View metadata, citation and similar papers at core.ac.uk brought to you by CORE providedORIGINAL by Elsevier - ARTICLE Publisher Connector

Histamine Enhances the Production of in Human Keratinocytes

Naoko Kanda and Shinichi Watanabe Department of Dermatology,Teikyo University, School of Medicine, 11-1,Tokyo, Japan

Nerve growth factor induces innervation and epidermal hancement of nerve growth factor secretion, promoter hyperplasia in in£ammatory skin diseases like psoriasis. activity, activator 1 transcriptional activity, and Nerve growth factor production by keratinocytes is in- c-Fos expression was suppressed by H1 antagonist pyri- creased in the in£ammatory lesions. Nerve growth lamine, protein kinase C inhibitor calphostin C, and factor induces histamine release from mast cells. We PD98059, an inhibitor of mitogen-activated protein ki- examined the in vitro e¡ects of histamine on nerve nase kinase 1. Histamine induced the translocation of growth factor production in human keratinocytes. His- protein kinase C activity from cytosol to membrane, tamine enhanced nerve growth factor secretion, which was suppressed by phospholipase C inhibitor mRNA expression, and promoter activity in keratino- U73122. It stimulated the phosphorylation of extra- cytes. Two TPA-response elements on the nerve growth cellular signal-regulated kinase, which was blocked by factor promoter were responsible for the activation by pyrilamine, calphostin C, and PD98059. These results histamine. Histamine enhanced transcriptional activity suggest that histamine may enhance nerve growth and DNA binding of activator protein 1 at the two TPA- factor production by inducing c-Fos expression in response elements. It shifted the TPA-response-element- keratinocytes. These e¡ects may be mediated by the binding activator protein 1 composition from c-Jun H1--induced signaling cascade of phospholi- homodimers to c-Fos/c-Jun heterodimers. Histamine pase Cprotein kinase Cmitogen-activated protein transiently induced c-Fos mRNA expression, which kinase kinase 1extracellular signal-regulated kinase. was not detectable in unstimulated keratinocytes, Key words: activator protein 1/c-Fos/extracellular signal-regu- whereas c-Jun mRNA expression was constitutive and lated kinase/H1 receptor/protein kinase C. J Invest Dermatol was not altered by histamine. Histamine-induced en- 121:570 ^577, 2003

erve growth factor (NGF) plays an important role NGF also promotes the histamine release of mast cells (Bruni in the development of in£ammatory skin diseases et al, 1982; Pearce and Thompson, 1986; Matsuda et al,1991; like psoriasis, atopic dermatitis, or allergic contact Horigome et al, 1993; 1994). In the early phases of psoriasis or ato- dermatitis (Raychaudhuri et al,1998;Kinkelin pic dermatitis, mast cells are increased in the dermis (Jiang et al, et al, 2000; Toyoda et al, 2002). In these skin dis- 2001; Alenius et al, 2002). Mast cell-derived histamine induces Neases, NGF production in keratinocytes is enhanced (Raychaud- keratinocytes to produce various in£ammatory cytokines or huri et al,1998;Kinkelinet al, 2000; Toyoda et al, 2002). NGF chemokines such as interleukin-6 or interleukin-8 (Kohda et al, potentiates the proliferation of neurons, which may result in hy- 2002). It is thus plausible that histamine may also induce NGF perinnervation and itch sensation in in£ammatory skin diseases production in keratinocytes. A previous study reported that his- (Kinkelin et al, 2000). NGF also enhances the growth of keratino- tamine in vitro enhanced NGF production by murine transformed cytes having the high a⁄nity NGF receptor trkE (Di Marco et al, keratinocytes PAM212 and 3T3 ¢broblasts (Matsuda et al,1998).It 1993; Pincelli et al, 1994), which may lead to epidermal hyperplasia is unknown, however, what level of NGF production, i.e. trans- associated with psoriasis or atopic dermatitis (Toyoda et al,2002). lational or transcriptional, histamine may potentiate. In addition, previous studies did not show intracellular signaling events responsible for the NGF potentiation by histamine. Histamine binds to cell surface receptors coupling to guanine Manuscript received February 19, 2003; revised April 1, 2003; accepted nucleotide-binding protein (G-protein) and induces various in- for publication April 3, 2003 tracellular signaling pathways. Four isotypes of histamine recep- Reprint requests to: Naoko Kanda, Department of Dermatology,Teikyo tors have been identi¢ed: H1, H2, H3, and H4 receptors (Oda University, School of Medicine, 11-1, Kaga-2, Itabashi-Ku,Tokyo 173-8605, et al, 2000). H1 receptor is linked to the activation of phospholi- Japan; Email: [email protected] pase C (PLC), which generates diacylglycerol activating protein Abbreviations: AG1478, 4-(3-chloroanilino)-6,7-dimethoxyquinazoline; kinase C (PKC) (Megson et al, 2001). Stimulation of H2 receptor AP-1, activator protein 1; EMSA, electrophoretic mobility shift assay; ERK, extracellular signal-regulated kinase; G-protein, guanine nucleo- mostly activates adenylate cyclase, which generates a cAMP signal tide-binding protein; H-89, N-[2-((p-bromocinnamyl)aminoethyl)-5-iso- that activates protein kinase A; however, H2 receptor is also quinolinesulfonamide; MAPK, mitogen-activated protein kinase; MEK, linked to PLC in certain cell types (Del Valle and Ganz, 1997). MAPK kinase; TRE, TPA response element; U73122, 1-[6-((17b-3-meth- H3 and H4 receptors are coupled to the inhibition of adenylate oxyestra-1,3,5(10)-trien-17-yl)amino)hexyl]-1H-pyrrole-2,5-dione. cyclase (Coge et al, 2001). These histamine-receptor-mediated

0022-202X/03/$15.00 . Copyright r 2003 by The Society for Investigative Dermatology, Inc.

