Functionality and Transduction Condition Evaluation of Recombinant Klf4 for Improved Reprogramming of iPS Cells

Abstract

The induced pluripotent stem cell (iPSC) technology holds great potential in regenerative therapy. iPSCs could be induced by proteins (piPSC) linked with poly-arginine cell-penetrating peptides (CPPs) without the risk of genomic alteration, athough with extremely low efficiency and delayed reprogramming. We aimed to evaluate the reprogramming potency of purified mouse Klf4 proteins linked with the CPP of HIV transactivator of transcription (TAT) or Drosophila Penetratin protein at the N- or C-terminus. Eukaryotically expressed recombinant Klf4 targeted cell nucleus while the purified proteins localized in the cytoplasmic and peri-nuclear region. However, using a combined transduction of Klf4 protein and retroviruses expressing Oct4, Sox2, and c-Myc (OSM), we found both TAT- and penetratin-linked Klf4 proteins significantly induced mouse iPSC formation at the nanomolar level in 2 to 4 weeks. Klf4 protein with TAT at the N-terminus showed no reprogramming activity while the fusion protein, with Discosoma red fluorescent protein (DsRed) between TAT and Klf4, exhibited significant iPSC induction function. The iPSCs induced by Klf4 protein and retroviral OSM combinations were pluripotent. Using the protein/retroviral OSM reprogramming assay, we further evaluated Klf4 protein transduction conditions and showed that four continued transductions by purified Klf4 proteins are sufficient for effective iPSC induction. Our results demonstrated for the first time that TAT- and Penetratin-linked Klf4 proteins can effectively replace viral Klf4 in reprogramming fibroblasts, and provided a valuable strategy to evaluate recombinant proteins and transduction conditions for the improvement of piPSC induction efficiency.

Introduction

The induced pluripotent stem cell (iPSC) technology is a ground-breaking advancement in stem cell therapy. It circumvents the ethical concern of embryonic stem cell (ESC) derivation from embryos and has been demonstrated to generate patient- and disease-specific pluripotent cells (Hanna et al., 2007; Okita et al., 2007; Park et al., 2007; Takahashi and Yamanaka, 2006; Takahashi et al., 2007b; Wernig et al., 2007; Yu et al., 2007). Proof-of-principle studies showed that iPSCs could be generated from mouse and human fibroblasts by either purified reprogramming proteins with valproic acid or by whole-cell extracts containing proteins for Oct4, Klf4, Sox2, and c-Myc (OKSM) linked to poly-arginine cell-penetrating peptides (CPPs) (Kim et al., 2009; Zhou et al., 2009). Generation of iPSC by proteins (piPSC) alleviates concerns over genomic alteration by viral integration or by other DNA vectors, which can lead to increased incidence of tumorigenicity (Duinsbergen et al., 2009; Okita et al., 2007). However, the reported piPSC induction efficiencies were considerably low, with 0.006% for mice and 0.001% for humans (Kim et al., 2009; Zhou et al., 2009), compared to 0.05–0.1% and 0.01–0.02%, respectively, by viral induction (Takahashi et al., 2007b; Wernig et al., 2007; Yu et al., 2007). Additionally, the reprogramming time was delayed, with over 30 days and 56 days for mouse and human piPSCs (Kim et al., 2009; Zhou et al., 2009), versus 20 days and 20–30 days for mouse and human iPSCs induction by viruses, respectively (Takahashi et al., 2007b; Wernig et al., 2007; Yu et al., 2007). These problems hinder the practicability of piPSC generation as routine procedures, and highlight the necessity to systematically evaluate the piPSC induction conditions and the potency of recombinant proteins.

An important question to consider when improving piPSC generation is whether the N-terminal or C-terminal linking of CPP to a reprogramming protein would have less adverse impact to its function. Although previous piPSC studies used CPP of poly-arginines (R9 and R11) fused to the C-termini of all OKSM proteins (Kim et al., 2009; Zhou et al., 2009), purified Nanog, Oct4, and Sox2 proteins with R9 fused to the N-termini showed cognate DNA-binding activities, and R9-Sox2 protein upregulated the expression of its target gene Myb upon cell transduction (Yang et al., 2009), indicating certain functionality for these constructs and justifying a systematic comparison of differential tagging.

The R9 and R11 CPPs are similar to the transduction segment of human immunodeficiency virus transactivator of transcription (HIV TAT) (Frankel and Pabo, 1988; Frankel et al., 1988). Bacterially expressed Oct4 and Sox2 with TAT fused to the C-termini exhibited target DNA binding activity, and could rescue the phenotype of mouse ESCs with cellular Oct4 or Sox2 expression knocked down by siRNA (Bosnali and Edenhofer, 2008). However, whether the TAT-fused reprogramming proteins can function in piPSC induction remained undetermined (Bosnali and Edenhofer, 2008; Pan et al., 2009).

We aimed to verify the functionality of TAT and the CPP of Drosophila antennapedia protein, penetratin (Derossi et al., 1994; Edenhofer, 2008) fused reprogramming proteins to induce piPSC from mouse embryonic fibroblasts (MEFs). To this end, we developed a strategy of combined protein/retroviral reprogramming approach using MEFs containing the GFP expression cassette driven by Pou5f1/Oct4 promoter to evaluate differentially constructed reprogramming proteins. We describe here the purification and comparison of TAT- or penetratin-linked Klf4 proteins, which can replace viral Klf4 in reprogramming and promote iPSC induction at nanomolar level. Our results provide proof-of-principle data that the combined protein/viral induction assay is an effective and simple strategy to evaluate and compare the reprogramming potency, as well as the transduction conditions for purified recombinant proteins, in order to improve the piPSC induction efficiency.

Materials and Methods

Plasmid construction

We cloned mouse Klf4 cDNA from pMXs-Klf4 plasmid (Addgene, Cambridge, MA, USA) into both pEcoli–Nterm–6xHN and pEcoli–Cterm–6xHN vectors (Clontech, Mountain View, CA, USA), by PCR, restriction enzyme digestion, ligation, and bacteria subcloning techniques. DNA sequences for TAT peptide (RKKRRQRRR) (Green and Loewenstein, 1988; Joliot and Prochiantz, 2004), penetratin peptide (RQIKIWFQNRRMKWKK) (Derossi et al., 1994) and DsRed monomer (Clontech) were subsequently engineered into these constructs (Fig. 1A and Table 1). The entire DNA fragments encoding the complete protein sequences for constructs TRK, PRK, TR, and PR (Fig. 1A) were subcloned into pIRES plasmid using In-Fusion Advantage PCR Cloning Kit (both from Clontech). All plasmid constructs were verified by DNA sequencing.

FIG. 1.

