Endocrine Disrupting Chemical Atrazine Causes Degranulation through G_q/11 Protein-Coupled Neurosteroid Receptor in Mast Cells

Abstract

We studied the effects of representative endocrine-disrupting chemicals on β-hexosaminidase release from mast cells and their putative neurosteroid receptor involvement. Some endocrine-disrupting chemicals, such as amitrol, benzophenon, bisphenol A, pentachlorophenol, and tetrabromophenol A did not cause hexosaminidase release from RBL-2H3 cells, but they blocked the release by dehydroepiandrosterone sulfate, a representative neurosteroid agonist. On the contrary, atrazine, which is a widely used herbicide, caused a rapid and concentration-dependent degranulation in the range between 10 nM and 1 μM in RBL-2H3 and peritoneal mast cells. Atrazine-induced degranulation was also evaluated by Alexa 488-annexin V binding to the phosphatidylserine, which is externalized during degranulation, and these actions were blocked by BSA-conjugated (membrane-impermeable) progesterone (PROG-BSA). The atrazine-induced β-hexosaminidase release was characterized by various inhibitors including antisense-oligodeoxynucleotide for Gα_q/11, pertussis toxin, phospholipase C inhibitor U-73122, inositol 1,4,5-triphosphate receptor inhibitor xestospongin C and Ca²⁺ channel blocker lanthanum chloride. These analyses revealed that the degranulation is mediated by putative metabotropic neurosteroid receptor, G_q/11, phospholipase C and Ca²⁺ mobilization from intracellular stores. Having documented progesterone receptor-modulation of atrazine-induced mast cell degranulation in vitro, this response was evaluated in mice. Atrazine caused pain responses when injected in the foot pads of mice, and they were antagonized by local administration of PROG-BSA or diphenhydramine. Atrazine also caused PROG-BSA-reversible plasma extravasation. All these findings strongly suggest that herbicide atrazine exerts inflammatory activity through activation of putative G_q/11-coupled neurosteroid receptor and phospholipase C.

degranulation, endocrine-disrupting chemical, extravasation, mast cell, neurosteroid

Endocrine disrupting chemicals (EDCs) are known to mimic the actions of sex steroid hormones and to cause serious dysfunction of reproductive organs through disruption of endocrine homeostasis (McLachlan, 2001). These actions are largely attributed to genomic actions through nuclear steroid receptors. However, some other biological actions of endocrine disrupting chemicals include non-genomic and acute responses (Chen, 2001). These are thought to be mediated through the central nervous system, therefore, several studies have been carried out on the psychoactive actions of neurosteroids (Dohi et al., 2004).

Neurosteroids are synthesized de novo in the nervous system and have a wide variety of functions (Baulieu, 1989; Rupprecht and Holsboer, 1999). They exert their actions not only through classical nuclear steroid hormone receptors, but also through plasma membrane receptors (Baulieu, 1989; Nadal et al., 2001). Although many studies have demonstrated that the non-genomic actions of neurosteroids are due to allosteric actions on ligand-gated channels such as GABA_A (Majewska et al., 1986) and NMDA receptors (Wu et al., 1991), we have previously shown in reconstitution experiments that some neurosteroids share pharmacological activity with a putative G_i-coupled σ receptor in the brain (Ueda et al., 2001b). Recently, we have demonstrated that pregnenolone sulfate and dehydroepiandrosterone sulfate (DHEAS), two representative neurosteroids, induce nociceptive flexor responses through two novel types of neurosteroid receptors, termed neuronal NS1/σ-type and non-neuronal NS2-type (Ueda et al., 2001a). In that study we proposed that the latter NS2-type neurosteroid receptor exists on mast cells and mediates histamine release, which, in turn, stimulates nociceptor endings. Other studies have demonstrated that neurosteroids cause extravasation as well as hyperalgesia, which were abolished by diphenhydramine, a representative H1 receptor antagonist (Uchida et al., 2003). It is of great significance that 1,1-dichloro-2,2-bis (p-chlorophenyl) ethylene (p,p′-DDE), which is a metabolite of dichlorodiphenyltrichloroethane (DDT) and listed as an EDC, blocks these actions by acting as a pure antagonist. More recently, we found that neurosteroid-induced extravasation are attributed to degranulation of mast cells (Uchida et al., 2003). Although representative EDCs, nonylphenol and 4-octylphenol showed no extravasation by themselves, they have antagonist effects. Therefore, as far as acute inflammatory effects are concerned, these EDCs without extravasation effects may cause no serious ecological problems.