570 VOL. 121, NO. 3 SEPTEMBER 2003 HISTAMINE ENHANCES NGF PRODUCTION IN KERATINOCYTES 571 signals regulate cytokine or chemokine in target polybrene-dimethylsulfoxide (8.5%70.9%), and thus was adopted in this cells (Kanda and Watanabe, 2002). study. Keratinocytes were plated in 10 cm dishes and grown to about 60% In this study, we examined the e¡ects of histamine on NGF con£uence. Twenty-four hours before the transfection, the medium was production in cultured human keratinocytes. We found that his- changed to KBM. Keratinocytes were incubated for 6 h with 5 mgof pNGF luc or p4 TRE1-TATA-luc or p4 TRE2-TATA-luc and 1 mgof tamine potentiates NGF production in keratinocytes via H1 re- Rous sarcoma virus b-galactosidase vector (pRSV-b-gal), premixed with ceptors. We further examined the precise mechanism for this enhancer, transfection bu¡er, and E¡ectene. The transfected cells were e¡ect focusing on the histamine-induced intracellular signaling washed and incubated in KBM for 18 h and then treated with histamine. pathways. After 6 h, cell extracts were prepared and luciferase activities were quanti- ¢ed using the luciferase assay system (Promega). The same cell extracts were assayed for b-galactosidase activity using the chemiluminescent MATERIALS AND METHODS Galacto-Light kit (Tropix, Bedford, MA). All readings were taken using a Lumat 9501 luminometer (Berthold, Wildbach, Germany). The results Reagents Histamine dihydrochloride, cimetidine, pyrilamine, and obtained in each transfection were normalized for b-galactosidase activity calphostin C were obtained from Sigma (St Louis, MO). Thioperamide and expressed as relative luciferase activity. maleate was from Biomol Research Laboratories (Plymouth Meeting, PA ). 1 - [6 - ( (17 b-3-methoxyestra-1,3,5(10)-trien-17-yl)amino)hexyl]-1H-pyrrole-2, EMSA EMSA was performed as described previously (Cartwright et al, 32 5-dione (U73122), N-[[2-(p-bromocinnamylamino)ethyl]-5-isoquinol- 1992). The probes used were P-labeled annealed double-stranded DNA inesulfonamide (H-89), and 4-(3-chloroanilino)-6, 7-dimethoxy- containing TRE1 or TRE2 from human NGF promoter. The sequence of quinazoline (AG1478) were obtained from Calbiochem (La Jolla, CA). the TRE1probe is 50 -ATTTGGAGCGTGTGACTCAGGAGTACGGGAG-30 Recombinant human tumor necrosis factor a was from R&D Systems (consensus sequence underlined) and that of the TRE2 probe is 50 - (Minneapolis, MN). Antibodies used in the electrophoretic mobility shift GGAGCGCAGCGGTGAGTCAGGCTGCCCCGAG-30.Forgelshift assay (EMSA) were from Santa Cruz Biotechnology (Santa Cruz, CA). assays, 2^5 mg of nuclear protein extracts were incubated at room temperature for 5 min with a mixture containing 6 mM HEPES (pH 7.9), Culture of keratinocytes Human neonatal foreskin keratinocytes were 0.4 mM ethylenediamine tetraacetic acid (EDTA), 125 mM KCl, 10% cultured in serum-free KGM medium (Clonetics, Walkersville, MD) glycerol, 0.05 mg per mL poly dI-dC, 1 mM dithiothreitol, 2.5 mM sodium consisting of basal medium MCDB153 supplemented with 0.5 mg per ml pyrophosphate, 1 mM b-glycerophosphate, 1 mM Na3VO4, 10 mM NaF, hydrocortisone, 5 ng per ml epidermal growth factor, 5 mg per ml insulin, 50 mg per ml aprotinin, 50 mg per ml leupeptin. One nanogram of and 0.5% bovine pituitary extract. The cells in third passage were used. labeled probe was added and the reactions were incubated for another 20 min. In antibody supershift experiments, the nuclear extracts were preincubated with various antibodies for 30 min before the addition of NGF secretion Keratinocytes (2 10 4 per well) were seeded in triplicate probe. Reactions were then fractionated on a nondenaturing 5% into 24-well plates in 1 ml KGM and adhered overnight; the medium was polyacrylamide gel, and visualized with phosphorimager (Molecular then changed to basal KBM depleted of hormones, growth factors, and Dynamics, Sunnyvale, CA). bovine pituitary extract, and the cells were incubated for 24 h. The medium was removed and the cells were incubated with histamine in Western blot analysis The dual threonine/tyrosine phosphorylation of KBM for 24 h. The supernatants were assayed for NGF by ELISA extracellular signal-regulated kinase 1/2 (ERK1/2), c-Jun N-terminal (Promega, Madison,WI). kinase (JNK), and p38 mitogen-activated protein kinase (MAPK) was assessed by western blot as described previously (Gee et al,2002). RT-PCR Keratinocytes were incubated as above for indicated periods, Keratinocytes were incubated with 10^6 M histamine or 10 ng per mL and cellular mRNA was extracted using an mRNA puri¢cation kit tumor necrosis factor a for 10 min. The cells were lyzed with lysis bu¡er (Pharmacia, Uppsala, Sweden) according to the manufacturer’s (50 mM HEPES (pH 7.5), 150 mM NaCl, 10% glycerol, 1% Triton X-100, instruction. cDNA was made from mRNA samples as described 1.5 mM MgCl2, 100 mM NaF,100 mM sodium orthovanadate, and 1 mM previously (Tjandrawinata et al, 1997). Primers for ampli¢cation and the b 0 0 0 ethyleneglycol-bis( -aminoethyl ether)-N,N,N ,N -tetraacetic acid (EGTA, sizes of respective PCR products were as follows: NGF, 5 -CACACTGAG- pH 7.7)) and centrifuged for 20 min at 14,000 g at 41C. The supernatant GTGCATAGCGT-30 and 50 -TGATGACCGCTTGCTCCTGT-30 for 0 0 0 were subjected to sodium dodecyl sulfate polyacrylamide gel 352 bp; c-Fos, 5 -GGCTTCAACGCAGACTACGAGG-3 and 5 -CTCCT- electrophoresis (SDS-PAGE) and transferred to a nitrocellulose GTCATGGTCTTCACAACG-30 for 340 bp; c-Jun, 50 -TTCACCTTCTC- 0 0 0 membrane. The membrane was blocked and incubated with antiphospho- TCTAACTGC-3 and 5 -TCACTCACTGAGCGCTCTTC-3 for 580 bp; ERK1/2 or antiphospho-JNK antibodies (Santa Cruz Biotechnology) or glyceraldehyde-3-phosphate dehydrogenase (GAPDH), 50 -GCAGGGGG- 0 0 0 antiphospho-p38 antibody (New England Biolabs, Mississauga, Ontario, GAGCCAAAAGGG-3 and 5 -TGCCAGCCCCAGCGTCAAAG-3 for Canada), followed by peroxidase-conjugated secondary antibodies (Bio- 566 bp (Lane et al, 1998; Risse-Hackl et al, 1998; Iannone et al, 2002; Kanda Rad, Hercules, CA). The blots were developed with an enhanced and Watanabe, 2002). PCR involved one denaturing cycle of 951Cfor chemiluminescence kit (Amersham, Arlington Heights, IL). The blots 3 min, 30 cycles of denaturation at 951C for 30 s, annealing at 581Cfor were stripped and reprobed with rabbit polyclonal antibodies speci¢c for 30 s, and extension at 721C for 30 s, and a ¢nal extension at 721C for 3 each of unphosphorylated ERK1/2, JNK, and p38 MAPK (Santa Cruz min. The PCR products were analyzed by electrophoresis on 2.5% Biotechnology), as described above. agarose gels and stained with ethidium bromide, viewed by ultraviolet light. Densitometric analysis was performed by scanning the bands into Assays of PKC PKC activity was analyzed as described previously Photoshop and performing densitometry with NIH Image Software. (Migliaccio et al, 1998). Keratinocytes were incubated with 10^6 M histamine for 10 min. The cells were harvested and homogenized in a Plasmids and transfections The ¢re£y luciferase reporter plasmids bu¡er containing 20 mM TrisHCl (pH 7.4), 2 mM EDTA, 10 mM EGTA, driven by human NGF promoter (^600/ þ 250 bp relative to the 250 mM sucrose, 5 mM dithiothreitol, 1 mM phenylmethyl- transcriptional start site) were constructed by PCR and insertion into sulfonyl £uoride, and 0.24 mM leupeptin. Nuclei and debris were removed pGL3 basic vector (Promega) as described previously (Cartwright et al, from the samples by centrifugation (500 g, 5 min) and this post 1992), and are denoted pNGF luc. Site-speci¢c mutation of the promoter nuclear fraction was then centrifuged (105,000 g, 90 min). The super- was created by multiple rounds of PCR using primers with altered bases as natant was saved as the cytosolic fraction. The pellet was homogenized in described previously (Li and Kolattukudy, 1994). p4 TPA response the same bu¡er, except for containing 0.1% Triton X-100. The samples element 1 (TRE1)-TATA-luc and p4 TRE2-TATA-luc were constructed were mixed continuously for 1 h at 41C and then centrifuged as described by inserting four copies of TRE1 (50 -GCGTGTGACTCAGGAGT-30, above. This supernatant was saved as the membrane fraction. The cytosolic consensus sequence underlined) or TRE2 (50 -AGCGGTGAGTCA- or membrane fractions were loaded onto columns packed with GGCTG-30) from human NGF promoter in front of the TATA box diethylaminoethyl-sephacel resin and were washed extensively with a upstream of ¢re£y luciferase reporter as described previously (Kanda and solution of 20 mM TrisHCl (pH 7.4), 20 mM NaCl, 0.5 mM EDTA, Watanabe, in press). Transient transfections were performed with E¡ectene 0.5 mM EGTA, and 10 mM 2-mercaptoethanol. PKC was eluted with the (Qiagen, Tokyo, Japan) as described previously (Zellmer et al, 2001). The same bu¡er except that the NaCl concentration was increased to 100 mM. e⁄ciency of transfection into keratinocytes by this method was Aliquots of puri¢ed PKC from cytosolic or membrane fractions were 28.5%73.1% (mean7SEM, n ¼ 9) as determined by £ow cytometry assayed for protein contents using the Bio-Rad protein determination using b-galactosidase vector. E¡ectene produced a higher transfection assay. PKC activity was assessed using a reaction mixture containing 32 e⁄ciency than Lipofectin (8.1%71.2%), Lipofectamine (9.5%71.1%), or 20 mM TrisHCl (pH 7.4), 8 mM MgCl2,16mMATP,[g- P]ATP, and 572 KANDA AND WATANABE THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

