Elucidation of the key role of HIF in the induction of vascular growth in response to hypoxia has led to several approaches to manipulate HIF as a therapeutic strategy which include the administration of modified, hypoxia-independent versions of HIF, as well as alternate strategies that target prevention of proteosomal degradation of HIF-1 a. These approaches are validated by clinical observations such as a correlation of HIF-1a genotype with the extent of collateral vessel development in patients with coronary artery disease . Similarly, in patients with critical limb ischemia, HIF-1a, HIF-2a and VEGF gene expression are up-regulated  while HIF-1 a protein levels and microvessel density are both enhanced in the ischemic muscle of these patients .
To experimentally evaluate neovascularization induced by sustained activation of HIF-1, either wild type (WT) HIF-1 a, or a version of HIF-1 a in which the oxygen-dependent degradation (ODD) domain had been removed (HIF-1aAODD), were expressed selectively in the skin of transgenic mice using a tissue-specific keratin-14 (K14) promoter . Animals expressing K14-(WT)HIF-1a exhibited no increased skin vascularity compared to non-transgenic controls despite confirmed transgene expression, due to the robust cellular mechanisms that regulate unrestrained HIF activity when oxygen is not limiting. In contrast, K14-HIF-1aAODD mice, which lack the oxygen-sensitive degradation domain, exhibited prominent vessels and reddening in exposed skin. Microscopic analysis of sections of skin from these animals documented an increase in the number and diameter of vessels, compared to either control or K14-(WT)-HIF-1a overexpressing mice. As the K14-HIF-1aAODD mice aged, although cutaneous rubor increased with time, there was no evidence of ulceration or angioma formation as had previously been reported by these authors in mice over-expressing a VEGF-A isoform (K14-VEGF-A164; ). In addition, there was no increase in vascular permeability in K14-HIF-1aAODD mice, whereas this had been a prominent feature of mice overexpressing either VEGF-A164  or VEGF-A120 alone .
A therapeutic transgene expressing a constitutively active form of HIF-1 a (HIF-a/VP16) has been constructed by truncating HIF-1a at amino acid 390, thereby removing the oxygen-dependent degradation domain as well as the endogenous HIF-1 activation domains, and replacing them with a constitutively active herpes virus VP16 transactivation domain. The resulting construct, HIF-1a/VP16, initiates expression of a number of genes involved in angiogenesis and possibly arteriogenesis [72, 92-95]. Using a similar approach, Semenza and colleagues  have generated an adenoviral HIF-1a construct termed AdCA5; the HIF-1 a gene within AdCA5 has been modified via deletion and point mutations within the ODD domain that result in constitutive HIF activity. Gene expression profiling using either SAGE in the case of a recombinant adenovirus expressing HIF-1 a/VP16  or microarray analysis in the case of AdCA5  yielded a similar spectrum of expression of genes involved in angiogenesis, such as marked increases in selected angiopoietins, as well as up-regulation of VEGF-A isoforms, PlGF, and PDGF among others. There were a number of differences as well, likely in part due to the fact that the Ad/HIF-1aVP16 was examined in human fetal cardiac myocytes, whereas AdCA5 was studied in human pulmonary endothelial cells. For example, the gene most induced by HIF-1a/VP16 was brain natriuretic protein (BNP), by over 100-fold, while the gene activated most by AdCA5 was angiopoietin-like protein 4, by approximately 15-fold. Both BNP and ANP have been shown to play a role in angiogenesis, in addition to their known vasoactive effects in the vasculature and antiapoptotic as well as anti-hypertrophic activity in cardiac muscle [99-101].
