If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Address reprint requests to Allan Lawrie, Ph.D., Department of Cardiovascular Science University of Sheffield, The Medical School, Beech Hill Rd., Sheffield, S10 2RX, United Kingdom
Department of Cardiovascular Science, University of Sheffield, Sheffield, United KingdomCardiovascular Biomedical Research Unit, National Institute for Health Research, Sheffield, United Kingdom
Department of Cardiovascular Science, University of Sheffield, Sheffield, United KingdomCardiovascular Biomedical Research Unit, National Institute for Health Research, Sheffield, United Kingdom
Department of Cardiovascular Science, University of Sheffield, Sheffield, United KingdomCardiovascular Biomedical Research Unit, National Institute for Health Research, Sheffield, United Kingdom
Cardiovascular Biomedical Research Unit, National Institute for Health Research, Sheffield, United KingdomSheffield Pulmonary Vascular Disease Unit, Royal Hallamshire Hospital, Sheffield, United Kingdom
Department of Cardiovascular Science, University of Sheffield, Sheffield, United KingdomCardiovascular Biomedical Research Unit, National Institute for Health Research, Sheffield, United Kingdom
Inflammatory mechanisms are proposed to play a significant role in the pathogenesis of pulmonary arterial hypertension (PAH). Previous studies have described PAH in fat-fed apolipoprotein E knockout (ApoE−/−) mice. We have reported that signaling in interleukin-1–receptor–knockout (IL-1R1−/−) mice leads to a reduction in diet-induced systemic atherosclerosis. We subsequently hypothesized that double-null (ApoE−/−/IL-1R1−/−) mice would show a reduced PAH phenotype compared with that of ApoE−/− mice. Male IL-1R1−/−, ApoE−/−, and ApoE−/−/IL-1R1−/− mice were fed regular chow or a high-fat diet (Paigen diet) for 8 weeks before phenotyping for PAH. No abnormal phenotype was observed in the IL-1R1−/− mice. Fat-fed ApoE−/− mice developed significantly increased right ventricular systolic pressure and substantial pulmonary vascular remodeling. Surprisingly, ApoE−/−/IL-1R1−/− mice showed an even more severe PAH phenotype. Further molecular investigation revealed the expression of a putative, alternatively primed IL-1R1 transcript expressed within the lungs but not aorta of ApoE−/−/IL-1R1−/− mice. Treatment of ApoE−/− and ApoE−/−/IL-1R1−/− mice with IL-1–receptor antagonist prevented progression of the PAH phenotype in both strains. Blocking IL-1 signaling may have beneficial effects in treating PAH, and alternative IL-1–receptor signaling in the lung may be important in driving PAH pathogenesis.
Pulmonary arterial hypertension (PAH) is a life-threatening condition with high morbidity and mortality.
Progressive pulmonary vascular remodeling that is characterized by disorganized cellular proliferation leads to an increase in pulmonary artery pressure, right ventricular hypertrophy, failure, and death.
American College of Cardiology Foundation Task Force on Expert Consensus Documents, American Heart Association, American College of Chest Physicians, American Thoracic Society Inc, Pulmonary Hypertension Association ACCF/AHA 2009expert consensus document on pulmonary hypertension a report of the American College of Cardiology Foundation Task Force on Expert Consensus Documents and the American Heart Association developed in collaboration with the American College of Chest Physicians; American Thoracic Society, Inc.; and the Pulmonary Hypertension Association.
Right ventricular function and failure: report of a National Heart, Lung, and Blood Institute Working Group on cellular and molecular mechanisms of right heart failure.
The importance of peroxisome proliferator-activated receptor γ (PPARγ) signaling (which is associated with insulin resistance) in PAH was shown by Hansmann et al
when describing PAH in Western diet–fed apolipoprotein E–deficient mice (ApoE−/−). Treatment with a PPARγ agonist reversed disease in this model and mice with the targeted deletion of PPARγ in either endothelial or smooth muscle cells develop PAH. Further work subsequently has linked this mechanism to aberrant bone morphogenetic protein (BMP), platelet-derived growth factor signaling, and reduced adiponectin.
Interactions between inflammatory signaling and vascular cells are a key aspect of vascular injury/repair and disease and are considered to have an important role in the pathogenesis of PAH. Previous studies have shown that numerous inflammatory cytokines, including those from the interleukin family, are up-regulated in PAH.
Interestingly, overexpression of IL-6 in the lungs of mice is sufficient to induce an increase in right ventricular systolic pressure (RVSP), pulmonary vascular remodeling, and right ventricular hypertrophy (RVH).
Furthermore, treatment of rats with the naturally occurring IL-1–receptor antagonist (IL-1ra) has been shown to protect against the development of monocrotaline-induced PAH, although interestingly not against hypoxia-induced PAH.
further supporting the concept that the IL-1 pathway also may be an important link between PPAR signaling and PAH.
Our group recently showed that feeding a high-fat, high-cholate diet (Paigen diet) to ApoE−/− mice for 8 weeks doubled the atherosclerotic lesion size in the aortic sinus compared with mice fed a high-fat (Western) diet. Furthermore, Paigen diet–fed mice that were doubly deficient for both ApoE and the IL-1–receptor type 1 (IL-1R1) (ApoE−/−/IL-1R1−/−) had a significantly reduced lesion size and lower systemic blood pressure compared with ApoE−/− mice on the same diet.
Because mice fed a Western diet have been shown to develop pulmonary hypertension, we hypothesized the following: feeding ApoE−/− mice the Paigen diet for 8 weeks would induce a more severe form of PAH than previously reported with the Western diet; given the known links with inflammation and PAH, ApoE−/−/IL-1R1−/− mice would show a reduced or protected PAH phenotype.
