Shiga Toxin Type 1 Activates Tumor Necrosis Factor-α Gene Transcription and Nuclear Translocation of the Transcriptional Activators Nuclear Factor-κB and Activator Protein-1

THE SHIGA TOXIN family of bacterial protein toxins consists of Shiga toxin (Stx), produced by Shigella dysenteriae serotype 1, and a group of closely related toxins designated Stx1, 2, 2c, and 2e produced by Escherichia coli.All members of the Stx family are AB₅ holotoxins, consisting of a single enzymatic A-subunit in noncovalent association with a pentamer of identical B-subunits. Stx binding to susceptible cells is mediated by B-subunit interaction with the neutral glycolipid globotriaosylceramide (Gb₃).1,2 The toxins are internalized by a receptor-mediated endocytic mechanism and undergo retrograde translocation through the Golgi stacks to the endoplasmic reticulum, where a fragment of the A-subunit may traverse the ER membrane and associate with ribosomes.3 The A-subunit cleaves a single adenine residue located in a prominent loop structure of the 28 S rRNA component of eukaryotic ribosomes, and depurination results in protein synthesis inhibition.4 5 

Stxs may cause disease in humans by damaging intestinal, renal, and central nervous system capillary blood vessels, resulting in an exacerbation of colonic ulceration and bloody diarrhea and the development of acute renal failure, seizures, and death.6Glomerular vascular lesions caused by Stxs are characterized by endothelial cell swelling and detachment from glomerular basement membranes and the deposition of micro-thrombi within glomeruli. However, a direct cytotoxic effect of purified Stxs on human vascular endothelial cells in vitro was minimal, unless the target cells were cultured in the presence of the proinflammatory cytokines tumor necrosis factor-α (TNF-α) or interleukin-1β (IL-1β).7,8 Subsequently, van der Kar et al9showed that TNF-α acts on human endothelial cells to upregulate the expression of the toxin-binding glycolipid Gb₃, suggesting that the host response to Stxs may participate in the development of vascular damage. We found that human vascular endothelial cells do not synthesize TNF-α or IL-1β when treated with purified Stxs in vitro, suggesting that other cell types may be necessary to produce the cytokines involved in glycolipid modulation on target cells. Murine peritoneal macrophages, human peripheral blood monocytes, and human monocytic cell lines are relatively resistant to the cytotoxic action of Stxs, express low levels of membrane Gb₃, and respond to toxin stimulation by secreting TNF-α, IL-1, and IL-6.10-12 However, the mechanism by which Stxs mediate cytokine induction is not known.

TNF-α gene expression is tightly controlled, including regulation of transcription initiation, mRNA processivity, and translational and posttranslational regulatory controls. In the experiments reported here, we investigated Stx1-mediated human monocyte TNF protein production, TNF-α mRNA induction, and the cellular signaling events that may be involved in cytokine induction. We focused our studies on the transcriptional activators nuclear factor-κB (NF-κB) and activator protein-1 (AP-1), protein complexes that have been shown to translocate to the nucleus and upregulate transcription in response to a variety of stimuli.13 14 

Human peripheral blood monocytes (PBMn) were derived from blood collected from healthy volunteers. Mononuclear cells were separated by Histopaque 1077 (Sigma, St Louis, MO) gradient centrifugation, and plastic nonadherent cells were removed after 1 hour of incubation at 37°C. The human myelogenous leukemia cell line THP-115 was purchased from ATCC (Rockville, MD). All monocytic cells were maintained in RPMI-1640 (GIBCO-BRL, Grand Island, NY) supplemented with penicillin (100 U/mL), streptomycin (100 μg/mL), amphotericin B (2.0 μg/mL), and 10% fetal bovine serum (FBS; Hyclone Laboratories, Logan, UT) at 37°C in humidified 5% CO₂. Differentiated THP-1 cells have been shown to share many of the physiological functions of primary monocyte-derived macrophages.16 In all the experiments reported here, THP-1 cells (1 × 10⁶ cells/mL) were induced to differentiate to the mature macrophage-like state by treatment with 12-0-tetradecanoylphorbol-13-acetate (TPA; Sigma) at 50 ng/mL for 48 hours in 100-mm culture dishes. Differentiated, plastic-adherent cells were washed twice with cold Dulbecco's phosphate-buffered saline (PBS; GIBCO-BRL) and incubated with fresh medium lacking TPA for 3 to 4 days with daily medium changes before use in assays. Although TPA treatment of THP-1 cells resulted in transient increases in TNF production and NF-κB nuclear translocation, these activities returned to baseline (unstimulated) levels 3 to 4 days after TPA treatment (data not shown). The murine fibroblast cell line L929 was maintained in Iscove's modified Dulbecco's medium (IMDM; Celox Corp, Hopkins, MN) containing 5% FBS at 37°C in humidified 5% CO₂.

