Activation of Akt signaling is sufficient to maintain pluripotency ... - Nature

Oncogene (2006) 25, 2697–2707

& 2006 Nature Publishing Group All rights reserved 0950-9232/06 $30.00 www.nature.com/onc

ORIGINAL ARTICLE

Activation of Akt signaling is sufficient to maintain pluripotency in mouse and primate embryonic stem cells S Watanabe1, H Umehara1, K Murayama2, M Okabe3, T Kimura2 and T Nakano1,2 1 Graduate School of Frontier Biosciences, Osaka University, Suita, Osaka, Japan; 2Department of Pathology, Graduate School of Medicine, Osaka University, Suita, Osaka, Japan; and 3Genome Information Research Center, Osaka University, Suita, Osaka, Japan

Embryonic stem (ES) cells can self-renew indefinitely without losing their differentiation ability to any cell types. Phosphoinositide-3 kinase (PI3K)/Akt signaling plays a pivotal role in various stem cell systems, including the formation of embryonic germ (EG) cells from primordial germ cells and self-renewal of neural stem cells. Here, we show that myristoylated, active form of Akt (myr-Akt) maintained the undifferentiated phenotypes in mouse ES cells without the addition of leukemia inhibitory factor (LIF). The effects of myr-Akt were reversible, because LIF dependence and pluripotent differentiation activity were restored by the deletion of myr-Akt. In addition, myr-Akt-Mer fusion protein, whose enzymatic activity is controlled by 4-hydroxy-tamoxifen, also maintained the pluripotency of not only mouse but also cynomolgus monkey ES cells. These results clearly demonstrate that Akt signaling sufficiently maintains pluripotency in mouse and primate ES cells, and support the notion that PI3K/Akt signaling axis regulates ‘stemness’ in a broad spectrum of stem cell systems. Oncogene (2006) 25, 2697–2707. doi:10.1038/sj.onc.1209307; published online 9 January 2006 Keywords: Akt; ES cells; pluripotency; stem cell system

Introduction Embryonic stem (ES) cells are cells derived from the inner cell mass of blastocysts and sustain pluripotency, the capacity to differentiate into all three germ layers and germ lineage (Chambers and Smith, 2004). Pluripotency of mouse and primate ES cells is regulated by distinct signaling pathways. Mouse ES cell pluripotency can be maintained by leukemia inhibitory factor (LIF), which activates the Janus kinase-signal transducer and activator of transcription-3 (JAK/STAT3) signaling Correspondence: Dr T Kimura or Professor T Nakano, Department of Pathology, Graduate School of Medicine, Osaka University, 2-2 Yamada-oka, Suita, Osaka 565-0871, Japan. E-mail: [email protected] or E-mail: [email protected] Received 26 July 2005; revised 2 November 2005; accepted 4 November 2005; published online 9 January 2006

pathway. Withdrawal of LIF and inhibition of JAK/ STAT3 signaling rapidly induce differentiation in mouse ES cells (Boeuf et al., 1997; Niwa et al., 1998). Introduction of a conditionally active version of STAT3 maintains the pluripotency of ES cells without LIF stimulation (Matsuda et al., 1999). In spite of its essential and sufficient roles in mouse ES cells, LIF/ JAK/STAT3 signaling is not involved in the maintenance of primate ES cell pluripotency (Daheron et al., 2004; Humphrey et al., 2004; Sumi et al., 2004). Furthermore, signaling by bone morphogenetic proteins (BMP) works together with LIF to maintain the pluripotency of mouse ES cells (Ying et al., 2003a; Qi et al., 2004). However, Noggin, an inhibitor of BMP signaling, sustains undifferentiated states of human ES cells in the presence of basic FGF (fibroblast growth factor) (Xu et al., 2005). In contrast to JAK/STAT3 and BMP signalings, Wnt/b-catenin signaling participates in the control of pluripotency in both mouse and human ES cells (Kielman et al., 2002; Sato et al., 2004). A variety of growth factors, cell adhesion molecules and chemokines activate phosphoinositide 3-kinase (PI3K), which generates the second messenger molecule, phosphatidylinositol (3,4,5)-triphosphate (PtdIns (3,4,5)P3) from PtdIns(4,5)P2 (Cantley, 2002). PtdIns (3,4,5)P3 transmits the signals through downstream effectors including Akt, a serine/threonine kinase implicated in the regulation of cell cycle progression, cell death, adhesion, migration, metabolism and tumorigenesis (Brazil et al., 2004). The tumor suppressor PTEN is a lipid phosphatase that catalyses the dephosphorylation from PtdIns(3,4,5)P3 to PtdIns(4,5)P2, thereby antagonizing the physiological and pathological processes of PI3K/Akt signaling (Stiles et al., 2004). Involvement of PI3K/Akt signaling in the regulation of stem cell systems has been proposed from the studies of conditional Pten-deficient mice. The production of embryonic germ (EG) cells is enhanced in primordial germ cell (PGC)-specific mutant mice (Kimura et al., 2003b) and self-renewal of neural stem cells is increased in brain-specific mutant mice (Groszer et al., 2001). Recently, it has been reported that inhibition of PI3K and Akt induces differentiation of mouse and human ES cells, suggesting that PI3K/Akt signaling is necessary for the maintenance of ES cell pluripotency. Here, we demonstrate that the activation of Akt signaling is

