Stem Cells
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

First published online December 9, 2005
Stem Cells Vol. 24 No. 3 March 2006, pp. 575 -586
doi:10.1634/stemcells.2005-0256; www.StemCells.com
© 2006 AlphaMed Press

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Data
Right arrow All Versions of this Article:
2005-0256v1
2005-0256v2
24/3/575    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Reprints/Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Sakurai, H.
Right arrow Articles by Nishikawa, S.-I.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Sakurai, H.
Right arrow Articles by Nishikawa, S.-I.
EMBRYONIC STEM CELLS

In Vitro Modeling of Paraxial and Lateral Mesoderm Differentiation Reveals Early Reversibility

Hidetoshi Sakuraia,b, Takumi Eraa, Lars Martin Jakta,c, Mitsuhiro Okadaa,c, Shigeru Nakaib, Satomi Nishikawaa, Shin-Ichi Nishikawaa

a Laboratory for Stem Cell Biology, RIKEN Center for Development Biology, Kobe, Japan;
b Science of In-Home Medicine, Health and Community Medicine, Nagoya University Graduate School of Medicine, Nagoya, Japan;
c Prefecture Collaboration of Regional Entities for the Advancement of Technological Excellence, Japan Science and Technology Corporation, Discovery Research Laboratory, Tanabe Seiyaku Co., Ltd., Osaka, Japan

Key Words. Embryonic Stem cell • Mesoderm • Reversibility

Correspondence: Takumi Era, M.D., Ph.D., Laboratory for Stem Cell Biology, RIKEN Center for Development Biology, 2-2-3 Minatojimaminamimachi, Chuo-ku, Kobe, 650-0047, Japan. Telephone: 81-78-306-1893; Fax: 81-78-306-1895; e-mail: tera{at}cdb.riken.jp

Received on June 7, 2005; accepted for publication on September 6, 2005.


   ABSTRACT
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
 References
 
Endothelial cells (ECs) are thought to be derived mainly from the vascular endothelial growth factor receptor 2 (VEGFR-2)+ lateral mesoderm during early embryogenesis. In this study, we specified several pathways for EC differentiation using a murine embryonic stem (ES) cell differentiation culture system that is a model for cellular processes during early embryogenesis. Based on the results of in vitro fate analysis, we show that, in the main pathway, committed ECs are differentiated through the VEGFR-2+ platelet-derived growth factor receptor {alpha} (PDGFR-{alpha})– single-positive (VSP) population that is derived from the VEGFR-2+PDGFR-{alpha}+ double-positive (DP) population. This major differentiation course was also confirmed using DNA microarray analysis. In addition to this main pathway, however, ECs also can be generated from the VEGFR-2–PDGFR-{alpha}+ single-positive (PSP) population, which represents the paraxial mesodermal lineage and is also derived from the DP population. Our results strongly suggest that, even after differentiation from the common progenitor DP population into the VSP and PSP populations, these two populations continue spontaneous switching of their surface phenotype, which results in switching of their eventual fates. The rate of this interlineage conversion between VSP and PSP is unexpectedly high. Because of this potential to undergo fate switch, we conclude that ECs can be generated via multiple pathways in in vitro ES cell differentiation.


   INTRODUCTION
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 Disclosures
 References
 
The formation of the vascular system is an essential process for the neogenesis of all tissues during embryogenesis and postnatal development [1, 2]. Because this process requires the recruitment of endothelial cells (ECs), how a sufficient number of ECs is enlisted to the region in need is one of the fundamental questions facing vascular biology. It is widely accepted that ECs are generated from vascular endothelial growth factor receptor 2 (VEGFR-2)+ lateral mesodermal cells at the initial phase of vasculogenesis [3], and their subsequent recruitment is supported by the proliferation of differentiated ECs. Recent studies have also suggested alternative sources for these cells [4 – 6].

Our studies have focused on the earliest process of mouse EC differentiation [7–9], for which murine embryonic stem (ES) cell differentiation culture provides an ideal tool, as it allows detailed analysis of the cell specification process from the pluripotent stage [7, 10, 11]. We have attempted to define the process of EC differentiation in terms of the expression of surface markers and showed that VEGFR-2, an essential molecule for all neoangiogenic processes [12], is the key marker required for a detailed examination of this process [7]. In our previous studies, we proposed that ECs are derived from an E-cadherin (ECD)–VEGFR-2+ population [7, 13]. In this concept, ECD expression is used to monitor the process of exfoliation, as its expression is downregulated during the exfoliation of mesoderm cells from the primitive streak [14]. In vitro fate analysis demonstrated that ECD–VEGFR-2+ cells derived from ES cells have the potential to produce endothelial cells, hematopoietic cells, and {alpha} smooth muscle actin–positive pericytes [7, 9, 15], indicating that ECD–VEGFR-2+ cells do indeed correspond to lateral mesodermal cells. In contrast to these findings, previous studies have shown that platelet-derived growth factor receptor {alpha} (PDGFR-{alpha}) is expressed in paraxial mesoderm during mouse embryogenesis [16, 17]. Although PDGFR-{alpha}+ cells were observed in ES cell differentiation culture together with VEGFR-2+ cells [7], the question of whether PDGFR-{alpha} expression specifies the paraxial mesodermal lineage remains unanswered, as does the potential of the ECDlow PDGFR-{alpha}+ population to give rise to ECs during in vitro ES cell differentiation.

A recent study has interestingly demonstrated that cartilage can be differentiated not only from PDGFR-{alpha}+ but also from VEGFR-2+ cells [18]. A lineage analysis by Sato and his colleagues further showed that Crerecombinase expressed under the promoter of the VEGFR-2 gene marked, in addition to the endothelial cell lineages, skeletal muscle cells, which are supposed to be derived from paraxial mesoderm [19]. In contrast to this, mounting evidence suggests that a portion of ECs can be derived from somites [6].

These observations, taken together, are a strong argument against our previous simple model, which specified VEGFR-2+ lateral mesoderm as the sole pathway for EC differentiation [7, 13]. There are two possibilities that may account for the expression of VEGFR-2 in the non-EC differentiation pathway. One possibility is that VEGFR-2 is expressed in uncommitted mesoderm that is subsequently able to give rise to paraxial mesodermal lineages. Alternatively, lineage switching may occur between the early stages of lateral and paraxial mesoderm.

The current research shows that both these possibilities do indeed contribute to the case for EC differentiation. Here, we present a new differentiation pathway in which the PDGFR-{alpha}+VEGFR-2+ double-positive (DP) population gives rise to both the VEGFR-2+ single-positive (VSP) population and the PDGFR-{alpha}+ single-positive (PSP) population. By both DNA microarray analysis and in vitro fate analysis, we demonstrate that the ES cell-derived VSP and PSP populations correspond to the lateral and paraxial mesodermal lineages, respectively. Although the PSP population represents the paraxial mesodermal lineage, it is able to give rise to ECs; our results show that this phenomenon is due to interlineage conversion between the VSP and PSP populations following divergence.


   MATERIALS AND METHODS
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 Disclosures
 References
 
Construction and Cell Line
Green fluorescent protein (GFP)-expressing vectors were constructed by inserting an eGFP [20] cDNA downstream of the CAG promoter [21]. In this vector, the G418 resistance gene is driven by the PGK promoter. The constructs were transfected into CCE ES cells by electroporation [22]. Stably transduced ES cells were established as G418-resistant colonies. After the colonies were picked up, the expression of GFP in undifferentiated and day 4 differentiated ES cells was confirmed by fluorescence-activated cell sorting (FACS) analysis. We established six clones that stably expressed GFP.

Cell Culture and In Vitro ES Cell Differentiation
Murine CCE, TT2 ES cells and OP9 stromal cells were maintained as described previously [7, 23, 24]. To maintain the expression of GFP, we continuously cultured the ES cells carrying GFP in 200 µg/ml G418 (Gibco-BRL, Gaithersburg, MD, http://www.gibcobrl.com).

