You are on page 1of 12

The FASEB Journal FJ Express Full-Length Article

A negative feedback loop of transcription factors that controls stem cell pluripotency and self-renewal
Guangjin Pan,1 Jun Li,1 Yali Zhou, Hui Zheng, and Duanqing Pei2
Institute of Pharmacology, Department of Biological Sciences & Biotechnology, State Key Laboratory of Biomembrane and Membrane Biotechnology, Institutes of Biomedicine, School of Medicine, Tsinghua University, Beijing, China
ABSTRACT

Embryonic stem (ES) cells possess the ability to renew themselves while maintaining the capacity to differentiate into virtually all cell types of the body. Current evidence suggests that ES cells maintain their pluripotent state by expressing a battery of transcription factors including Oct4 and Nanog. However, little is known about how ES cells maintain the expression of these pluripotent factors in ES cells. Here we present evidence that Oct4, Nanog, and FoxD3 form a negative feedback loop to maintain their expression in pluripotent ES cells. First, Oct4 maintains Nanog activity by directly activating its promoter at sub-steady-state concentration but repressing it at or above steady-state levels. On the other hand, FoxD3 behaves as a positive activator of Nanog to counter the repressive effect of Oct4. The expression of Oct4 is activated by FoxD3 and Nanog but repressed by Oct4 itself, thus, exerting an important negative feedback loop to limit its own activity. Indeed, overexpression of either FoxD3 or Nanog in ES cells failed to increase the concentration of Oct4 beyond the steady-state concentration, whereas knocking down either FoxD3 or Nanog reduces the expression of Oct4 in ES cells. Finally, overexpression of Oct4 or Nanog failed to compensate the loss of Nanog or Oct4, respectively, suggesting that both are required for ES self-renewal and pluripotency. Our results suggest the FoxD3-Nanog-Oct4 loop anchors an interdependent network of transcription factors that regulate stem cell pluripotency.Pan, G., Li, J., Zhou, Y., Zheng, H., Pei, D. A negative feedback loop of transcription factors that controls stem cell pluripotency and self-renewal. FASEB J. 20, E1094 E1102 (2006) Key Words: embryonic stem cells

The successful derivation and establishment of human embryonic stem (ES) cell lines have generated strong interest in harnessing the therapeutic potential of these pluripotent cells to treat human diseases (1). Derived from the inner cell masses of blastocysts, these pluripotent ES cells retain two key characteristics found in embryonic progenitor cells: self-renewal and pluripotency, i.e., the ability of a ES cell to generate exact copies of itself and the capacity of ES cells to differenE1094

tiate into all cell types in the body (25). At the present, there are considerable challenges in maintaining these ES cells in a pluripotent state or triggering ES cells to differentiate into a particular cell type (2, 6). In fact, the precise mechanism that regulates stem cell selfrenewal and pluripotency remains largely unknown (6). Thus, investigation into the molecular and cellular mechanisms of stem cell self-renewal and pluripotency should help meet these challenges and provide the necessary tools to harness the regenerative potential of embryonic stem cells for therapeutical purposes (25). The homeodomain transcription factors Oct4 and Nanog have been proposed as master regulators of stem cell self-renewal and pluripotency (2, 710). Oct4decient embryos fail to develop pluripotent inner cell mass (ICM) cells, which in fact differentiate along the extraembryonic trophoblast lineage (7). Unlike those transcription factors that act in a binary on-off mode, Oct4 appears to regulate cell fates in a quantitative fashion (11). Oct4 must be maintained at a critical concentration to sustain ES cell self-renewal (11). For example, over-expression of Oct4 causes differentiation into endoderm lineage, while its suppression triggers ES cells to become trophectoderm (11). In contrast, over-expression of Nanog, a self-renewal transcription factor isolated by expression cloning, did not induce ES cell differentiation but actually augmented the selfrenewal capacity of ES cells in the absence of LIF (9,10). Ablation of Nanog in ICMs or ES cells, on the other hand, led to the loss of self-renewal and the differentiation of endoderm-like cells (10). Recently, it has been reported that these two master regulators, along with the HMG-containing transcription factor Sox2, co-occupy a substantial portion of their target genes in human ES cells and thus may collaborate to regulate downstream genes important for stem cell self-renewal and differentiation (12). However, it is not clear how ES cells maintain the critical activity of these two and other potential master regulators at the transcription concentration.
These authors contributed equally to this work. Correspondence: GIBH, Chinese Academy of Sciences Guangzhou 510663, China. E-mail: pei-duanqing@gibh.ac.cn doi: 10.1096/fj.05-5543fje
2 1

0892-6638/06/0020-1094 FASEB

Analysis of individual promoters of Oct4 and Nanog revealed a complex set of cis-acting elements and transcription factors that may control their levels of expression. For example, Oct4 promoter contains conserved distal and proximal enhancers that can either repress or activate its expression depending on the binding factors occupying these sites (for review, see ref 2). On the other hand, the Nanog promoter has been suggested as a direct target of the Oct4/Sox2 complex through ChiP analysis, in vitro binding experiments, and RNAi-mediated knockdowns (1214). However, the functional outcome of these potential regulatory potentials remains undened. Furthermore, the dynamic relationship between Oct4 and Nanog has not been explored enough to explain their purported collaboration to regulate stem cell self-renewal and pluripotency. Nanog was originally proposed as a transcription repressor that inhibits the expression of genes important for cell differentiation (9, 10). Our recent studies revealed that it actually contains two unusually strong transactivators at the C-terminus, suggesting that it primarily acts to activate those genes directly involved in maintaining stem cell pluripotency (15, 16). On the other hand, the Oct4 protein appears to have a dual role in mediating both activation and repression of target genes, mostly through interactions with other protein factors such as Sox2 and FoxD3 (2, 17, 18). The winged helix protein FoxD3, isolated as a transcription repressor with restrictive expression in embryonic stem cells (19), is also required for the maintenance of embryonic cells in early mouse embryos (20), suggesting that it may play a role in regulating stem cell self-renewal and pluripotency alongside Oct4 and Nanog. Despite the fact that these transcription factors are all required for the identity of early embryonic progenitor cells, little is known about how Oct4, Nanog, and FoxD3 can regulate the activity of each other and coordinate the expression of genes required for stem cell self-renewal and pluripotency. Here we present the rst such attempt to construct a regulatory loop anchored by Oct4, Nanog, and FoxD3. We found that Oct4 regulates Nanog activity in a dosedependent fashion. Oct4 activates Nanog promoter when expressed below steady-state yet represses it at or above steady-state concentration in ES cells, thus, offering a mechanism to explain the observed dose dependence of Oct4 to regulate stem cell pluripotency (11). We also identied FoxD3 as a novel activator of Nanog, apparently to counteract the repressive activity of Oct4 on Nanog promoter at the steady-state. On the other hand, Nanog and FoxD3 were found to activate Oct4 promoter. Interestingly, Oct4 appears to serve as its own repressor, thus, exerting a negative feedback pressure in the FoxD3-Nanog-Oct4 regulatory loop. Given the established roles of both Nanog and Oct4 in maintaining stem cell pluripotency, we propose that this negative feedback loop anchors a large regulatory

network of transcription factors that control stem cell pluripotency.

