We have previously characterized silvestrol as an inhibitor of translation initiation in vitro and in vivo [4]. To determine how extensive this property is among flavaglines, we tested members of the cyclopenta[b]benzofurans, cyclopent[bc]benzopyrans, and benzo[b]oxepines family for their potential to inhibit protein synthesis. (Fig. 1 and Figs. S1 and S2). Among the compounds tested were rocaglaol (Fig. 1A, compound #4) [18], that has been shown to induce apoptosis and cell cycle arrest in LNCaP cells [7]; aglafolinformylester (compound #7) [19], which has been shown to inhibit cap specific translation initiation (4) and silvestrol, which can sensitize lymphomas to the cytotoxic action of doxorubicin (4). In vitro translation reactions in extracts programmed with the bicistronic mRNA, FF/HCV/Ren (Fig. S1A), revealed that although compound #7 inhibited FF translation to the greatest extent when compared to compounds 1–11 (Fig. 1B and Fig. S1B), it was not as potent as silvestrol in vitro (Figs. 1C) or in vivo (Fig. 1D). The closely related cyclopenta[bc]benzopyrans and benzo[b]oxepines flavagline-type compounds did not show significant activity as inhibitors of translation (Fig. S2). We focused the remainder of our studies on silvestrol since it displayed the most potent inhibition of protein synthesis among the flavaglines that we have identified to date.

We have previously demonstrated that silvestrol acts as a chemical inducer of dimerization – promoting the interaction between eIF4A and RNA [4]. We confirmed that silvestrol increases the RNA binding properties of eIF4A using nitrocellulose binding assays (Fig. 2A). In this assay, ³²P-labeled RNA is retained on nitrocellulose filters only when bound to proteins. Very little RNA was retained by nitrocellulose when RNA was incubated only in the presence of eIF4A (Fig. 2A, compare lane 2 to 1). In the presence of pateamine, a CID that induces eIF4A_f-RNA interaction, a significant proportion of ³²P-labeled RNA was retained (Fig. 2A, compare lane 3 to 2) [20]. Addition of silvestrol to the eIF4A-RNA binding reactions increased the retention of ³²P-labeled RNA on nitrocellulose (compare lanes 5–8 to 2); an event that was inhibited by hippuristanol (compare lane 9 to 7) – a compound that binds to the C-terminal domain of eIF4A to inhibit RNA binding [21]. These results indicate that silvestrol enhances binding between eIF4A_f and RNA.

We have previously shown that RNA binding of eIF4A_c is also increased by silvestrol [4], and we now demonstrated that this was a cap-dependent phenomenon (Fig. 2B). Crosslinking of eIF4E and eIF4A_c from RSW to ³²P-labeled mRNA cap structures was inhibited by the presence of m⁷GDP (Fig. 2B, compare lane 2 to 1). As previously documented [22], crosslinking of eIF4A, but not eIF4E, was ATP-dependent (Fig 2B, compare lane 3 to 1). The presence of silvestrol in the reaction stimulated the crosslinking of eIF4A_c, but not eIF4E (compare lane 4 to 1). Crosslinking of eIF4A_c in the presence of silvestrol was inhibited by m⁷GDP and required ATP (compare lanes 5–7 to 4).

One consequence of increased RNA binding of eIF4A in the presence of silvestrol is that the eIF4F complex could become depleted of eIF4A_c – thus reducing cap-dependent translation. We therefore tested this possibility by purifying eIF4E from RSW containing silvestrol (Fig. 2C; left panel) and from cell extracts that had been prepared from MDA-MB-231 silvestrol-treated cells (Fig. 2C; right panel) using m⁷GTP-Sepharose affinity chromatography. We then probed for the presence of eIF4E and co-purifying eIF4A in m⁷GTP eluates. In both cases, the levels of eIF4A_c in the eIF4F complex were reduced (Figs. 2C).

To determine if components of the eIF4F complex (i.e.-eIF4G) co-sequestered with eIF4A to RNA in the presence of silvestrol, we performed pull-down assays with poly(rG)-agarose (Fig. 2D). Increased eIF4A levels were retained on poly(rG)-agarose in the presence of silvestrol and pateamine, however, this did not result in a concomitant increase of eIF4G associating with poly(rG)-agarose (Fig. 2D, compare lanes 2 and 3 to 1). [The small amount of eIF4G binding to poly(rG)-agarose may be due to this proteins' intrinsic RNA binding.]

