Patterning and folding of intestinal villi by active mesenchymal dewetting.

Tissue folds are structural motifs critical to organ function. In the intestine, bending of a flat epithelium into a periodic pattern of folds gives rise to villi, finger-like protrusions that enable nutrient absorption. However, the molecular and mechanical processes driving villus morphogenesis remain unclear. Here, we identify an active mechanical mechanism that simultaneously patterns and folds the intestinal epithelium to initiate villus formation. At the cellular level, we find that PDGFRA+ subepithelial mesenchymal cells generate myosin II-dependent forces sufficient to produce patterned curvature in neighboring tissue interfaces. This symmetry-breaking process requires altered cell and extracellular matrix interactions that are enabled by matrix metalloproteinase-mediated tissue fluidization. Computational models, together with in vitro and in vivo experiments, revealed that these cellular features manifest at the tissue level as differences in interfacial tensions that promote mesenchymal aggregation and interface bending through a process analogous to the active dewetting of a thin liquid film.

Spin glasses, error correcting codes, and synchronization of human stem cell organoids.

Magnetic spins, pendulum clocks, and fireflies all self-organize into coherent collectives when arranged into groups of spatially coupled and interacting individuals. Ramanathan and colleagues demonstrate that spatial coupling of human stem cell organoids induces coherent progression through developmental transitions, allowing the dissection of molecular circuits underlying human development.

Recovery of missing single-cell RNA-sequencing data with optimized transcriptomic references.

Single-cell RNA-sequencing (scRNA-seq) is an indispensable tool for characterizing cellular diversity and generating hypotheses throughout biology. Droplet-based scRNA-seq datasets often lack expression data for genes that can be detected with other methods. Here we show that the observed sensitivity deficits stem from three sources: (1) poor annotation of 3' gene ends; (2) issues with intronic read incorporation; and (3) gene overlap-derived read loss. We show that missing gene expression data can be recovered by optimizing the reference transcriptome for scRNA-seq through recovering false intergenic reads, implementing a hybrid pre-mRNA mapping strategy and resolving gene overlaps. We demonstrate, with a diverse collection of mouse and human tissue data, that reference optimization can substantially improve cellular profiling resolution and reveal missing cell types and marker genes. Our findings argue that transcriptomic references need to be optimized for scRNA-seq analysis and warrant a reanalysis of previously published datasets and cell atlases.

Pluripotency factors in embryonic stem cells regulate differentiation into germ layers.

Cell fate decisions are fundamental for development, but we do not know how transcriptional networks reorganize during the transition from a pluripotent to a differentiated cell state. Here, we asked how mouse embryonic stem cells (ESCs) leave the pluripotent state and choose between germ layer fates. By analyzing the dynamics of the transcriptional circuit that maintains pluripotency, we found that Oct4 and Sox2, proteins that maintain ESC identity, also orchestrate germ layer fate selection. Oct4 suppresses neural ectodermal differentiation and promotes mesendodermal differentiation; Sox2 inhibits mesendodermal differentiation and promotes neural ectodermal differentiation. Differentiation signals continuously and asymmetrically modulate Oct4 and Sox2 protein levels, altering their binding pattern in the genome, and leading to cell fate choice. The same factors that maintain pluripotency thus also integrate external signals and control lineage selection. Our study provides a framework for understanding how complex transcription factor networks control cell fate decisions in progenitor cells.

Controlling organization and forces in active matter through optically defined boundaries.

Living systems are capable of locomotion, reconfiguration and replication. To perform these tasks, cells spatiotemporally coordinate the interactions of force-generating, 'active' molecules that create and manipulate non-equilibrium structures and force fields of up to millimetre length scales1-3. Experimental active-matter systems of biological or synthetic molecules are capable of spontaneously organizing into structures4,5 and generating global flows6-9. However, these experimental systems lack the spatiotemporal control found in cells, limiting their utility for studying non-equilibrium phenomena and bioinspired engineering. Here we uncover non-equilibrium phenomena and principles of boundary-mediated control by optically modulating structures and fluid flow in an engineered system of active biomolecules. Our system consists of purified microtubules and light-activatable motor proteins that crosslink and organize the microtubules into distinct structures upon illumination. We develop basic operations-defined as sets of light patterns-to create, move and merge the microtubule structures. By combining these operations, we create microtubule networks that span several hundred micrometres in length and contract at speeds up to an order of magnitude higher than the speed of an individual motor protein. We manipulate these contractile networks to generate and sculpt persistent fluid flows. The principles of boundary-mediated control that we uncover may be used to study emergent cellular structures and forces and to develop programmable active-matter devices.

