Myeloid ELF-1-like Factor Up-regulates Lysozyme Transcription in Epithelial Cells*

Lysozyme is an important component of innate immunity against common pathogens at mucosal surfaces. We previously cloned and characterized the bovine lysozyme 5A (lys5A) promoter with the purpose of determining cis- andtrans-acting elements controlling airway epithelial cell-specific expression. We found that such expression is controlled by protein binding to an ETS consensus sequence located approximately at −46 to −40 bp from the transcription start site. The identity of the ETS-related protein responsible for gene transactivation was unknown. In this study, we screened six ETS-related proteins by transient transfection into epithelial cells and fibroblasts. Results showed that among these factors, the myeloid Elf-1-like factor (MEF) was the most potent. Gel shift analysis of epithelial cell nuclear extracts using a lys5A probe including the ETS-binding site (−50/−31) yielded a single band with retarded mobility. This band was supershifted by an antibody directed against MEF. Supporting the possibility that MEF is responsible for functional transactivation of lysozyme in epithelial cells, we found that antisense MEF mRNA decreased lys5A promoter activity and that MEF overexpression in stably transfected cells increased lysozyme mRNA and protein expression. We conclude that MEF is required for epithelial cell transactivation of lysozyme.

Lysozyme, also known as muramidase, is an enzyme catalyzing the hydrolysis of ␤1-4-glycosidic bonds between N-acetylmuraminic acid and N-acetyl-D-glucosamine, constituents of the cell walls of most bacteria. Its antibacterial properties render it an important participant in host defense at mucosal surfaces and in leukocytes (1)(2)(3). Recent work has shown that inhibition of cationic antimicrobial proteins such as lysozyme predisposes the airway epithelium to infection (4).
At mucosal surfaces, lysozyme expression is confined to specialized epithelial cells, including the serous gland cell of the respiratory epithelium (5,6) and the Paneth's cell of the gastrointestinal epithelium. In hematopoietic cells, lysozyme expression is confined to the macrophage and granulocyte lineages (7). Mechanisms responsible for cell-specific expression are poorly understood. One feature that appears to be critical for expression in both cell types, however, is transactivation by ETS (E26 transforming-specific) family proteins (8). ETS proteins were initially described in the E26 avian leukemia virus (9 -11), and they have now been identified in many species from Drosophila to man (12). Many myeloid-specific genes require ETS protein binding to DNA (e.g. CD4 and macrophage colonystimulating factor) (13,14). That ETS proteins play important roles in myeloid cell development was demonstrated by experiments showing that the targeted disruption of the PU.1 or ETS-1 gene has profound effects on development (15,16).
Previously, we showed that epithelial cell-specific lysozyme expression required an intact ETS-binding site in the proximal promoter of the bovine lysozyme gene 5A (17). This was consistent with results indicating a requirement for binding of the ETS family transcription factor PU.1 to mediate lysozyme transcription in chicken macrophages (18). Although the identity of the lysozyme-associated ETS protein in epithelial cells was unknown, it was clearly not PU.1 since PU.1 is not expressed in epithelial cells (19).
In the studies reported here, we examined the role of various ETS family members in the control of lysozyme transcription in epithelial cells. Our results indicate that the myeloid Elf-1-like factor (MEF) 1 is necessary for lysozyme expression in epithelial cells and is sufficient to mediate lysozyme transcription in fibroblasts.

EXPERIMENTAL PROCEDURES
Cell Culture and Transfections-Cell lines were grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum at 37°C in a humidified 5% CO 2 and 95% air atmosphere.