Silvestrol inhibited protein synthesis in MDA-MB-231 breast and PC-3 prostate cancer cell lines with approximately the same IC₅₀ (∼60 nM) following a 1 h exposure (Fig. 3A). We next determined the extent to which 25 nM silvestrol would affect protein synthesis rates in these cells. Translation rates were monitored as a function of time-post-exposure to silvestrol by labeling proteins with ³⁵S-Met 15 min before harvest (Fig. 3B). A biphasic response was noted with a precipitous drop occurring over the first 8 h, followed by a slower reduction in translation occurring from 8–72 h (Fig. 3B). The reduction in translation rates was not a consequence of silvestrol-induced apoptosis since this was only observed for MDA-MB-231 cells during the last 24 h of the experiment (40% reduction in viability) (Fig. 3B). We noticed that during the first 8 h following inhibition of protein synthesis by silvestrol there was an ∼2 fold decrease in ³⁵S-met incorporation into newly synthesized proteins for MDA-MB-231 (Fig. S3A: between 0–4 h for MDA-MB-231 cells - compare lane 2 to 1) and PC-3 (Fig. S3A: between 4–8 h for PC-3 cells: compare lane 3 to 2). By 24 h, we noted a reduction in the labeling of specific proteins (Fig. S3A: denoted by open boxes). In contrast, cycloheximide, an inhibitor of elongation, completely blocked protein synthesis and reduced ³⁵S-Met incorporation equivalently in all proteins (data not shown).

These results indicate that treatment of cells with silvestrol might lead to mRNA discrimination during translation. To explore this, we generated reporter constructs in which a G-quadruplex had been engineered into the 5′ untranslated region of the mRNA 6 nucleotides from the site of transcription initiation (Fig. S4A, G-Q(+6)/RL). The G-quadruplex was modeled on one previously described to be present within the 5′ UTR of NRAS RNA and shown to mediate translational repression [23]. A control construct, [CAA]₁₀/RL, generated a reporter transcript having a 5′ UTR consisting of [CAA] tracts and harboring minimal secondary structure with lowered eIF4F dependency [24]. As an internal control, we utilized a construct in which the HCV IRES was driving expression of firefly luciferase, an element known to recruit ribosomes in an eIF4F-independent manner [25]. Translation of G-Q(+6)/RL was more cap-dependent than [CAA]₁₀/RL in Krebs extracts, as judged by inhibition of translation in the presence of m⁷GDP (Fig. S4B). In addition, G-Q(+6)/RL was more sensitive to reduced levels in eIF4A activity, as determined by inhibition of translation in the presence of the eIF4A inhibitor, hippuristanol [26]. In the presence of increasing concentrations of silvestrol, translation of G-Q(+6)/RL was inhibited to a greater degree than [CAA]₁₀/RL mRNA (Fig. S4C), being consistent with inhibition of translation by silvestrol leading to mRNA discriminatory effects. This discriminatory effect was also observed with another flavagline, compound #7 (Fig. 1A), as well as on another set of reporter constructs in which the highly structured HIV TAR element was placed upstream of the chloramphenicol acetyl transferase (CAT) coding region (Fig. S5A). The PLTAR element is inhibitory to translation initiation and renders translation more sensitive to ionic concentrations [27] (Fig. S5B; compare lane 7 to 4). PLTAR CAT mRNA translation proved to be more sensitive to compound #7, than the CAT reporter (Fig. S5B, compare lanes 8–11), indicating that the results obtained with silvestrol are not compound- or reporter- specific (Fig. S4).

