Biosynthetic Crossover of 5-Lipoxygenase and Cyclooxygenase-2 Yields 5-Hydroxy-PGE2 and 5-Hydroxy-PGD2

Abstract

The biosynthetic crossover of 5-lipoxygenase (5-LOX) and cyclooxygenase-2 (COX-2) enzymatic activities is a productive pathway to convert arachidonic acid into unique eicosanoids. Here, we show that COX-2 catalysis with 5-LOX derived 5-hydroxy-eicosatetraenoic acid yields the endoperoxide 5-hydroxy-PGH₂ that spontaneously rearranges to 5-OH-PGE₂ and 5-OH-PGD₂, the 5-hydroxy analogs of arachidonic acid derived PGE₂ and PGD₂. The endoperoxide was identified via its predicted degradation product, 5,12-dihydroxy-heptadecatri-6E,8E,10E-enoic acid, and by SnCl₂-mediated reduction to 5-OH-PGF_2α. Both 5-OH-PGE₂ and 5-OH-PGD₂ were unstable and degraded rapidly upon treatment with weak base. This instability hampered detection in biologic samples which was overcome by in situ reduction using NaBH₄ to yield the corresponding stable 5-OH-PGF₂ diastereomers and enabled detection of 5-OH-PGF_2α in activated primary human leukocytes. 5-OH-PGE₂ and 5-OH-PGD₂ were unable to activate EP and DP prostanoid receptors, suggesting their bioactivity is distinct from PGE₂ and PGD₂.

Scheme 1. Biosynthesis and Transformation of Endoperoxides in the Reaction of Cyclooxygenases with Arachidonic Acid and 5-HETE^,

Scheme 2. Reaction of the C-8 Carbinyl Radical Intermediate 1 in COX-2 Catalysis with 5-HETE Determines the Formation of Hemiketals (HK) or 5-Hydroxy-Prostaglandins as Enzymatic Products

Figure 1

LC-MS detection of HKE₂ and HKD₂ (m/z 399 ion trace) and novel products 3 and 4 (m/z 367) in a reaction of recombinant human COX-2 with 5-HETE substrate (m/z 319).

Figure 2

Indirect identification of 5-OH-PGH₂. (A) Treatment with SnCl₂ of a 5 min reaction of COX-2 with 5-HETE generated product 9 with m/z 369 that had the same retention time as a standard of 5-OH-PGF_2α formed by reduction (NaBH₄) of 5-OH-PGD₂ (inset). (B) Treatment of the enzymatic reaction with hematin yielded product 10 identified as 5,12-dihydroxy-heptadecatri-6E,8E,10E-enoic acid (5,12-diHHT) due to its UV spectrum characteristic of a conjugated triene in E,E,E configuration (inset) and (C) fragment ions derived from m/z 295.4 (10) in the MS2 spectrum. (D) Products of the reaction of 5-OH-PGH₂ (2) with SnCl₂ and hematin.

Figure 3

Base-catalyzed dehydration of prostaglandins. (A) Dehydration of PGE₂ yields PGA₂ that rearranges to PGB₂. (B) Time-course of dehydration of PGE₂ in NaOH (200 mM) illustrates formation of PGA₂ (λmax 217 nm) and PGB₂ (λmax at 278 nm). UV/vis spectra were acquired every 2 min. (C) Degradation of 5-OH-PGE₂ in 10-times dilute NaOH (20 mM) indicating complete conversion to a product with λmax 295 nm at the time of the first scan. (D) Degradation of 5-OH-PGE₂ in Tris-HCl (50 mM, pH 8). (E) Degradation of 5-OH-PGD₂ in NaOH (20 mM), with t = 0 min indicating conversion to a product with λmax 295 nm. The UV/vis scanning rate was 1/min in panels C–E.

Figure 4

5-OH-PGF₂ diastereomers formed by NaBH₄ reduction of (A) a 5 min reaction of 5-HETE with COX-2, (B) synthetic 5-OH-PGD₂, and (C) synthetic 5-OH-PGE₂. The ion trace for m/z 369 from LC-MS analysis of the reactions is shown. Products in panel A are expected to be mainly derived from reduction of 5-OH-PGH₂.

Figure 5

Formation of 5-OH-PGs and HKs by activated human leukocytes. Cells were treated with LPS for 45 min and then with A23187 for 15 min, followed by reduction or not with NaBH₄, extracted, derivatized with AMPP, and analyzed using LC-MS in the positive ion SRM mode. Ion traces for the detection of (A) 5-OH-PGF_2α, (B) 5-OH-PGD₂ and 5-OH-PGE₂, (C) 5,8,9,11,12,15-hexahydroxy-eicosadienoic acid 13, (D) HKE₂, and (E) HKD₂ are shown. (F) Quantification of the eicosanoids shown to the left in human leukocyte samples (n = 3) treated with or without NaBH₄ (*, p < 0.05). The arrows in panel B indicate the retention times of 5-OH-PGE₂ (§) and 5-OH-PGD₂ (#), respectively.