Volume 165, Issue 5 p. 1501-1514
Free Access

Non-bronchodilating mechanisms of tiotropium prevent airway hyperreactivity in a guinea-pig model of allergic asthma

KS Buels

KS Buels

Department of Physiology and Pharmacology

Search for more papers by this author
DB Jacoby

DB Jacoby

Division of Pulmonary and Critical Care Medicine, Oregon Health & Science University, Portland, OR, USA

Search for more papers by this author
AD Fryer

Corresponding Author

AD Fryer

Division of Pulmonary and Critical Care Medicine, Oregon Health & Science University, Portland, OR, USA

Allison D. Fryer, Mail Code UHN67, Oregon Health & Science University, 3181 SW Sam Jackson Park Rd., Portland, OR 97239, USA. E-mail: [email protected]Search for more papers by this author
First published: 22 August 2011
Citations: 40

Abstract

BACKGROUND AND PURPOSE Asthma is characterized by reversible bronchoconstriction and airway hyperreactivity. Although M3 muscarinic receptors mediate bronchoconstriction, non-selective muscarinic receptor antagonists are not currently recommended for chronic control of asthma. We tested whether selective blockade of M3 receptors, at the time of antigen challenge, blocks subsequent development of airway hyperreactivity in antigen-challenged guinea-pigs.

EXPERIMENTAL APPROACH Ovalbumin-sensitized guinea-pigs were pretreated with 1 µg·kg−1 of a kinetically selective M3 receptor antagonist, tiotropium, or 1 mg·kg−1 of a non-selective muscarinic receptor antagonist, atropine, and challenged with inhaled ovalbumin. Animals were anaesthetized, paralyzed, ventilated and vagotomized 24 h later. We measured vagally mediated bronchoconstriction and i.v. ACh-induced bronchoconstriction.

KEY RESULTS Electrical stimulation of both vagus nerves induced frequency-dependent bronchoconstriction in sensitized animals that was significantly increased after antigen challenge. Antigen-induced hyperreactivity was completely blocked by tiotropium pretreatment but only partially blocked by atropine pretreatment. Surprisingly, although tiotropium blocked bronchoconstriction induced by i.v. ACh, it did not inhibit vagally-induced bronchoconstriction in sensitized controls, suggesting that tiotropium does not block hyperreactivity by blocking receptors for vagally released ACh. Rather, tiotropium may have worked through an anti-inflammatory mechanism, since it inhibited eosinophil accumulation in the lungs and around nerves.

CONCLUSIONS AND IMPLICATIONS These data confirm that testing M3 receptor blockade with exogenous ACh does not predict vagal blockade. Our data also suggest that selective blockade of M3 receptors may be effective in asthma via mechanisms that are separate from inhibition of bronchoconstriction.

Abbreviations

  • i.s.
  • insufflated
  • Introduction

    In the lungs, airway smooth muscle tone is controlled by ACh released from parasympathetic nerves. ACh stimulates M3 muscarinic receptors on airway smooth muscle to induce smooth muscle contraction, resulting in bronchoconstriction. Parasympathetic nerves not only maintain airway tone, but they also mediate reflex bronchoconstriction (Canning, 2006). Diverse chemical and physical stimuli, including inhaled antigens, histamine and cold air, activate afferent sensory nerves in the lungs to initiate a reflex that increases ACh release from parasympathetic nerves and bronchoconstriction (Gold et al., 1972; Sheppard et al., 1982; Canning, 2006).

    Humans with asthma are hyperreactive to bronchoconstrictor stimuli. Exposure to antigens (antigen challenge) causes inflammation and airway hyperreactivity that develop within 24 h and persist for up to a week in both humans with asthma and sensitized animals (Cockcroft et al., 1977; Chung et al., 1985; Brown et al., 1998). In antigen-challenged animals, airway hyperreactivity is mediated by increased ACh release from the vagus nerves (Evans et al., 1997, 2001; Verbout et al., 2007). Thus, anticholinergics, which block the action of ACh at M3 muscarinic receptors, would be expected to be beneficial in asthma. However, anticholinergics are not currently recommended for long-term control of asthma (EPR-3, 2007). This is because many studies that used the anticholinergics atropine or ipratropium in humans with asthma found they were unable to effectively block bronchoconstriction (Fryer and Jacoby, 2008). This may be due, in part, to the low doses of anticholinergics used, since benefits were found in studies with higher doses (Ward et al., 1981; Sheppard et al., 1982).

    Another possible explanation is that atropine and ipratropium are not selective for M3 muscarinic receptors over other muscarinic receptor subtypes. M1–M3 muscarinic receptors are expressed on structural cells in the lungs, including smooth muscle, nerves, mucous glands, epithelial cells, fibroblasts and endothelial cells, and all five muscarinic receptors (M1–M5) are found on inflammatory cells (Wessler and Kirkpatrick, 2008). Non-selective blockade of all muscarinic receptor subtypes could counteract the inhibition of bronchoconstriction through M3 muscarinic receptor blockade in asthma. For example, M2 muscarinic receptors found on prejunctional parasympathetic nerves are stimulated by ACh to limit further release of ACh from the nerves onto airway smooth muscle (Fryer and Jacoby, 2008). Blocking these inhibitory M2 receptors can increase ACh enough to overcome M3 receptor blockade and increase bronchoconstriction, as has been shown in guinea-pigs treated with atropine and ipratropium (Fryer and Maclagan, 1987).

    Experiments in this paper were designed to test whether selectively blocking M3 muscarinic receptors at the time of antigen challenge reduces subsequent airway hyperreactivity. Tiotropium bromide, a kinetically selective muscarinic receptor antagonist, was used to block M3 muscarinic receptors in sensitized guinea-pigs at the time of antigen challenge, and airway hyperreactivity was measured 24 h later. The results show that selective blockade of M3 muscarinic receptors prevents airway hyperreactivity through a mechanism that is separate from direct inhibition of bronchoconstriction, and that this blockade is associated with inhibition of the inflammatory response to antigen challenge.

