Published online before print December 8, 2006
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,
* Division of Endocrinology, Metabolism and Gerontology, Department of Medicine and Graduate Program in Immunology, Stanford University School of Medicine, Stanford, California, USA;
Department of Pathology and Molecular Pathology Graduate Program, University of California at San Diego, La Jolla, California, USA; and
Deparment of Medicine, Harvard Massachusetts General Hospital, Boston, Massachusetts, USA
1 Correspondence: Division of Endocrinology, Metabolism and Gerontology, Department of Medicine and Graduate Program in Immunology, Stanford University School of Medicine, Stanford, CA 94305-5103, USA. E-mail: achawla{at}stanford.edu
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: cell line • alternative activation • classical activation
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
It is interesting that recent studies indicate that immune and innate activation of macrophages contributes to the pathogenesis of various chronic inflammatory diseases. For instance, macrophages are critical effector cells that promote tissue inflammation and damage in autoimmune (e.g., multiple sclerosis), inflammatory (e.g., psoriasis), and metabolic (e.g., atherosclerosis and type 2 diabetes mellitus) diseases [5 6 7 ]. Despite a greater appreciation for the pathogenic activities of these cells in chronic inflammatory states, molecular studies of macrophage activation programs have been hampered by the lack of good, tractable genetic systems. Currently, experimental macrophages are derived from two sources: tumor-derived cell lines (e.g., RAW 267.4, J774, and THP-1 cells) and primary cells {e.g., bone marrow-derived macrophages (BMDM) and elicited peritoneal macrophages [8 ]}. The former source is endowed with several propitious characteristics, including an unlimited replicative potential, genetic tractability, and population homogeneity. However, the use of macrophage-like cell lines is limited by their relative phenotypic immaturity and their inherent distance from normal biology as a result of their derivation from tumors. In contrast, BMDM and elicited-peritoneal macrophages are more reflective of quiescent and activated tissue macrophages, respectively. As such, however, the replicative potential, genetic tractability, and population homogeneity are all quite poor for these cells.
Embryonic stem (ES) cells, which can be differentiated into macrophages, provide a potential genetic system to study macrophage biology. However, conventional methods of generating ES cell-derived macrophages (ESDM) are tedious and give low yields, and the target population exhibits tremendous functional heterogeneity [9 10 11 ]. To develop a robust macrophage differentiation system, we investigated whether ES cell-derived myeloid progenitors could be isolated and expanded in vitro. Specifically, we asked whether oncoproteins, which arrest myeloid differentiation, could promote growth factor-dependent self-renewal of ES cell-derived myeloid progenitors. By testing various protocols for progenitor isolation, cytokine cocktails for progenitor renewal and differentiation, and different oncogenic proteins, we developed a simple, highly reproducible method for infinite expansion of ES-cell derived myeloid progenitors. In this report, we demonstrate that ectopic expression of estrogen-regulated HoxA9 protein (HoxA9-ER) conditionally immortalizes ES cell-derived myeloid progenitors, which are capable of differentiating synchronously into macrophages. It is important that immunologic and expression profiling of HoxA9-ER ESDM revealed a high degree of maturational and functional coherence. In contrast to other established culture systems, HoxA9-ER ESDM are cytokine-responsive and can express the macrophage programs of classical activation, alternative activation, and deactivation. Thus, these new methods provide a dramatic improvement in efficiency, purity, and homogeneity of cultured ESDM, making it an ideal system for high-throughput genetic analysis of macrophage functions.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, null ES cells were cultured on feeder cells as described [12
, 13
]. Embryoid bodies (EBs), generated as described previously [14
], were harvested on Days 6–8, trypsinized, and replated in IMDM, supplemented with murine stem cell factor (mSCF; 25 ng/ml) and mIL-3 (5 ng/ml). Two days later, nonadherent, hematopoietic precursors were collected, counted, and infected with ecotropic retroviruses encoding the conditional myeloid oncoprotein, HoxA9-ER [15
]. The immortalized progenitors, J1-HoxA9-ER or PPAR, null HoxA9-ER, were expanded and maintained in proliferation medium: IMDM supplemented with 10% FCS, SCF (25 ng/ml), mIL-3 (1.25 ng/ml), and ß-estradiol (1 µM). Macrophage differentiation was induced by withdrawing ß-estradiol and SCF and adding 20% L929 conditioned medium, a source of M-CSF. Media were subsequently changed every 2 days, and differentiated macrophages developed over 6–7 days. Genetic rescue experiments were carried out by infecting proliferating PPAR null HoxA9-ER cells with ecotropic retroviruses encoding full-length Flag-PPAR or empty vector and subjecting the selection with puromycin (4 µg/ml).
