Inhibition of Cell Cycle Progression by the Novel Cyclophilin Ligand Sanglifehrin A Is Mediated through the NF (cid:1) B-dependent Activation of p53*

Sanglifehrin A belongs to a novel family of immu-nophilin-binding ligands. Sanglifehrin A is similar to cyclosporin A in that it binds to cyclophilins. Unlike cyclosporin A, however, the cyclophilin-sanglifehrin A complex has no effect on the calcium-dependent protein phosphatase calcineurin. It has been previously shown that sanglifehrin A specifically blocks T cell proliferation in response to interleukin 2 by inhibiting the appearance of cell cycle kinase activity cyclinE-Cdk2. How sanglifehrin A treatment leads to the cell cycle blockade has remained unknown. We report that sanglifehrin A is capable of activating the tumor suppressor gene p53 at the transcription level, leading to up-regulation of p21 that then binds and inhibits the cylcinE-Cdk2 complex. Further analysis of different elements in the p53 promoter showed that sanglifehrin A activates p53 transcription primarily through the activation of the transcription factor NF (cid:1) B by activating I (cid:1) B kinase in a manner that is similar to several genotoxic agents. Unlike other genotoxic drugs, sanglifehrin A does not cause DNA damage, making it a unique natural product that is capable of activating the NF (cid:1) B signaling pathway without affecting DNA.

The immunosuppressive drugs cyclosporin A, FK506, and rapamycin constitute a unique family of natural products that work by an unusual mechanism (1)(2)(3)(4)(5)(6)(7)(8). They serve as natural dimerizers that bring together two proteins, suppressing the function of both proteins as a consequence. Thus, CsA 1 is known to bind to the cyclophilin family of proteins (9) while FK506 and rapamycin are known to bind to the FKBP family of proteins (10,11). Each of the immunophilin-drug complexes specifically interacts with and inhibits the function of their ultimate target. The CsA-cyclophilin and the FKBP-FK506 complexes specifically inhibit the protein phosphatase calcineurin (12,13) while the FKBP-rapamycin complex specifically inhibits the function of the protein known as FRAP/RAFT/ TOR (14 -18). Inhibition of the phosphatase activity of calcineurin prevents the dephosphorylation of a critical tran-scription factor, NFAT, thereby blocking the transcription of a number of cytokine genes (5,8). The binding of FKBP-rapamycin to RRAP/RAFT/TOR interferes with the function of TOR, leading to a blockade of cell cycle at the G 1 phase of the cell cycle.
Sanglifehrin A (SFA) is a new member of the immunophilin ligand superfamily. It was discovered through a screen for novel cyclophilin ligands that block T cell activation (19,20). Similar to CsA, SFA binds to cyclophilin with high affinity. Unlike CsA, however, the cyclophilin-SFA complex has no effect on the phosphatase activity of calcineurin (21). Although SFA was shown to inhibit mouse and human mixed lymphocyte reactions (19), we and others have recently found that SFA does not affect T cell receptor-mediated signal transduction pathways leading to the production of cytokines such as IL-2 (21,22). Instead, SFA inhibits IL-2-dependent T cell proliferation, similar to rapamycin (23,24). Moreover, SFA blocks the cell cycle progression of T lymphocyte in response to IL-2 at the G 1 phase of the cell cycle, an effect also exhibited by rapamycin (22). Unlike rapamycin, however, the activation of the p70 s6k activity was unaffected by SFA (22). These results indicate that SFA has a novel mechanism of action that is distinct from the other known immunophilin ligands, CsA, FK506, and rapamycin.
