The Novel Tubulin-Targeting Agent Pyrrolo-1,5-Benzoxazepine-15 Induces Apoptosis in Poor Prognostic Subgroups of Chronic Lymphocytic Leukemia

B-Chronic lymphocytic leukemia (CLL) is a common leukemia characterized by the accumulation of long-lived CD5-positive monoclonal B cells in the bone marrow, peripheral lymphoid organs, and blood (1). It is believed that CLL is primarily due to defects in apoptotic mechanisms in malignant cells rather than defects in control of lymphoproliferation, although this is the subject of some reevaluation (2). The clinical course of CLL is heterogeneous, with some patients displaying stable disease, which often requires no treatment other than “watchful waiting,” whereas other patients have aggressive disease necessitating early intervention. Several poor prognostic markers have been identified, including presence of unmutated immunoglobulin heavy chain genes (U-IgV_H), cell surface CD38 expression, cytoplasmic zeta-associated protein 70 (ZAP-70) expression, and chromosomal abnormalities such as deletion in 17p (del17p) and del11q23. In contrast, favorable outcome is associated with presence of del13q and mutated IgV_H (M-IgV_H). Current treatment paradigms for CLL involve use of alkylating agents, such as fludarabine, and monoclonal antibodies (mAb; ref. 3).

Pyrrolo-1,5-benzoxazepines (PBOX) are a family of novel compounds, which we have previously shown to induce cell cycle arrest and apoptosis in leukemia cell lines (4, 5) and inhibit tumor growth in an in vivo breast carcinoma model (6). More recently, we have shown that two of the more active members of the family, pyrrolo-1,5-benzoxazepine-6 (PBOX-6) and pyrrolo-1,5-benzoxazepine-15 (PBOX-15), cause depolymerization of microtubules and disassembly of tubulin in vitro (7).

Microtubules, key components of the cytoskeleton, are extremely important in mitosis and cell division and are therefore an important target for anticancer drugs. A chemically diverse group of antimitotic drugs target microtubules and have been used with great success in the treatment of cancer (8). These can be broadly classified either as tubulin polymerizers (e.g., taxanes) or depolymerizers (the Vinca alkaloids, colchicines, and nocodazole). Proliferating cancer cells treated with microtubule targeting agents (MTA) arrest in the G₂-M phase of the cell cycle and eventually undergo apoptotic cell death. Several well-characterized MTAs have been shown to induce apoptosis in CLL cells, including colchicine (9, 10), vincristine (11), and nocodazole (12), as well as novel agents with microtubule targeting activity (13, 14). There seems to be a specific effect of microtubule depolymerization agents in CLL, as Taxol has been reported not to induce apoptosis in CLL cells (12). The mechanism by which MTAs induce apoptosis in CLL is unclear. As circulating CLL cells are noncycling, the mechanism of action cannot be accounted for by induction of mitotic arrest. However, microtubules have been shown to be involved in B-cell receptor signaling via interaction of tubulin with the tyrosine kinase Syk and downstream signaling molecules Cbl and Vav (15). B-cell receptor signaling differs between the IgV_H subgroups of CLL cells, with U-IgV_H cells exhibiting signal transduction following cross-linking of B-cell receptor, whereas cells with M-IgV_H are unresponsive (16). This difference in B-cell receptor signaling may be due to ZAP-70, a critical T-cell receptor–associated tyrosine kinase (17), which is associated with U-IgV_H (18).

In this study, we show that PBOX-15 potently and selectively induces apoptosis in CLL cells. We show that PBOX-15 is equipotent in poor prognostic subgroups of CLL (CD38⁺, ZAP-70⁺, and U-IgV_H genes), including fludarabine-resistant cells with del17p. We investigated the mechanism of action and show that PBOX-15–induced apoptosis occurs via a caspase-8–dependent mechanism. In addition, we suggest a role for c-jun NH₂-terminal kinase (JNK) activation in PBOX-15–induced apoptosis in cells with favorable prognostic indicators (M-IgV_H and ZAP-70⁻).

PBOX-15 [pyrrolobenzoxazepine 4-acetoxy-5-(1-naphthyl)naphtha pyrrolo[1,4]-oxazepine; Fig. 1A] was synthesized as described previously (19, 20).

