Hesperetin-7,3'-O-dimethylether selectively inhibits phosphodiesterase 4 and effectively suppresses ovalbumin-induced airway hyperresponsiveness with a high therapeutic ratio

Background Hesperetin was reported to selectively inhibit phosphodiesterase 4 (PDE4). While hesperetin-7,3'-O-dimethylether (HDME) is a synthetic liposoluble hesperetin. Therefore, we were interested in investigating its selectivity on PDE4 and binding ability on high-affinity rolipram-binding sites (HARBs) in vitro, and its effects on ovalbumin-induced airway hyperresponsiveness in vivo, and clarifying its potential for treating asthma and chronic obstructive pulmonary disease (COPD). Methods PDE1~5 activities were measured using a two-step procedure. The binding of HDME on high-affinity rolipram-binding sites was determined by replacing 2 nM [3H]-rolipram. AHR was assessed using the FlexiVent system and barometric plethysmography. Inflammatory cells were counted using a hemocytometer. Cytokines were determined using mouse T helper (Th)1/Th2 cytokine CBA kits, and total immunoglobulin (Ig)E or IgG2a levels were done using ELISA method. Xylazine (10 mg/kg)/ketamine (70 mg/kg)-induced anesthesia was performed. Results HDME revealed selective phosphodiesterase 4 (PDE4) inhibition with a therapeutic (PDE4H/PDE4L) ratio of 35.5 in vitro. In vivo, HDME (3~30 μmol/kg, orally (p.o.)) dose-dependently and significantly attenuated the airway resistance (RL) and increased lung dynamic compliance (Cdyn), and decreased enhanced pause (Penh) values induced by methacholine in sensitized and challenged mice. It also significantly suppressed the increases in the numbers of total inflammatory cells, macrophages, lymphocytes, neutrophils, and eosinophils, and levels of cytokines, including interleukin (IL)-2, IL-4, IL-5, interferon-γ, and tumor necrosis factor-α in bronchoalveolar lavage fluid (BALF) of these mice. In addition, HDME (3~30 μmol/kg, p.o.) dose-dependently and significantly suppressed total and ovalbumin-specific immunoglobulin (Ig)E levels in the BALF and serum, and enhanced IgG2a level in the serum of these mice. Conclusions HDME exerted anti-inflammatory effects, including suppression of AHR, and reduced expressions of inflammatory cells and cytokines in this murine model, which appears to be suitable for studying the effects of drugs on atypical asthma and COPD, and for screening those on typical asthma. However, HDME did not influnce xylazine/ketamine-induced anesthesia. Thus HDME may have the potential for use in treating typical and atypical asthma, and COPD.


Background
Phosphodiesterases (PDEs) are classified according to their primary protein and complementary (c)DNA sequences, co-factors, substrate specificities, and pharmacological roles. It is now known that PDEs comprise at least 11 distinct enzyme families that hydrolyze adenosine 3',5' cyclic monophosphate (cAMP) and/or guanosine 3',5' cyclic monophosphate (cGMP) [1]. PDE1~5 isozymes, which are calcium/calmodulin dependent (PDE1), cGMP stimulated (PDE2), cGMP inhibited (PDE3), cAMP specific (PDE4), and cGMP specific (PDE5), were found to be present in the canine trachea [2], guinea pig lungs [3], and human bronchi [4]. PDE3 and PDE4 were identified in the guinea pig airway [5], but other isozymes might also be present. PDE4 may adopt two different conformations which have high (PDE4 H ) and low (PDE4 L ) affinities for rolipram, respectively. In general, it is believed that inhibition of PDE4 H is associated with adverse responses, such as nausea, vomiting, and gastric hypersecretion, while inhibition of PDE4 L is associated with anti-inflammatory and bronchodilating effects. Therefore the therapeutic ratio of selective PDE4 inhibitors for use in treating asthma and chronic obstructive pulmonary disease (COPD) is defined as the PDE4 H /PDE4 L ratio [6,7].
