Romidepsin and Belinostat Synergize the Antineoplastic Effect of Bortezomib in Mantle Cell Lymphoma

Published OnlineFirst January 12, 2010; DOI: 10.1158/1078-0432.CCR-09-1937 Published Online First on January 12, 2010 as 10.1158/1078-0432.CCR-09-1937

Cancer Therapy: Preclinical Clinical Cancer Research Romidepsin and Belinostat Synergize the Antineoplastic Effect of Bortezomib in Mantle Cell Lymphoma

Luca Paoluzzi1, Luigi Scotto3, Enrica Marchi3, Jasmine Zain3, Venkatraman E. Seshan2, and Owen A. O'Connor3

Abstract Purpose: Romidepsin and belinostat are inhibitors of histone deacetylases (HDACI). HDACIs are known to induce cell death in malignant cells through multiple mechanisms, including upregulation of death receptors and induction of cell cycle arrest. They are also known to be prodifferentiating. Mantle cell lymphoma (MCL) is an aggressive subtype of non–Hodgkin lymphoma characterized by the t(11;14) (q13;q32) translocation leading to the overexpression of cyclin D1. Experimental Design: Assays for cytotoxicty including mathematical analysis for synergism, flow- cytometry, immunoblottings, and a xenograft severe combined immunodeficient beige mouse model were used to explore the in vitro and in vivo activity of romidepsin and/or belinostat alone or in combination with the proteasome inhibitor bortezomib in MCL. Results: In vitro, romidepsin and belinostat exhibited concentration-dependent cytotoxicity against a panel of MCL cell lines. Both HDACI showed strong synergism when combined with the proteasome inhibitor bortezomib in MCL. An HDACI plus bortezomib also induced potent mitochondrial membrane depolarization and apoptosis, whereas no significant apoptosis was observed in peripheral blood mononuclear cells from healthy donors with the combination. These events were associated with a decrease in

cyclin D1 and Bcl-XL, and an increase in accumulation of acetylated histone H3, acetylated α-tubulin, and Noxa in cell lines. In a severe combined immunodeficient beige mouse model of MCL, the addition of belinostat to bortezomib enhanced efficacy compared with either drug alone. Conclusions: Collectively, these data strongly suggest that HDACI such as romidepsin or belinostat in combination with a proteasome inhibitor could represent a novel and rationale platform for the treatment of MCL. Clin Cancer Res; 16(2); 554–65. ©2010 AACR.

Mantle cell lymphoma (MCL) represents a distinct sub- ubiquitin E3-ligase that targets p27 for proteolytic degra- type of aggressive lymphoma, characterized by gross cell dation, may contribute to the loss of cell cycle inhibition cycle dysregulation. The pathognomonic lesion in the dis- (3). The collective importance of these genetic lesions has ease is the reciprocal translocation t(11;14)(q13;q32) lead- been supported by additional lines of data showing the ing to the juxtaposition of the cyclin D1 gene downstream importance of proliferation in MCL. Data by Rosenwald of the immunoglobulin heavy chain gene promoter (1). and colleagues (4) identified a “proliferation signature” This genetic event leads to constitutive overexpression of in MCL, in which the patients with the most highly prolif- cyclin D1. In addition, many forms of MCL exhibit loss erative disease experienced the poorest prognosis, whereas of the cyclin-dependent kinase (cdk) inhibitors p27 and patients with the least proliferative disease exhibited a su- p21, leading to further dysregulation of cell cycle control perior survival. This biology has also been validated by Ki- (2). Although the precise mechanism explaining the loss 67 immunohistochemical staining, which also correlated of p27 remains to be clarified, several lines of evidence overall survival with the proliferative index in MCL (5) suggest that overexpression of Skp2, a component of the These data suggest that by merely “lowering” the proliferative signature of the disease, it might be possible to alter the natural history and overall prognosis of the disease. Authors' Affiliations: 1Herbert Irving Comprehensive Cancer Center, In addition to the lesions involved in cell cycle control, Columbia University; 2Biostatistics Shared Resources, Herbert Irving Comprehensive Cancer Center; and 3NYU Cancer Institute, NYU it is also clear that aberrant expression of various Bcl-2 Langone Medical Center, New York, New York family members, including Mcl-1, and deficiencies in Note: Supplementary data for this article are available at Clinical Cancer BH3-only proteins such as Bim and Noxa may render Research Online (http://clincancerres.aacrjournals.org/). these cells relatively resistant to apoptosis (6). Pharmaco- Corresponding Author: Owen A. O'Connor, NYU Langone Medical Cen- logic strategies that directly address these specific lesions ter, 522 First Avenue, Smilow Room 1101, New York, NY 10016. Phone: represent an innovative strategy for targeting “targeted 212-263-9884; Fax: 212-263-9210; E-mail: owen.o'[email protected]. therapy” to a specific disease context. doi: 10.1158/1078-0432.CCR-09-1937 The proteasome inhibitor bortezomib was approved for ©2010 American Association for Cancer Research. the treatment of relapsed or refractory MCL based on the

