The chromatin landscape and transcription factors in T cell programming

Abstract

T cell development from multipotent progenitors to specialized effector subsets of mature T cells is guided by the iterative action of transcription factors. At each stage, transcription factors interact not only with an existing landscape of histone modifications and nucleosome packing, but also with other bound factors, while they modify the landscape for later-arriving factors in ways that fundamentally affect the control of gene expression. This review covers insights from genome-wide analyses of transcription factor binding and resulting chromatin conformation changes that reveal roles of cytokine signaling in effector T cell programming, the ways in which one factor can completely transform the impacts of previously bound factors, and the ways in which the baseline chromatin landscape is established during early T cell lineage commitment.

Keywords: CD4(+) T cell subsets; Cis-regulatory element; T cell development; genomics; histone modification.

Figure 1. Decision points and transcription factors in T cell development

A simplified schematic of T-cell development is shown with the main stages of development within the thymus labeled, key checkpoints and choice points indicated in red labels, and selected transcription factors indicated at major points when they must act. The figure shows diagrammatically which events occur in the thymus and which ones occur either prethymically or in the periphery after T-cell maturation. Major decisions and stages that depend on GATA-3, PU.1, RORγt, T-bet, Foxp3, and multiple STAT factors are indicated. Runx, Ets, and bHLH family factors have continuing roles throughout these events, although the precise family members involved and their roles can change. T-lineage entry: migration of cells to the thymic microenvironment, which is rich in Notch-activating molecules. Commitment: relinquishment of potential to give rise to anything except T cells. β-selection, positive selection: two stages of TCR-dependent developmental checkpoint where cells with failed TCR status will die. Also shown are the stages when continuing Notch pathway signaling in the thymus is critical for T-cell specification: these are stages when the cells are “double negative”, i.e. lacking CD4 and CD8 expression. The stages within this immature series are ETP (Early T-cell Precursor), c-Kit⁺⁺ CD44⁺ CD25⁻; DN2a, c-Kit⁺⁺ CD44⁺ CD25⁺; DN2b, c-Kit^{+(intermediate)} CD44⁺ CD25⁺; DN3a, c-Kit⁻ CD44⁻ CD25⁺. TCR gene rearrangement does not result in expression of a TCRβ chain until the DN3a stage, but this triggers β-selection. The cells quickly downregulate CD25 and upregulate CD8 and CD4 through the series of stages shown as DN3b, DN4, ISP in transition to “DP”, or CD4⁺ CD8⁺ double-positive. These cells continue TCR gene rearrangement until they finally express TCRαβ. They are subject to stringent TCR-based MHC restriction and selection and only a minority are allowed to mature as CD4 SP or CD8 SP cells (Positive selection, CD4/CD8 choice). For simplicity, alternative intrathymic fates such as TCRγδ cells, invariant NKT cells (iNKT) or natural Tregs (natural Tregs) are not shown. Four major types of effector subsets of CD4⁺ cells, which develop from mature CD4 SP cells in the periphery, are shown, with some of the transcription factors that play major roles in their development. Despite the clear distinctions between the regulatory programs of these effector subsets, they remain conditionally plastic depending on environmental signals. Gray dashed-line arrows indicate several well-established options for interconversion. This feature contrasts the effector cell lineage decisions with the decision to commit to T-cell fate, which is irreversible.

Figure 2. Multiple tiers of transcription factor action generate an active enhancer

The figure shows three kinds of transcription factor occupancy that can have distinct effects on the activity of an enhancer element. The steps in converting a closed, silent cis-regulatory region to an active enhancer are shown for an element in which three factors (or groups of factors) act in sequence. Only the Tier A factors are capable of recognizing their sites in the DNA from a compact chromatin configuration. Whether or not the Tier A factors themselves recruit the enzymes that mark local histones as “accessible”, however, they make it easier for Tier B factors to bind. Tier B factors are those that elicit activating histone marks and displace nucleosomes to create a DNase-accessible site. Note that the effect of Tier B occupancy is to create a small “volcano”-like shape of histone modification consisting of a wide, hollowed-out H3K4me1 peak with a narrow, cratered peak of H3K4me2 in its center. Still, despite this modification, the cis-regulatory element may not be active yet. Then, as Tier C factors arrive and bind via DNA and/or protein-protein interactions, a quorum is finally complete for the recruitment of the histone acetyltransferase EP300. Tier C factors may bind DNA directly or, as in the case of Foxp3, can determine the functional output of an enhancer even without extensive DNA contact. The large EP300 protein catalyzes modification of proximal histones to H3K27Ac forms and completes the activation of the cis-regulatory element. Whether or not the EP300 remains associated at the site, these fully activated, occupied elements are now competent to interact with the appropriate transcription start site to enhance transcription.