The OP9-DL1 System: Generation of T-Lymphocytes from Embryonic or Hematopoietic Stem Cells In Vitro

The OP9-DL1 System: Generation of T-Lymphocytes from Embryonic or Hematopoietic Stem Cells In Vitro

Roxanne HolmesandJuan Carlos Zúñiga-Pflücker1

Sunnybrook Research Institute and Department of Immunology, University of Toronto, Toronto, Ontario M4N 3M5, Canada

1 Corresponding author (jczp@sri.utoronto.ca )

INTRODUCTION

Differentiation of mouse embryonic stem cells (ESCs) or hematopoietic stem cells (HSCs) from fetal liver or bone marrow into T-lymphocytes can be achieved in vitro with the support of OP9-DL1 cells, a bone-marrow-derived stromal cell line that ectopically expresses the Notch ligand, Delta-like 1 (Dll1). This approach provides a simple, versatile, and efficient culture system that allows for the commitment, differentiation, and proliferation of T-lineage cells from different sources of stem cells. This article contains a series of protocols, the first of which describes the establishment, maintenance, and storage of OP9 and OP9-DL1 cells. Subsequent protocols detail how to co-culture the OP9 and OP9-DL1 cells with either ESCs or HSCs from fetal liver or bone marrow, leading to in vitro differentiation of the stem cells into lymphocytes.

RELATED INFORMATION

Preparation of the OP9 and OP9-DL1 cells should be started ~1 wk prior to initiating co-cultures with ESCs or HSCs. Protocols 3 and 4 use HSCs isolated from murine fetal liver or bone marrow. HSC isolation protocols are widely available (Klug and Jordan 2002 ; Schmitt et al. 2004 ; de Pooter et al. 2006 ; Bunting 2008 ). For a protocol to isolate and maintain mouse embryo fibroblasts (mEFs), see Preparing Mouse Embryo Fibroblasts (Nagy et al. 2006a ). For a protocol to derive or maintain ESCs, see De Novo Isolation of Embryonic Stem (ES) Cell Lines from Blastocysts (Nagy et al. 2006b ).

MATERIALS

Reagents

Anti-CD24 monoclonal antibody (mAb) (J11d clone) (for Protocol 3)

Use either culture supernatant that contains anti-CD24 mAb or purified anti-CD24 mAb (see Step 55).

BDPharmLyse (red blood cell lysing reagent; BD Biosciences 555899) (for Protocol 4)

Buffer for cell staining and cell sorting (for Protocol 4; see Step 72)

Complement, reconstituted from rabbit (Cedarlane CL3331) (for Protocol 3)

ESC medium (for Protocols 1, 2)

Freezing medium for OP9 cells (for Protocols 1, 2)

Leukemia inhibitory factor (LIF; 10 µg/mL; Chemicon LIF2010) (for Protocol 2; see Step 20)

Lympholyte-M (Cedarlane CL5120) (for Protocol 3)

Mice, 4-8 wk old (for Protocol 4; see Step 69)

Mice, fetal, at embryonic day 13-15 (for Protocol 3; see Step 52)

Mouse embryonic fibroblast (mEF) cells, irradiated and growing in a 6-cm dish (for Protocol 2)

Alternatively, a gelatin-coated 6-cm dish can be used. For a protocol to isolate and maintain mEFs, see Preparing Mouse Embryo Fibroblasts (Nagy et al. 2006a ).

Mouse embryonic stem cells (ESCs), stored frozen in liquid nitrogen (for Protocol 2)

R1 ESCs can be obtained from ATCC (SCRC-1036). For a protocol to derive or maintain ESCs, see De Novo Isolation of Embryonic Stem (ES) Cell Lines from Blastocysts (Nagy et al. 2006b ).

OP9 or OP9-DL1 cells, stored frozen in liquid nitrogen

OP9 cells can be obtained from the Riken Laboratory Cell Repository (Japan), and then transduced with a retroviral construct encoding Delta-like-1 to generate OP9-DL1 cells (Schmitt and Zúñiga-Pflücker 2002 ), or they can be requested from the Zúñiga-Pflücker laboratory.

OP9-DL1 or OP9 cells growing on a 6-well plate (for Protocol 2; see Step 39)

OP9-DL1 cells are used to generate T-lineage cells, and OP9 cells are used to generate B-lineage and myelo-erythroid cells.

