Self-Renewal Versus Lineage Commitment of Embryonic Stem Cells: Protein Kinase C Signaling Shifts the Balance§

Authors

  • Debasree Dutta,

    1. Department of Pathology and Laboratory Medicine, Institute for Reproductive Health and Regenerative Medicine, University of Kansas Medical Center, Kansas City, Kansas, USA
    Search for more papers by this author
  • Soma Ray,

    1. Department of Pathology and Laboratory Medicine, Institute for Reproductive Health and Regenerative Medicine, University of Kansas Medical Center, Kansas City, Kansas, USA
    Search for more papers by this author
  • Pratik Home,

    1. Department of Pathology and Laboratory Medicine, Institute for Reproductive Health and Regenerative Medicine, University of Kansas Medical Center, Kansas City, Kansas, USA
    Search for more papers by this author
  • Melissa Larson,

    1. Transgenic and Gene Targeting Institutional Facility, University of Kansas Medical Center, Kansas City, Kansas, USA
    Search for more papers by this author
  • Michael W. Wolfe,

    1. Department of Molecular and Integrative Physiology, University of Kansas Medical Center, Kansas City, Kansas, USA
    Search for more papers by this author
  • Soumen Paul

    Corresponding author
    1. Department of Pathology and Laboratory Medicine, Institute for Reproductive Health and Regenerative Medicine, University of Kansas Medical Center, Kansas City, Kansas, USA
    • Institute for Reproductive Health and Regenerative Medicine, Department of Pathology and Laboratory Medicine, University of Kansas Medical Center, Kansas City, Kansas 66160, USA
    Search for more papers by this author
    • Ph: 913-588-7236; Fax: 913-588-8287


  • Disclosure of potential conflicts of interest is found at the end of this article.

  • Author contributions: D.D., S.R., P.H., and M.L.: performed experiments; M.W.W.: provided reagents and designed experiments; S.P.: designed experiments and wrote the manuscript.

  • §

    First published online in STEM CELLSEXPRESS February 3, 2011.

Abstract

The intricate molecular mechanisms that regulate ESC pluripotency are incompletely understood. Prior research indicated that activation of the Janus kinase–signal transducer and activator of transcription (STAT3) pathway or inhibition of extracellular signal-regulated kinase/glycogen synthase kinase 3 (ERK/GSK3) signaling maintains mouse ESC (mESC) pluripotency. Here, we demonstrate that inhibition of protein kinase C (PKC) isoforms maintains mESC pluripotency without the activation of STAT3 or inhibition of ERK/GSK3 signaling pathways. Our analyses revealed that the atypical PKC isoform, PKCζ plays an important role in inducing lineage commitment in mESCs through a PKCζ–nuclear factor kappa-light-chain-enhancer of activated B cells signaling axis. Furthermore, inhibition of PKC isoforms permits derivation of germline-competent ESCs from mouse blastocysts and also facilitates reprogramming of mouse embryonic fibroblasts toward induced pluripotent stem cells. Our results indicate that PKC signaling is critical to balancing ESC self-renewal and lineage commitment. STEM Cells 2011;29:618–628

INTRODUCTION

Pluripotent stem cells deal with two critical, yet opposing forces, to self-renew and to respond to differentiation signals. Over the years, several culture conditions have been used to maintain ESC pluripotency and to derive new ESCs and induced pluripotent stem cells (iPSCs) [1–6]. Interestingly, all of these strategies lead to the establishment of a core transcription factor network that maintains the ESC-specific chromatin structure and gene expression pattern [7]. Induction of ESC differentiation is associated with the alteration of epigenetic mechanisms leading to expression of developmental regulators, which in turn dictate lineage commitment. Although multiple signaling pathways have been implicated in maintaining ESC self-renewal versus differentiation, our understanding regarding the inter-relationship among different signaling pathways or mechanisms downstream to a distinct signaling pathway is incomplete.

The protein kinase C (PKC) family is involved in multiple signaling pathways and consists of several serine/threonine protein kinases [8] that are divided into three subfamilies; (a) classical PKCs (isoforms α, β1, β2, γ, calcium and phospholipid-dependent), (b) novel PKCs (isoforms δ, ε, η, θ, calcium-independent and phospholipid-dependent), and (c) atypical PKCs (isoforms ζ, ι/λ). Earlier, we found that pharmacological inhibition of PKC signaling by 3-[1-[3-(dimethylamino)propyl]-5-methoxy-1H-indol-3-yl]-4-(1H-indol-3-yl)-1H-pyrrole-2,5-dione (Gö6983, henceforth mentioned as PKCi), a highly selective inhibitor of PKC isoforms α, β, γ, δ, ζ, and μ [9, 10], inhibits angiogenic signal-mediated gene expression in endothelial cells [11]. On the basis of this observation, we hypothesized that PKC signaling might be involved in regulating gene expression during formation of the endothelial cell lineage. In investigating the role of PKC during endothelial differentiation, however, we serendipitously found that inhibition of PKC signaling is sufficient to maintain, derive, and propagate pluripotent ESCs and also facilitates reprogramming of differentiated cells to induce pluripotency. Prior to this study, PKC isoforms have been studied during ESC differentiation in different perspectives [12–16]. However, the involvement of the PKC signaling pathway in ESC pluripotency is largely unknown. Therefore, our study uncovered a yet unknown function of PKC signaling pathway, in which PKC isoforms, specifically PKCζ, induces lineage commitment in ESCs.

MATERIALS AND METHODS

Inhibitors

PKCi was purchased from three different companies (Sigma, St. Louis, MO, www.sigmaaldrich.com/united-states.html, # G1918; EMD Chemicals Inc., Gibbstown, NJ, www.emdchemicals.com, # 365251; and Tocris Biosciences, Ellisville, MO, www.tocris.com, # 2285). In all experiments, except the concentration profile, PKCi was used at 5 μM final concentration. Gö6976 (0.1-2 μM, # G1171) was purchased from Sigma. Rottlerin (0.1-5 μM, # 557370) and RO-318425 (0.1-5 μM # 557514) were purchased from EMD Chemicals. PD0325901 (1 μM, # 444966) and LY294002 (10 μM, # L9908) were from Calbiochem and Sigma, respectively. CHIR99021 (3 μM, # 04-0004) was purchased from Stemgent (Cambridge, MA, www.stemgent.com/commerce/pages/home).

