The C-terminal Pentapeptide of Nanog Tryptophan Repeat Domain Interacts with Nac1 and Regulates Stem Cell Proliferation but Not Pluripotency

Abstract

Overexpression of Nanog in mouse embryonic stem (ES) cells has been shown to abrogate the requirement of leukemia inhibitory factor for self-renewal in culture. Little is known about the molecular mechanism of Nanog function. Here we describe the role of the tryptophan repeat (WR) domain, one of the two transactivators at its C terminus, in regulating stem cell proliferation as well as pluripotency. We first created a supertransactivator, W2W3×10, by duplicating repeats W2W3 10 times and discovered that it can functionally substitute for wild type WR at sustaining pluripotency, albeit with a significantly slower cell cycle, phenocopying Nanog(9W) with the C-terminal pentapeptide (WNAAP) of WR deleted. ES cells carrying both W2W3×10 and Nanog(9W) have a longer G₁ phase, a shorter S phase in cell cycle distribution and progression analysis, and a lower level of pAkt(Ser⁴⁷³) compared with wild type Nanog, suggesting that both mutants impact the cell cycle machinery via the phosphatidylinositol 3-kinase/Akt pathway. Both mutants remain competent in dimerizing with Nanog but cannot form a complex with Nac1 efficiently, suggesting that WNAAP may be involved in Nac1 binding. By tagging Gal4DBD with WNAAP, we demonstrated that this pentapeptide is sufficient to confer Nac1 binding. Furthermore, we can rescue W2W3×10 by placing WNAAP at the corresponding locations. Finally, we found that Nanog and Nac1 synergistically up-regulate ERas expression and promote the proliferation of ES cells. These results suggest that Nanog interacts with Nac1 through WNAAP to regulate the cell cycle of ES cells via the ERas/phosphatidylinositol 3-kinase/Akt pathway, but not pluripotency, thus decoupling cell cycle control from pluripotency.

Recent advances have identified Oct4, Sox2, and Nanog as core factors for the mammalian pluripotency program (1). Remarkably, some of these pluripotent factors have also been successfully utilized to reprogram somatic cells back to the pluripotent state through the iPS or induced pluripotent stem cell protocol (2–6).

Nanog is a relatively new arrival into the pluripotent factor family (7, 8). Discovered by its in vitro ability to sustain ES² cell self-renewal in the absence of LIF, Nanog was recently shown to possess reprogramming potential during the generation of human iPS cells, suggesting that it possesses power similar to that of other core regulators, such as Oct4 and Sox2. Paradoxically, recent work from Chambers et al. (9) has demonstrated that Nanog works to safeguard, but is not required for, pluripotency and appears to play a more direct role in germ line maintenance. Through high throughput technologies, several groups have identified the downstream targets of Nanog in the genome as well as proteins with which Nanog interacts (10, 11). Although these prominent studies illustrate the potential complexity of the function networks Nanog regulates, they describe very little how Nanog achieves these activities. The structural basis of Nanog function remains largely undefined.

Embyonic stem cells can undergo unlimited self-renewal, so that the cell cycle appears to be less controlled than the somatic ones. For example, although RB plays a key role in the progression of somatic cell cycle through its phosphorylation by cyclin D/CDK4 or cyclin D/CDK6 and subsequent release of E2F to allow the expression of downstream genes critical for the progression through the G₁/S checkpoint, embryonic stem cells execute cell cycles independent of RB phosphorylation and contain only a low level of cyclin D. In addition, although the Ras/extracellular signal-regulated kinase pathway promotes cell cycle progression in somatic cells, extracellular signal-regulated kinase signaling is dispensable for cell cycle progression in embryonic stem cells. Last, p53 is an important check point to induce cell apoptosis in somatic cells, whereas ES cells lack such a checkpoint (12). Until now, the only known regulator controlling the cell cycle of embryonic stem cells is the phosphorylation status of Akt at Ser⁴⁷³, which is activated by PI3K and is not regulated by mitogen stimulation (13, 14).

We investigated the structure-function relationship of Nanog in a series of studies. Based on these results, Nanog is divided into the N-terminal domain, DNA binding homeodomain, C-terminal domain 1, tryptophan repeat (WR) domain, and C-terminal domain 2 (CD2) (Fig. 1A). We first demonstrated that Nanog is a transcription activator possessing two strong transactivators, WR and CD2 (15, 16). Reporter assays demonstrated that deletion of WR had little effect on transcription activity of Nanog, whereas removal of CD2 reduced its activity severely, suggesting that WR is dispensable for Nanog function (15, 16). Consistently, we demonstrated that CD2, apparently endowed by a few critical aromatic residues with its strong transactivation activity, is required for Nanog to mediate LIF-independent ES cell self-renewal (17). Unexpectedly, we recently proved that WR plays an important role in Nanog-mediated LIF-independent ES cell self-renewal despite its apparent lack of contribution to Nanog transcription activity.³ The detailed function of other domains, especially the N-terminal domain and C-terminal domain 1, are largely unknown.

FIGURE 1.

Nanog mutants lacking the C-terminal pentapeptide of the WR domain slow down cell proliferation, whereas they maintain pluripotency. A, schematic illustration of full-length Nanog and its WR-truncated mutant Nanog(WR−). B, sequences of wild type WR and its mutants, W2W3×10, W3W2×10, WR×2, and 9W, in which repeated units are highlighted. C, morphology of ES cells constitutively expressing none, Nanog, Nanog(W2W3×10), Nanog(W3W2×10), Nanog(WR×2), Nanog(9W), and Nanog(WR−), after being cultured with or without LIF for 5 days. D, Western blot analysis of mock, Nanog, Nanog(W2W3×10), Nanog(W3W2×10), Nanog(WR×2), Nanog(9W), and Nanog(WR−) ES cells in the presence of LIF. Nanog and its mutants were detected with anti-Nanog antibody and normalized by endogenous β-actin with its specific antibody. The molecular weight of endogenous Nanog is indicated on the right. E, growth curves of mock, Nanog, Nanog(W2W3×10), Nanog(W3W2×10), Nanog(WR×2), Nanog(9W), and Nanog(WR−) ES cells in the presence of LIF for 4 days. Cell numbers were counted every day and shown as mean ± S.E. (n = 6, p < 0.01). F, quantitative PCR analysis of pluripotent genes (nanog, pou5f1, and zfp42) of ES cells in C. The relative expression levels of nanog (containing both endogenous and exogenous expression), pou5f1, and zfp42 were measured as mean ± S.E. (n = 6, p < 0.01), and all of the relative mean values of mock ES cells were arbitrarily designated as 1. The transcript level of β-actin was used as an internal reference. G–K, quantitative PCR analysis of differentiated and pluripotent genes in mock, Nanog, Nanog(9W), and Nanog(W2W3×10) ES cells for 0 days (EB0) or 5 days (EB5) after EB-induced differentiation. The relative expression levels of lamininB1 (endoderm marker; G), islet-1 (ectoderm marker; H), t (mesoderm marker; I), pou5f1 (pluripotent marker; J), and nanog (containing both endogenous and exogenous expression; K) were measured as mean ± S.E. (n = 3, p < 0.05), and all of the relative mean values of mock ES cells (EB0) were arbitrarily designated as 1. The transcript level of β-actin was used as an internal reference. IB, immunoblot.

