Pristimerin induces apoptosis in imatinib-resistant chronic myelogenous leukemia cells harboring T315I mutation by blocking NF-κB signaling and depleting Bcr-Abl

Pristimerin (Figure 1A, purity > 98%, HPLC) was obtained from Paypaytech Inc. (Shenzhen, China). Pristimerin was dissolved in DMSO (Sigma, Shanghai) at a stock concentration of 10 mM, and stored at -20°C. Imatinib (STI571) was purchased from Alexis Biochemicals (Plymouth Meeting, PA). Annexin V-FITC, and mouse monoclonal antibody against actin were from Sigma-Aldrich (Sigma, Shanghai). MG-132 was from Calbiochem (San Diego, CA). Recombinant human TNFα was from Peprotech (Rocky Hill, NJ). Antibodies against p65, IκBα, c-Abl (C-19), proliferating cell nuclear antigen (PCNA), IKKα, cyclin D1 (C-20), apoptosis-inducing factor (AIF, H300), Bax, Bcl-X_L, and Mcl-1 (S-19) were from Santa Cruz Biotechnology (Santa Cruz, CA). Mouse antibodies against poly(adenosine diphosphate [ADP]-ribose) polymerase (PARP, clone 4C10-5), caspase-3, and active caspase-3 (CM1), XIAP, cyclooxygenase-2 (Cox-2), and cytochrome c (clone 6H2.B4) were from BD Biosciences (San Jose, CA). Antibodies against phospho-IKKα (Ser180)/IKKβ (Ser181), phospho-IκBα at Ser32, phospho-p65 at Ser536, phospho-c-Abl at Y245, phospho-Erk1/2 (T202/Y204), Erk1/2, phospho-AKT at Ser473 and AKT were from Cell Signaling Technology (Beverly, MA). Antibodies against phospho-STAT5A/B (Y694/Y699) (clone 8-5-2), STAT5A, and Bcl-2 were from Upstate Technology (Lake Placid, NY). Anti-phospho-RNA polymerase II (S5) was from Bethyl Laboratories (Montgomery, TX); Anti-survivin was from Novus Biol (Littleton, CO). Anti-mouse immunoglobulin G and anti-rabbit immunoglobulin G horseradish peroxidase-conjugated secondary antibodies were from Pierce Biotechnology (Rockford, IL).

Imatinib-sensitive KBM5 cells bearing 210 kDa wild-type Bcr-Abl, were grown in Iscove's modified Dulbecco's medium (Invitrogen, Guangzhou, China) supplemented with 10% heat-inactivated fetal calf serum (FCS), as described previously [22, 23]. Imatinib-resistant KBM5-T315I cells bearing T315I mutation in Bcr-Abl were routinely maintained in the same medium but with 1.0 μM imatinib, which was removed before experiments with a wash-out period of 2-3 days. KBM5 and KBM5-T315I had different sensitivities to imatinib; IC₅₀ values were 0.28-0.53 and 5.04-5.40 μM, respectively [22, 23]. The 32D myeloid cells stably expressing either 210 kDa wild-type Bcr-Abl (32D-Bcr-Abl) or T315I Bcr-Abl (32D-T315I) were established and maintained in RPMI 1640 with 10% FCS as described previously [23]. K562 cell was grown in RPMI 1640 with 10% FCS. U2OS cells were cultured in DMEM with 10% FCS. Cells in logarithmic phase were used in all experiments starting with 2.0 × 10⁵ cells/mL.

Peripheral blood samples were obtained from 7 patients with leukemia [5 CML, 1 JMML (juvenile myelomonocytic leukemia), 1 ALL (acute lymphoid leukemia), Additional file 1, Table S1] and 4 healthy adult donors in Sun Yat-sen Memorial Hospital of Sun Yat-sen University, and The First Affiliated Hospital of Sun Yat-sen University after informed consent according to the institutional guidelines and the Declaration of Helsinki principles. Mononuclear cells were isolated by Histopaque gradient centrifugation (density 1.077 g/mL; Sigma-Aldrich, Shanghai, China). Contaminating red cells were lysed in 0.8% ammonium chloride solution for 10 minutes. After washing with PBS, cells were suspended in RPMI 1640 supplemented with 10% FCS.

