Bacterial Peptidoglycan-Induced tnf-α Transcription Is Mediated Through the Transcription Factors Egr-1, Elk-1, and NF-κB

Materials

Soluble PGN (sPGN), a polymeric uncross-linked PGN (average M_r = 125,000), released from Staphylococcus aureus grown in the presence of penicillin, was purified by vancomycin affinity chromatography and analyzed as before (23). Micrococcus luteus (American Type Culture Collection, Manassas, VA; ATCC 4698) was used as a prototypic Gram-positive nonpathogenic (i.e., easily eliminated by the host innate immune system) bacterium with unmodified unsubstituted PGN readily accessible on its surface. sPGN and Micrococcus contained <24 pg and <500 pg endotoxin/mg, respectively, determined by the Limulus lysate assay (23). Protein-free LPS from Salmonella minnesota Re 595 (ReLPS, a minimal naturally occurring endotoxic structure of LPS) obtained by phenol-chloroform-petroleum ether extraction, was purchased from Sigma (St. Louis, MO). sPGN, Micrococci, and LPS were used at optimal stimulatory concentrations (10 μg/ml, 40 μg/ml, and 10 ng/ml, respectively). All other chemicals were from Sigma-Aldrich (St. Louis, MO), unless otherwise indicated.

Cell culture and TNF-α assay

Murine macrophage RAW264.7 cell line, obtained from American Type Culture Collection, was cultured in DMEM with 10% defined FCS (HyClone Laboratories, Logan, UT; endotoxin content <6 pg/ml). Human monocytic cell lines THP-1 expressing hCD14 (THP-1/hCD14) were obtained from P. Tobias (The Scripps Research Institute, La Jolla, CA) and cultured as described previously (10). To determine whether there was any effect of serum on the activation of Egr-1 and Elk-1, Northern analysis of Egr-1 mRNA and phosphorylation of Elk-1 were initially performed in the presence and absence of serum. For serum-free experiments, all serum was removed from the cells 12 h before stimulation. There was no difference in the results between the serum and serum-free experiments; therefore, all data presented are from experiments carried out in the presence of serum. TNF-α concentrations in culture supernatants were determined using the L929 cytotoxicity assay as described previously (12).

Plasmids

The plasmids with serial deletions in the tnf-α promoter, pTNF(−615)Luc, pTNF(−295)Luc, pTNF(−95)Luc, and pTNF(−36)Luc, and the mutants CRE_m, AP-1_m, and AP-2_m in pTNF(−615)Luc were provided by Dr. J. Economou (24). The plasmids pTNF(−182)Luc, pTNF(−161)Luc, and the mutants Egr-1_m and κB3_m in pTNF(−615)Luc were provided by Dr. N. Mackman (18). The plasmids expressing dominant-negative forms of CREB (25), ATF-1 (25), and the constitutive repressor IκBΔN (11) have been described previously. The plasmid pSRE-Luc has five serum response element (SRE) domains upstream of a luciferase reporter gene driven by a basic promoter element (Stratagene, La Jolla, CA) and the plasmid expressing dominant-negative form of Elk-1 was provided by Dr. P. Shaw (26).

The plasmids pEgr1-chloramphenicol acetyl transferase (CAT) with full length egr-1 promoter, −606 to −7 bp of the putative transcription start site and various deletions of the egr-1 promoter, −480, −387, −235, and −180 bp cloned into the pCAT vector have been described (27). The region from −606 to −7 bp of the egr-1 promoter, amplified with the Egr-1 forward and reverse primers (Table I⇓), was subcloned into KpnI and HindIII sites of the vector pGL3 basic (Promega, Madison WI) and designated pEgr1(−600)Luc. The human egr-1 promoter contains an SRE site (SRE-1) from nucleotides −418 to −391 (27). Plasmids containing site specific mutations in the left Ets binding site (EBS-L_m), right EBS (EBS-R_m), and CArG (CArG_m) domains of SRE-1 were constructed in pEgr1(−600)Luc using the site directed mutagenesis kit from Clontech Laboratories (Palo Alto, CA) and the forward EBS-L_m, EBS-R_m, or CArG_m primers and the reverse trans AlwnI/SpeI primer (Table I⇓). All constructs were analyzed by restriction digests and by sequencing (on an ABI Prism 377XL automated DNA sequencer at the University of Chicago Cancer Center DNA Sequencing Facility, Chicago, IL).

