Progress Report 2009: The Quest for Better Schizophrenia Treatment
Serendipity and Science

by Hakon Heimer

January, 2009

In the early 1950s, a chance discovery helped transform schizophrenia from mystical affliction to medical disorder. French psychiatrists discovered that chlorpromazine, a drug used to make surgical patients less anxious, also relieved the symptoms of psychosis. The subsequent discovery that chlorpromazine targeted a brain messenger molecule called dopamine kicked off a large research effort into dopamine dysfunction in schizophrenia.

A half-century later, another avenue of research has yielded exciting results. Two recent studies—one a clinical drug trial and another a basic science study in laboratory mice—have helped turned the focus to a different messenger molecule, or neurotransmitter, called glutamate. In September 2007, researchers at Eli Lilly published a study showing that an experimental compound that inhibits glutamate signaling was able to reduce psychosis.1 Although the trial awaits confirmation, and it remains to be seen whether this particular compound will be any more effective or have fewer side effects than the older drugs, the results validate the basic neuroscience research and purposeful drug development that offered up the first successful new drug target in more than half a century.

Reinforcing this new emphasis on non-dopamine causes of schizophrenia, researchers at the University of California, San Diego, reported in December 2007 that interfering with glutamate signaling in their mouse model also disrupted brain cells that use yet another neurotransmitter, this one called gamma-aminobutyric acid (GABA).2 The fact that the GABA cell alterations mimicked those seen in schizophrenia may help to unite two prominent, and competing, theories of schizophrenia causation.

Schizophrenia Without Drugs

Although chlorpromazine (later sold as Thorazine in the United States) rescued schizophrenia sufferers from failed treatment strategies such as electroshock therapy, induced insulin shock, and frontal lobotomy, it did not restore full functionality to patients, as disabling cognitive and motivational symptoms persisted. Indeed, even today researchers are only in the infancy of understanding a disorder that was reported in historical texts as early as Pharaonic Egypt.

An important turning point in understanding psychotic disorders came around the turn of the twentieth century, when the German psychiatrist Emil Kraepelin distinguished two types of disorders featuring delusions, hallucinations, and other thought disruptions. The major distinction between “dementia praecox” and “manic depression,” Kraepelin postulated, was that although people with manic depression (now called bipolar disorder) might experience psychosis during manic periods, they return to relatively normal cognitive function when they come down from the mania. For people with dementia praecox, later termed “schizophrenia” by Kraepelin’s countryman Eugen Bleuler, psychosis is an ongoing state, often accompanied by profound deterioration in the ability to process information or interact socially.

The modern diagnosis of schizophrenia requires the persistence of psychotic, also called “positive,” symptoms for at least six months, without evidence of mood cycling. However, psychosis is not the only symptom of schizophrenia. Most people with the disorder also exhibit poor working memory (information stored temporarily during a task) and are unable to quickly recognize new situations and rules, an ability termed cognitive “flexibility.” These cognitive features contribute greatly to the chronic disability of most patients, as does a third symptom domain, that of “negative” symptoms. Negative symptoms describe aspects of normal behavior that are subtracted by the disease process—typically motivation, the display of emotion, or the desire to interact with other people. Thus, the most severely afflicted find themselves in a constant state of confusion about the events going on around them, without the capacity to have normal social interactions.

The Chlorpromazine Puzzle

The work of the German classifiers and their contemporaries had no direct benefit for people with schizophrenia and other psychotic disorders. Indeed, the next half-century saw some horrifically misguided attempts to alleviate the suffering of patients and their families. Treatments such as the surgical disconnection of major brain pathways with frontal lobotomy were the result of physicians’ moving forward with slim scientific evidence.

Finally, in the second half of the twentieth century, antipsychotic drugs provided a logic and a strategy for looking for chemical or structural changes in the brains of people with schizophrenia. If a single molecule—chlorpromazine—could reduce and in some cases eliminate the complex manifestations of schizophrenia, then it stood to reason that there was a chemical imbalance in the brain.

