Fas Ligand Interaction Contributes to UV-Induced Apoptosis in Human Keratinocytes

EXPERIMENTAL CELL RESEARCH ARTICLE NO.

232, 255–262 (1997)

EX973514

Fas/Fas Ligand Interaction Contributes to UV-Induced Apoptosis in Human Keratinocytes MARTIN LEVERKUS, MINA YAAR,

AND

BARBARA A. GILCHREST1

Department of Dermatology, Boston University School of Medicine, Boston, Massachusetts 02118-2394

Keratinocytes in human skin undergo apoptosis during various inflammatory processes and after ultraviolet (UV) irradiation. To determine if keratinocyte apoptosis may be mediated by the Fas/APO-1 receptor (CD95), a signal transduction pathway known to initiate programmed cell death of lymphocytes, we investigated Fas expression, modulation, and function in keratinocytes. Keratinocytes constitutively expressed the 2.5- and 1.9-kb Fas transcripts, as well as the 43-kDa Fas protein. Treatment of interferon-g-stimulated keratinocytes with Fas agonistic antibody significantly promoted their cell death, indicating that Fas in keratinocytes is functional. UV irradiation induced Fas mRNA expression within 16 to 24 h and Fas protein within 24 h and through 48 h after irradiation. Furthermore, keratinocytes constitutively expressed Fas ligand (FasL) mRNA and protein. UV irradiation induced FasL mRNA as early as 4 h after irradiation and elevated FasL mRNA levels were maintained for at least 24 h postirradiation. Moreover, a FasL neutralizing antibody significantly reduced UV-induced apoptosis of IFN-g-treated keratinocytes. Our data strongly suggest that the Fas system contributes to keratinocyte apoptosis in UV-irradiated human skin. q 1997 Academic Press

INTRODUCTION

Apoptosis is a distinct form of cell death, in which damaged cells activate a genetic program that leads to the destruction of their DNA [1]. Apoptosis plays a major role during embryogenesis [2] and in many diseases including neurodegenerative disorders [3], AIDS [4], and cancer [5]. In the skin it has been speculated that terminal differentiation of keratinocytes is apoptotic in nature [6, 7], and apoptotic keratinocytes have been recognized morphologically in many skin diseases such as erythema multiforme, graft-versus-host disease, and 1 To whom correspondence and reprint requests should be addressed at Boston University School of Medicine, Department of Dermatology, J-Bldg., 80 E. Concord Street, Boston, MA 02118-2394. Fax: 617-638-5550.

lichenoid drug eruption [8, 9]. As well, ultraviolet (UV) irradiation produces apoptotic-like keratinocytes, socalled ‘‘sunburn cells,’’ in the epidermis [10], and recent studies in vitro have proven that UV-irradiated keratinocytes display DNA fragmentation characteristic of apoptosis [11, 12]. Although some conditions that provoke keratinocyte apoptosis have been identified, little is known regarding the mechanisms that regulate this process. In the immune system, apoptosis is mediated by signaling molecules such as tumor necrosis factor (TNF) [13] and the recently described Fas ligand (FasL) [14]. When expressed by lymphoid cells, these signaling molecules can induce apoptosis in either an autocrine or a paracrine manner by activating their specific transmembrane receptors [14–18]. The TNF receptors 1 and 2 and the Fas/APO-1 receptor are members of a superfamily of receptors characterized by cysteine-rich motifs in the extracellular domain of the molecule [17]. The Fas receptor is a transmembrane protein of 43 kDa molecular weight [19]. Binding of FasL leads to Fas activation and the induction of intracellular signals that eventually result in apoptotic cell death [17, 19, 20]. FasL has been shown to be expressed in the immune system [17], and recent work demonstrated FasL expression also in macrophages [21], the anterior segment of the eye [22], and the Sertoli cells of the testis [23]. In human skin, Fas receptors have been detected in the epidermis in various inflammatory, infectious, and autoimmune diseases, including lichenoid drug eruption, erythema multiforme, bullous pemphigoid, and herpes zoster [24]. Pretreatment of cultured keratinocytes with interferon-g (IFN-g) was reported to induce Fas receptors on the cell surface [24–26]. It was not known whether FasL is expressed in the skin. We now report that keratinocytes express Fas mRNA and protein, that Fas is functional in these cells, that FasL is constitutively expressed in keratinocytes, and that both Fas and FasL mRNA and protein are induced in keratinocytes after UV irradiation. Furthermore, we show that addition of neutralizing FasL antibodies [27] inhibits UV-induced apoptosis of IFN-g-treated keratinocytes. Taken together our data support a role for

255

AID

ECR 3514

/

6i1f$$$$21

04-02-97 10:42:00

0014-4827/97 $25.00 Copyright q 1997 by Academic Press All rights of reproduction in any form reserved.