U2OS cells (2 × 10⁵) were transfected with reporter plasmids encoding NF-κB-TATA-luc (60 ng) and pEFRenilla-luc (10 ng) or together with plasmids encoding the desired genes by use of LipofectAMINE 2000 (Invitrogen). After 24 hours, cells were left untreated or exposed to pristimerin for 6 hours. The cell supernatants were prepared immediately after stimulation with TNFα (0.1 nM), and luciferase activities were measured with use of dual-luciferase assay kits (Promega, Madison, WI) as described previously [24]. NF-κB activities were determined by normalization of NF-κB-dependent firefly luciferase to Renilla luciferase activity. The pNF-κB-Luc plasmid was from Stratagene (La Jolla, CA). The plasmids pCMV5-IKKα, pCMV5-IKKβ, pCMV5-IKKγ, pCMV5-p65, and pCMV5-myc-TAK1 used for co-transfection experiments were described previously [24].

EMSA involved use of the LightShift Chemiluminescent EMSA kit (Pierce Biotechnology, Shanghai) according to the manufacturer's instructions [24]. The oligonucleotides for NF-κB were from Promega (Shanghai, China): forward, 5-AGT TGA GGG GAC TTT CCC AGG C-3; reverse, 5-G CCT GGG AAA GTC CCC TCA ACT-3. The binding specificity was examined by competition with a 200-fold excess of the unlabeled oligonucleotide probe (cold competitor, Figure 2A, lanes 7 and 14).

Control or drug-treated cells were pelleted by centrifugation and rinsed with PBS. The whole lysates were then prepared with RIPA buffer [22, 25]. The cytosolic fraction was prepared with digitonin extraction buffer to detect the level of cytochrome c and AIF in the cytosol, as described previously [22, 25].

TNFα- and/or pristimerin-treated cells were pelleted by centrifugation and rinsed with PBS. Pellets were then resuspended in 200 μl ice-cold lysis buffer (10 mM Hepes pH 7.9, 10 mM KCl, 0.1 mM EDTA, 0.4% NP 40 with 1 mM DTT, 0.5 mM PMSF, 1 mM sodium orthovanadate and Complete Protease Inhibitor Mix) by pipetting up and down (without bubbling) about 10 times [24, 26]. After incubation on ice for 10 minutes, the lysates were centrifuged at 10,000 g. The supernatants were transferred to fresh tubes and referred to as cytoplasmic extracts. After washing with the lysis buffer, the pellets were vigorously resuspended in nuclear protein extraction buffer with inhibitors (20 mM Hepes, pH 7.9, 0.4 M NaCl, 1 mM EDTA with 1 mM DTT, 0.5 mM PMSF, 0.2 mM sodium orthovanadate and Complete Protease Inhibitor Mix) and centrifuged for 10 minutes at 14,000 g speed at 4°C. The resultant supernatants were kept as nuclear fractions [24, 26].

After being preincubated with or without pristimerin for 6 hours, KBM5 cells were stimulated with TNFα. The cells were harvested and cytospinned onto glass slides. After fixation, and permeabilization, immunofluorescence staining was performed as described previously [27].

DNA-free total RNA was prepared with RNeasy mini-columns, using the on-column DNAase digestion step (Qiagen, Valencia, CA) and reverse transcribed into cDNA following the instruction of Invitrogen SuperScript III First-Stand Kit. Quantitative real-time PCR was performed with Roche LightCycler 480 according to the manufacture's recommended protocol. Optimal reaction conditions for amplification of both Bcr-Abl and 18S were as follows: 40 cycles of a two-step PCR (95°C for 15 s, 60°C for 20 s) after initial denaturation (95°C for 30s) with 5 μl of cDNA reaction by TaKaRa SYBR Premix Ex Taq Kit. The PCR-primers was as follow: Bcr-Abl - sense: 5'-TCCACTCAGCCACTGGATTTAA-3', antisense:

5'-TGAGGCTCAAAGTCAGATGCTACT-3'; Sirt1 - sense:

5'-GAGTGGCAAAGGAGCAGATT-3', antisense:

5'-ATGTAACGATTTGGTGGCAA-3'; 18S - sense:

5'-AAACGGCTACCACATCCAAG-3', antisense:

5'-CCTCCAATGGATCCTCGTTA-3'. The mRNA expression of Bcr-Abl or Sirt1 was normalized as described previously [28]. The results of three independent experiments are reported.

Small interfering RNA (siRNA) duplexes against p65 were from Cell Signaling Technology (Beverly, MA); ON-TARGETplus SMARTpoolsiRNA duplexes against human PDGFRa or ON-TARGETplus Non-Targeting Pool siRNA control were from Dharmacon RNA Tech. (Lafayette, CO) [27]. Transfection of either anti-p65, PDGFRα or the control synthesized siRNA duplexes into K562 cells was performed with the Cell Line Nucleofector Kit T (Amaxa, Gaithersburg, MD) and program O-17 according to the manufacturer's instruction [27]. Twenty-four hours after siRNA transfection, the cells were analyzed with Western blotting.

Cell viability was determined by MTS assay (CellTiter 96 Aqueous One Solution Cell Proliferation assay; Promega, Madison, WI) [22, 24]. An amount of 100 μl cells (2.0 × 10⁵/mL) were seeded in 96-well plates and incubated with various concentrations of pristimerin for 72 hours. Four hours before culture termination, 20 μL of MTS solution was added to each well. Absorbance was read on a 96-well plate reader at a wavelength of 490 nm. Control cells received DMSO (< 0.1%) containing medium. The drug concentration resulting in 50% inhibition of cell growth (IC₅₀) was determined.

To evaluate anchorage-independent growth of tumor cells, CML cells (2 × 10⁵/ml) were treated with increasing concentrations of pristimerin or diluent (DMSO, control) for 24 hours, then washed with PBS; the same amounts of viable cells (as determined by trypan blue exclusion assay and a hemocytometer) were subsequently seeded in 6-well plates in Iscove's medium containing 0.3% agar and 20% FCS in the absence of drug treatment. After incubation for 10 ∼ 14 days at 37°C, colonies composed of more than 50 cells were counted using an inverted phase-contrast microscope, as described previously [22, 24].

Male nu/nu BALB/c mice were bred at the animal facility of Sun Yat-sen University. The mice were housed in barrier facilities with a 12-h light dark cycle, with food and water available ad libitum. 3 × 10⁷ of KBM5-T315I cells were inoculated subcutaneously on the flanks of 4- to 6-week-old male nude mice. Tumors were measured every other day with use of calipers. Tumor volumes were calculated by the following formula: a² × b × 0.4, where a is the smallest diameter and b is the diameter perpendicular to a. Pristimerin was dissolved in DMSO and diluted with vehicle [30% Cremophor EL/ethanol (4:1), 70% PBS]. Mice in the control group received the same amount of vehicle treatment. The body weight, feeding behavior and motor activity of each animal were monitored as indicators of general health. The animals were then euthanized, and tumor xenografts were immediately removed, weighed, stored and fixed. All animal studies were conducted with the approval of the Sun Yat-sen University Institutional Animal Care and Use Committee.

Formalin-fixed xenografts were embedded in paraffin and sectioned according to standard techniques. Tumor xenograft sections (4 μm) were immunostained using the MaxVision kit (Maixin Biol, Fuzhou, China) according to the manufacturer's instructions [22, 23].

GraphPad Prism 4.0 software (GraphPad Software, San Diego, CA) was used for statistical analysis. Comparisons between 2 groups involved two-sided Student's t test, and comparisons among multiple groups involved one-way ANOVA with post-hoc intergroup comparisons using Tukey test. P < 0.05 was considered statistically significant.