Akt signaling in pluripotency of ES cells S Watanabe et al

2698 myr-Akt #11

98.92%

myr-Akt #12

b

control

myr-Akt #9

cell number

a

98.13%

98.74%

myr-Akt #9 #11 #12

p-Akt Akt EGFP

c

β-actin

(%) 100

+LIF

-LIF d4

-LIF d8

* * *

* * *

% of colonies

80 differentiated partially differentiated undifferentiated

60 40 20

+LIF

d

nt r EG ol FP

#9 #11 #12 myr-Akt

co

nt r EG ol FP

#9 #11 #12 myr-Akt

co

co

nt r EG ol FP

0 #9 #11 #12 myr-Akt

-LIF d4

-LIF d8

control

myr-Akt

e

+LIF

-LIF

f

control

control

myr-Akt

myr-Akt

BF

Oct3/4

Figure 1 LIF-independent maintenance of undifferentiated phenotypes in ES cells expressing constitutively active Akt. (a) EGFP expression in parental and myr-Akt-expressing ES cell lines (nos. 9, 11 and 12). EGFP expression of the ES cells carrying the expression plasmid encoding active Akt followed by IRES-EGFP was analysed by flow cytometry. Dotted lines, parental cells; solid lines, myrAkt-expressing cell lines. (b) Western blot analysis using anti-Akt, anti-phospho-Akt and anti-b-actin antibodies. Open and closed arrowheads represent the endogenous and the introduced constitutively active Akt, respectively. (c) Differentiation states of the parental, EGFP- and myr-Akt-expressing ES cells. The cells were seeded at low density and cultured in the presence or absence of LIF. After 4 and 8 days, the colonies were stained with the ALP substrate FastRed. The percentages of undifferentiated ALP-positive and differentiated ALP-negative colonies were calculated (mean7s.d., n ¼ 3). The colonies that had differentiated cells only at the margins of colonies were counted as partially differentiated colonies. In the absence of LIF, the percentages of the completely undifferentiated colonies of the myr-Akt-expressing ES cells were significantly higher than those of parental and control EGFP-expressing ES cells (*Po0.0001 by Student’s t-test). (d) ES cell colonies stained with FastRed. The myr-Akt-expressing ES cells were stained red with FastRed after 4 and 8 days culture without LIF, showing the undifferentiated phenotype. (e) Morphology of day 8 colonies. The myr-Akt-expressing cells retained the characteristic features of undifferentiated ES cells, that is, round morphology, low cytoplasmto-nuclei ratio and piled-up colonies. (f) Oct-3/4 immunostaining of colonies on day 8. Strong Oct-3/4 expression was observed in myr-Akt-expressing cells but not in parental cells. BF, bright field. Open bars, 200 mm; closed bars, 100 mm.

Oncogene


2699

sufficient to maintain the pluripotency of mouse and human ES cells and provides compelling evidence that PI3K/Akt signaling regulates ‘stemness’ in various stem cell systems.

a

control 0

15

myr-Akt

30

60

0

15

30

60 (min.)

p-STAT3

Results

STAT3

control LIF

+

-

myr-Akt +

-

control +

myr-Akt

-

+

-

Nanog Oct-3/4

b

control 0

15

30

myr-Akt 60

0

15

30

60

(min.)

p-ERK

ERK

c

control

myr-Akt

β-catenin

d

35 30 relative luciferase activity

Undifferentiated phenotype of myr-Akt-expressing mouse ES cells in the absence of LIF First, we investigated the potential impact of PI3K/Akt signaling on the pluripotency of ES cells. An expression plasmid encoding the myristoylated, constitutively active form of Akt (myr-Akt) followed by IRES-EGFP (internal ribosome entry site linked to enhanced green fluorescent protein) and a control plasmid with IRESEGFP alone were introduced into a feeder-free mouse ES cell line, E14tg2a. Three ES clones that exhibited high and uniform expression of EGFP were selected and used for the following analyses (Figure 1a). The expression level of the constitutively active form of Akt was higher than that of endogenous Akt (Figure 1b). Although LIF activates PI3K/Akt signaling, the phosphorylation level of myr-Akt was much higher than that of endogenous Akt. To examine the differentiation capacities of myr-Aktexpressing ES cells, these ES cells were seeded at a low density and cultured in the absence of LIF. The differentiation status was examined by cell morphology and the activity of alkaline phosphatase (ALP), one of the specific markers of undifferentiated ES cells. When

TOP FOP

25 20 15 10 5

Rex-1

0

PGC7/Stella TNAP Col. IV Brachyury GAPDH RT(+)

RT(-)

Figure 2 Gene expression pattern in myr-Akt-expressing ES cells. Total RNAs were extracted from EGFP-expressing and myr-Aktexpressing ES cells (no. 12) in the presence of LIF and 8 days after the withdrawal of LIF. The RNAs were subjected to semiquantitative RT-PCR to examine the expression of various marker genes. Nanog, Oct-3/4, Rex-1, PGC7/Stella and TNAP are the markers of the undifferentiated ES cells, whereas Collagen IV and Brachyury are the markers of endodermal and mesodermal cells, respectively.

control

myr-Akt

β-catenin

Figure 3 Activities of STAT3, ERK and b-catenin in myr-Aktexpressing ES cells. (a and b) Activation kinetics of STAT3 and ERK by LIF stimulation. Parental and myr-Akt-expressing ES cells were cultured in the absence of LIF for 24 h. Activation kinetics of STAT3 (a) and ERK (b) by LIF stimulation was examined by Western blot analysis using anti-phospho-STAT3 and anti-phospho-ERK antibodies, respectively. (c) Subcellular localization of b-catenin. Immunostaining shows that, both in the parental and myr-Akt-expressing ES cells, b-catenin was predominantly localized to the plasma membrane. Bar, 300 mm. (d) Transcriptional activity of b-catenin/TCF complex. ES cells were transfected with the TOP reporter plasmid carrying tandem repeats of the b-catenin/TCF binding sites and the control FOP reporter plasmid. The luciferase activities (mean7s.d., n ¼ 3) were analysed by Student’s t-test; activation of the TOP reporter gene was not significantly increased in myr-Akt-expressing cells (P ¼ 0.20). The ES cells expressing b-catenin with a stabilized mutation were used as a positive control.

