Since ARF's discovery, biomedical researchers have had considerable interest in this tumor suppressor. In one two-month period in late 1999, four papers on ARF by the Sherr/Roussel collaboration racked up 50 citations among them. One year ago, those papers helped place both Dr. Roussel and Dr. Sherr on this publication's annual roundup of the year's hottest researchers (see Science Watch, 11[2]:1-2, March/April 2000). And the team's papers continue to be cited at a good clip. One 1999 paper, for example, "Nucleolar ARF sequesters MDM2 and activates p53" (J.D. Weber, et al., Nature Cell Biology 1:20-6, 1999), was cited 14 times during November-December 2000 and has now received upwards of 70 citations overall.

Born in France and raised in Equatorial and West Africa, Dr. Roussel, 50, received her bachelor’s degree in biology from the Universite de Tours in 1971. She went on to earn a master's degree from the Universite des Sciences de Paris VII in 1974, and a doctorate from the Universite de Lille four years later. Dr. Roussel’s promise as a researcher was demonstrated back in the late 1970s when, working in the Paris laboratory of Dominique Stehelin, she helped discover the Myc oncogene. The subsequent 1979 Nature article (M. Roussel, et al., "Three new types of viral oncogenes of cellular origin specific for hematopoietic cell transformation," Nature, 281[5731]: 452-5, 1979) has since garnered nearly 350 citations. She began working with Dr. Sherr, to whom she’s now married, in 1983, while at the National Cancer Institute. In 1983 the two set up a laboratory at St. Jude (where Dr. Sherr is supported by the Howard Hughes Medical Institute). In 1989, Dr. Roussel also joined the University of Tennessee, where she is now a professor in the biochemistry department.

How was ARF discovered?

Roussel: That’s an interesting story. We were interested in the regulation of cyclin D-dependent kinases, cdk4 and cdk6. These cdks are responsible for the initial phosphorylation of retinoblastoma protein (Rb), which is a tumor-suppressor gene. Rb requires phosphorylation for cells to replicate DNA. INK4s are specific inhibitors of cdk4 and cdk6 activities and prevent Rb phosphorylation. We were interested in INK4 genes because one member, INK4a, was shown to be mutated in melanomas, and its locus was implicated in many cancers.

Then the discovery of ARF was made by Dr. Sherr and one of his post-docs in the lab, Dawn Quelle. Dawn was working in collaboration with David Beach and Manuel Serrano at Cold Spring Harbor Lab, trying to clone the mouse INK4a gene, and she was interested in its regulation. She used a cDNA library to clone mouse INK4a. That's when she found cDNAs that had the INK4a sequence, which she was using as a probe, but she found that it contained extra sequences, different from the known INK4a gene. When she sequenced these cDNAs she realized that the entire sequence was read as one single phrase (in frame), suggesting it was real—it was not a piece of DNA that was added by accident, or an artifact of cloning. And then she realized that it encoded a novel protein. She cloned those cDNAs into vectors and overexpressed the cDNAs into naive cells. And that’s when she realized that when this protein was expressed, it stopped cells from growing.

So "Alternative Reading Frame," which is what ARF stands for, means that ARF starts at a different point on the DNA than INK4a but still uses the same genetic information to encode its protein?

Roussel: Exactly. INK4a, also called p16, has three exons. ARF starts from an "alternative" exon, far away from the locus, 15 kb away. That little exon, called 1-beta, is transcribed as an RNA which then splices into the second exon of p16, but in a different frame than the frame used by p16. So exon two anchors two different proteins. And because ARF is in a different reading frame, as a different phrase, and starts from somewhere else, the protein encoded has nothing to do with p16. It’s therefore encoded by a completely different sequence and is completely unrelated immunologically. And yet both of those proteins, when put into cells, stop the cells from growing.

That sounds like a bizarre arrangement. Is it unusual?

Roussel: It is. In mammalian cells, it’s unprecedented. This economy of space is seen frequently in viruses, which have a minimum amount of DNA that can be packaged, so they have to use all their reading frames somehow.

So ARF is a tumor suppressor like p16?

Roussel: Both p16 and ARF stop cells from growing. They both affect cell-cycle progression or cell division. When we started to transfer ARF into different cells, we realized that the effect on cell-cycle progression occurred only when the p53 gene, also a tumor suppressor, was functional. p53 and the INK4a/ARF locus are the two loci that are most frequently mutated or deleted in human cancer. When we decided to delete ARF in mice, to make knockouts with ARF deleted independently of p16—this was done by Dr. T. Kamijo—we found the mice were born normally, but they developed tumors as would have been expected if ARF is a tumor suppressor. It’s really fascinating. p16 is interacting with the Rb pathway and ARF is interacting with the p53 tumor-suppressor pathway. So this locus encodes two proteins that regulate two important tumor suppressors, and thus it is very important for cells.
continued