In this study we present an in vitro culture system that can efficiently generate pDCs by co-culturing common lymphoid progenitors with AC-6 feeder cells in the presence of Flt3 ligand.
Plasmacytoid dendritic cells (pDCs) are powerful type I interferon (IFN-I) producing cells that are activated in response to infection or during inflammatory responses. Unfortunately, study of pDC function is hindered by their low frequency in lymphoid organs, and existing methods for in vitro DC generation predominantly favor the production of cDCs over pDCs. Here we present a unique approach to efficiently generate pDCs from common lymphoid progenitors (CLPs) in vitro. Specifically, the protocol described details how to purify CLPs from bone marrow and generate pDCs by coculturing with γ-irradiated AC-6 feeder cells in the presence of Flt3 ligand. A unique characteristic of this culture system is that the CLPs migrate underneath the AC-6 cells and become cobblestone area-forming cells, a critical step for expanding pDCs. Morphologically distinct DCs, namely pDCs and cDCs, were generated after approximately 2 weeks with a composition of 70-90% pDCs under optimal conditions. Typically, the number of pDCs generated by this method is roughly 100-fold of the number of CLPs seeded. Therefore, this is a novel system with which to robustly generate the large numbers of pDCs required to facilitate further studies into the development and function of these cells.
Dendritic cells (DCs) are professional antigen-presenting cells that play an important role in controlling immune responses1. While DCs are heterogeneous, they can be broadly classified into two populations, conventional DCs (cDCs) and plasmacytoid DCs (pDCs)2,3. In addition to lymphoid organs, cDCs and pDCs are also found in tissues including lung, intestine, and skin2. The morphology of cDCs and pDCs differs, with cDCs exhibiting dendrite-like projections and the shape of pDCs being more plasma cell-like. In addition, the common mouse DC marker, CD11c, is more highly expressed on cDCs than on pDCs. Moreover, cDCs can be further divided into CD11b+CD24- cDCs and CD11b-CD24+ cDCs, both of which express higher levels of MHC class II and costimulatory molecules than do pDCs2. Mature pDCs, on the other hand, have been shown to selectively express Siglec-H and BST24. Functionally, cDCs are better antigen presenting cells than are pDCs; however, pDCs can produce a vast amount of type I interferon upon virus infection or inflammatory stimulation5.
Both cDCs and pDCs are short-lived, and therefore, must be constantly replenished from progenitors within the bone marrow (BM)6,7. Adoptive transfer of common myeloid progenitors (CMPs) and common lymphoid progenitors (CLPs) into lethally-irradiated mice demonstrates that cDCs and pDCs can be generated from both lineages8-10. However, there is a subset of pDCs that express RAG1/2 and possess rearranged D-J segments at the IgH locus11,12. These cells also share molecular similarities with B lymphocytes, including expression of the B220 marker, nucleic acid-sensing receptors (TLR7/TLR9), and transcription factors (Spib and Bcl11a)13, features all believed to be signatures of the lymphoid lineage. Therefore, CLPs may be good choices for in vitro generation of pDCs because of lineage similarity.
While the frequency of both cDCs and pDCs in humans and mice is very low6, DCs, particularly cDCs, can be generated in vitro from BM or progenitors in the presence of cytokines, such as GM-CSF11,14 or Flt3 ligand (FL) using feeder-free systems 11,15,16. Unfortunately, however, it is not possible to produce large numbers of pDCs in vitro using FL11,15,16. Previously we demonstrated that pDCs can be efficiently generated in vitro from CLPs using the AC-6 feeder system17. The advantage of using the AC-6 stromal cell line in the culture system is that it provides cell-cell contact and secretion of cytokines that support the generation of large numbers of pDCs from CLPs. Although production with this system is quite robust, careful replication of the procedures described below is required in order to ensure efficient generation of pDCs.
C57BL/6 wild-type mice were purchased from the National Laboratory Animal Center (NLAC), Taiwan. All mice were bred and kept under specific pathogen-free conditions. Protocols and animal use procedures were reviewed and approved by the Institutional Animal Care and Use Committee of National Taiwan University College of Medicine (Permit Number: 20120075). In addition, researchers made every effort to reduce the potential for pain, suffering, or distress in the animals while performing experiments. All procedures described were carried out at RT while wearing gloves.