Fifty-five patients referred to St. James's Hospital, Dublin were recruited to this study. All patients exhibited morphologic and immunophenotypic criteria of CLL (CD5/19⁺, CD23⁺, FMC7⁻, weak surface immunoglobulin, and light chain restriction), had stage A disease, and had not undergone previous treatment. Written informed consent was obtained before blood collection. In a concurrent study (21), we analyzed IgV_H mutational status, karyotype, and CD38 and ZAP-70 expression in a subset of these patients (Table 1). Blood and bone marrow were also obtained from healthy donors. The study was approved by the St. James's Hospital and Adelaide and Meath incorporating the National Children's Hospital Ethics Committee. Peripheral blood mononuclear cells (PBMC) were isolated from herparinized peripheral blood by Lymphoprep (Axis-Shield) density gradient centrifugation and incubated in RPMI 1640 with Glutamax supplemented with 10% FCS, 100 units/mL penicillin, and 100 μg/mL streptomycin (Invitrogen; complete medium). EBV-transformed CLL cell lines, PGA1 (M-IgV_H) and HG3 (U-IgV_H), were obtained from Prof. Anders Rosén (Linköping University, Linköping, Sweden; ref. 23). Cell lines were maintained in complete media without antibiotics.

The colony-forming potential of normal donor bone marrow progenitor cells was assessed using granulocyte-macrophage colony-forming unit (CFU-GM) assays. Bone marrow mononuclear cells were isolated as described above. Mononuclear cells (1 × 10⁶) were incubated with PBOX-15 in 1 mL of complete medium. After incubation, a 200-μL aliquot was removed and added to 1.8 mL of CFU-GM MethoCult methylcellulose medium (StemCell Technologies) and plated in triplicate in 2-cm² cell culture dishes. Colonies with ≥30 cells were counted after 14 d.

An MTT assay (Roche) was used to quantify cell viability. PBMCs (1 × 10⁶) were seeded in 96-well plates, and MTT assay was carried out according to the manufacturer's protocol. IC₅₀ values were calculated using GraphPad Prism version 4.

Apoptosis was quantified by Annexin V binding assays as described previously (24). PBMCs (5 × 10⁶) or PGA1 and HG3 cells (5 × 10⁵) were treated with PBOX-15, 2-fluoroadenine-9-β-d-arabinofuranoside (fludarabine), or nocodazole (Sigma-Aldrich) in 1 mL of complete medium in 24-well plates. Caspase and JNK inhibitors (Calbiochem) were added 1 h before treatment. At the appropriate time point, cells were stained with Annexin V-FITC (IQ Products) and propidium iodide (PI; Molecular Probes) before collection of data via flow cytometry and analysis by CellQuest software (FACSCalibur, BD Biosciences). Apoptosis in normal donor B cells was assessed by incubation of PBMCs with phycoerythrin-labeled CD19 antibody (BD Biosciences) and subsequent analysis of Annexin V binding in CD19⁺ B cells. Mitochondrial integrity was assessed by incubating cells with the membrane potential–sensitive dye rhodamine-123 (Molecular Probes; 50 ng/mL in complete medium) for 30 min and analysis by flow cytometry.

PGA1 cells (4 × 10⁵) were incubated with 1 μmol/L Accell SMARTpool caspase-8 small interfering RNA (siRNA) or Accell nontargeting siRNA pool (Dharmacon) in 1 mL Accell siRNA delivery medium containing 3% FCS in 24-well plates. Following incubation for 72 h, PBOX-15 was added and cells were incubated for a further 16 h before analysis of caspase-8 expression by immunoblotting and of apoptosis induction by Annexin V/PI staining.

Immunoblotting of whole-cell lysates was done as previously described (25). Proteins were resolved by SDS-PAGE and transferred onto polyvinylidene difluoride membrane (Bio-Rad). After probing for Bcl-2 (mouse mAb; Calbiochem), caspase-8 (mouse mAb; Cell Signaling), or phospho-JNK (rabbit mAb; Cell Signaling), blots were incubated with the appropriate horseradish peroxidase–conjugated secondary antibody. All membranes were reprobed with anti-actin antibody to confirm equal loading.

Following treatment, cells were cytospun, air-dried, and fixed in 100% methanol. Fixed cells were stained with anti–α-tubulin mouse mAb (clone DM1A), FITC-conjugated goat anti-mouse secondary antibody (Sigma-Aldrich), and Hoechst 33258 (Molecular Probes) as described previously (26). Projected images from a z-series of images were captured using Olympus 1X81 microscope coupled with Fluoview Ver 1.5 software (Olympus).

