JOURNAL OF CELLULAR PHYSIOLOGY 195:92–98 (2003)

Cyclooxygenase and Cytochrome P-450 PathwaysInduced by Fetal Calf Serum Regulate Wound Closure

in 3T6 Fibroblast Cultures Through the Effect ofProstaglandin E2 and 12 and 20Hydroxyeicosatetraenoic Acids

JUAN JOSE MORENO*

Department of Physiology, Faculty of Pharmacy, Barcelona University,Barcelona, Spain

Wound-induced injury of 3T6 fibroblast cultures initiated a repair process stimu-lated by fetal calf serum (FCS) that restored the integrity of cell cultures. In theseexperimental conditions, FCS induced arachidonic acid (AA) release andeicosanoid production. Our results show that the inhibition of the cyclooxygenase(COX) and/or cytochrome P-450 pathways significantly decreases the woundclosure, whereas that of the lipoxygenase pathway does not modify the woundrepair process. Both EP1 and EP4 receptors of prostaglandin E2 (PGE2)mediate PGE2stimulated 3T6 fibroblast wound closure. Our data suggest that calcium and cAMPare involved in the signaling event induced by PGE2 during the 3T6 fibroblastwound repair process. On the other hand, we show that ketoconazole, a cyto-chrome P-450 inhibitor, hinders the wound closure induced by FCS in wounded3T6 fibroblast cultures. 12 and 20 Hydroxyeicosatetraenoic acids (HETEs), whicharekeyAAmetabolites synthesizedbycytochromeP-450,partially revert the effectsof ketoconazole on the wound repair process. Thus, the COX and cytochromeP-450 pathways of the arachidonate cascade are involved in 3T6 fibroblast woundclosure. J. Cell. Physiol. 195: 92–98, 2003. � 2003 Wiley-Liss, Inc.

Arachidonic acid (AA) which is predominantly este-rified in the phospholipids of membranes is mobilized byphospholipases and oxidized via three major metabolicpathways: the cyclooxygenase (COX) pathway, whichproduces prostaglandins and thromboxanes; the lipox-ygenase pathway, which forms leukotrienes, midchainhydroxyeicosatetraenoic acids (HETEs), and lipoxins;and the cytochrome P-450 monooxygenase pathway,which synthesizes HETEs and epoxyeicosatetrienoicacids (EETs). Thus, AA is the precursor of a large num-ber of biologically active molecules collectively namedeicosanoids (Needleman et al., 1986). These metabolicpathways play a key role on wound healing, fibrosis,rheumatoid arthritis, and similar physiological andpathological processes. However, the effect of eicosa-noids, on these processes is unclear.

The wound healing process involves the movement ofneighboring cells to cover the injury area and subse-quent cell proliferation. Cell spreading and migrationmay be initiated by the AA metabolites of lipoxygenaseand cytochrome P-450 pathways that release leuko-trienes and HETEs, which are the most potent attrac-tants to fibroblasts (Rieger et al., 1990). Thus, the cellssurrounding the damaged area migrate to re-establishculture integrity. However, cellular proliferation intothe denuded surface is necessary to complete the re-pair process. This step may be regulated by external

elements such as growth factors, extracellular matrixcomponents, and eicosanoids. Growth factors exert theirmitogenic effect by coupling with specific high-affinitycell-surface receptors, which initiate a variety of signaltransduction pathways like the AA cascade (Domin andRozengurt, 1993). Thus, AA release and prostaglandinE2 (PGE2) production are enhanced in wounded fibro-blast cultures and may be involved in the regulation of3T6 fibroblast cultures wound repair (Moreno, 1997).Prostanoids are involved in the wound closure process ina broad variety of tissues including corneal endothelium(Joyce and Meklir, 1994), keratinocytes (Kaneko et al.,

� 2003 WILEY-LISS, INC.

Contract grant sponsor: Spanish Ministry of Education; Contractgrant number: PM98-0191; Contract grant sponsor: Generalitatde Catalunya; Contract grant number: 1999SGR00266.

*Correspondence to: Juan Jose Moreno, Department of Physiol-ogy, Faculty of Pharmacy, Barcelona University, Avda. JoanXXIII s/n, E-08028 Barcelona, Spain.E-mail: moreno@farmacia.far.ub.es

Received 4 June 2002; Accepted 6 November 2002

DOI: 10.1002/jcp.10226

1995), skin (Talwar et al., 1996), and intestinal epithe-lium (Zushi et al., 1996). Furthermore, PGE2 and otherprostanoids have also been associated with woundclosure induced by growth factors such as epidermalgrowth factor (Zushi et al., 1996), hepatocyte growthfactor (Takahashi et al., 1996), and transforming growthfactor-b (Zushi et al., 1996).