eca

256

LEVERKUS, YAAR, AND GILCHREST

Fas/FasL in UV-induced keratinocyte apoptosis in human skin. MATERIAL AND METHODS Materials. Serum-free culture medium (K-Stim) was obtained from Becton–Dickinson (Two Oak Park, Bedford, MA). The monoclonal antibody mouse anti-human Fas IgM (CH11, purified by IgM purification column, Pierce, Rockford, IL) and anti-human Fas IgG (ZB4) were obtained from MBL International Corp. (Watertown, MA). Monoclonal anti-FasL antibody (mouse IgG) was obtained from Transduction Laboratories (Lexington, KY), polyclonal anti FasL antibody from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA), and neutralizing anti FasL antibody (NOK-1) was obtained from Pharmingen (San Diego, CA). Horseradish peroxidase-tagged rat antimouse IgM and an FITC-tagged F(ab)2 fragment of rabbit anti-mouse IgG were obtained from Zymed Laboratories (San Francisco, CA). An FITC-tagged F(ab)2 fragment of goat anti-rabbit IgG was obtained from Jackson ImmunoResearch Laboratories, Inc. (West Grove, PA). Horseradish peroxidase-tagged sheep anti-mouse IgG was obtained from Amersham Corp. (Arlington Heights, IL). Recombinant human IFN-g was obtained from Genzyme (Cambridge, MA). Ionomycin, phorbol 12-myristate 13-acetate (PMA), and adenine were obtained from Sigma Chemical Co. (St. Louis, MO). F12 nutrient mixture was obtained from Gibco (Gaithersburg, MD). Tissue culture. Primary keratinocyte cultures were prepared from newborn foreskins as described [28]. Briefly, dermis and epidermis were separated by trypsin treatment and the epidermal cell suspension was seeded onto a feeder layer of lethally irradiated 3T3 fibroblasts in a 75:25 mixture of Dulbecco’s modified Eagle’s medium and F-12 nutrient mixture containing 10% FBS, 1.8 mM adenine, 10 ng/ml EGF, and 1.4 mM hydrocortisone. At near confluence cultures were passed at a density of 1–2 1 106 per 60-mm dish into serumfree growth factor-supplemented medium at a calcium concentration of 0.03 mM, conditions favoring proliferation rather than differentiation. All cells were used for experiments at first and second passage. Jurkat cells (American Type Culture Collection (ATCC), clone E6-1) were maintained in RPMI 1640 medium containing 10% FBS. All cultures were maintained at 377C in 7% CO2 and supplied with fresh medium twice weekly. UV irradiation. UV irradiation was performed using a 1000-W solar simulator (Spectral Energy Corp., Westwood, NJ) metered at 285 { 5 nm, as described [28]. Cells were irradiated with physiologic UV doses (20–80 mJ/cm2) in PBS through the plastic petri dish cover. After irradiation, cells were provided fresh medium. Sham-irradiated cultures were handled identically, except that they were shielded with aluminum foil during the irradiation. PCR generation of cDNA probes. Human Fas, FasL, and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) cDNAs were generated from newborn keratinocyte RNA. Briefly, 5 mg of total cellular RNA was reverse transcribed according to the manufacturer’s recommendations (Pharmacia Fine Chemicals). Five microliters of the resulting cDNA was amplified with 15 pmol of each of the following primer pairs under otherwise identical conditions: ACAGCAGAACAGAAAGTTCA (forward primer) and GACCAAGCTTTGGATTTCAT (reverse primer), amplifying the COOH-terminal end of human Fas cDNA (bp 1003–1200), [19]. Primers for FasL were CTGGAATGGGAAGACACCTA and GTAGCTCATCATCTTCCCC, amplifying the COOH-terminal of human FasL cDNA (bp 480–700), [14]. Primers for human GAPDH were GTCATCATCTCTGCCCCCTC and AGCCCCGCGGCCATCACGCC, amplifying a fragment of the human GAPDH cDNA (bp 1050–1300) [29]. Amplification was performed for 40 cycles of denaturation at 947C for 2 min, primer annealing at 507C for 4 min, and DNA polymerization at 727C for 6 min in a microcycler (Eppendorf North America, Inc., Madison, WI). Sequence identity of the resulting products with the published se-

