDRl disease-specific element. These inventions provide a method for improving gene expression from the HCMV IE promoter alone and in combination with a disease-specific DNA element. In all cases, each promoter chimera has a unique DNA sequence not described in the scientific literature and not existing in nature.

It appears that only the Y-box mediates the increased promoter activity of the C3 chimeric construct (Figure 3), while both the Y-box and the penultimate GC-box are required for activity in the hMDRl promoter (Sundseth et al. (1997)). It is believed that this is the first example in which a single, simple human promoter element has been effectively combined with the HCMV IE P/E-intron A to create a more potent expression vector. Because sequences contained within the proximal hMDRl promoter also mediate induction by genotoxic stress such as ultraviolet radiation and chemotherapeutic drugs (Asakano et al. (1994) Biochem. Biophys. Res. Commun. 199:1428-35; Nguyen et al. (1994) Oncol. Res. 6:71-77; and Hu et al. (2000) J. Biol. Chem. 275:2979-85), the subject chimeric vectors containing the hMDRl element can be induced by these stimuli.

Two other interesting observations related to the subject chimeric constructs were that (1) the location of the hMDRl element relative to the HCMV IE sequences was important (Figure 1) and (2) multimerization of hMDRl sequences failed to augment chimeric promoter activity (Figure 3). Interactions between hMDRl and HCMV IE sequences depend on the location of the hMDRl element relative to the transcription start site of the vector. Although the specific HCMV IE elements that interact with the hMDRl element were not defined, HCMV IE sequences remaining in construct DEL- 13 (-584 to -349 and -243 to +72) contain putative binding sites for many transcription factors including AP-1, AP-2, CREB/ATF,

NF-kappaB, and Spl (Jeang et al. (1987) J. Virol. 61:1559-70; Sambucetti et al. (1989) EMBO J 8:4251-58; and Hunninghake et al. (1989) J. Virol. 63:3026-33). Thus, the factors that bind the hMDRl element may interact with some of these proteins. Multimerization of the hMDRl element decreased chimeric promoter activity and was the converse of what we expected. Creating multiple copies of defined elements in the promoter chimeras may provide a means to differentially target disease and normal tissues.

The hMDRl element seems to have enhancer-like characteristics in the promoter chimera; it is active both in reverse orientation (C3mut5, Figure 3) and when placed in three different locations relative to the HCMV IE P/E-intron A (Figure 1). In addition, promoter activity decreases as more copies of the hMDRl element are added (i.e., the promoter activity of C3 > C3dimer > C3trimer > C3tetramer, data not shown). Mechanistically, this effect does not seem to be related to spacing between the hMDRl element and downstream sequences, as introduction of a longer element from the hMDRl gene promoter did not reduce promoter activity to the same extent (data not shown). It is possible that multiple hMDRl elements cause nucleation of protein complexes that interfere with efficient transcription originating from the transcription start site. Alternatively, construction of multiples of the hMDRl element may induce repression. Both NF-Y and Spl are required for hMDRl activation by genotoxic stress in some cells (Hu et al. (2000)) and can exist as heterodimers (Roder et al. (1999) Gene 234:61-69; and Yamada et al. (2000) J. Biol. Chem. 275:18129-37). It seems unlikely that Spl acts through the 52-bp element in the promoter chimera. The GC-boxes are dispensable, and the hMDRl element clearly functions in a different way in combination with HCMV IE P/E-intron A. Other proteins can bind to Y-boxes (Carett et al. (2000) J. Mol. Biol. 302:539-52) and cooperative interactions between Y-box and cis-acting elements from the HCMV IE promoter could be inhibited by multimerization.

The HCMV IE P/E-intron A has been widely used in gene therapy vectors because of , its ability to drive high-level expression in numerous tissues and cell lines (Foecking et al. (1986) Gene 45:101-5; and Schmidt et al. (1990) Mol. Cell. Biol. 10:4406-11). There is little published work involving modification of this promoter. There are several potential advantages to a minimized HCMV IE P/E-intron A that retains full promoter activity. Most importantly, a minimized promoter allows for the insertion of larger therapeutic genes. This is particularly critical when using viral delivery systems with limited genetic packaging capacities like adeno-associated virus (Hermonat et al. (1997) FEBS Lett 407:78-84). A reduction or simplification of HCMV IE sequences can facilitate combination with tissue-specific or regulatory promoter elements, leading to the production of promoter chimeras that can more effectively target a given tissue or tumor. Altering the DNA sequence of the HCMV IE promoter or promoter chimeras of viral and disease-specific DNA sequences may cause changes in biological factors that bind to DNA and regulate transcription. These changes may be responsible for targeting or differential expression by these promoter chimeras. The biochemical and physical mechanisms for the improved properties of differential gene expression of said promoter chimeras described in this invention have not been defined. This invention provides a method for construction of said promoter chimeras without bias of mechanism of action for gene activation or repression.

Gene therapy vectors can have vastly different in vivo and in vitro activities (Barnhart et al. (1998) Hum. Gene Ther. 9:2545-53; and Li et al. (1999) Nat. Biotechnol. 17:241-45). Thus, it is imperative to evaluate vectors using in vivo model systems as well as traditional in vitro assays. In vivo models are more difficult to establish and may not adequately represent clinical malignancies (Winograd et al. (1987) In Vivo 1:1-13; Bepler et al. (1990) In Vivo 4:309-15; Gould, M (1995) Semin. Cancer Biol. 6:309-15; Manthorpe et al. (1993) Hum Gene Ther. 4:419-31). The promoter chimeras of the instant invention retain activity in vivo. In summary, the examples demonstrate that the HCMV IE P/E-intron A can be effectively combined with a small human promoter element from the hMDRl gene promoter to produce promoter chimera having improved properties of gene expression in vitro and in vivo. It is also shown that the HCMV IE P/E-intron A can be minimized without a reduction in in vitro promoter activity and that intron A contributes to cell-specificity of gene expression.

