SEARCH

SEARCH BY CITATION

Keywords:

  • Melanoblasts;
  • Kit;
  • Melanocytes;
  • Neuron;
  • Glial cells;
  • Neural Crest

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  9. References
  10. Supporting Information

Melanoblasts, precursor of melanocytes, are generated from the neural crest and differentiate into melanocytes during their migration throughout the entire body. The melanoblasts are thought to be progenitor cells that differentiate only into melanocyte. Here, we show that melanoblasts, even after they have already migrated throughout the skin, are multipotent, being able to generate neurons, glial cells, and smooth muscle cells in addition to melanocytes. We isolated Kit-positive and CD45-negative (Kit+/CD45−) cells from both embryonic and neonate skin by flow cytometry and cultured them on stromal cells. The Kit+/CD45− cells formed colonies containing neurons, glial cells, and smooth muscle cells, together with melanocytes. The Kit+/CD45− cells expressed Mitf-M, Sox10, and Trp-2, which are genes known to be expressed in melanoblasts. Even a single Kit+/CD45− cell formed colonies that contained neurons, glial cells, and melanocytes, confirming their multipotential cell fate. The colonies formed from Kit+/CD45− cells retained Kit+/CD45− cells even after 21 days in culture and these retained cells also differentiated into neurons, glial cells, and melanocytes, confirming their self-renewal capability. When the Kit signal was inhibited by the antagonist ACK2, the Kit+/CD45− cells did not form colonies that contained multidifferentiated cells. These results indicate that melanoblasts isolated from skin have multipotency and self-renewal capabilities. STEM CELLS 2009;27:888–897


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  9. References
  10. Supporting Information

Melanoblasts are melanocyte precursors that are derived from neural crest cells (NCCs) that emerge from the dorsal region of the fusing neural tube. Streams of NCCs travel dorsolaterally at the dermamyotome of the embryo toward the ventrum. During this migration, NCCs are thought to become committed progenitors, melanoblasts, before invading the ectoderm to finally colonize the skin and hair follicles, where they differentiate into mature melanocytes [1]. At around embryonic day 8.5 (E8.5), melanoblasts emerge from the dorsal neural tube and migrate ventrally through the developing dermis from E10.5. At E14.5, they begin to invade the overlying epidermis and then migrate into the developing hair follicle, where they continue to proliferate and differentiate, before starting to synthesize pigment at around postnatal day 4 [2, 3].

Melanoblasts depend on numerous signaling systems and transcriptional factors for both their survival and migration. These include the Wnt signaling pathway [4, 5], the G-coupled endothelin B receptor and its ligand endothelin-3 (EDN3) [6–8], the tyrosine kinase receptor Kit and its ligand stem cell factor (SCF, also called Kit-ligand, steel factor) [9, 10], and the transcriptional factors Pax3 [11, 12], Sox10 [12–14], and Mitf [15, 16]. Mutation in these genes interferes with normal generation of melanocytes and leads to pigment dilution, white spotting, or total lack of coat pigmentation [17–19]. Especially, the Kit and SCF mutant mice exhibit striking coat color phenotypes due to the loss of viable melanocytes [20, 21]. For example, unpigmented spots were observed in mice with mutant Kit gene such as KitWv/+, KitW37/W37, KitW/+, or KitlSl/+; and the complete loss of hair pigmentation was observed in homozygous KitW or KitlSl mutant mice [9, 10,22]. These phenotypes could be caused by any one of a number of possible events occurring in melanoblasts such as decreased survival, impaired proliferation, defective differentiation/maturation, or impaired migration.

The immediate descendants of NCCs whose differentiation potential was restricted only to melanocytes have been identified in avian and mammalian embryos. In the quail embryo, the NCCs that express the Kit molecule after emerging from the neural tube represent precursors that differentiate into melanocytes only [23, 24]. In the mouse embryo, the Kit-positive NCCs that emerge at the dorsal midline of the trunk neural tube at E9, precisely in the premigratory neural crest region, migrate only into the ectoderm and develop into melanocytes [25]. Therefore, melanoblasts, which have already migrated out from the neural crest, expressed the Kit molecule, and have moved into the embryonic skin, are thought to be fate-restricted precursors that differentiate only into melanocytes.

Recently, multipotent precursor cells were identified in the skin or the appendages of fetus and adult, and these cells have a differentiation capability similar to that of neural crest stem cells (NCSCs). Several laboratories have reported that cells isolated from the adult skin of mice and humans have self-renewal features and differentiate into neurons, glial cells, smooth muscle cells, melanocytes, chondrocytes, and adipocytes [26–28]. Other studies have shown that cells in adult hair follicles are also capable of differentiating into derivatives of NCCs and these cells were thought to be derived from NCCs according to the cell lineage analysis [29–31].

Furthermore, some reports have indicated that cells differentiated from NCCs also have multipotential cell fate. Pigmented cells isolated from the skin of quail embryos were able to generate glial cells and myofibroblastic cells, which are derivatives of NCCs, when exposed to EDN3 act as a mitotic signal [32–35]. Moreover, it was recently shown that Schwann cells isolated from quail embryonic nerve generated myofibroblastic cells [36]. These reports suggest that the NCC-derived cells, such as melanocytes or glial cells, show unstable NCC phenotypes and can dedifferentiate and then redifferentiate into other NCC derivatives [36]. In this context, it is possible that the melanoblasts, which are thought to be restricted in their fate to melanocytes, might have a multipotential cell fate, even when they have already migrated toward the target sites in the skin.

