Resolution of sickle cell disease-associated inflammation and tissue damage with 17 R-resolvin D1

Abstract

Resolvins (Rvs), endogenous lipid mediators, play a key role in the resolution of inflammation. Sickle cell disease (SCD), a genetic disorder of hemoglobin, is characterized by inflammatory and vaso-occlusive pathologies. We document altered proresolving events following hypoxia/reperfusion in humanized SCD mice. We demonstrate novel protective actions of 17R-resolvin D1 (17R-RvD1; 7S, 8R, 17R-trihydroxy-4Z, 9E, 11E, 13Z, 15E, 19Z-docosahexaenoic acid) in reducing ex vivo human SCD blood leukocyte recruitment by microvascular endothelial cells and in vivo neutrophil adhesion and transmigration. In SCD mice exposed to hypoxia/reoxygenation, oral administration of 17R -RvD1 reduces systemic/local inflammation and vascular dysfunction in lung and kidney. The mechanism of action of 17R-RvD1 involves (1) enhancement of SCD erythrocytes and polymorphonuclear leukocyte efferocytosis, (2) blunting of NF-κB activation, and (3) a reduction in inflammatory cytokines, vascular activation markers, and E-selectin expression. Thus, 17R-RvD1 might represent a new therapeutic strategy for the inflammatory vasculopathy of SCD.

Figure 1.

Lipidometabolomic signature of AA and SS mouse spleens. (A) Representative multiple reaction monitoring chromatograms, from spleen samples of sickle cell mice exposed to hypoxia (8% oxygen; 10 hours) and followed by reoxygenation (21% oxygen; 3 hours), used to identify LMs. (B) MS/MS fragmentation spectra used for identification of RvD1, LXA₄, and PGE_2.

Figure 2.

PCA of LMs, specialized proresolving mediator profiles, in healthy and sickle cell mice exposed to H/R stress. (A) Three-dimensional score plot of sickle cell mice under normoxia (red; n = 3) or exposed to 10 hours of hypoxia (8% oxygen) and followed by 18 hours of reoxygenation (yellow; n = 3) for murine spleen samples. Ellipses mark 95% confidence regions. (B) Three-dimensional loading plot. (C) Quantitative pathway network of sickle cell murine spleen samples. Node size represents the mean values (in picograms) of sickle cell normoxia spleen samples (n = 3). Node color denotes the fold changes in sickle cells exposed to hypoxia (10 hours 8%), followed by 18 hours of reoxygenation, vs sickle cell normoxia.

Figure 3.

