You are on page 1of 9

Metabolism of oxygenated derivatives of arachidonic acid by Caco-2 cells

T. E. Riehl,*+ J. Turk&** and W. F. Stenson*+
Department of Gastroenterology,*Jewish Hospital of St. Louis, and Departments of Medicine? and Pathology,** Washington University School of Medicine, St. Louis, M O 63110

Abstract Monolayers of Caco-2 cells, a humanenterocyte cell line, were incubated separately with "8-labeled prepaof inflammation: 5rations of three different lipid mediators hydroxyeicosatetraenoic acid, 12-hydroxyeicosatetraenoic acid, and leukotriene &. Both ['Hs]ihydroxyeicosatetraenoic and ['H81 12-hydroxyeicosatetraenoicacids were taken

tion is marked by infiltration of the mucosa with lymphocytes, macrophages,neutrophils, mast cells, and eosinophils (6, 7). Elevated levels of the lipoxygenase products 5-HETE,12-HETE,15-HETE, and LTB4 as well as the cyclooxygenase products, prostaglandin E2 up and metabolized b y Caco-2 cells, but ['H]leukotriene B4 and thromboxane B2, are found in the mucosa in inremained unmetabolized in the incubation medium. ['H]5flammation (8, 9). Incubation of colonic mucosa obhydroxyeicosatetraenoic acid was esterified into cellular tained from surgical resections of patients with phospholipids (15%) and triglycerides (4%) but did not uninflammatory bowel disease with exogenous AA leads dergo bxidation. When ['H] 12-hydroxyeicosatetraenoic the to synthesis of these lipoxygenase and acid was incubated withCaco-2cells, 14% underwent two cycles of Paxidation form to ['H]8-hydroxyhexadecacyclooxygenase products, whereas innormal colonic trienoic acid, and 3% underwent three cycles of P-oxidation mucosa AA is esterified into phospholipids and to form ['HI6 hydroxytetradecadienoic acid, both of which triglycerides (9). werereleased into the media. ['H]12-HydroxyeicosatetraeThe Caco-2 cell line is derived from a human noic acid was also esterified into cellular phospholipids colonic adenocarcinoma but functionally and his(13%),but none was esterified into cellular triglycerides.Riehl, T .E . , J. Turk, and W. F. Stenson. Metabolism of tologically resembles human ileal epithelium (10 12). oxygenated derivatives of arachidonic acidby Caco-2 cells. J. The lipid metabolic capabilities of Caco-2 cells include Lipid Res. 1992. 33: 323-331. the uptake of fatty acids and the synthesis and secretion of lipoproteins (11, 13-15) and the synthesis of Supplementary key words: 5-HETE I2-HETE 15-HETE LTB4 cholesteryl esters by acyl CoAcholesterol acyltransinflammatorybowel disease 6hydroxytetradecadienoic acid 8hydroxyhexadecatrienoic acid ferase (12). We demonstrated previously that one lipid mediator of intestinal inflammation, 15-HETE, was metabolized extensively by Caco-2 cells, primarily by fL 12-(S)-hydroxy-5,8, 10, lkicosatetraenoic acid (12oxidation (16). This finding suggested that the intesHETE) is a metabolite of arachidonic acid (AA) tinal epithelium may act to modify intestinal through the 12-lipoxygenase pathway in platelets, inflammation by degrading inflammatory mediators. neutrophils, macrophages, vascular smooth muscle, In inflammatory bowel disease the intestinal vascular endothelium, and erythrocytes (1). 12-HETE epithelium is also exposed to high concentrations of 5plays crucial roles in modulating secretory responses HETE, 12-HETE, and LTB4. In the present study we in pancreatic islet cells and adrenal glomerulosa cells, have examined the metabolism of these mediators by and it possesses chemoattractant activity forhuman Caco-2 cells. polymorphonuclear leukocytes (1). 5-Hydroxy-6, 8, 11, lkicosatetraenoic acid (5-HETE) and leukotriene B4 (LTB4) are metabolites of AA through the 5lipoxygenase pathway in neutrophils, mast cells, and other cell types (2, 3). 5-HETEis a chemotactic and Abbreviations: BSA, bovine serum albumin; DME, Dulbecco's chemokinetic agent for neutrophils and stimulates the modified Eagle's medium; TLCthin-layer chromatography; HETE, hydroxyeicosatetraenoic acid; 142 (BOH), 6hydroxytetrasynthesis of platelet-activating factor in neutrophils decadienoic acid; 16:3 (&OH), &hydroxyhexadecatrienoic acid; A A , (1). LTB4 is a potent chemotactic agent for arachidonic acid; LTB4, leukotriene h;HPLC, high performance neutrophils, eosinophils, and monocytes (4).LTB4is liquid chromatography; PFBE, pentafluorobenzyl ester; GCMS, gas also the major neutrophil chemotactic agent in ulcerachromatography mass spectrometry; NICI, negative ion chemical ionization; ME, methyl ester. tive colitis mucosal extracts (5). Intestinal inflamma-

Downloaded from by guest, on January 14, 2014

Journal of Lipid Research Volume 33, 1992


