|Home | About | Journals | Submit | Contact Us | Français|
Endoplasmic reticulum (ER) stress has been implicated in alveolar epithelial type II (AT2) cell apoptosis in idiopathic pulmonary fibrosis. We hypothesized that ER stress (either chemically induced or due to accumulation of misfolded proteins) is also associated with epithelial–mesenchymal transition (EMT) in alveolar epithelial cells (AECs). ER stress inducers, thapsigargin (TG) or tunicamycin (TN), increased expression of ER chaperone, Grp78, and spliced X-box binding protein 1, decreased epithelial markers, E-cadherin and zonula occludens–1 (ZO-1), increased the myofibroblast marker, α–smooth muscle actin (α-SMA), and induced fibroblast-like morphology in both primary AECs and the AT2 cell line, RLE-6TN, consistent with EMT. Overexpression of the surfactant protein (SP)–C BRICHOS mutant SP-CΔExon4 in A549 cells increased Grp78 and α-SMA and disrupted ZO-1 distribution, and, in primary AECs, SP-CΔExon4 induced fibroblastic-like morphology, decreased ZO-1 and E-cadherin and increased α-SMA, mechanistically linking ER stress associated with mutant SP to fibrosis through EMT. Whereas EMT was evident at lower concentrations of TG or TN, higher concentrations caused apoptosis. The Src inhibitor, 4-amino-5-(4-chlorophenyl)-7-(t-butyl)pyrazolo[3,4]pyramidine) (PP2), abrogated EMT associated with TN or TG in primary AECs, whereas overexpression of SP-CΔExon4 increased Src phosphorylation, suggesting a common mechanism. Furthermore, increased Grp78 immunoreactivity was observed in AT2 cells of mice after bleomycin injury, supporting a role for ER stress in epithelial abnormalities in fibrosis in vivo. These results demonstrate that ER stress induces EMT in AECs, at least in part through Src-dependent pathways, suggesting a novel role for ER stress in fibroblast accumulation in pulmonary fibrosis.
This study demonstrates that endoplasmic reticulum (ER) stress due to both chemical induction and overexpression of mutant surfactant protein C that is associated with protein misfolding leads to epithelial–mesenchymal transition in alveolar epithelial cells, suggesting a novel and direct role for ER stress in the pathogenesis of pulmonary fibrosis.
Idiopathic pulmonary fibrosis (IPF), one of the most common forms of chronic interstitial lung disease, is characterized by fibroblast/myofibroblast accumulation and extracellular matrix (ECM) remodeling that lead to disruption of alveolar architecture and progressive fibrosis (1–2–3). Although chronic inflammation has traditionally been viewed as key to the pathogenesis of IPF (4), it is not a prominent feature in biopsies of patients with IPF/usual interstitial pneumonia (UIP), and anti-inflammatory and immunosuppressive therapies have been largely unsuccessful, leading to reassessment of the role of inflammation in this disorder (3–4–5). A resulting paradigm shift suggests a central role for alveolar epithelium in disease pathogenesis, in which IPF is believed to result from repeated episodes of alveolar epithelial cell (AEC) injury in conjunction with release of proinflammatory and profibrotic mediators that lead to fibroblast activation, exaggerated ECM deposition, and progressive fibrosis reminiscent of abnormal wound repair (3–4–5). In addition to the notion of dysregulated epithelial–fibroblast crosstalk and abnormal repair in promoting fibrosis, recent studies suggest that AECs themselves can give rise to fibroblasts/myofibroblasts through the process of epithelial–mesenchymal transition (EMT) (6–7–8).
Several reports have revealed an association between mutations in the surfactant protein (SP)–C gene (SFTPC) and familial pulmonary fibrosis, including cases of UIP, the pathological correlate of IPF (9–10–11). Overexpression of SFTPC constructs with C-terminal mutations in human (A549) and mouse lung epithelial cells indicates that disruption of intracellular processing of mutant SP-C precursor protein leads to accumulation of misfolded protein and induction of endoplasmic reticulum (ER) stress and epithelial cell apoptosis, suggesting a role for ER stress–induced epithelial injury in the pathogenesis of pulmonary fibrosis (13,14). Although genetic mutations in SFTPC are thought to be uncommon in nonfamilial forms of IPF (15), two recent studies suggest that chronic ER stress due to other causes (e.g., viral infection) may contribute to epithelial abnormalities observed in sporadic IPF (13, 16), suggesting ER stress as a common factor underlying epithelial abnormalities in IPF.
