A Distinctive Physiological Role for IκBβ in the Propagation of Mitochondrial Respiratory Stress Signaling*

The NFκBs regulate an array of physiological and pathological processes, including propagation of mitochondrial respiratory stress signaling in mammalian cells. We showed previously that mitochondrial stress activates NFκB using a novel calcineurin-requiring pathway that is different from canonical or non-canonical pathways. This study shows that IκBβ is essential for the propagation of mitochondrial stress signaling. Knock down of IκBβ, but not IκBα, mRNA reduced the mitochondrial stress-mediated activation and nuclear translocation of cRel:p50, inhibiting expression of nuclear target genes RyR1 and cathepsin L. IκBβ mRNA knock down also reduced resistance to staurosporine-induced apoptosis and decreased in vitro invasiveness. Induced receptor switching to insulin-like growth factor-1 receptor and increased glucose uptake are hallmarks of mitochondrial stress. IκBβ mRNA knock down selectively abrogated the receptor switch and altered tubulin cytoskeletal organization. These results show that mitochondrial stress signaling uses an IκBβ-initiated NFκB pathway that is distinct from the other known NFκB pathways. Furthermore, our results demonstrate the distinctive physiological roles of the two inhibitory proteins IκBβ and IκBα.

NFB transcription factors play critical roles in the regulation of genes associated with T-cell differentiation, immunity, inflammatory response, cell proliferation/transformation, apoptosis, and metastasis. The NFB pathway responds to a battery of extracellular and intracellular stimuli (for a comprehensive review see Ref. 1), and the downstream transcriptional activators can be classified into two main groups. The first consists of RelA, RelB, and cRel, all of which contain an N-terminal Rel homology domain that has important roles in protein dimerization and DNA binding. The second group consists of p52 and p50, which are processed from the larger p100 and p105, respectively, by partial ubiquitin-mediated degradation. Two major pathways have been described for the activation of NFB, namely the canonical and non-canonical pathways. The canonical pathway involves the activation of RelA, cRel, p50 het-erodimers that are held in the cytosol by inhibitory IB proteins, including IB␣, IB␤, and IB⑀ (2,3). The physiological functions of different inhibitors and their specificity for various Rel proteins remain unclear. The non-canonical pathway is initiated by the IKK␣-mediated phosphorylation of p100, which provides the signal for ubiquitination of p100 and generation of the active p52:RelB dimer (2)(3)(4)(5)(6).
The canonical NFB pathway is stimulated by interleukins, interferons, or chemokines and mediated through phosphorylation and degradation of inhibitory proteins, particularly IB␣. In response to stimulation, IB␣ undergoes IKK␤-dependent phosphorylation and ubiquitin-mediated degradation, liberating the NFB heterodimer. The active heterodimer with unmasked nuclear localization signal is then translocated to the nucleus to carry out its transcriptional activity (2)(3)(4)(5)(6). Many studies of the canonical pathway have focused on IB␣ and its interaction with heterodimeric RelA/p50 proteins. It has been generally assumed that the same mechanism of regulation by inhibitor degradation applies to IB␤.
The many implied roles of the NFB pathway and its response to diverse stimuli (3,7,8) suggest additional mechanisms of activation of this pathway. For example, an IKK-independent pathway involving CKII or tyrosine kinase-mediated phosphorylation of IB␣ at sites other than the IKK target sites has been reported. The precise physiological roles of different pathways and their selectivity for different Rel proteins remain unclear (9 -13). Most of the NFB dimers activate common target genes that coordinate inflammatory response, immune regulation, cell cycle, cell survival, and tumorigenesis.
