Neurocircuit for physiological and pathophysiological brain
JS4-2-2-1
Dynamic regulation of fear memory after retrieval; therapeutic targets for the treatiment of PTsD
○喜田聡1,2
○Satoshi Kida1,2
東京農業大学応用生物科学部バイオサイエンス学科1
Dept. of Bioscience, Tokyo Univ. of Agriculture1, CREST, JST2

Memory retrieval initiates two opposite processes; memory reconsolidation and extinction. Memory reconsolidation is thought to be a process to update, strengthen or maintain, whereas memory extinction is a process to weaken the original fear memory. These memory phases induced by memory retrieval are thought to serve as useful therapeutic targets for the treatment of emotional disorders such as post-traumatic stress disorder (PTSD) and phobias. Indeed, exposure therapy for the treatment of emotional disorders is thought to reflect some aspect of the biological basis of memory extinction. Therefore, from this clinical view, it is important to understand mechanisms underlying the regulation of fear memory after retrieval. We have investigated mechanisms by which the fate of memory is determined after retrieval; retrieved memory is either reconsolidated or extinguished. Using contextual fear conditioning and inhibitory avoidance tasks, we found that (1) memory reconsolidation and extinction are not independent processes; these memory phases interact one another at behavioral, anatomical and molecular levels, (2) memory reconsolidation and extinction are regulated by distinct neural circuits and (3) memory reconsolidation and extinction display distinct molecular signatures. These our findings indicate the dynamic nature of fear memory after retrieval. These observations also emphasize the importance of estimating whether the status of memory is in the reconsolidation or extinction phases when clinical treatments, such as exposure therapy combined with drugs, are performed. Therefore, it is important to further understand the mechanisms by which retrieved memory is reconsolidated or extinguished at the molecular level from a clinical perspective.
JS4-2-2-2
What machine learning can do for neurophysiology and neuropathology
○銅谷賢治1
○Kenji Doya1
沖縄科学技術大学院大学1
Neural Computation Unit, Okinawa Institute of Science and Technology1

Machine learning algorithms are now widely used in internet services and intelligent devices. The success has prompted their use in neuroscience as well. One obvious way is to use statistical machine learning algorithms for processing massive data from DNA sequencing, mass spectrometry, neural recording, and brain imaging. A typical example is the use of supervised learning algorithms for brain-machine interface. Another way of usage is model parameter fitting to experimental data, from intracellular signaling to neural networks. Fro example, Bayesian inference algorithms have been used for reconstructing network connection structures behind observed multi-neuron spike trains. And last but not least, machine learning frameworks like reinforcement learning and Bayesian inference serve as the models of adaptive computation by the brain, and allow us to explore how it can be implemented in the brain and how it can go wrong. The theory of reinforcement learning has played a major role in understanding the functions of the basal ganglia circuit and the neuromodulators like dopamine and serotonin. I will introduce our recent efforts in 1) diagnosing and detecting subtypes of psychiatric disorders by combining MRI, genotype, and other patient data; 2) constructing multi-scale model of the basal ganglia network to reproduce its functions in normal and pathological states; 3) clarifying the role of serotonin in the regulation of patience and impulsivity.
JS4-2-2-3
Optogenetic stimulation of basal forebrain parvalbumin-positive neurons entrains cortical gamma band oscillations
○Robert McCarley1, Choi J.2, Kocsis B.3, Deisseroth K.4, Strecker R. E.1, Basheer R.1, Brown R. E.1
VA Boston Healthcare System/Harvard Med. Sch., Brockton, MA1, Korea Inst. of Sci. and Technol., Seoul, Korea2, Beth Israel Deaconess Med. Center/Harvard Med. Sch., Boston, MA3, Stanford Univ., Stanford, CA4

Objective: Cortical gamma band frequency oscillations (GBO, ~ 40 Hz in EEG recordings) are important for cognitive functions such as attention, perception, and memory formation, and are known to be abnormal in neuropsychiatric conditions such as schizophrenia. However, although the cortical activation-promoting role of the Basal Forebrain (BF) is well known, a specific role of the BF parvalbumin-containing GABAergic neurons (PV) in inducing and modulating cortical GBO is unknown. We thus selectively stimulated BF PV neurons optogenetically.

Method: Cre-inducible AAV vectors with DIOChR2eYFP were injected into the BF of PV-Cre knock-in mice. To evaluate the effects of specific activation of BF PV neurons on cortical GBO, the BF was stimulated by blue laser light pulse trains of 2 to 60 Hz.

Results: BF entrainment of cortical GBO was remarkably pronounced when the BF stimulation was at and near the gamma-band frequency of 40Hz. Several other features of the response were consistent with BF entrainment of a cortical oscillator tuned to ~40 Hz including: 1) the cortical EEG power response at and near a stimulation frequency of 40 Hz gradually increased over time during the course of stimulation; 2) Second, there was a significant 40 Hz second harmonic response to 20 Hz stimulation; and 3) Third, the oscillation in response to 40 Hz stimulation persisted for at least four 40 Hz cycles (100ms) following cessation of stimulation. BF PV stimulation also increased wakefulness and decreased nonREM sleep.

Conclusion: These data imply a novel cell-specific role for BF PV neurons in controlling cortical GBO. We believe the findings represent: 1) an important new clue to the origin of cortical gamma rhythms; 2) show the wake-promoting effect of BF PV neurons; and 3) an intriguing possible link to the pathophysiology of gamma rhythms in schizophrenia. These data are congruent our VA/Harvard EEG and sMRI/MRS clinical evidence of cortical abnormalities in SZ.
JS4-2-2-4
Brain connectivity in imaging genetics
○Andreas Meyer-Lindenberg1, Heike Tost1, Edda Bilek1
Department of Psychiatry and Psychotherapy, Central Institute of Mental Health, Medical Faculty Mannheim, University of Heidelberg, Mannheim, Germany1

Objective: In the past decade, imaging genetics has evolved into a highly successful neuroimaging discipline with a variety of sophisticated research tools. To date, several neural systems mechanisms have been identified that mediate genetic risk for mental disorders linked to common candidate and genome-wide-supported variants. In particular, the examination of intermediate connectivity phenotypes has recently gained increasing popularity.

Method: Neuroimaging of connectivity (both structural and functional) combined with genotyping of common variants in healthy controls, patients with psychosis and their relatives.

Results: Both candidate polymorphisms (5-HTTLPR, MAOA) and genome-wide significant variants (ZNF804A, CACNA1C) associated with psychosis risk impact on brain connectivity, brain network topology, and functional interactions between brain regions, highlighting systems such as prefrontal-hippocampal and prefrontal-amygdala-limbic networks that may mediate genetic risk for psychosis.

Conclusion: Indices of neural network organization help identify the neural substrate of genetic susceptibility for mental illness both for the effects of candidate genes and genome-wide supported risk variants in brain structure and function.

上部に戻る 前に戻る