We first examined whether pristimerin affected the TNFα-induced NF-κB-dependent reporter gene transcription. One day after cotransfection with pNF-κB-TATA-Luc and pEFRRenilla-Luc, U2OS cells were exposed to pristimerin at increasing concentrations for 6 hours or a fixed concentration (200 nM) for different durations. Prior to the termination of culture, TNFα was added for 10 minutes. The luciferase activity detected was increased by TNFα (Figure 1B); but pristimerin inhibited the TNFα-induced NF-κB reporter activity in a dose- and time-dependent manner (Figure 1B and 1C).

In the canonical NF-κB activation pathway, TAK1 and IKK are the major upstream regulators of IκBα. To determine the steps at which pristimerin acted, U2OS cells were cotransfected with plasmids to express IKKα, IKKβ or IKKγ, along with an NF-κB-TATA-Luc reporter plasmid. The luciferase activity of NF-κB-TATA-Luc reporter was significantly increased when cotransfected with p65, IKKα, IKKβ, or IKKγ constructs (Figure 1D) compared with transfection with reporter alone. However, addition of pristimerin significantly inhibited the NF-κB transcriptional activity (Figure 1D). Therefore, pristimerin could block NF-κB activation induced by IKK overexpression.

Because TAK1 is critical upstream regulator of IKK [29], we assessed the effect of pristimerin on cotransfection of a TAK1 construct along with NF-κB-TATA-Luc reporter plasmid. TAK1 significantly elevated NF-κB reporter luciferase activity (Figure 1E), and pristimerin significantly blocked TAK1-induced NF-κB activation.

We next examined whether pristimerin interfered with the binding of NF-κB to DNA by EMSA. KBM5 cells were preincubated with or without 200 nM pristimerin for 6 hours; TNFα was added for the indicated times, then nuclear extracts were assayed for NF-κB DNA binding activity by EMSA with a probe representing an NF-κB-binding site. After stimulation with TNFα, the levels of the NF-κB-DNA complex were steadily increased over time in the absence of pristimerin (Figure 2A, lanes 2-5 versus lane 1). With the same durations of stimulation with TNFα, NF-κB-DNA complex were not formed in the presence of 200 nM pristimerin (Figure 2A, lanes 8-12). Competition with an excess (200-fold) of unlabeled probe led to disappearance of the TNFα-induced bound complex (Figure 2A, lanes 7 and 14), which confirmed the binding specificity of this assay. Pretreatment for 6 hours with increasing concentrations of pristimerin abrogated TNFα-induced NF-κB-DNA complex formation in a dose-dependent manner (Figure 2B).

To address whether pristimerin exerted a direct inhibitory effect on the binding of NF-κB to DNA, nuclear extracts prepared from untreated KBM5 cells or cells stimulated with TNFα were incubated in a cell-free reaction system in the presence of increasing concentrations of pristimerin. Pristimerin, even at a concentration as high as 20 μM, did not block the TNFα-induced formation of NF-κB-DNA complex (Figure 2C), suggesting that pristimerin did not directly interfere with binding of NF-κB to DNA in a purified nuclear extract.

