Protective Roles of NF-κB for Chromium(VI)-induced Cytotoxicity Is Revealed by Expression of IκB Kinase-β Mutant

To delineate the molecular mechanisms of NF-κB-mediated regulation of chromium(VI)-induced cell death, the signaling pathway leading to the activation of NF-κB was interrupted by stable transfection of a kinase-mutated form of IκB kinase β (IKKβ-KM). Here we demonstrate a novel role for the NF-κB transcription factor in inhibiting chromium(VI)-induced cell death. Inhibition of NF-κB by IKKβ-KM or IKKβ gene deficiency resulted in a spontaneous cleavage of Bcl-xl antiapoptotic protein due to the elevated caspase-3 activity. DNA microarray assay suggested a decreased expression of genes encoding antiapoptotic proteins, cIAP1 and cIAP2, in the cells overexpressing IKKβ-KM. Chromium(VI) treatment of these NF-κB-inhibited cells induced necrotic-like cell death. Such chromium(VI)-induced cell killing could be partially inhibited by expression of exogenous cIAP1, an inhibitor of caspases, indicating that caspases along with others may be involved in chromium(VI)-induced cell death. These results suggest that NF-κB is essential for inhibiting toxic metal-induced cytotoxicity. Such inhibition may involve up-regulation of the expression of anti-death proteins including cIAP1 that prevents spontaneous caspase activation and subsequent cleavage of Bcl-xl protein.

A wide range of signals, many of which are thought to be related to cellular stress, induce expression of early response genes through the NF-B family of transcription factors (1)(2)(3)(4). In resting cells, NF-B is retained in cytoplasm in its inactive form by interaction with one of a number of inhibitory molecules including IB␣, IB␤, IB⑀, p105, and p100. Activation of the NF-B signaling cascade results in a complete degradation of IB or carboxyl-terminal partial degradation of the p105 and p100 precursors, allowing nuclear translocation of the NF-B complexes. Activated NF-B binds to specific DNA sequences in target genes, designated as B elements, and regulates transcription of genes mediating inflammation, carcinogenesis, and pro-or antiapoptotic reactions. IB␣ is the most abundant inhibitory protein for NF-B (5). The mechanisms of signalinduced IB␣ degradation involve phosphorylation of two serine residues, Ser 32 and Ser 36 . This phosphorylation leads to polyubiquitination of two specific lysines on IB␣ (Lys 21 and Lys 22 ) by an SCF-␤-TrCP complex and its degradation by the 26 S proteasome (6). The phosphorylation is accomplished by a specific IB kinase (IKK) 1 complex containing two catalytic subunits, IKK␣ and IKK␤, and a structural component named NEMO/IKK␥/IKKAP (3,5). IKK␣ and IKK␤ share 50% sequence homology. Both proteins contain an amino-terminal kinase domain, a carboxyl-terminal region with a leucine zipper, and a helix-loop-helix domain. In vitro and in vivo studies indicate that both IKK␣ and IKK␤ are capable of phosphorylating IB␣ on Ser 32 and Ser 36 , but IKK␤ is more potent in IB␣ phosphorylation induced by proinflammatory stimuli. Recent studies by several groups indicate the existence of an additional IKK-like kinase complex in T cells, named IKKi/⑀, which shares 27% homology with IKK␣ and IKK␤ and possibly mediates NF-B-activating kinase signaling and phorbol 12-myristate 13-acetate/protein kinase C⑀-induced Ser 36 phosphorylation of IB␣ and thus NF-B activation (7)(8)(9)(10)(11).
Increasing evidence indicates that NF-B is either a pro-or antiapoptotic transcription factor regulating a variety of apoptotic responses (12). NF-B is activated in response to several proapoptotic stimuli including oxidative stress, cytotoxic drugs, and ionizing radiation (13,14). Consistent with this notion, the gene encoding Fas ligand (FasL) has been shown to be transcriptionally regulated by NF-B in response to T-cell activation signals and to chemotherapeutic agents (15,16). The evidence that NF-B is also an antiapoptotic transcription factor is mainly provided by gene knockout studies of NF-B family members and IKK kinase subunits (17)(18)(19). RelA (p65)-deficient mice die during embryonic development through apoptosis of hepatocytes (17). IKK␤ gene knockout mice die as embryos and show massive liver cell apoptosis, a response similar to that of NF-B p65 gene knockout mice (19). Male mice with an inactivated X-linked gene encoding IKK␥/NEMO, an essential modulator of the IKK complex for NF-B activation, die at midgestation due to a massive cortical and medulla lymphocyte apoptosis in the thymus in addition to degeneration of the liver (20,21). Thus, in certain situations, NF-B is proapoptotic, but in alternative situations and cell types, NF-B inhibits apoptosis and contributes to cell proliferation or transformation. Therefore, cell type and inducing stimuli appear to determine whether NF-B is a causal or secondary event in apoptosis.
