Ex Vivo Expansion and Genetic Manipulation of Mouse Hematopoietic Stem Cells in Polyvinyl Alcohol-Based Cultures

The hematopoietic system supports a range of essential processes in mammals, from oxygen supply to fighting pathogens, through specialized blood and immune cell types. Continuous blood production (hematopoiesis) is required to support blood system homeostasis, which is sustained by hematopoietic stem and progenitor cells (HSPCs)¹. The most primitive hematopoietic cell is the hematopoietic stem cell (HSC), which has unique capacities for self-renewal and multilineage differentiation^2,3. This is a rare cell population, mainly found in the adult bone marrow⁴, where they occur at a frequency of just approximately one every 30,000 cells. HSCs are thought to support life-long hematopoiesis and help to re-establish hematopoiesis following hematological stress. These capacities also allow HSCs to stably reconstitute the entire hematopoietic system following transplantation into an irradiated recipient⁵. This represents the functional definition of an HSC and also forms the scientific basis for HSC transplantation therapy, a curative treatment for a range of blood and immune diseases⁶. For these reasons, HSCs are a major focus of experimental hematology.

Despite a large focus of research, it has remained challenging to stably expand HSCs ex vivo⁷. We recently developed the first long-term ex vivo expansion culture system for mouse HSCs⁸. The approach can expand transplantable HSCs by 234-899-fold over a 4 week culture. In comparison to alternative approaches, the major change in the protocol was the removal of serum albumin and its replacement with a synthetic polymer. Polyvinyl alcohol (PVA) was identified as an optimal polymer for the mouse HSC cultures⁸, which has now also been used to culture other hematopoietic cell types⁹. However, another polymer called Soluplus (a polyvinyl caprolactam-acetate-polyethylene glycol graft copolymer) has also recently been identified, which appears to improve clonal HSC expansion¹⁰. Prior to the use of polymers, serum albumin in the form of fetal bovine serum, bovine serum albumin fraction V, or recombinant serum albumin were used, but these had limited support for HSC expansion and only supported short-term (~1 week) ex vivo culture⁷. However, it should be noted that HSC culture protocols that retain HSCs in a quiescent state can support a longer ex vivo culture time^11,12.

In comparison with other culture methods, a major advantage of PVA-based cultures is the number of cells that can be generated and the length of time the protocol can be used to track HSCs ex vivo. This overcomes several barriers in the field of experimental hematology, such as the low numbers of HSCs isolatable per mouse (only a few thousand) and the difficulty to track HSCs over time in vivo. However, it is important to remember that these cultures stimulate HSC proliferation, while the in vivo HSC pool is predominantly quiescent at a steady state¹³. Additionally, although the cultures are selective for HSCs, additional cell types do accumulate with the cultures over time, and transplantable HSCs only represent approximately one in 34 cells after 1 month. Myeloid hematopoietic progenitor cells appear to be the major contaminating cell type in these HSC cultures⁸. Nevertheless, we can use these cultures to enrich for HSCs from heterogeneous cell populations (e.g., c-Kit⁺ bone marrow HSPCs¹⁴). It also supports transduction or electroporation of HSCs for genetic manipulation^14,15,16. To help identify HSCs from the heterogeneous cultured HSPC population, CD201 (EPCR) has recently been identified as a useful ex vivo HSC marker^10,17,18, with transplantable HSCs restricted to the CD201⁺CD150⁺c-Kit⁺Sca1⁺Lineage^- fraction.

This protocol describes methods to initiate, maintain, and assess PVA-based mouse HSC expansion cultures, as well as protocols for genetic manipulation within these cultures using electroporation or lentiviral vector transduction. These methods are expected to be useful for a range of experimental hematologists.

All animal procedures, including breeding and euthanasia, must be performed within institutional and national guidelines. The experiments detailed below were approved by the UK Home Office. See the Table of Materials for a list of all materials, reagents, and equipment used in this protocol.