Nac1 is a BTB domain-containing protein related to Drosophila bric-a-brac/tramtrack, which prevents inappropriate neural gene expression (18, 19). Recent studies revealed that Nac1 is a protein-interacting partner of Nanog and may participate in a regulatory network for sustaining pluripotency (20, 21).

In this report, we describe our findings that Nanog interacts with Nac1 through a pentapeptide WNAAP unit to regulate the proliferation of mouse embryonic stem cells via the ERas/PI3K/Akt pathway but not pluripotency.

MATERIALS AND METHODS

Plasmids

All of the WR mutants, W2W3×10, W3W2×10, WR×2, 9W, and W2W3×10mu, were substituted for the Nanog WR domain to construct Nanog(W2W3×10), Nanog(W3W2×10), Nanog(WR×2), Nanog(9W), and Nanog(W2W3×10mu), respectively. These five Nanog WR mutants, wild type Nanog, and WR-truncated mutant Nanog(WR−), with or without the C-terminal FLAG tag, were subcloned into pPyCAGIP (generously provided by Dr. Chambers) by XhoI and NotI sites. N-terminal Myc-tagged Nanog, Nac1, and Zfp281 were subcloned into pCBA-hrGFP (kindly provided by Dr. Kim) or pCAG-IRES-Neo (generously provided by Dr. Niwa) by XhoI and NotI sites. WNAAP and 2×WNAAP were ligated into EcoRV site at the C-terminal end of FLAG-Gal4DBD in pCR3.1 (Invitrogen) to construct FLAG-Gal4DBD-WNAAP and FLAG-Gal4DBD-2×WNAAP.

Cell Culture

HEK293T cells were cultured in Dulbecco's modified Eagle's medium (Hyclone) supplemented with 10% fetal bovine serum (Invitrogen) and antibiotics (penicillin and streptomycin; 100 μg/ml). CGR8 ES cells were cultured on 0.1% gelatin-coated substrates in ES medium consisting of Glasgow minimum essential medium (Sigma) supplemented with 20% fetal bovine serum (Invitrogen), 100 mm nonessential amino acids (Invitrogen), 0.55 mm β-mercaptoethanol (Invitrogen), 2 mm l-glutamine (Invitrogen), and 1,000 units/ml human recombinant LIF (Chemicon). The final concentration of DMSO (Sigma) or LY294002 (Sigma) used in inhibition of Akt phosphorylation was 10 μm.

Co-immunoprecipitation

3 μg of pCBA-hrGFP-c-Myc-Nanog, pCBA-hrGFP-c-Myc-Zfp281, or pCBA-hrGFP-c-Myc-Nac1 together with/without 1 μg of pPyCAGIP-Nanog-FLAG, were co-transfected with 1 μg of pPyCAGIP, pPyCAGIP-Nanog-FLAG, pPyCAGIP-Nanog(WR−)-FLAG, pPyCAGIP-Nanog(W2W3×10)-FLAG, pPyCAGIP-Nanog(9W)-FLAG, pPyCAGIP-Nanog(W2W3×10mu)-FLAG, pCR3.1, pCR3.1-FLAG-Gal4DBD, pCR3.1-FLAG-Gal4DBD-WNAAP, or pCR3.1-FLAG-Gal4DBD-2xWNAAP to HEK293T cells cultured in a well of a 6-well plate by using the calcium phosphate method. 48 h after transfection, cells were washed two times with PBS buffer and lysed by 200 μl of TNE buffer (100 mm Tris, pH 7.5, 150 mm NaCl, 0.5% Nonidet P-40, 1 mm EDTA) plus 5 μl of protease inhibitor mixture (Sigma). Whole cell lysates were cleared by centrifugation at 15,000 × g for 5 min at 4 °C. A volume of 40 μl of cleared whole cell lysates was reserved for Western blot analysis, and the remaining supernatants were transferred to a new 1.5-ml tube containing 15 μl of anti-FLAG-conjugated agarose beads (Sigma) equilibrated by TNE buffer before use. After rocking for 2 h, at 4 °C, the anti-FLAG beads were then washed five times with TNE buffer. After all of the TNE buffer was removed, the beads were boiled for 5 min in 40 μl of 1× SDS loading buffer (with 5% β-mercaptoethanol) to elute the bound antibodies and antigens. After centrifuging at 12,000 × g for 5 min, supernatants (8 μl) were detected by immunoblotting.

Western Blot and Antibodies

For Western blot analysis of mouse ES cells, the cells were lysed in radioimmune precipitation lysis buffer (50 mm Tris-HCl, pH 8.0, 150 mm NaCl, 0.5% sodium deoxycholate, 1% Triton X-100, 0.1% SDS). The prepared samples were loaded on 10% SDS-PAGE and then blotted onto polyvinylidene difluoride membrane for detection by using specific first antibody and second antibody. The antibodies we used included the following: Akt antibody (catalog number 9272; Cell Signaling), pAkt(Ser⁴⁷³) antibody (catalog number ab27773; Abcam), Nanog antibody (catalog number ab21603; Abcam), β-actin antibody (catalog number A2066; Sigma), FLAG antibody (catalog number 2368; Cell Signaling), c-Myc antibody (catalog number K03145R; Biodesign), alkaline phosphatase-labeled anti-rabbit IgG antibody (catalog number A3687; Sigma), and peroxidase-labeled anti-rabbit IgG antibody (catalog number 074-1506; KPL).