    Methods

    Animals

    Specific pathogen-free female Hartley guinea-pigs (Elm Hill Labs; Chelmsford, MA, USA) were housed in high-efficiency particulate-filtered air with ad libitum access to food and water. All animal care and experimental procedures were in accordance with the National Institutes of Health (NIH) guidelines, and were approved by the Oregon Health & Science University Institutional Animal Care and Use Committee.

    Sensitization and challenge with antigen

    All guinea-pigs (150–200 g) were sensitized to Grade II ovalbumin (20 mg·kg−1, i.p., Sigma-Aldrich, St. Louis, MO, USA) on days 1, 3 and 6. Treatments and challenge were given 21 days after the last injection. Some animals were challenged with an aerosol of 5% ovalbumin containing 0.2% antifoam Y-30 emulsion (Sigma-Aldrich) in sterile PBS for 10 min or until signs of respiratory distress appeared, in which case antigen challenge was immediately stopped (three of 27 animals).

    Treatment with insufflated tiotropium and lactose

    Tiotropium is a kinetically selective M3 receptor antagonist that dissociates more slowly from M3 (human t½= 27–34.7 h) muscarinic receptors than M2 (human t½= 2.6–3.6 h) or M1 (human t1/2= 10.5–14.6 h) muscarinic receptors (Disse et al., 1993; Casarosa et al., 2009). Tiotropium is the active ingredient in Spiriva (HandiHaler capsules, Boehringer Ingelheim Pharmaceuticals Inc., Ridgefield, CT, USA), which contains α lactose monohydrate filler. Tiotropium was administered as a powder to some sensitized animals using a DP-4 dry powder insufflator (Penn-Century Inc., Philadelphia, PA, USA) to mimic the method of delivery in humans. Since 2–5 mg of powder must be loaded in this insufflator, Spiriva was mixed with additional D-(+)-lactose monohydrate (Mallinckrodt Baker Inc., Phillipsburg, NJ, USA). Guinea-pigs were anaesthetized with ketamine (30 mg·kg−1, i.m., JHP Pharmaceuticals LLC, Rochester, MI, USA) and xylazine (5 mg·kg−1, i.m., Vedco Inc., St. Joseph, MO, USA) and supported vertically while the insufflator delivery tube was inserted into the trachea with the aid of a small animal laryngoscope (WelchAllyn, Skaneateles Falls, NY, USA). Tiotropium bromide 0.2 µg·kg−1 (0.05 mg Spiriva·kg−1) or 1 µg·kg−1 (0.25 mg Spiriva·kg−1) was then insufflated (i.s.) into the lungs. Vehicle control animals for these experiments received an equivalent amount of lactose powder. The actual dose administered was confirmed by weighing the insufflator sample chamber before and after delivery.

    Treatment with atropine

    Some sensitized animals were treated with the non-selective muscarinic receptor antagonist, atropine (1 mg·kg−1, i.p.) 1 h before challenge and again 6 h after challenge as previously described (Verbout et al., 2007).

    Experimental groups

    To determine the effect of selectively blocking M3 receptors with tiotropium at the time of challenge on airway hyperresponsiveness, physiological measurements were made in seven groups of animals. Three groups of animals were sensitized: (i) sensitized controls; (ii) sensitized animals treated with lactose; and (iii) sensitized animals treated with 1 µg·kg−1 tiotropium. In vivo physiology was measured 48 h after tiotropium or lactose administration in these animals. Four groups of animals were sensitized and challenged: (i) sensitized and challenged animals; (ii) sensitized animals treated with lactose as a vehicle control and challenged 24 h later; (iii) sensitized animals treated with 1 µg·kg−1 tiotropium and challenged 24 h later; and (iv) sensitized animals treated with atropine and challenged 1 h later. In vivo physiology was measured 24 h after challenge with inhaled ovalbumin in these groups, which corresponds to 48 h after tiotropium or lactose administration and 25 h after the first injection of atropine.

    Physiological measurements were also made at the time of challenge (24 h after treatment with lactose or tiotropium) in four groups of animals: (i) sensitized controls (anaesthetized with ketamine and xylazine); (ii) sensitized animals treated with lactose (vehicle control); (iii) sensitized animals treated with 0.2 µg·kg−1 tiotropium; and (iv) sensitized animals treated with 1 µg·kg−1 tiotropium.

    Measurement of pulmonary inflation pressure and vagal reactivity

    Guinea-pigs were anaesthetized with urethane (1.7 g·kg−1 i.p., Sigma-Aldrich Chemical Co.), and temperature was maintained at 37°C. Jugular veins were cannulated for drug administration, and heart rate and blood pressure were measured via a carotid artery cannula to ensure adequate levels of anaesthesia. Animals were chemically sympathectomized with guanethedine (2 mg·kg−1, i.v., Bosche Scientific, New Brunswick, NJ, USA), paralysed with succinylcholine chloride (5 µg·min−1, i.v., Sigma-Aldrich), and mechanically ventilated via a tracheal cannula (tidal volume 2.5 mL, 100 breaths·min−1). Guinea-pigs were vagotomized by crushing both vagus nerves, and distal portions of both vagi were placed on platinum electrodes and submerged in mineral oil. Pulmonary inflation pressure was measured via a sidearm of the tracheal cannula.

    Both vagus nerves were electrically stimulated simultaneously (1–25 Hz, 10 V, 0.2 ms pulse duration, for 5 s at 45–60 s intervals) to produce frequency-dependent bronchoconstriction, measured as an increase in pulmonary inflation pressure in mmH20, and bradycardia. Atropine (1 mg·kg−1, i.v.) was given at the end of each experiment to confirm that cholinergic nerves mediated the vagally-induced bronchoconstriction.

    Measurement of post-junctional muscarinic receptor function

    Following measurement of vagal reactivity, ACh (1–10 µg·kg−1, i.v., Acros Organics, Ceel, Belgium) was administered to test the function of post-junctional M2 receptors in cardiac muscle and post-junctional M3 receptors in airway smooth muscle.

