By David N. Leff

As recently as Monday, a national television news program reported that the average age of puberty in the U.S. has taken a sudden dip from 9 years of age toward 8 or 7.

That¿s the sequence of events whereby a child becomes a young adult, physiologically capable of parenting kids of her own. ¿In female development,¿ observed molecular geneticist and microbiologist Tanya Paull, ¿the breast and ovarian epithelial tissue undergo very rapid proliferation. Those tissues are dividing quickly during puberty. The clonal expansion of their cells in that short amount of time requires those cells to replicate fast, and through many rounds of expansion.¿

Paull made this observation in connection with the apparently linked occurrence of familial breast and ovarian cancers, caused by the two well-known genes, BRCA1 and BRCA2. Far less well known are the tumor suppressors they express ¿ Brca1 and Brca2.

¿Why are these BRCA1 and BRCA2 genes specializing in mammary and ovarian cancers?¿ Paull asked rhetorically. ¿That¿s still a major question, and nobody really has a good answer for it. Both genes are expressed everywhere in the body, and one hypothesis is that with the onset of puberty, their mammalian and ovarian tissues move into overdrive.¿

Paull is first author of an interim report in the Proceedings of the National Academy of Sciences (PNAS), dated May 22, 2001, but released today. Its title: ¿Direct DNA binding by Brca1.¿

¿We found that the Brca1 protein itself can bind directly to DNA,¿ Paull told BioWorld Today, ¿and that it has a particular affinity for some branched and unusual DNA structures. Brca1 has been implicated in a number of different processes,¿ she continued, ¿and one of those is repairing double-strand breaks in the DNA helix.

¿You can acquire DNA breaks just through the replication cycle,¿ she observed. ¿So if you¿re partially deficient in some way in repairing those breaks, then multiple rounds of expansion ¿ as in female puberty ¿ could be particularly bad for a tissue. That idea is consistent with the mouse knockout model. A number of groups have knocked out the BRCA1 gene in mice. That causes lethality. The animals die during prenatal development, but at a stage right where they¿re about to go into very rapid proliferative growth.¿

Clues To Cancer Connection

¿And that,¿ Paull suggested, ¿seems to be consistent with proliferation of cells in general. It gives another sense of why this process may be linked to breast and ovarian cancer. Most tumor cells,¿ she added, ¿are dividing rapidly. But these in particular relate to puberty as possibly linked to their hyper-speedy division. Of course, it¿s just a hypothesis. So I don¿t want to make too much of it,¿ she allowed. ¿It could be very important, but we¿re only at the beginning of our work.

¿Double-strand DNA breaks,¿ she went on, ¿can be introduced through a number of means. One is ionizing radiation; another, exposure to carcinogenic chemicals. In addition, it¿s becoming clear that double-strand breaks can occur in the absence of any such exogenous agents. They can happen just through self-replicating a cell¿s own DNA. This occurs during DNA replication. The strands are unwound and then each of them is a template for new strands. That replication fork,¿ she recounted, ¿moves through an individual¿s entire genome. Some do so simultaneously.

¿That DNA is called a fork,¿ Paull explained, ¿because it¿s shaped like a letter Y, as the two strands are being unwound from one end and are still together on the other end. This will enable new strands to be synthesized on each of the original strands. So it moves, unwinding as it goes, through the genome.

¿It¿s become clear,¿ she pointed out, ¿that even during a single replication cycle in a cell, that fork may become blocked, or stalled ¿ like a train on a track hitting a wall ¿ because of some kind of impediment or discontinuitiy in the template. The human genome is so large than on average we¿re going to have many forks blocked in the process of replication. And those stalled forks need to be restructured and restarted to successfully get through the cell cycle. Brca1 has been implicated in this process of recovering the replication forks. We found that that protein, in addition to binding strongly to normal helical DNA, especially prefers to bind to those branched structures, and perhaps initiates a repair response.

¿If the whole genome,¿ Paull recounted, ¿which needs to remain intact, starts having mutations and rearrangements, disrupting important genes, that can have deleterious consequences. As you look at tumor cell lines, you know that this is true for virtually every one of them ¿ that their genome is completely rearranged ¿ a total genomic disaster in cancer cells. You have multiple copies of some chromosomes, no copies of other chromosomes. You have chromosomal rearrangements that may be half No. 3 and part No. 7.

¿One of the ideas that links cancers to DNA repair,¿ Paull went on, ¿is that early in tumor progression if you lose a factor that¿s important for some type of repair, that increases statistically the rate of mutation across your entire genome. And that in turn will lead to an increased probability of losing a gene that¿s involved in growth control. That is how we think lots of DNA repair is linked to cancer. We have very strong and redundant systems for repairing DNA, so one of the ways a tumor can progress is by losing some element of those protective systems in the genome. That¿s why it¿s important for us to understand these processes. They¿re really at the initiating stages of lots of these systems in cancer.¿

Clinical Utility Not In Sight

Molecular geneticist Martin Gellert, a section chief at the National Institute of Diabetes and Digestive and Kidney Diseases in Bethesda, Md., is the PNAS paper¿s senior author. Paull did the research reported in the paper in Gellert¿s lab before moving last year to a faculty position at the University of Texas in Austin, where she continues to work on the Brca1 project..

Asked what putative clinical payoffs the work may have some day, Gellert replied, ¿Right now I¿d say a flat zero. The level of non-knowledge here,¿ he concluded, ¿is very profound.¿