Research projects

Behavioral-driven epigenetic dynamics for regulation of vocal learning and its sensitive period

Faculty of Science, Hokkaido University
Kazuhiro Wada

For learning and memory formation, especially sensory-motor learning, a voluntarily generated active-learning processes is a critical driving factor against passive stimulus. It has been considered that the existence of molecular mechanisms for storage the functional information of learning, such as when, how much and often the animals spend the time for learning.

In this study, we test a hypothesis, “behavioral driven epigenetic dynamics as a molecular accumulator for regulation of the sensitive period of vocal learning”, by focusing on song learning in songbirds as a behavioral model for sensory-motor learning. Our studies have suggested the critical contributions of vocalization-driven and sensitive period-specific genes for song development. Many epigenetic regulation-related genes were also identified as vocalization-driven and sensitive period-specific genes. Therefore, by direct manipulation of epigenetic states in song nuclei, we will examine the effects on regulation of neural plasticity-related genes and morphological changes in the dendritic spines, and further for the contribution of vocal learning and its maintenance through the development.

pagetop

Analyses of the effect of sleep quality on memory by genetic manipulation of REM sleep

WPI-IIIS, University of Tsukuba
Yu Hayashi

Mammalian sleep is composed of rapid eye movement (REM) sleep and non-REM (NREM) sleep. NREM sleep produces cortical slow wave activity (SWA) and contributes to memory consolidation, whereas the roles of REM sleep are less clear. Using mouse genetics, we established a method to manipulate neurons that either induce or inhibit REM sleep. Genetic reduction or elongation of REM sleep affected SWA during subsequent NREM sleep, implicating REM sleep in the regulation of NREM sleep quality. These result further raise the question of whether REM sleep actually contributes to learning and memory via regulation of NREM sleep quality. In this study, we will execute behavioral experiments to test this possibility. Furthermore, we will examine the effect of REM sleep manipulation on memory decline caused by aging or neurological diseases.

fig

pagetop

Mechanisms of age-dependent memory impairment and metabolic homeostasis

Graduate School of Pharmaceutical Sciences, Chiba University
Ayako Tonoki

Aging impairs the formation of memory and metabolic homeostasis. Numerous reports have shown that the mechanisms of memory formation and the regulation of metabolic homeostasis are evolutionarily well conserved. We aim to elucidate the mechanisms of age-dependent memory impairment with alteration in metabolic homeostasis, using Drosophila olfactory memory as a model system. For this purpose, we will combine various techniques including genetics, behavior analysis, calcium imaging, and transcriptome analysis. The research plan in this project is as below.

  1. Elucidate the effect of metabolic homeostasis on memory formation.
  2. Elucidate the effect of age-dependent impairment in metabolic homeostasis on memory formation.
  3. Identify neural circuits affected by age-dependent impairment in metabolic homeostasis.
  4. Identify molecular network affected by age-dependent impairment in metabolic homeostasis.

fig

pagetop

Understanding the network function of multiple olfactory areas in the odor-mediated learning and memory

Kochi Medical School
Masahiro Yamaguchi

Olfactory system underlies fundamental animal behaviors including food search and intake and escape from predators. To behave properly in changing odorous circumstances, olfactory learning and memory is critically important. While olfactory system is a highly plastic structure including the contribution of adult-born new neurons, how functional plasticity of the olfactory network underlies olfactory learning and memory remains poorly understood.

We have recently shown that plastic change of the olfactory system occurs across multiple olfactory areas in association with wake-sleep cycles. For example, electrophysiological interaction between olfactory bulb and olfactory cortex is strengthened along olfactory learning, and top-down inputs from the olfactory cortex to the olfactory bulb are enhanced during post-task sleep period. In addition, selection between survival and death of adult-born olfactory bulb neurons is determined by the interplay of bottom-up sensory inputs from periphery and top-down inputs from the olfactory cortex. Further, olfactory learning that elicits strong emotion and motivation activates particular domain in the olfactory tubercle, which is a part of the olfactory cortex and also a part of the ventral striatum.