Heptane extracts containing released radiolabeled free fatty acids weredried under nitrogen. The area contained by the filter and bathing the apical surface of the Caco-2 cells is designated as the “inner well”.0 pm pore polycarbonate membrane filters (Costar). Iowa City. v/v) to resolve methyl esters containing 0 to 3 double bonds.. N Y ) . eluted. 6.MATERIALS AND METHODS Analysis of lipid cell extract The cell homogenate was extracted by the method of Bligh and Dyer (17).and hydroxylatedfatty acids were eluted with 100% by guest. 6.” Cells were maintained in DME medium with 20% fetal calf serum. the medium was collected separately from the inner and outerwells. IL). MO). 6. or Fisher Scientific (St. 15sH112-HETE (143 Ci/mmol) . 9. Department of Medicine. (Lake Placid. 8. 92% hexane-8% ethyl ether. standards were used to identify the lipid classes appearing in each fraction. Chemicals Dulbecco’s modified Eagle’s medium (DME) was obtained from Sigma (St. IL). Inc. Triglycerides were eluted with 95% hexane-5% ethyl ether. Triglycerides and nonhydroxylated fatty acids were eluted with chloroform. on January 14.W). the area bathing the underside of the filter is designated as the “outer well. [5. 10% methanol in chloroform. 9.5 ml) was replaced with an equal volume of medium containingthe radiolabeled fattyacid plus albumin ( 1 mg/ml). Louis. When cells were to be used in a radiolabeling experiment. 8. and these cellswere combined with the first. 11. the medium in the inner well (1. 15 -3H] LTB4 (222 Ci/mmol) were obtainedfrom Amersham (Arlington Heights. and [ 5.2. followed by autoradiography. All other chemicals and solvents werehigh purity commercial materials obtained from Sigma Chemical Co. and solvent system 4 (chloroformmethanol-water80:20:2. After incubating Caco-2 cells with the radiolabeled fatty acid. The remainder of the triglyceride fraction was used to identify and quantify the radiolabeled fatty acid classes esterified in triglycerides. To initiate an experiment. An aliquot of the chloroform extract was separated by thin-layer chromatography (TLC) using solvent system 1 (hexane-ethyl etheracetic acid 80:20:1. 11. and phospholipids were eluted with methanol. 14. 12. nonhydroxylated fattyacidswere eluted with 92% hexane-8% ethyl ether. 9. Cells were harvested by scraping with a spatula. The remainder of the chloroform extract was separated on a 0. 15-3H]5-HETE (210 Ci/mmol). Standards were passed through the column to identify the lipid classes appearing in each fraction. taken up inethyl etherand separated by TLC using solvent system 2 (toluene-ethyl etherethanol-acetic acid 50:40:2:0.8 cm x 17 cmsilicicacid(Bio-Si1 A 100-200 mesh) column eluted sequentially with chloroform. Radioactive bands were scraped from the TLC plate and extracted with chloroform methanol 1:l and counted. the medium was replaced overnight with serum-free medium.jlr. The final cell density was 3 X 1oj cells/cm*. As above. and methanol. The UV visualizing reagent Phenacyl-8 was obtained from Pierce (Rockford. were grown to 12 days post-confluence at 37°C in a 10% CO2 atmosphere (10). C u l t u r i n g of cells Caco-2cells. An aliquot of the triglyceride fraction was subjected to basehydrolysis (16). Aldrich Chemical (Milwaukee. was separated on a similar silicic acid column eluted sequentiallywith 95% hexane-5% ethyl ether. and the fatty acid methyl esters were separated on the basis of the number of carbon-carbon double bonds by argentation TLC using solvent system 3 (chloroform-methan01 99:1. Jeffrey Field. in turn. 2014 Preparation of radiolabeled media and incubation with cells Radiolabeled medium was prepared by incubating the radiolabeled fatty acid in serum-free medium containing fatty acid-free bovine serum albumin (BSA) ( 1 mg/ml) at 37OC for 30 min to allow the fatty acid to bind to albumin. [5. MO) and fetal bovine serum was from Upstate Biotechnology. This procedure allowed for the separation and quantification of the radiolabeled hydroxylated and nonhydroxylated free fatty acids released from triglyceride. hydroxylated fatty acids were eluted with 10% methanol in chloroform. a gift from Dr. v/v/v/v). 14. Louis. 324 Journal of Lipid Research Volume 33. An aliquot of the neutral lipid fraction. Cells were grown as monolayers on 24mm Transwell 3. Bands representing different classes of fatty acids were scraped from the TLC plate. 12. Downloaded from www. University of Iowa. v/v/v). the filter was rinsed with 1 m1 distilled water. 14. v/v/v) to resolvemethylesters contianing 4 to 6 double bonds (18). 8. The cell suspension was kept on ice and was homogenized by sonication. followed by autoradiography. and separated into individual fattyacids by reverse phase HPLC eluted with 80% aqueous acetonitrile (19). and 100% ethyl ether. 12. The triglyceride fraction was subjected to base-methanolysis (16). and the medium in the outer well (3 ml) was replaced with an equal volume of medium containing albumin (1 mg/ml) but without the radiolabeled fatty acid. 11. 1992 . Fatty acids were identified by autoradiography and quantified as described above.