Atrazine, an herbicide, is still commonly used in the U.S. (US EPA, 2003), despite the report that atrazine has disruptive effects on the reproductive system in wild animals and humans (Hayes et al., 2002). It should be noted that up to 92% of male leopard frogs show gonadal abnormalities, such as retarded development and hermaphroiditism, on exposure to water-borne atrazine contamination. In the course of extensive study to evaluate the effects of more EDCs on the mast cell degranulation, we found that atrazine causes potent degranulation and related inflammatory actions, which may further warn the ecological problems. Here we focused our concerns on the mechanisms underlying these acute and non-genomic actions.

MATERIALS AND METHODS

Chemicals.

The following chemicals were obtained from Sigma (St. Louis, MO): Compound 48/80, dehydroepiandrosterone sulfate (DHEAS), progesterone (PROG), PROG 3-(O-carboxymethyl) oxime:BSA (PROG-BSA) (steroid:BSA, 38:1), PROG-BSA-fluorescein isothiocyanate conjugate (PROG-BSA-FITC), lanthanum chloride, N-acetyl-β-D-glucosaminide, quercetin and RU486. ICI-182,780, Evans blue were purchased from Wako (Osaka, Japan). Atrazine was purchased from Kanto Kagaku (Tokyo, Japan). Amitrol, bisphenol, benzophenone, pentachlorophenol, and tributylphenol A were gifts from Dr. Funae (Osaka City University, Japan). Xestospongin C, which is also called araguspongine E (Kobayashi et al., 1989) was a gift from Dr. Kobayashi (Osaka Univ., Japan). Alexa 488 annexin-V-fluorescein reagent was obtained from Funakoshi (Tokyo, Japan). The antisense oligodeoxynucleotide (AS-ODN, 5′-ATGGACTCCAGAGT-3′) for rat Gα_q/11 and its mismatch oligodeoxynucleotide (MS-ODN, 5′-AGTGACCTCAGGAT-3′) were synthesized as described previously (Ueda and Inoue, 2000). All oligodeoxynucleotides (ODNs) were purchased from QIAGEN (Tokyo, Japan). Anti-Gα_q/11 was obtained from NEN Life Science Products (Boston, MA).

Cell culture.

RBL-2H3 cells were maintained in minimum essential medium (MEM) supplemented with 10% fetal bovine serum, L-glutamine and penicillin/streptomycin. In most experiments, unless otherwise stated, RBL-2H3 cells were treated with 30 μM quercetin for 48 h. Quercetin, prepared as a 60 mM stock solution in propylene glycol was added to up-regulate G_i, since this treatment makes it easier to detect the β-hexosaminidase release by G_i-mediated mechanisms, as seen with Compound 48/80, known to degranulate mast cells (Senyshyn et al., 1998). Although the deletion of quercetin from the culture did not change the DHEAS-induced β-hexosaminidase release (Supplementary Data), the present study was carried out in the presence of quercetin in order to compare the signal transduction with the previous report (Mizota et al., 2005).

For each experiment, the cells were incubated overnight in complete growth medium and then with pertussis toxin (PTX; 100 ng ml⁻¹) for 12 h, or U-73122 (100 nM), U-73343 (100 nM), xestospongin C (1 μM), lanthanum chloride (50 μM), wortmannin (1 μM), RU-486 (1 μM), ICI-182,780 (1 μM) for 10 min before addition of the test drugs. Test drugs were dissolved in 100% methanol to 0.02 M, and diluted to make 1 μM in HEPES-Tyrode buffer (140 mM NaCl, 2.7 mM KCl, 1.8 mM CaCl₂, 5.6 mM glucose, 12 mM NaHCO₃, 0.37 mM NaH₂PO₄ · 2H₂O, 25 mM HEPES, 0.49 mM MgCl₂, 0.1% BSA, pH 7.4 adjusted with NaOH). The antisense-oligodeoxynucleotide (AS-ODN) or its mismatch-oligodeoxynucleotide (MS-ODN) for Gα_q/11 (20 μM) was added to the culture of RBL-2H3 cells every 8 h for 48 h, followed by wash with serum-free MEM and with HEPES-Tyrode BSA assay buffer. The treated cells (5 × 10⁴ cells/well) were suspended in 200 μl of HEPES-Tyrode BSA buffer and used for the β-hexosaminidase release assay. Western blot analysis was performed as reported previously (Ueda and Inoue, 2000), using the antiserum against Gα_q/11 (1:1000 dilution).

