Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

In response to injury caused by various pollutants, toxicants, and carcinogens, proximal and distal airway epithelial cells lose their normal secretory functions and express squamous and keratinizing properties (1). This phenomenon, which is thought to be a protective response against toxicants and pollutants, may lead to epithelial cell transformation and bronchial carcinogenesis if not properly restored (1, 2). Although much progress has been made in defining the regulatory mechanisms of terminal differentiation in keratinocytes, the signal transduction pathways and downstream effectors governing induction of the abnormal squamous differentiation in airway epithelium remain enigmatic. Therefore, delineating the molecular and cellular mechanisms involved in the induction of squamous differentiation will enable a better understanding of toxicant-induced epithelial injury-repair and differentiation processes, and the development of respiratory pathogenesis.

Clara cells, which are located in the distal bronchiolar region, are the major targets for pulmonary toxicity induced by various environmental pollutants and toxicants (3). Clara cells secrete various proteins, such as surfactants and a 10-kD Clara cell protein, CC10 (3), which play an important role in respiratory function. However, upon exposure to toxicants and pollutants, such as phorbol 13-myristate 12-acetate (PMA), tobacco smoke, naphthalene, and other cytotoxicants, Clara cells loose their differentiation markers and express squamous properties both in vivo and in vitro (4). Small proline-rich proteins (SPRRs) are highly abundant in various squamous tissues; however, the expression patterns are distinct. Their expression in general is high in epithelia of oral tissues, such as tongue, esophagus, and stomach, in contrast to external dry epithelia, such as skin (9). SPRRs belong to a multi-gene family consisting of two SPRR1 (1A and 1B) genes, seven SPRR2 (2A-2F) genes, and one each of SPRR3 and SPRR4 (10). At their N- and C-terminal regions SPRRs contain lysine and glutamine residues, which are cross-linked to themselves and/or other proteins by transglutaminase during formation of cornified cell envelope. The cornified cell envelope is a terminal phenotype of squamous cell and acts as a barrier to various external and internal epithelia.

In contrast to squamous epithelia, the expression of SPRRs is low or undetectable in airway epithelium, which normally express mucous properties (11). Among the genes studied, SPRR1B was originally identified as vitamin A- suppressed gene in airway epithelial cells (11). However, SPRR1B expression is highly inducible in airway epithelium both in vitro and in vivo by a variety of agents, which potently induce airway squamous cell metaplasia, such as PMA, carcinogens, and tobacco smoke (12). Recently, we have demonstrated that PMA, a stable functional analog of diacylglycerol that is produced in vivo by hydrolysis of phospholipids, selectively upregulates the expression of SPRRs in human Clara-like H441 cells (15). The induction of SPRR1B expression in proximal (BEAS-2B clone S-6) and distal (H441) airway epithelial cells is primarily regulated at the transcriptional level, mainly by the AP-1 protein complex composed of Jun/Fra-1 (15, 16). Further studies revealed that extracellular signal-regulated kinase (ERK)- like kinase(s) mediate PMA-inducible SPRR1B expression in S-6 and H441 cells (15, 17). Surprisingly, ERK1 and ERK2 do not participate in this process in both cell types.

Recent studies indicated that, in addition to ERK1 and ERK2, ERK5 (also known as big mitogen-activated kinase 1, BMK1), plays a regulatory role in proliferation and differentiation in various cell types (18). However, little information is available about the role of ERK5 in growth factor and/or toxicant-induced injury-repair and differentiation processes in airway epithelium. We present here, for the first time, evidence for the involvement of ERK5 in PMA-inducible SPRR1B transcription in Clara-like H441 cells. PMA significantly induced ERK5 activation in H441 cells, whereas overexpression of dominant-negative (dn) form of ERK5 strongly suppressed both basal and PMA-inducible SPRR1B transcription. Consistent with this, overexpression of dn mutant of MEK5, an upstream activator of ERK5, significantly blocked PMA-inducible promoter activation. Thus, the activation of MEK-ERK5 pathway probably plays a regulatory role in AP-1-dependent gene expression in the distal bronchiolar region.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Cell Culture and Reverse Transcriptase-Polymerase Chain Reaction Analysis

