Neural mechanisms for learning hierarchical structures of information.

Spatial and temporal information from the environment is often hierarchically organized, so is our knowledge formed about the environment. Identifying the meaningful segments embedded in hierarchically structured information is crucial for cognitive functions, including visual, auditory, motor, memory, and language processing. Segmentation enables the grasping of the links between isolated entities, offering the basis for reasoning and thinking. Importantly, the brain learns such segmentation without external instructions. Here, we review the underlying computational mechanisms implemented at the single-cell and network levels. The network-level mechanism has an interesting similarity to machine-learning methods for graph segmentation. The brain possibly implements methods for the analysis of the hierarchical structures of the environment at multiple levels of its processing hierarchy.

Multiscale representations of community structures in attractor neural networks.

Our cognition relies on the ability of the brain to segment hierarchically structured events on multiple scales. Recent evidence suggests that the brain performs this event segmentation based on the structure of state-transition graphs behind sequential experiences. However, the underlying circuit mechanisms are poorly understood. In this paper we propose an extended attractor network model for graph-based hierarchical computation which we call the Laplacian associative memory. This model generates multiscale representations for communities (clusters) of associative links between memory items, and the scale is regulated by the heterogenous modulation of inhibitory circuits. We analytically and numerically show that these representations correspond to graph Laplacian eigenvectors, a popular method for graph segmentation and dimensionality reduction. Finally, we demonstrate that our model exhibits chunked sequential activity patterns resembling hippocampal theta sequences. Our model connects graph theory and attractor dynamics to provide a biologically plausible mechanism for abstraction in the brain.

Extended Temporal Association Memory by Modulations of Inhibitory Circuits.

Hebbian learning of excitatory synapses plays a central role in storing activity patterns in associative memory models. Interstimulus Hebbian learning associates multiple items by converting temporal correlation to spatial correlation between attractors. Growing evidence suggests the importance of inhibitory plasticity in memory processing, but the consequence of such regulation in associative memory has not been understood. Noting that Hebbian learning of inhibitory synapses yields an anti-Hebbian effect, we show that the combination of Hebbian and anti-Hebbian learning can significantly increase the span of temporal association between correlated attractors as well as the sensitivity of these states to external input. Furthermore, these effects are regulated by changing the ratio of local and global recurrent inhibition after learning weights for excitation-inhibition balance. Our results suggest a nontrivial role of plasticity and modulation of inhibitory circuits in associative memory.

Unsupervised Detection of Cell-Assembly Sequences by Similarity-Based Clustering.

Neurons which fire in a fixed temporal pattern (i.e., "cell assemblies") are hypothesized to be a fundamental unit of neural information processing. Several methods are available for the detection of cell assemblies without a time structure. However, the systematic detection of cell assemblies with time structure has been challenging, especially in large datasets, due to the lack of efficient methods for handling the time structure. Here, we show a method to detect a variety of cell-assembly activity patterns, recurring in noisy neural population activities at multiple timescales. The key innovation is the use of a computer science method to comparing strings ("edit similarity"), to group spikes into assemblies. We validated the method using artificial data and experimental data, which were previously recorded from the hippocampus of male Long-Evans rats and the prefrontal cortex of male Brown Norway/Fisher hybrid rats. From the hippocampus, we could simultaneously extract place-cell sequences occurring on different timescales during navigation and awake replay. From the prefrontal cortex, we could discover multiple spike sequences of neurons encoding different segments of a goal-directed task. Unlike conventional event-driven statistical approaches, our method detects cell assemblies without creating event-locked averages. Thus, the method offers a novel analytical tool for deciphering the neural code during arbitrary behavioral and mental processes.

Neuronal stability in medial frontal cortex sets individual variability in decision-making.

In the brain, decision making is instantiated in dedicated neural circuits. However, there is considerable individual variability in decision-making behavior, particularly under uncertainty. The origins of decision variability within these conserved neural circuits are not known. Here we demonstrate in the rat medial frontal cortex (MFC) that individual variability is a consequence of altered stability in neuronal populations. In a sensory-guided choice task, rats trained on familiar stimuli were exposed to unfamiliar stimuli, resulting in variable choice responses across individuals. We created a recurrent network model to examine the source of variability in MFC neurons, and found that the landscape of neural population trajectories explained choice variability across different unfamiliar stimuli. We experimentally confirmed model predictions showing that trial-by-trial variability in neuronal activity indexes the landscape and predicts individual variation. These results show that neural stability is a critical component of the MFC neural dynamics that underpins individual variation in decision-making.