1. Preparation of AC-6 Feeder cells
Note: The AC-6.21 (AC-6 in short) stromal cell line18 (provided by I. Weissman, Stanford University) should be maintained in RPMI supplemented with 15% heat-inactivated fetal bovine serum (FBS). To serve as feeder cells, AC-6 cells are γ-irradiated with 3,000 rad (30Gy) one day before the co-culture to prevent their proliferation. Note that AC-6 cells tend to lose their differentiation-supporting ability if left overcrowded during maintenance, or if their passage number is over 20.
- Wash AC-6 cells once with 1 ml Dulbecco's Phosphate-Buffered Saline (DPBS) and remove DPBS by aspiration. Treat AC-6 cells with 0.8 ml trypsin solution (0.05 % trypsin and 0.5 mM EDTA in DPBS) for 3-5 min at 37oC, 5% CO2 and then stop the reaction by diluting it with 10 volumes of culture medium to make single cell suspension.
- Seed AC-6 cells at 5.9x104 cells/well into 12-well plates and incubate at 37oC, 5% CO2 O/N.
Note: The density of AC-6 cells is very important as a lower density will favor cDC generation. Seeding at a density of 5.9x104 cells/well will allow AC-6 to reach confluency by the next day, which is optimal for derivation of pDCs from CLPs.
- Irradiate AC-6 cells at 3,000 rad (30 Gy) using γ-irradiator.
Note: The dose used to irradiate AC-6 is optimized to prevent the cells from proliferating and yet keep them viable long enough to provide the cytokines and cell-cell contact required for differentiation of DCs from CLPs.
- Aspirate medium and replace with 1 ml complete RPMI (RPMI with 10 % heat-inactivated FBS, 50 µg/ml gentamycin and 50 µM β-mercaptoethanol).
2. Isolate BM Cells from Mice
- Sacrifice mice by CO2 asphyxiation and cervical dislocation. Place the mice on a dissection tray and sterilize with 70% ethanol. Make an incision at the mid-abdomen and remove the skin from the distal part of the mouse including the skin covering the lower extremities.
- Dissect mice in a semi-sterile hood. To release the femur and tibiae, clip the femur above and tibiae below the knee joint. Detach the muscles from the bones using scissors and place the bones in a 6 cm petri dish containing 5 ml complete RPMI.
- Move to a sterile culture hood. Fill a 3 ml syringe connected to a 27G needle with complete RPMI. Cut off both ends of the bones using scissors, and insert the 27G needle to flush the marrow out of the bones by gently injecting complete RPMI once from each end.
- Prepare a single cell suspension by gentle pipetting the cells up and down 3-5 times in the culture dish using the syringe without the needle.
- Centrifuge the cells for 5 min at 500 x g and decant the supernatant.
- Lyse red blood cells with 1 ml ACK buffer (150 mM NH4Cl, 10 mM KHCO3, 0.1 mM EDTA) incubating for 1 min and stopping the reaction by diluting with 10 volumes of complete RPMI. Allow cells to stand for 5 min in order to pellet down dead cell clumps and tissue debris by gravity.
Note: Do not stand more than 5 min as this will result in cell loss due to settling of viable cells.
- Slowly decant cell-containing supernatant to a new tube leaving debris behind in the original tube.
- Centrifuge for 5 min at 500 x g to pellet the cells, then remove and discard the supernatant.
- Resuspend total BM cells (typically 4-6 x 107 BM cells are obtained from one mouse) in 100 µl FACS buffer (1x PBS+ 2% FBS+ 1mM EDTA).
3. FACS Analysis and Sorting of CLPs
- Add anti-CD16/32 (hybridoma supernant clone 2.4G2, 50 µl/reaction or 1-2 µg/reaction) into cells in FACS buffer (step 2.9) for 1-2 min in order to block Fc receptors.
Note: Anti-CD16/32 is also called Fc block, which is added to prevent non-specific binding of antibodies to cells expressing Fc receptors including granulocytes, monocytes, and B lymphocytes.
- Simultaneously stain cells with the following fluorescent dye-labeled antibodies (for 4x107 cells) for 15 min on ice, avoiding ambient light. Antibodies: PE-conjugated lineage markers (0.2 µg each) including anti-CD3 (17A2), anti-CD8 (53-6.7), anti-B220 (RA3-6B2), anti-CD19 (eBio1D3), anti-CD11b (M1/70), anti-Gr-1 (RB6-8C5), anti-Thy1.1 (HIS51), anti-NK1.1 (PK136), anti-TER119 (TER-119), and anti-MHC-II (NIMR-4), 0.2 µg anti-c-Kit-PerCP-Cy5.5 (2B8), 1 µg anti-Sca-1-FITC (D7), 0.4 µg anti-M-CSFR-APC (AFS98), and 0.2 µg anti-IL-7Rα-PE-Cy7 (A7R34).
