Defects in genes crucial for the process of DNA repair by homology-directed DNA repair (HDR), such as BRCA1 and BRCA2, are well-known contributors to cancer pathogenesis as well as an Achilles' heel that can be exploited therapeutically. BRCA1/2-deficient cells are exquisitely sensitive to agents that stall replication forks, such as PARP inhibitors and platinating drugs, presumably due to the inability to repair double-stranded breaks that form as a consequence of replication fork collapse. BRCA1/2 also promote tolerance to DNA replication stress by protecting replication forks from nucleolytic degradation. Both biological endpoints involve the deposition of RAD51 onto single-stranded DNA (ssDNA) for homology searching and strand exchange during HDR repair, as well as protection of newly synthesized DNA from nucleolytic degradation (i.e., replication fork protection). In this issue of Cancer Research, Panzarino and colleagues performed multiple separation-of-function studies and identify the lesion most associated with intolerance to replication stress in BRCA1/2-deficient cells is persistent ssDNA gaps in newly synthesized DNA, resulting from a failure to restrain DNA replication. Mechanisms that suppress gap formation are closely associated with chemoresistance, and the authors’ findings challenge the paradigm that lack of HR repair or fork protection underlie the phenotype known as BRCAness.
See related article by Panzarino et al., p. 1388
The term “BRCAness” refers to any cancer that exhibits defects in homology-directed DNA repair (HDR) and is, therefore, predicted to be hypersensitive to PARP inhibitor (PARPi) and platinating agents (1). Inhibiting PARP activity or trapping the enzyme on single-stranded DNA (ssDNA) breaks, thereby inhibiting ssDNA break repair, creates lesions that impede the normal progression of DNA replication, causing replication forks to collapse and double-stranded breaks (DSB) to form. Platinating agents, such as cisplatin and carboplatin, physically restrict normal DNA replication through the covalent intrastrand crosslinking of DNA, thereby increasing replication stress. Platinating agents also create a small percentage of interstrand crosslinks in the DNA double helix. Repair of these lesions is complex, involving multiple repair pathways that include the Fanconi anemia (FA) pathway, nucleotide excision repair, and HDR. HDR is considered an error-free process by employing a sister chromatid as a template to accurately fix a DSB. BRCA1 governs the resection of DNA in a cell-cycle–dependent manner by restricting access of the nonhomologous DNA end-joining factor 53BP1 to the ends of DSBs, thereby permitting the generation of recessed DNA with 3′ overhangs (2). BRCA1 also binds PALB2, which in turn directs BRCA2 to promote RAD51 filament formation on the recessed DSB, a critical step necessary for homology searching, strand invasion, and repair synthesis during HDR.
Replication stress is generally defined as the slowing or stalling of replication fork progression and is often associated with stretches of ssDNA due to the continued action of helicase parental strand unwinding despite DNA polymerase inhibition (3). Persistent ssDNA attracts replication factor A (RPA), which serves as a platform to activate the ATR kinase. ATR orchestrates multiple responses including cell-cycle checkpoint control, inhibition of new origin firing, and the stabilization of stalled replication forks. Specifically, how replication forks are stabilized or protected has become an intense area of investigation beginning with the seminal finding that BRCA2 prevents the formation of DSBs within regions of active DNA replication in cells treated with hydroxyurea (4). Employing DNA fiber analysis, which measures the continuity of DNA replication at the molecular level, Schlacher and colleagues discovered that BRCA2 prevents nucleolytic degradation of newly synthesized DNA by the MRE11 nuclease (5). The ability of BRCA2 to deposit RAD51 onto ssDNA in place of RPA is crucial for this activity and is genetically separable from BRCA2′s role in HDR. The RAD51-mediated protection of nascent DNA from MRE11 nucleolytic processing also involves BRCA1 and key members of the FA pathway (2). Thus, recruitment of MRE11 to stalled replication forks can be toxic, and multiple factors cooperate to limit MRE11 access to these lesions.
Consistent with the model that HR repair deficiencies render BRCA1/2-deficient cancers sensitive to cisplatin and/or PARPi, acquisition of resistance to these agents often results from secondary mutations that restore HDR repair. However, drug resistance in the absence of restored HDR has also been associated with the loss of factors that stimulate MRE11's recruitment to stalled replication forks (e.g., PIP1, PARP1, and CHD4) or promoters of replication fork reversal (e.g., SMARCAL1 DNA translocase; refs. 2, 6, 7). Fork reversal is one of several DNA damage tolerance mechanisms where, upon replication stress, translocases backtrack the replication fork into four-way structures: the two nascent DNA strands anneal, forming a fourth arm that requires BRCA1/2 and RAD51 for protection from nucleases such as MRE11. Of clinical relevance, low expression of PIP1 and SMARCAL1 predict shorter progression-free survival in patients with BRCA1-mutated cancers, whereas low PIP1 and CHD4 levels correlate with reduced survival in patients with BRCA2 mutations (6–8).
