Histamine enhances the production of human β-defensin-2 in human keratinocytes

Naoko Kanda, Shinichi Watanabe

Abstract

The anti-microbial peptide human β-defensin-2 (hBD-2), produced by epidermal keratinocytes, plays pivotal roles in anti-microbial defense, inflammatory dermatoses, and wound repair. hBD-2 induces histamine release from mast cells. We examined the in vitro effects of histamine on hBD-2 production in normal human keratinocytes. Histamine enhanced TNF-α- or IFN-γ-induced hBD-2 secretion and mRNA expression. Histamine alone enhanced transcriptional activities of NF-κB and activator protein-1 (AP-1) and potentiated TNF-α-induced NF-κB and AP-1 activities or IFN-γ-induced NF-κB and STAT1 activities. Antisense oligonucleotides against NF-κB components p50 and p65, AP-1 components c-Jun and c-Fos, or H1 antagonist pyrilamine suppressed hBD-2 production induced by histamine plus TNF-α or IFN-γ. Antisense oligonucleotide against STAT1 only suppressed hBD-2 production induced by histamine plus IFN-γ. Histamine induced serine phosphorylation of inhibitory NF-κBα (IκBα) alone or together with TNF-α or IFN-γ. Histamine induced c-Fos mRNA expression alone or together with TNF-α, whereas it did not further increase c-Jun mRNA levels enhanced by TNF-α. Histamine induced serine phosphorylation of STAT1 alone or together with IFN-γ, whereas it did not further enhance IFN-γ-induced tyrosine phosphorylation of STAT1. The histamine-induced serine phosphorylation of STAT1 was suppressed by MAPKK (MEK) inhibitor PD98059. These results suggest that histamine stimulates H1 receptor and potentiates TNF-α- or IFN-γ-induced hBD-2 production dependent on NF-κB, AP-1, or STAT1 in human keratinocytes. Histamine may potentiate anti-microbial defense, skin inflammation, and wound repair via the induction of hBD-2.

  • NF-κB
  • activator protein-1
  • STAT1
  • TNF-α
  • IFN-γ

the anti-microbial peptide human β-defensin-2 (hBD-2) is produced by epidermal keratinocytes, and its production is enhanced in cutaneous infection, inflammatory dermatoses like psoriasis vulgaris, and wounds (9, 34). hBD-2 kills gram-negative bacteria and protects skin from their infection (10). Moreover, hBD-2 potentiates skin inflammation and wound repair; hBD-2 binds CC chemokine receptor-6 and induces chemotaxis of memory T cells or immature dendritic cells (49) or induces the production of IL-6, CC chemokine ligand-20 (CCL20), CXC chemokine ligand-10, or CCL5 in keratinocytes (34). hBD-2 accelerates wound repair by inducing migration and proliferation of keratinocytes (34).

Dermal mast cells are increased and activated in cutaneous infection, inflammatory diseases, or wounds and regulate innate and adaptive immune responses in these situations (6, 7, 36). The chemical mediator histamine is mainly released from dermal mast cells; however, in late phases of infection or allergic inflammation (44, 47), histamine is also newly synthesized and released from neutrophils via the induction of histidine decarboxylase, a rate-limiting enzyme in histamine synthesis. Histamine manifests proinflammatory effects; histamine induces the production of IL-6, IL-8, or granulocyte macrophage colony-stimulating factor (GM-CSF) in keratinocytes, alone or together with TNF-α, IL-1, or IFN-γ (25, 12, 20, 21). Histamine accelerates wound repair by inducing angiogenesis or recruitment of macrophages (36). Histamine binds the cell surface G protein-coupled receptors H1, H2, H3, or H4 (4). Individual histamine receptors are coupled to different G proteins like Gq/11, Gi/o, or Gs and to different signals; H1 receptors couple to the phosphatidylinositol-specific PLC (PI-PLC)/PKC pathway, and H2 receptors couple to cAMP/PKA or PI-PLC/PKC pathways, depending on the cell type, while stimulation of H3 or H4 receptors inhibits the cAMP/PKA pathway (11, 43). Histamine thus manifests differential effects on different cell types according to the relative expression of individual receptors or linkage to G proteins (4, 29, 43). H1 and H2 receptors are detected in epidermal keratinocytes (12).

