3.1. Effect of doxorubicin and sildenafil on cell cycle progression in PC-3 and DU145 cells
The cell population of the cell cycle in both PC-3 and DU145 cells treated with doxorubicin and sildenafil was determined by PI staining and analyzed using FACScan flow cytometry. Results showed that sildenafil alone did not affect the cell cycle distribution (Fig. 1). However, it significantly potentiated doxorubicin-induced increase of sub-G1 phase population, while decreased both S and G2/M cell population in PC-3 cells (Fig. 1). Similar effects in increasing sub-G1 population were observed in DU145 cells (Fig. 2). Since sub-G1 cell population was considered as apoptotic cells with DNA fragmentation followed by loss of DNA content, the validation of apoptotic cell death was conducted.
3.2. Validation of sildenafil-mediated sensitization of doxorubicin-induced apoptosis
Several assays were performed to confirm whether sildenafil could synergistically enhance apoptotic cell death induced by doxorubicin. The results of microscopic examination showed that, in contrast to doxorubicin alone, the combinatory treatment of doxorubicin and sildenafil for 24 h caused a profound increase of cell shrinkage and apoptotic bodies in PC-3 cells (Fig. 3A). It was worth noting that although sildenafil was able to sensitize PC-3 cells to doxorubicin, sildenafil itself did not have any cytotoxic effect on PC-3 cells. Besides, cell death detection ELISAPLUS kit which detects nucleosomal DNA fragments in the cytoplasm of apoptotic cells was employed to validate cell apoptosis. As a consequence, sildenafil alone did not induce an increase of
nucleosomal DNA fragments in PC-3 cells, but significantly increased doxorubicin-induced effects by 1.67 times (doxorubicin, 1.68-fold; combination, 2.80-fold) (Fig. 3B). Moreover, the expression of apoptotic markers in intrinsic and extrinsic pathway was monitored by Western blot in PC-3 cells. The apoptotic markers including cleaved caspase-9 and -8 which served as initiator caspases in intrinsic and extrinsic pathways, respectively, and their common downstream substrates, effector caspase-3 and PARP-1, were examined. The results demonstrated that the combinatorial treatment for 24 hours significantly increased the expression of cleaved caspase-3, 8, 9 and PARP-1 as compared to doxorubicin alone (Fig. 3C). These results suggested that sildenafil was able to potentiate doxorubicin-induced cell apoptosis by activation of both intrinsic and extrinsic apoptotic pathways.
3.3. Effect of doxorubicin or/and sildenafil on the expression of Bcl-2 family proteins
Bcl-2 family proteins, consisting of anti-apoptotic and pro-apoptotic members, are gatekeepers of mitochondria and crucial regulators of apoptosis particularly in the intrinsic pathway to govern the mitochondrial outer membrane permeability. Since the combinatorial treatment enhanced the activation of caspase-9, a key initiator caspase in the mitochondria-involved intrinsic apoptotic pathway, the effect of the combinatorial treatment on the expression of Bcl-2 family proteins was further examined by Western blot. Consequently, the combinatorial treatment for 24 h could further reduce doxorubicin-mediated decease of anti-apoptotic Bcl-2 family proteins, including Mcl-1, Bcl-xL and Bcl-2, but not those of pro-apoptotic members, such as Bax and Bak, in PC-3 cells (Fig. 4A and 4B). Moreover, combinatorial treatment further downregulated the protein level of Bid pro-form (Fig. 4A and 4B). The decreased Bid pro-form was
indicative of the caspase-8 activation since Bid served as a downstream substrate of caspase-8. Taken together, the data suggest that sildenafil may augment the mitochondria-dependent apoptosis induced by doxorubicin through modification of specific members of Bcl-2 family.
3.4. Effect of doxorubicin or/and sildenafil on ROS production
Since mitochondrial dysfunction is often associated with an increase in ROS production, the intracellular ROS levels in PC-3 cells after combinatorial treatment were determined by DCFH-DA assay. The data revealed that the short-term (e.g., 3 h) exposure to doxorubicin or sildenafil alone or combinatorial treatment significantly elevated cellular ROS levels. All the increased levels of ROS production were dramatically abolished in the presence of ROS scavengers, NAC and trolox (Fig. 5). To further determine whether the oxidative stress contributed to the apoptotic sensitization, the flow cytometric analysis of PI staining was performed to determine apoptotic sub-G1 population. As a result, neither NAC nor trolox significantly blunted the synergistic cell apoptosis caused by the combinatorial treatment (Fig. 6). The data suggest that the increase of oxidative stress is not responsible for the apoptotic sensitization mechanism.