A second aliquot was taken directly for separation on RP-HPLC. After esterification of the carboxylic acid.0% of the counts.. The extract was separated on TLCbysolventsystem 2 to separate hydroxylated from nonhydroxylated fatty acids.d.01 m1 of pyridine. and water was added.02 ml). The mixture was centrifuged at 10.6 mm X 25 cm Beckman Ultrasphere column eluted with a linear gradient from 27 to 100% acetonitrile in water adjusted to pH 2.Carboxylicacidswereconverted either to methyl esters (ME) or to pentafluorobenzyl esters (PFBE). adding 0.and the residue was taken up in 1. The G a s chromatography Samples were introduced into a Hewlett-Packard 5890 gas chromatograph via a Grob-type injector (temperature 250°C) operated in the splitlessmode and were analyzed on either an HP-1 capillary column (cross-linked methylsilicone.31 mm.2 ml).01 m1of N.and the sample was reconstituted in heptane (0. and Stenson Metabolism of oxygenated derivatives of AA by Cat-2 cells 325 .d.02 ml).Additional acetonitrile (2 ml) was added and the mixture was heated at 80°C while stirring for 30 min.05 m1 of tetramethylammonium hydroxide.17 um) interfaced with a Hewlett-Packard 5988 mass spectrometer operated in the negative ion (methane) chemical ionization mode.15 m1 of methanol) and 0.01 m1 of solution B (0.Obis( trimethylsilyl) trifluoroacetamide. 0. The mixture was centrifuged at 1800 rpm for 10 min and the supernatant was dried under a stream of NZ. Turk. Methyl esters were formedby dissolving the analyte in 0. using a C18 analytical column. The supernatant was dried and taken up in 3 m 1 methanol. The sample was then concentrated to dryness under N2. Excess NaHCOs was added to convert hydroxybutyrate to its sodium salt. and incubated 15 min at room temperature.1 m1 of ethereal diazomethane. and incubating 2 min at room temperature.5 min after injection the oven temperature was then Riehl.6 mm x 25 cm Beckman Ultrasphere column eluted with a linear gradient from 27 to 100% acetonitrile in water adjusted to pH by guest. hydroxyl groups wereconvertedtotrimethylsilyl(TMS) ether derivatives.02 m1 of methanol.5 with phosphoric acid (19). 0.An aliquot of the phospholipid fraction from the silicic acid column was subjected to base-methanolysis. An aliquot of the postincubation medium was extracted by the method of Bligh and Dyer (17). Extraction was performed twice with CH2C12 (0.aspreviously described (22).The liberated fatty acids were separated into individual fatty acids by RP-HPLC using a 4. in).1 m 1 of pentafluorobenzyl bromide and 0.2 mm i. i. An excess of NaBH4 was added to reduce any acetoacetate to khydroxybutyrate.005 mmol/ml) in acetonitrile form to the phenacyl esters of hydroxybutyrate and longer chainfattyacids. The organic extract was evaporated under NZ. A d y s i s of lipids in the media phenacyl ester of ghydroxybutyrate eluted at 12 min.0 m1 of phenacyl bromide (0.1 m1of solution A (0. Recovery from Bligh and Dyer extraction followed by R€-HPLC was 72% for HETEs and 84% for palmitate.000 rpm (12. film thickness 0.1 mmol/ml) and crown ether (0. 0. as previously described (23). A second aliquot was subjected to basehydrolysis and the released fatty acids were separated into hydroxylated and nonhydroxylated fractions on a TLC plate withsolventsystem 2.000 g) for 20 min. The methanol was evaporated under a stream of NZ. by dissolving the analyte in 0. 0.3 m 1 of N. At 0. head pressure 4 lb/sq. Samples were concentrated todryness under N2 and reconstituted in heptane (0. Pentafluorobenzyl esters were formed. 2014 Ketone body measurements Absolute ethanol (10 ml) was added to the postincubation medium (2 ml) and left on ice for 30 min.02 ml) . adding 0. Recovery of a standard solution of hydroxybutyrate after derivatization and RP-HPLC was 76%. The protein pellet was counted for radioactivity and always contained less than 1. Helium was the carrier gas (total flow10 ml/min.jlr.8 m1 of N. as previously described (20). A third aliquot was alsosubjected to base hydrolysis. Downloaded from www.Ndimethylacetamide). on January 14.Ndimethylacetamide. by dissolving the analyte in 0. 12m length. An isocratic mobile phase of 40% methanol in water was run at 1. argentation TLC. G a s chromatography mass spectrometry derivatization Before GC-MS analysis. The sample was then concentrated to dryness under N2 and reconstitutedin heptane (0. analytes were converted to suitablederivatives. and the phenacyl esters of 5-HETE and 12-HETE eluted at 72 min. as previously described (21). and R€-HPLCas describedabove from the treatment of the triglyceride extract.5with phosphoric acid. Initial oven temperature was 85°C. A portion of the medium extract was separated by reverse phase HPLC using a 4. The residue was taken up in 2 m 1 methanol and an aliquot was counted.33 micron film thickness) interfaced with a HewlettPackard 5970 mass spectrometer operated in the electron impactmode or anHPUltraperformance capillary column (8 m length. This reaction was allowedto proceed overnight at room temperature. cross-linked methylsilicone.0 ml/min for the first 20 min. followedby a 60-min linear gradient up to 100% methanol. and 0. and incubating for 30 min at 6 O O C .