A number of studies, including ours, have shown that mitochondrial respiratory stress induced by multiple causes, including mitochondrial respiratory inhibitors, partial or complete mtDNA depletion (14 -19), mtDNA mutations (20,21), suppression of mitochondrial transcription (22), and hypoxia (23), induce a mitochondrial stress signaling pathway that is analogous to the retrograde signaling pathway described in yeast cells (24). In contrast to the multifunctional Rtg factors in yeast cells (25)(26)(27)(28)(29), the mitochondrial stress signaling in mammalian cells occurs through increased cytosolic [Ca 2ϩ ] c and activation of cytosolic protein phosphatase calcineurin (Cn). 2 Recently, the * This work was supported by National Institutes of Health Grant CA-22762.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1  mitochondrial dysfunction and associated respiratory stress signaling have been proposed to play a role in aging and age-related pathologies (24). Activation of Cn, which is a critical upstream effecter of the mitochondrial respiratory stress pathway (14,27,28), causes preferential activation and nuclear localization of cRel:p50 dimers and also a number of other Ca 2ϩ -responsive factors (28 -30). Knock out of Calcineurin A␣ (CnA␣) or inhibition of Cn activity by FK506 increased the levels of phosphorylated IB␤ in the cytoplasm and reduced nuclear levels of cRel and p50. We showed that the binding of Cn to the IB␤-cRel:p50 complex and subsequent dephosphorylation of IB␤ causes the release and nuclear translocation of active cRel:p50 heterodimer. This pathway is independent of IKK but likely involves CKII-dependent phosphorylation of IB␤ at Ser-313 and Ser-315 of the C-terminal PEST II domain. Dephosphorylation of IB␤ at these sites in the PEST II domain by Cn is a critical and necessary step for the mitochondrial stress-mediated activation of cRel:p50 heterodimer, and mutations S313A and S315A of IB␤ severely curtailed NFB activation through this pathway (28). In conjunction with other transcription factors that are activated as part of the mitochondrial respiratory stress pathway, cRel:p50 heterodimers activate an array of target genes involved in Ca 2ϩ regulation, glucose metabolism and, most importantly, tumor promotion (15,16,(27)(28)(29)(30)(31). These targets are not considered classical NFB-responsive genes. Studies by our and other laboratories show that mitochondrial respiratory stress signaling induces a metabolic shift through activation of IGF-1R-phosphatidylinositol 3-kinase pathway and also AKT. Inhibition of IGF-1R and phosphatidylinositol 3-kinase induced apoptosis in cells subjected to mitochondrial stress (31)(32)(33)(34).
In this report, we have elucidated the unique and distinctive role of IB␤ in the propagation of mitochondrial respiratory stress signaling, activation of several marker genes, and development of stress-induced invasive phenotypes. Our results show that many of these characteristics of mtDNA-depleted cells are reversed by knock down of IB␤. Although it is generally believed that the roles of IB␣ and IB␤ in NFB signaling are indistinguishable, the present study shows that IB␤ and IB␣ have distinct physiological roles in responding to mitochondrial respiratory stress.

MATERIALS AND METHODS
Cell Lines and Culture Conditions-Murine C2C12 skeletal myoblasts (ATCC CRL1772) were grown in Dulbecco's modified Eagle's medium (Invitrogen) supplemented with 10% fetal bovine serum and 0.1% gentamycin as described before (14). Partial depletion of mtDNA was carried out by treatment with ethidium bromide (100 ng/ml) as described before (14). Selected clones containing ϳ15% mtDNA contents were grown in the presence of 1 mM sodium pyruvate and 50 g/ml uridine. Reverted cells represent mtDNA-depleted cells (with ϳ15% mtDNA contents) grown for 30 cycles in the absence of ethidium bromide until the mtDNA content was reverted back to ϳ80% of control cells. The mtDNA contents were monitored either by real-time PCR or Southern blot hybridization as described before (14,27).
Stable Expression of siRNA and Knock Down of IB␤ and IB␣ mRNAs-Two siRNA sequences targeting mouse IB␤ (5Ј-GAC-TGGAGGCTACAACTAG-3Ј and 5Ј-CAGAGATGAGGGCG-ATGAA-3Ј) were designed and cloned into pSilencer 1.0-U6 vector (Ambion). Empty pSilencer 1.0-U6 vector was used as control. Control and siRNA vector were co-transfected with neomycinresistant pcDNA3 vector (Invitrogen) into control and depleted cells. G418 (1 mg/ml) was added to the medium for selection. Two siRNA sequences against mouse IB␣ (5Ј-AGGCCAGCGTCTG-ACATTA3-Ј and 5Ј-GGCCAGCGTCTGACATTAT-3Ј) were cloned in pSUPER retro puro vector (OligoEngine). 293T cells were infected with the retroviral clones and the vector alone. The medium was used to infect control and mtDNA-depleted cells in the presence of 6 g/ml polybrene. Selection was done in the presence of puromycin (10 g/ml) for control cells and 100 g/ml for depleted cells because these cells were resistant to lower dose of the antibiotic. After 3 weeks, well separated individual colonies were picked and grown. The protein levels of IB␤ and IB␣ were checked by Western blot using antibodies against respective protein (Santa Cruz Biotechnology). Cells with significantly lower level of IB␤ or IB␣ compared with their respective controls were taken as stable knockdown cells for further study.