554 Clin Cancer Res; 16(2) January 15, 2010

HDAC Inhibitors in Mantle Cell Lymphoma

through, vorinostat, a hydroamic acid analogue, was the Translational Relevance first and, to date, only HDACI approved for the treatment of cancer (cutaneous T-cell–lymphoma; refs. 24–26). Al- Mantle cell lymphoma (MCL) is a relatively rare, although it has been difficult to ascribe a singular mecha- – though challenging, form of non Hodgkin's lympho- nism of action to the HDACI in any specific biological ma. Recently, several promising new drugs with context, the biological effect of these drugs has been important single-agent activity have begun to teach shown to include cell cycle arrest, terminal differentiation, us about the unique biology of this disease. The com- and induction of apoptosis, depending on the concentra- bination of a proteasome inhibitor and HDAC inhibition of drug and the cell model studied. tors represents a novel and theoretically highly targeted The immediate rationale for combining these two clas- drug combination suited to address the unique patho- ses of drugs in MCL revolves around recent observations genetic lesions indicative of MCL. In this article, we that HDACIs, such as vorinostat and LBH-589, may be show the remarkable synergy these two classes of drugs able to “turn off” cyclin D1. Ellis et al. (27) showed in a have in MCL, establishing the fact they are able to phase 1 experience that LBH-589 consistently produced “ ” “ ” turn off cyclin D1 expression and turn on NOXA downregulation of cyclin D1 in punch biopsies of patients in vitro in vivo in a manner that is synergistic and . These with cutaneous T-cell lymphoma treated with LBH-589 by data establish the class effects of these agents providing gene expression array. Similarly, vorinostat, trichostatin A, a clear-cut rational for moving these combinations in- and even the aliphatic acids have been shown to reduce to clinical trials for patients with MCL. In addition, intracellular cyclin D1 protein in MCL cell lines, but not they provide insights into possible biomarkers of activ- in cell lines of acute myeloid leukemia. These observations ity that could be explored in future clinical studies. were shown to be dependent on concentrations of drugs that also induced accumulation of acetylated histone H3 (28, 29). Although the precise mechanism of this downregulation remains to be established, the observation creates consistent observations from independent phase 2 studies a unique opportunity to both turn off cyclin D1 and “turn (7–10) including one registration-directed phase 2 study on” cdk inhibitors such as p21 and p27 in a disease char- (11). In addition, bortezomib has been shown to comple- acterized by these very lesions. Belinostat and romidepsin ment a host of other targeted drugs including Bcl-2– are both pan-HDACIs that now in advanced stage clinical targeted agents such as obatoclax mesylate, ABT-737, and trials. Given the activity of proteasome inhibitors in MCL, oblimersen sodium (7, 12–14). Although there are a mul- and its ability to enforce accumulation of cdk inhibitors, titude of biological effects induced by inhibition of cellu- we sought to explore the merits of this combination in lar proteasome, accumulation of the cell cycle–dependent models of MCL as well as in normal peripheral blood kinase inhibitors such as p27, p21, p53; inhibition of NF- mononuclear cells (PBMC) from healthy donors. κB; and accumulation of proapoptotoic proteins such as Noxa have been consistently shown in many cell line Materials and Methods models (15). The acetylation and deacetylation of histones regulate Cells and cell lines. HBL-2, Jeko-1, and Granta-519 are transcriptional activation by mediating chromatin conden- well-characterized MCL lines (30, 31). Mononuclear cells sation and decondensation (16). The acetylation status of from PBMC samples of healthy donors were purchased histones and nonhistone proteins is determined by the from Allcells. All cell lines were grown as described previ- balance between histone acetyl transferase (HAT) and his- ously (12, 13). tone deacetylase (HDAC) activity. In addition to histone, Materials. All reagents for Western blotting were ob- at least 50 nonhistone proteins have been documented to tained from Bio-Rad Laboratories and Pierce Biotechnol- be substrates for the enzymatic reactions mediated by his- ogy, Inc.; DMSO was obtained from Sigma. Drugs were tone acetyl transferases and HDACs (17–20). For example, obtained as follows: romidepsin was from Gloucester nonhistone protein targets of HDACs include transcrip- Pharmaceuticals, Inc.; belinostat was from TopoTarget; tion factors (such as p53, c-Myc, Bcl-6, NF-κB), transcrip- and bortezomib was from the institutional research tional regulators (such as Rb), signal transduction pharmacy. mediators (STAT3), DNA repair enzymes, nuclear import Cytotoxicity assays. For all in vitro assays, cells were regulators, chaperone proteins (HSP90), structural pro- counted, incubated, and processed as described previ- teins (α-tubulin), and mediators of inflammation (21). ously (12, 13). Romidepsin and belinostat were diluted Maintaining acetylation of histone as a therapeutic ratio- in DMSO that was maintained at a final concentration nale in cancer was considered after the discovery of the of <0.5%. Romidepsin and belinostat were added at HDAC inhibitor (HDACI) trichostatin A as early as 1990 concentrations from 1 nmol/L to 10 μmol/L. For com- (22). Since then, many different HDACIs have been iden- bination experiments with bortezomib, the final concen- tified and developed, including noncancer-related drugs trations of each drug were selected to approximate the not previously recognized as having HDACI activity IC10-60 for each drug. For all cytotoxicity experiments, (23). Although many of these agents are progressing Cell-Titer-Glo Reagent (Promega Corp.) and a Synergy

www.aacrjournals.org Clin Cancer Res; 16(2) January 15, 2010 555

Paoluzzi et al

HT Multi-Detection Microplate Reader (Biotek Instru- separated into treatment groups of 9 to 10 mice each. Tu- ments, Inc.) were used as described previously (12, 13). mors were assessed using the two largest perpendicular ax- Each experiment was done at least in duplicate and repeated es (l, length; w, width) as measured with standard calipers. at least twice. Data are presented as averages ± SD. Tumor volume was calculated using the formula 4/3r3, Flow cytometry. Cells were seeded at a density of 6 × 105/ where r =(l + w)/4. Tumor-bearing mice were assessed mL and incubated with concentrations of romidepsin or for weight loss and tumor volume at least thrice weekly. belinostat with or without bortezomib at concentrations Animals were sacrificed when one-dimensional tumor approximating the IC10-30 for 24 h. To determine changes diameter exceeded 2.0 cm, or when the tumor volume in the transmembrane mitochondrial membrane potential exceeded 2,000 mm3 in accordance with institutional (Δψm), cells were stained with 1.25 μg/mL of JC-1 dye guidelines. Complete response was defined as nonpalp- (Invitrogen) for 30 min at 37° in normal growth medium, able tumor. In a first series of experiments, mice not bear- then washed once in warm PBS, resuspended in 200 μLof ing tumors were treated with belinostat plus bortezomib medium, and analyzed using a FACSCalibur Flow Cyt- to explore the toxicity of the combination. Belinostat was ometer (BD). Carbonyl cyanide m-chlorophenylhydra- administered by i.p. injection at a dose up to 40 mg/kg/d zone (Sigma Immunochemicals) was used as a positive for 7 d [highest dose level explored in the study from control for loss of membrane potential. A minimum of Plumb et al. (ref. 32)]. Bortezomib was administered by 3×105 events were acquired from each sample. Median i.p. injection at 0.5 mg/kg on days 1, 4, 8, and 11. In a values obtained from the FL1 and FL2 channels after stan- subsequent xenograft experiment, belinostat was adminis- dard gating of forward and side scatter were used to calcu- tered at 35 mg/kg/d for 7 d and bortezomib was given by late the normalized Δψm. For detection of apoptosis, i.p. injection at 0.5 mg/kg on days 1, 4, 8, and 11. Belino- Yo-Pro-1 and propidium iodide (PI) were used (Invitro- stat was added to a mixture of 10% DMSO and 90% gen). After incubation with romidepsin or belinostat plus sterile water. Control groups were treated with the vehicle or minus bortezomib, cells were washed and resus- solution. pended in cold PBS. One microliter of Yo-Pro-1 and 1 μL Statistical analysis. The IC50s were calculated by fitting of PI were added to each 1 mL of cell suspension. The dose-response curves (33). Confidence intervals (CI) are fluorescence signals were acquired by a FACSCalibur Sys- shown in between parenthesis. Drug-drug interactions tem. All data from flow cytometry were analyzed with were computed using the relative risk ratio (RRR) analysis the Flowjo software. Each experiment was done at least (GraphPad),4 with a RRR of <1 defining synergism, a RRR in duplicate and repeated at least twice. Data are pre- of 1 defining additivity, and a RRR of >1 defining antago- sented as averages ± SD. nism. Data from the in vitro assays were analyzed using a t Western blot analysis. Cells were incubated with the ap- test with a robust variance estimate. For the mouse experi- proximate IC10-30 of each drug alone (bortezomib, romi- ments, the tumor volumes and area under the curves were depsin, and belinostat) and in combination (HDACI ± log transformed and evaluated using ANOVA for four-way bortezomib) under normal growth conditions for 24 h. comparison and Wilcoxon test for pairwise comparisons. Proteins from total cell lysates were resolved on 12% to Median absolute deviation was used as a measurement of 20% SDS-PAGE and transferred to nitrocellulose mem- variability. All significance testing was done at the P < 0.05 branes. Membranes were blocked in phospho-buffered sa- level. line and 0.05% Triton X-100 containing 5% skim milk powder, and were then probed overnight with specific pri- Results mary antibodies. Antibodies were detected with the corresponding horseradish peroxidase–linked secondary Romidepsin and belinostat interact synergistically with antibodies. Blots were developed using SuperSignal West bortezomib in MCL cell lines. The IC50 of romidepsin and Pico chemiluminescent substrate detection reagents. The belinostat after a 24 hours of exposure across a panel of membranes were exposed to X-ray films for various time MCL cell lines was in the 10 to 100 nmol/L range for romi- intervals. The images were captured with a GS-800 cali- depsin, and 1 to 100 μmol/L range for belinostat (Fig. 1A-B). brated densitometer (Bio-Rad), and the ratios were quan- For HBL-2, Jeko-1, and Granta-519, the IC50 values for ro- tified by densitometric analyses within the linear range of midepsin were 4.3 nmol/L (CI, 3.6-5.2), 11 nmol/L (CI, each captured signal. The monoclonal and polyclonal 6.7-18.2), and 58.5 nmol/L (CI, 26-131), respectively. For antibodies used were as follows: cyclin D1, acetylated his- belinostat, the corresponding IC50 values for the same lines α tone H3, acetylated -tubulin 1, Bcl-XL (all from Cell Sig- were 0.4 μmol/L (CI, 0.3-0.7), 0.2 μmol/L (CI, 0.1-0.7), and naling Technology), Noxa (Imgenex Corp.), and β-actin 56.3 μmol/L (CI, 30.7-103.5), respectively. (clone AC-74, Sigma). Formal synergy analyses were done using HBL-2, Jeko-1, Mouse xenograft models. In vivo experiments were done and Granta-519 cells treated with different concentrations as described previously (12, 13). In brief, 6- to 8-wk-old of romidepsin or belinostat (approximating the IC10-60)in beige mice with severe combined immunodeficiency combination with bortezomib at 3, 3.5, or 4 nmol/L for (SCID; Charles River Laboratories) were injected with 1×107 HBL-2 cells in their flanks through a s.c. route. 3 When tumor volumes approached 50 mm ,micewere 4 http://www.graphpad.com