Hesperetin (5,7,3'-trihydroxy-4'-methoxyflavanone), one of the most-common flavonoids in Citrus, is also present in herbal medicine as glycosides. For example, hesperidin and neohesperidin are abundantly present in the fruit peel of Citrus aurantium L. (Rutaceae), a wellknown traditional Chinese medicine (TCM) called "Chen-Pi"; they are used as an expectorant and stomach tonic, and contain vitamin P, a remedy for preventing capillary fragility and hypertension [8]. These glycosides are easily hydrolyzed by glycosidase to form hesperetin after ingestion. Based on lung histopathological studies using hematoxylin and eosin and alcian blue-periodic acid-Schiff staining, hesperidin was recently reported to inhibit inflammatory cell infiltration and mucus hypersecretion compared with the ovalbumin-induced group of mice in a murine model of asthma [9]. Men with higher hesperetin intake have lower mortality from cerebrovascular disease and lung cancer, and lower incidences of asthma [10]. Because hesperetin was reported to selectively inhibit PDE4 activity [11], it was used as a lead compound to synthesize hesperetin-7,3'-O-dimethylether (HDME), a more-liposoluble derivative of hesperetin. Therefore, we were interested in investigating the PDE4 H /PDE4 L ratio and suppressive effects of HDME on ovalbumin (OVA)-induced airway hyperresponsiveness (AHR), and clarifying its potential for treating asthma and COPD. Although both asthma and COPD are associated with an underlying chronic inflammation of the airways, there are important differences with regard to the inflammatory cells and mediators involved.
The key inflammatory cells in COPD are macrophages, CD8+ T-lymphocytes and neutrophils. Macrophages are strongly increased in the airway lumen, lung parenchyma and bronchoalveolar lavage fluid. In the airway wall and lung parenchyma, the ratio of CD8+/CD4+ T-lymphocytes increases. Neutrophils are increased in sputum and their number grows with the progression of the disease. In contrast, the key inflammatory cells in asthma are mast cells, eosinophils and CD4+ T-lymphocytes. Both diseases are sensitive to steroids. However, COPD shows a limited response to inhaled corticosteroids as compared to the efficacy achieved in asthma. Owing to the side effects of steroids, other therapeutics such as selective PDE4 or dual PDE3/4 inhibitors are developing. However, these developing inhibitors are also limited for the use of asthma and COPD in clinic because of their emetic side effect. This side effect can be easily assessed in non-vomiting species, such as rats or mice, in which selective PDE4 inhibitors reduce the duration of xylazine/ketamine-induced anesthesia [12,13].

Materials and methods
Reagents and animals HDME (mol wt., 330.27) was synthesized according to a previous method [14] in our laboratory and identified by spectral methods, including ultraviolet, infrared, mass spectroscopy, and nuclear magnetic resonance spectroscopic techniques. The purity of the compound exceeded 98% as determined by high-performance liquid chromatography. OVA, methacholine (MCh), aluminum sulfate hexadecahydrate, dimethylsulfoxide (DMSO), chloralose, urethane, Tris-HCl, Bis-Tris, benzamidine, phenylmethanesulfonyl fluoride (PMSF), d,l-dithiothreitol, polyethyleneimine, ethylenediaminetetraacetic acid (EDTA), bovine serum albumin (BSA), cAMP, cGMP, calmodulin, Dowex resin, Crotalus atrox snake venom, xylazine, and ketamine were purchased from Sigma Chemical (St. Louis, MO, USA). Vinpocetine, erythro-9-  Male Hartley guinea pigs (500~600 g) and female BABL/c mice at 8~12 weeks old were purchased from the Animal Center of the National Science Council (Taipei, Taiwan), and housed in ordinary cages at 22 ± 1°C with a humidity of 50%~60% under a constant 12/12-h light/dark cycle and provided with food and water ad libitum. Under a protocol approved by the Animal Care and Use Committee of Taipei Medical University, the following in vivo and in vitro experiments were performed.
