SEARCH

SEARCH BY CITATION

Keywords:

  • Calcium flux;
  • Tissue regeneration;
  • Neuron;
  • Embryonic stem cells

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

A clear understanding of cell fate regulation during differentiation is key in successfully using stem cells for therapeutic applications. Here, we report that mild electrical stimulation strongly influences embryonic stem cells to assume a neuronal fate. Although the resulting neuronal cells showed no sign of specific terminal differentiation in culture, they showed potential to differentiate into various types of neurons in vivo, and, in adult mice, contributed to the injured spinal cord as neuronal cells. Induction of calcium ion influx is significant in this differentiation system. This phenomenon opens up possibilities for understanding novel mechanisms underlying cellular differentiation and early development, and, perhaps more importantly, suggests possibilities for treatments in medical contexts.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Embryonic stem (ES) cells are pluripotent cells that can, in vivo and ex vivo, give rise to cells of different fates. Through induction, ES cells form embryoid bodies (EBs), and, in this way, ES cells are capable of differentiating into a variety of tissue types such as extraembryonic endoderm and neural and muscle tissue [1, [2]3]. Many growth factors and signaling pathways initiate differentiation and modulate the course of cellular differentiation [4, [5], [6]7]. Several reports show that, with the application of certain growth factors or the alteration of culture conditions, ES cells can differentiate in a relatively efficient manner into specific neuronal cell types that are destined to become certain neuronal tissues [5, 8, [9], [10]11].

Another important factor that possibly influences developmental processes and is central for cellular homeostasis is transmembrane ion distribution. Unequal distributions of ions between the intra- and extracellular compartments yield electrical potential, which is crucial for neural transmission. Ions also play a role in shaping neuronal circuits during development via neural transmission; ions induce functional and structural refinement of synapses and neuronal networks by modulating activity-dependent gene transcription [12, 13]. Despite the important role of ionic density in development, very little information is available on the roles of intra- and extracellular ionic density in cell-fate determination. Here, we report that electrical stimulation can bias the fate of differentiating ES cells toward neuronal lineages. Growth factor-induced ES cells usually differentiate into rather restricted neuronal cell types. In contrast, electrically induced ES cells that ultimately differentiate into neurons are plastic in their capacity to differentiate into a wide variety of specific cell types. These ES cells are pluripotent, capable of differentiating into any neuronal lineage found within the various neuronal tissue types we examined.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Fluorescent ES Cells

Venus-expressing ES cells were prepared by transfecting R1 ES cells with a construct containing Venus driven by a CAGGS promoter. R1 ES cells were a gift from Dr. Andreas Nagy. The Venus construct was made from plasmids provided by Dr. Jun-ichi Miyazaki [14, 15].

Differentiation Method in Cell Culture

The protocol for ES cell differentiation is schematized in Figure 1. Embryoid bodies (EBs) were made by culturing R1 ES cells with Dulbecco's modified Eagle's medium (DMEM) containing 10% fetal calf serum (FCS) without leukemia-inhibitory factor (LIF) in noncoated bacterial petri dishes (Nunc, New York, http://www.nalgenunc.com). After 3 days of culture, electrical stimulation was applied to cells in a 4-mm gap cuvette under several voltage conditions (0, 5, 10, and 20 V; see supplemental data). One train of five pulses (950-millisecond interpulse interval) was delivered with an electroporater (CUY21E; Tokiwa Science, Tokyo, http://www.tokiwakagaku.jp). For cell culture experiments, stimulated EBs were maintained in DMEM with 10% FCS (GIBCO, Carslbad, CA, http://www.invitrogen.com) on poly(d-lysine)-coated plates (BD, Franklin Lakes, NJ, http://www.bd.com). Ten days after stimulation, cells were fixed in paraformaldehyde in phosphate-buffered saline (pH 7.4) for immunocytochemical analyses. For animal experiments, Venus-positive EBs were stimulated similarly (10 V, same five-pulse train) and then dissociated with trypsin-EDTA for 3 minutes. Dissociated cells were injected into either mouse embryos or adult spinal cords.

thumbnail image

Figure Figure 1.. Experimental design. Embryoid bodies (EBs) were made by culturing embryonic stem (ES) cells with DMEM containing 10% FCS in noncoated bacterial petri dishes (Nunc). Electrical stimulation was applied to cells in a 4-mm gap cuvette under several voltage conditions (0, 5, 10, and 20 V; see supplemental data). For cell culture experiments, stimulated EBs were maintained in DMEM with 10% FCS on poly-d-Lysine-coated plates. For animal experiments, Venus-positive EBs were stimulated similarly (10 V, same 5-pulse train) and then dissociated with trypsin-EDTA for 3 minutes. Dissociated cells were injected into C57BL/6 blastocysts. Abbreviations: DMEM, Dulbecco's modified Eagle's medium; FCS, fetal calf serum.