ESC Cultures

E14Tg2a (E14), R1, and Stat3−/− ESCs were used in this study. Cells maintained for 3–5 days under different culture conditions were used for immunofluorescence study, whereas for other purposes, they were analyzed under steady state condition, that is, within 8–12 hours of leukemia inhibitory factor (LIF) and other inhibitor treatment. Detailed experimental protocol for ESC culture under different experimental conditions is described below.

E14 ESCs

Cells were grown in embryonic stem cell–Iscove's modified Dulbecco's medium (ES-IMDM) media (Lonza, Walkersville, MD, www.lonza.com) in a feeder-free condition. ES-IMDM was supplemented with 15% serum, 105 U/100 ml of LIF (ESGRO, Millipore, Temecula, CA, www.millipore.com) and 0.0124% monothioglycerol (Sigma Aldrich). For inducing differentiation in monolayer culture and to determine the effect of PKCi in preventing differentiation, embryonic day 14 (E14) cells were cultured on gelatin-coated plastics for 8 days without LIF. The details of experiments that are done at clonal-density are described below. To maintain E14 cells with PKCi, serum-supplemented ES-IMDM containing PKCi was used. For all assays involving elucidation of different signaling mechanism responsible for maintenance of pluripotency, cells from a ∼70% confluent ES culture plate were washed two times with 1× phosphate-buffered saline, trypsinized and plated on a six-well tissue culture plate, and treated with or without LIF or PKCi or different inhibitors for ∼10 hours followed by preparation of protein lysates and RNA. The PKCζ knocked down E14 cells that were generated in this study were maintained in serum-supplemented ES-IMDM on gelatin-coated plates in the absence of PKCi or LIF or any other inhibitor. We continuously cultured E14 cells for 18 consecutive passages (>58 days) on gelatin-coated plates without significant differentiation.

R1 ESCs

R1 cells were maintained on mouse embryonic fibroblast (MEF) feeder in ES-IMDM media supplemented with 15% ESC quality serum, 105 U/100 ml of LIF. Cells were grown for 3–5 days with change of medium every day. For inducing differentiation in monolayer culture and to determine the effect of PKCi, cells were cultured at feeder-free condition and without LIF for 7 days.

Stat3−/− ESCs

Passage 12 Stat3−/− ESCs (obtained from Dr. Austin Smith, Wellcome Trust Center for Stem Cell Research, Cambridge, U.K.) were maintained on N2B27 medium with 1 μM PD0325901 (mitogen-activated protein kinases/extracellular signal-regulated kinase [ERK] kinase [MEK] inhibitor) and 3 μM CHIR99021 (glycogen synthase kinase 3 [GSK3] inhibitor) and passaged every 3–4 days. For studies involving PKCi, around 1–2 × 104 cells were plated on six-well plates having N2B27 medium with or without 5 μM PKCi and analyzed for colony morphology and the expression of pluripotency markers.

Quantitative Clonal Assay

ESCs and iPSCs were dissociated into single cells using 0.05% trypsin/EDTA and 2–20 cells were plated on each well of a 96-well culture plate. The cells were cultured for 6–7 days, colonies were stained for Nanog, and the number of Nanog-positive colonies was counted. For determining maintenance of self-renewal for multiple passages at clonal density with PKCi, E14 cells were cultured at clonal density in 96-well plates with PKCi; cells from undifferentiated colonies were trypsizined after day 6, and again plated at clonal density with PKCi. This procedure was repeated for five consecutive passages (>30 days).

ESC Differentiation on Collagen IV and with Retinoic Acid

To differentiate ESCs in monolayer culture on collagen IV, ∼3 × 104 cells per well were transferred to collagen IV-coated six-well plates (354428, BD Biosciences, Franklin Lakes, NJ, www.bdbiosciences.com/home.jsp) cultured for 5 days in ES differentiation medium containing Dulbecco's modified Eagle's medium (DMEM; Invitrogen), 15% fetal bovine serum (FBS; selected for endothelial differentiation, Stem Cell Technologies, Vancouver, BC, www.stemcell.com), sodium pyruvate and L-glutamine with or without LIF and PKCi, and were recovered by cell dissociation buffer (BD Biosciences). For culturing multiple passages with PKCi on collagen IV, the recovered cells were again plated at a density of ∼3 × 104 cells per well and cultured again for 5 days. For our experiments, we continuously cultured E14 cells with PKCi (without LIF) up to eight consecutive passages on collagen IV plates without any noticeable differentiation. Besides, E14 cells maintained on collagen IV with PKCi for seven consecutive passages efficiently generated chimeras (showed in Fig. 2C). For retinoic acid (RA)-induced differentiation on monolayer culture, ESCs were treated with all-trans-retinoic acid (R2625-100MG, Sigma) in ethanol at a concentration of 1 μM with or without LIF and PKCi for 6 days followed by study of expression of pluripotency markers (by immunofluorescence or Western blot or real-time polymerase chain reaction [RT-PCR] analysis).

ESC Culture Under Serum-Free Condition

E14 cells were maintained in serum-free N2B27 medium containing DMEM/F12 (10565, Invitrogen, Carlsbad, CA, www.invitrogen.com/site/us/en/home.html), Neurobasal media (21103, Invitrogen), B27 supplement (17504-044, Invitrogen), N2 supplement (17502-048, Invitrogen), BSA fraction V (15260, Invitrogen), 2-mercaptoethanol (M7522-100ML, Sigma), and LIF/bone morphogenetic protein 4 (BMP4) (R & D Systems, Minneapolis, MN, www.rndsystems.com) with change of medium on alternate days and passed in every 2–4 days. For experiments with PKCi, ∼5 × 104 E14 cells were plated on each well of a gelatin-coated six-well plate having N2B27 medium without LIF/BMP4 and with or without 5 μM PKCi and cultured for 3–4 days before passing. Cells were analyzed for expression of different markers (by RT-PCR or immunofluorescence study). For experiments at clonal density, initially, E14 cells were cultured at high density in N2B27 medium with PKCi alone for four passages, then ∼200 cells were plated in each wells of a 96-well plate with PKCi in N2B27 and cultured for 7 more days. After 7 days, colonies were analyzed for Oct4 staining.

