Oxidation of IκBα at Methionine 45 Is One Cause of Taurine Chloramine-induced Inhibition of NF-κB Activation

A band shift of IκBα was observed in Western blots with Jurkat cells treated with 1 mm taurine chloramine (TauCl) for 1 h. TauCl treatment inhibited tumor necrosis factor α (TNFα)-initiated nuclear factor κB (NF-κB) activation. TauCl did not inhibit either the upstream of IκB kinase (IKK) activation or IKK itself but did inhibit NF-κB activation induced by IKK overexpression. Deletion experiments showed that a TauCl modification site causing the band shift of IκBα is Met45. High performance liquid chromatography and mass spectrometry analyses of a small peptide containing Met45 revealed that TauCl oxidizes Met45. A mutant of IκBα whose Met45 was converted to alanine did not generate a band shift upon TauCl treatment and degraded in response to TNFα stimulation. However, a reporter assay revealed that NF-κB-dependent luciferase expression was not fully recovered in cells transfected with this mutant. These results indicate that Met45 oxidation of IκBα is a molecular mechanism underlying the TauCl-induced inhibition of NF-κB activation. A similar band shift was observed when HL-60 cells expressing myeloperoxidase were treated with 100 μm hydrogen peroxide for 5 min. When rat neutrophils were incubated with bacteria, intracellular taurine decreased interleukin-8 production. Therefore, taurine may help suppress excessive inflammatory reaction in neutrophils.

Taurine, 2-aminoethanesulfonic acid, is a ␤-amino acid that is synthesized from sulfur-containing amino acids in the body.
A well known function of this amino acid is to increase solubility and intestinal absorption of ingested lipids as taurine conjugated with bile acid (1). In neutrophils, taurine is the most abundant free amino acid (2). Neutrophils are one of the white blood cells that have an important role in the host defense against bacterial infection. When neutrophils phagocytose invading bacteria, nuclear factor B (NF-B) 1 is activated (3). NF-B consisting of heterodimer of RelA and p50 is retained in the cytoplasm as an inactive tertiary complex with inhibitory proteins known as IBs. For example, when tumor necrosis factor ␣ (TNF␣) binds to its receptor, IB kinase (IKK) is activated through intracellular signal pathway (4). This activated kinase phosphorylates IB␣ protein on Ser 32 and Ser 36 (5), which leads to ubiquitination on Lys 21 and Lys 22 followed by proteasome-dependent degradation (6). The degradation and dissociation of IB␣ from NF-B leads to the nuclear transfer of NF-B, and subsequent gene expression is initiated (7,8).
In neutrophils, phagocytosis of bacteria also triggers biosynthesis of hypochlorous acid (HClO) from hydrogen peroxide (H 2 O 2 ) and chloride ion. This reaction is mediated by myeloperoxidase (MPO) (9). The neutrophil is the major tissue in which this oxidase is distributed. HClO is a strong oxidant and kills phagocytosed bacteria. Taurine in the neutrophil seems to help protect this tissue against HClO, because taurine instantaneously reacts with HClO and forms taurine chloramine (TauCl) that is by far less toxic than HClO to the neutrophil (10). However, the further biological activity of TauCl has been left unclear.
Recently, it has been shown that TauCl inhibits production of inflammatory cytokines (11,12). Because gene expressions of inflammatory cytokines are known to be regulated by NF-B (7,8), one assumes that TauCl inhibits NF-B activation. However, the molecular mechanism of this inhibition is unknown. The results in this study show that one molecular mechanism underlying TauCl-induced inhibition of NF-B activation is oxidization of IB␣ at Met 45 . It has also been shown that intracellular taurine decreases NF-B-dependent interleukin-8 (IL-8) production stimulated by bacteria in rat neutrophils. Therefore, the biological function of taurine in the neutrophil is not only reduction of the cytotoxicity of HClO but also mitigation of excessive inflammatory reaction. We assume that taurine may act as an intermediate regulator between two distinct events in neutrophils during their phagocytosis of bacteria, synthesis of HClO, and NF-B activation.

EXPERIMENTAL PROCEDURES
Cell Culture-Human acute T cell leukemia Jurkat T and HEK293 cells were purchased from the American Type Culture Collection (Manassas, VA). Promyelocytic leukemia HL-60 cells were obtained from Health Science Research Resources Bank (Osaka, Japan). These three kinds of cells were routinely maintained at 37°C in a humidified atmosphere of 95% air, 5% CO 2 and passed to new culture at a split ratio of 1:10 every week. The composition of culture medium for Jurkat cells was RPMI 1640 (JRH Biosciences, Lenexa, KS) containing 10% fetal bovine serum (Invitrogen), 100 units/ml penicillin, and 100 g/ml streptomycin (Invitrogen). The medium pH was adjusted to 7.3 with sodium bicarbonate (Kanto Chemical, Tokyo, Japan). The culture medium for HL-60 cells was made by supplementing one for Jurkat cells with 2 mM glutamine (Kanto Chemical). A culture medium for HEK293 cells was Dulbecco's modified Eagle's medium containing 10% fetal bovine serum, 100 units/ml penicillin, and 100 g/ml streptomycin, and 6 mM glutamine. The pH of this medium was also adjusted to 7.3 with sodium bicarbonate.