Download figure to PowerPoint

Immunostaining

Cultured cells or histological sections were processed for immunostaining using the following antibodies: anti-TuJ1 mouse monoclonal antibody (mAb; 1:500; BAbCO, Berkeley, CA, http://www.babco.com), anti-GFP rat mAb (Nacalai, Kyoto, Japan, http://www.nacalai.co.jp), anti-Hu human polyclonal antibody (pAb; 1:1,000; a gift from Dr. Robert Darnell), anti-Ki67 rat mAb (DAKO, Glostrup, Denmark, http://www.dako.com), anti-MAP2 mouse mAb (1:200; Chemicon, Temecula, CA, http://www.chemicon.com), anti-ChAT rabbit pAb (1:200; Chemicon), anti-Islet1 mAb (1:400; Developmental Studies Hybridoma Bank [DSHB]), anti-Pax6 mAb (1:200; DSHB), anti-Pax7 mAb (1:400; DSHB), anti-MNR2 mAb (1:400; DSHB), and anti-Nkx2.2 mAb (1:400; DSHB). Histological sections were stained with the protocol described previously [16], and immunostained images were obtained with an LSM-510 confocal laser microscope (Carl Zeiss, Jena, Germany, http://www.zeiss.com/) by sequential scanning and analyzed with adjunctive software attached to the LSM-510. The thickness of the histological sections was less than 7 μm, and the z-axis sampling of the confocal images was less than 1 μm.

Measurement of Intracellular Ca2+ Concentration

ES cells or EBs were loaded with the Ca2+ fluorescence indicator fura-2 by incubating the cells in Hank's balanced salt solution (HBSS) containing 2 μM fura-2 AM (Molecular Probes Inc., Eugene, OR, http://probes.invitrogen.com) and 0.01% cremophor-EL (Sigma Chemicals, St. Louis, MO, http://www.sigmaaldrich.com/) at room temperature for 30 minutes. After loading, cells were washed in fresh HBSS and incubated an additional 15+ minutes before analysis of intracellular Ca2+ concentration ([Ca2+]i). [Ca2+]i was analyzed with an inverted fluorescent microscope (IX-70, Olympus, Tokyo, http://www.olympus-global.com) equipped with a filter exchanger (Lambda 10-2, Sutter Instruments, Novato, CA, http://www.sutter.com/index.html) and a cooled charge-coupled device camera (MicroMax, Roper Scientific, http://www.roperscientific.com/). One train of five pulses (950-millisecond interpulse interval) was delivered with an electroporater (CUY459G20; Nepa Gene, Chiba, Japan, http://www.nepagene.jp). The following optics were used: excitation filter, 340HT15, and 380HT15; dichroic mirror, 430DCLP; emission filter, 510WT40 (all from Omega Optics, Austin, TX, http://www.omegaoptics.com/); and objective lens, Uapo/340 20x/0.75 (Olympus). Metafluor 5.0 software (Universal Imaging) was used to control the system and analyze acquired images [17].

Spinal Cord Injury Model

Spinal cord injury was induced with a modified NYU impactor as described previously [18]. Briefly, female C57BL/6J mice were anesthetized with an intraperitoneal injection of ketamine (100 mg/kg) and xylazine (10 mg/kg). After laminectomy at the T9 level, the dorsal surface of the dura matter was exposed. The vertebral column was stabilized with fine forceps and clamps at the T7 and T10 spinous processes and ligament, and then the animal's body was lifted. A 3-g weight (1.2-mm diameter tip) was allowed to drop from a height of 25 mm onto the dorsal surface of the dura matter. The muscles and the incision were then closed in layers, and the animals were placed in a temperature-controlled chamber until thermoregulation was reestablished. Manual bladder evacuation was performed twice per day until reflex bladder emptying was reestablished.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Cell Fate Determination of Electrically Stimulated Cells in Culture