- Wash cells with 3 ml FACS buffer, and centrifuge for 5 min at 500 x g.
- Resuspend the cells in 300 µl FACS buffer and filter cells through a 40 µM cell strainer.
- Wash the tube with an additional 100 µl FACS buffer to recover any remaining cells and filter into a sterile FACS sorting tube using the same cell strainer.
- Perform flow cytometric analysis immediately after the staining using appropriate filters and voltages for signal detection. Use 488 nm laser to detect the FITC-, PerCP-Cy5.5-labeled antibodies, 561 nm laser to detect the PE- and PE-Cy7-labelded antibodies and 633 nm laser to detect the APC-labeled antibody.
- Sort out CLPs according to the following markers lin-c-kitintSca-1intM-CSFR-IL-7Rα+ (as stained in step 3.2) using a cell sorter. Collect the sorted cells in a 15 ml tube containing 8 ml of complete RPMI as a cushion.
- Record the absolute number of sorted cells as shown by the cell sorter at the end of the run. Typically, 5x104 CLPs can be obtained from the BM of one mouse.
4. Coculture CLPs and AC-6
- Centrifuge the sorted cells from step 3.7 for 10 min at 500 x g, remove the supernatant and resuspend the cell pellet with enough complete RPMI to obtain a cell density around 5x104 cells/ml. Seed 500 cells/well into the 12-well plate containing feeder cells prepared in step 1.4.
- Add 100 ng/ml FL19 (hu-Flt3L-Ig generated in-house using an expression system provided by M. Manz, University Hospital Zürich, Switzerland) and incubate at 37oC, 5% CO2 with periodic visual monitoring of DC development under a microscope.
Note: Commercially available recombinant human FL or mouse FL can be used as a substitute for hu-Flt3L-Ig to support DC development.
- Collect the cells in the supernatant at day 12 and wash the wells once with 0.5 ml complete RPMI medium combining the resulting supernatants. Add 0.5 ml fresh medium and scrap the adherent cells with a cell scraper.
- Combine the cell-containing medium from both parts and centrifuge for 5 min at 500 x g.
- Decant supernatant and resuspend cells in 50 µl FACS buffer. Add 50 µl anti-CD16/32 hybridoma supernatant and incubate for 1-2 min to block Fc receptors.
- Enumerate and stain all the cells with the following antibodies (0.05 µg each) anti-CD11c-APC (N418), anti-CD11b-FITC, and anti-B220-PE.
- Wash and centrifuge the cells as described in step 3.3.
- Resuspend the cells in 100 µl FACS buffer, gate on CD11c+ and analyze for cDCs (CD11c+CD11b+B220-) and pDCs (CD11c+CD11b-B220+)17.
A total of 4-6x107 BM cells are typically isolated from femurs and tibiae of one 6-8 wk-old, wild-type C57BL/6 mouse. To sort out CLPs, total BM cells are stained with PE-conjugated antibodies against lineage markers (CD3, CD8, B220, CD19, CD11b, Gr-1, Thy1.1, NK1.1, TER119, and MHC-II), anti-c-Kit-PerCP/Cy5.5, anti-Sca-1-FITC, anti-M-CSFR-APC and anti-IL-7Rα-PE/Cy7, analyzed and sorted with a cell sorter. The sorting strategy for CLP is shown in Figure 1. Typically, 5x104 CLPs are obtained from one C57BL/6 mouse. To obtain more CLPs, BM cells from 3-4 mice may be combined and protocol adjusted accordingly before sorting.
Following the coculture of 500 CLPs with 5.9x106 AC-6 feeder cells, cobblestone area-forming cells (CAFC) start to appear by day 3-4. By day 8, round-shaped pDCs appear and suspend on top of the CAFC (Figure 2A). The number of pDCs peaks around day 12-14 at which point pDCs start to undergo apoptosis following their full development, and the numbers begin to gradually decrease. cDC colonies appear later than pDC colonies, at day 6-8, and the morphology of cDCs are larger, spindle-like, and adherent (Figure 2B).