Transient transfections were performed with Transfectam (Promega) according to the manufacturer's recommendations. Specifically, 10 l of Transfectam reagent and 4 g of total DNA in Dulbecco's modified Eagle's medium were incubated for 10 min before the mixture was applied to subconfluent cells on six-well plates. Cotransfection of various plasmids was performed with 2 g of reporter and 2 g of each effector. Empty vector (pCB6) was added where necessary to ensure a constant amount of input DNA. Cotransfection with the pRL-CMV vector (10 ng in each sample), which expresses Renilla luciferase (Promega), verified that differences in firefly luciferase reporter gene expression were not due to differences in transfection efficiency. Cells were incubated for 2 h with the DNA mixtures, at which time additional medium was added. Forty-eight hours after transfection, the medium was removed, and cells were harvested. Luciferase activity was measured using a dual-luciferase reporter assay system (Promega) and a luminometer (Lumat LB9507, EG&G Berthold). Absolute light emission generated from the luciferase enzyme reaction was determined. Relative luciferase activity is plotted and represents the -fold induction of activity generated by experimental treatments with respect to activity associated with basic vector alone. Values are shown as means Ϯ * This work was supported by grants-in-aid for scientific research from the Ministry of Education, Science, and Culture of Japan and by National Institutes of Health Grants R0143762 and P0124136 (to C. B.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
S.E. (n ϭ 6). To normalize expression levels of the transcription factors, we measured mRNA by Northern blotting using the transcribed sequence of pCB6 as a probe.
For generation of stably transfected clones, introduction of expression constructs into the human lung adenocarcinoma cell line A549 (RCB0098) was performed by electroporation. Approximately 1 ϫ 10 6 cells were transfected with 100 g of MEF cDNA in pCB6 linearized with ApaLI. Electroporations were performed using an ECM 600 apparatus (BTX Inc.) at 500 V and 1350 microfarads. Cells were then cultured on six-well plates at 1 ϫ 10 5 /well. At 50% confluency, 1 mg of G418 sulfate (Calbiochem)/ml was added. G418-resistant clones were picked 1 week later and analyzed individually or as pools of several hundred clones. Stably overexpressed cells were selected by Northern blotting.

FIG. 1. Transcriptional activation of the lys5A promoter in A549 cells (A), Caco-2 cells (B), and NIH3T3 cells (C) by the ETS gene family members.
The cells were cotransfected with the indicated ETS protein expression vectors and luciferase vectors containing the lys5A promoter (lys5A-(Ϫ100/ϩ10) or mutated lys5A-(Ϫ100/ϩ10)) or the parental pGL2-basic vector. Luciferase activity in the lysates was determined 48 h after transfection and is expressed as -fold activation over the pGL2-basic vector.
(Ϫ100 bp)E-M is the construct (Ϫ100 bp)WT with a mutated ETS site. It was prepared using a Transformer site-directed mutagenesis kit (CLONTECH) and BLP30-sense as a mutation primer. It contains Ϫ60 to Ϫ31 bp, with a mutated ETS site indicated in lowercase boldface letters, CCAGTCACATAAGAAGGtcGTGAAAAGATG.
(Ϫ50 bp)E-M is the construct (Ϫ50 bp)WT with a mutated ETS site. It was prepared using a Transformer site-directed mutagenesis kit and BLP30-sense as a mutation primer.
A full-length cDNA (1992 bp) for human MEF (20) was obtained by reverse transcription-PCR. This was done with a GeneAmp RNA PCR kit (Perkin-Elmer) and the following primers: 5Ј-primer, CGGGATC-CCGCATGGCTATTACCCTACAGC; and 3Ј-primer, GGAATTCCT-TATATGTCATGGGGCTCCATC. The PCR product was cloned into the pCR2.1 vector using the Original TA cloning kit (Invitrogen). After confirmation of the sequence, it was cloned into the BglII-HindIII site of pCB6 downstream of the cytomegalovirus promoter. Full-length human Ets-1 (1325 bp), Ets-2 (1410 bp; a gift from Dr. D. K. Watson), Elf-1 (1870 bp; cloned by us), PEA3 (a gift from Dr. J. A. Hassel), and ESE-1 (a gift from Dr. T. A. Libermann) were cloned into the KpnI-XbaI site of pCB6 downstream of the cytomegalovirus promoter. All constructs were verified by DNA sequencing.