Subcellular Fractionation and Immunoblot Analysis-The subcellular fractions were prepared essentially as described before (14,28) in the presence of protease and phosphatase inhibitors. Proteins (30 g each) were resolved on 10 or 12% SDS-polyacrylamide gels and detected by Western blot analysis as described as before (14). Blots were developed using Super Signal West Femto maximum sensitivity substrate (Pierce).
Measurement of Mitochondrial Membrane Potential-The mitochondrial membrane potential (⌬⌿m) was assayed spectrofluorometrically by loading the cells with a cationic dye, Mito Tracker Orange CM-H 2 TMRos (MTO) (Molecular Probe Inc.) as described before (15). The rate of dye uptake was recorded as a measure of ⌬⌿m using a Delta RAM PTI spectrofluorometer at 525 ex/575 em as fluorescence units/min.
Calcium Release Assay-Ca 2ϩ release was measured essentially as described before (15,35). The cells were suspended in intracellular medium (20 mM HEPES-Tris, pH 7.2, 120 mM NaCl, 5 mM KCl, 1 mM KH 2 PO 4 ) passed through a Chelex (Sigma) column to remove residual Ca 2ϩ . The cells were loaded with membrane-permeable Fura 2FF/FA (1 M) for 20 min at room temperature, pelleted, and resuspended in 1 ml of intracellular medium for measuring both basal [Ca 2ϩ ] as well as the RyR1-specific Ca 2ϩ release in response to 20 mM caffeine and 10 M acetylcholine. Fluorescence was monitored at excitation 340/380 nm and emission 510 nm. Calibration of Fura 2FF/FA signal was carried out using a calibration buffer (10 mM EGTA-Tris-HEPES, pH 8.5, and 5 mM CaCl 2 ).
Microarray Analysis-Total RNA from control, mtDNA-depleted, IB␤ knock down (KD)/Control, and IB␤ KD/Depleted C2C12 cells was isolated using TRIzol according to the manufacturer's instructions. RNA was analyzed on MOE430A chips (Affymetrix). Statistical significance established with Partek (two-way analysis of variance significant to Ͻ0.0005). Gene selection was performed using the Spotfire program. The total number of spots matching criteria was 43, representing 25 Gene Ontology header-mapped genes with known functions.
Real-time PCR-Total RNA (5 g) from cells was reversetranscribed using the High Capacity cDNA Archives kit (Applied Biosystems, Inc.). Real-time amplifications were performed using specific primers in an ABI 7300 real-time PCR machine using SYBR Green Master Mix (ABI). Each 25-l reaction contained 25 ng of cDNA and 200 nM forward and reverse primers. Two-step reverse transcription PCR was carried for 40 cycles. Data were analyzed using ABI Relative Quantitation analysis software. ␤-Actin served as an internal control. Target gene expression was presented as -fold increase over control levels.
Glucose Uptake Assay-Glucose uptake using 2-[ 3 H]deoxyglucose was measured as described before (31,36). 1 ϫ 10 6 cells grown in 6-well plates were serum-starved for 4 h and incubated for 30 min in glucose-free medium. 1 Ci of 2-[ 3 H]deoxyglucose was added and incubated for 15 min. Cells were rapidly chilled to 4°C, washed four times at 4°C, and transferred to scintillation vials for counting.
