Approaches of Transcriptional Regulation

A. Prior Rationale: Universal Promoters

Several universal promoters have been utilized to attempt to maximize gene expression. The LTR, CMV, and RSV promoters were isolated from Maloney retrovirus, cytomegalovirus, and Rous sarcoma virus, respectively.

These promoter elements were used because of the universal transcriptional activation over a broad host range. This universal transcription allowed for excellent but nondiscriminatory gene transcription and subsequent transgene expression. Because of the high levels of gene expression within several DNA constructs (i.e., viruses, cosmids, plasmids, etc.), these promoters are still used daily throughout the scientific community to test hypotheses which require uniform and high-level gene transcription. These were the promoters utilized in the first wave of gene therapy clinical trials, which focused on maximal gene expression and used local injection techniques to control the region of gene expression achieved. The LTR promoter was used to control herpes simplex virus thymidine kinase (HSV-TK) expression in a retroviral vector by placing retroviral producer cells into residual brain tumors to confer TK expression to the brain tumor, which could lead to conversion of a prodrug and subsequent tumor cell death. The CMV promoter was used in a replication-deficient adenovirus to deliver p53 gene expression after intralesional delivery to patients with both lung and head and neck tumors and is still under clinical investigation. The RSV promoter was employed to express HSV-TK after intralesional delivery in patients with several different tumor types.

B. Current Rationale of Tissue-Specific Promoters

A major challenge facing gene therapy is to generate vectors capable of achieving tissue- or tumor-specific expression. Initial gene therapy strategies utilized universal promoters that demonstrated gene transfer, but were associated with toxicity associated with nonspecific gene transduction (section III.A, above). Tissue-specific promoters offer a novel approach to developing transcriptionally targeted viral vectors with enhanced potential for human gene therapy applications as described below. Several important characteristics are required to develop a tissue/tumor-specific strategy for a particular disease. Fortunately, the recent explosion in our understanding of molecular events that are present in a variety of disease processes has simplified the identification of suitable promoters. Additionally the completion of the genome project and the utilization of microarray technology have enhanced the development of tissue- or tumor-specific promoters by allowing for the identification of novel but specific molecules associated with a particular disease (e.g., cancer). The advancements in molecular cloning techniques (e.g., PCR) has allowed the investigator to extract regulatory sequences from genomic DNA and evaluate each component through site directed mutagenesis analysis in plasmid expression vectors. Additionally, the development of luciferase and green fluorescent protein as well as other quantifiable transgenes has enabled the investigator to test the tissue- or tumor-specific nature of a particular promoter.

To illustrate the concept and utility of a tissue/tumor-specific promoter five such promoters have been selected from Table I. The basic rationale for

Table I

Gene Therapy Applications Of Tissue-Specific Promoters for Transcriptional Targeting

Promoter

Tissue-Specificity

Transgene

Vector

References

AFP

HCC

HSV-TK

Adenoviral

[27, 126,

134]

CD, IL-2

Adenoviral

[28, 30]

El A

Adenoviral

[133]

Albumin

Liver

factor VIII

Adenoviral

[135, 136]

oi-Actin

Muscle

GHRH

Nonviral

[137]

a-Lactalbumin

Breast cancer

CD

Adenoviral

[138]

P-Lactoglobulin

Breast cancer

HSV-TK

Adenoviral

[139]

P-Globin

Erythroid cells

P-globin

Retroviral

[140]

c-erbB2

Breast and

HSV-TK

Adenoviral

[141, 142]

pancreatic cancer

CEA

Breast, pancreatic,

HSV-TK, Cre

Adenoviral

[15, 19,21]

lung, and

H-ras mutant

colorectal

carcinoma

Egr-1

Radiation induced

TNF-a, LacZ

Adenoviral

[143, 144]

E-Selectin

Tumor endothelium

TNF-a

Retroviral

[145]

Fit-1

Vascular endothelial

Luciferase

Adenoviral

[146]

growth factor

receptor type-1

GFAP

Glial cells

FasL

Adenoviral

[147, 148]

TH

Retroviral

Grp78 (BIP)

Anoxic/acidic tumor

HSV-TK

Adenoviral

[149, 150]

tissue

HSV-TK

Retroviral

hAAT

Hepatocytes

FactorIX

Nonviral

[151]

HGH and HGPH-a

Pituitary

HSV-TK,

Adenoviral

[152, 153]

HIF-la/HRE

Hypoxia inducible

Erythropoietin

Nonviral

[154]

hK2

Prostate

EGFP, E1A, E1B

Adenoviral

[132, 155,

156]

HSP

Heat induced

p53, TNF-a

Nonviral

[157]

Hybrid ERE-HRE

Breast Cancer

Harakiri

Adenoviral

[158]

L-Plastin

Epithelial tumors

LacZ

Adenoviral

1159]

MBP

Oligodendrocytes

Caspase 8

Adenoviral

[160]

GFP

AAV

[161]

MCK

Undifferentiated

LacZ

Adenoviral

[162]

muscle

MMTV-LTR

Prostated cancer

antisense

Retroviral

[163]

c-myc

MN/CA9

Renal cell carcinoma

E1A

Adenoviral

[164, 165]