Competitive inhibition of PDE1, PDE3, and PDE4 activities Activities of PDE1~5 in the homogenate of guinea pig lungs or hearts were measured by a two-step procedure according to the previous method [15], using cAMP with [ 3 H]-cAMP or cGMP with [ 3 H]-cGMP as substrates. In the Lineweaver-Burk analysis, the reaction mixture contained 10 μl of vehicle or inhibitors, at various concentrations of HDME or selective PDE1, PDE3, and PDE4 inhibitors, such as vinpocetine [16], milrinone [17], and Ro 20-1724 [18] as reference drugs. The reagents and homogenate were mixed on ice, and the reaction was initiated by transferring the mixture to a water bath at 37°C. Following a 30-min incubation, the reaction was stopped by transferring the reaction vessel to a bath of boiling water for 3 min. After cooling on ice, 20 μl of a 1 mg/ml solution of Crotalus atrox snake venom was added to the reaction mixture, and the mixture was incubated at 37°C for 10 min. Unreacted [ 3 H]-cAMP or [ 3 H]-cGMP was removed by the addition of 500 μl of a 1-in-1 Tris-HCl (40 mM) buffer suspension of Dowex resin (1 × 8-200) with incubation on ice for 30 min. Each tube was then centrifuged at 3700 g for 2 min, and 150 μl of the supernatant was removed for liquid scintillation counting. Less than 10% of the tritiated cyclic nucleotide was hydrolyzed in this assay. The total protein in each fraction used was assayed according to a previous method [19]. PDE activities are reported as nmol/mg/min.

Determination of PDE4 H values
When the above-mentioned guinea pigs were sacrificed, the whole brains were removed and homogenized with a glass/Teflon homogenizer (Glas-Col, Terre Haute, IN, USA) in 10 volumes of cold medium (pH 6.5) containing 20 mM Bis-Tris, 2 mM benzamidine, 2 mM EDTA, 50 mM sodium chloride, 0.1 mM PMSF, and 1 mM dithiothreitol. At 4°C, the homogenate was centrifuged at 170 g for 5 min to remove connective tissues and blood vessels. The suspended homogenate was then recentrifuged at 40,000 g for 30 min to separate the cytosolic and particulate portions. The particulate portion was re-suspended in a suspension at a concentration of 400 mg/ml (wet weight/volume), after washing three times with homogenizing buffer. The particulate portion mainly consisted of cell membranes. The binding ability of HDME (3~300 μM) to high-affinity rolipram-binding sites (HARBSs) of guinea pig brain cell membranes was determined by replacing 2 nM [ 3 H]-rolipram in a reaction buffer at 30°C for 1 h, according to the method described by previous investigators [20,21] and modified by us. Briefly, the reaction buffer consisted of 50 mM Tris-HCl and 5 mM MgCl 2 (pH 7.5). The total volume of the reaction mixture was 25 μl, consisting of 10 μl of the particulate suspension, 10 μl of [ 3 H]-rolipram, and 5 μl of HDME or Ro 20-1724 (1~10,000 nM), a reference drug. After 1 h, the reaction was terminated by moving the reaction vessel into crushed ice. Then the reaction mixture was transferred onto Whatman GF/B glass-fiber filters, which were soaked in a 0.3% polyethyleneimine solution in a mini-funnel. The reaction mixture was filtered by centrifugation at 90 g for 10 s, and the filtrate was collected into a 1.5-ml Eppendorf tube with the top adapted to the outlet of the mini-funnel. The filters were washed with 300 μl of reaction buffer three times each in the same way, and transferred into 2 ml of cocktail for radiation counting (total binding) using a β-scintillation counter (Beckman, Fullerton, CA, USA). Nonspecific binding, which was defined in the presence of 10 μM Ro 20-1724, was subtracted from total binding to yield specific binding. Effective concentration (EC 50 ) values of HDME and Ro 20-1724, at which a half of the [ 3 H]-rolipram that was bound onto HARBSs of cell membranes was displaced, were defined as PDE4 H values, and these were related to any adverse effects, such as nausea, vomiting, and gastric hypersecretion [7].

