Hypoxic response elements and Tet-On advanced double-controlled systems regulate hVEGF165 and angiopoietin-1 gene expression in vitro

Angiogenesis in ischemic tissue is a complex and multi-gene event. In the study, we constructed hypoxic response elements (HRE) and the Tet-On advanced double-controlled systems and investigated their effects on the expression of hVEGF165 and angiopoietin-1 (Ang-1) genes in rat cardiomyocytes exposed to hypoxia and pharmacologic induction. We infected neonatal rat cardiomyocytes with recombinant rAAV-rtTA-Rs-M2/rAAV-TRE-Tight-Ang-1 and rAAV-9HRE- hVEGF165. Our results indicated that the viral titer was 1×1012 vg /mL and the viral purity exceeded 98%. hVEGF165 expression was induced by hypoxia, but not by normoxia (P < 0.001). Ang-1 expression was evident under doxycycline induction, but undetectable without doxycycline induction (P < 0.001). Immunofluorescence staining showed that positively stained hVEGF165 and Ang-1 protein appeared only under both hypoxia and doxycycline induction. We demonstrate here that HRE and the recombinant Tet-On advanced double gene-controlled systems sensitively regulate the expression of hVEGF165 and Ang-1 genes in an altered oxygen environment and under pharmacological induction in vitro.


INTRODUCTION
Therapeutic angiogenesis is a therapeutic intervention that introduces exogenous angiogenic growth factors to enhance collateral blood flow in the infarcted myocardium and improve global heart performance [1,2] . Although increasing interest in therapeutic angiogenesis has focused on vascular endothelial growth factor (VEGF), one of the most potent angiogenic factors, some studies indicate that VEGF, as an early-phase angiogenic factor, often produces immature vessels that are unable to create functional collateral development [3][4][5] . Moreover, uncontrolled long-term expression of VEGF delivered by recombinant adeno-associated virus (rAAV) vector in vivo may result in side effects such as hemangioma formation, retinopathy or arthritis [6] .
It is well documented that angiogenesis is a complex event in which multiple angiogenic factors may exert their actions at different phases of angiogenesis [7] . Therefore, for therapeutic angiogenesis, in addition to early angiogenic factors, introduction of late phase angiogenic factors to modify immature vascu-lature and enhance neovessel function has attracted considerable interest. Data indicate that angiopoietin-1 (Ang-1) plays a critical role in promoting maturation of VEGF-induced vessels in the late phase of angiogenesis [8] . However, uncontrolled early Ang-1 expression in ischemic tissues is likely inadequate to remodel VEGF-induced capillaries and does not benefit functional angiogenesis [9,10] . Clinically, due to limitations in gene delivery technology, multiple genes are usually transferred simultaneously. However, expression of multiple angiogenic genes at the same time is not consistent with the physiology of angiogenesis. Therefore, timely, controlled expression of these genes in vivo is required. Thus far, the ideal multi-gene expression control system in ischemic heart is still under development.
Our preliminary experiment confirmed that the hypoxic response element (HRE), as an effective gene switch, reliably induced the expression of the hVEGF 165 gene expression under hypoxia and re-oxygenation both in vivo and in vitro [11][12][13] . In the present study at the cellular level, we employed recombinant Tet (tetracycline)-On together with the HRE system and investigated the feasibility of regulated expression of hVEGF 165 and Ang-1 genes in cardiomyocytes in an altered oxygen environment and under pharmacological induction. We aimed to explore a new approach of therapeutic gene control for further selected multiple gene therapy in ischemic heart disease in vivo.

Tissues and animals
A healthy adult lung was obtained from the National Human Genome Center in Shanghai, China. Total human RNA was extracted using TRIzol Reagent (Invitrogen, USA).
Neonatal Sprague-Dawley rats (1-3 d old, weighing 5-7 g, mean 6.3±0.6 g) were obtained from The Experimental Animal Centre of Xuzhou Medical College. The study protocol was approved by the local Institutional Review Board at the authors affiliated institution. Acquisition of human tissue specimens was carried out in accordance with the institution guide line. All animals received humane care in compliance with the Guideline for Care and Use of Laboratory Animals published by Jiangsu Province, China.