It has recently been reported that hBD-2 binds unidentified receptors on mast cells and enhances their chemotaxis and histamine release (31, 33), indicating a relationship between hBD-2 and histamine. However, whether histamine might alter hBD-2 expression in epidermal keratinocytes has not been investigated. In this study, we investigated the in vitro effects of histamine on hBD-2 production in normal human keratinocytes. We found that histamine enhanced TNF-α- or IFN-γ-induced hBD-2 production in these cells.

MATERIALS AND METHODS

Reagents.

Histamine, pyrilamine, cimetidine, thioperamide, and 2′-amino-3′-methoxyflavone (PD98059) were purchased from Sigma-Aldrich (St. Louis, MO). Recombinant human IFN-γ and TNF-α were from R&D Systems (Minneapolis, MN). 4-(4-Fluorophenyl)-2-(4-hydroxyphenyl)-5-(4-pyridyl)-1H-imidazole (SB202190) and pertussis toxin were from Calbiochem (La Jolla, CA).

Culture of keratinocytes.

Human neonatal foreskin keratinocytes (Clonetics, Walkersville, MD) were cultured in serum-free keratinocyte growth medium (KGM) (Clonetics) consisting of keratinocyte basal medium (KBM) supplemented with 0.5 μg/ml hydrocortisone, 5 ng/ml epidermal growth factor, 5 μg/ml insulin, and 0.5% bovine pituitary extract. Cells in the third passage were used.

hBD-2 secretion.

Keratinocytes (5 × 104/well) were seeded in triplicate into 24-well plates in 0.4 ml of KGM, adhered overnight, and washed and were induced to differentiate by incubation with a high-calcium concentration (1.25 mM) of KBM for 48 h, since hBD-2 production is enhanced in differentiated keratinocytes (35). The cells were incubated with 10 ng/ml TNF-α or IFN-γ and/or indicated concentrations of histamine in high-calcium KBM for 48 h. The hBD-2 concentration of the culture supernatants was measured by ELISA (Peprotech, Rocky Hill, NJ). In some experiments, keratinocytes were preincubated with indicated concentrations of signal inhibitors for 30 min or pertussis toxin for 16 h before the addition of histamine or cytokines. The concentration of TNF-α (10 ng/ml) was adopted in this study because it did not reduce the viability of keratinocytes (>95% viable) and was generally used for in vitro hBD-2 production in keratinocytes in previous studies (10, 40), and it corresponds to the TNF-α concentration in supernatants from lipopolysaccharide-stimulated human macrophages or zymosan-stimulated human monocytes (3, 26). Thus this level of TNF-α may be likely in skin wounds, infections, or inflammatory diseases. The concentrations of MAPKK (MEK) inhibitor PD98059 (10 μM) and p38 MAPK inhibitor SB202190 (1 μM) were chosen according to the IC50 values for respective signals, 2 or 0.35 μM, respectively (38, 18). The concentration of histamine receptor antagonists (10 μM) was adopted because individual antagonists at this concentration distinctively inhibited the respective receptor-mediated cytokine or chemokine production in keratinocytes or chemotaxis of mast cells in vitro (14, 25, 29). The concentration of pertussis toxin (100 ng/ml) was adopted because this concentration is generally used for the specific inhibition of Gi/o protein-mediated effects (14, 29). The indicated concentrations of various inhibitors or antagonists did not reduce the viability of keratinocytes (>93% viable).

RT-PCR.

Keratinocytes were incubated as above for 30 min or 12 h to analyze the mRNA levels of c-Fos and c-Jun or of hBD-2, respectively. Total cellular RNA was extracted and reverse transcribed to produce cDNA. The cDNA was thermocycled for PCR using primers for c-Fos, c-Jun, hBD-2, and GAPDH as described previously (20, 28). PCR products were analyzed by electrophoresis, and densitometric analysis was performed by ATTO lane analyzer, version 3 (ATTO, Osaka, Japan). c-Fos, c-Jun, and hBD-2 mRNA levels were normalized to those of GAPDH and are shown as a fold induction.