OP9 medium

Phosphate-buffered saline (PBS; Hyclone SH30256)

Recombinant human Flt-3/Flk-2 ligand (R&D Systems 308-FK) (for Protocols 2-4)

Recombinant murine IL-7 (Peprotech 217-17) (for Protocols 2-4)

Trypsin solution, 0.25% (Invitrogen 15090046) (for Protocols 1, 2)

Dilute stock trypsin to 0.25% with PBS.

Equipment

Biosafety cabinet

Cell strainers (40-µm pore size; BD Falcon 352340) (for Protocols 2-4)

Centrifuge

Flow cytometer (for Protocols 2-4)

Forceps, 4-in. straight tips (Electron Microscopy Sciences 72991-4S) (for Protocol 4)

Forceps, curved fine points Dumont #7 (Electron Microscopy Sciences, 72800-D), and super fine points Dumont #5 (Electron Microscopy Sciences, 72700-D) (for Protocol 3)

Glass stopper, No. 24 (for Protocol 4)

Incubator, humidified (37°C and 5% CO2 )

Liquid-nitrogen cell storage system

Magnetic assisted cell sorter (MACS) (optional; see Protocol 4, Step 73)

Microscope

Plunger from a 3-mL syringe (for Protocol 3)

Scissors, dissecting (Electron Microscopy Sciences 72940) (for Protocols 3, 4)

Tissue culture plasticware

Carry out all procedures using standard aseptic technique with sterile plasticware (tissue-culture-treated 10-cm and 6-well plates, 15-mL and 50-mL centrifuge tubes, 2-mL cryovials, serological pipettes, and pipette filtered tips).

Water bath pre-set to 37°C (for Protocols 1, 2)

METHOD

Protocol 1: Preparation of OP9 Cells for Co-culture

Thawing OP9 or OP9-DL1 Cells

1. Add 12 mL of OP9 medium to a 15-mL centrifuge tube.

2. Thaw OP9 or OP9-DL1 cells quickly in a 37°C water bath.

3. Pipette the cell suspension slowly and gently from the cryovial tube, and transfer the contents to the 15-mL tube containing medium.

4. Centrifuge the cells at 400g (1500 rpm) for 5 min at 4°C. Resuspend the cells in 10 mL of OP9 medium.

5. Transfer the resuspended cells into a 10-cm dish, and place the dish in an incubator.

Maintaining OP9 or OP9-DL1 Cells

6. To passage the cells, remove the medium from the 10-cm dish.
The cells will become confluent in the dish within 2-3 d, depending on the method used to freeze the cells. Passage the cells before they reach ~80% confluency ( Fig. 1 ).

View larger version (59K): [in this window] [in a new window]

Figure 1. Photomicrographs of ESC/OP9 co-culture. (a ) Undifferentiated ES cells on mEF. (b ) Monolayer of OP9 cells. (c ) Day 0 ESC/OP9 co-culture. (d ,e ) Day 5 ESC/OP9 mesoderm-like colonies. (f ) Day 8 ESC/OP9 small, round clusters of cells. (g ) Day 12 ESC/OP9-DL1 hematopoietic cells. (h ) Day 16 ESC/OP9-DL1 hematopoietic and early T-lineage cells. (i ) Day 20 ESC/OP9-DL1 T-lineage cells.

7. Wash the plate with 4 mL of PBS. Discard the PBS.

8. Trypsinize the cells with 4 mL of 0.25% trypsin solution, and incubate the cells for 5 min at 37°C.

9. Disaggregate the cells from the dish by pipetting them up and down, and transfer the cell suspension into a 15-mL tube containing 4 mL of OP9 medium.

10. Wash the plate with PBS to remove any remaining cells, and transfer these cells to the same 15-mL tube.

11. Centrifuge the cells at 400g (1500 rpm) for 5 min at 4°C, and resuspend them in 4 mL of OP9 medium.

12. Transfer 1-mL aliquots of cells to four 10-cm dish each containing 9 mL of OP9 medium.
Maintain the cells by splitting them at a ratio of 1-to-4, and passaging the cells every 2 d. Do not keep the cells in continuous culture for more than 6 wk.
The cells can be used in subsequent protocols; e.g., Steps 28, 59, and 74.

Freezing OP9 or OP9-DL1 Cells

13. Passage the cells as described in Steps 6-11, except resuspend the cells (~8-10 x 105 cells), in 2 mL of freezing medium for OP9 cells per 10-cm dish.
Preferably, freeze cells within the first two to three passages.

