UVB Activates ERK1/2 and p38 Signaling Pathways viaReactive Oxygen Species in Cultured Keratinocytes

Dominik Peus, Remus A. Vasa, Astrid Beyerle, Alexander Meves, Carsten Krautmacher, andMark R. PittelkowDepartments of Dermatology, and Biochemistry and Molecular Biology, Mayo Clinic/Foundation, Rochester, Minnesota, U.S.A.

We have previously shown that hydrogen peroxideis an important mediator of ultraviolet B inducedphosphorylation of the epidermal growth factor recep-tor in human keratinocytes. Here we demonstratethat physiologic doses of ultraviolet B and hydrogenperoxide stimulate activation of two related but distinctmitogen-activated protein kinase pathways: extracellu-lar regulated kinase 1 and 2 (ERK1/2), as well as p38,the mammalian homolog of HOG1 in yeast which isa major kinase for a recently identified stress-inducedsignaling pathway. The time-dependent activation ofERK1/2 and p38 are distinct, and ultraviolet B-inducedERK1/2 activation is downregulated more rapidly thanp38. Using dihydrorhodamine or Amplex as specificfluorescent dye probes, we show that ultravioletB-induced peroxides can be inhibited by ascorbicacid. Ascorbic acid strongly blocks ERK1/2 and p38activation by ultraviolet B and hydrogen peroxidewhereas pyrrolidine dithiocarbamate and butyl hydro-

Ultraviolet radiation (UVR) is the primary cause forthe vast majority of cutaneous malignancies eachyear in the U.S.A. (Miller and Weinstock, 1994).UVR is therefore the most efficient environmentalcarcinogen known and acts as both a tumor initiator

as well as a tumor promotor (Romerdahl et al, 1989; Matsui andDeLeo, 1991). Skin exposure to UVB wavelength radiation, a minorbut highly carcinogenic constituent of sunlight (Tornaletti andPfeifer, 1996), induces a range of biologic responses such aserythema, inflammation, hyperpigmentation, hyperplasia, skin can-cer (Norris et al, 1993; Young, 1993; Gilchrest et al, 1996), DNAdamage and mutations, induction of early and late gene responses(Sachsenmaier et al, 1994; Herrlich et al, 1994), systemic immuno-suppression (Kripke, 1994), and activation of dormant viruses(Herrlich et al, 1992).

UVR is a potent inducer of reactive oxygen species (ROS)including superoxide radical (·O2

–), hydrogen peroxide (H2O2)and hydroxyl radical (·OH), which have been implicated in

Manuscript received February 3, 1998; revised January 25, 1999; acceptedfor publication February 1, 1999.

Reprint requests to: Dr. Mark R. Pittelkow, Mayo Clinic, DermatologyResearch, Gugg. 411, 200 First Street SW, Rochester, MN 55905.

Abbreviations: BHA, butyl hydroxyanisole; DHR, dihydrorhodamine123; ERK, extracellular regulated kinase; PDTC, pyrolidinedithiocarba-mate; ROS, reactive oxygen species.

0022-202X/99/$10.50 · Copyright © 1999 by The Society for Investigative Dermatology, Inc.

751

xyanisole are less effective. Pyrrolidine dithiocarba-mate was unable to inhibit ultraviolet B-induced p38activation. Cell death was increased after ultraviolet Bwhen ERK1/2 activation was attenuated by the specificinhibitor PD098059. The distinct time courses andextents of activation and inhibition of ERK1/2 andp38 indicate that these pathways are separate andregulated independently in keratinocytes. Specifictypes of reactive oxygen species induced by ultravioletB as well as selective activation or inhibition of specificphosphatases may mediate these responses inkeratinocytes. These findings demonstrate that reactiveoxygen species are important multifunctional medi-ators of ultraviolet B-induced ERK1/2 and p38 sig-naling transduction pathways and suggest that ERK1/2may play an important part in protecting keratinocytesfrom cell death following oxidative stress. Key words:extracellular regulated kinase/hydrogen peroxide/keratino-cytes/mitogen-activated protein kinase/p38/reactive oxygenspecies/ultraviolet B. J Invest Dermatol 112:751–756, 1999

cutaneous aging as well as cancer and various inflammatory disorders(Cerutti, 1985; Darr and Fridovich, 1994). Among ROS, H2O2as well as UVR have been reported to upregulate expression ofgenes such as c-fos and c-jun (Devary et al, 1992; Bender et al,1997). Transcription of many early genes is mediated by thesequential activation of cytoplasmic protein kinases, and themitogen-activated protein kinases (MAPK) play a major part intriggering and coordinating these gene responses (Karin, 1995).

