DEVELOPMENTAL DYNAMICS 216:450–458 (1999) Derivation of Melanocytes From Embryonic Stem Cells in Culture TOSHIYUKI YAMANE,1 SHIN-ICHI HAYASHI,1 MASAKO MIZOGUCHI,2 HIDETOSHI YAMAZAKI,1 AND TAKAHIRO KUNISADA1* 1Department of Immunology, School of Life Science, Faculty of Medicine, Tottori University, Yonago, Japan 2Department of Dermatology, St. Marianna University School of Medicine, Kawasaki, Japan ABSTRACT We report that embryonic stem (ES) cells were efficiently induced to differentiate to melanocytes in vitro. When undifferentiated ES cells were cocultured with a bone marrow– derived stromal cell line, a very small but significant number of melanocytes were reproducibly generated. This process was greatly enhanced by addition of dexamethasone to the culture and strictly depended on steel factor, the ligand for the c-Kit receptor tyrosine kinase. Expression of c-Kit on the precursor cells was confirmed by using SCL/tal-1-/- ES cells, which are defective for producing hematopoietic cells, which were thus ruled out as possible sources of nonmelanogenic c-Kit-expressing cells. The morphology, reactivity to growth factors, and expression of melanogenic markers of the cells generated all indicated unequivocally that these cells were melanocytes. This culture system may provide a potent tool for the study of development and function of melanocytes. Dev Dyn 1999;216:450–458. r 1999 Wiley-Liss, Inc. Key words: melanocyte; ES cell; stromal cell; dexamethasone; steel factor; c-Kit; endothelin 3; mouse INTRODUCTION Mouse embryonic stem (ES) cells derived from the inner cell mass of 3.5-day postcoitum (dpc) blastocysts are totipotent cells (Evans and Kaufman, 1981; Martin, 1981) closely resembling the primitive ectoderm of the very early postimplantation embryo. As in vitro models for cell differentiation, ES cells can generate various cell lineages, such as parietal and visceral endoderm, cardiac muscle (Doetschman et al., 1985), vascular endothelial cells (Vittet et al., 1996), hematopoietic cells (Doetschman et al., 1985), and neuronal cells (Wobus et al., 1988). Although ES cells can thus be induced to differentiate into cells of all three germ layers, until now there has been no solid evidence that ES cells could also be induced to generate neural crest lineage in vitro. During the neurulation stage, neural crest cells migrating away from the closing neural tube develop into a variety of cell lineages that include bone and cartilage cells of the head, neurons and glia of the sensory and autonomic nervous systems, and all melar 1999 WILEY-LISS, INC. nocytes of the skin, inner ear, and choroid (Le Douarin, 1982; Bronner-Fraser, 1995). Among the various types of neural crest–derived cell lineages, we first chose melanocytes for induction attempts from ES cells, because melanocytes are derived mainly from the neural crest. Migration, survival, proliferation, and differentiation of the neural crest– derived melanocytes are dependent on steel factor (SLF), the ligand for the c-Kit receptor tyrosine kinase expressed on the surface of melanocytes and their precursors (Kunisada et al., 1998b). On the other hand, pigmented epithelial cells of the retina and iris, which are sometimes called melanocytes, are formed from the neuroepithelium of the optic vesicle. These cells are histologically distinguishable from those of neural crest origin, because they form a simple epithelial layer of hexagonal cells without dendritic processes (Silvers, 1979). Dopaminergic neurons of the locus ceruleus and substantia nigra are also known to contain characteristic melanin granules, called neuromelanin (Breathnach, 1988). Importantly, these pigment cells of nonneural crest origin are not affected by null mutations of the Slf or c-Kit genes (Motro et al., 1996; Kunisada, unpublished observations) and thus they can be discriminated from melanocytes of neural crest origin not only by their morphology but also by their SLF-independent growth. To induce melanocytes from ES cells, we took advantage of the ST2 stromal cell line, which was first developed for the establishment of long-term bone marrow cultures (Ogawa et al., 1989). Also, co-culture of ES cells with ST2 cells generated hematopoietic cell lineages in response to appropriate growth factors or cytokines (Yamane et al., 1997); therefore, ST2 monolayers are also expected to provide the environment necessary for melanocyte differentiation. In this culture system, undifferentiated ES cells are inoculated directly onto ST2 monolayers, without forming cell aggregates or embryoid bodies, and thus the whole process of cell development can easily be monitored. Here, we report efficient and reproducible induction