Gene expression studies
Total RNA was prepared using TRIzol reagent and used in Northern blot analysis as described previously [16
]. For quantitative PCR (QPCR), total RNA was treated with DNase (1 U/ml) and reverse-transcribed using a first-strand cDNA synthesis kit, and QPCR reactions were performed on an Opticon 2 DNA engine. Primers used in these studies are: CD68 (f: GCGGTGGAATACAATGTGTC, r: TGGAGCTCTCGAAGAGATGA), F4/80 (f: GCATCATGGCATACCTGTTC, r: AGTCTGGGAATGGGAGCTAA), M-CSF receptor (M-CSFR; f: ATCATGAGTCACCTGGGACA, r: CCTTCCTTCGGAGAAAGTTG), mannose receptor (MR; f: TGATTACGAGCAGTGGAAGC, r: GTTCACCGTAAGCCCAATTT), PU.1 (f: TCTGATGGAGAAGCTGATGG, r: TTTCTTGCTGCCTGTCTCC), and L32 (f: TGCTGATGTGCAACAAATCTT, r: GTTGGGATTGGTGACTCTGA). QPCR data were normalized to L32 levels using the 
C(t) method. Fold change was calculated relative to the expression levels of proliferating HoxA9 cells.
Protein analyses
Western blot analyses were performed using standard procedures. Cells were lysed in radioimmunoprecipitation assay buffer containing 1 mM Na2VO4, 1 mM NaF, 1 mM Na4P2O8, and 1x protease inhibitor cocktail. Total cell lysate (30 µg) was analyzed by immunoblotting for Flag-PPAR with anti-Flag primary antibody (1:1000, Sigma Chemical Co., St. Louis, MO, USA) and ß-actin (1:5000, Sigma Chemical Co.) and detected on an Odyssey infrared detection system with IRDye 800-conjugated secondary antibody (1:30,000, Rockland Immunochemicals, Gilbertsville, PA, USA). Arginase assays were preformed by well-established methodologies [16
, 17
]. For ELISAs, cells were plated in 48- or 96-well plates and stimulated as indicated for 6–24 h. TNF-
, IL-6, and IL-12 in culture supernatants were quantified per the manufacturer’s protocols (BD PharMingen, San Diego, CA, USA). All assays were done in triplicate, and results are indicative of at least three independent experiments. For flow cytometry, cells were washed with PBS, harvested, and resuspended in PBS containing 2% FCS and 0.2 mM EDTA. Samples were blocked with 200 µg/ml normal mIgG and stained on ice with fluro-conjugated antibodies against CD34, c-kit, stem cell antigen 1 (Sca-1), CD16/32, CD11b, F4/80, CD80, CD3, B220, Ter-119, GR-1, MR, and PD-L2 and analyzed on a Becton Dickinson LSR. For nitrite production, macrophages were stimulated with interferon-
(IFN-) (100 U/ml) and lipopolysaccaride (LPS) (5 ng/ml), and nitrite production was quantified by a Greiss reagent in culture supernatants.
Uptake studies
Macrophages were incubated with 1 mg/ml dextran-FITC (m.w. 70,000), 1 mg/ml Lucifer yellow, or 2 µm FITC-labeled latex beads in differentiation medium for 15 min at 37°C and 4°C. Cells were washed thrice with PBS, harvested, and analyzed immediately by FACS. Microbial uptake studies were performed with DH5 Escherichia coli, Salmonella typhimurium strain SLI344 (Stanley Falkow, Stanford University, CA, USA), or Candida albicans clinical isolate SC5314 [Alexander "Sandy" Johnson, University of California San Francisco (UCSF), USA]. Early log-phase cultures were incubated in Luria-Bertani broth (E. coli and S. typhimurium) or yeast-peptone-dextrose (C. albicans) for 15 min with 5 µM 5-chloromethylfluorescein diacetate (CMFDA; Molecular Probes, Eugene, OR, USA), washed thrice with PBS, and added directly to macrophage cultures at a microbe:macrophage ratio of 10:1 for 15 min at 37°C.