Peritoneal mast cell isolation.

The experiment was performed as previously described (Fish et al., 2005; Mori et al., 2000). Male ddY mice (5 weeks) were euthanized by CO₂-asphyxia. Normal Hepes-buffered Ringer solution (HR; 118mM NaCl, 4.7 mM KCl, 2.5mM CaCl₂, 1.13mM MgCl₂, 1.0 mM NaH₂PO₄, 10 mM D-glucose, 10 mM HEPES, MEM, 0.1% bovine serum albumin; pH 7.4, adjusted with NaOH; approximately 10 ml) was injected intraperitoneally and the abdomen was gently massaged for approximately 5 min. The HR-buffer from the peritoneal cavity was centrifuged at 2000 rpm for 5 min, and the eluted peritoneal cells containing the mast cells were harvested. After the addition of several ml of fresh HR-buffer, the cell suspensions were plated, grown to confluence, in 24-well plates coated with lysine-collagen.

Measurement of β-hexosaminidase release.

The experiment was performed as previously described (Hong-Geller and Cerione, 2000; Zussman and Sagi-Eisenberg, 2000). RBL-2H3 or peritoneal mast cells cells were seeded in 24-well plates, grown to confluence, incubated in growth medium containing 30 μM quercetin and washed twice with HEPES-Tyrode BSA buffer, and suspended in 200 μl of HEPES-Tyrode BSA buffer at 37°C. The cells were then stimulated with various drugs for 10 min or the indicated periods at 37°C and harvested by spin-down. Hundred μl aliquots of the supernatants were collected and incubated with 50 μl 10 mM N-acetyl-β-D-glucosaminide in 0.05 M citrate buffer (pH 4.5) for 1 h at 37°C. To determine the total amount of β-hexosaminidase, the cells were lysed with 0.1% Triton X-100. The absorbance at 405 nm was read in a microplate reader (Bio-Rad 550; Nippon Bio-Rad Laboratories, Osaka, Japan).

Alexa 488-Annexin-V staining assay.

The experiment was performed as previously described (Demo et al., 1999; Windmiller and Backer, 2003). To study mast cell degranulation in single cells, RBL-2H3 cells (5.0 × 10⁴ cells well⁻¹) grown on lysine–collagen coated coverslips were cultured for 24 h, washed in HEPES-Tyrode BSA buffer and stimulated by test drug for 10 min at 37°C in the presence of 1:10 dilution of Alexa 488-labeled annexin-V, which binds to phosphatidylserine in the outer leaflet of the plasma membrane of degranulated mast cells. The cells were washed with phosphate-buffered saline (PBS), fixed in 4% (w/v) paraformaldehyde for 30 min at room temperature, washed three times with PBS and then mounted. The Alexa 488-annexin V positive cells were counted using a fluorescence microscope (Axiovert 135, Zeiss, Germany). Images were acquired with NIH image 1.63 analysis software. The Alexa 488-annexin V-based assay accurately reflects ligand-stimulated degranulation of RBL-2H3 cells.

PROG-BSA-FITC binding assay.

The experiment was performed as previously described (Benten et al., 1999; Nadal et al., 2000). RBL-2H3 cells (1 × 10⁵ cells well⁻¹) cultured on polylysine-coated coverslips for 24 h were fixed with 2% (w/v) paraformaldehyde for 15 min, and washed three times with PBS. The cells were incubated with 1 μM PROG-BSA-FITC at room temperature for 2 h and washed. Endocrine disrupting chemicals were added to the cell for 30 min before the start of the PROG-BSA-FITC binding. FITC-labeled cells were analyzed using a confocal laser scanning microscope (Axiovert 135, Zeiss, Germany), after excitation of FITC fluorescence by a 488 nm argon laser.