14. Aliquot 1 mL of cell suspension per cryovial.

15. Freeze the cells at -80°C, and then transfer them to liquid nitrogen for storage.

Protocol 2: In Vitro Generation of T-Lymphocytes from ESCs

Thawing ESCs

16. Prepare a 15-mL tube containing 12 mL of ESC medium.

17. Thaw the ESCs quickly in a 37°C water bath, and transfer the thawed cells slowly into the 12 mL of ESC medium.

18. Centrifuge the cells at 400g (1500 rpm) for 5 min at 4°C, and resuspend in 3 mL of ESC medium.

19. Seed ESCs onto a 6-cm dish containing irradiated mEF cells or a gelatin-coated 6-cm dish.

20. To keep the ESCs from differentiating, add 10 ng/mL of LIF when grown on mEF cells or 20 ng/mL of LIF when grown on gelatin.
The mEF cells can be irradiated up to 2 d before using them as feeder cells. It is important that the mEF cell layer completely covers the surface of the tissue culture dish, because ESCs will begin to undergo differentiation within mEF cell-free areas.

Maintaining ESCs

21. The following day, change the ESC medium and again add the appropriate concentration of LIF.

22. Passage the ESCs the next day with trypsin-mediated disaggregation.
Trypsin-mediated passage is used to break up large ESC colonies, allowing the culture to expand.

i. Remove the medium from the 6-cm dish. Wash the plate with 3 mL of PBS. Discard the PBS.

ii. Trypsinize the cells with 4 mL of 0.25% trypsin solution, and incubate the cells at 37°C for 5 min.

iii. Disaggregate the cells from the dish by pipetting them up and down, and transfer the cell suspension into a 15-mL tube containing 3 mL of ESC medium.

iv. Wash the plate with PBS to remove any remaining cells, and transfer these cells to the same 15-mL tube.

v. Centrifuge the cells at 400g (1500 rpm) for 5 min at 4°C, and resuspend them in 3 mL of ESC medium.

23. Seed ESCs in 3 mL of ESC medium onto irradiated mEF cells or gelatin-coated plates, and add LIF.

24. Repeat Steps 22 and 23 until the ESCs are needed for co-culturing.
The ESCs grow as colonies. If they become too crowded, passage the cells more sparsely ( Fig. 1a ).

Freezing ESCs

25. Passage the ESCs as described in Step 22. Resuspend the cells in 3 mL of freezing medium for OP9 cells.

26. Aliquot 1 mL of cell suspension per cryovial (~3-6 x 105 cells/cryovial).

27. Freeze the cells at -80°C, and then transfer them to liquid nitrogen for storage.

Establishing ESC/OP9-DL1 Cell Co-cultures

Day 0: Initiation of Co-culture
See Figure 2 for a schematic of the ESC/OP9 co-culture.

View larger version (18K): [in this window] [in a new window]

Figure 2. Schematic overview of ESC/OP9(-DL1) co-culture system, with key steps highlighted.

28. For each 10-cm dish containing 80%-90% confluent OP9 cells, remove the medium and replace it with 9 mL of fresh OP9 medium.
The OP9 cells were prepared in Steps 1-12.

29. Harvest ESCs as a single cell suspension by trypsin-mediated disaggregation (Step 22). Resuspend the cells to a concentration of 5 x 104 cells per mL of OP9 medium.

30. Seed 1 mL of ESCs per 10-cm dish of OP9 cells, and return the dishes to the incubator (Fig. 1c ).
Use OP9 cells (not ectopically expressing Dll1) to initiate the culture, because a more robust population of mesoderm colonies grows on these stromal cells as opposed to OP9-DL1 cells. The cells will be transferred onto OP9-DL1 cells at day 8 to induce T-lineage differentiation ( Fig. 2 ).
See Troubleshooting.

Day 3: Medium Change

31. Remove the old medium, replace it with 10 mL of OP9 medium, and return cells to the incubator.
The ESCs should have less defined borders and should start forming mesoderm colonies.

Day 5: Trypsin-Mediated Passage and Pre-plating

32. Remove the medium and trypsinize the cells (see Steps 7 and 8).
Mesoderm colonies should be visible ( Fig. 1d ,e), and some colonies will appear three-dimensional with a wagon wheel shape, whereas others will appear like small cells clustered in a pile or as bunches. These colonies will become visible without the aid of a microscope and will appear as white circles in the dish.

33. Resuspend the trypsin-disaggregated cells with 4 mL of OP9 medium.

34. Return the dish to the incubator for 30 min.
This pre-plating step will allow OP9 cells to readhere to the dish, which will limit the number of OP9 cells that are transferred.