AID

ECR 3514

/

6i1f$$$$22

04-02-97 10:42:00

quences of the human Fas, FasL, and GAPDH cDNAs was confirmed by automated sequencing (Applied Biosystems 373A, fluorescencebased sequencer). Northern blot analysis. Before and at different intervals after stimulation, total cellular RNA was collected with Tri-Reagent (Molecular Research Center, Cincinnati, OH) and RNA isolation was carried out according to the manufacturer’s instructions. Twenty micrograms of RNA was electrophoresed and transferred to a nylon membrane (Hybond N, Amersham Corp.) and hybridization with 32Plabeled Fas or GAPDH cDNA was performed as described before [28]. Blots were autoradiographed with intensifying screens at 0707C with Kodak X-Omatic AR films. RT–PCR and Southern blot analysis. Before and at different intervals after stimulation, total cellular RNA was collected, purified, and reverse transcribed as described above. Five microliters of the resulting cDNA was amplified with either FasL (20 cycles) or GAPDH (10 cycles) primers as an internal control. Preliminary experiments confirmed that amplification for this number of cycles was exponential under these selected amplification conditions for both FasL and GAPDH (data not shown). As a negative control, 1 mg of RNA was amplified identically. The resulting PCR products were electrophoresed on a 1% agarose gel, transferred, and hybridized as described above with radiolabeled cDNAs for FasL or GAPDH, respectively. Blots were autoradiographed with intensifying screens at 0707C for 1–6 h. Densitometric analysis. Autoradiograms were scanned in a computerized densitometer (Molecular Dynamics). Bands were manually defined, and band intensity was determined using the MD Image Quant 3.2 program. Western blot analysis. Total cellular proteins from keratinocyte cultures were collected as described before [28]. Protein (20–50 mg) was electrophoresed (10 or 12.5% SDS–PAGE) and transferred to nitrocellulose membranes (Schleicher & Schuell, Keene, NH). Membranes were reacted with anti-Fas antibody (CH11) or anti-FasL antibody (Transduction Lab.). After incubation with horseradish peroxidase-tagged secondary antibody, bands were visualized using the ECL detection kit (Amersham Corp.). FACScan analysis. Subconfluent keratinocytes were detached from the plates by EDTA treatment followed by brief treatment with 0.25% trypsin. Cells were reacted with 20 mg/ml monoclonal antiFas antibodies (ZB4, MBL) or isotyped control IgG (Sigma Chemical Co.) followed by an FITC-conjugated rabbit anti-mouse F(ab*)2 fragment (Zymed Laboratories Inc.). Cells (105) were analyzed by FACScan (Becton–Dickinson & Co., San Jose, CA) and data were processed with the LYSYS software package (Becton–Dickinson & Co., San Jose, CA). Staining of keratinocytes with 5 mg/ml polyclonal rabbit anti-FasL antibody (Santa Cruz) or rabbit IgG was performed identically. Induction of cell death by anti-Fas agonistic antibodies. Preconfluent keratinocytes were supplemented with 1000 U/ml recombinant IFN-g or diluent alone for 24 h. After the cells were washed in PBS, 1 mg/ml anti-Fas (CH 11) was added and the cultures were maintained at 377C for an additional 24 h. As controls for anti-Fas-treated cells, cultures were supplemented with the same concentrations of mouse IgM (Sigma). After 24 h, the plates were processed for terminal deoxynucleotidyl transferase-mediated nick end labeling (TUNEL) staining (see below). Neutralization of FasL-mediated apoptosis. Preconfluent keratinocytes were supplemented with 1000 U/ml recombinant IFN-g or diluent alone for 24 h. After the cells were washed twice with PBS, they were UV or sham irradiated as described above. Immediately after UV irradiation, cells were provided fresh medium containing either 1 mg/ml FasL neutralizing antibody (NOK-1, ref. 27) or identical concentrations of mouse IgG. After 24 h, the plates were processed for TUNEL staining (see below). After 48 h, cultures were washed twice with PBS, trypsinized, and counted with a particle counter

eca

FAS/FAS LIGAND AND UV-INDUCED APOPTOSIS IN KERATINOCYTES

257

(Coulter Electronics Inc., Miami, FL). Mean cell number of duplicate dishes for each condition was determined. DNA extraction. Total cellular DNA of UV- or sham-irradiated keratinocytes was extracted as described [30], except that attached and detached cells in the medium were collected for extraction. DNA was separated on an 1% agarose gel and stained with ethidium bromide. Assessment of DNA fragmentation by nick end labeling of apoptotic DNA. TUNEL assay of apoptotic DNA in cells was performed with the In Situ Cell Death Detection kit (Boehringer Mannheim, Germany) according to the manufacturer’s recommendations. Briefly, cells were fixed in 4% paraformaldehyde and permeabilized in 0.1% sodium citrate, 0.1% Triton X-100. After washing, plates were incubated for 1 h at 377C in the TUNEL reaction mixture containing terminal deoxynucleotidyl transferase and FITC-labeled dUTP. Plates were analyzed by immunofluorescent microscopy using a inverted Nikon Diaphot microscope (Nikon, Garden City, NY). Negative control cells were handled identically except that the enzyme was omitted. For analysis, representative pictures of dark-field and phase-contrast field were taken. The percentage of TUNEL-positive cells was determined in four representative fields for each condition.