RvD1 reduces ex vivo human neutrophil adhesion and in vivo neutrophil counts in humanized SCD mice, which show decreases in plasma RvD1 values after DHA administration. (A) Kinetics of DHA conversion to proresolving mediator RvD1 following oral administration in AA and SS mice. Levels of RvD1 were determined, using a competitive enzyme immunoassay, in plasma collected from AA and SS mice at the indicated times following DHA gavage. Data are mean ± SD (n = 3). *P < .05 vs baseline for AA mice. (B) Adhesion of neutrophils (green) to TNF-α–activated human microvascular endothelial cell line (HMEC). Blood samples from a healthy donor (AA) and an SCD patient (SS) were incubated for 10 minutes with vehicle or 17R-RvD1 (1 μM) (left panels). A second dose of vehicle or 17R-RvD1 (1 μM) was added before perfusing blood for 40 minutes (at 1 dyn/cm²) through the biochip channels containing activated HMEC-1 monolayers. Percentage of neutrophil adhesion to TNF-α–activated HMECs from AA (n = 6) and SS (n = 5) blood samples incubated with vehicle or 17R-RvD1 (right panel). ***P < .001 vs the corresponding vehicle group. (C) Representative images showing reduced neutrophil recruitment after 17R-RvD1 administration in sickle cell mice (left panels). All of these experiments suggest that 17R-RvD1 influences neutrophil recruitment in SS mice, especially in the context of sustained and intense neutrophil recruitment specific to experimental SCD. Neutrophil adhesion density, defined as the number of adherent neutrophils per square millimeter of endothelial surface, in TNF-α–inflamed venules after 17R-RvD1 treatment in SS mice (SS TNF-α 17R-RvD1) or after dual endothelin receptor antagonism (SS TNF-α bosentan) (upper right panel). Data are mean ± standard error of the mean. P < .001 for SS mice treated with TNF-α and 17-RvD1 vs with TNF-α and vehicle alone (SS TNF-α 17R-RvD1), 2-way analysis of variance (ANOVA). *P < .05, **P < .01 vs vehicle-treated SS with TNF-α, 2-way ANOVA followed by the Tukey multiple-comparison test. Extravascular volume in inflamed venules after 17R-RvD1 or bosentan vs vehicle administration in SS mice (lower right panel). Emigrated neutrophils were visualized and quantified by optical sectioning and 2-dimensional maximum intensity projection. Twenty venules (5 mice) were analyzed in each group. Data are mean ± standard error of the mean. P < .001 for SS mice treated with TNF-α and 17R-RvD1 vs with TNF-α and vehicle alone, 2-way ANOVA. Data for heterozygous AS littermates are provided as control reference for neutrophil adhesion and emigration to tissue after TNF-α. *P < .05, **P < .01 vs vehicle-treated SS with TNF-α, 2-way ANOVA, followed by the Tukey multiple-comparison test.

Figure 4.

17R-RvD1 enhances phagocytosis of sickle RBCs and PMNs by spleen macrophages. In vitro phagocytosis of carboxyfluorescein diacetate succinimidyl ester–labeled aged RBCs (at 4°C, 18 h; A) or PMNs (37°C, 18 h; B) from AA and SS mice by spleen MΦs (F4/80⁺ cells), as assessed using flow cytometry. Data are mean ± standard deviation from 6 independent experiments. (C-D) Phagocytosis of RBCs and PMNs by spleen MΦs in vivo in vehicle- or 17R-RvD1–treated AA and SS mice undergoing H/R (left panels). Percentages of MΦs engulfing RBCs (F4/80⁺ Ter-119⁺; C) or PMNs (F4/80⁺ Ly6G⁺; D) were calculated using flow cytometry. Results are mean ± standard deviation from 3 mice per group. Representative flow cytometry dot plots and gate criteria (right panels). ^§P < .05 vs AA MΦs + vehicle, *P < .05 vs vehicle-treated cells. MFI, mean fluorescence intensity.

Figure 5.

17R-RvD1 prevents H/R activation of acute inflammatory response pathways and reduces SCD vascular vulnerability through multimodal action on NF-κB activation. (A) BAL protein content (upper panel) and leukocyte content (lower panel) from AA and SS mice under normoxia and treated with vehicle or 17R-RvD1 (100 ng) and exposed to H/R: hypoxia (8% oxygen; 10 hours), followed by reoxygenation (21% oxygen; 3 hours) (upper panel). All data are mean ± SD (n = 6). (B) Immunoblot analysis, using specific antibodies against phosphorylated (P-)NF-κB, NF-κB, P-Nrf2, and Nrf2, in lung from AA and SS mice treated as in (A) (left panel). Vertical line(s) in NF-κB, P-p65 gel have been inserted to indicate a repositioned gel lane. One representative gel from 6 gels with similar results is presented. Densitometric analysis of immunoblots (right panels). Data are mean ± SD (n = 6 in each group). (C) Expression of miR-126 (mmu–miR-126-5p), as determined using quantitative polymerase chain reaction, in the lungs of AA and SS mice undergoing H/R and 17R-RvD1 treatment. Results are mean ± SD from 3 to 6 mice per group. (D) Immunoblot analysis, using specific antibodies against HO-1, IL-6, ET-1, ICAM-1, PDGF-B, and TXAS-1, of lung from AA and SS mice treated as in (B). One representative gel from 6 gels with similar results is shown. Densitometric analysis immunoblots are shown in supplemental Figure 2A. *P < .05 vs normoxia, °P < .05 vs healthy mice (AA), ^P < .05 vs vehicle.