Expression vectors and plasmid construction
The rAAV-9HRE-hVEGF 165 plasmid was a gift from Dr. Hua Su (the Cardiovascular Research Institute, University of California, USA). Recombinant AAV vectors were prepared with a three-plasmid cotransfection system as described previously [13,14] . The titer of rAAV-9HRE-hVEGF 165 was 2×10 12 vector genomes (vg/mL).
For the Tet-On advanced system, a reverse Tet Transactivator (rtTA)-Rs-M2, a p-tetracycline response element (pTRE)-Tight, a pTRE-Tight-luc and AAV expression vectors (pSNAV) were purchased from Clonetech, Inc. (USA) and Vector Gene Technology Ltd (Beijing, China). All vector constructs were confirmed by DNA sequencing.
To facilitate cloning, F and R primers contained a BamHI site at the 5' end of the coding sequence of F and a HindIII site at the 5' end of R (underlined above). After being recovered from the gel using an agarose gel DNA purification kit (TaKaRa, Japan), the PCR-amplified DNA was added with a poly "A" tail with the DNA A-tailing kit (TaKaRa, Japan) and ligated into the pMD18-T simple vector (TaKaRa, Japan). The primary structure of the insert was confirmed by direct sequencing. The fragment of coding Ang-1 was released from the pMD18-T-Ang-1 by digestion with BamHI and HindIII and subcloned into the expression vector pTRE-Tight as described previously [15] . The obtained recombinant eukaryotic expression vector pTRE-Tight-Ang-1 was identified by PCR and then analyzed by the restriction enzymes BamHI/HindIII.
The purified plasmids were used as standards, and the probes were labeled by PCR. The serially diluted virus stock suspension and standards preparations were determined by the dot-blot method using digoxigenin-labeled gene as probe. A viral titer was generated from the detected hybridization signals. The protein traps were obtained from 10% SDS-PAGE electrophoresis and the viral purity was also determined.

Cardiomyocyte culture and identification
Following ether inhalation anesthesia, thoracotomy was performed. The animal heart was rapidly collected, cut into small pieces and digested with trypsin (1.25 mmol/L) for 10 min. The separated cells were centrifuged at 200 g for 7 min and then suspended in DMEM for 2 h (1% CO 2 and 99% air). To inhibit fibroblast growth, unattached cells were treated with 5-bromodeoxyuridine (Brdu, 25 mmol/L) for 24 h before infection and then maintained in DMEM supplemented with 10% FBS.
For determination of cellular purity, cardiac troponin-I (cTnI) staining was completed with a rabbit polyclonal antibody against cTnI (Chemicon, USA) and secondary goat anti-rabbit IgG antibody (Sigma, USA). After hematoxylin and eosin (HE) staining, cardiomyocytes chosen randomly in six visual fields were visualized under a light microscope. The number of cTnI positively stained cells (N2) and nuclei with HE positive staining (N1) were counted. Cardiomyocyte purity was determined based on the following formula: cellular purity = N2/N1×100%.
The cardiomyocytes were divided into eight groups.

Western blotting studies
The cells were solubilized in lysis buffer (100 mmol/L Tris-HCl, 4% SDS, 20% glycerine, 200 mmol/L DTT and protease inhibitors, pH 6.8). Total cellular protein was denatured by boiling for 10 min with an equal volume of 2×Tris-glycine SDS buffer. Protein was separated by 10% SDS-PAGE and transferred to nitrocellulose membrane (Millipore, USA). After blocking with 5% non-fat milk/PBS-T for 3 h at room temperature, the membranes were incubated with a goat anti-Ang-1 antibody (Santa Cruz, USA) and a mouse anti-hVEGF 165 antibody (Sigma, USA), respectively. Then, fluorescently labeled secondary antibody (Rockland, USA) was added for 1 h and subsequently scanned by the Odyssey Infrared Imaging System (Li-Cor Biosciences, USA).