of the differentiation of ES cells into melanocytes. The melanocytes *Correspondence to: Takahiro Kunisada, Department of Immunology, School of Life Science, Faculty of Medicine, Tottori University, 86 Nishi-machi, Yonago 683-8503, Japan. E-mail: tkunisad@grape. med.tottori-u.ac.jp Received 6 July 1999; Accepted 20 September 1999 INDUCTION OF MELANOCYTES FROM ES CELLS 451 Fig. 1. Genotype of tyrosinase of ST2 cell. A: Restriction maps of mouse tyrosinase coding region subjected to polymerase chain reaction (PCR) amplification with primers described in experimental procedures. DdeI restriction enzyme site is indicated by D. B: DNAs prepared from ST2 cells or mouse strains indicated on each lane were amplified by PCR and then cut by DdeI restriction enzyme for restriction fragment length polymorphism analysis. produced express markers of the melanocyte lineage, including c-Kit, and require SLF, suggesting that these cells correspond to skin melanocytes originated from neural crest cells. Numbers indicated in A and B are the length of the DNA fragments (bp). RESULTS Generation of Melanocytes From Undifferentiated ES Cells in Culture Fig. 2. Effects of various factors on melanocyte differentiation from embryonic stem (ES) cells. Two thousand D3 ES cells were inoculated onto ST2 monolayers and incubated in the presence of the indicated factors. On day 21 of culture, cells were trypsinized, and the number of melanocytes was counted. The number represented by 100% is 29,700, the number of melanocytes generated in the presence of Dex only (bar 9). Each bar represents the average value of triplicate cultures in a representative experiment, and vertical bars indicate standard deviations. Differentiation of mature melanocytes from their precursors in vitro requires supporting feeder cells, such as keratinocytes (Hirobe, 1994). Because monolayers of stromal cells are used for the induction of hematopoietic cells from ES cells, we first tested whether the ST2 stromal cell line is capable of supporting melanocyte development. To confirm the genotype of tyrosinase of ST2 cell line, a part of tyrosinase coding region was amplified by polymerase chain reaction (PCR) and tested for its restriction fragment length polymorphism (RFLP), according to Shibahara et al. (Fig. 1A, based on the result of Shibahara et al., 1990). As shown in Fig. 1B, the amplified fragment from ST2 had two DdeI cutting sites, and this RFLP pattern was identical with that of BALB/c albino mouse. C57BL/6 mouse was confirmed to have the wild-type tyrosinase gene containing a single DdeI site. Thus, any pigmented cells observed in the culture are derived from ES cells. We initially inoculated 2,000 undifferentiated D3 ES cells onto ST2 monolayers, and then maintained the cultures for 21 days in medium containing 10% calf serum. Under these conditions, small numbers of melanocytes were reproducibly generated (110 cells per well in four independent experiments performed in duplicate), therefore, we performed further improvement of the culture system. The effects of basic fibroblast growth factor (bFGF), cholera toxin (CT), and 12-O-tetradecanoyl phorbol acetate (TPA), agents known to promote generation of melanocytes from migrated neural crest cells (Ito et al., 1993) and of dissociated epidermal skin tissue (Hirobe, 1992; Sviderskaya et al., 1995), were tested. TPA caused moderate enhancement (6,500 melanocytes per well in duplicate cultures) when added alone, but no other factor or combination of factors enhanced the generation of melanocytes (bars 1–8 of Fig. 2). Dexamethasone 452 YAMANE ET AL. (Dex) very strongly promoted the generation of melanocytes (23,000 cells per well in 3 independent sets of duplicate cultures). The addition of bFGF, CT, or TPA slightly increased the number of Dex-induced melanocytes (bars 9–16 of Fig. 2). When added alone, 1a, 25-dihydroxyvitamin D3 (1a, 25-(OH)2D3), which is known to promote melanin synthesis in mouse melanoma cells (Hosoi et al., 1985), was 40 times less potent than Dex. Treatment with bFGF, CT, or TPA in addition to 1a, 25-(OH)2D3 caused significant further enhancement of generation of melanocytes, but only to a level less than one-tenth of that induced by Dex alone (bars 17–20 of Fig. 2). In the presence of 1a, 25-(OH)2D3, TPA was inhibitory when combined with CT, or CT and bFGF (bars 21–24 of Fig. 2). Dex plus 1a, 25-(OH)2D3 was slightly more effective than Dex alone, but the further addition of the other three factors singly or in any combination had very little effect (bars 25–32). The degree of enhancement of induction of melanocytes by Dex was the same at concentrations of 1028 to 1025 mol/L Dex in the presence of 