The guardian of the genome

The archetypal tumour-suppressor gene P53, alias Guardian of the Genome, bestrides the intricate machinery of the cell cycle, ever alert for early signs of mutation or genomic stress that could tip a cell into cancerous growth.

P53 is best known as the master controller of apoptosis. But how does the Guardian learn of brewing genetic mischief? And how does it communicate through subsidiary gene networks to force pre-cancerous cells to die?

Molecular geneticist Professor Scott Lowe, of New York's Cold Spring Harbour Laboratory, a plenary speaker at the Lorne Cancer conference in February, was one of a number of international researchers who showed that P53 reacts to mutation overload by shutting down the damaged cell and activating cell-slaying caspase enzymes before it can turn cancerous and proliferate.

While P53 is most familiar as the master regulator of apoptosis, Lowe says it eliminates pre-cancerous cells by several other mechanisms, including forcing them into irreversible senescence.

"Our work on P53 has taken us in several directions," he says. "We're investigating DNA damage that activates P53 - our interest is not in the mechanism, but in the consequences of lesions that disable part of the P53 network, because it can affect the success of chemotherapy.

"Drugs that damage P53 can make a tumour resistant to P53. That got us interested in why some cells are sensitive to chemotherapy, while others are resistant - it's not explained by [inactivation of] P53 alone.

"We're also interested in the deregulation of oncogenes that can turn on P53 and make cells grow. Why would oncogenes turn on a tumour-suppressor gene? "We think cells have inbuilt safeguards that prevent them becoming cancerous. If those safeguards fail, tumours progress."

Lowe says his team's interest in P53-mediated senescence reflects new evidence that senescence may be just as potent a mechanism as apoptosis in programmed cell death. "We became interested because less is known about senescence than about apoptosis. If a cell senesces, it never grows again, and can't form a cancer."

The Lowe team is focusing on the involvement of two other tumour-suppressor genes that operate downstream of P53: P16 and Rb (retinoblastoma), which work together to initiate senescence. Lowe says P16 regulates production of Rb protein; the loss of either gene can cause cancer.

After P53, which is inactivated in some 60 per cent of cancers, P16 is the most frequently mutated gene in cancer.

"If either gene is lost, cells are likely to stop growing in response to stress, and go on to form a cancer."

Lowe's interest in the dysregulation of the genetic network of senescence stems from his primary interest in the genetic mechanisms of drug sensitivity and resistance.

"The integrity of this network affects the outcome of cancer therapy," he says. "We're trying to identify as many of its components as we can, and ask how they can be broken, how a tumour develops, and how it responds to therapy. We want to exploit that information to improve treatment of cancers."

The Lowe team is working with Cold Spring Harbor colleague and RNA-interference expert Professor Greg Hannon to probe the senescence network in mouse models.

Homologous recombination is the standard technique for engineering knockout mice, but the CHS researchers have pioneered the use of RNAi knockdown in mice.

In addition to single-gene RNAi knockdown mice, they have engineered 'double whammy' knockdown mice, in which a one gene of interest is constitutively suppressed, while a second is conditionally inactivated until a chosen time, for example in the mammary gland, after puberty.

In this way, they can selectively probe interactions between key genes in the P53-mediated senescence pathway.

"With RNA libraries, we can target any gene, and we use gene-expression microarrays to identify genes involved in the processes we're interested in," he says. "We're very excited about the technique - it's cutting edge.

"We can turn any gene on or off at will, to affect a particular process at will. This is about speed, and the rate at which we can acquire information about gene function, and the involvement of different genes in cancer."

The endpoint of such research will be personalised treatment of an individual's cancer with specific combinations of novel, yet-to-be-developed drugs.

"Everyone is waiting for a magic bullet, but that's my personal view of where cancer therapy is headed," Lowe says. "We will understand cancer genetics, and tumour cell behaviour.

"We'll be able to scan the entire cancer genome, and with comparative hybridisation, tell if any gene is being over- or under-expressed relative to the norm. And with RNAi, ultimately, we'll be much better able to diagnose and stratify cancers, and develop appropriate therapies."