RESULTS

UV irradiation induces keratinocyte apoptosis. To determine if keratinocytes undergo apoptosis after UV irradiation, we examined DNA fragmentation in keratinocytes 24 h after exposure to physiologic doses of solar-simulated light. In UV-irradiated keratinocytes, the characteristic DNA ladder was detected, whereas sham-irradiated cells showed intact DNA (Fig. 1A). Moreover, UV-treated cultures showed 23.6 { 6.1% TUNEL positivity, whereas sham-treated control cultures had only 8.95 { 3.46% TUNEL-positive cells (Figs. 1B and 1C), therefore confirming previous reports that keratinocytes undergo apoptosis after UV irradiation [11, 12]. Keratinocytes express Fas mRNA and protein. To investigate Fas mRNA expression in keratinocytes, total keratinocyte RNA was isolated and examined by Northern blotting using a 0.2-kb fragment of human Fas cDNA as a probe. The 2.5- and 1.9-kb Fas transcripts [19] were constitutively expressed in keratinocytes (Fig. 2A). To determine if keratinocytes express Fas protein, total cellular proteins were extracted and analyzed by Western blotting using an anti-Fas monoclonal antibody. The 43-kDa Fas protein was detected by anti-Fas antibody (Fig. 2B). FACScan analysis of keratinocytes showed surface expression of Fas receptor protein (Fig. 2C). Keratinocytes express FasL mRNA and protein. To investigate FasL mRNA expression in human keratinocytes, semiquantitative RT–PCR of keratinocyte cDNA was performed. The expected 220-bp band was present in keratinocytes after 20 cycles of amplification with FasL-specific primers, followed by Southern blot analysis with 32P-labeled FasL cDNA (Fig. 3A). To rule out the possibility that the source of FasL is contaminating cells occasionally present in keratinocyte cultures, the

AID

ECR 3514

/

6i1f$$$$22

04-02-97 10:42:00

FIG. 1. Induction of apoptosis after UV irradiation. (A) Total cellular DNA was collected 24 h after 20 mJ/cm2 UV or sham irradiation. The characteristic DNA ladder was seen only in the UV-irradiated sample (U), whereas the sham-irradiated sample (S) shows mainly intact DNA. (B) Keratinocytes were UV irradiated with 20 mJ/cm2 (U) or sham irradiated (S) and processed for TUNEL staining after 24 h. Shown are identical representative fields of phase-contrast and immunofluorescent fields of sham- or UV (20 mJ/cm2)irradiated keratinocytes (U) 24 h after irradiation. (C) Graphic representation of quantitative result of (B). Four representative fields for each condition were examined in duplicate plates. UV-treated plates contained 23.6 { 6.1% TUNEL-positive cells, whereas sham-irradiated control cultures contained only 9.0 { 3.5% positive cells.

highly differentiated squamous carcinoma cell line SCC12F was processed identically and showed the same result (data not shown). Western blots of total keratinocyte proteins reacted with anti-FasL antibody (Transduction Laboratories) recognized a single band of 38–40 kDa. The Jurkat lymphoma cell line (ATTC), stimulated with 10 ng/ml PMA and 500 ng/ml ionomycin for 6 h to induce FasL expression [31], was used

eca

258

LEVERKUS, YAAR, AND GILCHREST

FIG. 2. Constitutive Fas expression in keratinocytes. (A) Twenty micrograms of total cellular RNA was harvested from keratinocytes cultured under basal conditions. Fas transcripts of 1.9 and 2.5 kb are constitutively expressed (arrows). (B) Fifty micrograms of total keratinocyte proteins was processed for Western blotting and reacted with 10 mg/ml of anti-Fas antibody (lane 1) or with an isotypematched control antibody (filtered and centrifuged mouse ascites containing the indicated concentration of monoclonal IgM-k) (lane 2). The expected band of 43 kDa is seen only in the lane hybridized with monoclonal anti-Fas antibody (CH11). A band of approximately 68 kDa is seen in the lane hybridized with matched IgM monoclonal control antibody, probably representing nonspecific binding of the nonpurified control antibody. (C) Keratinocytes were reacted with 20 mg/ml anti-Fas antibody (ZB4) or identical concentrations of isotypematched IgG and were analyzed by FACScan. Keratinocytes constitutively express Fas receptor.

as a positive control and showed a band of the same molecular weight (Fig. 3B). Fas mRNA and protein are upregulated after UV irradiation. To determine Fas modulation after UV irradiation, total cellular RNA from UV- or sham-irradiated keratinocytes was isolated up to 72 h after irradiation and analyzed by Northern blotting. Within 24 h UV-irradiated cells showed a two- to fivefold induction of Fas mRNA compared to sham-irradiated cells at the same time (Figs. 4A and 4B). This upregulation was observed through 48 h and returned to baseline at 72 h. FACScan analysis of UV-irradiated keratinocytes showed consistent induction of Fas protein 24 through 48 h after irradiation (Fig. 4C). FasL mRNA and protein are upregulated after UV irradiation. To investigate if UV irradiation modifies FasL mRNA levels, keratinocyte RNA was processed for semiquantitative RT–PCR. Within 4 h after UV irradiation, UV-treated samples showed a three- to fivefold increase in FasL mRNA (Figs. 5A and 5B). Increased FasL mRNA persisted at least 24 h after irradiation. Consistent with the Western blotting data (Fig. 3B), FasL protein as determined by FACScan analysis