    Inhibition of bronchoconstriction following i.v. administration of lactose and tiotropium in non-sensitized guinea-pigs

    Vagus nerve-induced bronchoconstriction was elicited by electrically stimulating both vagus nerves every 2 min for 5 s (10 Hz, 10 V, 0.2 ms pulse duration) in anaesthetized animals. ACh (4 µg·kg−1, i.v.)-induced bronchoconstriction was elicited in the same animals following every third vagal stimulation. Gallamine (3 mg·kg−1, i.v.) was administered to block neuronal M2 receptors. Once reproducible baseline bronchoconstrictions were achieved, tiotropium bromide (0.3–10 µg·kg−1) was given by administering Spiriva dissolved in PBS i.v. Each increasing dose of tiotropium was given after inhibition of bronchoconstriction reached a plateau (time determined in early experiments and maintained throughout). Time/vehicle control animals received equivalent amounts of i.v. lactose. Following each dose, the last two i.v. ACh-induced bronchoconstrictions were averaged and presented as a percentage of baseline. The second and third bronchoconstrictions following vagal stimulation were similarly averaged and presented as a percentage of baseline.

    Bronchoalveolar lavage analysis

    At the end of each experiment, lungs were lavaged with five 10 mL aliquots of PBS containing 10 µg·mL−1 isoprenaline (Sigma-Aldrich). Total cells were counted using a haemocytometer. Cells were also cytospun onto slides and stained with Hemacolor (EMD Chemicals Inc., Gibbstown, NJ, USA) to obtain differential cell counts.

    Histopathological evaluation of eosinophils

    Animals were then perfused with PBS and exsanguinated. Lungs were removed, inflated with zinc-buffered formalin (Anatech Ltd, Battle Creek, MI, USA), and fixed overnight at 4°C. Two transverse sections of the proximal region of each of two lobes were embedded in paraffin and processed for immunohistochemistry. Airway nerves were detected with a monoclonal antibody specific for PGP9.5 and eosinophils were stained with chromotrope 2R [modified from Evans et al. (2001)]. Antigen retrieval was carried out by microwaving slides in antigen unmasking solution (Vector Laboratories Inc., Burlingame, CA, USA). Endogenous peroxidase activity was quenched with 3% H2O2 diluted in cold methanol (10 min). Sections were permeabilized with 0.05% Tween 20 in PBS (2 × 3 min), blocked in 10% normal goat serum (30 min room temperature, Vector Laboratories Inc.), incubated with mouse anti-PGP9.5 (5 µg·mL−1, 1 h room temperature, AbD Serotec, Cat no. 7863-2004, Raleigh, NC, USA), and incubated with biotinylated goat anti-mouse IgG (7.5 µg·mL−1, 30 min room temperature, Vector Laboratories Inc., BA-9200). Antibody staining was detected with a streptavidin-linked horseradish peroxidase substrate (Vectastain ABC kit, Vector Laboratories Inc.) and Vector SG chromagen (Vector Laboratories Inc.). Normal mouse IgG or no primary antibody served as controls. Slides were then stained with 1% chromotrope 2R (Sigma-Aldrich) in diH2O for 1 min, air dried overnight and mounted.

    Eosinophils within airway walls (large bronchioles) were quantified by an investigator blinded to treatment. Overlapping 400× photographs of each airway were taken and overlaid to make a single image for each airway using a digital camera and NIH Image J Software [version 1.42q, NIH, Bethesda, MD, USA (Rasband, 1997–2009)] with Mosaic J plugin (Thevenaz and Unser, 2007). Total airway area between the epithelial basement membrane and surrounding alveoli was measured using MetaMorph imaging software (version 7.1.2.0, Molecular Devices, Sunnyvale, CA, USA). Total eosinophils and nerve-associated eosinophils (within 8 µm of a PGP9.5-positive nerve) were counted and expressed as eosinophils·mm−2 airway area. Values were averaged across three airways for each animal, and these numbers were used to calculate the mean for each group.

    Statistical analysis

    Values are means ± SEM. Baseline pulmonary and cardiovascular parameters were compared using one-way anova with Bonferroni's correction for multiple comparisons. Frequency-response curves to nerve stimulation, dose-response curves to i.v. ACh, i.v. tiotropium or i.v. lactose were compared using two-way anova for repeated measures with Bonferroni's correction. Bronchoalveoloar lavage leukocyte and tissue eosinophil data were log-transformed to equalize variances and then analysed using one-way anova with Bonferroni's correction. A Kruskal–Wallis test was used to compare lymphoctye means in Figure 8 because variances could not be equalized. Significant P-values are reported as *P < 0.05, **P < 0.01 and ***P < 0.001. Statistical data were analysed with GraphPad Prism (version 5, GraphPad Software, Inc., La Jolla, CA, USA).

    Details are in the caption following the image

    Neutrophils in bronchoalveolar lavage remain increased 48 h after powder insufflation. Inflammatory cells were counted in bronchoalveolar lavage of sensitized guinea-pigs and sensitized and challenged guinea-pigs. Neutrophils increased following pretreatment with lactose vehicle or 1 µg·kg−1 i.s. tiotropium relative to sensitized controls that received no treatment. Antigen challenge did not increase neutrophils further in animals that received no pretreatment, lactose, tiotropium, or atropine (1 mg·kg−1, i.p.) pretreatment. Antigen challenge did increase eosinophils, and this was significant in animals pretreated with lactose vehicle. Values are means ± SEM, n= 5–7. Mean leukocyte numbers in sensitized animals pretreated with lactose or tiotropium were compared with sensitized animals that received no treatment; significant changes are noted with *. Each antigen-challenged group was also compared with the appropriate unchallenged control group; significance is noted with †.

    Results

    Effect of M3 muscarinic receptor blockade at the time of antigen challenge on subsequent airway hyperreactivity

    To test the effect of selectively blocking M3 muscarinic receptors during antigen challenge on the development of airway hyperreactivity, tiotropium bromide was i.s. into the lungs of ovalbumin-sensitized guinea-pigs. Tiotropium (1 µg·kg−1) was administered 24 h before antigen challenge to allow it to dissociate from M1 and M2 receptors, while still blocking M3 muscarinic receptors at the time of antigen challenge. In separate experiments, atropine (1 mg·kg−1, i.p.) was used to block all muscarinic receptor subtypes during antigen challenge and during the early response post-antigen challenge. Since atropine is short acting (elimination t½= 2.4 h (Kentala et al., 1990)), it was administered 1 h before and 6 h after challenge. All physiological measurements were made 24 h after antigen challenge.