35. Wash the non-adherent cells off the dish, filter the cells through a 40-µm cell strainer, and centrifuge the cells at 400g (1500 rpm) for 5 min at 4°C.

36. Resuspend the cells in OP9 medium, seed 5 x 105 cells per 10-cm dish containing ~80% confluent OP9 cells in 10 mL of OP9 medium, and add 5 ng/mL of recombinant human Flt-3L. Return the dish to the incubator.
The number of 10-cm dishes to be seeded should be equal to, or greater than, the number of analysis time points required. This ensures that the cultures remain undisturbed and there are enough samples for analysis (e.g., flow cytometry). Each 10-cm dish at day 5 yields enough cells to seed six to eight 10-cm dishes.

Day 8: Harvesting Semiadherent Hematopoietic Cells

37. Using the medium that is in the dish, wash the cells gently enough to keep the OP9 cell monolayer intact, but use enough force to remove the semiadherent cells.
Clusters of round, shiny cells should be visible as small groups, either semiattached to the OP9 cells or in suspension ( Fig. 1f ). Check the dish under the microscope to ensure that all the round, shiny cells have been removed.
See Troubleshooting.

38. Filter the cells through a 40-µm cell strainer, and centrifuge at 400g (1500 rpm) for 5 min at 4°C.

39. Resuspend the cells in 3 mL of OP9 medium per 10-cm dish initially seeded on day 5, and seed 3 mL per well of a 6-well plate containing either OP9-DL1 cells to generate T-lineage cells or OP9 cells to generate B-lineage and myelo-erythroid cells.

40. Add cytokines to each well; final concentrations are 5 ng/mL Flt-3L and 1 ng/mL.

Day 10: Medium Change

41. Remove the medium carefully so as to not disturb the co-culture.
Large clusters of round, shiny cells should be present, with some cells detached in suspension. Collect all of the medium (which may contain differentiating lymphocytes) and centrifuge these cells at 400 g (1500 rpm) for 5 min at 4°C, thus limiting cell loss during the medium change.

42. Centrifuge the medium at 400g (1500 rpm) for 5 min at 4°C. Resuspend the cells present in the spent medium with 3 mL of OP9 medium containing 5 ng/mL Flt-3L and 1 ng/mL IL-7.

43. Transfer the medium and cells to the same well, and return the plates to the incubator.

Day 12: No-Trypsin Passage

Large clusters of round, shiny cells should be present throughout the dish, with some cells detached in suspension ( Fig. 1g ). The OP9 cells may lift as a sheet, which will be broken up by forceful pipetting.

45. Filter the cells through a 40-µm cell strainer.

46. Wash the well with 3 mL of PBS, filter the suspension through the same cell strainer, and centrifuge the cells at 400g (1500 rpm) for 5 min at 4°C.

47. Resuspend the cells in 3 mL of OP9 medium, and seed the cells onto a new well of OP9 or OP9-DL1 cells, with the addition of 5 ng/mL Flt-3L and 1 ng/mL IL-7. Return the plates to the incubator.
This is a good time point to analyze by flow cytometry to determine the presence of different lympho-hematopoietic cell populations. As shown in Figure 3 , for ESCs cultured on OP9 cells, the majority of cells will be erythro-myeloid Ter119 + or CD11b + , while for ESCs cultured on OP9-DL1 cells, most of the cells will resemble DN1 (CD44 + CD25- ) stage thymocytes, although some DN2 (CD44 + CD25 + ) and DN3 (CD44- CD25 + ) stage cells may be present.

View larger version (20K): [in this window] [in a new window]

Figure 3. Flow cytometry analysis for the indicated cell surface markers of ESC/OP9 co-cultures and ESC/OP9-DL1 co-cultures at the indicated time points and culture conditions.

Day 14: Medium Change

48. Repeat Steps 41-43.

Day 16: No-Trypsin Passage

49. Repeat Steps 44-47.
By this time point, the dish should be replete with differentiating lymphocytes ( Fig. 1h ). For ESCs cultured on OP9, CD19 + cells will be starting to emerge. For ECSs cultured on OP9-DL1, cells that are mostly at the DN3 and DN4 stage of thymocyte development will be present, and some CD4 + CD8 + (double positive, DP) may start to appear ( Fig. 3 ).

Day 18 and Onward, at 2-d to 4-d Intervals: Medium Change

50. Repeat Steps 41-43.

Day 20 and Onward, at 4-d to 5-d Intervals: No-Trypsin Passage

51. Repeat Steps 44-47.
The differentiating lymphocytes expand rapidly ( Fig. 1i ), requiring daily monitoring to determine whether to simply change medium or to passage/split the co-cultures into additional dishes. For ESCs cultured on OP9, the majority of the cells will be CD19 + B-lineage cells. For ESCs cultured on OP9-DL1, the cells are mostly CD4 + - and CD8 + -expressing T-lineage cells ( Fig. 3 ).