Preparation of TauCl-TauCl was freshly prepared by adding equimolar NaOCl (Kanto) dropwise to 5 ml of 80 mM taurine (Kanto) in 0.05 M sodium phosphate buffer (pH 8.3) before use, as described previously (13). Each preparation of TauCl was monitored by ultraviolet absorption spectra (200 -300 nm) to confirm the production of monochloramine as well as the absence of dichloramine or unreacted NaOCl. The concentration of TauCl was determined using the molar extinction coefficient (⑀ 252 nm ϭ 415 M Ϫ1 cm Ϫ1 ).
Western Blot Analysis-Cell extracts were prepared as described previously (14). The extracts were boiled and fractionated by SDS-PAGE (12.5%). To detect a small change in the molecular weight of IB like its molecular weight increase due to phosphorylation, we used a large gel consisting of ϳ10 cm separating gel and overran until IB came close to the bottom of the gel (14). In most experiments, the primary and secondary antibodies used were rabbit anti-IB␣ antibody (Santa Cruz Biotechnology, Inc., Santa Cruz, CA) and donkey antirabbit IgG antibody linked to horseradish peroxidase (Amersham Biosciences), respectively. When mouse anti-FLAG antibody was used as a primary antibody, the secondary antibody was rabbit anti-mouse IgG horseradish peroxidase-conjugated antibody. Proteins on the membrane were visualized by an enhanced chemiluminescent method with a kit (Amersham Biosciences).
Reporter Assay-A reporter gene consisted of a plasmid containing the luciferase gene in the downstream of three binding sites of NF-B (3ϫB-tk-luc) or five binding sites of cAMP-responsive element-binding protein (5ϫCRE-tk-luc). A plasmid containing ␤-galactosidase gene, which has a ␤ actin promoter (pA5c␤gal) was co-transfected to estimate the transfection efficiency. Jurkat cells (2 ϫ 10 6 ) were transfected with 1 g of a reporter gene and 0.5 g of pA5c␤gal by the DEAE-dextran method. Cells were cultured for 2 days after transfection, treated with 1 mM TauCl for 1 h, and stimulated with 20 ng/ml TNF␣ for 1.5 h. HEK293 cells (3 ϫ10 5 ) seeded a day before were co-transfected with 0.1 g of 3ϫB-tk-luc and 5 g of pRc␤act-3HAIKK␤ by the calcium phosphate method. Cells were cultured for 2 days and treated with 1 mM TauCl for 1 h. Luciferase activity in the cells was measured with the PicaGene Luminescence kit (Wako, Tokyo, Japan). For determining the ␤-galactosidase activity, 50 l of each cell lysate was diluted with 450 l of 100 mM sodium phosphate buffer (pH 7.4) containing 10 mM KCl, 1 mM MgSO 4 , and 50 mM ␤-mercaptoethanol and mixed with 100 l of a substrate solution. The composition of this solution was 2 mg/ml o-nitrophenyl-␤-D-galactopyranoside (Sigma) in 100 mM sodium phosphate buffer (pH 7.0). When the mixture turned yellow after a certain incubation at 37°C, the reaction was stopped with 500 l of 1 M Na 2 CO 3 . The ␤-galactosidase activity was spectrophotometrically determined at 430 nm. Because the transfection efficiency estimated by ␤-galactosidase activity did not widely deviate, luciferase activity was not normalized with the transfection efficiency.
IKK Assay-After Jurkat cells (8 ϫ 10 6 ) were treated with TauCl and TNF␣, cell lysate was made as described previously (17). The lysate was centrifuged at 13,000 ϫ g for 15 min. The supernatant was cleaned by a 1-h incubation with 2 g of normal rabbit serum and 30 g of a 50:50 (v/v) slurry of Protein G-Sepharose beads and centrifuged again at 2,000 ϫ g for 1 min. An anti-IKK␤ (1.5 g) antibody (Santa Cruz Biotechnology) was added to the resultant supernatant. Then 20 l of the 50:50 (v/v) slurry of Protein G-Sepharose beads was added again. The mixture was rotated for 1 h and centrifuged at 2,000 ϫ g for 1 min. The final pellet was incubated at 37°C for 30 min with 3 g of IB␣ in the presence of 20 M ATP and 5 Ci of [␥-32 P]ATP. IB␣ was fractionated on 10% polyacrylamide-SDS gel and phosphorylation was detected by imaging analysis with BAS 3000 (Fuji, Tokyo, Japan).