10% calf serum, 20 pmol/L bFGF, and 10 pmol/L CT (Fig. 3A). We next tested the period during which Dex facilitates differentiation of melanocytes. As shown in Fig. 3B, the addition of Dex from day 0 to day 6 slightly enhanced the number of melanocytes, indicating that Dex was effective in the early phase of melanocyte development from ES cells. However, the addition of Dex from day 12 until day 21 exerted the same magnitude of effect as when Dex was added throughout the culture period, from day 0 until day 21. Thus, although Dex may affect the entire developmental process of melanocytes from ES cells, the number of melanocytes that appeared was mostly affected by Dex at a later stage, apparently from day 12 or later. In the medium containing calf serum and Dex, a small number of what appeared to be black, dendritic melanocytes first appeared at around day 14 of culture, and the number increased thereafter until day 21. In the presence of bFGF and/or CT, mature melanocytes were observed 1 or 2 days earlier in the culture. Based on these results, we set the standard conditions for the generation of melanocytes from D3 ES cells as: 10% calf serum, 1027 mol/L Dex, 20 pmol/L bFGF, and 10 pmol/L CT. The morphology of ES cell–derived melanocytes is shown in Fig. 4. On day 6, single ES cells had multiplied and formed colonies (Fig. 4A and B); then, on days 12 and 13, dendritic or bipolar pigmented cells that looked like mature melanocytes had migrated out of the colonies (Fig. 4C–F, and K). On day 21, most of the culture dish was covered with these mature melanocytes (Fig. 4G–J, and L). In this later stage of the culture, dendritic cells observed were fully pigmented. Electron microscopic analysis revealed typical melanocytes with long dendrites and numerous stage IV mature melanosomes (Fig. 4M and N). Some melanocytes were actively producing melanosomes, because they contained stage II and III melanosomes in their cytoplasm (m2 and m3 in Fig. 4N). Melanosome com- Fig. 3. The effects of Dex dose and period of Dex treatment in embryonic stem cell cultures on melanocyte differentiation. Cultures were prepared and maintained in the standard conditions except that various doses of Dex (A) were added for 21 days. Similarly, cultures were supplemented by 1027 mol/L Dex for the indicated periods (B). plexes made by cells phagocytosing their own dendrites were occasionally found in the cytoplasm of melanocytes. These ultrastructures confirmed that fully developed active melanocytes were generated from ES cells in this culture system. Effect of SLF and Endothelin 3 (ET3) on Induction of Melanocytes From ES Cells The factors thus far genetically proven to be indispensable for melanocyte development are SLF (Kunisada et al., 1998a; Kunisada et al., 1998b and references therein) and ET3 (Baynash et al., 1994). We therefore tested the INDUCTION OF MELANOCYTES FROM ES CELLS Fig. 4. Morphology of the cells differentiated from embryonic stem (ES) cells. D3 ES cells were cultured under the standard conditions (with Dex, basic fibroblast growth factor and cholera toxin). A, B: Photomicrographs from day 6 of the culture. Cell colonies, relatively small (A) and large (B), were observed. (C–F) On day 12, pigmented melanocytes were not yet detected in most of the colonies. G–J: On day 21, many melanocytes were observed inside and outside of the colonies. K: On day 13, pigmented melanocytes, indicated by arrows, started to appear at the edge of each colony. L: On day 21, melanocytes were highly pigmented 453 and dendritic in shape. M and N: To further define the structure of these melanocytes, cells from day 21 of culture were examined by electron microscopy, as previously described (Okura et al., 1995). D, F, H, and J: Bright-field images matching the phase-contrast images in (C), (E), (G), and (I), respectively. Scale bars: A and B 5 400 µm; C–J 5 200 µm; K and L 5 100 µm; M 5 5 µm; N 5 1 µm. Abbreviations in (M) and (N): d, dendrite; m2, m3, and m4: stage II, stage III, and stage IV melanosomes, respectively. Asterisk indicates a cell that may be a melanoblast without melanosomes. 