3.5. Effect of sildenafil on doxorubicin-induced DNA double-strand break signaling and repair system
Doxorubicin is a DNA damaging drug known to cause DNA double-strand breaks (DSB). The effects of sildenafil on doxorubicin-mediated DSB signaling and repair system were examined. We first employed the alkaline comet assay to monitor the chromosomal DNA integrity. Comet tail moment was utilized as a scoring parameter to
assess the DNA damage level in individual cells. The data revealed that although doxorubicin alone induced a rapid and profound increase of comet tail moment, sildenafil did not augment the DNA damage levels caused by doxorubicin (Fig. 7). The results of Western blot also showed that sildenafil did not increase doxorubicin-induced phosphorylation of histone H2A.X at Ser139 (γ-H2A.X) and Chk2 phosphorylation at Thr68 (the hallmarker and transducer kinase of DSB, respectively) at an early exposure time (e.g., 3 hours) (Fig. 8A, B). Of note, sildenafil significantly increased the levels of doxorubicin-induced γ-H2A.X formation at a longer exposure time (e.g., 24 hours) (Fig.
8A). The results indicate that sildenafil is unable to potentiate direct DNA damage induced by doxorubicin, but can ultimately sensitize the DNA damaging effect and apoptosis through certain programmed mechanism, such as impairment of DNA repair systems. Accordingly, several markers involved in DNA repair were examined. The data demonstrated that sildenafil blunted the initial RPA32 hyperphosphorylation induced by doxorubicin (Fig. 8C and 8D) and lowered doxorubicin-elicited DNA-PKcs phosphorylation (Thr2609) (Fig. 8F and 8G). In addition, sildenafil further decreased doxorucibin-induced down-regulation of Rad51 (Fig. 8C and 8E). Similar effects were observed in DU145 cells with combinatorial treatment (Fig. 9). It has been well recognized that RPA32 binds to ssDNA during the initial phase of homologous recombination (HR) pathway and its hyperphosphorylation plays a critical role in promoting DSB repair to maintain genome stability in response to DNA damage57, 58. Rad51 plays a major role in homologous recombination (HR) for repairing DSB59. Differently, DNA-PKcs is required for non-homologous end joining (NHEJ) pathway of DNA repair, which rejoins DSBs60. Altogether, the data suggest that sildenafil may impair both HR and NHEJ pathways of DNA repair during doxorubicin-induced DNA damage effect.
3.6. Effect of combinatorial treatment on DNA end-binding capacity and protein expression of Ku80
Ku heterodimers (Ku70/Ku80) are crucial DSB sensors in NHEJ pathway. They can bind directly to the ends of DSB, and serve as a molecular scaffold to recruit the core NHEJ machinery including DNA-PKcs to repair DSB. To monitor the activity of NHEJ, the DNA end-binding capacity of Ku80 was determined. The data demonstrated that the combinatorial treatment of PC-3 cells with doxorubicin and sildenafil significantly, although moderately, decreased the DNA end-binding capacity of nuclear Ku80 (Fig. 10A). Total and nuclear Ku80 levels were detected by Western blot analysis, and the Ku80 expression was not modified in the presence of doxorubicin and sildenafil (Fig. 10B). The results indicate that the combinatorial treatment can reduce the NHEJ activity by decreasing the DNA end-binding capacity of nuclear Ku80 without changing its protein expression.
3.7. Effect of combinatorial treatment on nuclear foci formation and expression of Rad51
Rad51, a recombinase, has been well recognized to play a central role in HR pathway to accurately repair DSB by catalyzing homology searching and strand exchange reactions. Successful HR repairing requires the formation of Rad51 nucleofilaments on single-strand DNA, which can be seen as Rad51 foci in nucleus using immunefluorescence imaging. Therefore, we further examined the nuclear Rad51 foci formation in PC-3 cells. The data showed that doxorubicin dramatically induced the formation of nuclear Rad51 foci; however, the effect was significantly inhibited by sildenafil (Fig. 11A-11C). Besides, Western blot of nuclear extract also showed the diminished expression of Rad51 after the combinatorial treatment for 24 hours (Fig.