Further studies revealed that SFA inhibited the hyperphosphorylation of Rb and abrogated the appearance of the G 1 cyclin-dependent kinase cyclinE-Cdk2 upon IL-2 stimulation (22). However, how SFA inhibits cell cycle progression has remained unknown. We report here that the cell cycle effect of SFA is not specific to T cells. Using the tumor cell line HCT 116 as a model system, we show that SFA stimulates the transcription of p53 and consequently p21, leading to a significant increase of p21 that is likely to be responsible for the G 1 cell cycle blockade. The role of p53 as a mediator of SFA is underscored by the demonstration that p53 null mouse embryo fibroblasts gain significant, albeit not complete, resistance to SFA. Furthermore, we demonstrate that the NFB binding site in the p53 promoter is required for the SFA-induced transcriptional activation of p53. SFA was found to activate NFB through the activation of the IB kinase activity. Analysis of the integrity of the genomic DNA ruled out the possibility that SFA activates NFB via DNA damage, making SFA a unique inducer of NFB and p53.
Plasmids-Human p53 luciferase reporter comprised of the 2.4-kilobase or 356-bp p53 promoter region were generated by excising the 2.4-kilobase XbaI fragment and the 356-bp XbaI-BamHI fragments of p53 promoter from pICAT vector and subcloning each into the SmaI site of pGL basic luciferase reporter vector from Promega after making each end blunt. pG13pyLuc was provided by Dr. Lazano (32). Mutations of the B or HoxA5 binding sites in the 356-bp p53 reporter construct were generated by polymerase chain reaction-mediated mutagenesis and confirmed by sequencing. NFB luciferase reporter construct was previously described (22). Expression vectors for mutant IB␣, dominant negative IKK␣, and GST-IB␣ were kindly provided by Dr. Warner Greene (Univ. of California San Francisco).
Cell Culture-Wild type and knockout mouse embryos fibroblasts (MEFs) were prepared from day 13.5 mouse embryos as previously described (26). Human colon cancer cell line HCT 116 (p53ϩ/ϩp73ϩ/ϩ, ATCC CCL247) was maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum, 2 mM glutamine and penicillin/streptomycin (50 IU/ml and 50 g/ml, respectively), and 6 mM HEPES. The HCT 116 cell line was purchased from American Type Culture Collection (Manassas, VA) and grown in a humidified incubator at 37°C in 5% CO 2 .
Fluorescence-activated Cell Sorter Analysis-Following treatment of cells with test agents, the 200 ϫ g cell pellet was fixed in 2 ml of cold absolute ethanol at 4°C for 1 h and then washed twice with cold PBS. The cells were resuspended in 1.76 ml of PBS to which 200 l of RNase (1 mg/ml in PBS) and 40 l of propidium iodide (2.5 mg/ml in PBS) were added. The cell suspension was incubated in the dark for 15 min and kept at 4°C until analyses. The propidium iodide fluorescence of individual nuclei was determined using a Becton-Dickson FACScan (emission at 675 nm with excitation at 488 nm). Cell cycle distribution was analyzed with CellQuest TM v3.1 acquisition software and the ModFit LT v2.0 program.
The protein concentration in each sample was determined with Bradford assay. Equal amounts of protein were denatured by heating to 95°C in Laemmli's sample buffer and were resolved by 12% SDSpolyacrylamide gel electrophoresis, followed by transfer to polyvinylidene difluoride membranes. The membranes were probed with indicated antibodies and detected using an ECL system per manufacturer's instruction.
Immunoprecipitation-Kinase Assay-Whole cell lysate (200 g of protein in 0.5 ml lysis buffer) was mixed with 4 g of polyclonal rabbit anti-IKK␣ antibodies (H-744) and incubated at 4°C on a rotator for 4 h. Protein G/A-Sepharose beads (30 l) were added and the incubation was continued for an additional 1 h. Immunoprecipitates were isolated by centrifugation and washed twice with 500 l of ice-cold lysis buffer and once with 500 l of ice-cold kinase buffer (25 mM Tris-HCl, pH 7.5, 5 mM ␤-glycerophosphate, 2 mM dithiothreitol, 0.1 mM Na 3 VO 4 , 10 mM MgCl 2 ). Pellets were then mixed with 50 l of kinase buffer containing 1 g of recombinant glutathione-S-transferase (GST)-IB␣ (1-62), 50 M cold ATP, and 5 Ci [␥-32 P]ATP (specific activity 3,000 Ci/mmol) on ice, followed by incubation at 30°C for 15 min. The reaction was terminated by the addition of 25 l of 3 ϫ Laemmli's sample buffer and boiling in water for 5 min. The 32 P-labeled GST-IB␣ was resolved by 12% SDS-polyacrylamide gel electrophoresis and detected by autoradiography.
