Source: The Scientist
Date: 21 Jan 2002

Phenotype Offers New Perception on Cocaine

Researchers say glutamate is more essential to addiction than dopamine

By Tom Hollon

In cocaine research, dopamine and glutamate make a brilliant star and supporting player, respectively. One takes center stage, the anointed crowd-pleaser; the other, though a leading actor in other productions, is so overshadowed that admiration of its performance is relegated to an acquired taste. A quick PubMed search recently disclosed their perceived importance: 3,628 abstracts on cocaine and dopamine, 178 for cocaine and glutamate.

Now, however, perceptions may shift - not that dopamine descends from the firmament, but that glutamate will sparkle as brightly. Recent knockout mouse evidence1 from researchers led by François Conquet, CEO of Addex Pharmaceuticals in Geneva, Switzerland, reveals that glutamate's role in cocaine dependence is even more central than dopamine's.

The case for dopamine's centrality remains airtight. Cocaine binds the dopamine transporter, blocking reuptake of dopamine into presynaptic neurons. Blockade increases dopamine concentration in synapses, an event responsible for cocaine's pleasurable effects and suggested as key to developing drug dependence. But although loss of the transporter and dopamine receptors in knockout mice may alter behavior toward cocaine, always the drug remains addictive. When the dopamine transporter is lost, for instance, mice may still become cocaine dependent through cocaine's ability to bind the serotonin transporter. This is not necessarily surprising, observes Peter Kalivas, of the Medical University of South Carolina in Charleston, who is a leading investigator of the glutamate-cocaine relationship. "The ability of an organism to predict rewarding stimuli in the environment is absolutely critical to survival," says Kalivas, "so there probably is some redundancy."

Contrast this redundancy to what Conquet finds when metabotropic glutamate receptor mGluR5 disappears: Without mGluR5, mice turn up noses and whiskers to cocaine, even though their dopaminergic systems respond to cocaine as usual. "These are the first knockout mice completely unresponsive to this powerfully addictive drug," says Conquet, who engineered the knockout mice when he was at GlaxoSmithKline in Lausanne, Switzerland. From this phenotype emerges a new picture of dopamine and glutamate: Sustaining cocaine-seeking behavior requires both neurotransmitters, while only glutamate is indispensable for cocaine dependence.

The Consolation Prize

Glutamate, the main excitatory neurotransmitter, is associated with learning and memory. Its receptors divide into ionotropic and metabotropic forms important to this function. "Learning occurs in part from changes in both ionotropic and metabotropic signals," Kalivas explains. "You adjust both in order to change the way cells communicate." Ionotropic receptors are also called ligandgated ion channels.

Ligand binding opens the channel so ions can pass through the cell membrane. Generally these are ion channels first, receptors second, controlling very quick changes in membrane current. In comparison, metabotropic glutamate receptors bring slower changes; largely they modulate signals from other neurotransmitters, acting through second messenger systems. They belong to the seven-transmembrane, G-protein linked superfamily of receptors. Conquet studies metabotropic receptors mGluR1 and mGluR5, which act through the phospholipase C signaling pathway.

Conquet's discovery of mGluR5's role in cocaine addiction originates in his second-place finish in a race to make mGluR5 knockout mice. Conquet was at the time head of Glaxo's experimental pathology unit, where his job was to make knockout mice deficient in various neuronal receptors. He lost to John Roder, of the Hospital for Sick Children, in Toronto. By showing that mGluR5 mutant mice perform poorly in the Morris water maze test and in fear-related learning, Roder implicated mGluR5 in spatial learning and memory.2 Roder's experiments suggest that mGluR5 is involved in long-term potentiation (LTP) within the hippocampus. Scooped, Conquet had to ask himself if Roder's description of the phenotype was complete. A possibility Roder might have missed, Conquet decided, was how the mice would react to cocaine.

The possibility of a connection between mGluR5 and cocaine appealed to Conquet's sense of drug dependence as a form of learned behavior. He knew that cocaine increases glutamate concentration in the nucleus accumbens, a brain region associated with cocaine dependence and stimulation of locomotor activity, and the location for the natural reward circuitry for food, water, mating and maternal behavior. Kalivas and his associates have demonstrated that mGluR5 receptors are down regulated following chronic cocaine administration.3

Conquet turned for help to his colleague Christian Chiamulera, who works on psychiatric drugs in a Glaxo laboratory in Verona, Italy. Willing to take a flyer on a wild idea, but wanting to avoid weeks of work with nothing to show for it, Chiamulera suggested a quick-and-dirty observation of cocaine as a psychostimulant: Inject cocaine into the bellies of the knockouts, then look for hyperactivity.

When Conquet watched the first injections, immediately he worried something was wrong. Instead of frenzied exploration of their surroundings, the mGluR5 knockouts lolled about as if nothing had happened. They verified in fact that the mice had received walloping doses. To their excitement, wild type mice on cocaine behaved as expected - no sleepwalkers or indolent beachcombers here. These creatures were ready to jitterbug 'til dawn at Mardi Gras. Conquet and Chiamulera were ready to join them. Chiamulera would now follow up with more elaborate experiments. For a test that approximates cocaine addiction in humans, Conquet brought in Mark Epping-Jordan, another Glaxo scientist, to do cocaine self-administration experiments.