We examined the influence of inter- and intracellular ionic balance on differentiation fate by application of weak electrical pulses. Embryoid bodies were stimulated at one of several intensities via an electrode (Fig. 1). The EBs were cultured for 10 days, and then fixed to assess differentiation fate (Fig. 1). Assessment for various markers (i.e., mainly for muscle and neural tissue) indicated that R1 ES cells showed no neuronal or myocytic differentiation, regardless of whether they were electrically stimulated (Fig. 2A, 2B, 2M, 2N). In contrast, EBs receiving electrical stimulation showed robust neuronal differentiation; control EBs (i.e., those receiving no stimulation) showed little differentiation (Fig. 2). Almost all colonies of EBs receiving 10-V stimulation contained cells immunoreactive for TuJ1, a marker for early committed neuronal cells, whereas less than 10% of control colonies contained TuJ1-positive cells derived from EBs receiving 0-V stimulation (Fig. 2A, 2C). We confirmed the neuronal identity of the cells from colonies that received 10-V stimulation: The majority of these cells were MAP2 immunoreactive (supplemental Fig. S1), whereas 20%–30% of cells showed immunoreactivity to MAP2 by retinoid treatment.

thumbnail image

Figure Figure 2.. Effect of electrical stimulation on embryonic stem (ES) cell differentiation in culture. (A–D): Increasing, mild electrical stimulation disproportionately biases ES cell differentiation toward a neuronal fate. Percentage of colonies containing cells that express the neuronal marker TuJ1 (A), and those that express the muscle marker α-actinin (B): Black, filled circles, original ES cells; white open circles, ES cells cultured for 1 day to make embryoid bodies (EBs; before electrical stimulation); blue, filled circles, ES cells cultured for 2 days to make EBs (before electrical stimulation); and red, filled circles, ES cultured for 3 days to make EBs (before electrical stimulation). Number of TuJ1-positive cells per colony (C) and α-actinin-positive cells per colony (D), both as a function of stimulation intensity. Daggers indicate p < .001 and asterisks indicate p < .05 compared to TuJ1-positive cells in zero-volt condition. Statistical differences between groups were assessed with Student's t test. A p value of at least p < .05 was considered significant. Numbers in parentheses in (C) and (D) indicate the number of colonies containing TuJ1-positive cells. (E–G): Appearance of unstimulated control EBs. Although the majority of colonies did not contain TuJ1-positive cells (E), a few TuJ1-positive cells were present (G). (F): Nuclear staining of EBs in (E) shows the density of cells. (H–L): Appearance of stimulated EBs. Anti-TuJ1 immunostaining of EBs subjected to either 5- (H), 10- (I), or 20-V (J) pulse stimulation. (K): Higher magnification of anti-TuJ1 immunostained EBs stimulated with 10 V. (L): Anti-α-actinin immunostained EBs stimulated with 10 V. (M): Nuclear staining demonstrates the existence of cells and lack of anti-TuJ1 immunostaining of ES cells. (N): Neuronal differentiation of ES cells was not induced. Nuclear stain was propidium iodide. Scale bars = 1 mm (E–G), 100 μm (H–J, L), 500 μm (K), and 200 μm (M, N).

Download figure to PowerPoint

It is noteworthy that the neuronal cells in our system differentiated in a significantly shorter time than did those in most of other systems that use growth factors to initiate cell differentiation [5, 10, 11]. The differentiation efficiency decreased slightly in cells that received 20-V stimulation compared to that in cells that received milder stimulation. Although the morphology of these cells also clearly differed (e.g., thicker dendritic processes than in the ones receiving milder stimulation), all expressed TuJ1 (Fig. 2J). The size and number of colonies produced from cells stimulated with 20 V did not clearly differ from the size and number of colonies produced from the unstimulated cells. The size and number of colonies produced from cells stimulated with 5–15 V also did not differ. We note that the cell number and cell death did not show significant difference among EBs with or without electrical stimulation after outgrowth on the poly(d-lysine) plate (tunnel assay showed 11%–13% of cell death at 1 day and 3 days after outgrowth of EBs with electrical stimulation and without stimulation).

In general, we observed few muscle progenitors as a result of electrical stimulation, even though a slight but insignificant, increase in muscle progenitors was observed for EBs receiving 5-V stimulation (Fig. 2B, 2D). In addition, we did not observe cells differentiating into glial cells in our system (i.e., no glial fibrillary acidic protein [GFAP]-immunoreactive cells; supplemental Fig. S1). Electrical stimulation induced EBs to differentiate somewhat specifically into neuronal cells. However, because we failed to observe various differentiation markers (e.g., Islet1, Pax6, Pax7, MNR2, Nkx2.2, tyrosine hydroxylase, GAD65, Islet1) for specific neuronal cell types within 10 days of culture in immunocytochemical analysis, whereas we detected slight elevation of transcription of Pax6, NeuroD1, and some LIM-homeodomain genes by RT-PCR after 10 days of culture (data not shown), it is most likely that the TuJ1-positive cells of this ex vivo system did not reach the stage at which such neuronal markers are expressed.

