13 Nov 2015
November 14, 2015. At a Tuesday, October 20, 2015, afternoon session of the Society for Neuroscience meeting in Chicago, there were a number of schizophrenia-focused talks in a symposium titled "Major Mental Disorders: Novel Approaches for Patient Evaluation."
First up was a talk that addressed both GABA- and glutamate-related hypotheses of schizophrenia etiology. The status of GABA and glutamate signaling in the schizophrenia brain, as measured by magnetic resonance spectroscopy (MRS), remains muddy, with studies disagreeing on whether, and in which regions, these neurotransmitters are increased or decreased (see SRF related news report; SRF news report). In a new attempt to get some traction on this question using "ultra-high" (7.0 Tesla) field strength MRS, Katharine Thakkar of the University of Michigan in East Lansing reported that both GABA and glutamate concentrations were reduced in occipital cortex in people with schizophrenia relative to healthy controls. Healthy first-degree relatives of the schizophrenia subjects, who are presumed to carry some genetic liability for the disorder, also showed reduced glutamate, though not GABA. This suggests that there is an underlying glutamate abnormality in people with schizophrenia that is not due to illness state or medication.
Another schizophrenia talk in the session, by Fabian Tremeau of the Nathan Kline Institute in Orangeburg, New York, described a framework for studying the ambivalent reactions that people with schizophrenia have to complex stimuli, and which may underlie the motivational component of negative symptoms. According to models from affective neuroscience, when humans encounter stimuli that have both positive and negative characteristics, two separate neural systems compete, with one typically winning out by inhibiting the other.
In Tremeau's study, even when schizophrenia patients rated pictures as extremely pleasant or extremely unpleasant, they reported more frequent ambivalent feelings than control subjects, suggesting that the cross-inhibition systems are not working properly. Moreover, the "negative" system seems to be particularly inefficient at inhibiting the positive one.
Noah Elkins of Johns Hopkins University in Baltimore, Maryland, described a new approach to the question of whether oxidative stress plays a role in schizophrenia. He and his colleagues detected higher levels of naturally occurring autofluorescence in blood-derived lymphoblasts from people with schizophrenia than from controls. Intriguingly, the levels of autofluorescence correlated with executive function in the schizophrenia patients. Autofluorescence stems in part from reactive oxygen species (ROS), and Elkins and colleagues found that the schizophrenia lymphoblasts had high ROS levels; when the ROS levels were reduced, autofluorescence dropped.
The researchers have explored this further in an animal model of mental illness, the disrupted in schizophrenia 1 (DISC1) dominant negative mouse, which exhibits oxidative stress. The mice also show elevated autofluorescence, in prefrontal cortex, and their cognitive deficits were reversed by interfering with downstream effects of ROS, specifically inhibiting glyceraldehyde-3-phosphate dehydrogenase (GAPDH).
New ways to subtype schizophrenia?
Aaron Sampson of the Salk Institute in La Jolla, California, previewed a new method that employs nonlinear dynamics (aka chaos theory) to unlock more data from the venerable electroencephalogram. Working with EEG data from the Consortium on the Genetics of Schizophrenia (COGS) project, Sampson and colleagues found that delay differential analysis (DDA) improved on traditional analysis methods of EEG data distinguishing patients with schizophrenia from controls on a sensory test of mismatch negativity. He also reported that DDA analysis allowed the researchers to identify schizophrenia "typical" patients from a group whose brain responses were more like those of controls.
The last schizophrenia talk of the session was from Leighton Hinkley of the University of California, San Francisco, who is interested in whether brain oscillations can tell us something useful about people with schizophrenia and their cognitive processes. Using magnetoencephalographic imaging (MEGI) of people during implicit learning tasks, he and his colleagues were able to correlate brain activity with acquisition of the task: A subgroup of people with schizophrenia who learn as well as control subjects (dubbed the schizophrenia learner, or SZ-L, group) nonetheless do not show the same brain activity as control subjects. That is, they achieve the learning through some compensatory process that the schizophrenia non-learner (SZ-NL) group does not have. In particular, Hinkley found that activity in the beta (12-30 Hz) and high gamma bands (65-115 Hz) in a network of sensorimotor cortical regions was suppressed relative to controls during the learning process for SZ-Ls, something not found in the SZ-NLs.
However, after the SZ-NLs had completed a computerized training paradigm, they acquired an ability to do the implicit learning task more like controls and SZ-Ls, and this was correlated with the capacity to suppress the high gamma band activity in cortical regions. The results support the notion that with the right kind of cognitive training, some forms of learning can be improved in people with schizophrenia, and this might generalize to greater functional improvement.—Hakon Heimer.