In multi-cellular organisms, biological function emerges when heterogeneous cell types form complex organs. Nevertheless dissection of tissues into mixtures of cellular subpopulations is currently challenging. We introduce an automated massively parallel single-cell RNA sequencing approach for analyzing in vivo transcriptional states in thousands of single cells. Combined with unsupervised classification algorithms, this facilitates ab initio cell type characterization of splenic tissues. Modeling single-cell transcriptional states in dendritic cells and additional hematopoietic cell types uncovers rich cell-type heterogeneity and gene-modules activity in steady-state and after pathogen activation. Cellular diversity is thereby approached through inference of variable and dynamic pathway activity rather than a fixed pre-programmed cell-type hierarchy. These data demonstrate single-cell RNA-Seq as an effective tool for comprehensive cellular decomposition of complex tissues.

Chromatin modifications are crucial for development, yet little is known about their dynamics during differentiation. Hematopoiesis provides a well-defined model to study chromatin state dynamics, however technical limitations impede profiling of homogeneous differentiation intermediates. We developed a high sensitivity indexing-first chromatin immunoprecipitation approach (iChIP) to profile the dynamics of four chromatin modifications across 16 stages of hematopoietic differentiation. We identify 48,415 enhancer regions and characterize their dynamics. We find that lineage commitment involves de novo establishment of 17,035 lineage-specific enhancers. These enhancer repertoire expansions foreshadow transcriptional programs in differentiated cells. Combining our enhancer catalog with gene expression profiles, we elucidate the transcription factor network controlling chromatin dynamics and lineage specification in hematopoiesis. Together, our results provide a comprehensive model of chromatin dynamics during development.

A widespread feature of extracellular signaling in cell circuits is paradoxical pleiotropy: the same secreted signaling molecule can induce opposite effects in the responding cells. For example, the cytokine IL-2 can promote proliferation and death of T cells. The role of such paradoxical signaling remains unclear. To address this, we studied CD4(+) T cell expansion in culture. We found that cells with a 30-fold difference in initial concentrations reached a homeostatic concentration nearly independent of initial cell levels. Below an initial threshold, cell density decayed to extinction (OFF-state). We show that these dynamics relate to the paradoxical effect of IL-2, which increases the proliferation rate cooperatively and the death rate linearly. Mathematical modeling explained the observed cell and cytokine dynamics and predicted conditions that shifted cell fate from homeostasis to the OFF-state. We suggest that paradoxical signaling provides cell circuits with specific dynamical features that are robust to environmental perturbations.

The molecular mechanism governing affinity-based B cell selection for germinal center colonization is unclear. Zaretsky et al. show that B cell ICAMs promote efficient B cell selection for clonal expansion by supporting sustained interactions with T follicular helper cells.

During cell differentiation, progenitor cells integrate signals from their environment that guide their development into specialized phenotypes. The ways by which cells respond to complex signal combinations remain difficult to analyze and model. To gain additional insight into signal integration, we systematically mapped the response of CD4 + T cells to a large number of input cytokine combinations that drive their differentiation. We find that, in response to varied input combinations, cells differentiate into a continuum of cell fates as opposed to a limited number of discrete phenotypes. Input cytokines hierarchically influence the cell population, with TGFβ being most dominant followed by IL-6 and IL-4. Mathematical modeling explains these results using additive signal integration within hierarchical groups of input cytokine combinations and correctly predicts cell population response to new input conditions. These findings suggest that complex cellular responses can be effectively described using a segmented linear approach, providing a framework for prediction of cellular responses to new cytokine combinations and doses, with implications to fine-tuned immunotherapies.C ell differentiation is controlled by complex gene regulatory networks that determine the expression of a large number of genes in response to external stimuli. CD4 + T cells, central regulators of immune responses, can differentiate into a number of phenotypes (or cell states), each defined by the expression of a panel of genes, such as lineage-specifying transcription factors (TFs) and cytokines (1). Differentiation outcome depends on a number of factors, including strength of antigen stimulation (2, 3), interactions with antigen presenting cells, and the signaling environment defined by secreted cytokines (1). By applying specific cytokine signals, it is possible to direct antigen-activated T cells toward differentiation into a number of specific phenotypes. Although the molecular interactions and regulatory networks that govern the differentiation process are being revealed at increasing resolution (4, 5), understanding the logic of operation of these complex networks and developing predictive mathematical models for cellular responses remain a challenge.Mathematical models that describe complex cellular processes are difficult to construct because of a lack of complete information about the underlying networks of molecular interactions and in particular, missing quantitative data regarding the parameters of biochemical interactions within these networks. Alternatively, a "black box" approach can be used, constructing a model of the system by analyzing its response to different input conditions (Fig. 1A). This technique enables study of complex systems in a simplified manner, because its implementation merely requires measurable inputs and outputs, without quantifying the response of each component inside the box. Approaches based on this concept are highly useful for describing engineered systems and were also applied for studying som...