    Antigen challenge did not change baseline pulmonary inflation pressure, heart rate or blood pressure measured in vagotomized guinea-pigs 24 h later relative to sensitized guinea-pigs (Table 1). Similarly, baseline pulmonary and cardiac parameters were not changed in those sensitized or sensitized and challenged animals pretreated with tiotropium, lactose vehicle or atropine (Table 1). Tiotropium, however, caused a small increase in baseline pulmonary inflation pressure in sensitized guinea-pigs.

    Table 1. Baseline pulmonary and cardiovascular parameters for data in vagotomized guinea-pigs for 1, 2, 6, 8, 9
    Pretreatment n Pulmonary inflation pressure (mmH2O) Heart rate (beats·min−1) Blood pressure (mmHg)
    Systolic Diastolic
    Sensitized
     None 6 87 ± 4 291 ± 3 51 ± 4 29 ± 4
     Lactose 6 103 ± 6 291 ± 10 49 ± 0.4 28 ± 2
     Tiotropium 7 107 ± 5* 284 ± 8 44 ± 5 26 ± 3
    Sensitized and challenged (24 h after challenge of sensitized guinea pigs)
     None 7 106 ± 2 278 ± 9 54 ± 2 29 ± 1
     Lactose 6 98 ± 4 299 ± 13 50 ± 6 32 ± 5
     Tiotropium 6 90 ± 3 293 ± 8 51 ± 2 30 ± 3
     Atropine 7 97 ± 5 293 ± 10 49 ± 2 28 ± 2
    • Values are means ± SEM. n, number of guinea-pigs. All baseline parameter means were compared with the sensitized no treatment group; statistically significant changes are noted with *.

    Antigen challenge of sensitized guinea-pigs caused airway hyperreactivity that was mediated by the vagus nerves (Figure 1). Electrical stimulation of both vagus nerves induced frequency-dependent bronchoconstriction in vagotomized, sensitized guinea-pigs (Figure 1A) that was significantly potentiated 24 h after antigen challenge (Figure 1B). Pretreatment with lactose vehicle did not change this vagally induced bronchoconstriction in either sensitized (Figure 1A) or sensitized and challenged (Figure 1B) guinea-pigs. Pretreatment with tiotropium did not reduce vagally-induced bronchoconstriction in sensitized guinea-pigs (Figure 1A). However, tiotropium pretreatment completely prevented the development of airway hyperreactivity in sensitized and challenged guinea-pigs (Figure 1B). In contrast, atropine pretreatment only partially inhibited airway hyperreactivity following antigen challenge (Figure 1B). Thus, tiotropium pretreatment prevented subsequent vagally mediated airway hyperreactivity in antigen-challenged animals without reducing bronchoconstriction in sensitized control animals.

    Details are in the caption following the image

    Selectively blocking M3 receptors with tiotropium at the time of antigen challenge prevents development of subsequent airway hyperreactivity but does not inhibit bronchoconstriction. Electrical stimulation of both vagus nerves (2–25 Hz, 10 V, 0.2 ms, 5 s pulse train) produced frequency-dependent bronchoconstrictions (measured as an increase in pulmonary inflation pressure) in vagotomized, sensitized guinea-pigs (A, open circles; also shown in B as a dashed line) that were potentiated 24 h after antigen challenge (B, solid circles). Lactose vehicle did not change vagally induced bronchoconstriction 48 h after administration in either sensitized (A, squares) or sensitized and challenged guinea pigs (B, squares). Tiotropium (1 µg·kg−1, i.s.) did not change vagally-induced bronchoconstriction in sensitized guinea pigs (A) measured 48 h after administration but did prevent potentiation of vagally-induced bronchoconstriction in sensitized and challenged guinea-pigs (B). Unlike tiotropium, atropine (1 mg·kg−1, i.p.) administered 1 h before and 6 h after challenge only partially prevented the potentiation of vagally-induced bronchoconstriction in sensitized and challenged guinea-pigs (B). Data are expressed as means ± SEM, n= 6–7. All frequency response curves were compared with the sensitized, no treatment group (open circles) and statistically significant changes are noted with *.

    ACh released from parasympathetic nerves in the lungs stimulates M3 muscarinic receptors on airway smooth muscle to induce bronchoconstriction. Smooth muscle M3 receptor function was measured with exogenous i.v. ACh. All of these animals were vagotomized to remove the confounding effects of reflex-induced bronchoconstriction. Antigen challenge did not increase i.v. ACh-induced bronchoconstriction (filled circles, Figure 2B) relative to sensitized controls (open circles, Figure 2A). Neither i.s. lactose nor atropine pretreatment had any affect on i.v. ACh-induced bronchoconstriction (Figure 2). Tiotropium inhibited i.v. ACh-induced bronchoconstriction in both sensitized (Figure 2A) and sensitized and challenged (Figure 2B) guinea-pigs 48 h after administration. Thus, some airway smooth muscle M3 receptors are still blocked 48 h after tiotropium treatment, while no M3 receptors are blocked 18 h after atropine treatment.

    Details are in the caption following the image

    Tiotropium still blocks some M3 muscarinic receptors on airway smooth muscle 48 h after administration in both sensitized (A) and sensitized and challenged (B) guinea-pigs. In vagotomized, sensitized guinea-pigs, i.v. ACh caused dose-dependent bronchoconstriction measured as an increase in pulmonary inflation pressure (A; also shown in B as a dashed line) that was not changed by antigen challenge (B). Bronchoconstriction was not changed by lactose vehicle 48 h after administration (A and B). Atropine (1 mg·kg−1, i.p.) administered 1 h before and 6 h after antigen challenge also did not change i.v. ACh-induced bronchoconstriction 24 h after challenge (B). In contrast, tiotropium (1 µg·kg−1, i.s.) inhibited i.v. ACh-induced bronchoconstriction in both sensitized (A) and sensitized and challenged (B) guinea-pigs when measured 48 h after administration. Data are expressed as means ± SEM, n= 6–7. All dose-response curves were compared with the sensitized, no treatment group (open circles). P-values approaching significance are reported following Bonferroni correction for six comparisons.