Allogeneic lymphocyte stimulation
HoxA9 macrophages were cultured in triplicate in a 96-well flat-bottom plate at the indicated numbers overnight in medium containing IFN- (10 U/ml). Wells were washed thrice with PBS, and 105 BALB/c allogeneic or SV129 syngeneic splenocytes in DMEM containing 10% FCS were added to each well. After 72 h, the cultures were pulsed with 0.5 µCi/well [3H]methyl-thymidine for 16 h, harvested onto glass fiber filters using the TomTec MachIII plate harvester, and counted in the Wallac 1450 Microbeta/TriLux detector. Proliferation is reported as 3H CPMwell – CPMbackground.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
 View larger version (39K): [in a new window] |
Figure 1. Generation of macrophage differentiating system by conditional immortalization of ES-derived myeloid progenitors. (a) Schematic representation for isolating and immortalizing ES-derived myeloid progenitors. (b–d) Differentiation of ES-derived myeloid progenitors yields mature macrophages, as assessed by morphology (b), progressive changes in gene expression (c), and expression of cell surface markers (d). For flow cytometry histograms, isotype-stained cells are represented by the thin, black line; proliferating cells by the blue line; and differentiated cells by the red line. LBD, Ligand-binding domain; E2, estradiol; Lin, lineage.
|
) but were negative for the hematopoietic stem cell marker Sca-1 and other lineage markers (Gr1, CD3, B220, and Ter-119; Fig. 1d
) [20
, 21
]. It is notable that maturation of J1-HoxA9-ER cells into macrophages was accompanied by down-regulation of progenitor markers (CD34: 7.4-fold; c-kit: 16-fold) and up-regulation of markers of mature macrophages (F4/80: 7.4-fold; CD11b: 28.4-fold; CD16/32: 3.5-fold; CD80: threefold), as shown in Figure 1d
. Although the cell surface phenotype of J1-HoxA9-ER progenitors is similar to the one ascribed to common myeloid progenitors, these immortalized progenitors only gave rise to macrophage-containing colonies in the in vitro colony-forming assays (data not shown), in agreement with the reported oncogenic activity of HoxA9 in acute myeloid leukemia and adult bone marrow progenitors [15
, 22
].
Intact IFN- signaling and classical activation in HoxA9-immortalized ESDM
Many extant macrophage-like cell lines lack the ability to activate appropriately when stimulated with IFN- [23
], and as such, their use in studies of macrophage biology is limited. To determine whether the J1-HoxA9-ER macrophages can enact the program of classical activation, these cells were stimulated with Th1 cytokine IFN- and assayed for expression of the proinflammatory phenotype. Immunoblots in Figure 2a
showed that treatment of cells with IFN- led to robust and rapid phosphorylation of STAT1 and strong up-regulation of key proinflammatory mRNAs, including TNF-, IL-6, and inducible NO synthase (iNOS; Fig. 2b
). Consistent with the known inflammatory priming effects of IFN-, J1-HoxA9-ER macrophages costimulated with IFN- and LPS had a marked increase in nitrite production (Fig. 2c)
, phlogistic cytokine response (Fig. 2d)
, and cell surface expression of costimulatory molecule PD-L1 (Fig. 2e)
[3
].
