Generation and Metabolism of Lipoxygenase Products in Normal and Membrane-Damaged Cultured Human Keratinocytes Floyd A. Green, M.D. Departments of Medicine and Microbiology, State University of New York at Buffalo and Veterans Administration Medical Center, Buffalo, N ew York, U.S.A. The production and metabolism of lipoxygenase eicosanoids nous lS-HETE by normal keratinocytes was observed. Mea­ were srudied in cultured human keradnocytes. The identiry surable quantities of esterified l S-HETE were found after 1 of these eicosanoid structures was estabLished by a variety of min, but by 18 - 20 h all the esterified IS-HETE was de­ chromatographic and analyti cal techniques. Normal cul­ graded to the extent that 80% o[ the recovered radioacti vity tured kerarinocyrcs did not produce lipoxygenase cicosanoids was found in water-soluble form. In contrast, when JabeJed either spontaneously or when given arachidonic ac id in the or unlabeled 5-HETE was used a much larger fraction was presence of permeabilizing concentrations of ethanol or di ­ esterified intact (30% as opposed to 10%) and at the end of methyl sulfox.ide. Freeze-thawing of human neonatal and 18 - 20 hours a substantial reak of esterified S- HETE re­ adult keratinocytes resulted in a rapid release of linoleic and mained. Intact esterified [3H HETE were recovered only in arachidonic acids over rime. Activation of a latent 15-lipoxy­ the triacy lglycerol fraction. The key findings that W-6Iipox­ genase was demonstrated by the synthesis of l S-hydroxyei­ ygenase products are generated but not esterified by mem­ cosatetraenoic acid (15-HETE) and 13-hydroxyoctadeca­ brane-damaged keratinocytes, whereas these products are es­ dienoic acid. and both these products were greatly increased rerified bur not genera[("d by normal keratinocytes, may be of in amoUJ1[ when the corresponding fatty acid precursor was importance in transcelJular metabolic control.] Intlcst Derma- added. Eicosanoid production by cells of newborn and adult 10/93:486 - 491 , 1989 origin was indistinguishable. Rapid metabolism of exoge- ince the first description of mammalian Jipoxygenases in undertaken with usc of defined cultures. Membrane damage of the 1974 (I} there has heen continuing interest in the role of ke: ratlnocytes by a freez.e-thaw techni<:lue was enlploycd to activate lipoxygenase . J?roducts as inflammatory mediators in the lipoxygenases. The use of intact cul tured keratinocyres also mammalian 12J as well as in amphibian inflammatory allowed examination in vitro of the metabolic fate of eicosanoids models [3J. Following the initial report of Hammarstrom produced by activated Iipoxygenases where such transcellular me­ eralS that 12-HETE could be recovered from the psoriaric epidermis tabolism could also be expected to occur in vivo. [4]. a variety of different eicosanoids has been reported to be presenr MATERIALS AND METHODS in lesional and non-iesional epidermis 15 - 7}. I nte r pr~tatian of these findings has been complicated by the difficulty in distinguishing Cells, IncubOltions. and HPLC Separations Keratinocy te cul­ between eicosanoid products formed by the epidermis, dermis. in­ tures (from both human adult and newborn donors) Wert' obtained filtrating inflammatory cells, and in addition, by co-processi ng from Dr. Howard Green (Depa,rtmenr of Physiology and Bi ophys­ among the different cell types. As a first step in understanding the ics. Harvard Medical School, Boston, MA) [8- 10]. The cultures rale of lipaxygenasc{s) in human dermatologic diseases. a careful were sh ipped in medium-filled flasks, were confluent at the time. of examination of Iipoxygcnase activity in human keratinocyces was arrival, and continued ro divide until used. On the day of the experi­ ment. the medium waS changed to protein and growth facmr-free Medium 199 containing 40 mM pH 7.6 Hepes buffer. The cells were washed twice .in 10 ml of this media at 37 0 for 30 min. For the Manuscript received December 20. t 988: accepted for publication March freeu-thaw experiments. 1 m! of fresh medium was employed to 20. 1989. just cover the cells, and the plates were air frozen at -26°C for a This nudy was supported by NIH GUilt HL24009 and by the Vetcram; minimum of 1 h. In some experiments, rhe cells were then lyophi­ Administration. lized. The frozen cells were rh en rhawed at 370 and studied with or Reprint requests to: Floyd A. Green, M.D., Department of Medicine. without the add.ition of exogenous arachidonic or linoleic acids Division of Rheumatology, VA Medical Center. 