    M3 receptor blockade at the time of antigen challenge

    Since vagally-induced bronchoconstriction was not blocked in sensitized guinea-pigs 48 h after tiotropium treatment (Figure 1A), we tested whether smooth muscle M3 receptors were blocked at the time of antigen challenge. Either lactose (vehicle control) or one of two different doses of tiotropium (0.2 µg·kg−1 or 1 µg·kg−1) was i.s. into the lungs of ovalbumin-sensitized guinea-pigs, and physiological measurements were made 24 h later. Baseline pulmonary inflation pressure, heart rate and blood pressure were not significantly changed by either dose of tiotropium or by lactose vehicle relative to sensitized guinea-pigs that received no additional treatment (Table 2).

    Table 2. Baseline pulmonary and cardiovascular parameters in vagotomized guinea-pigs for data in 3, 5, 7
    Pretreatment n Pulmonary inflation pressure (mmH2O) Heart rate (beats·min-1) Blood pressure (mmHg)
    Systolic Diastolic
    Sensitized
     None 3 107 ± 9 315 ± 10 41 ± 7 19 ± 2
     Lactose 3 100 ± 0 317 ± 16 41 ± 4 26 ± 3
     Tiotropium, 0.2 µg·kg−1 4 108 ± 6 305 ± 9 43 ± 3 26 ± 5
     Tiotropium, 1 µg·kg−1 4 125 ± 16 293 ± 4 43 ± 2 24 ± 1
    • Values are means ± SEM. n, number of guinea pigs. None of the baseline parameter means were statistically different from the sensitized, no treatment group.

    Bronchoconstriction induced by i.v. ACh in sensitized guinea-pigs was not changed by pretreatment with i.s. lactose vehicle (Figure 3A). However, i.v. ACh-induced bronchoconstriction was inhibited in a dose-dependent manner in sensitized animals pretreated with tiotropium (Figure 3A). The higher dose of tiotropium significantly decreased bronchoconstriction, by approximately 75%, demonstrating that 1 µg·kg−1 tiotropium still blocked M3 muscarinic receptors on airway smooth muscle 24 h after administration, which was the time of antigen challenge in 1, 2.

    Details are in the caption following the image

    Tiotropium blocks smooth muscle M3 receptors 24 h after administration when tested with i.v. ACh but not vagally released ACh. In vagotomized, sensitized guinea-pigs, i.v. ACh caused dose-dependent bronchoconstriction measured as an increase in pulmonary inflation pressure (A), and electrical stimulation of both vagus nerves caused frequency-dependent bronchoconstriction (B; 2–25 Hz, 10 V, 0.2 ms, 5 s pulse train). Bronchoconstriction was not changed by lactose vehicle 24 h after administration (A and B). Tiotropium, 0.2 µg·kg−1 i.s. and 1 µg·kg−1 i.s., inhibited bronchoconstriction induced by i.v. ACh (A) but not bronchoconstriction induced by vagal stimulation (B) 24 h after administration. Data are expressed as means ± SEM, n= 3–4. All dose- and frequency-response curves were compared with the no treatment group (open circles); statistically significant changes are noted with *.

    Electrical stimulation of both vagus nerves caused frequency-dependent bronchoconstriction that was not inhibited by lactose vehicle in sensitized guinea-pigs (Figure 3B). Surprisingly, and in contrast to i.v. ACh, neither dose of tiotropium inhibited bronchoconstriction induced by electrical stimulation of both vagus nerves in sensitized guinea-pigs (Figure 3B). Since vagally-induced bronchoconstriction could be blocked with i.v. atropine (data not shown), it must have been mediated by ACh release onto muscarinic receptors. These data suggest that i.s. tiotropium did not block M3 receptors at the neuromuscular junction with parasympathetic nerves.

    Determining whether the route of tiotropium administration affects blockade of vagally induced bronchoconstriction

    It is possible that i.s. tiotropium was unable to inhibit bronchoconstriction induced by electrical stimulation of the vagus nerves (Figure 3B) because the delivery method (insufflation) prevented tiotropium from distributing to M3 muscarinic receptors at the junction with parasympathetic nerves. Therefore, we tested whether i.v. administration of tiotropium blocked equally M3 muscarinic receptors stimulated by i.v. ACh and vagally released ACh. Reproducible baseline bronchoconstrictions to i.v. ACh (4 µg·kg−1) and electrical stimulation of the vagus nerves (10 Hz) were measured in vagotomized, non-sensitized guinea-pigs treated with gallamine to block neuronal M2 muscarinic receptors. Cumulative doses of i.v. tiotropium inhibited bronchoconstriction in a dose-dependent manner (Figure 4). The lowest dose of tiotropium (0.3 µg·kg−1) significantly inhibited i.v. ACh-induced bronchoconstriction by more than 50%, whereas only the highest dose of tiotropium (10 µg·kg−1) significantly inhibited vagally induced bronchoconstriction (relative to the appropriate i.v. lactose controls). At a minimum, i.v. tiotropium was 10-fold more potent at inhibiting i.v. ACh-induced bronchoconstriction than vagally-induced bronchoconstriction. Thus, lower doses of i.v. tiotropium were required to block M3 muscarinic receptors involved in exogenous ACh-induced bronchoconstriction than were required to block endogenous ACh-induced bronchoconstriction.