Purification and characterization of TAT- and penetratin-fused recombinant proteins. (A) Schematic design and abbreviations of the TAT- and penetratin-fusion protein constructs (6xHN: 6 tendon repeats of histidine/asparagine tag, T: TAT, PNT or P: penetratin, R: DsRed, K: Klf4). (B) Coomassie stain of purified proteins for purity assessment. (C) Western blot for verification of purified TAT- and penetratin-fused Klf4 protein using Klf4 antibody. The top bands in TRK and PRK lanes represent TRK and PRK proteins, the lower strong bands in these two lanes possibly represent protein translations from the original start codon of Klf4, which is embedded in the middle of TRK and PRK coding sequences, and have much higher affinity for the Klf4 antibody because it recognizes an epitope at the very N-terminal of Klf4 protein. (D) EMSA assays for recombinant protein/DNA binding using SYBR Green DNA Stain. Upper panel: EMSA using wild-type Klf4 binding oligos; lower panel: EMSA using oligos with mutated Klf4 binding sites. (Wt oligo, wild-type oligo control; Mt oligo, mutant oligo control). (E) Cytoplasmic and peri-nuclear localization of purified recombinant proteins. MEFs were transduced with 60-nM purified proteins and protein localizations were visualized under a UV-fluorescence microscope at 72-h posttransduction (short arrows indicate cell nuclear position, long arrows indicate background fluorescence of debris in culture; bar = 35 μm). (F) pIRES-TRK or PRK plasmids were transfected into Oct4/GFP MEFs for 24 h, after an additional 20-h cell culture the TRK and PRK protein expressions in MEFs were imaged under a fluorescence microscope (arrows indicate the DsRed and DAPI color overlay; bar = 35 μm). (G) pIRES-TRK or PRK plasmids were cotransfected with Klf4-Luciferase Reporter plasmids into HEK293H cells for 24 h. After an additional 20-h cell culture, the firefly luciferase activity in each treatment was measured, with renilla luciferase activity as the internal control. The scale of the Y-axis was arbitrarily set so that the ratio of firefly versus renilla luciferase activity of pIRES vector control equals to 1 unit. Error bars denote standard error. (*p < 0.01 compared to TR, PR, and pIRES vector controls, N = two independent experiments).

Table 1.

Primers and DNA Oligos for Plasmid Construction

Name	Sequence (5′–3′)
Klf4F-S1	TACGGTCGACATGGCTGTCAGCGACGCTC
Klf4R-E1	CGGGAATTCTTAAAAGTGCCTCTTCATGTGTAAGGC
Klf4R-E2	CGGGAATTCGCAAAGTGCCTCTTCATGTGTAAGGC
RedF-E1	GGCGGAATTCTGACAACACCGAGGACGTCATCAAG
RedR-S1	GGCGGTCGACCTGGGAGCCGGAGTGGCG
InFusion-F	CTAGCCTCGAGAATTCAGAAGGAGATATACCATGGGTC
InFusion-R	GGCCGCCCGGGTCGACAGCCAACTCAGCTTCCTCTAG
N-terminal TAT	AACGCTGCAGGTtatggcaggaagaagcggagacagcgacgaagaGGTGTCGACCAG
C-terminal TAT	AACGGCGGCCGCGGAGGAGGTtatggcaggaagaagcggagacagcgacgaagaTAATCTAGAGGCG
Penetratin	AACGCTGCAGGTcgccagatcaagatctggttccagaatcgtcgcatgaagtggaagaagGGTGTCGACCAG

Notes: Sequences highlighted in upper case: restriction digestion sites or fusion sequences for plasmid cloning; lower case: coding sequences for TAT or penetratin.

Primers Klf4F-S1 and Klf4R-E1 were used to clone TRK, TK, PRK, and PK constructs while primer pair Klf4F-S1 and Klf4R-E2 were used to clone KT construct by PCR. RedF-E1 and RedR-S1 were used to subclone DesRed fragment into all constructs.

Recombinant protein purification, verification, and quantification

Protein expression plasmids were transformed into Rosetta 2 (DE3) bacteria (EMD). After overnight inoculation at 37°C in 5 mL LB with antibiotics, cultures were expanded to 50 mL LB to continue growth at 37°C until OD₆₀₀ reached 1.0–1.2. The protein expression was induced with 1 mM isopropyl-β-D-1-thiogalactopyranoside (IPTG, Fisher Scientific, Pittsburgh, PA, USA) and the culture was continued at 37°C for 5 h. Bacteria were collected by centrifugation at 3000 × g for 15 min at 5°C and stored at −80°C until further processing. Recombinant proteins were purified and refolded using the Protein Refolding Kit (Novagen, Gibbstown, NJ, USA) according to manufacturer's instructions. Briefly, bacteria were lysed using lysozyme and sonication. Bacteria inclusion bodies were collected by centrifugation at 10,000 × g for 10 min at 4°C, then solubilized in 20 mM Tris buffer (pH 8.5) with 0.3% N-Lauroylsarkosine and 1 mM dithiothreitol (DTT). Protein refolding was carried out in Tris-buffer and 0.1 mM DTT. Further promotion of disulfide bond formation was performed by dialyzing against 1 mM reduced and 0.2 mM oxidized glutathione (Calbiochem, Gibbstown, NJ, USA) in Tris-buffer at 4°C overnight after removal of DTT. Proteins were further concentrated using the Amico Ultra Centrifugal Filter Unit (Millipore, Billerica, MA, USA) and centrifuged at 4500 × g at room temperature (RT), and stored at −20°C in the presence of 50% glycerol. The identities of purified proteins were verified using 10% SDS-polyacrylamide gel electrophoresis (SDS-PAGE) followed by Western blot using anti-Klf4 rabbit antibody (1:1000, ABR) and HRP conjugated goat antirabbit IgG secondary antibody (1:50,000; Jackson ImmunoResearch, West Grove, PA, USA) or using the Universal His Western Blot Kit 2.0 (Clontech). The purities of the recombinant proteins were verified by SDS-PAGE and Coomassie Stain. Proteins were quantified against protein standard bovine serum albumin (BSA;, Pierce, Rockford, IL, USA) using the Bio-Rad Gel Doc 2000 System and the densitometry module of the Quantity One software (Bio-Rad, Hercules, CA, USA). Two to three independent sample preparation and analysis were performed.

Cell culture, retroviral preparation, and MEF reprogramming assay

Male Oct4/GFP and CD1 MEFs were generated from E13.5 mouse embryos of either a hybrid background of B6;CBA-Tg(Pou5f1-EGFP) 2Mnn/J × B6D2F1/J mice (Jackson Laboratory, Bar Harbor, ME, USA), or CD1 mice (Charles River, Wilmington, MA, USA), respectively. MEFs up to passage 5 were used for reprogramming study. pMXs-mouse Oct4, Sox2, and c-Myc plasmids (Addgene) were transfected into PlatE cells (Morita et al., 2000) for retroviral packaging based on published protocol (Takahashi et al., 2007a). MEFs were plated on six-well-plates at 8.4 × 10⁵ cells per plate, with 10% fetal bovine serum (FBS; Hyclone, South Logan, UT, USA) in DMEM (Invitrogen, Carlsbad, CA, USA). The next day, cells were transfected with retroviral Oct4, Sox2, and c-Myc (OSM) combination. After 24 h of viral transfection, cells were trypsinized and passaged onto two 12-well-plates preseeded with mitomycin C-treated CD1 MEF feeders. On the following day, recombinant proteins were diluted in reprogramming medium and applied to cells. The reprogramming medium was a 1:1 mix of no-serum ESC medium (76% KO-DMEM, 20% KSR, 1% 100 × glutamax, 1% 100 × nonessential amino acids, and 0.5 × penicillin/streptomycin from Invitrogen, supplemented with 1% 100 × β-mercaptoethanol and 1,000 U/mL LIF from Millipore), with serum containing ESC medium (76% DMEM, 20% ESC-qualified FBS from Hyclone, 1% 100 × glutamax, 1% 100 × nonessential amino acids, 0.5 × penicillin/streptomycin, 1% 100 × β-mercaptoethanol, and 1000 U/mL LIF). Protein transductions were carried out for four to eight times, with a 48-, 72-, or 96-h interval between two transductions (Table 2). The GFP expressing colonies and total colonies were counted between 2 to 4 weeks under a Nikon fluorescence microscope, and mature iPSC colonies were picked and expanded at 3 weeks postretroviral transduction in no-serum ESC medium. All statistics were performed using one-way ANOVA with Tukey's multiple comparisons.