Three structurally related but biochemically and functionallydistinct MAPK signal transduction pathways have been identifiedand include the extracellular receptor kinases (ERK), c-jun N-terminal kinases (JNK) and p38 (also known as CSBP, RK, Mkp2),which is the mammalian homolog of yeast HOG1 (Su and Karin,1996). ERK are primarily activated in response to growth factorsand phorbol esters, whereas stress or inflammatory cytokines poorlyactivate the ERK (Whitmarsh et al, 1995; Xia et al, 1995).JNK, by contrast, participate in growth-factor signaling as well asresponding to various stress events that also activate p38, includingenvironmental stresses such as UVR and inflammatory cytokines(Raingeaud et al, 1995; Canman and Kastan, 1996). Epidermalgrowth factor (EGF) and phorbol esters have been reported to bepoor activators of p38 (Derijard et al, 1995; Rouse et al, 1994).

In comparison with the well-established function of ERK withinthe Ras/Raf/MEK/ERK signaling cascade, the mechanism ofactivation of p38 by UVB as well as its biologic role(s) remain tobe elucidated. The focus of this investigation was to identify more

752 PEUS ET AL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

precisely the role of ROS, in particular H2O2, on the activationand regulation of ERK1/2 and p38 signal transduction pathwaysin normal human keratinocytes following exposure to physiologicdoses of UVB radiation.

MATERIALS AND METHODS

Materials and chemicals The UVB source was a FS20 lamp(Westinghouse, Pittsburgh, PA) that emits an energy spectrum with highfluence in the UVB region and a peak at 313 nm (Peus et al, 1988). Toblock UVC emission a WG-295 long pass filter was used (Schott GlassTechnologies, Duryea, PA). Experiments using a WG-295 long pass filtershowed that differences in phosphorylated levels of p38 and ERK1/2 areonly minimal when unfiltered and filtered UVB-radiation sources werecompared. With longer incubation periods (30 min) a slightly strongerresponse was observed after exposure to unfiltered UVB radiation. Theemitted dose was regularly quantitated by a UVB radiometer and photode-tector (IL 443 and SEE 240, International Light, Newburyport, MA).Ascorbic acid (AA), butylated hydroxyanisole (BHA), and pyrolidinedithi-ocarbamate (PDTC), were purchased from Sigma (St. Louis, MO).Dihydrorhodamine 123 (DHR), Amplex Red reagent and LIVE/DEADViability/Cytotoxicity Kit were obtained from Molecular Probes(Eugene, OR).

Culture and treatment of normal human keratinocytes Normalhuman keratinocytes were isolated from neonatal foreskin specimens, andprimary cultures were initiated and maintained in a replicative state withcomplete, serum-free MCDB 153 medium. Keratinocytes from primarycultures were plated into secondary culture at 1–10 3 103 cells per cm2.Complete medium was supplemented with 0.1 mM calcium, 0.2%(vol/vol) bovine pituitary extract (BPE), EGF (10 ng per ml), insulin (5 µgper ml), hydrocortisone (5 3 10–7 M), ethanolamine (1 3 10–4 M),phosphoethanolamine (1 3 10–4 M), and additional amino acids (Willeet al, 1984; Pittelkow and Scott, 1986). Cells were grown to confluenceand fed with standard medium (without growth factors) for at least 48 hto obtain quiescent cells with low levels of activated ERK1/2 and p38prior to experiments examining UVB irradiation or H2O2 treatment. Allother assays were performed in confluent keratinocytes cultured in completemedium. No significant differences were observed whether cells wereirradiated in phosphate-buffered saline or standard medium.

Detection of intracellular H2O2 by flow cytometric analysisIntracellular levels of H2O2 were analyzed using DHR as a specificfluorescent dye probe as described previously (Peus et al, 1998). Briefly,confluent keratinocytes in a 24 well tissue culture plate were loaded withDHR (5 µM) for 45 min and washed before treatment with UVB. Duringthe intracellular release of H2O2, reduced DHR is irreversibly oxidizedand converted to the red fluorescent compound rhodamine 123 (Royaland Ischiropoulos, 1993). Plates were read on a Cytofluor II, fluorescenceplate reader (Perspectives Biosystems, Framingham, MA) with an excitationwavelength of 485 nm and an emission wavelength of 530 nm. Each barrepresents the mean value of triplicate wells 6 SD.

Extracellular H2O2 assay Measurement of extracellular H2O2 followingUVB exposure was performed using Amplex a derivative of dihydrophenox-azine, which becomes highly fluorescent upon oxidation by H2O2 in thepresence of horseradish peroxidase. Fluorescence intensity was detectedusing the microplate fluorometer. The excitation/emission wavelengthfilter was set at 590/615 nm. Fluorescence intensity of Amplex wasmeasured within the linear region of the H2O2 dose–response curve. Wellswithout cells containing the same substrates and exposed to the sametreatment represented controls. The control fluorescent values and back-ground fluorescence of untreated cells were monitored over time andsubtracted from the relative fluorescence intensity of treated samples. Thesevalues were converted into absolute pmoles using an H2O2 standard curve.All H2O2 measurements were performed in 24-well plates containingconfluent keratinocytes, 100 µM Amplex and 1 U per ml horseradishperoxidase (Mohanty et al, 1997). Each bar represents the mean value oftriplicate wells 6 SD.

