Genetic approach to track neural cell fate decisions using human embryonic stem cells
- First online:
With their capability to undergo unlimited self-renewal and to differentiate into all cell types in the body, human embryonic stem cells (hESCs) hold great promise in human cell therapy. However, there are limited tools for easily identifying and isolating live hESC-derived cells. To track hESC-derived neural progenitor cells (NPCs), we applied homologous recombination to knock-in the mCherry gene into the Nestin locus of hESCs. This facilitated the genetic labeling of Nestin positive neural progenitor cells with mCherry. Our reporter system enables the visualization of neural induction from hESCs both in vitro (embryoid bodies) and in vivo (teratomas). This system also permits the identification of different neural subpopulations based on the intensity of our fluorescent reporter. In this context, a high level of mCherry expression showed enrichment for neural progenitors, while lower mCherry corresponded with more committed neural states. Combination of mCherry high expression with cell surface antigen staining enabled further enrichment of hESC-derived NPCs. These mCherry+ NPCs could be expanded in culture and their differentiation resulted in a down-regulation of mCherry consistent with the loss of Nestin expression. Therefore, we have developed a fluorescent reporter system that can be used to trace neural differentiation events of hESCs.
KeywordsNestin knock-in human embryonic stem cells neural progenitor cells
Human embryonic stem cells (hESCs) undergo unlimited self-renewal and retain pluripotency to differentiate into all cell types of the body, making them invaluable tools for development of cellular therapeutics (Fu and Xu, 2011). Modeling human disease has further become possible with advances in genetic modification in hESCs either through homologous recombination (Song et al., 2010; Hockemeyer et al., 2011) or gene knockdown (Ordonez et al., 2012), as well as reprogramming of cells having disease-causing mutations into an ESC-like state (human induced pluripotent stem cells or hiPSCs) (Unternaehrer and Daley, 2011). Given the limitation of using mouse systems to develop human therapeutics, the use of these approaches can provide more physiologically relevant insight into disease pathology and help to identify more efficacious drug targets. These approaches additionally permit disease modeling of tissues that are previously unattainable or that show pathologies too advanced in donor tissues. This includes studies on the mechanisms underlying neuronal degeneration associated with central nervous system (CNS) diseases (e.g. Parkinson’s, Alzheimer’s or Amyotrophic lateral sclerosis).
Studies using human pluripotent stem cell derived neurons to model these diseases or generate neural tissue replacements rely upon the effectiveness of protocols that recapitulate normal embryonic development to direct the differentiation of hESCs into the neural cell types afflicted by the disease (Aas et al., 1996). A key step in this process is to differentiate hESCs into neural stem cells (NSCs) that can be further differentiated into neurons, astrocytes and glia. While significant progress has been made in improving the differentiation of hESCs into NSCs (Reubinoff et al., 2001; Itsykson et al., 2005; Wu et al., 2007; Chambers et al., 2009), this step remains an imperfect process with resulting cultures comprising a mixture of undifferentiated ESCs, NSCs, neurons, glia and non-CNS cell types including neural crest. Subsequent differentiation of these mixed cultures would fail to generate single pure lineages needed for transplantation, modeling or population based gene expression studies (e.g. RNA-seq).
To circumvent these issues, several groups have had success in using combinatorial cell surface marker staining to enrich for NSCs (Yuan et al., 2011; Israel et al., 2012). However this method requires further sorting strategies for continued monitoring of resultant cell differentiation states. Promoter-based viral reporter systems for the neural progenitor marker Nestin has also been used to specifically label NSCs within the human fetal brain and could be used to purify these cells for expansion or for further differentiation both in culture and following transplantation (Keyoung et al., 2001). Several groups have used Nestin reporter systems to identify and track neural progenitors both in vivo and in vitro (Sawamoto et al., 2001; Lenka et al., 2002; Mignone et al., 2004; Noisa et al., 2010). These transgenic or viral based strategies, however, involve integration of the exogenous DNA into random genomic loci that can modulate the promoter activity, depending on the epigenetic state around the integration site. To address this issue, we decided to generate knock-in cell lines using a Nestin-based fluorescent reporter system that would permit not only a simpler sorting strategy for isolating NPCs/NSCs but also would enable their live tracking during expansion or differentiation. Unlike transgenic approaches (Placantonakis et al., 2009), this system would ensure reporter expression specifically driven by the endogenous Nestin promoter and/or enhancer, and would not introduce any other genetic alterations that arise from random integration strategies. This would permit specific mCherry expression during neural differentiation of hESCs and the ability to generate enriched NSC cultures where multi-potentiality could be easily monitored. These cells will provide a unique opportunity for studies requiring purer neural cultures and live cell tracking, including transplantation, disease modeling or whole transcriptome analyses (Peljto and Wichterle, 2011).