Airway hyperresponsiveness (AHR) in vivo
According to the schedule (Figure 1), ten female BABL/ c mice in each group were sensitized by an intraperitoneal (i.p.) injection of 20 μg of OVA emulsified in 2.25 mg of an aluminum hydroxide gel, prepared from aluminum sulfate hexadecahydrate, in a total volume of 100 μl on days 0 and 14. On day 21, these mice were injected with (i.p.) 100 μl of a mixture of 1% OVA and Freund's complete adjuvant (1:1). Mice were challenged via the airway using 1% OVA in saline for 30 min on days 28, 29, and 30 by ultrasonic nebulization. After the last of OVA challenges [22], AHR was assessed on day 32 (48 h after 1% OVA provocation) in each group. Each group of mice was orally (p.o.) administered the vehicle (control) or 3~30 μmol/kg of HDME 2 h before and 6 and 24 h after OVA provocation. For comparison, sham-treated mice were challenged with saline instead of 1% OVA (non-challenged). The vehicle, a mixture of DMSO: ethyl alcohol: PEG 400: saline (0.5: 0.5: 1: 8, v/ v), or HDME was administered (p.o.) at a volume of 0.01 ml/g of body weight. AHR was assessed using two methods: (1) in anesthetized ventilated mice, AHR was assessed as previously described [23] by measuring changes in the airway resistance (R L, cmH 2 O/ml/sec) and lung dynamic compliance (C dyn , ml/cmH 2 O) after challenge with aerosolized methacholine (MCh, 0.78~25 mg/ml) using the FlexiVent system (SCIREQ, Montreal, Quebec, Canada). Anesthetized (urethane 600 mg/kg and chloralose 120 mg/kg, i.p.), tracheostomized (stainless-steel cannula, 18 G) mice were mechanically ventilated (at 150 breaths/min, with a tidal volume of 10 ml/ kg, positive end-expiratory pressure of 3 cmH 2 O). (2) in unrestrained animals by barometric plethysmography [24] using a whole-body plethysmograph (WBP) and analyzed using software of Life Science Suite P3 Analysis Modules (Gould, LDS Test and Measurement LLC, Valley View, OH, USA). Mice were placed into the main chamber of the WBP, and the baseline enhanced pause (P enh ) value was determined. Then mice were first nebulized with phosphate-buffered saline (PBS), and subsequently with increasing doses (6.25~50 mg/ml) of MCh for 3 min for each nebulization, followed by readings of breathing parameters for 3 min after each nebulization to determine P enh values.

Inflammatory cells, cytokines and immunoglobulins
Twenty-four hours after P enh determination, these mice were anesthetized with pentobarbital (50 mg/kg, i.p.), and the lungs were lavaged via a tracheal tube with PBS (1 × 1.0 ml, 37°C). After lavage, blood was collected from the jugular vein and allowed to sit so that it would coagulate. The collected bronchoalveolar lavage fluid (BALF) and coagulated blood were respectively centrifuged at 630 g for 7 min and at 3700 g for 10 min at 4°C. After centrifugation, the BALF and serum supernatants were stored at -20°C until determination of cytokines, including interleukin (IL)-2, IL-4, IL-5, tumor necrosis factor (TNF)-α, and interferon (IFN)-γ by flow cytometric methods [25] using mouse T helper (Th)1/ Th2 cytokine CBA kits, and of total immunoglobulin (Ig)E or IgG 2a using ELISA kits (Pharmingen, San Diego, CA, USA) according to the respective recommendations of the manufacturers. OVA-specific IgE was measured as described previously [26]. Wells were coated with 100 μl of OVA (20 μg/ml) instead of the capture antibody. Levels are expressed in arbitrary units, where 1 arbitrary unit equals the optical density of the sample divided by the optical density of unchallenged mouse serum or BALF (standard). The BALF pellet was resuspended in ACK lysing buffer (1.658 g NH 4 Cl, 0.2 g KHCO 3 and 1.44 mg EDTA in 200 ml of water) to lyse the residual erythrocytes in each sample. The number of inflammatory cells was counted using a hemocytometer (Hausser Scientific, Horsham, PA, USA). Cytospun slides were stained and differentiated in a blinded fashion by counting at least 100 cells under light microscopy. All undetectable data (< 1 pg/ml) of cytokines were taken as 0 pg/ml.

Statistical methods
All values are given as the means ± SEM. Differences among values were statistically calculated by one-way analysis of variance (ANOVA), and then determined by Dunnett's test. The difference between two values, however, was determined by the use of Student's t-test. Differences with p < 0.05 were considered statistically significant.