Chimera Generation and Germ Line Transmission

Prepubertal (3 weeks of age) donor C57BL/6 female mice (Jackson Laboratory, Bar Harbor, ME, jaxmice.jax.org) were superovulated, mated overnight to intact C57BL/6 stud males, and euthanized by cervical dislocation on day 3.5. Uteri were collected after euthanasia and flushed with M2 medium (Millipore) for the collection of blastocysts. Eight to ten ESCs or iPSCs were injected into the blastocoel of each blastocyst. After injection, the blastocysts were surgically transferred into recipient females that were pseudopregnant by mating with vasectomized males. Recipient females carried the pups to term and nursed until weaning at 3 weeks. High-level chimeras, generated from newly derived ESCs, were mated with C57BL/6 adult mice to test for germline transmission.

De Novo Derivation of New ESCs with PKCi

Blastocysts from 6-week-old female 129S2/SvPasCrl (129/Sv) strain (code 476, Charles River Laboratory, Wilmington, MA, www.criver.com/en-us/prodserv/bytype/resmodover/Pages/home2.aspx) mice were isolated at 3.5-day post coitum. Isolated blasotocysts were plated on gelatin-coated plastic in medium containing PKCi and FBS and without LIF. To obtain ESC colonies, blastocyst outgrowths were disaggregated with trypsin and replated on gelatin-coated wells with PKCi and FBS. ESC colonies were expanded by replating with PKCi at clonal or higher density. For in vitro analysis, cells were cultured following the same protocol mentioned earlier. After six passages with PKCi-condition, chromosome numbers were checked by karyotyping, and injected in C57BL/6 blastocysts for chimera generation.

Generation of iPSCs

iPSCs were derived using lentiviral vectors (Stemgent # 00-0004), expressing Oct4, Sox2, Klf4, or c-Myc under the control of the doxycycline (Dox)-inducible tetO operator. 129/Sv MEFs were infected with viral mix following the manufacturer's protocol. Twenty hours post-transduction, cells were reseeded in ESC growth medium with either LIF or PKCi at a density of 5 × 104 cells per well of a six-well plate and treated with Dox (2 μg/ml) for 48 hours and then replaced with medium without Dox. For further analysis, iPSC colonies were manually picked at different days starting at day 10. Cells were trypsinized, replated at clonal density in medium with PKCi (for PKCi-derived colonies) or LIF (for LIF-derived colonies) and the expression of pluripotent markers Nanog and Rex1 were tested by RT-PCR and immunofluorescence to determine true pluripotent iPSCs. For quantitation, iPSC colonies, which we were able to expand at clonal density and expressed both Nanog and Rex1 were considered. Total number of such colonies obtained with PKCi at day 20 was considered as 100%. RNAs were also analyzed by RT-PCR for the expression of E-Cadherin, Snail, Slug, and Twist to determine Epithelial to mesenchymal or mesenchymal to epithelial (MET) transition during reprogramming of MEFs to iPSCs in the presence of LIF or PKCi. For chimera generation, cells from two different PKCi-derived iPSC colonies were propagated at clonal density with PKCi, expanded for four passages and were injected into blastocysts.

RNA Interference and Rescue of PKCζ Expression in PKCζkd Cells

Short hairpin RNAs (shRNAs) targeting mouse PKCζ mRNA were cloned in pLKO1 (Addgene, Cambridge, MA, www.addgene.org). Lentiviral supernatants were produced in HEK-293T cells as mentioned earlier [11, 17]. E14 cells were transduced with lentiviral supernatants and were selected by the addition of 1 μg/ml of puromycin (Sigma). Construct with target sequence ATCC CGGTAAGTTCTGTTG, corresponding to the 3′untranslated region (UTR) region of PKCζ mRNA, specifically knocked down PKCζ expression. For rescue of PKCζ expression, PKCζkd cells were reinfected with lentiviral particles having PKCζ cDNA (subcloned from plasmid pMTH PKCζ, Addgene) into the pLKO.3G vector (Addgene) under a hU6 promoter and selected by the expression of green fluorescence protein (EGFP), which is expressed from the same vector under the control of an hPGK promoter (Supporting Information Fig. S6). Another construct with target sequence GGACCTCTGTGAG GAAGTG (shRNA2, Supporting Information Fig. S5), corresponding to the amino acid coding region of PKCζ mRNA, also specifically knocked down PKCζ. Construct with target sequence GTTGAGGACGAAGCAAGCC specifically knocked down PKCδ expression.

Additional experimental procedures are mentioned in the Supporting Information.

RESULTS

Inhibition of PKC Isoform Signaling Is Sufficient for Maintenance and De Novo Derivation of ESCs

To understand the function of PKC signaling during ESC differentiation, we cultured E14 mESCs with PKCi in the absence of LIF. We found that, E14 cells efficiently maintains undifferentiated colony morphology (Fig. 1A) when they are propagated at clonal density for five consecutive passages (Fig. 1B) with PKCi in the absence of LIF and protein analyses showed that PKCi-treatment maintains expression of pluripotency marker Oct4 (Fig. 1C) without induction of differentiation markers (Supporting Information Fig. S1A). To further test whether pluripotency is maintained in mESCs for higher passages in PKCi culture condition, we cultured E14 cells for 18 consecutive passages with PKCi in the absence of LIF and tested for colony morphology and expression of pluripotency markers. We found that, through out the culture period, the undifferentiated colony morphology as well expression of pluripotency markers Oct4, Nanog, and Sox2 are maintained similar to E14 cells that are cultured with LIF (Fig. 1D, 1E). We found a ∼ 30% loss in the Rex1 expression in PKCi-cultured cells compared with the LIF-cultured cells. However, in the PKCi condition, the expression of Rex1 was maintained at a significantly higher level compared with cells that were cultured in the absence of both LIF and PKCi (Fig. 1E). A concentration profile indicated that PKCi prevents ESC differentiation in a concentration-dependent manner (Supporting Information Fig. S1B).

