Dihydroxyoctadecamonoenoate esters inhibit the neutrophil respiratory burst

Abstract

The leukotoxins [9(10)-and 12(13)-EpOME] are produced by activated inflammatory leukocytes such as neutrophils. High EpOME levels are observed in disorders such as acute respiratory distress syndrome and in patients with extensive burns.Although the physiological significance of the EpOMEs remains poorly understood,in some systems, the EpOMEs act as a protoxin,with their corresponding epoxide hydrolase metabolites,9,10-and 12,13-DiHOME, specifically exerting toxicity.Both the EpOMEs and the DiHOMEs were also recently shown to have neutrophil chemotactic activity.We evaluated whether the neutrophil respiratory burst,a surge of oxidant production thought to play an important role in limiting certain bacterial and fungal infections,is modulated by members of the EpOME metabolic pathway.We present evidence that the DiHOMEs suppress the neutrophil respiratory burst by a mechanism distinct from that of respiratory burst inhibitors such as cyclosporin H or lipoxin A4,which inhibit multiple aspects of neutrophil activation.

Figure 1

(A) Representative cytochrome P450 and sEH metabolites of linoleic acid and arachidonic acid. (B) One of the monoepoxide derivatives of linoleic acid, 9(10)-EpOME and the corresponding sEH metabolite, 9,10-DiHOME. (C) Structure of a dihydroxy derivative of arachidonic acid, 5,6-DiHET.

Figure 2

9(10)-/12(13)-EpOME stimulates the neutrophil respiratory burst. Granulocytic HL-60 cells were treated with the indicated concentration of 9(10)-/12(13)-EpOME or linoleic acid, and cotreated with 7.5 μM EGTA as indicated. WST-1 reduction was evaluated as described in Materials and methods. Adjusted A440 values reflect the extent of WST-1 reduction which was inhibited by 200 U/ml superoxide dismutase. Error bars represent population standard error of two to three independent experiments. Analysis of the EpOME dose–response curve by linear regression from 10 nM to 200 μM yielded a statistically significant (P = 0.01) dose–response.

Figure 3

Methyl 9,10-/12,13-DiHOME (Me DiHOMEs) inhibit the neutrophil respiratory burst. Granulocytic HL-60 cells were treated with 10 nM PMA in conjunction with 200 μM of the indicated compound (unless otherwise indicated) or with vehicle. Nitroblue tetrazolium reduction was evaluated as described in Materials and methods. Error bars represent population standard error of two to three independent experiments. The vehicle (no PMA stimulation) control value was typically 0–2%. The P values compared to vehicle were P < 0.01 (Me DiHOMEs, 20 μM), P < 0.01 (Me DiHOMEs, 200 μM), P = 0.075 (Me EpOMEs), and P < 0.05 (Me linoleate).

Figure 4

Methyl 9,10-/12,13-DiHOME inhibits PMA-induced lucigenin-dependent chemiluminescence. Granulocytic HL-60 cells were stimulated as indicated with 10 nM PMA or the PMA vehicle (−) and cotreated with 200 μM of methyl 9,10-/12,13-DiHOME or with the corresponding vehicle (−). Integrated (30 min) lucigenin-dependent chemiluminescence (expressed in arbitrary units) was evaluated as described in Materials and methods. Error bars represent the population standard error of two independent experiments. The P values compared to cells treated only with PMA were P < 0.01 (vehicle/vehicle) and P = 0.01 (treatment with PMA and Me DiHOMEs).

Figure 5

Methyl 9,10-DiHOME and methyl 12,13-DiHOME inhibit the respiratory burst at similar concentrations. Granulocytic HL-60 cells were treated with 10 nM PMA in conjunction with the indicated concentration of the compound mentioned or with vehicle. WST-1 reduction was evaluated as described in Materials and methods; adjusted A440 represents absorbance at 440 nM after subtracting absorbance at 440 nm of cells cotreated with 10 nM PMA and 200 U/ml superoxide dismutase.