MUC1 (DF3)

Breast cancer

E1A

Adenoviral

[166]

HSV-TK

Retroviral

1167]

Nestin

Glioma,

Cre, LacZ

Adenoviral

[168]

glioblastoma

Table I {continued)

Promoter

Tissue-Specificity

Transgene

Vector

References

NSE

Neurons

FasL,

Adenoviral

[169]

BDNF

AAV

[170]

Osteocalcin

Osteosarcoma

HSV-TK

Adenoviral

[77, 79,

171-173]

Prostate

HSV-TK

Adenoviral

[174-176]

E1A

Adenoviral

[177|

PEPCK

Hepatocytes

Insulin

Adenoviral

[1781

PSA

Prostate

Nitroreductase

Adenoviral

[46,48,49]

HSV-TK, PNP

Adenoviral

Preproenkephalin

CNS

LacZ

HSV

[179]

Probasin

Prostate

E1A

Adenoviral

[156, 180]

Caspase 9

Prolactin

Pituitary

HSV-TK

Adenoviral

[181]

lactotrophic cells

SLPI

Ovarian, cervical

HSV-TK

Nonviral

[182]

carcinoma

SM22a

Smooth muscle cells

LacZ

Adenoviral

[183]

Surfactant protein C

Respiratory

HSV-TK

Adenoviral

[184]

epithelium

Tyrosinase

Melanocytes

Luc, PNP

Nonviral

[185, 186]

GALV-FMG

Retroviral

Tyrosine hydroxylase

Sympathetic nervous

LacZ

HSV

[122]

system system selection, in vitro and in vivo laboratory investigation and the clinical testing associated with each, will be briefly reviewed below.

  1. Carcinoembryonic Antigen (CEA) Promoter a. Rationale Carcinoembryonic antigen is a 180-kDa cell surface glycoprotein overexpressed in 90% of gastrointestinal malignancies, including colon, gastric, rectal, and pancreatic tumors, 70% of lung cancers, and about 50% of breast cancers [7]. Thompson et al. [8] initially reported on the molecular cloning of the CEA gene from a human genomic library. Subsequently, Schrewe et al. [9] also isolated and characterized a cosmid clone containing the entire coding region of the CEA gene including its promoter. The CEA promoter region encompasses 400 bp upstream of the translational start site and is known to confer tissue-specific CEA expression. Hauck and Stan-ners [10] demonstrated that the CEA promoter region located between —403 and —124 bp upstream of the translational initiation site is capable of directing high levels of gene expression in CEA-expressing human colon cancer CRC cells. Chen et al. [11] showed the CEA promoter region to lie between —123 and —28 bp upstream from the transcriptional start site and have demonstrated the presence of SP1 and upstream stimulatory factor binding sites. According to Richards et al. [12] the CEA promoter is located between —90 and +69 bp upstream from the transcriptional start site and the essential sequences of the CEA promoter reside between —90 and —17 bp upstream of the transcriptional start site of the CEA gene. Cao et al. [13] compared the CEA core promoter regions between —135 and +69 bp isolated from human colorectal carcinoma and normal adjacent mucosa and found that both the sequences were identical and without any mutations. Taking advantage of this fact, various studies have suggested the potential utility of the CEA promoter for restricted expression of heterologous genes (14, 10, 12).
  2. In Vitro and in Vivo Experiments with CEA Promoter Takeuichi et al. [15] demonstrated that an adenoviral vector encoding a CEA promoter-driven N116Y dominant-negative H-Ras mutant was capable of suppressing liver metastasis by the human pancreatic cancer cell line PCI-43 in a nude mice model. Lan et al. [16, 17] demonstrated successful adenoviral-mediated transduction of E. coli cytosine deaminase (CD) in vitro as well as in an immunodeficient in vivo model of MKN45 gastric carcinoma. As compared to an adenoviral vector in which CD expression is driven by the constitutive CAG promoter, the expression of CD under the control of CEA promoter was confined to tumor xenografts. However, the reduction in tumor burden by AdCEA-CD/5-fluoro-cytosine (5FC), although significant, was not as great as that induced by AdCAG-CD/5FC. In fact, the CEA promoter was shown to be 200 times less active than the CAG promoter.

Similar results have been described in mice bearing xenografts that were transfected with CEA-CD constructs and subsequently treated with 5-FC (18, and 12). Tanaka and colleagues [19] have used the CEA promoter sequence located between —424 and —2 bp upstream of translational start site to generate an adenoviral vector expressing HSV-TK and examined its efficacy in killing CEA-producing cancer cells in vitro and in vivo. By employing intratumoral Ad-CEA-TK injection and gancyclovir (GCV) administration, the growth of the tumors was inhibited by 20% as compared to untreated tumors. Brand et al. [20] have used the CEA promoter (—296 to +102 bp with respect to transcriptional start site) to drive the expression of HSV-TK in an adenoviral vector. Their results indicate that the CEA promoter was active in several human and rat tumor-derived cell lines but not in rat primary hepatocytes and in mouse liver, while the CMV promoter was highly active in all cell types. Although the CEA promoter-driven TK expression was less, it was sufficient to kill 100% of cancer cells, indicating a significant bystander effect. Treatment of subcutaneous tumors in SCID mice with Ad-CEA-TK was able to significantly reduce tumor growth and the tail vein injection of a high dose of this virus caused no side-effects in the liver.

