UVB-induced activation and internalization of keratinocyte growth factor receptor

UVB-induced activation and internalization of keratinocyte growth factor

receptor

Cinzia Marchese*,1,3, Vittoria Maresca2,3, Giorgia Cardinali2, Francesca Belleudi1, SimonaCeccarelli1, Marinella Bellocci2, Luigi Frati1, Maria Rosaria Torrisi1,2 and Mauro Picardo2

1Dipartimento di Medicina Sperimentale e Patologia, Universita di Roma ‘La Sapienza’, viale Regina Elena 324, Rome 00161, Italy;2Istituto Dermatologico Santa Maria e San Gallicano. IRCCS, Rome, Italy

Ultraviolet irradiation of mammalian cells induces severalevents that include activation of growth factor receptorsand triggering of signal transduction pathway. Most of theUV responses are mediated by the production of reactiveoxygen species (ROS) and can be blocked by antiox-idants. In this study, we analysed the effect of UVBirradiation at physiologic doses and that of the pro-oxidant agent cumene hydroperoxide (CUH) on theactivation of the receptor for keratinocyte growth factor(KGF), a key mediator of epithelial growth and differ-entiation. Exposure to both UVB (30–150mJ/cm2) andCUH (200 lm of NIH3T3 KGFR (KGF receptors)transfectants caused a rapid tyrosine phosphorylationand activation of KGFR similar to that induced by KGF,and internalization of the activated receptor. The KGFRexpression appeared unmodified by the treatments.Ultrastructural observations of both UVB- and CUH-treated cells showed a normal morphology of the plasmamembranes and intracellular organelles. The antioxidantN-acetylcysteine inhibited UVB-induced receptor phos-phorylation. The generation of an intracellular oxidativestress was detected as a decrease of catalase activity andof vitamin E, and reduced glutathione levels, whereassuperoxide dismutase activity was not significantlymodified. A peroxidation of polyunsaturated fatty acidsof cell membranes was observed after both treatments,associated with the intracellular oxidative stress. Similarbiochemical events were observed on NIH3T3 untrans-fected control cells, suggesting that KGFR activationfollows intracellular generation of ROS and is notassociated with a scavenging effect. Taken together ourresults demonstrate that exposure to UVB and to oxidantstimuli induces a rapid intracellular production of ROS,which in turn are capable of triggering KGFR activationand internalization, similar to those induced by KGF.Oncogene (2003) 22, 2422–2431. doi:10.1038/sj.onc.1206301

Keywords: UVB; keratinocyte growth factor receptor;oxidative stress; tyrosine phosphorylation

Introduction

Ultraviolet irradiation (UV) of human skin is known tobe involved in the pathogenesis of several disorderscharacterized by inflammation, hyperpigmentation andhyperproliferation, and possibly leading to tumorigenesis(Fisher et al., 1997; Glenn McGregor, 1999; Wikonkaland Brash, 1999). Exposure to UV radiation causes acomplex cellular response, which includes the regulationof gene expression, DNA damage, lipid peroxidation andinduction of apoptosis (Schwarz et al., 1995; Rosette andKarin, 1996; Sheikh et al., 1998; Glenn McGregor, 1999;Wikonkal and Brash, 1999). UV radiation, causingDNA damage by either direct interactions or by indirectpathways, may play a pivotal role in the process ofcarcinogenesis (for a recent review, see de Gruijl, 2000).The molecular mechanisms involved in the triggering ofsuch cellular response are still largely unknown; how-ever, most of the effects induced by UV seems to dependon the production of reactive oxygen species (ROS) thatmay act as second messengers (Fuchs and Packer, 1990;Dimon-Gadal, 1998).The earliest step in response to UV irradiation and

subsequent ROS generation involves ligand-indepen-dent phosphorylation of growth factor receptors such asepidermal growth factor receptor (EGFR) (Sachsenma-ier et al., 1994; Coffer et al., 1995; Huang et al., 1996;Peus et al., 1998, 1999), signal transduction (Sachsen-maier et al., 1994; Rosette and Karin, 1996) and rapidexpression of early growth response genes (Egr-1)(Huang and Adamson, 1995).Keratinocyte growth factor (KGF/FGF7) is a mem-

ber of the fibroblast growth factor (FGF) family andappears to represent a key mediator of epithelial growthand differentiation. Secreted from cells of mesenchymalorigin, it elicits its activity specifically on epithelial cells(Finch et al., 1989; Rubin et al., 1989) through bindingto the KGF receptor (KGFR), a splicing transcriptvariant of the FGFR2 (Miki et al., 1991, 1992). Oncultured human keratinocytes, KGF acts not only as apotent mitogen (Marchese et al., 1990), but alsopromotes their early differentiation (Marchese et al.,1990) and inhibits their terminal differentiation andapoptosis (Hines and Allen-Hoffman, 1996). In vivo,KGF appears to play a role in experimental (Werner

Received 28 October 2002; revised 26 November 2002; accepted 5December 2002

*Correspondence: C Marchese; E-mail: [email protected] equally to this work.

