Transcriptional programs of in vitro-expanded HSPCs with acute disruption of Lmo2 before and after co-culture with OP9-DLL1

Though Lmo2 is a well-established member of the hematopoietic stem/progenitor program, its function in normal T cell development has remained unclear and difficult to study. Lmo2's quick downregulation in thymic-seeding progenitor cells makes it particularly challenging to precisely perturb for mechanistic analyses; however, a recent in vitro HSPC expansion method provides a platform for probing a greater number of progenitor cells during an extended window of time through the T cell DN stages. Consequently, we performed loss-of-function experiments in expanded HSPCs and analyzed subsequent gene expression with bulk RNA-seq and single cell RNA-seq to reveal Lmo2's previously unknown role in early T cell development. Remarkably, premature loss of Lmo2 in expanded HSPCs results in the specific upregulation of the T cell program. Bulk RNA-seq at day 5 of OP9-Dll1 co-culture reveals that the DN1 (CD44+CD25-) population derived from expanded HSPCs with the Lmo2 knockout (KO) have higher expression of T cell-related genes, such as the TCR signaling genes Lat, Itk, Lck, Zap70, Cd3e, and Cd3g, and higher transcriptional activation of the T cell receptor (TCR)-C beta locus when compared to DN1 controls. These findings are supported by single cell RNA-seq (scRNA-seq), where Lmo2-KO cells have higher expression of T cell program members like Tcf7 and higher transcriptional activation of the (TCR)-C gamma locus by the day 3 timepoint. Interestingly, before entering a Notch environment, there is minimal difference between Lmo2 KO and their control counterparts by scRNA-seq. Furthermore, the d3 and d5 datasets do not strongly indicate that Lmo2 KO causes a downregulation of the stem/progenitor program. The scRNA-seq dataset also introduces the impact of Flt3L pre-treatment for 4 days of expansion (Flt3L-priming) before seeding onto OP9-DLL1 co-cultures. Differences due to Flt3L-priming can only be seen in a subpopulation of samples before entering a Notch signaling environment; by the day 3 timepoint, cells are virtually indistinguishable by priming, suggesting that developmental acceleration caused by Flt3L-priming is distinct from Lmo2-related mechanisms. Taken together, these datasets provide gene expression profiles of control-derived and Lmo2 KO-derived DN populations to present Lmo2 as an important controller of early T cell developmental kinetics. Overall design: In vitro HSPC expansion, retroviral infection and early T cell development: Lineage negative (lin-) bone marrow from 8-10 weeks old Cas9 x Bcl2 mice was harvested and obtained via lineage depletion with biotinylated antibodies (TCRb, TCRgd, CD19, NK1.1, CD49b, CD11b, CD11c, Ly6G/C), streptavidin microbeads (Miltenyi Biotec) and MACS LS columns (Miltenyi Biotec). These lin- 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 10-14 days. Expanded HSPCs were enriched for the LSK population by Sca1-positive selection using PE-conjugated anti-Sca1 and anti-PE microbeads (Miltenyi Biotec) (bulk RNA-seq), or else sorted for ELSK phenotype (scRNA-seq). 3 different sgRNA expressing vectors against Lmo2 were created by inserting a 19-mer sgRNA, designed using the CHOPCHOP web tool, into the empty E42 control vector with an mTurqoise2 reporter (Addgene #189799). These 3 sgRNA expressing vectors were pooled for packaging in Phoenix-Eco cells using Fugene 6 Transfection Reagent (Promega). Retroviral supernatant was then harvested and added to a non-tissue culture plate coated with 50 µg/mL RetroNectin (Takara bio) (1×PBS at 4°C overnight and then removed of unbound RetroNectin). After centrifuge at 2000xg at 32°C for 2 hours, thawed expanded HSPCs were incubated in HSPC expansion medium on the retroviral vector-bound plate at 37°C overnight in a 5% CO2 environment. Infection with Lmo2 KO vectors was compared to infection with control vectors to ensure the appropriate baseline was used for analysis. Infected cells were harvested after overnight incubation and seeded onto OP9-Dll1 stromal cells in OP9 medium (alphaMEM, 20% FBS, 1×PSQ, and 50 microM beta-mercaptoethanol) supplemented with 10 ng/mL human Flt3L and 10 ng/mL human IL-7. Bulk RNA-seq: Expanded HSPCs after Sca1-positive selection were frozen in Bam Banker solution (FUJIFILM) and later thawed in the same HSPC expansion media conditions for 4-5 days before retroviral infection with control or Lmo2 KO vectors. At day 5 of co-culture, cells were harvested and FACS sorted for Live Lin- CD45+ infection+ (mTurqoise2+) DN populations: DN1 (CD25-) and DN2 (CD25-Thy1+). Total RNA was isolated from 80,000 – 300,000 cells and processed for bulk RNA-sequencing. Two independent experiments were conducted (Exp1 and Exp2). Single Cell RNA-seq: After 10-14 days of expansion, HSPCs were ELSK-sorted and then placed in "unpriming" or "Flt3L-priming" conditions for 4 days total, where at day 2 of priming they underwent retroviral infection with control or Lmo2 KO vectors. Two timepoints were collected: before seeding cells onto OP9-DLL1 (preNotch: LiveLin-infection+) and after 3 days of co-culture on OP9-DLL1 (d3: LiveLin-CD45+infection+), making 8 conditions total. Cell from each of the 8 conditions were stained with unique hashtag oligos (HTOs) and harvested on the same day, where approximately 20,000 cells total (2,500 cells per condition) were loaded onto the same 10x Genomics Chromium Controller.