The findings that pristimerin inhibited TNFα-induced NF-κB transcriptional function and formation of an NF-κB-DNA binding complex predicted that pristimerin would affect IκBα phosphorylation and p65 translocation. To test this hypothesis, KBM5 or KBM5-T315I were pretreated with 200 nM pristimerin for 6 hours, then stimulated with TNFα; cytoplasmic and nuclear extracts were examined by Western blot analysis with antibodies against total and phosphorylated forms of TAK1, IKKα/β, IκBα and p65, respectively. In the absence of pristimerin (Figure 3A, left, control), the levels of phosphorylated TAK1 and phosphorylated IKKα/β in the cytoplasmic fraction of KBM5 cells were appreciably increased starting at 5 minutes after TNFα stimulation as compared with untreated cells. The levels of phosphorylated IκBα appeared as a pulse signal that peaked at 5 minutes after TNFα stimulation and then decreased back to baseline (Figure 3A) as previously reported by Anand et al. [30]. Subsequently (starting 5 minutes after TNFα stimulation), the levels of total IκBα protein in the cytoplasmic fractions decreased, which was consistent with activation of ubiquitination and proteasomal degradation triggered by the phosphorylation of IκBα [31]. In contrast, pristimerin completely abolished the phosphorylation of IKKα/β and IκBα induced by TNFα (Figure 3A, left, pristimerin). Accordingly, the TNFα-induced IκBα degradation was abrogated by pristimerin. The levels of phosphorylated and total p65 in the nuclear fractions were increased in TNFα-stimulated cells (Figure 3B, left, control). The increase in nuclear p65 was markedly abrogated by the pristimerin (Figure 3B, left, pristimerin). Similar findings were obtained in KBM5-T315I cells (Figure 3A and 3B, right panels). Further, pristimerin time- and dose-dependently abrogated the TNFα-induced IκBα degradation and p65 translocation (Figure 3C). The inhibitory effect of pristimerin on TNFα-induced translocation of p65 was further confirmed by immunofluorescent microscopy (Figure 3D).

Because IαBκ phosphorylation is a critical step to tag the protein and trigger ubiquitination-proteosome-dependent degradation of IκBα [32], we next examined whether pristimerin inhibited the degradation of IκBα via inhibiting IκBα phosphorylation. TNFα stimulation elevated the levels of phosphorylated IκBα in KBM5 cells (Figure 3E), which was potentiated by the presence of the proteosome inhibitor MG-132. However, pristimerin completely blocked the TNFα-induced phosphorylation of IκBα even in the presence of MG-132.

It is believed that the pro-survival function of NF-κB is mediated through NF-κB-regulated expression of cell survival proteins whose genes containing an NF-κB binding site in their promoters (e.g., Bcl-X_L, survivin, Mcl-1, cyclin D1, c-myc, Bcl-2, and Cox 2) [30]. Because pristimerin inhibits TNFα-induced activation of NF-κB, we hypothesized that it would also inhibit TNFα-induced expression of cell survival proteins. K562 cells were treated with or without 400 nM pristimerin for 6 hours, followed by stimulation with TNFα for up to 24 hours. The levels of Bcl-X_L, survivin, Mcl-1, cyclin D1, c-myc, Bcl-2, and Cox 2 as detected by immunoblotting were increased after TNFα stimulation (Figure 3F), and the TNFα-induced upregulation of these genes was abolished by pristimerin (Figure 3F).

We also examined the effect of pristimerin on the expression of Bcr-Abl in CML cells by immunoblotting. Twenty-four-hour treatment with pristimerin dose- and time-dependently decreased levels of Bcr-Abl protein in KBM5, KBM5-T315I or K562 cells (Figure 4A). Bcr-Abl can activate multiple signal transduction pathways (e.g., CrkL, STAT5, Erk1/2, and PI3K/AKT) which can promote cell proliferation and resistance to apoptosis [3]. We reasoned that the loss of total Bcr-Abl protein would lead to a decrease in its kinase activity. We measured autophosphorylation by immunoblotting with the phospho-specific antibody against the phosphorylated Y245 c-Abl as described previously [33]. Basal phosphorylation of Bcr-Abl was detectable in KBM5, KBM5-T315I and K562 cells (Figure 4A). As expected, pritstimerin potently decreased the amount of phosphorylated Bcr-Abl in accord with decrease in the total Bcr-Abl protein (Figure 4A). To better define the effect of pristimerin on Bcr-Abl signaling, we followed the phosphorylation of Bcr-Abl and its downstream targets over time. The levels of Bcr-Abl and phosphorylated Bcr-Abl started to decrease as early as 3 hours after pristimerin treatment, and continued to decrease up to 24 hours after pristimerin treatment (Figure 4B). Concomitantly, the phosphorylation of CrkL, STAT5, Erk1/2 and AKT decreased (Figure 4B). The total protein levels of STAT5 and AKT did decrease at later time points. Taken together, pristimerin decreased Bcr-Abl and inhibited its downstream signaling. Notably, the decrease in Bcr-Abl protein levels preceded the onset of apoptosis, as indicated by specific PARP cleavage and caspase-3 activation after pristimerin treatment (Figure 4C).