    Details are in the caption following the image

    Intravenous administration of tiotropium blocks bronchoconstriction induced by i.v. ACh at lower doses than are required to block bronchoconstriction induced by electrically stimulating the vagus nerves. Bronchoconstriction, induced by either electrically stimulating both vagus nerves (10 Hz, 10 V, 0.2 ms, 5 s pulse train, at 2 min intervals) or by i.v. ACh (4 µg·kg−1, every 6 min between vagal stimulations), was measured in vagotomized guinea-pigs that were not sensitized or challenged. Bronchoconstriction before i.v. drug administration is shown on the left as an increase in pulmonary inflation pressure in mmH2O (A), and on the right as a percentage of this bronchoconstriction (B). All animals received gallamine (3 mg·kg−1, i.v.) to block neuronal M2 muscarinic receptors. Cumulative doses of tiotropium inhibited bronchoconstriction in a dose-dependent manner (solid lines, B), but it was significantly more effective against i.v. ACh-induced bronchoconstriction than it was against vagally induced bronchoconstriction. Lactose (dashed lines, B) was administered as a control and concentrations are shown on the upper axis. Bronchoconstriction was also inhibited with lactose, although the effect was not significantly different between i.v. ACh-induced bronchoconstriction and vagally-induced bronchoconstriction (B). Data are expressed as means ± SEM, n= 3. Inhibition of i.v. ACh-induced bronchoconstriction was compared with inhibition of vagally-induced bronchoconstriction for each treatment (lactose and tiotropium); significance is noted with †. Inhibition by tiotropium of bronchoconstriction induced by vagal stimulation or i.v. ACh was also compared at each dose to inhibition in the lactose control; significance is noted with *.

    For these experiments, lactose (filler in Spiriva) was used as a control and was given to guinea-pigs in separate experiments i.v. (dashed lines, Figure 4). We found no difference between the effect of lactose on i.v. ACh-induced bronchoconstriction and vagally induced bronchoconstriction. Yet we did find that overall bronchoconstriction appears to be inhibited in the lactose controls. Whether this inhibition is due to time, since the experiments lasted close to 2 h, or to lactose concentration was not tested. Regardless, the inhibitory effect of tiotropium was substantially greater than lactose.

    Determining whether tiotropium blocks cardiac M2 receptors

    In the heart, ACh stimulates M2 muscarinic receptors to cause bradycardia. Neither i.s. tiotropium nor i.s. lactose vehicle affected bradycardia induced by i.v. ACh or by electrical stimulation of the vagus nerves relative to untreated sensitized controls 24 h after insufflation into the lungs (Figure 5), which would be at the time of antigen challenge. These data indicate that cardiac M2 receptors were not blocked by tiotropium at the time of antigen challenge.

    Details are in the caption following the image

    Tiotropium does not block M2 muscarinic receptors in the heart 24 h after administration. In vagotomized, sensitized guinea-pigs, i.v. ACh caused dose-dependent bradycardia measured as a fall in heart rate in beats·min−1 (A). Electrical stimulation of both vagus nerves caused frequency-dependent bradycardia (B; 2–25 Hz, 10 V, 0.2 ms, 5 s pulse train). Neither lactose vehicle, 0.2 µg·kg−1 i.s. tiotropium, nor 1 µg·kg−1 i.s. tiotropium affected bradycardia induced by i.v. ACh (A) or vagal stimulation (B) 24 h after administration. Data are expressed as means ± SEM, n= 3–4. None of the dose- and frequency-response curves were statistically different from the no treatment group (open circles).

    Similarly, in the antigen challenge model, bradycardia induced by either i.v. ACh or electrical stimulation of the vagus nerves in sensitized (Figure 6, A and C) and sensitized and challenged (Figure 6, B and D) guinea-pigs was unchanged by antigen challenge, lactose pretreatment, tiotropium pretreatment or atropine pretreatment (Figure 6). Thus, cardiac M2 muscarinic receptors were not blocked by tiotropium or atropine at the time physiological measurements were made (48 h after tiotropium and 18 h after atropine administration).

    Details are in the caption following the image

    Lactose, tiotropium and atropine pretreatment each had no affect on bradycardia in sensitized or sensitized and challenged guinea-pigs. In vagotomized, sensitized guinea-pigs, i.v. ACh caused dose-dependent bradycardia (A) and electrical stimulation of both vagus nerves (2–25 Hz, 10 V, 0.2 ms, 5 s pulse train) caused frequency-dependent bradycardia (C), both measured as a fall in heart rate in beats·min−1. Antigen challenge did not change bradycardia (B and D). Neither lactose nor tiotropium (1 µg·kg−1, i.s.) changed bradycardia measured 48 h after administration in sensitized (A and C) or sensitized and challenged guinea-pigs (B and D). Similarly, atropine (1 mg·kg−1, i.p.) administered 1 h before and 6 h after antigen challenge also did not change bradycardia in sensitized and challenged animals. Data are expressed as means ± SEM, n= 6–7. None of the dose- and frequency-response curves were statistically different from the sensitized, no treatment controls (open circles).

    Effect of M3 receptor blockade during antigen challenge on airway inflammation

    We tested whether tiotropium pretreatment had any effect on airway inflammation. Insufflating powder into the lungs, regardless of whether it contained tiotropium, increased total bronchoalveolar cells (Figure 7). Both 24 h (Figure 7) and 48 h (Figure 8) after lactose or tiotropium powder administration, neutrophils were significantly increased relative to untreated sensitized controls. Antigen challenge did not further increase this neutrophilic inflammation (Figure 8).

    Details are in the caption following the image

    Neutrophils in bronchoalveolar lavage were increased 24 h after insufflation of lactose powder regardless of whether it also contained tiotropium. Inflammatory cells were counted in bronchoalveolar lavage of sensitized guinea-pigs. Lactose vehicle, 0.2 µg·kg−1 i.s. tiotropium, and 1 µg·kg−1 i.s. tiotropium all increased total and neutrophil cell numbers in sensitized guinea-pigs. Data are expressed as means ± SEM, n= 3–4. Mean leukocyte numbers were compared with the no treatment group; statistically significant changes are noted with *.

    Antigen challenge increased eosinophils in bronchoalveolar lavage of sensitized guinea-pigs regardless of whether they received no treatment, lactose, tiotropium or atropine pretreatment. However, this was only statistically significant following correction for multiple comparisons in animals treated with lactose 24 h prior to challenge (Figure 8).