NCI-H441 (Clara-like bronchiolar pulmonary adenocarcinoma) cell line was obtained from the ATCC and maintained in RPMI culture medium supplemented with 5% serum, 1% streptomycin and penicillin, gentamicin (250 ng/ml) and fungizone (125 ng/ml). Human primary bronchial epithelial cultures and cell lines, HBE-1, BEAS-2B clone S-6, and A549 cells were propagated as described elsewhere (16). For reverse transcriptase-polymerase chain reaction (RT-PCR) analysis, total RNA (750 ng) isolated using Trizol reagent was reverse transcribed into cDNA. PCR amplification was performed at 94°C for 30 s, 53°C for 30 s, and 72°C for 1 min with an aliquot of cDNA using ERK5 (F: 5'-CCTCTGA AAGCCTTGAGGAG-3'; R: 5'-GCAGCCC-ACAGACCAGAG GTC-3') or -actin (F: 5'-GAGAAAATCTGGCACCACAAC-3'; R: 5'-TACCCCTCGTAGATG-GGCAC-3') gene-specific forward (F) and reverse (R) primer pairs. The amplified ERK5 and -actin cDNA fragments, 780 and 201 bp, respectively, were separated on 1.2% agarose gel and photographed using the Gel Doc 2,000 System (Bio-Rad, Hercules, CA).

Expression Vectors, Reporter Plasmids, and Transient Transfections

The wild-type (wt) and dn (AEF; mutation of Thr218 and Tyr 220 with Ala and Phe) ERK5 cDNAs cloned in pcDNA3 expression vectors (19) were kindly provided by Jiing-Dwan Lee (The Scripps Research Institute, La Jolla, CA). The dn-MEK5 (MEK5DD; mutation of Ser311 and Thr315 to Asp) cDNA cloned in pCEFL expression vector (20) was generously provided by Silivio Gutkind (National Institute of Dental and Craniofacial Research, NIH, Bethesda, MD). The constitutive active c-Raf-1 (Raf-BXB) expression vector was kindly provided by Stephen Ludwig (Institut für Medizinische Strahlenkunde und Zellforschung, Würzburg, Germany). c-Jun dn-Ras (Ras-N17) and ca-Ras (Ras-V12) were generated as previously described (15). The promoter of -150 to +12 bp SPRR1B fused to luciferase (Luc) gene was described elsewhere (15). AP-1-Luc reporter plasmid (Stratagene, La Jolla, CA) containing seven copies of AP-1-responsive elements (TGACTAA), henceforth referred to as AP-1-Luc. The 165 to +19 bp promoter of human IL-8, which contains the functional motifs such as nuclear factor B (21), subcloned into CAT3 plasmid vector was kindly provided by Reen Wu (University of California at Davis). IL-8 promoter was PCR amplified and directionally subcloned into pGL-3 vector (Promega, Madison, WI) at the restriction sites of SacI and NheI. The orientation and sequence of the promoter was verified by DNA sequencing. DNA transfections were performed using a Fugene transfection reagent (Roche Applied Science, Indianapolis, IN). Cells were grown on 48-well plates at 70-80% confluence and then transfected with 100 ng of promoter reporter construct, 50 ng of cytomegalovirus- -galactosidase (-gal) DNA, and 50-200 ng of parental empty or expression vectors as indicated. Approximately 24 h after transfection, cells were treated with either dimethyl sulfoxide (DMSO) or PMA (20 ng/ml) for 5-12 h. Cells were lysed and Luc activity was measured using a commercially available kit (Promega). Luc activity of individual samples was normalized against -gal activity and/or total protein as described previously (17). Promoter activity for every construct was analyzed in duplicates and all experiments were repeated two to three times.