454 YAMANE ET AL. Expression of c-Kit in ES Cell Culture Fig. 5. Effects of steel factor (SLF) and ET3 on generation of melanocytes from D3 ES cells. SLF and/or ET3 were added to culture medium containing Dex or Dex, basic fibroblast growth factor (bFGF), and cholera toxin. In the experiment represented by the bottommost bar, 30 µg/ml of ACK 2 monoclonal antibody (Nishikawa et al., 1991) was added throughout the culture period. Each bar represents the average value of triplicate cultures in a representative experiment, and vertical lines indicate standard deviations. effects of these factors in this culture system. Addition of SLF had no significant effect on the induction of melanocytes from ES cells cultured in medium containing Dex (bars 1 and 2 of Fig. 5). In the presence of ET3, five times more melanocytes were generated (bars 1 and 3 of Fig. 5) and this enhancement did not change by the further addition of SLF (bar 4 of Fig. 5). This ET3-induced enhancement was also observed in the standard culture containing Dex, bFGF, and CT (bars 5 and 7 of Fig. 5). Again, SLF did not induced any enhancement of melanocyte induction (bars 6 and 8 of Fig. 5). ET3-mediated promotion of proliferation and differentiation of melanocytes from embryonic neural crest has been reported (Reid et al., 1996; Lahav et al., 1996; Opdecamp et al., 1997). On the other hand, addition of SLF did not have any effect on melanocyte development. This is not surprising because ST2 cells are known to produce SLF (Ogawa et al., 1991), and differentiated heterogeneous cell populations other than melanocytes may also produce SLF. SLF-dependent melanocyte development of the present culture system was shown by the fact that the addition of ACK2 antibody, which prevents binding of SLF to its receptor, c-Kit (Nishikawa et al., 1991), abolished melanocytes from ES cell culture (bar 9 of Fig. 5). Elimination of melanocyte precursors by ACK2 treatment was observed in cultures of isolated neural tubes (Reid et al., 1995) or in developing embryos (Yoshida et al., 1993). The effect of ACK2 is specific to melanocyte lineage, because osteoclasts appeared normally in the presence of ACK2 (Yamane et al., 1997). The fact that induction of melanocytes from ES cells is SLF-dependent leads us to expect that precursors expressing the SLF receptor, c-Kit, might transiently appear in the culture. Because of the totipotency of ES cells, the present culture system generates multiple cell lineages in addition to melanocytes. In fact, SLFdependent, c-Kit1 precursors of hematopoietic cells are generated in the ES cultures on ST2 monolayers (Yamane, unpublished observation). To prevent the emergence of c-Kit1 hematopoietic cells in the present ES cell cultures, we took advantage of SCL/tal-1-/- ES cells, from which no hematopoietic cells are induced in vivo (Porcher et al., 1996) or in vitro (Porcher et al., 1996). SCL/tal-1-/- ES cells generated mature melanocytes morphologically indistinguishable (Fig. 6A) from those generated from normal ES cells (Fig. 6B). Thus, c-Kit1 cells detected on day 9 of the culture are highly likely to be melanocyte precursors or melanoblasts (Fig. 6C, D). The appearance of these c-Kit1 cells was prevented in the SCL/tal-1-/- ES cell culture by ACK2 treatment (Fig. 6E, F). Expression of Melanocyte Lineage Markers To follow the expression of some critical melanogenic markers, we used reverse transcriptase polymerase chain reaction (RT-PCR) to analyze expression of Trp-1 (Jimenez et al., 1991), Trp-2 (Tsukamoto et al., 1992), and tyrosinase (Jimenez et al., 1989), at various times in the cultures. Trp-2 and tyrosinase mRNAs were first detected on day 6 of culture, and the less-abundant Trp-1 mRNA was also detected on day 6. The expression of these markers reached a maximum on day 12 (Fig. 7A). Expression of MITF, a basic-helix-loop-helixzipper transcription factor whose mutations are associated with loss of neural crest–derived melanocytes, was monitored using anti-MITF antibody (Opdecamp et al., 1997). Cells stained by the anti-MITF antibody were detected on day 6 of the culture (Fig. 7B, C). Thus, melanogenic markers need at least 6 days to become detectable in this culture system. Functional Analysis of Melanocyte Precursors Generated From ES Cells Expression of c-Kit protein in vitro was detected as early as 2 days after explanation of the 9 dpc neural tube (Reid et al., 1995), corresponding to 11 dpc in vivo, when C-kit mRNA expression was found in the embryo along melanocyte migration pathways (Manova and Bachvarova, 1991; Duttlinger et al., 1993). SLF dependence of neural crest cells in vitro begins 2 days after explanation and lasts about 4 days (Morrison-Graham and Weston, 1993). Blocking of any endogenous SLF activity by ACK2 treatment of neural tube cultures resulted in the loss of c-Kit-expressing melanocyte precursors, indicating that these precursor cells are SLF dependent (Reid et al., 1995). Generation of melanocytes from ES cells is also SLF dependent, as shown INDUCTION OF MELANOCYTES FROM ES CELLS in the previous sections. We tried to determine the timing of the differentiation of ES cells into SLFdependent precursor cells. For this purpose, ACK2 was added from the initiation of the ES cultures for various