Given these findings, this project aims to understand the network plasticity across multiple olfactory areas in the odor-mediated learning and memory. In particular, we focus on the contribution of wake-sleep cycles, adult-born neurons in the olfactory bulb, and the motivation-related functional domain of the olfactory tubercle.

 

fig

pagetop

Neural circuit mechanisms integrating observation and real experience

Graduate School of Pharmaceutical Sciences, Hokkaido University
Hiroshi Nomura

Integration of our own experience and observation is important for learning and memory. Watching aggressive adults makes children aggressive. If a chimpanzee see a use of a drinking straw, the chimpanzee quickly learns to drink through a straw. Children often observe their parents’ behavior, and then they learn from their own experience. Prior observation affects a future memory formation of the own experience. There are some behavioral studies about observational learning, but neural circuits for observational learning is poorly understood. Especially, neural circuits integrating observation and real experience is unclear. We have studied neuronal ensembles encoding real experience. Based on the knowledge, we hypothesized that information about observation and real experience converges on a neuronal ensemble and that this convergence reinforces memory formation. In this study, we will test this hypothesis. We use in vivo calcium imaging from freely-behaving mice to examine detailed patterns of neuronal activity. Moreover, we employ optogenetics to manipulate the neurons activated at a specific time point. Through a series of experiments, we examine neural circuit mechanisms integrating observation and real experience.

fig

pagetop

Deciphering long-term memory mechanisms regulated downstream of CREB-Arc signaling

Graduate School of Medicine, The University of Tokyo
Haruhiko Bito

  Synaptic transmission triggers generation of EPSP as well as chemical signaling such as increase in intracellular Ca2+ concentration. Ca2+ signals play critical roles both for induction of long-term synaptic plasticity and for the signaling from the synapses to the nucleus. We previously showed an essential role of CaMKK-CaMKIV phosphorylation cascade in triggering the latter. Furthermore, we showed that this cascade tightly controlled the spatiotemporal dynamics of CREB phosphorylation, a key mechanism underlying cellular consolidation during long-term memory. Here, we will investigate the physiological significance of activity-induced Arc gene expression, downstream of CREB activation. In particular, we will elucidate in in vivo brains the information flow underlying: postsynaptic NMDA receptor activation => Ca2+ influx => CaMK activity => CREB phosphorylation => Arc induction => postsynaptic Arc targeting. We will further perform live imaging of active engrams using promoters of activity-dependent genes. Results of these studies are expected to shed light on spatiotemporal characteristics of widespread circuit dynamics activated by CREB-Arc signaling during long-term memory formation and maintenance.

fig

pagetop

Development of analysis method for multi-resolution Calcium imaging data

School of Computing, Tokyo Institute of Technology
Toru Aonishi

Calcium imaging is one of important methods for measuring multi-cellular activities in vivo and vitro. The rapid progress of imaging systems such as two-photon excitation microscopy has made it possible for us to simultaneously record activates of more than a thousand cells with high spatial-temporal resolution for a long time. However, the extreme increase in amount of multi-cellular activity data makes us face with a difficulty of data analysis. Some research groups try to develop machine learning methods to automatically detect cells from large-scale Calcium imaging data. In this proposal, we aim to develop a novel automatic method of cell detection using multi-resolution techniques recently focused on due to the efficacy of Deep learning methods. We extend non-negative matrix factorization to handling multi-resolution imaging data, and improve its cell detection performance by this extension.  Furthermore, we try to establish a super-resolution method to extract both high-spatial resolution information (i.e. morphology) and high-temporal resolution (i.e. activity) from two different data, high-frame-rate, low-spatial-resolution image and low-frame-rate, high-spatial-resolution image. By realizing this method, we address a dilemma of imaging system that we cannot improve both spatial and temporal resolutions simultaneously because of the limitation of its scanning speed.