Measurement of Expansion in Cell Numbers

We plated ES cells at 2 × 10⁵ cells/well in 24-well plates and cultured them for 4 days in ES medium. The cells were trypsinized and counted with a blood count plate (sample = 6) every day.

Early Apoptotic Marker Detection

ES cells cultured in ES medium were trypsinized, washed in PBS, and treated with the ApopNexin Annexin V fluorescein isothiocyanate apoptosis kit (catalog number APT750; Chemicon) according to the manufacturer's instructions and analyzed by a FACScalibur flow cytometer (BD Biosciences).

Cell Cycle Distribution Analysis

ES cells cultured in ES medium were harvested, fixed in cold 70% ethanol, washed in PBS, treated with 50 μg/ml propidium iodide (Sigma) and 5 μg/ml RNase A (Sigma) at 37 °C for 30 min, and analyzed by a FACScalibur flow cytometer (BD Biosciences).

BrdUrd Pulse Analysis

ES cells with 70% confluence were pulse-labeled with 10 μm BrdUrd (Sigma) for 30 min. After they were washed twice, labeled ES cells were cultured for 0, 7, and 14 h in ES medium. Cells were trypsinized, washed in PBS, and fixed in cold 70% ethanol for 30 min. Cell pellets were then incubated in 4 n HCl for 30 min and washed in 0.1 m sodium borate (pH 8.5) followed by a PBS wash. Cells were then resuspended in 50 μl of PBS-TB (PBS containing 0.5% Tween 20 and 0.5% bovine serum albumin) and a 1:250 dilution of anti-BrdUrd antibody (BU-33; Sigma) for 30 min at room temperature. After two washes in PBS-TB, cells were incubated with fluorescein isothiocyanate-conjugated goat anti-mouse IgG secondary antibody (catalog number sc-2010; Santa Cruz Biotechnology, Inc., Santa Cruz, CA) with a 1:100 dilution in PBS-TB, 30 mm at room temperature. Cells were again washed twice in PBS-TB and then resuspended in PBS containing 5 μg/ml propidium iodide (Sigma) and 1 μg/ml RNase A and analyzed on a FACScalibur flow cytometer (BD Biosciences).

Constitutive Expression of Transgenes in Mouse ES Cells

CGR8 ES cells cultured in ES medium were plated in 3.5-cm dishes and transfected with 2 μg of pPyCAGIP, pPyCAGIP-Nanog, pPyCAGIP-Nanog(W2W3×10), pPyCAGIP-Nanog(W3W2×10), pPyCAGIP-Nanog(WR×2), pPyCAGIP-Nanog(9W), pPyCAGIP-Nanog(WR−), or pPyCAGIP-Nanog(W2W3×10mu) by Lipofectamine 2000 (Invitrogen). 24 h after transfection, the cells were passaged by 1:50 and plated in new 6-cm dishes for puromycin (2 μg/ml; Invitrogen) selection. 10 days later, a single clone was picked up and expanded in a 12-well plate and then further identified by Western blot analysis and real time reverse transcription (RT)-PCR analysis. Through detection by anti-Nanog antibody and normalization by endogenous β-actin, different molecular weight and band signal could be used to discriminate expression levels of the wild type versus mutant Nanog in stable cell lines. The total Nanog mRNA level (containing both endogenous and exogenous expression) revealed by real time RT-PCR would also help for this discrimination. The positive clone was named after its transgene.

For establishment of ES cell lines carrying both transgenic Nanog and Myc-Nac1, 2 μg of pCAG-IRES-Neo or pCAG-Myc-Nac1-IRES-Neo was transfected into mock or Nanog ES cells, which were then selected by both puromycin (1 mg/ml) and G418 (500 μg/ml) and identified by Western blot with anti-Nanog and anti-Myc antibody.

Real Time RT-PCR Analysis

2 μg of total RNA was reverse transcribed in a final volume of 20 μl, as previously described (22). Real time RT-PCRs were undertaken using the real time PCR master mix (SYBR GREEN) reagent kit (TOYOBO), according to the manufacturer's protocol. PCR was performed in a 15-μl total volume for 40 cycles. The primers were as follows: β-actin, forward (5′-agtgtgacgttgacatccgt-3′) and reverse (5′-tgctaggagccagagcagta-3′); nanog, forward (5′-ctcaagtcctgaggctgaca-3′) and reverse (5′-tgaaacctgtccttgagtgc-3′); pou5f1, forward (5′-ggaagccgacaacaatgagaa-3′) and reverse (5′-tcgggcacttcagaaacatg-3′); zfp42, forward (5′-cagccagaccaccatctgtc-3′) and reverse (5′-gtctccgatttgcatatctcctg-3′); lamininB1, forward (5′-ttggatagcatcaccaagtatttcc-3′) and reverse (5′-ctcgcgtgagggcagact-3′); islet-1, forward (5′-tgtggacattactccctcttacagat-3′) and reverse (5′-tgggagacatgggcgatccacc-3′); t, forward (5′-atgcggacaattcatctgctt-3′) and reverse (5′-caggcactccgaggctagac-3′); ERas, forward (5′-ccaagacgcggcaaggt-3′) and reverse (5′-cctcctgggccctctga-3′).

Statistical Analysis

All of the data from quantitative RT-PCR, cell number count, and cell cycle distribution were shown as mean ± S.E. and analyzed by t test. p < 0.05 or p < 0.01 was considered to indicate statistical significance.