Plasmid and transfection.

pNF-κB-luc, pγ-activated site (GAS)-luc, and plasmid activator protein-1 (pAP-1)-luc containing four copies of NF-κB, two copies of STAT1, and seven copies of AP-1-binding sequences, respectively, in front of the TATA box upstream of the firefly luciferase reporter were purchased from Clontech (San Jose, CA). Transient transfection was performed with Fugene 6 (Roche, Indianapolis, IN) as described previously (19). Keratinocytes were plated in 35-mm dishes and grown to ∼60% confluence. pNF-κB-luc, pGAS-luc, or pAP-1-luc and herpes simplex virus thymidine kinase promoter-linked Renilla luciferase vector (pRL-tk, Clontech) mixed with Fugene 6 were added to the keratinocytes. After 6 h, the cells were washed and incubated in high-calcium KBM for 48 h and then treated with 10 ng/ml IFN-γ or TNF-α and/or 1 μM histamine. After 18 h, firefly and Renilla luciferase activities of the cell lysates were quantified by the dual-luciferase assay system (Promega, Madison, WI). The transcriptional activities of NF-κB, STAT1, and AP-1 were expressed as ratios of firefly-to-Renilla luciferase activity.

Cell-based ELISA.

The phosphorylation status of STAT1 or inhibitory NF-κBα (IκBα) was analyzed by cell-based ELISA as described previously (19), using specific antibodies (Raybiotech, Norcross, GA, for Ser727- or Tyr701-phosphorylated or pan STAT1; SuperArray Bioscience, Frederick, MD, for Ser32/Ser36-phosphorylated or pan IκBα). Keratinocytes (1 × 104/well) were seeded in triplicate into 96-well plates in 0.1 ml of KGM, adhered overnight, washed, and incubated with high-calcium KBM for 48 h. The cells were preincubated with signal inhibitors for 30 min and then incubated with 1 μM histamine and/or 10 ng/ml TNF-α or IFN-γ for 5 or 15 min to analyze phosphorylation of IκBα or STAT1, respectively. The cells were fixed, quenched, blocked, and incubated with primary antibodies and then with horseradish peroxidase-conjugated secondary antibodies; they were then washed and developed. The phosphorylation status is presented as the absorbance ratio of phosphorylated protein to total protein at 450 nm. The protein levels of NF-κB p50 and p65, STAT1α, and AP-1 components c-Jun and c-Fos were examined after 1 h of incubation by cell-based ELISA as above, using monoclonal antibodies against these proteins (Santa Cruz Biotechnology, Santa Cruz, CA). The levels of these proteins are shown as the absorbance ratio of individual protein to GAPDH.

Treatment with antisense oligonucleotides.

Antisense oligonucleotides against NF-κB p50 and p65, STAT1, c-Jun and c-Fos, and control scrambled oligonucleotides were synthesized by phosphoroamidite chemistry as described previously (19, 22). Keratinocytes were transfected with the indicated oligonucleotides (0.2 μM each) premixed with Fugene 6 in KGM for 6 h. The concentration of antisense oligonucleotides (0.2 μM) was selected because this concentration did not reduce the viability of keratinocytes (>94% viable) and selectively reduced the levels of respective proteins in our previous study (22). The cells were treated with high-calcium KBM for 48 h and then incubated with 1 μM histamine and/or 10 ng/ml IFN-γ or TNF-α. In some experiments, keratinocytes were transfected with antisense oligonucleotides together with luciferase reporter vectors.

Western blot analysis.

Keratinocytes incubated with 1 μM histamine for 1 h were lysed with lysis buffer [50 mM HEPES (pH 7.5), 150 mM NaCl, 10% glycerol, 1% Triton X-100, 1.5 mM MgCl2, 100 mM NaF, 100 mM sodium orthovanadate, and 1 mM EGTA (pH 7.7)], followed by centrifugation for 20 min at 14,000 g at 4°C. The supernatant proteins were subjected to SDS-PAGE and transferred to a nitrocellulose membrane. The membrane was blocked and incubated with anti-human c-Fos, c-Jun, NF-κB p50 or p65, STAT1, or GAPDH antibodies (Santa Cruz Biotechnology) followed by peroxidase-conjugated secondary antibodies (Santa Cruz Biotechnology). The blots were developed with an enhanced chemiluminescence kit (Amersham, Arlington Heights, IL).

Statistical analyses.

One-way ANOVA with Dunnett's multiple comparison test was used, as shown in Fig. 1A. One-way ANOVA with the Tukey-Kramer multiple comparison test was used, as shown in Figs. 2 and 47. A P value < 0.05 was considered significant.

