Coronary collateral development relies on the coordinated secretion of growth factors. However, alone they are insufficient for permanent collateral growth. We utilized proteomics to identify other important proteins in the extracellular environment that could facilitate collateralization. Chronically instrumented dogs developed coronary collaterals by the repetitive occlusion method. Subendocardial (0.19 ± 0.04, 0.27 ± 0.06, 0.48 ± 0.10, and 0.81 ± 0.11 ml·min−1·g−1 on days 1, 7, 14, and 21, respectively) and subepicardial (0.14 ± 0.01, 0.36 ± 0.06, 0.51 ± 0.07, and 0.71 ± 0.08 ml·min−1·g−1 on days 1, 7, 14, and 21, respectively) blood flow increased in animals subjected to repetitive occlusion. Sham animals exhibited no changes in blood flow. Myocardial interstitial fluid (MIF) from both groups was analyzed by two-dimensional electrophoresis with matrix-assisted laser desorption/ionization time-of-flight identification. The acute-phase protein haptoglobin was identified in the group subjected to repetitive occlusion. ELISA of MIF showed haptoglobin to be elevated at all time points of collateral development compared with sham, with maximal production on day 7. Purified haptoglobin dose dependently stimulated endothelial cells to form tubes and vascular smooth muscle cells to migrate. Purified haptoglobin did not stimulate proliferation of either cell type. The relative contribution of haptoglobin to the chemotactic properties of MIF was tested using a neutralizing antibody. Neutralized MIF could not stimulate smooth muscle cells to migrate at any time during collateral development. Endothelial cell tube formation was inhibited after the midpoint of collateralization. Therefore, the acute-phase protein haptoglobin plays a critical role during coronary collateralization.
- coronary circulation
coronary artery disease is characterized by plaque deposition, which ultimately reduces blood flow and leads to the pathophysiological sequelae of ischemic heart disease. Clinical data demonstrate that development of the coronary collateral circulation can ameliorate complications of coronary disease. Perez-Castellano et al. (33) associated reduced patient mortality from myocardial infarction with the extent of collateral development. Coronary collaterals can also reduce infarct size, preserve myocardial architecture, and improve cardiac function (5, 19).
Coronary collaterals are preexisting arterial anastomoses between coronary arteries. These vessels cannot conduct blood flow sufficient to prevent ischemia in the event of coronary occlusion because of their small caliber and high resistance in their native state (10). However, collaterals can dramatically increase in diameter and muscularity when stimulated (10). Vascular endothelial growth factor (VEGF) has been identified as important in collateralization, and administration of VEGF or transfection in vivo enhances angiogenesis and collateral growth in many models (4, 23, 24, 26). Basic fibroblast growth factor (bFGF), a mitogen and chemotactic factor for vascular smooth muscle cells (VSMC), has been identified as a contributor to limb collateralization (23). Intracoronary administration of this protein enhances development of the collateral circulation. Although these proteins have produced promising observations in the laboratory, the efficacy of these proteins in clinical trials has been less than desired. Failures of clinical investigations are likely due to the oversimplification that collateralization can be achieved by a single or a few stimulatory proteins.
We used the discovery approach of proteomics to identify other putative candidate growth factors and modulators that alter coronary blood vessel growth. Our rationale for using proteomic methods was to approach the analysis of expressed proteins in an unbiased manner without any a priori judgments about which proteins are involved in the process. Myocardial interstitial fluid (MIF) during stimulation of collateral growth was collected, and the expression of unique proteins during coronary collateral growth was analyzed using two-dimensional electrophoresis and matrix-assisted laser desorption/ionization time-of-flight spectroscopy. One of the identified proteins was haptoglobin, which is an acute-phase protein described as a cell migration factor involved in vascular remodeling (11). Because it is expressed in sufficient amounts during collateral growth to exert biological responses on cell migration, we speculate that haptoglobin plays a critical role in coronary collateral growth.
MATERIALS AND METHODS
Experimental procedures and protocols were approved by the Animal Care and Use Committee of the Medical College of Wisconsin. All procedures conformed to the “Guiding Principles for Research Involving Animals and Human Beings” (1) of the American Physiological Society and were in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals (2).
