Symposium 34
Cutting-edge translational research in schizophrenia -from neurophysiological perspective-
シンポジウム34
統合失調症のトランスレーショナル研究最前線 -神経生理学的研究的観点から- |
|---|
| SY34-1
Neurophysiological investigations into early psychosis using translatable brain markers
トランスレータブル脳指標を用いた早期精神病の神経生理学的研究
Kasai Kiyoto（笠井 清登）
Department of Neuropsychiatry, The University of Tokyo
Schizophrenia research has been limited in reciprocally translating findings between rodent models and human neuroimaging. Here we propose a concept of “translatable” brain markers (Okano et al., 2016) and describe cutting-edge efforts. The onset of schizophrenia is usually at adolescence and young adulthood, and the cognitive dysfunction persists for life-long in some patients. Structural MRI studies have shown progressive decrease of neocortical gray matter volume in early stages of schizophrenia that was coupled with an abnormality of neurophysiological index of glutamatergic neurotransmission called auditory mismatch negativity (MMN) (Kasai et al., 2003; Salisbury et al., 2007; Nagai et al., 2013; 2017). Patients show abnormal auditory steady state gamma-band oscillations (ASSR) (Tada et al., 2016), which is thought to be associated with dysfunction in GABA interneurons. On the other hand, rodent model and human postmortem studies have indicated that insult to dendritic spines through glutamatergic/GABAergic dysfunction may underlie the peri-onset progressive pathology in schizophrenia. However, there has been no direct evidence of synaptic dysfunction in schizophrenia, a missing link between animal/postmortem and in vivo human studies. To bridge the gap, “translatable” brain markers should be developed using neuroimaging and electrophysiological indices that can be commonly measured in animals and humans. We here present MMN and gamma oscillation data on primates, humans, and patients with schizophrenia. The bidirectional animal and human research using the translatable brain markers such as MMN/ASSR will facilitate identification of effective molecular targets for early intervention for schizophrenia. | | SY34-2
Functional connectivity rs-fMRI marker for schizophrenia spectrum disorder
Giuseppe Lisi
ATR Computational Neuroscience Laboratories, Department of Brain Robot Interface, Kyoto, Japan
Neuroimaging and machine learning have been recently used to identify abnormal functional connections associated with mental disorders. For instance, classifiers were built to discriminate between healthy control individuals and patients with a specific psychiatric condition. However, prior to our recent work on autism (Yahata et al. 2016) , classifiers have only been trained and tested using data from a single fMRI scanner, omitting a strict validation on independent cohorts. Such validation remains challenging due to the curse of dimensionality and to the confounding effects of varying scanning conditions and demographic distributions. To alleviate these issues, our method (Yahata et al. 2016) imposes sparsity constraints on a cascade of canonical correlation analysis and logistic regression models. This allows to train a classifier while discarding irrelevant features and simultaneously reducing the effects of confounding variables (e.g. scanning site, age, sex) that may cause catastrophic over-fitting. In our new study, we trained the classifier on data from 68 patients with schizophrenia spectrum disorders (SSD) , mainly at a chronic stage, and 102 healthy controls in Japan. Then, we investigated generalizability to various independent cohorts from different countries, stages of SSD (i.e. chronic and first episode) , and categories of mental illness. Results show that the classifier does generalize to data from chronic patients of other countries, but does not to first episode SSD or to other psychiatric disorders. Additionally, we explored the possibility of using classifiers trained on different psychiatric disorders to study cross-disorder overlap. One of the most intriguing results suggests an overlapping but asymmetrical relationships between SSD and autism. | | SY34-3
Neural Oscillation Abnormalities in Psychosis
精神病における神経同期活動異常
Hirano Yoji（平野 羊嗣），鬼塚 俊明，神庭 重信
Department of Neuropsychiatry, Kyushu University
The diagnosis of psychiatric disorders has been made largely based on narrative interactions between patients and clinical professionals. Such methods without any neurobiological perspective could run the risk of producing a high variance in diagnosis or a diagnosis without solid biological validity. Although studies of psychosis such as schizophrenia have traditionally focused on deficits in higher order cognition that affect psychosocial outcome in this disorder, there is an increasing evidence that deficits can be found even at the level of early sensory processing including oscillatory activities in their brain. These deficits are suitable for translational research and represent potential novel targets for clinical applications. In this presentation, I would like to review neurophysiological findings in psychosis, especially in auditory event-related potential or magnetic field and spectral analysis, which can provide effective clinical applications both for characterizing the deficits in psychosis and for linking them to underlying pathophysiological mechanisms. | | SY34-4
Spontaneous Gamma Activity Indexes Synaptic E/I Imbalance in Schizophrenia
Spencer Kevin M.
Department of Psychiatry, Harvard Medical School, USA
Broadband, spontaneous gamma (SG; 30-100 Hz) activity in the EEG may reflect the balance of synaptic excitation and inhibition (E/I balance) in the cortex, which is disrupted in schizophrenia (SZ) and other neuropsychiatric disorders. I will review work from our laboratory concerning SG in SZ and its implications for cortical circuit abnormalities. In this work, gamma activity was measured from scalp EEG recordings in healthy controls (HC) and individuals with chronic SZ. MRI was used to measure gray matter volume. 1) We found a modality dependence of increased SG, such that it was increased in SZ compared to HC during auditory but not visual tasks. 2) In auditory cortex, SG and cortical volume were inversely correlated in SZ, but the coupling of SG power to low frequency phase was unaffected in SZ. 3) During auditory oddball performance, SG was increased in SZ only in younger participants, although SG increased with age in HC but not SZ. These studies suggest that cortical E/I imbalance in SZ may be most prevalent in the auditory system. While gamma deficits are commonly ascribed to inhibitory interneurons, the relationship between SG and auditory cortex volume in SZ suggests that increased SG and hence abnormal E/I imbalance are actually due to reduced synaptic connectivity. This link is supported by the normal phase-amplitude coupling of SG in SZ, implying that inhibitory interneuron function is preserved in SZ auditory cortex. The occurrence of increased SG in younger SZ and the lack of an increase in SG with age in SZ imply that E/I imbalance is decoupled from age-related changes in the cortex in SZ. In sum, increased SG in SZ appears to reflect a modality-specific E/I imbalance due to reduced synaptic connectivity that is maintained during aging. | |
|
|