MATERIALS AND METHODS


Cell lines and plasmids NIH3T3, P19, and F9 cells were cultured in Dulbeccos modied Eagles medium (DMEM; Invitrogen) supplemented with 10% FBS (Hyclone) and antibiotics (penicillin and streptomycin, 100 g/ml) as described previously (16, 21). Mouse embryonic stem cells or ESCs were maintained on MEFs in ESC medium, which contains DMEM supplemented with 20% FBS, nonessential amino-acids (100 mM), and 0.55 mM 2-mercaptothanol (Invitrogen). The expression plasmids pCR3.1-NanogF and pCR3.1-Oct4 were prepared as described previously (16, 21). The FoxD3 expression plasmid was prepared by inserting its coding sequence obtained by reverse transcription (RT)-polymerase chain reaction (PCR) to the EcoRV site in frame to a FLAG coding sequence as described previously (16, 21). Nanog and Oct4 promoter fragments were amplied by PCR from the mouse liver genome DNA and inserted to the Sma lI site of a promoter less luciferase reporter vector pGL-basic (Promega). The primers were as follows: forward: 5 -acagacaggactgctgggctgcagg-3 , reverse: 5 tggaaagacggctcacctaggg-3 for Oct4 promoter, for a product of 2170 bp; and forward: 5 -tggaaataggaagatcaggagt-3 , reverse: 5 -acagtggtagcaacggtggtagtggca-3 for Nanog promoter with a product of 1120 bp. Nanog siRNA and Oct4 siRNA constructs were prepared by inserting annealed oligonucleiotides, which encode the double strands RNA toward the Oct4 and Nanog mRNA to pBSU6 vector, respectively, as described previously (22). The target sequences are gggtggattctcgaacctggc for Oct4 and gggaacgcctcatcaatgcct for Nanog. A complete GFP expression cassette including cytomeglovirus promoter, GFP coding sequence, and poly(A) was obtained by PCR from a GFP expression vector-pCR3.1 (21) and inserted to the siRNA vectors or to the parental pBSU6 vector (U6-GFP). TK-3(E2)-luci, which contain three copies of ES cell-specic like elements from Nanog promoter, were made by the method as described previously (21). Transfection, RT-PCR, reporter assay, and stable line selection For RT-PCR analysis of lineage markers, ES-R1 cells cultured on 6 cm dishes were transfected with siRNA-GFP expression vector (8 g), Oct4, or Nanog or FoxD3 constructs (8 g each) plus U6-GFP (2 g). GFP-positive cells were sorted 4 days after transfection by FACS and lysed by Trizon (Invitogen), and the lineage markers were detected by RT-PCR. The endogenous transcripts of Nanog and FoxD3 in NIH3T3, P19, F9, ES-R1 cells, and EBs were determined by RT-PCR using GAPDH as a total RNA control. For reporter assay, cells seeded in 24-well plate were transiently transfected with Nanog promoter or Oct4 promoter reporters (0.1 g each) and effector plasmids with increasing doses (0.1 0.4 g) using Lipofectamine2000 (Invitrogen) according to the manufacturers instruction. pCMV-Renilla plasmids (0.005 g per transfection, Promega) were cotransfected in each well as internal references, and the DNA concentrations for all transfections were normalized to equal amounts by adding pCR3.1 empty vector. Thirty-six hours later, cells were washed by PBS and lysed in 70 l of 1 PLB buffer (Promega). Luciferase activity was measured using the Dual-luciferase
E1095

INTERDEPENDENCE BETWEEN Oct4, Nanog, AND FoxD3 IN ES CELLS

Reporter Assay System (Promega) and TD2020 Luminometer (Turner Design). Each transfection was carried out in duplicate and repeated at least twice. For the ES stable line selection, feeder free ES cells maintained in ES medium containing 103 unit/ml LIF (Chemicom) were seeded in 3.5 cm dishes and transfected with 2 g of each expression plasmid. Twenty-four hours after transfection, the cells were divided by 1:15 and seeded to new 10 cm dishes for selection. G418 (500 g/ml, Invitrogen) was added to the medium for the selection. Chromatin immunoprecipitation assay For chromatin immunoprecipitation (Chip) assays, ES-R1 cells cultured in 6 cm dishes were transfected with 8 g of indicated expression constructs. Thirty-six hours later, cells were washed twice with PBS and crosslinked by incubation in 1% formaldehyade for 10 min. The crosslink was then stopped by adding glycine to a nal concentration of 0.125 M. The cells were then washed twice with PBS and lysed in 0.5 ml lysis buffer (0.25% SDS, 200 mM NaCl, 50 mM Tris, pH 8.0, 100 g/ml salmon sprerm DNA, and 1 proteinase inhibitor cocktail from Sigma, St. Louis, MO). Crosslinked chromatin was then sonicated to an average fragment length of 600 bp. The sonicated chromatin was cleared by centrifugation and preabsorbed with 30 l protein A agarose. The preabsorbed chromatin solutions was diluted with two volumes chip buffer (1% Nonidet P-40, 350 mM NaCl) and divided to three portions. One was incubated with 1 g of anti-Oct4 antibody (Ab; 21), one with 1 g of antiag Ab (Sigma), and the other without any rst Ab as control at 4C overnight. The complexes were absorbed by adding 30 l protein A resin to each tube and incubated for 1 h. The resin was then washed three times by wash buffer (1% TrionX100, 1% deoxycholate 0.2 M Tris, pH 8.1, 0.1 M EDTA, and 0.25 M NaCl) and two times by TE buffer. The complexes were eluted by 250 l elution buffer (1% SDS, 0.1 M NaHCO3), and the crosslink was reversed by adding 20 l of 5 M NaCl and a further incubation at 65C for 4 h. After treatment with proteinase K for 1 h at 45C, the DNA was analyzed by PCR using the following primers: forward: 5 -cttttgcattacaatgtccatgg-3 , reverse: 5 -gtcagtgtgatggcgagggaagggat-3 for the amplication of the Chip product by anti-Oct4; and forward: 5 ttgggcatggtggtagacaagc3 , reverse: 5 -gtcagtgtgatggcgagggaaggg-3 for the Chip product by anti-Flag. Three separate PCR reactions were performed, analyzed, and presented in Fig. 1E and Fig. 2F to demonstrate consistency.

