By David N. Leff

One might say that the p53 gene holds a Kevorkian gift certificate for assisting undesirable cells in the body to commit suicide.

Just how the near-ubiquitous p53 tumor-suppressing protein carries out its apoptosis-promoting role is only partly understood — but not for lack of worldwide research.

One such focus of investigative interest in p53's molecular machinery for programming cancer cells to die is the Johns Hopkins University's Oncology Center, in Baltimore. There, as reported in today's Nature, dated Sept. 18, 1997, the two senior authors, molecular geneticists Bert Vogelstein and Kenneth Kinzler, have tracked down p53 from gene to protein to tumor-cell suicide.

Their paper bears the cryptic title: "A model for p53-induced apoptosis."

Actually, the co-authors note, an apoptotic death sentence is only one of two possible verdicts carried by p53 on tumor cells. The alternative to capital punishment is growth arrest, which locks up the injured cell, often for long periods.

"The mechanisms underlying the development of p53-dependent apoptosis," they write, are largely unknown." To scope this mechanism, they started with the knowledge that p53 launches its assisted-suicide program by spawning large numbers of purpose-directed subservient gene sequences.

Whereupon, they asked: How many genes? And for what purposes?

For answers, together with the Nature paper's first author, Kornella Polyak, they began by putting the p53 gene into a laboratory line of colorectal cancer cells. Half of these cells are known to go belly-up in response to the tumor-suppressing protein.

To single out these PIGs — p53-induced genes — from the colorectal cells' genome, they employed a proprietary technique called SAGE — Serial Analysis of Gene Expression. Its lead inventors are Vogelstein and Kinzler; the university licensed their invention to PharmaGenics Inc., of Allendale, N.J. (See BioWorld Today, Oct. 25, 1995, p.1.)

PharmaGenics and its SAGE technology were purchased by Genzyme Corp., of Cambridge, Mass., earlier this year. (See BioWorld Today, Feb. 4, 1997, p.1.)

SAGE quantitates cellular messenger RNA populations, a measure of gene expression. From 101,694 expressed sequence tags, averaging 11 base pairs in length, the team identified 7,202 different gene transcripts. Only 14 of these putative PIGs were expressed at significantly higher levels in p53-expressing cells than in controls.

Seven of the 14 had never been described before. All 14 death-determining sequences gave hints as to the genes' function in the cell, and several pointed at one clue in particular: oxidation-reduction, a.k.a. redox. Of the 14 PIGs, seven had molecular ties to redox oxidative stress (ROS), which disrupts and destroys nuclear DNA in cells.

Ionizing radiation, a form of cancer therapy (and antibacterial treatment), acts through ROS. Conversely, so does cancer-causing radiation.

Vogelstein, Kinzler and Polyak (all members of the Howard Hughes Medical Institute) sum up their p53 apoptotic three-step as: (1) redox-related genes; (2) ROS formation; and (3) mitochondrial degradation.

The co-authors "propose that p53 transcriptionally activates a specific subset of genes . . . long before any morphological or biochemical evidence of cell death. The proteins encoded by these genes then collectively increase the content of ROS, which in turn damages mitochondria."

In a Nature editorial, titled "Clues in the p53 murder mystery," commenting on the Johns Hopkins paper, Scottish oncologist Andrew Wyllie, at the University of Edinburgh, wrote, apropos the mitochondrial connection: "[T]here is striking new evidence for the importance of signals from mitochondria in the initiation of apoptosis; these organelles are well placed to be sensors of oxidative damage."

Wyllie also made the point that beyond cancer, the paper's finding "raises the possibility that pharmacological regulation of p53 action might significantly modify the extent of [ROS] hypoxic tissue injury in myocardial infarction and stroke."

He concluded: "Perhaps . . . the game of molecular whodunit is not yet played out. Are all p53-dependent deaths effected by the same means? And does the murder take place in the nucleus or the mitochondrion?" *