Intercellular interactions during T cell activation and differentiation

Cells of the immune system are not acting autonomously; rather their responses strongly rely on signals that they receive from other cells in the system. Communication between immune cells is necessary to control immune responses in time and space, and probably takes an important part in immune recognition and information processing as well. The complexity of the intercellular communication network obscures analysis and understanding of its function. A possible approach towards simplification of the description of the system is by recognizing and studying recurrent modules, which may serve as simpler building blocks of the entire network. Our research focuses on the study of simple building blocks of the immune intercellular communication network, combining experimental approaches and mathematical modeling. ​

1. Extracellular feedback: roles of IL-2 in T-cell activation

Extracellular feedback occurs when a cell expresses receptors to a cytokine that it secretes. This is the simplest module of the intercellular cytokine communication network, and is abundantly found throughout the network. We study extracellular feedback by the cytokine IL-2 in the context of T-cell activation. IL-2 is secreted by T-cells upon stimulation of their T-cell receptor (TCR). IL-2 signaling drives expression of the alpha subunit of the IL-2 receptor (IL-2Rα), resulting in a positive feedback; and downregulation of IL-2 expression, resulting in a negative feedback. In addition, IL-2 shows a pleiotropic activity, enhancing both cell proliferation and cell death.
We study the consequences of IL-2 mediated extracellular feedback both theoretically and experimentally. A theoretical model that we developed, in collaboration with Yoni Savir and Tsvi Tlusty from the department of physics of complex systems, shows that IL-2 can result in cooperation or competition between activating T cells, depending on their relative activation strength and timing. Hence, IL-2 may serve to balance speed and accuracy of T cell responses (Savir, Waysbort et. al., 2012). In an additional work, performed in collaboration with Yuval Hart and Uri Alon from the department of molecular cell biology, we show that IL-2 pleiotropic activity can result in homeostasis of cell numbers, where the final cell concentration is independent of initial concentration over a large range (Hart et. al., 2012). We were able to demonstrate experimentally these theoretical predictions and identify cellular mechanisms that are important for maintaining homeostasis.

2. Decision making in CD4+ T-cell differentiation

Upon activation, naïve CD4+ T-cells can differentiate into one out of few possible lineages, each invoking a specific immune response. This decision is influenced by the spectrum of cytokines that the cells sense. The differentiating cells also actively alter these signals by producing relevant cytokines, thus influencing the differentiation process through extracellular feedbacks. While differentiation is usually studied under polarizing conditions, applying only one signal at a time, cells in vivo are expected to sense a more complex environment, in which they may be simultaneously exposed to combinations of cytokines. How do cells respond under such mixed conditions is not well understood.
We explore differentiation under mixed input conditions by exposing cells to varied combinations of cytokines, experimentally mapping the "decision space" of the differentiation process at the single cell level. Our data indicates that CD4 T cells can be found in a continuum of states, spanning the “canonical” differentiated states. Moreover, we show that cell state can be tuned into a specific state in this continuum by altering input signal combinations. We build a mathematical model describing the system, which can reproduce our experimental data under specific assumptions regarding the underlying molecular network structure and kinetic constants.
Our experiments provide information also on levels of heterogeneity in the population in response to mixed conditions. We find that heterogeneity in cytokine expression upon restimulation is much larger under mixed conditions than under a single cytokine input. These findings support a model by which decision under mixed conditions follows a fuzzy logic, where cells project their internal state onto the output phenotype stochastically, with probabilities which depend on the system’s input.