SEARCH

SEARCH BY CITATION

Keywords:

  • Human embryonic stem cells;
  • Lineage selection;
  • Human neurons;
  • Cryopreservation

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References

A major prerequisite for the biomedical application of human embryonic stem cells (hESC) is the derivation of defined and homogeneous somatic cell types. Here we present a human doublecortin (DCX) promoter-based lineage-selection strategy for the generation of purified hESC-derived immature neurons. After transfection of hESC-derived neural precursors with a DCX-enhanced green fluorescent protein construct, fluorescence-activated cell sorting enables the enrichment of immature human neurons at purities of up to 95%. Selected neurons undergo functional maturation and are able to establish synaptic connections. Considering that the applicability of purified hESC-derived neurons would largely benefit from an efficient cryopreservation technique, we set out to devise defined freezing conditions involving caspase inhibition, which yield post-thaw recovery rates of up to 83%. Combined with our lineage-selection procedure this cryopreservation technique enables the generation of human neurons in a ready-to-use format for a large variety of biomedical applications.

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


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References

Author contributions: J.L.: conception and design, collection and/or assembly of data, data analysis and interpretation, manuscript writing; P.K.: conception and design, provision of study material, data analysis and interpretation, manuscript writing; J.L. and P.K. contributed equally to this work. E.E.: collection and/or assembly of data; B.M.: collection and/or assembly of data; T.O.: collection and/or assembly of data, data analysis and interpretation; S.C.-D.: provision of study material; L.A.: provision of study material; O.B.: conception and design, financial support, administrative support, data analysis and interpretation, manuscript writing, final approval of manuscript.

Key prerequisites for the biomedical application of embryonic stem cell-derived somatic cell types are purity and cell type specification. Furthermore, to avoid laborious differentiation procedures and for having cells on demand, efficient storage protocols that do not impair the differentiation potential of the cells are necessary. Although numerous in vitro differentiation protocols are available to enrich for distinct tissues and cell types, these approaches typically do not permit the selective derivation of a distinct cellular subtype at a defined stage of differentiation.

As a result of their potential for neural regeneration, disease modeling, and compound screening for neurodegenerative disorders, human embryonic stem cell (hESC)-derived neurons represent a highly desired cell population. The preparation of this cell type in a form suitable for immediate application poses particular challenges. Although existing procedures permit the generation of purified and proliferating hESC-derived neural precursors, these cells give rise to a mixture of restricted progenitors, neurons, and glia [1, 2]. Fully mature neurons, on the other hand, are largely resistant to further in vitro processing, cryopreservation, and tissue integration [3, [4]5]. Thus, hESC-derived neurons should ideally be available at an immature yet committed state of differentiation.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References

Generation of hESC-Derived Neural Precursors

hESC (line H9.2 passages 32–61) were maintained on irradiated mouse embryonic fibroblasts (MEFs) at 5% CO2 in medium containing Knockout-DMEM (KO-DMEM, Invitrogen, Carlsbad, CA, http://www.invitrogen.com), 20% serum replacement, 1% nonessential amino acids, 1 mM l-glutamine, 0.1 mM β-mercaptoethanol, and 4 ng/ml fibroblast growth factor 2 (FGF2) (all Invitrogen). Cultures were manually passaged at a 1:3–1:5 split ratio every 4–5 days. Neural differentiation was performed as previously described [2, 6, 7]. Briefly, 4-day-old embryoid bodies were transferred to polyornithine-coated tissue-culture dishes and propagated in ITSFn medium (DMEM/F12; Invitrogen) containing 25 μg/ml insulin, 100 μg/ml transferrin, 5 ng/ml sodium selenite (all Sigma-Aldrich, St. Louis, http://www.sigmaaldrich.com), 2.5 μg/ml fibronectin (MP Biomedicals, Irvine, CA, http://www.mpbio.com), and 20 ng/ml FGF2 (R&D Systems Inc., Minneapolis, http://www.rndsystems.com). Within 10 days, neural tube-like structures developed in the embryoid body outgrowths. These structures were mechanically isolated and propagated as free-floating neurospheres in DMEM/F12 supplemented with one vol/% N2 supplement (Invitrogen) containing 10 ng/ml FGF2 (R&D Systems) for 1–3 weeks. Spheres were triturated to single cells in the presence of trypsin/EDTA and plated onto polyornithine/laminin (Sigma-Aldrich) coated plastic dishes. Cells were further proliferated in DMEM/F12 containing N2 supplement (Invitrogen), 10 ng/ml FGF2, and 10 ng/ml epidermal growth factor (EGF; R&D Systems). Daily media changes were performed during the first 7 days. Cells were passaged every 2–4 days. Differentiation was induced by growth factor withdrawal and propagation in DMEM/F12 (one vol% N2 supplement) and Neurobasal (two vol/% B27 supplement) mixed at a 1:1 ratio; 100 ng/ml cyclic adenosine monophosphate (Sigma-Aldrich) was added to the media. Long-term survival of neurons was promoted by adding 10 ng/ml brain-derived neurotrophic factor (BDNF) and 2 ng/ml glial cell line-derived neurotrophic factor (GDNF; R&D Systems) to the differentiation media. Alternatively, cells were cocultured with primary mouse astrocytes. Cocultures were performed by either plating the sorted cells directly onto the astrocytes or as shared-medium cultures in transwell chambers containing 0.4 μm pore size cell culture inserts (Falcon; BD Biosciences, San Diego, http://www.bdbiosciences.com).