Induction of ES cell differentiation was performed as described previously [7]. For reculture studies, 1.0–5.0 x 105 cells purified by FACS were recultured in a confluent OP9 cell layer on 24-well plates using {alpha} minimal essential medium ({alpha}-MEM) (Gibco-BRL) supplemented with 10% fetal calf serum (FCS) and 50 µM 2-mercaptoethanol (2ME). Twenty-four hours later, the cells were harvested and collected for examination of the surface markers and gene expression.

Antibodies, Cell Staining, and FACS Analysis
The rat monoclonal antibodies (moAbs) APA5 (anti-PDGFR-{alpha}) [25], ECCD2 (anti-ECD) [26], AVAS12 (anti-VEGFR-2) [13], and VECD1(anti-VE-cadherin) [27] were prepared as reported previously [7]. Phycoerythrin-conjugated streptavidin (Pharmingen, Franklin Lakes, NJ, http://www.bdbiosciences.com) was used for detecting biotinylated-APA5 Ab and biotinylated anti-CD34 moAb (Pharmingen). ECCD2, AVAS12, and VECD1 moAbs were directly conjugated using standard methods and Alexa488 and Alexa405 (Molecular Probes, Eugene, OR, http://probes.invitrogen.com), allophycocyanin (APC), and Alexa680 (Molecular Probes), respectively.

Cultured cells were harvested and collected using either 0.05% trypsin-EDTA (Gibco-BRL) or a dissociation buffer (Gibco-BRL), which was shown not to affect the surface expression levels of ECD and VE-cadherin (VECD) [7]. Single-cell suspensions were stained as previously described [7] and analyzed or sorted by FACSCalibur, FACSVantage, or FACS-Aria (Becton, Dickinson and Company, Franklin Lakes, NJ, http://www.bd.com).

Culture Conditions for Myogenesis, Osteogenesis, Chondrogenesis, Vasculogenesis, and Hematopoiesis
ES cell-derived mesodermal cells purified by FACS were recultured in distinctive conditions specific to individual lineages. For myogenesis, sorted cells were cultured on a collagen type I–coated dish in KnockOut Dulbecco’s modified Eagle’s medium (D-MEM; Gibco-BRL) supplemented with 5% horse serum and 2 ng/ml insulin-like growth factor-1 [28]. On day 14, expression of myocyte-specific markers was analyzed using immunohistochemistry. For osteogenesis, each fraction was re-cultured on a gelatin-coated dish in KnockOut D-MEM supplemented with 10% FCS, 0.1 µM dexamethasone (Sigma-Aldrich, St. Louis, http://www.sigmaaldrich.com), 50 µM ascorbic-acid-2-phosphate (Sigma), 10 mM ß-glycerophosphate (Sigma), and 10 ng/ml BMP4 (R&D Systems Inc., Minneapolis, http://www.rndsystems.com) [29]. On day 28, Alizarin Red staining was performed as previously described [30]. For chondrogenesis, the cells were resuspended at a density of 8 x 106 cells/ml in {alpha}-MEM supplemented with 10% FCS, 0.1 µM dexamethasone, and 0.17 mM ascorbic-acid 2-phosphate [18]. Ten microliters of this cell suspension was spotted onto a 24-well culture plate and incubated for 30 minutes, after which 1 ml of the medium containing 10 ng/ml transforming growth factor-ß3 (R&D) was added to each well [18]. On day 7, the medium was replaced with a medium containing 10 ng/ml BMP2 (R&D). On day 21, Alcian Blue staining was performed as previously described [30]. For vasculogenesis, 500 cells purified using a cell sorter were recultured onto a confluent OP9 cell layer on each well of a 24-well plate with {alpha}-MEM, 10% FCS, and 50 µM 2ME [8]. Two days later, VECD was immunohistochemically stained by anti-VECD antibody, and the number of endothelial colonies was counted. For hematopoiesis, the cells purified using FACS were recultured on an OP9 cell layer in {alpha}-MEM supplemented with 10% FCS and 10 ng/ml mouse recombinant erythropoietin (R&D). Five days later, floating cells were gently harvested and used to analyze both morphology and gene expression. May-Giemsa staining was carried out for morphological analysis. Benzidine staining for confirming the presence of the erythrocyte lineage was performed as previously described [31].

Real-Time Reverse Transcription-Polymerase Chain Reaction Analysis
Total RNA was isolated using TRIzol reagent (Invitrogen, Carlsbad, CA, http://www.invitrogen.com) according to the protocol recommended by the manufacturer. Residual genomic DNA was digested and removed using DNase I (Roche Diagnostics, Basel, Switzerland, http://www.roche-applied-science.com) treatment. First-strand cDNA was synthesized using the Superscript First-Strand Synthesis System (Invitrogen) for reverse transcription-polymerase chain reaction (RT-PCR). We used a mixture of both oligo d(T)12–18 and random hexamers to generate the first-strand cDNA. Ten microliters of cDNA (total, 100 ng) was mixed with a reaction buffer consisting of 2.5 µl primer mix (2 µM each) and 12.5 µl 2x SYBR Green Reaction Mix (Applied Biosystems, Foster City, CA, http://www.appliedbiosystems.com). Quantitative RT-PCR was performed using the ABI PRISM 7000 system (Applied Biosystems) according to the manufacturer’s instructions. We used GAPDH and Ubiquitin as the invariant controls. After being normalized by division to the value of the control, each value of the specific genes was divided by the maximum value in the experiment for standardization. The primers for the analyses are shown in Supplemental Table S1.

Immunohistochemical Analyses of Cultured Cells
For histochemical analysis, cultured cells were fixed with 2% paraformaldehyde in phosphate-buffered saline (PBS). Antibodies were diluted in 1.5% skim milk and 0.1% Triton X-100 in PBS as follows: rabbit anti-skeletal myosin (Sigma) at 1:500, mouse anti-myogenin (Pharmingen) at 1:500, and rat moAb VECD1 (anti-VECD) at 1:500. horseradish peroxidase (HRP)-conjugated anti-rabbit IgG (Biosource, Camarillo, CA, http://www.biosource.com), HRP-conjugated anti-mouse IgG (Zymed Laboratories, San Francisco, http://www.invitrogen.com), and HRP-conjugated anti-rat IgG (Zymed) were used as secondary antibodies. The substrate used was DAB-Ni for the detection of signals as previously described [25].

DNA Microarray Data Processing
Probe intensity data were taken from Affymetrix CEL files and internally normalized by subtracting the minimum intensity found on the array and dividing by the adjusted median probe intensity. Those data were stored in a Postgresql database (http://www.postgresql.org/) and analyzed and visualized using our own analysis system [32] (eXintegrator, http://www.cdb.riken.jp/scb/documentation/index.html). Individual probe pair differences were normalized across an experimental series by adjusting their means and variances to the median values for that probe set. Expression values were then estimated by taking the mean value of the normalized probe pair differences for a given experimental point. Alternatively, we used the affy suite of the Bioconductor package (http://www.bioconductor.org) to calculate expression values using the robust multi-array average (RMA) method as recommended. As the RMA method produces expression values in log space these values were first transformed to the linear space using ex.

Probe sets with relevant expression patterns were identified as described in the text using the tools in the eXintegrator suite. Similarities to specified profiles were calculated as the mean of Euclidean distances to the set of probe pair profiles in the probe sets. Selections were made either by choosing appropriate thresholds based on the distribution of Euclidean distances or by inspecting the raw data for some number of probe sets ordered by their distances. In addition, the reliability of expression profiles was gauged by the extent of covariation shown by individual probe profiles of given probe sets. This was calculated using the anova score for variation between samples as opposed to variation within samples, and thresholds determined as above for Euclidean distances.

Triangle Plot
The relative expression of genes in the three studied populations were visualized by calculating positions within an equilateral triangle formed from the points of three vectors of unit length radiating from the origin at 120-degree intervals. In this plot, each vector (0:1, –cos(30):–sin(30), cos(30):–sin(30)) represents expression in one of the populations. The three expression measurements (calculated as described above) were first transformed to have a sum of one. If any of the expression measurements was negative, then this value was subtracted from all the measurements for that probe set resulting in a minimum value of 0. The resulting expression measurements were then mapped onto their respective vectors and the vector sums were calculated. This results in a set of points within an equilateral triangle, in which genes expressed in only one sample are located on one of the three points of the triangle, genes expressed in only two samples locate on the edges, and genes expressed equally in all three samples are located in the center of the triangle. The triangle was arbitrarily divided into seven sections representing the seven possible binary combinations, and genes were classified by which area they fell into. Software implementing this mapping is available on request.