To gain additional insight into BRCA1/2 function at the replication fork, Panzarino and colleagues adopted a modified DNA fiber assay to analyze altered replication dynamics in cells challenged with replication stress (9). When assessing replication fork protection, cells are sequentially incubated with the thymidine analogs iodo-deoxyuridine (IdU) and chloro-deoxyuridine (CldU), and then challenged with agents that create replication stress, such as hydroxyurea. Following differential staining of denatured DNA fibers, the lengths of each tract are compared. Normally, shortened CldU tracts are presumed to have resulted from degradation, because CldU incorporation occurs just prior to hydroxyurea exposure. Instead, Panzarino and colleagues performed the CldU labeling step during replication stress rather than before it. Using this approach, they unexpectedly found that BRCA1/2-deficient cells produced longer CldU tracts, suggesting an inability to slow down DNA replication. These longer tracts are also sensitive to S1 nuclease digestion, indicating that the hyper-replicated tracts contained ssDNA gaps. A patient-derived xenograft model selected for cisplatin resistance also exhibited reduced gap formation, indicating that gap suppression may be relevant to acquired resistance in vivo. These observations raised the question as to whether gap formation during replication stress is an important determinant of chemosensitivity and if fork protection plays a role in gap suppression.
Because depletion of the chromatin remodeler CHD4 can restore fork protection and increase chemoresistance in BRCA2-deficient cancers, Panzarino and colleagues tested whether CHD4 depletion could suppress gap formation. Indeed, CHD4 depletion reduced the formation of ssDNA gaps during replication stress in BRCA2-deficient cells despite hyperactivating DNA replication. In addition, the group identified additional CHD4 interactors (i.e., EZH2, FEN1, and ZFXH3) that phenocopy CHD4 loss in promoting gap suppression and chemoresistance, as well as predicting patient survival. Given the protective effect loss of CHD4 conveys in BRCA2-deficient cells, the group used several model systems to tease apart contributions of fork protection versus gap suppression to chemoresistance. BRCA2-deficient cells expressing the BRCA2 S3291A mutant, which lacks fork protection activity, exhibit chemoresistance along with ssDNA gap suppression. In addition, disrupting nascent DNA degradation by depletion of MRE11 or the SMARCAL1 fork remodeler failed to rescue gap suppression or to confer cisplatin resistance.
Is HDR repair still relevant to chemoresistance? A RAD51 dominant-negative mutation identified in a FA patient (RAD51T151P) is proficient in HDR, but defective in protecting replication forks from MRE11-mediated degradation (2). Despite functional HDR, these cells remain hypersensitive to DNA crosslinking agents. Using this as a model system, Panzarino and colleagues found that these cells produce long replication tracts that are sensitive to S1 nuclease digestion during conditions of replication stress consistent with their model. They then asked the question as to whether chemoresistance could be acquired if fork protection were restored. Depletion of RADX, a RAD51-negative regulator, restores fork protection in RAD51T131P or BRCA2-deficient cells by permitting enhanced RAD51 activity at stalled replication forks, thereby inhibiting MRE11 access (2). Depleting RADX restored fork protection in RAD51T131P cells as expected, but failed to increase chemoresistance or suppress ssDNA accumulation. Thus, the creation of ssDNA gaps most closely correlates with chemosensitivity in multiple model systems.
Through the use of multiple separation-of-function experiments, Panzarino and colleagues identified ssDNA gaps formed during unrestrained DNA replication as the lesion most associated with chemosensitivity in BRCA1/2-deficient cells. They also provide convincing arguments that inability to repair DSBs by HDR fails to explain hypersensitivity to therapies used to selectively target BRCA1/2-mutant cancers. Instead, the authors provide substantive evidence that appearance of DSBs could in fact be the byproduct of apoptosis. Consistent with their model, the inability to suppress ssDNA gaps most closely correlates with sensitivity to PARPi (10). These findings raise numerous questions and will certainly stimulate new directions of research. For example, it remains unclear how BRCA1 and BRCA2 suppress replication fork progression under conditions of replication stress and prevent ssDNA gap formation. The roles the CDH4 chromatin remodeler and associated partners play in supporting hyper-replication and the associated accumulation of gaps remain to be clarified. Translesion DNA synthesis may play a role in filling in the gaps left behind from unrestrained replication as the authors suggest, however, it is unclear how all these pathways interact with one another to suppress gap formation leading to chemoresistance.
In summary, Panzarino and colleagues provide convincing evidence that the failure to prevent ssDNA gap formation is the key vulnerability in cancer cells displaying BRCAness and identify new factors that may better predict patient response. This study also raises the possibility that targeting factors that promote gap suppression may restore chemosensitivity to resistant tumors.
No disclosures were reported.