periods, as indicated in Fig. 8. For the first 3–6 days of culture, ACK2 showed no significant effect on the induction of melanocytes. When ACK2 was added for the first 9 days of the culture, the number of melanocytes on day 21 was decreased to 32% of control, and when ACK2 was added for 12 days or longer, virtually no melanocytes were generated. These results indicate that SLF-dependent melanocyte precursors did not emerge until around day 6, then emerged within the next 3 days, and increased their number to the maximum level by day 12. DISCUSSION In this study, we have recapitulated the whole developmental process of melanocyte cell lineage, starting from a totipotent ES cell line. Individual ES cells first proliferate in a tightly associated form, as shown in Fig. 4A and B, and then melanocytes emerge inside (Fig. 4I and J) or outside (Fig. 4G and H) of these colony-like structures. Apparently, cells other than melanocytes, such as endothelial cells, macrophages, and neuronlike cells (data not shown) exist in these colony-like structures, suggesting that interaction between cells derived from ES cells is important. In the present culture system, the most critical requirement for the induction of melanocytes was Dex, which is not usually used for the culture of melanocytes. Instead, other reagents known to be important for the induction of melanocytes from the neural tube, or epidermal skin, such as bFGF, CT, and TPA, were not as effective as Dex. Although the effect of Dex is prominent during the later stages of the ES cell culture, a small but significant effect was observed even when Dex was added only during the first 6 days of the culture. The effect of Dex on melanocyte development needs to be further investigated. Because melanocytes are mainly derived from neural crest cells, we assume that induction of the neural crest might have occurred in the culture. The possibilities that pigmented cells induced in this culture are derived from epithelial cells or neurons of nonneural crest origin appear not to be the case for two reasons. First, the pigmented cells induced are highly dendritic in shape and very different from the cells of pigment epithelium or neurons, as shown in Fig. 4L and M. Second, pigmented cells induced in the present system were dependent on c-Kit signaling, as is also true of the cells of neural crest origin. On the other hand, the pigment epithelium of retina and iris are totally independent of c-Kit signaling (Nishikawa et al., 1991). No developmental defects are reported in neuromelanincontaining neurons of the locus ceruleus and substantia nigra in c-kit-deficient W mutant mice (Motro et al., 1996). However, although very unlikely, it is also possible that melanocytes in the present ES cell culture are 455 directly derived from ES cells without any relevance to the normal differentiation pathway of melanocytes from trunk neural crest. In cultures started from 9.5 dpc neural tubes, early melanocyte lineage markers, such as MITF and TRP-2, appear as early as day 1 of culture and, in vivo, MITF-positive cells are observed at around 10.5–11 dpc (Opdecamp et al., 1997). In our culture system, melanocyte lineage markers first appeared as early as day 6 of the culture (Fig. 7). We think that this difference between the lag times before the onset of expression of melanocyte lineage markers in ES cell cultures and neural crest cell cultures is likely because ES cells closely resemble the inner cell mass of 3.5 dpc, and need the 6 days in culture to differentiate into the melanocyte lineage. In accordance with expression of melanocyte-specific markers on day 6 of the culture, the generation of SLF-dependent precursors was detected during days 6–9 of the culture (Fig. 8). Regardless of the developmental process by which they were derived, the melanocytes generated were strictly dependent on SLF, as are those developed in vivo (Nishikawa et al., 1991) or maintained in vitro (Murphy et al., 1992). Because c-Kit1 cells were generated from SCL/tal-1-/- ES cells, which are defective for hematopoiesis, precursors of melanocytes generated in the present ES culture are most likely to express c-Kit, as expected from the fact that induction of melanocytes in the culture is SLF dependent. ET3 is also indispensable for the proper differentiation of melanocytes, and the addition of ET3 to the cultures increased the number of melanocytes. These results show that the melanocytes generated are normal, as judged by their growth factor dependency. In this report, we have shown the induction of melanocyte lineage cells from ES cells. By introducing or removing genes necessary for the development and function of melanocyte