RESULTS

Nanog with Super-WR Decouples Proliferation from Pluripotency

We reported previously that Nanog possesses two potent transactivation domains, WR and CD2 (15, 16) (Fig. 1A). The WR domain, a 10-tandem pentapeptide repeat starting with a tryptophan residue, has been demonstrated to be indispensable for Nanog to mediate pluripotency (23, 24). Furthermore, we have uncovered the mechanism of transactivity of the WR domain, based on which we designed a superactive WR mutant, W2W3×10 (Fig. 1B).³

To test the effect of this superactive mutant on ES cells, we replaced the Nanog WR domain with this mutant to construct Nanog(W2W3×10) (Fig. 1B) and transfected mouse ES cells with Nanog and Nanog(W2W3×10), respectively. After stable expression clones were selected, we verified their expression level by immunoblotting (Fig. 1D). When culturing mock, Nanog, and Nanog(W2W3×10) ES cells in the presence of LIF for 5 days, Nanog(W2W3×10) ES cells formed smaller colonies than both mock and Nanog ES cells (Fig. 1C). This phenomenon suggests that Nanog(W2W3×10) would attenuate the proliferation of ES cells. To confirm it, we counted the cell number for a continuous 4 days in the presence of LIF. The slower proliferation rate of Nanog(W2W3×10) cells was consistent with its smaller colony size (Fig. 1E). Furthermore, in order to investigate whether this proliferation decrease is caused by alteration of Nanog-mediated pluripotency, we cultured these three cell lines both in the absence and presence of LIF for 5 days. As shown in Fig. 1C, Nanog and Nanog(W2W3×10) ES cells displayed pluripotent morphology with different colony sizes, whereas mock ES cells exhibited flat morphology, dispersed growth, and other differentiated features. The pluripotent state of these ES cells was also confirmed by quantitative RT-PCR analysis of two pluripotent markers, pou5f1 and zfp42. As shown in Fig. 1F, both pou5f1 and zfp42 maintained a high level in Nanog(W2W3×10) ES cells, medium level in Nanog ES cells, and low level in mock cells when cultured without LIF. These results indicated that duplicating WTNPTWSSQT 10 times in Nanog could decouple ES cell proliferation from pluripotency.

The obvious difference between W2W3×10 and wild type WR is the length of total WR domain (50 amino acids for wild type WR versus 100 amino acids for W2W3×10). To investigate whether the longer WR domain is the key to mediate slow proliferation, we established mouse ES cell lines constitutively expressing Nanog mutant containing two tandem WRs (Fig. 1B) and then confirmed the protein expressed by these cells (Fig. 1D). Both colony sizes and cell numbers (Fig. 1, C and E) clearly showed that ES cells expressing Nanog with double WR domains proliferated at a rate similar to that of mock- or wild type Nanog-expressed cell lines. When cultured in the absence of LIF, Nanog(WR×2) cells exhibited pluripotent morphology and high expressions of pluripotent markers, such as pou5f1 and zfp42, that are comparable with the cells expressing wild type Nanog (Fig. 1, C and E). These results suggest that the doubling of wild type WR length in Nanog has no effects to reduce proliferation and pluripotency.

Besides the length, the major difference distinguishing W2W3×10 from wild type WR is positional alteration between WTNPT and WSSQT. To exclude the possibility that relative sequence alteration slows down ES proliferation, we made another artificial repeat W3W2×10 (Fig. 1B) to substitute wild type WR and stably transfected it into mouse ES cells. The protein level of this mutant is shown in Fig. 1D. When cultured in the presence of LIF, Nanog(W2W3×10) and Nanog(W3W2×10) cell lines proliferate indistinguishably, as indicated by both colony size (Fig. 1C) and cell number (Fig. 1E). Even in absence of LIF, these two cell lines show similar morphology (Fig. 1C) and expression level of pluripotent marker (Fig. 1F). These findings suggested that either Nanog(W2W3×10) or Nanog(W3W2×10) can slow down proliferation of ES cells without losing pluripotency.

Deleting the C-terminal Pentapeptide of Nanog WR Domain Slows Down Cell Proliferation without Losing Pluripotency

By comparing sequences among W2W3×10, W3W2×10, and WR×2, we found their major difference lay on the two pentapeptides (WNGQP and WNAAP) at the C-terminal end of WR. A previous report comparing the Nanog protein sequence in four species showed that the latter one (WNAAP) was more conserved in these two pentapeptides (25), suggesting that WNAAP may play a more significant role than other repeating units. To test this possibility, we made another two Nanog mutant constructs lacking the last pentapeptides (Nanog 9W) or the entire WR domain (Nanog(WR−)) and selected their stable expression cell lines from ES cells (Fig. 1, A and B). After we confirmed their expression levels (Fig. 1D), we found that Nanog(9W) ES cells grew slowly and formed small colonies, whereas the cells expressing Nanog lacking the entire WR appeared to proliferate normally (Fig. 1, C and E). Interestingly, the proliferation rate of Nanog(9W) ES cell lines ranked between the ES cells expressing wild type Nanog and Nanog (W2W3×10) based on data on colony sizes (Fig. 1C) and cell number (Fig. 1E). These findings implicated that WNAAP but not the other nine pentapeptides of WR plays a specific role in retarded proliferation in ES cells.

Furthermore, when we withdrew LIF, the Nanog(9W) ES cell line exhibited pluripotent morphology (Fig. 1C) and marker gene expressions (Fig. 1E) comparable with those of the wild type Nanog cell line, whereas ES cells expressing Nanog lacking the entire WR showed significant differentiation (Fig. 1, C and E). These data suggest that the last WNAAP has little effect on pluripotency and just regulates proliferation. Indeed, upon embryonic body (EB) formation, all ES cell lines that express mock, Nanog, Nanog(9W), and Nanog(W2W3×10) were able to express differentiation markers, such as lamininB1, islet-1, and t (representing three germ layers) and down-regulated pou5f1 (indicating pluripotency) (Fig. 1, G–K), which indicated that ES cells expressing Nanog and its mutants kept the potential to differentiate, as mock ES cells did. In fact, compared with mock ES cells, the Nanog, Nanog(9W), and Nanog(W2W3×10) cell lines actually suppressed differentiation to certain levels as evidenced by expressing higher pou5f1 and lower differentiated markers at 5 days EB (Fig. 1, G–K, EB5), illustrating indistinguishable functions among Nanog and its two mutants in preventing EB-induced differentiation. Taking all of the data together, we could conclude that the C-terminal pentapeptide of the WR domain plays a critical role in regulating ES cell proliferation, not sustaining pluripotency.

