Background
The normal tracheal airway epithelial layer is composed primarily of pseudostratified ciliated, basal and secretory cells that maintain contact with each other and to a thin basement membrane [
1]. Molecules comprising the airway extracellular matrix (ECM) consist of fibrous (e.g., collagens and elastin) and structural proteins (e.g., fibronectin and laminins) embedded in a hydrated polysaccharide gel containing several glycosaminoglycans (e.g., hyaluronic acid). Laminins are one of many basement membrane ECM molecules that can contribute to cell support and signalling of the airway epithelium [
2]. Laminin was initially coined as a term to describe a single ECM protein but has come to encompass a family of heterotrimeric ECM proteins made up of single α, β and γ chains. To date, there are five α, three β and three γ chains that are known to form at least 16 laminin trimers and a variety of proteolytic fragments [
3]. Laminins can be produced by lung epithelial cells, including bronchial cells [
4,
5]. A variety of laminins are expressed by lung epithelial cells during development and in adult tissue [
6‐
11], including LM-332 (formerly Laminin-5) [
5,
12‐
14]. Differential LM-332/integrin interaction has been shown to be involved in airway epithelial wound responses in culture [
15] and
in vivo [
13]. It is possible that the remodeling of ECM, including LM-332, by protein cleavage or structural changes can expose and/or eliminate ECM receptor binding sites and promote changes in signalling and cellular activity [
16], however, direct studies on the effects of LM-332 on signalling of conducting airway cells are limited. In addition to ECM rearrangements, breach of the epithelial layer causes a redistribution of intercellular connections that are restored after reformation of the pseudostratified epithelial layer [
17,
18]. As a part of normal airway defense, epithelia coordinate cellular responses to prevent damage/toxicity. Airway epithelial cells rely on paracrine signalling and gap junctional communication to coordinate defence-related activities. Gap junctions are formed at points of cell-cell contact where each cell contributes a hexameric hemi-channel made up of connexins (Cx) [
19,
20]. Connexin proteins can convey unique permeability properties upon the gap junction channels, thus, alterations in connexin expression patterns can directly change the types of cell-cell communication between neighbouring cells, and contribute to local tissue response [
21,
22]. Direct studies on the effect of LM-332 on intercellular signalling of conducting airway epithelial cells have not been performed.
There is a complex pattern of connexin isoform expression in airway epithelial cells with at least eight different connexins expressed at various stages of differentiation and development: Cx26, Cx30.3, Cx31.1, Cx32, Cx37, Cx40, Cx43, and Cx46 [
23‐
27]. Changes in connexin expression in upper airway epithelial cells have been associated with developing or post-injury airways
in vivo [
24,
25].
In vitro, functional gap junctional intercellular communication has been traditionally monitored by transfer of low molecular weight fluorescent dyes, or by measurement of electrical conductance. Although these techniques are recognized as valuable experimental tools to identify cellular coupling, they do not always reflect transfer of physiologically significant molecules through gap junctions [
21,
26]. An alternative way to demonstrate gap junctional coupling in cultured airway epithelial cells is through monitoring of coordinated intracellular Ca
2+ concentration ([Ca
2+]
i) changes in response to mechanical stimulation of a single cell [
28]. However, diffusion of second messenger molecules/ions through gap junctions is not the only way Ca
2+ waves can be propagated [
29]. Following mechanical stimulation, cultured conducting airway epithelial cells can release nucleotides (e.g., ATP or UTP) into extracellular spaces resulting in the activation of Ca
2+ signalling pathways via plasma membrane purinergic receptors [
30]. These pathways need not be mutually exclusive: we have shown in primary cultures of rat alveolar epithelial cells that addition of LM-332 to collagen matrices alters the mechanism of coordinating [Ca
2+]
i changes among neighbouring cells [
26,
31‐
34]. These changes in the coordination of Ca
2+ waves were accompanied by alterations of connexin isoform expression patterns and affected by cellular differentiation.
In this study we grew ciliated rabbit tracheal epithelial cells (RTEC) on substrates of LM-332/collagen or collagen alone and monitored functional dye coupling, propagation of intercellular Ca2+ waves following mechanical stimulation, and alterations in connexin isoform expression. We found that, independent of the matrix substratum, ciliated RTEC were functionally coupled by low molecular weight dyes, although the incidence of dye coupling was increased by LM-332. Ciliated RTEC propagated intercellular Ca2+ waves in response to mechanical stimulation on both matrices tested. However, cells grown on LM-332/collagen matrix propagated Ca2+ waves via an extracellular nucleotide pathway whereas cells grown on collagen alone propagated Ca2+ waves via gap junctions. Direct immunocytochemical staining of connexins showed a cellular rearrangement of at least three isoforms, Cx26, Cx43 and Cx46, in response to LM-332/collagen matrix. We suggest that similar changes of extracellular matrix proteins in vivo (e.g., during development, wound repair or disease) lead to changes in intercellular signalling that are important in coordinating upper airway epithelial tissue function.
