Here a detailed protocol to isolate and characterize bone marrow microenvironmental populations from murine models of myelodysplastic syndromes and acute myeloid leukemia is presented. This technique identifies changes in the non-hematopoietic bone marrow niche, including the endothelial and mesenchymal stromal cells, with disease progression.
The bone marrow microenvironment consists of distinct cell populations, such as mesenchymal stromal cells, endothelial cells, osteolineage cells, and fibroblasts, which provide support for hematopoietic stem cells (HSCs). In addition to supporting normal HSCs, the bone marrow microenvironment also plays a role in the development of hematopoietic stem cell disorders, such as myelodysplastic syndromes (MDS) and acute myeloid leukemia (AML). MDS-associated mutations in HSCs lead to a block in differentiation and progressive bone marrow failure, especially in the elderly. MDS can often progress to therapy-resistant AML, a disease characterized by a rapid accumulation of immature myeloid blasts. The bone marrow microenvironment is known to be altered in patients with these myeloid neoplasms. Here, a comprehensive protocol to isolate and phenotypically characterize bone marrow microenvironmental cells from murine models of myelodysplastic syndrome and acute myeloid leukemia is described. Isolating and characterizing changes in the bone marrow niche populations can help determine their role in disease initiation and progression and may lead to the development of novel therapeutics targeting cancer-promoting alterations in the bone marrow stromal populations.
The bone marrow microenvironment consists of hematopoietic cells, non-hematopoietic stromal cells, and the extracellular matrix1,2. This microenvironment can promote hematopoietic stem cell self-renewal, regulate lineage differentiation, and provide structural and mechanical support to the bone tissue1,2,3,4,5. The stromal niche includes osteolineage cells, fibroblasts, nerve cells, and endothelial cells6, while the hematopoietic niche consists of the lymphoid and myeloid populations1,2,3. In addition to supporting normal HSCs, the bone marrow microenvironment can also play a role in the development of hematopoietic stem cell disorders such as MDS and AML7,8,9,10,11. Mutations in osteolineage cells have been shown to promote the development of MDS, AML, and other myeloproliferative neoplasms10,12,13,14.
Myelodysplastic syndromes are a group of pre-leukemic disorders that arise from mutations in hematopoietic stem cells. MDS is frequently associated with a block in HSC differentiation and the production of dysplastic cells, which can often lead to bone marrow failure. MDS is the most commonly diagnosed myeloid neoplasm in the United States and is associated with a 3-year survival rate of 35%-45%15. MDS is often associated with a high risk of transformation to acute myeloid leukemia. This can be a fatal complication, as MDS-derived AML is resistant to most therapies and likely to relapse. AML that arises de novo due to translocations or mutations in hematopoietic stem and progenitors is also often resistant to standard chemotherapy16,17. Since MDS and AML are primarily diseases of the elderly, with the majority diagnosed over the age of 60 years, most patients are ineligible for curative bone marrow transplants. There is, thus, a significant need to identify novel regulators of disease progression. Since the bone marrow microenvironment can provide support for malignant cells14, defining changes in the bone marrow niche with disease progression may lead to the identification of novel therapeutics aimed at inhibiting tumor niche remodeling. There is, therefore, a significant need to identify novel regulators of disease progression. To this end, it is critical to identify and characterize changes in the bone marrow stromal cell populations that may provide support for the malignant cells.
Several murine models of AML and MDS have been generated and can be used to study changes in the bone marrow microenvironment during disease initiation and progression6,1,19,20,21,22. Here, protocols to identify changes in the bone marrow stromal cell populations using murine models of retrovirally induced AML6,20, as well as the commercially available Nup98-HoxD13 (NHD13) model of high-risk MDS to AML transformation19, are described. Mice transplanted with de novo AML cells succumb to the disease in 20-30 days6. The NHD13 mice develop cytopenias and bone marrow dysplasia around 15-20 weeks, which eventually transforms into AML, and nearly 75% of the mice succumb to the disease around 32 weeks. To analyze the murine model bone marrow microenvironment populations, bones are harvested, bone marrow and bone spicules are digested using enzymatic digestion, and the cells are then enriched for CD45-/Ter119- non-hematopoietic populations by magnetic sorting. While similar analyses have been previously described11,13,22,23,24,25, they often focus on either the bone marrow or the bone and do not incorporate cells from both sources in their analyses. The combined characterization of these populations, in conjunction with gene expression analyses, can provide a comprehensive understanding of how the cellular hematopoietic microenvironment provides support for disease initiation and progression (Figure 1). While the protocol described below focuses on retrovirally induced AML model and a genetic MDS model, these strategies can be easily adapted to study changes in the bone marrow niche of any murine model of interest.