Fig. 1.

Stimulatory effects of histamine on TNF-α- or IFN-γ-induced human β-defensin-2 (hBD-2) production. A: keratinocytes were incubated with medium alone or with 10 ng/ml TNF-α or IFN-γ in the presence of the indicated concentrations of histamine. At 48 h, hBD-2 secretion was analyzed. Data are means ± SD of triplicate cultures and represent 4 experiments. *P < 0.05 vs. TNF-α alone. †P < 0.05 vs. IFN-γ alone. B: keratinocytes were incubated with medium alone or with 10 ng/ml TNF-α or IFN-γ and/or 1 μM histamine. At 12 h, hBD-2 and GAPDH mRNA levels were analyzed. The results represent 4 separate experiments.

Fig. 2.

Effects of antisense oligonucleotides on histamine- and/or TNF-α-induced hBD-2 secretion and activities or protein levels of transcription factors. AC: keratinocytes were transfected with antisense oligonucleotides (AS) against transcription factors or control (Con) scrambled oligonucleotides (0.2 μM each), together with pRL-tk and pNF-κB-luc, pAP-1-luc, or pGAS-luc. The cells were treated with 10 ng/ml TNF-α and/or 1 μM histamine. Transcriptional activities were analyzed at 18 h. DI: keratinocytes were transfected with antisense oligonucleotides without luciferase (luc) vectors and were treated as above. hBD-2 secretion was analyzed at 48 h (D). Protein levels were analyzed at 1 h (EI). Data are means ± SE (n = 4). AP-1, activator protein-1. *P < 0.05 vs. controls. †P < 0.05 vs. TNF-α alone. ‡P < 0.05 vs. TNF-α plus histamine.

Fig. 3.

Effects of antisense oligonucleotides on protein levels of transcription factors. Keratinocytes were transfected with antisense oligonucleotides against transcription factors or control scrambled oligonucleotides and treated with 1 μM histamine for 1 h. Cell lysates were analyzed for the indicated transcription factors or GAPDH by Western blot. Results are representative of 4 separate experiments.

Fig. 4.

Effects of antisense oligonucleotides on histamine- and/or IFN-γ-induced hBD-2 secretion and activities or protein levels of transcription factors. AC: keratinocytes were transfected with antisense oligonucleotides against transcription factors or control scrambled oligonucleotides (0.2 μM each), together with pRL-tk and pNF-κB-luc, pAP-1-luc, or pGAS-luc. The cells were treated with 10 ng/ml IFN-γ and/or 1 μM histamine. Transcriptional activities were analyzed at 18 h. DI: keratinocytes were transfected with antisense oligonucleotides without luciferase vectors and were treated as above. hBD-2 secretion was analyzed at 48 h (D). Protein levels were analyzed at 1 h (EI). Data are means ± SE (n = 4). *P < 0.05 vs. controls. †P < 0.05 vs. IFN-γ alone. ‡P < 0.05 vs. IFN-γ plus histamine.

RESULTS

Enhancement of TNF-α- or IFN-γ-induced hBD-2 production by histamine.

We examined the effects of histamine on TNF-α- or IFN-γ-induced hBD-2 production in keratinocytes. Histamine did not alter basal hBD-2 secretion. However, it dose-dependently increased TNF-α- or IFN-γ-induced secretion (Fig. 1A); the stimulatory effect of histamine manifested at 0.1 μM and was maximized at 1 μM, which increased TNF-α- or IFN-γ-induced hBD-2 secretion 4.0-fold or 6.6-fold, respectively. The addition of histamine did not increase the number of keratinocytes compared with TNF-α or IFN-γ alone: (mean ± SE) (n = 4) 96 ± 10 or 95 ± 12%, respectively. In parallel with protein secretion, histamine enhanced TNF-α- or IFN-γ-induced hBD-2 mRNA expression (Fig. 1B).