Supression of airway hyperresponsiveness in vivo
R L values at the baseline for the control (vehicle), nonchallenged, and 3, 10, and 30 μmol/kg HDME groups were 1.03 ± 0.03, 1.04 ± 0.05, 1.06 ± 0.05, 1.08 ± 0.06, and 1.09 ± 0.07 cmH 2 O/ml/s, respectively, and these values did not significantly differ from each other. R L values of PBS nebulization for each group were 1.05 ± 0.04, 1.06 ± 0.05, 1.10 ± 0.07, 1.06 ± 0.04, and 1.07 ± 0.06 cmH 2 O/ml/s, respectively, which also did not significantly differ from each other. Administration of nebulized PBS did not affect the R L values of the baseline in each group. However, MCh (6.25~25 mg/ml) concentration-dependently and significantly increased R L values ( Figure 5A), and decreased C dyn values ( Figure 5B) in the control sensitized and challenged group compared to the non-challenged group. HDME (3~30 μmol/kg, p. o.) significantly suppressed these changes ( Figure 5).
P enh values at the baseline for the control (vehicle), non-challenged, and 3, 10, and 30 μmol/kg HDME groups were 2.32 ± 0.04, 2.41 ± 0.05, 2.45 ± 0.04, 2.36 ± 0.05, and 2.43 ± 0.03, respectively, and these values did not significantly differ from each other. P enh values with PBS nebulization for each group were 2.42 ± 0.05, 2.41 ± 0.04, 2.43 ± 0.05, 2.38 ± 0.06, and 2.44 ± 0.06, respectively, which also did not significantly differ from each other. Administration of nebulized PBS did not affect the P enh value of the baseline in each group. However, MCh (6.25~50 mg/ml) concentration-dependently increased P enh values from 1-fold with PBS exposure to 1.85 ± 0.20-fold in control sensitized and challenged mice ( Figure 5A). P enh values of MCh at 25 and 50 mg/  ) and Ro 20-1724 (B) in high-affinity rolipram binding sites of guinea pig brain particulate. Each value represents the mean ± SEM. The experimental number for HDME at each concentration was 6, and for Ro 20-1724 was 4. Figure 5 Effects of HDME (3~30 µmol/kg, p.o.) on airway resistance. Effects of HDME (3~30 µmol/kg, p.o.) on airway resistance (R L , A) and lung dynamic compliance (C dyn , B) in sensitized and challenged mice which received aerosolized methacholine (MCh, 0.78~25 mg/ml) 2 days after primary allergen challenge. ### P <0.001, compared to the non-challenged group. * p < 0.05, ** p <0.01 and *** p <0.001, compared to the control (vehicle) group. The number of mice in each group was 10. PBS, phosphate-buffered saline. ml in control sensitized and challenged mice were significantly enhanced compared to those in non-challenged mice. HDME (3~30 μmol/kg, p.o.) dosedependently and significantly attenuated the enhancement of P enh values induced by 25 and 50 mg/ml MCh ( Figure 6A).

Suppression of inflammatory cells and cytokines in the BALF
In this special animal model, the number of neutrophils in the bronchoalveolar lavage fluid of control sensitized and challenged mice was significantly more than that of eosinophils. The numbers of total inflammatory cells, macrophages, lymphocytes, neutrophils, and eosinophils from the BALF of control sensitized and challenged mice significantly increased compared to those of nonchallenged mice ( Figure 6B). HDME (3~30 μmol/kg, p. o.) significantly suppressed the increases in numbers of total inflammatory cells, macrophages, lymphocytes, neutrophils, and eosinophils ( Figure 6B). Noticeably, the numbers of eosinophils were abolished by HDME at various doses used. Compared to those in non-challenged mice, levels of cytokines, such as IL-2, IL-4, IL-5, IFN-γ, and TNF-α, in the BALF of control sensitized and challenged mice significantly increased ( Figure 6C). HDME (3~30 μmol/kg, p.o.) also significantly suppressed increases in levels of IL-2, IL-4, IL-5, and TNF-α, but enhanced the level of IFN-γ at 30 μmol/kg ( Figure 6C).

Suppression of IgE and IgG 2a in the serum and BALF
The level of total IgG 2a in the serum of control sensitized and challenged mice was significantly supressed compared to that of non-challenged mice. HDME (3~30 μmol/kg, p.o.) dose-dependently and significantly enhanced this supression ( Figure 7A). However, levels of total and OVA-specific IgE in the BALF and serum of control sensitized and challenged mice were significantly enhanced compared to those of non-challenged mice. HDME (3~30 μmol/kg, p.o.) dose-dependently and significantly suppressed these enhancements ( Figure 7B-E).