Dissecting heterogeneous cell populations across drug and disease conditions with PopAlign.

Single-cell measurement techniques can now probe gene expression in heterogeneous cell populations from the human body across a range of environmental and physiological conditions. However, new mathematical and computational methods are required to represent and analyze gene-expression changes that occur in complex mixtures of single cells as they respond to signals, drugs, or disease states. Here, we introduce a mathematical modeling platform, PopAlign, that automatically identifies subpopulations of cells within a heterogeneous mixture and tracks gene-expression and cell-abundance changes across subpopulations by constructing and comparing probabilistic models. Probabilistic models provide a low-error, compressed representation of single-cell data that enables efficient large-scale computations. We apply PopAlign to analyze the impact of 40 different immunomodulatory compounds on a heterogeneous population of donor-derived human immune cells as well as patient-specific disease signatures in multiple myeloma. PopAlign scales to comparisons involving tens to hundreds of samples, enabling large-scale studies of natural and engineered cell populations as they respond to drugs, signals, or physiological change.

Reinforcement learning reveals fundamental limits on the mixing of active particles.

The control of far-from-equilibrium physical systems, including active materials, requires advanced control strategies due to the non-linear dynamics and long-range interactions between particles, preventing explicit solutions to optimal control problems. In such situations, Reinforcement Learning (RL) has emerged as an approach to derive suitable control strategies. However, for active matter systems, it is an important open question how the mathematical structure and the physical properties determine the tractability of RL. In this paper, we demonstrate that RL can only find good mixing strategies for active matter systems that combine attractive and repulsive interactions. Using analytic results from dynamical systems theory, we show that combining both interaction types is indeed necessary for the existence of mixing-inducing hyperbolic dynamics and therefore the ability of RL to find homogeneous mixing strategies. In particular, we show that for drag-dominated translational-invariant particle systems, mixing relies on combined attractive and repulsive interactions. Therefore, our work demonstrates which experimental developments need to be made to make protein-based active matter applicable, and it provides some classification of microscopic interactions based on macroscopic behavior.

Environmental statistics and optimal regulation.

Any organism is embedded in an environment that changes over time. The timescale for and statistics of environmental change, the precision with which the organism can detect its environment, and the costs and benefits of particular protein expression levels all will affect the suitability of different strategies--such as constitutive expression or graded response--for regulating protein levels in response to environmental inputs. We propose a general framework-here specifically applied to the enzymatic regulation of metabolism in response to changing concentrations of a basic nutrient-to predict the optimal regulatory strategy given the statistics of fluctuations in the environment and measurement apparatus, respectively, and the costs associated with enzyme production. We use this framework to address three fundamental questions: (i) when a cell should prefer thresholding to a graded response; (ii) when there is a fitness advantage to implementing a Bayesian decision rule; and (iii) when retaining memory of the past provides a selective advantage. We specifically find that: (i) relative convexity of enzyme expression cost and benefit influences the fitness of thresholding or graded responses; (ii) intermediate levels of measurement uncertainty call for a sophisticated Bayesian decision rule; and (iii) in dynamic contexts, intermediate levels of uncertainty call for retaining memory of the past. Statistical properties of the environment, such as variability and correlation times, set optimal biochemical parameters, such as thresholds and decay rates in signaling pathways. Our framework provides a theoretical basis for interpreting molecular signal processing algorithms and a classification scheme that organizes known regulatory strategies and may help conceptualize heretofore unknown ones.

Motor-driven microtubule diffusion in a photobleached dynamical coordinate system.

Motor-driven cytoskeletal remodeling in cellular systems can often be accompanied by a diffusive-like effect at local scales, but distinguishing the contributions of the ordering process, such as active contraction of a network, from this active diffusion is difficult to achieve. Using light-dimerizable kinesin motors to spatially control the formation and contraction of a microtubule network, we deliberately photobleach a grid pattern onto the filament network serving as a transient and dynamic coordinate system to observe the deformation and translation of the remaining fluorescent squares of microtubules. We find that the network contracts at a rate set by motor speed but is accompanied by a diffusive-like spread throughout the bulk of the contracting network with effective diffusion constant two orders of magnitude lower than that for a freely-diffusing microtubule. We further find that on micron scales, the diffusive timescale is only a factor of approximately 3 slower than that of advection regardless of conditions, showing that the global contraction and long-time relaxation from this diffusive behavior are both motor-driven but exhibit local competition within the network bulk.