Table 2.

Experimental Conditions for Protein Construct Screening and Transduction Optimization Using Combined Klf4 Protein/Retroviral OSM Reprogramming Assay

Experiments	Protein constructs	Transduction dosage	Transduction interval	Protein treatment duration per cycle	Number of transduction cycles
Constructs comparison	TRK, TK, KT, TR, PRK, PK	50 nM, 65 nM	72 h	72 h	7
Protein dosage comparison	TRK	50 nM, 70 nM, 90 nM	72 h	72 h	7
Transduction frequency comparison	TRK	50 nM	48 h, 72 h, 96 h	48 h, 72 h	7
Transduction time comparison	TRK	50 nM	72 hrs	24 h, 72 h	7
Transduction cycle comparison	TRK	50 nM	72 h	72 h	4, 5, 6, 7, 8

Electrophoretic mobility shift assay (EMSA)

For EMSA, 20 pmol of purified proteins were incubated with 20 pmol of Klf4-binding DNA oligo (5′-CCTCACCCAACAATGCACACACCAAGGGGGAACACACCCGAAGCC-3′), or the oligo with Klf4-binding sites mutated (5′- CCTCGTTTAACAATGCACACGTTAAGGGGGAACACGTTTGAAGCC-3′) (Patel et al., 2006), with 1 μg BSA and 1.0 to 1.3 × of 10 × binding buffer (100 mM Tris, 10 mM EDTA, 500 mM KCl, 10 mM DTT, and 50% glycerol) in 15 μL total volume. After 1 h equilibration at RT, the mixtures were resolved on 8% native PAGE gels with 0.5 × TBE. DNA and protein were stained by SYBR Green and SYPRO Ruby respectively per manufacturer's instructions for EMSA Kit (Molecular Probes, Carlsbad, CA, USA), and visualized under the UV light of Bio-Rad Gel Doc 2000 System.

Luciferase reporter assay and recombinant protein localization study

HEK293H cells (Invitrogen) were seeded at 2.5 × 10⁶ cells per 48-well plate. The next day, cells were transfected with the Cignal Klf4 Reporter (luc) plasmids (SABioscience, Frederick, MD, USA) at 100 ng per well, together with 300 ng of pIRES plasmid constructs for 24 h using Fugene 6 transfection reagent. Cells were allowed to grow for 20 h more after transfection before proceeding to luciferase activity measurements using the Dual-Luciferase Reporter Assay Kit (Promega, Madison, WI, USA) and a TD 20/20 luminometer. To verify the localization of recombinant proteins, pIRES plasmid constructs were transfected to Oct4/GFP MEFs in 48-well plates using Fugene 6 reagent for 24 h. After an additional 20 h, cells were stained with DAPI (Thermo Scientific, Logan, UT, USA) and the expression of DsRed fusion proteins was visualized and imaged under a Nikon fluorescence microscope.

MTT assay

Oct4/GFP MEFs (8.4 × 10⁵) were plated into two 96-well plates. After 2 days, proteins were added at 0 to 80 nM to each well for 24 h. Cell mitochondrial dehydrogenase activity in each well was measured using Tox1 Kit (Sigma, St. Louis, MO, USA) and a BioTek 96 well plate reader at OD₅₇₀ nm per manufacturer's instructions.

Embryoid body (EB) formation

Established iPSC lines (passage 6) were passaged on CD1 MEF feeders. After colonies grow for several days, cells were trypsinized and replated on the original plate for 2 h to allow MEFs to attach. The iPSCs in the medium were collected and subsequently plated onto Petri-dish with regular medium (10% FBS in DMEM) without LIF. Upon 1 week of differentiation, the EBs were plated back to 0.1% gelatin (Invitrogen)-coated cell culture dish, the cells were allowed to reattach and continue differentiation for 1 week before proceeding to RNA extraction and RT-PCR as described bellow.

PCR, RT-PCR, alkaline phosphatase (AP) staining, and immunostaining

Genomic DNA was isolated from iPSCs using DNeasy Mini Kit (Qiagen, Valencia, CA, USA). To verify the OSM retroviral DNA integration, 150 ng of genomic DNA was used for each PCR reaction with primers specific for both the pMXs vector and Oct4, Sox2, Klf4, and c-Myc cDNA sequences, using Taq DNA Polymerase kit (Invitrogen) and 38 PCR cycles. Total RNA was isolated from iPSCs or EBs using RNeasy Mini Kit (Qiagen). Semiquantitative RT-PCR reaction was carried out using 300 ng of total RNA and SuperScript III One-Step RT-PCR System with Platinum Taq (Invitrogen), to evaluate endogenous and viral gene expression using specific primers (Table 3). AP staining was done using Vector Red Alkaline Phosphate Substrate Kit I (Vector Laboratories, Burlingame, CA, USA). For immunostaining, cells were fixed in 4% paraformaldehyde with 1% sucrose in PBS for 15 min at RT, permeabilized in 0.5% TX-100, then incubated for 1 h in 5% goat serum with anti-SSEA1 IgM (1:50), anti-Sox2 IgG (1:100), or anti-Nanog IgG (1:100) antibodies (Chemicon, Billerica, MA, USA), washed in PBS, and then blotted with Alex Fluor 546 or 594 conjugated goat secondary antibodies (1:500; Jackson ImmunoResearch). After wash, cells were stained with DAPI and mounted. Fluorescent images were taken using the Nikon fluorescence microscope.

Table 3.

Primers for RT-PCR and Viral DNA PCR Genotyping (GT)

Name	Sequence (5′–3′)	Accession number
GT-pMXs-S1811	GACGGCATCGCAGCTTGGATACAC
GT-pMXs-S2	CCCTTGAACCTCCTCGTTCGACC
GT-Oct4-R	TCAGTTTGAATGCATGGGAGAGCCCA	NM_013633
GT-Klf4-R	TTAAAAGTGCCTCTTCATGTGTAAGGC	NM_010637
GT-Sox2-R	GCTTCAGCTCCGTCTCCATCATGTT	NM_011443
GT-cMyc-R	GTCGAGGTCATAGTTCCTGTTGG	NM_010849
RT-Oct4-F	TCT TTC CAC CAG GCC CCC GGC TC	NM_013633
RT-Oct4-R	TGC GGG CGG ACA TGG GGA GAT CC
RT-Sox2-F	TAG AGC TAG ACT CCG GGC GAT GA	NM_011443
RT-Sox2-R	TTG CCT TAA ACA AGA CCA CGA AA
RT-Nanog-F	AGCCTCCAGCAGATGCAAGA	AB093574
RT-Nanog-R	GCACTTCACCCTTTGGTTTTGAA
RT-Klf4-F	GCGAACTCACACAGGCGAGAAACC	NM_010637
RT-Klf4-R	TCGCTTCCTCTTCCTCCGACACA
RT-c-Myc-F	TGACCTAACTCGAGGAGGAGCTGGAATC	NM_010849
RT-c-Myc-R	AAGTTTGAGGCAGTTAAAATTATGGCTGAAGC
RT-β-Actin-F	GATGGTGGGAATGGGTCAGA	NM_007393
RT-β-Actin-R	CGTCCCAGTTGGTAACAATGC
RT-GAPDH-F	TGTGTCCGTCGTGGATCTGA	XM_001473623
RT-GAPDH-R	GATGCCTGCTTCACCACCTT
RT-AFP-F	GGCTTTCTAAACACCCATCG	NM_007423
RT-AFP-R	AGTGCGTGACGGAGAAGAAT
RT-ALB-F	AGGCGACTATCTCCAGCAAA	NM_009654
RT-ALB-R	AGTTGGGGTTGACACCTGAG
RT-SMA-F	TGTGAAGAGGAAGACAGCACA	NM_007392
RT-SMA-R	ACATACATGGCGGGGACAT
RT-Brachyury-F	CCGGTGCTGAAGGTAAATGT	NM_009309
RT-Brachyury-R	CCTCCATTGAGCTTGTTGGT
RT-Nestin-F	AGCAGGAGAAGCAGGGTCTA	NM_016701
RT-Nestin-R	TGGGAACTTCTTCCAGGTGT
RT-Musashi-F	CAGCCAAAGGAGGTGATGTC	NM_008629
RT-Musashi-R	GCGCTGATGTAACTGCTGAC

