Novel AKT phosphorylation sites identified in the pluripotency factors OCT4, SOX2 and KLF4

Abstract

The four OSKM factors OCT4, SOX2, KLF4 and c-MYC are key transcription factors modulating pluripotency, self-renewal and tumorigenesis in stem cells. However, although their transcriptional targets have been extensively studied, little is known about how these factors are regulated at the posttranslational level. In this study, we established an in vitro system to identify phosphorylation patterns of the OSKM factors by AKT kinase. OCT4, SOX2, KLF4 and c-MYC were expressed in Sf9 insect cells employing the baculoviral expression system. OCT4, SOX2 and KLF4 were localized in the nucleus of insect cells, allowing their easy purification to near homogeneity upon nuclear fractionation. All transcription factors were isolated as biologically active DNA-binding proteins. Using in vitro phosphorylation and mass spectrometry-based phosphoproteome analyses several novel and known AKT phosphorylation sites could be identified in OCT4, SOX2 and KLF4.

Abbreviations

BSA

bovine serum albumin

CV

column volume

GST

glutathione S-transferase

EMSA

electrophoretic mobility shift assay

MALDI

matrix-assisted laser desorption/ionization

MOI

multiplicity of infection

MS

mass spectrometry

OSKM

OCT4,

SOX2

KLF4,

c-MYC;

UTF1, undifferentiated embryonic cell transcription factor 1

Introduction

Stem cells require intricate regulatory networks governed by a limited number of key transcription factors in order to maintain self-renewal and pluripotency. Essential for these processes are the so-called OSKM factors including the transcription factors OCT4, SOX2, KLF4 and c-MYC.¹ The ectopic expression of solely OSKM factors is sufficient to reprogram terminally differentiated cells back to pluripotency in both human and mouse systems.^2,3 OSKM transcription factors are involved not only in pluripotency and differentiation of stem cells, but also in tumorigenesis.^4-11

The expression of OSKM factors is controlled in a coordinated manner by several feedback loops and requires a tight regulation to maintain stem cell potential. A slight change in the amount of the proteins can lead to drastic alterations in cell fate that may eventually result in tumorigenesis. For instance, the level of OCT4, also known as POU5F1, is crucial for cell fate transitions and embryonic development.^12,13 If OCT4 protein is elevated more than 2-fold compared to normal conditions, stem cells loose their pluripotency and start differentiation into endoderm and mesoderm.¹⁴ Moreover, a moderate increase of OCT4 is sufficient to cause to tumorigenesis e.g. of gonadal tumors.¹⁵

Both the expression levels and transcriptional activity of OSKM factors are fine-tuned by various posttranslational modifications, such as sumoylation, acetylation, methylation, O-GlcNAcylation and phosphorylation.^16-22 Recent evidence suggests that the serine/threonine kinase AKT plays a critical role not only in tumorigenesis, but is also required for self-renewal and pluripotency of normal cells.^23-25 AKT1 has been shown to directly phosphorylate OCT4, SOX2 and KLF4, modulating their nuclear localization, stability and transcriptional activity.^26-28 In particular, AKT phosphorylates OCT4 at T235 leading to enhanced apoptosis resistance and tumorigenic potential in mouse embryonic carcinoma cells.²⁹ Likewise, SOX2 phosphorylation at T118 by AKT results in decreased proteasomal degradation of SOX2 protein and enhanced self-renewal capacity of mouse embryonic stem cells.^27,30 In contrast, AKT-mediated phosphorylation of human KLF4 on T429 accelerates its degradation and thereby impairs stemness.³¹ In addition to triggering posttranslational modifications, indirect links between AKT and certain OSKM factors, such SOX2, have been described. In mice, Akt has been suggested to repress Sox2 transcription via a regulatory circuit involving FoxO1.²⁸ Moreover, AKT was reported to regulate SOX2 transcriptional activity via p27 and miR-30a in nasopharyngeal cancers.³² AKT mediates posttranslational modifications of the OSKM factors but, conversely, posttranslational modifications of OCT4 might also modulate AKT activity, thereby forming a positive feedback loop.^26,29 Thus, these data point to multiple mutual links between AKT and the OSKM factors, which might be critical for maintenance of stemness in malignant and non-malignant stem cells.