We next asked whether exposure of cells to silvestrol would affect the expression of malignancy-related proteins (Fig. 3C). We chose to evaluate the oncogenes cyclin D1 and c-myc and the anti-apoptotic proteins Bcl-2, survivin, and Mcl-1 since these have all been reported to be eIF4E-dependent. Levels of cyclin D1, Bcl-2 and c-myc diminished within the first 8 h of exposing MDA-MB-231 cells to silvestrol (Fig. 3C, compare lanes 1–3). Levels of survivin and Mcl-1 decreased substantially after 48 h, whereas tubulin and GAPDH expression was largely unaffected (Fig. 3C), consistent with the notion that silvestrol selectively affects production of growth-related proteins. Treatment of cells with vehicle for 72 hrs did not change levels of any of these proteins (Fig. 3C, compare lane 7 to 8). The levels of tubulin and GAPDH seen at 48–72 hrs in silvestrol treated cells are also similar to levels observed in untreated cells and thus cannot be attributed to selective enhancement of their synthesis by silvestrol (Fig. 3C, compare lanes 5 and 6 to 1). RT-PCR analysis of RNA isolated from unbound and polysome-bound fractions from DMSO- and silvestrol-treated MDA-MB-231 cells indicated that silvestrol leads to a redistribution of mRNA from polysomes into unbound fractions for these malignancy-related mRNAs (Fig. S3B). This is consistent with silvestrol inhibiting translation initiation of these transcripts.

We have previously shown that eIF2α is not phosphorylated in response to exposure of cells to silvestrol [4], and wanted to assess if inhibition of translation by silvestrol induced stress granule formation, a common event associated with inhibition of translation initiation. Arsenite, a known inducer of SGs caused relocalization of eIF4A and eIF4E into SGs, as determined by co-localization with the SG marker, G3BP (Fig. S6A, B). Exposure of HeLa cells to silvestrol also induced colocalization of eIF4A, eIF4E and G3BP into SGs (Fig. S6).

The cap-dependent translation process is a target of silvestrol. Injection of silvestrol for two consecutive days into mice caused a reduction in liver protein synthesis, as assessed by polysome analysis in liver tissue isolated 3 and 6 h after the second injection (Fig. 4A). These results clearly indicate that silvestrol can inhibit protein synthesis in vivo. We then administered silvestrol to non-tumor bearing mice once a day for 8 consecutive days and assessed if there was an effect on red blood cells, lymphocytes, or monocyte/granulocyte counts from both bone marrow (BM) and spleen (SP) of control and silvestrol-treated mice (Fig. 4B). Vehicle- and silvestrol-treated animals displayed similar blood cell profiles (Fig. 4B). As well, there was no change in body weight of mice treated with silvestrol (Fig. S7) or any signs of illness or distress. Additionally, there was no appreciable change in liver or spleen weight (Fig. 4C). In vivo toxicity can be monitored by the appearance of enzymes in the serum such as ALT and AST. However, rather than an elevation of ALT or AST levels as expected for liver or muscle damage, we observed a slight reduction in AST levels and a 50% decrease in ALT levels, likely reflecting the inhibition of protein synthesis exerted in vivo by silvestrol (Fig. 4C). Collectively, these data indicate that silvestrol appears well tolerated in normal tissues.

The potential consequences of silvestrol on angiogenesis have not been explored, so we evaluated if it could suppress the response of endothelial cells to angiogenic stimuli. Silvestrol inhibited protein synthesis in HUVEC cells (Figs. 5A, B) and did not induce apoptosis even when cells were exposed to compound for 72 h (Fig. 5B). We also evaluated the ability of silvestrol to prevent or delay formation of tube or chord-like structures by HUVECs. Though this in vitro model does not fully recapitulate the features of in vivo angiogenesis, it is none-the-less useful for evaluating the response of endothelial cells to angiogenic stimuli. Silvestrol prevented the ability of HUVECs cultured on Matrigel to form vessel-like structures (Fig. 5C). This was quantitated by counting the number of tubes per field and demonstrated that silvestrol caused a dose-dependent reduction in tube formation (Fig. 5D).