l B ) . respectively. 22% of the radioactivity was cell-associated. TMS derivatives) or at a rate of 20"C/min to a final temperature of 240°C (for PFBE. there were three new compounds that eluted prior to 12-HETE on RP-HPLC (Fig. Metabolism of ['HI 12-HETE. 1A) and ['H]LTB4. the small amount of cell-associated radioactivity all migrated as unmetabolized ['H] LTB4. The same two metabolites were derivatized to form the pentafluorobenzyl ester (PFBE). The radiolabeled compoundsin the media wereanalyzed byRP-HPLC. and methyl stearate standards. Twoof these 12-HETE metabolites. the only radiolabeled compounds in the media were ["H]5HETE (Fig. with [3H]5-HETE 40% was cell-associated.3 0 9 f3. and accounted for 3.9.In media from conditions in whichcells were incubated with [:jH] 12-HETE. These products eluted at 25.3 0 98 0 0 4 f 1.jlr. RESULTS Radiolabeled compounds in the cells Caco-2 cells actively took up [:jH]12-HETE and [3H]5-HETEbut not r3H]LTBq (Table 1 ) . the peaks eluting at 25 and 29 min (Fig. and methane was used as reagent gas at a source pressure of 1. Fatty acid methyl esters were separated by degree of unsaturation by argentation TLC and in every case all of the radioactivity comigrated with the band of Fatty acid methyl esters containing four carbon-carbon doublebonds(datanot shown). methyl palmitate. Source temperature was 200"C. 2014 Radiolabeled compounds i n the media After 4 hincubation ofCaco-2cellswith [ " H ] 12HETE. (9 nM). L3H]5HETE. trimethylsilyl ether (TMS) derivatives of thecompoundsinthe 25-min peak andthe 29-min peak had elution times reflecting equivalent chain lengths of by guest.+ ( S PM). Both instruments were operated via a Hewlett-Packard RTE-A data system. and 34 min. phospholipids. After 4 h of incubation with [ " H ] 12-HETE. TMS derivatives and analyzed by gas chromatography mass spectroscopy in the negative ion-chemical ionization mode and their retention times were compared with those of the PFBE TABLE 1.increased either at a rate of SO"C/min to a final temperature of 175°C (for analysis of ME.or ['H]LTB. 1B). whereas after incubation with [3H]12-HETE the radioactivity comigrated onlywith phospholipids and hydroxylated free fatty acids. Mass spectrometry The gas chromatograph was interfaced with either a Hewlett-Packard 5970 mass spectrometer (for electron impact analyses) or with a Hewlett-Packard 5988 mass spectrometer (for negative ion-chemical ionization analyses). with ["H]5-HETE 60% of the radioactivity was in the medium. 78% of the radioactivity was in the medium.iHETE. and hydroxylated free fatty acids. Base-methanolysis of cell lipid extracts released radiolabeled fatty acid methyl esters from phospholipids after the 4 h incubation with C3H]12-HETE or ["]5-HETE and also from triglycerides after incubation with ["H]5-HETE. and ionization energy 70 eV for electron impact analyses. 326 Journal of Lipid Research Volume 33.5 torr. and 3% of the total radioactivity.2 0 (trice) 0 60f4.respectively. 14. on January 14.6 0 Caco-2 cells were incubated for 4 h at 37°C in Ihlbecco's modified Eagle's medium(DME)containinghovineserumalbumin (1 mg/ml)and ['H]12HETE ( ~ P M ) ["HISHETE . Both products were methylated and silylated and analyzed by gas chromatography-mass spectrometry in electron-impact mode and their retention times were compared with those of methyl myristate. Downloaded from www. and of two experiments for LTB4. and ["H]LTB4 by Caco-2 cells after 4 h incubation o/u Recovered Radioactivitv 5HETE LT& 12-HETE Media Ketone bodies Hydroxylated fatty acids Polar lipids Cells Free fatty acids Hydroxylated fatty acids Nonhydroxylated fatty acids Triglycerides Hydroxylated fatty acids Nonhydroxylated fatty acids Phospholipids Hydroxylated fatty acids Nonhydroxylated fatty acids 0 78k5.6 0 0 15f2. The radiolabeled fatty acids incorporated into cellular phospholipids were further analyzed by isolating the phospholipid fraction from the total cell lipid extract by silicic acid chromatography and then subjecting the phospholipid fraction to basehydrolysiswith separation of the released fatty acids by RP-HPLC.3 0 21 k 7.. 29. respectively. but with ['H]LTB4 only 2%.and with [JH]LTB4 98% remainedinthe medium(Table 1). After incubation with [3H]5-HETE the cell-associated radioactivity migrated with triglycerides. Values represent the mean ? SEM of three experiments for 12-HETE and .5 2 0 0 0 0 0 0 13f2.6 and 17. 1992 . This analysis demonstrated that effectively all of the radiolabeled fatty acid incorporated into cellular phospholipids was unaltered ["]15-HETE or ["]12HETE (data not shown). were further analyzed. TMS derivatives). The methyl ester (ME). After incubation with ["H]LTB4. Source temperature was 100°C for negative ion-chemical ionization analyses. After 4 h incubation with ["H]5-HETE or ["H]LTBs.