Both Mutants Lacking WNAAP Shift Cell Cycle to Longer G₁ and Shorter S Phases with Lower pAkt(Ser⁴⁷³)

The smaller colony sizes and lower cell numbers we observed in Nanog(9W) and Nanog(W2W3×10) could be due to cell apoptosis. To certainly exclude this possibility, the early stage apoptotic marker, phosphatidylserine, was measured. As shown in Fig. 2A, we did not see significant differences in apoptotic populations (Annexin⁺PI⁻) among the cell lines we had tested thus far (mock, 2.55%; Nanog, 3.67%; Nanog(9W), 1.76%; Nanog(W2W3×10), 1.32%). Thus, the attenuated proliferations of Nanog(9W) and Nanog(W2W3×10) ES cells were not contributed to by apoptosis. To further understand the apparent slowdown of growth in Nanog(9W) and Nanog(W2W3×10) cells, we analyzed the cell cycle phase distribution of the above four cell lines (Fig. 2B). Statistical analyses indicated that Nanog(W2W3×10) and Nanog(9W) ES cells have significant longer G₁ phase (∼12 and ∼6%, respectively) and shorter S phase (∼12 and ∼6%, respectively) compared with Nanog or mock ES cells (Fig. 2B). Furthermore, the longer G₁ and shorter S phases for Nanog(W2W3×10) ES cells compared with those of Nanog(9W) (Fig. 2B) are quite consistent with the lower growth rate observed for Nanog(W2W3×10) ES cells. In addition to counting cell number, we also pulse-labeled four cell lines with BrdUrd and examined cell cycle progression through flow cytometry (Fig. 2C). As shown, 7 h after BrdUrd pulse labeling, cells from all four cell lines mainly accumulated at 2 n (40–45%) and 4 n (35–40%) based on DNA content, indicating that mitosis was proceeding (Fig. 2C). However, 14 h after BrdUrd pulse labeling, significant different distributions were found in Nanog(9W) and Nanog(W2W3×10) cell lines (2 n, ∼25 and ∼40%; 2–4 n, ∼25 and ∼20%; 4 n, ∼50 and ∼40%, respectively) compared with that from both mock and Nanog cell lines (2 n, ∼5%; 2–4 n, ∼30%; 4 n, ∼65%). These data suggest that cells in Nanog(9W) and Nanog(W2W3×10) undergo arrest at G₁/S transition (Fig. 2C).

FIGURE 2.

Both mutants lacking WNAAP shift cell cycle to longer G₁ and short S phases with lower pAkt(Ser⁴⁷³). A, mock, Nanog, Nanog(9W), and Nanog(W2W3×10) ES cells were flow cytometrically analyzed for membrane integrity and phosphatidylserine accessibility by staining with propidium iodide and fluorescein isothiocyanate-conjugated annexin V (Annexin V-FITC). The percentage of each population (annexin V⁺PI⁻, annexin V⁺PI⁺, or annexin V⁻PI⁺ cells) is indicated. The data were analyzed and represented by FCS Express software. B, flow cytometric analysis of cell cycle distribution of mock, Nanog, Nanog(9W), and Nanog(W2W3×10) ES cells by PI staining. The data were analyzed and represented by FCS Express software. The percentages of G₁, S, and G₂-M phases in each cell line were calculated as mean ± S.E. (n = 4, p < 0.05). C, expression level of Akt and phosphorylated Akt at Ser⁴⁷³ of mock, Nanog, Nanog(9W), and Nanog(W2W3×10) ES cells cultured with LIF. Akt and pAkt(Ser⁴⁷³) were detected by anti-Akt and anti-pAkt(Ser⁴⁷³) antibodies, respectively, and normalized by endogenous β-actin with its specific antibody. All of the bands were quantified by QuantityOne (Bio-Rad). D, mock, Nanog, Nanog(9W), and Nanog(W2W3×10) ES cells were pulse-labeled with BrdUrd for 30 min and harvested at the time intervals indicated on the left. The cells were then stained with anti-BrdUrd antibody and propidium iodide and cytometrically analyzed to measure DNA synthesis (BrdUrd; y axis) and DNA content (propidium iodide; x axis). The data were analyzed and represented by FCS Express software. E, growth curves of mock, Nanog, Nanog(9W), and Nanog(W2W3×10) ES cells cultured with LIF plus 10 μm DMSO (left) or 10 μm LY294002 (right) for 4 days. Cell numbers were counted every day and shown as mean ± S.E. (n = 6, p < 0.01). F, the ratio of cell number in the presence of LY294002 versus those of control at each day in E was calculated and plotted. G, expression level of Akt and phosphorylated Akt at Ser⁴⁷³ of mock, Nanog, Nanog(9W), and Nanog(W2W3×10) ES cells cultured with LIF plus 10 μm DMSO (left) or 10 μm LY294002 (right). Akt and pAkt(Ser⁴⁷³) were detected by anti-Akt and anti-pAkt(Ser⁴⁷³) antibodies, respectively, and normalized by endogenous β-actin with its specific antibody. All of the bands were quantified by QuantityOne (Bio-Rad). IB, immunoblot.

To identify a potential mechanism associated with this cell cycle delay, we examined the phosphorylation of Akt at Ser⁴⁷³, an indicator of PI3K/Akt activity, known as an important regulatory pathway for the G₁/S transition during ES cell proliferation (13,14). As shown in Fig. 2D, overexpression of Nanog in ES cells appears to have elevated Akt phosphorylation at Ser⁴⁷³ (∼1.3-fold), whereas overexpression of Nanog (W2W3×10) or Nanog(9W) significantly reduced this phosphorylation by ∼50 or ∼75%, respectively (Fig. 2D). In contrast, the amounts of Akt expression were nearly indistinguishable in all four ES cell lines (Fig. 2D). Compared with both mock and Nanog ES cells, delayed G₁/S transitions in Nanog(W2W3×10) or Nanog(9W) ES cells, corresponding to reduction of cell proliferation, was consistent with their lower phosphorylation of Akt at Ser⁴⁷³. These results suggested that the reduction in PI3K/Akt activity may be responsible for the observed slowdown of G₁/S transition for ES cells expressing Nanog mutants lacking WNAAP.

