|Home | About | Journals | Submit | Contact Us | Français|
Neurogenesis occurs in a stem cell niche in which vascular elements, including endothelial cells (ECs), are thought to play an important role. Using co-culture experiments, we investigated the effect of ECs on proliferation and functional neuronal differentiation of human embryonic stem (ES) cell-derived neuronal precursor cells (NPCs). NPCs were cultured for 5 days in medium containing fibroblast growth factor-2 (FGF-2), with or without ECs. FGF-2 and ECs were then removed, and NPCs were maintained in culture for additional periods. Compared to control NPC cultures, EC-treated NPC cultures showed increased cell proliferation at short intervals (5 days) after withdrawal of FGF-2 and larger tetrodotoxin-sensitive inward membrane currents at longer intervals (10–14 days), but a similar pattern of development of neuronal differentiation markers. The effects of ECs appeared to result from the release of soluble factors rather than from cell contact, because they were observed despite the physical separation of NPCs from ECs by a cell-impermeable membrane. These findings indicate that ECs can regulate the proliferation and electrophysiological neuronal differentiation of human NPCs.
Adult stem cells in two brain regions—the sub-granular zone (SGZ) of the hippocampal dentate gyrus (DG) and the most rostral extent of the forebrain subventricular zone (SVZ)—generate new neurons through the process of neurogenesis. Adult neurogenesis normally proceeds at low basal levels, which are presumed to be sufficient to replace cells lost via physiological cell death in the olfactory bulb, or to accommodate continued learning and memory in the hippocampus. However, following brain injury, such as stroke in humans, neurons proliferate at an increased rate and migrate to the site of injury, where they may replace lost cells or enhance the function of surviving cells [1,2].
The observation that new neurons arise in restricted areas of the adult brain suggests that a specific stem cell microenvironment, or niche, may be required for neurogenesis [3–5]. Further evidence includes the diminished potential of adult neuronal precursor cells (NPCs) engrafted into nonneurogenic brain regions , and the fact that cells from normally nonneurogenic regions can generate new neurons when implanted into neurogenic zones . The environmental factors that account for these observations are referred to collectively as the stem cell niche and may include cells, extracellular matrix, and other tissue constituents . In the adult brain, blood vessels appear to be important components of the neural stem cell niche, because neurogenesis occurs in a vascular environment in vivo  and certain neurogenesis-promoting signals may be mediated through endothelial cells (ECs) . ECs may also have a role in the homing of newborn neurons to sites of injury, as these neurons are found in close proximity to blood vessels in the ischemic cerebral cortex [1,11].
In a previous study, Shen and colleagues  showed that co-culturing neural stem cells from mouse embryonic cortex together with mouse brain or bovine pulmonary artery ECs promoted neural proliferation, inhibited neuronal differentiation, and, upon removal of the ECs, led to increased production of βIII-tubulin-, microtubule-associated protein-2 (MAP2-), and NeuN-expressing neurons. These effects appear to involve the release of soluble factors from cultured ECs. We have investigated the interaction between cultured ECs and NPCs to determine: (1) if ECs also regulate the proliferation of human NPCs, and (2) whether this is associated with changes in the development of electrophysiological neuronal properties.
NPCs derived from human embryonic (hES) stem cell line H9 were obtained from Aruna Biomedical Inc. (Athens, GA). Cells were seeded on polyornithine- and laminin-coated dishes and cultured in proliferation medium, consisting of Neurobasal medium with B27 supplementation  containing 2 mM l-glutamine and 50 μg/ml penicillin/streptomycin (all from Invitrogen), and 10 ng/ml leukemia inhibitory factor (LIF) plus 20 ng/ml fibroblast growth factor-2 (FGF-2) (both from R&D Systems) . Cells were propagated further in proliferation medium and, upon reaching 90–100% confluence, were triturated to detach them from the dish. After centrifugation at 200 × g for 4 min, cells were resuspended in fresh medium and replated. To induce differentiation, cells were washed with phosphate-buffered saline (PBS) containing Ca2+ and Mg2+, and cultured in differentiation medium (proliferation medium without FGF-2), which was replaced every 3 days.
For co-culture experiments, mouse brain ECs (ATCC #CRL-2299) were used at passages 14–16. Three days before co-culture, ECs were plated in Dulbecco modified Eagle medium (DMEM) with 10% fetal bovine serum (FBS) onto 6.5-mm Transwell membrane inserts (Costar) at 200–300 cells per Transwell. Four hours before use, Transwells were well rinsed and transferred to serum-free Neurobasal medium containing 20 ng/ml FGF-2 and 10 ng/ml LIF. The Transwells were placed above freshly plated NPCs, and cultures were fed every other day with serum-free medium. After co-culture for 4–5 days, Transwells and FGF-2 were removed, and cells were cultured for up to 14 additional days.