Importance of Calcium Ion Influx for Modulation of Differentiation Fate

The mechanism that modulates fate determination in this system is unknown. To determine this mechanism, first, we examined the role of calcium by measuring intracellular Ca2+ concentration in ES cells and EBs before and after electrical stimulation (Fig. 3). It is well established that Ca2+ is an important signal transducer or modulator [19]. ES cells or EBs loaded with the fluorescent Ca2+ indicator fura-2 were electrically stimulated [17], and the resulting fluorescence was measured. ES cells showed no change in fluorescence emission after electrical stimulation, indicating that electrical stimulation did not induce release or uptake of Ca2+ in ES cells (Fig. 3B). In contrast, numerous cells within EBs showed small but significant changes in fluorescence after electrical stimulation, indicating that this stimulation induced an increase in Ca2+ concentration in EBs (Fig. 3A). Addition of the Ca2+ chelator EGTA (25 mM) to the EB culture medium before electrical stimulation prevented any measurable change in fura-2 fluorescence (Fig. 3C), indicating that the source of Ca2+ was extracellular rather than intracellular.

thumbnail image

Figure Figure 3.. Electrical stimulation-induced calcium influx into cells. (A–C): Changes of [Ca2+]i in embryoid bodies (EBs) (A, B) and in ES cells (C) after electrical stimulation. Ratio of fluorescence excitation intensities at 340 nm to 380 nm is plotted as a function of time. Filled arrowheads indicate time of electrical stimulation. Three typical patterns are displayed for each plot. Inset in (A) shows a fluorescent image of fura-2-loaded EBs excited at 380 nm. The culture medium contained either 2 mM Ca2+(A, C) or 25 mM EGTA (B). Cells were stimulated at 30 V (5–6 W) in this experiment. Due to the different buffer composition for this experiment, a higher voltage was required to produce comparable power (5–6 W) to that produced in the 10–15-V condition of the experiments presented in Figure 1. This stimulus intensity (one associated with 5–6 W) induces neuronal differentiation. (D): Mean number of neuronal cells counted per colony in the presence and/or absence of the calcium chelator EGTA and stimulation with electric pulses (EP). The number of neuronal cells decreased dramatically when Ca2+ was absent from the medium. The same stimulation condition that yielded TuJ1-positive cells failed to yield neuronal cells when 25 mM EGTA was added to the medium. Open bars show data for incubation without EGTA and closed bars with EGTA. Number of colonies counted is indicated in parentheses. Again, due to the difference in culture conditions in this experiment, cell growth was largely disturbed in comparison with the conditions used in the experiments of Figure 1. The absolute number of cells after culturing was approximately one-tenth of the number of cells in the experiments presented in Figure 1. Statistical differences between groups were assessed with the Student's t test. Error bars are SEM. p = .00057 for comparison (∗) between stimulated EBs with EGTA present and stimulated EBs with EGTA absent.

Download figure to PowerPoint

To determine the significance of Ca2+ influx for electrically induced neuronal differentiation, EBs were electrically stimulated in the presence or absence of EGTA, and then these were cultured in the absence of EGTA and monitored for signs of differentiation fate. EBs electrically stimulated in the presence of EGTA failed to assume a neuronal fate, whereas those stimulated in the absence of EGTA assumed a neuronal fate (Fig. 3D). These results indicate that Ca2+ influx is, at the very least, necessary for neuronal cell fate determination in this system.

It is possible that this Ca2+ influx was not mediated by ionic channels, but instead, simply by physical disruption of the cell membranes caused by electrical stimulation. We believe that the latter is unlikely, because the emission ratio for fura-2 was significantly lower in electrically stimulated cells compared to that in cells with membrane fractures (Fig. 3A), suggesting that the increased Ca2+ influx did not result from passive influx caused by membrane fractures. In addition, we observed no blue staining in the cells when we applied trypan blue to the culture medium at the time of stimulation with parameters that successfully induced neuronal differentiation. However, we did observe staining in cells stimulated with higher voltage pulses (data not shown). The lack of trypan blue staining with milder stimulation parameters indicates that our parameters were sufficiently weak to avoid membrane fractures. Thus, the increased Ca2+ influx we observed was most likely mediated by Ca2+ ion channels, which probably play a key role in fate determination in our model system. However, we could not determine the type of Ca2+ channel at work in this system by applying inhibitors such as nifedepine or ω-conotoxin MVIIC. In addition, we could not detect significant expression of subunits from either of Ca2+ channels L, N, and P/Q types, by reverse transcriptase polymerase chain reaction (RT-PCR) analysis on EB after 3 days of aggregation (data not shown).

