Human colon carcinoma cell lines HT29, SW620, and HCT116 were obtained from the American Type Culture Collection and were grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum and antibiotics. Cell cultures were maintained at 37°C in 5% CO₂ and 95% air. Monoclonal antibodies to actin, GSH, and p53 were purchased from Millipore Corp. 1620 (wt-p53) and 240 (mt-p53) were purchased from Calbiochem. Piperlongumine was purchased from Cayman Chemical (CAS registry no. 20069-09-4, ≥98%, Ann Arbor, MI, USA). Stock solutions of PL were prepared in DMSO.

After trypsinization, the cell pellets were washed with cold phosphate-buffered saline (PBS) and subjected to sonication in 40 mM Tris-HCl (pH 8.0) containing 1% glycerol, 1 mM EDTA, 0.5 mM DTT, 1 mM PMSF and 1 mM benzamidine and centrifuged. Equal protein amounts from different treatments were electrophoresed on 12% sodium dodecyl sulfate (SDS)-polyacrylamide gels and transferred onto PVDF membranes (Millipore). The membranes were blocked with 5% non-fat dry milk in Tris-buffered saline (pH 8.0) containing 0.1% Tween-20 for 3 h, and subsequently incubated with appropriate primary antibodies. Antigen-antibody complexes were visualized by enhanced chemiluminescence.

Female Nu/Nu mice were obtained from Charles River Laboratories (Wilmington, MA, USA) and fed ad libitum. Female mice are traditionally used for developing xenografts because the males often tend to be aggressive and bite their littermates. They may also destroy the tumors present on their fellow cage mates or kill them. The animals were allowed to acclimatize to a 12 h light/12 h dark cycle, and all procedures were performed under the guidelines of the institutional animal care and use committee (IACUC). HT29 cells (5×10⁶) were injected subcutaneously into the right and left flanks. Tumor volume was calculated every 2–3 days with a caliper using the following formula: Volume = [length × (width)²]/2. The mice weighing 25 g were randomized into three groups and were administered i.p. injections of 5 and 10 mg/kg/day of PL dissolved in 1X PBS saline (6 animals per groups), when the agent was used alone. In the second study, the mice were randomized into six groups and they received 7.5 mg/kg/day PL, 1 mg/kg/day cisplatin, 0.75 mg/kg/day doxorubicin and the combinations of PL with cisplatin or doxorubicin. In both cases control mice received only PBS. The animals were sacrificed on day 21 after drug administration. The tumors were excised, washed and homogenized using a polytron in 40 mM Tris-HCl buffer (pH 8.0) containing 5% glycerol, 1 mM EDTA, 20 μM spermidine, 0.5 mM DTT, 1 mM PMSF, 1 mM benzamidine, 0.5% Triton X-100, and 1 mM sodium vanadate. All samples were centrifuged, and the resulting extracts were used for western blot analyses.

Nuclear extracts were prepared using the nuclear extract lysis buffer containing 25 mM Tris-HCl (pH 8.0), 5% glycerol, 1 mM EDTA, 0.5 mM DTT, 1 mM PMSF and 1 mM benzamidine and centrifuged. Five milligrams of each nuclear extract was used in a binding assay with 20 ng/μl of a biotin-labeled p53 probe (5′ Biotin AGACATGCCTAGACATGCCT-3′) (Signosis, Inc.). The nuclear extract was incubated with p53 probe at room temperature (20–23°C) for 30 min and the protein/DNA complexes were separated on a non-denaturing 6.5% polyacrylamide gel. The gel was transferred to a nylon membrane, and the biotin-labeled oligonucleotides were detected using streptavidin-HRP and a chemiluminescent substrate according to the manufacturer's protocol.

Binding of p53 proteins to their respective consensus recognition sequences was determined by the biotin-labeled oligonucleotide pull-down assays followed by western blotting in a procedure called DNA affinity immunoblotting (22). Nuclear extracts were prepared, and DNA-binding reactions were performed according to our published procedure (20). The following double-stranded consensus recognition sequences were prepared by annealing a strand labeled with biotin at 5′ end with an unlabeled complementary sequence; p53 used were 5′-TACAGAACATGTCTAAG CATGCTGGGGACT-3′. The duplex oligonucleotides (50 nM) were bound to streptavidin magnetic beads in TE buffer (pH 8.0) containing 100 mM NaCl. The beads were washed with TE buffer containing 1 M NaCl and then without salt. The oligo-bound beads were subsequently suspended in a binding buffer [20 mM Tris-HCl (pH 8.0), 75 mM NaCl, 5 mM MgCl₂, 1 mM EDTA, 2.5 mM DTT, 20 mM KCl, 1 μg poly(deoxyinosinic-deoxycytidylic) acid and 5% glycerol] and incubated with nuclear extracts at room temperature for 30 min. The beads were then washed with 300 mM NaCl and the DNA-bound protein was eluted by boiling the beads in SDS sample buffer, electrophoresed on 10% SDS-polyacrylamide gel and western blotting using appropriate antibodies.