   RESULTS
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 Disclosures
 References
 
The Diverging Point to Paraxial and Lateral Mesodermal Cells in ES Cell Differentiation Culture
In our previous study, we proposed that ECs are derived from the VEGFR-2+ population in ES cell culture [7]. However, the relationship between the DP population and the VSP population that are present together was left undetermined. Moreover, the fate of the PSP population, in particular, its potential to give rise to ECs, was not addressed. We thus wanted to complete a map of the differentiation pathway of the VSP and PSP populations in relation to the DP population.

First, we investigated the time course of generation of the DP, VSP, and PSP populations (Fig. 1Go). Differentiation of CCE ES cells was induced according to the method previously described [7] and the proportions of the DP, VSP, and PSP fractions were assessed by FACS on various culture days. Under our culture conditions, the DP and PSP populations appeared almost simultaneously on day 3. The proportion of the DP fraction increased to a peak on day 4, then rapidly decreased over the next 2 days, whereas the VSP and PSP fractions reached their peaks later. The peak of each fraction was reached successively: DP, VSP, and PSP on culture days 4, 5, and 6, respectively. Along with this process, ECD expression progressively decreased.


Figure 1
View larger version (50K):
[in this window]
[in a new window]
  		Figure 1. Mesodermal differentiation in in vitro embryonic stem (ES) cell differentiation. CCE ES cells were cultured on collagen type IV–coated dishes with differentiation medium in the absence of leukemia inhibitory factor. Three to six days after induction, differentiated ES cells were harvested and the expression of E-cadherin (ECD), vascular endothelial growth factor receptor 2 (VEGFR-2), and platelet-derived growth factor receptor {alpha} (PDGFR-{alpha}) were investigated using fluorescence-activated cell sorting. A few ECDlowPDGFR-{alpha}+VEGFR-2– (PDGFR-{alpha} single-positive [PSP]) and ECD–PDGFR-{alpha}+VEGFR-2+ (PDGFR-{alpha} and VEGFR-2 double-positive [DP]) colonies appeared on days 3–3.5. ECD–PDGFR-{alpha}–VEGFR-2+ (VEGFR-2 single-positive [VSP]) colonies appeared one half day later than the DP and PSP populations. The DP fraction increased to a peak on day 4 and then dramatically decreased on day 5. The VSP fraction increased to a peak on day 5. The PSP fraction increased over time at least until day 6. Almost all differentiated ES cells expressed ECD on day 3, and the percentage of ECD+ cells gradually decreased as differentiation progressed.

 
In Vitro Fate Analysis
We next attempted to determine whether the fate of each population was consistent with our hypothesis that the VSP and PSP populations represent lateral and paraxial mesoderm, respectively.

The DP, PSP, and VSP fractions were purified using FACS (Supplemental Fig. S1) and cultured under conditions for inducing myocytes, osteocytes, and chondrocytes, as well as conditions for ECs and hematopoietic cells. In contrast to our expectations, all fractions produced descendants of paraxial mesodermal lineages, such as myocytes, osteocytes, and chondrocytes, when they were cultured under conditions suited for the induction of these lineages (Fig. 2A–DGo). Although not based upon a quantitative assay, we had the impression that more osteocytes and myocytes were present in DP and PSP cultures than in VSP cultures. To confirm this impression, we performed quantitative RT-PCR analyses [33] of lineage markers for myocytes [34], osteocytes, and chondrocytes [35], derived from paraxial mesoderm (Fig. 2E–GGo). In agreement with our impression, the highest expression level of markers for lineages derived from somites was observed in PSP cultures, whereas medium and low expression levels of these markers were detected in DP cultures.


Figure 2
View larger version (32K):
[in this window]
[in a new window]
  		Figure 2. Fate of the embryonic stem (ES) cell–derived mesodermal cells in vitro to paraxial and lateral mesodermal descendants. All three populations—the platelet-derived growth factor receptor {alpha} (PDGFR-{alpha}) single-positive (PSP) population, the vascular endothelial growth factor receptor 2 (VEGFR-2) single-positive (VSP) population, and the PDGFR-{alpha} and VEGFR-2 double-positive (DP) population—showed the ability to differentiate into myocytes, osteocytes, and chondrocytes under specific conditions (A–D). (A, B): Myogenic potential of mesodermal subsets. Myogenin-positive (A) and skeletal myosin–positive (B) cells (dark brown) were derived from all mesodermal populations. (C): Osteocytic differentiation of ES cell–derived mesoderm. Alizarin red–positive calcium matrixes, shown as orange-colored areas, were detected in cultures derived from all mesodermal populations. (D): Chondrogenic potential of the DP, PSP, and VSP populations. To detect the sulfated glycosaminoglycans that are one of the major components of chondrocytes, Alcian Blue staining was performed. Positive chondrocytes expressing a blue color were generated from the three ES cell–derived mesodermal populations. (E–G): The expression profile of myogenesis-related (E), osteogenesis-related (F), and chondrogenesis-related genes (G) in the progeny of ES cell–derived mesoderm. The three populations were cultured under distinct conditions that allow the differentiation of myocytes, osteocytes, or chondrocytes. After differentiation, RNA was purified and the expression levels of the individual specific markers were measured using quantitative reverse transcription-polymerase chain reaction (RT-PCR). (E): Culture cells derived from the PSP fraction showed the highest expression levels of myogenesis-related genes, such as Myf5, MyoD, and myogenin. (F, G): Cultured cells derived from the PSP and DP fractions also expressed osteogenesis-related (Bglap1 and Bglap2) and chondrogenesis-related (col2a1 and col10a1) genes at higher levels than those from the VSP population. (H–K): The DP, PSP, and VSP populations bear the ability to differentiate into vascular endothelial cells and hematopoietic cells. (H): Colony assay of endothelial cells (ECs). Five hundred sorted cells were cultured on a confluent OP9 cell layer for 3 days. EC colonies were visualized using VE-cadherin (VECD) immunostaining. The left panel in each figure shows colonies of ECs (arrow heads). VECD+ colonies were present in all cultures of the three mesodermal fractions. The right panel in each figure displays a high-magnification view of a single EC colony. (I): Number of EC colonies derived from different mesodermal populations. ES cell–derived mesodermal cells were cultured as described in (H). The number of VECD+ colonies was counted in each well of the 24-well plates. (Error bars represent standard deviation.) The frequency of the DP fraction is almost the same as that of the VSP fraction. The frequency of endothelial progenitors in the PSP fraction was one fourth of those of the DP and the VSP fractions. (J): Hematopoietic cell differentiation of each fraction. Sorted cells (2.0 x 105) were cultured on a confluent OP9 cell layer in six-well plates for 5 days with erythropoietin. The morphology of floating cells was shown using Giemsa staining (large panels). (arrow head, erythroblast; arrow, erythrocyte; red arrow head, megakaryocyte.) The presence of erythroid cells was confirmed using Benzidine staining (blue cells in inset panels). (K): Expression level of ßH1 in the culture of each sorted fraction. The expression level of ßH1 was measured by quantitative RT-PCR and normalized by GAPDH expression level. While ßH1 was detected in all fractions, except for the ECD+PDGFR-{alpha}–VEGFR-2– double-negative population, VSP cultures exhibited the highest expression of ßH1. Scale bars: 500 µm (C); 100 µm (A, B, D, H right); 5 mm (H left); 20 µm (J).

 
Similar to the osteochondrogenic and myogenic results, ECs that we presumed to be derived from the lateral mesoderm were induced from all three aforementioned fractions, but not from the ECD+PDGFR-{alpha}–VEGFR-2– double-negative (DN) population (Fig. 2HGo).