200 mg per ml III-S at 301C for 3 min. Activity was determined in the presence or absence of 100 mMCaCl2,8mg per ml phosphatidyl serine, and 2 mg per ml diolein. Samples were ¢ltered withWhatman GF/C ¢lters, and 32P incorporation into lysine-rich histone was quanti¢ed by liquid scintillation.

Measurement of histamine release Keratinocytes (2 10 4 per well) were seeded into 24-well plates in KGM, adhered overnight, and incubated with KBM for 24 h. The medium was replaced with fresh KBM, and the cells were incubated for 24 h. The supernatant histamine amount was measured by ELISA (Cosmo Bio,Tokyo, Japan).

RESULTS Histamine enhanced NGF secretion in keratinocytes Keratinocytes constitutively secreted a low amount of NGF (21.572.8 pg per ml, mean7SEM, n ¼ 5), and the secretion was concentration-dependently increased by histamine; the Figure1. Dose dependence for the e¡ects of histamine on NGF stimulatory e¡ect of histamine was manifested at 10^7 Mand ^6 secretion and the inhibition by H1 or H2 receptor antagonists or was maximized at 10 M, which increased the secretion 4.5-fold pertussis toxin on the e¡ects of histamine. Keratinocytes were preincu- of controls (Fig 1). The stimulatory e¡ect of histamine was not bated with pertussis toxin (PTX; 16 h, 100 ng per ml), pyrilamine (PR), or suppressed by pertussis toxin, indicating that the e¡ect may be cimetidine (CM) (each 10^5 M, 30 min), and then incubated with indicated mediated via pertussis-toxin-resistant G-proteins. H1 receptor doses of histamine for 24 h in the presence of the inhibitors. The culture antagonist pyrilamine blocked the stimulatory e¡ect of supernatants were assayed for NGF.Values are mean7SD of triplicate cul- Ã histamine (Fig 1) whereas H2 antagonist cimetidine (Fig 1)or tures. po0.05 versus control cultures, by one-wayANOVAwith Dunnett’s H3 and H4 antagonist thioperamide (data not shown) did not. multiple comparison test. The data shown in the ¢gure are representative of These results suggest that H1 receptor but not H2, H3, or H4 ¢ve separate experiments. receptors may mediate histamine-induced enhancement of NGF secretion in keratinocytes. As 10^6 M of histamine was optimal for NGF induction, this concentration was used in the following experiments. As examined by ELISA, constitutive histamine secretion by keratinocytes was 9.7570.83 nM (mean7SEM, n ¼ 5), which was less than the threshold (100 nM) for NGF induction (Fig 1). Pyrilamine, cimetidine, and thioperamide did not reduce the constitutive NGF secretion in the absence of exogenous histamine (data not shown). This suggests that endogenous histamine secreted by keratinocytes may not contribute to NGF secretion. We next examined if histamine may enhance NGF mRNA expression in keratinocytes.

