This study demonstrates that the expression of two goblet cell–associated biomarkers, MUC5AC/Muc5ac and hCLCA1/Gob5, are differentially regulated by IL-13, depending on whether the study is performed in vivo or in vitro. When IL-13 was intratracheally instilled, both Muc5ac and Gob5 messages were elevated in mouse lungs after 24 h, and were presumably expressed by surface goblet cells of mouse airways. This elevated expression was abrogated by Stat6-null mutation, indicating a Stat6-dependent event for both. However, when isolated primary airway epithelial cells from mice and humans were treated with IL-13 in vitro for 24 h, only hCLCA1/Gob5A was found to be elevated, but not MUC5AC/Muc5ac. The lack of stimulation of MUC5AC/Muc5ac in vitro by IL-13 is not related to the activation of STAT6, the transcription factor known to be the main downstream signaling molecule for IL-4 and IL-13. In this study, we showed that STAT6 was readily phosphorylated in human airway epithelial cells by IL-13 within 1 h (). Subsequently, dual immunofluorescent staining with antibodies specific to these two biomarkers revealed the differential nature of the expression of these proteins by different airway cells ().
MUC5AC/Muc5ac and hCLCA1/Gob5 are well known markers for surface goblet cells in the airway epithelium (8
,). IL-13 is the T helper type 2 cytokine that is believed to affect the mucous cell metaplasia on airway epithelium present in allergic-mediated asthma models (20
). The logical conclusion to draw would be that IL-13 acts on airway epithelial cells to induce expression of MUC5AC/Muc5ac and hCLCA1/Gob-5, which then leads to the metaplastic change of cells into a mucous type. But if this model is correct, how can we explain the discrepancy between the ability of IL-13 to induce both Muc5ac and Gob-5 expression in vivo
, but only Gob-5 in vitro
? The first factor that we considered was whether the in vitro
cultures that we used are an acceptable model for airway epithelial cells in vivo
. The primary cell culture system used in this study has a heterogeneous cell population that, under ALI, develops into well differentiated cells with morphologies very close to that of the native airway epithelium (31
). Although there are limitations in all in vitro
systems, and our cultures cannot totally mimic the cells in the intact animal, we believe that our system is among the best for studying biological mechanisms, and provides conditions that are as close as possible to those in vivo
. Although our mouse and human cultured TBE cells did not increase MUC5AC expression in response to IL-13, as they did in vivo
, they were still able to respond to this cytokine by increasing hCLCA1/Gob-5 expression and the activation of STAT6 signaling. These results suggest that our primary TBE cells have a preserved ability to respond to IL-13 and mediate its intracellular signaling cascades, but that the regulation of hCLCA1/Gob5 and MUC5AC/Muc5ac are different. Our hypothesis is that shortly after ligand binding by IL-13 to its receptor on airway epithelial cells, there is an activation of a signaling cascade of the JAK/STAT pathway. The Gob5/hCLCA1 gene is the target of this activated JAK/STAT signaling cascade. However, for MUC5AC/Muc5ac induction in vivo
, the increased expression may depend on more than simple IL-13/epithelial interaction. Additional factors or signaling pathways that are present in vivo
, but missing in vitro
, may be required. These factors/signalings may lead to either the stabilization of the message (38
) or the activation of transcriptional factors, such as NF-κB, activator protein 1, surfactant protein 1, etc., that are needed for the transcription of MUC5AC/Muc5ac gene in airway cells (39
). Other investigators have found the NF-κB pathway to be important in the regulation of MUC5AC (39
). Our results with peptidoglycan support the possibility that the NF-κB pathway may play a role in the regulation of the MUC5AC/Muc5ac gene because signaling through TLR2 by the peptidoglycan could activate NF-κB through the MyD88/TIRAP/TRAF6 pathways (40
). Interestingly, one research group has found that IL-13's ability to induce Muc5ac in vivo
may depend on its cooperation with cysteinyl leukotrienes and chemokines (35
), further suggesting the significance of coordinated signaling in the regulation of gene expression. This possibility can be tested, but is beyond the scope of this study.