Myofibroblasts, key effector cells in IPF (17), have been suggested to arise from resident lung fibroblasts, through differentiation of circulating bone marrow–derived progenitors, and/or directly from AECs that have undergone EMT (6, 17, 18). EMT has been increasingly implicated in fibrosis after injury in a number of organs, including kidney and lung (19, 20). Treatment of both primary AECs and lung epithelial cell lines with transforming growth factor (TGF)–β1 induces EMT (8), and EMT has been reported to contribute to the fibroblast population in murine models of pulmonary fibrosis (7). Together with evidence for EMT in IPF lung biopsies (7, 8, 21), these studies support a role for EMT in the pathogenesis of lung fibrosis.
Association of mutant SPs with familial forms of interstitial lung disease, including pathologic UIP, demonstration that accumulation of misfolded SP-C protein in epithelial cells induces ER stress, evidence for ER stress in AECs in both familial and sporadic forms of IPF, and association of ER stress with apoptosis in IPF suggested to us that, regardless of the initiating factor(s), ER stress may serve not only as the basis for apoptosis, but may contribute to other epithelial derangements observed in IPF. In the current study, we demonstrate that both chemical induction of ER stress by thapsigargin (TG) or tunicamycin (TN) and overexpression of mutant SFTPC leads to EMT in AECs. Induction of EMT by overexpression of a misfolded protein serves as proof of concept that ER stress, regardless of cause, may contribute to fibrosis through EMT even in nonfamilial IPF. Furthermore, the level of ER stress to which cells are subjected appears to be a critical determinant of whether cells undergo apoptosis or EMT. These findings demonstrate that ER stress of diverse causes is associated with EMT in AECs, suggesting a novel mechanism whereby ER stress may contribute to fibroblast accumulation in pulmonary fibrosis.
Details of antibodies and reagents are provided in the online supplement.
Alveolar epithelial type II (AT2) cells were isolated from adult male Sprague-Dawley rats by elastase disaggregation, as previously described (22). See the online supplement for details.
Transepithelial resistance (Rt [KΩ-cm2]) was measured in AECs on Days 3 and 5 after plating using a rapid screening device (MilliCell-ERS; Millipore, Bedford, MA), as previously described (23).
RLE-6TN cells and A549 cells were purchased from American Type Culture Collection (Manassas, VA). Details of cell culture are provided in the online supplement.
RNA was harvested using an RNeasy Mini Kit (Qiagen, Valencia, CA). cDNA was synthesized with random hexamers and Superscript III (Invitrogen, Carlsbad, CA). Details of PCR conditions are provided in the online supplement.
Primary AEC monolayers grown on polycarbonate filters or RLE-6TN cells and A549 cells grown on chamber slides were fixed with 4% paraformaldehyde in PBS (pH 7.4) at room temperature for 10 minutes and processed for immunofluorescence staining, as described in the online supplement.
Western analysis was performed as previously described (22), and blots were analyzed with an Alpha Ease RFC Imaging System (Alpha Innotech, San Leandro, CA). See the online supplement for detailed methods.
Terminal deoxynucleotidyltransferase-mediated dUTP nick end labeling (TUNEL) assay (Roche Molecular Biochemicals, Indianapolis, IN) was performed according to the manufacturer's instructions as detailed in the online supplement.
cDNA for wild-type SP-C (SP-Cwild-type) and BRICHOS mutant SP-CΔexon4 (deletion of exon 4) were subcloned from previously published vectors (24) and inserted into the multiple cloning sites upstream of enhanced green fluorescent protein (EGFP) (25) to generate SinhCMV-mcs(r)-pre-cppt-IG SP-Cwildtype and SinhCMV-mcs(r)-pre-cppt-IG SP-CΔexon4. See the online supplement for details.
Details of lentiviral infection are provided in the online supplement.