Detection of activated ERK1/2 and p38 Normal human keratinocyteswere extracted in a lysis buffer supplemented with protease and phosphataseinhibitors. Endogenous, activated forms of ERK1 and ERK2 were detectedusing 20 µg of total protein and the anti-ACTIVE MAPK antibodypolyclonal antibody (Promega, Madison, WI) that selectively recognizesactivated ERK1/2 kinase (Sonoda et al, 1997). For detection of activatedp38, phospho-specific p38 polyclonal antibody was used (Wilhelm et al,

Figure 1. Differential stimulation of ERK1/2 and p38 by UVB andH2O2. Confluent, resting human keratinocytes were exposed to (a) UVB(200 J per m2) or (b) H2O2 (200 µM) and monitored for 5–120 min. Totalcell protein was extracted at the indicated times, separated by sodiumdodecyl sulfate–polyacrylamide gel electrophoresis, transferred tonitrocellulose membranes and immunoblotted with anti-ACTIVE ERK1/2, anti-phospho-p38 antibody or total p38 antibody as described in Materialsand Methods. One representative experiment is shown of three separateexperiments.

1997). A nonspecific band of higher molecular weight than p38 was alsoobserved but was not detected with newer lots of the phospho-specificp38 antibody. Phospho-specific and total p38 antibodies were purchasedfrom New England BioLabs (Beverley, MA). ERK1/ERK2 proteins weredetected using the anti-ERK1 C-16 antibody (Santa Cruz Biotechnology,Santa Cruz, CA), and equivalent loading of ERK1/2 and p38 proteins wasdocumented for each series of experiments. Samples were separatedelectrophoretically by 10% sodium dodecyl sulfate–polyacrylamide gelelectrophoresis. Proteins were electrotransferred to Immobilon-p(Millipore, Bedford, MA), and after blocking and incubation with secondaryantibody coupled to peroxidase, bands were visualized using the ECLdetection system (Amersham, Arlington Heights, IL).

Cell death assay Cell death was determined using the ethidium homo-dimer/calcein acetoxymethyl ester Viability/Cytotoxicity Kit (MolecularProbes, Eugene, OR) combination of vital dyes. Confluent keratinocytesin 24 well cell culture plates were treated with UVB, and 24 h laterfluorescence was analyzed in the Cytofluor II microplate fluorometer.Results are representative of three independent experiments (mean 6 SD).

RESULTS

Differential activation of ERK1/2 and p38 by UVB andH2O2 For selected cell types, ERK1/2 and p38 signaling pathwayshave been shown to be biochemically separate and distinct. Tocharacterize specific cellular responses mediated by these pathwaysin keratinocytes, activation of ERK1/2 and p38 by UVB andH2O2 over time was evaluated. Levels of activated ERK1/2 andp38 were examined 5–120 min following treatment of keratinocyteswith 200 J per m2 UVB (Fig 1a) or 200 µM H2O2 (Fig 1b).Within 5 min of UVB treatment, ERK1/2 activation was maximallyupregulated. By 30 min after irradiation, the signal declined andapproached baseline level by 60 min (Fig 1a). In contrast, as shownin Fig 1a, activated p38 only reached a maximum level by30 min with minimal decrease after 2 h. Interestingly, unlike UVBstimulation, H2O2 treatment showed similar profiles of ERK1/2and p38 activation, with an increase by 5 min and maximum

VOL. 112, NO. 5 MAY 1999 UVB INDUCES ERK1/2 AND P38 ACTIVATION VIA ROS 753

Figure 2. Concentration-dependent ERK1/2 and p38 activation byUVB and H2O2. Keratinocytes were treated with the indicated doses of(a) UVB (50–800 J per m2) for 15 min or (b) H2O2 (10–400 µM) for30 min. Immunoblotting with anti-active ERK1/2, phospho-specific p38or total p38 antibody was performed as described in Materials and Methods.One representative experiment is shown of three separate experiments.

phosphorylated levels reached by 15 min (Fig 1b). Decrease in p38activation was observed within 45 min versus 60 min for ERK1/2activation. Total ERK1/2 and p38 protein levels at all time pointswere equivalent as measured by immunoblot of the same strippedmembranes using antibodies against total ERK1/2 (data not shown)and p38 (Fig 1b) as described in Materials and Methods. These resultsdemonstrate marked differences in the activation of ERK1/2 andp38 following UVB exposure, but similar activation after H2O2treatment in keratinocytes.

