About 4 years ago, Arnab Ray Chaudhuri, Ph.D., joined the lab of Andre Nussenzweig, Ph.D., as a postdoctoral researcher to study DNA mutations in breast cancers. Back then, the lab at the Center for Cancer Research at the National Cancer Institute was beginning to investigate how BRCA1 and BRCA2 mutations affect the repair of DNA double-strand breaks (DSBs) and the mechanisms that underlie their resistance to chemotherapeutic drugs.
Now, Ray Chaudhuri is part of the team’s decade-old hunt for a completely unexpected chemoresistance mechanism—how tumors with BRCA1 and BRCA2 mutations recruit a mix of proteins involved in DNA replication pathways to resist some common DNA cross-linking chemotherapies such as cisplatin or poly (ADP–ribose) polymerase, or PARP, inhibitors (PARPi).
Ray Chaudhuri was the lead author of the study (Nature 2016;535:382–7; doi:10.1038/nature18325). The work changes how scientists understand the complex interplay between normal and tumor DNA replication and could lead to new drug-testing strategies and therapeutic targets.
“[Using our discovery] oncologists can test and screen patients who have the propensity to become resistant by this mechanism and can devise new strategies for treatment when they already know that the patients are going to be resistant in the long term,” Ray Chaudhuri said.
DNA Replication Forks
In normal cells, BRCA1 and BRCA2 sense, survey, and facilitate the repair of damaged DNA upon replication by homologous recombination (HR). However, in people with BRCA1/2 mutations, HR is defective and the alternative error-prone repair mechanisms lead to malignancies (Nat. Med. 2013;19:1381–8; doi:10.1038/nm.3369). Mutations in BRCA1/2 account for 5%–10% of all breast cancers and up to 25% of hereditary breast cancers.
DNA cross-linking drugs disrupt tumors’ ability to repair breaks in the replicated DNA. However, sometimes tumors quickly become resistant to those drugs.
“Patients do not always carry this resistance phenotype, nor do the oncologists always know which mechanism is behind the patient’s negative response to the treatment,” Ray Chaudhuri said.
Tumors become chemoresistant because drugs put tumors under selective pressure and tumors evolve to resist them (Nature 2012;481:306–13; doi:10.1038/nature10762). In particular, tumors with BRCA1/2 mutations resist chemotherapy by reducing uptake of the drugs, more efficiently pumping drugs out of the cell, or restoring the HR pathways to repair the DSBs (Cancers 2014;6:1769–92; doi:10.3390/cancers6031769). The researchers looked into that last mechanism.
HR pathways in tumor DNA replication involve a horde of proteins and enzymes that help to stabilize and protect the DNA replication fork and its movement. In normal cells, when some proteins and DNA structures stall the replication fork, the BRCA1 and BRCA2 proteins take over and protect the newly formed DNA strands. But that protective process doesn’t take place in BRCA-deficient cells, disrupting the replication fork. Nucleases would then degrade newly formed DNA strands, creating genome instability. When DNA-damaging drugs attack the unstable genome, the tumor becomes sensitive to the drugs.
“[But] the tumors figure out insidious mechanisms to find a way to stabilize the DNA replication forks by limiting the nuclease degradation of the replication forks,” said Alan D’Andrea, M.D., who was not involved in the study. D’Andrea is Alvan T. and Viola D. Fuller American Cancer Society Professor of Radiation Oncology at Harvard Medical School and the Dana–Farber Cancer Institute, both in Boston.
The Protein Cocktail
In the study, Ray Chaudhuri and colleagues knocked out the genes that code for proteins that help stabilize the DNA replication forks without BRCA1/2 proteins. The absence of one gene gave them an unexpected result.
The group knocked out the ptip gene from the BRCA2-mutated tumors in cells as well as in mice and treated them with the drugs cisplatin and PARPi. Loss of the PTIP protein protected the cells from the drugs by stabilizing replication forks. Using DNA fibre spreading, they observed the replication forks and, using isolation of proteins on nascent DNA (iPOND) and flow cytometry, the recruitment of proteins onto the DNA replication forks. Ultimately, the team found that these replication forks prevent the recruitment of nucleases to degrade them.
PTIP was the usual target for Ray Chaudhuri’s team, because they were already studying 53BP1, another protein that interacts with PTIP. Although chemoresistance occurred after knocking out PTIP in BRCA1/2-deficient conditions, DSB repair did not. “This was the first surprise that we got,” Ray Chaudhuri said.
Then the team realized that the interaction of PTIP with the replication fork was independent of its role in degrading DSB repair—that is, the knockouts did not need BRCA2 to restore. Then it struck them that other mechanisms should be involved that do not require restoring DSB repair.
So the team moved on to find other factors that orchestrate chemoresistance in the same way: by stabilizing and protecting the replication forks. They found that two other proteins in addition to 53BP1, MLL3 and MLL4, interact with PTIP. MLL3 and MLL4 are histone methyl transferases that play a role in transcription. Moreover, researchers found that BRCA2-deficient cells lacked CHD4, a chromatin remodelling protein, which they observed helped stabilize the replication forks.
“This is also a surprise in a way because now we can link how DNA replication and transcription machinery basically talk to each other in bringing about genome instability,” Ray Chaudhuri said. The other protein that is part of the process is PARP1. The team also found that losing both PTIP and PARP1 not only protects the replication fork. The loss also restores the viability of BRCA-2 knockout embryonic stem cells to participate in chemoresistance.“We are learning that even though tumors initially respond to PARPi, they rapidly develop resistance to them. This is a novel and a scary mechanism.”
According to Philippe Pasero, Ph.D., principal investigator at the Institute of Human Genetics in Montpellier, France, the study challenges all the published reviews and textbooks so far that refer to BRCA2 as a gene essential for repairing chromosome breaks. To him the study suggests that the only essential function of BRCA2 is to protect the replication forks, not repair chromosome breaks. “This is new and important, because this is a totally different way of seeing the function of BRCA2,” Pasero said.
Way Forward
Though Ray Chaudhuri and his team know that PTIP, CHD4, and PARP1 are involved in chemoresistance, they do not yet know how—whether those proteins work independently or “talk to each other.” So now they are set to figure out how those proteins protect replication forks and bring about chemoresistance.
The team also aims to collaborate with cancer clinics and screen samples from patients for the stability of replication forks. “The idea in the long term is to screen patient tumors during or before the start of the treatment regimen. If we can test the stability of replication forks already in the laboratory, then we can predict if the patient’s tumor will be resistant to chemotherapy in clinics,” Ray Chaudhuri said. That is a big milestone to achieve. However, he said, the team is looking for a way to use that mechanism as a biomarker to predict chemoresistance.
Interest has grown in using PARPi against breast, ovarian, and even prostate cancers, which often have underlying defects in HR repair. About 10 years ago, scientists showed that PARPi can fight those cancers. In October, the drug company Tesaro announced the latest trial using PARPi. “But the problem is we are already learning that even though tumors initially respond to PARPi, they rapidly develop resistance to them,” D’Andrea said. Researchers already know a few mechanisms by which tumors resist PARPi. “[And] this is a novel and a scary mechanism,” D’Andrea said.