Matrigel Invasion Assay-The in vitro invasion assays were carried out as described previously (16). 4 ϫ 10 4 cells were layered into the invasion chambers containing 1:3 diluted Matrigel (BD Biosciences) and incubated in wells (12-well plates) containing 1 ml of growth medium in each well for 24 -48 h at 37°C. Matrigel layers with non-invading cells were removed, and the membranes with invaded cells were stained with Meyer's hematoxylin, cut, and mounted on slides to view under the bright field Olympus BX61 microscope.
Adherence-independent Growth Assay on Soft Agar-The cells (2 ϫ 10 3 /well) were suspended in soft agar (2%) mixed with growth medium and plated in 12-well plates. After 48 h of incubation the plates were imaged and photographed using a bright field imaging microscope.
Immunocytochemistry-Cells were grown on coverslips and processed for antibody staining essentially as described before (14,28). Cells were immunostained with 1:100 dilution of primary antibody for 1 h and 1:100 dilutions of Alexa Fluor-conjugated secondary antibody for 1 h at 37°C. Fluorescence microscopy was carried out using an Olympus BX61 fluorescence microscope.

Distinct Roles of IB␤ and IB␣ in the Mitochondrial
Respiratory Stress-mediated Activation of NFB-To establish the roles of inhibitory IB proteins in the propagation of mitochondrial respiratory stress signaling, we generated cell lines deficient in IB␣ or IB␤ by siRNA expression (named IB␣ KD and IB␤ KD, respectively) from control and depleted C2C12 murine skeletal myoblasts (14). As shown in Fig. 1A, IB␣ and IB␤ knock down (KD) by mRNA silencing in control and mtDNA-depleted cells (hereafter called depleted cells) markedly diminished the levels of these respective proteins without affecting the levels of the other protein. Silencing the IB␤ mRNA in mtDNA-depleted cells (hereafter called IB␤ KD/Depl) caused reduced nuclear levels of cRel and p50 proteins when compared with mtDNA-depleted cells with mock transfection (Fig. 1B). On the other hand, cRel and p50 levels in the cytoplasm were increased by IB␤ silencing in depleted cells as compared with control cells although the level of cytoplasmic CnA␣ remained the same in IB␤ KD/Depl cells as compared with the depleted cells alone. Knock down of IB␣ in depleted cells (hereafter called IB␣ KD/Depl), by contrast, did not affect the nuclear or cytoplasmic levels of cRel, p50, or CnA␣ (Fig. 1B). Level of p65 in the nucleus was not significantly affected by knock down of IB␤ but marginally increased by IB␣ mRNA knock down. We also observed an increase in levels of cytosolic p65 in these samples. These results show that IB␤ may be selectively involved in responding to mitochondrial respiratory stress.
Temporal Order of Mitochondrial Respiratory Dysfunction, Changes in Ca 2ϩ Homeostasis, and Calcineurin-dependent NFB Activation-To understand the sequence of events leading to NFB activation, we studied mitochondrial membrane potential (⌬⌿m) and cellular Ca 2ϩ levels in IB␤ KD/Depl and IB␣ KD/Depl cells. Mitochondrial ⌬⌿m was measured as the rate of uptake of Mitotracker Orange CM-H 2 TMROS in control cells and depleted cells as well as IB␤ KD/Depl and IB␣/ Depl cells. The reduced form of the dye is taken up by mitochondria in proportion to ⌬⌿m, which fluoresces upon oxidation inside respiring mitochondria. Control cell mitochondria exhibited a steeper membrane potential as indicated by a time-dependent increase in fluorescence ( Fig. 2A). Depleted cells, and also cells treated with the mitochondrial ionophore, carbonyl cyanide-m chlorophenylhydrazone (CCCP), showed decreased fluorescence indicative of disrupted ⌬⌿m. IB␤ and IB␣ mRNA knock down had no effect on ⌬⌿m in depleted cells. Additionally, IB␤ and IB␣ mRNA knock down in control cells also did not affect the rate of increase in fluorescence (data not shown). These results suggest that perturbation of ⌬⌿m is an upstream event that marks the initiation of respiratory stress.
