Free hemoglobin induction of pulmonary vascular disease: evidence for an inflammatory mechanism

Abstract

Cell-free hemoglobin (Hb) exposure may be a pathogenic mediator in the development of pulmonary arterial hypertension (PAH), and when combined with chronic hypoxia the potential for exacerbation of PAH and vascular remodeling is likely more pronounced. We hypothesized that Hb may contribute to hypoxia-driven PAH collectively as a prooxidant, inflammatory, and nitric oxide (NO) scavenger. Using programmable micropump technology, we exposed male Sprague-Dawley rats housed under room air or hypoxia to 12 or 30 mg per day Hb for 3, 5, and 7 wk. Blood pressure, cardiac output, right ventricular hypertrophy, and indexes of pulmonary vascular remodeling were evaluated. Additionally, markers of oxidative stress, NO bioavailability and inflammation were determined. Hb increased pulmonary arterial (PA) pressure, pulmonary vessel wall stiffening, and right heart hypertrophy with temporal and dose dependence in both room air and hypoxic cohorts. Hb induced a modest increase in plasma oxidative stress markers (malondialdehyde and 4-hydroxynonenal), no change in NO bioavailability, and increased lung ICAM protein expression. Treatment with the antioxidant Tempol attenuated Hb-induced pulmonary arterial wall thickening, but not PA pressures or ICAM expression. Chronic exposure to low plasma Hb concentrations (range = 3-10 μM) lasting up to 7 wk in rodents induces pulmonary vascular disease via inflammation and to a lesser extent by Hb-mediated oxidation. Tempol demonstrated a modest effect on the attenuation of Hb-induced pulmonary vascular disease. NO bioavailability was found to be of minimal importance in this model.

Fig. 1.

Infusion pump, plasma, and urine analysis. A: mean plasma heme concentrations in high-dose normoxic Hb-infused rats obtained at 4 time points each day are represented as bars. B: mean urinary heme concentrations obtained in high-dose Hb-infused rats at 4 time points each day are represented as bars. Total excreted heme or area under urine concentration-vs.-time curves are shown for heme iron redox states and hemichrome. C: kidney heme oxygenase-1 (HO-1) expression. D and F: kidney iron deposition in saline- and hemoglobin-infused rats. E and G: kidney collagen deposition in saline- and hemoglobin-infused rats. Black arrows show regions of collagen (blue) deposition. All data are shown as means ± SE.

Fig. 2.

Pulmonary artery and pulse pressures. A and B: pulmonary arterial pressures of normoxic and hypoxic rats chronically infused with hemoglobin. C and D: pulmonary pulse pressures of normoxic and hypoxic rats chronically infused with hemoglobin. Data are represented as fold change ± SE vs. saline-infused cohorts. *P < 0.05 vs. saline-infused cohorts; †P ≤ 0.01 vs. saline-infused cohorts; ††P ≤ 0.001 vs. 3- and 7-wk hemoglobin-infused groups.

Fig. 3.

Lung histopathology after chronic free hemoglobin infusion. Microphotograph images of lung sections stained with hematoxylin and eosin taken from animals exposed to either normoxic (NX) or hypoxic (HX) environments and chronically infused with hemoglobin. Black arrows show regions of extravascular macrophage or neutrophil infiltration. Original magnification ×20.

Fig. 4.

Morphology analyses. Muscularization of small arteries in normoxic (A) and hypoxic (B) rats chronically treated with hemoglobin. Vessel wall thickening of small arteries of normoxic (C) and hypoxic (D) rats chronically treated with hemoglobin. *P ≤ 0.04 vs. saline-infused cohorts; †P < 0.001 vs. saline-infused cohorts; ††P < 0.01 vs. 3- and 5-wk hypoxia-alone groups.

Fig. 5.

Plasma hemoglobin and nitric oxide relationship. Representative Western blot from each group spliced from a contiguous lane. Lanes were comprised of 3 normoxic Hb-treated, 3 hypoxic Hb-treated, normoxic sham, and hypoxic sham. Densitometry analyses of endothelial nitric oxide synthase (eNOS) at 3 wk (A), 7 wk (B), and 5 wk (C). Nitric oxide bioavailability in high-dose rats. D: plasma nitrite (NO₂⁻) plotted against days. E: plasma nitrite plotted against plasma heme. *P = 0.45 vs. normoxic control. In instances that densitometry was compared across gels, samples were derived at same time and processed in parallel.