Differences in apoptosis between treated and untreated cells were analyzed by Student's t test. Association between prognostic markers and apoptosis was analyzed by Mann-Whitney U test and Fisher's exact test. Correlations were determined by Pearson's method. For all tests, P < 0.05 was considered significant.

Initially, we investigated the cytotoxic effects of PBOX-15 using an MTT assay. As CLL cells are growth arrested in G₀-G₁ (data not shown), this assay quantifies cell viability and not proliferation. Cell viability was determined after treatment with a range of PBOX-15 concentrations (3 nmol/L–10 μmol/L) for up to 72 hours (Fig. 1B), and a sigmoidal dose response allowed cytotoxicity to be expressed as an IC₅₀ value. In this typical result, maximal PBOX-15–induced cytotoxicity was induced after 48 hours; however, there was no significant increase in IC₅₀ values after 24 hours. In 19 patient samples analyzed, IC₅₀ values ranged between 0.18 and 1.1 μmol/L (mean, 0.55 μmol/L; Fig. 1C). We have previously shown that PBOX-15 targets tubulin polymerization in cancer cell lines and T lymphoctyes (7, 27). To determine whether PBOX-15 exerts a similar effect on CLL cells, the microtubule architecture was analyzed. Control cells exhibited an organized microtubule network radiating from the centrosome. However, following treatment with 1 μmol/L PBOX-15 for 24 hours, this microtubule organization was completely disrupted with markedly decreased tubulin staining (Fig. 1D).

Next, we investigated the type of cell death mediated by PBOX-15. CLL cells isolated from 11 patients were treated with 0.1 to 10 μmol/L of PBOX-15 for 24 and 48 hours, and a dose response induction of apoptosis was observed (Fig. 2A). Significant induction of apoptosis was induced in cells treated with 1 μmol/L PBOX-15 for 24 hours, with no further significant increase observed following treatment for 48 hours. Analysis of 46 patients' samples showed that 1 μmol/L PBOX-15 significantly induced apoptosis in CLL cells (Fig. 2B). Although there was a wide range in both spontaneous (3–43%) and PBOX-15–induced apoptosis (22–90%), apoptosis increased in every CLL sample following treatment (range of increase, 9–79%; median increase, 48%). Importantly, PBOX-15 induced similar levels of apoptosis in cells with markers for poor clinical outcome: U-IgV_H (Fig. 2C,i), CD38⁺ (Fig. 2C,ii), ZAP-70⁺ (Fig. 2C,iii), and del17p (Fig. 2C,iv). There was no correlation between spontaneous or PBOX-15–induced apoptosis and karyotypic status, IgV_H mutational status, or CD38 or ZAP-70 expression. PBOX-15 induced apoptosis in two CLL cell lines: PGA1 (M-IgV_H) and HG3 (U-IgV_H; Fig. 2D).

We investigated the effects of PBOX-15 on normal bone marrow donor progenitor cells using a CFU-GM assay. CFU-GM assays are physiologically relevant toxicology models that have a high predictive ability to detect myelosuppression due to chemotherapeutic agents (28). The effects of exposure of bone marrow progenitor cells isolated from three normal donors to 1 and 5 μmol/L PBOX-15 for up to 72 hours are shown in Fig. 3A,i–iii. There was a wide range of colony-forming ability between donor bone marrow progenitor cells following treatment, reflecting the heterogeneity of the donors. Treatment with 1 μmol/L PBOX-15 for 24 hours resulted in no significant decrease in colony-forming ability in any of the three donor samples. Further analysis of another five normal bone marrow donor samples showed that, whereas colony-forming ability was reduced in four of eight samples following 1 μmol/L PBOX-15 treatment for 24 hours, the overall colony-forming ability was not significantly reduced following treatment (Fig. 3A,iv). This relatively low cytotoxicity in normal bone marrow progenitor cells may be due to these cells recovering from PBOX-15–induced cell cycle arrest, resulting in no sustained effect on proliferative capacity. However, these results suggest that prolonged exposure to PBOX-15 may have significant cytotoxic effects on bone marrow progenitor cells. Further studies, including the use of animal models, are required before the bone marrow toxicity of PBOX-15 can be fully assessed. We also investigated whether PBOX-15 exhibited selective apoptosis induction in CLL cells compared with normal B cells. No significant increase in apoptosis was detected in B cells following treatment with 1 μmol/L PBOX-15 for up to 72 hours (Fig. 3B).

