Critical Role of Histamine H₄ Receptor in Leukotriene B₄ Production and Mast Cell-Dependent Neutrophil Recruitment Induced by Zymosan in Vivo

Histamine exerts its numerous physiological and pathophysiological functions by acting on at least four different receptors, designated H₁ through H₄, which all belong to the heptahelical G-protein-coupled receptor family (Hough, 2001). Although the specific functions of H₁ through H₃ receptor are well characterized (reviewed in Bakker et al., 2002; Repka-Ramirez and Baraniuk, 2002; Winbery and Lieberman, 2002; and Barocelli and Ballabeni, 2003), the role of H₄ receptor remains largely obscure. Cloned and characterized by various groups between 2000 and 2002 (Oda et al., 2000; Nakamura et al., 2000; Liu et al., 2001; Morse et al., 2001; Nguyen et al., 2001; Zhu et al., 2001; O'Reilly et al., 2002), its predominant expression on leukocyte populations suggested potential immunological and inflammatory functions. Indeed, some inflammatory functions mediated by H₄ receptor have recently been identified in vitro. Chemotaxis of human eosinophils (O'Reilly et al., 2002), IL-16 release from human CD8⁺ lymphocytes (Gantner et al., 2002), and calcium mobilization and chemotaxis of murine mast cells (Hofstra et al., 2003) are controlled by H₄ receptor.

Considerable expression of H₄ is also seen on lung tissue and lung cells (Coge et al., 2001; Morse et al., 2001; Zhu et al., 2001; Gantner et al., 2002), which makes H₄ receptor an attractive target for inflammatory lung disorders. Of particular interest is allergic asthma, where a causal role of mast cells, eosinophils, and lymphocytes is implied. Histamine itself, a long-known mediator of acute allergic reactions, has a variety of airway actions potentially relevant to the pathophysiology of asthma, such as airway smooth muscle contraction, vasodilation, the regulation of neurotransmitter release from airway nerve endings, and mucus production (Barnes et al., 1998). In addition, a plethora of different pro- and anti-inflammatory effects on immune cells are known for histamine (Jutel et al., 2002; Schneider et al., 2002), so its overall effect under chronic inflammatory conditions is hard to predict. However, recent evidence from mice carrying a disrupted gene for l-histidine carboxylase, thereby having much lower histamine tissue levels, suggests that the net effect of histamine in the allergic lung response is primarily in mediating eosinophilia but not hyperreactivity of the airways (Koarai et al., 2003). The relative histamine receptor contribution to this outcome, however, is still unknown.

Based on an overall homology of 37 to 43%, the human H₄ receptor is most closely related to H₃ receptor (Hough, 2001). This structural similarity is also reflected pharmacologically, as several compounds initially described as H₃ receptor-selective agonists or antagonists turned out to interact with H₄ receptor in a similar fashion. The apparent expression profile of H₃ and H₄ receptor, however, allows for the use of those pharmacological tools to investigate H₃ and H₄ receptor-mediated functions separately. In an attempt to elucidate the role of H₄ receptor in vivo, we studied two acute models of neutrophilia evoked by the inflammatory stimuli zymosan and carrageenan, respectively, and tested the H_3/4 receptor antagonist thioperamide for its potential modulation of neutrophil influx and mediator release. Experimental evidence for mast cell-driven, H₄ receptor-dependent leukotriene B₄ release and neutrophil recruitment is presented.

Materials and Methods

Chemicals. Histamine, thioperamide, pyrilamine, cimetidine, and zymosan were purchased from Sigma-Aldrich (St. Louis, MO). Carrageenan was purchased from Wako Pure Chemicals (Osaka, Japan). Fetal calf serum was obtained from JRH Biosciences (Lenexa, KS). RPMI 1640 medium was purchased from Invitrogen (Carlsbad, CA). Anti-TLR2 antibody and Ig-matched control antibody were purchased from eBioscience (#16-9021; San Diego, CA). Sodium azide was purchased from Nacalai Tesque (Kyoto, Japan). All bulk chemicals not further specified were purchased from Wako Pure Chemicals.

Mice. BALB/c and WBB6F1-W/W^v mice were purchased from Charles River Japan (Yokohama, Japan). MyD88^–/– mice (Adachi et al., 1998) were generated in the laboratory of S. Akira (Osaka University, Osaka, Japan).

Animals were kept under standard conditions in a 12-h day/night rhythm with free access to food and water ad libitum. All animals received humane care, and the studies were approved by the internal ethics committee in accordance with the guidelines recommended by the Japanese Association of Laboratory Animal Science.