1. Preparing stock solutions

1. PVA stock solution
   1. Take 50 mL of tissue culture quality water in a small glass bottle (suitable for autoclaving). Warm the water to near boiling in a microwave.
   2. Weigh out 5 g of PVA powder and add to the water.
      NOTE: It is recommended to only dissolve 1-5 g of PVA. Dissolving larger amounts may result in incomplete reconstitution.
   3. Close the lid tightly and shake to mix. Then, loosen the lid.
      NOTE: Be careful when re-opening, as the pressure may change due to the changing water temperature.
   4. Autoclave and allow to cool. Close the lid tightly and shake to mix. Make aliquots in sterile tubes and store for up to 3 months at 4 °C.
2. Dissolve lyophilized cytokines in F-12 medium containing 1 mg/mL PVA. Reconstitute the stem cell factor to 10 µg/mL and thrombopoietin to 100 µg/mL to generate 1:1,000 stocks. Make aliquots in sterile tubes and store long-term at -80 °C. Alternatively, store the aliquots at 4 °C for up to 1 week.
   NOTE: Avoid the reconstitution of cytokines in bovine serum albumin due to the negative impact on mouse HSC expansion.

2. HSC bone marrow extraction and c-Kit⁺ enrichment

1. Dissect the femurs, tibias, pelvises, and spine from freshly euthanized 8-12-week-old C57BL/6 mice (via CO₂ asphyxiation and/or cervical dislocation) and place the bones in PBS on ice.
   NOTE: Ensure the surfaces and tools are sterilized with 70% ethanol.
2. Clean the bones using lint-free delicate task wipes, removing muscle and spinal cord, and add to a mortar with ~3 mL of PBS.
   NOTE: Ensure the pestle and mortar are sterilized with 70% ethanol and then washed out once with PBS.
3. Crush the bones with the pestle without grinding to minimize shearing forces. Break up large bone marrow fragments released into the PBS using a 19 G needle attached to a 5 mL syringe. Transfer the cell suspension through a 70 µm filter into a 50 mL conical tube.
4. Once the suspension has been transferred, repeat with fresh PBS until the bones are bleached and no marrow is visible. Aim for an end volume of ~30 mL per mouse and ~50 mL for two mice.
5. Mix the bone marrow cells, collect 10 µL of cells, and count using Türks' solution at a dilution of 1:10-1:20 with a hemocytometer.
   NOTE: One mouse is expected to yield 2-5 × 10⁸ whole bone marrow cells.
6. Spin at 450 × g for 5 min at 4 °C, discard the supernatant, and resuspend the pellet in cold PBS (350 µL for one mouse or 500 µL for two mice).
7. Add 0.2 µL of allophycocyanin (APC) anti-c-Kit antibody per 10 million cells and incubate for 30 min in the dark at 4 °C.
8. Add 5 mL of cold PBS to the incubated cells to wash off excess antibody and filter through a 50 µm filter into a fresh 15 mL conical tube.
9. Wash the original tube with 7 mL of cold PBS and transfer through the filter. If filter clogging occurs, scratch the filter surface with a P1000 tip.
10. Spin at 450 × g for 5 min at 4 °C, discard the supernatant, and resuspend the pellet in cold PBS (350 µL for one mouse or 500 µL for two mice).
11. Add 0.2 µL of anti-APC microbeads per 10 million cells and incubate for 15 min in the dark at 4 °C. Add 12 mL of sterile PBS to wash off excess microbeads.
12. Spin down the cells at 450 × g for 5 min at 4 °C and resuspend in 2 mL of cold PBS. While the cells are spinning in this step, proceed to step 2.13.
13. Prepare for column enrichment by placing a magnetic filtration column into the magnet of a magnetic column separator, with a 50 µm filter on top and a 15 mL conical tube below.
14. Run 3 mL of sterile PBS through the 50 µm filter and filtration column. Once the PBS has run through, pass the cell suspension through the column, followed by three washes of 3 mL of cold PBS each time. For each wash, wait for the column to stop dripping before adding the next wash.
15. Remove the column from the magnet and place it on top of a fresh 15 mL tube. Add 5 mL of cold PBS, fit the column plunger onto the column, and elute the cells by pushing the plunger.
16. Mix the c-Kit-enriched cells, collect 10 µL of cells, and count using Türks' solution at a dilution of 1:2 with a hemocytometer. The typical yield from one mouse is 2-5 × 10⁶ c-Kit⁺ cells.
    NOTE: At this point, the cells can be directly seeded into HSC media or prepared for HSC fluorescence-activated cell sorting (FACS) purification.