PGE2 is one of the major eicosanoid synthesizedby fibroblasts and secreted into the culture medium(Hyman et al., 1982). Prostanoid effects are initiated onplasma membrane localized GTP-binding protein-coupled receptors (Coleman et al., 1994a). The eighttypes and sub-types of prostanoid receptors, de-fined pharmacologically, have recently been cloned(Narumiya et al., 1999). Among them, the four types ofPGE2 receptors are encoded by distinct genes. Thebinding of PGE2 to its receptors alters the levels ofthe second messengers like intracellular calcium con-centration (i[Ca2þ ]) (Yamashita and Takai, 1987) andcAMP (Rozengurt et al., 1983) that mediate the cellularresponse of prostanoids. On the other hand, HETEsaffect intracellular calcium concentration (Amlal et al.,1996) and the activity of kinases such as calcium/calmodulin kinase II (Muthalif et al., 1998) and MAPkinase (Sun et al., 1999).

The objective of this study was to determine the roleof prostanoids, lipoxygenase, and cytochrome P-450pathway metabolites in the regulation of wound closureusing wounded confluent 3T6 fibroblast cultures.

MATERIALS AND METHODSMaterial

[Methyl-3H]thymidine (20 Ci/mmol) was obtainedfrom Du Pont-New England Nuclear (Boston, MA). Cellculture medium RPMI 1640, fetal calf serum (FCS),trypsin/EDTA, penicillin-G, and streptomycin weresupplied by Life Technologies (Gaithersburg, MD).Ketoprofen, ketoconazole, forskolin, thapsigargin, bai-calein, nordihydroguaiaretic acid (NDGA), Fura-II/AM,ethidium bromide, and acridine orange were from Sigma(St. Louis, MO). AH6809 (6-isopropoxy-9-oxoxanthene-2-carboxylic acid) and AH23848B ([1 a (z), 2b, 5 a]-(�)-7-[5-[[(1,10-biphenyl-4-yl]metoxy]-2-(4-morphonilyl)-3-oxocyclopentil]-4-heptenoic acid) were kindly provid-ed by Glaxo Wellcome Medicines Research Centre(Stevenage Hertfordshire, UK). PGE2, 12 (R)-HETE,and 20-HETE were obtained from Cayman ChemicalCo. (Ann Arbor, MI). Compounds were dissolved indimethylsulfoxide and diluted in medium to keep thefinal concentration of dimethylsulfoxide below 0.1%. Alltreatments were used to non-cytotoxic concentrationsassayed by ethidium bromide/acridine orange staining.

In vitro wound closure model

Murine Swiss 3T6 fibroblasts (CCL 96) from AmericanType Culture Collection (Manassas, VA) were routinelymaintained inRPMI1640supplementedwithFCS(10%)and antibiotics (penicillin 100 U/ml and streptomycin100 mg/ml). For experimental use, cells were removedby treatment with trypsin/EDTA and seeded on 13-mmround Thermanox tissue culture coverslips (MilesLaboratories, Naperville, IL), which were placed into12-well plates (Costar, Cambridge, MD), and grown toconfluence in RPMI 1640 medium supplemented with

antibiotics and 10% FCS. Then, 48 h before wounding,cultures were transferred to medium supplementedwith FCS at 2.5% instead of 10% to decrease thepotential influence of serum-derived factors on woundrepair responses while maintaining the health andstability of the cultures. Thereafter, two 2-mm-widewounds were made radially in a cell monolayer using asterile Teflon rake. The wound area affected 50 mm2 ofcell surface (approximately 20% of the total surface)Finally, wounded monolayers on coverslip were incu-bated in RPMI 1640 containing 10% FCS and the testsubstances at 378C in a humidified chamber supple-mented with 95% air/5% CO2 and the regeneration of thecell monolayer was then assessed.

Measurement of the wound area

Images of the wounds were collected at specific timeswith a Nikon Diaphot 300 inverted microscope equipp-ed with an integrating charge-coupled device cameraand a digital image processor. The wounded area wasmeasured using a computer-assisted image analysissystem (Soltaire, Seescan, VIDS III, Analytical Measur-ing Systems). Images were obtained at the initial time ofwounding and at various times after wounding. Imagesof the wounds in all wells were obtained at three pointsin each well; measurements from these two images wereaveraged, and the results are presented as a simple datapoint for that experiment and expressed as % recoveryusing the formula: % recovery¼ (1�wound area attx/wound area at to)�100%, where tx is the number ofhours post-injury and to is time of injury. At the end ofthe experiments with inhibitors or antagonists, cellswere stained with ethidium bromide/acridine orange inorder to assess cell viability.

[3H]Thymidine incorporation assay

DNA synthesis was measured as described elsewhere(Martınez et al., 1997). In vitro wound cultures wereincubated with the treatments in the presence of[3H]thymidine (1 mCi/well). At the end of the experi-ments [3H]thymidine-containing media were aspirated,cells were overlaid with 1% Triton X-100, and scrapedoff the coverslips, and the radioactivity present in thecell fraction was then measured by liquid scintillationcounting.

cAMP levels

cAMP from cell cultures was extracted with ethanoland measured as by suggested the manufacturer of theenzyme immunoassay kit (Amersham Life Science,Buckinghamshire, UK).

