Decoding Regulatory T Cell Transcription Factor Networks: From Identity to Diversity

Abstract

Foxp3+CD4+ regulatory T cells (Tregs) are dominant controllers of immunological and organismal homeostasis. Tregs play diverse functional roles in lymphoid tissues, where they can adopt specialized phenotypes in response to varied immunological stimuli, and in non-lymphoid tissues, where they promote tissue homeostasis and repair. Reflecting their diverse functions, Tregs can adopt remarkably heterogenous molecular programs. While previous work, discussed in Chapter 1, had identified individual transcription factors (TFs) whose expression, along with that of the Treg lineage-defining TF FoxP3, enabled differentiation into specialized sub-phenotypes, how the many TFs expressed in Tregs were systematically organized to determine Treg identity and diversity remained unclear. To address this question, in Chapter 2, we performed single-cell Assay for Transposase-Accessible Chromatin using sequencing (scATAC-seq) of splenic Tregs to relate the activity of open chromatin regions (OCRs), and the TFs that modulate them, to variation in Treg states. As opposed to previous “one TF-one state” models which placed the control of each Treg subpopulation under the purview of individual context-specific master TFs, we instead found Treg chromatin programs to be shaped by interwoven, combinatorial TF inputs. We parsed this combinatorial complexity by combining machine learning, natural genetic variation, and Treg-specific TF knockouts to resolve the architecture of the Treg TF network. The resulting network captured chromatin changes across tissues, parsed heterogenous responses to acute immunological stimuli (IL2), and enabled the discovery and validation of a novel factor (Smarcc1) regulating Treg-state-specific chromatin accessibility. FoxP3 had a profound impact on this network, with either repressive and activating components across distinct chromatin programs and partner TFs. We also discovered that one population of Tregs, Rorγ+ Tregs, was independent of FoxP3 for its differentiation and homeostasis. Thus, dissecting the structure of the Treg TF network showed that Treg diversity arises not from the action of individual master TFs but rather graded multi-TF inputs with variable dependence on FoxP3. In Chapter 3, we focused our analysis on TF control of Treg diversity among colonic Treg populations. Several subsets of Tregs have been identified in the colon based on the expression of defining TFs (e.g., Helios, Rorγ, cMaf, Gata3). However, the organization of TF control and relationships between these subsets was previously unknown. Using a battery of immunological, genomic, and microbiologic assays on Tregs from mice carrying Treg-specific deletions of each of these four TFs, we uncovered the varied regulatory strategies and related homeostatic setpoints controlling each of these populations. Finally, we examined changes in Treg regulatory networks in disease-relevant contexts. In Chapter 4, we used a panel of mice carrying Foxp3 mutations found in human patients with Immune dysregulation, polyendocrinopathy, enteropathy, X-linked (IPEX) syndrome. Each mutation elicited different inflammatory phenotypes and, using bulk and single-cell genomic assays, we identified differing mechanisms for mutations involving Forkhead versus non-Forkhead FoxP3 domains. These results illustrated the unique molecular functions of each FoxP3 domain, which may partially explain disease heterogeneity. In Chapter 5, we associated alterations in human Treg phenotypes with severity of viral disease manifestations in samples from human COVID-19 patients. We found an increase in both Treg proportions and in the levels of FoxP3 protein among Tregs from patients with severe disease. This corresponded to increased expression of effector transcripts and overlapped strikingly with a signature of tumor-infiltrating Tregs. We screened and validated several candidate agents that reproduced this signature in vitro. Thus, this analysis highlighted a specific example of human disease-associated alterations in the Treg regulatory network. In summary, this thesis dissects the regulatory organization of Treg identity, diversity, and function, setting the stage for targeted Treg manipulation and engineering.