RESULTS Oct4 induced mouse ES cells differentiation by suppressing Nanog Oct4 has been shown to determine the fate of embryonic progenitor cells in vivo and stem cells in vitro (8, 11, 18). Specically, down-regulation of Oct4 triggers the mouse ES cells to differentiate into trophectoderm, while over-expression of Oct4 causes the differentiation of mouse ES cells into endoderm tissues (11). However, the precise mechanism through which Oct4 mediates such diverse effects remains unknown. To this end, we compared the phenotypes of ES cells that either overexpress Oct4 or underexpress Nanog. As shown in Fig. 1A, ES cells over-expressing Oct4 are morphologically identical to those underexpressing Nanog, suggesting
E1096 Vol. 20 August 2006

that both factors may have opposing functions inside ES cells. Indeed, analysis of pluripotent and differentiation markers revealed that these two cell populations expressed identical markers of endoderm origin (Fig. 1B), suggesting that ES cells over-expressing Oct4 phenocopy those with Nanog knockdown. Interestingly, we also noticed that the expression of Nanog was suppressed in ES cells with elevated Oct4 (Fig. 1B, top panel), arguing that over-expressed Oct4 mediates its effect via repressing Nanog directly. To test this possibility, we cloned the promoter region of Nanog and investigated its regulation by Oct4 in pluripotent cells. A 926 bp fragment upstream the transcription start site was cloned into a reporter vector, pGL-basic, as shown in Fig. 1C. As shown in Fig 1C, Nanog promoter functioned as expected with high activities in F9 and ES cells and low and no activities in P19 and NIH3T3, respectively, paralleling the endogenous levels of Nanog in these cells. To see if Oct4 can suppress Nanog promoter directly, we cotransfected Oct4 with this Nanog reporter into F9 cells. As a control, we also included a reporter that carries six copies of Oct4-binding sites. As shown in Fig. 1D, Oct4 suppressed the activity of Nanog promoter in a dosedependent fashion, while it activated the control reporter bearing Oct4 binding sites. As these assays were performed transiently within 48 h before the expected differentiation triggered by Oct4 over-expression, these data suggested that over-expressed Oct4 act as a repressor of Nanog, consistent with our prediction from Fig. 1A and B. There is one Oct4 binding site at 181 position within the Nanog promoter (1214), and we subsequently conrmed its occupation by Oct4 by ChiP as shown in Fig. 1E. Two deletion constructs were made, one deleting sequences upstream of 554 position and another deleting upstream of the 181 position but all retaining the Oct4 binding site. As expected, both constructs were repressed by Oct4 in non pluripotent NIH3T3 cells (Fig. 1F). However, a mutation at the Oct4 binding site rendered the 926 bp promoter insensitive to coexpressed Oct4 or Oct4 siRNA, suggesting that Oct4 regulates Nanog promoter specically (Fig. 1G). These data strongly suggested a negative role of Oct4 in regulating Nanog. Indeed, the concentration of Nanog in ES cells over-expressing Oct4 is also reduced signicantly as monitored by Western blot to detect endogenous Nanog (Fig. 1H, upper panel) and transfected Oct4 (Fig. 1H, middle panel). Together, these data demonstrate that elevated Oct4 induces ES cell differentiation by suppressing Nanog. Biphasic regulation of Nanog by Oct4 in ES cells The repressive effect of Oct4 we observed on Nanog promoter appears to be contradictory to the recent report that Oct4 is required to maintain Nanog activity in ES cells by RNAi-mediated knockdown (14). To reconcile this difference, we monitored Nanog promoter activity in a range of Oct4 siRNA concentrations.
PAN ET AL.

The FASEB Journal

Figure 1. Dose-dependent regulation of Nanog by Oct4. A) ES cells with overexpressed Oct4 or down-regulated Nanog share same morphology. R1 ES cells were transfected with U6-GFP, siNanog-GFP, or Oct4F plus U6GFP as indicated. GFP-positive cells were sorted by FACS and cultured further in the presence of LIF. Viable colonies were obtained and passaged. Morphology of obtained clones was photographed in bright elds (left panels) or in uorescence (right panels) as presented. CK, control (in all gures). B) Overexpression of Oct4 and down-regulation of Nanog induced identical differentiation pattern. ES cells transfected with GFP vector (lane 2), Oct4 (lane 3), Nanog siRNA vector (lane 4), and day 6 embryonic bodies (EB6, lane 5) were analyzed by RT-PCR for the expression of pluripotent as well as differentiation markers as indicated. Note identical pattern of marker expression between Oct4 and Nanog siRNA transfected cells. T: brachyury; TM: thrombomodulin. C) Nanog promoter ( 1 kb) recapitulates the endogenous expression of Nanog. Endogenous Nanog transcripts in NIH3T3, P19, F9, and ES cells were analyzed by RT-PCR. Activity of Nanog promoter was also analyzed in these cell lines to show its specicity in pluripotent cells as indicated as described previously (16). D) Oct4 suppresses the activity of Nanog promoter in F9 cells. Nanog promoter construct (NanP, lanes 25) and Oct4 reporter (6w-lu, lanes 6 9) were cotransfected with Oct4 expression constructs with increasing doses into F9 EC cells and luciferase activities were measured and analyzed as described previously (16). E) Binding of Nanog promoter by Oct4 in ES cells. Chromatin immuno-precipitation analysis was performed to demonstrate the in vivo binding of Nanog promoter by Oct4 as described in Materials and Methods. F) Oct4 suppresses Nanog promoter constructs in NIH3T3 cells. Cotransfecton of Oct4 with reporters carrying deletions in the Nanog promoter was performed and analyzed as described in D. G) a mutation in the Oct4 binding site abolished the regulation of Nanog promoter by Oct4. Reporters carrying wild-type and mutant promoters were cotransfected with Oct4 or Oct4 siRNA vectors and analyzed as in D. H) Overexpressed Oct4 reduces the expression of Nanog in ES cells. ES or Oct4 transfected ES cells were lysed, and lysates were separated by SDS-PAGE (10%). Endogenous Nanog or transfected Oct4 protein were detected by anti-Nanog or anti-FLAG (Oct4 is FLAG tagged) antibodies, respectively. Levels of actin in the cell lysate were detected by antiactin Ab as a loading control. I) Regulation of Nanog by Oct4 siRNA in F9 cells. Nanog promoter construct (NanP, lanes 2 6) was cotransfected with increasing doses of Oct4 siRNA construct (lanes 3 6) into F9 cells, and activities were analyzed by luciferase assay. J) Oct4 binding site is required for activity of Nanog promoter in F9 cells. Wide-type and mutant Nanog promoter constructs (lanes 2 and 3) were transfected to F9 cells, and their activities were evaluated by luciferase assay as in D.

