Principal investigators in the collaberative project are (l to r): ROBERT GREEN, GRANCINE BENES, ROBERT MCCARLEY and MPRTHA SHENTONAn oft-told story in the annals of psychiatry, and one as tragic as any in literature, tells of a young man or woman who, without any warning or perhaps only a few tell-tale signs, is struck down by schizophrenia.
"What we typically see is young people seemingly in the prime of their lives as adolescents-athletes, scholars, musicians-and they suddenly start withdrawing," says Francine Benes, Harvard Medical School associate professor of psychiatry. "They just withdraw to their rooms, and quietly and very surreptitiously become progressively dysfunctional. By the time their parents recognize the problem and take them to the hospital, they have florid hallucinations and delusions and thought disorders. They're already very sick."
For the past century, psychiatrists and neurologists have suspected that so devastating a psychiatric illness must be due to some defect in the brain. Yet, for the most part, they have been unable to link the hallmarks of schizophrenia-jumbled thoughts, bizarre delusions, and hearing voices-with any demonstrable brain pathology. Nor have they been able to explain why schizophrenia emerges so suddenly during adolescence and young adulthood.
"Schizophrenia typically comes on about 18 or 19 years of age," says Robert McCarley, professor of psychiatry and deputy chief of staff for mental health services at Brockton/West Roxbury Veterans Administration Medical Center.
Now, McCarley and Benes, along with an interdisciplinary team of researchers drawn from HMS and Harvard College, may have an answer. If they are right, schizophrenics are born with subtle wiring defects in certain regions of the brain. These defects become exacerbated in schizophrenics as a result of stressful events around the time of birth and then again at adolescence. Ultimately, the result is an actual breakdown in the abnormal neural circuits, which, in turn, leads to the behavioral symptoms associated with schizophrenia.
From this vantage point, schizophrenia is the result of a complicated interplay of genetic and environmental factors. "I think it's healthy for us to view some of our findings as conceivably being related to a gene and some of the findings as being related to the effects of an insult, and that the two together are what make for a fully dysfunctional [neuronal] circuit," says Benes, director of the laboratory for structural neuroscience at McLean Hospital.
The Harvard researchers recently received a five-year, $2 million grant from the federal Veterans Affairs Medical Research Office to establish one of three national "Research Centers for Basic and Clinical Neuroscience Studies of Schizophrenia." The center, which will be located at the Brockton/West Roxbury Veterans Administration Medical Center, builds upon work done by HMS researchers over the past decade in psychiatry, neurobiology, psychology and radiology. The unique convergence of basic and clinical research could lead to a variety of treatments for people at risk for the disease, as well as for those already afflicted with schizophrenia.
An estimated 2.5 million Americans currently suffer from schizophrenia, making it one of the most common major psychiatric illnesses. Yet it is a commonly misunderstood disease, often confused with split personality. Although it means "split mind," schizophrenia is more accurately described as a split between thought and feeling.
"People will be unaware of why they have thoughts. They feel like sudden thoughts are planted in their minds or there are extraterrestrials who tell them they should go and kill somebody," says McCarley. "They may have strange feelings-unearthly terror-without having any idea about what's going on."
After nearly a century of fruitless searching for the biological causes of schizophrenia-an effort hampered by a lack of refined methods for detecting the subtle nature of the brain defects-schizophrenia research began to take off in the 1970s and early 1980s. McCarley, Benes and other researchers were spurred on by new methods and imaging tools for studying brain structure and function.
In 1983, McCarley embarked on a series of experiments in which he used multiple scalp electrodes to record brain activity. In an area located just behind the left ear, known as the temporal lobe, he detected an abnormal response in schizophrenics when they were processing stimuli that were unusual or surprising. Normally, people respond to novel stimuli quickly, and try to incorporate the new stimuli into a stable world view, a process known as cognitive updating. But in schizophrenics the response was slower and less marked, indicating some impairment in the updating process. "If you don't update correctly, maybe that has something to do with delusions," McCarley reasons.
Thinking that the other hallmarks of schizophrenia-jumbled word associations and hearing voices-might also be due to an abnormality in this area, McCarley, working together with Martha Shenton, HMS associate professor of psychology, began looking for structural defects in the temporal lobes of schizophrenics. They found that the left Sylvian fissure, the cleft just above the temporal lobe, was unusually large in right-handed schizophrenics. This widening of the fissure appeared to be due to a shrinking of the temporal lobe, though it was not clear where exactly the shrinking occurred.