Protocol 3: Fetal Liver-Derived HSC Differentiation on OP9-DL1 Cells

Day 0: Initiation of Fetal Liver Co-culture

52. Isolate liver tissue from eight to 10 fetal mice at embryonic day 13-15 (E13-E15), using scissors and curved forceps.

53. Force the fetal livers through a 40-µm cell strainer using the rubber end from a 3-mL syringe plunger, and wash the strainer with OP9 medium.

54. Centrifuge the cells at 400g (1500 rpm) for 5 min at 4°C, and resuspend them in 4 mL of OP9 medium in a 15-mL tube.

55. To enrich for HSCs, add 1 mL of reconstituted rabbit complement and either 5 mL of culture supernatant containing anti-CD24 mAb (J11d clone) or 5 ml of OP9 medium containing 1 ng of purified anti-CD24 mAb per 106 cells.
CD24 is expressed in all Lineage + cells in E13-E15 fetal liver cells.

56. Incubate for 30 min at 37{ring}C.

57. Slowly underlay 7 mL of Lympholyte-M to the bottom of the cell suspension, and centrifuge at 580g (1800 rpm) for 10 min at room temperature.

58. Carefully transfer the interphase layer to a 50-mL tube, fill it with OP9 medium, and centrifuge again at 450g (1600 rpm) for 10 min at room temperature.
These cells will contain ~80%-95% CD117 + hematopoietic progenitor cells, including an enriched fraction of CD117 + Sca-1 + Lineage neg cells. These CD24-depleted fetal liver cells can be seeded directly onto OP9 or OP9-DL1 cells for co-culture. Alternatively, the HSC-enriched fetal liver cell suspension can be stained and sorted for CD117 + Sca-1 hi Lin neg cells by flow cytometric cell sorting.

59. Seed 4-6 x 104 cells in 10 mL of OP9 medium per 10-cm dish of 80%-90% confluent OP9 or OP9-DL1 cells, and add 5 ng/mL Flt-3L and 1 ng/mL IL-7. Place the dishes in the incubator.
The OP9 and OP9-DL1 cells were prepared in Steps 1-12.

Day 4: Medium Change
Steps 60-62 are optional if low numbers (<104 ) of HSCs are used at the start of the co-culture.

60. Pipette the medium off carefully, and centrifuge at 400g (1500 rpm) for 5 min at 4°C.
Clusters of round, shiny cells should be present, with some cells detached in suspension.

61. Resuspend cells in 10 mL of OP9 medium containing cytokines.

62. Transfer the medium and cells to the same dish and return to the incubator.

Day 7: No-Trypsin Passage

63. Disaggregate cells without the use of trypsin by pipetting the cells up and down until the OP9 cell monolayer is completely disrupted from the plate and broken into small pieces.
Large clusters of round, shiny cells should be present throughout the dish, with some cells detached in suspension.

64. Filter the cells through a 40-µm cell strainer to remove clumps of OP9 cells.
Some of the OP9 cells may pass through the cell strainer, but these will not affect the co-culture.

65. Wash the 10-cm dish with 6 mL of PBS, and filter the PBS through the same cell strainer.

66. Centrifuge the cells at 400g (1500 rpm) for 5 min at 4°C, and resuspend them in 10 mL of OP9 medium containing cytokines.

67. Seed cells onto new 10-cm dishes of 80%-90% confluent OP9 or OP9-DL1 cells. Return the cells to the incubator.
Split the co-culture 1-to-4 or 1-to-10 (depending on the starting number of HSCs).

Day 12 and Onward, at 4-d to 5-d Intervals: No-Trypsin Passage

68. Follow Steps 63-67.
At these time points, it is best not to split the culture more than 1-to-4. Refer to Figure 4 for the expected results at the different time points.

View larger version (18K): [in this window] [in a new window]

Figure 4. Flow cytometry analysis for the indicated cell surface markers of fetal liver/OP9 or OP9-DL1 co-cultures at the indicated time points and culture conditions.

Protocol 4: Bone-Marrow-Derived Hematopoietic Stem Cell Differentiation on OP9-DL1 Cells

Day 0: Initiation of Bone Marrow Co-Culture

69. Obtain femur and tibia pairs from 4- to 8-wk-old mice, using scissors and forceps.

70. Crush the bones using a glass stopper against a tissue culture dish containing 3 mL of OP9 medium to release the marrow. Filter the bone fragments and cells through a 40-µm cell strainer.