Protein determination

Total protein was measured by the Bradford method(Bradford, 1976) by means of the Bio-Rad detergent-compatible protein assay (Bio-Rad, Hercules, CA), usingBSA as standard.

Measurement of intracellular calciumconcentrations

Cells grown on glass slices were incubated with Fura-II/AM (3 mM) and pluronic F-127 (0.02% v/v) in RPMIat 378C for 45 min. Fluorescence was monitored on aspectrofluorometer with dual excitation wavelengths of

WOUND CLOSURE AND EICOSANOIDS 93

340 and 380 nm and an emission wavelength of 520 nm.i[Ca2þ] was calculated as described by Grynkiewicz et al.(1985).

Statistical analysis

For the cell culture experiments, data are expressedas the mean� standard error of the mean (SEM). Allexperiments were performed at least three times toensure consistency of the observations. The significanceof differences between data points and controls wasdetermined using a two-tailed Student’s t-test withP<0.05 confidence limits.

RESULTSInhibition of COX and cytochrome P-450pathways limit wound closure induced

by FCS in 3T6 fibroblast cultures

Based on previous reports that AA contribute towound closure of 3T6 fibroblast cultures, and assumingthat AA is metabolized mainly by three pathways: COX,lipoxygenase, and cytochrome P-450, we suggest thatAA metabolites produced by these pathways regulatewound healing in 3T6 cultures. To confirm this hypoth-esis, we inhibited the COX pathway with ketoprofen(Moreno et al., 1990), the lipoxygenase pathway withbaicalein (Sekiya and Okuda, 1982) or NDGA (Hopeet al., 1983; Argentieri et al., 1994), and the cytochromeP-450 pathway with ketoconazole (Maurice et al., 1992).Ketoprofen significantly inhibited wound closure in adose-dependent manner, reaching a plateau at 5 mM,when the percentage of recovery of wound closure was50–60% (Fig. 1). Ketoconazole also inhibited wound

closure by up to 50–60%. Moreover, ketoprofen andketoconozole partially reduced the time-course of theclosure (Fig. 2). When ketoprofen and ketoconazolewere used simultaneously, wound repair was abolished(Fig. 2). These results suggest that both branches of theAA cascade are involved in 3T6 fibroblast wound repairprocess. As several authors have proposed that lipox-ygenase metabolites are required for wound closure(Nakamura et al., 1994; Kobayashi and Arakawa, 1995),we examined the influence of this AA pathway on 3T6fibroblast wound repair. Baicalein and NDGA, lipox-ygenase inhibitors, did not significantly inhibit thewound closure (Fig. 1).

Wound closure was correlated with cell growth andan increase in [3H]thymidine incorporation to cells, anindicator of DNA synthesis (Table 1). To determinewhether DNA synthesis was enhanced by AA or itsmetabolites, the effect of ketoprofen and ketoconazole on[3H]thymidine incorporation was evaluated. The keto-profen and ketoconazole treatments that inhibitedwound closure also decreased DNA synthesis in wound-ed 3T6 fibroblast cultures (Table 1). Interestingly, theimpairment of [3H]thymidine incorporation inducedby ketoprofen and ketoconazole was partially revertedwhen we exogenously added PGE2 and HETEs(12 HETE and 20 HETE) to the culture medium,respectively.

PGE2 and 12, 20 HETEs stimulate woundclosure of 3T6 fibroblast cultures

PGE2, added to cell cultures, reduced the inhibitoryeffect of ketoprofen on wound repair (Table 1), especially

Fig. 1. Dose-dependent effect of arachidonic acid (AA) metabolisminhibitors on wound recovery of 3T6 fibroblast injured culturesinduced by fetal calf serum (FCS). Wounded cultures were incubatedin RPMI 1640 with 10% FCS in presence of ketoprofen (&),ketoconazole (*), NDGA (*), or baicalein (~) for 12 h. The percent-age of wound recovery was then determined. Results are the mean(�SEM) of two determinations performed in triplicate. *P< 0.05,significantly different from control values.

Fig. 2. Time-course of the percentage of recovery of 3T6 fibroblastmonolayer after wound-induced injury. Wounded cultures wereincubated in RPMI 1640 with 10% FCS in presence of ketoprofen(5 mM, &), ketoconazole (5 mM, *), ketoprofen (5 mM)þketoconazole(5 mM) (~), and baicalein (50 mM, !) or in absence of inhibitors(control condition, &). The percentage of wound recovery was thendetermined. Results are the mean (�SEM) of two determinationsperformed in triplicate. *P< 0.05, significantly different from controlvalues.

