Research Communication Mediators of Inflammation, 12(5), 277 /283 (October 2003)

cytokine whose levels are elevated in inflamed mucosa, causes part of its effect on intestinal epithelial cells (IEC) through inducing ceramide production. Aim : To study the role of nuclear factor-kB (NF-kB), a pro-inflammatory and anti-apoptotic factor, in IL-1treated IEC. Methods : NF-kB activity and levels of apoptotic proteins were assessed by electrophoretic mobility shift assay and RNA-protection assay, respectively. Results : IL-1 and ceramide, which have been shown to partially mediate IL-l effects on IEC, activated NFkB levels significantly. This activation was due to a decrease in IkB-a and IkB-b protein levels. Moreover, the ratio of mRNA levels of anti-apoptotic to proapoptotic proteins was significantly increased in IL-1treated IEC. Conclusion : NF-kB may play a key role in the regulation of the expression of pro-inflammatory and/or apoptotic genes in inflammatory bowel disease, making this protein an attractive target for therapeutic intervention.


Introduction
Transcriptional factors of the nuclear factor (NF)-kB/ Rel family regulate the expression of genes involved in inflammation, cell proliferation and apoptosis. 1,2 NF-kB is rapidly activated in response to a wide variety of stimuli such as viral and bacterial infections or the interaction of the inflammatory cytokines with their receptors. 3 It is a heterodimer of the p50 (NF-kB-l) and p65 (Rel-A) subunits; however, proteins that constitute the NF-kB family form a variety of functional homodimers and heterodimers. 4 Á 6 Under resting conditions, the complex is present in the cytoplasm sequestered in an inactive form due to its association with an inhibitory protein belonging to the IkB family. Activation of NF-kB involves the dissociation of the Rel-A subunit from the inhibitory proteins, such as IkB-a, and the translocation of Rel-A from the cytoplasm to the nucleus, where it dimerizes with the p50 or c-Rel subunits and interacts with specific target genes. 7 The dissociation of Rel-A from IkB occurs following a cascade of intracellular events that leads to specific phosphorylation of the inhibitory subunit 7 on serine or tyrosine residues. 8,9 This phosphorylation is follwed by ubiquitination and proteolysis of the inhibitory protein via a nonlysosomal, adenosine triphosphate (ATP)-dependent proteolytic proteasome. Inflammatory cytokines such as tumor necrosis factor (TNF)-a and interleukin-1 (IL-1) have been shown to be elevated in the mucosa of patients with inflammatory bowel disease. These cytokines possess diverse biological functions that include cell proliferation, inflammation and apoptosis. They contribute to activation of intestinal epithelial cells (IEC) activation as well as the activation of immune cells that get recruited after the inflammatory insult. We have shown that in IEC, IL-1 activates sphingomyelinase, an enzyme that catalyzes the hydrolysis of sphingomyelin to ceramide. 10 In contrast to what has been observed in many cell types, however, ceramide did not induce apoptosis in IEC, neither did IL-l, but rather induced the synthesis of cyclooxygenase-2 (COX-2) leading to increased levels of pro-inflammatory mediators such as prostaglandin E 2 . 11 NF-kB consensus regions have been shown to be present on the COX-2 promoter region and these regions have been shown to be involved in COX-2 mRNA induction by cytokines. 12 Moreover, NF-kB seems to possess a significant anti-apoptotic role. 13,14 Indeed, some known pro-apoptotic extracellular signals can induce NF-kB that in turn induces the expression of genes that are anti-apoptotic. 14 Mediators that can induce NF-kB could therefore protect against apoptotic elimination of affected cells. Interestingly, this anti-apoptotic role of NF-kB has only been described for Rel-A but not for c-Rel or p50, which possess pro-apoptotic and anti-proliferative properties. 14,15 In this study we investigate the effects and mechanism of action of IL-1 and/or ceramide in regulating NF-kB in IEC. Activation of NF-kB by IL-1 may be responsible for the lack of apoptosis observed in cells treated with IL-1 and/or ceramide. 11

Cell culture
Murine intestinal epithelial cell 1ine Mode-K 16 was obtained from Dr P.B. Ernst (University of Texas, Medical Branch, Galveston, Texas, USA). Mode-K cells were maintained in DMEM (low glucose) containing 10% fetal bovine serum (FBS), 10 mM sodium pyruvate, and non-essential amino acids. T-84 cells, human colonic cell line, were grown in DMEM/F-12 medium supplemented with 10% FBS. Cell viability was measured using the trypan blue exclusion assay.

