Long-term Incubation with Proteasome Inhibitors (PIs) Induces IκBα Degradation via the Lysosomal Pathway in an IκB Kinase (IKK)-dependent and IKK-independent Manner*

Background: IκB, a cytoplasmic inhibitor of NF-κB, is degraded via the proteasome. Results: Paradoxically, proteasome inhibitors (PIs) induce IκBα degradation via the lysosome in an IKK-dependent and IKK-independent manner. Conclusion: PI-induced IκBα degradation results in NF-κB activation that confers resistance to PI-induced cancer cell death. Significance: This provides a molecular mechanism to enhance the anti-cancer efficacy of PIs. Proteasome inhibitors (PIs) have been reported to induce apoptosis in many types of tumor. Their apoptotic activities have been suggested to be associated with the up-regulation of molecules implicated in pro-apoptotic cascades such as p53, p21Waf1, and p27Kip1. Moreover, the blocking of NF-κB nuclear translocation via the stabilization of IκB is an important mechanism of PI-induced apoptosis. However, we found that long-term incubation with PIs (PS-341 or MG132) increased NF-κB-regulated gene expression such as COX-2, cIAP2, XIAP, and IL-8 in a dose- and time-dependent manner, which was mediated by phosphorylation of IκBα and its subsequent degradation via the alternative route, lysosome. Overexpression of the IκBα superrepressor (IκBα-SR) blocked PI-induced NF-κB activation. Treatment with lysosomal inhibitors (ammonium chloride or chloroquine) or inhibitors of cathepsins (Z-FF-FMK or Z-FA-FMK) or knock-down of LC3B expression by siRNAs suppressed PI-induced IκBα degradation. Furthermore, we found that both IKK-dependent and IKK-independent pathways were required for PI-induced IκBα degradation. Pretreatment with IKKβ specific inhibitor, SC-514, partially suppressed IκBα degradation and IL-8 production by PIs. Blockade of IKK activity using insolubilization by heat shock (HS) and knock-down by siRNAs for IKKβ only delayed IκBα degradation up to 8 h after treatment with PIs. In addition, PIs induced Akt-dependent inactivation of GSK-3β. Inactive GSK-3β accelerated PI-induced IκBα degradation. Overexpression of active GSK-3β (S9A) or knock-down of GSK-3β delayed PI-induced IκBα degradation. Collectively, our data demonstrate that long-term incubation with PIs activates NF-κB, which is mediated by IκBα degradation via the lysosome in an IKK-dependent and IKK-independent manner.

The 26S proteasome, a multicatalytic enzyme complex that is expressed in the nucleus and cytoplasm of all eukaryotic cells, has emerged as a novel putative target for cancer therapy. It is the main intracellular, nonlysosomal, ATP-dependent proteolytic system by which various proteins involved in signal transduction, cell-cycle regulation, and apoptosis are degraded (1)(2)(3). Inhibition of this ubiquitin-mediated degradation of several regulatory proteins has been reported to induce cellular apoptosis in several types of cancer including lung cancer (4), colon cancer (5), breast cancer (6), and pancreatic cancer (7) as well as multiple myeloma (8). Proteasome inhibitors such as bortezomib (N-pyrazinecarbonyl-L-phenylalanine-L-leucine boronic acid; known as PS-341) 2 and synthetic peptide aldehydes MG132 show anti-tumor activity. Their molecular mechanisms of anti-tumor activity include the rapid accumulation of p53 and p27 kip1 , phosphorylation of c-Jun NH 2 -terminal kinase (JNK) and c-Jun, stabilization of BH3 pro-apoptotic proteins BIK, NOXA, and BIM, down-regulation of anti-apoptotic proteins Bcl-2 and Bcl-X L and stabilization of cyclin D, E, and A (9 -11). Moreover, proteasome inhibitors inhibit NF-B activity by blocking the degradation of its cytoplasmic inhibitor, IB (12).
