Biology subjects

Taira, M.

Publications and source records attributed to Taira, M..

12 recordsLinked to original sources

bioRxiv2026

Transition from model-free to structure-informed decision making in dynamic environments

Reinforcement learning theory formulates distinct decision-making strategies, including reactive model-free and deliberative model-based strategies. This study investigates how mice adjust their reinforcement learning strategies while learning decision-making in dynamic environments. Unlike previous studies that focused on behaviors after extensive training periods, we analyzed changes in learning strategies in the course of training of a two-step decision-making task with probabilistic state transition and fluctuating reward probabilities. Our statistical behavioral analysis showed that the stay-probability following common and rare transitions diverged with training, a signature of strategies that utilize knowledge of task structure. We fit various reinforcement learning strategies to behavioral data and found that structure-informed strategies became increasingly dominant in their behaviors during training. Whereas previous studies emphasized transition from goal-directed to habitual strategies after extensive training, which were often associated with model-based and model-free strategies, respectively, our results newly demonstrate a shift from model-free to structure-informed strategies in early training in mice. Author summaryReinforcement learning theory allows us to examine how we make decisions and what approaches we use to optimize rewards. Most previous research, however, has examined animal behavior only after extensive training. Here we analyzed how mice adjust their reinforcement learning strategies as they are trained in a two-step decision-making task. Initially, mice relied on reactive model-free strategies, but as training progressed, their behavior began to incorporate knowledge of task structure. While previous studies suggested transition from model-based to model-free strategies with extensive training, our study revealed the opposite in the early stage of training.

animal behavior and cognition↗

bioRxiv2026

Neural mechanisms of self-conformity

We conform not only to the opinions and behaviors of others but also to those of ourselves, suggesting that the motivation to appear consistent leads us to treat ourselves like others. However, the neural mechanisms of self-conformity remain unclear. We used a sequential facial attractiveness-rating task with ostensible reminders of participants own previous ratings to investigate self-conformity and compare it against group conformity. Rating data showed that participants on average conformed to both themselves and the group. While some individuals conformed excessively (over-conformity), others went against their previous behavior (anti-conformity), similarly in self and group conditions. Neurally, dorsomedial and ventrolateral prefrontal cortex activity correlated with both self- and group conformity. Moreover, self-versus group conformity engaged more medial versus lateral prefrontal regions, and over-versus anti-conformity more ventral versus dorsal regions. These results suggest that self- and group conformity rely on both common and dedicated neural mechanisms.

neuroscience↗

bioRxiv2026

Continuous Strategy Adaptation and Discrete Switching Driven by Environment and Internal State in Meta-Learning

AO_SCPLOWBSTRACTC_SCPLOWBehavioral strategies can change in response to environmental and internal states, either gradually or abruptly, enabling flexible adaptation. Such strategy regulation is central to meta-learning, the ability to learn to learn. Previous studies analyzed temporal or condition-dependent strategy change using models and theories that assume continuous or discrete changes. Here, we analyze the mices behavior in a two-step decision task using four different approaches: stay-switch choice probability analysis; generalized linear mixed model (GLMM) of choice and reaction time (RT) given preceding task events; fitting a reinforcement learning (RL) model with time-varying meta-parameter by a novel multiple-step particle filtering method; and fitting a finite internal state (FIS) model that produces choice and RT depending on discrete state transition. Together, the stay probability and GLMM analyses reveal that learning progress encourages a shift toward a model-based, value-based learning strategy, accompanied by elevated choice perseveration. More uncertain reward settings or changes in them lead to random, exploratory behavior. Meta-parameter dynamics show faster learning, greater involvement of a model-based strategy, higher choice stochasticity, and more rapid development of choice perseveration with less contribution to the final decision as learning progresses. Exploratory behavior in the face of uncertain reward settings or changes in those settings is underpinned by slower forgetting and greater model-based contribution. FIS modeling discovered a trial-level switch between an optimal value-based learning state and a suboptimal self-repeating state. Meta-parameter dynamics reflect continuous strategy changes, while state transitions capture abrupt, discrete strategy switches. At an intermediate timescale, when reward settings change, two processes interact: mice persist in a self-repeating state, leading to attempts at model-based strategy with incomplete adaptation.