In this study, we reinvestigated the differentiation of melanoblasts fulfilling the available criteria as the precursors of melanocytes. From E12.5 embryos to P6 neonates, we isolated Kit-positive melanoblasts from the skin and cultured them on monolayers of ST2 stromal cells. They differentiated into not only pigmented melanocytes, but also into neurons, glial cells, and smooth muscle cells, which were not previously thought to be generated from melanoblasts. Our findings suggest that melanoblasts retain the potential to differentiate into other NCC derivatives.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  9. References
  10. Supporting Information

Mice

C57BL/6 mice were purchased from Chubu-Kagakushizai Co. (Nagoya, Japan, http://www.cksdan.com/) KitV602A transgenic mice were established in our laboratory, as described previously [37]. KitW/W mice (purchased from Riken Bioresource Center, Tsukuba, Japan, http://www.brc.riken.jp) and DOPAchrome tautomerase-LacZ (Dct-LacZ) transgenic mice (a gift from Dr. I. Jackson, MRC Western General Hospital, U.K. [38]) were maintained in our animal faculty. At noon of the day the vaginal plug detected was designated as day 0.5 of gestation (E0.5). The developmental stages of embryos were judged by their morphological appearance as described in “The Mouse” [39]. All animal experiments were performed in accordance with the Regulations of Animal Experiments in Gifu University.

Preparation of Cell Suspensions from Mouse Embryonic Skin, Neonatal Skin, and Hair Follicles

The skin of E12.5, E14.5, and E16.5 embryos was removed from the dorsal lateral trunk region with fine forceps. Skin samples were incubated for 6 minutes at 37°C in Dispase II (diluted 1:10 in Ca- and Mg-free PBS; Roche, Diagnostics, Basel, Switzerland, http://www.roche-applied-science.com) and gently dissociated by passing through 18- to 21-gauge needles. The digestion was quenched with two volumes of staining medium (SM: PBS containing 3% FCS). In the case of P0.5 and P6 neonates, the back skin was removed with scissors and incubated for 40 minutes at 37°C in 0.25% Collagenase type I (Sanko Jyunyaku Co., Ltd., Tokyo, Japan, http://www.sanko-junyaku.co.jp/) in PBS. Epidermal sheets were peeled mechanically with fine forceps from digested dermal tissues during observation through a binocular microscope (Stereomicroscope DV4; Carl Zeiss, Jena, Germany, http://www.zeiss.de/micro). The peeled epidermal sheet and dermal tissues were washed with 5 ml of Hanks' balanced salt solution containing 0.005% DNaseI (Roche) and 20% FCS (JRH Biosciences, Inc., Lenexa, KS, http://www.jrhbio.com) [40]. The samples were cut finely, incubated in cell-dissociation buffer (Invitrogen, Carlsbad, CA, http://www.invitrogen.com) at 37°C for 10-15 minutes and dissociated by passage through 18- to 21-gauge needles. The digestion was quenched with two volumes of SM. For isolation of vibrissa follicles, the upper lip containing the vibrissa pad of P6 neonates was cut to expose its inner surface. The vibrissa follicles were gently dissected from the vibrissa pad under a microscope. The isolate follicles were then washed and incubated for 30 minutes at 37°C in 0.05% trypsin (Invitrogen). Some vibrissa follicles were cut into two pieces, that is, upper region and lower region (Fig. 3) and incubated in 0.05% trypsin. The samples were gently dissociated by passage through 21-gauge needles. The digestion was quenched with two volumes of SM.

Flow-Cytometric Analysis, Cell Sorting, and Cell Culture

The dissociated cells were washed with SM and blocked with rat anti-mouse Fc gamma receptor (2.4-G2, BD Biosciences, San Diego, http://www.bdbiosciences.com) on ice for 30-40 minutes. After another wash with SM, the cells stained with FITC-conjugated rat anti-mouse CD45 (30-F11; BD Bioscience) and allophycocyanin-conjugated rat anti-mouse Kit (2B8, BD Bioscience). The cells were washed and resuspended in SM containing 3 μg/ml propidium iodide (PI) (Calbiochem, San Diego, http://www.emdbiosciences.com). All cell sorting and analysis were performed with a FACS Vantage (Becton-Dickinson, Franklin Lakes, NJ, http://www.bd.com). Sorted cells (100-200 or single ones) were directly inoculated into six-well (Nunc, Rochester, NY, http://www.nuncbrand.com) and 96-well plates (Nunc), respectively, by using the CloneCyt Plus v3.1 System (Beckton-Dickinson). The plates were previously seeded with ST2 stromal cells and contained α-MEM (Invitrogen) supplemented with 10% FCS, 10−7 M dexamethasone (Dex; Sigma-Aldrich, St. Louis, MO, http://www.sigmaaldrich.com), 20 pM fibrobrast growth factor-2 (bFGF; R&D Systems Inc., Minneapolis, http://www.rndsystems.com), 10 pM cholera toxin (CT; Sigma), and 100 ng/ml human recombinant EDN3 (Peptide Institute, Inc., Osaka, Japan, http://www.peptide.co.jp/) The cultures were incubated in a 5% CO2 incubator at 37°C, and the medium was changed every 2 days. The day when the sorted cells were seeded onto ST2 monolayers was defined as day 0. ST2 stromal cells were maintained as previously described [41]. For some experiments, 100 ng/ml BQ788 (Peptide Institute, Inc.) or 10 μg/ml of anti-Kit monoclonal antibody (ACK2, a gift from Dr. S. Nishikawa) was added for all days of culture. To investigate the self-renewal property, we incubated colonies formed from cultured Kit+/CD45− cells for 7 minutes at 37°C in Dispase II, mechanically dissociated them, and used the methods described earlier to examine this property. In a clonal density culture experiment, 50 sorted cells were directly inoculated into 6-well plates (Nunc) bearing a monolayer of ST2 cells.