We report here that Paigen diet–fed ApoE−/− mice develop moderately severe PAH with obliterative pulmonary vascular remodeling. Surprisingly, ApoE−/−/IL-1R1−/− mice showed a more severe phenotype showing lesions with dysregulated elastin, subsequent medial hypertrophy, and neointimal formation similar to that found in the human disease. After further investigation into potential molecular mechanisms, we discovered a putative, alternatively primed IL-1R1 transcript
expressed within the lungs of the ApoE−/−/IL-1R1−/− double-null mice. Treatment of ApoE−/− and ApoE−/−/IL-1R1−/− mice with IL-1ra, as disease was developing, prevented an increased RVSP and pulmonary vascular remodeling in both models, suggesting that targeting IL-1 may have potential benefit in PAH.
Materials and Methods
Animals
All mice were on a C57BL/6 background. IL-1R1−/− (JAX 3245) and ApoE−/− (JAX 2052) were obtained from Jackson Laboratories (Bar Harbor, ME), and ApoE−/−/IL-1R1−/− mice were generated as previously described.
Male mice, 10 to 12 weeks of age (7 per group), were fed normal chow (4.3% fat, 0.02% cholesterol, and 0.28% sodium) or a Paigen diet (18.5% fat, 0.9% cholesterol, 0.5% cholate, and 0.259% sodium) for 8 weeks. Diets were supplied by Special Diet Services (Braintree, Essex, UK). All animal experiments were approved by the University of Sheffield Project Review Committee and conformed to UK Home Office ethical guidelines. Where stated, human IL-1Ra or placebo (MTA 200517250-001; Amgen, Inc., Thousand Oaks, CA) was administered by Alzet (Cupertino, CA) 1004 osmotic mini pumps (100-μL reservoir; delivery rate, 0.1 μL/h) set up to deliver 10 μg/h for 4 weeks, after an initial 4 weeks on diet. Pumps were primed and implanted subcutaneously as per the manufacturer's instructions.
Echocardiography
Echocardiography was performed using the Vevo 770 system (Visual Sonics, Toronto, Canada) using the RMV707B scan head. Mice were placed on a heated platform and covered to minimize heat loss. Rectal temperature, heart rate, and respiratory rate were recorded continuously throughout the study. Anesthesia was induced and maintained using isoflurane through oxygen, maintaining heart rates at around 450 to 500 beats per minute whenever possible. The mice were depilated and preheated ultrasound gel was applied (Aquasonics 100 Gel; Parker Labs, Inc., Fairfield, NJ). From the right parasternal long axis view, right ventricle free wall parameters were determined using M-mode. Standard parameters of the left ventricle were measured using 2-dimensional, M-mode and Doppler pulse wave in the short axis view at the level of the papillary muscles. Cardiac output was derived from flow and annulus diameter at the junction between the outflow tract and aortic valve. Cardiac index then was normalized by body weight. Analysis was performed offline using the accompanying software (Vevo 770, V3.0; Visual Sonics). Measurements were taken during the relevant phase of the cardiac cycle that did not coincide with inspiration artifact.
Cardiac Magnetic Resonance Imaging
Anesthesia was induced and maintained using isoflurane, and mice were placed in a custom-built acrylic magnet capsule for imaging. Inside the capsule, a nonmagnetic ceramic heated hot air system (SAII–MR-compatible Heater System for Small Animals, Small Animal Instruments, Stony Brook, NY) and rectal probe, integrated into the physiological monitoring system, maintained the temperature of the animal. Mice were imaged in a 7-Tesla magnet (BioSpecAVANCE, 310-mm bore, magnetic resonance imaging system B/C 70/30; Bruker, Coventry, UK), with a pre-installed, 12-channel, RT-shim system (B-S30) and fitted with an actively shielded, 116-mm inner-diameter, water-cooled, 3-coil gradient system (BioSpin magnetic resonance imaging GmbH B-GA12; 400 mT/m maximum strength per axis with 80-μs ramps; Bruker) to assess ventricular size after 8 weeks. A hydrogen-1 (1H) birdcage volume resonator (300 MHz, 1 kW maximum, outer diameter, 114 mm; inner-diameter, 72 mm; Bruker) placed at the iso-center of the magnet was used for both radiofrequency transmission and reception. A workstation configured for use with ParaVision 4.0 software (Bruker) operated the spectrometer. After field shimming, off-resonance correction and radiofrequency gain setting a triplane self-gated (in slice navigator) Fast Low Angle SHot (FLASH) sequence (TR, 44.5 ms; TE, 2.3 ms; flip angle, 7.5°; field of view, 60 mm × 60 mm; slice thickness, 1 mm; matrix, 128 × 128) was used for subject localization. A further self-gated FLASH sequence (Repetition Time, 5.8 ms; Echo Time, 2.3 ms; flip angle, 7.5°; field of view, 35 mm × 35 mm; slice thickness, 1.5 mm; matrix, 128 × 128) was used for single-slice cine magnetic resonance imaging of the cardiac long axis. This was used to plan for short-axis cine magnetic resonance imaging (TR, 12.6 ms; TE, 2.3 ms; flip angle, 7.5°; field of view, 35 mm*35 mm; slice thickness, 1.5 mm; matrix, 128*128). An in-parallel saturation slice was applied to null the blood signal and increase ventricle wall contrast. For data processing, ParaVision IntraGate software (Bruker) using Fourier filtering techniques
was used for navigator analysis and reconstruction of k-space data.
Cardiac Catheterization
After echocardiography, left and right ventricular catheterization was performed using a closed chest method via the right internal carotid artery and the right external jugular vein under isoflurane-induced anesthesia. Data were collected using a Millar ultra-miniature pressure-volume PVR-1045 1F catheter (Millar Instruments, Inc., Houston, TX) coupled to a Millar MPVS 300 and a PowerLab 8/30 data acquisition system (AD Instruments, Oxfordshire, UK) and recorded using Chart v7 software (AD Instruments). Pressure volume analysis was performed using PVAN v2.3 (Millar Instruments, Inc.).