3495 Bailey Avenue. Buf­ (2-20 Jig in 2-100,tt1 of ethanol or dimethylsu lfoxide in 1 ml me­ falo . NY 14215. dium). The ex tent of eicosanoid producti on was fou nd not to be Abbreviations: affected by the presence of 1%-9% ethanol or dimethylsulfoxide, IS-HETE: 15(S)-Hydroxycicosa-SZ,8Z.IIZ.1 3E-tctraenoic acid and in some experiments the exogenous fatty acid was dispersed in a 5-H£TE: 5(S)pH ydroxycicosa-6E.8Z, lIZ.14Z-temenoic acid small volume of 15 mM NH 0 H (prior to media-buffering) with GC/MS: Gas chromatography/ mass spectrometry 4 HPLC, LC: High performance liquid chromatography identical results. T he timed incuba.tions w ere stopped by a.dditions 0 13-HOOD: 13(S)-Hydroxyoctadeca-9Z,11 E-dicnoic acid of 3 volumes of ethanol followed by cooling to -25 and transferral TM$: Trimcthylsil yl to plastic tubes. After the addition of prostaglandin Bz or labeled 0022-202X/89/S03.50 Copyright@ 1989 by The Socic:ty fo r Investigative Dermarology, Inc . 486 VOL. 9J. NO. 4 OCTOBER 1989 KERATINOCYTE IS-LIPOXYGENASE .. 87 H ETE as an imernal standard, tile samples were cenuifuged and the incubation of exogenous labeled or unlabeled eicosanoids with in­ supnnacaOf was rotoevaporared. tnnsferred in methanol, and ulti­ tact keratinocytes was tenninated by rhe addition of 1 volume of mately dissolved in mobile phase for reversed-phase HPlC in a methanol and transferred to glass tubes where the methanol was Hewlett-Packard 1090 lC. This was performed isocratically on gently blown off with argon. The samples were then frozen and Hewlett-Packard ODS-Hypcrsil columns, 10 or 20 em in lyophilized. following lyophilization. the samples were treated by length X 4.6 mm at a flow rate of 0.4 rol / min. The- mobile phase the anhydrous methanolic NaOH method of Kates to guantita­ for the l1lonohyd roxyeicosanoids was 75-80: 25-20: 0.1. metha­ tively transesterify aJI esterified fatry acids and their derivatives to nol/water/ acctic acid; and for leukotrienes 45: 37: 18: 1, water/ their respective methyl esters [141. A single modi.fication was made acetonitrile/ methanol/acetic acid. Slight modifications of these in order to partition any unesterifled fatry acids or their hydroxyl­ conditions arc given in the text. The free acids collected during the ated derivatives into the organic layer: after the addition of water to lC runs were methylated with ethereal di:lZomethane. and rechro­ separate the phases. 100,u1 of 2 M ammonium fonnate buffer (pH matographed on the same HlllC columns (delayed retemion rime 3.2) was added. After washing with methanoljwater (10: 9) saru­ of approximately 15 min). All the lC srudies were run at a wave­ rated with chloroform, the lower phase was reconstituted into the length of236 nm, but recall was available at 270 run and 301 nm as mobile phases indicated above. With this procedure, free hydroxyl­ well as in situ UV spectroscopy. T he ordinate scale is given in ated fatry acids were unchanged and eluted at their nonnaJ times. milliabsorption Ul1it5 throughout, and al l ch romatograms a.re repro­ whereas tht HETE, which had been enzymatically esterified by the duced dirccdy from the LC computer plots. f or each iIlustrative keratinocytes. were el uted as their methyl esters. Retention times chromatogram shown at least three other similar experiments were were confirmed by mixing experiments with authentic labeled free performed which showed virtually identical results. In no case and methylated HETE and by in situ UV spectroscopy and ce. shown was rhe main observation of the experiment not demon­ Estimation of Cell Numbers In some experiments the cells strated in every case. Reproducibiliry of the results of studies with were separated from surface culrures by trypsin-EDTA treatment these cell cultures was a noteworthy feature. Retention rimes (RT) followed by counting in a ZBI Coulter Counter. In other experi­ of :mthentic standards are indicated in the figure legends. Software ments aft er incubation the cell proteins were solubilized in 0.5 N programs for guantitation using published molar extinction coeffi­ sodium hydroxide containing 0. 1% TritOn X t OO followed by esti­ cients were cmplo)'ed to measure concentrations as previously de­ mation of proteins by tbe Folin method [15]. scribed [111 . Unlabeled eicosanoids were obtained from BioMol Re:.earch Labs (Plymouth Meeting. PA) and fatry acids from Nu­ GC/ MS Studies were carried our using a KratOs MS 80 RFA mass C hek Prep. Inc. (Elysian. MN). spectrometer at 50-70 eV. Studies of tbe R elease of Arachidonic and Linoleic Acid Fol­ RESULTS lowing Freezing For these srudies [he incubatjons were inter­ Normal Keratinocytes Normal kerarinocytes did not produce ruptC'd by {he addition of3 volumes of methanol containing 50 pi of any lipoxygenase eicosanoids either spontaneously or .1fter feeding acetic acid . After centrifugation and transfer to glass tubes, ex trac­ arachidonic or linoleic acids in the prese.nce of pt:rmeabilizing con­ tion was performed with chlorofoml/ methanol 2: 1. Water was centrations of erhanol or dimethylsulfoxide or aft er dispt:rsion of added to produce rwo phases, and the lower chloroform layer was the fatty acid in dilute NH .