Animals and behavioral studies.

The pain-producing substance-induced biting and licking (pain-like behavior) test was performed as described previously (Uchida et al., 2003). Male ddY mice weighing 20–22 g were used in all experiments. The experimental procedures were approved by the Nagasaki University Animal Care Committee and complied with the recommendations of the International Association for the Study of Pain (Zimmermann, 1983). Test drugs were administered intraplantarly into the hind paw. Mice were adapted to individual transparent plastic cages, which served as observation chambers, for 1 h prior to intraplantar (i.pl.) injection. Immediately after injection, each mouse was re-placed in the cage over a mirror and behavioral testing was initiated. The total amount of time showing pain responses, such as biting and licking, was measured for 10 min after i.pl. injection. All drugs were first dissolved in 100% methanol to make 20 mM and diluted to make 10 fmol in 20 μl physiological saline. As the dilution is so high, no significant in vivo effect of methanol was expected. All drugs were dissolved in physiological saline, and administered by i.pl. injection in a volume of 20 μl. All mice were only used once.

Evans blue extravasation test.

Evans blue extravasation assay was performed as described previously (Uchida et al., 2003). Mice were anaesthetized with pentobarbital (50 mg kg⁻¹, ip) and injected with Evans blue (50 mg kg⁻¹, iv) into the tail vein. Antagonist PROG-BSA, diphenhydramine (DPH) and atrazine were administered by i.pl. injection 10 and 15 min later, respectively. Twenty-five min after atrazine injection, the plantar skin of the paw was removed, dried of excess liquid, weighed (0.03–0.05 g wet weight) and incubated in formamide for 24 h at 55°C. Extravasated Evans blue was measured by spectrophotometry at 620 nm. Results are expressed as the ratio of arbitrary fluorescence units for Evans blue of drug-treated paw to that of vehicle-treated paw.

Statistical analysis.

Results of all experiments were expressed as the mean ± SE and analyzed by one-way ANOVA with Scheffe's test. Significance was set at p < 0.05.

RESULTS

Atrazine-Induced Rapid Degranulation in RBL-2H3 Cells

Degranulation of mast cells was determined by measuring the release of granule-associated β-hexosaminidase from RBL-2H3 cells, as previously reported (Mizota et al., 2005), in which dehydroepiandrosterone sulfate (DHEAS)-induced β-hexosaminidase release was inhibited by co-administration of 10 μM nonylphenol, 4-octylphenol, or PROG-BSA (10 μM). As shown in Figure 1A, DHEAS and atrazine at 1 μM showed significant degranulation, while other EDCs, such as amitrol, benzophenon, bisphenol A, pentachlorophenol, and tetrabromophenol A did not. However, these EDCs showed potent antagonist activity against DHEAS- or atrazine-induced degranulation (Figs. 1B and 1C). As shown in Figures 1D and 1E, atrazine-induced β-hexosaminidase release was dose-dependent between 10 nM and 1 μM, and significant change was observed as early as 10 min after the addition. As seen in the case with DHEAS, atrazine-induced β-hexosaminidase release was blocked by PROG-BSA (10 μM), a membrane impermeable neurosteroid receptor antagonist (Fig. 1F). However no EDCs caused any cytotoxic effects on the cells, as evaluated by the Trypan blue exclusion assay (data not shown). Similar atrazine-induced degranulation and its PROG-BSA reversibility were observed in peritoneal mast cells (Figs. 1G and 1H). DHEAS and atrazine at 1 μM caused significant release by 4.6 and 3.8% of total contents, respectively from these cells.

FIG. 1.