Purified Stx1 was prepared as previously described.10Before use, toxin preparations were passed through ActiClean Etox columns (Sterogene Bioseparations, Arcadia, CA) to remove traces of endotoxin contaminants. Purified pentameric Stx1 B-subunits were the kind gift of Dr David Acheson (Tufts University School of Medicine, Boston, MA). Purified lipopolysaccharides (LPS) derived fromEscherichia coli 0111:B4 were purchased from Sigma. Murine monoclonal IgM antibody pK002, directed against Gb_3, was purchased from Accurate Chemical Corp (Westbury, NY).

TNF bioactivity in THP-1 supernatants from untreated control cells and cells treated with Stx1 and/or LPS was determined by the lysis of actinomycin-D (act-D; Sigma) –treated L929 murine fibroblasts as previously described.10 Briefly, L929 cells were cultured in 96-well microtiter plates at a density of 2 × 10⁵cells/mL in IMDM supplemented with 5% FBS at 37°C in 5% humidified CO₂. Dilutions of macrophage supernatants or recombinant human TNF-α (R&D Systems, Minneapolis, MN) were made in IMDM such that the final act-D concentration was 1.0 μg/mL. The dilutions were added to 5 replicate wells of L929 cells in microtiter plates and incubated for 18 hours. Twenty-five microliters of a 5.0 mg/mL stock solution of 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT; Sigma) was added to each well and incubated at 37°C for 2 hours. The cells were lysed and the formazan dye was extracted from the cells with 50% N,N-dimethylformamide and 20% sodium dodecyl sulfate (SDS). A₅₇₀ was measured (Dynatech MR5000; Dynatech Laboratories, Chantilly, VA) and L929 survival values were determined. TNF bioactivity in the macrophage supernatants was calculated by substituting the corresponding sample values in linear regression equations generated by the recombinant human TNF-α standard curve. Direct treatment of L929 cells with act-D and Stx1 consistently resulted in less than 10% cytotoxicity compared with untreated cells. Statistical analyses using Student's paired t-test were performed with Microsoft Excel version 5.0 software (Microsoft Corp, Redmond, WA).