556 Clin Cancer Res; 16(2) January 15, 2010 Clinical Cancer Research

HDAC Inhibitors in Mantle Cell Lymphoma

Fig. 1. Bortezomib interacts synergistically with romidepsin and belinostat in MCL cell lines. A, cytotoxicity after exposure to romidepsin or (B) belinostat for 24 h in three cell lines of MCL.

24 hours. In all cell lines, a synergistic cytotoxic effect was 80% (P = 0.001) in Jeko-1 (Fig. 3A). The combination of observed when combining either HDACI with bortezomib belinostat plus bortezomib increased the same percentage through a range of concentrations. The RRR analysis up to 80% in both cell lines (P = 0.002-0.003 for HBL-2, showed very strong synergism (RRR < 0.3) in virtually all P = 0.001 for Jeko-1; Fig. 3B). combinations of HDACI and bortezomib (Table 1A-B). Romidepsin or belinostat plus bortezomib enhance disrup- The observed RRR were as follows: belinostat + bortezo- tion of Δψm in MCL cell lines. Changes in Δψm represent mib: RRR ≤ 0.7 in HBL-2, RRR ≤ 0.5 in Jeko-1, and an early event in the induction of apoptosis, and likely RRR ≤ 0.4 in Granta-519; romidepsin + bortezomib: capture the effects of agents on various aspects of Bcl-2 RRR ≤ 0.8 in HBL-2, RRR ≤ 0.8 in Jeko-1, and RRR ≤ family members. Treatment of Jeko-1 and HBL-2 cells with 0.1 in Granta-519 (Fig. 1B). The combination of an romidepsin (5-6 nmol/L) or belinostat (200-600 nmol/L) HDACI plus bortezomib at concentrations approximating plus bortezomib (3.5 nmol/L) decreased the normalized the IC10-60 for each drug resulted in a cell viability rang- Δψm in a time-dependent manner (Fig. 4A). After incuba- ing between 9% and 30% in Jeko-1, 2% and 40% in HBL- tion with either romidepsin or belinostat plus bortezomib 2, and 3.5% and 18% for Granta-519. Figure 2 presents the for 16 hours, both cell lines did not show any significant cell viability data after treatment with the single agents or change in Δψm compared with the single agents and con- the combination with a fixed concentration of bortezomib trol. After 20 and 24 hours of incubation, the combination (3.5 or 4 nmol/L depending on the cell line). All the ex- groups in both cell lines showed more than 60% and 80% plored combinations showed impressive cytotoxicity with decrease in Δψm, respectively, compared with the untreat- no more than 20% to 25% cell viability at 24 hours in all ed controls. This level of effect was comparable with that three cell lines. The corresponding RRRs revealed very observed after treatment with the positive control carbonyl strong synergism in all cases (RRR ≤ 0.3). cyanide m-chlorophenylhydrazone. In particular, for ro- Romidepsin or belinostat plus bortezomib enhance apopto- midepsin + bortezomib, the P values in HBL-2 were sis in MCL cell lines. The combination of romidepsin or ≤0.003 (20 hours) and ≤0.002 (24 hours), and ≤0.002 belinostat (IC10-40) and bortezomib (IC10-20) for 24 hours (20 hours) and ≤0.001 (24 hours) in Jeko-1. For belino- showed potent induction of apoptosis in HBL-2 and Jeko-1 stat + bortezomib, the P values in HBL-2 were ≤0.003 (20 cell lines. When HBL-2 or Jeko-1 cells were treated with ro- hours) and ≤0.004 (24 hours), and ≤0.002 (20 hours) and midepsin (2.5-6 nmol/L) or belinostat (100-600 nmol/L) ≤0.001 (24 hours) in Jeko-1 (Fig. 4A). These data are con- plus bortezomib (3-3.5 nmol/L), statistically significant sistent with the cytotoxicity data that support the conten- apoptosis was observed in all the combination groups tion that a strong synergistic effect of the HDACI plus compared with the individual drugs and the control in bortezomib requires relatively longer durations of expo- both cell lines (P ≤ 0.007 for HBL-2, P ≤ 0.001 for sure (up to 24 hours). Jeko-1; Fig. 3). The combination of romidepsin plus bor- Romidepsin or belinostat plus bortezomib do not enhance tezomib increased the percentage of apoptotic plus dead apoptosis to PBMC from healthy donors. Concentrations cells up to about 70% in HBL-2 (P = 0.001-0.007) and of romidepsin and belinostat that produced synergy in