Figure 1.

Inhibition of protein kinase C isoform by PKCi supports self-renewal of ESCs. (A): Images of embryonic day 14 (E14) ESCs showing undifferentiated colony morphology with PKCi. (B): The plot shows percentage of undifferentiated colonies when E14 cells were cultured at clonal density for five consecutive passages with PKCi. (C): Western blots showing sustained Oct4 expression in PKCi-treated E14 cells. (D): E14 cells were maintained with PKCi for higher passages and the images show the undifferentiated colony morphology and nanog expression (inset) in cells of different passages. (E): qRT-PCR analysis (mean ± SE; three independent experiments) of pluripotency gene expression in E14 cells, maintained for different passages with PKCi. Statistically significant (p ≤ .05) changes were indicated in the plot. (F): E14 cells were cultured at clonal density in N2B27 with PKCi alone. Images show undifferentiated ESC colony morphology (left) and expression of Oct4 (right). (G): qRT-PCR analysis (mean ± SE; three independent experiments) of inhibitor-of-differentiation gene expression in E14 cells, cultured in N2B27 medium with PKCi alone or LIF/BMP4. (H): Oct4 staining in E14 cells after five passages (25 days) on collagen IV with PKCi. (I): qRT-PCR analysis (mean ± SE; three independent experiments) of lineage-specific marker expression in E14 cells that were cultured on collagen IV for 5 days with or without PKCi and LIF. Expression of lineage-specific markers was significantly (p ≤ .01) inhibited in PKCi-cultured cells compared with cells that were cultured on collagen IV without LIF and PKCi. Abbreviations: BMP4, bone morphogenetic protein 4; DAPI, 4′,6-diamidino-2-phenylindole; LIF, leukemia inhibitory factor; PKCi, 3-[1-[3-(dimethylamino)propyl]-5-methoxy-1H-indol-3-yl]-4-(1H-indol-3-yl)-1H-pyrrole-2,5-dione; qRT-PCR, quantitative real-time polymerase chain reaction.

As PKC signaling regulates cell proliferation and survival in multiple contexts [18], we tested the effect of PKC inhibition on mESC proliferation, cell-doubling time, cell cycle distribution pattern, and cell death. We found that at 2.5–5 μM concentration of PKCi, which efficiently inhibits ESC differentiation (Supporting Information Fig. S1B), cell proliferation was inhibited by ∼30%–40% (Supporting Information Fig. S2A). Similarly, we found an increase in cell doubling time in PKCi-cultured cells compared with cells, cultured with LIF (Supporting Information Fig. S2B). However, no increase in cell death was observed at 5 μM concentration of PKCi (Supporting Information Fig. S2C) and the cell cycle distribution pattern were also very similar between PKCi-cultured and LIF-cultured mESCs (Supporting Information Fig. S2D).

We also tested whether PKCi prevents mESC differentiation in the absence of serum. We found that, mESCs can be maintained in an undifferentiated state in serum-free N2B27 medium [4] with PKCi alone (Fig. 1F) and this maintenance is not associated with the induction of inhibitor-of-differentiation genes (Fig. 1G).

To test whether PKC inhibition maintains pluripotency independent of mESC types, we used R1 ESCs that are normally maintained at undifferentiated state with LIF on MEF feeder. We found that PKCi also inhibits differentiation of R1 ESCs (Supporting Information Fig. S3A). Next, we induced mESC differentiation in the presence of additional differentiation cues by culturing them on collagen IV or treating with RA [19, 20] and tested whether PKCi prevents mESC differentiation despite the presence of additional differentiation cues. We found that PKCi inhibits mESC differentiation on collagen IV (Fig. 1H). Analysis of mRNA expression showed that, similar to culturing with LIF, PKCi treatment completely inhibited induction of lineage-specific gene expression on collagen IV (Fig. 1I) and maintained expression of pluripotency markers (Supporting Information Fig. S3B, S3C). PKCi also efficiently inhibited RA-induced differentiation of mESCs (Supporting Information Fig. S3D, S3E).

We tested whether mESCs, cultured with PKCi for multiple passages, maintain multilineage differentiation potential in vitro and developmental potential in vivo. Intriguingly, we found that on withdrawal of PKCi, mESCs, maintained for five passages with PKCi, readily form embryoid bodies and differentiate on collagen IV (Fig. 2A) with induction of lineage-specific genes (Fig. 2B). Furthermore, when injected into the blastocysts, PKCi-maintained ESCs readily yielded chimeric mice (Fig. 2C). These results confirmed that culture in PKCi does not compromise multilineage differentiation and developmental potency.

Figure 2.

Multidifferentiation and developmental potency of ESCs that are maintained at undifferentiated state by inhibiting protein kinase C isoforms. (A): Experimental strategy to determine the differentiation potency of ESCs that were cultured in the presence of PKCi for multiple passages. The diagram shows that embryonic day 14 (E14) cells, maintained for five passages in PKCi, was differentiated on collagen IV (for 5 days) and generated EBs in the absence of PKCi. The bar graph shows the relative efficiency (mean ± SE; three independent experiments) of EB formation of PKCi-maintained E14 cells in comparison with cells that were maintained with LIF on plastic. (B): Multidifferentiation potency on collagen IV was determined by measuring mRNA expression (mean ± SE; three independent experiments) of pluripotency and lineage markers. mRNA levels were measured after culturing on collagen IV for 5 days in the absence and or presence of PKCi. The plot shows significant (p ≤ .01) loss in pluripotency gene expression and increase in lineage-specific markers in the absence of PKCi. (C): Chimeric mice, generated with E14 cells that were cultured for seven passages (35 days) with PKCi on collagen IV. The table below shows number of chimeras obtained. Abbreviations: EBs, embryoid bodies; LIF, leukemia inhibitory factor; PKCi, 3-[1-[3-(dimethylamino)propyl]-5-methoxy-1H-indol-3-yl]-4-(1H-indol-3-yl)-1H-pyrrole-2,5-dione.