Fig. 6.

Western blot analyses of lung HO-1. Representative blot from each group spliced from a contiguous lane. Lanes were comprised of 3 normoxic Hb treated, 3 hypoxic Hb treated, normoxic sham, hypoxic sham, and densitometry analyses: 3 wk (A), 7 wk (B), and 5 wk (C) of hemoglobin infusion. Data are represented as fold change ± SE vs. normoxic saline-infused cohorts. *P ≤ 0.045 vs. saline-infused cohorts; †P < 0.001 vs. saline-infused cohorts. In instances that densitometry was compared across gels, samples were derived at the same time and processed in parallel.

Fig. 7.

Representative microphotographs of stained lung sections for HO-1. A: merged image of lung: HO-1 expression (red), smooth muscle cell actin (green), and 4,6-diamidino-2-phenylindole (DAPI; blue). Original magnification ×20. White arrows show area of HO-1 expression. B: image of pulmonary macrophages expressing HO-1; HO-1 red, CD163 green, and DAPI blue. Original magnification ×60. White arrows point to macrophages expressing HO-1.

Fig. 8.

Lipid peroxidation products. A and B: data of malondialdehyde (MDA) and 4-hydroxynonenal (4-HNE) in normoxic and hypoxic rats infused with free hemoglobin. Data are represented as fold change ± SE vs. saline-infused cohorts. C: Hb main effect: data of Hb-infused vs. saline-infused animals. *P ≤ 0.045 vs. saline-infused cohorts; †P < 0.001 vs. saline-infused cohorts.

Fig. 9.

Hemoglobin reactivity with hydrogen peroxide in the presence Tempol. A: spectral analysis of ferrous hemoglobin at time 0 (black lines) and after 90 min of being exposed to hydrogen peroxide (H₂O₂, 10 μM) generated via glucose and 5 mU glucose oxidase (gray line). B: addition of sodium sulfide (Na₂S) 2 mM shows the formation of sulfhemoglobin with an absorbance at 620 nm. C: spectral analysis of ferrous hemoglobin at time 0 (black lines) and after 90 min of being exposed to H₂O₂ (gray line) in the presence of Tempol. D: addition of Na₂S indicates that Tempol limited the formation ferryl hemoglobin or accelerated its reduction to ferric hemoglobin. E: conversion of ferrous hemoglobin to ferric hemoglobin in the presence of hydrogen peroxide with (black) and without (gray). F: concentration of ferryl formation from the reaction between hydrogen peroxide and hemoglobin in the presence or absence of Tempol.

Fig. 10.

Tempol treatment attenuated free hemoglobin-induced pulmonary wall thickening and lipid peroxidation products. A: vessel wall-to-lumen ratio of normoxic and hypoxic rats chronically infused with free hemoglobin and treated with Tempol. B: data of MDA and 4-HNE in rats infused with free hemoglobin and treat with Tempol. Data are represented as means ± SE. †P < 0.0001 vs. rats not treated with Tempol.

Fig. 11.

Western blot analyses of intercellular adhesion molecule-one (ICAM-1) concentration. Representative blot from each group spliced from a contiguous lane. Lanes were comprised of 3 normoxic Hb treated, 3 hypoxic Hb treated, normoxic sham, hypoxic sham, and densitometry analyses 5 wk of free hemoglobin infusion. Data are represented as fold change ± SE vs. normoxic saline-infused cohorts. White arrows show regions of interest for endothelial factor 8 and ICAM staining and merge of the stains. *P ≤ 0.045 vs. saline-infused cohorts; †P < 0.001 vs. saline-infused cohorts. In instances that densitometry was compared across gels, samples were derived at same time and processed in parallel.

Fig. 12.

Representative microphotographs of stained lung sections for ICAM. Pulmonary images of ICAM-1 expression with factor 8 (endothelial marker). ICAM expression (green) and factor 8 (red), DAPI blue merge creates orange/yellow. Original magnification ×40. White arrows show regions of endothelial ICAM expression.