Preparation of Human IB␣ Mutants-Deletion mutants of IB␣ were prepared by a PCR method. All PCR products of IB␣ cDNA fragments were verified by nucleotide sequencing. To subclone IB␣ deletion mutants, pGEX (Amersham Biosciences) was used. Their overproduction as glutathione S-transferase fusion proteins was carried out with Escherichia coli BL21 strain (Promega, Madison, WI). IB␣-glutathione S-transferase fusion protein was purified by glutathione-Sepharose 4B (Amersham Pharmacia Biotech). IB␣ mutants were released from glutathione S-transferase by digestion with thrombin protease (Amersham Pharmacia Biotech) and concentrated using Centricon (Millipore Corp., Bedford, MA).
A point mutation of IB␣ whose Met 45 was converted to alanine (IB␣ M45A) was made by site-directed mutagenesis using a method previously described by Kunkel et al. (18). A wild type deletion mutant consisting of the amino acids 1-280 (IB␣ 1-280 (wild type)) and a corresponding point-mutated IB␣ (IB␣ 1-280 (M45A)) were prepared by PCR as described above. A vector pME was used to express wild type and mutants of IB␣ in Jurkat cells (14).
High Performance Liquid Chromatography (HPLC) and Mass Spectrometry Analyses-A peptide fragment consisting of IB␣ 43-50 was synthesized as one with acetylated N-terminal end (Ac-EQMLKELQ). This peptide fragment at 1 mM was incubated with 400 M TauCl for 30 min at 37°C. The reaction solution was separated by HPLC with an octadodecyl sulfate C18 column (Senshu Scientific Co., Tokyo, Japan). The separation run was recorded with a UV detector set at 215 nm. Solvents A and B were water and acetonitrile, respectively. Both solvents contained 0.1% trifluoroacetic acid. Solvent A was eluted at a flow rate of 1 ml/min with a gradient of solvent B from 5 to 25% in 20 min and maintained for an additional 5 min at 25% of solvent B. Peaks fractionated by HPLC were analyzed by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry for the molecular weight determination of AcEQMLKELQ. 2,5-Dihydroxybenzoic acid (Aldrich) was used as a matrix.
Measurement of MPO Activity and Taurine Concentration in Jurkat and HL-60 Cells-Cells (3 ϫ 10 6 ) in 50 mM potassium phosphate buffer (pH 6.0) were homogenized for 10 s three times with 5-s intervals using a Polytron-type homogenizer. The MPO activity was determined using o-dianisidine (Sigma) as a substrate (19). The concentration of protein in the cell homogenates was determined by the Lowry method (20).
Cells (1 ϫ 10 7 ) were rinsed twice with PBS(Ϫ) to remove extracellular taurine and digested with 5% trichloroacetic acid. Cell debris was removed by a centrifugation at 20,000 ϫ g for 5 min. The amount of taurine in the supernatant was determined by an amino acid analyzer (Hitachi, Tokyo, Japan). The cell volume (1 ϫ 10 7 cells) was estimated by subtracting freeze-dried weight from wet weight. The intracellular concentration of taurine was obtained by dividing the amount of intracellular taurine by the cell volume.
Isolation of Rat Neutrophils and Quantification of IL-8 -Sprague-Dawley male rats (10 weeks old) were intraperitoneally administered with 40 ml of 1% casein (Wako) in PBS(Ϫ). Eighteen hours later, neutrophils were collected from the abdominal cavity by washing with PBS(Ϫ). Collected neutrophils were rinsed twice with PBS(Ϫ), resuspended, and incubated for 30 min in a hypotonic buffer to reduce intracellular taurine. The composition of hypotonic buffer was 95 mM KCl, 1.3 mM CaCl 2 , 0.5 mM MgCl 2 , and 10 mM HEPES/Tris, pH 7.3, whose osmolarity was 198 Ϯ 1 mosM⅐kg Ϫ1 (n ϭ 3). To reload taurine, cells whose taurine content was reduced were incubated for 30 min with a hypertonic buffer (150 mM NaCl, 1.3 mM CaCl 2 , and 0.5 mM MgCl 2 containing 100 mM taurine, buffered with 10 mM HEPES/Tris, pH 7.3, 416 Ϯ 4 mosM⅐kg Ϫ1 (n ϭ 6)). Taurine was replaced with raffinose in the hypertonic buffer for cells whose taurine content was kept low. Cells were resuspended in RPMI 1640 containing 0.25% bovine serum albumin (Sigma), whose pH was adjusted to 7.3 with 5 mM sodium bicarbonate. An aliquot of neutrophil suspension (200 l of 1 ϫ 10 6 cells/ml) was mixed with the same volume of E. coli K88 suspension (200 l of 1 ϫ 10 7 cells/ml) to initiate phagocytosis. Therefore, the multiplicity of infection was 10. After incubation for the desired time, neutrophils and bacteria were precipitated by centrifugation at 20,000 ϫ g for 5 min. The supernatant was used to determine IL-8 by an enzyme-linked immunosorbent assay with the rat GRO/CINC-1 detection kit (Amersham Biosciences).