Nucleofection of hESC-Derived Neural Precursors

hESC-derived neural precursors were stably transfected by nucleofection (nucleofection programs B-016 or A-033; Amaxa Biosystems, Gaithersburg, MD, http://www.amaxa.com) with the phuDCX3509-enhanced green fluorescent protein (EGFP) or the phuDCX3509-DsRED2 vector [8]. By 24 hours postnucleofection and replating on polyornithine/laminin-coated (Sigma-Aldrich) plastic dishes, cells were selected with 150 mg G418/ml (activity: 690 mg/g, Sigma-Aldrich) for 2–3 weeks. Individual colonies were manually removed with a 100 μl pipette tip (Eppendorf AG, Hamburg, Germany, http://www.eppendorf.com), replated, and further propagated in the selection media. Recapitulation of doublecortin (DCX) expression by the EGFP reporter was confirmed after 6 days of differentiation via immunofluorescence. Selected clones were further propagated in the presence of FGF2 and EGF; aliquots were frozen in dimethyl sulfoxide (DMSO)-free freezing media (Sigma-Aldrich) and stored in liquid nitrogen.

Fluorescence-Activated Cell Sorting (FACS)

hESC-derived neural precursors stably expressing EGFP under the human DCX promoter and cultured in differentiation media for 8 days were trypsinized, gently resuspended in CytoconBuffer II (Evotec/PerkinElmer Life and Analytical Sciences, Waltham, MA, http://www.perkinelmer.com) and 0.1% deoxyribonuclease (DNase, Invitrogen), and filtered through a 40-μm nylon mesh (Pall GmbH, Dreieich, Germany, http://www.pall.com). The cells were sorted at a concentration of 3,000,000 cells/ml on a fluorescence-activated cell sorter (FACS) Diva (Becton, Dickinson and Company, Franklin Lakes, NJ, http://www.bd.com) using FACSDiva software (BD Biosciences). Cells were selected by forward-angle and side-angle light scatter and for EGFP fluorescence intensity, using an argon-ion laser (Coherent Inc., Santa Clara, CA, http://www.coherent.com) operating at 488 nm. Sorted cells were replated and the purity of sorted fractions was determined visually and by FACS reanalysis. Before sorting, the nozzle, sheath, and sample lines were sterilized with 70% ethanol followed by washes with sterile water. A 70-μm ceramic nozzle (BD Biosciences), sheath pressure of 20–25 pounds per square inch and an acquisition rate of 2,000–3,000 events per second were used as conditions optimized for young neuron sorting.

Preparation of Primary Astrocytes

Primary astrocytes were prepared according to the primary cell isolation protocol from Polleux and Glosh [9]. In detail, mouse pups from postnatal day 3 were anesthetized and decapitated. Preparation of the brains was performed in 6-cm petri dishes containing R6 medium, using a stereo microscope under a sterile laminar flow hood. The brains were removed and rinsed in phosphate-buffered saline (PBS) containing 2% glucose. The specimens were incubated in 3 ml 10 × trypsin at 37°C for 3 minutes and rinsed again. Then R6 medium and 500 μl DNase (1%) were added to a total volume of 5 ml. The tissue was triturated to a cell suspension. Finally, the cells were filtered through a nylon mesh (40 μm) and centrifuged for 10 minutes at 1,000 rpm. Cells (2.5–5 × 106) were plated onto polyornithine/laminin-coated 10-cm dishes in MEF medium (see above). A confluent astrocyte cell layer appeared after 5–10 days.