Bone Marrow-Derived Mast Cell Cultures. Bone marrow-derived mast cells (BMMC) were generated from the femural bone marrow of 6- to 8-week old male mice and maintained in RPMI 1640 medium (Nacalai Tesque) supplemented with 10% fetal bovine serum (JRH Biosciences), 100 units/ml penicillin-streptomycin (Invitrogen), and 10 ng/ml recombinant murine IL-3 (PeproTech EC Ltd., London, UK) at a density of 1 × 10⁶ cells/ml. Prior to experimentation, BMMC were harvested, washed twice in PBS, and further used as indicated. Purity of mast cells was >95% as assessed by toluidine blue staining or fluorescence-activated cell sorting analysis of cell surface expression of c-kit and FcϵRI using specific antibodies (BD Biosciences Pharmingen, San Diego, CA and eBioscience).

Zymosan- and Carrageenan-Induced Pleurisy. Mice (BALB/c, Charles River Japan) received a single intrapleural injection with 0.2 ml of sterile saline containing κ-carrageenan (500 μg/mouse) or zymosan (100 μg/mouse) under anesthetization with ether. Test compounds [pyrilamine, cimetidine, thioperamide, dexamethasone, and vehicle (10% cremophor in PBS)] were administered p.o. (0.2 ml/head) 60 min prior to carrageenan injection. Four hours after injection, mice were euthanized, and pleural fluid was collected by washing the pleural cavity twice with 2 ml of PBS. The cell suspension was diluted to one-tenth in Turk's stain solution (Nacalai Tesque). The number of total cells in the sample was counted under the microscope using a hemocytometer. Cytospin specimens were stained with May-Gruenwald's (Merck, Darmstadt, Germany) and Giemsa's solution (Merck) for leukocyte typing. The distribution of each cell population (neutrophils, eosinophils, macrophages, lymphocytes, and others) was counted under microscopy by counting 200 to 300 cells.

Measurement of Soluble Mediators. Pleural fluid was collected by flushing the cavity with 2 ml of PBS. Samples were stored at –80°C until ELISA measurements according to the instructions of the manufacturer (KC, IL-6, TNF-α, IL-1β, IL-12p40, IL-4, interferon-γ, and granulocyte macrophage-colony stimulating factor were obtained from Genzyme-Techne, Minneapolis, MN, and LTB₄ was purchased from R&D systems, Minneapolis, MN).

Statistics. Data are expressed as means ± S.E.M. Statistical significance was determined using the unpaired Student's t test if applicable, or results were analyzed by using one-way ANOVA, and if variances were nonhomogeneous, differences between groups were assessed using Dunnett's method with commercially available statistic software (GraphPad Software Inc., San Diego, CA). P values <0.05 were considered statistically significant (★, P < 0.05; ★★, P < 0.01).

Results

Although zymosan-induced pleurisy is well established in the rat, the peritoneum is much more commonly used in the mouse to evoke a local inflammatory reaction by zymosan. To compare leukocyte recruitment evoked by zymosan and by carrageenan, another widely used trigger of neutrophilia, in the same body compartment, we first established a dose-response curve of zymosan and monitored the leukocyte influx into the pleural cavity of mice. As expected, massive cell infiltrates were found within 4 h after challenge in the pleural cavity, with neutrophils being the dominating population (80–90% of total cells). In parallel, activity of the neutrophil marker enzyme myeloperoxidase was detectable at time points greater than 2 h, peaked at 4 h, and remained elevated 24 h after zymosan injection (data not shown). The residual 10 to 20% of cells consisted of macrophages, eosinophils, mast cells, and a few lymphocytes, and the absolute numbers of these populations were comparable with those obtained from nonchallenged control mice. The dose-response relationship reached a maximum at 100 to 300 μg of zymosan per mouse. Neutrophilia evoked by a zymosan dose of 100 μg/mouse peaked at 4 to 6 h as shown in Fig. 1.

Because mast cells and histamine have been shown to be critical for zymosan-induced inflammation in rodents (Tarayre et al., 1989; Kolaczkowska et al., 2001b; Kolaczkowska, 2002), we next evaluated their contribution to neutrophil influx into the pleural cavity. Mice pretreated with maximally tolerable doses of the histamine H₁ receptor antagonist, pyrilamine, or with the H₂ receptor antagonist, cimetidine, showed no change in the neutrophil influx. In contrast, pretreatment of mice with thioperamide, an H_3/4 antagonist, significantly reduced the inflammatory response in a dose-dependent manner (Fig. 2).