Reverse Transcription-Polymerase Chain Reaction Analysis

Skin Kit+/CD45− or Kit−/CD45− cells was isolated from embryonic of various stages by using the FACS. For the Kit+/CD45− cells cultured for 21 days, the total cells including ST2 cells were collected from the culture dishes. The total RNA was purified by using Isogen (Nippon Gene, Tokyo, http://www.nippongene.com), and first-strand cDNA synthesis was carried out with Superscript III (Invitrogen). The first-strand cDNA mixture was used for PCR conducted with recombinant Taq polymerase (Takara, Otsu, Japan, http://www.takara.co.jp). PCR reactions were performed under the following conditions: 94°C, 2 minutes; 35 to 40 cycles of 94°C for 30 seconds; gene-specific annealing temperature for 30 seconds; and 72°C for 60 seconds. The primers used for PCR are listed in supporting information Materials and Methods.

Immunohistochemical Analysis

Colonies were sequentially fixed in 4% PFA in PBS for 15 minutes, made permeable by immersion in 0.1% Triton X-100 in 0.5% BSA PBS for 30 minutes, washed in PBS, and blocked in a mixture of 3% goat serum or 5% BSA in PBS for 30 minutes. Primary antibodies, diluted in 0.5% BSA PBS, were then added and allowed to react at room temperature. After having been washed in PBS, the cells were stained with the secondary antibodies in the same manner. Colonies were examined by using an Olympus IX-71 fluorescence microscope (Olympus, Tokyo, Japan, http://www.olympus-global.com. The antibodies are listed in supporting information Materials and Methods.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  9. References
  10. Supporting Information

Kit+/CD45− Cells Isolated from Mouse Fetal Skin are Multipotent

We previously reported that mouse ES cells differentiated into neural crest-like cells by coculture with ST2 cell line derived from mouse bone marrow stromal cell [42]. This finding suggested that ST2 cells might have the ability to support the differentiation of embryonic NCCs. In fact, in dissected neural tubes from E10.5 mouse embryos cultured on ST2, delamination of NCCs and subsequent emergence of a number of pigmented melanocytes, β-tubulin (TuJ-1)-positive neuron and glial fibrillary acidic protein (GFAP)-positive glial cells were observed (supporting information Fig. 1), indicating that the ST2 cells were capable of supporting the development of embryonic NCCs.

Next, we isolated melanoblasts from E12.5 fetal skin and estimated the differentiation ability of these cells cultured on ST2 monolayers. In fetal skin, melanoblasts and also hematopoietic cells express Kit molecules on their cell surface [43]. To eliminate the hematopoietic cells, we isolated the Kit-positive and hematopoietic cell-specific marker CD45-negative cells (Kit+/CD45−) from E12.5 fetal skin. In flow cytometric analysis, approximately 3.42% of the total skin cells were Kit+/CD45− cells (Fig. 1A). We isolated and seeded these cells on ST2 cells in αMEM-based differentiated medium supplemented with EDN3, bFGF, CT, and Dex. After 21 days, the cells formed colonies that contained pigmented melanocytes (M, Fig. 1B). Immunostaining showed that some colonies contained TuJ-1-positive neurons (N) and GFAP-positive glial cells (G) together with the M (Fig. 1B, 1C). The Kit+/CD45− cells also formed colonies composed of two cell types (M/N, M/G, and N/G) and of one cell type (M, N, and G). It is thus conceivable that the formerly designated melanoblasts, like NCSCs, have the potential to differentiate into multilineage cells.

thumbnail image

Figure 1. Analysis and isolation of skin Kit+/CD45− cells by flow cytometry. (A): Flow cytometry analysis of Kit+/CD45− cells in skin from E12.5 embryo to P6 neonate. The skin cells were stained with anti-Kit-APC and anti-CD45-FITC antibodies. “Negative” indicates no staining. P0.5 and P6 neonatal skin were separated into epidermis and dermis, and cells from each were analyzed. Kit+/CD45− cells were sorted by flow cytometry (gray line gate) from skin of each embryonic day, and inoculated at 200 cell per well onto ST2 monolayers in a 6-well dish. After 21 days in culture, the colonies were immunostained for neuronal marker β-tubulin (TuJ-1) and glial marker GFAP. Melanocytes were detected as pigmented cells. Frequencies of the production of different types of colonies from Kit+/CD45− cells were calculated. “C” indicates the plating efficiency of the inoculated cells that actually formed colonies. M/N/G indicates that the colonies contained M, N, and G. M/N, M/G, and N/G refer to colonies containing two types of cell; and M, N, and G, to those containing a single cell type. The percentages were calculated as follows: (number of wells containing the indicated types of cells)/(total number of wells containing colonies) × 100. The experiment was performed three times. (B): Kit+/CD45− cells were isolated from E12.5 skin by flow cytometry (A, gray line gate) and cultured on ST2 monolayers for 21 days; and the colonies that emerged were then immunostained for neuronal marker β-tubulin (TuJ-1) and glial marker GFAP. The colonies derived from Kit+/CD45− cells contained TuJ-1-positive neurons (N), GFAP-positive glial cells (G) together with pigmented melanocytes (M). Merge indicate the merged image. The same visual field is shown in each photo. Scale bar = 200 μm. (C): The colonies derived from Kit+/CD45− cells contained bipolar morphology neurons or typical morphology glial cells. Scale bar = 200 μm. (D): Expression of melanoblast marker genes in the Kit+/CD45− cells. Kit+/CD45− cells (K+) and Kit-/CD45− cells (K−) were collected from the skin at each embryonic or postnatal day by flow cytometry. Expression of the indicated genes was analyzed by reversetranscription polymerase chain reaction analysis using RNA from 30,000 cells of each type. Der and Epi indicate dermis and epidermis, respectively. Abbreviations: GAPDH, glyceraldehyde-3-phosphate dehyrogenase; GFAP, glial fibrillary acidic protein; G, glial cell; M, melanocyte; N, neuron.