Kijima et al. [21] have utilized a novel Cre-lox-based strategy to achieve enhanced antitumor effect against CEA-producing human lung and colon cancer cell lines. Their strategy involved generation of two recombinant adenoviral vectors: one expressing the Cre recombinase gene under the control of the CEA promoter while the second adenoviral vector is designed to express HSV-TK gene from the CAG promoter only after Cre excises the neomycin resistance gene (inserted between the CAG promoter and HSV-TK) in a loxP site-specific manner. (Cre recombinase derived from bacteriophage PI mediates site-specific excisional deletion of a DNA sequence that is flanked by a pair of loxP sites composed of 34 nucleotides.) This novel approach requires simultaneous infection of a cell by the two adenoviral vectors. Using this approach, a CEA-producing human cancer cell line was rendered 8.4-fold more sensitive to GCV than infection by Ad-CEA-TK alone. Intratumoral injection of Ad-CEA-Cre along with Ad-lox-TK followed by GCV treatment almost completely eradicated CEA-producing tumors in an athymic subcutaneous tumor model, whereas intratumoral injection of Ad-CEA-TK with GCV treatment showed reduced tumor growth.

2. a-Fetoprotein (AFP) Promoter a. Rationale The human AFP gene is developmentally regulated and is expressed at high levels in the fetal liver but its transcription declines rapidly after birth and is barely detectable in adult life [22, 23]. However, overexpression of the AFP gene is a characteristic feature of human hepatocellular carcinoma. The human AFP gene is about 20 kb long and contains 15 exons and 14 introns [24], The cap site is located 44 nucleotides upstream of the translation initiation site and the TATA box is located 27 nucleotides upstream from the cap site and is flanked by sequences with dyad symmetry. Other sequences in the 5' untranslated region include a CCAAC pentamer, a 14-bp enhancer-like sequence, a 9-bp sequence homologous to the glucocorticoid responsive element, a 90-bp direct repeat, and several alternating purine/pyrimidine sequences.

The AFP promoter is 200 bp upstream of the transcriptional start site. It is regulated by hepatocyte nuclear factor 1 (HNF1), nuclear factor 1 (NF1), and CCAAT/enhancer binding protein (C/EBP). The human AFP enhancer is located between —4.9 and —3.0 kb upstream of the transcriptional start site and consists of at least two functional domains designated A and B which have binding sites for at least four transcription factors, including HNF1, HNF3, HNF4, and C/EBP. The domain B is located at —3.7 to —3.3 kb upstream of the transcriptional start site and is solely responsible for typical enhancer effects, but maximum enhancer activity is observed together with domain A located at —5.1 to —3.7 kb. A hepatoma-specific nuclear factor termed AFP1 is known to bind to an AT-rich sequence, TGATTAATAATTACA, in the B domain of the human AFP enhancer. The AFP enhancer plays a critical role in enhancing AFP gene expression in the fetal liver as well as in hepatocellular carcinoma. The AFP silencer, which is a negative cis-acting element with a consensus sequence, 5'-CTTCATAACCTAATACTT- 3', has been identified [25]. Two transcriptional silencer elements have been identified: the proximal silencer which contains a single copy of the consensus sequence at —0.31 kb and the distal silencer at —1.75 kb which carries four copies of the consensus sequence. Of the two silencers, the distal silencer, exhibits a higher suppressive activity than the proximal silencer. The silencer activity is manifested only when the silencer is located downstream of the enhancer and upstream of the promoter. An inverse correlation exists between the silencer activity and the AFP expression levels in hepatocellular carcinoma cell lines, thereby suggesting the role of the silencer in downregulating the level of AFP expression.