These studies represent a novel strategy for combining strong viral promoters with defined, human promoter elements to create promoter chimeras with unique characteristics. The subject data provide an example of eliminating promoter sequences from the HCMV IE promoter that repress activity of the promoter chimera. By using the approach of combining cellular and viral promoter elements, identifying the sites of cooperation that elevate promoter activity, altering the organization of these unnatural promoter chimera, and potentially multimerizing elements that contribute positively or negatively to the activity or the specificity of gene targeting, a rational strategy is provided for improving currently available vectors for human gene therapy and for recombinant protein expression.

As further exemplified herein, in vivo studies reveal that the promoter chimera of the subject invention can enhance selective expression in target tumor tissue. Combining a defined hMDRl promoter element with the HCMV IE promoter and intron A generates promoter chimeras having improved targeting properties in lung, heart, and hepatic tissues after systemic injection with cationic lipids. Expression in lung can be improved 6- to 50-fold in vivo in a murine model. The selectivity for target gene expression in lung was also improved 6- to 10-fold with the Cl and C2 plasmids. It was also found that the orientation of the hMDRl element in the plasmid backbone affects the expression properties of the plasmid in vivo. This may be caused by differences in nuclear factors that associate with these complex promoters. Minimization, multimerization, and combination of such viral and cellular promoter elements provide a rational strategy for improving the targeting properties of both viral and nonviral vectors.

The HCMV IE promoter has served as the benchmark for eukaryotic expression, both for recombinant proteins and for human gene therapy trials. The HCMV IE promoter contains a strong enhancer that lies between nucleotides -524 to -118 (Boshart et al. (1985) Cell 41:521- 30). The region of the HCMV IE enhancer that most strongly contributes to heterologous gene expression in cultured cells lies between -407 to -138 and relative to the transcription start site. The subject promoter chimeras that were constructed contain a portion of the HCMV IE enhancer. The subject data show that this region of the HCMV IE promoter is necessary to maintain reporter gene expression in vitro and in vivo with the promoter chimeras evaluated herein. It is surprising that deletion of approximately one-half of the HCMV IE promoter/enhancer (about 300-bp) from the hMDRl promoter chimera having the longest DNA sequence used herein yields a vector that retains significant potency of reporter gene expression. The subject data provide a model for combining enhancers from viral and mammalian genes to produce unnatural promoters having improved properties of expression and altered targeting properties in animals. The hMDRl promoter-modified vectors can be useful for recombinant protein expression in a variety of cell lines, for gene therapies or vaccine development. The mechanism responsible for enhancing gene expression from the hMDRl/HCMV

IE promoter chimeras is not understood. The HCMV IE promoter/enhancer is composed of unique and repetitive sequence motifs that interact with different host nuclear proteins. The major enhancer (-530 to -120) is composed of several repeat sequence motifs that bind NF-kB, NF-1 and CREB/ATF-1 (CRE) (Rideg et al. (1994) Differentiation 56:119-29; Jeang et al. (1987) J. Virol. 61:1559-70; Fickenscher et al. (1989) Gen. Virol. 70:107-123; Niller et al. (1991) NAR 19:3715-21; Prosch et al. (2000) Virology 272:357-65). Intron A also has multiple NF-1 sites (Henninghausen et al. (1986) EMBO J. 5:1367-71). These elements are required to maintain HCMV IE promoter strength (Weston et al. (1988) Virology 162:406-16). It is possible that the hMDRl element cooperates with one or more of these transcription factor binding sites.

The plasmids having the hMDRl element display improved gene expression in lung when delivered intravenously with cationic lipids. The mechanisms for this increase are not understood. Both the level of gene expression and the selectivity of expression in lung were modified by changing the orientation of the hMDRl element or the contribution of the HCMV IE promoter. Expression of the marker gene luciferase in heart and liver was also improved. Expression of the marker genes may not reflect accurately the deposition of DNA in tissues. For example, it has been reported that the major site of DNA accumulation is liver, whereas the major site of luciferase expression is in lung (Liu et al. (1995); Liu et al. (1997)).

There are at least two reasons that lung is the major site of gene expression in these experiments. First, lung is the first capillary bed encountered after intravenous injection, and proteoglycans that provide receptors for these complexes may be expressed at a higher level in lung tissue. Second, the hMDRl/HCMV IE promoter chimeras may also contribute to lung-specific expression because of cellular factors that activate this unique promoter. An important and critical process for gene delivery is receptor-mediated uptake of lipid-DNA complexes. Direct injection of lipid-DNA complexes into the cytoplasm or nucleus of cells yields no gene transfer (Zabner et al. (1993) Cell 75:207). Cationic-lipid DNA complexes accumulate within the interior of lung tissue within one hour (Hui et al. (1996) Biophys. J. 71:590-99; Legendre et al. (1992) Pharm. Res. 9:1235-42). Prior to this time, lipid-DNA complexes can be displaced by anionic liposomes, suggesting that their entry into lung is rate-limiting (Barron et al. (1999) Hum. Gene Ther. 10:1683-94). Lung is also the major site of gene expression when naked DNA is injected intravenously prior to injection of cationic liposomes, showing that lipid-DNA complexes can form in situ. In the subject invention, modified promoters are constructed to improve tumor- and tissue-specific targeting properties of vectors without bias as to mechanism of gene expression. The subject invention demonstrates that by using a rational strategy of modifying the strong viral HCMV IE promoter and a disease-specific DNA sequence that may or may not have enhancer-like properties, the delivery of a target gene to lung can be improved by at least 50-fold.