To assess the EC potential of each fraction in a more quantitative manner, we measured the frequency of endothelial progenitors in each fraction (Fig. 2IGo). Using this method, the DP and VSP fractions contain nearly the same number of EC progenitors, whereas the PSP fraction contained one fourth fewer EC progenitors than the other two fractions. In order to examine hematopoietic potential, we harvested floating cells for cytological analysis and RT-PCR measurement of the expression level of ßH1 embryonic-type globin (Fig. 2J, KGo). Although mature hematopoietic cells including both erythrocytes and megakaryocytes were observed in the cultures of all fractions, the VSP fraction had a higher potential to generate mature hematopoietic cells than the other fractions. Likewise, expression of ßH1 was detected in all fractions, whereas the VSP fraction exhibited the highest expression level among them.

These results suggest that the VSP and PSP populations represent the lateral and paraxial mesodermal lineages, respectively, and that the differentiation potential of the DP population is inherently different from those of the PSP and VSP populations, though the functional capacity for generating EC colonies is similar to that of the VSP population. However, the fate of each population has not yet been fully determined.

Gene-Expression Profiles During Differentiation of ES Cell–Derived Mesoderm
In order to further characterize the DP, PSP, and VSP populations, in a comprehensive manner, we examined their gene expression using the Affymetrix DNA microarray [32]. In order to visualize the expression of a number of genes in the three populations, we used a simple algorithm that converts the relative expression levels of genes into coordinates within a triangle. To obtain sensible data, it is necessary to select sets of genes that are expressed in at least one of the three cell types, but are not generally expressed. This can be achieved in many different ways, using both arbitrary thresholds for parameters that are related to these properties and semiautomated methods. The genes displayed in Figure 3AGo were selected by searching for genes that are selectively expressed in at least one of the three populations studied, compared to ES cells. The raw data from these lists were then inspected, and probe sets with high-quality expression data (as evidenced by covariation in individual probe pair profiles) were selected by manual inspection. In order to establish that this did not result in a bias in the overall pattern, we also selected probe sets using several different automated methods as well as using a different method of calculating expression values. Although these methods do not select exactly the same set of genes, and the calculated coordinates are not identical, they all essentially display the same overall pattern as the semiautomated method used here (data not shown).


Figure 3
View larger version (25K):
[in this window]
[in a new window]
  		Figure 3. Gene-expression profile of embryonic stem (ES) cell–derived mesoderm. (A): DNA microarray analysis of ES cell–derived mesoderm. The position close to the apex of the triangle indicates expression only in the platelet-derived growth factor receptor {alpha} (PDGFR-{alpha}) and vascular endothelial growth factor receptor 2 (VEGFR-2) double-positive (DP) population (light blue dots; group 1), and the positions close to the left point and to the right point indicate expression only in the PDGFR-{alpha} single-positive (PSP) population (black dots; group 5) and the VEGFR-2 single-positive (VSP) population (green dots; group 7), respectively. The center area of the triangle indicates expression in all three mesodermal populations (red dots; group 3). The left middle trapezoidal area and the right middle trapezoidal area indicate expression in both the DP and PSP populations (light green dots; Group 2) and in both the DP and VSP populations (purple dots; Group 4), respectively. The bottom trapezoidal area indicates expression in both the PSP and VSP populations (blue dots; group 6). Some sample genes located in the PSP-specific area (black dots numbered 1–5) and in the VSP-specific area (green dots numbered 6–10) are shown in Table 1Go. The total number of genes in each area is indicated in brackets. All the gene names are shown in Supplemental Figure S2 and Supplemental Table S2. (B): Gene-expression profile of lineage-specific markers in ES cell–derived mesoderm. Each mesoderm fraction was isolated and purified by fluorescence-activated cell sorting on day 4. The expression levels of lineage-specific markers were investigated using quantitative reverse transcription-polymerase chain reaction (error bars indicate standard deviation; n = 3). Vascular endothelial growth factor receptor 2 (VEGFR-2) and platelet-derived growth factor receptor {alpha} (PDGFR-{alpha}) expression levels in the three mesodermal populations indicate that the purities of these samples were reliable enough for quantitative analyses. Paraxial mesodermal markers, such as Tbx6, Mesp2, and Fst, were highly expressed in the PSP fraction. In contrast, lateral and extraembryonic mesodermal markers, such as GATA2 and Tal1, were selectively expressed in the VSP fraction.

 
In Figure 3AGo, the position close to the apex of the triangle indicates expression specifically in the DP population and the position close to the left point indicates expression exclusively in the PSP population. The expression data reveal that almost all the genes expressed in the DP population are coexpressed in either the VSP or PSP population, although the DP-specific genes, which distribute in the apex, are few. Moreover, the VSP-specific group consists of genes such as VECD, Flt4, GATA2, Tal1, and Ikaros that are known to be expressed in differentiated ECs or hematopoietic cells [36–40], whereas the PSP-specific group contains genes such as Tbx6, Mesp2, Follistatin (Fst), Fgf8, and Dll1, which are well known to be expressed in paraxial mesoderm and somites [41–45] (right and left corners of Fig. 3AGo and Table 1Go). This result suggests that the genes that are expressed exclusively in either the VSP or PSP populations display the prospective fate of the two populations. An almost complete lack of genes that are expressed equally in both the PSP and VSP populations indicates that the VSP and PSP are fully separated.


View this table:
[in this window]
[in a new window]
  		Table 1. Genes indicated with a number in Figure 3AGo

 
To confirm the results of the DNA microarray analysis, we compared them with the results of a quantitative (Q)-PCR analysis (Fig. 3BGo). From microarray data, we selected GATA2, VECD, and Tal1, which are known to be expressed in blood cells and ECs [46, 47], for lineage markers of lateral mesoderm. Likewise, Tbx6, Fst, and Mesp2 were selected as markers for paraxial mesoderm because they are widely accepted to be expressed in the paraxial mesoderm, presomitic mesoderm, and somites [42, 43, 48]. Indeed, the three genes isolated from the microarray data of the VSP fraction are selectively expressed in the VSP fraction using Q-PCR analysis (Fig. 3BGo). On the other hand, Tbx6, Fst, and Mesp2 were also detectable in the DP and VSP fractions, at lower levels than in the PSP fraction (Fig. 3BGo). This is not a result of contamination of unwanted fractions in the FACS-purified samples because the expression levels of PDGFR-{alpha} and VEGFR-2 mRNA, respectively, in the FACS-sorted VSP and PSP fractions were negligible (Fig. 3BGo). Actually, Tbx6 and Fst are also known to be expressed in nascent mesoderm in the primitive streak during mouse embryogenesis [41, 43], and a weak expression of Mesp2 has been shown for most mesodermal cells during early embryogenesis [42]. We consider that this discrepancy is a result of the difference in detection sensitivity between the DNA microarray and Q-PCR. All these results indicate that: (a) the microarray data are consistent with the Q-PCR data and (b) the lateral and paraxial mesodermal markers are specifically expressed in the VSP and PSP populations, respectively.

Evidence for Interlineage Conversion as the Underlying Mechanism for Phenotype Switching Between VSP and PSP
Gene-expression profiling suggests that the PSP and VSP fractions represent distinctive populations corresponding to the paraxial and lateral mesoderm, respectively. However, our in vitro fate analysis suggests that their fates are not fully restricted. Given that both the VSP and PSP fractions generated in day 4 ES cell culture maintain some flexibility in their fate, it is expected that they first undergo a switch in surface phenotype from VSP to PSP, or vice versa, before completing an irreversible fate determination.

Because an OP9 cell layer can support clonogenic proliferation of single EC progenitors [8], we used this to investigate whether switching of surface phenotype between VSP and PSP can occur. Because this experimental setting requires a method for distinguishing inoculated cells from OP9 cells, we prepared the ES cell lines harboring eGFP cDNA driven by chicken ß-actin promoter (GFP-ES).