Immunocytochemistry was performed as described previously [15,16]. Primary antibodies were rabbit polyclonal anti-Sox2 (1:200; Abcam, Cambridge, UK); rabbit polyclonal anti-γ-aminobutyric acid (GABA, 1:400), mouse monoclonal anti-OX4 (1:400), rabbit polyclonal anti-TUC-4 (1:250), guinea pig polyclonal anti-doublecortin (DCX) (1:500), rabbit polyclonal anti-synapsin I (1:500), rabbit polyclonal anti-N-methyl-D-aspartate receptor subunit NMDAR2A (1:200), and mouse monoclonal anti-nestin (1:200) (Chemicon International, Temecula, CA); mouse monoclonal anti–MAP2 (1:100) and rabbit polyclonal anti-glial fibrillary acidic protein (GFAP, 1:400) (Sigma-Aldrich, St. Louis, MO); rabbit polyclonal anti-calbindin (1:400; Upstate); rabbit polyclonal anti-βIII-tubulin (TuJ1 antibody, 1:1000; Covence, Berkeley, CA); and mouse monoclonal anti-Hu (1:200; Invitrogen, Carlsbad, CA). Secondary antibodies were fluorescein isothiocyanate (FITC)-conjugated goat anti-mouse IgG, FITC-conjugated rabbit anti-rat IgG, and FITC-conjugated rabbit anti-rabbit IgG (1:200; Vector Laboratories, Burlingame, CA), and FITC-conjugated pig anti-goat IgG, rhodamine-conjugated rat-absorbed donkey anti-mouse IgG, rhodamine-conjugated rat-absorbed donkey anti-rabbit IgG, and rhodamine-conjugated rat-absorbed donkey anti-sheep IgG (1:200; Jackson ImmunoResearch Laboratories Inc., West Grove, PA). 4′,6-Diamidine-2-phenylindole dihydrochloride (DAPI; Vector Laboratories) was used to counterstain nuclei, and fluorescence signals were detected with a Nikon E800 microscope at excitation/emission wavelengths of 535/565 nm (rhodamine, red), 470/505 (FITC, green) and 360/400 (DAPI, blue). Results were recorded with a Magnifire digital camera (ChipCoolers, Warwick, RI). Controls included omitting or preabsorbing the primary or omitting the secondary antibody.
Current and voltage-clamp recordings in cultured cells were conducted using the whole-cell patch-clamp technique [17,18]. Patch electrodes were constructed from thin-walled, 1.5-mm outside diameter borosilicate glass on a Flaming/Brown P-97 pipette puller (Sutter Instruments, Novato, CA), to a final inside diameter of 1–2 μm. The patch electrodes had resistances of 4–5 MΩ when filled with intracellular solution containing 140 mM K-gluconate, 0.1 mM CaCl2, 2 mM MgCl2, 1 mM EGTA, 2 mM K2ATP, 0.1 mM Na3GTP, and 10 mM HEPES (pH adjusted to 7.25 with KOH). An Axopatch 200B amplifier was used to control pipette potentials or inject current, a Digidata 1322A interface to convert A/D values at 5–50 kHz (filtered at 2 kHz), and pClamp 9.2 software to control both the amplifier and the converter (Axon Instruments, Foster City, CA).
Data were filtered at 2 kHz and digitized on-line using Digidata 1320A DAC units. The online acquisition was analyzed with pClamp software (version 9.2, Axon Instruments). Cells were cultured on 12-mm glass coverslips. Recording coverslips were placed in a chamber on a Nikon Eclipse E600FN microscope and perfused at 1.0–1.8 ml/min. The external (rat Ringer's) bath solution consisted of: 140 mM NaCl, 3 mM KCl, 1 mM MgCl2, 2 mM CaCl2, 10 mM HEPES, and 10 mM glucose, pH 7.4. Cells with small somata and two to three dendrite-like processes were selected to record. Action potentials were elicited by applying depolarizing current pulses (100 ms) ranging from −0.06 to 0.16 nA. Membrane currents were record with 10-mV voltage steps from −120 to +170 mV (holding potential, −70 mV). Mean current–voltage relationships were calculated by averaging the individual current–voltage relationship obtained from each cell in the study. All recordings were performed at room temperature.
hES cell-derived NPCs maintained in FGF-2-containing proliferation medium (Fig. 1A) expressed both nestin and Sox-2 (Fig. 1B). Cell counting in immunostained cultures showed that >90% of cells were positive for each marker (Fig. 1C). In contrast, only very few cells expressed lineage-specific neuronal markers such as Hu and TUC-4. Markers of mature cell types, including the neuronal marker calbindin, the astroglial marker GFAP, and the oligodendroglial marker OX4, were absent. Whole-cell patch-clamp recordings showed that cells in these cultures exhibited an outward current (Fig. 1D) that activated rapidly at about −10 mV, reached a peak at about + 60 mV, and then decreased (Fig. 1E). Neither inward currents nor action potentials were found in these cells, consistent with only primitive stages of neuronal differentiation.