    In lung tissue, quantitative analysis of eosinophils in airway smooth muscle and adventitia, and around airway nerves demonstrated that antigen challenge of sensitized guinea-pigs increased total and nerve-associated eosinophils in lactose (vehicle)-pretreated animals (Figure 9). In tiotropium-pretreated guinea-pigs, antigen challenge did not significantly increase either total eosinophils or nerve-associated eosinophils (Figure 9). Thus, tiotropium inhibited eosinophilic inflammation following antigen challenge in airway tissues. In the absence of lactose, there was a similar increase in total airway eosinophils from 264 ± 33 eosinophils·mm−2 in sensitized guinea-pigs to 437 ± 73 eosinophils·mm−2 in sensitized and challenged guinea-pigs. Additionally, eosinophils specifically associated with airway nerves were also increased from 67 ± 13 eosinophils mm−2 in sensitized guinea-pigs to 134 ± 27 eosinophils·mm−2 in sensitized and challenged animals. In the presence of atropine, there were 379 ± 48 total airway eosinophils mm−2 and 126 ± 28 eosinophils·mm−2 associated with nerves in sensitized and challenged guinea-pigs.

    Details are in the caption following the image

    Tiotropium reduces total and nerve-associated eosinophils in the lungs of antigen-challenged guinea-pigs. Total (A) and nerve-associated (B, within 8 µm of a nerve) eosinophils were counted per square millimetre airway smooth muscle and adventitia in tissue sections. Nerves were labelled with antibody to protein gene product 9.5, and eosinophils were stained with chromotrope 2R. Eosinophils are found in the lungs and around nerves in sensitized guinea-pigs pretreated with lactose vehicle or tiotropium (1 µg·kg−1, i.s.). Antigen challenge increased total and nerve-associated eosinophils in sensitized guinea-pigs pretreated with lactose. In contrast, in sensitized guinea-pigs that received tiotropium 24 h prior to challenge, total and nerve-associated eosinophils were not significantly increased following antigen challenge. Data are expressed as means ± SEM, n= 6–7. Comparisons were made between all groups; statistically significant changes are noted with *.

    Discussion

    The data presented here demonstrate that selectively blocking M3 muscarinic receptors during antigen challenge prevents subsequent development of airway hyperreactivity in guinea-pigs. Muscarinic receptor antagonists with 100-fold higher affinity for M3 receptors over other muscarinic receptor subtypes are not readily available; therefore, we used tiotropium, a clinically used (GOLD, 2010; Peters et al., 2010; Vogelmeier et al., 2011), kinetically selective, M3 receptor antagonist. Tiotropium was administered 24 h before antigen challenge to take advantage of its slow dissociation from M3 receptors relative to M2 and M1 receptors (Disse et al., 1993; Casarosa et al., 2009). Functional studies in guinea-pig trachea show M3 receptor function begins to return 9 h after tiotropium washout, while M2 receptor function returns fully within 2 h (Takahashi et al., 1994). In our study, we found no evidence for pharmacological blockade of M2 muscarinic receptors 24 h after tiotropium administration since i.v. ACh-induced bradycardia was not reduced. However, as previously shown for guinea-pigs (Villetti et al., 2006) tiotropium still blocked M3 receptors on airway smooth muscle when tested with i.v. ACh.

    Baseline heart rate was not changed with antigen challenge or pretreatments. However, there was a significant 20 mmH2O increase in baseline pulmonary inflation pressure in sensitized guinea-pigs pretreated with tiotropium. This increase is unlikely to be physiologically significant, since greater increases (150 mmH2O) following ozone do not suppress bronchoconstriction (Verhein et al., 2008). We were unable to determine whether tiotropium pretreatment reduced vagal tone in our animals, since baseline pulmonary inflation pressure was measured after vagotomy.

    We have demonstrated that airway hyperreactivity develops within 24 h of antigen challenge in sensitized guinea-pigs (Evans et al., 1997, 2001; Fryer et al., 2006; Verbout et al., 2007). This was confirmed in these experiments where bronchoconstriction induced by electrical stimulation of both vagus nerves was significantly increased in sensitized and challenged guinea-pigs relative to sensitized controls. Airway hyperreactivity was mediated by the vagus nerves, since airway smooth muscle contraction to i.v. ACh, which bypasses the nerves in vagotomized animals, was not similarly increased. Tiotropium pretreatment completely prevented hyperreactivity in antigen-challenged guinea-pigs, but did not block smooth muscle M3 muscarinic receptors innervated by the vagus nerves. This suggests that tiotropium prevents airway hyperreactivity in antigen-challenged guinea-pigs through a mechanism that is separate from inhibition of bronchoconstriction.

    Eosinophilic inflammation correlates with asthma severity (Bousquet et al., 1990). In patients with eosinophilic asthma, reducing lung eosinophils significantly decreases asthma exacerbations and allows decreased steroid use (Haldar et al., 2009; Nair et al., 2009). In antigen-challenged guinea-pigs, hyperreactivity is mediated by eosinophil recruitment to airway nerves and subsequent activation and release of eosinophil major basic protein (Costello et al., 1997; Evans et al., 1997, 2001; Verbout et al., 2007). Tiotropium reduced eosinophil accumulation in airway tissue and around airway nerves following antigen challenge. This could be a mechanism for preventing airway hyperreactivity, since airway hyperreactivity is also prevented by treatments that inhibit eosinophil association with nerves or inhibit deposition of eosinophil major basic protein on nerves (Evans et al., 1997, 2001; Fryer et al., 2006; Nie et al., 2009). It is noteworthy that bronchoalveolar neutrophils were increased by powder insufflation into the lungs. However, increased neutrophils were not associated with hyperreactivity, so the presence of neutrophils alone is not sufficient to cause airway hyperreactivity.

    Tiotropium has already been shown to reduce airway remodelling, including smooth muscle thickening, smooth muscle hypercontractility and mucous gland hypertrophy. It also reduces Th2 cytokines and lung eosinophils associated with chronic antigen challenge (Gosens et al., 2005; Bos et al., 2007; Ohta et al., 2010). Furthermore, tiotropium's anti-inflammatory effects are not limited to antigen challenge since tiotropium also reduces cytokine production and neutrophil accumulation following cigarette smoke exposure in mice (Wollin and Pieper, 2010). Together, these studies support our data that tiotropium may prevent vagally mediated airway hyperreactivity by reducing inflammation.

