Blocking macrophage leukotriene b4 prevents endothelial injury and reverses pulmonary hypertension

Abstract

Pulmonary hypertension (PH) is a serious condition that affects mainly young and middle-aged women, and its etiology is poorly understood. A prominent pathological feature of PH is accumulation of macrophages near the arterioles of the lung. In both clinical tissue and the SU5416 (SU)/athymic rat model of severe PH, we found that the accumulated macrophages expressed high levels of leukotriene A4 hydrolase (LTA4H), the biosynthetic enzyme for leukotriene B4 (LTB4). Moreover, macrophage-derived LTB4 directly induced apoptosis in pulmonary artery endothelial cells (PAECs). Further, LTB4 induced proliferation and hypertrophy of human pulmonary artery smooth muscle cells. We found that LTB4 acted through its receptor, BLT1, to induce PAEC apoptosis by inhibiting the protective endothelial sphingosine kinase 1 (Sphk1)-endothelial nitric oxide synthase (eNOS) pathway. Blocking LTA4H decreased in vivo LTB4 levels, prevented PAEC apoptosis, restored Sphk1-eNOS signaling, and reversed fulminant PH in the SU/athymic rat model of PH. Antagonizing BLT1 similarly reversed established PH. Inhibition of LTB4 biosynthesis or signal transduction in SU-treated athymic rats with established disease also improved cardiac function and reopened obstructed arterioles; this approach was also effective in the monocrotaline model of severe PH. Human plexiform lesions, one hallmark of PH, showed increased numbers of macrophages, which expressed LTA4H, and patients with connective tissue disease-associated pulmonary arterial hypertension exhibited significantly higher LTB4 concentrations in the systemic circulation than did healthy subjects. These results uncover a possible role for macrophage-derived LTB4 in PH pathogenesis and identify a pathway that may be amenable to therapeutic targeting.

Figures

Fig. 1. Increased macrophage LTB4 biosynthesis during evolution of experimental PH

(A) Summary of the LT pathways. (B and C) LTB4 and LTC4 concentrations in (B) BALF and (C) serum from dimethyl sulfoxide (DMSO)–treated (vehicle control) and SU-treated (PAH: 3 weeks after SU administration) animals with experimental PH as measured by liquid chromatography–tandem mass spectrometry (LC-MS/MS). (D) BLT1 (major LTB4 receptor) and CysLT1 (major CysLT receptor) mRNA levels measured by reverse transcription polymerase chain reaction (RT-PCR). (E to H) Representative immunofluorescence images from lung sections stained with p5-LO (green) and CD68 (red) from (E) DMSO, (F) 1week after SU, (G) 2 weeks after SU, and (H) 3 weeks after SU. 4′,6-Diamidino-2-phenylindole (DAPI) (blue) identifies nuclei. Differential interference contrast (DIC) highlights alveolar and vascular structures. Green arrows point to the center of occluded arterioles. Merged panels show all three stains. (I) Quantitation of the data in (E) to (H) (expressed as p5-LO+ macrophages per high-power field). n = 6 per group. Scale bar, 50 µm. Kruskal-Wallis test followed by Dunn’s multiple comparisons test for post hoc analyses was used. Data are means ± SEM. NS, not significant.

Fig. 2. Increased macrophage LTA4H over time in developing PH

(A to D) Representative immunofluorescence images from lung sections stained with antibody to LTA4H (green) and CD68 (red) from (A) DMSO, (B) 1 week after SU, (C) 2 weeks after SU, and (D) 3 weeks after SU. DAPI (blue) identifies nuclei. Green arrows point to the center of occluded arterioles. Merged panel shows all three stains. (E)Quantitation of LTA4H staining over time in PH in pulmonary macrophages. LTA4H+, CD68+ macrophages were counted per high-power field. (F) LTA4H mRNA, as measured by RT-PCR. (G) p5-LO+ or LTA4H+ macrophages were counted and grouped as <50 µm, 50 to 100 µm, or 100 to 150 µm from the center of the small pulmonary arterioles. n = 6 per group. Scale bar, 50 µm. Kruskal-Wallis test followed by Dunn’s multiple comparisons test for post hoc analyses was used. Data are means ± SEM.

