Diagnostics & Imaging Week Contributing Writer

LONDON – A gene that acts as a wolf in sheep’s clothing – or perhaps the other way around – is forcing researchers to reconsider the tenet that genetically tailored chemotherapy for cancer is the ideal therapeutic approach.

Benjamin Rowland and his colleagues at the Netherlands Cancer Institute (Amsterdam) have established that a molecule known as a tumor-suppressor gene can also act as an oncogene, depending on which other genes are active in the cell. The finding means that oncologists must thoroughly understand all the molecular roles of cancer genes before targeting individual genes with cancer therapies.

The work is described in the Oct. 23 issue of Nature Cell Biology.

Daniel Peeper, group leader at the Netherlands Cancer Institute and senior author of the paper, told Diagnostics & Imaging Week’s sister publication, BioWorld International: “We might need to reconsider how we think about oncogenes and tumor-suppressor genes. It is now clear that some genes can act as oncogenes in some settings and as tumor-suppressor genes in others.”

In terms of therapy, he added, that could mean that it might not always be a big advantage to treat patients systemically with highly specific drugs designed to block certain oncogenes. “Such therapy might destroy the tumor, but it might give other cells a selective growth advantage,” Peeper added.

Peeper’s lab specializes in identifying novel genes involved in cancer. He and his colleagues wanted to identify genes that interact functionally with the RAS oncogene. That is mutated and activated in about 30% of human cancers, yet when a mutant RAS oncogene is introduced into normal cells, it triggers arrest of the cell cycle and sends the cells into premature senescence.

Using a genome-wide expression library to search for genes in cells that continue to proliferate despite the presence of mutant RAS, Peeper and his colleagues identified one encoding a molecule called Kr ppel-like factor 4 (KLF4).

Like mutant RAS, KLF4 also has two functions, depending on where it is expressed. In most human breast cancers, it is present at high expression levels, suggesting that it acts as an oncogene. Yet in a proportion of human colon cancers it is present at only very low levels or absent altogether – a feature that is a hallmark of tumor-suppressor genes.

Studies of transgenic mouse models are equally conflicting. In a transgenic mouse in which KLF4 is expressed at high levels in the skin, tumors are common, while in a tissue-specific knockout mouse in which there is no expression of KLF4 in the gastric epithelium, the mice also develop tumors.

Experiments carried out by Peeper and his colleagues, and reported in the Nature Cell Biology paper, confirm the paradox. Although they clearly identified KLF4 as having an oncogenic function in their screen, when they put the gene into normal cells, it acted as a potent growth-inhibitor.

Further investigations by the group showed that KLF4 represses expression of p53, while inducing expression of the protein p21. Peeper said: “Our experiments showed that it is the induction of p21 that is responsible for the induction of growth arrest by KLF4 in normal cells.”

When the team added KLF4 to cells lacking a functional gene for p21, the cells continued to proliferate and transformed into cancer cells. Peeper explained, “So in the absence of p21, the repression of p53 by KLF4 becomes dominant and now there is an oncogenic effect.”

In an additional set of experiments, the group investigated the function of KLF4 in breast cancer cells, where it is often expressed at high levels. The scientists depleted KLF4 using RNA interference. They found that the cells, which formerly had low levels of p53 (presumably due to KLF4 over-expression), were then able to restore p53 protein levels. As a result, those cells died.

“You might argue from this that KLF4 would be an ideal drug target,” Peeper said, “but this is only going to be true in specific settings – that is, where p21 activity is compromised. If you inactivate KLF4 in cells where p21 is functional, then you may well convert those cells to early tumor cells.”