Cooperative effect of chidamide and chemotherapeutic drugs induce apoptosis by DNA damage accumulation and repair defects in acute myeloid leukemia stem and progenitor cells

Abstract

Background: Many conventional chemotherapeutic drugs are known to be involved in DNA damage, thus ultimately leading to apoptosis of leukemic cells. However, they fail to completely eliminate leukemia stem cells (LSCs) due to their higher DNA repair capacity of cancer stem cells than that of bulk cancer cells, which becomes the root of drug resistance and leukemia recurrence. A new strategy to eliminate LSCs in acute myeloid leukemia (AML) is therefore urgently needed.

Results: We report that a low-dose chidamide, a novel orally active benzamide-type histone deacetylase (HDAC) inhibitor, which selectively targets HDACs 1, 2, 3, and 10, could enhance the cytotoxicity of DNA-damaging agents (daunorubicin, idarubicin, and cytarabine) in CD34⁺CD38^- KG1α cells, CD34⁺CD38^- Kasumi cells, and primary refractory or relapsed AML CD34⁺ cells, reflected by the inhibition of cell proliferation, induction of apoptosis, and increase of cell cycle arrest in vitro. Mechanistically, these events were associated with DNA damage accumulation and repair defects. Co-treatment with chidamide and the DNA-damaging agent IDA gave rise to the production of γH2A.X and inhibited posttranslationally but not transcriptionally the repair gene of ATM, BRCA1, and checkpoint kinase 1 (CHK1) and 2 (CHK2) phosphorylation. Finally, the combination of chidamide and IDA initiated caspase-3 and PARP cleavage, but not caspase-8 and caspase-9, and ultimately induced CD34⁺CD38^- KG1α cell apoptosis. Further analysis of AML patients' clinical characteristics revealed that the ex vivo efficacy of chidamide in combination with IDA in primary CD34⁺ samples was significantly correlated to peripheral blood WBC counts at diagnosis, while LDH levels and karyotype status had no effect, indicating that the combination regimen of chidamide and IDA could rapidly diminish tumor burden in patients with R/R AML.

Conclusions: These findings provide preclinical evidence for low-dose chidamide in combination with chemotherapeutic agents in treating recurrent/resistant AML as an alternative salvage regimen, especially those possessing stem and progenitor cells.

Keywords: Acute myeloid leukemia; Chidamide; DNA damage; Drug resistance; Leukemia stem and progenitor cells.

Fig. 1

Chidamide enhanced IDA, DNR, or Ara-C in cytotoxic effects on CD34⁺CD38⁻ KG1α and Kasumi cells. CD34⁺CD38⁻ KG1α (a) and Kasumi (b) cells were exposed to the indicated concentrations of IDA, DNR, or Ara-C with or without 0.75 μM chidamide for 24, 48, and 72 h, after which the cell viability effect was analyzed by CCK-8 assay. Percent of viability is normalized with DMSO-treated control. The values in the figure are expressed as the mean ± S.D. from three independent experiments

Fig. 2

Chidamide synergized IDA-, DNR-, or Ara-C-induced apoptosis in both CD34⁺CD38⁻ KG1α cells and primary relapsed or refractory AML CD34⁺ cells. CD34⁺CD38⁻ KG1α cells were exposed to the indicated concentrations of IDA, DNR, or Ara-C with or without 0.75 μM chidamide for 24, 48, and 72 h (a), with or without 0.5 or 0.75 μM chidamide for 72 h (b), after which flow cytometric analysis was performed to determine the percentage of Annexin V⁺ cells. Horizontal lines represent the mean ± S.D. from three independent experiments. (c) Primary CD34⁺ AML cells were exposed to the 20 nM IDA with or without 0.75 μM chidamide, after which apoptotic ratios were determined by Annexin V staining and flow cytometry. (d) Representative data for flow cytometric analysis of hCD34 and Annexin V/PI staining in primary cells after exposed (48 h) to 20 nM IDA with or without 0.75 μM chidamide. *P < 0.05

Fig. 3

Chidamide intensified IDA-induced cell cycle arrest in CD34⁺CD38⁻ KG1α cells. (a) CD34⁺CD38⁻ KG1α cells were exposed to 5 or 10 nM IDA with or without 0.5 or 0.75 μM chidamide for 72 h, after which flow cytometric analysis was performed to determine the cell cycle. Horizontal lines represented the mean ± S.D. from three independent experiments. (b) Representative data for flow cytometric analysis of PI staining in CD34⁺CD38⁻ KG1α cells after exposed (72 h) to 5 or 10 nM IDA with or without 0.5 or 0.75 μM chidamide

Fig. 4

Chidamide increased IDA-induced DNA damage in CD34⁺CD38⁻ KG1α cells. CD34⁺CD38⁻ KG1α cells were treated with or without 0.75 μM of chidamide in combination with 40 nM IDA for 24 h. (a) Immunostaining was performed with γH2A.X and counterstained with DAPI. Each experiment was performed in triplicate. (b) The expression of γH2A.X was examined in the presence or absence of chidamide with or without IDA for 24 h followed by flow cytometry. Filled-area histograms represent medium treated cells, whereas color histograms represent different treated group, and the compiled results from three experiments are shown in (c). (d) The expression of γH2A.X and acetylation of histone 3 were examined in the presence or absence of chidamide with or without IDA for 24 and 48 h followed by western blot

Fig. 5

Chidamide posttranslationally but not transcriptionally inhibited the repair of IDA-induced DNA damage in CD34⁺CD38⁻ KG1α cells. CD34⁺CD38⁻ KG1α cells were incubated with 0.75 μM chidamide ±40 nM IDA for 24 and 48 h, followed by qRT-PCR to evaluate the gene expression of BRCA1, ATM, CHK1, and CHK2 (a), or western blot analysis to monitor the expression of p-BRCA1, p-ATM, p-CHK1, and p-CHK2 (b). β-actin was used as a loading control

Fig. 6

Chidamide initiated caspase-3 and PARP cleavage in CD34⁺CD38⁻ KG1α cells. Western blotting analysis of the expression of caspase-3, caspase-8, caspase-9, and PARP and their cleavage in CD34⁺CD38⁻ KG1α cells exposed to 0.75 μM chidamide with or without 40 nM IDA treatment for 48 h, with the untreated group as the control. β-actin was used as a loading control