www.aacrjournals.org Clin Cancer Res; 16(2) January 15, 2010 557

Paoluzzi et al

combination with bortezomib in cell lines of MCL were two MCL cell lines (24-hour exposure). Following treat- also tested in PBMC from healthy donors to explore po- ment with romidepsin at 4 or 6 nmol/L plus bortezomib, tential toxicity against healthy cells. The observed synergis- the expression of cyclin D1 was significantly decreased in tic effect in cell lines was specific to malignant cells, since HBL-2, whereas a similar effect was observed in Jeko-1 af- the combination of romidepsin (2.5-5 nmol/L) or belino- ter treatment with romidepsin at 6 nmol/L with or with- stat (400-600 nmol/L) plus bortezomib (3-5 nmol/L) did out bortezomib. Belinostat decreased the expression of not show any significant apoptosis compared with the sin- cyclin D1 in HBL-2, producing a near complete absence gle agents alone and the control (P ≥ 0.82). Apoptotic plus of this protein in the combination at 600 nmol/L. In dead cells were ∼40% in all groups (Fig. 4B). Jeko-1, the combination of belinostat at 400 or 600 Influence of HDACIs plus bortezomib on proteins involved nmol/L plus bortezomib and the group treated with beli- in cell cycle regulation and apoptosis. Supplementary Fig. S1 nostat alone at 600 nmol/L produced a significant de- and Fig. 5 show the results related to the immunoblottings crease in cyclin D1 compared with all other groups. An for cyclin D1, acetylated histone H3, acetylated α-tubulin, increase in histone H3 acetylation was observed after Bcl-XL, Mcl-1, and Noxa before and after treatment with treatment with romidepsin or belinostat alone or in com- romidepsin or belinostat plus or minus bortezomib in bination with bortezomib. In Jeko-1, the combination

Table 1. RRR analysis in three cell lines of MCL after treatment with bortezomib plus belinostat (A) or romidepsin (B) at different concentrations for 24 h

A. RRR analysis in three cell lines of MCL after treatment with bortezomib plus belinostat at different concentrations for 24 h Belinostat (μmol/L) Bortezomib (nmol/L) 3 3.5 4

HBL-2 0.06 0.69 0.37 0.09 0.12 0.30 0.11 0.04 0.25 0.61 0.11 0.08 Granta-519 1.2 0.38 0.25 0.23 2.5 0.09 0.09 0.07 5 0.08 0.05 0.10 10 0.22 0.12 0.08 Jeko-1 0.01 0.39 0.34 0.49 0.03 0.29 0.32 0.40 0.07 0.24 0.17 0.25 0.15 0.19 0.16 0.25

B. RRR analysis in three cell lines of MCL after treatment with bortezomib plus romidepsin at different concentrations for 24 h

Romidepsin (nmol/L) Bortezomib (nmol/L)

3 3.5 4

HBL-2 1 0.22 0.81 0.39 2 0.09 0.04 0.04 3 0.09 0.07 0.03 Granta-519 10 0.12 0.10 0.07 20 0.08 0.07 0.002 30 0.09 0.07 0.06 40 0.14 0.09 0.05 Jeko-1 1 0.75 0.31 0.41 2 0.43 0.10 0.14 3 0.48 0.17 0.12 4 0.56 0.21 0.16

NOTE: All ratios are <1 (synergy).

558 Clin Cancer Res; 16(2) January 15, 2010 Clinical Cancer Research

HDAC Inhibitors in Mantle Cell Lymphoma

Fig. 2. Bortezomib interacts synergistically with romidepsin and belinostat in MCL cell lines. Cytotoxicity after exposure to bortezomib (B)at3or 3.5 nmol/L plus belinostat (P) or romidepsin (R) for 24 h in three cell lines of MCL. RRRs for belinostat: 0.16 to 0.32 in Jeko-1 (A), 0.05 to 0.25 in Granta-519 (B), and 0.04 to 0.21 in HBL-2 (C). RRRs for romidepsin: 0.1 to 0.21 in Jeko-1 (A), 0.07 to 0.1 in Granta-519 (B), and 0.03 to 0.39 in HBL-2 (C). Cytotoxicity was evaluated by luminometric acquisition after cell TiterGlo assay; columns, mean of experiments repeated at least twice; bars, SD. groups with romidepsin (4 or 6 nmol/L) or belinostat zomib compared with the other groups in Jeko-1, where- (400 or 600 nmol/L) showed some increased accumula- as in HBL-2, some decrease was observed in the groups tion of acetylated H3 compared with romidepsin or beli- treated with belinostat alone or in combination. The ex- nostat alone, respectively. In HBL-2, the combination of pression of the antiapoptotic protein Mcl-1 did not pro- belinostat at 400 nmol/L with bortezomib showed an induce any significant change in either cell line. Noxa, a crease in acetylated histone H3 compared with the con- BH3-only proapoptotic protein that counteracts the anti- trol, belinostat, and bortezomib alone. The abundance apoptotic effect of Mcl-1, accumulated after treatment of the antiapoptotic protein Bcl-XL showed some decrease with romidepsin or belinostat plus bortezomib in both after treatment with romidepsin at 6 nmol/L plus borte- cell lines. In Jeko-1, significant accumulation of Noxa

www.aacrjournals.org Clin Cancer Res; 16(2) January 15, 2010 559

Paoluzzi et al

was also observed in the groups treated with belinostat mib was statistically superior to bortezomib alone (P ≤ alone. The expression of acetylated α-tubulin showed an 0.009), belinostat alone (P ≤ 0.035), and the control increase after treatment with belinostat plus or minus (P ≤ 0.003; Table 2A-B). One mouse in the combination bortezomib in both cell lines. In HBL-2, there was a trend group experienced a complete response on day 8 that toward significant increase of Noxa in the combination was lost by day 12, whereas an additional mouse experi- groups with romidepsin compared with the drugs given enced a complete response by day 12. Of note, no animals alone or the control. In Jeko-1 the combination with ro- in any of the single agent or control groups experienced a midepsin at 6 nmol/L seemed to slightly increase this complete remission. Thirty percent of the mice in the com- protein compared with the other groups. bination group experienced significant weight loss by day Belinostat enhances the activity of bortezomib in vivo. 12 (one mouse from day 3, one from day 8, and one from When belinostat was given at 40 mg/kg/d in combination day 12), but no deaths were observed. with bortezomib at 0.5 mg/kg, 1 out 10 mice died on day 7 after receiving the last dose of treatment of drug, and 2 additional mice experienced significant weight loss (>10% Discussion of the initial weight). A dose reduction of ∼10% (35 mg/ kg/d) was subsequently explored and found to be better The increasing clarification of the distinct molecular fea- tolerated with only one mouse experiencing significant tures of different NHL subtypes, coupled with an ever ex- weight loss on day 8, which fully recovered by day 12. panding understanding of the molecular pharmacology of The in vivo efficacy of belinostat at 35 mg/kg/d for 7 days many new drugs, has created a unique opportunity to for- in combination with bortezomib was investigated in a xe- mulate new concepts about the application of novel tar- nograft model of MCL (HBL-2; Fig. 6A). Statistical analysis geted drugs in cancer. The tailoring of these new agents (days 3, 8, and 12) from the beginning of the experiment based on their pharmacologic effects to specific disease revealed that the combination of belinostat and bortezo- contexts represents a rational strategy for developing novel