Oncogene


2700

control ES cells were cultured, approximately 70% of the colonies differentiated within 4 days and the vast majority of the colonies had completely differentiated by day 8 (Figure 1c). The differentiated cells exhibited a high cytoplasm-to-nuclei ratio and were negative for ALP activity (Figure 1d and e). The control ES cell lines carrying only the IRES-EGFP transgene also differentiated similarly. In contrast, when the myr-Akt-expressing ES cells were cultured, more than 90 and 60–80% of the colonies were ALP positive on days 4 and 8 after LIF withdrawal, respectively. The myr-Akt-expressing cells retained a low cytoplasm-to-nuclei ratio and formed round and multi-layered colonies, which are characteristics of undifferentiated ES cells (Figure 1d and e). In addition, Oct-3/4 expression, the hallmark of pluripotent ES cells, was retained in the myr-Aktexpressing cells after 8 days in culture without LIF (Figure 1f). A significant proportion of the myr-Aktexpressing ES cells maintained a morphology similar to undifferentiated ES cells, and the expression of ALP and Oct-3/4 in the absence of LIF for at least for 1 month (data not shown). These results show that myr-Aktexpressing ES cells self-renew with undifferentiated characteristics even in the absence of LIF. Gene expression profile of myr-Akt-expressing mouse ES cells To investigate the differentiation status of myr-Aktexpressing ES cells in more detail, we next analysed the expression of several marker genes of control ES cells along with myr-Akt-expressing ES cells in the presence of LIF and 8 days after the withdrawal of LIF. Nanog, Oct-3/4, PGC7/Stella and Rex-1 are stem cell markers exclusively expressed in pluripotent cells and germ lineages (Ben-Shushan et al., 1998; Pesce and Scholer, 2001; Saitou et al., 2002; Sato et al., 2002; Chambers et al., 2003; Mitsui et al., 2003). TNAP (tissuenonspecific alkaline phosphatase) encodes ALP, which is preferentially expressed in ES cells. Collagen IV and Brachyury are differentiation markers for endoderm and mesoderm, respectively. As shown in Figure 2, expres-

sion of all the stem cell markers was maintained and differentiation markers were not induced in myr-Aktexpressing ES cells on day 8 after the LIF withdrawal. We further examined the global gene expression pattern of the cells using Affymetrix microarrays (Supplementary Figure 1; accession numbers, GSM45082–GSM45085). When differentiation was induced in the control ES cells by the deprivation of LIF, expression of 139 and 200 genes was up- and downregulated more than 2.5-fold, respectively. However, approximately 90% of these genes exhibited only less than 2.5-fold change in Akt-Mer-expressing cells after the removal of LIF. Expression of the representative genes was validated by semiquantitative RT-PCR. Thus, the overall gene expression profile of the control ES cells in the presence of LIF was quite similar to that of the myr-Akt-expressing ES cells in the absence of LIF. Taken together, these results indicate that activation of Akt signaling prevents the differentiation and maintains the undifferentiated state in ES cells in the absence of LIF. STAT3, ERK and b-catenin signaling in myr-Akt-expressing mouse ES cells Various signaling pathways control pluripotency in mouse ES cells. We examined the possibility that Akt signaling maintains the undifferentiated state by affecting these signaling cascades. The activation kinetics of STAT3 in myr-Akt-expressing cells was analysed with a phospho-STAT3-specific antibody because the phosphorylation of STAT3 is sufficient to maintain mouse ES cell pluripotency (Niwa et al., 1998; Matsuda et al., 1999). No significant difference was observed between the parental and myr-Akt-expressing ES cells (Figure 3a). Inhibition of ERK signaling might be able to prevent the diffrentiation of ES cells, as ERK signaling negatively regulates the self-renewal of ES cells (Burdon et al., 1999). However, in myr-Aktexpressing ES cells, ERK signaling was activated and its phosphorylation level was enhanced rather than inhibited (Figure 3b).

Figure 4 Multi-lineage differentiation capacities of myr-Akt-expressing ES cells. (a) Structure of the plasmid, plox-myr-Akt. The plasmid harbors a floxed-myr-Akt expression cassette followed by an EGFP expression cassette. The EB3 ES cells were transfected with this construct and the puromycin-resistant clones were selected. By transfection of the Cre-expressing plasmid, the myr-Akt-expressing cassette can be deleted while EGFP is expressed in the resulting myr-Akt-deleted ES cells. (b) EGFP expression before and after Cremediated recombination. The ES cells with the plox-myr-Akt, LoxP-myr-Akt cells (no. 2), remained undifferentiated after culturing without LIF for 11 days, but the control ES cells differentiated completely. After transient transfection with the Cre-expressing plasmid, the EGFP-positive clones (Dmyr-Akt (no. 2) and Dmyr-Akt (no. 3)) were selected. Dotted lines, LoxP-myr-Akt cells (no. 2); solid lines, Dmyr-Akt (no. 2) and Dmyr-Akt (no. 3) cells. (c) Western blot analysis of Akt expression. Akt expression was not detected after Cre-mediated recombination. Open and closed arrowheads represent the endogenous and the introduced constitutively active Akt, respectively. (d) Restoration of LIF dependence after removal of myr-Akt cDNA. The percentages of undifferentiated and differentiated colonies in the presence and absence of LIF (mean7s.d., n ¼ 3) were analysed as described in Figure 1c (Student’s t-test; *Po0.0001). The clones in which myr-Akt had been deleted (Dmyr-Akt (no. 2) and Dmyr-Akt (no. 3)) differentiated as parental ES cells. (e) LIF dependency of the ES cells containing the plox-myr-Akt before and after Cre recombination. Representative colonies were stained with FastRed. Bar, 200 mm. (f) In vitro differentiation induction of the DAkt cells by an OP9 system. The Dmyr-Akt cells generated mesodermal colonies on day 5 (upper panel) and various hematopoietic cells on day 12 (lower panel). May–Giemsa staining showed megakaryocytes, erythrocytes and neutrophils. Open bar, 100 mm; closed bar, 25 mm. (g) In vitro neuronal differentiation of the Dmyr-Akt cells. The differentiated cells were positive for the TuJ1 neuronal marker. Bar, 50 mm. (h) Teratoma formation of the DmyrAkt cells in nude mice. Teratoma tissues were stained with hematoxylin and eosin (HE). Chondrocytes, a hair follicle and a sebaceous gland are shown in the upper, middle and bottom panels, respectively. Bars, 50 mm. (i) Contribution of the Dmyr-Akt cells to normal development. Microinjection of the Dmyr-Akt cells to blastocysts showed that the Dmyr-Akt cells contributed to chimera mice. EGFPpositive cells were distributed in the whole body of the E9.5 embryo (left: non-chimeric embryo; right: chimeric embryo). Bar, 600 mm. Oncogene