Despite the increasing evidence for the interaction of AKT with OSKM factors, the AKT-specific phosphorylation sites of the OSKM factors are still incompletely understood. Previous in vitro phosphorylation studies have been solely performed with recombinant OSKM proteins expressed in bacteria.^29,33 This approach exhibits several limitations, because the transcription factors are mostly expressed in inclusion bodies and need to be denatured and refolded in vitro. However, refolding of proteins from inclusion bodies is challenging and usually results in poor yields of natively folded, bioactive proteins.³⁴

In the present study we developed a reliable in vitro tool to identify the AKT phosphorylation sites, based on the baculoviral expression of OCT4, SOX2, KLF4 and c-MYC in insect cells. This expression system is capable of performing most mammalian post-translational modifications, which are crucial for the regulation of transcriptions factors.³⁵ We show that all OSKM factors can be purified in a simple manner to near homogeneity as native proteins and – similar to their mammalian counterparts – retain essential activities, such as nuclear translocation and DNA binding. Using in vitro kinase assays coupled with mass spectrometry-based phosphoproteomics we could not only confirm previously reported phosphorylation sites, but were able to identify several new AKT phosphorylation sites within OCT4, SOX2 and KLF4. The described approach will be also suitable to explore additional posttranslational modifications of OSKM transcription factors.

Results

Recombinant OSKM factors translocate to the nucleus of Sf9 insect cells

A major advantage of the Sf9 baculovirus system is that, unlike bacterial expression systems, proteins can be produced in a native form retaining the correct subcellular compartmentalization and posttranslational modifications. To produce human OSKM proteins their cDNAs were cloned into the N-terminal glutathione S-transferase (GST) fusion plasmid pAcG2T. After cotransfection of the vector with linearized wildtype baculoviral DNA high-titer virus stocks were produced. To optimize protein production we first analyzed several multiplicities of infection (MOI) and incubation periods postinfection. In first fractionation experiments we noticed that OCT4, SOX2 and KLF4 were predominantly localized in the nucleus of Sf9 cells, whereas c-MYC was distributed both in cytoplasmic and nuclear fractions, as revealed by Coomassie staining and Western blotting (Fig. 1A). For the purification of OCT4, SOX2 and KLF4 we therefore developed a single-step nuclear extraction protocol, allowing an easy enrichment of the transcription factors by nuclear fractionation, whereas c-MYC was purified from whole cell lysates.

Figure 1.

Recombinant OCT4, SOX2, and KLF4 are enriched in nuclear fractions of Sf9 insect cells. (A) Nuclear localization of OSKM factors: Sf9 cells were infected with baculoviruses encoding human OCT4, SOX2, KLF4 and c-MYC. Two days postinfection for SOX2, KLF4 and c-MYC, and 3 d postinfection for OCT4 cytosolic and nuclear fractions were prepared and subjected to SDS-PAGE. The upper panel shows a Coomassie blue staining of the gel with dotted boxes highlighting the protein of interest in the nuclear fractions. The lower panel shows an immunoblot using a GST antibody for the detection of OCT4, SOX2, and KLF4 and a MYC antibody for detection of c-MYC. (B) Representative microscopical pictures of the nuclear extraction. The nucleus (N) of Sf9 cells is shown before and after disruption of the cell membrane. 200x magnification. (C) SDS-PAGE of the purified OSKM factors following affinity chromatography on glutathione-coupled sepharose columns. The full-length proteins are marked with a close arrowhead, and the 27 kD GST fragment of the fusion proteins with an open arrowhead.

To prepare nuclear fractions Sf9 cells were swollen in hypotonic buffer and broken up with a Dounce homogenizer. Representative microscopic pictures show the successful nuclear extraction after disruption of the cell membrane (Fig. 1B). Following lysis of the nuclei for OCT4, SOX2 and KLF4 or of whole cells for c-MYC GST-affinity chromatography was performed. As revealed by silver staining, all transcription factors could be purified to near homogeneity (Fig. 1C). Purified KLF4 showed minor protein bands at 50 and 70 kDa, which were recognized by the KLF4 antibody and, as revealed by matrix-assisted laser desorption/ionization (MALDI) analysis, represented fragments of the transcription factor (data not shown). In addition, a minor GST band was found at 27 kDa, which was most pronounced for c-MYC, presumably due to its purification from cell lysates (Fig. 1C). Importantly, from Sf9 suspension cultures considerable protein yields of the transcription factors could be obtained: i.e. OCT4: 6.1 mg/l; SOX2: 1.5 mg/l; c-MYC: 2.8 mg/l and KLF4: 3.8 mg/l.