  1. In Vitro and in Vivo Experiments with the AFP Promoter Because of its tissue-specific nature, the AFP promoter has been used in adenoviral vectors for transcriptional targeting of suicide genes in AFP-producing hepatocellular carcinoma (HCC) cells in vitro as well as in vivo. Kaneko et al. [26] developed adenoviral vectors using either the 4.9-kb AFP promoter (Avl AFPTK1) or RSV promoter (AvlTKl) to express HSV-TK gene. In vitro and in vivo cell-specific killing was observed in AFP-producing HuH7 hepatocellular carcinoma cells transduced with AvlAFPTKl and treated with GCV. In contrast to HuH7 tumors, AFP-nonproducing hepatocellular carcinoma SK-Hep-1 cells did not show complete regression when treated with AvlAFPTKl. AvlTKl was able to cause complete regression in SK-Hep-1 tumors. Using a similar approach, Kanai et al. [27] developed adenoviral vectors by incorporating AFP enhancer domains A and B (—4.0 to —3.3 kb) and a 0.17-kb AFP promoter to drive the expression of HSV-TK. These vectors conferred cell-specific killing in AFP-producing HuH-7 and HepG2 cell lines but not in non-AFP-producing HLE and HLF cell lines. Kanai et al. [28] have also reported on the development of adenoviral vectors in which the expression of E. coli CD is driven by the AFP promoter. These vectors were capable of causing regression of HCC xenografts following treatment with 5FC. Arbuthnot et al. [29] analyzed in vitro and in vivo cell-specific expression of the nuclear P-galactosidase using adenoviral vectors containing transcriptional elements derived from either rat AFP or the human insulin-like growth factor II genes. Their results indicate hepatoma cell-specific expression using AFP promoter; however, primary hepatoma cells were poorly infected by these adenoviral vectors. Bui et al. [30] compared adenoviral vector-mediated expression of IL-2 under the transcriptional control of murine AFP promoter and CMV promoters for the treatment of established human hepatocellular xenografts in CB-17/SCID mice. Intratumoral injection of these adenoviral vectors resulted in growth retardation and regression in a majority of animals but with a wider therapeutic index and less systemic toxicity using the AFP vector. Using the AFP promoter and cre-lox based approach Sato et al. [31] were able to achieve strictly tissue-specific expression of LacZ in AFP-producing cells in vitro as well as in vivo in nude mice bearing AFP-producing tumor xenografts.
  2. Prostate-Specific Antigen (PSA) Promoter a. Rationale The gene for prostate-specific antigen, a member of the glandular kallikrein family, was independently characterized by Riegman et al. [32, 33] and Lundwall [34] from a human genomic library. The gene contains five exons and is located on the long arm of chromosome 19, in the region ql3.3-qter [33]. The gene is 7130 bp long and includes 633 bp of 5' and 639 bp of 3' flanking sequence. The promoter region contains a variant TATA box (TTTATA) at position -28 to -23, a GC box at -53 to -48, a CACCC box at -129 to -125. An imperfect palindromic sequence (AGAACAGCAAGTGCT) closely related to the reverse complement of the consensus sequence for steroid hormone receptor binding (TGTACANNNTGTC/TCT), is found at position -170 to —156. In addition, GGGAGGG and CAGCCTC repeats are located in the region —123 to —72. Expression of PSA is primarily detected in human prostate [35-37]. Further, PSA expression has been shown to be androgen-responsive [38], This is achieved by several transcription factors that are involved in regulating prostate-specific antigen gene.

Two functionally active androgen receptor-binding sites or androgen response elements (AREs) have been identified at positions —170 (ARE-I) and —394 (ARE-II) [38-41]. Cleutjens et al. have identified a complex, androgen-regulated 440-bp enhancer (—4366 to —3874) which contains a high-affinity AR-binding site, ARE-III (5'-GGAGGAACATATTGTATCGAT-3'), at position —4200. In subsequent studies, a 6-kb PSA promoter fragment has been shown to confer prostate-specific and androgen-regulated expression of p-galactosidase in transgenic mice [42]. Pang et al. [43] identified an 822-bp PSA gene regulatory sequence, PSAR which when combined with the PSA promoter (PCPSA-P) exhibited an enhanced luciferase activity in LNCaP cells. Upon stimulation with 10 to 100 nM dihydrotestosterone, a more than 1000fold increase in expression was observed as compared to androgen-negative controls. Their studies further suggest that this 822-bp sequence alone could serve as a promoter, thereby indicating that the complete PSA promoter contains two functional domains: a proximal promoter and a distal promoter, which can also function as an enhancer.

Yeung et al. [44] have identified two czs-acting elements within the 5.8kb PSA promoter that are essential for the androgen-independent activity of the PSA promoter in prostate cancer cells. Their studies provide evidence that androgen-independent activation of the PSA promoter in the androgen-independent prostate cancer cell line C4-2 involves two distinct regions, a 440-bp AREc and a 150-bp pN/H, which are responsible for upregulation of the PSA promoter activity by employing two different pathways. AREc confers high basal PSA promoter activity in C4-2 cells, while pN/H is a strong AR-independent positive-regulatory element of the PSA promoter in both LNCaP and C4-2 cells. Further, a 17-bp RI fragment within the pN/H region was identified as the key cis element, which interacts with a 45-kDa prostate cancer cell-specific transcription factor to mediate androgen- and AR-independent transcriptional activation of the PSA promoter. By juxtaposing AREc and pN/H, a chimeric PSA promoter has been created that exhibits 2- to 3-fold higher activity than wild-type PSA promoter in both LNCaP and C4-2 cells. Oettgen et al. [45] have identified a novel prostate epithelial-specific Ets transcription factor, PDEF, that is involved in PSA gene regulation and acts as a coregulator of AR. PDEF acts as an androgen-independent transcriptional activator of the PSA promoter. It also directly interacts with the DNA-binding domain of AR and enhances androgen-mediated activation of the PSA promoter. Thus, strong tissue-specificity of the PSA promoter makes it an ideal candidate for prostate cancer gene therapy. Latham [46] compared tissue-specific expression of luciferase reporter vectors by employing PSA, human glandular kallikrein (hKLK2), and CMV promoters in PSA-positive LNCaP and PSA-negative CoLo320, DG75, A2780, and Jurkat cells. Their studies revealed that minimal 628-bp PSA and hKLK2 promoters showed only low-level androgen-independent expression in both PSA-positive and PSA-negative cell lines. Tandem duplication of the PSA promoter slightly increased expression in LNCaP cells. Addition of the CMV enhancer upstream of the PSA or hKLK2 promoter led to substantially enhanced and nonspecific luciferase expression in all the cell lines. By placing a 1455-bp PSA enhancer sequence upstream of either the PSA or the hKLK2 promoter, a 20-fold increase in tissue-specific luciferase expression was observed. Tandem duplication of the PSA enhancer increased the expression 50-fold higher than either promoter while retaining tissue specificity. The expression from all the enhancer constructs was 100-fold above the basal levels upon induction with androgen dihydrotestosterone.