Incorporation of Electrically Stimulated ES Cells into Neural Tissues of Embryos

To determine whether electrically stimulated EBs were capable of differentiating into mature neurons in vivo, stimulated or unstimulated EBs were injected into mouse embryos, and the fate of these cells was traced during the course of embryonic development. In this set of experiments, we used an ES cell line ubiquitously expressing Venus, an enhanced yellow fluorescent protein derivative [14]. When injected into mouse blastocysts, line LCVL10 ES cells incorporated equally into the bodies of embryos and neonatal mice. As before, ES cells expressing Venus were induced to make EBs, which were then either stimulated with 10-V pulses or were unstimulated, and then injected into mouse blastocysts (supplemental Fig. S2). Thirteen embryos and 15 embryos were recovered from the 10-V and unstimulated conditions, respectively.

Unstimulated EBs failed to incorporate into specific tissues in most embryos tested (10 of the 11-days-postcoitum [dpc] embryos and five of the 13-dpc embryos; supplemental Fig. S2). We observed fluorescence nonspecifically across the embryo and in extraembryonic structures, such as in the yolk sac of two embryos (one from 11-dpc and the other from 13-dpc embryos; supplemental Fig. S2; supplemental Table S2). The 11-dpc embryo that ES cells heavily contributed showed abnormality (supplemental Fig. S2). In two other 11-dpc embryos, strong fluorescence was observed only in the yolk sac, whereas virtually none was observed in the embryo proper (data not shown). In the remaining 11 embryos, EBs did not incorporate (data not shown).

Cells arising from the stimulated EBs (EPs) tended to incorporate into neural tissue (supplemental Fig. S2; supplemental Table S2). We recovered nine embryos at 11 dpc and four embryos at 13 dpc. All of these recovered embryos were morphologically normal. Upon closer examination in whole-mount preparations, fluorescence appeared to be primarily localized to dorsal structures (i.e., the central nervous system [CNS]) in seven embryos (four of the 11-dpc embryos and three of the 13-dpc embryos) with small pockets of fluorescence in other structures, including the peripheral nervous system (PNS). In addition, in two of the 11-dpc embryos, we observed minor fluorescence in the PNS. For one embryo acquired at 13 dpc, the heart emitted strong fluorescence, and the PNS emitted weak fluorescence (supplemental Table S2). In summary, we observed that ES-derived Venus positive cells contributed primarily to CNS in 7 embryos among 13 recovered embryos when electrical stimulation is applied.

To determine whether incorporated EBs differentiated into proper neurons, we assessed several neuronal markers immunohistochemically. Transverse sections of two embryos recovered as 11 dpc and one from 13-dpc embryo were prepared and double immunostained for GFP and one of the following neuronal makers: TuJ1, Islet1, Pax6 (Fig. 4), LIM3, Pax7, or MNR2 [20]. All three preparations of CNS structures clearly contained numerous double-immunostained cells, as shown in Figure 4, for Islet1 and Pax6 (i.e., cells that were immunoreactive for both GFP and one of the above-mentioned neuronal markers), suggesting that, in vivo, cells from electrically stimulated EBs can differentiate into proper functional neurons. Although electrically stimulated cells tended to incorporate into ventral spinal cord to form motor neurons and interneurons, they can, in principle, incorporate into various neural structures across the dorsoventral axis to form different types of mature neurons (Fig. 4). With respect to the anteroposterior axis, we also found GFP-labeled cells in forebrain, hindbrain, brain stem, and spinal cord (Fig. 4).

thumbnail image

Figure Figure 4.. Distribution of electrically stimulated embryonic stem (ES) cells implanted in mouse embryos. An embryo at 11 days postcoitum (dpc) was sectioned to examine cell-type specificity of incorporated fluorescent ES cells in the central nervous system (CNS). Incorporation of ES cells into brain (A–C) and spinal cord (D–Q), shown in transverse sections. Green fluorescent puncta are anti-GFP-positive cells expressing Venus (A, C, D, G, H, K), red are TuJ1-positive cells, indicating differentiated neurons (B, E, G, I, J), and blue shows nuclear staining with TO-PRO3 (Molecular Probes) (F, G, J, K). Small boxes in D–G show areas of high magnification presented in H–K, respectively. (L–Q) High magnification images of a 13-dpc embryo showing that ES cells differentiated into a variety of neuron types, including motor neurons, interneurons, or their precursors. Green signals represent GFP, and red signals, Islet1 (L–N), Pax6 (O–Q). Scale bars = 500 μm (A–C), 250 μm (D–G), 50 μm (H–K), and 40 μm (L–Q). Abbreviation: GFP, green fluorescent protein.