RVH
RVH was measured using a modification of the Fulton index
by expressing the weight of the right ventricle free wall relative to the left ventricle and septum.
Immunohistochemistry and Immunoblotting
Immediately after harvest, the left lung was perfusion fixed via the trachea with 10% formal buffered saline by inflation to 20 cm of H2O. The lungs then were processed into paraffin blocks for sectioning. Paraffin-embedded sections of lung were histologically stained for Alcian Blue elastin van Gieson, immunohistochemically stained for α-smooth muscle actin (M0851; Dako, Cambridgeshire, UK) to visualize smooth muscle cells, von Willebrand factor (A0082; Dako) to visualize endothelial cells, CD3 (ab5690; Abcam, Cambridge, UK) to visualize T-lymphocytes, CD4 (ab51312; Abcam) to visualize helper T-lymphocytes, and F4/80 (ab6640; Abcam) to visualize macrophages. Antibodies to osteoprotegerin (OPG) (ab73400; Abcam) and tumor necrosis factor–related apoptosis-inducing ligand (TRAIL) (ab2435; Abcam) also were used to localize protein expression to pulmonary vascular lesions. Standard immunohistochemical techniques were applied. Mouse aortic tissue was used as a positive control for both von Willebrand factor and α-smooth muscle actin immunoreactivity, mouse spleen and human tonsil were used as a positive control for all other antibodies. In all cases, both IgG and no primary antibody-negative controls were used. For immunoblotting to detect IL-1R1 protein expression, whole-lung lysates were run on Invitrogen (Paisley, UK) 4% to 12% Bis-Tris gels as previously described,
and incubated with an IL-1R1 antibody (ab40774; Abcam) as per the manufacturer's instructions.
Quantification of Pulmonary Vascular Remodeling
To assess the degree of pulmonary arterial remodeling, microscopic images of Alcian Blue elastin van Gieson– and α-smooth muscle actin–stained lung sections were analyzed using a NIS-Elements Basic Research software (Nikon, Kingston upon Thames, Surrey, UK). The degree of muscularization and the percentage of affected vessels were calculated based on previously published methods.
In each lung section pulmonary arteries were categorized as occluded (ie, only a slit-like, or no, lumen remaining), muscularized (ie, with crescent, or complete rings of muscle), or nonmuscular (ie, no apparent muscle). The degree of muscularization was measured using the area of the media divided by the cross-sectional area of the whole artery (media/CSA) and the percentage of muscularization was calculated based on the number of affected pulmonary arteries divided by the total number of arteries multiplied by 100. Pulmonary arteries also were categorized based on their external diameter and divided into 3 groups: small pulmonary arterioles with a diameter less than 50 μm, medium pulmonary arteries with a range in diameter from 51 to 100 μm, and large pulmonary arteries with a diameter greater than 100 μm.
Enzyme-Linked Immunosorbent and Cytometric Bead Assays
To assess the levels of soluble cytokines in mouse serum we used Cytometric Bead Assay Flex sets (BD Biosciences, Oxford, UK) for IL-1 β, IL-6, and regulated on activation normal T-cell expressed and secreted (RANTES), and an enzyme-linked immunosorbent assay for OPG (R&D Systems, Abingdon, UK).
qPCR
To assess the level of gene expression in RNA isolated from mouse lung and aorta, RNA was extracted using RNA/DNA/protein isolation kits (Norgen Biotek, Thorold, Ontario, Canada), and reverse transcribed using Superscript III (Invitrogen). Gene expression was measured by performing TaqMan PCR using Gene Expression MasterMix and Gene Expression Assays (Applied Biosystems, Warrington, Cumbria) for IL-1R1 [assay ID Mm00434237_m1 (exon 6/7) and Mm01226959_m1 (exon 1/2)], OPG (assay ID Mm00435452_m1), and TRAIL (assay ID Mm00437174_m1). Gene expression was normalized to 18S ribosomal RNA (assay ID 999999011) using the ΔΔCT comparative quantification method as previously described.
Right Ventricular Hemodynamic Measurements Are Abnormal in Paigen Diet–Fed ApoE−/− Mice and More Exaggerated in ApoE−/−/IL-1R1−/− Mice
Fat feeding (Paigen diet) to IL-1R1−/− mice had no effect on RVSP (Figure 1A) or right ventricular end-diastolic pressure (Figure 1C). In contrast, ApoE−/− mice fed the high-fat diet (Paigen) developed a significant increase in both RVSP (Figure 1A) and right ventricular end-diastolic pressure (Figure 1C) compared with chow-fed littermates. Surprisingly, double-deficient ApoE−/−/IL-1R1−/− fed the same diet also showed a significantly increased RVSP (Figure 1A) and right ventricular end-diastolic pressure (Figure 1C) compared with chow-fed controls. Intriguingly, the RVSP observed in the Paigen diet–fed ApoE−/−/IL-1R1−/− mice was significantly higher than in ApoE−/− mice on the same diet (Figure 1A). Increased RV contractility, as measured by dP/dtmax (derivative of the maximal rate of ventricular pressure rise during isovolumetric contraction phase), also was increased significantly in Paigen diet–fed ApoE−/− mice and further increased in ApoE−/−/IL-1R1−/− mice (Figure 1E). Doppler ultrasound assessment of pulmonary artery acceleration time, which has an inverse relationship to RVSP, also was decreased in both the Paigen diet–fed ApoE−/− and ApoE−/−/IL-1R1−/− mice, compared with chow-fed littermates (Figure 1H).
