Lorne 2009 profile: David Vaux

The first chapter in Professor David Vaux’s tale concerns the evolution of cell death and begins late last century, probably on a cold Melbourne day. Vaux and a few colleagues were thinking about the big issues we all eventually ponder – life and death and how they evolved.

In Vaux’s case, it was how and why programmed cell death began. “We knew that many of the mechanisms for cell suicide are highly conserved amongst all the metazoans, but not when and how they evolved,” Vaux says.

It seemed unlikely – and against Darwinian theory – that physiological cell death could have evolved in single-celled organisms such as bacteria. If one evolved a mechanism to kill itself, how could it pass on its genes to the next generation?

However, Vaux says, it is possible that such organisms might have evolved a mechanism “not for programmed cell death (one that is obligatory), but rather one for contingent cell death, that doesn’t necessarily get activated.”

For example, a bacterium infected with a bacteriophage – a bacteria-infecting virus – or living under conditions of limited nutrients could activate its suicide mechanism. This scenario obviously wouldn’t help that individual organism, but its relatives carrying the same genes might be then at a survival advantage, so Darwinian theory would still work for the group.

Vaux contends that this is how the mechanisms for cell death evolved, as a contingent defence mechanism.

“It was then only later in evolution, with the advent of multicellular organisms, that the same apoptosis mechanisms were adopted for sculpting the body and maintaining cell number by balancing mitosis,” he says.

Vaux then got to thinking about this clear relationship between apoptosis and viruses. “I decided that viruses would be a fantastic place to look for new tools and ways to dissect the process of cell death. Viruses need to take over a cell’s machinery to replicate and a good way for the cell to defend against this is by committing suicide.”

However, viruses can evolve very rapidly and can pick up host genes, so there is evolutionary selective pressure on viruses to pick up genes that inhibit host cell suicide, giving them more time to replicate.

Viral genomes would therefore be the ideal place to look for new genes that regulate cell death, and this notion led into the next chapter: insect viruses.

---PB--- Grim, Reaper and Sickle

Vaux, an invited speaker last week’s Lorne Cancer conference, recalls that the American scientist Lois Miller had recently discovered a couple of baculovirus proteins that were able to inhibit apoptosis in the host insect cell.

“One was called p35, and is now known to be a caspase inhibitor, and the other one was called IAP, or inhibitor of apoptosis protein,” Vaux says. On reading Miller’s paper, Vaux wanted to know whether these insect virus cell-death inhibitors could function similarly in mammalian cells.

Vaux obtained the baculoviral IAP from Miller and his PhD student Chris Hawkins, now at La Trobe University, expressed it in human cells in tissue culture treated to induce apoptosis. The IAP could indeed inhibit cell death in the mammalian cell line.

“Then, because viruses hijack cellular genes, we thought that maybe the host insect cells also bore IAPs, and that perhaps mammals did also,” Vaux says.

“By database searching, we found a few IAPs in Drosophila and yeast, as well as a bunch in mammals. Miha Pakusch found a mammalian one now called XIAP because the gene is on the X chromosome, and Anthony Uren found cIAP1 and 2.”

Other groups were also identifying IAPs and the search for their mechanisms of action was hotting up.

Soon after, studies of mutant flies in which cell death fails to occur identified a key region in the fly genome encoding four small pro-apoptotic signalling proteins called Grim, Hid, Reaper and Sickle. These proteins were shown to bind insect IAPs and in doing so, reverse their inhibition of apoptosis.

There were no obvious homologues of these pro-apoptotic proteins in mammals, but because the baculoviral IAPs could inhibit cell death also in mammalian cells, Vaux hypothesised that similar molecules with an analogous function would exist in mammals.

Postdoctoral student Anne Verhagen then went looking for mammalian proteins that were able to bind to IAPs. Using the IAPs as bait, she did some biochemical fishing for binding partners in human cells, and found DIABLO, a protein that interacts with both baculoviral and mammalian IAPs and seemed to function in a similar manner to the insect pro-apoptotic molecules such as Grim and Reaper.

Xiaodong Wang’s group in Texas found the same protein independently and called it Smac, with both findings reported back-to-back in Cell in 2000.

Verhagen and another of Vaux’s postdocs, John Silke, then noticed that although DIABLO/Smac was very unlike the insect pro-apoptotic molecules, all of them had a similar four-amino acid region at the N-terminus.

“We hypothesised that this small bit at the end of the proteins must be the key binding site for IAPs,” Vaux says.

Labs in the US found that XIAP was a potent inhibitor of some caspases, so it seemed possible that DIABLO/Smac might cause apoptosis by stopping XIAP from inhibiting the activated proteases.

---PB--- IAP inhibitors

By now, the drug companies were getting interested in this potential to modulate cell death, particularly given that several cancer types show high expression of specific IAPs.

“So, a whole bunch of companies including Genentech, Novartis and Abbott started making compounds resembling those first four amino acids of DIABLO/Smac to try and inhibit IAPs in cancer cells and thereby induce cell death,” he says.

Preliminary testing of these agents looked promising – they seemed to be good at killing selected cancer cell lines without harming non-cancer cells, and if injected into mice grafted with a bit of human tumour, the cancerous cells were specifically killed.

Of course, nothing is ever that simple, and more conundrums were thrown up by subsequent work from Vaux’s group and others a few years ago. For example, knockout mice for XIAP and DIABLO had no phenotype, and Guy Salvesen in San Diego showed that cIAP1 and cIAP2 were not direct caspase inhibitors.

In collaboration with a start-up company in Pennsylvania which was making these IAP antagonist compounds, Vaux’s group showed that rather than countering XIAP, they caused cancer cell death by increasing the level of pro-apoptotic signalling from TNF receptor superfamily members.

At the moment Vaux’s group is concentrating on structure-function analyses of IAP and IAP-antagonist interactions by making IAP mutants with specific defects. The results will address what is happening in the cell at the molecular level when the IAP antagonist compound is added, and will help in refining these compounds to work as well as possible in clinical trials and ultimately, in patients.

In the meantime, Genentech has taken its IAP antagonist agent into Phase I human trials.

“This is a really nice story that goes from something that seems very arcane, namely insect viruses, that doesn’t seem to have any application at all to the development of compounds for treating humans with cancer,” Vaux says.

“It is a good illustration of the importance of fundamental research and how basic research can yield unexpected potential benefits for patients.”