NF-B, a pleiotropic transcription factor, is normally sequestered in the cytoplasm in an inactive form, bound to the inhibitory proteins (IBs). NF-B promotes the expression of various cytokines such as interleukin-6 (IL-6), IL-8, and TNF-␣ and adhesion molecules such as ICAM-1 and VCAM-1 as well as anti-apoptotic proteins such as survivin, IAP1/2, Bfl-1/A1, and cFLIP. Upon cell stimulation by a broad variety of stimuli including viral infection, growth factors, cytokines, or chemotherapeutic agents, IKK␣/␤ is activated. Active IKK directly phosphorylates IB␣ at Ser 32 and Ser 36 residues, leading to ubiquitination at Lys 21 and Lys 22 , and degradation of IB␣ through the 26S proteasome, resulting in nuclear translocation of the NF-B subunit complexes (13)(14)(15). In the nucleus, * This work was supported by Grant 0320030170 from the Seoul National NF-B binds to its cognate site, B element, and transactivates the downstream genes (13)(14)(15). Many types of cancer show constitutive or increased activity of NF-B and increased activation of NF-B confers resistance to cell death (16). Therefore, NF-B has been suggested to be related to increased survival in many tumor cells.
However, in this study, we found that long-term incubation with proteasome inhibitors (PS-341 or MG132) induces irreversible degradation of IB␣ via an alternative pathway, lysosome. After treatment with PIs, the IKK-dependent mechanism during early phase and IKK-independent mechanism during late phase are responsible for PI-induced IB␣ degradation, and inactive GSK-3␤ is involved in phosphorylation and degradation of IB␣ by PIs.
Quantitative Real-time PCR-Total RNA from NCI-H157 cells was isolated using the RNeasy kit (Qiagen, Hilden, Germany). cDNA was synthesized from 1 g of total RNA using the Reverse Transcription system (Promega, Madison, WI). PCR amplification was performed with 2ϫ TaqMan gene expression master mix (Applied Biosystems, Carlsbad, CA). The primer information is as follows: cIAP2 (Hs00985031_g1), IL-8 (Hs00174103_m1), GAPDH (Hs99999905 _m1). The primers were obtained from Applied Biosystems.
Western Blot Analysis-Proteins were resolved by 4 -12% SDS-polyacrylamide gel electrophoresis (PAGE) and transferred to nitrocellulose membranes. The membranes were blocked with 5% skim milk-PBS-0.1% Tween 20 for 1 h before being incubated overnight at 4°C with primary antibodies in 5% skim milk-PBS-0.1% Tween 20. The membranes were then washed three times in 1ϫ PBS-0.1% Tween 20 and incubated with horseradish peroxidase-conjugated secondary antibodies in 5% skim milk-PBS-0.1% Tween 20 for 1 h. After successive washes, the membranes were developed using an ECL kit.
Determination of Cytokine Secretion-Cytokine levels in culture supernatants were determined using a commercially available ELISA kit for IL-8, according to the manufacturer's instructions.
20S Proteasome Activity Assay-Proteasome activity was determined using a 20S proteasome activity assay kit (Chemicon, Temecula, CA) according to the manufacturer's specifications. In brief, cell lysates were incubated with proteasome substrate Suc-LLVY-AMC for 1 h at 37°C. The fluorescence from the mixture was quantified using a 380/460 nm filter set in a fluorometer.
Kinase Assay-The activities of IKK or GSK-3␤ were assessed by an in vitro kinase assay. In brief, the IKK or GSK-3␤ complex was immunoprecipitated with anti-IKK␣ or anti-GSK-3␤ antibodies. The immunoprecipitates were incubated at 30°C for 30 min in a kinase buffer containing 0.5 g of recombinant IB␣ (or 0.1 g/l recombinant Tau) and 0.2 mM of ATP. The kinase reaction products were subjected to SDS-PAGE in 4 -12% gels and then was transferred to a nitrocellulose membrane and analyzed by immunoblotting with anti-p-IB␣ or anti-p-Tau antibodies.
