PHENOTYPE OFFERS NEW PERCEPTION ON COCAINE

Researchers Say Glutamate Is More Essential To Addiction Than Dopamine

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 Francois 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
(thollon@starpower.net) is a freelance writer in Rockville, Md.

References

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.