Notes: Primers containing “GT” in their names denote those used for detection of retroviral DNA integration by PCR, and were also used in RT-PCR to detect viral gene expression.

Primers containing “RT” in their names denote those used in RT-PCR for endogenous gene expression.

Chimera formation and genital ridge isolation

CD1 female mice (Charles River) were superovulated and mated to CD1 males. Zygotes were isolated from plugged females and cultured to blastocysts in KSOM medium (Millipore). Blastocysts were injected with iPSCs and transferred to 2.5 days postcoitum (dpc) pseudopregnant recipient female CD1 mice. Cesarean sections were performed at 13.5 dpc and the genital ridges of the embryos were isolated under a dissection microscope and GFP expression was visualized under a Nikon fluorescence microscope. All animal work described here have been approved by International Animal Care and Use Committee at the University of Connecticut.

Results

Specific DNA binding and subcellular localization of recombinant Klf4 protein constructs

We constructed, expressed, and purified the TAT- and penetratin-fused Klf4 constructs TRK, TK, KT, PRK, and PK, with TR, PR as controls (Fig. 1A). The recombinant proteins showed relative high purity (Fig. 1B), and the identity of the proteins was verified by Western blot (Fig. 1C). We first tested the target DNA binding activities of our purified proteins. Only PRK showed detectable specific DNA binding by eletrophoretic mobility shift assay (EMSA) (Fig. 1D). These results indicated that the majority of the protein constructs (except for PRK) were not refolded properly during purification, even though they were purified under the same condition.

Klf4 has been shown to be exclusively nuclear-localized due to the presence of two potent nuclear localization signals within its C-terminus (Shields and Yang, 1997; Shields et al., 1996). To verify the cellular location of the various forms of recombinant Klf4, we determined whether the prokaryotically expressed Klf4 constructs and their eukaryotically expressed counterparts were able to target the nucleus. We transduced purified PRK and TRK proteins into Oct4/GFP MEF cells and examined their cellular localization after 24 to 72 h by the live DsRed fluorescence under a UV-fluorescence microscope. A high percentage of MEFs showed positive DsRed signal within cells, and TRK and PRK both localized in the cytoplasmic and peri-nuclear region, no obvious nuclear localization was observed (Fig. 1E). This is consistent with many other studies that showed apparent cytoplasm and/or peri-nuclear localization for the purified reprogramming proteins, including Oct4, Sox2, and Nanog fused with TAT or R9 at either the N- or C-terminus as a result of the endosomal transportation (Bosnali and Edenhofer, 2008; Yang et al., 2009).

When TRK and PRK were expressed directly in MEFs by plasmid transfection, however, we observed exclusive nuclear localization of these proteins in transfected cells (Fig. 1F). Furthermore, using the Klf4-luciferase reporter assay, we found that both TRK and PRK expressed in HEK293H cells showed significant transcriptional activities compared to the TR and PR controls (Fig. 1G). These data demonstrate that the TRK and PRK constructs with their native protein conformations were able to enter the nucleus and exert transcription functionality similar to the wild-type Klf4. Taking together, our data indicate that the reasons why our purified Klf4 proteins could not effectively target cell nucleus is not because of inappropriate construct design, but rather because of either protein misfolding or the endosomal sequestration for the majority of transduced proteins due to inadequate intracellular trafficking as previously suggested (Caron et al., 2004; Wadia et al., 2004).

Purified recombinant Klf4 proteins exert reprogramming activity in combined protein/retroviral induction assay

Although we did not observe an obvious nuclear targeting for our purified Klf4 proteins, it had been shown that a small percentage of internalized TAT-linked proteins could escape the endosomal vesicles and translocate into the nucleus (Bosnali and Edenhofer, 2008; Tunnemann et al., 2006), and piPSCs were generated by purified proteins with apparent cytoplasmic localizations (Zhou et al., 2009). We therefore reasoned that the small percentage of proteins entering the nucleus may be sufficient to promote reprogramming. We then tested the iPSC induction function of our purified Klf4 proteins using retroviral OSM transfected Oct4/GFP MEFs. We started with a seven-cycle protein transduction protocol (Fig. 2A and Table 2) with 72-h intervals between each transduction, by applying TRK, PRK, TK, and TR proteins to MEFs from the second day of retroviral infection. We used 50 nM protein for each transduction, based on both MTT assay data, which showed cytotoxicity to MEFs at 80 nM TRK but not under 60 nM (Fig. 2B), and our observation on MEF cell morphological changes under different Klf4 concentrations (data not shown). Colonies with ESC-like morphology emerged from TRK and PRK treatments 1 week after viral transduction, and at 2 weeks the GFP+ colonies could be seen in TRK and PRK treated wells under fluorescence microscopy (Fig. 2C). We counted the GFP+ colonies on days 19 and 23. Similar to human fetal fibroblasts (Park et al., 2007), viral OSM induced iPSC colonies (Fig. 2D) at a low efficiency (0.006 ± 0.005% at day 23). Transduction with TRK and PRK recombinant proteins, however, significantly promoted iPSC colony generation (Fig. 2D and E), with the induction efficiencies of 0.042 ± 0.003% and 0.031 ± 0.003%, respectively, at day 23. Transduction by Klf4 proteins also induced a number of ESC-like but GFP negative (GFP−) colonies, which represent intermediate reprogramming stage (Fig. 2F). A dynamic increase in both the numbers of fully (GFP+) and partially (GFP−) reprogrammed iPSC colonies by TRK treatment was observed from days 19 to 23 (Fig. 2F and Table 4). However, transduction with TK at 50 nM showed no effect on iPSC induction (Fig. 2D and E).

FIG. 2.