Atrazine-induced β-hexosaminidase release through a putative membrane receptor in mast cells. Experiments for β-hexosaminidase release were performed with RBL-2H3 cells (A–F) or peritoneal mast cells (G, H). Test drugs were used at 1 μM unless otherwise stated. (A) Significant increase in β-hexosaminidase release by 1 μM dehydroepiandrosterone sulfate (DHEAS) and 1 μM atrazine (ATZ), but not by 10 μM amitrol (AMI), benzophenone (BP), bisphenol A (BPA), pentachlorophenol (PCP), or tetrabromophenol A (TBPA). (B, C) Antagonisms of DHEAS- or atrazine-induced β-hexosaminidase release by various EDCs. (D, E) Concentration-dependency and time-course of atrazine-effects. (F) Antagonism of atrazine-effects by membrane impermeable neurosteroid receptor antagonist progesterone-BSA (PROG-BSA) of 10 μM. (G) Concentration-dependency of atrazine-effects in peritoneal mast cells. (H) 10 μM PROG-BSA-reversibility of 1 μM DHEAS- and 1 μM atrazine-effects. Data represent the mean ± SE from more than three separate experiments. *p < 0.05, compared with the vehicle (Veh)-treated group.

Single-Cell Assay for Mast Cell Degranulation

It is reported that degranulation is accompanied by externalization of outer surface membrane phosphatidylserine, which is labeled by Alexa 488-labeled annexin-V (Demo et al., 1999; Windmiller and Backer, 2003). The addition of Alexa 488-annexin V alone to RBL-2H3 cells showed no significant fluorescence, whereas further addition of atrazine (10 μM) caused some dotted fluorescent signals in most of cells (Fig. 2A). As shown in Figure 2B, similar membrane externalization was also observed with DHEAS (10 μM), but not PROG-BSA, while PROG-BSA abolished both atrazine- and DHEAS-mediated binding (Fig. 2B).

FIG. 2.

Atrazine-induced increase in Alexa 488-annexin V binding in RBL-2H3 cells. Test drugs were used at 10 μM. (A) Representative pictures of atrazine-induced Alexa 488-annexin V binding. Dotted fluorescent signals show the phosphatidylserine externalization, representing degranulation. (B) Antagonism of DHEAS- and atrazine-induced Alexa 488-annexin V binding by PROG-BSA (100 μM). Cells showing vesicle-type fluorescent signal are evaluated as positive ones. Results were represented as percentage of positive cells to total cells. Data are the mean ± SE from three separate experiments. *p < 0.05, compared with Veh, #p < 0.05, compared with DHEAS or atrazine alone.

Atrazine-Reversible PROG-BSA-FITC Binding

This experiment was determined as previously reported (Mizota et al., 2005). RBL-2H3 cells cultured on polylysine-coated coverslips were fixed with paraformaldehyde, washed three times with PBS and used for binding experiments with 1 μM PROG-BSA-FITC. FITC-labeled cells were analyzed using a confocal laser scanning microscope. Fluorescence was only observed at the cell surface, and the addition of high concentration atrazine (10 μM) completely abolished the binding, as shown in Figure 3. Similar inhibition was observed with 10 μM pentachlorophenol, which showed the strongest inhibition of DHEAS-induced degranulation (Fig. 3). Amitrol, benzophenon, also blocked the binding (data not shown).

FIG. 3.

Atrazine- or pentachlorophenol-reversible PROG-BSA-FITC binding. PROG-BSA-FITC (1 μM) binding experiments to RBL-2H3 cells were carried out in the absence or presence of atrazine (10 μM) or pentachlorophenol (10 μM).

Signal Transduction of Atrazine-Induced β-Hexosaminidase Release

Atrazine-induced β-hexosaminidase release was characterized by the use of various inhibitors. Pretreatment with pertussis toxin (100 ng ml⁻¹), which inhibits compound 48/80 (10 μg ml⁻¹)-induced degranulation (Mizota et al., 2005), had no effect (Fig. 4A). In contrast, pretreatment with AS-ODN for Gα_q/11, but not its MS-ODN, markedly reduced atrazine-induced β-hexosaminidase release (Fig. 4B). As shown in the inset of Fig. 4B, AS-ODN significantly reduced the amount of Gα_q/11 protein in RBL-2H3 cells. Atrazine-induced degranulation was significantly abolished by phospholipase C (PLC) inhibitor U-73122 at 100 nM, but not by its inactive derivative U-73343 at 100 nM (Fig. 4A). Although xestospongin C at 1 μM, an allosteric inositol trisphosphate receptor antagonist (Gafni et al., 1997) also blocked the release, lanthanum chloride at 50 μM, which inhibits Ca²⁺ influx in cells (Aussel et al., 1996), did not. These results suggest that Ca²⁺ mobilization from intracellular stores, but not Ca²⁺ influx, is involved in this mechanism. On the other hand, neither nuclear PROG receptor antagonist RU486 (1 μM) nor estradiol receptor antagonist ICI-182,780 (1 μM) (Baulieu, 1989; Howell et al., 2000) caused any significant reduction in atrazine-induced degranulation. Therefore, known genomic effects are unlikely involved in this degranulation.