Reverse Transcription-Polymerase Chain Reaction (RT-PCR)-Ti-tan TM One Tube RT-PCR System was used to detect p53 and p21 mRNA. Total RNA (1 g) from control and SFA-treated cells was reversed transcribed to synthesize the first-strand cDNA. Amplification was performed using primers specific for human p53, p21, or actin. PCR conditions were 94°C for 30 s (denaturation), 55°C for 30 s (annealing), and 68°C for 45 s (elongation) for a total of 30 cycles. PCR products were analyzed on 1.8% agarose gel and visualized by ethidium bromide staining. PCR reactions with either the reverse transcriptase or the template left out served as negative controls.
Reporter Plasmid Assay-For DNA transfection, HCT 116 cells were plated in 6-well plates (2 ϫ 10 5 /well/2 ml Dulbecco's modified Eagle's medium) and incubated until the cells were 50 -80% confluent. The cells were transfected with 1-2 g of various luciferase reporter plasmids and 1 g of a pSV-␤-gal control vector by LipofectAMINE reagent (Life Technologies, Inc.). After transfection, the cells were incubated in 10% fetal calf serum/Dulbecco's modified Eagle's medium overnight. Cells were then treated in the presence or absence of SFA for 12 h. After SFA treatment, cells were harvested and lysed for luciferase assay according to manufacturer's instructions (Promega Co., Madison, WI).
Proliferation Assays-A total of 2 ϫ 10 4 MEF cells were seeded in triplicate to 96-well microtiter plates. Varying concentrations of SFA were added to each well to a final volume of 200 l of medium. Following incubation at 37°C in a humidified atmosphere with 5% CO 2 for 20 h, [ 3 H]thymidine (1 Ci/well) was added, and the incubation was continued for an additional 4 h. Cells were then harvested onto glass fiber mats using a 96-well harvester (TomTech, Orange, CT). The mats were dried and sealed in pouches with scintillation mixture (EcoScint, ICN, Costa Mess, CA), and radioactivity was determined in a Wallace 1205 Betaplate liquid scintillation counter (Gaithersburg, MD). The median inhibition concentration (IC 50 ) was defined as the drug concentration required to inhibit proliferative responses of cells by 50% and calculated from linear regression analysis.
Determination of Cellular DNA Damage-To determine the extent of DNA damage in cells, alkaline micro-gel electrophoresis or the "Comet" assay was performed under dim light as described previously with slight modifications (27,28). Upon incubation with drugs, cells were washed and suspended in 1% agarose (low melting point) in PBS. Then 75 l of this mixture was pipetted onto a microscope slide that had been pre-coated with 75 l of 1% agarose (normal melting point). Without delay, a glass coverslip was placed on top of the slide, and the agarose/ cell mixture was allowed to congeal at 4°C. Upon removal of the coverslip, the slides were immersed in ice-cold lysis solution (10 mM Tris-HCl, 2.5 M NaCl, 0.1 M EDTA, 10% Me 2 SO, and 1% Triton X-100) for 1 h. Slides were then placed in an electrophoresis unit containing 0.3 M NaOH and 1 mM EDTA, pH 13, for 40 min before electrophoresis at 20 V (300 mA) for 20 min. Following electrophoresis, the slides were immersed in neutralizing buffer (0.4 M Tris-HCl, pH 7.5) before the DNA was stained with 4Ј,6-diamidino-2-phenylindole for observation with fluorescence microscopy.