Deregulated eIF4F activity has been postulated to contribute to the oncogenic process in breast cancers [28]. We therefore chose to evaluate the impact of silvestrol administration on tumor growth in MDA-MB-231 breast cancer xenografts (Fig. 6). At 11 days after implantation, silvestrol was administered intraperitonally at 0.5 mg/kg once per day for 8 consecutive days. Following treatment, mice were monitored for up to 2 months and we noted that silvestrol dramatically suppressed the growth of these tumors (Fig. 6A and B). In these experiments, we used two different passages of MDA-MB-231 cells, an early passage that appeared to be slower growing in vivo (Fig. 6A) and a late passage, faster growing, line (Fig. 6B). In contrast to silvestrol treatment, administration of doxorubicin or rapamycin did not inhibit tumor growth (Fig. 6B). However, synergy with doxorubicin was observed for both rapamycin and silvestrol in vitro when tested against MDA-MB-231 cells (Table 1). An additional xenograft model that we used was the PC-3 human prostate cancer model, since the AKT/mTOR signaling pathway is frequently upregulated in these cancers [29], [30]. Growth of PC-3 xenografts was also significantly reduced in mice dosed with 0.5 mg/kg silvestrol for 8 consecutive injections following appearance of the tumors 24 days after implantation (Fig. S8). This observation contrasts with the results obtained with Dox at 5 mg/kg, which showed little effect.

MDA-MB-231 tumors harvested at the end of the study were analyzed to determine the fraction of apoptotic and dividing cells (Fig. 6C). TUNEL staining showed a significant increase in silvestrol-treated tumor cells staining positive for apoptosis and proliferation of silvestrol-treated tumor cells was greatly reduced as judged by staining for Ki67 (Fig. 6C).

All animal studies were approved by the McGill University Faculty of Medicine Animal Care Committee

Silvestrol was resuspended in DMSO and stored at −70°C. Doxorubicin (Sigma) was dissolved in water and stored at 4°C. Rapamycin (LC Laboratories, Woburn, MA) was resuspended in 100% ethanol and stored at −70°C.

Malignant metastatic human mammary epithelial MDA-MB-231 cells were obtained from the American Type Culture Collection (ATCC, Rockville, MD) and grown as monolayers in L15 medium (Invitrogen) supplemented with 10% fetal bovine serum and 100 U/ml penicillin/streptomycin. PC-3 cells were obtained from the American Type Culture Collection (ATCC, Rockville, MD) and grown as monolayers in F12-K medium (Invitrogen) supplemented with 10% fetal bovine serum and 100 U/ml penicillin/streptomycin. HUVEC cells were obtained from Lonza (Walkersville, MD) and propagated in EBM-2 medium supplemented with EGM-2.

The bicistronic reporter, FF/HCV/Ren, has been previously described [46]. For generating reporter constructs containing a G-quadruplex or a [CAA] tract for in vitro translation assays, the plasmid phRL-null was used (Promega). This plasmid was linearized with NheI/NcoI and gel purified. Two sets of annealed oligonucleotides were inserted into these sites. One set [sense: ^5′CTAG[CAA]₁₀C^3′ and antisense: ^5′CATGG[TTG]₁₀ ^3′ ] generated [CAA]₁₀/RL, whereas Set II [sense: ^5′CTAGGGGAGGGGCGGGTCTGGG[CAA]₄C^3′ and antisense: ^5′CATG[GTT]₄CCCAGACCCGCCCCTCCC^3′ ] generated G-Q(+6)/RL. Plasmid G-Q(+6)/RL contains the G quadruplex from NRAS positioned six nucleotides downstream from the T7 RNA polymerase transcription start site [23]. Plasmids encoding CAT or PLTARCAT have been previously described [47].

For in vitro translations, constructs G-Q (+6)/RL and [CAA]₁₀/RL were linearized with XbaI followed by in vitro transciption with T7 RNA polymerase. Plasmid pSP72/HCV/Luc A+ (a gift from Dr. N. Sonenberg [McGill University]) was linearized with BamHI and subsequently in vitro transcribed with T7 RNA polymerase. In vitro translations where performed with 0.2 ng/µl in vitro transcribed RNAs in micrococcal nuclease treated Krebs extract as described previously [46] with the indicated concentrations of silvestrol for 1 h at 30°C. FF and Ren luc activity (RLU) were measured on a Berthold Lumat LB 9507 luminometer. In vitro translations of FF-HCV-Ren in RRL (Promega) were performed according to the manufacturer's instructions.

Filter binding and chemical crosslinking assays were performed as previously described [22], [48]. For crosslinking assays, a 30 µL reaction containing 10 µL ribosomal salt wash (RSW) (1.2 µg/µL) was incubated under standard conditions (25 mM Hepes [pH 7.5], 70 µM GTP, 11 µM of each of the amino acids, 2 mM DTT, 60 µM PMSF and 0.5 mM Mg(OAc)₂) with 0.9 mM ATP (unless indicated otherwise) in the presence of oxidized ³²P-labeled CAT RNA (50,000 cpm). Reactions were incubated for 10 min at 30°C and then crosslinked using 20 mM NaBH₃CN overnight at 4°C. After treatment with RNAse A, proteins were separated by 10% SDS-PAGE and visualized by autoradiography (Kodak X-Omat).

