- Email: [email protected]
Reoxygenation Promotes Myocardial Angiogenesis via anÈNF κ B-dependent Mechanism in a Rat Model of Chronic Myocardial Infarction
J Mol Cell Cardiol 33, 283–294 (2001) doi:10.1006/jmcc.2000.1299, available online at http://www.idealibrary.com on
Hypoxia/Reoxygenation Promotes Myocardial Angiogenesis via an NFB-dependent Mechanism in a Rat Model of Chronic Myocardial Infarction Hiroaki Sasaki1, Partha S. Ray1, Li Zhu2, Hajime Otani2, Takayuki Asahara3 and Nilanjana Maulik1 1
Department of Surgery, University of Connecticut Health Center, Farmington, Connecticut 060301110, USA, 2Department of Thoracic and Cardiovascular Surgery, Kansai Medical Center, Osaka, Japan and 3Tufts University School of Medicine , St. Elizabeth’s Medical Center, Boston, Massachusetts 02135, USA (Received 3 August 2000, accepted in revised form 8 November 2000, published electronically 8 January 2001) H. S, P. S. R, L. Z, H. O, T. A N. M. Hypoxia/Reoxygenation Promotes Myocardial Angiogenesis via an NFB-dependent Mechanism in a Rat Model of Chronic Myocardial Infarction. Journal of Molecular and Cellular Cardiology (2001) 33, 283–294. Therapeutic angiogenesis achieved either through the use of discreet angiogenic proteins or by gene therapy is fast emerging as a highly attractive treatment modality for ischemic heart disease. Herein we examine a novel method of stimulating myocardial angiogenesis by hypoxic preconditioning at both capillary and arteriolar levels, and the potential role of NFB in mediating such a response. We also investigate the functional relevance of such treatment by assessing whether the induced neovascularization can help preserve left ventricular contractile functional reserve in the setting of developing heart failure secondary to myocardial infarction. Male Sprague–Dawley rats were randomly divided into eight groups: normoxia+sham surgery (NS), normoxia+permanent left anterior descending coronary artery (LAD) occlusion (NMI), hypoxic preconditioning+sham surgery (HS), hypoxic preconditioning+permanent LAD occlusion (HMI), PDTC (NFB inhibitor)+hypoxic preconditioning+LAD occlusion (PHMI), PDTC+normoxia+LAD occlusion (PNMI), PDTC+hypoxic preconditioning+sham surgery (PHS) and PDTC+normoxia+sham surgery (PNS). Rats in the preconditioned groups were subjected to systemic hypoxemic hypoxic exposure (10±0.4% O2) for 4 h followed by a 24-h period of normoxic reoxygenation prior to undergoing LAD occlusion. Rats in the normoxia groups were time matched with the preconditioned group and maintained under normoxic conditions for the 28-h period prior to LAD occlusion. The HMI group displayed significant increases in capillary as well as arteriolar density after 2, 4 and 7 days post-operation compared to the NMI. Prior PDTC administration prevented such increases in the PHMI group and effectively abolished the pro-angiogenic effect of hypoxic preconditioning (HP). One week after sham surgery or LAD occlusion, rats underwent a pharmacological stress test with dobutamine in progressively increasing doses which revealed significantly elevated values of dp/dtmax at each dose point in the HMI group compared to the NMI or PHMI groups. Hypoxic preconditioning also decreases endothelial cell injury as determined by the extent of endothelial cell apoptosis using anti-VWF factor labelling and TUNEL assay. The results suggest that HP stimulates myocardial angiogenesis via redox-regulated transcription factor, NFB-dependent pathway to an extent sufficient to exert significant preservation of contractile functional reserve in a rat model of myocardial infarction progressing to heart failure. 2001 Academic Press K W: Angiogenesis; Myocardium; Infarction; Nuclear factor-kappa B; Hypoxia; Reoxygenation.
Please address all correspondence to: Nilanjana Maulik, Ph.D., Molecular Cardiology Laboratory, Cardiovascular Division, Department of Surgery, University of Connecticut School of Medicine, Farmington, CT 06030-1110, USA. Fax: (860) 679-2451. E-mail: [email protected]
0022–2828/01/020283+12 $35.00/0
2001 Academic Press
284
H. Sasaki et al.
Introduction Therapeutic angiogenesis achieved either through the use of discreet angiogenic proteins or by gene therapy is fast emerging as a highly attractive treatment modality for ischemic heart disease. It is well known that coronary collateral vessels help protect or reduce damage to ischemic myocardium after coronary obstruction. Factors which stimulate collateral growth can therefore be expected to exert a protective effect against myocardial infarction. Among the various triggers of angiogenesis, tissue hypoxia has been identified as being a particularly important stimulus for the induction of new vessel growth, particularly at the capillary level.1 Capillary endothelial cells play the most important role in giving rise to several types of functionally and morphologically distinct vessels: under the control of angiogenic stimuli they can re-enter the cell cycle, degrade basement membrane and migrate, forming capillary sprouts which then differentiate to form new vessels.2,3 In the myocardium, hypoxia usually occurs during ischemic episodes. Such ischemic hypoxia is often followed by reperfusion. It is during this subsequent reperfusion period that reoxygenation occurs and reactive oxygen intermediates are formed.4,5 In observations by Kuroki et al., reactive oxygen intermediates were found to increase vascular endothelial growth factor (VEGF) expression in vitro.6 Reoxygenation of human retinal epithelial cells in vitro and ocular reperfusion in vivo increased retinal VEGF mRNA levels.6,7 Administration of antioxidants prior to reoxygenation/reperfusion effectively inhibited such increases in VEGF mRNA. Lelkes et al. have shown that hypoxia/reoxygenation, but not hypoxia alone, cause the formation of reactive oxygen species (ROS).8 In their study, the rate of tubular morphogenesis by human microvascular cells was increased three-fold by hypoxia/reoxygenation. Tube formation in response to hypoxia/reoxygenation as well as under normoxic conditions was inhibited by various ROS antagonists, in a dose-dependent manner, indicating the essential role of ROS in initiating angiogenesis.9 A wide array of pathological conditions are known to generate ROS including inflammation, atherosclerosis, ischemia and reperfusion, etc. The vascular endothelium is thus a target of oxidative stress under a variety of conditions. In fact, Lelkes et al. have demonstrated that administration of ROS inhibitors such as pyrollidine dithiocarbamate (PDTC) inhibit activation of NFB and tube formation in vitro.8,9 Recently, it has been shown that following myocardial infarction various angiogenic growth factors
are upregulated in the human heart as an intrinsic adaptive response.10 Clinical gene trials with VEGF are showing promise as a potential means of augmenting such adaptive angiogenic responses. However, in order to achieve clinical relevance, the enhanced angiogenesis must be of sufficient extent so as to effectively reduce loss of cardiac function after ischemic injury. In this study, we sought to evaluate the functional relevance of stimulated myocardial angiogenesis by utilizing an animal model of chronic myocardial infarction progressing to heart failure and examining left ventricular response during pharmacological stress testing with dobutamine as a measure of cardiac reserve. This study constitutes the first report in vivo which demonstrates: (1) non-lethal moderate hypoxic challenge can decrease endothelial cell injury after myocardial infarction; (2) such preconditioning stimulates myocardial angiogenesis at both capillary and arteriolar levels; (3) the redox-sensitive transcription factor NFB plays an essential role in the initiation of such myocardial angiogenesis; and (4) myocardial angiogenesis stimulated by prior hypoxia/reoxygenation helps to significantly improve postmyocardial infarction (MI) left ventricular contractile functional reserve.
Materials and Methods All animals used in this study received humane care in compliance with the principles of laboratory animal care formulated by the National Society for Medical Research and Guide for the Care and Use of Laboratory Animals prepared by the National Academy of Sciences and published by the National Institutes of Health (publication no. NIH 85-23, revised 1985). The experimental protocol was performed after receiving approval from the Institutional Animal Care Committee.
In vivo hypoxia/reoxygenation Two hundred and eighty-eight rats weighing 275–300 g were divided randomly into eight groups as outlined in Figure 1. Based on previous experiments performed in our laboratory, we have determined that a 4-h duration of systemic hypoxemic hypoxia followed by 24 h of reoxygenation induces a very strong induction of angiogenic factors such as VEGF, Ang-1/Ang-2 and their receptors Flk-1, Flt-1 and Tie-2 determined by Western blot analysis11 (Table 1). We therefore used this duration
Hypoxia/Reoxygenation Induces Myocardial Angiogenesis via NFB
285
Figure 1 Outline of experimental protocol. Rats randomly assigned to various experimental groups (n=12) were subjected to LAD ligation following different exposure protocols. NMI group: Normoxia+LAD occlusion; HMI group: Hypoxia/Reoxygenation+LAD occlusion; PHMI group: PDTC+Hypoxia/Reoxygenation+LAD occlusion; PNMI group: PDTC+Normoxia+LAD occlusion. Rats which were assigned randomly to the next four groups served as the respective sham-operated controls for groups NMI through PNMI, respectively. NS group: Normoxia+Sham surgery; HS group: Hypoxia/Reoxygenation+Sham surgery; PHS group: PDTC+Hypoxia+Sham surgery; PNS group: PDTC+Normoxia+Sham surgery.
Table 1 Relative protein expression profile compared to the normoxic control (% normoxic control) after different duration of hypoxia followed by 24 h of reoxygenation Angiogenic factors and their receptors
1 h H/R (%)
2 h H/R (%)
4 h H/R (%)
VEGF Flk-1 Flt-1 Ang-1 Ang-2
50.8 41.5 325.0 23.8 31.3
51.9 22.0 201.0 30.0 14.8
Tie-1
533.0
586.7
Tie-2
93.9
120.5
50.0 12.8 149.9 50.0 Same as control Same as control 53.4
of hypoxic exposure to effect hypoxic preconditioning in all our preconditioned animal groups. Rats in the hypoxic preconditioned group (HMI) were subjected to a 4-h hypoxic challenge,
using a gas mixture (10% O2/90% N2),12 in an anesthesia chamber. An in-flow aperture was used for the gas mixture and an out-flow aperture connected to a gas absorption cannister (A.M. Bickford Inc., Wales Center, NY, USA). A KE-25 galvanic cell O2 sensor (Kent Scientific, Litchfield, CT, USA) was used to monitor continuously the %O2 concentration, the signal being amplified through a TRN-005 Amplifier (Kent Scientific). The O2 concentration was gradually lowered from an ambient %O2 of 20.9% down to 10% over the course of 30 min. Once stabilization was obtained, the inflow rate was adjusted so as to maintain the %O2 at 10±0.4% for 4 h. These rats were then maintained in a normoxic environment for 24 h until preoperative preparation. Rats in the PDTC+hypoxic preconditioned group (PHMI) were treated similarly except for PDTC administration (150 mg/kg b.w. i.p. injection)12 1 h prior to the hypoxic challenge. Rats in the PDTC control group
286
H. Sasaki et al.
(PNMI) were administered PDTC and were thereafter maintained in a normoxic environment in a time-matched manner (29 h) until preoperative preparation. Rats in the control group (NMI) were similarly maintained in a normoxic environment in a time-matched manner (29 h) until preoperative preparation. Six animals were excluded from each permanent occlusion group either due to post-op mortality or morbidity within 24 h or based on infarct size being less than 50%. Thus sample size for permanent occlusion groups was n=6. Although the hypoxic-preconditioned permanent occlusion group did display improved survival postop, the small sample size (despite the large number of total animals used) did not permit analysis for assessing any interventional benefit in reducing morbidity or mortality outcomes. No animals were excluded from sham-operated groups. Thus sample size for sham-operated groups was n=12 for function comparisons but n=6 for morphometric analysis chosen randomly.