UVB and H2O2 activate ERK1/2 and p38 in a concentration-dependent manner Human keratinocytes were irradiated with50–800 J UVB per m2 and incubated for 15 or 30 min. Concentra-tion-dependent increase in kinase phosphorylation was observedthat was more marked for p38 than for ERK1/2 activation (Fig 2a).Increasing concentrations of H2O2 (10–400 µM) stimulated ERK1/2 and p38 activation in a concentration-dependent manner. ERK1/2 phosphorylation was enhanced at 30 µM H2O2, whereas signific-ant p38 activation was observed only by 100 µM H2O2 (Fig 2b).Peak ERK1/2 activation was reached by 100 µM H2O2 versus200 µM H2O2 for p38 activation. Further increase in ERK1/2 orp38 activation by higher concentrations of H2O2 was not observed(data not shown). Total p38 protein levels remained unchangedover the time course (Fig 2b). These results show that both ERK1/2 and p38 are activated by UVB or H2O2 in a concentration-dependent manner.

Intracellular H2O2 production and H2O2 release induced byUVB is inhibited by ascorbic acid in a concentration-dependent manner We have previously shown that H2O2 isgenerated rapidly and in a concentration-dependent manner byhuman keratinocytes irradiated with UVB, and is inhibited by theanti-oxidants PDTC and BHA (Peus et al, 1998). To determinethe effects of the anti-oxidant AA on intracellular H2O2 production

Figure 3. Inhibition of UVB-induced H2O2 production by anti-oxidant AA. (a) To determine intracellular H2O2 generation, DHR-loaded keratinocytes were irradiated with 400 J UVB per m2 and analyzedfor fluorescence intensity using a Cytofluor II microplate reader as describedin Materials and Methods. H2O2 production after UVB (400 J per m2)irradiation and incubation for 30 min is inhibited by AA (30–300 µM).(b) Extracellular H2O2 release was determined using Amplex (100 µM), adihydrophenoxazine derivative and horseradish peroxidase (1 U per ml)that were added to confluent keratinocytes in 24-well plates. Thirty minutesafter exposure to different doses of UVB, fluorescence intensity wasmeasured by a Cytofluor II fluorescence plate reader and converted intoabsolute H2O2 concentration using a standard curve. AA was added to themedium 30 min before treatment in all experiments and no significantchange was observed in baseline fluorescence intensity of cells treated withAA only. Each bar represents the mean value of triplicate wells 6 SD.

in human keratinocytes, cells were loaded with DHR, pretreatedwith AA then irradiated with UVB and incubated for 30 min. Therelative fluorescent intensity of cells exposed to 400 J UVB per m2

was specifically inhibited in a concentration-dependent manner by30–300 µM AA (Fig 3a). In addition, levels of extracellular H2O2produced and released in unstimulated human keratinocytes werealso attenuated concentration-dependently by AA after UVB

754 PEUS ET AL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

Figure 4. Antioxidants AA, PDTC, and BHA modulate ERK1/2and p38 activation by UVB. Keratinocytes were pretreated for 30 minwith (a) different doses of AA or (b) with PDTC (3 µM) or BHA (10 µM)before exposure to UVB (200 J per m2) and incubation for 15 min.Activation of ERK1/2 and p38 was analyzed as described in Materialsand Methods. One representative experiment is shown of three separateexperiments.

(Fig 3b). These results provide direct evidence that AA decreasesH2O2 levels in UVB-stimulated keratinocytes.

Antioxidants AA, PDTC, and BHA modulate ERK1/2 andp38 phosphorylation by UVB Human keratinocytes pretreatedfor 30 min with 100 or 200 µM of AA prior to 200 J UVB perm2 exposure resulted in potent inhibition of p38 and ERK1/2phosphorylation (Fig 4a). Addition of 10 µM BHA to the medium30 min before exposure to 200 J UVB per m2 attenuated ERK1/2 and p38 activation (Fig 4b). Three micromoles of PDTC,however, potently inhibited ERK1/2 activation but failed to inhibitUVB-induced p38 activation. Both PDTC and BHA inhibitintracellular peroxide production but are less effective in blockingH2O2-induced ERK1/2 phosphorylation (data not shown). Allthree anti-oxidants had no significant effect on the basal level ofactivated ERK1/2 in unirradiated cells. The anti-oxidants AA andBHA attenuate levels of ROS induced in keratinocytes by UVBexposure and also down-regulate activation of ERK1/2 and p38,whereas PDTC failed to inhibit p38 activation. These findingsprovide further evidence for differential regulation of ERK1/2 andp38 activation in keratinocytes.

Specific inhibition of ERK1/2 by PD098059 after UVBor H2O2 exposure PD098059 has been reported to inhibitspecifically ERK1/2 activation in different cell types (Alessi et al,1995b; Dudley et al, 1995). We determined whether this compoundinhibits activation of ERK1/2 following pretreatment of humankeratinocytes for 30 min and exposure to 400 J UVB per m2 or200 µM H2O2. PD098059 potently inhibited ERK1/2 activationstimulated by UVB (Fig 5a) and H2O2 (Fig 5b) in a concentration-dependent manner with maximum inhibition at 30 or 3 µM,respectively. No significant inhibition of p38 phosphorylation wasobserved. Therefore, UVB- or H2O2-induced activation of ERK1/

Figure 5. Inhibition of ERK1/2 by PD098059 after UVB or H2O2exposure. Keratinocytes were pretreated for 30 min with (a) PD098059(3–30 µM) before exposure to UVB (400 J per m2) or (b) PD098059 (0.3–3 µM) before treatment with H2O2 (200 µM) and incubated for 15 min.Immunoblotting with anti-active ERK1/2 or phospho-specific p38antibody was performed as described in Materials and Methods. Onerepresentative experiment is shown of three separate experiments.

