T-cell receptor repertoire analysis of CD4-positive T cells from blood and an affected organ in an autoimmune mouse model.

One hallmark of some autoimmune diseases is the variability of symptoms among individuals. Organs affected by the disease differ between patients, posing a challenge in diagnosing the affected organs. Although numerous studies have investigated the correlation between T cell antigen receptor (TCR) repertoires and the development of infectious and immune diseases, the correlation between TCR repertoires and variations in disease symptoms among individuals remains unclear. This study aimed to investigate the correlation of TCRα and ß repertoires in blood T cells with the extent of autoimmune signs that varies among individuals. We sequenced TCRα and ß of CD4+ CD44high CD62Llow T cells in the blood and stomachs of mice deficient in autoimmune regulator (Aire) (AIRE KO), a mouse model of human autoimmune polyendocrinopathy-candidiasis-ectodermal dystrophy. Data analysis revealed that the degree of similarity in TCR sequences between the blood and stomach varied among individual AIRE KO mice and reflected the extent of T cell infiltration in the stomach. We identified a set of TCR sequences whose frequencies in blood might correlate with extent of the stomach manifestations. Our results propose a potential of using TCR repertoires not only for diagnosing disease development but also for diagnosing affected organs in autoimmune diseases.

Forward-Backward Sweep Method for the System of HJB-FP Equations in Memory-Limited Partially Observable Stochastic Control.

Memory-limited partially observable stochastic control (ML-POSC) is the stochastic optimal control problem under incomplete information and memory limitation. To obtain the optimal control function of ML-POSC, a system of the forward Fokker-Planck (FP) equation and the backward Hamilton-Jacobi-Bellman (HJB) equation needs to be solved. In this work, we first show that the system of HJB-FP equations can be interpreted via Pontryagin's minimum principle on the probability density function space. Based on this interpretation, we then propose the forward-backward sweep method (FBSM) for ML-POSC. FBSM is one of the most basic algorithms for Pontryagin's minimum principle, which alternately computes the forward FP equation and the backward HJB equation in ML-POSC. Although the convergence of FBSM is generally not guaranteed in deterministic control and mean-field stochastic control, it is guaranteed in ML-POSC because the coupling of the HJB-FP equations is limited to the optimal control function in ML-POSC.

Decentralized Stochastic Control with Finite-Dimensional Memories: A Memory Limitation Approach.

Decentralized stochastic control (DSC) is a stochastic optimal control problem consisting of multiple controllers. DSC assumes that each controller is unable to accurately observe the target system and the other controllers. This setup results in two difficulties in DSC; one is that each controller has to memorize the infinite-dimensional observation history, which is not practical, because the memory of the actual controllers is limited. The other is that the reduction of infinite-dimensional sequential Bayesian estimation to finite-dimensional Kalman filter is impossible in general DSC, even for linear-quadratic-Gaussian (LQG) problems. In order to address these issues, we propose an alternative theoretical framework to DSC-memory-limited DSC (ML-DSC). ML-DSC explicitly formulates the finite-dimensional memories of the controllers. Each controller is jointly optimized to compress the infinite-dimensional observation history into the prescribed finite-dimensional memory and to determine the control based on it. Therefore, ML-DSC can be a practical formulation for actual memory-limited controllers. We demonstrate how ML-DSC works in the LQG problem. The conventional DSC cannot be solved except in the special LQG problems where the information the controllers have is independent or partially nested. We show that ML-DSC can be solved in more general LQG problems where the interaction among the controllers is not restricted.

Memory-Limited Partially Observable Stochastic Control and Its Mean-Field Control Approach.

Control problems with incomplete information and memory limitation appear in many practical situations. Although partially observable stochastic control (POSC) is a conventional theoretical framework that considers the optimal control problem with incomplete information, it cannot consider memory limitation. Furthermore, POSC cannot be solved in practice except in special cases. In order to address these issues, we propose an alternative theoretical framework, memory-limited POSC (ML-POSC). ML-POSC directly considers memory limitation as well as incomplete information, and it can be solved in practice by employing the technique of mean-field control theory. ML-POSC can generalize the linear-quadratic-Gaussian (LQG) problem to include memory limitation. Because estimation and control are not clearly separated in the LQG problem with memory limitation, the Riccati equation is modified to the partially observable Riccati equation, which improves estimation as well as control. Furthermore, we demonstrate the effectiveness of ML-POSC for a non-LQG problem by comparing it with the local LQG approximation.