Generation of a knock-in Nestin reporter cell line
mCherry fluorescence follows Nestin expression dynamics
Following out-migration from EB or teratoma tissues, mCherry positive cells displayed properties consistent with neural progenitor cells. They could be further expanded in culture for several passages (data not shown); showed significant proliferation as indicated by Ki-67 staining (Fig. 3D); and co-expressed endogenous Nestin (Fig. 3E). Furthermore, high mCherry expression was localized to neural rosette-like structures, as was previously reported for a transgenic Nestin reporter (Noisa et al., 2010), and co-localized with the neural stem cell marker Sox2 (Fig. 3F). Therefore, mCherry expression can faithfully label Nestin positive cycling progenitors during differentiation of hESCs into the neural lineage both in vitro and in vivo.
mCherry levels dropped during NPC differentiation
One significant advantage to the labeling of NPCs with mCherry would be the potential for sorting and enriching this population. Furthermore, given the potential heterogeneity of neural progenitors in terms of stage of commitment and potentiality, it would be advantageous to specifically isolate or enrich bone fide neural stem cells. To determine the specificity of mCherry to NSCs or NPCs compared to more committed cell lineages, we used cell surface markers that highlight these different neural subpopulations. Previous reports have shown the use of multiple surface markers to identify NSCs based on a CD184+CD271-CD44-CD24+ staining signature (Yuan et al., 2011). Both CD271 and CD44 identify undifferentiated hESCs, as well as neural contaminants including neural crest stem cells, peripheral neural cells, more differentiated neuronal cell types, astrocyte progenitors and glia (Morrison et al., 1999; Luo et al., 2002; Liu et al., 2004; Lee et al., 2007). Alternatively, both CD184 and CD24 have been shown to be expressed on the surface of NSCs. Consistent with mCherry high expressers representing a more neural stem cell state, the CD184+CD44-CD24+ population localized more within the mCherryHi expressing cells in contrast to the CD184-CD24- population (Fig. 5). Furthermore, the CD271-CD44-CD24+ cells also mapped more to the mCherryHi expressers unlike a similar CD24- population (Fig. 5). These results indicate that mCherry high expression is in part predicative of a stem cell state, but is not specific to this population (Fig. 5). This is likely the result of continued high Nestin expression in more committed neural progenitors as well as a potential lag in the mCherry expression compared to endogenous Nestin that is a result of their differing protein stabilities.
mCherry sorting strategy for NSC enrichment
Our studies demonstrate a knock-in approach that can be used together with existing strategies of purifying NPCs/NSCs derived from hESCs while additionally providing a means of tracking these cells live, both in vitro and in vivo. Transgenic or viral Nestin reporters have been widely used to label neural progenitors (Keyoung et al., 2001; Sawamoto et al., 2001; Lenka et al., 2002; Mignone et al., 2004; Noisa et al., 2010) and Nestin positivity remains an important marker identifying hESC-derived NSCs (Reubinoff et al., 2001; Zhang et al., 2001; Itsykson et al., 2005; Joannides et al., 2007; Chambers et al., 2009). Our transgene-free Nestin:mCherry reporter hESC line provides the unique ability to monitor the differentiation status of our cultures with the sensitivity to distinguish high and low sub-populations of Nestin expression and without requiring Nestin immunostaining. While mCherryLo cells likely correspond to more committed neural lineages or non-neural contaminants, the mCherryHi population appears to enrich for neural progenitors with multi-lineage capabilities. Furthermore, the combination of mCherryHi with CD184, CD44 and CD271 surface marker staining permitted significant enrichment of NSC-specific markers (Sox1, Sox2 and Pax6). These results indicate that our reporter system can be combined with additional strategies for specifically isolating or enriching NSCs including and not limited to: surface marker staining; additional knock-in strategies conferring NSC or lineage-specific drug resistance; or knock-in fluorescent reporters of additional neural lineage markers. Furthermore, our knock-in line was able to achieve significant enrichment of NSCs with a smaller subset of surface markers (CD184+CD44-CD271-) as has been previously used (Yuan et al., 2011), providing a clear advantage over wild-type hESCs. The potential to replace the CD24 surface marker with intrinsic fluorescent reporter activity further provides a means to monitor the NSC cultures and track terminal differentiation. Our lineage-specific knock-in approach will enable further studies tracing the commitment of neural progenitors derived from hESCs; to perform population based studies; and to improve current protocols for directed differentiation of cell types needed for cell replacement therapy.