Download figure to PowerPoint

Multipotent Kit+/CD45− Cells Persist Even in Neonatal Skin

To test whether the multipotency of Kit+/CD45− cells was retained in the neonatal skin, we isolated Kit+/CD45− cells from E12.5 fetal skin to P6 neonatal skin and cultured them on ST2. The Kit+/CD45− cells from all stages formed colonies that contained N, G, and M (Fig. 1A, supporting information Fig. 2). Also, we detected α-smooth muscle-positive cells together with G and M (supporting information Fig. 3). These results suggest that the multipotency of Kit+/CD45− cells was retained even in neonatal skin. Although the colony-forming efficiency of the neonatal Kit+/CD45− cells was increased, the percentage of M/N/G colonies and of colonies that contained G was decreased in the cultures from the P6 neonate (Fig. 1A). In contrast, the percentage of M/N colonies increased continuously with age (Fig. 1A). Epidermal and dermal Kit+/CD45− cells underwent a similar cell fate in P0.5 neonate; however, only epidermal Kit+/CD45− cells generated M/N/G colonies in P6 neonates. These results suggest that the Kit+/CD45− cells retained their multipotency even in the neonate.

Kit+/CD45− Cells are Properly Categorized as Melanoblasts

To investigate the gene expression of Kit+/CD45− cells, we used reverse transcription polymerase chain reaction analysis (RT-PCR) to analyze cells purified from E12.5 skin to P6 skin. The Kit+/CD45− cells from all stages expressed Mitf-M, Sox10, Pax3, Trp-2, and Trp-1, all of which are known to be expressed in melanoblasts (Fig. 1D) [44, 45]. The lack of expression of the Nestin gene, which is expressed in neuronal cells, suggests that neuronal cells did not contaminate the Kit+/CD45− fraction.

To further investigate whether the Kit+/CD45− cells were indeed melanoblasts, we manipulated the Kit signal, which is a survival signal for proliferating melanoblasts [46]. First, we cultured the separated Kit+/CD45− cells under conditions inhibiting the Kit signal by adding the antagonistic antibody ACK2. When the Kit signal was blocked, the colony-forming efficiency was drastically decreased, and the generated colonies did not contain M, N, or G (Fig. 2). Next, we separated Kit+/CD45− cells from the skin of KitW/W mouse, which expresses loss-of-function forms of Kit receptor, and induced the cells to form colonies on ST2 monolayers. In KitW/W mouse, the Kit+/CD45− cells did not emerge; and even after the culture of nonseparated KitW/W mouse skin, colonies containing M, N, or G did not emerge (supporting information Fig. 4A). In the KitW/+ mouse with a reduced Kit signaling, purified Kit+/CD45− cells formed colonies that contained M and N (supporting information Fig. 4A). Finally, we separated Kit+/CD45− cells from the skin of a KitV620A transgenic mouse [37], in which a dominant-negative form of the Kit receptor was induced and whose Kit signaling was thus reduced. Although a small number of Kit+/CD45− cells were purified, the colonies that originated from KitV620A transgenic mice contained M and N (supporting information Fig. 4B). All these results suggest that Kit+/CD45− cells should be designated as melanoblasts, based on the widely accepted concept that melanoblasts or the melanocyte lineage cells are Kit signal dependent.

thumbnail image

Figure 2. Kit+/CD45− cells cultured under conditions inhibiting Kit signaling and EDN3. The Kit+/CD45− cells in E12.5 embryonic skin were cultured with BQ788, ACK2, or both. Colonies that emerged were stained with TuJ-1 and anti-glial fibrillary acidic protein. Melanocytes were detected as pigmented cells. Frequencies of the production of different types of colonies in each type of culture were calculated. The percentages were calculated as in Figure 1. The experiment was performed three times. Abbreviations: C, plating efficiency; G, glial cells; M, melanocytes; N, neurons.

Download figure to PowerPoint

Next, we used the Dct-LacZ transgenic mouse whose melanoblasts may be traced as LacZ-positive cells [38]. By flow cytometric analysis, only 0.34% of the E12.5 embryonic skin cells were Kit+/Dct+ cells (supporting information Fig. 5A). We isolated the Kit−/Dct−, Kit−/Dct+, Kit+/Dct−, and Kit+/DCT+ cells and cultured them on ST2. No colonies emerged from the cultures of Kit−/Dct− or Kit−/Dct+ cells (data not shown). In contrast, both Kit+/Dct− and Kit+/Dct+ cells were found to form M/N/G, N/G, and N colonies (supporting information Fig. 5B, 5C). Kit+/Dct− and Kit+/Dct+ cells formed a similar percentage of M/N/G or N/G colonies. These data suggest that Kit+/Dct+ and Kit+/Dct− cells are to be classified as melanoblasts.