2 can be potently and selectively blocked by PD098059 inkeratinocytes.

Inhibition of ERK1/2 results in enhanced cell death afterUVB To examine the biologic significance of ERK1/2 blockade,we analyzed the effect of PD098059 on cell survival after UVBexposure. Using the sensitive Viability/Cytotoxicity Kit (MolecularProbes), we observed that 24 h after irradiation with 400 J UVBper m2 cell death increased significantly in cells treated withPD098059 compared with irradiated control without the ERK1/2 inhibitor (Fig 6). In separate experiments the ratio of apoptoticto necrotic cells was determined by using DAPI and propidiumiodide staining and enumeration using fluorescence microscopicvisualization. Twenty-four hours after UVB irradiation, the ratioof apoptotic to necrotic cells was µ2:3, and remained unchangedwith or without PD098059 (data not shown). The results indicatethat ERK1/2 activation following UVB exposure plays an importantpart in mediating keratinocyte survival in the UVB response.

DISCUSSION

The biologic mediators of various cellular responses to UVR areonly beginning to be identified and characterized. Our observationsprovide new evidence for the regulation of ERK1/2 and p38 signaltransduction pathways in keratinocytes exposed to UVB or H2O2.Previous studies have used UVC irradiation, largely at supraphysio-logic doses, and tested immortalized or carcinoma cell lines whereintracellular signal transduction pathways are frequently perturbedas part of the transformation process. Therefore, this investigationwas designed to examine doses of 50–800 J UVB per m2 that arein the range of a minimal erythema dose (MED) under standardskin testing conditions for skin types I–III and typically induce thephysiologic response of solar erythema (Norris et al, 1993). As a

VOL. 112, NO. 5 MAY 1999 UVB INDUCES ERK1/2 AND P38 ACTIVATION VIA ROS 755

Figure 6. ERK1/2 inhibition leads to increased cell death afterUVB. Live and dead cells were determined 24 h after 400 J per m2 UVBby loading cells with 2 µM calcein acetoxymethyl ester and 2 µM ethidiumhomodimer-1 as described in Materials and Methods. Cells were pretreatedwith 20 µM PD098059 for 30 min before UVB exposure. p , 0.05versus control.

result, the dose range of UVB that induces ERK1/2 and p38pathway activation in cultured keratinocytes has physiologic relev-ance for epidermis and skin.

In addition to characterizing the concentration- and time-dependent activation of ERK1/2 and p38 pathways following UVBor H2O2 treatment, we examined potential mechanisms by whichthese pathways are activated and regulated. Recently, it has beenshown that H2O2 induces ERK activation in different immortalizedcell lines (Guyton et al, 1996; Rao, 1996; Zhao et al, 1996). It isnot clear, however, how cellular stress selectively triggers p38, andto a lesser extent ERK1/2, pathway activation in the cytoplasm.It has been shown that MAPK are activated by dual phosphorylationon threonine and tyrosine within protein kinase subdomain VIII(Davis, 1994). ERK are activated by MEK1 and MEK2 (Cobb andGoldsmith, 1995), and p38 by MKK3, MKK4, and MKK6 (Derijardet al, 1995; Raingeaud et al, 1996). Upon stimulation, these kinasestranslocate into the nucleus where they selectively activate severaldifferent transcription factors and are then rapidly inactivated bydual specific phosphatases (reviewed in Keyse, 1995).

ROS have been implicated as potential mediators of signalingfrom the membrane to nucleus, including phosphorylation of EGFreceptors (EGFR) (Peus et al, 1998), association of SOS, Grb2, andshc with EGFR (Huang et al, 1996), and ERK1/2 activationfollowing UV exposure (Assefa et al, 1997). UVB-induced EGFRphosphorylation, similar to ligand-induced autophosphorylation ofEGFR (Peus et al, 1997) was inhibited by anti-oxidants and catalase,and these effects were mimicked by H2O2, which was found tobe generated within keratinocytes after UVB exposure (Peus et al,1998). H2O2 is generated in the cell following UVB irradiation,stimulates phosphorylation of EGFR and therefore plays an import-ant part as a cellular mediator. This hypothesis is supported bythe observations that structurally different anti-oxidants includingPDTC, BHA, and AA inhibit UVB-induced H2O2 formation ina concentration-dependent manner, and this reduction closelycorrelates with inhibition of EGFR phosphorylation (Peus et al,1998) and ERK1/2 activation (Fig 4). Therefore, we suggest thatUVB-induced H2O2 may mediate ERK1/2 activation throughupstream substrates such as ras, raf, and MEK1 and 2 as a result ofEGFR phosphorylation. In addition, the distinct levels and kineticsof activation and inhibition of ERK1/2 and p38 strongly implicateindependent mechanisms of regulation of each pathway inkeratinocytes.

