Airway smooth muscle cells (ASMC) mediate bronchial narrowing and therefore play a key role in bronchial hyperresponsiveness (BHR). In asthma, increased ASMC mass has been reported by several investigators 123. However, the question arises whether changes in contractile properties of the ASMC 4 or in the signaling cascades that mediate agonist-induced contraction might also contribute to asthma 5. Calcium is a ubiquitous signaling molecule that is involved in the regulation of a broad variety of cellular events in almost all mammalian cell types 6789. In ASMC, contractile stimuli promote contraction by elevating cytoplasmic calcium-concentration ([Ca²⁺]_c). Calcium binds to calmodulin, thereby activating myosin light chain (MLC) kinase, which phosphorylates MLC and initiates actomyosin cross-bridge cycling. Consequently, agonist-induced Ca²⁺-signaling has been regarded as a possible target in the pathophysiology of BHR 1011. Indeed, in hyperresponsive inbred rats enhanced Ca²⁺-mobilization was correlated with BHR 1213, though, on the other hand, differences in acetylcholine (ACH)-induced airway narrowing were not associated with differences in ACH-induced Ca²⁺-signaling when we compared three different mouse strains 14. However, several studies have shown that Ca²⁺-responses can be modulated by a multitude of stimuli including the cytokines IL-1β, TNF-α, IL-4 and IL-13 1516171819202122. In addition, CD38/cyclic ADP-ribose-mediated Ca²⁺-signaling has been found to contribute to ASMC hyperresponsiveness 2324. But, in all of these studies BHR was investigated either by comparing innate properties of inbred strains or by using single mediators known to be involved in asthma.

T-bet is a T_H1-specific transcription factor, which has the ability to direct T_H2- into T_H1-cells 25. Naïve mice that have been target-deleted of the T-bet gene (T-bet KO-mice) spontaneously develop airway remodeling very similar to that seen in asthma and demonstrate multiple functional and inflammatory features characteristic of this disease 26. Recently, we developed a system consisting of thin lung slices in combination with confocal microscopy, which enables the analysis of ASMC Ca²⁺-signaling in an environment closely resembling the in vivo situation 2728. In the present study, we applied this system to T-bet KO-mice as a complex and multi-faceted asthma model to study alterations of Ca²⁺-homeostasis in ASMC. As a result, we found that lung slices from T-bet KO-mice preserved the key characteristics of the living animal in terms of increased baseline airway tone (BAT) and BHR. Increased BAT was correlated with an increased incidence of spontaneous changes in [Ca²⁺]_c, whereas increased BHR correlated with elevated acetylcholine (ACH)-induced Ca²⁺-transients and a higher proportion of ASMC showing Ca²⁺-oscillations. Emptying intracellular Ca²⁺-stores induced higher Ca²⁺-elevations in ASMC from T-bet KO- compared to wild-type mice. We therefore propose that a higher Ca²⁺-content of the intracellular Ca²⁺-stores is involved in the pathophysiology of these changes.

Cell culture reagents were obtained from Life Technologies (Eggenstein, Germany). Other reagents were bought from Sigma-Aldrich (Deisenhofen, Germany) unless stated otherwise. Balb/C mice were purchased from Harlan-Winkelmann (Borchen, Germany) and T-bet KO-mice on a Balb/C background from Charles River (Charles River Breeding Labs, Needham, MA). All procedures had been approved by the Ethics Committee of the Ludwig-Maximilians-University Munich.

Lung slices were prepared as described previously 27. Briefly, after sacrificing the mice, lungs were inflated with 2% agarose-sHBSS and the agarose was gelled by placing the mouse preparation at 4°C. For the use with video-microscopy, slices ~200 μm thick were cut with an EMS-4000 Tissue Slicer (Electron Microscopy Sciences, Fort Washington, PA). For better loading with Ca²⁺-indicator dyes, slices were cut ~100 μm thick for the use with two-photon microscopy. The slices were maintained by floating them in DMEM supplemented with Penstrep^® (2.4 ml/l, Penicillin 10.000 U/ml + Streptomycin 10.000 μg/ml) and Fungizone^® (4.8 ml/l, Amphotericin B 250 UG/ml) at 37°C in 5 % CO₂. Experiments were performed on day 2 to 4 of culture and each slice was used for one experiment only. For each group of experiments, slices from 5 to 6 different mice were used.