Cryopreservation of Purified Human Neurons

DCX/EGFP-positive sorted cells were centrifuged in a batch of 5 × 106 cells at 1,000 rpm for 5 minutes in a Megafuge 1.0R (Heraeus; EquipNet Inc., Canton, MA, http://www.equipnet.com). For caspase inhibition the cells were then exposed to 500 nM of the general caspase inhibitor benzyloxycarbonyl-V-A-D-O-methyl fluoromethyl ketone (z-VAD-FMK, R&D Systems) and incubated for 30 minutes before freezing. The cell suspension was frozen in 1.5 ml freezing media containing 10% DMSO (Hybri-Max; Sigma-Aldrich), 20% 500 mM myo-inositol (Sigma-Aldrich), 0.25 V% polyvinyl alcohol stock solution (Merck & Co., Whitehouse Station, NY, http://www.merck.com), and 65% GIBCO Knockout Serum Replacement (Invitrogen). Cells were directly transferred into a Nalgene Cryo freezing container (Nalge Nunc International, Rochester, NY, http://www.nalgenunc.com) and placed at −80°C to achieve a −1°C per minute rate of cooling. Final temperature was reached after freezing in liquid nitrogen. Thawing was carried out by dropping a frozen sample into a 37°C water bath for 2 minutes. Survival rates were determined using standard trypan blue exclusion (trypan blue stain 0.4%, Gibco, Invitrogen.com). Caspase-3 activity was measured 20 hours after thawing and replating of purified cryopreserved cells using a luminescence assay (Caspase-Glo-3/7Assay; Promega, Madison, WI, http://www.promega.com) according to the manufacturer's instructions.

Immunofluorescence Analysis

Cells were fixed in 4% neutral-buffered paraformaldehyde for 20 minutes. Cells were permeabilized with 0.1% Triton X-100 (Sigma-Aldrich) in PBS for 20 minutes. Blocking was performed with 10% fetal calf serum (FCS; Invitrogen) in PBS for 1 hour. Samples were incubated with primary antibodies to human DCX (1:200; #sc8067; Santa Cruz Biotechnology Inc., Santa Cruz, CA, http://www.scbt.com), β-III tubulin (1:2,000; #MM5435P; BAbCo, Cherry Hill, NJ, http://www.babco.com), Green Fluorescent Protein (GFP, 1:2,000; #ab296; Abcam, Cambridge, MA, http://www.abcam.com), human nestin (1:500; #MAB1259; R&D Systems), MAP2ab (1:500; #MAB378; Chemicon, Temecula, CA, http://www.chemicon.com), NeuN (1:100; #MAB377; Chemicon), γ-aminobutyric acid (GABA, 1:500; #A-2052; Sigma-Aldrich), GAD67 (1:250; #MAB 5406; Chemicon), or tyrosine hydroxylase (TH, 1:1,000; #T1299; Sigma-Aldrich) at room temperature for 3–4 hours, washed twice, incubated for 45 minutes with appropriate secondary antibodies (1:500), labeled with CY3 or fluorescein isothiocyanate, counterstained with 4′,6-diamidino-2-phenylindole, and mounted with Vectashield mounting solution (Vector Laboratories, Burlingame, CA, http://www.vectorlabs.com).

Slice Cultures

Hippocampal slices were prepared and maintained in interphase cultures as described [10, 11]. Using a Hamilton syringe (Hamilton Co., Reno, NV, http://www.hamiltoncompany.com) connected to a stereotaxic frame, 1 μl of cell suspension [100,000 cells/μl concentrated in Cytocon Buffer II (Evotec/PerkinElmer Life and Analytical Sciences), 0.1% DNase (Invitrogen)] was deposited on the slice surface. Slices were cultured for up to 4 weeks before electrophysiological recordings were performed.