As a test of the hypothesis that exogenous administration of a plasmid encoding HIF-1 a/VP16 could enhance collateral vessel formation, and to compare the potency of HIF-1 a/VP16 with that of VEGF-A as an angiogenic therapy, a plasmid expressing HIF-1a/VP16 has been tested in a rabbit hindlimb ischemia model . Naked plasmid DNA encoding either the HIF-1a/VP16 hybrid gene (pHIF-1a/VP16) or human VEGF165 (pVEGF165) was administered via injection to the hindlimb muscles of rabbits in which ischemia had been induced 10 days previously via excision of a portion of the femoral artery. Physiological findings such as blood pressure and blood flow in the affected limb suggest that HIF-1 a/VP16-induced angiogenesis is at least as effective as that achieved following treatment with human VEGF-A165. The anatomic evidence of revascularization in response to administration of HIF-1a/VP16 was observed both via angiography and histological analyses. Quantification of angiographic recordings established that the number of angiographically visible collateral arteries (i.e., arteries >200^m diameter) in the HIF-1 a/VP16-treated animals was similar to that achieved with phVEGF-A165 and exceeded that of the controls. Histological examination documented an increase in vascularity at the capillary level that also exceeded that in controls for both HIF-1a/VP16 and phVEGF-A165, with a larger increase in the HIF-1 a/VP16-treated animals than in those that received phVEGF-A165. Results of these studies suggest that administration of naked DNA encoding the HIF-1a/VP16 transcription factor may represent a viable treatment strategy for tissue ischemia. The bioactivity of a recombinant adenoviral vector expressing HIF-1 a/VP16 (Ad2/HIF-1a/VP16) has also been documented in the rabbit ischemic hindlimb model. Two doses of Ad2/HIF-1a/VP16 (1 x 109 and 1 x 1010 viral particles) were evaluated and found to be comparable pHIF-1a/VP16 with respect to the ability to stimulate therapeutic angiogenesis (Genzyme Corporation: unpublished data).
Experiments have also been performed in animal models of coronary ischemia to determine whether expression of the constitutively active HIF-1 a/VP16 construct would enhance blood flow. In a rat acute myocardial infarction (MI) model, it was found that intramyocardial injection of a plasmid expressing HIF-1a/VP16 resulted in decreased infarct size and increased myocardial blood flow . Similarly, the effect of intramyocardial injections of Ad2/HIF-1a/VP16 targeted to the central ischemic zone also has been examined in a porcine ameroid constrictor model. Injections of Ad2/HIF-1a/VP16 resulted in improvements in both blood flow in the ischemic territory, and myocardial function when compared to a negative control . Studies in transgenic mouse models have further supported the therapeutic potential of HIF-1 a in coronary ischemia. In an acute MI model, overexpression of HIF-1 a from a cardiac specific promoter resulted in increased vessel density in and surrounding the infarcted tissue, resulting in reduced infarct size and improved cardiac function . Conversely, cardiomyocyte-specific deletion of HIF-1 a caused a reduction in HIF target gene expression, contractile function and defects in cardiac myocyte energy metabolism under normoxic conditions .
Recently published experiments  describe the ability of the constitutively active AdCA5 HIF construct, discussed above, to increase blood flow in a novel rabbit hindlimb ischemia model in which the femoral artery is occluded by an endovascular approach. Both increased blood pressure and decreased angiographic perfusion time were observed in the affected limb of Ad5CA-treated animals relative to control animals at 14 days post-treatment. Evidence for arteriogenesis (i.e., an increase in vessel diameter due to vascular remodeling) was obtained both angiographically and histologically. Capillary density was also increased in the Ad5CA-treated animals, suggesting that angiogenesis had also occurred. Administration of AdCA5 to normal rabbit adductor muscle resulted in enhanced expression of multiple pro-angiogenic cytokines, including PDGF-B, PlGF, VEGF, MCP-1 and SDF-1 genes. A similar HIF-1a mutant containing amino acid substitutions at the two conserved proline residues (P402A, P563A) and expressed from an adeno-associated virus vector was tested in the murine hindlimb  and found to promote enhanced capillary number and sprouting in the transduced muscle. Increased vascular perfusion was observed in the absence of increased perivascular edema, which is more typical in vessels treated with a VEGF-A isoform alone. Co-expression of either Ang-1 or PDGF-B with the mutated HIF-1 a did not impact vessel permeability or function. Administration of a gene encoding a stabilized form of HIF-1 a lacking the ODD (HIF-1 a/AODD) via peptide-DNA nanoparticles embedded in a fibrin matrix promoted angiogenesis in a murine wound healing model . Compared with the control, VEGF-A165 protein in fibrin, HIF-1a/AODD-containing nanoparticles induced a higher proportion of mature vessels. Together, the results of these studies provide further support for the notion that therapeutic neovascularization results from the coordinated expression of several angiogenic growth factor genes.