For generation of antisense mRNAs of MEF and ESE-1, the cDNA in pCB6 was cut with EcoRI and religated to get a reverse orientation of MEF and ESE-1. The resulting plasmid was sequenced to verify insert orientation.
Electrophoretic Mobility Shift Assay-A549 cells (0.5-1 ϫ 10 6 ) were collected, washed with 10 ml of Tris-buffered saline, and pelleted by centrifugation at 1500 ϫ g for 5 min. The pellet was resuspended in 1 ml of Tris-buffered saline, transferred into an Eppendorf tube, and pelleted again by spinning for 15 s in a microcentrifuge. Tris-buffered saline was removed, and the cell pellet was resuspended in 400 l of cold buffer containing 10 mM HEPES (pH 7.9), 10 mM KCl, 0.1 mM EDTA, 0.1 mM EGTA, 1 mM dithiothreitol, and 0.5 mM phenylmethylsulfonyl fluoride by gentle pipetting. The cells were allowed to swell on ice for 15 min, after which 25 l of a 10% solution of Nonidet P-40 (Nakarai) was added, and the tube was vigorously vortexed for 10 s. The homogenate was centrifuged for 30 s in a microcentrifuge. The nuclear pellet was resuspended in 50 l of ice-cold buffer containing 20 mM HEPES (pH 7.9), 0.4 M NaCl, 1 mM EDTA, 1 mM EGTA, 1 mM dithiothreitol, and 1 mM phenylmethylsulfonyl fluoride, and the tube was vigorously rocked at 4°C for 15 min on a shaking platform. The nuclear extract was centrifuged for 5 min in a microcentrifuge at 4°C, and the supernatant (ϳ55 l) was frozen in aliquots at Ϫ80°C.
The following complementary oligonucleotides were synthesized and used in electrophoretic mobility shift assays: wild-type oligonucleotide (which contains lys5A promoter sequences from Ϫ50 to Ϫ31 bp), AA-GAAGGAAGTGAAAAGATG); and ETS mutant (which contains Ϫ50 to Ϫ31 bp, with a mutated ETS site indicated in lowercase boldface letters), AAGAAGGtcGTGAAAAGATG. The wild-type oligonucleotide was labeled with [␥-32 P]ATP (220 TBq/mmol) using T4 polynucleotide kinase. Binding assays were done by preincubating 5-10 g of crude nuclear extract, 2 g of poly(dI-dC) (Amersham Pharmacia Biotech), and 300 g/ml bovine serum albumin, with or without the indicated unlabeled DNA or a double-stranded oligonucleotide as a competitor, in 20 l of buffer (12 mM HEPES (pH 7.9), 4 mM Tris-HCl (pH 7.9), 1 mM EDTA, 1 mM dithiothreitol, and 12% glycerol) for 5 min at room temperature; 10,000 cpm (40 ng) of the labeled fragment was then added and incubated for 15 min at room temperature. The reaction products were analyzed by electrophoresis on a 4.5% polyacrylamide gel at 4°C, followed by autoradiography at Ϫ80°C.
Production of Recombinant MEF and Antisera-The MEF cDNA was subcloned into the pGEX2T plasmid, and GST-MEF protein was expressed in bacterial DH5␣. After induction of MEF protein expression using 1 mM isopropyl-␤-D-thiogalactopyranoside, cells were collected by centrifugation, resuspended in 500 mM NaCl and 20 mM EDTA, and disrupted by sonication; the soluble fraction was recovered after another centrifugation step. The expression of MEF protein in DH5␣ was confirmed by SDS-polyacrylamide gel electrophoresis.
To generate MEF antiserum, bacterially expressed GST-MEF protein was used to immunize a rabbit. For use in electrophoretic mobility shift assay experiments, IgG was purified using protein A-Sepharose (Amersham Pharmacia Biotech) as recommended by the manufacturer.