GFP-ES cells were cultured for 4 days to induce mesoderm cells, and the DP, PSP, and VSP fractions were purified and recultured with OP9 stromal cells. After additional incubation for 24 hours, the cells were harvested and the surface expression of VEGFR-2 and PDGFR-{alpha} in the GFP+ population was analyzed (Fig. 4AGo). Under this condition, the DP culture produced equal numbers of VSP and PSP cells, confirming that the PSP and VSP populations diverge from the DP population. Interestingly, a high level of phenotype switching in both directions was observed between the VSP and PSP cultures within 24 hours (Fig. 4AGo). The purified PSP culture at day 4 produced 44% PSP cells and 20% VSP cells, and cells the VSP culture differentiated into 12% PSP cells and 51% VSP cells. It should be noted that DP cells could not be detected in cultures of any population, including the DP culture itself. Thus, the phenotypic switch between VSP and PSP occurs directly rather than via an immature DP stage. The same switch between VSP and PSP was also observed in the cultured TT2 ES cell line (data not shown).


Figure 4
View larger version (33K):
[in this window]
[in a new window]
  		Figure 4. Lineage conversion of embryonic stem (ES) cell–derived mesodermal populations. A: Surface marker analyses of purified ES cell–derived mesodermal populations after additional incubation for 24 hours on a confluent OP9 cell layer. The platelet-derived growth factor receptor {alpha} (PDGFR-{alpha}) and vascular endothelial growth factor receptor 2 (VEGFR-2) double-positive (DP), PDGFR-{alpha} single-positive (PSP), and VEGFR-2 single-positive (VSP) populations were derived from green fluorescent protein (GFP)-ES cells to allow them to be distinguished from OP9 feeder cells using GFP expression. All three mesodermal populations retained the ability to differentiate into both PSP and VSP cells. (B): The gene-expression patterns of the PSP and the VSP populations derived from day 4 VSP and PSP cultures, respectively, were similar to those of the day 4 PSP and VSP populations. RNA was isolated from day 4 ES cell–derived mesoderm and mesodermal cells were recultured on an OP9 cell layer for one day. Gene-expression levels of specific markers were measured using quantitative reverse transcription-polymerase chain reaction. The lateral and extraembryonic mesodermal markers, such as GATA2 and Tal1, were expressed in both the day 4 VSP population and the day 4 PSP culture–derived VSP population. Similarly, the day 4 PSP population and the PSP population derived from the day 4 VSP population showed dominant expression of paraxial markers, such as Tbx6 and Mesp2. (C): Vasculogenic potentials of cells generated in secondary cultures. Day 4 mesodermal fractions were recultured on an OP9 cell layer, and the recultured PSP, VSP, and DP fractions were purified again using fluorescence-activated cell sorting (FACS). One thousand cells of each population were cultured on an OP9 cell layer in one well of a 24-well plate to induce EC differentiation. The mean and standard deviation of endothelial colony number visualized by VE-cadherin (VECD) staining are shown in (C). VECD+ colonies were observed in the VSP population derived from every mesodermal subset, but they were not observed in the other two populations. (D): Hematopoietic potentials of cells generated in secondary cultures. Day 4 mesodermal fractions were recultured on an OP9 cell layer, and the PSP and VSP fractions were isolated using FACS. Cells (2.0 x 105) from each population were cultured on an OP9 cell layer in one well of six-well plate to induce hematopoietic differentiation. Four days after induction, floating cells were harvested and counted (upper panel). The same cells were subjected to quantitative RT-PCR analysis for ßH1. Expression was observed exclusively in the VSP fraction, irrespective of its origin. (E): Osteogenic potential of lineage-converted populations. Each lineage-converted population was recultured under conditions favoring osteocytic differentiation for 28 days. Progeny of the PSP cells derived from the DP population and the VSP population had widely spread calcium matrixes (red area in upper left and lower right panels, Alizarin red staining). In contrast, little calcium deposition was observed in the cells differentiated from the VSP population. Scale bar = 500 µm.

 
In order to determine whether the switch in surface phenotype reflects a simple fluctuation in the expression of VEGFR-2 and PDGFR-{alpha} or a more global process involving other molecules, we analyzed the expression of the six aforementioned markers (Fig. 3BGo) that are expressed in either paraxial or lateral mesoderm cells (Fig. 4BGo). PSP and VSP cells were purified and recultured on an OP9 stromal cell layer. After additional incubation for 24 hours, newly generated PSP and VSP cells were purified again from the PSP and VSP cultures, respectively. The expression levels of the six genes in each population were assessed by quantitative RT-PCR. During the phenotype switch between VSP and PSP, the expression profile of these marker genes changed completely from one type to the other (Fig. 4BGo). Thus, the phenotype switch between VSP and PSP is not the result of a simple fluctuation in the expression levels of VEGFR-2 and PDGFR-{alpha}.

In order to further determine whether the phenotype switch between VSP and PSP is accompanied by an actual switch in fate, we evaluated the in vitro differentiation potential of each fraction in the secondary culture. DP, VSP, and PSP cells were prepared from ES cell differentiation cultures at day 4 and incubated on an OP9 feeder layer for 24 hours. Each PSP and VSP fraction was separately purified from the secondary cultures and cultured again on either an OP9 feeder layer, to measure the frequency of endothelial progenitors and the expression level of embryonic-type globin, or under osteogenesis-inducing culture conditions. Regardless of the initial phenotype of the cells in the secondary culture, only the VSP fraction could produce ECs (Fig. 4CGo). In order to confirm that the VSP cells derived from the PSP cultures still maintained the characteristics of lateral mesoderm, we assessed their erythropoietic potential. As shown in Figure 4DGo, erythropoiesis was observed only in the VSP culture, irrespective of its origin. Similarly, osteogenesis was only induced in the PSP culture, regardless of the phenotype of the initial population (Fig. 4EGo).

Time- and Stage-Dependent Restriction of Interlineage Conversion Capability
In order to determine how long the capability of undergoing a phenotype switch is maintained in each fraction, we examined phenotypic changes in VSP and PSP cultures until day 6 (Fig. 5AGo). The incidence of phenotype switching from PSP to VSP decreased remarkably with a longer incubation time. On day 6, the potential for switching from PSP to VSP was almost completely lost. On the other hand, the potential for VSP cultures to give rise to PSP cells was preserved for longer.


Figure 5
View larger version (26K):
[in this window]
[in a new window]
  		Figure 5. Commitment point from embryonic stem (ES) cell–derived mesoderm to specific lineage progenitors. A: Differentiation into platelet-derived growth factor receptor {alpha} (PDGFR-{alpha}) single-positive (PSP) cells and vascular endothelial growth factor receptor 2 (VEGFR-2) single-positive (VSP) cells at various time points of in vitro ES cell culture. The PDGFR-{alpha} and VEGFR-2 double-positive (DP), PSP, and VSP fractions were purified from differentiated green fluorescent protein (GFP)-ES cells on day 3.5, day 5, and day 6 of culture and were recultured on a confluent OP9 cell layer. One day later, the expression of surface markers in the GFP+ population was analyzed by fluorescence-activated cell sorting (FACS). Although the DP fraction decreased dramatically from day 4 to day 6 (Fig. 1Go), the DP population retained its ability to give rise to both PSP and VSP cells even on day 6 (left column). The VSP population also retained its ability to differentiate into PSP cells through the time course of the experiment (right column). In contrast to these two populations, the PSP population showed a dramatic reduction in conversion potency through the progression of differentiation, and cells in this population could no longer convert to the VSP phenotype on day 6 (center column). (B): Lack of secondary switching capability in the population generated by lineage switching. Day 4 mesodermal fractions were recultured on an OP9 cell layer, and the lineage-switched populations—the VSP fraction derived from PSP cultures and the PSP fraction derived from VSP cultures—were purified again using FACS. Each switched population was recultured again on an OP9 cell layer for 24 hours. After secondary culture, the expression levels of PDGFR-{alpha} and VEGFR-2 were analyzed. The VSP cells derived from the PSP fraction could no longer convert to the PSP phenotype (i). Similarly, the PSP cells derived from the VSP fraction could not convert to the VSP phenotype (ii). (C): VE-cadherin (VECD) expression in the VSP population implies a commitment to the endothelial lineage. i: Expression pattern of VECD in the VSP population on day 4. The VSP population contains both VECD+(10%) and VECD– (90%) cells. Each subpopulation was purified and recultured on an OP9 cell layer for 24 hours. ii: Expression of PDGFR-{alpha} and VEGFR-2 in GFP+ cells after reculture. The VECD+ subpopulation had lost the potential for lineage conversion, whereas the VECD– VSP fraction could give rise to PSP cells.