Many studies performed in cultured cell models support a mechanism for hMDRl gene activation through the promoter element we used to modify the vector (Sundseth et al. (1997); Cornwell et al. (1993); Hu, et al. (2000) J. Biol. Chem. 275:2979-85; Asakuno, et al. (1994)). The nuclear factors that bind to this element mediate drug- and UV-induction of the gene. The MDR1 gene is also activated in patients having non-resectable, sarcoma pulmonary metastases after treatment with doxorubicin (Abolhoda et al. (1999) Clin. Cancer Res. 5:3352-56). Thus, combining chemotherapy or radiation treatments with vectors that incorporate hMDRl/HCMV IE promoter chimeras may provide a way to further augment gene expression in lung, or other normal or diseased tissues. The HCMV IE promoter and enhancer is also stress-inducible in some cells (Prosch et al. (2000) Virology 272:357-65). The modified promoter chimeras may be particularly useful for combining chemotherapy and gene therapy. Plasmid vectors can induce immune responses (Yew et al. (2000) Mol. Ther. 1 :255-62). If directed appropriately, such bystander effects of vectors can be advantageous for cancer therapy. For example, systemic delivery of DNA complexed to cationic lipids activates IL-12 in lung (Freimark et al. (1998) J. Immunol. 160:4580-86; Blezinger et al. (1999) Hum. Gene Ther. 10:723-31), and taxol treatment activates IL-8 expression in lung cancer (Collins et al. (2000) Cancer Immunol. Immunother. 49:78-84). Thus, combining conventional cancer therapies with therapeutic gene delivery under the control of stress-inducible promoters may provide another novel approach for treatment of lung cancer or tumors that metastasize to lung. Overall, the subject invention demonstrates that a small promoter element from the hMDRl gene can significantly augment expression of an already potent viral promoter. The subject invention also shows that the position of the human promoter element relative to the HCMV IE P/E affects its expression patterns. The HCMV IE P/E used in the subject promoter chimera is composed of about 1.6 kb of DNA. Thus, a 3% addition of regulatory DNA from the hMDRl promoter had a profound effect on the potency and selectivity of tissue-specific or disease-specific expression.

EXAMPLES Example 1

Construction of Promoter Chimeras having an hMDRl disease-specific DNA element and deletion of HCMV IE sequences not contributing to the activity of the promoter chimera. A Renilla luciferase reporter vector, pRL-TK, was employed for analyzing the transcription activity of the promoter chimeras in cultured cell lines. All novel constructs were verified by sequence analysis. The vector employed for analyzing the transcription activity of the promoter chimeras in cultured cell lines contains approximately 1.6 kb of promoter sequences from the HCMV immediately early gene and a luciferase reporter. All restriction and modifying enzymes were purchased from New England Biolabs (Beverly, MA) unless otherwise noted. All novel constructs were verified by sequence analysis. Cl, C2, and C3 (see Figure 1) were created by ligating hybridized oligonucleotides containing the 52-bp hMDRl promoter element (-88 to -36 bp, sense 5'-GTGAGGCTGATTGGCTGGGCAGGAACAGCGCCGGGGCGTGGGCTGAGCACAG-3' (SEQ. ID NO. 1), antisense

5'-CTGTGCTCAGCCCACGCCCCGGCGCTGTTCCTGCCCAGCCAATCAGCCTCAC-3') (SEQ. ID NO. 2) with defined restriction enzyme overhangs into appropriately digested vector.

Cl was created by introducing the hMDRl element upstream of the HCMV P/E into BstAPLMscI vector. C2 was created by introducing the hMDRl element in between the HCMV P/E and intron A into SacII digested vector.

C3 was created by introducing the hMDRl element downstream of intron A into Pstl/ECORV digested vector. Constructs containing discrete mutations of hMDRl sequences (see Figure 3) were created by ligating hybridized oligonucleotides into Pstl/EcoRV digested vector.

C3mutl was created with the following oligonucleotides: sense 5'-GACAGCGCCGGGGCGTGGGCTGAGCACAGGTGAGGCTGATTGGCTGGGCAG GAGAT-3' (SEQ. ID NO. 3); antisense 5'-ATCTCCTGCCCAGCCAATCAGCCTCACCTGTGCTCAGCCCACGCCCCGGCGCTGT

C TGCA-3' (SEQ. ID NO. 4).

C3mut2 was created with the following oligonucleotides (mutations shown in lower case): sense 5'-GGTGAGGCTGATTGGCTGGGCAGGAACAGCGCCGGGGaGTGaaCTGAGCACAGG

AT-3' (SEQ. ID NO. 5); antisense

5'-ATCCTGTGCTCAGttCACtCCCCGGCGCTGTTCCTGCCCAGCCAATCAGCCTCACC

TGCA-3' (SEQ. ID NO. 6).

C3mut3 was created with the following oligonucleotides (mutations shown in lower case): sense

5'-GGTGAGGCTGATgtGCTGGGCAGGAACAGCGCCGGGGCGTGGGCTGAGCACA

GGAT-3' (SEQ. ID NO. 7); antisense

5'-ATCCTGTGCTCAGCCCACGCCCCGGCGCTGTTCCTGCCCAGCacATCAGCCTCAC

CTGCA-3' (SEQ. ID NO. 8). C3mut4 was created using the following oligonucleotides (mutations shown in lower case): sense

5'-GGTGAGGCTGATgtGCTGGGCAGGAACAGCGCCGGGGaGTGaaCTGAGCACAGGA

T-3' (SEQ. ID NO. 9); antisense

5'-ATCCTGTGCTCAGttCACtCCCCGGCGCTGTTCCTGCCCAGcacATCAGCCTCACCT GCA-3' (SEQ. ID NO. 10).

C3mut5 was created with the following oligonucleotides: sense

5'-GGACACGAGTCGGGTGCGGGGCCGCGACAAGGACGGGTCGGTTAGTCGGAG

TGGAT-3' (SEQ. ID NO. 11); antisense

5'-ATCCACCCTGACTAACCGACCCGTCCTTGTCGCGGCCCCGCACCCGACTCGTGTC CTGCA-3' (SEQ. ID NO. 12).