Total cellular RNA was isolated by the acid guanidinium isothiocyanate extraction method17 using Ultraspec II RNA isolation kits (Biotecx Laboratories, Houston, TX). RNA purity was assessed by OD₂₆₀/OD₂₈₀ readings. Ten micrograms of RNA per lane was electrophoresed through 0.8% agarose-2 mol/L formaldehyde gels in 1× MOPS running buffer at 50 V for 2 to 3 hours. RNA was then transferred onto positively charged nylon membranes (GeneScreen Plus; NEN Dupont, Boston, MA) using the Turboblotter Rapid Downward Transfer System (Schleicher & Schuell, Keene, NH) and cross-linked by exposing the membrane to UV light (254 nm) for a total dose of 120 mJ/cm² using a UV cross-linker (Bio-Rad, Hercules, CA). A human TNF-α probe (5′-ATCTCTCAGCTCCACGCCATTGGCCAGGAG-3′; Clonetech, Palo Alto, CA) was end-labeled with [γ³²P]-ATP as per the manufacturer's instructions. 18S RNA antisense control template (Ambion, Inc, Austin, TX) was random prime labeled using Megaprime DNA labeling kits (Amersham Corp, Arlington Heights, IL). After labeling, unincorporated nucleotides were removed using G-25 Sephadex columns (5 prime 3 prime, Inc, Boulder, CO). The membranes were prehybridized for 15 minutes at 42°C for the human TNF-α probe or 65°C for the 18S RNA probe in 7 to 10 mL of Rapid-hyb Hybridization Buffer (Amersham). After prehybridization, 10⁶ cpm TNF-α probe per milliliter of hybridization buffer was added and the hybridization was performed for 2 to 4 hours at 42°C. The membranes were washed in 2× SSC, 0.1% SDS for 10 minutes at room temperature and exposed to a phosphoimager screen overnight. The screen was analyzed on a phosphoimager (Molecular Dynamics, Sunnyvale, CA) and the sums of counts above background were calculated using ImageQuant software (Molecular Dynamics). The membranes were stripped by boiling in 0.1× SSC/0.1% SDS twice for 15 minutes and reprobed at 65°C with 18S antisense RNA as an internal control of RNA loading. Ratios of counts of TNF-α mRNA:18S RNA were calculated. To minimize interassay variability, relative levels of TNF-α mRNA from three separate experiments are shown by expressing the ratios as percentages above basal levels using the formula: ([stimulated ratio of counts TNF-α:18S] − [unstimulated ratio of counts TNFα:18S])/(unstimulated ratio of counts TNFα:18S) × 100.

TPA-differentiated THP-1 cells (1 × 10⁶/mL) were stimulated with different concentrations of Stx1 (10 to 800 ng/mL) for dose-response experiments and with 400 ng/mL of Stx1 for kinetics studies. The cells were washed twice with cold PBS before nuclear extract preparation. Nuclear extracts were prepared according to the methods of Dignam et al18 and Schreiber et al.19 Protein concentrations of nuclear extracts were determined by the Bradford method (Pierce Chemical Co, Rockford, IL²⁰). A NF-κB binding site-specific oligonucleotide containing two tandemly repeated HIV-1 long terminal repeat enhancers (5′-ATCAGGGACTTTCCGCTGGGGACTTTCCG-3′²¹) and a second mutant oligonucleotide lacking the NF-κB binding sites (5′-AGGATGGGAGTGTGATATATCCTTGAT-3′) were synthesized. An AP-1 consensus sequence oligonucleotide containing a binding site for C-Jun homodimer and Jun/Fos heterodimer complexes (5′-CGCTTGATGACTCAGCCGGAA-3′) and a corresponding AP-1 oligonucleotide with a CA to TG substitution in the AP-1 binding motif were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). All oligonucleotides were end-labeled with [³²P]-ATP using T4 polynucleotide kinase and unincorporated [³²P]-ATP was removed with Sephadex G-25 spin columns. DNA-protein interactions were performed by incubating 5.0 μg of nuclear extract with 20,000 cpm of [³²P]-end labeled double-stranded NF-κB site-specific or AP-1 site-specific probe in the presence of 1.0 to 2.0 μg of poly [dI-dC] (Pharmacia, Piscataway, NJ) in binding buffer (10 mmol/L Tris HCl [pH 5], 10% glycerol, 1.0 mmol/L EDTA, 40 mmol/L KCl, 1.0 mmol/L dithiothreitol, and 4.0 mmol/L MgCl₂) for 30 minutes at room temperature. In some experiments, nuclear extracts were incubated with 25 or 50 molar excess of unlabeled probe or with end-labeled mutated oligonucleotide with substitutions in binding motifs to examine the specificity of NF-κB or AP-1 binding to the DNA. After incubation, samples were loaded onto nondenaturing 4% polyacrylamide gels (acrylamide:bis-acrylamide 30:1 [wt:wt]) and electrophoresis was performed using 0.25× Tris-borate running buffer (0.089 mol/L Tris-borate, 0.089 mol/L boric acid, and 0.002 mol/L EDTA [pH 8]) at 180 V for 1 to 2 hours at 4°C. The gels were dried, radiolabeled bands were visualized by phosphoimager, and counts were analyzed using ImageQuant sofware.