Lineage EM algorithm for inferring latent states from cellular lineage trees.

SUMMARY: Phenotypic variability in a population of cells can work as the bet-hedging of the cells under an unpredictably changing environment, the typical example of which is the bacterial persistence. To understand the strategy to control such phenomena, it is indispensable to identify the phenotype of each cell and its inheritance. Although recent advancements in microfluidic technology offer us useful lineage data, they are insufficient to directly identify the phenotypes of the cells. An alternative approach is to infer the phenotype from the lineage data by latent-variable estimation. To this end, however, we must resolve the bias problem in the inference from lineage called survivorship bias. In this work, we clarify how the survivorship bias distorts statistical estimations. We then propose a latent-variable estimation algorithm without the survivorship bias from lineage trees based on an expectation-maximization (EM) algorithm, which we call lineage EM algorithm (LEM). LEM provides a statistical method to identify the traits of the cells applicable to various kinds of lineage data. AVAILABILITY AND IMPLEMENTATION: An implementation of LEM is available at https://github.com/so-nakashima/Lineage-EM-algorithm. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Connection between the Bacterial Chemotactic Network and Optimal Filtering.

The chemotactic network of Escherichia coli has been studied extensively both biophysically and information theoretically. Nevertheless, connection between these two aspects is still elusive. In this work, we report such a connection. We derive an optimal filtering dynamics under the assumption that E. coli's sensory system optimally infers the binary information whether it is swimming up or down along an exponential ligand gradient from noisy sensory signals. Then we show that a standard biochemical model of the chemotactic network is mathematically equivalent to this information-theoretically optimal dynamics. Moreover, we demonstrate that an experimentally observed nonlinear response relation can be reproduced from the optimal dynamics. These results suggest that the biochemical network of E. coli chemotaxis is designed to optimally extract the binary information along an exponential gradient in a noisy condition.

Forward and Backward Bellman Equations Improve the Efficiency of the EM Algorithm for DEC-POMDP.

Decentralized partially observable Markov decision process (DEC-POMDP) models sequential decision making problems by a team of agents. Since the planning of DEC-POMDP can be interpreted as the maximum likelihood estimation for the latent variable model, DEC-POMDP can be solved by the EM algorithm. However, in EM for DEC-POMDP, the forward-backward algorithm needs to be calculated up to the infinite horizon, which impairs the computational efficiency. In this paper, we propose the Bellman EM algorithm (BEM) and the modified Bellman EM algorithm (MBEM) by introducing the forward and backward Bellman equations into EM. BEM can be more efficient than EM because BEM calculates the forward and backward Bellman equations instead of the forward-backward algorithm up to the infinite horizon. However, BEM cannot always be more efficient than EM when the size of problems is large because BEM calculates an inverse matrix. We circumvent this shortcoming in MBEM by calculating the forward and backward Bellman equations without the inverse matrix. Our numerical experiments demonstrate that the convergence of MBEM is faster than that of EM.

Reliable target ligand detection by noise-induced receptor cluster formation.

Intracellular reactions are intrinsically stochastic. Nonetheless, cells can reliably respond to the changing environment by sensing their target molecules sensitively and specifically, even with the existence of abundant structurally-similar non-target molecules. The mechanism of how the cells can balance and achieve such different characteristics is not yet fully understood. In this work, we demonstrate that these characteristics can be attained by a ligand-induced stochastic cluster formation of receptors via the noise-induced symmetry breaking, in which the intrinsic stochasticity works to enhance sensitivity and specificity. We also show that the noise-induced cluster formation enables cells to detect the target ligand reliably by compensating the abundant non-target ligands in the environment. The proposed mechanism may lead to a deeper understanding of a biological function of the receptor clustering and provide an alternative candidate for the reliable ligand detection to the kinetic proofreading.

Long-term hindlimb unloading causes a preferential reduction of medullary thymic epithelial cells expressing autoimmune regulator (Aire).