The Swedish scientist Arvid Carlsson had discovered dopamine in the early 1950s, and a decade later he and his colleagues determined that antipsychotic drugs worked by blocking dopamine from attaching to its receptor molecules. This finding dovetailed with another serendipitous finding: as early as the 1930s, it had been noted that amphetamine could cause psychosis. Amphetamine and other psychostimulants, it turns out, boost the activity of dopamine. Thus, the “dopamine hypothesis” of schizophrenia was born.

For the next several decades, researchers focused on trying to understand how dopamine systems were disturbed in the disorder. However, despite some significant refinements to chlorpromazine, especially reductions of some side effects, this line of research has been disappointing. According to Joseph Coyle of Harvard University, one of the first schizophrenia researchers to turn their attention to glutamate, 70 to 80 percent of patients with schizophrenia treated with dopamine drugs remain profoundly impaired by cognitive and negative symptoms. Moreover, neither a clear understanding of how blocking dopamine receptors curbs psychosis nor any new molecular targets have emerged from this line of research. Most psychiatry researchers are currently of the opinion that the dysfunction of dopamine neurotransmission in schizophrenia results from, or compensates for, a more fundamental or “upstream” disturbance of the nervous system, perhaps in glutamate signaling.

The Biggest Little Neurotransmitter You’ve Never Heard Of

While the public has had many opportunities to learn about the important role of dopamine in the brain, especially in regard to Parkinson’s disease and the rewarding effects of sex, drugs, and chocolate, glutamate remains relatively unknown. In fact, it is the most common neurotransmitter in the brain and the signaling molecule of choice of the powerful pyramidal neurons. So named for their shape, these cells send information shooting around the cerebral cortex and other brain areas that control behavioral functions, rapidly combining sensory input with stored information and emotions.

One reason for glutamate’s anonymity in the public mind is that its status as a neurotransmitter was demonstrated only some twenty years ago. Researchers had long known that it was abundant in the brain, but it is involved in numerous other cell activities as well. In order to qualify glutamate as a true neurotransmitter, scientists had to establish that it was released by nerve cells at the ends of long fibers called axons. Researchers also showed that glutamate, like dopamine and all other neurotransmitters, crosses a narrow space beyond the axon called the synapse and binds to receptor molecules on the surface of other neurons, triggering rapid electrical or chemical activity in the second cell.

Once scientists had established the status of glutamate, especially in the cortex, they were quick to explore the possibility of glutamate dysfunction in schizophrenia. Here another bit of serendipity came into play. As with amphetamine and the dopamine hypothesis, it involved a drug that caused psychosis.

Remarkably, the drug phencyclidine was another gift from the anesthesiologists. Developed as an anesthetic in the 1950s, phencyclidine (PCP, called “angel dust” as a street drug) was soon pulled from regular use because it caused psychotic symptoms during recovery. But whereas amphetamine produces only the positive (psychotic) features of schizophrenia, PCP and chemically similar anesthetics such as ketamine produce both negative and cognitive symptoms as well. David Lodge and colleagues at the University of London supplied the link to glutamate in 1983, when they found that PCP and ketamine bind to one particular type of glutamate receptor called the N-methyl-D-aspartate (NMDA) receptor.


Activation of the N-methyl-D-aspartate (NMDA) receptor occurs when either glutamate (Glu) or NMDA and glycine (Gly) bind to the receptor molecule, shown on the left. A channel within the receptor complex enables molecules to cross the cell membrane. Magnesium (Mg) blocks this channel. When Mg is removed from the channel and the receptor is activated, calcium (Ca++) and sodium (Na+) ions enter the cell and potassium ions (K+) leave.  (J.D. Thomas and E.P. Riley / NIH National Institute on Alcohol Abuse and Alcoholism)

Researchers soon began to advance theories about how glutamate dysfunction might play a role in schizophrenia, led by Daniel Javitt of the Nathan Kline Institute in New York in 1987, as well as Joseph Coyle and his colleagues and John Olney of Washington University in St. Louis. If PCP and ketamine produced a schizophrenia-like state by interfering with normal NMDA receptors, then perhaps these receptors were performing poorly in the disorder. Evidence emerged from studies of postmortem brain tissue—some, but not all, such studies have found modest evidence of alterations in glutamate-related molecules in the brains of people with schizophrenia. The glutamate hypotheses have also gained support from genetic research—among the genes that have the strongest support as schizophrenia susceptibility candidates are several that code for proteins that influence glutamate signaling. In particular, a meta-analysis of genetic studies, by Lars Bertram and colleagues at Massachusetts General Hospital and published in 2008, found that variation in one of the subunits that makes up the NMDA receptor increases the risk for the disease.3