Dendritic processing of spontaneous neuronal sequences for single-trial learning.

Spontaneous firing sequences are ubiquitous in cortical networks, but their roles in cellular and network-level computations remain unexplored. In the hippocampus, such sequences, conventionally called preplay, have been hypothesized to participate in learning and memory. Here, we present a computational model for encoding input sequence patterns into internal network states based on the propagation of preplay sequences in recurrent neuronal networks. The model instantiates two synaptic pathways in cortical neurons, one for proximal dendrite-somatic interactions to generate intrinsic preplay sequences and the other for distal dendritic processing of extrinsic signals. The core dendritic computation is the maximization of matching between patterned activities in the two compartments through nonlinear spike generation. The model performs robust single-trial learning with long-term stability and independence that are modulated by the plasticity of dendrite-targeted inhibition. Our results demonstrate that dendritic computation enables somatic spontaneous firing sequences to act as templates for rapid and stable memory formation.

Recurrent network model for learning goal-directed sequences through reverse replay.

Reverse replay of hippocampal place cells occurs frequently at rewarded locations, suggesting its contribution to goal-directed path learning. Symmetric spike-timing dependent plasticity (STDP) in CA3 likely potentiates recurrent synapses for both forward (start to goal) and reverse (goal to start) replays during sequential activation of place cells. However, how reverse replay selectively strengthens forward synaptic pathway is unclear. Here, we show computationally that firing sequences bias synaptic transmissions to the opposite direction of propagation under symmetric STDP in the co-presence of short-term synaptic depression or afterdepolarization. We demonstrate that significant biases are created in biologically realistic simulation settings, and this bias enables reverse replay to enhance goal-directed spatial memory on a W-maze. Further, we show that essentially the same mechanism works in a two-dimensional open field. Our model for the first time provides the mechanistic account for the way reverse replay contributes to hippocampal sequence learning for reward-seeking spatial navigation.

Correction to: Reproducibility and Validity of a Questionnaire Measuring Treatment Burden on Patients with Type 2 Diabetes: Diabetic Treatment Burden Questionnaire (DTBQ).

In the original publication, second author's name was incorrectly published as Hiroki Shin. The correct name should read as 'Koki Shin'.

Reproducibility and Validity of a Questionnaire Measuring Treatment Burden on Patients with Type 2 Diabetes: Diabetic Treatment Burden Questionnaire (DTBQ).

INTRODUCTION: To measure the burden of pharmacotherapy on patients with type 2 diabetes mellitus (T2DM), we developed the Diabetes Treatment Burden Questionnaire (DTBQ), a patient-administered questionnaire composed of 18 questions, and evaluated its reproducibility and validity. METHODS: We enrolled 240 patients with T2DM under pharmacotherapy over 20 years of age at seven institutes in Japan. Their physicians filled out report forms on patient backgrounds, and the patients answered both the DTBQ and the Diabetes Treatment Satisfaction Questionnaire (DTSQ). For evaluation of reproducibility, 48 of the enrolled subjects completed a 2nd DTBQ at home after leaving the medical institutes. RESULTS: Statistical analyses were performed for two sets of subjects, the validity analysis set (N = 236) and the reproducibility analysis set (N = 47). Factor analysis found a simple structure in the DTBQ item scores using a three-factor model with varimax rotation; the three subscales were designated as "implementation burden", "flexibility burden", and "blood glucose control burden". All intraclass correlation coefficients for the subscale scores were 0.8 or higher, indicating high reproducibility. Negative correlations were observed between the DTSQ satisfaction score and the DTBQ subscale scores. Moreover, as the dosing frequency of diabetic medicines increased, the DTBQ total score (total burden score) also became higher. Likewise, expected associations were observed between patient backgrounds and DTSQ scores. CONCLUSION: The DTBQ has adequate reproducibility and validity as a measurement scale for treatment burden on T2DM patients. TRIAL REGISTRATION: University Hospital Medical Information Network (UMIN) 000026382. FUNDING: Eli Lilly Japan.

Temporal relation between neural activity and neurite pruning on a numerical model and a microchannel device with micro electrode array.