Purified Klf4 proteins significantly promote the MEF cell reprogramming. (A) Schematic design of combined protein/retroviral iPSC induction approach. Cells were transduced with proteins for four to eight times in different experiments as specified in the main text. GFP+ and/or total colonies were counted at days 16, 19, 22, or 26 under a fluorescence microscope. (B) Oct4/GFP MEFs were transduced with TRK proteins from 0 to 80 nM, MTT assays were performed 24 h after transduction, the absorbance of dissolved MTT formazan was measured at 570 nm using a 96-well plate reader. The error bars denote standard error derived from quantification of three separate wells of cells. (C) GFP+ colonies from Oct4/GFP MEFs by combined TRK or PRK and retroviral OSM reprogramming (bar = 300 μm). (D) Number of GFP+ colonies induced by combination of viral OSM and 50-nM recombinant proteins. Error bars denote standard error. (*p < 0.01 compared to TR control and TK, N = two independent experiments). (E) Alkaline phosphatase staining show colonies induced under different treatment conditions at 32 days postviral OSM transfection. (F) Numbers of GFP+ and total ESC-like colonies induced by the combination of viral OSM and 50 nM to 65 nM of recombinant proteins. The error bars denote standard error derived from quantification of three separate wells of cells. (G) Number of GFP+ colonies induced by combination of viral OSM and 50-nM recombinant proteins. TRK-1 and TRK-2 denotes two batches of independently purified TRK proteins. The error bars denote standard error derived from quantification of three separate wells of cells.

Table 4.

Number and Percentage of GFP+ Colonies Induced by Combined Protein/Retroviral Transduction (Mean ± Standard Error)

		TRK		PRK		TR control
GFP positive colonies		50 nM	65 nM	50 nM	65 nM	50 nM	65 nM
Day 19	Number^a	5.3 ± 0.9	4.0 ± 0.6	4.7 ± 0.7	2.7 ± 0.9	0.3 ± 0.3	0.0
	Percentage^b	12.7 ± 2.3	13.1 ± 2.1	12.3 ± 1.4	7.5 ± 2.4	—	—
Day 23	Number^a	14.0 ± 2.7	11.0 ± 2.9	11.7 ± 2.9	8.5 ± 0.4	0.7 ± 0.7	0.3 ± 0.3
	Percentage^b	21.6 ± 4.4	13.7 ± 2.5	26.8 ± 3.8	25.0 ± 1.2	—	—

Number of GFP+ colonies per 35,000 MEF cells.

Percentage of GFP+ colonies out of the total number of induced ESC-like colonies.

We then compared the induction efficiency among TRK, KT and PK proteins at 50 nM using another batch of Oct4/GFP MEFs, and counted the GFP+ iPSC colonies at days 16 and 22 postretroviral transfection. The overall GFP+ colony numbers induced from this batch of MEFs appear to be higher; however two independently prepared batches of TRK proteins showed almost identical induction efficiency (Fig. 2G). KT protein also induced more iPSC colonies than controls, but PK treatment showed much less induction efficiency than TRK and KT (Fig. 2G).

The fact that TRK and KT proteins did not cause mobility shift in EMSA but could reprogram MEFs suggests a functional restoration for these proteins similar to what evidenced previously (Kwon et al., 2000) during cell transduction. We found KT but not TK protein showed reprogramming activity, which indicates that direct linking of TAT, a strong cationic/amphipathic peptide to the N-terminus of Klf4 may interfere with its function. Interestingly, TRK protein, which has DsRed inserted between the N-terminal linked TAT and Klf4 showed significant reprogramming activity (Fig. 2D and G). These data suggest a sequence dependent impact to Klf4 function by its N-terminal linkers.

iPSCs induced by recombinant Klf4 and viral OSM are pluripotent

We picked TRK and PRK induced GFP+ colonies at 3 weeks postviral transduction and tested their ability to form iPSC lines. Eight TRK iPSC lines out of eight and three PRK iPSC lines out of three picked colonies were established, and termed TRK-iPSCs or PRK-iPSCs, respectively. These cells grew similarly as mouse ESCs, maintaining bright Oct4/GFP expression and ESC-like colony morphology (Fig. 3A). Because TRK- and PRK-iPSC lines were morphologically indistinguishable from each other, we further characterized the TRK-iPSC lines. Genomic DNA PCR confirmed the Oct4, Sox2, and c-Myc but no Klf4 viral DNA integration in these cells (Fig. 3B). Semiquantitative RT-PCR analysis revealed that the TRK-iPSCs (passage 6) expressed pluripotent genes Oct4, Sox2, and Nanog as well as Klf4 and c-Myc endogenously at levels similar to mouse ESCs (Fig. 3C). The Oct4/GFP MEFs expressed Klf4 and c-Myc endogenously, which is similar to previous findings for mouse and human fibroblasts (Park et al., 2007; Rowland et al., 2005). The Oct4 retroviral expression was completely silenced, whereas various levels of viral c-Myc and low levels of viral Sox2 were still expressed in some of the iPSC lines (Fig. 3D). Immunostaining showed that these cells had strong AP activity and expressed pluripotency markers including Oct4 (indicated by Oct4/GFP expression), Sox2, SSEA-1, and Nanog (Fig. 4A). In vitro differentiation analysis by EB formation showed that these iPSCs formed EBs upon withdrawal of LIF and differentiated into cells of three germ layers (Figs. 4B and C). Most importantly, the chimera study demonstrated that these iPSCs incorporated into mouse embryos and formed black eyed chimeric embryos at 13.5 dpc (Fig. 4D). Contribution of iPSCs to the germline was shown by the Oct4/GFP expression of germ cells derived from TRK-iPSCs in both male and female gernital ridges of 13.5 dpc embryos (Fig. 4E). Taking together, our data demonstrated that the iPSCs we generated by Klf4 protein/retroviral OSM transduction are pluripotent, similar to the ESCs.

FIG. 3.

Characterization of iPSCs induced by combined Klf4 protein and viral OSM transduction—Part one. (A) Established TRK-iPSC line at passage 4 (bar = 140 μm) and PRK-iPSC line at passage 2 (bar = 350 μm) expressing Oct4/GFP and with ESC-like colony morphology. (B) Genomic DNA PCR indentifies OSM viral integration for TRK-iPSC lines #4 and #8, pMXs OKSM plasmids served as positive controls, and genomic DNA from Oct4/GFP MEFs was the negative control. (C) RT-PCR detecting endogenous gene expression for R1 ESC, TRK-iPSC lines #4 and #8, and Oct4/GFP MEF. β-actin was used for sample normalization. (D) RT-PCR detecting viral gene expression for Oct4/GFP TRK-iPSC lines#4 and #8. pMXs OSM plasmids served as positive controls, and RNA from Oct4/GFP MEFs as negative control.

FIG. 4.

Characterization of iPSCs induced by combined Klf4 protein and viral OSM transduction—Part two. (A) AP stain and immunostaining of TRK-iPSC line#8. Pluripotency markers (red or orange) were detected by AP stain or by Alex Fluor 546 or 594 conjugated secondary antibodies and visualized under a fluorescent microscope, Oct4 expression was indicated by GFP fluorescence (bar = 70 μm). (B) EB bodies formed from TRK-iPSC lines#4 and #8 at days 1, 5, as well as at day 12 when differentiated cells already reattached (bar = 200 μm). (C) RT-PCR showing the expression of marker genes for three germ layers in cells derived from TRK-iPSC formed EBs. (D) TRK-iPSCs incorporated into 13.5 dpc mouse embryos. Left picture: chimera mouse embryo with black eyes; right picture: wild-type CD1 mouse embryo with pink eyes. (E) Oct4/GFP+ cells derived from TRK-iPSCs incorporated into the genital ridges of 13.5 dpc male (left picture) and female (right picture) mouse embryos.