Materials and methods
Construction of BAC-based targeting vector
The Nestin BAC clone was purchased from Life Technologies Inc. and the targeting vector was constructed by recombineering as we previously described (Song et al., 2010).
Cell culture of hESCs
The HUES9 hESCs were cultured as described (Cowan et al., 2004). Briefly, hESCs were cultured on mouse embryonic fibroblast (MEFs) feeder layer in knockout medium supplemented with 10% knockout serum replacement, 10% plasmanate, 1 mmol/L glutamine, 0.1 mmol/L nonessential amino acids, 10 ng/mL bFGF, and 0.1 μmol/L β-mercaptoethanol. To passage hESCs, confluent culture was washed with phosphate buffered saline (PBS), dissociated for 5 min with TrypLE and resuspended into single cells with culture medium. All tissue culture reagents were purchased from Life Technologies Inc. except where indicated otherwise.
Differentiation of hESCs into neural progenitors
The neural differentiation of hESCs was performed with established protocols (Reubinoff et al., 2001; Zhang et al., 2001). Briefly, hESC colonies were gently scraped from the plate and cultured in suspension as embryoid bodies (EB) on ultra-low cluster plates in EB formation medium (DMEM/F12-Glutamax, 10% characterized FBS, 1% Pen-Strep, 100 μmol/L β-mercaptoethanol). The media was changed to NPC induction medium (NPC medium: DMEM/F12-Glutamax, 0.5% N2, 1% B27, 1% Penicillin-Streptomycin, 10 ng/mL bFGF). After two days, the EBs were replated onto poly-L-ornithine/laminin coated plates. After neural cells migrated away from the EBs and proliferated to confluence, the culture was treated with 0.05% trypsin for 1–2 min at room temperature, resuspended into single cells, and either sorted or re-plated onto poly-L-ornithine/laminin coated dishes in NPC medium. In one sort experiment, BMP inhibitor Noggin (500 ng/mL, R&D Systems) and TGF-beta inhibitor SB431542 (10 μmol/L, Sigma-Aldrich) were included during EB differentiation stages prior to their attachment onto poly-L-ornithine/laminin coated dishes.
Teratoma formation of hESCs in SCID mice
Two million hESCs were suspended in the knockout medium supplemented with 30% matrigel (BD Biosciences) and injected subcutaneously into severe combined immunodeficiency (SCID) mice. After six to eight weeks, teratomas were recovered. For histological assessment, teratomas were fixed in 10% buffered formalin, embedded in paraffin, sectioned and stained with hematoxylin and eosin. To derive NPCs from teratomas, teratomas were chopped into small pieces and cultured in NPC medium. After several passages, the adherent culture became homogenous with mCherry positive cells.
Differentiation of hESC-derived NPCs into mature neurons
The hESC-derived neural progenitors were transduced with lentivirus harboring synapsin promoter-driven EGFP as described (Kim et al., 2011; Lake et al., 2012). The NPCs were cultured in the neuronal differentiation medium containing DMEM/F12 supplemented with N2/B27, BDNF (10 ng/mL, Peprotech), GDNF (10 ng/mL, Peprotech), Penicillin-Streptomycin (Invitrogen), glutamax (Invitrogen), cyclic AMP (1 μmol/L, Sigma Aldrich). The cells were cultured in differentiation media for two to three weeks with 50% of the medium replaced every other day.