To investigate the interaction between ECs and NPCs, NPCs were plated in culture wells with ECs seeded above them on Transwell inserts, and maintained for 5 days in proliferation medium (Fig. 1F). Proliferation of NPCs was measured by bromodeoxyuridine (BrdU) incorporation and cell counting. Co-culture with ECs increased the number of BrdU-labeled cells by ~100% and of total cells by ~70% (Fig. 2A–C), suggesting that ECs stimulated NPC proliferation. Because the 0.4-μm-diameter pores of the Transwell membranes used preclude cell migration between compartments, the proliferative effect of ECs was likely mediated through secreted soluble factors. Whole-cell patch clamp studies on NPCs after 5 days in culture with, but not without, ECs demonstrated a small inward current that activated at about −40 mV, reached a peak (~ 100 pA) at about −10 mV, and had a reversal potential near 0 mV (Fig. 2D–F). Immunocytochemistry showed that NPCs cultured either with or without ECs almost all expressed nestin and Sox-2, whereas markers of more differentiated neuronal phenotype were barely detectable or absent (Fig. 2G).
Next we asked whether prior exposure to ECs could also enhance NPC differentiation at longer intervals. EC-containing Transwell inserts were removed, proliferation medium was replaced with differentiation medium lacking FGF-2, and NPCs were cultured for 10 more days. Compared to NPCs cultured in proliferation medium, in either the absence (Fig. 1A–C) or presence (Fig. 2A) of ECs, cultures maintained in differentiation medium showed increased numbers of cells with morphological features of neurons (cellular polarity, neurite extension), which also expressed mature neuronal markers (Fig. 3A). Whole-cell patch-clamp studies showed only outward current in NPCs cultured in differentiation medium without ECs, but also inward current, which activated quickly at about −50 mV, reached a peak at about −40 mV, and reversed near 0 mV, in the presence of ECs (Fig. 3B,C). This inward current was increased in peak amplitude (~200 pA) compared to that measured in cultures exposed to ECs in proliferation medium (Fig. 2F). Despite their differing electrophysiological properties, NPCs cultured in differentiation medium with or without ECs showed similar patterns of neuron-lineage marker expression (Fig. 3D). These included proteins associated with differentiation toward a neuronal phenotype, such as βIII-tubulin and doublecortin (DCX), which were not expressed by NPCs in proliferation medium (Fig. 2G).
After 14 days in differentiation medium, NPCs cultured without ECs began to show a small inward current (Fig. 4A,B). However, NPCs co-cultured with ECs had much larger inward currents (peak amplitude ~400 pA), which activated at about −60 mV, reached a peak at about −30 mV, showed a reversal potential near 0 mV, and were blocked by 500 nM tetrodotoxin (TTX). In current-clamp mode, action potentials were demonstrable in only 3/8 cells cultured without ECs, but in 8/8 cells maintained in the presence of ECs. Expression of neuronal markers was similar after 14 days (Fig. 4C) compared to 10 days (Fig. 3D) in differentiation medium. In addition, at 14 days, NPCs cultured in differentiation medium expressed neuronal markers to similar extents in the absence or presence of ECs (Fig. 4C).
The main finding of this study is that ECs modify the proliferation of hES cell-derived, human NPCs and their development of electrophysiological neuronal properties. When Sox-2- and nestin-expressing NPCs were cultured in the presence of FGF-2, they proliferated and showed only outward membrane currents. Removal of FGF-2 led to biochemical differentiation, manifested by the expression of βIII-tubulin, DCX, Hu, calbindin, and GABA by 10 days, and electrophysiological differentiation, evidenced by the acquisition of a small inward current with a reversal potential close to the equilibrium potential for Na+, by 14 days. NPCs co-cultured with ECs showed a similar time course of neuronal marker protein expression, but accelerated development of inward current, which was detectable 5 days after removal of ECs and withdrawal of FGF-2, and which was larger in magnitude at 14 days than the inward current in NPCs cultured without ECs. In addition, action potentials were seen more frequently at 14 days in NPC-EC co-cultures. We conclude that ECs enhance cell proliferation and hasten the development of electrical but not biochemical neuronal properties in NPC cultures, and that the effect of ECs is likely to depend on diffusible mediators rather than cell–cell contact, because NPCs and ECs in co-culture were separated by a barrier that precluded transmigration.