 

fig

pagetop

The functional role of dentate gyrus maturity in infant amnesia

Life Science Research Center, University of Toyama
Keizo Takao

Processing and fixation of memory are essential for the formation of long-term memory, which is one of the most important functions of the brain that allows humans to socialize. Human infants can remember their experiences over short-term, however, by the time they become adults, they often loose their memory for events that occurred before they were approximately three years old. This phenomenon is called infant amnesia and is observed in a wide variety of species, including humans and rodents. We screened more than 160 mutant mouse strains using a large-scale comprehensive behavioral test battery and identified several strains with behavioral traits related to psychiatric disorders. The molecular and electrophysiological features of dentate gyrus (DG) neurons in the adult hippocampus of these mouse strains were similar to those of immature DG neurons in normal mice, a phenomenon termed the immature DG (iDG).

Mice possessing the iDG phenotype commonly display impairments in remote memory. Therefore, iDG might play an important role in remote memory impairment. Homeostatic synaptic plasticity is considered to be a mechanism by which neurons maintain function in the face of perturbations. The DG might become immature through homeostatic synaptic plasticity mechanisms induced by the application of electroconvulsive shock. Our study is based on the hypothesis that immaturity of DG causes infant amnesia. We will manipulate the maturity of the DG bidirectionally by changing activity of DG neurons in mice using optogenetics and other techniques, and will elucidate the functional role of maturity of the DG in infant amnesia.

fig

pagetop

The mechanism of positive memory erasing by hypothalamic neurons

RIEM, Nagoya University
Akihiro Yamanaka

We can retain some memory forever, but some memory is disappeared. It is still unknown how this difference arises. Here we found the evidence that the brain positively erasing memory. We focused on melanin concentrating hormone (MCH) producing neurons in the hypothalamus. Amino acid sequence of MCH is conserved among wide range of species suggests MCH is involved in important physiological role. Initially, MCH was thought to be a positive feeding regulator. However, recent studies showed that MCH is involved in sleep/wakefulness regulation. We previously showed that activation of MCH neurons using optogenetics increased time in REM sleep. And MCH neurons ablated mice showed good memory compare with control mice. The aim of this study is to reveal the molecular and cellular mechanism of memory erasing by MCH neurons by using optpgenetics and pharmacogenetics and behavior pharmacology.

 

fig

pagetop

Revealing regulatory mechanism of learning and memory through the analysis of C. elegans thermotaxis as a behavioral paradigm

Graduate School of Science, Nagoya University
Ikue Mori

C.elegans remember cultivation temperature and migrate to that temperature after placed on a temperature gradient without food for an hour. In the neural circuit essential for this behavior (thermotaxis), AFD sensory neuron perceives and memorizes the temperature, and its downstream interneurons AIY, AIZ and RIA are needed to associate the memorized temperature with food availability (Figure 1). After being placed on a temperature gradient without food for several hours, the animals that had been staying at around the cultivation temperature started to disperse to explore the temperature gradient (Figure 2). We use this switch between exploitation/exploration behavioral strategies as a simple model of decision making, and attempt to understand molecular, cellular and neural circuit bases underlying learning, memory and decision making (Figure 3).

fig

pagetop

Study of novel neurons related to the prediction error, which contribute to memory consolidation and modification

MIC, Kyoto University
Yukinori Hirano

Animals learn and form memories when the predicted outcome is different from the reality, which is referred to as a prediction error. Memories are further consolidated to long-term memory (LTM), which are mediated by de novo gene expression. How the prediction errors are generated, and affect memory consolidation, maintenance and extinction are the one of the important issues, which ensure consistency between the internal predictions and the reality. We found that, in Drosophila, there are neurons showing activities correlated with the prediction errors, and those neurons are required for, not learning itself, but memory consolidation to LTM by activating gene expression. In this study, we will focus on the issues, 1: the mechanism regulating the activities of those neurons, 2: the mechanism of gene expression mediated by those neurons, and 3: the identification of neurons related to the prediction errors when the memory is extinguished. This work could help understanding of, not only the mechanism of memory consolidation, but also the role of memory in animals’ behavior, which fills the gaps between the internal and external world.