We also assessed steady state Ca 2ϩ levels and caffeine-evoked Ca 2ϩ release in control, IB␤ KD/Depl, and IB␣ KD/Depl cells (Fig. 2B). Both IB␣ KD/Depl and IB␤ KD/Depl cells had a significant degree of Ca 2ϩ release in response to caffeine. These calcium pools represent the RyR1 channelspecific Ca 2ϩ stores and are comparable with the responses observed in depleted cells as depicted in Fig. 2C (14). Both cell lines also displayed increased basal Ca 2ϩ compared with control cells, which is another characteristic feature of depleted cells. In confirmation of our previous results (14), control cells responded only to acetylcholine, an agonist of the inositol triphosphate channel, and did not respond to caffeine. Therefore, this suggests that the marked change in the Ca 2ϩ homeostasis and/or steady state levels of Ca 2ϩ in response to mitochondrial stress are upstream events from the activity of IB␤ in mitochondrial stress signaling.
Role of IB␤ in the Propagation of Mitochondrial Stress Signaling and Expression of Nuclear Target Genes-In our previous study we showed that mitochondrial stress-mediated activation of NFB leads to transcriptional activation of a set of genes that are not regarded as classical NFB targets (28) . Fig. 3, A and B, shows immunoblots of proteins from post-mitochondrial supernatant fraction and immunohistograms of control and mtDNA-depleted C2C12 cells as well as IB␤ KD/Depl and IB␣ KD/Depl cells. The levels of RyR1 and cathepsin L (Cath L) were markedly reduced in IB␤ KD/Depl cells as compared with the parent mtDNA-depleted cells (Fig. 3A). The effect of IB␤ and IB␣ mRNA knock down on the levels of RyR1, Cath L, and transforming growth factor ␤, another target gene of the stress-signaling pathway, was further validated by immunohistochemical staining. As seen from Fig. 3B, knock down of IB␣ mRNA did not have a significant effect on the levels of these proteins as judged by the intensity of immunostaining. Expression of RyR1 and Cath L was further measured in terms of mRNA levels using real-time PCR. In line with the immunohistochemical data, results of real-time PCR (Fig. 3C) also show that the steady state levels of Cath L and RyR1 mRNAs were significantly reduced in IB␤ KD/Depl cells as compared with the parent depleted cells. IB␣ KD/Depl cells had only a marginal effect on the levels of Cath L and RyR1 mRNAs.
To assess the distinctive functional attributes of the two inhibitory proteins, we measured the TNF␣-induced expres-  A, steady state levels of Cathepsin L (Cath L) and ryanodine receptor 1 (RyR1) proteins were detected in total homogenates (30 g) of control, mtDNA-depleted, and IB␤ KD/Depl cells by immunoblot analysis. Na ϩ /K ϩ ATPAse (ATPase) was used as loading control. B, levels of RyR1, CathL, and transforming growth factor ␤ (TGF␤) proteins were determined by immunofluorescent labeling of the indicated cell lines. The cells were grown on coverslips and labeled with the indicated antibodies as described under "Materials and Methods." C, mRNA levels of target genes cathepsin L and RyR1 were measured by real-time PCR of total RNA isolated from the indicated cells lines as described under "Materials and Methods." Values represent average of triplicates and were normalized against ␤-actin as an internal control also run in triplicates. sion of the classical NFB targets IL-6, MnSOD, and TNF␣. IL-6 mRNA was induced 2.5-fold in response to TNF␣ treatment in mtDNA-depleted cells, while the extent of induction was severely curtailed in IB␣ KD/Depl cells (Fig. 4A). IB␤ mRNA knock down had no effect upon the extent of TNF␣mediated IL-6 mRNA induction. Surprisingly, IB␤ mRNA knock down caused a nearly 9-fold induction of MnSOD. This suggests that IB␤ may have a specific negative modulatory effect on the expression of this gene. IB␣ knock down also curtailed the TNF␣-mediated autoregulation of gene expression, while IB␤ knock down caused a 2-fold higher level of induction over mtDNA-depleted cells treated with TNF␣.