Recombinant OSKM factors bind to their DNA consensus motifs in vitro

We next tested the DNA-binding activity of the purified factors using electrophoretic mobility shift assays (EMSAs) with different DNA consensus motifs (Fig. 2). Even a small amount (75 ng) of recombinant OCT4 was sufficient to reveal strong DNA binding to an octamer-binding site from the Ig heavy chain enhancer (1W motif). A slightly weaker DNA binding of OCT4 was detected to the OCT4-binding site from the enhancer of undifferentiated embryonic cell transcription factor 1 (UTF1) which contains an additional adjacent SOX2-binding site (Fig. 2). In addition, also DNA binding of recombinant SOX2 could be demonstrated to the classical DNA consensus site present in the miR-302 promoter as well as to the UTF1 site.³⁶ Moreover, KLF4 bound efficiently to the typical DNA consensus sequence of the Nanog promoter. DNA binding of recombinant c-MYC to the classical E-box element was somewhat weaker but clearly detectable. For all transcription factors DNA-binding was abolished by competition with unlabeled oligonucleotides (data not shown). These results therefore suggest that recombinant OSKM factors retain not only their typical nuclear localization, but also their DNA-binding activity.

Figure 2.

Recombinant OSKM factors retain DNA-binding activity. The purified OSKM factors were subjected to EMSA analysis. The DNA-binding activity of OCT4 was analyzed using oligonucleotides containing the OCT4 1W consensus motif or UTF1 enhancer sequence. SOX2 DNA binding was analyzed using oligonucleotides containing the SOX2 consensus motif or the UTF1 enhancer sequence. Similarly, the DNA-binding activity of KLF4 and c-MYC was tested with oligonucleotides containing the DNA consensus sequences of the transcription factors. Bovine serum albumin (BSA) was used as a negative control. The closed arrowheads indicate the specific protein-DNA complexes and the open arrowheads the unbound oligonucleotides.

Identification of novel AKT phosphorylation sites in OCT4, SOX2 and KLF4

In order to investigate AKT phosphorylation sites of the different transcription factors, we incubated all 4 OSKM factors with active baculovirus-derived recombinant AKT1 in an in-vitro kinase assay. Subsequent immunoblotting of the reactions mixtures with an anti–phospho-(Ser/Thr) AKT substrate antibody, which recognizes AKT phosphorylation motif sequences, revealed that OCT4, SOX2 and KLF4 were efficiently phosphorylated by AKT (Fig. 3A). OCT4 showed several degradation products that were strongly phosphorylated by AKT, whereas no phosphorylation of c-MYC could be detected.

Figure 3.

OCT4, SOX2 and KLF4, but not c-MYC are phosphorylated by AKT in vitro. (A) In vitro kinase assay: The purified OSKM factors were incubated with active AKT1 protein in the absence or presence of ATP, and immunoblotted with the indicated antibodies. (B) Representative mass spectrum demonstrating AKT-mediated phosphorylation of OCT4 at T225 following digestion with LysC.

To identify the AKT phosphorylation sites of the different transcription factors we subjected in vitro phosphorylated as well the control OSKM factors to mass spectrometry (MS)-based phosphoproteomic analysis. As expected, we identified several intrinsic phosphorylation sites on OCT4 and KLF4 that were phosphorylated in the absence of ATP and AKT, indicating that the proteins were modified by the intrinsic phosphorylation machinery of Sf9 cells. Most phosphorylation sites however were dependent on AKT and their MS/MS spectra are depicted in Supplemental Figures 1–9. A representative spectrum of phosphorylation at of OCT4 at T225 is shown in Figure 3B. As published previously,²⁹ OCT4 was found to be further phosphorylated by AKT at T235. Importantly, our phosphoproteomic analysis identified 4 further unreported sites in OCT4 that were phosphorylated by AKT, namely, S136, T159, T225 and S236 (Fig. 4). In addition to OCT4, we found that AKT phosphorylated SOX2 at S83, which has been also yet not described. KLF4 was phosphorylated by AKT at 5 different sites: the already known residue T429, which is homologous to T399 in mouse,³¹ and 4 further novel sites, namely, S19, T33, S234 and S326 (Fig. 4, Supplemental Figures 6–9).

Figure 4.

AKT phosphorylation sites in OCT4, SOX2 and KLF4. The scheme shows the novel and previously reported AKT phosphorylation sites with the functional domains of OCT4, SOX2 and KLF4. OCT4 contains POU-specific and homeodomain (HD) DNA-binding domains as well as transactivation domains located at the N-terminus (N-TAD) and C-terminus (C-TAD). SOX2 has a high mobility group (HMG) DNA-binding domain and a transactivation domain (TAD). KLF4 contains a single N-terminal TAD, a central proline-rich region (PR) and 3 zinc-finger (ZF) DNA-binding domains. NLS marks the nuclear localization sequences of the transcription factors.