TauCl-induced Interruption of NF-B Activation-To inves-
tigate the effect of TauCl on NF-B activation, we paid attention to IB␣, which is a key protein in the cascade of NF-B activation. We used the Jurkat T cell line, because this cell line was one of the most characterized cell lines with regard to molecular mechanisms of NF-B activation. Western blot analysis of IB␣ was performed for 15 min after stimulation with 20 ng/ml TNF␣. When the cells were stimulated with TNF␣, a band shift of IB␣ due to its phosphorylation by IKK was observed at 5 min, and both original and shifted bands disappeared at 10 and 15 min (Fig. 1A). Treatment with 1 mM TauCl for 1 h prior to TNF␣ stimulation brought about two shifted bands of IB␣ even at 0 min, but these two shifted and original bands did not disappear for 15 min after TNF␣ stimulation. Although IB␣ also survived for 15 min when cells were pretreated for 1 h with 100 M PSI and 300 M PDTC, transient patterns of IB␣ after TNF␣ stimulation were different from that observed in TauCl pretreatment. These results indicate that TauCl pretreatment increased resistance of IB␣ to degradation induced by TNF␣. However, TauCl seems to inhibit IB␣ degradation differently from PSI or PDTC.
We studied the effect of TauCl on IB␤. Jurkat cells were stimulated with a combination of PMA and ionomycin to initiate the degradation of IB␣ and IB␤ at one time (Fig. 1B,  Control). IB␣ and IB␤ disappeared for 30 min and 2 h after stimulation, respectively. When cells were pretreated with 1 mM TauCl for 1 h, IB␤ degraded, whereas IB␣ did not. The results indicate that the modification of IB by TauCl is specific to ␣ isoform.
To examine if TauCl-induced inhibition of IB␣ degradation leads to interruption of nuclear transfer of NF-B, an electrophoretic mobility shift assay was performed. After TNF␣ stimulation, the amount of NF-B translocated into the nucleus increased for 30 min (Fig. 1C). However, a 1-h pretreatment with 1 mM TauCl drastically decreased NF-B translocation into the nucleus. We performed a reporter assay to reveal that TauCl inhibits NF-B-dependent gene expression. Luciferase activity in cells transfected with 3ϫB-tk-luc was decreased by TauCl in a dose-dependent manner, while no effects of TauCl on luciferase transcription were observed in cells transfected with 5ϫCRE-tk-luc (Fig. 1D). These results indicate that inhibition of IB␣ degradation by TauCl leads to inhibition of nuclear transfer of NF-B and subsequent NF-B-dependent transcription.
We examined effects of TauCl treatment on IKK activity and NF-B activation triggered by overexpression of IKK␤. IB␣ was phosphorylated by in vitro incubation with IKK immunoprecipitated from Jurkat cells stimulated with TNF␣ ( Fig. 2A). A similar phosphorylation was observed, although IKK was extracted from cells treated with 1 mM TauCl for 1 h prior to TNF␣ stimulation. This means that TauCl does not inhibit either the signal pathway upstream of IKK activation triggered by TNF␣ stimulation or IKK itself. When IKK␤ was overexpressed in HEK293 cells, NF-B was activated, and luciferase Anti-IB␣ and -IB␤ antibodies were used for Western blots. C, electrophoretic mobility shift assay was performed with nuclear extracts from cells (2 ϫ 10 6 ) treated with 1 mM TauCl for 1 h, followed by 20 ng/ml TNF␣ stimulation. D, reporter assay was performed with a plasmid 3ϫB-tk-luc (closed circle) or 5ϫCRE-tk-luc (open circle). A plasmid pA5c␤gal was co-transfected to estimate the transfection efficiency. Jurkat cells (2 ϫ 10 6 ) were transfected with 1 g of a reporter gene and 0.5 g of pA5c␤gal by a DEAE-dextran method and cultured for 2 days. Then they were treated with TauCl for 1 h and stimulated with 20 ng/ml TNF␣ for 1.5 h. The concentration of TauCl was altered in the range of 0 -1 mM. The data represent means Ϯ S.D. (n ϭ 3). expression was increased by ϳ30-fold (Fig. 2B). TauCl treatment inhibited two-thirds of this increase of NF-B-dependent luciferase expression. The results indicate that TauCl interacts with the signal pathway downstream rather than upstream of IKK activation.
