In this issue of ONCOLOGY, Comen and Robson provide a timely overview of poly(ADP-ribose) polymerase (PARP) inhibitors and their potential for the treatment of breast cancer. The authors highlight the recent demonstration of synthetic lethality between PARP inhibition and loss of either of the breast cancer susceptibility genes, BRCA1 and BRCA2, as well as the development of PARP inhibitors that are suitable for clinical therapy. However, many questions pertaining to both the basic biology of PARP inhibition and the potential clinical implications of PARP inhibitors still need to be addressed. In the following commentary, we highlight some of these remaining challenges.
PARP Inhibition: An Incomplete Picture
PARPs are a family of proteins that catalyze the conversion of nicotinamide(Drug information on nicotinamide) adenine dinucleotide (NAD+) into long poly(ADP-ribose) (PAR) chains. These highly negatively charged PAR chains can be either covalently or noncovalently attached to acceptor proteins, thereby altering their function. The most extensively studied member of the PARP family, PARP1, has been implicated in a variety of cell processes. In particular, PARP1 acts as a sensor of DNA damage and initiator of the base excision repair pathway.[1] It is this role in base excision repair that is postulated to be responsible for the synthetic lethality of PARP inhibition and BRCA1/2 loss. It is important to realize, however, that this model neglects the other roles of PARP1 in cellular survival. PARP1 also has a role in restarting stalled replication forks, inhibiting nonhomologous end-joining repair, regulating transcription, initiating a unique cell death pathway, and modulating cellular bioenergetics.[2] Elucidating how inhibition of PARP1 affects these processes is critical to fully understanding the effects of PARP inhibitors.
In this context we want to draw attention to one potentially important long-term effect of PARP inhibition that has not been widely discussed. Older studies demonstrated that genetic ablation of PARP1 in conjunction with p53 knockout increases the cancer incidence in mice.[3] Thus, PARP1 may function as a tumor-suppressor protein in cells; and PARP inhibition could conceivably increase the risk of secondary malignancies, particularly when PARP inhibitors are combined with genotoxic agents such as platins or topoisomerase I inhibitors. Although few previous studies have addressed the long-term effects of PARP inhibitors, this will become important as these agents progress through clinical trials. These long-term safety studies are particularly important because preclinical results suggest a potential use of PARP inhibitors as preventive therapy for patients with demonstrated heterozygosity at a BRCA locus.[4]
It is also important to emphasize that effects of PARP inhibitors are unlikely to be limited to PARP1. To date, 17 members of the so-called PARP family have been identified based on sequence homology to the catalytic domain of PARP1; and six of these enzymes are known to catalyze PAR synthesis.[2,5] Aside from PARP1 and PARP2, which are both involved in the response to DNA damage, the functions of many of these PARP enzymes remain incompletely understood. Both PARP1 and PARP2 are inhibited by the current generation of PARP inhibitors, but the effects of the inhibitors on the other family members remain to be studied. These investigations are required to assess the possibility that varying effects of PARP inhibitors on the other PARPs might contribute to the different side-effect profiles of the PARP inhibitors now in the clinic. In addition, further study of the other PARPs might also help identify new drug targets. Synthetic lethality was recently demonstrated between the BRCA genes and Tankyrase1, one of the less intensively studied PARP family members,[6] suggesting that BRCA1/2 mutant tumors might also be successfully targeted without inhibiting PARP1.
An Expanding Role for PARP Inhibitors
Even with the recent demonstration that PARP inhibitors are active against BRCA1/2-mutant tumors, one might legitimately ask whether clinical testing of six different inhibitors is required to thoroughly study this target in the clinic. In fact, it appears that not all PARP inhibitors are created equal. In addition to potential differences in their effects on enzymes other than PARP1, there are other potential differences between these inhibitors. One PARP inhibitor undergoing clinical testing (4-iodo-3-nitrobenzamide, now known as BSI-201) inhibits PARP1 and possibly other enzymes through an irreversible, covalent modification.[7] As a consequence, restoration of PARP1 activity requires resynthesis of the enzyme.
This contrasts with the other PARP inhibitors currently in development, which are thought to compete with the enzyme substrate NAD+ at the active site. The unique mechanism of BSI-201 may account for its apparently unique ability to synergize with gemcitabine(Drug information on gemcitabine) (Gemzar).[8] In addition, the irreversible inhibition of PARP1 reported with BSI-201 justifies the decision to test BSI-201 against tumors expressing high levels of PARP1. In contrast, the same reasoning would dictate that the noncovalent, reversible PARP inhibitors might be better suited for treating tumors expressing low levels of PARP1. Whether the unusual mechanism and structure of BSI-201 translate into observable differences in clinical efficacy and/or toxicity will be interesting to watch.
Although Comen and Robson emphasize the potential role of PARP inhibitors in BRCA1/2-mutant tumors, recent preclinical studies suggest that PARP inhibition might be beneficial in a much larger subset of tumors than initially thought. Cells harboring a variety of genetic impairments in homologous recombination—not just BRCA1/2 loss—are remarkably sensitive to PARP inhibitors in vitro.[9] Perhaps most intriguingly, cells that lack the tumor suppressor PTEN are highly sensitive to PARP inhibitors because of downregulation of Rad51, a critical homologous recombination component.[10] This is particularly exciting considering the high incidence of PTEN inactivation in human tumors. As a consequence, we look forward to seeing the results of clinical trials of PARP inhibitors in PTEN-deficient tumors in the near future.
As the story on PARP inhibitors continues to unfold, it is clear that much remains to be learned. However, the potential of these agents to produce clinical benefits in BRCA-deficient tumors, and perhaps other cancers, is tantalizing.
Financial Disclosure: The authors have no significant financial interest or other relationship with the manufacturers of any products or providers of any service mentioned in this article.