3. Initiating cell cultures with c-Kit-enriched HSPCs

1. Prepare fresh medium (Table 1) for the number of cells/wells required. Seed 0.5-1 million cells per mL for c-Kit-enriched HSPCs.
2. Spin down the c-Kit-enriched cells and resuspend in HSC medium at the desired cell density.
3. Transfer the cells to fibronectin-coated or negative surface charged plates, at 200 µL per 96-well plate or 1 mL per 24-well plate.
4. Place the cells in a tissue culture incubator set to 37 °C and 5% CO₂.

4. Initiating cell cultures with FACS purified HSCs

1. Prepare an appropriate volume of the biotinylated lineage antibody stain: 3 µL of master mix (Table 2) per 10 million cells, diluted 1:100 in PBS.
2. Spin down the c-Kit-enriched cells and resuspend in the lineage antibody stain for 30 min at 4 °C.
3. Wash with 10 mL of sterile PBS and spin down at 450 × g for 5 min at 4 °C.
4. Prepare an appropriate volume of the fresh HSC antibody stain (Table 3): 300 µL per 10 million cells. Alongside this sample staining, prepare appropriate staining control samples for compensation and gating.
   NOTE: Work in a tissue culture hood with the light off when using dye-conjugated antibodies.
5. Resuspend the cells in the HSC antibody stain and incubate at 4 °C for 90 min. Mix the cells every 20-30 min by tapping to prevent cell pelleting.
6. Wash with 10 mL of sterile PBS and spin down at 450 × g for 5 min at 4 °C.
7. Aspirate the supernatant, flick the pellet to disrupt, and resuspend in sterile PBS with 0.5 µg/mL propidium iodide (PI).
8. Prepare fresh medium (Table 1) for the number of wells required, and plate into fibronectin or negative surface charged plates (200 µL per 96-well plate or 1 mL per 24-well plate).
9. Prepare the FACS machine for sorting and sort CD150⁺CD34^-c-Kit⁺Sca1⁺Lineage^- HSCs directly into media-containing wells (see Figure 1 for the standard FACS gating strategy used here). Sort up to 200 cells per 96-well plate well or up to 1,000 cells per 24-well plate well.
   NOTE: FACS machines should be operated by a trained scientist. It is recommended that users contact their local FACS facility to discuss this sorting strategy if they are not experienced in FACS isolation of mouse HSCs.

5. Performing media changes

1. For cell cultures initiated from c-Kit-enriched cells, begin media changes after 2 days. For cell cultures initiated from FACS-isolated HSCs, begin media changes after 5 days.
2. Prepare sufficient fresh prewarmed (~37 °C) HSC medium (Table 1) for all wells.
3. Gently remove the plate from the tissue culture incubator.
   NOTE: As HSPCs on negative surface charged plates are more easily disturbed than those on fibronectin, extra care should be taken when changing the medium on negative surface charged plates to avoid disturbing the cell cultures.
4. Using a pipette or vacuum pump, slowly remove ~90%-95% of the medium from the well meniscus.
   NOTE: Avoid drawing up the medium from the base of the well, otherwise many cells will be removed.
5. Add 200 µL (for 96-well plates; 1 mL for 24-well plates) of fresh medium to the well.
6. Return the plate to the tissue culture incubator.
7. Repeat steps 5.1-5.6 every 2-3 days until the experimental end point.
8. For cell cultures initiated with c-Kit-enriched HSPCs (section 3), split the cultures at a ratio of 1:2-1:3 after ~3 weeks. For cell cultures initiated with FACS purified HSCs (section 4), split the cultures at a ratio of 1:2-1:3 after ~3 weeks and when the cultures are >90% confluent.
   ​NOTE: The exact timeline will depend on the experimental interests. These cultures have been characterized for 4-8 weeks⁸, but additional culture lengths may be possible.

6. Electroporating cultured HSPCs

NOTE: This protocol is for electroporation of Cas9/sgRNA ribonucleoprotein (RNP), but could be adapted for electroporation of mRNA or other recombinant proteins. Initiate cultures with sufficient numbers of cells in order to perform this at the desired experimental time point.