Apoptosis is a process in which cell death is initiated and completed in an orderly fashion through the activation of various apoptotic pathways (22,23). However, in cases of severe injury, cells may instead undergo necrosis, a passive death resulting in cellular lysis (23). Most apoptotic cells are characterized by unique morphological features, such as membrane blebbing, cell shrinking, cytosolic and nuclear condensation, and breakdown of chromosomal DNA. In contrast, cells dying by necrosis are characterized by cellular edema and loss of cell membrane integrity. Depending on the involvement of caspases or reactive oxygen species, cell death can be apoptotic, necrotic, or both (24). In fact, under many circumstances, different death pathways can co-exist in the same cell and are switched on by specific stimuli. A number of studies have revealed that when a cell dies by a typical apoptotic process, usually a late phase necrosis also occurs (25)(26)(27)(28)(29).
Cr(VI) compounds, widely used in industry, have been shown to have serious toxic and carcinogenic effects on humans. Although the biochemical features of the signals that associate Cr(VI) with NF-B activation and cell death have so far remained unclear, both reactive oxygen species (ROS)-dependent and ROS-independent mechanisms have been proposed (30 -32). The importance of NF-B as an antiapoptotic factor is evident mainly from the studies of gene knockout mice and the apoptotic pathways of tumor necrosis factor ␣ signaling (17)(18)(19)33). Much less is known concerning the role of NF-B in Cr(VI)induced cell death. The objective of the present investigation was to clarify the involvement of NF-B in Cr(VI)-induced cell death and to determine if NF-B plays a protective or promoting role in cell death triggered by Cr(VI).

MATERIALS AND METHODS
Cells and Reagents-The human bronchial epithelial cell line, BEAS-2B, from American Type Culture Collection (ATCC, Manassas, VA) was cultured in keratinocyte basal medium (Sigma) supplemented with 30 g/ml bovine pituitary extract and 5 ng/ml human epidermal growth factor. Mouse embryo fibroblasts (MEF) derived from wild-type mice and IKK␤ gene knockout mice were a gift from Dr. Michael Karin (University of California, San Diego, La Jolla, CA) and cultured in Dulbecco's modified Eagle's medium (Invitrogen) supplemented with 10% fetal bovine serum. Cr(VI) was purchased from Aldrich. The luciferase assay kit was from Promega (Madison, WI). All antibodies against NF-B family members, IKK␤, procaspase-3, Bcl-xl, and Myc tag were from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA) or Upstate Biotechnology, Inc. (Lake Placid, NY). Anti-FLAG monoclonal antibody was from Sigma. ECL Western blotting detection reagents were from Amersham Biosciences.
Cell Transfection-pCR-FLAG-IKK␤ and pCR-FLAG-IKK␤-KM (K44A) were gifts from Dr. Hiroyasu Nakano (Juntendo University, Japan). pcDNA3-myc-IAP1 was provided by Dr. John C. Reed (The Burnham Institute, La Jolla, CA). pEGFPluc vector was purchased from CLONTECH Laboratories, Inc. (Palo Alto, CA). BEAS-2B cells were plated in six-well tissue culture plates at 5 ϫ 10 5 cells/well for 2 days. The cells were transfected with a control vector (pCR3) or indicated expression vectors along with a 2ϫ B-dependent luciferase reporter construct using LipofectAMINE (Invitrogen) as previously described (34). Single clones of BEAS-2B cells, stably transfected with the control vector (pCR3), wild-type IKK␤, or IKK␤-KM, and luciferase reporter genes, were isolated in 700 g/ml G418 for 3 weeks and tested by Western blotting and luciferase activity assay for expression of the transfected genes. Stably transfected cells were maintained in regular culture medium supplemented with 200 g/ml G418. To minimize possible clone variations during the course of selection, several independently derived cell lines expressing each transfected vector with similar expression levels were pooled together for the experiments described below.