Typically, 5-6x104 cells can be generated from 500 CLPs after 2 weeks of coculture. After staining the cells with anti-CD11c-APC, anti-CD11b-FITC and anti-B220-PE, analysis by flow cytometry shows that the percentage of pDCs (CD11c+CD11b-B220+) and cDCs (CD11c+CD11b+B220-) is 91% and 6%, respectively (Figure 3). Therefore, with this method pDCs can be expanded as much as 100-fold.
Figure 1. Gating strategies for CLPs. BM cells isolated from one mouse were stained with fluorescently-labeled antibodies and analyzed for CLPs, which were defined as lin-c-Kitint Sca-1intM-CSFR-IL-7Rα+. Please click here to view a larger version of this figure.
Figure 2. Morphology of cDC and pDC colonies in CLP-AC-6 cocultures. 500 CLPs were cocultured with γ-irradiated AC-6 feeder cells in the presence of 100 ng/ml FL. (A) pDC colonies at day 8, 10, and 12, and (B) cDC colonies at day 8, 12, and 16 were photographed under light microscopy at 200 x (scale bar 50 µm). Please click here to view a larger version of this figure.
Figure 3. Flow cytometric analysis of CLP-derived pDCs and cDCs. Five hundred CLPs and γ-irradiated AC-6 feeder cells (5.9x104/ml) were cocultured in vitro for 12 days in the presence of FL (100 ng/ml). The cells were stained with anti-CD11c-APC, anti-CD11b-FITC, and anti-B220-PE and analyzed by flow cytometry. Cells were first gated on CD11c+ and then analyzed for cDCs (CD11c+CD11b+B220-) and pDCs (CD11c+CD11b-B220+). Please click here to view a larger version of this figure.
Here we describe an in vitro culture system for the robust generation of DCs, and pDCs in particular, from a small number of CLPs. The uniqueness of this culture system is due to the use of AC-6 cells, a stromal cell line, as feeders. This approach has been shown to provide not only the cytokines, such as IL-7, SCF, M-CSF and FL20, but also the cell-cell contact21 necessary to support DC development. AC-6 cells have been used previously to facilitate the study of in vitro DC development from several progenitors8,16,22. We found that it is critical to seed an optimal density of AC-6 feeder cells to support the efficient differentiation of pDCs from CLPs. Specifically, seeding 5.9x104 γ-irradiated AC-6 cells in 12-well plates results in confluency the next day, which is suitable for pDC development. A lower AC-6 cell density (3.9x104/well in 12-well plate) tends to support cDC development24. Therefore, it is recommended to first seed several concentrations of AC-6 cells in order to optimize conditions when scaling up the culture system23. Moreover, it is also important that the AC-6 cells be scrupulously maintained, including careful attention to passage number, and never allowed to reaching confluency. If left overcrowded or if their passage number is greater than 20, these cells tend to lose the ability to support hematopoiesis in vitro. Additionally, the concentration of FL also affects DC development. Specifically, at low doses (25-50 ng/ml) cDCs are preferentially developed, while at high doses (100-200 ng/ml) pDCs are predominantly generated24. In addition to AC-6 cell density and FL dosage, careful selection of a batch of FBS that efficiently supports DC development in this culture system is central to the success of the procedure.
While in vitro-derived cDCs and pDCs can be distinguished by their surface markers, namely CD11c+CD11b+B220- and CD11c+CD11b-B220+, respectively, their morphologies are also discernible by microscopy. cDCs are larger, spindle-shaped and are adherent, while pDCs are smaller, round-shaped and grow in suspension (Figure 2). The kinetics of DC differentiation also differs, with pDCs appearing earlier and cDCs later. Additionally, the process of pDC development is significantly influenced by AC-6 cell density with higher AC-6 density speeding up the developmental process. It should be noted that pDCs start to die following complete development and thus, the ratio of cDCs versus pDCs in this culture system gradually increases at later time points. Therefore, periodic monitoring of DC development by microscopy and flow cytometry is highly recommended in order to determine when pDCs start to undergo apoptosis.
Of note, the described method can also be applied to other DC progenitors, such as CDPs; however, the main DC population generated from these cells are cDCs17, even at high AC-6 densities or FL doses. The low pDC potential of CDPs is consistent with reports from different groups using either CDP-feeder11 or feeder-free16 systems, two different in vitro culture systems. These results suggest that the AC-6 feeder system can faithfully reflect the differentiation potential of different progenitors.