Figure 1Hemodynamic characterization. Bar graphs show RVSP in mm Hg (A), left ventricular end-systolic pressure (LVESP) in mm Hg (B), right ventricular end-diastolic pressure (RVEDP) in mm Hg (C), left ventricular end-diastolic pressure (LVEDP) in mm Hg (D), right ventricular dP/dtmax in mm Hg/s (E), left ventricular dP/dtmax in mm Hg/s (F), cardiac index in mL/min/g (G), and pulmonary artery acceleration time in ms (H). A–F: All measurements were made using closed chest cardiac pressure volume catheterization, and measurements for G and H were made using Vevo 770 echocardiography equipment. Bars represent mean ± SEM, n = 4 to 7. *P < 0.05, **P < 0.01 compared with chow-fed littermates. Statistical differences between Paigen diet–fed mice are shown as assessed by analysis of variance followed by Bonferroni post hoc analysis. ***P < 0.001.
To confirm this increase in RVSP was arising from the pulmonary circulation, left ventricular pressures also were measured. No significant changes in left ventricular end-systolic pressures or left ventricular end-diastolic pressures were observed (Figure 1, B and D). Similarly, there were no changes in LV contractility, as measured by dP/dtmax, in any group (Figure 1F). There was a significantly reduced cardiac index between the chow-fed ApoE−/− and ApoE−/−/IL-1R1−/− mice but there was no significant effect of the Paigen diet on the cardiac index of either genotype (Figure 1G).
Right Ventricular Remodeling in Paigen Diet–Fed ApoE−/− and ApoE−/−/IL-1R1−/− Mice
Volume analysis of the right ventricle, using a conductance catheter, showed a significant increase in dV/dtmax in the Paigen diet–fed ApoE−/− mice, and ApoE−/−/IL-1R1−/− mice compared with chow-fed littermates (Figure 2A), although a statistically significant increase in RVH was observed only in Paigen diet–fed ApoE−/−/IL-1R1−/− mice (Figure 2B). In addition, the increase in dV/dtmax and the degree of RVH was significantly greater for the Paigen diet–fed ApoE−/−/IL-1R1−/− mice compared with Paigen diet–fed ApoE−/− mice. Significant atrial enlargement also was observed in the Paigen diet–fed ApoE−/−/IL-1R1−/− mice compared with chow-fed littermates, although we did not dissect down further to determine whether this primarily was caused by right atrium enlargement (Table 1). Echocardiographic imaging revealed a trend for an increase in the internal diameter of the right ventricle in diastole (Figure 2C), with no observed changes in the left ventricle (Figure 2D). Image processing of H&E-stained hearts, and live cardiac magnetic resonance imaging at the level of the papillary muscles, further illustrated the increase in internal right ventricular size without thickening of the right ventricular wall, in the Paigen diet–fed ApoE−/, and particularly the ApoE−/−/IL-1R1−/− mice (Figure 2E).
Figure 2Assessment of right ventricular volume and hypertrophy. Bar graphs show right ventricular dV/dtmax in μL/s as measured by closed chest pressure volume cardiac catheterization (A), Fulton Index of RVH (B). Right ventricular internal diameter at diastole (RVIDd) (C), left ventricular internal diameter at diastole (LVIDd) measured in mm as assessed by M-mode echocardiography (D), and histologic and live cardiac magnetic resonance images at the level of the papillary muscles (E). Right ventricle internal area is colored blue. Bars represent mean ± SEM, n = 4 to 7. *P < 0.05, ***P < 0.001 compared with chow-fed littermates. Statistical differences between Paigen diet–fed mice are shown as assessed by analysis of variance followed by Bonferroni post hoc analysis.
Increased Pulmonary Vascular Remodeling in Fat-Fed ApoE−/− and ApoE−/−/IL-1R1−/− Mice
Histologic and immunohistochemistry analysis of serial lung tissue samples revealed no evidence of any pulmonary vascular remodeling in either the ApoE−/− or ApoE−/−/IL-1R1−/− mice fed a regular chow diet. Similarly, there was no evidence of any effect of the Paigen diet on pulmonary vascular remodeling in the IL-1R1−/− mice (data not shown). Analysis of lungs from Paigen diet–fed ApoE−/− and ApoE−/−/IL-1R1−/− mice did, however, reveal evidence of a severe pulmonary vasculopathy, including heavily muscularized and obliterative lesions in the small resistance pulmonary arteries (<50 μm) in both genotypes. Further assessment of the extent of pulmonary vascular remodeling by quantifying the degree of muscularization as previously described
subsequently was performed. Analysis of the media/CSA ratio of the whole artery or arteriole, and the percentage of nonmuscularized, muscularized, and obliterated vessels, was assessed in small (<50 μm), medium (51 to 100 μm), and large pulmonary arteries (>100 μm). The media/CSA ratio of the small pulmonary arteries and arterioles (<50 μm) in both the Paigen diet–fed ApoE−/− and ApoE−/−/IL-1R1−/− mice was increased significantly (Figure 3A) compared with chow-fed controls. Although there was no significant difference in media/CSA between the Paigen diet–fed mice, segmentation of small pulmonary arteries into those with no signs of muscularization, muscularized, or those obliterated/fully occluded showed a significant decrease in nonmuscular arteries, and a significant increase in muscularized lesions in the Paigen diet–fed ApoE−/−/IL-1R1−/− mice compared with the ApoE−/− mice (Figure 3B). In the medium-sized pulmonary arteries (51 to 100 μm), there was a significant increase in the media/CSA in the Paigen diet–fed ApoE−/− and this was increased further in the ApoE−/−/IL-1R1−/− mice (Figure 3C). There were no signs of muscularization in the large (>100 μm) pulmonary arteries in the Paigen diet–fed ApoE−/− mice but there was a significant increase in the media/CSA of the Paigen diet–fed ApoE−/−/IL-1R1−/− mice (Figure 3D). We have hypothesized that the more severe pulmonary hemodynamic changes and RVH seen in the Paigen-fed ApoE−/−/IL-1R1−/− mice compared with the Paigen diet–fed ApoE−/− mice are related to the more extensive remodeling of the small resistance vessels, although changes seen in the more proximal pulmonary vasculature also may contribute.