2701

loxP

pCAG

cell number

b

c

loxP IRES puro pA

myr-Akt

∆myr-Akt#2

EGFPpA

∆myr-Akt#3

control

a

also inhibits GSK3b through the phosphorylation of Ser9, b-catenin signaling may be involved in the Aktmediated maintenance of undifferentiated states. Ser9 phosphorylation level of GSK3b was enhanced in the myr-Akt-expressing cells (Supplementary Figure 2a). However, ehnanced Akt signaling did not activate bcatenin signaling in ES cells by the following reasons. First, b-catenin localized to the plasma membrane, but

LoxP myr-Akt#2

Wnt/b-catenin signaling plays a role in maintaining the pluripotency of ES cells (Kielman et al., 2002; Sato et al., 2004). Wnt blocks proteasome-mediated degradation of b-catenin through the inhibition of glycogen synthase kinase 3b (GSK3b) (Moon et al., 2004). Consequently, b-catenin accumulates in the nucleus, forms a heterodimer with transcription factor T-cell factors (TCF) and activates the target genes. As Akt

∆myr-Akt #2 #3

EGFP

d (%)

+LIF

-LIF

*

100

e

80 % of colonies

+LIF

differentiated partially differentiated undifferentiated

-LIF

control 60 40

LoxP myr-Akt#2

c m on yr tro -A l kt #2

0

#2 #3

c m on yr tro -A l kt #2

20

∆myr-Akt

Lo xP

Lo xP

∆myr-Akt

∆myr-Akt#2

#2 #3

f

h

i

d5

d12

g TuJ1

Oncogene


2702

did not accumulate in the nucleus both in the control and myr-Akt-expressing ES cells (Figure 3c). Secondly, the luciferase reporter gene carrying the tandem binding sites for b-catenin/TCF was not activated in myr-Akt-expressing cells (Figure 3d), whereas the reporter was induced upon Wnt3a treatment in the myr-Akt-expressing cells (Supplementary Figure 2b). Finally, treatment of control ES cells with the PI3K inhibitor LY294002 did not affect the Wnt3a-induced luciferase reporter activation (Supplementary Figure 2c). Collectively, Akt signaling maintains the undifferentiated state independent of the STAT3, ERK and b-catenin signaling pathways. Maintenance of the pluripotency of mouse ES cells by Akt signaling A unique property of ES cells is the ability to differentiate into multiple cell lineages. We next investigated whether myr-Akt-expressing ES cells cultured in the absence of LIF maintained pluripotency. For this purpose, the plasmid plox-myr-Akt, which carries the myr-Akt expression cassette flanked by loxP sites at both ends followed by the EGFP expression cassette, was transfected into EB3 feeder-free ES cells (Figure 4a). In the ES cells with the plox-myr-Akt, the myr-Akt-expressing cassette can be excised through Cremediated recombination, which results in the expression of EGFP cDNA (Figure 4b and c). Thus, the developmental capacities of the cells cultured in the absence of LIF can be assessed following the Cre-mediated deletion of myr-Akt. When the parental ES cells were cultured without LIF for longer than 10 days, the cells differentiated completely and no morphologically identifiable ES cells could be recovered by the addition of LIF. In contrast, the ES cells containing the plox-myr-Akt maintained an undifferentiated state in the absence of LIF. The myrAkt-expressing cassette was then deleted by the introduction of the Cre expression plasmid (Figure 4b and c). The in vitro and in vivo differentiation activity of the cells after culturing in the absence of LIF for 11 days was examined by several methods. The ES cells from which the floxed-myr-Akt was excised by Cre recombinase were karyotypically normal and differentiated upon the withdrawal of LIF, which indicates that LIF dependency had been restored (Figure 4d and e). In an in vitro hematopoietic differentiation system using OP9 stromal cells (Nakano et al., 1996), these cells formed mesodermal colonies on day 5 after differentiation induction and produced a variety of hematopoietic cells, such as erythrocytes and granulocytes, on day 12 (Figure 4f). In neuronal differentiation conditions (Ying et al., 2003b), the cells differentiated into TuJ1-positive neurons with long axons (Figure 4g). When transplanted into nude mice, these cells produced teratomas composed of various differentiated cells (Figure 4h). These ES cells were also incorporated into development and contributed to chimeric mice when injected into blastocysts (Figure 4i). These results clearly demonstrate that Akt signaling sufficiently maintains the pluripotency of ES cells in the absence of LIF. Oncogene