Synapse elimination and neurite pruning are essential processes for the formation of neuronal circuits. These regressive events depend on neural activity and occur in the early postnatal days known as the critical period, but what makes this temporal specificity is not well understood. One possibility is that the neural activities during the developmentally regulated shift of action of GABA inhibitory transmission lead to the critical period. Moreover, it has been reported that the shifting action of the inhibitory transmission on immature neurons overlaps with synapse elimination and neurite pruning and that increased inhibitory transmission by drug treatment could induce temporal shift of the critical period. However, the relationship among these phenomena remains unclear because it is difficult to experimentally show how the developmental shift of inhibitory transmission influences neural activities and whether the activities promote synapse elimination and neurite pruning. In this study, we modeled synapse elimination in neuronal circuits using the modified Izhikevich's model with functional shifting of GABAergic transmission. The simulation results show that synaptic pruning within a specified period like the critical period is spontaneously generated as a function of the developmentally shifting inhibitory transmission and that the specific firing rate and increasing synchronization of neural circuits are seen at the initial stage of the critical period. This temporal relationship was experimentally supported by an in vitro primary culture of rat cortical neurons in a microchannel on a multi-electrode array (MEA). The firing rate decreased remarkably between the 18-25 days in vitro (DIV), and following these changes in the firing rate, the neurite density was slightly reduced. Our simulation and experimental results suggest that decreasing neural activity due to developing inhibitory synaptic transmission could induce synapse elimination and neurite pruning at particular time such as the critical period. Additionally, these findings indicate that we can estimate the maturity level of inhibitory transmission and the critical period by measuring the firing rate and the degree of synchronization in engineered neural networks.

Molecular properties of the high-affinity choline transporter CHT1.

This article summarizes molecular properties of the high-affinity choline transporter (CHT1) with reference to the historical background focusing studies performed in laboratories of the author. CHT1 is present on the presynaptic terminal of cholinergic neurons, and takes up choline which is the precursor of acetylcholine. The Na(+)-dependent uptake of choline by CHT1 is the rate-limiting step for synthesis of acetylcholine. CHT1 is the integral membrane protein with 13 transmembrane segments, belongs to the Na(+)/glucose co-transporter family (SLC5), and has 20-25% homology with members of this family. A single nucleotide polymorphism (SNP) for human CHT1 has been identified, which has a replacement from isoleucine to valine in the third transmembrane segment and shows the choline uptake activity of 50-60% as much as that of wild-type CHT1. The proportion of this SNP is high among Asians. Possible importance of choline diet for those with this SNP was discussed.

Interaction of the muscarinic acetylcholine receptor M2 subtype with G protein Gα(i/o) isotypes and Gßγ subunits as studied with the maltose-binding protein-M2-Gα(i/o) fusion proteins expressed in Escherichia coli.

We expressed the fusion proteins of the muscarinic acetylcholine receptor M2 subtype (M2 receptor) with a maltose-binding protein (MBP) and various G protein α subunits (Gα(i1-i3/o)) at its N- and C-terminals, respectively (MBP-M2-Gα(i/o)), in Escherichia coli, and examined the effect of G protein ßγ subunits (Gßγ) on the receptor-Gα interaction as assessed by agonist- and GDP-dependent [(35)S]GTPγS binding of the fusion proteins. We found that (i) Gßγ promoted both the agonist-dependent and -independent [(35)S]GTPγS binding with little effect on the guanine nucleotide-sensitive high-affinity agonist binding, (ii) the specific [(35)S]GTPγS binding activity was much greater for MBP-M2-Gα(oA) than for MBP-M2-Gα(i1-i3) in the absence of Gßγ, whereas Gßγ preferentially promoted the agonist-dependent decrease in the affinity for GDP of MBP-M2-Gα(i1-i3) rather than of MBP-M2-Gα(oA), and (iii) the proportion of agonist-dependent [(35)S]GTPγS binding was roughly 50% irrespective of species of Gα and the presence or absence of Gßγ. These results demonstrate that receptor-Gα fusion proteins expressed in E. coli could be useful for studies of receptor-G interaction.

Role of the third intracellular loop in the subtype-specific internalization and recycling of muscarinic M2 and M4 receptors.

