A surprisingly vast molecular emergency team responds to the everyday genomic nicks and dings caused by chemicals, foods, sunlight and other radiation, and oxygen byproducts, reports a new study. The research expands the sparse ranks of known DNA damage–response proteins from a couple dozen to a legion of more than 700.
“The list,” as researchers refer to the new database in the lab where it was developed, includes the usual suspects, such as BRCA1 (short for breast cancer 1). When defective, this protein may confer such a high risk of breast cancer that some women with certain gene variations and a strong family history of breast cancer choose prophylactic mastectomies. The database also contains surprises. For instance, an insulin-signaling pathway shared so many players that the researchers wonder if chronic DNA damage contributes to diabetes and other age-associated metabolic disorders.
“A cell spares no expense to modulate everything it needs to make sure it can optimally repair the genome,” said Stephen Elledge, a Howard Hughes investigator and the Gregor Mendel professor of genetics and of medicine at HMS, who is a senior author of the paper on the DNA damage–response network and two related papers. “We have a much more expansive view of what cells do now when they are threatened by a broken chromosome or other insults to the DNA. It’s a much broader landscape than we anticipated.”
In the two other papers, the list helped one researcher identify the last of 13 known genetic mutations responsible for the devastating childhood condition Fanconi anemia. Another researcher found a new molecular partner of BRCA1, which may lead to fresh insights into cancer mechanisms and possible new treatments.
Thanks to a longstanding fascination for the double helix and what it represents, scientists know a lot about how it is replicated, maintained, and repaired, Elledge said. A complex signaling apparatus senses damage, pauses the cell cycle, and repairs the DNA. Specialized repair pathways can remove and replace a mismatched base or an entire patch around the nucleotide, reconnect a broken chromosome, and remove the cross-links that block replication and transcription. No matter what the damage, a pair of related DNA damage sensors, ATM and ATR, sets the response in motion.
Postdoctoral fellow Shuhei Matsuoka started at the top of this signaling network by asking, What proteins are ATM and ATR calling into action? It took a decade for researchers in the Elledge lab and others to tease out the first 25 damage response molecules known to be directly activated.
Matsuoka devised a shortcut to round up what he figured were the few remaining players by combining unique techniques and tools from many labs. He rounded up 68 antibodies to the phosphorylation sites in the two known activation patterns on many of the proteins.
New mass spectrometry techniques developed by Steven Gygi, an HMS associate professor of cell biology and co–senior author on the paper, fingered hundreds of activated proteins in the irradiated cells compared to those from undamaged cells that had been labeled with a stable isotope. A stunning 700 proteins carried a fourfold higher phosphorylation count. The results were convincing, but for further validation Matsuoka and his colleagues tested 37 of the proteins, blocking them one by one with small interfering RNA in damage assays. Of them, 35 proteins seemed to be important, suggesting a 95 percent accuracy of the list.
“You still have to test them all,” Elledge said. “You have to do experiments where you change one variable at a time.”
Of the 700, only 421 had biological purposes known by science. Many of those made sense in the DNA damage response, especially the ones playing some part in replication, recombination, or repair. The big surprise was the proteins in other categories, such as RNA splicing or protein degradation or protein synthesis or intracellular protein traffic.
“This is going to happen more and more,” Elledge predicted. “Lots of proteins have multiple functions, some of which are completely unrelated.”
The list may speed research into the DNA damage response and related diseases, including cancer. “We cannot work on 700 proteins in one lab; studying one protein a year would take 700 years,” Matsuoka said. But in the scientific community, “each researcher has a protein of interest. We have published the list for others. If 700 people each work on one protein for a year, they can finish the list. They have a chance to find new genes involved in a tumor or disease.”
One unknown protein on the list caught the attention of colleague Agata Smogorzewska, a clinical pathology resident at Massachusetts General Hospital on a research fellowship in the Elledge lab. She helped test the sample of 37 proteins on the list. Cells missing one particular uncharacterized protein limped along after double-stranded DNA breaks from gamma radiation and fared miserably after exposure to a chemotherapy-like cross-linking agent. The pattern was characteristic of Fanconi anemia, a cancer-prone condition of extreme sensitivity to DNA cross-linking agents that often leads to bone marrow failure in childhood.
“I became somewhat obsessed with finding out why it was a strong phenotype,” said Smogorzewska, who worked at night on the computer while her days were occupied by another project. “In science, once you find something striking, you have to follow it up.”
She searched in silico for a similar protein in databases of sequenced genomes ranging from pond scum to people. One sea urchin protein in that creature’s version of the Fanconi repair pathway popped out, confirming her instincts. Smogorzewska tested her protein in cells from a group of Fanconi anemia patients with an unknown gene mutation. The protein restored a healthy DNA damage response and earned the name FANCI, the name given to that subset of Fanconi anemia.
“This fits in with the canonical DNA damage–response protein,” Smogorzewska said, “but we still don’t know why and how others on the list are important.” Her paper was published in the April 20 Cell. All the other papers appear in the May 25 Science.
Meanwhile, Bin Wang, a postdoctoral fellow in the Elledge lab, tried a variation of Matsuoka’s technique to look for other proteins that normally hook up with BRCA1 at a crucial spot, known as the BRCT domain, site of most cancer-causing mutations. Hundreds of proteins made contact, but a cross-reference with Matsuoka’s DNA damage–response database narrowed the pool to three obvious candidates.
Two of the proteins form mutually exclusive liaisons with BRCA1 and work to suppress tumors, others had shown. But the third protein, Abraxas, was a new find. Wang showed it forms a third kind of complex and may also be a tumor suppressor. Further experiments found another protein, Rap80, on board and necessary for BRCA1’s DNA repair response. (Other researchers have shown Rap80 binds to the estrogen receptor, perhaps explaining why certain mutations translate to breast and ovarian cancer.)
The three different protein complexes may each direct one of the different BRCA1 functions—DNA repair, transcription, and a cell cycle checkpoint, Wang said. “We are really interested in knowing whether these proteins are involved in breast cancer progression” and “whether breast cancer patients have mutations in these other proteins in the pathway.”
Two companion papers in Science from other groups, including one from the DFCI lab of David Livingston, independently demonstrate the importance of Rap80 in conveying BRCA1 to the small clusters in the nucleus that are somehow essential to a cell’s normal DNA repair. BRCA1 proteins congregate at spots of damaged DNA, but how they get to the foci and whether those dots are storage depots, work sites, or both is a mystery. The ubiquitin binding domains of Rap80 properly dock BRCA1 to the dots at a polyubiquitinated target protein, helping “cells become more resistant to the potentially lethal effects of DNA damage,” Livingston said. “These papers give us a little molecular insight into how these dots form.”