All animal experiments were conducted in accordance with protocols approved by the University of Rochester University Committee on Animal Resources. Mice were bred and maintained in the animal care facilities at the University of Rochester. To model high-risk MDS, the commercially available NHD13 murine model19 is employed. In this model, bone marrow stromal cells are analyzed in female NHD13 mice at 8 weeks of age, before disease onset. De novo AML is generated as previously described6,11,20. The oncogenes used to induce AML, such as MLL-AF9 and NRas, are tagged with GFP or YFP, allowing for the analysis of the non-leukemic GFP- bone marrow populations using flow cytometry. In brief, 10-week-old female C57BL/6J mice are transplanted with murine GFP/YFP+ AML cells, and the bone marrow is harvested 2 weeks post-transplant. While female mice are used in this study for demonstration purposes, this protocol can be conducted with either male or female mice. It can also be carried out using either one femur or all long bones.
1. Bone marrow harvesting
NOTE: For details on the animal dissection protocol, please refer to Amend et al.26.
2. Digestion of bone marrow
3. Digestion of bone spicules
4. Staining
5. Depletion of sample by magnetic sorting
NOTE: This step is carried out using a commercially available manual magnetic separator according to the manufacturer's instructions. This step can also be performed with an automated separator (see Table of Materials).
6. Osteo-analysis/endothelial panel stain
NOTE: Compensation should be performed following standard flow cytometry protocols, including all appropriate staining and gating controls.
This article describes a flow cytometry-based method for analyzing bone marrow microenvironmental populations, such as the endothelial and mesenchymal stromal cells, from MDS and leukemia murine models (Figure 1). Figure 2 depicts the gating strategy for detection of populations of interest, beginning with the selection of cells (P1) in the digested and CD45/Ter119 depleted fraction through forward and side scatter profile. Example gating of cells in a leukemia sample are shown in P1 (Figure 2A). Singlets are selected and doublets are excluded from this analysis, P2 (Figure 2B). Figure 2C shows gates to select propidium iodine negative live cells, P3. To focus on non-hematopoietic stromal populations, cells that are CD45-/Ter119-, P4 (APC, Y-axis) vs. SSC-A are selected (Figure 2D). This initial gating strategy is common for all samples to be analyzed.
The leukemia murine model is used to illustrate gating for endothelial cells in Figures 2E–G. Figure 2E depicts CD31 positive (Pe-Cy7, Y-axis) vs. SSC-A cells in a non-cancer control mouse as well as a leukemic mouse, gated through P4. Cells positively labeled with CD31 are endothelial cells, P5. In Figure 2F, gated through P5, arteriolar endothelial cells are identified as CD45-Ter119-CD31+Sca1+, P6, and sinusoidal endothelial cells are identified as CD45-Ter119-CD31+Sca1-, P7 (BV421, Y-axis) vs. SSC-A.
Although this analysis focusses on non-leukemic bone marrow microenvironmental cells, it is also helpful to determine tumor burden. This can be done with a small fraction of undigested/non-depleted sample prior to initiating the experiment. In this experimental control (dashed lines) Figure 2G, P8 represents tumor burden in the sample and P9 represents the non-cancerous cells.
The MDS murine model is used to illustrate analysis of mesenchymal stromal cells in Figure 2H–J. Figure 2H depicts CD31 negative (Pe-Cy7, Y-axis) vs. SSC-A cells, gated through P4. Mesenchymal stromal cell populations in a wild-type control mouse and MDS mouse are shown in Figure 2I. Gated through P10, mesenchymal stromal cells are identified as CD45-Ter119-CD31-CD51+CD140a+, P113 (Pe, Y-axis; Pe-Cy5, X-axis). The experimental control (dashed lines), Figure 2J shows examples of single-color gating controls, CD51 single stain and CD140a single stain used to gate for the MSCs shown in the experimental samples in Figure 2I.