The finding that PDTC and BHA differently inhibit UVB- andH2O2-induced ERK1/2 and p38 activation (Fig 4b) as well as thedose disparity of other anti-oxidants required for inhibition (Fig 4a)indicate that the regulation of these pathways by ROS is distinct,and their responsiveness may differ depending on the types of ROSinvolved. Within 30 min of UVB exposure, ERK1/2, but notp38, activation is markedly downregulated (Fig 1a), whereas bothERK1/2 and p38 activation induced by H2O2 are attenuatedwithin 60 and 30 min after treatment, respectively (Fig 1b).Dephosphorylation of activated ERK1/2 and p38 may be mediatedby the recently isolated and characterized family of dual-specificity(Thr/Tyr) MAPK phosphatases (MKP), comprised mainly of MKP-1–4 (Chu et al, 1996; Muda et al, 1996) which are inducible bycellular stress and rapidly localize to the nucleus (Keyse, 1995).ERK, however, are more likely than p38 to be inactivated byprotein phosphatase 2A and a protein tyrosine phosphatase distinctfrom MKPs (Alessi et al, 1995a), as MKP-1, 2, and 4 predominantlyinactivate JNK and p38 MAPK (Muda et al, 1996; Hirsch andStork, 1997). Further investigation is required to delineate moreprecisely the mechanisms and specific regulators of stress-inducedMAPK inactivation. Our observations suggest that both the ERK1/2 and p38 pathways are not only distinct and differentially regulatedby oxidative stress in keratinocytes, but different types of ROSmay preferentially mediate UVB-induced activation of ERK1/2or p38 signaling pathways.

The role of ERK1/2 activation in cell survival remains controver-sial. In cardiac myocytes, inhibition of ERK increased apoptoticcell death (Aikawa et al, 1997) and cardiotropin 1 promotes survivalvia activation of a signaling pathway requiring ERK (Sheng et al,1997). In hematopoietic cells, activation of ERK partially preventsapoptotic death (Kinoshita et al, 1997). By contrast, ERK1/2inhibition had no significant effect on the insulin- or brain-derivedneurotrophic factor-induced survival of cerebellar granule cells(Gunn-Moore et al, 1997). In this study with keratinocytes,PD098059 specifically inhibits ERK1/2 (Fig 5) as has been reportedfor other cell types (Alessi et al, 1995b; Dudley et al, 1995). Herewe show that inhibition of ERK1/2 increases UVB-induced celldeath, indicating that ERK1/2 plays an important part in protectingkeratinocytes from cytotoxicity following oxidative stress.

Based on these findings, we speculate that the balance betweenpro-oxidative and anti-oxidative mechanisms within the ker-atinocyte will determine the levels of activation of downstreamsignaling cascades such as EGFR/ERK1/2 and p38 that aredifferentially triggered by UVB irradiation. Collectively, theseresults provide strong evidence that ROS, in particular H2O2, actas mediators to activate ERK1/2 and p38 pathways and inducecellular responses following UVB exposure. These observationswill be useful in focusing future investigations on specific UVR-related intracellular signaling pathways of keratinocytes and theepidermis. Early activation of signaling pathways in response toUVR may be involved in the erythema and inflammatory reactions,photoaging various photodermatoses, and carcinogenesis.

We thank Parke-Davis Research Division (Ann Arbor, MI) for the gift ofPD098059. The excellent technical assistance provided by M. Pott, K. Squillace,N.N. Swanson and M. Anderson is gratefully acknowledged. D. Peus wassupported by a grant (PE 635/1-2) from the Deutsche Forschungsgemeinschaft(DFG). R. Vasa was funded by a grant from the Heinrich-Hertz Stiftung,Dusseldorf, Germany. Additional support was provided by the Mayo Foundation.

REFERENCESAikawa R, Komuro I, Yamazaki T, et al: Oxidative stress activates extracellular

signal-regulated kinases through Src and Ras in cultured cardiac myocytes ofneonatal rats. J Clin Invest 100:1813–1821, 1997

Alessi DR, Gomez N, Moorhead G, Lewis T, Keyse SM, Cohen P: Inactivation ofp42 MAP kinase by protein phosphatase 2A and a protein tyrosine phosphatase,but not CL100, in various cell lines. Curr Biol 5:283–295, 1995a

Alessi DR, Cuenda A, Cohen P, Dudley DT, Saltiel AR: PD 098059 is a specific

756 PEUS ET AL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

inhibitor of the activation of mitogen-activated protein kinase kinase in vitroand in vivo. J Biol Chem 270:27489–27494, 1995b

Assefa Z, Garmyn M, Bouillon R, Merlevede W, Vandenheede JR, Agostinis P:Differential stimulation of ERK and JNK activities by ultraviolet B irradiationand epidermal growth factor in human keratinocytes. J Invest Dermatol 108:886–891, 1997

