High-Efficiency RNA Interference in Human Embryonic Stem Cells

Authors

  • Holm Zaehres,

    1. Harvard Stem Cell Institute and Harvard Medical School, Department of Biological Chemistry and Molecular Pharmacology and Division of Hematology/Oncology, Children's Hospital, Boston, Massachusetts, USA
    Search for more papers by this author
  • M. William Lensch,

    1. Harvard Stem Cell Institute and Harvard Medical School, Department of Biological Chemistry and Molecular Pharmacology and Division of Hematology/Oncology, Children's Hospital, Boston, Massachusetts, USA
    Search for more papers by this author
  • Laurence Daheron,

    1. Harvard Stem Cell Institute and Harvard Medical School, Department of Biological Chemistry and Molecular Pharmacology and Division of Hematology/Oncology, Children's Hospital, Boston, Massachusetts, USA
    Search for more papers by this author
  • Sheila A. Stewart,

    1. Washington University School of Medicine, St. Louis, Missouri, USA
    Search for more papers by this author
  • Joseph Itskovitz-Eldor,

    1. Technion, Rambam Medical Center, Haifa, Israel
    Search for more papers by this author
  • George Q. Daley M.D., Ph.D.

    Corresponding author
    1. Harvard Stem Cell Institute and Harvard Medical School, Department of Biological Chemistry and Molecular Pharmacology and Division of Hematology/Oncology, Children's Hospital, Boston, Massachusetts, USA
    • Harvard Stem Cell Institute and Harvard Medical School, Department of Biological Chemistry and Molecular Pharmacology and Division of Hematology/Oncology, Children's Hospital, Boston, Massachusetts 02115, USA. Telephone: 617-919-2013; Fax: 617-730-0222
    Search for more papers by this author

Abstract

RNA interference methodology suppresses gene expression, thus mimicking loss-of-function mutation and enabling in vitro and in vivo gene function analysis. In this study, we used retroviral and lentiviral vectors to deliver small interfering RNAs and report high-efficiency silencing of a green fluorescent protein (GFP) trans gene and the stem cell–specific transcription factors Oct4/POU5F1 and Nanog in human embryonic stem cells. Gene knockdown of Oct4 and Nanog promotes differentiation, thereby demonstrating a role for these factors in human embryonic stem cell self-renewal.

Introduction

Human embryonic stem (hES) cells are able to differentiate along all embryonic and adult cell lineages, making them valuable tools in the study of early developmental processes and a promising source of cells for gene and tissue replacement therapies. Vectors for efficient transgene expression in hES cells and their progeny are necessary to evaluate hES cell self-renewal and directed differentiation into specified lineages. Lentiviral vectors are emerging as an efficient tool for transgenesis [1, 2] because they escape much of the transgene silencing that is observed with oncoretrovirus-based vectors in hematopoietic and ES cells [3].

Recently, the use of lentiviral vectors has been described for transgene expression in hES cells [4, 5]. In this study, we extend the use of lentiviral gene transfer to achieve high-efficiency loss of function in hES cells. Classical strategies to knockout genes in mouse ES cells rely on gene targeting by homologous recombination. This technique was successfully applied to hES cells [6]. Stable expression of small interfering RNAs (siRNAs) provides a faster and more convenient method to knockdown gene expression, but high-efficiency RNA interference is required to approximate gene knockout. siRNA duplexes have been successfully used to suppress multiple genes in mammalian cells [7, 8], but to date RNA Interference (RNAi) has not been exploited with high efficiency in hES cells. Retroviral and lentiviral vectors have been described that stably express stem-loop cassettes under the control of RNA polymerase III U6 promoters [9, 10] that give rise to inhibitory RNAs after intracellular processing. Using these vectors, we have produced concentrated viral stocks that enable highly efficient transduction of siRNA in hES cells to effect gene silencing. We demonstrate high-efficiency silencing of a green fluorescent protein (GFP) transgene and the stem cell–specific transcription factors Oct4/POU5F1 and Nanog in hES cells.

