Extract from Asteraceae Brachylaena ramiflora induces apoptosis preferentially in mutant p53-expressing human tumor cells

Carcinogenesis vol.31 no.6 pp.1045–1053, 2010 doi:10.1093/carcin/bgq084 Advance Access publication April 28, 2010

Extract from Asteraceae Brachylaena ramiflora induces apoptosis preferentially in mutant p53-expressing human tumor cells Masoud Karimi, Francesca Conserva, Salah Mahmoudi, Jan Bergman1, Klas G.Wiman and Vladimir J.N.Bykov
Department of Oncology-Pathology, Cancer Center Karolinska, Karolinska University Hospital, Karolinska Institutet, SE-171 76, Stockholm, Sweden and 1 Department of Biosciences and Nutrition, Karolinska Institutet, Novum, SE-141 57 Huddinge, Sweden
To whom correspondence should be addressed. Tel: þ46 8 517 70 870; Fax: þ46 8 517 70 870; Email: [email protected]

Introduction The p53 tumor suppressor regulates cell cycle progression and cell survival in response to cellular stress (1). DNA damage or oncogenic stress induces p53 protein levels, allowing elimination of incipient tumor cells by apoptosis (2,3). p53 is frequently mutated in human tumors (4) (see www-p53.iarc.fr; p53.free.fr). Mutant p53 proteins are usually expressed at high levels in tumor cells. Moreover, mutant p53-carrying tumors are often more resistant to conventional chemotherapy and radiotherapy (5,6). Therefore, mutant p53 is an attractive target for novel cancer therapy. Since around half of all human tumors harbor p53 mutations (see www-p53.iarc.fr; p53.free.fr), the availability of drugs that reactivate mutant p53 might revolutionize treatment of human cancer. p53 mutation allows evasion from apoptosis in tumors and may increase resistance to chemotherapy. Pharmacological restoration of p53 function in tumors carrying mutant p53 should restore p53dependent apoptosis and thereby efficiently eliminate tumors in vivo. This notion is supported by studies showing that reconstitution of wild-type (wt) p53 expression induces tumor regression in vivo (7,8). We previously identified the mutant p53-reactivating compound p53 reactivation and induction of massive apoptosis (PRIMA-1) in a cellular screening of the Diversity set from National Cancer Institute (NCI), based on an assay that scores for mutant p53-dependent growth inhibition (9). The growth suppression profile of PRIMA-1 is distinct from that of conventional chemotherapeutic drugs in that PRIMA-1 preferentially targets mutant p53-expressing cells (10). In contrast, most chemotherapeutic drugs, particularly DNA damaging agents, preferentially target wtp53-harboring cells. Only taxol shows some selectivity for mutant p53-expressing tumor cells, as reported (11), but to a lesser extent than PRIMA-1 (10). Abbreviations: FACS, fluorescence-activated cell sorting; MIRA-1, mutant p53-dependent induction of rapid apoptosis; mRNA, messenger RNA; NAC, N-acetylcysteine; NCI, National Cancer Institute; PCC, pairwise correlation coefficient; PRIMA-1, p53 reactivation and induction of massive apoptosis; ROS, reactive oxygen species; wt, wild-type.

Materials and methods Cells, kits and antibodies The human Saos-2 osteosarcoma and H1299 lung adenocarcinoma cells are p53 null. The sublines Saos-2-His273 and H1299-His175 carry the indicated tetracycline-regulated mutant p53 expression constructs (Tet-Off). Human HCT116 colon carcinoma cells carry wtp53 (p53þ/þ) and the isogenic HCT116 (p53/) cells are p53 null. HCT116 p53 248/ express Trp248 mutant p53 and HCT116 p53 248/þ express Trp248 mutant p53 and wtp53. SW480 colon carcinoma cells express endogenous His273/Ser309 mutant p53. CaspaTag TM Pan Fluorescein Caspase (VAD) activity kit was from Intergen (Oxford, UK), carboxyfluorescein (FAM)-labeled caspase inhibitor FAMDEVD-FMK was from Immunohistochemistry Technologies (Bloomington, MN), polyclonal rabbit anti-p53 and anti-Bax antibodies were from Santa Cruz Biotechnology (Santa Cruz, CA), and fluorescein isothiocyanate-conjugated anti-rabbit immunoglobulin was from Vector Laboratories (Burlingame, CA). Extracts from terrestrial plants and marine invertebrates Extracts from Asteraceae B.ramiflora, N37063 (terrestrial plant); Porifera Demospongiae Thorectidae Ircinia, C3483 (marine invertebrate); Flacourtiaceae Flacourtia indica, N12727 (terrestrial plant) and Asteraceae Vernonia garnieriana, N38283 (terrestrial plant) were obtained from NCI, Developmental Therapeutic Program (Bethesda, MD, http://dtp.nci.nih.gov/index.html). Growth suppression assay Mutant p53 expression in cells harboring a tetracycline-controlled expression cassette was shut off by incubation with doxycycline for 1 week (5 lg/ml). Cells were grown in 96-well plates at a density of 3000 cells per well with or without doxycycline and treated with tested substances for 96 h. Cell viability was then assessed by the WST-1 cell proliferation reagent (Roche, Stockholm, Sweden). Western blotting Cells were seeded in six-well plates at a density of 10 000 cells/cm2 and treated with N37063 the following day. After 24 h, cells were harvested by trypsinization, lysed and analyzed by western blotting. Equal amount of total proteins were separated by electrophoresis on a sodium dodecyl sulfate–polyacrylamide gel and then transferred on a nitrocellulose membrane using the iBlotTM Dry Blotting System (Invitrogen, Stockholm, Sweden). Membrane was blocked and probed with corresponding antibodies. Protein bands were visualized using SuperSignalÒ West Femto Maximum Sensitivity Substrate (Pierce, Rockford, IL). Flow cytometry For DNA fragmentation assays, cells were grown in 12-well plates at an initial density of 40 000 cells per well. After treatment with plant extract, cells were harvested by trypsinization, fixed with 70% ethanol, treated with RNase A (0.25 mg/ml) and stained with propidium iodide (0.05 mg/ml). Samples were