AID

ECR 3514

/

6i1f$$$$23

04-02-97 10:42:00

was constitutively expressed, with approximately 30% of cells positive under all control conditions (sham-irradiated cultures). UV irradiation moderately induced expression of FasL to approximately 40% positivity between 16 and 48 h after irradiation. Maximum induction was seen 24 h postirradiation. Fas mediates cell death in keratinocytes. To determine if Fas is functional in epidermal cells, we examined the induction of apoptosis in keratinocytes after treatment of the cells with the agonistic anti-Fas monoclonal antibody CH 11 (MBL International), known to act as a Fas pseudo-ligand and to activate the receptor [32]. The agonistic anti-Fas antibody or control IgM antibody was added to keratinocytes pretreated for 24 h with IFN-g (1000 U/ml) to increase their sensitivity to Fas-mediated apoptosis [24]. Cultures were maintained for an additional 24 h at 377C and then in situ cell death was determined by the TUNEL assay. Fas agonistic antibody-treated plates showed 20.9 { 5.5% TUNEL-positive cells, whereas IgM-treated control cultures had only 8.79 { 2.2% positive cells (Figs. 6A and 6B). Neutralization of FasL reduces UV-induced keratinocyte apoptosis. To determine whether the Fas/FasL system participates in UV-induced apoptosis, cells were cultured in the presence of IFN-g 24 h prior to UV irradiation in order to increase sensitivity to Fasmediated apoptosis, as described [24]. Cells were then irradiated with 20–80 mJ/cm2 UV. Immediately after UV irradiation, one mg/ml of either FasL neutralizing antibody (NOK-1) or an irrelevant mouse IgG antibody

FIG. 3. Constitutive FasL expression in keratinocytes. (A) RT– PCR and Southern blot analysis of keratinocyte RNA. Five micrograms of RNA was reverse transcribed and amplified with specific primers for a 220-bp fragment of the extracellular domain of human FasL. To exclude contamination with genomic DNA, identical concentrations of non-reverse-transcribed RNA were used as a negative control. After 20 cycles of amplification, PCR products were processed for Southern blot analysis and hybridized with FasL cDNA. The expected band of 220 bp was detected only in cDNA, but not in RNA of human keratinocytes. Blots were exposed for 6 h at room temperature. (B) Twenty micrograms of total keratinocyte proteins was processed for Western blot analysis (lane 1); 4 mg of Jurkat cell protein, stimulated with 10 ng/ml PMA and 500 ng/ml ionomycin for 6 h, was used as a positive control (lane 2). FasL protein of 38–40 kDa was recognized by anti-FasL antibody in both samples.

eca

FAS/FAS LIGAND AND UV-INDUCED APOPTOSIS IN KERATINOCYTES

259

FIG. 4. Fas modulation after UV irradiation. (A) Twenty micrograms of total cellular RNA was harvested from keratinocytes at 24, 48, and 72 h after sham (S) or 25 mJ/cm2 UV irradiation (U). UV-irradiated cells showed a 2.3-fold induction of Fas mRNA compared to shamirradiated cells at 24 and 48 h after irradiation. Within 72 h, in UV-irradiated cells, Fas mRNA levels returned to baseline and were comparable to those of sham-irradiated cells and untreated control cells (C). Comparable induction of Fas mRNA after UV irradiation (2to 5-fold) was observed in four independent experiments. (B) Graphic representation of densitometric analysis shown in (A). (B) Keratinocytes were UV (U) (20 mJ/cm2) or sham (S) irradiated and analyzed by FACScan at 24 and 48 h postirradiation after incubation with 20 mg/ml anti-Fas antibody. UV-irradiated samples showed induction of Fas immunoreactivity when compared to sham-irradiated samples after 24 h, persisting through 48 h postirradiation. Shown is one representative experiment of a total of three experiments with similar results.

was added to the culture medium and cells were incubated for additional 24–48 h. Plates treated with FasL neutralizing antibody showed a significant decrease in the number of TUNEL-positive cells (Fig. 7A). Furthermore, cell yields determined 48 h after irradiation showed increased numbers of cells in cultures treated with FasL neutralizing antibody compared to control antibody-treated cells, indicating increased survival after UV irradiation (Fig. 7B, P õ 0.005). DISCUSSION