1. Perform a medium change 1 day before electroporation, as described in section 5.
   NOTE: A minimum of an overnight culture before electroporation is recommended. However, cells are typically cultured for 1-3 weeks before transduction.
2. On the day of electroporation, set up the nucleofector. Turn on the machine. On the touchscreen, select the X module and then the cuvette size used.
3. Prepare sufficient P3 solution (according to the manufacturer's instructions) for the scale of the electroporation (100 µL per cuvette or 20 µL per microcuvette) and allow to equilibrate to room temperature.
4. Prepare sufficient fresh medium (Table 1) for the number of wells being plated and add 500 µL of medium for a 24-well plate well or 100 µL of media to a 96-well plate well.
5. Thaw out sgRNA (prediluted to 2 µg/mL in RNase-free water) on ice and mix 16 µg of sgRNA with 30 µg of Cas9 enzyme (at 10 µg/mL) in a sterile PCR tube. Include an extra PCR tube containing only Cas9 protein as a control. Mix by flicking the tube, then briefly spin down. Incubate at 25 °C for 10 min in a thermocycler to complex the RNP, and then keep the tubes on ice.
   NOTE: This can be scaled down fivefold if performing electroporation in microcuvettes.
6. Mix and transfer the HSPCs for electroporation into a tube; collect 10 µL of the cells and count using Türks' solution at a dilution of 1:2 with a hemocytometer.
   NOTE: It is recommended to electroporate 1-5 million cells per 100 µL cuvette (scaled down fivefold for the microcuvette).
7. Spin down the appropriate number cells in a 1.5 mL tube at 450 × g for 5 min at 4 °C. Aspirate as much of the supernatant as possible and resuspend with 100 µL of nucleofection buffer.
8. Transfer the cell suspension immediately into the PCR tube containing the complexed RNP and pipette up and down slowly to gently mix and transfer to a 100 µL electroporation cuvette. Eject the mixture into the cuvette slowly and in one fluid motion to avoid the formation of air bubbles in the cuvette.
9. On the electroporator, select the position of the wells being electroporated. Using the touchscreen, select the cell type program CD34⁺, human, or type in the pulse code EO100. Press the OK button.
10. Transfer the cuvettes to the electroporator. Press the Start button on the touchscreen to initiate electroporation. Directly after electroporation, add 500 µL of the culture medium to the cuvette (100 µL if using a microcuvette).
11. Gently transfer the cells to the prepared plate and return to the tissue culture incubator.
12. For procedures employing an AAV6 donor template, add the vector immediately after the cells have been transferred to a fibronectin-coated plate at a concentration of 5,000 vectors/cell.
13. After 6-18 h, prepare fresh medium and perform a medium change as described in section 5. For this medium change, remove only 80%-90% of the medium.
    NOTE: Avoid drawing up the medium from the base of the well.
14. Analyze the cells by flow cytometry after 2 or more days (see Section 8 for details).
15. Continue to perform medium changes every 2 days, as described in section 5, until the experimental end point is reached.
    ​NOTE: Editing rates of CRISPR/Cas9 using this method rely highly on the targeting efficiency of the designed sgRNA. Editing rates up to 95% have been observed with an efficient guide¹⁵. Guidelines for sgRNA design have been previously detailed elsewhere^19,20.

7. Transducing cultured HSPCs with lentiviral vector

NOTE: Initiate cultures with sufficient numbers of cells, to perform this at the desired experimental time point.

1. Generate and titer lentiviral vector (depending on the experimental goals). Thaw lentiviral vector on ice.
2. Perform a medium change (as in section 5); then, mix and transfer the HSPCs for transduction into a tube. Collect 10 µL of the cells and count using Türks' solution at a dilution of 1:2 with a hemocytometer.
   NOTE: Cells can be transduced immediately following plating. However, cells are typically cultured for 1-3 weeks before transduction.
3. Replate the required dose of cells for transduction (typically 100,000 cells per 96-well plate). Separately, plate un-transduced negative control cells.
   NOTE: If using negative surface charged plates, transfer the cells to fibronectin-coated plates for lentiviral vector transduction.
4. Add lentiviral vector to each well of cells: add 20 transduction units per cell to achieve ~30% transduction efficiency. However, determine the lentiviral vector dose empirically, depending on the experimental requirements.
   NOTE: Ensure lentiviral vector is disposed of according to institutional guidelines.
5. Return the cells to the tissue culture incubator for 6 h. After, perform a medium change as described in section 5.
   NOTE: The supernatant contains live virus. Dispose of according to institutional guidelines.
6. Analyze the cells by flow cytometry (e.g., for GFP expression) after 2 days or more. See section 8 for details.
7. Continue to perform medium changes every 2 days, as described in section 5, until the experimental end point is reached.

