Transcriptional and epigenetic programs of in vitro-expanded HSPCs and their T cell progenitors progeny

Hematopoietic stem and progenitor cells (HSPCs) in the bone marrow are the ultimate sources of all hematopoietic lineage cells, including T cells. However, gene expression programs and chromatin dynamics that guide the bone marrow progenitor cells to enter the T-development programs are not fully understood due to limited cell numbers and population heterogeneity. By exploiting the in vitro HSPC expansion approach, which effectively expands HSPCs with high T cell potentials, we monitored the gene expression programs and chromatin accessibility changes underlying the transition from the bone marrow progenitor stages to early T cell development stages. Notably, expanded HSPCs displayed strikingly similar chromatin accessibility profiles with early-stage T cell progenitors, representing their shared hematopoietic chromatin landscapes. However, a select set of genomic regions and target genes were specifically regulated as cells first received the strong Notch signaling and engaged with T-development conditions. These events included a robust chromatin opening and transcriptional activation of the T cell receptor (TCR)-C beta locus. In addition, well-known stem and progenitor-associated transcription factors were sharply repressed, often concerted with broad chromatin accessibility losses at those loci. These gene regulation targets were not an artifact of in vitro expanded HSPC-derived pro-T cells. The progeny of expanded HSPCs and freshly isolated HSPCs share the same T-lineage developmental trajectory at the single-cell transcriptome level, and their gene expression programs were highly similar. However, expanded HSPC-derived pro-T cells showed temporal differences in early T-development speed and progressed through pre-commitment stages slowly. From cytokine and chemokine screening, we found that a brief Flt3L pre-treatment during the 4-5 days of the expansion period could moderately accelerate the T-development kinetics of expanded HSPCs. Thus, we compared the chromatin accessibility programs, H3K27ac and H3K27me3 histone marks, and gene expression programs upon Flt3L treatment. Although chromatin state and transcriptional features were mostly not altered by Flt3L treatment at the bulk population levels, scRNA-seq results showed that a set of activation and stress-response genes were upregulated upon Flt3L stimulation. However, Flt3L-primed HSPCs developed through a normal T cell pathway. Together, these datasets (1) provide comprehensive gene expression and chromatin accessibility profiles of expanded HSPCs and their progeny pro-T cells, (2) reveal molecular events accompanied by bone marrow progenitor cells transition to the T cell program, and (3) suggest a slight modification of the expansion protocol for the use of T cell biology studies by adding acute Flt3L treatment. Overall design: In vitro HSPC expansion and early T cell development for bulk RNA-seq, ATAC-seq, and CUT&RUN (C&R): To expand HSPCs, lineage (CD3e, TCRb, TCRgd, CD19, NK1.1, CD49b, CD11b, CD11c, Ly6G/C)-negative- Kit+ Sca1+ (“LSK”) EPCR+ CD150+ bone-marrow progenitor cells were obtained from 8-10 weeks old C57BL/6J wild-type (WT) animals or Bcl11b-mCitrine reporter mice by FACS sorting. Then, the sorted EPCR+ CD150+ LSK (“ELSK”) cells were expanded in HSPC expansion medium (F12 DMEM, 1 mg/mL of PVA, 1× Insulin–transferrin–selenium–ethanolamine, 10 mM HEPES (pH 7.3-7.5) , 1× Penicillin–streptomycin–glutamine) supplemented with 100 ng/mL of mouse thrombopoietin (TPO) and 10 ng/mL of mouse stem cell factor (SCF) for 2-4 weeks. Expanded HSPCs were enriched for the LSK population by Sca1-positive selection using PE-conjugated anti-Sca1 and anti-PE microbeads (Miltenyi Biotec), or else sorted for ELSK phenotype. To compare Flt3 ligand (Flt3L)-treatment effects, ELSK-sorted expanded HSPCs were treated with 10 ng/mL of human Flt3L along with 100 ng/mL TPO and 10 ng/mL SCF for 4 days prior to harvest. The unprimed HSPCs were cultured for 4-5 days with 100 ng/mL TPO and 10 ng/mL SCF in the absence of Flt3L. In vitro early T cell development was achieved by co-culturing Sca1+ expanded HSPCs with OP9-DLL1 stroma cells with human Flt3L and human IL-7 (10 ng/mL each from day 0-day 5, 5 ng/mL each from day 5-day 7, 1 ng/mL each after day 7) in OP9 medium OP9 medium (alphaMEM, 20% FBS, 1×PSQ, and 50 microM beta-mercaptoethanol). To distinguish DN2 pro-T cells before and after commitment, we used expHSPC from Bcl11b-mCitrine transgenic mice, which activate Bcl11b-mCitrine during commitment. Thus, ETP (Lin-, cKithi, CD25-, Bcl11b-mCitrine-), Bcl11b- DN2a (Lin-, cKithi, CD25+, Bcl11b-mCitrine-), and Bcl11b+DN2a/DN2b (Lin-, cKithi/int, CD25+, Bcl11b-mCitrine+) populations were FACS sorted on day 5 or day 9. The onset of Bcl11b expression is correlated with T-cell lineage commitment (Kueh et al. 2016 Nat Immunol PMID: 27376470). To measure immediate gene expression changes of expanded HSPCs in response to OP9-DLL1, d10-d14 expanded HSPCs were sorted for LSK. Then, they were subjected to unprimed or Flt3L-primed conditions for 4 days, and re-sorted for LSK phenotype (“d0 expHSPC LSK cells”. Each input population was co-cultured with OP9-DLL1 for 1 day or for 2 days. After 1 day or 2 days of OP9-DLL1 co-culture, expHSPC progeny cells were sorted for live Lin-negative markers (e.g., “d1 unprimed expanded HSPC progeny” or “d2 unprimed expanded HSPC progeny”) for bulk RNA_seq. For bulk RNA-seq, total RNA was isolated from 25,000 – 300,000 cells. ATAC-seq was performed using 50,000 cells. To measure H3K27ac and H3K27me3 marks, 250,000 cells were subjected to C&R. For bulk RNA-seq, total RNA was isolated from 50,000 – 300,000 cells. ATAC-seq was performed using 50,000 cells. To measure H3K27ac and H3K27me3 marks, 250,000 cells were subjected to C&R. Single-cell RNAseq: ELSK-sorted expanded HSPCs were cultured under “unpriming” or “Flt3L-priming” conditions for 4 days. After 4 days of treatment, unprimed expanded HSPCs were sorted for ELSK (Exp1) or LSK (Exp2). Flt3L-primed expanded HSPCs were sorted for LSK phenotype (Exp1 and Exp2). Then, these cells were co-cultured with OP9-Dll1 cells for 5 days (Exp1 and Exp2) or 9 days (Exp2). To compare in vitro T-development pathways that fresh bone marrow progenitor cells utilize, ELSK cells were sorted from freshly isolated bone marrow progenitors and co-cultured with OP9-Dll1 cells for 5 days or 9 days (Exp1 and Exp2). To measure transcriptional programs of input populations, ELSK cells (Exp1) or LSK cells (Exp2) were sorted from unprimed expanded HSPCs, and LSK cells were sorted from Flt3L-primed expanded HSPCs (Exp1 and Exp2). Similarly, ELSK cells (Exp1 and Exp2) and LSK cells (Exp2) were isolated from fresh bone marrow progenitor cells. Experiments were designed with staggered starting times for the cultures, in a way that all of these cells within the same experiment could be harvested on the same date. Each population was stained with unique hashtag oligos (HTOs) and loaded to the same 10x Genomics Chromium Controller. Two independent experiments (Exp1 and Exp2) were conducted.