Researchers, particularly Javitt and Coyle, have conducted clinical trials to boost the function of NMDA receptors. Although negative and cognitive symptoms improved in these small preliminary trials, the results were not strong enough to induce the pharmaceutical industry to pursue drug development efforts. However, a breakthrough at the turn of the new millennium has led to a revitalizing of these approaches.

A Different Window onto the Glutamate Synapse

While Coyle, Javitt, and others were focused on modulating the NMDA receptor directly, Bita Moghaddam at Yale University had turned her attention to a different class of glutamate receptor. Called metabotropic glutamate receptors (mGluRs), they do not rapidly convey information at glutamate synapses, as NMDA receptors do. Rather, they influence how the glutamate synapse operates in various and subtle ways. In a paper published in 1999, Moghaddam and colleague Barbara Adams took advantage of the fact that PCP and other NMDA-interfering drugs can be used in animal models, where they produce effects strikingly like the negative and cognitive symptoms of schizophrenia patients.4 When they gave rats a drug that activates only mGluRs, the researchers found that the cognitive effects of PCP—e.g., working memory impairment—were significantly reduced. In 2005, John Krystal of Yale University and his colleagues replicated this finding in humans, showing that the same mGluR receptor drug could reverse the cognitive effects of ketamine in healthy volunteers.

These studies set the stage for Eli Lilly to try the mGluR drug in people with schizophrenia. As the Lilly researchers reported in their 2007 paper in Nature Medicine, in a double-blind, placebo-controlled trial conducted in Russia with nearly 200 patients, they found that the experimental drug was significantly better than the placebo in treating positive symptoms, the first non-dopamine blocker to achieve this distinction.1 They did not report on whether they had found effects on cognitive measures, as Moghaddam and Krystal had in their experiments. The study is now being repeated in a different group of patients, with different doses of the mGluR drug. The Lilly trial provides not just proof of concept evidence for the target, but also proof of concept evidence that the strategy can yield new drugs that may turn out to be effective, according to David Lewis of the University of Pittsburgh’s Western Psychiatric Institute, one of the researchers who had demonstrated changes in glutamate-related molecules.

It’s Not All Excitatory

Another significant glutamate study published in 2007 pointed out both the significant advances and the remaining challenges in understanding schizophrenia pathology. Margarita Behrens, Laura Dugan, and their colleagues at the University of California, San Diego, reported in Science that interfering with glutamate NMDA signaling in mice can reproduce one of the most well-supported findings in schizophrenia: disruptions of cells called interneurons, which signal using the neurotransmitter GABA.2

GABA is the yin to glutamate’s yang. While glutamate is the neurotransmitter of choice for the pyramidal neurons, which use it to excite electrical activity in the neurons they contact, GABA is typically employed by a more modest group of cells, the interneurons. These cells confine their axons to their local areas, in which their bursts of GABA inhibit the activity of the pyramidal neurons. Researchers including David Lewis, Francine Benes at McLean Hospital in Belmont, Massachusetts, and others have found changes in GABA-related proteins in schizophrenia but only in a select population of interneurons. A recent study in genetically altered mice also points to GABA cell problems in schizophrenia. Akira Sawa and colleagues inserted a mutant form of the schizophrenia susceptibility gene called “disrupted in schizophrenia 1” (DISC1) into mice.5 When they examined the brains of the mice, the researchers found that the same interneurons affected in schizophrenia are altered in the mice with mutant DISC1. The researchers described their work in a 2007 paper in Proceedings of the National Academy of Sciences.

Lewis and his collaborators are now testing a drug that may normalize GABA interneuron function in the brain of people with schizophrenia, perhaps with a beneficial effect on symptoms.