Discussion
Allergic asthma (type-I allergic airway disease) is a chronic respiratory disease characterized by AHR, mucus hypersecretion, bronchial inflammation, and elevated IgE levels. Th2 cells, together with other inflammatory cells such as eosinophils, B cells, and mast cells are thought to play critical roles in the initiation, development, and chronicity of this disease [27]. This clinical Figure 6 Effects of HDME (3~30 μmol/kg, p.o.) on the enhanced pause (P enh ). Effects of HDME (3~30 μmol/kg, p.o.) on the enhanced pause (P enh ) (A), inflammatory cells (B), and cytokines (C) in sensitized mice which received aerosolized methacholine (6.25~50 mg/ml) 2 days after primary allergen challenge. # p <0.05, ## p <0.01, and ### p <0.001, compared to the non-challenged group. * p <0.05, ** p <0.01, and *** p <0.001, compared to the control (vehicle) group. The number of mice in each group was 10. Total, total cells; Mac, macrophages; Lym, lymphocytes; Neu, neutrophils; Eos, eosinophils; IL, interleukin; IFN, interferon; TNF, tumor necrosis factor. definition fails to account for the atypical and often more severe phenotype found in a considerable proportion of asthmatics who have increased neutrophil cell counts in the airway as a distinguishing trait. Neutrophilic inflammation is a hallmark of another type of allergic airway pathology, hypersensitivity pneumonitis. Considered as an immune counterpart of asthma, hypersensitivity pneumonitis is a prototypical type-III allergic inflammatory reaction involving the alveoli and lung interstitium, steered by Th1 cells and IgG and, in its chronic form, accompanied by fibrosis [28]. Thus, this animal model appears to be suitable for studying the effects of drugs on the atypical asthma and COPD, and for screening those on typical asthma. One hypothesis emphasizes an imbalance in Th cell populations favoring expression of Th2 over Th1 cells in typical asthma. Cytokines released from Th2 cells are IL-4, IL-5, IL-6, IL-9, and IL-13, and those from Th1 cells are IL-2, IL-12, IFN-γ, and TNF-α [29,30]. In the present results, HDME (3~30 μmol/kg, p.o.) significantly decreased R L ( Figure 5A), and increased C dyn (Figure 5B), and also attenuated P enh values ( Figure 6A) suggesting that it significantly suppresses AHR. The numbers of all types of inflammatory cells examined, including total inflammatory cells, macrophages, lymphocytes, neutrophils, and eosinophils in the BALF of sensitized and challenged mice were reduced by HDME (3~30 μmol/kg, p.o.) (Figure 6B). It is well known that after oral administration and digestion of hesperidin, a flavanone glycoside Figure 8 Effects of subcutaneously administered rolipram (A) and HDME (B). Effects of subcutaneously administered rolipram (A) and HDME (B) on the duration of xylazine (10 mg/kg, i.p.)/ketamine (70 mg/kg, i.p.)-induced anesthesia in mice. Rolipram was administered 0.25 h and HDME 1 h before anesthesia. ** p <0.01, *** p <0.001, compared to the vehicle (control). Each value represents the mean ± SEM. The number of mice in each group was 10.

Figure 7
Effects of HDME (3~30 μmol/kg, p.o.) on total IgG 2a and IgE. Effects of HDME (3~30 μmol/kg, p.o.) on total IgG 2a (A), total IgE (B) and ovalbumin-specific IgE (C) levels in the serum and bronchial alveolar lavage fluid (D, E) of sensitized mice which had received aerosolized methacholine (6.25~50 mg/ml) 2 days after primary allergen challenge. # p <0.05, ## p <0.01 and ### p <0.001, compared to the non-challenged group. * p <0.05, ** p <0.01 and *** p <0.001, compared to the control (vehicle) group. Each value represents the mean ± SEM. The number of mice in each group was 10. comprised of the flavanone hesperetin and the disaccharide rutinose, forms hesperetin. Similarly, hesperetin is also formed by demethylation of HDME after oral administration. However, whether the effects of HDME on lung tissue are similar to those of hesperidin [9] needs to be further investigated. It also suppressed levels of IL-2, IL-4, IL-5, and TNF-α, but significantly enhanced the level of IFN-γ ( Figure 6C). These results suggest that HDME fully suppresses Th2 cells and partially activates Th1 cells, and ameliorates this imbalance occurred in typical asthma. However, this partial activation of Th1 cells may offset, at least in a part, some anti-inflammatory effects of HDME, by which IL-2 and TNF-α released from Th1 cells were reduced. However, the number of neutrophils was significantly reduced by HDME, suggesting that it may have a benefit for treating atypical asthma. Similarly, the numbers of macrophages and neutrophils were reduced by HDME, suggesting that it may ameliorate COPD too.