This data indicates that arteriolar endothelial cells significantly expand in the AML microenvironment, with a concomitant loss in the sinusoidal endothelial populations (Figure 2F), consistent with earlier studies using patient derived xenografts in immunodeficient mice28. It is likely that the small expansion in the mesenchymal stromal cells seen in the NHD13 mice at 8 weeks of age (Figure 2I) may increase at 16-20 weeks, when these mice start to display characteristics of MDS25. Although only the leukemia model is used to illustrate the endothelial cell population and the MDS murine model is used to illustrate the mesenchymal stromal cell populations, similar staining and gating strategies can be used to analyze the different microenvironmental populations in either of these models, or indeed, in any genetically engineered murine model of interest.
Figure 1: Isolation of bone marrow stromal cells. Schematic shows the process of isolating non-hematopoietic bone marrow stromal cells from control and leukemic mice. Briefly, bone spicules and bone marrow are digested separately and then pooled. The CD45-Ter119- population is enriched by magnetic sorting, and stained with antibody panels against populations of interest, such as mesenchymal stromal cells and endothelial cells. The cells are then analyzed by flow cytometry. Please click here to view a larger version of this figure.
Figure 2: Gating strategy for non-hematopoietic bone marrow and stromal cells. (A–D) Flow cytometry gating strategy used to select digested, CD45-/Ter119-, non-hematopoietic populations (P4) in both the leukemia and MDS murine model. (E) Leukemia bone marrow endothelial cells (P5, CD31+) can be sub-divided as (F) Sca1+ arteriolar (P6) or Sca1- sinusoidal endothelial cells (P7). (G) Leukemia engraftment in an AML sample where P8 represents engraftment while P9 represents non-cancer cells. Dotted lines indicate that this plot is an experimental control. (H) Gated through P4, MDS bone marrow CD31- cells, P10 (I) Mesenchymal stromal cells (CD51+/CD140a+, P11). (J) Example single color gating controls, CD51 (left) and CD140a (right) used to determine the gates for mesenchymal stromal cells in (I). Please click here to view a larger version of this figure.
Solution | Reagent | Concentration | Amount to add |
FACS buffer | HBSS | 10x | 100 mL |
EDTA | 0.5 M | 4 mL | |
Fetal Bovine Serum | – | 50 mL | |
Water | – | 848 mL | |
Bone marrow digestion mixture | HBSS | 1x | 2 mL |
DNase 1 | 1 µg/mL in 1x DPBS | 20 µL | |
Dispase II Powder | – | 4 mg | |
Collagenase Type IV | – | 2 mg | |
Bone spicule digestion mixture | DPBS | 1x | 320 mL |
Collagenase Type 1 | – | 1 g | |
Fetal Bovine Serum | – | 80 mL | |
MACs buffer | BSA | 66 g/mol | 5 g |
DPBS | 10x | 100 mL | |
EDTA | 0.5 M | 4 mL | |
Water | – | 896 L |
Table 1: Composition of solutions and buffers used in the present study.
Murine leukemia models have been extensively used to identify cell intrinsic and niche-driven signals that promote aggressive myeloid leukemia progression6,19,21. Here, a comprehensive flow cytometry-based protocol to define the cellular composition of the bone marrow microenvironment in murine models of MDS and AML is presented.
Prior to acquiring flow cytometric data from experimental samples, it is important to carefully compensate for fluorescence overlap. It is also essential to include all appropriate staining and gating controls. These steps will allow the experimenter to confirm that positive or negative antibody staining represents accurate expression of cell markers of interest and is not an artefact of fluorescence spectral overlap or autofluorescence. Although this protocol describes selected cell surface markers, the antibody panels can be expanded based on experimental need. For example, CD144 (Ve-Cadherin) can be administered in vivo before harvesting the mice and can serve as an additional specific marker of endothelial cells5,29. While the fluorophores and antibody clones indicated here can be changed, titrations should be carried out to determine their ideal dilution. To categorically define all cell populations in the bone marrow niche, single cell RNA-sequencing can be used to establish the bone marrow niche landscape during MDS/AML initiation and progression23,24,30.