RESULTS
Cell Cycle Arrest at G 1 -S Phase Is General-We have previously shown that SFA inhibits IL-2-dependent T cell proliferation at the G 1 phase of the cell cycle (22). To determine whether SFA is specific for T lymphocyte, we examined several transformed cell lines. As shown in Fig. 1, the human colon cancer cell line HCT 116 is sensitive to SFA. Like CTLL2, the cell cycle progression of HCT 116 is also blocked in the G 1 phase of the cell cycle. To simplify subsequent studies of the cell cycle effect of SFA, we decided to use HCT 116 as a model system.
SFA Induces p53 mRNA Synthesis-We have shown that SFA blocks the appearance of growth factor-dependent cyclinE-Cdk2 activity, likely explaining the G 1 effect of SFA (22). This inhibition is not due to down-regulation of either CyclinE or Cdk2 proteins as determined by Western blot analysis (22). When we examined the expression of the p21 family of CyclinE-Cdk2 inhibitor proteins, it was found that SFA treatment of HCT 116 cells led to a significant increase in the level of p21 protein expression (Fig. 2A). When p21 mRNA was examined, it was also induced dramatically by SFA (Fig. 3A). As p21 transcription is known to be regulated by p53, we checked whether p53 expression is affected by SFA.
We first examined the effect of SFA on p53 at the protein level. Indeed, p53 protein is significantly induced by SFA ( Fig.  2A). Given that p53 is known to be regulated at multiple levels, including protein stability and mRNA synthesis, we deter-mined whether SFA stabilized p53 protein. We treated HCT 116 cells with SFA for 3 h to induce p53. Cycloheximide was then added into the cell culture to block de novo protein synthesis for an addition 1-8 h. The SFA-induced p53 was degraded over time upon treatment with cycloheximide even in the continued presence of SFA (Fig. 2B). This result clearly indicated that the increase in p53 in the presence of SFA is not due to stabilization of p53 protein. When p53 mRNA synthesis was examined by RT-PCR, it was found to be significantly induced by SFA (Fig. 3A), indicating that SFA activates p53 primarily at the mRNA level either by induction of its transcription or by stabilization of its mRNA or both. SFA Activates p53 Promoter-To determine whether SFA affected p53 transcription, we examined its effect on a luciferase reporter gene under the control of the p53 promoter. Indeed, SFA activated the p53-luciferase reporter gene in a dosedependent manner (Fig. 3B). As a control, we included a p21 reporter, pG13py.luc, which was activated by SFA as expected. In addition to the full-length p53 reporter gene, we also examined another reporter gene containing a 356-bp fragment of the p53 promoter with several important transcription factor binding sites responsive to DNA-damaging agents (29 -31). Similar to the full-length p53 reporter, the minimal p53 reporter, p53Luc (356 bp), is induced by SFA in a dose-dependent manner (Fig. 3B), suggesting that the most important SFA-responsive elements lie within this 356-bp fragment.
NFB Is Required for the Activation of p53 Transcription by SFA-Several transcription factor binding sites within this 356-bp fragment have been previously shown to be important for p53 transcription, including a HoxA5 and an NFB binding site (29,32). To determine whether either of the transcription factor binding sites is important for the activation of the p53 FIG. 1. SFA blocks HCT116 cell cycle progression at G 1 -to-S transition. HCT116 cells were starved in 0.1% serum for 2 days before they were stimulated with 10% serum in the presence or absence of 500 nM SFA for an additional 12 h. Cells were harvested and subjected to FACScan analysis of DNA content upon staining with propidium iodide. The numbers of cells in the G 0 /G 1 , S, G 2 , and M phases of the cell cycle (expressed as a percentage of the total cell population in each case) are indicated. FIG. 2. SFA induces p53 and p21 proteins. A, HCT116 cells were treated with 500 nM SFA for various times as indicated under low serum conditions (2% fetal calf serum). Cell lysates were prepared, subjected to SDS-polyacrylamide gel electrophoresis followed by Western blot analysis using anti-p53 monoclonal antibody (Pab1801), anti-p21 antibody (N-20), or anti-actin antibody. B, HCT116 cells were treated with 500 nM SFA for 0, 1, 3, or 6 h. For a subset of cell cultures, treatment with cycloheximide (CHX, 25 g/ml) ensued for an additional 1-8 h upon incubation with SFA for 3 h. Cells were lysed immediately, and cell lysates were subjected to Western blot analysis.