IL-4 and IL-13 were shown to induce AHR in mouse asthma models [31,32]. IL-4 has three primary effects. First, IL-4 promotes B cell differentiation to plasma cells that secrete antigen-specific IgE antibodies. Second, IL-4 promotes mast cell proliferation. Third, increased IL-4 upregulates endothelial cell expression of adhesion molecules for eosinophils [33]. IL-5 mobilizes and activates eosinophils, leading to the release of a major basic protein, cysteinyl-leukotriene, and eosinophil peroxidase that contribute to tissue damage and AHR [32,34]. Phosphoinositide 3-kinase δ (p110δ) was shown to play a crucial role in the development, differentiation, and antigen receptor-induced proliferation of mature B cells [35,36], and inhibition of p110δ attenuates allergic airway inflammation and AHR in a murine asthma model [35,37]. In addition, IL-4 and IL-13 are important in directing B cell growth, differentiation, and secretion of IgE [38]. However, IFN-γ released from Th1 cells preferentially directs B cell switching of IgM to IgG 2a and IgG 3 in mice [39,40]. HDME (3~30 μmol/kg, p.o.) herein dose-dependently and significantly enhanced total IgG 2a level in the serum and suppressed total and OVA-specific IgE levels in the BALF and serum of sensitized and challenged mice, suggesting that HDME has immunoregulatory and antiallergic asthmatic effects.
In the present results, HDME selectively inhibited PDE4 activity with the IC 50 and K i values of 3.0 and 2.1 μM, respectively. Selective PDE4 inhibitors specifically prevent the hydrolysis of cAMP, a 3',5'-cyclic nucleotide, and therefore have broad anti-inflammatory effects such as inhibition of cell trafficking and of cytokine and chemokine release from inflammatory cells. The increased cAMP levels induced by these selective PDE4 inhibitors subsequently activate cAMP-dependent protein kinase which may phosphorylate and inhibit myosin light-chain kinase, thus inhibiting contractions [41]. The precise mechanism through which relaxation is produced by this second-messenger pathway is not known, but it may result from decreased intracellular Ca 2+ ([Ca 2+ ] i ). The decrease in [Ca 2+ ] i may be due to reduced influx of Ca 2+ , enhanced Ca 2+ uptake into the sarcoplasmic reticula, or enhanced Ca 2+ extrusion through cell membranes [41]. Thus selective PDE4 inhibitors may have bronchodilatory effects. The second-generation PDE4 inhibitors, cilomilast and roflumilast, have reached the clinical trial stage and exhibit some beneficial effects in treating asthma and COPD [42]. The effectiveness of these PDE4 inhibitors may be limited by their clinical potency when using doses that have minimal adverse effects such as headaches, diarrhea, nausea, and abdominal pain. The PDE4 H /PDE4 L ratios of cilomilast and roflumilast were respectively reported to be 117.8 nM/ 120 nM (1), and 2.4 nM/0.8 nM (3) [21,43], which are considerably greater than that (0.01~0.001) of rolipram [7]. Owing to its adverse effects or lack of efficacy, cilomilast was discontinued for use against asthma after phase II clinical trials in 2003 [42]. In terms of tolerability over 6 months with 15 mg twice daily for COPD in a phase III study, cilomilast was reported to be associated with higher frequencies of diarrhea and nausea than a placebo [42]. Roflumilast was evaluated for asthma and COPD in phase III clinical trials, and was reported to reduce those adverse effects after longer-term treatment at 0.5 mg once daily [42]. Roflumilast, compared to a placebo, was reported to significantly improve the mean pre-and post-bronchodilator forced expiratory volumes in 1 s (FEV 1 ) in patients with moderate-to-severe COPD. However, nausea, diarrhea, weight loss, and headaches were more frequent in patients in the roflumilast group. These adverse events were associated with increased patient withdrawal [44,45]. Recently, roflumilast was approved by the European Commission as an add-on to bronchodilator therapy for maintenance treatment of severe COPD associated with chronic bronchitis in adults with a history of frequent exacerbations. However, the