Fig. 3. LTB4 induction of PAEC apoptosis and promotion of human PASMC (hPASMC) growth and hypertrophy

(A and B) Interstitial macrophages from (A) DMSO and (B) SU rat lungs were cocultured with PAECs for 24 hours. PAEC apoptosis was analyzed by flow cytometry for annexin V and by confocal microscopy for cleaved caspase-3 staining (green). (C and D) Effect of LTB4 on endothelial injury. PAECs were treated with (C) exogenous LTB4 (200 nM) or (D) LTB4 (200 nM) and U75302 (1 mM). (E) S1P concentrations in culture medium from different cells and treatments, as measured by LC-MS/MS. (F) Effect of LTB4 on smooth muscle cells. Various concentrations of LTB4 were added to hPASMC culture with or without U75302 (1 mM) for 24 hours. hPASMC proliferation was measured by MTT assay. Cellular hypertrophy was determined by protein/DNA ratio. n = 3 per group. Scale bar, 50 µm. Representative flow cytometry plots and confocal images are shown. (F) LTB4-induced hPASMC growth as measured by Western blot of PCNA (proliferating cell nuclear antigen). n = 3 per group. Kruskal-Wallis test followed by Dunn’s multiple comparisons test for post hoc analyses was used. Data are means ± SEM. NS, not significant.

Fig. 4. LTB4 induction of PAEC apoptosis through the inhibition of endothelial Sphk1-eNOS pathway

(A to F) PAECs cultured with or without macrophages were examined for pSer225 Sphk1, total Sphk1, pSer1177 eNOS, and total eNOS expression by Western blot. Because of differences in antibody specificity and sensitivity, blots were developed at various exposure times to best represent the signal of interest. Densitometry of the Western blots is summarized in fig. S8. Western blots of PAECs cultured with (A) macrophages isolated from DMSO and SU-PH athymic rat lungs, (C) transfected macrophages treated with or without LTB4 (200 nM) or S1P (1 µM), and (E) exogenous LTB4 (200 nM) with or without U75302 (1 µM) and with or without S1P (1 µM). NO production from culture medium in (A), (C), and (E) is summarized in (B), (D), and (F), respectively. NO levels are expressed relative to background fluorescence. n = 3 experiments per group. Data are means ± SEM. NS, not significant.

Fig. 5. The reversal of established PH by blockade of LTB4 signaling

Rats were treated with montelukast (MON), bestatin (BE), JNJ-26993135 (JNJ), or LY293111 (LY) starting 3 weeks after SU administration. Animals were monitored by echocardiography weekly and sacrificed for hemodynamic measurements at week 5. (A) Right ventricular systolic pressure (RVSP) in DMSO, SU, and four different treatment protocols at week 5 after SU treatment (doses of the drugs used: montelukast, 10 mg/kg, orally, daily; bestatin, 1 mg/kg, orally, daily; JNJ-26993135, 30 mg/kg, orally, twice daily; LY293111, 0.5 mg/kg, orally, daily). RV hypertrophy (RVH) was assessed by the RV/[left ventricle (LV) + septum (S)] weight ratios. (B) Kaplan-Meier plot of survival of the groups of rats treated as in (A). The bestatin-, JNJ-26993135–, and LY293111-treated groups overlap with the DMSO group. (C) Representative immunohistochemistry images of pulmonary arterioles stained for α-smooth muscle actin (SMA) in lung tissues at week 5. Black arrows, occluded arterioles. (D) Medial wall thickness of α-SMA–positive small vessels (<100 µm in external diameter) from rats described in (A). Data are expressed as % total occlusion. (E) Representative immunofluorescence images of pulmonary arterioles stained for α-SMA (red) and von Willebrand factor (vWF) (green) in lung tissues at week 5 from rats described in (A). DAPI (blue), nuclei. (F) Proportion of nonmuscularized (N), partially muscularized (P), or fully muscularized (M) pulmonary arteries, as a percentage of the total pulmonary arteriolar cross-sectional area (sized <100 µm) from rats described in (A). (G) LTB4, LTC4, and LXA4 in BALF as measured by LC-MS/MS in rats described in (A). SU data were pooled from contemporaneous control data (n = 15) as comparator because only one of the SU (PH) rat survived to 5 weeks in this experiment, as reflected in survival curve in (B). n = 6 per group. Two-way analysis of variance (ANOVA) with Bonferroni multiple comparisons test for post hoc analyses was used. Data are means ± SEM. NS, not significant. ‡, LXA4 levels in the experiment groups were below the level of detection.