Immunoprecipitation and In Vitro ERK5 Kinase Assays

H441 cells were cultured to 80-90% confluence and then treated with PMA (20 ng/ml) for the indicated time periods. Cells were washed three times with chilled phosphate-buffered saline containing 1 mM Na₃VO₄ (sodium orthovanadate) and scraped into a lysis buffer containing 20 mM Tris (pH 7.5), 150 mM NaCl, 1 mM ethylene diamine tetraacetic acid, 1 mM ethyleneglycol-bis-(-aminoethyl ether)-N,N'-tetraacetic acid, 1% Triton, 2.5 mM sodium pyrophospate, 1 mM Na₃VO₄, 5 mM -glycerolphosphate, and 1 µg/ml leupeptin. Lysates were sonicated for 15 s and centrifuged for 10 min at 10,000 × g at 4°C to remove cellular debris. Protein concentration in the lysates was determined using the BCA reagent (Pierce, Rockford, IL). An equal amount of cellular protein (200 µg) was immunoprecipitated with anti-ERK5 antibodies (0.4 µg, Cat # E1523; Sigma, St. Louis, MO) and washed extensively. Half of the immunoprecipitate was used for analyzing kinase activity, and the other half was used for immunoblot analysis to determine ERK5 content. ERK5 kinase activities of various samples were determined using 2.5 µg myelin basic protein as substrate in a reaction mixture containing 10 µCi -³²P (adenosine triphosphate), 25 mM HEPES (pH 7.4), 15 mM MgCl₂, and 1 mM dithiothreitol (17). After 15 min incubation, the in vitro kinase assays were terminated with 2× sodium dodecyl sulfate loading buffer. The reaction products were separated on sodium dodecyl sulfate/polyacrylamide gel electrophoresis, transferred to polyvinylidene difluoride membrane, and substrate phosphorylation was determined by PhosphorImager analysis. Total ERK5 content of the immunoprecipitates was analyzed by the Western blot analysis using ERK-5 specific antibodies. ERK5 activity was normalized against total ERK5 content of the respective samples. The value of the untreated sample was considered as one unit.

Statistical Analysis

Data are expressed as the mean ± standard error. The StatView program was used to perform analysis of variance between different samples. Statistical significance was accepted at P < 0.05. All assay samples were performed in duplicates and each experiment was repeated at least two times.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

PMA Activates ERK5 in H441 Cells

To examine a role for ERK5, H441 cells were treated with PMA, which induces epithelial injury and repair process in distal bronchiolar region (22), for different time periods as indicated. Whole cell extracts were immunoprecipitated with anti-ERK5 antibodies (developed against amino acid 789-802 of the c-terminus region of human ERK5) and in vitro kinase assays were performed using myelin basic protein as a substrate as described previously (19). As shown in Figure 1, following PMA treatment the ERK5 kinase activity rapidly increased within 5 min (> 3-fold) and was maximal at 30 min (> 4-fold). ERK5 activation remained elevated for 3 h. The increase in kinase activity following PMA treatment is nearly comparable to that observed with exposure to puromycin, which stimulates ERK5 activity in various cell types (23).

View larger version (34K): [in this window] [in a new window]   Figure 1.   PMA stimulates ERK5 activity in H441 cells. Cells were grown to 70-80% confluence and treated with vehicle (DMSO) or PMA (20 ng/ml) for different time periods as indicated. Cellular extracts (200 µg) were immunoprecipitated with 0.4 µg of anti-ERK5 antibodies. (A) The kinase activity of the immunoprecipitates was analyzed using myelin basic protein as substrate (top). As a positive control for ERK5 activation, H441 cells were treated with puromycin (Pu, 2 µg/ml for 30 min). The autoradiogram (bottom) depicts the Western blot analysis of total ERK5 content in the immunoprecipitates used for in vitro kinase assay. (B) The ERK5 kinase activity was normalized against total ERK5 content. The value of untreated samples was designated as one.

Expression of ERK5 in Airway Epithelial Cells

To confirm the expression of ERK5 in airway epithelium, total RNA was isolated from various human airway epithelial cells, and the message levels were amplified using ERK5 gene-specific primers. As shown in Figure 2, the mRNA expression of ERK5 is detectable in proximal (HBE-1 and S-6) and distal (Clara-like H441 and alveolar-like type II epithelial cell line, A549) airway epithelial cell lines, indicating a broad expression pattern of ERK5. Moreover, ERK5 mRNA expression was also detectable in primary cultures derived from tracheobronchial epithelium (lanes 8 and 9). The amplified PCR product was purified and digested with EcoRI to yield 415 and 365 bp DNA fragments (GenBank Accession # U25278) confirming the specificity of ERK5 cDNA amplification (data not shown).