Histamine increased NGF mRNA levels At 3 h of incubation, histamine increased the NGF mRNA level 6.1-fold of controls, which was reversed by pyrilamine but not by cimetidine or pertussis toxin (Fig 2). Thus histamine increased NGF production via H1 receptors at pretranslational level. We then examined if histamine may enhance NGF promoter activity in keratinocytes.

Histamine enhanced NGF promoter activity It is reported that human NGF gene contains two TREs, TRE1 (^62/^56 bp) and TRE2 ( þ 35/ þ41 bp), and these may act as enhancer elements for NGF transcription (Cartwright et al, 1992) (Fig 3a). We transiently transfected NGF promoter (^600/ þ 250 bp) linked to luciferase reporter into human keratinocytes. Histamine treatment increased NGF promoter activity 5.1-fold Figure 2. The e¡ects of histamine on NGF mRNA level in kerati- of controls, and the e¡ect was blocked by pyrilamine (Fig 3b). nocytes. Keratinocytes were preincubated with pertussis toxin (PTX; The mutation of TRE1 or TRE2 each reduced the basal 16 h, 100 ng per ml), pyrilamine (PR), or cimetidine (CM) (each 10^5 M, promoter activity and histamine-induced enhancement of the 30 min), and then incubated with 10^6 M histamine (HA). After 3 h, activity. The mutation of both TRE completely abrogated the mRNA was isolated and RT-PCR was performed. The intensity of the basal and histamine-induced promoter activities. These results band for NGF was corrected to that for GAPDH. The lower graph shows suggest that both TRE1 and TRE2 may be required for basal the corrected intensities relative to that in control cells cultured with med- and histamine-induced NGF promoter activities. We then ium alone (set as 1.0). The results shown in the ¢gure are representative of analyzed if histamine may enhance transcriptional activities ¢ve separate experiments. dependent on TRE1 or TRE2.

Histamine enhanced transcription from TRE1 or front of the TATA box. Histamine increased the TRE1- or TRE2 Keratinocytes were transiently transfected with TRE2-dependent transcriptional activities, and the e¡ects were luciferase reporter linked to four repeats of TRE1 or TRE2 in reversed by pyrilamine (Fig 4). VOL. 121, NO. 3 SEPTEMBER 2003 HISTAMINE ENHANCES NGF PRODUCTION IN KERATINOCYTES 573

Figure 3. The e¡ects of histamine on the activities of wild-type or Figure 5. The e¡ect of histamine on binding to TRE. Keratinocytes were preincubated with 10^5 M pyrilamine (PR) for mutated NGF promoters. (a) Schematic representation of human NGF ^6 promoter. The locations of cis-elements are shown with their sequences, 30 min, and then incubated with 10 M histamine (HA) for 60 min. Nu- and substituted bases for mutation are indicated in italics. The nucleotide clear extracts were prepared and incubated with TRE1-containing probe. positions are relative to the transcriptional start site. (b) Keratinocytes were In supershift assays, nuclear extracts were incubated with antibodies against transiently transfected with wild-type (WT) or mutated pNGF luc to- c-Fos or c-Jun for 30 min before the addition of the probe. The asterisks gether with pRSV-b-gal. The cells were preincubated with or without 10^ indicate the supershifted complexes. The results shown in the ¢gure are re- 5 M pyrilamine (PR) for 30 min, and then treated with 10^6 M histamine presentative of four separate experiments. (HA). After 6 h, relative luciferase activities were analyzed. The data are mean7SEM of four separate experiments. Values at the right indicate the à fold induction versus basal promoter activity. po0.05 versus control values, suggests that histamine may increase transcription factor binding wpo0.05 versus values with histamine alone, by one-way ANOVA with to TRE in keratinocytes via H1 receptors. In unstimulated Sche¡e’s multiple comparison test. keratinocytes, anti-c-Jun did but anti-c-Fos antibody did not supershift the complex (Fig 5, lanes 5, 6), whereas in histamine- stimulated keratinocytes both antibodies supershifted the complex (Fig 5, lanes 7, 8). Antibodies against the other Fos family (FosB, Fra-1, Fra-2) or Jun family (JunB, JunD) proteins did not supershift the complexes by nuclear extracts from unstimulated or histamine-stimulated keratinocytes (data not shown). These results suggest that histamine may shift the AP-1 composition from c-Jun homodimers to c-Fos/c-Jun hetero- dimers at TRE1. Similar results were obtained by using the TRE2 probe (data not shown).