To further test this idea, we cultured these four ES cell lines in the presence of LY294002, a specific inhibitor of PI3K, which could reduce Akt phosphorylation at Ser⁴⁷³. As indicated in Fig. 2E, all of the ES cell lines treated with LY294002 grew more slowly than control groups treated with DMSO as expected; Nanog ES cells grow faster in all four cell lines cultured with LY294002 and exhibited some resistance to this inhibitor, which was consistent with its highest level of phosphorylation of Akt at Ser⁴⁷³ (Fig. 2D); and the calculated ratio of cell number in the presence of LY294002 versus those of control at each day in Fig. 2F suggested that Nanog confers partial resistance to LY294002 inhibition, whereas Nanog(W2W3×10) and Nanog(9W) ES cells with lower phosphorylation of Akt at Ser⁴⁷³ are in fact more sensitive to this drug. Western blot analysis of Akt phosphorylation essentially confirmed this observation (Fig. 2G). These results further demonstrated that Nanog mutants lacking WNAAP reduce activity of PI3K/Akt, which in turn attenuates G₁/S transition and cell growth.

WNAAP Mediates Nanog-Nac1 Interaction

The experimental evidence presented so far strongly suggests that the last pentapeptide, WNAAP, mediates ES cell proliferation. Recently, the WR domain has been shown to be involved in transactivation (15, 16), dimerization (23, 24), and interaction for Sal4, Nac1, Zfp281, and Dax1 (24). Since we have found that the WR mutant devoid of the last two pentapeptides has transcriptional activity similar to that of wild type WR,³ we focus on dimerization and interaction with other proteins. First, we tested whether Nanog(W2W3×10) or Nanog(9W) could dimerize with Nanog. The results revealed that Nanog(W2W3×10) or Nanog(9W) could form a dimer with Nanog, as did wild type Nanog (Fig. 3, A and D). As negative control, Nanog(WR−) dimerized with Nanog obviously weakly (see Fig. 5, A and D). These observations excluded the possibility that WNAAP participates in Nanog dimerization. Then we tested if WNAAP mediates interaction with other pluripotency sustaining proteins, in which Nac1 and Zfp281 were the most likely candidates, because knockdown of either would decrease mouse ES cell proliferation (20). Due to the profound reduction of growth rate for Nanog(W2W3×10) ES cells, we tested first whether Nanog(W2W3×10) could interact with Nac1 or Zfp281. After immunoprecipitation and immunoblot, the results revealed indistinguishable interactions of Nanog, Nanog(W2W3×10), and Nanog(WR−) with Zfp281 (Fig. 3C), whereas Nanog(W2W3×10) interacted weakly with Nac1 (Fig. 3B). Indeed, Nanog(9W) shows the same ability as Nanog(WR−) to bind Nac1(Fig. 3E), which is consistent with its intermediate Akt phosphorylation and proliferation rate between Nanog and Nanog(W2W3×10) (Figs. 1E and 2D). Thus, Nac1 could be the interaction partner for WNAAP. In our experiments, Nanog(WR−) could interact with Nanog weakly and form a heterodimer with Zfp281 as does wild type Nanog, which conflicts with previous reports (24), and these result might be caused by the enhanced overexpression of the CAG promoter we used.

FIGURE 3.

WNAAP mediates Nanog-Nac1 interaction. A, Myc-Nanog (labeled at the top) was co-transfected with vector, Nanog-FLAG, Nanog(WR−)-FLAG, or Nanog(W2W3×10)-FLAG (indicated at the bottom) into HEK293T cells. B, Myc-Nac1 (labeled at the top) was co-transfected with vector, Nanog-FLAG, Nanog(W2W3×10)-FLAG, or Nanog(WR−)-FLAG (indicated at the bottom) into HEK293T cells. C, Myc-Zfp281 (labeled at the top) was co-transfected with vector or Nanog-FLAG, Nanog(W2W3×10)-FLAG, or Nanog(WR−)-FLAG (indicated at the bottom) into HEK293T cells. D, Myc-Nanog (labeled at the top) was co-transfected with vector, Nanog-FLAG, Nanog(WR−)-FLAG, or Nanog(9W)-FLAG (indicated at the bottom) into HEK293T cells. E, Myc-Nac1 (labeled at the top) was co-transfected with vector, Nanog-FLAG, Nanog(9W)-FLAG, or Nanog(WR−)-FLAG (indicated at the bottom) into HEK293T cells. Whole-cell lysates were analyzed with anti-Myc antibody (top). After immunoprecipitation by anti-FLAG-conjugated agarose beads, the samples were detected by anti-FLAG or anti-Myc antibody (two lower panels). All of the bands were quantified by QuantityOne (Bio-Rad). F, schematic illustration of FLAG-tagged Gal4DBD, Gal4DBD-W10, and Gal4DBD-2×W10. G, Myc-Nac1 (labeled at the top) was co-transfected with vector, FLAG-Gal4DBD, FLAG-Gal4DBD-W10, or FLAG-Gal4DBD-2xW10 (indicated at the bottom) into HEK293T cells. Whole cell lysates were analyzed with anti-Myc antibody (top). After immunoprecipitation by anti-FLAG-conjugated agarose beads, the samples were detected by anti-FLAG or anti-Myc antibody (two lower panels). All of the bands were quantified by QuantityOne (Bio-Rad). IB, immunoblot; IP, immunoprecipitation.

FIGURE 5.