promoter by SFA, we mutated both sites and constructed the corresponding mutant reporter genes (Fig. 4A). Mutation of the HoxA5 binding site had negligible effect on the response to SFAmediated activation (Fig. 4B). In contrast, mutation of the NFB binding site completely abolished the response of the reporter gene to SFA. These data strongly suggest that the NFB binding element in the p53 promoter is essential for the activation of p53 transcription by SFA.
The signal transduction pathways leading to NFB activation have been largely elucidated. A number of stimuli have been shown to converge on the upstream kinases, IKKs, to activate NFB. The IKKs phosphorylate the cytoplasmic-anchoring protein IB, which leads to the degradation of IB and nuclear translocation of NFB. To determine which step in the upstream NFB activation pathway SFA affects, we repeated the p53 reporter gene assay in the presence of a dominant negative IKK␣ and a mutant IB␣ that is incapable of being phosphorylated by IKKs. We thus cotransfected HCT 116 cells with the p53Luc reporter gene along with the dominant negative IKK␣ (dnIKK␣) and the mutant IB␣ (mIB␣) and examined the activation of the p53Luc reporter gene by SFA. Both mIB␣ and dnIKK␣ abolished the activation of the p53Luc reporter by SFA, suggesting that SFA activated NFB at a step upstream of IKKs (Fig. 4C).
To confirm that SFA is capable of activating NFB, we examined the effect of SFA on an NFB luciferase reporter gene comprising of multimerized NFB binding sites. SFA stimulated the NFB reporter by about 4-fold, consistent with its stimulation of the p53Luc reporter (Fig. 5A). Like the activation of the p53Luc reporter gene, the activation of the NFB reporter gene by SFA is also inhibited by both dominant negative IKK␣ and the mutant form of IB␣ (Fig. 5A). We next determined whether SFA induced IB degradation and IKK activation. Of the two isoforms of IB examined, SFA induced IB␣ degradation in a dose-dependent manner, while IB␤ appeared to be less responsive, with appreciable degradation observed only at the highest dosage of SFA used (Fig. 5B). When we checked the IKK activity by immunoprecipitationkinase assay, SFA was found to activate IKK activity in a dose-dependent manner (Fig. 5C). At the highest dose used, SFA induced a similar amount of IKK activity as 20 ng/ml of TNF␣. It can thus be concluded that SFA activates NFB at a step upstream of IKK.
P53 Is an Important Mediator of Inhibition of Cell Proliferation by SFA-We have shown that SFA induces p53 transcription, which appeared to be sufficient to cause cell cycle arrest at the G 1 phase through the induction of p21. It is not clear, however, whether p53 is in fact necessary for the cell cycle blockade caused by SFA. To address this question, we took advantage of the availability of MEFs that are deficient in p53. Similar to T cells and HCT 116 cells, the proliferation of wild type MEFs is sensitive to SFA (Fig. 6). In contrast, p53 Ϫ/Ϫ MEFs exhibited significant resistance to SFA. It is noteworthy, however, that the resistance of p53 null MEFs to SFA is not complete, indicating that there exist proteins that play a redundant role with p53.
