Promotion Of Therapeutic Angiogenesis By Peptide Competitively Blocking The Binding Of Hypoxia Inducible Factor-1α And VHL Protein

AIMSIschemic disease including mainly ischemic heart disease, stroke and peripheral vascular disease is a major cause of morbidity and mortality in the world. The initial treatments include modifying risk factors and reducing oxygen demand with medical therapy. When these treatments fail, revascularization to restore blood supply by percutaneous coronary intervention or coronary artery bypass grafting is often necessary. However, a significant number of patients have diffuse coronary artery disease, absent conduits after previous bypass surgery, small distal vessels, and comorbidities that preclude either procedure. This patient population has been estimated to comprise up to 12% of individuals presenting to interventional referral centers. With the widespread use of revascularization , it is likely that the number of patients who will not be suitable for revascularization in the future will increase significantly. Since there is an inverse correlation between infarct size and collateral blood flow, therapeutic angiogenesis presents an exciting new method of improving blood supply to the ischemic myocardium and therefore providing symptomatic relief to a large and growing population of patients.Vascular endothelial growth factors ( VEGF) and fibroblast growth factors (FGF) have been shown to play key roles in therapeutic angiogenesis and are the most extensively investigated and utilized in clinical and preclinical studies. It has been demonstrated that they promote angiogenesis both in animal models and in clinical trials either by gene transfer or by protein therapy. However, the resulting capillaries are often leaky and accompanied by edema, inflammation, and spontaneous hemorrhagic ulcers. Administration of VEGF in combination with other angiogenic factors alleviates some of the vascular defects. Hypoxia in-ducible factor -1 ( HIF -1) is a transcriptional factor playing a key role in an-giogenesis by activating gene transcription of VEGF and other proangiogenic factors. It has also been shown that the resulting blood vessels from HIF -1 were structurally indistinguishable from normal capillaries in both morphology and function. Importantly, as a transcription factor, HIF - 1 can not only promote angiogenesis, but also mediate cellular and systemic homeostatic responses to reduced 02 availability by promoting the transcription of dozens of its target genes, thus contributing to the survival of ischemic/hypoxia cells. Therefore, increasing HIF -1 activity may provide a new promising therapeutic strategy for ischemic disease.HIF -1 is a heterodimeric complex of HIF - la and HIF -1 p. In contrast to HIF - 1(3, both HIF - la protein expression and transactivation activity are regulated by changes in cellular oxygen level. Under normoxic conditions, HIF- la is rapidly degraded by 26S proteasome, which is mediated by von Hippel- Iindau tumor suppressor gene product ( pVHL) , that together with elongin B and C and a RING finger protein, form a complex harboring ￡3 ubiquitin ligase activity. ITie binding of pVHL to HIF - la is dependent on the hydroxylation of proline 563 and 402 of HIF - la by prolyl hydroxylases. pVHL can also form a ternary complex with HIF - la and the co - repressor FIH -1 (factor inhibiting HIF -1). Both VHL and FIH -1 recruit histone deacetylases that may contribute to the loss of HIF - la transcriptional activity under non - hypoxic conditions. Accordingly, the binding of pVHL to HIF - la is critical for the degradation of HIF - la protein and the depression of HIF - la activity at normoxia. Interference with the binding of pVHL and HIF - 1 a may represent a new method for the treatment of ischemic disease.In this project, we first determined the smallest effective binding domain of HIF - la and pVHL, overexpressed the VHL binding domain by transient trans-fection, observed the effect of overexpressed VHL binding domain on the endogenous HIF - la protein level and activity under normoxic conditions, and finally, chemically synthesized peptide containing VHL binding domain and HIV TAT protein, evaluated the effect of synthesized peptide on angiogenesis in vivo.METHODS1. Plasmid Constructs.