Nanog and Nac1 synergistically up-regulate ERas expression in ES cells, and both Nanog mutants lacking WNAAP might down-regulate ES cell proliferation in a dominant negative way. A, quantitative PCR analysis of ERas mRNA level in mock, Myc-Nac1, Nanog, and Nanog + Myc-Nac1 ES cells. The relative expression levels of ERas were measured as mean ± S.E. (n = 6, p < 0.01), and the relative mean value of mock ES cells was arbitrarily designated as 1. The transcript level of β-actin was used as an internal reference. B, growth curves of mock, Myc-Nac1, Nanog, and Nanog + Myc-Nac1 ES cells in the presence of LIF for 4 days. Cell numbers were counted every day and are shown as mean ± S.E. (n = 6, p < 0.01). C, Western blot analysis of mock, Myc-Nac1, Nanog, and Nanog + Myc-Nac1 ES cells in the presence of LIF. Nanog, Myc-Nac1, Akt, and pAkt(Ser⁴⁷³) were detected with anti-Nanog, anti-Myc, anti-Akt, and anti-pAkt(Ser⁴⁷³) antibodies, respectively, and normalized by endogenous β-actin with its specific antibody. The molecular weight of Myc-Nac1 is indicated on the right. All of the quantitative bands were calculated by QuantityOne (Bio-Rad). D, both Myc-Nac1 and Nanog-FLAG (labeled at the top) were co-transfected with vector, Nanog(WR−), Nanog(9W), Nanog(W2W3×10), and Nanog(W2W3×10mu) (indicated at the bottom) into HEK293T cells. Whole cell lysates were analyzed with anti-Nanog or anti-Myc antibody (two upper panels). After immunoprecipitation by anti-FLAG-conjugated agarose beads, the samples were detected by anti-FLAG or anti-Myc antibody (two lower panels). Molecular weights of Nanog-FLAG, Nanog(WR−), Nanog(9W), Nanog(W2W3×10), and Nanog(W2W3×10mu) are indicated. All of the quantitative bands were calculated by QuantityOne (Bio-Rad). E, quantitative PCR analysis of ERas mRNA level in mock, Nanog(WR−), Nanog(9W), Nanog(W2W3×10), and Nanog(W2W3×10mu) ES cells. The relative expression levels of ERas were measured as mean ± S.E. (n = 6, p < 0.01), and the relative mean value of mock ES cells was arbitrarily designated as 1. The transcript level of β-actin was used as an internal reference. F, schematic illustration of the proposed mechanism by which WNAAP, the C-terminal pentapeptide of the Nanog WR domain, regulates mouse ES cell proliferation but not pluripotency. WNAAP mediates tight interaction between Nanog and Nac1, which controls ERas expression, which in turn regulates ES cell proliferation through PI3K/Akt activity. Furthermore, Nanog mutant lacking WNAAP might down-regulate ES cell proliferation through interfering with the interaction between Nanog and Nac1. IB, immunoblot; IP, immunoprecipitation.

To further support the idea of WNAAP mediating interaction between Nanog and Nac1, we fused WNAAP or two tandem WNAAP to the C terminus of the Gal4 DNA binding domain (Fig. 3F) and tested whether these two fusion proteins interact with Nac1. Pull-down assays showed that Gal4DBD-WNAAP or -2×WNAAP endows Nac1 binding with increasing affinity, whereas Gal4DBD alone does not (Fig. 3G). Thus, these results strongly suggest that WNAAP, the C-terminal pentapeptide in WR, is sufficient to mediate the interaction of Nanog with Nac1.

Pentapeptide WNAAP Can Restore Normal Proliferative Potential to Nanog(W2W3×10) Mutant

To further confirm the function of WNAAP, we then replaced the 10th and 20th pentapeptides (WSSQT) in Nanog(W2W3×10) with WNAAP and tested whether this replacement can restore normal function to Nanog(W2W3×10mu) (Fig. 4A). First, we performed co-immunoprecipitation and immunoblot to verify that this WNAAP replacement rescued its interaction with Nac1, nearly comparably as wild type Nanog did (Fig. 4B). To evaluate the role of this replacement mutant on ES cells, we introduced Nanog(W2W3×10mu) into ES cells, and enhanced expression was detected in Fig. 4D. Relative to low Akt phosphorylation at Ser⁴⁷³ in Nanog(W2W3×10) ES cells, Nanog(W2W3×10mu) elevated Akt phosphorylation at Ser⁴⁷³ up to the phosphorylation level in mock ES cell line (Fig. 4D). Consistent with the enhancement of Akt phosphorylation, analysis of colony size (Fig. 4C) and cell number (Fig. 4E) also demonstrated that Nanog(W2W3×10mu) ES cells grew in a fashion similar to the mock or Nanog ES cell line, in contrast to the Nanog(W2W3×10) ES cell at a very low rate.

FIGURE 4.

The pentapeptide WNAAP can restore normal proliferative potential to the Nanog(W2W3×10) mutant. A, sequences of wild type WR and its mutants, W2W3×10 and W2W3×10mu, in which repeated unit and substitutive WNAAP were highlighted. B, Myc-Nac1 (labeled at the top) was co-transfected with vector or Nanog-FLAG, Nanog(W2W3×10mu)-FLAG, or Nanog(WR−)-FLAG (indicated at the bottom) into HEK293T cells. Whole cell lysates were analyzed with anti-Myc antibody (top). After immunoprecipitation by anti-FLAG-conjugated agarose beads, the samples were detected by anti-FLAG or anti-Myc antibody (two lower panels). All of the bands were quantified by QuantityOne (Bio-Rad). C, morphology of ES cells constitutively expressing none, Nanog, Nanog(W2W3×10), and Nanog(W2W3×10mu), after they were cultured with or without LIF for 5 days. D, Western blot analysis of mock, Nanog, Nanog(W2W3×10), and Nanog(W2W3×10mu) ES cells in the presence of LIF. Nanog and its mutants, Akt, and pAkt(Ser⁴⁷³) were detected with anti-Nanog, anti-Akt, and anti-pAkt(Ser⁴⁷³) antibodies, respectively, and normalized by endogenous β-actin with its specific antibody. The molecular weight of endogenous Nanog is indicated on the right. All of the quantitative bands were calculated by QuantityOne (Bio-Rad). E, growth curves of mock, Nanog, Nanog(W2W3×10), and Nanog(W2W3×10mu) ES cells in the presence of LIF for 4 days. Cell numbers were counted every day and are shown as mean ± S.E. (n = 6, p < 0.01). F, quantitative PCR analysis of pluripotent genes (nanog, pou5f1, and zfp42) of ES cells in C. The relative expression levels of nanog (containing both endogenous and exogenous expression), pou5f1, and zfp42 were measured as mean ± S.E. (n = 6, p < 0.01), and all of the relative mean values of mock ES cells were arbitrarily designated as 1. The transcript level of β-actin was used as an internal reference. IB, immunoblot; IP, immunoprecipitation.

To further evaluate whether this replacement mutation affects the role of Nanog in sustaining ES cell pluripotency independently of LIF, we cultured the Nanog(W2W3×10mu) cell line in the absence of LIF. After 5 days, this cell line displayed pluripotent morphology as Nanog ES cells (Fig. 4C) and maintained a relatively high level pluripotent marker like the Nanog(W2W3×10) ES cell line (Fig. 4F). Based on this observation, we confirmed that the last pentapeptide, WNAAP, participates only in mediating cell proliferation, not pluripotency in ES cells.