 
We also compared the potential to undergo a phenotype switch in VSP and PSP cells derived directly from ES cells with those generated by the phenotype switches. As shown in Figure 5BGo, this switch potential was detected only in primary VSP and PSP cultures derived directly from ES cells and not in those generated in the secondary culture.

As some of the VSP cells differentiate into VECD+ ECs during culture, we investigated whether or not the potential to switch the surface phenotype to PSP is maintained after completing EC differentiation (Fig. 5CGo). The VSP population was divided into VECD– VSP and VECD+ VSP fractions. The VECD+ VSP population could no longer produce PSP cells, whereas this potential was maintained in the VECD– population (Fig. 5CGo). These results suggest that the potential for phenotypic switching depends on the time and the stage of cell culture.


   DISCUSSION
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
Disclosures
 References
 
The aim of this study was to dissect the early process of divergence of the EC differentiation pathway away from other mesodermal lineages. Combining PDGFR-{alpha}, as a marker for the paraxial mesodermal lineage, with VEGFR-2, as a marker for the lateral mesodermal lineage, we have found multiple EC differentiation pathways in ES cell development.

An in vitro ES cell differentiation system is advantageous to analyze the process of cell specification during early embryogenesis, particularly when obtaining enough cells from the actual embryo is difficult. Conversely, the lack of positional information that is available for specifying cell types in the embryo is a disadvantage of ES cell culture. Consequently, for analyzing the divergence point between lateral mesoderm and paraxial mesoderm in vitro, it is necessary to redefine the two lineages according to their individual features. Because our previous study in gastrulating embryos demonstrated that VEGFR-2 and PDGFR-{alpha} are expressed in the lateral and extraembryonic mesoderm and the paraxial mesoderm, respectively [13], we speculated that surface expression of PDGFR-{alpha} and VEGFR-2 could define these mesodermal lineages in ES cell culture. Differentiated ES cells that expressed VEGFR-2 and PDGFR-{alpha} were divided into three populations—DP, VSP, and PSP. In the present study, we characterized these three populations in a comprehensive way using a DNA microarray. To track the dynamic change in gene-expression profiles during diversification, we developed a new method that can display the genes according to their expression levels in three populations. Compared with the most popular method, which simply classifies genes into groups, in our method, each dot in Figure 3AGo corresponds to a value reflecting the relative specificity of the expression of a given gene among the three populations. Using this new method, the current study demonstrates two important findings. First, we found two groups of genes that are expressed exclusively by either the VSP or PSP population. The VSP-specific group consists of genes that are known to be expressed in differentiated ECs or blood cells, whereas the PSP-specific group contains genes that are known to be expressed in somites. Of note is that the expression of only a few genes overlaps in these two populations, though many more genes are found to be expressed in all three populations. Thus, the VSP and PSP populations are likely to represent fully segregated populations. Second, many genes expressed in the DP population are coexpressed by either the VSP or PSP population. In contrast, DP-specific genes are few. This distribution pattern of genes fits well into a model of lineage divergence in which a common progenitor population, that is, the DP population, separates into two more mature stages, PSP and VSP.

Based on the results of the DNA microarray analysis, we hypothesized that the DP population is located at the divergence point from which PSP paraxial and VSP lateral mesoderm are generated. The results of in vitro fate analyses demonstrated that DP cells have the potential to differentiate into the lateral mesodermal lineage (ECs) and the paraxial mesodermal lineage (myocytes, osteocytes, and chondrocytes). DP cells can directly give rise to both VSP and PSP cells after short-term reculture on an OP9 feeder layer. Taken together, our results lead us to concluded that the VSP and PSP populations diverge from the common DP progenitor population, as illustrated in Figure 6Go.


Figure 6
View larger version (39K):
[in this window]
[in a new window]
  		Figure 6. Differentiation process of mesoderm in in vitro an embryonic stem (ES) cell differentiation system. The analyses of differentiated ES cells revealed three types of mesodermal cells—an E-cadherin (ECD)–/platelet-derived growth factor receptor {alpha} (PDGFR-{alpha})+/vascular endothelial growth factor receptor 2 (VEGFR-2)+ population (the PDGFR-{alpha} and VEGFR-2 double-positive [DP] population, green), an ECDlowPDGFR-{alpha}+VEGFR-2– population (the PDGFR-{alpha} single-positive population [PSP] population, dark blue), and an ECD–PDGFR-{alpha}–VEGFR-2+ population (the VEGFR-2 single-positive [VSP] population, yellow). The DP population is the most immature and can give rise to both PSP and VSP cells. Although both the VSP and the PSP populations exhibit the specific properties of paraxial and lateral mesoderm, respectively, these populations can be converted into each other at an early stage. This conversion is dependent on the time course of the culture and the stage of the mesodermal cells. The VE-cadherin–positive subpopulation, which is illustrated as though forming a drop that is derived from, but discontinuous with, the VSP population, has lost this potential and has irreversibly committed to the endothelial lineage.

 
Our results also revealed that the VSP and PSP populations correspond to the lateral and paraxial mesodermal lineages, respectively. In vitro fate analyses revealed that the VSP population has a greater capacity to produce ECs and hematopoietic cells than does the PSP population. Likewise, the PSP population has higher myogenic, osteogenic, and chondrogenic potentials than does the VSP population. The genes expressed exclusively by the VSP population contain markers for ECs and hematopoietic cells. Similarly, the genes that are specifically expressed by the PSP population contain the genes expressed in embryonic somites and paraxial mesoderm.

Though the expression of VEGFR-2 and PDGFR-{alpha} can define the lateral and paraxial mesodermal lineages, respectively, in ES cell culture, we also demonstrated that VSP and PSP cells can give rise to progeny of the other lineage. This unexpected result is due to lineage switching between VSP and PSP cells. Our result shows that ECs are generated from not only DP and VSP cells but also from PSP cells. It is remarkable that the frequency of endothelial progenitors in the PSP fraction harvested from day 4 ES cell differentiation cultures is as high as one fourth of those in the VSP fraction from the same culture. Similarly, long-term in vitro fate analyses have revealed that myocytes, osteocytes, and chondrocytes, which are progeny of paraxial mesoderm, are generated from VSP cells. With respect to the potential of the PSP population to give rise to ECs, it has already been reported that some committed somitic cells can give rise to ECs [49, 50]. However, what we describe here is the transition of PSP cells to multipotent VSP cells that can give rise to not only ECs but also to hematopoietic cells. Considering that somitic cells cannot give rise to hematopoietic cells, it is likely that VSP cells derived from PSP cells correspond to lateral mesoderm.

It is generally considered that the decision of fate is such an ambiguous process that the selected fate is reversible for a while after lineage diversion [51, 52]. Even so, it is remarkable that a phenotype switch of such magnitude—20% of the cells generated in the PSP culture were VSP cells and 10% of the cells generated in the VSP culture were PSP cells—occurs within 24 hours. Because our culture conditions allow phenotype switch in both directions between the VSP and PSP populations, it is unlikely that this is a result of selectivity of the culture conditions to expand a particular population. Thus, a considerable proportion of nascent VSP and PSP cells maintain flexibility in their fate.