71. Centrifuge the cells at 400g (1500 rpm) for 5 min at 4°C, resuspend them in 5 mL of BDPharmLyse to remove red blood cells, and incubate for 5 min at room temperature.

72. Fill the tube with OP9 medium, centrifuge again at 400g (1500 rpm) for 5 min at 4°C, and resuspend the cells in the appropriate buffer for staining and cell sorting.

73. Using flow cytometric cell sorting, isolate CD117+ Sca-1+ Lineageneg (CD4, CD8, CD11b, CD19, CD45R, CD161, GR.1, Ter119) progenitor cells or HSCs.
The sorting step can be expedited by first depleting the Lineage + cells with MACS prior to flow cytometric cell sorting.

74. Seed 2-5 x 105 cells in 10 mL of OP9 medium per 10-cm dish of 80%-90% confluent OP9 or OP9-DL1 cells, and add 5 ng/mL Flt-3L and 1 ng/mL IL-7.
The OP9 and OP9-DL1 cells were prepared in Steps 1-12.

75. Place the dishes in the incubator.

Day 5: No-Trypsin Passage

76. Disaggregate cells without the use of trypsin by pipetting the cells up and down until the OP9 cell monolayer is completely disrupted from the plate and broken into small pieces.
Large clusters of round, shiny cells should be present throughout the dish, with some cells detached in suspension. Bone marrow co-cultures underperform when split or passaged too sparsely.

77. Filter cells through a 40-µm cell strainer.

78. Wash the 10-cm dish with 6 mL of PBS, filter through the same cell strainer, and centrifuge at 400g (1500 rpm) for 5 min at 4°C.

79. Resuspend the cells in 10 mL of OP9 medium containing cytokines, and seed the cells onto 10-cm dishes of 80%-90% confluent fresh OP9 or OP9-DL1 cells.

80. Return the cells to the incubator.

Day 8 and Onward, at 4-d to 5-d Intervals: No-Trypsin Passage

81. Follow Steps 76-80.
Refer to Figure 5 for the expected results at the different time points.

View larger version (17K): [in this window] [in a new window]

Figure 5. Flow cytometry analysis for the indicated cell surface markers of bone marrow/OP9 or OP9-DL1 co-cultures at the indicated time points and culture conditions.

TROUBLESHOOTING

Problem: The OP9 cells are more than 80%-90% confluent.

Solution: It is important when creating working stocks of OP9 cells for freezing that the cells are never allowed to become more than 80%-90% confluent. Monitor cell density to avoid exceeding this level of confluency. If the OP9 cells are too confluent when a co-culture is in progress, it is fine to use these OP9 cells for an ongoing experiment. However, do not use these cells for making stocks.

Problem: The OP9 cells are becoming adipocytic (full of fatty globules).

Solution: This occurs when the OP9 cells become more than 80%-90% confluent. Monitor cell density to avoid exceeding this level of confluency. Although the presence of some adipocytic cells will not affect the culture, large numbers of these cells will have a negative impact. Some lots of fetal bovine serum (FBS) can make the cells turn adipocytic more quickly. In addition, avoid FBS substitutes or growth supplements, as these will increase the number of adipocytic cells.

Problem: The OP9 cells are growing too quickly.

Solution: Reduce the amount of serum. When an almost confluent plate is split 1-to-4 or 1-to-5, the cells should be confluent again in 2 d. If the cells are growing more rapidly than this, reduce the serum and discard these cells from the working stocks.

Problem: The OP9 cells did not recover well after thawing them.

Solution: Do not store the frozen OP9 cells at �80°C for extended periods, because this will decrease the recovery of the cells when thawed. It is best to store them in liquid nitrogen. Freeze cells in 90% FBS and 10% DMSO (i.e., freezing medium for OP9 cells).

Problem: The OP9 cells are not irradiated and will become overconfluent.

Solution: It is more important that the cells are not overconfluent when seeding with hematopoietic cells than at later time points. The co-cultures are passaged every 4-5 d to take care of this issue, because the OP9 cells will only stop proliferating when fully confluent, due to contact inhibition. During the passage steps, the cells are filtered out with cell strainers, to remove the excess number of OP9 cells. Although this does not eliminate all of the OP9 cells, it does limit them, and the remaining cells will not affect the co-culture.