fig

pagetop

Molecular mechanism of local circuit formation for generating memory

FBS, Osaka University
Takeshi Yagi

Memory is dynamically generated by neural activities of cell assembly in the brain. The cell assembly is thought to be based on structures of complex neural circuits. It has not yet revealed mechanisms for building the structures of complex neural networks for generating memory. We have identified diverse cell adhesion molecules, clustered protocadherins (cPcdh), from the mammalian brain. They display differential and combinatorial expression in each individual neurons, suggesting that they are possible to regulate complex neural circuits at individual cell level. So far we have revealed that the cPcdhs contribute to memory function and regulate axonal and/or dendrite pattern formation. In this project, we will analyze molecular mechanism of cPcdhs for generating memory. This approach will reveal molecular mechanisms to generate structures of complex neural networks and will realize how to generate memory by neural activity of cell assembly .

fig

Manipulation of memory formation

pagetop

Understanding of the operating principle of the nervous system underlying mating preferences for familiar individuals

Graduate School of Natural Science and Technology, Okayama University
Hideaki Takeuchi

Within group-living animals, individuals appropriately tailor their attitudes and responses to other group members according to the social context and external environment. The territory between sensory input and behavioral output, which comprises the integrative circuits underlying decision-making processes, however, is vast and mysterious. To address this issue, we have focused on medaka fish, a model animal used mainly in the field of molecular genetics. Recent studies in the filed of behavioral ecology demonstrated the ability to recognize conspecific individuals in some fish species such as guppies and cichlids. Previously we also reported that medaka females recognize familiar males following prior visual exposure, and social familiarity influences female mating receptivity. Medaka females exhibit a positive response (high receptivity) to familiar males, and a negative response (low receptivity) to unfamiliar males. In addition, we revealed the essential role of a subpopulation of gonadotropin-releasing hormone-producing neurons (GnRH3 neurons) in switching from low to high female receptivity. Our aim here is to describe the neural network of the integrative circuits associated with individual recognition and the regulation of female receptivity.

fig

pagetop

A novel principal regulatory mechanism of synaptic plasticity by AMPA-type glutamate receptor glycosylation

Dept. Neuroscience, University of Miyazaki Faculty of Medicine
Kogo Takamiya

Glutamate is a main excitatory neurotransmitter in central nervous system and its receptor, glutamate receptors, play important roles in many neuronal functions including memory. In particular, AMPA-type glutamate receptor (AMPA-R) plays central roles in not only main excitatory neurotransmission, but also expression of synaptic plasticity. To date many interacting proteins with AMPA-R cytoplasmic domain have been reported on their many functions in the regulation of synaptic plasticity. However the regulation of AMPA-R function in extracellular side still remains unclear.

In this project, we will focus on N-glycosylation of AMPA-R, and we aim to elucidate their functions in synaptic transmission and plasticity. We already found that non-glycosylated AMPA-R is preferentially localized in lipid raft on the cell membrane and this localization has altered channel property, missing desensitization. We also revealed that there are raft-positive and  -negative synapses in the same neuron, proposing the hypothesis that glycosylation of AMPA—R decides the synaptic localization, resulting in the regulation of synaptic strength in synaptic plasticity. We will verify this hypothesis and study the involvement in memory and learning.   

fig

pagetop

Development of manipulation technology for synapses in pre-learning.

Dept. Physiol. Sch. Med. Yokohama City University
Kiwamu Takemoto

To elucidate the mechanism of learning and memory formation in vivo, manipulation of molecular function is an key technology together with in vivo imaging. In our research, we will develop the synaptic manipulation technology to examine the effect of default synapse networks in pre-learning for efficiency of learning. To address this issue, We will develop the light induced-loss of function technology for neurotransmitter receptor which is expressed in a silent synapse. We developed the effective CALI (chromophore assisted light inactivation) method by eosin (Takemoto et al. ACS.Chem.Biol. 2011) and succeeded to explore the CALI method of AMPA receptor, GluA1, which get delivered in synapse in response to learning and experience in vivo. We hope to utilize this strategy for neurotransmitter receptor which is expressed in a silent synapse. Our method will be important to identify the default synapse network and structure for effective learning in vivo.

fig

pagetop

Novel molecular mechanism regulating age-dependent dynamism of motor memory

Keio University School of Medicine
Wataru Kakegawa

Molecular mechanisms which underlie learning & memory vary throughout life. Such age-dependent memory dynamism has so far been intensively studied using excellent model organisms including Drosophila and C. elegans, and many kinds of genetically engineered mice. However, its mechanisms are poorly understood because neuronal circuits for learning & memory are complicated especially for vertebrate.