Download figure to PowerPoint

Together, our data demonstrate that electrically stimulated EBs differentiate primarily into early committed neuronal cells. These cells are plastic and can differentiate into a variety of specific neuronal cell types in accordance with the environment.

Electrically Stimulated EBs Contribute to Injured Adult Spinal Cords

Our embryo experiments prompted us to examine whether electrically stimulated EBs are capable of differentiating into neurons when grafted into the injured spinal cords of adult mice. Stimulated EBs produced from Venus-expressing ES cells were injected into spinal cords of mice 7 days after injury, and their survival and phenotype within the spinal cord was assessed 50 days after transplantation (Fig. 5D). As control experiments, we also grafted unstimulated EBs or unstimulated ES cells into the spinal cords of adult mice (Fig. 5A–5C). In each animal, 106 cells were injected in a volume of 5 μl. Of the five animals injected with stimulated EBs, one failed to survive due to injury, and all of the others showed evidence of incorporation into the spinal cord as neuronal cells. Two of five animals injected with unstimulated EBs and three of six animals injected with unstimulated ES cells died within 50 days. All of the survivors from both of these control experiments showed tumorigenic or pathogenic features (Fig. 5B, 5C).

thumbnail image

Figure Figure 5.. Longitudinal sections of injured, adult mouse spinal cords injected with electrically stimulated or nonstimulated cells. (A–D): Hematoxylin-eosin staining. Injured spinal cords 57 days after trauma was induced without injection of ES cells (ES; control) (A). (B–D): Untreated or treated ES cells were injected 7 days after injury. Histological analysis was performed 50 days after injection with unstimulated ES cells (B), with nonstimulated EBs (C), or stimulated EBs (EP) (D). Scale bar = 1 mm. (D): Incorporation of electrically stimulated EBs into injured adult spinal cord. (Db): Cells derived from stimulated EBs (EP) are positive for GFP. (Dc): ES cells from stimulated EBs differentiated into Hu-positive neuronal cells. Abbreviations: EB, embryoid body; EP, electropulsed; ES, embryonic stem; GFP, green fluorescent protein.

Download figure to PowerPoint

Immunostaining revealed that a large part of electrically stimulated EBs were capable of differentiating into Hu-positive neuronal cell types (Fig. 5D; Fig. 6A–6D). We detected no cells that differentiated into GFAP-positive astrocytes or NG2-positive oligodendrocytes (data not shown). In these experiments, approximately 80% of grafted cells assumed a neuronal lineage, clearly expressing neural markers including Hu (Fig. 6A–6D, 6M). Furthermore, Venus-positive cells also immunostained for MAP2, indicating that the exogenous cell population differentiated into neurons (Fig. 7A–7C). On the contrary, even surviving animals possessed few Hu-positive neuron-like cells derived from unstimulated EBs or ES cells (Fig. 6E–6M). Notably, in all the survivors, we observed pathology within the grafts (e.g., infiltration of inflammatory cells, such as macrophages, or ectopically formed tubular structures surrounded by epithelia-like cells; Fig. 5B, 5C). In both cases, many of the injected cells were reactive for phosphorylated histone H3 (phospho-H3) or Ki67 antibodies, suggesting that the grafted cells maintained proliferative activity and thus can be tumorigenic (Fig. 6E–6L, 6N). On the other hand, stimulated EBs showed essentially no phospho-H3 and Ki67 immunoreactivity (Fig. 6A–6D, 6N).

thumbnail image

Figure Figure 6.. Stimulated EBs frequently adopted the appearance of neuronal cells when injected into spinal cord. (A–D): Almost all the cells derived from stimulated EBs (EP) displayed Hu immunoreactivity but not Ki67 immunoreactivity, whereas cells derived from unstimulated EBs (EB; E–H) or ES cells (I–L) displayed Ki67 immunoreactivity but not Hu immunoreactivity. (M): Graph showing the percentage of cells that coexpress both GFP and Hu. (N): Graph showing the percentage of cells that coexpress both GFP and Ki67. Percentages indicated were average of percentages of Hu- or Ki67-positive cell counts obtained by tallying the GFP-positive cells observed in more than 10 different focal planes of each EP, EB, and ES examined. Green, blue, and red signals indicate GFP, Hu, and Ki67 immunoreactivity, respectively (A–L). Scale bar = 50 μm. Abbreviations: EB, embryoid body; EP, electropulsed; ES, embryonic stem; GFP, green fluorescent protein.