Nuclear extract preparation
Cells were harvested and collected by centrifugation at 200 g for 10 min and washed once with PBS. Cells were lysed by rapid freezing in ethanol/dry ice and thawed by resuspension in a hypotonic ice-cold buffer containing 10 mM KCl, 1.5 mM MgCl 2 , 1 mM dithiothreitol (DTT), and 10 mM Hepes. The nuclei were centrifuged at 1250 g for 10 min at 48C and the nuclear pellets were gently extracted in a hypertonic solution containing 0.4 mM NaCl, 1.5 mM MgCl 2 , 0.2 mM EDTA, 1 mM DTT, 0.5 mM PMSF, 20 mM Hepes and 25% glycerol, for 30 min at 48C, and then centrifuged for 20 min at 20,000 g to collect the nuclear proteins in the supernatant. The supernatant was diluted with 30 ml of buffer containing: 50 mM KCl, 20% (v/v) glycerol, 0.2 mM EDTA, 1 mM DTT, 0.5 mM PMSF and 20 mM Hepes, and stored at (/708C. Protein concentrations were determined using the Bio-Rad assay.

Western blotting
Total protein extracts were run on a 12% SDSpolyacrylamide gels. Gels were transferred to PVDF membranes overnight at 48C. Following transfer, membranes were washed with PBS containing 0.1% Tween 20 (TPBS) and blocked with 5% non-fat dry milk for 1 h at room temperature. Antibodies were then added in TPBS and incubated for 2 h at room temperature. Unbound antibodies were washed three times with TPBS and the anti-rabbit IgG HRPconjugate was added at 1:5000 dilution for 1 h at room temperature. The membranes were washed and incubated with luminol reagents and then exposed to autoradiography.
Electrophoretic mobility shift assay NF-kB consensus oligonucleotide was end-labeled with g-32 P ATP using T4 polynucleotide kinase. The hybridization reaction was performed using 10Á/20 mg of nuclear extract, 1 mg of poly(dI.dC), 1 mg of poly(dN6) as a non-specific competitor, and 10 mg of bovine serum albumin in 20 mM Hepes, 50 mM KCl, 1 mM EDTA and 5 mM DTT. The reaction was diluted with water to a v/v ratio of 1:20 of the labeled probe. The mixture was incubated for 30 min, then stopped by adding 6 ml of 15% Ficoll solution containing the indicator dyes bromophenol blue and xylene cyanol. The reaction mixture (20 ml of each sample) was electrophorized on a 5% non-denaturating polyacrylamide gel. The gel was transferred to Whatman filter paper, dried at 808C under vacuum for 2 h and processed for autoradiography at (/808C overnight. The specific NF-kB band was determined by competition experiments using a mutant oligonucleotide that has lost its ability to bind to the transcription factor. Subunit specificity was determined using specific antibodies to the NF-kB components (anti Rel-A, p50 and c-Rel) in the incubation step, which results in a supershift of the specific band due to the bound antibody.

RNase protection assay
Radioactive RNA probes were synthesized from DNA templates using T7 RNA polymerase and radioactive UTP as a substrate. The reaction was terminated after 1 h incubation at room temperature by the addition of DNase. The labeled RNA probes were extracted once with phenol:chloroform-isoamyl alcohol and once with chloroform-isoamyl alcohol. The probes were then precipitated in equal volume of 4 M ammonium acetate and 5 volumes of ice-cold absolute ethanol, incubated at (/808C for 30 min, centrifuged at full speed for 15 min at 48C in a microcentrifuge, washed once with 90% ethanol, allowed to air-dry and then solubilized in hybridization buffer and quantified using a scintillation counter. RNA samples (2Á/20 mg) were dried under vacuum and hybridized with the probe overnight.
Unhybridized RNA were digested using RNase A and the reaction was terminated by the addition of Proteinase K. Hybridized RNA were then extracted using phenol:chloroform-isoamyl alcohol, precipitated in ammonium acetate and absolute ethanol, washed once with 90% ethanol and then air-dried. The pellets were solubilized in loading buffer (supplied by the manufacturer) and were ran on a 5% acrylamide-urea gel using 0.5 )/tris borateÁ/EDTA buffer (containing 45 mM tris-borate, and 1 mM EDTA, pH 8.0) as a running buffer, dried and exposed to autoradiography. Yeast tRNA was used as a negative control.