Transfection of siRNAs-Transfection of siRNAs targeting LC3B, IKK␤, or GSK-3␤ genes (Cell Signaling, Danvers, MA) was carried out using Lipofectamine 2000 according to the manufacturer's specifications. 100 nM siRNA was sufficient to mediate silencing. After 48 h, the cells were used in the experiments indicated.
Transduction of Adenoviruses or Transfection of Plasmid Vectors-Cells were plated in a 6-well tissue culture plate. After overnight incubation, cells were transduced at multiplicities of infection (MOI) of 50 by adenovirus vector expressing IB␣ superrepressor (IB␣-SR) cDNA in which serine 32/36 was substituted with alanine in complete RPMI for 2 h with gentle shaking, and then washed with PBS and incubated with growth medium at 37°C, 5% CO 2 until use. Cells were transfected with plasmid vectors expressing HA-tagged WT-Akt, dominant negative Akt (DN-Akt), WT-GSK-3␤, or GSK-3␤ cDNA in which serine 9 was substituted with non-phosphorylatable alanine (S9A). After 48 h, the cells were used in the experiments indicated.
Heat Treatment-Heat stress (HS) was induced by incubating cells in a water bath at 43°C. After HS treatment, the culture medium was replaced with fresh medium. Cells were allowed to recover in a 5% CO 2 incubator at 37°C.
Statistical Analysis-Data were subjected to Student's t test for analysis of statistical significance, and a p value of Ͻ0.05 was considered to be significant.

PIs Increased NF-B-regulated Gene
Expression-To determine the impact of proteasome inhibition on the NF-B pathway, we first analyzed the dose-and time-dependent expression of NF-B-regulated genes by PS-341 or MG132. Upon PI stimulation, transcripts of COX-2, cIAP2, and IL-8 were induced (Fig. 1A). COX-2 was hardly detectable in the basal state. More than 10 nM of PS-341 or 1 M of MG132 was required to increase COX-2 protein expression. COX-2 started to increase at 4 h after treatment with PS-341 or MG132 and increased further in a time-dependent manner ( Fig. 1, B and C). The time course of PI-induced cIAP2 expression was similar to that of COX-2 expression (Fig. 1D). Moreover, Fig. 1, E and F show a significant increase in XIAP expression and the secretion of pro-inflammatory cytokine IL-8 in PS-341-treated cells. These results indicate that PIs up-regulate the expression of NF-Bregulated genes.

PI-induced Up-regulation of NF-B-regulated Proteins Was Associated with IB␣ Phosphorylation and Its Subsequent
Degradation-As NF-B exists in an inactive form in the cytoplasm bound to inhibitory protein IB␣, degradation of IB␣ through the 26S proteasomal pathway is a prerequisite for the activation of NF-B. To test the possibility that IB␣ degradation is required for PI-induced up-regulation of NF-B-regulated proteins, we measured IB␣ protein levels from control and PI-treated cells by Western blot analysis. IB␣ was markedly degraded after 4 h of incubation with PS-341 or MG132, and did not recover up to 24 h (Fig. 2, A and B). Because IB␣ degradation is preceded by phosphorylation of two serine residues (Ser 32 and Ser 36 ), we tested the levels of phosphorylated  NOVEMBER 8, 2013 • VOLUME 288 • NUMBER 45 IB␣ from PI-treated cells. As expected, IB␣ phosphorylation occurred earlier than IB␣ degradation. Phosphorylation of IB␣ was increased 4 h after PS-341 or MG132 treatment and returned to basal levels at 8 h (Fig. 2B). We next evaluated whether IB␣ phosphorylation is necessary for the PI-induced degradation of IB␣ and COX-2 induction. Cells were infected with either control adenovirus or adenovirus expressing IB␣superrepressor (Ad-IB␣-SR) at the dose of 50 MOI for 48 h, and then stimulated with PS-341 (50 nM) for 24 h in NCI-H157 and A549 cells. IB␣ superrepressor (IB␣-SR) in which serine 32/36 was substituted with alanine is not phosphorylated, and it was not degraded by PS-341. This result indicates that IB␣ phosphorylation is necessary for its degradation by PS-341 (Fig.  2C). PS-341-induced up-regulation of COX-2 was completely suppressed in Ad-IB␣-SR-infected cells, which implies that IB␣ degradation is necessary for COX-2 induction by PS-341 (Fig. 2C). Overexpression of IB␣-SR has been well-known to block NF-B activation (17). However, the possibility that strong overexpression interferes with the translation of COX-2 cannot be excluded. Taken together, these observations indicate that PI-induced upregulation of NF-B-regulated proteins is associated with IB␣ phosphorylation and its subsequent degradation.