Figure 3Quantification of pulmonary artery muscularization. Bar charts showing the degree of medial wall thickness as a ratio of total vessel size (media/CSA) (A, C, and D), and the percentage of affected vessels (B). A and B: Data from pulmonary arteries less than 50 μm in diameter. C: Data from vessels from 51 to 100 μm in diameter. D: Data from vessels more than 100 μm in diameter. Bars represent mean ± SEM, n = 4 to 7. *P < 0.05, **P < 0.01, ***P < 0.001 compared with chow-fed littermates. Statistical differences between Paigen diet–fed mice are shown as assessed by analysis of variance followed by Bonferroni post hoc analysis.
The lesions show many features of human disease including dysregulated elastin and increased collagen as shown by the Alcian Blue elastin van Gieson staining (Figure 4) and disorganized smooth muscle cell proliferation. Increased smooth muscle cell accumulation (α-smooth muscle actin) was observed in hypertrophied medial layers, and within the neointima of lesions. All lesions showed an intact endothelium (von Willebrand factor) (Figure 4). To determine whether these obliterative pulmonary arterial lesions were associated with infiltration of inflammatory cells we performed immunohistochemistry for macrophages and T lymphocytes. By using F4/80 as a marker for macrophages we failed to see a significant association of F4/80-positive cells with any of the remodeled pulmonary arteries (data not shown). The vascular and perivascular regions of lung from the chow-fed mice showed very little T-lymphocyte infiltration, however, there was an influx of CD3+ cells to perviscular regions of remodeled pulmonary arteries in fat-fed ApoE−/− and ApoE−/−/IL-1R1−/− mice (Figure 4). Only a few isolated CD4+ cells were observed in the perivascular regions of both Paigen diet–fed ApoE−/− and ApoE−/−/IL-1R1−/− mice (Figure 4).
Figure 4Representative photomicrographs of serial lung sections from chow- and Paigen diet–fed mice. Sections were stained with Alcian Blue elastic van Gieson (ABEVG) or immunostained for α-smooth muscle actin (α-SMA), von Willebrand factor (vWF), CD3, or CD4. The ApoE−/−/IL-1R1−/− mice fed either chow or Paigen diet displayed some nonspecific CD3 immunoreactivity in the airway epithelium. Representative heavily muscularized arteries are shown. Original magnification, ×400. Scale bar = 50 μm.
Inflammatory Mediators Are Up-Regulated in the Paigen Diet–Fed ApoE−/−/IL-1R1−/− Mice
To investigate possible mechanisms that might be driving disease pathogenesis, particularly in the Paigen diet–fed ApoE−/−/IL-1R1−/− mice, we performed a series of serum cytokine cytometric bead assays and enzyme-linked immunosorbent assays. Analysis of inflammatory mediators in these mice revealed that feeding of the Paigen diet was associated with significantly higher levels of IL-1β (Figure 5A) and IL-6 (Figure 5B) in both ApoE−/− and ApoE−/−/IL-1R1−/− mice. Although there was a trend for increased RANTES in the Paigen-fed ApoE−/− mice this was not significant (Figure 5C). Interestingly, this response was heightened significantly in the Paigen diet–fed ApoE−/−/IL-1R1−/− mice compared with ApoE−/− mice for IL-6 (Figure 5B) and RANTES (Figure 5C), suggesting that the Paigen diet–fed ApoE−/−/IL-1R1−/− mice show an exaggerated inflammatory phenotype despite lacking IL-1R1.
Figure 5Serum cytokine levels are increased in Paigen diet–fed ApoE−/−/IL-1R1−/− mice. Bar graphs show serum levels of IL-1β (A), IL-6 (B), and RANTES (C). Bars represent mean ± SEM, n = 4 to 7. *P < 0.05, **P < 0.01 compared with chow-fed littermates. Statistical differences between Paigen diet–fed mice are shown as assessed by analysis of variance followed by Bonferroni post hoc analysis.
Putative Biomarkers of PAH Are Increased in Paigen-Fed ApoE−/− and ApoE−/−/IL-1R1−/− Mice
We previously reported that human pulmonary vascular lesions from patients with idiopathic PAH are associated with medial immunoreactivity for both TRAIL and OPG.
We investigated expression of these molecules in this mouse model of severe PAH. Little OPG expression was observed in chow-fed mice. Diffuse medial staining and strong cellular immunoreactivity for OPG was observed within lesions from both Paigen diet–fed ApoE−/− and ApoE−/−/IL-1R1−/− mice (Figure 6A). Weak TRAIL immunoreactivity was observed within the epithelial and endothelial cells of chow-fed ApoE−/− and ApoE−/−/IL-1R1−/− mouse lungs (Figure 6A). There were increased TRAIL-positive cells within the media and perivascular tissue of remodeled arteries from both the Paigen diet–fed ApoE−/− and ApoE−/−/IL-1R1−/− mice (Figure 6A). This appeared to be de novo expression because TaqMan PCR on whole-lung extracts from these mice showed TRAIL expression to be increased significantly in both Paigen diet–fed ApoE−/− and ApoE−/−/IL-1R1−/− mice (Figure 6B). OPG expression also was increased significantly in both Paigen diet–fed ApoE−/− and ApoE−/−/IL-1R1−/− mice, and the level of gene expression in Paigen diet–fed ApoE−/−/IL-1R1−/− mice was significantly higher than the ApoE−/− mice (Figure 6C). Because OPG is a naturally secreted molecule and we previously reported increased serum OPG in patients with Idiopathic (IPAH),
we next performed an OPG enzyme-linked immunosorbent assay on the mouse serum. OPG levels were significantly higher in both Paigen diet–fed ApoE−/− and ApoE−/−/IL-1R1−/− mice compared with chow-fed littermates. The circulating levels of OPG in the Paigen diet–fed ApoE−/−/IL-1R1−/− mice were also significantly higher than the ApoE−/− mice (Figure 6D). These data confirm our initial studies in human PAH and provide further evidence for a role of both molecules in disease pathogenesis, and for OPG as a potential biomarker of disease. BMP-R2 mutations are clearly important in human forms of IPAH
Altered growth responses of pulmonary artery smooth muscle cells from patients with primary pulmonary hypertension to transforming growth factor-beta(1) and bone morphogenetic proteins.
