The underlying mechanism for the PARP and BRCA synthetic lethality: clearing up the misunderstandings

Abstract

Poly (ADP-ribose) polymerase (PARP) inhibitors effectively kill tumours defective in the BRCA1 or BRCA2 genes through the concept of synthetic lethality. It is suggested that PARP inhibitors cause an increase in DNA single-strand breaks (SSBs), which are converted during replication to irreparable toxic DNA double-strand breaks (DSBs) in BRCA1/2 defective cells. There are a number of recent reports challenging this model. Here, alternative models that are not mutually exclusive are presented to explain the synthetic lethality between BRCA1/2 and PARP inhibitors. One such model proposes that PARP inhibition causes PARP-1 to be trapped onto DNA repair intermediates, especially during base excision repair. This may in turn cause obstruction to replication forks, which require BRCA-dependent homologous recombination to be resolved. In another model, PARP is directly involved in catalysing replication repair in a distinct pathway from homologous recombination. Experimental evidence supporting these novel models to explain the PARP-BRCA synthetic lethality are discussed.

Figure 1

Base excision repair (BER) is a separate process from DNA single‐strand break (SSB) repair in mammalian cells, although the two processes share proteins. (A) SSB repair: PARP‐1 has a high affinity for SSBs and will be amongst the first proteins to bind to the lesion. In turn PARP recruits factors to start end processing and finally ligation, normally through short patch repair and through long patch repair where the lesions are more difficult to repair. (B) Two‐step model for BER: Different base lesions are recognised by different glycosylases (Gly), which are excised before SSB incision by the AP‐endonuclease (APE). These SSBs are then left unprotected and recognised in a separate process by PARP‐1 that will then initiate SSB repair. (C) One‐step model for BER: The glycosylase interacts with proteins involved in the early BER incision step and excises the damaged base shortly before APE incision. The half‐life of the SSB intermediate is very short and rapidly ligated by short patch repair, which switches to long patch repair in case of ligation difficulty. PARP‐1 has no role in BER, but can transiently bind the SSB intermediate. When PARP‐1 is inhibited, it can be trapped on the SSB intermediate and prevent the ligation step.

Figure 2

Pathways underlying PARP‐BRCA synthetic lethality. (A) SSB replication run‐off model. PARP‐1 is involved in repair of SSBs, which may in the presence of a PARP inhibitor persist and collapse a replication fork into a one‐ended DSB. Since BRCA defective cancer cells lack HR, the resulting DSBs would be selectively toxic to the cancer cells. (B) PARP‐1 trapping model. PARP inhibitors trap PARP‐1 onto SSBs formed spontaneously or as an intermediate during BER. Trapped PARP‐1 may pose an obstacle to replication that would require HR to bypass. (C) Replication restart model. In the case of normal replication, forks will stall owing to lack of replication factors or by obstacles on the DNA template. PARP and HR are activated at stalled forks and mediate distinct pathways for restart.