Figure 6

(A) Methyl 9,10-/12,13-DiHOME specifically inhibits the neutrophil respiratory burst. Granulocytic HL-60 cells were treated with 20 nM PMA and/or with 200 μM of either the fatty acid (F) or methyl ester (M) of 9,10-/12,13-DiHOME (DiHOMEs). The respiratory burst was assessed by microscopic evaluation of nitroblue tetrazolium reduction. Error bars represent standard deviation. Results are representative of two independent experiments. (B) Methyl stearate diols do not inhibit the neutrophil respiratory burst. Granulocytic HL-60 cells were treated with 20 nM PMA in conjunction with 200 μM of the indicated lipid. Nitroblue tetrazolium reduction was assessed as described in Materials and methods. DHO represents dihydroxyoctadecanoic acid or (for esters) dihydroxyoctadecanoate. DiHOMEs represent a mixture of the indicated derivatives of 12,13-dihydroxyocta-9-enoate and 9,10-dihydroxyocta-12-enoate. Me (methyl), Pr (propyl), and Bu (butyl) refer to the corresponding esters. Error bars represent the population standard error of two independent experiments. The vehicle control value was typically 0–2%. (C) Methyl DHETs inhibit the respiratory burst. Granulocytic HL-60 cells were treated with the indicated concentration of methyl 5,6-, 8,9-, 11,12-, or 14,15-DHET and with 10 nM PMA. WST-1 reduction was evaluated as described in Materials and methods. Adjusted A440 values represent WST-1 reduction inhibited by 200 U/ml superoxide dismutase and scaled based on the value observed with 10 nM PMA (this value varied substantially (from 0.181 to 0.584) between experiments since granulocytic HL-60 cells gradually lose responsiveness to PMA with passaging). Error bars represent population standard error of three to four independent experiments. Within each experiment, variability of replicate samples (10 nM PMA) was less than 5%. Inset: Granulocytic HL-60 cells were treated with 20 nM PMA and evaluated for respiratory burst activation as described above. Symbols represent methyl 5,6,14, 15-tetrahydroxyeicosadienoate(circles),8,9,14,15-tetrahydroxyeico-sadienoate (squares), a mixture of methyl 9-, 11-, and 12-hydroxyeicosatetraenoates (triangles), and 8-hydroxyeicosatetraenoate (diamonds). Error bars represent the population standard error of three to four independent experiments.

Figure 7

(A) Methyl 9,10-/12,13-DiHOME inhibits fMLF-induced nitroblue tetrazolium reduction. Granulocytic HL-60 cells were incubated with 1 μM fMLF and cotreated with either 200 μM methyl 9,10-/12,13-DiHOME or vehicle as indicated. Nitroblue tetrazolium reduction was evaluated as described in Materials and methods. Error bars represent population standard error of two independent experiments. The P value is 0.05. (B) Methyl 9,10-/12,13-DiHOME does not inhibit fMLF-induced β-glucuronidase release. Granulocytic HL-60 cells were stimulated with 1 μM fMLF and cotreated with 200 μM methyl 9,10-/12,13-DiHOME. β-Glucuronidase release was evaluated as described in Materials and methods. Results represent three independent experiments. Error bars represent population standard error. Since the magnitude of β-glucuronidase varied substantially between experiments, results were normalized based on the minimal and maximal response in each data set. P values (with respect to the vehicle control) were P < 0.01 for both fMLF-treated groups. The P value with respect to comparison between fMLF-treated and fMLF- and MeDiHOME-treated cells was P > 0.1.

Figure 8

Methyl 9,10-/12,13-DiHOME does not inhibit PMA-induced p47phox phosphorylation. (A) HL-60 cells were lysed and p47phox immunoprecipitated from the indicated amount of cell lysate protein using anti-p47phox primary antibody and protein A-Sepharose beads. Protein was eluted with Laemmli sample buffer and subjected to SDS-PAGE and subsequent immunoblotting for p47phox as described in Materials and methods. Line arrows indicate the positions of the heavy and light antibody chains; the closed arrow indicates the band corresponding to p47phox. (B) Granulocytic HL-60 cells were treated with 20 nM PMA and with 200 μM methyl 9,10-/12,13-DiHOME as indicated. After 14 min, cells were harvested. Cell lysates were subjected to NEPHGE (a variant of isoelectric focusing gel electrophoresis); the gel was immunoblotted for p47phox as described in Materials and methods. pI standard locations are indicated at the margins of the image. Data from two representative experiments are shown. (C) India ink staining of the immunoblot membrane shown in C was used to verify that equal masses of protein were loaded for each sample.