Electrophysiology

Cells grown on 13-mm diameter plastic coverslips (Nunc, Rochester, NY, http://www.nuncbrand.com) or on rat hippocampal slice cultures were transferred to a chamber that was mounted to an x-y stage and continuously superfused with artifical cerebrospinal fluid (aCSF) at 1–2 ml/min. This aCSF contained the following (in mM): 140 NaCl, 3 KCl, 2 CaCl2, 1 MgCl2, 25 d-glucose, and 10 Hepes/NaOH (pH 7.35, 305–315 mOsmol/kg). Recordings were performed at room temperature. Cells were visualized using an upright microscope equipped with near-infrared differential interference contrast and ×40 water-immersion objective (Carl Zeiss, Jena, Germany, http://www.zeiss.com). In slice cultures transplanted cells were identified by their EGFP fluorescence [11]. Whole cell current-clamp and voltage-clamp recordings were carried out with an Axopatch-200B amplifier (Axon Instruments/Molecular Devices Corp., Union City, CA, http://www.moleculardevices.com) that was interfaced by an A/D-converter (Digidata 1320, Axon Instruments) to a personal computer running pClamp software (version 9, Axon Instruments). For recordings of membrane potential or current, the patch pipette (tip resistance 3–5 MΩ) contained the following (in mM): 120 potassium gluconate (C6H11O7K), 20 KCl, 10 NaCl, 10 EGTA, 1 CaCl2, 4 Mg-ATP, and 0.4 Na-GTP, 10 HEPES/KOH (pH 7.2, 280–290 mOsmol/kg). Command potential in voltage-clamp recordings was corrected for a 13-mV junction potential. For some recordings of postsynaptic currents, another pipette filling solution was used (in mM): 110 cesium methanesulfonate (CH3O3SCs), 10 CsCl, 10 TEA-Cl, 5 QX-314 Cl, 10 EGTA, 1 CaCl2, 4 Mg-ATP, and 0.4 Na-GTP (pH 7.2, 280–290 mOsmol/kg). For the latter solution holding potential was corrected for a 9-mV junction potential. Signals were filtered at 2 kHz and recorded at a rate of 10 kHz.

Statistics

Quantification of marker expression during in vitro differentiation, post-thaw survival, and caspase-3 activity is based on at least three independent experiments performed on duplicate samples. Statistical analysis was performed using Student's t test.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References

Doublecortin: A Reliable Marker for Young Human Neurons

An attractive candidate marker for immature neurons is doublecortin (DCX), a 40-kDa microtubule-associated protein specifically and abundantly expressed in young neurons [12, 13]. In proliferating hESC-derived neural precursors generated according to established conditions [6, 7], expression of DCX was found to be restricted to very occasional spontaneously differentiating neurons (<1%; Fig. 1A). When differentiation was induced by growth factor withdrawal, DCX expression increased alongside the appearance of the neuronal marker β-III tubulin (Fig. 1B, 1E). At this stage, DCX expression was restricted to cells with polar neuronal morphology. Upon further differentiation DCX could also be detected in neurons expressing MAP2ab (Fig. 1C). DCX was barely detectable in neurons positive for the more mature neuronal marker NeuN (Fig. 1D), suggesting that DCX is downregulated in terminally differentiated neurons as observed during normal in vivo development. No expression of the astrocytic glial fibrillary acidic protein or the oligodendroglial antigen O4 was noted in DCX-positive cells (data not shown). Taken together, DCX expression in this hESC-based system delineates an early but postmitotic stage of neuronal differentiation. Accordingly, Ki67 immunostaining and BrdU labeling showed no evidence of cell proliferation within the DCX-positive population (data not shown). Thus, during neural differentiation of hESC, DCX is specifically expressed in young neurons.

thumbnail image

Figure Figure 1.. Expression profile of neural precursors and DCX-positive cells during in vitro differentiation. (A): Cultures of proliferating neural precursors exhibit only very occasional DCX-positive neurons. The expression of DCX and nestin is mutually exclusive. (B): Two days after induction of differentiation, DCX is coexpressed in emerging β-III tubulin-positive neurons. (C): At later stages of neuronal differentiation, DCX-positive neurons coexpress MAP2ab. (D): DCX is hardly detectable in NeuN-positive cells, suggesting downregulation upon terminal differentiation. (E): Quantification of marker expression during the first 8 days of in vitro differentiation of human embryonic stem cell-derived neural precursors. Scale bars: 50 μm. Abbreviation: DCX, doublecortin.

Download figure to PowerPoint

Neuronal Lineage Selection Using a DCX-EGFP Transgene

Based on the highly selective expression of DCX in neurons, we transfected hESC-derived neural precursors with a construct carrying the EGFP gene under control of the human DCX promoter alongside a constitutively expressed neomycin resistance gene. Clones were generated by subsequent selection, expanded, and further investigated for faithful coexpression of DCX protein and the EGFP reporter. Four of 27 clones analyzed showed strong and exclusive expression of the EGFP transgene in DCX-immunoreactive neurons. The remaining clones showed either background expression in nonneuronal cells (n = 5) or an EGFP-expression level insufficient for cell sorting (n = 18). A detailed in vitro differentiation study of the four promising clones revealed that EGFP expression during proliferation was restricted to sporadic spontaneously differentiating cells with immature neuronal morphology (Fig. 2A). When stained for DCX, differentiating cells showed coexpression of DCX and EGFP (Fig. 2B). Expression of nestin and EGFP was largely exclusive (Fig. 2C) and EGFP-labeled cells were consistently positive for β-III tubulin (Fig. 2D). With increasing maturation, they also expressed MAP2ab (Fig. 2E). Thus, the expression of EGFP in the selected differentiated hESC-derived neural precursor clones reflects the endogenous expression profile of the DCX protein. Consolidated clones of DCX-EGFP-transduced cells could be further propagated without losing their neurogenic potential. For example, cells expanded in FGF2 and EGF for 4 and 11 passages and subsequently differentiated for 8 days yielded, respectively, 26 ± 2.2% and 25 ± 2.5% EGFP-positive, DCX-immunolabeled neurons. Expanded clones could be frozen and stored in liquid nitrogen without loss of their proliferative and neurogenic potential.