Histamine induced c-Fos expression in keratinocytes We then examined if histamine may enhance c-Fos or c-Jun expression in keratinocytes. Histamine rapidly and transiently induced c-Fos mRNA. In unstimulated cells, c-Fos mRNA was undetectable; however, in histamine-stimulated cells, the Figure 4. The e¡ects of histamine on TRE-dependent transcription. expression was induced at 15 min, maximized at 30 min, Keratinocytes were transiently transfected with p4 TRE1-TATA-luc or reduced at 1 h, and disappeared at 3 h (Fig 6). c-Jun mRNA p4 TRE2-TATA-luc or enhancerless pTATA-luc, together with pRSV- expression was constitutively detectable and was not enhanced b-gal. The cells were preincubated with 10^5 M pyrilamine (PR) for 30 ^6 by histamine. The results indicate that histamine-induced NGF min, and then incubated with 10 M histamine (HA). After 6 h, relative production may be mainly attributable to the induction of c-Fos luciferase activities were analyzed. The results represent mean7SEM of expression. four separate experiments. Values at the right indicate the fold induction à versus basal activity. po0.05 versus control values, wpo0.05 versus values PLC^PKC^MAPK kinase 1 (MEK1)^ERK pathway is with histamine alone, by one-way ANOVA with Sche¡e’s multiple com- involved in histamine-induced NGF production and c-Fos parison test. expression Histamine is known to induce various intracellular signaling pathways: the PLCPKC pathway leading to the activation of ERK (Megson et al, 2001) or the adenylate Histamine induced c-Fos/c-Jun binding to TRE It is known cyclaseprotein kinase A pathway (Shayo et al, 1997). We thus that TRE is mostly bound by transcription factor activator tried to identify the signaling pathway involved in histamine- protein 1 (AP-1) (Allegretto et al, 1990). We then examined if induced enhancement of NGF production. PKC inhibitor histamine may enhance DNA binding of AP-1 at TRE on the calphostin C completely suppressed histamine-induced enhance- NGF promoter. At 1 h of incubation, histamine increased ment of NGF promoter activity (Fig 7a), NGF secretion the amount of DNAprotein complex with TRE1 probe, and (Fig 7b), TRE1-dependent transcriptional activity (Fig 7c), the e¡ect was reversed by pyrilamine (Fig 5, lane 4). This and c-Fos expression (Fig 7d). Histamine-induced c-Fos 574 KANDA AND WATANABE THE JOURNAL OF INVESTIGATIVE DERMATOLOGY expression was also completely inhibited by PD98059, which membrane whereas cimetidine and pertussis toxin did not selectively blocks ERK activity by inhibiting MEK1 to phos- (Fig 8b). These results suggest that histamine-induced phorylate ERK1/2 (Fig 7d). PD98059 also suppressed histamine- membrane translocation of PKC may be mediated via H1 induced enhancement of NGF promoter activity (Fig 7a), NGF receptor and PLC but may be independent of H2 receptor and secretion (Fig 7b), and TRE1-dependent transcriptio- insensitive to pertussis toxin. ERK is known to be activated by nal activity (Fig 7c) by 60%^65%. These results suggest MEK1 via dual phosphorylation on threonine and tyrosine the requirement of PKC and ERK for histamine-induced residues (Widmann et al, 1999). We thus examined the enhancement of NGF production. In contrast, histamine- phosphorylation status of ERK after histamine treatment. At 10 induced NGF promoter activation, NGF secretion, TRE1- min, histamine induced dual phosphorylation of ERK1 (44 dependent transcriptional activity, and c-Fos expression were kDa) and ERK2 (42 kDa) (Fig 9a), and the phosphorylation not suppressed by protein kinase A inhibitor H-89, indicating was suppressed by calphostin C, PD98059, and pyrilamine, that protein kinase A may be dispensable for the histamine- indicating that histamine-induced ERK activation may require induced e¡ects. We then examined if histamine may activate PKC, MEK1 activity and may be mediated via H1 receptor. PKC or ERK in keratinocytes. In unstimulated keratinocytes, Recent studies have demonstrated that certain G-protein- most PKC activity resided in the cytosolic fraction (Fig 8a). At coupled receptors activate ERK via transactivation of epidermal 10 min after histamine treatment, PKC activity in the membrane growth factor receptor (Zwick et al, 1999). fraction was enhanced, indicating the membrane translocation This possibility is rather unlikely in histamine-stimulated of PKC. Pyrilamine and PLC inhibitor U73122 suppressed keratinocytes, however, as AG1478, an inhibitor of epidermal the histamine-induced enhancement of PKC activity in the growth factor receptor tyrosine kinase, did not block histamine- induced ERK phosphorylation (Fig 9a). In contrast to ERK, histamine did not stimulate the dual phosphorylation of other MAPKs, JNK, and p38 MAPK in keratinocytes (Fig 9b, c).

DISCUSSION In this study, histamine enhanced NGF production in human keratinocytes via H1 receptors, which is consistent with previous studies on rat astrocytes (Lipnik-Strangelj et al, 1998; Lipnik- Strangelj and Carman-Krzan, 2000; 2001). Histamine stimulated NGF transcription in keratinocytes by enhancing transcriptional activity of AP-1. Histamine induced rapid and transient c-Fos mRNA expression. The induced c-Fos may heterodimerize with c-Jun and the heterodimers may bind to two TREs on NGF pro- moter and drive NGF transcription. On the other hand, c-Jun homodimers appeared to bind to TRE in unstimulated keratino- Figure 6. The e¡ects of histamine on c-Fos or c-Jun mRNA expres- cytes. It is reported that c-Fos/c-Jun heterodimers much more sion in keratinocytes. Keratinocytes were incubated with 10^6 M hista- avidly bind to DNA and have much higher transcriptional acti- mine. At indicated time-points, mRNA was isolated and RT-PCR was vity than c-Jun homodimers (Allegretto et al, 1990; Suzuki et al, performed. The results shown in the ¢gure are representative of ¢ve sepa- 1991). Thus histamine may enhance AP-1 binding and transcrip- rate experiments. tional activity by inducing c-Fos expression and thus shifting

Figure 7. The inhibition by several signal inhibitors of histamine-induced NGF promoter activity (a), NGF secre- tion (b), TRE-dependent transcriptional activity (c), and c-Fos mRNA expres- sion (d). (a), (c) Keratinocytes were transi- ently transfected with pNGF luc (a)or p4 TRE1-TATA-luc (c), together with pRSV-b-gal, then preincubated with 10^6 M H-89,10^6 M calphostin C (CalC), or 10^5 M PD98059 for 30 min, and then incubated with 10^6 M histamine (HA). After 6 h, - tive luciferase activities were analyzed. (b), (d) Keratinocytes were preincubated with in- hibitors, and then incubated with histamine as above. Culture supernatants were har- vested after 24 h whereas mRNA was ex- tracted after 30 min. In (a), (b), (c), values are mean7SEM of ¢ve separate experi- Ã ments. po0.05 versus control values, wpo 0.05 versus values with histamine alone, by one-way ANOVA with Sche¡e’s multiple comparison test. The results shown in (d)are representative of ¢ve separate experiments. VOL. 121, NO. 3 SEPTEMBER 2003 HISTAMINE ENHANCES NGF PRODUCTION IN KERATINOCYTES 575