Malignant metastatic human mammary epithelial MDA-MB-231 cells were obtained from the American Type Culture Collection (ATCC, Rockville, MD) and grown as monolayers in L15 medium (Invitrogen) supplemented with 10% fetal bovine serum and 100 U/ml penicillin/streptomycin. PC-3 cells were obtained from the American Type Culture Collection (ATCC, Rockville, MD) and grown as monolayers in F12-K medium (Invitrogen) supplemented with 10% fetal bovine serum and 100 U/ml penicillin/streptomycin. HUVEC cells were obtained from Lonza (Walkersville, MD) and propagated in EBM-2 medium supplemented with EGM-2.

To measure the rate of ³⁵S-Met incorporation into protein, 60,000 cells/well were seeded in a 24-well plate. The following day, the medium was removed, cells washed with PBS and exposed to silvestrol at the indicated concentrations in methionine-free DMEM supplemented with 10% dialyzed serum for 1 hr. For the last 15 min, cells were labeled with ³⁵S-methionine. Medium was removed, cells washed in PBS and lyzed in RIPA buffer (20 mM Tris [pH 7.5], 100 mM NaCl, 1 mM EDTA, 1 mM EGTA, 0.1% NP-40, 0.5% sodium desoxycholate, 0.1% SDS, 20 mM ß-glycerophosphate, 10 mM NaF, 1 mM PMSF, 4 µg/ml aprotinin, 2 µg/ml leupeptin, 2 µg/ml pepstatin) for 20 min with shaking at 4°C. The protein was TCA precipitated and the radioactivity quantitated by scintillation counting. Protein content in the cell lysates was measured using the Bio-Rad D_C ProteinAssay (Bio-Rad Laboratories) and used to standardize the counts obtained by TCA precipitation.

To visualize ³⁵S-methionine labeled proteins, equal amounts of extracts were resolved on 10% SDS-polyacrlamide gels, stained with Coomassie Blue to verify equivalent loading, treated with En³Hance, dried and exposed to X-OMAT X-ray film (Kodak).

To monitor cellular viability, 200,000 cells/well were seeded in a 6-well plate and treated with 25 nM silvestrol for the indicated times. At the end of the treatment, the cell media was collected, cells were washed with 1 ml PBS and trypsinized with 200 µl trypsin. Cells were collected and pooled with the initial media and PBS wash. Samples were spun at 4°C for 2 min at 2000 rpm in a Sorval LegendRT table centrifuge. The pellet was resuspended in 2 ml cold PBS followed by another 2 min spin at 2000 rpm. The pellet was resuspended in 100 µl Annexin V binding buffer (10 mM HEPES [pH 7.4], 140 mM NaCl, 2.5 mM CaCl₂) followed by the addition of propidium iodide to a final concentration of 5 µg/ml. After addition of 5 µl Annexin V-FITC (BD- Biosciences), samples were incubated for 15 min at RT in the dark followed by the addition of 400 µl of Annexin V binding buffer. FACS analyses were performed using a FACScan instrument from BD Biosciences and CELLQUEST software.

For Western blot analysis, cells were grown in 6-well plates, washed with PBS, harvested with a rubber policeman, and collected by brief centrifugation. Cell pellets were lyzed in RIPA buffer and separated on a 10% SDS-polyacrylamide gel followed by transfer to a PVDF membrane (Millipore). Primary Antibodies used were anti-Cyclin D1 (Cell Signaling, Danvers, MA), anti-Bcl-2 (Cell Signaling, Danvers, MA), anti-c-myc (Santa Cruz, Santa Cruz, CA), anti-Mcl-1 (Rockland, Gilbertsville, PA), anti-Survivin (Novus, Burlington, ON), and anti-tubulin (Sigma-Aldrich, Oakville, ON) and anti-GAPDH (Abcam, Cambridge, MA). Secondary antibodies were from Jackson Immuno Research (Burlington, ON).