hBD-2 promoter contains NF-κB, STAT1, or AP-1-binding sequences, and these transcription factors may transactivate this gene (30). We thus examined whether histamine alone or together with TNF-α or IFN-γ alters NF-κB, AP-1, or STAT1 activities. Histamine alone increased NF-κB and AP-1 activities (2.0- and 2.57-fold compared with the controls, respectively) (Fig. 2, A and B) but did not alter STAT1 activity (Fig. 2C). TNF-α alone increased NF-κB and AP-1 activities 6.13- and 1.68-fold compared with the controls, respectively, but did not alter those of STAT1. Histamine further enhanced TNF-α-induced NF-κB (Fig. 2A) and AP-1 activities (Fig. 2B) but did not alter STAT1 activity in the presence of TNF-α (Fig. 2C). Transcription factor AP-1 is composed of homodimers of Jun family proteins or heterodimers of Fos and Jun family proteins (2, 6). The protein level of the AP-1 component c-Jun was increased by TNF-α (Fig. 2G), and histamine did not alter the c-Jun level either in the presence or the absence of TNF-α. The protein level of the AP-1 component c-Fos was constitutively low (Fig. 2H). Histamine potently and TNF-α modestly increased c-Fos protein levels, and the addition of both had additive effects on the c-Fos levels, whereas the protein levels of NF-κB p50 or p65 or STAT1 were not altered by histamine and/or TNF-α (Fig. 2, E, F, and I). Antisense oligonucleotides against p50, p65, c-Jun, c-Fos, or STAT1 selectively reduced the protein levels of respective transcription factors (Figs. 2, EH, and 3). Antisense p50, p65, c-Jun, or c-Fos suppressed hBD-2 secretion by TNF-α or TNF-α plus histamine (Fig. 2D) in parallel with the specific inhibition of the activities of respective transcription factors (Fig. 2, A and B). On the other hand, antisense oligonucleotide against STAT1 did not suppress TNF-α- or TNF-α-plus-histamine-induced hBD-2 secretion (Fig. 2D), although it specifically reduced the activity or protein level of STAT1 (Fig. 2, C and I). Control scrambled oligonucleotides did not reduce hBD-2 secretion or activities or protein levels of transcription factors. These results indicate that hBD-2 production induced by TNF-α or TNF-α plus histamine is dependent on NF-κB and AP-1, and that TNF-α and histamine may promote hBD-2 production by activating NF-κB and AP-1. The activation of AP-1 was associated with the increased levels of c-Jun by TNF-α and those of c-Fos by histamine and/or TNF-α.

IFN-γ alone increased the transcriptional activities of NF-κB (Fig. 4A) or STAT1 (Fig. 4C) 2.48- or 17.53-fold compared with the controls, respectively, but did not alter AP-1 activity or c-Jun or c-Fos protein levels (Fig. 4, B, G, and H). AP-1 activity in the presence of IFN-γ alone was reduced by antisense c-Jun but not by antisense c-Fos (Fig. 4B), indicating that AP-1 activity in keratinocytes with IFN-γ alone may be mediated by c-Jun but not by c-Fos. hBD-2 secretion induced by IFN-γ alone was suppressed by antisense oligonucleotides against NF-κB p50 or p65 or STAT1 but not by antisense c-Jun or c-Fos (Fig. 4D), indicating the requirement for NF-κB and STAT1 but not for AP-1. Histamine synergistically enhanced IFN-γ-induced NF-κB and STAT1 activities (Fig. 4, A and C), while the activity of AP-1 (Fig. 3B) or c-Jun and c-Fos levels (Fig. 4, G and H) in IFN-γ-plus-histamine-stimulated cells was similar to that of histamine alone. IFN-γ-plus-histamine-induced hBD-2 secretion was reduced by antisense c-Jun or c-Fos in addition to antisense p50, p65, and STAT1 (Fig. 4D), in parallel with the inhibition of protein levels (Fig. 4, EI) and activities (Fig. 4, AC). These results suggest that hBD-2 production induced by IFN-γ alone is dependent on NF-κB and STAT1, whereas that induced by IFN-γ plus histamine is dependent on NF-κB, STAT1, and AP-1. This indicates that IFN-γ plus histamine may promote hBD-2 production by activating NF-κB, STAT1, and AP-1.

Involvement of H1 receptor in the enhancement of hBD-2 production by histamine.