Figure 9. The e¡ects of histamine on ERK1/2 (a), JNK (b), or p38 MAPK (c) phosphorylation. Keratinocytes were preincubated with 10^5 M pyrilamine (PR), 10^6 M calphostin C (CalC), 10^5 M PD98059, or 10^6 M AG1478 for 30 min, and then incubated with 10^6 M histamine (HA) for Figure 8. The histamine-induced e¡ects on PKC. (a) Keratinocytes ^6 10 min. Crude protein extracts (50 mg) were subjected to SDS-PAGE fol- were incubated with medium alone or with 10 M histamine (HA) for 10 lowed by western blot analysis. The membranes were blotted with anti- min. PKC activity was measured in cytosolic and membrane fractions. The 7 à phospho-ERK antibody (upper panel), and then stripped and reprobed data are mean SEM of ¢ve separate experiments. po0.05 versus control with anti-ERK antibody (lower panel). (b), (c) Keratinocytes were incubated values (Con), by paired t test. (b) Keratinocytes were preincubated with ^6 ^5 ^5 with 10 M histamine for 10 min, and cellular protein extracts were pyrilamine (PR; 10 M, 30 min), cimetidine (CM; 10 M, 30 min), probed with antiphospho-JNK (b) or antiphospho-p38 antibody (c)(upper U73122 (10^6 M, 30 min), or pertussis toxin (PTX; 100 ng per ml, 16 h), ^6 panels), and then reprobed with anti-JNK or anti-p38 antibodies (lower pa- and then incubated with 10 M histamine (HA) for 10 min. PKC activity nels), respectively. As a positive control, keratinocytes were incubated with in the membrane fraction was analyzed. The data are mean7SEM of ¢ve à 10 ng per ml tumor necrosis factor a (TNF-a) for 10 min, and were ana- separate experiments. po0.05 versus control values with medium alone, w lyzed as above. The results shown in the ¢gure are representative of ¢ve po0.05 versus values with histamine alone, by one-way ANOVA with separate experiments. Sche¡e’s multiple comparison test.