Hindlimb unloading (HU) of rodents has been used as a ground-based model of spaceflight. In this study, we investigated the detailed impact of 14-day HU on the murine thymus. Thymic mass and cell number were significantly reduced after 14 days of hindlimb unloading, which was accompanied by an increment of plasma corticosterone. Although corticosterone reportedly causes selective apoptosis of CD4+CD8+ thymocytes (CD4+CD8+DPs) in mice treated with short-term HU, the reduction of thymocyte cellularity after the 14-day HU was not selective for CD4+CD8+DPs. In addition to the thymocyte reduction, the cellularity of thymic epithelial cells (TECs) was also reduced by the 14-day HU. Flow cytometric and RNA-sequencing analysis suggested that medullary TECs (mTECs) were preferentially reduced after HU. Moreover, immunohistochemical staining suggested that the 14-day HU caused a reduction of the mTECs expressing autoimmune regulator (Aire). Our data suggested that HU impacts both thymocytes and TECs. Consequently, these data imply that thymic T cell repertoire formation could be disturbed during spaceflight-like stress.

Systematic time-dependent visualization and quantitation of the neurogenic rate in brain organoids.

Organoids mimicking the formation of the brain cortex have been demonstrated to be powerful tools for developmental studies as well as pathological investigations of brain malformations. Here, we report an integrated approach for the quantification of temporal neural production (neurogenic rate) in organoids derived from embryonic brains. Spherical tissue fragments with polarized cytoarchitectures were incubated in multiple cavities arranged in a polymethylmethacrylate chip. The time-dependent neurogenic rate in the organoids was monitored by the level of EGFP under the promoter of Tbr2, a transcription factor that is transiently expressed in neural fate-committed progenitors during corticogenesis. Importantly, our monitoring system exhibited a quick response to DAPT, a drug that promotes neural differentiation. Furthermore, we successfully quantified the temporal neurogenic rate in a large number of organoids by applying image processing that semi-automatically recognized the positions of organoids and measured their signal intensities from sequential images. Taken together, we provide a strategy to quantitate the neurogenic rate in brain organoids in a time-dependent manner, which will also be a potent method for monitoring organoid formation and drug activity in other tissue types.

Single cell visualization of yeast gene expression shows correlation of epigenetic switching between multiple heterochromatic regions through multiple generations.

Differences in gene expression between individual cells can be mediated by epigenetic regulation; thus, methods that enable detailed analyses of single cells are crucial to understanding this phenomenon. In this study, genomic silencing regions of Saccharomyces cerevisiae that are subject to epigenetic regulation, including the HMR, HML, and telomere regions, were investigated using a newly developed single cell analysis method. This method uses fluorescently labeled proteins to track changes in gene expression over multiple generations of a single cell. Epigenetic control of gene expression differed depending on the specific silencing region at which the reporter gene was inserted. Correlations between gene expression at the HMR-left and HMR-right regions, as well as the HMR-right and HML-right regions, were observed in the single-cell level; however, no such correlations involving the telomere region were observed. Deletion of the histone acetyltransferase GCN5 gene from a yeast strain carrying a fluorescent reporter gene at the HMR-left region reduced the frequency of changes in gene expression over a generation. The results presented here suggest that epigenetic control within an individual cell is reversible and can be achieved via regulation of histone acetyltransferase activity.

Fluctuation Relations of Fitness and Information in Population Dynamics.

Phenotype switching with and without sensing environment is a common strategy of organisms to survive in a fluctuating environment. Understanding the evolutionary advantages of switching and sensing requires a quantitative evaluation of their fitness gain and its fluctuation together with the conditions for the switching and sensing strategies being adapted to a given environment. In this work, by using a pathwise formulation of the population dynamics, we show that the optimal switching strategy is characterized by a consistency condition for time-forward and backward path probabilities. The formulation also clarifies the underlying information-theoretic aspect of selection as a passive information compression. The loss of fitness by a suboptimal strategy is also shown to satisfy a fluctuation relation, which provides us with the information on how environmental fluctuation impacts the advantages of the optimal strategy. These results are naturally extended to the situation that organisms can use an environmental signal by actively sensing the environment. The fluctuation relations of the fitness gain by sensing are derived in which the multivariate mutual information among the phenotype, the environment, and the signal plays the role to quantify the relevant information in the signal for the fitness gain.