Ó The Author 2010. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected]

1045

Downloaded from http://carcin.oxfordjournals.org at Karolinska Institutet on June 1, 2010

The p53 tumor suppressor gene is inactivated by point mutation in a large fraction of human tumors, allowing evasion of apoptosis and tumor progression. p53 mutation is often associated with increased resistance to therapy. Pharmacological reactivation of mutant p53 is an attractive therapeutic strategy. We previously identified p53 reactivation and induction of massive apoptosis, a low-molecular weight compound that suppresses the growth of cancer cells in a mutant p53-dependent manner. Here, we report the identification and characterization of an extract from the terrestrial plant Brachylaena ramiflora (Asteraceae) that preferentially induces apoptosis in human tumor cells expressing mutant p53. Further analysis of this extract and identification of active compounds may provide novel structural scaffolds for the development of mutant p53-targeting anticancer drugs.

Several mutant p53-reactivating small molecules have been identified in recent years (12). Some, like CP-31398, were identified by in vitro screening for compounds that preserve native conformation of p53 protein under denaturing conditions (13). In contrast, PRIMA-1 (9) and mutant p53-dependent induction of rapid apoptosis (MIRA-1) (14) were identified in cellular screens for substances that suppress cell growth in a mutant p53-dependent manner. In addition, a peptide derived from the C-terminus of p53 (15), and CDB3 peptide derived from the p53-binding protein, 53BP2, have been shown to reactivate mutant p53 (16). The NCI has collected a large number of different organic molecules and natural extracts from all over the world. Most of these substances have been tested on a panel of 60 human cancer cell lines (see www.dtp.nci.nih.gov). We performed an in silico screening of the NCI natural product database for compounds and plant extracts that suppress growth of human tumor cells with a profile similar to that of PRIMA-1. We identified an extract from the terrestrial plant Brachylaena ramiflora (Asteraceae) that has mutant p53-dependent pro-apoptotic activity in cell culture and restores some wt properties to mutant p53.

M.Karimi et al.

analyzed on a FACScan flow cytometer (Becton Dickinson, San Jose, CA) according to standard procedures. For caspase activation assay, cells were harvested by trypsinization and labeled with FLICA reagent (CaspaTag TM Pan Fluorescein Caspase (VAD) activity kit, Intergen, UK) according to the manufacturer’s instructions. Samples were analyzed on a FACScan. For reactive oxygen species (ROS) determination, cells were stained with 2,7dichlorofluorescein diacetate 5 lg/ml for 30 min at 37°C, harvested by trypsinization and analyzed by fluorescence-activated cell sorting (FACS). All the FACS data were analyzed by the CellQuest software, version 3.2.1 and by WinMDI 2.9 software. Glutathione-based depletion of N37063 N37063 was dissolved in phosphate-buffered saline at a concentration of 10 mg/ml, mixed with glutathione immobilized on agarose beads (Thermo Fisher Scientific, Waltham, MA) and incubated on rocking platform at room temperature for 1 h. Then suspension was loaded on a spin column (Thermo Fisher Scientific) and centrifuged at 2000 r.p.m. for 2 min. The eluate was collected and stored at 20°C.

Real-time reverse transcription–polymerase chain reaction Saos-2 and Saos-2-His273 cells were treated with 5 lg/ml of N37063 and harvested after 6 h. RNA was extracted using the RNeasy Mini Kit (QIAGEN, Sollentuna, Sweden) according to the manufacturer’s instructions. Complementary DNA was synthesized according to standard procedures and 100 ng complementary DNA were added to 10 ll of 2 TaqMan Universal PCR Master Mix (Applied Biosystems, Carlsbad, CA) and 1 ll 20 TaqMan Gene Expression Assay and appropriate TaqMan probe. Amplification consisted of a first step with 2 min at 50°C and 10 min at 95°C followed by 40 cycles with 15 s denaturation at 95°C and 1 min annealing/extension at 60°C. Reactions were performed in 96-well optical PCR plates (Applied Biosystems) using the ABI PRISM 7700 Sequence Detection System (Applied Biosystems). Results were analyzed with the comparative Ct method using glyceraldehyde 3-phosphate dehydrogenase as the endogenous control. Fractionation One milligram of N37063 was dissolved in 200 ll water and mixed with 400 ll of n-hexan (Sigma-Aldrich, Stockholm, Sweden) by vortexing for 1 min. Then, mixture was centrifuged at 2000 r.p.m. for 10 min. The organic fraction, the interphase and aqueous fractions were collected separately. The extraction was repeated three times. The corresponding fractions were pulled together, samples were frozen at 20°C and freeze-dried in SpeedVac (Savant SVC100H, Ramsey, MN). Statistical analysis Data for 60 human tumor cell lines were extracted from the NCI database (http//dtp.nci.nih.gov). We selected 34 lines using the following criteria: availability of sequence information for p53 status, data on p53 protein levels, growth suppression profiles for all compounds selected for analysis and data for at least two cell lines carrying wtp53 and two lines carrying mutant p53 for each cancer type. We extracted growth inhibition profiles for 44 drugs representing major mechanisms of activity: alkylating agents, topoisomerase I inhibitors, topoisomerase II inhibitors, antimitotic agents, DNA antimetabolites and DNA/RNA antimetabolites. Statistical analysis was carried out by Statistica 8.0 software (StatSoft, Tulsa, OK). The GI50 profiles were compared by using the Compare algorithm of the NCI database. GI50 is defined as the concentration of a tested compound that inhibits cell growth by 50%. The obtained GI50 values were analyzed by the Wilcoxon matched pairs test. Cluster analysis was performed according to Ward’s method that evaluates distance between clusters through the analysis of variance by minimizing the sum of squares of any two