Discussion

The self-renewal of pluripotent stem cells is dependent on a coordinated network of a key set of transcription factors, which require a tight regulation of their expression to maintain pluripotency. OSKM factors have not only attracted increased interest for their role in stemness and embryonic pathways, but have been also implicated in tumorigenesis. So far, the mechanisms that regulate the levels of OSKM factors are poorly understood and only partially controlled by transcriptional events. For instance, SOX2 and OCT4 promote their own transcription by cooperative binding to adjacent DNA sites in the promoter regions of their genes.³⁷ Recent evidence highlights an important role of posttranslational modifications in regulating the levels and activity of pluripotency factors.³⁸ However, although the transcriptional targets of these factors have been extensively studied, very little is known about how the proteins are regulated at the posttranslational level.

The main intention of the present work was to establish an efficient in vitro system to study posttranslational modifications of the OSKM factors. To this end, we employed a baculoviral expression system for GST fusion proteins. We demonstrate that, with the exception of c-MYC, all OSKM factors were localized in the nuclear compartment and could be therefore efficiently enriched from nuclear fractions of Sf9 cells. In contrast to our study, previous in vitro phosphorylation studies used recombinant transcription factors expressed in bacteria.^29,33 Bacterial expression systems exhibit several limitations, because the recombinant proteins are mostly localized in inclusion bodies, requiring protein denaturation and an often inefficient refolding.³⁴ In line with this notion, we found that several commercial preparations of bacterially expressed OSKM factors lacked robust DNA-binding activity. In contrast, consistent with their function as transcription factors, the baculovirally expressed OSKM factors revealed strong DNA-binding activity to their consensus sequences.

We chose to study AKT-mediated phosphorylation of the OSKM factors, because, in addition to its established function as a survival factor, AKT is regarded as an important regulator of stemness.^25,29 Moreover, increasing evidence indicates that AKT exerts an essential function in cancer stem cell biology.^39,40 By combining in vitro phosphorylation and phosphoproteomic analyses, we were able to identify several novel putative AKT phosphorylation sites in the OSKM factors with the exception of c-MYC. For OCT4 we confirmed not only the previously reported AKT phosphorylation site T235,^29,33 but also found new phosphorylation sites at S136, T159, T225 and S236. Likewise, for KLF4 we identified the reported AKT target site at S429 ³¹ as well as novel phosphorylation sites at S19, T33, S234 and S326. Despite several attempts, however, we were unable to verify the reported T116 site of SOX2 (equivalent to T118 in the murine protein),²⁷ but identified a novel SOX2 phosphorylation site at S83.

It was reported that AKT-mediated phosphorylation of OCT4 at T235 promotes self-renewal, survival and the tumorigenic potential of embryonal carcinoma cells.²⁹ Mechanistically, AKT-mediated phosphorylation prevents its nuclear export and subsequent cytosolic degradation, resulting in increased stability and transcriptional activity of OCT4. Similarly, phosphorylation of OCT4 at the adjacent S236 site is likely to influence OCT4 activity.⁴¹ Interestingly, the newly discovered site T225 in OCT4 is localized in POU DNA-binding domain, suggesting that phosphorylation at T225 might influence the DNA-binding and transcriptional activity of OCT4. Similar to OCT4, also the novel phosphorylation site of SOX2 at S83 is located within the DNA-binding domain as well as in one of the 2 nuclear localization sequences of SOX2. This might hint at the possibility that AKT-mediated phosphorylation of SOX2 might influence its nuclear import and DNA-binding activity. It should be noted that murine Sox2 can be also phosphorylated at T118 by AKT which not only promotes its stability, but also its activity to reprogram mouse embryonic fibroblasts.²⁷ How the AKT-mediated phosphorylation of KLF4 affects nuclear localization or transcriptional activity remains unknown. Interestingly, both the reported phosphorylation site at T429, which is equivalent to mouse T399,³¹ as well as the newly identified T33 site are located in one of the 2 nuclear localization sequences of KLF4.

With the exception of S326 in human KLF4, all phosphorylation sites that we identified in OCT4, SOX2 and KLF4 are also present in the mouse proteins. The novel sites discovered in this study are therefore valuable candidates for in vivo validation and further functional analyses. Recent evidence demonstrates that, in addition to phosphorylation events, further posttranslational modifications, including ubiquitination, sumoylation, methylation, acetylation or O-GlcNAcylation, may regulate the levels and activity of pluripotency factors.³⁸ Our in vitro system will be also useful to study these posttranslational modifications as well as their potential crosstalk in the OSKM factors. All constructs used in this study will be therefore made available to the scientific community.