Electrophoretic Mobility Shift Assay (EMSA)-For nuclear protein extraction, cells were harvested and resuspended in hypotonic buffer A (10 mM HEPES (pH 7.6), 10 mM KCl, 0.1 mM EDTA, 1 mM dithiothreitol (DTT), 0.5 mM phenylmethylsulfonyl fluoride) as previously described (35). Briefly, cells were incubated in buffer A for 10 min on ice and then vortexed for 10 s. Nuclei were pelleted by centrifugation at 12,000 ϫ g for 20 s and were resuspended in buffer C (20 mM HEPES (pH 7.6), 25% glycerol, 0.4 M NaCl, 1 mM EDTA, 1 mM DTT, 0.5 mM phenylmethylsulfonyl fluoride) for 30 min on ice. The supernatants containing nuclear proteins were collected after centrifugation at 12,000 ϫ g for 2 min and stored at Ϫ70°C. For EMSA, 4 g of nuclear extract were mixed with the 32 P-labeled double-stranded oligonucleotide containing a B sequence (5Ј-CAACGGCAGGGGAATTCCCCTCTCCTT-3Ј). The reaction solution was incubated at room temperature for 30 min and electrophoresed on a native 5% polyacrylamide gel in 0.25ϫ TBE buffer for 2-3 h. The DNA-binding proteins were visualized by autoradiography.
Kinase Activity Assay-The IKK activity assay was performed by the method reported by Geleziunas et al. (36) with minor modifications. Briefly, transfected BEAS-2B cells, seeded at a concentration of 5 ϫ 10 6 cells/ml and cultured for 2 days, were treated with indicated agents and lysed in a lysis buffer containing 1% Nonidet P-40, 250 mM NaCl, 50 mM HEPES (pH 7.4), 1 mM EDTA, 0.5 mM phenylmethylsulfonyl fluoride, 1 mM DTT, 10 g/ml aprotinin, and 10 g/ml leupeptin. After centrifugation of the lysate at 16,000 ϫ g for 20 min at 4°C, the supernatant was incubated with anti-IKK␤ antibody H-470 or anti-FLAG antibody with rotation for 4 h at 4°C, followed by the addition of 20 l of Protein A-agarose and incubation at 4°C for an additional 2 h. The immunoprecipitate was collected by centrifugation at 2,000 ϫ g and washed three times with lysis buffer and two times with kinase buffer containing 20 mM HEPES (pH 7.4), 20 mM ␤-glycerophosphate, 1 mM MnCl 2 , 5 mM MgCl 2 , 2 mM NaF, and 1 mM DTT. To monitor the kinase reaction, the immunoprecipitate was incubated in 20 l of kinase buffer supplemented with 5 Ci of [␥-32 P]ATP and 1 g of GST-IB␣ (1-54) (CLON-TECH, Palo Alto, CA) for 30 min at 30°C. The reaction was stopped by the addition of SDS sample buffer. The samples were separated by SDS-polyacrylamide gel electrophoresis (SDS-PAGE), which was then transferred onto a nitrocellulose membrane and subjected to autoradiography.
Clonogenic Survival Assay and Cell Death Assay-Logarithmically growing cells stably transfected with indicated expressing vectors were harvested by typsinization. Cell suspensions were seeded into six-well tissue culture plates with a concentration of 10 3 cells/well. After allowing cells to adhere for 12 h, the cells were treated with various concentrations of Cr(VI) for an additional 12 h. After the treatment, cells were washed and incubated for 1 week in tissue culture medium containing 5% fetal bovine serum. At the end of culture, the cell colonies were washed and fixed by the addition of water/methanol (1:1, v/v) containing crystal violet (1 mg/ml) and counted under a microscope. The clonogenic survival rate was calculated based on the number of colonies that grew and the number of cells plated into each well. For the analysis of cell death, stably co-transfected cells with the pEGFPluc and/or other indicated vectors were cultured in six-well tissue culture plates for 48 h before the experiments. The percentage of green cells was determined by fluorescence microscopy. Five independent counts in each experiment were used to determine a mean and S.D.