Margarita Behrens and colleagues found that abusing ketamine, which inhibits the NMDA receptor, can result in symptoms indistinguishable from schizophrenia in the mouse prelimbic cortex. At right, ketamine has reduced the expression in the prelimbic cortex of parvalbumin (the large, light-colored spots), a molecule reduced in GABA cells in schizophrenia. (M. Margarita Behrens, Ph.D. / University of California, San Diego)

However, Behrens, Dugan, and colleagues’ intriguing finding is that the NMDA blocker ketamine selectively damages this same group of GABA interneurons. The researchers suggest that the glutamate deficit might therefore be “more primary,” or “upstream” of the GABA deficit. The intermediate step, their report suggests, may be the production of destructive molecules called “free radicals.” These results have not been confirmed, so the conclusions vis-à-vis schizophrenia remain speculative. But in 2008 Behrens, working with John Lisman of Brandeis University, added supporting evidence for the link between glutamate and GABA disruptions.6 As they reported in the Journal of Neurophysiology, they were able to directly record altered electrical activity in GABA interneurons that had been disrupted with an NMDA receptor blocker.

Searching Between Drugs and Behavior

The results of Behrens and colleagues highlight the need to work out the complex set of relationships between the different types of neurons in the brain, said Coyle. The strategy of giving different compounds to animals or people and studying how their behavior changes, which has been so productive in the case of the NMDA-blocking drugs, still leaves a fuzzy area between the drug input and the behavior output. In addition to the ongoing debate on the relative importance of glutamate versus GABA disruption in schizophrenia, researchers disagree on whether the neurons of most interest are those in the higher-reasoning areas of the cortex, in areas that connect the cortex with sensory or movement regions of the brain, or both.

Scientists will now attack the neuronal circuits from different entry points: they will explore metabotropic glutamate receptors (of which eight variations have been identified), as well as the manipulation of other glutamate receptors and molecules that help control the amount of glutamate floating about in synapses. Researchers will also focus on the cells on the receiving side of glutamate neurotransmission, principally the GABA interneurons and their connections back to the glutamate-releasing pyramidal cells. Other neurotransmitters, such as dopamine, acetylcholine, and serotonin, will receive attention because they subtly alter communication between glutamate and GABA cells. It remains to be seen whether any single approach will lead to a drug that effectively treats schizophrenia, or whether different compounds, targeting separate neurotransmitter systems for the different symptom domains, will be needed.


1. Patil ST, Zhang L, Martenyi F, Lowe SL, Jackson KA, Andreev BV, Avedisova AS, Bardenstein LM, Gurovich IY, Morozova MA, Mosolov SN, Neznanov NG, Reznik AM, Smulevich AB, Tochilov VA, Johnson BG, Monn JA, and Schoepp DD. Activation of mGlu2/3 receptors as a new approach to treat schizophrenia: a randomized Phase 2 clinical trial. Nature Medicine 2007 13(9):1102–1107.

2. Behrens MM, Ali SS, Dao DN, Lucero J, Shekhtman G, Quick KL, and Dugan LL. Ketamine-induced loss of phenotype of fast-spiking interneurons is mediated by NADPH-oxidase. Science 2007 318(5856):1645–1647.

3. Allen NC, Bagade S, McQueen MB, Ioannidis JP, Kavvoura FK, Khoury MJ, Tanzi RE, and Bertram L. Systematic meta-analyses and field synopsis of genetic association studies in schizophrenia: The SzGene database. Nature Genetics 2008 40(7):827–834.

4. Moghaddam B and Adams BW. Reversal of phencyclidine effects by a group II metabotropic glutamate receptor agonist in rats. Science 1998 281(5381):1349–1352.

5. Hikida T, Jaaro-Peled H, Seshadri S, Oishi K, Hookway C, Kong S, Wu D, Xue R, Andradé M, Tankou S, Mori S, Gallagher M, Ishizuka K, Pletnikov M, Kida S, and Sawa A. Dominant-negative DISC1 transgenic mice display schizophreniaassociated phenotypes detected by measures translatable to humans. Proceedings of the National Academy of Sciences USA 2007 104(36):14501–14506.

6. Zhang Y, Behrens MM, and Lisman JE. Prolonged exposure to NMDAR antagonist suppresses inhibitory synaptic transmission in prefrontal cortex. Journal of Neurophysiology 2008 100(2):959–965.