FIG. 4.

G_q/11 and PLC-mediated β-hexosaminidase release by atrazine. Concentrations of inhibitors were stated in the text. (A) Effects of various inhibitors on atrazine (1μM)-induced β-hexosaminidase release from RBL-2H3 cells. (B) Effects of AS-ODN for Gα_q/11. Inset: significant reduction of Gα_q/11 levels in Western blot analysis (1.5 μg protein per lane, 12% acrylamide gel) after treatments with AS- or MS-ODN. Data represent the mean ± SE from four to five separate experiments. *p < 0.05, compared with Veh.

Atrazine-Induced Rapid Inflammatory Effects

Previously, we have reported that intraplantarly administered neurosteroids cause pain-like responses, such as biting and licking behaviors, in a diphenhydramine-reversible manner (Mizota et al., 2005). In these studies, DHEAS (1–30 fmol, i.pl.) induced pain behaviors for 5–25 s in 10 min observation. As shown in Figure 5A, atrazine caused similar acute pain responses at a dose of 3 fmol per 20 μl, while other EDCs (amitrol, benzophenon, bisphenol A, pentachlorophenol, and tetrabromophenol A) alone caused no significant responses, but showed significant blockade of DHEAS- or atrazine-effects (Figs. 5B and 5C). Atrazine-induced pain responses occurred immediately and ceased within 5 min after the administration (Fig. 5D). The responses were dose-dependent between 1 and 10 fmol (i.pl.), blocked by PROG-BSA or DPH (Figs. 5E–5G).

FIG. 5.

Atrazine-induced pain responses. (A) Pain responses by DHEAS and atrazine, but not by other EDCs. Test drugs were used at 3 fmol per 20 μl (i.pl.). (B, C) Antagonisms of DHEAS (10 fmol, i.pl.)- or atrazine (10 fmol, i.pl.)-induced pain responses by various EDCs (3 fmol, i.pl.) (B) and antagonism of DHEAS (10 fmol)-induced responses by various endocrine disrupting chemicals (3 fmol). (D) Time course of atrazine (3 fmol, i.pl.)-induced pain responses. (E–G) Dose-dependency, PROG-BSA- and diphenhydramine (DPH)-reversibility of atrazine-induced pain responses. DPH was used at 30 fmol (i.pl.). Results represent the mean ± SE of the total period(s) showing pain-like biting and licking behaviors from four to five separate experiments. *p < 0.05, compared with Veh, #p < 0.05, compared with the group without each corresponding antagonist.

Atrazine (10 fmol) increased the vascular permeability in terms of blue color at the injection site of the paw skin, when observed 10 min after systemic injection of Evans blue (Fig. 6A). This atrazine-induced vascular extravasation was abolished by PROG-BSA or DPH, which have no effects themselves in Evans blue extravasation.

FIG. 6.

Atrazine-induced plasma extravasation. (A) Representative result of atrazine (10 fmol)-induced extravasation of Evans blue dye. (B) Antagonism of atrazine-induced extravasation by PROG-BSA or DPH. Atrazine, PROG-BSA, and DPH were used at 10, 100, and 100 fmol (i.pl.), respectively. Extravasation of Evans blue dye by atrazine-treatment was evaluated as the percentage of Veh-treatment. The Evans blue dye amount in the paw skin of Veh-treated mice was 0.2 ± 0.1 mg per g of dried skin (n = 5). Results represent the mean ± SE from more than six separate experiments. *p < 0.05, compared with Veh, #p < 0.05, compared with atrazine alone.