Genefilter Microarray and RT-PCR-The Genefilter membrane (gf2l1) from Research Genetics (Huntsville, AL), which covers 3,965 genes, was used for mRNA expression profiling following the manufacturer's instructions. Briefly, 1 g of total RNA extracted from transfected cells was incubated with 2 g of oligo(dT); 1.5 l of reverse transcriptase; 20 mM dATP, dGTP, and dTTP; and 100 Ci of [ 32 P]dCTP in 30 l of diethylpyrocarbonate-treated water for 90 min at 37°C. After purification through a Bio-Spin 6 chromatography column, labeled probe was mixed with prehybridization solution and incubated with Genefilter membranes overnight at 42°C. To minimize possible variations among individual membranes, the same membrane was stripped and rehybridized with a second probe after the first round of hybridization. To verify the microarray data, some of the differentially regulated genes in the transfected cells, wild-type or IKK␤ Ϫ/Ϫ MEF, were analyzed by RT-PCR. The primers used for RT-PCR were designed by using Primer3 software (available on the World Wide Web at wwwgenome.wi.mit.edu/cgi-bin/primer/primer3_www.cgi) and indicated in Table I.
Western Blotting-Whole cell extracts were mixed with 3ϫ SDS-PAGE sample buffer and then subjected to SDS-PAGE in 10 or 16% gels. The resolved proteins were transferred to a nitrocellulose membrane. Western blotting was performed using antibodies against IKK␤, FLAG, Myc tag, Bcl-xl, caspase-3, and anti-rabbit IgG-horseradish peroxidase conjugates.

RESULTS
Inhibition of IKK␤ Blocks NF-B Activation-IKK␤ has been considered as the major IB␣ kinase in response to a variety of stimuli (3,5). To determine whether overexpression of a kinase-mutated form of IKK␤ (IKK␤-KM) can lead to inhibition of NF-B, we characterized BEAS-2B cell clones stably expressing either wild-type IKK␤ or IKK␤-KM along with an NF-Bdependent luciferase reporter construct. BEAS-2B cells transfected with the empty vector pCR3 were employed as a control. To exclude the potential problem associated with overexpression, we selected clones with a range of expression of the exogenous proteins relative to the endogenous IKK␤ and identified clones with comparable levels of expression of wild-type IKK␤ and IKK␤-KM. We first confirmed the previously observed inhibition of NF-B in the cells expressing IKK␤-KM (37). The nuclear proteins were prepared from the transfected clones in the absence or presence of various doses of Cr(VI) for 1 h and subjected to EMSA. Fig. 1A shows that NF-B DNA binding activity in the cells transfected with a control vector or wild-type IKK␤ could be induced by Cr(VI) in a dose-dependent manner. In contrast, no or very marginal induction of NF-B DNA binding activity by Cr(VI) could be observed in the cells transfected with IKK␤-KM (Fig. 1A, lanes 6 -10, upper panel). The same nuclear extracts were also analyzed for the Sp1 DNA binding activity. As shown in Fig. 1A, overexpression of IKK␤-KM did not alter the Sp1 DNA binding activity in the cellular response to Cr(VI) (Fig. 1A, bottom panel).
To verify that the inhibition of NF-B was a result of the functional disruption of IKK in cells expressing IKK␤-KM, we examined IKK kinase activity in these cells in the absence or presence of Cr(VI). Cell extracts prepared at a 40-min time point after treatment with Cr(VI) were immunoprecipitated using IKK␤ antiserum and subjected to an immune complex kinase assay using GST-IB␣ (amino acids 1-54) as the substrate. As depicted in Fig. 1B, Cr(VI) stimulated IKK kinase activity in the cells transfected with a control vector or wildtype IKK␤ (Fig. 1B, lanes 1-3 and lanes 7-9, top panel). Only marginal IKK kinase activity was induced by Cr(VI) in the cells stably expressing IKK␤-KM (Fig. 1B, lanes 4 -6, top panel). Essentially equal amounts of IKK␤ proteins were present in the extracts from the cells transfected with vector, IKK␤-KM, or IKK␤ as verified by immunoblot using anti-FLAG and anti-IKK␤ antibodies (Fig. 1B, middle and bottom panels). Since the transfected IKK␤ and IKK␤-KM were consistently of the expected size in the immunoblot using anti-FLAG antibody (Fig.  1C, third panel), it seemed unlikely that the IKK␤-KM coding region had undergone mutation or rearrangement during plasmid amplification or integration into genomic DNA. Thus, these results suggest that the IKK kinase activity is indeed inhibited in the cells expressing a kinase-mutated form of IKK␤, IKK␤-KM.