animal behavior and cognition↗

bioRxiv2025

Fictive Learning in Model-based Reinforcement Learning by Generalized Reward Prediction Errors

Reinforcement learning (RL) is a normative computational framework to account for reward-based learning. However, widely used RL models, including Q-learning and its variants, fail to capture some key behavioral phenomena observed in animal experiments, particularly the dynamic switch between model-free and model-based control in the two-step task. A fundamental discrepancy is that learning is restricted to experienced outcomes in the models, whereas biological agents may generalize learning to unvisited options based on internal world models, so-called fictive learning. We propose a simple, brain-inspired fictive learning rule and conduct the rodent two-step task to examine whether fictive learning could explain the observed behavior. The learning rule uses a generalized reward prediction error to update both experienced and non-encountered states and actions. The factual prediction error is scaled by the event correlation inferred from the internal model for fictive update. The generalized reward prediction error might be supported by brain-wide dopaminergic broadcasting. Through simulations, we show that this model reproduces key behavioral traits in the two-step task, including stay probabilities and regression analysis, which common RL models fail to explain. Model fitting validates its superior fit over existing alternatives. Furthermore, the model replicates dopaminergic dynamics observed in the same task. This framework bridges normative RL theory and biological learning, offering new insights into adaptive behavior.

animal behavior and cognition↗

bioRxiv2025

Identification of N-acetyl-rich heparan sulfate binding motifs and their role in expanding BMP distribution and signaling by Cerberus

Secreted signaling molecules called morphogens including BMP and Wnt provide positional information by their concentration gradient. Although binding to an extracellular matrix, heparan sulfate (HS), is essential for morphogen gradient formation, it is still unclear how different ranges of gradients can be achieved. We previously found that two types of HS, N-acetyl-rich HS (NAc-HS) and N-sulfo-rich HS (NS-HS), are differentially clustered on the cell membrane, and bind to Frzb and Wnt8, respectively. Furthermore, we demonstrated that Frzb expands Wnt8 distribution by transferring Wnt8 from NS-HS to NAc-HS. This indicates that NAc-HS or its unsulfated form, heparan, which has been considered as a precursor of HS including N-sulfo HS, has an essential role in morphogen distribution. Here, we identified a so-called BMP antagonist, Cerberus (Cer), as another NAc-HS-binding protein. Binding of Cer to NAc-HS was enhanced by knockdown and repressed by overexpression of ndst1, which catalyzes the conversion of NAc-HS to NS-HS, suggesting that Cer binds to NAc-HS and/or heparan as does Frzb. We then explored NAc-HS binding motifs using deletion or alanine-substitution mutants of Frzb and Cer, and revealed that clusters of basic amino acids, Lys and Arg, are critical for their binding to NAc-HS. Functionally, Cer was able to expand distribution and signaling ranges of BMP4 through relocation of BMP4 from NS-HS to NAc-HS as has been shown with Wnt8 and Frzb. For this Cer activity, its NAc-HS binding motifs were essential. These results suggest that NAc-HS clusters and NAc-HS binding motifs constitute an interacting unit to modulate extracellular distribution of BMP4 as well as Wnt8.

developmental biology↗

bioRxiv2024

The differential effect of optogenetic serotonergic manipulation on sustained motor actions and stationary waiting for future rewards in mice

Serotonin is an essential neuromodulator that affects behavioral and cognitive functions. Previous studies have shown that activation of serotonergic neurons in the dorsal raphe nucleus (DRN) promotes patience to wait for future rewards. However, it is still unclear whether serotonergic neurons also regulate persistence to act for future rewards. Here we used optogenetic activation and inhibition of DRN serotonergic neurons to examine their effects on sustained motor actions for future rewards. We trained mice to perform stationary waiting and repeated lever-pressing tasks with variable reward delays and tested effects of optogenetic activation and inhibition of DRN serotonergic neurons on task performance. Interestingly, in the lever-pressing task, mice tolerated longer delays as they repeatedly pressed a lever than in the stationary waiting task, suggesting that lever-pressing actions may not simply be costly, but may also be subjectively rewarding. Optogenetic activation of DRN serotonergic neurons prolonged waiting in the stationary waiting task, consistent with previous studies, but it did not affect lever pressing time or numbers. While optogenetic inhibition decreased waiting, it did not affect lever pressing time or numbers. In generalized linear model analysis that incorporated the time during each session and the number of sessions, however, optogenetic activation negatively affected the number and the speed of lever pressing. These results revealed that the necessity of motor actions may increase motivation for delayed rewards and that DRN serotonergic neurons more significantly promote stationary waiting rather than persistent motor actions for future rewards.