D-Serine (D-Ser), a kind of D-amino acid, is well-known as a co-agonist for NMDA-type ionotropic glutamate receptors (NMDA receptors), which regulate synaptic plasticity and a certain form of learning & memory in the forebrain. In contrast, we recently demonstrated that D-Ser in the cerebellum specifically binds to delta2-type ionotropic glutamate receptors (GluD2 receptors) to promote synaptic plasticity and cerebellum-dependent motor learning (D-Ser-GluD2 signaling; Kakegawa et al., Nat Neurosci, '11). Interestingly, D-Ser in the cerebellum exists abundantly but transiently during postnatal development, which suggests that these unique and dramatic changes of D-Ser amount in the cerebellum may support the age-dependent memory dynamism. In this study, we aim to clarify the detailed molecular mechanisms for the novel D-Ser signaling.

fig

pagetop

Brain-wide mechanisms for visual object memory retrieval

Graduate School of Medicine, Juntendo University
Masaki Takeda

The question of how we recall objects from memory has always attracted the interest of philosophers and is now one of the major research problems for neuroscientists who investigate the circuits of the brain. Using monkeys performing memory tasks, previous electrophysiological studies have found that neurons in the inferior temporal cortex have a fundamental role in the memory retrieval of visual objects. However, the mechanism by which these memory-related neurons interact with each other and functionally connect with other brain regions is still under debate. In addition, it remains to be elucidated whether the interaction of these neurons has a causal role for memory retrieval.

In this study, we aim to identify the brain-wide neuronal circuit involved in memory retrieval by recording neuronal activity simultaneously from multiple memory-related regions, including the inferior temporal cortex. We will investigate the brain-wide memory mechanism at six-laminar resolution using multi-contact linear array electrodes. We will furthermore try to unravel the causal role of the identified neuronal circuit for memory retrieval using optogenetic approaches and behavior analyses.

fig

pagetop

Design and synthesis of new chemical tools to manipulate neuronal signaling pathways

Faculty of Science, Toho University
Toshiaki Furuta

Caged compounds are synthetic molecules whose biological activities are temporally masked by covalent attachment of an appropriately designed photo-removable protecting group. We have designed and synthesized caged compounds of biologically related molecules including neurotransmitters, second messengers, mRNAs and DNAs, which can be photo-chemically reactivated under one and two photon excitation conditions. The compounds have been used to control gene expression, signal transduction and neuronal signaling. Examples of application include local stimulation of an intracellular signaling pathway within subcellular spatial resolution.

In this project, we will develop new chemical tools designed to manipulate a neuronal circuit involved in memory and learning. (1) We have designed new caged second messengers which enable cell type specific photo-manipulation of an intracellular signaling pathway. (2) We have developed a facile and practical method for preparation of caged oligonucleotides and applied the method to photo-mediated gene expression in cultured mammalian cells. We will elucidate whether the newly developed caged oligonucleotides can manipulate photo-chemically local protein synthesis within dendrites.

fig

pagetop

Analysis of cell assembly activity realizing formation and reformation of various contents of memory

Graduate School of Brain Science, Doshisha University
Yoshio Sakurai

Cognitive psychology has suggested that memory items are remembered and consolidated as networks of information items. It has also indicated that changes and reconstruction of such information networks make it possible to remember almost unlimited amount of memory in spite of the limited amount of neuronal networks in the brain. However, it is still unknown what neuronal networks actually realize the dynamic changes of information networks and how they work in the brain. How are functional groups of neurons, i.e., cell assemblies, which code information items constructed during formation of memory? How do such constructed cell assemblies change when new memory is formed and/or old memory is erased or reformed? The present study aims to clarify the activities and changes of cell assemblies underlying dynamics of memory formation and reformation. We record spike activities from closely neighboring multiple neurons with special electrodes during all processes when the same rat is acquiring multiple memory tasks (conditioned discrimination tasks) (Figure 1). We then separate the recorded multi-neuronal activities to activities from individual neurons by our spike-sorting method using independent component analysis (ICA) (Figure 2). We will finally show changes of firing rates of individual neurons and firing synchrony in large and small groups of neurons during the processes of memory formation and reformation. To-be-answered questions are 1) How do cell assemblies change during formation of memory A → memory B → memory C → memory A? 2) How different are such changes of cell assemblies in prefrontal cortex, motor cortex, and hippocampus? 3) How do macro circuits across the prefrontal cortex, motor cortex and hippocampus change during the dynamic formation of different memory?