The following oligonucleotides were used to create both the C3dimer and C3trimer constructs: sense

5'-GGTGAGGCTGATTGGCTGGGCAGGAACAGCGCCGGGGCGTGGGCTGAGCACAG

G AT-3' (SEQ. ID NO. 13); antisense 5'-ATCCTGTGCTCAGCCCACGCCCCGGCGCTGTTCCTGCCCAGCCAATCAGCCTCAC

C TGCA-3' (SEQ. ID NO. 14). The insert for C3dimer was created by hybridizing these oligonucleotides, blunting the newly created dsDNA with Klenow fragment, and then phosphorylating this insert with T4 polynucleotide kinase. This insert was then ligated into EcoRV digested C3. C3trimer was created by ligating identically prepared oligonucleotides into EcoRV digested C3dimer.

The following deletion constructs (see Figure 3) were derived from C3 using convenient restriction sites within HCMV P/E. To create DEL-3, C3 was digested with BstAPI and Spel. The digested vector was blunted with Klenow fragment and then religated. To create DEL- 1, C3 was digested with Ndel and then religated. To create DEL-4, C3 was digested with BstAPI and SnaBI. The digested vector was blunted with Klenow fragment and then religated. To create DEL-5, C3 was digested with Ndel and SacII. The digested vector was blunted with Klenow fragment and then religated. To create DEL- 10, C3 was digested with SnaBl and SacII. The digested vector was blunted with Klenow fragment and then religated.

Intron deletions (see Figure 4) were generated as follows. To create DEL-11, C3 was digested with SacII and PflMI. The digested vector fragments were blunted with Klenow fragment and then religated. To create DEL- 12, C3 was digested with PflMI and Hpal. The digested vector fragments were blunted with Klenow fragment and then religated. To create DEL- 14, C3 was digested with SacII and Pstl. The digested vector was blunted with Klenow fragment and then religated. To create DEL- 15, C3 was digested with Pstl and Hpal. The digested vector was blunted with Klenow fragment and then religated.

Minimal promoter constructs (see Figure 5) were generated as follows. To create DEL-9, DEL-3 was digested with Ndel and SnaBl . The digested vector was blunted with Klenow fragment and then religated. To create DEL-13, DEL- 12 was digested with BstAPI and Spel, blunted with Klenow fragment and religated. This plasmid was digested with Ndel and SnaBl, and then blunted with Klenow fragment and religated. To create DEL- 16, DEL- 13 was digested with Pstl and EcoRV. The digested vector was blunted with Klenow fragment and then religated.

Cell lines and culture: The colorectal carcinoma cell line, HCT 116, the epidermoid carcinoma cell line, KB, and the hepatocellular carcinoma cell line, Hep G2, were purchased from American Type Culture Collection (Manassass, VA) and maintained per ATCC instructions under standard conditions. Transient transfection and luciferase assay: One day prior to transfection, 2 ml of medium containing 2 x 10(5) HCT 116, 4 x 10(5) KB or 5 x 10(5) Hep G2 cells, were plated into six- well plates. Up to 1 microgram experimental reporter plasmid and 1.5 microgram of control pRL-TK were diluted to 0.1 microgram/microliter HEPES-buffered saline (HBS; 20 mM HEPES pH 7.4, 150 mM NaCl). A total of 12.5 microgram DOTAP in 42 microliters HBS was added to the DNA then incubated 15 min at room temperature. The DNA-DOTAP mixture was then mixed with 2 ml of complete medium and added to each well. Forty-eight h after transfection, cells were washed three times with 1 ml of ice cold phosphate-buffered saline (PBS; 137 mM NaCl, 2.7 MM KCI, 4.3 mM Na2HP04, 1.4 mM KH2P04) and lysed with 0.25 ml lysis buffer and 15 min of gentle rocking. Lysate was cleared by centrifugation (13,000 rpm for 5 min) then stored at -70 °C until assayed. Luciferase activities were measured in the same 20 microliter aliquots according to the manufacturer's instructions (Dual-Luciferase Reporter Assay System, Promega) using a MLX Microtiter Plate Luminometer (Dynex Technologies; Chantilly, VA). In brief, firefly luciferase activity was measured by adding 100 microliters Luciferase Assay Reagent II (LAR II). The addition of 100 microlitres Stop & Glo Reagent quenched the firefly luciferase reaction and initiated the Renilla luciferase (RL) reaction. All transfections were performed in triplicate and repeated at least once.

Statistical analysis: pRL-TK contains the Renilla luciferase (RL) gene under the control of the thymidine kinase (TK) promoter and was co-transfected in all in vitro experiments, allowing normalization of firefly luciferase activity to constitutive pRL-TK activity. Data are depicted as mean = standard error of the mean (SEM). Fold induction is defined as the ratio of a test constructs activity to the activity of the C3 or the parent construct or the parent construct. Differences in promoter activity were analyzed using Student's unpaired t-tests. P<0.05 was considered significant.