The pDCs generated from this culture system are CD11c+CD11b-B220+. Moreover, they express relatively high levels of Tcf4 (encodes E2-2) and Rag1, two pDC-specific genes, but lower levels of Id2, a cDC-specific gene, than do cDCs17. Even though these pDCs lack Siglec-H and BST2, markers considered characteristic of murine pDCs, they are still able to produce higher levels of IFN-I than do cDCs and upregulate CD86 when infected with vesicular stomatitis virus (VSV)24. This suggests that these pDCs, although incompletely differentiated, are still functional. Moreover, we have already successfully used this method to delineate a synergistic role of IFN-I and FL in pDC development from CLPs17, confirming that this technique is valuable for the study of DC development.
In conclusion, here we present a simple, but efficient, method to generate pDCs from a small number of CLPs using the AC-6 feeder coculture system, which can facilitate developmental or functional studies of pDCs.
Authors have nothing to disclose.
We are grateful to Drs. Markus Manz and Irving Weissman for providing reagents. We also acknowledge the service provided by the Flow Cytometric Analysis and Cell Sorting Core Facility of the First and Second Core Laboratory at National Taiwan University College of Medicine and NTU hospital, respectively. This work was supported by the Ministry of Science and Technology, Taiwan (MOST 102-2320-B-002-030-MY3) and the National Health Research Institutes, Taiwan (NHRI-EX102-10256SI and NHRI-EX103-10256SI).
|Anti-mouse Ly6g/Ly6c (PE), clone RB6-8C5||Biolegend||108408||linage marker|
|Anti-mouse NK1.1 (PE), clone PK136||Biolegend||108708||linage marker|
|Anti-mouse CD11b (PE), cloneM1/70||Biolegend||101208||linage marker|
|Anti-mouse CD19 (PE), clone eBio1D3||Biolegend||115508||linage marker|
|Anti-mouse B220 (PE), clone RA3-6B2||Biolegend||103208||linage marker/FACS|
|Anti-mouse CD3 (PE), clone 17A2||Biolegend||100308||linage marker|
|Anti-mouse CD8a (PE), clone 53-6.7||Biolegend||100707||linage marker|
|Anti-mouse MHC-II (PE), clone NIMR-4||Biolegend||107608||linage marker|
|Anti-mouse Ter119 (PE), clone TER-119||Biolegend||116208||linage marker|
|Anti-mouse Thy1.1 (PE), clone HIS51||eBioscience||12-0900-83||linage marker|
|Anti-mouse M-CSFR (APC), clone AFS98||Biolegend||135510||FACS|
|Anti-mouse c-Kit (PerCP/Cy5.5), clone 2B8||Biolegend||105824||FACS|
|Anti-mouse Sca-1 (FITC), clone D7||Biolegend||108106||FACS|
|Anti-mouse IL-7Ra (PE/Cy7), clone A7R34||Biolegend||135014||FACS|
|Anti-mouse CD11c (PerCP/Cy5.5), clone N418||Biolegend||117328||FACS|
|Anti-mouse CD11b (FITC), clone M1/70||Biolegend||101206||FACS|
|FACSAria III||BD Biosciences||Cell sorter|
|FACS sorting tube||BD Biosciences||352054|
|FlowJo||FlowJo LLC||Flow analysis sofrware|
- Hartwig, C., et al. Fcgamma receptor-mediated antigen uptake by lung DC contributes to allergic airway hyper-responsiveness and inflammation. Eur. J. Immunol. 40, 1284-1295 (2010).
- Merad, M., Sathe, P., Helft, J., Miller, J., Mortha, A. The dendritic cell lineage: ontogeny and function of dendritic cells and their subsets in the steady state and the inflamed setting. Annu. Rev. Immunol. 31, 563-604 (2013).
- Mildner, A., Jung, S. Development and Function of Dendritic Cell Subsets. Immunity. 40, 642-656 (2014).
- Swiecki, M., Colonna, M. Unraveling the functions of plasmacytoid dendritic cells during viral infections, autoimmunity, and tolerance. Immunol. Rev. 234, 142-162 (2010).
- Liu, Y. J. IPC: professional type 1 interferon-producing cells and plasmacytoid dendritic cell precursors. Annu. Rev. Immunol. 23, 275-306 (2005).