-1) pCMV - FLAG - GAL4 constructs were constructed by using PCR fragments generated with primer pairs carrying EcoRI and BamHI ends. The PCR fragments were inserted inframe into EcoRI - BamHI - restricted pFLAG -GAL4.2) pSP72 - FLAG - GAL4 constructs were generated by inserting Sad -Smal fragments from the corresponding p CMV - FLAG - GAL4 construct into Sad - EcoRV - digested pSP72.3 ) pCMV - FLAG - GFP - mHIF -1 a ( 559 - 573 ) construct were generated by insterting Hind III - EcoRV GFP fragment from pCMX - SAH/Y145F into NotI - EcoRV - digested, Klenow complemented pCMV - FLAG - GAL4 -mHIF -la(559 -573).4) Amino acid mutations were introduced using the QuickChange site -directed mutagenesis kit ( Stratagene) according to the instructions of the manufacturer, and positive mutants were screened by sequencing.5 ) The inserts generated by PCR and mutations were completely sequenced using the Dyenamic sequencing kit ( Amersham - Pharmacia).2. Cell Culture and Transient Transfections:Mouse brain endothelial cells ( MBEC) and human heptoma cells (HepG2) were routinely maintained in F -12 (HAM) and Dulbeccos minimal essential medium respectively supplemented with 10% fetal calf serum plus penicillin (50 IU/ml) and streptomycin (50 mg/ml). The cells were transfected u-sing Fugene according to the manufacturers recommendations.3. HRE driven luciferase activity assay:Increasing amount of constructs containing different wide - type and mutated VHL binding domain were contransfected with HRE luciferase reporter gene. 44 hours after transfection, whole cell extracts were prepared and luciferase activity were measured using Luciferase Assy Kit. HRE driven luciferase activityrepresents the endogenous HIF - la activity.4. Immunoprecipitation Assays:In vitro immunoprecipation assays were performed to analyze the binding of different VHL binding domain and pVHL. In in vivo immunoprecipitation assays were performed to measue the endogenous HIF - la protein level. Hie precipitated proteins were detected either by autoradiography or by immunoblotting assays.5. Electrophoresis and Immunoblotting Assays:Whole cell extract proteins or immunoprecipitated proteins were separated by SDS - PAGE and blotted onto nitrocellulose filters. The expressions of trans-fected VHL binding domain were detected by anti - FLAG antibody, and the endogenous HIF - la protein was detected using anti - HIF - la monoclonal or polyclonal antibody.6. Confocal Microscopy:HepG2 cells were cultured on cover slips and were fixed by 4% paramalde-hyde at room tempreture after 44 hours transfection. The coverslips were then mounted and the distribution of GFP constructs was observed. Totally about 200 cells were counted and divided into 3 groups. N;absolutely nuclear distribution;N > C: distribution in nucleus is more than that in cytoplasm;N = C: distribution in the nucleus and cytoplasm is equal.7. Mouse Corneal Angiogenesis Assay:According to the method described by Cao, the peptides were implanted under the mouse cornea. 5 days later, corneal angiogenesis was detected by slit - lamp.8. Statistic Analysis:All values were presented by mean ± SD. Student t test was performed a-mong groups. P <0.05 is significan.RESULTS 1. Effect of overexpressed VHL binding domain on endogenous HIF -la protein and activity.1) In vivo immunoprecipitation assays showed that endogenous HIF — 1 a protein was hardly detected under normoxic conditions. Overexpression of VHL binding domain mHIF - la (546 -574) stabilized endogenous HIF - la protein, about 120 KDa.2 ) HRE driven luciferase assay demonstrated that, as a control, increasing amount of overexpressed FLAG - GAL4 had no effect on luciferase activity;however, the overexpression of VHL binding domain mHIF - la (546 -573) increased the endogenous mHIF - la (546 -574) activity, in a dose - response manner. There was significant difference between the two groups.2. Mechanism of the effect of VHL binding domain on endogenous HIF - la protein and activity.In vitro immunoprecipitation assay showed that VHL binding domain with proline 563 mutated to alanine couldnt bind