As shown in Fig. 1I, an initial up-regulation of Nanog was observed at lower doses of Oct4 siRNA (lanes 3, 4 vs. 2), while higher doses suppressed Nanog activity as reported previously (14). These data demonstrated that Oct4 is essential in maintaining the expression of Nanog. To further conrm this observation, we com-

pared the activities of reporter plasmids carrying the wild-type promoter or a mutated one at the Oct4 binding site in pluripotent cells. As shown in Fig 1J, the mutant plasmid has much lower activity than the wildtype promoter (Fig. 1J, lanes 3 vs. 2), suggesting that Oct4 does play an essential role in maintaining Nanog
E1097

INTERDEPENDENCE BETWEEN Oct4, Nanog, AND FoxD3 IN ES CELLS

expression at steady-state concentration. Taken all these observations together, we can conclude that Oct4 regulates Nanog biphasically, i.e., a steady-state concentration maintains its expression, while an elevated concentration suppresses its expression. FoxD3 as an activator of Nanog The fact that Oct4 represses Nanog activity in ES cells raises the possibility that there must be positive activators in ES cells to sustain the expression of Nanog. Indeed, Nanog is expressed in Oct4 decient embryos (10). One of the candidate transcription factors is FoxD3, a forkhead family transcription factor highly expressed in mouse ES cells and pluriopotent cells in early embryos (19, 20). FoxD3 null embryos died shortly after implantation with a loss of epiblast, a

phenotype similar to Nanog decient embryos (20), suggesting a likely regulatory relationship between these two factors. We analyzed the expression pattern of FoxD3 in pluripotent cells and differentiated EBs. As shown in Fig. 2A, upper panel), FoxD3 is expressed among pluripotent cells such as P19, F9, and ES cells but not in nonpluripotent NIH3T3 cells. During the differentiation of embryonic bodies, FoxD3 is expressed from day 0 to day 4 and then downregulated in days 5 and 6 (Fig. 2A, lower panels). These data suggest that FoxD3 may play an important role in maintaining ES cell pluripotency in a similar fashion as Nanog. It is also apparent that FoxD3 and Nanog are coexpressed from day 0 to day 3 (Fig. 2A, lower panel), indicating a likely regulatory relationship between them. To test this possibility, we performed cotransfection of Nanog promoter reporter with FoxD3 expression

Figure 2. FoxD3 activates expression of Nanog. A) Expression of FoxD3 is restricted in pluripotent cells and down regulated during the differentiation of EBs. FoxD3 was analyzed by RT-PCR in indicated cells and EBs to show its restriction in pluripotent cells. B, C, D) FoxD3 activates Nanog only in puripotent cells. Nanog promoter construct was transfected alone (lane 2) or with FoxD3 (lane 35) with increasing doses to F9 cells (B), ES-R1 cells (C), and NIH3T3 cells (D). Activities were evaluated by luciferase assay as described (16). E) A cis-acting element between 269 and 554 on the Nanog promoter is responsible for the observed activation. A series of 5 deletion constructs of Nanog promoter were made and cotransfected with FoxD3 in a dose fashion and the activities were measured by luciferase assay as before. F) FoxD3 binds to the Nanog promoter. F9 cells were transfected with FoxD3-Flag construct and the cell lysate was analyzed by Chip assay using antiag Ab as described in Materials and Methods. G) An ES cell specic enhancer (E2) like element is present between 269 299 on Nanog promoter. The entire 926 bp of Nanog promoter is presented with deletion sites marked by downward vertical arrows and E2 element by a reverse arrow. H) FoxD3 suppresses the reporter gene bearing 3 copies of E2 like element from Nanog promoter. An alignment between E2 ES cell specic enhancer and the fragment of 269 299 on Nanog promoter were shown. E2 like element was inserted upstreamly to luciferase reporter carrying a minimal TK promoter. This reporter was cotransfected with FoxD3 in a dose fashion to F9 cells and luciferase activity was evaluated as before.
E1098 Vol. 20 August 2006 The FASEB Journal PAN ET AL.

vector in F9 EC cells. As shown in Fig. 2B, FoxD3 activated Nanog promoter in a dose-dependent fashion in F9 cells (Fig. 2B). To further conrm this nding, we performed the same experiments in ES cells and NIH3T3 cells. As expected, FoxD3 strongly up-regulated the activity of Nanog in ES cells (Fig. 2C), but surprisingly, had no effect on the same reporter in nonpluripotent NIH3T3 cells (Fig. 2D). These ndings suggest that FoxD3 may activate Nanog in a pluripotency-specic manner. FoxD3 was originally identied as a transcription repressor through binding an AT rich cis-element (20). To map the cis-elements that FoxD3 utilizes to activate Nanog, a series of deletions were made in Nanog promoter and their activities analyzed. As shown in Fig. 2E, FoxD3 strongly activated reporters bearing 554, 676, and 733 deletions but failed to activate a construct deleting sequences upstream of 269, suggesting that there is a FoxD3 binding site between 269 and 544 in Nanog promoter. Indeed, ChiP analysis conrmed the binding of FoxD3 within this region in vivo in F9 cells (Fig. 2F). A closer examination of the sequences in the 269 and 554 region identied an AT rich element, which is highly homologous to the identied ES specic enhancer (Fig. 2G, H). This

enhancer was shown as a FoxD3 binding site in vitro (20). To further investigate this AT element, we inserted three copies of this element upstream of a minimal TK promoter driving a luciferase reporter (Fig. 2H). When coexpressed with FoxD3, we found that this element is indeed regulated by FoxD3, albeit negatively as reported previously (Fig. 2H; ref 19). Given our nding that FoxD3 activates Nanog expression transcriptionally, we suggest that FoxD3 may assume a dual role in regulating downstream genes, either activation or repression, depending on the context of the target promoter. In the case of Nanog, we showed here that FoxD3 is an activator and, thus, may play an important role in sustaining the expression of Nanog in ES cells. Oct4 activity is maintained at the steady-state concentration by a negative feedback loop in ES cells It is well established that Oct4 must be maintained at a rigid concentration for ES cells to remain pluripotent (8, 11). To date, there has been no attempt to understand how ES cells achieve such a rigid control over Oct4 expression. The apparent biphasic regulation of Nanog by Oct4 observed in Fig. 1 prompted us to

Figure 3. Regulation of Oct4 by Nanog, FoxD3, and Oct4. A) Oct4 promoter is highly active in pluripotent cells and not in nonpluripotent cells. A 2 kb promoter fragment was isolated for Oct4 and analyzed in P19, F9 or NIH3T3 cells by luciferase assay. B, C, D, E) Nanog, FoxD3 can positively regulate the expression of Oct4 while Oct4 suppresses itself. The Oct4 promoter reporter was transfected alone or with Nanog (B, lanes 35), FoxD3 (C, lanes 35), Oct4 (D, lanes 35), and Oct4 siRNA (E, lanes 35) to F9 cells and the activity were evaluated by luciferase assay as described previously (16). F) Modulation of endogenous Oct4 by Nanog, FoxD3 and their siRNAs in ES cells. R1 ES cells were transfected with vector alone (lane 2) or with FoxD3 (FoX, lane 3), Nanog (Nan, lane 4), and siRNAs for Nanog (SiN, lane 5) or FoxD3 (SiF, lane 6), respectively. RNAs were extracted from these transfected cells and analyzed by RT-PCR for expression of Nanog and Oct4 in ES cells.