Manipulation of pluripotency in mouse ES cells by myr-Akt-Mer fusion protein To manipulate the differentiation status of ES cells in a conditional manner, we engineered mouse E14tg2a ES cells to express myr-Akt-Mer, which is a fusion protein consisting of the active form of Akt and the modified ligand binding domain of the estrogen receptor (Mer) (Figure 5a) (Kohn et al., 1998). In these cell lines, myrAkt-Mer was expressed at the levels comparable to endogenous Akt (Figure 5b). The enzymatic activity of myr-Akt-Mer can be controlled by the addition of the ligand of Mer, 4-hydroxy-tamoxifen (4OHT). myr-AktMer is activated in the presence of 4OHT, whereas it is inactive in the absence of 4OHT (Figure 5b). Neither STAT3 nor b-catenin signaling pathways was activated by the addition of 4OHT in the myr-Akt-Mer-expressing cells (Supplementary Figure 3). In the absence of LIF, the ES cells expressing myr-Akt-Mer maintained undifferentiated phenotypes in the presence but not in the absence of 4OHT (Figure 5c and d). The myr-AktMer-expressing cells that had been cultured in the absence of LIF sustained the capacity to differentiate into hematopoietic cells and neruonal cells after the withdrawal of 4OHT (Figure 5e and f). Furthermore, these cells generated teratomas composed of various differentiated cells (Figure 5g) and contributed to chimeric mice when injected into blastocysts (Figure 5h). Thus, myr-Akt-Mer enabled us to control the differentiation status of mouse ES cells by conditional regulation of Akt signaling. Conditional regulation of cynomolgus monkey ES cell pluripotency by myr-Akt-Mer LIF/STAT3 does not have the function of maintaining the pluripotency of primate ES cells (Daheron et al., 2004; Humphrey et al., 2004; Sumi et al., 2004). Instead, feeder cells or conditioned medium of the feeder cells is required, which indicates that the factors derived from feeder cells are necessary for maintaining primate ES cell pluripotency. To examine the effects of Akt signaling on primate ES cell pluripotency, cynomolgus monkey ES cell lines expressing myr-Akt-Mer were generated (Figure 6a and b). When cultured on the feeder cell layers, the myr-Akt-Mer-expressing cynomolgus monkey ES cells were indistinguishable from the parental ES cells, either in the absence or presence of 4OHT, with respect to their morphology and also the expression of ALP and Oct-3/4 (data not shown). However, when cultured without feeder cells, the control ES cells differentiated within 5 days. In contrast, significant populations of myr-Akt-Mer-expressing ES cells remained undifferentiated in the presence but not in the absence of 4OHT (Figure 6c and d). Undifferentiated phenotypes under the feeder-free conditions continued for at least 1 month (data not shown). Embryoid bodies were generated after culture in the presence of 4OHT without feeder cells for 5 days. A variety of morphologically differentiated cells were observed after the embryoid bodies became adherent to the culture dishes (data not shown). RT-PCR analysis of these cells


2703 myr-Akt-Mer #21

cell number

a

b

myr-Akt-Mer #42

97.3 %

97.7 %

control #21 4OHT - + - +

#42 - +

p-Akt

Akt EGFP β-actin

c

+LIF (%) - 4OHT +4OHT 100

-LIF - 4OHT +4OHT * *

d

-4OHT

+4OHT

control

% of colonies

80 60 #21 40 myr-Akt-Mer 20 #42 l ro

#21#42

nt

#21#42 co

l ro

nt

l ro

#21#42 co

co

co

nt

#21#42

nt

ro

l

0

differentiated partially differentiated undifferentiated

e

g

h

d5

d12

f TuJ1

Figure 5 Manipulation of pluripotency in mouse ES cells by myr-Akt-Mer. (a) Expression of EGFP in parental and myr-Akt-Merexpressing mouse ES cell lines. Dotted lines, parental ES cells; solid lines, Akt-Mer-expressing ES cell lines, nos. 21 and 42. (b) Western blot analysis using anti-Akt, anti-phospho-Akt and anti-b-actin antibodies. The ES cells were cultured in the absence or presence of 1 mM 4OHT for 3 days. Open and closed arrowheads represent the endogenous Akt and myr-Akt-Mer, respectively. (c) LIF dependency of myr-Akt-Mer-expressing ES cells. LIF dependency was analysed as described in Figure 1c. In the absence of LIF, the percentages of the completely undifferentiated colonies of the myr-Akt-Mer-expressing ES cells cultured with 4OHT were significantly higher than those of parental ES cells and the myr-Akt-Mer-expressing ES cells cultured without 4OHT (mean7s.d., *Po0.0001 by Student’s t-test, n ¼ 3). (d) LIF dependency of the ES cells expressing myr-Akt-Mer. Representative colonies stained with FastRed. The myr-Akt-Mer-expressing ES cells were stained red with FastRed after 8 days of culture without LIF. Bar, 200 mm. (e, f) In vitro differentiation potentials of the recovered ES cells. The myr-Akt-Mer-expressing ES cells completely differentiated in the absence of 4OHT, but remained undifferentiated in the presence of 4OHT, after culturing without LIF for 11 days. When these ES cells were applied to in vitro differentiation assays in the absence of 4OHT, they differentiated to hematopoietic cells (e) and neuronal cells (f). Open bars, 100 mm; closed bar, 25 mm. (g) Teratoma formation in nude mice. Neuronal cells, a mucosal gland and a hair follicle are shown in the upper, middle and bottom panels, respectively. Bar, 50 mm. (h) Chimeric mice containing the recovered myr-Akt-Merexpressing ES cell-derived tissues. The myr-Akt-Mer-expressing ES cells that had maintained undifferentiated states by the addition of 4OHT contributed to chimeric mice. EGFP-positive cells were distributed in the whole body of the E9.5 embryo (left: non-chimeric embryo; right: chimeric embryo). Bar, 600 mm.

Oncogene


a

cell number

2704

b

95.83%

control 4OHT

-

+

myr-Akt -Mer -

+

p-Akt EGFP Akt β-actin

c

- 4OHT

(%) 100

+4OHT

* differentiated partially differentiated undifferentiated

% of colonies

80 60 40 20 0

control myr-Akt control myr-Akt -Mer -Mer

- 4OHT

d

ALP

+4OHT

control

myr-Akt -Mer - 4OHT

myr-Akt-Mer

myr-Akt -Mer +4OHT

e

myr-Akt myr-Akt control -Mer -Mer ES EB ES EB ES EB ES EB control

Albumin AFP αMHC cTnT Musashi1 Oct-3/4 GAPDH RT(+)

Oncogene

RT(-)

Oct-3/4


2705

revealed that the myr-Akt-Mer-expressing ES cells, which had been cultured in the presence of 4OHT without feeder cells, differentiated into three germ layers (Figure 6e). These results demonstrate that conditional activation of Akt signaling supports the pluripotency of primate ES cells as well as that of mouse ES cells.