with pVHL. Concomitantly, in cells overexpressing this proline 563 mutated VHL binding domain, endogenous HIF — la? protein degraded rapidly (5 min) during the reoxygenation process after hypoxia, while the overexpression of wide — type VHL binding domain inhibit the degradation. In HRE driven luciferase assay, the mutated VHL binding domain lost the ability to enhance the endogenous HIF - la activity as the wide - type did.3. Identification of essential amino acids among the VHL binding domain.In HRE driven luciferase assay performed in both HepG2 and MBEC cells, in addition to P563 and PYI — DDD mutants that have been shown can not bind with pVHL and lost the positive effects on endogenous HIF - la activity, the mutants D570A, L573A and FL571/573AA abolished the effect as well, indicating that D570 and L573 were playing an important role in the effect of VHL binding domain.4. Delineation of the smallest effective VHL binding domain1). In vitro immnoprecipitation assay showed that full - length HIF - la and HIF - la (546 -574) binded with pVHL very weU, while HIF - la rrag-ment with QL deleted couldnt bind with pVHL. While HIF - la (559 -573) could still bind with pVHL, the ability of HIF-la (560-573,561-573) to bind with pVHL decreased dramatically. 2). HRE driven luciferase assay showed that HIF - la (559 -573) was most effctive in MBEC cells and still effective in HepG2 cells. Its effects were higher than 18 - mer domain encompassing P402.These results indicated that HIF -la (559 -573) is the smallest effective VHL binding domain.5. Impact of the distribution of overexpressed VHL binding domain on its effect1) Endogenous HIF - la degradation assay showed that endogenous HIF -la (559 -573) degraded in both nucleus and cytoplasm with a half - life 6 minutes and 4 minutes respectively. MG132 treatment inhibited the degradation.2) FLAG -GFP - HIF - la(559 -573) was distributed equally in cytoplasm and nucleus and it increased endogenous HIF - la activity dose respon-sively. However, FLAG - GFP - NLS - HIF - la(559 - 573 ) was mainly distributed in the nucleus and its effect on endogenous HIF -1 a activity was significantly less than that of FLAG - GFP - HIF - la(559 - 573 ). While the NLS was mutated and lost the intranuclear signal function, the mutated FLAG - GFP- NLS - HIF - la( 559 - 573 ) was distributed all over the cell and regained the effect on endogenous HIF - la activity.6. Chemically synthesized peptide containing VHL binding domain inhibited the binding of HIF - la and pVHL and performed proangiogenic effectTo test its angiogenic effect in vivo, we chose the peptide transduction domain from the HTV tat protein to allow the intracellular delivery of the chemically synthesized peptide corresponding to mouse HIF - la (559 -573) with P563 in a hydroxylated or unhydroxylated form. We found that in the absence of whole cell extracts, the P563 hydroxylated peptide impaired binding of pVHL and HIF -1 a in vitro, but not the unhydroxylated peptide. In vitro hydroxylation of the unhydroxylated peptide by incubation with whole cell extracts reconstituted thiseffect, but with much lower efficiency. In the in vivo mouse cornea angiogenesis assay, the hydroxylated peptide was able to induce angiogenesis, but not the un-hydroxylated peptide.CONCLUSIONS1. Intracellularly overexpressed VHL binding domain of HIF - la stabilized endogenous HIF - la protein and increased its activity.2. VHL binding domain of HIF - la stabilized endogenous HIF - la protein and increased its activity by competitively binding with pVHL. Mutants of VHL binding domain that could not bind with pVHL lost the effects.3. In addition to P563 and PYI motif, D570 and L573 also played an important role in the effect of VHL binding domain.4. mHIF - la (559 -573) was the smallest effective VHL binding domain within HIF-la.5. It was necessary for VHL binding domain to be distributed all over the cell to exert the best effect. It was not necessary to include a NLS in the peptide designed for proangiogenic agent by upregulating HIF - la passway.6. Chemically synthesized peptide containing VHL binding domain inhibited the binding of HIF -1 a and pVHL and performed proangiogenic effect.