This study is the first to demonstrate acute receptor-dependent, modulatory effects of nicotine on AM. The nAChR involved in this process differ from subtypes reported previously to be involved in "cholinergic anti-inflammatory pathways" outside the lung. Although the effects of nicotine are receptor mediated, these receptors do not form a classical ion channel known from neuronal cells.
Importantly, we detected neither mRNA nor protein of α7 nAChR in AM in contrast to easily detectable α7 subunit mRNA in sensory neurons, brain and in the whole lung homogenate. This is consistent with reported data on the lack of the expression of α7 nAChR in the murine AM cell line MH-S  and our previous work on expression of nAChR in freshly isolated murine AM . In contrast, binding of the polyclonal nAChR α7 antibody H-320 to murine AM has been reported by Su et al. [8, 9] and was also noted by our group in a previous study in the absence of α7 subunit mRNA detection . This antibody, however, produces identical staining in immunohistochemistry and western blotting of the mouse brain, clearly demonstrating lack of specificity at least in the nervous system , so that these findings have to be considered with caution unless corresponding controls on mouse lungs from α7 nAChR-/- mice have been successfully performed.
Still, there might be species differences and plasticity in receptor expression under pathological conditions, since a low level of basal expression of α7 nAChR subunit mRNA in AM isolated from healthy volunteers and an increase in AM isolated from smokers has been reported . Also, α7 nAChR are essential for systemic cholinergic anti-inflammation since the beneficial effects of nicotine in endotoxemia are abrogated in α7 subunit gene-deficient mice . Accordingly, two potent α7 nAChR agonists, GTS-21 and PNU-282987 [22, 23], inhibit LPS-induced TNFα release and reduce acid-induced acute lung injury, respectively, in the mouse lung [8, 10]. Their potency on the most prevalent nAChR subunits identified in our present study on AM, i.e. α9 and α10 nAChR that generally share many pharmacological properties with α7 nAChR , yet has not been determined. Without doubt, however, α7 nAChR is expressed in the lung as demonstrated by RT-PCR in this and previous studies [25, 26]. Functional data show increases in acid-induced excess lung water and vascular permeability in α7 nAChR deficient mice . Endothelial cells may account for this effect . However, since all α7 nAChR antibodies tested so far produce immunohistochemical labeling also in organs taken from α7 nAChR deficient mice [20, 28], immunohistochemistry alone cannot decipher the cell-type specific α7-subunit distribution in the lung, and this issue remains to be solved.
Instead of α7 we observed expression of nAChR subunits α9, α10, β1, and β2, and to a variable extent α2, α3, α5, in rat AM. Mouse AM expressed nAChR subunits α9, α10, β2, and β4. To form classical, ion-conducting nAChR, α subunits combine as heteropentamers with β subunits or build α heteropentamers of α9α10 and homopentamers of α7 and α9 (for review, see ). The subunits detected in AM in the present study would allow combining the following nAChR pentamers: α3β2, α3α5β2, α9α10, and α9 as homopentamer. Since there is a constant expression of subunits α9 and α10 in AM, this combination as homo- or heteropentamer seems to be most likely, if pentamer formation occurs at all.
These subunits have been best characterized in the inner ear, where they form Ca2+-permeable ion channels involved in efferent modulation of hair cell function [30, 31]. Our whole-cell patch clamp recordings and [Ca2+]i measurements in rat AM, however, neither revealed changes in membrane current in response to ACh nor in [Ca2+]i in response to nicotine. Similarly, a subpopulation of human T-lymphocytes expresses α9 and α10 nAChR subunits but fails to show transmembrane currents triggered by ACh , and nicotine does not cause alteration of [Ca2+]i in the rat AM cell line NR8383  and in rat intravascular mononuclear leukocytes obtained from isogenic kidney transplants . Thus, α9α10 nAChR subunits apparently do not form classical ionotropic receptors in cells of the immune system. Still, α9α10 nAChR subunits confer intracellular effects as our data demonstrate an acute α-bungarotoxin sensitive modulatory effect of nicotine upon ATP-induced calcium release from intracellular stores. In general, although to a much smaller extent than in rat cells, this effect was also present in macrophages isolated from C57BL6N and α7 nAChR subunit deficient mice, demonstrating its independency from the α7 nAChR subunit. Similarly, we recently identified a methyllycaconitine sensitive modulatory effect of nicotine upon ATP-induced rise in [Ca2+]i in rat mononuclear leukocytes obtained by vascular perfusion of isogenic kidney transplants . In line with this observation, α9 subunit containing nAChR in outer hair cells of the inner ear do not exclusively assemble into ionotropic receptors, but form metabotropic receptors as well. Here, ACh also reduces ATP-induced rise in [Ca2+]i at a concentration that alone is insufficient to impact [Ca2+]i, and again this effect is α-bungarotoxin sensitive .
Atypical, non-ionotropic effects have also been reported for the nAChR α7 subunit. In T cells, this subunit fails to form a ligand-gated Ca2+ channel but interacts with CD3ζ to modulate TCR/CD3 function . Notably, α7 subunits in this complex exhibit a different agonist/antagonists profile than neuronal ionotropic α7 nAChR. Methyllycaconitine and α-bungarotoxin, both potent inhibitors of ionotropic α7 nAChR, indeed are strong agonists at T cells expressing nAChR α7 subunits . Correspondingly, epibatidine, a highly potent agonist at ionotropic nAChR, failed to mimic the nicotine effect in our present experiments on rat AM.
In contrast to the well-characterized channel properties of nAChR, the mechanisms of atypical nAChR signaling are currently only poorly understood. In peritoneal macrophages, coupling of α7 nAChR to the Jak2-STAT3 signaling pathway resulting in STAT3 phosphorylation has been reported  which we could not observe in rat AM predominantly expressing α9 and α10 subunits. Membrane bound nAChR subunits have been demonstrated to interact with and to modulate signaling by β-arrestin , phosphatidyl-inositol-3-kinase , CD3ζ , and purinergic P2X-receptors [36, 37]. The latter are involved in ATP-induced increase in [Ca2+]i by extracellular influx in human AM, since initial Ca2+ transients are reduced by 40% in Ca2+-free medium . In our present study of rat AM, however, the ATP-induced initial increase in [Ca2+]i and the modulatory effect of nicotine persisted in Ca2+-free solution, demonstrating interference of atypical nAChR with P2Y-receptor mediated Ca2+-release from intracellular stores. In support, we observed expression of P2Y purinergic receptors on AM, among them P2Y2 that mediates Ca2+-release from the endoplasmatic reticulum in mouse macrophages .
Extracellular ATP is well recognized as a "danger" or "host tissue damage" signal and is mostly regarded to promote inflammation [39, 40]. In human AM, it couples to [Ca2+]i increases and stimulates IL-1β and IL-6 release albeit suppressing TNFα production . In the rat AM cell line NR8383, ATP induces P2Y2- and Ca2+-dependent increase in CCL2 synthesis and release . The CCL2-CCR2 axis is a crucial regulator of inflammatory cell influx into the murine lung [42, 43]. Hence, the presently observed nicotinic attenuation of ATP-induced rise in [Ca2+]i can be considered as an anti-inflammatory mechanism triggered by atypical nAChR.