Fig. 3. Enhanced apoptosis of bortezomib (B) combined to an HDACI in MCL cell lines. A, treatment of HBL-2 and Jeko-1 cells with romidepsin (R) plus bortezomib; P ≤ 0.004, P ≤ 0.007, P ≤ 0.002 for combination group versus bortezomib alone, romidepsin alone, or control, respectively. B, treatment of HBL-2 and Jeko-1 cells with belinostat (P) plus bortezomib; P ≤ 0.001, P ≤ 0.002, P ≤ 0.002 for combination group versus bortezomib alone, belinostat alone, or control, respectively. Apoptosis was evaluated by fluocytometric analysis of Yo-Pro-1 and PI. Columns, mean; bars, SD. *, the comparison of the combination group to the correspondent single drugs and control.

560 Clin Cancer Res; 16(2) January 15, 2010 Clinical Cancer Research

HDAC Inhibitors in Mantle Cell Lymphoma

Fig. 4. The combination of an HDACI and bortezomib significantly changes the Δψm in MCL cell lines without effect on PBMC. A, HBL-2 and Jeko-1 cells were incubated with romidepsin (R) at 5 or 6 nmol/L and belinostat (P) at 200 or 600 nmol/L plus or minus bortezomib at 3.5 nmol/L for up to 24 h. Both combination groups were statistically significant to any of the single group and control (P ≤ 0.004) after 20 and 24 h but not after 16 h in both carbonyl cyanide m-chlorophenylhydrazone (CCCP; B) PBMCs were treated with bortezomib at 3 or 5 nmol/L and/or romidepsin at 2.5 or 5 nmol/L or belinostat at 400 or 600 nmol/L for 24 h. Each combination group did not shown significant apoptosis compared with any single drug given alone (P ≥ 0.82). Apoptosis was evaluated by cytofluorimetric analysis of Yo-Pro-1 and PI. Columns, mean; bars, SD. *, statistical significance for the combination groups compared with the correspondent single drugs and control.

platforms for treatment. MCL is a great example of a dis- 20, 35–37). The biological effects of HDACIs seem to de- ease that has been largely redefined over the past few pend on a variety of factors, including the nature of the years. Understanding the molecular pathogenesis of the HDACI, its concentration and duration of exposure, and disease, coupled with an appreciation of the molecular importantly, the cellular context. At least 12 different pharmacology of the new agents, has created the opportu- HDACI are undergoing clinical trials in patients with he- nity to develop rational combinations of agents in this matologic and solid tumors. Vorinostat (suberoylanilide disease-specific context. hydroxamic acid; 38) was the first-in-class agent approved MCL is a disease characterized by gross cell cycle dysre- by the Food and Drug Administration for the treatment of gulation. At the core of this dysregulation is the constitu- cutaneous T-cell lymphoma (26). Belinostat and romidepsin tive overexpression of cyclin D1, and the enhanced are two new drugs now in registration-directed clinical trials, proteolytic degradation of cdk inhibitors such as p27 both of which are pan-Class I and II HDACIs (39). and p21. Lesions in apoptosis related to an imbalance in A third rational that supports the potential combination Mcl-1 and BH3-only mimetic proteins also contribute to of these drug classes revolves around the role of HDAC6. an elevated apoptotic threshold. Although HDACIs and HDAC6 is a class II HDAC enzyme that deacetylates α- proteasome inhibitors may be associated with a plethora tubulin, leading to increased cell motility. Importantly, of biological effects, the observation that the former can several investigators have shown that the induction of turn off cyclin D1, and the latter force accumulation of HDAC6 by proteasome inhibitors such as bortezomib cdk inhibitors, create a rational platform for the combina- may also account for some of the acquired drug resis- tion of these two drug classes in MCL. tance seen in cell lines treated with proteasome inhibi- Supporting the rationale further is the observation that tors, (40) establishing the notion that inhibition of both proteasome and HDACIs have been shown to have HDAC6 may circumnavigate the development of resis- therapeutic value in the treatment of lymphoid malignan- tance to proteasome inhibitors such as bortezomib. cies (7, 24–26, 34). In addition to affecting cell cycle reg- HDAC6 plays an essential role in aggresomal protein ulatory proteins, HDACIs have been shown to induce cell degradation, a proteasome-independent pathway that death by activating the intrinsic and extrinsic pathways of eliminates misfolded polyubiquitinated proteins. HDAC6 apoptosis, inducing mitotic catastrophe and autophagic can bind both polyubiquitinated proteins and dynein cell death, and generating reactive oxygen species (19, proteins, recruiting protein cargo to dynein motors that

www.aacrjournals.org Clin Cancer Res; 16(2) January 15, 2010 561

Paoluzzi et al

Fig. 5. Modulation of proteins involved in apoptosis and cell cycle before and after treatment with romidepsin (R) plus or minus bortezomib (B). Cyclin D1 (CyD1), acetylated histone 3 (Ac-HH3), acetylated α-tubulin (Ac-α Tub), BCL-XL, Mcl-1, and Noxa expression before and after treatment with romidepsin (R) at 4 or 6 nmol/L and bortezomib (B) at 3.5 nmol/L was analyzed by Western blot. β-Actin was used to normalize protein loading.

transports misfolded proteins to aggresomes (41). Over- dependent pathways with bortezomib and the aggresome expression of HDAC6 leads to deacetylation of tubulin pathway in tumor cells with a HDACI induces greater accu- and increases cell motility (42, 43). Specific inhibition mulation of polyubiquitinated proteins resulting in in- of HDAC6 activity or its downregulation has been shown creased cellular stress and apoptosis. In addition to these to increase tubulin and HSP90 acetylation, which reduces effects of HDACI on the aggresome, some studies restricted cellular motility, and induces HSP90 client protein degra- to in vitro models of leukemia have suggested that HDACIs dation, cell growth inhibition, and cell death (44). Inhi- can induce apoptosis through inactivation of antiapoptotic bition of HDAC6 by either specific or pan-HDACIs can Bcl-2 family members (45, 46). trigger different mechanisms of cell death. Hideshima The collective experimental data presented here sup- et al. (40) showed that targeting both proteasome- port the potent activity of the new HDACIs romidepsin

Fig. 6. Enhanced activity of belinostat combined to bortezomib in a xenograft SCID beige mouse model of MCL (HBL-2). The combination of i.p. belinostat (P)at 35 mg/kg/d for 7 d plus i.p. bortezomib (B) at 0.5 mg/kg on days 1, 4, and 11 has shown significant activity with one complete response on day 8 and one on day 12.

