Electrophilic fatty acid species inhibit 5-lipoxygenase and attenuate sepsis-induced pulmonary inflammation

Abstract

Aims: The reaction of nitric oxide and nitrite-derived species with polyunsaturated fatty acids yields electrophilic fatty acid nitroalkene derivatives (NO2-FA), which display anti-inflammatory properties. Given that the 5-lipoxygenase (5-LO, ALOX5) possesses critical nucleophilic amino acids, which are potentially sensitive to electrophilic modifications, we determined the consequences of NO2-FA on 5-LO activity in vitro and on 5-LO-mediated inflammation in vivo.

Results: Stimulation of human polymorphonuclear leukocytes (PMNL) with nitro-oleic (NO2-OA) or nitro-linoleic acid (NO2-LA) (but not the parent lipids) resulted in the concentration-dependent and irreversible inhibition of 5-LO activity. Similar effects were observed in cell lysates and using the recombinant human protein, indicating a direct reaction with 5-LO. NO2-FAs did not affect the activity of the platelet-type 12-LO (ALOX12) or 15-LO-1 (ALOX15) in intact cells or the recombinant protein. The NO2-FA-induced inhibition of 5-LO was attributed to the alkylation of Cys418, and the exchange of Cys418 to serine rendered 5-LO insensitive to NO2-FA. In vivo, the systemic administration of NO2-OA to mice decreased neutrophil and monocyte mobilization in response to lipopolysaccharide (LPS), attenuated the formation of the 5-LO product 5-hydroxyeicosatetraenoic acid (5-HETE), and inhibited lung injury. The administration of NO2-OA to 5-LO knockout mice had no effect on LPS-induced neutrophil or monocyte mobilization as well as on lung injury.

Innovation: Prophylactic administration of NO2-OA to septic mice inhibits inflammation and promotes its resolution by interfering in 5-LO-mediated inflammatory processes.

Conclusion: NO2-FAs directly and irreversibly inhibit 5-LO and attenuate downstream acute inflammatory responses.

FIG. 1.

Comparison of the effects of OA, LA, NO₂-OA, and NO₂-LA on the activity of 5-LO in intact human PMNL. (A–C) Concentration-response curves of human PMNL stimulated with A23187 (2.5 μM) in the absence of arachidonic acid (AA, A) and in the presence of 2 μM (B) as well as 20 μM AA (C). (D and E) Concentration-response curve of human PMNL stimulated with NaCl (300 mM, D) or fMLP (100 nM) after GM-CSF (1 nM) priming (E). The graphs summarize data from three to six independent experiments each using a different batch of human PMNL/eosinophil preparations; *p<0.05, **p<0.01, ***p<0.001 versus OA. fMLP, N-formyl-methionine-leucine-phenylalanine; GM-CSF, granulocyte macrophage colony-stimulating factor; LA, linoleic acid; LO, lipoxygenase; OA, oleic acid; PMNL, polymorphonuclear leukocytes.

FIG. 2.

Comparison of the effects of NO₂-FAs and parent fatty acids on the 5-, 12-, and 15-LO-activity after stimulation in the recombinant system or intact human cells. (A) Concentration-response curve of recombinant human 5-LO in the presence of all fatty acids. (B) Comparison of the effects of increasing amounts of arachidonic acid (AA) on the NO₂-FA-mediated 5-LO inhibition. Recombinant 5-LO was incubated with the indicated concentrations of AA in the presence of 5 μM OA, LA, NO₂-OA, or NO₂-LA and 5-LO metabolites were measured. (C, D) Effect of fatty acids on 12-HETE formation in human platelets (C) and 15-HETE formation in human eosinophils (D). (E) Reversibility of the effect of fatty acids (10 μM) on PMNL 5-LO activity. BW (BW4AC, 1 μM) and U73122 (10 μM) were used as positive controls for a reversible and irreversible inhibition, respectively. (F) Comparison of the effects of NO₂-FAs (3 μM) on the wild-type (WT) enzyme and different 5-LO cysteine mutants. The enzyme activities were WT: 0.8±0.1 μmol·min⁻¹·mg⁻¹, C4: 0.8±0.1 μmol·min⁻¹·mg⁻¹, C159S: 0.7±0.1 μmol·min⁻¹·mg⁻¹, C300S: 0.8±0.2 μmol·min⁻¹·mg⁻¹, C416S: 0.5±0.1 μmol·min⁻¹·mg⁻¹, and C418S: 0.9±0.1 μmol·min⁻¹·mg⁻¹. (G) Consequence of NO₂-FA stimulation on the iron content of WT and C418S 5-LO. Values were normalized to the iron content of the corresponding solvent-incubated enzyme. The graphs summarize data from three to six independent experiments each using a different batch of PMNL eosinophils, platelets or recombinant human 5-LO; *p<0.05, **p<0.01, and ***p<0.001 versus corresponding non-nitrated fatty acid (A, B, G), wash out (E), or WT (F). NO₂-FA, nitro-fatty acids; NO₂-LA: nitro-linoleic acid; NO₂-OA: nitro-oleic acid.

FIG. 3.

