To the editor:
We read with great interest the article by Hedge et al,1 reporting that Δ12-prostaglandin (PG) J3 (Δ12-PGJ3) has antileukemic activity in mice. Anti-inflammatory and antineoplastic activity has also been reported for 15-deoxy-Δ12,14-prostaglandin J2 (15d-PGJ2).2 We agree with Hedge et al1 that one of the most important questions is whether sufficient quantities of Δ12-PGJ3 are formed in vivo to exert any biologic activity. Here, we comment on this eminently crucial issue from pharmacologic and nutrition perspectives.
PGJ3 and PGJ2 are the dehydrated products of PGD3 and PGD2 formed in vivo from eicosapentaenoic acid (EPA) and arachidonic acid (ARA), respectively, by the catalytic action of cyclooxygenase (COX). PGJ3 and PGJ2 are further dehydrated and isomerized to produce Δ12-PGJ3 and 15d-PGJ3 and 5d-PGJ2, respectively. Common feature of Δ12-PGJ3 and 15d-PGJ2 is the highly reactive cyclopentenone ring, which is readily attacked by low- and high-molecular-mass thiols to form thioethers (Figure 1). Thiolation of Δ12-PGJ3 and 15d-PGJ2 is likely to reduce both availability and bioactivity of Δ12-PGJ3 and 15d-PGJ2. So far, there are no data about excretion of Δ12-PGJ3 and 15d-PGJ3. We (Figure 1) and others3 found only pM-concentrations of 15d-PGJ2 in human urine, while PGJ3 metabolites including 15d-PGJ3 were below the detection limit of our method (30 pM) in urine. This may suggest that basal PGJ3 biosynthesis from EPA is several fold lower than PGJ2 from ARA. Dietary EPA has been shown to increase formation of prostaglandin I3 (PGI3) and thromboxane A3 (TxA3), but EPA, even at very high doses, did not increase PGI3 and TxA3 synthesis to a degree comparable with that of PGI2 and TxA2 from ARA.4
![Figure 1. Excretion of 15d-PGJ2 in human urine and its in vitro conjugation with glutathione, L-cysteine and N-acetylcysteine. (A) Reaction of 30μM 15d-PGJ2 with each 1110μM glutathione (GSH), L-cysteine (Cys) or N-acetylcysteine (NAC) in 100mM phosphate buffer (pH 7.4) resulted in formation of the corresponding conjugates and concomitant decrease of 15d-PGJ2 as measured by high-performance liquid chromatography (HPLC). Retention time was 12.7, 3.6, 2.8 and 1.2 minutes for 15d-PGJ2 and the 15d-PGJ2-NAC, 15d-PGJ2-Cys, and 15d-PGJ2-GSH conjugates, respectively. Reaction of 15d-PGJ2 with Cys was accompanied by a shift of the maximum wavelength from 318 nm to 312 nm and an increase in absorbance at 230 nm. (B,C) The HPLC fractions of the above mentioned conjugates were collected and subjected to catalytical hydrogenation/desulfurization as described elsewhere for the cysteinyl leukotriene E4.5 After derivatization with pentafluorobenzyl (PFB) bromide (PFB-Br) followed by N-methyl-N-trimethylsilyl-trifluoroacetamide (MTSFA) in the presence of NH4I and 2-mercaptoethanol (ME), gas chromatography-mass spectrometry (GC-MS) spectra were generated in the electron-capture negative-ion chemical ionization mode (B). The precursor ion at m/z 397 [M-PFB]− was subjected to collision-induced dissociation (CID) to generate GC-tandem MS (GC-MS/MS) spectra (C). Expectedly, virtually identical GC-MS and GC-MS/MS mass spectra were obtained from all thiol (RSH) conjugates of 15d-PGJ2. Inserts in panels B and C indicate schematically part of the analytical procedure used and the proposed structures for the ions obtained. (D) Excretion of 15d-PGJ2 and the isoprostane 15(S)-8-iso-PGF2α (iso-PGF2α) was measured in fresh spot urine samples of 12 healthy volunteers (4 females) by GC-MS/MS using 2H4-15d-PGJ2 and 2H4-15(S)-8-iso-PGF2α as internal standards. 