US Food and Drug Administration voted against using roflumilast to treat COPD. The PDE4 H / PDE4 L ratio of AWD 12-281, another selective PDE4 inhibitor, was reported to be 104 nM/9.7 nM (approximately 11) [46]. AWD 12-281 was undergoing clinical development phase IIa trials for COPD, and was reported to be a unique potential drug for the topical treatment of asthma and COPD [47]. AWD 12-281 was reported to be a very promising drug candidate for treating lung inflammation when administered by inhalation and for treating atopic dermatitis [48]. However, AWD-12-281 was also discontinued in clinical trials for both asthma and COPD owing to a lack of efficacy [49,50]. Many compounds that are in development will not reach the market as monotherapies unless their emetic liability is reduced [51], although inhaled GSK256066 demonstrated efficacy in trials in asthma [52] and oral apremilast was clinically reported to be effective for treating severe plaque-type psoriasis [53]. PDE4 subtypes (A~D) may be considered for drug development of new PDE4 inhibitors. PDE4D inhibition in non-target tissues promotes emesis, since PDE4D knock-out mice showed reduction of xylazine/ketaminetriggered anesthesia which is used as a surrogate marker for emesis in mice, a non-vomiting species [13]. Recently, small-molecule allosteric modulators of PDE4D that do not completely inhibit enzymatic activity were reported to reduce emesis and have therapeutic benefits of a brain distribution, for such entities as Alzheimer's disease, Huntington's disease, schizophrenia, and depression [54]. In contrast to PDE4D, selective inhibition of PDE4A and/or PDE4B in proinflammatory and immune cells is believed to evoke the therapeutically desired effects of these drugs [55]. Cilomilast has a higher potency for PDE4D compared to PDE4A and PDE4B, while roflumilast is non-selective for these four PDE4 subtypes with similar degrees of inhibition [56].
There is no literature about AWD 12-281's inhibition of PDE4 subtypes until now. However, whether HDME selectively inhibits the PDE4 subtype also needs to be further investigated.
In the present results, the PDE4 H /PDE4 L ratio of HDME was calculated to be 35.5, which is considerably greater than that of AWD 12-281. In addition, HDME did not influnce xylazine/ketamine-induced anesthesia. However, rolipram, a selective PDE4 inhibitor, reversed the anesthesia. The reversing effect may occur through presynaptic α 2 -adrenoceptor inhibition [57], because MK-912, an α 2 -adrenoceptor antagonist, was reported to reverse xylazine/ketamine-induced anesthesia in rats [12] and trigger vomiting in ferrets [57]. In contrast, clonidine, an α 2 -adrenoceptor agonist, prevented emesis induced by PDE4 inhibitors in ferrets [57]. The present results also suggest that HDME may have few or no adverse effects, such as nausea, vomiting, and gastric hypersecretion.

Conclusions
In conclusion, HDME exerted anti-inflammatory effects, including suppression of AHR, and reduced expressions Figure 9 Mechanisms of action of HDME. HDME selectively inhibits PDE4 activities and results in an increase in cAMP, which activates cAMPdependent protein kinase (PKA) and increases calcium extrusion from the intracellular space and uptake to sarcoplasmic reticula (SR). Therefore, HDME largely decreases the concentration of intracellular calcium ([Ca 2+ ] i ) and results in bronchodilatation. The increase in cAMP also has antiinflammatory and immunoregulatory effects. AC, adenylate cyclase; Th, T-helper cells; Ig, immunoglobulin; IL, interleukin; IFN, interferon; TNF, tumor necrosis factor. Up and down arrows respectively indicate increases and decreases.
of inflammatory cells and cytokines in this murine model, which appears to be suitable for studying the effects of drugs on atypical asthma and COPD, and for screening those on typical asthma. Its mechanisms are summarized in Figure 9. However, HDME did not influnce xylazine/ketamine-induced anesthesia. Thus, HDME may have the potential for use in treating typical and atypical, at least in part, asthma, and COPD.