CUDC-907 blocks multiple pro-survival signals and abrogates microenvironment protection in CLL

Abstract

CUDC-907, a dual PI3K/HDAC inhibitor, has been proposed to have therapeutic potential in hematopoietic malignancies. However, the molecular mechanisms of its effects in chronic lymphocytic leukaemia (CLL) remain elusive. We show that CLL cells are sensitive to CUDC-907, even under conditions similar to the protective microenvironment of proliferation centres. CUDC-907 inhibited PI3K/AKT and HDAC activity, as expected, but also suppressed RAF/MEK/ERK and STAT3 signalling and reduced the expression of anti-apoptotic BCL-2 family proteins BCL-2, BCL-xL, and MCL-1. Moreover, CUDC-907 downregulated cytokines BAFF and APRIL and their receptors BAFFR, TACI, and BCMA, thus blocking BAFF-induced NF-κB signalling. T cell chemokines CCL3/4/17/22 and phosphorylation of CXCR4 were also reduced by CUDC-907. These data indicated that CUDC-907 abrogates different protective signals and suggested that it might sensitize CLL cells to other drugs. Indeed, combinations of low concentrations of CUDC-907 with inhibitors of BCL2, BTK, or the NF-κB pathway showed a potent synergistic effect. Our data indicate that, apart from its known functions, CUDC-907 blocks multiple pro-survival pathways to overcome microenvironment protection in CLL cells. This provides a rationale to evaluate the clinical relevance of CUDC-907 in combination therapies with other targeted inhibitors.

Keywords: CLL; HDAC; PI3K; microenvironment.

Figure 1

CUDC‐907 inhibits CLL cell growth and induces apoptosis. A, MEC‐1 cell viability measured by MTS assay, normalized to Control samples (0 μmol/L). Cells were treated with various small molecule inhibitors (0.001‐10 μmol/L) or DMSO (0 μmol/L) for 48 h. Experiments were performed in triplicate and repeated at least 2 times. Graphics show mean values and error bars represent standard deviation, as in all the other panels of this figure. B, Induction of cell death in MEC‐1, as measured by FACS analysis of by PI‐stained cells. Cells were treated with inhibitors for 48 h. Graphics show percentage of PI positive (dead) cells. C, Same as (A), using primary CLL cells. Cells were cultured for 24 h in the presence of 10 ng/mL IL‐4 and CD154 and then treated with inhibitors for 48 h. The number of primary samples used for each inhibitor is: 23 (CUDC‐907), 13 (IMD0354), 6 (CAL‐101), 11 (Ibrutinib), 8 (PLX‐4720), 6 (Entospletinib), and 7 (ABT‐199). Experiments were performed in triplicate. Graphics show mean values and error bars represent standard deviation. D, Same as (B), using primary CLL cells. The number of primary samples used for each inhibitor are: 8 (CUDC‐907), 6 (IMD0354), 5 (CAL‐101), 5 (Ibrutinib), 5 (PLX‐4720), 5 (Entospletinib), and 5 (ABT‐199)

Figure 2

CUDC‐907 inhibits pro‐survival signals in CLL cells. A, Western blots showing protein expression in lysates of MEC‐1 cells treated with CUDC‐907 at various concentrations or DMSO as a control (0) for 12 h. AKT, p‐AKT (Ser473), p‐p70S6K (Thr389), Ac‐H3K9, ERK, p‐ERK (Thr202/Tyr204), p‐MEK (Ser217/221), STAT3, p‐STAT3 (Tyr705), MCL‐1, BCL‐2, BCL‐xL, and PARP were detected with specific antibodies. β‐actin was used as a loading control. B, Representative Western blot showing PARP, caspase‐9, and caspase‐8 cleavage in CLL primary cells from two patients. Cells were cultured as in Figure 1C and treated with different concentrations of CUDC‐907 for 12 h. C, mRNA expression levels of BAFF, APRIL, BAFFR, TCIA, and BCMA, as measured by qRT‐PCR, in CLL patient cells cultured in the presence of 10 ng/mL IL‐4 and CD154 for 24 h then treated with CUDC‐907 for 12 h. The expression of target genes was normalized to an internal control, GAPDH. Experiments were performed in triplicates and repeated at least 3 times. Data were expressed as relative to control (no inhibitor). D, Representative Western blot of lysates of CLL patient cells stimulated with BAFF (100 ng/mL) for 24 h. NF‐κB(p65), NF‐κB(RelB), IκBα, p‐IκBα, and β‐actin were detected using specific antibodies. The graph shows quantitation the Western blot bands normalized to β‐actin and expressed as relative to control (0 h) treatment

Figure 3

Effects of CUDC‐907 on the expression of microenvironment chemokines. A, mRNA expression of CCL3, CCL4, CCL17, and CCL22, as measured by qRT‐PCR in CLL patient cells cultured with 10 μg/mL anti‐IgM and treated with CUDC‐907 for 24 h. B, Secretion levels of CCL3/4, as measured by quantitative ELISA, in supernatants of CLL patient cells cultured with anti‐IgM and treated with CUDC‐907. C, Expression of smCXCR4 in CLL patient cells not stimulated (left) or stimulated with 200 ng/mL CXCL12 (right), and treated with CUDC‐907 for 12 h. DMSO was used in controls. Cells were incubated with a CXCR4 primary antibody and a fluorescent labelled secondary antibody. The smCXCR4 signalling was measured with FACS. D, Representative Western blot showing levels of total and phosphorylated CXCR4 measured in lysates of CLL cells from two patient treated with CUDC‐907 for 12 h. E, Representative Western blot showing levels of ERK and phospho‐ERK in CLL patient treated as in (C)

Figure 4

Combination of CUDC‐907 with other specific inhibitors show synergistic effects. A, Percentage of cell death, as measured by PI staining, in MEC‐1 cells treated with different concentrations of CUDC‐907 for 48 h, alone or in combination with IMD‐0354, ibrutinib or ABT‐199. Graphs show average of three independent experiments. B, Same in CLL patient cells cultured in the presence of 10 ng/mL IL‐4 and CD154 for 24 h before treated with drug combinations