IKK␤ Inhibition Enhances Cell Death-Evidence that cells lacking NF-B activity undergo apoptosis suggests that NF-B activation provides protection against apoptotic signals (17). The above data show that NF-B activation in response to Cr(VI) is defective in the cells expressing IKK␤-KM. We next determined whether NF-B inhibition by expression of IKK␤-KM sensitized cells to apoptosis in response to Cr(VI). To our surprise, Cr(VI) (5 M) treatment for 12 h induced a necrotic-like, rather then apoptotic, cell death of IKK␤-KM cells. Morphologic analysis of phase-contrast images of cells indicates that only a few control vector-transfected cells or wildtype IKK␤-expressing cells exhibited partial cell shrinkage and condensation after the treatment with Cr(VI) (Fig. 2, F and G). In contrast, after the same treatment, IKK␤-KM-expressing cells manifested cell blebbing, swelling, and loss of membrane integrity, characteristics similar to those seen in cells undergoing necrosis (Fig. 2H).
The role of IKK␤ and NF-B in controlling Cr(VI)-induced cytotoxicity was further investigated genetically using a knockout MEF cell line lacking IKK␤ subunits. A dramatic loss of cell viability in response to Cr(VI) was observed in IKK␤ Ϫ/Ϫ MEF (Fig. 2J) but not in wild-type MEF (Fig. 2I). Thus, these results excluded the potential artifacts associated with the use of dominant negative IKK␤ kinase mutant in overexpression experiments (Fig. 2H).
To further assess the cytotoxic effect of Cr(VI) on the cells in which NF-B was inhibited due to overexpression of IKK␤-KM or deficiency of the IKK␤ gene, IKK␤-expressing cells, IKK␤-KM-expressing cells, wild-type MEF, and IKK␤ Ϫ/Ϫ MEF were treated with increasing concentrations of Cr(VI). Cytotoxicity was determined by both LDH release analysis and clonogenic survival assay. As indicated in Fig. 3, A and B, compared with their wild-type counterparts, a substantial increase of LDH release was observed in IKK␤-KM cells (Fig. 3A) and in IKK␤ Ϫ/Ϫ cells (Fig. 3B) in response to various doses of Cr(VI). Consistent with this observation, the clonogenic survival assay indicated that exposure to increasing amounts of Cr(VI) inhibited clonogenic survival in IKK␤-KM cells and IKK␤ Ϫ/Ϫ MEF more effectively than in the cells expressing wild-type IKK␤ or wild-type MEF (Fig. 3C). Fig. 3D depicts a representative clonogenic survival experiment.

Spontaneous Cleavage of Bcl-xl in IKK␤-KM Cells or IKK␤ Ϫ/Ϫ Fibroblasts-It has been demonstrated that the bcl-x
gene is a transcriptional target of NF-B in both mouse and human cells (35,38). Enhanced NF-B activity has been correlated with the up-regulated expression of Bcl-xl, an important antiapoptotic protein that can stabilize mitochondrial membranes and prevent the release of cytochrome c and apoptosis-inducing factor (39 -41). A possible explanation for the increased vulnerability of IKK␤-KM-expressing cells in response to Cr(VI) is that these cells may lack sufficient antideath proteins, such as Bcl-xl, due to the impairment of NF-B signaling. Decreased expression of Bcl-xl can cause either apoptosis due to the increase of mitochondrial membrane permeability or necrosis due to the collapse of fragile mitochondria (42). However, gene expression profiling showed no difference of bcl-xl gene expression between IKK␤ and IKK␤-KM expressing cells (data not shown). Unexpectedly, spontaneous cleavage of Bcl-xl protein was observed in IKK␤-KM-expressing cells but not in control vector-or wild-type IKK␤-transfected cells (Fig.  4A, left panel). A 17-kDa fragment occurred concomitant with a disappearance of the 30-kDa intact Bcl-xl protein band in nonstimulated or Cr(VI)-stimulated IKK␤-KM-expressing cells. There are two potential cleavage sites of caspase-3 (HLAD61/S and SSLD76/A) that are located in the loop region between the BH4 and BH3 domains of the Bcl-xl protein (43,44). Cleavage of these sites by activated caspases releases a C-terminal product that lacks the BH4 domain, an antiapoptotic domain of Bcl-xl protein. The spontaneous cleavage of Bcl-xl in IKK␤-KM cells indicated possible activation of caspases in these cells. Indeed, immunoblotting shows a basal activation of caspase-3 as judged by the cleavage of the 32-kDa precursor caspase-3 with the appearance of a 12-kDa activated caspase-3 fragment (Fig. 4A, right panel). Cr(VI) treatment did not further alter the cleavage of Bcl-xl and activation of caspase-3, indicating that Cr(VI) itself has no effect on proteases responsible for the cleavage of Bcl-xl or the activation of caspase-3.