fig

pagetop

Functional change of cerebellar neuronal circuit underlying the formation and transfer of memory

Graduate School of Brain Science, Doshisha University
Shin-ya Kawaguchi

Critical roles of synaptic plasticity in learning and memory in animals have been clarified. However, many issues regarding how synaptic change brings about learning and memory remain unclear, as follows: the impact of synaptic plasticity on each neuronal computation, the assembly of synaptic plasticity for memory encoding, and the mechanism for sustained maintenance of memory.

The cerebellum contributes to fine motor control, and the motor learning is thought to be mediated mainly by long-term depression at parallel fiber synapses on a Purkinje cell in the cerebellar cortex. It was also suggested that the motor learning memory formed in the cortex is consolidated by its transfer to the cerebellar nuclei region. In this study, focusing on motor learning memory in the cerebellum, we attempt to clarify the whole memory dynamics underlying the memory formation and stabilization initiated by the plasticity of elementary synapses. For this aim, we develop two genetically-encoded fluorescent probes: One is the visualization of synapses which have undergone long-term depression, and the other is a fluorescent indicator of membrane potential. These new techniques together should unveil adaptive information processing in the cerebellar neuronal circuit at a high spatio-temporal resolution, demonstrating the essential process of learning and memory in animals.

fig

Using fluorescent probes for long-term depression (LTD) and membrane potential (Vm),
the neuronal circuit mechanism for establishment and stabilization of motor learning is studied.

pagetop

Significance of synaptic instability (dynamism) as studied by newly developed optogenetic tools.

Supportive center for brain research, National Institute for Physiological Sciences
Hideji Murakoshi

Synapse is a structure to connect cell to cell for communication, and people believe that this is the basic unit of memory. Synapse is often compared with magnetic material used in computer hard disk for the digital information storage. However, since synapses consist of biomolecules such as protein and lipid, the state of synapses are modulated and instabilized temporally. Why did brain choose such an unstable structure as memory unit? In this project, we define the dynamic behavior and instability of synapses as "synaptic dynamism", and we will seek the biological significance of response variability of synapse and its relationship with memory and learning by using newly developed optogenetic tools. Recently, the development and popularization of channel rhodopsin led to the accumulation of the tremendous knowledge about neuronal networks. However, optogenetic manipulation of signaling molecules at the level of single-synapse has not been established. Therefore, we will develop optogenetic tools with high spatiotemporal resolution for the manipulation of synaptic ensemble by utilizing LOV2 domain from Phototropin1 and apply them to study memory and learning in vivo.

fig

pagetop

Neural mechanisms of context-specific reward-related learning

GraduateSchool of Medicine, Kyoto University
Masaaki Ogawa

In order to live better, animals, including human, must make a search for important reward-related information to be learned, and subsequently use that information to control one’s behavior, while maintaining motivation. The orbitofrontal cortex (OFC), a prefrontal cortex region, could be a higher control center for this kind of reward-related learning and memory, but the precise role of the OFC in such “memory dynamism” is not clear. Rodents and primates’ orbitofrontal cortex are similar in terms of the pattern of connectivity with other brain regions such as amygdala and striatum. Recent evidence has suggested that fundamental functions of OFC to regulate reward-related behavior could be common between the species. As such, OFC in rodents is a good model system to address a fundamental neural mechanism of reward-related “memory dynamism”.