To measure airway cross-sectional area, lung slices were placed in culture dishes immersed in sHBSS and held in position by a piece of a nylon mesh. Phase-contrast images were recorded using a digital CCD camera (AxioCam MRm, Carl Zeiss Vision, Munich, Germany). Frames were captured in time-lapse (0.5 frame/s) and the cross-sectional area of the airway was measured by pixel summing using the image analysis software "Scion" (Scion Corporation, Frederick, Maryland). The mean cross-sectional area of all airways used was 39326 ± 9390 μm² without significant differences between T-bet KO- and wild-type mice. This size reflects an airway level below the segmental bronchi and above the respiratory bronchioles. Contraction velocity was defined as maximal change in cross-sectional area per second.

To determine [Ca²⁺]_c in ASMC, slices were loaded for 1 h at 37°C with 10^-5 M Fluo-4-AM (Molecular Probes, Eugene, OR) in sHBSS containing 0.2 % Pluronic (Pluronic F-127, Calbiochem, La Jolla, CA) and 10^-4 M sulfobromophthalein. Sulfobromophthalein was used to block unspecific ion pumps and cation channels thereby reducing the loss of de-esterified Fluo-4. After loading, slices were incubated for at least 30 min in sHBSS containing 10^-4 M sulfobromophthalein to allow for complete dye de-esterification. The bath solution for all experiments was sHBSS without sulfobromophthalein. The slices were placed in a custom made Plexiglas chamber and held in position by a platinum mesh. Detection of fluorescent signals of single ASMC within the slices was performed with a custom-built two-photon microscope based on an Olympus microscope (BX51WI, Olympus, Hamburg, Germany, for details see 2930). Briefly, the 790 nm laser lane of a Ti:Saphir femtosecond laser (Spectra Physics, Darmstadt, Germany) is scanned across the specimen with 2 oscillating mirrors (for X- and Y-scan). The density of photons is only in the focal plane high enough to ensure simultaneous excitation of fluorescent molecules by two photons. Out of the focal plane, the energy of one photon is not sufficient to excite fluorescent molecules and therefore, all emitted light originates from the focal plane. The resultant fluorescence (> 510 nm) is detected by a photomultiplier tube and an image is formed using the recording software "Video Savant" (Cosyco, Germering, Germany). Images were recorded in time lapse-mode (1 f/sec). In the case of ACH-induced Ca²⁺-plateaus, the acquisition rate was increased to 10 f/sec (sufficient to detect Ca²⁺-oscillations with a frequency of up to 500/min) to ensure the detection of potentially high frequency Ca²⁺-oscillations. Regions of interest of 10 × 10 pixels were defined in single ASMC excluding the nucleus and fluorescence intensities were analyzed frame by frame using custom written macros in the image analysis software "Scion". Final fluorescence values were expressed as fluorescence ratio (F/F_o) normalized to the fluorescence immediately prior to the addition of an agonist (F_o).

To measure the Ca²⁺-content of the intracellular Ca²⁺-stores, ASMC in lung slices were exposed to 10^-3 M caffeine (to open ryanodine-receptor Ca²⁺-channels) or to 10^-6 M cyclopiazonic acid (CPA, to inhibit SERCA-pumps), and the resulting increase in cytoplasmic Ca²⁺-concentration was measured. To prevent Ca²⁺-entry by store operated channels, the slices were placed in phosphate-buffered saline without calcium containing 0.02 % EDTA immediately prior to the experiments.

One-way or two-way ANOVA or ANOVA on ranks (combined with pairwise multiple comparisons) were performed using the "Sigma Stat" software (Jandel Scientific, Chicago, IL). A P value of less than 0.05 was considered statistically significant.

In this study, we investigated the Ca²⁺-homeostasis of ASMC in lung slices from T-bet KO-mice. Based on previous knowledge 26, these mice served as a genetically engineered animal model of asthma. Lung slices from T-bet KO-mice preserved the contractile characteristics of the living mice by showing increased BAT and BHR compared to wild-type mice. BHR could be mimicked incubating lung slices from wild-type mice with IL-13 indicating that the increased levels of IL-13 found in T-bet KO-mice might have mediated the changes in bronchial reactivity (see also 31). Increased BAT was related to an increased incidence of spontaneous changes in [Ca²⁺]_c, while BHR correlated with elevated ACH-induced Ca²⁺-transients and a higher proportion of ASMC showing Ca²⁺-oscillations. Baseline fluorescence values and the amplitude as well as the frequency of the ACH-induced Ca²⁺-oscillations were not different. The fact that caffeine and CPA induced higher Ca²⁺-elevations in ASMC from T-bet KO-mice suggests that a higher content of the intracellular Ca²⁺-stores in T-bet KO-mice contributed to the alterations in the Ca²⁺-homeostasis.