Northern analysis

RAW264.7 cells were cultured at 0.35–0.4 × 10⁶/ml in 6-well plates (4.0 ml/well) for 16–20 h and were stimulated as indicated in the figure. Total RNA was prepared using the RNeasy Purification kit (Qiagen, Valencia, CA). Ten micrograms of total RNA from each sample was separated on 1% agarose formaldehyde gel, and transferred to Hybond nylon membrane (Amersham, Arlington Heights, IL). The probe used was a 3.2-kb EcoRI fragment from the Egr-1 cDNA clone zif/268 (ATCC 63027). The fragment was purified from an agarose gel using the QIAquick PCR purification kit (Qiagen), labeled with ³²P using the random primer labeling method (28) and purified on ChromaSpin columns (Clontech Laboratories). Prehybridization, hybridization, and washes were performed as before (12). The blots were stripped and reprobed with GAPDH as a control, obtained by polymerase chain reaction using total RNA as template and with primers described before (11).

EMSAs

Cells were cultured at 0.35–0.4 × 10⁶/ml in 24-well plates (1.0 ml/well) for 16–20 h and stimulated as indicated in the figures. Nuclear extracts were prepared as before (8). Five micrograms of nuclear proteins were incubated with ³²P-labeled oligonucleotides for 30 min at 22°C as described before (8). All samples were then separated on 6% nondenaturing polyacrylamide gels and the DNA-protein complexes were visualized by autoradiography. Double stranded oligonucleotides with sequences homologous to human tnf-α promoter (CRE, Egr-1, and κB3) or egr-1 promoter (SRE-1) and containing binding sites for the indicated transcription factors are listed in Table I⇑. To show specificity of binding to SRE-1 nuclear extracts were preincubated for 10 min at 22°C in the presence of 50-fold excess of unlabeled self oligonucleotide. For supershift experiments, 1 μg of Ab was preincubated with nuclear extracts for 10 min for Egr-1 and 20 min for all other transcription factors, before the addition of the labeled oligonucleotide. Abs against NF-κB p50 and p65 and CREB were from Upstate Biotechnology (Lake Placid, NY) and all other Abs were from Santa Cruz Biotechnology (Santa Cruz, CA).

Phosphorylation of Elk-1 and Western blots

RAW264.7 cells were seeded at 0.35–0.4 × 10⁶/ml in 24-well plates (1 ml/well) and cultured for 16–20 h. Cells were activated with the stimulants as indicated in the figures and then washed and lysed as before (12). Cell lysates were separated on 12% SDS-PAGE and transferred to Immobilon P (12). Phosphorylation of Elk-1 protein was determined by Western blotting with 0.5 μg/ml rabbit anti-pElk-1 Ab (Santa Cruz Biotechnology), and detected by the ECL system. This Ab specifically recognizes the Elk-1 protein phosphorylated at serine 383. Control blots were done with 0.5 μg/ml anti-Elk-1 (Santa Cruz Biotechnology) which recognizes the nonphosphorylated form of Elk-1.