It is critical to prepare the sample carefully as per the steps described in this protocol. When homogenizing the bones, it is crucial that the mortar and pestle are chilled, and all steps are carried out on ice to prevent cell death and to ensure high cell recovery. It is very important to remove all tissue surrounding the bones before homogenizing to prevent contamination of other undesired cell types. During RBC lysis and digestions, it is important to stop the enzymatic reaction with an appropriate amount of FACS Buffer and resuspend thoroughly in fresh buffer, or cells will continue to digest and eventually die.
Separating the bone and bone marrow fractions is essential since the bone spicules require a different mix of enzymes for digestion buffer and need longer time to digest. In bone digestions, the collagenase type 1 is useful for digesting collagenase fibrils which are commonly found in extracellular matrix and collagen fibers31. Additionally, some bone marrow cells located close to the endosteum will remain attached to the bone after homogenizing and are only released by enzymatic digestion. When digesting bone marrow, collagenase type IV is used to digest the basement membrane of epithelial and endothelial cells within the bone marrow32, while dispase mainly cleaves fibronectin31. The bone marrow requires less time to digest as the associations between cells and matrix are weaker. Incubating the bone marrow fraction in similar conditions can damage the populations of interest. Using two different digestion buffers significantly increases the number of stomal populations that can be detected, and thus provides a larger data set to analyze.
Most available protocols to analyze murine bone marrow microenvironmental populations use a syringe to flush the bone marrow from only the long bones13,24. The current method of crushing bones provides the ability of acquiring data from other bones such as the pelvis, significantly reduces sample preparation time, and mitigates the possibility of injuries with sharp needles. Given that the bone consists of mature osteoblasts and other cell populations that can provide support to normal and malignant hematopoietic cells, this method of including cells from the bone spicules enables a more accurate representation of the bone marrow microenvironment in the sample. While one previous study digested bone marrow and bone spicules separately11, it did not demonstrate staining of osteolineage and endothelial cells in the same sample. A second study used commercially available proprietary enzyme mixes to digest bone marrow and bone spicules23, and analyzed the data using the significantly more expensive single cell RNA-sequencing technology. This method of enriching for CD45-/Ter119- cells selects for cells of interest, and significantly reduces the time needed to acquire data on the flow cytometer. Thus, compared to other state of the art methods such as single-cell RNA sequencing, this flow cytometry-based method is more accessible, cost-effective and does not require sophisticated analysis by trained bioinformaticians23,24,30.
It is important to note that this protocol can be used to characterize the bone marrow microenvironment of not only any of the available murine models of MDS and AML but any genetic mouse model. Similar methods can also be effective in analyzing the changes in the murine bone marrow niche of patient derived xenograft (PDX) models. These methods can be useful for studies aimed at determining the mechanisms by which MDS/AML affect their bone marrow niche. Given the technical challenges associated with sampling large quantities of bone marrow from human patients, these analyses of murine models are an effective tool to further the understanding of the malignant bone marrow microenvironment and define its role in disease progression.
The authors have nothing to disclose.