Reprogramming by recombinant Klf4 is dosage dependent and four continued protein transductions are sufficient for effective iPSC induction

We reasoned that suboptimal protein transduction conditions contributed to the low induction efficiency for piPSCs (Kim et al., 2009; Zhou et al., 2009); therefore, factors like transduction dosage, frequency, treatment time, and number of transduction cycles needed to be further evaluated. We had observed decreased number of GFP+ iPSC colonies by TRK and PRK induction at 65 nM compared to those at 50 nM (Fig. 2F and Table 4). We further confirmed this dosage dependence of reprogramming by TRK protein transduction from 50 to 90 nM using the same protein/retroviral reprogramming assay described earlier (Fig. 2A and Table 2). 50 nM TRK induced a greater number of GFP+ iPSC colonies than 70 nM and 90 nM TRK treatments (Fig. 5A). This corresponds with our observed cell toxicity by TRK at 80 nM (Fig. 2B), and suggests that a balance between transduction dosage and cellular toxicity is crucial to effective reprogramming by recombinant proteins.

FIG. 5.

Evaluation of Klf4 protein transduction conditions for improved reprogramming. (A) Number of GFP+ colonies induced by the combination of viral OSM and 50 nM to 90 nM TRK proteins. The error bars denote standard error derived from quantification of three separate wells of cells. (B) Schematic design of iPSC induction assay for TRK protein transduction frequency and treatment duration tests. TRK proteins (50 nM) were applied with different intervals or times. Arrows indicate the days non-TRK containing medium was applied. GFP+ colonies were counted on days 16 and 22 under a fluorescence microscope. (C) Number of GFP+ colonies induced by 50 nM TRK at 48-, 72-, and 96-h intervals. The error bars denote standard error derived from quantification of three separate wells of cells. (D) Number of GFP+ colonies induced by 50 nM TRK with 72-h continued transduction or 24-h transduction plus 48 h no-TRK medium to cells, for a total of seven cycles. The error bars denote standard error derived from quantification of three separate wells of cells. (E) Number of GFP+ colonies induced by 50 nM TRK for four to eight times with 72-h transduction interval. The error bars denote standard error derived from quantification of three separate wells of cells.

We then compared the effect to reprogramming efficiency by different protein transduction frequencies, with intervals of 48, 72, or 96 h between two transductions. For treatment with 96-h intervals, 50 nM TRK was added to MEFs for 72 h and then replaced by no-TRK reprogramming medium for an additional 24 h to avoid exhaustion of nutrients (Fig. 5B and Table 2). Transduction by TRK with 48-h interval induced slightly less GFP+ iPSC colonies than that of the 72-h interval; however, transduction with 96-h intervals resulted in very few GFP+ colonies similar to the control (Fig. 5C). Because the 96-h interval treatment was essentially the same as the 72-h treatment with the exception of an extra 24-h no-TRK medium, this result suggested that a continued presence of TRK in the media is required for efficient MEF cell reprogramming. To further verify this, we compared the effect to iPSC induction by different TRK treatment duration within each cycle, with either 72 h continued protein exposure to cells or the discontinued transduction, which is 24-h TRK transduction followed by 48 h of no-TRK medium to MEFs (Fig. 5B and Table 2). Again, we observed a sharp decrease in the number of GFP+ colonies when MEFs were exposed to 50 nM TRK for only 24 h, compared to the 72-h TRK exposure for each cycle (Fig. 5D). This result argues the notion that the presence of extracellular Klf4 proteins in a continued fashion is crucial to effective MEF cell reprogramming.

In our Klf4 reprogramming study, GFP+ colonies usually emerged at 2 weeks postviral transfection, following with a rapid climb in numbers of GFP+ colonies within 1 week. It has been shown that 8 to 12 days continued expression of viral reprogramming factors is sufficient for iPSC induction (Brambrink et al., 2008; Stadtfeld et al., 2008). We then asked whether our protein induction follows the same time frame, and whether additional Klf4 transductions are necessary for the rapid increase of GFP+ colonies between 2 and 3 weeks. We treated viral OSM infected MEFs by four to eight cycles of 50 nM continued TRK transduction with a 72-h interval, and counted the GFP+ colonies at days 16, 22, and 26 (Fig. 2A and Table 2). We found that a four-cycle transduction regimen (from days 2–14) resulted in similar GFP+ colony number as the seven-cycle transduction until day 22 (Fig. 5E), suggesting that additional TRK is not necessary for the rapid GFP+ colony number increase between weeks 2 and 3. In fact, the GFP+ colonies continued to develop, and at day 26 the average number of GFP+ colonies by the four-cycle transduction is approximately 1.5 × more than the seven- and eight-cycle transductions (Fig. 5E), yielding an iPSC induction efficiency of 0.22 ± 0.05%. These results correlate well with the aforementioned 8- to 12-day time frame for reprogramming factors (Brambrink et al., 2008; Stadtfeld et al., 2008), and demonstrate that a 12-day TRK exposure (four-cycle transduction at 72-h intervals) is sufficient for effective MEF reprogramming. It is interesting that the five- to eight-cycle TRK transductions all resulted in less GFP+ colonies than the four-cycle transduction, indicating a suppressive effect to reprogramming in our assay by additional Klf4 exposure after day 14. This may have resulted from an imbalance between core pluripotency regulatory proteins including Oct4, Sox2, and Klf4 (Huang et al., 2009; Liu et al., 2008; Sridharan et al., 2009) due to the gradual retroviral OSM silencing in MEFs while the extraneous Klf4 protein continued to be present in the medium.

Discussion

Klf4 belongs to the Krüppel-like factor family, which is characterized by three C2H2 zinc-finger DNA binding domains at the very C-terminus (Shields et al., 1996). It synergizes with Oct4, Sox2, and c-Myc during the reprogramming process by co-binding to the promoters of genes that regulate core developmental signal networks for pluripotency (Huang et al., 2009; Liu et al., 2008; Sridharan et al., 2009). In pluripotent cells, Klf4 regulates ESC self-renew and the pluripotency maintenance, at least partially by regulating Nanog expression (Jiang et al., 2008; Li et al., 2005; Zhang et al., 2010). We report here the purified Klf4 proteins attached with TAT or Penetratin have significant reprogramming effects at nanomolar level, with an average of five to seven times improvement in iPSC induction efficiency compared to control of viral OSM induction in our assay. We found Klf4 protein with direct fusion of TAT to its C- but not N-terminus showed reprogramming activity, whereas the protein construct with DsRed between N-terminal TAT and C-terminal Klf4 exhibited consistent reprogramming function upon cell transduction.

We also showed that the reprogramming by Klf4 protein is dosage dependent and requires a continued presence of Klf4 in the medium. Previous piPSC study reported the use of 8 μg/mL of each purified protein for transduction (Zhou et al., 2009), which equals to about 150 nM Klf4 in the medium, and is three times greater than the dosage used here. Frequent protein transduction at high dosage can lead to cell toxicity, as observed by others (Kim et al., 2009) and by us, which may explain why the discontinued protein transduction strategy (overnight transduction followed by 36-h or 7-day cell culture in no-reprogramming protein medium, respectively) was favored in those pioneer piPSC studies (Kim et al., 2009; Zhou et al., 2009). However, we found that discontinued Klf4 protein transduction resulted in poor reprogramming efficiency even though the MEFs endogenously express Klf4. It is noticeable that the intracellular level of transduced protein decreased more than 40% by 8 h after withdrawal of the protein from culture medium (Caron et al., 2004). These data suggest that the continued high level of reprogramming protein in the medium is necessary during the early stage of piPSC induction. However, the protein dosage and cellular toxicity effects need to be balanced for all-protein piPSC induction approach.