Histamine binds four different G protein-coupled receptors, H1–H4 (43). We examined the receptor subtypes involved in histamine-induced enhancement of hBD-2 production and NF-κB, AP-1, or STAT1 activities. H1 antagonist pyrilamine suppressed histamine-induced enhancement of NF-κB (Fig. 5A), AP-1 (Fig. 5B) or STAT1 activities (Fig. 5C), hBD-2 secretion (Fig. 5D), or mRNA expression (Fig. 5E) alone or together with TNF-α or IFN-γ, whereas H2 antagonist cimetidine or H3/4 antagonist thioperamide or pertussis toxin did not. These results suggest that H1 receptor and pertussis toxin-resistant G protein(s) may mediate histamine-induced enhancement of hBD-2 production and NF-κB, AP-1, or STAT1 activities.

Fig. 5.

Effects of histamine receptor antagonists on histamine-induced NF-κB, AP-1, or STAT1 activities and hBD-2 production. A–C: keratinocytes transfected with pRL-tk and pNF-κB-luc, pAP-1-luc, or pGAS-luc were preincubated with 10 μM pyrilamine, cimetidine, or thioperamide (H1, H2, or H3/4 antagonist, respectively) for 30 min or with 100 ng/ml pertussis toxin (PT) for 16 h. The cells were incubated with 10 ng/ml TNF-α or IFN-γ and/or 1 μM histamine. At 18 h, transcriptional activities were analyzed. D and E: keratinocytes without transfection were preincubated and incubated as above. hBD-2 secretion (D) or mRNA levels (E) were analyzed at 48 or 12 h, respectively. Data in AD are means ± SE (n = 4). Results in E represent 4 experiments. *P < 0.05 vs. controls. †P < 0.05 vs. histamine alone. ‡P < 0.05 vs. TNF-α alone. §P < 0.05 vs. TNF-α plus histamine. ¶P < 0.05 vs. IFN-γ alone. ||P < 0.05 vs. IFN-γ plus histamine.

Mechanisms for activation of AP-1, NF-κB, and STAT1 induced by histamine in concert with TNF-α or IFN-γ.

NF-κB p50/p65 is normally sequestered in cytoplasm by its interaction with IκB, and on stimulation, IκB is phosphorylated, ubiquitinated, and degraded, leading to the release of active NF-κB and its nuclear translocation and DNA binding (41). IKK is reported to catalyze serine phosphorylation of IκB (16). We thus examined whether histamine may induce phosphorylation of IκB, alone or together with TNF-α or IFN-γ. Histamine alone induced Ser32/Ser36 phosphorylation of IκBα (Fig. 5A). TNF-α alone more potently induced Ser32/Ser36 phosphorylation of IκBα. Histamine plus TNF-α had additive effects on phosphorylation of IκBα. IFN-γ alone enhanced Ser32/Ser36 phosphorylation of IκBα (Fig. 6A). Histamine plus IFN-γ had additive effects on phosphorylation of IκBα. These results suggest that histamine enhances phosphorylation of IκBα, alone or together with TNF-α or IFN-γ.

Fig. 6.

Histamine- and/or TNF-α- or IFN-γ-induced IκBα phosphorylation or c-Fos or c-Jun mRNA expression. Keratinocytes were incubated with 10 ng/ml TNF-α or IFN-γ and/or 1 μM histamine. Serine phosphorylation of IκBα (A) or c-Fos and c-Jun mRNA levels (B) were analyzed at 5 or 30 min, respectively. Data in A are means ± SE (n = 4). Results in B represent 4 experiments. *P < 0.05 vs. controls. †P < 0.05 vs. TNF-α alone. ‡P < 0.05 vs. IFN-γ alone.

mRNA levels of AP-1 component c-Fos were constitutively low and were potently increased by histamine (Fig. 6B). TNF-α increased c-Fos mRNA levels less potently than histamine, and histamine plus TNF-α additively increased c-Fos mRNA levels. On the other hand, mRNA levels of another AP-1 component, c-Jun, were constitutively higher than those of c-Fos and were moderately increased by TNF-α (Fig. 6B). Histamine did not alter c-Jun levels in either the presence or the absence of TNF-α. IFN-γ did not alter c-Fos or c-Jun mRNA levels in either the presence or the absence of histamine (Fig. 6B). These results suggest that histamine may increase c-Fos levels, while TNF-α may increase c-Fos and c-Jun levels. These effects of histamine and TNF-α may converge on the transcriptional activity of AP-1.