LTB4. 2B) and are formed from the molecular ion by loss of CH3 and by loss of CH3 plus CH30H. Themolecular ion (m/z 326) is not seen but the ion m/z 311 formed by the molecular ion loss of CH3 is clearly visualized.9. The amount of radioactivity in 12-HETE in the media decreased progressively over the 4 h period whereas the amount of radioactivity in 16:3(8-OH) in the media increased progressively as did the amount of radioactivity in the cells (Fig. is not actively taken up or metabolized by Caco-2 cells.8. 25) for the ME. and 4 h. Downloaded from www.12. In addition. Reverse ohase HPLC elution Drofiles of fattv acids conL tained in the media after a 4 h incubat’ion at 37°C ofCaco-2 cells in DME media containin bovine serum albumin ( 1 mg/ml) and 9. l2HETE at 39 min.12. Themolecular ion (m/z 352) is not formed in high abundance. At the end of the incubations the medium and the cells were extracted separately and the amount of radioactivity in each was assessed. The electronimpact mass spectrum of the (ME. The third eicosanoid studied. Caco-2 cell monolayers were incubated with [3H]12HETE for 15 min. 10)-trienoic acid [16:3(8-OH)].8.5 with phosphoric acid) over 50 by guest. Turk.6. as previously reported (24. 25).MINUTES 1. 337) is generated by loss of the pentafluorobenzyl moeity and reflects the carboxylate anion of the TMS derivative of the parent compound. Thebase ion in this spectrum ( m / . Lesser amounts of 15-HETE . TMS) derivative of the metabolite eluting at 25 min is displayed in Fig. was assessed. and is characteristic of the methane negative ionchemical ionization mass spectra of fatty acid derivatives (20-23). arachidonic acid at 48 min. The base ion in this spectrum (m/z 311) is generated by the loss of the pentafluorobenzyl moiety and reflects the carboxylate ion of the TMS derivativeof the parent compound. The base ion m/z 241 reflects fragmentation between carbons 8 and 9 and loss of C H S ( C H ~ ) ~ ( C H ) ~ moiety. and palmitate at 53 min.4 and is similar to that reported for the ME. 4). The methane negative ionchemical ionization mass spectrum of the (PFBE. 2 h.8)dienoic acid [14:2(6OH)] (24).11.15(A). The electron impact mass spectrum of the (ME.11. TMS) derivative of this metabolite is displayed in Fig. whereas that of the compound in the 28min peak was between these of 16:O and 18:O standards. derivatives of the fatty acids listed above. Riehl. TMS) derivative of the metabolite eluting at 29 min is displayed in Fig. TMS) derivative of this metabolite is displayed in Fig. The base ion m/z 215 reflects fragmentation between carbons 6 and 7 and lossof the C H S ( C H ~ ) ~ ( C H ) moiety ~ C H ~ as illustrated in the structural diagram inFig. Media were removed and extracted. 3 and supports the identification of the metabolite as 16:3(8-0H).TMS derivative of the 12-HETE metabolite 8-hydroxy-hexadeca-( 4.Hj5-hydroxyeicosatetraenoic acid (3 nM) either [5.6. 2.14. The methane negative ion-chemical ionization mass spectrum of the (PFBE. respectively. 5-HETE at 41 min. and Stenson Metabolism of oxygenatedderivatives of AA by Caco-2 cells 327 . The elution time of the derivative of the compound in the 25-min peak was between those of 14:O and 16:O standards.15-H]12-hydroxyeicosatetraenoic acid (6 PM) (B). or [5. 1 h. 3A and supports the identification as 14:2(6OH). 305 are clearlyvisualized (Fig. 2014 DISCUSSION The present study demonstratesthat Caco-2cells take up 12-HETE and metabolize it by esterification into cellular phospholipids and by poxidation. on January 14. 12-HETE and 16:3(8-OH). In this HPLC system LTB4 elutes at 26 min.jlr. We previously reported that Caco-2 cells metabolized 15-HETE primarily by bxidation with each molecule going through at least eight cycles of poxidation. and the lipid extracts were separated by reverse phase HPLC eluted (with a linear gradient from 27 to 100% acetonitrile in water adjusted to pH 2. 5-HETEis taken up and esterified into cellular phospholipids and triglycerides but is not bxidized. 2B and is essentially identical to that reported (24. C H ~ as illustrated in the structural diagram in Fig.6. the lipids in the extract of the medium were separated by HPLC and the amount of radioactivity in the two major peaks. but the ions m/z 337 and m / . Incorporation of12-HETE into cellular phospholipids also increased progressively over the 4 h period (data not shown).9.14. In contrast. TMS derivativeof the 12-HETE metabolite 6hydroxytetradeca-(4.were esterified into triglycerides and phospholipids (16). 3B. 2A.

however in macrophages some 12-HETE underwent additional Poxidation to form 12:1(4OH). one exception is murine cerebral endothelial cells in which not only 12-HETE but also some 16:3(8-0H) is esterfied into cell phospholipids (26). platelets. murine cerebral microvascular en- dothelium (26). Why does the k x i d a tion of 12-HETE largelystop after two cycles? 328 Journal of Lipid Research Volume 33. 5-HETE did not undergo Poxidation. TMS) derivatives of two ofthemetabolitesof 12-HETE formed by Caco-2 cells. 12-HETE also undergoes Poxidation to form 16:3(8-0H) in human umbilical vein smooth muscle cells (25). MDCK cells (24). 1B) were obtained on a Hewlett-Packard 5970 mass spectrometer with an ionization energy of 70 eV at a source temperature of 200°C as described in the Methods section.One metabolic fate of 5-. Why are 12-HETE and 15-HETE substrates for b x i d a tion where as 5-HETE is not? 2. and 15-HETE are esterified into phospholipids and 5. Macrophages. Electronimpactmassspectra of the (ME. Esterification of these compounds into phospholipids and triglycerides has been reported previously in neutrophils. MDCK cells.(CH3) = 31 1 31l M = 352 M"(CH3) = 337 M"(CH3 + CHJOH) = 305 l 360 380 400 2 0 260 280 300 320 340 41a 2 395 337 260 280 300 320 340 3 10 MASS TO CHARGE RATIO (m/z) MASS TO CHARGE RATIO ( d z ) Fig. macrophages. 12-HETE underwent two tothree cyclesof poxidation. 2. 1992 . and 15-HETE by Caco-2cellsraise two questions. A detailed investigation of 12-HETE metabolism in mouse peritoneal macrophages revealed that bxidation is a major metabolic pathway (28).m12 241 Downloaded from www. 16:3(&OH) is released into the medium. Studies in other cell systems have revealed considerable variety in the patterns of uptake and metabolism of 5-.and 15-HETE are esterified into triglycerides (16). A second potential metabolic fate for the HETEs is poxidation. the product of four cycles of poxidation. on January 14. and normal skin fibroblasts (27). 2014 84 M = 326 M . and HPLC 29-min elution peak (B) (see Fig. In most of these cell types. The product of two cycles of 12-HETE Poxidation in Caco-2 cells is 16:3(&OH) and the product of three cyclesis 14:2(&0H).org by guest.and 15-HETE underwent at least eight cycles.B 137 I 241 I 80 I 80 o o 40 H r l55 l ll ll l I 60 40 20 0 a 80 120 l60 200 240 c r . 12-. as in Caco-2 cells. The electron impact mass spectra of HPLC elution 25-minpeak (A). and 15-HETEsis esterification into triglycerides and phospholipids. mast cells. like Caco-2 cells.jlr. and 15-HETEs (1). The results of the studies of Poxidation of 5-. In Caco-2 cells 5-. and vascular endothelial cells (1). In Caco-2 cells. 1. released large amounts of 16:3(8-OH) and small amounts of 14:2(6OH) into the media. 12-. 12-. 12-. demonstrating that much of the @oxidation stopped after two or three cycles.