b. In Vitro and in Vivo Experiments with the PSA Promoter These enhancer sequences were incorporated in adenoviral vectors to express enhanced green fluorescent protein (EGFP) and nitroreductase. The results indicate low-level expression of EGFP by PSA enhancer promoter in LNCaP cells and no expression in non-PSA-producing EJ cells when compared with CMV promoter-driven EGFP. However, the PSA enhancer promoter was able to direct expression of comparable levels of nitroreductase in a tissue-specific manner in LNCaP cells alone. These transduced LNCaP cells upon treatment with CB1954 exhibited cytotoxicity. A replication-competent adenoviral vector CN706 in which the E1A gene is under the transcriptional control of the PSA enhancer/promoter has been shown to exhibit selective toxicity toward PSA-expressing prostate cancer cells [47]. Martinello-Wilks et al. [48] examined the efficacy of adenoviral vectors with a 630-bp PSA promoter-driven HSV-TK and E. coli purine nucleoside phosphorylase (PNP) genes for their ability to kill androgen-insensitive prostate cancer cell line PC-3 tumor xenografts in a nude mouse model. Both HSV-TK and E. coli PNP-expressing adenoviral vectors were able to achieve significant tumor regression in vivo following

GCV or 6MPDR treatment. Gotoh et al. [49] developed transcriptionally targeted recombinant adenoviral vectors by incorporating either 5837-bp long or 642-bp short PSA promoter elements to drive the expression of HSV-TK. The long PSA promoter was shown to have superior activity over the short promoter and was more active in C4-2 cells than in LNCaP cells. In vitro expression of TK conferred marked killing of C4-2 cells upon acyclovir treatment. Administration of this virus in an in vivo subcutaneous C4-2 tumor model, followed by acyclovir treatment, revealed significant inhibition of tumor burden. Lee et al. [50] demonstrated tissue-specific growth suppression of PSA-positive and -negative cell lines by transfecting PSA promoter enhancer-driven p53 tumor-suppressor genes. Recently, human prostate cancer- and tissue-specific genes P503, P540S, and P510S have been identified using a combination of cDNA library subtraction and high-throughput microarray screening by Xu et al. [51]. It would be interesting to characterize the promoter region of these genes and use it in developing transcriptionally targeted adenoviral vectors.

  1. Osteocalcin (OC) Promoter (Fig. 3)
  2. Rationale Osteocalcin (bone y-carboxyglutamic acid (Gla)-containing protein (BGP)) is a 50-amino-acid, 5.8-kDa, major noncollagenous protein found in adult bone and has been shown to be transcriptionally regulated by 1,25-dihydroxyvitamin D3 [52, 53]. The human, rat, and murine osteocalcin genes have been cloned and each consists of four exons and three introns [54-57]. Montecino [58] reported that the key promoter elements are located in two DNase I-hypersensitive sites. The proximal hypersensitive site ( — 170 to —70) includes sequence motifs that specifically interact with basal transcription factors such as Msx [59-61], HLH protein Id-1 [62], AP-1 [63], a bone-specific nuclear-matrix-associated protein, NMP-2 [64],

Nonprostatic Ceil

Prostate Cancer Cell

Figure 3 The specific ability of a tissue specific promoter such as the osteocalcin promoter to produce HSV-TK in a prostate cancer allowing cell death on prodrug (ACV) administration while sparing nonprostate cell by not allowing osteocalcin promoter activation.

and a member of the AML family of transcription factors [65, 66, 71]. The distal hypersensitive domain (—600 to —400) contains the vitamin D-responsive element (VDRE, —465 to —437), which interacts with the VDR-RXRa complex in a ligand-dependent manner [67-69]. Montecino et al. [70] have demonstrated that the promoter segment —343 to —108 is critical for inducing both proximal nuclease hypersensitivity and basal transcriptional activity and the DNase I hypersensitivity at —600 is not essential for vitamin D-dependent transcriptional upregulation. Two additional NMP-2 sites (site A, -604 to -599; site B, -440 to -435) have been identified in the sequences flanking the distal DNase I-hypersensitive domain that might support specific interactions between the nuclear matrix and the OC gene promoter [64, 71]. Analysis of the 5' flanking sequence of rat osteocalcin gene reveals a modular organization of the promoter consisting of the TATAAAA sequence between —31 and —25 and the CCAAT sequence between —92 and