View larger version (23K): [in this window] [in a new window]   Figure 2.   Expression of ERK5 in human airway epithelial cells. Total RNA isolated by standard protocol was subjected to RT-PCR using gene-specific primers as detailed in MATERIALS AND METHODS. Lane 1: H441; lanes 2 and 3: A549; lanes 4 and 5: BEAS-2B clone S-6; lanes 6 and 7: HBE1; lanes 8 and 9: primary cultures of bronchial epithelium.

ERK5 Regulates PMA-Inducible SPRR1B Transcription in H441 Cells

Recently, we have shown that PMA-stimulated SPRR1B mRNA expression (data not shown) and promoter activation (15) in H441 cells is strongly inhibited by PD98059, which prevents the activation of MKK1/2 (24). Surprisingly, overexpression of dn forms of ERK1 and/or ERK2, which are downstream effectors of MKK1/2, did not suppress PMA-stimulated SPRR1B transcription. A similar observation was also made in BEAS-2B clone S-6 cells, indicating the involvement of a different ERK-like kinase in this pathway (17). It was reported that PD98059 and U0126 compounds also inhibit ERK5 activation (25, 26). Therefore, we examined whether ERK5 mediates PMA-inducible SPRR1B transcription in H441 cells. H441 cells were exposed to 2.5-100 ng/ml of PMA, and SPRR1B promoter-driven luciferase expression was analyzed as described in MATERIALS AND METHODS. PMA increased SPRR1B promoter activity in a dose-dependent manner, with the maximal response seen at 20 ng/ml (Figure 3A). Although higher doses of PMA (50-100 ng/ml) caused significant stimulation of the promoter compared with the untreated controls, these PMA doses caused a slight reduction of the promoter activity compared with cells treated with 20 ng/ml PMA. Therefore, we have used the dose of 20 ng/ml of PMA throughout the study. Moreover, we have previously shown that over a comparable range PMA stimulates SPRR1B expression in H441 cells (15). In support of a role for ERK5 in SPRR1B regulation, overexpression of dn ERK5 strongly suppressed PMA-stimulated reporter gene expression (Figure 3B). Similarly, dn-ERK5 also suppressed PKC activator indolactam-stimulated SPRR1B transcription (Figure 3C). Consistent with this, overexpression of wild-type ERK5 significantly increased SPRR1B promoter activation, which could be augmented with PMA treatment (Figure 3D).

View larger version (30K): [in this window] [in a new window]   Figure 3.   ERK5 regulates PMA-inducible SPRR1B transcription in H441 cells. Upon reaching 70-80% confluence, cells were cotransfected with 100 ng of -150 to +12 bp SPRR1B promoter- driven Luc reporter construct and pCMV--gal (50 ng). After ~ 24 h transfection, cells were treated either with vehicle (control) or with different doses of PMA (5-100 ng/ml; A) or with indolactam (Indo, 5 nM; B). Cells were transfected with SPRR1B promoter reporter construct and pCMV--gal vector in the presence of 100 ng either empty parental, dominant-negative ERK5 mutant (dn-ERK5; C), or wild-type ERK5 (wt-ERK5; D) expression plasmids. After ~ 24 h transfection, cells were treated either with vehicle (control) or PMA (20 ng/ml) in a fresh culture medium. Reporter expression was analyzed as described in MATERIALS AND METHODS. Promoter activity was calculated using the values of parental empty vector transfected cells as one. A, B, and D: open bars, control; solid bars, PMA. C: open bars, control; solid bars, indolactam.

MEK5 Regulates PMA-Inducible SPRR1B Transcription in H441 Cells

We next examined the role of MEK5, an upstream activator of ERK5, in the regulation of SPRR1B expression. Cells were cotransfected with dn-MEK5 expression vector along with SPRR1B promoter reporter construct. After overnight incubation, cells were treated with DMSO or PMA and reporter gene expression was analyzed. Overexpression of dn-MEK5 significantly suppressed PMA-stimulated SPRR1B transcription (Figure 4). Thus, these results indicate that the MEK5-ERK5 pathway mediates PMA-stimulated SPRR1B transcription in H441 cells.

View larger version (23K): [in this window] [in a new window]   Figure 4.   MEK5 regulates SPRR1B transcription in H441 cells. Cells were cotransfected with SPRR1B-Luc reporter and -gal constructs in the presence of empty (pCEFL, 100 ng) or dominant-negative MEK5 (dn-MEK5) expression vector and treated with vehicle (open bars) or PMA (solid bars). The total DNA amount was kept constant by adding empty vector. *P < 0.01 compared with PMA-treated empty vector-transfected group.