Survival surgery and LAD occlusion After the 24-h reoxygenation period, time-matched normoxic period rats were anesthetized with Ketamine HCl (100 mg/kg i.p.) and Xylazine (10 mg/ kg i.p.).13–16 Cefazolin (25 mg/kg i.p.) was administered by i.v. injection as preoperative antibiotic cover. The animals were placed in the supine position and body temperature was maintained at 37°C by means of a water-circulating thermalheating pad. A midline neck incision was made and the trachea was identified. Tracheostomy was performed by making a vertical incision approximately 2 mm long in the trachea. A polyethylene tube (2 mm o.d.)17 was introduced through the opening and positive pressure ventilation with room air was commenced (stroke volume 12 ml/ kg,17–19 tidal volume 2.5 ml20 and respiratory rate 70/min21) using a small animal respirator (model 683, Harvard Apparatus, Holliston, MA, USA). The tube was secured in place by means of stay sutures. An intercostal thoracotomy in the left fourth intercostal space was performed.13,14,16,22 The heart was exposed and the pericardium incised. The left coronary artery was identified between the left atrial appendage and the pulmonary artery outflow tract. A 6-0 suture was passed beneath the vessel 1–2 mm from its origin. Care was taken not to enter the ventricular cavity. In the sham-operated group (NS), the suture was left in place up until chest closure but was not ligated. In the permanent
coronary-occlusion groups (NMI, HMI) electrophysiological changes. After completion of all protocols, the chest wall was closed in three layers using 4-0 silk sutures. A catheter fashioned from PE-190 tubing attached to a 12-ml syringe with stopcock was used as a chest tube to evacuate air from the pleural space as the last stitch was tightened. Post-operative pain relief was achieved by the administration of Buprenorphine (0.05–2.5 mg/kg s.c.).
Non-survival surgery and baseline cardiac function One week after the operative intervention, the rats were anesthetized and ventilated as described above. A small incision was made to the right of the midline in the neck. The right internal jugular vein was identified and a PE 50 catheter was introduced into the vein. The proximal end of the catheter was connected to a low-pressure transducer. The inserted catheter tip was then advanced until it reached the right atrium and the central venous pressure (CVP) signal was obtained. The same procedure was used to place a catheter in the right carotid artery where the catheter was used to monitor aortic blood pressure. The inserted tip of this catheter was advanced down until it reached the left ventricular lumen and the left ventricular pressure (LVP) signal was obtained. All pressure signals were monitored, analysed and recorded in real time using the Digimed data acquisition and analysis system (Micromed, Louisville, KY, USA). Heart rate (HR), developed pressure, and dp/dtmax were all calculated from the continuously obtained LVP signal.
Cardiac stress testing with dobutamine After baseline left ventricular functional parameters were recorded, rats were subjected to pharmacological cardiac stress testing with dobutamine infusion in progressively incremental doses to reveal the extent of left ventricular contractile functional reserve. Dobutamine was infused i.v. through the venous catheter by means of a microinjection pump (Harvard Apparatus) to achieve doses of 1, 2, 3 and 5 g/kg/min. Based on previous studies in our laboratory, we have found that administration up to a dose of 5 g/kg/min is necessary to observe the plateau of the dose-response curve. Administration was maintained at each dose for a duration of 2 min. dp/dtmax and other left ventricular functional parameters were recorded for each dose 90 s after the start of each dose.
Hypoxia/Reoxygenation Induces Myocardial Angiogenesis via NFB
Tissue retrieval and processing After measurement of all functional parameters, the animals were heparinized systemically (Heparin Sodium, 500 IU/kg b.w., i.v. injection, Elkins-Sinn Inc., Cherry Hill, NJ, USA) to prevent intravascular coagulation. The animals were killed with 1 ml of 30 m KCl injected i.v. Hearts were excised, the ventricular lumen was flushed clean with Ringer’s lactate solution and the ventricles were either frozen in liquid nitrogen for later use in biochemical studies (Gel Shift assay for NFB) or were sectioned approximately 5–6 mm from the apex into approximately 2-mm thick transverse sections. The sections were then fixed in 10% buffered formalin and embedded in paraffin using standard procedures.
Measurement of capillary and arteriolar density The standard deparaffinization protocol was used. Endothelial cells were labeled using mouse monoclonal anti-CD31/PECAM-1 (1:100, Pharmingen) followed by a biotinylated horse anti-mouse secondary antibody (1:200 dilution). The reaction product (brown) was visualized with DAB substrate using the Vector ABC Vectastain Elite Kit (Vectorlabs, Burminghame, CA, USA) and was counterstained with methyl green (Vectorlabs). On separate slides, vascular smooth muscle cells were labeled using mouse monoclonal anti-smooth muscle actin (1:50, Biogenex, San Ramon, CA, USA) followed by a biotinylated rat-adsorbed horse anti-mouse secondary antibody (1:200 dilution). The reaction product (violet) was visualized with VIP substrate using the Vector ABC Vectastain Elite Kit (Vectorlabs). All sections were dehydrated in two changes of 100% ethanol, cleared in Toluene and permanently mounted using Polymount (Poly Scientific R&D, Bayshore, NY, USA). Capillaries were identified and recognized by their size, a thin layer of endothelial cells and a lack of smooth muscle cells. Similarly, arterioles were identified and recognized by their size and a thin layer of smooth muscle cells. At a total magnification of 400×, eight non-overlapping random fields were selected from endocardial regions of non-infarct/borderline zones of the left ventricle of two sections from each heart (16 fields per region per heart, 64 fields per region per group). Images were captured and stored in digital .tif file format for later image analysis. Counts of capillary density and arteriolar density/ mm2 were obtained after superimposing a calibrated morphometric grid on each digital image using
287
Adobe Photoshop Software. Counting was performed in a blind manner by two separate investigators. Counts of particular type (e.g. capillary) in the same region (e.g. epicardial) from hearts obtained from animals in the same experimental group (e.g. hypoxic preconditioned+LAD occlusion) were averaged to yield the numerical value of the particular density measurement for the group. Counts of capillary density and arteriolar density/ mm2 were obtained.
Measurement of percent circumferential left ventricular infarct size After following the standard deparaffinization protocol described above, the extent of left ventricular myocardial infarct was delineated histochemically by utilizing Mallory’s Trichrome staining technique for collagen present in the infarct scar.23 The hearts which underwent 50% infarction by LAD occlusion were considered for the experimental evaluations. Electrophoretic mobility assay (EMSA) Nuclear proteins were isolated from the heart to estimate NFB translocation.24 In short, 150 mg of left ventricle from heart tissue was homogenized in ice-cold Tris-buffered saline (TBS) and centrifuged at 3000×g for 5 min at 4°C. The pellet was resuspended by gentle pipetting in 1.5 ml of ice-cold hypotonic buffer containing 10 m HEPES, pH 7.9, 10 m KCL, 0.1 m EDTA, 0.1 m EGTA, 1 m DTT, 0.5 m PMSF, and 1 each of aprotonin, pepstatin and leupeptin. The solution was allowed to swell on ice for 15 min after addition of 100 l of 10% Nonidet P-40, the tube was vortexed vigorously for 45 s. This homogenate was centrifuged for 30 s at 4°C in a microcentrifuge tube. The supernatant contained the cytoplasmic protein. The nuclear pellet was resuspended in a solution containing 20 m HEPES, pH 7.9, 0.4 NaCl, 1 m EDTA, 1 m EGTA, 1 m DTT, 1 m PMSF and 1 each of aprotonin, pepstatin and leupeptin. The tubes were shaken vigorously at 4°C for 30 min on a shaking platform. The nuclear extracts were stored at −70°C. Protein concentration was estimated by using Pierce protein assay kit (Pierce Chemical Company, Rockford, IL, USA). NFB oligonucleotide (AGTTGAGG-GGACTTTCCCAGG) [2.5 l (20 ng/l)] was labeled using T4 polynucleotide kinase as previously described. The binding reaction mixture contained in a total volume of 20.2 l, 0.2 l DTT (0.2 ), 1 l BSA (20 mg/ ml), 4 l PdI-dC (0.5 g/l), 2 l Buffer D+ 4 l,
288
H. Sasaki et al.
Buffer F, 2 l 32P-oligo (0.5 ng/l) and 7 l extract containing 10 g protein. Composition of Buffer D+ was 20 m HEPES, pH 7.9, 20% glycerol, 100 m KCl, 0.5 m EDTA, 0.25% NP 40 while Buffer F contained 20% Ficoll 400, 100 m HEPES, pH 7.9, and 300 m KCl. Incubation was carried out for 20 min at room temperature. Ten microlitres of the solution was loaded onto a 4% acrylamide gel and separated at 80 V until the dye hit the bottom. After electrophoresis, gels were dried up and exposed to Kodak X-ray film at −70°C. The purity of the nuclear extracts was examined using lactate dehydrogenase (LDH) as a cytosolic marker, and it was found that 1.5% of total myocardial tissue LDH was present in the nuclear fraction. The super shift assays and the competition assays were performed to measure that the signal was specific for NFB. For the supershift assays, anti-p65 or anti-p50 antibodies were added, separately or together, to the reaction mixtures immediately after addition of radiolabeled probe. For the competition assays, increasing molar excess of unlabelled NFB oligonucleotide was added into separate reaction mixtures.
Evaluation of endothelial cell apoptosis The formaldehyde-fixed left ventricle was embedded in paraffin, cut into transverse sections (4-m thick) and deparaffinized with a graded series of xylene and ethanol solutions. The sections are treated with 40 g proteinase K for 15 min at 37°C and then incubated in absolute methanol containing 0.3% hydrogen peroxide for 30 min. Non-specific binding is blocked with normal goat serum. The sections are assayed for in situ terminal transferase labeling (TUNEL) using Apop Tag Plus (Oncor Inc., Gaithersburg, MD, USA). Negative control slides are processed with the TdT enzyme excluded. The occurrence of apoptosis in endothelial cells are demonstrated in the sections which are first stained with TUNEL (FITC staining). The sections were then incubated with rabbit polyclonal anti-von Willebrand factor (Dako) as a primary antibody followed by incubation with tetrarhodamine isothiocyanateconjugated goat anti-rabbit IgG as a secondary antibody and viewed with double immunofluorescence confocal laser microscopy. For the quantitative purpose, the number of TUNEL-positive endothelial cells was counted on 60 high power fields (HPF, magnification ×600) from the endocardium through the epicardium of the left ventricular free wall in three sections from each heart. Representative confocal images shows von Wil-
lebrand factor-positive endothelial cells (strong red staining in their cytosol) which are negative for TUNEL staining (absence of green staining in the nucleus) as well as those positive for TUNEL staining (magnification ×1200).
Statistical analysis Results are expressed as mean±standard error of the mean (...). Differences between groups were tested for statistical significance by one-way analysis of variance (ANOVA) and Student’s t-test, with the Bonferroni adjustment for multiple comparisions. The dose–response curves to dobutamine of the various groups were tested for significant differences by repeated measure ANOVA. Differences were considered significant at P<0.05.
Results Effect of hypoxia/reoxygenation (H/R), H/R/PDTC and PDTC only on microvascular density Quantitative measurement of capillary density At a total magnification of 400×, eight non-overlapping random fields each were selected from the endocardial regions of non-infarcted zones of the left
Figure 2 Left ventricular endocardial capillary density. Tissue sections were processed for CD31 staining and eight non-overlapping random fields were selected from endocardial regions of non-infarcted of the left ventricle of two sections from each heart (16 fields per region per heart, 64 fields per region per group, magnifications of 400× were used, n=6). Images were captured and stored in digital .tif file format for image analysis. Where ∗ P<0.001 compared to sham control, † P<0.001 compared to HMI. Ε Sham, ∆ NMI, HMI, ; PHMI, Φ PNMI.