We then immunostained isolated Kit+/CD45− cells or Kit+/Dct− cells with anti-Mitf antibodies. Mitf is another melanoblast marker [16], and approximately 73% of the Kit+/CD45− cells and 74% of the Kit+/Dct− cells were Mitf+ (supporting information Fig. 6A, 6B). These data suggest that most part of the Kit+/CD45− cells should be classified as melanoblasts.

Terminal Migrated Kit+/CD45− Melanoblasts are also Multipotencial

In the mouse, developing melanoblasts migrate to cover the whole body and enter the hair follicles [46]. Melanoblasts are maintained as melanocyte stem cells in the upper bulge area of hair follicles, and they differentiate into pigmented melanocytes during their migration to the lower hair bulb. To test whether the hair follicles contain multipotential cells or not, we dissected both the upper and the lower region of the hair follicles from mouse vibrissa (Fig. 3A), dissociated the cells from each part, and cultured them on ST2. The cells from both parts formed colonies having M, N, and G in comparable percentages (Fig. 3A, 3B). This finding suggests that the multipotential cells were distributed rather uniformly in the hair follicle. Then, we isolated Kit+/CD45− cells from the hair follicles (Fig. 3C). Although the percentage of Kit+/CD45− cells was as low as 0.1% of the total hair follicle cells (data not shown), the cells formed colonies containing N and M.

thumbnail image

Figure 3. Terminally migrated Kit+/CD45− cells also have multipotency. (A): Hair follicle cells were harvested from both the upper region (bulge area) and the lower region (hair bulb) of mouse vibrissa, and the cells from each region were cultured on ST2 cells. After 21 days in culture, the colonies were stained with TuJ-1 and anti-GFAP. Melanocytes were detected as pigmented cells. Frequencies of the production of different types of colonies from cells from each region were calculated. (B): A typical M/N/G colony generated from the hair follicle cells is shown in each panel. Merge indicates the merged image. The same field is shown in each panel. Scale bar = 200 μm. (C): Kit+/CD45− cells were isolated from hair follicles and cultured on the ST2 cells. Frequencies of the production of different types of colonies from Kit+/CD45− cells were calculated. The percentages were calculated as in Figure 1. The experiment was performed three times. Abbreviations: G, glial cells; GFAP, glial fibrillary acidic protein; M, melanocytes; N, neurons.

Download figure to PowerPoint

Single Kit+/CD45− Cells Formed M/N/G Colonies

To clarify that Kit+/CD45− cell could precisely exert multipotency, we isolated single Kit+/CD45− cell from E12.5, P0.5, and P6 neonate skin and cultured them on ST2. From these single Kit+/CD45− cells obtained from E12.5 skin, colonies arose consisting of M, N, and G; but those from P0.5 or P6 skin did not form M/N/G colonies (Table 1). In contrast, the single cells from P0.5 and P6 formed a higher percentage of M/N colonies (Table 1). Epidermal and dermal Kit+/CD45− cells from P0.5 skin formed M/N colonies in similar proportion; but in the case of P6 skin, only epidermal Kit+/CD45− cells formed such colonies. These results show that Kit+/CD45− cells were multipotential even in the isolated single cell state; however, the Kit+/CD45− cells tend to be restricted to M or N cell fates along with increasing age.

Table 1. A single Kit+/CD45− cell generated multilineage colonies
inline image

Self-Renewal Capacity of Kit+/CD45− Cells

To investigate whether the multipotent Kit+/CD45− cells had self-renewal capacity, we analyzed the persistence of the Kit+/CD45− cells in the colonies derived from Kit+/CD45− cells after 21 days in culture (indicated as “secondary” in Fig. 4A). Flow cytometric analysis showed that 9.30% of the colonies formed were Kit+/CD45− cells (Fig. 4A). Next, we isolated these Kit+/CD45− cells (secondary Kit+/CD45− cells, Fig. 4A) from the colonies and cultured them on ST2. The colony-forming efficiency of these secondary Kit+/CD45− cells was 11%, and various colonies containing three types (M/N/G), two types (N/G, M/N, M/G), or a single type of cell were generated (Fig. 4B). From these secondary Kit+/CD45− cells, a lower percentage of M/N/G colonies was generated compared with that for the first colonies, and instead, the percentages of M/N and M colonies were increased (Fig. 4B). We further analyzed Kit+/CD45− cells in the colonies derived from secondary Kit+/CD45− cells; however, no Kit+/CD45− cells were detected after 21 days (designated as tertiary cells in Fig. 4A). By transferring all of the cells harvested from the colonies from secondary Kit+/CD45− cells onto ST2, no colonies were formed (data not shown). These results suggest that the Kit+/CD45− cells have limited but significant self-renewal capacity as multipotental precursors even under in vitro culture conditions.