Spatial transcriptomics defines injury specific microenvironments and cellular interactions in kidney regeneration and disease.

Kidney injury disrupts the intricate renal architecture and triggers limited regeneration, together with injury-invoked inflammation and fibrosis. Deciphering the molecular pathways and cellular interactions driving these processes is challenging due to the complex tissue structure. Here, we apply single cell spatial transcriptomics to examine ischemia-reperfusion injury in the mouse kidney. Spatial transcriptomics reveals injury-specific and spatially-dependent gene expression patterns in distinct cellular microenvironments within the kidney and predicts Clcf1-Crfl1 in a molecular interplay between persistently injured proximal tubule cells and their neighboring fibroblasts. Immune cell types play a critical role in organ repair. Spatial analysis identifies cellular microenvironments resembling early tertiary lymphoid structures and associated molecular pathways. Collectively, this study supports a focus on molecular interactions in cellular microenvironments to enhance understanding of injury, repair and disease.

Engineering programmable material-to-cell pathways via synthetic notch receptors to spatially control differentiation in multicellular constructs.

Synthetic Notch (synNotch) receptors are genetically encoded, modular synthetic receptors that enable mammalian cells to detect environmental signals and respond by activating user-prescribed transcriptional programs. Although some materials have been modified to present synNotch ligands with coarse spatial control, applications in tissue engineering generally require extracellular matrix (ECM)-derived scaffolds and/or finer spatial positioning of multiple ligands. Thus, we develop here a suite of materials that activate synNotch receptors for generalizable engineering of material-to-cell signaling. We genetically and chemically fuse functional synNotch ligands to ECM proteins and ECM-derived materials. We also generate tissues with microscale precision over four distinct reporter phenotypes by culturing cells with two orthogonal synNotch programs on surfaces microcontact-printed with two synNotch ligands. Finally, we showcase applications in tissue engineering by co-transdifferentiating fibroblasts into skeletal muscle or endothelial cell precursors in user-defined micropatterns. These technologies provide avenues for spatially controlling cellular phenotypes in mammalian tissues.

Commensal bacteria promote type I interferon signaling to maintain immune tolerance in mice.

Type I interferons (IFNs) exert a broad range of biological effects important in coordinating immune responses, which have classically been studied in the context of pathogen clearance. Yet, whether immunomodulatory bacteria operate through IFN pathways to support intestinal immune tolerance remains elusive. Here, we reveal that the commensal bacterium, Bacteroides fragilis, utilizes canonical antiviral pathways to modulate intestinal dendritic cells (DCs) and regulatory T cell (Treg) responses. Specifically, IFN signaling is required for commensal-induced tolerance as IFNAR1-deficient DCs display blunted IL-10 and IL-27 production in response to B. fragilis. We further establish that IFN-driven IL-27 in DCs is critical in shaping the ensuing Foxp3+ Treg via IL-27Rα signaling. Consistent with these findings, single-cell RNA sequencing of gut Tregs demonstrated that colonization with B. fragilis promotes a distinct IFN gene signature in Foxp3+ Tregs during intestinal inflammation. Altogether, our findings demonstrate a critical role of commensal-mediated immune tolerance via tonic type I IFN signaling.

Unexplored regions of the protein sequence-structure map revealed at scale by a library of foldtuned language models.

Nature has likely sampled only a fraction of all protein sequences and structures allowed by the laws of biophysics. However, the combinatorial scale of amino-acid sequence-space has traditionally precluded substantive study of the full protein sequence-structure map. In particular, it remains unknown how much of the vast uncharted landscape of far-from-natural sequences consists of alternate ways to encode the familiar ensemble of natural folds; proteins in this category also represent an opportunity to diversify candidates for downstream applications. Here, we characterize sequence-structure mapping in far-from-natural regions of sequence-space guided by the capacity of protein language models (pLMs) to explore sequences outside their natural training data through generation. We demonstrate that pretrained generative pLMs sample a limited structural snapshot of the natural protein universe, including >350 common (sub)domain elements. Incorporating pLM, structure prediction, and structure-based search techniques, we surpass this limitation by developing a novel "foldtuning" strategy that pushes a pretrained pLM into a generative regime that maintains structural similarity to a target protein fold (e.g. TIM barrel, thioredoxin, etc) while maximizing dissimilarity to natural amino-acid sequences. We apply "foldtuning" to build a library of pLMs for >700 naturally-abundant folds in the SCOP database, accessing swaths of proteins that take familiar structures yet lie far from known sequences, spanning targets that include enzymes, immune ligands, and signaling proteins. By revealing protein sequence-structure information at scale outside of the context of evolution, we anticipate that this work will enable future systematic searches for wholly novel folds and facilitate more immediate protein design goals in catalysis and medicine.