1046

Results In silico screening of NCI library We screened the NCI Natural product repository database for substances that inhibit tumor cell proliferation in a manner similar to that of PRIMA-1. The Compare program yielded a number of extracts based on maximal pairwise correlation coefficients (PCC) with PRIMA-1. The highest ranked products were C9904 (PCC 5 0.70), N37063 (PCC 5 0.69), C3483 (PCC 5 0.68), N12727 (PCC 5 0.67) and N38283 (PCC 5 0.67). These five best matches were further analyzed for their selectivity, based on GI50 values extracted from the NCI database, toward mutant or wtp53-expressing cells as described (10). Detailed information about the cell lines used in the analysis and GI50 measurements is presented in the supplementary Tables I, II and III (available at Carcinogenesis Online). Figure 1A shows mutant or wtp53 selectivity of the studied extracts in five cancer types, non-small-cell lung cancer, colon cancer, ovarian cancer, renal cancer and melanoma. Figure 1B shows mutant versus wtp53 selectivity of the identified extracts based on average GI50 values for each cancer type. Comparisons were performed according to the Wilcoxon matched pairs test. Extracts N37063 and C3483 showed mutant p53 selectivity in growth inhibition at P 5 0.04. A correlation between levels of mutant p53 expression and GI50 values was observed for both N37063 (r 5 0.56, P 5 0.02) and C3483 (r 5 0.61, P 5 0.009) according to the linear regression model. Mutant p53-dependent inhibition of cell growth by N37063 Extracts N38383, C3483, N37063 and N12727 were requested from the NCI Natural product repository as the most selective for mutant p53-expressing cells and were tested in WST-1 cell proliferation assays in H1299 and H1299-His175 cells at concentrations up to 25 lg/ml. N38383 and C3483 did not affect cell growth and were therefore not studied further. As shown in Figure 2A, N37063 had a more pronounced effect in the mutant p53-expressing Saos-2 cells according to our WST-1 assays. Treatment with 2.75 lg/ml of N37063 resulted in significant growth inhibition or cell death (38% survival) in H1299-His175 cells but had little or no effect on the p53-null parental H1299 cells. N3483 was inactive in both H1299 and H1299-His175 cells at concentrations up to 80 lg/ml. We also analyzed the effect of N37063 on cell cycle distribution using FACS (Table I and Figure 2B). To allow a more stringent analysis of the role of mutant p53 expression in the same cellular background, Saos-2 and Saos-2-His273 cells were cultured without or with doxycycline that switches off expression of the mutant p53 construct. The cells were then treated with 5 lg/ml of the Asteraceae extract for 72 h and analyzed by FACS. We observed only minor cell death, 3.5 ± 0.5% and 2.1 ± 0.6%, respectively, in p53-null Saos-2 cells in the absence and the presence of doxycycline. However, the N37063 extract induced substantial cell death, 48.5 ± 0.5%, in Saos-2His273 cells in the absence of doxycycline (mutant p53 expressed) but had a significantly reduced effect, 33 ± 4% cell death, in Saos-2His273 cells in the presence of doxycycline (mutant p53 switched off) (Figure 2B). Similarly, treatment of p53-null H1299 cells with 5 lg/ml of N37063 resulted in 4 ± 2% and 3.5 ± 3.5% cell death in the absence or presence of doxycycline, respectively. For H1299-His175 cells, we observed significant cell death, 34.5 ± 2.5%, in the absence of doxycycline (mutant p53 expressed) but only 14.5 ± 0% cell death in the presence of doxycycline (mutant p53 switched off) (Figure 2B). Mutant p53 expression was significantly reduced by treatment with doxycycline. However, some residual expression remains, particularly in Saos-2-His273 cells (Figure 2B, right panel). N12727 did not show

Downloaded from http://carcin.oxfordjournals.org at Karolinska Institutet on June 1, 2010