To identify specific NF-κB proteins involved in DNA binding, 5.0 μg of nuclear extracts was incubated with 1.0 μL of rabbit antisera directed against human p50, p65, Rel B, or c-Rel proteins (kind gift of Dr Nancy R. Rice, National Cancer Institute, Frederick Cancer Research and Development Center, Frederick, MD²²) for 10 minutes at room temperature. After incubation, labeled NF-κB specific probe was added and the reaction was allowed to proceed for 30 minutes at room temperature before electrophoretic mobility shift assays were performed. Specificity of super-shifts was tested by adding 25 to 50 molar excess of the NF-κB binding site-specific double-stranded unlabeled oligonucleotide before incubation with the labeled probe.

Western immunoblot analyses were performed using 5.0 μg of cytoplasmic extracts prepared from Stx1-treated cells by the method of Dignam et al.18 Proteins were resolved in 10% tris-glycine gels and electrotransferred to nitrocellulose membranes. The membranes were blocked with 5% bovine serum albumin in PBS/0.1% Tween and probed with rabbit polyclonal anti–IκB-α antisera (kind gift of Dr Nancy Rice23). Primary antibody binding was detected by using peroxidase-conjugated antirabbit Ig antibodies. Blots were developed by the addition of peroxidase substrate with enhancement by ECL solution (Amersham) and exposed to blue sensitive films (Midwest Scientific, St Louis, MO). Bands were quantitated by densitometric scanning using Alpha-imager 3.21 software (Innotech, San Leandro, CA). Correlations between NF-κB translocation and Iκ-B degradation were evaluated by the Pearson product moment correlation test.

We previously demonstrated that human PBMn and differentiated THP-1 cells were relatively insensitive to the cytotoxic action of purified Stxs, expressed low levels of Gb₃, and respond to Stxs by secreting proinflammatory cytokines.11 The dose response of TNF production by differentiated THP-1 cells incubated with serial dilutions of purified Stx1 is depicted in Fig 1. Stx1 appeared to be a less potent inducer of TNF production in comparison to LPS, because treatment of THP-1 cells with 800 ng/mL Stx1 generated approximately 75% of the soluble TNF bioactivity stimulated by 200 ng/mL of LPS. The holotoxin molecule was necessary to induce TNF secretion, because Gb₃binding by purified Stx1 B-subunits (400 ng/mL) or by an anti-Gb₃ monoclonal antibody (10 μg/mL) did not trigger TNF synthesis and secretion above basal levels (∼25 pg/mL). Northern blot analyses showed that the dose-dependent production of soluble TNF correlated with increased levels of TNF-α mRNA transcripts isolated from toxin-treated cells (Fig 2). Thus, increased TNF production stimulated by Stx1 is mediated, at least in part, at the transcriptional level.

The kinetics of TNF production were examined by incubating THP-1 cells with purified Stx1, LPS, or both for varying timepoints and then measuring TNF bioactivity in macrophage supernatants using the L929 cytotoxicity assay as described in the Materials and Methods. Treatment of THP-1 cells with Stx1 or LPS alone resulted in the rapid induction of TNF expression, with bioactivity peaking between 3 and 6 hours (Table 1). Stx1- and LPS-induced TNF activity then decreases to 72% to 77% of peak values over the last 6 hours of the experiment. Treatment of THP-1 cells with the combination of Stx1 and LPS resulted in synergistically increased induction of soluble TNF bioactivity compared with treatment with either stimulant alone. However, the kinetics of TNF protein production were similar to that of Stx1 or LPS alone (Table 1).