Anti-eIF4A and anti-G3BP antibodies have been previously described [49], [50]. Anti-eIF4E antibody was a gift of S. Kimball and has been described previously [51]. Cells were processed for immunofluorescence as previously described [50]. Essentially, cells were fixed in 3% paraformaldehyde and permeabilised with 0.1% Triton-X100/PBS. Slides were incubated with primary antibodies diluted in 0.1% normal goat serum for 1 h at RT. Following washing, the slides were incubated with goat anti-mouse/rabbit IgG (H+L) secondary antibodies coupled to goat Alexa Fluor 488/594. Fluorescence microscopy was performed using a Zeiss Axiovision 3.1 microscope equipped with Axiocam HR (Zeiss) digital camera. Images were compiled using Adobe Photoshop software.

RSW was incubated in the presence of 0.5% of DMSO or 50 µM silvestrol for 1 h at 30°C, followed by the addition of 50 µL of 50% m⁷GTP-Sepharose beads (G.E. Healthcare). The reactions were incubated for 2 h end-over-end at 4°C, after which the beads were washed three times with 240 µL of LCB (20 mM HEPES [7.5], 100 mM KCl, 0.2 mM EDTA), and twice with 240 µL of LCB+ 200 µM GDP. Proteins were eluted with 120 µl of LCB+ 200 µM m⁷GTP for 10 min on ice.

For eIF4F pulldown experiments from cell extracts, 8×10⁶ MDA-MB-231 cells were seeded into 15 cm² dishes and the next day treated with 25 nM silvestrol for 4 h. Cells were washed in cold PBS, scraped with a rubber policeman and spun down for 5 min at 2500 rpm. Cell pellets were resuspended in lysis buffer (20 mM Tris [7.5], 100 mM KCl, 1 mM DTT, 1 mM EDTA, 0.2% Tween20, 20 mM ß-glycerophosphate, 10 mM NaF, 1 mM PMSF, 4 µg/ml aprotinin, 2 µg/ml leupeptin, 2 µg/ml pepstatin) and immediately put on dry ice. Following 3 freeze-thaw cycles, extracts were centrifuged for 10 min at 14,000×g to remove cell debris. Pulldowns were performed with 1 mg of total protein extract and 50 µl of 50% m⁷GTP-Sepharose beads (GE Healthcare) for 2 h end-over-end at 4°C. Beads were washed 3 times with lysis buffer, once with LCB+ 500 µM GDP and proteins eluted with LCB+ 500 µM m⁷GTP for 10 min on ice. m⁷GTP elutions were analyzed on a 10% polyacrylamide gel, followed by transfer to a PVDF membrane.

For poly (rG) pulldown experiments, 500 µL of Krebs-2 extract was incubated with 0.5% DMSO, 10 µM pateamine or 50 µM silvestrol for 10 min at 4°C, followed by the addition of 50 µl of 50% Poly(rG) agarose (Sigma). Pulldowns were incubated for 1 h at 4°C. After incubation, the beads were washed twice with 10 volumes wash buffer (20 mM HEPES [7.5], 250 mM KOAc, 0.1% NP-40, 1 mM DTT). SDS elutions were resolved on a 10% SDS-polyacrylamide gel and transferred to PDVF membranes (Millipore). Antibodies used were anti-eIF4A [10] and anti-eIF4GI (Bethyl, Montgomery, TX), respectively. Secondary antibodies were from Jackson Immuno Research (Burlington, ON).

MDA-MB-231 cells were seeded at 10,000 cells/well in a 96- well plate. The cells were allowed to adhere to the bottom of the plates for 24 h in complete media and then exposed to drug or vehicle in fresh media for the indicated periods of time. At the end of treatment, cells were washed with PBS followed by addition of 200 µl PBS to each well. Cell proliferation was monitored using the sulforhodamine B (SRB) assay [52]. Drug interaction was assessed by the combination index method of Chou and Talalay using CalcuSyn software (BioSoft, Cambridge, UK) [53].