  • 88 [72], Lian et al. [55] identified a 24-nucleotide regulatory sequence, 5'-ATGACCCCCAATTAGTCCTGGCAG-3', in the proximal promoter region with a CAAT motif as a central element, and have designated this sequence as an osteocalcin (OC) box since only two nucleotide substitutions are found in the rat and human osteocalcin genes in this region. Hoffman et al. [59] reported that the OC box is located at nucleotide positions between —99 and —76 and TATA box containing a consensus glucocorticoid-responsive element (GRE) between —44 and —31. The stimulation of osteocalcin gene expression by 1,25-dihydroxyvitamin D3 is associated with sequence-specific binding of nuclear factors to a 26-bp sequence, 5'-CTGGGTGAATGAGGACATTACTGACC-3', located between —462 and —437. This sequence contains a region of hyphenated dyad symmetry and shares homology with consensus steroid-responsive elements. The promoter region has been shown to contain two sites of an E-box motif (a consensus binding site for HLH proteins) termed OCE1 (CACATG at -102) and OCE2 (CAGCTG at -149) [62], Mutagenesis studies have indicated that osteoblastic-specific gene transcription is regulated via the interaction between certain E-box binding transcription factors in osteoblasts and the OCE1 sequence in the promoter region of the osteocalcin gene. Banerjee et al. [63] demonstrated that an AML-1 binding sequence within the proximal promoter (nt —138 to —130) contributes to 75% of the level of osteocalcin gene expression. The promoter region is not GC-rich and does not contain a consensus sequence for the SP1 binding site [73]. Theofan et al. [74] performed a detailed analysis of the BGP promoter region. Three regulatory elements that share partial homology with the consensus sequence for the GRE have been identified at nucleotide positions —356, —178, and —68, respectively. In addition, two sequences related to the consensus sequence for the metal ion-responsive element (MRE) have been identified at positions
  • 190 and —143. An octanucleotide sequence, TGCAGTCA, is located directly adjacent 3' to the second MRE. Two other sequences that share homology with the cAMP-responsive element are found at —437 (TGAGGACA) and —392 (TCACGGCA). The BGP promoter region also contains several pairs of inverted repeat sequences that form regions of dyad symmetry. Three particularly long regions of imperfect dyad symmetry are located between -523 and -504, -234 and -214, and -51 and -28. An octanucleotide palindromic sequence from —134 to —127 partially overlaps both a putative MRE and a cAMP-responsive element. A short sequence, GCAG, or its complement, CTGC, is repeated 17 times. A region of alternating purines and pyrimidines at location —90 to —81 from the CAT box has the potential to form a Z-DNA structure which may be important in gene regulation. A 7-bp osteocalcin silencer element, 5'-TGGCCCT-3', has been located between +29 and +35 position in the first exon of the human osteocalcin gene, while two silencer elements, 5'-CCTCCT-3' (nt +106 to +111 and +135 to +140) and 5'-TTTCTTT-3' (nt +118 to +124), have been located in the first intron of the rat osteocalcin gene [75, 76].

b. In Vitro and in Vivo Experiments with the OC Promoter Ko et al. [77] developed an osteocalcin-promoter-driven TK-expressing recombinant adenoviral vector to achieve tissue-specific killing of osteosarcoma cells in experimental animal model. Administration of this vector followed by acyclovir treatment led to a significant growth inhibition of osteosarcoma in an experimental animal model. Cheon et al. [78] used a chemogene therapy approach by combining OC-promoter-driven TK expression and acyclovir with a methotrexate treatment regimen in nude mice bearing either subcutaneous human osteosarcoma (MG-63) or rat osteosarcoma (ROS) xenografts. Their results indicate that osteosarcoma tumor growth was more efficiently inhibited due to synergistic effects of combined methotrexate and acyclovir treatment. Shirakawa et al. [79] further demonstrated the potential utility of an adenoviral osteocalcin promoter-mediated suicide gene therapy for osteosarcoma pulmonary metastasis in nude mice. Hou et al. [80] demonstrated osteoblast-specific gene expression in adherent bone marrow cells using a 1.7-kb rat OC-CAT gene. Recipient mice were shown to be positive for osteoblast-specific expression following bone marrow transplantation.

Using a replication-defective adenovirus, Ad-OC-TK, we have completed a phase I clinical trial that demonstrated the expected safety profile and gene transfer that we expected. Eleven men with recurrent or metastatic prostate cancer were enrolled in a phase I intralesional dose-escalating trial, combining two Ad-OC-TK injections with 3 weeks of valacyclovir administration. In summary, this was well tolerated at all doses reaching a maximum of 5 x 1010 pfu (or 1 x 1012 vp) in patients in the high-dose group. Viral distribution studies revealed that after intralesional administration the patients demonstrated a measurable viremia for 2-3 days. Despite the presence of viral particles at these time points, no patient demonstrated hepatotoxity with valacyclovir administration. This is in direct contrast to intralesional delivery of Ad-RSV-TK

to the prostatic recurrence, in which patients experience hepatotoxicity upon prodrug administration. Finally, comparison of biopsy specimens prior to the first (day 0) and second (day 7) injection and at the end of the study (day 30) revealed successful gene transfer at day 7 by immunohistochemical staining for HSV-TK and some evidence of tumor destruction by day 30. These expected and encouraging results have led us to propose a phase I trial to test the transcriptional ability of the osteocalcin promoter to regulate adenoviral replication in a similar format.