Download figure to PowerPoint

thumbnail image

Figure Figure 7.. Stimulated embryoid bodies (EBs) adopted the appearance of various types of neuronal cells when injected into spinal cord. (A–C): Almost all the cells derived from stimulated EBs (electropulsed [EP]) displayed MAP2 immunoreactivity, indicating that they had differentiated into mature neurons. Some of the cells derived from EPs expressed ChAT, a motor neuron marker (D–F), or parvalbumin, an inhibitory neuron marker (G–I). Scale bar = 50 μm.

Download figure to PowerPoint

In summary, stimulated EBs contributed robustly to form neurons within the spinal cord, whereas unstimulated and stimulated ES cells, as well as unstimulated EBs, formed few neurons (Fig. 5). In addition, grafts appeared to display pathological features (Fig. 5).

To analyze in more detail the fate of these neural cells, we performed additional immunohistochemical analyses using antibodies against several neuronal markers. ChAT and Islet1, markers found in spinal motor neurons, colocalized within Venus-expressing cells (Fig. 7D–7F; data not shown) [21], indicating that grafted cells have the potential to differentiate into motor neurons. We detected a few parvalbumin- and γ-aminobutyric acid-positive cells among the Venus-expressing cells that displayed typical neuron-like morphology (Fig. 7G–7I; data not shown). Stimulated EBs incorporated quite well into spinal cord tissues, differentiating into a variety of neuronal cell types and mixing with the recipients' own cells (Fig. 7).

These experiments with injured spinal cord established that stimulated EBs are capable of differentiating into mature neurons when inserted into an in vivo environment. It is possible that injury to the spinal cord provided environmental cues necessary to direct differentiation. This hypothesis is supported by findings that adult neuronal stem cells differentiate into glial cells in response to spinal cord injury [22, [23]24]. Growth factors and cytokines are released, and these factors, in turn, can modulate proliferation and differentiation of neuronal stem cells [25, [26], [27]28].

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

ES cells that have initiated differentiation steps will preferentially assume a neuronal fate when electrically stimulated. In comparison with other systems that produce neural cells from ES cells, the neuronal cells in our system differentiated in a significantly shorter period than did those produced by other methods [5, 10, 11]. Moreover, these stimulated ES cells could be easily transplanted into animal tissues. According to our observations, the differentiation process can be divided into three steps: (a) destabilization of undifferentiated ES cells, (b) modulation of cell fate direction, and (c) differentiation to a mature, terminal cell state. Unlike other induction methods, electrical stimulation seems to work only during the second step. There are several reports on electrical stimulation-induced neurite extension and growth cone guidance cue of PC12 cells or some neural cells [29, [30], [31], [32]33], which are mostly destined to neural fate, but not pluripotent, as ES cells are. In these reports, electrical stimulation alters their morphology, which probably corresponds to the third or even later phase of developmental process in this scheme.

These cells possess terminal differentiation plasticity, requiring further steps to determine a neuronal destiny. They failed to express any specific markers indicative of mature neurons ex vivo, and were flexible or plastic; thus, the differentiation course of these multipotent cells can be further molded by the environmental context and can differentiate into cells suitable for the region in which they have been transplanted. Indeed, the cells derived from treated EBs showed varied neural specificities, both when injected into blastocysts and when implanted into adult spinal cords. Electrically stimulated EBs seem to incorporate into the CNS with no specific preference for anterior-posterior or dorsoventral axes, as we observed that the injected cells contributed to the entire CNS in our blastocyst injection experiments. Some induction protocols may specifically direct cells to preferentially contribute to certain tissues. For example, the protocol that uses retinoids directs cells to differentiate toward cell types located in posterior structures [34].

A large proportion of the cells (approximately 80%) derived from stimulated EBs displayed neuronal identities when injected into injured spinal cord, an environment that is non-neurogenic (Fig. 5). This amount of neural differentiation is extremely high in comparison with that observed by Ogawa et al. [35], who adopted a similar experimental design with in vitro expanded neural stem/progenitor cells derived from fetal spinal cord. In the present study, the cells we grafted differentiated into various types of neurons, as indicated by expression of several markers including neurotransmitters. Importantly, stimulated EBs showed almost no proliferative activity 50 days after transplantation, unlike ES cells or unstimulated EBs, which showed proliferative and pathogenic features (Fig. 6). Because cells derived from stimulated EBs were extremely efficient in assuming a course of neuronal differentiation in an in vivo environment, they hold much promise for use in therapeutic applications.