Muscarinic M2, M4, and M2-M4 chimera receptors were transiently expressed in HEK-293 tsA201 cells, and agonist-dependent internalization of these receptors and recycling of internalized receptors were examined by measuring the amount of cell-surface receptors as [3H]N-methylscopolamine (NMS) binding activity. Coexpression of a dominant negative form of dynamin (DN-dynamin,dynamin K44A) greatly reduced the agonist-dependent internalization of M4 receptors but not of M2 receptors, as was reported by Vögler et al. (J Biol Chem 273, 12155-12160, 1998).The agonist-dependent internalization of M2/M4-i3/M2 chimera receptors (M2 receptors with the i3 loop replaced by that of M4 receptors) was greatly reduced by co-expression of DN-dynamin as was the case for M4 receptors, whereas the agonist-dependent internalization of M4/M2-i3/M4 chimera receptors was hardly affected by co-expression of DN-dynamin as was the case for M2 receptors.Internalized M2/M4-i3/M2 receptors as well as internalized M4 receptors were shown to be recycled back to the cell surface after removal of agonists, whereas no recycling was observed for M4/M2-i3/M4 receptors as well as M2 receptors. These results indicate that the i3 loops of M2 and M4 receptors take a major role in their agonist-dependent internalization and recycling.

Chemokine receptor CCR8 is required for lipopolysaccharide-triggered cytokine production in mouse peritoneal macrophages.

Chemokine (C-C motif) receptor 8 (CCR8), the chemokine receptor for chemokine (C-C motif) ligand 1 (CCL1), is expressed in T-helper type-2 lymphocytes and peritoneal macrophages (PMφ) and is involved in various pathological conditions, including peritoneal adhesions. However, the role of CCR8 in inflammatory responses is not fully elucidated. To investigate the function of CCR8 in macrophages, we compared cytokine secretion from mouse PMφ or bone marrow-derived macrophages (BMMφ) stimulated with various Toll-like receptor (TLR) ligands in CCR8 deficient (CCR8-/-) and wild-type (WT) mice. We found that CCR8-/- PMφ demonstrated attenuated secretion of tumor necrosis factor (TNF)-α, interleukin (IL)-6, and IL-10 when stimulated with lipopolysaccharide (LPS). In particular, LPS-induced IL-10 production absolutely required CCR8. CCR8-dependent cytokine secretion was characteristic of PMφ but not BMMφ. To further investigate this result, we selected the small molecule compound R243 from a library of compounds with CCR8-antagonistic effects on CCL1-induced Ca2+ flux and CCL1-driven PMφ aggregation. Similar to CCR8-/- PMφ, R243 attenuated secretion of TNF-α, IL-6, and most strikingly IL-10 from WT PMφ, but not BMMφ. CCR8-/- PMφ and R243-treated WT PMφ both showed suppressed c-jun N-terminal kinase activity and nuclear factor-κB signaling after LPS treatment when compared with WT PMφ. A c-Jun signaling pathway inhibitor also produced an inhibitory effect on LPS-induced cytokine secretion that was similar to that of CCR8 deficiency or R243 treatment. As seen in CCR8-/- mice, administration of R243 attenuated peritoneal adhesions in vivo. R243 also prevented hapten-induced colitis. These results are indicative of cross talk between signaling pathways downstream of CCR8 and TLR-4 that induces cytokine production by PMφ. Through use of CCR8-/- mice and the new CCR8 inhibitor, R243, we identified a novel macrophage innate immune response pathway that involves a chemokine receptor.

Estimation of templates and timings of spikes in extracellular voltage signals containing overlaps of the arbitrary number of spikes.

Development of methods to detect and classify neural spikes in extracellular voltage signals (e.g. commonly referred to as spike sorting) have been one of important subjects in neuroscience and neural engineering. Most of previous spike sorting methods suffer from unresolved overlaps of spike waveforms which make timings and shapes of spikes unclear. Some methods have got a handle on this problem, but they had restrictions about the type of electrodes or complexity of overlaps. In this paper, we attempted to develop a spike sorting method for the signal containing overlaps of the arbitrary number of spikes recorded with arbitrary electrodes. We estimated templates and timings of spikes by the inference based on hidden Markov model. In order to avoid the problem of too high computational cost and too much decomposition caused by assuming arbitrary overlaps, we imposed the weak probabilistic penalty on overlaps in the model and reduced computation of estimation by approximating low probabilities to zero. As the result of assessments using simulated signal and real extracellular recordings, we showed that proposed method could robustly detect and sort complexly overlapped spikes.

Efficient sequential Bayesian inference method for real-time detection and sorting of overlapped neural spikes.