15(S)-8-iso-PGF2α was extracted from urine (1 mL) by immunoaffinity column chromatography.6 15d-PGJ2 was extracted from acidified (pH 4.5) urine samples by solid-phase extraction and purified by isocratic reverse phase HPLC. In the urine samples no 15d-PGJ3 was detectable. 15(S)-8-iso-PGF2α was measured because it is considered a COX-independent metabolite, analogous to 15d-PGJ2 and 15d-PGJ3.](https://ash.silverchair-cdn.com/ash/content_public/journal/blood/119/12/10.1182_blood-2012-01-403105/4/m_zh89991288600001.jpeg?Expires=1735516409&Signature=UvCJunxLodJetauiO1Z0sbbhsbg-9pSkHWqQ92yuZzNIfCw5ST2qSYEIYa7PpjbEomQ8EuBorED6mRITaVLvZSdE3wveC1rR2Qp9~2~R6WhabK5VZGHIAvd4bjQTAgi8RsJ2GFGu2S-I9g-GeMJkVeQNraT9M6JGt2CzCteLiDMwmv45s8rEelxZBH~awsJZqc0GjSSuGtCostEx9hcSumvyE9UUUaXvtpnlz6ajYJgmF9KBqCAcD5p1DwPTschtFa5BBLQBVUKGFjp27M44TiCBFEsrjDl4jfZUIJHSW5JydssrM1Veht2mTF-PIwwSokiL2AC-1lZe1phI1UyVrQ__&Key-Pair-Id=APKAIE5G5CRDK6RD3PGA)
Excretion of 15d-PGJ2 in human urine and its in vitro conjugation with glutathione, L-cysteine and N-acetylcysteine. (A) Reaction of 30μM 15d-PGJ2 with each 1110μM glutathione (GSH), L-cysteine (Cys) or N-acetylcysteine (NAC) in 100mM phosphate buffer (pH 7.4) resulted in formation of the corresponding conjugates and concomitant decrease of 15d-PGJ2 as measured by high-performance liquid chromatography (HPLC). Retention time was 12.7, 3.6, 2.8 and 1.2 minutes for 15d-PGJ2 and the 15d-PGJ2-NAC, 15d-PGJ2-Cys, and 15d-PGJ2-GSH conjugates, respectively. Reaction of 15d-PGJ2 with Cys was accompanied by a shift of the maximum wavelength from 318 nm to 312 nm and an increase in absorbance at 230 nm. (B,C) The HPLC fractions of the above mentioned conjugates were collected and subjected to catalytical hydrogenation/desulfurization as described elsewhere for the cysteinyl leukotriene E4.5 After derivatization with pentafluorobenzyl (PFB) bromide (PFB-Br) followed by N-methyl-N-trimethylsilyl-trifluoroacetamide (MTSFA) in the presence of NH4I and 2-mercaptoethanol (ME), gas chromatography-mass spectrometry (GC-MS) spectra were generated in the electron-capture negative-ion chemical ionization mode (B). The precursor ion at m/z 397 [M-PFB]− was subjected to collision-induced dissociation (CID) to generate GC-tandem MS (GC-MS/MS) spectra (C). Expectedly, virtually identical GC-MS and GC-MS/MS mass spectra were obtained from all thiol (RSH) conjugates of 15d-PGJ2. Inserts in panels B and C indicate schematically part of the analytical procedure used and the proposed structures for the ions obtained. (D) Excretion of 15d-PGJ2 and the isoprostane 15(S)-8-iso-PGF2α (iso-PGF2α) was measured in fresh spot urine samples of 12 healthy volunteers (4 females) by GC-MS/MS using 2H4-15d-PGJ2 and 2H4-15(S)-8-iso-PGF2α as internal standards. 15(S)-8-iso-PGF2α was extracted from urine (1 mL) by immunoaffinity column chromatography.6 15d-PGJ2 was extracted from acidified (pH 4.5) urine samples by solid-phase extraction and purified by isocratic reverse phase HPLC. In the urine samples no 15d-PGJ3 was detectable. 15(S)-8-iso-PGF2α was measured because it is considered a COX-independent metabolite, analogous to 15d-PGJ2 and 15d-PGJ3.