Oncogene (2003) 22, 2422–2431& 2003 Nature Publishing Group All rights reserved 0950-9232/03 $25.00

www.nature.com/onc

et al., 1992, 1994; Staiano-Coico et al., 1993) and human(Marchese et al., 1995) wound healing. UVB irradiationof human cultured keratinocytes induces, as a late event(24 h), drastic downmodulation of KGFR expression,and it has been proposed that this KGFR down-regulation may account for the therapeutic effect ofgrowth inhibition induced by UVB in the treatment ofhyperproliferative skin disorders such as psoriasis (Zhouet al., 1996; Finch et al., 1997).Here we analysed the very early effects of physiolo-

gical doses of UVB irradiation on KGFR expression,phosphorylation and activation using NIH3T3 cellstransfected with KGFR. In addition, since it has beenproposed that UVB-induced phosphorylation of EGFRis mediated by the production within minutes ofhydrogen peroxide (H2O2) (Peus et al., 1999), wecompared the effects induced by UVB with thoseexerted by treatment with cumene hydroperoxide(CUH), a synthetic and stable hydroperoxide able toinduce the production of H2O2 (Vessey et al., 1995;Kayanoki et al., 1996). The modification of someenzymatic and nonenzymatic antioxidants and thepattern of membrane fatty acid were evaluated to studythe generation of an oxidative stress. Moreover, in orderto evaluate whether KGFR activation by UV couldexert a free radical scavenger role, untransfectedNIH3T3, grown in the same conditions, were alsostudied.

Results

UVB induces tyrosine phosphorylation and activation ofKGFR

To analyse the effects of physiologic doses of UVBirradiation on the expression and phosphorylation ofKGFRs, we exposed KGFR-transfected NIH3T3 cellsto increasing doses of UVB (20–150mJ/cm2) andincubated for 5–20min, followed by cell viabilityanalysis and immunofluorescence with antiphosphotyr-osine antibodies (data not shown). We selected the UVBdoses (20–50mJ/cm2) and the time point of incubationafter UVB exposure (10min) corresponding to anevident immunofluorescence signal on the cell surfaces.We then exposed NIH3T3 KGFR transfectants to UVB(20, 30 and 50mJ/cm2) and incubated the cells for10min before immunoprecipitation with anti-Bekmonoclonal or polyclonal antibodies, which specificallyrecognize the intracellular domain shared by the twosplicing isoforms, FGFR2 and KGFR, and immuno-blotting with antiphosphotyrosine antibodies. A proteinreacting with anti-Bek antibodies and corresponding tothe molecular weight of the KGFR was visible inKGFR-transfected NIH3T3 cells (Figure 1a), as ex-pected (Marchese et al., 1998), and no reactivity wasdetected in control nontransfected NIH3T3 cells(Marchese et al., 1998 and data not shown). As shownin Figures 1–3, the KGFR protein levels were similar inuntreated as well as in UVB-treated cells. Although inour experimental conditions the KGFR expression

appeared unmodified, exposure to UVB as aboveinduced receptor tyrosine phosphorylation, as assessedby either immunoprecipitation with anti-Bek followedby immunoblotting with antiphosphotyrosine(Figure 1a) or immunoprecipitation with antiphospho-tyrosine followed by immunoblotting with anti-Bek(Figure 1b), which appeared to reach the plateau atthe UVB dose of 30mJ/cm2. Similar tyrosine phosphor-ylation of KGFR was induced by treatment with KGF(100 ng/ml), as expected (Marchese et al., 1998)(Figure 1b). Weak tyrosine phosphorylation of KGFRwas detected also in KGF-untreated cells (Figure 1a, b),most likely as a consequence of autocrine receptoractivation, since NIH3T3 cells are known to producemouse KGF (Miki et al., 1991).Since the UVB exposure is known to generate ROS

and to induce an oxidative stress, we exposed the cells toUVB (30mJ/cm2) in the presence of antioxidants such asN-acetyl cysteine (NAC) (10mm) or pyrrolidine dithio-carbamate (PDTC) (50 mm) and the UVB-inducedreceptor phosphorylation appears to be inhibited(Figure 2). Since treatment with antioxidants have beenshown to inhibit the UVB-induced KGFR phosphor-ylation, we compared the effects of UVB on KGFRswith those exerted by treatment of the NIH3T3 KGFRcells with the pro-oxidant agent CUH. Immunoprecipi-tation followed by immunoblotting experiments re-vealed that a similar tyrosine phosphorylation ofKGFR was induced by UVB (30mJ/cm2) and CUH(100 and 200 mm), and that KGFR protein levels weresimilar in CUH-treated cells compared with theuntreated and the UVB-treated cells (Figure 3).To demonstrate that KGFR phosphorylation induced

by UVB irradiation is a general phenomenon occurringalso in cells endogenously expressing KGFR, we usedhuman keratinocyte HaCaT cells grown at confluence toexpress higher levels of KGFR (Capone et al., 2000).Immunoprecipitation followed by immunoblotting ana-lysis showed that UVB exposure (30 and 50mJ/cm2)stimulated, similar to KGF, phosphorylation of aprotein species corresponding to the molecular weightof KGFR (Figure 4).To evaluate if UVB irradiation and CUH treatment

at the doses used could determine possible alterationsin cell morphology, we examined at ultrastructurallevel the UVB-treated and CUH-treated NIH3T3KGFR-transfected cells. Electron microscopicobservations revealed a normal unaltered morphologyin treated cells compared to the untreated cells (data notshown).To demonstrate that UVB exposure induces KGFR

activation and signalling by the receptor, NIH3T3KGFR transfectants treated with KGF or irradiatedwith UVB were analysed by immunoprecipitation andWestern blot to detect possible phosphorylation of theFGF receptor substrate 2 (FRS2) (Hadari et al., 2001)and of PLCg, which is known to represent the majorsubstrate of KGFR (Shaoul et al., 1995), as well as thephosphorylation and activation of extracellular signal-regulated-kinase (ERK1), a signal transduction elementdownstream to KGFR (Shaoul et al., 1995). Immuno-

UVB-induced KGFR activation and internalizationC Marchese et al

2423

Oncogene

precipitation with anti-FRS2 polyclonal antibodiesfollowed by immunoblot with antiphosphotyrosineantibody or immunoprecipitation with antiphosphotyr-osine antibody followed by immunoblot with anti-PLCgmonoclonal antibody mAb, as well as Western blot withantiphospho-(ERK1) mAb, confirmed that UVB ex-posure is able to induce activation and signal transduc-tion of KGFR similar to that elicited by the KGF ligand(Figure 5).Thus, KGFRs can be efficiently tyrosine phosphory-

lated and activated by UVB treatment as well as by anoxidant stimulus and this event occurs shortly after theexposure and in the absence of detectable morphologicalalterations. Since NAC and PDTC are able to inhibitUVB-induced phosphorylation of the receptor, themechanism of UVB-dependent KGFR phosphorylationappears to involve intracellular production of ROS.