What then underlies this phenotype switch between PSP and VSP? The first possibility is that the switch is restricted to the expression of VEGFR-2 and PDGFR-{alpha}, while the actual fate remains unaffected. Our result suggests that this is unlikely. VSP populations derived from PSP cultures contain nearly the same number of EC progenitors as those derived from VSP cultures, or DP populations derived directly from ES cells. Likewise, PSP populations derived from DP and VSP cultures do give rise to progeny of the paraxial mesoderm, such as osteocytes. Therefore, the switch in surface phenotype between PSP and VSP is associated with a switch in prospective fate. In accordance with the fate analysis, along with the switch from PSP to VSP, the expression pattern of six markers for paraxial and lateral mesoderm shifted from one type to the other. These results suggest that phenotype switching between PSP and VSP indeed represents a lineage conversion between lateral and paraxial mesodermal cells. Our results also suggest that this lineage conversion is not a result of dedifferentiation, because the DP population, which is located at the divergence point of the PSP and VSP populations in the normal differentiation course, was hardly detected during the process of phenotype switch in culture. Thus, what is being observed here is a direct lineage conversion between VSP and PSP. Our findings suggest that in the actual embryo, this lineage conversion may occur spontaneously after commitment of the lateral and paraxial mesoderm. Further investigation is necessary to resolve this question.

We have demonstrated that the ability to undergo lineage conversion is restricted to an early stage of mesodermal cell differentiation. The capacity to undergo lineage conversion is quickly lost over time, as the PSP fraction harvested from day 6 ES cell differentiation cultures is irreversibly committed to a paraxial mesoderm lineage. The ability to undergo a phenotype switch is maintained longer in the VSP population, but it is completely lost when VECD is expressed on the cell surface upon differentiation to ECs. These results suggest that the lineage conversion observed in this study is a phenomenon inherent to early mesodermal cells and may not be relevant to the plasticity of fully differentiated cells.

Taking our new results into account, we corrected our previous model of EC differentiation to the one illustrated in Figure 6Go. In this scheme, the DP population is placed at the divergence point of the VSP and PSP populations. While VSP and PSP cells represent fully diverged populations in terms of their gene-expression profiles, some of them maintain the ability to undergo interlineage conversion for a short period of time after divergence to VSP and PSP cells. Concerning EC differentiation pathways, ECs are generated from not only the DP population but also from the PSP population via the VSP population, and surface expression of VECD coincides with the timing of irreversible commitment.


   DISCLOSURES
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
References
 
The authors indicate no potential conflicts of interest.


   ACKNOWLEDGMENTS
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
 References
 
We thank Dr. K. Nakao for a kind gift of the TT2 ES cells. We also thank Dr. H. Yoshida and Dr. H. Kurata for technical support. This work was supported by grants from the Ministry of Education and Science of Japan (No. 12219209 to N.S. and No. 16606005 and No. 17045039 to E.T.; and the Project for Realization of Regenerative Medicine to N.S.).


   REFERENCES
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
 References
 

  1. Jain RK. Molecular regulation of vessel maturation. Nat Med 2003;9: 685–693.[CrossRef][Medline]

  2. Cines DB, Pollak ES, Buck CA et al. Endothelial cells in physiology and in the pathophysiology of vascular disorders. Blood 1998;91:3527–3561.[Free Full Text]

  3. Shalaby F, Ho J, Stanford WL et al. A requirement for Flk1 in primitive and definitive hematopoiesis and vasculogenesis. Cell 1997;89:981–990.[CrossRef][Medline]

  4. Le Grand F, Auda-Boucher G, Levitsky D et al. Endothelial cells within embryonic skeletal muscles: A potential source of myogenic progenitors. Exp Cell Res 2004;301:232–241.[CrossRef][Medline]

  5. Bailey AS, Jiang S, Afentoulis M et al. Transplanted adult hematopoietic stems cells differentiate into functional endothelial cells. Blood 2004; 103:13–19.[Abstract/Free Full Text]

  6. Ambler CA, Nowicki JL, Burke AC et al. Assembly of trunk and limb blood vessels involves extensive migration and vasculogenesis of somite-derived angioblasts. Dev Biol 2001;234:352–364.[CrossRef][Medline]

  7. Nishikawa SI, Nishikawa S, Hirashima M et al. Progressive lineage analysis by cell sorting and culture identifies FLK1 + VE-cadherin + cells at a diverging point of endothelial and hemopoietic lineages. Development 1998;125:1747–1757.[Abstract]

  8. Hirashima M, Kataoka H, Nishikawa S et al. Maturation of embryonic stem cells into endothelial cells in an in vitro model of vasculogenesis. Blood 1999;93:1253–1263.[Abstract/Free Full Text]

  9. Yamashita J, Itoh H, Hirashima M et al. Flk1-positive cells derived from embryonic stem cells serve as vascular progenitors. Nature 2000;408: 92–96.[CrossRef][Medline]

  10. Era T, Witte ON. Regulated expression of P210 Bcr-Abl during embryonic stem cell differentiation stimulates multipotential progenitor expansion and myeloid cell fate. Proc Natl Acad Sci U S A 2000;97:1737–1742.[Abstract/Free Full Text]

  11. Lee SH, Lumelsky N, Studer L et al. Efficient generation of midbrain and hindbrain neurons from mouse embryonic stem cells. Nat Biotechnol 2000;18:675–679.[CrossRef][Medline]

  12. Dumont DJ, Fong GH, Puri MC et al. Vascularization of the mouse embryo: A study of flk-1, tek, tie, and vascular endothelial growth factor expression during development. Dev Dyn 1995;203:80–92.[Medline]

  13. Kataoka H, Takakura N, Nishikawa S et al. Expressions of PDGF receptor alpha, c-Kit and Flk1 genes clustering in mouse chromosome 5 define distinct subsets of nascent mesodermal cells. Dev Growth Differ 1997;39:729–740.[CrossRef][Medline]

  14. Burdsal CA, Damsky CH, Pedersen RA. The role of E-cadherin and integrins in mesoderm differentiation and migration at the mammalian primitive streak. Development 1993;118:829–844.[Abstract]

  15. Wang L, Li L, Shojaei F et al. Endothelial and hematopoietic cell fate of human embryonic stem cells originates from primitive endothelium with hemangioblastic properties. Immunity 2004;21:31–41.[CrossRef][Medline]

  16. Orr-Urtreger A, Bedford MT, Do MS et al. Developmental expression of the alpha receptor for platelet-derived growth factor, which is deleted in the embryonic lethal Patch mutation. Development 1992;115:289–303.[Abstract]

  17. Takakura N, Yoshida H, Ogura Y et al. PDGFR alpha expression during mouse embryogenesis: Immunolocalization analyzed by whole-mount immunohistostaining using the monoclonal anti-mouse PDGFR alpha antibody APA5. J Histochem Cytochem 1997;45:883–893.[Abstract/Free Full Text]

  18. Nakayama N, Duryea D, Manoukian R et al. Macroscopic cartilage formation with embryonic stem-cell-derived mesodermal progenitor cells. J Cell Sci 2003;116:2015–2028.[Abstract/Free Full Text]

  19. Motoike T, Markham DW, Rossant J et al. Evidence for novel fate of Flk1+ progenitor: Contribution to muscle lineage. Genesis 2003;35: 153–159.[CrossRef][Medline]

  20. Zhang G, Gurtu V, Kain SR. An enhanced green fluorescent protein allows sensitive detection of gene transfer in mammalian cells. Biochem Biophys Res Commun 1996;227:707–711.[CrossRef][Medline]

  21. Niwa H, Yamamura K, Miyazaki J. Efficient selection for high-expression transfectants with a novel eukaryotic vector. Gene 1991;108:193–199.[CrossRef][Medline]

  22. Era T, Takagi T, Takahashi T et al. Characterization of hematopoietic lineage-specific gene expression by ES cell in vitro differentiation induction system. Blood 2000;95:870–878.[Abstract/Free Full Text]

  23. Nakano T, Kodama H, Honjo T. Generation of lymphohematopoietic cells from embryonic stem cells in culture. Science 1994;265:1098–1101.[Abstract/Free Full Text]

  24. Yagi T, Tokunaga T, Furuta Y et al. A novel ES cell line, TT2, with high germline-differentiating potency. Anal Biochem 1993;214:70–76.[CrossRef][Medline]

  25. Takakura N, Yoshida H, Kunisada T et al. Involvement of platelet-derived growth factor receptor-alpha in hair canal formation. J Invest Dermatol 1996;107:770–777.[CrossRef][Medline]

  26. Shirayoshi Y, Nose A, Iwasaki K et al. N-linked oligosaccharides are not involved in the function of a cell-cell binding glycoprotein E-cadherin. Cell Struct Funct 1986;11:245–252.[Medline]