lineage, the present ES cell cultures may contribute to elucidate more precise mechanisms for melanocyte differentiation. Furthermore, in combination with strategic insertional mutagenesis study of ES cells, previously unknown genes could possibly identified. EXPERIMENTAL PROCEDURES Genotype of Tyrosinase of ST2 Cell Line Genomic DNAs purified from ST2 cells, kidney of C57BL/6 and BALB/c mice were amplified by PCR. Primers used were 58CTTCAAAGGGGTGGATGACC38 and 58GAACTTATTCTTTTCGGAGACACTC38. PCR cycle was 94°C for 5 min, then 36 cycles of 94°C for 45 sec, 58°C for 1 min, and 72°C for 1 min. The amplified fragments were cut by DdeI restriction enzyme and separated by 3% of Agarose X (Nippon Gene, Japan) gel. ES Cell Culture D3 ES cells, SCL/tal-1-/- ES cells (established from the J1 ES cell line) and J1 ES cells were maintained on mouse embryonic fibroblasts in the presence of leukemia inhibitory factor as previously described (Yamane 456 YAMANE ET AL. Fig. 7. Expression of melanocyte lineage markers. A: Embryonic stem (ES) cell culture was carried out under the standard conditions, and on the days indicated, cells were harvested for the purification of total RNA, and then reverse transcriptase polymerase chain reaction was performed to detect expression of melanogenic marker genes. The lane labeled ‘‘ES’’ shows RNA that was prepared from undifferentiated ES cells (HPRT, hypoxanthine phosphoribosyl transferase). B and C: Immunocytochemical detection of MITF was carried out by the method described previously (Kunisada et al., 1996). The antisera were used at 150-fold dilutions. On day 6, MITF became detectable in the nucleus (B). The phase-contrast image shows that these MITF1 cells were growing tightly attached to ST2 cells (C). Scale bar: B, C 5 25 µm. Fig. 6. Melanocyte differentiation from SCL/tal-1-/- embryonic stem (ES) cells. SCL/tal-1-/- ES cells (A) and J1 ES cells, from which SCL/tal-1-/cells were derived (B), were cultured on ST2 cells for 21 days under the standard conditions. To identify c-Kit1 precursor cells, cultures of SCL/tal1-/- ES cells were fixed on day 9 and stained with anti-c-Kit monoclonal antibody ACK4 (C, bright-field image; D, phase-contrast image). The appearance of these c-Kit1 cells was prevented by the addition of ACK2-blocking antibody during days 0–9 of culture (E, bright-field image; F, phase-contrast image). Note that ACK4 and ACK2 have different antigen recognition sites and do not interfere with the binding to each other’s determinants. Scale bars: A, B 5 25 µm; C, F 5 50 µm. et al., 1997). Two thousand undifferentiated ES cells were harvested by trypsinization and inoculated into six-well plates previously seeded with ST2 cells, and differentiated in a-minimum essential medium (MEM) supplemented with 10% calf serum (Hyclone) in 5% CO2 at 37°C. For the experiments shown in Fig. 2, the following factors: 1027 mol/L Dex (Sigma), 1028 mol/L 1a, 25-(OH)2D3 (Biomol Research Laboratories), 20 pmol/L bFGF, 50 nmol/L TPA (Sigma), or 10 pmol/L CT (Sigma), were added in some experiments as indicated. The medium was changed twice a week. For the experiments shown in Fig. 5, 100 ng/ml of recombinant mouse SLF (Yasunaga et al., 1995), 100 ng/ml of ET3 (Peptide Institute Inc.), and 30 µg/ml of ACK2 monoclonal antibody (Nishikawa et al., 1991) were added as indicated. Immunohistochemistry Cells grown on ST2 monolayers in six-well plates were fixed with 4% paraformaldehyde, washed twice with phosphate-buffered saline (PBS) and incubated with 0.3% H2O2 in methanol to inactivate endogenous peroxidase, and then incubated with blocking medium (10% normal goat serum in PBS). Next, 0.8 ml of INDUCTION OF MELANOCYTES FROM ES CELLS 457 PCR conditions for tyrosinase were: 94°C, 5 min; 58°C, 2 min; 72°C, 3 min for the primary cycle, followed by 36 cycles of 94°C, 0.75 min; 58°C, 2 min; 72°C, 1.5 min; and then extension at 72°C for 6 min. For other primers, a 94°C, 3-min denaturation step was followed by 36 cycles of 94°C, 0.75 min; 58°C, 1 min; 72°C, 1.5 min; and then extension at 72°C for 3 min. ACKNOWLEDGMENTS Fig. 8. Emergence of steel factor–dependent melanocyte precursors in the embryonic stem (ES) cell culture. ACK2 blocking antibody (10 µg/ml) was added for the indicated periods in the ES cultures prepared in six-well plates under the standard condition. The numbers of melanocytes were counted on day 21. We thank T. Baba for technical assistance for electron microscopy; S.H. Orkin for SCL/tal-1-/- ES cells; T. Nakano for providing D3 ES cells; E. Nakajima for reading the manuscript; and T. Shinohara for secretarial assistance. This work was supported by grants from the Special Coordination Funds for Promoting