8. Flow cytometric analysis of HSPC cultures

1. Prepare a concentrated cultured ex vivo HSC antibody mix in PBS containing 2% FBS (Table 4). Store the mixture at 4 °C in the dark for up to 1 month.
   NOTE: Work in a tissue culture hood with the lights off when using dye-conjugated antibodies.
2. Add 2 µL of concentrated antibody mix to 50 µL of cells. Incubate for 30 min at 4 °C in the dark. Alongside this sample staining, prepare appropriate staining control samples for compensation and gating.
3. Add 200-1,000 µL of PBS containing 2% FBS and centrifuge at 450 × g for 5 min at 4 °C. Remove as much supernatant as possible and resuspend in 100-500 µL of PBS containing 2% FBS and 0.5 µg/mL PI.
4. Set up the flow cytometer and record at least 10,000 live cells per samples.
   NOTE: Flow cytometers should be operated by a trained scientist. Users must contact their local FACS facility to discuss this analysis if they are not experienced in flow cytometry.
5. Export the data in FCS format and analyze the data using appropriate flow cytometry analysis software. See Figure 2 for the standard gating strategy used here.

For the FACS purification of HSCs, we expect that within the c-Kit-enriched bone marrow, ~0.2% of the cells are the CD150⁺CD34^-c-Kit⁺Sca1⁺Lineage^- population for young (8-12-week-old) C57BL/6 mice (Figure 1). However, it is likely that transgenic mice or mice of different ages display differing HSC frequencies. After 4 weeks of culture, we expect the CD201⁺CD150⁺c-Kit⁺Sca1⁺Lineage^- fraction to be ~10% (Figure 2). These results are similar for cell cultures initiated from c-Kit enriched bone marrow or FACS purified HSCs. Following transduction with a GFP-expressing lentiviral vector (at 20 transduction units/cell), we expect ~30% GFP⁺ cells (Figure 3).

When confluent, we expect the cultures to be at ~2 million cells/mL. Within the culture, we expect to see mainly small round cells, although it is normal to see a small frequency of larger round (megakaryocyte-like) cells. Within the cell cultures initiated from c-Kit-enriched bone marrow, some initial cell death is expected¹⁴. Approximately 50% cell death is expected within the first 24-48 h, before the first medium changes are performed. Initial dead cell debris in these cultures likely comes from cell death within the cultures rather than bone debris (from the crushed bones), since the c-Kit-enrichment step should deplete such bone debris. Cell numbers, however, return to 80%-100% of the seeded cell numbers after 1 week (see Table 5 for expected cell density values). If poor results are seen, we recommend troubleshooting the protocol (Table 6). In this case, it can be most simple to batch-test reagents using the c-Kit-enriched bone marrow protocol (Section 3), as this avoids complications associated with FACS-sorting. Note that the representative results are based on the use of reagents and equipment detailed in the Table of Materials; similar results may be achieved using reagents from different vendors, however, validation (and titration) of new reagents is likely to be necessary.

Figure 1: Gating strategy for FACS purification of HSCs from c-Kit-enriched bone marrow. Sequential gating used to identify CD150⁺CD34^-c-Kit⁺Sca1⁺Lineage^- cells from c-Kit-enriched bone marrow. Abbreviations: FACS = fluorescence-activated cell sorting; HSCs = hematopoietic stem cells; SSC-A = side scatter-peak area; FSC-A = forward scatter-peak area; FSC-W = forward scatter-peak width; FSC-H = forward scatter-peak height; SSC-H = side scatter-peak height; PI = propidium iodide; PE = phycoerythrin; APC = allophycocyanin; FITC = fluorescein isothiocyanate. Please click here to view a larger version of this figure.

Figure 2: Gating strategy for flow cytometric analysis of cultured HSPCs. Sequential gating used to identify CD201⁺CD150⁺c-Kit⁺Sca1⁺Lineage^- cells from HSPC cultures. Abbreviations: HSPCs = hematopoietic stem and progenitor cells; SSC-A = side scatter-peak area; FSC-A = forward scatter-peak area; FSC-W = forward scatter-peak width; FSC-H = forward scatter-peak height; SSC-H = side scatter-peak height; PI = propidium iodide; PE = phycoerythrin; APC = allophycocyanin. Please click here to view a larger version of this figure.