To determine the effect of UVB exposure on DNAsynthesis and cell proliferative activity, we performedthe 50-Bromo-deoxyuridine (BrdU) incorporation assay.NIH3T3 KGFR and NIH3T3 untransfected controlcells were serum-starved for 12 h and then exposed toUVB 30mJ/cm2 and replaced for an additional 12 h inthe medium alone before BrdU incorporation. TheDNA synthesizing cells were then visualized with anti-BrdU monoclonal antibody. When cells were serum-starved for 12 h and kept in the medium alone for anadditional 12 h, the percentage of cells showing nucleipositively stained for BrdU (Figure 6, arrows) was low(43 and 25% for NIH3T3 KGFR and NIH3T3 cells,respectively), compared with the percentage of cells keptin medium plus serum (70 and 58% for NIH3T3 KGFRand NIH3T3, respectively). UVB exposure after 12 h ofstarvation induced a drastic decrease (becoming ap-

Figure 1 Tyrosine phosphorylation and expression of KGFR in NIH3T3 KGFR-transfected cells and NIH3T3 control cells exposedto different doses of UVB (20, 30 and 50mJ/cm2): immunoprecipitation with anti-Bek antibodies, which specifically recognize theintracellular domain shared by the two splice variants, FGFR2 and KGFR, followed by incubation for 10min and immunoblottingwith antiphosphotyrosine antibodies or with anti-Bek antibodies (a) or immunoprecipitation with antiphosphotyrosine antibodiesfollowed by immunoblotting with anti-Bek antibodies (b) show that KGFR tyrosine phosphorylation is induced by UVB irradiation.KGFR tyrosine phosphorylation is similarly induced by treatment with KGF (100 ng/ml) and a weak band is detected also in KGF-untreated cells, most likely as a consequence of autocrine receptor activation


2424

Oncogene

proximately 1%) in the number of BrdU-positive nuclei,suggesting that UVB irradiation causes a block of theproliferative activity and the UVB-induced KGFRactivation and signal transduction do not stimulateDNA synthesis.

UVB irradiation induces internalization of activatedKGFRs

To visualize the phosphorylation of KGFR in responseto UVB irradiation at the cell plasma membranes and toanalyse the possible internalization of the activatedreceptors, NIH3T3 KGFR transfectants were analysedby immunofluorescence microscopy. Cells were exposedto UVB as above and fixed (Figure 7b) or warmed to371C for 30min to allow receptor internalization(Figure 7c). Alternatively, cells were treated with KGFfor 2 h at 41C before fixation to induce ligand bindingand subsequent receptor activation and phosphorylation(Figure 7a). KGFRs were immunolabelled with anti-

Figure 2 Inhibition of tyrosine phosphorylation of KGFR in NIH3T3 KGFR-transfected cells following exposure to UVB (30mJ/cm2) in the presence of the antioxidants NAC (10mm) or pyrrolidine dithiocarbamate (PDTC) (50 mm) and incubation for 10minbefore immunoprecipitation with anti-Bek polyclonal antibodies followed by immunoblotting with antiphosphotyrosine or with anti-Bek antibodies. The KGFR protein levels are similar in untreated as well as in cells exposed to UVB or to UVB in the presence ofantioxidants

Figure 3 Tyrosine phosphorylation of KGFR in NIH3T3 KGFR-transfected cells and NIH3T3 control cells exposed to UVB (30mJ/cm2) or treated with CUH (100 and 200mm) detected by immunoprecipitation with anti-Bek polyclonal antibodies followed byimmunoblotting with antiphosphotyrosine or with anti-Bek antibodies

Figure 4 Tyrosine phosphorylation of KGFR following exposureto UVB (30 and 50mJ/cm2) or treatment with KGF (100 ng/ml) inHaCaT keratinocytes revealed by immunoprecipitation with anti-Bek polyclonal antibodies followed by immunoblotting withantiphosphotyrosine or with anti-Bek antibodies


2425

Oncogene

Bek antibodies (red signal) and double stained withantiphosphotyrosine antibodies (green signal) to identi-fy, by colocalization of the two signals (in yellow),KGFRs tyrosine phosphorylated at the cell plasmamembranes (Figure 7a, b). Although anti-Bek anti-bodies do not discriminate among the two FGFR2splice variants, transfected cells expressing only unde-tectable level of KGFR among the KGFR overexpres-sing ones (Figure 7, arrowheads) were virtuallyunstained, indicating that the anti-Bek antibodiesprimarily detected the exogenously overexpressedKGFRs. After UVB exposure and warming to 371C,the immunofluorescence signals colocalized in intracel-lular dots located in both the peripheral and central areaof the cell (Figure 7c), suggesting receptor internaliza-tion and intracellular accumulation of the activated

receptors. Thus, short physiological UVB exposure wasable to induce tyrosine phosphorylation, activation andinternalization of KGFR similar to the KGF ligand.

