Inflammation perturbs hematopoiesis by remodeling specific compartments of the bone marrow niche

Hematopoietic stem and progenitor cells (HSPC) are regulated by interactions with stromal cells in the bone marrow (BM) cavity, which can be segregated into two spatially defined central marrow (CM) and endosteal (Endo) compartments. However, the importance of this spatial compartmentalization for BM responses to inflammation and neoplasia remains largely unknown. Here, we extensively validate a combination of scRNA-seq profiling and matching flow cytometry isolation that reproducibly identifies 7 key CM and Endo populations across mouse strains and accurately surveys both niche locations. We demonstrate that different perturbations exert specific effects on different compartments, with type I interferon responses causing CM mesenchymal stromal cells to adopt an inflammatory phenotype associated with overproduction of chemokines modulating local monocyte dynamics in the surrounding microenvironment. Our results provide a comprehensive method for molecular and functional stromal characterization and highlight the importance of altered stomal cell activity in regulating hematopoietic responses to inflammatory challenges. Overall design: Combined stromal and hematopoietic cells isolation Mice were euthanized in a rising concentration of carbon dioxide, followed by cervical dislocation. For each mouse, both femurs, tibiae, hemipelves, and humeri were dissected and thoroughly cleaned using KimWipes (Kimtech) and were used for combined isolation of stromal and hematopoietic cells. To isolate central marrow (CM) stromal cells, the proximal and distal epiphyses were removed from 1 femur. Intact marrow plugs were flushed with Hank's balanced saline solution (HBSS) without calcium or magnesium into 5 ml polypropylene tubes by inserting a 3 ml syringe with 22G needle into the distal end of the femur. The plugs were digested with 1 ml of a solution of 3 mg/ml type I collagenase (Worthington) dissolved in HBSS for 10 min at 37 ?C with 110 rpm shaking. The tube was then vortexed briefly and the supernatant was removed into a new tube through a 100 ?m mesh, taking care not to disturb the marrow plug. A further 1 ml of 3 mg/ml type I collagenase solution was then added to the marrow plug and the digestion repeated. After the second incubation, the plug was dissociated by pipetting up and down with a P1000 pipette before passing the cell suspension through a 100 ?m mesh into the tube containing the previous digestion supernatant. To isolate endosteal (Endo) stromal cells and hematopoietic cells, the flushed femur was combined with the other bones, gently crushed up to 10 times using a mortar and pestle, and thoroughly washed with 10 ml HBSS until all the non-adherent BM cells were removed and collected in a separate 15 ml polypropylene tube. The bone chips were then digested in a 15 ml polypropylene tube with 3 ml of 3 mg/ml type I collagenase solution for 1 hour at 37 ?C with 110 rpm shaking. After digestion, the tube was vortexed briefly and the cell suspension was filtered through a 100 ?m mesh into a new tube. The bone chips were then washed with HBSS, and the washing was collected into the same tube as the digestate. For cell suspensions acquired for both stromal compartments, red blood cells were removed by adding 1 ml ACK lysis buffer (150 mM NH4Cl and 10 mM KHCO3) and incubated on ice for 3 minutes before washing with HBSS containing 4% fetal bovine serum (HI FBS, Gibco). During the optimization of the protocol outlined above, the digestion with 3 mg/mL type I collagenase was performed in parallel to and compared with 3 mg/mL type I collagenase + 4 mg/mL type II Dispase (Roche), 250 µg/mL Liberase DL (Roche) + 200 U/mL DNAse I (Sigma), and mechanical dissociation with P1000 pipette. The BM cells collected from crushed bones were resuspended in HBSS with 2% FBS and RBCs were removed by lysis with ACK buffer. The BM cell suspension was further purified on a density gradient (Histopaque 1119, Sigma-Aldrich), by layering 2 ml of the cell suspension under 2 ml of Histopaque solution. For profiling of CM plug and Endo marrow, only femurs were used and were flushed once with a 3 ml syringe and 21G needle to collect BM plugs, while the flushed bones were crushed to collect the remaining