8 Dec 2014
December 9, 2014. On Tuesday, November 18, participants in the 44th annual Society for Neuroscience meeting gathered to hear the latest electroencephalogram (EEG) and magnetic resonance imaging (MRI) studies, along with a smattering of cellular research, in schizophrenia and bipolar disorder. Session moderator Akira Sawa of Johns Hopkins University in Baltimore, Maryland, watched the clock in order to keep up with the ambitious schedule of 13 presenters.
Considering cognitive deficits
Rachel So of the University of Connecticut in Storrs discussed abnormalities in neural synchrony patterns during working memory in schizophrenia. She analyzed the phase-locking values, a measure of synchrony, between the dorsolateral prefrontal cortex (DLPFC) and other brain regions during a modified version of the Sternberg working memory task. In the left hemisphere, patients exhibited increased synchrony between the DLPFC and occipital lobe as task difficulty increased, while in the right hemisphere, they showed increased synchrony between the DLPFC and the temporal lobe. So suggested the neural patterns may reflect patients' use of a vision-based compensatory strategy to improve working memory deficits.
So's advisor, Chi-Ming Chen, discussed the use of machine learning to identify the features of EEG that are associated with working memory performance in schizophrenia. The approach was used to classify correct versus incorrect trials as well as identify diagnosis across the correct and incorrect trials. Overall, the approach achieved an "acceptable" accuracy of 74 percent. Chen also noted that the EEG data identified by the machine-learning approach recapitulated many of the findings in the healthy control and schizophrenia literature, further strengthening the utility of the approach.
Hamdi Eryilmaz of Massachusetts General Hospital in Boston discussed his event-related fMRI study that observed altered error-related activation during working memory in schizophrenia. Using the Sternberg Item Recognition paradigm to assess working memory at four different levels of difficulty, the researchers found that schizophrenia subjects displayed less deactivation of the medial prefrontal cortex (mPFC) than controls during incorrect performances. The degree of this mPFC deactivation correlated with task accuracy during incorrect trials: the schizophrenia subjects whose performance was more accurate displayed greater mPFC deactivation. Eryilmaz and colleagues noted that their data underscore the importance of taking performance differences into account when analyzing disease effects.
Elizabeth Andersen of the University of North Carolina, Chapel Hill, examined whether the attentional and affective processing deficits of schizophrenia extend to first-degree relatives (FDRs) and whether any EEG correlates of these deficits could be used as markers of susceptibility. Andersen and colleagues recorded EEGs during an oddball event-related potential paradigm and found that the schizophrenia group demonstrated the most disrupted selective attention, while the FDRs' performance was between that of patients and controls. As for affective processing, FDRs displayed higher late positive potential processing and elevated evoked potential in the beta frequency range that was not observed in either patients or controls, suggesting that these parameters could be used in the early detection of psychosis.
Next, another member of the lab, Alana Campbell, described her work using fMRI and EEG to explore executive and affective neural connectivity in patients with schizophrenia and FDRs. Both groups showed disrupted neural and functional connectivity, though the pattern was milder in FDRs. In addition, impaired connectivity during working memory worsened under emotional conditions. EEG synchrony was associated with functional frontal-to-frontal and cingulate-to-subcortical connectivity. Taken together, the results suggest that aberrant brain connectivity is associated with executive and affective deficits in schizophrenia, and, to a lesser degree, in FDRs.
Probing positive symptoms
Navin Gupta of Georgia State University in Atlanta presented evidence that reality distortion symptoms are associated with gray matter loss in schizophrenia. Using source-based morphometry to examine gray matter concentration, the researchers found that the amount of superior frontal and superior middle gyri gray matter was significantly correlated with reality distortion symptoms measured with the Scale for Assessment of Positive Symptoms (SAPS). In contrast, no association was observed with disorganized symptoms.
Hao Tan of the Lieber Institute for Brain Development in Baltimore examined the brain activity correlates of delusions in schizophrenia by using a numerical fMRI task in which subjects were instructed to respond when the numerical context changed. When inferring context change, patients showed less prefrontal cortical connectivity than controls. While assessing feedback about their predictions, patients had higher levels of prefrontal cortical-to-midbrain connectivity, resulting in greater midbrain activity. Tan said the results suggest that patients develop an enhanced learning of uncertain information that, when reinforced inappropriately, contributes to delusions.
Signatures in induced neuronal cells
In the few molecular studies of the session, a trio of speakers described recent data from Sawa's lab. First up was Yasue Horiuchi, who presented a molecular signature in olfactory neuronal epithelial cells that correlated with cognitive impairment in schizophrenia. Cells were collected using nasal biopsy, enriched using laser-capture microdissection, and gene expression examined using microarray. In particular, the researchers highlighted the SMAD family of proteins. There was a correlation between the levels of SMAD5 and cognitive function, and several members of the family were dysregulated in the microarray study, which will lead Sawa's group to continue examining the SMAD pathway in schizophrenia.
Next, Eleonora Passeri highlighted her improvement of the generally poor conversion rate of fibroblasts to induced neuronal cells (iNs) through the addition of valproic acid and a prostaglandin metabolite. She then discussed an analysis of the cellular and molecular signature of iN cells derived from patients with psychiatric illness-associated copy number variants. A duplication of 16p11.2 and a deletion of 22q11 affected iN cell conversion rate and subsequent neurite growth.
The third member of the Sawa lab, Hiroshi Yukitake, described the upregulation of micro-RNA 124 (miR-124) in both olfactory cells and postmortem brain tissue from subjects with schizophrenia and bipolar disorder. He also showed data indicating that mice overexpressing miR-124 exhibited an enhanced sensitivity for amphetamine, as well as an increased sensitivity to dopamine release in the nucleus accumbens following amphetamine treatment.
Rebecca Shafee of Harvard Medical School in Boston discussed an approach to estimate the heritability of intracranial volume (ICV) as well as the volume of several subcortical structures (using MRI) from single-nucleotide polymorphisms (SNPs) of unrelated participants. The researchers found statistically significant common SNP heritability for ICV as well as both the left and right caudate, suggesting that these genetic variations contribute to the scaling of the brain with head size. In addition to increasing sample size, future studies will separate this heritability according to the functional categories of SNPs.
Matthew Hudgens-Haney of the University of Georgia in Athens presented data on neural synchrony during pro- and anti-saccade tasks in a subset of the Bipolar-Schizophrenia Network on Intermediate Phenotypes (B-SNIP) consortium. Synchronization of the steady state visually evoked potential was assessed using intersensor phase coherence (ISC). In the period immediately preceding the presentation of the cue in the anti-saccade task, control subjects exhibited greater ISC than either schizophrenia or bipolar disorder subjects. However, the opposite was true during the same time period in the pro-saccade task. Taken together, these results suggest that the modulation of synchronized neural activity before cue presentation is altered in both illnesses.
Anouk Marsman showed her magnetic resonance spectroscopy findings of glutamate and GABA levels in the prefrontal cortex in schizophrenia (Marsman et al., 2014). Consistent with their hypothesis, the researchers showed a reduction in GABA/creatine in schizophrenia. This ratio was significantly negatively associated with cognitive functioning in patients. However, unlike previous studies, there was no change in glutamate levels. Marsman and colleagues attribute this lack of change to the relatively young age of the study participants.—Allison Marin.