One-Step, Two-Step Regulation of Therapeutic Genes

In February The Scientist ran a timely article by Eugene Russo discussing the new and exciting advances in the regulation and targeting of genes for gene therapy procedures.1 The elegant studies from research groups led by James Wilson, Michael Gilman, and Bert O'Malley showed that it is possible to deliver a putative therapeutic gene to specific tissues, in a dormant state, and turn the gene on and off at will by orally administering a second drug. The second drug orchestrates the formation of an active transcription factor complex on the promoter of the transferred gene.2,3 Tests were carried out in both mice and rhesus monkeys over periods of up to one year. The groups used different viral vectors, regulatory systems, activator drugs, and therapeutic genes, but the kinetics, magnitude, and duration of transferred gene expression were comparable in the two studies, and both studies demonstrated physiological impact; in one case increased hematocrit in response to elevated erythropoietin, and in the other, increased body weight responding to induced levels of growth hormone. The value of these results lies not only in the potential for strict and rapid control of therapeutic transgenes by exogenous drugs (that lack detectable side effects), but also in the apparent stability of the injected viral genes. The type of regulation described will have a significant positive impact on the future of gene therapy, and it should be applicable to numerous human conditions that will respond to augmentation, substitution, or even inhibition of a specific gene product. The technique may be viewed as a two-step regulation, the first step being the deposition and intracellular integration of the inactive genetic material and the second step, the introduction of the activator drug to turn on the gene when it is required. One possible limitation of the method in some instances may be the requirement for the "second-step" activator drug, which will require additional monitoring and testing of patients to determine the dose for appropriate regulation of the transgene. Although this is certainly not prohibitive for many medical procedures, there may be conditions when the treatment would be improved by a more direct communication between the disease and the therapeutic gene. Work from several groups, including our own, suggests that it may be possible in some instances to circumvent the second step of regulation, and activate the dormant transgene directly by an endogenous symptom of the targeted disease. This will be possible under conditions when a disease is stringently linked to a molecular phenotype that can have a direct impact on gene expression. I will discuss below the possibility that hypoxia (oxygen deficit) may be such an endogenous signal that could replace the "second step" of regulation in the gene-based treatment of diseases such as myocardial ischemia, stroke, and diabetic retinopathy. Cells and tissues respond to hypoxia by rapidly, and sometimes strongly, inducing a specific subset of genes. Often these genes improve the functions of the cell or tissue experiencing the hypoxia. Genes involved in anaerobic energy production such as glucose transporters and glycolytic enzymes are induced. Also, genes required for erythrogenesis (erythropoietin) and angiogenesis (vascular endothelial growth factor, or VEGF) are induced, possibly within minutes after the cells become hypoxic. The rapid induction is mediated through the increased activity of one or more specific transcription factors that are highly sensitive to oxygen tension.4 One such factor, called hypoxia-inducible factor 1 (HIF-1), is undetectable in aerobic cells but accumulates within minutes after the cells are made hypoxic. The rapid appearance of HIF-1 is possible because HIF-1 is synthesized continuously; it is rapidly and completely degraded under conditions of normal oxygenation, but the degradation process is blocked, and HIF-1 accumulates immediately under hypoxia. Conversely, when hypoxic cells are reoxygenated, HIF-1 is equally rapidly degraded. The rapid kinetics of these responses are compatible with a possible role for the HIF-1 pathway in the regulation of integrated vectors containing therapeutic genes. A second compatible feature is the high specificity of gene regulation by hypoxia; HIF-1 binds to and activates only genes that have a consensus site containing the sequence element ACGTG and compatible flanking sequences, known collectively as hypoxia responsive enhancers (HREs). Could tissue hypoxia be used to regulate the expression of a therapeutic gene and replace the exogenous "second-step" regulation with an endogenous regulation? The feasibility of such a procedure must be considered within the context of the specific target diseases. The most suitable candidate diseases are ischemia related, where blood flow to a region of tissue is restricted or eliminated, and hypoxia is a central and consistent feature of the disease. This disease category includes, although is not necessarily limited to, myocardial ischemia, stroke, peripheral (limb) ischemia, diabetic retinopathy, and solid tumors. In all of these conditions the affected tissues are subjected to hypoxia. A number of criteria need to be met for ischemia/hypoxia to be a viable endogenous signal to regulate transferred foreign genes. These include: efficient delivery of a tissue-specific vector to the target tissue; long-term stability and viability of expression; low expression of the transferred gene under aerobic (uninduced) conditions; and strong and rapid activation of the gene by (patho) physiological ischemia or hypoxia in vivo. The first three criteria are also applicable to the two-step regulation described above and in references 2 and 3; in one of these studies, an adeno-associated viral (AAV) vector