In the present study, our aim was to analyze molecularly the inhibitory effect of BDZs on MC activity by comparison of the effects of the two BDZs Ro5-4864 and clonazepam. The two drugs differ markedly in their affinities for the archetypical BDZ recognition sites, i.e., the GABAA receptor and the TSPO, previously termed peripheral-type BDZ receptor. Ro5-4864 is an agonist at TSPO and has only low affinity to the GABAA receptor , whereas clonazepam is a high-affinity GABAA receptor agonist, but has only low affinity for TSPO . Ro5-4864 concentration-dependently inhibited Ag-triggered degranulation in BMMCs and PMCs, whereas clonazepam was essentially ineffective in this respect. In accordance with this observation, Ro5-4864 suppressed Ca2+ mobilization, production of ROS and activation of the PI3K pathway (as measured by phosphorylation of Akt at Ser473), which are all important signaling events in the positive regulation of the secretory response [26, 27]. These data suggest that Ro5-4864 and structurally related compounds might be applicable as versatile MC stabilizing drugs in MC-dependent diseases, e.g., hypersensitivity diseases, asthma, and systemic MCAD [18, 40]. This was also shown by the inhibition of allergen-induced bronchoconstriction in rat precision-cut lung slices. In this context it is interesting to note that Ro5-4864 did not change the basal activation of the MCs indicating a selective action of BDZs at (pathologically) activated MCs.
The question arises for the target sites at which Ro5-4864 and other 1,4-benzodiazepines mediate their inhibitory effects on MCs. Since the selective GABAA receptor agonist clonazepam did not mimic the effects of Ro5-4864, an action of Ro5-4864 at classical GABAA receptors is rather unlikely. One potential candidate structure is TSPO, a transmembrane protein located in the OMM and enriched in OMM-IMM contact sites. It is a central component of a multimeric protein complex, comprising amongst others TSPO, VDAC1, ANT, and PRAX-1, and is associated with the mitochondrial permeability transition pore (mPTP) . Therefore, functions of TSPO in regulating apoptotic processes have been discussed. Indeed, Ro5-4864 has been reported to induce apoptosis in some human and murine cancer cell lines and thymocytes, in particular by interfering with the mitochondrial membrane potential [41–45]. In these studies, cells were treated with Ro5-4864 for many hours to days, i.e. for a much longer time span compared to our experiments, which were performed within minutes to 3 h. The analysis of the effect of Ro5-4864 on BMMC survival showed only subtle apoptotic effects after 24 h (data not shown) excluding apoptotic effects within the significant shorter time windows of our MC activation experiments. Since we were able to reduce the pre-incubation time with Ro5-4864 to 1 min without loosing inhibitory efficiency, a mechanism via plasma membrane-located target sites instead of mitochondria-resident TSPO seems more likely. Interestingly, in certain cell types TSPO expression has also been detected in the plasma membrane [21, 22]. By expressing a fluorescent TSPO fusion protein and using confocal microscopy, however, we were not able to detect the TSPO-containing fusion protein in the plasma membrane of MCs. Though we consider the technique used sufficiently sensitive, we did not have the material to detect endogenous TSPO and, thus, cannot totally exclude expression of the endogenous protein in the plasma membrane.
The concentration-dependent inhibition of Ag-triggered degranulation by Ro5-4864 could be due to suppression of mitochondrial Ca2+ uptake. Ag-triggered degranulation is dependent on influx of extracellular Ca2+ ions through SOC channels [46, 47]. We found that Ro5-4864 suppressed SOC entry, whereas intracellular Ca2+ release from the ER appeared unaltered. Recently, Farsky and colleagues reported on a similar attenuation of fMLP-induced Ca2+ mobilization in neutrophils by Ro5-4864 . Optimal SOC entry requires efficient emptying of the ER. To prevent immediate re-uptake of Ca2+ into the ER via the SERCA, mitochondria are able to take up Ca2+ at the moment of release from the ER by the uniporter channel. Ro5-4864 treatment might suppress this mitochondrial Ca2+ uptake mechanism. If so, passive release of Ca2+ from the ER by means of treatment with the SERCA inhibitor thapsigargin should be independent of such a mitochondrial buffering mechanism. Indeed, thapsigargin-induced SOC influx was not suppressed by Ro5-4864 treatment, which would be in agreement with this idea of an interaction of Ro5-4864 with mitochondrial Ca2+ uptake.