Altered bone morphogenetic protein and transforming growth factor-signaling in rat models of pulmonary hypertension: potential for activin receptor-like kinase-5 inhibition in prevention and progression of disease.
Because we previously reported that levels of OPG and TRAIL can be regulated by impaired or reduced BMP-R2 expression as well as increased IL-1 signaling,
we next sought to determine whether the increased expression of these proteins was associated with reduced expression of BMP-R2 in this model. BMP-R2 levels were reduced significantly in both Paigen diet–fed ApoE−/− and ApoE−/−/IL-1R1−/− mice compared with chow-fed littermates (Figure 6E).
Figure 6OPG and TRAIL increase with PAH. A: Representative sections of lung immunostained for OPG and TRAIL. TaqMan expression of OPG (B), and TRAIL in whole lung normalized using ΔΔCT with 18S rRNA as the endogenous control gene (C). D: Serum expression of OPG. E: TaqMan expression of BMP-R2 in whole lung normalized using ΔΔCT with 18S rRNA as the endogenous control gene. Bars represent mean ± SEM, n = 4 to 7. *P < 0.05, **P < 0.01, ***P < 0.001 compared with chow-fed littermates. Statistical differences between Paigen diet–fed mice are shown as assessed by analysis of variance followed by Bonferroni post hoc analysis. Original magnification, ×400. Scale bar = 50 μm.
Inflammatory Mechanisms Drive PAH Phenotype in Fat-Fed ApoE−/−/IL-1R1−/− Mice
Interestingly, the Paigen diet–fed ApoE−/−/IL-1R1−/− compared with Paigen diet–fed ApoE−/− mice showed increased levels of IL-1β as well as increased expression of molecules downstream of IL-1 signaling such as IL-6, RANTES (Figure 5, B and C), and OPG (Figure 6C). TaqMan PCR for IL-1R1 mRNA in the lungs of both the ApoE−/− and ApoE−/−/IL-1R1−/− mice revealed significant expression of IL-1R1 mRNA in RNA isolated from the lungs of all mice (Figure 7A). This led us to speculate that an IL-1R1 mRNA and a putative protein may be produced in the lungs of ApoE−/−/IL-1R1−/− mice. Further investigation revealed that the IL-1R1 probe in the TaqMan assay annealed across the exon 6/7 boundary, and that the IL-1R1−/− mouse within our experiments was generated by removing the sequence for amino acids 4 to 146, equating to exons 1 and 2.
We confirmed that these exons were indeed missing from the ApoE−/−/IL-1R1−/− mice using an alternative TaqMan probe set (Figure 7B). It is known that there are multiple promoter sites in the human
IL-1R1 gene that can regulate tissue-specific expression. We were able to show lung-specific expression of this IL-1R1 form using a TaqMan assay for exon 6/7 and expression of IL-1R1 was evident in the lungs but not aorta of ApoE−/−/IL-1R1−/− mice (Figure 7C). This indicates a degree of tissue specificity for the alternative IL-1R1 transcript. As expected, mRNA for exon 2/3 was not found in ApoE−/−/IL-1R1−/− lung or aorta (Figure 7D). We next sought to determine whether this putative mRNA transcript present within the lungs of these mice resulted in protein expression. Western immunoblotting of whole-lung lysates from both ApoE−/− and ApoE−/−/IL-1R1−/− mice using an anti–IL-1R1 antibody revealed immunoreactivity in both strains of mice, indicating the presence of translated protein. These data show that in this model, lung-specific expression of a putative alternate IL-1R1 receptor form may contribute to the severe PAH phenotype observed in these mice.
Figure 7Expression of an alternative IL-1R1 transcript in ApoE−/−/IL-1R1−/− lungs. Gene expression of IL-1R1 using TaqMan primer/probe sets that cross exons 6/7 (A), and exons 2/3 in whole-lung RNA samples (B). Comparison of lung and aorta expression of IL-1R1 using TaqMan primer/probe sets that cross exons 6/7 (C), and exons 2/3 (D). E: Western immunoblot for IL-1R1 in whole-lung lysates isolated from three ApoE−/− and three ApoE−/−/IL-1R1−/− mice. All data are normalized using ΔΔCT with 18S rRNA as the endogenous control gene. Bars represent mean ± SEM, n = 4 to 7. Statistical differences between groups are shown as assessed by analysis of variance followed by Bonferroni post hoc analysis. *P < 0.05 and ***P < 0.001.