Equal protein amounts in lysates from different treatments were precleaned with 5 μl protein-A agarose beads, and immunoprecipitated using 240 p53 mutant (PAb240) and 1620 p53 wt (PAb1620) antibodies. The immuno-complexes were solubilized in non-reducing SDS-sample buffers, subjected to electrophoresis on 12% gels followed by western blotting with D-1 antibodies for p53 which recognizes both the wt and mutant proteins.

Intracellular ROS production was determined by 2′,7′-dichlorofluorescin diacetate (DCFDA) staining followed by fluorescence detection using a Biotek plate reader (Model-Synergy 2SL) with excitation and emission wavelengths set at 485 and 535 nm, respectively. Briefly, cells were incubated with 10 μM DCF-DA solution at 37°C for 0.5 h, washed with PBS twice, treated with different concentrations of PL for 1, 3 and 6 h followed by fluorescence measurement.

For cell viability assays, the yellow tetrazolium dye [(3-(4,5-dimethylthiazolyl-2)-2, 5-diphenyl tetrazolium bromide) (MTT)] was used. HT29 cells were seeded at a density of 7,000 cells per well in 96-well plates and were treated post-24 h with PL at concentrations specified. In some cases, cells were treated with PL for 24 h. They were treated with PL alone or with PL followed by exposure to BCNU or temozolomide or doxorubicin. The cells were washed, and MTT (10 μl of 5 mg/ml) in 1X PBS was added. The plates were incubated for 4 h followed by the addition of DMSO, stored in the dark for 2 h before reading the absorbance at 570 nm.

All experiments were performed three times independent of each other. Results were assessed by Student's t-test. Significance was defined as P<0.05. Power analysis was used to calculate the number of animals required to achieve a statistical power of >80%.

PL, an electrophilic small molecule identified in cell-based, high-throughput screening assays was shown to selectively kill cancer cells without harming the normal epithelial cells (3). Structurally, PL has two active double bonds (Fig. 1A) and can conjugate with small-molecule thiols. Adams et al synthesized 80 structural analogs of PL to investigate some lead compounds from which they concluded that C2–C3 olefin is critical for the electrophilicity of the molecule. They observed that elimination of C2–C3 olefin did not result in ROS elevation or decreased viability of cancer cells while removal of the C7–C8 olefin was associated with decreased toxicity without affecting the generation of ROS. Several analogs of piperlongumine such as piperine and piperlonguminine also have active double bonds that are located in proximity to the carbonyl group but they lack the crucial C2–C3 olefin in their structures. As a result, although these analogs still possess the ability to elevate cellular ROS, they exhibit reduced toxicities (4). Therefore, the therapeutic properties of PL can be attributed to both the olefins present in the molecule. These structural features of PL also appear to underlie the feeble p53 reactivating properties described in this report.

Glutathionylation is a posttranslational modification where low molecular weight thiols such as the glutathione (GSH or GSSG) form mixed disulfides with protein-bound reactive (anionic) cysteines, apparently, as a defensive/protective strategy during oxidative stress (23). This modification is reversible through an increase in GSH/GSSG ratio or enzymatic reactions involving glutare-doxin, thioredoxin or sulfiredoxins which restore the protein sulfhydryl groups back to their reduced state (24). Similar to phosphorylation, glutathionylation is known to inhibit enzyme activities and transcription protein functions and alter protein-protein interactions (25). Our previous study demonstrated that human p53 is a substrate for glutathionylation in vitro and in cells, and p53 function may be modulated by glutathionylation (20,21). Another study suggested that the thiolation induced by PL, and its analogs might play an important role in inducing cancer cell death (4). Because PL induces a redox imbalance, we examined the bulk glutathionylation of proteins as a biochemical marker of oxidative stress. Conjugation of glutathione or glutathione disulfide with bulk proteins was determined by western blotting using a monoclonal antibody against glutathione. In HT29 and SW620 cells, we observed a large increase in protein glutathionylation after PL treatment where a de novo or increased glutathionylation of numerous proteins was observed as reflected by enhanced band densities when compared with the untreated cells (Fig. 1B).