Next, we investigated whether blockade of PKC signaling is sufficient for de novo derivation of ESCs. We plated E3.5 blastocysts from 129/Sv mice on gelatin-coated plates with PKCi and serum and without LIF and feeder cells. We readily derived several ESC colonies from blastocyst outgrowths (Fig. 3Aa, 3Ab). Although we obtained multiple ESC colonies from 10 blastocysts in presence of serum and PKCi, we selected 10 ESC colonies for subsequent analyses. Interestingly, without PKCi, and with serum alone, we were unable to obtain any undifferentiated ESC colony from >15 blastocysts. The newly PKCi-derived ESC lines were successfully propagated in an undifferentiated state when cultured at clonal density with PKCi (Fig. 3Ac, 3Ad). After six passages, two different PKCi-derived ESC lines were injected into the blastocysts to generate chimera (Fig. 3B). The PKCi-derived ESCs successfully generated adult chimera, which produced germline offsprings (Fig. 3C, 3D) when crossed with C57BL/6 adults. These results indicate that inhibition of PKC signaling is sufficient to derive germline-competent pluripotent ESCs from mouse blastocyst.

Figure 3.

De novo derivation of pluripotent ESCs by inhibiting PKC isoforms. (A): Blastocyst outgrowth (A′) and establishment of ES colony (B′) from 129/Sv mice with PKCi. Phase (C′) and fluorescence (D′) images showing Nanog expression in 129/Sv ESCs after four passages with PKCi. (B): Chimeric mice produced from PKCi-derived 129/Sv ESCs cultured for six passages with PKCi. (C): Germline offsprings produced from chimeras that were generated from PKCi-derived 129/Sv ESCs. (D): The table shows number of chimeras that produced germline offsprings. Abbreviation: PKCi, 3-[1-[3-(dimethylamino)propyl]-5-methoxy-1H-indol-3-yl]-4-(1H-indol-3-yl)-1H-pyrrole-2,5-dione.

Atypical PKC Isoform PKCζ Promotes mESC Differentiation

PKCi selectively inhibits six different PKC isoforms (α, βI, βII, γ, δ, and ζ) and at a higher concentration (>20 μM) inhibits isoform PKCμ [9]. As 2.5 μM of PKCi inhibits mESC differentiation (Supporting Information Fig. S1B), we concluded that PKCμ function is dispensable for ESC differentiation. Western blot analysis showed that all of the other six PKC isoforms are expressed in mESCs (Supporting Information Fig. S4). Western blot analyses further showed that PKC α, βI, δ, and ζ are phosphorylated in mESCs and their phosphorylation were strongly inhibited by PKCi (Fig. 4A). Because of our inability to obtain specific antibody, we were unable to definitively determine the phosphorylation state of PKC βII and PKC γ in mESCs. We next used a series of PKC inhibitors possessing different specificities to narrow our search for the PKC isoform responsible for mESC differentiation. However, Gö6976 (inhibits PKC α, βI, βII [21]; Fig. 4B, left), Rottlerin (inhibits PKCδ, [22]; Fig. 4B, middle), and Rö-31-8425 (inhibits PKC α, βI, βII, γ and ε, [23]; Fig. 4B, right), could not prevent differentiation of mESCs in the absence of LIF. Therefore, we predicted that the atypical PKC, PKCζ, might be important for mESC differentiation. Although Western blot analysis showed that PKCζ is phosphorylated in mESCs and phosphorylation is strongly inhibited by PKCi (Fig. 4A), to further validate that PKCi impairs PKCζ function, we looked at PKCζ target proteins. PKCζ directly phosphorylates the serine 311 (S311) residue of the RelA subunit of nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB) [24] and the lethal giant larvae 1 and 2 (LGL1/2) proteins at conserved serine residues [25, 26]. We found that PKCi inhibits the phosphorylation of RelA and LGL1/2 (Fig. 4C) in mESCs, confirming that activity of PKCζ is disrupted with PKCi treatment.

Figure 4.

LIF-independent maintenance of self-renewal of PKCζ-depleted ESCs. (A): Western blots showing inhibition of phosphorylation of different PKC isoforms by PKCi in embryonic day 14 (E14) cells. (B): Images showing differentiation of E14 cells in the presence of selective PKC isoform/s inhibitors. (C): Western blots showing loss of RelA(S311) and LGL1/2(S650/S654) phosphorylation in PKCi-treated E14 cells. (D): Western blots showing knockdown of PKCζ and loss of RelA(S311) phosphorylation in E14 cells with shRNA molecules, targeted against the 3′ UTR region of PKCζ mRNA. The blots also show rescue of PKCζ expression and RelA phosphorylation with ectopic expression PKCζ from an RNA interference (RNAi) immune construct. (E): Images showing maintenance of undifferentiated ESC colony morphology (top left) and expression of Oct4 (top right) in PKCζkd cells after five passages without LIF. Images showing rescue of differentiation of PKCζkd cells with ectopic expression of RNAi immune PKCζ (bottom). Ectopic expression of PKCζ in differentiated cells (inset) was confirmed from the expression of EGFP (expressed from the same construct). (F): The plot shows percentage of undifferentiated E14 ESC colonies (mean ± SE; three independent experiments) when PKCζkd cells, with or without ectopic expression of RNAi-immune PKCζ, were cultured at clonal density. The data indicate significant loss (p ≤ .01) of undifferentiated colony formation with ectopic expression of RNAi-immune PKCζ. (G): qRT-PCR analysis (mean ± SE; three independent experiments) showing significant (p ≤ .01) loss of pluripotency gene expression in PKCζkd cells on ectopic expression of PKCζ from the RNAi-immune construct. Abbreviations: DAPI, 4′,6-diamidino-2-phenylindole; LGL, lethal giant larvae; LIF, leukemia inhibitory factor; PKC, protein kinase C; PKCi, 3-[1-[3-(dimethylamino)propyl]-5-methoxy-1H-indol-3-yl]-4-(1H-indol-3-yl)-1H-pyrrole-2,5-dione; UTR, untranslated region.