We would like to thank the URMC Flow Cytometry Core. This work was supported by American Society of Hematology Scholar Award, Leukemia Research Foundation award and NIH grants R01DK133131 and R01CA266617 awarded to J.B.
1 mL pipette Tips | Genesee Scientific | 24-165RL | |
1.7 mL Microcentrifuge Tubes | AVANT | L211511-CS | |
10 µL pipette Tips | Genesee Scientific | 24-140RL | |
10 mL Individually Wrapped Sterile Serological Pipettes | Globe scientific | 1760 | |
1000 mL Vacuum Filtration Flask | NEST | 344021 | |
15 mL Centrifuge Tube | VWR | 10026-076 | |
2 mL Aspirating Pipette | NEST | 325011 | |
200 µL pipette Tips | Genesee Scientific | 24-150-RL | |
25 mL Individually Wrapped Sterile Serological Pipettes | Globe scientific | 1780 | |
5 mL Individually Wrapped Sterile Serological Pipettes | Globe scientific | 1740 | |
5 mL Polystyrene Round-Bottom Tube 12 x 75 mm style | Falcon | 352054 | |
5 mL Polystyrene Round-Bottom Tube with Cell Strainer Cap 12 x 75 mm style | Falcon | 352235 | |
50 mL Centrifuge Tube | NEST | 602052 | |
6 Well, Flat Bottom with Low Evaporation Lid | Falcon | 353046 | |
Absorbent Underpads with Waterproof Moisture Barrier | VWR | 56616-031 | |
APC MicroBeads | Miltenyi | 130-090-855 | |
autoMACS Pro Separator | Miltenyi Biotec GmBH | 4425745 | |
BD Pharmingen Purified Rat Anti-Mouse CD16/CD32 (Mouse BD Fc Block) | BD Biosciences | 553141 | 0.5 mg/mL |
Bovine Serum Albumin | Sigma-Aldrich | A7906 | 66.000 g/mol |
Brilliant Violet 421 anti-mouse Ly-6A/E (Sca-1) Antibody (D7) | Invitrogen | 404-5981 | 0.2 mg/mL |
C57BL/6J Mice | Jackson Laboratory | 664 | |
Carbon Dioxide Gas Tank | Airgas | CD50 | |
CD31 (PECAM-1) Monoclonal Antibody (390), PE-Cyanine7 | Invitrogen | 25-0311-82 | 0.2 mg/mL |
CD45 Monoclonal Antibody (30-F11), APC | Invitrogen | 17-0451-82 | 0.2 mg/mL |
Cell Strainer 70 µm Nylon | Falcon | 352350 | |
Cole-Parmer Essentials Mortar and Pestle; Agate, 125 mL | Cole-Parmer | EW-63100-62 | |
Collagenase from Clostridium histolyticum | Sigma-Aldrich | C5138-500MG | |
Collagenase Type I | STEMCELL | 7415 | |
Corning Mini Centrifuge | CORNING | 6770 | |
Corning Stripettor Ultra Pipet Controller | Corning | 4099 | |
Deoxyribonuclease I from bovine pancreas | Sigma-Aldrich | D4513 | |
Dispase II, powder | Gibco | 117105041 | |
DPBS 10x | gibco | 14200-075 | |
eBioscience 1x RBC Lysis Buffer | Invitrogen | 00-4333-57 | |
Ethanol absolute, KOPTEC, meets analytical specification of BP, Ph. Eur., USP (200 Proof) | VWR | 89125-174 | |
Fine scissors – sharp | Fine Science Tools | 14061-10 | |
Foundation B Fetal Bovine Serum | GeminiBio | 900-208 | |
Gilson PIPETMAN L Pipette Starter Kits | FisherScientific | F167370G | |
Graefe Forceps | Fine Science Tools | 11051-10 | |
Hank's Balanced Salt Solution (HBSS) 10x | gibco | 14185-052 | |
Hemocytometer | Fisher | 02-671-10 | |
Incubator | BINDER | C150-UL | |
Kimwipes | KIMTECH | K222101 | |
LABGARD Class II, Type A2 Biological Safety Cabinet | Nuaire | NU-425-400 | |
LD Columns | Miltenyi Biotec GmBH | 130-042-901 | |
LSE Vortex Mixer | CORNING | 6775 | |
LSRII/Fortessa/Symphony A1 | Becton, Dickinson and Company | 647800L6 | |
MACS MULTI STAND | Miltenyi Biotec GmBH | 130-042-303 | |
MACsmix Tube Rotator | Miltenyi Biotec GmBH | 130-090-753 | |
mIgG | Millipore-Sigma | 18765-10mg | 2 mg/mL |
Nup98-HoxD13 (NHD13) Mice | Jackson Laboratory | 010505 | |
PE anti-mouse CD51 Antibody (RMV-7) | Biolegend | 104106 | 0.2 mg/mL |
PE/Cyanine5 anti-mouse CD140a Antibody (RUO) | Biolegend | 135920 | 0.2 mg/mL |
Penicillin-Streptomycin | Gibco | 15140122 | 10,000 U/mL |
Plastipak 3 mL Syringe | Becton, Dickinson and Company | 309657 | |
Propidium Iodide – 1.0 mg/mL Solution in Water | ThermoFisher Scientific | P3566 | |
QuadroMACS Separator | Miltenyi Biotec GmBH | 130-090-976 | |
Sorvall X Pro / ST Plus Series Centrifuge | Thermo Scientific | 75009521 | |
TER-119 Monoclonal Antibody (TER-119), APC | Invitrogen | 17-5921-82 | 0.2 mg/mL |
Trypan Blue Solution 0.4% | Gibco | 15-250-061 | |
Ultrapure 0.5 M EDTA, pH 8.0 | Invitrogen | 15575-038 |