FIG. 5. Fas L modulation after UV irradiation. (A) Total cellular RNA was harvested from keratinocytes 4 and 24 h after UV (25 mJ/ cm2) or sham irradiation. Five micrograms of the resulting RNA was reverse transcribed and amplified for 20 cycles with FasL-specific primers or for 10 cycles with GAPDH-specific primers. PCR products were processed for Southern blotting and hybridized with FasL or GAPDH cDNA. After exposure of the blots for 1 h at room temperature, constitutive expression of FasL mRNA was detectable in the control lane (C). When compared with sham-irradiated samples (S), a threefold induction of FasL mRNA is detected as early as 4 h postirradiation and persists through 24 h postirradiation (U). Amplification of the constitutively expressed GAPDH gene confirms comparable amplification of all samples. (B) Graphic representation of densitometric analysis of the experiment shown in (A).

AID

ECR 3514

/

6i1f$$$$23

04-02-97 10:42:00

Our experiments confirm the recent report [11] that UV irradiation induces apoptosis in keratinocytes. We documented apoptosis by two techniques, with TUNEL staining of keratinocyte cultures showing a less pronounced increase in apoptotic cells than anticipated from examining nucleosomal size DNA fragmentation (DNA ladder). This apparent discrepancy may be explained, however, by differences between the techniques, in that total DNA in the culture dishes was harvested for gel electrophoresis, while only attached cells were processed for the TUNEL assay. Because apoptotic cells eventually detach and are released into the culture medium [33], our TUNEL data probably

eca

260

LEVERKUS, YAAR, AND GILCHREST

induced apoptosis in human keratinocytes sensitized by pretreatment with IFN-g. Our data therefore confirm previous reports [24–26] that IFN-g-treated cultured keratinocytes are sensitive to Fas-mediated apoptosis. The reason for the increased sensitivity of IFN-g-treated keratinocytes to Fas-mediated apoptosis is not clear. However, some studies showed induction of Fas protein [24] or Fas mRNA [26] after treatment of keratinocytes with IFN-g, suggesting increased sensitivity through increased number of receptors. Our demonstration that normal human keratinocytes express FasL mRNA and protein adds to the rapidly growing literature suggesting a general role for Fas/FasL in mediating apoptosis in many organ systems, as has been demonstrated in lymphoid cells and more recently in macrophages [21], the anterior segment of the eye [22], and the Sertoli cells of the testis [23]. Interestingly, in vitro keratinocytes constitutively express both Fas and FasL. This constitutive expression appears not to be associated with constitutive apoptosis, at least under the conditions of these experiments, in which keratinocytes are maintained as relatively undifferentiated monolayer cultures, since FasL neutralizing antibodies did not significantly affect cell number or the number of TUNEL-positive cells in control cultures. Recently, a similar constitutive coexpression of Fas and FasL on normal human corneal epithelium without apparent apoptosis [36] was reported. These findings are consistent with previously published reports concerning another member of the Fas

FIG. 6. Induction of apoptosis after treatment with agonistic anti-Fas antibody. (A) The agonistic Fas monoclonal antibody CH11 or control IgM antibody was added to cells pretreated for 24 h with IFN-g (1000 U/ml). Cultures were maintained for an additional 24 h at 377C and then plates were processed for in situ cell death detection by the TUNEL assay. Anti-Fas antibody-treated plates showed 20.9 { 5.5% TUNEL positivity, whereas IgM-treated control cultures were 8.79 { 2.2% positive. (B) The same experiment as (A), showing representative fields of phase-contrast and immunofluorescent field pictures of IgM-treated control cultures (IgM) or agonistic anti-Fas antibody-treated keratinocytes (anti-Fas) 24 h after antibody provision.

underestimate the percentage of apoptotic keratinocytes in our experiments. In order to evaluate the role of the Fas/FasL system in UV-induced keratinocyte apoptosis, we first demonstrated that cultured human keratinocytes constitutively express Fas mRNA and protein. The 2.5- and 1.9-kb transcripts and the 43-kDa protein we observed are in agreement with the previously reported molecular sizes of Fas mRNA and protein in other cell types [19, 34, 35]. The detected Fas receptor is functional, since treatment with an agonistic anti-Fas antibody

AID

ECR 3514

/

6i1f$$$$23

04-02-97 10:42:00

FIG. 7. Neutralization of FasL after UV irradiation. Keratinocytes were cultured in the presence of IFN-g 24 h prior to UV irradiation. Cells were then irradiated with 20–80 mJ/cm2 UV. Immediately after UV irradiation, 1 mg/ml of either FasL neutralizing antibody (N) or a control mouse IgG antibody (C) was added to the culture medium and cells were incubated for an additional 24–48 h. (A) After 24 h, plates were stained for TUNEL. Plates treated with FasL neutralizing antibody showed a significant decrease in the number of TUNEL-positive cells in all examined UV doses. Decrease of apoptosis attributable to FasL neutralization was statistically significant (P õ 0.003, nonpaired t test). (B) After 48 h, plates were washed twice with PBS and cells were then counted with a particle counter. Increased cell number in cultures treated with FasL neutralization was statistically significant (P õ 0.005, nonpaired t test).