Discussion In this study, we demonstrate that activation of Akt signaling can maintain the pluripotency of mouse and cynomolgus monkey ES cells. Although it has been shown that Wnt/b-catenin signaling also regulates pluripotency in mouse and primate ES cells (Kielman et al., 2002; Sato et al., 2004), Akt signaling functions independently of Wnt/b-catenin signaling, as neither nuclear accumulation of b-catenin nor b-cateninmediated transactivation was detected in the myr-Aktexpressing ES cells. In addition, introduction of the constitutively active GSK3b mutant, which is resistant to Akt-mediated phosphorylation, did not revert the phenotype in myr-Akt-expressing ES cells (data not shown). Hence, Akt signaling is a signaling pathway that functions commonly in mouse and primate ES cells as does Wnt/b-catenin signaling. Treatment of mouse ES cells with the PI3K inhibitor LY294002 inhibits the proliferation and induces cell cycle arrest at the G1 phase (Jirmanova et al., 2002). Both the proliferative and tumorigenic potential are enhanced in Pten-deficient ES cells, which is reverted by the additional deletion of Akt-1 (Sun et al., 1999; Stiles et al., 2002). Furthermore, the proliferative and tumorigenic activites of mouse ES cells are suppressed by the deletion of a small GTPase ERas, which activates PI3K/ Akt signaling, and this phenotype is complemented by the introduction of an active form of PI3K (Takahashi et al., 2003). Thus, it has been considered that PI3K/Akt signaling in ES cells is important for the promotion of cell growth and tumorigenesis. Our results cast a new light on Akt signaling, namely, Akt signaling regulates the pluripotency as well as the proliferation of ES cells. Recently, inhibition of PI3K/Akt signaling in mouse and human ES cells was reported to induce differentia-

tion of the ES cell in the presence of LIF and feeder cells, respectively (Paling et al., 2004; Kim et al., 2005), which supports the claim for the importance of Akt signaling in ES cell differentiation. PI3K/Akt signaling is activated by the growth factors and cell adhesion molecules, which participate in ES cell pluripotency. For example, LIF activates PI3K/Akt signaling in mouse ES cells (Jirmanova et al., 2002; Paling et al., 2004). In human ES cells, basic FGF sustains undifferentiated proliferation in collaboration with Noggin, the inhibitor of BMP signaling (Xu et al., 2005). In addition, the treatment with blocking antibodies against a6b1 integrin induces the differentiation of human ES cells (Kim et al., 2005). Thus, multiple signaling would cooperatively regulate the ES cell pluripotency through the activation of PI3K/Akt signaling. Akt controls various celluar processes through the phosphorylation of a number of target proteins (Brazil et al., 2004). The mTOR (mammalian target of rapamycin)–eIF4E (eukaryotic translational initiation factor 4E) axis is a critical downstream effector for the PI3K/Akt-mediated regulation of cell size, metabolism and tumorigenesis (Hay and Sonenberg, 2004). However, the addition of rapamycin, an inhibitor of mTOR, to the Akt-expressing mouse ES cells inhibited proliferation, but did not induce differentiation (data not shown). Additionally, forced expression of Rheb, an activator of mTOR, and eIF4E, was not sufficient for the LIF-independent maintenance of pluripotency (data not shown), suggesting that the mTOR–eIF4E axis is not a major downstream effector. Cell cycle control differs in ES and differentiated cells. The G1 phase is short in ES cells owing to the absence of the cell cycle inhibition mediated by the CDK inhibitor INK4 and tumor suppressor RB. ES cells acquire these inhibitory mechanisms when the cells undergo differentiation upon LIF withdrawal (Burdon et al., 2002). Akt promotes a G1 to S phase transition by facilitating the formation of cyclin/CDK complexes (Brazil et al., 2004). The inability to gain the inhibitory mechanisms may be the key role of Akt signaling for the differentiation block of ES cells. Consistent with this notion, aberrant cell cycle progression has been observed in Pten-deficient cells, as discussed below. However, this

Figure 6 Regulation of pluripotency in cynomolgus monkey ES cells by myr-Akt-Mer. (a) Expression of EGFP in parental and myrAkt-Mer-expressing cynomolgus monkey ES cell lines. Dotted line, parental ES cells; solid line, myr-Akt-Mer-expressing ES cells. (b) Western blot analysis using anti-Akt, anti-phospho-Akt and anti-b-actin antibodies. The ES cells were cultured in the absence or presence of 4OHT (1 mM) for 3 days. Open and closed arrowheads represent the endogenous Akt and myr-Akt-Mer, respectively. (c) Feeder cell dependency of myr-Akt-Mer-expressing monkey ES cells. The parental and the myr-Akt-Mer-expressing monkey ES cells were cultured under the feeder-free condition with or without 4OHT for 5 days. The percentages of the completely undifferentiated colonies of the myr-Akt-Mer-expressing ES cells cultured with 4OHT were significantly higher than those of parental ES cells and the myr-Akt-Mer-expressing ES cells cultured without 4OHT (mean7s.d., *Po0.01 by Student’s t-test, n ¼ 3). (d) Representative colonies of the monkey ES cells expressing myr-Akt-Mer under the feeder-free condition. The myr-Akt-Mer-expressing cells retained the characteristic features of undifferentiated ES cells in the presence of 4OHT. These cells were positive for ALP and Oct-3/4. (e) In vitro differentiation potentials of the myr-Akt-Mer-expressing monkey ES cells. The myr-Akt-Mer-expressing ES cells that had been cultured in the presence of 4OHT without feeder cells were subjected to embryoid body formation assay. The parental ES cells cultured on feeder cells were used as a control of normal differentiation. After 10 days culture, the embryoid bodies were adherent to the culture dishes and further differentiated for 10 more days. The expression of the marker genes of the three germ layers was analysed by RT-PCR. Albumin and AFP (a-fetoprotein), endoderm markers; aMHC (a-myosin heavy chain) and cTnT (cardiac troponin T), mesoderm markers; Musashi1, ectoderm marker. Bars, 100 mm. Oncogene