Characterization of TauCl-induced Modification of IB␣-Because TauCl treatment caused a band shift of IB␣ similar in molecular size to that of phosphorylated IB␣ at Ser 32 /Ser 36 , we assumed that TauCl might modify IB␣ at these serine residues. A FLAG-tagged IB␣ mutant whose Ser 32 /Ser 36 are converted to alanine (FLAG-IB␣ (S32A/S36A)) was overexpressed in Jurkat cells, treated with 1 mM TauCl for 1 h, and immunoprecipitated with anti-FLAG antibody. Western blot analysis with anti-FLAG antibody revealed that TauCl treatment induced band shift of FLAG-IB␣ (S32A/S36A), indicating that modification of IB␣ by TauCl does not occur at Ser 32 / Ser 36 (Fig. 3A).
To examine whether TauCl-induced band shift is due to phosphorylation, cell extracts were treated for 30 min with 10 units of alkaline phosphatase before electrophoresis. Alkaline phosphatase treatment dephosphorylated IB␣ that had been phosphorylated at Ser 32 /Ser 36 by IKK activated by TNF␣ (Fig.  3B). However, a similar alkaline phosphatase treatment of the extract from cells pretreated for 1 h with 1 mM TauCl did not affect shifted bands of IB␣, showing that phosphorylation is not a cause of TauCl-induced IB␣ band shift.
Cell extracts were first boiled for 5 min to inactivate enzymes and then treated for 30 min with TauCl in the range of 0 -0.4 mM. Boiling reduced the sensitivity of IB␣ to TauCl, but TauCl still generated the band shift (Fig. 3C). This experiment suggests that any enzymatic reaction may not be involved in band shift formation of IB␣ by TauCl.
Identification of TauCl-modified Site in IB␣-The modification site of IB␣ by TauCl was investigated to gain insight into molecular mechanisms of TauCl-mediated inhibition of IB␣ degradation. Two deletion mutants of IB␣ 43-180 and 47-180 synthesized in bacteria were treated with 400 M TauCl in vitro and fractionated on 12.5% polyacrylamide/SDS gel. The deletion mutant of IB␣ 47-180 did not show its band shift upon TauCl treatment, although band shift was observed with wild type IB␣ 1-317 and the other mutant of IB␣ 43-180 (Fig. 4A). We speculated that one of the TauCl modification sites exists in the region of amino acid residues 43-46 of IB␣. Therefore, we made three more deletion mutants of IB␣ 44 -180, 45-180, and 46 -180 and repeated a similar TauCl treatment experiment. Mutants of IB␣ 44 -180 and 45-180 exhibited band shift, whereas IB␣ 46 -180 did not (Fig. 4B). We assumed that a TauCl modification site might be Met 45 of IB␣.
Analysis of TauCl-induced Modification of IB␣ Met 45 -A small peptide of IB␣ 43-50 with acetylated N-terminal end, Ac-EQMLKELQ, was used to understand how IB␣ Met 45 is modified by TauCl. The chromatogram of this peptide showed a major peak at the retention time of 22 min (Fig. 5A, Control).

FIG. 2. Effects of TauCl on IKK activity.
A, Jurkat cells (8 ϫ 10 6 ) were pretreated with 1 mM TauCl for 1 h and stimulated with 20 ng/ml TNF␣ for 10 min. Cells were harvested, and IKK complex was immunoprecipitated with anti-IKK␤ antibody. After recombinant IB␣ (3 g) was incubated with IKK complex (200 g), phosphorylated IB␣ was fractionated by 10% SDS-PAGE and detected by an imaging analysis. B, HEK293 cells (3 ϫ 10 5 ) seeded and cultured for 1 day were co-transfected with 0.1 g of 3ϫB-tk-luc and 1 g of pRc␤act-3HAIKK␤ by the calcium phosphate method. After cells were cultured for 2 days, they were treated with 1 mM TauCl for 2 h, and luciferase activity was determined. The data represent means Ϯ S.D. (n ϭ 3).
This was designated as the intact peptide peak. A small peak at 2 min was due to phosphate in a buffer. Treatment of the peptide for 30 min with 400 M TauCl resulted in the appearance of another peak at 15 min with decreasing the intact peptide peak (Fig. 5A, TauCl). This means that TauCl treatment increases the hydrophilicity of the peptide.
For further analysis, the intact peptide peak at 22 min and another peak at 15 min generated by TauCl treatment were collected and subjected to matrix-assisted laser desorption/ ionization time-of-flight mass spectrometry. TauCl treatment increased the molecular weight of the peptide by a factor of ϳ16 (Fig. 5B). Because the increased molecular weight of 16 is equal to that of one oxygen atom, the modification of the peptide by TauCl is estimated to be oxidation of IB␣ Met 45 to methionine sulfoxide. Other small peaks usually observed in this type of mass spectrometry are ghosts.