thumbnail image

Figure 4. Characterization of Kit+/CD45− cells. (A, B): Kit+/CD45− cells have self-renewal property. (A): The Kit+/CD45− cells were isolated from E12.5 embryonic skin by flow cytometry, analyzed (primary, gray line gate), and then cultured on stromal cells for 21 days. The colonies that emerged from these primary Kit+/CD45− cells were dissociated and analyzed. The Kit+/CD45− cells were present even after 21 days in culture (secondary). These secondary Kit+/CD45− cells were isolated and cultured for another 21 days. The colonies that emerged from these secondary Kit+/CD45− cells were dissociated and analyzed (tertiary). (B): Colonies that emerged from primary and secondary Kit+/CD45− cells were stained for TuJ-1 and anti-glial fibrillary acidic protein (GFAP). Melanocytes were detected as pigmented cells. Frequencies of the production of different types of colonies from Kit+/CD45− cells were calculated. (C, D): Kit high-expressing cells retained their multipotency. (C): The Kit high-expressing (Kit high) and low-expressing (Kit low) cells were isolated from E12.5 embryonic skin by flow cytometry (primary, gray line gate) and cultured on stromal cells for 21 days. The colonies that emerged from primary Kit+/CD45− cells were dissociated; and again, the Kit high- and low-expressing cells were isolated (secondary, gray line gate). (D): Colonies emerged from Kit-high and Kit-low cells of primary and second cultures were stained TuJ-1 and anti-GFAP. Melanocytes were detected as pigmented cells. Frequencies of the production of different types of colonies from each type of cell were calculated. The percentages were calculated as in Figure 1. The experiment was performed three times. Abbreviations: C, plating efficiency; G, glial cells; M, melanocytes; N, neurons.

Download figure to PowerPoint

Kit Expression and the Multipotency

As we noticed a broad range of distribution of the amount of Kit expression, the populations with higher Kit expression (Kit high) and lower Kit expression (Kit low) were separated from E12.5 skin (Fig. 4C), cultured on ST2, and identified for types of differentiated cells by immunohistochemistry. The colony-forming efficiency and the percentage of colony-types generated were similar between Kit high and Kit low populations (indicated as “Primary” in Fig. 4D). This finding indicates that the number of cell-surface Kit molecules did not affect the fate of Kit+/CD45− cells. Furthermore, we isolated the Kit-high and Kit-low populations from the primary colonies (indicated as “secondary Kit-high” and “secondary Kit-low” in Fig. 4C) and cultured them on ST2. The colony efficiency was similar between secondary Kit-high and secondary Kit-low populations. From the former population, M/N/G colonies and M/N colonies were generated at 14% and 46%, respectively (Fig. 4D). In contrast, from the secondary Kit-low population, no M/N/G colonies were generated; M/N colonies decreased significantly in number, and instead, M colonies increased to 62% (Fig. 4D). Thus, higher Kit expression in primary colonies is likely to be related to the enhanced multipotency of the precursor cells during the differentiation of the secondary colonies.

Next, we measured the mRNA level of the Mitf, Dct, and Sox10 of the primary and secondary sorted Kit high+ population and Kit low+ population by real-time PCR. Opdecamp et al. reported that the Kit positive melanoblasts can be categorized into three types: melanoblast precursor (Kit low+/Mitf+/Dct−), early melanoblast (Kit low+/Mitf+/Dct+), and late melanoblast (Kit high+/Mitf+/Dct+) [16]. In the primary sorted cells, the mRNA levels of the Mitf, Dct, and Sox10 were lower in both Kit high+ population and Kit low+ populations in comparison with those in melanoblast cell line, Melb-a [47] (supporting information Fig. 7). This finding suggests that the primary Kit+ population is likely to be early melanoblast, not melanoblast precursor. In the case of the secondary Kit+ cells, both Kit high+ and Kit low+ population expressed a similar level of the Mitf and Sox10 mRNA as the Melb-a cells, but a higher level of the Dct than them (supporting information Fig. 7). These data suggest that the secondary Kit+ population is likely to be the late melanoblast. For the secondary Kit+ cells, only the Kit high+ population formed M/N/G colonies (Fig. 4D); whereas, the Kit low+ population tended to form more M colonies (Fig. 4D). This finding suggests that the developmental potential of the late melanoblast population is related to the Kit expression.

Kit Signal and Multipotency of Kit+/CD45− Cells

To investigate how critical the microenvironment was for the multipotency of the Kit+/CD45− cells, we cultured the cells on another stromal cell line, PA6. Although the least number of colonies containing M, N, and G were formed, greatly reduced numbers of colonies were generated on PA6 monolayers, indicating that the cues provided by the stromal cells determined the efficiency of cell differentiation (supporting information Fig. 8). Next, to test the roles of signal transduction system in the multipotency and differentiation of Kit+/CD45− cells, we manipulated EDN3 and Kit-ligand (SCF), both of which are critical for the regulation of melanocyte development [48]. The colony-forming efficiency slightly decreased by the inhibition of the EDN3 signal (17% added with EDN3 inhibitor BQ788 and 25% in the control, Fig. 2), and the percentage of M/N/G colonies generated decreased to 28% compared with the 58% for the control (Fig. 2). More colonies (39%) containing only N were formed by the inhibition of the EDN3 signal compared with the percentage for the control (21%). In contrast, when the SCF signal was inhibited by the addition of the antagonistic antibody ACK2, the colony-forming efficiency drastically decreased to 6%; and the generated colonies did not contain N or G (Fig. 2). When both signals were blocked, the colony-forming efficiency also drastically decreased to 7%, and no N or G were generated (Fig. 2). These results suggest that the SCF signal is critical for the maintenance of multipotency and differentiation of the cells. On the other hand, EDN3 signaling appears to have a less important role for the multipotency of the Kit+/CD45− cells, rather regulating the differentiation of the Kit+/CD45− cells.