To address a fundamental role of neural circuits involving OFC as a higher control center for context-specific reward-related learning, we will manipulate activity of rodents’ OFC and its neural circuits in a context-specific manner in reward-related learning. We will use cutting-edge optogenetics combined with in vivo electrophysiology. Further, we will develop a technology to address molecular mechanisms of such context specific roles of OFC.

fig

pagetop

Neural circuit model of dynamic interactions between the hippocampus and cortical working memory

RIKEN, Brain Science Institute
Tomoki Fukai

When the brain senses, analyzes, memorizes and responses to sensory stimuli, various processes emergent in different cortical and subcortical areas need to be adequately harmonized to produce a consistent picture of the external world and generate an adequate behavioral output. To uncover the mechanisms of cross-area communications is therefore essential for understanding the underlying mechanisms of information processing in the whole brain. It was recently shown in a memory-based spatial navigation task that projections from the entorhinal cortex (EC) to the hippocampal CA1 are crucial for the successful readout of task-related information from cortical working memory, and that theta (3-7 Hz) and fast gamma (> 60 Hz) oscillations play active roles in this communication (Yamamoto et al., Cell 2014). To uncover the underlying circuit mechanisms, we model the local circuit of EC and CA1, and simulate mutual interactions between them through multiple synaptic pathways (Fig. 1).     

We will utilize the modeled network to clarify the role of theta/gamma oscillations in coordinating the transfers of task-related information between the hippocampus and EC, and to elucidate the network mechanism that dynamically controls information flow between the two local network systems. Based on the modeling study, we will provide testable predictions for experiment and accumulate useful knowledge for future modeling of large-scale brain-like artificial intelligence systems.

fig

pagetop

The spatial memory of self and other

BSI, RIKEN
Teruko Danjo

When animals explore an open field, how do they recognize their place and other various positions in the field? Previous studies have shown that such spatial memory is founded on the hippocampus; the place cells in the hippocampus, activated specifically when the self is located in its place field, are known to play an essential role in spatial memory. The assemblies of the place cells are regarded to establish the ‘cognitive map’, which is pivotal to animals’ efficient navigation in the environment.

Spatial navigation is supposed to be based on two distinctive mechanisms: the ‘self-referenced’ and the ‘map-based’ navigation. In the ‘self-referenced’ navigation, an animal identifies its position by computing the speed, time duration, and direction of its motion. On the other hand, in the ‘map-based’ navigation, one’s position is identified by referring cognitive maps in the brain, same as we usually find a place by referring a map. This way of navigation requires neurons representing place as itself, regardless of one’s own location. In other words, there assumed to be neurons that response to the place specifically when animals draw attentions to it. However, neurons with such a feature have never been reported yet.

     The main purpose of this study is to identify the mechanisms of how animals represent the spatial information of other animals in the brain. To this end, we record electrophysiological activities from the hippocampus while the rat is required to observe the other rat, and then isolate neuronal activities that contribute to represent the position of the other. Through quantified analysis using the information theory, we compare the spatial representation of ‘the self’ with ‘the other’, and further examine whether there is a common feature for the spatial representation that can be applied both to the position of the self and other.

fig

pagetop

Noradrenaline circuit mechanisms for regulation of emotional learning

RIKEN Brain Science Institute
Joshua Johansen

Emotional learning and memory is critical for survival, but we also need mechanisms to reduce emotional responding when it is not appropriate to facilitate behavioral and cognitive flexibility. The brainstem locus coeruleus (LC)-noradrenaline neuromodulatory system sends widespread projections throughout the brain and is important for both emotional learning and behavioral flexibility. In this grant we will explore how the anatomical organization of the LC-noradrenaline system mediates its role in distinct forms of learning and behavior. Specifically, we will examine whether different populations of LC-noradrenaline neurons projecting to specific brain regions participate in either emotional learning or overriding emotional memories when they are no longer appropriate. To examine these questions we will utilize cutting edge viral tracing and manipulation tools combined with in-vivo neuronal readout technologies to define how noradrenaline cell populations are connected, encode information (see example below, Fig. 1) and functionally participate in different aspects of behavioral learning and memory. Together these studies will reveal novel circuit mechanisms for neuromodulatory control of learning and behavior and potentially lead to new treatment avenues for psychiatric disorders associated with exaggerated and persistent emotional memories.