562 Clin Cancer Res; 16(2) January 15, 2010 Clinical Cancer Research

HDAC Inhibitors in Mantle Cell Lymphoma

Table 2. Multiple comparison analysis for tumor volume as function of time (HBL-2)

A. Multiple comparison analysis of the median AUCs (MAD) at day 3, 8, and 12 Treatment Median (mm3) ± MAD Day 3 Day 8 Day 12

Control 81.6 ± 26.8 427.8 ± 265.8 1,352 ± 1,217.1 B 84.2 ± 66.7 276.9 ± 179.2 866.5 ± 864.3 P 83.3 ± 40.1 291.1 ± 121,7 752.9 ± 261.3 B + P 40.6 ± 27 53.8 ± 73.2 236.8 ± 317.8

B. The P values for all comparison are shown. Statistically significant tumor shrinkage for the combination group compared with the single drugs and the control was observed at all time points (P ≤ 0.03 in bold)

Treat 1 Treat 2 Day 3 P Day 8 P Day 12 P

Control B 0.740 0.529 0.394 Control P 0.278 0.212 0.357 B P 0.243 0.720 0.843 Control B+P 0.001 0.003 0.003 B B+P 0.001 0.004 0.009 P B+P 0.035 0.003 0.023

NOTE: All significance testing is done at the P < 0.05 level. Abbreviations: AUC, area under the curve; MAD, median absolute deviation; B, bortezomib; P, Belinostat.

and belinostat in combination with the proteasome in- combination may also be associated with a good thera- hibitor bortezomib in MCL. The cytotoxicity assays estab- peutic index. lished IC50 values in the nanomolar range for romidepsin To explore the mechanistic basis for the synergy be- and low micromolar range for belinostat across all MCL tween the proteasome and HDACIs, a series of immuno- cell lines. To quantitate the synergistic interaction be- blottings for proteins involved in cell cycle regulation and tween the HDAC and proteasome inhibitor, intentionally apoptosis were done. Although more comprehensive stud- subtherapeutic concentrations of each drug were studied. ies on the effect of the drugs on the proteome need to be In the cytotoxicity assays, the combination of romidepsin considered, these experiments suggest a prominent effect or belinostat plus bortezomib showed potent synergism on cyclin D1, histone H3, HDAC6, Bcl-XL, and Noxa. De- in all three cell lines of MCL. The RRR analysis showed spite slight differences between the cell lines and the HDA- very strong synergism (RRR < 0.1) in virtually all of the CI used, there was a consistent reduction in cyclin D1 and concentrations studied for the combination. RRR in this Bcl-XL with a consistent increase in acetylated histone H3, range are unusual even for the most established antineo- acetylated α-tubulin, and Noxa in the cell lines treated plastic doublets. These observations were further sup- with the HDACI, bortezomib, and the combination. ported by the changes in the Δψm as a function of the From a mechanistic perspective, high levels of expres- drug exposures. These experiments showed statistically sion of the antiapoptotic protein Mcl-1 has been shown significant changes in the Δψm with the combination to correlate with high-grade morphology and a high proof drugs that was time sensitive. The observed effect on liferative state in MCL (47). Noxa, a BH3-only proapopto- the Δψm was most prominent after 20 hours of exposure tic protein that specifically interacts with and inactivates with >60% of cells exhibiting Δψm in the combination Mcl-1, seems to be emerging as a critical regulator of apo- groups in both cell lines. When used alone, neither HDA- ptosis in MCL (48). It has been reported that Noxa accu- CI nor bortezomib produced a significant change in mulation following bortezomib disrupts the Mcl-1–Bak Δψm or induction of apoptosis. The potent induction complex by displacing free proapoptotic Bak. This event of Δψm and apoptosis in the MCL lines was restricted leads to conformational changes of Bak with activation to the combination treatments. Importantly, when the of the intrinsic or mitochondrial pathway of apoptotis. same combinations of an HDACI plus bortezomib were Several studies have firmly shown strong induction of evaluated on PBMC from healthy donors, none of the apoptosis after treatment with bortezomib irrespective of combinations were found to be more cytotoxic than p53 status through a mechanism that involves increased any drug alone (P ≥ 0.82). These data suggest that this accumulation of Noxa (49, 50). The immunoblottings

www.aacrjournals.org Clin Cancer Res; 16(2) January 15, 2010 563

Paoluzzi et al

performed here seem to suggest that a similar mechanism belinostat plus bortezomib confirmed the in vitro obser- of apoptosis my be operative with the combination in vations. Phase I to II studies have begun to explore this MCL as well. combination, with a focus on relevant pharmacokinetic Before exploring the in vivo activity of belinostat and ro- and pharmacodynamic relationships. The consequence midepsin in combination with bortezomib in a SCID beige of many lines of data around the pivotal role of Noxa xenograft model of MCL, we performed exploratory studies in MCL, especially as it relates to the use of proteasome to optimize the toxicity profile of each HDACI given in and HDACIs, suggests that Noxa and the Δψm could be combination. Plumb et al. (32) showed that a dose of be- important biomarkers of response. Although no studies linostat up to 40 mg/kg/d administered i.p. for 1 week is to date have addressed the potential value of identifying safe in a nude mouse model. When giving belinostat at such biomarkers, it is clear that such surrogate markers of 40 mg/kg/d for 7 days in combination with bortezomib activity could, or perhaps should, be used to guide the in a SCID beige model, we found that a 10% lower dose pharmacologic dosing or schedule design of these agents of belinostat was required to avoid excessive toxicity (no to optimize the synergistic interaction of these novel mouse deaths together with no more than 30% of the mice drug classes. experiencing significant weight loss). In a subsequent xenograft model of MCL (HBL-2), the combination of belinostat Disclosure of Potential Conflicts of Interest at 35 mg/kg with bortezomib showed acceptable toxicity (transient weight loss in 3 of 10 mice but no deaths) com- O.A. O'Connor has received a sponsored research pared with other doses and schedules. A statistically signif- grant from Millenium, Allos, Merck, Abbott, TopoTarget, icant advantage for the combination group was shown (P ≤ Gloucester, Novartis. The other authors disclosed no po- 0.03), with two complete responses out of 10 mice. tential conflicts of interest. To date, the in vivo experience with romidepsin is limit- ed. Although exploratory dosing studies with romidepsin have been conducted, the optimal doses and schedules Acknowledgments have yet to be identified. From a proof-of-principle perspective, the in vivo data of bortezomib plus belinostat We thank Gloucester and TopoTarget for their advice support the in vitro data. Clearly, additional studies across and supply of romidepsin and belinostat, respectively. the panoply of drugs now being developed as protesome This study was supported in part by a grant from the Lym- and HDACIs need to be conducted. phoma Research Foundation for Mantle Cell Lymphoma. In conclusion, the combination of HDACIs such as ro- The costs of publication of this article were defrayed in midepsin or belinostat and a proteasome inhibitor has part by the payment of page charges. This article must shown to be markedly active across a panel of MCL cell therefore be hereby marked advertisement in accordance lines without producing excessive toxicity in normal with 18 U.S.C. Section 1734 solely to indicate this fact. PBMCs from healthy donors. In addition, in vivo xeno- Received 7/22/09; revised 10/6/09; accepted 10/9/09; graft studies exploring the combination of the HDACI published OnlineFirst 1/12/10.

