The effect of PARP inhibition on DNA repair

Read Also

                As cells which have PARP-1 inactivated show an increase in RAD51 foci, it is thought that HR serves as the primary repair mechanism in situations where PARP-1 is inactivated (27).  Because of this redundancy, loss of PARP-1 activity alone does not appear to represent a lethal event; for example, PARP-1 deficient mice embryos are viable and survive gestation despite increased sensitivity to alkylating agents and radiation (28-30).  In contrast, in cells with non-functional HR-repair, loss of PARP function represents a lethal event as this redundancy has been lost.  This concept was clearly demonstrated in two separate reports exploring the effects of PARP inhibition in cells with inactive HR due to loss of key components of the HR complex.  Bryant and colleagues demonstrated that cells deficient in HR due to loss of XRCC2 and XRCC3, two proteins associated with the RAD51 complex, showed decreasing viability when exposed to increasing concentrations of the PARP inhibitors NU1025 and AG14361; this sensitivity was reversed when XRCC2 and XRCC3 activity was restored, confirming their hypothesis that HR-deficient cells are unable to compensate for BER loss.  Bryant and colleagues then demonstrated in vitro and in vivo that BRCA2 deficient tumors, either by mutation or siRNA, showed similar reduced viability with loss of PARP enzymatic activity (31).  Farmer and colleagues showed that BRCA1 and BRCA2-deficient cells demonstrated loss of viability when exposed to the PARP inhibitor KU0058684 in both in vitro and in vivo models.  Farmer also showed that PARP inhibition in BRCA1 and BRCA2-deficient cells resulted in significant genomic instability with complex chromatid rearrangements, suggestive that HR-deficient cells respond to loss of HR activity by repairing DSBs via NHEJ.  Due to the loss of genetic integrity from NHEJ, cellular viability is rapidly loss due to accumulation of genetic mutations in critical genes, leading to either G2/M cell cycle arrest or apoptosis (32).  PARP inhibition in HR-deficient cells is representative of a concept known as synthetic lethality, which is defined as the situation where the combination of loss of activity in two different genes results in cell death while the loss of either activity does not affect the viability of the cell (Figure 3) (33).

 

                Although it is thought that PARP inhibitors act by preventing the initiation of BER at SSBs, this is not universally agreed upon as some data suggests that PARP-1 activity may not essential for BER repair to take place.  First, mouse knockouts of BER repair proteins APE1, Pol-beta, and XRCC1 are not viable while PARP-1 knockouts survive, suggesting that BER is still functional despite the loss of PARP-1 (34).  Second, the alkylating agent dimethylsulfate resulted in accumulation of SSBs in cells exposed to PARP inhibitors but not in cells exposed to PARP-1 siRNA; this is unexpected as PARP inhibitors and loss of PARP-1 through siRNA should result in similar, not different, phenotypes (35).  Third, the steady state levels of SSBs do not appear to increase in either wild type or BRCA2 defective cells exposed to PARP inhibitors (36).  Overall, these observations are suggesting that PARP-1 plays a non-essential role in BER.  One caveat is that PARP inhibitors inhibit both PARP-1 and PARP-2 while PARP-1 siRNA does not.  Since PARP-1 and PARP-2 knockout mice are not viable (37), it is possible that the failure for PARP-1 siRNA to produce the same effects as a PARP inhibitor may be secondary to PARP-2 compensation.

                Based on the above observations, alternative mechanisms of action for PARP inhibitors have been proposed.  These proposals, as well as the evidence supporting them, are discussed in detail by Helleday and the reader is referred to his manuscript if greater detail on this topic is desired (34).  These alternative models for the mechanism of action for PARP inhibitors are illustrated in Figure 4.  First, it has been observed that overall BER kinetics can be significantly reduced in the presence of activated PARP-1 enzyme (38); therefore, one model is that PARP inhibitors work by preventing the release of PARP-1 from the SSB intermediate, trapping the complex at the site of repair.  Failure to release PARP-1 from the SSB repair intermediate physically prevents resolution of the lesion, resulting in replication fork arrest which can only be repaired by HR mechanisms (39).  A second possibility is that stalled replication forks can be repaired by either DSB-repair HR or a PARP-1-dependent HR-repair mechanism.  It has been observed that PARP depletion results in reduced recruitment of MRN, RPA and RAD51 to collapsed replication forks but not to HR-mediated DSB repair (18); this data suggests that stalled replication forks can either be repaired by a PARP-dependent HR process or the DSB-mediated HR repair process.  In cells which are HR-deficient, inhibition of the PARP activity prevents PARP-mediated restart of stalled replication forks, resulting in replication failure and synthetic lethality (34).