91av

LATE last year, researchers at pharmaceutical giant Eli Lilly’s labs in Indianapolis, Indiana, suffered a rude shock. Mice involved in tests of one of the company’s most promising drugs – a new treatment for anxiety – were having seizures. Lilly put the drug on hold and suspended a clinical trial involving 1900 people. It is still trying to work out what went wrong, and whether trials of the drug can be revived.

Talk to Lilly and you could be forgiven for thinking that the drug’s woes came as a huge surprise. But perhaps the real surprise is that the drug ever got so far. The compound Lilly was testing was designed to plug into a brain circuit that pharmacologists have long considered a no-go area. Fiddle with it, the orthodoxy goes, and all you get for your troubles is hallucinations, psychosis and seizures.

Yet Lilly, and most of its competitors, now believe that this circuit is the key to a new class of molecule that will revolutionise the treatment of mental illness – including many currently intractable or poorly treated diseases such as addiction, anxiety, schizophrenia, epilepsy and chronic pain. Almost every major pharmaceutical company is developing similar molecules. And while none is yet close to the market, some neuroscientists believe they will be the biggest shake-up in central nervous system medicine for decades.

The compounds that are getting neuroscientists excited are based on glutamate, the brain’s primary neurotransmitter, or communication molecule. Just about every circuit in the central nervous system uses glutamate, so in theory, drugs that target glutamate signalling have the potential to treat almost any brain disorder.

But there is a catch. Glutamate signalling is so pervasive in the brain that interfering with it usually leads to horrendous side effects. And so neuroscientists working in the pharmaceutical industry have generally steered clear of it. But not any more. As researchers discover more about the glutamate system, they have found a promising way to get a handle on it without causing side effects. For the first time, there is the real prospect of taking control of the brain’s master switch.

“It’s a huge conceptual leap,” says Bita Moghaddam, a neuroscientist at the University of Pittsburgh in Pennsylvania. “For the past few decades, the concepts behind treatment haven’t changed. We are stuck on serotonin for depression, dopamine for schizophrenia and GABA for anxiety. This is the first time we’re going beyond these old ideas.”

That is not to deny the importance of serotonin, dopamine and GABA (gamma aminobutyric acid). GABA is the brain’s main inhibitory molecule – the “off” signal for neurons – while serotonin and dopamine help fine-tune communication between brain cells. But none is as important as glutamate. Whenever you have a thought or take an action, glutamate is right at the heart of it, transmitting “on” signals from neuron to neuron. The sender releases a puff of glutamate which diffuses across the synapse, binding to receptors on the other side.

Until recently, neuroscientists thought that there was only one type of glutamate receptor. Known as ionotropic receptors, they function like the lock on a gate. Binding of glutamate molecules opens the gate, allowing charged particles to rush into the neuron and trigger the electrical current or nerve impulse.

It was investigations of these receptors that earned glutamate a reputation as a no-go area. Glutamate blockers were tested as a treatment for stroke, but were scrapped because they induced psychosis. Similarly, the illicit drug PCP (also known as angel dust) was found to induce its hallucinogenic and psychotic effects by blocking ionotropic glutamate receptors. “Glutamate is too broad a hammer on too many circuits,” says Jeffrey Conn, a neuroscientist at Vanderbilt University in Nashville, Tennessee, who has been studying glutamate receptors for 20 years.

But in the mid-1980s, neuroscientists began to suspect that there was a different class of glutamate receptor in the brain – the so-called metabotropic receptor. This class of receptor was already well known to neuroscientists, though it was always associated with neurotransmitters other than glutamate. Serotonin and dopamine, for example, work largely by acting on metabotropic receptors. In 1991 the hunch was confirmed when groups in the US and Japan cloned the first metabotropic glutamate receptor (mGluR).

Metabotropic receptors have a slower, more subtle effect than ionotropic receptors. If these are an on-off switch, then metabotropic receptors are a dimmer, turning the strength of the signals up or down. They do this by triggering a cascade of metabolic reactions inside the neuron that either increases or decreases its excitability. For example, they may alter the ion channels that have an impact on the electrical properties of the cell, or affect enzymes that make neurotransmitters.

This action to modulate signal strength makes metabotropic receptors particularly good as drug targets. “They are having a more subtle effect than drugs that directly target glutamate,” says Conn. “You’re not hitting [the brain] with a sledgehammer and causing the toxicity you would get when you go after the main circuits.” Hence the pharmaceutical industry’s focus on serotonin and dopamine.