For xenograft models, 5×10⁶ MDA-MB-231 or PC-3 cells were injected with matrigel sub-cutaneously into the right flank of 4–6 weeks old female Balb/c nu/nu mice. Tumor growth was monitored every day using calipers. Treatments were started when tumors had reached 25–50 mm³. Nude mice bearing tumors were dosed I.P. with vehicle (5.2% PEG400/5.2% Tween80), silvestrol (0.5 mg/kg), doxorubicin (5 mg/kg), or rapamycin (4 mg/kg). Tumor growth was then monitored for the remainder of the experiment and no further injections were made. For fixation, tumors were fixed in 10% formalin and embedded in paraffin for further analysis. TUNEL staining was performed using the In Situ Cell Death Detection Kit, POD (Roche) following the manufacturer's protocol. For Ki67 staining, the rabbit monoclonal Ki67 (Clone SP6) antibody was purchased from Thermo Scientific. Antigen retrieval was performed by boiling samples for 15 min in 10 mM Citrate buffer [pH 6.5]. Sections were incubated with primary antibody for 1 h at RT. The Ultravision Detection System Anti Rabbit, HRP,DAB (Thermo Scientific) was used according to the manufacturer's instructions.

To assess cytotoxicity of silvestrol, six male wt Balb/c mice, 8 weeks old, were treated with vehicle (5.2% PEG 400/5.2% Tween-80) or 0.2 mg/kg silvestrol for 8 consecutive days. Fresh cell suspensions of bone marrow (BM) and spleen (SP) were prepared in PBS+ 2% FBS. Nonspecific binding was blocked by incubation of the samples with purified anti-CD16/CD32 antibody (clone: 2.4G2; BD Biosciences) for 5 mins on ice before labeling of the cells with a combination of fluorochrome conjugated substrate specific antibodies. Antibodies used to identify monocytes and granulocytes were: Ly-6G/Ly-6C (Gr-1) PECy7 (clone RB6-8C5; Biolegend) and CD11b PE (clone M1/70; BD Biosciences). Antibodies used to identify T and B lymphocytes were: CD4 FITC (clone RM4-5; BD Biosciences), CD8 CyChrome (clone 53-6.7; BD Biosciences) and CD45R/B220 APC (clone RA3-6B2; BD Biosciences). Red blood cells were identified with the antibody Ter119 Biotin (clone ter119; Biolegend) followed by a streptavidin Pacific blue conjugated antibody (Invitrogen/Molecular Probes). Incubation was performed in the dark on ice for 30 min before data acquistion and analysis was conducted on a FACSAria machine (BD Biosciences) by using CELLQUEST (BD Biosciences) or FlowJo (Treestar) softwares.

Polysome profiling analysis on liver extracts was performed on male wt Balb/c mice, 6–8 weeks old. Mice were injected with vehicle (5.2% PEG 400/5.2% Tween-80) or 0.2 mg/kg silvestrol for 2 days. Three and six hours after the second injection, animals were sacrificed, and the livers excised and washed in cold PBS containing 100 µg/ml cycloheximide. The dissected livers were homogenized in a Eurostar Power-b homogenizer (IKA Liver Labortechnik, Staufen, Germany) in 3 volumes of lysis buffer (40 mM HEPES [7.5], 100 mM KCl, 5 mM MgCl₂, 100 µg/ml cycloheximide). Homogenates were then centrifuged for 10 min at 1,200×g at 4°C. The supernatant was transferred into a new tube and to 150 µl of extract, 300 µl of lysis buffer containing 0.5% Triton X-100 and 0.5% sodium deoxycholate were added and the sample centrifuged briefly. Samples were loaded on 10–50% sucrose gradients and centrifuged in an SW40 rotor at 35,000 rpm for 2 h. Gradients were analyzed by piercing the bottom of the tubes with a Brandel tube piercer and passing 60% sucrose through the bottom. Recording of the data was performed using InstaCal Version 5.70 and TracerDaq Version 1.9.0.0 (Measurement Computing Corporation, Norton, MA).

While this manuscript was under review, Lucas et al. (Lucas, DM, Edwards, RB, Lozanski, G, et al. The novel plant-derived agent silvestrol has B-cell selective activity in chronic lymphocytic leukemia and acute lymphoblastic leukemia in vitro and in vivo. Blood, 2009; [Epub ahead of print]) also showed reduction of Mcl-1 protein following silvestrol treatment of acute lymphoblastic and chronic lymphocytic leukemias.