However, the effects of Ro5-4864 on Ag-triggered signaling in BMMCs deficient for the negative regulator SHIP1 point to another target site of Ro5-4864. Intriguingly, whereas Ro5-4864 concentration-dependently suppressed degranulation in SHIP1-deficient BMMCs, Ag-triggered activation of Akt was not altered in these cells. Moreover, there was only a slight reduction of extracellular Ca2+ influx (Additional file 3: Figure S3). It is known that SHIP1-deficient MCs are less sensitive to drugs inhibiting PI3K compared to wild-type MCs . Since Akt phosphorylation and Ca2+ mobilization are PI3K-dependent [26, 49] suppression of PI3K activation by Ro5-4864 was not as effective in SHIP1-deficient BMMCs. These data suggested that Ro5-4864 very likely did not interfere with the mitochondrial Ca2+ buffering mechanism (such an effect should be observable in SHIP1-deficient BMMCs as well) and that a target site located more upstream should be involved in Ro5-4864-mediated regulation of the secretory response.
1,4-Benzodiazepines have been reported to inhibit SFKs , which are known to play multiple important roles in MC activation, in particular via the FcεRI . Amongst the first signaling events in MCs in response to FcεRI crosslinking are activation of the SFK Lyn, subsequent tyrosine phosphorylation of the FcεRI β-chain and γ-chain ITAMs by Lyn and activation of the tyrosine kinase Syk via interaction with the phosphorylated γ-chains and phosphorylation by Lyn . Moreover, immediate activation of the SFK Fyn leads to the activation of the PI3K pathway . Thus, pharmacological interference with SFK activation would have a negative impact on most FcεRI-mediated signaling pathways. Indeed, Ro5-4864 attenuated Ag-triggered tyrosine phosphorylation events in both wild-type and SHIP1-deficient MCs. Using a GST-SH2(Lyn) fusion protein to pull-down specific tyrosine-phosphorylated proteins, reduced phosphorylation of FcεRIβ and Syk was observed, indicating early interference of FcεRI signaling by Ro5-4864. Though both FcεRIβ and Syk are known targets of Lyn, involvement of other SFKs cannot be excluded, all the more since particularly Fyn seems involved in regulation of Ag-triggered degranulation and activation of the PI3K pathway . In this respect, enhanced Ag-induced phosphorylation of Akt in Lyn-deficient BMMCs was markedly suppressed by Ro5-4864, clearly indicating Lyn-independent effects of this BDZ and suggesting a Fyn-dependent effect (Additional file 4: Figure S4).
Whereas degranulation is a fast response after Ag triggering of MCs occurring within a few minutes, production of pro-inflammatory cytokines takes place with slower kinetics. Important signaling pathways for cytokine production downstream of the FcεRI include the PI3K, p38, and NFκB pathways . All of these pathways were attenuated by Ro5-4864 treatment in wild-type MCs underlining the role of central signaling elements, e.g. SFKs, being blocked by this drug. Intriguingly, Ro5-4864 also concentration-dependently suppressed cytokine production in response to stimulation of receptor systems such as Kit and TLR4. Also this effect can be explained by the inhibition of SFKs. Both Lyn and Fyn have been reported to play positive regulatory roles in the context of Kit signaling [52, 53]. In addition, a recent publication by Avila et al. has demonstrated the importance of Lyn for the production of TNF-α in response to LPS in MCs . Thus, all effects observed in the present study with the 1,4-benzodiazepine Ro5-4864 are explainable by attenuation of SFK activity.