thumbnail image

Figure Figure 2.. Expression of the DCX-EGFP transgene in differentiating hESC-derived neural precursors. (A): EGFP-positive cells emerging between still undifferentiated precursors. (B): Simultaneous detection of EGFP and DCX protein confirms faithful recapitulation of DCX expression by the reporter construct. (CE): DCX-EGFP-positive cells are negative for nestin (C) but coexpress the neuronal markers β-III tubulin (D) and MAP2ab (E). Cells were captured during proliferation (A) and at 2 (C), 8 (B, D), and 12 (E) days of in vitro differentiation. Scale bars: (A, B): 100 μm; (CE): 50 μm (B, insert): 25 μm. Abbreviations: DCX, doublecortin; EGFP, enhanced green fluorescent protein.

Download figure to PowerPoint

A central issue in neuronal lineage selection is the identification of a time point where the neurons are present in sufficient numbers but still immature enough to tolerate sorting, replating, or cryopreservation. Quantitative analysis of DCX-EGFP-positive cells showed a continuous increase in the numbers of EGFP-positive cells within the first 3 weeks of differentiation (from 28 ± 2.8% by day 7 up to 55 ± 6% by day 21; Fig. 2B). Within the first week of differentiation, the DCX-EGFP-positive cells displayed a still immature morphology with most of them exhibiting only two to three unbranched processes. Less than 50% of the EGFP-positive cells expressed the more mature neuronal marker MAP2ab, indicating a still immature stage of differentiation. Based on these observations, we chose day 8 ± 1 as the ideal time window for sorting. Within this time period, an average of 31 ± 2.8% of the cells expressed the DCX-EGFP transgene.

Preparative FACSorting was performed with EGFP fluorescence intensity as the only sorting criterion. The population of EGFP-positive neurons was readily distinguishable from other cell types (undifferentiated or glial cells) by virtue of their autofluorescence (Fig. 3A). Analysis of the sorted population showed efficient enrichment of the EGFP-positive cells (Fig. 3B). After replating, most of the cells survived and exhibited strong EGFP fluorescence (Fig. 3C), but were negative for nestin (Fig. 3D). Twenty-four hours after plating, up to 95% (92.3 ± 2.5%) of the cells were found to co-express β-III tubulin and EGFP (Fig. 3E). The replated cells initiated neurite outgrowth within the first 24 hours after attachment (Fig. 3C–3F). Four days after FACS the cells exhibited distinct neuronal phenotypes with extension of long neurites forming a network-like architecture (Fig. 3G). Comparable results were obtained with a DCX-promoter construct carrying a DsRED2 gene as a fluorescence detection marker instead of EGFP (data not shown).

thumbnail image

Figure Figure 3.. Fluorescence-activated cell sorting (FACS)-based lineage selection of DCX/EGFP-positive human neurons. (A): After 7–9 days of in vitro differentiation, a large fraction of fluorescent cells is detected in the DCX/EGFP cultures by FACS. (B): Reanalysis reveals enrichment of EGFP-positive cells to up to 95%. (C, D): Twenty-four hours after FACS and replating, sorted cells maintain strong EGFP fluorescence and are largely negative for nestin (D) and positive for β-III tubulin (E). (FH): Neurite outgrowth 24 hours (F), 4 days (G), and 1 month (H) after replating. Cells in (H) were propagated as shared media cultures with primary mouse astrocytes to promote long-term survival. (I, J): Expression of GABA (I) and GAD67 (J) 3 weeks after FACSorting and differentiation in astrocyte-conditioned medium (K). Occasional neurons were found to express tyrosine hydroxylase. Note downregulation of GFP in this neuron, which might be attributable to advanced differentiation. Scale bars: (C, E–H): 50 μm; (D and I): 100 μm; (J and K): 10 μm. Abbreviations: DCX, doublecortin; GFP, EGFP, denotes fluorescence signal from enhanced green fluorescent protein.