Animal instrumentation was performed according to Tessmer et al. (39). Coronary collateral development was induced with brief (2-min) repetitive (1/h; 8/day for 21 days) left anterior descending coronary artery occlusions. Systemic and coronary hemodynamics were monitored and recorded on a polygraph. Radioactive microspheres were administered during the first 2-min coronary occlusion on experimental days 1, 7, 14, and 21. Sham dogs were instrumented identically but were not subjected to repetitive coronary occlusions (with the exception of a 2-min occlusion on days 1, 7, 14, and 21 for measurements of coronary collateral blood flow).
Collection of MIF.
A miniature multiport intramyocardial catheter can be implanted into the myocardium to enable sampling of interstitial fluid on a periodic basis (Micro-Renathane MRE 080, Braintree Scientific, Braintree, MA; 2.0 mm OD, 1 mm ID). Forty 25-gauge needle holes were punctured in a 2-cm area, and the volume in this punctured area was ∼200 μl. The total dead space of the system, including the connecting catheters, was ∼500 μl. The MIF was collected daily from this indwelling intramyocardial catheter, starting on postoperative day 1. Lactated Ringer solution (4 ml) containing 10 mg of gentamicin was simultaneously infused and withdrawn, resulting in a 20-fold dilution of MIF. The 20-fold dilution was calculated on the basis of the structure of the intramyocardial catheter. The area of the catheter was perforated for the movement of the interstitial fluid, accounting for a volume of ∼200 μl. The total volume of saline in the catheter was 1 ml. Each day another 4 ml of saline were pushed through the catheter and simultaneously drawn out, resulting in the 20-fold dilution of the 200-μl accounted volume of the interstitial fluid.
The MIF was filtered, placed on ice, aliquoted into vials containing protease inhibitor cocktail tablets (Complete Mini, EDTA free, Roche), and frozen at −80°C until analysis.
Two-dimensional electrophoresis of MIF.
MIF collected from animals subjected to repetitive occlusion and sham-operated animals (those instrumented but not occluded) was analyzed using the isoelectric focusing (Bio-Rad). MIF (30 μl) was centrifuged to remove insoluble material and added to rehydration buffer (8 M urea, 2% CHAPS, 50 mM DTT, and 0.2% ampholytes; all purchased from Bio-Rad). The first-dimension separation protocol was 150 V for 2 h (rapid), 250 V for 15 min (rapid), 5,000 V for 2.5 h (slow), and 8,000 V for 35,000 V-h (slow). After separation, gel strips were equilibrated with SDS buffer containing DTT and then iodoacetamide. The second-dimension was standard SDS-PAGE with 10% and 18% polyacrylamide. Gels were stained using Silver Stain Plus (Bio-Rad). Gels were imaged on an Epson scanner with 24-bit color and 600 dpi and analyzed with PDQuest (version 6.1, Bio-Rad).
Preparation of proteins for mass spectroscopy.
Proteins of interest were excised and destained with equal volumes of potassium ferricyanide and sodium thiosulfate (Sigma). The gel slices were subsequently washed with deionized water. Spots were trypsinized and extracted according to Shevchenko et al. (36), with the exception that peptides were desalted using an equilibrated Zip Tip C18 column (Millipore). Matrix-assisted laser desorption/ionization time-of-flight experiments were performed using a Kratos Analytical Axima-CFR spectrometer.
ELISA for haptoglobin.
A 1:1,000 dilution of sheep anti-human haptoglobin antibody (Novus) in 50 mM sodium carbonate (pH 9.0) was added to each well and incubated overnight at 4°C. Wells were blocked with 3% bovine serum albumin in PBS for 2 h at room temperature. Purified human haptoglobin (100 ng/ml to 10 pg/ml; Sigma) diluted in blocking buffer was used for the standard curve. MIF was diluted 1:10 or 1:100 in blocking buffer. Samples were incubated at room temperature for 1 h and then washed three times for 5 min each with PBS and 0.05% Tween 20 (Bio-Rad). A 1:1,000 dilution of rabbit anti-human haptoglobin antibody (Sigma) in blocking buffer was added and incubated at 37°C for 30 min. Samples were washed and then incubated in 1:5,000 goat anti-rabbit alkaline phosphatase-conjugated antibody diluted in blocking buffer for 30 min at 37°C. After standard washing, there was a final wash in substrate buffer (0.9 M diethanolamine, pH 9.6, 1 mM magnesium chloride). Fifty microliters of p-nitrophenyl phosphate (1 mg/ml; Calbiochem) in substrate buffer were added to each well, and the samples were incubated for 20 min at room temperature. The reaction was stopped with 0.1 M EDTA (pH 7.5). Absorbances were read at 405 nm on a spectrophotometric plate reader. All samples were performed in triplicate.