Figure 3: Representative results from lentiviral vector transduction. Representative GFP expression following cultured HSPC transduction with a GFP-expressing lentiviral vector (20 transduction units/cell) at 48 h after transduction. Abbreviations: GFP = green fluorescent protein; HSPCs = hematopoietic stem and progenitor cells; FSC-H = forward scatter-peak height. Please click here to view a larger version of this figure.

Table 1: HSC media composition. Media reagent volumes for 1 mL of complete medium. Abbreviations: PVA = polyvinyl alcohol; SCF = stem cell factor; TPO = thrombopoietin; ITSX = insulin-transferrin-selenium-ethanolamine.

Table 2: Biotinylated antibody cocktail. Antibody volumes for the stock of the biotin antibody cocktail.

Table 3: Fresh HSC antibody cocktail. Antibody volumes for the fresh HSC antibody cocktail. Abbreviations: PE = phycoerythrin; APC = allophycocyanin; FITC = fluorescein isothiocyanate.

Table 4: Cultured HSC antibody cocktail (100x). Antibody volumes for the 100x stock of the cultured HSC antibody cocktail. Abbreviations: PE = phycoerythrin; APC = allophycocyanin; FITC = fluorescein isothiocyanate.

Table 5: Expected results. Expected cell densities and frequencies of c-Kit⁺Sca1⁺Lineage^- cells following (1) seeding of c-Kit⁺ bone marrow, (2) electroporation, and (3) lentiviral vector transduction based on seeding at 1 × 10⁶ cells mL^-1. Abbreviations: KSL = c-Kit⁺Sca1⁺Lineage^-; HSPC = hematopoietic stem and progenitor cell.

Table 6: Common troubleshooting issues. Summary of common issues and suggested troubleshooting.

We hope that this protocol provides a useful approach to investigate HSC biology, hematopoiesis, and hematology more generally. Since the initial development of the PVA-based culture method for FACS-purified HSCs⁸, the method has been extended. For example, the method has been shown to work with c-Kit enriched with bone marrow and with negative surface charged plates¹⁴. Its compatibility with transduction and electroporation has also been demonstrated^14,15. The in vivo validation of these HSC and c-Kit⁺ HSPC cultures can be found in these publications, while readers can refer to other published protocols for in vivo transplantation protocols²¹. We do not see major differences in lineage chimerism between fresh and cultured HSPCs following transplantation, however, the reconstitution potency of individual HSCs after ex vivo expansion is yet to be determined in detail. The large numbers of HSCs generated by this approach opens up new ways to interrogate HSCs using molecular or biochemical assays that require large cell numbers. Additionally, being able to generate genetically modified HSCs within these culture systems should allow us to further probe the mechanisms regulating HSC activity and the hematopoietic system. For example, these systems are amenable to performing genetic screens¹⁶.

Mechanistically, we do not yet fully understand why PVA and other polymers can replace serum albumin and support efficient HSC expansion; we believe PVA at least partially replaces serum albumin through stabilizing cytokines within the media⁹. Additionally, the lack of poorly defined bioactive contaminants found in serum albumin products appears to reduce HSC differentiation. The use of synthetic polymers should also help to reduce batch-to-batch variability and confounding effects associated with these bioactive contaminants when studying HSCs ex vivo²². There is also still more to be learnt regarding why certain polymers provide better support for HSCs ex vivo.

While already powerful, further optimization and characterization of these culture protocols should be possible. In particular, it would be useful to improve the purity of HSCs within these cultures. Additional markers distinguishing the long-term HSC compartment from these cultures would also help track and isolate these cell types. It is also of interest to see whether this system can be eventually used to quantify HSC activity ex vivo, without the need for in vivo transplantation assays. We also do not yet understand for how long HSCs can be expanded ex vivo, although it is certainly longer than 6-8 weeks if maintained properly⁸. Finally, while these cultures provide a useful model to study mouse HSCs, it is also important to develop equivalent culture systems for human HSCs to provide a more tractable system to study human HSC biology and hematopoiesis, and eventually, to generate HSCs for clinical stem cell transplantation therapies.