Nanog and Nac1 Synergistically Up-regulate ERas Expression in ES Cells, and Both Nanog Mutants Lacking WNAAP Might Down-regulate ES Cell Proliferation in a Dominant Negative Way

ERas is specifically expressed by ES cells and has been shown to be crucial to mediate PI3K/Akt activity (14). Microarray analysis of transcription factor targets in whole genome scale, followed by chromatin immunoprecipitation, indicated that ERas could be regulated by Nanog (21). Together with our results shown here, we reasoned that Nanog and Nac1 might regulate transcription of ERas in ES cells. To test this hypothesis, we introduced Nanog and Nac1 into ES cells, either individually or in combination. The expressions of these transgenes were verified by Western blot (Fig. 5B). Quantitative RT-PCR analysis showed that ERas mRNA levels were slightly elevated (1.2–1.4-fold) in ES cells carrying either Nanog or Nac1 alone compared with mock cells (Fig. 5A), whereas they were more dramatically up-regulated in ES cells carrying both Nanog and Nac1 (up to ∼2.5-fold; Fig. 5A). Consistent with the up-regulated ERas, phosphorylated Akt at S473 also increased accordingly in either Nanog or Nac1 alone or co-expressed cell lines (Fig. 5D), and again, ES cells expressing both Nanog and Nac1 showed higher levels of phosphorylated Akt than those expressing Nanog or Nac1 alone (Fig. 5D). We then examined the proliferation rate of these cell lines. As shown in Fig. 5C, ES cells expressing both Nanog and Nac1 grew much faster than other cell lines, presumably due to higher levels of ERas and phosphorylated Akt. These findings indicated that ERas might be an immediate downstream effector for Nanog and Nac1 to mediate ES cell proliferation. In addition, it is implicated that an unknown mechanism makes ES cell proliferation tolerant of small changes in the ERas/PI3K/Akt cascade, which would explain the similar growth rate between ES cells carrying either Nanog or Nac1 alone and mock ES cells.

Since the mutant Nanog(9W) and Nanog(W2W3×10) kept the capability to dimerize with wild type Nanog, we reasoned that these mutants might interfere with wild type Nanog to bind its partner Nac1. We then co-transfected wild type Nanog and Nac1 to HEK293T cells and along with various Nanog WR mutants, as indicated in Fig. 5D. The interactions between wild type Nanog and Nac1 from various transfection combinations were examined by immunoprecipitation. As shown in Fig. 5D, Nanog-Nac1 interaction was reduced to ∼60 and ∼30%, respectively, when co-expressed with Nanog(9W) and Nanog(W2W3×10) compared with control. As expected, Nanog lacking the entire WR that presumably cannot bind wild type Nanog showed no effects of interference with the interaction (Fig. 5D). Surprisingly, Nanog(W2W3×10mu) with restored capability to bind Nac1 (Fig. 4D) showed little effect to alter the interaction (Fig. 5D). Nanog(9W) and Nanog(W2W3×10) can interfere with wild type Nanog to bind Nac1, so they might serve as dominant negative mutants upon expression in ES cells. We then expressed Nanog with various WR mutations in ES cells and examined the expression levels of ERas by quantitative RT-PCR. As shown in Fig. 5E, expressions of Nanog(9W) and Nanog(W2W3×10) in ES cells reduced ERas expression to ∼60 and ∼30%, respectively, compared with mock cells, although ES cells expressing Nanog(WR−) or Nanog(W2W3×10mu) showed little change upon ERas expression. The decreased levels of ERas transcript in Nanog(9W) and Nanog(W2W3×10) ES cells could explain the lower pAkt(S473) and reduced proliferation that we showed earlier, and also in Fig. 1E, ES cells expressing Nanog(WR−) growing at a normal rate could be explained by the normal ERas levels showed in Fig. 5E. Based on all of the results thus far, we could conclude that Nanog(9W) and Nanog(W2W3×10) are dominant negatives that reduce ES cell proliferation through down-regulation of ERas, an immediate downstream effector of the Nanog-Nac1 complex.

Taking all of our data together, we propose that ES cells engage separate mechanisms to govern cell proliferation and pluripotency. The interaction between WNAAP, C-terminal pentapeptide of the Nanog WR domain, and Nac1 contributes to the cell cycle control of ES cells through mediation of the ERas/PI3K/Akt pathway, whereas the Nanog mutant lacking WNAAP might slow down ES cell proliferation in a dominant negative way (Fig. 5F).

DISCUSSION

Nanog is the first in the class of transcription factors that can confer a LIF-independent self-renewal property to mouse ES cells in culture (7, 8). In vivo evidence suggests that Nanog plays a critical role in the inner cell mass formation of early embryo and the maintenance of germ line pluripotency (9). These properties have generated tremendous interest in the mechanisms through which Nanog can execute these functions both in vitro and in vivo. We have taken a structure-function approach to dissect the various components of Nanog protein and determine the related functional correlates. So far, we have been able to divide Nanog into five distinct domains, the N-terminal domain, the DNA binding homeodomain, C-terminal domain 1, WR, and CD2. Functionally, we were able to assign WR and CD2 as transactivators embedded in its C terminus (15, 16). Our further analysis has demonstrated that CD2 is the dominant transactivator and that its activity is absolutely required for Nanog-mediated LIF-independent ES cell self-renewal (17). We report here that WNAAP from the WR domain appears to bind to Nac1 and impact the cell cycle machinery rather than the pluripotency maintenance or self-renewal function of ES cells. Moreover, the interaction between endogenous Nanog and Nac1 appears to be critical to regulate ERas/PI3K/Akt pathway responsible for sustaining proliferation in mouse ES cells.

Our findings have several implications for stem cell biology. First, since the cell cycle of embryonic stem cells is not well characterized, our finding that cell cycle and self-renewal can be investigated separately may become an important tool to delineate the molecular circuits regulating ES cell cycle. Second, WR is uniquely conserved in Nanog throughout evolution. The fact that the last pentapeptide, WNAAP, is more conserved in Nanog signifies that it may participate in a critical function. Our finding that it interacts with Nac1 and impacts the cell cycle may reveal more regulatory circuits beyond ERas/PI3K/Akt pathways. Last, if WNAAP plays such an important role, one may argue that the other nine pentapeptides may have similar or different roles in ES cell self-renewal or other physiological property. Further work is needed to further delineate their contribution to Nanog function.