IKK-dependent and -independent IB␣ Degradation by PIs
PI-mediated Degradation of IB␣ Is through the Lysosomal Pathway-Proteasome inhibitors have been shown to block proteolysis of the phosphorylated IB␣ through the 26S protea-some in response to IL-1␤ and TNF-␣ (18). Paradoxically, longterm incubation with PIs induced IB␣ degradation in this study. To confirm whether proteasome activity was effectively blocked by PI, a time-dependent proteasome activity was measured after treatment with PS-341. Proteasome activity was effectively suppressed by PS-341 for up to 24 h (Fig. 3A). As proteolysis of ubiquitinated proteins via the proteasomal pathway was blocked, ubiquitinated proteins in whole cell lysates were significantly increased in PS-341-treated cells (Fig. 3B). Moreover, short-term incubation (1 h) of MG132 blocked TNF-␣-induced IB␣ degradation (Fig. 3C), which supports that PIs effectively block proteasome activity.
Eukaryotic cells have two major systems for protein degradation: the proteasome by which the majority of proteins (more than 80%) are degraded and the lysosomal apparatus. To test whether the lysosomal pathway is involved in this process, cells were pretreated with lysosomal inhibitors, chloroquine (100 and 200 M) and ammonium chloride (NH 4 Cl, 5-50 mM) for 2 h before addition of PS-341 and TNF-␣. Pretreatment with chloroquine or NH 4 Cl stabilized IB␣ in response to PS-341 (Fig. 3, D and E). In contrast, both lysosomal inhibitors had no effect on TNF-␣-induced degradation of IB␣ (Fig. 3F). Moreover, treatment with inhibitors of lysosomal digestive enzymes such as cathepsins (Z-FF-FMK or Z-FA-FMK) suppressed PIinduced IB␣ degradation (Fig. 3G).
Autophagy is one of the intracellular protein degradation mechanisms through the lysosomal machinery. During autophagy, light chain 3B (LC3B)-II expression is elevated. LC3B-II accumulation is a hallmark of autophagy activation (19). Both PS-341 and MG132 induced time-dependent LC3B accumulation and knockdown of LC3B using siRNAs suppressed IB␣ degradation in PS-341-or MG132-treated NCI-H157 cells (Fig. 3H). These results indicate that PI-induced IB␣ degradation requires the activity of lysosomal hydrolases.