containing the therapeutic gene was shown to be stably expressed for at least one year after injection (confirming previous reports). In the second, a gutless adenoviral (Ad) vector containing the therapeutic gene directed by a liver-specific promoter, was also shown to be stable for at least three months after intravenous delivery. Other work has demonstrated efficient delivery of genetic material to the myocardium, brain, and vasculature using AAV, Ad, and herpes simplex viral vectors. We can conclude that criteria 1-3 can be satisfactorily met. Fulfillment of the fourth criterion involves properties of both the hypoxia-regulated vector and the target tissue. The vector containing the therapeutic gene is required to have low expression in normal (normoxic) tissue and be rapidly activated after the onset of ischemia. Work from a number of groups, including Greg Semenza (Johns Hopkins University), Peter Ratcliffe (Oxford University), and Frank Bunn (Harvard University), as well as our own, indicate that this criterion will also be met. Studies on both isolated cells and whole animals indicate that there is a graded response of hypoxia-activated genes to hypoxia. The effect is apparent in whole animals exposed to 10 percent oxygen or less, and is maximal in cells cultured in vitro under 0.5 percent oxygen. Pathophysiological oxygen tensions are likely to fall within this range, with severely ischemic tissue at the lower end. The level of induced gene expression is gene specific, determined by the characteristics of the individual hypoxia-responsive promoter. It ranges from a low three- to fourfold induction of some glycolytic enzyme promoters, through a median of >10-fold induction of promoters with more active or multiple-copy HREs, to the more than 100-fold activation of the endogenous erythropoietin gene. Therefore, there is little doubt that signals from hypoxic/ischemic tissues are sufficient to exert strong regulation over gene promoters that contain an appropriate arrangement of HREs. In our laboratory, we tested the kinetics of the hypoxia/ischemia-regulatory system directly in the context of myocardial ischemia.5 We injected rabbit left ventricles with a vector containing the luciferase gene linked to a minimal heart-specific promoter containing four tandem HRE repeats. After one week we subjected the same left ventricles to a single 15-minute period of ischemia by ligating the coronary artery, and measured luciferase at intervals after reperfusion. The short ischemic episode initiated a rapid activation of the transferred gene that was close to maximum after one hour of reperfusion, peaked within four hours, and returned to base within eight hours. In parallel experiments, there was no change in the expression of a control vector lacking the HREs. In these experiments we used a medium-strength HRE combination that produced a fivefold maximum induction. More powerful HRE combinations are capable of at least 20-fold activation over basal expression, and the induced expression approaches that of a strong promoter. Keith A. Webster, Ph.D. (kwebster@chroma.med.miami.edu), is professor of pharmacology at the University of Miami and director of molecular cardiology at SRI International, Menlo Park, Calif. He also is the author of U.S. Patent #5,834,306, entitled "Tissue specific hypoxia regulated therapeutic constructs," issued Nov. 10, 1998. These results support the feasibility of a hypoxia/ischemia-regulatory system that could replace "second-step" drugs for the regulation of gene therapy vectors. The system would be applicable to the treatment of conditions in which ischemia and/or hypoxia are integral features of the targeted disease. The dual regulation provides for tissue-specific transgene expression that is restricted physically and temporally to diseased (ischemic) tissue. A specific advantage of such a regulation is the inherently rapid response that will result when the medication is regulated directly by the disease. This may be particularly important in the treatment of diseases such as myocardial ischemia and stroke, where the window of opportunity for treatment is small. In acute stroke patients, only tissue plasminogen activator (tPA) has been proven to provide neurologic improvement, and only when it is administered within three hours of the stroke.6 Other neuroprotective agents such as antioxidants and the antiapoptotic protein Bcl2 are considered promising candidates, again if they can be administered soon enough. Similarly, candidates for the treatment of myocardial ischemia include antioxidants, antiapoptotic gene products, vasodilators, and thrombolitics. As with stroke, early treatment of myocardial ischemia is a critical parameter. It is not difficult to visualize how both of these conditions would benefit from a therapeutic gene product that is synthesized and released as soon as the symptoms of ischemia appear. E. Russo, "New gene therapy systems: advancement in drug delivery," The Scientist, 13[3]:1, Feb. 1, 1999. X. Ye et al., "Regulated delivery of therapeutic proteins after in vivo somatic cell gene transfer," Science, 283:88-91, Jan. 1, 1999. M.M. Burcin et al., "Adenovirus-mediated regulable target gene expression in vivo," Proceedings of the National Academy of Sciences, 96:355-60, Jan. 19, 1999. H.F. Bunn, R.O. Poyton, "Oxygen sensing and molecular adaptation to hypoxia," Physiological Reviews, 76:839-85, 1996. H. Prentice et al., "Regulated expression of a foreign gene targeted to the ischemic myocardium," Cardiovascular Research, 35:567-74, 1997. J.A. Zivin, "Factors determining the therapeutic window for stroke," Neurology, 50:599-603, March 1998. D.W. Losordo et al., "Gene therapy for myocardial ischemia," Circulation, 98:2800-4, Dec. 22, 1998.

One-Step, Two-Step Regulation of Therapeutic Genes

Interested in reading more?

Become a Member of