ATM inhibition overcomes resistance to histone deacetylase inhibitor due to p21 induction and cell cycle arrest

Abstract

The antiproliferative effect induced by histone deactylase inhibitors (HDACi) is associated with the up-regulated expression of the cyclin-dependent kinase inhibitor p21. Paradoxically, the increased expression of p21 correlates with a reduced cell killing to the drug. The direct targeting of p21 is not feasible. An alternate approach could selectively target factors upstream or downstream of p21 that affect one or more specific aspects of p21 function. HDAC inhibitors appear to activate p21 expression via ataxia telangiectasia mutated (ATM) activity. KU60019, a specific ATM inhibitor, has shown to decrease the p21 protein levels in a concentration dependent manner. We explored the potential synergistic interaction of the ATM inhibitor with romidepsin, given the potential complementary impact around p21. A synergistic cytotoxic effect was observed in all lymphoma cell lines examined when the HDACi was combined with KU60019. The increase in apoptosis correlates with decreased expression of p21 due to the ATM inhibitor. KU60019 decreased expression of the cyclin-dependent kinase inhibitor at the transcriptional level, compromising the ability of HDACi to induce p21 and cell cycle arrest and ultimately facilitating a shift toward the apoptotic phase. Central to the increased apoptosis observed when romidepsin is combined with KU60019 is the reduced expression of p21 and the absence of a G2/M cell cycle arrest that would be exploited by the tumor cells to evade the cytotoxic effect of the HDAC inhibitor. We believe this strategy may offer a promising way to identify rational combinations for HDACi directed therapy, improving their activity in malignant disease.

Keywords: ATM inhibitor; HDAC inhibitor; cell cycle; lymphoma; p21.

Figure 1. Romidepsin affects expression of proteins involved in cell cycle regulation in MCL.

(A) Growth inhibition curves were generated for a panel of MCL cell lines following 24 hours treatment with increased concentrations of romidepsin. (B) Protein expression analysis of Jeko-1 cells following 24 hours treatment with bortezomib or romidepsin (upper panel). Emi 1 and p21 protein expression in a panel of MCL cell lines following 24 hours romidepsin exposure (lower panel). (C) Cell cycle analysis of Jeko-1 cells following 24 hours romidepsin treatment. Error bars represent SD.

Figure 2. The ATM inhibitor KU60019 and Romidepsin are highly synergistic in in vitro models of MCL.

(A) Growth inhibition curves were generated for the panel of MCL cell lines following 48 hours of KU60019 exposure. (B) Cytotoxicity effects observed at 48 hours of exposure to single agents and combination for Jeko-1, Maver-1, and Z-138 MCL cell lines. (C) Cytotoxicity effects observed at 48 hours of exposure to single agents and combination for PBMC isolated from a patient diagnosed with classic MCL, blastoid variant of MCL and a healthy donor. Also shown are RRR values at 24, 48, and 72 hours of exposure to romidepsin in combination with KU60019. Error bars represent SD.

Figure 3. Romidepsin plus KU60019 enhances apoptosis in MCL.

(A) Cytofluorimetric detection of apoptosis in Jeko-1, Maver-1 and Z-138 MCL cell lines after 48 hours of exposure to single agents and combination. (B) Expression analysis of p21, pH2AX, pro- and anti-apoptotic proteins in Jeko-1, Maver-1, and Z-138 MCL cell lines after 48 hours of exposure to single agents and combination.

Figure 4. The ATM inhibitor KU60019 and Romidepsin are highly synergistic in in vitro models of DLBCL, CTCL, TALL, ATLL.

(A) Romidepsin and KU60019 were used as single agents and combination to assess their cytotoxicity on a panel of human derived lymphoma cell lines. The relative risk ratio values at 24, 48, and 72 h of exposure are shown. RRR values < 1 represent the synergistic effect of the 2 drugs; values equal to 1 indicate the mean additive effect of the drugs; and values > 1 represent an antagonistic effect. (B) Protein expression of PARP and pH2AX after 48 hours of exposure to single agents and combination. One lymphoma cell line per tumor type is shown.

Figure 5. KU60019 affects the ability of romidepsin to induce expression of p21.

(A) Protein and RT-PCR expression analysis of p21 and transcription factors Sp1, Sp3 in Jeko-1, Maver-1, and Z-138 cells after 24 hours of exposure to single agent and combination. (B) Transient transcription analysis of p21 promoter luciferase gene fusion constructs. Reporter plasmids were transiently transfected in Jeko-1 cells and luciferase activity was measured after 24 hours exposure to single agents and combination. Schematic representation of p21 promoter region with depicted Sp1 binding sites and cloned DNA regions. Sp3 primers were designed so to identify both alternative transcripts. Luciferase basal activity of the two constructs is arbitrarily set to 1.0. Error bars represent SD.

Figure 6. Inhibition of ATM mediated activation of the G2/M checkpoint by KU60019 facilitates mitotic exit.

(A) Protein expression analysis of ATM, H3, and CDC2 in Jeko-1, Maver-1, and Z-138 cells exposed 24 hours to romidepsin and KU60019 as single agents and combination (B) KU60019 inhibit HDACi mediated p21 transcriptional induction facilitating G2 to M transition. Premature mitotic exit induced by romidepsin promote apoptosis (C) Cell cycle analysis of Jeko-1 cells following 24 hours exposure to romidepsin and KU60019 as single agents and combination.

Figure 7. The combination of romidepsin and KU60019 is synergistic in a xenograft model of MCL.

(A) The in vivo therapeutic efficacy of romidepsin and KU60019 as single agents and combination was evaluated using a SCID Beige xenograft mouse model of MCL. The effectiveness of romidepsin in combination with KU60019 was evaluated in a 3 week cycle administration. Romidepsin was administered weekly at days 1, 8, and 15. KU60019 was given at days 1, 2, 3, 4, and 5 of each week. The data are expressed as average tumor volume (mm³) per group as a function of time. A two-tailed t test confirmed that combination was statistically superior to both the single agents and the control in inhibiting tumor growth (0.001 < P < 0.05). (B) Statistical analysis of survival following the 3 weeks administration cycle. When compared to control, only the romidepsin plus KU60019 combination was statistically significant. P-value, standard error of the mean (SEM), coefficient of determination (R squared), and coefficient of correlation (r) of comparisons.