Sensitizing thermochemotherapy with a PARP1-inhibitor

Abstract

Cis-diamminedichloroplatinum(II) (cisplatin, cDDP) is an effective chemotherapeutic agent that induces DNA double strand breaks (DSBs), primarily in replicating cells. Generally, such DSBs can be repaired by the classical or backup non-homologous end joining (c-NHEJ/b-NHEJ) or homologous recombination (HR). Therefore, inhibiting these pathways in cancer cells should enhance the efficiency of cDDP treatments. Indeed, inhibition of HR by hyperthermia (HT) sensitizes cancer cells to cDDP and in the Netherlands this combination is a standard treatment option for recurrent cervical cancer after previous radiotherapy. Additionally, cDDP has been demonstrated to disrupt c-NHEJ, which likely further increases the treatment efficacy. However, if one of these pathways is blocked, DSB repair functions can be sustained by the Poly-(ADP-ribose)-polymerase1 (PARP1)-dependent b-NHEJ. Therefore, disabling b-NHEJ should, in principle, further inhibit the repair of cDDP-induced DNA lesions and enhance the toxicity of thermochemotherapy. To explore this hypothesis, we treated a panel of cancer cell lines with HT, cDDP and a PARP1-i and measured various end-point relevant in cancer treatment. Our results demonstrate that PARP1-i does not considerably increase the efficacy of HT combined with standard, commonly used cDDP concentrations. However, in the presence of a PARP1-i, ten-fold lower concentration of cDDP can be used to induce similar cytotoxic effects. PARP1 inhibition may thus permit a substantial lowering of cDDP concentrations without diminishing treatment efficacy, potentially reducing systemic side effects.

Keywords: PARP1-inhibitor; RAD51; cDDP; hyperthermia; synthetic lethality.

Figure 1. Sensitivity of cells to hyperthermia

(A) Cell cycle analysis were determined via FACS analysis after BrdU incorporation. A G2-arrest is observed after HT treatment. (B) Apoptosis levels were studied using the Nicoletti assay. HT induced apoptosis in all cell lines. (C). Representative pictures of co-localization of γ-H2AX and RAD51 foci on α-irradiation tracks in untreated cells and after HT treatment. RAD51 is no longer detected 30 min after HT, indicating that HR is not active. The bar graph with the standard error of the mean shows the mean of at least three independent experiments. For each condition more than 300 cells were analyzed.

Figure 2. Effects of PARP1-i (100 μM NU1025/continuously), HT (42°C/1 h), cDDP (5 μM/1 h)

(A) Overview of treatment schemes for different experiments represented in B–E. (B) Clonogenic assays were conducted in order to study the cell survival after 10–12 days post-treatment. No significant differences were found between HT+cDDP and cDDP + HT + PARP1-i. (R1: p = 0.10, SiHa: p = 0.12, HeLa: p = 0.10). (C) DNA DSBs were analyzed using the γ-H2AX assay. The induction of DSBs in SiHa and HeLa after cDDP treatment is 3-4 times higher than in R1 cells. Nonetheless, all three cells lines show around 70–80 DSBs after cDDP+HT or cDDP+HT+PARP1-i. (D) The Nicoletti assay was performed to observe apoptosis levels after different treatments. The triple combination treatment gave the highest levels of apoptosis. (E) Cell cycle was determined via FACS analysis after BrdU incorporation. cDDP+HT and cDDP+HT+PARP1-i show a slight increase in cells in S-phase. Graph bars represent mean of at least three experiments with standard error of the mean. Asterisks indicate the significant differences compared to the untreated sample (ctrl), this was tested using the non-parametric Mann-Whitney test. *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 3. Time-lapse microscopy analysis 0 and 96 h after combination treatments with PARP1-i

(100 μM NU1025/continuously), HT (42°C/1 h), cDDP (5 μM/1 h). (A) Representative pictures of cells directly after treatment (0 h) and four days after treatment (96 h). (B) Cell division times of two generations post treatment. Cells that did not divide for over 50 h, but did not die, were regarded to be in ‘cell cycle arrest’. R1 cells have a shorter cell division time than SiHa and HeLa cells. But no significant differences were found after any of the treatments compared to the untreated cells. Graph bars represent mean with standard error of the mean of three independent experiments with the standard error of the mean of three independent experiments. (C) Graph bars with means represent cell status 96 h after treatment with examples pictures of each cell status on the right. Each condition has been performed in triplicate.

Figure 4. Addition of PARP1-i (100 μM NU1025/continuously) to the commonly used dose of cDDP (5 μM/1 h; cDDP/C) and HT (42°C/1 h) permits lowering of cDDP concentration (0.5 μM/1 h; cDDP/L)

(A) To study apoptosis levels, the Nicoletti assay was performed 48 h after different treatments. No significant differences were observed between cDDP/L+HT+PARP1-i and cDDP/C+HT (R1: p = 0.10, SiHa: p = 0.86, HeLa: p = 0.22). (B) Time-lapse microscopy was performed 0–96 h after treatments. Pictures represent cells directly after treatment (0 h) and at the end of the analysis (96 h). (C) Clonogenic assays were conducted to study the effect 10-12 days after treatments. No significant differences were observed between cDDP/L+HT+PARP1-i and cDDP/C+HT (R1: p = 0.10, SiHa: p = 0.70, HeLa: p = 0.20). Graph bars show mean of at least three independent experiments with the standard error of the mean. Asterisks indicate the significant differences compared to the untreated sample (ctrl), tested using the non-parametric Mann-Whitney test. *p < 0.05, **p < 0.01, ***p < 0.001.