Download figure to PowerPoint

Plated neurons survived up to 3 weeks in standard media (containing B27 supplement but no neurotrophic factors). Supplementation of the media with BDNF and GDNF or astrocyte-conditioned medium promoted the development of complex neuronal networks and enhanced the survival of neurons (Fig. 3H). For example, a quantification performed 3 weeks after plating and propagation in astrocyte-conditioned medium revealed that 40 ± 4% of the neurons had survived. These neurons could be further maintained for a total period of more than 2 months without overt cell loss.

The majority of the neurons acquired an GABAergic phenotype. Immunohistochemical analysis performed 3 weeks after FACSorting and replating in astrocyte-conditioned medium showed that 61 ± 5.4% of the purified neurons expressed GABA. These cells displayed a small bi- to multipolar morphology and were partially positive for GAD67 (Fig. 3I, 3J). Other neuronal subtypes such as TH-positive neurons were detected only very occasionally (<0.1%; Fig. 3K), This strong preponderance for GABAergic differentiation was independent of the passage number of the DCX-EGFP cells and corresponds well with the notion that growth factor-expanded neural precursors tend to acquire a GABAergic phenotype [7, 14, [15]16].

Functional Maturation of Purified hESC-Derived Neurons

To assay functionality of the selected cells, we performed whole cell patch-clamp recordings on DCX-sorted neurons. To facilitate long-term survival and functional maturation the cells were cocultured with primary mouse astrocytes. Neurons gained fast transient inward currents sensitive to the selective Na+ channel blocker tetrodotoxin (300 nM) and complex outward currents. The latter could be shown to consist of at least two components: a fast activating and inactivating one sensitive to 3 mM 4-aminopyridine, resembling characteristics of A-type K+ current, and a slowly activating sustained component reminiscent of delayed rectifier type (n = 3, Fig. 4A–4D). All neurons tested (n = 13) fired repetitive action potentials upon sustained depolarization (Fig. 4E). Furthermore, the neurons displayed surface expression of AMPA/kainate and GABAA receptors (n = 3, Fig. 4F) as prerequisite for the formation of glutamatergic and GABAergic synapses.

thumbnail image

Figure Figure 4.. Electrophysiological properties of DCX/EGFP purified neurons differentiated in vitro. Voltage-dependent whole cell current (A) consisted of a fast transient inward component that could be blocked by 300 nM TTx (B) and a complex outward current. The latter consisted of a 4-AP-insensitive, slowly activating, and during the test pulse (90 ms) noninactivating component (C). Digital subtraction of the traces shown in (B) and (C) revealed a 4-AP-sensitive fast activating and inactivating current (D). Neurons displaying the current pattern shown in (A) were able to fire multiple action potentials upon long-lasting (1 second) depolarizing current injection (E). Brief (500 ms) application of either the AMPA/KA receptor agonist kainic acid or GABAA receptor agonist muscimol elicited a clear current response (F). (GI): Postsynaptic currents in DCX-EGFP-selected neurons propagated on rat hippocampal slice cultures. Four weeks after deposition, spontaneous PSCs could be observed that reversed between −60 and −20 mV (G). Electrical stimulation of presynaptic fibers evoked monosynaptic PSCs (H). Plotting PSC amplitude over holding potential revealed a reversal potential of −48 mV (I). Abbreviations: AMPA, α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid; 4-AP, 4-aminopyridine; DCX, doublecortin; EGFP, enhanced green fluorescent protein; KA, kainic acid; PSCs, postsynaptic currents; TTx, tetrodotoxin.

Download figure to PowerPoint

To further explore the ability of the selected neurons to engage in synaptic circuits, we transplanted purified neurons onto hippocampal slices prepared from P10 rats, a paradigm known to promote long-term survival and functional maturation of embryonic stem cell (ESC)-derived neurons [17]. Functional integration was evaluated by patch-clamp recordings 4 weeks after transplantation on those neurons with faint GFP expression (Fig. 4G–4I), given that GFP downregulation indicates a more mature neuronal stage. The cells exhibited both transient and persistent outward and large-amplitude transient inward currents (data not shown). All of the neurons recorded (n = 9) were able to fire repetitive action potentials upon long-lasting depolarization (data not shown). In addition, in eight of nine neurons spontaneous postsynaptic currents (PSCs) could be recorded. In some recordings (n = 3) we used a Cs+-based pipette solution for a closer characterization of those PSCs. Recordings at different holding potentials revealed a reversal potential of postsynaptic currents between −20 and −60 mV (Fig. 4G). Large PSCs in cells located in the CA1 region could be evoked by electrical stimulation with a stimtrode placed in the stratum radiatum (Fig. 4H). Those evoked PSCs reversed polarity at −48 mV (Fig. 4I), close to the calculated ECl = −43 mV, suggesting synaptic input via GABAA receptors. Taken together, these data indicate that the population of human cells selected with the DCX-EGFP transgene acquires functional properties of mature neurons.