References 1. Raffeld M, Jaffe ES. bcl-1, t(11;14), and mantle cell-derived lympho- 9. Strauss SJ, Maharaj L, Hoare S. Bortezomib therapy in patients with mas. Blood 1991;78:259–63, Review. No abstract available. relapsed or refractory lymphoma: potential correlation of in vitro sen- 2. Shapiro GI. Cyclin-dependent kinase pathways as targets for cancer sitivity and tumor necrosis factor α response with clinical activity. J treatment. J Clin Oncol 2006;24:1770–83, Review. Clin Oncol 2006;24:2105–12, Epub 2006 Apr 10. 3. Lim MS, Adamson A, Lin Z, et al. Expression of Skp2, a p27(Kip1) 10. Belch A, Kouroukis CT, Crump M. A phase II study of bortezomib in ubiquitin ligase, in malignant lymphoma: correlation with p27(Kip1) mantle cell lymphoma: the National Cancer Institute of Canada Clin- and proliferation index. Blood 2002;100:2950–6. ical Trials Group trial IND.150. Ann Oncol 2007;18:116–21, Epub 4. Rosenwald A, Wright G, Wiestner A, et al. The proliferation gene ex- 2006 Sep 13. pression signature is a quantitative integrator of oncogenic events that 11. Fisher RI, Bernstein SH, Kahl BS, et al. Multicenter phase II study of predicts survival in mantle cell lymphoma. Cancer Cell 2003;3:185–97. bortezomib in patients with relapsed or refractory mantle cell lym- 5. Katzenberger T, Petzoldt C, Höller S. The Ki67 proliferation index is a phoma. J Clin Oncol 2006;24:4867–74. quantitative indicator of clinical risk in mantle cell lymphoma. Blood 12. Paoluzzi L, Gonen M, Gardner JR, et al. Targeting Bcl-2 family mem- 2006;107:3407. bers with the BH3 mimetic AT-101 markedly enhances the therapeu- 6. Rummel MJ, de Vos S, Hoelzer D, Koeffler HP, Hofmann WK. Altered tic effects of chemotherapeutic agents in in vitro and in vivo models apoptosis pathways in mantle cell lymphoma. Leuk Lymphoma of B-cell lymphoma. Blood 2008;112:2906–16, Epub 2008 Jun 30. 2004;45:49–54, Review. 13. Paoluzzi L, Gonen M, Bhagat G, et al. The BH3-only mimetic ABT-737 7. O'Connor OA, Wright J, Moskowitz C, et al. Phase II clinical experi- synergizes the antineoplastic activity of proteasome inhibitors in lym- ence with the novel proteasome inhibitor bortezomib in patients with phoid malignancies. Blood 2008;112:2906–16, Epub 2008 Jun 30. indolent non-Hodgkin's lymphoma and mantle cell lymphoma. J Clin 14. Pérez-Galán P, Roué G, Villamor N, Campo E, Colomer D. The BH3- Oncol 2005;23:676–84, Epub 2004 Dec 21. mimetic GX15-070 synergizes with bortezomib in mantle cell lym- 8. Goy A, Bernstein SH, Kahl BS, et al. Bortezomib in patients with re- phoma by enhancing Noxa-mediated activation of Bak. Blood lapsed or refractory mantle cell lymphoma: updated time-to-event 2007;109:4441–9, Epub 2007 Jan 16. analyses of the multicenter phase 2 PINNACLE study. Ann Oncol 15. Hoeller D, Dikic I. Targeting the ubiquitin system in cancer therapy. 2009;20:520–5. Nature 2009;458:438–44, Review.

564 Clin Cancer Res; 16(2) January 15, 2010 Clinical Cancer Research

HDAC Inhibitors in Mantle Cell Lymphoma

16. Lehrmann H, Pritchard LL, Harel-Bellan A. Histone acetyltrans- effect analysis for obtaining dose-response curves for in vitro che- ferases and deacetylases in the control of cell proliferation and dif- mosensitivity testing. Int J Clin Pharmacol Ther 1999;37:189–92. ferentiation. Adv Cancer Res 2002;86:41–65, Review. 34. O'Connor OA, Smith EA, Toner LE, et al. The combination of the pro- 17. Glozak MA, Sengupta N, Zhang X, Seto E. Acetylation and deacety- teasome inhibitor bortezomib and the bcl-2 antisense molecule ob- lation of non-histone proteins. Gene 2005;363:15–23, Epub 2005 limersen sensitizes human B-cell lymphomas to cyclophosphamide. Nov 11. Review. Clin Cancer Res 2006;12:2902–11. 18. Marks PA, Dokmanovic M. Histone deacetylase inhibitors: discovery 35. Ruefli AA, Ausserlechner MJ, Bernhard D, et al. The histone deace- and development as anticancer agents. Expert Opin Investig Drugs tylase inhibitor and chemotherapeutic agent suberoylanilide hydro- 2005;14:1497–511, Review. xamic acid (SAHA) induces a cell-death pathway characterized by 19. Bolden JE, Peart MJ, Johnstone RW. Anticancer activities of histone cleavage of Bid and production of reactive oxygen species. Proc Natl deacetylase inhibitors. Nat Rev Drug Discov 2006;5:769–84, Review. Acad Sci U S A 2001;98:10833–8, Epub 2001 Sep 4. 20. Minucci S, Pelicci PG. Histone deacetylase inhibitors and the prom- 36. Qiu L, Burgess A, Fairlie DP, Leonard H, Parsons PG, Gabrielli BG.