W/W^v mice deficient in mast cells, the major cellular source of histamine, also showed an 80% reduction in neutrophilia upon zymosan challenge (Fig. 3A). Maximum efficacy was similar to that observed in thioperamide-treated mice, which prompted us to speculate that thioperamide might act on mast cell-expressed H₄ receptor (Morse et al., 2001; Gantner et al., 2002; Hofstra et al., 2003). In support of this hypothesis, no further reduction of neutrophilia was seen when thioperamide was given to mast cell-deficient animals (Fig. 3B). Neither mast cell nor H₄ receptor dependence was seen when neutrophil recruitment into the pleural cavity was evoked by carrageenan, i.e., neutrophilia was unchanged in W/W^v mice, and thioperamide pretreatment showed no effect either (Fig. 3, A and C).

To further prove the mast cell dependence of the zymosan-induced reaction, we tested the neutrophil influx in mice lacking MyD88, a pivotal signaling module linked to toll-like receptors, including the mast cell-expressed zymosan receptor, TLR2. In accordance with our findings in mast cell-deficient mice, total cell influx (data not shown) and neutrophilia were dramatically reduced in zymosan-challenged MyD88^–/– mice (Fig. 4A). In addition, a similar reduction of neutrophilia was observed upon pretreatment of mice with an antibody directed against TLR2, whereas an Ig-matched control Ab was without effect (Fig. 4B). In line with our hypothesis, neutrophilia was inducible when W/W^v or MyD88^–/– mice were reconstituted with BMMC from wt or +/+ mice of the respective control strain, evidencing the necessity of zymosan-activated mast cells to recruit neutrophils (Fig. 4, C and D).

In contrast, carrageenan-induced neutrophilia remained unchanged in mice lacking MyD88 (data not shown). Dexamethasone, however, was similarly active in both models and reduced zymosan- and carrageenan-induced neutrophilia to a similar extent in a dose-dependent manner (Fig. 4, E and F).

To investigate the potential link between mast cell- and H₄ receptor-mediated neutrophil recruitment after zymosan stimulation, we looked for potential chemoattractant mediators being released into the pleural cavity of zymosan-challenged mice. Specifically, we collected pleural fluid within 1.5 h after challenge, i.e., before significant neutrophilia was observed. Increasing amounts of a variety of mast cell- or monocyte-derived factors, such as TNF-α, IL-1β, IL-6, IL-10, IL-12, KC, and LTB₄, were found over 90 min, whereas typical T cell cytokines, such as IL-4, IL-5, interferon-γ, or granulocyte macrophage-colony stimulating factor, were not detectable at any time point investigated (Fig. 5 and data not shown).

We next investigated whether H₄ receptor might be involved in zymosan-induced mediator release in vivo and measured mediator concentrations in the pleural cavity of thioperamide-pretreated animals 90 min after challenge. A significant reduction in LTB₄ of about 75% was noted in H₄ receptor antagonist-treated animals compared with vehicle-treated controls. Likewise, LTB₄ levels were similarly reduced in mast cell-deficient mice (Fig. 6). A 20 to 30% reduction was also noted for KC (data not shown), whereas all other mediators remained unchanged upon thioperamide pretreatment.

Discussion

The involvement of histamine and mast cells in the model of zymosan-induced peritonitis is well established. In the present report, we expand these findings to neutrophil recruitment into the pleural cavity and demonstrate that the recently identified receptor for histamine, H₄ receptor, is mediating this response.

Not surprisingly, the inflammatory response triggered by zymosan injection into the pleural cavity resembles the one observed in the peritoneum (Rao et al., 1994; Kolaczkowska et al., 2001a,b; Kolaczkowska, 2002). Neutrophils constitute the vast majority of all leukocytes recruited (cf. Fig. 1), mast cell deficiency diminishes (cf. Fig. 3A), whereas mast cell reconstitution in W/W^v mice restores neutrophilia (Fig. 4C), and a similar set of soluble mediators of inflammation are released within minutes and hours (cf. Fig. 5). Notably, the absolute number of mast cells recovered from the pleural cavity did not change upon zymosan challenge compared with naive animals (ca. 15,000 mast cells/cavity). This low number of mast cells apparently is still enough to control the recruitment of neutrophils upon zymosan challenge.

