Mapping genomic responses to kit receptor tyrosine kinase signaling in erythroid progenitors

Abstract Chromatin signatures at cis-regulatory elements can predict how they regulate transcription. However, accurate cell type-specific prediction of functional cis-elements remains a significant computational challenge due to the lack of context regarding the role of environmental stimuli that may alter activity. Most existing datasets use static molecular and sequence features which do not evaluate cis-element responses to extracellular cues. Identifying functional cis-elements which regulate key hematopoietic processes (e.g. BCL11A repression of gamma-globin) have guided therapies for sickle cell disease. To capture dynamic cis-elements controlled by key signaling pathways, we built a multi-factorial strategy that identifies, annotates, ranks and functionally tests candidates. We evaluated cis-element activity in response to the Kit receptor tyrosine kinase (activated by Stem Cell Factor (SCF)), given its crucial role in hematopoietic and erythroid progenitor cell (EPC) survival and lineage commitment. To investigate chromatin changes mediated by Kit/SCF, we performed ATAC-seq and RNA-seq in acutely SCF stimulated HUDEP-2 cells and mapped enhancer-promoter interactions using existing datasets. ATAC-seq Footprinting (TOBIAS) predicted an increased occupancy of 984 AP-1 sites and 397 Early Growth Response (EGR) sites, and a decreased occupancy of 487 GATA sites. Next, we identified molecular features (occupancy, accessibility) enriched at Kit-sensitive vs. Kit-insensitive cis-elements. 55 transcription factors correlated at Kit-sensitive sites. Elements containing the enriched molecular features were termed “Kit response elements”, KREs), We hypothesize that KREs are required for the transcriptional activity at linked genes in erythroid progenitors. Our in-silico model predicted Kit response activity at 88% accuracy. Among 3,379 possible KREs, the top Kit-predictive features included Inflammatory Response (JUN, JUND, ATF2, ATF3, EGR1), blood cell maintenance (RUNX1) and chromatin remodelers/transcription factors (THAP1, MXI1, PKNOX1, CTCF). EGR1 - a transcription factor which regulates proliferation and differentiation - was 37-fold upregulated after SCF treatment. Combined with ATAC-seq footprinting data, this suggested Kit transcriptional targets may be EGR1 dependent. To test this, we blocked Kit-mediated EGR1 transcriptional upregulation (8-fold, p<0.0001) using CRISPR-interference. We identified 1487 EGR1-sensitive KREs using our previously implemented in-silico approach. 5 EGR1-sensitive predictors were associated with transcriptional regulation and chromatin remodeling activities (ZNF143, REST, KLF16, CTCF, CTCFL). These findings suggested EGR1-sensitive and EGR1-insensitive networks operate downstream of Kit signaling. To distinguish EGR1-sensitive from EGR1-insensitive KREs, we compared feature occupancy and selected evolutionarily conserved regions in each category. We associated KREs with established blood cell functions – BCL11A, DUSP5 (EGR1-insensitive) and negative feedback regulators of RTK – SPRED1 signaling mediators, and immediate early response regulator - NAB2 (EGR1-sensitive) revealing an interconnected network that rewired erythropoiesis. CRISPR/Cas9-mediated deletion of KREs lowered associated transcript levels significantly (p<0.05), demonstrating that these cis-elements are required. To test the cellular consequences of crippling KRE activity in erythroid-committed primary human cells, we are using colony forming and proliferation assays. Previously unstudied KREs at evolutionary conserved sites are dynamically activated in acute anemia, so these sites may control anemia responses in acute/chronic contexts. Kit activated regions contain naturally occurring polymorphisms that are associated with erythroid traits. A KRE was identified at an intergenic region near the BCL11A gene which contains a red cell-linked SNP. Another SNP was identified in a KRE near the ZBTB7A gene which has been associated with anemia in embryonic conditions. Understanding KREs' roles at a systems-level can also provide insights regarding how chromatin occupancy and transcription reorganizes signal responses. Dysfunctional KREs may contribute to chromatin misregulation in hematologic malignancies, or chemoresistance. Many of our predicted KRE target genes have poorly described or unknown functions, which provide numerous in-roads to discover genetic, molecular, and cellular regulators of erythropoiesis.