図

Figure 1: Optogenetic identification of locus coeruleus (LC)-noradrenaline neurons for in-vivo electrophysiological recordings during learning. (a) Cre dependent opsin expression (green) in tyrosine hydroxylase (TH, immunostained in red) noradrenergic LC neurons (merged, in yellow) in a TH-cre rat. Schematic shows experimental strategy for optogenetic identification of noradrenaline neurons to determine their neural coding properties during learning and memory. (b) Example of an optogenetically identified LC-noradrenaline neuron responding to a fear-inducing auditory cue following fear learning.

pagetop

Long-term observation of cell assembly in hippocampus during memory formation, consolidation and recall.

RIKEN Brain Science Institute
Yasunori Hayashi

Hippocampus is well accepted as the central player in memory formation from both clinical cases and experimental animals with hippocampal lesion while it is required for retrieval of memory. However, recent optogenetic experiments cast doubt on this view and claim that hippocampus is indeed required for memory retrieval as well. In order to settle the discrepancy among different studies, we decided to establish a system to directly visualize the activity of -1000 neurons in pyramidal cells in dorsal hippocampal CA1 regions chronically for months by using virtual reality, Ca2+-imaging, and two-photon microscopy. This will allow us to visualize true dynamics of hippocampal cell assembly. Using this system, we will address the basic questions of cell assembly which was not possible before, such as; how cell assembly is formed during memory formation, whether the same cell assembly is reactivated by recent and remote memory, how cell assembly is reorganized as the memory itself is reorganized or extinguished, whether re-establishment of the same memory uses cell assembly or not, and finally, how reward system reinforces memory. Through these experiments, we will address memory dynamisms at a cellular precision, which had never been attained before.

fig

Figure left: Virtual reality system used in this study.
Figure right: Example of virtual place cells detected in dorsal hippocampal CA1 region.

pagetop

Investigation of a functional interaction between multiple brain sites to regulate working memory in macaque monkeys by using a gene-manipulation technique

BSI, RIKEN
Takayasu Higo

It has been suggested that the higher brain functions in primates including human are regulated by a highly developed cerebral cortex. A large number of researches using primates in particular macaque monkeys have actively investigated to clarify the functions of each area and anatomical connections between areas. However, it has been considered that a specific function is rarely generated by information processing in one site of the brain, and is mainly regulated by functional interaction across multiple sites.  For example, “working memory” is the cognitive function of focusing on specific (memory) information in the brain, while ignoring other irrelevant information, and it has been proposed top-down control in which prefrontal cortex positively and negatively regulates neural activities in other brain areas such as temporal and parietal cortices in response to a necessity of the information. However, the underlying circuit mechanisms of the higher functions remain elusive, because of the difficulties to generate a method of selectively manipulating the interaction between brain sites.

 In this research project, we mainly focus on a functional interaction between prefrontal cortex and inferior temporal cortex to regulate working memory, and in order to identify circuits and neural activities, we will analyze monkeys undergoing surgeries to manipulate a specific gene expression by behavior and electrophysiological studies.

 

fig

pagetop

Memory consolidation and regulation of neuronal exitability by REM sleep

Faculty of Medicine, Osaka City University
Kenji Mizuseki

 Sleep is composed of an alternating sequence of rapid eye movement (REM) sleep and non-REM (NREM) sleep. REM sleep is suggested to be important for memory consolidation, but the mechanism is not known. Recently we found that REM sleep reduces firing rate of pyramidal neurons and increases population synchrony in the hippocampal CA1 region (Figure; Grosmark, et al., Neuron, 2012). However, it is not known if REM sleep has the same role in other brain regions. Further, it is not clear if the reduction of firing rates and increase of synchrony are related to memory consolitation.

 In this grant, usigng silicone probe and optogenetics techniques in behaving and sleeping rodents, we will address the following questions. ①We will examine how REM sleep controls neuronal correlation and representation of information in the hippocampus. ②We will examine how REM sleep regulates neuronal activity and population synchrony in prefrontal cortex, amygdala, basal ganglia and cerebellum. ③By optogenetically manipulating REM sleep, we will eculidate roles of REM sleep in neuronal plasticity and memory consolidation.

 

fig

pagetop

PAGE TOP