INTERDEPENDENCE BETWEEN Oct4, Nanog, AND FoxD3 IN ES CELLS

E1099

investigate the regulation of Oct4 by FoxD3 and Nanog. To this end, we cloned the promoter region of Oct4 and analyzed its expression in the context of Nanog, FoxD3, and Oct4 itself. First, we evaluated the activity of this promoter in the P19 and F9 cells and NIH3T3 cells (P19 and F9 are positive for Oct4 and NIH3T3 negative by RT-PCR, data not shown). As expected, the Oct4 promoter has robust activity in pluripotent P19 and F9 cells but not in NIH3T3 cells (Fig. 3A), suggesting that this promoter region is regulated similarly as endogenous Oct4. When analyzed with Nanog and FoxD3, the Oct4 promoter was activated strongly by both factors in a dose-dependent fashion (Fig. 3B and C). Interestingly, when cotransfected with Oct4, we found that Oct4 protein appeared to be a repressor of its own promoter (Fig. 3D). These data suggested that Oct4 had a feedback loop to limit its own quantity. We then performed siRNA knockdown experiment to downregulate endogenous Oct4 and conrmed the inhibitory effect of Oct4 even at the endogenous concentration (Fig. 3E). Thus, we concluded that Oct4 behaves as a persistent repressor of its own promoter in contrast to its biphasic regulation of Nanog (see Fig. 1). The negative autoregulation of Oct4 promoter by Oct4 protein raises the possibility that Oct4 works as a negative feedback mechanism to ensure its own expression set at a steady state in ES cells. To test this possibility, we examined the concentration of Oct4 expression in ES cells over-expressing both FoxD3 and Nanog. As shown in Fig. 3F, neither FoxD3 nor Nanog was able to increase the expression concentration of Oct4 in ES cells (lanes 3 and 4 vs. 2). On the other hand, knockdown of either FoxD3 or Nanog by siRNAs signicantly reduced the expression of Oct4 as expected (Fig. 3F, lanes 5, 6 vs. 2). Similar results were also obtained in F9 cells (data not shown). These data demonstrate a potential feedback inhibitory loop that counterbalances the activation potential of FoxD3 and Nanog on Oct4. The suppression of Oct4 by Oct4 itself is sufcient to achieve the critical concentration of Oct4 at equilibrium, a hallmark function of Oct4 in regulating stem cell pluripotency. Taken together, our results demonstrate a regulatory circuitry among FoxD3, Nanog, and Oct4, which ensures the proper expression of Oct4 in ES cells. Neither Oct4 nor Nanog is dispensable in sustaining the self-renewal of mouse ES cells The interdependence between Nanog and Oct4 to maintain their transcription activity suggest that both master regulators may cooperate to sustain stem cell pluripotency. Indeed, either gene is required for early embryo development and stem cell pluripotency (7, 9, 10). Unlike Oct4, overexpression of Nanog has been demonstrated to positively promote stem cell self-renewal by bypassing the requirement of LIF/Stat pathway (9, 10), raising the possibility that over expression of Nanog may be able to compensate the activity of Oct4. Alternatively, Oct4 and Nanog may function in
E1100 Vol. 20 August 2006

Figure 4. Both Oct4 and Nanog are required for the ES self-renewal. A) Oct4 and Nanog are required for the selfrenewal of ES cells. ES cells were transfected with the constructs for pCR3.1 and U6-GFP (a), pCR3.1 and siOct4 (b), pCR3.1 and siNanog (c), Nanog and siOct4 (d) or Oct4 and siNanog (e) as indicated and treated with G418 selection. Plates were photographed to show differences in pluripotency. Number of drug resistant colonies was counted after 2 wk selection and presented in f. B) A proposed model for Oct4, Nanog, FoxD3 regulatory circuit. FoxD3 and Nanog work positively to regulate Oct4 while Oct4 negatively regulate itself. FoxD3 can up-regulate Nanog while Oct4 can suppress the expression of Nanog.

parallel or tandem as master regulators of stem cell pluripotency. To test these possibilities, we attempted to rescue the loss of Oct4 activity by overexpressing Nanog and vice versa. We cotransfected the RNAi vectors for Oct4 or Nanog either alone or with Nanog or Oct4 expression vectors, respectively, and then analyzed the self-renewal activity of these transfected cells. As expected, stem cells expressing siRNAs for Oct4 or Nanog failed to undergo self-renewal (Fig. 4A,b, c vs. a). Interestingly, the overexpressed Nanog failed to compensate the loss of Oct4 in sustaining self-renewal (Fig. 4A, d vs. a). Conversely, overexpression of Oct4 also failed to rescue the loss of Nanog (Fig. 4A, e vs. a). These data demonstrate that two master regulators, Oct4 and Nanog, need to function in parallel in maintaining ES cell self-renewal and that neither is dispensable and nor capable of compensating the role of the another.

DISCUSSION We report here that Oct4, Nanog, and FoxD3 form an interactive and interdependent circuit for the regulaPAN ET AL.