TRILL: ORCHESTRATING MODULAR DEEP-LEARNING WORKFLOWS FOR DEMOCRATIZED, SCALABLE PROTEIN ANALYSIS AND ENGINEERING.

Deep-learning models have been rapidly adopted by many fields, partly due to the deluge of data humanity has amassed. In particular, the petabases of biological sequencing data enable the unsupervised training of protein language models that learn the "language of life." However, due to their prohibitive size and complexity, contemporary deep-learning models are often unwieldy, especially for scientists with limited machine learning backgrounds. TRILL (TRaining and Inference using the Language of Life) is a platform for creative protein design and discovery. Leveraging several state-of-the-art models such as ESM-2, DiffDock, and RFDiffusion, TRILL allows researchers to generate novel proteins, predict 3-D structures, extract high-dimensional representations of proteins, functionally classify proteins and more. What sets TRILL apart is its ability to enable complex pipelines by chaining together models and effectively merging the capabilities of different models to achieve a sum greater than its individual parts. Whether using Google Colab with one GPU or a supercomputer with hundreds, TRILL allows scientists to effectively utilize models with millions to billions of parameters by using optimized training strategies such as ZeRO-Offload and distributed data parallel. Therefore, TRILL not only bridges the gap between complex deep-learning models and their practical application in the field of biology, but also simplifies the orchestration of these models into comprehensive workflows, democratizing access to powerful methods. Documentation: https://trill.readthedocs.io/en/latest/home.html.

Spatio-temporal patterning of extensile active stresses in microtubule-based active fluids.

Microtubule-based active fluids exhibit turbulent-like autonomous flows, which are driven by the molecular motor powered motion of filamentous constituents. Controlling active stresses in space and time is an essential prerequisite for controlling the intrinsically chaotic dynamics of extensile active fluids. We design single-headed kinesin molecular motors that exhibit optically enhanced clustering and thus enable precise and repeatable spatial and temporal control of extensile active stresses. Such motors enable rapid, reversible switching between flowing and quiescent states. In turn, spatio-temporal patterning of the active stress controls the evolution of the ubiquitous bend instability of extensile active fluids and determines its critical length dependence. Combining optically controlled clusters with conventional kinesin motors enables one-time switching from contractile to extensile active stresses. These results open a path towards real-time control of the autonomous flows generated by active fluids.

Motor processivity and speed determine structure and dynamics of microtubule-motor assemblies.

Active matter systems can generate highly ordered structures, avoiding equilibrium through the consumption of energy by individual constituents. How the microscopic parameters that characterize the active agents are translated to the observed mesoscopic properties of the assembly has remained an open question. These active systems are prevalent in living matter; for example, in cells, the cytoskeleton is organized into structures such as the mitotic spindle through the coordinated activity of many motor proteins walking along microtubules. Here, we investigate how the microscopic motor-microtubule interactions affect the coherent structures formed in a reconstituted motor-microtubule system. This question is of deeper evolutionary significance as we suspect motor and microtubule type contribute to the shape and size of resulting structures. We explore key parameters experimentally and theoretically, using a variety of motors with different speeds, processivities, and directionalities. We demonstrate that aster size depends on the motor used to create the aster, and develop a model for the distribution of motors and microtubules in steady-state asters that depends on parameters related to motor speed and processivity. Further, we show that network contraction rates scale linearly with the single-motor speed in quasi-one-dimensional contraction experiments. In all, this theoretical and experimental work helps elucidate how microscopic motor properties are translated to the much larger scale of collective motor-microtubule assemblies.

D-SPIN constructs gene regulatory network models from multiplexed scRNA-seq data revealing organizing principles of cellular perturbation response.