Hypoxia/Reoxygenation Induces Myocardial Angiogenesis via NFB
289
numerical value of the particular density measurement. No statistically-significant differences in arteriolar density between groups were demonstrable in hearts evaluated 2 days or 4 days post-op (except for the HMI and PHMI groups). The arteriolar density was, however, significantly elevated in the hypoxic preconditioned group (HMI) after 1 week post-op (2.23±0.12 v 1.53±0.02 in NMI and 1.51±0.08 counts/mm2 in PNMI) (Fig. 3). No such increase in arteriolar density was observed in the group which had been administered PDTC prior to hypoxic preconditioning (1.6±0.02 counts/mm2, PHMI group). Figure 3 Left ventricular endocardial arteriolar density. Left ventricular tissue sections were labelled using monoclonal anti-smooth muscle actin and eight non-overlapping random fields were selected from endocardial region of the left ventricle. Where ∗ P<0.001 compared to sham control, † P<0.001 compared to HMI. Ε Sham, ∆ NMI, HMI, ; PHMI, Φ PNMI.
ventricles. Quantitatively the HMI group displayed a statistically significant increase in capillary density as early as 2 days post-operation when compared to sham-operated groups and/or the control group (1862±67 counts/mm2 v 1642±18 counts/mm2) (Fig. 2). The capillary density continued to increase in the HMI group and was maintained at a higher level 4 days and 7 days post-operation (2035±53 and 1963±45 counts/mm2). Capillary density was maintained close to baseline levels in the NMI group after 2 days post-operation (1591±38 counts/ mm2), but thereafter displayed a significant decrease after 4 days and 7 days post-operation (1217±47 and 1313±55 counts/mm2, respectively) of hypoxic preconditioning. The incremental effect on capillary density was not displayed in the group which had received PDTC treatment prior to hypoxic preconditioning (PHMI). This was evident from the significant decrease in capillary density observed in this group after 2 days and 4 days post-op (1122±60 and 1430±70 counts/mm2 respectively), returning to a baseline-comparable value only after 7 days post-op (1663±78 counts/ mm2) (Fig. 2).
Quantitative measurement of arteriolar density At a total magnification of 200×, five non-overlapping random fields were selected from the endocardial regions of non-infarcted zones of the left ventricles. Arteriolar counts (purple color) for each field in the same region was averaged to yield the
Effect of H/R, H/R/PDTC and PDTC only on cardiac function: dose–response study of pharmacological stress by dobutamine infusion on left ventricular pressure (LVP), maximal dP/dt, left ventricular end diastolic pressure (EDP) and heart rate (HR) Indices of baseline cardiac contractile function (as assessed after 1 week post-operation, or after a 1week time-matched duration in case of control rats) were significantly different in all operation groups (NMI, HMI, PHMI and PMI) irrespective of preoperative treatment protocol, when, compared to control rats which had not been subjected to any survival surgery procedure. The LVSP was significantly depressed and the LVEDP significantly increased compared to normal baseline values in the unoperated rat (Table 2). Baseline values of HR, EDP, LVP and dp/dtmax in the sham-operated groups (NS, HS, PHS, PNS), however, did not differ significantly from those of the control rats which did not undergo survival surgery. Sham functional data were not shown because there were no statistically significant differences between them and their inclusion would distract from the data. There were also no demonstrable statistically significant differences between functional data of any of the sham groups and the control baseline group. Therefore, functional data only from the control baseline group has been included in the figure and table as being representative of all the sham groups. Differences between sham groups were only demonstrable in capillary density (CD): CD was significantly higher in the hypoxic preconditioned sham group after 7 days (1814±56 counts/mm2) compared to the control baseline group after 7 days (1642±18 counts/mm2). Again, there were no statistically significant differences between CD measurements in any of the other sham groups with the control baseline group. In the figure on
290
H. Sasaki et al.
Table 2 Measurement of hemodynamic parameters under dobutamine stress test Dobutamine (g/min/kg)
0
Heart rate (bpm) Sham 235.1±11.0 NMI 197.7±16.0 HMI 207.6±8.4 PHMI 213.1±11.2 PMI 178.4±11.9 Left ventricular systolic pressure (mmHg) Sham 105.3±11.5 NMI 96.9±4.9 HMI 85.9±7.3 PHMI 92.9±5.2 PMI 92.5±2.6 Left ventricular end-diastolic pressure (mmHg) Sham 5.93±0.66 NMI 14.93±1.54∗ HMI 13.15±2.38∗ PHMI 14.23±1.42∗ PMI 15.92±1.56∗
1
2
3
5
229.5±9.7 227.9±12.1 223.3±9.2 224.0±13.5 202.3±5.3
253.5±14.9 238.5±13.4 237.4±10.9 231.5±12.0 213.2±6.0
263.1±13.7 251.3±15.0 254.2±13.0 237.2±13.5 215.7±10.1
242.2±9.9 260.5±14.1 269.4±14.6 243.9±11.0 231.7±11.8
119.7±4.8 111.3±6.3 100.5±8.9∗ 100.8±6.7∗ 105.1±4.3
146.2±5.6 115.8±9.4∗ 107.8±8.6∗ 109.1±7.3∗ 110.0±7.0∗
158.0±6.2 124.1±10.2∗ 111.1±7.6∗ 114.0±7.3∗ 117.5±5.2∗
153.1±3.4 130.7±6.1∗ 114.2±7.3∗ 117.4±5.0∗ 114.7±6.8∗
6.28±0.68 17.68±2.04∗ 16.75±2.14∗ 12.92±1.34∗ 18.17±2.56∗
7.28±0.41 18.07±2.61∗ 18.08±1.99∗ 14.33±1.23∗ 19.10±3.22∗
8.18±0.66 23.3±1.96∗ 20.88±1.73∗ 18.73±1.38∗ 22.97±2.27∗
9.05±0.89 24.87±2.55∗ 22.58±2.22∗ 21.48±1.85∗ 26.55±2.67∗
∗ P>0.05 compared with sham.
capillary density, only permanently occluded groups are compared.
Response to dobutamine Pharmacological cardiac stress testing with dobutamine infusion in incremental doses revealed differences in the extent of cardiac contractile reserve between groups. This was evident from differences in the extent of change of dp/dtmax values displayed by the various groups during the course of such stress testing (Fig. 4). The hypoxic preconditioned group displayed significantly elevated contractile reserve at each dose point of evaluation compared to all other operation groups, but the extent of such reserve was lower than that displayed by sham-operated groups or by control rats (all of which exhibited similar and statistically indifferent values of dp/dtmax throughout the course of stress testing). This enhanced preservation of contractile reserve in the hypoxic preconditioned group was apparent at the 1 g/kg/min dose (6026±444 mmHg/s in HMI group v 5075±355 mmHg/s in NMI group, 4403±384 mmHg/s in PHMI group and 5031±151 mmHg/s in PMI) and this trend persisted at all subsequent doses up to 5 g (7630±729 mmHg/s in HMI v 6084±531 in NMI, 5292±665 in PHMI and 6111±537 in PMI). This capability of hypoxic preconditioning to better preserve contractile functional reserve in the infarcted heart was, however, not evident in the
Figure 4 Dose–response curves of dp/dtmax during dobutamine stress test. Changes in dp/dtmax in mm of mercury/s (mmHg/s) from baseline with increasing dobutamine boluses (0–5 g/kg/min) in 7 days MI hearts. The dose–response curve for dobutamine for the various doses were tested for significant differences by repeated measure ANOVA. Differences were considered significant at P<0.05. Where ∗ P<0.01 compared to Sham, † P<0.01 compared to NMI, PHMI, PNMI.
group which had received PDTC treatment prior to the hypoxic preconditioning (PHMI). The performance of this latter group on such cardiac stress testing resembled that of the NMI group, as also the group which had received only PDTC and no hypoxic preconditioning (PMI).
Hypoxia/Reoxygenation Induces Myocardial Angiogenesis via NFB
Figure 5 Effect of hypoxia/reoxygenation and PDTC on NFB/DNA binding activity in rat myocardium subjected to whole-body hypoxic preconditioning for various periods of hypoxia. Rats were subjected to 1 h (H1), 2 h (H2) and 4 h (4 h) of whole-body hypoxia followed by 2 h of reoxygenation as described in the Materials and Methods. Gel-shift analysis revealed that baseline control (BL) hearts have minimal NFB activity whereas hypoxically adapted hearts demonstrated significantly increased DNA binding activity of NFB (A) in H1 and H2 heart samples whereas H4 group of hearts showed normal baseline DNA binding activity like BL. PDTC prior to hypoxic treatment (B) inhibited NFB binding activity completely as shown in 1 h and 2 h hypoxic group pretreated with PDTC PH1 and PH2.
291
Figure 6 Estimation of apoptotic cells. Apoptotic cells expressed as a percentage of total number of endothelial apoptotic cells in the various groups. Results are expressed as mean±... for n=6 per group. Each assay was determined in triplicates. ∗ P<0.05, compared to the baseline control group, † P<0.05 compared to NMI group. Where the extent of apoptosis was shown after 2 days, 4 days and 1 week of post-op. BL Ε, NMI ∆, HMI , PHMI ;, PNMI Φ.
cells when compared to the HMI group after 2, 4 days and also after 1 week. PNMI treatment groups also have similar results as shown in Fig. 6.
Effect of H/R, H/R/PDTC and PDTC only on DNA binding activity of NFB
Discussion Gel-shift analysis revealed the time course of hypoxia-induced NFB/DNA binding activity in rat myocardium. NFB/DNA binding activity was detectable at low levels in control hearts, increased significantly after 1 h of hypoxic exposure followed by 2 h of reoxygenation to 30% which was further increased (40%) after 2 h of whole body hypoxia and 2 h of reoxygenation (Fig. 5). PDTC completely suppressed the hypoxia/reoxygenation induced NFB/DNA binding activity completely when administered i.p. 1 h before hypoxia.