    Since smooth muscle M3 receptors innervated by the nerves were not blocked at the time of antigen challenge, it is possible that tiotropium prevented airway hyperreactivity by blocking ACh released from non-neuronal sources such as epithelial cells or macrophages (Wessler and Kirkpatrick, 2008). Tiotropium could reduce eosinophil trafficking into tissue by blocking M3 receptors on airway epithelial cells or alveolar macrophages that normally induce release of eosinophil chemoattractants (Koyama et al., 1998; Sato et al., 1998; Buhling et al., 2007). Tiotropium could also reduce eosinophil accumulation by blocking M3 muscarinic receptors on eosinophils (Verbout et al., 2006). But, the role of M3 muscarinic receptors on eosinophils is unknown. In addition, tiotropium could block M3 receptors on smooth muscle cells not innervated by nerves to inhibit pro-inflammatory cytokine release (Gosens et al., 2009) following antigen challenge.

    In contrast to selectively blocking M3 receptors, blockade of all muscarinic receptors with atropine during antigen challenge was less effective in preventing airway hyperreactivity, suggesting that blockade of all muscarinic receptors may counteract the benefit of blocking M3 receptors. For example, M1 muscarinic receptors on human airway mast cells inhibit evoked histamine release (Reinheimer et al., 2000), and it may be important to maintain the activity of these inhibitory muscarinic receptors. Alternatively, differences in atropine and tiotropium kinetics might explain differences in prevention of airway hyperreactivity, since tiotropium, but not atropine, blocked some smooth muscle M3 receptors when airway responsiveness was measured. However, this seems unlikely because neither drug blocked smooth muscle M3 receptors innervated by vagal nerves (Figure 1A and Verbout et al., 2007).

    We have previously shown that atropine at the time of antigen challenge makes airway hyperreactivity significantly worse (Verbout et al., 2007, 2009). Although, we were unable to see similar potentiation in this study, atropine was not completely protective either. Thus, tiotropium, but not the non-selective atropine, prevented development of hyperreactivity. Lack of atropine-induced hyperreactivity in our study may be explained by the additional use of antihistamines and isoprenaline (β-agonist) in the previous studies, while neither drug was used here.

    During the course of these studies, we also observed that i.s. tiotropium blocked bronchoconstriction induced by i.v. ACh but did not block bronchoconstriction induced by vagally released ACh in sensitized guinea- pigs. These data are consistent with previous studies using atropine in humans with asthma and in anaesthetized dogs (Sheppard et al., 1982, 1983; Holtzman et al., 1983) that demonstrate there are several variables contributing to the efficacy of a competitive muscarinic antagonist. These include dose of antagonist, route of antagonist administration and regional differences in agonist concentration. In these previously published studies, low doses of inhaled atropine blocked bronchoconstriction induced by inhaled ACh (or methacholine) and also maximally decreased baseline vagally mediated cholinergic tone. However, much higher doses of atropine (8- to 16-fold) were required to block bronchoconstriction induced by either indirect (cold air) or direct (electrical) stimulation of the vagus nerves (Sheppard et al., 1982; Holtzman et al., 1983). Thus, similar to atropine, we show that doses of i.s. tiotropium that block exogenous ACh do not block vagally-induced bronchoconstriction. This is important because our data and these papers confirm that blockade of exogenous ACh-induced bronchoconstriction is not an adequate measure of blockade of ACh from its physiological source, the vagus nerves. Furthermore, these data demonstrate that inhibition of vagally-induced bronchoconstriction is not necessary to prevent subsequent development of airway hyperreactivity.

    The difference in atropine's ability to inhibit bronchoconstriction induced by inhaled ACh and vagally-released ACh only occurs when atropine is administered by inhalation (Holtzman et al., 1983; Sheppard et al., 1983). Intravenous atropine distributes evenly to all airway smooth muscle muscarinic receptors and equally blocks bronchoconstriction induced by exogenous and endogenous ACh (Holtzman et al., 1983). In contrast, in our study, i.v. tiotropium did not equally block exogenous ACh and vagally-induced bronchoconstriction; significantly higher doses were required to block vagally-induced bronchoconstriction. These data indicate that the dose of tiotropium is critical for bronchodilatation, independent of delivery route.

    Tiotropium is a competitive antagonist (Casarosa et al., 2009), but its slow dissociation kinetics make it functionally irreversible at M3 receptors. Our data suggest a model where tonic neuronal stimulation maintains high local concentrations of ACh at smooth muscle M3 receptors, while M3 receptors further away from the nerves are exposed to lower endogenous ACh concentrations. Tiotropium would have to compete with ACh for receptor sites, and this would require more tiotropium around nerves where agonist concentrations are high. Thus, even though i.s. tiotropium blocked bronchoconstriction induced by exogenous agonists, it would require a higher dose to block vagally-induced bronchoconstriction.

    In summary, the data presented here demonstrate that selectively blocking M3 muscarinic receptors with tiotropium at the time of antigen challenge prevents subsequent vagally-mediated airway hyperreactivity. Importantly, airway hyperreactivity is prevented by a dose of tiotropium that is unable to inhibit vagally-induced bronchoconstriction. Furthermore, our data suggest that the mechanism may be anti-inflammatory, and that selective blockade of airway M3 muscarinic receptors may be an effective treatment for asthma separate from bronchodilatation.

    Acknowledgements

    We would like to thank Malcolm Johnson for helpful insight and advice related to the research in this manuscript. This work was supported by the US NIH [NIH HL61013 (DBJ), HL71795 (DBJ), AI75064 (DBJ), HL055543 (ADF), ES014601 (ADF), ES017592 (ADF), and T32HL83808 (KSB)]; an Oregon Health and Science University Tartar Trust Research Fellowship (KSB); an Achievement Awards for College Scientists Foundation scholarship (KSB) and a grant from GlaxoSmithKline (DBJ).

      Conflicts of interest

      This research was partially funded by a grant from GlaxoSmithKline.