Collision-induced dissociation spectra identifying NO₂-OA adducts on Cys416 and Cys418 in human 5-LO. Precursor ions of peptide EQLIC₄₁₆EC₄₁₈GLFDK alkylated by iodoacetamide with a mass of 1511.69 Da [M+H]⁺ or alkylated with NO₂-OA at Cys416 or Cys418 with masses of 1781.92 Da [M+H]⁺ were isolated and fragmented by collision-induced dissociation in the linear ion trap. Generated product ions were evaluated for carbamidomethylation or NO₂-OA adducts. (A) Collision-induced dissociation spectra of the native peptide mapping relevant b- and y-type fragment ions. Both cysteines are carbamido methylated and exhibit a 57.02 Da mass shift to the corresponding fragment ions b5, b6, and y7. (B) The fragment ions b5 and y8 showed a mass shift of 327.24 Da that indicated the adduction of NO₂-OA at Cys416. (C) The fragment ions y6 and y7 exhibited a mass shift of 327.24 Da that indicated the adduction of NO₂-OA at Cys418. Due to higher hydrophobicity, NO₂-OA alkylated peptides elute 30 min later than native peptide. They also exhibit spectra of reduced intensity.

FIG. 4.

Consequences of OA- and NO₂-OA treatment on LPS-induced lung injury and cellular infiltration in wild-type mice. C57Bl6/J mice treated with vehicle (DMSO), OA, NO₂-OA, or zileuton were challenged with solvent (sol, PBS) or LPS (20 mg/kg). After 16 h, lung and blood samples were taken for subsequent analysis. Quantification of infiltrating (A) neutrophils and (B) monocytes in the lung. (C) Representative hematoxylin- and eosin-stained lung sections of solvent and LPS-challenged mice treated with vehicle, OA, NO₂-OA, or zileuton. The bar represents 100 μm. (D–G) Consequences of OA-, NO₂-OA-, and zileuton treatment on LPS-induced (D) LTB₄-, (E) 5-HETE (F) 12-HETE, and (G) PGE₂ levels in lung tissue. The graphs summarize data from 3 to 8 mice/group; *p<0.05, **p<0.01, and ***p<0.001 versus LPS+vehicle.

FIG. 5.

Consequences of OA- and NO₂-OA treatment on LPS-induced lung injury and cellular infiltration in 5-LO-deficient mice (5-LO^−/−). Wild-type (WT) and 5-LO^−/− (−/−) treated with vehicle (veh, 50% DMSO), OA, or NO₂-OA were challenged with solvent (sol) or LPS. After 16 h, lung and blood samples were taken for subsequent analysis. Quantification of infiltrating (A) neutrophils and (B) monocytes in the lung. (C) Representative hematoxylin- and eosin-stained lung sections of solvent- or LPS-challenged WT and 5-LO^−/− mice. (D) Representative hematoxylin- and eosin-stained lung sections of solvent- or LPS-challenged 5-LO^−/− mice treated with vehicle, OA, NO₂-OA, or zileuton. The bar represents 100 μm. (E, F) Consequences of OA- and NO₂-OA treatment on LPS-induced neutrophils (E) and monocyte mobilization (F) into blood of WT and 5-LO⁻/⁻ mice. The graphs summarize data from 4 to 7 mice/group. *p<0.05, and ***p<0.001 versus WT+Sol, ^#p<0.05, ^##p<0.01, and ^###p<0.001 versus 5-LO^−/−+Sol.

FIG. 6.

Consequences of LA-, OA-, NO₂-LA-, and NO₂-OA treatment on the LPS-induced expression of TNF-α, MCP1, 5-, 12-, and 15-ALOX in human WBC. Freshly isolated human WBC were incubated with solvent (CTL), the indicated fatty acids (5 μM), or an inhibitor of the I(B kinase (BMS, 8 μM) for 8 h in the presence and absence of LPS (100 ng/ml) and qRT-PCR was performed. (A) TNF-α, (B) MCP-1, (C) ALOX-5, (D) ALOX-12, and (E) ALOX-15 mRNA expression. The graphs summarize data from three independent experiments each using a different cell donor; *p<0.05, and ***p<0.001 versus CTL+LPS; ^#p<0.05, ^##p<0.01, and ^###p<0.001 versus CTL+Sol. LPS, lipopolysaccharide; TNF-α, Tumor necrosis factor-α; WBC, white blood cell.

FIG. 7.

NO₂-FA-mediated 5-LO inhibition. (A) Reaction of 9-NO₂-OA with a cysteine or histdine residue within target proteins. Electrophilic adduction of NO₂-OA to a thiolate of a cysteine (upper scheme) or imidazol residue of histidine (lower scheme) leads to alkylation of the corresponding residue. (B) Ribbon diagram of human 5-LO (PDB code 3o8y (2)) showing the FY-cork (orange), the domain containing the iron coordinating H367 and 372 (green), as well as the domain harboring C418 (yellow); the C2-like domain (light pink) and the catalytic domain (gray). The detailed view of the 5-LO catalytic domain shows the F177 and Y181 residues that form the FY-cork as well as the iron coordinating residues H367 and 372 and C418. The alkylation of C418 (a) along with the alkylation of H367 and H372 leads to the release of iron (red sphere; b) and inhibition of the 5-LO. The adduction of NO₂-OA to C418 may enable the entry of NO₂-FA to the catalytic domain before the modification of H367 and 372 (c). The ribbon diagrams were prepared using PyMol (www.pymol.org).