The Klf4 protein transcription activation domain resides in its N-terminus, which is rich in acidic residues and interacts with p300/CBP coactivators, part of the histone acetyltransferase complex (Geiman et al., 2000). The Klf4 zinc finger-containing C-terminus was also shown to interact with Oct4 and Sox2 directly (Wei et al., 2009). Therefore, it is important to verify the impact of N- and C-terminal linking of CPP to Klf4 protein functions. We found that the purified recombinant TK showed no reprogramming activity, which suggests TAT may negatively affect Klf4 function when linked directly to its N-terminus, probably by interfering with the clusters of acidic residues of Klf4 through its strong cationic/amphipathic property. PK protein, with highly cationic penetratin linked to the N-terminus also showed much less reprogramming activity than KT and PRK. It is not likely that the N-terminal 6xHN tag in TK and PK affects Klf4 function, because Klf4 with a similar 6xHis tag at the N-terminus demonstrated strong transcription regulation and target DNA binding activity (Brembeck and Rustgi, 2000; Jenkins et al., 1998). Moreover, we found that TRK protein, which has DsRed inserted between the N-terminal TAT and C-terminal Klf4, showed significant reprogramming activity. Insertion of DsRed may have reduced the potential interference between TAT and N-terminus of Klf4. However, we could not exclude the possibility that it is a phenomenon associated solely with the prokaryotic expression of Klf4 proteins. Evaluating the function of eukaryotic expression of TK and PK, such as by reporter assay would be helpful for further clarification.

TAT and penetratin have been the most widely used CPPs in protein delivery studies (Fischer et al., 2005). It has been suggested that recombinant proteins can undergo refolding in target cells upon cell transduction, probably through endogenous chaperons (Kwon et al., 2000; Nagahara et al., 1998; Schwarze et al., 2000). The fact that TRK and KT proteins showed significant reprogramming activity in MEFs even though we could not detect any positive DNA binding by EMSA may be explained by this. Similar to what we observed for purified Klf4, other studies had shown that the purified reprogramming proteins mainly remained in the cytosolic and peri-nuclear region and were hardly detectable in the nucleus after transduction (Bosnali and Edenhofer, 2008; Caron et al., 2004; Yang et al., 2009), although we had shown clearly that TRK and PRK localize exclusively in cell nucleus when expressed directly by MEFs. Inadequate intracellular trafficking and limited endosomal release had been suggested to be the cause (Caron et al., 2004; Wadia et al., 2004), and represent a bottleneck for the recombinant protein transduction by CPP-mediated endocytosis (Fischer et al., 2005). Reagents or peptides interfering with the endosomal integrity may help to improve the nuclear targeting of our proteins and further boost the Klf4 protein reprogramming efficiency.

It has been reported that the small molecule kenpaullone can replace Klf4 in iPSC induction from MEFs in combination with viral OSM factors (Lyssiotis et al., 2009). However, the reprogramming time is prolonged to 25–30 days, compared to only about 2 weeks by viral OKSM induction. In our study, Oct4/GFP+ colonies induced by Klf4 proteins appeared at 2 weeks similar to all-viral induction. Kenpaullone does not directly induce endogenous Klf4 expression (Lyssiotis et al., 2009), which may explain the delay in its reprogramming.

Viral based iPSC induction assays have been used to screen new reprogramming factors or small molecules for replacement of current factors by many groups (Heng et al, 2009; Huangfu et al., 2008; Ichida et al., 2009; Shi et al., 2008); however, no assay had been reported for evaluation of purified reprogramming proteins. We reported here the functional test for TAT- and penetratin-linked Klf4 recombinant proteins, using combined protein/retroviral OSM transduction. We demonstrate that it is a feasible and less tedious method to determine and compare the potency of various reprogramming protein constructs, as well as to evaluate the protein transduction conditions, compared to the all-protein-induction approach. We are currently evaluating the recombinant Oct4, Sox2, and Myc proteins using similar assays. Our finding that four cycles of continued Klf4 exposure at nanomolar level with a 72-h transduction interval is sufficient for effective iPSC induction provides further knowledge to help in optimizing the piPSC induction strategy.

Footnotes

Acknowledgments

We greatly appreciate Dr. David Goldhamer, Dr. Carolyn Teschke, Dr. Lawrence Silbart, and Dr. Yinghong Ma for their critical input to this project. We thank Dr. Chul Kim and Dr. Tomokazu Amano for MEF cell preparation, and Dr. Toshio Kitamura in the University of Tokyo for PlatE cells. This project was supported by USDA Grant 1265-31000-091-02S.

Author Disclosure Statement

The authors indicate herein that they have no potential conflicts of interest.

Author contributions

Y.T.: conception and design, provision of study material, collection, and/or assembly of data, data analysis, and interpretation, and manuscript writing. C-J.L.: provision of study material, collection, and analysis of data. X.C.T.: conception and design, data analysis, and interpretation, financial and administrative support, and manuscript writing.

References

Bosnali

, Edenhofer

2008. Generation of transducible versions of transcription factors Oct4 and Sox2. Biol. Chem., 389:851–861.

Brambrink

, Foreman

, Welstead

G.G.

et al. 2008. Sequential expression of pluripotency markers during direct reprogramming of mouse somatic cells. Cell Stem Cell, 2:151–159.

Brembeck

F.H.

, Rustgi

A.K.

2000. The tissue-dependent keratin 19 gene transcription is regulated by GKLF/KLF4 and Sp1. J. Biol. Chem., 275:28230–28239.

Caron

N.J.

, Quenneville

S.P.

, Tremblay

J.P.

2004. Endosome disruption enhances the functional nuclear delivery of Tat-fusion proteins. Biochem. Biophys. Res. Commun., 319:12–20.

Derossi

, Joliot

A.H.

, Chassaing

et al. 1994. The third helix of the Antennapedia homeodomain translocates through biological membranes. J. Biol. Chem., 269:10444–10450.

Duinsbergen

, Salvatori

, Eriksson

et al. 2009. Tumors originating from induced pluripotent stem cells and methods for their prevention. Ann. N. Y. Acad. Sci., 1176:197–204.

Edenhofer

2008. Protein transduction revisited: novel insights into the mechanism underlying intracellular delivery of proteins. Curr. Pharm. Des., 14:3628–3636.

Fischer

, Fotin-Mleczek

, Hufnagel

et al. 2005. Break on through to the other side-biophysics and cell biology shed light on cell-penetrating peptides. Chembiochem., 6:2126–2142.

Frankel

A.D.

, Pabo

C.O.

1988. Cellular uptake of the tat protein from human immunodeficiency virus. Cell, 55:1189–1193.

10.

Frankel

A.D.

, Bredt

D.S.

, Pabo

C.O.

1988. Tat protein from human immunodeficiency virus forms a metal-linked dimer. Science, 240:70–73.

11.

Geiman

D.E.

, Ton-That

, Johnson

J.M.

et al. 2000. Transactivation and growth suppression by the gut-enriched Kruppel-like factor (Kruppel-like factor 4) are dependent on acidic amino acid residues and protein-protein interaction. Nucleic Acids Res., 28:1106–1113.

12.

Green

, Loewenstein

P.M.

1988. Autonomous functional domains of chemically synthesized human immunodeficiency virus tat trans-activator protein. Cell, 55:1179–1188.

13.

Hanna

, Wernig

, Markoulaki

et al. 2007. Treatment of sickle cell anemia mouse model with iPS cells generated from autologous skin. Science, 318:1920–1923.

14.

Heng

J.C.