Treatment with peroxidative agents induces alterations inthe membrane pattern of fatty acids and in the levels ofenzymatic and chemical antioxidants in both NIH3T3KGFR transfectants and NIH3T3 cells

In order to verify if peroxidative treatments were able toproduce alterations in the membrane pattern of fattyacids and in the intracellular levels of enzymatic andchemical antioxidants, NIH3T3 KGFR-transfected cellsand NIH3T3 cells were irradiated with UVB (30mJ/cm2) or incubated in a culture medium containing CUH(200 mm).

Figure 5 (a, b) Tyrosine phosphorylation of FRS2 and PLCg in NIH3T3 KGFR-transfected cells following exposure to UVB (30mJ/cm2) or treatment with KGF (100 ng/ml), detected by immunoprecipitation with anti-FRS2 antibodies followed by immunoblottingwith antiphosphotyrosine antibody, where the equal loading was assessed by reprobing with anti-FRS2 antibodies (a), or byimmunoprecipitation with antiphosphotyrosine antibody followed by immunoblotting with anti-PLCg antibody (b). (c) UVB-inducedactivation of ERK1 in NIH3T3 KGFR cells. Low levels of ERK1 phosphorylation appear to be induced by UVB exposure also inNIH3T3 control cells


2426

Oncogene

After UVB irradiation, the percentage of polyunsa-turated fatty acids (PUFA) of the cell membranes, oneof the main targets of peroxidative damage, decreasedup to 73.3 and 84.6% of the control in both transfectedand untransfected cells, respectively (see Tables 1and 2). A similar effect was observed with regard tothe levels of chemical antioxidants; in fact after UVBirrediation, vitamin E concentration decreased up to82.2 and 85.1% and GSH up to 68.1 and 70.2%, inboth NIH3T3 KGFR and NIH3T3 cells (see Tables 1and 2).Treatment with CUH produced analogous modifica-

tions in the PUFA percentage in the cell membranes andin the intracellular concentration of chemical antiox-idants, in both transfected and untransfected cells (seeTables 1 and 2).When antioxidant enzymatic activities were evalu-

ated, before and after the treatment with UVB, adecrease in CAT activity, up to 71.3 and 78.8%, wasobserved in untransfected and transfected cells, respec-tively, and similar results were obtained after CUHtreatment (see Tables 1 and 2). On the contrary,superoxide dismutase (SOD) activity, in both trans-fected and untransfected cells, was not statistically

modified by the peroxidative treatments with UVB orCUH (see Tables 1 and 2). Therefore, the alteration ofantioxidants observed in both NIH3T3 KGFR-trans-fected cells and NIH3T3 cells reflects the presence ofoxidative stress induced by UVB and CUH treatmentand could be the basis for an intracellular generation ofH2O2, as previously described in keratinocyte cultures(Peus et al., 1999).

Discussion

Ligand-independent phosphorylation of receptor tyro-sine kinases, induced by UV irradiation, is known torepresent a rapid, membrane-associated event (Sachsen-maier et al., 1994; Coffer et al., 1995; Huang et al., 1996;Rosette and Karin, 1996). Here we show that also thereceptor for KGF, which plays a fundamental role inregulating keratinocyte growth and differentiation, israpidly phosphorylated by exposure to physiologicaldoses of UVB with timing and efficiency similar to thatinduced by binding of the ligand to the receptor. Nodifferences in KGFR expression are observed at 10minafter UVB exposure, a time point too early to expectreceptor degradation and downregulation.

Figure 6 BrdU incorporation assay in NIH3T3 KGFR and NIH3T3 cells performed 12 h after UVB irradiation. Compared withunexposed culture, either serum-starved and kept in medium alone or grown in the continuous presence of serum, UVB exposureblocks BrdU incorporation, as shown by the drastic decrease in the number of BrdU-positive nuclei (arrows). Bar: 10mm


2427

Oncogene

We have recently demonstrated that KGF binding tothe receptor triggers a rapid receptor clustering in theclathrin-coated pits and internalization of ligand–receptor complexes (Marchese et al., 1998). Here weshowed that UVB exposure also induces a rapidendocytosis of the phosphorylated KGFRs which,30min after irradiation, appear localized in intracellularperipheral and perinuclear structures, presumably re-presenting early and late endosomes. These findings not

only are in agreement with previous observations onother receptor tyrosine kinases, such as EGFR andPDGFR (Rosette and Karin, 1996), but also mayexplain the KGFR downregulation observed at 6–24 hafter UVB exposure (Zhou et al., 1996; Finch et al.,1997) and the decrease in the proliferative activity andDNA synthesis observed in our experiments 12 h afterirradiation.The parallel receptor phosphorylation induced by

irradiation with UVB and treatment with CUH, as wellas the significant reduction of the UVB-triggeredphosphorylation, observed after cotreatment with theantioxidant NAC, strongly imply that the UVB effect onKGFR is mediated by the intracellular generation ofH2O2 and ROS, as reported for EGFR (Peus et al.,1998, 1999). The biochemical results obtained indicatethe mechanism of H2O2 generation inside the cells.Antioxidants operate in combination: SOD dismutatessuperoxide anion radicals generating H2O2 and oxygenand catalase (Cat) and GSH Px remove the H2O2

produced. However, Cat is the main enzyme involvedduring a significant oxidative stress. When the ratio ofthese enzymes is altered, H2O2 is accumulated inside thecells and can produce OH?, which rapidly reacts withlipids, protein and DNA, propagating the dangerouseffect. Chemical antioxidants inhibit or reduce theseevents (Fuchs and Packer, 1990). The alteration of theratio of SOD and Cat activities (ratio SOD/Cat),observed immediately after both UV and CUH treat-ments, suggests that H2O2 is generated through thismechanism. Moreover, the peroxidation of the unsatu-rated lipids compound of the cell membrane, associatedwith the significant depletion of the chemical antiox-idants, indicates that ROS generation and the conse-quent oxidative stress takes place inside the cells. In thisconnection, the overexpression of Cat in keratinocytesby electroporation significantly reduces the amount ofintracellular H2O2 and the phosphorylation of theEGFR following UVB irradiation (Peus et al., 1999).Growing evidence indicates that H2O2 may serve as

second messenger that can activate transcription factors,such as NF-kB (Huang et al., 1996), and it could bespeculated that these events may also have a scavengingeffect. The evaluations performed in comparison be-tween transfected and nontransfected NIH3T3 revealedthat the expression of the KFGR was associated with a