  27. Matsuyoshi N, Toda K, Horiguchi Y et al. In vivo evidence of the critical role of cadherin-5 in murine vascular integrity. Proc Assoc Am Physicians 1997;109:362–371.[Medline]

  28. Ewton DZ, Florini JR. Effects of the somatomedins and insulin on myoblast differentiation in vitro. Dev Biol 1981;86:31–39.[CrossRef][Medline]

  29. Herbertson A, Aubin JE. Dexamethasone alters the subpopulation make-up of rat bone marrow stromal cell cultures. J Bone Miner Res 1995;10:285–294.[Medline]

  30. Muraglia A, Corsi A, Riminucci M et al. Formation of a chondro-osseous rudiment in micromass cultures of human bone-marrow stromal cells. J Cell Sci 2003;116:2949–2955.[Abstract/Free Full Text]

  31. Cooper MC, Levy J, Cantor LN et al. The effect of erythropoietin on colonial growth of erythroid precursor cells in vitro. Proc Natl Acad Sci U S A 1974;71:1677–1680.[Abstract/Free Full Text]

  32. Jakt LM, Okada M, Nishikawa SI. An open source client-server system for the analysis of Affymetrix microarray data. Genome Inform 2003; 14:267–277.

  33. Heid CA, Stevens J, Livak KJ et al. Real time quantitative PCR. Genome Res 1996;6:986–994.[Abstract/Free Full Text]

  34. Parker MH, Seale P, Rudnicki MA. Looking back to the embryo: Defining transcriptional networks in adult myogenesis. Nat Rev Genet 2003;4:497–507.[Medline]

  35. Aubin JE, Liu F, Malaval L et al. Osteoblast and chondroblast differentiation. Bone 1995;17:77S–83S.[Medline]

  36. Breier G, Breviario F, Caveda L et al. Molecular cloning and expression of murine vascular endothelial-cadherin in early stage development of cardiovascular system. Blood 1996;87:630–641.[Abstract/Free Full Text]

  37. Kallianpur AR, Jordan JE, Brandt SJ. The SCL/TAL-1 gene is expressed in progenitors of both the hematopoietic and vascular systems during embryogenesis. Blood 1994;83:1200–1208.[Abstract/Free Full Text]

  38. Silver L, Palis J. Initiation of murine embryonic erythropoiesis: A spatial analysis. Blood 1997;89:1154–1164.[Abstract/Free Full Text]

  39. Dumont DJ, Jussila L, Taipale J et al. Cardiovascular failure in mouse embryos deficient in VEGF receptor-3. Science 1998;282:946–949.[Abstract/Free Full Text]

  40. Georgopoulos K, Moore DD, Derfler B. Ikaros, an early lymphoid-specific transcription factor and a putative mediator for T cell commitment. Science 1992;258:808–812.[Abstract/Free Full Text]

  41. Chapman DL, Agulnik I, Hancock S et al. Tbx6, a mouse T-Box gene implicated in paraxial mesoderm formation at gastrulation. Dev Biol 1996;180:534–542.[CrossRef][Medline]

  42. Saga Y, Hata N, Koseki H et al. Mesp2: A novel mouse gene expressed in the presegmented mesoderm and essential for segmentation initiation. Genes Dev 1997;11:1827–1839.[Abstract/Free Full Text]

  43. Albano RM, Arkell R, Beddington RS et al. Expression of inhibin subunits and follistatin during postimplantation mouse development: Decidual expression of activin and expression of follistatin in primitive streak, somites and hindbrain. Development 1994;120:803–813.[Abstract]

  44. Crossley PH, Martin GR. The mouse Fgf8 gene encodes a family of polypeptides and is expressed in regions that direct outgrowth and patterning in the developing embryo. Development 1995;121:439–451.[Abstract]

  45. Bettenhausen B, Hrabe de Angelis M, Simon D et al. Transient and restricted expression during mouse embryogenesis of Dll1, a murine gene closely related to Drosophila Delta. Development 1995;121:2407–2418.[Abstract]

  46. Orkin SH. Diversification of haematopoietic stem cells to specific lineages. Nat Rev Genet 2000;1:57–64.[CrossRef][Medline]

  47. Risau W, Flamme I. Vasculogenesis. Annu Rev Cell Dev Biol 1995;11: 73–91.[CrossRef][Medline]

  48. Chapman DL, Papaioannou VE. Three neural tubes in mouse embryos with mutations in the T-box gene Tbx6. Nature 1998;391:695–697.[CrossRef][Medline]

  49. Pardanaud L, Dieterlen-Lievre F. Manipulation of the angiopoietic/hemangiopoietic commitment in the avian embryo. Development 1999; 126:617–627.[Abstract]

  50. Pardanaud L, Luton D, Prigent M et al. Two distinct endothelial lineages in ontogeny, one of them related to hemopoiesis. Development 1996; 122:1363–1371.[Abstract]

  51. Miyamoto T, Iwasaki H, Reizis B et al. Myeloid or lymphoid promiscuity as a critical step in hematopoietic lineage commitment. Dev Cell 2002;3:137–147.[CrossRef][Medline]

  52. Prohaska SS, Scherer DC, Weissman IL et al. Developmental plasticity of lymphoid progenitors. Semin Immunol 2002;14:377–384.[CrossRef][Medline]




This article has been cited by other articles:


Home page
 Stem CellsHome page
H. Sakurai, Y. Okawa, Y. Inami, N. Nishio, and K.-i. Isobe
Paraxial Mesodermal Progenitors Derived from Mouse Embryonic Stem Cells Contribute to Muscle Regeneration via Differentiation into Muscle Satellite Cells
Stem Cells, July 1, 2008; 26(7): 1865 - 1873.
[Abstract] [Full Text] [PDF]

Home page
 Stem CellsHome page
T. Era, N. Izumi, M. Hayashi, S. Tada, S. Nishikawa, and S.-I. Nishikawa
Multiple Mesoderm Subsets Give Rise to Endothelial Cells, Whereas Hematopoietic Cells Are Differentiated Only from a Restricted Subset in Embryonic Stem Cell Differentiation Culture
Stem Cells, February 1, 2008; 26(2): 401 - 411.
[Abstract] [Full Text] [PDF]

Home page
 Stem CellsHome page
N. Izumi, T. Era, H. Akimaru, M. Yasunaga, and S.-I. Nishikawa
Dissecting the Molecular Hierarchy for Mesendoderm Differentiation Through a Combination of Embryonic Stem Cell Culture and RNA Interference
Stem Cells, July 1, 2007; 25(7): 1664 - 1674.
[Abstract] [Full Text] [PDF]

Home page
 J. Cell Sci.Home page
M. Takenaga, M. Fukumoto, and Y. Hori
Regulated Nodal signaling promotes differentiation of the definitive endoderm and mesoderm from ES cells
J. Cell Sci., June 15, 2007; 120(12): 2078 - 2090.
[Abstract] [Full Text] [PDF]

Home page
 Stem CellsHome page
K. Schenke-Layland, E. Angelis, K. E. Rhodes, S. Heydarkhan-Hagvall, H. K. Mikkola, and W. R. MacLellan
Collagen IV Induces Trophoectoderm Differentiation of Mouse Embryonic Stem Cells
Stem Cells, June 1, 2007; 25(6): 1529 - 1538.
[Abstract] [Full Text] [PDF]

Home page
 BloodHome page
F. Anjos-Afonso and D. Bonnet
Nonhematopoietic/endothelial SSEA-1+ cells define the most primitive progenitors in the adult murine bone marrow mesenchymal compartment
Blood, February 1, 2007; 109(3): 1298 - 1306.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Data
Right arrow All Versions of this Article:
2005-0256v1
2005-0256v2
24/3/575    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Reprints/Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Sakurai, H.
Right arrow Articles by Nishikawa, S.-I.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Sakurai, H.
Right arrow Articles by Nishikawa, S.-I.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
STEM CELLS THE ONCOLOGIST CME ALPHAMED PRESS JOURNALS