IL-1 targets the NF-kB pathway
To establish whether IEC are functionally coupled to biological signals that lead to the activation of NF-kB by IL-1, cells were treated with IL-1 at different concentrations and time periods, and the NF-kB DNA binding activity was measured in nuclear extracts by eletrophoretic mobility shift assay using a NF-kB consensus oligonucleotide. We first confirmed the constitutive expression of NF-kB in Mode-K cells. Two positive complexes were detected by the radiolabeled NF-kB consensus oligonucleotide (Fig.  1). These complexes were not changed in the presence of a mutant oligonucleotide (data not shown).
IL-1 has been shown to activate the sphingomyelin metabolic pathway, which will cause increased accumulation of ceramide in IEC. 10 Ceramide acts as a second messenger in IEC and seems to mediate, at least in part, the effects of IL-l observed on IEC function. Thus the effect of ceramide on NF-kB in the presence or absence of IL-l was tested. Treatment of IEC by IL-l and/or ceramide caused a significant increase in NF-kB levels in Mode-K cells (Fig. 2). Supershift assays using anti-p50 and c-Rel-specific antibodies showed little variation in the amount of the DNA-bound proteins with the various treatments, while Rel-A-containing heterodimers (activated NF-kB state) were supershifted, indicating that Rel-A is the major subunit that is translocated to the nucleus after cell activation by IL-1.

Regulation of IkB subunits by IL-1
Effect of IL-1 and ceramide on IkBa protein levels We next examined the effect of IL-1 on the activation of IkB to determine which subunit is responsible for Effect of IL-1 and ceramide on IkBb protein levels IL-1 and ceramide also induced a decrease in the levels of IkBb levels in T-84 cells. The decrease, which was time (Fig. 5) and concentration dependent (data not shown), was maximal around 1 h with both treatments. The decrease observed in the levels of IkBa that was maximal at 4 h for IL-1 (10 ng/nl) and ceramide (10 mM) was much greater than that observed with Ik-Bb for the same treatment (80% decrease at 4 h versus 40% at 1 h in ceramide-treated cells, for example). This suggests that both IkBa and IkBb may be involved in releasing NF-kB from the cytosolic complex to translocate into the nucleus; however, the effect of treatment on IkBb proceeds that of IkBa. Similar results were obtained in Mode-K cells (data not shown).

Effects of IL-1 and ceramide on IkBo protein levels
The effect of ceramide and/or IL-1 on IkBo were also investigated. Neither treatment had any significant effect on the levels of IkBo (for Mode-K cells tested up to 24 h; Fig. 6) in either IEC line used, suggesting that it is not involved in the action of IL-1 on NF-kB activation. Equivalent amounts of nuclear extract from IEC were prepared and analyzed for NF-kB binding activity with a 32 P-labeled oligonucleotide probe containing NF-kB binding sites. Supershift assays using anti-p50-specific antibodies showed little but significant variation in the amount of the DNA-bound proteins, while Rel-A containing heterodimers (activated NF-kB state) were greatly supershifted, indicating that Rel-A is the major subunit that is translocated to the nucleus after cell activation by IL-1.

Effects of IL-1 and ceramide on IkBa and IkBb protein levels in the presence of ZLL
ZLL is a potent and selective reversible proteasome inhibitor. It inhibits NF-kB activation by preventing IkB degradation. Cells were treated with ZLL for different time periods before IL-1 or ceramide treatment, and the levels of IkBa and IkBb were measured (Fig. 6). ZLL reversed the effect induced by IL-1 or ceramide on both IkB levels, thus preventing their degradation, and subsequently no translocation of NF-kB was observed. These data suggest therefore that translocation of NF-kB is dependent on the activation of proteasome that leads to the degradation of IkBa and IkBb, thus freeing NF-kB to translocate to the nucleus for its action on activating different genes.
Regulation of mRNA levels of apoptotic proteins by IL-1 treatment Using the RNA protection assay, the mRNA levels of apoptotic proteins were shown to be regulated by IL-1 treatment. The mRNA levels of the anti-apoptotic proteins bcl-x and blc-w were increased while bcl-2 was unchanged; and the levels of the pro-apoptotic proteins bax, bak and bad were significantly reduced (Fig. 7). Similar results were obtained when cells were treated with ceramide. These data provide supportive evidence that IL-1 and/or ceramide do not induce apoptosis in intestinal epithelial cells and, rather, they feed into the prostanoid synthetic pathway through the induction of COX-2. 11