in the studyof 12-HETE metabolism by mouse macrophages mentioned above. occurs in peroxisomes. Methane negative ion chemical ionization mass spectra of the (PFBE.and 15-HETEsuggests that they contain peroxisomes. Turk. and in 15-HETE at carbons . 1B) were obtained on a Hewlett-Packard 5988 mass spectrometer with a methane source pressure of 1. Products of mitochondrial oxidation are tightly bound within the mitochondria whereas release of products during chain shortening is characteristic or peroxisomal oxidation (28). 2014 I 220 240 260 200 300 320 340 360 MASS TO CHARGE RATIO (mlz) Fig.11.jlr. and Stenson Metabolism of oxygenated derivatives of M by Caco-2 cells 329 . Figard. Gordon. on January 14. trans diene system that occurs in 5-HETE at carbons 6. 8. 31). bxidation of very long chain saturated fatty acids and longchainunsaturated fatty acids is begun in peroxisomes. The ability of HETEs to undergo bxidation is proportional not only to the distance of their hydroxyl groups from the carboxyl terminus but also to the distance of their conjugated double bonds from the carboxyl terminus. In addition. The ability of Caco-2 cells to p oxidize 12. trans double bonds on peroxisomal oxidation has not been Riehl. The hydroxyl groups on 12. and presumably also the initial oxidation of ISHETE. The release of 16:3(&OH) into the media is more consistent with peroxisomal than mitochondrial oxidation. however the effect of conjugated cis. These studies suggest that the Boxidation of 12-HETE. Previous studies have demonstratedthatthe rates of peroxisomal oxidation are influenced by the degree of unsaturation of the fatty acid substrate and by the geometric configuration of the double bonds (30. and Spector (27) demonstrated that skin fibroblasts from a patient with Zellwegers syndrome did not metabolize 12-HETE whereas fibroblasts from normal individuals bxidized 12-HETE to 16:3(80H). Zellwegers syndrome is an autosomal dominant disorder characterized by the a b sence of peroxisomes. 3.OSi (CH3I3 1 0 0- 3 I 7 331 Y z a 3 z m 00- Y 3 z 3 a W go60- 2 60U ‘“1 20 2 40- 203 221 I I 236 I 255 I 0 Il . The methane negative ion chemical ionization massspectra of HPLC elution 25-minpeak (A).org by guest. Why is 5-HETE not a substrate for bxidation? The absence of 5-HETE P+xidation in Caco-2 cells suggests that 5-HETE CoA is not a good substrate for peroxisomal acyl-CoA oxidase. The hydroxyl group on the k a r b o n may interfere with thebinding of 5HETE-CoA to peroxisomal acyl-CoA oxidase. 1 I I 3 2 20-1247 r” MASS TO CHARGE RATIO (NZ) Downloaded from www. The presence of peroxisomes in Caco-2 cells is not unexpected since they are found in abundance in guinea pig and rat enterocytes (29). the other distinguishing characteristics of the HETEs is their conjugated cis.and 15-HETE are further from the carboxyl group and may not interfere with the interaction with the oxidase. TMS) derivatives of two of the metabolites of 12-HETE formed byCaco-2 cells. in 12-HETE at carbons 8. In addition to the hydroxyl groups. bxidation was only minimallyinhibited by methyl palmoxirate. 13. After the long chain fatty acidsare chainshortened they aretransferred to mitochondriafor further Boxidation.5 torr and a source temperature of 100°C as described in the Methods section. and29-minpeak (B) (see Fig. 10. an inhibitor of mitochondrial bxidation.