Efficient Cryopreservation of hESC-Derived Neurons

An essential prerequisite for the future widespread use of hESC-derived neurons is the development of efficient cryopreservation protocols. When we applied conventional slow-cooling and fast-thawing protocols (e.g., a −1°C per minute cooling rate and 37°C thawing temperature) involving commercial serum-free freezing formulations or standard protocols including 90% FCS and 10% of DMSO, sorted DCX-EGFP-positive cells exhibited relatively low survival rates between 30 and 35% as detected by trypan blue exclusion. We then formulated a freezing solution containing DMSO, the cryoprotectant myo-inositol, polyvinyl alcohol as ice-controlling agent, and serum replacement with an osmolarity of 300–400 mOsm/kg (see Materials and Methods). Using this formula, we were able to enhance postfreeze survival up to 47% (40 ± 5%).

To further increase post-thaw survival we applied the caspase inhibitor z-VAD-FMK to our cells right before freezing. Apoptosis mediated through a caspase-3-dependent mechanism has been shown to be responsible for a significant cryopreservation-associated cell loss, and caspase inhibition has been successfully used to improve cell survival upon freezing of undifferentiated hESC and other cell types [18, [19]20]. We found that preincubation of DCX-EGFP-sorted neurons with 500 nM z-VAD-FMK for 30 minutes significantly reduced the number of cells undergoing cell death, resulting in an increased survival rate of up to 83% (70.1 ± 13.1%) as detected by trypan blue exclusion directly after thawing (Fig. 5A). Furthermore, a caspase-3 assay performed 20 hours post-thawing showed a 3.5-fold reduced caspase-3 activity in z-VAD-FMK-treated neurons compared to untreated control cells (Fig. 5B). Two days after thawing and plating, z-VAD-FMK-treated cells still showed strong EGFP expression (Fig. 5C) and a neuronal morphology with prominent neurites and immunoreactivity to DCX and β-III tubulin (Fig. 5D, 5E). Maintenance of functionality was again confirmed by patch-clamp analysis, which revealed that thawed neurons cocultured for 4 weeks with mouse primary astrocytes generate functional ion channels, fire repetitive action potentials, and display functional AMPA and GABA receptors, thus indicating that cryopreservation does not impair the physiological properties of the cells (data not shown).

thumbnail image

Figure Figure 5.. Cryopreservation of human embryonic stem cell-derived neurons. (A): Post-thaw cell survival of cryopreserved DCX-EGFP-sorted cells with and without pretreatment with z-VAD-FMK. *, p < .05. (B): Quantification of caspase-3 activity of cryopreserved DCX-EGFP-sorted cells with and without z-VAD-FMK pretreatment as detected 20 hours after cryopreservation and replating. *, p < .01. (CE): Two days after thawing sorted and z-VAD-FMK-treated cells maintain strong expression of the DCX-EGFP transgene (C), DCX protein (D), and the neuronal marker β-III tubulin (E). Scale bars: (C and E): 50 μm; (D): 100 μm. Abbreviations: DCX, doublecortin; EGFP, enhanced green fluorescent protein; z-VAD-FMK, benzyloxycarbonyl-V-A-D-O-methyl fluoromethyl ketone.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References

The results of this study demonstrate that expression of the EGFP gene under control of the human DCX promoter permits efficient FACS-based lineage selection of immature hESC-derived neurons. Stable transfection and subsequent selection enabled the generation of hESC-derived neural precursors that, upon differentiation, express EGFP exclusively in neurons. These neurons are amenable to FACS-based enrichment, yielding purities of >90%, and undergo functional maturation. Finally, we developed a cryopreservation protocol that permits long-term storage of the DCX-EGFP-positive cells with post-thawing mean survival rates of 70%, thus enabling banking and on-demand differentiation of highly purified human neurons.

In the past, several markers have been explored as potential candidates for the purification of ESC-derived neural cells. Immunoisolation has successfully been applied to neurons [21, 22] and glia [23] in mouse and human systems. Recently, Pruzak and colleagues [24] published a comprehensive analysis on the applicability of surface epitope-mediated selection of human ESC-derived neural lineages. Although selection based on surface markers bypasses genetic modification and can thus be applied to a variety of cell sources, it is often hampered by high costs for large-scale applications, lack of specificity of the individual antibodies, and heterotopic expression of the target epitopes in unwanted cell types. For example, expression of the widely used neural selection epitopes neural cell adhesion molecule (CD56), NG2, and CD133 has also been observed in lymphocytes, melanocytes, and hematopoietic cells, respectively [25, [26]27].