Figure 6.

17R-RvD1 protects kidney from sickle cell–related acute injury, prevents inflammatory vascular activation, and positively modulates antifibrotic let7c miRNA. (A) Plasma creatinine (upper panel) and blood urea nitrogen (BUN) (lower panel) levels in AA and SS mice under normoxic conditions or treated with vehicle or 17R-RvD1 (100 ng) and exposed to H/R: hypoxia (8% oxygen; 10 hours), followed by reoxygenation (21% oxygen; 3 hours). Data are mean ± SD (n = 6). *P < .05 vs normoxia, °P < .05 vs healthy mice (AA), ^P < .05 vs vehicle. Hematoxylin and eosin–stained sections of kidney tissue from AA and SS mice treated with vehicle [AA (B); SS (D-E)] or 100 ng of 17R-RvD1 [AA (C); SS (F)] exposed to H/R: hypoxia (8% oxygen; 10 hours) followed by reoxygenation (21% oxygen; 3 hours) (original magnification ×400). Sections of kidney from mice given 17R-RvD1 (C,F) show less glomerular inflammatory cellular infiltrate (in the form of lymphocytes, neutrophils, and plasma cells), glomerular sclerosis (arrow), and thrombi (arrow) compared with vehicle-treated mice (B,D-E); also see Table 1). (G) Immunoblot analysis, using specific antibodies against phosphorylated (P-)NF-κB, NF-κB, P-Nrf2, and Nrf2, in kidney from AA and SS mice treated as in (B). One representative gel from 6 gels with similar results is shown. Densitometric analysis of immunoblots is shown in supplemental Figure 2A. (H) Immunoblot analysis, using specific antibodies against HO-1, IL-6, ET-1, VCAM-1, and TXAS-1, of kidney from AA and SS mice treated as in (B). One representative gel from 6 gels with similar results is shown. Vertical line(s) in ET-1 gel have been inserted to indicate a repositioned gel lane. Densitometric analysis immunoblots are shown in supplemental Figure 2A. (I) Effect of 17R-RvD1 on kidney let7c expression. Levels of miRNA let7c were quantified, using real-time polymerase chain reaction, in kidneys collected from AA and SS mice that were treated as above. Results are mean ± SD from 3 to 6 mice per group.

Figure 7.

17R-RvD1 reduces vascular vulnerability in SCD mice during acute VOCs. (A) Immunoblot analysis, using specific antibodies against HO-1, IL-6, ET-1, VCAM-1, and TXAS-1, of isolated aorta from AA and SS mice under normoxic conditions and treated with vehicle or 17R-RvD1 (100 ng) and exposed to H/R: hypoxia (8% oxygen; 10 hours) followed by reoxygenation (21% oxygen; 3 hours) (left panel). One representative gel from 6 gels with similar results is shown. Vertical line(s) have been inserted to indicate a repositioned gel lane. Densitometric analysis of immunoblots (right panels). All data are mean ± SD (n = 6). *P < .05 vs normoxia, °P < .05 vs healthy mice (AA), ^P < .05 vs vehicle. (B) VCAM-1 expression was evaluated in aortas isolated from AA or SS mice exposed to H/R and treated with 17R-RvD1, as above (left panels). Following incubation with the indicated antibodies, fluorescence intensities of VCAM-1 staining (green channel) were quantified in digital images (2 to 6 microscopic fields per sample taken at a ×630 magnification scale) by selecting green stained vessels using the Magic Wand Tool in Adobe Photoshop. Semiquantitative analysis of the number of pixels in the selected fields is shown in a bar graph (right panel). *P < .05.