The FASEB Journal

tion of stem cell pluripotency. Nanog is a direct target of Oct4, and the suppression of Nanog by Oct4 can help explain the previous observation that overexpression of Oct4 leads to endoderm-like differentiation in ES cells (11). Furthermore, we demonstrated that FoxD3 also participates in the regulation of stem cell self-renewal and pluripotency by activating the expression of the two known master regulators, Nanog and Oct4, in ES cells. We propose that this FoxD3-NanogOct4 circuit may anchor a much larger network of transcription factors dedicated to stem cell self-renewal and pluripotency. Self-renewal and pluripotency are two critical properties of embryonic progenitor cells in vivo and embryonic stem cells in vitro, which should be under strict control at the molecular concentration. Given their signicant role in development, considerable amount of resources must have been devoted to maintain these two important properties and ensure the continuity of life. At the core of this investment, transcriptional factors such as Oct4 and Nanog have been shown to play a critical role at the individual basis in controlling stem cell self-renewal and pluripotency. Yet, the regulatory logic among these important factors remains unresolved. To this end, the interdependent relationship among Oct4, Nanog, and FoxD4 illustrated in Fig. 4B represents the rst step toward a rational understanding of stem cell self-renewal and pluripotetncy. This regulatory circuit also reveals an important negative feedback loop. Whereas a positive feedback system leads to unsustainable escalation or downward spiral of activities, a negative feedback system is destined to achieve a state of equilibrium and a sustaining xed concentration of output. Most of the biological processes tend to adapt a negative feedback loop. In the case of ES cell self-renewal and pluripotency, a negative feedback loop presented in Fig. 4B appears to evolve to control these important properties for animal development. Anchored by Oct4, this negative feedback loop may also serve as an integrator or converter for all intracellular and extracellular signals that destined to regulate stem cell pluripotency. This is in agreement with a previous report that only a critical concentration of Oct4 can maintain stem cells at pluripotent state, both under- and overexpression of Oct4 lead to cell differentiation (11). In contrast, Nanog does not have a self-regulatory loop (data not shown). Its overexpression has been shown to further enhance the stemness of ES cells and prevent the differentiation of ES cells induced by retinoic acid (9), suggesting that it may regulate other critical genes in addition to Oct4. This will also explain why they can not rescue each other because this loop is interrupted when one of them is missing from the circuit. We believe that this circuit operates in the ES cells to ensure the activity of the critical master regulators, namely Nanog and Oct4.
This work was supported in part by the Tsinghua University BaiRen Scholar Program, NSFC 30270287 and 30470839, the

973 Project-2001CB5101 (P.I. Lingsong Li) from The Ministry of Science and Technology of China, and the Tsinghua Yue-Yuen Medical Fund. D. Pei is a CheungKong (Changjiang) Scholar of the Li Ke-Shing Foundation and Ministry of Education, China. We acknowledge the support of Professor Nanming Zhao at Tsinghua University. The assistance from M. Chen, Y.Q. Guo, and the rest of the Pei laboratory made this study possible.

REFERENCES
1. Thomson, J. A., Itskovitz-Eldor, J., Shapiro, S. S., Waknitz, M. A., Swiergiel, J. J., Marshall, V. S., and Jones, J. M. (1998) Embryonic stem cell lines derived from human blastocysts. Science 282, 11451147 Pan, G. J., Chang, Z. Y., Scholer, H. R., and Pei, D. (2002) Stem cell pluripotency and transcription factor Oct4. Cell Res. 12, 321329 Audet, J. (2004) Stem cell bioengineering for regenerative medicine. Expert. Opin Biol Ther. 4, 631 644 Constantinescu, S. (2003) Stemness, fusion and renewal of hematopoietic and embryonic stem cells. J. Cell Mol. Med. 7, 103112 Koike, N., Fukumura, D., Gralla, O., Au, P., Schechner, J. S., and Jain, R. K. (2004) Tissue engineering: creation of long-lasting blood vessels. Nature 428, 138 139 Chambers, I., and Smith, A. (2004) Self-renewal of teratocarcinoma and embryonic stem cells. Oncogene 23, 7150 7160 Nichols, J., Zevnik, B., Anastassiadis, K., Niwa, H., Klewe-Nebenius, D., Chambers, I., Scholer, H., and Smith, A. (1998) Formation of pluripotent stem cells in the mammalian embryo depends on the POU transcription factor Oct4. Cell 95, 379 391 Pesce, M., and Scholer, H. R. (2000) Oct-4: control of totipotency and germline determination. Mol. Reprod. Dev. 55, 452 457 Chambers, I., Colby, D., Robertson, M., Nichols, J., Lee, S., Tweedie, S., and Smith, A. (2003) Functional expression cloning of Nanog, a pluripotency sustaining factor in embryonic stem cells. Cell 113, 643 655 Mitsui, K., Tokuzawa, Y., Itoh, H., Segawa, K., Murakami, M., Takahashi, K., Maruyama, M., Maeda, M., and Yamanaka, S. (2003) The homeoprotein Nanog is required for maintenance of pluripotency in mouse epiblast and ES cells. Cell 113, 631 642 Niwa, H., Miyazaki, J., and Smith, A. G. (2000) Quantitative expression of Oct-3/4 denes differentiation, dedifferentiation or self-renewal of ES cells. Nat. Genet. 24, 372376 Boyer, L. A., Lee, T. I., Cole, M. F., Johnstone, S. E., Levine, S. S., Zucker, J. P., Guenther, M. G., Kumar, R. M., Murray, H. L., Jenner, R. G., et al. (2005) Core transcriptional regulatory circuitry in human embryonic stem cells. Cell 122, 947956 Kuroda, T., Tada, M., Kubota, H., Kimura, H., Hatano, S. Y., Suemori, H., Nakatsuji, N., and Tada, T. (2005) Octamer and Sox elements are required for transcriptional cis regulation of Nanog gene expression. Mol. Cell. Biol. 25, 24752485 Rodda, D. J., Chew, J. L., Lim, L. H., Loh, Y. H., Wang, B., Ng, H. H., and Robson, P. (2005) Transcriptional regulation of nanog by OCT4 and SOX2. J. Biol. Chem. 280, 2473124737 Pan, G. J., and Pei, D. Q. (2003) Identication of two distinct transactivation domains in the pluripotency sustaining factor nanog. Cell Res. 13, 499 502 Pan, G., and Pei, D. (2005) The stem cell pluripotency factor NANOG activates transcription with two unusually potent subdomains at its C terminus. J. Biol. Chem. 280, 14011407 Guo, Y., Costa, R., Ramsey, H., Starnes, T., Vance, G., Robertson, K., Kelley, M., Reinbold, R., Scholer, H., and Hromas, R. (2002) The embryonic stem cell transcription factors Oct-4 and FoxD3 interact to regulate endodermalspecic promoter expression. Proc. Natl. Acad. Sci. USA. 99, 36633667

2. 3. 4. 5. 6. 7.

8. 9.

10.

11. 12.

13.

14. 15. 16. 17.

INTERDEPENDENCE BETWEEN Oct4, Nanog, AND FoxD3 IN ES CELLS

E1101

Pesce, M., and Scholer, H. R. (2001) Oct-4: gatekeeper in the beginnings of mammalian development. Stem Cells 19, 271278 19. Sutton, J., Costa, R., Klug, M., Field, L., Xu, D., Largaespada, D. A., Fletcher, C. F., Jenkins, N. A., Copeland, N. G., Klemsz, M., and Hromas, R. (1996) Genesis, a winged helix transcriptional repressor with expression restricted to embryonic stem cells. J. Biol. Chem. 271, 23126 23133 20. Hanna, L. A., Foreman, R. K., Tarasenko, I. A., Kessler, D. S., and Labosky, P. A. (2002) Requirement for Foxd3 in maintaining pluripotent cells of the early mouse embryo. Genes Dev. 16, 2650 2661

18.