VSMC migration assay.
A rat pulmonary artery smooth muscle cell line (PAC1) at passages 30–33 was used in all studies. Cells were seeded into six-well culture dishes (25,000 cells/well; Costar) with 10% fetal bovine serum (FBS) in Dulbecco's modified Eagle's medium (DMEM) supplemented with 1% l-glutamine, 1% sodium pyruvate, 1% streptomycin, and 1% penicillin. Cultures were incubated at 37°C and equilibrated in 95% air-5% CO2 for 72–96 h to reach 90% confluence; then growth was arrested in 0.1% DMEM for 48 h. The cell monolayer was scraped and washed in PBS to remove debris, and 2 ml of the proper medium were added. The cells were monitored at 15 and 30 h to assess migration. Ten images of each well were captured using Scion Image or MetaMorph.
The mitogenic activity of purified haptoglobin was assessed by measurement of the proliferative response of growth-arrested human aortic endothelial cells (HAEC) and PAC1 in culture. Ten thousand cells per well were seeded in DMEM containing 5% FBS overnight. Cells were growth arrested for 72 h in 0.5% FBS for HAEC and in 0.1% FBS for PAC1 in DMEM. The cells were then incubated in 10, 50, or 100 μg/ml purified human haptoglobin from pooled plasma (Sigma) in 0.1% FBS in DMEM. Other cultures were treated with 0.5 or 0.1% FBS as negative control and 10% FBS as positive control for the same interval. The number of cells was determined with a hemocytometer after 72 h.
Haptoglobin neutralization of MIF.
MIF (200 μl) obtained on experimental days 1–3, 6–8, 13–15, or 19–21 was added to 1.8 ml of 0.1% DMEM. One sample of MIF included the addition of 3 μl of sheep anti-human haptoglobin antibody (Novus). Samples were sterilized using a syringe filter (2 μm diameter; Amicon) and added to PAC1 cells prepared as described above.
Endothelial cell tube formation and haptoglobin neutralization.
HAEC were grown to confluence in Ham's F-12 medium supplemented with 15% FBS, 500 U/ml penicillin, 50 μg/ml streptomycin, and 100 μg/ml heparin (all from GIBCO) and 100 μg/ml endothelial cell (EC) growth supplement (BD Biosciences). Cells were incubated at 37°C and equilibrated in 95% air-5% CO2. Fibrin gels were prepared using endotoxin- and plasminogen-free bovine fibrinogen (5 mg/ml; Calbiochem) dissolved in serum-free medium and filtered through a 0.2-μm filter (Millex GS, Millipore). Fibrin matrices were prepared by polymerization of the fibrinogen solution using low concentrations of α-thrombin (2.5 U/ml; Sigma). After polymerization, gels were soaked in culture medium containing 10% FBS for 2 h at 37°C to inactivate the thrombin. HAEC were seeded on the surface of the three-dimensional matrix in 24-well plates in Ham's F-12 medium and cultured for 48 h in a 10% dilution of MIF in serum-free medium. Negative control was 15% FBS Ham's F-12 medium; 50 ng of VEGF were used as a positive control. Alternatively, one sample of MIF included 3 μl of sheep anti-human haptoglobin antibody (Novus). Images were captured using Scion Image or MetaMorph. Angiogenic properties were determined by calculating the total area of tubes formed after the image was superimposed on a predetermined grid, and the number of squares (0.25 μm2) of intersecting tubes was counted.
Statistical analysis of data within and between groups was performed with analysis of variance for repeated measures followed by the Student-Newman-Keuls modification of the two-tailed Student's t-test. Values are means ± SE and were considered statistically significant when P < 0.05.
Repetitive coronary occlusions stimulate collateral development.
There were no differences in heart rate or mean arterial blood pressure between groups under baseline conditions (Table 1). No temporal changes in heart rate or pressure were observed in control and sham dogs.