Figure 7 Immunofluorescence analysis of the activation and inter-nalization of KGFRs induced by UVB exposure. NIH3T3 KGFRtransfectants were incubated with KGF for 2 h at 41C or exposed toUVB and either immediately fixed or warmed to 371C for 30min toallow receptor internalization. Double immunofluorescence stainingwith anti-Bek polyclonal antibodies (red: rhodamine staining) and withantiphosphotyrosine mAb (green: fluorescein staining) shows coloca-lization of the two signals (in yellow after overlapping of the singleimages) at the cell plasma membranes following KGF binding at 41C(a) or UVB exposure (b) and in intracellular dots after receptorinternalization induced by warming to 371C after UVB irradiation (c).Arrowheads point to NIH3T3 cells that do not express KGFRs andtherefore appear unstained or very weakly stained by both anti-Bekand antiphosphotyrosine antibodies. Colocalization of the two signals,assessed by overlapping the single images, is shown in yellow. Bars:5 mm


2428

Oncogene

different enzymatic antioxidant pattern and of thePUFA percentage (Table 1 and 2) even if the culturedconditions were the same. We are not able to explaincompletely this phenomenon at this moment; however,different growth factors induce intracellular generationof H2O2 (Ohba et al., 1994; Bae et al., 1997) and theautocrine loop, in NIH3T3 KGFR-transfected cells,may lead to a decrease of Cat activity. Alternatively, theexpression of KGFR may prone the cells to a differentsensitivity to external stimuli. In any case, however, theoxidative stress, generated following UVB or CUHexposition, as evaluated by the biochemical parameters,was similar in percentage in both transfected anduntransfected cells, suggesting that KGFR activationfollows intracellular generation of ROS and is notassociated with a scavenging effect.In conclusion, our data demonstrate that UVB

irradiation is able to activate KGFR through thegeneration of intracellular ROS in a manner similar tothat observed for other growth factor receptors. SinceKGFRs, in contrast to EGFRs, appear to be upmodu-lated during keratinocyte differentiation (Marcheseet al., 1997), we may suggest that similar UVB-inducedphosphorylation of both EGFR and KGFR, differentlyexpressed on basal and superbasal skin layers, isrequired for a complete response involving eitherproliferating or differentiated cells.

Materials and methods

Cell lines

NIH3T3 cells transfected with the pCEV27 vector containinghuman KGFR cDNA (Miki et al., 1991; LaRochelle et al.,1995) and untransfected NIH3T3 cells were cultured inDulbecco’s modified Eagle’s medium (DMEM) supplementedwith 10% bovine calf serum plus antibiotics. HaCaT cells were

cultured in DMEM supplemented with 10% FCS andantibiotics, and allowed to grow for 11 days to reach theconfluence.For KGF treatment, all cells were serum starved for 12 h

and then incubated with KGF (UBI, Lake Placid, NY, USA)(100 ng/ml in a medium containing 0.3m NaCl) at 371C for10min.Radiation of cells, between 270 and 400 nm, peaking at

310 nm, was delivered either from UVB lamp, Philips TL20W/12 (Philips GmbH, Hamburg, Germany), housed in a UV800U (Waldmann GmbH, VS-Schwenningen, Germany) orfrom VL6-M Biotronic Device (Vilber Lourmat, Marne LaVallee, France). The irradiance was regularly quantified by acalibrated radiometer equipped with an SCS 280 photodetec-tor (International Light, Newburyport, MA, USA) and was0.59mW/cm2 at a distance of 25 cm. UVB irradiation wasperformed with a dose of 20, 30, 50mJ/cm2 in the presence orabsence of NAC (10mm, BDH-chemicals, Poole, Dorset, UK)or PDTC (50 mm, Sigma Chemical Co, St Louis, USA). Forperoxidative treatment, CUH (Fluka Chemika, AG, Buchs,Switzerland) was dissolved in ethanol at 1mg/ml and diluted inthe culture medium. Cells, plated in 75 cm2 flasks, wereincubated with a concentration of 200 mm, CUH for 10minin a medium without serum. After treatment, cell layers werewashed twice in PBS without Ca2+ and Mg2+, scraped with arubber policemen and centrifuged at 10 000 g for 10min at 41C.Cell pellets were then processed for analysis of antioxidantsand fatty acids of cell membranes.

Immunoprecipitation and Western blot analysis

Subconfluent cultures of NIH3T3 KGFR transfectants werelysed in RIPA buffer (10mm Tris pH 7.4; 50mm NaCl; 1mm

EDTA; 10mm KCl; 1% NP40; 0.1% SDS; 0.05% Tween 20)supplemented with protease inhibitors (10 mg/ml aprotinin,1mm PMSF, 10mg/ml leupeptin) and phosphate inhibitors(25mm, sodium orthovanadate, 20mm, sodium pyropho-sphate, 0.5m Naf); 50 mg of total proteins were resolved underreducing conditions by 7% SDS–PAGE and transferred toreinforced nitrocellulose (BA-S 83, Scheleider & Schuell,

Table 1 PUFA percentage and antioxidants levels in NIH3T3 KGFR-transfected cells before and after treatment with pro-oxidant agents

PUFA (% tot. fatty acids) SOD (U/mg prot.) Cat (U/mg prot.) Vit. E (ng/mg prot.) GSH (nmol/mg prot.)