To investigate whether the proteosome pathway was involved in the pristimerin-mediated downregulation of Bcr-Abl, we pretreated K562 cells with a sub-cytotoxic concentration (0.5 μM) of the proteosome inhibitor MG-132, but found no effect on the decrease in Bcr-Abl protein levels (Figure 4D). Similarly, the pristimerin-induced decrease of Bcr-Abl protein was not rescued by the presence of caspase-3 inhibitor z-DEVD-fmk (data not shown).

We next examined whether pristimerin inhibited Bcr-Abl at the transcriptional levels. KBM5 and KBM5-T315I cells were exposed to various concentrations of pristimerin for 24 hours (left) or 600 nM for different durations (right); reverse transcription quantitative real-time PCR (RT-qPCR) revealed that the mRNA levels of Bcr-Abl were decreased after treatment with pristimerin (Figure 4E) while the mRNA levels of Sirt1 gene (an unrelated gene control) were not decreased, suggesting a selective activity of pristimerin against Bcr-Abl transcription.

The possibility of interdependence of the two actions of pristimerin (i.e., NF-κB inactivation and depletion of Bcr-Abl) was tested. First, we examined whether the TNFα-induced NF-κB activation was dependent on Bcr-Abl. We pretreated KBM5 cells with 1.0-3.0 μM imatinib for 1 hour to inactivate Bcr-Abl, the cells were then stimulated with TNFα (0.1 nM, 5 minutes), followed by analysis of NF-κB activation. Imatinib treatment led to dephosphorylation of Bcr-Abl, which, however, did not abrogate the TNFα-induced IκBα phosphorylation, and subsequent relocation of p65 (Figure 4F). Second, we employed the siRNA approach to determine the effect of p65 on Bcr-Abl in CML cells. The levels of p65 were appreciably decreased in K562 cells 24 hours after transfection with siRNA duplexes specifically against p65 when compared with control siRNA and PDGFRα siRNA (unrelated gene control) (Figure 4G). Silencing p65 did not affect the levels of Bcr-Abl. The above results together implicate that NF-κB inactivation and Bcr-Abl inhibition may be parallel independent pathways.

We then evaluated the effect of pristimerin on the growth of CML cells. Human CML cell lines (KBM5, KBM5-T315I and K562), and a pair of murine myeloid cells 32D stably transfected with either the wild-type or T315I human Bcr-Abl were exposed to escalating concentrations of pristimerin for 72 h, followed by the MTS assay. Cell viability of all 5 lines of CML cells was inhibited, with IC₅₀ values of 199 nM, 135 nM, 450 nM, 242 nM and 387 nM, respectively (Figure 5A). These data suggest similar sensitivities of imatinib-sensitive and imatinib-resistant CML cells to pristimerin.

Because clonogenicity is believed to better reflect malignant behavior of tumor cells, we compared the impact of pristimerin on clonogenicity in CML cells with that of imatinib. We examined the clonogenicity of KBM5 and KBM-T315I cells after a 24 hour-treatment of imatinib. The IC₅₀ values of KBM5 and KBM5-T315I to imatinib were 0.45 μM and 4.10 μM, respectively (data for dose-response curves not shown). The results confirmed the resistance of KBM5-T315I cells to imatinib. In contrast, pristimerin treatment potently inhibited the number of surviving clonogenic KBM5, KBM5-T315I and K562 cells in a dose-dependent manner, with IC₅₀ values of 43.7 nM, 35.9 nM and 36.5 nM, respectively (Figure 5B).