Material and Methods

Recombinant protein expression in Sf9 cells

The cDNAs of human OCT4, SOX2, KLF4 and c-MYC were amplified using Pfx polymerase (Life Technologies) and flanked with a flag tag and BamH1 site from the 5′ end and EcoR1 site from the 3′ end. After digestion with EcoR1 and BamH1 (Thermo Scientific) and ligation with T4 ligase, the DNA amplicons were cloned into the insect cell transfer vector pAcG2T (BD Biosciences) encoding N-terminal GST fusion proteins. Following transformation of competent NEB 5-α E. coli the correct cloning of the inserts was verified by restriction analysis using PvuII. Sf9 insect cells were cultured at 27°C in Ex-cell 420 medium (Sigma-Aldrich) supplemented with 10% fetal calf serum (PAA Laboratories). Recombinant baculoviruses were generated through homologous recombination after calcium phosphate transfection of the recombinant pAcG2T transfer vector and Baculogold Bright Linearized DNA (BD Biosciences). After virus amplification and production of high-titer virus stocks protein production of the transcription factors was performed in Sf9 cell suspension cultures. Different multiplicities of infections (MOIs) and incubation times postinfection were explored for each transcription factor to optimize protein production.

Purification of recombinant OSKM factors

Sf9 cells were harvested 2 d postinfection for SOX2, KLF4 and c-MYC, and 3 d postinfection for OCT4 at 2500 g for 10 min. OCT4, SOX2 and KLF4 were isolated from purified Sf9 cell nuclei. To this end, the cell pellet was washed with PBS and resuspended in a hypotonic buffer containing 10 mM Na-HEPES, pH 7.9, 10 mM KCl, 0.1 mM EDTA, 0.1 mM EGTA, 1 mM dithiothreitol and 0.5 mM PMSF. After incubation on ice for 15 min, cells were homogenized with 25 strokes in a Dounce homogenizer. After centrifugation at 7000 g for 10 min at 4°C, the nuclear pellet was resuspended in 5 pellet volumes of extraction buffer (20 mM Na-HEPES, pH 7.9, 400 mM KCl, 1 mM EDTA, 1 mM EGTA, 10% glycerol, 1 mM dithiothreitol, 10 nM aprotinin, 10 µM leupeptin, 0.5 µg/ml pepstatin and 0.5 mM PMSF), and incubated for 30 min at 4°C with rotation. For c-MYC, whole cell extraction was performed by incubating the cells for 45 min in ice-cold RIPA lysis buffer (25 mM Tris-HCl, pH 7.6, 150 mM NaCl, 1% NP-40, 1% sodium deoxycholate, 0.1% SDS) containing protease inhibitors. Both the nuclear and whole cell lysates were centrifuged at 50,000 g for 15 min at 4°C to remove cellular debris and filtered through a 0.45 µm filter. Protein purification was performed using GSTrap FF 1-ml columns (GE Healthcare) coupled to a peristaltic pump. The column was equilibrated with 5 column volumes (CV) of PBS, before the lysates were loaded with a flow-rate of 1 ml/min. After sample application, the column was washed with 10 CV PBS and proteins were eluted in 0.5 ml fractions with 5 CV glutathione elution buffer (10 mM GSH, 50 mM Tris-HCl, pH 8.0) at a flow rate of 0.2 ml/min. All steps were performed at 4°C.

Coomassie/silver staining and immunoblotting

The recombinant proteins were separated by 10% SDS-PAGE. For Coomassie staining, gels were stained for 1 h in staining solution (0.1% Coomassie brilliant blue R-250, 50% methanol and 10% acetic acid) and destained overnight in 10% acetic acid, 10% isopropanol and 10% methanol. For silver staining, gels were fixed for 30 min in 10% acetic acid and 40% ethanol, sensitized for 30 min in 40% ethanol, 8 mM Na₂S₂O₃ x 5 H₂O, 500 mM sodium acetate, and washed thrice for 5 min in H₂O. After staining in 30 mM silver nitrate solution for 30 min gels were washed in H₂O and developed in 100 ml of a solution containing 235 mM Na₂CO₃, 50 µl 37% formaldehyde and 25 µl 10% Na₂S₂O₃ x 5 H₂O. For Western blot analysis proteins were transferred onto polyvinylidenedifluoride membranes (Amersham Biosciences).^42,43 Membranes were blocked in PBS containing 4% BSA and 0.05% Tween-20 for 1 h, followed by an overnight incubation with the primary antibodies in blocking buffer at 4°C. The following antibodies were used: mouse anti-GST (Santa Cruz, clone 1E5, 1:1000), mouse anti-c-MYC (Santa Cruz, clone A-14, 1:1000) and rabbit anti-phospho-(Ser/Thr) AKT substrate (RxRxxS/T) antibody (Cell Signaling, 1:1000). After washing the membrane in TBS/0.05% Tween and incubation with peroxidase-coupled secondary antibodies for 1 h proteins were visualized using ECL reagents (Amersham Biosciences).