Basal levels 1572 4.670.22 11.572.2 225722 0.2270.018% 100 100 100 100 100UVB 1172* 5.170.67 8.271.5* 185731* 0.1570.027**% 73.3 110.8 71.3 82.2 68.1CUH 1271.9* 5.470.6 8.171.8* 190721* 0.1770.025**% 80.1 117.3 70.4 84.4 77.2

Samples were irradiated with a doses of UVB of 30mJ/cm2, peroxidative treatment was performed incubating the samples, for 10min, in CUH200mm. Each result represents the mean7s.d. of three different experiments in duplicate. In the treated samples results were also expressed as apercentage of modification in comparison with the values observed in nontreated samples. Po0.001, *Po0.05, **Po0.005.

Table 2 PUFA percentage and antioxidants levels in NIH3T3 cells before and after treatment with pro-oxidant agents

PUFA (% tot. fatty acids) SOD (U/mg prot.) Cat (U/mg prot.) Vit E (ng/mg prot.) GSH (nmol/mg prot.)

Basal levels 3972.2 4.670.28 21.371.4 200723 0.270.02% 100 100 100 100 100UVB 3371.8** 4.370.45 16.870.75** 170722* 0.1470.04*% 84.6 93.4 78.8 85.1 70.2CUH 3471.17** 4.1170.5 17.570.9** 165718* 0.1570.03*% 87.2 89.3 82.1 82.5 75.1

*Po0.001 vs NIH3T3 KGFR-transfected cells. See footnote of Table 1.


2429

Oncogene

Keene, NH, USA). The membranes were blocked with 5%nonfat dry milk in PBS 0.1% Tween 20, and incubated withantiphospho-extracellular signal-regulated-kinase (ERK)monoclonal antibody (Cell Signaling Technology Inc., Bev-erly, MA, USA) diluted 1 : 2000 overnight at 41C, followed byenhanced chemiluminescence detection (ECL) (AmershamCorp., Arlington Heights, IL, USA). To estimate the proteinequal loading, the membrane was exstensively washed in PBSTween 20 and reprobed with antitotal-ERK rabbit antibody(K-23, Santa Cruz Biotechnology Inc, CA) diluted 1 : 1000 for1 h at 251C, followed by ECL detection. For immunoprecipi-tation/immunoblotting experiments, NIH3T3 KGFR cells,NIH 3T3 or HaCaT cell line, after treatments, were lysed in1% Triton X-100, 50mm HEPES buffer containing 150mm

NaCl, 1% glycerol, 1.5mm MgCl2, 5mm EGTA and proteaseand phosphatase inhibitors as above. Total protein (1mg) wasimmunoprecipitated with 4mg/ml rabbit anti-Bek polyclonalantibodies, diluted 1 : 200 (C-17, Santa Cruz Biotecnology Inc,CA, USA), raised against the carboxy-terminal region of theintracellular domain of FGFR2/KGFR, or with antipho-sphotyrosine, mAb, diluted 1 : 1000 (Upstate Biotechnology,Lake Placid, NY, USA). Immunocomplexes, aggregated with50 ml of g-bind Protein-A Sepharose (Pharmacia, Uppsala,Sweden) were washed four times with 0.6ml of buffer. Pelletswere boiled in Laemly buffer for 5min and the proteinsresolved under reducing conditions by 7–10% SDS–PAGEand transferred to reinforced nitrocellulose (PROTRAN,Schleider & Schuell, Keene, NH, USA). The membranes wereblocked with 3% BSA in PBS (0.05% Tween 20) overnight at41C and probed for 1 h at 251C with anti-PLCg mAb, diluted1 : 500 (Upstate Biotechnology, Lake Placid, NY, USA) oranti-FRS2 rabbit polyclonal antibodies, diluted 1 : 500 (H-91Santa Cruz Biotecnology Inc, CA, USA), followed by ECL. Insome experiments, in order to estimate the protein equalloading, the membranes were rehydrated by washing in PBSTween 20, stripped with 100mm mercaptoethanol and 2%SDS for 30min at 551C, and reprobed again with anti-Bekpolyclonal antibodies or antiphosphotyrosine mAb andprocessed for immunoblotting as above.

Immunofluorescence microscopy

Cells, grown on coverslips and treated with KGF or exposedto UVB as above, were fixed in methanol at �201C for 4min.For double staining experiments, cells were incubated withanti-Bek polyclonal antibodies (1 : 10 in PBS) (Santa CruzBiotechnology, Inc., CA, USA) and antiphosphotyrosine mAb(1 : 100 in PBS) (PY69 Immunobiological Costa Mesa, CA,USA) for 1 h at 251C. The primary antibodies were visualizedwith goat anti-mouse IgG-FITC (1 : 10 in PBS) (CappelResearch Products Durham, NC, USA) and goat anti-rabbitIgG-Texas Red (1 : 50 in PBS) (Jackson Immuno ResearchLaboratories Inc., PA, USA) after appropriate washing withPBS. Colocalization of the two fluorescence signals wasanalysed by recording and merging single stained images usinga cooled CCD color digital camera SPOT-2 (DiagnosticInstruments Incorporated, MI, USA) and FISH 2000/H1software (Delta Sistemi, Roma, Italy).