TransAM assay Cells (1.2  106) were grown in 10 cm plates and treated next day with 5 lg/ml of N37063 for 18 h. Cells were harvested by trypsinization and nuclear protein fractions were prepared by a nuclear extract kit (Active Motif, Rixensart, Belgium). Equal amount of total nuclear protein, 10 lg, was loaded on to a 96-well plate coated with an immobilized oligonucleotide containing a p53 consensus binding site (TransAM, p53 Transcription Factor Assay Kit; Active Motif). Nuclear extract of H2O2-treated MCF7 cells (provided by the manufacturer) served as a positive control. Anti-p53 and anti-rabbit horseradish peroxidase antibodies were used to quantify the amount of bound p53 protein. The horseradish peroxidase signal was developed by a substrate provided by the manufacturer and samples were analyzed in an enzyme-linked immunosorbent assay reader at 450 nm. In addition, nuclear protein fractions extracted from untreated H1299 and H1299-His175 cells were incubated for 1 h at þ4°C with 2 and 5 lg/ml of N37063. Equal amounts, 11 lg, of total nuclear protein were assayed with the TransAM p53 kit as above.

hypothetical clusters that can be formed at each step. Ward’s amalgamation method yielded more coherent clusters as compared with other average linkage methods available in our statistical package. As a linkage distance between clusters, we have chosen the Pearson correlation coefficient.

Extract from Asteraceae induces apoptosis

any mutant p53-dependent effect in H1299/H1299 His175 or Saos2/Saos-2-His273 cells (data not shown). Treatment with 3 lg/ml of N37063 caused a marked increase in the G2/M fraction of cells both in the absence and presence of mutant p53 expression. Nonetheless, N37063 induced a significant mutant p53dependent cell death in both Saos-2 and H1299 cells. Importantly, inhibition of mutant p53 expression by doxycycline reduced N37063induced cell death, confirming that mutant p53 is critical for this effect. Caspase activation assay Next, we assessed whether the N37063 extract induced cell death by apoptosis using a caspase activation assay. As shown in Figure 2C, treatment with 5 lg of N37063 for 24 h resulted in a significantly higher fraction of active caspase-positive Saos-2 cells in the presence of mutant p53. We observed 9.7 ± 1.2% active caspase-positive Saos-

2 cells, 18 ± 6% active caspase-positive Saos-2-His273 cells and 2.8 ± 0.3% active caspase-positive Saos-2-His273 cells upon treatment with doxycycline that downregulates mutant p53 expression (Figure 2C). We obtained even more striking mutant p53-dependent caspase activation in H1299 cells. Treatment with 3 lg/ml N37063 induced 42% active caspase-positive H1299-His175 cells but no detectable increase in caspase activity was observed in the parental H1299 cells (Figure 2C). N12727 did not trigger any mutant p53dependent activation of caspases in H1299/H1299-His175 and Saos2/Saos-2-His273 cells (data not shown). N37063 suppresses growth of colon carcinoma cells expressing mutant and wtp53 To extend our analysis to other tumor types, we tested the effect of N37063 on human colon carcinoma cells carrying wt or mutant p53. Figure 2D shows growth suppression induced by N37063 in

1047

Downloaded from http://carcin.oxfordjournals.org at Karolinska Institutet on June 1, 2010

Fig. 1. Identification of N37063. (A) Growth inhibition profiles of five identified extracts and PRIMA-1 in cell lines derived from colon, non-small-cell lung cancer (NSCLC), ovarian cancer, renal cancer and melanoma. Data are the calculated factors of selectivity against wtp53 or mutant p53-carrying lines. Factor of selectivity (F) of an agent for wt or mutant p53-carrying cells was determined as follows: if GI50wt . GI50mt, F 5 GI50wt/GI50mt; if GI50wt , GI50mt, F 5 GI50mt/GI50wt. F . 1 indicates that an agent is preferentially targeting tumor cells carrying mutant p53. F , 1 indicates that an agent is preferentially targeting tumor cells carrying wtp53. (B) Mutant p53-dependent growth suppression of the identified extracts. GI50 values (y-axis) were calculated for wt and mutant p53-carrying tumor cell lines for each extract.

M.Karimi et al.

Downloaded from http://carcin.oxfordjournals.org at Karolinska Institutet on June 1, 2010

Fig. 2. N37063 induces mutant p53-dependent apoptosis in human tumor cells. (A) N37063 preferentially suppresses growth of mutant p53-expressing Saos-2 and H1299 cells according to the WST-1 cell proliferation assay. (B) N37063 induces apoptosis in Saos-2-His273 and H1299-His175 cells as demonstrated by the appearance of cells with fragmented DNA (sub-G1 DNA content) in FACS analysis. Quantitative data after treatment with 3 lg/ml of N37063 are shown in the right panel. Right panel, bottom: regulation of p53 expression by doxycycline. (C) N37063 induces active caspase-positive cells in a mutant p53-dependent manner in Saos-2/Saos-2-His273 and in H1299/H1299-His175 cells according to the CaspaTag assay (FACS). Quantitative data after treatment with 3 lg/ml of

1048

Extract from Asteraceae induces apoptosis

Table I. Cell cycle analysis Saos-2

H1299

P53 null

0 N37063

His273

P53 null

His175

G1 (%)

S (%)

G2/M (%)

G1 (%)

S (%)

G2/M (%)

G1 (%)

S (%)

G2/M (%)

G1 (%)

S (%)

G2/M (%)