We next ascertained the activity of pristimerin against imatinib-resistant primary CML cells. Mononuclear cells in peripheral blood from 1 CML patient who acquired clinical resistance to imatinib versus 4 CML patients that responded to imatinib were exposed to various concentrations of pristimerin for 72 hours. The primary CML cells were sensitive to pristimerin as well, with IC₅₀values 6-578 nM (Table 1). In particular, the primary leukemia cells from 1 CML patient with acquired clinical resistance to imatinib were also sensitive to pristimerin (IC₅₀value 578 nM). Of interest, primary leukemia cells from 1 patient with JMML (juvenile myelomonocytic leukemia) and 1 patient from ALL (acute lymphoid leukemia) were also sensitive to pristimerin (Table 1).

We also examined the effect of pristimerin on normal bone marrow cells. Normal bone marrow cells collected from 4 healthy donors were treated with increasing concentrations of pristimerin for 72 h; cell viability assayed by MTS indicated that pristimerin was cytotoxic to normal bone marrow cells with a median IC₅₀ value of 740.5 nM (range, 681-999 nM). These data were in line with the report by Costa et al. [19] who showed that pristimerin inhibited the cell viability of peripheral blood mononuclear cells with an IC₅₀ value of 880 nM. Therefore, bone marrow suppression may be a significant side effect of pristimerin.

We next evaluated the in vivo activity of pristimerin against imatinib-resistant CML cells using the nude mouse xenograft model. Twenty six nu/nu BALB/c mice were injected with KBM5-T315I cells. Five days after inoculation of tumor cells, when the size of tumor reached approximately 50 mm³, mice inoculated with KBM5-T315I cells were randomized to receive intratumoral injection with either 50 μL vehicle [30% Cremophor EL/ethanol (4:1), 70% PBS] or 1.0 mg/kg pristimerin in vehicle daily during days 6-19 after inoculation of KBM5-T315I cells. The growth curves (the estimated tumor size calculated from the tumor dimension versus time) are shown in Figure 5C. Pristimerin potently inhibited the growth of KBM5-T315I tumors. The weight of tumors was significantly lower in the treated group than in the control (vehicle) group (Figure 5D, P < 0.0001). Immnunohistochemical analysis with an anti-c-Abl antibody (to detect both c-Abl and Bcr-Abl) revealed that c-Abl immunoreactivity was markedly decreased by pristimerin treatment (Figure 5E). These data again pointed to the antitumor activity of pristimerin.

After exposing CML cells to various concentrations of pristimerin for 24 hours, cell cycle analysis was conducted by using flow cytometry with propidium iodide staining. The results revealed no significant cell cycle alteration in CML cells treated with various concentrations of pristimerin for 24 hours except for the appearance of the sub-G1 apoptotic population (Figure 6).

The ability for pristimerin to induce apoptosis in CML cells was assessed by flow cytometry after staining with Annexin V and propidium iodide. After treatment with pristimerin apoptosis was induced in a dose- and time-dependent manner (Figure 7A, 7B) in the CML cells. In parallel, Western blotting analysis revealed that pristimerin induced a dose-dependent specific cleavage of PARP, which is a hallmark of apoptosis (Figure 7C). Concomitantly, pristimerin led to a decline of the precursor form of caspase-3, reflecting its activation.

Most chemotherapeutic agents induced apoptosis by triggering release of cytochrome c and AIF from mitochondria into the cytosol. To evaluate the apoptosis pathway activated by pristimerin, CML cells were exposed to pristimerin, cytochrome c and AIF in the cytosolic fraction was examined by Western blotting at different time points. Cytochrome c and AIF were undetectable in the cytosol of control cells, but were released from the mitochondria into the cytosol after pristimerin treatment (Figure 7D). The results implied that pristimerin triggered the mitochondrial pathway of apoptosis. Pristimerin activated the same apoptotic pathway in breast cancer cells as reported by Wu C et al. [21] The effect of pristimerin on antiapoptotic proteins (Bcl-2, Bcl-X_L, Mcl-1, survivin) of the Bcl-2 gene family regulating the mitochondrial pathway of apoptosis was also examined. Pristimerin treatment decreased the levels of Bcl-X_L and Mcl-1 in a dose- and time-dependent manner without affecting Bcl-2, XIAP, Bax and survivin (Figure 7E and 7F).