Feedback repression is required for mammalian circadian clock function.

Direct evidence for the requirement of transcriptional feedback repression in circadian clock function has been elusive. Here, we developed a molecular genetic screen in mammalian cells to identify mutants of the circadian transcriptional activators CLOCK and BMAL1, which were uncoupled from CRYPTOCHROME (CRY)-mediated transcriptional repression. Notably, mutations in the PER-ARNT-SIM domain of CLOCK and the C terminus of BMAL1 resulted in synergistic insensitivity through reduced physical interactions with CRY. Coexpression of these mutant proteins in cultured fibroblasts caused arrhythmic phenotypes in population and single-cell assays. These data demonstrate that CRY-mediated repression of the CLOCK/BMAL1 complex activity is required for maintenance of circadian rhythmicity and provide formal proof that transcriptional feedback is required for mammalian clock function.

Melanopsin-dependent photo-perturbation reveals desynchronization underlying the singularity of mammalian circadian clocks.

Singularity behaviour in circadian clocks--the loss of robust circadian rhythms following exposure to a stimulus such as a pulse of bright light--is one of the fundamental but mysterious properties of clocks. To quantitatively perturb and accurately measure the dynamics of cellular clocks, we synthetically produced photo-responsiveness within mammalian cells by exogenously introducing the photoreceptor melanopsin and continuously monitoring the effect of photo-perturbation on the state of cellular clocks. Here we report that a critical light pulse drives cellular clocks into singularity behaviour. Our theoretical analysis consistently predicts and subsequent single-cell level observation directly proves that desynchronization of individual cellular clocks underlies singularity behaviour. Our theoretical framework also explains why singularity behaviours have been experimentally observed in various organisms, and it suggests that desynchronization is a plausible mechanism for the observable singularity of circadian clocks. Importantly, these in vitro and in silico findings are further supported by in vivo observations that desynchronization underlies the multicell-level amplitude decrease in the rat suprachiasmatic nucleus induced by critical light pulses.

Information geometric bound on general chemical reaction networks.

We investigate the convergence of chemical reaction networks (CRNs), aiming to establish an upper bound on their reaction rates. The nonlinear characteristics and discrete composition of CRNs pose significant challenges in this endeavor. To circumvent these complexities, we adopt an information geometric perspective, utilizing the natural gradient to formulate a nonlinear system. This system effectively determines an upper bound for the dynamics of CRNs. We corroborate our methodology through numerical simulations, which reveal that our constructed system converges more rapidly than CRNs within a particular class of reactions. This class is defined by the count of chemicals, the highest stoichiometric coefficients in the reactions, and the total number of reactions involved. Further, we juxtapose our approach with traditional methods, illustrating that the latter falls short in providing an upper bound for CRN reaction rates. Although our investigation centers on CRNs, the widespread presence of hypergraphs across various disciplines, ranging from natural sciences to engineering, indicates potential wider applications of our method, including in the realm of information science.

Mitochondrial protein C15ORF48 is a stress-independent inducer of autophagy that regulates oxidative stress and autoimmunity.

Autophagy is primarily activated by cellular stress, such as starvation or mitochondrial damage. However, stress-independent autophagy is activated by unclear mechanisms in several cell types, such as thymic epithelial cells (TECs). Here we report that the mitochondrial protein, C15ORF48, is a critical inducer of stress-independent autophagy. Mechanistically, C15ORF48 reduces the mitochondrial membrane potential and lowers intracellular ATP levels, thereby activating AMP-activated protein kinase and its downstream Unc-51-like kinase 1. Interestingly, C15ORF48-dependent induction of autophagy upregulates intracellular glutathione levels, promoting cell survival by reducing oxidative stress. Mice deficient in C15orf48 show a reduction in stress-independent autophagy in TECs, but not in typical starvation-induced autophagy in skeletal muscles. Moreover, C15orf48-/- mice develop autoimmunity, which is consistent with the fact that the stress-independent autophagy in TECs is crucial for the thymic self-tolerance. These results suggest that C15ORF48 induces stress-independent autophagy, thereby regulating oxidative stress and self-tolerance.

Acute irradiation causes a long-term disturbance in the heterogeneity and gene expression profile of medullary thymic epithelial cells.