Science and Technology from the Science and Technology Agency of Japan, the Ministry of Education, Science, Sports and Culture of Japan, Cellular Technology Institute, Otsuka Pharmaceutical Co. Ltd, and the Uehara Memorial Foundation. T.Y. is a JSPS Research Fellow. REFERENCES primary antibody solution diluted with blocking medium (1:400 for rabbit anti-mouse TRP-2; 1: 150 for rabbit anti-mouse MITF; 10 µg/ml for rat anti-mouse c-Kit monoclonal antibody, ACK4, Ogawa et al., 1991) was added and incubated at room temperature for 12 hr. Cells were washed in PBS and then incubated for 30 min in 0.8 ml of second antibody solution (1 µg/ml of horseradish peroxidase [HRP]-conjugated goat antirabbit IgG or HRP-conjugated goat anti-rat IgG, both were purchased from Vector). Cells were again washed in PBS, and 0.8 ml of diaminobenzidine (DAB) solution was added to develop the peroxidase reaction. After 10 min, cells were washed several times with PBS to stop the reaction. Anti-TRP2 antibody and anti-MITF antibody were gifts from Drs. V. Hearing (Tsukamoto et al., 1992) and H. Arnheiter (Opdecamp et al., 1997), respectively. RT-PCR Analysis Total RNA was purified from cultures at various times using Isogen (Nippon Gene). After treatment with DNase I (Pharmacia), first-strand cDNA synthesis was carried out using Super Script reverse transcriptase (Gibco-BRL) primed with random hexamer in a 20-µl reaction mixture containing 1 µg of total RNA. One and one-half microliter of the first-strand cDNA mixture was used for PCR with Taq polymerase (Toyobo) performed in a 25-µl volume. Ten microliter of each PCR product was electrophoresed on an agarose gel and stained with ethidium bromide. Primers used for PCR were as follows: TRP-1: 58GCCCCAACTCTGTCTTTTCTCAAT38/58GATCGGCGTTATACCTCCTTAGC38; TRP-2: 58GGACCGGCCCCGACTGTAATC38/58GTAGGGCAACGCAAAGGACTCAT38; tyrosinase: 58GGGCCCAAATTGTACAGAGAAGC38/58CTGCCAGGAGGAGAAGAAGAAGGATG38; hypoxanthine phosphoribosyl transferase (HPRT): 58GTAATGATCAGTCAACGGGGGC Baynash AG, Hosoda K, Giaid A, Richardson JA, Emoto N, Hammer RE, Yanagisawa M. 1994. Interaction of endothelin-3 with endothelin-B receptor is essential for development of epidermal melanocytes and enteric neurons. Cell 79:1277–1285. Breathnach AS. 1988. Extra-cutaneous melanin. Pigment Cell Res 1:234–237. Bronner-Fraser M. 1995. Origins and developmental potential of the neural crest. Exp Cell Res 218:405–417. Doetschman TC, Eistetter H, Katz M, Schmidt W, Kemler R. 1985. The in vitro development of the blastocyst-derived embryonic stem cell lines: formation of visceral yolk sac, blood islands and myocardium. J Embryol Exp Morphol 87:27–45. Duttlinger R, Manova K, Chu TY, Gyssler C, Zelenetz AD, Bachvarova RF, Besmer P. 1993. W-sash affects positive and negative elements controlling c-kit expression: ectopic c-kit expression at sites of kit-ligand expression affects melanogenesis. Development 118:705– 717. Evans MJ, Kaufman MH. 1981. Establishment in culture of pluripotent cells from mouse embryos. Nature 292:154–156. Hirobe T. 1992. Basic fibroblast growth factor stimulates the sustained proliferation of mouse epidermal melanoblasts in a serum-free medium in the presence of dibutyryl cyclic AMP and keratinocytes. Development 114:435–445. Hirobe T. 1994. Keratinocytes are involved in regulating the developmental changes in the proliferative activity of mouse epidermal melanoblasts in serum-free culture. Dev Biol 161:59–69. Hosoi J, Abe E, Suda T, Kuroki T. 1985. Regulation of melanin synthesis of B16 mouse melanoma cells by 1 a, 25-dihydroxyvitamin D3 and retinoic acid. Cancer Res 45:1474–1478. Ito K, Morita T, Sieber-Blum M. 1993. In vitro clonal analysis of mouse neural crest development. Dev Biol 157:517–525. Jimenez M, Maloy WL, Hearing VJ. 1989. Specific identification of an authentic clone for mammalian tyrosinase. J Biol Chem 264:3397– 3403. Jimenez M, Tsukamoto K, Hearing VJ. 1991. Tyrosinase from two different loci are expressed by normal and by transformed melanocytes. J Biol Chem 266:1147–1156. Kunisada T, Yoshida H, Ogawa M, Shultz LD, Nishikawa S-I. 1996. Characterization and isolation of melanocyte progenitors from mouse embryos. Dev Growth Differ 38:87–97. Kunisada T, Lu SZ, Yoshida H, Nishikawa S, Nishikawa SI, Mizoguchi M, Hayashi SI, Tyrrell L, Williams DA, Longley BJ. 1998a. Murine cutaneous mastcytosis and epidermal melanocytosis induced by keratinocyte expression of transgenic stem cell factor. J Exp Med 187:1565–1573. 