The Nrf-2 (nuclear factor erythroid 2-related factor) is a cytoprotective transcription factor activated during oxidative stress and functions to restore the redox homeostasis. Nrf2 controls the basal and induced expression of an array of antioxidant response element-dependent genes such as the γ-glutamylcysteine synthase, glutathione S-transferases, and NAD(P)H oxidoreductase 1 to regulate the physiological and pathophysiological outcomes of oxidant exposure (26). Consistent with the properties known for PL, we observed an increase in the Nrf-2 protein levels in HT29 and SW620 cells (Fig. 1C). Therefore, an increase in Nrf-2 demonstrates the development and maintenance of oxidative stress by PL in cancer cells. The ability of PL to induce oxidative stress in tumor cell lines with mutant p53 was further explored using the redox probe DCF-DA. DCF-DA oxidizes rapidly to a highly fluorescent 2′, 7′-dichlorodihydrofluorescein (DCF) in stressed cells and the fluorescence intensity is proportional to the cytosolic ROS levels (27). The increase in ROS levels by PL was confirmed when the HT29 and SW620 cells were incubated with DCF-DA for 30 min, followed by PL treatment for 1, 3 and 6 h (Fig. 2B and C). However, there was no significant difference in ROS levels in MCF10a cells after PL treatment for the same time periods (Fig. 2A) indicating that ROS production is entirely selective for cancer cells, not for normal epithelial cells.

p53 is the most commonly mutated gene in human cancers and hence has emerged as a huge target for novel cancer therapies (28). These mutations occurring in the DNA-binding domain abrogate binding of p53 protein to its target sequences, leading to a suppression of target gene products involved in cell cycle arrest, apoptosis and a multitude of oncogenic pathways (29). Additionally, p53 mutations are also associated with the inhibition of two other p53 family members, p63 and p73 through oligomerization, and this leads to the inhibition of apoptosis (30). Reactivation or restoration of functional properties to the mutant p53 proteins is expected to sensitize the human tumors to therapy-induced DNA damage and facilitate a greater antitumor efficacy through a pronounced apoptosis. Most compounds reported to impart at least some wild-type properties to mutant p53 proteins are all electrophilic with great affinity for nucleophilic groups such as the protein bound cysteines or glutathione present in abundance (19). PL is a reactive compound as well (4); therefore, we hypothesized that anticancer effects of PL may stem at least partially from the oxidative stress-induced modification of the p53 protein. Our own findings implicating the reactive cysteines 124, 141 and 182 in the DNA-binding domain of p53 as targets for oxidation and thiolation further supported the above hypothesis (20,21).

We used a number of approaches to validate the functionalization of mutant p53. Chief among these were the conformation specific monoclonal antibodies for p53. pAb 1620 can recognize the wild-type structure, and many cancer mutations perturb this conformation (31). The epitope comprising Arg156, Leu209, Arg 209, and Gln/Asn210 in p53 DNA-binding domain is recognized by pAb1620 (32). On the other hand, a highly conserved, denaturation-resistant epitope on p53 is targeted by the antibody pAb240; it binds to p53 in the mutant conformation and also to denatured p53. Both antibody epitopes are located away from the p53-DNA interface (33).

As a first step to verify the reactivation of mutant p53, the cell lysates from PL-treated and untreated HT29 and SW620 cells were immunoprecipitated with pAb1620 or pAb240 followed by immunoblotting with the DO1 anti-p53 antibodies. The western blots in Fig. 3A show that in both cell lines, the bands corresponding to the wt-like p53, which were absent in control cells appeared after PL-treatment. Further, a concomitant decrease of the mutant p53 protein was evident in extracts immunoprecipitated with pAb240. Next, the above results were verified by direct western blotting using the p53 conformation specific antibodies. The untreated HT29 and SW620 cells which harbor a mutant p53 exhibited a very weak immunoreactivity for pAb 1620, which, however, increased significantly in a time-dependent manner after PL treatment. On the contrary, the intrinsic mutant p53 in these cells decreased gradually as evident from the diminished pAb240 immunoreactivity (Fig. 3B).

To further establish the conformational transition and acquisition of DNA binding of the R273H mutant p53 present in HT29 an SW620 cells, we performed DNA-affinity immunoblotting (DAI) and EMSA, both of which detect the protein specifically associated with DNA. The representative results from these experiments are shown in Fig. 4. The nuclear extracts (NE) from control cells failed to bind with DNA, whereas the NE from PL-treated cells showed a time-dependent increase in DNA binding in the DAI assays (Fig. 4A). Similar data showing the binding of p53 with DNA in both HT29 and SW620 cells were obtained in EMSAs; the bands corresponding to DNA-bound proteins were diminished by the addition of the non-biotinylated cold probe, demonstrating the specificity of interaction (Fig. 4B). Consistent with the mutant p53 conversion to its functional form, we observed upregulation of the target genes, MDM2 and Bax in both cell lines after PL treatment (Fig. 5). A significant increase in the levels of apoptotic markers, cleaved caspase-3 and cleaved PARP was also observed (Fig. 5), suggesting that the oxidative milieu may prime and usher the tumor cells along the cell death pathways.