94 MORENO

at the optimal concentration of 0.1 mg/ml, similar to thePGE2 levels present in the supernatant of 3T6 fibroblastcultures with FCS (Lloret et al., 1996b). Likewise,AA metabolites of the cytochrome P-450 pathway, like12 and 20 HETEs (Makita et al., 1996) reverted thewound closure due to ketoconazole in a dose-dependentmanner. Thus, although other eicosanoids may con-tribute to the stimulation of wound healing, PGE2 and12 and 20 HETE are effective in our experimentalconditions.

PGE2 regulates cellular proliferation through inter-action with specific receptors and may modify the levelsof second messengers like calcium and cAMP (Sanchezand Moreno, 2002). Here, i[Ca2þ] and cAMP levelsincrease during wound closure (Table 1). Our resultssuggest that COX pathway metabolites like PGE2

regulated i[Ca2þ] and cAMP, whereas 12 and 20 HETEscontrol i[Ca2þ]. Thus, ketoprofen treatment decreasedcAMP levels and i[Ca2þ], whereas ketoconazol impairedi[Ca2þ] (Table 1).

PGE2 acts through the EP1 and EP4

receptors to induce the wound repairof 3T6 fibroblast cultures

PGE2 mediates its biological activities through fourdistinct receptors EP1, EP2, EP3, and EP4 which coupleto specific guanine-nucleotide binding(G)-proteins anddownstream signal transduction systems (Negishiet al., 1995). Both PGE2 subtype receptors, EP1 andEP4, have been found in murine fibroblasts (Watanabeet al., 1996). To determine which PGE2 receptorsare involved in wound repair, 3T6 fibroblast woundedmonolayers were treated with the EP1 antagonist,AH6809 (Coleman et al., 1985) and the EP4 antagonist,AH23848B (Coleman et al., 1994b). Figure 3 shows thatboth treatments blocked the wound closure stimulatedby FCS. Furthermore, EP1 and EP4 antagonists sig-nificantly reduced the i[Ca2þ] and cAMP levels inducedby FCS, respectively (Table 2).

To determine the implication of calcium and cAMP inthe control of the wound repair process induced by FCSin 3T6 fibroblast cultures, we used thapsigargin, anendoplasmic reticule ATPase inhibitor (Takemura et al.,1989), and forskolin, an adenylate cyclase activator(Seamon and Daly, 1986), to recover i[Ca2þ] and cAMPand thus revert the inhibitory effect of EP1 and EP4

antagonists on wound repair. Thapsigargin and for-skolin significantly reduced the inhibitory effect ofAH6809 and AH23848B on wound recovery, respec-tively (Fig. 4). We would like to highlight that theseeffects were correlated with the recovery of i[Ca2þ] andcAMP levels, respectively, in the wounded 3T6 fibro-blast cultures (Table 2).

DISCUSSION

AA is predominantly esterified in the phospholipids ofmembranes and its mobilization by phospholipases,mainly phospholipases A2 (PLA2), which occurs as aresult of a variety of physiological stimuli, is the rate-limiting step in the AA cascade. A key regulatory factorin spreading, migration, and proliferation is cell adhe-sion to the substrate and neighboring cells. Adhesion to a

TABLE 1. PGE2 and 12, 20 HETEs stimulate wound repair in 3T6 fibroblast cultures

Wound recovery(%)

[3H]Thymidineincorporation i[Ca2þ ] cAMP

Non-wounded cultures 78�5 126�11 19� 3Wounded cultures 97� 3 273�16 432�36 48� 5Ketoprofen 58� 3* 161� 9* 258�21* 12� 2*KetoprofenþPGE2 79� 5** 212�11** 386�24** 52� 4**Ketoconazole 61� 3* 158�10* 139�13* N.D.Ketoconazo-

leþHETEs82� 4** 223�11** 297�19** N.D.

Wounded cell cultures were incubated with ketoprofen (5 mM) or ketoconazole (5 mM) and PGE2 (100 pg/ml) or 12, 20HETEs (1 mM) in presence of fetal calf serum (FCS) (10%) for 12 h. The percentage of wound recovery was then determined.[3H]Thymidine incorporation (Dpm� 103), i[Ca2þ ] (nM) and cAMP levels (pmol/mg protein) were measured 6 h after cellmonolayers were injured. Measurements were performed in triplicate and expressed as the mean�SEM of twoexperiments. N.D., not determined.*P< 0.05, significantly different from wound cultures.**P< 0.05, significant different from cells incubated with ketoprofen or ketoconazole.

Fig. 3. Dose-dependent effect of EP1 and EP4 antagonists on thepercentage of recovery of 3T6 fibroblast monolayer after wound-induced injury. Wounded cultures were incubated with RPMI 1640with 10% FCS in presence of AH6809 (&) or AH23848B (*) for 12 h.The percentage of wound recovery was then determined. Results arethe mean (�SEM) of two determinations performed by triplicate.*P<0.05, significantly different from control values.