ERK5 Mutant Suppresses Ras- and Raf-Inducible SPRR1B Transcription in H441 Cells

Recently, we have shown that Ras regulates PMA-inducible SPRR1B transcription in H441 cells (15). To determine whether ERK5 mediates Ras-inducible gene expression, cells were cotransfected with constitutively active (ca-Ras) expression vector along with SPRR1B promoter construct in the presence or absence of dn-ERK5 plasmid DNA. Consistent with our previous results, overexpression of dn-Ras strongly suppressed PMA-stimulated activity nearly to the basal level (data not shown). In contrast, overexpression of ca-Ras significantly stimulated promoter activity (~ 2.5-fold) compared with an empty vector-transfected control (Figure 5). However, coexpression of dn-ERK5 mutant significantly suppressed such activation. Together, these results indicate that Ras inducible SPRR1B transcription in H441 cells is mediated in part by ERK5.

View larger version (19K): [in this window] [in a new window]   Figure 5.   ERK5 mutant suppresses Ras-inducible SPRR1B promoter activation. Cells were transiently cotransfected with 150-SPRR1B-Luc (100 ng) and -gal (50 ng) vectors along with 100 ng of either empty parental or constitutively active Ras (ca-Ras) expression vectors in the presence or absence of an equal amount of dn-ERK5 plasmid DNA. *P < 0.01 compared with ca-Ras-transfected group.

We next tested whether downstream effectors of Ras, which include Raf-1 and MEKK1 (27), were involved in PMA-induced SPRR1B expression. Although MEKK1 does not appear to activate MEK5 (28), recent studies showed the involvement of Raf-1 in ERK-5-mediated gene regulation (28). The involvement of c-Raf-1 in ERK-5-mediated PMA-inducible SPRR1B expression was demonstrated as overexpression of constitutively active form of c-Raf-1 (ca-Raf-1) markedly upregulated SPRR1B transcription in H441 cells (Figure 6A). In support of ERK5 as downstream target of Raf-1, cotransfection of an equal amount of dn-ERK5 strongly suppressed active c-Raf-1-enhanced SPRR1B promoter activity. In contrast, cotransfection of dn-MEK5 did not suppress c-Raf-1-inducible SPRR1B promoter activation (Figure 6B). These data suggest that c-Raf-1 inducible promoter activity is mediated via ERK5, whereas MEK5 does not appear to participate in it.

View larger version (16K): [in this window] [in a new window]   Figure 6.   ERK5, but not MEK5, mutant suppresses c-Raf-1-inducible SPRR1B transcription. (A) Cells were transiently cotransfected with SPRR1B-Luc and -gal vectors along with 100 ng of either empty parental or constitutively active c-Raf-1 (ca-Raf-1) in the presence or absence of an equal amount of dn-ERK5 (A) or dn-MEK5 (B) plasmid DNA. *P < 0.001 compared with ca-Raf-1- transfected group.

Overexpression of c-Jun Restores Basal and PMA-Stimulated SPRR1B Transcription Suppressed by ERK5 Mutant

We have previously shown that c-Jun, an immediate-early gene, strongly upregulates SPRR1B transcription in BE cells. Because ERK5 regulates PMA-inducible SPRR1B transcription and activates c-Jun transcription (29), H441 cells were cotransfected with SPRR1B promoter construct along with dn-ERK5 in the presence or absence of c-Jun expression vector. Cells were treated with DMSO or PMA and reporter gene expression was analyzed. As shown in Figure 7, overexpression of dn-ERK5 strongly suppressed both basal and PMA-stimulated promoter activity. However, no significant effect was noticed when c-Jun was coexpressed along with mutant ERK5. In fact, SPRR1B transcription remains robustly elevated in the presence of c-Jun.

View larger version (11K): [in this window] [in a new window]   Figure 7.   ERK5 mutant does not suppress c-Jun-enhanced SPRR1B transcription. (A) Cells were transfected with SPRR1B promoter reporter construct along with varying amounts of c-Jun expression vector in the presence or absence of 100 ng of dn-ERK5 plasmid. DNA concentration was kept equal in all samples by adding appropriate amount of pcDNA3 vector. (B) Cells were cotransfected with SPRR1B promoter and 100 ng Jun expression vector in the presence or absence of 200 ng of dn-ERK5. Cells were treated with DMSO (open bars) or PMA (solid bars) and Luc expression was determined. The value of the empty vector- transfected untreated samples was taken as one.