As the purine analogue fludarabine is a frontline agent in CLL therapy, we compared the sensitivity of CLL cells to PBOX-15 and fludarabine. Fludarabine-induced apoptosis was assessed using 2-fluoroadenine-9-β-d-arabinofuranoside, the active nucleoside form of fludarabine. A typical result comparing the effects of PBOX-15 and fludarabine shows that 1 μmol/L PBOX-15 is more potent than the therapeutically relevant concentration of 20 μmol/L fludarabine (29) in inducing apoptosis up to 72 hours (Fig. 4A). PBOX-15 at 1 μmol/L was significantly more potent in inducing apoptosis after 24 hours compared with 50 μmol/L fludarabine (P < 0.02, n = 14; Fig. 4B). Del17p results in loss of expression of p53 and is associated with short survival time and resistance to purine analogues, such as fludarabine (30, 31). The three samples with del17p were assessed for fludarabine- and PBOX-15–induced apoptosis after 24 hour treatment (Fig. 4C). Del17p cells exhibited complete resistance to spontaneous and fludarabine-induced apoptosis (levels of apoptosis <10%). However, del17p cells were sensitive to PBOX-15–induced apoptosis with levels of apoptosis ranging from 25% to 63%.

To investigate the mechanism of PBOX-15–induced apoptosis in CLL cells, we investigated the effect of PBOX-15 on mitochondrial membrane depolarization, which is indicative of activation of the intrinsic pathway of apoptosis. We show mitochondrial membrane depolarization at 24 hours, but not at 2 hours, after treatment with PBOX-15 (Supplementary Fig. S1), suggesting that activation of this apoptotic pathway is not an early event in PBOX-15–induced apoptosis.

We have previously shown that proapoptotic PBOX compounds result in phosphorylation and inactivation of the antiapoptotic protein Bcl-2 in a number of cancer cell lines (26, 32). Bcl-2 is phosphorylated in unstressed cells undergoing mitosis; however, hyperphosphorylation of Bcl-2 occurs in response to treatment with MTAs (33), including nonproliferating CLL cells (12, 34). Accordingly, we investigated phosphorylation and expression levels of Bcl-2 in CLL cells following treatment with PBOX-15 (Fig. 5A). No change in protein levels of Bcl-2 was detected in any of four patient samples following 24 hours of treatment with PBOX-15. Treatment of the chronic myelogenous leukemia cell line K562 with PBOX-15 resulted in extensive phosphorylation of Bcl-2; however, no phosphorylation was detected in PBOX-15–treated CLL samples, suggesting that PBOX-15–induced apoptosis occurs independent of Bcl-2.

Previously, we showed that PBOX-6 can induce either caspase-dependent or caspase-independent apoptosis depending on the cell type (4, 25). To investigate the role of caspases in mediating PBOX-15–induced apoptosis in CLL cells, we used a series of specific pharmacologic inhibitors of caspase activity (Fig. 5B). Pretreatment of cells with the pan-caspase inhibitor z-VAD-fmk significantly inhibited PBOX-15–induced apoptosis (n = 21, P < 0.0001), suggesting a requirement for caspase activity. Interestingly, pretreatment with caspase-3 inhibitor, Ac-DMQD-CHO, and caspase-9 inhibitor, Ac-LEHD-CMK, had no significant effect, whereas caspase-8 inhibitor, z-IETD-fmk, significantly abrogated PBOX-15–induced apoptosis (n = 8, P < 0.003; Fig. 5B,i). Pretreatment of the CLL cell line PGA1 with z-VAD-fmk and z-IETD-fmk significantly inhibited PBOX-15–induced apoptosis (Fig. 5B,ii). The tubulin depolymerization agent nocodazole has been shown to induce apoptosis in CLL cells primarily by activation of caspase-9 (12), and here, we show that z-VAD-fmk and Ac-LEHD-CMK significantly inhibit nocodazole-induced apoptosis in PGA1 cells. To confirm the requirement for caspase-8 activation in PBOX-15–induced apoptosis, we used siRNA to knock down expression in PGA1 cells. We achieved partial knockdown of caspase-8 and observed significant inhibition of PBOX-15–induced apoptosis compared with control siRNA–transfected cells (Fig. 5C). In addition, immunoblot analysis of caspase-8 in six patient samples undergoing PBOX-15–induced apoptosis showed down-regulation of full-length procaspase-8 in five CLL samples (nos. 15, 20, 26, 28, and 40) and increased detection of cleaved caspase-8 fragments in three samples (nos. 28, 37, and 40), suggesting PBOX-15–induced activation of caspase-8 in these cells (Fig. 5D).