Gene regulatory networks within cells modulate the expression of the genome in response to signals and changing environmental conditions. Reconstructions of gene regulatory networks can reveal the information processing and control principles used by cells to maintain homeostasis and execute cell-state transitions. Here, we introduce a computational framework, D-SPIN, that generates quantitative models of gene-regulatory networks from single-cell mRNA-seq data sets collected across thousands of distinct perturbation conditions. D-SPIN models the cell as a collection of interacting gene-expression programs, and constructs a probabilistic model to infer regulatory interactions between gene-expression programs and external perturbations. Using large Perturb-seq and drug-response datasets, we demonstrate that D-SPIN models reveal the organization of cellular pathways, sub-functions of macromolecular complexes, and the logic of cellular regulation of transcription, translation, metabolism, and protein degradation in response to gene knockdown perturbations. D-SPIN can also be applied to dissect drug response mechanisms in heterogeneous cell populations, elucidating how combinations of immunomodulatory drugs can induce novel cell states through additive recruitment of gene expression programs. D-SPIN provides a computational framework for constructing interpretable models of gene-regulatory networks to reveal principles of cellular information processing and physiological control.

Configurational entropy is an intrinsic driver of tissue structural heterogeneity.

Tissues comprise ordered arrangements of cells that can be surprisingly disordered in their details. How the properties of single cells and their microenvironment contribute to the balance between order and disorder at the tissue-scale remains poorly understood. Here, we address this question using the self-organization of human mammary organoids as a model. We find that organoids behave like a dynamic structural ensemble at the steady state. We apply a maximum entropy formalism to derive the ensemble distribution from three measurable parameters - the degeneracy of structural states, interfacial energy, and tissue activity (the energy associated with positional fluctuations). We link these parameters with the molecular and microenvironmental factors that control them to precisely engineer the ensemble across multiple conditions. Our analysis reveals that the entropy associated with structural degeneracy sets a theoretical limit to tissue order and provides new insight for tissue engineering, development, and our understanding of disease progression.

Spatial transcriptomics defines injury-specific microenvironments in the adult mouse kidney and novel cellular interactions in regeneration and disease.

Kidney injury disrupts the intricate renal architecture and triggers limited regeneration, and injury-invoked inflammation and fibrosis. Deciphering molecular pathways and cellular interactions driving these processes is challenging due to the complex renal architecture. Here, we applied single cell spatial transcriptomics to examine ischemia-reperfusion injury in the mouse kidney. Spatial transcriptomics revealed injury-specific and spatially-dependent gene expression patterns in distinct cellular microenvironments within the kidney and predicted Clcf1-Crfl1 in a molecular interplay between persistently injured proximal tubule cells and neighboring fibroblasts. Immune cell types play a critical role in organ repair. Spatial analysis revealed cellular microenvironments resembling early tertiary lymphoid structures and identified associated molecular pathways. Collectively, this study supports a focus on molecular interactions in cellular microenvironments to enhance understanding of injury, repair and disease.

Extreme differences in SARS-CoV-2 viral loads among respiratory specimen types during presumed pre-infectious and infectious periods.

SARS-CoV-2 viral-load measurements from a single-specimen type are used to establish diagnostic strategies, interpret clinical-trial results for vaccines and therapeutics, model viral transmission, and understand virus-host interactions. However, measurements from a single-specimen type are implicitly assumed to be representative of other specimen types. We quantified viral-load timecourses from individuals who began daily self-sampling of saliva, anterior-nares (nasal), and oropharyngeal (throat) swabs before or at the incidence of infection with the Omicron variant. Viral loads in different specimen types from the same person at the same timepoint exhibited extreme differences, up to 109 copies/mL. These differences were not due to variation in sample self-collection, which was consistent. For most individuals, longitudinal viral-load timecourses in different specimen types did not correlate. Throat-swab and saliva viral loads began to rise as many as 7 days earlier than nasal-swab viral loads in most individuals, leading to very low clinical sensitivity of nasal swabs during the first days of infection. Individuals frequently exhibited presumably infectious viral loads in one specimen type while viral loads were low or undetectable in other specimen types. Therefore, defining an individual as infectious based on assessment of a single-specimen type underestimates the infectious period, and overestimates the ability of that specimen type to detect infectious individuals. For diagnostic COVID-19 testing, these three single-specimen types have low clinical sensitivity, whereas a combined throat-nasal swab, and assays with high analytical sensitivity, was inferred to have significantly better clinical sensitivity to detect presumed pre-infectious and infectious individuals.