The major advantage of lung slices is that the in situ organization of the lung tissue and the contractility of the ASMC are maintained for several days. Lung slices in combination with video-microscopy have been used before to study airway responses 3233. Using conventional imaging techniques, observation of ASMC on a single cell-level is prohibited by the thickness of the slice. Therefore, we combined the lung slice technique with confocal microscopy 142728, which allows the observation of single cells in a slice by excluding out of focus light 30. In this study, we refined this technique by using a custom-build two-photon instead of a confocal microscope, which allowed to reduce photo-bleaching while increasing spatial resolution 29. With this approach, we were able to combine the advantages of a complex tissue culture system preserving many of the alterations observed in vivo with the ability to observe single cell properties, which is usually only rewarded by single cell preparations.

Similar to previous observations 26, we found the thickness of the ASMC-layer to be elevated in histological sections from T-bet KO-mice. Although an increased ASMC mass alone could have accounted for elevated airway contraction, the contraction velocity of airways from T-bet KO-mice was also increased and this finding pointed towards intrinsic alterations in the contractile properties of the ASMC. A thickened ASMC-layer probably affected the stiffness of the airway wall and therefore lung compliance. This may have contributed to the different contractile properties of airways in lung slices from T-bet KO- and wild-type mice. But, we believe that this is an advantage rather than a negative aspect of the lung slice preparation making it even closer to the in vivo situation.

Finotto et al. reported elevated levels of TH₂-cytokines especially IL-13 in the lavage of T-bet KO-mice 26 and recently, the same group showed that IL-13 mediated the asthmatic changes seen in these mice 31. Furthermore, IL-13 has been shown to increase the Ca²⁺-response to a variety of agonists 1922. In our study, incubating lung slices of wild-type mice with IL-13, we found increased contraction and contraction velocity in response to ACH closely resembling the changes seen in lung slices from T-bet KO-mice. To our knowledge, a direct link connecting the genetic alteration in the T-bet gene with BHR has not been established. We therefore propose that the shift from TH₁- to a TH₂-phenotype mediated by the loss of the T-bet gene function results in a predominance of TH₂-cytokines especially IL-13, which in turn mediates BHR.

ASMC from T-bet KO- and wild-type mice showed spontaneous changes in [Ca²⁺]_c. Depending on the frequency, these changes could be regarded as spontaneous Ca²⁺-oscillations or as spontaneous Ca²⁺-transients, although sometimes the differentiation between slow frequency Ca²⁺-oscillations and Ca²⁺-transients was arbitrary. The higher incidence of changes in [Ca²⁺]_c in T-bet KO-mice may have led to an increased BAT by either a frequency modulated process integrating Ca²⁺-oscillations or simply by increasing overall [Ca²⁺]_c.

In a previous study, we proposed that the initial Ca²⁺-transient evoked by ACH determines the level of contraction while the subsequent Ca²⁺-oscillations keep the airway in the narrowed state 27 and this idea has been supported by recent data 3435. In accordance with this hypothesis, the augmented Ca²⁺-transient in ASMC from T-bet KO-mice led to increased contraction. However, although the contraction was different, the frequencies of the ACH-induced Ca²⁺-oscillations were similar in T-bet KO- and wild-type mice. After the initial Ca²⁺-transient, the level of [Ca²⁺]_c has to be high enough to prevent dephosphorylation of MLC and thereby maintaining contraction. The latch state describes a situation, where MLC is dephosphorylated but nevertheless remains attached to actin, which results in a much slower relaxation rate 3. The present data suggest that once ASMC are in the latch state, comparable frequencies of Ca²⁺-oscillations or a sustained, adequately high steady-state Ca²⁺-plateau may be sufficient to maintain different levels of contraction set by different Ca²⁺-transients.

We found the percentage of ASMC showing ACH-induced Ca²⁺-oscillations to be higher in T-bet KO-mice compared to ASMC from wild-type mice. Even though the frequency of the Ca²⁺-oscillations was comparable, an enhanced recruitment of ASMC showing Ca²⁺-oscillations might have contributed to the maintenance of the higher contraction in ASMC from T-bet KO-mice.

Using two different pharmacological interventions, we found the Ca²⁺-content of intracellular Ca²⁺-pools of ASMC to be elevated in lung slices from T-bet KO- compared to wild-type mice. However, the question arises how an increased Ca²⁺-content of intracellular Ca²⁺-pools of ASMC may be related to the altered spontaneous and ACH-induced Ca²⁺-signaling found in T-bet KO-mice?

