E1A Inhibition of Radiation-induced NF-κB Activity through Suppression of IKK Activity and IκB Degradation, Independent of Akt Activation¹

IR³ therapy is one of the most common cancer treatments. However, the therapeutic efficacy of this method decreases when cancer cells develop resistance to radiation. Ad5 E1A expression has been shown to sensitize cells to IR, anticancer agents, TNF-induced apoptosis (1, 2, 3, 4, 5, 6). Ad5 E1A was also shown to function as a tumor suppressor, suppressing transformation, tumorigenicity, and metastasis (7, 8, 9). Induction of apoptosis may be one of the important functions of E1A tumor suppression (8, 9, 10). However, how E1A mediates inhibition of radiation-induced NF-κB activity is not clear.

NF-κB has been shown to play a critical role in blocking apoptosis induced by a variety of stimuli, including TNF, chemotherapeutic compounds, and γ-radiation (1, 6, 11). The most common form of NF-κB is a heterodimer of p65 and p50 subunits (12). In resting cells, NF-κB is sequestered in a latent form in the cytoplasm through association with IκBs, the endogenous inhibitors of NF-κB. In the usual pathway of NF-κB activation, IκBα is phosphorylated at Ser32 and Ser36 by the IKK complex, which consists of the IKKα, IKKβ, and IKKγ/NF-κB essential modulator subunits (12). The phosphorylated IκBα is additionally ubiquinated and degraded, resulting in NF-κB translocation into the nucleus to activate target genes. Some of these genes products, such as c-IAP1, c-IAP2, TRAF1, TRAF2, and Bf1/A1 have been shown to exhibit antiapoptotic functions (13, 14).

Recently, activation of the growth-factor-regulated serine/threonine protein kinase Akt (also known as protein kinase B) was shown to provide a survival signal that protects cells from apoptosis induced by various stresses (14). Activation of Akt was also shown to provide a critical cell survival signal required for tumor progression (15). Akt is downstream of PI3k. Activation of PI3k results in increase of the 3′-phosphorylated phosphatidylinositides P(3, 4, 5)P3 and P(3,4)P2. These phosphatidylinositides bind to the pleckstrin homology domain of Akt and induce translocation to the plasma membrane, where Akt is phosphorylated on Thr308 and Ser437 by phosphoinoside-dependent kinase 1 and phosphoinoside-dependent kinase 1/protein kinase C-related protein kinase-2 complex (16). The antiapoptotic effects of activated Akt include the phosphorylation of Bad, caspase 9, forkhead transcription factors, and IKKα (14, 17). The involvement of Akt activity in TNF-induced NF-κB activation is controversial (17, 18). It is not yet clear whether Akt might be involved in radiation-induced NF-κB activity.

In this study, we examined how E1A suppresses radiation-induced NF-κB activity, including IKK activity and IκBα degradation. In addition, we examined the effect of E1A on Akt phosphorylation and the effect of IR on Akt phosphorylation and its relationship with NF-κB activity.

The parental human ovarian carcinoma cell line SKOV3.ip1, ip1-E1A2, the Ad5 E1A transfectant of SKOV3.ip1, and ip1-Efs, a stable transfectant of SKOV3.ip1 cells carrying a 2-bp frameshift deletion in the E1A coding region, were established and cultured as described previously (1). dA17/i and dA31/i, which are dominant-negative mutant Akt (dn-Akt, K179M) transfectants of SKOV3.ip1, were established by cotransfection of the SKOV3.ip1 cells with a HA-tagged dn-Akt expression vector and pcDNA3. The culture conditions for the dn-Akt cells were the same as for the other cells, except that the medium was supplemented with 500 μg/ml G418. One week before the experiment, the G418-containing medium was replaced with the regular medium. Cells were irradiated with a ¹³⁷Cs source emitting a fixed dose of irradiation (Irradiator model 0103; United States Nuclear Corporation).

Total cell lysates were subjected to 10% SDS-PAGE and blotted onto nitrocellulose membranes. Polyclonal antibodies against NF-κB (p65), IKKβ, and IKKα (Santa Cruz Biotechnology, Santa Cruz, CA), and Akt or p-Akt on Ser437 (Cell Signaling, Beverly, MA) were used.

The immunocomplex kinase assay was performed as described previously (6). Briefly, ip1-E1A2, ip1-Efs and SKOV3.ip1 cells were treated with IR (5 Gy) or not irradiated. The cell lysates were incubated with anti-IKKβ antibody at 4°C overnight. The immunoprecipitates were collected for the kinase assay with a glutathione S-transferase-IκBα substrate.

The electrophoretic mobility shift assays were performed as described previously (1). Briefly, the dn-Akt transfectants and SKOV3.ip1 cells were exposed to 5 Gy of IR, or SKOV3.ip1 cells were treated with 30 ng/ml of TNF (R & D Systems Inc., Minneapolis, MN) for 30 min. Nuclear extract (5 μg of protein) was incubated with 1 μg of poly(dI-dC; Pharmacia, Piscataway, NJ) on ice for 20 min, and a ³²P-labeled double-stranded oligonucleotide containing the κB site from the human immunodeficiency virus was added. The probe was allowed to bind at room temperature for 20 min. The resulting complexes were resolved by electrophoresis on a 4% nondenaturing polyacrylamide gel. The mutant κB competitor was a double-stranded oligonucleotide containing mutations in the κB site (6).