The Effects of SFA Is Not Mediated through DNA Damage-The ability of SFA to activate p53 via the NFB signaling pathway is reminiscent of several cytotoxic drugs or carcinogens that act by causing DNA damage (31,33). For example, it has been previously reported that the anticancer drug daunomycin exhibited similar effects on cells through activation of NFB and p53 (33). We have extended those observations by demonstrating that activation of p53 and NFB reporter genes by the DNA damaging agent doxorubicin is sensitive to inhibition by both dominant negative IKK␣ and mutant IB␣, similar to SFA (Figs. 4C and 5A). These observations raised the possibility that SFA may manifest its cell cycle effect through DNA damage. To test this possibility, we turned to the so-called Comet assay that allows for quantitative determination of DNA damage at individual cell level (27,28). As expected, doxorubicin caused extensive DNA damage to HCT 116 cells (Fig. 7). However, SFA had little effect on the integrity of chromosomal DNA at concentrations that are sufficient to cause cell cycle arrest. These results clearly distinguish SFA from other known cytotoxic DNA damaging drugs and indicate that SFA is a unique agent that interacts with immunophilins on the one hand and activates NFB and p53 on the other without causing DNA damage. DISCUSSION SFA was discovered as a novel immunosuppressive agent (19,20). It was recently shown that SFA blocks IL-2-dependent T cell proliferation at the G 1 phase of the cell cycle, similar to FIG. 3. SFA induces p53 and p21 at the level of transcription. A, p53 and p21 mRNA levels in HCT116 cells were determined by RT-PCR after cells were treated with SFA at indicated concentrations for 3 h. B, HCT116 cells were transiently transfected with one of the following luciferase reporter constructs (1 g): pG13.pyLuc, p53Luc (2.4 kilobase) or p53Luc (356 bp). Transfected cells were allowed to recover for 24 h. They were then incubated in the presence or absence of SFA for 12 h before they were harvested for the determination of luciferase activity. the immunosuppressive drug rapamycin (21,22,24,34). The site of action of SFA, however, is different from that of rapamycin, as it has no effect on p70 s6k activity (21,22). In this study, we found that the cell cycle effect of SFA is not confined to lymphocytes; it is equally effective at inhibiting the proliferation of several cancer cell lines, including HCT 116. Using HCT 116 as a model system, we have uncovered a pathway that appears to mediate the cell cycle effect of SFA. This pathway includes the IB kinases that phosphorylate IB, leading to activation of NFB. The activated NFB then induces the transcriptional activation of p53, which in turn activates the transcription of p21. The signaling cascade from IB kinase activation to p21 expression accounts for the inhibition of the critical G 1 kinase CyclinE-Cdk2 (22). These findings shed new light on the mechanism of action of SFA and further distinguish SFA from rapamycin or CsA in their modes of action.
As a cellular gate keeper for growth and division, p53 plays an essential role in sensing various stress signals and serves as a focal point of signal integration to decide whether cells will undergo growth arrest or apoptosis (35)(36)(37). A multitude of stress signals are known to activate p53, including ionizing irradiation, UV, hypoxia, nucleotide deprivation, and chemotherapeutic agents. While most chemotherapeutic drugs are known to activate p53 through DNA damage, other small molecules capable of activating p53 have been reported. Among those is the fumagillin family of angiogenesis inhibitors that FIG. 4. NFB-binding site is required for the activation of the p53 promoter by SFA. A, schematic representations of wild type and mutant p53 promoter-luciferase reporter genes. The DNA sequences for two transcription factor binding sites, NFB and HoxA5, in human p53 promoter were shown. The mutated sequences were highlighted in bold. B, responsiveness of wild type and mutant p53 luciferase reporter genes to SFA stimulation. HCT116 cells were transfected with various luciferase reporter constructs (1 g), and the transfected cells were allowed to recover for 24 h. Cells were incubated in the presence or absence of SFA for 12 h before determination of luciferase activity. C, inhibition of SFA-induced p53 promoter transactivation in cells expressing unresponsive IB␣ mutant or dominant negative IKK␣. HCT116 cells were cotransfected with the 356-bp p53 promoter-Luc plasmid (1 g) and IB␣ mutant or dominant negative IKK␣ expression plasmids (1 g). After recovery for 24 h, the cells were treated with SFA, doxorubicin (Dox) or control solvent for 12 h, before they were harvested for determination of luciferase activity.
block endothelial cell cycle progression by activating p53 (38,39). The ability of SFA to induce p53 in a variety of cells suggests that it may have potential as an inhibitor of angiogenesis in addition to its immunosuppressive activity.