Northern blot analyses were performed to correlate the kinetics of TNF-α gene transcriptional activation with the appearance of TNF bioactivity in macrophage supernatants. Maximal TNF-α mRNA levels after Stx1 or LPS treatment of THP-1 cells preceded maximal soluble TNF bioactivity, peaking at 2 hours and 0.5 hours, respectively (Fig 3). Whereas LPS stimulated a transient increase in TNF-α transcripts, with mRNA levels returning to basal values within 3 hours, Stx1 stimulated a prolonged elevation of TNF-α mRNA. When both Stx1 and LPS were used as stimulants, peak levels of TNF-α mRNA were detected at 2 hours, and the levels of induced transcripts were higher than the levels induced by Stx1 or LPS alone. The kinetics of LPS + Stx1 mediated TNF-α mRNA induction more closely resembled that induced by treatment of THP-1 cells with Stx1 alone in that mRNA levels remained elevated over the course of the experiment.

To elucidate the mechanism(s) by which Stx1, a potent protein synthesis inhibitor in toxin-sensitive cells, induces the transcriptional activation of the TNF-α gene in PBMns and THP-1 cells, we measured the nuclear translocation of two transcriptional activator complexes, NF-κB and AP-1, in response to toxin treatment. Treatment of macrophages with LPS or pharmacologic agents, such as phorbol diesters, is known to induce NF-κB and AP-1 nuclear translocation, and both NF-κB and AP-1 are involved in the regulation of many genes encoding cytokines, cytokine receptors, and acute-phase proteins involved in inflammatory responses.13,14 PBMns and differentiated THP-1 cells were incubated with increasing concentrations of Stx1 for 2 hours. Nuclear extracts were prepared and incubated with a radiolabeled oligonucleotide containing two tandem NF-κB binding sites.21 Translocation of NF-κB from the cytosol to the nucleus after toxin exposure was assessed by electrophoretic mobility shift in nondenaturing polyacrylamide gels associated with the binding of NF-κB to the radiolabeled target DNA as outlined in the Materials and Methods. Nuclear extracts prepared from untreated control cells contained low basal levels of NF-κB binding activity (Fig 4). However, upon Stx1 treatment, THP-1 cells (Fig 4A) and PBMn (Fig 4B) displayed concentration-dependent increased NF-κB binding activity. Stx1 concentrations as low as 100 ng/mL induced detectable NF-κB binding activity, which reached a maximum at 800 ng/mL Stx1. A control oligonucleotide containing mutated NF-κB binding sites and competition with excess unlabeled NF-κB-specific oligonucleotide were used to determine the specificity of DNA binding.

Earlier studies have shown that the translocation of an NF-κB complex to the nuclei of LPS-treated THP-1 cells occurs within 30 minutes.24 To investigate the time course of nuclear localization of NF-κB in response to Stx1, TPA-differentiated THP-1 cells were stimulated with 400 ng/mL of Stx1 for various time points and nuclear extracts analyzed for NF-κB binding. In accordance with the kinetics of Stx1-mediated soluble TNF activity and TNF-α mRNA induction, inducible NF-κB binding activity was detectable at 60 minutes and was maximal by 120 minutes (Fig5A and C). The response began to decrease by 3 hours after Stx1 treatment; however, the level of NF-κB induction remained elevated compared with basal values. Cold oligonucleotide competition and mutant oligonucleotide treatment demonstrated that NF-κB binding activity was specific. The kinetics of the loss of immunoreactive IκB-α in the cytoplasm were directly related to the kinetics of NF-κB nuclear translocation (r = −.98, P ≥ .01; Fig 5B and C).

The transcriptional activator complex NF-κB may be composed of homodimeric or heterodimeric proteins designated p50, p65 (relA), c-rel, relB, and p52. All members of the NF-κB/rel family share an approximately 300 amino acid region of homology with the c-Rel protooncogene that is essential for complex dimerization and nuclear translocation.13 To assess the composition of the nuclear factors translocating to nuclei and binding NF-κB sites in response to Stx1 treatment of THP-1 cells, a panel of rabbit antibodies against specific proteins of the human NF-κB/Rel family was used in electrophoretic mobility shift assays.22 Nuclear extracts incubated with p50 and p65 antisera displayed an altered electrophoretic mobility pattern (supershift) consistent with the binding of the antisera to specific NF-κB proteins (Fig 6). No shifts in electrophoretic mobility were observed with either relB or C-rel antisera (data not shown). Increasing the concentration of anti-p50 or anti-p65 antibodies did not change the supershift pattern. Supershifts observed with p50 and p65 antisera were specific, because the patterns could be eliminated by treatment with excess unlabeled oligonucleotide (Fig 6, lanes 4 and 8).