To rule out the possibility that above observations are artifacts due to overexpression of IKK␤-KM, we next examined the status of Bcl-xl proteins and caspase-3 in MEF cells derived from both wild-type mice and IKK␤ gene knockout mice. As depicted in Fig. 4B, IKK␤ protein is absent in IKK␤-deficient MEF (IKK␤ Ϫ/Ϫ ; Fig. 4B, top panel). However, these cells express comparable levels of IKK␣ as observed in wild-type cells (Fig. 4B, the second panel). The spontaneous cleavage of Bcl-xl protein and activation of caspase-3 are evident in IKK␤ Ϫ/Ϫ cells (Fig. 4B, third and bottom panels, respectively).  1 and 3) or presence (lanes 2 and 4) of 5 M Cr(VI) for 12 h. Total cellular proteins were extracted and subjected to immunoblot using antibodies against IKK␤, IKK␣, Bcl-xl, and caspase-3.

Decreased cIAP Expression in IKK␤-KM Cells-
The spontaneous activation of caspase-3 in IKK␤-KM cells implied an impaired antiapoptotic function in these cells. It is known that NF-B may regulate the expression of several antiapoptotic genes, such as cIAP1 and cIAP2. The failure of IAP antibody to detect IAP proteins in our system prompts us to analyze the basal gene expression profile of both wild-type IKK␤ and IKK␤-KM-expressing cells by DNA microarray. Both wild-type IKK␤and IKK␤-KM-expressing cells were cultured in medium for 12 h. cDNA probes were generated from the RNAs of both cell lines and used for sequential hybridization with the human Genefilter gf2l1, which contains 3,965 sequence-verified known human genes. The majority of these genes were expressed at similar levels in cells stably expressing either wild-type IKK␤ or IKK␤-KM. In IKK␤-KM cells, several genes encoding proteins involved in the P450 function/cellular redox regulation, protein degradation, cell cycle, and transforming growth fac-tor-␤ signaling were up-regulated by more than 2.5-fold in comparison with IKK␤ cells (Fig. 5). Thus, these data indicate that NF-B may negatively regulate the expression of these genes. At least two recent reports also demonstrated that NF-B suppressed the expression of the P4501A1 (cyp1a1) gene (45) and a proteasome C3 subunit gene (46). Under the basal condition, many of the documented NF-B target genes, such as cytokines and chemokines, were not changed (data not shown). However, we did note a decreased expression of both cIAP1 and cIAP2 genes in IKK␤-KM cells. Both cIAP1 and cIAP2 have been originally identified as direct inhibitors for caspases, especially for caspase-3, caspase-7, and caspase-9 (47). In addition, the expression of genes encoding transcription factor E2F5, keratin 18, and an antioxidant protein, protein disulfide isomerase-related protein, is decreased in the IKK␤-KM cells. Therefore, the observed spontaneous activation of caspase-3 in IKK␤-KM cells may be explained as the lack of sufficient endogenous caspase inhibitors, such as cIAP1 and cIAP2.