In collaboration with radiologists at Brigham and Women's Hospital, McCarley and Shenton used magnetic- resonance-imaging techniques to determine that the shrinkage occurred in a structure known as the planum temporale. This region of the temporal lobe was abnormally small in about half of the schizophrenics studied.
"The degree of volume reduction in these people was highly correlated with their degree of thought disorder [as measured by a standardized test]. That was really exciting," McCarley says. The researchers also observed a loss of volume-but no strict correlation with behavioral symptoms-in two other areas of the left temporal lobe.
These findings raised a new round of intriguing questions. Is the reduction in the temporal lobe due to the wholesale loss of tissue or to the loss of certain kinds of cells? Is the reduction in size of the temporal-lobe structures a primary cause of schizophrenia or the consequence of more basic defects in other brain regions? Answers to these questions will come as researchers further explore the neural circuitry in the brain.
Meanwhile, Benes has been exploring a nearby structure of the brain, known as the anterior cingulate cortex. Her findings, based on postmortem examinations of nearly 40 brains of schizophrenics, provide a major clue as to the kinds of cellular changes that might be occurring in schizophrenics, and also what might be causing them. Whereas McCarley identified abnormalities in regional areas of the brain, Benes has focused on defects in the connections between neurons.
"I realized you could characterize thought disorders and illogical thinking in terms of altered cortical circuits. This just captured my imagination," says Benes.
One of the first things she observed in her anatomical investigations of the anterior cingulate cortex was a decrease in the number of neurons. This reduction, upon closer examination, was notable in two ways. First, the loss was most prominent in layer two of the cingulate cortex (like all structures in the human cortex, the cingulate has six layers). And it seemed to be restricted to a particular kind of neuron, known as an inhibitory interneuron.
Inhibitory interneurons regulate the excitability of other neurons, known as excitatory neurons, by releasing the transmitter gamma-aminobutyric acid (GABA), which essentially prevents the excitatory neurons from firing. Through this process, the nay-saying inhibitory interneurons modulate the flow of information through a neural circuit.
"I had been reading about the schizophrenic's inability to filter out extraneous stimuli and I thought you could really conceptualize that in terms of diminished inhibitory modulation in a cortical circuit. That was it-I knew that's where I had to go," recounts Benes.
To see if the missing neurons were producing GABA, and were therefore truly inhibitory interneurons, Benes looked for an increase in GABA receptors on the excitatory neurons. "You'd expect a compensatory increase of GABA receptors on [those] cells to optimize the cell's ability to respond to the transmitter that's diminished," she says. Sure enough, she and her colleagues found there was an increase-and the increase was most striking in layer two.
In another case of reality matching expectation, Benes speculated that the spaces left vacant by the loss of inhibitory neurons might be occupied by axons sweeping in from other areas of the cortex. She found that there were 70 percent more incoming axons in layer two of the anterior cingulate of schizophrenics. "I was just kind of shaking my head. And I thought, 'God bless hypothesis generation and testing,'" she says.
Principal investigators in the collaberative project are (l to r): ROBERT GREEN, GRANCINE BENES, ROBERT MCCARLEY and MPRTHA SHENTON
The picture of schizophrenia that is emerging from Benes' research is one of imbalance at the cellular level: too much excitation due to an increase in incoming axons, and too little inhibition due to the loss of inhibitory neurons. Moreover, this picture is gaining support from other researchers in the Harvard group who have been looking at two nearby structures, the amygdala and hippocampus.
In both of these areas, inhibitory interneurons and excitatory neurons are connected by a simple yet ingenious feedback loop (see diagram on next page). The loop is set in motion when the excitatory neuron fires, releasing glutamate into the synapse between it and the inhibitory interneuron. (Both the inhibitory and excitatory neurons are covered with special receptors for glutamate, known as NMDA receptors.) If enough glutamate is released-and this may require more than one burst of activity by the excitatory neuron-the inhibitory neuron will then fire, releasing the GABA transmitter into the synaptic cleft (and thus inhibiting the excitatory neuron). This process, in which the excitatory neuron inhibits its own actions via the inhibitory interneuron, is known as recurrent inhibition.
