The novel plant-derived agent silvestrol has B-cell selective activity in chronic lymphocytic leukemia and acute lymphoblastic leukemia in vitro and in vivo

Abstract

Therapeutic options for advanced B-cell acute lymphoblastic leukemia (ALL) and chronic lymphocytic leukemia (CLL) are limited. Available treatments can also deplete T lymphocytes, leaving patients at risk of life-threatening infections. In the National Cancer Institute cell line screen, the structurally unique natural product silvestrol produces an unusual pattern of cytotoxicity that suggests activity in leukemia and selectivity for B cells. We investigated silvestrol efficacy using primary human B-leukemia cells, established B-leukemia cell lines, and animal models. In CLL cells, silvestrol LC(50) (concentration lethal to 50%) is 6.9 nM at 72 hours. At this concentration, there is no difference in sensitivity of cells from patients with or without the del(17p13.1) abnormality. In isolated cells and whole blood, silvestrol is more cytotoxic toward B cells than T cells. Silvestrol causes early reduction in Mcl-1 expression due to translational inhibition with subsequent mitochondrial damage, as evidenced by reactive oxygen species generation and membrane depolarization. In vivo, silvestrol causes significant B-cell reduction in Emu-Tcl-1 transgenic mice and significantly extends survival of 697 xenograft severe combined immunodeficient (SCID) mice without discernible toxicity. These data indicate silvestrol has efficacy against B cells in vitro and in vivo and identify translational inhibition as a potential therapeutic target in B-cell leukemias.

Figure 1

Silvestrol cytotoxicity. (A) Silvestrol cytotoxicity in chronic lymphocytic leukemia (CLL) patient cells: CD19⁺ cells from CLL patients (n = 14) were incubated with or without silvestrol at various concentrations for 72 hours, and viability was determined by MTT assay. Viability was calculated relative to time-matched untreated controls. (B) Time dependency of silvestrol cytotoxicity: CLL patient samples (n = 4) were incubated with 0 or 80 nM silvestrol for 4, 24, or 72 hours. At each time point, cells were centrifuged and resuspended in fresh media without drug. Incubations were each continued to a total of 72 hours, at which point viability was determined by MTT assay. Results for each exposure time are shown relative to time-matched untreated cells, set at 100%. Bars show plus or minus SD. (C) Silvestrol cytotoxicity in del(17p13) cells: CLL cells from patients with or without the del(17p13) chromosomal abnormality (n = 7 for each subset) were incubated with silvestrol at 8 nM (approximate LC₅₀) or 1 μM F-ara A, and viability was determined by MTT assay at 72 hours. Results are shown relative to time-matched untreated cells. Bars show plus or minus SD.

Figure 2

Silvestrol produces B-cell selective cytotoxicity. (A) Silvestrol effects on isolated normal B and T lymphocytes: B cells (n = 4, 74%-88% CD19⁺) and T cells (n = 4, 81%-86% CD3⁺) were selected from peripheral blood of healthy volunteers and incubated in media with or without silvestrol. Viability was assessed at 48 hours by MTT assay and results were calculated relative to the time-matched untreated samples in each case. Bars show plus or minus SD. (B) Silvestrol effects on B and T cells in whole blood: Peripheral blood from either CLL patients (left, n = 5) or healthy volunteers (right, n = 4) was incubated at 37°C with mixing, with or without 80 nM silvestrol () or 1 μM F-ara A (). At the indicated times, an aliquot was analyzed by flow cytometry for CD19⁺ cells (striped bars) and CD3⁺ cells (solid bars). Data are shown relative to time-matched untreated samples. Bars represent plus or minus SD.

Figure 3

Silvestrol effects on Mcl-1 expression. (A) Silvestrol mediates caspase-independent Mcl-1 reduction: CLL cells treated with 80 nM silvestrol, 100 μM cycloheximide (CHX), or 25 μM F-ara A for 16 hours with or without caspase inhibitors Z-VAD-fmk or Boc-D-fmk (100 μM) were analyzed for Mcl-1 protein expression by immunoblot. Results are shown from 1 of 4 identical experiments. All samples tested to date showed substantial Mcl-1 protein reduction by this time point after silvestrol treatment. (B) Silvestrol-mediated Mcl-1 reduction in 697 acute lymphoblastic leukemia (ALL) cells is distinct from caspase activity: 697 ALL cell line was incubated for 8 hours with or without 80 nM silvestrol or 200 nM flavopiridol, and lysates were examined by immunoblot. (C) Silvestrol-mediated Mcl-1 reduction is independent of proteasome activity: CLL patient cells were treated with 80 nM silvestrol, 100 μM cycloheximide (CHX), or 200 nM flavopiridol for 4 or 8 hours, with or without 50 nM bortezomib. Lysates were analyzed for PARP and Mcl-1 expression by immunoblot.

