Combining Chemical Genetics and Cancer Immunotherapy

Two years ago, a team at Stanford described a new "chemical genetics" technique--a more precise way to use small molecules to perturb gene products and living systems. Now they have published more data on the system. And of acute interest to drug developers, they have linked the work to therapeutic applications in cancer immunotherapy, showing the ability to regulate secreted proteins such as IL-2 and TNF.

Mark L. Ratner

Two years ago, a team at Stanford University’s Stanford University School of Medicine described a new "chemical genetics" technique—a way to use small molecules to perturb gene products and living systems—in the journal Cell. The group engineered a small, inherently unstable domain that triggered rapid protein degradation and then demonstrated they could regulate the destabilizing function in a highly controlled manner using a small molecule. This "destabilizing domain"/small molecule combo was intended primarily as a research tool. Now, in a September 28 online paper in Nature Medicine, the researchers have published more data on the system, including its effectiveness in an animal model. And importantly, of acute interest to drug developers, they have also linked their work to therapeutic applications in cancer immunotherapy, showing the ability to regulate secreted proteins such as the potent immune system modulator, interleukin-2, and tumor necrosis factor in vivo.

Chemical genetics, or chemogenomics, has been around since the 1990s. It was the founding premise of Vertex Pharmaceuticals Inc. as well as other biotechs that pioneered high-throughout screening. (See "Vertex: Sticking To Its Story," IN VIVO, October 2002 (Also see "Vertex: Sticking To Its Story" - In Vivo, 1 Oct, 2002.).) But while there are assorted examples of how well chemical genetics can work, generalizing the approach is difficult, explains Stanford assistant professor Tom Wandless, PhD, whose lab developed the destabilizing domain system. "We tried to engineer a second wave," he says, which required an investment in molecular biology or genetic manipulation--to go beyond the blunt-force screening of proteomics approaches, which often depend on small molecules that lack specificity. The strategy pursues "exactly what you want to target and ideally nothing else," he notes.

The destabilizing domain is a point mutation in the FK protein family targeted by drugs including the immunosuppressants rapamycin and FK506. Those drugs, unlike the small molecule in Wandless’s system, also target mTOR, a regulator of a range of cell signaling activities, and calcineurin, which is involved in stimulating a T-cell response, respectively. For its purpose, however, the Stanford team needed a small molecule that would be biologically silent; that is, it had to diffuse into mammalian cells and not do anything. They also needed the domain to be unstable in the absence of the small-molecule ligand and stable in the presence of it. They engineered mutations that gave the FK domain the behavior they sought and a ligand, called Shield-1, which met their criteria. The resulting destabilizing domain approach is more "tune-able" than other genetic methods for regulating protein expression, such as the use of inducible promoters or suicide genes.

After completing initial work in cultured cells, Wandless approached Stanford colleague Chris Contag, PhD, an expert in molecular imaging, seeking his opinion on the best way to measure how quickly Shield-1 made it to different organs and tissues in an animal, and what the pharmacokinetics and so forth would be. Contag immediately saw more potential applications for it than as just a research tool.

"When Tom published his Cell paper, I thought this would be perfect for gene therapy," Contag recalls. "When we started working together on this recent paper my whole objective was to use it in a therapeutic approach."

The team first demonstrated that the system could be used in vivo in conjunction with IL-2, a cytokine whose functions include the recruitment of immune cells (natural killer cells, in particular). "We wanted to show we could begin to control IL-2 expression," says Contag. "Turn it on and off at the time we wanted to recruit immune cells to the tumor. Not just as gene therapy but [also] immune cell therapy." They introduced the genes encoding the destabilization domain and IL-2 into cultured tumor cells, then grafted the cells into mice able to produce functional B and natural killer cells. Adding Shield-1 "restabilized" the system and turned on the IL-2, preventing the tumors from developing in the animals. The researchers then tested a different protein. They used vaccinia virus to deliver TNF-alpha and the destabilization domain into mice bearing large tumors; as with IL-2, introduction of Shield-1 turned on the TNF-a and the tumors regressed.

Contag is particularly taken with the idea of using vaccinia virus to deliver the genes encoding the destabilizing domain and a therapeutic protein.

"We used vaccinia in the study because we’re trying to combine immune cell therapy with vaccinia [gene] therapy for cancer treatment," he explains. "It’s a highly lytic virus that we can engineer to be tumor selective." Contag has been working with cytokine induced killer (CIK) cells as an immune cell therapy and wants to add IL-2 to provide an additional immune cell boost. But CIK cells have to be enriched for three weeks in the culture dish. That amount of time in culture "presents problems for the patient, it takes time, there’s more chance of contamination, and it’s expensive," he points out. Being able to introduce the IL-2 in vivo using the destabilization domain approach "would save quite a bit of effort," he says.

The destabilization domain system is now sold as a research reagent kit by Clontech Laboratories Inc. (a division of Takara Shuzo Co. Ltd.’s Takara Bio Inc.) under the name ProteoTuner. (Stanford will also license its use for commercial [therapeutic] applications nonexclusively.) Until now, the method’s value as a tool for target validation has been "underappreciated" by pharma, Wandless believes. "I think we’ve only had two companies license it." That could change somewhat with the animal data validating its potential use to deliver a therapy.

The Wandless lab has now developed a second destabilization domain system using a completely different protein-small molecule combination – also validated in vivo -- so that researchers can regulate two proteins or two families of proteins independently. Although there are very few proteins for which the initial FK-based system doesn’t work, its use with GPCRs – transmembrane proteins – is limited. The second system appears to work with GPCRs and ion channels, which are difficult to manipulate in drug discovery. A paper describing the second system should be completed in the next few months, according to Wandless. That could further pique the interest of pharma and biotech.