IKK Activity Is Required for the Rapid PI-induced IB␣ Degradation-IB␣ degradation through the proteasomal pathway requires IB␣ phosphorylation by IKK. However, its role in PI-induced IB␣ degradation through the lysosomal pathway is not clear. Thus, we next evaluated whether IKK activation is required in PI-induced IB␣ degradation through the lysosomal pathway. To evaluate the effect of PI treatment on IKK activity, IKK activities were measured by immune complex kinase assays after PS-341 or MG132 stimulation. PS-341 and MG132 activated IKK (Fig. 4, A and B) up to 8 h. When IKK activity was suppressed by SC-514 (IKK␤ specific inhibitor) pretreatment, PS-341-induced IB␣ degradation and IL-8 production were partially blocked (Fig. 4C). Interestingly, phosphorylation of IB␣ was delayed (Fig. 4C), which suggests that IKK-independent mechanism(s) might be associated with PI-induced phosphorylation of IB␣. To further confirm if IKK activity is necessary for PI-induced IB␣ degradation, IKK activity was blocked by two different approaches: by insolubilizing IKK and by knockdown of IKK␤ using siRNAs. Neither IKK␣ nor IKK␤ expression was detected in soluble extracts up to 10 h of recovery after heat shock (HS) at 43°C for 1 h as previously reported (20). PS-341-induced IB␣ phosphorylation and degradation were delayed in HStreated cells (Fig. 5A), by which PS-341-or TNF-␣-induced IKK activity was completely blocked (Fig. 5, B and C). In accordance with this, knockdown of IKK␤ in cells transfected with IKK␤ siRNAs delayed PS-341-induced IB␣ phosphorylation and its subsequent degradation (Fig. 5, D and E). In contrast, TNF-␣-induced IB␣ degradation, which is known to be dependent on IKK activation, was suppressed by HS treatment or by knockdown of IKK␤ (Fig. 5, C and F). Taken together, these results imply that IKK activity is required for PI-induced IB␣ degradation during early phase, and an IKK-independent mechanism might regulate PI-induced IB␣ phosphorylation and degradation during late phase.
Inactivation of GSK-3␤ via the PI3K/Akt Pathway Is Related to the Acceleration of PI-induced IB␣ Degradation-The role of the PI3K/Akt pathway was evaluated in PI-induced IB␣ degradation. Active phosphorylated Akt was up-regulated after incubation with PIs for 4 h, and returned to baseline at 8 h or 24 h (Fig. 6, A and B). When PI-induced activation of Akt was blocked by treatment with PI3K/Akt pathway inhibitors (LY294002 or wortmanin) or overexpression of dominant-negative Akt (DN-Akt) plasmid vector, IB␣ degradation by PIs was delayed (Fig. 7, A and B).
Because GSK-3␤ is one of the well-known downstream molecules of Akt, we investigated whether PS-341-induced IB␣ degradation is mediated through GSK-3␤. The inactive phosphorylated GSK-3␤ at serine 9 (p-GSK-3␤) was increased by PIs time-dependently (Fig. 6, A and B). GSK-3␤ is known to be constitutively active and is inactivated by phosphorylation at serine 9. To confirm that increased p-GSK-3␤ is associated with decreased activity, GSK-3␤ activity was measured using an in vitro immune complex kinase assay after treating cells with PS-341. GSK-3␤ was constitutively active, and its activity was reduced by PS-341 treatment (Fig. 6C). When Akt activation was blocked with wortmanin, GSK-3␤ inactivation by PS-341 was markedly suppressed (Fig. 6D). These findings suggest that PS-341 inactivates GSK-3␤, and it was mediated by activation of the PI3K/Akt pathway.

DISCUSSION
NF-B activation is associated with the resistance of lung cancer cells to TNF-␣-induced apoptosis. Proteasome inhibition prevents NF-B activation and enhances TNF-␣-induced cell death (16). Moreover, inducible activation of NF-B suppresses the apoptotic response to irradiation and chemotherapy. Treatment with PS-341 blocks activation of NF-B induced by chemotherapy agents such as SN-38, the active metabolite of the topoisomerase I inhibitor, in human colorectal cancer cells (5) or gemcitabine, a nucleoside analog, in nonsmall cell lung cancer (NSCLC) cells (22). However, in this study, treatment with PS-341 or MG132 induced rather than inhibited NF-B activity and increased the expression of NF-B-regulated genes COX-2, cIAP2, IL-8, and XIAP. This result is consistent with other observations indicating that different proteasome inhibitors (epoxomicin or ALLN) induce NF-B nuclear translocation and transcriptional activity on endometrial carcinoma cell lines (23) and colon cancer cells (24). PS-341 treatment increased ubiquitinated proteins in whole cell lysate. The activity of the proteasome started to decrease 30 min after stimulation with PS-341 and was sustained up to 24 h. IB␣ is one of the selected proteins affected by inhibition of the ubiquitin-proteasome pathway. Proteasome inhibition elevates IB␣ levels and leads to inhibition of NF-B activity. However, we showed that PS-341-induced activation of NF-B was mediated by IB␣ phosphorylation and subsequent degradation. This suggests that other proteolytic systems, apart from the 26S proteasome, might be involved in PIinduced IB␣ degradation.