  1. MN/CA9 Promoter a. Rationale The human MN/CA9 gene has been isolated, sequenced, and characterized by Opavsky [81]. This gene is a member of the carbonic anhydrase (CA) family, which codes for a diverse group of catalysts of the reversible conversion of carbon dioxide to carbonic acid. MN/CA9 expression has been detected in several types of carcinomas including renal, ovarian, and cervical, as well as in normal gastric mucosa [82-85]. The complete genomic sequence of the MN/CA9 gene including the 5'-flanking region encompasses 10.9 kb with a coding sequence comprising of 11 exons. The MN/CA9 protein contains 459 amino acids with a molecular weight ranging from 54 to 58 kDa. MN displays CA activity and binds zinc [86]. The nucleotide sequence close to the 5' end shows 91.4% sequence homology to the U3 region of the long terminal repeats (LTRs) of the human HERV-K endogenous retroviruses [87]. This LTR-like sequence is 222 bp long with an A-rich tail at its 3' end. Analysis of the MN/CA9 promoter region between —507 and +1 upstream of the transcription initiation site indicates that despite the presence of 60% GC residues, the additional features of TATA-less promoters are absent, but the presence of consensus sequences for API, AP2, and p53 transcription factor binding sites has been demonstrated [88-90]. Functional characterization of the 3.5-kb MN 5' upstream region by deletion analysis led to the identification of —173 to +31 fragment as the MN promoter. The promoter region lacks the CpG-rich islands that are typical for TATA-less promoters but contains two nonoverlapping consensus initiator sequences required for promoter activity.
  2. In Vitro and in Vivo Experiences with MN Initial in vitro studies with this promoter driving luciferase expression demonstrated tumor specificity for both renal cell carcinoma and cervical carcinoma. Based on the expression assays, we have constructed an oncolytic adenovirus with the MN promoter which has demonstrated 40- to 100-fold increased killing in human renal cell carcinomas compared to control cell lines not expressing this promoter activity. We are currently evaluating this oncolytic vector in animal models of human renal cell carcinoma.
  3. Inducible Transcription

The ability to precisely regulate spatial and temporal expression of a particular gene is likely to have a significant impact in the field of human gene therapy. In order to be effective, such an approach must necessarily fulfill several criteria, including: (1) biological safety, (2) ease of administration, (3) low basal expression, (4) high and gene-specific inducibility, (5) reversibility, and (6) (preferably) of human origin to minimize immunogenicity. A wide variety of inducible systems for regulating gene expression have been developed. These include the use of metal response promoter [91], heat-shock promoter [92], the glucocorticoid-inducible promoter [93], IPTG-inducible lac repressor/operator system [94, 95], tetracycline-inducible system [96], RU486-inducible system [97], ecdysone-inducible system [100], FK506/rapamycin-inducible system [101], hypoxia-inducible factor 1 system [102], radiation-inducible system [103], and the tamoxifen-inducible system [104]. It is beyond the scope of this chapter to provide in-depth information on all of the above-mentioned inducible systems. Consequently, we will focus on those inducible systems that might have the greatest potential for human gene therapy applications.

1. Tetracycline-Inducible System

The tet-inducible system originally developed by Bujard and coworkers [94, 105] is widely used to regulate gene expression. The tet-inducible system is based on the tetracycline resistance operon of E. coli. The system utilizes the specificity of the tet repressor (tetR) for the tet operator sequence (tetO), the sensitivity of tetR to tetracycline, and the potent transactivator function of herpes simplex virus protein VP16. The system is based upon the concept of negatively regulating the transcription of the bacterial resistance gene by tetR protein binding to tetO DNA sequences. Addition of tetracycline or doxycycline causes derepression by binding to the tetR protein, thereby allowing transcription to proceed. This has been achieved by employing a tet transactivator (tTA) which is a chimeric tetracycline-repressed transactivator generated by fusing the carboxy terminal of tetR protein to the carboxy terminal 127 amino acids of VP16. The tTA, when bound to tetracycline, is prevented from binding to seven copies of tetO sequences, which are juxtaposed upstream of a minimal human cytomegalovirus promoter, thereby selectively turning off the transcription of the gene in question. Removal of tetracycline results in binding of tTA to the tetO sequences in the tet-inducible promoter, following which the VP16 moiety of tTA transactivates the target gene by promoting assembly of a transcriptional initiation complex, thereby selectively turning on the gene expression. A recent modification of this system allows for selective induction of gene expression in the presence of tetracycline. In this strategy, a mutated tetR, called reverse tTA (rtTA), has been generated by incorporating 4 amino acid changes into tTA, thereby facilitating rtTA