JNK has been reported to be constitutively activated in CLL (35), and we have previously shown a requirement for early activation of JNK in apoptosis induced by PBOX-6 in several leukemic cell lines (36). To investigate a role for JNK activation in PBOX-15–induced apoptosis in CLL cells, we pretreated cells with the specific JNK inhibitor SP600125 before incubation with PBOX-15 (Fig. 6A). Inhibition of JNK activity significantly inhibited PBOX-15–induced apoptosis (n = 17, P < 0.0005). However, a large heterogeneity in the inhibition of apoptosis was observed. Interestingly, when samples were subgrouped according to IgV_H mutational status and ZAP-70 expression, it was observed that JNK inhibition completely inhibited apoptosis in M-IgV_H and ZAP-70⁻ cells (P < 0.0006 and P < 0.001, respectively) while having little effect on U-IgV_H and ZAP-70⁺ cells (Fig. 6B). As expected, there was a high level of concordance with IgV_H mutational status and ZAP-70 expression (eight of nine patients with M-IgV_H were ZAP-70⁻ and four of six patients with U-IgV_H were ZAP-70⁺). To confirm this role for JNK activation, we used siRNA to knock down expression in PGA1 (M-IgV_H) and HG3 (U-IgV_H) cells; however, we were unable to achieve satisfactory knockdown of JNK using this approach (data not shown). This may be due to our finding that pharmacologic inhibition of JNK induced apoptosis in these cell lines (data not shown), suggesting that JNK may have a role as a survival factor in proliferating CLL cells. We detected activation of JNK following treatment with PBOX-15 in six CLL patient samples by immunoblotting using a phospho-JNK antibody (Fig. 6C). PBOX-15 seemed to selectively activate the p46 JNK isoform because little phosphoylated p54 JNK was detected. Increased activation of JNK was detected in CLL cells that were either U-IgV_H or ZAP-70⁺ compared with cells that were M-IgV_H and ZAP-70⁻.

We have previously shown that PBOX-15, a novel tubulin depolymerization agent, causes disruption of the microtubule network (7) and, similar to another tubulin depolymerization agent, nocodazole, induces mitotic arrest in G₂-M before onset of apoptosis in proliferating cell lines (26). However, microtubules are critical in many cellular processes as well as cytokinesis, and therefore, disruption of microtubule dynamics is likely to result in significant cellular effects. We recently showed that PBOX-15 inhibits integrin-associated T-lymphocyte migration (27), and other MTAs have been shown to have anti-inflammatory properties (13, 37). Microtubules are involved in B-cell signaling (15), and consequently, microtubule disruption may influence B-cell survival. This led us to investigate the activity of PBOX-15 in nonproliferating CLL cells.

We showed PBOX-15–induced destruction of microtubule structure in CLL cells, consistent with our previous findings that PBOX-15 is a tubulin depolymerization agent (7, 27). We observed this loss of microtubule architecture in all CLL cells treated with PBOX-15, even though in some instances the number of cells undergoing apoptosis was less than 50%. This suggests that PBOX-15 disrupts microtubules in CLL cells through direct effects on tubulin and not secondary to apoptosis induction. Indeed, our previous work has shown decreased tubulin polymerization and posttranslational modifications following PBOX-15 treatment in lymphocytes (27).

Here, we show that PBOX-15 induced apoptosis in CLL cells, including those with poor prognostic indicators: U-IgV_H, CD38⁺, and ZAP-70⁺. Of particular note is the sensitivity of del17p cells to PBOX-15, resulting in loss of p53. These cells were resistant to fludarabine-induced apoptosis but exhibited significant apoptosis when exposed to PBOX-15. Cells without del17p underwent fludarabine-induced apoptosis, but were more sensitive to PBOX-15–induced apoptosis. These results suggest that fludarabine-induced apoptosis is p53 dependent, whereas PBOX-15–induced apoptosis is primarily p53 independent. This is a relevant observation in the context of combating fludarabine resistance in high-risk CLL patients.