Example 2 - A 52-bp element from the hMDRl promoter augments HCMV P/E promoter activity

The HCMV IE P/E drives high-level expression in numerous tissues and cell lines (Foecking et al. (1986); and Schmidt et al. (1990) Mol. Cell. Biol. 10:4406-11), and has been extensively employed in mammalian expression vectors. The vector used to test the promoter chimera contains approximately 685 base pairs (bp) of the HCMV IE P/E and approximately 900 by of intron A and is similar to one described in Hartikka et al. (1996) Hum. Gene Ther. 7:1205-17. By introduction of an hMDRl disease-specific element (-88 to -36) into different locations of the HCMV P/E we sought to develop a promoter both sensitive to hMDRl activation and having a superior transcription activity than the hMDR element alone (Sundseth et al. (1977)). The hMDRl element was introduced either 5' to HCMV P/E (construct Cl), between HCMV P/E and intron A (construct C2), or 3' to intron A (construct C3, see Figure 1 A). Promoter activity of these constructs was then evaluated in transient transfection assays (Figure IB). Similar patterns of promoter activity were observed in two human cell lines that overexpress hMDRl . Introduction of the hMDR.l element upstream of the HCMV P/E (Cl) or in between the HCMV P/E and intron A (C2) resulted in a 1.2 to 1.4 fold increase in promoter activity. In contrast, introduction of the same sequence downstream of intron A (C3) resulted in significantly higher promoter activity (1.5-fold for Hep G2 cells and 2- fold for HCT 116 cells). In KB epithelial carcinoma cells, constructs Cl and C2 displayed about 2-fold more promoter activity than the parent construct, while construct C3 again demonstrated significantly higher promoter activity than either Cl or C2. These data demonstrate that a small human promoter element can significantly augment expression of an already potent viral promoter and that the placement of the element in relation to HCMV P/E sequences can affect promoter activity.

The activity of the promoter chimeras was evaluated by transient transfection in HCT 116 colon carcinoma cell lines (Table 1). Replicate transfections are shown (A and B). The luciferase activity was dose responsive in each case (compare 0.5 and 1.0 ug DNA). When 1.0 ug of DNA was transfected, luciferase activity from the promoter chimeras Cl, C2, and C3 in HCT 116 cells was elevated 1.4, 1.9, and 2.6-fold, respectively. Thus, the potency of the promoter chimera was altered by changing the orientation of the hMDRl element relative to the HCMV promoter cassette. As the hMDRl element is cloned closer to the reporter gene, the reporter activity improves in this cultured cell line.

Parent Cl C2 C3

0.5 1.0 0.5 1.0 0.5 1.0 0.5 1.0

HCT 3,081.4 5,432.1 3,209.1 8,627.0 4,250.8 9,955.7 4,736.5 14,158.6 SET A ± 71.4 ± 164.0 ± 726.5 ± 821.5 ± 91.1 ± 126.6 ± 46.8 ± 424.9

HCT 2,689.3 5,668.2 3,511.1 7,557.0 4,863.0 11,365.5 6,093.8 13,870.5 SET B ± 30.3 ± 59.0 ± 38.5 ± 559.8 ± 65.4 ± 934 ± 69.3 ±158.2

Table 1. Transient transfection of a modified promoters into cultured cells. The methods for transfection are described in our published work (Sundseth et al., 1997). Results from two separate transfections are shown for each condition (8 replicates each). DNA was transfected at 0.5 and 1.0 ug. The light units/mg protein and standard errors are given below the bar graph.

The modified plasmids were further evaluated in three different cell lines to compare their potencies by correcting for transfection efficiency using a dual reporter assay with luciferase in a plasmid under the control of a cellular promoter from a gene constitutively expressed in all cells (Table 2). Luciferase in the Cl, C2, and C3 plasmids was under the control of the hMDRl/HCMV promoter chimeras described above. The absolute luciferase activities for all the plasmids having hMDRl/HCMV promoter chimeras was significantly higher in HCT 116 cells. The C3 promoter chimera was most potent in all cell lines, having 200-fold, 90-fold, and 115-fold better activity than the cellular promoter in HCT 116, KB, and Hep G2 cells, respectively. The C3 chimera was approximately four times more potent in HCT 116 cells (range of 264 to 421 LU/pg protein) than in KB cells (range of 55 to 108 LU/pg protein) when the correction for transfection efficiency was omitted. The activity of the C3 promoter chimera is approximately 2.5, 4.8, and 1 J times higher than the parent plasmid in HCT 116, KB, and Hep G2 cells, respectively (data not shown). The hMDRl/HCMV promoter chimeras are also active in lung tumor cell lines such as A549 and Calu 6 (data not shown). These data support the conclusion that the hMDRl/HCMV promoter chimeras will have general utility for recombinant protein expression in vitro.

Table 2. Transient transfection of two reporter vectors into cultured cells under the control of the hMDRl promoter chimera or a promoter from a cellular gene. Column 1 shows reporter gene activity from the hMDRl promoter chimera, and column 2 shows the activity from an alternate cellular gene that is constitutively expressed to a similar level in all cells. Column 3 is the fold-induction of the promoter chimera relative to the cellular promoter. KB cells that do not express the hMDRl gene have a lower fold-induction compared to two other cell lines that overexpress the hMDRl gene.

Example 3 - Mutation of the liMDRl Y-box and two overlapping GC-boxes to identify DNA sequences responsible for augmenting transcription activity of the promoter chimera

The hMDRl element that increases HCMV P/E promoter activity contains two GC- boxes and a Y-box (inverted CAATT box) (Cornwell et al. (1993); Goldsmith et al. (1993); and Sundseth et al. (1977)). A series of discrete mutations of this element was introduced into C3 to determine which part(s) of the hMDRl element interacts with HCMV sequences (Figure 3). Introduction of a mutation (C3mut2) that abolishes Spl binding to a hMDRl GC-box-probe and reduces promoter activity of hMDRl expression vectors in a transient transfection assay, (Sundseth et al. (1997)) had no effect on the activity of the promoter chimera in HCT 116 and Hep G2 cells, and modestly increased promoter activity in KB cells. In contrast, introduction of a mutation (C3mut3) that abolishes NF-Y binding to a hMDRl Y-box probe (Sundseth et al. (1997)) reduced promoter activity in all cell lines by approximately 50%. A construct containing both of these mutations (C3mut4) had a similar activity as C3mut3. These data suggest that in the C3 chimera the Y-box, but not the GC-boxes is responsible for augmenting the activity of the HCMV P/E.