 View larger version (27K): [in a new window] |
Figure 2. J1-HoxA9-ER macrophages undergo classical activation in response to IFN-. (a) Stimulation with IFN- leads to robust phosphorylation of STAT1. Mature HoxA9 ESDM were stimulated for 30 min with IFN- (100 u/ml), and total cellular protein was probed for pY701 STAT1 (pSTAT1) and total STAT1. (b) IFN- induces inflammatory gene expression in HoxA9 ESDM. Cells were stimulated with IFN- (100 U/ml) for 24 h, and gene expression analyses were performed by Northern blotting. (c, d) Combination of IFN- (100 u/ml) and LPS (5 ng/ml) promotes classical activation of HoxA9 ESDM. Nitrite production in culture supernatants (c) or inflammatory cytokine secretion (d) was monitored by Griess reagent and ELISAs, respectively. (e) Induction of cell surface expression of PD-L1 by IFN- in HoxA9-immortalized ESDM. Isotype, Black; vehicle, blue; IFN-, red. Results are representative of at least three different independent experiments. *, P = 0.04; **, P < 0.005.
|
30-fold (Fig. 3c)
and dramatically enhanced cell surface expression of MR and dectin-1 by 2.2- and threefold, respectively (Fig. 3d)
. PD-L2, a high-stringency marker of macrophage alternative activation, is not expressed by any extant macrophage cell line in response to IL-4. It is remarkable that the Th2 cytokine IL-4 enhanced cell surface PD-L2 expression by threefold in J1-HoxA9-ER macrophages (Fig. 3d)
[24
]. Thus, unlike the existing macrophage cell lines, such as RAW264.7 or J774, the conditionally immortalized ES-derived macrophages can faithfully execute the cellular programs of classical and alternative activation.
 View larger version (31K): [in a new window] |
Figure 3. J1-HoxA9-ER macrophages undergo alternative activation in response to IL-4. (a) Treatment with IL-4 (10 ng/ml for 30 min) induces phosphorylation of STAT6 in HoxA9 ESDM. (b) IL-4 stimulates expression of mRNAs associated with the alternative activation phenotype. (c) HoxA9-immortalized ESDM induces arginase activity upon stimulation with IL-4 (10 ng/ml) for 24 h. (d) Cell surface expression of alternative activation markers. Flow cytometric analyses were performed on macrophages treated with vehicle or IL-4 (10 ng/ml) for 24 h. Isotype control, Black; vehicle, blue; IL-4, red. (e, f) IL-10 potently inhibits inflammatory activation of J1-HoxA9-ER macrophages. Cells were treated with LPS (5 ng/ml) in the presence or absence of IL-10 (10 ng/ml), and cytokine release was monitored by ELISAs. Results are representative of at least three different independent experiments. **, P < 0.005.
|
and IL-6 from J1-HoxA9-ER macrophages, indicating intact IL-10 signaling in these cells.
HoxA9-immortalized ESDM can perform essential macrophage functions
Resident and elicited macrophages continually sample their local environment. Large volumes of fluid phase can be ingested by pinocytosis, whereas endocytosis and phagocytosis mediate engulfment of larger particles, cellular debris, and invading microorganisms [29
]. To determine whether the J1-HoxA9-ER macrophages were capable of performing these essential macrophage functions, cells were incubated with labeled substrates at 4°C or 37°C, and the increase in fluorescence was monitored by flow cytometry. As shown in Figure 4a
, incubation of J1-HoxA9-ER macrophages at 37°C with Lucifer yellow or dextran-FITC (fluorescent molecules used to monitor pinocytosis and endocytosis, respectively) led to a strong increase in cellular fluorescence. Similarly, although J1-HoxA9-ER macrophages phagocytosed FITC-labeled latex beads rapidly, physiologically relevant phagocytosis was demonstrated by uptake of CMFDA-labeled C. albica, E. coli, and S. typhimurium (Fig. 4a
and data not shown).
 View larger version (23K): [in a new window] |
Figure 4. J1-HoxA9-ER macrophages are functionally patent for antigen uptake, identification, and presentation. (a) Cells were incubated with labeled endocytic, pinocytic, and phagocytic substrates for 15 min, and specific uptake (red) was followed by fluorescence gain over-reactions carried out at 4°C (blue). (b) TLR ligation induces TNF- secretion in HoxA9-immortalized ESDM. Cells were stimulated with tripalmitoyl-S-glycero-Cys-(Lys)4 (Pam3Cys; 100 ng/ml), polyinosinic:polycytidylic (pI:pC; 25 µg/ml), CpG DNA (1 µM), zymosan (1 µg/ml), or LPS (10 ng/ml) for 6 h, and TNF- secretion was monitored in supernatants by ELISA. (c) MLR between HoxA9-immortalized ESDM and allogenic or syngeneic splenocytes. Allogeneic (blue,  ) or syngeneic (green,  ) splenocytes (105) were added to HoxA9 macrophages activated overnight with IFN- (10 U/ml). Cellular proliferation was monitored 72 h later with [3H]methyl-thymidine.
|
and IL-6 (Fig. 4b
and data not shown).