Resistance of Oxidized Met 45 to TNF␣-induced IB␣ Degradation-To confirm that Met 45 of IB␣ is an amino acid residue that is modified by TauCl, Western blot analysis was performed with Jurkat cells transfected with pME-IB␣ 1-280 (M45A). For control, pME-IB␣ 1-280 (wild type) was used. These deletion mutants were used to distinguish mutant IB␣ from endogenous full-length IB␣. When cells expressing IB␣ 1-280 (wild type) and (M45A) were stimulated with 20 ng/ml TNF␣, these IB␣s degraded for 20 min (Fig. 6A). Upon 1-h pretreatment with 2 mM TauCl, two shifted bands of IB␣ 1-280 (wild type) were generated. Either shifted or original band did not degrade like endogenous full-length wild type IB␣. However, IB␣ 1-280 (M45A) was degraded by TNF␣ stimulation regardless of a similar treatment with TauCl.
Using 3ϫB-tk-luc, reporter assay was performed in Jurkat cells transfected with pME-IB␣ 1-280 (wild type and M45A). Treatment with 2 mM TauCl for 1 h inhibited TNF␣-induced luciferase activity by a factor of 90% in cells expressing IB␣ 1-280 (wild type) (Fig. 6B). In cells expressing IB␣ 1-280 (M45A), inhibition by TauCl was less (70%), although not completely reversed. To evaluate expression levels of IB␣ 1-280 (wild type) and (M45A), a Western blot was performed. Band densities were less than that of endogenous IB␣ (Fig. 6C). The low level of expression of IB␣ (M45A) relative to endogenous IB␣ may account for failure in full recovery of luciferase expression in cells transfected with pME-IB␣ 1-280 (M45A).  (21), HL-60 cells were used to determine whether TauCl that is intracellularly synthesized modifies IB␣ like TauCl that is extracellularly administered. The intracellular concentration of taurine and the activity of MPO were compared between Jurkat and HL-60 cells. The intracellular concentration of taurine was ϳ1.0 and 0.8 mM in Jurkat and HL-60 cells, respectively (Fig. 7A). Whereas the MPO activity was none in Jurkat cells, that activity was ϳ8 units/mg of protein in HL-60 cells.

Treatment of HL-60 Cells with H 2 O 2 -Because HL-60 cells express MPO
When HL-60 cells were stimulated with 20 ng/ml TNF␣, IB␣ was degraded for 10 min after the stimulation (Fig. 7B). However, when HL-60 cells were pretreated with 100 M H 2 O 2 for 5 min prior to TNF␣ stimulation, two band shifts of IB␣ were generated, and they did not degrade for 10 min. Furthermore, when HL-60 cells were pretreated for 30 min with a specific MPO inhibitor, ABAH, at 500 M and treated with H 2 O 2 in the presence of this inhibitor at the same concentration, the band shift of IB␣ was not generated, and IB␣ degraded in response to TNF␣ stimulation. In Jurkat cells, IB␣ degraded upon TNF␣ stimulation, regardless of H 2 O 2 or ABAH pretreatment. These results indicate that TauCl is synthesized by MPO in HL-60 cells treated with H 2 O 2 , and intracellularly synthesized TauCl can modify IB␣ as observed in Jurkat cells extracellularly administered TauCl.
Electrophoretic mobility shift assay was performed with HL-60 cells. NF-B translocated into the nucleus when the cells were treated with 20 ng/ml TNF␣ for 15 min. However, the nuclear transfer of NF-B was decreased by treatment for 5 min with 100 M H 2 O 2 but recovered by the 30-min pretreatment and presence of 500 M ABAH during TNF␣ stimulation. This suggests that IB␣ modified by intracellularly synthesized TauCl inhibits NF-B nuclear translocation.
Effect of Intracellular Taurine on IL-8 Release from Rat Neutrophils Stimulated with Bacteria-To demonstrate that intracellular taurine really affects the NF-B-dependent production of inflammatory cytokine, the release of IL-8 was measured with rat neutrophils. As shown in Fig. 8, the release of IL-8 initiated by bacteria was significantly higher in taurinereduced neutrophils (1 Ϯ 1 mM, n ϭ 11) than reloaded ones (26 Ϯ 9 mM, n ϭ 11) (p Ͻ 0.05). Furthermore, when taurinereloaded neutrophils were treated with 500 M ABAH before and during incubation with bacteria, the release of IL-8 was significantly increased for 3-7 h (p Ͻ 0.05). Because neutrophils express MPO, intracellular taurine reacts with HClO produced during phagocytosis of bacteria and forms TauCl. Intracellularly synthesized TauCl seems to decrease NF-B-dependent IL-8 production.