Although the possible negative modulatory role of IB␤ in MnSOD and TNF␣ gene expression was surprising, these results collectively show the distinctive physiological roles of the two inhibitory proteins being compared here. We carried out cDNA microarray analysis to understand the range of genes affected by the mitochondrial stress-activated NFB signaling and the involvement of IB␤ in the expression of these genes. Total RNA from control and depleted C2C12 cells with or without siRNAbased knock down of IB␤ was analyzed. We identified a set of genes that were up-regulated at least 2-fold by mtDNA depletion in a manner that was dependent upon the activity of IB␤. As represented in Fig. 4B and Table 1, ϳ40 genes fit these criteria. For these genes the effect of mtDNA depletion was  mediated by signaling through IB␤. Importantly, there was no significant change in the level of expression of these genes in control cells with IB␤ knock down. The up-regulated genes represent diverse roles in cellular metabolism, including regulating signal transduction, cellular redox function, ion transport, glucose metabolism, mitochondrial energetics, cell adhesion, cell cycle, and tumorigenesis (see Table 1). These results show that IB␤ plays a critical role in the mitochondrial stressinduced NFB activation and that this pathway affects the expression of a large number of nuclear genes associated with an array of critical cellular functions.

The NFB-dependent Mitochondrial Stress Response Pathway Modulates Glucose Uptake and Regulates the Expression of
Glut4 and IGF-1R-A characteristic feature of fast growing tumor cells is high utilization of glucose in glycolysis despite adequate oxygen utilization and mitochondrial electron transport function (33,34,37). Energy thus derived supports uninhibited tumor cell proliferation. Previously we and others have shown that the mtDNA-depleted cells have altered metabolism and invasive properties (16, 31, 33, 38 -40). Therefore, we examined the glucose uptake capability of the mtDNA-depleted cells and investigated whether IB␤ mRNA knock down affects glucose uptake. In keeping with our recent results, mtDNA-depleted cells showed significantly elevated levels of glucose uptake as compared with controls (Fig. 5A). IB␤ mRNA knock down blocked 60% of this increase. IB␣ mRNA knock down also reduced glucose uptake, although to a lesser extent.
In a recent study we showed that mtDNA depletion or treatment with mitochondrial ionophores like CCCP induced the IGF-1R-regulated pathway (31). Real-time PCR results in Fig. 4B show that the increase in Glut4 and IGF-1R mRNA in response to mitochondrial stress was markedly abated by IB␤ mRNA knock down, suggesting that mitochondrial stress-mediated metabolic shift requires functional IB␤-dependent NFB pathway. Notably, IB␣ knock down caused a further increase in Glut4 mRNA levels, suggesting its negative modulatory role on gene expression. In the case of IGF-1R mRNA, the effect of IB␣ knock down was comparatively marginal compared with the effect of IB␤ mRNA knock down (Fig. 5B). Further, we tested the effect of picropodophyllin, a specific inhibitor of the IGF-1 receptor (41) that induces apoptosis in mtDNA-depleted cells (31). The terminal deoxynucleotidyltransferase-mediated dUTP nick end-labeling (TUNEL) assay results in Fig. 5C show that the depleted as well as IB␣ KD/Depl cells had vastly increased number of apoptotic cells in response to picropodophyllin addition but IB␤ mRNA knock down had no effect on picropodophyllin susceptibility, similar to the control cells (Fig. 5C). These results confirm the importance of IB␤ in the regulation of nuclear gene expression and metabolic function of cells undergoing mitochondrial genetic stress.

IB␤-dependent NFB Activation Plays a Role in Mitochondrial Stress-induced Survival and Invasiveness-C2C12
skeletal muscle cells and A549 lung carcinoma cells subjected to mitochondrial stress developed increased invasiveness and resistance to staurosporine-(STP) and etoposidemediated apoptosis (15,27,30). In this study we tested the role of IB␤-dependent signaling in stress-induced resistance to apoptosis. Depleted cells showed markedly reduced STP-induced apoptosis compared with control cells but IB␤ knock down in these cells impeded resistance to apoptosis (Fig. 5D). IB␣ knock down, on the other hand, did not change the number of cells undergoing apoptosis. These results suggest that IB␤-dependent signaling is important in mediating mitochondrial stress-induced resistance to apoptosis.