Transfection and CAT and luciferase assays

RAW264.7 cells were cultured at 0.35–0.4 × 10⁶/ml in 48-well plates (0.5 ml/well) for 16–20 h and transfected with 200 μg/ml DEAE-dextran and 1 μg/ml DNA (optimal concentration) (25). In cotransfection experiments, when dominant-negative or mutant plasmids were not used, equivalent amounts of appropriate control vectors were included and, in addition, the total DNA concentration was brought up to 2.5 μg/ml with salmon sperm DNA. All plasmid DNA was prepared using the endotoxin-free DNA kit (Qiagen). Cells were allowed to recover for 24–48 h and then were left unstimulated or were stimulated as described in the figures. Lysates were prepared and assayed for luciferase activity using the Luciferase Reporter kit (Promega) and the Reporter Microplate Luminometer (Turner Instruments, Sunnyvale, CA), or for CAT activity as described (25).

sPGN-induced activation of tnf-α promoter requires CRE, Egr-1, and κB3 binding sites and transcription factors c-Jun, Egr-1, and NF-κB, but not CREB and ATF-1

To identify the transcription factors that regulate the expression of tnf-α in cells stimulated with bacterial cell wall components, we first tested activation of a tnf-α promoter construct, in macrophages transiently transfected with the plasmid pTNF(−615)Luc. This plasmid has 615 bp of the tnf-α promoter placed directly upstream of the luciferase coding region and the promoter activity is assayed based on the amount of luciferase produced. sPGN and LPS induced tnf-α promoter activity by an average of 6.3- and 8.9-fold, respectively, over unstimulated cells (Fig. 1⇓A). The increase in transcription in these cells is accompanied by an increase in the amount of TNF-α secreted (Fig. 1⇓B). To identify the regions of the tnf-α promoter that are involved in the regulation of transcription, we next analyzed the activity of constructs with serial deletions in the tnf-α promoter. In cells stimulated with sPGN, transcription of tnf-α gene was regulated by sequences in the region between −615 and −295 bp, and by sequences proximal to −182 bp (Fig. 1⇓C). The deletion between −615 and −295 bp reduced transcriptional induction of tnf-α in sPGN stimulated cells by 50% (Fig. 1⇓C), which indicates that sequences within this region are required for full induction of tnf-α in sPGN-stimulated cells. The loss of sequences between −182 and −95 further reduced tnf-α induction to levels that are no higher than unstimulated samples (Fig. 1⇓C). By contrast, in cells stimulated with ReLPS, only sequences proximal but not distal to −182 bp contain regulatory elements that were required for transcriptional induction of tnf-α (Fig. 1⇓C).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

sPGN induces transcription of tnf-α gene and secretion of TNF-α, and transcriptional induction requires binding sites CRE, Egr-1, and κB3. RAW264.7 cells were transfected with pTNF(−615)Luc or empty vector pXP1 (A); the indicated deleted promoter constructs or empty vector pXP1 (C); tnf-α promoter mutants (D); and pTNF(−615)Luc plus dominant-negative forms of ATF-1, CREB, or the constitutive repressor IκBΔN (E). A and C–E, Cells were stimulated with sPGN or LPS and cell lysates were assayed for luciferase activity. B, RAW264.7 cells were stimulated with sPGN or LPS or left unstimulated for 16 h and supernatants were assayed for TNF-α. The results are means of duplicate samples from one of three similar experiments (A); means of three experiments (B); and means of the percentage of pTNF(−615)Luc activity (C and D) or the percentage of control vector from three experiments (E). The percent of control vector = (mean cpm stimulated with dominant negative − mean cpm unstimulated with dominant negative) × 100/(mean cpm stimulated with empty vector − mean cpm unstimulated with empty vector).