21.

Pan, G., Qin, B., Liu, N., Scholer, H. R., and Pei, D. (2004) Identication of a nuclear localization signal in OCT4 and generation of a dominant negative mutant by its ablation. J. Biol. Chem. 279, 3701337020 22. Sui, G., Soohoo, C., Affar el, B., Gay, F., Shi, Y., and Forrester, W. C. (2002) A DNA vector-based RNAi technology to suppress gene expression in mammalian cells. Proc. Natl. Acad. Sci. USA. 99, 55155520 Received for publication January 15, 2006. Accepted for publication March 20, 2006.

E1102

Vol. 20

August 2006

The FASEB Journal

PAN ET AL.

The FASEB Journal FJ Express Summary

A negative feedback loop of transcription factors that controls stem cell pluripotency and self-renewal
Guangjin Pan,1 Jun Li,1 Yali Zhou, Hui Zheng, and Duanqing Pei2
Institute of Pharmacology, Department of Biological Sciences and Biotechnology, State Key Laboratory of Biomembrane and Membrane Biotechnology, Institutes of Biomedicine, School of Medicine, Tsinghua University, Beijing, China To read the full text of this article, go to http://www.fasebj.org/cgi/doi/10.1096/fj.05-5543fje

SPECIFIC AIMS The successful derivation and establishment of human embryonic stem (ES) cell lines have generated strong interest in harnessing the therapeutic potential of these pluripotent cells to treat human diseases. At the present, there are considerable challenges in maintaining these ES cells in a pluripotent state or triggering ES cells to differentiate into a particular cell type. In fact, the precise mechanism that regulates stem cell selfrenewal and pluripotency remains largely unknown. The homeodomain transcription factors Oct4 and Nanog have been proposed as master regulators of stem cell self-renewal and pluripotency. However, it is not clear how ES cells maintain the critical activity of these two and other potential master regulators at the transcription level. We would like to determine the molecular mechanism that governs the regulation of Oct 4 and ES cells.

PRINCIPAL FINDINGS 1. Oct4 induced mouse ES cells differentiation by suppressing Nanog Our initial aim was to dene the relationship between Nanog and Oct4, two transcription factors implicated in stem cell self-renewal and pluripotency. More specifically, we would like to dene if Oct4 and Nanog regulate each other in ES cells. To this end, we compared the phenotypes of ES cells that either overexpress or underexpress Oct4 or Nanog by their expression constructs or siRNA constructs. To our surprise, ES cells over-expressing Oct4 are morphologically identical to those with Nanog knocked down by siRNA (Fig. 1A), suggesting that both factors may have opposing function inside ES cells. Indeed, analysis of pluripotent and differentiation markers revealed that these two cell populations expressed identical markers of endoderm origin (Fig. 1B), suggesting that ES cells over-expressing Oct4 phenocopy those with Nanog knockdown. Interestingly, we also noticed that the expression of Nanog was suppressed in ES cells with elevated Oct4
1730

(Fig. 1B, top panel), arguing that over-expressed Oct4 mediates its effect via repressing Nanog directly. To test this possibility, we analyzed the effect of Oct4 on the Nanog promoter directly. As shown in Fig 1C, the Nanog promoter functioned as expected with high activities in F9 and ES cells and low and no activities in P19 and NIH3T3, respectively, paralleling the endogenous levels of Nanog in these cells. Oct4 when coexpressed suppressed the activity of Nanog promoter in a dose-dependent fashion while activating the control reporter bearing Oct4 binding sites (Fig 1D). There is one Oct4 binding site at 181 position within the Nanog promoter of which we conrmed its occupation by Oct4 by chromatin immunoprecipitation assay (ChiP) as shown in Fig. 1E. Deletion analysis shown in Fig. 1F demonstrated that the two constructs bearing this Oct4 binding site were repressed by Oct4 in nonpluripotent NIH3T3 cells. Mutation at the Oct4 binding site rendered the 926bp promoter insensitive to coexpressed Oct4 or Oct4 siRNA, suggesting that Oct4 regulates Nanog promoter specically (Fig. 1G). These data strongly suggested a negative role of Oct4 in regulating Nanog. Indeed, the concentration of Nanog in ES cells over-expressing Oct4 is also reduced signicantly as monitored by Western blot (Fig. 1H). Together, these data demonstrate that elevated Oct4 induces ES cell differentiation by suppressing Nanog. 2. Biphasic regulation of Nanog by Oct4 in ES cells The repressive effect of Oct4 we observed on Nanog promoter appears to be contradictory to the recent report that Oct4 is required to maintain Nanog activity in ES cells by RNAi-mediated knockdown. To reconcile this difference, we monitored Nanog promoter activity in a range of Oct4 siRNA concentrations. Figure 1I demonstrated an initial up-regulation of Nanog by lower doses of Oct4 siRNA (lanes 3, 4 vs. 2), and higher doses suppressed Nanog activity as reported. Figure 1J
These authors contributed equally to this work. Correspondence: GIBH, Chinese Academy of Science, Guangzhou 510663, China. E-mail: pei-duanqing@gibh.ac.cn doi: 10.1096/fj.05-5543fje
2 1