IL-1ra (Anakinra) Prevents Progression of PAH in Paigen Diet–Fed ApoE−/− and ApoE−/−/IL-1R1−/− Mice
To further investigate IL-1–driven inflammatory mechanisms in both Paigen diet–fed models, and to determine whether the alternative IL-1R1 transcript may be functional, we implanted Alzet mini osmotic pumps to deliver either placebo or IL-1ra (anakinra) at 10 μg/h for 4 weeks. Treatment was started after an initial 4 weeks of feeding to allow the disease process to become established. Paigen diet–fed ApoE−/− and ApoE−/−/IL-1R1−/− mice treated with IL-1ra showed no hemodynamic evidence of PAH, with RVSP significantly lower than their placebo-treated littermates and not significantly higher than chow-fed mice (Figure 8A), and also showed a significant reduction in media/CSA of the small pulmonary arteries (<50 μm) to near-normal levels (Figure 8B). IL-1ra–treated Paigen-fed ApoE−/−/IL-1R1−/− mice also showed a significantly reduced RVH (Figure 8C). We next looked to see whether the increased levels of IL-1β, IL-6, RANTES, and OPG similarly were reduced by treatment with IL-1ra. We found no significant increase in serum levels of IL-1β (Figure 8D), IL-6 (Figure 8E), RANTES (Figure 8F), or OPG (Figure 8G) in the IL-1Ra-treated, Paigen diet-fed mice compared to chow-fed controls. There was also a significant reduction in OPG gene expression in whole-lung lysates from the IL-1ra–treated mice compared with placebo-treated littermates (Figure 8H). These data suggest that IL-1ra effectively blocks the progression of PAH in this model and that this is associated with reduced levels of key drivers of PAH pathogenesis, particularly IL-6 and OPG.
Figure 8Treatment with IL-1ra reversed PAH. Bar graphs show RVSP measured in mm Hg (A), the degree of medial wall thickness as a ratio of total vessel size (media/CSA) (B), RVH (C), and serum levels of IL-1β (D), IL-6 (E), RANTES (F), OPG (G), and TaqMan expression of OPG in whole lung normalized using ΔΔCT with 18S rRNA as the endogenous control gene (H). Bars represent mean ± SEM, n = 4 to 7. Statistical differences between groups are shown as assessed by analysis of variance followed by Bonferroni post hoc analysis. *P < 0.05, **P < 0.01, and ***P < 0.001.
We show that by feeding the more aggressive atherogenic Paigen diet, ApoE−/− mice develop severe PAH. A continuing challenge for research into the mechanisms important in the pathogenesis of PAH has been the absence of a murine, preclinical model that recapitulates accurately all of the clinical and pathologic features of human pulmonary hypertension (PH).
To our knowledge, the severity of PH in our mouse model is greater both hemodynamically and with regard to pulmonary vascular remodeling than previously published murine models, including (but not exclusively) schistosomiasis models,
Activin-like kinase 5 (ALK5) mediates abnormal proliferation of vascular smooth muscle cells from patients with familial pulmonary arterial hypertension and is involved in the progression of experimental pulmonary arterial hypertension induced by monocrotaline.
Of particular note in this model is that ApoE−/−/IL-1R1−/− mice have dissociation between the two inflammatory vascular phenotypes studied. In these animals, systemic atherosclerosis is 40% less at the aortic sinus than ApoE−/− counterparts,
and there is a 50% increase in RVSP compared with ApoE−/−. We show that a potential explanation for this unexpected effect is a putative, alternative IL-1R1 transcript expressed in the lungs but not aorta of ApoE−/−/IL-1R1−/− mice. In our model, this alternative IL-1R1 transcript appears to have the capacity to drive the pulmonary vascular phenotype. We have shown, at least in part, functionality of this receptor in these mice by blocking the progression of disease through the administration of IL-1ra (anakinra). These data provide further evidence that IL-1 is important in the pathogenesis of PAH and suggest that targeting IL-1 may have therapeutic potential in the setting of PAH.
The implication that IL-1 signaling may play an important role in PAH pathogenesis has been highlighted previously. Biomarker studies have shown increased serum levels of many inflammatory cytokines, including IL-1 and the downstream cytokine IL-6.
Treatment of rats with the naturally occurring IL-1ra has been shown to protect against development of monocrotaline-induced PAH, but interestingly not hypoxia,
The exaggerated pulmonary hypertension, despite reduced atherosclerosis, indicates that some lung-specific IL-1R1 signaling remains in our ApoE−/−/IL-1R1−/− mouse. IL-1R1−/− transcript analysis suggests that tissue specificity arises from alternative expression of these transcripts, and that an artificial alternative transcript is still expressed that actively signals IL-1, and is inhibited by IL-1ra. These data suggest that there are unique aspects of IL-1 signaling in the lung and we note that a recent report on a genome-wide association study of asthma implicated a single-nucleotide polymorphism affecting the IL-1R1 gene.
These data therefore raise significant opportunities for IL-1 blockade in PAH and other lung diseases, and also caution the interpretation of lung phenotype data that are generated using the current IL-1R1−/− mouse.
In keeping with published data from studies performed in end-stage human IPAH, we also observed an increase in gene expression and/or serum levels of proteins known to be activated by IL-1, and proposed to be important in the pathogenesis of PAH including IL-6,
Indeed, the co-localization of OPG and TRAIL expression to the remodeled pulmonary arteries in this new mouse model was almost identical to our findings in human PAH.
Further studies currently are underway to fully determine the role that these molecules, particularly OPG and TRAIL, play in disease pathogenesis of this and other established models of pulmonary hypertension.
Very recent reports have suggested that the ingredient cholate in the Paigen diet formulation is responsible for the formation of lung granulomas in a proposed mouse model of sarcoidosis
Although we observed some early granulomas within the lungs of our mice, we detected no difference in these across our mouse genotypes, suggesting that the two lung pathologies are not intrinsically linked, at least at the earlier time points used for this study (8 weeks on diet compared with 16 weeks).