binding to the tetO sequence in the presence of tetracycline. Another variation involves fusion of tTA with the KRAB repressor domain of the human zinc-finger protein Koxl. Upon binding to tetO sequences, this protein is capable of blocking transcription as far as 3 kb downstream [106]. A further variation has revealed that by placing two minimal promoters in opposite orientations on either side of the tetO sequences, it is possible to simultaneously regulate the expression of two genes from a single plasmid [107], Massie et al. [108] used the tet-inducible system to generate a recombinant adenoviral vector encoding a deletion in the RI subunit of the herpes simplex virus type 2 ribonucleotide reductase. Topical and tetracycline-inducible gene expression in transgenic mice carrying a gene under the tet-inducible promoter has been achieved by adenovirus mediated gene transfer and expression of tTA [109]. Rubinchik et al. [110] developed a tet-inducible, double recombinant adenoviral vector expressing a fusion of murine FasL and green fluorescent protein. In this virus, the tet-responsive element and the transactivator element are built into opposite ends of the same vector to avoid enhancer interference. The in vitro expression of FasL-GFP in various cell lines could be conveniently regulated by tetracycline or doxycycline in a dose-dependent manner.

2. FK506/Rapamycin-Inducible System

The latest in the armamentarium of inducible gene expression systems are the chemical dimerizers that rely upon drug-dependent recruitment of a transactivation domain to a basal promoter to drive the expression of the therapeutic gene. The strategy is based upon generating a genetic fusion composed of a heterologous DNA-binding domain and an activation domain with the drug binding domain, thereby enabling a bivalent drug to crosslink the two proteins and reconstitute an active transcription factor. This is achieved by using small cell-permeable immunosuppressive molecules, FK506, rapamycin, and cyclosporine, to bind members of the immunophilin family. The FK506 molecule binds tightly to the cellular protein, FKBP12, while FK1012, a synthetic dimer of FK506, causes dimerization of several chimeric proteins containing FKBP12 [111], Another synthetic compound, FKCsA, created by fusion between FK506 and cyclosporine A, binds with high affinity to FKBP12 and cyclophilin and has been used for inducible transcription of exogenous genes [112]. However, the most promising results have been obtained using the heterodimerizer rapamycin, which binds simultaneously to the human proteins FKBP and FRAP [113, 114], In this system, transcriptional activation is achieved through rapamycin induced reconstitution of a transcription factor complex formed by coupling of (a) a unique DNA-binding domain, ZFHD, genetically fused to FKBP and (b) the activation domain of the p65 subunit of nuclear factor kappa B (NFkB), fused with the rapamycin-binding domain of FRAP. This novel approach has been successfully utilized for stable in vivo delivery of secreted alkaline phosphate, murine erythropoietin and human growth hormone using eukaryotic expression vectors, adenoviral, retroviral, and adeno-associated viral vectors [115-117]. One of the limitations of this approach is the growth inhibitory and immunosuppressive activity of rapamycin which is due to the inhibition of endogenous FRAP activity [114]. This limitation can be overcome by nonimmunosuppressive analogs (rapalogs) of rapamycin by incorporating mutations in the FRAP domain that accommodate modified drugs [118, 119]. Considerable progress has also been made in designing novel synthetic dimerizers of the ligand for human FKBP12 and mutated FKBP [119-121], These studies are suggestive of the potential utility of this novel approach for human gene therapy applications.

Wang et al. [97] developed a novel regulated transcriptional activator consisting of a truncated ligand-binding domain of the human progesterone receptor, the DNA-binding domain of yeast transcriptional activator GAL4, and a C-terminal fragment of the herpes simplex virus VP16 transcriptional regulator protein. This novel transcriptional activator binds with high affinity to the synthetic progesterone antagonist RU 486 but binds very poorly to progesterone. In conjunction with the target gene containing four copies of the consensus GAL4 binding site, the gene expression was activated only in the presence of RU 486 [97, 98]. Wang et al. [99] also developed an inducible repressor system by substituting the KRAB transcriptional repressor domain for the VP16 transactivation domain. In addition to RU 486, this system can be activated by other synthetic progesterone antagonists at low concentration. The efficacy of this system has been demonstrated using an ex vivo transplantation approach in which cells containing stably integrated chimeric regulator GLVP and a target gene (tyrosine hydroxylase) were grafted in rats. One of the caveats of this system is the low but distinctive basal activity of the GAL4-responsive promoter in the absence of RU 486. Consequently, this system has been refined by designing a synthetic transcription factor which contains a 35-amino-acid truncation of the progesterone receptor rather than the 42-amino-acid truncation [123]. This system exhibits two- to threefold lower basal activities as compared to the earlier version.

10 Ways To Fight Off Cancer

10 Ways To Fight Off Cancer

Learning About 10 Ways Fight Off Cancer Can Have Amazing Benefits For Your Life The Best Tips On How To Keep This Killer At Bay Discovering that you or a loved one has cancer can be utterly terrifying. All the same, once you comprehend the causes of cancer and learn how to reverse those causes, you or your loved one may have more than a fighting chance of beating out cancer.

Get My Free Ebook


Post a comment