WOUND CLOSURE AND EICOSANOIDS 95

solid substrate results in AA mobilization, and metabo-lism through the COX pathway in 3T6 fibroblasts (Lloretand Moreno, 1996a). Stockton and Jacobson (2001) havereported an initial transient AA release concomitantlywith the cell attachment stage followed by a secondsustained AA release simultaneous to the spreadingstage. They have observed that fibroblast spreadingdepends on the metabolism of AA to leukotienes bylipoxygenases, whereas migration requires the oxida-tion of AA by COX to prostaglandins. Furthermore, theextent of AA release and the subsequent metabolism to

prostaglandins is dependent on adhesive interactionswith neighboring cells, the level of confluence, and theextent of cell–cell adhesion. Thus, there is a dramaticdecline in AA release and PGE2 synthesis induced byFCS in cultured 3T6 fibroblasts as cells progress from anactively proliferating preconfluent state to a contact-inhibited confluent monolayer, when the cells establishnumerous adhesive interactions with neighboringcells (Lloret and Moreno, 1996b; Sanchez and Moreno,2001a).

In our experimental conditions, the wounding of 3T6fibroblast cultures resulted in the release of AA inducedby FCS, which may be a stimulus of migration andproliferation after wounding. When a COX inhibitor wasused in this wound model, PGE2 synthesis and woundclosure were significantly inhibited impaired, whichwas reverted by adding exogenous PGE2. Thus, pro-stanoids like PGE2 may be directly involved in theregulation of migration and proliferation after wound-ing of 3T6 fibroblast cultures (Moreno, 1997). If weconsider that COX-2 was expressed by 3T6 fibroblasts(Moreno, 1996) and it was induced by growth factors(Martınez et al., 1997), it is plausible that COX-2 may beinvolved in the wound repair. In this way, we observedthat a COX-2 inhibitor reduced the wound repair pro-cess (Moreno, 1997).

Several studies have pointed to the role of PGE2

receptors in cellular proliferation. Thus, the EP3 sub-type PGE2 receptor is responsible for the proliferativeeffect of prostaglandins on rat hepatocytes (Hashimotoet al., 1997). Moreover, PGE2-interaction with EP1 andEP4 receptors may regulate progression through the cellcycle of 3T6 fibroblasts by modulating cyclin levels aswe reported recently (Sanchez and Moreno, 2002). Ourpresent results are the first evidence that EP1 and EP4

receptors subtypes are key transducers of the stimu-lus from PGE2 to induce wound closure in fibroblastcultures. Thus, AH6809 and AH23848B which antago-nize the interaction of PGE2 with EP1 and EP4 receptors,respectively, significantly inhibit the wound closureinduced by FCS presence.

Furthermore, each receptor exerts its effects throughchanges in specific cellular pathways. Thus, the EP1

receptor increases i[Ca2þ], whereas PGE2 interactionwith the EP4 receptor raises intracellular cAMP levels.Both second messengers are involved in the control of

Fig. 4. Effect of thapsigargin or forskolin on the percentage of woundrecovery of 3T6 fibroblast injured monolayers in presence of EP1 andEP4 antagonists. Wounded cultures were incubated in RPMI 1640with 10% FCS in presence of AH6809 (150 mM) or AH23848B (100 mM)or in absence of treatments (control condition, &) for 12 h.Thapsigargin (0 mM, ; 0.5 mM, ; and 5 mM, &) and forskolin(0 mM, ; 1 mM, ; and 10 mM, &) were added to cells treated withAH6809 and AH23848B, respectively. The experiments were per-formed for 12 h and the percentage of wound recovery was determined.Each bar represents the mean (� SEM) of two experiments performedfor triplicate. *P< 0.05, significantly different from control values.**P<0.05, significantly different from cells treated with AH6809 orAH23848B.

TABLE 2. EP1 and EP4 antagonists inhibit wound repair induced by FCS (10%) in 3T6 fibroblastcultures

Wound recovery(%)

[3H]Thymidineincorporation i[Ca2þ] cAMP

Wound cultures 95� 2 268� 18 456�38 52�4Thapsigargin 94� 3 252� 16 597�27* 55�4AH6809 18� 1* 43� 2* 112� 9* 47�3AH6809þ thapsigargin 43� 6 126�6** 398�26** 49�5Forskolin 36� 5 218� 11 472�29 109� 8*AH23848B 15� 2* 51� 3* 462�36 8� 2*AH23848Bþ forskolin 54� 4** 143�5** 423�29 39�4**

Wounded cell cultures were incubated with AH6809 (150 mM) or AH23848B (100 mM) and thapsigargin (0.5 mM) orforskolin (10 mM) in presence of FCS (10%) for 12 h. The percentage of wound recovery was then determined.[3H]Thymidine incorporation (Dpm� 103), i[Ca2þ] (nM) and cAMP (pmol cAMP/mg protein) levels were measured 6 h aftercell cultures were injured. Measurements were performed in triplicate and expressed as the mean�SEM of twoexperiments.*P< 0.05, significantly different from wound cultures.**P< 0.05, significantly different from cells incubated with AH6809 or AH23848B.

