Chemogenetics Method Uses Anti-Smoking Drug to Control Cells
Chemogenetics Method Uses Anti-Smoking Drug to Control Cells

Chemogenetics Method Uses Anti-Smoking Drug to Control Cells

A new set of engineered receptors responds to an FDA-approved drug to provide the most potent chemogenetic toolkit to date.

Ruth Williams
Ruth Williams
Mar 14, 2019

ABOVE: Mouse neurons targeted to selectively produce chemogenetic activator channel (red)

Researchers have come up with a new method to control brain cells in live animals using specially designed receptor proteins that respond to the drug varenicline. While drug-responsive receptors have been around for sometime, the new incarnations, described today (March 14) in Science, have been structurally optimized, as has the drug itself, to create a novel repertoire of precise and powerful chemogenetic resources.

“It really is an exciting new development that has great potential not only for basic research but potentially also in translation and applications for human use,” says neuroscientist Christian Lüscher of the University of Geneva who was not involved with the research.

“There is a tremendous need for novel medications that have higher selectivity . . . , higher potency at very low doses, and hence less side effects. And this technology potentially fits these needs,” adds neurologist Antonello Bonci of the National Institute on Drug Abuse, also not part of the research team, in an email to The Scientist. 

The aim of chemogenetic techniques is to enable researchers to activate or silence specific cell types at will. Applied typically to brain cell manipulations, the techniques employ specially designed receptors that only respond to particular ligands (drugs or molecules). Introducing the receptors into chosen cells thus allows drug-dependant control of those cells’ activities.

One of the principal chemogenetic systems—designer receptors exclusively activated by designer drugs (DREADDs)—has limitations. For one thing, a commonly used DREADD activator, CNO, was recently found to transform into the drug clozapine, which has widespread effects in the brain. Moreover, DREADDs are based on G-protein coupled receptors (GPCRs), explains Lüscher, meaning they must associate with ion channels in the cell to have an effect. “So if the cell doesn’t express the [necessary ion] channel it will simply not work,” he says.

Scott Sternson of the HHMI Janelia Research Campus and colleagues’ new system by contrast is based on fusion proteins consisting of an ion channel domain and a receptor domain—specifically that of the α7 acetylcholine receptor (α7nAChR). “They are in themselves already the effectors,” says Lüscher, meaning they can work in essentially any cell type regardless of the other ion channels present. “That’s a big advantage.”

Rather than making both a designer receptor and ligand straight away, Sternson’s team first focused on creating receptors that would respond to a given FDA-approved drug. “To use chemogenetics therapeutically you’re going to have a gene therapy component, which is the receptor, and then you have a chemical component, which obviously is a small molecule drug, and that, from a practical standpoint, creates certain regulatory challenges” in terms of clinical testing, Sternson explains. Starting with an approved drug would thus more likely result in a system ripe for translation into, for example, therapies for pain or epilepsy.

To that end, the team screened an array of safe, well-tolerated, brain-entering drugs for their ability to interact with a variety of α7 fusion receptors. Varenicline, an anti-smoking drug, stood out as a strong candidate, explains Sternson. To maximize the effect of varenicline the investigators then studied the crystal structure of the drug-receptor interaction and, with a great deal of educated tinkering and patience, tweaked the receptor until they produced optimized versions many times more responsive than the originals.

Indeed, in cultured mouse neurons as well as in live mice and monkeys, doses of varenicline substantially lower than that normally required for the drug’s nicotine-substitution effect were able to induce or suppress the activity (depending on the ion channel domain spliced to the receptor) of cells presenting the optimized receptors.

We’re getting to a point where we can hack the brain.

—Gordon Fishell,
Harvard Medical School

This is important for future translation of the system to human use, says Lüscher. “If you can use such low doses of varenicline then I anticipate there should be minimal side effects.”

Sternson’s team has also tinkered with varenicline itself to make versions of the drug that interact with the optimized receptors more specifically, or that offer yet more potency. In mice, one of these varenicline variants, when used at a three-fold lower dose than the original drug, could just as effectively suppress the activity of neurons expressing the engineered receptor and alter the animals’ behavior.

While these varenicline variants, like the α7 receptors, are so far only for research purposes, “I am rather cautiously optimistic that this will change the landscape of our ability to use chemogenetics in a way that has the potential to fairly quickly be applied in clinical contexts,” says neuroscientist Gordon Fishell of Harvard Medical School who was not part of the research team. “We’re getting to a point where we can hack the brain.”

C. J. Magnus et al., “Ultrapotent chemogenetics for research and potential clinical applications,” Science, doi:10.1126/science.aav5282, 2019.