DISCUSSION

Our previous study demonstrated that some neurosteroids, such as pregnenolone sulfate and DHEAS cause degranulation in RBL-2H3 cells through putative G_q/11-coupled neurosteroid receptors, while PROG or its BSA-conjugate (PROG-BSA) block these actions (Mizota et al., 2005). It should be noted that nonylphenol and bisphenol A have antagonist effects. The fact that EDCs have antagonist effects on cell surface neurosteroid receptors is also supported by Uchida et al. (2003), who showed that p,p'-DDE, a metabolite of DDT, antagonizes neurosteroid-induced hyperalgesia and extravasation. Similarly in the present study, we observed that some other EDCs such as amitrol, pentachlorophenol, benzophenon, bisphenol A, or tetrabromophenol A did not cause degranulation in terms of β-hexosaminidase release from RBL-2H3 cells, but abolished the DHEAS-induced actions.

However, we found an exception, in which atrazine, a herbicide EDC caused degranulation and related inflammatory actions. Atrazine-induced β-hexosaminidase release was blocked by membrane-impermeable BSA-conjugate PROG-BSA (Fig. 1F). Thus, it is evident that atrazine effects are mediated through putative membrane receptors. The atrazine-induced degranulation is unlikely due to cell damage, since it is also observed in a different assay using Alexa 488-annexin V binding to label phosphatidylserine externalized in the plasma membrane during degranulation (Demo et al., 1999; Smith et al., 2001). This atrazine-induced degranulation seems to be mediated through specific membrane receptors, since it was also reversible by PROG-BSA, but not by nuclear PROG or E₂ receptor antagonists. Indeed, pharmacological characterization using specific inhibitors revealed that the signal transduction involved in this degranulation is mediated through G_q/11, PLC and mobilization of intracellular Ca²⁺, and all these characteristics are identical to those with neurosteroids. As previously reported (Mizota et al., 2005), neurosteroid-induced degranulation level was only 1–2% of the total content, which is comparable with atrazine-induced one, but much less than that through IgE receptors (10–15%). As a little more release by 1 μM DHEAS (4.6%) or atrazine (3.8%) was observed in peritoneal mast cells, however, such weak potencies in RBL-2H3 cells may be attributed to the low density of neurosteroid receptors.

Atrazine-induced degranulation was also supported by the in vivo studies (Figs. 5 and 6). Atrazine-induced pain responses were observed within as early as 5 min after the administration. As these effects were abolished by PROG-BSA and DPH, which is a histamine receptor antagonist, it is evident that they are caused by the neurosteroid receptor-mediated histamine release from mast cells. This view was supported by the atrazine-induced plasma extravasation assay using Evans blue. Throughout the present study, it is evident that atrazine also drives neurosteroid receptor-mediated mechanisms and causes some inflammatory effects, pathophysiological roles, in vivo. It is an important issue to be discussed that in vivo DHEAS or atrazine doses (10 fmol/20 μl, equivalent to 0.5 nM) required for induction of inflammatory actions are less than the in vitro concentrations (more than 10 nM) required for the degranulation. Although there is no clear answer, it might be possible that the mast cells existing in foot pad space may be more sensitive than those in RBL-2H3 or peritoneal mast cells. Alternatively, the histamine release required for induction of pain responses may be enough even if the mast cell degranulation is only partial. As reported in our previous study, such behavioral changes are related to the local amplification mechanism (Inoue et al., 1998). Taken this hypothesis, the initial weak histamine release would be enough to generate forthcoming bigger pain fiber stimulations.

In conclusion, we demonstrate that the herbicide atrazine induces some inflammatory effects, including nociception and plasma extravasation, and the underlying mechanisms are mediated through mast cell degranulation and activation of the G_q/11-protein coupled neurosteroid receptor, PLC and intracellular Ca²⁺ mobilization.

We thank Prof. Motomasa Kobayashi for providing xestospongin C. Parts of this study were supported by Health Sciences Research Grants for Research on Environmental Health from the Ministry of Health, Labor and Welfare of Japan, Grants for the Nagasaki Prefecture Collaboration of Regional Entities for the Advancement of Technological Excellence, JST and Grants-in-Aid from the Ministry of Education, Science, Culture and Sports of Japan. This study was also supported by grants from the CREST Project of JST for Endocrine Disruption on Action of Brain Neurosteroids.