Evaluation of endothelial cell injury by examining the extent of endothelial cell apoptosis under the exposure of H/R, H/R/PDTC and PDTC only We were unable to detect any apoptotic endothelial cells in the control baseline samples (Fig. 6). However, significant numbers of apoptotic cells were visible in the NMI groups after 2 and 4 days after post-op compared to the HMI groups. After 1 week the number of apoptotic cells went down in the NMI group but remained significantly higher when compared to the HMI group. PHMI group of rat hearts revealed a significant number of apoptotic
The present study reports for the first time that systemic hypoxemic hypoxic exposure induces myocardial angiogenesis in a rat model of myocardial infarction as evidenced by increased capillary/arteriolar density. Myocardial adaptation to intermittent hypoxia appears to be a highly promising approach to reduce cellular injury due to ischemia and reperfusion. It has been shown that hypoxic adaptation could prevent myocardial stress, ischemic damage, and cardiac arrhythmias in both animals and humans.25 Such adaptation also restricts the impairment of cardiac electric stability and contractility in acute myocardial infarction.26 Hypoxia has the potential to promote the induction of angiogenesis by altering angiogenic regulatory proteins, metabolic enzymes and transcription factors.27–29 Tissue hypoxia, once achieved, exerts such a proangiogenic action through various angiogenic factors, the most notable being vascular endothelial growth factor, which has been chiefly associated with initiating the process of angiogenesis through the recruitment and proliferation of endothelial cells. However, according to Lelkes et al.8 brief exposure to hypoxia followed by reoxygenation significantly accelerates the formation of capillary
292
H. Sasaki et al.
formation by HMVEC cultured on Matrigel and hypoxia/reoxygenation mediated tube formation does not involve upregulation of PKC or the release of angiogenic growth factors, but rather it depends on the generation of reactive oxygen species (ROS) and the subsequent activation of redox regulated nuclear transcription factor, NFB. Evidence indicates that oxidative stress/free radicals lead to the activation of NFB, which in turn induces the expression of genes.30 NFB is the first eukaryotic transcription factor shown to be influenced directly by oxidative stress.31 Recently it was demonstrated in mouse model that NFB activation is obligatory for retinal angiogenesis to occur and it was also documented that the administration of pyrrolidine dithiocarbamate (PDTC) suppressed such retinal neovascularization.32 The present report demonstrates that hypoxic preconditioning triggers myocardial angiogenesis via an NFB-mediated pathway in a rat model of myocardial infarction. Significant increases in capillary as well as arteriolar density were observed in the 2, 4 and 7 days HMI myocardial sections when compared to the NMI, PHMI and PMI samples. Another interesting finding was to verify the hypothesis that non-lethal moderate hypoxic challenge can decrease endothelial cell injury after myocardial infarction as was determined by the relative percentage of endothelial apoptotic cells under various conditions as shown in Figure 6. Interestingly, we found that hypoxic challenge followed by reoxygenation can reduce apoptosis significantly and decrease endothelial cell injury in vivo. Supplementary to the above findings, pharmacological cardiac stress testing with dobutamine proved successful in effectively revealing the ability of hypoxia/reoxygenation pretreatment to exert a pronounced and marked effect in maintaining LV contractile reserve. Such means of measuring LV contractile reserve is especially crucial to detect effective reserve cardiac function in animal models where heart failure is the final pathophysiological manifestation, which in this case was still in the process of developing secondary to chronic MI. Differences in LV contractile function between operated groups were not apparent from baseline functional criteria, as might be expected in the early post-MI state when intrinsic compensatory mechanism are activated on account of the necessity of the animal to maintain a minimal level of circulatory function necessary to stay alive, independent and regardless of preoperative treatments. Differences, however, may become apparent at a later time period when heart failure is no longer sufficiently compensated. Interestingly, although the extent of myocardial
infarction (50%) was statistically insignificant between the various groups compared, the HMI group still demonstrated a significantly enhanced level of contractile function. This would seem to suggest that in the HMI group, the major contributor to such preservation of contractile function was not a reduction in the extent of myocardial necrotic cell death or infarction, but perhaps an enhanced contractile ability of the non-infarcted myocardium in the HMI group. The capillary density and arteriolar density data suggest that the enhanced angiogenesis displayed in the non-infarcted zone of the HMI group has some role to play in effecting the enhanced contractile function. Furthermore, all potentially beneficial and cardioprotective effects of hypoxia/reoxygenation were effectively reversed by prior treatment with PDTC indicating the essential and critical role of NFB in mediating hypoxia/ reoxygenation induced myocardial angiogenesis. It seems NFB—iNOS—NO pathway may be also involved in the fascinating process of angiogenesis. Taken together, it appears that myocardial angiogenesis is precisely controlled by a redox switch and that oxygen-free radicals appear to function as a signaling molecule in this process. In our rat model significant NFB binding activity was observed during hypoxia which was inhibited by PDTC (Fig. 5). In the PHMI group of rats significant inhibition of capillary density was observed when compared with the HMI group of rats (Fig. 2) after 1 week of coronary artery occlusion. This is the first in vivo evidence indicating the involvement of an inducible transcription factor NFB in myocardial angiogenesis. Moreover, the effect on contractile reserve observed in the HMI group at 1 week post-op may be attributed at least partially to the concurrent increase in capillary density. While the increase is clearly suggestive of new capillary growth, the maintenance of capillary density near normal 2 and 4 days post-op in the HMI group is perhaps best explained as enhanced endothelial cell survival early after LAD occlusion. This also explains why there was an acute and significant drop in capillary density values in the other operated groups (NMI, PHMI). The marginal yet significant increase in arteriolar density in the HMI group compared to NMI group of rats indicated that the process of neovascularization is not restricted to the capillary level but extends to the arteriolar level as well. PDTC clearly inhibited the hypoxic preconditioning mediated increased in arteriolar density. It is logical to propose that reoxygenation preceded by hypoxia induces a cascade of events involving VEGF and its receptors expression as well
Hypoxia/Reoxygenation Induces Myocardial Angiogenesis via NFB
as Ang-Tie system protein expression11 might also be one of the possible mechanisms besides NFB– iNOS–NO pathway by which hypoxia reoxygenation triggers myocardial angiogenesis in this model. It is very difficult at this point to speculate on the exact pathway by which hypoxic preconditioning stimulates myocardial angiogenesis. However, inhibition of angiogenesis by the administration of PDTC was successful in demonstrating the essential and important role of NFB in myocardial angiogenesis. This document clearly emphasizes the role of NFB activation in mediating angiogenesis in response to hypoxia in a rat model of chronic myocardial infarction.
Acknowledgement This study was supported by National Institutes of Health Grants HL 56803, HL 22559, HL 33889, HL 34360.
10.
11.
12.
13.
14.
15.
References 1. S D, I A, S D, K E. Vascular endothelial growth factor induced by hypoxia-initiated angiogenesis. Nature 1992; 359: 843–845. 2. M T, A K. Endothelial receptor tyrosine kinases involved in angiogenesis. J Cell Biol 1995; 129: 895–898. 3. D C, E JA, U H, H K, F N, W LT. The fms-like tyrosine kinase, receptor for vascular endothelial growth factor. Science 1992; 255: 989–991. 4. D DK, M N. Evaluation of antioxidant effectiveness in ischemic reperfusion tissue injury. Methods Enzymol 1994; 233: 601–610. 5. D DK, E RM. Phospholipids and myocardial reperfusion injury. In: Das DK (ed.). Pathophysiology of reperfusion injury. Florida: CRC Press, 1992: 149–179. 6. K M, V EE, A S, B LV, T S, T M, K RY, R RM, C KA, Y KT, A AP. Reactive oxygen intermediates increases vascular endothelial growth factor expression in vitro and in vivo. J Clin Invest 1996; 98: 1667–1175. 7. P’ J, S D, I I, H I, G H, K E. Hypoxia-induced expression of vascular endothelial growth factor by retinal cells is a common factor in neovascularizing ocular diseases. Lab Invest 1995; 72: 638–645. 8. L PI, H KA, K S, S DH. Hypoxia/reoxygenation enhances tube formation of cultured human microvascular endothelial cells: role of reactive oxygen species. In: Maragoudakis ME (ed.). Angiogenesis. New York and London: Plenum Press, 1998: 321–336. 9. L PI, W CR. Reactive oxygen species and angiogenesis. In: Rubanyi GM (ed.). Angiogenesis in
16.
17.
18. 19.
20.
21.
22.
23.
24.
293
Health and Disease: Basic Mechanisms and Clinical Applications. New York: Marcel Dekker, Inc., 2000; 199–214. L SH, W PL, E R, D R, J SW, T PA. Early expression of angiogenesis factors in acute myocardial ischemia and infarction. N Engl J Med 2000; 342: 626–633. R PS, E-H T, S H, Z L, M N. Early effects of hypoxia/reoxygenation on VEGF, Ang-1, Ang-2 and their receptors in the rat myocardium: implications for myocardial angiogenesis. Mol Cell Biochem, In press. G E, S N, G D. NFB induction during in vivo hypoxia in dorsocaudal brainstem of rat: effect of MK 801 and L-NAME. J Appl Physiol 1998; 85: 372–376. L SF, Y X, M AB. Inhibition of NFB activation by pyrrolidine dithiocarbamate prevents in vivo expression of proinflammatory genes. Circulation 1999; 100: 1330–1337. D RL, M MD, G RD, C RG, C KR, R J J. Cardiovascular effects of insulin-like growth factor-1 and growth hormone in chronic left ventricular failure in the rat. Circulation 1996; 93: 2188–2196. D RL, H S, M HR, C R, C KR, R J J. Insulin-like growth factor1 enhances ventricular hypertrophy and function during the onset of experimental cardiac failure. J Clin Invest 1995; 95: 619–627. O S, B V, O S, M L, H G, R HA, R J J. In vivo assessment of left ventricular remodelling after myocardial infarction by digital videocontrast angiography in the rat. Cardiovasc Res 1994; 28: 349–357. O B-H, O S, R HA, R J J. Myocardial hypertrophy in the ischemic zone induced by exercise in rats after coronary reperfusion. Circulation 1993; 87: 598–607. X Z, G M, B S, K T. The capillary of left ventricular tissue of rats subjected to coronary artery occlusion. Cardiovasc Res 1997; 33: 671–676. L J, Q MJ, P M, K , K RA. Adenovirus-mediated gene transfer into infarcted myocardium: feasibility, timimg and location of expression. J Mol Cell Cardiol 1996; 28: 2057– 2060. S JE, Q YZ, G GJ, K RC. The ischemia-selective KATP channel antagonist, 5-hydroxydecanoate, blocks ischemic preconditioning in the rat heart. J Mol Cell Cardiol 1997; 29: 1055–1060. G RJ, S JE, Y Z, K D, W DC, N PJ, G GJ. Antibody to platelet/ endothelial cell adhesion molecule-1 reduces myocardial infarct size in a rat model of ischemia-reperfusion injury. Circulation 1996; 94: 3327–3333. MN PH, D A, W JM. Glycogen depletion contributes to ischemic preconditioning in the rat heart in vivo. Am J Physiol 1996; 271: H2283–H2289. A M, T JR, B M, T J, L A, J A, D M, R L. Pattern of superoxide dismutase enzymatic activity and RNA changes in rat heart ventricles after myocardial infarction. Am J Pathol 1997; 151: 549–555. P MA, P JM, F MC, F PJ,
294
25.
26.
27.
28. 29.
H. Sasaki et al.
S J, K RA, B E. Myocardial infarct size and ventricular function in rats. Circ Res 1979; 44: 503–512. M N, S M, P PD, D DK. An essential role of NFB in tyrosine kinase signaling of p38 MAP kinase regulation of myocardial adaptation to ischemia. FEBS Letts 1998; 429: 365–369. E DT, W M, E RM, R JA, K E, K VE, M N, D DK. Hypoxic preconditioning preserves antioxidant reserve in the working rat heart. Cardiovasc Res 1995; 29: 133– 140. T MA, K DA, R SL, A CA, B CM. Characterization of human proximal tubular cells after hypoxic preconditioning: constitutive and hypoxia-induced expression of heat shock proteins HSP70 (A, B, and C), HSC70, and HSP90. Biochem Mol Med 1997; 60: 49–58. S RM. Tumor hypoxia and gene expression. Acta Oncol 1998; 37: 567–574. P EA, A L, F ED, A LP, S LE.
30.
31.
32. 33.
Vascular endothelial growth factor/vascular permeability factor expression in a mouse model of retinal neivascularization. Proc Natl Acad Sci USA 1995; 92: 905–909. L A, F C. Expression of vascular endothelial growth factor by cultured endothelial cells from brain microvessels. Biochem Biophys Res Commun 1993; 194: 799–803. D EJ, M WJ, F TM, F AS, R AB. Tumor necrosis factor alpha activates human immuno-deficiency virus type 1 through induction of nuclear factor binding to the NFB sites in the long terminal repeat. Proc Natl Acad Sci USA 1989; 86: 5974–5978. M S, V IM. REL/NF kB/IkB story. Adv Cancer Res 1995; 66: 255–292. Y A, Y S, I T, K M, I H. Suppression of retinal neovascularization by the NFB inhibitor pyrrolidine dithiocarbamate in mice. Invest Ophthalmol Vis Sci 1999; 40: 1624–1629.