12-. I I 2 / 0 HOURS F i g . In human neutrophils and macrophages. 2014 REFERENCES 1.Caco-2 cells were incubatedfor up to 4 h at37°CinDME media containing bovine serum albumin ( I mg/ml) and [5.14. Inability to form 16:3(&OH)-CoA would also explain the absence of 16:3(8-OH) esterification into triglycerides and phospholipids and the release of free 16:3(&OH) intothe media. W. The medium-chain acyl residues formed in peroxisomes can be transferred to carnitine by peroxisomal medium-chain carnitine acyltransferase and the resulting acylcarnitine moved to mitochondria for further oxidation. Parker. Doig. This suggests that 15-HETE is a substrate for peroxisomal oxidation and that the chain-shortened product is able to enter the mitochrondria.However.15~H]12hydroxyeicosatetraenoic acid (6 PM). M. 76: 21482152. and S. Sci. Arachidonic acid metabolism in polymorphonuclear leukoctyes: effectsof ionophore A23187. 5. Although 5-. on January 14. Bray. In addition. which lack the abundant low affinity LTB4 receptors found in human neutrophils. 16:3(&OH) may be a poor substrate for peroxisomal medium-chain carnitine acyl transferase. the less polar HETEs are able to cross the plasma membrane by permeation. trans double bonds interfere with further oxidation. Peroxisomes can only chain-shorten long-chain fatty acids to medium chain acyl-CoA thioesters. 1990. USA. 5-. 1992 . V. M. and 15-HETE and LTB4 are present in high concentrations in the intestinal mucosa in inflammatory bowel disease and also in ischemic colitis and other intestinal diseases marked by inflammation (5). Downloaded from www. Samuelsson. LTB4 remained almost totally unmetabolized in the media even at 4 by guest. In The 330 Journal of Lipid Research Volume 33. Manuscript received l Februaly1991 and i n raised f o n n 20 Noumber 1991.12. Leukotriene B. specifically addressed. F. Moore. the 16:3(8-OH) released into the cytosol may be a poor substrate for acyl-CoAsynthetase as theformation of 16:3(8-OH)-CoA in the cytoplasm would also allow for mitochondrial b x i d a tion. Cellularuptake and metabolismof 12-HETE. and B. Shipley.jlr. E. 1988. Thus the inability of Caco-2 cells to take up and metabolize LTB4 may relate to an absence of LTB4 receptors. Prog.11. 15-HETE undergoes at least eight cycles of Poxidation Caco-2 cells. 30: 175-199. Why does 12-HETE Poxidationstop after two or three cycles? The termination of poxidation after the formation of 14:2(6OH) or 16:3(&OH) again raises the possibility thatthe hydroxyl grouporthe conjugated cis. Rat neutrophils.areunableto take upand metabolize LTB4 (33). W. F. 1979. Thereareother possible explanations for the differences in the metabolic fates of the different HETEs. These cells contain LTB4 receptors and it is likely that LTB4 enters by a recep tor-mediated process. the most important of these lipid mediators of inflammation. W. Med. Stenson. 4. 4. Natl. Gordon.6. ( A ) 12-HETEin media. Borgeat. 1984. Znlern.. Adu.W. Alternatively. Smith.9. these cells cannot take up and metabolize LTB4. P. M. and C .. ( 0 ) 16:3(8-0H) in media. Acad. F. Spector. In contrast. A. J. Hydroxyeicosatetraenoic acids (HETEs). The greater distance of both the hydroxyl group and the conjugated double bond from the carboxyl terminus in 15-HETE than 12HETE may explain the greater @oxidation of 15HETE. 266: 264-265. 12-. A. A. chemotacticandaggregatingsubstancereleasedfrom polymorphonuclear leukocytes. 1980. This work was supported by NIH grants DK33165 (WFS) and DK34388 UT).8. 27: 271-323. a potent 3. Data are presented as nmol/filter: ( 0 ) cellular accumulation.‘*L . LTB4is taken up and undergoes P-oxidation to form first 20-OH LTB4 and then 20-COOH LTB4 (32). Leukotrienes. Differential competition for the formation of CoA derivatives or acylation reactions are also plausible explanations. A. Lipid Res. and J. Ford-Hutchinson. Nature. Eicosanoids in inflammatory bowel disease with special reference to leukotriene B4. A. 2. Proc. A. Stenson. and 15-HETE were all metabolized by Caco-2 cells.. The findings described here demonstrate that enterocytes are capable of down-regulating the inflammatory response by metabolizing 5-HETE and 12-HETEjust as an earlier study demonstrated the metabolism of 15-HETE by Caco-2cells. it may be that there is diffcdty intransferring the 16:3(8-OH)fromthe peroxisomes to the mitochondria.