Next, to test the importance of PKCζ activity during mESC differentiation, we tested differentiation potential of mESCs, in which PKCζ was knocked down by RNA interference (RNAi). We used the RNAi approach because the PKCζ (Prkcz)−/− ESCs are not available for our study. For RNAi, we designed a shRNA molecule that specifically targets the 3′ UTR of PKCζ and efficiently knocks down its expression in E14 cells (Fig. 4D). The loss of PKCζ function in PKCζ-knocked down (PKCζkd) cells was validated from the loss of RelAS311 phosphorylation (Fig. 4D). We found that, when cultured on gelatin-coated plates for multiple passages and without LIF, the PKCζkd cells maintain undifferentiated ESC colony morphology and expression of pluripotency markers (Fig. 4E). Similar results were obtained when PKCζ was specifically knocked down using a different shRNA construct, which targets the PKCζ coding sequence (data shown in Supporting Information Fig. S5). To validate that impaired mESC differentiation is specifically due to the loss of PKCζ function, we ectopically expressed an RNAi-immune PKCζ mRNA (without 3′UTR) in PKCζkd cells using a lentiviral vector. The viral vector also expressed an EGFP cDNA (Supporting Information Fig. S6) for monitoring ectopic expression of PKCζ. When PKCζ is ectopically expressed from the RNAi-immune construct (right lane of Fig. 4D), the PKCζkd cells readily differentiate in the absence of LIF (Fig. 4E). To further validate the role of PKCζ in mESC differentiation, we cultured PKCζkd cells at clonal density without LIF and found that PKCζkd cells maintain undifferentiated colony morphology at a >60% efficiency (Fig. 4F) with expression of pluripotency markers (Fig. 4G). However, they failed to do so when PKCζ was ectopically expressed from the RNAi immune construct. These results confirm that depletion of PKCζ promotes mESC self-renewal and implicate an active role of PKCζ in inducing lineage commitment.

Interestingly, when cultured on collagen IV without LIF, the majority of the PKCζkd cells undergo differentiation along with the presence of some undifferentiated ESC colonies (Supporting Information Fig. S7A). We reasoned that other PKC isoforms might compensate the loss of PKCζ function in the presence of additional differentiation cues on collagen IV. So, we knocked down PKCδ in PKCζkd cells (Supporting Information Fig. S7B). We chose PKCδ because of a recent report [27] that implicated PKCδ inhibition in maintaining mESCs self-renewal in vitro. Compared with the PKCζkd cells, the expression of pluripotency genes, Oct4, Nanog, and Sox2, were significantly induced when both PKCζ and PKCδ were knocked down (Supporting Information Fig. S7C). However, the expression of pluripotency genes was further induced significantly when double knocked down cells were cultured on collagen IV with PKCi (Supporting Information Fig. S7C). Thus, we concluded that function of PKCζ alone promotes differentiation in mESCs but a combinatorial function of other PKC isoforms along with PKCζ further potentiate lineage commitment of ESCs. However, detailed studies are needed to make definitive conclusions regarding contribution of other PKC isoforms toward mESCs differentiation.

Inhibition of PKC Signaling Inhibits NF-κB Activity in Mouse ESCs

Although PKCi is a selective PKC inhibitor, it might regulate other signaling pathways that are implicated in the maintenance of ESC pluripotency. Activation of Janus kinase–signal transducer and activator of transcription (JAK-STAT3) and PI (3)K-Akt pathways have been implicated in maintaining mESC pluripotency [2, 28]. We found that PKCi does not induce STAT3 phosphorylation in E14 cells (Fig. 5A). Furthermore, PKCi efficiently prevents differentiation of Stat3−/− ESCs [5] (Fig. 5B), indicating that PKCi-mediated inhibition of ESC differentiation is independent of JAK-STAT3 signaling pathway. We also tested whether PKCi activates PI (3)K-Akt signaling in mESCs. Similar to an earlier observation [29], we found that LIF induces Akt phosphorylation in mESCs. However, unlike LIF, PKCi-mediated inhibition of mESC differentiation is not associated with Akt phosphorylation (Fig. 5C). Moreover, LY294002, a potent inhibitor of PI3 kinase [29], does not prevent the PKCi-mediated maintenance of mESC self-renewal (Fig. 5D). Thus, PKCi-mediated maintenance of mESC self-renewal is not associated with the activation of the PI (3)K-Akt pathway.

Figure 5.

Effect of PKC isoform inhibition on signaling pathways that are implicated in ESC pluripotency. (A): Western blots showing lack of STAT3 (Tyrosine (Y)705) phosphorylation in PKCi-treated E14 ESCs in the absence of LIF. (B): Images of Stat3−/− ESCs, cultured with PKCi, showing undifferentiated ESC colony morphology (left) and expression of Nanog (right). (C): Western blot analysis of Akt phosphorylation in E14 ESCs. (D): Images showing undifferentiated ESC colony morphology (left) and expression of Oct4 (right) in E14 ESCs, cultured with PKCi and PI3 kinase inhibitor, Ly294002, in the absence of LIF. (E): Western blot analysis of GSK3 phosphorylation in E14 cells when cultured with or without LIF and PKCi. (F): Western blot analysis of β-catenin in the cytosolic fractions of E14 cells, cultured with or without LIF, and in the presence of PKCi with or without Wnt3A conditioned medium. (G): TopFlash luciferase reporter analysis (mean ± SE; three independent experiments) in E14 cells, cultured with or without LIF, and in the presence of PKCi with or without Wnt3A conditioned medium (*, p ≤ .05). (H): Western blots showing c-Myc phosphorylation in E14 cells, cultured with or without LIF, and in the presence of PKCi with or without GSK3 inhibitor CHIR99201. (I): Western blots showing ERK1/2 and RSK1 phosphorylation in E14 cells, cultured in the presence and absence of LIF, and in the presence of PKCi with or without ERK inhibitor PD0325901. (J): qRT-PCR analysis (mean ± SE; three independent experiments) of NF-κB target genes expression in E14 ESCs, cultured with or without LIF and PKCi, and in PKCζkd cells, with or without ectopic expression of RNA interference-immune PKCζ cDNA (*, p ≤ .05). (K): Analysis (mean ± SE; three independent experiments) of NFκ5x-Luc reporter activation in cells, analyzed in (J). *, p ≤ .05. Abbreviations: ERK, extracellular signal-regulated kinase; GSK3, glycogen synthase kinase 3; Igfpb2, insulin-like growth factor binding protein 2; LIF, leukemia inhibitory factor; PKCi, 3-[1-[3-(dimethylamino)propyl]-5-methoxy-1H-indol-3-yl]-4-(1H-indol-3-yl)-1H-pyrrole-2,5-dione; Plaur, plasminogen activator, urokinase receptor; STAT3, signal transducer and activator of transcription; Vim, vimentin.