Chiamulera confirmed his initial results. Wild type mice responded to cocaine in a dose-dependent manner: The higher the dose, the more hyperactive they became. Knockouts remained unperturbed regardless of dose. Abolishing mGluR5 abolishes cocaine-induced hyperactivity.

Epping-Jordan began the self-administration experiments by first training knockout and wild type mice to press a lever in order to get food. Both groups learned lever pressing equally well. Then he substituted intravenous cocaine for food and watched what happened. Wild type mice responded enthusiastically to the new menu, and would press for cocaine a dozen or more times an hour. MGluR5 mutants ignored cocaine at every dose; within a few sessions they would learn levers no longer produced food and stop pressing.

It was possible that the connection between cocaine dependence and mGluR5 was indirect, that loss of mGluR5 during development altered molecules even closer to control of dependence. Conquet's group examined the issue by asking if 2-methyl-6-(phenylethynyl)-pyridine (MPEP), a selective mGluR5 antagonist, reduced cocaine self-administration in normal mice. It did—In dose-dependent fashion, MPEP decreased demand for cocaine by up to 50%. The link, then, is direct and essential. "Somehow, glutamate transmission at mGluR5 is critical for the animal to recognize the rewarding effects of cocaine," says Kalivas. "The surprising thing is that it must be a secondary effect, because cocaine does not act directly on glutamate transmission. There is no binding by cocaine directly to any protein that has to do with glutamate transmission."

Leaving Natural Reward Along

Loss of mGluR5 apparently leaves the dopaminergic system intact. Using microdialysis to measure dopamine in freely moving mice, the researchers found dopamine concentrations in the nucleus accumbens were the same for mutant and normal mice, with or without cocaine. Levels of D1 and D2-class dopamine receptors and dopamine transporters were also normal.

Most striking is that reward systems strongly influenced by dopamine - nourishment, mating, and nursing - were also unaffected by loss of mGluR5. Conquet emphasizes that no other knockout has behaved this way: "This is the first time a mammal has been found insensitive to cocaine while its other reward-based systems remain normal."

He continues, "Dopamine receptor knockouts fail to curb cocaine dependence because mGluR5 is still working. They just affect general dopaminergic activity," and with considerable "collateral damage". Experiments with dopamine receptor agonists also indicate that dopamine does not lie at the center of cocaine dependence: "People have shown that you can never induce dependence from scratch with dopamine agonists. But you can maintain the process with these compounds once dependence is ongoing, probably after mGluR5 has turned the system on."

Kalivas now distinguishes dopamine and glutamate by their short and long term effects. "The acute rewarding properties that keep people coming back to the drug are mediated by dopamine," he says. "The 'Once an addict, always an addict' kinds of folklore that really make an addict are probably long-term changes in glutamate transmission." In retrospect, this isn't surprising: "All of neuroscience has been pointing to glutamate transmission as the critical player in the brain's ability to adapt to the environment." Cocaine addiction is one such adaptation.

From Scientist to Entrepreneur

Conquet founded Addex only a few weeks ago, departing Glaxo for better opportunities to continue his work. Following Glaxo's merger with SmithKline, drugs against cocaine addiction seemed better markets for smaller companies. Glaxo's larger size demands larger markets if the pharma giant is to sustain itself. For a small firm like Addex, a new mGluR5 antagonist could be quite profitable. Why develop a new drug when MPEP exists? Because MPEP dissolves very poorly and barely crosses the blood-brain barrier, Conquet explains.

Conquet does not know if mGluR5 plays a role with ethanol and nicotine addiction. Self-administration experiments have not been done. Partly he hasn't had time, since he's busy starting Addex. Partly the mice haven't had time, since other drugs of abuse, especially alcohol, require longer training periods. Whether mGluR5 influences other so-called addictions, is a question left for the distant future.

If Addex does find a good mGluR5 antagonist, therapeutic possibilities may extend well beyond helping snorters and crackheads stay clean. Glutamate may be implicated in numerous psychiatric disorders, according to Kalivas. mGluR5 inhibitors have been suggested as possible treatments for Alzheimer Disease, Parkinsonian akinesia, muscle rigidity, stroke, anxiety, and inflammatory pain. But as always, Kalivas reminds, once a good drug candidate is in hand, only running the clinical experiments will tell for sure.
Tom Hollon ( is a freelance writer in Rockville, Md.

1. C. Chiamulera et al., "Reinforcing and locomotor stimulant effects of cocaine are absent in mGluR5 null mutant mice," Nature Neuroscience, 4:873-4, September 2001.

2. Z. Jia et al, "Gene targeting reveals a role for the glutamate receptors mGluR5 and GluR2 in learning and memory," Physiology and Behavior, 73:793-802, August 2001.

3. C.J. Swanson et al., "Repeated cocaine administration attenuates group I metabotropic glutamate receptor-mediated glutamate release and behavioral activation: a potential role for Homer," Journal of Neuroscience, 21:9043-52, Nov. 15, 2001.

01 02 03 04 05 06 07 08 09 10 11 12
13 14 15 16 17 18 19 20 21 22 23 24