Hypoxia/Reoxygenation Promotes Myocardial Angiogenesis via an NFB-dependent Mechanism in a Rat Model of Chronic Myocardial Infarction Hiroaki Sasaki1, Partha S. Ray1, Li Zhu2, Hajime Otani2, Takayuki Asahara3 and Nilanjana Maulik1 1
Department of Surgery, University of Connecticut Health Center, Farmington, Connecticut 060301110, USA, 2Department of Thoracic and Cardiovascular Surgery, Kansai Medical Center, Osaka, Japan and 3Tufts University School of Medicine , St. Elizabeth’s Medical Center, Boston, Massachusetts 02135, USA (Received 3 August 2000, accepted in revised form 8 November 2000, published electronically 8 January 2001) H. S, P. S. R, L. Z, H. O, T. A N. M. Hypoxia/Reoxygenation Promotes Myocardial Angiogenesis via an NFB-dependent Mechanism in a Rat Model of Chronic Myocardial Infarction. Journal of Molecular and Cellular Cardiology (2001) 33, 283–294. Therapeutic angiogenesis achieved either through the use of discreet angiogenic proteins or by gene therapy is fast emerging as a highly attractive treatment modality for ischemic heart disease. Herein we examine a novel method of stimulating myocardial angiogenesis by hypoxic preconditioning at both capillary and arteriolar levels, and the potential role of NFB in mediating such a response. We also investigate the functional relevance of such treatment by assessing whether the induced neovascularization can help preserve left ventricular contractile functional reserve in the setting of developing heart failure secondary to myocardial infarction. Male Sprague–Dawley rats were randomly divided into eight groups: normoxia+sham surgery (NS), normoxia+permanent left anterior descending coronary artery (LAD) occlusion (NMI), hypoxic preconditioning+sham surgery (HS), hypoxic preconditioning+permanent LAD occlusion (HMI), PDTC (NFB inhibitor)+hypoxic preconditioning+LAD occlusion (PHMI), PDTC+normoxia+LAD occlusion (PNMI), PDTC+hypoxic preconditioning+sham surgery (PHS) and PDTC+normoxia+sham surgery (PNS). Rats in the preconditioned groups were subjected to systemic hypoxemic hypoxic exposure (10±0.4% O2) for 4 h followed by a 24-h period of normoxic reoxygenation prior to undergoing LAD occlusion. Rats in the normoxia groups were time matched with the preconditioned group and maintained under normoxic conditions for the 28-h period prior to LAD occlusion. The HMI group displayed significant increases in capillary as well as arteriolar density after 2, 4 and 7 days post-operation compared to the NMI. Prior PDTC administration prevented such increases in the PHMI group and effectively abolished the pro-angiogenic effect of hypoxic preconditioning (HP). One week after sham surgery or LAD occlusion, rats underwent a pharmacological stress test with dobutamine in progressively increasing doses which revealed significantly elevated values of dp/dtmax at each dose point in the HMI group compared to the NMI or PHMI groups. Hypoxic preconditioning also decreases endothelial cell injury as determined by the extent of endothelial cell apoptosis using anti-VWF factor labelling and TUNEL assay. The results suggest that HP stimulates myocardial angiogenesis via redox-regulated transcription factor, NFB-dependent pathway to an extent sufficient to exert significant preservation of contractile functional reserve in a rat model of myocardial infarction progressing to heart failure. 2001 Academic Press K W: Angiogenesis; Myocardium; Infarction; Nuclear factor-kappa B; Hypoxia; Reoxygenation.
Please address all correspondence to: Nilanjana Maulik, Ph.D., Molecular Cardiology Laboratory, Cardiovascular Division, Department of Surgery, University of Connecticut School of Medicine, Farmington, CT 06030-1110, USA. Fax: (860) 679-2451. E-mail: [email protected]
0022–2828/01/020283+12 $35.00/0
2001 Academic Press
284
H. Sasaki et al.
Introduction Therapeutic angiogenesis achieved either through the use of discreet angiogenic proteins or by gene therapy is fast emerging as a highly attractive treatment modality for ischemic heart disease. It is well known that coronary collateral vessels help protect or reduce damage to ischemic myocardium after coronary obstruction. Factors which stimulate collateral growth can therefore be expected to exert a protective effect against myocardial infarction. Among the various triggers of angiogenesis, tissue hypoxia has been identified as being a particularly important stimulus for the induction of new vessel growth, particularly at the capillary level.1 Capillary endothelial cells play the most important role in giving rise to several types of functionally and morphologically distinct vessels: under the control of angiogenic stimuli they can re-enter the cell cycle, degrade basement membrane and migrate, forming capillary sprouts which then differentiate to form new vessels.2,3 In the myocardium, hypoxia usually occurs during ischemic episodes. Such ischemic hypoxia is often followed by reperfusion. It is during this subsequent reperfusion period that reoxygenation occurs and reactive oxygen intermediates are formed.4,5 In observations by Kuroki et al., reactive oxygen intermediates were found to increase vascular endothelial growth factor (VEGF) expression in vitro.6 Reoxygenation of human retinal epithelial cells in vitro and ocular reperfusion in vivo increased retinal VEGF mRNA levels.6,7 Administration of antioxidants prior to reoxygenation/reperfusion effectively inhibited such increases in VEGF mRNA. Lelkes et al. have shown that hypoxia/reoxygenation, but not hypoxia alone, cause the formation of reactive oxygen species (ROS).8 In their study, the rate of tubular morphogenesis by human microvascular cells was increased three-fold by hypoxia/reoxygenation. Tube formation in response to hypoxia/reoxygenation as well as under normoxic conditions was inhibited by various ROS antagonists, in a dose-dependent manner, indicating the essential role of ROS in initiating angiogenesis.9 A wide array of pathological conditions are known to generate ROS including inflammation, atherosclerosis, ischemia and reperfusion, etc. The vascular endothelium is thus a target of oxidative stress under a variety of conditions. In fact, Lelkes et al. have demonstrated that administration of ROS inhibitors such as pyrollidine dithiocarbamate (PDTC) inhibit activation of NFB and tube formation in vitro.8,9 Recently, it has been shown that following myocardial infarction various angiogenic growth factors
are upregulated in the human heart as an intrinsic adaptive response.10 Clinical gene trials with VEGF are showing promise as a potential means of augmenting such adaptive angiogenic responses. However, in order to achieve clinical relevance, the enhanced angiogenesis must be of sufficient extent so as to effectively reduce loss of cardiac function after ischemic injury. In this study, we sought to evaluate the functional relevance of stimulated myocardial angiogenesis by utilizing an animal model of chronic myocardial infarction progressing to heart failure and examining left ventricular response during pharmacological stress testing with dobutamine as a measure of cardiac reserve. This study constitutes the first report in vivo which demonstrates: (1) non-lethal moderate hypoxic challenge can decrease endothelial cell injury after myocardial infarction; (2) such preconditioning stimulates myocardial angiogenesis at both capillary and arteriolar levels; (3) the redox-sensitive transcription factor NFB plays an essential role in the initiation of such myocardial angiogenesis; and (4) myocardial angiogenesis stimulated by prior hypoxia/reoxygenation helps to significantly improve postmyocardial infarction (MI) left ventricular contractile functional reserve.
Materials and Methods All animals used in this study received humane care in compliance with the principles of laboratory animal care formulated by the National Society for Medical Research and Guide for the Care and Use of Laboratory Animals prepared by the National Academy of Sciences and published by the National Institutes of Health (publication no. NIH 85-23, revised 1985). The experimental protocol was performed after receiving approval from the Institutional Animal Care Committee.
In vivo hypoxia/reoxygenation Two hundred and eighty-eight rats weighing 275–300 g were divided randomly into eight groups as outlined in Figure 1. Based on previous experiments performed in our laboratory, we have determined that a 4-h duration of systemic hypoxemic hypoxia followed by 24 h of reoxygenation induces a very strong induction of angiogenic factors such as VEGF, Ang-1/Ang-2 and their receptors Flk-1, Flt-1 and Tie-2 determined by Western blot analysis11 (Table 1). We therefore used this duration
Hypoxia/Reoxygenation Induces Myocardial Angiogenesis via NFB
285
Figure 1 Outline of experimental protocol. Rats randomly assigned to various experimental groups (n=12) were subjected to LAD ligation following different exposure protocols. NMI group: Normoxia+LAD occlusion; HMI group: Hypoxia/Reoxygenation+LAD occlusion; PHMI group: PDTC+Hypoxia/Reoxygenation+LAD occlusion; PNMI group: PDTC+Normoxia+LAD occlusion. Rats which were assigned randomly to the next four groups served as the respective sham-operated controls for groups NMI through PNMI, respectively. NS group: Normoxia+Sham surgery; HS group: Hypoxia/Reoxygenation+Sham surgery; PHS group: PDTC+Hypoxia+Sham surgery; PNS group: PDTC+Normoxia+Sham surgery.
Table 1 Relative protein expression profile compared to the normoxic control (% normoxic control) after different duration of hypoxia followed by 24 h of reoxygenation Angiogenic factors and their receptors
1 h H/R (%)
2 h H/R (%)
4 h H/R (%)
VEGF Flk-1 Flt-1 Ang-1 Ang-2
50.8 41.5 325.0 23.8 31.3
51.9 22.0 201.0 30.0 14.8
Tie-1
533.0
586.7
Tie-2
93.9
120.5
50.0 12.8 149.9 50.0 Same as control Same as control 53.4
of hypoxic exposure to effect hypoxic preconditioning in all our preconditioned animal groups. Rats in the hypoxic preconditioned group (HMI) were subjected to a 4-h hypoxic challenge,
using a gas mixture (10% O2/90% N2),12 in an anesthesia chamber. An in-flow aperture was used for the gas mixture and an out-flow aperture connected to a gas absorption cannister (A.M. Bickford Inc., Wales Center, NY, USA). A KE-25 galvanic cell O2 sensor (Kent Scientific, Litchfield, CT, USA) was used to monitor continuously the %O2 concentration, the signal being amplified through a TRN-005 Amplifier (Kent Scientific). The O2 concentration was gradually lowered from an ambient %O2 of 20.9% down to 10% over the course of 30 min. Once stabilization was obtained, the inflow rate was adjusted so as to maintain the %O2 at 10±0.4% for 4 h. These rats were then maintained in a normoxic environment for 24 h until preoperative preparation. Rats in the PDTC+hypoxic preconditioned group (PHMI) were treated similarly except for PDTC administration (150 mg/kg b.w. i.p. injection)12 1 h prior to the hypoxic challenge. Rats in the PDTC control group
286
H. Sasaki et al.
(PNMI) were administered PDTC and were thereafter maintained in a normoxic environment in a time-matched manner (29 h) until preoperative preparation. Rats in the control group (NMI) were similarly maintained in a normoxic environment in a time-matched manner (29 h) until preoperative preparation. Six animals were excluded from each permanent occlusion group either due to post-op mortality or morbidity within 24 h or based on infarct size being less than 50%. Thus sample size for permanent occlusion groups was n=6. Although the hypoxic-preconditioned permanent occlusion group did display improved survival postop, the small sample size (despite the large number of total animals used) did not permit analysis for assessing any interventional benefit in reducing morbidity or mortality outcomes. No animals were excluded from sham-operated groups. Thus sample size for sham-operated groups was n=12 for function comparisons but n=6 for morphometric analysis chosen randomly.