96 MORENO

the cell cycle machinery of 3T6 fibroblasts (Sanchez andMoreno, 2002) and, thus in the wound repair process.On the other hand, the eventual effects of AA metabo-lites on cell migration appear to be at the level of actincytoskeleton polymerization. The enhancement of cAMPby prostaglandins leads to protein kinase A activation,stress fiber reduction and actin filament bundling(Peppenlenbosch et al., 1995; Gratacap et al., 2001),which are essential for cell migration and so for thewound repair process.

We have reported that a Ca2þ-independent cytosolicPLA2 is involved in serum-induced 3T6 fibroblast pro-liferation (Sanchez and Moreno, 2001b) and that thesubsequent inhibition of AA release induced 3T6 fibro-blast growth arrest. Taken together the effects of PLA2

inhibitor, COX inhibitors and EP antagonists on 3T6fibroblast growth appear that COX pathway is a pivotalelement in these events. However, other pathway in-volved in the wound repair process will be importantspecially when COX pathway was inhibited and free AAmight be metabolized through alternative pathway.Here, we observe that EP1 and EP4 antagonists blocked3T6 fibroblast FCS-stimulated wound closure, whereasthe inhibition of the COX pathway inhibited 3T6 fibro-blast wound closure only partially. These findingspoint to the participation of another pathway of the AAcascade in the control of 3T6 fibroblast growth and thusin the wound closure of injured 3T6 cultures. Therefore,other eicosanoids like leukotrienes, lipoxins, HETEsand EETs, which are the end products of AA metabolismby the lipoxygenase or cytochrome P-450 pathways, mayalso be involved in the repair of wounded 3T6 fibroblastcultures. These products may alter cell growth andstimulate chemotaxis, both of which are essential in thewound repair process. Our results show that ketocona-zole, a cytochrome P-450 inhibitor, hinders the woundclosure induced by FCS in wounded 3T6 fibroblastcultures, whereas baicalein or NDGA, which inhibit thelipoxygenase pathway, do not modify the repair process.Our data also show that exogenous adding of 12 HETEand 20 HETE, which are key AA metabolites synthe-sized by cytochrome P-450, can revert, at least partially,the effect of ketoconazole on the wound repair processinduced by FCS. HETEs are involved in some aspects ofadhesion (Rice et al., 1998). Thus, they may controlmigration and wound closure. Moreover, our findingssuggest that these cytochrome P-450 pathway metabo-lites participate in the control of i[Ca2þ], as proposed byAmlal et al. (1996) and Roman and Harder (1993). Theseevents may be involved in the control of DNA synthesis,cell growth, and the subsequent wound repair process inwounded 3T6 fibroblast cultures. However, we must alsoconsider that 20 HETE mediates calcium/calmodulin-dependent protein kinase II-induced mitogen-activatedprotein kinase activation in vascular smooth musclecells (Muthalif et al., 1998) and Sun et al. (1999) reportedthat 20 HETE modulates MAP kinase in renal arter-ioles, which may also influence 3T6 fibroblast growthand wound closure process.

In conclusion, COX pathway metabolites are essentialat the onset of 3T6 fibroblasts wound closure. Inhibitionof this pathway dramatically hinders the wound repairprocess, whereas the exogenous adding of PGE2 signi-ficantly accelerate closure. Both EP1 and EP4, receptors

of PGE2 mediate PGE2-stimulated 3T6 fibroblast woundclosure. Furthermore, our results reveal that calciumand cAMP are involved in the signaling events inducedby these eicosanoids in the wound repair process.Furthermore, the cytochrome P-450 pathway may bean alternative to AA metabolism, especially when COXpathway is blocked. Thus, 12 and 20 HETEs may alsoparticipate in the wound closure process of 3T6 fibro-blast cultures.

ACKNOWLEDGMENTS

The author is very grateful to Robin Rycroft for hisassistance in the preparation of the manuscript.AH6809 and AH23848B were kindly provided by Dr.S.G. Lister from Compound Supplies Officer, GlaxoWellcome Medicines Research Centre, Hertfordshire,UK.

LITERATURE CITED

Amlal H, Legoff C, Vernimmen C, Paillard M, Bichara M. 1996. Naþ–Kþ–(NH4

þ)–2Cl� contransport in medullary thick ascending limb:Control by PKA, PKC, and 20-HETE. Am J Physiol 271:C455–C463.

Argentieri DC, Ritchie DM, Ferro MP, Kircner T, Wachter MP,Anderson DW, Rosenthale ME, Capetola RJ. 1994. Tepoxalin: Adual cyclooxygenase/5-lipoxygenase inhibitor of arachidonic acidmetabolism with potent anti-inflammatory activity and a favorablegastrointestinal profile. J Pharmacol Exp Ther 271:1399–1408.