Bender K, Blattner C, Knebel A, Iordanov M, Herrlich P, Rahmsdorf HJ: UV-induced signal transduction. J Photochem Photobiol 37:1–17, 1997

Canman CE, Kastan MB: Three paths to stress relief. Nature 384:213, 1996Cerutti PA: Prooxidant states and tumor promotion. Sci 227:375–381, 1985Chu Y, Solski PA, Der Khosravi-Far RCJ, Kelly K: The mitogen-activated protein

kinase phosphatases PAC1, MKP-1, and MKP-2 have unique substratespecificities and reduced activity in vivo toward the ERK2 sevenmakermutation. J Biol Chem 271:6497–6591, 1996

Cobb MH, Goldsmith EJ: How MAP kinases are regulated. J Biol Chem 270:14843–14846, 1995

Darr D, Fridovich I: Free radicals in cutaneous biology. J Invest Dermatol 102:671–675, 1994

Davis RJ: MAPKs: new JNK expands the group. Trends Biochem Sci 19:470–473, 1994Derijard B, Raingeaud J, Barrett T, Wu IH, Han J, Ulevitch RJ, Davis RJ:

Independent human MAP-kinase signal transduction pathways defined byMEK and MKK isoforms. Science 267:682–685, 1995

Devary Y, Gottlieb RA, Smeal T, Karin M: The mammalian ultraviolet response istriggered by activation of Src tyrosine kinases. Cell 71:1081, 1992

Dudley DT, Pang L, Decker SJ, Bridges AJ, Saltiel AR: A synthetic inhibitor of themitogen-activated protein kinase cascade. Proc Natl Acad Sci, USA 92:7686–7689, 1995

Gilchrest BA, Park HY, Eller MS, Yaar M: Mechanisms of ultraviolet light-inducedpigmentation. Photochem Photobiol 63:1–10, 1996

Gunn-Moore FJ, Williams AG, Toms NJ, Tavare JM: Activation of mitogen activatedprotein kinase and P70S6 kinase is not correlated with cerebellar granule cellsurvival. Biochem J 324:365–369, 1997

Guyton KZ, Liu Y, Gorospe M, Xu Q, Holbrook NJ: Activation of mitogen-activated protein kinase by H2O2. Role in cell survival following oxidantinjury. J Biol Chem 271:4138–4142, 1996

Herrlich P, Ponta H, Rahmsdorf HJ: DNA damage-induced gene expression: signaltransduction and relation to growth factor signaling. Rev Physiol BiochemPharmacol 119:187–223, 1992

Herrlich P, Sachsenmaier C, Radler-Pohl A, Gebel S, Blattner C, Rahmsdorf HJ:The mammalian UV response: Mechanism of DNA damage induced geneexpression. Adv Enzyme Regul 34:381–395, 1994

Hirsch DD, Stork PJS: Mitogen-activated protein kinase phosphatases inactivatestress-activated protein kinase pathways in vivo. J Biol Chem 272:4568–4575, 1997

Huang RP, Wu JX, Fan Y, Adamson ED: UV activates growth factor receptors viareactive oxygen intermediates. J Cell Biol 133:211–220, 1996

Karin M: The regulation of AP-1 activity by mitogen-activated protein kinases.J Biol Chem 270:16483–16486, 1995

Keyse SM: An emerging family of dual specificity MAP kinase phosphatases. BiochimBiophys Acta 1265:152–160, 1995

Kinoshita T, Shirouzu M, Kamiya A, Hasmimoto K, Yokoyama S, Miyajima A:RAF/MAPK and rapamycin-sensitive pathways mediate the antiapoptoticfunction of P21RAS in IL3-dependent hematopoietic cells. Oncogene 15:619–627, 1997

Kripke ML: Ultraviolet radiation and immunology: Something new under the sun.Cancer Res 54:6102–6105, 1994

Matsui MS, DeLeo VA: Longwave ultraviolet radiation and promotion of skin cancer.Cancer Cells 3:8–12, 1991

Miller DL, Weinstock MA: Nonmelanoma skin cancer in the United States:incidence. J Am Acad Dermatol 30:774–778, 1994

Mohanty JG, Jaffe JS, Schulman ES, Raible DG: A highly sensitive fluorescentmicro-assay of H2O2 release from activated human leukocytes using adihydroxyphenoxazine derivative. J Immunol Methods 202:133–141, 1997

Muda M, Boschert U, Dickinson R, et al: Mkp-3, a novel cytosolic protein-tyrosinephosphatase that exemplifies a new class of mitogen-activated protein kinasephosphatase. J Biol Chem 271:4319–4326, 1996

Norris PG, Gange RW, Hawk JLM. Acute effects of ultraviolet radiation on the

skin In: Fitzpatrick TB, Eisen AZ, Wolff K, Freedberg IM, Austen KF (eds).Dermatology in General Medicine, 4th edn. New York: McGraw-Hill, 1993,pp 1651–1658