In this current study we have shown that treatment with IL-1ra as disease is developing prevents further disease progression. Although the ability of IL-1ra treatment to fully reverse human PAH is difficult to fully ascertain in animal models, prevention or slowing disease progression in patients would in itself be highly desirable. IL-1ra (anakinra/Kineret, Amgen, Thousand Oaks, CA) currently is used for the treatment of rheumatoid arthritis and is being studied for its efficacy in a number of clinical trials ranging from diabetes
VCU-ART investigators: interleukin-1 blockade with anakinra to prevent adverse cardiac remodeling after acute myocardial infarction (Virginia Commonwealth University Anakinra Remodeling Trial [VCU-ART] pilot study).
in Western-diet fed ApoE−/− mice. This relatively mild RVH appears to be out of proportion to the degree of pulmonary vascular remodeling and increased in right heart pressures. This is particularly surprising given that we can routinely measure RVH in the hypoxic mouse model that generates substantially lower RVSP. We have investigated this phenomenon intensively through the use of invasive pressure-volume catheterization and noninvasively with dedicated small animal ultrasound and magnetic resonance imaging to further verify this finding. All of the data obtained points to a right ventricle that, although being slightly enlarged, is coping well with the high demands made on it by the increased pulmonary vascular resistance. We hypothesize that this may be owing in part to the altered metabolic condition of these mice owing to the Paigen diet and this is something that we will continue to investigate. This phenomenon is not unique to this study, other investigators have reported increased right heart pressures in transgenic mice without the occurrence of RVH, particularly without exposure to hypoxia.
Studies examining the effect of norepinephrine infusion in congestive heart failure models also have highlighted significant increases in both RVSP and RV dP/dtmax with no detectable increase in RVH.
Interestingly, in this study Barth et al do report a significant increase in left ventricular hypertrophy despite no increase in left ventricular pressure and implicate IL-1/IL-6–mediated collagen I and III in this process. Further evidence for the importance of IL-1 in cardiac remodeling comes from murine studies describing spontaneous left ventricular hypertrophy leading to heart failure in transgenic mice after cardiac-specific overexpression of IL-1 α.
These studies highlight the complexity of cardiac remodeling and perhaps highlight other key drivers for hypertrophy other than hemodynamic changes alone. Whether differential IL-1 signaling in the two ventricles is responsible for this certainly warrants further investigation.
The accessibility of ApoE−/− mice and atherosclerotic diets, combined with the high degree of pulmonary vascular remodeling, make this a practical and alternative model for many laboratories interested in studying the role of a particular gene or drug on the pathogenesis of PAH. Furthermore, despite the severe hemodynamic and histopathologic phenotype, this models displays favorable right ventricular adaptation, and thus also may serve as an additional investigative tool to study the right ventricle. These findings highlight the potential role of IL-1 signaling in PAH and raise the possibility and importance of alternative signaling in the lung. Targeting IL-1, which is upstream of most vascular inflammatory pathways and has proven systemic anti-inflammatory effects, may be a viable alternative treatment strategy to slow disease progression in PAH, particularly when associated with other inflammatory disorders.
Acknowledgments
We thank Profs. Marlene Rabinovitch and Nicholas Morrell for helpful discussions throughout this work. We also acknowledge technical assistance from Mrs. Tatiana Vinogradova.
References
D'Alonzo G.
Barst R.
Ayres S.
Bergofsky E.
Brundage B.
Detre K.
Fishman A.
Goldring R.
Groves B.
Kernis J.
Levy P.
Pietra G.
Reid L.
Reeves J.
Rich S.
Vreim C.
Williams G.
Wu M.
Survival in patients with primary pulmonary hypertension.
American College of Cardiology Foundation Task Force on Expert Consensus Documents, American Heart Association, American College of Chest Physicians, American Thoracic Society Inc, Pulmonary Hypertension Association.
Right ventricular function and failure: report of a National Heart, Lung, and Blood Institute Working Group on cellular and molecular mechanisms of right heart failure.
Altered growth responses of pulmonary artery smooth muscle cells from patients with primary pulmonary hypertension to transforming growth factor-beta(1) and bone morphogenetic proteins.
Altered bone morphogenetic protein and transforming growth factor-signaling in rat models of pulmonary hypertension: potential for activin receptor-like kinase-5 inhibition in prevention and progression of disease.
Activin-like kinase 5 (ALK5) mediates abnormal proliferation of vascular smooth muscle cells from patients with familial pulmonary arterial hypertension and is involved in the progression of experimental pulmonary arterial hypertension induced by monocrotaline.
VCU-ART investigators: interleukin-1 blockade with anakinra to prevent adverse cardiac remodeling after acute myocardial infarction (Virginia Commonwealth University Anakinra Remodeling Trial [VCU-ART] pilot study).
Supported by a Medical Research Council Career Development Award (G0800318 to A.L.); a British Heart Foundation Clinical Research Training Fellowship (FS/08/061/25740 to A.G.H.); the National Institute for Health Research Sheffield Cardiovascular Biomedical Research Unit (N.A.); and the National Institute for Health Research via its Biomedical Research Units funding scheme (N.A. and D.C.).
Human IL-1Ra and placebo were obtained under MTA 200517250-001 from Amgen, Inc. (Thousand Oaks, CA).
Current address of D.C.C., Norwich Medical School, University of East Anglia, Norwich, UK.
P < 0.05 chow versus Paigen.
00659-6&id=gr1.jpg){kind=link}
00659-6&id=gr2.jpg){kind=link}
00659-6&id=gr3.jpg){kind=link}
00659-6&id=gr4.jpg){kind=link}
00659-6&id=gr5.jpg){kind=link}
00659-6&id=gr6.jpg){kind=link}
00659-6&id=gr7.jpg){kind=link}
00659-6&id=gr8.jpg){kind=link}