Nkx2.1-Cre mice in the C57BL/6 strain were kindly provided by Dr. Stuart Anderson (Cornell University) (26). R26R–β-galactosidase (LacZ) reporter mice (27) and C57BL/6J mice were purchased from Jackson Laboratory (Bar Harbor, ME). Nkx2.1-Cre;LacZ mice were generated by crossing Nkx2.1-Cre mice and R26R-LacZ reporter mice. See the online supplement for details.
LacZ activity was determined by 5-bromo-4-chloro-3-indolyl B-D-galactopyranoside (X-gal) (Sigma, St. Louis, MO) staining. X-gal–stained sections were processed for immunofluorescence for localization of Grp78. Details are provided in the online supplement.
Data are shown as means (±SEM) for the indicated number of experiments (n). Proteins were normalized to an internal control and expressed as percentage of controls treated with vehicle only. We used z tests to determine if ratiometric data were statistically different from control. Rt and quantitation of TUNEL staining and α–smooth muscle actin (α-SMA) expression were analyzed by one-way ANOVA. A P value of less than 0.05 was considered statistically significant.
TN elicits ER stress by inhibiting protein N-linked glycosylation, leading to accumulation of unfolded proteins and consequent unfolded protein response (UPR) activation (28). TG inhibits ER Ca2+-ATPase and blocks calcium reuptake into the ER lumen, leading to disruption of protein folding and subsequent induction of ER stress (29). Effects of TN and TG on ER stress in RLE-6TN cells and primary AECs were investigated by assessing the expression profile of ER chaperone Grp78 and splicing of mRNA encoding spliced X-box binding protein (XBP)–1 (a downstream target of inositol-requiring enzyme–1). Grp78 mRNA increased 15- to 20-fold in RLE-6TN cells (Figure 1A) and 30- to 50-fold in primary AECs (Figure 1D) after treatment with TN or TG. Western analysis demonstrated significant increases in Grp78 protein in RLE-6TN cells (Figure 1B) and primary AECs (Figure 1E) after TN or TG treatment. Consistent with induction of ER stress by these agents, an increase in the spliced form of XBP-1 mRNA (XBP-1S, 575 bp) was observed after TN or TG treatment in RLE-6TN cells (Figure 1C) and AECs (Figure 1F). Immunofluorescence microscopy of primary AECs on Day 3 for TN and on Day 7 for TG (Figure 1G) further demonstrated intense Grp78 staining after treatment with either TN or TG on Day 2 compared with controls, consistent with UPR activation.
Effects of TN and TG on expression of epithelial and mesenchymal markers were evaluated in RLE-6TN cells and primary AECs. In RLE-6TN cells treated with TN (Figure 2A) and TG (see Figure E1A in the online supplement), steady-state levels of zonula occludens (ZO)–1 and E-cadherin protein were markedly reduced to approximately 40% those of untreated cultures. To assess whether down-regulation of these epithelial markers reflected EMT, we also examined effects of TN and TG on cell morphology and α-SMA expression. After treatment of RLE-6TN cells with TN for 24 hours on Day 1 after plating, progressive loss of cuboidal, cobblestone appearance (Figure 2B, i) and acquisition of irregular spindle-shaped morphology (Figure 2B, ii) were observed on Day 3. Similar changes were seen after treatment of RLE-6TN cells with TG (Figure E1B, i, and Figure E1B, ii). Consistent with transition to a myofibroblast phenotype, intense staining for α-SMA in a fibril-associated pattern was observed on Day 3 in TN- (Figure 2C, ii) and TG- (Figure E1Cii), but not vehicle-treated RLE-6TN cells (Figure 2C, i and Figure E1Ci). Western analysis confirmed the increase in α-SMA after treatment of RLE-6TN cells with TN (Figure 2D).
These observations were extended to primary AECs using freshly isolated rat AT2 cells grown on polycarbonate filters for 2 days, followed by treatment with TN for 24 hours, and maintained in culture for an additional 4 days. Similar to observations with RLE-6TN cells, phase contrast microscopy demonstrated a change in morphology of treated AECs from a cuboidal shape (Figure 2E, i) to an irregular, spindle-shaped appearance (Figure 2E, ii). Similar results were observed after treatment of primary AECs with TG for 30 minutes on Day 2 (Figure E1D). Intense staining for α-SMA was observed in primary AECs treated with TN (Figure 2F, ii) or TG (Figure E1E, ii), but was absent in control cells (Figure 2F, i and Figure E1E, i). Immunoreactivity for ZO-1 (Figure 2G, ii) and E-cadherin localized to cell borders (Figure 2G, iv) was concurrently reduced in TN-treated compared with untreated AECs (Figure 2G, i and iii). Similar redistribution of epithelial markers was observed in AECs treated with TG (Figure E1F). Furthermore, AEC monolayers treated with TN or TG on Day 2 failed to develop transepithelial electrical resistance on Days 3 and 5, reflecting deterioration of tight junctions accompanying EMT and consistent with loss of the epithelial phenotype (Figure 2H and Figure E1G).