We used the Matrigel invasion assay to show that that mtDNA depletion causes a marked increase in invading cells (Fig. 6A). IB␤ knock down, but not IB␣ knock down, reversed the level of invasion to near control cell level, confirming the role of the IB␤-dependent NFB pathway in the propagation of mitochondrial stress signaling. It is also apparent from the growth pattern in soft agar (Fig. 6B) that mtDNA-depleted cells proliferated into an attachment-independent mass of cells while control and IB␤ knock down cells grew as monolayers on soft agar (Fig. 6B).
The Role of IB␤ in Mitochondrial Stress-induced Changes in Cytoskeletal Structure and Function-C2C12 cells differentiate to form multinucleated myotubes in response to serum deprivation. MtDNA depletion markedly affected the ability of cells to form multinucleated myotubes (Fig. 7A). Consistent with the altered myogenic property, mtDNA-depleted cells contained markedly reduced levels of ␣5␤1 integrins around the plasma membrane. In view of the suggested role of integrins in skeletal myogenesis (42), our results suggest that mitochondrial stress signaling disrupts myogenesis.
To further investigate the nature of cytoskeletal changes in response to mitochondrial stress, we examined the tubulin network in mtDNA-depleted cells. MtDNA-depleted cells have an abnormal cytoskeletal organization with extensively condensed tubulin organization (Fig. 7B). Tubulin bundles were concentrated around the filopodia-like structures. IB␤ (but not IB␣) mRNA knock down reverted the cytoskeletal disorganization, providing a possible mechanistic explanation for the reduction of invasiveness seen previously.

Role of IB␤ in Mediating Mitochondrial
Stress Signaling-In this study we have provided evidence for the intermediaries of the NFB pathway that propagate mitochondria-to-nucleus stress signaling and stress-mediated changes in cell morphology and phenotype. Specifically, our results show that IB␤ is a key regulatory molecule in modulating mitochondrial stress signaling and activating a set of genes that are distinct from classical NFB targets. The dependence on Cn for the activation of NFB in mitochondrial stress signaling is unique and appears to be specific to this signaling pathway (28). The results also point to the possibility that the IB␤-dependent pathway plays an important role in the activation of tumor-promoting genes through NFB/cRel, thus presenting an additional and specific molecular target for controlling oncogenic progression. In this study we have also demonstrated the importance of IB␤ in the mitochondrial stress-mediated metabolic switch and cytoskeletal reorganization.
A number of studies have implicated NFB signaling in tumor development and progression (43,44). Studies have also suggested that NFB-dependent tumorigenesis and invasion are cell-and tissue-specific (44 -46). Many of these studies point to the possibility that NFBs activate anti-apoptotic genes, providing protection against cell death and thus supporting the proliferation of tumor cells (47)(48)(49). Constitutive expression of NFB in tumors has been suggested to play critical roles in tumor cell chemoresistance, growth promotion, invasion, metastasis, and tumor angiogenesis (50 -53). Attempts to develop pharmacological interventions based upon NFB blockade for the treatment of cancer have been largely unsuccessful because of the side effects of NFB inhibition (4). Furthermore, the small molecule inhibitors of IKK or NFB that have been developed have not been sufficiently specific (50). The stained cells were visualized under bright field microscope as described under "Materials and Methods." B, the role of IB␤ in supporting the invasive property of mtDNA-depleted cells was investigated by adherence-independent growth of control (a), depleted (b), and depleted cells knocked down for IB␤ expression (IB␤ KD/Depl) (c). Effect of IB␤ knock down on cell growth was compared with that of IB␣ KD/Depl cells (d). 2 ϫ 10 3 cells were grown in 6-well plates containing layers of 2% agar. Adherence-independent growth was assessed as described under "Materials and Methods." FIGURE 7. Effect of IB␤ knock down on respiratory stress-mediated changes in cell morphology and cytoskeletal organization. A, fusion of myoblasts into multinucleated myotube structure was determined by serum deprivation in control and mtDNA-depleted cells (upper panels). Cells were grown on coverslips in regular (10% fetal bovine serum) growth medium and after 48 h replaced with differentiating Dulbecco's modified Eagle's medium containing 2% fetal bovine serum. Cells were grown for another 48 h, and the phase contrast images were captured using an Olympus BX61 microscope. The control cells showed formation of multinucleated myotubes whereas they were absent in depleted cells. The cells were also immunostained for integrin ␣5␤1 (lower panels) as described under "Materials and Methods," and images were taken using a fluorescence microscope. B, changes in cytoskeletal organization were detected by immunostaining of control, depleted, IB␤ KD/Depl, and IB␣ KD/Depl cells with ␤-tubulin antibody as described under "Materials and Methods." Arrows indicate the filopodia with newly formed tubulin network.