ise of epigenetic (and more) treatments for cancer. Nat Rev Cancer Histone deacetylase inhibitors trigger a G2 checkpoint in normal cells 2006;6:38–51, Review. that is defective in tumor cells. Mol Biol Cell 2000;11:2069–83. 21. Xu WS, Parmigiani RB, Marks PA. Histone deacetylase inhibitors: 37. Shao Y, Gao Z, Marks PA, Jiang X. Apoptotic and autophagic cell molecular mechanisms of action. Oncogene 2007;26:5541–52, death induced by histone deacetylase inhibitors. Proc Natl Acad Review. Sci U S A 2004;101:18030–5, Epub 2004 Dec 13. 22. Yoshida M, Kijima M, Akita M, Beppu T. Potent and specific inhibi- 38. Heider U, Von Metzel I, Kaiser M, et al. Synergistic interaction of the tion of mammalian histone deacetylase both in vivo and in vitro by histone deacetylase inhibitor SAHA with the proteasome inhibitor trichostatin A. J Biol Chem 1990;265:17174–9. bortezomib in mantle cell lymphoma. Eur J Haematol 2008;80:133–45. 23. Paoluzzi L, Scotto L, Marchi E, Seshan VE, O'Connor OA. The anti- 39. Dokmanovic M, Clarke C, Marks PA. Histone deacetylase inhibitors: histaminic cyproheptadine synergizes the antineoplastic activity of overview and perspectives. Mol Cancer Res 2007;5:981–9, Review. bortezomib in mantle cell lymphoma through its effects as a histone 40. Hideshima T, Bradner JE, Wong J, et al. Small-molecule inhibition of deacetylase inhibitor. Br J Haematol 2009;146:656–9. proteasome and aggresome function induces synergistic antitumor 24. O'Connor OA, Heaney ML, Schwartz L, et al. Clinical experience with activity in multiple myeloma. Proc Natl Acad Sci U S A 2005;102: intravenous and oral formulations of the novel histone deacetylase 8567–72, Epub 2005 Jun 3. inhibitor suberoylanilide hydroxamic acid in patients with advanced 41. Kawaguchi Y, Kovacs JJ, McLaurin A, Vance J.M., Ito A, Yao TP. The hematologic malignancies. J Clin Oncol 2006;24:166–73, Epub 2005 deacetylase HDAC6 regulates aggresome formation and cell viability Dec 5. in response to misfolded protein stress. Cell 2003;115:727–38. 25. Duvic M, Talpur R, Ni X, Zhang C, et al. Phase 2 trial of oral vorinostat 42. Hubbert C, Guardiola A, Shao R, et al. HDAC6 is a microtubule- (suberoylanilide hydroxamic acid, SAHA) for refractory cutaneous associated deacetylase. Nature 2002;417:455–8. T-cell lymphoma (CTCL). Blood 2007;109:31–9, Epub 2006 Sep 7. 43. Haggarty SJ, Koeller KM, Wong JC, Grozinger CM, Schreiber SL. Erratum in: Blood. 2007 Jun 15;109:5086. Domain-selective small-molecule inhibitor of histone deacetylase 6 26. Olsen EA, Kim YH, Kuzel TM, et al. Phase IIb multicenter trial of vor- (HDAC6)-mediated tubulin deacetylation. Proc Natl Acad Sci U S A inostat in patients with persistent, progressive, or treatment refracto- 2003;100:4389–94, Epub 2003 Apr 3. ry cutaneous T-cell lymphoma. J Clin Oncol 2007;25:3109–15, Epub 44. Kovacs JJ, Murphy PJ, Gaillard S, et al. HDAC6 regulates Hsp90 2007 Jun 18. acetylation and chaperone-dependent activation of glucocorticoid 27. Ellis L, Pan Y, Smyth GK, et al. Histone deacetylase inhibitor pano- receptor. Mol Cell 2005;18:601–7. binostat induces clinical responses with associated alterations in 45. Inoue S, Riley J, Gant TW, Dyer MJ, Cohen GM. Apoptosis induced gene expression profiles in cutaneous T-cell lymphoma. Clin Cancer by histone deacetylase inhibitors in leukemic cells is mediated by Res 2008;14:4500–10. Bim and Noxa. Leukemia 2007;21:1773–82, Epub 2007 May 24. 28. Kawamata N, Chen J, Koeffler HP. Suberoylanilide hydroxamic acid 46. Dai Y, Chen S, Kramer LB, Funk VL, Dent P, Grant S. Interactions (SAHA; vorinostat) suppresses translation of cyclin D1 in mantle cell between bortezomib and romidepsin and belinostat in chronic lym- lymphoma cells. Blood 2007;110:2667–73, Epub 2007 Jul 2. phocytic leukemia cells. Clin Cancer Res 2008;14:549–58. 29. Hu J, Colburn NH. Histone deacetylase inhibition down-regulates cy- 47. Khoury JD, Medeiros LJ, Rassidakis GZ, McDonnell TJ, Abruzzo LV, clin D1 transcription by inhibiting nuclear factor-κB/p65 DNA binding. Lai R. Expression of Mcl-1 in mantle cell lymphoma is associated Mol Cancer Res 2005;3:100–9. with high-grade morphology, a high proliferative state, and p53 over- 30. Abe M, Nozawa Y, Wakasa H., Ohno H, Fukuhara S. Characterization expression. J Pathol 2003;199:90–7. and comparison of two newly established Epstein-Barr virus-negative 48. Willis SN, Chen L, Dewson G, et al. Proapoptotic Bak is sequestered lymphoma B-cell lines. Surface markers, growth characteristics, by Mcl-1 and Bcl-xL, but not Bcl-2, until displaced by BH3-only pro- cytogenetics, and transplantability. Cancer 1988;61:483–90. teins. Genes Dev 2005;19:1294–305, Epub 2005 May 18. 31. De Leeuw RJ, Davies JJ, Rosenwald A, et al. Comprehensive whole 49. Fennell DA, Chacko A, Mutti L. BCL-2 family regulation by the 20S genome array CGH profiling of mantle cell lymphoma model gen- proteasome inhibitor bortezomib. Oncogene 2008;27:1189–97, Epub omes. Hum Mol Genet 2004;13:1827–37, Epub 2004 Jun 30. 2007 Sep 10. Review. 32. Plumb JA, Finn PW, Williams RJ, et al. Pharmacodynamic response 50. Pérez-Galán P, Roué G, Villamor N, Montserrat E, Campo E, Colomer and inhibition of growth of human tumor xenografts by the novel his- D. The proteasome inhibitor bortezomib induces apoptosis in mantle- tone deacetylase inhibitor PXD101. Mol Cancer Ther 2003;2:721–8. cell lymphoma through generation of ROS and Noxa activation inde- 33. Pöch G., Vychodil-Kahr S, Petru E. Sigmoid model versus median- pendent of p53 status. Blood 2006;107:257–64, Epub 2005 Sep 15.

www.aacrjournals.org Clin Cancer Res; 16(2) January 15, 2010 565

Romidepsin and Belinostat Synergize the Antineoplastic Effect of Bortezomib in Mantle Cell Lymphoma

Luca Paoluzzi, Luigi Scotto, Enrica Marchi, et al.

Clin Cancer Res Published OnlineFirst January 12, 2010.

Updated version Access the most recent version of this article at: doi:10.1158/1078-0432.CCR-09-1937

Supplementary Access the most recent supplemental material at: Material http://clincancerres.aacrjournals.org/content/suppl/2010/01/18/1078-0432.CCR-09-1937.DC1

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/early/2010/01/10/1078-0432.CCR-09-1937. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.