Neurons and Glial Cells were Differentiated from Kit+/CD45− Melanoblasts, Not from Dedifferentiated Pigmented Cells

Quail pigmented cells isolated from embryonic skin are able to generate glial cells and myofibroblastic cells, which are derivatives of NCCs, when exposed to the mitogenic signal of EDN3 [32–35]. This phenomenon is thought to reflect dedifferentiation of pigmented cells into NCCs [34]. To investigate this possibility in mouse fetal skin, we cultured Kit+/CD45− cells isolated from E12.5 skin at clonal densities and immunostained them at days 1, 3, 6, 9, and 12 of culture. At 1, 3, and 6 day of culture, no pigmented cells were observed; on the other hand, a small number of GFAP+ glial cells (G) or TuJ-1+ neurons (N) was observed at 3 days of culture (Fig. 5A). At 9 days of culture, a small number of pigmented cells (P) were observed together with G and N cells. The differentiated cells formed cluster consisting of several cells. The clusters contained three types (P, N, and G), two types (P/N, N/G), or single type of cells (N, G; Fig. 5B). In this observation, differentiation to N and G was earlier or simultaneous with that to P cells; and N or G cells were never generated after emergence of the P cells. Next, to further investigate Kit+/CD45− cell differentiation, we isolated E16.5 fetal Kit+/CD45− skin cells from a Dct-LacZ mouse and immunostained the isolated cells at days 1, 6, 7, and 8 in culture. After 1 day in culture, only several LacZ-positive cells (L) were detected; on the other hand, P, N, or G cells were never detected. At 6 days in culture, the L cells formed clusters consisting of several cells. A number of the clusters contained only L cells, although several clusters contained P and L, N and L, or G and L cells (supporting information Fig. 9). Only a few cluster contained P, N, and L cells (supporting information Fig. 9). This observation also indicates that differentiation to N and G occurred earlier or simultaneously with that to P cells. These results suggest that N and G were directly differentiated from Kit+/CD45− melanoblasts, not from dedifferentiated pigmented cells as reported for quail pigmented cells.

thumbnail image

Figure 5. Differentiation of the Kit+/CD45− cells at clonal densities. Kit+/CD45− cells in the skin of E12.5 embryos were isolated by flow cytometry, cultured on ST2 monolayers at clonal densities, and immunostained with TuJ-1 (red) and anti-glial fibrillary acidic protein (green) at the days 3, 6 (A), 9 (B) of culture. The nuclei were stained with Hoechst 33258 (blue). Arrowheads show the pigmented cells. Both photos in “B” are the same visual field. Scale bar = 200 μm.

Download figure to PowerPoint

Kit+/CD45− Cells Differentiate into Autonomic Neuronal Precursors

To investigate the types of cells differentiated from the Kit+/CD45− cells, we immunostained the colonies for certain neuronal, glial, and melanocyte markers. Immunostaining showed that colonies contained a number of Dct- or S100β-positive melanocyte lineage cells (supporting information Fig. 10A, 10B) as well as Nestin-positive and TuJ-1-positive/Neurofilament-M-positive neuronal cells (supporting information Fig. 10C, 10D). Some TuJ-1+ cells also expressed Dct (supporting information Fig. 10D). GFAP-positive glial cells did not express S100β (supporting information Fig. 10B).

Next, we performed RT-PCR analysis of the neuronal marker genes to investigate the types of neurons finally differentiated from the Kit+/CD45− cells. The Kit+/CD45− cells cultured on ST2 expressed Mash-1 and Peripherin, which are known to be expressed in autonomic neuronal precursors [49] (supporting information Fig. 10E). Next, we transplanted 200-1,000 Kit+/CD45− cells directly into in vitro cultured fetal duodenum or ileum obtained from an E12.5 mouse and incubated the cultures for 7 days. The transplanted cells survived in the gut walls, but they did not differentiate into neurons (supporting information Fig. 10F). These results show that the Kit+/CD45− cells could differentiate into autonomic neuronal precursor in vitro culture but that the organ culture conditions presently used were not conducive to neuronal differentiation.

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  9. References
  10. Supporting Information

We showed herein that the Kit+/CD45− cells, regarded as melanoblasts, have multipotency even after having migrated out to cover the entire skin including the hair follicles. Isolated Kit+/CD45− melanoblasts differentiated into N, G, and smooth muscle cells together with M in vitro, even from a single Kit+/CD45− cell. Along with the in vitro differentiation in the developing colonies started from Kit+/CD45− cells, limited but significant self-renewal capacity of these cells was confirmed.

The RT-PCR results showed that the Kit+/CD45− cells were able to differentiate into autonomic neuronal precursors in culture (supporting information Fig. 10E). To investigate whether other types of neurons aside from autonomic neuronal precursors were differentiated from the Kit+/CD45− cells, we immunostained the colonies formed with antibodies against various neuronal markers. The colonies contained TuJ-1- or nestin-positive neurons but did not contain other neuronal markers such as tyrosine hydroxylase, Phox2b, or Brn3a (data not shown). Therefore, the Kit+/CD45− cells cultured on ST2 cells may have generated only immature neurons, such as autonomic neuronal precursors. We tried to culture the Kit+/CD45− cells under SKP cells (multipotent adult skin-derived precursor cells) culture conditions in which SKP cells differentiate into catecholaminergic neurons [26]; however, no neurons and no colonies were generated (data not shown). Furthermore, we transplanted the Kit+/CD45− cells directly into in vitro cultured mouse fetal gut. The transplanted cells survived in the gut walls, but they did not differentiate into neurons (supporting information Fig. 10F). Other supplementations or environments are apparently required for differentiation into mature neurons. On the other hand, some TuJ-1+ cells also expressed Dct, a known melanoblast marker (supporting information Fig. 10A), and some pigmented cells also expressed TuJ-1. The function or the role of the TuJ-1+/Dct+ neuronal cells is unknown. We are presently investigating the nature of the TuJ-1+/Dct+ cells and TuJ-1+ pigment cells.