Overlapping of extracellularly recorded neural spike waveforms causes the original spike waveforms to become hidden and merged, confounding the real-time detection and sorting of these spikes. Methods proposed for solving this problem include using a multi-trode or placing a restriction on the complexity of overlaps. In this paper, we propose a rapid sequential method for the robust detection and sorting of arbitrarily overlapped spikes recorded with arbitrary types of electrodes. In our method, the probabilities of possible spike trains, including those that are overlapping, are evaluated by sequential Bayesian inference based on probabilistic models of spike-train generation and extracellular voltage recording. To reduce the high computational cost inherent in an exhaustive evaluation, candidates with low probabilities are considered as impossible candidates and are abolished at each sampling time to limit the number of candidates in the next evaluation. In addition, the data from a few subsequent sampling times are considered and used to calculate the "look-ahead probability", resulting in improved calculation efficiency due to a more rapid elimination of candidates. These sufficiently reduce computational time to enable real-time calculation without impairing performance. We assessed the performance of our method using simulated neural signals and actual neural signals recorded in primary cortical neurons cultured on a multi-electrode array. Our results demonstrated that our computational method could be applied in real-time with a delay of less than 10 ms. The estimation accuracy was higher than that of a conventional spike sorting method, particularly for signals with multiple overlapping spikes.

Molecular properties of muscarinic acetylcholine receptors.

Muscarinic acetylcholine receptors, which comprise five subtypes (M1-M5 receptors), are expressed in both the CNS and PNS (particularly the target organs of parasympathetic neurons). M1-M5 receptors are integral membrane proteins with seven transmembrane segments, bind with acetylcholine (ACh) in the extracellular phase, and thereafter interact with and activate GTP-binding regulatory proteins (G proteins) in the intracellular phase: M1, M3, and M5 receptors interact with Gq-type G proteins, and M2 and M4 receptors with Gi/Go-type G proteins. Activated G proteins initiate a number of intracellular signal transduction systems. Agonist-bound muscarinic receptors are phosphorylated by G protein-coupled receptor kinases, which initiate their desensitization through uncoupling from G proteins, receptor internalization, and receptor breakdown (down regulation). Recently the crystal structures of M2 and M3 receptors were determined and are expected to contribute to the development of drugs targeted to muscarinic receptors. This paper summarizes the molecular properties of muscarinic receptors with reference to the historical background and bias to studies performed in our laboratories.

Transmembrane topology and oligomeric structure of the high-affinity choline transporter.

The high-affinity choline transporter CHT1 mediates choline uptake essential for acetylcholine synthesis in cholinergic nerve terminals. CHT1 belongs to the Na(+)/glucose cotransporter family (SLC5), which is postulated to have a common 13-transmembrane domain core; however, no direct experimental evidence for CHT1 transmembrane topology has yet been reported. We examined the transmembrane topology of human CHT1 using cysteine-scanning analysis. Single cysteine residues were introduced into the putative extra- and intracellular loops and probed for external accessibility for labeling with a membrane-impermeable, sulfhydryl-specific biotinylating reagent in intact cells expressing these mutants. The results provide experimental evidence for a topological model of a 13-transmembrane domain protein with an extracellular amino terminus and an intracellular carboxyl terminus. We also constructed a three-dimensional homology model of CHT1 based on the crystal structure of the bacterial Na(+)/galactose cotransporter, which supports our conclusion of CHT1 transmembrane topology. Furthermore, we examined whether CHT1 exists as a monomer or oligomer. Chemical cross-linking induces the formation of a higher molecular weight form of CHT1 on the cell surface in HEK293 cells. Two different epitope-tagged CHT1 proteins expressed in the same cells can be co-immunoprecipitated. Moreover, co-expression of an inactive mutant I89A with the wild type induces a dominant-negative effect on the overall choline uptake activity. These results indicate that CHT1 forms a homo-oligomer on the cell surface in cultured cells.

Identification of physiologically active substances as novel ligands for MRGPRD.

Mas-related G-protein coupled receptor member D (MRGPRD) is a G protein-coupled receptor (GPCR) which belongs to the Mas-related GPCRs expressed in the dorsal root ganglia (DRG). In this study, we investigated two novel ligands in addition to beta-alanine: (1) beta-aminoisobutyric acid, a physiologically active substance, with which possible relation to tumors has been seen together with beta-alanine; (2) diethylstilbestrol, a synthetic estrogen hormone. In addition to the novel ligands, we found that transfection of MRGPRD leads fibroblast cells to form spheroids, which would be related to oncogenicity. To understand the MRGPRD novel character, oncogenicity, a large chemical library was screened in order to obtain MRGPRD antagonists to utilize in exploring the character. The antagonist in turn inhibited the spheroid proliferation that is dependent on MRGPRD signaling as well as MRGPRD signals activated by beta-alanine. The antagonist, a small-molecule compound we found in this study, is a potential anticancer agent.