Peering into the early stages of cancer is tricky business. Oxidative stress, proliferation signals, and loss of suppressive signals from the microenvironment have all been fingered as initiating elements in the process. As DNA damage is often seen in full-blown tumors, its role has been considered, but whether it is cause or consequence of ramped-up cellular proliferation had been unclear.
In 2005, teams led by Thanos Halazonetis, now at the University of Geneva, and Jiri Bartek at the Institute of Cancer Biology in Copenhagen revealed evidence that cellular responses to DNA damage, specifically to double-strand breaks, are activated early in precancerous lesions.
Polling Cancers for Answers
Both groups examined clinical specimens from different stages of human cancer. Bartek and his colleagues looked at tumors of the urinary bladder, breast, lung, and colon, while Halazonetis and his collaborators studies lung hyperplasias. They found that all these cancers expressed markers of DNA-damage response, including the phosphorylated kinases ATM and Chk2, as well as phosphorylated histone H2AX and p53. This included the precancerous lesions, where cells had neither become fully invasive nor displayed signs of gross chromosomal instability.
"People had supposed that DNA damage occurred late in tumorigenesis. The finding that it was this early was a surprise," says Fabrizio d'Adda di Fagagna at the FIRC Institute of Molecular Oncology Foundation in Milan.
To see what might trigger such a DNA-damage response, Halazonetis and his colleagues induced hyperproliferation of human skin cells grafted onto the backs of immunodeficient mice using viral vectors carrying genes for growth factors. Bartek and his collaborators overexpressed oncogenes such as cyclin E in cultured cells. Vassilis Gorgoulis at the University of Athens, a coauthor with Halazonetis, says that genomic instability was evident within four weeks after administering growth factors.
The teams proposed a model for cancer development: Oncogene activation disturbed DNA replication, which in turn led to DNA damage and genomic instability. The DNA-damage response normally activated p53 to help keep precancerous lesions under control. "The impact of these papers has been huge," d'Adda di Fagagna says. "You cannot possibly study tumor formation now without considering DNA breaks."
Supporting the initial findings are two recent papers, one from Halazonetis, Bartek, Gorgoulis, and their colleagues, and another from d'Adda di Fagagna and his collaborators.
Halazonetis notes that DNA replication stress and the DNA damage response could "distinguish the majority of cancers from normal tissues." This raises the possibility that detecting the DNA-damage response in cells could help physicians spot cancer early, and that somehow bolstering the DNA-damage response could help impede cancer.
Remaining Concerns
Still, researchers question whether the DNA-damage response actually does trigger p53 tumor suppression. Teams led by Evan and by Manuel Serrano at the Spanish National Cancer Center in Madrid independently found that ARF, a tumor suppressor triggered by oncogenic disruption of the cell cycle and not by DNA damage, was sufficient to activate p53.
"Although it makes perfect sense that DNA-damage signaling should be a tumor suppression mechanism, I think that the evidence for this is still missing," Serrano says. For instance, he points out that "p53 is the integrator of so many stresses that it is not possible to say that the key stress that activates p53 in cancer is DNA-damage signaling." He also notes: "It remains to be demonstrated that cells or mice with enhanced DNA-damage signaling are better protected from cancer."
Evan emphasizes that their results do not exclude the possibility that the DNA-damage response is important in tumor suppression. d'Adda di Fagagna notes, "Observations concerning DNA-damage response are being reproduced more and more with different tumor types and oncogenes. But the picture is very complicated. It seems that the DNA-damage response may play a stronger role in some tumors and less with others."