In A549 cells, overexpression of mutant SP-C leads to ER stress and α-SMA is increased in TGF-β–induced EMT (14,30). We used both A549 cells and primary AECs to investigate the contribution of ER stress due to misfolded protein accumulation to lung fibrosis through induction of EMT. Transduction of A549 cells with mutant SP-CΔexon4 increased expression of Grp78 protein approximately 1.5-fold compared with cells transduced with either SP-Cwild-type or EGFP control (Figures 3A and 3B). Importantly, transduction with mutant SP-CΔexon4 increased expression of α-SMA by approximately 50% (Figures 3A and 3B) and decreased membrane localization of ZO-1 (Figure 3C), changes that were not observed with SP-Cwild-type compared with EGFP control. In primary AECs, overexpression of mutant SP-CΔexon4 resulted in loss of cobblestone epithelial appearance and transition to fibroblastic-like morphology (Figure 3D), together with a decrease in ZO-1 and E-cadherin and an increase in α-SMA protein expression (Figure 3E) compared with EGFP control. The finding that overexpression of the mutated form of SP-C induces α-SMA concurrent with loss of epithelial characteristics further suggests that, in interstitial lung disease associated with SFTPC mutations, ER stress induced by misfolded protein may, in part, contribute to fibrosis through induction of EMT. To confirm that lentivirus constructs encoding EGFP (control), SP-Cwild-type (WT) and SP-CΔexon4 (Δexon4) were equally overexpressed, we performed FACS analysis of EGFP in A549 cells and Western blotting of EGFP in primary AECs that were transduced with virus. As shown in Figures E2A and E2B, EGFP expression of all three constructs was similarly distributed in both A549 cells and primary AECs. EGFP was not detected in non-transduced cells (data not shown). When cells were transduced with lentivirus constructs encoding SP-Cwild-type or SP-CΔexon4, pro-SPC (26 kD) or a truncated SP-C were observed (Figure E2B), confirming appropriate expression of the specific construct.
Induction of the UPR is designed to be protective against ER stress; however, sustained or excessive ER stress can lead to apoptosis (12). Our current studies demonstrating EMT in primary AECs in response to ER stress were performed using concentrations of 1 μg/ml TN and 0.2 μM TG. To investigate dose-dependent effects of ER stress, AECs were treated with increasing concentrations of TN (1–10 μg/ml) or TG (0.2–0.6 μM) on Day 2 and analyzed by TUNEL staining on Day 3. Exposure to higher concentrations of TN (5–10 μg/ml) (Figure 4A) or TG (0.6 μM) (Figure E3A) induced AEC apoptosis. Concurrent reduction in EMT, as evidenced by less expression of α-SMA (Figure 4B) at higher doses, supports the notion that the extent of ER stress is critical in determining whether AECs undergo EMT or commit to apoptosis.
To further investigate cell fate decisions in response to ER stress, expression of C/EBP homologous protein (CHOP), a UPR marker that is implicated in apoptosis, was evaluated after treatment of primary AECs with TN (1 μg/ml) or TG (0.2 μM) on Day 2. Representative Western blots demonstrated CHOP induction at early time points (Day 3) at lower concentrations of TN (Figure 4C) or TG (Figure E3B) that, in the current study, induced EMT. However, up-regulation of CHOP was transient and had decreased markedly by Day 5. In contrast, increases in Grp78 evident on Day 3 were sustained through Day 5 (Figure 4C and Figure E3B). Spliced XBP-1, a downstream target of the UPR, was also present on both Days 3 and 5, although at lower levels on Day 5 than on Day 3 (Figure 4D and Figure E3C). Despite sustained activation of the UPR under conditions of mild ER stress, these transient increases in CHOP were insufficient to induce significant apoptosis, suggesting that the balance between different arms of the UPR determines the decision between apoptosis and survival.