DISCUSSION
In this study, we found that when Jurkat cells were treated with TauCl, IB␣ was band-shifted and became resistant to TNF␣ stimulation (Fig. 1A). A similar treatment of Jurkat cells with TauCl caused the suppression of NF-B nuclear translocation and consequent NF-B-dependent gene expression (Fig.  1, C and D). Recently, Barua et al. (22) have shown that TauCl inhibits NF-B-dependent gene expression of inducible nitricoxide synthase and TNF␣ in activated alveolar macrophages, because IB␣ degradation is inhibited by TauCl treatment. In this respect, our observation is consistent with their results (22). They speculated that TauCl may interact with a certain upstream step of IKK because TauCl did not directly inhibit IKK activity in an in vitro assay. In Fig. 1B, IB␤ was degraded by TNF␣ stimulation in TauCl-treated Jurkat cells, while IB␣ did not degrade. An expressed deletion mutant of IB␣ 1-280 (M45A) also degraded after TNF␣ stimulation regardless of a similar treatment with TauCl (Fig. 6A). Since the activation of IKK is required for degradation of IB␤ and the IB␣ mutant, Electrophoretic mobility shift assay was performed with nuclear extracts (5 g each) as described in the legend of Fig. 1C.   FIG. 8. Effects of intracellular taurine on IL-8 release from rat neutrophils. Rat neutrophils whose taurine content was changed were incubated with bacteria for 1, 3, 5, and 7 h. When ABAH was used, cells were treated with 500 M of this MPO inhibitor for 30 min prior to bacterial stimulation. ABAH was also present during incubation with bacteria. The supernatant was used to determine IL-8. The data are given as means, with bars representing S.D. (n ϭ 3). Open circles, taurine-reduced cells; filled circles, taurine-reloaded cells; half-filled circles, taurine-reloaded cells treated with ABAH.
we do not agree with their speculation that TauCl negatively regulates IKK in its upstream. Furthermore, Fig. 2, A and B, clearly shows that TauCl does not inhibit either the signal pathway upstream of IKK activation or IKK itself but interrupts the signal pathway after IKK activation. In our in vitro experiments with cell extracts and synthesized IB␣ fragment peptides, TauCl treatment also caused IB␣ band shift, showing that TauCl can directly modify the IB␣ molecule (Fig. 3, B  and C, and Fig. 4). We used a large gel whose length was ϳ10 cm and overran to separate shifted bands from the original band of IB␣ as previously described (14). When we used a minigel, we could not observe the band shift of IB␣ (data not shown). Our results show that a direct modification of IB␣ by TauCl that is a cause of its band shift is a crucial step for increase in IB␣ resistance to degradation initiated by TNF␣ stimulation.
When we first observed TauCl-induced band shifts of IB␣, we thought that TauCl inhibited IKK or proteasome, modified IKK phosphorylation sites Ser 32 /Ser 36 , or caused phosphorylation of other sites, because the band shifted by TauCl was similar to that by phosphorylation in molecular size. When Jurkat cells were treated with the combination of PMA and ionomycin to degrade IB␣ and IB␤ together, IB␤ degraded even in the presence of TauCl, whereas IB␣ did not (Fig. 1B). This result precluded the possibility of inhibition of IKK or proteasome by TauCl, because IKK and proteasome both are required to degrade IB␤. When Jurkat cells expressing FLAG-IB␣ (S32A/S36A) were treated with TauCl, IB␣ mutant exhibited the band shift (Fig. 3A). Therefore, it is very hard to think that phosphorylation sites of IB␣, Ser 32 /Ser 36 , were modified by TauCl. An idea of phosphorylation at other sites by TauCl was also excluded by the results shown in Fig. 3B. Fig.  3C indicates that the modification of IB␣ by TauCl is a nonenzymatic reaction. We carried out in vitro experiments using synthesized IB␣ deletion mutants and fragment peptide and came to the conclusion that the oxidation of IB␣ at Met 45 by TauCl is one cause of its band shift (Figs. 4 and 5). Because methionine sulfoxide is a product when one oxygen molecule reacts with methionine, we deduced that Met 45 is oxidized to methionine sulfoxide.
It is still unclear whether Met45 is the only oxidation site by TauCl. Because two shifted bands of IB␣ were observed when cells were treated with TauCl or H 2 O 2 (Figs. 1A, 6A, and 7B), TauCl may oxidize another methionine residue of IB␣, besides Met 45 . If one of two shifted bands resulted from phosphorylation, the number of the shifted bands would decrease when cell extracts were treated with alkaline phosphatase. Therefore, we expect that another methionine residue is oxidized. As shown in Fig. 1B, TauCl treatment did not generate a band shift of IB␤, whereas a similar treatment generated two shifted bands of IB␣. We think that this IB␣-specific oxidation results from different location of methionine residues between IB␣ and -␤. IBs consist of a centrally located ankyrin repeat domain (ARD); a signal-receiving domain (SRD) from the amino terminus to ARD; and a proline-, glutamic acid-, serine-, and threonine-rich region from ARD to the carboxyl terminus (23). IB␣ contains six methionine residues, Met 1 , Met 13 , Met 37 , and Met 45 in SRD and Met 91 and Met 279 in ARD. On the other hand, IB␤ has a Met 1 in SRD and all of the others, Met 87 , Met 222 , Met 271 , and Met 282 , in ARD. Probably methionine residues in SRD are more susceptible to oxidation than those in ARD because SRD is jutting out on the surface of IB␣⅐NF-B complex. Thus, besides Met 45 , Met 13 , or Met 37 of IB␣ may be a feasible oxidation site. In Fig. 4A, only one shifted band was observed when full-length IB␣ (residues 1-317) was treated with 400 M TauCl in vitro. We assume that one shifted band is the upper one of two shifted bands probably due to a high concentration of TauCl at 400 M in in vitro treatment, because two shifted bands of IB␣ were observed by in vitro treatment with 100 M TauCl (Fig. 3C). Furthermore, the distance between the original and shifted band of full-length IB␣ is larger than that of a deletion mutant of IB␣ 43-180 (Fig. 4A). These data support an idea that another oxidation site is Met 13 or Met 37 .