61 53

27 27

12 20

53 39

27 28

20 33

69 53

10 13

21 34

60 44

15 13

25 43

Induction of p53 target genes As a sign of mutant p53 reactivation, we assessed expression of the known p53 target genes MDM2, p21 and PUMA. Treatment with 5 lg/ml of N37063 for 24 h induced PUMA expression according to immunostaining in Saos-2-His273 cells but not in p53-null Saos-2 cells (Figure 3A). Similar results showing mutant p53-dependent induction of PUMA were obtained in H1299-His175 cells but not in the parental p53-null H1299 cells (data not shown). Treatment of Saos-2-His273 cells with 5 lg/ml of N37063 for 6 h induced p21 messenger RNA (mRNA) by 1.4-fold, of PUMA mRNA by 1.3fold and of MDM2 mRNA by 2.4 fold. No changes in the expression of the correspondent mRNAs were detected in Saos-2 cells (Figure 3B). Next, we verified induction of p53 target genes after treatment with N37063 in Saos-2-His273 and H1299-His175 cells by western blotting. MDM2, p21 and PUMA were induced in the mutant p53expressing Saos-2-His273 and H1299-His175 cells (Figure 3C), but such changes in p53 target protein expression were either less pronounced or undetected in the corresponding p53-null cells. For instance, induction of p21 in Saos-2-His273 cells was 3.3-fold, whereas only a 2-fold induction was detected in Saos-2 cells. Interestingly, N37063 enhanced mutant p53 expression in both Saos-2His273 and H1299-His175 cells. N37063 promotes an oxidative environment in tumor cells expressing mutant p53 Since we observed that N-acetylcysteine (NAC) blocks the effect of N37063 and since NAC can inhibit the formation of ROS, we investigated whether the extract induces ROS. We treated Saos-2 and Saos-2-His273 cells with N37063 for 24 h, stained cells with 2,7-

dichlorofluorescein diacetate, a reduced form of fluorescein that fluoresces upon oxidation, and analyzed samples by FACS (17). We found that N37063 induced ROS in mutant p53-expressing cells at the time and the concentrations used but not in the corresponding p53-null cells (Figure 3C, left and middle panels). Of note, the induction of ROS triggered by N37063 occurs prior to DNA fragmentation as judged by sub-G1 cell population (Figure 3D, right panel). Role of reactive products in the biological effect of N37063 We have shown previously that several mutant p53-reactivating small molecules, i.e. MIRA-1, SH group targeting and induction of massive apoptosis (STIMA-1) and PRIMA-1 act via modification of cysteines in mutant p53 and that their activity is blocked by NAC (18–20). To examine if N37063 might contain substances with a similar mode of action, we first treated H1299-His175 cells with 5 mM of NAC followed by the N37063 at different concentrations. Control samples were treated with plant extract alone or with NAC alone. Treatment with 8 lg/ml of N37063 resulted in a substantial cell death (74.5 ± 4%). However, NAC completely blocked cell death induced by N37063 according to FACS–propidium iodide (Figure 3E). NAC alone did not have any effect on cell survival. Next, we have depleted N37063 of sulfhydryl-reactive products by incubating the plant extract with glutathione immobilized on agarose beads. After depletion, the extract and untreated N37063 were incubated with H1299-His175 cells for 4 days. The number of living cells was estimated by the WST-1 cell proliferation assay. The depletion of sulfhydryl-reactive products clearly reduced the ability of N37063 to suppress growth of H1299-His175 (Figure 3F). p53 DNA-binding assay Our findings that N37063 selectively affects mutant p53-expressing cells and that it triggers activation of the p53 target genes p21, MDM2 and PUMA, suggested that the extract restores wtp53-dependent transcription. Therefore, we asked whether the N37063 extract is able to restore DNA binding to His175 mutant p53 expressed in H1299His175 cells. We addressed this question using the TransAM DNA-binding assay. Treatment with 5 lg/ml of N37063 increased the DNA-binding capacity of His175 mutant p53 by 40% (data not shown). This increase corresponded to the induction of mutant p53 expression observed in H1299-His175 (Figure 3B). Next, we treated protein extracts isolated from H1299 and H1299-His175 cells with 2 and 5 lg/ml of N37063 during 1 h on ice. No induction of the DNA binding was detected according to the TransAM assay, suggesting again that the observed increase in DNA binding was indeed due to the induction of mutant p53 levels. Characterization of active components in Asteraceae plant extract Two substances active against A2780 ovarian carcinoma cells were previously isolated from B.ramiflora and structurally characterized

N37063 are shown in the right panel. (D) Analysis of the effect of N37063 on HCT116 colon carcinoma cells with different p53 status using the WST-1 assay. Both wt and Trp248 mutant p53 expression confer increased sensitivity to N37063. (E) Western blot analysis showing expression of p53, PUMA and p21 in HCT116 cells lacking p53 or expressing either wtp53 or Trp/wtp53 or Trp/null after 20 h of treatment with 5 lg/ml of N37063.