Fig. 6. Components of the eicosanoid pathway in experimental PH

(A) Eicosanoid synthesis pathways, with the site of action of various inhibitors indicated: PLA2 inhibitor, arachidonyl trifluoromethyl ketone (ATK); 5-LO inhibitor, Zileuton; FLAP inhibitor, MK886; COX-1 and COX-2 inhibitor, aspirin; EET inhibitor, DCU. (B) RV parameters in rats treated with the inhibitors described in (A) 3 weeks after SU administration. See table S2 for complete dosing regimens. Animals were monitored by echocardiography weekly and sacrificed for hemodynamic measurement at week 5. RVSP and RV/(LV + S) measurements after treatments with DMSO, SU, bestatin, and five inhibitors were assessed at week 5 of SU administration. (C) Kaplan-Meier curve showing survival of rats after treatment. n = 6 per group. Bestatin treatment group overlaps with the DMSO, Zileuton, and MK886 groups. Data are means ± SEM. ‡, all animals were dead by the day of terminal cardiac catheterization.

Fig. 7. LTA4H expression and serum LTB4 concentrations in PAH patients

(A to D) Confocal images of human lung tissues stained for CD68 (green) and LTA4H (red). (A) Healthy control. (B) Plexiform lesion from lung of patient with iPAH. (C) Plexiform lesion from patient with PAH associated with systemic sclerosis. (Inset) LTA4H+CD68+ macrophages in the alveolar space. (D) Plexiform lesion from systemic sclerosis–associated PAH lung. DAPI (blue), nuclei; red, LTA4H+CD68+ macrophages; white arrows, occluded arterioles. Scale bars, 50 µm. (E and F) Serum LTB4 concentrations in healthy controls (n = 6) and treatment-naïve PAH patients (n = 19). Gray shaded area in chart indicates CTD-APAH patients. (Right) Data from chart presented as concentrations (points) and means ± SEM in figure (lines and whiskers). P value reflects nonparametric one-way ANOVA with the Kruskal-Wallis test. H, Hispanic; C, Caucasian; A, Asian; AA, African American. Diagnosis: iPAH, idiopathic PAH; MCTDz-APAH, mixed connective tissue disease–associated PAH; SLE-APAH, systemic lupus erythematosus–associated PAH; Sscl-APAH, systemic sclerosis–associated PAH; NYHA, New York Heart Association functional class. WHO, World Health Organization; NTproBNP, N-terminal pro-B–type natriuretic peptide.

Fig. 8. Model illustrating how macrophage-derived LTB4 may induce vascular remodeling and contribute to PH

(A) Endothelial injury causes a local immune response. (B) With immune dysregulation because of a lack of normal Treg activity, there is an exuberant inflammatory response that includes perivascular inflammation with mast cells, B cells, anti–endothelial cell antibodies, and macrophages. In macrophages, 5-LO is phosphorylated by p38 MAPK in the nucleus. p5-LO converts AA to LTA4, which is subsequently catalyzed by LTA4H to produce LTB4. LTB4 is secreted by macrophages, binds to BLT1 on PAECs, and inhibits the Sphk1-eNOS survival signal in the PAEC. LTB4 signaling through BLT1 also promotes PASMC proliferation and hypertrophy. (C) This process causes accelerated endothelial cell apoptosis and smooth muscle cell layer growth. (D) Ongoing endothelial apoptosis and smooth muscle growth can result in luminal occlusion and PH.