seaborg921

Lawrence C. Brody, Ph.D.
When it comes to antitumor drugs, both the physicians who prescribe them and the patients who receive them would agree that better chemotherapeutic agents are needed. Drugs in common use were designed to take advantage of general characteristics of tumor cells such as hormone responsiveness or high rates of cell division. Unfortunately, these properties are also found in some healthy cells. This general lack of specificity has led to the use of drug doses that straddle a knife edge between efficacy and toxicity. Two recent studies, one by Bryant etal.1 and the other by Farmer et al.2 suggest that it is possible to achieve specificity without toxicity by adopting a new approach to targeting tumor cells.
Much time and energy have been spent searching for the biologic bits that selectively mark tumors, with the aim of achieving drug specificity. In the molecular age, attention has been focused on isolating and characterizing gene products that are specific to cancer cells. There have been three high-profile success stories of drugs targeted to tumorspecific molecular changes: imatinib mesylate (Gleevec) for chronic myelogenous leukemia and gastrointestinal stromal tumors, gefitinib (Tarceva and Iressa) for non–small-cell lung cancer, and trastuzumab (Herceptin) for breast cancer. Although each drug has serious shortcomings, they share a common mechanism in that each targets a specific molecule present in a tumor whose activity drives tumor growth.

Might there be a yin to this yang? Is it possible to target a weakness — rather than a strength — of the tumor cell? Consider what is known about genes responsible for inherited forms of cancer; most are tumor-suppressor genes. Inherited mutations eliminate or suppress activities that prevent tumor formation.
In the case of breast cancer, mutant copies of BRCA1 and BRCA2 may be responsible for 3 to 8 percent of all invasive breast tumors. The protein products of both genes participate in DNA repair.Tumors arise from cells that have lost their remaining good copy of the gene and, hence, the ability to repair specific lesions in DNA (resulting in cells that are genetically unstable).
Can this compromised repair pathway be further stressed so as to push these unstable cells over the edge? Lesions in DNA can take several forms.To ensure that the genome remains intact, evolution has delivered an array of DNA-repair systems.
BRCA1 and BRCA2 are components of a complex that identifies and repairs breaks in double-stranded DNA. Bryant et al.1 and Farmer et al.2 hypothesized that cells deficient in the ability to repair breaks in double-stranded DNA might be killed by flooding them with breaks. The classic generator of double-strand breaks is ionizing radiation, but radiation (and the chemotherapeutic drugs that mimic it) is nonselective and also damages normal cells, paradoxically increasing the probability that they will be transformed into tumors.
The two groups drew on the fact that an independent complex of proteins is responsible for repairing a second type of DNA damage, single-strand breaks (Fig. 1). Previous work had shown that mice engineered to lack a key enzyme, called poly–adenosine diphosphate–ribose polymerase 1 (PARP1),in the pathway that mediates the repair of singlestrand breaks are healthy and fertile.3 Cells from these mice have large numbers of unrepaired singlestrand breaks, which on DNA replication are converted to double-strand breaks and then tidied up by the complex that repairs double-strand breaks.
What would happen to a cell lacking the ability to repair both types of DNA breaks? With PARP1 inhibitors and dishes of cells in hand, Bryant et al.1 and Farmer et al.2 answered this question. As predicted,wild-type cells shrugged off the loss of PARP1 activity and continued happily to divide. In contrast, dishes of cells lacking either of the BRCA genes were killed by very low doses of the PARP1
inhibitors. Some cell lines were 1000 times as sensitive as wild-type cells. Another important finding was that cells with only one good copy of BRCA1 or BRCA2
(these represent noncancerous cells in persons carrying mutant BRCA1 or BRCA2
) had the same behavior as wild-type cells.
Would this trick work in vivo? Although details differ, both groups took the logical next step. They implanted cells with BRCA2 and cells without BRCA2 into mice. Because they started with transformed cells, all xenografts formed tumors. An identical set of mice received the grafts and a PARP1 inhibitor. In treated mice, tumors with BRCA2 continued to grow but BRCA2-deficient tumors did not grow, regressed, or disappeared. No side effects related to the PARP inhibitor were noted, leading one to speculate that these drugs may have no side effects if used as an experimental therapy.
The studies by Bryant et al.1 and Farmer et al.2 could influence clinical practice at several levels.Most immediately, they suggest that the breast and ovarian tumors that develop in women who have inherited BRCA mutations might have a response to treatment with PARP1 inhibitors. If the model holds true, nontumor cells (retaining one copy of a BRCA gene) would be spared. Some data suggest that the BRCA genes are turned off in noninherited forms of breast cancer. If this silencing is stable, PARP1 inhibitors could also be used to treat more prevalent,
nonfamilial forms of breast cancer. In addition,these agents might be used more broadly to treat tumors with defects in other components of the complex that repairs double-strand breaks. Finally, the approach of targeting pathways that tumors lack may warrant more attention. Seeing what is not there requires us to look at tumors in different ways.

Figure 1. Model of Tumor-Cell Killing by PARP1
Inhibitors.
The strategy used by Bryant et al.1 and Farmer et al.2 takes advantage of the mechanism through which damaged
DNA is repaired by the cell and invokes the adage about “the straw that broke the camel’s back.” BRCA1 and BRCA2 proteins are part of a complex that recognizes and repairs double-strand breaks. Cells deficient in these proteins (such as those with mutations in both copies of BRCA2) are unable to repair double-strand
breaks and therefore become genetically unstable and prone to tumorigenesis. Bryant et al.1 and Farmer et al.2 hypothesized that cells would die if their ability to repair both single-strand breaks, which are caused by environmental stress, and double-strand breaks, was compromised. PARP1 is a vital part of a complex that repairs single-strand breaks. Consistent with the authors’ hypothesis is the arrested progression of BRCA2-deficient tumors in mice treated with PARP1 inhibitors.

1.Bryant HE, Schultz N, Thomas HD, et al. Specific killing of BRCA2-deficient tumours with inhibitors of poly(ADP-ribose) polymerase. Nature 2005;434:913-7.
2. Farmer H, McCabe N, Lord CJ, et al. Targeting the DNA repair defect in BRCA mutant cells as a therapeutic strategy. Nature 2005; 434:917-21.
3. Conde C, Mark M, Oliver FJ, Huber A, de Murcia G, Menissierde Murcia J. Loss of poly(ADP-ribose) polymerase-1 causes increased tumour latency in p53-deficient mice. EMBO J 2001;20: 3535-43.