To investigate the selective anticancer effects of PL in the context of the DNA contact mutants of p53 and our observations that PL can impart wt-like the protein conformation to the mutant proteins (Figs. 3–5), we performed cell survival assays with PL alone or PL in combination with some clinically used drugs against cancers. PL by itself, over a concentration range of 2.5–15 μM for 24 h, showed significant anti-proliferative effects against cell lines of various cancer types. These included the breast cancer MCF-7, colon cancer HT29, HCT116, SW620, and the Mia Paca of pancreatic origin (Fig. 6). However, consistent with earlier results (3) PL was 10–12-fold less toxic to MCF10a normal breast epithelial cells as also shown in Fig. 6. Based on these observations, we chose 7.5 μM PL, which caused 35–55% cell-kill for 24 h as the preincubation period for testing the potentiation of cytotoxicities of temozolomide (TMZ), 1,3 bis(2-chloroethyl)-1-nitrosourea (BCNU) and doxorubicin (Dox). Three cell lines, the HT29, HCT116 and SW620 were tested to determine the drug potentiation by PL. In this setting, the TMZ (250–1,000 μM range) + PL combination showed 2- to 3-fold increased cytotoxicity compared with TMZ alone (Fig. 7). For BCNU and doxorubicin, the potentiation was ~2-fold in different cell lines. Our findings suggest that addition of PL to the anticancer regimens will be beneficial and result in increased tumor cell killing. The enhanced anticancer effects are likely to occur at least in part through the functional restoration of the p53 tumor suppressor observed in this study.

To probe whether PL induces a conversion of mutant p53 protein to its wt-like counterpart in vivo, we developed subcutaneous xenografts by injecting the HT29 cells in nude mice. PL (5–10 mg/kg) was administered every day to the tumor bearing mice for 20 days. A significant and PL dose-dependent delay of tumor growth was observed (Fig. 8A). No changes were discernible in the body weight of animals indicating a lack of toxicity (Fig. 8B). Further, when the lysates from the excised tumors were immunoblotted, there was an apparent decrease in the R273H mutant p53 protein levels (Fig. 8C). This decrease was accompanied by enhanced levels of both the cleaved PARP and cleaved caspase-3 (Fig. 8C), verifying that PL can trigger apoptotic initiation in tumor tissues as well.

To explore the possibility of combining PL with current chemotherapy drugs, we evaluated the tumor growth delay in HT29 xenografts given cisplatin or doxorubicin along with PL (7.5 mg/kg/day). As shown in Fig. 9, PL treatment alone produced significant tumor regression. Simultaneous treatment of PL with cisplatin or doxorubicin resulted in a vigorous and significant reduction of tumor volume, compared with either agent alone. Statistical analyses indicated that the antitumor effects were synergistic (Fig. 9A). Changes in body weights were not significantly different between the control and PL-treated groups (P>0.5) indicating that no apparent adverse effects were associated with the chemotherapy (Fig. 9B). Western blot analyses again revealed a reduction in mutant p53 levels, unaltered total p53 levels (reactive with the pan DO-1 Ab), along with a marked upregulation of apoptotic proteins such as the Bax and cleaved PARP (Fig. 9C). Taken together, these findings suggest that oxidative stress mediated by PL can potentiate the cytotoxic effects of many clinically used anticancer drugs.

In continuation of our research interests in oxidative stress signaling (34), and regulation of p53 functions by redox imbalance in human cancers (20,21), the present study investigated the biochemical effects of PL on the R273H mutant p53 present in HT29 and SW620 cells and the therapeutic consequences thereof. Fig. 10 summarizes our findings. We propose that PL, as an electrophile, can either react directly with the protein-bound anionic cysteines or conjugate with the SH-group of glutathione (4,5); the glutathione S-transferases may facilitate the latter reaction. The ROS generation by PL and the thiol conjugations are likely to deplete the cellular GSH levels and trigger protein thiolation. We hypothesize that mutant p53 proteins are efficient substrates for glutathionylation, which in turn can induce structural perturbations in the defective DNA-binding domain of the tumor suppressor and restore some functionality (Fig. 10). While the PL-induced p53 reactivation is somewhat feeble, and the mechanism(s) involved are yet to be proven, our recent study with a hybrid C-7 aryl piperlongumine derivative indeed suggests that induction of redox-imbalance may serve as a general platform to rescue the cancer mutations in p53 (35). In conclusion, the findings made here and the evidence that oxidative stress may prevent distant metastasis (36) highlight the therapeutically beneficial and exploitable aspects of tumor redox biology.