To further evaluate interactions between the hMDRl element and HCMV P/E intron A, three other constructs were created. C3mut5 demonstrates that the 52-bp element can function in reverse (3' to 5') orientation. Thus, this element demonstrates some enhancer-like behavior within the promoter chimera. When the Y-box and GC-boxes are transposed but individually retain their 5' to 3' orientation (C3mutl), promoter activity is modestly reduced in HCT 116 and Hep G2 cells, but unaffected in KB cells. This suggests that the spacing of the Y-box in relation to HCMV sequences may be important for full chimeric promoter activity in a cell line-dependent manner. Surprisingly, the introduction of multiple copies of the hMDRl element (C3trimer) drastically reduced promoter activity (to -10% of C3). Similar repression was obtained for constructs containing two or four copies of the hMDRl element (data not shown). The mechanistic reason for these dramatic decrements in promoter activity when multiples of the liMDRl are created is unclear.

Example 4 - Deletion analysis of the HCMV promoter, enhancer and intron A to identify sequences required for full activity of the promoter chimera

To evaluate the regions of the HCMV P/E and intron A required for full activity in the C3 promoter chimera, the 5' region of the HCMV promoter or intron A was deleted (Table 4). Deletion of 312-bp from the HCMV promoter (Ndel/Ndel) decreased promoter activity by approximately 20 to 40%_o in each cell line tested (DELI, Table 3). Thus, the 5' region of the HCMV promoter is necessary to maintain optimal gene expression levels. Deletion of the entire intron A (DEL 6) reduced the potency of the C3 promoter chimera in all cell lines from 20 to 70% (Table 3). The potency was affected most in KB cells and least in Hep G2 cells. In transgenic mice, the addition of intron A to the HCMV promoter/enhancer caused muscle-specific expression that was absent in a vector lacking intron A (Schmidt et al., 1990). These data support the conclusion that intron A contributes to tissue selective expression both in vitro and in vivo. Sequences that are 5' to the major HCMV enhancer and intron A contribute to the activity of the modified promoter chimeras. Deletion of intron A from the promoter chimeras generates a novel promoter chimera with differential expression in cells that express the hMDRl gene (HCT 116 and Hep G2) and those that do not (KB). CELL LINES

HCT 116 KB Hep G2

Construct

RLU ± SD fold induct. RLU± SD fold induct. RLU± SD fold induct.

A

C3 421.1 ±126.0 1 ±0.29 108.3±11.2 I ±O.IO 171.8±28. 1 ±0.16

DEL-1 256.5± 63.1 0.60±0.14 66.9±24.5 0.61 ±0.22 142.0±24. 0.82±0.14

B

C3 280.6±28.1 I ±O.IO 5.8±5.4 1 ±0.09 156.3±19. 1 ±0.12

DEL-6 142.6±45.2 0.50±0.16 16.3±2.6 0.29±0.04 128.4±34. 0.82±0.22

Table 3: Normalized and corrected values representing the activity of HCMV P/E/I deletions in HCT 116, KB, and Hep G2 cell lines. Transfections were done in triplicate and repeated twice.

Deletion of 150-bp from 5' portion of the HCMV promoter (113-bp) and a portion of the plasmid bordering the promoter (DEL3) had little effect on luciferase activity when tested in HCT 116 cells. Thus, this region of the HCMV promoter is dispensable. Deletion of 312-by (DELI) or 491-bp (DEL4) reduced the potency by about 20% to 40%. Complete deletion of the HCVM promoter/enhancer (DEL5), but retaining both intron A and the hMDRl element, nearly abolished the activity of the promoter chimera. When intron A is deleted (DEL6) but the HCMV promoter is retained, approximately 30% of the activity remains. Thus, the combination of hMDRl and HCMV promoters yields promoter chimeras having improved expression properties in vitro. Complete elimination of the HCMV regulatory cassette yielded a plasmid that retained the hMDRl element but had very low activity in vitro (about 1%). Thus, both the HCMV promoter/enhancer and intron A contribute to the activity of the promoter chimeras. In addition, a 5' region of the HCMV promoter can be deleted from the C3 chimera (DEL3) without significant loss of activity. Several deletion constructs of the prompter chimera retain significant activity (DELI, DEL6, and DEL4).

Deletion construct Total b removed % Activity of C3

DEL I -312-b 77

DEL 3 -150-b 116

DEL 4 -491-b 77

DEL 5 -HCMV P/E 1 -807-b

DEL 6 -Intron A 31 764-b

DEL 8 -P/E/I 1

Table 4: Summary of deletion constructs of hMDRl/HCMV promoter chimeras Example 5 - The 3' and 5' splice sites of intron A are required for the full activity of the C3 promoter chimera

Additional deletions of the chimeric C3 construct were made (Figure 4). Deletion of the entire intron A (DEL6 and DEL14) reduced promoter activity between 30-40% in HCT 116 and KB cell lines (Tables 2, 3 and 4). However, the same construct had increased promoter activity relative to C3 in Hep G2 cells, suggesting that the function of intron A in regulating promoter activity is cell line-dependent. Deletion of either 5' intron sequences (DEL- 11) or 3' intron sequences (DEL- 15) reduced promoter activity by 80-95% in all cell lines tested. These deletions, which eliminate the 5' or 3' intron splice sites respectively, suggest that HCMV P/E activity requires appropriate intron A splicing. In contrast, internal intron A sequences between the PflM 1 site and the Hpal site could be deleted without a reduction in promoter activity (DEL- 12). Promoter activity was actually augmented by deletion of this fragment in both HCT 116 and Hep G2 cell lines. These data demonstrate that while intron A splice sites are required for full promoter activity, sequences between the PflMI site and the Hpal site are dispensable and may inhibit expression in some cell lines. The DEL12 promoter chimera retains some differential transcription activity in cells that do or do not express the hMDRl gene.