11D). The results indicate that sildenafil may impair doxorubicin-elicited HR-mediated DSB repair by lowering the total protein level of Rad51, thereby decreasing the formation of nuclear Rad51 foci.
3.8. Effect of other PDE5 inhibitors on doxorubicin-induced cell apoptosis
To realize whether inhibition of PDE5 activity was indispensable for the sensitization mechanism, two other FDA-approved PDE5 inhibitors, vardenafil and tadalafil, were examined in PC-3 cells in this study. The data showed that tadalafil alone but not vardenafil alone induced a small but significant production of nucleosomal DNA fragments (apoptosis). Furthermore, both PDE5 inhibitors sensitized doxorubicin-induced effects although vardenafil exhibited a much higher sensitization activity than that of tadalafil (Fig. 12A). The data of Western blot analysis also showed that both PDE5 inhibitors potentiated doxorubicin-induced caspase-3 activation and the cleavage of its downstream substrate PARP-1 (Fig. 12B and 12C).
The synergism of doxorubicin and PDE5 inhibitors (vardenafil or tadalafil) on apoptosis was further substantiated in DU145 using PI staining and FACScan flow cytometry for the detection of sub-G1 population (apoptotic cells). The results revealed that both vardenafil (Fig. 13A) and tadalafil (Fig. 13B) could significantly potentiate sub-G1 population induced by doxorubicin in a dose-dependent manner.
Taken together, the data suggest that the inhibition of PDE5 activity may play a crucial role in the sensitization mechanism when combined with doxorubicin.
3.9. Effect of other PDE5 inhibitors on HR-mediated repair of doxorubicin-induced DSB
To investigate whether two other PDE5 inhibitors, vardenafil and tadalafil, could
also reduce the expression of HR-related proteins in response to doxorubicin, the Western blot analysis was performed to detect protein levels of hyperphosphorylated RPA32 (p-RPA32) and Rad51 in PC-3 cells. The results revealed that only vardenafil but not tadalafil could reduce the doxorubicin-induced hyperphosphorylation of RPA32 (Fig. 14 A and B). However, both vardenafil and tadalafil could lead to a decrease in Rad51 protein level when combined with doxorubicin although vardenafil was more effective than tadalafil (Fig. 14 A and B). Notably, there was a strong negative correlation (R2 = 0.9304) between the levels of Rad51 protein expression and nucleosomal DNA fragments in the presence of doxorubicin and different PDE5 inhibitors (sildenafil, vardenafil or tadalafil) (Fig. 15), supporting that the down-regulation of Rad51 protein and the impaired DNA repair might contribute to the sensitization effect on cell apoptosis.
3.10. Effect of PDE5 knockdown on doxorubicin-induced cell death and DSB signaling and repair
Although sildenafil, vardenafil and tadalafil are all potent and selective PDE5 inhibitors approved by FDA, they have also been reported to have inhibitory effects on other PDE isoenzymes, such as PDE1, 6 or 11. Thus, to determine the role of PDE5 but not the other subtypes in the sensitization mechanism is critical. To this end, we knocked down PDE5 using small interfering RNA (siRNA) to further check doxorubicin-mediated effect. The data showed that PDE5 knockdown did not change the cell cycle distribution of the control group in PC-3 cells but decreased both S and G2/M cell population, while significantly increased the sub-G1 population induced by doxorubicin (Fig. 16A and B). However, PDE5 knockdown unexpectedly reduced the doxorubicin-induced caspase-3 cleavage but had no effect on PARP-1 cleavage (Fig.
16C). The PDE5 knockdown also showed different regulation from that of PDE5 inhibitors on several proteins markers of DNA repair (Fig. 17). Altogether, the data indicate that the mechanism of the sensitization effect caused by PDE5 knockdown was not similar to that of PDE5 inhibitors.
3.11. Effect of sildenafil on the sensitization of apoptosis induced by other topoisomerase inhibitors
In addition to doxorubicin, several other topoisomerase inhibitors were also examined for the effect of sildenafil-mediated chemosensitization on cell apoptosis. Cell apoptosis was determined by PI staining and flow cytometric analysis for the detection of sub-G1 population. The data in Figure 18 showed that sildenafil could significantly potentiate cell apoptosis induced by two other topoisomerase Ⅱ inhibitors, etoposide and mitoxantrone, but not by topoisomerase I inhibitor, camptothecin (Fig. 18).