We next identified specific sequences that are required for tnf-α induction in sPGN-stimulated cells. In this study, we focused on the region between −182 and −95 bp. We are currently analyzing the region between −615 and −295 bp and the results will be presented separately. However, we do know that the NF-κB site at −587 bp is not involved in transcriptional induction of tnf-α in sPGN-stimulated cells (data not shown). The region proximal of −182 bp contains binding sites for several different transcription factors, including an Egr-1 site at −172 bp, a CRE site at −106 bp, a κB3 site at −97, and an AP-1 site at −66 bp. To determine whether these sites are required for tnf-α induction in sPGN-stimulated cells, we tested for tnf-α promoter activity using promoter-luciferase construct with site specific mutations in the binding sites for these transcription factors. Mutations within the CRE, Egr-1, and κB3, but not the AP-1 site, greatly reduced tnf-α induction for both sPGN- and LPS-stimulated cells (Fig. 1⇑D). To confirm the role of specific transcription factors in the regulation of tnf-α expression, we tested the effect of dominant-negative forms of the transcription factors ATF-1 and CREB and a constitutive repressor of the inhibitory protein IκB, IκBΔN. Dominant-negative ATF-1 and CREB did not inhibit tnf-α promoter activity in sPGN- or LPS-stimulated cells (Fig. 1⇑E). These results along with the data in Fig. 1⇑D indicate that the CRE site on the tnf-α promoter is required for tnf-α induction; however, the transcription factors ATF-1 and CREB are not involved in this induction. The constitutive repressor IκBΔN strongly inhibited tnf-α induction (Fig. 1⇑E), which confirms the role for NF-κB in tnf-α transcription in sPGN- and LPS-activated macrophages.

To determine whether proteins from sPGN- and LPS-stimulated nuclear extracts bind to the CRE, Egr-1, and κB3 sites on the tnf-α promoter, we performed EMSAs. Oligonucleotides homologous to the region of the tnf-α promoter containing binding sites for CRE, Egr-1, and κB3 were used for these assays. Both sPGN and LPS induced time-dependent binding to the CRE, Egr-1, and κB3 sites (Fig. 2⇓, A–C). We next determined by supershift assays which nuclear proteins from sPGN- and LPS-stimulated macrophages bound these motifs. The CRE sequence can bind a variety of dimers from the CRE/ATF and the AP-1 family of transcription factors and we tested Abs against CREB, ATF-1, and c-Jun. Abs against c-Jun, but not against CREB and ATF-1, decreased binding of proteins to the CRE oligonucleotide, indicating that c-Jun protein, but not CREB and ATF-1 proteins, bind CRE. Abs against CREB and ATF-1 do have the capacity to bind to the appropriate transcription factors and induce inhibition of binding to appropriate oligonucleotides or supershift of the oligonucleotide-protein complexes (25).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

sPGN induces binding of c-Jun, Egr-1, and NF-κB to tnf-αpromoter. RAW264.7 cells were stimulated with sPGN or LPS and nuclear extracts were incubated with radioactively labeled oligonucleotides homologous to the tnf-α promoter with binding sites for CRE (A), Egr-1 (B), and κB3 (C). For supershift experiments, nuclear extracts were preincubated with the indicated Abs. The results are from one of two similar experiments.

An Ab to Egr-1 decreased binding of nuclear proteins to the Egr-1 motif from both sPGN- and LPS-stimulated nuclear extracts (Fig. 2⇑B). To identify the proteins bound to the κB3 site, we tested Abs against p50, p65, and c-rel. Abs against p50 and p65 supershifted the protein-DNA complex (Fig. 2⇑B). These results indicate that both sPGN and LPS stimulation induces binding of the proteins c-Jun, Egr-1, and NF-κB p50 and p65 to the CRE, Egr-1, and κB3 sites, respectively, on the tnf-α promoter.

To determine whether the same transcription factors are similarly activated by sPGN in human monocytes, we performed EMSAs with nuclear extracts from the human monocytic cell line THP-1/hCD14. Nuclear proteins from both sPGN- and LPS-stimulated cells induced time-dependent binding to the CRE, Egr-1, and κB3 sites (Fig. 3⇓).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

sPGN induces binding of proteins to the CRE, Egr-1, and κB3 sites in human monocytic THP-1 cells. THP-1/hCD14 were stimulated with sPGN or LPS and nuclear extracts were incubated with radioactively labeled oligonucleotides homologous to the tnf-α promoter with binding sites for CRE, Egr-1, and κB3.

sPGN and Micrococci induce transcription of egr-1 gene and this induction requires SRE-1 sequence