1049

Downloaded from http://carcin.oxfordjournals.org at Karolinska Institutet on June 1, 2010

HCT116 cells expressing either wtp53 or Trp248 mutant p53 alone or both or lacking p53. We found that the half maximal inhibitory concentration value for HCT116 wtp53þ/þ cells is 3.1 lg/ml, whereas the half maximal inhibitory concentration value for HCT116 p53/ cells is 5.8 lg/ml. For mutant p53-expressing HCT116 p53 Trp248/ and HCT116 p53 Trp248/wt cells, the half maximal inhibitory concentration values are 3.3 and 2.7 lg/ml, respectively. Thus, both wtp53 and Trp248 mutant p53 expression confers sensitivity to N37063. In the case of Trp248 mutant p53, this effect is observed regardless of whether the mutant is expressed in a p53-null or a wtp53 background. Interestingly, we did not observe any significant induction of sub-G1 cell population in HCT116 p53þ/þ (12.4 ± 0.3%) and in HCT116 p53/ (15.6 ± 1.4%) cells after 72 h treatment with 4 lg/ml of N37063. Figure 2E shows PUMA and p53 expression in HCT116 cell lines after treatment with 5 lg/ml of N37063. PUMA was not induced in HCT116 p53þ/þ cells and HCT116 p53/ cells, although some induction was detected in HCT116 p53 Trp248/wt cells. Levels of wtp53 were not changed after the treatment with N37063.

M.Karimi et al.

Downloaded from http://carcin.oxfordjournals.org at Karolinska Institutet on June 1, 2010

Fig. 3. Mutant p53-dependent induction of p53 target genes. (A) induction of PUMA in Saos-2/Saos-2-His273 according to immunofluorescence staining. (B) Real-time reverse transcription–polymerase chain reaction showing induction of p21, PUMA and MDM2 mRNAs after treatment with 5 lg/ml of N37063. (C) Western blot showing induction of the indicated proteins in Saos-2/Saos-2-His273 and H1299/H1299-His175 cells after treatment with 5 lg/ml of N37063. (D) N37063 induces ROS in a mutant p53-dependent manner as shown by 2,7-dichlorofluorescein diacetate staining and FACS analysis. Right panel, generation of ROS precedes induction of cell death after treatment of H1299-His175 cells with 5 lg/ml of N37063. (E) NAC prevented DNA fragmentation induced by N37063 in H1299-His175 cells. (F) Depletion of sulfhydryl-reactive products in N37063 by immobilized glutathione has reduced its cytotoxic properties in H1299-His175 cells.

1050

Extract from Asteraceae induces apoptosis

(21). One of the first steps of isolation was hexane extraction of the aqueous phase of extract. In full agreement with the reported data, we observed that the cytotoxic activity against H1299-His175 and Saos-2-His273 cells is present exclusively in the organic phase after extraction with hexane (Figure 4A). Notably, the organic phase was not only cytotoxic for mutant p53-expressing cells but also had mutant p53-dependent activity. After extraction with hexane, the aqueous fraction lost its cytotoxic properties (Figure 4A). We obtained the two known compounds that were isolated from the organic phase, kairatenyl palmitate and hopenyl palmitate (Figure 4B), and tested their activity in H1299 and H1299-His175 cells (Figure 4B). Both the tested compounds showed mutant p53-dependent growth inhibition similar to that of the hexane extraction phase and similar to the whole N37063 extract, although with 25% lower potency. Moreover, NAC inhibited the activity of kairatenyl palmitate by 10% and the activity of hopenyl palmitate by 50% in our WST-1 proliferation assay (data not shown). Clustering of the Asteraceae plant extract with PRIMA-1 and MIRA-1 in the NCI database By using cluster analysis, we generated a dendrogram based on the growth suppression profiles of 46 known anticancer drugs across a panel of 34 cell lines as described in supplementary Tables I and

II and as reported before (10). Most of the selected compounds either have been or are in clinical uses or are in clinical trials (Figure 5). Major groups of compounds were not as easily recognizable as in the case of analysis of all 118 compounds in all cell lines in the NCI database (11). However, clustering according to the mechanism of action was achieved. Some drugs, e.g. camptothecins, formed a tight cluster, whereas others formed groups with each other in two to three different clusters per type of agent. As shown in Figure 5, the Asteraceae plant extract clustered together with the previously identified mutant p53-reactivating compounds PRIMA-1 and MIRA-1. This confirms our experimental data indicating that the extract selectively inhibits growth and survival of mutant p53expressing cells. Discussion Terrestrial and marine plants are large sources of chemicals that have successfully made their way into the clinic. In fact, 70% of all drugs used in chemotherapy today are either products of natural origin or are based on natural product pharmacophores (22). With this in mind, we decided to use database mining to identify novel substances in the products isolated from the Earth biosphere. We applied in silico methods to search for compounds with similar growth inhibition pattern as PRIMA-1 across NCI’s panel of human

1051

Downloaded from http://carcin.oxfordjournals.org at Karolinska Institutet on June 1, 2010

Fig. 4. Characterization of active components in N37063. (A) Mutant p53-dependent growth suppression is associated with the organic phase obtained by n-hexan extraction of N37063. (B) Structural formulas of kairatenyl and hopenyl palmitates. Both substances show mutant p53-dependent growth suppression in H1299-His175 cells as demonstrated by the WST-1 assay.