Problem: There is no precise seeding cell count for the OP9 cells.

Solution: It is difficult to use a particular number of OP9 cells because the cells will grow at different rates. The important issue is if the cells are near confluent, rather than if there is a specific number of cells on the plate.

Problem: The ESC single cell suspension contains mEF feeder layer.

[Step 30]

Solution: Because the mEF feeder layer is irradiated, these cells will not interfere with the co-culture.

Problem: The monolayer is breaking up as the semiadherent cells are being washed off the dish.

[Step 37]

Solution: It is acceptable if some of the monolayer is washed off the dish. Because the cells will be filtered, using a cell strainer, these clumps will not pass though the filter.

Problem: The OP9 monolayer peels away from the dish.

Solution: The OP9 cells were overconfluent at the time of seeding; the co-cultures should be passaged onto new OP9 cells.

Problem: The OP9-DL1 cells appear in the FL1 channel.

Solution: The OP9-DL1 cells also express green fluorescent protein (GFP) as part of the integrated retroviral bicistronic construct from which both Dll1 and GFP are coexpressed. During flow cytometric analysis, the OP9-DL1 will fluoresce in the FITC (FL1) channel. These cells can be gated-out using forward and side-scatter criteria, because these cells are larger and more granular than hematopoietic cells. Additionally, staining for CD45 expression can be used to conclusively detect hematopoietic cells as opposed to OP9-DL1 cells (see Fig. 3 ).

Problem: Bone marrow cultures show delayed kinetics generating CD4+ CD8+ cells.

Solution: The recommended IL-7 final concentration is 1 ng/mL (Step 74). However, this concentration can be lowered to 0.1-0.5 ng/mL starting on day 12 of the culture. Reduction of IL-7 will enhance the differentiation of CD4+ CD8+ cells from the earlier progenitor subset.

Problem: Natural killer (NK) cells do not proliferate or appear in the culture.

Solution: Adding IL-2 (1-10 ng/mL) or IL-15 (5-10 ng/mL) to the culture starting at day 8 for the ESC cultures and day 0 for the bone marrow and fetal liver cultures will allow for more a robust NK cell population to be generated.

Problem: The cultures appear to be contaminated.

Solution: Discard the co-cultures, and decontaminate the work areas both inside and outside the biosafety cabinet and incubator. To prevent contamination when working with ex vivo material (such as bone marrow), normocin and fungizone can be used, but these agents may interfere with ESC cultures. Gentamicin can be added to the OP9 medium to limit potential contamination in the cultures.

DISCUSSION

The differentiation of T- and B-lineage cells from multiple sources of stem cells can be readily studied in vitro using the OP9-DL1 or OP9 co-culture system, respectively (Zúñiga-Pflücker 2004 ). This methodology has been used by several hundred laboratories to address numerous questions, because it is an effective way to determine lymphocyte lineage commitment, regulation of lymphocyte differentiation, and other aspects of lymphocyte development (de Pooter and Zúñiga-Pflücker 2007 ). In particular, the OP9-DL1 system has been useful in addressing questions about the cellular and molecular regulation of T-lymphocyte lineage commitment, pre-T-cell receptor signaling (β-selection), functional characteristics of progenitor T-cells, and maturation of functional CD8 T-cells (for review, see de Pooter and Zúñiga-Pflücker 2007 ). However, a few challenges remain, such as whether CD4 T-cells can be readily generated and determination of the rules regarding positive and/or negative selection of T-cells in this culture system. Although the present protocol only addresses the use of mouse stem cells, this system has been successfully adapted for the generation of T-lineage cells from human HSCs (La Motte-Mohs et al. 2005 ; Awong et al. 2008 ).

Although OP9 cells provide a robust stromal cell monolayer to support hemato-lymphopoiesis, OP9 cells should not be kept in continuous culture for extended periods of time (>6 wk) before initia, ting a co-culture. Extended periods of culturing will increase the likelihood that the cells will deteriorate and become less effective. Thus, it is highly recommended that a co-culture be started within 2 wk of thawing the OP9 cells, which has the added advantage that the same OP9 cells can be used for the entire co-culture period. Another variable to manage is the FBS used in the co-cultures, which can affect the growth and proliferation of the co-cultures. Test several lots of serum to determine the one that performs well and yields results similar to those shown in these protocols (Figs. 3 , 4 , and 5 ).