Caution has to be put on the interpretation of the pharmacological results obtained by the use of histamine receptor antagonists. We present here compelling evidence for H₄ receptor being critical for mast cell-dependent neutrophil recruitment. Thioperamide, a well characterized H₄ antagonist, but not antagonists specifically targeting H₁ or H₂ receptor, significantly reduced zymosan-induced neutrophilia. Although able to potently block H₃ receptors, thioperamide most likely acts on H₄ receptor expressed on mast cells that completely lack, as do most leukocytes studied so far, H₃ receptor. This conclusion is further supported by novel data obtained with the recently described H₃ receptor-inactive, H₄-selective antagonist JNJ 7777120, which showed identical efficacy in zymosan-induced mast cell-dependent neutrophilia in the mouse (oral presentation by Dr. R. Thurmond at the European Histamine Research Society meeting in Leiden, The Netherlands, May 2003; see patent WO 02072548 for details).

In contrast to our findings, partial inhibition of neutrophilia was reported in the zymosan-peritonitis model when mice were pretreated with H₁ receptor antagonists, such as mepyramine (Kolaczkowska et al., 2001b). This effect, however, was only seen at an early time point (2 h after challenge) but not 6 h after challenge, when neutrophilia was maximal, or thereafter. The major effect of H₁ receptor antagonists was seen on vascular permeability, a histamine effect known to be mediated primarily by H₁ receptor (reviewed in Hill et al., 1997), and thus, the diminished cell counts found in the peritoneum might be a consequence of such a response. Like thioperamide, mepyramine pretreatment had no effect on the levels of TNF-α excluding reduced TNF levels as the mechanism leading to reduced neutrophilia after blockade of histamine action. In contrast, H₄ receptor antagonism clearly diminished the levels of LTB₄, one of the most potent chemoattractants known for neutrophils. This result fits well to the protective effect of leukotriene synthesis inhibitors (Rao et al., 1994) and the diminished zymosan response of mice lacking 5-lipoxygenase (Byrum et al., 1999) or the receptor for LTB4, BLTR (Haribabu et al., 2000). The fact that LTB₄ levels were also reduced in mast cell-deficient animals further supports the hypothesis of mast cell-expressed H₄ receptor as the primary target of thioperamide action. In line with this, we measured mediator release at 90 min after zymosan challenge, i.e., at a time point before neutrophils were recruited to the pleural cavity. This again argues for mast cells being the source of LTB₄ and the target of thioperamide action.

Of course, direct thioperamide effects on neutrophils cannot be fully excluded, although among murine leukocytes, only mast cells, eosinophils, and basophils have been reported to express H₄ receptor (Hofstra et al., 2003). Detailed analyses of H₄ receptor expression in mouse neutrophils isolated from various compartments and under various activation states are currently under investigation in our laboratory. All currently known functions of H₄ receptor are related to the control of leukocyte trafficking: histamine-induced chemotaxis of eosinophils (O'Reilly et al., 2002) and mast cells (Hofstra et al., 2003), the release of IL-16, a chemoattractant of CD4⁺ cells, from CD8⁺ lymphocytes (Gantner et al., 2002), and, finally, the zymosan-induced recruitment of neutrophils in vivo (this study).

Mechanistically, the neutrophil recruitment evoked by the two inflammatory stimuli, zymosan and carrageenan, fundamentally differs. Although both responses are sensitive to dexamethasone pretreatment (Fig. 4), the inflammatory response evoked by carrageenan obviously does not depend on H₄ receptor or mast cells (cf. Fig. 3, A and C), which is line with findings by others (Horakova et al., 1980; Lo et al., 1982; Damas and Remacle-Volon, 1986). Furthermore, preliminary studies reveal that the neutrophil populations infiltrating the pleural cavity seem to be recruited from different body compartments (data not shown). A more detailed analysis of the similarities and disparities underlying these two models is currently under investigation in our laboratory (unpublished data) and thus shall not be further discussed here.

Collectively, we describe here a role for the novel histamine H₄ receptor in vivo. We propose a model in which histamine, released from mast cells after TLR2-triggered MyD88 activation by zymosan, acts in an autocrine fashion on mast cell-expressed H₄ receptor, thereby leading to the release of chemoattractants for neutrophils among which LTB₄ has been identified as a likely candidate for mediating cell recruitment. The model of zymosan-induced pleurisy is a suitable model to selectively study the efficacy of H₄ receptor antagonists that might become future therapeutics to treat mast cell- and histamine-dependent disorders, such as allergy and asthma.