Ca2+ is one of the most important signaling ions involved in various biological activities [19]. Ca2+ also appears to play an important role in the system we studied. We found that Ca2+ influx was necessary for neural fate determination of EBs. Our finding indicates that this influx did not result from cell membrane fissures; rather, it appears that Ca2+ channels may be responsible for Ca2+ uptake in our system for the three following reasons: (a) passive influx via membrane fractures results in a much higher magnitude of intracellular fluorescence than what we observed in this study, (b) dye in the culture medium failed to enter the cells upon electrical stimulation, and (c) the original ES cells that did not take on a neuronal fate after electrical stimulation showed no evidence of Ca2+ influx (Fig. 3C). We attempted to identify these Ca2+ channels by culturing electrically stimulated EBs in media containing agents that block three major Ca2+ channels. Treatment with nifedepine, an L-type Ca2+ channel blocker, and ω-conotoxins, N- and P/Q-type blockers, failed to block Ca2+ influx. Although the existence of these major Ca2+ channels in EBs could not be excluded, we conclude that EBs most likely express other minor Ca2+ transporters and that these channels mediate the Ca2+ influx necessary for neuronal cell fate determination in this system. We conclude that EB formation is required for activation of these Ca2+ ion transporters, because, unlike EBs, ES cells showed no change in intracellular Ca2+ density after electrical stimulation.

Ca2+ is also known to be involved in the noncanonical Wnt signaling pathway. In Xenopus and zebrafish, some Wnt ligands stimulate release of intracellular Ca2+ during development [36]. This Ca2+ release, however, induces differentiation into elements (e.g., non-neuronal cells) fated to become ventrally located structures [37]. This is in stark contrast to our system, in which Ca2+ induces differentiation into elements (e.g., neuronal cells) fated to become dorsally located structures.

In spite of our efforts, we do not fully understand the mechanisms for this differentiation system. Although we successfully demonstrated that the Ca2+ flux is required for the induction (Fig. 3), this will not fully interpret the system. It is possible that other ions, such as potassium or sodium, can also be involved [38]. We are underway to further understand the mechanisms of neuronal induction ability of electrical stimulation on ES cells.

Signaling pathways that transmit information out of cells into the environment are usually activated by receptor-ligand recognition. However, in the present experiments, physical alteration of cell surface membranes may initiate signaling, even though the normal signaling molecule takes over later. This primitive signaling pathway may be a prototype that mediates environmental effects on cells. Receptor-ligand signaling systems may have evolved for stabilization and refinement of environmental cues impinging on cells. This simple system may possibly be invoked during early development and neuronal regeneration. In neuronal tissues, ionic currents are continuously flowing, and these currents may instruct and ensure that undifferentiated cells assume the fate of neuronal progenitors. Ionic flux, therefore, may be a novel category of differentiation signals. Indeed, recent findings show that excitatory neural activity induces adult neural stem cells to adopt a neural fate [39, 40]. In our study, we observed that cells affected by ionic currents were very close to being in an undifferentiated state. The effective differentiation step, therefore, corresponds to states that occur earlier than the neural stem cell state, because unstimulated cells could adopt various cell fates. Even though further studies are required to investigate the physiological role of ionic flow in early embryos, our observations suggest that transmembrane ion flow may play an integral role in early stages of development.

In summary, we described a simple and novel procedure for producing neural cells from ES cells. Further studies are needed to improve our procedure so that parameters are appropriate for therapeutic applications.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

We thank A. Nagy, J. Miyazaki, and R. Darnell for the gift of reagents. We also thank S. Itohara for sharing the ES cell facility, and N. Nakatani for useful discussions. Monoclonal antibodies against Islet1, Pax6, Pax7, MNR2, Nkx2.2, and Lim3 developed by T.M. Jessell were obtained from the Developmental Studies Hybridoma Bank under the auspices of NICHD and maintained by University of Iowa. T.K. thanks M. Muramatsu for continuous encouragement, I. Naganuma for secretarial work, and the late G. Matsumoto for help in establishing the laboratory. M.Y. is a recipient of a grant-in-aid for Young Scientists from The Ministry of Education, Culture, Sports, and Technology (MEXT) of Japan. T.K. is supported by a grant from the Human Frontier Scientific Program Organization and a grant-in-aid for Scientific Research on Priority Areas from MEXT of Japan. All mice were maintained at the animal facilities of RIKEN-BSI according to the Institution's guidelines.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References
  10. Supporting Information
FilenameFormatSizeDescription
Figure_1.pdf63KSupplemental Figure 1
Figure_2.pdf91KSupplemental Figure 2
Supp_Figure_Legends.pdf9KSupplemental Legends
Supp_Table_1.pdf8KSupplemental Table 1
Supplemental_Table_s2.pdf13KSupplemental Table 2

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.