458 YAMANE ET AL. Kunisada T, Yoshida H, Yamazaki H, Miyamoto A, Hemmi H, Nishimura E, Shultz LD, Nishikawa SI, Hayashi SI. 1998b. Transgene expression of steel factor in the basal layer of epidermis promotes survival, proliferation, differentiation and migration of melanocyte precursors. Development 125:2915–2923. Lahav R, Ziller C, Dupin E, Le Douarin NM. 1996. Endothelin 3 promotes neural crest cell proliferation and mediates a vast increase in melanocyte number in culture. Proc Natl Acad Sci USA 93:3892– 3897. Le Douarin NM. 1982. The neural crest. Cambridge: Cambridge University Press. Manova K, Bachvarova RF. 1991. Expression of c-kit encoded at the W locus of mice in developing embryonic germ cells and presumptive melanoblasts. Dev Biol 146:312–324. Martin GR. 1981. Isolation of a pluripotent cell line from early mouse embryos cultured in medium conditioned by teratocarcinoma stem cells. Proc Natl Acad Sci USA 78:7634–7638. Morrison-Graham K, Weston JA. 1993. Transient steel factor dependence by neural crest-derived melanocyte precursors. Dev Biol 159:346–352. Motro B, Wojtowicz JM, Bernstein A, Van Der Kooy D. 1996. Steel mutant mice are deficient in hippocampal learning but not longterm potentiation. Proc Natl Acad Sci USA 93:1808–1813. Murphy M, Reid K, Williams DE, Lyman SD, Bartlett PF. 1992. Steel factor is required for maintenance, but not differentiation, of melanocyte precursors in the neural crest. Dev Biol 153:396–401. Nishikawa S, Kusakabe M, Yoshinaga K, Ogawa M, Hayashi SI, Kunisada T, Era T, Sakakura T, Nishikawa SI. 1991. In utero manipulation of coat color formation by a monoclonal anti-c-kit antibody: two distinct waves of c-kit-dependency during melanocyte development. EMBO J 10:2111–2118. Ogawa M, Nishikawa S, Ikuta K, Yamamura F, Naito M, Takahashi K, Nishikawa S, Nishikawa SI. 1989. B cell ontogeny in murine embryo studied by a culture system with the monolayer of a stromal cell clone, ST2: B cell progenitor develops first in the embryonal body rather than in the yolk sac. EMBO J 7:1337–1343. Ogawa M, Matsuzaki Y, Nishikawa S, Hayashi SI, Kunisada T, Sudo T, Kina T, Nakauchi H, Nishikawa SI. 1991. Expression and function of c-kit in hematopoietic progenitor cells. J Exp Med 174:63–71. Okura M, Maeda H, Nishikawa SI, Mizoguchi M. 1995. Effects of monoclonal anti-c-kit antibody (ACK2) on melanocytes in newborn mice. J Invest Dermatol 105:322–328. Opdecamp K, Nakayama A, Nguyen M-TT, Hodgkinson CA, Pavan WJ, Arnheiter H. 1997. Melanocyte development in vivo and in neural crest cell cultures: crucial dependence on the Mitf basic-helixloop-helix-zipper transcription factor. Development 124:2377–2386. Porcher C, Swat W, Rockwell K, Fujiwara Y, Alt FW, Orkin SH. 1996. The T cell leukemia oncoprotein SCL/tal-1 is essential for development of all hematopoietic lineages. Cell 86:47–57. Reid K, Nishikawa S-I, Bartlett PF, Murphy M. 1995. Steel factor directs melanocyte development in vitro through selective regulation of the number of c-kit1 progenitors. Dev Biol 169:568–579. Reid K, Turnley AM, Maxwell GD, Kurihara Y, Kurihara H, Bartlett PF, Murphy M. 1996. Multiple roles for endothelin in melanocyte development: regulation of progenitor number and stimulation of differentiation. Development 122:3911–3919. Shibahara S, Okinaga S, Tomita Y, Takeda A, Yamamoto H, Sato M, Takeuchi T. 1990. A point mutation in the tyrosinase gene of BALB/c albino mouse causing the cysteine–serine substitution at position 85. Eur J Biochem 189:455–461. Silvers WK. 1979. The coat colors of mice. New York: Springer Verlag. Sviderskaya EV, Wakeling WF, Bennett DC. 1995. A cloned, immortal line of murine melanoblasts inducible to differentiate to melanocytes. Development 121:1547–1557. Tsukamoto K, Jackson IJ, Urabe K, Montague PM, Hearing VJ. 1992. A second tyrosinase-related protein, TRP-2, is a melanogenic enzyme termed DOPAchrome tautomerase. EMBO J 11:519–526. Vittet D, Prandini MH, Berthier R, Schweitzer A, Martin-Sisteron H, Uzan G, Dejana E. 1996. Embryonic stem cells differentiate in vitro to endothelial cells through successive maturation steps. Blood 88:3424–3431. Wobus AM, Grosse R, Schoneich J. 1988. Specific effects of nerve growth factor on the differentiation pattern of mouse embryonic stem cells in vitro. Biomed Biochim Acta 47:965–973. Yamane T, Kunisada T, Yamazaki H, Era T, Nakano T, Hayashi SI. 1997. Development of osteoclasts from embryonic stem cells through a pathway that is c-fms but not c-kit dependent. Blood 90:3516– 3523. Yasunaga M, Wang FH, Kunisada T, Nishikawa S, Nishikawa SI. 1995. Cell cycle control of c-kit1IL-7R1 B precursor cells by two distinct signals derived from IL-7 receptor and c-kit in a fully defined medium. J Exp Med 182:315–323. Yoshida H, Nishikawa SI, Okamura H, Sakakura T, Kusakabe M. 1993. The role of c-kit proto-oncogene during melanocyte development in mouse. Dev Growth Differ 35:209–220.