The transcriptional activity of STAT1 is regulated by phosphorylation; phosphorylation of STAT1 at Tyr701 induces its DNA binding, while that at Ser727 enhances its transactivation potential (13). Histamine alone induced Ser727 phosphorylation of STAT1, which was suppressed by MEK inhibitor PD98059 (Fig. 7A), indicating the involvement of MEK/ERK in the Ser727 phosphorylation by histamine. IFN-γ alone induced Ser727 phosphorylation of STAT1, which was suppressed by p38 MAPK inhibitor SB202190 but not by PD98059, indicating the involvement of p38 MAPK in the Ser727 phosphorylation by IFN-γ. Histamine plus IFN-γ had additive effects on Ser727 phosphorylation of STAT1. IFN-γ alone induced Tyr701 phosphorylation of STAT1 (Fig. 7B). Histamine did not alter Tyr701 phosphorylation of STAT1 in either the presence or absence of IFN-γ. These results suggest that histamine may enhance Ser727 phosphorylation of STAT1 alone or together with IFN-γ without altering its Tyr701 phosphorylation. Ser727 phosphorylation by histamine may potentiate transcriptional activity of STAT1 induced by IFN-γ.

Fig. 7.

Histamine- and/or IFN-γ-induced serine or tyrosine phosphorylation of STAT1. Keratinocytes were preincubated with 10 μM PD98059 (PD) or 1 μM SB202190 (SB) for 30 min and then incubated with 10 ng/ml IFN-γ and/or 1 μM histamine. Serine (A) or tyrosine phosphorylation (B) of STAT1 was analyzed at 15 min. Data are the means ± SE (n = 4). *P < 0.05 vs. controls. †P < 0.05 vs. histamine alone. ‡P < 0.05 vs. IFN-γ alone.

DISCUSSION

Histamine enhances hBD-2 production in synergy with TNF-α or IFN-γ in human keratinocytes. Histamine alone enhances transcriptional activities of AP-1 and NF-κB. Histamine further augments TNF-α-induced AP-1 and NF-κB activities and, as a result, upregulates hBD-2 production via the cooperation of both transcription factors (Fig. 8A). On the other hand, histamine potentiates IFN-γ-induced NF-κB and STAT1 activities in addition to its own effects on AP-1 and, as a result, promotes hBD-2 production via cooperation of the three factors (Fig. 8B). Thus histamine may use partially overlapping but distinct mechanisms for hBD-2 induction that are dependent on co-stimuli. Although histamine alone increases AP-1 and NF-κB activities, it cannot by itself induce hBD-2 expression. Presumably, the magnitude of the increases in AP-1 or NF-κB activities is small and insufficient, and/or some other stimuli from TNF-α or IFN-γ may be simultaneously required for hBD-2 expression. TNF-α or IFN-γ may activate transcriptional co-activators like cAMP response element-binding protein (CREB)-binding protein (CBP)/p300 or general transcriptional machinery like TATA-binding proteins and thus enhance their interaction with individual transcription factors (NF-κB, AP-1, STAT1, or other factors) to form a stable transcriptional complex (37, 7). TNF-α or IFN-γ may also enhance the histone acetylation on the gene promoter by CBP/p300 or p/CAF, which may promote the access of various transcription factors to the promoter and the recruitment of RNA polymerase II to the promoter (17). Such effects of TNF-α or IFN-γ may induce hBD-2 transcription in cooperation with histamine. TNF-α or IFN-γ may also enhance hBD-2 production at posttranscriptional levels; they may enhance the stability of hBD-2 mRNA and/or protein (8, 23, 24).

Fig. 8.

A schematic model for the pathways in TNF-α- (A) and IFN-γ-plus-histamine-induced hBD-2 expression (B). Black, white, and gray arrows indicate histamine-induced, TNF-α- or IFN-γ-induced, and common pathways, respectively. H1R, histamine 1 receptor; TNFR, TNF receptor; IFN-γR, IFN-γ receptor.

We (21) and Matsubara et al. (29) previously found that histamine binds H1 receptor on keratinocytes and thus stimulates PKCα, which activates both c-Raf/MEK/ERK and IKK pathways, leading to the activation of AP-1 and NF-κB, respectively. These signaling cascades result in gene expression that is dependent on AP-1 and/or NF-κB, such as GM-CSF or IL-8 genes (21, 29). Similar mechanisms may work in the histamine-induced enhancement of hBD-2 production in synergy with TNF-α or IFN-γ. Histamine-activated PKCα may directly or indirectly activate IKK, which catalyzes serine phosphorylation of IκBα (21, 29), and may lead to the activation of NF-κB in concert with TNF-α or IFN-γ.