eca

FAS/FAS LIGAND AND UV-INDUCED APOPTOSIS IN KERATINOCYTES

receptor family. Keratinocytes constitutively express low levels of both the p55-TNF receptor [37] and TNF [38, 39]. Nevertheless these cells do not undergo extensive apoptosis unless stimulated by UV irradiation, known to induce the p55-TNF receptor and TNF [37, 39]. We examined the potential role of the Fas/FasL system specifically in UV-mediated keratinocyte apoptosis. UV irradiation upregulated Fas mRNA, suggesting that UV irradiation enhances Fas transcription and/or mRNA stability. Induction of Fas protein was also observed within 24 h postirradiation and persisted through at least 48 h postirradiation. FasL was induced as early as 4 h after UV irradiation and through at least 24 h postirradiation. Consistent with the mRNA data, FasL protein was mildly induced on the cell surface by 16 h postirradiation, with maximum induction 24 h postirradiation. The lesser induction of FasL protein on the cell surface than at the mRNA level suggests the possibility of increased cleavage and secretion of FasL from the membrane, as has been demonstrated in lymphoid cells stimulated by metalloproteinases [27, 31]. Because metalloproteinases are known to be activated in skin after UV irradiation [40], this mechanism may be germane to keratinocytes as well. If only secreted FasL can crosslink Fas receptors on the membrane, this would explain the failure of cells expressing both Fas and FasL to undergo apoptosis in spite of constitutive expression on the cell surface. Direct involvement of the Fas/FasL system in UVinduced keratinocyte apoptosis is established by the fact that neutralizing anti-FasL antibodies [27], when provided immediately after UV irradiation, decrease UV-induced apoptosis of keratinocytes. Neutralizing FasL did not completely block UV-induced apoptosis; however, it is unlikely that this is due to insufficient amounts of the neutralizing antibodies, since we used a concentration of FasL neutralizing antibodies 10-fold greater than that necessary for complete protection against FasL-mediated apoptosis in Fas-transfected cells [27]. Rather, we believe that other molecules such as TNF and its receptor, known to be involved in UVinduced apoptosis [41], also contribute to the apoptotic process. Such redundancy is consistent with the critical homeostatic role of apoptosis in removing possibly mutated cells from severely damaged tissues [42]. The authors thank M. Eller, S. Tyagi, T. Maeda, and A. Schwaaf for helpful and critical discussions and Professor E. B. Broecker for continuing support. M. Leverkus is a postdoctoral fellow of the German Research Foundation.

REFERENCES 1. Steller, H. (1995) Science 267, 1445–1449. 2. Koseki, C., Herzlinger, D., and Al-Awqati, Q. (1992) J. Cell Biol. 119, 1327–1333. 3. Isacson, O. (1993) Trends Neurosci. 16, 306–308.