So far, melanoblasts were thought to be unipotent precursors that only differentiate into melanocytes. Especially, once NCCs express Kit, the NCCs migrate only toward the epidermis along the dorsolateral pathway and differentiate into M only [25]. We showed here that the Kit+/CD45− melanoblasts maintained their multidifferentiation ability even after they had migrated into the epidermis. It is possible that melanoblasts are multipotential in nature and that the skin microenvironment continuously restricts their fate to M. If the appropriate condition allows the melanoblasts to be released from the suppressive environment, these cells are likely to show their multipotency. The ST2 cells and culture medium used in our present experiments may provide a permissive environment to allow the differentiation into N, G, and smooth muscle cells in addition to M. The fact that Kit+/CD45− cells cultured on PA6 cells, which were derived from mouse cranial skull, showed greatly reduced numbers of M/N/G colonies (supporting information Fig. 6) might be indicative of such environmental suppression. As we confirmed that Kit+/CD45− cells formed M/N/G colonies even in cultures without Dex or CT or bFGF, which are all supplements in our culture (data not shown), these supplements are likely not to be important for the cancellation of suppression. Extremely, NCCs specified to be other than the unipotent precursors might be eliminated, as is likely in the avian embryo in which neuronal NCCs that faultily migrated into the dorsolateral pathway are removed by apoptosis [50].

Alternatively, melanoblasts might have dedifferentiated into NCCs and then redifferentiated into M, N, and G like avian pigmented cells. However, we observed that Kit+/CD45− cells directly differentiated into M, N, and G, not via dedifferentiated NCCs originated from pigmented melanocytes in the clonal culture (Fig. 5). Experiments using Kit+/CD45− cells isolated from Dct-LacZ mouse also showed that Dct+ cells have directly differentiated into M, N, and G (supporting information Fig. 9). Furthermore, the Kit+/CD45− cells formed colonies containing M, N, and G cells even in the absence of EDN3, which reportedly induces the reversion of avian melanocytes to NCCs [32] (Fig. 2). These results support the notion that melanoblasts are likely to directly differentiate into M, N, and G.

What role does the multipotency of the Kit+ melanoblasts play in vivo? Nevocellular nevus is a benign melanocytic neoplasm that occurs in the dermo-epidermal junction and consists of melanocyte-like cells in the epidermis and Schwann cell-like cells in the dermis. One hypothesis for the origin of nevocellular nevus is that these cells are derived from common pluripotential stem cells or NCCs [51]. Taking into account the multipotency of the Kit+ melanoblasts revealed presently, the origin of the nevocellular nevus might be these Kit+ melanoblasts. Because Kit+ melanoblasts have the ability to differentiate into M and G in vitro, they might differentiate in a disordered manner and form nevocellular nevus, if the suppression of the multipotency was released in vivo under certain circumstances.

In this study, we have demonstrated the multipotency of Kit-positive melanoblasts in the skin. Although the mechanisms allowing the multipotency and the biological significance of the multipotency are yet largely unknown, melanoblasts revealed as multipotent cells could be used as cell sources for developmental studies on the stem cells and also for cell therapy.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  9. References
  10. Supporting Information

We thank Dr. Shin-Ichi Nishikawa for the ACK2 antibody, Dr. I. Jackson for Dct-LacZ mice, and other members in our laboratory for discussion and critical reading of the manuscript. This study was supported by a grant from the program Grants-in-Aid for Scientific Research (C) from the Japan Society for Promotion for Science.

References

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  9. References
  10. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  9. References
  10. Supporting Information

Additional supporting information available online.

FilenameFormatSizeDescription
STEM_20080678_sm_suppinfo1.doc76KSupporting Information 1
STEM_20080678_sm_suppinfo2.doc72KSupporting Information 2
STEM_20080678_sm_suppinfofigure1.eps3616KSupporting Information Figure 1
STEM_20080678_sm_suppinfofigure2.eps12133KSupporting Information Figure 2
STEM_20080678_sm_suppinfofigure3.eps5147KSupporting Information Figure 3
STEM_20080678_sm_suppinfofigure4.eps361KSupporting Information Figure 4
STEM_20080678_sm_suppinfofigure5.eps3036KSupporting Information Figure 5
STEM_20080678_sm_suppinfofigure6.eps2036KSupporting Information Figure 6
STEM_20080678_sm_suppinfofigure7.eps278KSupporting Information Figure 7
STEM_20080678_sm_suppinfofigure8.eps2446KSupporting Information Figure 8
STEM_20080678_sm_suppinfofigure9.eps4733KSupporting Information Figure 9
STEM_20080678_sm_suppinfofigure10.eps6669KSupporting Information Figure 10
STEM_20080678_sm_suppinfofigure11.eps3616KSupporting Information Figure 11
STEM_20080678_sm_suppinfofigure12.eps12133KSupporting Information Figure 12
STEM_20080678_sm_suppinfofigure13.eps5147KSupporting Information Figure 13
STEM_20080678_sm_suppinfofigure14.eps361KSupporting Information Figure 14
STEM_20080678_sm_suppinfofigure15.eps3036KSupporting Information Figure 15
STEM_20080678_sm_suppinfofigure16.eps2036KSupporting Information Figure 16
STEM_20080678_sm_suppinfofigure17.eps285KSupporting Information Figure 17
STEM_20080678_sm_suppinfofigure18.eps2446KSupporting Information Figure 18
STEM_20080678_sm_suppinfofigure19.eps4733KSupporting Information Figure 19
STEM_20080678_sm_suppinfofigure20.eps6669KSupporting Information Figure 20

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.