Distinctive Features of IB␤-mediated NFB Pathway-
There is increasing evidence that the interaction of Rel proteins with IB inhibitory proteins is a necessary regulatory step for activation of the NFB pathway (54 -56). Absence of these inhibitory proteins severely affects the nuclear translocation of NFB/Rel transcription factors to the nucleus and activation of NFB-responsive genes. Consistent with this view, we have shown that the IB␤-dependent pathway plays a critical role in the mitochondrial respiratory stress-mediated resistance to apoptosis and in tumor cell invasion.
Because of its slow turnover rate and relatively unchanging steady state levels in response to cytokine and chemokine treatment, IB␤ is believed to regulate the constitutive NFB pathway while IB␣ is involved in the inducible phase of regulation (57,58). In cells treated with TNF␣ or interleukins, IB␣ is transcriptionally induced within minutes of treatment and undergoes rapid turnover at the end of signaling (59 -61). The level of IB␤, on the other hand, undergoes only a marginal change in response to these inducers (27,57,62). Detailed biochemical and crystal structure studies suggest that binding and affinity of IB␣ to p65 homodimers significantly differ from that of IB␤. IB␤ binds to Rel factors more tightly, masking their nuclear translocation signal and preventing nuclear entry of oligomeric complexes (60). Chemical cross-linking of cytosolic fractions from mtDNA-depleted cells suggested that IB␤ largely exists in a complex with c-Rel and p50 heterodimers (28).
The results of this study confirm the functional distinction between IB␤ and IB␣ in mitochondrial stress signaling. In our experiments, the knock down of IB␤, but not IB␣, decreased the nuclear localization of p50 and cRel as well as the expression of target gene RyR1 and Cath L. IB␤ knock down also blocked the changes in morphology, viability, and glucose uptake found in mtDNA-depleted cells. Additionally, knock down of IB␤ reversed many of the changes in nuclear gene expression induced by mtDNA depletion seen in transcriptional arrays, implicating this protein in mitochondria-to-nucleus communication. Remarkably, IB␤ knock down also enhanced the TNF␣-mediated increase in downstream mRNA expression, in marked contrast to the effect of IB␣ knock down. These results provide compelling evidence for different physiological roles for the two major IB proteins, although there may be some overlap in the nuclear targets of these factors as observed in the expression of Glut4 and IGF-1R.
Permissive Role of IB␤ as Opposed to Inhibitory Role of IB␣-Many studies have shown that IB␣ degradation following cytokine or chemokine induction triggers a marked increase in the nuclear localization of p65:p50 Rel proteins and induced expression of target genes. IB␣ knock down in mtDNA-depleted cells also resulted in the increased nuclear localization of p65:p50 in keeping with its well acknowledged "inhibitory" function. In contrast, knock down of IB␤ mRNA caused a marked reduction in the nuclear cRel:p50 levels, suggesting that IB␤ does not function as a classical inhibitor. In fact, cytosolic IB␤ and Cn are required for the release of active cRel:p50 and their nuclear translocation (28). Therefore, IB␤ is more accurately described as having a permissive role in NFB signaling in response to mitochondrial stress. In sum-mary, we present compelling evidence for the distinct physiological roles of IB␤ and IB␣ in mediating the NFB signaling. The calcineurin-regulated activation of IB␤ in response to mitochondrial stress is an important landmark of this extensive signaling pathway that is interrupted by blocking the IB␤ mRNA expression. As outlined in the model presented in Fig. 8, IB␤ knock down inhibits the various morphologic and functional changes seen in response to mitochondrial stress.