, Feng

, Han

et al. 2009. The nuclear receptor Nr5a2 can replace Oct4 in the reprogramming of murine somatic cells to pluripotent cells. Cell Stem Cell., 6:167–174.

15.

Huang

, Chen

, Liu

et al. 2009. More synergetic cooperation of Yamanaka factors in induced pluripotent stem cells than in embryonic stem cells. Cell Res., 19:1127–1138.

16.

Huangfu

, Maehr

, Guo

et al. 2008. Induction of pluripotent stem cells by defined factors is greatly improved by small-molecule compounds. Nat. Biotechnol., 26:795–797.

17.

Ichida

J.K.

, Blanchard

, Lam

et al. 2009. A small-molecule inhibitor of tgf-Beta signaling replaces sox2 in reprogramming by inducing nanog. Cell Stem Cell, 5:491–503.

18.

Jenkins

T.D.

, Opitz

O.G.

, Okano

et al. 1998. Transactivation of the human keratin 4 and Epstein-Barr virus ED-L2 promoters by gut-enriched Kruppel-like factor. J. Biol. Chem., 273:10747–10754.

19.

Jiang

, Chan

Y.S.

, Loh

Y.H.

et al. 2008. A core Klf circuitry regulates self-renewal of embryonic stem cells. Nat. Cell Biol., 10:353–360.

20.

Joliot

, Prochiantz

2004. Transduction peptides: from technology to physiology. Nat. Cell Biol., 6:189–196.

21.

Kim

, Kim

C.H.

, Moon

J.I.

et al. 2009. Generation of human induced pluripotent stem cells by direct delivery of reprogramming proteins. Cell Stem Cell., 4:472–476.

22.

Kwon

H.Y.

, Eum

W.S.

, Jang

H.W.

et al. 2000. Transduction of Cu,Zn-superoxide dismutase mediated by an HIV-1 Tat protein basic domain into mammalian cells. FEBS Lett., 485:163–167.

23.

, McClintick

, Zhong

et al. 2005. Murine embryonic stem cell differentiation is promoted by SOCS-3 and inhibited by the zinc finger transcription factor Klf4. Blood, 105:635–637.

24.

Liu

, Huang

, Chen

et al. 2008. Yamanaka factors critically regulate the developmental signaling network in mouse embryonic stem cells. Cell Res., 18:1177–1189.

25.

Lyssiotis

C.A.

, Foreman

R.K.

, Staerk

et al. 2009. Reprogramming of murine fibroblasts to induced pluripotent stem cells with chemical complementation of Klf4. Proc. Natl. Acad. Sci. USA, 106:8912–8917.

26.

Morita

, Kojima

, Kitamura

2000. Plat-E: an efficient and stable system for transient packaging of retroviruses. Gene Ther., 7:1063–1066.

27.

Nagahara

, Vocero-Akbani

A.M.

, Snyder

E.L.

et al. 1998. Transduction of full-length TAT fusion proteins into mammalian cells: TAT-p27Kip1 induces cell migration. Nat. Med., 4:1449–1452.

28.

Okita

, Ichisaka

, Yamanaka

2007. Generation of germline-competent induced pluripotent stem cells. Nature, 448:313–317.

29.

Pan

, Lu

, Chen

et al. 2009. Reprogramming human fibroblasts using HIV-1 TAT recombinant proteins OCT4, SOX2, KLF4 and c-MYC. Mol. Biol. Rep., 37:2117–2124.

30.

Park

I.H.

, Zhao

, West

J.A.

et al. 2007. Reprogramming of human somatic cells to pluripotency with defined factors. Nature, 451:141–146.

31.

Patel

, Xi

Z.F.

, Seo

E.Y.

et al. 2006. Klf4 and corticosteroids activate an overlapping set of transcriptional targets to accelerate in utero epidermal barrier acquisition. Proc. Natl. Acad. Sci. USA, 103:18668–18673.

32.

Rowland

B.D.

, Bernards

, Peeper

D.S.

2005. The KLF4 tumour suppressor is a transcriptional repressor of p53 that acts as a context-dependent oncogene. Nat. Cell Biol., 7:1074–1082.

33.

Schwarze

S.R.

, Hruska

K.A.

, Dowdy

S.F.

2000. Protein transduction: unrestricted delivery into all cells? Trends Cell Biol., 10:290–295.

34.

Shi

, Desponts

, Do

J.T.

et al. 2008. Induction of pluripotent stem cells from mouse embryonic fibroblasts by Oct4 and Klf4 with small-molecule compounds. Cell Stem Cell, 3:568–574.

35.

Shields

J.M.

, Yang

V.W.

1997. Two potent nuclear localization signals in the gut-enriched Kruppel-like factor define a subfamily of closely related Kruppel proteins. J. Biol. Chem., 272:18504–18507.

36.

Shields

J.M.

, Christy

R.J.

, Yang

V.W.

1996. Identification and characterization of a gene encoding a gut-enriched Kruppel-like factor expressed during growth arrest. J. Biol. Chem., 271:20009–20017.

37.

Sridharan

, Tchieu

, Mason

M.J.

et al. 2009. Role of the murine reprogramming factors in the induction of pluripotency. Cell, 136:364–377.

38.

Stadtfeld

, Maherali

, Breault

D.T.

et al. 2008. Defining molecular cornerstones during fibroblast to iPS cell reprogramming in mouse. Cell Stem Cell, 2:230–240.

39.

Takahashi

, Yamanaka

2006. Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors. Cell, 126:663–676.

40.

Takahashi

, Okita

, Nakagawa

et al. 2007a. Induction of pluripotent stem cells from fibroblast cultures. Nat. Protoc., 2:3081–3089.

41.

Takahashi

, Tanabe

, Ohnuki

et al. 2007b. Induction of pluripotent stem cells from adult human fibroblasts by defined factors. Cell, 131:861–872.

42.

Tunnemann

, Martin

R.M.

, Haupt

et al. 2006. Cargo-dependent mode of uptake and bioavailability of TAT-containing proteins and peptides in living cells. FASEB J., 20:1775–1784.

43.

Wadia

J.S.

, Stan

R.V.

, Dowdy

S.F.

2004. Transducible TAT-HA fusogenic peptide enhances escape of TAT-fusion proteins after lipid raft macropinocytosis. Nat. Med., 10:310–315.

44.

Wei

, Yang

, Zhang

et al. 2009. Klf4 interacts directly with Oct4 and Sox2 to promote reprogramming. Stem Cells, 27:2969–2978.

45.

Wernig

, Meissner

, Foreman

et al. 2007. In vitro reprogramming of fibroblasts into a pluripotent ES-cell-like state. Nature, 448:318–324.

46.

Yang

W.C.

, Patel

K.G.

, Lee

et al. 2009. Cell-free production of transducible transcription factors for nuclear reprogramming. Biotechnol. Bioeng., 104:1047–1058.

47.

, Vodyanik

M.A.

, Smuga-Otto

et al. 2007. Induced pluripotent stem cell lines derived from human somatic cells. Science, 318:1917–1920.

48.

Zhang

, Andrianakos

, Yang

et al. 2010. Kruppel-like factor 4 (Klf4) prevents embryonic stem (ES) cell differentiation by regulating Nanog gene expression. J. Biol. Chem., 285:9180–9189.

49.

Zhou

, Wu

, Joo

J.Y.

et al. 2009. Generation of induced pluripotent stem cells using recombinant proteins. Cell Stem Cell, 4:381–384.