ERK5 Mutant Suppresses PMA-Stimulated AP-1, but Not IL-8, Promoter Transcription

To further determine the specificity of the ERK5-dependent gene expression, we have transfected H441 cells with AP-1-Luc or with human IL-8 promoter reporter constructs in the presence or absence of dn-ERK5 plasmid DNA. As shown in Figure 8A, coexpression of dn-ERK5 strongly suppressed both basal and PMA-stimulated reporter gene expression driven by the consensus AP-1-responsive elements. Note that at the higher concentration, dn-ERK5 can suppress the PMA-inducible promoter activity near to the basal levels. In contrast, dn-ERK5 did not have any significant effect on PMA-inducible IL-8 gene transcription, which appears to be mainly regulated by NF-B (Figure 8B). These results suggest that the ERK5 MAPK pathway probably plays a major role in AP-1-dependent gene expression in bronchiolar cells.

View larger version (20K): [in this window] [in a new window]   Figure 8.   ERK5 mutant suppresses AP-1- but not NF-B-dependent gene transcription. Cells were transfected with AP-1-Luc (A) or IL-8-Luc (B) reporter plasmid (see MATERIALS AND METHODS) in the presence or absence of dn-ERK5 expression vector as indicated. After 24 h incubation, cells were treated with DMSO (open bars) or PMA (20 ng/ml; solid bars) for 5 h. Luc expression was determined as in Figure 3.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The present study clearly establishes a novel role for the MEK5-ERK5 MAPK pathway in PMA-inducible transcriptional stimulation of SPRR1B expression in Clara-like H441 cells. ERK5 is a novel member of the MAPK family (18). Like ERK1/2, ERK5 shares the Thr-X-Tyr sequence and is activated by dual phosphorylation within a TXY motif. However, unlike other MAPKs, such as ERK1/2, p38, and JNKs, ERK5 contains a distinctly larger (by 400 amino acids) C-terminal region (thus named as a Big MAPK) as well as a unique 12-domain loop containing two proline-rich regions (18). The unique C-terminal region of ERK5 can serve as a localization and/or regulatory domain. For example, ERK5 can be localized in the nucleus (18). Also, it was recently shown that C-terminal region of ERK5 contains potent transcriptional activation and interacting domains, which are required for coactivation of myocyte enhancer-binding factor-2D (MEF-2D) during calcium signaling in T cells (30). The loop 12-domain contains several small Pro-Ala and Pro-Thr repeats, similar to that of myosin light chain kinase, which apparently play an important role in cytoskeletal organization (31). Indeed, a recent study showed a role for MEK5-ERK5 pathway in the serial assembly of sarcomeres (32). Therefore, it appears that ERK5 has functions distinct from those of ERK1/2, p38, and JNK MAPKs.

ERK5 activation is induced more pronouncedly by oxidative stress signals, such as H₂O₂ and hyperosmolarity, than by growth factors such as epidermal growth factor and serum (23). Various agents that potently induce ERK1/2, such as TNF , PMA, and PDGF, are apparently weak activators of ERK5 (23). Though much is being learned in other cell types, the mode(s) of activation and functional significance of MEK5/ERK5 pathway in airway epithelium, a major target of various toxicants and pollutants, are not known. By RT-PCR analysis, we were able to detect ERK5 message levels in both primary bronchial epithelial cultures and various cell lines that represent both upper and distal region of the lung (Figure 2). More importantly, in vitro kinase assays of the H441 cellular extracts immunoprecipitated with anti-ERK-5 antibodies revealed stimulation of ERK5 activation in H441 cells by PMA (Figure 1) and H₂O₂ (data not shown). Furthermore, the dn form of ERK5 suppresses PMA-stimulated reporter gene expression, driven by the SPRR1B promoter harboring two functional PMA-responsive elements (or AP-1 sites) or by the consensus AP-1 sites. This suggests a role for ERK5 in the regulation of AP-1-dependent gene expression in the distal bronchiolar region.