We have shown that SFA is capable of activating p53 transcription, though it remains to be determined whether SFA also affects the stability of p53 mRNA. Several transcription factors have been identified that appear to play a role in the activation of the p53 promoter, among which are NFB and HoxA5 (29,32). Of these two transcription factors, NFB has been shown to mediate the transcriptional activation of p53 in response to DNA damage, though the extent to which NFB contribute to p53 transcription has remained somewhat controversial (30,33). It was reported that an NFB-binding element overlaps with a minimal p53 promoter that responds to chemotherapeutic agents (30). But NFB was reported to play an insignificant role in the activation of p53, as evidenced by the lack of activation of a p53 reporter by expressed p65 and the lack of effect by N-acetyl cysteine on genotoxic stress-induced p53 activation. In contrast, both benzo[a]pyrene, a carcinogen, and daunomycin, an anticancer drug, have been shown to induce p53 transcription via activation of NFB (31, 33). Our FIG. 5. SFA activates the NFB activation cascade. A, SFA induces NFB luciferase reporter gene activation. HCT116 cells were transfected with the NFB-Luc plasmid (1 g) alone or cotransfected with IB␣ mutant or dominant negative IKK␣ expression plasmids (1 g). After recovery for 24 h, the cells were treated with SFA, doxorubicin (Dox) or control solvent for 12 h, before the cells were harvested for determination of the luciferase activity. B, HCT116 cells were treated with various concentrations of SFA for 6 h. Cytosolic cell lysates were prepared and analyzed by Western blot using antibodies against IB␣ and IB␤. C, immunocomplex kinase assay for IKK in HCT116 cells. HCT116 cells were treated with indicated amounts of TNF␣ or SFA for 1 h. Whole cell lysates were prepared and immunoprecipitated with the anti-IKK␣ antibody H740. The kinase reactions were carried out with immunoprecipitates using 1 g of recombinant GST-IB␣ as the substrate. observations that mutation of the NFB site within the p53 promoter led to complete abrogation of reporter gene activation by SFA and that dominant negative form of IKK␣ and mutant IB␣ inhibited activation of p53 reporter gene corroborate with the conclusion that NFB can play an essential role in the transcriptional activation by certain reagents.
Although activation of p53 seems sufficient for the SFAmediated cell cycle arrest at the G 1 phase, it is not absolutely necessary (Fig. 6). When MEFs deficient in p53 were tested for their sensitivity to SFA, they were much less sensitive to SFA than wild type MEFs (Fig. 6), suggesting that p53 is indeed involved in the action of SFA. Interestingly, the resistance to SFA upon p53 deletion is incomplete; their proliferation can still be inhibited by high doses of SFA (Fig. 6). It is likely that redundant proteins exist that can compensate for the lack of p53. It will be interesting to test whether the recently discovered p53 homologs, p63 and p73, may play such a role.
A number of genotoxic agents, be they anticancer drugs or carcinogens, have been shown to be capable of activating p53. And NFB has been shown to be important for the activation of p53 transcription by several agents (31,33). To determine whether SFA is also a genotoxic agent, we determined the ability of SFA to cause DNA damage. Using the Comet assay, we have unambiguously shown that SFA, in contrast to doxorubicin, has no effect on the integrity of genomic DNA, distinguishing SFA from other genotoxic agents. Using dominant negative IKK␣, we were able to block SFA-stimulated p53 luciferase reporter (Fig. 4C) as well as NFB luciferase reporter (Fig. 5A), suggesting that SFA acts upstream of IKK to activate the NFB. Indeed, SFA was found to be capable of activating IKK like TNF␣ (Fig. 5C). The precise mechanism of activation of IKK by SFA will have to await the identification of the direct molecular target of SFA, which may unravel an entry point for pharmacological activation of the NFB pathway and p53.