2706

explanation does not seem sufficient, as the overexpression of cyclin D, the amount of which is regulated transcriptionally and post-transcriptionally by Akt signaling, did not maintain the undifferentiated phenotype of ES cells (data not shown). Thus, the critical target molecule(s) remain to be identified. Stem cell systems depend on a balance between self-renewal and commitment to differentiation. The involvement of the PI3K/Akt signaling pathway in stem cell regulation has been proposed in studies with conditional Pten knockout mice, in which Akt signaling is upregulated. One study was conducted with PGCspecific Pten-null mice (Kimura et al., 2003b). PGC is the germ cell precursor and source of another pluripotent stem cell, EG cells. In PGC-specific Pten-deficient mice, the proliferation of PGC and the production of EG cells were enhanced in vitro. In vivo, testicular teratomas developed, presumably by de-differentiation from PGCs to pluripotent cells. The other study was in neural cell-specific Pten-null mice (Groszer et al., 2001). Self-renewal and proliferation of neural stem cells were increased in the Pten-deficient mice. Taken together with the present results, the activation of PI3K/Akt signaling pathway skews the balance of self-renewal and differentiation, and maintains the immature stem cell state.

Materials and methods Cell culture The feeder-free mouse ES cells, E14tg2a and the derivatives were maintained as described previously (Niwa et al., 1998). To examine the LIF dependency, the cells were seeded at a density of 100 cells/cm2 and stained using an ALP-staining kit (Sigma, St Louis, MO, USA). For the experiments described in Figures 4 and 5, the floxed-myr-Akt-expressing cells and myrAkt-Mer-expressing cells were seeded at a density of 100 cells/ cm2 and cultured without LIF for 5 days. These cells were then passaged at the same density and cultured without LIF for a further 6 days. The parental and myr-Akt-Mer cells cultured without 4OHT differentiated completely during this culture, and ES cells were never recovered by the addition of LIF. However, floxed-myr-Akt- and myr-Akt-Mer-expressing cells that were cultured in the presence of 4OHT remained undifferentiated, and their developmental capacities were examined. To excise the myr-Akt expression cassette, a Cre expression plasmid was transiently transfected into 1 105 floxed-myr-Akt-expressing cells using Lipofectamine 2000 (Invitrogen, Rockville, MD, USA). The colonies expressing EGFP were picked up and the absence of myr-Akt expression was confirmed by Western blot analysis. The cynomolgus monkey ES cell line CMK6 was purchased from Asahi Technoglass Corporation (Tokyo, Japan). The CMK6 cells were cultured on mitomycin C-treated mouse embryonic fibroblasts as described (Suemori and Nakatsuji, 2003). Transfection was performed by electroporation as described (Furuya et al., 2003). For the feeder-free culture, CMK6 cells were dissociated into small clumps by a dissociation solution (Suemori and Nakatsuji, 2003). To eliminate feeder cells, these clumps were cultured on a gelatin-coated dish for 30 min. The floating ES cells were seeded to a Matrigelcoated dish and cultured in the absence or presence of 1 mM 4OHT for 5 days. For the embryoid body assays, the ES cell clumps were cultured for 10 days in Petri dishes. Oncogene

Plasmids The cDNA encoding myristoylated-human Akt lacking the PH domain (myr-Akt) was cloned into the pCAGGS-IRESEGFPpA vector to produce the active Akt expression plasmid (Niwa et al., 1991; Masuyama et al., 2001). To construct an estradiol-dependent Akt expression vector, the myr-Akt fragment fused to Mer in-frame was cloned into the pCAGGS-IRES-EGFPpA vector. To construct the floxedAkt expression vector, myr-Akt-IRES-puropA was cloned into a pBS246 vector that carried two loxP sequences (Invitrogen). The EGFP expression unit was next incorporated into this plasmid, resulting in the generation of the cassette composed of floxed-myr-Akt-IRES-puropA followed by the EGFP expression unit. The whole cassette was finally cloned into the pCAGGS expression vector. Antibodies Antibodies against phospho(Ser473)-Akt, Akt, phospho (Tyr705)-STAT3, phospho(Ser9)-GSK3b, phospho(Thr202/ Tyr204)-ERK1/2 and ERK1/2 were purchased from Cell Signaling Technology (Beverly, MA, USA). Anti-STAT3, anti-GSK3b, anti-Oct-3/4 and TuJ1 were purchased from Transduction labs (Lexington, KY, USA), Upstate Cell Signaling Solutions (Waltham, MA, USA), Santa Cruz Biotechnology (Santa Cruz, CA, USA) and R&D Systems Inc. (Minneapolis, MN, USA), respectively. RT-PCR Total RNA was isolated by an RNeasy mini kit (Qiagen, Valencia, CA, USA), and 1 mg of total RNA was used for cDNA synthesis. The reverse transcription was performed using the ThermoScript RT-PCR system (Gibco BRL, Rockville, MD, USA) as described (Kimura et al., 2003a). PCR reactions were optimized to allow semiquantitative comparisons within the log phase of amplification. For the mouse genes, the primer sequences and cycles are listed in Supplementary Table 1. The primers for the monkey genes were the same as designed previously (Nagata et al., 2003; Sumi et al., 2004). Luciferase assay Cells (5 104) were transfected on 24-well culture plates with 150 ng pTOPFLASH or pFOPFLASH vector and 20 ng luciferase from Renilla areniformis. After 48 h, luciferase activities were measured in a luminometer and normalized for the data for the transfection efficiency by means of a Dual Luciferase Assay System (Promega, Madison, WI, USA). In vitro differentiation assays In vitro differentiation induction to hematopoietic and neuronal cells from ES cells was performed as described (Nakano et al., 1996; Ying et al., 2003b). Teratoma ES cells (5 106) were injected subcutaneously to nude mice. After 3 weeks, tumors were analysed histologically. All animal studies were conducted according to the guidelines of Osaka University. Acknowledgements We thank Dr H Suemori for generous help of culturing monkey ES cells, and Drs Y Gotoh, T Akiyama, H Niwa, S Takada and S Akira for providing materials and supporting microarray analysis. We also thank Dr E Morii for histological inspection and Ms M Ikeuchi, Y Fujita, T Asada, A Mizokami


2707 and A Kawai for assistance. SW is supported by scholarship from the Japanese Society for Promotion of Science. This work is supported in part by grants from the Ministry of Education,

Science, Sports and Culture, Support Program for Technology Development on the Basis of Academic Findings (NEDO) and Uehara Memorial Foundation.