M.Karimi et al.

tumor cell lines. N37063 and N12727 were selected as the extracts showing the highest mutant p53-dependent growth suppression in the NCI database of Natural products. As judged by growth suppression, induction of DNA fragmentation and caspase activation, N37063 was considered the most promising extract and was thus chosen for further characterization. The selectivity of N37063 for mutant p53-expressing cells was similar to that of PRIMA-1 (9). We also have found that wtp53 expression sensitizes tumor cells to N37063. N37063 is an extract of the plant B.ramiflora, a member of the Asteraceae family, which is widespread in the rainforests of tropical

1052

Downloaded from http://carcin.oxfordjournals.org at Karolinska Institutet on June 1, 2010

Fig. 5. Dendrogram showing hierarchical clustering of 43 different anticancer agents, along with PRIMA-1, MIRA-1 and N37063. A cluster tree of 46 compounds was obtained based on data for cell lines with wt and mutant p53. The metric distance used (1-Pearson correlation coefficient) was expressed as a percentage of the maximum distance between clusters. Clustering was performed according to Ward’s method.

Africa and Madagascar. A number of sesquiterpenoids have been isolated from plants of the Brachylaena family and some of these compounds possess antibacterial activity (23). Interestingly, several terpenes that are cytotoxic against tumor cells have previously been identified in the extract from B.ramiflora (21). The cytotoxic activity against ovarian carcinoma cells was retained in the hexane fraction of the extract (21). The mutant p53-dependent activity that we have identified here is also retained in the organic fraction, in agreement with the notion that one or several of the previously identified terpenes might contribute to this activity. We found that treatment with N37063 induces the p53-regulated genes MDM2 and PUMA. The MDM2 protein targets p53 for proteasome degradation (24). p53-mediated upregulation of PUMA triggers mitochondrial apoptosis (25). Moreover, we could confirm that these genes are also upregulated at mRNA level in a mutant p53-dependent manner. Thus, our results argue that N37063 is able to restore wtp53 function to His175 and His273 mutant p53 proteins, leading to p53-dependent apoptosis. Expression of wtp53 confers sensitivity to N37063 in HCT116 cells. However, that was related only to growth suppression but not to apoptosis or to the induction of apoptosisrelated protein PUMA. Treatment of H1299-His175 cell extract with N37063 did not result in any changes in His175 binding to DNA (data not shown). N37063 treatment enhanced expression of mutant p53 in cells, and this is probably responsible for the observed enhanced DNA binding of mutant p53 in cellular experiments. Thus, transcription-independent mechanisms of mutant p53-dependent apoptosis, as reported for PRIMA-1 (26), should not be ruled out. We have found that N37063 induced ROS preferentially in cells expressing mutant p53. Since induction of cell death after 24 h is a relatively late event compared with the kinetics of ROS induction, it is unlikely that ROS induction is a consequence of cell death. It is plausible that induction of ROS by N37063 is directly related to restoration of wtp53 activity, as previous studies have shown (27– 29). Interestingly, we found that NAC completely blocked the cytotoxic activity of N37063. NAC harbors one free thiol group that endows it with potent antioxidant activity. Thus, NAC captures ROS and disrupts disulfide bonds that could cause protein aggregation. NAC traps reactive alkylating compounds via adduct formation. We have characterized several mutant p53-reactivating compounds and shown that both MIRA-1 (14) and STIMA-1 (19) are thiol-reacting Michael acceptors. We recently demonstrated that PRIMA-1 is converted to methylene quinuclidinone, a compound with Michael acceptor reactivity (18). The mutant p53-dependent apoptotic activity of all these three compounds can be completely blocked by NAC. Our finding that the activity of N37063 is also blocked by NAC suggests that the mechanism of action of N37063, PRIMA-1, MIRA-1 and STIMA-1. Moreover, depletion of sulfhydryl-reactive products from N37063 significantly reduced biological activity of the extract, further arguing in favor of the presence of reactive substances in the extract. However, no reactive groups that could participate in the reactions of nucleophilic addition are present in the structures of the kairatenyl and hopenyl palmitates from N37063 tested here, and NAC was not able to completely inhibit their biological activity. Therefore, it is probably that N37063 contains a number of different compounds with biological activity, including some that have thiol-modifying activity and some that act via other mechanisms. A role of thiol-modifying compounds in N37063 is further supported by our cluster analysis that indicated similarity to the maleimide MIRA-1. The observation that the isolated kairatenyl and hopenyl palmitates were active at the same concentration range as the whole extract indicates that N37063 contains other substances with even greater mutant p53-dependent activity. It is also plausible that some substances in the extract enhance the potency of the kairatenyl and hopenyl palmitates. In conclusion, our database mining approach has yielded a plant extract that contains several mutant p53-reactivating substances. These compounds may serve as leads for the development of efficient and selective mutant p53-targeting drugs and ultimately improved treatment of mutant p53-carrying tumors.

Extract from Asteraceae induces apoptosis

Supplementary material Supplementary Tables I, II and III can be found at http://carcin .oxfordjournals.org/ Funding Swedish Cancer Society (Cancerfonden); Magnus Bergvalls Stiftelse; Karolinska Institutet; EU 6th Framework Program. Acknowledgements

Conflict of Interest Statement: K.G.W. and V.J.N.B. are co-founders and shareholders of Aprea AB, and K.G.W. is a member of its board.