Several prior studies have documented the transition of NPCs to more functionally mature cells of neuronal lineage in vitro. In rodents, cultured ES cells [19,20] and NPCs from embryonic [21,22], fetal , or adult [24,25] brain develop inward and outward membrane currents and action potentials consistent with the expression of voltage-gated Na+ and K+ channels. Similar observations have been made for NPCs cultured from embryonic  or adult [27–29] human brain. The electrophysiological properties of undifferentiated hES cells and hES cell-derived dopaminergic neurons maintained in culture have also been described. Undifferentiated hES cells showed depolarization-activated, tetraethylammonium-sensitive delayed rectifier K+ currents, but neither Na+ nor Ca2+ currents . In contrast, hES cells induced to undergo dopaminergic differentiation by co-culture with stromal cells that overexpress sonic hedgehog developed fast inward currents and delayed outward currents with slight inactivation, rapidly inactivating A-type currents and action potentials .
As noted previously, Shen et al.  reported that the in vitro proliferation of murine NPCs could be enhanced by co-culture with ECs, providing a possible mechanistic basis for the neurovascular stem-cell niche . When Transwell inserts containing primary bovine pulmonary or clonal mouse brain ECs were added to mouse embryonic cortical cell cultures, the latter expanded more rapidly, but the percentage of cells expressing the neuronal marker βIII-tubulin was lower. Removal of the inserts was associated with an increase in the percentage of cortical cells that went on to express the neuronal marker βIII-tubulin, compared to pure cortical cultures, or to cortical cells co-cultured with fibroblasts or vascular smooth muscle cells, without change in the percentage of GFAP+ astrocytes. The authors concluded that endothelium-derived factors stimulate the proliferation of NPCs and inhibit neuronal differentiation. Our findings are substantially in agreement, in that co-culture of ECs with hES cell-derived NPCs led to an increase in cell proliferation but not in biochemical neuronal differentiation, as determined by marker protein expression. One difference between the two studies is that we withdrew FGF-2 from NPC cultures when EC-containing Transwell inserts were removed, whereas Shen et al.  did not. Therefore, the differences that we observed between control and EC-treated NPC cultures could depend on concurrent FGF-2 withdrawal. Another difference is that we studied not only protein marker expression, but also the development of electrical neuronal properties. These studies revealed a possible role of ECs in promoting the electrical, as opposed to biochemical, neuronal differentiation of NPCs. Alternatively, the sole effect of ECs could be to increase NPC proliferation, and the more prominent TTX-sensitive currents in EC-treated cultures could simply reflect the presence of more neurons.
The induction of electrical neuronal properties in hES cell-derived NPCs by other cell types has also been described. Co-culture with astrocytes from neonatal or adult hippocampus, or culture on astrocyte-conditioned substrates, promoted the development of TTX-sensitive action potentials in adult rat hippocampal NPCs . In another study, Johnson et al.  followed the development of neuronal electrical properties in cultured hES cells induced to undergo neuronal differentiation in the absence of FGF-2 and the presence of brain- and glial-derived neurotrophic factors. Neuronal differentiation was accompanied by the emergence and subsequent maturation of TTX-sensitive Na+ action potentials, which showed progressively increased amplitude and decreased duration. Co-culture of hES cell-derived NPCs with astrocytes had little effect on cell-intrinsic neuronal properties, but accelerated the development of synaptic activity, manifested by spontaneous, α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) and GABA receptor-mediated postsynaptic currents. In contrast to our results and those of Shen at al.  regarding ECs, the effect of astrocytes appeared to depend on cell–cell contact, because it was abolished if NPCs and astrocytes were plated on opposite sides of a Transwell insert.
The stem-cell niche in which neurogenesis occurs comprises a variety of cell types and extracellular components, which contribute to the production of neurons through diverse mechanisms. In addition to NPCs, which are the direct predecessors of new neurons, these include cells that regulate neuronal differentiation. This study and previous work cited above [12,32] suggest important roles for ECs and astrocytes in this latter process. In particular, soluble factors released from ECs contribute to the expression of ion channels underlying neuroexcitability, whereas contact with astrocytes helps guide the subsequent development of transsynaptic neuronal activity. The roles of other constituents of the stem-cell niche, including vascular smooth muscle cells, pericytes, circulating blood cells, and extracellular matrix remain to be clarified, especially in the context of human neurogenesis.
This work was supported by National Institutes of Health (NIH) grants NS44921 (D.A.G.) and AG21980 (K.J.) and by the Buck Institute for Age Research.