Our results show that tnf-α induction in sPGN-stimulated macrophages is regulated by the transcription factors c-Jun, Egr-1, and NF-κB. We have previously studied the activation and regulation of c-Jun (25) and NF-κB (11); therefore, we investigated regulation of Egr-1 in this study. Macrophages activated with sPGN and Micrococci showed time-dependent induction of egr-1 transcription; levels of egr-1 transcripts increased within 15 min and reached a maximum by 30 min (Fig. 4⇓A). LPS, as shown previously (29), also induced expression of the egr-1 gene. To determine the regulation of egr-1 transcription in sPGN-stimulated macrophages, we first constructed and tested the expression of an egr-1 promoter luciferase plasmid, pEgr1(−600)Luc, in macrophages transiently transfected with the plasmid. This plasmid contains −606 to −7 bp of the egr-1 promoter fused directly upstream of the coding sequence for luciferase. sPGN and LPS induced egr-1 promoter activity by an average of 5.5- and 5.7-fold, respectively, over unstimulated cells (Fig. 4⇓B). To identify regulatory elements within the egr-1 promoter that are required for sPGN- and LPS-induced transcription of the gene, we next analyzed the expression of deletion constructs. A deletion from −600 to −480 bp resulted in a slight increase in egr-1 promoter activity, however, a deletion from −480 to −387 bp reduced sPGN- and LPS-induced egr-1 transcription by 77 and 86%, respectively (Fig. 4⇓C). Further deletions of the egr-1 promoter did not result in any additional significant changes of sPGN- and LPS-induced egr-1 transcription (Fig. 4⇓C). Our results indicate that sequences between −480 and −387 bp of the egr-1 promoter are required for sPGN- and LPS-induced expression of the egr-1 gene.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

sPGN, LPS, and Micrococci induce transcription of egr-1 gene and this induction requires sequences from −480 to −387 nucleotides in egr-1 promoter. A, RAW264.7 cells were stimulated with sPGN, LPS, or Micrococci, and total RNA was analyzed for Egr-1 transcripts by Northern blots. B and C, RAW264.7 cells were transfected with pEgr1(−600)Luc or empty vector pGL3 (B), and pEgr1(−600)Luc or the indicated deleted constructs (C). Cells were stimulated and cell lysates were assayed as described in Fig. 1⇑. The results for A are from one of two similar experiments; the results of B are means of duplicate samples from one of three similar experiments; and the results of C were calculated as in Fig. 1⇑D and are means from two experiments.

The region between −480 and −387 bp of the egr-1 promoter contains SRE-1 from −418 to −391 bp with a core CArG box and EBS on either side. To determine whether SRE-1 is required for sPGN- and LPS-induced egr-1 transcription, we introduced site specific mutations within the SRE-1 domain of pEgr1(−600)Luc. We constructed plasmids with mutations within the CArG box and either of the two flanking EBSs, called CArG_m, EBS-L_m, and EBS-R_m recepectively (Fig. 5⇓A). All three mutants showed a decrease in the induction of egr-1 expression, with the greatest loss caused by the mutation in the EBS sequence on the right of the CArG box in both sPGN- and LPS-stimulated cells (Fig. 5⇓B).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 5.

sPGN- and LPS-induced expression of egr-1 gene requires SRE-1 binding site. A, Sequences of the wild type and mutant SRE-1. B, RAW264.7 cells were transfected with wild type or mutant pEgr1(−600)Luc DNA. Cells were stimulated and cell lysates were assayed as described in Fig. 1⇑. The results are means from two experiments.