To elucidate mechanisms mediating ER stress–induced EMT in AECs, we performed experiments with pharmacologic inhibitors of candidate signaling pathways, including inhibitors of JNK (SP600125), phosphoinositide 3-kinase/Akt (LY294002), Akt, Rho kinase (Y27632), Raf-1 kinase (GW5074), Alk-5 (SB431542), Smad3 phosphorylation (SIS3), p38 kinase (SB202190), NF-κB (JSH23) and Src (PP2). When assessed by immunofluoresecence, PP2 abrogated effects of TN (Figure 5A) and TG (Figure 5B) on cell shape and expression of epithelial and mesenchymal markers in primary AECs. Although induction of α-SMA was only partially inhibited by PP2, particularly after treatment with TG, expression and localization of epithelial markers was preserved, suggesting an important role for Src activation in ER stress–induced EMT. Treatment with JSH23 slightly decreased α-SMA expression without affecting cell morphology (data not shown). To exclude the possibility that reduction of α-SMA was caused by potential confounding inhibitory effects of PP2 on p38 mitogen-activated protein kinase (30), we evaluated effects of the p38 kinase inhibitor, SB202190, on TN-induced EMT. SB202190 did not affect TN-induced α-SMA expression (data not shown), suggesting that it was the inhibitory effects of PP2 on Src signaling that abrogated ER stress–induced EMT. Other inhibitors examined had no effect, and were not investigated further.
As shown in representative Western blots and quantitative analysis (Figures 5C), TN increased tyrosine phosphorylation of c-Src, which peaked at 1 hour and, as expected, was inhibited by PP2. Similar effects on c-Src phosphorylation were observed after treatment with TG (Figure E4), confirming activation of Src as the result of chemical induction of ER stress. Interestingly, mutant SP-CΔexon4 also induced phosphorylation of c-Src approximately 1.5-fold compared with SP-Cwild-type (Figure 5D), implicating this pathway in induction of EMT in response to ER stress, regardless of the initiating factor.
The potential contribution of ER stress to epithelial abnormalities in the bleomycin model of lung fibrosis was evaluated using Nkx2.1-Cre;LacZ mice in which AECs are permanently labeled to express β-galactosidase to allow identification of cells of epithelial origin. As shown in Figure 6A, minimal expression of Grp78 was observed in frozen lung sections from Nkx2.1-Cre;LacZ mice 12 days after intranasal administration of saline. In contrast, Grp78 protein was dramatically increased in lung sections from reporter mice 12 days after administration of bleomycin and colocalized with X-gal–positive cells, confirming up-regulation of Grp78 in AECs (Figure 6B). These results indicate that ER stress is induced in AECs in this model of lung injury and support a role for ER stress in epithelial abnormalities observed in pulmonary fibrosis in vivo.
Activation of the UPR and ER stress have been demonstrated in AECs in patients with both familial and sporadic IPF, whereas, in sporadic IPF, markers of the UPR have been colocalized in AECs with markers of apoptosis (13,16). Expression of mutant SP-C protein in AECs in vitro has also been shown to evoke both the UPR and apoptosis, further suggesting a role for ER stress in epithelial abnormalities observed in IPF, and implicating ER stress in disease pathogenesis (16). We previously demonstrated colocalization of epithelial and mesenchymal markers in hyperplastic AT2 cells in IPF lung tissue, suggesting that AECs undergo EMT and contribute to fibroblast accumulation in pulmonary fibrosis (8). In the current study, we report, for the first time in the lung, that chemical induction of an ER stress response is associated with EMT in both primary AECs and an immortalized AT2 cell line, and is mediated, at least in part, in a Src kinase–dependent manner. Furthermore, induction of the UPR and acquisition of a mesenchymal phenotype through overexpression of mutant SFTPC in AECs in vitro, and up-regulation of Grp78 in AT2 cells after bleomycin-induced lung injury, support a role for this pathway in epithelial abnormalities observed in vivo in pulmonary fibrosis. These findings suggest that ER stress may contribute to lung fibrosis through induction of EMT.