animal behavior and cognition↗

bioRxiv2024

Histological and gene-expression analyses of pyloric sphincter formation during stomach metamorphosis in Xenopus laevis

During anuran metamorphosis from herbivorous tadpoles to carnivorous frogs, the gastrointestinal (GI) tract undergoes drastic remodeling, such as the formation of the stomach-intestine boundary and the development of the pyloric sphincter at the posterior end of the stomach. However, the morphogenetic process and molecular mechanisms of how the pyloric sphincter is formed during metamorphosis, instead of during embryogenesis as in amniotes, are largely uninvestigated. Using the African clawed frog Xenopus laevis, we histologically examined the development of the pylorus region from embryonic to froglet stages and performed spatiotemporal gene expression analyses. We found that the pyloric sphincter is formed at a flexure within the pyloric region during metamorphic climax, and that the pyloric and duodenal epithelia, which are morphologically indistinguishable before sphincter formation, become clearly demarcated by the sphincter at the end of metamorphosis. Consistent with these morphological changes, expression domains of a stomach marker barx1 and an intestine marker cdx2 overlapped until late metamorphic climax, but became separated after metamorphosis. Despite the absence of the sphincter before metamorphosis, various genes crucial for sphincter formation in amniotes were already expressed in the pylorus region of Xenopus embryos. RNA-sequencing analysis at pre-metamorphic and metamorphic-climax stages suggest unappreciated roles of genes, such as those for retinoic acid signaling and various transcription factors, in suppressing or promoting sphincter formation. These data provide histological and molecular insights into the heterochrony of the pyloric sphincter formation in amniotes and anurans.

developmental biology↗

bioRxiv2023

Zbtb11 interacts with Otx2 and patterns the anterior neuroectoderm in Xenopus

The zinc finger and BTB domain-containing 11 gene (zbtb11) is expressed in the Xenopus anterior neuroectoderm, but the molecular nature of the Zbtb11 protein during embryonic development remains to be elucidated. Here, we show the role of Zbtb11 in anterior patterning of the neuroectoderm and the cooperative action with the transcription factor Otx2. Both overexpression and knockdown of zbtb11 caused similar phenotypes: expanded expression of the posterior gene gbx2 in the neural plate, and later microcephaly with reduced eyes, suggesting that a proper level of zbtb11 expression is necessary for normal patterning of the neuroectoderm, including eye formation. Co-immunoprecipitation assays showed that Zbtb11 formed a complex with itself and with a phosphomimetic and repressive form of Otx2, suggesting that Zbtb11 forms a dimer or oligomer and interacts with Otx2 in a phosphorylation-dependent manner. Reporter analysis further showed that Zbtb11 enhanced the activity of the phosphomimetic Otx2 to repress a silencer element of the posterior gene meis3. These data suggest that Zbtb11 coordinates with phosphorylated Otx2 to specify the anterior neuroectoderm by repressing posterior genes.

developmental biology↗

bioRxiv2022

Optogenetic activation of dorsal raphe serotonin neurons induces a brain-wide response in reward network

Serotonin is a neuromodulator that affects multiple behavioral and cognitive functions. Nonetheless, how serotonin causes such a variety of effects via brain-wide projections and various receptors remains unclear. Here we measured brain-wide responses to optogenetic stimulation of serotonin neurons in the dorsal raphe nucleus (DRN) of the male mouse brain using functional MRI with an 11.7T scanner and a cryoprobe. Transient activation of DRN serotonin neurons caused brain-wide activation, including the medial prefrontal cortex, the striatum, and the ventral tegmental area. The same stimulation under anesthesia with isoflurane decreased brain-wide activation, including the hippocampal complex. These brain-wide response patterns can be explained by DRN serotonergic projection topography and serotonin receptor expression profiles, with enhanced weights on 5-HT1 receptors. Together, these results provide insight into the DR serotonergic system, which is consistent with recent discoveries of its functions in adaptive behaviors.