- Merad, M., Manz, M. G. Dendritic cell homeostasis. Blood. 113, 3418-3427 (2009).
- Ghosh, H. S., Cisse, B., Bunin, A., Lewis, K. L., Reizis, B. Continuous expression of the transcription factor e2-2 maintains the cell fate of mature plasmacytoid dendritic cells. Immunity. 33, 905-916 (2010).
- Karsunky, H., Merad, M., Cozzio, A., Weissman, I. L., Manz, M. G. Flt3 ligand regulates dendritic cell development from Flt3+ lymphoid and myeloid-committed progenitors to Flt3+ dendritic cells in vivo. J. Exp. Med. 198, 305-313 (2003).
- Manz, M. G., Traver, D., Miyamoto, T., Weissman, I. L., Akashi, K. Dendritic cell potentials of early lymphoid and myeloid progenitors. Blood. 97, 3333-3341 (2001).
- D'Amico, A., Wu, L. The early progenitors of mouse dendritic cells and plasmacytoid predendritic cells are within the bone marrow hemopoietic precursors expressing Flt3. J. Exp. Med. 198, 293-303 (2003).
- Sathe, P., Vremec, D., Wu, L., Corcoran, L., Shortman, K. Convergent differentiation: myeloid and lymphoid pathways to murine plasmacytoid dendritic cells. Blood. 121, 11-19 (2013).
- Shigematsu, H., et al. Plasmacytoid dendritic cells activate lymphoid-specific genetic programs irrespective of their cellular origin. Immunity. 21, 43-53 (2004).
- Reizis, B., Bunin, A., Ghosh, H. S., Lewis, K. L., Sisirak, V. Plasmacytoid dendritic cells: recent progress and open questions. Annu. Rev. Immunol. 29, 163-183 (2011).
- Inaba, K., et al. Generation of large numbers of dendritic cells from mouse bone marrow cultures supplemented with granulocyte/macrophage colony-stimulating factor. J. Exp. Med. 176, 1693-1702 (1992).
- Naik, S. H., et al. Development of plasmacytoid and conventional dendritic cell subtypes from single precursor cells derived in vitro and in vivo. Nat. Immunol. 8, 1217-1226 (2007).
- Onai, N., et al. Identification of clonogenic common Flt3+M-CSFR+ plasmacytoid and conventional dendritic cell progenitors in mouse bone marrow. Nat. Immunol. 8, 1207-1216 (2007).
- Chen, Y. L., et al. A type I IFN-Flt3 ligand axis augments plasmacytoid dendritic cell development from common lymphoid progenitors. J. Exp. Med. 210, 2515-2522 (2013).
- Whitlock, C. A., Tidmarsh, G. F., Muller-Sieburg, C., Weissman, I. L. Bone marrow stromal cell lines with lymphopoietic activity express high levels of a pre-B neoplasia-associated molecule. Cell. 48, 1009-1021 (1987).
- Onai, N., Obata-Onai, A., Tussiwand, R., Lanzavecchia, A., Manz, M. G. Activation of the Flt3 signal transduction cascade rescues and enhances type I interferon-producing and dendritic cell development. J. Exp. Med. 203, 227-238 (2006).
- Szilvassy, S. J., et al. Leukemia inhibitory factor upregulates cytokine expression by a murine stromal cell line enabling the maintenance of highly enriched competitive repopulating stem cells. Blood. 87, 4618-4628 (1996).
- Arcanjo, K., et al. Biochemical characterization of heparan sulfate derived from murine hemopoietic stromal cell lines: a bone marrow-derived cell line S17 and a fetal liver-derived cell line AFT024. J. Cell. Biochem. 87, 160-172 (2002).
- Onai, N., et al. A clonogenic progenitor with prominent plasmacytoid dendritic cell developmental potential. Immunity. 38, 943-957 (2013).
- Whitlock, C. A., Muller-Sieburg, C. E. Long-term B-lymphoid cultures from murine bone marrow establishment and cloning by using stromal cell line AC 6.21. Methods Mol. Biol. 75, 231-248 (1997).
- Chen, Y. -L., Chang, S., Chen, T. -T., Lee, C. -K. Efficient Generation of Plasmacytoid Dendritic Cell from Common Lymphoid Progenitors by Flt3 Ligand. PLoS ONE. 10, (8), e0135217 (2015).