But going after serotonin and dopamine is inherently limited because both are just bit-part players in the brain. Only about 10,000 of the brain’s 100 billion neurons produce dopamine, and serotonin circuits are confined mainly to the midbrain. That’s why the discovery of metabotropic glutamate receptors holds such promise. Like serotonin and dopamine receptors, they act as a dimmer switch on glutamate signalling. But they are much more widespread than the serotonin and dopamine receptors. Wherever you have glutamate circuits, there are metabotropic glutamate receptors modulating them: “mGluRs give the opportunity of direct but subtle manipulation of brain circuits,” says Conn.

The receptors have another important advantage: they come in several different flavours. Eight types have been identified so far and, crucially, these are not uniformly distributed throughout the brain. Some are found only in certain parts. The receptor known as mGluR4, for example, is confined largely to the basal ganglia, an area that is damaged in Parkinson’s disease. And mGluR5 is linked to an ionotropic glutamate receptor, the malfunction of which is strongly linked to schizophrenia. So by designing drugs that target the different mGluR receptors, pharmacologists believe they will be able to modulate glutamate signalling in highly selective and valuable ways.

According to Conn, the pharmaceutical potential of mGluRs was obvious 20 years ago. Only now, though, is the research approaching fruition, with pharmaceutical companies racing to develop compounds that target them. The field is secretive but a number of big players have declared their hands. For example, Merck recently published research on mGluR4 and Parkinson’s disease. Novartis is developing a mGluR5 antagonist for pain and anxiety, and Lilly is working with compounds that act on mGluR2, which have potential to treat anxiety, addiction and possibly pain. These trial drugs include the one that caused seizures in mice last year.

Such is the secrecy in the field that Lilly will not specify how far the compound, called LY544344, had progressed before being put on hold, other than saying it was in “late-stage development”. Lilly began testing the drug in humans in the mid-90s and says the clinical results looked promising. The compound is the only mGluR-targeting drug that has been tested in humans to date. Lilly says the significance of the mouse seizures for humans remains unknown.

The drug represents a wholly new approach to treating anxiety. Existing medications, such as diazepam (Valium) and the other benzodiazepines, quiet the brain by ramping up the inhibitory GABA system. But they also cause grogginess and can be addictive. LY544344 decreases glutamate’s excitatory action by binding to pre-synaptic mGluR2 receptors and inhibiting glutamate release.

The seizures occurred in mice given chronic high doses, a routine part of testing for drugs intended to treat lifelong conditions such as anxiety. It is no real surprise that the molecule induces seizures: LY544344 has a degree of selectivity for mGluR2 receptors, but it also binds to other glutamate receptors so it lacks the specificity drug makers search for. Lilly says it has not given up on the drug and is making new derivatives to test. It is also working with the FDA to design new clinical trials.

But even failure will not mean the end of the mGluR story. Other researchers are taking a different tack that promises to make the problems encountered by LY544344 look like a hiccup rather than a terminal blow. “The molecule is a breakthrough, but it’s not exactly the type of molecule we want to develop,” says Vincent Mutel, president of Addex Pharmaceuticals, a drug development company in Geneva, Switzerland, which focuses on addiction and is developing mGluR drugs. He says the way forward is not to go straight for the glutamate binding sites, but to tinker with them indirectly. “We are more interested in modulators,” he says. “They are a safer and softer way.”

Imagine the receptor as a lock and glutamate as the key. LY544344 and glutamate act directly on the lock. But other compounds bind at sites on the receptor molecule close to the lock, changing how easily it is opened or how long it stays open. These spots, known as allosteric binding sites, vary from receptor to receptor, and each has its own key.

According to Conn, allosteric binding sites have previously been found on ionotropic receptors, but their discovery on metabotropic receptors is very recent. No one knows what, if anything, binds to the sites naturally. It’s possible that they have no normal function, but just happen to bind drug molecules.

What seems clear, though, is that the allosteric binding sites are a golden opportunity waiting to be exploited. Because the sites do not bind glutamate itself, they are all different. The result is a rich diversity of dimmer switches – each with a different compound to turn them up or down. What is more, allosteric modulators have already been exploited elsewhere, with some success. Diazepam, for example, is an allosteric modulator of the GABA ionotropic receptor, which acts to make the receptor more sensitive.

Most mGluR researchers now agree the future lies in allosteric modulators. “We have proof-of-concept drugs, and lots of companies are racing to develop other drugs,” says Mutel. But they admit these are still some way off entering human trials. Darryle Schoepp of Lilly Research Labs in Indianapolis, Indiana, who helped develop LY544344, adds: “Until we get more compounds moved into clinical trials and results start coming out, we won’t have an understanding of their potential.”

But already, researchers have promising early results for Parkinson’s disease, anxiety, addiction, chronic pain and schizophrenia (see “Glutamate: what can it do?”). Many of these compounds are certain to fall by the wayside, and any that succeed will take several years to reach the market. But a field of biomedicine that has had no new concepts for decades can surely wait a little while longer.