Example 6 - Deletion of nonessential HCMV sequences from the promoter chimera

Sequences found in the deletion analysis to negatively contribute to transcription in cultured cells were eliminated to generate a smaller promoter chimera that retains transcription activity. The distal 100 bp of the HCMV promoter in the chimera (to the Spel site) and the internal HCMV promoter sequences that failed to contribute to the C3 promoter chimera activity (Ndel to SnaBl) were deleted (DEL9). This promoter chimera (DEL9) retained activity in all cell lines tested comparable to the full length chimera. When an internal fragment from intron A was eliminated from the DEL9 promoter chimera (DELI 3), the transcription activity increased (20% to 65%). These data show that nonessential sequences can be eliminated from the promoter chimera and retain activity in cultured cells. To confirm that the disease-specific hMDRl element contributed to transcription activity in the reduced version of the promoter chimera, it was deleted from DEL13 (DEL16). This promoter chimera had a reduced activity in the two cell lines that overexpress the hMDRl gene, and a similar activity to DEL 13 in a cell line that does not express hMDRl . The DEL 13 promoter chimera has a modest differential expression in cells that overexpress the hMDRl gene. Example 7 - HCMV/frMDRl promoter chimeras display improved properties of gene expression in lung and other tissues

To evaluate the efficacy of gene delivery by hMDRl/HCMV promoter chimeras, four modified plasmids (40 ug) and the parent having only the HCMV P/E and intron A were injected into the tail veins of mice using a formulation with the cationic lipid DOTAP (1 ug DNA/ 36 umol lipid) (Song, 1997). After 24 hours, luciferase expression was measured in several different tissues (lung, liver, colon, heart). There is a significant improvement in luciferase expression in lung with all four of the modified promoter chimeras when compared to the parent promoter that lacks the hMDRl element (Figures 6 and 7). The C2 chimera and the DELI promoter chimera had lower levels of expression in lung than the Cl and C3 promoter chimeras. The location of the hMDRl element relative to the HCMV P/E may affect the cellular factors that operate on the modified promoter template (compare C3 and Cl with C2). Although the DELI promoter chimera showed a significantly improved luciferase expression compared to the parent vector, its expression in lung is reduced approximately 15 to 17-times. These data imply that in the DELI promoter chimera a critical region of the HCMV promoter/enhancer required for efficient gene expression in lung subsequent to systemic injection was deleted.

The C3 promoter chimera showed the highest levels of luciferase expression in all tissues (not shown) but colon after systemic injection (about 50-fold in lung, about 4-fold in heart, about 8 -fold in liver). None of the promoter chimeras provided improved reporter gene expression in colon after systemic injection, although each is highly potent in a cultured colon carcinoma cell line (HCT 116). Only the C3 promoter chimera had a significantly improved expression in liver tissue. Three of the plasmids having promoter chimeras (Cl, C2, and C3) display significantly improved properties of luciferase expression in heart. Deletion of the 5' region of the HCMV P/E eliminates or reduces this improvement in heart and lung. Deletion of the 5' portion of the HCMV P/E significantly improved the activity of the promoter chimera in liver, supporting the notion that modification of the HCMV P/E can alter the targeting properties of promoter chimeras subsequent to systemic injection with cationic lipids. By combining a small, defined cellular element from the hMDRl promoter with the

HCMV P/E and intron A, we generated one plasmid having improved properties of expression in liver, three in heart, and four in lung. The addition of a small, defined element from the hMDRl gene promoter improved gene delivery to lung approximately 50-fold (C3). Selectivity for gene expression in lung was also improved by 6- to 10-fold with the Cl, C2, and DELI promoter chimeras. The invention describes rational methods for generating novel promoter chimeras from disease-specific sequences and viral promoter sequences. More than a dozen novel promoter chimeras have been engineered having unique DNA sequence and containing a disease-specific element from the hMDRl gene. Gene targeting and expression can be improved and modified by engineering unnatural promoter chimeras composed of truncated, mutated, rearranged, or multiples of DNA sequences between components of different origins.

Claims

What is claimed is:

1. A promoter chimera for in vivo or in vitro expression comprising: a) a viral promoter, intron, or other unit transcription regulatory sequence of any length, and b) a unit disease-specific promoter, intron, or other unit transcription regulatory sequence of any length sensitive to genotoxic stress.

2. The chimera of claim 1 , wherein the unit transcription regulatory sequences within the promoters or introns are multiples of the same sequence, ligated sequence or mutated sequence in forward or reverse orientation.

3. The chimera of claim 1, wherein: a) the viral promoter, intron or other unit transcription regulatory sequence is from the HCMV immediately early (IE) gene promoter or transcription regulatory sequence including intron A (HCMV IE P/E/I); and b) the unit disease specific promoter, intron or other transcription regulatory sequence is 100 base pair or less in length.

4. The chimera of claim 3 wherein the sequences are multiples of the same sequence and/or mutated sequence in forward or reverse orientation.

5. The chimera of claim 1 and 3, wherein: the disease specific promoter, intron, or other transcription regulatory sequence is an hMDRl promoter selected from the group consisting of: i) a sequence located at -88 to -36 of the human hMDRl gene, wherein said sequence is located at any position in forward or reverse orientation within or flanking the viral promoter sequence, and wherein said unit transcription regulatory sequence may be multiples of the same sequence and/or mutated; ii) a Y-box sequence from the hMDRl promoter (-88 to -36) in forward or reverse orientation; iii) a sequence located at -137 to +158 of the hMDRl gene, wherein said unit transcription regulatory sequence may be multiples of the same sequence, ligated sequence, or mutated sequence in forward or reverse orientation; and iv) a sequence located -5,000 to +250 of the hMDRl gene, wherein said unit transcription regulatory sequence may be multiples of the same sequence, ligated sequence, or mutated sequence in forward or reverse orientation.