References 1. Riley,T. et al. (2008) Transcriptional control of human p53-regulated genes. Nat. Rev. Mol. Cell Biol., 9, 402–412. 2. Vogelstein,B. et al. (2000) Surfing the p53 network. Nature, 408, 307–310. 3. Levine,A.J. et al. (2006) The P53 pathway: what questions remain to be explored? Cell Death Differ., 13, 1027–1036. 4. Olivier,M. et al. (2002) The IARC TP53 database: new online mutation analysis and recommendations to users. Hum. Mutat., 19, 607–614. 5. Bergh,J. et al. (1995) Complete sequencing of the p53 gene provides prognostic information in breast cancer patients, particularly in relation to adjuvant systemic therapy and radiotherapy. Nat. Med., 1, 1029–1034. 6. Borresen-Dale,A.L. (2003) TP53 and breast cancer. Hum. Mutat., 21, 292–300. 7. Ventura,A. et al. (2007) Restoration of p53 function leads to tumour regression in vivo. Nature, 445, 661–665. 8. Xue,W. et al. (2007) Senescence and tumour clearance is triggered by p53 restoration in murine liver carcinomas. Nature, 445, 656–660. 9. Bykov,V.J. et al. (2002) Restoration of the tumor suppressor function to mutant p53 by a low-molecular-weight compound. Nat. Med., 8, 282–288. 10. Bykov,V.J. et al. (2002) Mutant p53-dependent growth suppression distinguishes PRIMA-1 from known anticancer drugs: a statistical analysis of

Received August 21, 2009; revised April 15, 2010; accepted April 25, 2010

1053

Downloaded from http://carcin.oxfordjournals.org at Karolinska Institutet on June 1, 2010

This publication reflects the author’s views and not necessarily those of the EC. The information in this document is provided as is and no guarantee or warranty is given that the information is fit for any particular purpose. The user thereof uses the information at its sole risk and liability. The Community is not liable for any use that may be made of the information contained herein. We thank Bert Vogelstein, Johns Hopkins Oncology Center, for HCT116 cells; Peter Chumakov, Engelhard Institute of Molecular Biology, Moscow, for H1299-His175 cells; David Kingston, Virginia Polytechnic Institute & State University, Blacksburg, VA, for samples of kairatenyl and hopenyl palmitate, and the Drug Synthesis & Chemistry Branch, Developmental Therapeutics Program, Division of Cancer Treatment and Diagnosis, National Cancer Institute, for plant extracts.

information in the National Cancer Institute database. Carcinogenesis, 23, 2011–2018. 11. Weinstein,J.N. et al. (1997) An information-intensive approach to the molecular pharmacology of cancer. Science, 275, 343–349. 12. Bykov,V.J.N. et al. (2003) Small molecules that reactivate mutant p53. Eur. J. Cancer, 39, 1828–1834. 13. Foster,B.A. et al. (1999) Pharmacological rescue of mutant p53 conformation and function. Science, 286, 2507–2510. 14. Bykov,V.J.N. et al. (2005) Reactivation of mutant p53 and induction of apoptosis in human tumor cells by maleimide analogs. J. Biol. Chem, 280, 30384–30391. 15. Selivanova,G. et al. (1997) Restoration of the growth suppression function of mutant p53 by a synthetic peptide derived from the p53 C-terminal domain. Nat. Med., 3, 632–638. 16. Friedler,A. et al. (2002) A peptide that binds and stabilizes p53 core domain: chaperone strategy for rescue of oncogenic mutants. Proc. Natl Acad. Sci. USA, 99, 937–942. 17. Bass,D.A. et al. (1983) Flow cytometric studies of oxidative product formation by neutrophils: a graded response to membrane stimulation. J. Immunol., 130, 1910–1917. 18. Lambert,J.M. et al. (2009) PRIMA-1 reactivates mutant p53 by covalent binding to the core domain. Cancer cell, 15, 376–388. 19. Zache,N. et al. (2008) Mutant p53 targeting by the low molecular weight compound STIMA-1. Mol. Oncol., 2, 70–80. 20. Bykov,V.J. et al. (2005) Reactivation of mutant p53 and induction of apoptosis in human tumor cells by maleimide analogs. J. Biol. Chem., 280, 30384–30391. 21. Chaturvedula,V.S. et al. (2002) Two new triterpene esters from the twigs of Brachylaena ramiflora from the Madagascar rainforest. J. Nat. Prod., 65, 1222–1224. 22. Cragg,G.M. et al. (2006) Natural product extracts of plant and marine origin having antileukemia potential. The NCI experience. J. Nat. Prod., 69, 488–498. 23. Viera,P.C. et al. (1991) Sesquiterpenoids from Brachylaena hutchinsii. J. Nat. Prod., 54, 416–420. 24. Kruse,J.P. et al. (2009) Modes of p53 regulation. Cell, 137, 609–622. 25. Chipuk,J.E. et al. (2009) PUMA cooperates with direct activator proteins to promote mitochondrial outer membrane permeabilization and apoptosis. Cell Cycle, 8, 2692–2696. 26. Chipuk,J.E. et al. (2003) Pharmacologic activation of p53 elicits Baxdependent apoptosis in the absence of transcription. Cancer Cell, 4, 371–381. 27. Sablina,A.A. et al. (2005) The antioxidant function of the p53 tumor suppressor. Nat. Med., 11, 1306–1313. 28. Johnson,T.M. et al. (1996) Reactive oxygen species are downstream mediators of p53-dependent apoptosis. Proc. Natl Acad. Sci. USA, 93, 11848– 11852. 29. Polyak,K. et al. (1997) A model for p53-induced apoptosis. Nature, 389, 300–305.