The mechanism of PBOX-15–induced apoptosis in CLL seems to be different from that of the tubulin depolymerization agent nocodazole. Here, we show that PBOX-15–induced apoptosis is caspase-8 dependent, whereas nocodazole-induced apoptosis is caspase-9 dependent. In addition, nocodazole has been reported to have a differential effect on Bcl-2 in CLL cells: decreasing Bcl-2 expression in some samples and inducing phosphorylation in others after as little as 1-hour exposure (12). However, we detected no changes in expression levels or phosphorylation of Bcl-2 following treatment with PBOX-15, although we cannot discount the possibility of an earlier transient phosphorylation of Bcl-2 (38).

The inhibition of apoptosis by a caspase-8 inhibitor and siRNA knockdown suggests caspase-8 activation as the primary mechanism of PBOX-15–induced apoptosis in CLL. Caspase-8 activation is involved in the death receptor apoptotic pathway; however, earlier evidence has suggested that this pathway is not functional in CLL. CLL cells have been shown to be resistant to CD95/Fas ligand– and tumor necrosis factor–related apoptosis-inducing ligand (TRAIL)–induced apoptosis (39, 40), and activation of caspase-8 has been shown to be low in B-lymphocytic leukemia cells (41). However, more recently, it has been reported that activation of caspase-8 can induce apoptosis and sensitize CLL cells to TRAIL-induced apoptosis (42). Activation of a caspase-8–dependent apoptosis pathway in CLL cells by PBOX-15 is potentially very promising as it may overcome chemoresistance to fludarabine and rituximab, which induce apoptosis via activation of caspase-9 and the mitochondrial pathway (43, 44).

Here, we provide evidence of a role for JNK activation in PBOX-15–induced apoptosis in ex vivo CLL cells. Our results suggest that this PBOX-15 effect is JNK dependent in M-IgV_H and ZAP-70⁻ cells and JNK independent in U-IgV_H and ZAP-70⁺ cells. We have previously shown a role for JNK-mediated phosphorylation and inactivation of antiapoptotic Bcl-2 proteins, which is required for PBOX-6–induced apoptosis in several leukemic cell lines (32). However, the role of JNK in PBOX-15–induced apoptosis in CLL cells seems to be somewhat different: pharmacologic inhibition of JNK resulted in apoptosis in CLL cell lines, and no evidence of PBOX-15–induced Bcl-2 phosphorylation was seen in ex vivo CLL cells.

Although requiring further confirmation, this differential role for JNK activation in high-risk and low-risk CLL is interesting. Although CLL cells have impaired signaling via the B-cell receptor compared with normal B cells, U-IgV_H and ZAP-70⁺ CLL cells have increased B-cell receptor signaling in response to antigen stimulation compared with M-IgV_H and ZAP-70⁻ cells, suggesting that this plays an important role in the clinical behavior of this disease (16, 45). Indeed, downstream JNK, extracellular signal–regulated kinase, and Akt activation has been reported to be more prevalent in CLL cells with U-IgV_H, compared with M-IgV_H, following activation of the B-cell receptor (46), and here, we show greater relative levels of JNK activation in cells that were either U-IgV_H or ZAP-70⁺ compared with cells that were M-IgV_H and ZAP-70⁻. Differences in prosurvival signaling in the subgroups of CLL may be reflected in the role of JNK in PBOX-15–induced apoptosis shown here. JNK plays a complex role in B-cell survival (47); therefore, this role of JNK in CLL may have further therapeutic significance.

Our data extend the anticancer activity of PBOX-15, a novel tubulin-targeting agent, to primary CLL cells, indicating its ability to induce apoptosis in a predominantly nonproliferating population of tumor cells. Importantly, PBOX-15 induces apoptosis in CLL cells with poor prognostic markers, including fludarabine-resistant cells with del17p, indicating its potential as an anticancer agent in chemoresistant disease. Based on these data, novel tubulin-depolymerizing agents with distinct mechanisms of action, such as PBOX-15, may indicate new therapeutic approaches for treatment of CLL and other malignancies characterized by low proliferation rates.

No potential conflicts of interest were disclosed.

Grant support: Enterprise Ireland, Cancer Research Ireland, and the Higher Education Authority of Ireland Programme for Research in Third-Level Institutions.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

We thank Professor Anders Rosén (Department of Clinical and Experimental Medicine, Linköping University, Linköping, Sweden) for the kind gift of PGA1 and HG3 cell lines, and Dr. Orla Hanrahan (School of Biochemistry and Immunology, Trinity College Dublin, Dublin, Ireland) for help with microscopy.