There are several circumstances in which participation of other protein degradation systems have been described. The calcium-activated calpain system (25-26), caspases (27), lyso- somes (28), and unknown proteinases have been suggested to be responsible for IB degradation. In this study, lysosomal inhibitor (chloroquine or NH 4 Cl) and cathepsin inhibitors (Z-FF-FMK or Z-FA-FMK) suppressed PI-induced IB␣ degradation, but did not affect TNF-␣-mediated IB␣ degradation in NCI-H157 cells and lipopolysaccharide (LPS) or CpG-oligodeooxynucleotide (CpG-ODN)-induced IB␣ degradation in macrophage cell line (data not shown). TNF-␣-induced IB␣ degradation in NCI-H157 cells or LPS (or CpG-ODN)-induced IB␣ degradation in macrophages were blocked by short-term incubation (1 h) with PIs (data not shown), which supports that PIs effectively block proteasome activity, and prolonged incubation with PIs degrades IB␣ through the non-proteasomal pathway. Blocking of autophagy activation by knockdown of LC3B expression suppressed PI-induced IB␣ degradation, which confirm that PIs induce IB␣ degradation via the autophagylysosomal pathway. A previous report shows that a portion of intracellular IB␣ is located in the lysosome as well as in the cytosol and microsomes and that the transport of IB into lysosome is selective (28). In the report, ubiquitination and phosphorylation of IB␣ is not required for its targeting to lysosome under conditions of nutrient deprivation. In our study, overex-pression of IB-SR suppressed PI-induced IB␣ degradation and subsequently inhibited NF-B activation, which suggests that phosphorylation of IB␣ is necessary for PI-induced IB␣ degradation. However, to evaluate if translocation to the lysosome needs ubiquitination of IB␣ in PI-treated cells, further detailed study is required. Moreover, a recent report suggests that PS-341 induces caspase-independent, leupeptin-sensitive protease-independent, but calpain-dependent IB␣ proteolysis (29). IB degradation liberates NF-B for nuclear translocation, where it drives transcription of the downstream genes, including cIAP, Bcl-2, Bcl-X L , COX-2, XIAP, and others (30). NF-B activates cIAP1 and cIAP2 to inhibit TNF-␣-induced apoptosis by blocking caspase-8 activity. COX-2 has been known to promote angiogenesis and is found up-regulated in some cancer cells. In agreement with this, we showed that IB␣-SR gene transfer blocked PS-341-induced increase in COX-2 expression. Pretreatment with DHMEQ, which inhibits DNA binding affinity of NF-B and knockdown of p65 using siRNAs, suppressed COX-2 induction by PS-341 (data not shown). These results suggest that PIs activate the NF-B cascade via lysosomal degradation of IB␣, resulting in induction of anti-apoptotic genes, such as cIAP2 and COX-2 in lung cancer cells. . IKK activity is required for the rapid PI-induced IB␣ degradation (2). A-C, NCI-H157 cells were exposed to HS at 43°C for 1 h and then stimulated with PS-341 (50 nM) for the indicated times or TNF-␣ (10 ng/ml) for 5 min. D-F, cells were transfected with siRNAs targeting IKK␤ gene and control siRNAs. Forty-eight hours after transfection, cells were treated with PS-341 (50 nM) for the indicated times and TNF-␣ (10 ng/ml) for 30 min. IKK assays were performed as described under "Experimental Procedures." Whole cell lysates were immunoblotted with antibodies against IB␣, p-IB␣, IKK␣, IKK␤, and actin. Results are representative of three separate experiments.