Material and Methods

hES Cell Culture

The hES cell line WA09/H9 [11] (passage ∼60, karyotype analysis showed no evidence of chromosomal abnormalities) was cultured on feeder layers (primary mouse embryo fibro-blasts [MEFs] from Specialty Media [Phillipsburg, NJ]) in knockout Dulbecco's modified Eagle's medium (Gibco, Grand Island, NY) supplemented with 20% serum replacement (Knockout SR [Gibco]), 1 mM L-glutamine, 1% non-essential amino acids, 0.1 mM beta-mercaptoethanol, and 4 ng/ml human basic fibroblast growth factor (bFGF) (Peprotech, Rocky Hill, NJ) as described [12]. For the differentiation assays, H9 cells were grown in MEF-conditioned medium supplemented with bFGF (4 ng/ml) on Matrigel basement membrane matrix (BD, Franklin Lakes, NJ) [13].

siRNA Vector Design

An oligonucleotide encoding a stem-loop structure targeting the GFP with the targeting sequence GGC.TAC.GTC.CAG.GAG.CGC.ACC was cloned under the control of the human U6 promoter in the vectors Retrohair (RH) and Lentihair (LH) [9]. Two oligonucleotides encoding stem-loop structures targeting the human Oct4/POU5F1 gene with the targeting sequences AAC.ATG.TGT.AAG.CTG.CGG.CCC and AAG.GAT.GTG.GTC.CGA.GTG.TGG and one oligonucleotide targeting the human Nanog gene with the targeting sequence AAG.GGT.TAA.GCT.GTA.ACA.TAC were designed using the siRNA Selection Program at the Whitehead Institute for Biomedical Research (http://jura.wi.mit.edu/siRNAext/) [14]. These were cloned under the control of the U6 promoter in the vector Lentilox 37 [10] and verified by DNA sequencing.

Retroviral and Lentiviral Vector Production and hES Cell Transduction

Retroviral and lentiviral vectors, pseudotyped with the vesicular stomatitis virus (VSV) G protein, were produced in 293T cells as described [9, 10, 15]. Viral supernatants were concentrated by ultracentrifugation to produce virus stocks with titers of 1 × 108 to 1 × 109 infectious units per milliliter. Titers were determined on 293 T cells. A total of 1 × 105 H9 hES cells were transduced on MEFs or in conditioned media on Matrigel by single round infections for 24 hours. Multiplicities of infection (MOI) of 10, 50, and 100 were used to transduce H9-GFP or H9 with the siRNA expression vectors.

Fluorescence Microscopy, Immunostaining, and Flow Cytometry

H9-GFP colonies and H9 cultures were analyzed by fluorescence microscopy and flow cytometry (Becton Dickinson FACScalibur with Cellquest software).

H9 cells were stained with the primary antibody TRA-1-60 [16]. Phycoerythrin (PE)-conjugated goat anti-mouse immunoglobulin (IgM) (BD) was used as the secondary antibody.

Immunoblotting

Protein extracts (40μg) were prepared from hES cells and run on 15% SDS-PAGE as described [12]. The antibodies used were Oct3/4 (Santa Cruz Biotechnology, Santa Cruz, CA), hActin (Oncogene Research Products, Boston), and horse-radish peroxidase–linked anti-mouse IgM (Amersham Biosciences, Piscataway, NJ). Peroxidase activity was detected with the ECL Detection Kit (Amersham Biosciences).

Quantitative Real-Time Reverse Transcription Polymerase Chain Reaction Analysis

Quantitative real-time reverse transcription–polymerase chain reaction (QRT-PCR) analysis was performed in triplicate for each primer set and in each cell line. Total cellular RNA was isolated using RNA STAT-60 (Tel-Test, Friendswood, TX) according to the manufacturer's instructions. cDNA synthesis was performed with 3 μg total RNA and Superscript II reverse transcriptase (Invitrogen, Carlsbad, CA). QRT-PCR was then performed using an ABI Prism 7700 Sequence Detection System (ABI, Foster City, CA). Amplification was performed using the TaqMan PCR system (ABI) per the manufacturer's instructions. Specific primers and probes were designed and obtained from Integrated DNA Technologies (Coralville, IA). The overall number of threshold amplification cycles per primer set was first calibrated against a control of actin cDNA. The cycle number for the RNAi vector–transduced cells was then standardized against cells transduced with the control vector. For each evaluated marker, the amplification threshold obtained with the H9 vector control cell line was defined as a fold value of 1 (100%). The primers used are listed in Table 1.