An increased Ca²⁺-load of the SR sensitizes both RyR and inositol-3-phosphate receptors (IP₃R), which leads to increased elemental Ca²⁺-events like Ca²⁺-sparks and Ca²⁺-puffs 8. The elevated rate of elemental Ca²⁺-events might result in an elevated rate of spontaneous changes in whole cell [Ca²⁺]_c as observed in ASMC from T-bet-KO mice.

RyR and IP₃R play a critical role in the induction of Ca²⁺-oscillations because these Ca²⁺-channels repetitively allow or inhibit Ca²⁺-release from the Ca²⁺-pools. Shifting the Ca²⁺-sensitivity of RyR and IP₃R by increasing the Ca²⁺-content of the Ca²⁺-pools could sensitize the Ca²⁺-signaling apparatus to allow a higher proportion of cells to display Ca²⁺-oscillations in response to ACH as seen in ASMC from T-bet KO-mice.

The ACH-induced initial Ca²⁺-transient has been shown to be caused by Ca²⁺-release from Ca²⁺-pools with Ca²⁺-influx across the cell membrane playing a minor role 27. Therefore, the increased Ca²⁺-transient in ASMC from T-bet KO-mice presumably was caused by the higher Ca²⁺-load of the Ca²⁺-pools. On the other hand, our results do not preclude the possibility that voltage dependent or independent Ca²⁺-influx contributed to the different Ca²⁺-signaling in T-bet KO-mice.

At the highest ACH concentrations used in our study, ASMC from T-bet KO-mice showed a sustained Ca²⁺-plateau rather than Ca²⁺-oscillations after the initial Ca²⁺-transient. A higher Ca²⁺-load of the Ca²⁺-pools results in a higher Ca²⁺-gradient between the Ca²⁺-pools and the cytoplasm. Ca²⁺-pumps force calcium against this gradient out of the cytoplasm into the Ca²⁺-pools. Therefore, a higher gradient would impair the reuptake of calcium into the Ca²⁺-pools. In the case of Ca²⁺-oscillations with high frequencies and hence short intervals between oscillations, the reduced Ca²⁺-reuptake in-between single oscillations could lead to a Ca²⁺-plateau rather than Ca²⁺-oscillations.

The machinery that establishes Ca²⁺-homeostasis in ASMC is complex and, therefore, it seems unlikely that alteration of a single component, e.g. the content of the Ca²⁺-pools, accounted for all the differences in Ca²⁺-signaling observed in T-bet KO-mice. But, as pointed out, it is plausible that the altered content of the Ca²⁺-pools at least contributed to the increased BAT and BHR of T-bet KO-mice.

A detailed analysis of the Ca²⁺-signaling in asthma seems particularly relevant when considering potential implications for clinical studies. As L-type Ca²⁺-channel-blockers such as nifedipine failed to provide benefits in patients with asthma 36, Ca²⁺-signaling as a therapeutic target has been neglected for a long time. Ca²⁺-influx across the cell membrane via L-type Ca²⁺-channels, however, appears to play a minor role in ASMC 2737 and this easily explains the failure of L-type Ca²⁺-channel-blockers. On the other hand, the key role of Ca²⁺-homeostasis in ASMC is progressively being acknowledged. We believe that the combination of T-bet KO-mice, lung slices and two-photon microscopy as presented in this study constitutes a promising tool to establish a deeper understanding of the mechanisms underlying the altered Ca²⁺-signaling. A deeper understanding of these mechanisms may lead to development of new drugs specifically targeting the mechanisms of altered Ca²⁺-homeostasis relevant in asthma.

Lung slices from T-bet KO-mice as an asthma model preserve the key characteristics of the living animal in terms of increased BAT and BHR. Increased BAT is correlated with an increased incidence of spontaneous changes in [Ca²⁺]_c, whereas increased BHR correlates with elevated ACH-induced Ca²⁺-transients and a higher proportion of ASMC showing Ca²⁺-oscillations. We propose that a higher Ca²⁺-content of the intracellular Ca²⁺-stores is involved in the pathophysiology of these changes. Pharmacological interventions targeting altered Ca²⁺-homeostasis may therefore constitute a new therapeutic tool in asthma.

The author(s) declare that they have no competing interests.

AB laid out the principal conception and design of the study, carried out experiments on Ca²⁺-signaling in lung slices and drafted the manuscript. JK performed studies on airway contraction in lung slices and contributed substantially to the analysis and interpretation of the data. AK also performed in vitro studies and contributed substantially to the analysis and interpretation of the data. FG revised the manuscript and gave valuable intellectual input in the course of the study. RMH participated in the study design and made substantial contributions to the interpretation of the data. All authors read and approved the final manuscript.