The thymus has the ability to regenerate from acute injury caused by radiation, infection, and stressors. In addition to thymocytes, thymic epithelial cells in the medulla (mTECs), which are crucial for T cell self-tolerance by ectopically expressing and presenting thousands of tissue-specific antigens (TSAs), are damaged by these insults and recover thereafter. However, given recent discoveries on the high heterogeneity of mTECs, it remains to be determined whether the frequency and properties of mTEC subsets are restored during thymic recovery from radiation damage. Here we demonstrate that acute total body irradiation with a sublethal dose induces aftereffects on heterogeneity and gene expression of mTECs. Single-cell RNA-sequencing (scRNA-seq) analysis showed that irradiation reduces the frequency of mTECs expressing AIRE, which is a critical regulator of TSA expression, 15 days after irradiation. In contrast, transit-amplifying mTECs (TA-mTECs), which are progenitors of AIRE-expressing mTECs, and Ccl21a-expressing mTECs, were less affected. Interestingly, a detailed analysis of scRNA-seq data suggested that the proportion of a unique mTEC cluster expressing Ccl25 and a high level of TSAs was severely decreased by irradiation. In sum, we propose that the effects of acute irradiation disrupt the heterogeneity and properties of mTECs over an extended period, which potentially leads to an impairment of thymic T cell selection.

Theoretical aspects of cellular decision-making and information-processing.

Microscopic biological processes have extraordinary complexity and variety at the sub-cellular, intra-cellular, and multi-cellular levels. In dealing with such complex phenomena, conceptual and theoretical frameworks are crucial, which enable us to understand seemingly different intra- and inter-cellular phenomena from unified viewpoints. Decision-making is one such concept that has attracted much attention recently. Since a number of cellular behavior can be regarded as processes to make specific actions in response to external stimuli, decision-making can cover and has been used to explain a broad range of different cellular phenomena [Balázsi et al. (Cell 144(6):910, 2011), Zeng et al. (Cell 141(4):682, 2010)]. Decision-making is also closely related to cellular information-processing because appropriate decisions cannot be made without exploiting the information that the external stimuli contain. Efficiency of information transduction and processing by intra-cellular networks determines the amount of information obtained, which in turn limits the efficiency of subsequent decision-making. Furthermore, information-processing itself can serve as another concept that is crucial for understanding of other biological processes than decision-making. In this work, we review recent theoretical developments on cellular decision-making and information-processing by focusing on the relation between these two concepts.

Comparative Study of Repertoire Classification Methods Reveals Data Efficiency of k -mer Feature Extraction.

The repertoire of T cell receptors encodes various types of immunological information. Machine learning is indispensable for decoding such information from repertoire datasets measured by next-generation sequencing (NGS). In particular, the classification of repertoires is the most basic task, which is relevant for a variety of scientific and clinical problems. Supported by the recent appearance of large datasets, efficient but data-expensive methods have been proposed. However, it is unclear whether they can work efficiently when the available sample size is severely restricted as in practical situations. In this study, we demonstrate that their performances can be impaired substantially below critical sample sizes. To complement this drawback, we propose MotifBoost, which exploits the information of short k-mer motifs of TCRs. MotifBoost can perform the classification as efficiently as a deep learning method on large datasets while providing more stable and reliable results on small datasets. We tested MotifBoost on the four small datasets which consist of various conditions such as Cytomegalovirus (CMV), HIV, α-chain, ß-chain and it consistently preserved the stability. We also clarify that the robustness of MotifBoost can be attributed to the efficiency of k-mer motifs as representation features of repertoires. Finally, by comparing the predictions of these methods, we show that the whole sequence identity and sequence motifs encode partially different information and that a combination of such complementary information is necessary for further development of repertoire analysis.

Machine Learning Approaches to TCR Repertoire Analysis.

Sparked by the development of genome sequencing technology, the quantity and quality of data handled in immunological research have been changing dramatically. Various data and database platforms are now driving the rapid progress of machine learning for immunological data analysis. Of various topics in immunology, T cell receptor repertoire analysis is one of the most important targets of machine learning for assessing the state and abnormalities of immune systems. In this paper, we review recent repertoire analysis methods based on machine learning and deep learning and discuss their prospects.