Other investigators have shown that IL-13 is capable of inducing the formation of goblet cells under in vitro
conditions in primary cell cultures similar to ours (25
). Although these studies did not always look specifically at MUC5AC/Muc5ac (41
), it is very likely that its induction occurred in their systems, as MUC5AC/Muc5ac is a well-established marker for mucous cell metaplasia (7
). The likely reason for the discrepancy between our results and those of other investigators is the differences in culture conditions and/or doses and time of treatment. Our dose treatment of IL-13 is similar to what other investigators have used, although one group did find that typical doses of IL-13, such as 10 ng/ml, actually inhibited goblet cell formation, but small doses, such as 1 ng/ml, promoted it (41
). Although we did not include doses as low as 1 ng/ml, we did go as low as 2 ng/ml, and did not see any induction of MUC5AC/Muc5ac. Our treatment course (24 h) was much shorter than that used by most other investigators (1–2 wk) in the treatment of primary cells to induce goblet cell formation and/or MUC5AC/Muc5ac induction (25
). However, we chose our time based on our observation that intratracheal induction of MUC5AC/Muc5ac by IL-13 can occur at 24 h. This time course is consistent with that found by other investigators after a single intratracheal injection of IL-13. If IL-13 can induce MUC5AC/Muc5ac within 24–48 h in vivo
), we believe that it is reasonable for our in vitro
treatments to parallel the same time course.
Several studies have suggested that hCLCA1/Gob5 may play a regulatory role in stimulating MUC5AC/Muc5ac expression (13
). Our data do not directly address the issue of whether MUC5AC expression is dependent on hCLCA1/Gob5 expression. Instead, our focus was aimed at understanding the relationship of these two genes in their responses to IL-13. Nevertheless, we suspect that if MUC5AC/Muc5ac expression was critically dependent on hCLCA1/Gob-5, there would not be a discrepancy seen between their coordinated upregulation by IL-13 in vivo
compared with the isolated upregulation of hCLCA1/Gob5 in vitro
. The ability to induce MUC5AC without the induction of hCLCA1 by peptidoglycan, as demonstrated in this study, further supports this notion.
One group of investigators has found a critical importance of Stat6 expression in airway epithelial cells for the induction of Muc5ac in vivo
). Using an ingenious method, whereby mice that were Stat6-deficient but had it reconstituted, only on airway epithelial cells, through a CC10 promoter, this group showed that overexpression of IL-13 could induce mucous cell metaplasia only in mice with reconstituted epithelial Stat6. These data demonstrate convincingly that IL-13's ability to induce Muc5ac and Gob5 depends critically on Stat6. Because no other mediator than IL-4/IL-13 has been identified that can activate STAT6 (42
), it is reasonable to conclude that IL-13's induction of Muc5ac, Gob-5, and mucous cell metaplasia occurs through its direct effects on airway epithelial cells. Our in vivo
data involving the intratracheal injection of IL-13 are consistent with those of this study (36
). Although we did not see an induction of MUC5AC/Muc5ac by IL-13 in vitro
, this does not contradict the notion that IL-13 can induce MUC5AC/Muc5ac and mucous cell metaplasia directly in vivo
. But our data do suggest, as we hypothesized previously here, that IL-13 may require additional factors or signals that are present in vivo
to induce Muc5ac. These factors and signals could still be present in the reconstituted epithelial STAT6 of mice noted previously here.
What are the roles for these two gene products in mucous cell metaplasia and mucus hypersecretion in the airways? The expression of both genes are increased in goblet cells (8
), and both gene products have been localized to mucin granules on goblet cells (16
). However, our data, with the double immunostaining for hCLCA1 and MUC5AC in primary TBE cultures, did not often show colocalization of these proteins. Most cells that expressed either of these proteins appeared to express either one or the other, but not both. Cells that did express both proteins were fairly uncommon in our sections. What can explain this apparent discrepancy? One possibility is that these proteins are expressed by cells of different lineage. We believe this is unlikely, as both proteins have well-established associations with goblet cells. A more likely possibility is that cells stained for hCLCA1 and MUC5AC are cells of the same lineage but along different time-points of differentiation. It is possible that early on, cells destined for differentiation into goblet cells may express hCLCA1, but then express mucin genes later on as they mature. The different staining that we see representing cells at different time points of this maturation process with the rare costained cells are in a transition phase period.
In summary, through an intratracheal instillation, we demonstrated that both Muc5ac and Gob5 were elevated by IL-13 in mouse lungs. These elevations are Stat6-dependent. However, the stimulatory phenomena could not be reproduced for MUC5AC/Muc5ac in primary TBE cultures, including those from human tissues. This negative result is not due to the impairment of STAT6 signaling in these primary TBE cultures, because STAT6 phosphorylation still occurs and JAK inhibitors could still partially block hCLCA1/Gob5 expression. In addition, we have shown that it is possible to selectively stimulate MUC5AC, but not hCLCA1, expression by a bacterial product, peptidyloglycan, in vitro. This suggest to us that, in IL-13 induction of mucous cell metaplasia, the induction of the two biomarkers MUC5AC/Muc5ac and hCLCA1/Gob5 may be differently regulated, with additional factors required for the former compared with the latter.