Peus D, Hamacher L, Pittelkow MR: EGF-receptor tyrosine kinase inhibitioninduces keratinocyte growth arrest and terminal differentiation. J Invest Dermatol109:101–106, 1997

Peus D, Vasa R, Meves A, Pott M, Beyerle A, Squillace K, Pittelkow MR: H2O2is an important mediator of UVB-induced EGF-receptor phosphorylation incultured keratinocytes. J Invest Dermatol 110:966–971, 1998

Pittelkow MR, Scott RE: New techniques for the in vitro culture of human skinkeratinocytes and perspectives on their use for grafting of patients with extensiveburns. Mayo Clin Proc 61:771–777, 1986

Raingeaud J, Gupta S, Rogers JS, Dickens M, Han J, Ulevitch RJ, Davis RJ: Pro-inflammatory cytokines and environmental stress cause p38 mitogen-activatedprotein kinase activation by dual phosphorylation on tyrosine and threonine.J Biol Chem 270:7420–7426, 1995

Raingeaud J, Whitmarsh AJ, Barrett T, Derijard B, Davis RJ: MKK3- and MKK6-regulated gene expression is mediated by the p38 mitogen-activated proteinkinase signal transduction pathway. Mol Cell Biol 16:1247–1255, 1996

Rao GN: Hydrogen peroxide induces complex formation of SHC-Grb2-SOS withreceptor tyrosine kinase and activates Ras and extracellular signal-regulatedprotein kinases group of mitogen-activated protein kinases. Oncogene 13:713–719, 1996

Romerdahl CA, Stephens LC, Bucana C, Kripke ML: The role of ultravioletradiation in the induction of melanocytic skin tumors in inbred mice. CancerCommun 1:209–216, 1989

Rouse J, Cohen P, Trigon S, et al: A novel kinase cascade triggered by stress andheat shock that stimulates MAPKAP kinase-2 and phosphorylation of the smallheat shock proteins. Cell 78:1027–1037, 1994

Royal JA, Ischiropoulos H: Evaluation of 29,79-dichlorofluorescin anddihydrorhodamine 123 as fluorescent probes for intracellular H2O2 in culturedendothelial cells. Arch Biochem Biophys 302:348–355, 1993

Sachsenmaier C, Radler-Pohl A, Zinck R, Nordheim A, Herrlich P, RahmsdorfHJ: Involvement of growth factor receptors in the mammalian UVC response.Cell 78:963–972, 1994

Sheng ZL, Knowlton K, Chen J, Hoshijima M, Brown JH, Chien KR: Cardiotropin1 (CT-1) inhibition of cardiac myocyte apoptosis via a mitogen-activatedprotein kinase-dependent pathway – divergence from downstream CT-1 signalsfor myocardial cell hypertrophy. J Biol Chem 272:5783–5791, 1997

Sonoda Y, Kasahara T, Yamaguchi Y, Kuno K, Matsushima K, Mukaida N:Stimulation of interleukin-8 production by okadaic acid and vanadate in ahuman promyelocyte cell line, an hl-60 subline – possible role of mitogen-activated protein kinase on the okadaic acid-induced Nf-kappa-B activation.J Biol Chem 272:15366–15372, 1997

Su B, Karin M: Mitogen-activated protein kinase cascades and regulation of geneexpression. Curr Opin Immunol I 8:402, 1996

Tornaletti S, Pfeifer GP: UV damage and repair mechanisms in mammalian cells.Bioessays 18:221–228, 1996

Whitmarsh AJ, Shore P, Sharrocks AD, Davis RJ: Integration of MAP kinase signaltransduction pathways at the serum response element. Science 269:403–407, 1995

Wilhelm D, Bender K, Knebel A, Angel P: The level of intracellular glutathione isa key regulator for the induction of stress-activated signal transduction pathwaysincluding jun n-terminal kinases and p38 by alkylating agents. Mol Cell Biol17:4792–4800, 1997

Wille JJ Jr, Pittelkow MR, Shipley GR, Scott RE: Integrated control of growth anddifferentiation of normal human prokeratinocytes cultures in serumfree medium:Clonal analyses, growth kinetics and cell cycle studies. J Cell Physiol 121:31–44, 1984

Xia Z, Dickens M, Raingeaud J, Davis RJ, Greenberg ME: Opposing effects ofERK and JNK-p38 MAP kinases on apoptosis. Science 270:1326–1331, 1995

Young AR. Chronic effects of ultraviolet radiation on the skin: Experimental aspects.In: Fitzpatrick TB, Eisen AZ, Wolff K, Freedberg IM, Austen KF (eds).Dermatology in General Medicine, 4th edn. New York: McGraw-Hill, 1993, pp1658–1660

Zhao Z, Tan Z, Diltz CD, You M, Fischer EH: Activation of mitogen-activatedprotein (MAP) kinase pathway by pervanadate, a potent inhibitor of tyrosinephosphatases. J Biol Chem 271:22251–22255, 1996