"It's as if the excitatory neurons were saying: 'If I get very excited, this is a way to prevent excess messages [from] being transmitted.' So it says, 'Well, I may be able to say once or twice to you something, but not a third, fourth, fifth or sixth time,'" explains Donald Rainnie, instructor in psychology.
Rainnie, along with Robert Greene, associate professor of psychiatry, and Heinz Grunze, research fellow in psychiatry, have found that the inhibitory interneurons are, in fact, exquisitely sensitive to drugs that block the NMDA receptors. And, tellingly, there is evidence that schizophrenics have higher levels of naturally occurring NMDA- blocking agents.
Guochuan Tsai, clinical fellow in psychiatry and Joseph Coyle, Eben S. Draper professor and head of the Department of Psychiatry, found increased levels of a naturally occurring NMDA-blocking agent, known as n- acetylaspartylglutamate (NAAG), in the hippocampus of postmortem brains of schizophrenics. Thus, it appears that loss of inhibition in the hippocampus may result from the inhibitory neurons-their NMDA receptors tied up-being cut off from the glutamate that normally triggers their firing.
The $2 million grant will enable the Harvard researchers to further investigate the biological underpinnings of schizophrenia. Benes, for example, will seek to understand what triggers the decreased numbers of GABA interneurons and the increased number of incoming axons in the anterior cingulate. She believes that the abnormal influx of axons may actually come first, killing the inhibitory GABA neurons.
"We think that these vertical fibers coming in may use glutamate, and glutamate can cause damage to neurons," says Benes.
She thinks that this abnormal influx of axons is genetically determined. To investigate if this is so, she is currently examining the brains of first-degree relatives of schizophrenic patients to determine if they have abnormally high numbers of axons. Yet the influx of axons may not by itself be enough to trigger schizophrenia. Rather it may need to be complemented by a series of stressful events beginning around the time of birth.
There is a higher than average incidence of birth- related complications among schizophrenics, Benes observes. "Nothing specific. They include anything from a viral infection in the mother to prolonged labor to a whole host of things."
These stressful events may trigger the release of hormones, known as glucocorticoids. Benes speculates that the stress hormones-which, in excess, can damage and even kill neurons-together with excess glutamate from the abnormal influx of axons may harm the inhibitory GABA cells in the anterior cingulate.
"All of these things are presumably dormant until some critical stage when they become manifest," Benes says.
The critical stage, she believes, occurs at adolescence, which would explain the sudden onset of schizophrenia among teenagers. In the anterior cingulate, neurons are still maturing at this time. The inhibitory GABA interneurons are increasing their synaptic contacts with cells producing the neurotransmitter dopamine. An exuberant release of dopamine, which can be triggered by stress, may prove damaging to those inhibitory GABA neurons that have already been impaired by earlier stress.
"So if you have circuits with intrinsic impairment in GABA interneurons, you could well imagine that such an individual under stress could just fall apart. The stress- induced dopamine would just blow the circuit to bits," says Benes.
This model of what goes awry in the anterior cingulate may apply to other areas of the brain. If so, stressful events could trigger impairment of inhibitory neurons in multiple areas of the temporal lobe, and the effect of the alterations in circuitry would be quickly apparent.
"The person could just fall apart. When schizophrenics are under stress, that is exactly what can happen. They can become acutely psychotic right before your eyes," Benes says.
Excitatory neurons inhibit their own activity by transmitting the excitatory amino acid (EAA) glutamate to the inhibitory interneuron.
This body of research by the Harvard scientists has obvious potential for the treatment of schizophrenics. For example, the researchers intend to pursue tests of a compound, d-cycloserine, which acts on the NMDA receptor, as a drug for schizophrenia.
About 20 percent to 30 percent of people who suffer a psychotic episode recover from it. The rest go on to develop schizophrenia. The Harvard researchers hope that by gaining a better understanding of the nature-nurture factors that trigger schizophrenia, and of the precise damage to cortical circuits, they may be able to develop novel treatments that can be applied early after onset of the disorder, or perhaps even before its appearance.
"Wouldn't it be very nice to have people in whom there is the vulnerability for schizophrenia, but the realization of its full pathological potential is thwarted?" asks Benes. "They could go on to live relatively normal lives."
--Misia Landau
Copyright 1995, President and Fellows of Harvard College. Multiple distribution or commercial reproduction by permission only.