Western Blot Analysis-After parental A549 and stably transfected A549 cells were confluent, cells were cultured in serum-free medium for 1 week. The culture medium was harvested and concentrated with ammonium sulfate. The concentrated medium was analyzed by SDSpolyacrylamide gel electrophoresis. Samples were electroblotted onto a polyvinylidene difluoride membrane (Millipore Corp.). The membrane was blocked in PBST (phosphate-buffered saline containing 0.05% (v/v) Tween 20) containing 5% nonfat milk. After three washes in PBST, the membrane was incubated for 1 h with primary antibody (sheep polyclonal anti-human lysozyme, IBL). After an additional three washes in PBST, the membrane was incubated for 1 h with secondary antibody (peroxidase-conjugated rabbit anti-sheep IgG (HϩL) antibody, Jackson ImmunoResearch Laboratories, Inc.).

RESULTS
A screen of ETS transcription factors (Elf-1, Ets-1, Ets-2, PEA3, ESE-1, and MEF) revealed that despite the presence of an ETS consensus binding sequence on the lys5A promoter, most ETS proteins were unable to activate lysozyme transcription. The only ETS proteins active in this regard were MEF and ESE-1 (Fig. 1). MEF stimulated transcription more strongly than did ESE-1 in lung epithelial cells (A549), colon carcinoma cells (Caco-2), and skin fibroblasts (NIH3T3) (Fig. 1). That this occurred through MEF binding to the ETS consensus sequence was indicated by the fact that activation was inhibited when the ETS consensus sequence in the lysozyme regulatory region was changed from 5Ј-GGAA-3Јto 5Ј-GGTC-3Ј. The reason for the weak inhibition by the mutation in Caco-2 cells is unclear.
To determine whether or not MEF was endogenously present in the nuclei of epithelial cells and could thereby mediate endogenous lysozyme expression, we performed gel shift and supershift assays. As a probe, we used the radiolabeled lys5A promoter Ϫ50/Ϫ31 oligonucleotide. We observed a single band with retarded mobility after probe incubation with A549 cell nuclear extract (Fig. 2). That the band represented a specific DNA/protein interaction was evident from experiments showing competitive inhibition of band formation using excess unlabeled wild-type probe, but not the probe mutated at the ETS consensus binding site. That MEF was present in the complex was indicated by assays showing that anti-MEF antibody, but not preimmune serum or antibodies directed against human Ets-1, Ets-2, PU.1, or PEA3 (Santa Cruz Biotechnology) (data not shown), caused a supershift of the DNA-protein complex.
To determine whether or not endogenous MEF is required for basal activity of the lys5A promoter, we produced antisense MEF mRNA in A549 cells using expression plasmid pCB6 with MEF cDNA in reverse orientation. The presence of antisense mRNA was associated with an 80% inhibition of luciferase activity for lys5A-(Ϫ50/ϩ10) and 30% inhibition for lys5A-(Ϫ100/ϩ10) (Fig. 3). In addition, the presence of antisense ESE-1 mRNA with antisense MEF mRNA strongly inhibited luciferase activity for lys5A-(Ϫ100/ϩ10). This construct contains enhancer-like activity in the region at Ϫ94 to Ϫ66 bp (17), although the precise DNA sequence and cognate binding protein remain to be identified.
We obtained further evidence for the role of MEF in lysozyme transcription by stably overexpressing MEF in A549 cells. The stable transfectants are referred to as line 71. MEF overexpression strongly stimulated transiently transfected lys5A promoter activity if cells were transfected with wild-type constructs, but not with constructs mutated at the MEF DNAbinding site (Fig. 4A). In addition, MEF overexpression upregulated endogenous lysozyme mRNA and protein expression, but not mRNA encoding the housekeeping protein glyceraldehyde-3-phosphate dehydrogenase (Fig. 4, B and C). DISCUSSION The studies reported here indicate that the ETS protein MEF is present in lung epithelial cells and is required for lysozyme transcription in these cells. This is the first information regard- FIG. 4. Effects of stably overexpressed MEF on lys5A promoter activity (A), lysozyme steady-state mRNA (B), and lysozyme protein expression (C). A, the cells were transfected with luciferase constructs ((Ϫ100 bp)WT) containing the lys5A promoter as well as the parental pGL2-basic vector. Luciferase activity in the lysates was determined 48 h after transfection and is expressed as -fold activation over the activity of (Ϫ100 bp)WT in parental A549 cells. B, semiquantitative reverse transcription-PCR of lysozyme and glyceraldehyde-3phosphate dehydrogenase (GAPDH) transcripts was carried out as described under "Experimental Procedures." C, shown is the stably transfected MEF enhanced expression of endogenous lysozyme protein.