Table Table 1.. Primers used in quantitative real-time reverse transcription–polymerase chain reaction studies
PrimerNCBI sequenceForward reverse dual-labeled probe
hActinNM_001101CGG.GAA.ATC.GTG.CGT.GAC.ATT.AAG
  TAC.TCC.TGC.TTG.CTG.ATC.CAC.ATC
  FAM - TGA.AGT.GTG.ACG.TGG.ACA.TCC.GCA.AA - TAM
hOct4NM_203289GGA.AGG.TAT.TCA.GCC.AAA.CGA.CCA
  CTC.ACT.CGG.TTC.TCG.ATA.CTG.GTT
  FAM -AAG.AAC.ATG.TGT.AAG.CTG.CGG.CCC.TT - TAM
hNanogNM_024865ACC.AGA.ACT.GTG.TTC.TCT.TCC.ACC
  CCA.TTG.CTA.TTC.TTC.GGC.CAG.TTG
  FAM - AAT.ACC.TCA.GCC.TCC.AGC.AGA.TGC.AA - TAM
hCGaNM_000735GTT.TCT.GCA.TGT.TCT.CCA.TTC.CGC
  GTG.GAC.TCT.GAG.GTG.ACG.TTC.TTT
  FAM - TGA.TGT.GCA.GGA.TTG.CCC.AGA.ATG.CA - TAM
hCGbNM_000737ATC.ACC.GTC.AAC.ACC.ACC.ATC.TGT.G
  AGA.GTG.CAC.ATT.GAC.AGC.TGA.G
  FAM - TCA.GGT.GGT.GTG.CAA.CTA.CCG.CGA.T - TAM
hAlbuminM12523AAC.CAC.TCT.AGA.GAA.GTG.CTG.TGC
  CTG.ACT.CTG.TCA.CTT.ACT.GGC.GTT
  FAM - TCC.GTG.GTC.CTG.AAC.CAG.TTA.TGT.GT - TAM
hCardiac ActinNM_005159GCA.AGG.ACC.TGT.ATG.CCA.ACA.ATG
  GCC.TCA.TCG.TAC.TCT.TGC.TTG.CTA
  FAM - TGC.AGA.AGG.AAA.TCA.CTG.CTC.TGG.CT - TAM
hNestinNM_006617TCT.CCA.GAA.ACT.CAA.GCA.CCA.CTG
  TCC.TGA.TTC.TCC.TCT.TCC.AGG.AGT\
  FAM - ACC.ACT.GGA.GTC.TGT.GGA.AGT.GAA.CCA.A - TAM
hNFHCNM_021076TGG.AAG.GTG.AAG.AGT.GTC.GGA.TTG
  GGA.CTT.GAC.CTC.AGC.TGG.AGA.TTT
  FAM - TTG.GCC.CAA.TTC.CTT.TCT.CGC.TTC.CA - TAM

Results and Discussion

We first modified a hES cell line to stably express an enhanced GFP (eGFP) transgene. We transfected 293T cells with a lentiviral vector expressing eGFP under the control of the human phosphoglycerate kinase promoter [17] and a plasmid encoding the VSV-G protein [15] to produce a pseudotyped viral supernatant. We then used this virus at a MOI of 50 to infect the human H9 hES cell line to produce a population that expressed GFP in 50% of cells (49.8% ± 4.4% [mean ± standard error]; n = 5). We next transduced the H9-GFP cells with increasing MOIs of the VSV-G pseudotyped vectors Retro-hair (RH) and Lentihair (LH) encoding stem-loop structures targeting the GFP gene (Fig. 1A). Infection of H9-GFP with the vectors RH-GFPi and LH-GFPi reduces GFP expression up to fivefold after 2 days and up to 140-fold after 7 days (measured by fluorescence-activated cell sorter histogram statistics) (Figs. 1B, 1C). Infected colonies maintained viability, and the reduction in GFP expression could be appreciated by fluorescence and phase-contrast microscopy, respectively (Fig. 1D). These data demonstrate that highly efficient gene silencing can be achieved when retroviral and lentiviral vectors are used to deliver interfering RNAs in hES cells.

Figure Figure 1..