The differentiation of lymphocytes from ESCs can be readily observed 10-12 d after the initiation of the co-cultures (Fig. 3 ). Prior to this stage, the majority of the cellular subsets are comprised of erythro-myeloid cell lineages. Beyond the second week of co-culture, B- or T-lineage cells (OP9 or OP9-DL1 co-cultures, respectively) predominate and expand rapidly. These cells display and follow a normal pattern of lineage differentiation, such as lineage-specific differentiation checkpoint events, to yield functional lymphocytes (Schmitt and Zúñiga-Pflücker 2002 ; Schmitt et al. 2004 ; de Pooter et al. 2006 ).

As seen in Figures 4 and 5 , the differentiation of fetal liver HSCs to the B- or T- cell lineage takes place with faster temporal kinetics than that of bone marrow HSCs. Additionally, ESC co-cultures, as expected, require a much longer period of time to yield lymphocytes because of the additional time involved in the initial hematopoietic induction and differentiation steps when starting from an ESC (Fig. 3 ). Although different stem cells give rise to lymphocytes with different kinetics, the overall differentiation steps follow a similar pattern. Additionally, if lymphocyte progenitors (CD117+ CD127+ ) from the bone marrow or thymus or more downstream progenitors such as pre-B- (CD117+ CD19+ ) or pre-T-cells (CD44+ CD25+ CD3� CD4� CD8� ) are used, then the differentiation kinetics are accelerated from those shown in Figures 4 and 5 .

ACKNOWLEDGMENTS

We thank the many members of the Zúñiga-Pflücker lab who over the years have helped to develop, test, and optimize the methods listed above. In particular, we are grateful to James Carlyle, Sarah Cho, Renée de Pooter, Dzana Dervovic, Ross La Motte-Mohs, Thomas Schmitt, Gladys Wong, and John Xu. Additionally, we thank Korosh Kianizad and Tina Wang for their helpful comments. We also appreciate the kind assistance and generosity of Tasuko Honjo and Toru Nakano for initially sharing the OP9 cells and expertise with us. These protocols were developed with support from the Canadian Institutes of Health Research and with funds from the Canadian Cancer Society.

REFERENCES

1. Awong, G., La Motte-Mohs, R.N., and Zúñiga-Pflücker, J.C. 2008. In vitro human T cell development directed by notch-ligand interactions. Methods Mol. Biol. 430: 135142.[Medline]
2. Bunting, K.D., ed. 2008. Hematopoietic stem cell protocols . Humana Press, Clifton, NJ.
3. de Pooter, R. and Zúñiga-Pflücker, J.C. 2007. T-cell potential and development in vitro: The OP9-DL1 approach. Curr. Opin. Immunol. 19: 163168.[Medline]
4. de Pooter, R.F., Schmitt, T.M., de la Pompa, J.L., Fujiwara, Y., Orkin, S.H., and Zúñiga-Pflücker, J.C. 2006. Notch signaling requires GATA-2 to inhibit myelopoiesis from embryonic stem cells and primary hemopoietic progenitors. J. Immunol. 176: 52675275.[Abstract/Free Full Text]
5. Klug, C.A. and Jordan, C.T., eds. 2002. Hematopoietic stem cell protocols . Humana Press, Totowa, NJ.
6. La Motte-Mohs, R.N., Herer, E., and Zúñiga-Pflücker, J.C. 2005. Induction of T-cell development from human cord blood hematopoietic stem cells by Delta-like 1 in vitro. Blood 105: 1431439.[Abstract/Free Full Text]
7. Nagy, A., Gertsenstein, M., Vintersten, K., and Behringer, R. 2006a. Preparing mouse embryo fibroblasts. Cold Spring Harb. Protoc. doi: 10.1101/pdb.prot4398.[Free Full Text]
8. Nagy, A., Gertsenstein, M., Vintersten, K., and Behringer, R. 2006b. De novo isolation of embryonic stem (ES) cell lines from blastocysts. Cold Spring Harb. Protoc. doi: 10.1101/pdb.prot4403.[Free Full Text]
9. Schmitt, T.M. and Zúñiga-Pflücker, J.C. 2002. Induction of T cell development from hematopoietic progenitor cells by delta-like-1 in vitro. Immunity 17: 749756.[Medline]
10. Schmitt, T.M., de Pooter, R.F., Gronski, M.A., Cho, S.K., Ohashi, P.S., and Zúñiga-Pflücker, J.C. 2004. Induction of T cell development and establishment of T cell competence from embryonic stem cells differentiated in vitro. Nat. Immunol. 5: 410417.[Medline]
11. Zúñiga-Pflücker, J.C. 2004. T-cell development made simple. Nat. Rev. Immunol. 4: 672.[Medline]