Last, in addition to orchestrating the cytokine milieu of the inflammatory response, macrophages bridge the innate and adaptive immune systems by presenting antigen for adaptive activation. The ability of J1-HoxA9-ER macrophages to activate adaptive immune responses was assessed by monitoring alloreactivity in a MLR. Data presented in Figure 4c
show that IFN--activated J1-HoxA9-ER macrophages supported robust proliferation of allogeneic, but not syngeneic, lymphocytes (Fig. 4c)
. In aggregate, these results suggest that the HoxA9-immortalized ESDM are not only adept at performing the critical functions of innate immunity but can also activate the adaptive immune response efficiently.
Generation of genetically engineered macrophages from targeted ES cells
As homologous recombination can be used to introduce deletions or mutations in ES cells, our procedures for isolating and immortalizing myeloid progenitors from EBs can potentially be used to generate large quantities of genetically modified macrophages. To test this hypothesis, we took advantage of genetically targeted ES cells to generate a line of PPAR null ESDM. Macrophages differentiated from these lines were indistinguishable from their wild-type counterparts in surface marker expression and in transcriptional programs of differentiation (data not shown). These cells demonstrated a lack of sensitivity to the synthetic PPAR ligand GW501516, as assessed by induction of adipose differentiation-related protein (ADRP; Fig. 5b
), a well-characterized PPAR target gene [13
]. To demonstrate that functional transgenes can be expressed ectopically in PPAR null HoxA9-ER macrophages, epitope-tagged PPAR or empty vector was introduced retrovirally into proliferating PPAR null HoxA9-ER progenitors, and infected cells were subjected to selection by puromycin. Immunoblot analysis of selected cells with Flag antibody revealed robust expression of Flag-tagged PPAR in the PPAR, null HoxA9-ER macrophages. It is important that responsiveness to GW501516 was restored in PPAR null ESDM expressing Flag-PPAR, as evidenced by ligand inducibility of ADRP mRNA expression. Collectively, these studies indicate that immortalized ESDM can be generated from genetically modified ES cells and used for transgene expression, thus making this model system ideal for biological pathway discovery by genetic manipulation.
 View larger version (26K): [in a new window] |
Figure 5. Genetic tractability of HoxA9-immortalized ESDM. (a) Rescue of PPAR–/– ESDM by ectopic expression of the Flag-PPAR. Immunoblots showing expression of Flag-PPAR in PPAR, null ESDM. (b) Ectopic expression of PPAR confers ligand responsiveness to PPAR, null ESDM. Mature ESDM were stimulated with synthetic PPAR ligand GW501516 (100 nM) for 24 h, and ADRP mRNA expression was monitored by Northern blot analysis. WT, Wild-type.
|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Previous reports have described methodologies for isolating various definitive hematopoietic lineages, including macrophages, from spontaneously differentiating EBs [10 , 14 , 33 ]. However, lack of cellular homogeneity, poor yields, and the elaborate protocols for culturing EBs make these procedures unsuitable for generating large quantities of macrophages for biochemical and molecular studies. Thus, to fully exploit the genetic tractability of murine ES cells, methodologies are needed that facilitate the conditional expansion of myeloid precursor cells in vitro. As deregulated expression of the Hox family of transcription factors can lead to a block in hematopoiesis or expansion of hematopoietic stem cells [15 , 22 , 34 ], we investigated whether the conditional oncoprotein HoxA9-ER could arrest myeloid differentiation in the ES cell-derived hematopoietic precursors. As reported here, ectopic expression of HoxA9-ER leads to self-renewal of ES cell-derived myeloid precursors which, when stimulated with M-CSF, can terminally differentiate into macrophages. This new methodology for generating ESDM is superior to the conventional protocols in three different ways. First, rather than having to isolate myeloid progenitor cells repeatedly from EBs, these new methods allow for indefinite expansion of the myeloid progenitor cell population. In fact, the J1-HoxA9-ER-proliferating cells have been maintained stably in culture for over 3 years (>400 passages). Second, although the conventional protocols yield 10–20 million macrophages/week [14 ], we can routinely obtain 200–300 million macrophages/week, reflecting a 10- to 15-fold increase in the efficiency of the differentiation process. Third, our new methodology yields a homogenous population of cells that displays much better activation coherence, a point of great importance in dissecting the regulatory mechanisms controlling macrophage activation.