Quantitative real-time PCR analysis
Total RNA from hESCs and hESC-derived neural cultures were isolated using RNAeasy Mini Kit (QIAGEN) or ZR RNA MicroPrep Kit (Zymo Research). Total RNA was reverse transcribed using the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems) and analyzed on a StepOnePlus™ quantitative real-time PCR system (Applied Biosystems) as previously described (Song et al., 2007). The primers used are as following: Sox1 (F-GAGTGGAAGGTCATGTCCGAGG; R-CCTTCTTGAGCAGCGTCTTGGT), Sox2 (F-ATGCACCGCTACGACGTGA; R-CTTTTGCACCCCTCCCATTT), Pax6 (F-TGTCCAACGGATGTGTGAGT; R-TTTCCCAAGCAAAGATGGAC), GAPDH (F-GATGACATCAAGAAGGTGGTGA; R-GTCTACATGGCAACTGTGAGGA). Data analysis was performed using Microsoft Excel.
Neural progenitor cells grown on poly-L-ornithine/laminin coated 4-well chamber plates were fixed in 4% paraformaldehyde for 15 min at room temperature, permeabilized with 0.3% Triton-X-100 in Triton Buffer Solution (TBS) for 10 min, blocked with 2% FBS in TBS-0.05% Tween-20 for 20 min and stained with primary antibodies in blocking buffer at 4°C overnight. Primary antibodies included: anti-Nestin (611658; BD Pharmingen); anti-SOX2 (561469; BD Pharmingen); anti-RFP (600-401-379; Rockland); anti-ki67-Alexa fluor-488 (51-9007231; BD Stemflow Human neural Lineage analysis Kit); TuJ1 (MMS-435P, Covance); GFAP (ab7779, Abcam). The next day, the cells were stained with secondary antibodies (Alexa fluor 568 donkey anti-rabbit; Alexa fluor 488 donkey anti-mouse; Life Technologies, Inc.) in blocking buffer for 45 min and nuclei were counterstained with DAPI. Images were acquired using an Olympus FV1000 confocal microscope. Live culture images were also acquired using an Inverted Axioscope and AxioCam MRm (Carl Zeiss, Inc.). Image assembly was performed using Adobe Photoshop CS5 (Adobe Systems, Inc.).
hESC-derived neural cultures were washed with PBS, trypsinized with 0.05% trypsin for 1–2 min at room temperature and resuspended into single cells using DMEM/F12 with Glutamax containing 10% FBS. For cell surface staining, cells were incubated with CD184-APC (560936; BD Pharmingen), CD24-PE-Cy7 (561646; BD Pharmingen), CD44-PE (51-9007231; BD Pharmingen) and/or CD271-Alexa fluor 647 or BV421 (560877 or 562562; BD Pharmingen) antibodies in PBS containing 1% bovine serum albumen (BSA) according to manufacturer’s protocol. Stained and unstained cells were analyzed using either a BD LSR-II or LSRFortessa flow cytometry machine (BD Biosciences). Control hES9 hESCs or hESC-derived NPCs were used for mCherry control and IgG-PE-Cy7, IgM-FITC, IgG-BV421 and IgG-APC (BD Pharmingen) were used for isotype controls. Cell sorting was performed using either BD FACSAria II or BD Influx cell sorter. Gating for all sorts was defined by isotype control staining. Flow cytometry data analysis was performed using FlowJo software (Tree Star, Inc.).
We thank Drs. Anirvan Ghosh, Ji-Eun Kim, Evan Snyder and Jean-Pyo Lee for help with neural differentiation of human ES cells. This work was supported by a grant from California Institute for Regenerative Medicine (RC1-00148) to YX and a grant from the National Natural Science Foundation of China (Grant No. 81172828) to XF. Histology was carried out by the UCSD Histology and Immunohistochemistry Shared Resource; confocal microscopy was performed within the UCSD Neuroscience Microscopy Shared Facility (Grant: P30 NS047101). FACS sorting was performed within UCSD Human Embryonic Stem Cell Core Facility.
Compliance With Ethics Guidelines
All hESC work in this study has been approved by the INSTITUTIONAL EMBRYONIC STEM CELL RESEARCH OVERSIGHT COMMITTEE (ESCRO) of University of California, San Diego.
All animal work in this study has been approved by Institutional Animal Care and Use Committee (IACUC) of University of California, San Diego. Xuemei Fu, Zhili Rong, Shengyun Zhu, Xiaocheng Wang, Yang Xu, and Blue B. Lake declare that they have no conflict of interest.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution License which permits any use, distribution, and reproduction in any medium, provided the original author(s) and the source are credited.