RT-PCR
Total RNA from cultured cells was extracted using RNAiso Reagent. RT-PCR was performed by using 1 μg of total RNA as described by the M-MLV (RNase H) procedure. The primer sequences of hVEGF165 were 5'-CTTGCCTTGCTGCTCTACCT-3' for the forward primer and 5'-CCTTGCAACGCGAGTCTGT -3' for the reverse primer. The primer sequences of Ang-1 were 5'-GCGGATCCATGACAGTTTCCTT-TCC-3' for the forward primer and 5'-GCAAGCTT-TCAAAAATCTAAAGGTCG-3' for the reverse primer.

Statistical analysis
All data are expressed as mean±SD and analyzed by one-way ANOVA and Student-Newman-Keuls test. P values less than 0.05 were considered statistically significant.
The representative photo for the purity of myocardial cells was provided to indicate the identification of myocardial cells (Fig. 2), and the purity of the cardiomyocytes was (90±3)%.

Optimal Dox concentration for induction of Ang-1 expression
Western blotting studies showed that a low basal level of Ang-1 protein expression was detected in the absence of Dox. A higher level of Ang-1 protein expression was achieved with 1 μg/mL Dox induction. Ang-1 expression was found in a dose-dependent manner with Dox concentration (Fig. 3A).
hVEGF 165 and Ang-1 expression RT-PCR determination indicated that hVEGF 165 mRNA bands of about 484 bp were found in group A, B and C, but not in the other groups (P < 0.001). Ang-1 mRNA bands about 1497 bp were found in group C, D and E where Dox was administered, but not identified in group F, G and H (P < 0.001). Both hVEGF 165 and Ang-1 mRNA expression was found in group C only (Fig. 4A).
Western blotting determination indicated that hVEGF 165 protein expression was detected in group A, B and C, but not in the other groups (P < 0.001), while Ang-1 protein expression was found in group C, D and E, but not identified in the other groups (P < 0.001). Both hVEGF 165 and Ang-1 protein expression appeared in group C only (Fig. 4B).
Immunofluorescence analyses also showed that both hVEGF 165 and Ang-1 protein immunofluorescence could be observed in group C. The expressed sites of hVEGF 165 and Ang-1 protein were located in the cytoplasm rather than in the nucleus (Fig. 5).