To determine whether AP-1 complex translocation is associated with Stx1 treatment of human monocytes, we examined AP-1 nuclear translocation by electrophoretic mobility shift assay using radiolabeled target DNA containing Jun/Jun and Fos/Jun binding motifs. Treatment of cells with 400 ng/mL of Stx1 for 2 hours resulted in translocation of AP-1 into nuclei (Fig 7). The mutant oligonucleotide binding (lane 6) and cold oligonucleotide competition (lane 7) experiments showed that the DNA-protein interaction is AP-1 specific. TPA and LPS treatments of THP-1 cells were used as the positive controls of AP-1 translocation.

Given the central role of the host response in the development of Stx-mediated disease, it would be useful to better understand the mechanisms by which Stxs trigger cytokine induction to devise interventional strategies limiting the development of life-threatening sequelae. We show here that the incubation of the THP-1 cell line with purified Stx1 resulted in a dose-dependent increase in TNF bioactivity detectable in cell supernatants. Northern blot analyses showed that increased soluble TNF activity was mediated, at least in part, by transcriptional activation of the gene encoding TNF-α. For comparative purposes, we also treated THP-1 cells with LPS, bacterial outer membrane constituents that have been shown to possess potent TNF-inducing capabilities. LPS or Stx 1 treatment stimulated TNF production by THP-1 cells with dissimilar dose responses; approximately fourfold more Stx1 was necessary to induce approximately 75% of the TNF bioactivity induced LPS. In accordance with earlier studies, we found that LPS induced a rapid transient increase in TNF-α transcripts that returned to basal levels within 3 hours of LPS stimulation. In contrast, TNF-α transcripts produced by Stx1 stimulation reached peak levels more slowly and remained elevated for longer periods of time (Fig 3).

When THP-1 cells were incubated with Stx1 and LPS, we noted a marked enhancement in TNF-α gene transcription and protein production. Although the precise mechanism of this augmented response remains to be defined, earlier studies demonstrated that treatment of macrophages with LPS and the protein synthesis inhibitor, cycloheximide, markedly upregulated TNF-α production, a phenomenon referred to as superinduction.25 Cycloheximide may inhibit the de novo synthesis of LPS-induced derepressor molecules or endogenous endonucleases involved in the degradation of cytokine transcripts.26-28 We show here that treatment of monocytic cells with LPS + Stx1 also results in a TNF superinduction effect. Reagents that simply bind Gb₃ (purified Stx1 B-subunits or monoclonal anti-Gb₃ antibody) do not trigger TNF synthesis. Whether the Stx1 A-subunit and its associated protein synthesis inhibitory activity are necessary for TNF induction and the LPS + Stx1 TNF superinduction effect is currently under investigation. Anti-LPS antibodies are frequently detected in HUS patients,29 a finding consistent with E coli- or Shigella dysenteriae-mediated gut mucosal damage. Thus, the presence of Stx and LPS in the circulation may stimulate proinflammatory cytokine production in a synergistic manner.