Lacroix.. Biol.R. Turk.A. M. Biol. Kayden. E. J. 1984. JSaduce. 32. M. and W. Hart.. Lipid Res. G. Kotagal. C h . Biochim. E. M. R. Morris. 33. Effect of eicosapentaenoic acid on triacylglycerol Biochim. Philadelphia. E. 31. Traber. Polarized secretion of newly synthesized lipoproteins by the Cace2 human intestinal cell line. and J. A. 10..jlr. 262: 3762-3767. J. Hull.W. S. and E. Lipid Res. 794 110-1 24. T. H. Kreisle. Lipid Res.Cell Biology of Inflammation in the Gastro-Intestinal Tract. Simon-Assman. Separation and quantitation of free fatty acids and fatty acid methyl esters by reverse phase high pressure liquid chromatography. McDaniel. 1987. Hughes. Zweibaum. P. Turk.1959. L. R. 1990. 1972. M. L. N. and A. Prokuski. 1983. Kluwer Academic Publishers. Fogh. 28: 1350-1 363. J. and J. 1987. Phamacol. Clin. R. Biochem. J. 18. N. A.. F. and M. Clin. 32: 293-304. 1984. and A. 1985. 28.. Identification and quantitation of lipoxygenase and cyclooxygenase products. Can. Field. Easom. V. 17. 28: 1057-1066. T. Mathur.. 1989. Figard. 29.. B. Hadjiagapiou. S. P. J. Peroxisomal & oxidation of long-chain fatty acids possessing different extents of unsaturation. 1989. and M. editor. Storch. A. Lipid Res. Hughes. Biochm. J. G. Peroxisomes in absorptive cells of mammalian small intestine. 0. Wolf. Trotter. E. A. and M. and A. Zimmerman. J. editors. J. J.A. Wolf. P. L. Lipid Res. C. H. J. A. B. Bligh. Biophys. K. Spector. 1045 transport inCaco-2cells.Animal models of inflammatory bowel disease. Acta.. and A. R. I. 2014 Riehl.. P. J. L. Stenson. 1988. Lamon-Fava.Appay. 26. 24. M. 2 4 83-93. Pinto. Ordovas. N. Biochmist?y. M. 1966. Kirsner and R. Physiol. Fatty acid oxidation and its by guest. J. In Inflammatory Bowel Diseases. Schaefer. A. 16. 273-281. Arachidonic acid metabolism in isolated pancreatic islets. J. McDaniel. and A. D. N.B.J. J. A.. Acta. 25.D. J. M. 30.. Lipzd Res. J. T. Parker. 1988.... and F. Turk. Osmundsen. 22. 12. J. J. 134: 33563363. Turk. S. 85: 1173-1 181. R.J. McDaniel. Acta. Turk. Enhanced synthesis of leukotriene B4 by colonic mucosain inflammatory bowel disease. McDaniel. Biol. E. 1987. and S. Field. P. F. 24: 11761182.. Easom. J. A. 1990.P. 13. Sharon. Hydroxyeicosatetraenoic acid metabolism in cultured human skin fibroblasts. Downloaded from www. Colcan. S. J. 1983. Numa. 3rd edition. J. chromatography. T. Metabolism of 15hydroxyeicosatetraenoic acid by Caco-2 cells. W. S. E. UK. N. J. 11. J. Hovik. P. Prostaglandins. Albright. L. 30: 731-738. N. on January 14. Biophys. 1001: 1624. 20. Rachmilewitz and J. A novel cell line (Caco-2) for the study ofintestinal lipoprotein synthesis. Immunol. J. Whittle.. and J. Stenson. Murthy. Murine cerebral microvascular endothelium incoporate and metabolize 12-hydroxyeicosatetraenoic acid. Elsevier Science Publishers B. A. editors. E. M. Forte.J. C. Regulation of cholesterol esterification by micellar cholesterol in Caco-2 cel1s. Haffen.Wolf. Mass spectrometric characterization of the fattyacyl content indicates multiple mechanisms of generation. Griffin. A. L. Invest. H. A. A. A rapid method of total lipid extraction and purification. and W. M. C h .G. Boughton-Smith. L. Robine-Leon. Stenson. Gordon. Invest. 97-164. Dordrecht. J. Horrocks. 34: 579-589. B. V. Figard. 23. In Fatty Acid Metabolism and its Regulation. 1991. Van Rollins. 1983. 47: 323330. Gastroenterology.C.1990.. Sasak. B i o c h . Separation of lipids by silver ion ' : 717-732. and L. 147-155. Diacylglycerol synthesis de novo from glucose by pancreatic islets from rats and humans.. 1990. Rindler. Peters. J. J. 137: 75-85. Senior. Stenson.J. 7 19. Bioibhys. B. 14. I. Physiol. Kedinger. J. Amsterdam. The enantiomeric composiacid indition of 12-hydroxy-5. F. Novikoff. Cut. 31: 773-780.8. Formation of 8 hydroxyhexadecatrienoic acid by vascular smooth muscle cells. and M. T. Mathur. 27. Fatty acid uptake and metabolism in a human intestinal cell line (Caco-2): comparison of apical and basolateral incubation.12Hydroxyeicosatetraenoic acid is metabolized by P-oxidation in mouse peritoneal macrophages. J. Can. S. Physiol. Cell. Identification of the major metabolite of 12-HETE produced by renal tubular epithelial cells. 15. Albright. N.A. Evidence for peroxisomal & 0xidation. Novikoff. C. 1990. L. K. In Inflammatory Bowel Disease. M. G. 37: 911-917. Spector. and W. Spector. andH. 247: 531-535.J. Figard. A. Easom. 28: 4291-4301. Cell Biol. S.. Bremer. Spector. 86: 453-460. and B. F. 67: 936942. J. Shorter. and F. Biochemistry ofthe lipoxygenase pathways in neutrophils. M. and W. R. N. Dyer. B. H. Riehl. 53: 532-560. Bass. 113-154.". Lea and Febiger. Triadon. editor. H. Corners Publications... H. F.10.. P. 1989. 265: 21048-21055. Osmundsen.J. Elson.Biosynthesis of lipoxygenase and cyclc-oxygenase products from '*C-arachidonicacid by human colonic mucosa.J. 1990. Field. V. Sprecher. The immunology of inflammatory bowel disease. 21. 71-78. Enterocytelike differentiation and polarization of the human colon carcinoma cell line Caco-2 in culture. M. Hawkey. W. and H. Aveldano. 85: 482-490.. Dussaulx. J... Gordon. H. Arachidonic acid metabolism in isolated pan- creatic islets.14icosatetraenoic cates synthesis by a 12-lipoxygenase rather than a monooxygenase. Mathur. J. Borgeat.. J. Lipd Res. and Stenson Metabolism of oxygenatedderivatives of AA by Caco-2 cells 331 . 1987. 1989. Secretagogue-induced diacylglycerol accumulation in isolated pancreatic islets. Hughes. Studies of leukotriene B 4 specific binding and function in rat polymorphonuclear leukocytes: absence of a chemotactic response. The Netherlands. 1987. P. Cell. H. 1984. R. Albright. J. Moore. Figard.