AID

ECR 3514

/

6i1f$$$$23

04-02-97 10:42:00

261

4. Meyaard, L., Otto, S. A., Jonker, R. R., Mijnster, M. J., Keet, R. P., and Miedema, F. (1992) Science 257, 217–219. 5. Hoffmann, B., and Liebermann, D. A. (1994) Oncogene 9, 1807. 6. Polakowska, R. R., and Haake, A. R. (1994) Cell. Death Differ. 1, 19–31. 7. Song, Y., Yaar, M., and Gilchrest, B. A. (1996) J. Invest. Dermatol. 106, 875. 8. Kanerva, L. (1990) J. Cutaneous Pathol. 17, 37–44. 9. Weedon, D. (1990) Apoptosis. Adv. Dermatol. 5, 243–256. 10. Young, A. R. (1987) Photodermatology 4, 127–134. 11. Casciola-Rosen, L. A., Anhalt, G., and Rosen, A. (1994) J. Exp. Med. 179, 1317–1330. 12. Zhai, S., Pincelli, C., Yaar, M., Gonsalves, J., and Gilchrest, B. A. (1995) J. Invest. Dermatol. 104, 573. 13. Greenblatt, M. S., and Elias, L. (1992) Blood 80, 1339–1346. 14. Takahashi, T., Tanaka, M., Inazawa, J., Abe, T., Suda, T., and Nagata, S. (1994) Int. Immunol. 6, 1567–1574. 15. Cohen, J. J., Duke, R. C., Fadok, V. A., and Sellins, K. S. (1992) Annu. Rev. Immunol. 10, 267–293. 16. Wright, S. C., Kumar, P., Tam, A. W., Shen, N., Varma, M., and Larrick, J. W. (1992) J. Cell Biochem. 48, 344–355. 17. Nagata, S., and Golstein, P. (1995) Science 267, 1449–1456. 18. Dhein, J., Walczak, H., Baeumler, C., Debatin, K. M., and Krammer, P. H. (1995) Nature 373, 438–441. 19. Itoh, N., Yonehara, S., Ishii, A., Yonehara, M., Mizushima, S., Sameshima, M., Hase, A., Seto, Y., and Nagata, S. (1991) Cell 66, 233–243. 20. Trauth, B. C., Klas, C., Peters, A. M. J., Matzku, S., Moeller, P., Falk, W., Debatin, K. M., and Krammer, P. H. (1989) Science 245, 301–305. 21. Arase, H., Arase, N., and Saito, T. (1995) J. Exp. Med. 181, 1235–1238. 22. Griffith, T. S., Brunner, T., Fletcher, S. M., Green, D. R., and Ferguson, T. A. (1995) Science 270, 1189–1192. 23. Bellgrau, D., Gold, D., Selawry, H., Moore, J., Franzusoff, A., and Duke, R. C. (1995) Nature 377, 630–632. 24. Sayama, K., Yonehara, S., Watanabe, Y., and Miki, Y. (1994) J. Invest. Dermatol. 103, 330–334. 25. Matsue, H., Kobayashi, H., Hosokawa, T., Akitaya, T., and Ohkawara, A. (1995) Arch. Derm. Res. 287, 315–320. 26. Takahashi, H., Kobayashi, H., Hashimoto, Y., Matsuo, S., and Iizuka, H. (1995) J. Invest. Dermatol. 105, 810–815. 27. Kayagaki, N., Kawasaki, A., Ebata, T., Ohmoto, H., Ikeda, S., Inoue, S., Yoshino, K., Okumura, K., and Yagita, H. (1996) J. Exp. Med. 183, 1777–1783. 28. Wintzen, M., Yaar, M., Burbach, J. P. H., Gilchrest, B. A. (1996) J. Invest. Dermatol. 106, 673–678. 29. Tso, J. Y., Sun, X. H., Kao, T., Reece, K. S., and Wu, R. (1985) Nucleic Acids Res. 13, 3181–3187. 30. Zhai, S., Yaar, M., Doyle, S., and Gilchrest, B. A. (1996) Exp. Cell Res. 224, 335–343. 31. Mariani, S. M., Matiba, B., Baeumler, C., and Krammer, P. H. (1995) Eur. J. Immunol. 25, 2303–2307. 32. Yonehara, S., Ishii, A., and Yonehara, M. (1989) J. Exp. Med. 169, 1747–1756. 33. Weller, M., Frei, K., Groscurth, P., Krammer, P. H., Yonekawa, Y., and Fontana, A. (1994) J. Clin. Invest. 94, 954–964. 34. Oehm, A., Behrmann, I., Falk, W., Pawlita, M., Maier, G., Klas, C., Li-Weber, M., Richards, S., Dhein, J., Trauth, B. C., Ponstingl, H., and Krammer, P. H. (1992) J. Biol. Chem. 267, 10709–10715.

eca

262

LEVERKUS, YAAR, AND GILCHREST

35. Enari, M., Hug, H., and Nagata, S. (1995) Nature 375, 78–81. 36. Wilson, S. E., Li, Q., Weng, J., Barry-Lane, P. A., Jester, J. V., Liang, Q., Wordinger, R. J. (1996) Invest. Ophthalmol. Vis. Sci. 37, 1582–1592. 37. Trefzer, U., Brockhaus, M., Loetscher, H., Parlow, F., Budnik, A., Grewe, M., Christoph, H., Kapp, A., Schoepf, E., Luger, T. A., and Krutmann, J. (1993) J. Clin. Invest. 92, 462–470. 38. Itoh, N., and Nagata, S. (1993) J. Biol. Chem. 268, 10932– 10943. 39. Koeck, A., Schwarz, T., Kirnbauer, R., Urbanski, A., Perry, P.,

Ansel, J. C., and Luger, T. A. (1990) J. Exp. Med. 172, 1609– 1614. 40. Fisher, G. J., Datta, S. C., Talwar, H. S., Wang, Q., Varani, J., Kang, S., and Voorhees, J. J. (1996) Nature 379, 335–339. 41. Schwarz, A., Bhardwaj, R., Aragane, Y., Mahnke, K., Riemann, H., Metze, D., Luger, T. A., and Schwarz, T. (1995) J. Invest. Dermatol. 104, 922–927. 42. Ziegler, A., Jonason, A. S., Leffell, D. J., Simon, J. A., Sharma, H. W., Kimmelmann, J., Remington, L., Jacks, T., and Brash, D. E. (1994) Nature 372, 773–776.

Received June 21, 1996 Revised version received January 27, 1997

AID

ECR 3514

/

6i1f$$$$24

04-02-97 10:42:00

eca