In unstimulated macrophages, the five members of the rel family of transcriptional activators form homodimers or heterodimers that are retained in the cytoplasm in an inactive state through association with one of a series of proteins called IκB. Upon stimulation with LPS, proinflammatory cytokines, or phorbol diesters, IκB proteins are phosphorylated, dissociate from NF-κB complexes, and undergo degradation. IκB dissociation shows nuclear translocation motifs on the dimers that then bind with high affinity to DNA containing NF-κB consensus sequences GGGRNNYYCC (where R = purines, Y = pyrimidines, and N = any nucleotide). NF-κB binding regulates the transcription of a large number of murine genes, including genes encoding cytokines, chemokines, proto-oncogenes, and leukocyte or endothelial cell adhesion molecules.13 In contrast to the murine TNF-α gene, the role of NF-κB in the activation of the human TNF-α gene is controversial. Multiple NF-κB binding sites have been characterized upstream of the human TNF-α transcription start site (TSS³⁰). Goldfield et al31 reported that each NF-κB site could be deleted without affecting LPS induction of human TNF-α gene expression in transiently transfected murine fibroblasts or monocytes. In contrast, several investigators demonstrated that LPS treatment of human monocytic cell lines resulted in the nuclear translocation of p65/p50 heterodimers and transcriptional activation.24 32-34 In this study we show that purified Stx1, like LPS, rapidly (within 60 minutes) stimulates the nuclear translocation of p65/p50 NF-κB complexes to PBMn and THP-1 nuclei and the degradation of cytoplasmic IκB-α. Examination of the kinetics of Stx1 induced NF-κB activation shows that nuclear translocation precedes TNF protein and transcript synthesis. We did not detect the translocation of p50 homodimers or p50/c-rel complexes in Stx1-stimulated THP-1 cells.

AP-1 complexes are sequence-specific transcriptional activators composed of homodimers or heterodimers of the Fos and Jun family of leucine zipper-containing proteins. The human TNF-α gene contains an AP-1 binding site 59 bp upstream of the TSS, and a number of studies suggest that AP-1, in combination with other transcriptional activators, is required for optimal gene expression. For example, the human TNF-α promoter contains a cAMP responsive element (CRE) 100 bp upstream of the TSS, and treatment of THP-1 cells with LPS resulted in the induction of TNF-α secretion and the selective binding of Jun/ATF heterodimers at the AP-1/CRE site.35 Recently, Yao et al36 demonstrated that multiple activators binding to AP-1/CRE, NF-κB (κB-3), and Sp1/Egr-1 sites are necessary for maximal LPS induction of TNF-α gene expression in THP-1 cells. Mackman et al37 showed that maximal activation by LPS of the gene encoding human tissue factor in THP-1 cells required the nuclear translocation of both NF-κB and AP-1. Thus, the transcriptional activation of eukaryotic genes may require cooperative binding of multiple activators at multiple cis-active sites within promoter regions.38 We show here that treatment of THP-1 cells with Stx1 activates the nuclear translocation of factors capable of binding NF-κB and AP-1 consensus sequences.

Although our studies are the first to show that Stxs may induce TNF production via a transcriptional activation mechanism involving NF-κB and AP-1, a number of infectious agents or toxins have been reported to trigger nuclear translocation of transcriptional activators. Invasion of HeLa cell monolayers by Shigella flexneri enhanced protein binding to AP-1, CREB, and NF-κB specific probes.39Spirochetal lipoproteins will trigger NF-κB translocation in THP-1 cells.40 Trede et al41,42 demonstrated that treatment of THP-1 cells with staphylococcal enterotoxin A (SEA) induced nuclear translocation of both NF-κB (p65/p50) and AP-1 complexes. Interestingly, p65/p50 heterodimers activated by SEA treatment appeared to preferentially bind the NF-κB binding site most proximal to the TNF-α TSS, whereas LPS treatment of Mono Mac 6 cells resulted in p65/p50 complexes binding to the NF-κB site most distal to the TSS.32 Staphylococcal superantigens are thought to signal cells via interaction with MHC class II molecules, whereas LPS has been shown to signal monocytes through CD14 or other membrane proteins. Whether Stx1, which binds to a membrane glycolipid, also uses intracellular signaling mechanisms activated by LPS or other bacterial products and whether Stx1 selectively triggers translocation of p65/p50 complexes to bind to specific NF-κB binding sites are currently being studied.

The authors thank David Acheson and Nancy Rice for sharing reagents necessary to perform these studies, Bharat Aggarwal for assistance with the electrophoretic mobility shift assays, and Gregory Foster for excellent technical assistance.