neuroscience↗

bioRxiv2020

Asymmetric requirement of Dpp/BMP morphogen dispersal in the Drosophila wing disc

How morphogen gradients control patterning and growth in developing tissues remains largely unknown due to lack of tools manipulating morphogen gradients. Here, we generate two membrane-tethered protein binders that manipulate different aspects of Decapentaplegic (Dpp), a morphogen required for overall patterning and growth of the Drosophila wing. "HA trap" is based on a single-chain variable fragment (scFv) against the HA tag that traps HA-Dpp to mainly block its dispersal, while "Dpp trap" is based on a Designed Ankyrin Repeat Protein (DARPin) against Dpp that traps Dpp to block both its dispersal and signaling. Using these tools, we found that, while posterior patterning and growth require Dpp dispersal, anterior patterning and growth largely proceed without Dpp dispersal. We show that dpp transcriptional refinement from an initially uniform to a localized expression and persistent signaling in transient dpp source cells render the anterior compartment robust against the absence of Dpp dispersal. Furthermore, despite a critical requirement of dpp for the overall wing growth, neither Dpp dispersal nor direct signaling is critical for lateral wing growth after wing pouch specification. These results challenge the long-standing dogma that Dpp dispersal is strictly required to control and coordinate overall wing patterning and growth.

developmental biology↗

bioRxiv2020

Uncovering the mesendoderm gene regulatory network through multi-omic data integration

Mesendodermal specification is one of the earliest events in embryogenesis, where cells first acquire distinct identities. Cell differentiation is a highly regulated process that involves the function of numerous transcription factors (TFs) and signaling molecules, which can be described with gene regulatory networks (GRNs). Cell differentiation GRNs are difficult to build because existing mechanistic methods are low-throughput, and high-throughput methods tend to be non-mechanistic. Additionally, integrating highly dimensional data comprised of more than two data types is challenging. Here, we use linked self-organizing maps to combine ChIP-seq/ATAC-seq with temporal, spatial and perturbation RNA-seq data from Xenopus tropicalis mesendoderm development to build a high resolution genome scale mechanistic GRN. We recovered both known and previously unsuspected TF-DNA/TF-TF interactions and validated through reporter assays. Our analysis provides new insights into transcriptional regulation of early cell fate decisions and provides a general approach to building GRNs using highly-dimensional multi-omic data sets. HighlightsO_LIBuilt a generally applicable pipeline to creating GRNs using highly-dimensional multi-omic data sets C_LIO_LIPredicted new TF-DNA/TF-TF interactions during mesendoderm development C_LIO_LIGenerate the first genome scale GRN for vertebrate mesendoderm and expanded the core mesendodermal developmental network with high fidelity C_LIO_LIDeveloped a resource to visualize hundreds of RNA-seq and ChIP-seq data using 2D SOM metaclusters. C_LI

genomics↗

bioRxiv2020

Quantitative analyses reveal extracellular dynamics of Wnt ligands in Xenopus embryos

The mechanism of intercellular transport of Wnt ligands is still a matter of debate. Here, to better understand this issue, we examined distribution and dynamics of Wnt8 in Xenopus embryos. While Venus-tagged Wnt8 was found on the surfaces of cells close to Wnt-producing cells, we also detected its dispersal over distances of 15 cell diameters. A combination of fluorescence correlation spectroscopy and quantitative imaging revealed that only a small proportion of Wnt8 ligands diffuses freely, whereas most Wnt8 molecules are bound to cell surfaces. Fluorescence decay after photoconversion showed that Wnt8 ligands bound on cell surfaces decreased exponentially, suggesting a dynamic exchange of bound forms of Wnt ligands. Mathematical modelling based on this exchange recapitulates a graded distribution of bound, but not free, Wnt ligands. Based on these results, we propose that Wnt distribution in tissues is controlled by a dynamic exchange of its abundant bound and rare free populations.

developmental biology↗