0892-6638/06/0020-1730 FASEB

Figure 1. Regulation of Nanog by Oct4. A) ES cells with overexpressed Oct4 or down-regulated Nanog share same morphology. R1 ES cells were transfected with U6-GFP, siNanogGFP, or Oct4F plus U6GFP as indicated. Morphology of obtained clones was photographed in bright elds (left panels) or in uorescence (right panels) as presented. B) Overexpression of Oct4 and down-regulation of Nanog induced identical differentiation pattern. ES cells transfected with GFP vector (lane 2), Oct4 (lane 3), Nanog siRNA vector (lane 4), and day 6 embryonic bodies (EB6, lane 5) were analyzed by reverse transcriptase-polymerase chain reaction for the expression of pluripotent as well as differentiation markers as indicated. Note identical pattern of marker expression between Oct4 and Nanog siRNA transfected cells. C) Nanog promoter ( 1 kb) recapitulates the endogenous expression of Nanog. Endogenous Nanog transcripts in NIH3T3, P19, F9, and ES cells were analyzed by reverse transcription (RT)-polymerase chain reaction (PCR). Activity of Nanog promoter was also analyzed in these cell lines to show its specicity in pluripotent cells as indicated as described (16). D) Oct4 suppresses activity of Nanog promoter in F9 cells. Nanog promoter construct (NanP, lanes 25) and Oct4 reporter (6w-lu, lanes 6 9) were cotransfected with Oct4 expression constructs with increasing doses into F9 EC cells, and luciferase activities were measured and analyzed as described previously (16). E) Binding of Nanog promoter by Oct4 in ES cells. Chromatin immuno-precipitation analysis was performed to demonstrate in vivo binding of Nanog promoter by Oct4. F) Oct4 suppresses Nanog promoter constructs in NIH3T3 cells. Cotransfecton of Oct4 with reporters carrying deletions in the Nanog promoter was performed and analyzed as described in D. G) A mutation in the Oct4 binding site abolished regulation of Nanog promoter by Oct4. Reporters carrying wild-type and mutant promoters were cotransfected with Oct4 or Oct4 siRNA vectors and analyzed as in D. H) Overexpressed Oct4 reduces the expression of Nanog in ES cells. ES or Oct4 transfected ES cells were lysed and fractioned by 10% SDS-PAGE. Endogenous Nanog or transfected Oct4 protein was detected by anti-Nanog or anti-FLAG (Oct4 is FLAG tagged) antibodies, respectively. Levels of actin in the cell lysate were detected by antiactin antibody as a loading control. I) Regulation of Nanog by Oct4 siRNA in F9 cells. Nanog promoter construct (NanP, lane2 6) was cotransfected with increasing doses of Oct4 siRNA construct (lane 3 6) into F9 cells, and activities were analyzed by luciferase assay. J) Oct4 binding site is required for activity of Nanog promoter in F9 cells. Wide-type and mutant Nanog promoter constructs (lanes 2 and 3) were transfected to F9 cells, and their activities were evaluated by luciferase assay as in D.

shows that the mutant plasmid bearing Oct4 site mutation has much lower activity than the wild-type promoter (Fig. 1J, lanes 3 vs. 2), suggesting that Oct4 does play an essential role in maintaining Nanog expression at steady-state concentration. Taken all these observations together, we can conclude that Oct4 regulates Nanog biphasically, i.e., a steady-state concentration

maintains its expression, while an elevated concentration suppresses its expression. 3. FoxD3 as an activator of Nanog The second aim was to identify additional factors that may positively regulate the expression of Nanog, thus,
1731

INTERDEPENDENCE BETWEEN Oct4, Nanog, AND FoxD3 IN ES CELLS

knockdown experiment to down-regulate endogenous Oct4 and conrmed the inhibitory effect of Oct4 even at the endogenous concentration. Thus, we concluded that Oct4 behaves as a persistent repressor of its own promoter, in contrast to its biphasic regulation of Nanog (see Fig. 1). The negative autoregulation of Oct4 promoter by Oct4 protein raises the possibility that Oct4 works as a negative feedback mechanism to ensure its own expression set at a steady state in ES cells. Indeed, our data demonstrate a feedback inhibitory loop that counterbalances the activation potential of FoxD3 and Nanog on Oct4. The suppression of Oct4 by Oct4 itself is sufcient to achieve the critical concentration of Oct4 at equilibrium, a hallmark function of Oct4 in regulating stem cell pluripotency. Taken together, our results demonstrate a regulatory circuitry among FoxD3, Nanog, and Oct4, which ensures the proper expression of Oct4 in ES cells.
Figure 2. Schematic diagram.

5. Neither Oct4 nor Nanog is dispensable in sustaining the self-renewal of mouse ES cells The interdependence between Nanog and Oct4 to maintain their transcription activity suggests that both master regulators may cooperate to sustain stem cell pluripotency. We cotransfected the RNAi vectors for Oct4 or Nanog either alone or with Nanog or Oct4 expression vectors, respectively, and then analyzed the self-renewal activity of these transfected cells. Stem cells expressing siRNAs for Oct4 or Nanog failed to undergo self-renewal. The overexpressed Nanog also failed to compensate the loss of Oct4 in sustaining self-renewal. Conversely, overexpression of Oct4 failed to rescue the loss of Nanog. These data demonstrate that two master regulators, Oct4 and Nanog, need to function in parallel in maintaining ES cell self-renewal and that neither is dispensable and nor capable of compensating the role of the another.

counteracting the repressive effects of Oct4 observed in the previous section. FoxD3 is expressed among pluripotent cells such as P19, F9, and ES cells but not in nonpluripotent NIH3T3 cells. We thus hypothesized that FoxD3 may play an important role in maintaining ES cell pluripotency in a similar fashion as Nanog. We then performed cotransfection of Nanog promoter reporter with FoxD3 expression vector in F9 EC cells and demonstrated that FoxD3 is a very potent activator of Nanog in a pluripotency specic manner, apparently through a FoxD3 binding site between 269 and 544 in Nanog promoter. 4. Oct4 activity is maintained at the steady-state concentration by a negative feedback loop in ES cells The third aim was to establish a regulatory circuit among FoxD3, Nanog, and Oct4 in ES cells. The apparent biphasic regulation of Nanog by Oct4 observed in Fig. 1 prompted us to investigate the regulation of Oct4 by FoxD3 and Nanog. To this end, we cloned the promoter region of Oct4 and analyzed its expression in the context of Nanog, FoxD3, and Oct4 itself. First, we evaluated the activity of this promoter in pluripotent P19 and F9 cells and nonpluripotent NIH3T3 cells. As expected, the Oct4 promoter has robust activity in pluripotent P19 and F9 cells but not in NIH3T3 cells. When analyzed with Nanog and FoxD3, the Oct4 promoter was activated strongly by both factors in a dose-dependent fashion. Interestingly, when cotransfected with Oct4, we found that Oct4 protein appeared to be a repressor of its own promoter. These data suggested that Oct4 had a feedback loop to limit its own quantity. We then performed siRNA

CONCLUSIONS AND SIGNIFICANCE We conclude that Oct4, Nanog, and FoxD3 form an interactive and interdependent circuit for the regulation of stem cell pluripotency. Nanog is a direct target of Oct4, and the suppression of Nanog by Oct4 can help explain the previous observation that overexpression of Oct4 leads to endoderm-like differentiation in ES cells. Furthermore, we demonstrated that FoxD3 also participates in the regulation of stem cell selfrenewal and pluripotency by activating the expression of the two known master regulators, Nanog and Oct4, in ES cells. We propose that this FoxD3-Nanog-Oct4 circuit may anchor a much larger network of transcription factors dedicated to stem cell self-renewal and pluripotency.

1732

Vol. 20

August 2006

The FASEB Journal

PAN ET AL.