AP-1 composition from c-Jun homodimers to c-Fos/c-Jun hetero- ever, H2 receptor may not be functionally linked to the dimers. PLCPKC pathway as cimetidine did not inhibit PKC activation Histamine-induced c-Fos expression was mediated via ERK by histamine (Fig 8b). activation, which is consistent with the previous study (Megson In monocytic U937 cells, histamine induces c-fos transcription et al,2001).c-fos promoter contains serum response element where by stimulating H2 receptor linked to the adenylate cyclase ternary complex factors like Elk1 are binding (Janknecht et al, protein kinase A pathway (Shayo et al, 1997). c-fos promoter con- 1995; Soh et al, 1999). It is known that ERK-mediated phosphory- tains cAMP response element, where cAMP-response-element- lation of Elk1 promotes its transcriptional activity (Marais et al, binding protein binds and drives c-fos transcription (Janknecht 1993). It is thus anticipated that Elk1 phosphorylation by ERK et al, 1995). The transcriptional activity of cAMP-response- may be involved in histamine-induced c-Fos expression. Hista- element-binding protein is promoted via phosphorylation by mine-induced activation of ERK was dependent on H1 receptor, cAMP-dependent protein kinase A (Lalli and Sassone-Corsi, PKC, and MEK1 (Fig 9a). It is reported that the activation of H1 1994). Protein kinase A was dispensable for histamine-induced c- receptor leads to the stimulation of PLCb, which generates dia- Fos expression in keratinocytes, however (Fig 7). Thus the me- cylglycerol that activates PKC (Hill et al, 1997). PKC is known chanism for c-Fos induction by histamine appears to vary with to phosphorylate and activate c-Raf (Kolch et al, 1993), and the cell type. This is possibly due to the cell type speci¢city in relative activated c-Raf may further phosphorylate and activate MEK1 expression of histamine receptor isotypes and the amounts, activ- catalyzing ERK phosphorylation. Thus the activation of H1 recep- ities, or receptor-coupling status of G-protein subfamilies. tor may trigger the signaling cascade of PLCPKCc- MEK1 inhibitor PD98059 completely inhibited histamine- RafMEK1ERK, and resultantly induce c-fos transcription. induced c-Fos expression whereas its inhibition on NGF pro- The activation of this cascade is insensitive to pertussis toxin, in- duction and AP-1 transcriptional activity was incomplete (Fig dicating that pertussis-toxin-insensitive G-proteins like Gq/11 may 7). On the other hand, these were completely inhibited by PKC link H1 receptors to PLC in keratinocytes. In certain cell types, inhibitor calphostin C (Fig 7). This indicates that PKC may H2 receptor, as well as H1 receptor, is known to be linked to c-Fos- and ERK-independently stimulate AP-1 activity and PLC activation (Del Valle and Ganz, 1997). In keratinocytes, how- NGF production. PKC is known to enhance the DNA binding 576 KANDA AND WATANABE THE JOURNAL OF INVESTIGATIVE DERMATOLOGY activity of c-Jun by dephosphorylating serine residues adjacent in human monocytic cells: Distinct involvement of c-jun N-terminal kinase in to the DNA binding domain (Boyle et al,1991;Goodeet al, LPS-induced CD44 expression. JImmunol169:5660^5672, 2002 Goode N, Hughes K, Woodgett JR, Parker PJ: Di¡erential regulation of glycogen 1992), which may be the e¡ect independent from ERK and synthase kinase-3b by protein kinase C isotypes. JBiolChem267:16878^16882, c-Fos. 1992 In contrast to c-Fos, c-Jun mRNA level was not increased by Hill SJ, Ganellin CR, Timmerman H, et al: International Union of Pharma- histamine (Fig 6), which may be because histamine failed to acti- cology XIII. Classi¢cation of histamine receptors. Pharmacol Rev 49:253^278, vate JNK (Fig 9b), which phosphorylates c-Jun at N-terminal 1997 serine residues. c-jun promoter contains atypical TRE where Horigome K, Pryor JC, Bullock ED, Johnson EMJ: Mediator release from mast cells by nerve growth factor. Neurotrophin speci¢city and receptor mediation. c-Jun/activating transcription factor 2 heterodimer constitutively JBiolChem268:14881^14887, 1993 binds, and phosphorylation of the heterodimer by JNK enhances Horigome K, Bullock ED, Johnson EMJ: E¡ects of nerve growth factor on rat peri- c-jun transcription (Karin, 1995). Though JNK increases transcrip- toneal mast cells. Survival promotion and immediate-early gene induction. tional activity of c-Jun, JNK may not be involved in histamine- JBiolChem269:2695^2702, 1994 induced activation of AP-1 in keratinocytes. Robinson and Iannone F, de Bari C: Dell’Accio F, Covelli M, Patella V, Lo Bianco G, Lapadula G: Dickenson (2001) also reported that histamine did not activate Increased expression of nerve growth factor (NGF) and high a⁄nity NGF re- ceptor (p140 TrkA) in human osteoarthritic chondrocytes. Rheumatology JNK in hamster smooth muscle cell line DDT1MF-2. On the 41:1413^1418, 2002 other hand, histamine activated JNK in H2-receptor-transfected Janknecht R, Cahill MA, Nordheim A: Signal integration at the c-fos promoter. human embryonic kidney cells (Wang et al, 2000). Thus JNK reg- Carcinogenesis 16:443^450, 1995 ulation via H1 or H2 receptor may vary depending on cell type. Jiang WY, Chattedee AD, Raychaudhuri SP, Raychaudhuri SK, Farber EM: Mast The level of endogenous histamine produced by keratinocytes cell density and IL-8 expression in nonlesional and lesional psoriatic skin. ^7 Int J Dermatol 40:699^703, 2001 was less than threshold (10 M) for NGF induction. More than Kanda N,Watanabe S: Histamine inhibits the production of interferon-induced pro- threshold level of histamine in vivo may possibly be derived from tein of 10 kDa in human squamous cell carcinoma and melanoma. JInvestDer- mast cells accumulated in the dermis of in£ammatory skin matol 119:1411^1419, 2002 lesions. Thus histamine released from mast cells may induce Kanda N,Watanabe S: 17b-estradiol inhibits MCP-1 production in human keratino- keratinocytes to produce NGF, and the induced NGF may in cytes. JInvestDermatol, 120:1058^1066, 2003 turn promote histamine release from mast cells. Such a positive Karin M: Regulation of AP-1 activity by mitogen-activated protein kinases. JBiol Chem 270:16483^16486, 1995 feedback loop of histamine/NGF may amplify in vivo skin in£ammat- Kinkelin I, Motzig S, Koltzenburg M, Brocker E-B: Increase in NGF content and ion as NGF induces neurons to synthesize immunomodulatory nerve ¢ber sprouting in human allergic contact eczema. Cell Tissue Res 302: neuropeptides like substance P or calcitonin-gene-related peptide 31^37, 2000 (Linsay and Harmar, 1989), or induces keratinocytes to produce Kohda F, Koga T, Uchi H, Urabe K, Furue M: Histamine-induced IL-6 and IL-8 in£ammatory cytokines or chemokines (Raychaudhuri et al, production are di¡erentially modulated by IFN-g and IL-4 in human kerati- 2000), or enhances the survival or proliferation of neutrophils nocytes. J Dermatol Sci 28:34^41, 2002 Kolch W, Heldecker G, Kochs G, et al: Protein kinase Ca activates RAF-1 by direct and T cells (Toyoda et al, 2002). H1 receptor antagonist may inhi- phosphorylation. Nature 364:249^252, 1993 bit histamine-induced NGF production and thus block the Lalli E, Sassone-Corsi P: Signal transduction and gene regulation: The nuclear re- in£ammation, epidermal hyperplasia, and hyperinnervation sponse to cAMP. J Biol Chem 269:17359^17362, 1994 in in£ammatory skin lesions like psoriasis, atopic dermatitis, Lane SJ, Adcock IM, Richards D, Hawrylowicz C, Barnes PJ, Lee TH: Corticoster- or allergic contact dermatitis. As coculture with murine 3T3 oid-resistant bronchial asthma is associated with increased c-fos expression in monocytes and T lymphocytes. JClinInvest102:2156^2164, 1998 ¢broblasts increased constitutive NGF secretion by human Li Y-S, Kolattukudy PE: Functional role of the cis-acting elements in human keratinocytes approximately 8-fold compared to controls in our monocyte chemotactic protein-1 gene in the regulation of its expression preliminary experiments, we are now studying whether hista- by phorbol ester in human glioblastoma cells. MolCellBiochem141:121^128, mine may also increase NGF production in the coculture system, 1994 which is more relevant to in vivo skin morphology. Linsay RM, Harmar AJ: Nerve growth factor regulates expression of neuropeptide genes in adult sensory neurons. Nature 337:362^364, 1989 Lipnik-Strangelj M, Carman-Krzan M: The e¡ects of histamine and interleukin-6 on NGF release from cortical astrocytes in primary culture. P£ugers Arch 440 This work was supported in part by aid from Kanae Medical Research Foundation. (5 Suppl.):R99^R100, 2000 Lipnik-Strangelj M, Carman-Krzan M: The in£uence of histamine H1-, H2-, and H3 -receptor antagonists on histamine stimulated NGF release from cultured REFERENCES astrocytes. In£amm Res 50 (Suppl. 2):S84^S85, 2001 Lipnik-Strangelj M, Juric D, Carman-Krzan M: Histamine-induced synthesis and Alenius H, Laouini D, Woodward A, et al: Mast cells regulate IFN-g expression in secretion of nerve growth factor from astrocytes. In£amm Res 47 (Suppl. the skin and circulating IgE levels in allergen-induced skin in£ammation. 1):S34^S35, 1998 J Allergy Clin Immunol 109:106^113, 20 02 Marais R, Wynne J, Treisman R: The SRF accessory protein Elk-1 contains a Allegretto EA, Smeal T, Angel P, Spiegelman BM, Karin M: DNA-binding activity growth factor-regulated transcriptional activation domain. Cell 73:381^393, of Jun is increased through its interaction with Fos. JCellBiochem42: 1993 193^206, 1990 Matsuda H, KannanY, Ushio H, KisoY, Kanemoto T, Suzuki H, KitamuraY: Nerve Boyle WJ, Smeal T, De¢ze LH, Angel P,Woodgett JR, Karin M, Hunter T: Activa- growth factor induces development of connective tissue-type mast cells in vitro tion of protein kinase C decreases phosphorylation of c-Jun at sites that nega- from murine bone marrow cells. JExpMed174:7^14, 1991 tively regulate its DNA-binding activity. Cell 64:573^584, 1991 Matsuda H, Koyama H, Sato H, et al: Role of nerve growth factor in cutaneous Bruni A, Bigon E, Boarato E, Mietto L, Leon A, To¡ano G: Interaction between wound healing: Accelerating e¡ects in normal and healing-impaired diabetic nerve growth factor and lysophosphatidylserine on rat peritoneal mast cells. mice. JExpMed187:297^306, 1998 FEBS Lett 138:190^192, 1982 Megson AC, Walker EM, Hill SJ: Role of protein kinase Ca in signaling from the Cartwright M, Martin S, D’Mello S, Heinrich G: The human nerve growth factor histamine H1 receptor to the nucleus. Mol Pharmacol 59:1012^1021, 2001 gene: Structure of the promoter region and expression in L929 ¢broblasts. Mol Migliaccio S,Washburn TF, Fillo S, Rivera H, Teti A, Korach KS, Wetsel WC: Mod- Brain Res 15:67^75, 1992 ulation of levels in mouse uterus by protein kinase C isoen- Coge F, Guenin S-P, Rique H, Boutin JA, Galizzi J-P: Structure and expression of zymes. Endocrinology 139:4598^4606, 1998 the human histamine H4-receptor gene. Biochem Biophys Res Commun 284: Oda T, Morikawa N, SaitoY, MasuhoY, Matsumoto S: Molecular cloning and char- 301^309, 2001 acterization of a novel type of histamine receptor preferentially expressed in Del Valle J, Ganz I: Novel insights into histamine H2 receptor biology. Am J Physiol leukocytes. J Biol Chem 275:36781^36786, 2000 273:G987^G996, 1997 Pearce F,Thompson HL: Some characteristics of histamine secretion from rat perito- Di Marco E, Mathor M, Bondanza S, Cutuli N, Marchisio PC, De Cancedda R, neal mast cells stimulated with nerve growth factor. JPhysiol372:379^393, 1986 Luca M: Nerve growth factor binds to normal human keratinocytes through Pincelli C, Sevianani C, Manfredini R, et al: Expression and function of nerve high and low a⁄nity receptors and stimulates their growth by a novel auto- growth factor and nerve growth factor receptor on cultured keratinocytes. crine loop. JBiolChem268:22838^22846, 1993 J Invest Dermatol 103:13^18, 1994 Gee K, Lim W, Ma W, Nandan D, Diaz-Mitoma F, Kozlowski M, Kumar A: Di¡er- Raychaudhuri SP, Jiang W-Y, Farber EM: Psoriatic keratinocytes express high levels ential regulation of CD44 expression by lipopolysaccharide (LPS) and TNF-a of nerve growth factor. Acta DermVenereol (Stockh) 78:84^86, 1998 VOL. 121, NO. 3 SEPTEMBER 2003 HISTAMINE ENHANCES NGF PRODUCTION IN KERATINOCYTES 577