Survival surgery and LAD occlusion After the 24-h reoxygenation period, time-matched normoxic period rats were anesthetized with Ketamine HCl (100 mg/kg i.p.) and Xylazine (10 mg/ kg i.p.).13–16 Cefazolin (25 mg/kg i.p.) was administered by i.v. injection as preoperative antibiotic cover. The animals were placed in the supine position and body temperature was maintained at 37°C by means of a water-circulating thermalheating pad. A midline neck incision was made and the trachea was identified. Tracheostomy was performed by making a vertical incision approximately 2 mm long in the trachea. A polyethylene tube (2 mm o.d.)17 was introduced through the opening and positive pressure ventilation with room air was commenced (stroke volume 12 ml/ kg,17–19 tidal volume 2.5 ml20 and respiratory rate 70/min21) using a small animal respirator (model 683, Harvard Apparatus, Holliston, MA, USA). The tube was secured in place by means of stay sutures. An intercostal thoracotomy in the left fourth intercostal space was performed.13,14,16,22 The heart was exposed and the pericardium incised. The left coronary artery was identified between the left atrial appendage and the pulmonary artery outflow tract. A 6-0 suture was passed beneath the vessel 1–2 mm from its origin. Care was taken not to enter the ventricular cavity. In the sham-operated group (NS), the suture was left in place up until chest closure but was not ligated. In the permanent
coronary-occlusion groups (NMI, HMI) electrophysiological changes. After completion of all protocols, the chest wall was closed in three layers using 4-0 silk sutures. A catheter fashioned from PE-190 tubing attached to a 12-ml syringe with stopcock was used as a chest tube to evacuate air from the pleural space as the last stitch was tightened. Post-operative pain relief was achieved by the administration of Buprenorphine (0.05–2.5 mg/kg s.c.).
Non-survival surgery and baseline cardiac function One week after the operative intervention, the rats were anesthetized and ventilated as described above. A small incision was made to the right of the midline in the neck. The right internal jugular vein was identified and a PE 50 catheter was introduced into the vein. The proximal end of the catheter was connected to a low-pressure transducer. The inserted catheter tip was then advanced until it reached the right atrium and the central venous pressure (CVP) signal was obtained. The same procedure was used to place a catheter in the right carotid artery where the catheter was used to monitor aortic blood pressure. The inserted tip of this catheter was advanced down until it reached the left ventricular lumen and the left ventricular pressure (LVP) signal was obtained. All pressure signals were monitored, analysed and recorded in real time using the Digimed data acquisition and analysis system (Micromed, Louisville, KY, USA). Heart rate (HR), developed pressure, and dp/dtmax were all calculated from the continuously obtained LVP signal.
Cardiac stress testing with dobutamine After baseline left ventricular functional parameters were recorded, rats were subjected to pharmacological cardiac stress testing with dobutamine infusion in progressively incremental doses to reveal the extent of left ventricular contractile functional reserve. Dobutamine was infused i.v. through the venous catheter by means of a microinjection pump (Harvard Apparatus) to achieve doses of 1, 2, 3 and 5 g/kg/min. Based on previous studies in our laboratory, we have found that administration up to a dose of 5 g/kg/min is necessary to observe the plateau of the dose-response curve. Administration was maintained at each dose for a duration of 2 min. dp/dtmax and other left ventricular functional parameters were recorded for each dose 90 s after the start of each dose.
Hypoxia/Reoxygenation Induces Myocardial Angiogenesis via NFB
Tissue retrieval and processing After measurement of all functional parameters, the animals were heparinized systemically (Heparin Sodium, 500 IU/kg b.w., i.v. injection, Elkins-Sinn Inc., Cherry Hill, NJ, USA) to prevent intravascular coagulation. The animals were killed with 1 ml of 30 m KCl injected i.v. Hearts were excised, the ventricular lumen was flushed clean with Ringer’s lactate solution and the ventricles were either frozen in liquid nitrogen for later use in biochemical studies (Gel Shift assay for NFB) or were sectioned approximately 5–6 mm from the apex into approximately 2-mm thick transverse sections. The sections were then fixed in 10% buffered formalin and embedded in paraffin using standard procedures.
Measurement of capillary and arteriolar density The standard deparaffinization protocol was used. Endothelial cells were labeled using mouse monoclonal anti-CD31/PECAM-1 (1:100, Pharmingen) followed by a biotinylated horse anti-mouse secondary antibody (1:200 dilution). The reaction product (brown) was visualized with DAB substrate using the Vector ABC Vectastain Elite Kit (Vectorlabs, Burminghame, CA, USA) and was counterstained with methyl green (Vectorlabs). On separate slides, vascular smooth muscle cells were labeled using mouse monoclonal anti-smooth muscle actin (1:50, Biogenex, San Ramon, CA, USA) followed by a biotinylated rat-adsorbed horse anti-mouse secondary antibody (1:200 dilution). The reaction product (violet) was visualized with VIP substrate using the Vector ABC Vectastain Elite Kit (Vectorlabs). All sections were dehydrated in two changes of 100% ethanol, cleared in Toluene and permanently mounted using Polymount (Poly Scientific R&D, Bayshore, NY, USA). Capillaries were identified and recognized by their size, a thin layer of endothelial cells and a lack of smooth muscle cells. Similarly, arterioles were identified and recognized by their size and a thin layer of smooth muscle cells. At a total magnification of 400×, eight non-overlapping random fields were selected from endocardial regions of non-infarct/borderline zones of the left ventricle of two sections from each heart (16 fields per region per heart, 64 fields per region per group). Images were captured and stored in digital .tif file format for later image analysis. Counts of capillary density and arteriolar density/ mm2 were obtained after superimposing a calibrated morphometric grid on each digital image using
287
Adobe Photoshop Software. Counting was performed in a blind manner by two separate investigators. Counts of particular type (e.g. capillary) in the same region (e.g. epicardial) from hearts obtained from animals in the same experimental group (e.g. hypoxic preconditioned+LAD occlusion) were averaged to yield the numerical value of the particular density measurement for the group. Counts of capillary density and arteriolar density/ mm2 were obtained.
Measurement of percent circumferential left ventricular infarct size After following the standard deparaffinization protocol described above, the extent of left ventricular myocardial infarct was delineated histochemically by utilizing Mallory’s Trichrome staining technique for collagen present in the infarct scar.23 The hearts which underwent 50% infarction by LAD occlusion were considered for the experimental evaluations. Electrophoretic mobility assay (EMSA) Nuclear proteins were isolated from the heart to estimate NFB translocation.24 In short, 150 mg of left ventricle from heart tissue was homogenized in ice-cold Tris-buffered saline (TBS) and centrifuged at 3000×g for 5 min at 4°C. The pellet was resuspended by gentle pipetting in 1.5 ml of ice-cold hypotonic buffer containing 10 m HEPES, pH 7.9, 10 m KCL, 0.1 m EDTA, 0.1 m EGTA, 1 m DTT, 0.5 m PMSF, and 1 each of aprotonin, pepstatin and leupeptin. The solution was allowed to swell on ice for 15 min after addition of 100 l of 10% Nonidet P-40, the tube was vortexed vigorously for 45 s. This homogenate was centrifuged for 30 s at 4°C in a microcentrifuge tube. The supernatant contained the cytoplasmic protein. The nuclear pellet was resuspended in a solution containing 20 m HEPES, pH 7.9, 0.4 NaCl, 1 m EDTA, 1 m EGTA, 1 m DTT, 1 m PMSF and 1 each of aprotonin, pepstatin and leupeptin. The tubes were shaken vigorously at 4°C for 30 min on a shaking platform. The nuclear extracts were stored at −70°C. Protein concentration was estimated by using Pierce protein assay kit (Pierce Chemical Company, Rockford, IL, USA). NFB oligonucleotide (AGTTGAGG-GGACTTTCCCAGG) [2.5 l (20 ng/l)] was labeled using T4 polynucleotide kinase as previously described. The binding reaction mixture contained in a total volume of 20.2 l, 0.2 l DTT (0.2 ), 1 l BSA (20 mg/ ml), 4 l PdI-dC (0.5 g/l), 2 l Buffer D+ 4 l,
288
H. Sasaki et al.
Buffer F, 2 l 32P-oligo (0.5 ng/l) and 7 l extract containing 10 g protein. Composition of Buffer D+ was 20 m HEPES, pH 7.9, 20% glycerol, 100 m KCl, 0.5 m EDTA, 0.25% NP 40 while Buffer F contained 20% Ficoll 400, 100 m HEPES, pH 7.9, and 300 m KCl. Incubation was carried out for 20 min at room temperature. Ten microlitres of the solution was loaded onto a 4% acrylamide gel and separated at 80 V until the dye hit the bottom. After electrophoresis, gels were dried up and exposed to Kodak X-ray film at −70°C. The purity of the nuclear extracts was examined using lactate dehydrogenase (LDH) as a cytosolic marker, and it was found that 1.5% of total myocardial tissue LDH was present in the nuclear fraction. The super shift assays and the competition assays were performed to measure that the signal was specific for NFB. For the supershift assays, anti-p65 or anti-p50 antibodies were added, separately or together, to the reaction mixtures immediately after addition of radiolabeled probe. For the competition assays, increasing molar excess of unlabelled NFB oligonucleotide was added into separate reaction mixtures.
Evaluation of endothelial cell apoptosis The formaldehyde-fixed left ventricle was embedded in paraffin, cut into transverse sections (4-m thick) and deparaffinized with a graded series of xylene and ethanol solutions. The sections are treated with 40 g proteinase K for 15 min at 37°C and then incubated in absolute methanol containing 0.3% hydrogen peroxide for 30 min. Non-specific binding is blocked with normal goat serum. The sections are assayed for in situ terminal transferase labeling (TUNEL) using Apop Tag Plus (Oncor Inc., Gaithersburg, MD, USA). Negative control slides are processed with the TdT enzyme excluded. The occurrence of apoptosis in endothelial cells are demonstrated in the sections which are first stained with TUNEL (FITC staining). The sections were then incubated with rabbit polyclonal anti-von Willebrand factor (Dako) as a primary antibody followed by incubation with tetrarhodamine isothiocyanateconjugated goat anti-rabbit IgG as a secondary antibody and viewed with double immunofluorescence confocal laser microscopy. For the quantitative purpose, the number of TUNEL-positive endothelial cells was counted on 60 high power fields (HPF, magnification ×600) from the endocardium through the epicardium of the left ventricular free wall in three sections from each heart. Representative confocal images shows von Wil-
lebrand factor-positive endothelial cells (strong red staining in their cytosol) which are negative for TUNEL staining (absence of green staining in the nucleus) as well as those positive for TUNEL staining (magnification ×1200).
Statistical analysis Results are expressed as mean±standard error of the mean (...). Differences between groups were tested for statistical significance by one-way analysis of variance (ANOVA) and Student’s t-test, with the Bonferroni adjustment for multiple comparisions. The dose–response curves to dobutamine of the various groups were tested for significant differences by repeated measure ANOVA. Differences were considered significant at P<0.05.
Results Effect of hypoxia/reoxygenation (H/R), H/R/PDTC and PDTC only on microvascular density Quantitative measurement of capillary density At a total magnification of 400×, eight non-overlapping random fields each were selected from the endocardial regions of non-infarcted zones of the left
Figure 2 Left ventricular endocardial capillary density. Tissue sections were processed for CD31 staining and eight non-overlapping random fields were selected from endocardial regions of non-infarcted of the left ventricle of two sections from each heart (16 fields per region per heart, 64 fields per region per group, magnifications of 400× were used, n=6). Images were captured and stored in digital .tif file format for image analysis. Where ∗ P<0.001 compared to sham control, † P<0.001 compared to HMI. Ε Sham, ∆ NMI, HMI, ; PHMI, Φ PNMI.