Discussion
The airway epithelium relies on intercellular communication to coordinate cellular behaviour into tissue function. Such communication is sensitive to changes in the local environment. In this study we used fluorescent dye transfer and intercellular Ca2+ wave coupling assays to elucidate alterations in cell-cell signalling of ciliated RTEC grown on either a collagen or a LM-332/collagen matrix. Diffusion of negatively charged low molecular weight dyes between cells was significantly increased in the RTEC grown on LM-332/collagen matrices. In contrast to the significant increases in dye coupling, gap junctional coupling for physiologically-relevant second messenger molecules that help to coordinate intercellular Ca2+ waves was severely restricted when cells were grown on the LM-332/collagen matrix. Direct analysis of three connexin isoforms – Cx26, Cx43 and Cx46 – displayed a spatial redistribution coincident with matrix and functional/physiological coupling changes. Taken together, ciliated epithelial cells have distinct intercellular signalling pathways that are responsive to alterations of ECM proteins such as those occurring during development, or in response to wounding or disease.
Molecules comprising the airway ECM consist of both fibrous (e.g., collagens and elastin) and structural (e.g., fibronectin and laminins) proteins. Laminins are one of many basement membrane extracellular matrix molecules that can contribute to cell support and signalling of the developing airway [
2,
5,
7,
9,
12]. The laminin isoform LM-332 can be remodelled in the conducting airway during injury or disease [
6,
43,
44]. We have shown that LM-332 has profound effects on cell signalling, development and morphology in primary cultured alveolar epithelial cells [
26,
31‐
34]. In the bronchial airway epithelium, LM-332 can contribute to hemidesmosome formation [
5], however, specific effects of LM-332 on cellular physiology of conducting airway epithelial cells remain ill-defined.
Direct cellular coupling through gap junctions has been traditionally monitored by transfer of low molecular weight fluorescent dyes or by measurement of electrical conductance. Both ciliated and aciliated RTEC have been shown to be electrically coupled [
45]. Initial experiments reported herein focussed on the effects of LM-332 on cell-cell coupling between RTEC using fluorescent tracer molecules. In our findings, RTEC grown on collagen were poorly coupled with LY and showed a low but significantly higher coupling with Alexa 350. When RTEC were grown on collagen matrices that included LM-332, significant increases in both LY and Alexa 350 dye transfer were observed. These shifts in dye coupling in response to LM-332 matrices are similar to increased gap junctional permeability of calcein (MW 622 Da; net charge = -3) in keratinocytes grown on LM-332 and collagen matrices [
46]. The fact that increase in gap junctional permeability to fluorescent markers after growth on LM-322 occurs across cell types may represent a general response to altered matrices.
Although dye coupling techniques are recognized as valuable experimental tools to identity functional gap junctions, it has become increasingly clear that gap junctions made of different connexin isoforms can also allow the differential transfer of physiologically relevant molecules [
21,
47,
48]. To evaluate potential differences in the transfer of physiologically significant molecules, we initiated mechanically-induced Ca
2+ waves between RTEC grown on collagen or LM-332/collagen matrices and used specific inhibitors to identify intercellular signalling pathways. A role for gap junctions in mechanically induced Ca
2+ waves in RTEC grown on collagen matrices has been firmly established [
28,
29,
35,
42,
49‐
51]. In this model, mechanical stimulation induces both the opening of Ca
2+ channels in the plasma membrane and an increase in 1,4,5-inositol trisphosphate (IP
3) concentrations in the stimulated cell [
50,
51] that can further increase [Ca
2+]
i of the stimulated cell through release of Ca
2+ from intracellular stores. The changes in [Ca
2+]
i in adjacent cells is through a gap junctional mediated, IP
3-dependent Ca
2+ release [
29,
35,
42,
51]. A role for paracrine signalling via mechanically-induced ATP or UTP release in primary cultured mouse and human airway cells has been established also [
30,
52,
53]. In this model, mechanical stimulation induces release of nucleotide triphosphate that diffuses in the extracellular environment and binds to purinergic receptors on adjacent cells, activating cellular signals that lead to increases in [Ca
2+]
i.