Raychaudhuri SP, Farber EM, Raychaudhuri SK: Role of nerve growth factor in Tjandrawinata RR, Dahiya R, Hughes-Fulford M: Induction of cyclo-oxygenase-2 RANTES expression by keratinocytes. Acta Derm Venereol 80:247^250, 2000 mRNA by prostaglandin E2 in human prostatic carcinoma cells. Br J Cancer Risse-Hackl G, Adamkiewicz J,Wimmel A, Schuermann M: Transition from SCLC 75:1111^1118, 1997 to NSCLC phenotype is accompanied by an increased TRE-binding activity Toyoda M, Nakamura M, Makino T, Hino T, Kagoura M, Morohashi M: Nerve and recruitment of speci¢c AP-1 proteins. 16:3057^3068, 1998 growth factor and substance P are useful plasma markers of disease activity in Robinson AJ, Dickenson JM: Activation of the p38 and p42/p44 mitogen-activated atopic dermatitis. Br J Dermatol 147:71^79, 2002 protein kinase families by the histamine H1 receptor in DDT1MF-2 cells. Br Wang LD,Wang M,Todisco A, Grand E, del Valle J: The human histamine H2 recep- J Pharmacol 133:1378^1386, 2001 tor regulates c-jun and c-fos in a di¡erential manner. AmJ Cell Physiol 278:1246^ Shayo C, Davio C, Brodsky A, et al: Histamine modulates the expression of c-fos C1255, 2000 through cyclic AMP production via the H2 receptor in the human promono- Widmann C, Gibson S, Jarpe MB, Johnson GL: Mitogen-activated protein kinase: cytic cell line U937. Mol Pharmacol 51:983^990, 1997 Conservation of a three-kinase module from yeast to human. Physiol Rev Soh J-W,Lee EH, Prywes R,Weinstein B: Novel roles of speci¢c isoforms of protein 79:143^180, 1999 kinase C in activation of the c-fos serum response element. MolCellBiol Zellmer S, Gaunitz F, Salvetter J, Surovoy A, Ressig D, Gebhardt R: Long-term ex- 19:1313^1324, 1999 pression of foreign genes in normal human epidermal keratinocytes after Suzuki T, Okuno H, Yoshida T, Endo T, Nishina H, Iba H: Di¡erence in transcrip- transfection with lipid/DNA complexes. Histochem Cell Biol 115:41^47, 2001 tional regulatory function between c-Fos and Fra-2. Nucl Acids Res 19:5537^ Zwick E, Hackel PO, Prenzel N, Ulrich A:The EGF receptor as central transducer of 5542, 1991 heterologous signaling systems.Trends Pharmacol Sci 20:408^412, 1999