The downstream effector(s) that mediate ERK5-regulated SRRR1B transcription in H441 cells are not known. The best-known targets that are phosphorylated and activated by ERK5 include MEF2 family transcription factors (18). Though mainly involved in muscle cell differentiation, the MEF2 family of proteins, MEF2A-D, are also expressed in different cell types, suggesting other regulatory functions for these transcription factors (29). It has been shown that ERK5 regulates c-Jun expression possibly through the activation of MEF2D (29), which bind at the c-Jun promoter to stimulate its transcription (33). Consistent with this, PMA-inducible SPRR1B transcription is mainly mediated by two functional AP-1 sites. Furthermore, PMA induces mRNA expression as well as DNA-binding activity of AP-1 proteins, which form strong protein-DNA complexes at consensus AP-1 sites as well as at the SPRR1B promoter (16, 34). The protein complex at the latter is preferentially occupied by Jun/Fra-1. Moreover, overexpression of c-Jun or Fra-1 robustly upregulates SPRR1B transcription in both H441 and S-6 cells. Therefore it is plausible that overexpression of dn-ERK5 suppresses c-Jun transcription, which results in downregulation of SPRR1B expression. Indeed, in support of this view, overexpression of mutant form of ERK5 did not suppress c-Jun-enhanced SPRR1B transcription (Figure 7A). Furthermore, coexpression of c-Jun blocks the suppressive effects of mutant ERK5 on PMA-inducible SPRR1B promoter activation (Figure 7B). Because ERK5 interacts and coactivates transcription factors (30), we cannot rule out the possibility of potential transcription activation and/or interaction of ERK5 with other proteins, such as ETS family member ESE-1, which also upregulates SPRR1B transcription in S-6 and H441 cells (unpublished data). Indeed, phosphorylation of Sap1a, an ETS family member, by ERK5 has been documented (25).

It is well established that MEK5 is the upstream activator of ERK5. Though overexpression of mutant MEK5 significantly suppressed PMA-inducible SPRR1B transcription, it did not completely suppress promoter activation, which suggests the involvement of other pathways. Indeed, our previous study demonstrated the involvement of Ras/MEKK1/JNK1 pathway in PMA-inducible SPRR1B transcription (17). Coexpression of ERK5 mutant also suppressed both constitutively active Ras and its downstream effector c-Raf-1-inducible promoter activation. However, overexpression of MEK5 mutant did not have any effect on c-Raf-1-inducible promoter activation. Consistent with this, previous studies have shown that Raf-1 and MEKK1, which are downstream effectors of Ras, did not activate MEK5 (28). Interestingly, Raf-1-dependent and -independent pathways have been shown to regulate ERK5-mediated gene regulation (35). For example, it was shown that c-Raf-1 binds in vivo with ERK5, but not with ERK2 or JNK1, suggesting specific protein-protein interactions between MAPKs also play a role in cell proliferation and transformation (28). Although the present study does not rule out such possibilities, a detailed investigation is warranted to further delineate upstream kinase modules that regulate MEK5-ERK5-mediated PMA-inducible SPRR1B transcription in H441 cells.

In summary, we have shown that PMA stimulates ERK5 activation. Furthermore, we have shown that PMA-inducible SPRR1B transcription in H441 cells is mediated in part by MEK5-ERK5 MAPK module probably through c-Jun. Based on this information and on our previous study (15), it appears that a coordinated complex network of MAPK modules, rather than a straightforward linear module, regulates the induction of squamous cell functions in the distal bronchiolar region (see Figure 9). Though regulation of AP-1-dependent gene expression by MEK5-ERK5 MAPK pathway adds a new dimension to the biology of airway epithelium, further investigation(s) is necessary to delineate other downstream effectors, including transcription factors and target genes, as well as upstream activating MAPK modules. This will provide additional insight to the mechanisms of toxicant-induced epithelial injury repair and transformation.

View larger version (16K): [in this window] [in a new window]   Figure 9.   The proposed model depicts the multiple MAPK signaling modules that regulate PMA-stimulated SPRR1B expression in Clara-like H441 cells. Question marks indicate unknowns. The broken arrow indicates upstream MAPK module(s) that activate MEK5 and AP1, AP-1 binding sites.