DISCUSSION
It is generally recognized that angiogenesis is a complex and multi-gene event. Making therapeutic angiogenesis more closely resemble physiological processes in tissue is extremely important. Thus, multi-angiogenic gene control is technically required to achieve this goal. Physiologically, angiogenesis in tissue relies on two sequential phases including an early phase that strictly depends on the presence of an early angiogenic factor (i.e. VEGF). During the VEGFdependent period, endothelial cells, mainly originating from the local pre-existing vasculature, become activated to form a set of immature and irregularly shaped vessels surrounded by a thin endothelial layer. This is A B Fig. 2 The purity of myocardial cells indicating the identification of myocardial cells. A: cTnI positive staining. B: Negative control (×200).  followed by a late phase where vessel maturation induced by a late angiogenic factor is required to ensure the proper acquisition of functional competence of the newly formed vasculature. During this period, VEGF is not only obsolete but is even detrimental for vessel functionality and pericyte regeneration [16,17] . Therefore, the approach developed here is an appealing alternative to the simultaneous delivery of VEGF together with other angiogenic factors. Ang-1, one of the late angiogenic factors, has been proven to be capable of promoting angiogenesis and, in synergy with VEGF, maintaining stability and integrity of the mature vasculature by mediating interaction between endothelial cells and their underlying support cells [18][19][20][21] . Further-more, Ang-1 is the first angiogenic factor identified to exert a protective effect on the vascular endothelial barrier by blocking the action of permeability-increasing mediators, such as VEGF [8] .
Lee et al. [22] reported that VEGF and Ang-1 expression was upregulated in human ischemic myocardium. VEGF expression progressively increased over time after the onset of acute ischemia, peaked at 6 weeks and was followed by Ang-1 expression. Therefore, during therapeutic angiogenesis with VEGF and Ang-1 genes, the accurate and timely control of Ang-1 gene expression behind early hypoxia-mediated VEGF expression may be a key point for multiple gene-induced angiogenesis in an ischemic heart [23,24] . So far, an ideal in- Fig. 3 Determination of inducible gene expression of Ang-1. A: High levels of Ang-1 expression were found with 1 μg/ mL Dox induction. Compared with groups 0, 0.01, 0.1, 5.0 and 10.0, * P < 0.05. B: There was no significant difference among ratios of 1∶1 to 1∶4 (rAAV-rtTA-Rs-M2 / rAAV-TRE-Tight-Ang-1). Compared with groups 1∶1, 1∶2, 1∶3 and 1∶4, * P < 0.05. C: There was no significant difference among concentrations of 10 11 , 5×10 10 and 10 10 (rAAV-TRE-Tight-Ang-1). Compared with groups 10 11 , 5×10 10 and 10 10 , * P < 0.05. D: There was no significant difference among concentrations of 10 11 , 5×10 10 and 10 10 (rAAV-9HRE-hVEGF 165 ). Compared with groups 10 11 , 5×10 10 and 10 10 , * P < 0.05. * * ducible expression control system for Ang-1 in vivo is the so-called TRE-regulated gene system. The Tet-On advanced system has several advantages over other regulated mammalian gene expression systems: low basal expression, high inducibility, no pleiotropic effects and high absolute expression levels. In our study, we successfully constructed a new Tet-On advanced system for Ang-1 gene inducible expression. We found that, at the cellular level, the Tet-On advanced system effectively activated Ang-1 gene expression with the addition of Dox, a tetracycline (Tc) derivative, to the culture medium and was rapidly turned off upon withdrawal of the antibiotic. Furthermore, Ang-1 expression was tightly regulated dosedependently in response to varying concentrations of Dox and was independent of the absolute amount of rAAV-TRE-Tight-Ang-1.Dox is a commonly used broad-spectrum antibiotic. In biology, Dox is usually employed in the inducible expression of Tet-on / Tet-Off system or as an inhibitor for suppression of target gene expresssion [25,26] . Our result suggests that Dox induction was fully necessary for enabling gene control over Ang-1 expression. We noted that the background gene expression of Ang-1 was extremely   A B C bp low in the absence of Dox due to the combined effect of the modified rtTA and TRE-Tight, while the maximal level of Ang-1 gene expression was comparable with that obtained from strong, constitutive mammalian promoters such as cytomegalovirus (CMV) [27,28] . This Tet-On advanced system gave us ready access to tightly regulated, high level Ang-1 gene expression. As for rAAV-rtTA-Rs-M2 and rAAV-TRE-Tight-Ang-1, an optimal proportion should meet the need of obtaining the highest level of Ang-1 expression with the lowest dose of Dox. We found no significant difference of Ang-1 expression in the range of 1:1 to 1:4 (rAAV-rtTA-Rs-M2/rAAV-TRE-Tight-Ang-1). In our recombinant Tet-On advanced system, rtTA-Rs-M2 acted as a control element for Dox. Therefore, under the premise of keeping the highest level of Ang-1 expression, our selected ratio of 1 1:4 guaranteed that the exact amount of rtTA-Rs-M2 was minimized. From the clinical point of view, a low dose of Dox for target gene induction would further decrease possible side effects and improve safety of antibiotic induction.
Concerning the combined efficiency of targeted gene control, hVEGF 165 and Ang-1 expression could be induced by altered oxygen and Dox administration, respectively. We noted that one gene control system had no significant interference on the inducible expression of the other target gene, although expression of hVEGF 165 or Ang-1 in the co-infected group was lower than that in the single gene infected group.
Based on our results, following early hypoxiamediated hVEGF 165 expression, it is feasible to realize timely control over Ang-1 gene expression with a Tet-On advanced system. Our findings present a strategy of enhanced therapeutic angiogenesis with the sequentially controlled expression of multiple angiogenic genes transferred simultaneously in vivo.
Nevertheless, although HRE and the Tet-On advanced systems acted as effective double gene switches in vitro, our new system needs to be validated in animal models in which actual ischemic myocardial pathologies can be reproduced. Additionally, the combined efficiency of VEGF and Ang-1-driven therapeutic angiogenesis should be further evaluated.
In conclusion, HRE and the Tet-On advanced systems have been verified to be new promising gene switches for VEGF and Ang-1-inducible expression control in vitro. Thus, this innovative double gene control system leads to a broader prospect for effective and safe angiogenic therapy in ischemic heart disease in vivo.