Nascent proteins that enter the ER must be properly folded with the assistance of ER chaperones before export to the Golgi (31). Protein folding may be disrupted by a number of processes, including, for example, alterations of calcium homeostasis, viral infection, expression of mutant proteins, and altered glycosylation, leading to retention of unfolded or misfolded proteins in the ER (32). In response, the UPR becomes activated through a process involving dissociation of bound Grp78 from ER transmembrane proteins (33). The UPR is executed through the actions of three proximal ER transmembrane proteins (pancreatic ER kinase [PKR]-like ER kinase, activating transcription factor [ATF]–6, and inositol-requiring enzyme–1) (33), activation of which leads to attenuation of the effects of unfolded protein accumulation by enhancing protein folding capacity (28), attenuating global protein synthesis (34), and inducing transcription of ER chaperones and genes that promote degradation of unfolded proteins (35). If the capacity of the UPR is exceeded, ER stress results, which, if excessive or prolonged, may lead to apoptosis through both caspase-dependent and -independent pathways (36).
Activation of the UPR and ER stress have been reported in both sporadic and familial IPF in either the presence or absence of SFTPC mutations, whereas mutations of SP-A2 associated with pulmonary fibrosis have been shown to lead to protein instability and ER stress in vitro (37). In addition, in sporadic IPF, UPR markers have been colocalized with activated caspase 3 and TUNEL staining in hyperplastic AT2 cells (16), suggesting that ER stress contributes to AEC apoptosis. Our current results demonstrate that ER stress in response to either chemical induction or overexpression of mutant proteins is associated with EMT in both RLE-6TN and A549 cells, as well as primary AECs, suggesting a novel mechanism whereby ER stress might contribute directly to fibrosis. Previous studies have demonstrated that, after bleomycin-induced lung injury, up to 30% of fibroblasts may be derived via EMT (7), although a possible role for ER stress in response to bleomycin has not been evaluated. Our current results demonstrating up-regulation of Grp78 on Day 12 after bleomycin-induced injury in AECs (Figure 6) suggest a possible role for ER stress in observed epithelial abnormalities (including EMT) associated with fibrosis in vivo. The factors that lead to ER stress in IPF have been suggested to include viral infection, gatroesophageal reflux, and accumulation of reactive oxygen species, all of which have an increased association with IPF in sporadic disease (16,38).
The UPR functions as an adaptive response to accumulation in the ER lumen of mis- or unfolded proteins (32). If protein accumulation is sustained and ER stress conditions do not resolve, UPR activation leads to apoptosis, although the precise mechanisms regulating transition from an adaptive, prosurvival response to a commitment to apoptosis are not fully understood (39). Rutkowski and colleagues (40) suggested that the adaptive response under conditions of mild ER stress was not the result of selective activation of specific UPR components, but rather due to early down-regulation of proapoptotic UPR components. In our studies in primary AECs under mild ER stress conditions (1–5 μg/ml TN or 0.1–0.5 μM TG), up-regulation of CHOP was transient, increasing on Day 3 in culture after treatment on Day 2 (Figure 4C and Figure E3B) and decreasing by Day 5. In contrast, increases in Grp78 were sustained through Day 5 (Figure 4C and Figure E3B), consistent with its continued adaptive role in restoration of normal ER function (35). Although both apoptosis and EMT are features of IPF, the mechanisms that determine the choice between these two cell fates likely depends on the duration and extent of AEC injury/ER stress. Demonstration in the current study that mild ER stress (0.2 μM TG or 1–5 μg/ml TN; Figures 4A and 4B, and Figure E3A) is associated with less apoptosis than more stringent conditions, leads us to speculate that EMT in IPF is part of an adaptive response that serves to protect cells against ER stress–induced apoptosis, while recognizing that this adaptive response to ER stress is associated with dysregulation of normal epithelial function and defective repair that may, itself, contribute to the pathogenesis of fibrosis (Figure 7).