The first step of IB␣ degradation involves phosphorylation of IB␣ by the IB kinase (IKK) complex. The IKK complex is composed of several kinases including IKK␣, IKK␤, and IKK␥, and it requires phosphorylation by NIK to become activated. Our previous study showed that heat stress (HS) insolubilizes IKKs, resulting in the loss of IKK activity (20). As expected, PS-341-or TNF-␣-induced IKK activity was completely suppressed by HS, and IKK␣/␤ levels were reduced in soluble extracts after HS. However, phosphorylation and degradation of IB␣ by PIs were delayed by HS as well as by knock-down of IKK␤. These findings suggest the possibility of other IKK-independent pathway(s) involvement regarding PI-induced degradation of IB␣. Similarly, a previous study showed that IB␣ degradation in response to anti-cancer reagents such as doxorubicin was mediated by an IKK-independent mechanism (31).
To determine the responsible factor(s), we evaluated the involvement of signaling mediators activated by PIs. Our data show that simultaneous activation of the PI3K/Akt pathway was involved in PS-341-induced IB␣ degradation. Treatment with PI3K/Akt inhibitor, LY294002 or wortmannin, and overexpression of dominant negative Akt (DN-Akt) delayed IB␣ degradation by PIs. Moreover, inactivation of GSK-3␤, one of the well-known downstream effectors of Akt, mediated the rapid degradation of IB␣ by PIs. Akt-mediated inactivation of GSK-3␤ has been involved in many other signaling pathways including the Wnt/␤-catenin pathway (32). In addition, the inhibitory effect of active GSK-3␤ on NF-B-dependent transcription has been reported to be related to prevention of IKK activity, which occurs as a result of competitive binding of GSK-3␤ to NF-B essential modifier (NEMO) (33). Inactiva-tion of GSK-3␤, however, has been suggested to prevent TNF-␣-induced IB␣ degradation, but had no effect on the IKK␤ activation (34). In our study, blockade of GSK-3␤ inactivation by overexpression of constitutively active GSK-3␤ (S9A) and knock-down of GSK-3␤, delayed PI-induced IB␣ degradation, which implies that inactive GSK-3␤ plays a positive role on PI-induced IB␣ degradation. IKK activity, phosphorylation of p65, or NF-B DNA binding activity are putative target steps of GSK-3␤ action on the NF-B pathway (21,35,36). In our study, PI-induced IB␣ phosphorylation was reduced in cells with decreased levels of GSK-3␤. However, IKK activity was not affected by knock-down of GSK-3␤ (data not shown).
To the best of our knowledge, this is the first study to find that PIs activate NF-B, which is mediated by IB␣ degradation via the autophagy-lysosomal pathway. The IKK-dependent mechanism in the earlier stage and IKK-independent mechanism in the late stage are required for PI-induced IB␣ degradation. Moreover, inactive GSK-3␤ as well as active IKK mediates PI-induced IB␣ degradation in lung cancer cells. Thus, prevention of IB␣ degradation by targeting its upstream reg- Total cellular extracts were subjected to Western blot analysis for p-Akt, Akt, p-GSK-3␤, and GSK-3. C, NCI-H157 cell were treated with PS-341 (10 -50 nM) for 24 h. GSK-3␤ was immunoprecipitated with an anti-GSK-3␤ antibody, and GSK-3␤ activity assays were performed as described under "Experimental Procedures." D, NCI-H157 cells were pretreated with wortmanin (WT, 50 nM) for 2 h followed by PS-341 (50 nM) for the indicated times. Cell lysates were subjected to Western blot analysis for p-GSK-3␤ and actin. Results are representative of three separate experiments. ulators or the responsible proteolysis pathway will augment anti-tumor activity of PIs.