Unlike the tumor-derived or virally transformed macrophage cell lines [8 , 35 ], HoxA9-immortalized ES cell progenitors retain many of the cellular and functional characteristics of primary monocytes and macrophages. For instance, although IL-3 is required for proliferation of ES cell-derived myeloid progenitors, M-CSF is essential for terminal differentiation and survival of ESDM. Furthermore, analyses of gene expression by QPCR and flow cytometry revealed a high degree of phenotypic and maturational coherence between ESDM and BMDM (Fig. 1) . The ability of ESDM to faithfully recapitulate the classical, alternative, and deactivation phenotypes of BMDM further affirms the superiority of this macrophage differentiation system over other commonly used tumor-derived lines (Figs. 2 and 3) . For example, treatment of HoxA9-immortalized ESDM with IL-4 led to robust expression of the complete alternatively activated phenotype (Fig. 3) , an activation program that is only partially expressed by tumor-derived macrophage cell lines. Furthermore, microarray analyses of classical or alternatively activated ESDM demonstrated the canonical cytokine responses [16 ]. Taken together, these results suggest that our novel ES cell-derived macrophage differentiation system provides a means to generate large numbers of genetically defined macrophages, which retain the salient features of quiescent and activated BMDM.
Introduction of heterologous genes or small interfering (si)RNA-mediated knockdown of endogenous proteins is a powerful tool in understanding the function of a particular gene in macrophage biology. As HoxA9-ER progenitors have an infinite replicative capacity, infection of proliferating cells with transgene-encoding retroviruses can be performed easily (Fig. 5)
. Indeed, we recently reported gain- and loss-of-function studies with the transcriptional proliferator-activated receptor- coactivator (PGC)-1ß in J1-HoxA9-ER macrophages [16
]. It is notable that transgenic expression of PGC-1ß in mice, via the CD68 promoter, yielded results that were similar to those obtained with J1-HoxA9-ER macrophages, thus validating the use of our novel macrophage differentiation system for rapid analysis of gene function in macrophage biology.
Finally, testing macrophage-specific promoters and transgenic constructs is an important area where this novel macrophage differentiation system can be exploited. The standard methods used to generate transgenic mice, including pronuclear injection, founder screening, and transgene expression analysis, are costly and time-consuming [12 ]. As transgenic constructs can be introduced easily into ES cells, the methods outlined in this manuscript provide an excellent means of identifying genetically engineered ES cells, which express the desired transgene in macrophages. By initially screening for transgene expression in HoxA9-immortalized ESDM, one can be assured that ES cells injected into blastocysts will give rise to chimeric mice fully capable of transgene expression. This adaptation of the transgenic technology not only provides a reliable means of generating transgenic mice but also a genetically identical macrophage differentiation system that can be manipulated further by siRNA or overexpression to dissect regulatory cascades governing macrophage activation.
Received September 28, 2006; revised November 9, 2006; accepted November 14, 2006.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
is a very low-density lipoprotein sensor in macrophages Proc. Natl. Acad. Sci. USA 100,1268-1273 dependent and independent effects on macrophage-gene expression in lipid metabolism and inflammation Nat. Med. 7,48-52[CrossRef][Medline] in macrophage differentiation and cholesterol uptake Nat. Med. 7,41-47[CrossRef][Medline]This article has been cited by other articles:
 |
|
 |
|
  S. Senju, M. Haruta, Y. Matsunaga, S. Fukushima, T. Ikeda, K. Takahashi, K. Okita, S. Yamanaka, and Y. Nishimura Characterization of Dendritic Cells and Macrophages Generated by Directed Differentiation from Mouse Induced Pluripotent Stem Cells Stem Cells, May 1, 2009; 27(5): 1021 - 1031. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