Hypoxia/Reoxygenation Induces Myocardial Angiogenesis via NFB
289
numerical value of the particular density measurement. No statistically-significant differences in arteriolar density between groups were demonstrable in hearts evaluated 2 days or 4 days post-op (except for the HMI and PHMI groups). The arteriolar density was, however, significantly elevated in the hypoxic preconditioned group (HMI) after 1 week post-op (2.23±0.12 v 1.53±0.02 in NMI and 1.51±0.08 counts/mm2 in PNMI) (Fig. 3). No such increase in arteriolar density was observed in the group which had been administered PDTC prior to hypoxic preconditioning (1.6±0.02 counts/mm2, PHMI group). Figure 3 Left ventricular endocardial arteriolar density. Left ventricular tissue sections were labelled using monoclonal anti-smooth muscle actin and eight non-overlapping random fields were selected from endocardial region of the left ventricle. Where ∗ P<0.001 compared to sham control, † P<0.001 compared to HMI. Ε Sham, ∆ NMI, HMI, ; PHMI, Φ PNMI.
ventricles. Quantitatively the HMI group displayed a statistically significant increase in capillary density as early as 2 days post-operation when compared to sham-operated groups and/or the control group (1862±67 counts/mm2 v 1642±18 counts/mm2) (Fig. 2). The capillary density continued to increase in the HMI group and was maintained at a higher level 4 days and 7 days post-operation (2035±53 and 1963±45 counts/mm2). Capillary density was maintained close to baseline levels in the NMI group after 2 days post-operation (1591±38 counts/ mm2), but thereafter displayed a significant decrease after 4 days and 7 days post-operation (1217±47 and 1313±55 counts/mm2, respectively) of hypoxic preconditioning. The incremental effect on capillary density was not displayed in the group which had received PDTC treatment prior to hypoxic preconditioning (PHMI). This was evident from the significant decrease in capillary density observed in this group after 2 days and 4 days post-op (1122±60 and 1430±70 counts/mm2 respectively), returning to a baseline-comparable value only after 7 days post-op (1663±78 counts/ mm2) (Fig. 2).
Quantitative measurement of arteriolar density At a total magnification of 200×, five non-overlapping random fields were selected from the endocardial regions of non-infarcted zones of the left ventricles. Arteriolar counts (purple color) for each field in the same region was averaged to yield the
Effect of H/R, H/R/PDTC and PDTC only on cardiac function: dose–response study of pharmacological stress by dobutamine infusion on left ventricular pressure (LVP), maximal dP/dt, left ventricular end diastolic pressure (EDP) and heart rate (HR) Indices of baseline cardiac contractile function (as assessed after 1 week post-operation, or after a 1week time-matched duration in case of control rats) were significantly different in all operation groups (NMI, HMI, PHMI and PMI) irrespective of preoperative treatment protocol, when, compared to control rats which had not been subjected to any survival surgery procedure. The LVSP was significantly depressed and the LVEDP significantly increased compared to normal baseline values in the unoperated rat (Table 2). Baseline values of HR, EDP, LVP and dp/dtmax in the sham-operated groups (NS, HS, PHS, PNS), however, did not differ significantly from those of the control rats which did not undergo survival surgery. Sham functional data were not shown because there were no statistically significant differences between them and their inclusion would distract from the data. There were also no demonstrable statistically significant differences between functional data of any of the sham groups and the control baseline group. Therefore, functional data only from the control baseline group has been included in the figure and table as being representative of all the sham groups. Differences between sham groups were only demonstrable in capillary density (CD): CD was significantly higher in the hypoxic preconditioned sham group after 7 days (1814±56 counts/mm2) compared to the control baseline group after 7 days (1642±18 counts/mm2). Again, there were no statistically significant differences between CD measurements in any of the other sham groups with the control baseline group. In the figure on
290
H. Sasaki et al.
Table 2 Measurement of hemodynamic parameters under dobutamine stress test Dobutamine (g/min/kg)
0
Heart rate (bpm) Sham 235.1±11.0 NMI 197.7±16.0 HMI 207.6±8.4 PHMI 213.1±11.2 PMI 178.4±11.9 Left ventricular systolic pressure (mmHg) Sham 105.3±11.5 NMI 96.9±4.9 HMI 85.9±7.3 PHMI 92.9±5.2 PMI 92.5±2.6 Left ventricular end-diastolic pressure (mmHg) Sham 5.93±0.66 NMI 14.93±1.54∗ HMI 13.15±2.38∗ PHMI 14.23±1.42∗ PMI 15.92±1.56∗
1
2
3
5
229.5±9.7 227.9±12.1 223.3±9.2 224.0±13.5 202.3±5.3
253.5±14.9 238.5±13.4 237.4±10.9 231.5±12.0 213.2±6.0
263.1±13.7 251.3±15.0 254.2±13.0 237.2±13.5 215.7±10.1
242.2±9.9 260.5±14.1 269.4±14.6 243.9±11.0 231.7±11.8
119.7±4.8 111.3±6.3 100.5±8.9∗ 100.8±6.7∗ 105.1±4.3
146.2±5.6 115.8±9.4∗ 107.8±8.6∗ 109.1±7.3∗ 110.0±7.0∗
158.0±6.2 124.1±10.2∗ 111.1±7.6∗ 114.0±7.3∗ 117.5±5.2∗
153.1±3.4 130.7±6.1∗ 114.2±7.3∗ 117.4±5.0∗ 114.7±6.8∗
6.28±0.68 17.68±2.04∗ 16.75±2.14∗ 12.92±1.34∗ 18.17±2.56∗
7.28±0.41 18.07±2.61∗ 18.08±1.99∗ 14.33±1.23∗ 19.10±3.22∗
8.18±0.66 23.3±1.96∗ 20.88±1.73∗ 18.73±1.38∗ 22.97±2.27∗
9.05±0.89 24.87±2.55∗ 22.58±2.22∗ 21.48±1.85∗ 26.55±2.67∗
∗ P>0.05 compared with sham.
capillary density, only permanently occluded groups are compared.
Response to dobutamine Pharmacological cardiac stress testing with dobutamine infusion in incremental doses revealed differences in the extent of cardiac contractile reserve between groups. This was evident from differences in the extent of change of dp/dtmax values displayed by the various groups during the course of such stress testing (Fig. 4). The hypoxic preconditioned group displayed significantly elevated contractile reserve at each dose point of evaluation compared to all other operation groups, but the extent of such reserve was lower than that displayed by sham-operated groups or by control rats (all of which exhibited similar and statistically indifferent values of dp/dtmax throughout the course of stress testing). This enhanced preservation of contractile reserve in the hypoxic preconditioned group was apparent at the 1 g/kg/min dose (6026±444 mmHg/s in HMI group v 5075±355 mmHg/s in NMI group, 4403±384 mmHg/s in PHMI group and 5031±151 mmHg/s in PMI) and this trend persisted at all subsequent doses up to 5 g (7630±729 mmHg/s in HMI v 6084±531 in NMI, 5292±665 in PHMI and 6111±537 in PMI). This capability of hypoxic preconditioning to better preserve contractile functional reserve in the infarcted heart was, however, not evident in the
Figure 4 Dose–response curves of dp/dtmax during dobutamine stress test. Changes in dp/dtmax in mm of mercury/s (mmHg/s) from baseline with increasing dobutamine boluses (0–5 g/kg/min) in 7 days MI hearts. The dose–response curve for dobutamine for the various doses were tested for significant differences by repeated measure ANOVA. Differences were considered significant at P<0.05. Where ∗ P<0.01 compared to Sham, † P<0.01 compared to NMI, PHMI, PNMI.
group which had received PDTC treatment prior to the hypoxic preconditioning (PHMI). The performance of this latter group on such cardiac stress testing resembled that of the NMI group, as also the group which had received only PDTC and no hypoxic preconditioning (PMI).
Hypoxia/Reoxygenation Induces Myocardial Angiogenesis via NFB
Figure 5 Effect of hypoxia/reoxygenation and PDTC on NFB/DNA binding activity in rat myocardium subjected to whole-body hypoxic preconditioning for various periods of hypoxia. Rats were subjected to 1 h (H1), 2 h (H2) and 4 h (4 h) of whole-body hypoxia followed by 2 h of reoxygenation as described in the Materials and Methods. Gel-shift analysis revealed that baseline control (BL) hearts have minimal NFB activity whereas hypoxically adapted hearts demonstrated significantly increased DNA binding activity of NFB (A) in H1 and H2 heart samples whereas H4 group of hearts showed normal baseline DNA binding activity like BL. PDTC prior to hypoxic treatment (B) inhibited NFB binding activity completely as shown in 1 h and 2 h hypoxic group pretreated with PDTC PH1 and PH2.
291
Figure 6 Estimation of apoptotic cells. Apoptotic cells expressed as a percentage of total number of endothelial apoptotic cells in the various groups. Results are expressed as mean±... for n=6 per group. Each assay was determined in triplicates. ∗ P<0.05, compared to the baseline control group, † P<0.05 compared to NMI group. Where the extent of apoptosis was shown after 2 days, 4 days and 1 week of post-op. BL Ε, NMI ∆, HMI , PHMI ;, PNMI Φ.
cells when compared to the HMI group after 2, 4 days and also after 1 week. PNMI treatment groups also have similar results as shown in Fig. 6.
Effect of H/R, H/R/PDTC and PDTC only on DNA binding activity of NFB
Discussion Gel-shift analysis revealed the time course of hypoxia-induced NFB/DNA binding activity in rat myocardium. NFB/DNA binding activity was detectable at low levels in control hearts, increased significantly after 1 h of hypoxic exposure followed by 2 h of reoxygenation to 30% which was further increased (40%) after 2 h of whole body hypoxia and 2 h of reoxygenation (Fig. 5). PDTC completely suppressed the hypoxia/reoxygenation induced NFB/DNA binding activity completely when administered i.p. 1 h before hypoxia.
