Keith Siklenka

/ Summary: A wealth of noncoding regulatory elements has been described across mammalian cell types, yet determining their functional role remains a challenge. Regulatory control of gene expression is critical during active processes such as the adaptive immune response. Upon antigen presentation, a naive CD4+ T cell undergoes major transcriptional and structural reorganization necessary for establishment of subset identity and immune function. In this study, we systematically measure the regulatory potential of candidate regulatory elements associated with open chromatin across five mouse CD4+ T cell subsets. Using ATAC-STARR-seq, we found that approximately one quarter of open chromatin regions demonstrate regulatory activity. Most exhibit shared functional potential across subsets, though we identify enhancers with activity that is restricted to specific cellular contexts. To distinguish regulatory potential from endogenous function, we performed CRISPR-based epigenome editing screens at noncoding regions of Th17 cells and identified a set of core elements essential for subset polarization. Integrating Region Capture Micro-C, we resolved precise 3D chromatin topologies that explain functional regulatory networks via physical contacts. We characterize examples of active regulatory hubs formed through multiple CTCF-independent interactions organized in a hierarchical architecture. Furthermore, we discover a critical Batf enhancer that operates via these contacts. Using targeted perturbations, we disrupt local chromatin topology and gene expression with profound consequence to downstream Th17 cell phenotypes. We confirm the physiological necessity of these functional enhancers in vivo , demonstrating the importance of noncoding elements for Th17 cell identity. Together, this work reveals how DNA sequence and chromatin cooperate to shape the regulatory logic of Th17 cells, with implications for cis-regulatory principles beyond the immune system.

Identifying genetic variants in noncoding DNA that impact gene expression and thereby contribute to disease risk remains a difficult but important challenge in genomic medicine. Modern reporter assays such as STARR-seq and MPRA provide an efficient and effective means of testing, in very high throughput, millions of variants captured directly from patient genomes. While these assays have previously been scaled to whole genomes and, separately, to populations, we report findings from the first whole-genome population-scale STARR-seq experiment performed on 100 individuals. In order to achieve that scale we devised a novel experimental design that partitions samples into pools so as to increase allele frequencies within pools and thereby reduce expected dropout and increase signal-to-noise ratio in experimental readouts. We show that this design produces more accurate estimates of variant effect sizes, and we provide a Bayesian model for robust estimation of those effect sizes that also reports full posterior distributions for assessment of confidence in estimates. Together, these methodological innovations facilitate the detection of functional regulatory variants, particularly rare variants, with much higher accuracy and at greater scale than previously possible. We demonstrate the utility of this approach on the task of functional annotation of quantitative trait loci such as eQTLs and caQTLs, and show concordance with patterns of constraint in transcription factor binding profiles.

The overwhelming majority of genetic associations with complex traits and disease involve non-coding genetic variants. Those variants are substantially enriched in gene regulatory elements, indicating that allelic impacts on gene regulatory element activity is a major contributor to those associations. As a step towards fine mapping the mechanisms within those associations, several studies have identified variants associated with gene expression and chromatin accessibility within those phenotypic associations. However, connecting those associations with functional impacts on gene regulatory element activity remains a major challenge. Here, we functionally measured allele-specific regulatory element activity across five human genomes using the genome-wide reporter assay STARR-seq. We identified tens of thousands of gene regulatory elements and estimated allele effects at ∼200,000 genetic variants therein, including ∼10,000 indels in our study population. Allelic effects on regulatory element activity correspond closely with predicted impacts on transcription factor binding motifs. The measured variant effects also allow us to fine map potential causal variants within eQTLs and chromatin QTLs from the same population. Together, these results provide an initial atlas of genome-wide variant effects across the human genome and demonstrate the potential for such approaches to prioritize causal variants for future mechanistic investigation.

Glucocorticoids are a widely used, potent class of anti-inflammatory drugs that modulate the expression of hundreds of genes across the genome. Although the glucocorticoid response is primarily carried out by the glucocorticoid receptor (NR3C1, a.k.a. GR), there are many glucocorticoid receptor co-factors that are also essential to the downstream effects. To identify novel factors necessary for the glucocorticoid gene expression response, we used a genome-wide CRISPR screen in A549 lung adenocarcinoma cells. In that screen, we knocked out every gene in the human genome, and measured the effect of expression of the glucocorticoid-induced leucine zipper (GILZ), a classic glucocorticoid-response gene. We identified two chromatin remodeling proteins, SMARCA2 and BPTF, that are essential for GILZ expression. We then evaluated the genome-wide effects of SMARCA2 and BPTF on glucocorticoid-mediated gene expression. BPTF had a highly specific role in the glucocorticoid response, affecting the expression of only a handful of genes, and having virtually no effect on dexamethasone-induced changes in chromatin accessibility. However, SMARCA2 was necessary for 27% of dexamethasone-induced transcriptional changes (152 genes), and ∼7% of dexamethasone-induced changes in chromatin accessibility (586 regions of the genome). Genomic regions with SMARCA2-dependent changes in chromatin accessibility were characterized by high dexamethasone-induced regulatory activity in a massively parallel reporter assay, and dexamethasone-induced increases in transcription factor binding and chromatin states. Taken together, these data suggest that SMARCA2 is critical for chromatin remodeling at a specific set of genomic regions with high regulatory activity, which in turn drive changes in expression for many glucocorticoid-responsive genes.