6. The chimera of claim 1, wherein: a) the viral promoter is any sequence from the HCMV immediately early gene promoter or transcription regulatory sequence including intron A (HCMV IE P/E/I), and b) the disease-specific promoter or transcription regulatory sequence is an hMDRl promoter selected from the group consisting of: i) a unit transcription regulatory sequence located at -88 to -36 of the human hMDRl gene, wherein said sequence is located at any position in forward or reverse orientation within or flanking the viral promoter sequence, and wherein said sequence may be multiples and/or mutated; ii) one or more Y-box sequences from the hMDRl promoter (-88 to -36) in forward or reverse orientation; iii) a unit transcription regulatory sequence located at -137 to +158 of the hMDRl gene, wherein said sequence may be multiples, ligated, or mutated in forward or reverse orientation; and iv) a unit transcription regulatory sequence located at -5,000 to +250 of the hMDRl gene, wherein said sequence may be multiples, ligated, or mutated in forward or reverse orientation.

7. The chimera of claim 6, further comprising: c) additional disease specific promoter or intron sequences or other franscription regulatory sequences of 100 base pair or less.

8. The chimera of claim 6, wherein any of the sequences may be multiples of the same sequence, ligated sequence or mutated sequence in forward or reverse orientation.

9. The chimera of claim 6, wherein any unit transcription regulatory sequence from the HCMV immediate early promoter or transcription regulatory sequence including intron A is present in one or more copies, in forward or reverse orientation, or is mutated relative to wild type.

10. The chimera of claim 6, wherein the hMDRl promoter, intron or other transcription regulatory sequence is the sequence located at -88 to -36 of the human hMDRl gene.

11. The chimera of claim 10, wherein said chimera is used in a vector having a structural or non-structural gene.

12. The chimera of claim 10, wherein the hMDRl promoter, intron, or other transcription regulatory sequence is located 5' to intron A and 3' to the enhancer of the HCMV IE P/E/I sequence.

13. The chimera of claim 10, wherein the hMDRl promoter, intron or other transcription regulatory sequence is located 5' to the HCMV IE P/E/I sequence.

14. The chimera of claim 10, wherein the hMDRl promoter, intron, or other transcription regulatory sequence is located at any position within or flanking the HCMV IE P/E/I sequence.

15. The chimera of claim 14, wherein the hMDRl promoter, intron, or other transcription regulatory sequence is oriented in reverse orientation in the HCMV IE P/E/I sequence.

16. The chimera of claim 14, wherein the unit franscription regulatory sequence from the HCMV IE P/E/I sequence is in forward or reverse orientation, and the hMDRl sequence is located at any position within or flanking the P/E/I sequence.

17. The chimera of claim 10, wherein the hMDRl promoter, intron, or other transcription regulatory sequence is multiples of the same sequence or ligated sequence, and is in forward or reverse orientation.

18. The chimera of claim 10, wherein: the HCMV IE P/E/I sequence is multiples of the same sequence, or ligated sequence, and is in forward or reverse orientation, and the hMDRl promoter, intron, or other franscription regulatory sequence is in forward or reverse orientation.

19. The chimera of claim 10, wherein the hMDRl promoter, infron, or other transcription regulatory sequence is mutated.

20. The chimera of claim 10, wherein the HCMV IE P/E/I sequence is mutated.

21. The chimera of claim 10, wherein the hMDRl promoter, intron, or other transcription regulatory sequence is one or more Y-box (-88 to -36) sequences in forward or reverse orientation.

22. The chimera of claim 6, wherein the hMDRl promoter, intron or other transcription regulatory sequence is located at -137 to +158 of the hMDRl gene, wherein said sequence may be multiples of the same sequence, ligated sequence, or mutated sequence in forward or reverse orientation.

23. The chimera of claim 6, wherein the hMDRl promoter, intron or other transcription regulatory sequence is located at -5,000 to +250 of the hMDRl gene, wherein said sequence may be multiples of the same sequence, ligated sequence, or mutated sequence in forward or reverse orientation.

24. A vector comprising a promoter chimera of any of claims 1-23, and further comprising:

(a) a coding sequence for a therapeutically relevant product operably linked to the promoter chimera; and

(b) sequences necessary for replication in a virus, plant, eukaryotic or prokaryotic cell.

25. A method for the generation of improved expression constructs for in vitro and in vivo expression of therapeutically-relevant gene products, the method comprising: a) identifying a human gene undergoing disease-specific transcriptional activation; b) identifying promoter and/or enhancer and/or intron sequences or other franscription regulatory sequences from said human gene that at least partially responsible for said disease-specific transcriptional activation; c) preparing a plurality of promoter chimeras comprising:

(i) one or more copies in forward or reverse orientation of sequences from a viral promoter and/or enhancer and/or intron; (ii) one or more copies in forward or reverse orientations of said human disease-specific sequences; wherein said sequences in (ii) are located at a variety of positions in said promoter chimera relative to said sequences in (i); d) preparing vectors for the expression of said therapeutically relevant gene product, each vector comprising a promoter chimera of c) operably linked to said therapeutically relevant gene product, and further comprising sequences necessary for replication; e) optionally mutating said human disease-specific sequences and/or said viral sequences in said vectors; f) selecting vectors from d) and e) that have a desirable expression level and/or expression site of said therapeutically relevant gene product; and g) optionally minimizing the size of the vectors selected as having desirable expression levels and/or sites in step f), wherein said minimized vector substantially retains a desirable expression level and/or site, and wherein minimization is achieved by deleting portions of said sequence from said human disease-specific sequences and/or said viral sequences.

26. The method of claim 25 wherein said viral sequences comprise the human cytomegalovirus intermediate early gene promoter/enhancer/intron A (HCMV IE P/E/I).

27. The method of claim 25 wherein said human disease specific sequence is hMDRl promoter, intron, or other franscription regulatory sequence.