References Ben-Shushan E, Thompson JR, Gudas LJ, Bergman Y. (1998). Mol Cell Biol 18: 1866–1878. Boeuf H, Hauss C, Graeve FD, Baran N, Kedinger C. (1997). J Cell Biol 138: 1207–1217. Brazil DP, Yang ZZ, Hemmings BA. (2004). Trends Biochem Sci 29: 233–242. Burdon T, Smith A, Savatier P. (2002). Trends Cell Biol 12: 432–438. Burdon T, Stracey C, Chambers I, Nichols J, Smith A. (1999). Dev Biol 210: 30–43. Cantley LC. (2002). Science 296: 1655–1657. Chambers I, Smith A. (2004). Oncogene 23: 7150–7160. Chambers I, Colby D, Robertson M, Nichols J, Lee S, Tweedie S et al. (2003). Cell 113: 643–655. Daheron L, Opitz SL, Zaehres H, Lensch WM, Andrews PW, Itskovitz-Eldor J et al. (2004). Stem Cells 22: 770–778. Furuya M, Yasuchika K, Mizutani K, Yoshimura Y, Nakatsuji N, Suemori H. (2003). Genesis 37: 180–187. Groszer M, Erickson R, Scripture-Adams DD, Lesche R, Trumpp A, Zack JA et al. (2001). Science 294: 2186–2189. Hay N, Sonenberg N. (2004). Genes Dev 18: 1926–1945. Humphrey RK, Beattie GM, Lopez AD, Bucay N, King CC, Firpo MT et al. (2004). Stem Cells 22: 522–530. Jirmanova L, Afanassieff M, Gobert-Gosse S, Markossian S, Savatier P. (2002). Oncogene 21: 5515–5528. Kielman MF, Rindapaa M, Gaspar C, van Poppel N, Breukel C, van Leeuwen S et al. (2002). Nat Genet 32: 594–605. Kim SJ, Cheon SH, Yoo SJ, Kwon J, Park JH, Kim CG et al. (2005). FEBS Lett 579: 534–540. Kimura T, Ito C, Watanabe S, Takahashi T, Ikawa M, Yomogida K et al. (2003a). Mol Cell Biol 23: 1304–1315. Kimura T, Suzuki A, Fujita Y, Yomogida K, Lomeli H, Asada N et al. (2003b). Development 130: 1691–1700. Kohn AD, Barthel A, Kovacina KS, Boge A, Wallach B, Summers SA et al. (1998). J Biol Chem 273: 11937–11943. Masuyama N, Oishi K, Mori Y, Ueno T, Takahama Y, Gotoh Y. (2001). J Biol Chem 276: 32799–32805. Matsuda T, Nakamura T, Nakao K, Arai T, Katsuki M, Heike T et al. (1999). EMBO J 18: 4261–4269.

Mitsui K, Tokuzawa Y, Itoh H, Segawa K, Murakami M, Takahashi K et al. (2003). Cell 113: 631–642. Moon RT, Kohn AD, De Ferrari GV, Kaykas A. (2004). Nat Rev Genet 5: 691–701. Nagata M, Takahashi M, Muramatsu S, Ueda Y, Hanazono Y, Takeuchi K et al. (2003). J Gene Med 5: 921–928. Nakano T, Kodama H, Honjo T. (1996). Science 272: 722–724. Niwa H, Burdon T, Chambers I, Smith A. (1998). Genes Dev 12: 2048–2060. Niwa H, Yamamura K, Miyazaki J. (1991). Gene 108: 193–199. Paling NR, Wheadon H, Bone HK, Welham MJ. (2004). J Biol Chem 279: 48063–48070. Pesce M, Scholer HR. (2001). Stem Cells 19: 271–278. Qi X, Li TG, Hao J, Hu J, Wang J, Simmons H et al. (2004). Proc Natl Acad Sci USA 101: 6027–6032. Saitou M, Barton SC, Surani MA. (2002). Nature 418: 293–300. Sato M, Kimura T, Kurokawa K, Fujita Y, Abe K, Masuhara M et al. (2002). Mech Dev 113: 91–94. Sato N, Meijer L, Skaltsounis L, Greengard P, Brivanlou AH. (2004). Nat Med 10: 55–63. Stiles B, Gilman V, Khanzenzon N, Lesche R, Li A, Qiao R et al. (2002). Mol Cell Biol 22: 3842–3851. Stiles B, Groszer M, Wang S, Jiao J, Wu H. (2004). Dev Biol 273: 175–184. Suemori H, Nakatsuji N. (2003). Methods Enzymol 365: 419–429. Sumi T, Fujimoto Y, Nakatsuji N, Suemori H. (2004). Stem Cells 22: 861–872. Sun H, Lesche R, Li DM, Liliental J, Zhang H, Gao J et al. (1999). Proc Natl Acad Sci USA 96: 6199–6204. Takahashi K, Mitsui K, Yamanaka S. (2003). Nature 423: 541–545. Xu RH, Peck RM, Li DS, Feng X, Ludwig T, Thomson JA. (2005). Nat Methods 2: 185–190. Ying QL, Nichols J, Chambers I, Smith A. (2003a). Cell 115: 281–292. Ying QL, Stavridis M, Griffiths D, Li M, Smith A. (2003b). Nat Biotechnol 21: 183–186.

Supplementary Information accompanies the paper on Oncogene website (http://www.nature.com/onc).

Oncogene