Retroviral delivery of RNA interference mediates gene silencing in human embryonic stem cells. Small interfering RNA molecules against the GFP transcript were introduced as stem-loop structures by retroviral and lentiviral vectors into a human embryonic stem cell line (WA09/H9) in which the GFP gene had been stably integrated. (A): Retroviral vector (RH)-GFPi and lentiviral vector (LH)-GFPi used to deliver the interfering RNA molecules to silence GFP expression. (B): Fluorescence-activated cell sorter histograms of H9-GFP cells 48 hours after transduction with increasing doses of the retroviruses RH-GFPi and LH-GFPi. The upper histogram represents the H9-GFP–expressing cell line before infection with the siRNA retroviruses. (C): Histogram statistics reveal reduction in GFP expression up to fivefold after 2 days and up to 140-fold after 7 days. Data are means ± standard error for two experiments. (D): Phase contrast (upper) and fluorescence (lower) microscopy of individual H9-GFP colonies before (left) and after (right) transduction (day 4) with the retrovirus LH-GFPi. Abbreviations: GFP, green fluorescent protein; LH, lentihair; RH, retrohair.

To demonstrate a biological effect of gene knockdown in hES cells, we targeted two genes implicated in maintenance of pluripotency in mouse ES cells: Oct4 [18], a developmentally regulated gene belonging to the POU class of helix-turn-helix transcription factors, and Nanog [19, 20], a homeodomain transcription factor. We developed lentiviral vectors with two oligonucleotides encoding stem-loop structures targeting the human Oct4/POU5F1 gene and one oligonucleotide targeting the human Nanog gene (Fig. 2A). We then transduced H9 hES cells with VSV-G pseudotyped and concentrated viruses by a single round of infection for 24 hours. The control Lentilox vector expressing eGFP from the cytomegalovirus promoter gave robust transduction efficiencies of 85% at a MOI of 50 (84.7% ± 3.1% [mean ± standard error]; n = 10) GFP-positive H9 cells. Infected cell cultures were maintained in MEF-conditioned media [13], conditions that would otherwise maintain the self-renewal of hES cells. After RNAi application, Western blot analysis showed complete extinction of the Oct4 protein (Fig. 2B). We interrogated the levels of expression of the TRA-1-60 antigen associated with undifferentiated ES cells [16] at different time points after RNAi application. TRA-1-60 levels fell, achieving a nadir by day 7, when TRA-1-60 was markedly reduced from 80% in the control population to 28% in the population infected with the Oct4-RNAi knockdown vectors. TRA-1-60 reduction was even greater (to 3%) in cells infected with the Nanog-RNAi knockdown vector (Fig. 2C). The differentiated cell populations transduced by the RNAi vector (GFP positive and TRA-1-60 negative) represent most cells at this time point. They are no longer refractile and assume a flattened morphology, typical of differentiating cultures of hES cells. After 10 days, however, nontransduced, undifferentiated hES cells overgrow the slower growing differentiated cells, suggesting that a selection strategy would be required to obtain pure, stable populations of differentiated cells after RNAi treatment. TRA-1-60 antigen expression remained stable and high when H9 hES cells were infected with the GFP knockdown vector LH-GFPi, demonstrating that siRNA expression per se does not promote hES cell differentiation.

Figure Figure 2..

Small interfering RNA molecules against the human Oct4 and Nanog transcripts were introduced as stem-loop structures by lentiviral vectors in the human embryonic stem cell line WA09/H9. (A): Lentiviral vectors LL-OCT4i and LL-NANOGi used to deliver the interfering RNAs targeting Oct4 and Nanog. (B): Western blot analysis of the Oct4 protein in cells after RNAi application at a MOI of 10, 50, and 100 at day 6. The blot was stripped and reprobed with actin as a loading control. (C): FACS histograms of TRA1-60–stained H9 at day 7 after transduction with the Lentilox37 control vector (right), the two lentiviral Oct4 RNAi vectors (middle), and the lentiviral Nanog RNAi vector (left) at a MOI of 50. Cells were stained with the primary antibody TRA-1-60 and detected with PE secondary antibody in the FL2 chanel (y-axis). The GFP positivity in the FL1 chanel (x-axis) marks the vector-transduced cell populations. The percentages indicate the cell populations in each of the four quadrants. (D): Quantitative real-time reverse transcription polymerase chain reaction analysis of human Oct4 and Nanog at days 4 and 7 (upper panel) and the trophectodermal lineage markers chorionic gonadotropin alpha and beta, the endodermal marker albumin, the mesodermal marker cardiac actin, and the ectodermal markers nestin and neuro-filament heavy chain at day 7 after transduction (lower panel). Data are means ± standard error for three experiments, each carried out in triplicate (upper panel), and a single representative experiment carried out in triplicate (lower panel). Abbreviations: MOI, multiplicity of infection; PE, phycoerythrin.