Evaluation of endothelial cell injury by examining the extent of endothelial cell apoptosis under the exposure of H/R, H/R/PDTC and PDTC only We were unable to detect any apoptotic endothelial cells in the control baseline samples (Fig. 6). However, significant numbers of apoptotic cells were visible in the NMI groups after 2 and 4 days after post-op compared to the HMI groups. After 1 week the number of apoptotic cells went down in the NMI group but remained significantly higher when compared to the HMI group. PHMI group of rat hearts revealed a significant number of apoptotic
The present study reports for the first time that systemic hypoxemic hypoxic exposure induces myocardial angiogenesis in a rat model of myocardial infarction as evidenced by increased capillary/arteriolar density. Myocardial adaptation to intermittent hypoxia appears to be a highly promising approach to reduce cellular injury due to ischemia and reperfusion. It has been shown that hypoxic adaptation could prevent myocardial stress, ischemic damage, and cardiac arrhythmias in both animals and humans.25 Such adaptation also restricts the impairment of cardiac electric stability and contractility in acute myocardial infarction.26 Hypoxia has the potential to promote the induction of angiogenesis by altering angiogenic regulatory proteins, metabolic enzymes and transcription factors.27–29 Tissue hypoxia, once achieved, exerts such a proangiogenic action through various angiogenic factors, the most notable being vascular endothelial growth factor, which has been chiefly associated with initiating the process of angiogenesis through the recruitment and proliferation of endothelial cells. However, according to Lelkes et al.8 brief exposure to hypoxia followed by reoxygenation significantly accelerates the formation of capillary
292
H. Sasaki et al.
formation by HMVEC cultured on Matrigel and hypoxia/reoxygenation mediated tube formation does not involve upregulation of PKC or the release of angiogenic growth factors, but rather it depends on the generation of reactive oxygen species (ROS) and the subsequent activation of redox regulated nuclear transcription factor, NFB. Evidence indicates that oxidative stress/free radicals lead to the activation of NFB, which in turn induces the expression of genes.30 NFB is the first eukaryotic transcription factor shown to be influenced directly by oxidative stress.31 Recently it was demonstrated in mouse model that NFB activation is obligatory for retinal angiogenesis to occur and it was also documented that the administration of pyrrolidine dithiocarbamate (PDTC) suppressed such retinal neovascularization.32 The present report demonstrates that hypoxic preconditioning triggers myocardial angiogenesis via an NFB-mediated pathway in a rat model of myocardial infarction. Significant increases in capillary as well as arteriolar density were observed in the 2, 4 and 7 days HMI myocardial sections when compared to the NMI, PHMI and PMI samples. Another interesting finding was to verify the hypothesis that non-lethal moderate hypoxic challenge can decrease endothelial cell injury after myocardial infarction as was determined by the relative percentage of endothelial apoptotic cells under various conditions as shown in Figure 6. Interestingly, we found that hypoxic challenge followed by reoxygenation can reduce apoptosis significantly and decrease endothelial cell injury in vivo. Supplementary to the above findings, pharmacological cardiac stress testing with dobutamine proved successful in effectively revealing the ability of hypoxia/reoxygenation pretreatment to exert a pronounced and marked effect in maintaining LV contractile reserve. Such means of measuring LV contractile reserve is especially crucial to detect effective reserve cardiac function in animal models where heart failure is the final pathophysiological manifestation, which in this case was still in the process of developing secondary to chronic MI. Differences in LV contractile function between operated groups were not apparent from baseline functional criteria, as might be expected in the early post-MI state when intrinsic compensatory mechanism are activated on account of the necessity of the animal to maintain a minimal level of circulatory function necessary to stay alive, independent and regardless of preoperative treatments. Differences, however, may become apparent at a later time period when heart failure is no longer sufficiently compensated. Interestingly, although the extent of myocardial
infarction (50%) was statistically insignificant between the various groups compared, the HMI group still demonstrated a significantly enhanced level of contractile function. This would seem to suggest that in the HMI group, the major contributor to such preservation of contractile function was not a reduction in the extent of myocardial necrotic cell death or infarction, but perhaps an enhanced contractile ability of the non-infarcted myocardium in the HMI group. The capillary density and arteriolar density data suggest that the enhanced angiogenesis displayed in the non-infarcted zone of the HMI group has some role to play in effecting the enhanced contractile function. Furthermore, all potentially beneficial and cardioprotective effects of hypoxia/reoxygenation were effectively reversed by prior treatment with PDTC indicating the essential and critical role of NFB in mediating hypoxia/ reoxygenation induced myocardial angiogenesis. It seems NFB—iNOS—NO pathway may be also involved in the fascinating process of angiogenesis. Taken together, it appears that myocardial angiogenesis is precisely controlled by a redox switch and that oxygen-free radicals appear to function as a signaling molecule in this process. In our rat model significant NFB binding activity was observed during hypoxia which was inhibited by PDTC (Fig. 5). In the PHMI group of rats significant inhibition of capillary density was observed when compared with the HMI group of rats (Fig. 2) after 1 week of coronary artery occlusion. This is the first in vivo evidence indicating the involvement of an inducible transcription factor NFB in myocardial angiogenesis. Moreover, the effect on contractile reserve observed in the HMI group at 1 week post-op may be attributed at least partially to the concurrent increase in capillary density. While the increase is clearly suggestive of new capillary growth, the maintenance of capillary density near normal 2 and 4 days post-op in the HMI group is perhaps best explained as enhanced endothelial cell survival early after LAD occlusion. This also explains why there was an acute and significant drop in capillary density values in the other operated groups (NMI, PHMI). The marginal yet significant increase in arteriolar density in the HMI group compared to NMI group of rats indicated that the process of neovascularization is not restricted to the capillary level but extends to the arteriolar level as well. PDTC clearly inhibited the hypoxic preconditioning mediated increased in arteriolar density. It is logical to propose that reoxygenation preceded by hypoxia induces a cascade of events involving VEGF and its receptors expression as well
Hypoxia/Reoxygenation Induces Myocardial Angiogenesis via NFB
as Ang-Tie system protein expression11 might also be one of the possible mechanisms besides NFB– iNOS–NO pathway by which hypoxia reoxygenation triggers myocardial angiogenesis in this model. It is very difficult at this point to speculate on the exact pathway by which hypoxic preconditioning stimulates myocardial angiogenesis. However, inhibition of angiogenesis by the administration of PDTC was successful in demonstrating the essential and important role of NFB in myocardial angiogenesis. This document clearly emphasizes the role of NFB activation in mediating angiogenesis in response to hypoxia in a rat model of chronic myocardial infarction.
Acknowledgement This study was supported by National Institutes of Health Grants HL 56803, HL 22559, HL 33889, HL 34360.
10.
11.
12.
13.
14.
15.
References 1. S D, I A, S D, K E. Vascular endothelial growth factor induced by hypoxia-initiated angiogenesis. Nature 1992; 359: 843–845. 2. M T, A K. Endothelial receptor tyrosine kinases involved in angiogenesis. J Cell Biol 1995; 129: 895–898. 3. D C, E JA, U H, H K, F N, W LT. The fms-like tyrosine kinase, receptor for vascular endothelial growth factor. Science 1992; 255: 989–991. 4. D DK, M N. Evaluation of antioxidant effectiveness in ischemic reperfusion tissue injury. Methods Enzymol 1994; 233: 601–610. 5. D DK, E RM. Phospholipids and myocardial reperfusion injury. In: Das DK (ed.). Pathophysiology of reperfusion injury. Florida: CRC Press, 1992: 149–179. 6. K M, V EE, A S, B LV, T S, T M, K RY, R RM, C KA, Y KT, A AP. Reactive oxygen intermediates increases vascular endothelial growth factor expression in vitro and in vivo. J Clin Invest 1996; 98: 1667–1175. 7. P’ J, S D, I I, H I, G H, K E. Hypoxia-induced expression of vascular endothelial growth factor by retinal cells is a common factor in neovascularizing ocular diseases. Lab Invest 1995; 72: 638–645. 8. L PI, H KA, K S, S DH. Hypoxia/reoxygenation enhances tube formation of cultured human microvascular endothelial cells: role of reactive oxygen species. In: Maragoudakis ME (ed.). Angiogenesis. New York and London: Plenum Press, 1998: 321–336. 9. L PI, W CR. Reactive oxygen species and angiogenesis. In: Rubanyi GM (ed.). Angiogenesis in
16.
17.
18. 19.
20.
21.
22.
23.
24.
293
Health and Disease: Basic Mechanisms and Clinical Applications. New York: Marcel Dekker, Inc., 2000; 199–214. L SH, W PL, E R, D R, J SW, T PA. Early expression of angiogenesis factors in acute myocardial ischemia and infarction. N Engl J Med 2000; 342: 626–633. R PS, E-H T, S H, Z L, M N. Early effects of hypoxia/reoxygenation on VEGF, Ang-1, Ang-2 and their receptors in the rat myocardium: implications for myocardial angiogenesis. Mol Cell Biochem, In press. G E, S N, G D. NFB induction during in vivo hypoxia in dorsocaudal brainstem of rat: effect of MK 801 and L-NAME. J Appl Physiol 1998; 85: 372–376. L SF, Y X, M AB. Inhibition of NFB activation by pyrrolidine dithiocarbamate prevents in vivo expression of proinflammatory genes. Circulation 1999; 100: 1330–1337. D RL, M MD, G RD, C RG, C KR, R J J. Cardiovascular effects of insulin-like growth factor-1 and growth hormone in chronic left ventricular failure in the rat. Circulation 1996; 93: 2188–2196. D RL, H S, M HR, C R, C KR, R J J. Insulin-like growth factor1 enhances ventricular hypertrophy and function during the onset of experimental cardiac failure. J Clin Invest 1995; 95: 619–627. O S, B V, O S, M L, H G, R HA, R J J. In vivo assessment of left ventricular remodelling after myocardial infarction by digital videocontrast angiography in the rat. Cardiovasc Res 1994; 28: 349–357. O B-H, O S, R HA, R J J. Myocardial hypertrophy in the ischemic zone induced by exercise in rats after coronary reperfusion. Circulation 1993; 87: 598–607. X Z, G M, B S, K T. The capillary of left ventricular tissue of rats subjected to coronary artery occlusion. Cardiovasc Res 1997; 33: 671–676. L J, Q MJ, P M, K , K RA. Adenovirus-mediated gene transfer into infarcted myocardium: feasibility, timimg and location of expression. J Mol Cell Cardiol 1996; 28: 2057– 2060. S JE, Q YZ, G GJ, K RC. The ischemia-selective KATP channel antagonist, 5-hydroxydecanoate, blocks ischemic preconditioning in the rat heart. J Mol Cell Cardiol 1997; 29: 1055–1060. G RJ, S JE, Y Z, K D, W DC, N PJ, G GJ. Antibody to platelet/ endothelial cell adhesion molecule-1 reduces myocardial infarct size in a rat model of ischemia-reperfusion injury. Circulation 1996; 94: 3327–3333. MN PH, D A, W JM. Glycogen depletion contributes to ischemic preconditioning in the rat heart in vivo. Am J Physiol 1996; 271: H2283–H2289. A M, T JR, B M, T J, L A, J A, D M, R L. Pattern of superoxide dismutase enzymatic activity and RNA changes in rat heart ventricles after myocardial infarction. Am J Pathol 1997; 151: 549–555. P MA, P JM, F MC, F PJ,
294
25.
26.
27.
28. 29.
H. Sasaki et al.
S J, K RA, B E. Myocardial infarct size and ventricular function in rats. Circ Res 1979; 44: 503–512. M N, S M, P PD, D DK. An essential role of NFB in tyrosine kinase signaling of p38 MAP kinase regulation of myocardial adaptation to ischemia. FEBS Letts 1998; 429: 365–369. E DT, W M, E RM, R JA, K E, K VE, M N, D DK. Hypoxic preconditioning preserves antioxidant reserve in the working rat heart. Cardiovasc Res 1995; 29: 133– 140. T MA, K DA, R SL, A CA, B CM. Characterization of human proximal tubular cells after hypoxic preconditioning: constitutive and hypoxia-induced expression of heat shock proteins HSP70 (A, B, and C), HSC70, and HSP90. Biochem Mol Med 1997; 60: 49–58. S RM. Tumor hypoxia and gene expression. Acta Oncol 1998; 37: 567–574. P EA, A L, F ED, A LP, S LE.
30.
31.
32. 33.
Vascular endothelial growth factor/vascular permeability factor expression in a mouse model of retinal neivascularization. Proc Natl Acad Sci USA 1995; 92: 905–909. L A, F C. Expression of vascular endothelial growth factor by cultured endothelial cells from brain microvessels. Biochem Biophys Res Commun 1993; 194: 799–803. D EJ, M WJ, F TM, F AS, R AB. Tumor necrosis factor alpha activates human immuno-deficiency virus type 1 through induction of nuclear factor binding to the NFB sites in the long terminal repeat. Proc Natl Acad Sci USA 1989; 86: 5974–5978. M S, V IM. REL/NF kB/IkB story. Adv Cancer Res 1995; 66: 255–292. Y A, Y S, I T, K M, I H. Suppression of retinal neovascularization by the NFB inhibitor pyrrolidine dithiocarbamate in mice. Invest Ophthalmol Vis Sci 1999; 40: 1624–1629.