Using small molecule inhibitors, it has been shown that inhibition of GSK3 and ERK1/2 signaling also promote mESC pluripotency. In addition, GSK3/Wnt/β-catenin pathway has also been shown to maintain undifferentiated phenotype of both mESCs and hESCs [3]. However, PKCi does not change phosphorylation of either GSK3α or GSK3β (Fig. 5E), does not stabilize β-catenin (Fig. 5F), or modulate β-catenin-mediated transcriptional activation of a canonical Wnt reporter (TOPflash reporter, Fig. 5G). We also found that PKCi does not inhibit GSK3β-mediated phosphorylation of c-Myc (Fig. 5H) at threonine 58(T58) residue, which induces degradation of c-Myc and contributes to mESC differentiation [30]. Rather, in the PKCi culture condition, addition of the GSK3 inhibitor, CHIR99021 [5], abolishes c-Myc phosphorylation (Fig. 5H). Next, we tested whether ERK signaling is functional in PKCi-treated ESCs. We found that PKCi does not inhibit phosphorylation of ERK1/2 or p90Rsk1 (Rsk1), a downstream target of ERK1/2 [31] (Fig. 5I). However, a combination of PKCi and PD0325901, a potent MEK inhibitor [5], almost completely inhibited ERK1/2 and Rsk1 phosphorylation (Fig. 5I). Therefore, GSK3 or ERK1/2 inhibition is not involved in maintenance of pluripotency in the PKCi-treated mESCs.

Recently, inhibition of NF-κB activity has been linked to stem cell pluripotency [32]. Studies in knockout mice indicated that PKCζ function is essential for NF-κB transcriptional activity in response to several signaling pathways [33]. Furthermore, it has also been shown that in PKCζ-deficient cells NF-κB is transcriptionally inactive due to impaired phosphorylation at S311 residue of the RelA subunit [24]. As both PKCi and PKCζ knockdown inhibit RelA phosphorylation in mESCs (Fig. 4C, 4D), and ectopic expression of PKCζ rescues the phosphorylation in PKCζkd cells (Fig. 4D), we tested whether PKCi downregulates NF-κB activity in mESCs. mRNA analysis showed that, in the absence of LIF, transcription of NF-κB target genes are activated, and PKCi inhibits their activation (Fig. 5J). Analysis with PKCζkd cells in the absence of LIF also showed similar results (Fig. 5J). However, ectopic expression of PKCζ rescues activation of NF-κB target genes in PKCζkd cells. To further evaluate impairment of NF-κB transcriptional activity, we used a reporter plasmid, in which Luciferase expression is regulated by five NF-κB binding motifs (NFκ5x-Luc reporter). Interestingly, we detected considerable reporter activity in E14 cells in the presence of LIF, and the reporter activity was further induced in the absence of LIF (Fig. 5K). However, reporter gene activation is strongly inhibited with PKCi or in PKCζkd cells. Furthermore, similar to NF-κB target gene activation, ectopic expression of PKCζ rescues reporter activity in PKCζkd cells (Fig. 5K). These multiple lines of evidence strongly indicate that in mESCs NF-κB is a target downstream of PKCζ and the involvement of a PKCζ-NF-κB signaling axis contributes to the regulation of lineage commitment in mESCs.

Inhibition of PKC Signaling Facilitates Reprogramming of Differentiated Cells

As PKCi condition maintains ESC pluripotency, next, we tested whether PKCi facilitates derivation of iPSCs [6]. We infected 129/Sv MEFs with lentiviral vectors encoding the four reprogramming factors Oct4, Sox2, Klf4, and c-Myc, and infected MEFs were cultured in the presence or absence of PKCi or LIF. Several ESC-like iPSC colonies appeared beginning 12 days after culturing with PKCi (Fig. 6A) and expression of pluripotency marker Nanog was confirmed in forming iPS colonies (Supporting Information Fig. S8). Those iPSC colonies were readily propagated at clonal density with PKCi and expression of pluripotency markers were confirmed (Fig. 6B). Finally, we injected PKCi-derived iPSCs into blastocysts and generated chimeric mice (Fig. 6C, 6D). Interestingly, compared with LIF, in the presence of PKCi, true iPSC colonies, defined by the expression of Nanog and Rex-1 after the cells are propagated at clonal density, were obtained at a significantly faster rate (Fig. 6E). Furthermore, the efficiency of obtaining iPSC colonies was ∼three fold higher with PKCi compared to LIF (Fig. 6F). As MET is crucial for the formation of iPSC colonies [34], we tested expression of genes implicated in MET. We found that, compared with MEFs, mRNA expression of E-Cadherin was induced in both PKCi-and LIF-induced iPSCs (Fig. 6G). However, E-Cadherin induction was much higher in PKCi-derived iPSCs compared with LIF-derived iPSCs (Fig. 6G). Moreover, the expression of Snail and Slug were more strongly repressed with PKCi compared with LIF (Fig. 6G). These results indicate an increased MET response in PKCi-cultured cells compared with the LIF-cultured cells. Collectively, our results indicate that inhibition of PKC by PKCi provides an efficient culture condition for reprogramming differentiated cells to iPSCs.