There were no differences in collateral blood flow between groups on day 1 (Fig. 1). Subepicardial (0.14 ± 0.01, 0.36 ± 0.06, 0.51 ± 0.07, and 0.71 ± 0.08 ml·min−1·g−1 on days 1, 7, 14, and 21, respectively) and subendocardial (0.19 ± 0.04, 0.27 ± 0.06, 0.48 ± 0.10, and 0.81 ± 0.11 ml·min−1·g−1 on days 1, 7, 14, and 21, respectively) collateral flow increased in dogs subjected to repetitive occlusion but remained unchanged in sham dogs.
Increased expression of haptoglobin during collateralization.
Experimental day 7 was of particular interest because of previous data identifying maximal growth factor production at this time. Approximately 236 unique protein spots were present on day 7 of collateralization (Fig. 2). Small amounts of protein prohibited identification in many cases; however, one unique protein identified was haptoglobin. ELISA determined relative differences in the concentration of haptoglobin in MIF throughout the experimental protocol. Figure 3 demonstrates haptoglobin production to be significantly increased on day 1, with peak expression on day 7. There was no increase in expression of haptoglobin over the course of the experiment in sham dogs.
Haptoglobin stimulates cell migration.
Haptoglobin stimulated HAEC to form tubes in culture in a dose-dependent fashion (Fig. 4A). Haptoglobin was also dose dependently chemotactic to vascular smooth muscle in a wound assay (Fig. 4B). Cell migration was modest at 15 h but was pronounced at 30 h. The monolayer in the 0.1% FBS negative control was intact, with minimal migration at 15 and 30 h.
The effect of haptoglobin inhibition on MIF-induced smooth muscle cell migration is shown in Fig. 5. VSMC migration was stimulated by MIF on day 2, was maximal through days 7–15, and returned toward baseline by day 21. When a haptoglobin antibody was coadministered, VSMC migration stimulated by MIF was effectively abolished.
The effect of haptoglobin on EC tube formation was strikingly different from results obtained in VSMC migration experiments. Maximal tube formation occurred in response to MIF from experimental day 7, with a gradual decline throughout the rest of the experiment (Fig. 6). Neutralization of haptoglobin in the MIF had no effect on tube formation at earlier time points (days 2 and 7); however, the response to MIF obtained on days 14 and 21 could be significantly attenuated.
Haptoglobin does not induce cellular proliferation.
Cellular proliferation assays performed using HAEC and PAC1 VSMC demonstrated no mitogenic actions of haptoglobin (Fig. 7).
The purpose of this study was to identify and characterize proteins previously undescribed in coronary collateral development. Haptoglobin was identified in MIF during coronary collateralization. We characterized haptoglobin as a chemotactic protein for VSMC and an inducer of tube formation by EC. Finally, we demonstrated that the quantity of haptoglobin in MIF is biologically relevant, as antibody neutralization of haptoglobin in MIF abolished VSMC migration and affected endothelial tube formation.
Haptoglobin is a tetrameric protein consisting of two α-chains and two β-chains. It is primarily expressed in the liver, but extrahepatic sites such as activated monocytes, eosinophils, pulmonary epithelial cells, fibroblasts, and tumor cell lines are sources of this protein (8, 11). Its expression is induced in response to stimulation with interleukin-6 (25).
Haptoglobin acts as an antioxidant by scavenging free hemoglobin (28). Haptoglobin has also been identified as an angiogenic protein. Serum collected from patients with chronic vascular inflammation induced EC to form tubes in culture, and haptoglobin was the protein responsible for this activity (6). Further testing with purified haptoglobin demonstrated that all phenotypes can elicit an angiogenic response in vivo, with the 2-2 phenotype producing the most robust generation of blood vessels (6). Other research on the role of haptoglobin focused on impaired lipopolysaccharide-induced migration of haptoglobin-null murine embryonic fibroblasts (11).