ing requirements for lysozyme expression in epithelial cells, as previous studies have focused on macrophages. Although lysozyme transcription in macrophages and epithelial cells is similar in requiring ETS protein transcription factors, the two cell types differ in their specific use of ETS proteins. Thus, whereas macrophages activate lysozyme transcription via PU.1, our data indicate that epithelial cells use, at least in part, the ETS family member MEF.
We found that in epithelial cells, MEF up-regulated not only the activity of a transiently transfected lys5A promoter, but also the transcription of the endogenous lysozyme gene and protein. Moreover, antisense experiments showed that inhibition of endogenous MEF in epithelial cells attenuated the baseline level of lysozyme expression. Taken together, these results indicate that MEF is required for base-line expression of lysozyme in epithelial cells.
The finding that antisense MEF constructs did not completely block lysozyme expression suggests that other factors are also involved in A549 cells. A strong candidate is the ETS family member ESE-1/ERT/ESX/ELF-3 (21-26) since antisense ESE-1 constructs also partly inhibited promoter activity, since it is known to exist in epithelial cells, and since ESE-1 overexpression was seen to up-regulate lys5A promoter activity in cotransfection experiments (Fig. 1). With respect to lysozyme up-regulation, we are aware of the existence of a second element (Ϫ100/Ϫ51) that is well conserved in the human lysozyme gene. Although the enhancer remains to be identified, MEF may require interaction with the enhancer-binding protein to activate lysozyme expression.
Our results showing a requirement for ETS family proteins in the expression of lysozyme in epithelial cells are consistent with previous data showing that the ETS protein PU.1 is required for lysozyme expression in chicken macrophages (18). The PU.1-binding site is present in an enhancer located 2.7 kilobases upstream of the transcription start site of chicken lysozyme. Mutation of this site abolishes enhancer activity in macrophages (18). Unclear at this point is the issue of whether epithelial cells in chicken, like those in mammals, use MEF and/or ESE-1 to activate the response element activated by PU.1 in macrophages.
A subtle contrast with the findings reported here are results regarding the mouse M lysozyme gene. This gene, although also regulated by ETS transcription factors, is flanked by ETS consensus sequences lying 3Ј, rather than 5Ј, to the gene coding region. The 3Ј-region containing ETS consensus sequence is co-extensive with DNase I hypersensitivity sites that undergo demethylation in lysozyme-expressing macrophages (27) and is therefore likely to be functionally important.
A major focus of our laboratory's research is the differentiation and function of respiratory tract epithelial cells. In this regard, it is interesting that the 5Ј-flanking regions of several genes selectively expressed in respiratory tract serous cells contain ETS protein-binding sites. Specifically, the motif is present flanking genes encoding proline-rich proteins (28), lactoferrin (29), the cystic fibrosis transmembrane conductance regulator (30), and secretory leukoprotease inhibitor (31). The presence of shared transcription factor-binding sites may denote the presence of a regulatory cascade that controls serous cell differentiation in a manner similar to that already described for more extensively studied tissues such as skeletal muscle (32).
In summary, these studies provide the first information identifying transcription factors controlling lysozyme expression in epithelial cells. As bacteria become progressively more resistant to exogenously applied antibiotics, information regarding mechanisms controlling innate immunity, such as that provided by epithelial lysozyme, may offer important new therapeutic approaches.