Bradford MM. 1976. A rapid and sensitive method for the quantifica-tion of microgram quantities of protein using the principles ofprotein-dye binding. Anal Biochem 72:248–254.

Coleman RA, Kennedy I, Sheldrick RLG. 1985. AH6809, a prostanoidEP1-receptor blocking drug. Br J Pharmacol 85:273P.

Coleman RA, Smith WL, Narumiya S. 1994a. International union ofpharmacology classification of prostanoid receptors: Properties,distribution, and structure of the receptors and their subtypes.Pharmacol Rev 46:205–229.

Coleman RA, Grix SP, Head SA, Louttit JB, Mallett A, Sheldrick RL.1994b. A novel inhibitory prostanoid receptor in piglet saphenousvein. Prostaglandins 47:151–167.

Domin J, Rozengurt E. 1993. Platelet-derived growth factor stimu-lates a biphasic mobilization of arachidonic acid in Swiss 3T3 cells.J Biol Chem 268:8927–8934.

Gratacap MP, Payrastre B, Nieswandt B, Offermanns S. 2001.Differential regulation of Rho and Rac through heterotrimeric G-proteins and cyclic nucleotides. J Biol Chem 276:47906–47913.

Grynkiewicz G, Poenie M, Tsien RM. 1985. A new generation of Ca2þ

indicators with greatly improved fluorescence properties. J BiolChem 260:3440–3450.

Hashimoto N, Watanabe T, Ikeda Y, Yamada H, Taniguchi S,Mitsui M, Kurokawa K. 1997. Prostaglandins induce proliferationof rat hepatocytes through a prostaglandin E2 receptor EP3 subtype.Am J Physiol 272:G597–G604.

Hope WC, Welton AF, Fiedler-Nagy C, Batula-Bernardo C, Coffey JW.1983. In vitro inhibition of the biosynthesis of slow reactingsubstance of anaphylaxis (SRS-A) and lipoxygenase activity byquercetin. Biochem Pharmacol 32:367–371.

Hyman BT, Stoll LL, Spector AA. 1982. Prostaglandin production by3T3-L1 cells in culture. Biochim Biophys Acta 713:375–385.

Joyce NC, Meklir B. 1994. Prostaglandin E2: A mediator of cornealendothelial wound repair in vitro. Am J Physiol Cell Physiol 266:C269–C275.

Kaneko F, Zhang JZ, Maruyama K, Nihei Y, Ono I, Iwatsuki K,Yamamoto T. 1995. Prostaglandin I1 analogues, SM-10902 and SM10906, affect human keratinocytes and fibroblasts in vitro in amanner similar to PGE1: Therapeutic potential for wound healing.Arch Dermatol Res 287:539–545.

Kobayashi K, Arakawa T. 1995. Arachidonic acid cascade and gastricmucosal injury, protection, and healing: Topics of this decade. J ClinGastroenterol 21:S12–S17.

Lloret S, Moreno JJ. 1996a. Stimulation of arachidonic acid mobiliza-tion by adherence of resident peritoneal macrophages to plasticsubstrate. Comp Biochem Physiol 113:C403–C408.

Lloret S, Moreno JJ. 1996b. Proliferation-dependent changes inarachidonic acid mobilization from phospholipids of 3T6 fibroblasts.Pflugers Archiv Eur J Physiol 432:655–662.

WOUND CLOSURE AND EICOSANOIDS 97

Makita K, Falck JR, Capdevila JH. 1996. Cytochrome P450, thearachidonic acid cascade, and hypertension: New vistas for an oldenzyme system. FASEB J 10:1456–1463.

Martınez J, Sanchez T, Moreno JJ. 1997. Role of prostaglandin Hsynthase-2-mediated conversion of arachidonic acid in controlling3T6 fibroblast growth. Am J Physiol Cell Physiol 273:C1466–C1471.

Maurice M, Pichard L, Daujat M, Fabre R, Joyeux H, Domergue J,Maurel P. 1992. Effects of imidazole derivatives on cytochromeP450 from human hepatocytes in primary culture. FASEB J 6:752–758.

Moreno JJ. 1996. Role of cyclooxygenase-2 in 3T6 fibroblast prolifera-tion. Biochem Soc Trans 24:551S.

Moreno JJ. 1997. Regulation of arachidonic acid release andprostaglandin formation by cell–cell adhesive interactions in woundrepair. Pflugers Archiv Eur J Physiol 433:351–356.

Moreno JJ, Calvo L, Fernandez F, Carganico G, Bastida E. 1990.Biological activity of ketoprofen and its optical isomers. Eur JPharmacol 183:2263–2264.

Muthalif MM, Benter IF, Karzoum N, Fatima S, Harper J, Uddin MR,Malik KU. 1998. 20-Hydroxyeicosatetraenoic acid mediates cal-cium/calmodulin dependent protein kinase II-induced mitogen-activated protein kinase activation in vascular smooth muscle cells.Proc Natl Acad Sci USA 95:12701–12706.