Even if Met 13 or Met 37 is oxidized by TauCl, one may think that this oxidation site may not be crucial in degradation of IB␣ because a mutant of IB␣ 1-280 (M45A) expressed in Jurkat cells did not generate a band shift upon TauCl treatment and degraded in response to TNF␣ stimulation (Fig. 6A). However, in a reporter assay, luciferase expression was not fully recovered in cells expressing this mutant (Fig. 6B). The density ratio of endogenous over exogenous IB␣ bands was ϳ3, indicating that expressed IB␣ 1-280 (wild type) and (M45A) were only 25% of total IB␣. Therefore, even if expressed IB␣ 1-280 (M45A) fully degrades upon TNF␣ stimulation after TauCl treatment, the maximal restoration of luciferase expression might not be more than 25%. However, unless the restoration of luciferase expression by overexpression of the mutant becomes 100%, there is still a possibility of other mechanisms, including oxidation of IB␣ at Met 13 or Met 37 . Therefore, we have to clearly state here that oxidation of IB␣ at Met 45 may be only one cause of TauCl-induced interruption of NF-B activation.
There are two possible reasons why oxidized Met 45 disturbs IB␣ degradation. One possibility is that oxidized Met 45 constructs a structural obstacle, which prevents IKK from approaching IB␣ phosphorylation sites, Ser 32 and Ser 36 . The other possibility is that even if IKK can phosphorylate oxidized IB␣, the F-box protein necessary for ubiquitination that occurs at Lys 21 and Lys 22 of IB␣ cannot recognize the phosphorylated IB␣ (24). Because Met 13 and Met 37 are also located close to IB␣ phosphorylation and ubiquitination sites, their oxidation may have a similar disturbing effect on IB␣ degradation. Further experiments are required to understand detail molecular mechanisms of IB␣ resistance induced by TauCl.
Reactive oxygen species (ROS) such as H 2 O 2 , hydroxyl radical, and superoxide anion are generated during the course of normal metabolism. These oxidants modify a variety of cellular constituents, protein, lipid, DNA, and so on, which causes cellular damage. All amino acids can be oxidized with ROS. However, methionine as well as cysteine residues possess sulfur-containing side chains that are the most sensitive to the ROS-mediated oxidation. Involvement of cysteine residues in intracellular redox regulation is well known. Oxidation of cysteine thiol most likely forms disulfide, which glutathione easily reduces. However, function of methionine residues in redox regulation has been left unclear. A recently published review article suggests that methionine residues play an important role in redox regulation via a redox of methionine and methionine sulfoxide (25). There are two representative proteins whose functional activity is regulated by this redox system, Shaker potassium channel (26) and calmodulin (27). Both are involved in cellular excitability. This study provides the first evidence for the regulation of NF-B activation by oxidation of a methionine residue of IB␣. Although TauCl is known to be able to oxidize proteins by in vitro treatment (28), it has not been demonstrated that methionine oxidation by TauCl takes a part in intracellular events. Because it has been reported that methionine sulfoxide reductase is present in white blood leukocytes (29), NF-B activation might be under the methionine redox control. Fig. 1B shows that TauCl selectively reacts with IB␣. IB␣ degradation is important in rapid response to phag-ocytosis, but IB␤ is related to continuous NF-B activation (30,31). Therefore, we assume that intracellular taurine may function to suppress undesired rapid inflammatory reaction in neutrophils. The concentration of taurine is ϳ50 M in the extracellular space (32). MPO is released to the extracellular space of inflammatory site when neutrophils are broken by dramatic activation. This situation leads to the extracellular formation of TauCl. Although the membrane permeability of TauCl is low, extracellular TauCl can diffuse into surrounding cells and inhibit NF-B activation (33). Taking these data together, we assume that in inflammatory sites as well as neutrophils, taurine can suppress the inflammation in which ROS is generated to an excessive extent.