Electrophoretic mobility shift assay (EMSA)

EMSA was performed using Odyssey® Infrared EMSA Kit (LI-COR Bioscience) according to the manufacturer's instructions. Briefly, the recombinant proteins were incubated with 1 µl of IRDye® 700 infrared dye-labeled double-stranded oligonucleotides, 2 µl of 10 × binding buffer, 2.5 mM DTT, 0.25% Tween-20 and 1 µg of poly(dI-dC) in a total volume of 20 µl for 20 min at room temperature in the dark. Samples were separated on 4% native polyacrylamide gels in 0.5 × Tris-borate-EDTA buffer. The gel was scanned by direct infrared fluorescence detection on the Odyssey® imaging system. The following oligonucleotides with high-affinity binding sites for the transcription factors were labeled from both 5′ and 3′ ends with IRDye 700: OCT4 (GCCGAATTTGCATATTTGCATGGCTG), UTF1 (CTGAAAGATGAGAGCCCTCATTGTTATGCTAGTGAAGTGCCAAGCTGA), SOX2 (CAGA TAGAAACACAATGCCTTTCTCGGC), KLF4 (GTAGGGGGTGTGCCCGCCAGGAGGGGTGGGTCTAAGGTGATAGAGCCTTC) and c-MYC (GTGTTAATTGGGAGCACGTGTAGGTC). The DNA-binding sites are written in bold. In the UTF1 enhancer element, the OCT4-binding site is written in bold italic and the SOX2-binding site in bold.

Kinase Assays

Kinase assays were carried out essentially as described ^44,45 in kinase buffer containing 25 mM Tris-HCl, pH 7.5, 5 mM β-glycerophosphate, 2 mM DTT, 0.1 mM sodium orthovanadate and 10 mM MgCl₂ with or without 200 μM ATP. Three µg of the recombinant substrates and 0.5 µg of active human AKT1 protein (Millipore) were incubated for 2 h at 30°C.

Mass spectrometry and data analysis

After the kinase assays proteins were processed for MS analyses as described ⁴⁶ with the following changes: In-solution digestion was performed with endoproteinases Lys-C, trypsin or AspN. LC-MS/MS analyses were performed on an EasyLC nano-HPLC (Proxeon Biosystems) coupled to an LTQ Orbitrap Elite mass spectrometer (Thermo Scientific) as described.⁴⁷ For the analyses, 15 most intense precursor ions were sequentially fragmented in each scan cycle (90 min, HCD, top15). The MS data of all experiments were processed using default parameters of the MaxQuant software version 1.2.2.9.⁴⁸ Peak lists were searched against a human target-decoy database (taxonomy id 9606), containing 84946 forward protein sequences and 248 common contaminants, and GST-tagged versions of OCT4, SOX2 and KLF4. The following criteria were applied for the database search: Endoproteinases Lys-C, trypsin or AspN were defined as proteases and 2 missed cleavages were allowed. Carbamidomethylation of cysteine was set as fixed modification; N-terminal acetylation, oxidation of methionine, and phosphorylation of serine, threonine and tyrosine were set as variable modifications. Initial precursor mass tolerance was set to 7 ppm and 20 ppm at the fragment ion level. Identified MS/MS spectra were further processed by MaxQuant for statistical validation and quantification of peptides and protein groups. A false discovery rate of 1% was set at the peptide, protein and phosphorylation site level.

Funding Statement

This study was supported by the Baden-Württemberg Foundation (Adult Stem Cells II Program), the Deutsche Forschungsgemeinschaft (GRK1302, SFB665) and the German Ministry for Education and Research (AID-NET; 01FP09104B).

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Acknowledgments

We thank Boris Maček and Ana Velic for the proteome analyses.