In this study we show that when RTEC are grown on a LM-332/collagen matrix, mechanically-stimulated Ca
2+ waves are conserved. However, inhibitor studies are consistent with a shift in the mechanism of coordination of Ca
2+ changes to a paracrine/purinergic signalling pathway. Although cultured RTEC cells grown on collagen [
38,
54,
55] or LM-332/collagen (data not shown) can respond to extracellular ATP or UTP by increasing [Ca
2+]
i, it is only the RTEC grown on LM-332/collagen that utilize purinergic signalling in response to mechanical stimulation to coordinate [Ca
2+]
i changes. This pronounced switch in communication mechanisms in RTEC cultures in response to LM-332 suggests that differences in mechanically-induced Ca
2+ communication between rabbit [
28,
29,
35,
42,
49‐
51] and mouse or human airway epithelial cell cultures [
30,
52,
53] may not be due to species-specific differences in airway signalling. Given the extensive remodelling of matrix during development, wound response and disease, mechanisms of cellular communication might also be "remodelled" at these crucial times for coordinated airway epithelial tissue function.
In an attempt to determine specific changes in gap junctions that contributed to the observed alterations in dye and second messenger coupling in RTEC, we examined directly the expression and spatial organization of three connexin isoforms: Cx26, Cx43 and Cx46. All of these isoforms showed mRNA and protein expression in RTEC after growth on either matrix, however, spatial distribution of each of these connexin isoforms was dependent on matrix. On LM-332/collagen matrices Cx26 and Cx43 isoforms were more prominent and Cx46 was less prominent at the cell membrane. These results are not entirely consistent with our previous report that examined connexin isoforms in RTEC grown on collagen [
42]. Using rabbit polyclonal antibodies we detected only a slight pericellular Cx26 staining pattern, an extensive pericellular staining of Cx32, and a lack of Cx43 isoform staining. Our experience with multiple antibodies for connexin isoforms [
56] allowed for a more direct probe of connexins in RTEC reported herein. The establishment of Cx26 or Cx43 gap junctions at the plasma membrane in RTEC grown on LM-332/collagen matrices may account for increased dye coupling; both Cx26 and Cx43 have been shown to increase LY transfer in transfected HeLa cells [
57]. In contrast, in experiments directed at testing isoform second messenger transfer through gap junctions, neither Cx26 nor Cx43 was as efficient as Cx32 in allowing transfer of IP
3 after microinjection [
48]. Similar to what is shown here, increases in dye transfer do not necessarily correspond to second messenger transfer via gap junctions. Because gap junctions made of Cx32 allow for transfer of IP
3 and Cx32-specific antibodies can directly inhibit Ca
2+ wave propagation in RTEC grown on collagen [
42], we suspect changes of this connexin isoform also occur after RTEC are grown on LM-332/collagen matrices. Additionally, we cannot rule out that Cx46 rearrangements shown herein contribute to the observed changes in second messenger coupling. As noted for dye coupling experiments above, there is precedence also for the regulation of connexin expression in response to LM-332 in the extracellular matrix [
31,
32,
46]. In primary cultured alveolar epithelial cells, LM-332/collagen induced a similar change in mechanism of Ca
2+ communication to that observed in RTEC cultures presented in this study [
31,
32]. In addition, an upregulation of Cx26 and a downregulation of Cx43 were reported in these cells, as well as significant changes in cell morphology [
31,
32,
34]. In RTEC, no apparent changes in cellular morphology, connexin protein or mRNA expression were noted when LM-332 was included in the collagen matrix. The observed LM-332 induced changes were more similar to those seen in transfected CHO cells, where a distinct translocation of Cx43 from the cytoplasm to the plasma membrane occurred in the presence of LM-332 [
46].
In summary, the observed connexin re-organization in RTEC grown on collagen or LM-332/collagen matrices does not inhibit fully cellular coupling as demonstrated by the LY and Alexa 350 dye transfer, but shifts the ways in which the cells communicate an increase in [Ca
2+]
i. It should not be discounted that a subtle change in the pattern of Ca
2+ signalling can have a significant change on the cellular physiology of the signal [
58]. In addition to the effects of changes in Ca
2+ signalling on cellular physiology, the direct transfer of several other second messenger/small metabolites could also be altered after connexin re-organization. ATP, ADP, glutathione, glutamate, IP
3, cAMP, cGMP have all been shown to have altered permeability through gap junctions made up of different connexin isoforms [
21,
47,
48,
59]. Thus, the documented changes in LM-332 during wounding, development or pathology could directly affect intercellular communication within the conducting airway epithelium to coordinate a variety of tissue responses.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
BEI contributed cell culture, initial microinjection experiments and was responsible for all immunocytochemistry experiments. CEO contributed to cell culture, microinjection experiments and was responsible for RT-PCR. SB contributed to cell culture and was responsible for digital imaging of [Ca2+]i. All authors contributed to design of experiments, drafting of the manuscript and approved of the final manuscript.