Endo-associated BM cells. No ACK RBC lysis was performed for these analysis. Both stromal and hematopoietic cells were finally counted using a Vicell automated cell counter (Beckman Coulter). RNA sequencing - Droplet-based scRNA-seq for stroma and hematopoietic analyses For Endo and CM stromal cell preparations, 30,000–50,000 Ter119-/CD45- cells were sorted on purity mode into 1.5 ml tubes containing 300 µl of 50% FBS and 50% HBSS using a BD FACS Aria II instrument (gating shown in Supplementary Fig. 2a, 3b). For whole BM cells, 40,000 live (propidium iodide negative) cells from n=2 mice were sorted into similar tubes from PBS or pIC-injected mice. Cells were rested for 1 hour on ice then pelleted at 350 x g for 5 minutes at 4°C before the supernatant was removed down to a volume of 40 µl. For sequencing of defined HSPCs, 8,000 cells of 10 different populations (HSC, stHSC, MPP2, Fc?R- MPP3, Fc?R+ MPP3, MPP4, GMP, MkP, CLP, CFU-E, see Supplementary Fig. 1 for sorting scheme) were sorted separately into 1.5 ml tubes. In this experiment, there were 2 biological replicates with BM from n=5 mice, and 2 technical replicates were generated per sample. Sorted cell populations were labelled with 0.05 µg of oligo-hashed antibodies specific for MHC class I and CD45 (BioLegend, TotalSeq B1-10) for 30 minutes on ice before washing three times with HBSS containing 2% FBS. All 10 populations were then pooled for subsequent partitioning. Cells were rested for 1 hour on ice then pelleted at 350 x g for 5 minutes at 4°C before the supernatant was removed down to a volume of 40 µl. For sequencing of purified stromal populations, samples of central marrow or endosteal stromal cells pooled from n=7 mice were labelled with oligo-hashed antibodies at a concentration of 0.375 µg of antibody per 106 cells. These antibodies were mixed with the fluorophore-conjugated surface antibodies, and cells were then washed three times with HBSS and 4% FCS. Specific stromal cell populations were then sorted by FACS (gating scheme shown in Fig. 1e), and mixed according to the following scheme to produce two combined samples, each containing a total of 25-40,000 cells: Sample type Oligo hash antibody Sample 1 (cells) Sample 2 (cells) CM TotalSeq B1 MSC-L (6899) SEC (2228) CM TotalSeq B2 mMPr (2775) SEC (2097) CM TotalSeq B3 mMPr (2406) SEC (2529) CM TotalSeq B4 MPr (847) SEC (1041) CM TotalSeq B5 MPr (802) SEC (1353) CM TotalSeq B6 MSC-L (7642) AEC (438) CM TotalSeq B7 - AEC (496) Endo TotalSeq B8 MPr (10048) MSC-S (8212) Endo TotalSeq B9 AEC (5046) MSC-L (2319) Endo TotalSeq B10 SEC (3537) mMPr (6823) Cells were rested for 1 hour on ice then pelleted at 350 x g for 5 minutes at 4°C before the supernatant was removed down to a volume of 40 µl. GEM generation and 3' RNA library preparation was performed according to 10X Genomics protocol CG000315 Rev E, targeting 5000 cell data recovery. RNA libraries were pooled 1:1:1:etc, sequenced on an Illumina NovaSeq 5000, and alignment was performed using Cellranger (v.7.0.1) to mouse genome mm10. For hashing samples, hashing libraries were pooled with RNA libraries at a ratio of 1:4. Library concentrations and fragment sizes were evaluated using Qubit dsDNA HS assay kit (ThermoFisher Scientific) and TapeStation D5000 DNA ScreenTape analysis (Agilent). RNA sequencing - Plate-based SMART-seq for stromal cells and MSC-L For Smart-seq analyses of Endo (AEC, MSC-S, mMPr, MPr) and CM (SEC, MSC-L) stromal cells isolated from young WT mice, we utilized our own published data (Mitchell et al., 2023) that was processed using the same analytical pipeline. For Smart-seq of iMSC-L from pIC-injected young WT mice and aged 24-month-old WT mice, we utilized the SMART-Seq Single Cell PLUS kit (96 reactions, Takara Bio). Single iMSC-L were sorted into 96 well plates (see Supplementary Fig. 3c for gating strategy) containing 12.5 µl of CDS sorting solution provided in the kit. Preparation and amplification of cDNA and single cell libraries were completed according to the manufacturer's instructions. Single cell libraries that passed quality control (n=43 from pIC-injected mice and n=30 from aged mice) were pooled at a 1:1:1etc ratio before sequencing on an Illumina NextSeq 500/550 instrument using a version 2 kit (Illumina). Fastq files were generated for each library in Illumina BaseSpace.