Figure 6.

PKC inhibition facilitates reprogramming and the derivation of induced pluripotent stem cells. (A): Images of a developing iPSC colony from MEFs in the presence of PKCi. (B): Immunostaining showing Nanog expression in iPSCs after replating with PKCi at clonal density. (C): Chimeric mice that were generated with iPSCs, which are derived and propagated (passage 4) with PKCi. (D): The table shows number of chimeras obtained with different iPSC clones. (E): Plot shows comparative time course of iPSC colony formation in the presence of LIF and PKCi. (F): Relative numbers of iPSC colonies that were derived in different culture conditions (mean ± SE; three independent experiments; *, p ≤ .05). (G): qRT-PCR analysis (mean ± SE; three independent experiments) showing significantly (p ≤ .05) higher expression of E-Cadherin mRNA and reduced expression of Snail and Slug mRNAs in PKCi-derived iPSCs compared with LIF-derived iPSCs. Abbreviations: iPSC, induced pluripotent stem cell; LIF, leukemia inhibitory factor; MEF, mouse embryonic fibroblast; PKCi, 3-[1-[3-(dimethylamino)propyl]-5-methoxy-1H-indol-3-yl]-4-(1H-indol-3-yl)-1H-pyrrole-2,5-dione.

DISCUSSION

In recent years, a major area of research has focused on studying the molecular basis of pluripotency. This study reports for the first time that inhibition of PKC isoform function is sufficient to maintain undifferentiated cultures of mESCs without affecting their multidifferentiation and developmental potency.

Activation of the ERK pathway downstream to fibroblast growth factor 4 or other stimuli has been indicated to be a key-signaling component to induce lineage specification in ESCs [5]. This is supported by the findings that small molecule inhibitors of this pathway along with inhibition of GSK3 is sufficient to sustain ESC pluripotency and to derive new ESCs from multiple species [5, 35]. In this study, we have clearly shown that, under different culture conditions, similar to a “neutralized” environment without external stimuli (in N2B27 medium) or in the presence of strong differentiation cues (collagen IV, RA-treatment), a single selective PKC inhibitor maintains ESC pluripotency without affecting ERK or GSK3-dependent pathways. Furthermore, we showed in this study that PKC inhibition by PKCi is an efficient strategy for derivation of new ESCs and reprogramming of differentiated cells to iPSCs. Therefore, it is attractive to propose that similar to blockade of ERK activation, inhibition of PKC isoforms also promote ground state of self-renewal in mouse ESCs.

How does blockade of two different signaling pathways independently promote ground state of ESC self-renewal? As the decision between ESC self-renewal versus differentiation to other cell types is executed through epigenetic regulations, we hypothesize that individual signaling pathways such as ERK or PKC regulate distinct and/or overlapping cellular components that modulate the epigenetic regulators in stem cells, thereby altering the gene expression program and leading to lineage commitment. Therefore, future studies investigating a role for PKC signaling in relation to regulation of epigenetic components will provide insight into cellular mechanisms that dictate stem cell self-renewal versus differentiation.

A significant finding of this study is the involvement of PKCζ-NF-κB pathway during ESC differentiation. We showed, here, that PKCζ function is involved in the activation of NF-κB pathway during ESC differentiation and inhibition of PKCζ signaling by PKCi or knock down of PKCζ inhibits NF-κB activity and target gene expression. These results strongly indicate that a PKCζ-NF-κB signaling axis contributes to lineage commitment in mESCs. PKCζ is known to interact with other proteins including prostate apoptosis response (PAR) proteins and is involved in regulation of other cellular functions including cell polarity [36]. Other cellular mechanisms, downstream to PKCζ function, might also be involved in inducing ESC differentiation. In that relation, it is worthwhile to mention that downregulation of both PKCζ and PAR3 in blastomeres of preimplantation mouse embryos increases their differentiation toward inner cell mass [37], the source of ESCs. We found that PKCζ is phosphorylated in undifferentiated ESCs that are maintained in the presence of LIF. Furthermore, PKCζ overexpression from the RNAi-immune construct was not sufficient to induce mESC differentiation when cultured with LIF (data not shown). These results indicate that LIF-signaling pathway in mESCs overrides the differentiation signals mediated via activated PKCζ. However, the mechanism of PKCζ activation in ESCs and other pathways downstream of PKCζ activation merits further study.

Although PKCζ is involved in ESC differentiation, depletion of PKCζ is not sufficient to inhibit multilineage differentiation of ESCs in the presence of strong differentiation cues, such as culturing on collagen IV or with RA. Rather, PKCi, which inhibits four additional PKC isoforms, prevents multilineage differentiation in diverse culture conditions. Thus, combinatorial function of other PKC isoforms along with PKCζ might be important to induce multilineage commitment of ESCs. Roles of individual PKC isoforms have been implicated in regulating self-renewal versus differentiation of lineage-specific stem or progenitor cells [38–41]. Thus, it will be interesting to understand whether an individual PKC isoform plays a distinct role in differentiation of pluripotent stem cells toward a specific lineage.

CONCLUSION

Inhibition of PKC signaling maintains mESC pluripotency without affecting their developmental potency. In mESCs, a PKCζ–NF-κB signaling axis contributes toward lineage commitment. In addition, inhibition of PKC signaling also facilitates induced pluripotency. So, PKC signaling directly contributes to mESC lineage commitment.

Acknowledgements

We thank Dr. Stuart H. Orkin for critical comments and Dr. Austin Smith for Stat3−/− cells. This work was supported by NIH grant (HL094892, P20 RRO24214, HD062546).

Disclosure of Potential Conflicts of Interest

D.D., S.R., and S.P have filed patent applications through University of Kansas Medical Center. The other authors indicate no potential conflicts of interest.