An alternate strategy has been delivery of the peptide PR39, an innate immunity associated protein enriched in proline and arginine that, in addition to its antimicrobial activity, also increases endogenous HIF-1 activity due to inhibition of HIF-1a degradation by the proteosome . Transgenic mice expressing PR39 selectively in the heart exhibited increased baseline vascularity and enhanced expression of VEGF, flt-1, flk-1, eNOS and FGFR1. These mice also exhibited significantly higher coronary blood flow at a given perfusion pressure than control animals. Infusion of the peptide in a murine infarct model of angiogenesis also resulted in an increase in vascularity along the infarct border zone. For a review of the role of PR39 and other components of innate immunity in the regulations of angiogenesis and arteriogenesis see .
The bioactivity and apparent lack of toxicity exhibited by PR39 and HIF-1 a/VP16 suggests that a small-molecule, pharmacological approach to therapeutic angio-genesis could also be successful. This strategy would involve stabilization of HIF-1 a via the use of small-molecule inhibitors of the HIF prolyl hydroxylases (PHD's). A known inhibitor of the related procollagen prolyl hydroxylases, FG0041, has been shown to inhibit the HIF PHD's and lead to the accumulation of transcriptionally active HIF in vitro . This compound had been tested as an anti-fibrotic in a rodent model of myocardial infarction and shown to have a positive effect on left ventricular function . The cardioprotective effect was likely due to HIF-1 a (or HIF-2a) stabilization as a similar study using a different inhibitor (FG2216) documented increased levels of HIF-2a protein in the heart . Stabilization of HIF-1 a by administration of known (procollagen) prolyl hydroxylase inhibitors has also been shown to induce HIF target gene expression and angiogenesis in a rat model . Ideally, any therapeutic candidate should be specific for the HIF PHD's, and the identification of compounds that inhibit the procollagen, but not the HIF prolyl hydroxylases, suggests that this might be possible . This goal will be facilitated by the elucidation of the crystal structure of the PHD's as has been recently reported for PHD2 . Alternatively, expression of the PHD's as well as FIH has been inhibited through the use of specific siRNA's . Potential therapeutic application of this latter approach was demonstrated by Natarajan et al.,  who documented up-regulation of HIF-1 a protein and protection from ischemia-reperfusion injury in mice following administration of a prolyl-hydroxylase targeted siRNA. Degradation of HIF-1 a under normoxic conditions has also been blocked with polypeptide inhibitors of the HIF-1 a/VHL interaction [118, 119]; these have been shown to promote an angiogenic response in vitro and in vivo .
2.1. Clinical Experience with a Constitutively Active HIF-1a Construct
A Phase I clinical trial-evaluating Ad2/HIF-1a/VP16 for the treatment of peripheral arterial disease has been completed (Rajagopalan, et al., manuscript in preparation). In the trial, patients were treated via intramuscular injection with Ad2/HIF-1a/VP16 at a range of doses. Additional Phase II trials to confirm and extend these results are have been initiated in peripheral vascular disease patients with severe intermittent claudication and in patients with critical limb ischemia. In addition, a Phase I study of patients with advanced coronary artery disease undergoing coronary artery bypass surgery, who have an area of viable ischemic muscle but no suitable target vessels for venous or arterial bypass, is also nearing completion. In summary, the use of analogues of HIF-1a that permit sustained - at least days to weeks - expression and activity of this transcription factor may prove beneficial in the treatment of refractory tissue ischemia.
Was this article helpful?