Excretion of 15d-PGJ2 in human urine and its in vitro conjugation with glutathione, L-cysteine and N-acetylcysteine. (A) Reaction of 30μM 15d-PGJ2 with each 1110μM glutathione (GSH), L-cysteine (Cys) or N-acetylcysteine (NAC) in 100mM phosphate buffer (pH 7.4) resulted in formation of the corresponding conjugates and concomitant decrease of 15d-PGJ2 as measured by high-performance liquid chromatography (HPLC). Retention time was 12.7, 3.6, 2.8 and 1.2 minutes for 15d-PGJ2 and the 15d-PGJ2-NAC, 15d-PGJ2-Cys, and 15d-PGJ2-GSH conjugates, respectively. Reaction of 15d-PGJ2 with Cys was accompanied by a shift of the maximum wavelength from 318 nm to 312 nm and an increase in absorbance at 230 nm. (B,C) The HPLC fractions of the above mentioned conjugates were collected and subjected to catalytical hydrogenation/desulfurization as described elsewhere for the cysteinyl leukotriene E4.5 After derivatization with pentafluorobenzyl (PFB) bromide (PFB-Br) followed by N-methyl-N-trimethylsilyl-trifluoroacetamide (MTSFA) in the presence of NH4I and 2-mercaptoethanol (ME), gas chromatography-mass spectrometry (GC-MS) spectra were generated in the electron-capture negative-ion chemical ionization mode (B). The precursor ion at m/z 397 [M-PFB]− was subjected to collision-induced dissociation (CID) to generate GC-tandem MS (GC-MS/MS) spectra (C). Expectedly, virtually identical GC-MS and GC-MS/MS mass spectra were obtained from all thiol (RSH) conjugates of 15d-PGJ2. Inserts in panels B and C indicate schematically part of the analytical procedure used and the proposed structures for the ions obtained. (D) Excretion of 15d-PGJ2 and the isoprostane 15(S)-8-iso-PGF2α (iso-PGF2α) was measured in fresh spot urine samples of 12 healthy volunteers (4 females) by GC-MS/MS using 2H4-15d-PGJ2 and 2H4-15(S)-8-iso-PGF2α as internal standards. 15(S)-8-iso-PGF2α was extracted from urine (1 mL) by immunoaffinity column chromatography.6 15d-PGJ2 was extracted from acidified (pH 4.5) urine samples by solid-phase extraction and purified by isocratic reverse phase HPLC. In the urine samples no 15d-PGJ3 was detectable. 15(S)-8-iso-PGF2α was measured because it is considered a COX-independent metabolite, analogous to 15d-PGJ2 and 15d-PGJ3.
Δ12-PGJ3 and 15d-PGJ2 are considered potentially useful therapeutic agents for the treatment of cancer.1,2 Dietary EPA supplementation is unlikely to produce nM-concentrations of Δ12-PGJ3 required for antileukemic activity, but topical administration of considerable amounts of synthetic Δ12-PGJ3 would be required.
Authorship
Acknowledgments: The authors thank B. Beckmann, K. Berg, A. Mitschke and M.-T. Suchy for laboratory assistance and Frank-Mathias Gutzki for performing GC-MS/MS analyses.
Contribution: D.T. and D.O.S designed and performed the study, analyzed the data, and wrote the manuscript.
Conflict-of-interest disclosure: The authors declare no competing financial interests.
Correspondence: Prof Dimitrios Tsikas, Institute of Clinical Pharmacology, Hannover Medical School, Carl-Neuberg-Str 1, 30625 Hannover, Germany; e-mail: tsikas.dimitros@mh-hannover.de.