Nakamura M, Fujihara T, Mibu H, Hikida M. 1994. Arachidonic acidstimulates corneal epithelial migration. J Ocul Pharmacol Ther10:453–459.

Narumiya S, Sugimoto Y, Ushikubi F. 1999. Prostanoid receptors:Structures, properties, and functions. Physiol Rev 79:1193–1226.

Needleman P, Turk J, Jakschik BA, Morrison AR, Lefkowich JB. 1986.Arachidonic acid metabolism. Annu Rev Biochem 55:69–102.

Negishi M, Sugimoto Y, Ichikawa A. 1995. Prostaglandin E receptors.J Lipid Mediat Cell Signal 12:379–391.

Peppenlenbosch MP, Qui RG, de Vries-Smits AM, Tertoolen LGJ,de Laat SW, Mc Cormick F, Hall A, Symons MH, Bos JL. 1995. Racmediates growth factor-induced arachidonic acid release. Cell 81:849–856.

Rice RL, Jang DG, Haddad M, Honn KV, Taylor JD. 1998. 12 (S)hydroxyeicosatetraenoic acid increases the actin microfilamentcontent in B16a melanoma cells: A protein kinase-dependentprocess. Int J Cancer 77:271–278.

Rieger GM, Hein R, Adelmann-Grill BC, Ruzicka T, Krieg T. 1990.Influence of eicosanoids on fibroblast chemotaxis and protein syn-thesis in vitro. J Dermatol Sci 1:347–354.

Roman RJ, Harder DR. 1993. Cellular and ionic signal transductionmechanisms for the mechanical activation of renal arterial vascularsmooth muscle. J Am Soc Nephrol 4:986–996.

Rozengurt E, Collins MK, Keehan M. 1983. Mitogenic effect of pro-staglandin E1 in Swiss 3T3 cells: Role of cAMP. J Cell Physiol 116:379–384.

Sanchez T, Moreno JJ. 2001a. Role of phospholipase A2 in growth-dependent changes in prostaglandin release from 3T6 fibroblasts.J Cell Physiol 189:237–243.

Sanchez T, Moreno JJ. 2001b. The effect of high molecular pho-spholipase A2 inhibitors on 3T6 fibroblast proliferation. BiochemPharmacol 61:811–816.

Sanchez T, Moreno JJ. 2002. Role of EP1 and EP4 PGE2 subtypereceptors in serum-induced 3T6 fibroblast cycle progression andproliferation. Am J Physiol Cell Physiol 282:C280–C288.

Seamon KB, Daly JW. 1986. Forskolin: Its biological and chemicalproperties. Adv Cyclic Nucleotide Protein Phosphorylation Res20:1–150.

Sekiya K, Okuda H. 1982. Selective inhibition of platelet lipoxygenaseby baicalein. Biochem Biophys Res Commun 105:1090–1095.

Stockton RA, Jacobson BS. 2001. Modulation of cell-substrate adhe-sion by arachidonic acid: Lipoxygenase regulates cell spreading andERK1/2-inducible cyclooxygenase regulates cell migration in NIH-3T3 fibroblasts. Mol Biol Cell 12:1937–1956.

Sun CW, Falck JR, Harder DR, Roman RJ. 1999. Role of tyrosinekinase and PKC in the vasoconstrictor response to 20-HETE inrenal arterioles. Hypertension 33:414–418.

Takahashi M, Ota S, Hata Y, Mikami Y, Azuma N, Nakamura T,Terano A, Omata M. 1996. Hepatocyte growth factor as a key tomodulate anti-ulcer action of prostaglandins in stomach. J ClinInvest 98:2604–2611.

Takemura H, Hughes AR, Thastrup O, Putney JW. 1989. Activation ofcalcium entry by the tumor promoter thapsigargin in parotid acinarcells. J Biol Chem 264:12266–12271.

Talwar M, Moyana TN, Bharadwaj B, Tan LK. 1996. The effect ofsynthetic analogue of prostaglandin E2 on wound healing in rats.Ann Clin Lab Sci 26:451–457.

Watanabe T, Satoh H, Togoh S, Taniguchi Y, Hashimoto Y, KurokawaK. 1996. Positive and negative regulation of cell proliferationthrough prostaglandin receptors in NIH-3T3 cells. J Cell Physiol169:401–409.

Yamashita T, Takai Y. 1987. Inhibition of prostaglandin E1-inducedelevation of cytoplasmic free calcium ion by protein kinase C-activating phorbol esters and diacylglycerol in Swiss 3T3 fibro-blasts. J Biol Chem 262:5536–5539.

Zushi S, Shinomura Y, Kiyohara T, Minami T, Sugimachi M,Higashimoto Y, Kanayama S, Matsuzawa Y. 1996. Role of prostag-landins in intestinal epithelial restitution stimulated by growthfactors. Am J Physiol 270:G757–G762.

98 MORENO