BrdU incorporation

NIH3T3 KGFR and NIH3T3 cells were serum-starved for12 h, exposed to UVB (30mJ/cm2) and replaced to 371C for anadditional 12 h in medium alone. BrdU (100mm) (Sigma) wasadded to the medium and kept for 1 h at 371C to allow BrdUincorporation. Cells were then washed extensively and fixed in4% formaldehyde in PBS for 30min at 251C, followed by

treatment with 0.1m glycin for 20min at 251C and with 3NHCl, 0.1% Triton X-100 for an additional 1 h at 251C to allowpermeabilization. After extensive washing in PBS, cells werebuffered with 0.1m Na2B4O7 and incubated with anti-BrdUmAb (1 : 50 in PBS) (Sigma) for 1 h at 251C in a humiditychamber, followed by goat anti-mouse IgG-FITC (1 : 10 inPBS) (Cappel Research Products Durham, NC, USA).

Fatty acid analysis

Cell pellets (5� 106 cells) were extracted twice in chloroformmethanol 1 : 1 in the presence of butylated hydroxytoluene(BHT, 50 mg) as antioxidant and 25mg of tricosanoic acid ethylester as internal standard. Fatty acids were trans-methylatedwith sodium metoxide (30% (w/v) in methanol and analysedby gas chromatography–mass spectrometry (GC–MS) oncapillary column (FFA-P, 60m� 0.32mm� 0.25mm, HewlettPackard). The results were obtained after time integration ofthe chromatogram and final processing of areas. The identityof each fatty acid has been determined by comparing the massspectrum of the peaks with those obtained using referencestandards (Picardo et al., 1996). Results are the mean7s.d. ofthree different experiments in duplicate.

Antioxidant enzyme assays

Cells (5� 106) were sonicated in phosphate buffer saline of50mm pH 7.4, centrifuged at 10 000 g for 10min at 41C andsupernatants were collected. Protein concentration was deter-mined by Bradford test. Enzymatic activities were evaluatedwith a Beckman DU 70 spectrophotometer. Cat activity wasmeasured by the disappearance of hydrogen peroxide (Clai-borne, 1985) and SOD activity by the inhibition of superoxideproduction by xantine–xantine oxidase system (O’Neil et al.,1988). Standard curves have been obtained by using purifiedhuman SOD and bovine Cat at different concentrations. Oneunit of Cat has been defined as the amount that degrades 1 mmH2O2 and one unit of SOD as the amount of enzyme thatinduces 50% inhibition of nitro blue of tetrazolium reduction.Results are reported as mean7s.d. of three different experi-ments in duplicate and expressed as units per milligram ofproteins.

Vitamin E analysis

Cells (5� 106) were extracted three times in hexane : ethanol 3 : 1in the presence of 125ng of g and d tocopherol as internalstandards. The solvent was evaporated to dryness under anitrogen stream. Tocopherols, derivatized with N,O-bis-(tri-methylsilyl)-trifluoroacetamide (BSTFA) with 1% trimethylchlorosilane as catalyst, were analysed by GC–MS on SPB1column (30m� 0.2mm ID� 0.25mm, Supelchem) by a selectedion (s) monitoring (SIM) technique (Picardo et al., 1996). Resultsare reported as mean7s.d. of three different experiments induplicate and expressed as nanogram per milligram of proteins.

Glutathione (GSH) analysis

The analysis of GSH was performed by means of capillaryelectrophoresis according to a modified method (Piccoli et al.,1994). Cells were lysed in 500ml of hyposmotic buffer and at100 ml of lysate, 100 ml of N-ethylmaleimide (NEM 10mm)were added in order to prevent the oxidation of GSH. 5-sulfosalicylic acid (50 ml at 10%) was added 30min later toprecipitate the proteins and the volume made up to 500ml with50mm Na2B4O7. The samples were centrifuged and thesupernatant was injected under pressure into the capillaryelectrophoretic apparatus. The electrophoretic separation of


2430

Oncogene

reduced and oxidized GSH was performed using an automatedP/ACE 5500 system (Beckman Instruments, Fullerton, CA,USA) equipped with a diode array detector and fitted at 15 kVhigh-voltage power supply. The detection wavelength was210 nm. The capillary was of untreated fused silica (57 cm,75mm ID, 375 mm OD) with the detector cell at 50 cm from theinlet end. The run buffer was KH2PO4 20mm and Na2B4O7

50mm, at pH 7.8. The amount of the GSH and its oxidizedform, GSSG, in the sample was calculated according to acalibration curve obtained from a standard solution of GSHand GSSG. Results are reported as mean7s.d. of threedifferent experiments in duplicate and expressed as nanomolesof GSH per milligram of proteins.

Abbreviations

KGF: keratinocyte growth factor; CUH: cumene hydroper-oxide; NAC: N-acetylcysteine; PDTC: pyrrolidine dithiocar-bamate; ROS: reactive oxygen species; SOD: superoxidedismutase; Cat: catalase.

Acknowledgements

This work was partially supported by grants from MURST,from Associazione Italiana per la Ricerca sul Cancro (AIRC),from CNR (Target Project on ‘Biotechnology’) and Ministerodella Sanita Italy.

References

Bae YS, Kang SW, Seo MS, Baines IC, Tekle E, Chock PBand Rhee SG. (1997). J. Biol. Chem., 272, 217–221.

Capone A, Visco V, Belleudi F, Marchese C, Cardinali G,Bellocci M, Picardo M, Frati L and Torrisi MR (2000). CellGrowth Differ., 11, 607–614.

Claiborne A. (1985). Handbook of Methods for Oxygen RadicalResearch. Greenwald RA (ed). CRC: Boca Raton, FL, pp.283–284.