Genetic lineage selection using genes encoding fluorescent proteins or antibiotic resistance is another widely used method. Transgenic murine ESCs harboring a marker gene have been successfully used to monitor and isolate multiple cell types including neurons [28] and even neuronal subtypes [29, 30]. Although these approaches have the disadvantage that the selected cells carry a genetic modification, the beauty of this system lies in its reliability and simplicity once established. Working with human ESC implies long differentiation protocols. The availability of proliferating and cryopreservable neural precursors stably expressing a cell type-specific selection marker largely facilitates the generation of purified somatic cell types, enabling—in principle—lineage selection on demand.

Considering that an immature, neuroblast-like stage of differentiation would be the most desired target for neuronal lineage selection, we found DCX to be a perfect marker. DCX is expressed in newborn and immature neurons in the developing and adult central nervous system (CNS) and peripheral nervous system [8, 12, 13, 31]. Previous studies in transgenic mice have shown that the DCX promoter enables transgene expression in immature neurons throughout the CNS [32, 33]. We could confirm this neuronal expression pattern in our hESC culture system because, in the selected clones, GFP fluorescence was strongly and specifically detected in early neuronal cells with subsequent downregulation during further maturation. Importantly, the neurogenic potential and the neuron-specific expression of GFP were maintained through further passaging and cryopreservation of the transduced cells, thereby rendering them a bankable resource of cells for on-demand lineage selection. As a result of their early stage of neuronal differentiation, DCX-EGFP-selected cells were robust enough to survive the harsh requirements associated with FACS such as enzymatic digestion and mechanical pressure during the sorting procedure—stress factors not well tolerated by mature neurons (data not shown). DCX-EGFP-based selection was found to be highly efficient. Starting from three to four 10-cm dishes with a total of 8 × 107 cells, a typical FACS session of 4–5 hours yielded more than 2 × 107 purified neurons. In contrast, genetic lineage selection using promoter elements of genes expressed in more differentiated neurons such as synapsin results in poor survival (Pruzak et al. [24] and the authors' own unpublished observations). Importantly, DCX-EGFP-sorted neurons retained functionality (i.e., they were able to generate action potentials and to receive synaptic input).

The immature state of the DCX-EGFP-sorted neurons was also an important prerequisite for their amenability to cryopreservation. To optimize cell survival, we supplemented our freezing solution with myo-inositol and polyvinyl alcohol. Myo-inositol is found in cyanobacteria, algae, fungi, and plants, where it functions as an osmolyte, to tolerate or avoid freezing [34]. Polyvinyl alcohol is well known to prevent or delay ice nucleation [35]. A further increase in post-thaw cell survival was achieved by preincubation with the caspase inhibitor z-VAD-FMK. Caspase-3 has been shown to be crucial in cryopreservation-mediated apoptosis [18, 20]. Although the inhibitor appears to have an in vivo half-lifetime of less than 24 hours [36], it suffices to protect the cells during the cryopreservation process as well as in the first hours after thawing and plating. However, the effect is temporary because staurosporine treatment 4 days after thawing showed no difference in induced apoptosis between z-VAD-FMK-treated and nontreated cells (data not shown). Thus, caspase-3 inhibitor treatment should not impair pharmacological toxicity studies in these cryopreserved hESC-derived neurons.

Conclusion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References

Our observations provide interesting prospects for the use of DCX-EGFP-selected cells in screening functional effects of pharmacological or toxicological compounds. Considering their immature stage of differentiation, DCX-EGFP-selected neurons might also represent an attractive donor source for neuronal transplants. In principle, our DCX-based neuronal lineage selection approach should be applicable to other human ESC-like cells. In this context, recently described alternative sources such as induced pluripotent stem cells obtained by transcription factor-based reprogramming of neonatal and adult human fibroblasts [37, [38], [39]40] might further extend potential applications of this technology.

Disclosure of Potential Conflicts of Interest

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References

The authors indicate no potential conflicts of interest.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References

We thank Joseph Itskovitz-Eldor and Michal Amit for providing the H9.2 cell line, Barbara Steinfarz for technical assistance, and Jerome Mertens for providing primary mouse astrocytes as well as editorial support. This work was supported by the European Union (LSHG-CT-2006-018739; ESTOOLS), the Deutsche Forschungsgemeinschaft (SFB TR3 D2), and the Hertie Foundation.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Conclusion
  8. Disclosure of Potential Conflicts of Interest
  9. Acknowledgements
  10. References