There is increasing evidence for association of mutations of the SFTPC gene with chronic interstitial lung disease in both children and adults (9, 11). The C-terminal domain of SP-C is critical for normal intracellular protein processing, and mutations in this region are thought to induce cellular toxicity due to protein folding abnormalities (41). Overexpression of BRICHOS domain mutant forms of SP-C in vitro induces ER stress and apoptosis (9,11, 13,14,24,41,42). Our studies show that activation of the UPR as a result of overexpression of the mutant protein SP-CΔexon4 is accompanied by changes consistent with EMT in both A549 cells and primary AECs (Figure 3). These results suggest that EMT serves as a mechanism whereby mutant SFTPC leads to fibroblast accumulation in familial pulmonary fibrosis, and supports the notion that ER stress, regardless of cause, may be a common mediator of epithelial abnormalities in pulmonary fibrosis.
Our investigations provide evidence to support a role for activated Src kinase in mediating ER stress–induced EMT in AECs. PP2 attenuated effects of TN and TG on EMT, and Src phosphorylation was induced by TN and TG, as well as overexpression of SP-CΔexon4. These results suggest a role for Src activation in EMT in response to both chemically induced ER stress and overexpression of misfolded SP, suggesting that Src is activated downstream of the UPR rather than due to direct effects of TN or TG. Although expression of α-SMA was not completely inhibited by PP2, especially after TG treatment, expression and localization of epithelial markers (especially E-cadherin) were preserved, suggesting an important role for Src activation in ER stress–induced EMT. Our findings provide the first evidence in lung epithelial cells that links ER stress to EMT, at least in part through Src activation, and are consistent with a recent study in thyroid PC CI3 cells showing that treatment with TN or TG promoted loss of differentiated phenotype and induction of changes consistent with EMT in a Src-dependent manner (43). Our current findings suggest that Src activation may be a generally applicable response to UPR activation and ER stress, particularly in cells with high protein synthetic function. However, the precise mechanism(s) that link the UPR to Src activation remain to be determined. c-Src has been shown to mediate cancer progression primarily through effects on cell adhesion, invasion, and motility, although evidence for induction of complete EMT is more limited (45–46–47). Src activation has also been shown to modulate cell–cell adhesion and cell migration through effects on both focal adhesions and adherens junctions (44), and specifically to induce tyrosine phosphorylation of β-catenin that leads to disruption of its association with E-cadherin, thereby disrupting cell–cell adhesion (45–46–47). Whether similar mechanisms mediate EMT in response to Src activation in the context of UPR activation and ER stress remains to be determined.
In conclusion, the demonstration that ER stress is associated with EMT in AECs suggests a novel and direct role for ER stress in the pathogenesis of pulmonary fibrosis. Although ER stress in association with SFTPC mutations is attributed to accumulation of mutant proteins, the precise factors resulting in ER stress associated with fibrosis in nonfamilial disease remain to be determined. Differential responses of AECs to various levels of ER stress further suggest that EMT may represent an adaptive response in which the level/duration of ER stress determines the choice between cell survival and apoptosis. Modulation of ER stress responses in AECs could serve as a novel approach to ameliorate epithelial abnormalities implicated in the pathogenesis of IPF.
E.D.C. is Hastings Professor and Kenneth T. Norris Jr. Chair of Medicine. Z.B. is Edgington Chair in Medicine. P.M. is Hastings Professor of Pediatrics. The authors thank Dr. Amy Lee for helpful discussions and Dr. Janice Liebler and Anahita Nersiseyan for technical help.
This work was supported by the Hastings Foundation, the Whittier Foundation, and by National Institutes of Health grants DE014183 (D.K.A.), DE010742 (D.K.A.), ES017034 (E.D.C.), ES018782 (E.D.C.), HL019737 (M.F.B.), HL038578(Z.B.), HL038621 (E.D.C.), HL056590 (P.M.), HL062569 (Z.B.), HL089445 (Z.B.) and HL095349 (P.M.).
This article has an online supplement, which is accessible from this issue's table of contents at www.atsjournals.org
Originally Published in Press as DOI:10.1165/rcmb.2010-0347OC on December 17, 2010
Author Disclosure: R.B. has served as a consultant for Cambridge Antibody Technology and Bayer, has served on the board for Boehringer-Ingelheim, InterMune Inc, Actelion, and Genzyme, and has received lecture fees from InterMune, Actelion, GlaxoSmithKline, and Astra-Zeneca. None of the other authors has a financial relationship with a commercial entity that has an interest in the subject of this manuscript.