Figure 4

Mechanism of silvestrol-mediated Mcl-1 reduction. (A) Effect of silvestrol on Mcl-1 transcription: RNA was extracted from CLL patient cells treated either with 80 nM silvestrol, 200 nM flavopiridol, or 100 μM cycloheximide (CHX) for 4 or 12 hours. n = 6 for untreated and silvestrol treated; n = 3 for flavopiridol and CHX treated. Mcl-1 message level was analyzed by real-time RT-PCR and was normalized relative to 18S RNA. Data are expressed as fold change in Mcl-1 message level over the time-matched untreated sample. Increases in Mcl-1 mRNA with silvestrol treatment were not significant. Bars show plus or minus SD. (B) Translation inhibition by silvestrol: In vitro translation reactions using in vitro–transcribed Mcl-1 mRNA were prepared in the presence or absence of silvestrol, cycloheximide, or flavopiridol as indicated. Reactions were separated by SDS-PAGE and detected using anti–Mcl-1 antibody. Translation assay control reactions contained luciferase mRNA or no mRNA. Lysate from the 697 cell line was included as an Mcl-1 protein control. This figure is from a single immunoblot; a vertical line was inserted to indicate the deletion of an irrelevant lane. (C) Silvestrol does not affect eIF2α phosphorylation: 697 ALL cells were treated with or without silvestrol (80 nM) or arsenic (200 μM) for 2.5 hours and lysates analyzed by immunoblot. HeLa cells were included as control. (D) Silvestrol does not affect 4EBP phosphorylation: 697 ALL cells were serum starved for 3 hours, then were treated with or without silvestrol (80 nM) or rapamycin (20 nM) for 2.5 hours in the presence of 20% fetal bovine serum. Lysates were analyzed by immunoblot.

Figure 5

Apoptotic mechanism of silvestrol. (A) Silvestrol causes increase in proapoptotic Bcl-2 family members in CLL cells: CLL patient cells were treated with silvestrol (80 nm), cycloheximide (CHX, 100 μM), or flavopiridol (200 nM) for 16 hours. Cells were analyzed by immunoblot for Bmf, Bak, and other Bcl-2 family members. Blot is representative of 4 CLL patient samples tested. Numbers below the lanes show densitometry data for Bmf and Bak levels in treated samples relative to untreated, normalized to the actin control. (B) Silvestrol induces reactive oxygen species (ROS) generation and mitochondrial depolarization in CLL cells: CLL patient cells (n = 3) were treated with silvestrol (80 nm) or flavopiridol (200 nM) for 16 hours. Cells were analyzed by flow cytometry using dihydroethidium (DHE) to detect ROS, JC-1 to detect mitochondrial depolarization, and propidium iodide (PI) to detect cell death. Percentages of live cells (PI negative) are shown as . Percentages of cells showing ROS generation (DHE positive) are shown as and percentages of cells with depolarized mitochondria (monomeric JC-1) are shown as ■. Bars show plus or minus SD. (C) Silvestrol-induced mitochondrial depolarization: CLL patient cells (n = 4) were treated as in panel A, with or without the caspase inhibitor Boc-D-fmk (100 μM). Cells were analyzed at 16 hours by flow cytometry using PI and JC-1. The percentages of cells with intact mitochondria are shown as , and PI-negative cells are shown as ■. Bars show plus or minus SD. Boc-D-fmk efficacy was confirmed by reversal of annexin-FITC positivity in treated cells (not shown). Although Boc-D-fmk reduced mitochondrial depolarization in silvestrol-treated CLL cells, this effect did not reach statistical significance relative to the untreated sample. (D) Silvestrol causes reduction in mitochondrial Smac/DIABLO: CLL patient cells treated with 80 nM silvestrol or 200 nM flavopiridol for 16 hours were subfractionated and analyzed by immunoblot. Voltage-dependent anion channel (VDAC) was used as a control for mitochondrial fractionation.

Figure 6

Silvestrol effects on caspase activation. CLL patient cells were treated either with 80 nM silvestrol or 200 nM flavopiridol for 16 hours. Caspases were assessed by immunoblot. Controls are Jurkat cells with or without ultraviolet irradiation. Representative results from 1 of 4 patient samples are shown.

Figure 7

In vivo efficacy of silvestrol. (A) B-cell selectivity of silvestrol in vivo: Eμ-TCL1 mice with leukemia were injected intraperitoneally either with vehicle or 1.5 mg/kg silvestrol (n = 4 each) daily for 5 days for 2 weeks. Percentages of B and T cells were assessed by CD3/CD19 flow cytometry and total lymphocytes by blood smear. Each shape represents 1 animal; circles represent B cells, and triangles represent T cells. Horizontal bars are the average for the group. B-cell reduction in silvestrol-treated mice relative to that in vehicle-treated control mice at week 2 was significant (P = .009). No significant changes in T cells were detected. (B) In vivo efficacy of silvestrol in a xenograft model of ALL: SCID mice that received a transplant of 697 ALL cells were injected intraperitoneally with vehicle only (n = 13) or silvestrol (n = 14) at 1.5 mg/kg every other day, starting 1 week after engraftment. Survival difference between the treated and untreated groups was significant (P = .002). * indicates 3 silvestrol-treated mice survived to 6 weeks after engraftment without symptoms of disease or toxicity. These mice were then followed without treatment for an additional 6 weeks and killed for examination. No evidence of disease or toxicity was found.