To verify the difference of gene expression observed by microarray analysis between IKK␤-and IKK␤-KM-expressing cells, we next performed RT-PCR using equal amount of total RNAs from IKK␤-expressing cells, IKK␤-KM-expressing cells, wild-type MEF, or IKK␤ Ϫ/Ϫ MEF. The results of the RT-PCR analysis confirmed decreased expressions of cIAP1 and cIAP2 and increased expression of endothelial nitric-oxide synthase and POH1 in IKK␤-KM cells (Fig. 5B). In fact, the cIAP1 expression appears to be undetectable in the cells stably expressing IKK␤-KM in this RT-PCR analysis (Fig. 5B, lane 4 of  the cIAP1 panel). In addition, we also compared the expression levels of endothelial nitric-oxide synthase, POH1, cIAP1, and cIAP2 between wild-type MEF and IKK␤ Ϫ/Ϫ MEF. Similar to the BEAS-2B cells stably expressing IKK␤-KM, the IKK␤ Ϫ/Ϫ MEF exhibited an increased expression of POH1 and decreased expression cIAP2 (Fig. 5B, lane 6). We failed to detect the expression of endothelial nitric-oxide synthase and cIAP1 in both wild-type and IKK␤ Ϫ/Ϫ MEF. For unknown reasons, we also failed to detect the expression of the XDH gene in both BEAS-2B cells transfected with different vector and MEF with different genetic backgrounds in several RT-PCR analyses (data not shown).
Cr(VI)-induced Cell Death Can Be Partially Inhibited by Exogenous cIAP1-To determine whether Cr(VI)-induced necrotic-like cell death of IKK␤-KM cells was in fact due to the reduced expression of cIAP1 genes, we tested whether overexpression of cIAP1 was capable of reducing Cr(VI)-induced cell death. The IKK␤-KM cells were further transfected with a control vector, pcDNA, or a vector expressing Myc-tagged cIAP1 and cultured for 48 h. Cells were then left untreated or treated with various concentrations of Cr(VI). After an additional 12 h, the caspase-3 activation, Bcl-xl cleavage, and LDH release were determined. As depicted in Fig. 6A, IKK␤-KM cells transfected with the control vector exhibited spontaneous activation of caspase and Bcl-xl cleavage as judged by the disappearance of procaspase-3 bands and intact Bcl-xl bands (Fig. 6A, lanes 4 -6, top and middle panels). In contrast, transfection of Myc-tagged cIAP1 significantly blocked caspase-3 activation and Bcl-xl cleavage under either basal or Cr(VI)treated conditions (Fig. 6A, lanes 1-3, top and middle panels).
The possible protective role of cIAP1 on Cr(VI)-induced cytotoxicity was also determined by cell viability analysis of IKK␤-KM cells co-transfected with pEGFPluc and Myc-tagged cIAP1 or pcDNA control vector (Fig. 6B) was observed in IKK␤-KM cells co-transfected with pEGFPluc and Myc-tagged cIAP1.

DISCUSSION
The results presented here provide evidence for a novel function of NF-B in inhibiting Cr(VI)-induced necrotic-like cell death. In the cells stably expressing IKK␤-KM, an essential component of NF-B signaling, IKK␤, is defective (Fig. 1B). EMSA indicates a pronounced decrease of NF-B DNA binding activity in these IKK␤-KM expression cells in response to Cr(VI) (Fig. 1A). Cell morphologic analysis demonstrates that treatment of the cells expressing IKK␤-KM with Cr(VI) induced a necrotic-like cell death (Fig. 2). Analysis of the protein expression levels for both Bcl-xl and caspase-3 shows that IKK␤-KM-expressing cells or IKK␤ gene knockout MEF exhibited spontaneous cleavage of Bcl-xl protein and activation of caspase-3 (Fig. 4, A and B). The gene expression profiling analysis reveals that inhibition of IKK␤ to block NF-B signaling decreased the expression of two important antiapoptotic genes, cIAP1 and cIAP2. Transfection of the cells expressing IKK␤-KM with cIAP1 partially prevents caspase-3 activation and Bcl-xl cleavage (Fig. 6A) and protects the cells from Cr(VI)induced cytotoxicity (Fig. 6B).