To identify proteins from sPGN-stimulated cells that bind the SRE-1 domain on the egr-1 promoter, we did gel shift assays using an oligonucleotide homologous to this region of the egr-1 promoter. Nuclear proteins from sPGN- or LPS-stimulated or unstimulated macrophages (Fig. 6⇓) bound the SRE-1 domain resulting in the formation of at least five different DNA complexes. These protein-DNA complexes were specific for SRE-1 as unlabeled SRE-1 oligonucleotide completely eliminated the formation of these complexes with the labeled SRE-1 (Fig. 6⇓). Abs against Elk-1, a member(s) of the Ets family of transcription factors, and against serum response factor (SRF) supershifted or reduced binding of one or more of the protein-DNA complexes (Fig. 6⇓), which indicates that these transcription factors bind SRE-1. Higher concentrations of these Abs had the same effect. These results indicate that SRE-1 is required for sPGN- and LPS-induced expression of the egr-1 gene and this induction may be mediated by the transcription factors SRF and Elk-1.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 6.

sPGN and LPS induce binding of Elk-1 and SRF to SRE-1 region of egr-1 promoter. RAW264.7 cells were stimulated or left unstimulated and nuclear extracts were prepared and incubated with radioactively labeled oligonucleotide homologous to the SRE-1 region of the egr-1 promoter in the presence and absence of excess unlabeled self competitor oligonucleotide. In supershift experiments nuclear extracts were preincubated with the indicated Abs, and assayed as in Fig. 2⇑. The arrows indicate the shift in the DNA-protein complexes upon Ab binding. The results are from one of three similar experiments.

sPGN induces phosphorylation and activation of Elk-1, and dominant-negative Elk-1 inhibits sPGN-induced transcription of egr-1 gene

Elk-1 is activated by phosphorylation at several different serine residues, including serine 383. To determine whether sPGN and LPS induce phosphorylation of Elk-1, we used an Ab that specifically recognizes Elk-1 phosphorylated at S383. In these experiments we also tested Micrococcus as a prototypic, Gram-positive bacterium. All three stimulants induced rapid but transient phosphorylation of Elk-1 (Fig. 7⇓A, upper panel). Identical samples analyzed with an Ab against both phosphorylated and nonphosphorylated Elk-1 showed equal amounts of the protein in all lanes (Fig. 7⇓A, lower panel). To determine whether phosphorylation results in functional activation of Elk-1, we tested the expression of a reporter plasmid that is regulated by SRE domains, pSRE-Luc. Both sPGN and LPS induced luciferase activity by 6.8- and 4.6-fold, respectively, in cells transfected with pSRE-Luc (Fig. 7⇓B), and this induction was inhibited 80 and 50%, respectively, by dominant-negative Elk-1 (Fig. 7⇓C). These results indicate that the induction of SRE activity in sPGN-stimulated cells is mostly due to Elk-1, whereas, the induction in LPS-simulated cells is only in part due to Elk-1. To confirm that sPGN- and LPS-induced expression of egr-1 is regulated by Elk-1, we tested the effect of dominant-negative Elk-1 on the expression of egr-1 in sPGN- and LPS-stimulated cells. Dominant-negative Elk-1 inhibited egr-1 induction by 70% in both sPGN- and LPS-stimulated cells (Fig. 7⇓D). These results indicate that increased expression of the egr-1 gene in macrophages stimulated with sPGN and LPS is regulated by the SRE-1 domain the egr-1 promoter and that the transcription factor Elk-1 is in part responsible for this activation.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 7.

sPGN and LPS induce phosphorylation of Elk-1, and dominant-negative Elk-1 inhibits transcription of egr-1 gene. A, RAW264.7 cells were stimulated as in Fig. 3⇑ and cell lysates were separated by SDS-PAGE and analyzed by Western Blots for phosphorylation of Elk-1 using an anti-phosphorylated Elk-1 Ab or total (phosphorylated and nonphosphorylated) Elk-1 protein using an anti-Elk-1 Ab. The results are from one of three similar experiments. B–D, RAW264.7 cells were transfected with pSRE-Luc or empty vector pCMV (B), pSRE-Luc and dominant-negative Elk-1 or its control vector (C), and pEgr1(−600)Luc and dominant-negative Elk-1 or its control vector (D). The results for B are means of duplicate samples from one of three similar experiments; the results for C and D were calculated as for Fig. 1⇑D and are means from two experiments.