Many clinical studies support the impact of haptoglobin on blood vessel growth. Peripheral diabetic retinopathy causes vision impairment by stimulating angiogenesis around the retina. Patients with a haptoglobin 1-1 phenotype, a phenotype found to be the least stimulative for cell migration, experience reduced ocular angiogenesis (29). The Belgian Interuniversity Research on Nutrition and Health Study found the overall death rate from myocardial infarction to be reduced in patients carrying the 2-2 haptoglobin genotype (9). In patients expressing haptoglobin 2-2, coronary bypass graft survival time is reduced because of increased intimal hyperplasia (12, 35). One study suggested that a haptoglobin 1 allele in diabetic patients improves coronary collateralization; however, this is likely to be an indirect effect of the improved antioxidant function of the 1-1 allele (16, 20). Densem et al. (13) investigated the influence of haptoglobin polymorphism on the incidence of coronary artery disease after transplantation. They report that the haptoglobin 2–1 phenotype is the most important predictor for the development of cardiac transplant vasculopathy.
The present results indicate the profound effect of haptoglobin inhibition on VSMC migration throughout coronary collateral development. VSMC migration is stimulated by growth factors such as bFGF or platelet-derived growth factor (23, 30, 34, 38). Previous experiments could not identify a significant quantity of bFGF in MIF (40). The paucity of potent VSMC chemotactic proteins in MIF leads us to believe that haptoglobin may play a significant role in VSMC migration.
Haptoglobin's impact on VSMC migration is greater than EC tube formation in our model of coronary collateralization. Neutralization of haptoglobin in MIF did not attenuate endothelial tube formation until after the midpoint of collateral development, in contrast to reductions in VSMC migration, which occurred in all phases of development. This may be secondary to the numerous EC chemotactic proteins identified in MIF, such as VEGF and angiopoietin-1 (26, 40). VEGF expression is under the control of hypoxia, as it contains a promoter that binds hypoxia-inducible factor-1α (14). VEGF production peaks at day 6 in the repetitive occlusion protocol and gradually decreases over time as the collateral circulation develops and ischemia wanes (27). Once VEGF levels decrease, haptoglobin stimulation of EC tube formation may become the predominant mechanism.
An interesting finding of our study was that haptoglobin's action was limited to cell migration. We tested EC and VSMC proliferation in vitro and observed no effect of haptoglobin. Few factors are capable of stimulating cell migration in the absence of proliferation. For example, insulin-like growth factor I has been shown to be a weak stimulator of smooth muscle cell proliferation but can potentiate cell migration (30).
Haptoglobin is an angiogenic protein, but it in some ways is not comparable to other growth factors that induce collateralization. The range of effective concentrations for haptoglobin is 10–100 μg/ml, well within normal plasma concentrations (0.3–3 mg/ml and extending to 8 mg/ml during inflammation) (21). Interstitial concentrations would be expected to be lower because of reduced permeability across the endothelium. However, on the basis of ELISA measurements, biologically relevant concentrations of haptoglobin exist in MIF when collateralization is stimulated. Because there is approximately a 20-fold dilution of MIF, realized values of haptoglobin exist at 20–40 μg/ml. Haptoglobin has a limited effective range (∼1 order of magnitude) in vitro, and the maximum effective concentration is ∼100 μg/ml. Physiological concentrations for growth factors such as VEGF and bFGF are in the nanogram and picogram range (15, 32, 37, 40). The effect of 100 μg/ml haptoglobin was comparable to 50 ng/ml VEGF (Fig. 4).
The present study highlights the contribution of an inflammatory regulated protein on coronary collateralization. Mounting clinical and experimental evidence has indicated that inflammation is intimately associated with collateralization. Secretion of inflammatory cytokines such as TNF-α and interleukins can induce cell migration and proliferation (31). Increases in neutrophils and monocytes are also associated with improved collateralization in the hindlimb (3, 17, 22). Patients with giant cell arteritis have increased interleukin-6 with a resultant increase in angiogenesis and reduced ischemic events (7, 18). Localized enhancement of inflammation as a means to direct blood vessel development may be a useful approach in therapeutic angiogenesis. The present investigation has shown haptoglobin, a protein increased in inflammation, to be essential for coronary collateral development.
This work was supported in part by National Institutes of Health Grants HL-054820, GM-008377, and GM-066730 (to D. C. Warltier) and HL-065203 (to W. M. Chilian).
We gratefully acknowledge John P. Tessmer and David A. Schwabe for technical assistance and Mary Lorence-Hanke for assistance in preparation of the manuscript.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2005 by the American Physiological Society