As assessed by QRT-PCR, application of the Oct4 knockdown vectors reduced Oct4 expression to 31% on day 4 and 15% on day 7, whereas the Nanog knockdown reduced Nanog expression to 14% on day 4 and 13% on day 7 (Fig. 2D). We also observed that Oct4-RNAi reduced Nanog expression and Nanog-RNAi reduced Oct4 expression. This reciprocal effect occurred already at day 4, when the culture shift into differentiation was just at its beginning, suggesting coordinate gene regulation of Oct4 and Nanog in hES cells. Decreased Oct4 expression has been shown to redirect mouse ES cells into the trophectodermal lineage [21], whereas Nanog expression seems to prevent differentiation into primitive endoderm [20]. Although undifferentiated H9 cells have been reported to express low levels of some trophoblastic markers [22], QRT-PCR analysis of Oct4-RNAi–treated H9 showed twofold to fourfold upregulation of the trophoblastic markers human chorionic gonadotropin alpha (hCGα) and beta (Fig. 2D). In the case of the Nanog knockdown, the endodermal marker albumin was upregulated 13-fold, confirming a role for Nanog in antagonizing endodermal differentiation. Interestingly, the most striking effect was 47-fold upregulation of hCGα, suggesting a role for Nanog in modulating trophectodermal differentiation as well. The expression of the mesodermal marker cardiac actin was modestly decreased, whereas we did not detect significant changes in the ectodermal markers nestin and neurofilament heavy chain for either of the knockdowns.

Our data indicate that Oct4 and Nanog are central to hES cell maintenance and play functionally conserved roles in hES and mouse ES cells [19, 20]. The profound consequence of the Nanog knockdown on loss of self-renewal supports a central role for Nanog in the transcription factor hierarchy that maintains the pluripotency of hES cells. Furthermore, our data demonstrate that downregulation of Oct4 and Nanog promotes hES cell differentiation under conditions that would otherwise foster self-renewal.

We have demonstrated that RNA interference is effective at inducing gene silencing in hES cells, both for the GFP transgene as well as the endogenous genes Oct4 and Nanog. Recent reports used transfection of chemically synthesized siRNAs against a GFP transgene [23] and Oct4 [24, 25] in hES cells. Comparison of these data with our results concerning GFP and Oct4 silencing, TRA-1-60 downregulation, as well as upregulation of trophectodermal and endodermal markers highlights the greater efficiency of viral delivery of siRNAs, whose effects at higher MOIs approximate gene knockout. A report targeting mouse ES cell pluripotency genes of the Src kinase families describes difficulties isolating stable siRNA-expressing clones [26]. Given the high titers that can be generated and corresponding high infection rates in cell populations, lentiviral vector–delivered RNAi allows evaluation of silencing of genes involved in self-renewal, differentiation, or apoptosis without having to select stable ES cell clones. When combined with large-scale libraries of interfering RNAs [27, 28] or the recently available lentiviral conditional RNA interference systems [29, 30], our rapid and convenient approach should be applicable to genome-wide screening experiments and conditional loss-of-function analysis in hES cells and their progeny.

Acknowledgements

We thank Michael McManus, Christopher Dillon, and Luk van Parijs for providing the LentiLox vector system; Peter Andrews for providing the Tra-1-60 antibody; Patricia Ernst for help with quantitative real-time PCR analysis; Asmin Tulpule for critical discussion; and James Cha for help with vector design. This work was supported by grants from the Bekenstein Family; the National Science Foundation MIT Biotechnology Process Engineering Center; and the National Institutes of Health (DK59279 and HL71265; the NIH Director's Pioneer Award of the NIH Roadmap for Medical Research). G.Q.D. is a recipient of the Burroughs Wellcome Fund Clinical Scientist Award in Translational Research, and the Birnbaum Scholar of the Leukemia and Lymphoma Society of America. M.W.L. is a fellow of the Leukemia and Lymphoma Society of America.

Ancillary