LSK immunophenotype hematopoietic stem cells in Dapp1 knockout mice

To investigate the impact of perturbing key differentiation driver genes on hematopoietic stem cell fate decision, we employed a Dapp1 gene knockout approach. This allowed us to observe changes in the subpopulations of hematopoietic stem cells at the LSK level in mice. Additionally, we assessed alterations in their fate decision and lineage differentiation tendencies We conducted an in-depth single-cell RNA sequencing study using the 10X Genomics platform to analyze mouse bone marrow LSK (Lin^⁻Sca-1^⁺c-Kit^⁺) cells. Starting with ethically euthanized mice, we extracted femurs and tibias to harvest the bone marrow. The marrow cells were flushed out using a syringe with PBS containing 2% FBS to create a single-cell suspension. This suspension was filtered through a 40 µm cell strainer to eliminate debris and cell clumps, ensuring a pure population of cells for analysis. To isolate our target LSK cells, we employed fluorescence-activated cell sorting (FACS), selecting for Lineage-negative, Sca-1-positive, and c-Kit-positive cells. We assessed the viability of these cells using trypan blue staining and adjusted the concentration to approximately 1,000 cells/µL, optimizing conditions for the 10X Genomics Chromium Controller. We then loaded the purified single-cell suspension, along with the necessary reagents and gel beads, into a 10X Genomics Single Cell 3' Chip. Running the chip in the Chromium Controller, we encapsulated individual cells into Gel Bead-In-Emulsions (GEMs), enabling the barcoding of mRNA from each cell during reverse transcription within the GEMs. Following emulsion breakage, we purified the barcoded cDNA and performed PCR amplification. The amplified cDNA was evaluated for quality and quantity using an Agilent Bioanalyzer to ensure downstream sequencing success. Constructing sequencing libraries from the high-quality cDNA, we followed the 10X Genomics protocol meticulously. These libraries were then sequenced on an Illumina NovaSeq platform, adhering to the recommended read lengths and depths to achieve comprehensive single-cell transcriptional coverage. The sequencing data generated were processed using the 10X Genomics Cell Ranger pipeline, which aligned reads and produced a gene expression matrix for each cell. For data analysis, we utilized various bioinformatics tools, prominently featuring our novel scRL (single-cell Reinforcement Learning) framework. This advanced analytical approach allowed us to identify distinct cell populations and gene expression patterns with high precision. Most critically, employing scRL enabled us to obtain quantitative indicators of the strength of cell lineage differentiation fate decisions—a level of insight not achievable with existing methods.