Discussion
Apoptosis is a highly conserved evolutionary event crucial for normal development and homeostasis. Deregulated cell death has been associated with disease entities such as cancer 17 and cardiovascular disease. 18 Cells defective for NF-kB signaling were found to be more sensitive to apoptotic-inducing signals than normal wild-type cells, 19,20 suggesting an anti-apoptotic role for NF-kB. NF-kB is a transcription factor involved in the regulation of many genes and is activated in response to a wide variety of stimuli including inflammation, viral and bacterial infections and stress factors. 21 In particular, NF-kB has been shown to modulate the expression of pro-inflammatory cytokines, chemokines, immune receptors, adhesion molecules, and enzymes that generate mediators of inflammation. 2 Activation of NF-kB involves the dissociation of the Rel-A (p65) subunit from inhibitory proteins, mainly IkBa, and its translocation from the cytoplasm to the nucleus, where it dimerizes with the p50 or c-Rel subunits and interacts with specific target genes. Activation of Rel-A or c-Rel containing NF-kB dimers results in the transcriptional stimulation of genes that are important for cell proliferation. NF-kB also possesses a significant anti-apoptotic role. 13,14 Indeed, the pro-apoptotic extracellular signals can also induce NF-kB that in turn induces the expression of genes that are antiapoptotic. 14 We have previously shown that IL-1 and ceramide did not cause any increase in apoptosis in Mode-K intestinal epithelial cells. 11 In this report, we show that treatment of cells with IL-1 caused activation of NF-kB and that this activation was completely dependent on its dissociation from the inhibitory proteins IkBa and IkBb. The release of NF-kB from IkBb proceeds its dissociation from IkBa.
The increase in the ratio of anti-apoptotic to proapoptotic mRNA provides an explanation for the lack of induction of apoptosis in IL-1-treated and/or ceramide-treated IEC. Both stimulators feed into the prostanoid synthetic pathway through the induction of COX-2, 11 whhich is thought to be due to activating NF-kB and translocation Rel-A to the nucleus where it can then bind on the COX-2 gene and induce its synthesis.
Mediators that can induce NF-kB could therefore protect against apoptotic elimination of affected cells. Interestingly, this anti-apoptotic role of NF-kB has only been described for Rel-A but not for c-Rel or p50, which possess pro-apoptotic and anti-proliferative properties. 14,15 This study shows that, upon activation of NF-kB, Rel-A was the subunit that was translocated preferentially from the cytoplasm to the nucleus, where it is capable of dimerizing with the p50 or c-Rel subunits and interacts with specific target genes, some of which are apoptotic proteins. Using the RNA protection assay we were able to show that in IL-1-treated or ceramide-treated IEC, the message levels of anti-apoptotic proteins were significantly increased while those of the pro-apoptotic proteins levels were significantly decreased. This correlated well with what we have observed earlier in IL-1treated IEC where it was shown that the protein levels of bcl-x were increased while protein levels of bax were significantly reduced. 11 Inhibition of NF-kB activation and subsequent translocation have been shown to promote apoptosis in tumor cells such as multiple myeloma cells. 22 This was achieved using a proteasome inhibitor PS-341, which prevents the degradation of IkBs by the proteasomes and hence the release of NF-kB from the IkB complex. PS-341 has also been recently used in clinical trials to resolve different types of malignancies. 23,24 Activation of NF-kB is an important event in numerous inflammatory processes of the gastrointestinal tract. Induction of this transcription factor has been reported in Helicobacter pylori gastritis, 25 inflammatory bowel disease, 26 and pancreatitis. 27 Recently, it has been shown that protein kinase C, intracellular calcium and activation of NF-kB are crucial for the induction of the COX-2 gene in the parietal cells in response to carbachol stimulation. 28 Earlier reports have stated that ceramide is a major second messenger for TNF-a-induced NF-kB activation, 29,30 while others have not observed any role for ceramide in IkBa degradation. 31 Ceramide was capable of activating NF-kB to the same extent as IL-1, providing more evidence that IL-1 causes some of its effects through the activation of the sphingomyelin pathway. 10,11 Neither IL-1 nor ceramide induced apoptosis in IEC but both induced COX-2 synthesis leading to increased levels of prostaglandin E 2 . 10,11 The COX-2 gene contains consensus regions for NF-kB that have been shown to be important for the induction of COX-2 mRNA levels by cytokines. 12 In human alveolar epithelial cells, 32 activation of mitogen-activated protein kinases by ceramide contributes to the TNF-a signaling that occurs downstream of neutral sphingomyelinase activation and results in the stimulation of IkB kinase (IKK1/2) and NF-kB in the COX-2 promotor, followed by initiation of COX-2 expression. Pro-inflammatory cytokines activate the IKK complex that phosphorylates the NF-kB inhibitors, triggering their conjugation with ubiquitin and subsequent degradation 33,34 via a non-lysosomal, ATP-dependent proteolytic proteasome. Freed NF-kB dimers translocate to the nucleus and induce target genes, such as COX-2, which catalyzes the synthesis of pro-inflammatory prostaglandins, in particular prostaglandin E 2 . At late stages of inflammatory episodes, however, COX-2 directs the synthesis of anti-inflammatory cyclopentenone prostaglandins, suggesting a role for these molecules in the resolution of inflammation.