Histamine induced c-Fos expression (Fig. 6B). That effect of histamine may be mediated by H1 receptor-mediated PKCα/c-Raf/MEK/ERK pathway. It is reported that ERK-mediated phosphorylation of ternary complex factors such as Elk1 or SAP1 promotes c-fos transcription by these transcription factors (39, 46). The constitutive AP-1 activity required c-Jun but not c-Fos, whereas histamine-induced AP-1 activity required both c-Jun and c-Fos (Fig. 4B). The results indicate that constitutive AP-1 activity may be mediated by c-Jun homodimers, while histamine-induced AP-1 activity may be mediated by c-Fos/c-Jun heterodimers. It is reported that c-Fos/c-Jun heterodimers much more avidly bind DNA and have much higher transcriptional activity than c-Jun homodimers (2, 6). The c-Fos induced by histamine and/or TNF-α and c-Jun induced by TNF-α may dimerize and bind the AP-1 element on the hBD-2 gene and induce the transcription.

In IFN-γ-plus-histamine-stimulated keratinocytes, STAT1 is activated and required for hBD-2 induction in addition to AP-1 and NF-κB (Fig. 4D). Histamine did not per se induce tyrosine phosphorylation but enhanced serine phosphorylation of STAT1 (Fig. 7, A and B) and, as a result, potentiated STAT1 activity in synergy with IFN-γ (Fig. 4C). The histamine-induced serine phosphorylation of STAT1 may be mediated by ERK or its downstream kinase(s), whereas that by IFN-γ may involve p38 MAPK (Fig. 7A). Kinases that serine phosphorylate STAT1 may thus differ with types of stimuli. However, serine phosphorylation without tyrosine phosphorylation of STAT1 may be insufficient for transcriptional induction, since histamine alone did not increase its transcriptional activity (Fig. 4C). Recent studies suggest that histamine induces tyrosine phosphorylation of STAT1 in murine splenocytes (42). In contrast, histamine is reported to suppress tyrosine phosphorylation of STAT1 via H4 receptors in human T cells (15). According to our present results, however, histamine may not affect STAT1 tyrosine phosphorylation, at least in vitro in human keratinocytes, indicating cell type-dependent effects of histamine.

Previous studies revealed that histamine acts on keratinocytes and induces the production of GM-CSF or IL-8 (21, 29), which enhances opsonophagocytosis of macrophages (5) or chemotaxis of neutrophils (27), respectively, both contributing to pathogen killing. In addition to these previous data, our present study demonstrates the histamine-induced hBD-2 production, a novel function linking cutaneous innate and adaptive immunity. In cutaneous infection, inflammatory skin diseases like psoriasis, and wounds, hBD-2 production of epidermal keratinocytes is enhanced (9, 34). In these situations, mast cells or neutrophils are accumulated, and histamine release from these cells is also enhanced (1, 36, 45). Thus our present results indicate that histamine may be one candidate stimulus for hBD-2 production in these situations and may contribute to anti-microbial defense, inflammatory dermatoses, and wound repair via hBD-2. The released hBD-2 may form lytic pores on the lipid bilayers of gram-negative bacteria and kill them or inhibit their growth (10). hBD-2 released from keratinocytes may in turn act on keratinocytes (34) and induce their migration, proliferation, and production of proinflammatory cytokines/chemokines in an autocrine/paracrine manner. Moreover, the released hBD-2 may induce histamine release from mast cells (33) or chemotaxis of TNF-α-activated neutrophils (32), another source of histamine, indicating positive-feedback control. Such a paracrine loop of hBD-2/histamine may amplify cross talk between keratinocytes and mast cells or neutrophils in cutaneous infection, inflammatory dermatoses, and wounds.

GRANTS

This work was supported in part by a grant from Japan Society for the Promotion of Science (grant no. 18244120).

Acknowledgments

We are grateful to Hiroko Sato for the maintenance of keratinocytes.

Footnotes

  • The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

REFERENCES

View Abstract