While the mechanism by which NF-B protects cells from death signals remains to be further investigated, it may be related to its transcriptional regulation on several antiapoptotic genes (33). The observations presented in this paper support the notion that NF-B plays a pivotal role in the expression of both cIAP1 and cIAP2 genes. These data also support a model for the consequent effects of NF-B inhibition on Cr(VI)-induced cell death (Fig. 7). The levels of cIAPs and Bcl-xl may determine whether necrotic cell death or apoptosis ensues in the cellular response to Cr(VI). In NF-B-inhibited cells, such as the expression of IKK␤-KM and IKK␤ gene knockout, caspase-3 was activated due to the reduced expression of cIAP1 and cIAP2. Activated caspase-3 cleaves Bcl-xl, which not only weakens the protective mechanism of Bcl-xl on the mitochondrial outer membrane but also converts this antiapoptotic protein to a killer molecule (40,41). Under this predisposed condition, Cr(VI) treatment may result in necrosis rather than apoptosis due to severe damage of mitochondria. Severely damaged mitochondria release an excessive amount of cytochrome c that interrupts electron transport in the inner membrane, causing ATP depletion and consequently switching the cells from apoptosis to necrosis. However, if the levels of cIAP1 and cIAP2 are maintained by a normal NF-B activation response, the cleavage of Bcl-xl will be prevented by IAPmediated inhibition of caspases (Fig. 6).
The protective effect of cIAP1 on Cr(VI)-induced death of IKK␤-KM cells is distinct from the previous reports indicating that peptidyl caspase inhibitors potentiate tumor necrosis factor ␣or double-stranded RNA-induced cytotoxicity (48,49). It should be noted that there are several substantial differences between cIAPs and peptidyl caspase inhibitors. In addition to their function as endogenous inhibitors for caspases, cIAP1 and cIAP2 have recently been shown to regulate several signal transduction pathways leading to the activation of NF-B and c-Jun N-terminal kinase (50, 51) and act as ubiquitin ligases modulating protein degradation (52,53). Thus, the observed Caspase-3 is spontaneously activated under this circumstance, which causes cleavage of Bcl-xl protein. Bcl-xl cleavage not only weakens the protective mechanism of Bcl-xl on mitochondrial outer membrane but also converts this antiapoptotic protein to killer molecules of mitochondria. Necrosis, rather then apoptosis, will occur upon persistent insults, such as Cr(VI) or overwhelming ROS. Small up and down arrows indicate increased and decreased activities, respectively. protection of cIAP1 from Cr(VI)-induced killing of IKK␤-KM cells might not only be the result of inhibition of caspases but also the result of regulation of intracellular signal transduction.
It has been proposed that Cr(VI)-induced cellular responses are both ROS-dependent and ROS-independent. A limited amount of ROS can be buffered in cells by glutathione and thioredoxin (54,55). This raises the possibility that the increased vulnerability of IKK␤-KM-expressing cells to Cr(VI) may be partially due to a reduced generation of oxidative buffering molecules. Indeed, the gene expression profiling study showed that the lowest expressed gene in IKK␤-KM cells, compared with that in IKK␤ cells, is the gene encoding proteindisulfide isomerase-related protein (Fig. 5), an important member of the thioredoxin superfamily participating in redox regulation (55). Lowered oxidative buffering could lead to oxidative stress. Under this circumstance, the mitochondrial respiratory chain would be easily disrupted. The cells would undergo necrosis rather than apoptosis due to the depressed activation of caspases by Cr(VI) or ROS. It has been demonstrated that activation of caspases requires ATP and reduction of cysteine in the essential active center of caspases (24). To support this, combined treatment of cells with cIAP1 and Nacetyl-L-cysteine to elevate intracellular thio-containing molecules, such as GSH, partially protected IKK␤-KM cells from Cr(VI)-induced killing. 2 In conclusion, we have demonstrated a novel function of NF-B in inhibiting Cr(VI)-induced cell death. The levels of cIAPs that are transcriptionally regulated by NF-B are critical in determining the activation and activity of caspases and the integrity of the Bcl-xl protein. Investigations are currently under way to address whether other oxidative stress inducers, such as H 2 O 2 and nitric oxide, exhibit a similar effect on the cells where NF-B was specifically inhibited by different approaches (e.g. gene knockout for IKK␤ or p65, transfection of degradation-resistant IB␣, or delivery of peptidyl inhibitors for the IKK complex).