Coffer PJ, Burgering BMT, Peppelenbosch MP, Bos JL andKruijer W. (1995). Oncogene, 11, 561–569.

de Gruijl FR. (2000). Methods Enzymol., 319, 359–366.Dimon-Gadal S. (1998). J. Biol. Chem., 273, 22833–22284.Finch PW, Murphy F, Cardinale I and Krueger JG. (1997)

Am. J. Pathol., 151, 1619–1628.Finch PW, Rubin JS, Miki T, Ron D and Aaronson SA.(1989) Science, 245, 752–755.

Fisher GJ, Wang ZQ, Datta S, Varani J, Kang S and VoorheesJJ. (1997). New Engl. J. Med., 337, 1419–1428.

Fuchs J and Packer L. (1990). Photodermatol. Photoimmunol.Photomed., 7, 920–925.

Glenn McGregor W. (1999). J. Invest. Dermatol., 4, 1–5.Hadari YR, Gotoh N, Kouhara H, Lax I and Schlessinger J.(2001). Proc. Natl. Acad. Sci. USA, 98, 8578–8583.

Hines MD and Allen-Hoffman BL. (1996). J. Biol. Chem., 271,6245–6251.

Huang RP and Adamson ED. (1995). Oncogene, 10, 467–475.Huang RP, Wu JX, Fan Y and Adamson ED. (1996). J. Cell.

Biol., 133, 211–220.Kayanoki Y, Fujii J, Islam KN, Suzuki K, Kawata S,Matsuzawa Y and Taniguchi N. (1996) J. Biochem.(Tokyo), 119, 817–822.

LaRochelle WJ, Dirsch OR, Finch PW, Cheon HG, May M,Marchese C, Pierce JH and Aaronson SA. (1995). J. CellBiol., 129, 357–366.

Marchese C, Chedid M, Dirsch OR, Csaky KG, Santanelli F,Latini C, LaRochelle WJ, Torrisi MR and Aaronson SA.(1995). J. Exp. Med., 182, 1369–1376.

Marchese C, Mancini P, Belleudi F, Felici A, Gradini R,Sansolini T, Frati L and Torrisi MR. (1998). J. Cell Sci.,111, 3517–3527.

Marchese C, Rubin JS, Ron D, Faggioni A, Torrisi MR,Messina A, Frati L and Aaronson SA. (1990). J. CellPhysiol., 144, 326–333.

Marchese C, Sorice M, De Stefano C, Frati L and Torrisi MR.(1997). Cell Growth Differ., 8, 989–997.

Miki T, Bottaro DP, Fleming TP, Smith CL, Burgess WH,Chan AML and Aaronson SA. (1992). Proc. Natl. Acad. Sci.USA, 89, 246–250.

Miki T, Fleming TP, Bottaro DP, Rubin JS, Ron D andAaronson SA. (1991). Science (Washington, DC), 251,72–75.

Ohba M, ShibanumaM, Kuroky T and Nose K. (1994). J. CellBiol., 126, 1079–1088.

O’Neil P, Davies S and Fielden EM. (1998). J. Biochem., 24,41–46.

Peus D, Meves A, Vasa RA, Beyerle A, O’Brien T andPittelkow MR. (1999) Free Rad. Biol. Med., 27, 1197–1202.

Peus D, Vasa RA, Meves A, Pott M, Beyerle A, Squillace Kand Pittelkow M. (1998). J. Invest. Dermatol., 110, 966–971.

Picardo M, Grammatico P, Roccella F, Roccella M, Grand-inetti M, Del Porto G and Passi S. (1996). J. Invest.Dermatol., 107, 322–326.

Piccoli G, Fiorani M, Biagiarelli L, Palma F, Potenza L,Amicucci A and Stocchi V. (1994). J. Chromatogr., 676, 239–246.

Rosette C and Karin M. (1996). Science (Washington, DC),274, 1194–1197.

Rubin JS, Osada H, Finch PW, Taylor W, Rudikoff S andAaronson SA. (1989). Proc. Natl. Acad. Sci. USA, 86, 802–806.

Sachsenmaier C, Radler-Pohl A, ZincK R, Nordheim A,Herrlich P and Rahmsdorf HJ. (1994). Cell, 78, 963–972.

Schwarz A, Bhardway R, Aragane Y, Mahnke K, Riemann H,Metze D, Luger TA and Schwarz T. (1995). J. Invest.Dermatol., 104, 922–927.

Shaoul E, Reich-Slotky R, Berman B and Ron D. (1995).Oncogene, 10, 357–366.

Sheikh MS, Antinore MJ, Huang Y and Fornace AJ. (1998).Oncogene, 17, 2555–2563.

Staiano-Coico L, Krueger GJ, Rubin JS, D’limi S, Vallat VP,Valentino L, Fahey T, Hawes A, Kingston G, Madden MR,Mathwich M, Gottlieb AB and Aaronson SA. (1993). J.Exp. Med., 178, 865–878.

Vessey DA, Lee KH and Boyer TD. (1995). J. Invest.Dermatol., 104, 355–358.

Werner S, Peters KG, Longaker MT, Fuller-Pace F, BandaMJ and Williams LT. (1992). Proc. Natl. Acad. Sci. USA, 89,6896–6900.

Werner S, Smola HM, Xiang L, Longaker MT, Krieg T,Hofschneider PH and Williams LT. (1994). Science, 266,819–822.

Wikonkal NM and Brash DE. (1999). J. Invest. Dermatol., 4,6–10.

Zhou Y, Lee HST, Kooshesh F, Fujisawa H, Sauder DN andKondo S. (1996). Exp. Dermatol., 5, 138–144.


2431

Oncogene

Documents

UVB-induced activation and internalization of keratinocyte growth factor receptor