a ß-gal = ß-galactosidase; GH = Growth Hormone; SEAP = Secreted Alkaline Phosphatase; VEGF = Epo = erythropoietin; GHRH = Growth Hormone Releasing Hormone.

a ß-gal = ß-galactosidase; GH = Growth Hormone; SEAP = Secreted Alkaline Phosphatase; VEGF = Epo = erythropoietin; GHRH = Growth Hormone Releasing Hormone.

control of an autoinducible promoter that consists of four GAL4 sites upstream of the tk minimal promoter. Because of the residual basal activity of the tk minimal promoter (see above), GS 3.1 was expressed in the absence of the ligand, but only at low levels. Treatment with RU486 activates GS 3.1, which in turn promotes its own transcription, as well as that of the codelivered reporter gene cloned downstream of the classical 6 x GAL4-TATA promoter. In this autoinducible configuration, GeneSwitch displays lower basal activity, and the fold induction thus increases by one order of magnitude following intramuscular delivery in mice. However, expression levels of the target gene are lower than those obtained from more conventional constructs (140).

4. Inducer Drug

RU486 is a small, orally bioavailable, hydrophobic molecule that diffuses passively through cell membranes and freely distributes to several tissues, including the brain. RU486 is approved in several countries in combination with prostaglan-dins as a treatment to terminate pregnancy (141). In this context it is important to observe that RU486 dosages required to activate GeneSwitch are significantly lower than those required to induce abortion (132,134).

RU486 readily activates the GeneSwitch: in general, transgene product is already detectable 3 h after administration of a single dose. A dose-response experiment in mice demonstrated that GeneSwitch is maximally and half-maximally stimulated by intraperitoneal doses of about 300 ^g/kg and 30 ^g/kg, respectively (134). Since the bioavailability of RU486 after oral dosing is 40% in rodents, we can assume that GeneSwitch would be maximally stimulated by an oral dose of 600 ^g/kg (142). A comparison between the pharmacokinetic properties of RU486 in rodents and humans allows the conclusion that maximal and half-maximal stimulation of the system in humans would be achieved by oral doses of approximately 30 mg and 3 mg, respectively (143). These dosages are significantly lower than those required to induce abortion (200-600 mg) and are well tolerated in humans (143). Nonetheless, even at these low dosages, RU486 affects the ovarian cycle and has a contraceptive activity related to the dosing schedule. For instance, a 2-mg daily dose inhibits ovulation, and 5 mg once weekly prevents implantation (144,145).

Finally, RU486 has been recently demonstrated to inactivate the human cytochrome P-405-3A4 (CYP3A4) and is thus expected to increase the bioavailability of several clinically used drugs that are metabolized by CYP-3A4 (146). The potential for drug-drug interaction must be taken into account in planning long-term treatment with RU486.

In conclusion, therefore, the contraceptive effect of RU486 at low doses but also the potential for drug-drug interactions may curtail the acceptance of GeneSwitch in clinical practice.

5. Perspectives

GeneSwitch has been constantly upgraded over the years in order to meet the stringent criteria required for gene therapy applications. Experiments in animals indicated that the system enables transgene regulation, but also suggested that addi tional improvements are required. A major drawback appears to be the ratio of potency vs. leakiness: GeneSwitch system is probably leakier and less potent than the most advanced versions of the tet system upon electroinjection of plasmids into mouse muscles. A direct comparison with the rapamycin-dependent system is not possible because the experimental settings in which the two systems have been so far tested are too different. The recently described autoregulatory circuit has reduced its leakiness but is also less potent and is not amenable for the construction of tissue-specific regulatory systems. In relation to this point, GS 4.0 may represent a major improvement.

The ligand-dependent activator has been partially humanized by using the p65 activation domain as a substitute for VP16: 82% of its amino acid sequence is now of human origin. Nonetheless, the GAL4 DBD may still be immunogenic in primates. In relation to this point, however, it is worth remembering that the system is highly flexible and can accept DBD of human origin (see below, 4-OHT-dependent regulatory system). Indeed, substituting the GAL4 DBD with artificial zinc fingers of predetermined specificity has generated novel RU486-inducible activators. In transient transfection experiments, these activators specifically induced transcription of reporter genes cloned downstream of multiple zinc-finger binding sites in an RU486-dependent manner (147,148).

C. HEA-3: A Humanized TAM-dependent Transcription Activator

1. General Principles

As outlined above, HBD-based chimeric activators have been generated, which carry synthetic polydactyl zinc fingers or the DBD of the yeast GAL4 transactivator. These DBDs enable specific transgene regulation because they recognize DNA sequences that are not the target of mammalian endogenous transcription factors. Nevertheless, these heterologous DBDs may be immunogenic in humans and therefore of limited efficacy for long-term gene therapy applications.

An HBD-based activator containing a DBD derived from a human protein is expected to have a lower potential for immunogenicity. DBDs derived from ubiquitous transcription factors are not appropriate for this purpose, because the responsive promoter would be constitutively activated in every tissue. In theory, however, customized molecular switches, which make use of human DBDs and effectively regulate gene expression in selected tissues, can be constructed. To this aim, two requirements must be fulfilled: first, the DBD must be derived from a transcription activator that is not expressed in the target cell/tissue; second, the inducible gene must be placed under the control of a synthetic promoter that carries only binding sites for that specific DBD. A fully humanized HBD-based activator (called HEA-3) specifically designed to work in muscle cells has been recently described and used to regulate gene expression in vivo (149).

2. Transcription Factor

HEA-3 is composed of three elements. At the N-terminus, it carries the DBD of the transcription factor Hepatocyte Nuclear Factor 1 (HNF-1). HNF-1a and its closely related variant HNF-1 p, are dimeric homeoproteins expressed mainly in hep-atocytes, where they are required for the expression of liver specific genes, and in a few other epithelial cells (150,151). The two proteins are not expressed in muscle (152). The DBD of human HNF-1 was fused in frame with the HBD of the human ERa. To prevent activation by endogenous steroids, an HBD containing the mutation glycine to arginine at residue 521 (G521R) was used. This mutant HBD binds the antiestro-gen 4-OHT, whose structure is shown in Fig. 15, but not endogenous estradiol (E2) (153,154). The activation domain from human p65 protein was added at the C-terminus (78).

3. Target Promoter

As for other regulatory switches, HEA-3 responsive promoters consist of multimeric binding sites cloned upstream of a minimal promoter. Seven tandem repeats of HNF1 binding sites

0 0

Post a comment