NF-κB activation requires degradation of IκBs to free NF-κB from the heterotrimeric IκB/NF-κB complex and to translocate it into the nucleus. IκBα is the major endogenous inhibitor of NF-κB activation. Because IR treatment has been shown to induce IKK activation and IκBα degradation, and E1A sensitizes cells to IR-induced apoptosis by inhibiting IR-induced NF-κB activity (1, 19), we asked whether suppression of IR-induced NF-κB by E1A may go through the IKK/IκBα pathway. To test whether E1A can inhibit IR-induced IκBα degradation, we examined the IκBα levels in parental SKOV3.ip1 cells, E1A transfectants, ip1-E1A2, and E1A frameshift mutant transfectants, ip1-Efs cells treated with IR. IR treatment reduced IκBα protein level in SKOV3.ip1 and ip1-Efs cells. However, in E1A-expressing ip1-E1A2 cells, IR did not change the IκBα protein level. Furthermore, the base level of IκBα is much higher in ip1-E1A2 cells. Similar results were also observed with the other sublines of E1A-transfected human ovarian cancer cells, SiE3 and SiE17 (data not shown). There was also more NF-κB protein in the ip1-E1A2 cells than in the control cells (Fig. 1,A). This change is functionally insignificant, because NF-κB is known to be inactive in ip1-E1A2 cells (1). This could be attributable to stabilization of NF-κB by increased IκBα. We also used cycloheximide to block de novo protein synthesis in the cells during IR treatment. Under those conditions, we found that E1A prolonged the half life of IκBα (Fig. 1,B). The graph in Fig. 1 B shows the quantification of IκBα levels after irradiation. More than 80% of the IκBα protein in ip1-Efs and SKOV3.ip1 cells was degraded in 1.5 h after irradiation. However, <20% of the IκBα protein in the ip1-E1A2 E1A transfectant cells was degraded at the same time. The results indicated that E1A inhibited IR-induced IκBα degradation.

Whereas IKK activity has been shown to be necessary for IκB degradation and NF-κB activation induced by IR (19), X-ray-induced NF-κB activity was reported not to require IκBα degradation in rat astrocytes and human brain tumor cells (20). To evaluate the possible inhibitory mechanism of E1A on IκBα degradation, we measured the IKK protein levels and activity. Treatment of E1A-transfected and parental cells with IR had no effect on the protein level of IKKβ (Fig. 2,A) and IKKα (data not shown). However, IR treatment increased IKKβ activity in E1A frameshift mutant transfected and parental cells but to a much lesser degree in E1A transfectants (Fig. 2 B), indicating that E1A inhibited IR-induced IKKβ activity. The TNF-treated parental cells were used as a positive control.

Recent studies have shown that the involvement of Akt in the NF-κB pathway is a cell-specific activity (19). To test whether E1A mediates changes in Akt and, thus, is involved in IR-induced activation of NF-κB, we compared the phosphorylation of Akt in E1A transfectants and parental cells with or without IR treatment. The results indicated that E1A inhibited basal Akt phosphorylation (Fig. 3,A). The data shown in Fig. 3,A have been repeated a few times to confirm down-regulation of p-Akt by E1A (data not shown). Although the Akt level seems to be elevated slightly in the E1A transfectants shown in Fig. 3,A, we did not observe the change in Akt protein level in other experiments. Thus, E1A inhibits phosphorylation of Akt but does not significantly affect Akt expression. Fig. 3,B shows that IR induced Akt phosphorylation and that E1A did not inhibit IR-induced Akt phosphorylation. Because E1A inhibits IR-induced NF-κB activity (1) but not IR-induced Akt phosphorylation, it is conceivable that separate pathways may exist for IR-induced NF-κB and IR-induced Akt activation. To additionally investigate this phenomenon, we established cell lines stably transfected with dn-Akt and examined the IR-induced NF-κB DNA binding activity of those cells. The result indicates that IR-induced NF-κB activity was not blocked in the transfected cells (Fig. 4,A), suggesting that Akt activation is not involved in IR induction of NF-κB activity. A similar result was observed with dn-Akt stable transfectants of the human breast cancer cell line MDA-MB-453 (data not shown). The bottom panel of Fig. 4,A shows the expression of dn-Akt in the cell lines. To additionally confirm these observations, we treated the cells with the PI3k inhibitor wortmannin and IR. The results indicated that wortmannin could not block IR-induced NF-κB DNA binding activity (data not shown). In addition to the DNA binding activity of NF-κB, we also confirmed that dn-Akt could not block IR-induced NF-κB activity as measured by a NF-κB-mediated promoter activity experiment (Fig. 4,B). IR treatment increased luciferase activity in cells transfected with either dn-Akt plus κB-luciferase or empty vector plus κB-luciferase construct, and the IR-induced luciferase activity was not significantly different in dn-Akt-transfected cells and vector-transfected cells (Fig. 4 B). Taken together, our results suggest that inhibition of IKK activity and IκBα degradation is involved in E1A-mediated inhibition of IR-induced NF-κB activity and that IR induces Akt phosphorylation, which is not required for IR-induced NF-κB activity.

We thank Dr. Mickey C-T. Hu and Zhenming Yu for technical discussion, and Dr. Dipak K. Giri for reading the manuscript.